Washington Academy of Sciences

Founded in 1898

EXECUTIVE COMMITTEE

President

John G. Honig

President-Elect

James F. Goff

Secretary

Jean K. Boek

Treasurer

John J. O’Hare

Members at Large

Conrad B. Link

Elaine Shafrin

John J. O’Hare

Michael J. Pelezar, Jr.

Jo-Anne Jackson

Grover C. Sherlin

Lavern S. Birks

BOARD OF MANAGERS

All delegates of affiliated

Societies (see Cover 3)

EDITORS

Irving Gray

Joseph Neale

ACADEMY OFFICE

1101 N. Highland St.

Arlington, Va. 22201

Telephone: (703) 527-4802

The Journal

This journal, the official organ of the Washington

Academy of Sciences, publishes historical articles, critical

reviews, and scholarly scientific articles; proceedings of

meetings of the Academy and its Board of Managers; and

other items of interest to Academy members. The Journal

appears four times a year (March, June, September, and

December)—the September issue contains a directory of

the Academy membership.

Subscription Rates

Members, fellows, and patrons in good standing receive

the Journal without charge. Subscriptions are available on

a calendar year basis only, payable in advance. Payment

must be made in U.S. currency at the following rates:

U.S.and Canada..sc5 5. $19.00

POFGIERY < 35 35 ocr 22.00

Single Copy Price >... = 7.50

UHURU Symposium—2 issues...........- $15.00

Delivery by Air... 2... 2.00

Single-copy price for Vol. 66, No. 1 (March, 1976) is $7.50.

Back Issues

Obtainable from the Academy office (address at bottom

of opposite column): Proceedings: Vols. 1-13 (1898-1910)

Index: To Vols. 1-13 of the Proceedings and Vols. 1-40 of

the Journal Journal: Back issues, volumes, and sets (Vols.

1-62, 1911-1972) and all current issues.

Claims for Missing Numbers

Claims will not be allowed if received more than 60 days

after date of mailing plus time normally required for

postal delivery and claim. No claims will be allowed

because of failure to notify the Academy of a change in

address.

Change of Address

Address changes should be sent promptly to the Academy

office. Such notification should show both old and new

addresses and zip number.

Published quarterly in March, June, September, and December of each year by the Washington

Academy of Sciences, 1101 N. Highland St., Arlington, Va. 22201. Second class postage paid at

Arlington, Va. and additional mailing offices.

INTRODUCTION

DAVID NISHIOKA

Organizer

Symposium: Cellular and Molecular Aspects of Fertilization

and Early Embryonic Development

The title of this symposium reflects a cooperation between many different disciplines.

As our understanding of cells and molecules has advanced, so has our understanding of

fertilization and the early embryo, and vice versa. This cooperation is a recurring theme

throughout this symposium. I open by reviewing work directed at determining the roles of

extracellular ions in the activations of the sperm and the egg. Besides providing insight

into these central questions, the work has provided important experimental tools which

may be used to dissect these processes into their component parts. Dr. Schatten follows

with a description and analysis of the many cellular movements involved in fertilization,

from the movement of the spermatozoon prior to fertilization to the movements of the

sperm and egg nuclei towards one another within the egg cytoplasm. Work related to the

question of how the genetically repressed sperm nucleus is transformed to the active state

within the egg is then discussed by Dr. Poccia. Finally, Dr. Belisle closes with a review and

experimental analysis of the cell division cycle in the early embryo. Cellular activation,

cellular movements, genetic derepression, and the cell cycle are some of the most actively

pursued areas of research in modern cell biology. In this symposium they lead to a closer

understanding of fertilization and early embryonic development.

AAGINEIM qiVAd

vesineg2O

meanest ha Bees oe “iy layto eet ii baw vast ah} o. ‘ouhaugory

igntaelavads ° ince patie Y chee Gane

' i 1 ODA GOON 6 AtoMl ey AME Cie a ite

a cié oe ; SIS ow alte " vt TAH Geo

> abt? .savey Se ae pa ‘Tenants bam

sy "e . ey ; seer ae Oe

rey ee) Oe a ee a)

, aw’, ah) .2cov teat ‘na

Journal of the Washington Academy of Sciences,

Volume 72. Number 1. Pages I-11, March 1982

The Ionic Basis of Fertilization

David Nishioka

Department of Biology, Georgetown University,

Washington, D.C. 20057 USA

ABSTRACT

Fertilization is considered in terms of two cellular activations. Experiments with sea urchins

gametes in which the ionic composition of the external medium is controlled, show that extra-

cellular sodium ions (Na’) are required for the activations of both the sperm and the egg.

These requirements are for the release of intracellular protons (H’) and result in increased

intracellular pH (pHi). Evidence is presented showing that these changes occur spontaneously

upon suspension of sperm in sea water and result in the initiation of sperm motility. In the egg,

these changes occur after fusion with sperm and result in the initiation of embryonic

development.

Introduction

How cells go from the metabolically re-

pressed to the metabolically active states

and how they go from the nondividing to

the dividing states have long been recog-

nized as central questions in our under-

standing of cells. They are crucial to our

understanding of multicellular organisms

as well. The immune response, wound heal-

ing, carcinogenesis, all involve these cellu-

lar transformations. In fact, as this review

and the three which follow point out, the

multicellular state in many organisms Is it-

self predicated on these types of cellular

transformations. The questions are clear in

the study of fertilization. How does the dif-

ferentiated sperm cell which is stored in a

metabolically repressed state undergo acti-

vation to an actively respiring, fully motile

State prior to fertilization; and how does

the egg, upon fusion with sperm, undergo

transformation from the nondividing to

the dividing states leading to the develop-

ment of a new individual?

An implicit concept used in answering

these questions is that these cells are closely

linked with their environments in predicta-

ble and specific ways. It would be difficult

to explain the activation of sperm in any

other terms. The sperm cell has a limited

amount of energy. If it does not fuse with

an egg in a limited amount of time, it

simply expends this energy and dies. Adult

male organisms must provide an environ-

ment which preserves this energy until

sperm are released. The composition of the

external medium then becomes the impor-

tant limiting factor in whether or not the

sperm is activated. The activation of the

egg is less direct. An unfertilized egg may

remain suppressed for extended periods of

time in a medium fully capable of support-

ing activation. Fusion with sperm is neces-

sary for the activation to proceed, but as

for sperm, the composition of the external

2 DAVID NISHIOKA

‘medium then becomes the limiting factor.

The central questions become more clear.

How are these cells linked with their envi-

ronments to bring about cellular activation?

In this review studies will be presented

suggesting that this linkage is provided by

specific ions which cross the plasma mem-

branes of these cells. These ionic flows re-

sult in changes in the ionic composition of

the intracellular environment where they

are translated into changes in cell metabo-

lism. For reasons that are obvious, the or-

ganism of choice in making these determi-

nations has been the sea urchin. A single

female, upon the gentle request of injecting

isosmotic KCL into the body cavity, may

shed as much as twenty mls of eggs, each ml

containing two million eggs. A single male

may shed as much as five mls of undiluted

semen, each ml containing 10-100 billion

sperm. Jn vitro fertilization is accomplished

simply by mixing sperm and eggs together

in sea water. A most important advantage

of this system in determining the involve-

ment of extracellular ions in fertilization is

that the ionic composition of the external

medium, sea water, is known and may be

totally defined. Fertilized eggs will reach

first cell division approximately 90 minutes

after insemination. The next four to five

divisions occur every 60 minutes and re-

main synchronous within each embryo and

within a given culture. Development pro-

ceeds synchronously through easily recog-

nizable embryonic stages (blastula, gas-

trula) to an early larval stage (pluteus)

within three days.

Activation of Sperm Metabolism and

Motility

Several early studies were directed at the

questions of how sea urchin sperm are re-

pressed in undiluted semen and activated

upon suspension in sea water. Gray’ first

described the activation as a ‘‘dilution ef-

fect,” explaining the repression in undi-

luted semen by simple mechanical crowd-

ing of the spermatozoa and the activation

as a release from this crowding. In support

of this idea, he reported that respiration

and motility are stimulated when semen is

further diluted with seminal fluid (later

confirmed by Hayashi).’ Forcing a reeval-

uation of this idea, Rothschild” reported

that when undiluted semen is experimen-

tally oxygenated the spermatozoa become

motile, and when it is deoxygenated motil-

ity is again inhibited. These results suggest

that the lack of oxygen (or CO: anoxia) is

the repressive factor in undiluted semen

and that oxygenation is at least one of the

factors involved in the activation of sperm

upon suspension in sea water. CO anoxia

can be expected if one views undiluted

semen as an extremely dense nonvascular-

ized suspension of living cells. The results

of Gray and Hayashi can be explained if,

during preparation, the seminal fluid be-

came oxygenated, which could occur in the

absence of sperm.

There is no doubt that oxygenation is

one of the factors required for the activa-

tion. Spermatozoa are not anaerobic. How-

ever, it cannot be the only factor involved.

If this were the case, suspension of undi-

luted semen in any oxygenated, isosmotic

medium should support activation. The re-

sults presented in Table | indicate that this

is not the case. When undiluted semen is

suspended in isosmotic solutions of the

major salts of sea water, NaCl, KCI, CaClo,

and MgSO,, sperm motility is inhibited in

all solutions except NaCl. Additionally,

sperm motility is inhibited in sodium-free,

choline-substituted sea water (ONa -SW)

in which all of the major ions of sea water

except Na’ are present in normal amounts.

When Na’ is added back, motility is stimu-

lated. These results suggest that external

sodium, the most plentiful cation in sea

water, is specifically involved in the activa-

tion of sperm metabolism and motility.

It is now known that sperm release acid

upon suspension in sea water* ° and that

this release is dependent upon extracellular

THE IONIC BASIS OF FERTILIZATION 3

Table 1—Sperm Motility in Isosmotic Salt Solutions®

Solution? Motility‘

ASW -

NaCl (0.51 M) -

KCL (0.51 M) -

CaCl, (0.34 M) —

MgCl) (0.34 M) —

MgSO, (0.51 M) =

ONa’*-SW -

ONa*-SW + 0.01 M NaCl -

“5.0 ul of undiluted semen added to 10.0 ml of solu-

tion with constant stirring.

’ASW (artificial sea water) is composed of 0.423 M

NaCl, 0.009 M KCl, 0.00927 M CaCl2, 0.02294 M

MgCl, 0.0255 M MgSOu, 0.00215 NaHCOs, pH 8.0.

Osmolarity = 1.02 OsM.

All salt solutions buffered at pH 8.0 with 0.01 M tris.

ONa’™-SW (Sodium-free sea water) is prepared ac-

cording to the same formula for ASW substituting

choline chloride for NaCl and KHCO; for NaHCO3.

“Scored as motile (+) or immotile (—) 30 seconds

after suspension.

7.5

7.4

sodium.°® As shown in Figure la, it can be

observed as a drop in pH of the suspension,

and is composed of two kinetically differ-

ent components, one fast which imme-

diately lowers the pH of the suspension

from 8.0 to 7.6, and one slow which further

lowers the pH to 7.4 over the three minute

time period monitored. The respiratory in-

hibitor cyanide completely inhibits the slow

component, while the fast component is

unaffected (Figure 1b). These results con-

firm that the release is composed of two

components, 1.e., they are not only kineti-

cally separable but also experimentally

separable; and they indicate that the slow

component represents increased respira-

tion and production of metabolic acid

(CO2). The latter indication is supported by

the fact that motility is inhibited in the

presence of cyanide. Neither component is

observed in ONa’-SW in which motility is

1 2 3 0 1 2 3

Minutes

Fig. 1. Acid release from sea urchin sperm. [a] 250 ul undiluted semen added to 2.0 ml ASW at 0 time

(arrow). [b] 250 ul undiluted semen added to 2.0 ml ASW containing 10° M potassium cyanide at 0 time

(arrow). [c] 250 ul undiluted semen washed and resuspended in 2.0 ml ONa’-SW;; at 0 time (arrow) NaCl, q.s. 10

mM, was added. The pH of these suspensions (extracellular pH) was measured continuously with a Corning

Model 12 expanded scale pH meter equipped with a Fisher Microprobe combination electrode and a Fisher

Series 5000 chart recorder.

4 DAVID NISHIOKA

inhibited and the pH of the suspension re-

mains constant at 8.0. When sodium is

added back, the normal two-component re-

lease proceeds (Figure Ic), and motility is

stimulated.

The picture which has emerged from

these facts is that the fast component rep-

resents a rapid exchange of extracellular

Na’ into the cell for intracellular protons

(H*) out, and the slow component repre-

sents increased metabolism. Viewed in this

way, the separability of the two compo-

nents becomes more important. Since the

fast component precedes the slow compo-

nent, it could control metabolism and mo-

tility. This idea finds support in the fact

that both components of acid release from

sperm suspended in ONa’-SW are propor-

tional to the amount of Na’ added back

(Figure 2). The results presented in Table 2

confirm that the rate of respiration is also

directly proportional. It is as if metabolic

activation may be titrated with extracellu-

lar Na’. In the presence of cyanide, Na*-H™

8.0

7.9

7.8

pH 7.7

7.6

7.5

Table 2—O2-Consumption by Sea Urchin Sperm in

ONa’-SW*

O2-consumption

mM Na’* added nM 02/2 X 10° sperm/min

0.0 1.062

0.5 2.165

1.0 4.330

1°5 25.980

2-5 54.125

5.0 64.950

10.0 73.610

“Measured with a Yellow Springs Model 5331 oxy-

gen electrode.

exchange occurs but metabolism is inhib-

ited at a later link ina chain of events, in the

case of cyanide, at the level of the cyto-

chrome system and electron transport.

An obvious dilemma that the Na -H’ ex-

change hypothesis raises is whether the met-

abolic switch is increased intracellular Na"

or decreased intracellular H’. It is now be-

coming clear that the latter alternative is

1 2 3 0 1

Minutes

Fig. 2. Acid release from sea urchin sperm. 250 yl undiluted semen was washed and resuspended in 2.0 ml

ONa’-SW. At 0 time (arrows) varying amounts of NaCl, q.s. 0.25 mM [a], 0.5 mM [b], and 1.0 mM [c], were

added. Measurements were made as described for Fig. 1.

THE IONIC BASIS OF FERTILIZATION 5

the important intracellular change. Several

methods for measuring intracellular pH

(pHi) in sperm have recently been devel-

oped’” and all show increases in pHi (de-

creased H”) which correlate directly with

increases in metabolism and motility. These

studies agree that upon addition of Na’ to

sperm suspended in ONa’-SW the pH in-

creases 0.4-0.6 pH units, although there is

some disagreement about the absolute pH;

values. Lee et a/. and Lee and Epel (unpub-

lished results) calculate pHi’s of 6.5 in rest-

ing sperm and 6.9-7.1 in activated sperm,

while Christen ef al.’ calculate respective

pHi’s of 7.0 and 7.4. Regardless of the abso-

lute values, there is good experimental evi-

dence that the increase stimulates metabo-

lism and motility. For example, Table 3

shows that when the pH of asperm suspen-

sion in ONa -SW is raised to pH 9.0 with

KOH, sperm motility is stimulated. The

OH moiety of KOH is the effective varia-

ble, since additions of KCI equal to the

amount of KOH used to raise the pH have

no effect, i.e., the sperm remain immotile.

Additionally, K* is already present in

ONa’-SW in normal amounts (ca., 0.009

M). When the pH of this same suspension is

returned to 8.0 with HCl, motility is again

inhibited. In this case the H” moiety of HC!

is the effective variable, since Cl is already

present in ONa’-SW in normal and much

larger amounts (ca., 0.496 M) than the

amount of HCl used to lower the pH. Thus,

sperm motility may be controlled by vary-

Table 3—Sperm Motility in ONa’-SW°

Treatment Motility‘

Adjusted to pH 9.0 with KOH a

Readjusted to pH 8.0 with HCl —

+10.0 mM NH.Cl oS

—NH.CI’ =

“5.0 ul indiluted semen added to 10.0 ml ONa*-SW

with constant stirring.

’Motile sperm in ONa*-SW + 10.0 mM NH,Cl

centrifuged and resuspended in ONa’-SW-NH,Cl.

‘Scored as motile (+) or immotile (—) 30 seconds

after suspension.

ing the extracellular pH in ONa’-SW.

These results are interpreted as a bypass of

normal Na’-H" exchange and a direct con-

trol over pHi. No Na’ is present and pH is

the only effective variable in the system.

This type of control is apparently not pos-

sible in normal Na’-containing sea water in

which Na’-H” exchange may occur and

sperm motility may be observed through-

out the pH range 6.5-9.5. It is interesting to

note that the lower limit of this pH range ap-

proximates the calculated pHi’s of resting

sperm. For the pHi to be raised to activat-

ing levels at this extracellular pH, protons

would have to be transported against a

concentration gradient.

Another means of bypassing Na’-H" ex-

change can be achieved by adding NH,Cl

(final concentration, 0.01 M) to sperm sus-

pended in ONa’-SW. Through a well-stud-

ied transmembrane equilibrium, extracel-

lular.NH,’ raises intracellular pH by equili-

brating with NH; outside the cell, entering

the cell as NH, and then re-equilibrating

with NH,’ inside the cell by taking up in-

tracellular protons.

Ht Hq”

4 NHz NH < a 4

C +

H H

extracellular plasma intracellular

space membrane space

As shown in Table 3, when NH,’ is added

to sperm suspended in ONa’-SW, motility

is stimulated. Removal of extracellular

NH,’ should shift the transmembrane equil-

ibrium in the reverse direction and lower

intracellular pH. When NH, ’-activated

sperm are centrifuged and resuspended in

ONa’-SW without NH,’ present, motility

is again inhibited. The results in Table 4

confirm that the stimulations of sperm mo-

tility in ONa’-SW induced by raising the

pH to 9.0 with KOH and by adding 10.0

mM NH,Cl are coupled with increased res-

piratory rates.

6 DAVID NISHIOKA

Table 4—O,-Consumption by Sea Urchin Sperm*

O2-consumption

(nanomoles 02/2 X 10°

Medium sperm/min)

ASW 88.33

ONa’-SW 1.06

ONa’™-SW + 10.0 mM NaCl 73.61

ONa’-SW, pH 9.0 56.29

ONa’™-SW + 10.0 mM NH.Cl 74.61

“Measured with a Yellow Springs Model 5331 oxy-

gen electrode.

All of the studies reviewed above are

consistent with the idea that intracellular

pH controls metabolism and motility. Con-

sider first the repressed sperm in undiluted

semen. If the seminal fluid were either low

in Na* (below 0.25 mM) or low in pH (be-

low 6.5) repression would be expected.

These possibilities have recently been elim-

inated by direct measurements which show

that the ionic composition of seminal fluid

is surprisingly similar to sea water (0.44 M

Na’, pH 7.4).’° To provide an alternative

explanation, it is necessary to go back to the

early studies with undiluted semen which

suggested that the sperm are repressed by

CO, anoxia. How CO? could generate pro-

CO, COp + HO

]

H5CO3

]

Nat H*+ HCO37

intracellular

space

extracellular

space

plasma

membrane

Fig. 3. The proposed scheme for the generation of

intracellular protons by extracellular CO, and the re-

lease of intracellular protons by Na’-H’ exchange. See

explanation in text.

tons and maintain a low pH; is shown in

Figure 3. CO, diffuses freely across the

plasma membrane. Once inside a cell, and

through catalysis by the enzyme carbonic

anhydrase, it could combine with water to

form carbonic acid which would imme-

diately dissociate to free protons and bi-

carbonate ions in the measured pHi range

(6.5-6.8). Replacement of CO2 with Qo, a

treatment which stimulates motility in un-

diluted semen,’ could raise the pHi in two

ways. It could simply reverse all of the equi-

libria or it could allow Na’-H” exchange to

proceed.

Now consider the sperm upon suspen-

sion in sea water. Extracellular CO; levels

are reduced, Na’ is present, O2 is provided

for respiration, and metabolism and motil-

ity are stimulated. Again, the pH; could be

raised through a simple reversal of the equi-

libria. The results presented here, however,

show that extracellular Na’ is required to

remove the protons, through Na’-H™ ex-

change. A recent and interesting finding is

that even after sperm have undergone acti-

vation in sea water and are actively swim-

ming, they may be rearrested by passing

CO> gas over the suspension.’ It is as if the

conditions in undiluted semen have been

experimentally reintroduced. Upon replac-

ing the CO with air, and since norma! levels

of extracellular Na” are present, metabo-

lism and motility are stimulated.

Activation of the Egg

Based on the number of things that the

egg must do after fertilization, egg activa-

tion is necessarily more complex. Sperm

are activated solely to fertilize the egg. The

egg is activated to produce a new organism.

This is not to say that the cause of activa-

tion is necessarily more complex; only that

its effects are. The mature unfertilized sea

urchin egg is metabolically quiescent and

repressed, having undergone both meiotic

reduction divisions and a period of in-

THE IONIC BASIS OF FERTILIZATION

tense synthetic activity and storage during

oogenesis. Fusion with sperm activates or

derepresses the unfertilized egg from this

relatively inactive state, initiating a series

of events which eventually results in cell

division and development of the embryo. A

chronological list of these events is shown

in Figure 4. Occurring within the first min-

ute of sperm-egg fusion are those events

termed ‘‘early” which include a Na’-de-

pendent depolarization of the egg plasma

membrane,’ the cortical reaction,” and

increased respiration.'* Beginning five min-

utes after sperm-egg fusion are the “‘late”’

events which include development of K*-

conductance,'”'* increased permeability to

phosphate,'° nucleosides, ° and amino

acids,’’ increased protein synthesis,’® and

initiation of DNA synthesis.’”

Knowing the times at which these events

are stimulated has provided great insight

into the activation process. For example, it

was first proposed by Mazia’’ and since

confirmed by numerous investigators’ ’ ~*

that treatment of unfertilized eggs with mil-

limolar concentrations of NH," bypasses

the early events and selectively turns on the

late events. An interesting and important

aspect of this activation is that these eggs

may be fertilized at any time after treat-

ment and by this definition, remain unfer-

tilized. Subsequent fertilization of these

eggs stimulates the early events. It is as

though the total activation of the egg is

under control of two regulatory master

switches, each activating a group of events,

and the order in which the switches are

thrown may be reversed.

It has become increasingly clear that the

switch for the late events is the same switch

implicated in the activation of sperm, in-

creased pHi. The fact that NH,’ is a stimu-

lator is itself evidence for this idea since it

could experimentally raise pHi through the

Same transmembrane equilibrium. Direct

measurements have now confirmed this ef-

fect.*> ** The means by which this switch is

thrown is apparently also similar to that in

sperm. Chambers” first reported a Na’-re-

Fusion of Sperm and

Egg Plasma Membranes

Minor Influx of Sodium lons

Membrane Potential Change

(Rapid Block to Polyspermy)

Liberation of Calcium lons

18... From Intracellular Depots

20-~— Cortical Reaction Release of Acid

(Permanent Block to Polyspermy)

30~— Conversion of NAD to NADP

40-—-—— Rise in Oxygen Consumption

60—— Na dependent acid release

100

Increase in Intracellular pH

(Decreased Acidity)

300

~— Increase in Protein Synthesis

~~ Activation of Transport

Systems

400

1,000 Fusion of Egg and

=— Sperm Nuclei

<— Initiation of DNA Synthesis

First ‘Debaer

6,000-—=—= irst Cell Division

Fig. 4. Chronological list of events occurring in

the sea urchin egg following sperm-egg fusion. Time is

indicated logarithmically in seconds. Modified after

Epel.*?

8 DAVID NISHIOKA

quirement for activation and soon thereaf-

ter J Johnson et al.*° reported that this re-

quirement is for acid release. Beginning

about one minute after insemination and

lasting approximately four minutes, fertil-

ized eggs release acid into the surrounding

sea water. For a two percent suspension of

eggs (2.0 ml packed eggs suspended in 100

ml sea water) it is detected as a drop in pH

from 8.0 to 7.6.*° This release is coupled

with a measurable increase in pHi.”””*”” As

for sperm the effect of NH,” may be inter-

preted as an experimental bypass of normal

Na’-H® exchange and a direct increase of

pHi. In the egg the effect is a turnon of the

late events.

What then is the switch for the early

events? This is perhaps the more important

question since it would be more imme-

diately related to the overall initiating

event, the fusion of sperm and egg plasma

membranes, and would lead to increased

pHi. Referring to Figure 4, it is noted that

the first detectable response of the egg to

the fertilizing spermatozoon is a Na -de-

pendent depolarization of the egg plasma

membrane. This response is due to a minor

influx of Na” which causes the electrical

charge inside the egg to go positive relative

to the outside.’' The depolarization itself is

apparently not the switch, because direct

injection of a depolarizing current through

an intracellular electrode does not activate

the unfertilized egg.*’ Instead it appears to

function independently as a rapid and tran-

sient block to polypermy, ’’ i.e., the fertiliz-

ing spermatozoon induces the depolariza-

tion preventing supernumerary sperm from

entering the egg until more permanent

blocks are established. Neither does the

minor influx of Na’ appear to be the switch,

because upon stimulating sperm motility in

ONa’-SW (described in the preceding sec-

tion) in the presence of eggs, fertilization

and activation of the early events proceed.°

These eggs are however polyspermic be-

cause the depolarization is prevented in the

absence of Na’.°

The current idea, and one that has ac-

cumulated a convincing body of evidence

(reviewed by Jaffe*’ and by Epel®’), is that a

transient increase in the amount of intracel-

lular free calcium ions (Ca”’) is the switch

for the early events. Mazia”’ first reported

this change in 1937 and it has since been

confirmed by Nakamura and Yasumasu>

and Steinhardt er al/.*° The two crucial

questions which these findings have raised

are (1) is the increase a cause of activation,

and (2) if so, how is the increase achieved.

Both of these questions were approached in

an important study by Steinhardt and

Epel’’ using the calcium ionophore A23187

which selectively permeabilizes membranes

to Ca’*. They showed that treatment of un-

fertilized eggs with A23187 stimulated vir-

tually all of the events normally stimulated

by fertilization. Unlike ammonia-treated

eggs, ionophore-treated eggs are not fertil-

izable because the cortical reaction, an

early event, is stimulated and establishes a

permanent block to polyspermy. This reac-

tion involves a massive reorganization of

the egg surface and is readily observable as

the fertilization coat lifts away from the egg

plasma membrane and forms a hard inpene-

trable barrier to sperm.'’? Vacquier® has

presented evidence that the cortical reac-

tion is stimulated directly by increased lev-

els of Ca**. An important finding related to

the second question, how increased Ca” is

achieved, is that activation will proceed in

Ca*’-free sea water.*’ This finding elimi-

nates the need for Ca’’-influx and has led to

the idea that unfertilized eggs contain a

store of sequestered Ca” available for re-

lease upon fertilization. The nature of this

sequestration remains unknown. Since the

ionophore works by causing selective

permeability across membranes, one may

postulate that Ca” resides in some mem-

brane-bound organelle. Alternatively or

additionally, Ca** could be bound directly

to molecules within the egg. Calcium bind-

ing proteins have been reported in various

types of cells and tissues’? and Nakamura

and Yasumasu” have provided evidence

for one in sea urchin eggs.

THE IONIC BASIS OF FERTILIZATION 9

Given the two regulatory master switches

are increased intracellular free Ca” for the

early events and increased pH; for the late

events, certain predictions are possible.

Treatment of unfertilized eggs with NH,”

was shown to bypass the Ca’’-requiring

early events and selectively turn on the pH

-dependent late events. The reciprocal re-

sult, a turnon of the early events with an ar-

rest of the late events should be possible by

increasing intracellular Ca** under condi-

tions which prevent increased pHj. Such a

situation has been experimentally imposed

in fertilized eggs transferred to ONa*-SW,

and in unfertilized eggs treated with A23187

in ONa’-SW. Here fertilization coats are

raised indicating that the Ca’ increase has

occurred but the eggs remain arrested be-

cause Na’-H’ exchange and increased pHi

are prevented. Upon addition of Na’, acid

release, increased pH; and stimulation of

the late events proceed.*”*°*! Raising the

pH of the ONa’-SW to 9.0 with KOH or

adding 0.01 M NHg also stimulate the late

events in these eggs.*”** These experiments

are analogous to those described for the ac-

tivation of sperm metabolism and motility

on ONa’-SW.

Another prediction is possible for eggs

fertilized in ONa*-SW by NH,’-activated

sperm. In this situation sperm motility has

been activated by NH,’, and fertilization

proceeds in ONa’-SW. Since NH,’ is pres-

ent the pHi in the egg should also be raised

and allow for complete activation. This re-

sult has been reported.® Apparently, in-

creased pH; in both the sperm and egg has

been experimentally induced under condi-

tions which prevent Na’-H” exchange. A

notable difference in these eggs is that in-

stead of dividing normally into two cells at

first division, they divide into many cells.

The reason for this abnormality is that

these eggs are polyspermically fertilized

under the Na’-free conditions imposed.

The rapid block to polyspermy has been

bypassed.

The results described here leave little

doubt that the early events are experimen-

tally separable from the late events. The

problem that remains is to determine how

they are related to one another during

normal uninterrupted fertilization. Per-

haps the solution is already evident. The

early events accomplish two things. (1)

They prevent polyspermy, ensuring that

normal development will ensue, and (2)

they result in a massive reorganization of

the egg surface (the cortical reaction), al-

lowing the egg to interact in new ways with

its environment. The important interaction

is Na*-H* exchange which begins ona time

course consistent with the completion of

the cortical reaction. The interpretation is

that the early events prepare the egg to un-

dergo Na ’-H’ exchange, increased pHi, and

activation of the late events by causing the

proper surface changes. Viewed in this way

it is easier to understand how the unfertil-

ized egg remains suppressed in a medium

fully capable of supporting activation. It

does so by sequestering intracellular stores

of Ca** and preventing the surface changes

from occurring. It is also easier to under-

stand how the fertilizing spermatozoon

triggers the activation process. It does so by

stimulating Ca’’-release.

Summary and Implications for Future Study

As reviewed above, evidence has accumu-

lated that Na’-H" exchange and increased

pHi result in the activations of both the

sperm and egg in the sea urchin. In the

sperm these changes occur spontaneously

upon dilution of semen in sea water. In the

egg they are stimulated indirectly by the

fertilizing spermatozoon through a tran-

sient increase of intracellular Ca’”. It is

often argued that the common trigger, in-

creased pHi, would be too simple and too

nonspecific to explain the activation of

cells as different as sperm and eggs. These

arguments are based on misunderstand-

ings. Increased pHiis only the trigger. How

each cell responds to it is determined by

10 DAVID NISHIOKA

how each cell is genetically and differen-

tially programmed to respond to it. The

spermatozoon responds by becoming mo-

tile. The egg responds by reorganizing,

assembling, and synthesizing all of the

materials needed for cell division and de-

velopment of the embryo. Specificity is nec-

essarily maintained in the response.

What is now obviously needed is to de-

termine how universally applicable these

findings with sea urchin gametes are to

other cells which undergo activation. For

reasons mentioned earlier, studies with sea

urchin gametes have far exceeded those in

other cell systems, especially in terms of the

roles of ions in cellular activation. There

are however studies which would suggest

that various elements of the story in sea ur-

chin gametes are present in the activations

of other cells. Wong et al.“ have recently

presented evidence for Na’-H” exchange

and increased pH; in the activation of

sperm motility in the rat. The Ca’’-iono-

phore A23187 has been shown to activate

many different types of eggs,’* as well as

resting lymphocytes in tissue culture.*” An

increase in the pHi of frog eggs after fertili-

zation has been reported.*° If one accepts

the idea that cell surface changes are re-

quired for promoting ionic exchange and

cellular activation as proposed here for the

egg, there is an enormous body of work

available correlating functional and visible

changes of the cell surface with changes in

the metabolic states of virtually every type

of cell studied.*’ The importance of the

story in the sea urchin eggs is that a link has

been provided between the cell surface

changes and changes inside the cell.

Acknowledgments

The author’s work is supported by a

grant from the National Science Founda-

tion (No. PCM 7923487).

References Cited

1. Gray, J. (1928). The effect of dilution on the activ-

ity of spermatozoa. J. Exp. Biol., 5: 337-344.

2. Hayashi, T. (1946). Dilution medium and survival

12.

13;

14.

16.

of the spermatozoa of Arbacia punctulata. II. Ef-

fect of medium on respiration. Biol. Bull., 90;

177-187.

. Rothschild, Lord. (1948). The physiology of sea

urchin spermatozoa. Lack of movement in semen.

J. Exp. Biol., 25: 344-352.

. Ohtake, H. (1976). Respiratory behavior of sea

urchin spermatozoa. I. Effect of pH and egg water

on the respiratory rate. J. Exp. Zool., 198:

303-311.

. Schackman, R. W., E. M. Eddy and B. M. Shapiro.

(1978). The acrosome reaction of Strongylocen-

trotus purpuratus sperm: ion requirements and

movements. Dev. Biol., 65: 483-495.

. Nishioka, D. and N. Cross. (1978). The role of ex-

ternal sodium in sea urchin fertilization. In: Cell

Reproduction. E. R. Dirksen, D. Prescott and C.

F. Fox, eds., Academic Press, New York, pp.

403-413.

. Lee, H. C., S. Schuldiner, C. Johnson and D. Epel.

(1980). Sperm motility initiation: changes in in-

tracellular pH, Ca*’ and membrane potential. J.

Cell Biol., 87: 39a.

. Lee, H. C., J. G. Forte and D. Epel. (1982). The

use of fluorescent amines for the measurement of

pHi: applications in liposomes, gastric micro-

somes, and sea urchin gametes. In: Intracellular

PH: Its Measurement, Regulation, and Utilization

in Cellular Processes. R. Nuccitelli and D. W.

Deamer, eds., Alan R. Liss Inc., New York.

. Christen, R., R. W. Schackman and B. M. Shapiro.

(1981). Coupling of sperm respiration, motility,

and intracellular pH (pHi) in Strongylocentrotus

purpuratus. J. Cell Biol., 91: 174a.

. Johnson, C. H. (1981). Activation of echinoderm

gametes: the role of intracellular pH. Ph.D.

Dissertation, Stanford University, Stanford,

California.

Steinhardt, R. A., L. Lundin and D. Mazia. (1971).

Bioelectric responses of the echinoderm egg to

fertilization. Proc. Nat. Acad. Sci. USA, 68:

2426-2430.

Runnstrom, J. (1966). The vitelline membrane and

cortical particles in sea urchin eggs and their func-

tion in maturation and fertilization. Advan. Mor-

phogen. 5: 221-325.

Epel, D. (1964). Simultaneous measurement of

TPNH formation and respiration following fertil-

ization of the sea urchin egg. Biochem. Biophys.

Res. Commun., 17: 69-73.

Steinhardt, R. A., S. Shen and D. Mazia. (1972).

Membrane potential, membrane resistance and

an energy requirement for the development of

potassium conductance in the fertilization reac-

tion of echinoderm eggs. Exp. Cell Res., 72:

195-203.

. Whiteley, A. H. and E. L. Chambers. (1966).

Phosphate transport in fertilized sea urchin eggs.

II. Effects of metabolic inhibitors and studies on

differentiation. J. Cell Physiol., 68: 309-324.

Piatigorsky, J. and A. H. Whiteley. (1965). A

20.

21.

22.

23:

24.

26.

27.

28.

29.

30.

<5 i

THE IONIC BASIS OF FERTILIZATION 11

change in permeability and uptake of '“C-uridine

in response to fertilization in Strongy/ocentrotus

purpuratus eggs. Biochim. Biophys. Acta, 108:

404-418.

. Epel, D. (1972). Activation of an Na’-dependent

amino acid transport system upon fertilization of

sea urchin eggs. Exp. Cell Res., 72: 74-89.

. Epel, D. (1967). Protein synthesis in sea urchin

eggs: a “‘late”’ response to fertilization. Proc. Nat.

Acad. Sci. USA, 57: 899-906.

. Hinegardner, R. T., B. Rao and D. E. Feldman.

(1964). The DNA synthetic period during early

development of the sea urchin egg. Exp. Cell Res.,

36: 53-61.

Mazia, D. (1974). Chromosome cycles turned on

in unfertilized sea urchin eggs after exposure to

NH,OH. Proc. Nat. Acad. Sci. USA, 71: 690-693.

Mazia, D. and A. Ruby. (1974). DNA synthesis

turned on in unfertilized sea urchin eggs by treat-

ment with NH,OH. Exp. Cell Res., 85: 167-172.

Steinhardt, R. A. and D. Mazia. (1973). Develop-

ment of K-conductance and membrane potentials

in unfertilized sea urchin eggs after exposure to

NH,OH. Nature, 241: 400-401.

Epel, D., R. Steinhardt, T. Humphreys and D.

Mazia. (1974). An analysis of the partial meta-

bolic derepression of sea urchin eggs by ammonia;

the existence of independent pathways in the pro-

gram of activation at fertilization. Dev. Biol., 40:

245-255.

Nishioka, D. and D. Mazia. (1977). The phospho-

rylation of thymidine and the synthesis of his-

tones in ammonia-treated eggs and egg fragments

of the sea urchin. Cell Biol. Internat. Rep., 1:

23-30.

Shen, S. S. and R. A. Steinhardt. (1978). Direct

measurement of intracellular pH during meta-

bolic derepression at fertilization and ammonia

activation of the sea urchin egg. Nature, 272:

253-254.

Winkler, M. M. and J. L. Grainger. (1978). Mech-

anism of action of NH,Cl and other weak bases in

the activation of sea urchin eggs. Nature, 273:

536-538.

Grainger, J. L., M. M. Winkler, S. S. Shen and R.

A. Steinhardt. (1979). Intracellular pH controls

protein synthesis rate in the sea urchin egg and

early embryo. Dev. Biol., 68: 396-406.

Johnson, C. H. and D. Epel. (1981). Intracellular

PH of sea urchin eggs measured by the dimethyl-

oxazolidinedione (DMO) method. J. Cell Biol.,

89: 284-291.

Chambers, E. L. (1976). Na is essential for activa-

tion of the inseminated sea urchin egg. J. Exp.

Zool., 197: 149-154.

Johnson, J. D., D. Epel and M. Paul. (1976). Intra-

cellular pH and activation of sea urchin eggs after

fertilization. Nature, 262: 661-664.

Jaffe, L. A. (1976). Fast block to polyspermy in

sea urchin eggs is electically mediated. Nature,

261: 68-71.

ae;

34.

36.

ST.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

Jaffe, L. F. (1980). Calcium explosions as triggers

of development. Annals New York Acad. Sci.,

339: 86-101.

Epel, D. (1980). Ionic triggers in the fertilization

of sea urchin eggs. Annals New York Acad. Sci.,

339: 74-85.

Mazia, D. (1937). The release of calcium in

Arbacia eggs on fertilization. J. Cell. Comp. Phy-

siol., 10: 291-308.

Nakamura, M. and I. Yasumasu. (1974). Mecha-

nism for increase of free calcium in fertilized sea

urchin eggs. A method for estimating intracellular

concentration of free calcium. J. Gen. Physiol.,

63: 374-388.

Steinhardt, R. A., R. Zucker and G. Schatten.

(1977). Intracellular calcium release at fertiliza-

tion in the sea urchin egg. Dev. Biol., 58: 185-196.

Steinhardt, R. A. and D. Epel. (1974). Activation

of sea urchin eggs by calcium ionophore. Proc.

Nat. Acad. Sci. USA, 71: 1915-1919.

Vacquier, V. D. (1975). The isolation of intact cor-

tical granules from sea urchin eggs: calcium ions

trigger granule discharge. Dev. Biol., 43: 62-74.

Siegel, F. L. (1973). Calcium-binding proteins. In:

Structure and Bonding, Vol. 17. J. D. Dunitz, P.

Hemme rich, J. A. Ibers, C. K. Jorgensen, J. B.

Neilands, D. Reinen, R. J. P. Williams, eds.,

Springer-Verlag, New York.

Nishioka, D. and N. F. McGwin. (1980). Relation-

ships between the release of acid, the cortical reac-

tion, and the increase of protein synthesis in sea

urchin eggs. J. Exp. Zool., 212: 215-223.

Nishioka, D. and L. S. Magagna. (1981). Increased

uptake of thymidine in the activation of sea ur-

chin eggs: specificity of uptake and dependence

on internal pH, the cortical reaction, and external

sodium. Exp. Cell Res., 133: 363-372.

Nishioka, D. and D. Epel. (1977). Intracellular pH

and the activation of sea urchin eggs at fertiliza-

tion. J. Cell Biol., 75: 40a.

Wong, P. Y. D., W. M. Lee and A. Y. F. Tsang.

(1981). The effects of extracellular sodium on acid

release and motility initiation in rat caudal epidi-

dymal spermatozoa in vitro. Exp. Cell Res., 131:

97-104.

Steinhardt, R. A., D. Epel, E. J. Carroll and R.

Yanagimachi. (1974). Is calcium ionophore a uni-

versal activator for unfertilized egg? Nature, 252:

41-43.

Crumpton, M. J., B. Perles and J. Auger. (1977).

Mitogen receptors and mitogen responsiveness.

In: International Cell Biology 1967-1977. B. R.

Brinkley and K. R. Porter, eds., The Rockefeller

University Press, New York, pp. 119-127.

Webb, D. J. and R. Nuccitelli. (1980). The intracel-

lular pH goes alkaline at fertilization in the

Xenopus egg. J. Cell Biol., 87: 137a.

The Cell Surface: Mediator of Developmental Proc-

esses. S. Subtelny and N. K. Wessells, eds., Aca-

demic Press, New York.

Journal of the Washington Academy of Sciences,

Volume 72, Number 1, Pages 12-23, March 1982

The Movements During Fertilization

Gerald Schatten

Department of Biological Science, The Florida State University,

Tallahassee, Florida 32306

ABSTRACT

The movements during fertilization, the systems responsible for each motion, the effects of

microtubule and microfilament inhibitors and the regulation of the associated motility are

reviewed using the sea urchin system as a paradigm. Following the swimming of the sperm to

the egg surface, mediated by the sliding of axonemal microtubules, polymerization of actin

present in the periacrosomal cap of the sperm results in the elongation of the acrosomal pro-

cess. The acrosomal process, extruded as a result of the formation of bundles of micro-

filaments, establishes the initial contact between the sperm and egg. The egg actively responds

to the successful sperm by elongating adjacent microvilli to surround and engulf the successful

sperm. These microvilli continue to enlarge, forming a bulbous structure, the fertilization

cone, which is responsible for sperm incorporation. Sperm incorporation occurs in two stages,

first the enlargement of the fertilization cone around the sperm and second the lateral dis-

placement of the sperm from the site of fusion into the egg cytoplasm proper. Sperm incorpo-

ration requires the assembly of egg cortical microfilaments. The lateral displacement of the

sperm along the egg cortex is terminated by the assembly of microtubules nucleated at the base

of the sperm head, the position occupied by the sperm centrioles. Microtubule assembly con-

tinues off of the sperm centrioles to form at first a radial structure, the sperm aster. The growth

and the formation of the sperm aster first displace the male pronucleus (sperm nucleus) from

the egg cortex towards the egg center. When the microtubules from the sperm aster contact the

surface of the female pronucleus (egg nucleus) these microtubules seem to undergo a selective

disassembly, resulting in the pulling of the female pronucleus to the center of the sperm aster.

Subsequently the now adjacent pronuclei are moved to the egg center by the continuing en-

largement of the sperm aster. Pronuclear fusion (syngamy) typically occurs at the egg center

after the sperm aster has disassembled. The first cell cycle is characterized by other cycles of

assembly and disassembly of cytoskeletal components, including bursts in microvillar length

and the microtubule-mediated formation of the interim apparatus or “‘streak.”’ In this system,

within an hour, the microtubules responsible for the formation of the mitotic apparatus form

and regress as do the microfilaments comprising the contractile ring that effect cytokinesis.

Fertilization represents a unique model in which to study cellular motility since the proper

functioning of each gamete requires microtubule- and microfilament-mediated activities and

since virtually every example of movement is manifested.

1. The Requirement for Movement

During Fertilization

For fertilization to be successful several

movements must occur. Sperm must be

transported to a position near the egg sur-

face, the sperm and egg must achieve the

close contact required to effect membrane

fusion and sperm must be drawn into the

egg cytoplasm proper. Once within the egg

cytoplasm the sperm and egg nuclei must

locate one another and move together to

establish a contact that will eventually re-

sult in the fusion of their nuclear envelopes.

Following the successful fusion of the ma-

ternal and paternal genomes, subsequent

events necessary to prepare the zygote for

cell division and embryogenesis must

THE MOVEMENTS DURING FERTILIZATION 13

shortly ensue (see accompanying review by

Dr. Belisle).

As a model for studying cellular motility,

fertilization is unique for a variety of rea-

sons. Perhaps foremost is polarity of each

gamete’s quest for survival; the nature of

the sperm’s motility is to propel it through

the suspending fluid first to the proximity

of and then into actual contact with the

egg. Contrasting with this cellular migra-

tion are the egg’s movements, which con-

sist entirely of intracellular translocations.

First the egg participates in sperm incorpo-

ration, 1.e., the motion bringing the at-

‘tached sperm from the egg exterior into its

cytoplasm, and then the intracellular mi-

grations of the sperm and egg nuclei that

result in syngamy. Simultaneous with these

motions of the sperm and egg, dramatic

- surface alterations involved with the estab-

lishment of the block to polyspermy and

metabolic activation occur. With the com-

pletion of pronuclear fusion the now fertil-

ized egg is prepared to begin to undergo

cleavages and the morphogenetic move-

ments required for proper development.

Though fertilizaton is indeed unexcelled as

a system for studying motion, few animal

systems readily lend themselves to the sort

of experimentation required for biochemi-

cal and microscopic analyses. Of the avail-

able systems in which in vitro fertilization is

routinely obtainable, the echinoderm sys-

tem is unrivaled. Closely mimicking the

events occurring in mammals, this system

permits the routine collection of hundreds

of millions of perfectly synchronized eggs,

which will fertilize at a near-perfected rate.

The synchrony, yield and routine availabil-

ity of this system, when coupled with the

glass-like transparency of eggs of some

species, has rendered sea urchin fertiliza-

tion the system of choice for cellular, bio-

chemical and molecular investigations for

Over a century. In this review the move-

ments during echinoderm fertilization will

be predominantly considered; readers in-

terested in mammalian fertilization are di-

rected to the recent reviews by Gwatkin'

and Yanagimachi.”

2. Sperm Incorporation

The ultrastructural features of sperm i1n-

corporation have been documented by

Longo and Anderson,’ Schatten and

Mazia,’ Schatten and Schatten,’ and Tilney

and Jaffe.° In Figure la, the attachment of

sperm to the egg shortly after insemination

is depicted following the binding of the

sperm to the egg surface by the apical re-

gion of the sperm head, the acrosome. As

observed in Figure 1b, microvilli adjacent

to the successful sperm elongate and cluster

to form the anlage of the fertilization cone.

The activity of the egg surface during

sperm incorporation is particularly well

documented in Figures Ic and d, in which

the extracellular surface coats were re-

moved to allow a direct view of the plasma

membrane; here the engulfment of the

sperm by the egg cortex and plasma mem-

brane is clearly visible because the over-

lying layers have been removed.

Recent advances in video microscopy

using differential interference contrast op-

tics have permitted the recording of the

movements during fertilization in the liv-

ing. ”* In Figure 2 the entire sequence of fer-

tilization is documented. Unlike electron

microscopy, which requires the study of

fixed, and therefore static, specimens, video

tape recording permits a relatively high de-

gree of resolution in the living state in

which the sequence of fertilization is di-

rectly observed rather than compiled from

a sequence of still micrographs. In Figure

2, the initial stages of sperm incorporation

are observed as involving first the attach-

ment of the sperm by the acrosomal proc-

ess to the egg surface. Following a varying

period of time, during which the sperm gy-

rates about its attachment site, the sperm

stands erect on the egg surface and the mo-

tility of the sperm tail ceases. Moments

later the elevation of the fertilization coat

around the successful sperm occurs and the

unsuccessful sperm attached to the egg sur-

face are lifted from the plasma membrane

by the elevation of this extracellular coat.

The fertilization cone begins to form around

Fig. la. Insemination observed by scanning electron microscopy. An early stage of insemination of an egg

glued to a polylysine-coated slide. Only the tops and sides of the egg are available for sperm binding. Strongylo-

centrotus purpuratus. Bar: 10 um. Reprinted, with permission, from ref. #4. lb. The egg membrane rises around

the spermhead. Microvilli elongate around the spermatozoon as the membrane derived from the sperm appears

slack and convoluted. S. purpuratus. Bar: 1 um. Reprinted, with permission, from ref. #4. lc. In these eggs,

devoid of their vitelline layers, the activity of the egg surface in engulfing the sperm is clearly apparent. Microvilli

have elongated, to 1.2 um, to completely surround the successful sperm. These microvilli will continue to elon-

gate to form the fertilization cone. L. variegatus. Bar: 1 um. Reprinted, with permission, from Schatten and

Schatten, 1980. 1d. The fertilization cone forms from these elongating microvilli, which surround the base of

the fertilization cone, and which continue to engulf the sperm. Note the microvilli surrounding the sperm tail.

L. variegatus. Bar: | um. Reprinted, with permission, from ref. #5.

Fig. 2. Movements during Fertilization Studied in Living Eggs. Time-lapse video microscopy of fertilization

with water immersion, differential interference contrast optics. Sperm-egg attachment occurs in A at 1:36 (min:

sec). The sperm tail becomes immotile in B and a second later the fertilization coat (white arrow) elevates over

the attached sperm (black arrow). The fertilization cone forms around and above the erect sperm in F-I. The

oe i a ie i ee

15 ty

MWY

Static sperm tail, which projects through the elevated fertilization coat, can be observed in E-H. The displace-

ment of the sperm nucleus (male pronucleus) within the egg cytoplasm occurs in I-P; the sperm tail beats errati-

cally at this stage. The sperm aster forms as the male pronucleus is moved centripetally (Q—-U). In V, the field has

been shifted to include the sperm aster (large black arrow) and the female pronucleus. Fibers radiating from the

sperm aster are denoted by black v’s in T-CC. The migration of the female pronucleus to the center of the sperm

aster occurs in W-Z; the female pronucleus is distorted from a sphere to an ovoid during this migration. Pro-

nuclear centration (AA-BB) occurs as the fibers of the sperm aster (black v’s) continue to elongate. Small parti-

cles (black triangles) appear on the nuclear surface in BB: these particles may represent the centrioles since they

are positioned along the presumptive axis for mitosis. Cleavage (DD) occurs parallel to the direction of pro-

nuclear centration. L. variegatus. Bar: 10 um. Reprinted, with permission. from ref. #8.

16 GERALD SCHATTEN

the perpendicularly oriented and static

spermatozoon. Shortly afterward, the sperm

rotates 90° to lie parallel with the egg cor-

tex, and then begins to undergo a lateral

displacement along the egg cortex from the

site of sperm-egg fusion. Concomitant with

this lateral displacement the erratic beating

of the sperm tail is observed, perhaps caus-

ing this displacement. The momentary ar-

rest in tail beating and the later resumption

of this example of ciliary motility may be

indicators of the changes in cytoplasmic

ionic conditons (see section 6). Following

the movement of the sperm along the egg

cortex, the sperm is discharged into the egg

cytoplasm proper with its mid-piece di-

rected towards the egg center.

In summary then, the events during

sperm incorporation following the swim-

ming of the sperm to the egg surface and

the contact by the exterior acrosomal proc-

ess of the sperm with the egg involve first

the formation of the fertilization cone

around the erect and static sperm and

then the rotation and displacement of the

sperm along the egg cortex, which dis-

charges it into the egg cytoplasm in a ro-

tated position so that its centriole faces to-

ward the egg center. This latter point is of

importance when the significance of the

centriole contributed by the sperm during

the pronuclear migrations is considered

and during later development (see section

3).

3. Pronuclear Migrations

In this section, the cytoplasmic migra-

tions of the male pronucleus (sperm nu-

cleus) and female pronucleus (egg nucleus)

will be traced from the moment when the

sperm leaves the egg surface following in-

corporation to that at which the pronuclei

fuse. Syngamy completes the fertilization

process. The terminology used throughout

this chapter will refer to the unincorpo-

rated sperm nucleus as a sperm nucleus; the

sperm nucleus within the egg cytoplasm

will be referred to as the male pronucleus;

the egg nucleus will be referred to as the

female pronucleus. The transition from the

genetically inactive sperm nucleus into the

active metabolic state is reviewed in this

monograph by Professor Poccia.

The documentation of the pronuclear

migrations has been a difficult undertaking

since most eggs are relatively opaque be-

cause of the presence of numerous yolk

platelets and since the incorporated sperm

nucleus migrates centripetally from the

surface, where it is visible, into the egg

center, where it is not. Modern sophis-

tications in optics have increased the visibil-

ity of the pronuclei and their motile struc-

ture, and quite importantly, the nearly

transparent egg of the Gulf coast sea urchin

Lytechinus variegatus has contributed great-

ly in the living documentation. The pro-

nuclear movements at fertilization involve

the formation of the sperm aster, at the in-

itial stages a radially symmetrical structure

emanating from the sperm centrioles at the

base of the rotated sperm mid-piece (see

Figures 2 and 3a). The formation of the

sperm aster moves the male pronucleus

centripetally at a rate of 4.9 um/min. Con-

comitant with this centrad motion the male

pronucleus begins to undergo chromatin

condensation (see accompanying review by

Professor Poccia). When the microtubules

of the sperm aster contact the surface of the

female pronucleus the next of the three

pronuclear migrations occurs, i.e. the mi-

gration of the female pronucleus (Figure

2). The movement of the female pronucleus

to the center of the sperm aster is the swift-

est and most dramatic of the pronuclear

migrations, occurring at a rate of 14.6

m/min, often traversing half the diameter

of the egg. The final movement of the now

adjacent pronuclei is dependent on the ex-

tension of the sperm astral microtubules,

which push the pronuclei to the egg center.

This final motion occurs at a rate of 2.6

m/min. The fusion of the pronuclei typi-

cally occurs at the egg center shortly after

THE MOVEMENTS DURING FERTILIZATION 17

Fig. 3. Antitubulin Immunofluorescence Micros-

copy. A. Growth of sperm aster. The sperm aster is

moved into the cytoplasm of the egg, accompanied by

the elongation of astral microtubules. Sperm nucleus

is visible as an area from which microtubules are ex-

cluded; all microtubules appear to be organized

around the sperm midpiece. A. punctulata. Many of

the fibers visible in these micrographs are of substan-

tially lower intensity than the sperm axoneme, sug-

gesting that microtubule bundles containing only a

few microtubules are visible by immunofluorescence

microscopy. Bars: 10 um. B. Mitotic Apparatus.

From unpublished work of Balczon and Schatten.

the sperm aster has reached its maximal

size.

4. Effects of Motility Inhibitors

Fertilization is a superb model for study-

ing the effects of selective inhibitors of mo-

tility. The beating of the sperm tail, which

propels the sperm to the egg surface, is an

example of the sliding of adjacent micro-

tubules.” The extrusion of the sperm acro-

somal process requires the assembly of mi-

crofilaments,'° as does the formation of the

fertilization cone required during sperm

incorporation.”'' '* The centripetal migra-

tion of the male pronucleus and the later

centration of the adjacent pronuclei’® are

systems of movement in which microtubules

appear to push the pronuclei away from the

surface. In this case the egg cortex, follow-

ing the migration of the female pronucleus

to the center of the sperm aster (Figure 3a),

appears to require the pulling of micro-

tubules; this is emphasized by the distor-

tion of the typically spherical female pro-

nucleus into an oblate spheroid during its

migration (Figure 2). The use of motility

inhibitors to study the mechanism respon-

sible for fertilization has been investigated

by a number of workers. Though the acro-

some reaction appears insensitive to cyto-

chalasin B,'° eggs treated with microfilament

inhibitors are unable to incorporate the

spermatozoon even though sperm-induced

egg activation occurs. In contrast, micro-

tubule inhibitors have been utilized to dem-

onstrate that syngamy requires microtubule

assembly and that pronuclear fusion and

the onset of DNA synthesis are independ-

ent processes./7~””

The effects of inhibitors of microfila-

ment assembly, microfilament disassem-

bly, microtubule assembly, and micro-

tubule disassembly are summarized in

Table 1. The inhibitors of microfilament

assembly, cytochalasins B, D and E, pre-

vent sperm incorporation if added prior to

or simultaneously with insemination, but

they have no effect on pronuclear migra-

tions if added after incorporation. This

pattern indicates that the functioning of the

egg cortical microfilaments is necessary for

the drawing of the sperm from the exterior

into the egg cytoplasm but that these micro-

filaments play no role during the subse-

quent movements of the pronuclei. The

inhibitor of microfilament disassembly

phalloidin slows the rate of sperm incorpo-

18 GERALD SCHATTEN

Table 1—Summary of Effects of Motility Inhibitors

Microfilament

Microtubule

Inhibitors Inhibitors

(Assembly) (Disassembly) (Assembly) (Disassembly)

Colcemid,

griseofulvin,

nocodazole,

maytansine,

vinblastine,

Cytochalasins Phalloidin et al. Taxol, D2O

Sperm-Egg Attachment and Fusion ate a + a5

Cortical Reaction, Fertilization Coat

Elevation =i = = ln

Fertilization Cone Formation -—— teste oie te

Lateral Displacement of Sperm

during Incorporation —— = aaa ts

Restructuring of Fertilized Egg

Cortex -—— ar =e +

Formation of Sperm Aster = + —— ate

Migration of Female Pronucleus ate Si ee oe

Pronuclear Centration tr SI — ——

Syngamy + at op —-—

Formation of Streak == 56 —- —

Mitosis SF = a ——

Cytokinesis ae — + +

Legend: —— (event blocked); — (event retarded); + (normal event); ++ (event enhanced).

ration and results in a larger persistant fer-

tilization cone; indications that an active

dynamic equilibrium between assembly and

disassembly of actin in microfilaments

might well be occurring in the fertilization

cone. Inhibitors of microtubule assembly,

colchicine, colcemid, griseofulvin, maytan-

sine, nocodazole, podophyllotoxin, and

vinblastine, all permit sperm incorporation

but prevent the subsequent migrations of

the pronuclei by blocking the assembly of

microtubules that form the sperm aster.

This clearly indicates that whereas sperm

incorporation does not involve the func-

tioning of microtubules, these subsequent

cytoplasmic movements of the pronuclei

do. The inhibitor of microtubule disassem-

bly taxol permits the initial formation of

the sperm aster but blocks its disassembly

with the result that the migration of the fe-

male pronucleus, perhaps moved by the

disassembly of microtubules, is inhibited.

Furthermore the sperm aster remains as an

almost crystalline structure and in the ab-

sence of its disassembly the centrioles can-

not separate and the mitotic apparatus

cannot form. This then indicates that, in

addition to the proper assembly of micro-

tubules, their subsequent disassembly is

also required for pronuclear fusion and

subsequent development. Insummary then

almost any inhibitor of motility will block

the normal repertoire of motility during

fertilization; the proper union of the sperm

and egg genomes requires an intricate and

orderly assemblage of motile components,

followed by their subsequent disassembly.

5. The Egg Cytoskeleton

The cytoskeleton arrays of the egg are as

exceptional as are its other features. The

unfertilized egg is the only higher cell com-

pletely devoid of any assembled micro-

THE MOVEMENTS DURING FERTILIZATION 19

tubules and microfilaments. Following in-

semination, cytoskeletal elements form de

novo, another extraordinary event. In the

normal case of fertilization, the sperm

brings with it the trigger to assemble micro-

filaments of the fertilization cone and the

microtubules of the sperm aster. These

localized assemblages are converted into

global events with the subsquent restruc-

turing of the fertilized egg cortex and the

microfilament-mediated extension of the

egg microvilli. Alternately the separation

of the pair of sperm centrioles leads to the

formation of the mitotic apparatus (Figure

3b) and the subsequent cytoskeletal reor-

ganizations during division, which result in

the mathematical paradox of multiplica-

tion by division. In addition to the cyto-

skeletal rearrangements active during cell

division, bursts in microvillar length”’ are

noted during the first cell cycle as is the

formation of a transient microtubule struc-

ture, the streak or interim apparatus. The

significance of the streak and the bursts in

microvillar elongation is not yet clear.

It is also instructive to analyze the cyto-

skeletal rearrangements during artificial

activation. Clues concerning the natural

regulation during fertilization can be easily

inferred and furthermore, in the absence of

a sperm, its direct contribution to the reor-

ganization of the cytoskeleton following

fertilization can be assessed. Microvilli of

activated eggs elongate as shown by Mazia

et al.*> Microtubules also form in artifi-

cially activated eggs.**”” However, in the

absence of a sperm centriole, these micro-

tubules do not develop into the sperm aster

but rather start as a subcortical disarray.

These microtubules elongate to form a ra-

dial shell that moves the female pronucleus

towards the egg center, and finally the shell

coalesces at the egg center to form the apo-

lar mitotic apparatus.~* It appears then in

the unfertilized egg that all the components

to form the microfilaments and micro-

tubules of the egg cytoskeleton exist. How-

ever, the seeds that normally result in the

proper polarity and orientation of the cyto-

skeletal components and the trigger signal-

ling the onset of polymerization are missing

and must be either artificially induced or

provided by the inseminating sperm.

6. Regulation of Motility During

Fertilization

The program of activation, reviewed by

Professor Nishioka in this monograph, has

been compiled as a result of nearly a cen-

tury of work. Though details of the scheme

are still under active investigation the es-

sential features are summarized as follows:

following the acrosome reaction the sperm

has greatly elevated levels of internal cal-

cium and intracellular pH. It is at this stage

that it fuses its membrane with the plasma

membrane of the egg and thereby triggers

the onset of development. It is presently

thought that the sperm enters the egg ef-

fectively as a calcium bomb (estimates of

the intracellular calcium concentration in

the acrosome reacted sperm are over

10mM) and perhaps as a pH bomb. An al-

ternative view is that the contribution of

the sperm membrane to the egg membrane,

with its presumed calcium channels serving

as endogenous ionophores, permits an in-

flux of sufficient external Ca™ to trigger ac-

tivation. The sudden and localized increase

in cytoplasmic calcium at the site of sperm-

egg fusion is sufficient to trigger the explo-

sive discharge of adjacent cortical granules.

This initial discharge stimulates a propa-

gating wave of released calcium from intra-

cellular stores,”° which is followed by the

secretion of the cortical granules. This cal-

cium transient is concluded within four

minutes of sperm-egg fusion when the cal-

cium is presumably resequestered. The in-

itial release of calcium triggers a sodium:

proton exchange, which then results 1n an

increased cytoplasmic pH. This alkaliniza-

tion of the egg cytoplasm appears to be the

pervasive ionic trigger for signaling the egg

that it is now fertilized.

20 GERALD SCHATTEN

For the study of the ionic regulation of

the egg cytoplasm, two signals appear at-

tractive: the intracellular calcium release

triggered by the sperm, and/or the subse-

quent increase in intracellular pH. Four

different sorts of experimental manipula-

tions can be used to pose the question,”’

‘‘Which is the dominant regulator for the

formation and functioning of the egg cyto-

skeleton during fertilization, the intracellu-

lar calcium release and/or the rise in intra-

cellular pH.”’

The first groups of experiments involve

cases in which both intracellular pH and

calcium fluxes occur: fertilization or acti-

vation with the divalent ionophore, A23187.

In these cases the cytoskeleton forms and

functions as indicated by the intracellular

translocations of the pronuclei.

In the second case for studying the ionic

manipulations during fertilization only the

intracellular release of calcium is permit-

ted. These are eggs in which either the

Na’: H” exchange has been blocked by the

removal of all extracellular Na” or sodium

acetate at acidic pH’s, which will diffuse

and reduce the intracellular pH to the un-

fertilized values, is added. In this group of

experiments, using either natural fertiliza-

tion or artificial activation with the dival-

ent ionophore A23187, no microfilament-

mediated and no microtubule-mediated

motion occurs and the cytoskeleton does

not form. This indicates that the intracellu-

lar release of calcium alone is insufficient to

permit the formation and functioning of

the egg cytoskeleton.

The third group of experiments pertain-

ing to this question involve cases in which

intracellular pH elevation alone is induced

using alkaline NH4Cl in the absence of ex-

ternal Ca”, procaine, or nicotine. In these

cases of artificial activation microfilament-

mediated events as indicated by microvillar

elongation observed with the scanning

electron microscope, microtubule-mediated

events observed by antitubulin microscopy,

and time-lapse video studies of female

pronuclear centration indicate the forma-

tion and motility of the egg cytoskeleton. It

appears that intracellular pH in the ab-

sence of an intracellular Ca™ release can re-

sult in cytoskeletal formation and activity.

The last group of experiments bearing on

the ionic regulation of motility during fer-

tilization are cases in which the intracellu-

lar Ca™ release is first triggered and then

subsequently the rise in intracellular pH is

permitted. The Ca” release is triggered by

the divalent ionophore A23187 or by sperm

and when the intracellular elevation is

blocked no motility is observed. However,

when the intracellular pH is either permit-

ted to rise or artificially elevated by diffus-

ible weak bases, motion then ensues and, in

the case of fertilization, development will

occur but is delayed by the time span dur-

ing which the intracellular pH rise was

prevented.

To summarize these experiments, intra-

cellular pH appears to bea primary regula-

tor of motility during fertilization.”’ How-

ever, it also appears reasonable to conclude

at this stage that secondary regulators and

Fig. 4. The Movements During Fertilization. Sperm attach to the egg surface (A) and gyrate about their

attachment sites (B) for varying times prior to fusion (C). Following a rapid cortical contraction radiating from

the fusion site, the fertilization coat elevates (D-F). Sperm incorporation is characterized by the formation of

the fertilization cone around the erect and stationary sperm; the sperm tail is immotile at this stage (D-F). The

sperm glides along the egg cortex during penetration (G, H). The formation of the sperm aster moves the male

pronucleus centripetally (I, J). The migration of the female pronucleus occurs when the fibers of the sperm aster

interconnect the pronuclei (K, L). The adjacent pronuclei are moved to the egg center by the continuing elonga-

tion of the sperm aster (M); the centrioles may separate during this motion, and the sperm aster appears to have

two focal points. Syngamy typically occurs at the egg center after the disassembly of the sperm aster (N). The

streak forms around and distorts the zygote nucleus (O). The axis of the streak (O) is usually identical to the axis

of the mitotic apparatus (Q). The streak is disassembled prior to the nuclear breakdown at prophase (P). Cleav-

age is perpendicular to the axis of the mitotic apparatus and is usually parallel to the egg radius passing through

the sperm entry site (Q). Reprinted, with permission, from ref. #8.

22 GERALD SCHATTEN

modulators of motility will be observed to

be active during fertilization. Fluxes in cal-

cium ions may well play a role in regulating

the fine tuning of each event, e.g. it is con-

ceivable that during the initial alkaliniza-

tion of the egg cytoplasm both microtu-

bules and microfilaments might assemble.

However, the prevailing [Ca™] might be

too high for microtubules to form and con-

sequently only the microfilament-mediated

motions are observed during the first cou-

ple of minutes. Following sequestration of

the released Ca™ the cytoplasmic concen-

tration is reduced to a level permitting the

formation and subsequent functioning of

the sperm aster.

7. Summary

The schematic diagram (Figure 4) sum-

marizes the movements during fertilization

and is based primarily on observations of

living recordings. The beating of the sperm

tail propels the spermatozoon to the egg

surface near, or perhaps at, the egg surface

the acrosome reaction occurs, whereupon

the acrosomal process is extruded from the

apex of the sperm head. This process estab-

lishes the initial contact between the ga-

metes by effectively harpooning the egg

surface. The sperm, attached by this acro-

somal process, continues to beat actively,

resulting in the gyration of the sperm about

its attachment site on the egg surface. A

varying time later sperm-egg fusion occurs,

characterized first by the sudden immobili-

zation of the sperm tail, with the sperm

head and mid-piece held in an erect and

perpendicular fashion on the egg surface.

The fertilization cone begins to form on the

egg surface at the site where the sperm head

is attached and the fertilization coat ele-

vates over the attached sperm and propa-

gates from the site of attachment to envelope

the now fertilized egg. Unsuccessful sperm

attached to the vitelline layer are physi-

cally removed from the egg surface by the

elevation of the fertilization coat. Sperm

incorporation involves first the formation

of the fertilization cone around the station-

ary and erect sperm and then later the rota-

tion and lateral displacement of the sperm

head, mid-piece and tail along the egg cor-

tex. Though the sperm tail is immotile at

the instant of sperm-egg fusion it begins to

beat during the latter stages of sperm in-

corporation and continues to beat within

the egg cytoplasm in an erratic fashion. It

should be noted that in virtually all recent

studies the sperm tail has been found to be

fully incorporated into the fertilized egg

cytoplasm. The pronuclear migrations begin

with the formation of the sperm aster

emanating from the base of the sperm head

and mid-piece. The sperm aster first pushes

the male pronucleus centripetally, and upon

contact with the female pronucleus, it pulls

the egg nucleus to the center of the sperm

aster. The now contiguous pronuclei are

pushed to the center of the egg cytoplasm,

whereupon pronuclear fusion occurs. The

remainder of the first cell cycle is character-

ized by another burst of microvillar elonga-

tion and the formation and regression of

the streak prior to the events at cell di-

vision: mitosis and cytokinesis.

The activity of microfilaments assem-

bling to form, in the sperm, the acrosomal

process and, in the egg, first the fertiliza-

tion cone and then the global rearrange-

ment of the egg microvilli is well docu-

mented. The sliding of the microtubules in

the sperm axoneme and the sequential as-

sembly and disassembly of microtubules to

form first the sperm aster, then the streak,

and finally the mitotic apparatus is equally

clear. The ionic regulation within the egg

for the assembly and subsequent function-

ing of the cytoskeleton appears to be pre-

dominately under the control of intracel-

lular pH, with, perhaps, the intracellular

calcium concentration as a modulator.

Acknowledgments

The support of the original research de-

scribed in this report by a research grant

THE MOVEMENTS DURING FERTILIZATION

(HD12913) and a Research Career Devel-

opment Award (HD00363) from the Na-

tional Institutes of Health is gratefully

acknowledged.

10.

References

. Gwatkin, R. B. L. (1978). Fertilization Mechanisms

in Man and Mammals. Plenum Press.

. Yanagimachi, R. (1978). Sperm-egg association in

mammals. Curr. Topics Develop. Biol., 4: 83-104.

. Longo, F. and E. Anderson. (1968). The fine struc-

ture of pronuclear development and fusion in the

sea urchin Arbacia punctulata. J. Cell. Biol., 39:

339-368.

. Schatten, G. and D. Mazia. (1976). The surface

events at fertilization: The movements of the

spermatozoon through the sea urchin egg surface

and the roles of the surface layers. J. Supramolec.

Struct., 5: 343-369.

. Schatten, H. and G. Schatten. (1980). Surface ac-

tivity at the egg plasma membrane during sperm

incorporation and its cytochalasin B sensitivity:

Scanning electron microscopy and time lapse

video microscopy during fertilization of the sea

urchin Lytechinus variegatus. Develop. Biol., 78:

435-339.

. Tilney, L. G. and L. A. Jaffe. (1980). Actin, micro-

villi and the fertilization cone of sea urchin eggs.

J. Cell Biol., 81: 608-623.

. Schatten, G. (1981la). The movements and fusion

of the pronuclei at fertilization of the sea urchin

Lytechinus variegatus. J. Morph., 167: 231-247.

. Schatten, G. (1981b). Sperm incorporation, the

pronuclear migrations and their relations to the

establishment of the first embryonic axis. Develop.

Biol. 86: 426-437.

. Gibbons, I. (1981). Cilia and flagella of euka-

ryotes. J. Cell Biol., 91: 1075-1245.

Tilney, L. G., D. P. Kiehart, C. Sardet and M. Til-

ney. (1978). Polymerization of actin. IV. Role of

Ca™ and H’ in the assembly of actin and in mem-

brane fusion in the acrosomal reaction of echino-

derm sperm. J. Cell Biol., 77: 536-550.

. Gould-Somero, M., L. Holland and M. Paul.

(1977). Cytochalasin B inhibits sperm penetration

into eggs of Urechis caupo. Develop. Biol., 58:

11-22.

. Longo, F. (1978). Insemination of immature sea

urchin (Arbacia punctulata) eggs. Develop. Biol.,

62: 271-291.

13.

ID:

16.

20.

22.

23:

24.

26.

Zt.

Byrd, W. and G. Perry. (1980). Cytochalasin B

blocks sperm incorporation but allows activation

of the sea urchin egg. Exp. Cell Res., 126:

333-342.

. Longo, F. (1980). Organization of microfilaments

in sea urchin (Arbacia punctulata) eggs at fertiliza-

tion: Effects of cytochalasin B. Develop. Biol., 74:

422-431.

Schatten, G. (1982). The movements of the nuclei

at fertilization. In: Cellular Biodynamics: Mitosis

and Cytokinesis. A. Zimmerman and A. Forer,

eds., Academic Press, New York. pp. 59-82.

Sanger, J. W. and J. M. Sanger. (1975). Polymeri-

zation of sperm actin in the presence of cytocha-

lasin B. J. Exp. Zool., 193: 441-447.

. Zimmerman, A. M. and S. Zimmerman. (1967).

Action of colcemid in sea urchin eggs. J. Cell

Biol., 34: 483-488.

. Aronson, J. F. (1973). Nuclear membrane fusion

in fertilized Lytechinus variegatus eggs. J. Cell

Biol., 58: 126-134.

. Schatten, H. (1977). Untersuchungen uber die

Wirkung von Griseofulvin in Seeigeleiern und Mam-

malierzellen. Dissertation, University of Heidel-

berg, West Germany.

Schatten, G. and H. Schatten. (1981). Effects of

motility inhibitors during sea urchin fertilization.

Exp. Cell Res., 135: 311-330.

Schatten, H., C. Petzelt, D. Mazia and G. Schat-

ten. (1982). Effects of griseofulvin on fertilization

and early development. Eur. J. Cell Biol., in press.

Schroeder, T. E. (1979). Surface area changes at

fertilization. Develop. Biol., 70: 306-326.

Mazia, D., G. Schatten and R. Steinhardt. (1975).

Turning on of activities in unfertilized sea urchin

eggs: Correlation with changes of the surface.

Proc. Natl. Acad. Sci. USA, 72: 4469-4473.

Paweletz, N. and D. Mazia. (1979). Fine structure

of the mitotic cycle of unfertilized sea urchin eggs

activated by ammoniacal sea water. Eur. J. Cell

Biol., 20: 37-40.

. Bestor, T. and G. Schatten. (1981). Anti-tubulin

microscopy of microtubules present during the

pronuclear movements of sea urchin fertilization.

Dev. Biol., 88: 80-91.

Steinhardt, R., R. Zucker and G. Schatten. (1977).

Intracellular calcium release at fertilization in the

sea urchin egg. Dev. Biol., 58: 185-196.

Schatten, G., T. Bestor, H. Schatten and C. Cline.

(1981). Regulation of motility during fertilization:

Intracellular pH as a primary modulator. J. Cell

Biol., 91: 184.

Journal of the Washington Academy of Sciences,

Volume 72, Number 1, Pages 24-34, March 1982

Biochemical Aspects of Sperm Nucleus Activation

by Egg Cytoplasm

Dominic L. Poccia

Department of Biology, Amherst College, Amherst, Massachusetts 01002

ABSTRACT

The sperm nucleus is highly unusual in structure and is genetically inactive. Beginning im-

mediately after penetration of the egg at fertilization, it is transformed into a more typical

nucleus. This transformation is brought about by action of the egg cytoplasm which acquires

this ability late in oogenesis. Molecular details of male nuclear activation are beginning to

emerge which may apply to the control of chromosome structure and gene expression in other

cell types as well.

I. Introduction

The act of fertilization creates a single to-

tipotent cell, the zygote, which in turn gives

rise through a series of divisions to every

other cell of the organism. Fertilization in

the sea urchin involves the fusion of two

highly differentiated cells, the spermato-

zoan and the mature oocyte or egg. It is an

unequal partnership. In the laboratory the

egg alone can give rise to an apparently

normal organism through parthenogenesis

(for example, by simple chemical stimula-

tion). Under normal conditions, however,

the sperm contributes two important ele-

ments to the zygote: a haploid set of

chromosomes (the paternal gene set) anda

pair of centrioles (organelles used in the or-

ganization of the apparatus used for segre-

gating chromosomes at cell division).

The remainder of the sperm provides a

means for efficient motility and for pene-

tration of and binding to the egg vestments,

processes required for successful delivery

of the chromosomal and centriolar pay-

load to the egg. Virtually all else used in

early embryonic development is provided

by the egg. This disparity in contribution of

the two gametes is suggested by the dispro-

portionality of relative sizes of the gametes

and is substantiated at the biochemical

level.

As development proceeds, the embry-

onic nuclei (all descendents of the zygote

nucleus to which the egg and sperm

DNA contribute equally) become increas-

ingly important in directing development

through information sent to the cytoplasm

as messenger RNA (mRNA). But the main

theme of the very earliest stages of devel-

Opment in most organisms is the impor-

tance of maternal cytoplasm. Its effects

may be mediated in two ways: 1) by direct

expression of information already present

in mature egg cytoplasm accumulated dur-

ing differentiation of the oocyte in oogene-

sis, and 2) by interaction of egg cytoplasm

with nuclei to alter the structure or activity

of the chromosomes or chromatin. The

early embryo is perhaps best viewed in

terms of nucleocytoplasmic interactions,

information flow occurring in both

directions.

The purpose of this paper is to review in-

formation relevant to the activation of the

dormant sperm nucleus following its pene-

ACTIVATION OF THE MALE PRONUCLEUS 25

tration of egg cytoplasm. I will first de-

scribe those features of the sea urchin

sperm nucleus which distinguish it from

other nuclei of the embryo or adult, then

discuss macromolecular properties of egg

cytoplasm which may bear upon its ability

to transform the male nucleus, and finally

treat some biochemical aspects of male nu-

clear activation. More comprehensive and

technical reviews on certain of these sub-

jects can be found in the bibliography.'~

References are given to illustrative papers

and no attempt has been made to be

comprehensive.

II. The Sperm Nucleus

A. Spermatogenesis

During the formation of the mature

sperm cell (spermatogenesis), the chroma-

tin of the precursor cell, the spermatogo-

nium, replicates to give rise to the chro-

matin of the primary spermatocyte. The

spermatocyte undergoes two meiotic divi-

sions to yield four spermatids which then

without division differentiate into the ma-

ture spermatozoan. The nuclei of sperma-

togonia look like typical somatic cell nuclei

in the electron microscope. However, the

nuclei of early spermatids are of an irregu-

lar shape and contain granular aggregates

of chromatin displayed in no particular

order against a background of more diffuse

chromatin.° Chromatin condensation con-

tinues progressively until in the mature

sperm the chromatin packing is extremely

dense. Dense packing of chromatin is usu-

ally a sign of genetic inactivity. The differ-

entiation of the spermatid (spermiogenesis)

has not been studied biochemically in the

sea urchin, but by analogy to other orga-

nisms, it is likely that during this period

RNA synthesis ceases and sperm-specific

nuclear proteins become associated with

the DNA.

B. Nuclear proteins

Chromosomes of all nucleated cells ex-

cept sperm contain DNA packaged with

highly basic nuclear proteins called his-

tones. There are five histone types: H2A,

H2B, H3, H4 and H1. The first four (core

histones) aggregate to form a structure

called a nucleosome. The nucleosome con-

sists of an octamer of two each of the core

histones. DNA wraps around the outside

of the core. A length of chromatin consists

of an unbroken string of DNA connecting

cores together. The length of DNA be-

tween cores Is called linker DNA and H1 is

probably associated with the linker.

Sperm cells from various organisms con-

tain a bewildering array of nuclear proteins

associated with the DNA. These range

from typical histones to proteins which are

not even basic.’ In the sea urchin sperm,

histones are present but three of the five

(H1, H2A, and H2B) differ from the so-

matic or embryonic forms electrophoreti-

cally and by amino acid composition and

sequence.* '' Sperm H3 and H4 differ little

if at all from their embryonic counterparts.

No multiple secondary modifications due

to phosphorylation, acetylation or methy]I-

ation of any of the histones has been de-

tected, unlike histones of virtually all other

cell types examined.* In addition levels of

non-histone chromosomal proteins are ex-

tremely low.

C. Genetic inactivity

Sperm chromatin is inactive in DNA or

RNA synthesis, lacks associated DNA or

RNA polymerases,'”’® and is associated

with little if any RNA.”

D. Chromatin structure

In the electron microscope sperm chro-

matin shows typical nucleosomal organi-

zation,’ and this organization is con-

firmed biochemically by nuclease digestion

studies.'> The typical 145 base pairs of

26 DOMINIC L. POCCIA

DNA associated with each core is separ-

ated on average by about 100 base pairs of

linker DNA, making the length of one re-

peating unit (240-260 base pairs) the long-

est known for chromatin in any cell of any

organism. The cores differ from embryonic

cores in physical properties such as thermal

stability and digestion rates or cutting sites

of various nucleases. ’°

E. Summary

The sperm nucleus differs from somatic

nuclei in its composition, structure, and ac-

tivity. Its chromatin contains histones only

found in sperm chromatin, and little if any

non-histone proteins or RNA. Its chro-

matin is highly compact and stable and or-

ganized into long repeat units. It is inactive

in DNA or RNA synthesis.

III. Maternal Storage of Macromolecules

A. RNA Synthesis During Oogenesis

As in spermatogenesis, the production of

a mature egg involves transformation of an

oogonium to a primary oocyte which

through meiotic division gives rise to four

ootids. In contrast to formation of sper-

matids, the divisions are unequal, produc-

ing one large cell and three small ones. In

the case of the sea urchin, the large celli is

the mature egg. Having completed meiosis

it contains a haploid amount of nuclear

DNA and is fertilized in this state. Its

growth (about 2000X in volume) occurs in

the primary oocyte stage, the latter part of

which (vitellogenesis) involves accumula-

tion of large amounts of yolk. During vitel-

logenesis large amounts of ribosomal RNA

(rRNA) are synthesized and accumulate.’’

The total sequence complexity of the RNA

present in the oocyte (the sum of the

lengths in nucleotides of all the qualita-

tively different RNA sequences present) in-

creases throughout development of the

primary oocyte.’

The nuclear DNA of higher organisms

falls into two classes: those nucleotide se-

quences which are present in essentially

one copy per haploid genome, and those

which are present in many copies (reiter-

ated). Most messenger RNA’s are comple-

mentary to the single copy DNA. By the

time it is mature, the sea urchin oocyte con-

tains RNA transcribed from about 6% of

its single copy DNA. Much of this corre-

sponds to structural gene sequences whose

mRNA’s are translated in the early em-

bryo.'® If all of this RNA (30-40 X 10° nu-

cleotides long) were mRNA it could code

for as many as 30,000 different average-

sized proteins.

Of the total RNA in the mature egg

about 80% is ribosomal RNA.’ Approxi-

mately 10’ ribosomes are present per egg

and maternal ribosomes account for virtu-

ally all the rRNA of early development. All

other factors needed for protein synthesis

such as tRNA’s and various enzymes are

abundant in the egg.

About 1-3% of the total mass of egg

RNA is mRNA. An important class of

maternal RNA is histone mRNA which is

derived from the reiterated DNA sequence

class. It accounts for 4-8% of the total

mRNA.’

Not only does the mature egg contain a

variety of different RNA sequences and a

large amount of RNA (3.3 ng/egg), but its

nucleus unlike that of the sperm is actively

engaged in RNA synthesis.’® *'

B. Precursor Pools

Mature eggs contain pools of low molec-

ular weight molecules which are precursors

in the synthesis of macromolecules. Sub-

stantial pools of ribonucleotides (RNA

precursors) and deoxyribonucleotides

(DNA precursors) have been measured.

Total pools of deoxyribonucleotides can

support embryonic DNA synthesis for sev-

eral cell cycles.”

Amino acid pools however are only suf-

ficient for a few minutes of protein synthe-

ACTIVATION OF THE MALE PRONUCLEUS 27

sis.”> They are presumably replenished by

breakdown of yolk protein which is 50-80%

of the total egg protein. About 1% of total

egg protein is converted per hour.

C. Protein Storage

Many enzymes and structural proteins

are stored in the egg including DNA and

RNA polymerases, and a variety of en-

zymes involved in nucleotide metabolism

and protein synthesis. Of particular inter-

est are the polymerases responsible for rep-

licating and transcribing the DNA. Even

though the number of nuclei/embryo in-

creases exponentially during the cleavage

stage of development,” the total activi-

ties/embryo of these enzymes remain

constant.*”*°

The presence of large amounts of deoxy-

ribonucleotides and DNA _ polymerase

allow the very rapid rates of DNA replica-

tion seen in the early embryo. All the new

DNA must be complexed with an equal

mass of histone. This is provided during

development by a combination of mater-

nally stored histone protein, translation of

stored mRNA coding for histones, and

translation of histone mRNA newly tran-

scribed from the embryonic genes. The

predominant type of histone found within

the first few cell cycles after fertilization is

of the so-called cleavage-stage (CS) variant

class. These are the only variant types

stored in the egg and are present in several

hundred haploid equivalents/egg. They are

stored therefore in great excess over the

amounts required by the male or female

nuclei.”*’ Evidence for a pool of non-his-

tone chromosomal proteins has also been

presented.”

D. Macromolecular Responses of Egg

Cytoplasm to Fertilization

The most thoroughly studied macromo-

lecular transition occurring in the zygote

following fertilization is the activation of

protein synthesis.”* The rate of synthesis

increases about 15-fold by two hours post-

fertilization.”” This stimulation is accom-

panied by a 30-fold increase in the fraction

of ribosomes found in polysomes but not

by changes in the efficiency of translation

(i.e., the rate of protein synthesis/poly-

somal mRNA/time).*” Protein synthesis

activation occurs even when transcription

of the maternal or paternal genes is blocked

with the drug actinomycin D, as well as in

enucleated egg cytoplasms. Therefore it

has been attributed to the recruitment of

maternal mRNA from an untranslated

store in the mature egg to an actively trans-

lating polyribosomal form. In the first two

hours of development 90% of the poly-

somal mRNA is maternal whereas by gas-

trula stage virtually all is newly synthesized

from embryonic genes.”

The maternal RNA is sufficient to sup-

port essentially normal development to the

blastula stage as demonstrated by the abil-

ity of embryos to grow in the presence of ac-

tinomycin D. Quantitatively, the most im-

portant proteins made during cleavage

stages (between fertilization and blastula)

are nuclear proteins.©’ Among cleavage

stage transcripts from embryonic genes,

histone mRNA accounts for a substantial

portion of all mRNA synthesis,” but by

utilizing maternal histone mRNA’s and

stored histone and non-histone chromo-

somal proteins the embryo can apparently

complete cleavage stages in the absence of

embryonic transcription. Blocking protein

synthesis, however, stops development

within a single cell cycle.

An interesting question regarding the re-

cruitment of maternal mRNA after fertili-

zation is whether it is selective. Can a qual-

itatively distinct set of mRNA’s be selected

for translation compared to the set already

being translated in the unfertilized egg? In

general, it appears that most of the abund-

ant mRNA’s being translated before or

after fertilization code for the same

proteins.»

An exception is provided by the histone

family. Although mRNA’s coding for both

28 DOMINIC L. POCCIA

the CS variants (predominant in early

cleavage chromatin) and a variants (pre-

dominant in late cleavage) are present in

the unfertilized egg as demonstrated by

translation in a cell-free system, in the liv-

ing egg only the CS forms appear to be

translated.** (It will be recalled that only

CS variants are stored maternally.) How-

ever, after fertilization alpha variants begin

to be synthesized, and so their mRNA’s

appear to be selectively recruited.

E. Summary

Oogenesis produces an extraordinarily

large egg cell containing substantial pre-

cursor pools, enzymes of nucleic acid me-

tabolism, histones, ribosomes and factors

involved in protein synthesis, and an

enormously diverse store of genetic infor-

mation in the form of mRNA molecules.

This maternal store of genetic information

is sufficient for early development without

major contribution by the embryo’s genes.

The egg cytoplasm can therefore control

the transformation of the male nucleus fol-

lowing fertilization and provide for the

very rapid production of chromosomes re-

quired to form a multicellular blastula em-

bryo from a unicellular zygote.

IV. Transformation of the Sperm Nucleus

by the Egg

A. Pronuclear Development

A critical event in normal early devel-

opment of all organisms is the morphologi-

cal and biochemical transformation of the

inactive sperm nucleus. In sea urchin egg

cytoplasm, the sperm nucleus becomes the

male pronucleus which fuses with the egg

nucleus (female pronucleus) to form the

zygote nucleus. The egg pronuclear chro-

matin is already decondensed and active.

Decondensation and swelling of the male

pronucleus in S. purpuratus occurs within

15-20 minutes post-fertilization proceed-

ing from the nuclear periphery towards the

core, ”’*° and during a fraction of this pe-

riod the nuclear envelope is absent (see

Figure 1), although the female pronucleus

retains its envelope. Thus the male chro-

matin is rather directly exposed to egg cy-

toplasm. Highly condensed chromatin is

characteristic of other genetically inactive

nuclei and reactivation of these in foreign

cytoplasms is usually accompanied by

swelling and ingress of cytoplasmic

proteins.

The conditions which promote decon-

densation appear to be absent from previ-

tellogenic or vitellogenic oocytes and only

begin to appear during meiotic maturation

divisions.”** In the mature egg, however,

maintenance of these conditions is not de-

pendent on the presence of the female pro-

nucleus, nuclear or mitochondrial RNA

synthesis, or protein synthesis.*”*° The

conditions persist for a time into

embryogenesis.”'

B. Nucleic Acid Synthesis

DNA synthesis normally follows pronu-

clear fusion but fusion is not required.

DNA synthesis in both pronuclei is in-

itiated at 20-30 minutes post-fertilization

depending on species.*” The 30 minute lag

is absent from the next few cell cycles as —

DNA synthesis follows immediately upon

mitotic chromosome decondensation. The

first mitotic condensation occurs at about

90 minutes (Figure 1). The early DNA syn-

thetic periods in monospermic eggs may

last only 5-15 minutes making them among

the most rapid known in higher organ-

isms.*? DNA synthesis is confined to the

Fig. 1. Diagrammtic representation of various male nuclear transitions in the first cell cycle following fertili-

zation of the sea urchin S. purpuratus at low degrees of polyspermy. Timing and amounts are approximate.

From A. Savic, P. Richman, P. Williamson, and D. Poccia (1981). Proc. Natl. Acad. Sci. USA 78: 3706-3710.

-©®@ O

pronuclear mitotic

decondensation condensation

NUCLEAR

MORPHOLOGY

DNA

SYNTHESIS

DNA 7 CELL

VARIANTS

H2B

VARIANTS

H2A

VARIANTS

SPACING

BASE PAIRS

NUCLEOSOME

0 30 60 90

TIME POST — FERTILIZATION ( MIN.)

30 DOMINIC L. POCCIA

nucleus. Although most DNA in the egg is

mitochondrial, this DNA does not repli-

cate in early development. It is made in

oogenesis.

Nuclear RNA synthesis though difficult

to detect apparently occurs in the first cell

cycle.*’ The nature of the initial transcripts

is not known. However, both paternal and

maternal nuclear transcripts of histone

mRNA can be detected as early as the two

cell stage.**

By inference entry of DNA and RNA

polymerases into male pronuclei must occur

in the first cell cycle. DNA synthesis in the

first cycle is largely independent of protein

synthesis following fertilization, although

protein synthesis seems to be required for

the next round of replication.** Severe in-

hibition of nuclear or mitochondrial RNA

synthesis has little effect on progression of

the first cycle.”

C. Nuclear Proteins

Biochemical transitions must occur in

the male pronucleus in the direction of

those properties that distinguish embry-

onic nuclei from sperm nuclei as outlined in

Section II. Most of these are known only by

comparison of blastula or later embryonic

nuclei with sperm. For several technical

reasons it has been virtually impossible to

study isolated male pronuclear chromatin

on a biochemical level in any organism.

These reasons are: 1) the high ratio of cyto-

plasm to nucleus in fertilized eggs which re-

sults in formidable purification problems;

2) difficulties in obtaining synchronously

developing fertilized eggs in sufficient

number to allow requisite quantities of

chromatin for biochemical analysis to be

isolated; and 3) the inability to distinguish

the maternal and paternal contributions to

the chromatin isolated from an egg fertil-

ized by only one sperm. These problems

have been largely circumvented in the sea

urchin by use of polyspermically fertilized

eggs. °°? Similar problems apply to analysis

of the unfertilized egg nucleus or female

pronucleus so biochemical information on

these is limited.

Our knowledge of transitions in sperm

specific nuclear proteins following fertili-

zation has been largely based on cytochem-

ical procedures such as staining reactions.

Most studies suggest that a period occurs

between the loss of sperm specific nuclear

proteins and the acquisition of “‘typical”’

adult histones during which the male chro-

matin has proteins different from either.*°

Using polyspermic eggs a much more de-

tailed description is now available for the

first cell cycle.’ The first transition to take

place is the complete replacement of sperm

H1 by CS H1. This occurs almost imme-

diately following fertilization and well be-

fore pronuclear decondensation is com-

plete (Figure 1). Roughly in parallel with

decondensation, a gradual loss of sperm

H2B’s and proportional gain of proteins O

and P occur. O and P are distinguished by

gel electrophoresis in the presence of a non-

ionic detergent, but have virtually the same

molecular weights as the sperm H2B’s. O

and P areas yet incompletely characterized

and may represent either modified forms of

the sperm H2B’s or stored egg histone var-

iants. During replication CS variants of

H2A and H2B accumulate in large amounts

on the chromatin, supplementing sperm

H2A and O, P respectively. Several other

changes are seen as well including accumu-

lation of a form of H3 called 3”. By the time

of the first chromosome condensation,

male pronuclear chromatin consists of a

heterogeneous group of histone variants

composed of proteins the sperm nucleus

brought into the egg and histone variants

from the maternal store. An important

point is that the chromatin remodeling that

occurs in the first cell cycle is accomplished

by a fairly complex set of transitions which

do not occur coordinately. The discrete

temporal segregation of specific transitions

Suggests distinct functions for each.

Whether earlier transitions must occur in

order for later transitions to take place is

not known. The extent to which given var-

ACTIVATION OF THE MALE PRONUCLEUS 31

iants are secondarily modified is also un-

known except for CS H1.*° In any event

male chromatin at the end of the first cycle

is considerably more complex in histone

composition than sperm chromatin.

All the histone transitions outlined will

take place in eggs blocked in protein syn-

thesis.”*’ In a polyspermic egg this is taken

as evidence for a minimal store of 25-50

functional haploid equivalents of each var-

iant/egg. Independent measurements sug-

gest an upper limit of several hundred hap-

loid equivalents.”’

Under conditions where DNA synthesis

is blocked by aphidicolin (a DNA polymer-

ase inhibitor) assembly of nucleosomes

from the pool apparently continues.*’ In

this case some of the sperm core histones

must be replaced by CS variants.

The most likely candidates to be in-

volved in decondensation of the chromatin

are the H1 and H2B classes since decon-

densation is complete before the other

transitions have progressed very far.

D. Chromatin Structure

During development of the sea urchin,

sets of histone variants appear sequentially

in the chromatin.** Once incorporated into

chromatin they are stable and thus passed

on to future cell generations. The elabora-

tion of this program then must result in an

increasing heterogeneity of chromatin

within the organism as development pro-

gresses. The functional outcome of this in-

creasing diversity is not known, but inter-

esting speculations have been presented

relating histone variant inheritance pat-

terns to simple proliferative cell division

and stem cell generation and mainte-

nance.”

Some circumstantial evidence exists link-

ing histone variant changes with differen-

ces in physical properties of the core his-

tones’® or changes in average nucleosomal

repeat lengths.”° In the first cell cycle the

repeat length of the male pronuclear chro-

matin remains at the high level characteris-

tic of sperm, then declines during replica-

tion to levels typical of embryonic

chromatins (Figure 1). Since Sp H1/CS H1

and SpH2B/O,P transitions take place

without change in spacing, they cannot at

least by themselves cause the repeat length

decline. H1, the noncore histone, has been

a leading candidate in the literature for set-

ting the repeat length.

The decline in repeat length coincident

with replication suggests a requirement for

DNA synthesis to allow adjustment of nu-

cleosome spacing, perhaps by altering

chromatin structure in such a way as to

allow sliding of cores along the DNA fiber.

Under conditions where DNA synthesis is

blocked, the decline in repeat length is

also blocked lending weight to this inter-

pretation.*’ Since the SpH2A/CSH2A and

O,P/CSH2B transitions occur in the ab-

sence of DNA synthesis they also can be

ruled out as sufficient cause of the altera-

tion in spacing.

The highest level of chromatin organiza-

tion in normal cell cycles is the packing of

chromatin into mitotic chromosomes. Con-

ditions for promoting chromosome con-

densation have been studied in mono- and

polyspermic sea urchin eggs.°’“° These

conditions are assayed either by observa-

tion of endogenous chromosomes or by

chemically activating the eggs to start the

maternal cell cycle and subsequently fertil-

izing at various times to test for the conver-

sion of sperm chromatin directly to

chromosomes (premature chromosome

condensation) instead of to decondensed

pronuclei. These studies show that chromo-

some condensing conditions can develop

even in egg halves devoid of the maternal

nucleus, although they are more ephemeral

in enucleated eggs suggesting a role of the

egg nucleus in stabilizing condensing con-

ditions. The conditions develop independ-

ently of RNA synthesis but require protein

synthesis prior to the end of replication.”

A biochemical transition occurring to

the histone complement during mitotic or

premature chromosome condensation,

32 DOMINIC L. POCCIA

also seen in other cell types, is extensive

phosphorylation of histone H1 (in this case

CS H1). H1 phosphorylation and chromo-

some condensation can however be un-

linked in the egg. If protein synthesis is in-

hibited, CS H1 is recruited from the stored

pool and becomes just as highly phosphor-

ylated as in controls from uninhibited cul-

tures. However chromosome condensation

is blocked.Thus phosphorylation of H1

cannot drive chromosome condensation by

itself. The experiment also shows that all

proteins needed for CS H1 phosphoryla-

tion and its timing are already present in

the unfertilized egg and suggest arolefor a

newly synthesized protein(s) in chromo-

some condensation.

V. Summary

Using the sea urchin polyspermic egg it

has now become possible to describe a va-

riety of macromolecular transitions occur-

ring in the male pronucleus to a degree of

detail never before approached. This should

allow an analysis of the relationship of var-

lous aspects of the reactivation of the

sperm nucleus such as protein composi-

tion, chromatin structure and genetic activ-

ity. In addition to contributing to our un-

derstanding of events of the first cell cycle

of the urchin, many of the lessons learned

are likely to be generally applicable to

problems of chromatin structure and the

activation or inactivation of replication

and gene expression in a variety of cell

types and organisms.

Acknowledgments

The author’s work is supported by grants

from the National Institutes of Health

(Nos. HD 096654 and HD 12982).

References Cited

1. Longo, F. J. and M. Kunkle. (1978). Transforma-

tions of sperm nuclei upon insemination. Curr.

Topics in Devel. Biol., 12: 149-184.

2. Guidice, G. (1973). Developmental Biology of the

Sea Urchin Embryo. Acad. Press, New York.

10.

18.

. Davidson, E. H. (1976). Gene Activity in Early De-

velopment, Second Edition. Acad. Press, New

York.

. Kedes, L. H. (1979). Histone genes and histone

messengers. Ann. Rev. Biochem. 48: 837-70.

. Hentschel, C. C. and M. L. Birnstiel. (1981). The

organization and expression of histone gene fam-

ilies. Cell 25: 301-13.

. Longo, F. J. and E. Anderson. (1969). Sperm dif-

ferentiation in the sea urchins Arbacia punctulata

and Strongylocentrolus purpuratus. J. Ultrast.

Res. 27: 486-509.

. Bloch, D. P. (1969). A catalog of sperm histones.

Genetics (Supp.) 61: 93-111.

. Easton, D. and R. Chalkley. (1972). High resolu-

tion electrophoretic analysis of the histones from

embryos and sperm of Arbacia punctulata. Exp.

Cell Res. 72: 502-508.

. Poccia, D., J. Salik and G. Krystal. (1981). Transi-

tions in histone variants of the male pronucleus

following fertilization and evidence for a mater-

nal store of cleavage-stage histones in the sea ur-

chin egg. Devel. Biol. 82: 287-296.

Strickland, W. N., M.S. Strickland, P. C. DeGroot

and C. van Holt. (1980). Primary-structure of his-

tone H2A from the sperm cell of the sea urchin

Parechinus angulosus. Eur. J. Biochem. 109:

151-58.

Strickland, W. N., M. Strickland, W. F. Brandt, C.

van Holt, A. Lehmann and B. Wittmann-Liebold.

(1980). The primary structure of histone H1 from

sperm of the sea urchin Parechinus angulosus. Se-

quence of the C-terminal CNBr peptide and the

entire primary structure. Eur. J. Biochem. 104:

567-78.

. Slater, J. P. and L. Loeb. (1970). Initiation of

DNA synthesis in eukaryotes: A model in vitro

system. Biochim. Biophys. Res. Comm. 41:

589-593.

. Paoletti, R. A. and R. C. Huang. (1969). Character-

ization of sea urchin sperm chromatin and its

basic proteins. Biochemistry 8: 1615-1625.

. Zentgraf, H., U. Muller and W. W. Franke. (1980).

Supranucleosomal organization of sea urchin

sperm chromatin in regularly arranged 40 to 50

nm large granular subunits. Eur. J. Biochem. 20:

254-64.

. Spadafora, C., M. Bellard, J. L. Compton and P.

Chambon. (1976). The DNA repeat lengths in

chromatins from sea urchin sperm and gastrula

cells are markedly different. FEBS Lett. 69:

281-285.

. Simpson, R. T. and L. W. Bergman. (1980). Struc-

ture of sea urchin sperm chromatin core particles.

J. Biol. Chem. 255: 10702-10709.

. Gross, P. R., L. I. Malkin and M. Hubbard. (1965).

Synthesis of RNA during oogenesis in the sea ur-

chin. J. Mol. Biol. 13: 463-481.

Hough-Evans, B. R., S. G. Ernst, R. J. Britten and

ert .

19.

20.

2h.

23.

24.

2a.

26.

27.

28.

30.

a1.

32.

a3.

34.

ACTIVATION OF THE MALE PRONUCLEUS 33

E. H. Davidson. (1979). RNA complexity in de-

veloping sea urchin oocytes. Devel. Biol. 69:

258-269.

Levner, M. H. (1979). RNA transcription in ma-

ture sea urchin eggs. Exp. Cell Res. 85: 296-302.

Brandhorst, B. P. (1980). Simultaneous synthesis,

translation and storage of messenger RNA includ-

ing histone mRNA in sea urchin L. pictus eggs.

Dev. Biol. 79: 139-48.

Ruderman, J. V. and M. R. Schmidt. (1981). RNA

transcription and translation in sea urchin oo-

cytes and eggs. Dev. Biol. 81: 220-228.

Nemer, M. (1962). Characteristics of the utiliza-

tion of nucleosides by embryos of Paracentrotus

lividus. J. Biol. Chem. 237: 143-149.

Fry, B. J. and P. R. Gross. (1970). Patterns and

rates of protein synthesis in sea urchin embryos.

II. The calculation of absolute rates. Devel. Biol.

21: 125-146.

Hinegardner, R. T. (1967). Echinoderms, In: Meth-

ods Developmental Biology. F. H. Wilt and N. K.

Wessells, eds., Cromwell-Collier, New York, pp.

139-155.

Loeb, L. A. and B. Fansler. (1970). III. intracellu-

lar migration of DNA polymerase in early devel-

oping sea urchin embryos. Biochim. Biophys.

Acta 217: 50-55.

Roeder, R. G. and W. J. Rutter. (1970). Multiple

RNA polymerases and ribonucleic acid synthesis

during sea urchin development. Biochem. 9:

2543-2553.

Salik, J., L. Herlands, H. P. Hoffmann and D.

Poccia. (1981). Electrophoretic analysis of the

stored histone pool in unfertilized sea urchin eggs:

Quantification and identification by antibody

binding. J. Cell Biol. 90: 385-395.

Kuhn, O. and F. H. Wilt. (1981). Chromatin pro-

teins of sea urchin embryos: Dual origin from an

oogenetic reservoir and new synthesis. Devel.

Biol. 85: 416-424.

Epel, D. (1967). Protein synthesis in sea urchin

eggs: A “‘late’’ response to fertilization. Proc.

Natl. Acad. Sci. USA 57: 899-904.

Humphreys, T. (1971). Measurement of RNA en-

tering polysomes upon fertilization of sea urchin

eggs. Devel. Biol. 26: 201-208.

Kedes, L. H., P. R. Gross, G. Cognetti and A. L.

Hunter. (1969). Synthesis of nuclear and chromo-

somal proteins on light polyribosomes during

cleavage in the sea urchinembryo. J. Molec. Biol.

45: 337-351.

Nemer, M. (1975). Developmental changes in the

synthesis of sea urchin embryo messenger RNA

containing and lacking polyadenylic acid. Cell 6:

559-570.

Brandhorst, B. (1976). Two-dimensional gel pat-

terns of protein synthesis before and after fertili-

zation of sea urchin eggs. Devel. Biol. 52: 310-317.

Herlands, L., V. G. Allfrey and D. Poccia. (1982).

35.

36.

37.

38.

35.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

Translational regulation of histone synthesis in

the sea urchin Strongylocentrotus purpuratus. J.

Cell Biol., in press.

Longo, F. J. and E. Anderson. (1969). The fine

structure of pronuclear development and fusion

in the sea urchin Arbacia punctulata. J. Cell Biol.

39; 339-368.

Poccia, D., G. W. Krystal, D. Nishioka and J.

Salik. (1978). Control of sperm chromatin struc-

ture by egg cytoplasm in the sea urchin. In: “ICN-

UCLA Symposium on Molecular and Cellular

Biology: Cell Reproduction XII (E. R. Dirksen,

D. M. Prescott, and C. F. Fox eds.), Acad. Press,

New York, pp. 197-206.

Harris, H. (1974). Nucleus and Cytoplasm, Clar-

endon Press, Oxford.

Longo, F. J. (1978). Insemination of immature sea

urchin (Arbacia punctulata) eggs. Devel. Biol. 62:

271-291.

Krystal, G. W. and D. Poccia. (1979). Control of

chromosome condensation in the sea urchin egg.

Exp. Cell Res. 123: 207-219.

Krystal, G. W. and D. L. Poccia. (1981). Phosphor-

ylation of cleavage stage histone HI! in mitotic

and prematurely condensed chromosomes. Exp.

Cell Res. 134: 41-48.

Sugiyama, M. (1951). Refertilization of the fertil-

ized eggs of the sea urchin. Biol. Bull. 101:

335-340.

Hinegardner, R. T., B. Rao and D. E. Feldman.

(1964). The DNA synthetic period during early

development of the sea urchin egg. Exp. Cell Res.

36: 53-61.

Longo, F. J. and M. Kunkle. (1977). Synthesis of

RNA by male pronuclei of fertilized urchin eggs.

J. Exp. Zool. 201: 431-438.

Maxson, R. E. and J. C. Egrie. (1980). Expression

of maternal and paternal histone genes during

early cleavage stages of the echinoderm hybrid S.

purpuratus and L. pictus. Devel. Biol. 74: 335-342.

Wagenaar, E. B. and D. Mazia. (1978). The effect

of emetine on first cleavage division in the sea ur-

chin Stronglyocentrotus purpuratus. In: “‘ICN-

UCLA Symposium on Molecular and Cellular

Biology: Cell Reproduction XII” (E. R. Dirksen,

D. M. Prescott and C. F. Fox, eds.), Acad. Press,

New York, pp. 539-545.

Rodman, T. C., F. H. Pruslin, H. P. Hoffmann and

V. G. Allfrey. (1981). Turnover of basic chromo-

somal proteins in fertilized eggs—a cytoimmuno-

chemical study of events in vivo. J. Cell Biol. 90:

351-361.

Poccia, D., J. Erickson, L. Nash, T. Greenough and

G. R. Green. Unpublished results.

Newrock, R. M., C. R. Alfageme, R. V. Nardi and

L. H. Cohen. (1978). Histone changes during

chromatin remodeling in embryogenesis. Cold

Spr. Harb. Symp. Quant. Biol. 42: 421-431.

Weintraub, H., S. J. Flint, 1. M. Leffak, M. Grou-

dine and R. M. Grainger. (1978). The generation

and propagation of varigated chromosome struc-

tures. Cold Spr. Harb. Symp. Quant. Biol. 42:

401-407.

Journal of the Washington Academy of Sciences.

Volume 72, Number |, Pages 34-42. March 1982

50. Savié, A., P. Richman, P. Williamson and D. Poccia.

(1981). Alterations in chromatin structure during

early sea urchin embryogenesis. Proc. Natl. Acad.

Sci. USA 78: 3706-3710.

The Cell Cycle in Early Embryonic Development

Barbara Wolfanger Belisle

Department of Biology, Georgetown University, Washington, D.C. 20057

ABSTRACT

The cell cycle in early cleavage stages of sea urchin embryos is reviewed. The embryonic cell

cycle consists of S, G2 and M (there is no G;). Various aspects of mitotic apparatus formation,

chromosome movement and the cleavage furrow are discussed. Cell cycle regulation is also

reviewed, including factors involved in the control of chromosome condensation cycles and

also in terms of cell cycle timing.

The eukaryotic cell cycle represents all of

the events occurring in sequence from one

cellular division to the next. Cells repeti-

tively undergoing the series of events lead-

ing to division are termed cycling cells,

while those cells involved in other aspects

of development, growth or maintenance

(for example, terminally differentiated cells)

are considered to be in a non-cycling state.

The mature, unfertilized sea urchin egg is

one cell which 1s in a resting or non-cycling

state. However, when the egg is fertilized

and becomes metabolically activated, the

first cell cycle is initiated, unleashing in the

egg an awesome potential for division that

can result in the production of as many asa

thousand cells in 8 hours.’ Add to this di-

vision potential the attraction of being able

to fertilize an entire population of cells in

such a way that all of the cells within the

population divide synchronously, and the

sea urchin egg becomes a useful model sys-

tem for cell cycle studies. The purpose of

this discussion is to review some of the

events of the sea urchin cell cycle, and some

of the proposed mechanisms by which the

tempo of the cell cycle may be regulated.

Cell cycle models derived from a study of

this system must be extrapolated to other

cellular systems with care. The egg is, after

all, a developmental system, with a re-

quirement for large numbers of divisions

over very short time spans, with little or no

interphase pause between successive cell

divisions.

The basic eukaryotic cell cycle is gener-

ally subdivided into four phases designated

as G; (Gap 1), S, G2 (Gap 2) and M. In this

scheme (Figure 1), interphase (encompass-

ing G,, S and G)) is actually a dynamic

state of the growing, metabolically active

cell. The G; phase marks the interval oc-

curring between the end of mitosis (M) and

the onset of the DNA synthetic phase (S). It

CELL CYCLES IN EMBRYONIC DEVELOPMENT 35

Fig. 1. The normal sequence of the eukaryotic cell

cycle. Interphase includes Gap 1 (Gi), DNA synthesis

(S) and Gap 2 (G2). Mitosis (M) includes normal mi-

totic chromosome movements designated as prophase

(pro), metaphase (meta), anaphase (ana) and telo-

phase (telo).

is during G; that synthesis and mobili-

zation of the various elements required for

DNA synthesis occurs.* The S phase rep-

resents the period when DNA replication

occurs,’ while G2 isa relatively nonvariable

interval occurring from S to mitotic onset.

Mitosis consists of the typical chromo-

somal stages ensuing as the cell progresses

in an orderly manner from prophase

through telophase.” Also associated with

the cycle phases on a temporal basis are

protein synthesis, beginning in telophase

and lasting through prophase, and RNA

synthesis, occurring in late telophase or

early interphase (reviewed in’).

In early sea urchin cleavage stages the

typical cell cycle phases are not distinct.

Hinegardner et a/.° monitored the cell cycle

in fertilized eggs of the purple sea urchin,

Strongylocentrotus purpuratus, using tritium

(H) labelled thymidine experiments. The

*H-thymidine is a specific precursor of

chromosomal DNA’ and its incorporation

into newly synthesized DNA can be moni-

tored by typical radioactive assay methods.

Using this approach, the first period of

DNA synthesis (S;) is shown to occur at

approximately 30 minutes post-fertilization,

about the time of pronuclear fusion. Syn-

thetic periods after this initial S phase begin

in telophase and end as ‘interphase’ begins.

The overlap of S with M indicates there is

an absence of G;.°” Interphase is very short

in these early cleavage divisions (20 min-

utes in S. purpuratus at 15°C”) and not well

defined.’ A recent study by Dan et al.’ has

further clarified the cell cycle of early

cleavage stages in fertilized sea urchin eggs.

Using ~H-thymidine labelling of eggs from

the Japanese sea urchin, Hemicentrotus

pulcherrimus, they verified the cell cycle

consists of S, M and a questionable G,

phase. In this species also, S overlaps with

the preceding M, beginning when the chromo-

somes are still in anaphase and continuing

through telophase (Figure 2).

The M phase is the best characterized

portion of the sea urchin cell cycle, at least

from an ultrastructural and morphological

standpoint. During this time chromosomes

Fig. 2. Thesea urchin cell cycle during early cleav-

age stages. Interphase (inter) is Gap 2 (G2). Mitosis

(M) consists of prophase (pro), metaphase (meta), an-

aphase (ana) and telophase (telo). The DNA synthetic

phase (s) overlaps with M, beginning in late ana or

early telo. There is no G).

36 BARBARA WOLFANGER BELISLE

become visibly condensed and associated

with other mitotic structures, such as the

spindle and mitotic asters. The differentia-

tion of the mitotic apparatus (the chromo-

somes, spindle and mitotic asters) of the

first cell cycle has been described in

detailer?”

From the time of fertilization and for a

period prior to the onset of first prophase,

the egg remains in interphase. Following

pronuclear fusion, the paternal chromatin

at first remains visible as a ball of chro-

matin, but eventually disappears as it dif-

fuses and mixes with the maternal chro-

matin.* Also prior to the onset of prophase,

some cytoplasmic elements (composed of

annulate lamelli, ribosomes, smooth endo-

plasmic reticulum and microtubules””’)

concentrate adjacent to the zygote nucleus

resulting in the appearance of a ‘streak’ on

the surface of the egg.’ The chromatin

becomes visibly condensed as the cell be-

gins early prophase. The chromosomes

continue to contract and condense during

prophase’ and the nuclear envelope be-

gins to vesiculate and breakdown (pro-

metaphase”). At metaphase the chromo-

somes appear small and have become equato-

rially arranged. Many microtubules are

observed among the chromosomes.’ The

sister chromatids then begin separating

toward opposite poles as anaphase is in-

itiated.”'* There is no distinct transition

from anaphase to telophase.” As chromo-

somes begin to enlarge and disperse, vesi-

cles form about decondensing regions of

chromatin, eventually fusing to form a new

interphase nucleus.”

Temporally paralleling these events is

formation of the sea urchin mitotic appara-

tus—the cellular structure postulated to

function in providing the motive force by

which chromosomes move during anaphase.

This mitotic apparatus is represented by

the mitotic spindle, composed of micro-

tubules, vesicles and possibly actin (re-

viewed by McIntosh et al.'*) and its

associated asters (microtubules, smooth en-

doplasmic reticulum*"*). Formation of the

aster and spindle has been studied by

both ultrastructural and immunofluores-

cent means”’”’*’°). The developing sperm

aster can be seen as early as 13 minutes

after fertilization.” During male pronu-

clear migration, microtubules and smooth

endoplasmic reticulum begin to surround

the sperm centrioles, forming a dense ma-

trix termed the centrosphere. As the sperm

pronucleus continues to differentiate, the

asters increase in size as microtubules ra-

diate outward from the centrioles.*® Shortly

before pronuclear fusion, a radially ar-

ranged system of microtubules is apparent

about the egg nucleus, ° and about the time

of pronuclear fusion this system contrib-

utes, with the sperm aster, to form the egg

monaster.’° This radial arrangement of fi-

bers is gradually lost as the fibers appear to

move to a position underlying and parallel

to the egg surface, forming a cortical spiral

array.'”'° This cortical array persists from

just prior to pronuclear fusion until the

‘streak’ stage. '° During this time, the grow-

ing interphase asters become visible, lo-

cated at either pole of the nucleus,’ and ap-

pearing to increase in size independently of

the cortical fiber array. ° Subsequent to the

streak stage, the interphase asters loose mi-

crotubules in a progressive manner from

the center. A short while later, mitotic aster

formation begins as microtubules accumu-

late and grow from the center outward.”

These mitotic asters will continue to grow

and eventually move to opposite poles of

the egg as spindle development occurs. The

spindle is also composed of microtubular

elements associated with a variety of pro-

teins, endoplasmic reticulum, enzymes, and

vesicles,'*"’"' and provides the pathway

along which chromosomes migrate during

their anaphase movements.

One intriguing aspect of mitotic nuclear

division is the motive force underlying

chromosome movement during anaphase.

Most models, derived from studies using

either lysed cells or the isolated mitotic ap-

paratus, identify spindle microtubules as

the force-generating structures.” '?*' Actin

CELL CYCLES IN EMBRYONIC DEVELOPMENT 37

has also been postulated to function in the

control of chromosome movement (see re-

view by Forer’’) but the involvement of

actin in the spindle is not well defined. It is

now generally accepted that chromosome

movement is regulated at least in part by

the controlled assembly and dissassembly

of microtubule subunits.”””’

Upon the completion of M phase the cell

is ready to divide, and this is accomplished

by the circumferential constriction of the

cell to form two new daughter cells (cyto-

kinesis). This constriction is referred to as

the cleavage furrow. Cytokinesis usually

occurs at telophase, although cleavage fur-

row formation begins in anaphase.”** The

furrow consists of microfilaments which

form a contractile ring around the cell

margin halfway between the spindle

poles.”***° Cytokinesis is thought to result

from the active contraction of cortical sub-

stances localized in the furrow.”’ Since the

cleavage plane usually coincides with the

plane of the metaphase plate, it has been

suggested that aster microtubules are in-

volved in stimulating the cleavage furrow.

However, Asnes and Schroeder” have dem-

onstrated that aster microtubules do not

penetrate the equatorial cortex soon enough

or in great enough numbers to stimulate

the presumptive furrow to constrict, and

the mechanism of constriction remains rel-

atively poorly understood.

Most of the cell cycle events discussed

are typical of the first five cleavage stages in

sea urchin embryos. It is during the early

Stages (divisions 1-4) that equal division

occurs, producing uniformly sized blasto-

mers.’ It is also during this portion of em-

bryonic development that a high degree of

division synchrony among blastomeres is

apparent.’ At the fourth cell division,

cleavage becomes unequal and embryos

containing three cell types separable based

on size (micromeres, mesomeres and mac-

romeres) are formed.'' A study of division

synchrony in cleavages after the fifth sug-

gests an overall increase in cycle duration,

and increasing asynchronies in cycle times

among the three main cell types (although

Synchrony appears to remain relatively

high within a particular cell type’). Asyn-

chronies among cell types are probably a

result of differences in the length of the G;

and Gp phases of the cycle, which become

incorporated into the cell cycle during

these later divisions.

While the sequence of morphological

and metabolic events occurring during early

cell divisions in sea urchin embryos has

been described extensively, the division

cycle has historically been less well studied

in terms of its regulation. Only in the past

few years has the sea urchin egg come into

vogue as a model for cell cycle control stud-

ies. Most studies using sea urchin eggs as a

model system have approached the prob-

lem of cycle regulation by asking two basic

questions: 1) What process triggers chromo-

some condensation; and 2) What factors

are involved in cell cycle timing?

Cell cycles in sea urchin eggs are usually

studied in eggs which have been activated

artificially. By treating the eggs with cer-

tain chemicals (particularly those of the

amine group) several of the activation

events normally initiated naturally by fer-

tilization are induced in the absence of any

sperm contact.~* *” One useful aspect of ar-

tificial activation is that eggs treated in this

manner can still be fertilized and will de-

velop into normal larval forms,”* although

care must be taken to use sperm concentra-

tions which reduce the risk of polyspermy

(the incorporation of more than one sperm

per egg). Because various cycle events are

initiated by the activators before sperm

contact, fertilization can be made to occur

at various points within the cell cycle

simply by varying the length of treatment

time before fertilization. By fertilizing the

eggs during different portions of the cell

cycle, and then determining the effect on

various cycle events (such as paternal chro-

mosome condensation, the length of the cell

cycle relative to control eggs, etc.) one can

address the question of phase dependencies

of various cleavage related events on spe-

38 BARBARA WOLFANGER BELISLE

cific cell cycle events. An alternative ap-

proach to the study of the interdependen-

cies of various cycle events is to study

division timing after the addition of spe-

cific metabolic inhibitors which affect one

or more cycle related events.

1. The Control of Chromosome

Condensation Cycles

The control of chromosome condensa-

tion in sea urchin eggs can be studied by cell

fusion involving either egg-sperm interac-

tions***° or by egg-egg hybridization tech-

niques.*° In both approaches, the results

are the same. Fusion of a cell in interphase

with a cell in mitosis results in stimulation

of chromosome condensation in the inter-

phase cells.

Treatment of unfertilized eggs with am-

monia stimulates *H-thymidine incorpo-

ration (DNA synthesis), and eggs treated

for extended periods of time display cycles

of *H-thymidine incorporation represent-

ative of repetitive S phases of the cell

cycle.*? Even though successive chromo-

some cycles occur, there is no spindle for-

mation, and normal anaphase movements

do not occur.*’ When these eggs are treated

for approximately 60 minutes, and then

fertilized, both the maternal and paternal

chromatin 1s visibly condensed by 90 min-

utes post-activation.** The maternal chro-

mosomes are condensing at the expected

time (90 minutes post-activation), but the

paternal chromatin is condensing an hour

prematurely (30 minutes post-insemina-

tion).***°?* Here fusion of sperm with an

egg already into the condensation cycle in-

duces premature condensation (PCC) of

the paternal chromatin. Poccia et al.** ana-

lyzed this response of the male chromatin

in an activated, cycling egg further in an at-

tempt to determine whether cytoplasmic

factors might be involved in controlling

PCC. When whole, activated eggs are fertil-

ized after maternal chromosomes have en-

tered prophase, the paternal chromatin

condensation cycle conforms to that of the

maternal pronuclear cycle. If unfertilized

eggs are broken into nucleated and non-

nucleated egg halves (a technique performed

by centrifugation of the eggs through su-

crose gradients'’*’) and treated with am-

monia, both eggs activate,” and chromo-

some cycles are visible in the nucleated

halves. If the halves are fertilized when the

nucleated halves are in prophase, both

halves show PCC at the same time, but

non-nucleated halves do not give as strong

a response, and the lifetime of PCC-pro-

moting activity is less than in nucleated

halves.*°** It seems from these results that

cytoplasmic components effect PCC-pro-

moting activity, but these effects are modu-

lated by the maternal pronucleus.*

Vacquier and Brandriff*° studied the ef-

fects of the activator, procaine hydrochlo-

ride, on chromosome cycles and on S. Pro-

caine, like NH4Cl, stimulates late phase

fertilization events, and also induces cy-

taster formation,’' the assembly and dis-

assembly of which follows the procaine in-

duced chromosome cycle.*” Vacquier and

Brandriff show that DNA synthesis can be

turned on and off by the addition and re-

moval of procaine. Also, when procaine is

removed by washing the treated eggs with

fresh sea water after chromosome conden-

sation has begun, chromosomes continue

to condense, but do not decondense.”’

These workers suggest it is chromosome

decondensation which is controlling en-

trance into S.

Removal of NH4OH does not appear to

turn off chromosome cycles once they have

been initiated.'* However, removal of

NH.,Cl gives procaine-like results. If NH4Cl

is removed after chromosome condensa-

tion has started, condensation will pro-

ceed, but decondensation does not occur.”

But, if chromosomes have started to de-

condense when the NH,Cl is removed, they

will continue to decondense (and may even

initiate a second cycle*’).

Preliminary experiments suggest that

CELL CYCLES IN EMBRYONIC DEVELOPMENT 39

chromosome cycles can be initiated by

activating eggs with NH,Cl in sodium-

calcium free sea water,’ ionic conditions

which are supposed to prevent any intra-

cellular increases in calcium concentration

at activation.** Also, treating sea urchin

eggs simultaneously with NH4Cl and pro-

tein synthesis inhibitors in either normal

sea water or sodium-calcium free sea water,

seems to allow chromosome condensation,

but prevents chromosome decondensation.™

These activator removal experiments and

inhibitor experiments suggest that chro-

mosomes cycles in sea urchins may be bi-

phasic, with each phase (condensation and

decondensation) independently regulated

by separate metabolic events.” It has pre-

viously been shown that eggs fertilized in

the presence of protein synthesis inhibitors

do not cycle and work by Wagenaar and

Mazia“° suggests that cytoplasmic factors

regulating chromosome cycles are in fact

proteins. They were able to demonstrate

that a specific sequence of protein synthesis

is required for complete and normal pro-

gression of the chromosome cycle and that

proteins produced at different times in the

cell cycle are required for chromosome

condensation, nuclear envelope breakdown,

and formation of a normal nuclear mitotic

apparatus.”°

2. Control of Cell Cycle Timing

Sea urchin eggs which have been acti-

vated with monogen or duponal, and then

fertilized, cleave earlier than untreated, fer-

tilized controls. This decrease in the time to

first cleavage is dependent on the length of

activator treatment.*”** Sea urchin eggs ac-

tivated with NH,CI or procaine show this

same time dependency (cleavage advance)

relative to control eggs.*”** If eggs are fer-

tilized before maternal pronuclear prophase

has begun, there is no cleavage advance.

However, as eggs begin to initiate chromo-

some condensation, a short advance oc-

curs, and increases to a maximum at atime

when 100% of the eggs contain maximally

condensed chromosomes. Cleavage advance

remains constant as chromosomes enter

the decondensation phases of the cycle, but

decreases again at interphase and early

prophase stages of the next cycle. This pat-

tern is repeated with successive cycles, and

suggests that cleavage advance is depend-

ent on the degree of chromosome conden-

sation, but not decondensation (deconden-

sation is not a factor in cycle timing) and

the capacity to cleave in advance is reset

with each successive cycle.*”

It is not likely that the timing of cellular

progression is directly controlled by chro-

mosome condensation,” and it is also un-

likely that cell division advance is due to

the physical process of chromosome con-

densation.”” Sluder® has shown that chro-

mosome cycles in sea urchins will proceed

to completion, even in eggs which have

been exposed to colcemid (a drug which

prevents microtubule assembly, including

assembly of spindle microtubules). Eggs

fertilized after treatment with colcemid

show complete chromosome cycles, but do

not form spindles and do not divide. The

cell cycle itself is slowed by colcemid treat-

ment, and this delay in the cycle appears to

reflect a prolongation of prometaphase.”

By irradiating colcemid treated eggs, the

colcemid is inactivated*’ and spindle as-

sembly occurs, but delayed relative to nu-

clear envelope breakdown (depending on

when after nuclear envelope breakdown ir-

radiation occurs). These delays in spindle

assembly are reflected as delays in the initi-

ation of anaphase and in the time of cell

division. Based on these results, Sluder

suggests that microtubule assembly cycles

‘set’ the tempo of the cell cycle, forming

part of the trigger that turns on the events

critical to completion of the cycle subse-

quent to nuclear envelope breakdown.”

It is true that microtubules assemble and

dissassemble, and that the pattern of as-

sembly can be correlated with the cell

cycle.°' And it is generally accepted that

40 BARBARA WOLFANGER BELISLE

microtubules are universally associated

with chromosomes during cell division. As

mentioned previously, the sea urchin aster

is primarily composed of microtubules and

smooth endoplasmic reticulum. While mi-

crotubules themselves comprise only a small

portion of the spindle apparatus (the mi-

totic gel is actually made up of membrane

bound vesicles within which the micro-

tubules are embedded), a close relationship

is maintained between these two structures

throughout the mitotic cycle.'* Micro-

tubule assembly is influenced by calcium,

and vesicles with calcium ion sequestering

activity have been identified in the isolated

sea urchin mitotic apparatus. * In addition,

a calcium activated AT Pase has been impli-

cated in the mitotic cycle. Mazia et al.°”

identified this enzyme in mitotic apparatus

isolated from echinoderm eggs, and Pet-

zelt’? showed initiation of a calcium-

ATPase activity cycle in fertilized sea urchin

eggs which can be correlated with the cell

cycle. Enzyme activity peaks near the start

of chromosome condensation and de-

creases when nuclear membranes are re-

formed.’ Harris” postulates that mitosis is

controlled by the temporal and spatial ef-

fects of calcium on microtubule assembly

and dissassembly. This theory suggests that

unknown factors (i.e., pH, critical tubulin

levels, cAMP, etc.) stimulate a self-pro-

pogated trigger wave of calcium release

which initiates at the aster center and ra-

diates outward, causing in its path a change

in the state of tubulin polymerization.

While attributing changes in tubulin poly-

merization to changes in calcium ion con-

centration, it is not clear from this model

which cytoplasmic microtubule elements

are forming mitotic microtubules or what

specific event is triggering the initial proc-

ess (i.e., what is resetting the timer).

It does not appear that a ‘central timer’ is

executing a short lived state which allows

cleavage to occur. When chromosome con-

densation is uncoupled from chromosome

decondensation by activator removal cleav-

age is not affected. As long as chromo-

somes remain condensed maximal cleavage

advance will occur,’ suggesting some per-

missive signal to cleavage may be turned on

indefinitely in these eggs. Chromosome de-

condensation, then, may be a reset signal

required for cell cycle continuity.*° A sur-

vey of results obtained from studies involv-

ing the manipulation of the cell cycle would

suggest that a model analogous to Hart-

well’s dependent pathway model (described

for yeast) may bean appropriate model for

the sea urchin egg also.”° Here two or more

‘loops’ (for example a chromosome loop

and a metabolic loop) are running in paral-

lel circuit, and close together at the appro-

priate time to initiate cleavage.

Acknowledgments

The author is supported by a grant from

the National Science Foundation (No.

PCM 7923487 to D. Nishioka).

The author wishes to thank D. Nishioka

for critical review of the manuscript.

References

1. Stearns, L. W. (1974). Sea Urchin Development:

Cellular and Molecular Aspects. Dowden, Hut-

chinson and Ross, Inc., Penn., pp. 339.

2. Wilson, G. B. (1966). Cell Division and the Mitotic

Cycle. Reinhold Publishing Corporation, New

York, pp. 98.

3. Baserga, R. (1969). The Biochemistry of Cell Di-

vision. Charles C. Thomas, Springfield, IIl., pp.

214.

4. Mitchison, M. (1971). The Biology of the Cell

Cycle. Cambridge University Press, New York,

pp. 313.

5. Hinegardner, R. T., B. Rao and D. E. Feldman.

(1964). The DNA synthetic period during early

development of the sea urchin egg. Exp. Cell. Res.

36: 53-61.

6. Brachet, J. (1974). Introduction to Molecular Em-

bryology. Springer-Verlag, New York, pp. 176.

7. Dan, K, S. Tanaka, K. Yamazaki and Y. Kato.

(1980). Cell cycle study up to the time of hatching

in the embryos of the sea urchin, Hemicentrotus

pulcherrimus. Develop, Growth and Differ. 22:

589-598.

8. Longo, F. J. and E. Anderson. (1968). The fine

structure of pronuclear development and fusion

in the sea urchin, Arbacia punctulata. J. Cell Biol.

395,339:

10.

ll.

12.

13.

14.

16.

17.

19.

20.

21.

22.

23.

CELL CYCLES IN EMBRYONIC DEVELOPMENT 41

. Longo, F. J. (1972). An ultrastructural analysis of

mitosis and cytokinesis in the zygote of the sea ur-

chin, Arbacia punctulata. J. of Morphol. 138:

207-238.

Harris, P., M. Osborn and K. Weber. (1980). Dis-

tribution of tubulin containing structures in the

egg of the sea urchin Strongylocentrotus purpura-

tus from fertilization through first cleavage. Exp.

Cell. Res. 84: 668-679.

Harvey, E. B. (1956). The American Arbacia and

Other Sea Urchins. Princeton University Press,

Princeton.

Mazia, D. (1974). Chromosome cycles turned on

in unfertilized sea urchin eggs exposed to NH,OH.

Proc. Natl. Acad. Sci., USA, 71: 690-693.

McIntosh, J. R., W. Z. Cande, E. Lazarides, K.

McDonald and J. A. Snyder. (1976). Fibrous ele-

ments of the mitotic spindle. In: Ce// Motility, Vol.

3, R. Goldman, T. Pollard and J. Rosembaum,

eds., Cold Spring Harbor Laboratories, USA, pp.

1261-1272.

Harris, P. (1975). The role of membranes in the

organization of the mitotic apparatus. Exptl. Cell

Res., 94: 409-425.

Mar, H. (1980). Radial cortical fibers and pro-

nuclear migration in fertilized and artificially ac-

tivated eggs of Lytechinus pictus. Devel. Biol., 78:

1-13.

Harris, P., M. Osborn and K. Weber. (1980). A spi-

ral array of microtubules in the fertilized sea ur-

chin egg cortex examined by indirect immuno-

fluorescent electron microscopy. Exptl. Cell Res.,

126: 227-236.

Hepler, P. (1972). In: Mechanisms and Control of

Cell Division. T. Rostand E. M. Gifford, Jr., eds.,

Springer-Verlag, New York, pp. 149.

. Silver, R. B., R. D. Cole and W. Z. Cande. (1980).

Isolation of mitotic apparatus containing vesicles

with calcium sequestration activity. Cell 19:

505-516.

Forer, A. (1974). In: Cell Cycle Controls. G. M.

Padilla, I. L. Cameron, A. Zimmerman, eds.,

Academic Press, New York, pp. 319-336.

McIntosh, J. R. (1977). Mitosis in vitro: Isolates

and models of the mitotic apparatus. In: Mitosis:

Facts and Questions. M. Little, N. Paweletz, C.

Petzelt, H. Ponstingl, D. Schroeter and H.-P.

Zimmerman, eds., Springer-Verlag, New York,

pp. 167-184.

Inuoe, S. (1981). Cell division and the mitotic

spindle. J. Cell Biol., 91: 131s.

Forer, A. (1976). Actin filaments and birefringent

spindle fibers during chromosome movements.

In: Cell Motility, Vol. 3, R. Goldman, T. Pollard

and J. Rosenbaum, eds. Cold Spring Harbor Lab-

oratories, USA, pp. 1273-1293.

Inuoe, S. (1976). Chromosome movement by re-

versible assembly of microtubules. In: Ce// Motil-

ity, Vol. 3, R. Goldman, T. Pollard and J. Rosen-

24.

20%

26.

27.

28.

29.

30.

Ski

32.

33.

34.

S35

36.

a7.

38.

39.

baum, eds., Cold Spring Harbor Laboratories,

USA, pp. 1317-1328.

Rappaport, R. (1975). Establishment and organi-

zation of the cleavage mechanism. In: Molecules

and Cell Movement. S. Inoué and R. E. Stephens,

eds., Raven Press, New York, pp. 287-302.

Schroeder, T. E. (1968). Cytokinesis: filaments in

the cleavage furrow. Exptl. Cell Res. 53: 272-318.

Asnes, C. and T. E. Schroeder. (1979). Cell cleav-

age. Ultrastructural evidence against equatorial

stimulation by aster microtubules. Exptl. Cell

Res., 122: 327-338.

Schroeder, T. E. (1976). Actin in dividing cells:

Evidence for its role in cleavage but not in mitosis.

In: Cell Motility, Vol. 1. R. Goldman, T. Pollard

and J. Rosenbaum, eds., pp. 265-278.

Harvey, E. N. (1909). Methods of artificial par-

thenogenesis. Biol. Bull., 18: 269-280.

Steinhardt, R. A. and D. Epel. (1974). Activation

of sea urchin eggs by a calcium ionophore. Proc.

Natl. Acad. Sci., US, 71: 1915-1919.

Epel, D., R. Steinhardt, T. Humphreys and D.

Mazia. (1974). An analysis of the partial meta-

bolic derepression of sea urchin eggs by ammonia:

The existence of independent pathways. Devel.

Biol., 40: 245-255.

Steinhardt, R. A. and D. Mazia. (1973). Develop-

ment of a K* conductance and membrane poten-

tials in unfertilized sea urchin eggs after exposure

to NH,OH. Nature, 241: 400-401.

Mazia, D. and A. Ruby. (1974). DNA synthesis

turned on in unfertilized eggs by treatment with

NH,OH. Exptl. Cell Res., 85: 167-172.

Brandriff, B, R. T. Hinegardner and R. Steinhardt.

(1975). Development and life cycle of the parthe-

nogenetically activated sea urchin embryo. J. of

Exptl. Zool., 192: 13-24.

Mazia, D. (1974). The Cell Cycle. Sci. Amer., 230:

54-64.

Krystal, W. and D. Poccia. (1979). Control of

chromosome condensation in the sea urchin egg.

Exptl. Cell Res., 123: 207-219.

Bennett, J. and D. Mazia. (1981). Interspecific fu-

sion of sea urchin eggs. Surface events and cyto-

plasmic mixing. Exptl. Cell Res., 131: 197-207.

Paweletz, N. and D. Mazia. (1978). The “‘Nuclear

Mitotic Apparatus” in sea urchin eggs. In: Cell

Reproduction. E. R. Dirkson, D. M. Prescott and

C. F. Fox, eds., Academic Press, Inc., pp. 495-504.

Poccia, D., G. Krystal, D. Nishioka and J. Salik.

(1978). Control of sperm chromatin structure by

egg cytoplasm in the sea urchin. In: Cell Repro-

duction. E. R. Dirkson, D. M. Prescott and C. F.

Fox, eds., Academic Press, New York, pp. 197-

206.

Nishioka, D. and D. Mazia. (1977). The phospho-

rylation of thymidine and the synthesis of his-

tones in ammonia-treated eggs and egg fragments

of the sea urchin. Cell Biol. Int. Rep. 1: 23-30.

40.

41.

42.

43.

44.

45.

46.

47.

BARBARA WOLFANGER BELISLE

Epel, D. (1975). The program and mechanism of

fertilization in the Echinoderm egg. Amer. Zool.,

15: 507-522.

Vacquier, V. D. and B. Brandriff. (1975). DNA

synthesis in unfertilized sea urchin eggs can be

turned on and turned off by the addition and re-

moval of procaine hydrochloride. Devel. Biol.,

47: 12-31.

Moy, G. W., B. Brandriff and V. D. Vacquier.

(1977). Cytasters from sea urchin eggs parthe-

nogenetically activated by procaine hydrochlo-

ride. J. Cell Biol., 73: 788-793.

Belisle, B. W. (1981). Mitotic triggers in the cell

cycle: phase dependency of cleavage advance on

the chromosome cycle in parthenogenically acti-

vated sea urchin eggs. Ph.D. Thesis, Louisiana

State University, Baton Rouge, La.

Belisle, B. W., W. Byrd and J. Wille. (1981).

Cleavage advance and chromosome cycles in

parthenogenically activated sea urchin eggs. J.

Cell Biol., 91: Ila.

Winkler, M., R. A. Steinhardt, J. L. Grainger and

L. Minning. (1980). Dual ionic controls for the ac-

tivation of protein synthesis at fertilization. Na-

ture, 287: 558-560.

Wagenaar, E. B. and D. Mazia. (1978). The effect

of emetine on first cleavage division in the sea ur-

chin, Strongylocentrotus purpuratus. In: Cell Re-

production. E. R. Dirkson, D. M. Prescott and C.

F. Fox, eds., Academic Press, New York, pp.

539-545.

Kojima, M. K. (1967). Acceleration of the cleavage

48.

49.

50.

Sf.

52:

33:

54.

55.

in sea urchin eggs by insufficient treatments with

activating reagents before fertilization. Embry-

ologia, 10: 75-82.

Kojima, M. K. (1969). Acceleration of the cleavage

in sea urchin eggs by repeated insufficient stimu-

lations with various activating reagents and sper-

matozoa. Embryologia, 10: 323-333.

Sluder, G. (1979). Role of spindle microtubules in

the control of cell cycle timing. J. Cell Biol., 80:

674-691.

Belisle, B. W., J. Wille and W. Byrd. (1982). Phase

coupling of cleavage advance to the chromosome

cycle following fertilization of parthenogeneti-

cally activated sea urchin eggs. (Manuscript

submitted).

Lloyd, C. (1979). Primitive model for cell cycle

control. Nature, 280: 631-632.

Mazia, D., Chr. Petzelt, R. O. Williams and I.

Meza. (1972). A Ca*?-activated ATPase in the mi-

totic apparatus of the sea urchin (isolated by a

new method). Exptl. Cell Res., 70: 325-332.

Petzelt, C. H. (1972). Ca*? activated ATPase dur-

ing the cell cycle of the sea urchin Strongylocen-

trotus purpuratus. Exptl. Cell Res., 70: 332-339.

Harris, P. (1978). Triggers, trigger waves, and mi-

tosis: A new Model. In: Cell Cycle Regulation. J.

R. Jeter, I. L. Cameron, G. M. Padilla and A. M.

Zimmerman, eds., Academic Press, Inc., New

York.

Hartwell, L. H., J. Culotti, J. R. Pringle and B. J.

Reid. (1974). Genetic control of the cell division

cycle in yeast. Science, 183: 46-51.

DELEGATES TO THE WASHINGTON ACADEMY OF SCIENCES,

REPRESENTING THE LOCAL AFFILIATED SOCIETIES

I SS James F. Goff

Mr EMICIEEY OF WASHINGTON . oc. cece ccc cect cee wesc cere cc encccuccece Ruth H. Landman

SECT occ ge ue viva vines caeneceneneeveneveussuwees William Ronald Heyer

NS) RO Oe PINE SE ee oe a ne es Anne Brown

ETE AVASININGTON «oun sees cease vas eceseeeuncccecencuceens Margaret Collins

I aa al ecu Rats wid WIG Si Rstresineh@ Wh Wad whl a eee aS wee T. Dale Stewart

I SISRINYERON 5 oo wn ec cack ecw caw eile caeadusceveeusecdians James V. O’Connor

nnn tar ae Pistrict Of Columbia... . 0.0.0... 0ccccccccvccccscuceveces Charles E. Townsend

eee 20k Soke Di ci eeg.es 6 anh wee ha cM RE HOEY Veo AE OO hs Paul H. Oehser

RSET 2 02 oi 5 sions. inca anonn soe hype w Kin le alle CoA eUmChw ne sibel iv ep wens Conrad B. Link

REIS te oe. wn. x WH asda Ca eee Oe een Wels bw Sd Sed pee Oa eemad oe Boyd W. Post

EEE ESIICIEGES. coi cies cs ccc Se hae cc seca sea aranbule cesautaeeee nas George Abraham

een teenie) and Electronics Engineers ............0cccccccccscavsvcecease George Abraham

EME MEGERAMICAl ENSINCETS . . 6 5. oe cee eee ee ecb nest ecestoewccneaces Michael Chi

ene mseicty Gl WaSHiINftONn ... 2... 2.6. cee eee es ranacoewsens Robert S. Isenstein

EERO MICEORIOIOPY |. os 6. ce cc cnc ces ce ccceceeeeeadesteeuenes Lloyd G. Herman

a MGRCAM Military ENGINCErDs..... 2... 5. ce eee eee ewes ccc eusecesvecucs H. P. Demuth

Mire MtUN PNGIICETS 22... 2 a ce cee cee nese beaeenaenenaeuees Wallace J. Cohen

eer eeectimental Biology and Medicine................0cseeccceucuccscves Cyrus R. Creveling

LEIS 0 cS A ale Sa 8s Od aa esd oe oda ea cee Ween Rae eee Charles G. Interrante

ee eemmeeteenne! Pctiital Research .... 2.6.5.0 0 eee eet cane nee ee nens William R. Cotton

Pamerican institute of Acronautics and Astronautics.............0.00cceeceeesccees Richard P. Hallion

PEER EPEC E oe oie co go 5 oon ae a win seb a aide Be dp a's epdretorw wie wore awl A. James Wagner

NCGS TV ASHINGION. . 0c. 5. we ee Reece ence ceedaeneeeveneedean Jack R. Plimmer

EMT NE Be 6) 5 8 co hse cae Sena a bie caine eS Uae daa e a Me ww Rees eele we me Richard K. Cook

IE ES elo oc aa his Simca t big ahaials 6 oe 6 wx odiagobeweu peice ee dee Mewewere Dick Duffey

INE SEIMBTIOPIGES 5 06 5 oc Gi 5.0% a wo Seles io a VG oS he's vv ee Oe Se Re eaccwses A. D. Berneking

ER RN cee oS or Ss oh de aS Rg RD whe A wiong hod vixiaEHle Sa See AMOS Edwin R. Fuller, Jr.

I I Poa gay see few eae kent a A OS a we ds A RR Bee Dee A Alayne A. Adams

rnenemermreeseicnce CMD . 3. 5's. ec ce blied ae ices ena coe esaceecewseewe Deborah J. Warner

nnnmnmeanmrst PHYSICS LEACHCTS . 2... ccc cn des ccna dckeceeneeadeactess nas Peggy A. Dixon

ON NARA a wso antd a bm Sige kori Wind's 4 ehade bod es dia Mw el George J. Simonis

ee neren rial PHYSIOIOPISES ..... ¢ ioc ccc secre cece bee e be ce cddaeescte Walter Shropshire, Jr.

rE eas ESEATCH COPE! «2 < «ais xs sie 5 on 0s Se wdc dade nie du os wets eeee John G. Honig

Se ee Jewel B. Barlow

American Institute of Mining, Metallurgical

EE RAEN Goa a) ee oe ed ae ego SORE ageis # ake awe dea eee Garrett R. Hyde

EINER S52 ase ho yo. oie ek Hahn cose alee ee RCs 5 dees die cee Benson J. Simon

ciation of Amicricad «2... s<.<0< 20% «cc endvacs de deaevcndeeedscceusned® Patrick Hayes

NR isc Sh eps oo is in ada im arta 00's Aw loins Geel ee wide ale wows Miloslav Racheigl, Jr.

ER Se AI RARTAYS 8s Vn wie <x kw ese 2 oidsiase av esewicnwawcedae ve H.N. Reynolds

nes © Att PeEchmical GTOUD ..- 02. dic cee esse ce cccceoscactcusssccuvecs Paul G. Campbell

INRA NICSE AE SCNICEN «55 oe ocx nn ces bonis wiv ews oie c emcee dae ve wenden Howard E. Waterworth

Pn SeCHCE Al SVStCIIS RESCAFEN 26s ccc cer cc ccc cecssesvecvcencctonve Ronald W. Manderscheid

PDR el EE RIE ent I iS ag sc hiice o Wid wwe vl ee he ele a vbw Stanley Deutsch

OME Ra RS ie ar gsnitcse’ so Sl d o's ¢ iu 6c, nieie Bese didi sl cm aeintals Irwin M. Alperin

meeetntion for science, lechnology and Innovation ...........0.s0ccescnensesecessunns Ralph I. Cole

Delegates continue in office until new selections are made by the representative societies.

2nd Class _

Washington Academy of Sciences

1101 N. Highland St. 4 at Arl ]

Arlington, Va. 22201 ~ ~and additional mail

Return Requested with Form 3579

11 VOLUME 72

W317 ’ Number 2

¢ NH , of the June, 1982

WaoHINGTON

ACADEMY... SCIENCES

ISSN 0043-0439

Issued Quarterly

at Washington, D.C.

CONTENTS

Articles:

BARRY BUNOW: All Things Flow and Change—Some Thoughts on the

fen or Psmision and Reaction in Biology sis... 6.46.6 c.s00 sc cels oe ces 43

L. A. McNICOL, T. BARKAY, M. J. VOLL and R. R. COLWELL:

Plasmid Carriage in Vibrionaceae and Other Bacteria Isolated from the

Aquatic Environment

see ee eee eee e eee eee eee eee eee eee eee eee er ee eeeoee

ALAN J. PUE: Operational Concepts for Automated Transportation

Systems

Washington Academy of Sciences

Founded in 1898

EXECUTIVE COMMITTEE

President

John G. Honig

President-Elect

James F. Goff

Secretary

Jean K. Boek

Treasurer

John J. O'Hare

Members at Large

Conrad B. Link

Elaine Shafrin

John J. O’ Hare

Michael J. Pelezar, Jr.

Jo-Anne Jackson

Grover C. Sherlin

Lavern S. Birks

BOARD OF MANAGERS

All delegates of affiliated

Societies (see Cover 3)

EDITORS

Irving Gray

Joseph Neale

ACADEMY OFFICE

1101 N. Highland St.

Arlington, Va. 22201

Telephone: (703) 527-4802

The Journal

This journal, the official organ of the Washington

Academy of Sciences, publishes historical articles, critical —

reviews, and scholarly scientific articles; proceedings of |

meetings of the Academy and its Board of Managers; and

other items of interest to Academy members. The Journal .

appears four times a year (March, June, September, and

December)—the September issue contains a directory of

the Academy membership.

Subscription Rates

Members, fellows, and patrons in good standing receive ~

the Journal without charge. Subscriptions are available on _

a calendar year basis only, payable in advance. Payment

must be made in U.S. currency at the following rates:

U.S. and Canada...... $19.00

POreiEne. 5 x00 bs cee 22.00

Single Copy Price tas... 7.50

UHURU Symposium—2 issues............

Delivery by Air ....20- 2.00

Single-copy price for Vol. 66, No. 1 (March, 1976) is $7.50.

Back Issues

Obtainable from the Academy office (address at bottom

of opposite column): Proceedings: Vols. 1-13 (1898-1910)

Index: To Vols. 1-13 of the Proceedings and Vols. 1-40 of

the Journal Journal: Back issues, volumes, and sets (Vols.

1-62, 1911-1972) and all current issues.

Claims for Missing Numbers

Claims will not be allowed if received more than 60 days

after date of mailing plus time normally required for j

postal delivery and claim. No claims will be allowed —

because of failure to notify the Academy of a change in

address.

Change of Address

Address changes should be sent promptly to the Academy

office. Such notification should show both old and new

addresses and zip number.

Published quarterly in March, June, September, and December of each year by the Washington

Academy of Sciences, 1101 N. Highland St., Arlington, Va. 22201. Second class postage paid at

Arlington, Va. and additional mailing offices.

Journal of the Washington Academy of Sciences,

Volume 72, Number 2, Pages 43-60, June 1982

All Things Flow and Change—Some Thoughts on

the Role of Diffusion and Reaction in Biology

Barry Bunow

Laboratory of Applied Studies, Division of Computer Research and

Technology, National Institutes of Health, Bethesda, Maryland 20205

ABSTRACT

Movement of chemical substances and their interconversion are two of the most important

processes occurring in living organisms. Interaction between these two processes leads to a

number of phenomena whose existance was unanticipated and which closely resemble many

fundamental operations occurring within living organisms, the mechanisms of which are cur-

rently obscure. It is tempting to suppose that reaction/diffusion might provide a particularly

simple explanation for a number of such processes, although no proof is provided.

Introduction

The title of this article comes from philos-

ophy, and dates back to the Golden Age of

Athens, almost twenty five hundred years

ago. Heraclitus was probably thinking more

about the creations of men than about the

natural substances with which we shall be

concerned here. Nevertheless, all things do

flow and change, and nowhere is this fact

more important than in the organization of

living matter. The essence of life is its cease-

less activity: nutrients are taken in, broken

down, put in new places, transformed to

new molecular configurations, and, ulti-

mately, re-expelled into the external envi-

ronment as waste products.

The specific mechanisms by which sub-

Stances are moved within living matter de-

pend upon the distance over which motion

is required. The many slips betwixt the cup

and the lip are performed by a mechanical

linkage. From the lip through the intestine,

motion is acombination of mechanical and

hydraulic. Substances move from the intes-

tine to the vascular system, over a distance

measured in tens of microns, ona molecule

at a time basis: by diffusion. In the vascular

system, motion is clearly hydraulic. Sub-

stances carried in the blood must then pass

into the cells which make up each kind of

tissue. Cells are typically of the order of ten

Or sO microns in dimension, and diffusion

i.e. transport of molecules one by one, is

the principal mechanism of motion at this

size scale. The means by which molecules

actually move is outside our topic. We take

the following phenomenological definition:

A substance is said to move by diffusion

when the rate at which it moves is propor-

tional to the gradient of concentration.

The cell is also the level where chemical

transformations typical of the life process

occur. Chemical reactions can be thought

of as flows taking substances from one

form to another rather than from one place

44 BARRY BUNOW

to another. The rate at which a chemical

reaction flows is an empirically determined

function of the concentrations of the sub-

stances which participate in the reaction.

Thus chemical reaction and diffusion are

two of the principal physical-chemical

phenomena which are involved in the op-

eration of living things at the size scale of

about ten microns.

In this article we shall explore some of

the effects which arise when the processes

of chemical reaction and diffusion simul-

taneously influence the concentration of

substances. The goal will be to attempt to

explain in a reductionist way some of the

behavior of living systems which might

otherwise appear to depend on some

unique aspect of whatever it really means

to be “alive.”

A Very Simple Model System

In Figure 1 is shown a simple, non-living

system, consisting of a bag, immersed in

and filled with an aqueous medium. The

medium contains one or more chemical

substances, we’ll call them A, B, etc., which

can pass through the walls of the bag by dif-

fusion. The rate of flow is proportional to

the concentration difference between the

two sides of the bag, Eq. (1).

Ja = Pa(Ao — A) (1)

co EOE.

Fig. 1. A highly oversimplified model of a cell,

consisting of a permeable bag containing an enzyme

in solution. The bag is immersed in a large stirred res-

ervoir, and its contents are also well-stirred.

where P, is called the permeability of the

bag, Aj is the external concentration, and

A is the internal concentration. We shall

suppose that the inside of the bag is well-

stirred so that the concentrations of any

substance inside is the same at all points.

Inside the bag are one or more enzymes.

Enzymes act as catalysts for specific reac-

tions, and are of biological origin. We sup-

pose that the volume outside the bag is suf-

ficiently large that movement of A in and

out of the bag cannot alter Ao. The same

holds for B, C, etc. The enzymes are present

only inside the bag, so reactions are limited

to its interior. This system constitutes the

simplest possible model of a living cell.

What can it do?

The enzyme E in the bag catalyzes a

chemical reaction converting A to B. The

rate of the reaction is a function of A alone

by Eq. (2).

Jr = R(A) (2)

Sometimes these reaction rate expres-

sions cn be very complex and depend upon

the concentrations of several components.

As the reaction progresses, diffusion of A

from the outside drives the concentration

of A towards Ao. After an initial transient

phase, these two competing processes take

place at equal rates, and the concentration

in the bag ceases to change, although both

processes continue. This condition is called

a steady state. The equality in the rates of

consumption and diffusion is expressed by

the condition of material balance for the in-

terior of the bag, Eq. (3).

Pa(Ao — A) = R(A) (3)

Equation (3) can be solved to obtain A, the

internal concentration, as a function of Ao,

numerically if not algebraically.

There are three distinct regimes in which

this system can operate: (1) A very small

gradient of concentration across the bag

leads to a flow sufficient to replace all of the

A consumed by the reaction. Then, to a

good approximation, the presence of the

bag may be ignored. (2) The flow across the

bag when the concentration inside is very

—————— —

eo ———

ee ee ee

REACTION AND DIFFUSION IN BIOLOGY 45

small is just enough to compensate for the

very slow rate of reaction which would

occur inside when there is not much mate-

rial to react. Then, to a good approxima-

tion, we can ignore the reaction altogether.

(3) Neither of these extremes holds, and the

reaction occurs at a considerable rate, with

an intermediate concentration gradient suf-

ficing to supply the reactant flowing across

the walls of the bag. Only in this third re-

gime can new effects arising from the

strong interaction between diffusion and

chemical reaction.

Now, we suppose that there is none of

the product, B, outside the bag. The reac-

tion is producing B inside the bag. The

equation of material balance for B, reflect-

ing the absence of B outside of the bag is

P,B = R(A) (4)

where Pp is the permeability of the bag to B,

and A is obtained by solving Eq. (3). One

effect arising from the presence of the bag is

that the chemical environment inside it is

decidedly different from that outside. This

Situation is central to the operation of

many processes within living cells, which

would be inactivated by the hostile condi-

tions of the external world, but operate

productively in the interior of cells.

A second, related effect is that concen-

tration gradients are created across the

walls of the bag. Concentration gradients

are one form of stored energy which may be

readily transformed for use in other proc-

esses. For example, two salt solutions at

different concentrations separated by an

ion exchange membrane will produce an

electric current through a pair of electrodes

inserted into the two solutions and con-

nected by a wire. Some of the energy con-

tained in the concentration gradient can,

for example, run an electrical machine

whose terminals are connected between the

electrodes. In the present case, the energy

stored in the concentration gradient came

from the free energy of the chemical reac-

tion catalyzed by the enzyme in the bag. In

living cells, the same effect is operating.

Cellular metabolic reactions lead to the ac-

cumulation of potassium ions inside the

cell, with sodium ions being expelled. As a

consequence of these concentration gra-

dients, an electrical potential develops

across the cell membrane. This potential is

used to energize the sending of signals by

nerve cells, for example.

A third effect is that the internal envi-

ronment of the bag, which we will call a cell

from now on, may be partially insulated

from the effects of variation of concentra-

tion outside. We illustrate the effect in the

following example. Suppose that substance

B is hydrogen ion, responsible for acidifica-

tion of your stomach, as well as polluted

rain. Many enzymes, including the one re-

sponsible for the reaction A — B in our

model cell, are quite sensitive to the con-

centration of hydrogen ion. In certain con-

centration ranges, the enzyme may be in-

creasingly inhibited by increasing hydrogen

ion concentrations. A phenomenological

rate expression is shown in Eq. (5).

R(A,H) = VmA+H/(Kn + H + H’/K})

(5)

Evidently, when H is small compared to

Kn, or large compared to Kj}, the reaction

rate is significantly depressed compared to

its rate when Ky < H < Kt}. The details of

how this might come about are not well

understood, but a comprehensible working

model can be understood by assuming that

the enzyme has multiple ionizable groups

with separated ionization constants. When

some but not others of these groups are

ionized, the enzyme is catalytically active.

This condition will hold only within a cer-

tain range of hydrogen ion concentrations.

For H in the vicinity of Kh, the required

condition will apply. The conditions of

material balance for this system, one for A

and one for H, are given in Eq. (6).

Pa(Ao ai A) a R(A,H)

Pi(Ho — H) = —R(A,H) (6)

where Phis the permeability to H, and Hols

the external hydrogen ion concentration.

Examining the behavior of our system as

H, varies, we might see the results shown in

the left side of the curve in Figure 2. As Ho

46 BARRY BUNOW

aoe ee ee eee ee Sa

0 2.9

=2.0

Fig. 2. Dependence of internal hydrogen ion con-

centration upon external hydrogen ion concentration

for a permeable bag containing an enzyme inhibited

by hydrogen ion, which is also a product of the reac-

tion. The internal concentration is highly buffered. A

ten-fold variation in external concentration is re-

flected by only a ten percent change in the internal

concentration.

varies over a ten-fold range, H varies only

by about 10 percent. How does this process

work? Without going into technical detail,

the idea is the following. When H is low,

the enzyme is active producing more H and

keeping the concentration up. As H, in-

creases, the enzyme starts to turn off, de-

creasing the rate of production. The effects

of increased entry from the outside and de-

creased production internally counterbal -

ance one another: His effectively buffered.

Living cells are likewise buffered in their in-

ternal composition, although it is unknown

whether this mechanism is actually respon-

sible for the effect.

Interestingly, the buffering effect just

demonstrated can be neatly turned around,

so that Hisa very rapidly varying function

of Ho. In the previous example, we no posit

that the enzyme is turned on, rather than

off, by increasing H over some concentra-

tion range. In terms of Eq. (5), this condi-

tion will be satisfied when H is the vicinity

of Kn. Now we might see the effect shown in

the right side part of the curve in Figure 3.

A ten percent variation in H, leads toa ten-

fold variation in H. Thus the cell is acting

as an amplifier of variations in Ho with a

Fig. 3. Dependence of internal hydrogen ion con-

centration upon external hydrogen ion concentration

for a permeable bag containing an enzyme activated

by hydrogen ion, which 1s also a product of the reac-

tion. The internal concentration varies much more

widely than the external concentration: a ten percent

change in external concentration leads to a ten-fold

change in the internal concentration.

gain of one hundred. This system works in

the following way: at low Ho, the enzyme is

turned off, and H and Hoare nearly equal.

As Hog is increased, the enzyme begins to

turn on, producing more H internally.

Both the increasing Ho and the increased

rate of internal production act in concert to

increase H. This increase further activates

the enzyme, which produces more H, in-

creasing the reaction rate even more. This

is a positive feedback situation, character-

istic of amplification proccesses.

Many enzymes are sensitive to hydrogen

10n concentration; the cell could simul-

taneously switch the state of function of

many different enzymes, thus altering the

flow through its metabolic reaction net-

work in response to small environmental

changes. There is nothing unique about the

role played by hydrogen ion in the present

model. Similar effects are easily designed

using the concentration of nutrients and

ions as well. In this example, as well as in

the previous one, the effect can also be ob-

served in the absence of the cell membrane,

using enzymes alone, but the strength of

the effect can be very much larger when dif-

fusion is involved.

REACTION AND DIFFUSION IN BIOLOGY 47

Another example will illustrate the ver-

satility of this simplest of models. A some-

what unusual type of enzyme has the prop-

erty that, at high concentrations of reactant,

the rate of reaction decreases with further

increases in reactant concentration. The

mechanism is not well understood, but a

phenomenological reaction rate expression

is given in Eq. (7).

R(A) = VmA/(Ka + A + A’/Ki’)_ (7)

In our cell, this enzyme may lead to the

occurence of multiple steady states. Figure

4 shows a case where, at the same Ao, there

are two observable values for the rate of

reaction inside the cell. First of all, what

does it mean to have two possible values?

After all, we are speaking about determin-

istic, not stochastic systems. The system

has a definite state. What we mean is that,

depending upon how the system is treated,

it may be found in either of the two states.

The upper state in Figure 4 is one in which

the rate of reaction is high, and as a result,

the concentration inside the cell, A, is low.

Atlow A, the enzyme is active, rapidly con-

suming A and keeping its concentration

low. The other state shown in Figure 4 is

one in which the rate of reaction is low. As

a result, A is not much lower than Ao. At

high concentration, A turns the enzyme

off, so that there is no mechanism to con-

sume A. The system can be put into one or

the other of these two states by slowly vary-

ing Ao. Starting at low values of Ao, there is

a unique steady state, which is connected to

the high reaction rate state smoothly as Ao

increases. Starting at high values of Ao, on

the other hand, the steady state is also

unique, but this time it is smoothly con-

nected to the low reaction rate state as Ais

decreased. |

As can be seen in the diagram, there are

two concentrations at which the curve of

reaction rate as a function of Ap becomes

vertical. If the concentration is varied

through the region of verticality, there is a

discontinuous change in the rate of reac-

tion. What role might this phenomenon of

bistability play in living cells? In engineer-

ing regulators, of which the thermostat ina

home heating system is a typical example,

bistability is employed to provide a range

of the quantity to be regulated over which

the regulator remains either on or off. This

greatly reduces the frequency with which

the transition between the on and off states

is carried out, increasing the lifetime of the

regulator. We simply don’t know if such

ideas have any relevance to the operation

of biological systems. In these systems,

many quantities, e.g. temperature, compo-

sition, electrical potential, etc. are regu-

lated. There is always a cost to changing the

state of the regulator, e.g. synthesis of new

proteins, so the utility of this idea at the cel-

lular level is a suggestive hypothesis.

Until this point, all of the systems had

time-invariant steady states. It may happen,

however, that the concentrations inside the

cell would not settle down in this manner.

One possibility is that the concentrations in

the cell oscillate periodically. This behavior

can be produced most simply by two types

of enzymes, both quite similar to ones we

have encountered previously. One case in-

volves an enzyme with hydrogen ion as a

reaction product. If the enzyme is increas-

ingly active as the hydrogen ion concentra-

tion, H, becomes small relative to that of

hydroxyl ion, O, then there is an additional

mechanism to control H: Ocan also diffuse

into the cell, neutralizing H via the water

buffer reaction, and further increasing the

activity of the enzyme. When the diffusion

of these ions is faster than the diffusion of

the reactant, temporal oscillations in con-

centration can result. The material balance

equations for this system are further com-

plicated by the fact that there is no steady

state, so that these equations are differen-

tial and not merely algebraic, as before.

The material balance conditions for this

system are given in Eq. (8).

V.dA/dt = Pa(Ao — A) — R(A,H)

V.dH/dt = Px(H. — H) (8)

+ P.0Kw(1/He — 1/H) + R(A,H)

where R(A,H) is given by Eq. (5), Piis per-

meability of the species, Ky is the ioniza-

tion product for water, and V, is the vol-

48 BARRY BUNOW

ume of the cell. An example of temporal

oscillations is shown in Figure 5.

A similar effect is possible with enzymes

inhibited by an excess of reactant when

there is a second reactant with no such in-

hibition whose diffusion is slower than the

first. Oscillations are sometimes seen in bio-

chemical systems. For many years, their

presence was ignored because a misunder-

standing of thermodynamics led biochem-

ists to believe that such effects were impos-

sible. More recently, careful experiments

have shown that coupled reaction systems

with certain kinds of feedback characteris-

tic of biochemical pathways can oscillate.

Incorporating diffusion enlarges dramati-

cally the variety of systems capable of

oscillation.

Many organisms are capable of telling

time, and show circadian periodic patterns

in their behavior even when placed in a

constant environment. In order to tell time,

oscillators are needed. This function can be

provided by our reaction-diffusion systems,

but no particular biological oscillator is

known to be responsible for synchronizing

any of these circadian rhythms.

UPPER STATE ~,

(STABLE)

a MIDDLE STATE

[UNSTABLE }

INTERNAL CONCENTRATION--S

LOWER STATE

(STABLE)

0 20 40 60 80

EXTERNAL SUBSTRATE CONCENTRAT 1ON--SO

Fig. 4. Multiple steady states in a permeable bag

containing an enzyme inhibited by an excess of sub-

strate. Depending upon the way in which the system is

exposed to variations in the external concentration,

one or the other of the two stable steady states can be

assumed.

The examples shown here were con-

structed out of just two parts, a permeable

membrane and a single enzymatic catalyst.

It is clearly a very long way to a living orga-

nism witha very large number of cells anda

large number of enzymes. Nevertheless, it

is remarkable that effects such as energy

conversion, homeostatic regulation, ampli-

fication, bistability, and oscillation can all

be achieved with this simplest case. Engi-

neering implementations of these functions

in hardware are generally achieved only

after elaborate design efforts and involve

many components. It appears that the

building blocks of biological systems may

be especially appropriate for the design of

devices to perform elaborate function needed

by living things.

Some Generalizations of the Simple Model

Clearly we can make only the most pre-

liminary of explorations in the direction of

generalization. There are just too many

possibilities. Among the simplest are the

following:

More Than One Reaction

Still in the context of the cell of Figure 1,

we now suppose that A and Bare present in

the external solution, while there are en-

zymes inside the cell which catalyze the se-

quential reactions A — B, with rate Ri(A),

and B — C, with rate R2(B). Eq. (9) gives

the material balance conditions for the

steady state of this system.

P,(Ao — A) = VaRWA)

P,(Bo — B) —VaRi(A) + VeR.(B) (9)

Poc = VpR2(B)

From our previous discussion, we know

that, as a result of reaction, the concentra-

tion of B in the cell will be elevated over its

level outside. This B is available for conver-

sion to C by the second enzyme. Compar-

ing the rate at which the second reaction

produces C to that at which it would be

REACTION AND DIFFUSION IN BIOLOGY 49

produced were the first enzyme absent, it is

evident that the rate is faster in the former

instance. One effect of the presence of the

two enzymes inside the cell is to speed up

the production of C. This is true even

though the first enzyme does not produce C

itself. There is a synergistic interaction

leading to an increased efficiency. The or-

ganization of the intricate web of chemical

reactions going on in real cells takes advan-

tage of this synergism at almost every turn

both to accelerate as well as to coordinate

the chemical flows of many substances

simultaneously.

Two Cells

We suppose that two cells as defined in

the previous section are in diffusional con-

tact, so that substances can diffuse from

one cell to another. We assume that the en-

zyme in both cells obeys the rate law of

Eq. (10).

R(A,B) = VmA-B/(Ka + A + A’/K3)

(10)

Both A and Bare reactants in this reaction,

and we denote the concentration of species

I in cell j by I;. There are four material bal-

ance conditions, one for each species in

each cell, given as Eq. (8).

Pa(Ao — Ai) + Qa(Ai — A2) = R(A:,B1)

Px(Bo — Bi) + Qu(Bi — Bz) = R(A:,B1)

(11)

Pa(Ao — Az) — Qa(Ai — A2) = R(A2,Bz2)

Pi(Bo — Bz) — Qi(Bi — Bz) = R(A2.Bz2)

In these equations, P; represents the

permeability of the barrier between the cell

and the external reservoir, while Qj repre-

sents the permeability of the barrier be-

tween the two cells. These four equations

are to be solved for A; and Bi, as functions

of A.and Bg. Since the two cells are identi-

cal, there is no reason to suppose that the

concentration inside one will differ from

that inside the other. Thus, even though

transport is possible between the cells,

there is no gradient, and therefore no flow.

It would appear, then that putting the two

cells in contact had no effect. Appearances

can be deceiving, however. In fact, this as-

sumption about communication suffices to

permit the two cells, which are identical in

their permeability and enzyme content to

differ in the rate at which reactions go on

inside them. The arrangement of the two

cells is shown in Figure 6. The kinetics

which lead to bistability in a single cell

permit the concentrations in the two cells

to differ. In effect, one cell is in one of the

two stable states while the other cell is in

the second. This condition can occur even

though the two cells are identical both in

their connection and initial state.

One process by which systems spontane-

ously lose their symmetry is called bifurca-

tion for the reason illustrated in Figure 7.

In the figure we have plotted AA, the con-

centration difference between the cells as a

function of Ao. For all values of Ao, there is

at least one solution for AA, namely zero.

There is also a range of values for A» within

which there are three solutions, while else-

where the solution is unique. Mathemati-

cians speak of the two new solutions which

appear as having bifurcated from the origi-

nal solution. Near the point where the new

solutions appear there is a certain resem-

blance to a three-pronged fork, with the

single solution acting as the handle on one

side of the point, and the central prong on

the other. In the regime where there are

three distinct solutions mathematically, the

system described by equations still has only

one solution. Which of the three ts selected

is controlled by considerations of stability

and by initial conditions.

We have seen in the case of the single cell

that the selection between the two solutions

shown there depended upon initial condi-

tions. Here, however, we assume that the

two cells were initially identical. How then

do they become different? There are two

aspects to this issue. In the regime where

the solution with equal concentrations in

the two cells is unique, it is also stable:

small perturbations in c relax back to the

original solution. In the regime of multiple

solutions, this solution is unstable to such

50 BARRY BUNOW

$60 en a a nn a a rn nn rn ne nr nnn nnn nnn nn ee - - === I --- ——— ---—- I-----

I I

I IX be ¢ Po | I

4.5 - xX XXX K XE x &3I-

I IX irr I ar

I I bo @ ¢ I Err D4 I

I IX be I

9.0 - KIX I IXXr I III I -

EX EX Ir b ? 4 1

Ir Xr xr gr I

»¢ xx xX I I x D ¢ I

I XY IX I I

3o5.0X x X I -

I x x I I I I

Ds x x I I I

I x I I I

3.0 - » 4 X I > 4 -

I I Dg b 4 D4 z I

I X Xr p I

I I I I I

x I I z I

22 > I I -

x X xX I I I

I I I D4 I

I D ¢ I »¢ I

I » 4 I D 4 =

2.0) = I X b >

I x x I I I I I

I b I I I

I Ir Ir rr I

1.5 - I 4 I I » 4 -

I » 4 I I rz I

I » © ¢ Ir Ir I

> Ix Ir I ar 1

I I Le 4 I » Ir I

1.0 - XX II ar -

I I bo ¢ I IImr I Irr I

I IXX III bo o4 I

I I Ixr x be @ 4 I Ir 1

0.5 - XXXXX IIIK EEIr -

p Ir bo ¢ ar z

I I

x I

CU) MOSSES DS SS LSSSS SDSS => eres 1 D Ieoee eo I----------- I-----------I

0.0 0.3 0.6 0.9 5 1.8 2.1 2.8 2.7 3.0

STATE1

@REACTION-DIFFUSION OSCILLATOR

VARIABLE

13(¢0)

Fig. 5. Time-periodic variation of concentration of A inside cell containing an enzyme catalyzing a reaction

with hydrogen ion as a product, where the enzyme is maximally active at moderate acidity. The hydrogen ion

concentration also oscillates.

perturbations. The other two solutions,

however, are stable to small perturbations.

Selection between these two solutions de-

pends upon the details of the perturbation.

For example, in one of the solutions, the

concentration is higher in one cell than in

the other. Initial perturbations from the

uniform state which have this character

will lead to that solution. Where do these

perturbations come from? Since the trivial

solution is unstable, any perturbation, no

matter how small, suffices to set the system

evolving away from that solution. Very

small perturbations are generated, for ex-

ample, by thermal fluctuations. We might

also suppose that some external influence

was being applied.

Another type of bifurcation, this time

involving two cells with the enzyme whose

kinetics led to oscillations in a single cell,

leads to oscillations in which both the

phase and amplitude of the oscillation may

be different in the two cells. In effect, there

is a kind of signalling going on between the

two oscillating cells. Differentiation proc-

esses, such as those involved in the aggrega-

tion of the slime molds Dictyostelium dis-

coidium, have been shown to involve this

kind of signalling, in which an oscillatory

emission by one cell is received by a second

cell of the same type, lending to synchroni-

zation of the signal production.

Reaction and Diffusion in Tissues

A tissue is composed of a very large

number of cells. To a good approximation,

a large number of small cells can be consid-

ered as a continuum. Fick’s law is used to

describe continuum diffusion, Eq. (12).

J =—D-.Ve (12)

REACTION AND DIFFUSION IN BIOLOGY 51

Fig. 6. Arrangement of two communicating cells,

each with identical membranes, and containing equal

concentrations of the same enzyme. Under appropri-

ate conditions, this system can spontaneously evolve

from astate in which the concentrations of reactants is

identical in the two cells to one in which they are sta-

bly different.

In Eq. (12) J is a vector field equal to the

rate of flow of substance c, De, which meas-

ures how flow responds to the gradient, is

called the diffusion coefficient, and Vc is

the local concentration gradient. A chemi-

cal reaction is a scalar field,

ta (ec) (13)

which depends upon the local concentra-

tion. The connection between reaction and

diffusion lies in the fact that material is

conserved: in any small region, anything

which diffuses in either reacts or accumu-

lates.

dc/dt = DAc — R(c) (14)

In general, there will be more than just

material c, and there is one equation for

each material. The connection between the

equations for the several materials lies in

the fact that R(c) will depend upon the con-

centrations of all of the materials which

Participate in the reaction. Usually, we are

concerned principally with the steady states,

so that the dc/dt term in Eq. (14) may be set

eae

Fig. 7. Typical bifurcation diagram for the two-

cell system. The vertical axis is the concentration dif-

ference between the cells, while the horizontal axis is

the external concentration. Stable solutions are indi-

cated by solid lines, and unstable solutions by dashed

lines.

to zero. Equations of the form of (14) are

generally solved numerically, the details of

which are not the subject of this presentation.

A Thread of Tissue

As a first model for a tissue, we consider

a thread of material immersed in a stirred

medium, illustrated in Figure 8. We sup-

pose, for mathematical simplicity that the

two ends of the thread permit no matter to

cross them. Transport from the medium

into the thread is also by diffusion. How-

ever, the thread is assumed to be so thin

that the flow in this direction may be re-

placed by a simple Fick term.

(15)

where cis the concentration at the center of

the thread. Thus the form of the reaction-

diffusion equation for this system is

dc/dt = D.d*c/dx?

“Pees = 'c}'— Ro)

J external = P(Co = €)

(16)

Suppose for the moment that the con-

centration, c is both uniform and constant

52 BARRY BUNOW

along the entire thread. Then, in Eq. (16),

terms containing spatial and temporal de-

rivatives vanish, and the equation becomes

indistinguishable from the material bal-

ance equation for a single cell treated pre-

viously. There is always at least one such

solution, which is uninteresting, in the

sense that it shows no new behavior besides

what was seen in the system with a single

cell. This trivial solution may also bifur-

cate, in much the same way we saw pre-

viously for the two cell system.

The result of bifurcation is a non-uni-

form concentration profile along the length

of the thread. Bifurcation theory is useful

in predicting the approximate form of these

concentration profiles, as well as the pa-

rameter values for which these solutions

can be expected. Figure 9 shows several ex-

amples of profiles which were computed

numerically. A modification of this prob-

lem, studied because it was feasible rather

than for any biologically relevant reason,

was to connect the ends of the thread, form-

ing a ring. Some solutions on the ring are

shown in Figure 10.

As in the cellular case, there are also bi-

furcations to time-periodic solutions. Some

examples are shown in Figure 11. In the fig-

ure, each horizontal curve is a snapshot of

the solution, and successive snapshots are

arranged behind one another. The vertical

coordinate is the concentration.

In the thread there is a new type of phe-

nomenon which could not be observed in

the one- and two-cell models: propagation

of a wave of steep change in concentration

along the thread. We suppose that the

thread is initially at a rest state. At one end,

one of the resulting substances is injected

for a short period, then the end is closed

off. In the vicinity of the injection, the con-

centration rises. When a threshold is crossed,

the zone of high concentration starts to

spread, much more rapidly than it would

by simple diffusion. Eventually, this zone

passes along the length of the thread. Even-

tually, however, the high concentration dif-

fuses out of the thread, restoring the initial

uniform concentration profile. An example

of such a wave solution is shown in Figure

12. Propagating waves of concentration .

have been hypothesized to play a role in

setting up a “coordinate system” within

which the cells of a developing embryo can

measure their distance from centers of or-

ganization. In this hypothesis, the distance

information is perceived as time delays be-

tween waves of different types. There is no

convincing evidence to support this idea,

but convincing evidence in embryology is

hard to come by, so its absence is not to-

tally damning. The action potential in

nerve is a more convincing example of a

propagating wave of activity. While meas-

urements in the nervous system are gener-

ally of electrical quantities, the process by

which the potentials are generated are fun-

damentally chemical. There is a very close

resemblance between commonly used models

for impulse propagation in nerve and the

reaction-diffusion model we have described

here.

A thread is a rather feeble model for a

tissue. Its use may be justified by the obser-

vation that many of the phenomena possible

in more elaborate geometries can be ob-

[_UNSTIRRED LAYER

1S +4 “acTIVE LAY

NNSASAAAASSANS AAAS SSAA AS NN NAAAAAAANAAASAS ASS BARAAS

Soe

og Dens es Se

Fig. 8. A thread of tissue, closed at the ends,

immersed in a reservoir. Reactants and products can

diffuse in and out of the thread through its surface in

common with the reservoir. They can also diffuse

along the length of the thread.

REACTION AND DIFFUSION IN BIOLOGY 53

served here. At the same time, computa-

tions are significantly shorter than for the

two-dimensional tissue models described

in the next section. For the latter, the cost

of computation remains a significant im-

pediment to research in this area.

A Two-Dimensional Sheet of Tissue

Some tissues, especially those of embry-

onic stages of primitive animals, have the

form of an extended sheet, only one or a

few cells in thickness, but tens or hundreds

of cells in extent. The cell layer is nourished

from the planar surface, and transport

through the edges of the sheet is unimpor-

tant. The embryo is growing, and as it does,

the size of the tissue sheet increases. Analy-

sis shows that the size of the sheet plays the

role of a bifurcation parameter. For suffi-

ciently small size, the only solution is con-

stant on the whole sheet. As the tissue

grows, a bifurcation occurs, anda non uni-

form pattern of concentration develops.

Further growth causes the pattern to in-

crease in amplitude and change shape, but

eventually, the pattern disappears, leaving

only the uniform solution. At a somewhat

larger size, a second pattern develops,

grows, and disappears as the size continues

to increase. Sometimes there is no range of

sizes without patterns intervening between

patterns of a given type; patterns succeed

one another as the tissue gets larger. Gen-

erally, the patterns increase in complexity

as the tissue grows. One sequence of pat-

terns is shown in Figure 13. In fact, many

other patterns were obtained in the calcula-

tion leading to this figure, but only a selec-

tion are shown there.

At an earlier stage of development than

the tissue sheets described above, the em-

bryo is simply a closed surface layer, sur-

rounding the yolk of the egg. Growth at

this stage consists principally of an increase

in the number of cells, each cell becoming

smaller. Diffusion of substances in the

layer becomes generally slower, since more

cell membranes must be crossed in order to

move the same distance. The diffusion

coefficient can serve as a bifurcation pa-

CONCENTRATION

fo)

PIS)

0) Ae 1

Fig. 9. Non-uniform concentration profiles along

the length of the tissue thread. These profiles develop

spontaneously from an initial state of uniform con-

centration. Which of several possible patterns is actu-

ally observed depends upon the physical and chemical

parameters characterizing the tissue, and, possibly, on

the form of the perturbation which is applied externally.

rameter insuchasystem. Figure 14shows a

selection of patterns computed on the sur-

face of an ellipse following growth under

these conditions.

The mechanism by which a fertilized egg

turns into an adult organism 1s almost

completely obscure. One undeniable ob-

servation is that an assemblage of cells

which are morphologically and chemically

similar in early stage becomes an assem-

blage in which one cell differs from another.

A second observation is that reaction and

diffusion processes play a central role in the

operation of living cells. We have seen that

the equations of reaction-diffusion possess

the property of spontaneously developing

solutions with spatial organization starting

from uniform initial conditions. It is very

tempting to suppose that this property lies

at the basis of differentiation in living or-

ganisms. Proving that this connection Is

Operating is rather more difficult. Biolo-

gists initially rejected the idea, on the

grounds that the patterns found in living sys-

tems were much too complex to have origi-

nated from some mathematical operation.

We have seen, however, that patterns of

considerable complexity do appear this

way. Probably models could be concocted

to display almost any desired pattern, so

54 BARRY BUNOW

Fig. 10. Non-uniform concentration profiles along the circumference of a circular thread of tissue. The cir-

cular shape is intended to show some possibilities for patterns in tissue shapes such as hollow organs or long tubes.

this objection is inappropriate. Biological

patterns are of two types, very stable, e.g.

the number of vertebral bones in the spine,

and highly variable within limits, e.g. the

stripes on a zebra or butterfly. Depending

On parameter choices, the patterns pro-

duced by the reaction-diffusion equations

may also be of these two types. It appears

to be rather difficult to exclude the hy-

pothesis that pattern formation in embry-

ology might use reaction-diffusion mecha-

nisms. At the present there are no other

theories which can be stated with sufficient

precision to even design experiments to test

them.

Discussion

Within the domain of biology, simul-

taneous reaction and diffusion play many

REACTION AND DIFFUSION IN BIOLOGY 55

Fig. 11. Time-periodic variation of concentration in tissue thread. Each curve is a snapshot at equally spaced

times. Successive curves are arranged with the earliest apparently nearest the observer. The height of the curve is

proportional to the concentration.

other well-established and important roles

which we have not even mentioned. Among

these the most thoroughly studied is the

transport of oxygen from blood capillaries

to its site of consumption in tissue. Capil-

laries are separated by distances of the

order of tens of microns. Each tissue cell is

no more than a few cell diameters from a

capillary, and nutrients must diffuse from

capillary to cell over this distance. Nu-

trients, hormones and pharmacological

agents must also be distributed by the same

mechanism, but almost nothing is known

about the details. Carbon dioxide pro-

duced in tissue passes out to capillaries bya

mechanism which is almost the exact in-

verse of that by which oxygen enters. The

Same must also be true of other metabolic

end products. Again, these problems are

almost untouched. The physiology of intes-

tinal absorbtion is an example of diffusion

and reaction at multiple locations simul-

taneously. Within the lumen of the intes-

tine, there are narrow crevices and pits

whose typical dimension is of the order of

tens of micra. Within these spaces, nu-

trients are enzymatically digested as they

diffuse toward the intestinal walls. To

reach the bloodstream, these nutrients must

pass across a layer of cells. Some consump-

tion and further enzymatic digestion goes

on in this layer. Many other examples

could be sited. The point is that there is a

broad area in the physiology of tissues and

organs where simultaneous reaction and

diffusion clearly plays an important, even

central role.

In each of the little model problems de-

scribed above, we have postulated an

‘“‘anatomy”’ (in the form of cells and their

O2-eeesmE Boedooe": lo-cocens oo Lo-ecoo"s

ionis

Fig. 12. Propagating wave-type of solution to reaction-diffusion equations ona thread. The tissue is initially at a uniform concentration. Then a perturbat

applied at one end. If the perturbation is both long and large enough, a wave of activity will sweep across the thread, even after the perturbation is removed. After

long times, the system returns to its initial condition, and a second wave can be initiated.

“UMOYS 21B YSIYM JO aWIOS A[UO ‘susaiied Jo satias vO} suIpeo]

‘eInjoNI}s NUOAIQUIA UB O} SUIPUOdsaII09 UOIZ|I JBUO!SUSWIP-OmM) IB[NBIIII UB UO susaIIed UOISNJIP UoNvAY Cr Shy

rnin. 6e/6 erin: aMERiaIGa Acie afc AIO Ca Bree : wee Ne KARAS

SE IO SOR) OT One ice ois rewenakancans

Rw Whe sie wo syateneteta sts Aiels RISES Bye nial sinh Sale ae tesa NASSSAAN ASSL

RN AISONCeya Uwe le ules Cre eNO ra acters DS SsisiSSinnisie PUN CRANKS wevewe RIDReN Sie ee eta ye

SANS <5 brn c'n'n, wR cisicre ie SRO QS SEER a “hshwss

ENN Ne Ne ee aN Ns Se ean DREN SERN ITE sa eeeee

mT ANN SA SAR SA se NSSSS NANA AAA ee . sweeney

SSNASNNNGN Seco benny: > tee “AS

Sor eesene = we eee ANN

fs LOG aRHE & Bower mer ehiiT ‘5 = Neu NGO

N SSQQy tet oT a eT 2: a

SS SITSISSSS— Wem tens 3 =: RNAeaR ERR TE

enciceee Zt TEELILisse VASA a taNg zz 3 FF] NAAN ANNAN eo rT

=e mS) HR ELACELECEC Yh #0 =} E NSARA AAA ee ee

ec itneenaseds REE Hiss Peon ae

ER MOm conte rere eendies i FFL is TEAR eae

Se Le ROKER: uC N = 3 Ege: TT NNNN Nene ee

Stic ed he Seo NN = 5 Se PP ONNNNN wee es

ee Oe hese ore Oa RMS, NN = Bs: see eee a. .

C10) SSSESEES E EEHSG EERE tore pee

NN ere LAT SSSEEEE = Es th eee eee ANNA eee

SS. Sviseeminmwe bin NAASEEES ERz:: Sn ce «eee

SSE =: NNN Sinta emit itt NSSEESE Ba eS WeeweeeNQNNcens

zit: es SSS=EL=== SG mesenieie os

Soe e ie NSSSNN S802 TEESSSEEESE 2 EEEEEE Win Silas RRO,

== am =f ze NNANSSNSANN § S SSie HANNE 4 miniyia ernie w Wie RNIN & or Sale

= = zt SS . SINNSSSSNANW See eG SSSEEEE Uuumeacane mute een

~ = N NN : 3 PSSSSiz DIEDIDEDESSSSSii fF

~ ; yf -- oD DSSS NNSSASNN See cesar te PES SSNS RRP ,

R -s : SSS + ANANAN--- ‘

N i Se oe we QNAANASSNS SSN NNN NN eee ee = ‘ SS

= ep aca SS ESR NSN ARES ra SC one se oe een My SU eee ne eee = HA NS

+e SeAR Sz : SN LSo a es era RR Hee cin es Dee mete panics Sven elec - BE Ee SERS SS Sas HEN SS

‘S, hep ee te o> a SSaieier sal aaa eaten stare aeeistaie eta che. =: Sioleta aietmmiminieiere stale wate aE te mmm ENNANAA Ree eens SNS N

seen Satoh Sieg Unilin Brand Cela ica sea aan Sac SRDS claret sia tials arerere a rae SSNCNGNGUSe See SSS S SMNEEREGGRRCE oe

— Bisiale pet theif laa Be NNN NAAN SAAN eee = NS==> MNNNA SNe siwieelwiu/win dies NSN N TANNINS Ne eee

tte FOR GS oe SE ESTE NN SSR SAAN RNASE ARON CC eles Soe : DENN NSS CS SEE i SSS Ng SRGSNSGA Suse aes

[ee] Sami wm ON Nereis in oer SF Et aig 6 Ow cA RRMIREN Sa See eee F === SSSSSSNS NSS mT aaaty H ANNES vaio one Gee ESS = A WGA Maine Gig Gn cals

PANNA eee e eee Bet pee pe en Oe werner e ene Se eee ve Si: SN NGS We Seiernenisy nie 333 Ss. NNNN 8 tet eee ees

na ara s{ Cis OAS wares : NAN iv se REA ums eiiBT ne ny Wi ence he SS SS SiGak wh eee by Cal aig

Zz Seka ot SUSE ENS SSS SDURN Nereis eee #35 SS tmaNANNe tee eeseeeees

Biv emnware canis Re SSeS SEAN ule sieiats pure are = NANG te eee eee eee

— hh. ere S =: Cpe eNWNeenctcccnee - 2 : VWNGweeacssehivocs

amttaie sp ats == AER OT OE aS EREERH LISRA WANS ww RVelwn Ee aes

bianco ss. Hie Nnis eee ia asae FFEEEH HhoassNte sees eee l:

Z Sateee oeSS SNEDDON G ek eects AEREH SSStttrteeees

Ne Wekcte wis SS SE SNe ae EESa> Sean eee ee

=3 Powe Ne £ & TNNSO eee

> tweens = eNSNe eens

i ites H =s3s

DN N

$ENNSies es eeemnc rere

= Mtawweseperece ses

Sevccnee bibs eae

ies ostawaweg eee

= Th ae

ay awe Vaan egiber qaneat 4 : Sttitstee te eee seas eeseness

Z SSeS 4 dW EEC ES no ONGR ACKER ; ziit Sra sste tees restesesesseceeess

Oo wee tenwee eee RAANRAAU eee = SSEEE aes see year ae hf eg dae

(@) A4 0 SE 6s neSanneuNape : SSS TARR Se ce Sunn eRe E eet eae

pom ale che ete leis pals ae : : SSSSSS SRE RNR RRS cena seca ne

RGAAT Coes sa aR ADTs Siete HESS PSPS aNNSANSAURSSSSAN SUNS ASARMARRARSONSA

ad se cc vANRS aes ESSE BNNS : TIT INSSTTENNTN NESS SSN SS

isan SSSz SSSSSSE BEST TE TEST SSSNNS SSS NSS SSNS SEE

S ee $3: " BSSSSSS Popitrsiiisfiiisiisiisissiestesteest

< SSS--- Site mai SSS : (GEST IRieA davis caw Recbenueen eters Keer

SANw acer : ees aa SSS = peteenes mia caubev Tansey fen

<3) Sashes HS EEE SaRice EEEESSS : Spa Os SER UNUAR

SAN AAA. NN Roi Soe ' PSS : : ‘ tenebuens +4

f= NANAAAAA- HNN AAAAA-- ‘ NN = SSE5 Pan Ce we a See

NSSSA NAN - eee, SSS VARMA Nee 671 N 35 SSSSSHLELL! : : H SSSA

SSSSSS SSN eee ee NNN eee NNN Ne ‘ N ss SESSSPPSISSLLL LLL: SESS re SESS SNS

SSSSSSSSSSE = ema HSS ses oa NNANANe ' = = SSSSSSSSSSSSSSPS ALLELE SESE SSS5z

TF PAN SAN ANS - SSSi=: WhisNie/o 8 CINGNNA eS Te : BEE E SE ESSSSSSSSSSIP PSPS SVS Sypgys SESS >

DIIIDSQEMSN ESET SSS SnunstecsNwys seca) : BEEEEEEEEN SN NNN

ee oe ee ' NNN Use ca pia Sle lecaxe WN NWN ie ’ = Se etiee te ee en ae wat oa oe JSST SSSSS SSS SSy ==

tee ee ANNAN 1 33> SIs {Ravi Sata GeO ew co tk = — = RESSESSSSSSSESSSSS=S=S=5=: =

tt tere ee ANA e ‘ NSS RAW Musi Teg da hG: NINN oe ey: = pac tplel eee SS:

: 2 €88o 8 aS RNIN aaa es Wty of NN Ushsikt ote Kiwi lg WINN R See oh (ia 5 SSERECSEEE EES rs

rS 74:8 G0 8s WiC ee RUNG pre i704 NN Wiad reie fairey aida wes cae ac 4 3 SESsnSssz ZESe

$s ES so oer ay NN Fe eee ee eee ee ANNAN ue g = 5 s

33 DIDTILITI I Deg SSe82sS- SS CMU oy A eR AN on 5 :

< 33 ess eeeseReaeQNQNaabe ye NN FROLIC STN RO me e 3

. =: *\e10\0 6 OS asia Wales RN ANE CT SNe MeN Uae. CTs ate Ba Gte Min ow 5 i

Se re to (96.8.8 «0S EROS, POR ee ee SSSz a a iri i ear, a are ' 8 =

bs apm <P tS A 3} 20 by 81's Veins c ia kEERe ASS SSS= Si ace ei R Likigiaiy wie oe R ‘ 4

Titre ai dae te a : 3 2 69.30» 00.0 Madoc KEAN Bm S§ ae RW ie bp a ele we GR, aa) =

was SE > : ee ne ay S8=s sete nne . we =

earl $333 i35 waar e's 1 dela maar aieee SSS: si aieiOahag Serer tS 3

oN sve 223 3S33 =e et THOUS A RSC LTTE RSE cease Neweweeeted FY

weer S2SSTSz gsc i ee ee et 8 ae NN aeeue bake LN 3

; 23533333 $$$ tipeer ene ecssenconenncce rit see ESR NeEnTHEN 3

e253 S$S3SF3 = ss TEDES pr espa ee ses Semen nth SNe ees ae :

PLETE SSSSREESS ze 35 ie oda pawitatiie Vand SE DNA resi tE ENS ee er a 3

aoe z=: cet sh a0 ope Nae ee ee me : i

* SR33SSS3 S838 PSS SSS SSS fees ee dene seneeeRNReamn tits . wre N H

USE SSSSESSSE ESSE SETS SIT TSE whet ce tees t ease eeseeaseNNaueersiis vee Pana HEN

333 i agua genie

SSUES pene OPE es PACE : triige

STPPPHF PPP TIIIT es ss aateks s Ades Seats 10 ein we TS

SSSESSSESSS TESTS AEE apt obs LsWeNiigees sews cee mitts 5

PPTSSSTETSPPP ese oe rr ye rar \ NN =

3333 333335 ce ge ORANG ae Foe mee R eee iSs E

3333335 A See ae pl get 4a s os tivetnie nea ESSS =e

333: é RAGRANTORP REEF 9.034.008 ePrice ne pcr Teer ESSS 5:

ete Bays: Danas ae ab Cee eee peewee RNNwwet SESE EE

Ot pwm men sennreee Cane ee reeeseQNkoatt : ESSSS EE

SLs cRMASaS RAR Gnie © pF tos bey SAR ees RMR SESS A

“eet pmwawannngncareccers Cede e eee ann neeens PESTSE e

rere rent Cerri ceewee Anan S85 =

wa esnnnenee a

BARRY BUNOW

“UMOYUS =

uonsajas B A[UO YotymM Jo *susayjed Jo satias B SI 91941 “IINBIY sNOTAaId SY} UI SY “9dBJINS 94} Ul UOISNYJIP ZuloNpal ‘sasvoiOUl 2devjAINs dy) UO AlIsUap |]99 dy) ‘spas001d

joasut Surdojaaap k Jo aoRJANs ay) JUasaidas 0} papuajU si WaIshs sty] ‘prosdi|[a UB Jo sdRJ1Ns 94] UO suiajjed uoisnjjip-uonoeay ‘pl “sl

yIMols sy ‘339

LEB

eae

oes <a

i rt

eer} Et

: Toe! TS

eas

ire Sr

ite ———

eh ——

Nina i

aaa a

ae TT

Fas a

bine

————

Tree Saas

HEH

REACTION AND DIFFUSION IN BIOLOGY 59

connections or regions with boundaries)

and a “biochemistry” (in the form of de-

tails of the kinetics of reactions going on in

the cells or tissue). From this information,

we have inferred a ‘“‘physiology” (the inte-

grated behavior of the composite system).

This logical approach is almost exactly the

opposite from that taken in biological in-

vestigations, where the integrated behavior

is the initial observation, with the goal

being to infer which structures and which

mechanisms are responsible. The present

approach is akin to that of an engineer

who, given a supply of components, as-

sembles them into a device. Usually he

knows in advance what properties the de-

vice will have, but sometimes the assem-

blage does the unexpected. Even morerarely,

this unexpected behavior will turn out to be

applicable. In order to have an idea of

where to look for mechanistic explanations

of physiological observations, it is essen-

tial to have a “handbook” of the possible

behaviors of the underlying components.

Some behaviors of components and as-

semblages which have been studied will

remain unknown: perhaps we failed to in-

clude the right ranges of physical and

chemical parameters, or perhaps we lacked

sufficient imagination. Nevertheless, such

studies provide a healthy stimulus to sub-

sequent workers.

We have seen here using this synthetic

approach that components whose charac-

teristics are like those present in living or-

ganisms could perform certain functions

which are like the functions observed in liv-

ing organisms. One clear limitation is that

any given function need not be performed

that way by a particular organism. Never-

theless, it may be useful to see that these

functions do not require complex struc-

tures in principle in order to operate at all.

One use is that the epistemological program

of biology is supported. If functions of the

type described here could only be per-

formed by very elaborate organization of

complicated components, then it would be

unlikely that laboratory experiment could

sort out how real biological systems oper-

ate. Another use might be to encourage ef-

forts to design and fabricate prosthetic de-

vices In order to restore or even enhance

normal operation of physiological systems.

Again, if such devices had to be too com-

plex, there would be little hope at succeed-

ing in their construction.

Finally, theory in the biological sciences

has generally been held in low esteem, in

striking contrast to its constructive posi-

tion in such sciences as physics and chemis-

try. This state of affairs is frequently justi-

fied by statements to the effect that biology

is too complicated to permit useful predic-

tions to be made on the basis of analyzing

only simple models. The study of reaction-

diffusion phenomena suggests that certain

kinds of biological complexity may emerge

from very simple physical and chemical

principles. It further suggests that the arro-

gant presumption of biologists to deny the

utility of theory may be destructive for the

progress of their science.

References Cited

This reference list includes reference only

to the author’s work, not because his is the

only good work in this area, but because

the references in the listed papers will pro-

vide all the references which the interested

reader will require to pursue this field

further. The alternative would have been a

reference list here which would be quite in-

complete or intolerably lengthy.

1. Bunow, B. (1974) ‘‘Enzyme Kinetics in Cells,”’ Bull.

Math. Biophys. 36, 150-168.

2. Bunow, B. and C. K. Colton (1975) ‘Substrate Inhi-

bition Kinetics in Cellular Assemblages,” Biosys-

tems 7, 160-171.

3. Bunow, B. and C. K. Colton (1975) ‘‘Steady State

Multiplicity in Cellular Arrays of Enzymes with

Hydrogen Ion as a Reaction Product,” in Analysis

and Simulation of Immobilized Enzyme Systems,

ed. by J. P. Kernevez and D. Thomas. North-Hol-

land, Amsterdam.

4. Bunow, B. (1978) ‘‘Chemical Reactions and Mem-

branes: A Macroscopic Basis for Active Transport

Facilitated Transport and Chemiosmosis” J. Theor.

Biol. 75, 51-78.

5. Bunow, B. (1978) ‘Chemical Reactions and Mem-

branes: A Macroscopic Basis for Active Transport,

Part II—Nonlinear Aspects,’ J. Theor. Biol. 75,

79-96.

6. Kernevez, J. P., G. Joly, M. C. Duban, B. Bunow and

D. Thomas (1979) ‘Hysteresis, Oscillations, and

Pattern Formation in Realistic Immobilized En-

zyme Systems,”’ J. Math. Biol. 7, 41-56.

7. Kernevez, J. P., G. Joly, D. Thomas and B. Bunow |

(1980) ‘‘Pattern Formation and Wave Propagation

in the SA-System,”’ in Nonlinear Eigenvalue Prob-

lems, ed. by C. Bardos, J. M. Lasry, and M.

Journal of the Washington Academy of Sciences,

Volume 72. Number 2, Pages 60-66, June 1982

Schatzman. pp. 201-221. Springer Lecture Notes

in Mathematics, No. 782.

8. Bunow, B., J. P. Kernevez, M. C. Duban, G. Joly and

D. Thomas (1980) ‘‘Pattern Formation by Reaction-

Diffusion Instabilities: Application to Morpho-

genesis in Drosophila,” J. Theor. Biol. 84, 629-649.

9. Bunow, B. (in press) ‘“‘Turing and the Physico-

chemical Basis of Biological Patterns,” in Turing

Memorial, ed. J. Prewitt, IEEE.

Plasmid carriage in Vibrionaceae and other bacteria

isolated from the aquatic environment’

L. A. McNicol, T. Barkay, M. J. Voll, and R. R. Colwell*

Department of Microbiology, University of Maryland,

College Park, Maryland 20742

ABSTRACT

Twenty-eight percent of a collection of Aeromonas hydrophila organisms isolated selectively

and non-selectively from the natural environment of Bangladesh and the Chesapeake Bay

contained extrachromosomal deoxyribonucleic acid (DNA) bands in agarose gel electro-

phoresis. Vibrio spp. isolated from the same waters had a similar frequency of plasmid carriage

(22%), while more of the Escherichia coli (53%) and Pseudomonas spp. (41%) isolated from the

same sources possessed plasmids.

The plasmids from environmental bacteria had cryptic functions, except in those cases

where the host organisms were selected on the basis of a plasmid-mediated trait. The plasmids

demonstrated heterogeneity in molecular weight, ranging from 2to 80 X 10°. The majority of

the bacterial strains examined in this study carried multiple plasmid species.

As part of long-term studies of the ecol-

ogy of bacteria of the Vibrionaceae family,

we have isolated and characterized a large

number of such strains from the natural

environment (McNicol et a/. 1980a, 1980b).

In our investigations, we have focused on

the comparison of isolates from two widely

'Presented in the Symposium on Aeromonas and

Vibrio as Pathogens of Fish at the 29th Annual Meet-

ing of the Canadian Society of Microbiologists, Vic-

toria, B.C., June 11-15, 1979. Portions of this data

have been published (McNicol er al/., 1980a).

* Author to whom reprint requests should be

addressed.

different geographic regions. The Dhales-

wari-Meghna rivers, which feed the delta

mouth of the Ganges system in Bangla-

desh, comprise a tropical estuary subject to

high levels of domestic pollution. This de-

veloping region supports an extremely large

human population, within which Vibrio

cholerae and Vibrio parahaemolyticus dis-

ease are endemic. In comparison, the Chesa-

peake Bay is a temperate estuary which

contains relatively unpolluted waters, but

does have point sources of human and/or

industrial pollution. In this region, V. para-

haemolyticus is widely distributed but dis-

PLASMID CARRIAGE IN VIBRIONACEAE 61

ease caused by V. parahaemolyticus is un-

common and occurs only sporadically

(Colwell et al. 1972; Barker et al. 1974). V.

cholerae has not been associated with human

infection in Maryland in this century, al-

though the organism is indigenous to the

Chesapeake Bay (Colwell et a/. 1977).

Earlier reports from our laboratory have

described the toxigenicity, antibiotic re-

sistance, and plasmid carriage of strains of

Aeromonas hydrophila isolated from the

two estuarine regions (Kaper, J., H. Lock-

man, and R. R. Colwell. 1979. Ecology and

toxigenicity of Aeromonas hydrophila in

Chesapeake Bay. Abst. Ann. Mtg. Amer.

Soc. Microbiol. p. 188; McNicol ef al.

1980a,b). This communication will present

an overview of our work on plasmid car-

riage by the Vibrionaceae and other genera

of bacteria recovered from the same waters.

The results of a survey of estuarine bac-

terial strains isolated in Bangladesh are

summarized in Table 1. The BE series of

organisms were isolated on antibiotic-

amended Mueller-Hinton agar (Difco), and

the aeromonads have been described by

McNicol et a/. (1980a). Vibrionaceae strains

(BV) were isolated from Thiosulfate Cit-

rate Bile Salts (TCBS) agar (Difco) plates

inoculated with water samples collected in

Bangladesh during the same season as the

BE strains. Strain BV89 is closely related

to V. cholerae but does not fall within the V.

cholerae taxon. V. parahaemolyticus strain

A348 (Table 1) was isolated by Dr. S. P. De

from a clinical specimen from a human

case of gastroenteritis. Among the strains

of A. hydrophila, none of the 3 BV strains

and half of the BE set contained extra-

chromosomal DNA. Four of the nine plas-

mid-carrying organisms (44%) contained

two distinct plasmid species of different

molecular weights. The mass of the plas-

mids varied from 2 to 68 X 10° daltons, but

eight of the nine strains shared a high mo-

lecular weight plasmid species, average

MW of 20 X 10°, Figure 1.

The twelve BE strains of E. coliin Table |

represent a random sample (21%) of the 56

strains identified in this set. As observed

for Aeromonas, half of the E. coli strains

carried plasmids. However, more of the E.

coli strains carried multiple plasmid spe-

cies, five of the six isolates (83%). Three of

these E. coli contained two plasmids, one

strain had three, and one had five plasmids.

Table 1.—Frequency of Occurrence of Plasmids in Estuarine Strains Isolated from Water Samples Collected in

Bangladesh.

No. of

strains

No. of possessing

strains multiple

No. of possessing plasmid

strains plasmids species

Species examined (%)° (%)°

Aeromonas hydrophila

BES 18 9 (50%) 4 (44%)

BV 3 0 0

Escherichia coli BE* 12 6 (50%) 5 (83%)

Vibrio parahaemo-

lyticus* 1 (100%) 0

Vibrio cholerae BV 19 4 (21%) 3 (75%)

BV 89 | 1 (100%) 1 (100%)

TOTAL 54 21 (38%) 13 (60%)

* Percent of total number of strains found to contain at least one extrachromosomal DNA species by agarose

gel electrophoresis (McNicol ef al. 1980a).

Percent of plasmid-containing strains possessing more than one plasmid.

“Selected for antibiotic resistance phenotype.

“Clinical isolate.

62 L. A. McNICOL, T. BARKAY, M. J. VOLL, AND R. R. COLWELL

100

RI (68 x 10®)

P307 (62 x10®)

ANNA~< 10407 (60 x10®)

A~<H1I0407 (40 x 10®)

AXE RP4 (34 x10®)

S-a (25 x10§)

MOLECULAR WEIGHT (x1I0§)

0.1 10

RELATIVE MOBILITY

Fig. 1. Aeromonas hydrophila plasmid molecular

weights. Plot of log molecular weight versus log rela-

tive mobility. o, A. hydrophila plasmids from Bangla-

desh; ®@, A. hydrophila plasmids from the Chesapeake

Bay; A, reference plasmids used to construct the

curve. E. coli JS3 with plasmids R1, RP4, and S-a were

from E. M. Lederberg, Plasmid Reference Center. E.

coli K12 strains P307 and H10407 were from S.

Falkow.

The E. coli plasmids were more hetero-

genous in molecular weight than those

found in Aeromonas, Figure 2. More plas-

mids of low molecular weight were present

in E. coli Bour of fifteen (27%) EL. coli

plasmids had a mass of 10 megadaltons or

smaller, whereas only two of thirteen (15%)

Aeromonas plasmids were of a comparable

size. Furthermore, the E. coli strains exam-

ined in this study did not contain a com-

mon plasmid species.

100

MOLECULAR WEIGHT (xIO§)

| ;

0.1 0.5 10

RELATIVE MOBILITY

Fig. 2. Escherichia coli plasmid molecular weights.

The standard curve is identical to that of Figure 1. o,

E. coli plasmids from Bangladesh; ®, E. coli plasmids

from the Chesapeake Bay.

The single V. parahaemolyticus strain iso-

lated from a clinical specimen contained

extrachromosomal DNA, with a mass of

20 X 10° daltons. Twenty-one percent of

the V. cholerae strains from Bangladesh

possessed a plasmid and 3 of the 4 plasmid-

containing strains had more than one mo-

lecular species. The unidentified Vibrio,

strain BV89, demonstrated at least six

plasmid bands on gel electrophoresis, the

largest number observed for any of the en-

vironmental isolates examined. The mo-

lecular weights of the Vibrio plasmids were

heterogeneous, Figure 3.

Results of screening for plasmids in

Chesapeake Bay bacteria are presented in

PLASMID CARRIAGE IN VIBRIONACEAE 63

MOLECULAR WEIGHT (x1I08)

re)

RELATIVE MOBILITY

Fig. 3. Vibrio plasmid molecular weights. The

standard curve is identical to that of Figure 1. 0, V.

cholerae plasmids from Bangladesh; @, V. cholerae

plasmids from the Chesapeake Bay; 0, V. parahae-

molyticus plasmid from clinical strain A348.

Table 2. Twenty-one percent of the A. hy-

drophila strains and 18% of the V. cholerae

strains contained plasmids. Seventeen strains

of Pseudomonas spp. isolated from the

Chesapeake Bay were examined for plas-

mids, and 41% of these demonstrated the

presence of extrachromosomal DNA

(Table 2). One strain possessed two plas-

mid species and another strain had five.

The Pseudomonas plasmids were hetero-

geneous in molecular weight, ranging from

3 to 56 X 10°, with no common species

being present in those strains carrying

plasmids.

The molecular weights of the Chesa-

peake Bay plasmids were lower than those

of isolates from Bangladesh. Twenty-five

of the 39 Chesapeake Bay plasmids (64%)

had molecular weights less than 10 X 10°

whereas only 12 of 40 Bangladesh plasmids

(30%) were as small.

In addition to studies of environmental

isolates, the incidence of plasmids in a

small group of representative wild type Vi-

brio organisms was examined. From Table 3

it can be seen that one of seven strains pos-

sessed extrachromosomal DNA. Clearly,

extra-chromosomal DNA is common in

Vibrio spp.

The standard cleared lysate technique of

Guerry et al. (1973) was employed for the

plasmid screening reported here. Although

this procedure may underestimate the fre-

quency of high molecular weight DNA,

certain comparative conclusions can be

drawn.

Our major observation 1s that plasmid

DNA 1s common in strains of the Vibrio-

naceae bacteria found in the natural envi-

ronment. A total of ninety-one environ-

mental strains of A. hydrophila and Vibrio

spp. were examined. Fifty percent of the

Bangladesh A. hydrophila series selected for

antibiotic resistance contained extrachro-

mosomal DNA, and 20% of those Vibrio-

naceae Strains which were not selected fora

plasmid-mediated trait contained plasmid

species. Also, 47 other bacterial strains iso-

lated at the same time as the vibrio family

organisms were studied, and 49% (23 strains)

yielded plasmid DNA.

Aeromonas hydrophila strains isolated

from Chesapeake Bay were found to carry

plasmids in 21% of the strains examined, a

rate of occurrence remarkably similar to

that reported by Aoki et a/. (1977) for R

factors in strains of Aeromonas pathogenic

for fish (22%). The Chesapeake Bay strains,

in general, carried small, multiple plasmid

species, which were cryptic in function. The

presence of plasmids was not correlated

with any antibiotic resistance pattern and

none of the plasmid-carrying strains pro-

duced a toxin (McNicol er al. 1980a). In

contrast, A. hydrophila strains from Ban-

gladesh carried larger plasmids. A high mo-

lecular weight species found in Aeromonas

64 L. A. McNICOL, T. BARKAY, M. J. VOLL, AND R. R. COLWELL

Table 2.—Plasmid Carriage in Estuarine Bacterial Isolated from Chesapeake Bay.

No. of

strains

No. of possessing

strains multiple

No. of possessing plasmid

strains plasmids species

Species examined (%)* (%)°

Aeromonas hydrophila‘ 39 8 (21%) 5 (63%)

Escherichia coli® 18 10 (56%) 4 (40%)

Vibrio parahaemo-

lyticus® ] 0 0

Vibrio cholerae® 11 2 (18%) 1 (50%)

Pseudomonas sp.° 17 7 (41%) 2 (29%)

TOTAL 86 27 (31%) 12 (44%)

’ Percent of total number of strains found to contain at least one extrachromosomal DNA species.

® Percent of plasmid-containing strains possessing more than one plasmid.

and McNicol er a/. (1980a). More recently isolated strains of Aeromonas organisms are also included in the study

reported here. V. parahaemolyticus strain Cr 2/2 originated from the Chester River in Chesapeake Bay.

“ An L Series of E. coli strains were isolated using a modified Escherichia coli (EC) medium (Difco) without bile

salts and with raffinose substituted for lactose. An R series of E. co/i strains were isolated on EC medium with

raffinose replacing lactose. The EC series of E. co/i strains were isolated from EC broth incubated at 44°C

(Kaper, J.. H. Lockman, and R. R. Colwell, in manuscript). Plasmid carriage was not affected by isolation

procedure.

* Three strains of Ps. aeruginosa, a Ps. stutzeri, and 13 unidentified strains of Ps. sp. were isolated from Bailey’s

Creek and the James River (Orndorff and Colwell 1980).

Strains isolated from the environment was

similar to that observed for Aeromonas,

20% versus 28%, respectively. The fre-

quency of V. cholerae plasmids does not

appear to vary between the two geographic

locations, although more strains will need

to be examined to confirm this. The V.

(see Figure 3) was correlated with a strep-

tomycin/tetracycline resistance pattern

(McNicol et a/. 1980a). Cumberbatch et al.

(1979) recently reported a study of human

clinical isolates of A. Aydrophila, using the

Same procedure as that employed in this

study and found that 32% of their strains

contained plasmid DNA. They also ob-

served that the plasmids present in the clin-

ical isolates were not associated with pro-

duction of toxin.

The incidence of plasmids in V. cholerae

cholerae strains harbored small, multiple,

plasmid species of cryptic function.

V. cholerae strains have been reported to

carry Pand V fertility plasmids (Bhaskaran

1960; Bhaskaran and Sinha 1971), but their

Table 3.—Plasmid Carriage in Vibrio spp.

Molecular

Biotype No. of weight

Organism Strain (serotype) plasmids en

Vibrio cholerae ATCC 14033 El Tor (Inaba) 1 ~

ATCC 14035 cholerae (Ogawa) 0

569B cholerae (Inaba) 0

Vibrio marinus ATCC 15382 0

Vibrio parahaemolyticus ATCC 17749 alginolyticus 0

ATCC 17802 parahaemolyticus 0

3D38 parahaemolyticus 0

—— SS

PLASMID CARRIAGE IN VIBRIONACEAE 65

incidence in natural populations is un-

known. Rahal e7 a/. (1978) have remarked

that plasmid carriage in V. cholerae is ‘‘very

low”. Hedges e7 al. (1977) examined 1,156

worldwide V. cholerae isolates and found

that only 0.47% contained transmissible R

plasmids. Our observed cryptic plasmid

carriage rate is similar to the 26% incidence

of plasmids seen in clinical isolates from

India (Dastidar er a/. 1977), although these

are disparate samples.

One of two strains of V. parahaemolyti-

cus examined during this study contained a

20 X 10° MW plasmid. It is interesting to

note that this strain, A348, possesses an

04:K8 serotype (Dr. S. P. De, personal

communication). Previous studies of plas-

mid incidence in Vibrio parahaemolyticus

have shown that 04:K8 serotypes frequently

carried a 62 and a 23 X 10° molecular

weight plasmid (Guerry and Colwell, 1977;

Isaki, L. 1979. Studies on haemolytic activ-

ity, and extrachromosomal deoxyribonu-

cleic acid of Vibrio parahaemolyticus. Ph.D.

Thesis, University of Maryland, College

Park, Maryland; J. Newland, unpublished

results).

E. coli and Pseudomonas spp. are not

members of the family Vibrionaceae, but

Strains of these taxa were recovered from

the same sources as the other isolates re-

ported here. Strains of these two taxa

showed higher plasmid carriage rates than

A. hydrophila or V. cholerae. Our observed

rate for environmental strains of E. coli

(53%) is higher than a reported 39% plas-

mid carriage in E. coli clinical isolates

(Christiansen ef al. 1973).

In conclusion, plasmids are commonly

found in bacteria isolated from the natural

environment. The frequency of occurrence

of plasmids varied from 28% to 50% for

Strains of the four bacterial taxa examined

in this study. Overall, of the 137 environ-

mental strains surveyed, 46, or 34%, con-

tained at least one plasmid. Furthermore,

24 of the 46 (52%) contained multiple

plasmid species (Tables 1 and 2). The 78

plasmids observed in the survey of envi-

ronmental strains were heterogeneous in

molecular weight, ranging from 2 to 80 X

10°. Except where selected for a plasmid-

related phenotype, the plasmids found in

this study remain cryptic in function until

further work can be done. The extent to

which plasmids contribute to gene exchange

between bacterial strains in the natural envi-

ronment is not yet clear. Nevertheless, the

high frequency of occurrence of plasmids

in environmental strains suggests that

plasmids contribute to the genetic plastic-

ity of natural populations.

Acknowledgments

This work was funded by NSF Grant

DEB-77-14646 and NIH Grant R22-

AI-14242.

References Cited

Aoki, T., T. Kitao and T. Arai. 1977. R plasmids in fish

pathogens. Jn: S. Mitshuhashi, L. Rosival, and V.

Krcmery eds., Plasmids. Medical and Theoretical

Aspects. Springer Verlag, Berlin, p. 39-45.

Barker, W. H., R. E. Weaver, G. K. Morris and W. T.

Martin. 1974. Epidemiology of Vibrio parahaemoly-

ticus infection in humans. Jn: D. Schlessinger, ed.,

Microbiology 1974. Amer. Soc. Microbiol., Be-

thesda, MD, p. 257-262.

Bhaskaran, K. 1960. Recombination of characters be-

tween mutant stocks of Vibrio cholerae, strain 162.

J. Gen. Microbiol. 23: 47-54.

Bhaskaran, K. and V. B. Sinha. 1971. Transmissible

plasmid factors and fertility inhibition in Vibrio

cholerae. J. Gen. Microbiol. 69: 89-97.

Christiansen, C., G. Christiansen, A Leth Bak and A.

Stenderup. 1973. Extrachromosomal deoxyribonu-

cleic acid in different enterobacteria. J. Bacteriol.

114: 367-377.

Colwell, R. R., T. Kaneko and T. Staley. 1972. Vibrio

parahaemolyticus—estuarine bacterium resident in

Chesapeake Bay. Proc. Mar. Tech. Soc., Food-

Drugs from the Sea, p. 87-94.

Colwell, R. R., J. Kaper and S. W. Joseph. 1977. Vibrio

cholerae, Vibrio parahaemolyticus, and other vi-

brios: Occurrence and distribution in Chesapeake

Bay. Science 198: 394-396.

Cumberbatch, N., M. J. Gurwith, C. Langston, R. B.

Sack and J. L. Brunton. 1979. Cytotoxic enterotoxin

produced by Aeromonas hydrophila: Relationship

of toxigenic isolates to diarrheal disease. Inf. Imm.

23: 829-837.

Dastidar, S. G., R. Podda, R. Kumar and A. N. Chak-

rabarty. 1977. Incidence and elimination of R plas-

mids in Vibrio cholerae. Antimicrob. Agts. Che-

mother. //: 1079-1080.

Guerry, P. and R. R. Colwell. 1977. Isolation of cryptic

plasmid deoxyribonucleic acid from Kanagawa-

positive strains of Vibrio parahaemolyticus. Inf.

Imm. 16: 328-334.

Guerry, P., D. J. LeBlanc and S. Falkow. 1973. Gen-

eral method for the isolation of plasmid deoxyribo-

nucleic acid. J. Bacteriol. 1/6: 1064-1066.

Hedges, R. W., J. L. Vialard, N. J. Pearson and F.

O’Grady. 1977. R plasmids from Asian strains of

Vibrio cholerae. Antimicrob. Agts. Chemother. JJ:

585-588.

Kaper, J., H. Lockman, R. R. Colwell and S. W. Joseph.

1979. Ecology, serology, and enterotoxin produc-

tion of Vibrio cholerae in Chesapeake Bay. Appl.

Environ. Microbiol. 37: 91-103.

MeNicol, L. A., K. M.S. Aziz, I. Hug, J. B. Kaper, H. A.

Lockman, E. F. Remmers, W. M. Spira, M. J. Voll

and R. R. Colwell. 1980a. Isolation of drug-resistant

Journal of the Washington Academy of Sciences,

Volume 72, Number 2, Pages 66-78, June 1982

Aeromonas hydrophila from aquatic environments.

Antimicrob. Agts. Chemother. 17: 477-483.

McNicol, L. A., J. B. Kaper, H. A. Lockman, E. F.

Remmers, W. M. Spira, M. J. Voll and R. R. Colwell.

1980b. R-factor carriage in a group F Vibrio iso-

lated from Bangladesh. Antimicrob. Agts. Che-

mother. /7: 512-515.

Orndorff, S. A. and R. R. Colwell. 1980. Distribution

and characterization of kepone-resistant bacteria in

the aquatic environment. Appl. Environ. Micro-

biol. 39: 611-622.

Rahal, K., G. Gerbaud and D. H. Bousnchaud. 1978.

Stability of R plasmids belonging to different in-

compatibility groups in Vibrio cholerae ‘“‘Eltor”.

Ann. Microbiol. 129: 409-414.

Operational Concepts for Automated

Transportation Systems

Alan J. Pue

Applied Physics Laboratory, Johns Hopkins University,

Laurel, Maryland 20707

ABSTRACT

Automated transportation systems are typified by driverless vehicles operating on exclusive

guideway networks. To achieve cost-effective and energy saving operation requires the devel-

opment of control policies that maximize the utilization of system capacity. This paper sum-

marizes several solution approaches to the various problems encountered when designing a

high capacity automated transportation system. In particular, concepts for vehicle longitudi-

nal control, the merging of vehicles at junctions, the control of vehicles within stations, and

the routing of vehicles through a network are discussed.

1. Introduction

Automated Guideway Transit (AGT)

systems are a relatively recent development

in transportation technology, intended to

serve urban regions and major activity cen-

ters such as airports, shopping districts,

and universities ({1]-[5]). They are charac-

terized by automatically controlled vehi-

cles (driverless) of small-to-moderate size,

operating on dedicated guideway networks

that have either on-line or off-line stations.

AGT systems can vary in complexity ac-

cording to vehicle size, system capacity,

network density, and whether routes are

fixed, selected at the time of trip request, or

dynamically selected as vehicles progress

through the network.

An urban wide AGT network may be

composed of hundreds of miles of con-

nected guideways, joining up to a hundred

stations (typically off-line), and contains

AUTOMATED TRANSPORT SYSTEMS 67

hundreds of vehicles. Consequently, the

problem of efficiently distributing and

routing vehicles between many origins and

destinations is of considerable importance

in terms of customer acceptance, cost effec-

tive operation, and energy use. Moreover,

this general problem encompasses several

ancillary problems involving overall net-

work operation, namely, the regulation of

vehicles’ velocities and spacings, the merg-

ing of vehicles at junctions, and the man-

agement of vehicles within stations. In all

cases, the objectives are to achieve min-

imum trip times and to maximize utiliza-

tion of available line capacity while assur-

ing safe and comfortable operation.

This paper reviews the above mentioned

control problems associated with a large

automated transportation system and de-

scribes several approaches that have been

proposed in recent years. First, Section 2

provides a general description of an auto-

mated transportation system, the network

layout and customer demand. We then

summarize various approaches to the spe-

cific problems of vehicle control, merging,

station operation and routing. Throughout

the paper, discussion is limited to the nor-

mal operational problems of an AGT sys-

tem. Questions related to operation during

emergency conditions, as well as, reliability

and maintainability are not addressed, but

discussion of these issues may be found in

the references ([{1]-[5]).

2. Network Description

Since the late 1960’s, numerous types of

AGT systems have been proposed, varying

in scale and complexity. To date, the opera-

tional systems may be characterized as rel-

atively small, involving several miles of

guideway, shuttle type service, and net-

work configurations that contain either

Straight connections or simple loops. Ex-

amples of systems [6] that are currently op-

erating in the United States are at Morgan-

town, West Virginia, Tampa and Atlanta

Airports, and the Airtrans system at Dal-

las-Fort Worth Airport. Other systems op-

erating or planned ([5],[6]) are in West

Germany, France, and Japan. Typical ve-

hicle sizes vary from 3 to 5 meters in length,

1 to 4kg in weight, and depending upon the

particular system, vehicle capacities range

from 12 to 70 passengers.

This paper describes more complex sys-

tems intended for urban deployment where

line capacity requirements are high and

vehicles are required to operate at head-

ways significantly below those of conven-

tional transit systems, dictating headways

in the range of 3 to 5 seconds. As a result,

substantial technical innovation, hardware

development, and performance evaluations

are needed before such systems can be

constructed, presenting many challenging

problems to system designers. A general

description of the operational policies and

physical characteristics of AGT systems is

given below.

The operational policies of an automated

transit system may be defined as the method

for scheduling trips and the technique em-

ployed for routing vehicles through the

network, including the dispatching of empty

vehicles to stations where demand exceeds

supply. The selected policies are largely de-

termined by the size of the network and the

level of demand. The policy options may be

broadly categorized as either dynamic

(demand actuated) scheduling or fixed

scheduling and dynamic routing (respon-

sive to current demands and congestion) or

fixed routing. The dynamic policies would

most likely be associated with large systems

(high density) while the fixed policies are

most appropriate for small systems (low

density) although demand actuation has

been used in very simple systems (e.g.,

elevators).

In a single party system a party enters a

station, makes its destination known and

requests service. An empty vehicle is pro-

vided and the destination 1s encoded on the

vehicle which then departs into main line

traffic. The vehicle is appropriately routed

through the network to the destination sta-

tion where passengers deboard. A system

68 ALAN J. PUE

of this type leads to the following problems

[7]:

1. Selecting a vehicle to respond to a re-

quest for service at a station, e.g., as-

signing a vehicle from station storage,

calling in a passing empty vehicle from

the main line, or waiting for the next in-

bound vehicle to arrive.

2. Disposing of an empty vehicle at a sta-

tion after its passengers have departed,

e.g., placing vehicle in storage, assign-

ing vehicles to another passenger, or

dispatching the empty vehicle to another

station.

3. Selecting a route for each vehicle (occu-

pied or empty) once it has been

dispatched.

Multiple party occupancy of each vehicle

also creates the following questions:

1. Once a vehicle has been assigned to one

party, will it wait for other passengers

and for how long?

2. Will service to other passengers be lim-

ited to those going to the same destina-

tion or to certain destinations that are

‘“‘along the same route’’?

3. Will the vehicle be permitted to stop at

stations along the way to pick up other

passengers or possibly diverted for pick-

ups?

4. Should transfers be permitted?

Several approaches to these problems

are summarized in Section 6. Below, net-

work geometry and trip demand, the major

constraints in the vehicle management

problem, are characterized.

The physical characteristics of the AGT

network can be described in terms of nodal

elements that are classified according to

five basic types [7], and links that represent

the route segments between node pairs.

The first nodal element is a station, points

where passengers board and deboard vehi-

cles, and are characterized by the number

of berths, berth configuration, length of the

deceleration lane, and length of the egress

lane into mainline traffic. Intersection nodes

are points where two lanes merge or a sin-

gle lane diverges while a yard designates the

location of the entrance to empty vehicle

storage facilities and maintenance area.

The final type of nodal element is a link

characteristic breakpoint where certain link

characteristics such as speed, grade, and

radius of curvature change. The guideway

links of the network are assumed to be con-

structed within the public right-of-way,

1.e., urban streets and highway systems and

are specified according to directionality,

speed, curvature, and grade. A link is con-

sidered to be one-way although two links of

opposite direction may parallel one another.

Thus, network layout consists of station

locations, station configuration, link con-

nectivity and geometry. These parameters

are typically dictated by known and pro-

jected population distribution in conjunc-

tion with cost-benefit studies ({1]-[3]). Asa

result, the design of the physical network is

performed separately and possibly deter-

mines the system operational design. Al-

though as noted in Section 5, policies for

moving vehicles within stations are closely

related to the station configuration.

The nature of trip requests may be speci-

fied by level of demand, pattern of demand,

arrival frequency, and diurnal variation of

demand. Network demand description

therefore has several characteristics [6].

First, the number of trips between various

stations varies throughout the day. For ex-

ample, morning trips are generally towards

employment centers while midday demand

may be between shopping areas. There is

also variation in total trip demand as a

function of the time of day. For downtown

urban systems there is heavy morning and

evening use while at activity centers such as

airports and universities, the patterns are

quite different.

Variations also occur in the rate at which

passengers arrive at a particular station.

For some stations, arrivals may be by car

or walking and by bus at other stations.

Thus, in the latter case large groups will ar-

rive with large interarrival times and in the

former case demand will be characterized

by small groups with short interarrival

times. Finally, the size of a party making a

trip is of importance, particularly for sys-

AUTOMATED TRANSPORT SYSTEMS 69

tems where several parties use the same

vehicle and the possibility arises that par-

ties may be split between vehicles.

Consequently, the demand for service at

each station will vary in a random manner

throughout the day and from day to day. In

distributing occupied and empty vehicles,

the ability to predict demand 1s likely to be

advantageous. For this purpose, it is useful

to note that demand variations will likely

fall into one of three categories: (1) recog-

nizable average trends in demand such as

morning and evening rush hours, (2) short-

term random fluctuations in demand char-

acterized by commonly occuring brief surges

or lulls and, (3) an occasional unantici-

pated long-term change due to, for exam-

ple, sporting or cultural events. The first

two categories may be termed routine while

the last may be known in advance but the

precise time and extent is often not known.

In an attempt to predict demand changes

the following kinds of data would be avail-

able for use: (1) cumulative past experience

such as daily trip records over past weeks

or months and, (2) current states of stations

in the system, that is, queues of waiting

passengers and trip requests received.

To quantify demand for analysis, many

investigators define a demand matrix, where

each element, dij, is equal to the average

demand from station i to station j in trip

requests per unit time. Passenger interarri-

val time is then exponential distribution

with mean, dj. In addition, dj may be a

function of time to account for temporal

variations in demand.

We now describe the various control

problems associated with an automated

transit system.

3. Vehicle Control

Vehicle control requires the regulation of

longitudinal and lateral motions of the ve-

hicle during normal operation. Longitudi-

nal control involves the adjustment of ve-

hicle velocities and spacings within a string

of several vehicles while lateral control [8]

is needed to center the vehicle on the

guideway and to control lateral motions of

the vehicle during turning and switching

maneuvers. Because longitudinal control is

critical to safe operation and achievement

of high capacity, most developers of AGT

systems have focused attention on the lon-

gitudinal control problem.

The longitudinal control laws that have

received the most study may be classified

according to one of two approaches. The

first, usually termed point following, as-

signs each vehicle to a moving cell, the cells

being propagated along the guideway net-

work at predetermined velocities and spac-

ings [9],[10]. In this case, propulsion com-

mands are generated so that each vehicle

maintains its location within its assigned

cell. The second approach, termed vehicle-

following, is a control scheme which allows

communication between successive vehi-

cles such that the motion of a given vehicle

is controlled in accordance with the motion

of its neighbors. In particular, the strategy

of greatest practical interest is where a ve-

hicle follows the motion of its immediate

predecessor, and is specifically considered

below. Note that a critical feature of a ve-

hicle-follower control system design is string

stability, that is, disturbances must propa-

gate through a string of vehicles with de-

creasing magnitude.

For a vehicle-following system, a typical

situation is depicted in Figure 1. A wayside

computer determines the position of each

vehicle within its jurisdiction (i.e., guide-

way section) by position markers, such as

induction coils imbedded in the guideway.

Vehicle velocities are also down-linked

through communication lines to the way-

side computer. A vehicle-follower control

law, generally based on the spacing and

relative velocity between a preceding and

trailing vehicle, generates acceleration

commands to the trailing vehicle in order

to maintain a specified spacing policy,

S(vt), typically a function of the trailing

vehicle velocity, vt. When the spacing be-

tween vehicles is large, a trailing vehicle

70 ALAN J. PUE

Uplink Downlink

computer

Fig. 1. Vehicle-follower control.

Operates in an open-loop velocity com-

mand mode where velocity commands are

transmitted from the wayside computer to

the vehicle, independent of any preceding

vehicle actions. As vehicle spacing falls

below some threshold value, the vehicle

must account for the presence of a preced-

ing vehicle and a transition to the spacing

policy, S(vt), is initiated. Under these cir-

cumstances the vehicle enters what is termed

a closed loop regulation mode.

Several spacing policies, S(vt), have been

suggested [10]. A constant spacing policy,

employed in point following, maintains a

constant nose to nose spacing between ve-

hicles while a constant time headway main-

tains a constant time between vehicles pass-

ing a fixed point on the guideway. A third

spacing policy defines a K-factor as the

ratio of vehicle separation (tail to nose) to

emergency stopping distance. Thus, this

represents a “‘brickwall” stopping criterion

where if a vehicle is operating with a K-fac-

tor greater than one, the trailing vehicle

may come to a stop without collision if the

preceding vehicle were to become a brick-

wall. Both constant headway and constant

K-factor policies may be used in either point-

following or vehicle-following systems. A

final policy designed for vehicle-following

systems is a kinematic spacing ([{11],[12])

scheme where vehicle spacing is based on a

relative stopping distance. This policy arises

in a vehicle following system because a

trailing vehicle cannot anticipate future de-

celeration maneuvers of a preceding vehi-

cle (due to downstream congestion). Thus,

a trailing vehicle must maintain a spacing

such that if the preceding vehicle suddenly

brakes on limits to a stop, the trailing vehi-

cle can respond without collision.

Many studies of vehicle-following have

focused on the regulation of vehicle speeds

and spacings during perturbations about

nominal values as determined by one of the

first three policies described above [10].

However, in general, a control capability

must exist to perform transient maneuvers

such as closing upon a slower moving vehi-

cle, switching from an open-loop velocity

command mode to a closed-loop regula-

tion mode, merging vehicles both on the

main line and from stations, and generat-

ing gaps in vehicle flows during such merg-

ing operations. A control law using a

kinematic spacing policy is suitable for all

of these maneuvers as reported in [12].

How a kinematic constraint determines

a safe vehicle spacing may be illustrated

with the aid of Figure 2, where a trailing

vehicle (T) is traveling at speed v; and over-

taking a preceding vehicle (P) traveling at

speed vp < vt. It is assumed there is a min-

imum allowable guideway speed, Vmin = 0,

and that it is possible for vehicle P to slow to

this speed on service acceleration and jerk

limits at any time as a result of downstream

congestion. This maneuver requires a dis-

tance of d? and atime of t? and is completed

at point Prin Figure 2. Similarly, for vehicle

T, there is a distance dy > d? and time

tr > tP required to slow to speed Vmin, Where

the inequalities hold since vt > Vp. We then

define the kinematic constraint as a min-

imum spacing such that the trailing vehicle

may slow to minimum velocity at a vehicle

length, L, behind the preceding vehicle

without exceeding service jerk or accelera-

tion limits. This minimum spacing is given by

Smin = di Fz ad — Vmin (tr ae y se (1)

Fig. 2. Minimum spacing as required by kinematics.

AUTOMATED TRANSPORT SYSTEMS 71

where d; and ty are nonlinear functions of

the vehicle states, service acceleration limit

(as), service jerk limit (js), and the maxi-

mum velocity (Vmax).

The quantity Smin may be considered a

spacing constraint for transient maneuvers

and it increases in significance as nominal

headway decreases.

To illustrate, the quantity, Smin/Vmax 1S

plotted in Figure 3 as a function of initial

preceding vehicle velocity, vp with maximum

line speed, Vmax, aS a parameter. It is as-

sumed that a trailing vehicle is traveling at

a constant speed Vmax and the preceding

vehicle decelerates from a constant vp to

zero velocity on jerk and acceleration limits

of 2.0 m/s’ and 1.5 m/s’, respectively. (A

slightly worse case can be found for an ac-

celerating trailing vehicle, but it does not

significantly alter the results.) Figure 3

shows the headway corresponding to a

trailing vehicle velocity of Vmax and a spac-

ing, Smin. For example, in a system with a

maximum line speed of 15 m/s, safe opera-

tion can be assured only for headways

greater than 5.4s when the kinematic con-

Straint is not taken into account. On the

other hand if the kinematic constraint Is

explicitly included in the vehicle-follower

control law [11], then this requirement can

be overcome for arbitrarily small head-

ways. Note, however, other factors such as

communication delays, variations in brak-

ing characteristics, measurement errors,

grade changes, and wind must be consid-

ered when determining a safe allowable

minimum headway. Several investigators

have proposed that a practical lower limit

on headway for vehicle-follower systems is

0.5 s [10].

4. Merge Control

Many merging techniques have been

Studied in terms of the effect upon vehicle

delay at a junction given various types of

input flows. The algorithms which have

been studied may be classified into one of

two general categories. First, synchronous

max

Minimum headway satisfying kinematic constraint (Si in/¥max- 3)

0 5 10 15 20 15

Initial preceding vehicle velocity, Vp (m/s)

Fig. 3. Required minimum headway to satisfy

kinematic constraint.

methods are those where all vehicles are al-

located to slots which move along the

guideway so that no conflicts occur in the

system, thus requiring a central controller.

Also note these methods would be asso-

ciated with a point-following system. A sec-

ond approach is to employ local control

where at a given distance upstream from

the merge junction, a merge controller

commands vehicle actions to assure a suc-

cessful merge of two vehicle strings. This

may be accomplished in two ways: (1)

movements of vehicles are controlled by

the relative movements of other vehicles on

the guideway (vehicle-following); or (2)

vehicles are assigned slots so that merging

occurs without conflict. The local strategy

may be broken down into a number of

schemes depending on how priorities are

assigned. The results of studies comparing

various schemes are now discussed.

A first come, first serve algorithm is sim-

ulated by Brown [13] where individual ve-

hicles through a merge are simulated using

a linear vehicle-follower control at a 4.0

second headway. This approach appears to

be a good candidate for a merging algo-

rithm because of its simplicity for imple-

mentation and the fact that it tends to treat

all vehicles equally. This latter result has

been shown by several investigators in the

point following case discussed below. To

implement first-come, first-serve a parallel

Parallel data region

Fig. 4. Merging control.

data (Figure 4) region upstream from the

merge junction 1s selected such that a merge

controller has knowledge of all vehicles

within this region. Each vehicle entering

the region is then assigned to follow the last

vehicle to have entered the region whether

it is a vehicle immediately in front or a ve-

hicle on the alternate guideway. For exam-

ple, in Figure 4 vehicle 5 is assigned to fol-

low vehicle 4 while vehicle 6, not yet having

entered the data region is also following

vehicle 4. When vehicle 6 enters the parallel

data region it will be assigned to follow ve-

hicle 5. Within the parallel data region ve-

hicle control proceeds as though the two

lanes are collapsed into a single lane with

overlapping vehicles. The vehicle follower

control law automatically resolves merge

conflicts by creating appropriate gaps on

each lane. The length of the parallel data

region must be greater than a stopping dis-

tance to assure no collisions can occur at

the junction.

There is more literature available con-

cerning merge control using point-follow-

ing, most likely because of its analytic

tractability. The most basic formulation of

the problem is given by Whitney [14] where

incoming streams of vehicles are described

by a binary word with a 1 representing an

occupied slot and a 0 representing an

empty slot. A terminal cost is computed by

squaring the net number of slots moved by

each vehicle, and summing over all vehicles

moved. In [15], it is proved that a first-

come, first-serve scheme minimizes this

cost.

The most detailed investigation of merg-

ing for a point following system is con-

tained in the thesis by Godfrey [16] where

he evaluates six merging strategies based

on queueing models and simulation. He

ALAN J. PUE

also found the first-come, first-serve

strategy to be consistently the best strategy

by measuring merge effectiveness in terms

of tail probabilities (1.e., Prob [Delay > N

slots]).

In [15], Sakasita simulates Godfrey’s six

merging strategies, compares resulting queue

lengths of each merge lane separately, and

comes to the same conclusion concerning

the desirability of first-come, first-serve.

Some specific algorithms and problems

which arise at intersections are considered

in [17] and [18].

5. Station Control

The impact of station design upon net-

work performance depends upon the sta-

tion guideway configuration and the sta-

tion operating policy ([{19],[20],[21]). The

configuration is the actual physical layout

of the station including deceleration and

accelerating ramps, number of berths, and

layout of the docks. The station operating

policies involve the management of vehi-

cles once they have entered the station, the

unloading and loading of passengers, and

the merging of vehicles onto the mainline.

It can be assumed that a station design is

performed to efficiently handle expected

demand at that location; the vehicle man-

agement design variables are the handling

of empty vehicles and the dispatching of

occupied vehicles.

To aid in station modeling we note the

events which occur [20] when a vehicle en-

ters a station: (1) time to switch off guide-

way, (2) deceleration time, (3) move from

entrance queue to dock, (4) open doors, (5)

unload time, (6) load time, (7) close doors,

(8) dispatch queueing time, (9) move from

dock to acceleration ramp, (10) acceleration

time, (11) switch onto main guideway.

The above sequence may be altered

when, for example, a vehicle unloads and

then advances to the first unoccupied berth

to await passengers for loading. The load-

ing and unloading times are random varia-

bles and have been modeled [20] using a

AUTOMATED TRANSPORT SYSTEMS 73

log-normal distribution. However, this may

also be represented as a deterministic min-

imum dwell time which is normally ex-

pected to be sufficient for all loading and

unloading of passengers.

The most important aspect concerning

stations is the possibility that there is no

room for an arriving vehicle and it must be

rejected. The vehicle would then be routed

to the nearest station or looped around

until a space is free. Alternatively, ‘‘back-

ups” onto the main guideway can be al-

lowed until space is free.

In a demand actuated AGT system an

essential feature of vehicle management is

to provide empty vehicles at stations to

serve trip requests [22]. Two questions

which naturally arise are where to obtain

the empty vehicles and how to efficiently

dispatch the vehicle to the station. The first

question deals with the arrangement of

vehicle storage facilities while the latter is

associated with vehicle routing. The basic

objectives are to minimize passenger wait-

ing time and to minimize the fraction of

line capacity used by empty vehicles.

The possibilities for vehicle storage facil-

ities are as follows [22]: (1) each station can

have its own storage, (2) network divided

into districts, each district having a com-

mon storage, and (3) moving storage with

vehicles circulating through a district ready

to be called upon. Many factors including

economic cost and real estate availability

enter into the selection of an appropriate

arrangement for vehicle storage facilities.

This selection process would be aided by

comparing network performance in terms

of figures of merit (discussed below) for a

given vehicle management strategy and

various facility configurations.

The major vehicle control problems within

a Station are precision stop at queue posi-

tions, advancement between queue posi-

tions, and merging of vehicles from dock

berths into mainline traffic. The final prob-

lem is of special interest because it is desir-

able (due to guideway construction costs)

to minimize the length of the acceleration

ramp without severely disturbing mainline

traffic flow. That is, there exists a trade-off

between length of the acceleration ramp

and the degree to which mainline vehicles

must decelerate to allow egressing vehicles

to enter the mainline guideway. It has been

found [23] that a kinematic constraint,

analogous to the vehicle-following con-

straint described in Section 3, can be used

to accommodate these conflicting problems.

6. Routing Control

In the previous sections we have de-

scribed the local control problems of vehi-

cle control, merging, and station control.

We now consider the combined problem of

network control, illustrated by the network

section in Figure 5. Passengers arrive at the

station loading dock and board a vehicle

that has either just unloaded arriving pas-

sengers, has been withdrawn from vehicle

storage, or has been sent as an empty from

another station. The vehicle merges into

mainline traffic and is controlled by a series

of local wayside computers that communi-

cate with the vehicle through sensors 1m-

bedded in the guideway. Each wayside

computer has jurisdiction over a region of

the guideway network and passes control

to adjacent wayside computers as the vehi-

cle passes through jurisdictions.

One approach to the problem of routing

the vehicle through the network is a reser-

vation system (or synchronous method)

that selects an entire unimpeded route for

each vehicle before it departs an origin sta-

tion. This selection could be a static func-

tion of the destination or determined at the

time of trip request. A second approach,

possibly more appropriate to large sys-

tems, dispatches vehicles on demand at

origin stations with the local wayside

computers at intersections solving the

merging and routing problems on-line.

Algorithms for this latter approach would

be structured by the following situation. As

each vehicle enters an intersection, the ve-

hicle is assigned to a particular outgoing

link based on its origin, destination and the

state of the network. The vehicle is then as-

74 ALAN J. PUE

signed to follow a preceding vehicle and

controlled according to specified merging

and spacing policies. It is often assumed

that these latter problems (merging and

spacing regulation) are of a local nature

and can be solved independently of the

routing strategy. As discussed earlier, an

additional concern is the distribution of

empty vehicles from stations where de-

mand is low to stations where demand is

high and should therefore be included as

part of the routing problem formulation.

From a practicality standpoint, there is

considerable advantage in determining

routing solution by local control. Suppose

routing strategy were directed by a central

control node in the network. Such a node

would periodically obtain information con-

cerning traffic congestion from all other

nodes in the network and solve the current

routing problem. Fora network with many

nodes and vehicles, the communication re-

quirements would be costly, and serious

problems could develop when either the

control node or any communication link

fails. Consequently, from an economic cost

and reliability viewpoint it is beneficial to

design a distributed routing control algo-

rithm. For example, each merge junction

may communicate only with its nearest

neighbors to determine the dynamic rout-

ing of vehicles throughout the network, the

Guideway

platform

objective being the avoidance of vehicle

bunching and excessive delay.

Thus, in a large network it may be desir-

able to make routing decisions on-line and

locally [24]. That is, as each vehicle ap-

proaches a diverge link the local wayside

computer makes the decision as to what

ongoing link the vehicle should travel. In

addition, the local wayside computer makes

this decision based on information ob-

tained from the adjacent wayside compu-

ters. Thus, using this approach we utilize

the communication links already in place

for the vehicle control problem and no ad-

ditional links are required.

It might be suggested that because we

have a completely automated system, we

only need information at stations. That is,

all behavior within the system is determin-

istic and all control policies are initiated in

response to trip demands. However, any

practical system will have occasional fail-

ures, requiring a re-start within the net-

work and therefore, an implementation

such as that in Figure 5 is needed. Other-

wise, a network control center would re-

quire communication lines to every guide-

way link and vehicle in the network, a

costly approach for a large network.

The above discussion implies a multi-

level structure to the control of an auto-

mated transportation network where the

computer

Fig. 5. Typical network section.

AUTOMATED TRANSPORT SYSTEMS 75

levels of control are interacting. The effects

of local vehicle control, merge control, and

distributed routing must be assessed in

terms of network performance. To quan-

tify this performance several figures of

merit may be employed as we now describe.

In describing figures of merit for a trans-

portation network we first note the distinc-

tion between a system optimizing index and

a user optimizing index. System optimiza-

tion involves assigning a cost to each sec-

tion based on system variables such as flow

or delay and then summing over all sec-

tions in the guideway network. User opti-

mization assigns costs associated with origin

to destination pairs.

System optimizing performance measures

would include:

1. Total Delay—average all link delays

where delay is defined as either (1) time

to traverse a link, or (2) extra time to

traverse link due to congestion, and is

weighted by the number of vehicles ex-

periencing delay.

2. Total Service—compute the product of

the number of vehicles on each link and

the average velocity on that link; sum

over all links and integrate over a time

interval T; the result is the total distance

moved by vehicles in the network dur-

ing time interval T.

3. Energy Use—vehicle accelerations,

number of empty vehicles traveling on

network, fleet size (1.e., total number of

vehicles required to fulfill a given level

of demand).

User optimizing performance measures

would include:

1. Average origin to destination delay where

delay is the time beyond the time re-

quired for an unimpeded trip (including

wait time).

Maximum origin to destination delay.

Dependability—variance in origin to

destination delay at a given time of day.

To complete a description of the control

problem, we conclude with a discussion of

the information that would be available

through sensor measurement.

The fundamental information needed

for control is position and velocity. A mul-

titude of systems have been conceived to

measure these quantities, but at this time

the most technically feasible type of sensor

appears to be one which is imbedded in the

guideway to measure vehicle position. Asa

vehicle passes over a sensor (e.g., inductive

loop) this event is recorded to determine

vehicle position with respect to a reference.

The spacing of these sensors thus deter-

mines the accuracy of the position meas-

urement. Vehicle velocity may be measured

by a tachometer onboard the vehicle or es-

timated from the position measurements.

Using these measurements, vehicle control

computation is performed either at way-

side, onboard or with an appropriate allo-

cation between wayside and vehicle [25].

The necessary communication links are

generally through the guideway by induc-

tive coupling.

Other possible sensing elements would

be: (1) measure demand at stations, (2)

measure time delay to traverse a link by

clock aboard vehicle and, (3) measure

number of vehicles on a link with up-down

counter.

We now consider solution approaches to

the routing problem.

The fundamental routing problem may

be viewed as a nonlinear multi-commodity

flow problem, a problem that has received

wide attention in the nonlinear program-

ming literature, although primarily for the

Static case. Recently, more research has

been devoted to the decomposition and ef-

ficient solution of large problems for both

the static and dynamic situations. In par-

ticular, applications have centered on traf-

fic, power, and computer communication

networks.

For the specific application of automated

transit systems, the major component de-

scribing the behavior and characteristics of

a transportation network is vehicle flow.

This flow is determined by the operational

policies, intervehicle spacing control scheme,

station geometries, and link connectivities

that form the overall network design. Thus,

the accurate modeling of vehicle flow and

the attendent description of network dy-

namics Is likely to be critical for a success-

76 ALAN J. PUE

ful routing control law design although rel-

atively little work has been accomplished in

this area for automated transit systems.

Research into developing traffic flow

models has concentrated on the problem of

automobile traffic in congested urban streets

and freeways. Many models are based on

the original continuum model of Lighthill

and Whitham [26] derived by using the

analogy to continuity of fluid flow and

forms the basis for many traffic flow mod-

els found in the literature. This fundamen-

tal flow model has the general form

Xi = [qi — gi+1]/Di (2)

where x; 1S vehicle density on section 1

(veh/m), qi is the flow in veh/sec into link

i and D; is the length of link i(m).

Fortunately, the modeling of an auto-

mated network does not contain the com-

plicating factors of an urban traffic net-

work such as passing, parked vehicles and

human behavior. The interactions between

vehicles in an automated system may be

described according to a well-defined lon-

gitudinal control law ([{10],[11]) that is

based on accurate measurements of vehicle

position and velocity. Thus, one could eas-

ily derive a microscopic traffic flow model

based on discrete vehicle behavior. How-

ever, such a model would be of extremely

high order and would lead to a complex

and costly control algorithm. As a matter

of judgement and practicality, it is likely

that some aggregate model using macro-

scopic variables such as density will prove

to be sufficient for the successful applica-

tion of on-line control laws. Such models

analogous to traffic flow models, have been

derived in [27] for the vehicle-follower lon-

gitudinal control scheme. These models

have the same form as the fundamental

flow model given in (2) but the vehicle

flows qi are computed as nonlinear func-

tions of the densities in adjacent links or

Xi = [Gil Xi. Xi me Geri(Xe Xie IZ Di 23)

where Xj-1, Xi+1 are the upstream and down-

stream link densities, respectively. This

basic link model is extended to include

merges, diverges, and stations.

Most analyses of automated transit net-

works has been exclusively limited to point-

following systems or a model which makes

no distinction between point-following and

vehicle-following. In [28], a steady-state

point-following model is assumed and flows

are assigned to each link based on minimi-

zation of a performance index given by the _

sum of fleet size cost, travel time cost, and

wait time cost. Tong and Morse [29] also

study a point-following system but divide

the system into cells of consecutive slots.

The vehicle management strategy exam-

ined employs a fixed routing policy where

each vehicle is assigned a particular route

before leaving a station. The vehicle is then

assigned to occupy the first cell leaving the

station which has an open slot fora vehicle

traveling the designated route. It is shown

in [29] that the system may be properly

phased so that as two distinct cells pass

through a merge, they merge together, slot

by slot, and may be locally merged without

conflict under certain explicit conditions.

In [30], a point-following network is ana-

lyzed from the viewpoint of queueing the-

ory for both local merge control and syn-

chronous control methods. The two methods

are compared by calculating expected delay

where for the local merge case, a station is

modeled to bea M/M/1 queue anda merge

is a M/M/1—Mgq queue where Mgq is the

maximum queue length. For the synchro-

nous case, the probability of securing a res-

ervation is computed. In [31], a model

based on conservation of flow is used to de-

termine a suboptimal control (i.e., routing

variables) which minimizes the maximum

value of the sum, over all links, of the densi-

ties squared. Roesler, et al. [7] have evalu-

ated a variety of static linear programming

and heuristic routing strategies for a spe-

cific network configuration.

The approach taken in [24] is to formu-

late an optimization problem by develop-

ing accurate dynamic models of vehicle

flow [27] for the vehicle-follower case and

formulating a performance index based on

AUTOMATED TRANSPORT SYSTEMS 77

total, time averaged, travel time in the net-

work. Thus, if 7(xj) is the travel time on link

i the total cost is

YS YS TWirilxi(k) }xi(k) te (4)

k=1 ie&

where K is the number of time steps, & is

the set of network links and tg is the final

time. The optimization problem is to min-

imize (4) subject to the dynamic flow con-

straints of the network as defined by (3).

This represents the first attempt at combin-

ing local vehicle dynamics with network

control in a systematic way, contrasting

with previous studies that have either con-

sidered only local vehicle dynamics while

ignoring the network problem or vice-versa.

In [24], the problem is solved by apply-

ing duality theory to decouple the overall

dynamics into vehicle type (origin destina-

tion pair) network flow constraints that are

decoupled in vehicle type state and control

but coupled through the interconnection

variables of total link density and total di-

verge link control. The resulting structure

of the subnetwork dynamics is then ex-

ploited to allow a distributed control com-

putation where each node in the network

only needs to communicate with neighbor-

ing nodes to optimize the dual function ob-

jective. An upper level coordinating con-

trol, localized to each link, seeks to satisfy

the interconnection constraint that the sum

of individual vehicle type densities is equal

to the total vehicle density on the link. Asa

result, all control computations can be per-

formed in acompletely decentralized manner

where information exchange only occurs

between physically adjacent wayside con-

trol computers. Convergence of the algo-

rithm has been shown theoretically and by

computer simulation. The principal benefit

of this approach is that communication

links needed for solution of the routing

problem correspond to communication links

in place for vehicle spacing and velocity reg-

ulation. Moreover, as noted earlier this

type of computational structure also has

the advantage of producing a significant

Savings in Communication costs when com-

pared to a central control node imple-

mentation.

7. Conclusions

We have reviewed the major control

problems that exist when designing an au-

tomated transportation system. In this space,

it is possible to briefly summarize only a

few of the accomplishments that have oc-

curred over the past decade. However, per-

haps the most important aspect is the devel-

opment of algorithms that allow safe short

headway operation. In particular, the in-

clusion of a kinematic constraint in the

longitudinal control law design permits

greater efficiency in utilizing guideway ca-

pacity at critical junctions in the system

where congestion occurs (i.e., merges and

stations). Other important areas of research

have included network flows and their ef-

fects upon network performance, station

design and its impact upon throughput,

station operations, lateral control system

designs, and communications technology.

Results of this work have produced a

greater understanding of how a transporta-

tion network should operate.

Finally, it should be noted that although

the implementation of automated trans-

portation networks is in its infancy, the

technology and techniques that have been

developed have wider applicability. For

example, problems in air-traffic control,

urban traffic networks, communication net-

works, and power networks have many

analogies to the automated transportation

problem.

References Cited

1. Anderson, J. E. et al. (eds.). 1972. ‘‘Personal

Rapid Transit’, University of Minnesota, Min-

neapolis, Minn.

2. Anderson, J. E. (ed.). 1974. ‘Personal Rapid

Transit II’, University of Minnesota, Minneapo-

lis, Minn.

3. Gary, D. A. (ed.). 1976. ‘‘Personal Rapid Transit

III’, University of Minnesota, Minneapolis, Minn.

y3.

14.

lp);

ALAN J. PUE

MacKinnon, D. (ed.). February 1979. Special Is-

sue on Automated Transportation, JEEE Trans.

Veh. Tech., Vol. VT-28, No. 1.

. MacKinnon, D. February 1975. ““High Capacity

Personal Rapid Transit System Developments”’,

IEEE Trans. on Veh. Tech., VT-24, No. 1, pp. 8-14.

. Proceedings International Conference on Trans-

portation Electronics, IEEE Vehicular Technology

Society Annual Conference, Dearborn, Michi-

gan, September 1980.

. Roesler, W. J..M. C. Waddel, B. M. Ford and E. A.

Davis. August 1971. ‘“‘Operating Strategies for

Demand-Actuated ACGV Systems”, Voll and II,

Johns Hopkins University Applied Physics Lab-

oratory Report TPR-019.

. Hisonger, R. L. and D. N. Wormley. February

1979. “Dynamic Performance of Automated

Guideway Transit Vehicles With Dual-Axle Steer-

ing’, JEEE Trans. Veh. Tech., VT-28, No. 1.

. Brown, S. J., Jr. May 1977. ‘‘Point-Follower Au-

tomatic Vehicle Control: A Generic Analysis’,

Johns Hopkins University Applied Physics Lab-

oratory Report CP 057/TPR-025.

. Garrard, W. L., R. J. Caudill, A. L. Kornhauser, D.

MacKinnon and S. J. Brown. 1978. ‘State of the

Art of Longitudinal Control of Automated Guide-

way Transit Vehicles’, High Speed Ground Trans.

Vol ig No. 2, 1978:

. Pue, A. J. December 1978. ‘‘A State-Constrained

Approach to Vehicle-Follower Control for Short-

Headway AGT Systems”, ASME Journal of Dy-

namic Systems, Measurement and Control.

. Pue, A. J.. H. Y. Chiu and S. J. Brown. August

1979. ‘““Operational Concepts and Implementa-

tion Techniques for Vehicle-Follower Control of

AGT Systems”, Johns Hopkins University Ap-

plied Physics Laboratory Report CP 074/TPR-041.

Brown, S. J. 1975. ‘‘Traffic Merging Under Vehicle-

Follower Control’’, 1975 International Confer-

ence on Personal Rapid Transit, Denver, Col.

Whitney, D. E. June 1972. ‘tA Finite State Ap-

proach to Vehicle Merging’, ASME Journal of

Dynamic Systems, Measurement and Control’, pp.

147-151.

Sakasita, M. 1975. ““An Analysis of Merge Con-

trol for the Automated Transportation System”,

1975 International Conference on Personal Rapid

Transit, Denver, Col.

. Godfrey, M. B. June 1968. ‘“‘Merging in Auto-

mated Transportation Systems’, Sc. D. Thesis,

Department of Mechanical Engineering, M.I.T.

. Caudill, R. J. and J. M. Youngblood. 1976. ‘‘Inter-

section Merge Control in Automated Transporta-

tion Systems’’, Trans. Res., Vol. 10, pp. 17-24.

. Brown, S. J. May 1977. ‘“‘Point-Follower Auto-

20.

21.

22°

23.

24.

22:

26.

28.

20:

30.

Si:

matic Vehicle Control: A Generic Analysis”’,

Johns Hopkins University Applied Physics Lab-

oratory Report CP 057/TPR-025.

. Liopiros, K. J. February 1974. “PRT Station Op-

erational Strategies and Capacities’’, 1973 Inter-

national Conference on Personal Rapid Transit”,

Minneapolis, Minnesota.

Sirbu, M. A., Jr. February 1974. ‘‘Station Config-

uration, Network Operating Strategy and Station

Performance”, 1973 International Conference on

Personal Rapid Transit, Minneapolis, Minnesota.

Johnson, R. E., H. T. Walker and W. A. Wild. June

1976. “Analysis and Simulation of Automated

Vehicle Stations”, 1975 International Conference

on Personal Rapid Transit, Denver, Col.

Waddell, M. C., M. B. Williams and B. M. Ford.

May 1974. ‘Deposition of Empty Vehicles in a

Personal Rapid Transportation System”, Johns

Hopkins University Applied Physics Laboratory

Report TPR-28.

Chiu, H. Y. February 1979. “‘A State-Contrained

Vehicle-Follower Approach to the Station-Egress

Problem”, JEEE Trans. on Veh. Tech., Vol.-28,

No. I.

Pue, A. J. 1981. ‘‘Vehicle Distribution and Rout-

ing in a Large Automated Transportation Net-

work”’, Ph.D Dissertation, Dept. of Electrical

Engineering, University of Maryland.

Pue, A. J. February 1979. ‘Implementation

Trade-Offs for a Short-Headway Vehicle-Follower

Automated Transit System”, JEEE Trans. on Veh.

Tech., Vol. VT-28, No. 1.

Lighthill, M. J. and G. B. Whitham. May 1955.

“On Kinematic Waves-II, A Theory of Traffic

Flow on Long Crowded Roads”’, Proc. Roy. Soc.

(London) 229A, pp. 317-345.

Pue, A. J. Vol. 16B, No. 2, pp. 125-142, 1982.

‘“Macroscopic Traffic Models for Vehicle-Fol-

lower Automated Transportation Systems”’, Trans.

Res., to appear.

Gershwin, S. June 1976. “‘A Design Tool and a

Routing and Scheduling Technique for Personal

Rapid Transit Systems’’, 1975 International Con-

ference on Personal Rapid Transit, Denver, Col.

Tong, Y. M. and A. S. Morse. “‘Decentralized Con-

trol of Automated Transportation Systems via

Cellular-Flow Coordination’, 1976 Conference

on Decision and Control.

Kornhauser, A. L. and P. McEvaddy. 1975. “A

Quantitative Analysis of Synchronous vs. Quasi-

Synchronous Network Operations of Automated

Transit Systems”’, Trans. Res., Vol. 9, pp. 241-248.

Cunningham, E. P. 1975. ‘‘A Dynamic Model for

Vehicle Routing in a Two-Way PRT Network”,

Trans. Res., Vol. 9, No. 6.

DELEGATES TO THE WASHINGTON ACADEMY OF SCIENCES,

REPRESENTING THE LOCAL AFFILIATED SOCIETIES

Cr MW RCMINGIONN .y.c 4 vce cts ewecesectawecweseweceanvcessbeccepives James F. Goff

En GIECY OF WOSNINBION. 12... cc ccc cs ccecanecasccuusesccceuseveuse Ruth H. Landman

ES William Ronald Heyer

TR Sas Scalia ign clul Odi atn 0's wwe a Paw ake Wik web eed owed dw weet Anne Brown

er GT WASMINGION ... ecu cewek cee veaveseedueusuacuewesseneeess Margaret Collins

i he iaio kids Wau id KM OR Ew be Re aieee aa ak aales ¢ T. Dale Stewart

MSN DION axis cis cans wed aicaw ees sb enenenwadaeseseuaeee James V. O'Connor

cm tne Pistrict Of Columbia... .0. 0.0 ccc ccc ccenenscnccacewecenas Charles E. Townsend

end Sa aig aad iw: Ne AW UmiNC Sea el Uh dca we way aia ew mle od ie aici tld anne Paul H. Oehser

NEN 6c cers oa 6 God SNe dn 69 db Wie died Male a kel aran nd aera ermlens Conrad B. Link

EES Boyd W. Post

EE EAIIEINIGOTS sw cc ese ee de ec ee ve ween enna we dntadeuenseses George Abraham

er eeeeired: and Electronics Engineers ..........ccccccsecncnccsccecdenees George Abraham

eee PCC ANICAl ENGINCETS ....... 6.2.2.2 ceca cece sscctncsenavessccssecees Michael Chi

i neICEY C1 WASMINPION .. 0.5.2.2. 2 ccc cece ue cs enesecceeesscseees Robert S. Isenstein

EE RMEREUIMIETORIOIOPY 2.0... ce ec bane cet ace ann naeescencsasvaaens Lloyd G. Herman

IONE ATY FNPINGETS .. . 5 os cs cee ces ince ecasaussek eee uae H. P. Demuth

EE RU ESTIDITICETS 7. i 5 oe ni anise win ce se ce Kelas eee e ba ele es viele ca biele Wallace J. Cohen

mn emnermmiental Biology and Medicine..........5.:.0.cccenescccssneseces Cyrus R. Creveling

REDS ace a. hee Ww leet 4 odho. be, Ska me Kas we War weve ww lee» Charles G. Interrante

irre er Dental Research .........-..0scur ec secee se wecesncduenes William R. Cotton

eemenit btistiute Of Acronautics and Astronautics.............0.000ceceeeeecnaes Richard P. Hallion

EE IEICE, fire cco ein os cle wa lasen cece weeeck Souacdaes ce alee A. James Wagner

EES SETS 1 eee Jack R. Plimmer

IE OUEREEIETRIOTE os so oss x oie bas ales whe dias Be ews baniele ome sos whe wire Richard K. Cook

PRA cae ye wo Sg ee A SIR ow oh Dia oo Shelia late SVefane Blac wave w are Sal etmmane Dick Duffey

ERE RISES 2 pe S. ok aye 8, wl a! wn ik aro 3m WE ju ine: Welt erase ORES wie ah A. D. Berneking

I IEE 5 sesso alan whe ao dled iets ewe ciara e wit baw nein epee disled are Edwin R. Fuller, Jr.

I Fe ooo ie i 0, 00 2/0 e a a eid. weieciw Ale & bins sd w Soe wR OS Alayne A. Adams

ee INOEEIEE ICHETICE CID 2. 2. 5 cia niece en te eee ees aaaeewencees Deborah J. Warner

UENCE PIVSICS TCACHEDS. «52.0. = ois-cjeis ed's oo ta oe need nee Ce eae wher Peggy A. Dixon

So SS) c So asas ais. 0 ak a dn G4 in @ Pee wre wwe aie wie hw leas oa he George J. Simonis

Perret Fiat PHYSIGIOPISIS ... 2... 2 ec cee ne eee ene cswennsnece Walter Shropshire, Jr.

ECAC MOCSCATEN COUNCH. ociic ce ck es acs ne views Damas ecasaeaace ees John G. Honig

EEE eS cs eee ee ee eer Jewel B. Barlow

American Institute of Mining, Metallurgical

IE Ea fo Sak aie casa Soe A eG MS al oso ae 8 ww A Garrett R. Hyde

SE ESR MEE eee ce Ie Pe ed Benson J. Simon

rear eALIEAEY CHIMUR LICH. «aig x co x cose 30/6’ ¢ aac ® u'a’Zlp oc eiclere nis ale(era eines ona dlenees Patrick Hayes

RAE ee ek gw aiaididia wa dk'cwne eh 6 de ks ole an seed Miloslav Racheigl, Jr.

AERIS rte chez Site cie aieie dia aid dou Sou al wa. bie sie awl Gow See eins H. N. Reynolds

rt EArt TECHNICAL GTOMD ...6.05c ona sa ctewvnccssauwardeicunebsunases Paul G. Campbell

ERICH ODICAL SEICIEDY (cic cn vice a duce vine seus dpwcaveceusacinesaus Howard E. Waterworth

TMserAl SVStCIIS RESCATCH «. 0.005. a scene cc dente svecceccvncenas Ronald W. Manderscheid

SE Pl Ge Gg ues tu ig. a Kwioh da’ vie wo i aiid wo vim'niai amend wm WO Stanley Deutsch

Ee ha Sian els Ginla,e wk x A ess wie coo Sre ws ede eWeek mn lm Irwin M. Alperin

in tor Science, SEchnOlopy ANG INNOVATION 6: .......006cescccscarccesnesteccen Ralph I. Cole

Delegates continue in office until new selections are made by the representative societies.

Washington Academy of Sciences 2nd Class Postage Paid

1101 N. Highland St.

Arlington, Va. 22201 and-.additi

Return Requested with Form 3579

REHDER, |

HARALD A. (DR)

5620 OGDEN ROAD Ge

BETHESDA, MD hate

20816). 2 (WAS-BB4. 00740)

at Arlington, Va.

ey (rete

ah ae

Dey:

ot Pyhe te

AR ai

% a a

Ra usinae

las

14 VOLUME 72

W317 Number 3

ec) _/ SH of the September, 1982

WASHINGTON

ACADEMY.-SCIENCES

ISSN 0043-0439

Issued Quarterly

at Washington, D.C.

LIBRARIES A

CONTENTS

Articles:

JOSEPH H. NEALE, KENNETH A. BLANK, RONALD J. PLISHKA and

PAUL M. SWEETNAM: The Opioid and Other Neuropeptides: Diverse

Elements in the Complex Equation of Brain Function................ 79

meeourR HABER: Personal Scientific Computers 2.0.6 2c os 5. cegew cans 87

RICHARD T. KORITZER, D.D.S.,. MALCOLM M. MARTIN, M.D.,

ARLINE L. A. MARTIN, M.D., LUCILE E. ST. HOYME, D. PHIL. and

JOHN J. CANARY, M.D.: Caries and Juvenile Onset Diabetes: A Meta-

ME IUERSEV IES © oe pac! 6 oxi Aa os Kees a Saw oR ekg BANE IS Re bhade ous 6 mia eine 98

SHERMAN ROSS: The Scientific Awards of the Academy: 1982 ..........

Washington Academy of Sciences

Founded in 1898

EXECUTIVE COMMITTEE

President

James F. Goff

President-Elect

Jean K. Boek

Secretary

John J. O’Hare

Treasurer

Ronald Manderscheid

Managers-at-Large

Jo-Anne Jackson

Grover C. Sherlin

Alfred Weissler

Elise A. Brown

James Wagner

Ralph I. Cole

BOARD OF MANAGERS

All delegates of affiliated

Societies (see Cover 3)

EDITORS

Irving Gray

Joseph Neale

ACADEMY OFFICE

1101 N. Highland St.

Arlington, Va. 22201

Telephone: (703) 276-9254

The Journal

This journal, the official organ of the Washington

Academy of Sciences, publishes historical articles, critical

reviews, and scholarly scientific articles; proceedings of

meetings of the Academy and its Board of Managers; and

other items of interest to Academy members. The Journal

appears four times a year (March, June, September, and

December)—the September issue contains a directory of

the Academy membership.

Subscription Rates

Members, fellows, and patrons in good standing receive

the Journal without charge. Subscriptions are available on

a calendar year basis only, payable in advance. Payment

must be made in U.S. currency at the following rates:

US: and Canada... >..'.- $19.00

ot | eee eee gear ae 22.00

Single Copy Price ..... 7.50

UHURU Symposium—2 issues............

Delivery by Air....... 2.00

Single-copy price for Vol. 66, No. 1 (March, 1976) is $7.50.

Back Issues

Obtainable from the Academy office (address at bottom

of opposite column): Proceedings: Vols. 1-13 (1898-1910)

Index: To Vols. 1-13 of the Proceedings and Vols. 1-40 of

the Journal Journal: Back issues, volumes, and sets (Vols.

1-62, 1911-1972) and all current issues.

Claims for Missing Numbers

Claims will not be allowed if received more than 60 days

after date of mailing plus time normally required for

postal delivery and claim. No claims will be allowed

because of failure to notify the Academy of a change in

address.

Change of Address

Address changes should be sent promptly to the Academy

office. Such notification should show both old and new

addresses and zip number.

Published quarterly in March, June, September, and December of each year by the Washingten

Academy of Sciences, 1101 N. Highland St., Arlington, Va. 22201. Second class postage paid at

Arlington, Va. and additional mailing offices.

Journal of the Washington Academy of Sciences,

Volume 72, Number 3, Pages 79-87, September 1982

The Opioid and Other Neuropeptides: Diverse Elements

in the Complex Equation of Brain Function

Joseph H. Neale, Kenneth A. Blank*, Ronald J. Plishka and Paul M. Sweetnam

Department of Biology and *Department of Obstetrics and Gynecology,

Georgetown University, Washington, D.C. 20057

SUMMARY

As many research disciplines have pursued the intricacies of brain function, a rigorous

framework has been established within which this tissue can be better understood. One of

these disciplines, the relatively new field of peptide neurobiology, has rapidly evolved during

the past decade. The distinction between peptide hormones, gastric peptides, sensory peptides

and traditional neurotransmitter molecules has been obscured by the observation that many

of these molecules are synthesized and released by nerve cells in the brain and spinal cord to

act upon distinct receptors on the surface of nearby cells. As peptides, however, they possess

characteristics which provide for considerable variation. This endows the neuropeptides, such

as the opioid group, with the potential to play diverse roles in the process of chemical com-

munication which is central to the acquisition, transmission, analysis and storage of informa-

tion by the nervous system.

Introduction

How is that a kilogram of tissue, the

human brain, executes the information

processing required for complex activities

from homeostatic regulation to the deriva-

tion of the Heisenberg uncertainty princi-

ple? The answer to this resides in under-

standing the biological events associated

with the reception, transmission, integra-

tion and storage of information by nerve

cells. Some aspects of the overall puzzle are

at this moment quite unclear. Surely, some

others are beyond our immediate imagina-

tion. Between the obvious and the obtuse,

many disciplines of neuroscience are effec-

tively replacing hypothesis with fact and

exposing provocative questions, as well as

solutions. Neuropeptide research is such a

field of study. It has established its credibil-

ity during the past decade with evidence of

the contribution of peptides to the com-

plexity of brain function, particularly that

of chemical communication. In this article,

a brief overview of nervous system function

preceeds a review of selected aspects of

peptide neurobiology which are both scien-

tifically intriguing and relevant to the un-

derstanding of the nervous system’s most

dynamic expressions.

Nervous System Design

The brain has often been correctly viewed

as analogous to a complex circuit. In man,

for instance, there are hundreds of billions

of individual nerve cells in this circuit. Each

of these cells has the potential to exchange

direct communication with hundreds of

80 NEALE, BLANK, PLISHKA, and SWEETNAM

thousands of other cells. These points of

contact and direct communication between

nerve cells are termed synapses and they

are the physical substrate through which

direct electrical or chemical signals are ex-

changed. Thus, the brain has enormous po-

tential for convergence and divergence in

information processing. The effective use

of these circuits requires that the hundreds

of billions of neurons be linked by rather

precise patterns of connections. While this

is clearly the case in the mature nervous

system, the internal and external cues which

specify these patterns and direct their de-

velopment, remain to be determined. Be-

yond the large number of cells involved in

this circuit and the patterns of their connec-

tions, the brain functions through the move-

ment of information over long distances

within individual nerve cells, as well as

through the execution of integrating deci-

sions regarding the termination or contin-

uance of signals from one cell to the next.

In terms of the circuit analogy, the decision

process is similar to the operation of an

off/on switch which integrates incoming

positive and negative signals and acts to

send the signals further when a threshold

value of input is reached.

To appreciate the role of peptides in ner-

vous system function, one must focus upon

the events which occur as signals are con-

veyed between nerve cells. This process,

known as synaptic transmission, may in-

volve the movement of charged ions di-

rectly from one cell to another through

specialized junctions. Alternatively, the

communication may involve the release of

a chemical neurotransmitter from the end

of the cytoplasmic process of one cell and

the diffusion of this chemical over a short

distance to the next nerve cell in the circuit.

Macromolecules, termed receptors, in the

membrane of the cell receiving the signal

may then recognize the chemical and bind it

in the same manner that an enzyme reacts

with its substrate or an antibody binds an

antigen. The transient binding of chemical

transmitter to receptor causes a change in

the permeability of the cell membrane to

specific ions, principally sodium, potassium

or chloride. With this change in permeabil-

ity of the receptor regulated ion channels,

an excitatory or inhibitory signal is trans-

mitted from one nerve cell to another. The

potential for complex information process-

ing is enhanced at this point with the ability

of different nerve cells to deliver different

chemicals signals and of the cells receiving

the chemical transmission to couple their

receptors to several types of ion channels.

The channels may differ not only with re-

spect to their ion selectivity, but also in the

rate of channel opening and in the interval

during which the channel will remain open.

Thus it is that one of the critical variables of

synaptic transmission which enhances the

capacity of the nervous system to process

information is the diversity of chemical

transmitters and the receptors which me-

diate their action.

It has long been appreciated that the

process of chemical neurotransmission is

similar to the interaction of hormones with

their target cells. Indeed, the major distinc-

tions between hormonal regulation and ner-

vous system control has simply been the dis-

tance over which the chemical must act and

the structure of the chemicals which are used

in these processes. This latter distinction

appears to be less clear, today. That is,

nerve cells and the endocrine system appear

to use similar peptides as transcellular mes-

sengers. This is most readily observed at

the interface between the brain and the en-

drocrine system, the hypothalamic-pituitary

axis. Information regarding homeostatic

requirements of the organism are channeled

from many circuits into the nerve cells of the

hypothalamus. These cells send long cyto-

plasmic projections out from the bottom of

the brain and into the pituitary. Some of

these hypothalamic nerve cells then release

chemical signals, such as vasopressin and

oxytocin, directly into the bloodstream

where they will eventually find target tissues

and mediate the hormonal regulation of

water resorption, mammary gland secretion

and uterine contraction. Alternatively, the

hypothalamic cell projections may release

their chemicals into the portal blood of the

median eminence for passage into the ante-

NEUROPEPTIDES AND BRAIN FUNCTION 81

rior pituitary, where the chemicals will reg-

ulate the release of trophic hormones, such

as adrenocortcotrophic hormone (ACTH),

luteinizing hormone (LH), follicle stimu-

lating hormone (FSH) and growth hor-

mone. The molecules from hypothalamic

nerve cells which regulate the anterior pi-

tuitary have been termed releasing factors

and they share with oxytocin and vasopres-

sin an important characteristic. They are

peptides which are secreted by nerve cells in

a manner that is indistinguishable from the

release of neurotransmitters elsewhere in

the nervous system.

Peptides as Neurotransmitters

Several factors converged during the

past ten years to provide impetus to the

study of neuropeptides. There were a num-

ber of distinguished and insightful research-

ers in this field for many years prior to this

and several neuropeptides had been thor-

oughly characterized. Initially, there was

little agreement as to the significance of

these molecules in central nervous system

function. Beyond the limited recognition of

these few peptides, however, there was, an

increasingly rigorous literature detailing

the structure and function of both hypo-

thalamic-pituitary and gastro-intestinal pep-

tides, as well as the development of analyti-

cal and immunochemical reagents for the

study of such molecules. During this inter-

val, many years of study of the mechanism

of opiate drug action reached fruition with

two discoveries. Several research groups’

demonstrated nerve cell receptors which

recognized and bound opiate alkaloids ina

stereospecific and saturable manner. Others

succeeded in the purification of endogenous

brain peptides* ° which could mimic the ac-

tions of opiate alkaloids and also react with

the opiate receptors in nerve cell mem-

branes. These observations quickly lead to

the conclusion that the opioid peptides play

a role in chemical transmission which is

similar to that of the traditionally recognized

neurotransmitters, such as acetycholine and

norepinephrine. Stimulated in part by these

achievements, there rapidly accumulated a

body of data which indicated that many

other peptides, previously studied for their

role in endocrine control, there regulation

of the gastrointestinal tract, or their pres-

ence in sensory nerve cells, were distributed

throughout the central nervous system,

often in anatomically or functionally dis-

crete pathways.”° With this research has

come the realization that during billions of

years of evolution the diversification of

peptides and their receptors has promoted

their participation in a broad spectrum of

neurophysiologic functions.

The Opioid Peptides

The opioid peptides provide an example

of neuropeptide diversity within a family of

structurally related molecules (Table 1).

The first members of this family to be iso-

lated from brain contained the amino acid

sequence, argenine-glycine-glycine-phenyl-

alanine, with either leucine or methionine

at the carboxy terminus.” They were termed,

leucine- and methionine-enkephalin. Not

surprisingly, these endogenous brain op-

iates have been localized within nerve cell

circuits'’’'? which had long been known to

mediate pain perception. The combined

application of immunochemistry and high

performance liquid chromatography has

resulted in the characterization of larger

opioid peptides. There appear to be two

classes based upon either the methionine-

or leucine-enkephalin amino terminus.

While it has not been unequivocally estab-

lished that each of these peptides is released

from nerve cells during synaptic communi-

cation, it seems likely that many of the

molecules listed in Table 1 will be shown to

serve neurotransmitter-like functions. Each

of the larger peptides also has the potential

to act as the precursor in the peptidase me-

diated synthesis of smaller opioid peptides.

Because of the propensity of conventional

endopeptidase enzymes for action at basic

amino acid residues, the argenine, argenine-

lysine and argenine-argenine positions seem

to be important locations for enzyme me-

82 NEALE, BLANK, PLISHKA, and SWEETNAM

Table I.—Opioid Peptides: Diversity from a Common Amino Terminus.

Peptide

Methionine-enkephalin

Met-enk heptapeptide

Met-enk octapeptide

alpha-endorphin

Thr(16)

gamma-endorphin

Thr-Leu(17)

beta-endorphin

FMRFamide

Leucine-Enkephalin

Dynorphin(1-8)

Dynorphin

Dynorphin B(1-13)

Dynorphin B(1-29)

beta-neo-endorphin

alpha-neo-endorphin

Post-translational modifications of opioid peptides.

Amino Acid Sequence (No. of Amino Acids)

Tyr-Gly-Gly-Phe-Met(5)

Tyr-Gly-Gly-Phe-Met-Arg-Phe(7)

Tyr-Gly-Gly-Phe-Met-Arg-Gly-Leu(8)

Tyr-Gly-Gly-Phe-Met-Thr-Ser-Glu-Lys-Ser-Gln-Thr-Pro-Leu-Val-

Tyr-Gly-Gly-Phe-Met-Thr-Ser-Glu-Lys-Ser-Gln-Tyr-Pro-Leu-Val-

Thr-Leu///(31)

Phe-Met-Arg-Phe(NH2)(4)

Tyr-Gly-Gly-Phy-Leu(5)

Tyr-Gly-Gly-Phe-Leu-Arg-Arg-Ile(8)

Tyr-Gly-Gly-Phe-Leu-Arg-Arg-Ile-Arg-Pro-Lys-Leu-Lys-Trp-

Asp-Asn-GIn(17)

Tyr-Gly-Gly-Phe-Leu-Arg-Arg-Gly-Phe-Lys- Val-Val-Thr(13)

Tyr-Gly-Gly-Phe-Leu-Arg-Arg-Gly-Phe-Lys-Val-Val-Thr-Arg-Ser-

Gln-Glu///(29)

Tyr-Gly-Gly-Phe-Leu-Arg-Lys-Tyr-Pro(9)

Tyr-Gly-Gly-Phy-Leu-Arg-Lys-Tyr-Pro-Lys(10)

a) Endopeptidase cleavage, particularly at basic amino acids pairs, such as Arg-Arg and Arg-Lys.

b) Aminopeptidase enkephalinase removal of tyrosine residue producing inactivation.

c) O-sulfation of amino terminal tyrosine residue producing inactivation.

d) N-acetylation of beta-endorphin producing inactivation and possible storage.

diated cleavage during the synthesis of the

enkephalins.

The biological synthesis of peptides with

more than two or three amino acids appears

to be achieved by a ribosomally dependent

process with post-translational splitting of

a polypeptide to produce the active product.

Extraordinarily rapid progress has been

made in determining the nature of the ma-

cromolecular precursors which are respon-

sible for the production of the opioid pep-

tides. Purified pro-opioid polypeptides have

been sequenced directly,’? and more re-

cently, three different precursors (Table 2)

which together are capable of producing all

of the known opioid peptides have been

characterized using complementary DNA

and gene cloning technology.'*'*'®'”'® The

endorphins which are concentrated in the

hypothalamic-pituitary axis are synthesized

from a precursor polypeptide which con-

tains at least seven peptide hormones, in-

cluding ACTH, melanocyte stimulating

hormone, lipotropin and beta-endorphin.

A gene containing precursor sequences

suitable for the synthesis of several copies

of methionine-enkephalin and one copy of

leucine-enkephalin has also been iso-

lated.'*'®'’ Neither the larger endorphin

sequences nor the dynorphin/neoendorphin

sequences are found in this gene. These lat-

ter two peptides are synthesized with dy-

norphin B in a third gene isolated from

brain.'® While confirming the heterogeneity

of the opioid peptide sequences in the ner-

vous system, these results stimulate consid-

erable speculation as to the role of the addi-

tional amino acid sequences found in these

cloned gene products. There appear to be

several peptides specified by the pro-opioid

genes for which a function exists only in

the imagination of today’s neuroscientists.

If the opioid neuropeptides are serving a

function in the chemical transmission be-

tween nerve cells, as they clearly appear to

be, what is their role and why have the large

number of related molecules in this family

evolved? The answers to these questions

remain speculative. On the surface, the evo-

lution of peptide molecules for both hor-

SS ee ee

a ae eS es

—————

NEUROPEPTIDES AND BRAIN FUNCTION 83

Table 2.—Cloned cDNA Genes Containing Opioid Peptide Sequences.

Precursor Size

Opioid Sequence

A. Pro-opio-melano-corpin 163AA

B. Preproenkephalin 400AA

C. Prodynorphin T76AA

alpha, beta, and gamma-endorphin, also:

Adrenocorticotropic Hormone, Melanocyte

Stimulating Hormone, Lipotropin.

131 AA sequence, function unknown.

4 copies of Methionine-enkephalin

1 copy of Leucine-enkephalin

1 copy of Methionine-enkephalin

heptapeptide

1 copy of Methionine-enkephalin

octapeptide

Dynorphin

alpha-and beta-neo-endorphin

Dynorphin B-(1-29)

The amino acid sequences of the opioid precursors were deduced from the nucleotide sequence of cloned

DNA complementary to the mRNA purified from bovine pituitary intermediate lobe (A), bovine adrenal tissue

(B) and porcine hypothalamus (C). The preproenkephalin cDNA from adrenal tissue has since been shown to be

complementary to brain mRNA by hybridization.

monal and neuronal communication ap-

pears rather straightforward. Among the

limited molecular pool available to the first

organisms on this planet were the amino

acids. Amino acids such as glycine, glutamic

acid, and gamma-aminobutyric acid, as well

as modified amino acids, such as dopamine,

norepinephrine and serotonin have been

established as potent chemical signals in

the nervous system. It would be a short

jump in the scheme of evolution to increase

the diversity of neurotransmission events

by coupling several amino acids together

for use by either the endocrine or nervous

system. Carnosine, an enzymatically syn-

thesized dipeptide, provides an interesting

example of a transmitter-like molecule

which bridges the amino acids and larger

peptides. The evolutionary force which

sustained this drive for neuropeptide syn-

thesis and diversification may well have

been the demands for increasingly more

sophisticated information processing. That

is, in agreement with one theory of neuro-

peptide action, these molecules in some

Cases may function not to communicate di-

rectly between cells, but rather to modulate

the ongoing communication process. For

instance, when glutamic acid is applied

from a micropipet onto a spinal cord nerve

cell in a manner analogous to the chemical

release during synaptic transmission, this

amino acid binds to its receptor and pro-

duces excitation by causing sodium channels

to open and allowing sodium ions to enter

the cell along their concentration gradient.

Methionine-enkephalin applied to the same

cell has no detectable effect as a signal pro-

ducing chemical. However, when glutamic

acid and the opioid peptide are applied to

the nerve cell simultaneously, the enke-

phalin causes a reduction in the ability of

the glutamate to excite the cell.'’ The peptide

modulates the effect of the amino acid

transmitter. Because cells in the nervous

system may receive hundreds of nearly si-

multaneous chemical signals from other

cells, the combination of chemical trans-

mission and chemical modulation may be

an effective mechanism for fine tuning the

information processing in these circuits.

Analysis of the physiologic role of specific

classes of peptides, such as the opioid

molecules, has substantiated the concept

that nerve cell circuits which mediate spe-

cific functions, such as pain perception,

anxiety, satiety or even the sense of well-

being, may use a rather restricted reper-

toire of molecules for chemical transmis-

sion. One role of the opiates in the circuit

which transmits information about painful

stimulus from the body surface to the brain

84 NEALE, BLANK, PLISHKA, and SWEETNAM

appears to be the modulation of perception

of the intensity of such stimuli. This modu-

lation of pain perception would permit the

detection of very low levels of stimulus

while, under different circumstances, re-

ducing the more intense pain signals to

bring them within effective limits.

There are a number of human conditions

in which pain may be a rather predictable

feature. Aggresive behavior by another or-

ganism is one. The “‘fight or flight”’ response

mediated by the release of adrenalin from

the adrenal medulla is one response to the

perception of impending danger. It would

seem quite useful to add to the alerting re-

sponse of adrenalin an increase in the pain

threshold and, indeed, this appears to be

the case. High concentrations of enkephal-

ins have been found in the adrenal medulla”

where they may also be released systemati-

cally to function in the response of this tissue

to stress.”””' As for the more specialized

painful biological events, there appear to be

other appropriate opioid mechanisms. For

example, these analgesic molecules have

been found in the human placenta, where,

it is hypothesized, they function during

childbirth.”

The many sites of action of the opioid

peptides may, in part, account for the di-

versity in their amino acid sequences. An-

other, equally significant component in this

heterogeneity is the presence in the nervous

system of several pharmacologically distinct

classes of opiate receptors (Table 3). These

molecules are situated in the membrane of

cells receiving opioid chemical signals. They

recognize the three-dimensional structure

of the peptide released nearby and react

with them in a reversible manner. Opiate

receptors in different tissues or regions of

the brain possess differing preferences or

reaction affinities for the opioid peptides

and alkaloid drugs.****” The complexity of

opioid peptide structure may be matched

by variations in receptors, thus providing

greater specificity in the translation of each

opioid chemical messenger.

Other Neuropeptides

There are groups of non-opioid neuro-

peptides which are equally intriguing.’°

They may be classified according to the

tissue in which they function or by their

structural relatedness (Table 4). As was in-

dicated earlier, the ‘““chormones”’ vasopres-

sin and oxytocin are synthesized and re-

leased by hypothalamic nerve cells. Despite

their divergent hormonal actions these

peptides are structurally similar with the

same amino acids in seven of the nine re-

sidues. Immunohistochemical studies have

demonstrated the presence of vasopressin

and oxytocin-like peptides throughout the

central nervous system,” a result which

suggests that like the opioid peptides they

are participating in diverse nerve cell com-

Table 3.—Proposed Classes of Opiate Receptors in the Nervous System.

Designation Preferred Drug Characterization

mu morphine 1 nM affinity; irreversibly blocked by naloxazone;

potent in guinea pig ileum; mediates analgesia and

euphoria.

mu2 morphine 10 nM affinity.

delta enkephalins potent in mouse vas deferens; mediates analgesia.

kappa enthylketocyclazocine mediates sedative effects.

dynorphin

sigma benzomorphans mediates disphoria, and hallucination;

SKF 10047 possible receptor for “‘angel dust.”

phencyclidine

These classes of opiate receptors have been proposed at different times to explain the data obtained from

competitive binding studies, bioassays and behavioral analysis of different opiate alkaloids and peptides, as well

as non-opioid drugs which appear to interact with opiate receptors. Many nervous system tissues appear to

contain several different receptors and this complicates the unequivocal demonstration of more than 2 or 3

distinct opiate receptors.

NEUROPEPTIDES AND BRAIN FUNCTION 85

Table 4.—Families of Neuropeptides.

Hypothalamic-Posterior Pituitary

Argenine vasopressin-in some cells with dynorphin

+Lysine vasopressin

+Oxytocin—in some cells with methionine-enkephalin

+Vasotocin

+Isotocin

+Mesotocin

Hypothalamic-Anterior Pituitary

Luteinizing hormone releasing factor

Corticotropin releasing factor

Thyrotropin releasing factor

Other Neuropeptides

Bradykinin

Bombesin

Carnosine

Avian pancreactic polypeptide—with somatostatin

Opioid peptides (Table 1)

Factor S, sleep peptide

Delta sleep inducing peptide

Endogenous Peptide—A reported Diazepam antagonist

Spinal Sensory Cells

Substance P

+Physaelemin

+Eleidosin

Somatostatin

Vasoactive intestinal peptide

Cholecystokinin, +Gastrin

Angiotensin II

Dynorphin (?)

Gastrointestinal Tract Neurons

Neurotensin

Substance P

Somatostatin

Cholecystokinin

Dynorphin

+Enkephalin

Vasoactive intestinal peptide

+Glucagon

+Gastric inhibitory peptide

+Secretin

Possible Hormone Neuropeptides

Insulin

ACTH

Thyroid stimulating hormone

Growth hormone

A summary of the neuroactive peptides is presented in which they are grouped according to the tissues with

which they are most prominently associated. Many of the peptides listed are believed to function within the

central nervous system.

(+) indicates peptides with similar amino acid sequences to molecule listed immediately above it.

munication processes. It is quite possible

that some of these circuits regulate func-

tions which are related to those that the

peptides control as hormones. This is most

apparent for the peptide hormones regulat-

ing the reproductive system.

Peptides with similar amino acid se-

quences have been purified from the gas-

trointestinal tract (Table 4) where they

function as local messengers regulating se-

cretion and motility. While several of these

gut peptides are found in the brain, they

also are prominent within the sensory nerve

cells which convey information from inter-

nal organs and the surface of the organism

to the spinal cord and eventually to the

brain. The function of these gut and sen-

sory peptides in the brain is currently being

determined with great interest, since it ap-

pears that one of these peptides, cholecys-

tokinin,”’ may mediate, centrally, the per-

ception of hunger and satiety”* in a manner

analogous to the opioid peptides’ role in

pain perception.

As was indicated earlier, hypothalamic

nerve cell peptides regulate the release of

trophic hormones from the anterior pitui-

tary. In this role the peptides regulate many

functions including growth, maturation,

homeostasis, response to injury and repro-

duction. One of these peptides, termed so-

matostatin because it inhibits the release of

growth hormone, demonstrates the diver-

gent utilization of these hypothalamic re-

leasing factors by the nervous system. So-

matostatin is found in nerve cells in the

digestive tract, spinal-sensory cells, retinal

nerve cells in the eye and in local circuits

throughout the brain.”° Indeed, this peptide

may be one of the single most prevalent

transmitter chemicals in the nervous sys-

tem. The number of functions in which it

may thus participate is extraordinary. An-

other releasing factor, LHRH, may have a

much more restricted function. Beyond its

role in the control of luteinizing hormone

release and ovulation, this peptide also

mediates a repertoire of behaviors related to

86 NEALE, BLANK, PLISHKA, and SWEETNAM

coitus.” While this has been demonstrated

in experimental animals, the involvement of

LHRH in the human sexual response has

not been established.

A simplifying principle in neurobiology

was formulated by Sir Henry Dale nearly

half a century ago. He suggested that indi-

vidual nerve cells use but a single neuro-

transmitter at each of their many synaptic

endings. This concept appeared to be valid

with respect to small amine transmitter

molecules, such as acetylcholine, norepi-

nephrine, glutamate, dopamine, serotonin,

and gamma-aminobutyric acid. The neu-

ropeptides, however, appear to often coex-

ist in nerve cells with the small amine

transmitters.’ This potential to release from

a single nerve ending both a traditional

neurotransmitter and one or more neu-

roactive peptides considerably enlarges the

scope and complexity of the communica-

tion process and provides yet another mech-

anism through which the circuits of the

mind can execute their extraordinary tasks.

Recent progress in peptide neurobiology

barely keeps pace with the imagination as

ever expanding roles for peptides in human

function are recognized. While researchers

probe the function of these molecules in

pain perception, sexual behavior, childbirth,

vision or hunger, still others are opening

even more novel vistas with reports of sleep

regulating peptides*”*’ or those which may

act to determine our sense of anxiety.°” And

these observations, one has the sense, re-

flect merely the beginning.

References

1. Pert, C. B. and S. H. Snyder. 1973. Opiate recep-

tor: demonstration in nervous tissue. Science,

179: 1011-1014.

2. Simon, E. J., J. M. Hiller and I. Edelman. 1973.

Stereospecific binding of the potent narcotic

analgesic (*H)-etorphine to rat brain homoge-

nate. Proc. Natl. Acad. Sci., 70: 1947-1949.

3. Terenius, L. 1973. Stereospecific interaction be-

tween narcotic analgesics and a synaptic plasma

membrane fraction of rat cerebral cortex. Acta

Pharmacol. Toxicol., 32: 317-320.

4. Hughes, J., T. W. Smith, H. W. Kosterlitz, L. A.

Fathergill, B. A. Morgan and H. R. Morris. (1975).

Identification of two related pentapeptides from

the brain with potent opiate agonist activity. Nature,

258: 577-579.

5. Ling, N., R. Burgus and R. Guilleman. 1976. Isola-

tion, primary structure, and synthesis of a-endorphin

and y-endorphin, two peptides of hypothalamic-

hypophysial origin with morphinomimetic activ-

ity. Proc. Natl. Acad. Sci. USA, 73: 3942-3946.

6. Feschli, W., A. Goldstein, M. Lowney, M. Hunka-

piller and L. Hood. 1982. Isolation and amino acid

sequence analysis of a 4,000-dalton dynorphin

from porcine pituitary. Proc. Natl. Acad. Sci.

USA, 79: 5435-5437.

7. Rossier, J., Y. Audingier, N. Ling, J. Cros and

S. Udenfriend. 1980. Met-enkephalin-Arg*-Phe’, pres-

ent in high amounts in brain of rat, cattle and

man, is an opioid agonist. Nature. 288: 88-90.

8. Kangawa, K., H. Matsuo and M. Igarishi. 1979. a-

neo-endorphin, a big leu-enkephalin with potent

opiate activity from porcine hypothalami. Bio-

chem. Biophys. Res. Commun., 86: 153-160.

9. Hokfelt, T., O. Johansson, A. Ljungdahl, J. M.

Lundberg and M. Schultzberg. 1980. Peptidergic

neurons. Nature, 284: 515-521.

10. Bloom, F. E. 1981. Neuropeptides. Sci. Amer.,

245: 148-168.

11. Johansson, O., T. Hékfelt, R. P. Elde, M. Schultz-

berg and L. Terenius. 1978. Immunohistochemical

distribution of enkephalin neurons. In: Advances

in Biochemical Psychopharmacology, Vol. 18. E.

Costa and M. Trabucchi, eds., Raven Press, New

York, pp. 51-70.

12. Uhl, G. R., R. R. Goodman, M. J. Kuhar and S. H.

Snyder. 1978. Enkephalin and neurotensin: Im-

munohistochemical localization and identifica-

tion of an amygdalofugal pathway. In: Advances

in Biochemical Psychopharmacology, Vol. 18. E.

Costa and M. Trabucci, eds., Raven Press, New

York, pp. 71-87.

13. Lewis, R. V., A. S. Stern, S. Kimura, J. Rossier, S.

Stein and S. Udenfriend. 1980. An about 50,000-

dalton protein in adrenal medula: a common pre-

cursor for (Met)- and (Leu)-enkephalin. Science,

208: 1459-1461.

14. Nakanishi, S., A. Inoue, T. Kita, M. Makamura,

A. C. Y. Chang, S. Cohen and S. Numa. 1979. Nu-

cleotide sequence of cloned cDNA for bovine cor-

ticotropin-B-lipotropin precursor. Nature, 278:

423-427.

15. Noda, M., Y. Furutani, H. Tskahashi, M. Toyosato,

T. Hirose, S. Inayama, S. Nakanishi and S. Numa.

1982. Cloning and sequence analysis of cDNA for

bovine adrenal preproenkephalin. Nature, 295:

202-206.

16. Gubler, U., P. Seeburg, B. J. Hoffman, L. P. Gage

and S. Udenfriend. 1982. Molecular cloning estab-

lishes proenkephalin as precursor of enkephalin-

containing peptides. Nature, 295: 206-208.

17. Comb, M., P. H. Seeburg, J. Adelmann, L. Eiden

and E. Herbert. 1982. Primary structure of the human

Met- and Leu-enkephalin precursor and its MRNA.

Nature, 295: 663-666.

18. Kakidani, H., Y. Furutani, H. Takahashi, M. Noda,

Y. Morimoto, T. Hirose, M. Asai, S. Inayama, S.

Nakanishi and S. Numa. 1982. Cloning and se-

quence analysis of cDNA for porcine B-neoen-

dorphin/dynorphin precursor. Nature, 298:

245-249.

19. Barker, J. L., J. H. Neale, T. G. Smith, Jr. and R. L.

Macdonald. 1978. Opiate peptide modulation of

amino acid responses suggests novel form of

neuronal communication. Science, 199: 1451-

1453.

20. Viveros, O. H., E. J. Diliberto, Jr., E. Hazum and

K-J. Chang. 1979. Molecular Pharmacology., 16:

1101-1108.

21. Lewis, J. W., M. G. Tordoff, J. E. Sherman, J. C.

Liebeskind. 1982. Adrenal medullary enkephalin-

like peptides may mediate opioid stress analgesia.

Science, 217: 557-559.

22. Houck, J. C., C. Kimball, C. Chang, N. W. Pedigo

and H. I. Yamamura. 1980. Placental B-endor-

phin-like peptides. Science, 207: 78-79.

23. Zukin, R.S. andS.R. Zukin. 1981. Multiple opiate

receptors: emerging concepts. Life Sci., 29: 2681-

2690.

24. Wolozin, B. L., S. Nishimura and G. W. Pasternak.

1982. The binding of x- and o-opiates in rat brain.

J. Neurosci., 2: 708-713

25. Wolozin, B. L. and G. W. Pasternak. 1981. Classi-

fication of multiple morphine and enkephalin

binding sites in the central nervous system. Proc.

Natl. Acad. Sci. USA, 78: 6181-6185.

Journal of the Washington Academy of Sciences,

Volume 72, Number 3, Pages 87-97, September 1982

26. Elde, R. and T. Hokfelt. 1979. Localization of hy-

pophysiotrophic peptides and other biologically

active peptides within the brain. Ann. Rev. Phys-

iol., 41: 587-602.

27. Rehfield, J. 1980. Cholecystokinin. Trends Neu-

rosci., 3: 65-67.

28. Nemeroff, C. B., A. J. Osbahr, G. Bissette, G.

Jahnke, M. A. Lipton and A. J. Prange. 1978.

Cholecystokinin inhibits tail-pinch-induced eat-

ing in rats. Science, 200: 793-794.

29. Moss, R. L. 1979. Actions of hypothalamic-hyypo

Physiotropic hormones on the brain. Ann. Rev.

Physiol., 41: 617-631.

30. Kruegar, J. M., J. R. Pappenheimer and M. L.

Karnovsky. 1982. The composition of sleep-

promoting factor isolated from human urine. J.

Biol. Chem., 257: 1664-1669.

31. Schoenenberger, G. A. and M. Monnier. 1977.

Characterization of a delta-electroencephalogram

(-sleep)-inducing peptide. Proc. Natl. Acad. Sci.

USA, 74: 1282-1286.

32. Guidotti, A., B. Ebstein and E. Costa. 1981. Purifi-

cation and characterization of an endogenous

brain peptide that competes with *H-diazepam

binding. Soc. Neurosci. Abstr., 7: 634.

Personal Scientific Computers*

Seymour Haber

National Bureau of Standards, Washington, D.C. 20234

ABSTRACT

This paper discusses the sorts of small, low-priced computers that are now available for

scientific work, and that are expected to become available during this decade. The background

technological developments are described briefly and projections are made of the capabilities

that may be expected in desktop computers. Software questions that are especially relevant to

the use of such computers are taken up. It is suggested that some software developments may

lead scientists to extend the ways in which they use computers.

* Note: Certain commercial products are identified

in this paper. Such mention does not imply recom-

mendation or endorsement by the National Bureau of

Standards, nor does it imply that the products identi-

fied are the best available for the purposes discussed.

tT This article was written in 1981.

Introduction

The Association for Computing Machin-

ery is one of the largest associations of

computer professionals in the world. When

it held its annual meeting last yearT, it was

88 SEYMOUR HABER

conscious of the fact that 1980 marked the

35th anniversary of the start of operation

of the ENIAC, the first electronic computer.

Now there were several hundred thousand

computers in the world, and the association

itself had about fifty thousand members.

The theme it chose for the annual meeting

was ‘‘Previewing the Computer Age.”’

Previewing?! The situation is certainly

remarkable. Computers have already exten-

sively affected scientific research. Through

credit card systems, computerized super-

markets, and electronic games, they have

some effect on the daily lives of most Amer-

icans. Giant corporations manufacture com-

puters. But it now seems that this 1s barely

the beginning. The number of computers is

expected to grow from about one-half mil-

lion in 1980 to at least 10 million, and per-

haps 100 million, by the year 2000.’ Pushed

by the immense advances of electronic

technology, computation is becoming less

and less expensive and also more and more

widely useful. One aspect of this develop-

ment is that small and inexpensive compu-

ters will be available for the individual use

of scientists. This paper attempts to de-

scribe the capabilities that those small com-

puters will have during the next five to ten

years, and some impacts they may have on

the ways scientists use computers. (I shall

focus on computation primarily and on

data processing, without discussing the use

of computers in the experimental labora-

tory—with which I am not familiar. Nor

shall I discuss, except incidentally, advances

that may take place in the use of large com-

puters. The views expressed in this paper are

my own and are not official conclusions of

the National Bureau of Standards.)

Small inexpensive computers already

exist. The most widely sold ones are the

Radio Shack and Apple computers, which

many scientists regard as toys; but they are

much more than that. J. Presper Eckert, Jr.,

one of the builders of the ENIAC, recently

described that machine as “‘the equivalent

to a simple Radio Shack computer of the

type that sells for $500”’.” Some present day

scientific problems are within the capaci-

ties of such machines, though of course

some are not. And better machines are

arriving.

The Technology

To a certain degree we can clearly foresee

the sorts of small computers that will soon

be made. There is a steady progression

from the laboratory demonstration of each

advance in electronic technology to the

production, in sample quantities, of inte-

grated circuits incorporating that advance;

and then to mass production of those cir-

cuits and production of computer circuit

boards containing them, and eventually to

their inclusion in complete computers that

have adequate software and may be pur-

chased off-the-shelf.

The transistor was invented in 1947.

Monolithic integrated circuits, consisting of

several functioning electrical circuit ele-

ments formed in a single piece (“‘chip”’) of

material, were first made in 1959. Ever since

then, the number of circuit elements per chip

has been growing. Today, chips containing

more than 50,000 transistors are in regular

production, and chips containing almost

500,000 have been shown at technical con-

ferences. The main route to increasing the

amount of circuitry that can be built (grown,

photoengraved) on a chip is to make the

circuit elements smaller. The densest chips

that are now in mass production have fea-

tures that are about 3 micrometers wide.

The microelectronics industry appears con-

fident that it can lower this to 1 micrometer

in this decade.’ This would allow a ninefold

increase in the number of circuit elements

on achip. But other improvements are ex-

pected to result in the reliable production of

larger chips, so that the number of circuit

elements will increase even more.

An example of what can be accomplished

by putting more circuitry on each chip is

building chips that contain larger and

larger amounts of memory. Figure 1, repro-

duced from Ref. 4, shows some of this devel-

opment. In 1970 integrated circuit memory

chips containing 256 bits of memory per

chip were being replaced by chips contain-

——

a oe ee eS —— eee ee eee eee

okoate

PERSONAL SCIENTIFIC COMPUTERS 89

COST PER BIT (CENTS)

mos

1973 1975 1977 1979 1931 1993

Fig. 1. Adiscription and projection, made in 1977,

of four successive generations of random-access

ing 1024 (=1K*) bits; then came 4K bit

chips and then 16K bit chips, and this year

those are being replaced, in the market-

place, by 64K bit chips.

Most important is the fact that the new

chips end up, each time, costing little more

than the old ones, and so each bit of mem-

ory comes to cost very much less. The

curves in Figure | illustrate a famous phe-

nomenon known as the “learning curve”’:

A new type of chip is quite expensive to

make at first—often so expensive that the

Same functions can be obtained more

cheaply using chips of the older type. How-

ever as production continues, the unit cost

goes down, declining roughly 20 to 30 per-

cent each time the cumulative output dou-

bles [Ref. 4, p. 67]. Soon the new chips are

more cost effective than the old, if they are

successful.

*In the world of computers, ‘*K”’ usually denotes

2'° = 1,024, rather than 1,000. However usage varies,

resulting in the number 65,536 (=64 X 1,024) for ex-

ample, being written variously ‘“‘65K"’, “66K”, or

““64K"’. I shall use the last form. In the same manner,

““M” (for ‘‘mega”’) represents not 10°, but 27°. A meg-

abit is thus 1,048,576 bits.

(Figure 1 was originally published in

1977, and it is interesting to compare its

projections with what happened afterward.

Despite rapid inflation, the cost of 64K bit

chips is now about what was projected. But

16K bit chips are much cheaper than

projected—about 0.01 cents per bit—and

the denser chips have just begun to displace

them.)

At present 64K bit memory chips are in

mass production, and samples of 256K bit

chips are beginning to appear. Megabit

chips are expected to be in mass production

late in the decade. It seems that we can ex-

pect that the cost of memory will con-

tinue to decline by a factor of 2—apart

from effects of inflation—every 2 or 3 years.

At the same time memory will become

physically smaller. How much memory is

desirable, in a single-user computer? (We

are considering, now, only the primary or

‘“‘fast’’ memory; computers also have sec-

ondary storage, as on magnetic tapes or

magnetic disks.)

I recall that the first computer I worked

on, twenty-five years ago, had 40K bits of

memory. We were using it for a wide range of

scientific problems, including solution of

partial differential equations, and definitely

felt cramped. When the memory was quadru-

pled to 160K bits, we all felt that was quite

sufficient. But as problems are solved, more

difficult problems are tackled. Nowadays

the memories of large scientific computers

are counted in megabytes (1 byte = 8 bits;

IM byte = 2”’ bytes = 8,388,608 bits), often

going as high as 32 megabytes. Such large

memories are needed by very few (though

important) scientific calculations—their

main purpose is to allow the computer to

efficiently serve a multitude of “‘time shar-

ing’’ users. Small, single-user computers

need not have quite so much memory.

The first “‘personal’’ computers came

with memories ranging in size from 4K to

16K bytes. This was in 1977, when personal

computers became available not only as kits

but as complete machines, with reasonable

software, ready to use. Many of the buyers

were engineers and scientists, and they

found the memory rather limiting. Today,

90 SEYMOUR HABER

memories of 48K or 64K bytes are com-

mon.* That size of fast working memory is

typically used as follows: 20K-25K bytes

are likely to be taken up by the machine’s

operating system and the interpreter for the

BASIC language. A program of 200 lines

may take up 4K bytes, leaving 35K-40K if

the memory is 64K. Each number used in

the calculation will take up 5-7 bytes, in-

cluding its addressing information. Thus

there is working space for perhaps 6,000 or

7,000 numbers. If the program draws graphs

or other figures, it must set aside space to

store graphic information—perhaps 8K

bytes. That would reduce the capacity of

the numerical working space to 4,000-6,000

numbers. This would allow, for an exam-

ple, calculating with matrices up to a size of

about 50 X 50.

If working memory were increased to

256K bytes, for example, that would allow

for perhaps 50,000 numbers. That is enough

for fairly large-scale computations; as desk-

top machines become faster, they will be able

to handle even larger computations. Ad-

vanced graphics (higher resolution, color)

would use considerable amounts of mem-

ory. A picture formed with a resolution

of 1000 points by 1000, with 16 colors or 16

levels of ‘‘grey scale,”’ can fill 256K bytes of

memory. Such capabilities and more are

not uncommon now, in graphics systems

that are used with large computers. I shall

discuss graphics on small computers below;

it seems very likely that personal compu-

ters will come to have advanced graphics

capabilities.

So it appears that reasonable memory

sizes for desktop computers will range from

64K bytes to a few megabytes. The 64K bit

*To describe these computers a little: They com-

monly cost about $500-$1500 at entry level—for that

you get the computer and some memory, a keyboard,

and either a T V-type screen for display or a device that

allows you to use a home television set. There is usu-

ally also an arrangement for storing programs, and

perhaps other files, on magnetic tape cassettes using

an ordinary cassette recorder. Most purchasers add a

printer and a better form of external memory—one or

two disk drives. That brings the price into the

$2,000-$3,000 range.

and even megabit memory chips will indeed

be of use.

Besides expanding memory, another thing

that can be done with very fine integrated

circuitry is to put the whole central process-.

ing unit of acomputer on asingle chip. This

remarkable accomplishment—a triumph

of the electronics industry—was first done

in 1971. The CPU’s of the first personal

computers—the famous 8080, 6502, and

Z80—belong to the second generation of

these ‘‘microprocessors’’.* These are “‘8-

bit’” microprocessors; they operate on a

single byte of information at a time and

they have very limited internal facilities for

arithmetic. The only calculation they do di-

rectly, as a single operation, is an 8-bit add

or subtract. A scientist programming one

of these machines does not, however, think

in terms of 8-bit binary numbers as a usual

thing. He will instead think in terms of the

numbers his version of BASIC—or per-

haps FORTRAN—uses: usually decimal

floating-point numbers of from 6 to 16 sig-

nificant figures. An addition or multiplica-

tion of such numbers is carried out by a

special program embedded in the compu-

ter’s software. It will involve hundreds of

separate operations at the microprocessor

level—memory accesses, logical operations,

additions and subtractions. As a result the

scientist sees arithmetic operations as taking

1 to 5 milliseconds, even though the cycle

time of his machine’s CPU may be less than

1 microsecond. Another limitation of those

microprocessors is that they have only 16

address lines for memory handling, which

allows just 2'° possible addresses. That is

why personal computers have generally

been limited to only 64K (=2'°) bytes of

memory.

The integrated-circuit technology that

produced the microprocessors mentioned

above, is about the same as that which pro-

duced the 4K bit memory chips. By now it

should be possible to produce far more

* This term gave rise to the term “microcomputer”

forasmall computer whose CPU is a microprocessor.

Unfortunately ‘“‘microcomputer” has a different us-

age in the electronics industry.

PERSONAL SCIENTIFIC COMPUTERS 91

complex microprocessors; ones which could,

say, do full-scale floating arithmetic oper-

ations directly. This is in fact being done,

but progress is not as fast as with memory

chips. The reason is that memory chips have

a simple, regular structure that is easy to

design. Designing a new-generation micro-

processor is another matter, and takes years.

The third generation of microprocessors—

**16-bit’’ machines, bearing such names as

8086, 9900, Z8000, MC68000—has been in

production for some time. Complete desk-

top computers that incorporate those micro-

processors have just begun to appear.

The 16-bit microprocessors generally have

faster cycle times than the 8-bit ones, and

also far more capable arithmetic units.

They add 16 bits at a time, and usually have

multiplication and division as single opera-

tions. They generally also have 24 address

lines, raising the memory limitation from

64K to 16M bytes, where it should not

bother anybody for a while.

Some 32-bit microprocessors have been

announced, though they are not yet in mass

production. Their arithmetic capabilities

begin to resemble those of the CPU’s of

large computers in the IBM 370 line. They

will not be seen in desk-top computers for

some years yet; designing has to be done

and software has to be written. But the line

of future progress is clear.

When it became evident, some years ago,

that these advances in the manufacture of

the core parts of computers—CPU’s and

memories—would take place, it was thought

that the cost of computer peripherals, such

as printers, plotters, and tape or disk mem-

ories, would not change much; so that

complete computer systems would remain

expensive. But that did not happen. To an

Outside observer it seems as though the

demand generated by the possibility of

cheap core computers brought forth an ex-

plosion of new cheap peripherals (perhaps

I should add: cheap only in price, not in

capability or reliability). There are now

desktop printers, printing 50-100 lines per

minute, for less than $500; floppy-disk

drives, storing 100-300K bytes on each

disk, for about $500; and fast-access ** Win-

chester technology” disks, storing 3-10M

bytes, for under $4,000. All the prices are

still dropping, as are the physical sizes of

the disk units.

An intriguing device that has not yet ap-

peared as a computer peripheral, but cer-

tainly will do so soon, is the videodisk

reader. It has enormous potential; a single

videodisk can store thousands of megabytes

of information, all accessible in seconds.

Readers may perhaps cost $1,000 or $2,000,

and disks cost very little. It remains to be

seen what libraries of programs and infor-

mation will be made available on videodisks.

(The ability to write onto videodisks lies

farther in the future.)

Graphics

One of the most attractive features of

personal computers is their graphics capabil-

ity. Almost every scientist often constructs—

by hand, on paper—plots of functions or

data points. We are all aware of the im-

mense advantage of visual representation

over numerical tables, for gaining under-

standing of clustering of data, varying rates

of change, and indeed most kinds of quan-

titative functional behavior. The primary

output medium of personal computers is

the CRT screen, and one thing they can do

with ease is plot points and curves on the

screen. The drudgery of graphing is largely

removed, and a plot can be obtained on the

screen about as quickly as the data points

can be entered into the computer, or pro-

duced by a program running inside the com-

puter. In addition, most of the printers now

being sold for use with micro-computers

can quickly copy onto paper whatever

image is present on the screen.

There are limitations, of course. The ease

with which graphs may be produced de-

pends strongly on the software available

for this purpose. We shall go into that when

we discuss software. The other limitations

have to do with resolution—the fineness of

the lines which may be drawn on the screen

and the closeness with which they may be

drawn without merging—and with color.

92 SEYMOUR HABER

A person working with pencil and paper

can draw lines of perhaps %4 mm width,

separated by less than 1 mm, with ease; and

these may slant in any direction. This gives

an 8%” X 11” sheet of paper an effective

resolution of better than 200 X 300; and

one can use larger sheets of graph paper to

get effective resolutions as great as 1000 X

1000 and more. Some low-priced personal

computers can plot only rectangular blocks

on the screen, to give effective resolutions as

low as 40 X 40. Others can plot smaller

dots, and resolutions of about 200 X 300

are common, with some going as high as

300 X 500. Even then, the result will often

not look as good as a hand-drawn curve on

84” X 11” paper. The reason is that the

most common computer graphics systems

place their dots only at points of a specified

rectangular grid, each grid point corre-

sponding to a specific bit in the machine’s

memory. An attempt to drawa straight line

segment that is very nearly horizontal will

result, because of the restriction to points of

the grid, in a succession of actually hori-

zontal segments, each slightly displaced

vertically from the preceding one. The

same applies to near-vertical lines; and this

effect also results in a roughness in curves of

continuously changing direction, when their

directions come near to vertical or hori-

zontal.

This “‘staircase effect’’ can be mitigated

by going to higher resolution—finer dots,

more closely placed. It can also be elimi-

nated in a different way, if each screen dot

may be varied in intensity, rather than only

turned on or off. Either approach increases

the memory requirements of the computer’s

graphics subsystem, and so could not be

used in the first generation of personal

computers, which were limited in memory.

But now memory is becoming very inex-

pensive, and the new microprocessors can

manage very large memories.

When graphics is done by hand, color can

easily be included; and color is a very valu-

able adjunct. It serves to focus attention,

and also to organize information. For ex-

ample, if data points obtained by different

experimenters are being shown together, a

distinct color may be assigned to each

source, so as to clearly show any differ-

ences in tendency or scatter among the sev-

eral sets of data. But adding color into

computer graphics runs into difficulties.

One is that color CRT’s of high resolution

have been very much more expensive than

monochrome ones. Another is the problem

of “Shard copy’’—how do you reproduce on

paper the colored image you have formed

on the screen? Until recently, the best

method was essentially to photograph the

screen and then make photographic prints!

But both problems are in the process of so-

lution. Medium-resolution (300 X 500)

color CRT’s are now available for under

$1,000, and the price of higher resolution is

dropping. Printers using a 3-color ribbon,

together with the inexpensive “dot-matrix”

technology that is common in the lower

priced black-and-white printers, are com-

ing down in price to the $2,000 range. A

new technology, announced by Sony Cor-

poration for making color prints in its new

filmless camera system, holds promise of

further lowering of costs.

A graphics facility that has never been

available to scientists working with pencil

and graph paper, but that personal compu-

ters are beginning to provide, is the capabil-

ity for dynamic graphics—display of mo-

tion, either continuous or through steps of

convenient size. This is made possible by

the computer’s ability to send information

very rapidly to its own CRT display unit, so

that it can send a new picture in ‘4o of a

second—the amount of time between suc-

cessive frames in television broadcasts. (In-

cidentally, this is an advantage desktop

computers have over time-sharing systems;

trying to send so much information froma

central computer to a terminal would over-

burden most communications lines.) At the

moment, dynamic graphics is being used

mostly for games. Indeed scientists are not

accustomed to thinking in terms of using it,

since the capability—traditionally the abil-

PERSONAL SCIENTIFIC COMPUTERS 9

ity to produce a “‘movie’’—has rarely been

available to them. Still, the dynamic nature

of most phenomena studied by scientists

strongly suggests that dynamic graphics

will prove valuable. (Though it is hard to

see how it could be included in a paper for

publication!)

Projections

I shall begin by describing the main

characteristics of two types of machines that

are available today (mid-1981); and then de-

scribe the sorts of desktop computers that I

expect to see in 1984 or thereabouts, and

about 1988. (I think that a small computer

““generation”’ will be about 3 or 4 years, for

a while.) In connection with these projec-

tions, let me first describe what a megabyte

of memory is like, and what I shall mean by

Table 1.—Two present-day computers.

the ‘‘speed”’ of computers: A megabyte is

the amount of memory space that will ac-

commodate about 200,000 numbers, or, if

one is working with higher precision—say

16 significant figures—about 125,000 num-

bers. Alternatively, it is the space taken up

by 40,000 or so lines of program code.

When English text is stored in a computer,

each character—letter, punctuation mark,

space—takes up one byte. A megabyte will

hold a medium-sized book, of about 350

pages. It’s a fairly large amount of memory,

for the purpose of working space that may

be accessed in microseconds.

As for computer speed, it is really not

quite a scalar quantity. Acomputer may do

one operation faster, and a second opera-

tion slower, than some other computer.

This is particularly so for very different

types of activities, such as massive arith-

metic on the one hand and managing time-

sharing on the other. The problem is per-

Price: $2,500

Speed: 1

Main Memory: 64K byte

Secondary Memory:

Printer:

100K byte floppy disk

75 LPM, low quality print

fair mechanical quality

$12,500

256K byte

250K byte floppy disk

120 LPM, medium quality print

good mechanical quality

Table 2.—Two 1984 computers.

“Price: $2,000

Speed: 10

Main Memory: 256K byte

Secondary Memory

Printer:

medium mechanical quality

250K byte floppy disk

100 LPM, medium quality

$10,000

between 20 and 50

1M byte

500K byte floppy disk, 10M byte

sealed disk

150 LPM, good quality print

plotter

good mechanical quality

Table 3.—Two 1988 computers.

“Price: $2,000

Speed: between 20 and 100

Main Memory: 256K byte

Secondary Memory: 1M byte floppy disk

Printer:

medium mechanical quality

150 LPM, good quality print

$10,000

between 100 and 1,000

2M byte

25M byte

300 LPM, good quality print

color plotter/printer

good mechanical quality

* These prices are in 1981 dollars; I make no attempt to estimate future inflation rates.

94 SEYMOUR HABER

haps alleviated a bit in the case of desktop

computers, which are primarily single-user

machines, and are not apt to be used in such

widely differing ways. The speed estimates

given below are intended to refer initially to

the classical type of scientific computing,

involving a lot of arithmetic and consider-

able use of functions such as square root,

cosine, etc. In any case they are to be re-

garded as quite rough figures.

As my unit of speed I shall take the speed

of current personal computers, such as the

TRS80, Apple II, or PET. Those are not all

the same speed, in fact, but for the accuracy

involved here I may gloss over the differ-

ences. This speed unit may be though of as

the ability to do about 20,000 arithmetic

Operations, or some similar-sized mix of

other computer operations, in one minute.

For comparison to larger computers, it is—

very roughly—one two-hundredth the speed

of an IBM 370/145.

(In these tables “‘low quality print’? means

that the letters are formed of distinctly vis-

ible dots and only in upper-case, or perhaps

also ina limited form of lower-case in which

the lower parts of the letters g,j, p,q, and y

do not descend below the line. ““Medium

quality” refers to printing in both upper

and lower case, with better-formed letters;

and “‘good quality” refers to results equal

to those produced by an office typewriter,

or better.)

This is a rosy picture! The amount of

computing that can be done on a machine

of speed 100—or even 10—and a good-sized

memory, is really large. After all, on a single-

user computer a run of ten to twenty hours

can be done any night; and runs of 100

hours or more can be made when needed.

But something important is missing from

the picture.

Software

Will the software that is available on large

computers be available on little ones? The

99 66

answer is “‘yes,”’ “‘maybe not,” or “better

than that,” depending on the type of soft-

ware.

The software that a scientist is most con-

scious of is the programming language—

FORTRAN, BASIC, APL, etc.—in which

he works. Here we may pause to consider a

striking contrast between hardware and soft-

ware: For all the advance there has been

and continues to be in computer hardware,

we are using the same software as 25 years

ago! It was about then that the “‘high level,”

‘problem oriented” languages FORTRAN

and COBOL were created. Before that,

programs had to be written directly in ma-

chine language (long strings of zeros and

ones; usually compressed into octal or hex-

adecimal numerals) or in assembly language.

With the high-level languages researchers

could step back from the machine, focus less

on the details of its actions, and give more

attention to their problems and their algo-

rithms for solving them. In the years since

then these languages have been improved

and many others, of similarly high level,

have been introduced; but there has not

been a general step to any higher level.

What would a “‘higher level’’ language

be like? It would contain commands which

come closer to just specifying the result

wanted, and involve less detailing of the

procedure the computer uses in obtaining It.

For example, BASIC now lets us write

LET C]254

where A, B, and C are to be floating-point

numbers occupying several bytes each, with-

out requiring us to explicitly instruct the

machine what to do with each byte, or ex-

actly where in memory each or those bytes

is. A higher-level language would perhaps

permit us to write

Let S = the sum, for i = —M(1)M and j

= O(2)2M, of. 1/@*1. +37) +

instead of having to write the nested loops

that specify the term-by-term formation of

the sum. There could be direct commands

PERSONAL SCIENTIFIC COMPUTERS 95

for common mathematical and statistical

operations, such as calculating a standard

deviation, or a definite integral, or the solu-

tion of a system of linear equations.*

There is now some progress in the devel-

opment of higher level languages. I expect

some of this development to significantly

extend the way scientists use computers, as

I shall explain below.

Returning to the current programming

languages, a point of interest for us is that

while FORTRAN is the language most

used by scientists when working with large

computers, the language most commonly

provided with desktop computers is BASIC.

BASIC is similar to FORTRAN in its hand-

ling of calculation, but is considerably

simpler and provides many conveniences to

the user in connection with input and output

and program testing and correcting. There

are some problems with BASIC, which will

be discussed below; and certainly a scientist

who wishes to work on both a large and a

desktop computer, would find it burden-

some to have to use different languages.

Another kind of software is the compu-

ter’s operating system. It is often largely in-

visible to the user, but a capability offered

by some operating systems is of special in-

terest for single-user computers. That is a

limited sort of time-sharing facility, some-

times called ‘‘foreground/background” op-

eration. It allows the user to do very long

calculations without interfering at all with

his other use of the computer. The long cal-

culation yields priority to every other use of

the computer, resuming when the latter is

concluded. It is to be hoped that some manu-

facturers of desktop computers will pro-

vide this facility.

A third (and fairly new) type of software

is becoming of increasing interest to scien-

tists, and poses a problem in connection

with small computers. That is ““mathemati-

cal software’’—packages, or libraries, of

* There are serious reliability problems involved in

having such commands, as some higher level opera-

tions are subject to errors far less predictable than

those of computer multiplication and division. But

the problems are not insuperable.

very carefully written subroutines for the

most frequently done kinds of mathemati-

cal calculation. LINPACK’ for solving sys-

tems of linear equations, EISPACK® for

eigen-value problems, andthe NAG library’,

are among the packages being produced by

numerical analysis groups around the world.

Numerical analysis methods have pro-

gressed greatly in the past thirty years (see,

for example, Ref. [8]) and these packages in-

corporate, at their best, the latest ideas.

Their subroutines are apt to be much more

efficient and accurate than those scientists

write themselves.

But they may not be available on small

computers! They are generally written in

standardized languages—mainly in FOR-

TRAN—for use on many different com-

puters. But FORTRAN is not usually pro-

vided on small computers. BASIC is a lan-

guage that is hardly standardized at all. Fur-

thermore most versions of BASIC contain

a technical obstacle—the lack of any provi-

sion for local variables in subroutines—that

has the practical effect of making it quite

impossible to write subroutine packages

for them.

There are several possible solutions that

may be implemented by the manufacturers.

One is to switch from BASIC to FORTRAN

as the prevalent programming language.

This would bring the additional advantage

of easing the transition to asmall machine,

for scientists who are already working in

FORTRAN on large computers. It would

also bea step backward, from the ease of use

and generally greater consideration for the

user that has been associated with BASIC.

Another possibility is to extend BASIC to

include subroutines with local variables.

This is now being done by some manufac-

tures, such as Hewlett-Packard and Texas

Instruments. It then becomes necessary to

rewrite the good mathematical software,

into BASIC; and until a standard extended

BASIC emerges, this would have to be

done over and over. A third possibility

perhaps, is to ““extend’” BASIC to become a

higher-level language in which many of

these high-quality subroutines would be

invoked by a single instruction. The sub-

96 SEYMOUR HABER

routines behind the instructions could then

be upgraded, periodically, to gain the benefit

of advances in numerical methods.

I might mention another possible solu-

tion for the manufacturer: ignore the prob-

lem, and with it those customers for whom

the problem is significant. I mention this

only to say that I do not think it will

happen. Technical people have come to

think of themselves as insignificant to com-

puter manufactures, but I do not believe

that that is so. Scientists and engineers

number in the hundreds of thousands in the

United States; if we add in usage in higher

education, it seems reasonable to estimate a

market for 500,000 personal scientific com-

puter systems. If they average about $5,000

in price, that is 24% billion dollars. The

prospect of gaining a major share of sucha

market is significant to most computer

manufacturers. In addition, desktop com-

puters are likely to be used in part (as we

shall see below) in conjunction with a firm’s

Or institution’s central system. The choice

an engineer or scientist makes among desk-

top systems will also influence his views on

those more expensive systems. Scientists

needn’t be bashful about telling computer

sellers what features they would like to have!

A software area especially relevant to

personal computers is that of software for

graphics. At present that is being handled

largely by the addition of graphics instruc-

tions to BASIC. These may be only low-

level, such as an instruction for showing a

dot at a specified screen location. Or they

may be relatively high-level: instructions for

forming axes, placing legends, translating

and rotating figures, filling areas with

color, etc. Dynamic graphics is still used

mainly for games, where it is programmed in

machine language by specialists. But it is

also available to general users, on some

computers, through high-level languages

such as GRASS and LOGO. In the future I

expect to see the capabilities of those

languages added into versions of BASIC or

of other languages designed for scientific

computation.

The last aspect of software that I wish to

discuss is the recent emergence of various

higher level software facilities that will, I

think, bring scientists to use computers in

new ways. These include word processing

programs, electronic mail facilities, and

very high level languages for information

management. In the past such things were

either unavailable or else of little interest to

scientists. Computerized file management,

for example, offers benefits such as auto-

matic cross indexing and searching; but if it

involves learning a cumbersome and hard-

to-remember system of commands, the ad-

vantages are not worth the trouble, and the

scientist will manage with card indexes or

indeed just by keeping the information in

his head. However, there are now higher-

level languages—INFORM, Mark V, TE-

DIUM, PEARL, for example—that are de-

signed to make it easy to form and use

organized information structures in a com-

puter, without learning a new language.

Some operate interactively: They tell the

user, via the display screen, what possibilities

are available to him, and ask him to choose.

When the user has, in this manner, speci-

fied just what he wants, the higher-level

language writes a working program in an

ordinary high-level language—say BASIC

or COBOL—which is afterwards used to

run the new information system. The scien-

tist can specify that the working program

include descriptive statements, forms, or

prompting messages, that allow him and

his associates to use the information system

without having to remember a lot of techni-

calities.

Computerized information, electronic

mail, and word processing gain in utility

from the possibility of interacting with each

other. Furthermore, it seems that offices in

general are moving toward less paper-

handling and more electronic information-

flow—that is what the terms “‘office of the

future’’ and ‘“‘paperless office” signify—

and this may provide additional reasons to

use the new software.

Since this new software functions through

interaction with the user, it should do so in

‘real time’’—that is it should process each

specification made, or piece of information

entered by the user, quickly enough to be

PERSONAL SCIENTIFIC COMPUTERS 97

ready for the user’s next input as he types it.

This is a reason for having desktop compu-

ters, rather than terminals which would

overload a central computer with demands

for rapid complex interactions. On the

other hand, electronic mail requires com-

munication lines, and computerized file

systems may need to use central data bases.

This suggests a picture of desktop compu-

ters connected to local computer networks

and perhaps also far-ranging ones.

To summarize: There are now software

advances taking place that should be of

value to scientists. The outlook for the

availability of the best software on little

computers is good; but a serious question

remains about mathematical software.

Closing Remarks

People seem to /ike computing on their

Own computers much more than on large

multi-user computers. It is a remarkable

phenomenon. I have seen scientists—I have

done it myself—positively enjoy learning

about and using tricky features of their

computers’ languages, when those were not

really necessary for the task at hand; al-

though those same scientists, when work-

ing on a large computer would positively

avoid learning anything more about the

machine than the absolute minimum that

they needed for their work. This “‘play”’

aspect—I’ll call it that for lack of any better

understanding—of working with personal

computers seems to turn what had been

drudgery into pleasure. It is very valuable; I

think it will move scientists to learn to use

computers more effectively. The “‘play”’

factor even seems to eliminate most of the

annoyance of typing for two-finger typists

like myself.

The steadily dropping prices of small

computers strangely inhibit some people

from buying them! They constantly feel that

now is the wrong time to buy—something

better or less expensive is around the corner.

And it does look pretty certain that better

and better things will indeed be coming

along for at least 10 or 15 more years. But it

is certainly wrong to conclude that we

should avoid buying computers, until the

manufacturers stop improving them! A

reasonable strategy, perhaps, is fora scien-

tist to acquire a computer as soon as a relia-

ble, affordable one appears that offers him

important benefits; keeping in mind that he

will probably replace it within five years by

one that is much more valuable to him.

References

1. A Technology Assessment of Personal Computers.

Office of Interdisciplinary Programs, University of

Southern California OIP/PCTA 80-3/1.

2. Computerworld, Vol. 15, No. 9, March 2, 1981,

Page |.

3. Lyman, J., Scaling the Barriers to VLSI’s Fine

Lines. Electronics, Vol. 53, No. 14, June 19, 1980,

pp. 115-126.

4. Noyce, R. N., Microelectronics. Scientific American,

Vol. 237, No. 3, September 1977, pp. 62-69.

5. Dongarra, J. J., C. B. Moler, J. R. Bunch, and G. W.

Stewart, LINPACK User’s Guide. Soc. for Industr.

and Appl. Math., Philadelphia, 1976.

6. Smith, B. T., et al. Matrix Eigensystem Routines

(EISPACK). Lecture Notes in Computer Science 6,

Springer, New York, 1976.

7. NAG FORTRAN Library manual Mark 8, NAG

(USA) Inc., 1250 Grace Court, Downer’s Grove,

Illinois.

8. Parlett, B., Progress in Numerical Analysis. SIAM

Review, Vol. 20, 1978, pp. 443-456.

Journal of the Washington Academy of Sciences,

Volume 72, Number 3, Pages 98-104, September 1982

Caries and Juvenile Onset Diabetes: A Metabolic Viewpoint

Richard T. Koritzer, D.D.S., Malcolm M. Martin, M.D., Arline L. A. Martin, M.D.,

Lucile E. St. Hoyme, D. Phil., and John J. Canary, M.D.

Departments of Fixed Prosthodontics, Pediatrics, Orthodontics and Medicine,

Georgetown University Medical Center, Washington, D.C., 20007, and

Department of Anthropology, Smithsonian Institution, Washington, D.C. 20560.

ABSTRACT

Insulin dependent diabetics, unlike controls, showed low caries incidence to age 8 years.

Timed, coincidentally perhaps, with the prepubertal growth spurt, the caries rate increases.

Except for about 30% of the control sample, who have fewer carious teeth than expected, con-

trols have approximately the caries rate predicted. Variability in caries experience, however, is

much greater in diabetics, with more than 50% of the sample having either too many or too few

carious teeth. Long term change in caries rate varies for well controlled insulin dependent

diabetics and controls. Controls (N = 31) and diabetics (N = 31) in this pilot study ranged in

age from 3 to 27 years, and were selected on the basis of good socio-economic level, similar

environment, and careful dental care.

Introduction

Carles experience in juvenile onset dia-

betics has been studied by numerous inves-

tigators.'° A metabolic entity such as dia-

betes may indeed be associated with host

susceptibility to many disease processes.

An association for diabetes and periodontal

pathology has been well established. How

diabetes relates to dental caries, especially

in children and young adults, has aroused

both curiosity and controversy. Both low’ ~

and high*” caries incidence in young dia-

betics, when compared to “‘normal”’ con-

trols, have been reported.

Caries frequency variations have been

associated with diabetes duration and age

at onset.”*° Lower caries rates in diabet-

ics have been attributed to relatively lower

carbohydrate and increased protein dietary

regimens.” However, the accuracy of die-

tary histories and the quality of patient

compliance has been questioned.” ° Con-

trolled and poorly controlled diabetics have

been studied.’* > For comparison, control

groups have been matched for age and sex’”

and school cohorts have been used.”° The

epidemiologic approach, using relatively

large numbers, has been employed.*°

Seldom, however, have clinical popula-

tions with wide age range been studied. In-

stead, samples have usually been limited to

readily available material, such as school

populations. As a consequence of this re-

stricted age range, it has rarely been possible

to observe caries rate changes with age.

Constant rate becomes an implicit assump-

tion in such studies. In this pilot study, our

samples range in age from 3 to 27 years.

This physiological range of physical differ-

entiation includes early childhood, adoles-

cence, and young adult life.

The data presented in this pilot study sug-

gest that caries rates for well controlled

diabetics and matched controls do indeed

vary with age, but in an unexpected way.

Materials and Methods

Age and caries experience were recorded

for 31 diabetics and 31 controls, ranging in

CARIES: A METABOLIC VIEWPOINT

age from 2 years, 10 months, to 27 years

(Table 1). Caries experience was recorded

by number of carious teeth rather than the

number of surfaces affected or the severity

of the pathology. No teeth had been lost in

either sample. Diabetics were all seen (by

RTK) in the morning, and did not know in

advance that they would have dental exam-

inations. Controls were examined (by RTK)

at random times of the day, but more fre-

quently after school when they came in for

their periodic dental checkups.

Based on clinical observation, periodon-

tal disease was minimal to non-existent in

the diabetics in this study. The entire sample

presented here was apparently free of bony

periodontal pathology even to age 27 and

only a few individuals evidenced minimal

soft tissue change. Controls exhibited more

marginal inflammation associated with

frank plaque accumulation.

Most caries for diabetics observed in this

study involved occlusal surfaces only, buta

few individuals exhibited smooth surface

Table 1.—Number of carious teeth in young diabetics (N = 31) and controls (N = 31).

Table 1la.—Up to 9 years.

Diabetics Controls Risk-years*

Age Carious teeth Age Carious teeth Age Total

eiews) me,» vobs. exp. Aad ey MOaie “OBS. lexne x? y. teeth = r-y r-y/teeth

1 2 10 0 1.65 1.65 x "9 0 2.4 2.4 1:25 12 2 .167

2 a 610 0 1.65 R65) 4". 0 0 2.65 2.65 1.5 16 5 313

3 . ee 0 3.15 35 aes 4 2.8 ol 2 20 13 .65

- 6 1 0 2.88 2.88 (rae: 1 2.95 1.29 3 20 Bi 1.65

5 oS 0 2.8 2.8 iddeke, 5 ¥2 1.01 4 20 53 2.65

6 ee 0 2.8 2.8 Cee: ] 3.37 1.67 5 20 73 3.65

7 6 10 0 2.7 Pl 7 ye 0 3.5 25 6 24 69 2.88

8 6 10 0 zat 7 a 8 5 6 4.21 76 7 24 65 2.17

9 a. 1 4.21 4.21 8 10 8 4.56" 26 8 24 89 3.71

10 8 10 5 4.56 4 9 24 113 4.71

Total 1 22.78 30 15.43

Mean 6 8 1 2:53 7 1 3 1.54

% %

Fewer carious teeth 100 40

Expected no. 0 50

More 0 10

Table 1b.—9 to 15 years.

Diabetics Controls Risk-years*

Age Carious teeth Age Carious teeth Age Total

No. y. mo. obs. exp. x? y. mo. obs. exp. x? ¥: teeth r-y r-y/teeth

10 ee | 7 4.22 1.83

11 ene 10 Ser) mas | 7 3.58 a27 10 24 98 4.08

12 mo oO2 1 4.0 S225) iad 0 4 2.34 DS is A 24 80 3.33

13 11 8 2 2.14 OEY ie BZ SS 0 DGS oo, OA. «12 28 60 2.14

14 ~. © 6 2a” | | RR | 2.65 LE a 28 88 3.14

15 myers 0 564 4 (364--13 "' 0 0 3: 1r aka 4 28 116 4.14

16 e075 0 3.64 3.64 13 4 2 3.47 620) BS 28 144 5.14

17 13. 6 3 Al 3.64 14 11 5 5.14 0

Total 29 26.36 19 11.87

Mean 11 9 3.63 3.3 0 era | 2.71 1.7

% %

Fewer carious teeth 37.5 29

Expected no. 25 57

More 37.5 14

100

Table 1c.—Over 15 years.

R. T. KORITZER, et al.

Diabetics Controls Risk-years*

Age Carious teeth Age Carious teeth Age

No. y. io: obs. exp. x? y. mo. obs. exp. x? y. teeth r-y r-y/teeth

18 15 1 0 5:34. , Sol AS » 0 12 S14), “9.16),,15 28 144 5.14

19 1S: oe 7 5.34 es ST nue 2 5:39) 2A 16 28 172 6.14

20 cee 4 7.14 PSs Vis 36 4 5.64 48 17 28 200 7.14

21 Ler = 0 11 13 2.1 Fs 1 5.89 406 18 32 228 7.13

22 ES!) US 0 7230." J7RO. SGo 24 2 6.47 3:09 18 32 260 8.13

23 18 10 14 8.0 4.5 16°. 23 7 6.64 D2 sar 20 32 292 9.13

24 +6 8 8.63 at a i | 10 7.13 Lie. 2! 32 324 10.13

25 19ers 7 1] 8.63 65 18 3 7 7.39 026. °F 22 32 356 11.13

26 pe | 16 9.13 SAF 28s kG 8 9.63 28, 23 32 388 12.13

27 20 +4 1] 9.46 a a | 9 8 10.88 76 24 32 420 13.13

28 7 a 6 B62) 273° 23) 9 6 12 12.63 03 25 32 452 14.13

29 7 12 14.88 IO, «25% ai 1 14.13 12.20 26 32 484 15.13

30 26 1 2 1515, - 1,39". 25 2 16 14.63 ee ae 32 516 16.13

31 2 6 16.13 6.36 26 1 13 13.13 .30

Total 108 48.63 103 33.82

Mean 20 3 eth SAT TSE OS 7.36 2.42

% %

Fewer carious teeth 36 29

Expected no. 43 64

More 21 7

Table 1d.— Whole series

Total 138 97.74 152 61.12

Mean 13 10 4.45 NO (gm 4.90 1.97

* To obtain an ‘“‘expected”* number for carious teeth, risk-years was estimated by summing the years since eruption for each

tooth and then dividing by the number of teeth present. As deciduous teeth are shed, the number of risk-years decreases accord-

ingly, to rise again as permanent teeth replace them. (It is assumed that teeth erupt at the usual time and that no teeth are missing.)

The resulting number was then divided by the number of teeth present to give a ratio “risk-years/teeth.”’ This ratio is the solid

line in the graph (Figure 1) with which the moving averages for diabetics and controls are compared, as well as the “expected” in

the tables.

caries. There was, indeed, more smooth

surface caries in controls.

The early-onset diabetics were drawn

from patients of a Georgetown University

Hospital (Washington, D.C.) private pedi-

atric practice. The controls were drawn

from a private dental practice in Glen

Burnie, Md., a Baltimore suburb about

twenty miles from the Washington metropol-

itan area. The samples chosen had roughly

similar socio-economic backgrounds. Wash-

ington, D.C., has had fluoridated water

since 1952, while Glen Burnie, Md., water

has been fluoridated since 1963. Incomes,

as a socio-economic indicator, were matched

in the two groups to eliminate access to

dental care as a possible contributor to varia-

bility. Annual family income range in each

sample was from 20 to 40 thousand dollars

USA.

A. The Diabetics

Medical “‘control”’ in these young, insulin-

dependent diabetics was defined by their

physician (MMM) as “‘approaching physi-

ologic norms.”’ Naturally, this ideal was not

always achievable in all of these patients.

Urine glucose was maintained as close as

possible to negative, based on frequent daily

patient testing and recording. Total 24-hour

urinary glucose excretion was monitored at

regular intervals with the goal of keeping it

to less than 10 gm a day. Glycosylated he-

moglobin values were also followed, and

an attempt was made to stabilize these levels

as close to the normal range as possible.

Dietary carbohydrate intake was regulated

to allow for normal energy and activity lev-

els. Thus carbohydrate intake was not re-

stricted, but rather arranged to provide

CARIES: A METABOLIC VIEWPOINT 101

adequate nutrition for growth and devel-

opment. The usual daily pattern consisted

of 6 measured carbohydrate ingestions, in 3

meals and 3 snacks, providing approxi-

mately 100 gm of carbohydrate daily plus

10 gm per year of age.

Most young children requiring less than

30 or at most 40 units per day (less than one

unit/kg) could be well controlled by asingle

injection of mixed regular and NPH insulin

daily. Those not well controlled with this

regimen were treated with 2 injections of

mixed regular and NPH insulin in a pro-

portion of 1:1 with approximately * of the

dose injected before breakfast and ‘4 before

dinner. Most patients over 10 years of age

were on the latter arrangement.

Some patients in the series were more

amenable to control by virtue of their

temperament, disposition, stage, or severity

of disease. Others were less cooperative,

had not been under the physician’s (MMM)

care for as long a time, or may have been

admitted to the practice at different ages or

different stage of their disease. Nevertheless,

the management, once instituted was more

intensive than in most such series, with visits

to the physician occurring every two months

in most cases.

A further measure of the effectiveness of

treatment was the growth pattern of these

childern. Insulin-dependent diabetic chil-

dren frequently fail to reach expected heights

and weights consistent with their arents’

heights or other expected values.”” In this

sample, overall medical management was

Strict, and the patients’ growth curve values

for height and weight were maintained at

their prediabetic slopes for norms as de-

termined by the National Center for Health

Statistics. Therefore, growth retardation

does not seem to have been a factor in the

diabetic childern in this study. In addition,

tooth eruption status in all patients was

neither retarded nor advanced for their

ages.

Generally, insulin-dependent diabetic

children cause concern to both parents and

clinicians entrusted with their care. At this

socio-economic level, such patients’ total

health needs are usually very carefully

overseen. One expects that regular dental

care would receive a higher priority in this

group than among the general population.

Both on the basis of dental history and clini-

cal examination, this premise proved valid

in this diabetic sample.

B. The Controls

After the diabetic series had been as-

sembled, control patients matching them

as closely as possible were chosen from pa-

tients coming in for dental checkups.

Pragmatically, no control group, identical

save for one factor, is ever available for

comparison. Diabetics are “‘different’’ by

virtue of their pathology. The disease pro-

cess and its effects are complex, going

beyond biology and affecting individual

socialization and family culture. In such a

situation, the really significant variables

may not be recognized. Nevertheless, after

matching age, race, geographic area, fluor-

idation, quality of dental care and family

income as a marker of socio-economic sta-

tus, it was hoped that the differences in car-

ies rates observed were associated with the

diabetic state.

C. The Statistical Analysis

A variety of statistical procedures were

employed to accommodate the problems

associated with variations in number of

teeth and years at risk. As deciduous teeth

erupt and are gradually replaced by per-

manent teeth, both the number, age, and

kind of teeth vary. For a newly erupted

tooth the cumulative exposure to caries is

less than for one already present. A table of

‘‘risk-years’’ was calculated, using as fac-

tors the modular ages of tooth eruption and

loss, and dividing total risk-years by the

number of teeth present to estimate the

‘“‘“expected”’ average years of exposure to

caries (Table 1). Using these figures, inter-

polating as needed, a Chi square score

[(o — e)’/e] was calculated for each indi-

vidual.

Because tooth eruption and growth pat-

102

terns were comparable in these diabetics

and controls, use of the same “‘expected”’ is

acceptable. This device, however, has limi-

tations: At 12 years of age, for example, the

“‘old”’ teeth, such as the relatively caries-

resistant incisors, tend to weight the risk-

years inordinantly, as pit and fissure caries

predominates in the molar and premolar se-

ries. Nevertheless, this ratio provides a use-

ful means of comparing caries experience.

A further problem is the great variability

in caries experience. In both diabetics and

controls, the number of carious teeth varied

from 0 to 16. Some deciduous dentitions had

many carious teeth; some permanent denti-

tions had none. This problem was dealt with

first by transforming both ages and numbers

of carious teeth, for both samples, into mov-

ing averages (nm = 6), and graphing these

transformed data, along with the risk-years

data just described (Figure 1). Here also

Chi square tests were used to estimate good-

ness of fit.

In analysing the data, the samples are di-

vided into three subseries: preadolescents

up to 8 years, 10 months of age; young ado-

lescents from 9 to 14 years, 11 months, and

15 MOVING AVERAGE (N=6) AGE vs NUMBER

14 OF CARIOUS TEETH

aS RISK YEARS/TEETH

12 ¢ DIABETICS.-------------

4 © CONTROLS “rstssetesseceeseeeee

10 4

WwW

e 8 @

”

yy & @

ro)

=n) e 4

Oo 5 e e wt eeecce %,

i . By >< a 4

3 ene a. °

r) *, ‘ 4~

stark cin (telat / Wnts

1 ist ee ¢ ¢ e

Of gece? —9------ re % 6 ec?

3 4 5 6 “| 8 8 10 11 1

2. a3) 44) ais

AGE (YEARS)

R. T. KORITZER, et al.

those 15 to 27 years old. These age divisions

were Selected because they represent both

hormonal events (the prepubertal growth

spurt and onset of puberty), and dental

events (the completion, except for the wis-

dom teeth, of the permanent dentition).

Mean ages for the whole series, as well as

the subseries, were not significantly different

(Table 1).

Results

Examination of Figure 1, in which the

moving averages (n = 6) for age and caries

are presented along with an “expected”

Caries estimate (derived from total risk-years

divided by number of teeth in the dentition)

shows a number of interesting features. In

both diabetic and controls, deviations from

the “‘expected”’ estimate are greatest up to

age 8 or 9. Although neither group has as

many carious teeth as expected, the controls

approach the expected more closely. After 9

years, the moving average graphs for both

series conform surprisingly well with the

““expected.’’ In fact, when the diabetic curve

¢ e <a

7 —

e =

@ e ¢ _

° nd. Pgs ay

pA mS 10

@ Ss ‘ —

H ne oe

! eee ae

EA ete ion

Up ACO an peo

YR OS =

° we. =

c wise

yf .

Jas % om

oe —5

o's Re /

Pe Cr

Coie e =

/ _—

e e ¢ =

e e —

o —oO

Fig. 1. Numbers of carious teeth in young diabetics and controls. Raw data and moving averages (n = 6) are

plotted against an ‘‘expected’’ number of carious teeth (solid line). See Table 1 for explanation.

CARIES: A METABOLIC VIEWPOINT 103

is compared with the control, little differ-

ence is seen, and deviations from the ex-

pected, as described by the Chi square total

are so small for each series as to suggest that

the curves may be quite similar.

If one looks only at the caries experience

for the total series (Table 1d), there seems

to be little difference between diabetics

(138 carious teeth, or 4.45 per person) and

non-diabetics (152 carious teeth, or 4.90

per person). However, when these series

are subdivided by age, differences appear.

Among the9 preadolescent diabetics, there

is One carious tooth (Table la), whereas the

10 controls had 30 carious teeth (0.1 vs. 3.0

per person). In the age range 9 to 14, the 8

diabetics have an average of 3.6 carious

teeth against 2.7 for the controls (Table 1b);

and in the 15 and over group, the diabetics

have an average of 7.7 versus 7.4 for the

controls (Table Ic).

The distribution of the raw data, how-

ever, plotted along with the moving aver-

ages, Suggests that patterns of variation

differ in the diabetics and the controls.

Closer inspection of the Chi square scores

in Table 1 confirms this impression. At all

ages, more diabetics than controls deviate

significantly from the “expected,” and the

deviations are more marked. For the total

series, average Chi square for the diabetics

is 3.15; for the controls 1.97.

Except for the youngest group, the “‘ex-

pected” proves an unexpectedly good pre-

dictor of caries experience in the controls:

Regarding an individual with a Chi square

score of 0-1.5 as falling into the “‘expected”’

range, 50% of the youngest control children

have the “‘expected’”” number of carious

teeth, with 40% having fewer and 10% more

than expected. In the 9-14 group, 57% have

the “expected,” with 29% having fewer and

14% more than the expected number of car-

ious teeth. In the oldest control group, 64%

have the ‘“‘expected’’ number of carious

teeth, with, again, 29% having fewer, and

7% more.

The diabetic Chi square scores are in

marked contrast to those of the controls.

All of the youngest diabetics, with an aver-

age Chi square of 2.53 (controls, 1.54),

have fewer than the expected number of

cavities. The older diabetic children, witha

mean Chi square of 3.3 (controls, 1.7), have

only 25% in the “expected’’ range, but

37.5% with fewer, and 37.5% with more

than the expected. The oldest diabetics, with

an average Chi square of 3.47 (controls,

2.42), have only 43% in the “expected”

range, with 36% having fewer and 21%

having more than the expected number of

carious teeth.

In short, although the total frequency of

carious teeth, and the mean number per

person may be similar in the diabetics and

the control series described in this pilot

study, distributions are quite different: In

each age group, fewer diabetics conform to

the expected, but have either significantly

more or significantly fewer carious teeth

than expected. Controls, on the other hand,

tend to conform more closely to expecta-

tions, except for some 30% in all age groups

who have fewer carious teeth than expected.

Discussion

Because both high and low caries rates

have been reported in young diabetics, all

potential working hypotheses have been

entertained in this pilot study. We must,

therefore, consider the possibility that dia-

betics might have more, the same as, or less

caries than controls. A related hypopthesis

is that caries rates may change with age, and

may be lower in diabetics than in controls at

one age, but equal or higher at another stage

of life. If this were the case, reported rate

differences might be partially an artifact of

the age ranges of the samples reported. It is

also possible that, at any given age period,

diabetics may be more deviant—some hav-

ing markedly more and some markedly

fewer carious teeth—than controls.

Lower caries rates in diabetics have been

attributed to retarded dental eruption, re-

ducing exposure to caries-producing en-

vironments.” This does not seem to have

been a factor in our study. Growth, includ-

ing tooth eruption, seems to have been

within normal limits in both diabetics and

104 R. T. KORITZER, et al.

controls. Caries resistance in diabetics at-

tributable to carbohydrate restriction, sim-

ilar to that reported in other studies, was

not observed here. Difference in carbohy-

drate intake does not seem to have been a

significant contributor to lower rates in the

youngest diabetics. Neither growth rate,

carbohydrate intake, or other environmen-

tal factor, can account for the marked vari-

ation in our diabetics from 9 to 27.

Among early onset diabetics, it becomes

clear that disease duration and age at onset

are not totally independent. In this pilot

study, age changes in caries rates appear to

differ for the two samples chosen. Assuming

the controls to be satisfactory, the differ-

ences observed may be construed, at least in

part, as related to the diabetic state. Al-

though our samples are small, we have, nev-

ertheless, attempted to remove conflicting

variability and to isolate the effect under

consideration. On the diabetic side of the

equation, duration and age at onset might

have caused variability."*° Although one

might, therefore, expect to find the smallest

effect of these variables in the youngest

children, this group is indeed the most

different.

The apparent caries effects previously

reported for diabetics seem to be related to

age selection. Since many reported samples

deal with a limited age range, followed fora

limited number of years, only portions of

the 2 to 27 year curve have been examined.

Thus, age changes in such samples are not

evident and their directionality is maxi-

mized. Such studies would indeed report

higher as well as lower caries rates as noted

in the curve crossovers seen in Figure 1.

Summary and Conclusions

To our knowledge, caries effects in dia-

betics associated with hormonal events

have not previously been suggested. The

age distribution of diabetics in previous

studies has probably not allowed observa-

tion of this phenomenon.

When the series were divided by age, var-

ious differences appeared: The youngest

diabetics had fewer carious teeth than the

controls of the same age, and the older dia-

betics, on the average, more. There seems

to be a natural division at about age 8 for

diabetics. The insulin-dependent diabetics

in this pilot sample, nevertheless, ultimately

achieved the same mean caries as the control

group. When individual caries experience,

however, is compared to that predicted on

the basis of mean risk years per tooth, differ-

ences are seen. Except for about 30% of the

control sample, who have fewer than ex-

pected, controls have approximately the

caries predicted. Variability in caries expe-

rience, however, is much greater in diabetics,

with more than 50% of the sample having

either too many, or too few carious teeth.

The succinct point of all this is that caries

rates of early onset insulin dependent dia-

betics and reasonably matched controls

fluctuate somewhat independently when

sufficient age range is included. The tenta-

tive conclusion is that caries rate variation

in these well controlled diabetics may be

associated with metabolic aspects of their

disease rather than with external factors.

Further study is required.

Literature Cited

1. Bernick, Sheldon M., D. Walter Cohen, Lester

Baker, and Larry Laster. 1975. Dental disease in

children with diabetes mellitus. J. Periodontology,

45(4), 241-245.

2. Matsson, L. and G. Koch. 1975. Caries frequency in

children with controlled diabetes. Scand. J. Dent.

Res., 83, 327-332.

3. Sterky, G., O. Kjellman, O. Hogberg and A. L.

Loforth. 1971. Dietary composition and dental dis-

ease in adolescent diabetics. Acta Paediatr. Scand.,

60, 461-464.

4. Stadtler, P., M. Sulzer and P. Petrin. 1978. Zum

Kariesbefall jugendlicher Diabetiker. Wien. kli-

nische Wochenschrift, 90, 844-847.

5. Vassileva, S. 1975. Clinical-statistical study upon

caries in children with diabetes mellitus. Stomato-

logija (Sofia), 57(1), 22-28.

6. Wegner, H. 1975. Increment of caries in young dia-

betics. Caries Research, 9, 91-96.

7. Craig, Oman. 1977. Childhood diabetes and its

management. Butterworths, R. J. Ackford, Ltd.,

Chichester, Sussex.

8. Grave, Gilman (ed.). 1979. Early detection of poten-

tial diabetics, the problems and the promise. Raven

Press, New York.

Journal of the Washington Academy of Sciences,

Volume 72, Number 3, Pages 105-107, September 1982

The Scientific Awards of the Academy: 1982

Sherman Ross

General Chairman

The scientific Achievement Awards of

the Academy were presented at the Annual

Meeting on May 19, 1982 at the Cosmos

Club, Washington, D.C. Two awards were

made for significant contributions to re-

search, and one award for science teaching.

This program of the Academy was started

in 1939 to recognize young scientists for

*. . . noteworthy discovery, accomplish-

ment, or publication in the Biological,

Physical, and Engineering Sciences.’’ An

award for Outstanding Teaching was added

in 1955 (renamed in 1979 as the Leo Schubert

Award), and in Mathematics in 1959. In

1975 the award for the Behavioral Sciences

the Berenice G. Lambert award for Teach-

ing of High School Science was started.

Engineering Sciences

Allen Plotkin recieved the B.S. (1963)

and M.S. degrees (1964) from Columbia

University, and the Ph.D degree in 1968

from Stanford University all in Applied

Mechanics. He started as an assistant pro-

fessor in 1968, was promoted to associate

professor in 1972, and made a full professor

in 1977 in the Department of Aerospace

Engineering at the University of Maryland,

College Park.

He has made research advances in a

broad range of problems in theoretical fluid

mechanics with application to aerodynam-

ics and hydrodynamics. He has published

widely in journals serving the aeronautics,

applied mathematics, mechanics and hydro-

Nautics communities and is currently a

member of a national fluid mechanics

committees of both the AIAA and SNAME.

A significant aspect of Dr. Plotkin’s re-

105

search efforts is the stress on the impor-

tance of analytical techniques and the tools

of applied mathematics in the solution of

complex fluid flow problems. The most ra-

pidly growing area in fluid mechanics re-

search is computational fluid dynamics,

with emphasis on large-scale computations

in many cases. In his current research into

the description of laminar flow separation,

one of the last major unsolved problems in

aerodynamics, he is using a semianalytical

spectral method to reduce the order of the

governing differential equation set to pro-

vide insight into the problem, which might

suggest efficient numerical approaches to

the solution.

Dr. Plotkin has used a wide variety of

techniques to deal with the fluid mechanics

of bodies moving through air and water. He

has made extensive use of the methods of

matched asymptotic expansions and per-

turbation theory to extract accurate (in

many cases higher order) analytical ap-

proximations to complicated viscous and

inviscid flow problems. He had also made

contributions in computational fluid dy-

namics beginning with his dissertation re-

search in the 1960s.

Some of the major problem areas that

have benefitted from Dr. Plotkin’s Contri-

butions are:

Laminar jet flow on curved surfaces

Variable-depth shallow water flows

Shallow-water ship motion theory

Thin-hydrofoil theory

Hydrofoil cavity flow

Laminar flow separation

Unsteady subsonic and transonic thin-

airfoil theory

106 SHERMAN ROSS

For these contributions, Dr. Plotkin has

been selected for the Scientific Achievement

Award in the Engineering Sciences.

Mathematics and Computer Sciences

Larry S. Davis is a graduate of Colgate

University (B.A. in Mathematics, 1970). He

completed graduate study at the University

of Maryland, College Park, and received an

MS and Ph.D. degree by 1976 in Computer

Science. From 1977-81 he served as assist-

ant professor in the Computer Sciences

Department at the University of Texas, and

in 1981 was appointed an Associate Profes-

sor at the University of Maryland.

Dr. Davis works in the 25 year old field of

digital image processing and analysis, which

deals with the computer manipulation,

description, and recognition of pictorial in-

formation. Examples of its applications in-

clude document processing (character rec-

ognition), biology and medicine (cytology,

radiology, etc.), industrial automation (in-

spection, robot vision), and remote sensing

of environmental factors, (natural re-

sources, crops, cultural features, etc.). This

area is now a major branch of computer

science.

Dr. Davis entered this field as a graduate

student less than ten years ago, and has

rapidly become one of the leading individ-

uals. His work has primarily been technique

oriented rather than application oriented,

and his list of publications includes contri-

butions to many different aspects of the

subject.

The basic task of digital image analysis is

to generate a description of an image (given

in the form of a discrete array of brightness

values) in terms of significant parts (regions,

features), and their properties (such as size,

shape, “‘visual texture’’, etc.). This process

of identifying significant parts is known as

segmentation. Dr. Davis has developed a

variety of basic segmentation techniques

based on the identification of regions that

are homogeneous either in brightness or in

texture. He has also done work on the con-

verse problem of detecting transitions be-

tween pairs of such regions, “‘edge detec-

tion’’. Several of his publications deal with

the detection of other types of distinctive

features in images, including lines, curves,

and “‘corners’’.

In the area of shape description and rec-

ognition, Dr. Davis’ contributions, which

began with his Ph.D. thesis, are quite fun-

damental. He introduced a method of hier-

archical polygonal approximation for rep-

resenting shapes at multiple levels of detail,

and for detecting partial symmetries. He

then applied this approach to the problem

of approximate shape matching. Another

basic contribution, in collaboration with

an architect, relates to the analysis of

shapes in terms of regions of visibility aae

of distances from the border.

More recently, Dr. Davis has done ex-

tensive work on the computer analysis of

visual testure. His principal contributions

concerned texture models based on random

geometric processes, texture analysis based

on second-order statistics of local image

features, and_ periodicity/directionality

analysis of textures.

He has also done work on adaptive image

quantatization, noise cleaning, constraint

analysis and its uses in computer vision,

pattern databases, and analysis of time-

varying imagery. His current research is

concerned with this last topic, as well as

with the development of “‘expert systems”’,

based on rules of inference, for the analysis

of satellite imagery.

These efforts are major contributions to

computer vision and digital image process-

ing. For these accomplishments Dr. Davis

has been selected for the Scientific Achieve-

ment Award in Mathematics and Computer

Sciences.

Berenice G. Lamberton Award

Mrs. Gloria B. Speroni is a member of

the science faculty at Anacostia High

School, Washington, D.C. She was born in

Milford, Massachusetts, completed her

undergraduate work at the College of St.

Elizabeth in New Jersey, graduating with

> oer DL > ae AE ee 2, ae

THE SCIENTIFIC AWARDS OF THE ACADEMY: 1982 107

honors in science in 1958. Her Master’s de-

gree in Biology in 1960 was from New York

University, where she served as a Teaching

Fellow. After a year of teaching in Sacra-

mento, California, she came to Ballou

High School in Washington. She has been

at Anacostia High School since 1968, teach-

ing Biology and Chemistry.

She exhibits the characteristics of an out-

standing teacher in many ways. She has

made notable efforts to individualize in-

struction for students with different skills,

interests, learning styles, and educational

goals. She has produced individualized

modules in Chemistry and Biology courses,

which use multi-media, large/small group

patterns, and self-pacing strategies. She has

begun the programming of module com-

ponents for use, when a computer terminal

becomes available. Her teaching behavior

is consistently friendly, enthusiastic, con-

cerned and professional.

She has published a paper on team teach-

ing in The Science Teacher, and on an indi-

vidualized approach to BSCS Biology in

The American Biology Teacher. She received

an Outstanding Biology Teacher Award

from the National Association of Biology

Teachers in 1975.

For these contributions, Mrs. Speroni has

been selected to receive the Berenice G.

Lamberton Award for the Teaching of High

School Science.

Acknowledgments

The contributions of the chairmen of the

various panels and their colleagues, who

carried out the difficult task of making the

selections, are acknowledged with sincere

thanks. The chairmen were:

Dr. John J. O’Hare—Behavioral

Sciences

Dr. Kun-Yen

Sciences

Dr. John D. Anderson, Jr.—Engi-

neering Sciences

Dr. Joan Rosenblatt— Mathematical

& Computer Sciences

Huang—Biological

Dr. Mary H. Aldridge—Physical

Sciences

Dr. Joseph B. Morris—Teaching of

Science

Thanks are due to the nominators and to

the sponsors of all the candidates.

On behalf of the Academy we com-

mend the recipients, whose work is

honored, and we wish them con-

tinued productive careers.

oer i ig Rl Ae in

| be omer ntiee Mair

Rh wes yan i : ane Ait

Age ,

LNs aa a aA

b: ama bla

nie Aa ‘aad at

at mat aT ee {ane net hs

my 7 q ‘a 0

>. neve Oe

aeiwerth se 4 ; ) -

di tee tad

| > THE WARING TOM &; apis

- NT bhi (5 THe ea vs i ‘ 4 '

DELEGATES TO THE WASHINGTON ACADEMY OF SCIENCES,

REPRESENTING THE LOCAL AFFILIATED SOCIETIES

NE SUHSUIEIGION cg tcc ccc nensieuedewsecucccvsauscusenacecevveees James F. Goff

ane acrety OF Washington ........0cccecscccvercncsccucuceececesners Ruth H. Landman

SE MUERITIIGEOT) civics ccs e cus svascencuvsccseeuareensveceecse William Ronald Heyer

ETON og 5c kc a oe bc vancvudaudWhinedesteceseevoeneveevane Anne Brown

EIEN €0E WOSPINGTON oo ccc cc cece ne ceccecccccencecceccuseeveneas Margaret Collins

en ee ee T. Dale Stewart

EE TUESIITICTON. ccc ccs ccc csccaccucedecececweccevucttcseens James V. O’Connor

ry Or the District Of Columbia ..........ccccccccccccccntcvvenacees Charles E. Townsend

I occ, wa. Wey a Gas Sa Nee he SAN AEE OWE MEE RES Se eaek Paul H. Oehser

ENT NLER ET 500.0) 655 50s Sok si incw's wins, Soe RU eee a Ne Ried Oa w ER a ee Rene Conrad B. Link

a e550, EG GIG bs Kina w ea WLR LRLG ao ARAM RUMI SANE a Boyd W. Post

ICES chee eid ors casein es sees dine wee eee oe kel Mente anew Wer George Abraham

Seeeeuectiical and Electronics Engineers ..........00cccccenscccscsescescees George Abraham

meneey Ot MeChAniCal ENPINECers...... 2... ccc cece ce ccsccccvcececuceccvecsce Michael Chi

ear erety OF WaSHINGtON ...... 0.6 cc csc rec c ene vencaeentetscweants Robert S. Isenstein

EGE NMICTOBIONOCY . 5. occ cc cece ence wens ceneewecececeuneens Lloyd G. Herman

Se EEMECM THIINCATY ENQINGEIS..... 2. c sec cee eeu cnueeeuseucessusecerens H. P. Demuth

Se PCMEROCIUIETIDINGEDS 5.00. sis cies cscs act ccceacecntvebesccovcceeenene Wallace J. Cohen

eee enerimental Biology and Medicine............ccccccccccecscvccvcccvcs Cyrus R. Creveling

NEUES ea a roe ood g ok asa aie eis See ad's Ow EE) ae Seb ea eee wee Charles G. Interrante

ememciaion Of Dental Research .<.........0.ccccencencccssvcccvcsesccs William R. Cotton

American Institute of Aeronautics and AstronauticS..........0cc cee cece eeeeeeeees Richard P. Hallion

I NUINIEMIE TE ICICIELY 2 wes csc sea ceded avis asia ed os ves bee ee S neice s A. James Wagner

ENON ASIPIPEON . 2c. 5. bos ce sce tbe gebddeecceeecscededuavepewent Jack R. Plimmer

RE NET as oie. 5S. a esas bade. veiw ole bee ewe ee bos sales We elas Ree Richard K. Cook

gon Sos ous weg odio we Pao ole e ss eos Ueno ed eeeueciny son tee Dick Duffey

Se EEMEMEMGDISS ecco ds Ue ae kes nda eee edb as Paes eine ede Cee wee A. D. Berneking

IE rao af chcs oa a5 a dw don aa aun winiia ReolGed Seale midland w wate Saale Meda Edwin R. Fuller, Jr.

a ty ica aie 'n apes ve. w ©. oops edanlea iol ha elB prem cane wlacaie Alayne A. Adams

RRP SCICHCE CHID: oj... 52 do cine cease noes vedas econeeceveeeeaans Deborah J. Warner

EPEC MeanCO? PHYSICS LEQCHers .... 2... .0.c cee cece eeecwsieescescuusesevcnse Peggy A. Dixon

NW AA. ok sv Ba cig acdiiate wldibni-ood oa hays wd weon uateterel w guarale George J. Simonis

emer Plant PHYSIGIOPISES :........06ecccnncdodeccuasnesssubeses Walter Shropshire, Jr.

Enews Research Council. .........0.ccccccccccccecccsccecancvensnns John G. Honig

EEE G02 2 oo ca han Goa 20s lacus Aieoy cc OW stwydia swine Swe eae a erage Jewel B. Barlow

American Institute of Mining, Metallurgical

RTC Sas erp he a hss asa aa cine shal leu alnaa. ¥ Riel be RAR PRN NIN Garrett R. Hyde

IE SSSI. © 55 acted oP ko Ck Ade on nk Ge athe ew eden Sub nsee Benson J. Simon

EMME ARES CIS AUINICLICA sss ove cis ed edie ae died dnt adpadedaenseawasiees ann Patrick Hayes

TEES Nr i GD a Miloslav Racheigl, Jr.

EERE SS MAS pe a ee ae H. N. Reynolds

ee Paint Techitiical Group .........ceccccccccsncvecvecsscensuscences Paul G. Campbell

Seemmtanathiological SOciely ..........cccccccccccccnccccecvsrecenses Howard E. Waterworth

Sen ascreral Systems Research ...........0cccccccccsccecercaccecces Ronald W. Manderscheid

EE EI on ae Stanley Deutsch

OR ei dc Gia diode c UaWics Ghee ws scan chadeeavewe kote Irwin M. Alperin

mssociation for Science, Technology and Innovation ...........2ssscceccccsccccssccsees Ralph I. Cole

Delegates continue in office until new selections are made by the representative societies.

Washington Academy of Sciences

1101 N. Highland St.

Arlington, Va. 22201

Return Requested with Form 3579

2nd Class Postage Paid

at Arlington, Va.

and additional mailing offices.

REHDER,

HARALD A. (DR)

5620 OGDEN ROAD

BETHESDA, MD 441

20816 WAS-F84 0074

) HSONTAy~

. / Fj \ VOLUME 72

N 400% Number 4

ral of the x JUN 6 € 464 December, 1982

_LIBRARIES

wAaSHINGTON

ACADEMY .-SCIENCES

ISSN 0043-0439

Issued Quarterly

at Washington, D.C.

CONTENTS

Articles:

women J. SEEGER: On the Humanism of Science ..........002.000008

JAMES B. COULTER: Bane of the Bay—Chlorinated Sewage ..............

Meena RAPHAEL: The World Energy Unit... .....0.00secceseeccns as

Washington Academy of Sciences

Founded in 1898

EXECUTIVE COMMITTEE

President

James F. Goff

President-Elect

Jean K. Boek

Secretary

John J. O’Hare

Treasurer

Ronald Manderscheid

Managers-at-Large

Jo-Anne Jackson

Grover C. Sherlin

Alfred Weissler

Elise A. Brown

James Wagner

Ralph I. Cole

BOARD OF MANAGERS

All delegates of affiliated

Societies (see Cover 3)

EDITORS

Irving Gray

Joseph Neale

ACADEMY OFFICE

1101 N. Highland St.

Arlington, Va. 22201

Telephone: (703) 276-9254

The Journal

This journal, the official organ of the Washington —

Academy of Sciences, publishes historical articles, critical

reviews, and scholarly scientific articles; proceedings of

meetings of the Academy and its Board of Managers; and

other items of interest to Academy members. The Journal

appears four times a year (March, June, September, and

December)—the September issue contains a directory of

the Academy membership.

Subscription Rates

Members, fellows, and patrons in good standing receive

the Journal without charge. Subscriptions are available on ~

a calendar year basis only, payable in advance. Payment

must be made in U.S. currency at the following rates:

U:S. anid Canada>..2.2. $19.00

FOEIOR Si. oi ouly Ceres 22.00

Single Copy, Price 25 .: 7.50

UHURU Symposium—2 issues............

Delivery by: Air 23/22 2.00

Single-copy price for Vol. 66, No. 1 (March, 1976) is $7.50.

Back Issues

Obtainable from the Academy office (address at bottom

of opposite column): Proceedings: Vols. 1-13 (1898-1910)

Index: To Vols. 1-13 of the Proceedings and Vols. 1-40 of

the Journal Journal: Back issues, volumes, and sets (Vols.

1-62, 1911-1972) and all current issues.

Claims for Missing Numbers

Claims will not be allowed if received more than 60 days

after date of mailing plus time normally required for

postal delivery and claim. No claims will be allowed ©

because of failure to notify the Academy of a change in

address.

Change of Address

Address changes should be sent promptly to the Academy —

office. Such notification should show both old and new

addresses and zip number.

Published quarterly in March, June, September, and December of each year by the Washington

Academy of Sciences, 1101 N. Highland St., Arlington, Va. 22201. Second class postage paid at

Arlington, Va. and additional mailing offices.

Journal of the Washington Academy of Sciences,

Volume 72, Number 4, Pages 109-119, December 1982.

On the Humanism of Science

Raymond J. Seeger

National Science Foundation, Retired

You and I live in one experiential world.

Each of us, as subject, experience a whole

view of the world as object. In every in-

stance, however, three different points of

view are distinguishable. Whether we look

at a planet, at a plant, or at a person, we

may do so from the overlapping viewpoints

of aesthetic enjoyment, scientific related-

ness, or technological use, or possibly also

of mystical participation. Three mountain

peaks afford different cultural outlooks.

On the one hand, lie the twin peaks of intel-

lectualism, comprising the humanities and

the sciences; on the other hand, is the

neighboring peak of practicality. The cele-

brated “‘Two Cultures” of Charles Percy

Snow are actually the academic twin peaks

(he admittedly does not consider the prac-

tical peak, which, indeed, is separated from

the intellectual ones by a greater valley

than that between the twin peaks them-

selves). In line with his vision people have

focussed their own attention upon the gap

between the twin peaks and pondered the

necessity of a bridge between them. I prefer

to fix our gaze upon the mountain roots in

the common valleys below, namely, the

underlap which forms a natural bridge be-

tween the peaks. One such underlap is the

humanism of science, which is our major

consideration today, particularly in view of

the dire prediction of Herbert Read in ‘‘The

World in 1984” (1966): ‘‘There will be light

everywhere except in the mind of man, and

the fall of the last civilization will not be

heard above the incessant din.”

Although each educational edifice will

differ in its functional design, it must con-

109

sist of the same essential components and

have the same structural foundation. The

omission of technological use from our dis-

cussion here should be regarded not as evi-

dence of the sufficiency of any one of these

types, but merely as an indication of the

limitation of the brief time allowable for

our consideration. Nevertheless, in pass-

ing, I do wish to call attention to two im-

portant aspects of the technological com-

ponent of all education.

In “‘The Scientific Estate’ (1965), Don

Price identifies four so-called estates: scien-

tific, professional, administrative, and po-

litical, all of which must work together in

modern society. Thus, science certainly is

becoming more and more a significant fac-

tor in social change, as evidenced in spread-

ing urbanization, in higher living stand-

ards, and in multiplying population.

Increasingly, it is being exploited for its po-

tential contribution to the solution of so-

cial problems such as developing countries,

population explosions, war and peace. With

a profound faith in science, modern pil-

grims travel to Utopia along a road paved

superficially with technological progress.

They chanta litany of George A Lundberg:

“Can Science Save Us?” In the dim light

of social science, he shouts hopefully,

*“Yes!’’ Price, however, cautions wisely

that ‘‘science alone cannot solve political

problems.”

Moral dangers, moreover, are intimately

associated with technological development;

for we all live in a moral world. We are

rightfully optimistic that the universe has

been revealed by science to be somewhat

110 RAYMOND J. SEEGER

intelligible, and by technology to be in-

creasingly habitable. Yet, we also cannot

help being anxiously pessimistic over the

practicability of man’s control of his tech-

nology. Rebellious Prometheus, you may

recall, although venerated by man because

of his theft of fire from the Greek gods, was

banished by them to a lonely rock in the

Caucasus because of their fear that man

would not be able to control this consum-

ing peril. Science, to be sure, is inherently

neutral. Yet, a knife in the hands of a bad

man, can stab a good heart; in the hands of

a good physician, that same knife can cut

out a bad appendix. The knife itself is

neither good nor bad, but it can be used for

good or for bad. So, too, science can be

used for benevolent or malevolent pur-

poses by people, including scientists, who

are necessarily members of social groups. It

was truly “‘we the people”’, all of us, who de-

vised, constructed, and dropped the atomic

bomb, not simply one evil-minded or psy-

chologically unbalanced estate of our

complex democracy. Technology, accord-

ingly, is potentially good or evil. Do you

recall the story of the little boy and the old

man, who was supposed to be very wise?

The boy thought to himself, “‘I will test the

old man. I will take a bird in my hand and

ask the old man if the bird is alive or dead.

If he says, ‘alive’, I will crush the bird at

once. If he says, ‘dead’, I will allow the bird

to fly away.’ So he came to the old man

and asked, “‘Is the bird in my hand alive or

dead?” Said the wise old man, “‘As you will,

my son; as you will.”

Finally, the foundations of our modern

educational structures, the images of life,

are rotting with moral termites: caricatured

truth exhibited brazenly in the ‘‘Advise and

Consent” of Allen Drury’s Washington,

psychopathic sex smoldering secretly in

**The Group” of Mary McCarthy’s Vassar

College, and ““The Graduate”’ of Charles

Webb’s Williams College, impotent reli-

gion confessed shamelessly by godless theo-

logians, new and old.

In education, therefore, as in life, all

three aspects of our one world of expe-

rience ultimately must be considered in re-

lation to one another. Let us turn now to

our understanding of the multivalent term,

“humanism”.

Students of the so-called humanities often

display a narrow humanism with respect to

science. Moody E. Prior, for example,

claims in ‘“‘Science and the Humanities”

(1962), “‘The proof that science and human

values are related cannot be found in the

conduct of scientists who have advanced

knowledge, or in the motives of those who

have participated in scientific learning.” In

‘““Science and the Creative Spirit” (1958),

John B. Priestley states categorically, “It is

the scientists who deny the validity of po-

etry and who ignore it.’ (The very opposite

is revealed in Jacob Bronowski’s “Insight”

(1964).) In the same book, Harcourt Brown

boasts, “‘We may describe the humanist as

one whose object of study is the creative

imagination of man, in its efforts to pro-

duce objects and material effects.’’ Nota

bene: he does not include the scientist

among humanists. And yet, with reference

to scientific connotations, one wonders

how literature can even be read meaning-

fully unless these are interpreted from the

scientific viewpoints both of the past and of

the present. In ‘‘The Place of Science in

Liberal Education”’ (1959), Ernest Nagel

concludes, “‘The study of science commonly

is not regarded as contributing significantly

to the development of mind, sensitive to

qualities that ennoble and enrich human

life. . . Science is too important to be left

only to the concern of scientists.” The hu-

manities, however, may be equally provin-

cial, as when they are concerned with mak-

ing a living more than a life. (In this

connection, may I express my aversion to

all so-called terminal courses which admit-

tedly do not even try to prepare people for

continuous growth in life.) In ““The Shame

of the Graduate Schools—A Plea for a

New American Scholar” (Harpers, March

1966), while William Arrowsmith deplores

the psuedo-scientific spirit common in

academic courses in literature, at the same

time, he reveals his own distorted view of

science: ‘“‘The sciences aim at knowledge,

and the student in the sciences is appro-

ON THE HUMANISM OF SCIENCE 111

priately an apprentice, his professor a

craftsman or technician. . . It [science ed-

ucation] is obviously effective and no more

barbarous than science education at the

undergraduate level.’’ The balanced judg-

ment of Martin Green in his “Science and

the Shabby Curate of Poetry” (1964), is

more enlightening, ““Modern physics has

presented itself as the ultimate achievement

of the impersonal mode of knowledge, and

thus at the opposite extreme of modern lit-

erature. . . they are both equally inacces-

sible to the average educated man, both

equally recalcitrant to humanistic treat-

ment.’’ No wonder Read predicts, “‘Poetry

already an archaic activity will have totally

disappeared.”’ Perhaps one should refer to

the report (1964) of the Commission on the

Humanities for a proper definition of that

subject: ““The humanities are the study of

that which is most human. . . the body of

knowledge is usually taken to include the

study of history, literature, arts, religion

and philosophy.”’ What! no science?

To me, all this is a false dichotomy, an ar-

tificial, man-made separation of our one

world of experience. Brown classifies ‘“‘the

humanist as self-communicating and vari-

able, and the scientist as objective and im-

personal.” Obviously he would not differ-

entiate between genetics and taxonomy in

this regard. In “‘Literature and Science”’

(1963), Aldous L. Huxley amplifies this

point of view, “His [scientist] primary con-

cern is not with the concreteness of some

unique event. . . but with abstracted gen-

eralization in terms of which all events of a

given class make sense.” Thus art is to be

regarded as concrete and unique, opposed

to abstract and discursive science. And yet,

in my Own experience some modern art is

certainly abstract and individualistically

discursive, to the degree that it might even

be called idiotic (i.e., very private—in the

Greek sense of that term). On the other

hand, what is more concrete and unique

than nature itself? Evidently Huxley, too,

has over-simplified the matter in attempt-

ing to force the changing mysterious uni-

verse into his own fixed classification molds.

There is, to be sure, a difference in empha-

sis. Art, indeed, is more subjective and

hence more individualistic; science is more

objective, owing to its allowance of toler-

ances, which thus makes available com-

mon standards for possible communica-

tion. At any rate, the self-styled humanities

and the natural sciences do not exhibit this

presumed contrast between creativity and

routine, between the imaginative and the

factual, respectively.

Let us consider, therefore, a broader

humanism, one that better reflects the his-

torical attitude of man himself. You may

recall the nosce te ipsum (“‘know thyself”’)

inscription in the Delphic temple, regarded

by Socrates as the message which he was

charged to deliver to his fellow-citizens. In

the American Scholar (1837), Ralph Waldo

Emerson interpreted this phrase to mean,

“Study nature. . . Thefirstin time and the

first in importance of the influences upon

the mind is that of nature.’’ The most cele-

brated statement is undoubtedly that of the

Roman Publius Terentius Afer: ‘““Homo

sum; humani nil a me alienum puto” (Iam

a man, I count nothing inhuman indiffer-

ent to me). And yet, every year the very

tourists who are most thrilled at the sight of

decaying Roman aqueducts are those who

have paid no attention whatsoever to the

greatly improved water supply in their own

cities. Another often quoted statement is

that of Alexander Pope in his ‘‘Essay on

Man” (1732-4): “The proper study of

mankind is man.”’ Rarely given, however,

is the relevant previous line which advises,

‘**Presume not God to scan.”’ The differen-

tiation here is between God and man, not

between man and his environment. Cer-

tainly, any separation of the latter, too, is

unnatural; yet, it is customary to idealize

glamorous Athens as the center of Greek

culture with its drama and philosophy, its

architecture and sculpture, but to ignore

science-breeding Ionia and its neighboring

isles. The Greeks never thought of these two

parts as culturally separated; they viewed

life as a whole. They were the first who

looked at nature and proclaimed, “‘It is

real, it is interesting, it is understandable!”

William Wordsworth’s musing unfortu-

112 RAYMOND J. SEEGER

nately still characterizes the popular atti-

tude of modern times: “‘The world is too

much with us; late and soon, Getting and

spending, we lay waste our powers: Little

we see in Nature that is ours.” As a working

definition of broad humanism, therefore,

let us follow Walter Pater in ““The Renais-

sance”’ (1873): ‘“‘The essence of humanism

. . nothing that has ever interested living

men and women can wholly lose its vital-

ity.”’ Evidently, this kind of humanism ra-

diates from an anthroprocentric point of

view. From this standpoint, then, what is

humanistic science?

Let us begin with the distinctive feature

of modern culture, the role of science and

technology, both of which have grown out

of humanistic soil, out of man’s interests,

propensities, and capacities, out of his

common humanity, Note, however, that

science and technology increasingly forma

single concept, a continuous spectrum with

distinctive ends, namely, science with un-

derstanding and intellectual abstraction,

and technology with use and moral power.

In the past, science and technology each

have initiated their own peculiar revolu-

tions. We can identify at least four major

technological revolutions, beginning with

the industrial revolution arising in the 18th

century. Man has sought power succes-

sively in mechanical energy (wind and

water), electrical energy, chemical energy,

and atomic (nuclear) energy. Out of such

revolutions have emerged many of our

practical problems, both social and moral.

So-called scientific revolutions, however,

have evolved largely out of ideas, begin-

ning with the renascence of science in the

late 16th century, when man realized that

nature itself could answer some of his en-

vironmental questions. Success bred the

19th century belief on the part of some in-

tellectuals that nature could answer all

human questions, and the 20th century

corollary that such answers can be ob-

tained only through the physical sciences.

Out of such revolutions, in turn, have

emerged theoretical problems, personal and

intellectual, intimately linked with philos-

ophy and religion. One other characteristic

of modern science and technology that is

becoming more and more important is the

decreasing time-lag of interaction between

both, so that some planning is now con-

ceivable, particularly in technology (but to

a less extent in science). Random seeding,

we are now aware, can produce only a ran-

dom harvest. Between the obvious ends of

this interrelated spectrum of science and

technology, there is an indistinct haze of

confusion hiding an ever threatening danger

that science may be regarded falsely, as

merely the handmaiden of technology.

Despite our vaunted new education,

public misunderstanding of science and

technology continues to spread like wild

fire. At home, for example, science com-

monly is assumed to be something foreign

to the pursuits of an average citizen, re-

served for unusually gifted persons like Al-

bert Einstein. On this premise, therefore,

one primarily needs to identify such rare

individuals and to encourage their profes-

sional development.

In the world at large, a scientist is fre-

quently portrayed as a freak with unusual

senses; with bulging eyes, or with spreading

ears, or with large hands, or with a brain

tumor (i.e., a swelled head). Caricatures of

scientists are publicized widely in cartoons

and science fiction. Even politicians take

time to make gibes. Upon reading an an-

nual report of the National Science Foun-

dation, certain Congressmen expressed

their amusement at titles of various re-

search projects being supported by the

Foundation. “‘Cabbage Sex Life’ was the

CATEGORY

FOOD

TRANSPORTATION

SPACE HEATING FROM SOLAR, ENVIRON-

WORLD ENERGY REQUIREMENT

QUADS/YEAR

NEVER CALCULATED

MENTAL, ANIMAL & BIOMASS SOURCES

ADDITIONAL HEATING AND INDUSTRIAL

PROCESSES

Had I been willing to put up with a

number of inconveniences, I could have

used the energy much more efficiently. In-

stead of using the coal to produce electric-

ity, for example, I could have burned it di-

rectly and toasted four slices of bread. Or,

if I had been around when the solar energy

originally reached the earth, I could have

intercepted it with a giant magnifying glass

and provided enough heat to toast many

loaves.

Of course there is one more possibility. I

could have practiced conservation and

skipped the toast entirely. After all, un-

toasted white bread isn’t that bad. I would

have only been a little bit uncomfortable.

Another fossil-fuel form of solar energy

which we exploit greatly today is oil, par-

ticularly when we convert it to gasoline and

use it as a motor fuel. Perhaps we covet this

liquid energy source because it adapts eas-

ily to our automobile engines, or perhaps

we produce the internal combustion engine

because gasoline is available; in either

event they adapt to each other well. We

therefore burn the liquid fuel, using the vol-

umetric changes caused by the released

chemical energy to push against pistons.

We than convert most of this energy to the

form of engine heat and vehicular motion.

200

Fig, 2

The latter is then further transformed

through brake and wheel friction and air

resistance to atmospheric heating and is ul-

timately again radiated into space. This

time we take advantage of the original solar

energy when it is in kinetic form, using it to

provide us with transportation.

We therefore see that various energy

forms are useful and important to man-

kind, providing us, among other things,

with toast and transportation. As a next

step, let us consider the ways that the world

presently uses energy, so that we know the

current demand before we begin to exam-

ine the supply possibilities.

How much energy does the world use?

The use of energy on earth can be broken

into four basic categories: to provide food,

motion, heat, and industrial processes. It is

not easy to measure the energy which we

use, or to quantify each of these categories.

Some limited quantification is possible, as

is summarized in Figure 2 and discussed

below.

(a) Food. Various life forms require a

number of substances in order to survive

and grow. These substances include carbon

and nitrogen. The carbon generally comes

THE WORLD ENERGY GLUT 131

from CO2, and the nitrogen indirectly from

the atmosphere. In the process of life de-

velopment, these substances are converted

into more complex compounds. In the

highest life forms, such as man, the re-

quired compounds include a variety of pro-

teins and vitamins. Ultimately, the life

forms die and the complex cells decompose

and decay until they return to their original

basic elements. There will have been no

creation or loss of matter, but energy will

have been used in the process. The source

of this energy for living beings is called

food, and the energy content is measured in

Calories.*

The average human being requires about

2000 to 2500 Calories per day to thrive.

Children require much less; large male la-

borers much more. The average U.S. diet

provides close to 3000 Calories per day.

The underdeveloped countries are rife with

Starvation and malnutrition, and half of

the world’s population ingests less than

2000 Calories daily. Let us assume here

that it is an objective to provide an average

of 2000 Calories for all the people on earth.

With a current world population of 43 bil-

lion people (and a conversion factor of 1

Btu = 252 Calories), we determine that the

food energy requirement to take care of the

entire world would be approximately 13

quads per year.

It is interesting to note that this energy

requirement for food is almost never

counted in the energy inventories which are

frequently published. Whenever an analy-

Sis is made of energy supply vs. demand,

the food energy requirement is generally

ignored. We’ll refer more to this shortly.

(6) Transportation. We have progressed

from walking to the use of beasts of

burden, to self-powered vehicles (e.g., ca-

noes and bicycles), and finally to land, sea,

air, and space vehicles which carry their

Own energy sources. Most of the land

* The Calorie (capital C) used for food energy mea-

surement, is equivalent to a kilocalorie, or 1000 cal-

Ories (small c), where the calorie is the common heat

measurement used by physicists to raise | cc of water

through | degree Celsius.

transportation vehicles adapt easily to oil,

which represents stored chemical energy

that is convenient to use and transport. As

a result, we have developed an economy

which is heavily oil based. and have be-

come critically dependent upon it.

In the United States, our chief artificial

mode of personal transportation today is

the automobile. There are approximately

300 million automobiles in the world today,

with about half of them in the United

States. In this country we average 10,000

miles per year, and 15 miles per gallon. In

the rest of the world, the averages are 7,000

miles per year and 15 miles per gallon.

Converting this to quantities of energy for

automobile fuel, we calculate the world

uses 25 quads per year. In other words, we

use twice as much energy to feed the

world’s automobiles as we do to feed the

world’s people.

The energy to fuel airplanes, trains, and

all the other modes of transportation is

equivalent to that used for automobiles, re-

sulting in a total energy use for transporta-

tion of 50 quads.

is used to provide a comfortable environ-

ment and to support industrial processes.

In this section, let us consider environmen-

tal control only. We obtain our warmth

primarily from the atmosphere which has

been heated by the sun, and we supplement

this as necessary by the burning of addi-

tional fuels. The thatched hut in the tropics

is warmed by the surrounding air via solar

energy, but this is not counted in standard

energy calculations. The Eskimo in his

igloo may burn some kerosene to provide

warmth; this is almost always counted. He

may supplement his heating by burning

whale blubber; this is rarely counted. The

native who lives on the fringes of a forest

area may burn some wood to give him

warmth; this is sometimes counted.

If we seek a logical pattern to these di-

verse reporting practices, we discover that

formal energy audits refer to fossil fuels

and nuclear power plants, whether they

provide energy directly or are used first to

create electricity. This is because it is easy

132 COLEMAN RAPHAEL

to measure the output of mines, oil fields,

utilities, and large production plants. The

use of solar radiation, biomass, and animal

sources for heating is much more difficult

to measure and record, and has much less

of an impact on our economic structure;

hence it is often ignored and sometimes

unrecognized.

Certainly these uncounted energy sources

represent hundreds of quads per year.

Since we get most of it “‘free,’’ and don’t

count it anyway, we have not identified a

number with it in Figure 2.

(d) Industrial processes. Finally, there

are all the other things we want to do. We

find it convenient to use electricity. We

want to light our cities at night. We want to

cast aluminum into pots and forge steel

into automobiles. It was shown above that

the world’s population needs 13 quads

worth of food, but we want the food in spe-

cial ways. First we must irrigate, fertilize,

and harvest. Then we process the food, put

it in boxes, cans, and bottles, freeze it, store

it, wrap it in cellophane, prepare it in TV

dinners, transport it, and display it. By the

time we’ve finished, food utilizes an annual

energy expenditure of 50 quads.

The fuels which are used for these var-

ious industrial processes are easily mea-

sured, consisting primarily of oil, coal, gas,

and nuclear energy, with asmall amount of

hydropower for electricity production. The

annual world expenditure for such pur-

poses is approximately 200 quads.

In summary, then, the world currently

uses about 250 quads of energy which is

primarily from fossil and nuclear sources.

As we examine our energy budgets and the

sources which are available, we must look

for ways in which the 250 quads can be

provided with the lease disruption and the

most benefit for society. Let us keep the

figure of 250 quads in mind as we take a

look at the various energy sources which

are available to us.

What energy sources must we consider?

Figure 3 divides all the available energy

sources into three categories: those which

emanate from our mother sun, the largest

particle in our solar system; those which

can be obtained from the atomic nucleus,

one of the smallest particles in that same

solar system; and those which exist in a

stored form, such as the earth’s heat, mo-

tion, and gravitational field. We shall now

consider some of the characteristics of each

of these sources, after which we shall at-

tempt to quantify them.

(a) Wood. The energy of the solar

photon which is captured through photo-

synthesis may be stored in microscopic

plants, grasses, flowers, vegetables, shrubs,

or trees. The hard fibrous substance that

makes up the greater part of the stems and

branches of shrubs and trees is called

wood, and it can continue to store the

chemical energy for many years until this

energy is recovered through oxidation,

which may be slow (decay) or fast (burn-

ing). From man’s “‘taming” of fire 150,000

years ago until the 19th century, wood was

the primary energy source used by man.

Even today, it is the chief source of energy

for some nations, and it represents about

15% of all the energy which is used in the

world.

(5) Biomass. ‘‘Biomass”’ refers to living

matter, and we will use it here to refer to all

the other forms of plant life which store

energy; biomass is sometimes distinguished

from wood, which remains as an energy

source long after its parent plant has died.

Consider the amount of space which is

used to intercept solar energy in the grow-

ing of a tree to maturity. At the end of 25

years, for example, an oak tree might cover

a ground diameter of 30 feet. During this

period, the energy capture efficiency varies,

depending upon the time of day, the cloud

cover, the season, and the amount of fo-

liage. For deciduous trees in particular,

there are annual energy losses. However, at

the end of 25 years, the energy then remain-

ing in the wood represents the net solar

energy intercepted and stored within the

30-foot diameter circular area during the

tree’s lifetime.

If we consider the purpose of the tree to

be solely for the collection of solar energy

THE WORLD ENERGY GLUT 133

ENERGY SOURCES

SOLAR ATOMIC NUCLEUS STORED

° woop e NUCLEAR FISSION (NATURAL) ° GEOTHERMAL

e BIOMASS e NUCLEAR FISSION (BREEDED) ° TIDES

e SOLID WASTE e NUCLEAR FUSION

e FOSSIL FUELS

— OIL

— NATURAL GAS

— COAL

— OIL SHALE

— TAR SANDS

e SOLAR RADIATION

— THERMAL PANELS

SOLAR CONCENTRATORS

PHOTOVOLTAIC

OCEAN THERMAL GRADIENTS

SOLAR PONDS

PHOTOCHEMICAL

e HYDROPOWER

WIND

Fig. 3

Over a 25-year period, we should now con-

sider whether there are better ways of using

this circular area over this long time period

for energy collection. For example, euca-

lyptus trees grow more rapidly than oak;

perhaps four eucalyptus trees could use the

Same circular area and grow to maturity in

8 years. This way we could accumulate the

wood of 12 eucalyptus trees in the same

time that we grow one oak tree. Carrying

this further, might we find other forms of

plant life (known as biomass) which grow

rapidly without using excessively large

areas? Crops such as hyacinth and alfalfa

grow rapidly; kelp, which flourishes under

water, has been known to grow as much as

1 foot per day. Crops such as these, cover-

ing a 30-foot circle, could be planted, har-

vested, dried, and accumulated for 25 years,

at which time the energy would be far

greater than that obtained from a single

Oak tree.

In the past, we have considered agricul-

ture primarily for the production of food.

In the future, we can expect a significant

agricultural thrust to be directed toward

energy crops.

omass is currently used directly for energy

production, including burning and food,

but much of it goes to other purposes such

as lumber, paper, textiles, and other indus-

tries. After the wood, paper, cloth, etc.,

have served there original purposes they

are deemed “‘solid waste’’ and are dis-

carded and/or destroyed. We must recog-

nize, however, that these are the products

of agriculture which still contain much of

the energy which was originally stored pho-

tosynthetically. In their normal course,

these products will eventually decay, re-

lease their energy, and end up primarily in

the form of heat, carbon dioxide and water.

Instead of permitting this to occur natu-

134 COLEMAN RAPHAEL

rally, we can hydrogenate, pyrolyze, or

burn this material to make full use of its

stored energy which would otherwise be

lost to us.

(d) Oil. The next five energy sources

are classified as “‘fossil’’ fuels, or forms of

stored solar energy (in plants, microscopic

animals, etc.) which have been locked away

for millions of years, which can now be re-

covered by extracting them from the ground

and burning them. The best known of these

is oil, a liquid organic material which is

generally trapped in permeable rocks. It

becomes useful to us when it is sufficiently

concentrated so that it becomes easy to ex-

tract. Although oil is only one of the many

energy sources discussed in this paper, it is

convenient for powering our automobiles,

and it is the basic energy source upon which

our world economy is based. Until re-

cently, it has been believed that over half

the free world oil is in the Middle East, with

some in Africa and the rest in the western

hemisphere. New discoveries in the North

Sea, the Gulf of Mexico, and off the Alas-

kan and South American coasts are now

leading to a reassessment of oil resource

potential. When a Middle East political cri-

sis occurs, reducing potential oil supplies,

the public becomes concerned about an

“energy crisis.”” When the oil stockpiles

build up, we speak of an “energy glut.”’ Ac-

tually, oil should not be that important to

the world. It has many problems associated

with it, and it is a short-time resource. It

has to be found, extracted, stored, pro-

cessed, and transported, all requiring mas-

sive resources in the form of pipelines,

tankers, deep-water ports, and refineries.

And yet we began using oil only a hundred

years ago, and will probably stop using it

less than a hundred years from now. Within

man’s possible span of a few million years

on earth, we live in the brief period of less

than 200 years during which oil is available

to us. Yet we are basing a massive economic

structure and sociological existence on the

deliberate burning of this precious and frag-

ile commodity which could be used as a

source of plastics, fertilizer, and general

chemical feedstock.

(e) Natural gas. Natural gas is another

form of fossile fuel in gaseous, rather than

liquid, form. It is a hydrocarbon which

burns easily and cleanly, and it is generally

found with or near oil. It was originally

considered economically unusable and was

flared off during oil production, until its

value was recognized around 1920. Its use

then increased rapidly, and in 1958 it

passed coal and became the second most

used energy source in the United States.

(f) Coal. Coal is the most plentiful of

the fossil fuels in the United States. It is

found in beds which vary in thickness and

in location, some being close to the surface,

while others are down thousands of feet.

Being a derivative of wood and biomass,

coal consists largely of carbon and hydro-

gen, and therefore can yield energy by

burning (combining with oxygen), similar

to oil and gas. The main problem with coal

is that it is dirty and contains sulfur, as well

as many minerals collectively known as

ash. The composition of coal varies widely,

including hardness, moisture content, py-

rites, percentage of ash, and form and

quantity of sulfur. Also, there are hazards

associated with the extraction of coal from

underground mines. If the seam is near the

surface so that it can be strip mined, there

can be significant environmental effects.

As aresult, in the last few decades the use of

coal has decreased as consumers have gone

to oil, natural gas, nuclear power, and die-

sel engines while supporting local and na-

tional clean air acts.

Because of the low cost and abundance

of coal, it is starting to make a comeback. A

main attraction of coal is its ability to be

used as a solid or converted to liquid or gas.

The key emphasis in its increased use will

be on methods of cleaning coal, removing

the impurities without significantly affect-

ing its energy content.

(g) Shale oil. Inaddition to developing

into gas, oil, and coal forms, a great deal of

organic fossil material has become an inte-

gral part of an impermeable rock material

known as oil shale. The organic elements of

this rock is a waxy material known as kero-

gen, and it can be extracted by mining the

THE WORLD ENERGY GLUT 135

shale, crushing it, and heating it to 400°C in

special retorts, thereby producing both a

gas and a dirty viscous liquid containing

much nitrogen. If the liquid is then cleaned,

the resulting oil is a high-energy material

useful for the production of jet fuel as well

as gasoline.

Most of the oil shale in the United States

is found in a single deposit known as the

Green River formation, at the intersections

of Wyoming, Utah, and Colorado. Much

of it is under salt water. The problems of

shale oil recovery are principally logistic,

but they are immense. They involve the

building of major industrial complexes and

the support communities, massive rock ex-

traction operations, bringing in great

quantities of cooling water over long dis-

tances, and construction of transportation

networks. The energy required for doing all

this and for processing the shale oil is

enormous, and the crushed shale will not fit

back into the hole from which it was taken,

so that a byproduct of the operation is the

building of the new mountains (but not in

my backyard).

Despite these severe logistic problems,

there is a lot of energy in oil shale, and it

must be considered as a potential major

source.

(h) Tar sands. In the case of tar sands,

the organic fossils have combined with

sand instead of with impermeable rock as

above. Extraction of the oil involves heat-

ing and separating the ore material. In the

cold weather this material is brittle and

hard to handle, and in the warm weather it

is viscous and hard to handle. Tar sands in

the western hemisphere are almost re-

Stricted to a region in Alberta, Canada,

Known as Athabasca. Some energy com-

Panies are starting to exploit this fuel

source, although the problems of extrac-

tion, refinement, and transportation are

almost as complex as those for shale oil,

and the potential source size is significantly

smaller.

(i) Thermal solar panels. The next six

energy sources discussed here are all forms

of solar radiation, but they are used in dif-

ferent ways. Before discussing any of these,

however, let us consider some general ob-

servations regarding solar radiation.

Solar energy which reaches the earth

may or may not be intercepted and put to

use. If it is not, the energy will heat the

ground or the atmosphere, and eventually

radiate back out to space. Later in this

paper we will consider how this energy

reaches and leaves the earth. For the pres-

ent, let us recognize that this energy will re-

side temporarily on earth whether or not

we choose to use it. One might therefore

consider it as a “‘free’’ fuel.

One characteristic of solar radiation is

that it is diffused. This can be an advantage

because it minimizes the energy transporta-

tion problems; on the other hand, it can

also be disadvantageous because the lack

of concentration results in lower tempera-

tures which therefore limit its application.

Another problem is that the energy is

available only during daylight and as a

function of cloud cover and sun angle.

Therefore, it requires storage. Finally, we

find that most solar energy applications are

capital-intensive, so that the “‘free”’ fuel

must be balanced off against the cost of

amortizing a large initial investment.

‘Thermal panels” refer to the flat collec-

tor plates which can be placed on the

ground or on rooftops and pointed in the

approximate direction of the sun. These

panels are usually made of aluminum or

copper with coverings of glass, and they

contain a fluid such as air or water which

travels through them, being heated by the

sun enroute. The technology is available

today, but at acommonly cited cost of $10

per square foot of installed thermal panels,

it requires an investment of over $15,000 to

save an oil-heating bill of $1,500, which

does not make economic sense at current

interest rates. But the technology is here.

(j) Solar Concentrators. Unlike the

thermal panels above, solar concentrators

are not flat. They consist of curved reflec-

tor plates, or a large number of flat reflec-

tors which are arrayed in such a pattern as

to forma large parabolic curve. In this way,

most of the solar energy which is inter-

cepted by these panels is reflected back and

136 COLEMAN RAPHAEL

concentrated in a small region where very

high temperatures are produced. This

energy can then be used to heat a fluid and

drive conventional generators to produce

electricity.

(k) Photovoltaic Solar Cells. The use-

fulness of solar radiation for direct conver-

sion to electricity has been demonstrated in

the space program, where solar cells have

been used as the primary energy source for

providing power to satellites in orbit. These

photovoltaic devices are made of materials

(such as silicon or selenium) which have the

characteristic of producing electricity (a

voltage gradient) when subjected to light

energy. At present the process of making

the silicon wafers and the covering filters is

very expensive, and a square foot produc-

ing 10 watts of satellite power has generally

cost $500 to $1,000. Today there is consid-

erable research in techniques for bringing

down photovoltaic collector costs by more

than an order of magnitude, at which point

it may begin to be of some greater eco-

nomic interest.

(1) Ocean Thermal Gradients. Thesun’s

energy is absorbed in the upper portions of

bodies of water, creating a significant

temperature difference between the surface

and deep water. Whenever such a tempera-

ture difference exists, it may be used as the

basis for a thermal cycle for extracting

energy. Although the temperature differ-

ence is relatively small, it can be made a

useful energy source by using large areas of

water so that floating islands or giant

tankers are required. The energy can then

be extracted from the ocean and shipped to

nearby land in the form of hydrogen, am-.

monia, or electricity.

(m) Solar Ponds. The concept of solar

ponds has received much attention in Is-

rael. As above, the system uses a thermal

gradient across a depth of water, but in-

stead of ocean depths, the pond may only

be a few feet deep. Minerals and/or salts

are introduced into the pond varying from

very low concentrations at the top to very

high concentrations at the bottom. Asa re-

sult, the solar energy is collected at the bot-

tom producing a reverse temperature gra-

dient in which the bottom of the solar pond

may be hundreds of degrees Fahrenheit

warmer than the top surface. The extreme

temperature gradient over a short distance

can lead to efficient thermal engines for

energy production.

(n) Photochemical. As we have dis-

cussed, the energy of the solar photon can

be captured through photosynthesis in a

living miniature factory in nature. It is

called a green leaf, and its raw materials are

carbon, hydrogen, oxygen, and a catalyst

called chlorophyl. During the photosyn-

thesis process the photon energy is stored

in the form of some chemical changes dur-

ing which carbon dioxide is absorbed and

oxygen is released.

Similarly, it is possible to create man-

made counterparts to the “‘leaf.’’ The con-

cept involves putting together a suitable

combination of certain materials which

possess this unique characteristic of “‘photo-

chemical-conversion”’ capability. Labora-

tory demonstrations have shown that this

method of capturing solar energy is feasi-

ble. At present, however, it has not been

developed beyond the early laboratory

stage, and it is questionable whether it ever

will become economically practical.

(0) Hydropower. Although one does

not generally think of it as such, hydro-

power is another common result of solar

radiation. In this case the solar energy

heats the surface of the planet causing some

of the water to evaporate and ice to subli-

mate and to rise and form mist or clouds.

These clouds are then moved by winds until

they reach an area where condensation and

precipitation occur. The resulting rain or

snow falls on mountains and high ground,

producing streams which flow into rivers,

heading down once more to ocean level. On

the way, they form rapids and waterfalls

and can be used to power water wheels or

turbines. The original solar energy has thus

been converted first to the gravity potential

of raised water and then some of this

energy is further converted to the kinetic

energy of falling water.

To capture and control hydropower, en-

gineers build dams which havea significant

effect on the earth’s ecology. Bodies of

water are changed in shape and size, silt

THE WORLD ENERGY GLUT 137

beds are created or destroyed, and salmon

are prevented from swimming upstream.

However, hydropower does represent about

4% of all the energy used in the world

today, and it can probably be doubled if we

wish to exploit the many waterfalls and riv-

ers in the world which are presently

untouched.

(p) Wind. Justas the effects of the solar

radiation and earth motion combine to

produce hydropower, they similarly are the

prime movers behind wind energy. Air

masses move vertically due to temperature

and pressure changes, and horizontally due

to atmospheric circulation caused, in part,

by earth motion and horizontal pressure

gradients. Asa result, ““winds”’ are created.

Many of these winds blow steadily andina

constant direction, such as across the great

plains from Montana to Texas, across the

Aleutians, and off the coasts of North

Carolina and New England. If, in these re-

gions a grid of large windmills were to be

constructed with 1-mile spacing, this could

provide all the electrical energy currently

used in the United States. The blades of

such windmills might be over 100 feet in di-

ameter, yet relatively light weight, based on

. recent NASA research and demonstrations.

There are those who argue that an array

of giant windmills throughout a large area

may have a feedback effect upon the

weather, and certainly there is an aesthetic

question concerning their appearance. It

can be argued, however, that a windmill

tower is no more visually objectionable

than a tower carrying electrical wires, and

these have been generally accepted. On a

smaller scale, windmills have been used as

an energy source since the 12th century

when the Dutch used them heavily. Today

there are probably 10,000 windmills in the

United States, used by farmers to run 1-kil-

Owatt generators which provide trickle

charges to 12-volt batteries.

(q) Nuclear Fission from Natural Sources.

The previous 16 energy sources are all, in

full or in part, forms of solar energy, mani-

fested in different ways. Let us now turn

our attention to nuclear power, a subject

which has many advocates and many ene-

mies. The viability of nuclear energy is de-

pendent, in part, upon the readiness of pub-

lic acceptance, and at present this is certainly

a questionable matter. We shall consider

here the two primary classes of nuclear

energy, fission and fusion. In fission, a

heavy atom is broken into two or more

lighter ones; in fusion, two light atoms typ-

ically are combined (fused) to form a heav-

ier one. In both cases, energy is released in

the process, so that the penalty for obtain-

ing nuclear energy is to change the form of

some existing matter on earth (E = mc’).

The concept of nuclear fission involves

finding a heavy atom which is fissile, or

easy to split. If this atom is then bom-

barded with a fast-moving neutron, it splits

into two or more smaller atoms, and simul-

taineously releases some energy, plus other

neutrons, which strike other atoms, thereby

continuing the process. In order to assure

that as many neutrons are released as are

absorbed, ina controlled “‘chain reaction,”

the fissile atoms must be packed closely to-

gether in a definable critical mass. During

the process of fission, enormous amounts

of energy are released, and the original fuel

material is ultimately spent, being con-

verted to other materials.

A principal problem with the concept of

nuclear fission is that there is but one spe-

cies found in nature which 1s capable of fis-

sioning under normal mild conditions. This

is an isotope of uranium called U-235. One

device which uses U-235 to produce energy

through nuclear fission is called a “‘light

water reactor,” and the energy is captured

through the heating of water similar to any

normal boiler. In order to obtain the nu-

clear fuel, we must first mine uranium

oxide (U308) and extract the uranium. Of

this, seven-tenths of one percent is U-235

and the rest is U-238, which is not fissile.

We therefore see that it takes a significant

amount of mining to produce a small

amount of fuel, and the total world fuel

source for this means of energy supply is

limited.

(r) Nuclear Fission from “‘Bred’’ Sources.

As we have just stated, U-235 is the only na-

tural-occurring material which is fissile.

However, there is a material known as

plutonium 239 (P-239) which does not exist

138 COLEMAN RAPHAEL

naturally but is also fissile, and this mate-

rial can be artificially produced when U-

238 is bombarded with the high-energy

neutrons in a reactor. Once the U-238 has

been converted to P-239 it then can fission,

but this time the reactor must be modified

to accomodate the conversion of U-238 to

P-239, which then becomes the primary

reactor fuel. A reactor which does this is

knownasa “breeder reactor.’’ The breeder

reactor breeds fuel at a faster rate than it

uses it for energy production, but of course

it is dependent upon the availability of the

initial supply of U-238. Since this latter iso-

tope represents 99.3% of normally mined

uranium, it does represent a plentiful fuel

source. In addition, there is another sub-

stance which can be used for practical ap-

plication in breeder reactors. Thorium-

232, a natural-occurring nonfissile material

can be converted in a breeder to U-233, an

artificial but fissile substance.

A principal concern with nuclear fission

is that it produces radioactive waste. When

the uranium (or plutonium) atom is split, it

may separate into different types of ele-

ments, since it does not always fission in the

same manner. Therefore there will be a

wide variety of materials in the waste pro-

ducts, some of which are radioactive. These

materials can include cesium 137, carbon-

14, radium, strontium-90, and others, as

well as traces of the original fuel, which is

plutonium or uranium in the breeder. The

intensity and duration of the radioactivity

varies widely, and can neither be controlled

nor eliminated. Therefore one must accept

that the use of nuclear fission implies the

continuous production of a long-lived haz-

ardous radioactive waste. The question of

disposal of such waste is a public issue

which has not yet been satisfactorily re-

solved. In addition, since P-239 is the pri-

mary element for the production of nuclear

weapons, questions regarding terrorism,

theft, nuclear blackmail, and the morality

of nuclear war are all inextricably linked

with the question of feasibility of nuclear

fission. The legitimacy of such questions

presently remains an arguable issue.

(s) Nuclear Fusion. Nuclear fusion re-

fers to the process of taking two atoms of a

light material such as hydrogen (either the

hydrogen isotope deuterium or tritium)

and fusing them together. During the proc-

ess, energy and high-velocity nuclear parti-

cles are again released, but this time the end

product is a simple benign substance, such

as helium or water. The concept of nuclear

fusion is extremely attractive not only be-

cause the radioactive waste problem is es-

sentially eliminated, but because the fuel

source, hydrogen, is virtually unlimited.

Unfortunately, the engineering problems

to achieve practical fusion are somewhat

overwhelming. The process _ requires

temperatures of millions of degrees at pres-

sures and densities which are extremely dif-

ficult to achieve, and it is expected that

there will be a level of radioactivity asso-

ciated with the containment boundaries.

There is a significant technical question of

whether or not the world will ever achieve a

truly practical fusion reactor.

(t) Geothermal. The final two energy

sources are associated with the physical

characteristics of the earth. First, there is a

great deal of stored thermal energy within

the earth’s core. We know that temperature

increases aS we penetrate the earth’s sur-

face, and that at depths of 100 miles it is in

excess of 1000°F. The most popular current

theory is that the temperature throughout

the molten core is relatively constant, pos-

sibly 3000 to 5000 degrees F, and that it is

maintained by the radioactive material and

activity in the earth’s outer crust. It would

obviously be desirable to tap this energy

source by drilling holes through the crust,

but mankind does not currently know how

to safely do this. However, there are many

places where the crust has cracked, or

where fissures and overlaps occur in the

tectonic plates, and the mantle and core

material have risen close to or through the

surface. Such phenomena are associated

with volcanoes, geysers, faults, or hot

springs, and we have been able to use the

heat in such areas to provide local energy

needs.

The specific way to use geothermal energy

is dependent upon the form in which it ap-

pears. Geothermal sources may be wet,

dry, or briny, and each of these requires dif-

a ee ee ee ee ee ee

THE WORLD ENERGY GLUT 139

ferent heat exchangers and different ways

of dealing with the many problems, such as

corrosion.

(u) Tidal Energy. Under the combined

action of the earth’s rotation and the lunar

gravitation force, the earth’s oceans are

subjected to tides in which the water levels

rise and fall predictably. If it were possible

to construct giant dams in high-tide loca-

tions, the water could be captured when the

level is high, and permitted to flow and

drive turbine wheels at low tide. The prac-

tical problem of damming the oceans and

the relatively small amount of potential

energy make this a somewhat unattractive

prospective energy source.

The Earth’s Energy Balance

We have just considered 21 different

well-known energy sources. Before we begin

to quantify them, certain general character-

istics of individual energy sources should

be noted, i.e., their interchangeability and

their associated efficiencies.

(a) Interchangeability. Although energy

can be changed from one form to another,

the process can be expensive and time-con-

suming, and it is obviously desirable to se-

lect an energy source which adapts to the

particular requirement with minimum con-

version . Therefore, if the application is for

plant growth or photovoltaic conversion,

we look for photons and light energy. If we

wish to produce electricity from turbogen-

erators, we need energy sources which can

produce high temperatures easily. If we’re

interested in transportation via cars, air-

planes or boats, it is convenient to use

energy stored as liquid petroleum products.

Fortunately, just as there are many energy

sources, there are also many energy needs.

By wisely tailoring the sources to the needs,

we can make the best use of the energy re-

sources available to us.

(b) Efficiencies. When we normally

speak of energy conversion efficiency, we

generally refer to a single process, assum-

ing that all of the energy which is not co-

verted in that process is considered wasted.

However, we should recognize that energy

“wasted” in a single process is not lost, but

may be used for a different purpose or ina

different process. For example, photosyn-

thesis is considered a low-efficiency proc-

ess, converting less than 1% of the incident

solar energy. This really means that 1%

(say) of the solar energy has been stored

chemically for use in foods, vegetables, and

plants.

The remaining 99% is not lost. Some of it

heats the atmosphere and produces am-

bient warmth. Some of the warmth pro-

duces moisture evaporation which leads to

hydropower. Some of it heats our bodies.

Some of the energy which went into our

bodies as food is later usable again as a

waste material for energy production.

Ultimately, all of the energy will radiate

away. In the meantime, it is used and

reused for many different purposes, in many

different ways, and a true measure of over-

all ‘‘efficiency”’ is hard to define.

In the following section, we shall con-

sider all the energy which reaches the earth

from the sun, and we shall see what happens

to it. This is known as an energy balance.

is 93 million miles away and radiates the

equivalent of 12 quadrillion quads per

year. This energy radiates from the sun into

space in all directions, and about one 2-bil-

lionth of it is intercepted by a spinning

8000-mile-diameter sphere called the earth.

The earth has a solid crust, an inner mantle,

a hot liquid core, and outer oceans and

gases. These gases vary from a dense at-

mosphere to a rarified ionosphere. The

temperature around the earth varies signif-

icantly, but if we were to average it over the

entire surface at sea level, we would obtain

an average temperature of 59°F. This aver-

age temperature has been the same for

many, many years, and the earth is essen-

tially in thermal equilibrium.

Near the equator, between N 38° and S

38° latitude, the earth is constantly absorb-

ing heat. Near the poles where the tempera-

ture is much lower, there is a loss of heat.

The net balances out to be approximately

zero.

The energy coming from the sun peaks at

a wavelength of .48 microns and consists

primarily of infrared, visible, and ultravi-

140

47% ABSORBED BY EARTH

~—-

— “gi nBSORE

COLEMAN RAPHAEL

SOLAR ENERGY COMING IN

€ee ENERGY FROM SUN = 100%

5.4(10)° QUADS/YR

Fig. 4

olet light. Figure 4 shows how this energy is

distributed. The total amount of solar

energy intercepted by the earth and its at-

mosphere is 5.4 million quads per year.

Some of this energy is reflected from the

atmosphere directly back to space, some of

it is absorbed in the clouds, and some of it

in the clear atmosphere. Nine percent of the

total energy reaches the earth’s surface, but

is of such wavelength that it is reflected

away as though it had struck a mirror. The

remaining 47% of the 5.4 million quads per

year reaches the earth and is absorbed into

the earth’s surface.

Figure 5 shows the energy which leaves

the earth. Due to the earth’s surface temper-

ature, the net radiation away from the sur-

face is approximately 6.2 million quads per

year, or 115% of the intercepted incoming

radiation (of 5.4 million quads/year). In

addition, 10% leaves the earth by convec-

tion (carried away by rising heated air) and

19% by moisture evaporation from ocean

surfaces.

We see then that the total heat leaving

the earth’s surface is much greater than the

incident solar energy. However, much of

this heat is trapped and reflected back from

the atmosphere and clouds, so that the net

efflux of energy from the earth’s surface is

equal to the 47% influx. Therefore the

earth is in energy balance and its average

temperature remains approximately con-

stant.

This conclusion contradicts early the-

ories that the earth was cooling off like a

baked apple, hottest at the center with heat

flowing out at the surface. If one analyzes

this model by taking the thermal gradient

at the crust and extending it into the center,

the temperature at the center is beyond Jjus-

tification. The more generally accepted

thesis today is that the temperature through-

out the earth’s core is constant and is in the

THE WORLD ENERGY GLUT 141

ENERGY LEAVING EARTH

TO SPACE

Fig. 5

order of 3000° to 5000°C. The heat in the

interior is produced by radioactivity con-

centrated in the earth’s mantle, witha small

amount escaping outward, but this amount

is negligible when compared to the incident

solar energy. This model of the earth is then

similar to an oven wherein the heating ele-

ments are distributed around the walls and

the temperature is constant inside.

How Much Energy is Available?

We have shown previously that the an-

nual world demand for measurable energy

is in the order of 250 to 300 quads. Keeping

this in mind, Figure 6 has been prepared

showing the potential and practical availa-

bility of the energy sources previously des-

cribed. In the case of solar-dependent non-

fossil sources such as wind, hydropower, or

thermal panels, they have all been lumped

into one category in Figure 6, titled “‘all

solar forms.’ However, the solar energy

collected in past years and available today

as fossil fuels are listed separately.

The center column of Figure 6 gives a

rough approximation of the annual energy

available from each of these sources. These

figures are independent of any considera-

tion of economic benefit or technological

ability to harness this energy today. The

purpose of this column is to identify the

total resource potential, possibly to be used

in future years as economic values change

and science continues to advance.

The right-hand column of Figure 6 takes

into account the practicalities of current

engineering know-how and public social

attitudes. If tapping a particular energy re-

source is beyond our present ability or so-

cially prohibited, such resources are not

shown. The costs have not been considered

142 COLEMAN RAPHAEL

SOURCE CURRENTLY

ENERGY POTENTIAL AVAILABLE

SOURCE QUADS QUADS

W/0 ECONOMIC CONSIDERATION

ALL SOLAR FORMS __ 5,000,000/YR 2,000/YR

(CURRENT)

OIL 70,000 13,000

NATURAL GAS 25,000 4,000

COAL 240,000 24,000

SHALE OIL 20,000 10,000

TAR SANDS 10,000 2,000

NUCLEAR FISSION 100,000,000 400,000

NUCLEAR FUSION —> 00 fh

GEOTHERMAL 100,000,000 2,000

TIDES AIYEAR es

Fig. 6

in preparing the right-hand column; in

other words, this energy is available and

acceptable to the world right now if we are

willing to pay for it.

Let us comment on each of these energy

sources briefly:

(a) All Solar Forms. Except for the fos-

sil fuels, all other forms of solar energy are

included here, regardless of the methods by

which we intercept and use them. We may

then look at the total solar radiation which

reaches the surface of the earth, and, as we

have shown in Figure 4, this is roughly 47%

of 5.4 million quads/year.

Obviously we cannot capture this solar

energy everywhere. But certainly a large

portion of it can be intercepted. Thirty per-

cent of all land surface is currently forest,

and significantly more could be set aside

for more forest and biomass production.

Rooftops of most buildings could be used

for thermal panels, and large desert areas

could represent thermal-electric collection

areas. Rural areas near cities could become

dedicated solar farms, and large areas of

ocean could be tapped for thermal gradient

utilization. With a world program dedi-

cated to the use of solar energy, it is cer-

tainly reasonable to assume that 2% of the

projected surface area can be used for the

capture of solar radiation, particularly if a

great portion of it is in the form of biomass,

thereby minimizing capital investment.

By “‘efficiency’’ we refer to the percen-

tage of the original solar energy which is

diverted to a form which satisfied man-

kind’s needs. Individual efficiencies vary

greatly. The amount of incident energy

captured in plants through photosynthesis

is very low (less than 1%), while thermal

capture efficiency is significantly higher.

Atmospheric heating represents a very high

efficiency. In some cases, the same energy

can be reused to serve several purposes,

such as atmospheric heating, ocean ther-

mal energy conversion, and hydropower,

or biomass and solid waste. Let’s apply

here an average efficiency value of 4% to all

the solar radiation which we can capture.

Combining these factors of 47%, 2%, and

THE WORLD ENERGY GLUT 143

4% with the total incident radiation of

5.4(10)° quads/year, we determine that

there could be available for our imminent

use a total of approximately 2000

quads/year.

One can distribute the exploitation of

this solar energy over the many different

techniques which have been discussed here.

For example, there are currently 4 trillion

cubic feet of wood available in the world’s

forests, covering 30% of land surface area,

with a total stored energy potential of 2000

quads. The average tree is approximately

25 years old, so that the growth rate of new

wood is 4%, or 50 quads/year. If we resolve

to stop all deforestation but make use of

new growth, and combine this with an in-

tensive worldwide planting and biomass

conversion program, hundreds of quads

per year of energy could be made available

from this solar source alone.

(5) Oil. The most optimistic estimate

of oil availability would take into account

all known, estimated, and yet undiscovered

resources in the free and communist worlds,

and would assume that modern plus more

advanced extraction mechanisms will be-

come available. On this basis, the total re-

source could be as high as 11 trillion bar-

rels, corresponding to 70,000 quads. If we

then consider the free world only, defined

reserves, and existing extraction techniques,

the immediate resource corresponds closer

to 13,000 quads.

in the United States are in the order of sev-

eral hundred TCF (trillion cubic feet). If we

assume that these reserves represent one-

tenth of the total U.S. resource, and that

the world resource is 10 times that of the

U:S., the resulting total potential is 25,000

quads. If we sought to extract this gas im-

mediately with present knowledge and tech-

nology, the available number of quads is

assumed to be 4,000.

(d) Coal. The U.S. coal resource has

been estimated to be about 3 trillion tons,

and it’s assumed that this is about one-

fourth of all the world’s coal. On that basis,

the coal source potential is 240,000 quads,

Significantly more than any of the other

fossil fuels. If we consider only that coal

which is immediately recoverable in forma-

tions similar to those being mined at pres-

ent, and we use present techniques and

equipment, about one-tenth of the resource,

or 24,000 quads, is immediately recover-

able. This could probably be increased rap-

idly by undertaking a concerted coal recov-

ery program.

(e) Shale Oil. The U.S. Goedetic Sur-

vey estimates that at the present time, 600

billion barrels of shale oil are recoverable

in the United States, and that with a further

increase in technology, this can ultimately

be doubled. If we then assume that the

world resource is double that of the U.S. we

obtain a potential of 20,000 quads, of

which possibly one-half could be recovered

right away if we choose to do so.

(f) Tar Sands. Tar sands are primarily

a Canadian resource in the western hemi-

sphere. Extracting and processing this energy

source for use in the United States ob-

viously will require political negotiations

and agreements. The numbers shown in

Figure 6 are rough estimates based upon

educated guesses.

(g) Nuclear Fission. To determine the

resource potential for nuclear fission, let us

consider that all the uranium and thorium

that we are able to mine can be used for

energy production, whether it is naturally

fissile or has to be conditioned through

breeding. On this basis, using these ele-

ments in the earth’s accessible crust regard-

less of the compounds or forms of ore in

which they appear, the very large resource

potential of 100 million quads is obtained.

For the right-hand column of Figure 6,

we must be more practical. In the case of

nuclear fission, more than the others, the

additional restraint of sociological accep-

tability becomes significant. However, as-

suming that society is willing to accept

breeder reactors, and limiting ourselves

only to the U-235 and U-238 which could

be extracted from known quantities of ura-

nium oxide (U3QOx), the practical availabil-

ity of this source is 400,000 quads.

(h) Nuclear Fusion. The prospects of

nuclear fusion are exciting. The raw mate-

144 COLEMAN RAPHAEL

rial, deuterium in the oceans, is essentially

limitless. Although the end product is non-

radioactive, the process does involve the

production of high energy neutrons, and

there will be temptations and pressures to

use them productively (leading to pluto-

nium, radioactive waste, etc.) rather than

merely absorb their energy. For the mo-

ment, however, this question need not be

faced, since the high temperatures, short

time periods, and material requirements

are such as to discount nuclear fusion as an

energy source which will be technologically

available within the next decade.

(i) Geothermal. The potential capacity

of geothermal energy ranks in magnitude

with solar energy and nuclear fusion. For

example, if we could utilize all the thermal

energy stored at temperatures over 150°F.

in athin shell of 30,000 feet of earth’s crust,

we would have a potential of over 100 mil-

lion quads. However, if we limit ourselves

to those geological areas where the magma

has penetrated or is close to the surface, the

immediate energy availability is in the

order of 2000 quads. For example, geo-

thermal energy is currently the primary

energy source in Reykjavik, Iceland, Lar-

derello in central Italy, and the Geysers

north of San Francisco.

(j) Tides. The potential of the ocean

tides as an energy source is very low, both

now and in the future.

Where Do We Go From Here?

Let us consider the points which have

been made so far. In addition to the “‘free”’

energy which warms our air and feeds our

bodies, mankind uses approximately 250

quads a year to suit his remaining needs

and desires. In Figure 6 we see that there is

far more energy available than man on

earth will ever be able to use.* In the 21st

* The values in Figure 6 do not represent hard

numbers which can be indisputably justified. They

are, rather, based upon an assortment of scientific and

nonscientific sources, expert opinions, recollection,

general observation, logical deduction, and ‘“‘gut

feel.’’ If the reader disagrees strongly and is so in-

clined, please multiply or divide any of the numbers

by two or three. Doing so does not significantly affect

the conclusions.

century we can expect that our primary

sources of energy will be solar, fusion, and

geothermal, all of which are safe and virtu-

ally limitless. At the present time however,

we are limited by technological constraints

to the right-hand column of Figure 6. Ata

rate of 250 quads per year, we have about

50 years of oil or 100 years of coal imme-

diately available to us. By the time we de-

velop advanced coal extraction techniques,

this resource could last closer to 1000 years.

And, if we choose to undertake a massive

assault on the use of solar energy in its

many forms, we could use this as our sole

energy source to fill our needs many times

over.

The basic purpose of this paper is to

present the argument that society’s selec-

tion of its prime energy sources is illogical

when considered in the light of source

availability or present and future needs. It

is time to reexamine our dependence on oil,

and consider whether it should be saved as

a source for plastics, new materials, fertil-

izer, and other valuable chemical products

instead of being indiscriminately burned.

Similarly, we must recognize that the other

fossil fuels are limited. Should we be using

combinations of these energy sources in

order to extend their availability? Should

we Switch to nuclear fission, or will this al-

ways be vetoed by society because of fears

of terrorists, sabotage, and/or nuclear waste

disposal? Should we consider bypassing the

questions of hazards and resource deple-

tion and go immediately to the use of re-

newable and clean solar energy to satisfy all

our needs? The general public (and many of

its elected representatives) are not aware of

the fact that solar energy ona massive scale

is technologically available today. There

are many forces and constraints against its

universal adoption; these are primarily

economic, political, and sociological, and

fall into three principal classes of argu-

ments. These arguments and the counter-

arguments are given below:

(1) Argument: Solar energy is too ex-

pensive at present.

Counterargument: This is true, so let’s

change the economics. We make “‘eco-

THE WORLD ENERGY GLUT 145

nomic” decisions every day based upon

price, but price is not a valid indication of

sociological value. Prices are determined

by supply, demand, and regulation. Regu-

lation may be blatant, or it may be subtle,

such as through the use of penalty and in-

centive taxes. If burning high-sulfur coal

produces air pollution, why not add the

clean-up costs to the cost of the coal? If the

burning of oil eliminates its availability as a

chemical feedstock, why not add a sur-

charge to oil representing the cost of devel-

oping replacement feedstocks? If the use of

solar energy can lead to independence from

Middle East pressures, why not provide

significant tax incentives to motivate the

U.S. public to demand this energy source?

Thermal solar panels could become eco-

nomically attractive to many new home

owners if the government were to providea

subsidy of at least a billion dollars a year.

Such a federal expenditure in a single year

would recover itself through savings in oil

import costs in less than 5 years.

We should make the decision to use solar

energy on the basis of health, security, and

quality of life, and then strive to adjust the

economic tools to make it financially at-

tractive. Considerations must be given to

international trade impact and to our con-

tinuing relationships with present fuel-

supplying nations. There must also be a na-

tional commitment and a_ national

understanding if the required legislation is

to be developed and enacted. This brings us

to the second of the three arguments.

(2) Argument: The application and the

State of technology of solar energy is not

understood.

Counterargument: Educate the world’s

peoples. Massive amounts of energy are

wasted because of ignorance and misuse.

Third world citizenry collect meager

amounts of wood or dung to burn, and

then lose over 90% of the energy because of

bad design of their earthen ovens. Fast-

growing plants, biomass conversion, and

conservation techniques must be developed

and disseminated. People in general must

be made aware of all the forms of solar

energy which they can use, and the fact that

much of this energy is technologically

available to them today. Such an education

program is possible, but it requires the ded-

ication and concerted efforts of our na-

tion’s leaders.

(3) Argument: The disruptive effects of

a sudden switch from oil would destroy the

miltinational oil companies and produce

worldwide economic catastrophe. There-

fore the big oil companies are not going to

let it happen.

Counterargument: There is no need for

the creation of newcorporate giants. Let us

give the solar energy responsibility to the

oil companies. The industrial structures are

in place for developing, servicing, and dis-

tributing energy to the world. We need the

know-how, management ability, and cooper-

ation of such organizations if we undertake

a universal program to exploit solar energy.

It makes no sense to destroy existing cor-

porations and create new ones which must

be just as large and just as powerful. With

proper regulation and profit motivation,

the oil-distributing multinationals can make

the switch to solar energy with the leader-

ship talent we need and with the least eco-

nomic disruption.

By now the reader may have inferred

(correctly) that the author is prejudiced

toward an accelerated utilization of solar

energy. But that is not the primary thrust of

this paper. The intent here has been to

show that the energy “‘crisis”’ is not one of

unavailability, but rather one of economic

imbalance. As far as energy sources are

concerned, we have, and will continue to

have, an energy glut. In making decisions

regarding the distribution and use of these

sources, it is time for mankind to make

judgments and take actions which are

based upon concern for the future, com-

passion for people, and hope for the com-

mon good of society rather than upon a

comparison of prices which are artificially

constrained by man’s inherent selfishness

and shortsightedness.

i !

t |

HH by |

i : “ ;

, { gil , ets) Py Ea ily a i iw

im] ¥ ! 1a 4 ait (A ’ ‘. 8 i, ¥

Ve ct gh UR WL GLU AIOD Sse dala Pan

ee oan 4 il Oo te J et ea he, ry ny jive ( | : iif

Nahar aii Ra wi Ne ee ee j \: oud ty fare f et } 7)

¥ " yt Mae tad a) pase oP WR Ree ay hh

ae Teena a ry

as 4 ‘ Ai , : eh ©) viby W * i

‘ ; 4 . Brod , | 4 0 AG nal 7 F : \

a |

v 9 ) , ¥ 4

a on M > ‘.) it ;, a yi sch uA - r

: rh mk a :

ye Cara tou ‘ p

a | sabe | Lv ight Sd a \+ ,

4 yp . ae Pak ti e aa ré. i ‘ x ,

& ) PL APY r, 7 ; i .

a i A ; ie

a © <i halal Poel co i ek as ' Pity a Bilas: had

a ¢ / hb i *, Ce er a: men ih ivi os we: aro) a heen Ve Re et * i

1% east & x Kathy, 4, iy

" P / .

, i .

, A

ray if

eters AFR wat y: oti "

DELEGATES TO THE WASHINGTON ACADEMY OF SCIENCES,

REPRESENTING THE LOCAL AFFILIATED SOCIETIES

IEEE TUGSIIIILOM cc cece ests vets a scvcunesvcsnescccsascecvecesees James F. Goff

nC EreGTOLY OF WASHINGTON... occ eves ce ceecenccceccucveccteccuseess Ruth H. Landman

EE MUMERIIIBEON soko ce eta te ccc eecnerecevnesserseenoues William Ronald Heyer

i acc ss aw Wis'e aermie kv x Gob dee awn nodes dcwdes obese vals Anne Brown

IE, WASIMINIDTON «5 15 cc cece cee cece eee nace veveusccscccuceveuas Margaret Collins

a, aha arsiain'e a bia’ Ww adhd OS UNG a edkb ace end Soc we eie eu be T. Dale Stewart

EI EEINIOLOY he saw eee c nese av envevesadeevienessueuesse James V. O’Connor

nn te pistrict Of Columbia..........02.cccccssccccsvcseceeseue Charles E. Townsend

re aie cca aig oe aise Saw weal win eos anm WG Koo, u le WR p Vio wre ata ice alee Paul H. Oehser

ERED Fo no fal Ss) Od og: tak do b.0/ Wk Hee OEM On PE mane nents Puente Conrad B. Link

EE oat ie ck Gad ¢ sla ele awe wins vow Sdv wiaalwinte Kiele'e wg ede nie'ws Gow eine Boyd W. Post

ES IESIFICETS cic ce cc ccc eee nee eeedesteeossnentusneneees George Abraham

emecettea and Electronics Engineers ..........0.ccceccceccccvaccveceree George Abraham

I RMCCHATICAL ENPINGETS. .. 0... c ccc cect cc catcaseccweccecesbecessvseve Michael Chi

CECH NV ASIINGTON . .. 0.0 aces cece cn ccccaccncwuede deawenses Robert S. Isenstein

EEE IVIICTOUIOIOLY «20.5 ccc ccc ete ees en ees eebncrcuceenecenenen Lloyd G. Herman

i MRIICTLY ESNGINCETS. . 0. ws ee vee c be cheese eneeeseccebeenens H. P. Demuth

MEE OULEMOUICCES .. . 6. ee a ee ee ee ce eee ene neewewseeenses Wallace J. Cohen

eee berinchial Biology and Medicine.............cceeecrcceveneccenaces Cyrus R. Creveling

ENS ol id ns Kin n’s os bine oo wv eat eens cee Oe pepe ble we meee Charles G. Interrante

menmnatienn OF Denial Research: ...........c cree ences nec secceasvensnces William R. Cotton

American Institute of Aeronautics and AstronauticS............cccceeeecccceeeeees Richard P. Hallion

ER EIEMESOCICEY occ. cc ok ecw tree edocs eetecewedeesteecs A. James Wagner

SG Jack R. Plimmer

MERTEN OA 00 oo gsi a ol 6 x'o nis cic wie an este ie when ebene vere bausweae cus Richard K. Cook

EN re Sk ss an eed te Reels eee dle ee eK deN eS Ra ee wale Sele Dick Duffey

ES SS A. D. Berneking

EEE, cack ees oie ks a aes cde wale cbs one cetwer ewes uaiepeis Edwin R. Fuller, Jr.

ee eon. ne can ave eiald one vide ee aie Ri eew edewes ce Alayne A. Adams

Se EERICHCS CLUB cc ccc cee ect esc a be ewe rene nn een cesewnnws Deborah J. Warner

I EIMEOETIVSIES FEACHEIS <5... 2.202 sec ce cece sedan seeeeres ced eneres Peggy A. Dixon

ES N62 25 81S 0 5 iin don Sila vie oak Persia's ONO eRe ew ee Ree George J. Simonis

OE Litit PHYSIONIOPIStS «2.22.22 oe cece ee eee eee es eneunenens Walter Shropshire, Jr.

eaiions Research Council...........2..cccceccecscrecscvccnccssesnss John G. Honig

CSIC has iia a es no oa vice 6 edie nse neeesneccenwuswemawane ned Jewel B. Barlow

American Institute of Mining, Metallurgical

I ESPUICCES soc keene cee ence cen cine ngecenaseencevesvoeunesen Garrett R. Hyde

EEE Soo, sc Pd dike os elena cn kicce oD sie dike ean soe neces ea ning ees Benson J. Simon

CECI Of AMICTICA ..... 2.000 0cce ccc cccccccccunccesscccsucevonsnene Patrick Hayes

fe oot wie ide wpa Ok ew ae eh wale en Se ees Miloslav Racheigl, Jr.

TEE AISSOCHALION «<i scin cnc c cess caceesescncacccccuscesebecessecense H. N. Reynolds

Seeeeemineton Paint Technical Group ..........----0--eccecceccscceececssseeses Paul G. Campbell

See ytopathological Society ..........--00scceccccncccceeceencescees Howard E. Waterworth

Seer Gzeneral Systems Research ............2ccscccccccccccccscescces Ronald W. Manderscheid

ee, (dei ccd cc cidien kr decdarddesenkeneaseunemesiee’ Stanley Deutsch

PEE TMIIEEY gle k se cece ence seeccecccetsccccsessececvasevessesns Irwin M. Alperin

Association for Science, Technology and Innovation ..........eeecceeceee reece eeeeeees Ralph I. Cole

Delegates continue in office until new selections are made by the representative societies.

ieee “a ey { ry

: 049, ee OS as f *~ |

Wee ae by eT

ar y *

a f i

aed Bs oo * i

“1

Washington Academy of Sciences

1101 N. Highland St.

Arlington, Va. 22201

Return Requested with Form 3579

and additional 1

VOLUME 73

Number 1

) : nal of the March, 1983

wASHINGTON

ACADEMY .. SCIENCES

ISSN 0043-0439

Issued Quarterly

at Washington, D.C.

CONTENTS

Articles

JOSEPH V. BRADY and HENRY H. EMURIAN: Experimental Studies of

Mima Groups in Programmed Environment ...... .....0-2eenceecanns

D. H. UBELAKER: Human Skeletal Remains From OGSE-MA-172: An Early

Suaneala Cemetery Site on the Coast.of Ecuador ............s0cces0

R. W. SWEZEY. SIEGFRIED STREUFERT. and JOHN MIETUS:

Development of an Empirically Derived Taxonomy of Organizational

ee Es oe niga ake 'e wis 2 mya eRe s Fee nie She Re al ee ae we

OE RRR Sn 8 ee onl hg ieee aie Be el a en

Washington Academy of Sciences

Founded in 1898

EXECUTIVE COMMITTEE

President

James F. Goff

President-Elect

Jean K. Boek

Secretary

John J. O’Hare

Treasurer

Ronald Manderscheid

Managers-at-Large

Jo-Anne Jackson

Grover C. Sherlin

Alfred Weissler

Elise A. Brown

James Wagner

Ralph I. Cole

BOARD OF MANAGERS

All delegates of affiliated

Societies (see Cover 3)

EDITORS

Irving Gray

Joseph Neale

ACADEMY OFFICE

1101 N. Highland St.

Arlington, Va. 22201

Telephone: (703) 276-9254

The Journal

This journal, the official organ of the Washington

Academy of Sciences, publishes historical articles, critical

reviews, and scholarly scientific articles; proceedings of

meetings of the Academy and its Board of Managers; and

other items of interest to Academy members. The Journal

appears four times a year (March, June, September, and

December)—the September issue contains a directory of

the Academy membership.

Subscription Rates

Members, fellows, and patrons in good standing receive ~ .

the Journal without charge. Subscriptions are available on

a calendar year basis only, payable in advance. Payment

must be made in U.S. currency at the following rates:

U.S. and Canada...... $19.00

POreiga . s228ss sees 22.00

Single Copy Price . >... 7.50

UHURU Symposium—2 issues............

Delivery by Aur ic. 2.5 2.00

Single-copy price for Vol. 66, No. 1 (March, 1976) is $7.50.

Back Issues

Obtainable from the Academy office (address at bottom

of opposite column): Proceedings: Vols. 1-13 (1898-1910)

Index: To Vols. 1-13 of the Proceedings and Vols. 1-40 of

the Journal Journal: Back issues, volumes, and sets (Vols.

1-62, 1911-1972) and all current issues.

Claims for Missing Numbers

Claims will not be allowed if received more than 60 days

after date of mailing plus time normally required for —

postal delivery and claim. No claims will be allowed

because of failure to notify the Academy of a change in

address.

Change of Address

Address changes should be sent promptly to the Academy

office. Such notification should show both old and new

addresses and zip number.

Published quarterly in March, June, September, and December of each year by the Washington

Academy of Sciences, 1101 N. Highland St., Arlington, Va. 22201. Second class postage paid at

Arlington, Va. and additional mailing offices.

Journal of the Washington Academy of Sciences.

Volume 73, Number |. Pages 1-15, March 1983

Experimental Studies of Small Groups

in Programmed Environments*

Joseph V. Brady and Henry H. Emurian

The Johns Hopkins University, School of Medicine

ABSTRACT

This paper describes a programmed environment research laboratory that was designed

and constructed to support functional analyses of individual and group performances viewed

conceptually within the context of a small-scale microsociety. Summarized are research re-

sults from experimental analyses of (1) conditions that sustain group cohesion and productiv-

ity and that prevent social fragmentation and individual performance deterioration, (2) moti-

vational effects produced by the programmed consequences of individual and group perform-

ance requirements, and (3) behavioral and biological effects resulting from changes in group

size and membership. The significance of these investigative undertakings is to be understood

in terms of emergent environmental, motivational, and social-interaction principles having

practical relevance for the establishment and maintenance of high levels of human perform-

ance within isolated and confined microsociety environments.

Introduction

The voluminous research literature on

the study of small groups that has devel-

oped over the past two decades’” clearly re-

flects the limitations imposed by investiga-

tive restraints in operational settings that

lack experimental control on the one hand,

and laboratory situations that seldom pro-

vide for extended observation under realis-

tic conditions, on the other. Investigative

initiatives on this important aspect of human

social behavior would be advantaged by

the development and application of an ef-

fective methodology for extended-duration

analyses of both the functional and topo-

graphic aspects of such group performance

Situations under conditions that provide

for operational assessments and evaluation

within the context of a self-contained,

comprehensive living and work setting.”

The present report describes an experimen-

*Supported by NASA(NAG 2-139), ONR(N-

00014-80-C-0467) AND NIDA(DA-00018).

tal methodology and representative results

derived in the course of developing a labor-

atory environment designed for the obser-

vation and measurement of human behav-

ior in small groups over extended periods

of time under conditions of continuous res-

idence by volunteer human participants. A

discursive rationale and preliminary model

have been developed for such continuously

programmed environments in human re-

search on the basis of extended experimen-

tal control, objective recording, and the

maintenance of realistic and naturalistic

incentive conditions for the assessment of a

broad range of individual and group be-

havioral processes.* ° The results of an ex-

tensive series of studies in this research set-

ting have been described in several previously

published reports.’

Methods and Procedures

The residential laboratory consists of

five rooms and an interconnecting corridor

2 JOSEPH V. BRADY AND HENRY H. EMURIAN

constructed within a wing of the Phipps

building at the Johns Hopkins University

School of Medicine. The overall floor plan

of the laboratory and its arrangement

within the external building shell are illus-

trated in Figure 1. The three identical pri-

vate rooms (P—each 2.6 X 3.4 X 2.4 m)

are similar to small efficiency apartments

containing kitchen and bathroom facilities,

bed, desk, chair, rug and other furnishings.

The social living area (SL—4.3 X 6.7 X

2.7 m) is equipped with tables, chairs, sofa

beds, storage cabinets, and a complete

kitchen facility. The workshop (WS—2.6 X

4.1 X 2.7 m) contains benches, stools,

storage cabinets, tools,and a washer-dryer

combination. A common bath (B, Figure

1) serves the social living area and the

workshop. Access to the exterior walls of

the laboratory is provided by a four- to six-

foot corridor between the residential cham-

bers and the external building shell that

permits transfer of supplies and materials

through two-way storage facilities accessi-

ble from both sides. Remotely controlled

solenoid locks on doors and cabinets

throughout the environment provide for

experimental programming of access to

various facilities and resources, though at

least one unlocked door in each compart-

ment permits departure from the labora-

tory at any time in case of emergency and

preserves the right of subjects to terminate

their participation in an experiment at any

time.

The electromechanical control devices

throughout the environment are interfaced

with a computer system located in an ad-

joining laboratory support facility that

provides for experimental monitoring,

programming, recording, and data analy-

sis. The computer is linked to a Cathode

Ray Tube (CRT) Display Device within

each of the private rooms of the residential

laboratory, and an alpha-numeric keyboard

with each display unit provides for direct

communication with the system control.

The communication panel, in each indi-

vidual chamber incorporates the CRT unit

and also includes both a cassette tape

player and a telephone intercom for ex-

changes between subjects within the envi-

ronment. Audio and video equipment in

each of the residential chambers permits

continuous monitoring during conduct of

an experiment.

Behavioral programming procedures have

been developed to establish and maintain

Fig. 1. Diagrammatic representation of the overall floor plan of the laboratory and its arrangements within

the external building shell.

SMALL GROUPS IN PROGRAMMED ENVIRONMENTS

BEHAVIORAL PROGRAM

een rrr. |

he —")

bebesteh sl ely “bet

INVENTORY OF ACTIVITIES

NOTATION

FULL NAME

HEALTH CHECK

PHYSICAL EXERCISE

TOILET OPERATIONS

AUTOGENIC BEHAVIOR

FOOD ONE

SLEEP

PRIVATE ARITHMETIC

PROBLEMS

GROUP ARITHMETIC

PROBLEMS

READING

WORK TWO

PUZZLE ASSEMBLY

MANUAL BEHAVIOR

REQUISITION

WORK THREE

FOOD TWO

FOOD THREE

MUSIC

PRIVATE GAMES

COMMUNI CATION

LIMITED TOILET

OPERATIONS

CONDITION B

BRIEF DESCRIPTION

TEMPERATURE, PULSE, WEIGHT, STATUS REPORT

300 CORRECT PRESSES ON AUTOMATED TASK

USE OF PRIVATE BATHROOM AND CONTENTS OF DRAWER

CONTAINING TOILETRIES

RELAXATION EXERCISES ON CASSETTE TAPE

TWO SELECTIONS FROM A LIST OF LIGHT FOODS

USE OF BED AND PRIVACY CURTAIN

150 CORRECT SOLUTIONS OF ARITHMETIC PROBLEMS

OPTIONAL, CONTRIBUTE CORRECT SOLUTIONS OF

PROBLEMS TO GROUP RATIO CRITERION

ACCESS TO BOOK

PROBLEMS, EXPERIMENTS, ASSEMBLY PROJECTS

ASSEMBLE A PUZZLE

ACCESS TO ART MATERIALS

EARN DELAYED DELIVERY OF TREATS

ACCESS TO WORKSHOP

PRIVATE MAJOR MEAL

MAJOR MEAL IN RECREATION ROOM, GAMES

EARN A CASSETTE TAPE

ACCESS TO SOLITARY GAMES

ACCESS TO INTERCOM

ACCESS TO ESSENTIAL TOILET FACILITIES

CHANGE IN PROGRAM CONDITION

Fig. 2. Diagrammatic representation of a typical behavioral program governing the sequential and contin-

gent relationship of activities.

stable performance baselines and to pro-

vide for systematic experimental manipula-

tion of performance interactions during ex-

tended residential studies in the laboratory

environment. A behavior program is de-

fined by: (1) anarray of activities or behav-

ioral units, and (2) the rules that govern the

relationship between these activities. Fig-

ure 2, for example, illustrates diagrammat-

ically the fixed and optional sequences that

characterize a typical behavioral program

used to establish baseline performances for

these experiments and the array of compo-

nent activities that make up such a pro-

gram. Variations in this program as re-

quired for specific experimental studies will

be described below. Each box in the dia-

gram represents a distinct behavioral unit

and performance requirement, with pro-

gression through the various activities pro-

grammed sequentially from left to right.

Regardless of the fixed or optional se-

quence selected, all behavioral units are

scheduled on a contingent basis such that

access to a succeeding activity upon satis-

faction of the requirement for the preced-

ing unit.

At the bottom of Figure 2 are two activi-

ties with more general rules. The Limited

Toilet Operations (LTO) activity, which al-

lows access to essential toilet facilities, is

the only activity that can be selected at any

time. The Communication (COM) activity

allows access to the intercom for intersub-

ject communications. A subject is permit-

ted to use the intercom to initiate or answer

4 JOSEPH V. BRADY AND HENRY H. EMURIAN

a communication only if he is between any

two program activities. Although the

Communication activity is available be-

tween activities, an actual conversation re-

quires at least two subjects’ simultaneous

presence within the Communication activ-

ity. Conversing subjects, however, whether

in pairs or all three at once, could be lo-

cated at different sequential positions within

the behavioral program. The CON B nota-

tion at the bottom of the diagram refers toa

program change determined by the require-

ments of a specific experiment, as described

below. A manual of instructions detailing

the program and use of environmental re-

sources 1s contained in each room of the

environment. Subjects follow the behav-

ioral program throughout the periods of

residence, and pairs of research assistants

monitor the experimental environment con-

tinuously with audio and video equipment

located, with the subject’s awareness, in

each room of the environment.

The selection of human research partici-

pants for these small group studies has

been based upon the response to recruit-

ment notices on local college bulletin boards

and in local newspapers. Over 150 male

and female volunteers have been screened

and accepted for participation in the re-

search following extensive evaluation of

their personal, educational, and medical

histories. Such research participants have

been thoroughly familiarized with the ob-

jectives of the small group studies, the op-

erational features of the laboratory, the

experimental methodology, and the per-

formances required during several daily 1n-

forming sessions prior to completion of a

witnessed written consent form and initia-

tion of an experiment.

Research Results

Over the past several years, volunteer re-

search participants have participated in a

series of experimental group studies involv-

ing continuous residence for varying peri-

ods within the programmed laboratory en-

vironment. Early experiments involved

simply confinement and isolation of two-

person groups for relatively brief 24-hour

periods to demonstrate the adequacy of the

equipment and to determine habitability

under conditions that required only min-

imal, and basically biological, activity se-

quences, é.g., eating, sleeping, group inter-

actions, etc. The major findings and

conclusions from these initial studies were

that the experimental equipment and resi-

dential setting were capable of sustaining

stress-free living conditions for at least

these brief 24-hour periods. The second

phase of the research involved extending

the length of these studies from one to

three, and then to ten days of continuous

residence in the laboratory and introducing

programmatic sequencing of performance

activities. The major findings and conclu-

sions were not only that such small groups

could be maintained under stress-free liv-

ing conditions for these more extended pe-

riods in the experimental environment but

also that the sequential contingency per-

formance program was supportive of both

individual and group behavioral produc-

tivity."°

These early studies also emphasized the

differential importance of selected program

components (e.g., social activities) in main-

taining individual and group performance

and the sensitivity of behavioral interac-

tions to experimental manipulations (e.g.,

program condition changes and reversals)

over the course of extended residential pe-

riods. Consequently, a series of more sys-

tematic and extensive studies’ was under-

taken to focus upon the motivational and

emotional effects of varying social interac-

tion conditions. A “‘cooperation”’ condi-

tion was in effect when all three subjects

were required to select simultaneous access

to a group area before it became available

for use. Typically, subjects would use the

intercom several activities in advance to

plan subsequent selection of a social activ-

ity. They would then pace their individual

schedules accordingly to arrive at the choice

point in the program at approximately the

same time. In contrast to the cooperation

contingency, a “‘non-cooperation”’ condi-

SMALL GROUPS IN PROGRAMMED ENVIRONMENTS 5

tion was in effect when the group area was

accessible singly, without regard to the

other subjects’ activity selections. For ex-

ample, a single subject could select an ap-

propriate activity at the choice point and

could then leave his private room imme-

diately and enter the chosen group area

even though the other subjects were en-

gaged in private activities. Of course, the

other subjects could also have access to the

same group area at the same time, but they

were not required to enter and leave together.

The results of this experiment showed

clearly that the systematic effects of such

contingency management procedures could

be discerned not only upon the social be-

havior of the group, but also upon collat-

eral individual behaviors that character-

ized performances within the continuously

programmed environment. Enhanced lev-

els of intersubject program synchroniza-

tion and intercom frequencies were ac-

companied by comparable increases in the

magnitude of triadic episodes during the

cooperation condition. Not only the per-

cent of total time spent in triadic social ac-

tivities, but the durations of triadic epi-

sodes combined with corresponding social

interaction measures suggested a poten-

tially important consequence of coopera-

tion contingencies in maintaining more du-

rable social interactions when continued

access to the group areas accrued primarily

as a result of the frequency of the social

interactions.

Cooperation contingency effects on tri-

adic conditions would seem to be of partic-

ular significance when considered in light

of group fragmentation effects observed

during non-cooperation conditions. The

distribution of dyadic percent times into

two high-pairing subjects and one low-

pairing subject within the groups suggested

development of a two-person in-group and

a relative social isolate during the non-

cooperation condition. And the extent to

which motivational and emotional interac-

tions participated in the social contingency

effects is suggested by the results observed

with the very first group when the change

from cooperation to non-cooperation con-

ditions was programmed. Within minutes

after the condition change was introduced

during the course of a triadic social episode

at the end of Day 4, Subject 2 became in-

volved in an altercation with the other two

subjects and abruptly returned to his indi-

vidual chamber. During the ensuing three

days of the non-cooperation condition,

Subjects | and 3 continued to engage in

frequent dyadic social interactions that ex-

cluded Subject 2. More importantly, the

performances of Subject 2 with respect to

the maintenance of ““housekeeping” chores

in his individual chamber deteriorated, and

significantly, the error rate reflected in his

“private mathematics” performances in-

creased dramatically during the period

immediately following the disruptive emo-

tional interaction.

A more extended analysis of such social

interaction contingency effects was under-

taken with four additional groups of three

male subjects who participated in a series

of 10-day experiments to evaluate further

the effects of subject pairing on individual

and social behavior.’ In addition to the tri-

adic contingencies studied previously, dyadic

contingencies were scheduled when simul-

taneous occupancy of a group area was

permitted to any combination of two, and

only two, subjects. Solitary access to group

areas also was permitted to parcel out the

reinforcing effects of social episodes, inde-

pendently of those attributable to the ac-

cessibility of a larger space. Additionally,

included in the behavioral program was a

group task that allowed individual contri-

butions to a group criterion that had to be

satisfied before triadic or dyadic episodes

could occur. A “‘triadic’’ program condi-

tion was in effect when either of two social

activities within group areas (/.e., WK3 or

FD3) was accessible only when all three

subjects selected it together. In contrast to

the triadic condition, a “dyadic”’ program

condition was in effect when WK3 and

FD3 were accessible for social activities by

any combination of two, and only two, sub-

jects. In both conditions, subjects were re-

quired to enter and leave the group areas at

the same time. Once a group area was oc-

6 JOSEPH V. BRADY AND HENRY H. EMURIAN

cupied by a dyad, access to that area by the

third subject was denied until the activity

was terminated by the dyad.

Throughout social episodes in the rec-

reation room, 10-second observational

samples were taken on a variable-interval

schedule averaging five minutes between

samples. During each observational sam-

ple. the subjects’ identification numbers

(7.e., 1,2,and 3) were recorded directly ona

schematic diagram of the room by two in-

dependent observers, giving the subjects’

exact location in the room and their prox-

imity to one another. On the basis of these

observations, a social distance scale was

computed for each subject reflecting his

physical proximity to the other two sub-

jects during triadic social episodes. A given

subject’s score for a single observational

sample was the sum of the distance between

himself and the other two subjects. The re-

cordings upon which the social distance

scores were based showed high inter-rater

reliability (correlation = +.96).

The results of this experiment showed

that the status of a closed three-person so-

cial system changed when social opportun-

ities were limited to dyads as compared to

the triad. Under such dyadic conditions,

durations of social contacts were briefer,

and performance schedules drifted apart as

reflected by decreased levels of harmony in

the selection and completion of behavioral

program activities. Additionally, daily re-

sponse outputs ona task having social con-

sequences were more often omitted during

dyadic conditions. These results illustrate

the group fragmentation effects previously

observed during a non-cooperation condi-

tion® in a situation in which triadic social

interactions were prohibited, rather than

being optionally available.

Although division of group members oc-

curred in the non-cooperation condition of

the previous experiment, all subjects con-

tinued to have both dyadic and triadic so-

cial interactions, and consequently, no sub-

ject was ever completely isolated from

group activities. In the present experiment,

however, group fragmentation effects were

stronger during the dyadic condition than

observed under the previous non-coopera-

tion condition. Under dyadic conditions,

three of the four groups in the present study

had a lone member who failed to have any

direct social contact for several successive

days. These differences may be attributa-

ble, at least in part, to the more demanding

contingencies that were in effect for social

contact under dyadic conditions, where re-

sponding was required on the group task,

and two subjects had to cooperate in the

choice of a group area before social behav-

ior could occur. That dyadic episodes oc-

curred at all when free access to the large

group areas was available shows the moti-

vational effects of even such minimal social

contact.

Of particular interest were the results ob-

tained when the mean social distance scores

for all subjects observed over triadic epi-

sodes were rank ordered from high to low

and plotted against corresponding percents

of time in dyadic social episodes during

dyadic conditions, as illustrated in Figure 3.

The correlation between these social dis-

tance scores and percents of time in dyadic

episodes during dyadic conditions revealed

a significant inverse relationship (r = —.79,

p < .01). These data show that as the phys-

ical distance between subjects increased for

a given subject within a triadic group set-

ting, the proportion of time he spent in so-

cial episodes decreased during dyadic con-

ditions, predicting with reasonable accuracy

the sociability of group members under

conditions requiring group fragmentation.

The robust effects of social contingencies

upon the behavior of small groups within

the continuously programmed environment

provided the basis for some extensions of

such group interaction analyses to investi-

gate the role of explicitly programmed mo-

tivational operations.’ Three 10-to-12 day,

three-person experiments incorporated a

‘“‘work unit’? completion contingency de-

termining the amount of group remunera-

tion for participation in the study. In all

previous experiments, individual subjects

received a fixed per diem allowance (e.g.,

$25) for participation regardless of their

performance. In contrast, this series of

SMALL GROUPS IN PROGRAMMED ENVIRONMENTS 7

range

VVC MWTCXC@X™W@M!t'!#tz=——

VMVVMVC@CCCMMMMtz

Vdd

YY YYT@@| CX@_ @ dane

MEAN SOCIAL DISTANCE SCORES (FEET)

WV VCCCMMMt###t-—

MVMCVVVVVVC@CCC€@t##I00--$£$

SIG]

N distance scores

B dyodic time 20

Vet

M.VMMMMM“E

VVC"

PER CENT TIME IN DYADIC EPISODES

$2G1 S3G] $2G3 SIG3 SIG4 $2G4 S3G4 $3G3 S$IG2 S3G2 $2G2

SUBJECT(S) & GROUP (G)

Fig. 3. All subjects’ mean social distance scores, rank ordered from high to low, plotted against correspond-

ing percents of time in dyadic social episodes during dyadic conditions. The ranges of social distance scores

between episodes are given for all groups except Group 2, which had only one triadic episode in the recreation

room.

motivational studies provided a program-

matically controlled amount of remunera-

tion for each completed work unit by an

individual subject in the form of a contri-

bution to a group “bank account,” with

group earnings divided evenly among the

Participants upon completion of the

experiment.

The basic fixed and optional compo-

nents of the behavioral program continued

to be in effect during these experiments

with a sequence of work unit activities

made available independently of the re-

maining sequentially arranged activities.

This work unit could be selected upon

completion of any activity within the full

behavioral program. Once a work unit had

been selected, all five activities had to be

completed before the subject could resume

the behavioral program at the location

where the work unit was voluntarily in-

itiated. The parameters for the several

component activities (/.e., mathematical

problem solving, manual lever operations,

physical exercise, and health check) were

chosen such that one to two hours were re-

quired for completion of a work unit.

The consequences of completing a work

unit were systematically varied to assess the

effects of alternative behavior-consequence

relationships under program. control.

Throughout the initial four days of the first

experiment, for example, a “‘positive”’ (/.e.,

appetitive) relationship was in effect whereby

completion of a work unit by an individual

subject produced a $10 deposit to the group

bank account. Throughout the next four

days of the experiment, a “‘negative” (/.e.,

avoidance) relationship was in effect such

that work units no longer produced $10 in-

crements in the group bank account, but

rather were required of the participants in

order to avoid withdrawals of similar mag-

nitude. That is, work performance require-

ments for Days 5 through 8 provided thata

$10 withdrawal be made from the group

bank account for each uncompleted work

sequence below an assigned daily total (/.e.,

20) determined on the basis of the group

productivity sequences completed per 24

hours. This group requirement could be

satisfied under any conditions of individual

work scheduling or distribution decided

upon by the group participants. Finally,

8 JOSEPH V. BRADY AND HENRY H. EMURIAN

the last two days of the experiment, Days 9

and 10, were programmed asa return to the

conditions in effect during the first four

days.

The work unit contingency maintained

substantial productivity levels for all sub-

jects throughout the course of the experi-

ment. No participant completed fewer than

five work sequences per day witha range of

5 to 14 units. A distinguishable and rela-

tively stable pattern of group work perform-

ances and social interactions emerged dur-

ing the first four “‘appetitive’’ days of the

experiment. Although all members of the

group were not contributing equally to the

group bank account (/.e., one of the three

participants consistently completed fewer

work units than the other two during this

period), a high degree of group cohesive-

ness was reflected in the social episodes, the

intercom exchanges, and the frequent use

of the ‘‘audit”’ option available to each par-

ticipant for monitoring the status of the

group bank account and the individual

contributions thereto.

In contrast, the second four-day segment

of the experiment (i.e., Days 5 through 8)

with work performances aversively main-

tained by avoidance of group monetary re-

sources diminution was characterized by a

dramatic change in the relatively stable

work-rest pattern observed during the first

four days, and by a progressive deteriora-

tion of group cohesiveness. Beginning with

Day 5, work schedules were drastically al-

tered by the group, and the two most pro-

ductive members of the group became

openly intolerant of the third participant’s

‘““below-par”’ performance. Asa result, this

low-productivity participant was progres-

sively isolated from the group and spent

Days 7 and 8 alone in his private chamber.

Concomitantly, all three members of the

group became openly hostile and vehe-

mently expressive of their displeasure with

the experimenters who were perceived as

responsible for this obviously ‘‘aversive”’

state of affairs.

Paradoxically, group productivity was

not materially affected by the change from

appetitive to aversive maintaining condi-

tions, and the absolute number of work

units completed by the low-productivity

group member actually increased slightly

during Days 5 through 8. Both the daily

work-unit frequency and the total number

of hours devoted to work by the group par-

ticipants were maintained at relatively sta-

ble levels throughout the two four-day in-

tervals, and they remained sufficiently high

during the avoidance contingency in effect

from Days 5 through 8 to prevent even a

single withdrawal from the group bank ac-

count. And on the basis of a more detailed

analysis of the several component tasks in

the work units, there was little evidence

that performance effectiveness was differ-

entially influenced by the two conditions.

This finding is in marked contrast to the

dramatic changes in group cohesiveness,

ratings of program control conditions, and

both interpersonal (e.g., “‘irritation’’) and

intrapersonal (e.g., ‘“‘“mood’’) ratings re-

corded by the subjects during Days 5

through 8.

Although these socially disruptive by-

products of the aversive control procedures

in effect during the avoidance segment of

the experiment did not produce obvious

decrements in either individual or group ef-

fectiveness on work unit performance,

changes did occur in the distribution of

work units as a function of the transition

from appetitive to aversive motivational

conditions. These effects are presented

graphically in Figure 4, which shows the

distribution of work unit time (shaded

segments) over successive days (depicted as

24-hour clocks) under each of the three

program conditions. During Days | thru 4,

the completion of one or two work units

was typically followed by a rest break dur-

ing which a social episode (e.g., communal

meal) would usually occur. Additional brief

work periods would then generally occur

interspersed with individual or social rec-

reational interludes before sleep. In con-

trast, Days 5 through 8 were characterized

by a dramatic change in this work-rest pat-

tern with comparable numbers of work

units compressed into more restricted time

segments as shown on the 24-hour clocks

SMALL GROUPS IN PROGRAMMED ENVIRONMENTS

APPETITIVE

ie ei arr, or, attr,

- , » Pa a A >

. . . » ; "

. 4 _ : “

. ; ‘ . ‘

ee \ , + ‘ ;

y y ; ,

y sy ¥

x Y

AVY OIDANCE

APPETITIVE

EXPERIMENTAL DAYS

Fig. 4. The distribution of work unit time (shaded segments) over successive days (depicted as 24-hour

clocks) under each of the three program conditions.

for this avoidance period. This alteration in

the temporal distribution pattern effectively

insured that the daily group performance

requirement (i.e., 20 work units) was com-

pleted before any social or recreational epi-

sodes occurred. Significantly, the progres-

Sive deterioration of group cohesiveness,

reflected in the marked decrease in triadic

social interaction over Days 5 through 8,

developed concurrently with this change in

the work-rest pattern.

The extremely high work rates reflected

in the 24-hour clock distribution for Days 9

and 10 following reversal to the appetitive

motivational conditions of Days | through

4 probably accounts, at least in part, for the

group fragmentation that persisted through-

Out these final two days of the experiment.

While this final burst of work activity can

be attributed to some combination of

“emotional” facilitation occasioned by the

condition change (i.e., from aversive to ap-

petitive control) on the one hand, and the

“motivational” potentiation produced by

temporal proximity of the behavior-main-

tained consequence (i.e., the “pay-off” at

the end of the experiment), it is noteworthy

that this extremely high work output (far

exceeding those levels observed during any

of the 25 previous experiments conducted

in our laboratory) occurred in the absence

of any deleterious side effects to the sub-

jects. In fact, the rating data along with the

‘‘Mood Scale” assessments obtained dur-

ing each Health Check activity reflected

virtually complete recovery to the positive

levels that characterized the first four days

of the experiment.

To provide a more detailed analysis of

such motivational and emotional interac-

tions under aversive and appetitive pro-

gramming conditions, as well as to control

for order effects attributable to the tem-

poral sequence in which these diverse pro-

gramming conditions were presented, two

additional 12-day experiments were con-

ducted, one with three male participants

and one with three females. The experi-

mental methodology and general program-

ming procedures were basically similar to

those for the experiment described above.

As with the first group, the work unit con-

tingency, requiring approximately one hour

for completion, maintained substantial

productivity levels for all subjects in each

of these two additional groups. No subject

10 JOSEPH V. BRADY AND HENRY H. EMURIAN

completed fewer than two work units per

day witha range of 2to 16 units. Within all

groups, the work unit outputs were more

evenly distributed among subjects during

the avoidance condition than during the

appetitive condition. A comparison between

the two conditions of the differences be-

tween the highest and lowest work fre-

quency for all subjects in each group (under

the assumption that such differences ap-

proach zero when variability is absent)

showed a significant effect.

Subjects with a relatively low daily work

unit output during the first appetitive con-

dition showed a work unit performance in-

crement during the succeeding avoidance

condition, and a marked increase in daily

work unit frequency was observed when

the appetitive condition was reintroduced.

With respect to the intrapersonal aspects of

the program condition effects, almost all

subjects reported mood changes between

program conditions on the Depression fac-

tor of the Lorr’s Outpatient Mood Scale

that was administered during each Health

Check activity. Eight of the nine subjects

showed the highest rating during the avoid-

ance condition, and, for a pooled analysis,

the avoidance condition was associated

with significantly higher depression rat-

ings. Additionally, on a 4-point scale re-

flecting degree of irritation (1 = none to

4 = extreme) with the program condition,

all subjects in each group displayed more

irritation during the avoidance condition

than during corresponding appetitive pro-

gram conditions. As with Group I, the

three male subjects in Group 2 showed

local effects of the avoidance condition in

the form of clear displays of aggression.

-Members within Group 2 evidenced de-

structive behaviors in relationship to lab-

Oratory property (e.g., kicking the walls

and damaging the furniture) and repeat-

edly failed to conform to the requirements

of the behavioral program. Incontrast, the

three female subjects in Group 3 displayed

no such aggressive or hostile behaviors,

even after six successive days under the

avoidance contingency, though their pro-

gram rating scores did show a modest de-

gree of intermittent irritation in the course

of this extended exposure to the ; aversive

avoidance condition.

The reliability and generality of the ef-

fects of aversive performance schedules

were further demonstrated in a 6-day sys-

tematic replication with three male sub-

jects.'* During days when subjects’ work

on Multiple Task Performance Battery had

the effect of preventing reductions in ac-

cumulated group earnings, all subjects

complained, one subject stopped working,

and another subject’s productivity declined.

When identical work had the effect of in-

crementing group earnings, such by-prod-

ucts of aversive control were absent. These

results confirmed the outcomes of our pre-

vious analyses of aversive reinforcement ef-

fects under programmed laboratory condi-

tions, and they suggest that under such

conditions where performance requirements

are continuous and challenging, a group

may fail to complete its mission (7.e., com-

pletion of assigned work).

The foregoing investigations clearly es-

tablished social variables as fundamental

contributors to the overall status of a con-

fined microsociety, and they emphasized

the sensitivity of such variables to a range

of experimental manipulations having op-

erational significance. Throughout such

studies, participants were observed to seek

social interaction under one set of condi-

tions (e.g., triadic social contingencies and

positive performance outcomes) and to

withdraw from such interaction under other

conditions (e.g., pairing social contingen-

cies and aversive performance outcomes).

Thus, the joining and leaving of a group by

participants under circumstances encom-

passing more than a single environmental

condition appeared to generate social ef-

fects reflecting important dynamic processes

requiring systematic experimental analysis.

Accordingly, six additional studies were

conducted to assess the effects on individ-

ual and group behavior of a novitiate par-

ticipant’s introduction into and subsequent

withdrawal from a previously established

SMALL GROUPS IN PROGRAMMED ENVIRONMENTS 11

and stable two-person social system.'' The

objectives of these studies were to focus

upon (1) the social mechanisms and tem-

poral properties associated with the inte-

gration of such a participant into an estab-

lished group, and (2) sources of group

disruption or cohesiveness fostered by his

or her presence. Additionally, measures of

hormonal levels based upon the collection

of total urine volumes throughout the

course of the studies focused upon changes

in the androgen testosterone as an endocrin-

ological index of demonstrated sensitivity

to social interaction effects in both ani-

mals'*’> and humans.'*'’ Such a behav-

ioral biological analysis was implemented

to provide a more comprehensive assess-

ment of the personal and social impact

generated by the introduction and with-

drawal of new members with an established

group.'*

The paradigm adopted for experimental

analysis of effects of changes in group size

and composition was.as follows. A two-

person group resided for ten successive

days within the programmed environment,

and members of that dyadic team operated

performance tasks for their earnings. Dur-

ing the course of that ten-day period, a

third “novitiate”’ participant was introduced

into the programmed environment for sev-

eral successive days, thereby increasing the

size of the group from two to three members.

A typical “introduction” period with three

group members lasted four days, and it us-

ually began on Day 4 or Day 7 of a ten-day

experiment.

The rule conditions of the behavioral

program that were associated with the no-

vitiate’s entrance into the group differed

across successive investigations. In some

Studies, the novitiate received a per diem al-

lowance, and he was not required to work

for compensation, although he was permit-

ted to contribute to the performance tasks

that advantaged the two established group

members. In other studies, the novitiate

was required to work for compensation by

competing with the two other group mem-

bers for access to the single work task con-

sole located within the workshop. Finally,

the series of investigations was undertaken

with both male and female novitiates and,

in some cases, with novitiates and dyadic

members who had previously participated

in earlier residential studies.

In studies where the novitiate’s presence

primarily served as additional social stimu-

lation for the established dyad and as a

source of information regarding current

events outside the laboratory, the two-per-

son group showed a resistance to granting

the novitiate permission to work, even

when such work would have provided relief

from operating a demanding task. Impor-

tantly, however, as the three-person condi-

tion continued over days, novitiates were

observed to contribute to work productiv-

ity to a degree that was almost equivalent

to the productivity of the dyadic members.

Since there were no external incentives for

a novitiate’s work in these first introduc-

tion studies, these findings emphasized the

influence of social processes alone in main-

taining performance productivity at least

within these cohesive group situations. In

contrast, transitions between two-person

and three-person conditions were not al-

ways smooth in groups where the novitiate

had to work for compensation. When a no-

vitiate forcefully intruded himself into the

dyad’s customary work schedule, his tes-

tosterone levels rose or fell generally in

close relationship to his success or failure,

respectively, to gain and maintain access to

the work station according to a schedule

that was least disruptive to his wake-sleep

cycles as determined during several baseline

days preceding his introduction into the

group. When sleep discipline was imposed,

and when a novitiate was cooperative in

negotiating an orderly sequence of using

the work task console, notable changes in

testosterone were not observed in any crew

participant. When a female novitiate was

introduced into a two-man group, wake-

sleep cycles and work periods were erratic

throughout the three-person condition. Such

effects were associated with the absence of

notable androgen changes, even by a dy-

12 JOSEPH V. BRADY AND HENRY H. EMURIAN

adic member who, as a novitiate in an ear-

lier study, had successfully maintained his

wake-sleep cycles and had shown a striking

increase in testosterone when he joined the

group.

The significance of these behavioral-bio-

logical interactions is to be understood in

terms of the completeness of the resulting

account of effects of the experimental vari-

able, 7.e., the introduction of a novitiate

into an established group. With regard to

the relevance of the endocrinological rela-

tionships observed under such conditions,

it seems reasonable to suggest that the

adaptive significance of any hormonal re-

sponse can be best interpreted in terms of

the consequences of that response at the

metabolic level. Although research on the

androgens has typically emphasized repro-

ductive functions, it has been established

that testosterone has potent ‘‘anabolic”

properties, promoting protein synthesis in

muscle and many other tissues’”~' and

potentiating some effects of insulin on car-

bohydrate metabolism.*” Whether these

“anabolic” effects of testosterone and the

androgenic metabolites play any apprecia-

ble role in general organic or energy me-

tabolism must, of course, await clarifica-

tion by further investigative analysis. But

at the very least, this series of experiments

emphasized the importance of a multidi-

mensional analysis of the behavioral and

biological interactions that determine the

adaptations and adjustments of small groups

in confined microsocieties.

Currently ongoing experiments were de-

signed to extend the research paradigm

from analyses of “introduction” effects to

the analysis of ‘‘replacement’’ effects.

Whereas the previous investigations changed

group size as an experimental variable or

treatment, the most recently initiated stud-

ies held group size constant to evaluate ef-

fects of replacing a member of an estab-

lished three-person group with a novitiate

participant.”” These replacement analyses

involved important elements of continuity

with the earlier studies in the manner of

being systematic replications of those in-

vestigations. In a research strategy based

upon systematic replications, as compared

with exact or direct replications, effects of

the experimental variable or treatment are

demonstrated by affirming the consequent,

in which case each successive replication

incrementally contributes to an understand-

ing of effects that can be reliably attributa-

ble to the antecedent condition (e.g., intro-

ductions or replacements). The generality

of the behavioral processes is assured by

showing similar relationships across a broad

range of circumstances (e.g., subjects, order

and duration of experimental conditions,

performance tasks, group size, etc.). This

research strategy as adopted by the pro-

grammed environment unit has proved to

be most productive and economical.

A typical replacement investigation pro-

ceeded as follows. An original three-person

group resided in the programmed envi-

ronment for five successive days. At the

end of Day 5, one of the original group

members was withdrawn, and he was re-

placed by a novitiate participant who,

along with the remaining two original

members, formed a new group for the next

five successive days. Consecutive studies

differed in terms of (1) the decision rule by

which an original group member was with-

drawn, and (2) the number of baseline days

that came before group formation.

In the first replacement experiment, for

example, three-person group members re-

sided in their private rooms for a two-day

baseline ‘“‘alone”’ period during which time

access to the intercom, to social activities,

and to the work station was prohibited.

This two-day period provided a necessary

hormonal reference against which to assess

endocrine responses in relationship to in-

itial group formation. On Day 3, all activi-

ties previously prohibited were made avail-

able to the group, and each member was

required to operate the work task for indi-

vidual compensation. As in the introduc-

tion experiments, there was only one work

task console located within the workshop,

and subjects occupied the workshop singly

on a self-determined rotational basis. This

procedure, then, permitted an evaluation

of the manner in which subjects occupied

SMALL GROUPS IN PROGRAMMED ENVIRONMENTS 13

the work station (e.g., duration of work pe-

riods, time-of-day of work periods, etc.) as

one of the principal dependent variables of

the experiment.

At the end of Day 5, whoever of the three

group members had earned the fewest

work task performance points, totaled

across Days 3-5, was withdrawn from the

experiment. This decision rule was known

by the group members before the experi-

ment began. The novitiate participant en-

tered the programmed environment on

Day 6, which wasa solitary baseline day for

all three subjects. On Day 7, the newly

formed group had access to intercom com-

munications, social activities, and the work

station that continued to be available

throughout Days 7-10. Thus, the two ten-

day participants were required to adjust to

the replacement of an original member,

and the novitiate member was required to

adjust to his entrance into an established

unit whose members shared a history of

having competed successfully to maintain

high levels of performance effectiveness.

The results of the first two replacement

analyses, which used the individual work

task console as the groups’ performance

requirement, emphasized the readjustments

that original participants underwent when

a change in team membership occurred.

When a male novitiate participant replaced

an Original team member, wake-sleep cy-

cles and time-of-day spent working were

somewhat less stable than observed when a

female subject, who had previously partic-

ipated in a 10-day residential study and

who had more experience than either of the

Original group members, replaced an origi-

nal male member. That such differences

were associated with interpersonal effects

was indicated by the negative ratings di-

rected toward the replacement participant

by original group members in the second

experiment.

The continuity of the behavioral—-biolog-

ical relationships observed in the “‘intro-

duction” studies was demonstrated in the

testosterone data from the first replace-

ment analysis. Figure 5 shows total urinary

testosterone for all subjects across succes-

sive days of that first experiment (G-1). The

novitiate participant is identified as ‘*S4.”’

With respect to the original group members,

S2 showed testosterone values that were

somewhat lower than the other two partic-

ipants. Importantly, these comparatively

lower values were evident during the first

two baseline days of the experiment. When

group members commenced working on

Day 3, S2’s values increased somewhat

over baseline levels, but they continued to

be below the values exhibited by the other

two members across Days 3-5. Signifi-

cantly, S2 was the group member who did

not compete successfully to remain within

the experiment for ten days, and he was

withdrawn at the conclusion of Day 5. Fi-

paris [J Alone

iit Ge Triad

I2O0 S|

lOO

2 18o

= eu

S 20

= oe

= 20

® 120 fe,

S 100

2 780

eS 9,

a 40

e ep

Le,”

2 4 6 8 lO

Days

Fig. 5. Total urinary testosterone for all subjects

across successive days of the experiment. The novi-

tiate participant is identified as “S4.”

14 JOSEPH V. BRADY AND HENRY H. EMURIAN

nally, across Days 7-10, testosterone levels

progressively declined for S3 in relation-

ship to his shift in work and sleep times.

This latter effect confirmed the outcomes

observed in the introduction studies, and it

demonstrated, by systematic replication,

the generality of the behavioral-biological

processes governing such effects.

Summary and Conclusions

The objectives of this research project on

individual and small group performance

focused upon the development of princi-

ples relevant to human social interactions

in a residential programmed laboratory

and upon the investigation of procedures

to enhance performance effectiveness under

conditions of relative isolation and inter-

personal stress. Initial research activities

were directed toward the design and devel-

opment of an experimental microsociety

environment for continuous residence by

three-person groups of human volunteers

Over extended time periods providing for

programmable performance and recreational

opportunities within the context of a bio-

logically and behaviorally supportive set-

ting. Studies were then undertaken to ana-

lyze experimentally (1) conditions that

sustain group cohesion and productivity

and that prevent social fragmentation and

individual performance deterioration, (2)

motivational effects produced by the pro-

grammed incentives maintaining individ-

ual and team performance requirements,

and (3) behavioral and biological effects re-

sulting from changes in crew size and com-

position. The significance of these investi-

gative endeavors is to be understood in

terms of emergent motivational, social-in-

teraction, and group composition princi-

ples having practical relevance to the estab-

lishment and maintenance of operational

performance effectiveness.

The results of these small-group studies

clearly established the applicability of be-

havioral technologies and methodologies

to the experimental analysis of individual

and group performances within the context

of a human microsociety. Additionally, the

development of behavioral programming

techniques was demonstrably effective in

generating and maintaining such individ-

ual and group performances for unobtru-

sive monitoring and measurement with

precision and regularity over time. Further-

more, the interplay between conditions

both internal and external to the program

provided the occasion for observations of

performance in relationship to realistic in-

centive schedules. The application of such

contingency management principles, along

with the technological guidelines that pro-

vided the basis for design and development

of the programmed microsociety environ-

ment, were shown to be capable of sustain-

ing individual and group performance ef-

fectiveness and cohesion without notable

biological or behavioral disruption under

conditions of spatial restriction, social sep-

aration, enforced intimacy, and high per-

formance requirements.

It was clearly evident in these studies that

both individual and group productivity can

be enhanced under confined microsociety

conditions by the application of contin-

gency management principles to designated

‘high-value’? component tasks embedded

within the overall performance program.

Similarly, group cohesiveness can be pro-

moted, and individual social isolation and/or

alienation (i.e., group fragmentation) can

be prevented by the application of these

principles to social-interaction segments of

the performance program. On the other

hand, conditions found to result in pro-

gressive deterioration of individual and

group performance effectiveness included

aversive motivational contingencies. The

by-products of aversive schedules that

emerged under such circumstances were

found to be quantifiable in measures of

verbal performance (e.g., behavioral pro-

gram ratings), interpersonal performance

(e.g., verbal confrontation and aggression),

work performance (e.g., diminished pro-

ductivity), and group morale (e.g., irritabil-

ity and dysphoric mood). In contrast, positive

incentive schedules effectively counteracted

SMALL GROUPS IN PROGRAMMED ENVIRONMENTS 15

the disruptive consequences of aversive

contingencies while at the same time sup-

porting high work productivity free from

negative side-effects.

Related research results emphasized the

prominent involvement of behavioral and

biological processes that were functionally

related to adjustments and reactions when

changes occurred in group size and team

composition. The experimental analysis of

such “introduction” and “replacement”

effects emphasized the critical importance

of providing a structure transition in the

form of orientation and training regimens

for both novitiate and established group

participants to minimize potentially dis-

ruptive effects of altering the interpersonal

and social dynamics of a _ confined

microsociety.

References

1. Shaw, M. E. 1971. Group Dynamics: The Psychol-

ogy of Small Group Behavior. McGraw Hill, New

York.

2. Hare, A. P. 1976. Handbook of Small Group Re-

search. Free Press of Glencoe, New York.

3. Thorndyke, P. W. and M. G. Weiner. 1980. Jmprov-

ing Training and Performance of Navy Teams: A

Design for a Research Program. RAND, Santa

Monica.

4. Findley, J. D. and J. V. Brady. 1963. Exposure to

Total and Continuous Environmental Control with

a Single Human Organism. Symposium at the Sev-

enth Annual Human Factors Society meeting,

Palo Alto, California.

5. Findley, J. D. 1966. Programmed environments

for the experimental analysis of human behavior.

In: Operant Behavior: Areas of Research and Ap-

plication. W. Honig,ed., Appleton-Century-Crofts,

New York.

6. Brady, J. V., G. E. Bigelow, H. H. Emurian and

D. M. Williams. 1975. Design of a programmed

environment for the experimental analysis of so-

cial behavior. In: Man-Environment Interactions:

Evaluations and Applications. 7: Social Ecology.

D. H. Carson, ed., Environmental Design Re-

search Association, Wisconsin, pp. 187-208.

7. Brady, J. V. and H. H. Emurian. 1979. Behavior

analysis of motivational and emotional interac-

tions in a programmed environment. In: Ne-

braska Symposium on Motivation. R. Dienstbier

and R. Howe, ed., University of Nebraska Press,

Lincoln.

8. Emurian, H. H., C. S. Emurian, G. E. Bigelow and

J. V. Brady. 1976. The effects of a cooperation

contingency on behavior in a continuous three-

21.

person environment. J. Exper. Anal. Behav.,

25(3): 293-302.

. Emurian, H. H., C. S. Emurian and J. V. Brady.

1978. Effects of a pairing contingency on behavior

in a three-person programmed environment. J.

Exper. Anal. Behav., 29: 319-329.

. Emurian, H. H., C. S. Emurian, F. R. Schmier and

J. V. ‘Brady. 1979. Notes on programmed envi-

ronment research. JSAS Catalog of Selected

Documents in Psychology, 9: 66.

. Emurian, H. H., J. V. Brady, J. L. Meyerhoff and

E. H. Mougey. 1981. Behavioral and biological in-

teractions with confined microsocieties in a pro-

grammed environment. In: Space Manufacturing

4. J. Grey and L. A. Hamden, eds., American In-

stitute of Aeronautics and Astronautics, New

York, pp. 407-421.

. Emurian, H. H., C. S. Emurian and J. V. Brady.

1982. Appetitive and aversive reinforcement

schedule effects on behavior: A systematic repli-

cation. Basic Appl. Soc. Psychol., 3(1): 39-52.

. Emurian, H. H., G. E. Bigelow, J. V. Brady and C.S.

Emurian. 1976. Small group performance main-

tenance in a continuously programmed environ-

ment. JSAS Catalog of Selected Documents in

Psychology, 167: 5.

. Eberhart, J. A., E. G. Keverne and R. E. Meller.

1980. Social influences on plasma testosterone

levels in male talopoin monkeys. Hormones Behav.,

14: 247-266.

. Bernstein, I. S., R. M. Rose, T. P. Gordon and C. L.

Grady. 1979. Agonistic rank, aggression, social

context, and testosterone in male pigtail monkeys.

Aggressive Behav., 5: 329-339.

. Scarmamella, T. J. and W. A. Brown. 1978. Serum

testosterone and aggressiveness in hockey play-

ers. Psychosom. Med., 40: 262-265.

. Elias, M. 1981. Serum cortisol, testosterone, and

testosterone-binding globulin responses to com-

petitive fighting in human males. Aggressive

Behav., 7: 215-224.

. Frankenhauser, M. 1979. Psychoneuroendocrine

approaches to the study of emotion as related to

stress and coping. 1978 Nebraska Symposium on

Motivation. University of Nebraska Press, Lincoln.

. Dorfman, R. I. and R. A. Shipley. 1956. Androgens.

Wiley, New York, pp. 218.

. Kochakian, C. D. 1964. Protein anabolic property

of androgens. Alabama J. Med. Sci., 1: 24.

Kochakian, C. D. 1975. Definition of androgens

and protein anabolic steroids. Pharm. Therap.

Behav., 1(2): 149-177.

Talaat, M., Y. A. Habib and M. Habib. 1957. The

effect of testosterone on the carbohydrate metab-

olism in normal subjects. Arch. Int. Pharma-

codyn., 111: 215.

. Emurian, H. H., J. V. Brady, J. L. Meyerhoff and

E. H. Mougey. 1983. Small groups in programmed

environments: behavioral and biological interac-

tions. Pavl. J. Biol. Sci. (in press).

. Sidman, M. 1960. Tactics of Scientific Research.

Basic Books, New York.

Journal of the Washington Academy of Sciences,

Volume 73, Number 1, Pages 16-27, March 1983.

Human Skeletal Remains From OGSE-MA-172

An Early Guangala Cemetery Site on the

Coast of Ecuador

D. H. Ubelaker

Department of Anthropology, National Museum of Natural History,

Smithsonian Institution, Washington, D.C. 20560

ABSTRACT

Recent salvage excavations sponsored by the Banco Central of Ecuador at site OGSE-MA-

172 on the Ecuadorean coast produced 27 burial features with an early Guangala cultural affil-

iation and radiocarbon dated at about 100 B.C. Analysis of the recovered skeletal sample in

November and December, 1981 revealed 30 individuals of both sexes and ranging in age from

newborn to greater than 50 years. Data are presented here on the skeletal content of each

feature, as well as demographic, morphological, pathological, and developmental informa-

tion derived from the entire sample.

Late in the 15th century AD, Europeans

unknowingly ventured into the Caribbean

and “‘discovered”’ the New World. In actu-

ality of course, the New World had been

““discovered”’ before and was occupied by

perhaps as many as 100 million persons

(Dobyns, 1966). Since European contact,

explorers, and more recently scholars have

debated the origin and biological history of

these first Americans. Answers to some of

the questions can be found in early descrip-

tions of the Americans by Europeans, but

most detailed biological information must

come from archeology, particularly mor-

tuary site excavation and human skeleton

analysis. Large numbers of well excavated

and well documented prehistoric skeletons

are now being assembled that enable us to

address some fundamental problems in

prehistoric skeletal biology relating to dis-

ease, demography, and population rela-

tionships.

In recent years, much of my own effort in

this regard has focused on the recovery and

analysis of human remains in Ecuador, a

country that shows remarkable prehistoric

cultural diversity for its size and a country

that has attracted considerable archeologi-

cal activity. Archeological hypotheses have

emerged on population movements and

subsistence that can be tested using biolog-

ical data, if adequate samples can be ac-

quired. The corpus of biological data avail-

able from Ecuador has grown rapidly in

recent years with the publication of new

data from the sites of Ayalan (Ubelaker,

1979, 1981), Cotocollao (Ubelaker, 1980b),

Sta. Elena (Ubelaker, 1980a) and Real Alto

(Klepinger, 1979). These new data augment

earlier studies by Munizaga (1965), Duck-

worth(1951) and Van Bork-Feltkamp (1965)

and offer an opportunity to understand

patterns of biological variation and evolu-

tion through time and space in ancient

Ecuador. Continuing archeological inves-

tigations sponsored by the Banco Central

of Ecuador promise to add to this under-

standing with the discovery of additional

human skeletal samples of established age

and cultural association. This paper fo-

cuses on new data collected recently froma

sample dated just after the period when ar-

HUMAN SKELETAL REMAINS FROM OGSE-MA-172 17

cheologists believe subsistence shifted to

agriculture. As such it not only adds an im-

portant documented sample to the series,

but offers a possible glimpse at some of the

biological impact of agriculture subsistence.

During the months of July and August

1980, Dr. Karen Stothert, with support

from the Museo Anthropologico, Banco

Central de Guayaquil, conducted a salvage

excavation in the coastal town of Valdivia,

at site OGSE-MA-172. The excavation fo-

cused on a 45 square meter area that was

threatened by proposed new construction.

Stothert’s timely excavation identified and

recovered 26 burial features with an early

Guangala cultural affiliation that have

subsequently been radiocarbon dated at

2030 + 60 years (Tx-4455) and 2370 + 60

years (Tx-4456) (Stothert, pers.comm.). At

the invitation of Stothert and the Banco

Central, I analyzed the recovered human

remains in Ecuador between December 1,

1981 and January 5, 1982. The human re-

mains were separated from the cultural

items and trucked to Quito for my analysis,

along with additional human remains from

the site of Real Alto. The material was ex-

amined in a makeshift laboratory outside

of Quito.

The skeletal remains were located in la-

beled plastic bags within the cartons. The

contents of each bag was first sifted through

a fine screen(1 mm mesh) to remove the as-

sociated soil. The remains were then washed

(water and soft brush), dried, reconstructed

with Duco cement and analyzed. Skeletal

material from each feature was processed

and analyzed individually to avoid any

possible mixing of material from different

features. All unpacking, processing, recon-

Struction, analysis, photography and re-

packing was done by the author. All rea-

sonably intact tibiae were separated for

radiographic analysis and then returned to

their individual bags and cartons after ra-

diographs were taken. Following analysis,

these cartons as well as those of Real Alto,

were returned to the Museo Antropologico,

Banco Central in Guayaquil by truck on

December 21, 1981.

The following presents the biological

analysis of the sample. Detailed informa-

tion on the site location, excavation tech-

nique, burial customs and cultural associa-

tions will be reported separately by Stothert.

All feature numbers were assigned by Sto-

thert and based on her field observations.

Information on feature location, depth,

position, etc. also was generously provided

by Stothert and will be published by her

later in greater detail. In general, the human

remains are well preserved but badly frag-

mented and in many cases incomplete. Ac-

curacy of estimating sex, age at death, liv-

ing Stature, pathology, etc. is diminished by

fragmentation and incomplete bone rep-

resentation. With some complete skeletons,

rather precise age estimates were possible,

while only general estimates could be made

from most isolated skeletal fragments.

General criteria used in estimating sex and

age are described by Ubelaker, 1978 and

references cited therein.

Sample Statistics

At least 30 individuals are represented in

the 26 features and human bone recovered

from disturbed areas. This estimate in-

cludes neither minor amounts of human

adult bone recovered from the disturbed

areas nor the occasional “‘extra’’ bones that

were found associated with primary skele-

tons. These bones were not counted as rep-

resenting separate individuals since they

may in fact have originated from features

already included. A seven to eight year old

child found in one disturbed area was in-

cluded since no other child of that age ap-

pears in the sample. This approach, of

course, assumes that all skeletal assem-

blages assigned separate feature numbers

are in fact separate individuals. This as-

sumption is supported by the fact that 21 of

the 26 features are primary skeletons. Of

the 30 individuals estimated to comprise

the sample, 21 (70%) are less than 10 years

of age and 17 of those probably died at

about the time of birth.

Nine adults are present, three males, one

female and five of undetermined sex. Adult

18 D. H. UBELAKER

age estimates ranged from about 21 years

(the female) to over 50 years, with an aver-

age adult age at death of about 39 years.

Summary age at death data are presented

in Table 1 in the form of a life table. Data

presented in this table describe the age

structure of the recovered sample and pres-

ent statistical demographic information

derived from the sample. Note, however,

that since the sample is so small (30 indi-

viduals) and the area excavated is relatively

limited, the sample recovered may not ac-

curately represent the entire cemetery sam-

ple, or the total number of deaths in the

population. Definitions of the symbols are

as follows: x = age interval in years; Dx =

actual number of deaths in each age inter-

val; dx = the percentage of deaths in each

age interval; lx = the percentage of survi-

vors entering each age interval; qx = the

probability of death in each age interval;

Lx = the number of years lived during

each age interval; Tx = the number of

years remaining in the lifetimes of all indi-

viduals entering each age interval; e°x = the

average number of years an individual en-

tering age interval x can expect to continue

to live.

Artificial Modifications of the Skeleton

Only one cranium, the adult male from

feature 76 shows evidence of artificial cra-

nial deformation. This individual displays

occipital flattening similar to that described

from other Ecuadorean sites dating after

about 2000 B.C. (Munizaga, 1976; Ube-

laker, 1980b, 1981). No other evidence of

artificial cranial deformation was discov-

ered although all other crania but one were

very fragmentary.

Observations of metatarsophalangeal al-

terations (Ubelaker 1979, 1981) are possi-

ble on only three adult individuals, two

males and one female. In the female from

Feature 45, alterations are present on both

first, the right second, the right third and

both fourth metatarsals. They are absent

on the left third, and both fifth metatarsals.

In the male from Feature 76, an alteration

is present on the right third metatarsal but

absent on all other metatarsals and both

first proximal foot phalanges. In the male

from Feature 108, alterations are present

on both first, both second and the left third

metatarsals and absent on the right fourth,

and both right fifth metatarsals. The one

first proximal foot phalanx present shows

the alteration. The presence of alterations

on these three individuals suggests their

habitual resting and/or work posture in-

volved kneeling with hyperdorsiflexion of

the toes (Ubelaker, 1979).

All teeth in the sample were examined

for the occurence of drilled perforations for

stone insets as described by Saville (1913)

and Ubelaker (1977) from other Ecuadorean

sites, for evidence of filing as found at Aya-

lan (Ubelaker, 1981) or for any other evi-

Table 1.—Life table, reconstructed from the OGSE-MA-172 sample.

X Dx dx Ix qx Lx Tx eux

0-.9 17 57 100 57 72 1308 13

1.0-4.9 3 10 43 23 153 1237 29

5.0-9.9 | 3 33 10 167 1083 33

10.0-14.9 0 0 33 0 167 O17 28

15.0-19.9 0 0 33 0 158 750 23

20.0-24.9 l 3 30 11 142 392 20

25.0-29.9 ] 3 Pe | 13 b25 450 17

30.0-34.9 l a 23 14 108 325 14

35.0-39.9 2 s| 20 33 83 217 1]

40.0-44.9 5 13 25 58 133 10

45.0-49.9 l 3 10 35 42 75 8

50.0-54.9 ] 3 7 50 25 33 5

55.0-59.9 l 3 3 1.00 8 8 3

HUMAN SKELETAL REMAINS FROM OGSE-MA-172 19

Table 2.—Measurements and observations of adult crania and mandibles.

Feature

39 45 76 108

Male Female Male Male

Measurement

Minimum frontal breadth 84

Nasal breadth 28

Maxillo-alveolar length 47

Maxillo-alveolar breadth 64

Palatal breadth 38

Bicondylar breadth 118 117

Bigonial breadth 116 86

Height of ascending ramus 60

Minimum breadth of ascending ramus 34 36 35 36

Height of mandibular symphysis 39 30

Corpal length of mandible 85 81

Observation

Mylohyoid bridge AA A AA rr

Accessory mental foramen AA A AA AA

Frontal groove AA AA

Supraorbital foramen AP A

Squamoparietal synostosis AA

Auditory exostosis AA PP AA

Marginal foramen of tympanic plate A AP AA

Tympanic plate dehiscence AA AA PP

dence of intentional dental modification.

No examples were found. One male from

Feature 39 does display an interproximal

groove between the mandibular right pre-

molars at the crown-root junction. The axis

of the groove runs buccal-lingually with a

slight mesial angle. Similar grooves have

been described from archeologically recov-

ered remains in North America (Ubelaker,

et al., 1969 Berryman et al., 1979) and

probably result from abrasion produced by

the repeated insertion of an instrument be-

tween the teeth in an attempt to relieve lo-

calized discomfort.

Living Stature

Living stature was estimated for five

adult individuals, three males, one female

and one individual of unknown sex. Male

Statures were 165 cm calculated from a fib-

ula length from Feature 39, 160 cm calcu-

lated from a tibia length from Feature 76,

and 157 calculated from a femoral length

from Feature 108. The female stature of

152 cm was calculated froma fibula length.

Feature 22 produced a stature estimate of

155 cm for an adult of unknown sex, calcu-

lated from a tibia length. The formulae of

Trotter and Gleser (1958) and Genoves

(1967), summarized by Ubelaker (1978:44-

45) were used to calculate the statures.

These estimated statures are very similar to

those estimated for other prehistoric Ecua-

dorean sites (Ubelaker, 1980a,b, 1981).

Measurements and Observations

Measurements and observations of the

crania and mandibles were recorded using

the system employed in previous investiga-

tions (Ubelaker, 1980a,b, 1981). Due to the

fragmentation and paucity of the adult re-

mains, data were collected from only three

males and one female. These data are pre-

sented in Table 2. Summary statistics are

not offered since the sample size is so small.

Pathology

A variety of expressions of disease occur

in the bones and teeth of this sample reflect-

20 D. H. UBELAKER

ing trauma, congenital abnormalities, in-

fectious disease and possible nutritional

disturbances.

Trauma

Two individuals show definite bony

modifications caused by trauma. The young

female of Feature 45 displays a healed frac-

ture of the right third metatarsal midshaft.

The 50-55 year old male of Feature 76

shows a circular depression 10 mm in di-

ameter on the right frontal (Figure 1). The

depression is located about 10 mm, right of

the midline, 53 mm above the right orbit.

A possible third example of trauma

comes from the young male of Feature 108.

A bony spur 15 mm in length occurs on the

superior medial surface of the left third

metatarsal. The spur may have been caused

by trauma to that area of the left foot.

Infectious Disease

Evidence of infectious disease occurs on

the left tibia of the adult in Feature 6. Well

remodeled periosteal apposition is located

on the ventral shaft extending superiorly 70

mm from a point 90 mm from the distal

end. In addition the male from Feature 76

shows considerable evidence of bony reac-

tion to infectious disease in the vertebrae.

The centra of the sixth cervical shows con-

siderable pitting, especially on the interior

surface. The centrum of the seventh cervi-

cal is nearly completely destroyed and that

of the first has been modified slightly. The

loss of the centrum of the seventh vertebra

has created a marked scoliosis at that

point. Centra of most upper thoracics are

not present for observation but the ninth

thoracic vertebra through the sacrum are

present and normal except for excessive os-

teophytic formation on the left side. This

Fig. 1. Circular Frontal Depression, Feature 76.

. |

HUMAN SKELETAL REMAINS FROM OGSE-MA-172 21

Fig. 2. Porotic Hyperostosis on Parietal from Feature 47.

individual also shows fusion between the

sixth and seventh cervical vertebrae be-

tween the fourth and fifth cervical verte-

brae, and between a proximal and middle

foot phalanx.

Porotic Hyperostosis

Evidence for porotic hyperostosis and

cribra orbitale is confined to infant and

child skeletons from three features. The

one year old infant of Feature 47 shows po-

rotic hyperostosis on both parietals (Fig-

ure 2), the squamosal area of both tempor-

als, the occipital and both orbits. An infant

from Feature 77/166 with an age at death

of 18 months shows fine perforations on

both parietals along the lambdoidal suture

and on the right parietal along the coronal

Suture, 20 mm from bregma. The three year

old from Feature 122 shows fine perfora-

tions in the upper left orbit (right orbit not

present).

Dental Caries

Observations on dental disease were re-

corded on 73 permanent teeth, 40 mandib-

ular and 33 maxillary. Carious lesions were

found on six teeth (8.22 percent of total)

from both males and females (Table 3). The

lesions were concentrated in the posterior

dention of both the maxilla and mandible.

Frequency of dental lesions is intermediate

between that reported for Buena Vista Val-

divia (0 percent) (Turner, 1978), Sta. Elena

preceramic (3 percent) (Ubelaker, 1980a),

Cotocollao (2 percent) (Ubelaker, 1980b),

and Ayalan non-urns (8 percent) (Ube-

laker, 1981) and the 11 percent reported for

the late Ayalan urn sample (Ubelaker,

1981).

22 D. H. UBELAKER

Table 3.—Frequency and distribution of caries in permanent teeth from OGSE-MA-172. (N = number of

teeth present. C = number of teeth with carious lesions.)

Mandibular

I € PM

N G N C N ¢

Male 4 0 4 0 8 0

Female 4 0 2 0 4 0

Sex unknown 0 0 0 0 l 0

Total 8 0 6 0 1s 0

Alveolar Abscess

Alveolar abscesses occurred in associa-

tion with 10 teeth (12 percent) from two in-

dividuals. The male in Feature 49 shows al-

veolar abscesses in the area of the three left

mandibular molars. The maxilla of unde-

termined sex from Feature 23 shows ab-

scesses associated with seven teeth, all inci-

sors and the left canine and premolars.

Alveolar abscesses were not present in as-

sociation with 76 teeth. The main mecha-

nism for the production of alveolar ab-

scesseS appears to be the exposure and

subsequent infection of the pulp cavity by

rapid attrition of the occlusal tooth surface.

Loss of Teeth

Fifty teeth (42 percent) had been lost

antemortem. One of these apparently repre-

sents a congenital absence (third molar).

The loss of all others appears to have been

caused by disease or related factors. Man-

dibular missing teeth consist of 6 incisors

(43 percent), 2 canines (25 percent), 4

premolars (25 percent) and 13 molars (54

percent). Maxillary missing teeth consist of

6 incisors (50 percent), 2 canines (7 per-

cent), 4 premolars (29 percent), and 12 mo-

lars (50 percent). Two factors probably

caused most of the tooth loss; alveolar ab-

scess and alveolar bone loss due to peridon-

tal disease.

Dental Hypoplasia

Only two individuals display marked

hypoplasia of teeth; the one year old of

Maxillary

M I c PM M

N Co. N. CoN) €C UNG ee

OL 2 0 2 0 3 1 0.)

DS yee - 0 2 0 is 0 Gee

2. e 0 0 1 0 1 0 SG

ee 6 0 5 0 7 | eye |

Feature 47 and the three year old of Fea-

ture 122. In the one year old, discoloration

occurs on the occlusal half of the crown of

the deciduous maxillary incisors, but not

on the other deciduous teeth. The discolor-

ation probably occurred when these teeth

were forming during the last months of inter-

uterine life. Hypoplastic deformation in

the three year old occurs on the crowns of

all mandibular deciduous teeth and on the

permanent mandibular canines (Figure 3).

The crowns of the deciduous teeth were

being formed at about the time of birth or

just before while the permanent canine

crown reflects a disturbance at about 2.5

years.

Dental Calculus

Dental calculus occurs on 30 (41 percent)

of the permanent teeth in this sample, 23

maxillary (30 percent) and 20 mandibular

(50 percent). In maxillary teeth calculus oc-

curs on 33 percent of the incisors, 40 per-

cent of the canines, 43 percent of the pre-

molars and 25 percent of the molars. In the

mandible, calculus occurs on 38 percent of

the incisors, 67 percent of the canines, 54

percent of the premolars and 46 percent of

the molars. Thus calculus generally is evenly

distributed throughout the dentition. Cal-

culus also shows no marked preference for

any particular tooth surface. On mandibu-

lar teeth, 19 deposits occurred on buccal

tooth surfaces while 22 occurred on lingual

surfaces. Maxillary teeth show 10 deposits

on buccal surfaces and nine on lingual sur-

faces. The deposits of calculus tend to be

slightly larger on the mandibular teeth. On

HUMAN SKELETAL REMAINS FROM OGSE-MA-172 23

Fig. 3. Hypoplasia of Permanent Canine Teeth, Feature 122.

mandibular teeth 7 deposits (17 percent)

were Classified as large and five (12 percent)

as medium. On the maxillary teeth no large

deposits were observed and only 2 (11 per-

cent) medium deposits were present.

Dental Anomalies

Three peg-shaped molars were observed

from two individuals. All are from maxil-

lary third molars. No other anomalous

conditions were noted.

Growth and Development

As noted earlier, 21 of the 30 individuals

in this sample are below the age of ten

years. Since many of these young skeletons

are primary and well preserved, the sample

offers an unusual opportunity to document

the growth and development of a prehis-

toric Latin American population. The sam-

ple size and preservation are still less than

ideal, but since no other comparative data

exist in South America the data need to be

recorded.

Only those individuals with both teeth

and measureable bones are included in this

study. Ages were first estimated by com-

paring the stage of dental formation with

the data summarized in Ubelaker’s

(1978:112-113) dental chart. Maximum

lengths of long bones without epiphyses

were then measured with a sliding caliper.

Measurements were taken of the humerus,

radius, ulna, femur, tibia and fibula. Meas-

urements were also recorded of the min-

imum length of the basilar, maximum

width of the ilium (anterior superior iliac

spine to posterior superior iliac spine)

ischium and pubis and the symphysis height,

corpal length, height of ascending ramus

and minimum breadth of ascending ramus

in the mandible. Data were collected from

16 individuals with both teeth and meas-

ureable bones, ranging in age from new-

born to about three years (Table 4). The

sample size is too small to produce detailed

growth charts but the data generally indi-

D. H. UBELAKER

09T

(aG|

ell

cCl

OOT7LE

C6 68

$6 cr

01 8 8

LY ase Ol

LI 91

9€

bl el

c9 v9

v9 89

£9 L9

91 78

sivdA Ul [eNPIAIPU] JO ady pue ainjeI4

aa|

6£

‘(uiu) SaU0g J[Npegns jo syugWaINSea||—'p 2/qQU],

snWel “Ose "1H

yisua] jedioy

9/qipuey

Ieyjiseg

siqnd

WINIyos]

wnt]

IOIARTD

einqi4

BIQIL

InwdJ

eulf)

snipey

sniswiny

u0g

a3V

o1njeo4

HUMAN SKELETAL REMAINS FROM OGSE-MA-172 25

cate that the rate of long bone growth from

birth to about three years is about the same

as that documented for some North Amer-

ican Indian groups but, at least for the

femur, greater than documented for Eski-

mos (Merchant and Ubelaker, 1977).

Some data are also offered in this sample

on the timing of union of the centra and

transverse processes of the vertebrae. In 12

infants between the ages of 0 and }5 years,

the three vertebral components remain sep-

arate. In the infant of Fe 10 of age .8 years,

the transverse processes are united to each

other but not to the centra. In another in-

fant, of age .8 years, (Feature 35b), the

transverse processes are united on the tho-

racics and lumbars but not on the cervicals.

No transverse processes are united to the

centra. In the 3.0 year old of Feature 122,

all transverse processes are united, but not

to centra.

Lines of Increased Density

All mostly intact tibiae (seven) were re-

moved during analysis for radiographic

study. These are from two subadults and

five adults of features 6, 39,45, 76, 108, 122

and 166. No lines of increased density and

no other abnormal radiographic conditions

were observed. Absence of lines does not

necessarily mean that lines were never

present. Some of the distal ends had been

broken post-mortem with some erosion of

the internal structure that may have elimi-

nated or reduced the areas of density.

Discussion

The small size and fragmentary nature of

this sample greatly limits its potential for

interpopulation comparison and generali-

zation about the biology of the early

Guangala population. This is especially

true of the demographic data that are most

Susceptible to sampling problems. The pres-

ence of all ages and both sexes indicates

that no major sampling problems exist, but

the very high proportion of infants and ab-

sence of deaths between the ages of 10 and

20 may shift if the sample could be ex-

panded by additional excavation.

The occipital flattening in Feature 76 is

not surprising since cranial deformation

has been documented from that approxi-

mate date in other Ecuadorean sites (Muni-

zaga, 1976; Ubelaker, 1980b, 1981) and

from the Guangala sites of Palmar (Van

Bork-Feltkamp, 1965) and La Libertad

(Duckworth, 1951). Observations of meta-

tarsophalangeal alterations on three adults

represent the first observation of that trait

in an Ecuadorean site other than Ayalan,

the site of the discovery of the trait. The in-

terproximal groove of the Feature 39 male

represents the first description of that al-

teration in an Ecuadorean sample.

Although living stature could be esti-

mated for only five individuals, the esti-

mates are very close to those derived for

Sta. Elena, Cotocollao, and Ayalan, sug-

gesting further that stature varied min-

imally in prehistoric Ecuador.

Just as in Van Bork-Feltkamp’s (1965)

Palmar sample, these skeletons show evi-

dence of porotic hyperostosis of the cranial

vault and cribra orbitale, both probably re-

flecting skeletal response to anemia pro-

duced by nutritional deficiency. Evidence

of trauma (fractured metatarsal of Feature

45 and frontal depressed fracture of Fea-

ture 76), and infectious disease are also

present.

The pattern of dental disease shows con-

siderable tooth loss and alveolar abscess-

ing. Most alveolar abscesses appear to be

caused by excessive attrition of the occlusal

surface, exposure of the pulp cavity and

subsequent infection. This process, as well

as peridontal disease, probably largely pro-

duced the excessive loss of teeth in the adult

population (42 percent loss ante-mortem).

Dental caries may also be a contributing

factor to tooth loss since 8.2 percent of ex-

isting permanent teeth show at least one

carious lesion. This caries frequency Is sub-

stantially higher than that documented for

26 D. H. UBELAKER

the Sta. Elena Vegas and Cotocollao sam-

ples, and only slightly lower than that of

the Ayalan urn sample.

Dental calculus occurs on 41 percent of

permanent teeth, is evenly distributed among

all tooth groups, and shows no marked

preference for either the lingual or buccal

tooth surfaces.

The growth data offer an important al-

though small South American supplement

to the literature on that subject. They tenta-

tively suggest that although estimated adult

living stature is relatively short for prehis-

toric Ecuador, long bone diaphyseal growth

between birth and three years is close to

that reported for North American aborigi-

nal populations.

This study considerably augments our

information concerning the skeletal biol-

ogy of Guangala populations on the pre-

historic Ecuadorean coast. Collectively,

the data seem to document many of the

possible negative effects of changing sub-

istence and social organization. Frequen-

cies of dental caries, infectious disease, po-

rotic hyperostosis and infant mortality are

elevated in comparison with those of pre-

agricultural samples. The need persists

however, for additional data from other

prehistoric Ecuadorean sites, especially from

the highland and inland areas.

Acknowledgments

I thank Olaf Holm and Karen Stothert of

the Banco Central of Guayaquil for their

invitation to examine this material and for

their help in facilitating the study. The fam-

ilies of Jaime Andrade Heymann and Jaime

Andrade Moscoso contributed laboratory

and writing facilities, considerable project

support and immeasurable hospitality to

“la familia Ubelaker” during the two month

stay in Quito. I acknowledge financial sup-

port from the Banco Central of Guayaquil

and the Smithsonian Institution and ex-

press appreciation to the Smithsonian ad-

ministration superiors Fiske, Challinor,

Hughes and Ripley and Acting Chairmen

Van Beek, Trousdale and Viola for allow-

ing me the time to conduct this study away

from the Smithsonian.

References Cited

Berryman, H. E., D. W. Owsley and A. M. Henderson.

1979. Non-Carious Interproximal Grooves in

Arikara Indian Dentitions. American Journal of

Physical Anthropology, 50(2): 209-212.

Dobyns, H. F. 1966. Estimating aboriginal American

population: an appraisal of techniques with a new

hemispheric estimate. Current Anthropology, 7:

395-416.

Duckworth, W. L. H. 1951. Human Skulls from Ecua-

dor. In The Archaeology of the Santa Elena Penin-

sula in South-west Ecuador by G. H. S. Bushnell.

Cambridge, pp. 144-148.

Genovés, Santiago. 1967. Proportionality of the Long

Bones and their Relation to Stature among Meso-

Americans. American Journal of Physical Anthro-

pology, 26: 67-77.

Klepinger, Linda L. 1979. Paleodemography of the

Valdivia III Phase at Real Alto, Ecuador. American

Antiquity, 44(2): 305-309.

Merchant, Virginia L. and Douglas H. Ubelaker. 1977.

Skeletal Growth of the Protohistoric Arikara.

American Journal of Physical Anthropology, 46(1):

61-72.

Munizaga, Juan R. 1965. Skeletal Remains from sites

of Valdivia and Machalilla Phases. Appendix 2 in

Early Formative Period of Coastal Ecuador: The

Valdivia and Machalilla Phases by B. J. Meggers, C.

Evans, and E. Estrada. Smithsonian Contributions

to Anthropology, Vol. 1. Smithsonian Institution,

Washington, D.C.

. 1976. Intentional Cranial Deformation in the

Pre-Columbian Populations of Ecuador. American

Journal of Physical Anthropology, vol. 45, pp.

687-694.

Saville, M. H. 1913. Pre-Columbian Decoration of the

Teeth in Ecuador, with Some Occurrence of the

Custom in Other Parts of North and South Amer-

ica. American Anthropologist, Vol. 15, No. 3, pp.

377-394.

Trotter, M. and G. C. Gleser. 1958. A Re-evaluation of

Estimation of Stature Based on Measurements of

Stature Taken During Life and of Long Bones after

Death. American Journal of Physical Anthropology,

Vol. 16, pp. 79-123.

Turner, Christy G. II. 1978. Dental Caries and Early

Ecuadorian Agriculture. American Antiquity, Vol.

43, No. 4, pp. 694-697.

Ubelaker, D. H. 1978. Human Skeletal Remains, Exca-

vation, Analysis, Interpretation. Taraxacum, Wash-

ington, D.C.

. 1979. Skeletal Evidence for Kneeling in Pre-

historic Ecuador. American Journal of Physical An-

thropology, Vol. 51, No. 4, pp. 679-685.

—

te a

eee

—a Fe

—— a

ee ee ee ee ee

—oe

——

See etal Sry a ene

. 1980a. Human Skeletal Remains from Site

OGSE-80 A Preceramic Site on the Sta. Elena Pen-

insula, Coastal Ecuador. Journal of the Washington

Academy of Sciences, Vol. 70. No. 1. pp. 3-24.

. 1980b. Prehistoric Human Remains from the

Cotocollao Site. Pichincha Province. Ecuador.

Journal of the Washington Academy of Sciences, Vol.

70, No. 2, pp. 59-74.

. 1981. The Ayalan cemetery. A Late Integra-

tion Period Burial Site on the South Coast of Ecua-

Journal of the Washington Academy of Sciences.

Volume 73. Number |. Pages 27-42. March 1983.

dor. Smithsonian Contributions to Anthropology,

No. 29. Washington, D.C.

Ubelaker, D. H., T. W. Phenice and W. M. Bass. 1969.

Aruficial Interproximal Grooving of the Teeth in

American Indians. American Journal of Physical

Anthropology, Vol. 30, No. 1. pp. 141-150.

Van Bork-Feltkamp, A. J. 1965. Squelettes de Palmar.

Nederlands Museum Voor Anthropologie et du De-

partement D' Anthropologie, No. 68. L‘Institut Royal

des Regions Tropicales, Amsterdam.

Development of an Empirically Derived Taxonomy

of Organizational Systems'

R. W. Swezey

Science Applications, Inc., McLean, Virginia

Siegfried Streufert

Pennsylvania State University College of Medicine, Hershey, Pennsylvania

John Mietus

Baltimore Gas and Electric Company, Baltimore, Maryland

The development of organizational tax-

onomies has been of interest to psycholo-

gists for some time. Warriner (1980), for

"Support for the work reported in this paper was

provided by the U.S. Army Research Institute for the

Behavioral and Social Sciences under Contract No.

MDA904-79-C-0699. The authors wish to thank T. O.

Jacobs for his continuing support and encouragement

on this project. This paper is based upon a larger re-

port developed during the course of the project

(Baudhuin, Swezey, Foster, and Streufert, 1980). This

Paper was presented at the 20th Congress of the Inter-

national Association of Applied Psychology, Edin-

burgh, Scotland, 1982.

Requests for reprints should be sent to Robert W.

Swezey, Behavioral Sciences Research Center, Science

Applications, Inc., 1710 Goodridge Drive. McLean.

Virginia, 22102, USA.

example, has recently identified three major

approaches used to classify organizational

research: traditional or common sense

classification systems, theoretical or heu-

ristic systems, and so-called *‘empirical”’

taxonomies. According to Warriner, tradi-

tional classification systems have been most

popular, but are known to have basic lim-

itations, since such systems do not typically

define the content(s) of classes, do not

show relationships among various classes,

and do not provide reproducible catego-

ries. Such intuitive approaches to classifi-

cation are bound by the limitations and bi-

ases of those doing the classification.

Theoretical or heuristic systems, sim-

ilarly, are usually based upon a theory of

28 R. W. SWEZEY ET AL.

interest to the individual developing the

taxonomy. The theory essentially dictates

the approach to classification. Such classi-

fication systems are of use to those who

subscribe to the particular theory upon

which the classification is based; they may

however, be of less value to others. More-

over, theoretical taxonomies depend di-

rectly upon the adequacy of the theory. If

the theory is weak, then the classification

system itself will also be weak.

In “‘empirical”’ taxonomic approaches,

the intention is to describe the array of

units to be categorized relative to a large

number of variables, and then to sort them

in accordance with their affinity or similar-

ity across the variable set. The major lim-

itations of such procedures are related to

limits of representativeness reflected in the

cases used for analysis. Such limitations

can, however, be overcome as computer

technology and capability increase. Empir-

ical taxonomies are derived from data

rather than from theory.

Recently, social and behavioral scientists

have expressed interest in the development

of empirically derived classification sys-

tems. McKelvey (1975) has provided a re-

view of multivariate approaches to empiri-

cal taxonomy development in the disciplines

of social and organizational psychology,

along with a series of guidelines for their

development. The underlying objective of

such efforts, in McKelvey’s view, is par-

simony of classification. This objective

however, is often difficult to achieve, given

the potentially large number of important

organizational characteristics which have

been identified in existing studies. (See, for

example, Sells, 1964—500 variables; Haas,

Hall, and Johnson, 1966—210 variables;

and Pugh, Hickson, Hinings, and Turner,

1963—64 variables.) Given such large num-

bers of potential variables, multivariate

approaches appear to be an effective way of

reducing the variable set toa smaller, more

meaningful array of attributes.

From the empirical perspective, taxo-

nomic development in the social and or-

ganizational sciences should be designed to

reduce large, complex attribute popula-

tions into smaller, more homogeneous di-

mensions. This is the aim of such multivar-

late statistical techniques as factor analysis

and of the numerical taxonomy approaches

developed by Sokal and Sneath (1963). The

basic question underlying this viewpoint

involves the extent to which a complex set

of diverse organizational dimensions may

be described by fewer, more homogeneous

factors. Recent work by Carper and Snizek

(1980) has reviewed nearly 20 such classifi-

cation systems for the purpose of synthesiz-

ing these systems into effective taxonomic

categories. Carper and Snizek have pro-

posed an evaluation scheme for existing

taxonomy studies, which use criteria pre-

sented by Sokal and Sneath (1963), to de-

fine requirements for any practical taxo-

nomic system. That study also provides

guidelines for research efforts, and dis-

cusses the nature of various problems asso-

ciated with taxonomic development.

The development of empirical approaches

to organizational classification shares com-

mon problems with virtually all other so-

cial and behavioral science classification

approaches. As expressed recently by such

organizational taxonomists as Warriner

(1980), these concerns focus mainly upon

such traditional social science research

issues as: control, random sampling, repre-

sentativeness of variables, equivalence and

independence of variables, theoretical bi-

ases, objectivity, validity, and generaliza-

bility. Despite such concerns, the develop-

ment of cogent organizational taxonomies

remains a fundamental element in the evo-

lution of organizational theory.

In a recent review of the literature re-

garding empirical taxonomies of organiza-

tions (Baudhuin, Swezey, Foster, and

Streufert, 1980), only a few studies were

identified in which a taxonomy of organi-

zations or of organizational variables was

developed using empirical research methods.

A majority of these studies used some form

of cluster analysis and focused on classify-

ing organizations based upon their similar-

ities to a list of previously selected attri-

butes. None of the research reviewed in that

effort met the requirements recommended

ay es _— tn, eee ie

EMPIRICALLY DERIVED TAXONOMY 29

by McKelvey (1975) or those suggested by

Sokal and Sneath (1963). Baudhuin er a/.,

concluded that taxonomy development in

organizational theory is essentially a multi-

dimensional enterprise requiring multi-

variate classification approaches.

The present effort was intended to de-

velop an empirically generated technique

for classifying organizational variables.

More specifically, the intent was to organ-

ize the areas of organizational psychology

and general systems theory into a coherent

framework for use in a subsequent litera-

ture review (Swezey, Davis, Baudhuin,

Streufert, and Evans, 1980).

The empirical classification approach

was Selected for this effort for several rea-

sons. First, organizations are themselves

complex, multidimensional units. Such

complexity demands an approach which is

capable of dealing with this multidimen-

sionality. Second, since a large number of

variables could potentially contribute to

the various classes with the chosen ap-

proach, an empirical technique using multi-

variate logic is appropriate. Such a tech-

nique also has the capability to address

interactions among the organizational var-

iables of interest. Third, the empirical ap-

proach and associated multivariate tech-

niques are compatible with the general

Systems theory viewpoint. Multivariate

_ approaches, by nature, are themselves

adapted from a general systems perspective

(Sells, 1964).

This study departs dramatically from

others which have attempted to classify or-

ganizations, since the intent here was to

classify variables from the arenas of organi-

Zational psychology and general systems

theory that affect the functioning of organi-

Zations, rather than to classify organiza-

tion per se. In support of this objective, the

present approach involved examining the

literatures on organizational psychology

and general systems theory to determine

the extent to which variables in these litera-

tures might be identified and organized

into a generic classification system. Once

Such a system was developed, it was em-

ployed as the basis for structuring a formal

review of those literatures (Swezey et al.,

1980).

Method

Selection of a generic set of organiza-

tional and systems theoretic attributes con-

stituted a major task in the taxonomy de-

velopment process. The objective of this

activity was to identify organizational psy-

chology or general systems theory varia-

bles, attributes, or characteristics which

could reasonably be expected to exert an

influence on, or contribute to, organiza-

tional behavior and performance. Various

areas were, however, excluded from con-

sideration. Excluded were: (1) methodo-

logical or analytical techniques, (such as

gaming/simulation, modeling, sampling,

surveys, testing instruments, training aids,

etc.), (2) specific theoretical modes of in-

quiry (such as field theory, contingency

theory, path-goal theory, etc.), and (3)

““meta-organizational,”’ or global, constructs

(such as administration, bureaucracy, so-

ciety, politics/economics, etc.).

The criteria established for inclusion or

exclusion of attributes were as follows:

@ Attributes selected should constitute,

as nearly as possible, a comprehensive rep-

resentation of all recognized facets of or-

ganizational behavior and performance.

On the other hand, where multiple terms

have accepted meanings that are virtually

synonymous, parsimony should be the over-

riding consideration.

@ Attributes selected must be an integral

component of, coincident with, or readily

adaptable to, accepted systems-theoretic

concepts.

@ Attributes selected should focus on

systemic/subsystemic attributes rather than

onessentially individual psychological states

or manifestations (e.g., alienation, attitudes,

bias, cognition, emotion, morale, motiva-

tion, etc.).

Five steps were followed in establishing

the list of attributes. Step | involved a re-

view of the indices and topical headings of

30 R. W. SWEZEY ET AL.

over 20 major texts in the field of manage-

ment, organization theory, and organiza-

tional psychology. From this, a prelimi-

nary listing of 350 attributes was formulated.

Four project team members then served as

the selection panel. Consensus was required

among all team members in order for any

attribute to be added to, deleted from, or

combined within, the master list. Step 2 in-

volved a review of these attributes to add,

delete, or combine terms as necessary. Step

3 involved eliminating attributes that ad-

dressed individual psychological states or

manifestations and thus were considered

inappropriate. Step 4, undertaken in the in-

terest of parsimony, involved combining

duplicative terms, 1.e., those which were

roughly, though not necessarily precisely,

synonymous (for example, development-

dynamic equilibrium-morphogenesis; inter-

action-cooperation-coordination, etc.). Step

5, then, involved a review, and consensual

agreement upon, the list of attributes. A

final list of 84 attributes was thus derived

from the original list of 350 attributes.

Selection of Articles

A group of over 1000 articles was se-

lected from the literature in five areas

which were generally defined as being rele-

vant to the overall research task. These

areas were: organizational psychology and

behavior, general systems theory, organi-

zational effectiveness and development,

simulation, and training.

Three criteria were developed for the se-

lection and retrieval of articles: currency,

completeness, and representativeness. In

order to meet each of these criteria in a

manageable way, certain trade-offs had to

be made. On the issue of currency, it was

determined that the articles should reflect

current state-of-the-art research and the-

ory on organizational behavior. Greatest

emphasis was placed, therefore, on select-

ing documents from the 1970-1980 time-

frame. Less, but substantial, emphasis was

then placed on obtaining literature cover-

ing the 1960-1969 period. A careful study

was made of the reference sections of major

articles written during this 20-year span;

any works consistently referenced were

identified and retrieved.

The criterion of completeness was ad-

dressed on two fronts:

(1) Sources. Over 100 primary sources

were surveyed, including refereed journals,

professional society journals and proceed-

ings, university publications, theses and

dissertations, books, and technical reports

of federally and privately funded research.

Special attention was paid to selecting the

most relevant works of key researchers/

theorists in each area.

(2) Breadth of coverage. Using the final

attribute list, over 80 aspects of organiza-

tional behavior, general systems theory,

organizational effectiveness and develop-

ment, simulation, and training were inves-

tigated; as well as the full spectrum of sys-

tems concepts related to human organizing.

The final criterion, representativeness,

was used as the basis for any required trade-

offs between completeness and manage-

ability/feasibility. While the data base is

not exhaustive of literature pertaining to

each constituent attribute, it is believed to

be representative of current thinking on the

problem.

In addition to being classified for taxo-

nomic development purposes, the litera-

ture in the data base was also classified by

document type according to a system mod-

ified from an approach developed pre-

viously for classifying instructional devel-

opment literature (Schumacher, Swezey,

Pearlstein, and Valverde, 1974). This sys-

tem involved classification into seventeen

categories of interest to the project, such as:

opinion articles, theoretical discussions,

methodological development articles, eval-

uative summaries, etc. Interested readers

should refer to Baudhuin et al. (1980) fora

complete description of this system.

Checklist Development

A checklist format was then developed

to aid reviewers in establishing the extent to

which each reviewed document addressed

EMPIRICALLY DERIVED TAXONOMY 31

the 84 attributes. The checklist was de-

signed to provide a basis for determining

the extent to which an author treated each

of the attributes, according to a four-point

rating scale as follows: 0 = the attribute

was not mentioned by the author(s). | = the

attribute was minimally mentioned; i.e., the

attribute may have been mentioned as one

of many in a literature review, for example,

but was not the major thrust of the discus-

sion. 2 = the attribute was discussed; i.e.,

one of several (five to ten) topics treated.

3 = the attribute was emphasized, i.e., one

of the major topics (one to four) discussed.

Table 1 shows the checklist format.

Pilot Studies

The effort included two pilot studies: one

associated with the evaluation instrument,

and a second associated with the factor

analysis procedure selected for use in con-

structing the taxonomy. The first pilot

study reflected a concern for the inter-rater

reliability of those doing the evaluation of

the articles. For this purpose, six raters

were employed to evaluate the articles on

the extent to which they addressed the 84

defined attributes. Since it was expected

that it would be extremely time consuming

for all raters to rate all articles in the total

data base, it was hoped that each article

could be evaluated by a single rater.

It was therefore, considered necessary to

estimate the extent to which general sys-

tems theoretic terms and/or organizational

attributes were assigned to articles in a con-

sistent fashion by the six raters. For this

purpose, a pilot inter-rater reliability study

was conducted. Five articles were selected

at random: one from each of the five cate-

gories previously listed. These articles were

Georgopoulos and Tannenbaum (1957)—

organizational effectiveness; Lucas (1979)—

simulation; Kast and Rosenzweig (1972)—

general systems theory; O’Reilly and Roberts

(1974)—organizational psychology; and

DeCotiis and Morano (1977)—training.

Five inter-rater reliability coefficients

were then computed following techniques

described in Winer (1971, pp. 283-289).

This method of computing inter-rater reli-

ability involves use of an analysis of var-

lance model to estimate reliability of as-

signment of the taxonomic categories by

the raters. Five ANOVA’s were computed:

one for each article, across the ratings as-

signed by the six raters. (All six raters rated

all five articles.)

Table 2 shows the analysis of variance re-

sults and the resulting inter-rater reliability

coefficients (r¢) for the five articles. As can

be seen, inter-rater reliability was very high

across the six judges, ranging from a low of

.85 to a high of .94. Such a level of reliabil-

ity indicates a high degree of consistency

across judges in their assessment of the ex-

tent to which the same topic areas were in-

cluded in each of the five randomly selected

sample articles upon which the pilot study

was conducted.

The second pilot study was conducted to

determine the utility of a factor analysis

model for use as the basic statistical ap-

proach to the empirical taxonomic devel-

opment activity. A factor analysis model

was Selected as the statistical technique for

empirical taxonomic development for two

basic reasons. First, the nature of the prob-

lem itself suggested this approach. The

overall objective of the effort was to pro-

vide an ordered set of organizational at-

tributes grounded in a system theoretic

perspective, in order to provide a frame-

work for a subsequent literature review.

With the potentially large number of at-

tributes which could be considered relevant

to the broadly defined field of organiza-

tional psychology and general systems the-

ory (we initially identified nearly 400 such

attributes), a multivariate technique which

systematically reviews the correlations or

affinity clusters among the variables and

reduces the set to a more organized and

manageable group of factors was needed.

Numerous previous research efforts have

suggested multivariate techniques as a way

of achieving parsimony in developing tax-

onomies of organizational variables (c.f.

McKelvey, 1975). Factor analysis, as has

been applied to organizational taxonomy

studies for example, by Pugh, Hickson,

R. W. SWEZEY ET AL.

Table 1.—Document Checklist Format.

Instructions: The checklist is designed to provide a basis for determining the extent to which an author

treats each of the attributes listed. Check the appropriate box for each attribute as follows:

O0—The attribute was not mentioned by the author(s).

1—The attribute was minimally mentioned, i.e., the attribute may have been mentioned as one of many

in a literature review, for example, but was not the major thrust of the discussion.

2—The attribute was discussed, 1.e., one of several (5-10) topics treated.

3—The attribute was emphasized, i.e., one of the major (1-4) topics discussed.

. Absenteeism

. Adaptability (adaptation, coping, flexibility)

Authority

Boundary

. Capability (capacity, potential)

. Centralization

. Certainty

. Change (innovation)

. Change Agent

10. Channel (network)

11. Climate (organizational climate/health/pathology/personality)

12. Closed System

13. Communications (bargaining, information exchange)

14. Communication Barriers/Filters

15. Competence

16. Complexity (dimensionality, dimensions)

17. Conflict (role conflict, value conflict, competition, confrontation)

18. Conflict Regulation

19. Consensus (agreement)

. Control (accountability, compliance, conformity, correction, maintenance, regulation)

. Creativity

. Decentralization

. Decision-Making (choice, problem-solving)

. Development (dynamic equilibrium, evolution, heterostasis, morphogenesis)

. Differentiation (compartmentation, division of labor, elaboration, specialization)

. Direction (intentionality)

. Disorganization (disorder, entropy)

. Efficiency

. Environment (situation)

. Equifinality

. Equilibrium (balance, homeostasis, morphostasis, stability, steady state)

. Feedback

. Goals (objectives, requirements)

. Goal Attainment (performance, productivity)

. Goal Displacement

. Goal Setting (expectancy, expectations)

. Goal Succession (ideal seeking)

. Group Dynamics

. Growth

. Hierarchy

. Incentive (reinforcement, reward)

. Independence (autonomy, totipotentiality)

. Influence

. Information (experience, knowledge, learning, variety)

. Initiative (proaction)

. Input (contribution, resources)

. Integration (symbiosis)

. Interaction (cooperation, coordination, human relations, participation)

. Interdependence (partipotentiality)

. Intervention (third party intervention)

. Job (function)

WOonan & WN

ya a es em fc |) ff fp a | ae yf a a ff an ee sem | afte} een

alalofe ae lalalelala af sae) s) Safeo sie) a ee ee) eo aoe ee epee ease ee eae ye yea eps

AnNBRRRHR HHP HR HAW WWW WWW WWWNNNNNNNN NY WV

HPD BABE EOD oDEB BRB nD BADE BRa PB ee eee SE SE se SE eis Bei Sel

apoleae ee lejele|e)s sles e) S|) Soe (ea) a) eee) ee) eo) see) aie) eee ele oo eae eyenoiasior—

KH COW NAM RWNHK TOA AIAMARWNK TUOWAIHDUNHWNK OS

es ee eee eee

——= >=

EMPIRICALLY DERIVED TAXONOMY 33

Table 1.

(continued)

. Job Enrichment/Enlargement

. Job/Task Analysis

. Management

Open System (permeability)

. Optimization

. Output (product)

. Power (coercion, dominance)

. Resource Allocation/ Distribution

. Response (reaction)

. Responsibility

. Rigidity (change resistance)

. Role (relationship)

. Simplicity (routinization)

Size

Task

. Technology (automation)

. Training

. Turnover

. Uncertainty (risk)

. Values

. Leadership

. Effectiveness

DOOR OCOODDRBOODBOoODUR Onna eel awae

i a at i a i a a a a a fa ak), tS ee

me he i af i a a a | ee

OOODOOODODODDOODODODDOOODOODODDOD OBO ateaoe

SCRRSSPAA RAS SP ISSA RA Pes e seat aneas

Hinings and Turner (1968), was determined

to be sufficiently flexible for this purpose,

and provided the additional advantage of

allowing for analysis of large quantities of

data; whereas cluster analysis, also used in

several previous studies (c.f. Pugh, Hick-

son, and Hinings, 1969) was viewed as a

technique which is very similar to factor

analysis, but having somewhat less statisti-

cal elegance (Lawlis and Chatfield, 1974).

Numerical taxonomy, a technique devel-

oped and used frequently for biological

and zoological taxonomies, was also con-

sidered (Sokal and Sneath, 1963). One type

of input often used in numerical taxonomy

programs is data from previously com-

puted factor matrices and correlation ma-

trices. Therefore, with a view toward addi-

tional analysis, if needed, factor analysis

. Performance Evaluation/Appraisal

. Plan/Planning (strategy/strategize)

. Maturity (maturation, organizational life)

. Organization (cohesion, negative entropy, order)

. Process (conversion, implementation, throughput, transformation)

. Sensing (cognition, forecasting, intelligence, scanning)

. Standards (critical variables, norms, regulations, rules)

. Structure (design, form formalization)

. Suboptimization (equity, satisficing)

. Subsystem (component, group, team)

. Synergism (gestalt, holism, organicism)

was selected as the technique of choice;

however, should this analysis not have

provided well defined and meaningful fac-

tors, it would have been possible to subject

a smaller matrix (approximately 40 X 40),

as derived from the factor analysis, to the

numerical taxonomy technique.

Second, the nature of the data to be ana-

lyzed appeared to be compatible with fac-

tor analytic procedures. Since the checklist

ratings were conceptualized as interval

scale measures, rather than as non-metric

yes/no type responses, factor analysis was

appropriate. Nonmetric cluster analysis

techniques are more appropriate for scal-

ing discrete variables (Lawlis and Chat-

field, 1974).

A pilot factor analysis was thus con-

ducted using the checklist evaluations for

R. W. SWEZEY ET AL.

1898 Svr6 7968" CS16 £688" 94

C660 88S SS E8C 8h 88S Gd ESC. “OECD 88S LEO P9ZE" BBS =§=£G6 16) = EPIC. 88S S97 OCP [BIOL

Tal Cs Och 6768 21 Och = bh HS 89 680€° Och =v 6CI OPT OcP cS 19 Ect. O¢r EL Sel [enpIsoy

CVG 2 § ToS? ~ Os’ S 96L'7 Coe a oe 9L'6 £9" S 7) Gt Leena <2 Dince sodpnf usamjog

8L9C vOS OO9EI PI vos veil 87 vOS Sori e€8¢t bos L9'v9 ce vOS $891 Sa10da}vd UY AA

Flos) vp SS S51 ESC —-p8 8P°CIT OL7 8 L8°9C7 SIS T= <8 9¢ LCI 667 8 SOLISZ Sal4odayeo usamiag

SIN JP SS SI JP SS SIN JP SS SW JP SS SIN JP SS soinos

g v t (6 I

apniy

*SJUIIIIJIOD ARYIQeIjay 19j81-19}UT SISAJBUY JO[IG—"Z IIqU]

~ greater than one. .

EMPIRICALLY DERIVED TAXONOMY 35

239 initially selected articles (as well as

books and book chapters). (Slightly over

40 articles were randomly selected from

each of the original five categories of

articles—organizational psychology and

behavior, general systems theory, organi-

zational effectiveness and development,

simulation, and training.) The BMDP-79

(1979) P4M factor analysis program was

used for analysis of the data. A principal

components method was used for the in-

itial factor extraction, followed by an or-

thogonal rotation to simple structure with

the varimax criterion. A 30-factor extrac-

tion was specified; in addition to the preas-

signed criterion for the number of factors

being those factors with eigenvalues greater

than unity.*

Although this termination point is an

arbitrary decision, it appeared that, in this

case, meaningfulness, rather than mathe-

matical elegance, should be the fundamen-

tal concern. Support for this criterion is

found in Harman’s (1967) citation of Kais-

ers recommendation that “‘after consider-

ing statistical significance, algebraically

necessary conditions, psychometric relia-

bility, and psychological meaningfulness

. the number of common factors should

be equal to the number of eigenvalues

.” (p. 198). Rummel

(1970) suggests that the generalizability of

a factor decreases as the eigenvalue falls

below unity and the variance accounted for

is small in comparison to other factors in

the matrix (pp. 344-364).

The results of the pilot factor analysis are

presented in Table 3. The criterion used for

selecting variables for each factor was a

factor loading of .40 or greater on one fac---

tor and no loadings greater than .30 on any

other factor. The results depicted in Table 3

show only those variables which were sorted

by the rotated factor matrix, in accordance

with this .40/.30 criterion. One result of

this pilot run was the finding that amount

of work space designated in control cards

* For this analysis, the raw data from the checklist

were changed from the 0, 1, 2,3 coding scheme, shown

in Table 1, to a 1, 2, 3, 4 code.

for the computer run was not sufficient to

reach the specified 30 factor criterion.

Hence, the factor extraction terminated

after 17 factors. However, Table 3 shows

that the eigenvalues begin to drop off be-

tween Factor VII and Factor VIII, and

level after Factor VIII.

Figure | depicts this factor variance

leveling in Scree test format (Cattell, 1966).

This pilot result suggested that a more

meaningful factor structure might be ob-

tained by computing analyses with maxi-

mum factor numbers, specified at 7, 6, and

5. Such analyses could illustrate where var-

iables moved and/or clustered in each suc-

cessive run. For the pilot factor analysis, no

attempt was made to name factors.

Results

A final factor analysis was conducted

using a randomly selected sample of 210

references. Based on the results of the pilot

study, factor extractions were specified at

7,6, and 5 factors. Table 4 lists the variable

loadings under each factor for each desig-

nated run. Only those variables with load-

ings of .40 or higher were listed under their

respective factors. Based ona review of the

factor structure through each of the desig-

nated rotations, it was determined that the

six-factor solution appeared to be most

meaningful for purposes of this study.

Since, in the final factor analysis, the seven

factor rotation resulted in one factor load-

ing group which barely exceeded the .40

criterion; the six factor solution, which re-

sulted in acceptably high loadings on all

factors, was selected. As is evident from a

comparison of Tables 3 and 4, the final fac-

tor solution was very similar to that ob-

tained in the pilot analysis.

Table 5 depicts the variable loading ma-

trix for the six factor solution along with

the eigenvalues and cumulative percent of

factor variance accounted for by each fac-

tor. Consideration of the variables loading

under each factor in the solution lead to the

following conclusions regarding taxonomic

categories.

Factor

II]

VII

XII

XIII

XIV

R. W. SWEZEY ET AL.

Table 3.—Pilot Analysis Sorted Factor Variable Matrix.

Loading/ Variable

.79 Integration

.74 Complexity

.74 Differentiation

.66 Information

.60 Decision Making

.57 Input

.57 Sensing

.55 Simplicity

.55 Environment

.71 Power

.65 Influence

.47 Authority

.72 Certainty

.72 Change

.65 Decentralization

.52 Size

.76 Closed System

.69 Open System

.69 Equifinality

.49 Subsystem

.64 Job/Task Analysis

.60 Job

.52 Task

.64 Communications

.62 Communications Barriers

.55 Direction

.71 Training

.63 Management

.54 Effectiveness

.67 Goal Attainment

.46 Input

.58 Process

.55 Output

.75 Growth

.75 Consensus

.45 Values

.70 Absenteeism

.68 Incentive

.54 Intervention

.65 Development

.63 Organization

.54 Disorganization

.69 Optimization

.51 Suboptimization

.47 Adaptability

.59 Role

.45 Interdependence

.48 Conflict

45 Conflict Reg.

.62 Responsibility

.51 Maturity

Eigenvalue

6.044

2.905

2.796

2.730

2.695

2.614

2.592

2.192

2.177

2.167

2.141

2.140

DAB?

2.053

2.019

Cumulative &

Factor Variance

14.2

25.9

32°

38.4

60.9

66.0

76.1

85.9

90.6

EMPIRICALLY DERIVED TAXONOMY

Table 3.—(continued)

Factor Loading/ Variable

Cumulative &

Eigenvalue Factor Variance

XVI .66 Goals

.61 Goal Succession

.43 Goal Setting

.69 Rigidity

.63 Change

.60 Change Agent

.44 Standards

XVII

@ Factor I was termed Multidimensional

Information Processing and accounted for

22.88% of the factor variance. The varia-

bles which load on this factor reflect botha

process systems model of organizations

and/or the individual/group/organization

processes associated with acquiring infor-

mation, processing information, and dis-

seminating that information (including de-

cision making) as components in complex

multidimensional environments. They also

address the structure of how information is

processed in organizations.

@ Factor II was called Organizational

Systems Dynamics and accounted for 17.0%

of the factor variance. The variables which

load on this factor represent the character-

7.0

6.0

5.0

4.0

3.0

Eigenvalues

1.0

a SY a - - S

2.007

1.975 100.0

istics of an organizational system relative

to its adaptation and flexibility as it copes

with its environment, attempts to maintain

a relatively steady state or balance, and

utilizes its resources to grow in more or less

planned ways.

@ Factor III was called Organizational

Change Technologies and accounted for

16.13% of the factor variance. The varia-

bles loading on this factor focus on those

techniques normally associated with the

organizational development/ organizational

effectiveness domain and reflect concerns

for individual growth and development in

organizations, personnel interface with jobs,

the organization, and the work process.

This factor identifies human resource tech-

os 16 17

1 de: tar WA

Factors

Fig. 1. Pilot Analysis Factor Variance Leveling.

Factor

II]

Vil

* These variables were not unidimensional. They are included in the matrix to illustrate relationships among

factors and variables; i.e., the .40/.30 criteria was not met by these variables indicating a factorially complex

variable. It should be emphasized that the number of factorially complex variables is very small considering the

R. W. SWEZEY ET AL.

Table 4.—Sorted Factor Variable Matrix.

Varimax Rotations to Seven, Six and Five Factor Solutions (Loadings/Variables)

.69 Integration

.69 Input

.66 Complexity

.65 Information

.62 Output

.58 Differentiation

.56 Sensing

.56 Decision Making

.52 Environment

.58 Subsystem

.54 Direction

.46 Adaptability

.42 Maturity

.44 Open System

.46 Rigidity

.42 Certainty

.64 Change Agent

.48 Intervention

.47 Job Enrichment

.44 Organization

.42 Process

.42 Training

.60 Influence

.50 Power

.40 Responsibility

.48 Hierarchy

.66 Independence

.66 Centralization

.65 Size

.54 Decentralization

.49 Interdependence

.58 Goal Setting

.57 Goal Attainment

.56 Goals

.53 Goal Succession

.46 Goal Displacement

.42 Consensus

.41 Optimization

.40 Resource Allocation

.71 Input

.69 Integration

.66 Complexity

.65 Output

.62 Information

.58 Differentiation

.56 Sensing

.55 Decision Making

.54 Environment

.55 Subsystem

.54 Equilibrium

.51 Open System

.49 Direction

.44 Growth

.47 Adaptability

.46 Closed System

.42 Rigidity

.66 Change Agent

.53 Feedback

.49 Intervention

.47 Job Enrichment

.45 Organization

.42 Process

.41 Training

.57 Influence

.52 Power

.42 Conflict

.48 Hierarchy*

.41 Interaction

.47 Authority*

.41 Role

.67 Independence

.64 Centralization

.64 Size

.55 Decentralization

.51 Interdependence

.45 Authority*

.61 Goal Setting

.56 Goals

.52 Goal Succession

.52 Goal Attainment

.44 Goal Displacement

.71 Integration

.70 Input

.66 Complexity

.63 Output

.61 Information

.58 Differentiation

.57 Decision Making

.56 Sensing

.54 Environment

.58 Authority*

.57 Change Agent

.52 Power

.45 Job Enrichment

.45 Management

.44 Competence

.46 Responsibility

.50 Role

.56 Equilibrium

.55 Subsystem

.52 Open System

.47 Direction

.42 Growth

.45 Adaptability*

.46 Closed System

.42 Authority*

.63 Centralization

.63 Independence

59 Size

.57 Decentralization

.42 Hierarchy

.48 Interdependence

.53 Absenteeism

.51 Goal Setting

.48 Goals

.46 Goal Attainment

.46 Goal Succession

.46 Turnover

size of the variable matrix and the complexity of the taxonomic research problem.

EMPIRICALLY DERIVED TAXONOMY

Table 5.—Six Factor Solution Variable Loading Matrix.

Factor

Loading/ Variable

Eigenvalue

Cumulative %

Factor Variance

Ill

.71 Input

.69 Integration

.66 Complexity

.65 Output

.62 Information

.58 Differentiation

.56 Sensing

.55 Decision Making

.54 Environment

.55 Subsystem

.54 Equilibrium

.51 Open System

.49 Direction

.44 Growth

.47 Adaptability

.46 Closed System

.42 Rigidity

.66 Change Agent

.53 Feedback

.49 Intervention

.47 Job Enrichment/Enlargement

.45 Organization

.42 Process

.41 Training

.57 Influence

.52 Power

.42 Conflict

.48 Hierarchy*

.41 Interaction

.57 Authority*

.41 Role

.67 Independence

.61 Centralization

.64 Size

.55 Decentralization

.51 Interdependence

.45 Authority*

.61 Goal Setting

.56 Goals

.52 Goal Succession

.52 Goal Attainment

.44 Goal Displacement

5.193

3.867

3.662

3.503

3.366

3.108

22.88

39.91

56.04

71.48

86.31

100.00

* These variables were not unidimensional. They are included in the matrix to illustrate relationships among

factors and variables: i.e.. the .40/.30 criteria was not met by these variables indicating a factorially complex

variable. It should be emphasized that the number of factorially complex variables is very small considering the

size of the variable matrix and the complexity of the taxonomic research problem.

40 R. W. SWEZEY ET AL.

nologies associated with enhancing indi-

viduals and work group perceptions regard-

ing job development and/or modification.

@ Factor IV was called Management

Authority/ Compliance Characteristics and

accounted for 15.44% of the factor var-

iance. The variables loading on this factor

are associated with the dimensions of influ-

ence and power as components in the su-

perior/subordinate organizational scheme

where compliance is required, for example,

from subordinates relative to their position

or level in the scaler chain. The variables

reflect status or hierarchical leveling attri-

butes found in more organizations nor-

mally associated with management control

procedures.

@ Factor V was called Organization

Coordination and Control and accounted

for 14.83% of the factor variance. The vari-

ables which loaded on this factor reflect

characteristics of organizations associated

with structure and those concerns leading

to the coordination and/or control of the

Organizational systems, subsystems and

subsidiaries. Because of the ‘“‘authority”

variable loading on this dimension as well

as under Factor IV one might speculate

that a relationship exists between the two

factors. The Management Authority factor

(Factor IV) may well describe the individ-

ual control dimension in organizations,

l.e., the manager influencing and control-

ling his subordinates, while the Organiza-

tional Coordinator and Control factor (Fac-

tor V) may describe those structural

organizational features related to coordi-

nation and control at the organization-

wide level.

@ Factor VI was called Goal Orientation

and accounted for 13.69% of the factor var-

lance. Variables loading under this factor

reflect those activities that organizations

and individuals engage in to determine de-

sired states that the organizational system

and its personnel are attempting to achieve

through planning, organizing, and control-

ling. Most organizations, by definition, are

goal directed and the variables loaded

under this factor focus on the range of goal

activities required by an organizational

system to determine priorities, to achieve

objectives, and to modify or replace those

objectives no longer important to the system.

Discussion

Factor analysis identified six distinct and

relatively stable factors influencing organi-

zational systems. With minor exceptions,

the composition of each factor remained

consistent over several designated rotations

to different numbers of factor extractions.

The first factor, multidimensional infor-

mation processing, was the most reliable

and consistent factor. This factor appears

in both the pilot and the final analyses, with

only minor changes in factorial composi-

tion. This finding suggests that substantial

consistency exists among the reviewed arti-

cles regarding the importance of the varia-

bles within this factor in organizational

theory. The analysis demonstrates how

these variables group together in defining

this factor. Factor I reflects organizational

(as well as individual) multidimensional

processes, where the dimensionality of in-

puts or information is sensed, differen-

tiated, and integrated through the structure

of the organization. Multidimensionality

(complexity) of decision making and or-

ganizational output is achieved via the

Same integrative and differentiative proc-

esses. Factor I is a process factor, applica-

ble to various levels and entities within an

organization (e.g., persons and their in-

formation processing, organizations and

their processing, the structure of the-organ-

izations themselves, etc.).

Since one of the objectives of this project

was to address systems theoretic issues, the

original list of terms included a sizable

number of such variables. It is not surpris-

ing, therefore, that an organizational sys-

tem dynamics factor appeared in the factor

solution. This factor also exhibited relative

stability and consistency in its factorial

composition. However, a change in posi-

tion from Factor II to Factor III was noted

between the six- and five-factor solutions.

An interesting result relative to this dimen-

— ~~ wit.

EMPIRICALLY DERIVED TAXONOMY 41

sion was the composition of this factor.

There appears to be an aspect of this factor

which combines organizational theory con-

structs with general systems theory con-

cepts. The composition of this factor, thus,

addresses those dimensions of organiza-

tions which allow them to adapt to varying

environmental conditions.

Another major aspect of the study in-

volved concern for the methods and/or

techniques associated with organizational

effectiveness. The third factor, organiza-

tional change technologies, is an outcome

of this frame of reference. The composition

of Factor III also shows relationship to the

systems theoretic framework, in its inclu-

sion of “‘organization” and “‘process”’ vari-

ables. Change technologies are thus viewed

as Ongoing, dynamic activities. This factor

remained stable through the six-factor so-

lution, where it reversed position with the

organizational system dynamic factor. One

might postulate an interaction between

these two factors based on this reversal in

the factor pattern.

The fourth and fifth factors, manage-

ment authority/compliance characteristics,

and organizational coordination and con-

trol, reflect the traditional organizational

variables found in the literature which

served as the input data to the factor analy-

sis. The final factor, goal orientation, shows

one of the more reliable structures in the so-

lution. All variables which loaded under

this dimension grouped together consist-

ently across solutions. The composition of

this factor seems well defined, highly rele-

vant to its name, and directly related toa

major dimension of both systems theory

and organizational psychology.

The primary rationale for selecting the

six-factor solution as the final solution re-

sided in the breakdown of factor composi-

tion in the five-factor solution for factors

IV and V. Further, the proportion of var-

lance accounted for in the six-factor solu-

tion did not change dramatically from the

Other solutions. It should be emphasized

that in this study, the selection of terms (at-

tributes) was based upon their use in the

literature by organizational psychology and

general systems theory authors. The result-

ing analyses, therefore, reflect current, not

potential thinking in these fields. The in-

tent here was to develop a technique for or-

ganizing existing concepts in these areas in

(hopefully) more meaningful ways.

References

Baudhuin, E. S., Swezey, R. W., Foster, G. and Streu-

fert, S. 1980. An empirically derived taxonomy of or-

ganizational systems. McLean, VA, Science Appli-

cations, Inc. SAI-80-091-178.

Carper, W. B. and Snizek, W. E. 1980. The nature and

types of organizational taxonomies: An overview.

Academy of Management Review, 5: 65-75.

Cattell, R. (Ed.). 1966. Handbook of multivariate

experimental psychology. Chicago: Rand McNally.

DeCotiis, T. A. and Morano, R. A. 1977. Applying job

analysis to training. Training and Development

Journal, 31: 20-24.

Georgopoulos, B. S. and Tannenbaum, A. S. 1957. A

study of organizational effectiveness. American So-

ciological Review, 22: 534-540.

Haas, J. E., Hall, R. H. and Johnson, N. J. 1966. To-

ward an empirically derived taxonomy of organiza-

tions. In R. V. Bowers (Ed.), Studies on behavior in

organizations: A research symposium. Athens, GA:

University of Georgia Press.

Harman, H. H. 1967. Modern factor analysis (2nd

Ed.). Chicago: University of Chicago Press.

Kast, F. E. and Rosenzweig, J. E. 1972. General sys-

tems theory: Applications for organization and

management. Academy of Management Journal, 15:

447-463.

Lawlis, G. F. and Chatfield, D. 1974. Multivariate ap-

proaches for the behavioral sciences: A brief text.

Texas Tech Press, Lubbock, Texas.

Lucas, J. C., Jr. 1979. Performance in a complex man-

agement game. Simulation and Games, 10: 61-74.

McKelvey, W. 1975. Guidelines for the empirical clas-

sification of organizations. Administrative Science

Quarterly, 20: 509-525.

O’Reilly, C. A., II] and Roberts, K. H. 1974. Informa-

tion filtration in organizations: Three experiments.

Organizational Behavior and Human Performance,

11: 253-265.

Pugh, D. S., Hickson, D. J., Hinings, C. R., Mac-

Donald, K. N., Turner, C. and Lupton, T. 1963. A

conceptual scheme for organizational analysis.

Administrative Science Quarterly, 8: 289-315.

Pugh, D. S., Hickson, D. J. and Hinings, C. R. 1969. An

empirical taxonomy of structures of work organiza-

tions. Administrative Science Quarterly, 14: 115-126.

Pugh, D. S., Hickson, D. J., Hinings, C. R. and Turner,

C. 1968. Dimensions of organizational structure.

Administrative Science Quarterly, 13: 65-105.

Rummel, R. J. 1970. Applied factor analysis. Evanston,

IL: Northwestern University Press.

Sells, S. B. 1964. Toward a taxonomy of organiza-

42 R. W. SWEZEY ET AL.

tions. In Cooper, W. W., Leavitt, H. J., and Shelly,

M. W. (Eds.), New perspectives in organization re-

search. New York: Wiley and Sons.

Shumacher, S. P., Swezey, R. W., Pearlstein, R. B. and

Valverde, H. H. 1974. Guidelines for abstracting

technical literature on instructional system devel-

opment. JASA catalog of selected documents in psy-

chology, 4: 113-114. MS 751.

Sokal, R. R. and Sneath, P. H. 1963. Principles of nu-

merical taxonomy. San Francisco: Freeman.

Swezey, R. W., Davis, E. G., Baudhuin, E. S., Streu-

fert, S. and Evans, R. A. 1980. Organizational and

systems theories: An integrated review. McLean, VA:

Science Applications, Inc.

Warriner, C. K. 1980. Organizational types: Notes on

the “‘organizational species” concept. University of

Kansas working papers in sociology. Lawrence,

Kansas: Charles K. Warriner.

Winer, B. J. 1971. Statistical principles in experimental

design (2nd Ed.). New York: McGraw Hill.

Journal of the Washington Academy of Sciences.

Volume 73. Number |. Pages 43-44. March 1983

Instructions to Contributors

Type manuscripts on one side of white

bond paper. Double space all lines, includ-

ing those in abstracts, tables, legends, quoted

matter, acknowledgments, and references

cited. Number all pages consecutively.

Page | should contain the title (not to ex-

ceed 100 characters), author’s name and af-

filiation, a running title (not to exceed 70

characters) and an indication to whom cor-

respondence is to be sent. In research pa-

pers concerning biological subjects, include

an indication of the order and family of the

taxa discussed.

Page 2 should contain an abstract which

should be intelligible without reference to

the text of the paper. Write an informative

digest of the significant content and con-

clusions, not a mere description. Gener-

ally, the abstract should not exceed 3% of

the text.

Footnotes should be used sparingly. On

each page use the symbols which follow to

indicate the footnotes for that page. The

order of use should follow the order in

which the symbols are listed herein. The

Same symbols may be used on separate

Pages but may not be re-used on the same

page. The footnotes must be typed on a

separate page. Please be sure to indicate

both the manuscript page number and the

symbol. The symbols are: *, f.

The quality of all original illustrations

must be high enough to facilitate good

offset reproduction. They should have ample

margins and be drawn on heavy stock or

fastened to stiff cardboard to prevent bend-

ing. They should be proportioned to col-

umn (1 X 3) or page (2 X 3) type-dimen-

sions. Photographs should have a glossy

finish. They reproduce best when the con-

trast is fairly high. Identify each illustration

with number and author in light pencil

marks on the unused lower or side margins.

Submit all illustrations separately—please

do not glue or clip them to the pages of the

manuscript.

Do not type or write legends directly on

the illustrations. Type legends on a sepa-

rate page or pages at the end of the

manuscript.

Tables should be included only when the

same information cannot be presented eco-

nomically in the text, or when a table pre-

sents the data in a more meaningful way.

Tables should be double spaced through-

out and contain no vertical lines. The table

should be organized from top down as fol-

lows: table number (arabic numerals), title,

body and table footnotes (use the same

footnote symbols as for general footnotes.)

References should be noted in the text by

superscript arabic numerals at the appro-

priate points. The citations should be typed

on a separate page headed “references”

and should be listed in numerical order.

The following illustrate the form to be

used in the list of references.

1. Coggeshall, R. E. 1967. A light and electron micro-

scope study of the central nervous system of the

leech, Hirudo medicinalis. J. Neurophysiol., 27:

229-289.

. DeVellis, J. and G. Kukes. 1973. Regulation of glial

cell function by hormones and ions. Tex. Rep. Biol.

Med., 31: 271-293.

. Mehler, W. R. 1966. Further notes on the center

median nucleus of Luys. In: The Thalamus. D. P.

Purpura and M. D. Yahr, eds., Columbia Univer-

sity Press, New York, pp. 109-127.

. Tremblay, J. P., M. Colonnier and H. McLennan.

1979. An electron microscope study of synaptic

contacts in the abdominal ganglion of Ap/ysia cali-

fornica. J. Comp. Neurol., 188: 367-390.

Abbreviations of journal titles should

follow those listed in the Jndex Medicus.

Responsibility for the correctness of the

44 INSTRUCTIONS TO CONTRIBUTORS

references lies with the author(s). Schedul-

ing pressures make it impossible for them

to be checked by either the Editors or the

publisher.

Send completed manuscripts and sup-

porting material to: The Editors, Journal

of the Washington Academy of Sciences,

Department of Biology, Georgetown Uni-

versity, 37th and O Streets, N.W., Wash-

ington, D.C. 20057.

DELEGATES TO THE WASHINGTON ACADEMY OF SCIENCES,

REPRESENTING THE LOCAL AFFILIATED SOCIETIES

ER OVESIINOTON occ c ean nte ccawcnscdeuausvesveuscwvecaceaveceur James F. Goff

rmnarery OF WASMINDION ... ees cca e sec ccuncncnncvetcececusecece Ruth H. Landman

ee William Ronald Heyer

ee MEREINIDEON cv ccc ccc tte ees veccesvacesucceavvcvevenausccvseseee Anne Brown

TIE OF WASMINGION .... 22. cece cere seus cevccccccccsccsseevecers Margaret Collins

EN Sn cs awe bbe Peale ree abo wR Raw eds dee als weak oWweeaeue T. Dale Stewart

EON TROIIIOCON: ccc cc cece enc was cecucesvesevdesucessvuaevave James V. O'Connor

me aa tne Pyistrict Of Columbia...........ccccccsvcccccnnecsveavece Charles E. Townsend

ese See kw in vcd Wi klble all Wika ed A De SUA Mabe Wie mle e eae Awe wme Paul H. Oehser

OTN aig Oo Wi we av ick a uN kb Wie w Weve Wels adldaKin-s wow ve bo ww ales Conrad B. Link

URIS cocci chee svi vole evan va vlicced seas obevdWeweaaseeedususbenne Boyd W. Post

ES George Abraham

netnctnanG EICCEONICS ENBINEErS ........0 0000s ccctencccencceccecsac George Abraham

rE MeeritAnICAl ENPINCETS . 2... 6.5. eect cece ce ncnteseecuscenceuteues Michael Chi

EE IGELY Gh WASHINGION .. 2. 6 ew cee rec eversweeaucecacee Robert S. Isenstein

i IELODIOIOLY . 2... 1 ce weet cca t cen tcnneuscecccbsccseee Lloyd G. Herman

SMC NINEATY ENPINCErS. ... 2... eee ccc eee ewtutvceccseceece H. P. Demuth

i CPMRORURNESROURECEES . 05... scl e c c ce cee ce enencceebenvanebtens@ne Wallace J. Cohen

een experimental Biolopy and Medicine.............0..ccccenscccscvcccces Cyrus R. Creveling

ERNE AIS. one sc rs vg ee hs ate cee necncnwanaceeseancvcaess Charles G. Interrante

eeeennaien Of Dental Research .......... 2... 0 eens cc cee e eee ceeresceces William R. Cotton

Semerican institute of Acronautics and Astronautics.............cceccccecceccecaee Richard P. Hallion

PE OMSIOR 6 oe ei oc, cng kasd ldcaivielba a tis ¢ nls be eels bus mee eeeien re’ A. James Wagner

RRM ASIP LON. ooo says oon ce otc cee eh et cece ce eee ce eeneesasevewinas Jack R. Plimmer

RNG CE, sos 5 ails enw 9 bY nae, w wie 'o Be wo lw om eles mie em Richard K. Cook

Ee cai 2 ain own au WS id wiara ew b:dopcik bie ale'e wie ele eR ER Dick Duffey

ES CO A. D. Berneking

I ni roc ic cS oo wid Wo win wdc wee vie dee cee een wks @ eeeenne Edwin R. Fuller, Jr.

oe ce a's Sa a wie'e Ge) wales Sade Dew ew ales ¥ Caeinw em Alayne A. Adams

EE RENCHCE CUID . os. oe cece cece klen esac scasswncereccenewadss Deborah J. Warner

EE EMILE HYSICS FEACHEFS «2.5.5. sie ee ccc cece ect a eteeseneseaauues Peggy A. Dixon

a SD 2 8 2S id cic eG scale Sele n'a io Sidiera’o eM okiScaewe Ae 4m «0G George J. Simonis

em Pint PiVSiOIOPistS ... ... <2. 2c cece eee n coe s eee ceasiseues Walter Shropshire, Jr.

mttions Research Council...........02.00cceecccecvenedecersvcusecns John G. Honig

CR Eee) ga. aide atolls a od a ed\v pols ainnin alee wieelew ee ea was lw’ Jewel B. Barlow

American Institute of Mining, Metallurgical

EEESTIOLIC OES 6 Soo cc eh oie oyslnle a a Kadlec aes Mee Peasy dewesuknevenn des Garrett R. Hyde

RE Gas oa) a old blu v.b vd nie Ein ob inte wide ie dials eee bile Some ede lp Benson J. Simon

Ste tA OS CAUSTICKICA oc os soca bane da cede rin de ds qenewadoseacedavesouns Patrick Hayes

Ea) Oy PF bok valica g ks va nhs pen asevmeke enced seekers Miloslav Racheigl, Jr.

EET SN ney ee eee H. N. Reynolds

me aint Lechtlical Group .........-..00cccccccscucsarceccesenseses Paul G. Campbell

NEMA OLOgiCal SOcicty .......--0--ccccnevccccccccscnveasccuces Howard E. Waterworth

Seeeteiieral Systems Research .........--0-cccccecccccccvsscsescucs Ronald W. Manderscheid

I st divs a) Malwa ea sie viewe Hae as dnie ta ween W Ske Rene eee Stanley Deutsch

EM sda civic wscvese deine seve cceevvecemesenvessecueuas Irwin M. Alperin

Association for Science, Technology and Innovation .........00ecceeeeeee nese eeeeeeeees Ralph I. Cole

ETT Siig NG e's sodas ch sees ciwsabectnerevonnes Ronald W. Manderscheid

Delegates continue in office until new selections are made by the representative societies.

2nd Class Postage Paid

at Arlington, Va.

and additional mailing offices.

Washington Academy of Sciences

1101 N. Highland St.

Arlington, Va. 22201

Return Requested with Form 3579

SMITHSONIAN INSTITUTION F,

LIBRARY ACQUISITIONS | : :

ROOM 51 °#£=\NHB | y

WASHINGTON, DC 20560 |

| 7

j ‘a

; 4g

| i

7 sf

j “a

<< 7

has 3

«ft

VOLUME 73

Number 2

j of the June, 1983

WaoHINGTON

ACADEMY .-. SCIENCES

ISSN 0043-0439

Issued Quarterly

at Washington, D.C.

SMITHSON

( NOVQ: )

\ 25 983 |

“ ?

S LIBRARIES _

CONTENTS

Articles:

D. J. GRIMES and R. R. COLWELL: Survival of Pathogenic Organisms in the

Anacostia and Potomac Rivers and the Chesapeake Bay Estuary ...........-.

DICK DUFFEY, PETER F. WIGGINS, and A. A. ELKADY: Gamma Ray Cali-

bration Standards Using Californium-252 Neutrons ..........-eeeeeeeeeeeees ;

A.G. WHEELER, JR.: Outbreaks of the Apple Red Bug: Difficulties in Identifying

a New Pest and Emergence of a Mirid Specialist .......-.seeeeeee seers eeees 60

RONALD W. MANDERSCHEID, ANNE K. McCARRICK, and SAM SIL-

BERGELD: The Assessment of Anxiety and Hostility in Dyadic Interaction ..

SHERMAN ROSS: The Scientific Awards of the Academy: 1983 .........+4+-

EIERINOTEITIETIESNICOES. . «. o.) die nto sc.iv vices ec bu Fw e een eteeree sean sewers

Washington Academy of Sciences

Founded in 1898

EXECUTIVE COMMITTEE

President

James F. Goff

President-Elect

Jean K. Boek

Secretary

John J. O’Hare

Treasurer

Ronald Manderscheid

Managers-at-Large

Jo-Anne Jackson

Grover C. Sherlin

Alfred Weissler

Elise A. Brown

James Wagner

Ralph I. Cole

BOARD OF MANAGERS

All delegates of affiliated

Societies (see Cover 3)

EDITORS

Irving Gray

Joseph Neale

| ACADEMY OFFICE

| 1101 N. Highland St.

| Arlington, Va. 22201

Telephone: (703) 276-9254

The Journal

This journal, the official organ of the Washington

Academy of Sciences, publishes historical articles, critical

reviews, and scholarly scientific articles; proceedings of

meetings of the Academy and its Board of Managers; and

other items of interest to Academy members. The Journal

appears four times a year (March, June, September, and

December)—the September issue contains a directory of

the Academy membership.

Subscription Rates

Members, fellows, and patrons in good standing receive

the Journal without charge. Subscriptions are available on

a calendar year basis only, payable in advance. Payment

must be made in U.S. currency at the following rates:

U.S. and Canada... 2°. $19.00

Foreign’ =... 20 cee 22.00

Single Copy Price ....- 7.30

UHURU Symposium—2 issues............ $15.00

Delivery by Air ..:..... 2.00

Single-copy price for Vol. 66, No. 1 (March, 1976) is $7.50.

Back Issues

Obtainable from the Academy office (address at bottom

of opposite column): Proceedings: Vols. 1-13 (1898-1910)

Index: To Vols. 1-13 of the Proceedings and Vols. 1-40 of —

the Journal Journal: Back issues, volumes, and sets (Vols.

1-62, 1911-1972) and all current issues.

Claims for Missing Numbers

Claims will not be allowed if received more than 60 days

after date of mailing plus time normally required for —

postal delivery and claim. No claims will be allowed ©

because of failure to notify the Academy of a change in

address. ;

Change of Address

Address changes should be sent promptly to the Academy |

office. Such notification should show both old and new

addresses and zip number.

Published quarterly in March, June, September, and December of each year by the Washington

Academy of Sciences, 1101 N. Highland St., Arlington, Va. 22201. Second class postage paid at

Arlington, Va. and additional mailing offices.

Journal of the Washington Academy of Sciences,

Volume 73, Number 2. Pages 45-50, June 1983.

“Survival of Pathogenic Organisms in the Anacostia

and Potomac Rivers and the Chesapeake Bay Estuary”’

D. J. Grimes and R. R. Colwell

Department of Microbiology, University of Maryland,

College Park, MD 20742

ABSTRACT

Highlights of microbiological studies of the Anacostia and Potomac Rivers and the Chesa-

peake Bay estuary, spanning a two-decade period, are presented, with emphasis on pathogenic

bacteria. Data collected document a seasonal distribution of several pathogens, including

Aeromonas, Escherichia, Salmonella, and Vibrio spp. These bacteria disappear from the water

column as temperatures approach 10°C. Laboratory studies designed to simulate in situ con-

ditions, and preliminary in situ experiments now permit a new interpretation of this seasonal

distribution. While a small number of cells die upon entering natural bodies of water, a fluo-

rescent antibody/direct viable counting procedure demonstrates that the large majority of the

cells remain viable, but non-culturable by routine bacteriological methods. This finding has

far reaching implications, not only for water quality studies of the Anacostia and Potomac

Rivers and the Chesapeake Bay, but for water quality surveillance programs in general.

Our laboratory has studied the microbi-

ology of the Anacostia and Potomac Rivers

and the Chesapeake Bay estuary for almost

2 decades. From our early work in the mid-

1960’s, to our present study of indicator

and pathogenic bacteria in the Potomac

and Anacostia Rivers we have accumulated

data that document the occurrence and

survival of pathogenic microorganisms in

aquatic systems. In this paper, we will pre-

sent highlights of studies involving patho-

gens in estuarine systems, with remarks

concerning preliminary data and the di-

rection of our current Potomac River

investigation.

We began surveying the upper (tidal) Po-

tomac River for fecal indicator bacteria

and enteric pathogens in 1965. Indicator

counts were high, ranging from several

hundred per 100 ml to several million, and

Salmonella enteritidis serotype typhimurium

was isolated from the Three Sisters Island

area (Colwell, unpubl. data). In 1977, we

did a retrospective computer study of bac-

teriological water quality data collected by

others for the Potomac River from 1950 to

1977.’ Conclusions were tenuous, because

of the variability between investigators

who collected the data. However, one gen-

eral irend was apparent. There was a signif-

icant decline in water quality, as measured

by total coliform counts, at the Great Falls,

Maryland sampling station over this 27-

year period. The obvious conclusion was

that urbanization was having a deleterious

effect on this stretch of the river.

Also in 1977, Kaper er al.” published re-

sults of a study of Salmonella species in

Chesapeake Bay. Of 72 water and sediment

samples collected, 17 (23.6%) yielded Sa/-

46 D. J. GRIMES AND R. R. COLWELL

monella spp. The largest number of salmo-

nellae were detected at Jones Falls, where

sediment was observed to have a most-

probable-number (MPN) index of 110 sal-

monellae per gram. A seasonal pattern was

observed for the incidence of salmonellae,

suggesting that they disappeared from both

water and sediment when water tempera-

tures dropped below 10°C. Our current in-

terpretation of this phenomenon is that the

cells may have become non-culturable, a

concept that will be addressed in more de-

tail later in this paper.

It is now accepted that Vibrio parahae-

molyticus is a ubiquitous member of the

bacterial community of estuarine and

coastal waters.’ This human pathogen has

been isolated from the uppermost reaches

of the Potomac River,’ and its distribution

is affected by temperature,” ° salinity,’ dis-

solved oxygen,” sedimentation,’ and inter-

action with higher organisms, such as

zooplankton’ and blue crabs.* Recently, V.

cholerae, causitive agent of human cholera,

has been added to the list of vibrios which

are autochthonous members of estuarine

bacterial communities. In 1977, we hypoth-

esized that V. cholerae is a normal inhabit-

ant of brackish water and estuarine envi-

ronments.’ Both 01 and non-01 serovars of

V. cholerae have been repeatedly detected

by us in Chesapeake Bay samples. Work

now underway in our laboratory strongly

suggests that V. cholerae becomes non-

culturable in aquatic environments, when

temperature, salinity, and/or nutrient con-

centrations become suboptimal.’ In other

words, the organisms can not be detected

by conventional cultural techniques, but

are, nevertheless, still viable. The following

two figures illustrate this most interesting

phenomenon. Figure | is a V. cholerae cul-

ture stained with an antiserum that con-

tains fluorescent-labelled antibody (FAB)

against V. cholerae; this is the technique we

have chosen to investigate non-culturabil-

ity. Figure 2 summarizes the results of sev-

eral experiments in which FAB counts of V.

cholerae suspended in sterile Chesapeake

Bay water at 4°C were compared with con-

ventional cultural enumeration techniques.

Finally, we have shown that V. cholerae can

attach to copepods (Figure 3), a feature

which may help explain the apparent “‘dis-

appearance” of this pathogen from aquatic

habitats."

|S be

Vibrio cholerae ATCC 14035 adsorbed with rabbit polyvalent antiserum, stained with fluoroisothio-

cyanate-labelled goat anti-rabbit antiserum, and viewed with epifluorescent microscopy (1000). Photo cour-

tesy of H.-S. Xu.

AQUATIC PATHOGENS 47

Log,, Number of Cells per ml

@FA

DAODC

eTCBS

OTSA

AAPB-MPN

ATSB-MPN

1 3 5 - 9 11

Days

Fig. 2. Survival of Vibrio cholerae ATCC 14035 in

sterile Chesapeake Bay water (4°C, 11°/o00 salinity) as

measured by acridine orange direct count (AODC);

fluorescent antibody (FA); plate counts on tryptic soy

agar (TSA) and thiosulfate-citrate-bile salts-sucrose

(TCBS) agar: and most-probable-number (MPN) de-

terminations in alkaline peptone broth (APB) and

tryptic soy broth (TSB).

Members of the genus Aeromonas are

also autochthonous members of the estu-

arine environment.’’ However, unlike vi-

brios, aeromonads, in general, do not re-

quire NaCl for growth. Certain Aeromonas

spp. are capable of causing human disease,

and these bacteria have, therefore, been

studied both as agents of water-borne dis-

ease’ and as possible indicators of water

pollution.’* Ina study of Aeromonas spp. in

the Anacostia River, we reported that inci-

dence of this group was directly related to

water temperature. During summer

months, Aeromonas concentrations rose to

a maximum of 30,000 cells per 100 ml of

water and 40,000 cells per gram of sedi-

ment; counts subsequently dropped by 2 to

4 logs when the water temperature de-

creased to 0°C. This same relationship to

temperature has been observed by other

investigators.'” '® Seasonal growth was

also examined ina study recently published

by LaLiberte and Grimes,'’ in which

Escherichia coli was shown to be capable of

in situ growth in sterile riverine sediment

and extended in situ survival in non-sterile

sediment (Figure 4). Interestingly, as shown

in Figure 5, strong correlations (r = 0.92)

were detected between Aeromonas and coli-

form concentrations in sediment.'* These

data strongly suggest that both Aeromonas

and E. coli coliforms grow in the natural

environment during the summer months.

In the same study,'* we examined Navy

divers conducting training exercises at the

U.S. Navy Yard in Washington, D.C. Pre-

and post-dive samples were collected from

divers’ masks, ears, noses, and throats.

Water and sediment samples were also col-

lected on two occasions, August and Oc-

tober, 1979. Results are presented in Table

1. Several of the Aeromonas isolates were

examined for virulence characteristics, using

both the YI adrenal cell and rabbit ileal

loop assays. Approximately 38% of the 365

aeromonads examined were cytotoxic for

Yl adrenal cells, and 8 of 17 cultures were

positive for fluid accumulation in the rab-

bit ileal loop. Thus, Anacostia River water

harbors human pathogens in greater num-

bers than traditional bacterial indices would

suggest and demonstrate virulence charac-

teristics.

Another study, which also involved

Aeromonas, was done with the cooperation

of Navy divers working in the Anacostia

River and illustrates the public health sig-

nificance of aeromonads in aquatic envi-

ronments.'> Approximately 30 min after

surfacing from a 10 min dive to 130 feet, a

Navy diver experienced a severe left tem-

poral headache and became very weak. The

only apparent medical problem was a

puncture wound to his left lower leg, sus-

tained the previous evening while scuba

diving in the Anacostia River. The wound

was red and swollen, with considerable

pain in the affected area, knee, and left

48 D. J. GRIMES AND R. R. COLWELL

Fig. 3. Oral region of a living copepod artificially contaminated with Vibrio cholerae serovar 01 (strain

CA401). Black bar inset is 10 um. Photo courtesy of A. Huq.

groin. Following re- and de-compression in

a compression chamber, the diver was

admitted to a hospital where an aspirate

collected from the puncture wound subse-

quently yielded a mixed culture of Aeromo-

nas hydrophila and A. sobria. Both cultures

were cytotoxic for Yl adrenal cells. The A.

sobria isolate was also enterotoxigenic

when tested ina rabbit ileal loop. The diver

was placed on tetracycline for 10 days, and

recovery was complete.

Recently, we studied the distribution and

toxigenicity of obligately anaerobic bacte-

ria in the Anacostia River.'* Several obli-

gate anaerobes were recovered, primarily

of the genera Bacteroides, the most preva-

lent anaerobe in human disease, and C/os-

tridium, species of which are causative

agents of botulism, tetanus, food poisoning,

and gaseous gangrene. Anaerobe counts

were similar in both water and sediment,

ranging from 10° cells/ml in the warmer

months to 10/ml in winter. Of interest was

the fact that even in the presence of dis-

solved oxygen, these anaerobes were pres-

ent and culturable. Also, several of the iso-

lates were shown to be cytotoxic and

enterotoxic. Thus, obligate anaerobes sur-

vive in, and present health hazards to per-

sons in contact with, Anacostia River.

We are now conducting a microbiological

survey of the Potomac and Anacostia Riv-

ers, with five stations being monitored: the

Pennsylvania Avenue and East Capital

Street bridges on the Anacostia River; and

water off Fletcher’s boat house, Three Sis-

ters Island, and the Tidal Basin inlet on the

Potomac. Water samples are analyzed for

standard plate count, MPN total coliform,

MPN fecal coliform, and MPN fecal strep-

tococcus indices. In addition, the samples

are also examined for: Salmonella (MPN

and FAB); Vibrio cholerae; Aeromonas hy-

drophila; Pseudomonas aeruginosa (MPN)

and Campylobacter jejuni (MPN). Prelimi-

nary data for the indicator bacteria listed

AQUATIC PATHOGENS 49

Log,, Number of Cells per g

b nr

20 40 60 _ 80

Hours

100

Fig. 4. Number of Escherichia coli cells in auto-

claved, inoculated sediment (A) and in nonsterile, in-

oculated sediment (B).

above confirm previous studies, in that the

Anacostia stations have significantly higher

counts of bacteria than Potomac River sta-

tions, especially during and after a heavy

rainfall. The station at which the lowest

concentrations were recorded to date Is the

Three Sisters Island station, at which FC

and FS counts of less than 50 per 100 ml

were recorded during periods of little or no

rainfall. Preliminary results indicate that

Salmonella, Pseudomonas, and Aeromonas

spp. can be isolated from water samples

collected from both the Anacostia and Po-

tomac Rivers.

@ Aeromonas

A Total coliforms

® Fecal coliforms

”

—_— 4

—

—=

e's

o 2

ig Bier, hi, Wa WAT Dig EE i ae

Month

Fig. 5. Total and fecal coliforms and number of

Aeromonas cells in sediment from the Anacostia

River.

In summary, we have clearly demon-

strated that pathogenic bacteria survive,

and in certain cases, inhabit and grow in,

the Potomac and Anacostia Rivers and in

the Chesapeake Bay. Furthermore, these

pathogens often out-number the fecal indi-

cator bacteria. Based on toxigenicity test-

ing, they demonstrate a potential for being

pathogenic for humans. With this informa-

tion. it is imperative that decisions as to the

appropriate use of the Anacostia and Po-

Table 1.—Incidence of Aeromonas in water and sediment samples from, and in divers and their masks before and

after swimming in, the Anacostia River.

No. of Aeromonas

Water cells per No. positive before ~ no. positive after/no. tested

temp. AAA Tie mTOR ‘

Month hs ml water g sediment Mask Ear Nose Throat

August 28 300 >10° Oe $3VES) VO 4/15 Ore A/S DAIS

October 13 50 >10° 0-2/10 O0-9/10 0-0/10 0-— 2/10

50 D. J. GRIMES AND R. R. COLWELL

tomac Rivers, relative to public health fac-

tors, be made with care.

Acknowledgements

This paper was presented at the Seminar

on Water Quality in the Potomac Estuary,

U.S. Geographical Survey, Reston, VA,

September 14, 1982. Results cited in this

manuscript were obtained from studies

supported by the National Science Foun-

dation (Grant No. DEB 77-14646), Na-

tional Sea Grant (NA81AA-D-0040), Of-

fice of Naval Research (N00014-8 1-K0638),

National Oceanic and Atmospheric Ad-

ministration (Contract 04-8-M01-71), and

Washington D.C. Department of Environ-

mental Services (Contract ES-82-31).

References

1. Colwell, R. R. and Belas, M. R., 1978. Microbio-

logical studies. Interstate Commission on the

Potomac River Basin, Bethesda, MD.

2. Kaper, J. B., Sayler, G. S., Baldini, M. M. and

Colwell, R. R., 1977. Ambient temperature pri-

mary nonselective enrichment for isolation of

Salmonella spp. from an estuarine environment.

Appl. Environ. Microbiol. 33: 829-836.

3. Joseph, S. W., Kaper, J. and Colwell, R. R., 1982.

Vibrio parahaemolyticus and related halophilic

vibrios. CRC Crit. Rev. Microbiol., CRC Press.

10: 77-124.

4. Kaper, J. B., Remmers, E. F., Lockman, H. and

Colwell, R. R., 1981. Distribution of Vibrio para-

haemolyticus in Chesapeake Bay during the

summer. Estuaries. 4: 321-327.

5. Kaneko, T. and Colwell, R. R., 1973. Ecology of

Vibrio parahaemolyticus in Chesapeake Bay. J.

Bacteriol. 113: 24-32.

6. Kaneko, T. and Colwell, R. R., 1978. The annual

cycle of Vibrio parahaemolyticus in Chesapeake

Bay. Microbial Ecol. 4: 135-155.

7. Sayler, G. S., Nelson, J. D., Jr., Justice, A. and

Colwell, R. R., 1976. Incidence of Salmonella spp.,

Clostidium botulinum, and Vibrio parahaemolyti-

10.

Le:

18.

cus in an estuary. Appl. Environ. Microbiol. 31:

723-730.

. Krantz, G. E., Colwell, R. R. and Lovelace, E.,

1969. Vibrio parahaemolyticus from the blue crab

Callinectes sapidus in Chesapeake Bay. Science

164: 1286-1287.

. Colwell, R. R., Kaper, J. and Joseph, S. W., 1977.

Vibrio cholerae, Vibrio parahaemolyticus and other

vibrios: occurrence and distribution in Chesa-

peake Bay. Science 198: 394-396.

Xu, Huai-Shu, Roberts, N., Singleton, F. L., Att-

well, R. W., Grimes, D. J. and Colwell, R. R., 1982.

Survival and viability of non-culturable Escheri-

chia coli and Vibrio cholerae in the estuarine and

marine environment. Microbial Ecol. 8: 313-323.

. Huq, A., Small, E. B., West, P. A., Hug, M. I.,

Rahman, R. and Colwell, R. R., 1983. Ecological

relationships between Vibrio cholerae and plank- |

tonic crustacean copepods. Appl. Environ. Mi-

crobiol. 45: 275-283.

. Kaper, J. B., Lockman, J. and Colwell, R. R., 1981.

Aeromonas hydrophila: ecology and toxigenicity

of isolates from anestuary. J. Appl. Bacteriol. 50:

359-377.

. Joseph, S. W., Daily, O. P., Hunt, S. W., Seidler,

R. J., Allen, D. A. and Colwell, R. R., 1979. Aero-

monas primary wound infection of a diver in pol-

luted waters. J. Clin. Microbiol. 10: 46-49.

. Seidler, R. J., Allen, D. A., Lockman, H., Colwell,

R. R., Joseph, S. W. and Daily, O. P., 1980. Isola-

tion, enumeration, and characterization of

Aeromonas from polluted waters encountered in

diving operations. Appl. Environ. Microbiol. 39:

1010-1018.

. Cavari, B. Z., Allen, D. A. and Colwell, R. R., 1981.

Effect of temperature on growth and activity of

Aeromonas spp. and mixed bacterial populations

in the Anacostia River. Appl. Environ. Microbiol.

41: 1052-1054.

. Hazen, T. C. and Fliermans, C. B., 1979. Distribu-

tions of Aeromonas hydrophila in natural and

man-made thermal effluents. Appl. Environ. Mi-

crobiol. 38: 166-168.

LaLiberte, P. and Grimes, D. J., 1982. Survival of

Escherichia coli in lake bottom sediment. Appl.

Environ. Microbiol. 43: 623-628.

Daily, O. P., Joseph, S. W., Gillmore, J. D., Col-

well, R. R. and Seidler, R. J., 1981. Identification,

distribution, and toxigenicity of obligate anae-

robes in polluted waters. Appl. Environ. Micro-

biol. 41: 1074-1077.

Journal of the Washington Academy of Sciences,

Volume 73. Number 2. Pages 51-59, June 1983.

Gamma Ray Calibration Standards Using

Californium-252 Neutrons

Dick Duffey*, Peter F. Wigginst, and A. A. Elkadyt

University of Maryland, College Park, MD 20742*,

U.S. Naval Academy, Annapolis, MD 21402+, and

Atomic Research Establishment, Cairo, United Arab Republict

ABSTRACT

A 252 Cf source of 10 micrograms (2.3 X 10’ neutrons/sec.) or less in a small easily shielded

arrangement with a few target elements can readily provide a series of gamma ray energy

standards from about | MeV to 10 MeV. This overlaps and extends the energy range of the

usual decay standards. Such a convenient assembly of moderate cost may be attractive for the

expanding calibration needs of nuclear spectroscopy.

Introduction

Photon energy measurements are often

required for many measurements in exper-

imental and applied science. In a nuclear

laboratory, this often involves a gamma

ray detector for spectroscopy, and conven-

ient standards are needed.” * The gamma

ray calibration energies available have been

tabulated by Marion;’ these are mostly

from decay sources but some gamma ener-

gies for neutron capture, and from other

nuclear reactions, are included.

The decay sources, which are small, port-

able, long lasting, and without significant

radiation safety and licensing requirements,

have served most calibration needs. Com-

Table I.—Decay Gamma Ray Standards.

Nuclide Half Life Energy

Na 2.6 yr 511 (annihilation)

Cs 30.2 yr 661

Co 5.26 yr 1.173 and 1.333

"x 1.28 X 10° yr 1.461

ad | 3.06 min 2.615 and .583

(7°*TI is from the decay chain of *’Th,

half life 1.41 X 10’° yr)

mon decay standards are listed in Table 1.

Of these, '?’Cs, °°Co, and the natural iso-

topes ““K and *°*T (thorium chain) are

widely used because of half life and availa-

bility. A thorium decay spectrum taken

with a Ge(l1) detector and showing particu-

larly the useful 2.615 Mev energy is in Fig-

ure 1. The lines of thorium, as well as of

potassium, from the structural and shield-

ing concrete often appear in the gamma

background of a laboratory and assist the

experimenter.

Two higher energy decay sources of

some use are °°Co (3.548 MeV, 72 day half

life) and °°Ga (4.805 MeV, 9.4 hr); how-

ever, the short life is limiting. The pluto-

nium beryllium neutron sources (as the

alloy Pu Be;3), available at many laborato-

ries has led to some use of the 4.439 MeV

gamma ray line that accompanies the reac-

tion of alpha particles on °Be to yield "°C

and a neutron.” In any case, the decay

standards are limited, and the increased in-

terest in somewhat higher energies has

created a need for additional standards. A

high energy line of some use is the 6.130

MeV photon of '°O from nuclear reactions

of fast neutrons of a reactor core, or other

neutron source, with water or oxides.

52 DICK DUFFEY, PETER F. WIGGINS AND A. A. ELKADY

- 2087, 9,51

- 228n- 0,9)

-228 Ac 0,98

1000 a

~ 100

100

- 2087, Z 0 45)

200

THORIUM OXALATE

Decay 10 MIN.

- 2087, (Tcl!) 2.61 Mev (F)

300

CHANNEL

Fig. 1. Thorium oxalate decay spectrum.

As indicated, the neutron capture gamma

rays can extend the decay energies availa-

ble. If the experiments involve reactor neu-

trons, capture gamma rays are readily pro-

duced for calibration by placing suitable

targets in the neutron field with an appro-

priately shielded detector nearby. Determi-

nation of gamma rays for calibration using

reactor neutrons have been reported by

Greenwood.” ° Decay neutron sources de-

pending on alpha reactions can also serve,

but the relatively low neutron intensity is a

disadvantage: *°’Cf with its relatively high

neutron yield is more useful.’

“aa

GAMMA CALIBRATION STANDARD 53

Table I1.—Capture Gamma Ray Standards and Sensitivities.

Cross Line

Section Atomic Energy Intensity Sensitivity

Element o barn Mass MeV I y/100 n. I a/A

Fe 2.62 55.85 9.298 3.85 181

7.646 22.14 1.04

7.632 27.19 1.27

6.018 8.08 .379

5.921 8.29 389

Si .160 28.09 4.934 70.55 .402

3.539 79.58 453

Al 233 26.98 7.724 20.10 ita

Ca .430 40.08 6.420 28.09 301

Cr 3.10 51.996 9.720 9.82 585

8.884 24.14 1.44

7.939 11.41 .680

Ni 4.6 58.71 8.999 41.65 5.2

8.533 18.74 >,

6.837 11.91 .93

Ti 6.09 47.90 6.760 54.07 6.87

6.418 36.47 4.64

Pb .170 207.2 7.368 94.77 .0777

N .075 14.007 10.828 15.00 .08

3:267 25.41 .136

H 333 1.0078 2.223 100.00 33.04

POLYETHYLENE (Boron)

GE(L1)

DETECTOR

252ce STORAGE DRUM IRRADIATION ARRANGEMENT

Fig. 2. **°Cf irradiation arrangement.

COUNTS

~ >

oOo > <

ar <x Vv

- YW ww

sa w a

Ose Ss :

1000 S

S ~

~ wo

fom) t N

N bat

e =~

= a

Bie gms | R =

ce Oo S

~ N oR

par 1 a a

== '

100

HP Zoe =

300 400 500

200

DICK DUFFEY, PETER F. WIGGINS AND A. A. ELKADY

me TRON PLATE

i=)

~ 133 Min. 10/n/sec, 2°2Cr

Wo =

= - wn

uw wo ~

oar \ x

N Lam]

on ~

“— wo . .

wo i=) N

a at ad 4

wn” w -

toe fod ~~ wo ™~

a N

Ciel : !

LAL oO wo

- 7,63, 765 (F)

ze 8.28 (p)

Ps 6135 (D) =

- 8.79 (s)

- 9,30 Mev (F)

600

CHANNEL

700 800 900 1000

Fig. 3. Iron plate neutron capture gamma ray spectrum.

Sensitivity and Energies

Sensitivity to capture gamma methods

depends on the neutron absorption cross

section, o, and the capture gamma ray yield

expressed as (I gamma rays per 100 neu-

trons absorbed). The product (Io/A) where

A is the mass number gives a sensitivity

index.® Table II gives values for some ele-

ments of interest for standards and their

more useful energies.

Utilization of capture gamma techniques

for standards require accurate values of en-

ergies. Groshev’ has published an extensive

tabulation. More recently, Rasmussen” at

the M.I.T. reactor, using a Ge(Li) detector

with Nal scintillators that allowed coinci-

dence counting, measured the thermal neu-

tron capture gamma ray spectra for most of

the elements and has reported energies and

yields. Both Groshev’s and Rasmussen’s

tabulations have been useful in our work

which has been mostly towards applica-

tions by industry. These MIT values are

used for the sensitivities reported in this

paper.

Equipment Procedure

A ’°Cf source of a few micrograms was

employed. A typical *°*Cf source is doubly

encapsulated in stainless steel, overall size

GAMMA CALIBRATION STANDARD

100

(S) AaW 12°6 -

(3) 88°8 -

(4) 02°8 -

(S288 =

900

oy Se" -

(9) 6h'Z -

800

(2) 26°9 -

Gl): 'SE.9 =

(2) 80°9 -

(2) tS 5. =

CHROMIUM TRIOXIDE

210 Min. 10’n/sec. 2°2Cr

mt *(4) 19° Uys -

200

(S) BAT H -

: : .

SLYNOD

CHANNEL

Fig. 4. Chromium trioxide neutron capture gamma ray spectrum.

56 DICK DUFFEY, PETER F. WIGGINS AND A. A. ELKADY

TITANIUM DIOXIDE

346 Min. 10’N/sec. 2°4CF

— > ox

an =

rice)

{ ow

f giesss x

et ' >

~ <

. oO

! lu

i] ~~

1000 os

2 =

= =

8 a

100

200 300

Fa ~~

(=)

— TF

(=) be

ar uw

int

—

1 = w

no ww

= LY —

= ana N wL

[ =) ~~ —Q e ee

ww - WO

ae) ae

a - 1 mS hE

° LN = <s

hk ' ~

1 NS =

Tr WO

(ve) |

‘oO

fos)

400 500 600 700

CHANNEL

Fig. 5. Titanium dioxide neutron capture gamma ray spectrum.

about 3/4” diameter and 3” long with a

stem for handling. The source storage

drum of steel filled with paraffin was easily

adapted for irradiations to yield standard

spectra; to do this asample cavity was built

at the top as in Figure 2.

After placing a target sample (perhaps a

few hundred grams of powder ina polyeth-

ylene bag) in the cavity shielded with poly-

ethylene blocks containing about a few

percent boron, the *°’Cf was raised into ir-

radiation positions by a string. About an

inch of polyethylene between the source

top and sample provided neutron modera-

tion. Irradiation time for an energy calibra-

tion spectrum varied depending on size of

source and kind of sample. The Ge(Li) unit

at the side, with some lead bricks to reduce

background, measured the gamma rays. A

typical Ge(L1) detector has a resolution of 2

Kev (FWHM) at the 1.333 MeV Co”

gamma ray line. Boron, in polyethylene or

as Boral, reduced the neutrons to the detec-

tor to limit possible damage and loss in sen-

sitivity. Data from a 1024 channel analyzer

was put ona tape which was processed ina

computer.

The Ge(Li) detectors for gamma rays

give not only the full energy peak (f) but

two lower peaks, namely, a single escape (Ss)

ee Ee eS

(

‘

;

8 la ea ai

£0 sie

GAMMA CALIBRATION STANDARD 57

- H 120 (p)

ALUMINUM PLATE

410 Min. 10’n/sec. 224Cr

= :

/ <a

Hs}

1000 my

: c

-—- =

n EAS

> Tg

oO wo Ap -_

on ww

rat rere)

i wo ~ =

100 a Nae A.

4 >

pat

Be Shee

(a a

200 300 400 500 600 700 800

CHANEL

Fig. 6. Aluminum plate neutron capture gamma ray spectrum.

at0.511 MeV less and a double escape (d) at

1.022 MeV less. This multiple peak phe-

nomena stems from the pair production of

the high energy gamma rays and annihila-

tion photons (.511 MeV) which may escape

from the detector. At energies above about

3 MeV, the double escape peak was most

pronounced with the detector used.

Results

To provide energy standards for a re-

search program, which was mostly toward

geologic materials, elements as pure metals,

oxides, carbonates, or other compounds

were irradiated, namely: NHzNO3, NaCOQOs,

MgO, Al2O3, SiO2, S, NaCl, K2CO3, CaO,

Sc203, T10), V20s, Cr203, M,OQO>2, Fe, iG.

Ni, Cu, Zn, Y203, BaO, Au, and Hg acetate.

Oxygen and carbon in these compounds

have a low sensitivity and add little to the

spectra. However, the large amounts of C

and H in polyethylene and the lead of the

shielding yield their lines, e.g., 4.945 and

3.684 MeV for C, 2.223 MeV for H, and

7.367 MeV for lead.

The spectrum for iron is in Figure 3. Iron

is an excellent calibration standard because

of its good response over a long energy

range from 9.30 MeV (f) down to the 1.201

MeV (d) line of hydrogen of the polyeth-

58 DICK DUFFEY, PETER F. WIGGINS AND A. A. ELKADY

DETECTOR CALIBRATION

Fe 9,30 MeV(rF)

it 9,00(F)

ENCHeY /Me\y

hey We)

S1 2.52(p)

S1 3,54(F)

S1 3,91(p)

§ 4,40(p)

C1 S,09(p)

Ti 5.74(p)

fy 6,22(p)

Al 6,70(p)

Cu 6,39(p)

Ni 7,51(p)

Ni 7,98(p)

Fe 8,28(p)

tlt 8,99(s)

i 2,22(F)

H 1.20(p)

200 400 600 300 1000

CHAWEEL

Fig. 7. Detector calibration curve.

ylene. Other useful elements for standards emitters) are available, Table III; all could

because of sensitivity, energy range, and __ serve in a calibration assembly. Radium-

spectral structure are chromium (Figure 4), beryllium sources are available for labora-

titanium (Figure 5),andaluminum (Figure — tory work; the half life is long but the ac-

6). The best standards are those withstrong companying gamma radiation isa problem.

isolated peaks with little interference from As indicated, the Pu-Be sources have been

the escape lines. widely available. All these decay sources of

A plot (Figure 7) of the energies against neutrons with their encapsulation provide

channels for the principal lines of the ele- some gamma background.

ments tested showed good linearity. Californium sources are inherently small

since a microgram supplies 2.3 X 10° neu-

trons/sec. Of course, the gamma rays from

Experimental Assembly Features the **Cf spontaneous fission and decaying

fission products also give some background.

Neutron Sources—A number of neutron Nevertheless, *’Cf appears to have advan-

sources of decaying isotopes (mostly alpha _ tages for such gamma ray standards.

Table III.—Isotopic Neutron Sources—1 curie (7)

Average

Isotope Neutron

Source Weight-gms Half Life Energy Neutron/Sec

a OE .0019 2.65 yr 2.3 Mev 4.4 xX 10°

*?>RaBe l 1620 yr 3.6 Mev S610;

**>CmBe .0003 163 d 4 Mev 4 x19"

**4CmBe 013 18.1 yr 4 Mev 4° GP

**8PuBe .06 87.4 yr 4 Mev 289010"

>!°Po Be .0002 138 d 4.3 Mev 2:5 10°

*4! AmBe 3 433 yr 4 Mev 2 Wali

>°PuBe 17 24.400 yr 4.5 Mev 2... aly

'4SbBe SASH 60 d 24 Mev 16 X10"

GAMMA CALIBRATION STANDARD 59

Assembly Materials—Spectral interfer-

ence from the Cr, Ni, and Fe of the stainless

steel encapsulation of the source can be

avoided by carefully shielding the detector

and source with lead, zirconium encapsula-

tion of the **’Cf can limit this. For modera-

tion, water and polyethylene are conven-

ient and provide a compact unit.

Polyethylene, which is easily fabricated

and does not leak, yields the carbon and the

H lines mentioned above. Water avoids the

capture gamma lines of carbon but adds

the 6.130 MeV line from '°O. Water proba-

bly allows the most flexibility in irradiation

geometry. For neutron shielding, boron is

useful since its main line (Doppler broaded)

is at .48 MeV, well below the energies of

most interest in capture gamma work.

Lithium, particularly °Li, is another useful

neutron shield. For gamma ray shielding,

lead is useful; however, care should be

taken to avoid impure lead bricks—many

contain Sb. If concrete is used for shielding,

quartz aggregate is preferable; concrete

may contribute lines of Si, Al, and Ca for

calibration.

Conclusions

A small *°’Cf source, its storage con-

tainer, and some shielding materials can be

arranged with a Ge(L1) detector for a cap-

ture gamma ray measurement. Irradiation

of a few common materials, e.g., Fe, Ni,

Cu, Mn, Ti, Al, Cl, Ni, and Si, can easily

provide a series of calibration energies up

to about 11 MeV. This extends the usual

decay standards and can assist measure-

ments in nuclear laboratories and in

industry.

References

1. Motz, H., 1970. ‘“‘Nuclear Capture Gamma Ray

Spectroscopy,” Annual Reviews of Nuclear

Science, Vol. 20.

2. Motz, N. and Backstrom, G., 1965. ‘‘Neutron Cap-

ture Radiation Spectroscopy, Alpha, Beta, and

Gamma Ray Spectroscopy.’ Editors K. Sieg-

bahn, N. Holland Pub. Co., Amsterdam.

3. Marion, J. B., 1968. “Gamma Ray Calibration

Energies,” Nuclear Data, Vol. A4, p. 301.

4. Mehta, K. K., 1966. ‘““Thermal Neutron Capture

Gamma Ray from Thulium 170,” Ph.D. Thesis,

University of Maryland.

5. Greenwood, R. C. and Chrien, R. E., 1978. ‘‘The

H (ny) Reaction Gamma Ray Energy: Revised

Bending Energies of *T, °C, '4C, and '°N,”’ Neu-

tron Capture Gamma Ray Spectroscopy Edited

by R. E. Chrien and Walter R. Kane, Plenum

Press, New York and London, p. 618.

6. Greenwood, R. C. and Chrien, R. E. Gamma Ray

Energies from '*N(n, y) Reaction,” IBID, p. 621.

7. Reining, W. C. and Evans, A. G., 1968. “‘Califor-

nium 252: A New Neutron Source for Activation

Analysis,’ The 1968 International Conference on

Modern Trends in Activation Analysis, National

Bureau of Standards, No. 33, p. 166.

8. Duffey, D., Elkady, A. A. and Senftle, F. E., 1970.

‘Analytical Sensitivities and Energies of Thermal

Neutron Capture Gamma Rays, Nuclear Instru-

ments and Methods,” Vol. 80, p. 149.

9. Groshey, L. V. et. al. 1959. “‘Atlas of the Spectra of

Gamma Rays from the Radiative Capture of

Thermal Neutrons,’ Pergamon Press, London

and New York.

10. Rasmussen, N. C. et. al. January 1969 “‘Thermal

Neutron Capture Gamma Ray Spectra of the

Elements,’ Report MITNE-85, Mass. Inst. Tech.

Journal of the Washington Academy of Sciences.

Volume 73, Number 2. Pages 60-64. June 1983.

Outbreaks of the Apple Red Bug:

Difficulties in Identifying a New Pest and

Emergence of a Mirid Specialist

A. G. Wheeler, Jr.

Bureau of Plant Industry, Pennsylvania

Department of Agriculture, Harrisburg, PA 17110

ABSTRACT

In the early 1900's two species of plant bugs (Hemiptera: Miridae), both undescribed, be-

came important pests of apple in New York and other northeastern states. M. V. Slingerland,

an economic entomologist at Cornell University, corresponded with the leading North Amer-

ican hemipterists in an attempt to obtain a name for one of the bugs, Heterocordylus malinus.

Although the so-called apple red bug was recognized as new to science, the actual description

was left to the European specialist O. M. Reuter. The problem in obtaining a name to be used

in a paper on life history reveals the immature state of mirid taxonomy in North America at

the beginning of the century. Correspondence between Slingerland and P. R. Uhler, E. P. Van

Duzee, and O. Heidemann is discussed, and it is suggested that the episode may have helped

influence H. H. Knight to become North America’s first specialist in Miridae. It also is shown

that Slingerland validated the name Heterocordylus malinus in a note on its injury to apple.

Because his paper predated the publication of Reuter’s formal description, authorship should

be credited to Slingerland.

In the first decade of the 20th century

two plant bugs of the family Miridae rose

from obscurity to threaten New York’s

apple industry. Heterocordylus malinus

Reuter, at first called the apple red bug, and

Lygidea mendax Reuter, the false apple red

bug,’ attracted the attention of orchardists

who sought help from entomologists at

Cornell University.

Cornell’s Mark Slingerland had been

aware of H. malinus since 1896 and had ob-

served its habits, but it took an outbreak

near Syracuse, New York, in spring 1908 to

prompt more intensive studies. Unfortu-

nately, his premature death kept him from

completing the work. After Slingerland’s

death in 1909,” Cyrus Crosby continued the

research and made extensive use of his col-

league’s preliminary notes and photographs

to provide the first life history informa-

tion.’ Soon entomologists in the Northeast

described their efforts to control the red

bugs; 43 papers appeared during 1915-19

at the peak of research activity.” Today

these once prominent pests are unfamiliar

to growers and to most entomologists as-

sociated with the apple industry; both mirid

species are suppressed by current manage-

ment practices.

Some 65 years after the initial outbreak,

a present-day entomologist, C. W. Schaefer,

reflected on and proposed explanations for

the red bugs’ “‘sudden rise, brief glory, and

swift fall.’’? As a native insect, H. malinus

most likely was associated with wild ro-

saceous hosts like species of hawthorn

(Crataegus). In western New York it may

have adopted apple as a host when this

immigrant tree was planted extensively in

the mid-19th century. Improved horticul-

tural practices, rather than the greater

number of trees, brought increased pro-

ductivity; thus the fruit and succulent fo-

liage of apple may have offered a more fa-

APPLE RED BUG HISTORY 61

vorable environment than provided by the

original hosts.°

Schaefer’s fascinating account of the rise

and fall of the apple red bugs complemented

the numerous and, necessarily, terse reports

of entomologists faced with the urgency of

combating a new insect problem. A story

yet untold is the difficulty Slingerland en-

countered in getting the bugs identified. It

involved more than merely sending speci-

mens “‘to an expert in far-away Finland.’”’

The account that follows describes the ef-

fort needed to secure a name for H. malinus,

admittedly a lesser pest than L. mendax,®

but the species for which old notes and cor-

respondence are available.’ My historical

research also revealed that Slingerland un-

intentionally validated the name Hetero-

cordylus malinus before Reuter’s formal de-

scription appeared; hence, authorship of the

species should be credited to Slingerland.

Identifying a New Pest

When Professor Slingerland encountered

the apple red bug in early 1896, he logically

turned to P. R. Uhler of Baltimore, Mary-

land, for an identification. Uhler was North

America’s first specialist in the Hemiptera-

Heteroptera and for many years the only

American authority available to make de-

terminations. “Like nearly all great natu-

ralists, he was a most helpful man; no

worker appealed to him in vain... .”"°

Slingerland sent material to Uhler and

learned that although the new pest had

been observed previously, it apparently

was undescribed.

The insect you have sent to me in the

little phial is a Capsid which has been

known to our eastern entomologists for

at least fifty years, but which appears to

be still unpublished. Some years since I

recognized it as forming a new genus for

which I selected the name Prodinia, hav-

ing reference to its bad habits as recog-

nized by Dr. Fitch, near Salem, N.Y. He

has not, however, published his obser-

vations on it. I have sent out a few spec-

imens under the name Prodinia strigata

Uhler, Ms. The description has long

been ready for the press, but I have held

it back to go in a longer paper on the

Capsidae. If you desire to describe it, do

so, but if not, I will send you the diag-

noses to print in your memoir. . . .

Can you, possibly, save a series of the

two sexes for me. It is a form not yet dis-

covered in Maryland, and it may belong

to the Canadian Fauna. I have two spec-

imens from northern Illinois. Fitch’s

manuscript name for it, Phytocoris albo-

punctatus, was absurd, and I have ac-

cordingly changed it. It is widely re-

moved from Phytocoris."'

Specimens of Heterocordylus malinus

bearing Uhler’s manuscript name Prodinia

strigata have not been located, and the con-

specificity of Asa Fitch’s insect and the one

Slingerland collected from apple in New

York perhaps could be questioned. But

Uhler was a good hemipterist, even if not a

specialist in the Miridae. Although he was

hampered by poor eyesight, an operation

in 1886 restored his sight, and it was not

until 1905 that his eyes again began to fail.’

The apple red bug did not yet threaten

apple production in New York, and Slinger-

land did not pursue detailed biological stud-

ies or publish a description of the species as

Uhler said he was free to do. But even before

the orchard near Syracuse sustained heavy

injury in 1908, Slingerland had sent speci-

mens collected at Syracuse to E. P. Van

Duzee, who succeeded Uhler as the leading

hemipterist in North America. The material

arrived in poor condition, but Van Duzee

gave Slingerland essentially the same opin-

ion Uhler had 11 years earlier: “I am very

sorry to have to report that this insect is en-

tirely new to me and is probably a still un-

described species of Lopidea. If I were you

I would get some perfect specimens and

send them to Mr. Otto Heinemann (sic),

National Museum, Wash., D.C., and ask

him to describe them if they are new. Iam

no authority on the Capsids but have a fair

collection representing something over 220

species of our native forms and have seen

many others but your bug is entirely new to

me.”?!3

62 A. G. WHEELER, JR.

Otto Heidemann, Honorary Custodian

of Hemiptera at the U.S. National Museum

(USNM), figures in the red bug story, al-

though Slingerland may not have sent spec-

imens to Washington. It appears that

Heidemann also recognized the red bug as

an undescribed species. A specimen from

Glen Ellyn, Illinois, housed in the USNM

collection, bears Heidemann’s handwritten

label: ““Capsid/new-/near; Orthocephalus.”’

Like Uhler and Van Duzee, Heidemann

did not describe the species, perhaps be-

cause he was unsure of the correct generic

placement, and sent the Illinois specimen

to the Finnish hemipterist O. M. Reuter. A

final judgment on the mirid’s status thus

was left to the acknowledged world’s au-

thority on the group,’ who also had re-

ceived material of the new species from Mr.

Van Duzee. After Van Duzee received

Reuter’s opinion, he informed Slingerland

of the results: “You may recollect that

about July Ist you send me for identifica-

tion a Capsid you said was injuring apples

which I reported to be probably an undes-

cribed species of Lopidea. A short time

after that I took a few examples on wild

apple trees at Colden, N.Y. and sent them

to Prof. Reuter of Abo, Finland and he re-

ports them to be a new species of Hetero-

cordylus which he will describe as H. mali-

nus. I send this notice so in case you wish to

publish any notice of the species you can

give it the proper name.””!

In spring 1908 when red bug populations

at Syracuse reached alarming levels, Slin-

gerland wanted to publish his biological

observations. He continued to ask Mr. Van

Duzee about the status of Reuter’s proposed

description and to send additional speci-

mens to Van Duzee for identification. In

June 1908 Van Duzee replied that he had

not yet received determined specimens from

Reuter but that Slingerland could proceed

with his plans for publication, using the

name “‘Heterocordylus malinus Reut., per-

haps with the statement that the description

is expected to appear in a paper to be pub-

lished by Dr. Reuter this summer.””’

Reuter’s description, which was not pub-

lished until the following year, was based

on specimens from Colden, New York, and

Glen Ellyn, Illinois.'’ However, before the

formal description had appeared, Slinger-

land used the name H. malinus in a talk given

to the Western New York Horticultural So-

ciety in January 1909. Slingerland’s paper,

published later that year in the society’s

Proceedings,'* predated Reuter’s descrip-

tion.’’ Slingerland referred to the new apple

pest as a “‘vermillion-red”’ bug that attacked

the unfolding leaves; later, the recently set

fruits.”” He noted that Reuter had named it

Heterocordylus malinus. Although Slinger-

land had no intention of validating the

name, his photograph of a late-instar

nymph establishes the identity of the spe-

cies, and his “‘description,” albeit brief, is

sufficient to validate the name.” I propose

that the correct designation for this species

is H. malinus Slingerland.

After Slingerland’s death, Prof. Crosby

sent additional specimens of both red bug

species to Van Duzee for identification.”

H.. malinus is quite variable in coloration,

and apparently Crosby was not certain

whether the extreme dark color form repre-

sented the same species.

Historical Interpretation

The difficulty Mark Slingerland encoun-

tered in obtaining a name for one of the

new apple pests reveals the immature state

of mirid taxonomy in the early 20th century.

Today, when a new insect problem arises in

North America it usually is easy to identify

the pest, or at least to place it to genus.

Such was not the case for the Miridae when

the red bugs gained prominence. P. R. Uhler

and E. P. Van Duzee worked with nearly all

groups of Hemiptera; both had described

new mirids, including some of the most

common North American species, but

neither specialized in the group.”

The struggle to get the red bug identified

may have significance for the effect it had

on Harry H. Knight, a student of Crosby’s

who was to become the first North Ameri-

can miridologist. He certainly was familiar

with the red bug outbreaks; he commented

APPLE RED BUG HISTORY 63

in 1915 that the ‘rapid development of

both species as pests has been quite

remarkable.””*

Knight entered Cornell in September

1910, having already completed two years

of undergraduate training at the State

Normal School in Springfield, Missouri

(now Southwest Missouri State University).

He began collecting mirids in 1911,”° the

year Crosby published the first substantial

biological data on apple red bugs. It prob-

ably was Crosby who encouraged Knight

to stay at Cornell for graduate work and to

characterize insect injury to apple fruit as

his thesis research.”°

Harry Knight soon encountered mirids

on apple other than the red bugs and discov-

ered that no one in North America could

identify the species he collected. To fulfill

what he saw as an urgent need for mirid

identifications, Knight decided to specialize

in the group.””

Knight soon dominated the family, his

accomplishments between 1915 and 1923

termed “‘one of the astonishing episodes in

the history of hemipterology.””* He con-

tinued active taxonomic work until his

death at the age of 87 on 6 September

1976.”° In his distinguished career he pub-

lished about 180 papers on the Miridae and

described some 1,200 new species. One

might wonder whether Harry Knight might

have turned to the systematics of some

other insect group, or even to another area

of entomology, had it not been for the

apple red bugs and the taxonomic problems

they posed for Professor Mark Slingerland.

Acknowledgments

Crucial to the completion of this paper

was the dating of Reuter’s description of

Heterocordylus malinus. Thanks to the ef-

forts of Antti Jansson, Zoological Museum,

University of Helsinki, Helsinki, Finland, I

was able to pinpoint the beginning of the

delivery for volume 36 of Acta Soc. Sci.

Fenn. lam grateful to him for his invaluable

assistance. I also thank the staff of the Na-

tional Agricultural Library, Beltsville,

Maryland, for supplying information which

established that Slingerland’s description

predated Reuter’s. For offering helpful

comments on earlier versions of the manu-

script I acknowledge T. J. Henry, System-

atic Entomology Laboratory, USDA, c/o

National Museum of Natural History,

Washington, DC; C. W. Schaefer, Biologi-

cal Sciences Group, University of Connect-

icut, Storrs, CT; and K. Valley, Bureau of

Plant Industry, Pennsylvania Department

of Agriculture, Harrisburg, PA. Access to

Slingerland’s ““Experiment File” at Cornell

University was provided by E. R. Hoebeke

and L. L. Pechuman, Department of En-

tomology, Cornell University, Ithaca, NY.

References Cited and Notes

1. When Heterocordylus malinus was referred to as

the apple red bug (or redbug), Lygidea mendax

often was called the bright or false apple red bug.

Probably because of its greater economic impor-

tance, the latter species was given the approved

common name “apple red bug” and H. malinus

was then called the “dark apple red bug.” Al-

though the Entomological Society of America re-

tains “‘apple red bug” for L. mendax and does not

currently list an approved name for H. malinus, |

will use “‘apple red bug”’ in its original sense, 1.e.,

for H. malinus.

2. Mark Vernon Slingerland, who died of Bright’s

disease on March 10, 1909 at the age of 44, wasa

prolific writer and a leader in the relatively young

field of economic entomology. J. H. Comstock,

‘*“Mark Vernon Slingerland,” J. Econ. Entomol.

2(1909), 195-196. Fora list of Slingerland’s writ-

ings, see M. D. Leonard, “‘A Bibliography of the

Writings of Professor Mark Vernon Slingerland,”

Cornell Univ. Agric. Exp. Stn. Bull. 348 (1914),

625-651.

3. C. R. Crosby, “‘Notes on the Life-History of Two

Species of Capsidae,”” Can. Entomol. 43 (1911),

17-20 and “‘The Apple Redbugs,”’ Cornell Univ.

Agric. Exp. Stn. Bull. 291 (1911), 213-230. Cyrus

R. Crosby (1879-1937) was a well-known eco-

nomic entomologist who worked in his spare time

on spider taxonomy. A. Mallis, American Ento-

mologists (New Brunswick, N.J.: Rutgers Univ.

Press, 1971), pp. 420-421.

4. C. W. Schaefer, ‘‘Rise and Fall of the Apple Red-

bugs,’ Conn. Entomol. Soc. Mem. (1974), 104.

. Ibid., p. 110.

6. Ibid., p. 110-111. See also R. A. Cushman, “The

Native Food-Plants of the Apple Red-Bugs,”’

Proc. Entomol. Soc. Wash. 18 (1916), 196; W. H.

18.

Jie)

A. G. WHEELER, JR.

Wellhouse, ‘““The Insect Fauna of the Genus Cra-

taegus,’ Cornell Univ. Agric. Exp. Stn. Mem. 56

(1922), 1054-1055; and P. J. Chapman and S. E.

Lienk, ‘‘Tortricid Fauna of Apple in New York

(Lepidoptera: Tortricidae).”’ N.Y. Agric. Exp. Stn.,

Geneva Spec. Publ. (1971), 4-8.

. M. V. Slingerland, “A Red Bug on Apple,” in

Proc. Fifty-fourth Annual Meeting of the Western

New York Horticultural Society, 1909, p. 91.

. The overwintered eggs of H. malinus hatch 7-10

days earlier than those of L. mendax. Thus, the

developing nymphs feed primarily on young fo-

liage and usually reach maturity before fruit is

large enough to be injured. H. H. Knight, “An In-

vestigation of the Scarring of Fruit Caused by

Apple Redbugs,”’ Cornell Univ. Agric. Exp. Stn.

Bull. 396 (1918), 198, 200.

. The Slingerland and Crosby material, part of ex-

tensive experiment files preserved in the Depart-

ment of Entomology, Cornell University, Ithaca,

N.Y., is labeled “‘Capsid (?) on Apple — — Expt.

No. 554.”

. L. O. Howard, “Philip Reese Uhler, L. L. D.,”’ En-

tomol. News 24 (1913), 436.

. P. R. Uhler to M. V. Slingerland, June 3, 1896.

. Howard, ‘“‘Philip Reece Uhler,” p. 437.

. E. P. Van Duzee to M. V. Slingerland, July 2, 1907.

. Odo M. Reuter (1850-1913), a linguist, poet, and

philosopher, was an outstanding entomologist

who published nearly 500 papers, including sev-

eral monographs on Miridae. O. Heidemann,

““O. M. Reuter,’’ Proc. Entomol. Soc. Wash. 16

(1914), 76-78. For a more detailed biography see

J. Sahlberg, ““Odo Morannal Reuter. Nagra Min-

nesord,” Entomol. Tidskr. 38 (1917), 62-96.

. E. P. Van Duzee to M. V. Slingerland, Oct. 26,

1907.

. E. P. Van Duzee to M. V. Slingerland, June 15,

1908.

. O. M. Reuter, ‘““Bemerkungen uber nearktische

Capsiden nebst Beschreibung neuer Arten,” Acta

Soc. Sci. Fenn. 36 (1909), 71. On p. 47, Lygidea

mendax was described on the basis of specimens

E. P. Van Duzee had collected at Colden, Go-

wanda, and Hamburg, N.Y.

Slingerland, ‘‘A Red Bug on Apple,” pp. 90-91.

Slingerland’s paper appeared early in 1909. A

copy in the National Agricultural Library, Belts-

ville, Maryland, bears an accession stamp; it was

received 8 April 1909. The publication date of

20.

22.

23:

24.

25:

26.

zie

28.

25;

Reuter’s description can be determined more pre-

cisely. In the minutes of the Societas Scientiarum

Fennicae for 20 September 1909 is the following

statement: “Printing of volumes 35 and 36 of Acta

Societas Scientiarum Fennicae have been com-

pleted and the volumes are ready for distribu-

tion.’ Antti Jansson to A. G. Wheeler, Jr., April 7,

1983.

Slingerland, ““A Red Bug on Apple,” p. 90.

Similarly, some of P. R. Uhler’s manuscript

names were validated by Otto Heidemann. Fora

discussion of what represents an adequate descrip-

tion, see A. G. Wheeler, Jr. and T. J. Henry, “‘Rec-

ognition of Seven Uhler Manuscript Names, with

Notes on Thirteen Other Species Used by Heide-

mann (1892) (Hemiptera: Miridae).”’ Trans. Am.

Entomol. Soc. 101 (1975), 355-356.

E. P. Van Duzee to C. R. Crosby, May 6, 1910.

See J. A. Slater, ‘““Harry H. Knight: an Apprecia-

tion and Remembrance,” Melsheimer Entomol.

Ser. 24 (1978), 1.

H. H. Knight, ‘“‘Observations on the Oviposition

of Certain Capsids,”’ J. Econ. Entomol. 8 (1915),

293:

H. H. Knight, ““A New Key to Species of Reutero-

scopus Kirk. with Descriptions of New Species

(Hemiptera, Miridae),”” Jowa State J. Sci. 40

(1965), 101.

H. H. Knight, ‘Studies on Insects Affecting the

Fruit of the Apple with Particular Reference to

the Characteristics of the Resulting Scars,” Cor-

nell Univ. Agric. Exp. Stn. Bull. 410 (1922),

447-498.

H. H. Knight, ‘“‘Forty Years of Progress on the

Classification of Family Miridae (Hemiptera),”

Proc. Twelfth Annual Meeting, North Central

Branch, Entomological Society of America (1958),

Slater, ‘‘Harry H. Knight: an Appreciation and

Remembrance,” p. 2.

T. A. Brindley, ‘““Harry H. Knight, 1889-1976,” J.

Econ. Entomol. 69 (1976), 792. For additional in-

formation on Knight’s mirid work see Slater,

‘Harry H. Knight: an Appreciation and Remem-

brance,”’ pp. 1-8; A. G. Wheeler, Jr., ““A Compar-

ison of the Plant-Bug Fauna of the Ithaca, New

York Area in 1910-1919 with That in 1978,” Jowa

State J. Res. 54 (1979), 29-35; and A. G. Wheeler,

Jr., “‘Plant Bugs at Cornell: a Changing Fauna,”

Cornell Plantations 36 (1980), 3-8.

Journal of the Washington Academy of Sciences,

Volume 73, Number 2, Pages 65-76, June 1983.

The Assessment of Anxiety and Hostility in

Dyadic Interaction

Ronald W. Manderscheid, Anne K. McCarrick, and Sam Silbergeld

Naticnal Institute of Mental Health

ABSTRACT

A need exists for quantitative measures of anxiety and hostility within dyadic relationships.

The present paper reports a study in which anxiety and hostility levels are measured with both

the Free Association Test (FAT), a content analysis procedure based on speech samples, and

the Dyadic Free Association Test (DFAT), an extension of the FAT to a dyadic format. Sub-

jects are seven married couples who participated in a research program On innovative modes

of mental-health service delivery. Patterns of anxiety and hostility in an individual context are

contrasted with those in a dyadic context. Findings indicate clinical and statistical differences

between the two contexts, and demonstrate the utility of the DFAT for assessment of the psy-

chodynamic aspects of a relationship. The DFAT might also be useful for training psycho-

therapists and for research in psychotherapy.

A need exists for quantitative measures

of anxiety and hostility within dyadic rela-

tionships. Such measures could have poten-

tial clinical utility for screening, diagnosis,

and assessment in a range of relation-

ships, e.g., husband-wife, parent-child, and

teacher-student. Moreover, in the therapist-

patient relationship, these measures could

be employed to improve training of thera-

pists, point out problem areas between

therapist and patient, and enhance learning

and interpersonal skills. Granted the po-

tential range of applications, it is surprising

that very little research has been conducted

on this topic.

Some time ago, the present researchers

conducted a small pretest to assess the feasi-

bility of generalizing the Free Association

Test (FAT), acontent analysis procedure

for deriving a range of anxiety and hostility

measures, from the standard subject-alone

application to a dyadic format for married

couples. The results of this research, which

were subsequently reported in the litera-

ture,” suggested the desirability of a more

comprehensive evaluation of the dyadic

technique for married couples. The pur-

pose of the present work is to report the re-

sults of a larger-scale, longitudinal pilot

study on this topic.

The FAT is acontent-analysis procedure

which can be applied to one or more verbal

samples collected from a subject. The

procedure focuses upon the implicit anx-

iety and hostility exhibited by a speaker,

rather than descriptions of the content

alone. The primary purpose ts to identify

and classify statements which serve as indi-

cators for varying levels of emotional states

which are present-time, specific, subject to

fluctuation, and influenced by environ-

mental details. Such affects may or may not

coincide with longer-term mood or per-

sonality variables.

In the standard, subject-alone applica-

tion of the FAT, each subject is requested

to sit alone in a room and record a five- to

seven-minute audiotape on personal expe-

riences reflective of feelings. Subsequently,

the tape is transcribed, and the resulting

66 RONALD W. MANDERSCHEID, ANNE K. MCCARRICK, AND SAM SILBERGELD

typescript is evaluated by means of content

analysis for specific forms of anxiety (death,

mutilation, separation, guilt, shame, dif-

fuse) and hostility (outward overt, outward

covert, inward, and ambivalent). Scores for

the six types of anxiety are combined intoa

summary measure, total anxiety. Summary

measures for total hostility outward and

total ambivalent and inward hostility are

obtained in a similar manner.

The protocol for the dyadic form of the

test (DFAT) differs from that of the subject-

alone FAT. In the husband-wife applica-

tion of the DFAT, spouses sit together in a

room and jointly record a 10-minute audio-

tape on personal experiences reflective of

feelings. As with the FAT, the tape is sub-

sequently transcribed. However, in distinc-

tion to the FAT, three sets of anxiety and

hostility scores are calculated: wife’s con-

tribution to the dyad, husband’s contribu-

tion to the dyad, and a dyad score obtained

by combining the contribution scores for

wife and husband.

The present analysis has been guided by

consideration of three focal issues. First,

do a subject’s anxiety and hostility scores

differ significantly between the spouse-

alone FAT and the conjoint FAT? If the

procedure is sensitive to details of the in-

terpersonal environment, scores may be

expected to vary between these two con-

texts. In addition, variability is anticipated

to be more pronounced if the relationship

with spouse is problematic. Second, do

scores for the two spouses differ to a larger

degree on the FAT or on the DFAT? Re-

sults from the earlier pretest* suggest that

scores are more congruent on the DFAT

than on the FAT. This would imply that

anxiety and hostility can be communicated

between marital partners. Third, do the

FAT and DFAT scores for a husband-wife

pair have utility for clinical diagnosis and

assessment of their relationship? Since FAT

scores have been shown to correlate well

with clinical observation and data,’ it may

be anticipated that the pattern of FAT and

DFAT scores for a couple will be useful for

analyzing the quality of their relationship.

Method

Design

Married couples were recruited on a self-

selection basis for participation in a re-

search program on innovative modes of

mental-health service delivery. The initial

phase of the program consisted of screen-

ing and history-taking for each spouse.

Spouses were required to have some college

background and no previous or concurrent

therapy. In addition, couples were excluded

from the program if either spouse was suici-

dal, psychotic, or not physically well.

The second phase of the program con-

sisted of a series of interviews with a psy-

chiatrist, spaced at convenient time inter-

vals of one week or longer, during which

each marital partner was interviewed sep-

arately, followed by a conjoint interview.

In the conjoint interview, spouses were

confronted by the therapist with their per-

sonal differences and problems. This

procedure had the advantage of permitting

the therapist to observe interactions be-

tween partners as part of the diagnostic

process, while, at the same time, preparing

the couple to enter group therapy.

The third phase of the program was

composed of 15 sessions of group psycho-

therapy. Sessions | and 2 were essentially

laissez-faire and designed to have minimal

therapist activity. In Sessions 3-12, the

therapists focused systematically and some-

what sequentially upon each marital

partner, with particular attention to behav-

ior, affect, and problems. Discussion con-

tent varied from session to session. For ex-

ample, the techniques of the therapist

involved active listening, offering alterna-

tive problem resolution strategies to deal

with conflicts, and numerous role-playing

situations designed to promote change to-

ward more effective dyadic behavior.

Throughout this process, verbal feedback

from all group members was selectively

utilized. The first hour of Session 13 was

filmed on videotape. The remainder of this

session was used to replay and discuss the

ANXIETY AND HOSTILITY IN DYADIC INTERACTION 67

resulting film. Participants’ feelings con-

cerning their appearance and their nonver-

bal performances were examined. Session

14 analyzed the previous separation anx-

iety experienced by participants and the re-

lationship between these experiences and

the forthcoming end of the group. In Ses-

sion 15, therapists and members provided a

spontaneous appraisal of each spouse’s

and each couple’s progress during the

group.

Subjects

Subjects were seven, volunteer, married

couples who participated in the research

program just described. The 14 partici-

pants ranged in age from 27 to 43 years,

with a mean of 36 for husbands and 35 for

wives. The mean length of marriage was

13.6 years, with a range of 7 to 21 years.

Mean educational level of husbands was 18

years, of wives, 14. One couple had three

children, one had none, and each of the re-

maining couples had two children. All sub-

jects were white. One wife was a home-

maker; all remaining participants were

employed outside the home, primarily in

white-collar occupations.

Procedure

For each of the seven couples in the sam-

ple, FAT and DFAT data were collected

prior to and following the first seven con-

secutive interview sessions in the second

phase of the research program. The follow-

ing procedure was employed for each of the

pre- and postsession data collection inter-

vals. First, the two spouses were placed in

separate rooms for the purpose of record-

ing individual, seven-minute audiotapes on

personal experiences reflective of feelings.

Second, after completing initial individual

taping, the spouses were brought together

immediately in a common room for the

purpose of recording a dyadic, 10-minute

audiotape on joint experiences reflective of

feelings. Subsequently, each tape was la-

beled with the identity of the spouse or

couple making it, the session number, and

a designation of whether it was a pre- or

postsession tape. A complete data set for a

single couple consisted of 14 individual

presession tapes, seven dyadic presession

tapes, 14 individual postsession tapes, and

seven dyadic postsession tapes. Note that

all data reported were collected prior to

group psychotherapy.

Each audiotape was transcribed. The re-

sulting typescript was “‘claused” according

to the Gottschalk, Winget, & Gleser?’

procedure, and each clause was rated for

anxiety and hostility. Each typescript from

a spouse-alone recording generated one set

of FAT scores; each typescript from a dy-

adic recording, three sets of DFAT scores

(contribution of wife, contribution of hus-

band, combined). Since a detailed proce-

dure is required to derive final scores,’ a

computer program was developed for this

purpose. Since resultant scores are ad-

justed for the total number of words spoken,

effects due to differential time length of the

FAT and DFAT tapes have been controlled.

Results

Table 1 shows results from two-way co-

variance analyses of FAT and DFAT data.

Independent variables are spouse (wife vs.

husband) and type of score (individual

FAT vs. contribution to DFAT); covar-

lates, session number and a dummy varia-

ble specifying whether the score derived

from a pression or postsession recording.

In each analysis, the actual score on one of

the 13 component or summary measures

serves as the dependent variable.

Several interesting contextual patterns

emerge in these analyses. Among the vari-

ables that exhibit significant differences,

anxiety scores are lower and hostility out-

ward scores are higher on the DFAT as

compared with the FAT. Specifically, scores

for separation, diffuse, and total anxiety

are lower, and those for hostility outward,

overt, covert, and total, are higher in the

DFAT context. Inward and ambivalent

ea)

—

Mad

eo}

100° > d= exe 10° > d = ax 60 >F= 400 >I =H,

a8) k LSOF8L0 IL04 £60 IL04960 080 F 88°0 [BIOL

x x 1) x 1 6v0 F890 OPOFLYO LSOFLIO0 7@0F 990 JUSTBAIQUIYy

dy : ~OFCO 6904220 8904990 890 F 89°0 pleMuy

40K IOTFCCI ILOF98O0 E€80F7II 14904060 [B10],

x ccOF8rO IE0FEVPO IS0F 650 OF OF 7ZHO WaAo0D pieMinC

40K 660 FETT 8904FSL0 CLOFL60 S90F I8'0 W9AQ PseMINC

AVIUISOLT

Pers x Pr 4K c60F 081 LOTFITZ CIF ELI OVI FOI? [BIOL

2k ok 44K mT 6K ELOFLII SOTFOSI SLOFLOI p80F SET asnjjid

* * 9LOF ESO 6804960 IL0F780 I80F €60 sWIBYS

890 FLO LIDOF6LO bLOF69O0 990F b2'0 WIND

4k ITOFSEO I9F€0F6E0 bp70F IPO 995049650 uoneiedag

4k * x * clO+ EEO SteOFOPO StEO0FPYO Pved0+F PO uoTe/ HN]

jCOF PEO ITEO0OFIEO OSTOFLEO LIOF PED yleoq

Ajaixup

SIJBLIBAOC) \SOd UOISS9G = =—s-: UO TOBIN} UT LVAC pueqsnyH LVAC LVA IVAd LVA S[QeLIeA

ylog SA old SA [VHA SA CRIT

SdJBIIVAOD puegsny OI

SIIJq IUBIIIUBIC

UOT]VIAIG PiepurIS F ues

‘dnoiy aijugq 10j e3eq [VAC pue Ly Jo sasAjeuy sduelseaoj—]| aque],

ANXIETY AND HOSTILITY IN DYADIC INTERACTION

"100° > d = exw ‘10 > d = on 60 > I= 4 Ol > =H),

* eS0+S590 OLOFL8O0 6904620 SLOFLLO [BIOL

4) * €$0+090 cro0+SSO 6904020 9904 £90 JUS|BAIQUIY

40k 6r0+cE0 c904+890 610F S50 LEOFSSO psemuy

4 x * CSOFIET 8904+ 680 eLOFLYPI 8904 €0'1 [BIOL

SCOF+EVPO OF 0F 660 CEOFLSO LIOFEPO W9Ao0D pleMino

4] x * psoFEecI 9SO0FSL0 CLOFSEI L904 001 MIAQ PIEMINO

AMUSO

Ok cSOF9ET LIOFELTI OFSOFI6O SOI F P6T [BIOL

* CSOF+90T e€LOFPCI t€9OF8L0 8604 0S'1 asnjflqd

* ce0F SPO 6904890 OCOFSEO 1904 L8'0 sWeYS

dq * cvOoO+rrO c90FILO O7TOFEEO SPOFLSO WIN’)

dt) ccOFEEO OLOFOEO cloO+r6c70 6004 L€0 uoneiedag

4] 4d] ecOFEEO CCOFLEO 610F 780 Iv04F 9F0 YON UN YAY

* I1T0+6c70 OLOFIEO 8004870 e104 6€0 A he

AJaIXUp

SOJBIIBAO’) 1SOd UOISS9G =: UOTJNB19}UT LVAC pueqsnyH LVAd LVA LVAd LVA 9[QelleA

ylog SA 31d SA [VA SA IJIAA

SOIBIIBAOD puergsny JIM

SIF JUBIIJIUBIS UONPIAIG PiepurIS F uvspy

*y ajdnoyd 10) saskjeuy sdueleAoj—'7Z afquy

70 RONALD W. MANDERSCHEID, ANNE K.

hostility also exhibit significant contextual

patterns. Scores for inward hostility are .

lower and those for ambivalent hostility

are higher on the DFAT.

Two variables show significant differ-

ences between spouses. Scores for mutila-

tion and separation anxiety are signifi-

cantly higher for wives; ambivalent hostility

exhibits a trend difference in the same di-

rection. A significant interaction between

the independent variables is not observed

in a single instance. However, a trend does

emerge for inward hostility. The latter re-

sult implies that the pattern of differences

between the FAT and DFAT on this varia-

ble is not the same for husbands and wives.

Although results in Table | give some in-

dication of overall patterns for the group,

more detailed analyses are necessary in

order to elicit information that has clinical

utility. Toward this goal, covariance anal-

yses identical to those displayed in Table |

have been performed separately for each

couple. Several of these by-couple analyses

are described below for _ illustrative

purposes.

Table 2 shows results for Couple A.

Findings for Couple A are of particular in-

terest because they reflect a symmetrical

pattern. Anxiety and hostility levels do not

differ significantly between husband and

wife on a single variable, and when signifi-

cant differences are observed between the

FAT and DFAT, both spouses change in

the same direction.

For both spouses of Couple A, each of

the anxiety variables that exhibits a signifi-

cant contextual effect is scored lower on the

DFAT. A similar pattern is observed for

hostility inward. By contrast, hostility

outward overt is higher on the DFAT.

Overall, the patterns observed for Couple

A are similar to those displayed by the

group as a whole.

Generally, results for Couple A imply

that the spouses are similar to each other in

patterns of anxiety and hostility, and that

presence or absence of the spouse exerts an

equivalent influence upon both partners.

Apparently, each spouse is more anxious

and self-critical when alone. These affects

MCCARRICK, AND SAM SILBERGELD

appear to be altered in the direction of

greater hostility outward when the spouses

are together. Such patterns prompt one to

observe that the negative affects in both

contexts may constitute a single system,

and that intervention to deal with affects in

the couple context may reduce the negative

affects experienced when the partners are

alone.

Table 3 displays results for Couple B. As

can be noted from the table, patterns for

Couple B are considerably different from

those for Couple A. The two partners from

Couple B exhibit significant differences on

several variables and show significantly

different change patterns between the FAT

and DFAT contexts. In fact, between the

FAT and DFAT contexts, the husband and

wife from Couple B do not change in the

same direction on a single variable.

The statistically significant findings for

Couple B can be summarized as follows.

The wife exhibits a somewhat higher level

of death anxiety than does the husband; the

husband, a considerably higher level of

shame anxiety than the wife, irrespective of

context. The large differential on shame

anxiety results in a higher level of overall

anxiety for the husband. The two remain-

ing wife-husband differentials cannot be

interpreted because each of the respective

variables interacts with the FAT/DFAT

contextual variable. The interaction effects

show that the two spouses do not change in

the same direction between the FAT and

DFAT. The patterning of the significant in-

teractions is instructive. Only one interac-

tion 1s observed among the anxiety varia-

bles: The wife’s level of mutilation anxiety

is lower; the husband’s about the same on

the DFAT compared with the FAT. Among

the hostility variables, the husband’s level

of hostility outward is higher; the wife’s

lower, on the DFAT compared with the

FAT. Inward hostility exhibits the opposite

pattern. Generally, the hostility totals re-

flect these relationships.

Unlike Couple A, Couple B appears to

exhibit a somewhat complementary pat-

tern of anxiety and hostility. When alone,

the husband experiences about three times

ANXIETY AND HOSTILITY IN DYADIC INTERACTION

100° > d = wax 10 > d = ae SO >I =4 01 >P=41,

* 4) 6$0F+690 680FE1I OSOFELO I€04+ SV0 [BIOL

LTO FEeeO sOO+r6CO POOFLEO t10+F 670 JUZTBAIQUIYY

* + 090FS590 680F ETI IS0F9O0 O0€04F 7F0 pseMuy

+ L90F+2490 s8tO0OF6¢0 crotFtrso e904 £80 [BIOL

SPOF LPO £00F 870 L7EOFS8EO 610F LEO WeAOD pseMINO

* 9v0F9SO0 st0F6EO0 I1E04F 870 £S042L0 W9IAQ PseMINO

ANUSOH

as +e * p90 F6CC CII F8E¢C C8OFLIT v8 0+ 107 [BIOL

ae oe . 990-960 cLOFCOT CLOFIOI Pp8OF IST asnjjld

4) * KOK 960-991 StTIFIZt c9O0FLIO0 ecSO+F E80 sweys

090+cL0 OLOF680 LYO+990 6F70+4 850 ying)

coO0O F870 £004 870 O10F0€0 I704 9€°0 uoneiedag

4 coo+ 870 £004 870 SsdOOFLTO O7TOF LED uonelhnnw

x OlrO0FIEO c0O0F 870 LOO0OFOPO I€04+ 70 yea

Ajaixup

CS EE a a a ee a a

S9IBIIBAOL) sOd uOIss9g = UOTJOBIIIUT =—ss LVAC pueqsnyH LY cl LVA LIVAd LVA a[QelLeA

ylog SA 31g SA IVA SA JJIM

Sd]BLIPAOD pueqsny IM

SIajJq UByIUsIS uoNeIANG piepuRIS F urs

aes

‘gq adnod 10} saskjeuy aduelBAOD—'¢ aque]

j=)

ea)

é)

[a4

ea)

—

ome

if)

Zz,

<—

i=]

fx)

77)

[a4

fe)

i=)

‘e)

100° > 4 = xe 10 > = xe 60 >IT =4 Ol >I =H,

OK OK * ek 0$04+990 6L0F 601 LLOFETI 8L0F 690 [B10

OOK xk 4k dy 4 0OS0O+F+SSO LPOFIVO 8L04F+ 160 £904 56°9'0 JUSTBATIQUIYY

Pen I@O+?@PO 9L04F960 LrPO0+790 9F0F PFD preMuy

14 OCI FIO psO0O+F680 LETFPTI 1804 680 [B10

* 91T04F+S€0 cro+FCrO pOT+F880 690 F190 WeAoD plemjino

J} SPI +09T SLOFISO 6604880 S904 790 IAQ Ps1eMINO

AUSOL

dy * Ok 4} OCT +te@c LII+68C SEL+99C 860+ 9ET [BIOL

+ * p8OFOPI IZTIFITZ SsOFSSI 09:0 + 980 asnjyiq

dq] * OOT+FE60 LLOF E60 STIFSEI 9804290 sueys

* L6OFIOT LLOFITI I01FSL0 €€0F IPO WIND

* vo0O+CHO pSO0OF6S0 OFO0FIIO vEO+F 8E0 uonviedas

* ClLO+9E0 e£€0F 860 I90F €20 CEOFEPVO uone] yn]

* cooO+FOEO O7TOFCEO I9104F9F0 I704F £0 yjeaq

Ajaixup

SOWUBAOD = SO = uOIssag_ Ss uoNoeINIUE «= LWA —s PueqsnyH Ivid LVA LVAd LVd a]qelied

ylog SA 31d SA LVA SA SJIM

Sd}VLIVAOD pueqsny TIM

S10IJJq JURIIUSIC

UOIeIAIG Plepuels TF uvspy

*?) aJdnozd 10j saskjeuy sduelIwVAOD—'p a1quy

ANXIETY AND HOSTILITY IN DYADIC INTERACTION 1

as much hostility inward as does the wife,

but only about half as much hostility out-

ward overt. Inthe DFAT context, husband

and wife show equivalent scores on hostil-

ity inward, due to a precipitous decrease

for the husband and a moderate increase

for the wife. Scores for hostility outward

overt also show transcontextual changes.

From the FAT to the DFAT, the wife’s

score decreases. By contrast, the husband’s

score increases to a level higher than that of

the wife. Among the anxiety variables, the

husband shows about twice as much shame

anxiety as the wife, irrespective of context.

One could infer the following dynamics

for Couple B. The results suggest that the

husband may convert his hostility inward

to hostility outward in the presence of his

wife. The wife appears to respond to this

with higher levels of hostility inward and

lower levels of hostility outward. This may

be the case because the wife is in a power

inferior position vis-a-vis the husband or

because she views the husband in a positive

manner. It also seems reasonable to specu-

late that the husband’s high level of hostil-

ity inward is correlated with his high level

of shame anxiety. A clinical question that

remains is the source of the husband’s

shame anxiety, 1.e., is it due to factors

within or outside the relationship? The

answer to this question could provide a key

to the observed anxiety and _ hostility

patterns.

Table 4 shows results for Couple C. The

patterns of anxiety and hostility for Couple

C are more complex than those for either

Couple A or Couple B. Not only does Cou-

ple C show significant wife-husband differ-

entials and significantly different spousal

changes between the FAT and DFAT, but

also significant contextual effects that are

congruent for the two spouses. It should be

noted, however that the contextual effects

do not achieve the same level of statistical

significance as the other differences.

Results for Couple C suggest a mixture

of interspousal symmetry, complementar-

ity, and difference. Between the FAT and

DFAT, trend level significance is observed

for hostility outward overt and total, and

for ambivalent hostility, with both spouses

showing higher levels on the DFAT. For

separation, diffuse, and total anxiety, and

for hostility inward and total, significant

interaction effects are observed. In each

case, scores for the wife increase from the

FAT to the DFAT, whereas those for the

husband decrease. In addition, the wife ex-

hibits significantly higher levels of death

and mutilation anxiety, and hostility out-

ward covert, with a trend toward higher

levels of ambivalent hostility, irrespective

of context. The husband, by contrast, shows

a significantly higher level of guilt anxiety.

The interspousal effects for diffuse and

total anxiety cannot be interpreted because

the interaction with the contextual variable

is significant in each case.

Both spouses from Couple C seem to di-

rect more hostility outward when they are

together, and to feel that more hostility is

being directed toward them. This appears

to have a more adverse impact on the wife,

since, for her, higher hostility outward

scores are associated with higher scores for

separation and diffuse anxiety, and for hos-

tility inward. The impact on the husband

seems to be salutary. Several transcontex-

tual patterns underlie these dynamics.

Generally, the husband experiences higher

levels of guilt anxiety; the wife, higher lev-

els of death and mutilation anxiety, hostil-

ity outward covert, and ambivalent hostil-

ity. The latter patterns suggest that the wife

feels she is the object of hostility, but rela-

tively powerless to do much about it.

The interpretation of the results just de-

scribed depends to a considerable degree

on inference. To provide some cross-vali-

dation, relational control patterns were ex-

amined for each of the three couples. Rela-

tional control refers to the definition of

‘‘who’s in charge” in a relationship. In the

present research, the typescripts produced

through the DFAT task were also coded

into the Ericson-Rogers Relational Coding

System.°

The Ericson-Rogers System derives from

Bateson’s’ concepts of symmetry and com-

plementarity. Symmetry refers to a similar-

ity of control, e.g., when a “‘one-up”’ mes-

74 RONALD W. MANDERSCHEID, ANNE K. MCCARRICK, AND SAM SILBERGELD

sage—an attempt to define oneself as in

charge—is countered by the other’s asser-

tion of being in charge; or when a “‘one-

down’ message—relinquishing control—

is followed by the other’s message of giving

control. Complementarity is the reverse;

One person seeks control and the other

yields it, or One person wants to be con-

trolled and the other takes charge. Ericson

and Rogers’ include a ‘“‘one-across”’ cate-

gory for messages which neither seek nor

give up control.

In the Relational Coding System, a cod-

ing unit 1s defined as each verbal interven-

tion of each member in a dialogue. Each

coding unit is viewed as being a response to

the message that preceded it, and in that

sense, a “‘definer”’ of that transaction.

After the transcripts are coded into the

Ericson-Rogers System, the resulting data

are analyzed with the log-linear procedure

from the Biomedical Data Series computer

program BMDP3F-.. Details of the method-

ology may be found in McCarrick, Man-

derscheid, Silbergeld, & McIntyre.’ This

type of anaylsis yields four pieces of infor-

mation about relational control relevant to

the present study, I.e., the most likely mes-

Sage style (one-up, one-down, or one-

across), the most likely response style, the

most likely response granted the message

style, and a determination of whether the

husband and wife differ in any of these

responses.

For the present study, the most appro-

priate log-linear model was selected, using

standard statistical criteria for this type of

analysis, and parameters were estimated

(results not shown; analysis available upon

request). Patterns of relational control be-

tween the spouses of Couples A, B, and C

show remarkable congruence with the pat-

terning of anxiety and hostility described

earlier. Note that the affect measures and

the relational control patterns are not

based on the same operational measure-

ments, although the DFAT transcripts are

the source for both measures.

Both husband and wife from Couple A

are significantly more likely to make a one-

up statement to the other spouse, and to

give a one-up response as well. The most

likely pattern of control between these

Spouses 1S competitive symmetry, 1.e., one-

up messages followed by one-up responses.

All differences described for Couple A are

Statistically significant (p < .001). Thus,

the patterning of relational control lends

support to the earlier description of Couple

A as a dyadic system in which anxiety and

criticism of self is lessened by focusing on

the shortcomings of the spouse.

The analysis for Couple B shows that

both spouses are likely to give messages in

the neutral mode, and to respond in a one-

down mode. Patterning is slightly comple-

mentary, but does not differ sufficiently

from randomness to attain statistical sig-

nificance. The wife is slightly more likely

(p < .10) to give one-down messages than

the husband. The congruence between

dyadic patterns of affect and relational

control is less marked for Couple B than

for Couple A, but some similarities are still

evident. Couple B is somewhat comple-

mentary in relational control but more

complementary in patterning of affect. The

wife is more apt to respond with one-downs

than is the husband, suggesting that she is

more likely to accept a one-down position

in the relationship.

Couple C shows a mixture of symmetry

and complementarity in patterning of af-

fect. In relational control, Couple C shows

complementary patterning, with the hus-

band much more likely to offer one-up

messages (p < .05). As noted earlier, the

patterning of affect suggests that the wife

feels she is the object of hostility, and rela-

tively powerless to do much about it. Rela-

tional control patterning suggests that she

is accurate in her perceptions, in that she is

a target of verbal “‘one-upmanship,”’ if not

hostility, by the husband. Her sense of futil-

ity is reflected in her failure to challenge the

husband’s one-ups.

Discussion

Marital and family clinicians and theo-

rists have long been interested in the pat-

ANXIETY AND HOSTILITY IN DYADIC INTERACTION 75

terning of anxiety and hostility in marital

relationships, e.g., Zinner'’ and Dicks,'!

but case studies have been the primary

source of data. Experienced marital thera-

pists are aware of the emotional under-

currents in relationships which allow per-

sons to appear healthier at the expense of a

spouse, for example, or which accentuate

conflict, or subtly influence the way spouses

treat one another. The technique used in

the current study seems to offer a promis-

ing method of extending knowledge of

marital psychodynamics by quantifying

differences in affect between individual and

dyadic contexts. Moreover, the validity of

interpretations derived from the procedure

are supported by the Ericson-Rogers Rela-

tional Coding System.

A question posed earlier was whether af-

fect scores would differ between the DFAT

and the FAT. Scores differed significantly;

anxiety and hostility inward scores were

lower and hostility outward scores higher

on the DFAT compared to the FAT. This

difference is evidence of the function of

blaming in a relationship. Many authors,

e.g., Ellis’? and Zinner,!° describe mutual

blaming patterns in couples and theorize

that blaming the other reduces self-blame

and thus anxiety. The current study pro-

vides empirical support for these assump-

tions. As mentioned earlier, one would ex-

pect variability to be greater in problematic

relationships such as those of couples seek-

ing marital therapy. The DFAT permits in-

vestigation of the degree to which mutual

blaming also occurs in satisfactory marital

relationships; there is some evidence’ that

this process occurs in non-problematic

marriages as well.

The second issue mentioned was one of

congruence; do spouses’ scores differ more

on the FAT or DFAT? Couples A and B do

show more congruence in the dyadic con-

text, but Couple C shows less. For at least

three couples, congruence seems to depend

on the dynamics of the particular dyad. Al-

though a larger sample would be required

to discern the range of possible dyadic

patterns, present findings suggest that the

range is limited and that the specific pat-

tern observed depends on the dyad and the

subset of variables under consideration.

Obviously, results are more complex than

originally anticipated.

The third issue was one of the utility of

the DFAT for clinical diagnosis and as-

sessment of marital relationships. In the

present study, the DFAT provided useful

insights into the interpersonal dynamics of

the dyads studied. The DFAT, given at in-

tervals throughout therapy, would permit

the measurement of change in the psycho-

dynamic aspects of a relationship. As

Gurman and Kniskern” note, effective as-

sessment of therapy will require measures

suited to very specific problems, goals, and

interventions, as well as global change

measures.

The DFAT could also be very useful in

the training of psychotherapists. Samples

of the interactions between patients and

therapists could be scored, and change in

anxiety and hostility levels correlated with

the different issues being considered. AIl-

though countertransference is sometimes

obvious to a trained therapist, the DFAT

would allow trainees to examine less ob-

vious instances. If the Ericson-Rogers Re-

lational Coding System was used simul-

taneously, some of the complexities of

therapy relative to power and anxiety or

hostility could be analyzed quantitatively.

For example, does control by the therapist

increase or decrease the patient’s anxiety?

In marital therapy, what happens to a ther-

apist’s anxiety level when spouses attack

each other? What kinds of therapeutic sit-

uations lead to an increase in therapist

hostility?

Finally, Gurman” argues that although

the primary measures in any evaluative ef-

fort should be congruent with the thera-

peutic approach under investigation, it

would be useful if researchers include at

least some measures specific to other types

of approaches. The DFAT is appropriate

for assessment of psychodynamic therapy,

but the procedure also permits coding of re-

lational control, which is appropriate for

structural, strategic, and behavioral ther-

apy. This combination of specificity and

versatility makes the DFAT, alone or in

combination with the Relational Control

System, a useful addition to the growing

collection of instruments with which one

may assess process as part of therapy or

outcome.

References

1. Gottschalk, L. A., 1979. The content analysis of

verbal behavior. New York: Halsted Press.

. Gottschalk, L. A. and Gleser, G. C., 1969. The

measurement of psychological states through the

content analysis of verbal behavior. Los Angeles,

CA: University of California Press.

. Gottschalk, L. A., Winget, C. A. and Gleser, G. C.,

1965. A manual for using the Gottschalk-Gleser

content analysis scales. Berkeley, CA: University

of California Press.

. Silbergeld, S. and Manderscheid, R. W., 1976.

Dyadic free association. Psychol. Rep., 39:

423-426.

. Rae, D. S., Pautler, C. P., Manderscheid, R. W.

and Silbergeld, S., 1977. Free association test

(FAT) scoring and analysis program. Behavior

Research Methods & Instrumentation, 9: 31-32.

. Ericson, P. M. and Rogers, L. E., 1973. New

procedures for analyzing relational communica-

tion. Fam. Process, 12, 245-267.

. Bateson, G., 1958. Naven (2nd ed.). Stanford, CA:

Stanford University Press.

Journal of the Washington Academy of Sciences.

Volume 73. Number 2. Pages 76-79, June 1983.

. Dixon, W. J. and Brown, M. B. (eds), 1978. BMDP-

77: Biomedical computer programs, P-Series (rev.

ed.). Les Angeles: CA: University of California

Press.

. McCarrick, A. K., Manderscheid, R. W., Silber-

geld, S. and McIntyre, J. J., 1982. Control pat-

terns in dyadic systems: Marital group psycho-

therapy as change agent. American Journal of

Family Therapy, 10, 3-14.

. Zinner, J., 1976. The implications of projective

identification for marital interaction. In: Contem-

porary marriage: Structure, dynamics, and ther-

apy. H. Grunebaum & J. Christ, eds., Boston: Lit-

tle. Brown, & Co.

. Dicks, H. V., 1967. Marital tensions: Clinical stud-

ies toward a psychological theory of interaction.

New York: Basic Books.

. Ellis, A., 1976. Neurotic interaction in marriage.

In: Handbook of marriage counseling. B. N. Ard,

Jr..& C.C. Ard, eds., Palo Alto, CA: Science and

Behavior Books.

. Whitehouse, J., 1981. The role of the initial at-

tracting quality in marriage: Virtues and vices.

Journal of Marital and Family Therapy, 61-67.

. Gurman, A. S. and Kniskern, D. P., 1981. Hand-

book of family therapy. New York: Brunner/Mazel.

. Gurman, A.S., 1978. Contemporary marital ther-

apies: A critique and comparative analysis of psy-

choanalytic, behavioral, and systems theory ap-

proaches. In: Marriage and marital therapy. T. J.

Paolino & B. S. McCrady, eds., New York:

Brunner/Mazel.

The Scientific Awards of the Academy: 1983

Sherman Ross

General Chairman, Washington Academy of Sciences

The Scientific Achievement Awards of

the Academy were presented at a meeting

on March 23, 1983 at The American Uni-

versity, Washington, D.C. Eight awards

were made for significant contributions to

research, and two awards for contributions

to science teaching. This program of the

Academy was started in 1939 to recognize

young scientists for “*. . . noteworthy dis- —

covery, accomplishment, or publication in ~

the Biological, Physical, and Engineering

Sciences.” An award for Outstanding —

Teaching was added in 1955 (renamed in ~

1979 as the Leo Schubert Award), and in

Mathematics in 1959.In 1975 theawardfor —

the Behavioral Sciences was added, as well

1983 SCIENTIFIC AWARDS 77

as the Bernice G. Lambert Award for

Teaching of High School Science.

Robert E. Davis, (Plant Virology Labora-

tory, U.S. Department of Agriculture,

Beltsville, MD.) was honored for funda-

mental contributions to the discovery

and characterization of Spiroplasmas—a

unique new class of microbes. In addition

to a number of new diagnostic approaches

mainly by optical microscopy, Dr. Davis

provided the evidence that the Spiroplasma

in infected vectors and plants is the causal

agent of corn stunt disease. He also showed

that helical morphology is a constant fea-

ture of Spiroplasmas in vitro and in vivo.

Dr. Davis received his B.S. degree in Bot-

any from the University of Rhode Island in

1961, anda Ph.D. in plant pathology-virol-

ogy from Cornell University in 1967. He

was a research associate in 1966-67 at the

Plant Virology Laboratory, and since 1967

he has been a research plant pathologist at

the Plant Virology Laboratory, Plant Pro-

tection Institute, Northeastern Region,

Beltsville Agricultural Research Center.

Eva Donaldson (Science Teacher, Dunbar

Senior High School, Washington, D.C.)

was selected for the Berenice G. Lambert

Award for the Teaching of High School

Science in recognition of her initiatives in

providing learning success and science ca-

reer awareness to her students. She has

been successful in the individualization of

the instructional program in chemistry and

physical science, which was derived from a

proposal for a mini-grant under the ESEA

(Title III) effort. Ms. Donaldson faced the

problems of poor attendance and low aca-

demic performance of students in her

classes.

Eva Donaldson is a graduate of Johnson

C. Smith University (B.S., 1948) and re-

ceived an M.S. degree from Howard Uni-

versity in 1953. She served as a Technician

in cardiovascular research at Freedman’s

Hospital (1952-53), as a chemist at NIH in

1954, and as a science teacher at Spingarn

High School from 1954-59. She has served

as a science teacher at Dunbar Senior High

School since 1962.

Robert J. Dooling (Department of Psy-

chology, University of Maryland, College

Park, MD) was recognized for his research

On comparative studies of hearing and

vocal learning. For the past several years

Dr. Dooling has been developing an animal

model for normal speech and language

development in humans. This effort is pro-

ceeding in three ways: |) an operant condi-

tioning paradigm for psychophysical stud-

ies of hearing in the parakeet, (2) computer

approaches to the study of acoustic com-

munication in animals, and 3) bird’s per-

ceptions of their own vocalizations. Using

evoked potentials and the cardiac orienting

response Dr. Dooling demonstrated an

early perceptual selectivity for conspecific

vocalizations in sparrows before any expo-

sure to conspecific song. Other parallels be-

tween human vocal learning and bird song

have been found including a sensitive pe-

riod for vocal learning, a babbling phase in

normal vocalization development, and a

left hemispheric dominance for vocal

control.

Dr. Dooling received a B.S. degree in Bi-

ology and Chemistry from Creighton Uni-

versity. He was awarded M.S. degrees (Bi-

ology and Psychology)—1969, anda Ph.D.

degree in Physiological Psychology in 1975

from St. Louis University. He held a post-

doctoral fellowship in the Behavioral

Sciences at Rockefeller University, and

served as an Assistant Professor from

1977-81. He has been at the University of

Maryland since 1981.

James W. Gentry (Department of Chem-

ical Engineering, University of Maryland,

College Park, MD) was selected for his

contributions to the classification and

measurement of non-spherical aerosols.

Dr. Gentry has made major, continuous

contributions to the study of the properties

of submicroscopic non-spherical particles,

their transport properties in gases, and

methods of separating fractions of aerosols

for detailed examination. His theoretical

and experimental work have involved the

analysis of particle size distribution from

penetration measurement, the development

of procedures for determining fiber lengths

as well as diameter from on-line penetra-

78 SHERMAN ROSS

tion measurements, the production of a

mononodisperse aerosol in the 0.03 to 0.06

um range (a size of biological importance),

and the delineation of collection mecha-

nisms for fibrous aerosols and in screen

filters.

Dr. Gentry recetveda B.S: degree’ im

Chemical Engineering in 1961 from Okla-

homa State University. His graduate study

in Chemical Engineering resulted in an

M.S. degree from the University of Birming-

ham (U.K.) in 1963, and a Ph.D. from the

University of Texas in 1969. He has served

as assistant professor (1969), associate pro-

fessor (1972), and professor of chemical

engineering since 1978 at the University of

Maryland at College Park.

James E. Hughey (Department of Chem-

istry, University of Maryland at College

Park, MD) was selected for the Leo Schu-

bert Award for Outstanding Teaching for

his contributions as a teacher and writer of

chemistry for high school to graduate

school. He is the author of the extremely

popular and successful textbook: ‘‘Jnor-

ganic Chemistry: Principles of Structure and

Reactivity,’ published in 1972. This book

and its second edition have had national

and international influence on the teaching

of chemistry. A short textbook he published

in 1978 has had a significant impact on the

teaching of high school chemistry in the

United States, Canada and Australia. He

has continued as an active researcher in

chemistry, as well as contributing to evolu-

tionary biology and to continued research

contributions on amphibians and reptiles.

Dr. Hughey received a B.S. degree from

the University of Cincinnati in 1957. His

graduate study in chemistry was at the

University of Illinois (M.S. 1959, and Ph.D.

1961). He served at the Worcester Polytech-

nic Institute as an assistant professor from

1961-65. Then he came tothe University of

Maryland, where he has been an assistant

professor (1965-68), associate professor

(1968-75), and a full professor since 1975.

Kenneth Kirk (Laboratory of Chemistry,

NIADDK, NIH) was selected for a scientific

achievement award for the synthesis and

biological evaluation of Fluorine substi-

tuted Biogenic Amines. For the past five

years Dr. Kirk has devised new techniques

for the synthesis of these amines and dem-

onstrated the biological properties of these

fluoro analogs. These compounds have

proven to be unique and useful and have

shown their potential as biological tools.

Dr. Kirk received a B.A. (Chemistry)

from the De Pauw University (1959), where

he was a Rector Scholar. In 1963 he was

awarded a Ph.D. in Organic Chemistry

from the University of Wisconsin. He spent

1963-64 at the Technische Hochschule

(Braunschweig), and 1964-65 at Cornell

University. He then became a Staff Fellow,

Research Chemist, and is now Senior Re-

search Chemist, Laboratory of Chemistry,

NIADDK, NIH.

William D. Phillips (National Bureau of

Standards, Washington, D.C.) was selected

for a scientific achievement award for his

pioneering studies of the laser cooling of

neutral atomic beams. Dr. Phillips’ contri-

butions have varied from the application of

modern technology to a classical field of

electrical metrology, to the determination

of an important fundamental constant, to

his recent work on the cooling of a neutron

beam of atoms using a counter propagating

laser beam. The goal of readily obtaining

very slow or “cold” atoms has been reached

by his research contributions.

Dr. Phillips received a B.S. in Physics

summa cum laude from Juniata College in

1970, anda Ph.D. in Physics from M.I.T. in

1976. He was a Postdoctoral Fellow at

M.I.T. from 1976-78, before his appoint-

ment at the National Bureau of Standards.

Elizabeth S. Raveche (Section on Cyto-

genetics, NIADDK, NIH) was selected for

outstanding scientific achievement for elu-

cidating the genetic basis for the spontane-

ous development of autoimmunity. Her re-

search on the genetics of autoimmunity in

mice has established classical Mendelian

approaches, and provided the basic type of

inheritance to many systems. In addition to

other contributions by her studies she has

shown that there is no autoimmunity gene,

but rather that autoimmunity arises from

the aggregate of independent abnormalities.

Dr. Raveche received a B.S. degree in

Chemistry from Seton Hill College in 1972,

and then a Ph.D. in Genetics from The

George Washington University in 1978.

She was a Postdoctoral Investigator

(1977-78), Scientist (1978-80), and since

1981 a Senior Investigator at NIADDK.

Paul B. Torrence was selected for the

Scientific Achievement Award for his con-

tributions on the involvement of 2-5A in

the antiviral actions of Interferon, and its

proposed role in cellular development or

differentiations: Chemical modification

studies have provided structural-activity

relationships, which have permitted the

synthesis of a2-5A analogue with enhanced

in vitro biological activity.

Dr. Torrence received a B.S. in Chemis-

try magna cum laude in 1965 from Geneva

College, anda Ph.D. in Chemistry from the

State University of New York at Buffalo in

1969. He served as a Staff Fellow and Senior

Staff Fellow at the Laboratory of Chemis-

try, NIAMDD, NIH from 1969-74. He has

been a research chemist in the Laboratory

Journal of the Washington Academy of Sciences.

Volume 73. Number 2. Pages 79-80. June 1983.

of Chemistry, NIADDK, NIH, since 1974.

Edward J. Wegman (Office of Naval Re-

search, Arlington, VA) was selected for the

Mathematics & Computer Sciences Award

for outstanding original and expository

work in the mathematical and statistical

theory of function estimation. Dr. Wegman

has made major contributions to mathe-

matical statistics, particularly in probabil-

ity density estimation, the relationship be-

tween spline estimation and estimation by

isotonic methods, theoretical computer

science, and recently a synthesis of func-

tional estimation procedures into a com-

mon functional optimization framework.

Dr. Wegman received a B.S. degree with

honors from St. Louis University, an M.S.

in 1967 and a Ph.D. in 1968 from the Uni-

versity of lowa. He was an assistant profes-

sor of Statistics (1968-73) and an associate

professor (1973-78) at the University of

North Carolina. In 1978 he became the D1-

rector, Statistics and Probability Program,

Office of Naval Research.

Instructions to Contributors

Type manuscripts on one side of white

bond paper. Double space all lines, includ-

ing those in abstracts, tables, legends, quoted

matter, acknowledgments, and references

cited. Number all pages consecutively.

Page | should contain the title (not to ex-

ceed 100 characters), author’s name and af-

filiation, a running title (not to exceed 70

characters) and an indication to whom cor-

respondence is to be sent. In research pa-

pers concerning biological subjects, include

an indication of the order and family of the

taxa discussed.

Page 2 should contain an abstract which

should be intelligible without reference to

the text of the paper. Write an informative

digest of the significant content and con-

clusions, not a mere description. Gener-

ally, the abstract should not exceed 3% of

the text.

Footnotes should be used sparingly. On

each page use the symbols which follow to

indicate the footnotes for that page. The

order of use should follow the order in

which the symbols are listed herein. The

same symbols may be used on separate

80 INSTRUCTIONS TO CONTRIBUTORS

pages but may not be re-used on the same

page. The footnotes must be typed on a

separate page. Please be sure to indicate

both the manuscript page number and the

symbol. The symbols are: *, Tf.

The quality of all original illustrations

must be high enough to facilitate good

offset reproduction. They should have ample

margins and be drawn on heavy stock or

fastened to stiff cardboard to prevent bend-

ing. They should be proportioned to col-

umn (1 X 3) or page (2 X 3) type-dimen-

sions. Photographs should have a glossy

finish. They reproduce best when the con-

trast is fairly high. Identify each illustration

with number and author in light pencil

marks on the unused lower or side margins.

Submit all illustrations separately—please

do not glue or clip them to the pages of the

manuscript.

Do not type or write legends directly on

the illustrations. Type legends on a sepa-

rate’ page’'or ‘pages ‘at the*'end of ‘the

manuscript.

Tables should be included only when the

same information cannot be presented eco-

nomically in the text, or when a table pre-

sents the data in a more meaningful way.

Tables should be double spaced through-

out and contain no vertical lines. The table

should be organized from top down as fol-

lows: table number (arabic numerals), title,

body and table footnotes (use the same

footnote symbols as for general footnotes.)

References should be noted in the text by

Superscript arabic numerals at the appro-

priate points. The citations should be typed

on a separate page headed ‘“‘references”’

and should be listed in numerical order.

The following illustrate the form to be

used in the list of references.

1. Coggeshall, R. E. 1967. A light and electron micro-

scope study of the central nervous system of the

leech, Hirudo medicinalis. J. Neurophysiol., 27:

229-289.

2. DeVellis, J. and G. Kukes. 1973. Regulation of glial

cell function by hormones and ions. Tex. Rep. Biol.

Med., 31: 271-293.

3. Mehler, W. R. 1966. Further notes on the center

median nucleus of Luys. In: The Thalamus. D. P.

Purpura and M. D. Yahr, eds., Columbia Univer-

sity Press, New York, pp. 109-127.

4. Tremblay, J. P., M. Colonnier and H. McLennan.

1979. An electron microscope study of synaptic

contacts in the abdominal ganglion of Aplysia cali-

fornica. J. Comp. Neurol., 188: 367-390.

Abbreviations of journal titles should

follow those listed in the Index Medicus.

Responsibility for the correctness of the

references lies with the author(s). Schedul-

ing pressures make it impossible for them

to be checked by either the Editors or the

publisher.

Send completed manuscripts and sup-

porting material to: The Editors, Journal

of the Washington Academy of Sciences,

Department of Biology, Georgetown Uni-

versity, 37th and O Streets, N.W., Washing-

fon, 19.: 20037.

Acknowledgments

The contributions of the chairmen of the

various panels and their colleagues, who

carried out the difficult task of making the

selections, are acknowledged with sincere

thanks. The chairmen were: Dr. James H.

Howard, Jr. (The Catholic University of

America)—Behavioral Sciences; Dr. C. R.

Creveling (NIADDK, NIH); Dr. John D.

Anderson, Jr. (University of Maryland at

College Park)—Engineering Sciences; Dr.

Joan Rosenblatt—(National Bureau of

Standards)—Mathematical & Computer

Sciences; Dr. Mary H. Aldridge (The

American University)—Physical Sciences;

Dr. Joseph B. Morris (Howard University )—

Teaching of Science.

Thanks are due to the nominators and to

the sponsors of all the candidates. On be-

half of the Academy we commend the re-

cipients, whose work is honored, and we

wish them continued productive careers.

DELEGATES TO THE WASHINGTON ACADEMY OF SCIENCES,

REPRESENTING THE LOCAL AFFILIATED SOCIETIES

IE NISEURTICIIOITD gcc cc uin's Siu cc un tebe tesnebstwesvwcsenctedbastwneye James F. Goff

EEMEEY OF WASIITIDION . 6... ccc ccc cence scvvceuentdespiuvscvceve Ruth H. Landman

EE MINIDTON oes cca sw ec ck vieveccaveedusenssaantentvvese William Ronald Heyer

TY CG cn cu fas Wa vnc o wieln 6a sen wee apiece s viuwudiaccgciainbes Anne Brown

ee WRSHINGTON 5.05 cies ics skecewcavaescuewbouresaudceununcenwes Margaret Collins

Ee oso ar 8a anid wich y aie Vein bd 8 dees da 4 Nee wea wee ee T. Dale Stewart

ESIC co ac vce ewe h ect e dave ppewesvevesescuntsetse’s James V. O'Connor

rn ome PuStrict Of Columbia........cccccsccccssevccecunscssecus Charles E. Townsend

digas Gin nis own wel & SYN. Wem, us oib,0: 0 VS we io a Meare dlwaleieies Paul H. Oehser

ee ae Conrad B. Link

IEEE oi nk vs Se sn eviews eae e nce db whe oidew ne vivis eee ew Wee we Boyd W. Post

TE co 8 ga x ew (aw le o's 0 arate Ale ma ee brats bie aimatn le nig Kaede bail George Abraham

eertcren! and BICCIrOnics ENPINCErDs ...... 2.66060 screen ccececsncecceaes George Abraham

in MECHANICA ENPINCETS .... 1... 20.0 cee veanecnwunscnnnevecnwenveseuns Michael Chi

reeset Gt WHASHINPION 0.3 << << oc oes cie cee pcan ee nee selene vad nae Robert S. Isenstein

I ICT ENNIOIONY ne hc cee ec edeseu na uns Sbeleeeueeailew aus Lloyd G. Herman

Cen MMIEATY ENIPINGETS.. 0.60. keene crac eaeucnetcaewrdeene H. P. Demuth

MERE DUE PPINCETS . .. . 5. one ec we ce ee ee eee eee eee ese ee Wallace J. Cohen

ee emeeiimenta! Biology and Medicine ...............0s00seccnneduceecees Cyrus R. Creveling

EE EMMEE AIS ono oie ose soe eae cee we wae eee keea skew eaweneee es Charles G. Interrante

eneeeimncm GO! Dental Research ..... 2.2.60. ccc cece cc en es eee teemscenes William R. Cotton

eee tnistituic Of Acronautics and Astronautics. .... cai ee cece ec e eens Richard P. Hallion

I ENG se 2 Fone fe Se, 6 ein ae eb Wea bY Oe xine BUR gee ae we A. James Wagner

I ASINITIDC OH. ooo oc o's coins cio. dic cased re ee ete ds seen vwad atic se waleea Jack R. Plimmer

I UETTOEN Go oscar oa oa whe sin dn in ca wae s bee eon cece u helnenn et pi mene Richard K. Cook

EE yet oa Gedo d sce he bg os ee aw D hs Mes no oe ara mud ed aie am RR Dick Duffey

EEE 2 ik ga is picks A sles vols oa saws sede a ee nice mane eeaes A. D. Berneking

EI IN oi 5 ia a loa sn ne was oles Sew sea dew a cede a eweme Edwin R. Fuller, Jr.

EE ec eo See cis om wield waive Mae diaee soem dld wae ome wenn Alayne A. Adams

Ee INCNE PE RINSIES PEACHEES 2.2 3 on oosce, ovcie Wp ws 2 wa ee So's peso dee ee waco Rn Peggy A. Dixon

EM ha Sood sl open cod owicta ws wk RSX BAe eee OR em oe George J. Simonis

rE eee TAME EMYSIOIOPISIS |... - 0.20 ec ee ce cite ene bo selewe ems s eee Walter Shropshire, Jr.

MER CCITE RC MPISICII 2.5 5c eis eds ois sien ce ws nek We ewe ese ene wee John G. Honig

REN aL pieignle arise o-9,n'e ule wxlaie ee Rae RG duns cd owe alerd Jewel B. Barlow

American Institute of Mining, Metallurgical

I MP A Ne Shae) ateiava'alwin 08 ea Store ad Mp ale am eee cd amen Garrett R. Hyde

SME kg gs A iin! wina. na DMs em oS silo der ol A lnve'e oie 6 ore a Benson J. Simon

ANI NUNS NE PUSTE NEA 0 dis.c sou sis iw af aie oe ale wie hes abn Wee Geen da nls e pak dalton Patrick Hayes

MNS eres cole wend cv Oo daeed od swans deine then ews Miloslav Racheigl, Jr.

PRE SERS sg fat 9's. en 0: wok, ots ocn ws ip in carve: «| simi, wma H. N. Reynolds

nn Sette WCCMMICAT USTOUD 2. 6. cede es funnest nen ances cnvesntenen a Paul G. Campbell

RTA TISMCICEED EOEIESY (oo. iy ial o ole diane visne weed akeue ne sud egies Howard E. Waterworth

RESIN SUSTCMIS RRESCAICH ci. cc csc cc cnc cb cce sp eedesaceneser Ronald W. Manderscheid

Re ne DR ca dg wiawn Kia gate wikig ales a walulen'e eae Stanley Deutsch

NE A RA hora 1, ali g0) x wh clpia/u di etiviw a didi n etd wie. viel (od ales Wie vida bl Irwin M. Alperin

me tar scieice, Lechnology and Innovation .........0.ccecscsercsesacsccesnns Ralph I. Cole

AE RM ta ob Caideetcn o V's ohne ben sees anndasb vice’ Ronald W. Manderscheid

Delegates continue in office until new selections are made by the representative societies.

2nd Class Post:

Washington Academy of Sciences

1101 N. Highland St.

Arlington, Va. 22201

Return Requested with Form 3579

ay *

‘a

W317 VOLUME 73

NH Number 3

al of the September, 1983

WASHINGTON

ACADEMY .- SCIENCES

ISSN 0043-0439

Issued Quarterly

at Washington, D.C.

A HSON lAn

o ;

D ’ re

CONTENTS

Articles

AT eONry J. SEEGER: On “A Patron for Pure Science” ............--

FRANK R. YEKOVICH, CAROL H. WALKER. and PAUL K. DUNAY: The

Use of Scripts in the Study of Knowledge-Based Comprehension of

teal (ake tat le ee ee oe) ee eee we ale a wld eae 2 as os he) ee Gere a6) Ss Bw Melee ee we WS

W.W.CANTELO and R. E. WEBB: An Examination of Selected Companion

Plant Combinations, and How Such Systems Might Operate ..........

east OF Members 2... 2... one tance eects ce dnmeeeeeensene

Washington Academy of Sciences

Founded in 1898

EXECUTIVE COMMITTEE

President

Jean K. Boek

President-Elect

Ralph I. Cole

Secretary

Elise A. Brown

Treasurer

Ronald Manderscheid

Managers-at-Large

Jo-Anne Jackson

Grover C. Sherlin, Office Manager

Alfred Weissler

Howard E. Sorrows

James Wagner

Simon W. Strauss

BOARD OF MANAGERS

All delegates of affiliated

Societies (see Cover 3)

EDITORS

Irving Gray

Joseph Neale

ACADEMY OFFICE

1101 N. Highland St.

Arlington, Va. 22201

Telephone: (703) 276-9254

The Journal

This journal, the official organ of the Washington _

Academy of Sciences, publishes historical articles, critical

reviews, and scholarly scientific articles; proceedings of

meetings of the Academy and its Board of Managers; and

other items of interest to Academy members. The Journal

appears four times a year (March, June, September,and

December)—the September issue contains a directory of

the Academy membership.

Subscription Rates

Members, fellows, and patrons in good standing receive

the Journal without charge. Subscriptions are available on

a calendar year basis only, payable in advance. Payment

must be made in U.S. currency at the following rates:

U.S. and ‘Canada. os... $19.00

Foreign (2. choco 22.00

Single Copy ‘Pree’ <.c2< 7.50

UHURU Symposium—2 issues............ $15.00

Delivery by Alf ...s2.% 2.00

Single-copy price for Vol. 66, No. 1 (March, 1976) is $7.50.

Back Issues

Obtainable from the Academy office (address at bottom E

of opposite column): Proceedings: Vols. 1-13 (1898-1910)

Index: To Vols. 1-13 of the Proceedings and Vols. 1-40 of

the Journal Journal: Back issues, volumes, and sets (Vols. _

1-62, 1911-1972) and all current issues.

Claims for Missing Numbers

Claims will not be allowed if received more than 60 days

after date of mailing plus time normally required for

postal delivery and claim. No claims will be allowed

because of failure to notify the Academy of a change in —

address.

Change of Address

Address changes should be sent promptly to the Academy

office. Such notification should show both old and new ~

addresses and zip number.

Published quarterly in March, June, September, and December of each year by the Washington

Academy of Sciences, 1101 N. Highland St., Arlington, Va. 22201. Second class postage paid at

Arlington, Va. and additional mailing offices.

Journal of the Washington Academy of Sciences,

Volume 73, Number 3, Pages 81-87, September 1983

On “A Patron for Pure Science’

Raymond J. Seeger

National Science Foundation (retired)

At dinner one evening my teenage

daughter asked, ““Daddy, why don’t you

ever talk about science instead of the office?”

I feel the same way about the book “‘A Pa-

tron for Pure Science,” “‘The National

Science Foundation’s Formative Years,

1945-1957” (NSF 1982) by J. Merton Eng-

land, NSF historian. I would have liked to

learn about NSF’s contributions to science

per Se, say, its contributions to the work of

Nobel Laureates—particularly before the

awards—for example, the grant of $25,000

for Donald E. Glaser in the early stage of

the development of the bubble chamber.

Later, the Chemistry Program made a

much greater contribution for the low-

temperature research of William F. Giauque

(a transfer of ONR responsibility). More

imaginative, but less productive, was a

grant given for the pursuit of the phantom

solitary magnetic pole, predicted by P.A.M.

Dirac.

This history of ‘The National Sciences

Foundation’s 1957,” is based largely upon

many cited documents. The author, as he

wrote me sometime ago, Is strongly averse

to interviews, which he regards ‘‘as usually

fallible and self-seeking.’’ Written records,

on the other hand, do not always contain

“the truth, the whole truth, and nothing

but the truth.”’ The author admits that he

did have some portions reviewed in draft

form; it would have been preferable to ob-

tain beforehand all possible viewpoints

rather than to prejudge their potential

value. (It is noticeable that more attention

is given to subjects of interest to the review-

ers.) I must confess I am always bothered

by a common practice of historians to in-

ject their own opinions by the casual use of

words such as “‘may, might, seems, appar-

ently, presumably, implicit, no doubt, evi-

dently, clearly, obviously,” et al.

‘Part I—The Long Debate, 1945-1950”

together with Chapter 15 on ‘“‘Policy for

Science’’ comprises about one-third of the

book. They give a very detailed (day by

day, blow-by-blow), but tedious presenta-

tion of interest primarily for their social as-

pects. In view of the author’s prevalent use

of statistical analysis it is regrettable that he

did not ascertain the academic background

of each Congressman involved—particularly

with respect to science.

I am surprised that he makes no refer-

ence to the unique use of the word Founda-

tion for a government agency. In this case it

connotes the legality of a public adminis-

tration to receive private funds—rarely

done. Because of its historical significance

the ““NSF Act of 1950” should have been

included in the Appendix to complement

‘““Executive Order (10521) Concerning

Government Scientific Research.”’ The au-

thor’s relating the provocative ““Science—

the Endless Frontier’ to what he calls

‘**America’s dominant myth—the frontier”

(F. G. Turner 1920) hardly honors Van-

nevar Bush, “‘the founding father of NSF.”

There is careful avoidance of any discus-

82 RAYMOND J. SEEGER

sion of the overlapping charges for gov-

ernmental advice on science given to the

National Academy of Sciences during the

Civil War, to the National Research Coun-

cilin World War I, and to NSF at the close

of World War II—not to mention conflicts

of interest of the very people involved. The

author has a tendency to elaborate in detail

preparatory moves towards a policy and

then to ignore its implementation, e.g.,

NSF’s housing the Interdepartmental Com-

mittee on Scientific Research and Devel-

opment, where he does not even mention

the. Executive Secretary Lawson M.

McKenzie, former head of ONR physics.

He neglects the worrisome problem faced

initially by the statutory Divisional Com-

mittees, which functioned between operat-

ing Advisory Panels and the overseeing Na-

tional Science Board.

Organizational problems between NSF

Divisions are treated lightly. There is some

discussion of those arising from the com-

plete separation of research and education,

not customary in academic practice.

Whereas the quarrels of the Biological and

Medical Sciences with Scientific Personnel

and Education are discussed, the good rela-

tions of the Mathematical, Physical, and

Engineering Sciences are ignored. To be

sure, in the latter case there were rarely any

formal consultations, but the staffs enjoyed

informal fellowship. MPE, indeed, stressed

the linkage ‘‘and”’ in research and educa-

tion—universities emphasizing research,

colleges teaching,—both public relations.

No mention is made either of the unusual

MPE conferences on undergraduate courses

from a research point of view.

NSF, I believe, is truly more than just

Presidential appointees; rather it is the to-

tality of these together with the entire staff,

the Divisional Committees, the Advisory

Panels, and the community of concerned

scientists. The author’s viewpoint is mostly

that of management. He is more impressed

with the social classification of people than

with their personal characteristics. For ex-

ample, he describes Father Patrick H. Yan-

cey, a member of the first NSB, as a white,

male, Catholic biologist from a small,

southern college, whereas President Tru-

man had actually appointed him primarily

for his unusual stimulation of youth in

science. He pays too little attention to staff

members who played an especially signifi-

cant role in the formative years—until

bureaucratic hardening of the arteries set

in. The book emphasizes skeletal features

of NSF, but it is the staff that gives life.

There is a good discussion of three man-

dates of the Act that had to be faced by the

first Director, the physicist Alan T. Water-

man, viz., (1) a national policy for basic re-

search and education in the sciences, (2)

appraisal of the impact of research upon

the general welfare, including industrial

development, (3) evaluation of science in

the Federal Government. The controver-

sial responsibility for military research re-

quested by the Department of Defense is

discussed, as well as NSF’s avoidance of

medical research, comprehended under the

successful National Institute of Health.

Agricultural research is ignored.

The slow process for recognition of the

social sciences is well presented—particu-

larly the political maneuverings. Regardless

of their national importance, however, the

author fails to recognize the valid concern

of natural scientists for the difficulties in-

herent in the methodology of many social

sciences (political science is still beyond the

cognizance of NSF). Although their nu-

merous variables present no greater com-

plexity than those in statistical physics or in

fluid dynamics (weather phenomena), their

interrelatedness is of such a magnitude that

the powerful method of successive approxi-

mations is not practicable so that one can-

not isolate major characteristics and thus

make controlled experimentation possible.

The author regards the establishment of a

Sociophysical Sciences Program in MPE as

merely a ploy of social scientists in NSF.

He does not appreciate that natural sci-

ences may have breadth as well as depth.

He makes no mention of notable MPE

grants such as the one to the American

Academy of Arts and Sciences for Philipp

G. Frank’s ‘‘Reasons for the Acceptance of

Scientific Theories.”

A PATRON FOR PURE SCIENCE 83

My own NSF experience is related to

“Part II—Beginning, 1950-1954” and ‘Part

I1I—Cold War Growth, 1954-1957.” It

was my good fortune to be appointed the

first MPE Program Director by Waterman,

who had introduced me at Yale to the fas-

cinating theory of metals and who had a

wholesome academic view of research and

education. (Actually, | was Program Direc-

tor of Physics and Astronomy, initially a

single NSF program.) I was fortunate also

to serve for a short while under Paul E.

Klopsteg, first Assistant Director for MPE,

who had many broad interests and profes-

sional contacts. All Program Directors had

freedom for creativity in this exciting new

venture. Unfortunately, the long delay later

in the appointment of a permanent MPE

Assistant Director encouraged a lack of

confidence in the too obliging acting one so

that a few aggressive subordinates took ad-

vantage of his apparent lack of authority.

What was truly disappointing was the se-

duction of most staff members away from

scholarly pursuits; a scientific publication

was a rarity.

Obsessed with administrative channels,

the author fails to appreciate the role of

junior staff; he does not realize that much

science policy at NSF was formulated de

facto. For example, the separation of phys-

ics from astronomy was not the result of an

order from the Director or of a demand

from the astronomers. It was initiated by

the physics staff, who believed that astron-

omy would develop better at NSF if it were

a separate program. Dr. England passes

over also some other apparently minor

changes that were actually highly signifi-

cant, e.g., change in the name of Engineer-

ing to Engineering Sciences (to emphasize

Scientific aspects) and in the name of

Mathematics to Mathematical Sciences (to

distinguish it from its traditional humanis-

tic role).

Although the book is not concerned

much with staffing, it does contain a mis-

Statement in this regard; it accuses MPE of

not having any women scientists. On the

contrary, it had one of the first research

Program Directors, namely, Helen Hogg, a

distinguished American-trained Canadian

astronomer. In addition, MPE was greatly

helped by part-time persons such as profes-

sors Dorothy Weeks of Wilson College and

Marguerite Wrisley of Randolph Macon

for Women. The advantage of permanent

staff members was continuity—and memory;

the advantage of rotational ones was the in-

jection of new approaches and fresh ideas

from academia. Everyone, of course, bene-

fited from the national outlook and con-

tacts afforded by NSF. The modest Walter

R. Kirner, first Program Director for Chem-

istry, tells of his embarrassment upon being

hailed by passersby at annual meetings of

the ACS, while his companion Roger

Adams, ACS President, was ignored. Pro-

gram Directors were often regarded exter-

nally as ambassadors at court so that occa-

sionally they considered their first loyalty

to their discipline. In some instances NSF

was considered merely an academic step-

ping stone.

Initially, all MPE Program Directors

met regularly to discuss democratically

issues and proposals. Eventually this free-

dom was exploited by some who made the

meeting a forum to insure their getting a

‘fair’ share of funds—hence divisiveness

set in. The author stresses their loud com-

plaints, but ignores the silent cooperative-

ness of the majority. Even the MPE Divi-

sional Committee could not find a simple

solution for the Division’s request for “‘a

rational distribution of funds.’’ (Such con-

cerns became less critical as NSF’s budgets

increased. )

The physical sciences were regarded by

some at NSF as having an understanding

group in top management. On the other

hand, as the author notes generally, Pro-

gram Directors, particularly in MPE, were

always vulnerable to the temptation of

those in authority to play the role of Pro-

gram Director.

The furtherance of astronomy was espe-

cially evident in the staff’s creation of a

separate panel for the new radioastronomy

inasmuch as the one for astronomy, the

first Advisory Panel in MPE, was interested

primarily in optical astronomy. At the first

84 RAYMOND J. SEEGER

meeting of the latter an excellent radio-

astronomy proposal from Bart Bok (Har-

vard) had been rejected because three tech-

nical prerequisites were needed. When these

improvements were promptly made, the

Program Director immediately recom-

mended a grant—much to the consterna-

tion of the Panel which expected to recon-

sider it at the next meeting a year later.

There is a good discussion of the Associ-

ation of Universities for Research in As-

tronomy owing to the very good compre-

hensive interviews (not cited) by Frank K.

Edmondson, an early Program Director

for Astronomy. Little attention, however,

is paid to the careful survey of all possible

sites by the Robert R. McMath Committee,

which selected Kitt Peak. The discussion of

Associated Universities, Inc. reflects accu-

rately the complexity of the somewhat bit-

ter struggle for the control of Green Bank.

In neither case is there adequate recogni-

tion of the role of the staff. For example,

the solution for the need of AURA to be

both under the jurisdiction of top academic

observatories and at the same time to be re-

sponsive to the needs of less prominent

schools was actually suggested informally

to Leo Goldberg by a MPE staff member;

viz., some at-large representatives on the

AURA Board—finally done.

Whether intentional or not, the author

shows a preference for illustrative mate-

rials about Divisions other than MPE. For

example, in discussing early research grants

he restricts himself to those in BMS and in

presenting the work of a typical Program

Director he selected one from BMS. He

casually dismisses two highly significant

MPE facilities programs with the comment

that they would be too “‘tedious”’ to exam-

ine. Both had been initiated by physics staff

through their close relationship with John

von Neumann, AEC Commissioner. At

that time high-speed computers were a nov-

elty on a few academic campuses. NSF of-

fered to supply them for general scientific

research. An applicant had to indicate its

probable use, including conditions for

availability. (A major university reported

that its psychologist had an urgent need for

tables of v = s/t.) In the case of campus

reactors for popular nuclear research the

AEC agreed to sponsor such research if

NSF provided the facility. These actions

exemplify the excellent relations between

MPE staff and their colleagues in other

federal agencies (program directors would

consult one another about similar pro-

posals submitted simultaneously to several

agencies—for insurance).

MPE did have some problems peculiar

to its disciplines; MPE fields were already

well sponsored by DOD and AEC. (Water-

man had regarded the Office of Naval Re-

search as a temporary surrogate until the

establishment of NSF. Initially, MPE had

no staff members from ONR—different

from BMS and SPE.) What was left for

MPE to do? The practical strategy, ignored

by the author, was to sponsor exploratory

conferences. Three questions were always

asked: (1) the status of research, (2) the

chief research needs, (3) the role of NSF, if

any, in meeting these needs. Thus policy

planning was actually performed at the ac-

tion level. For example, MPE was able to

make a unique contribution by sponsoring

an annual conference on high-energy re-

search under the able leadership of Robert

E. Marshak (AEC could not do so if for-

eigners were to attend). There was also an

Advisory Panel on High-Energy under the

knowledgeable chairman Leland J. Haw-

roth. (NSF was already assisting accelera-

tor-design studies by the Midwestern Uni-

versities Research Association—discussed

well in the book.) At one of its meetings a

staff member mentioned in a conversation

with Isidor I. Rabi the desirability of a con-

ference at which the USA and USSR could

jointly discuss their research progress in

nuclear physics for the first time as part of

President Eisenhower’s “‘Atoms for Peace”’.

This idea was relayed to the Director, who

requested a formal memorandum. The ul-

timate Geneva Conference of international

importance has been ignored by the author.

Another omission is the unique MPE

conferences sponsored at Liberal Arts Col-

leges to encourage research practicable at

such institutions. The first one, on low-

A PATRON FOR PURE SCIENCE 85

temperature physics, was held at Amherst

College under the excellent leadership of J.

Howard McMillen, first NSF Program Di-

rector for Physics. The second, on astron-

omy, was at Swarthmore College under

Peter van de Kamp, first Program Director

for Astronomy, himself an excellent exam-

ple of how to conduct such research. Chem-

istry was considered at Washington and

Lee, Earth Sciences at Beloit, Mathematics

at Colby. Colleges were generally advised

not to strive to compete with large labora-

tories; of course, one could associate with

those nearby. Rather, one should leisurely

cultivate neglected fields, e.g., shockwave

phenomena, where excellent research train-

ing is afforded. An unusual MPE confer-

ence was that at the American Philosophi-

cal Society on the “History, Philosophy,

and Sociology of Science,” not mentioned

by the author, who does not recognize such

interests as particularly germane to the

physical sciences despite their broad and

long history. Accordingly, he does not

mention a grant such as that for Leon Bril-

louin to complete his research-oriented

book on “Science and Information Theory”

(1956).

As other groups in NSF, MPE was

blessed with the enthusias’‘c interest and

responsible cooperation of countless

Scientists in the written reviews of pro-

posals. (I recall a hand-written one on hotel

Stationary by the busy Hans Bethe.) (The

book is sadly lacking also in discussion of

the role of scientific societies in assisting the

young NSF.) The methods of peer review

differed widely from program to program:

e.g., careful consideration of each proposal

by the Chemistry Panel to deliberation

about only major issues and critical prob-

lems by the Physics Panel. In an Appendix

the author gives an interview (his only one)

with William E. Benson, long-time Pro-

gram Director for Earth Sciences, about

procedures in his discipline—well presented.

It is regrettable that the author did not have

such interviews with other Program Direc-

tors so as to exhibit the spectrum of ap-

proaches. Reviewers included often refer-

ences to other relevant research, sometimes

unpublished, and even suggestions for

further research. Regardless of whether or

not a grant was made, all such comments

were transmitted anonymously to proposers

by MPE staff—called ‘‘good Samaritan

letters.”

Neglected fields, however, presented a

formidable problem, viz., to ascertain quali-

fied reviewers. I recall a proposal on bub-

bles—the best presented one I ever read.

On the other hand, NSF received proposals

with the curt message, ‘“‘Have need. Send

funds!”’ Reviewers, too, sometimes had to

be reviewed. An unforgettable case was the

one who himself submitted the same pro-

posal he had rejected the year previously.

Another proposal was reviewed by 33 physi-

cists and astronomers, including Einstein

and some foreigners. The proposer had in-

formed his Board (24 Congressmen) that

the potentiality of the research was greater

even than that of atomic energy; it dealt

with the possible screening of gravitation.

Two persons, including Wm. F. G. Swann,

former APS President, regarded it as worthy

of further investigation. Unfortunately, the

cost was orders of magnitude greater than

the whole MPE budget—even for the next

requisite data. (NSF was often regarded in

those day as meaning Not Sufficient Funds).

Now and then a single review would be out

of line with others. Fortunately, a reviewer

was always asked to indicate other experts.

In such cases these would be asked to re-

view the questionable proposal. In one in-

stance an MIT professor had rated a pro-

posal at the bottom, whereas three others

had given it the highest rating. Finally there

were nine top reviews. Upon meeting the

professor a year later, I learned that he

thought he had a better way of doing the

Same research.

One other significant difference between

BMS and MPE not noted by the author!

The BMS programs were not related to

traditional academic subjects such as bot-

any whereas the MPE ones were, e.g., phys-

ics. The BMS classification facilitates in-

terdisciplinary research, which was rarely

handled well in MPE. On the other hand,

MPE had the direct support—and pres-

86 RAYMOND J. SEEGER

sures—from the respective professional

societies. The engineers, for instance, kept

pushing for independence, often claiming

that physical scientists did not understand

engineering. Mathematicians and earth

scientists insisted that any distribution of

funds—obviously inadequate for every

program—was “‘unfair.”’ The author makes

no mention of how a Program Director

might solicit proposals to inflate the appar-

ent demand. The very availability of funds

for any particular purpose stimulates pro-

posals for them. It was difficult, moreover,

for any Program Director to ascertain the

support of his field by other government

agencies owing to their peculiar classifica-

tions; for example, NIH included all chem-

istry under biochemistry, ONR insisted

that all mathematics is potentially pro-

grammatic, the Office of Ordnance Research

(Army) claimed at one time that it had no

so-called basic research contracts, etc.

In view of continual Congressional con-

cern for justification of research expendi-

tures, it is surprising that the author fails to

mention MPE’s attempt to evaluate com-

pleted research. The investigation was made

in the case of the best managed MPE Pro-

gram, i.e., Chemistry. Kirner selected 5

grants in physical chemistry and 5 in or-

ganic chemistry. A chemistry staff member

and I visited each grantee to inquire about

specific research accomplishments. The re-

sults were disappointingly nebulous. The

grantee was asked also if he would be will-

ing to evaluate research results of other

grantees in his field; he was generally reluc-

tant to do so for fear of negative reactions

on reviews of his own proposals. Amazing

that research can be judged beforehand but

not afterwards! Proposals are frequently

being judged on ‘‘Who, where’’—not on

Twihtate?

The author fails to note the significant

importance of research grants for training

graduate students. Their financial support

allows less favored institutions to attract

some that cannot be absorbed by elite re-

search centers.

The author does not mention a major

difference in the approaches of MPE and of

SPE. In general, SPE (except in cases of na-

tional importance like PSSC) would make

public announcements inviting applicants.

MPE went further; it actually visited insti-

tutions, large and small, to encourage them

to apply. It sought out those geographically

isolated and those for minority groups

(women, blacks, religious colleges). For in-

stance, I myself would address a whole

science faculty and then meet with inter-

ested individuals. I urged the University of

Nevada to submit at least one proposal

(NSF had none at all from the entire state).

(All other factors being comparable, geog-

raphy was considered.) I assured mathema-

ticians in the southwest that they would be

given consideration, as well as those in the

northeast. The message was clear: “Only

one grant, to be sure, for every four pro-

posals; but no proposal meant certainly no

grant!’’ I was, however, shocked to learn

that some institutions were actually using

NSF to evaluate their own faculty—no

grant was wrongly interpreted as neces-

sarily a lack of scientific competence. The

author shows no interest in such promo-

tional or field activities—of paramount

importance in the “‘formative years.” In-

cidentally, there is no mention either of the

excellent Visiting Scientist Program of SPE.

(One physics staff member actually partici-

pated in it.)

Despite the many lacunae discussed criti-

cally above, the author is to be congratu-

lated upon producing a readable account

based upon the digestion of a mass of rec-

ords. One wishes, however, that he had

started with 1940 to continue the excellent

work of the historian A. Hunter Dupree,

‘Science in the Federal Government”

(1957), done under a NSF grant to the

American Academy of Arts and Sciences.

It had the merit of having an Advisory

Committee of eminent historians under the

chairmanship of I. Bernard Cohen. No

mention is made of the later grant made to

the University of California, Berkeley, for

Dupree to continue his study from 1940-

1960; it was to be done under an internal

NSF Advisory Committee headed by Lee

Anna Embrey. Although four preliminary

publications (not cited) were made, Dupree

had to forego completion of the project

Owing to personal problems. | do believe,

however, that an external advisory com-

mittee is always advisable to forestall the

usual criticisms of in-house publications

being merely justification of past perform-

ance. Surprisingly, there is no reference

either to the book on ‘“‘The National Science

Foundation” (1969) by Dorothy Schaffter

of the Legislative Reference Service of the

Library of Congress (1945-1962).

The author intimates that the controver-

sial issue of so-called “block grants” will be

a major concern of Waterman’s second

term—to be discussed in a second volume.

Journal of the Washington Academy of Sciences,

Volume 73, Number 3, Pages 87-99, September 1983

He does describe the grant to the National

Academy of Sciences for the Pacific Science

Board as leading in this direction. He neg-

lects, however, to explain the rejection of a

similar informal proposal from Sigma Xi

(it proposed matching funds, without any

administrative charges to the government

for small (under $2000) research grants to

individuals—the administrative cost of gov-

ernmental processing was a major handi-

cap). Some MPE staff, agreeing with the

Director at that time, were loth to see any

such simplified type of institutional grant

replace that for creative research by an

individual.

The Use of Scripts in the Study of Knowledge-Based

Comprehension of Text

Frank R. Yekovich

The Catholic University of America

Carol H. Walker

The Catholic University of America

Paul K. Dunay

University of Notre Dame

Most current models of reading assume

that text understanding is an interactive

process.’ ° At least three aspects of this

process are free to vary: the reader’s goals,

the reader’s knowledge base, and the char-

acteristics of the presented text. In the last

decade, significant progress has been made

in determining the impact of various text

characteristics on text comprehensibility.

A number of macro- and micro-structural

variables have been specified”*® and recent

models of text comprehension have incor-

porated various combinations of these vari-

ables into their predictions about the com-

prehensibility of particular texts.**? In

addition, computer simulations of these

88 FRANK R. YEKOVICH

10,11 :

human performance models are being

used to examine the contributions of vari-

ous text components to the reading proc-

ess. Examples of text variables used in these

simulations include word length, word fre-

quency, word familiarity, number of words

per sentence, number of propositions per

sentence, and number of reinstatements.

In contrast to the advances that have

been made in specifying the contribution of

text components, less is known about how

readers’ goals and readers’ world knowl-

edge affect the reading process.* Conse-

quently, while the reading comprehension

theorists acknowledge the importance of

readers’ goals and general world knowl-

edge, their models do not specify the details

of these processes. Our immediate goal and

the purpose of this paper is to describe

some of our preliminary work on the topic

of knowledge activation during reading.

Our long range goal is to develop a process-

ing model of text comprehension that in-

corporates knowledge-based components

of reading comprehension into an overall

theoretical framework.

The paper is organized into three sec-

tions. In the first section, we discuss the

notion of scripts and explain why they are

appropriate in the study of knowledge acti-

vation during text comprehension. In addi-

tion, we summarize the research on scripts

in text comprehension and memory. In sec-

tion two, we present the results of a pilot

study which investigated the effect of scripts

in immediate memory. In the third section

we discuss the relationship between repre-

sentational and processing issues. We will

also consider alternative methodologies for

studying knowledge-based comprehension

of text.

Scripts in Text Comprehension and Memory

The concept of scripts originated in the

field of artificial intelligence.'* According

to the AI conceptualization, scripts are se-

* Although the Center for the Study of Reading has

been working on these problems, no comprehensive

theory has emerged.

quences of causal chains that organize in-

formation about highly structured and

conventional events or action sequences.

Such stereotypical event sequences as going

to a restaurant, going to the doctor, and

going to a birthday party are often cited as

examples of scripted activites. The basic

idea is that while specific values (informa-

tion) within a script may vary from one res-

taurant encounter to another (e.g., restau-

rant location, details about the waiter,

quality of the food), the basic underlying

goal structure and the relations among the

generic components stay the same. Thus, a

typical sequence in a typical restaurant (ex-

cluding a fast-food chain) would include

entering, sitting down, ordering, eating,

paying, and leaving.

Scripts are assumed to exist in memory

because they are an economical way to

store information about action and event

sequences that tend to be encountered

again and again. It is not known exactly

how scripted information is represented in

memory. Anderson’’ recently proposed that

scripts are complex, schematically struc-

tured, cognitive units that are encoded and

retrieved in a unitized, “‘all-or-none”’ fash-

ion. Schank,'* on the other hand, has

argued that scripts are not stored and acti-

vated as precompiled chunks. Instead,

Schank maintains that “‘scenes”’ or even

parts of scenes are activated as needed,

based upon the characteristics of the input

text. Experiments by Bower, Black, and

Turner’’ have raised other interesting ques-

tions about the representation of scripts in

memory. It is not known, for example, at

what level of abstraction scripted informa-

tion is stored (e.g., visits to a particular

doctor; visits to doctors in general; visits to

health professionals in general, etc.) Another

question concerns the storage of script

components such as props and characters.

Is information about waiters stored as part

of a restaurant script or does a restaurant

script only contain “‘bare bones”’ that can

be instantiated with information about

waiters as needed? We will say more about

these and other script issues later in the

paper. We mention these questions now

KNOWLEDGE-BASED COMPREHENSION 89

primarily to point up the relation between

script representation and script activation.

More detailed information about the repre-

sentation of scripts will help us to specify

how scripts are activated and used during

reading.

Using Scripts to Study Knowledge Activation

We chose to use scripts in our study of

knowledge activation for four reasons.

First, it has been assumed that scripts are

activated and used during the comprehen-

sion of script-based texts.”'*'® While most

of the evidence pertaining to scripts has

been collected in memory experiments,

some comprehension time studies have

shown that stored scripts can facilitate

comprehension.'*'*'” Thus, at least initial

findings suggest that scripts would be an

appropriate knowledge base for our

purposes.

The second reason we chose scripts for

our study of knowledge activation is be-

cause scripts have hierarchical and sequen-

tial properties that can be specified using

conventional text analysis procedures.

Meyer’ has argued that the problem of

specifying variables on which text passages

are similar and different is crucial for re-

search on reading or learning from text.

Unless text variables are specified, results

obtained from one passage cannot be gen-

eralized to another passage. It would seem

that this argument would apply to the

study of knowledge structures as well.

Hence, scripts are attractive for our pur-

poses because the hierarchical and sequen-

tial properties of scripts as representational

structures have been well specified in Al,

and representational structures with well-

defined hierarchical and sequential proper-

ties can be analyzed using procedures that

have already been developed by text analysts.

The third reason we chose scripts is be-

cause they are conventional. By definition,

scripts are assumed to share certain stereo-

typic qualities across people. Thus, the

amount of person-to-person variation with

respect to scripted knowledge is relatively

low. While certain idiosyncratic variations

exist among readers, studies in script gen-

eration have shown a high degree of agree-

ment among subjects on both the hierar-

chical and sequential aspects of scripted

activities. '°

The final consideration that led to our se-

lection of scripts is the degree of specificity

or concreteness that seems to exist with re-

spect to scripted roles, characters, and

props. While other schema-level structures

are constrained by their stereotypic quali-

ties (e.g., narratives have setting, theme,

plot, resolution'’), scripts are constrained

not only by abstract structural properties,

but also by content. Thus, there is less po-

tential for variation within a script than

within a story. For example, a restaurant

script by definition should be set in a res-

taurant, whereas the setting possibilities

for a narrative are almost limitless. Simi-

larly, the role of waiter in a script seems to

have a more limited set of characteristic

features (e.g., a waiter is human, takes the

order, brings the food) than the role of hero

in a story (a hero can be man or beast, ani-

mate or inanimate, real or imagined, etc.)

This a priori “‘variable binding” quality of

scripts may allow the reader to establish a

limited domain of referents upon activa-

tion of the script.” Thus, scripted texts

might be easier to comprehend than non-

scripted ones by virtue of their implicit co-

herence properties. This aspect of scripts is

one we are beginning to study.

Script-Based Research

We have previously stated that we (as

well as other researchers) assume that

scripts and other schematic knowledge

structures are internal sets of knowledge

that are used to guide the encoding, stor-

age, and retrieval of information from ex-

ternal sources. Such a schema-theoretic

view is appealing because it can accommo-

date the inferences, intrusions, and elabo-

rations that typically occur during reading,

90 FRANK R. YEKOVICH

recognition, and recall. A number of schema

theories have been proposed to account for

the persistent tendency of readers to “un-

derstand” and ‘“‘remember”’ information

from text that has not been explicitly pre-

sented.!® 2) 7 23 While an in depth discus-

sion of schema theory in general is beyond

the scope of this paper, we do wish to dis-

cuss some of the findings that relate specifi-

cally to scripts (for a discussion of schema

theory, see Ref. 24). Most of these studies

investigated the effects of scripts on memory

rather than on comprehension. Neverthe-

less, it is still useful to summarize this re-

search since the findings provide valuable

information about the properties of scripts

as knowledge structures.

Bower established scripts as a psychologi-

cally valid concept with the publication of

seven experiments describing scripts in

memory for text.'° We will present some of

their more important findings, integrating

other research into our discussion where

relevant. In the course of the Bower et al.

experiments, four major dependent vari-

ables were utilized: generation, recall,

recognition, and reading time. We will con-

sider each of these separately.

Script generation. Since scripts are hy-

pothetical structures that represent people’s

knowledge about routine activites, one

possible way to collect data about scripts is

to ask people to describe scripted activities

in detail. It was not known initially whether

people would agree about the probable

characters, props, actions, and event se-

quences. A reasonable assumption was

that people’s conceptualizations of such

typical activities shared a “family resem-

blance”’ relationship analogous to that shared

by category members in the studies of natu-

of eating in a restaurant might be expected

to possess some or most of the typical res-

taurant actions, but no one particular action

is necessary to all instances. For example,

although eating would generally be consid-

ered to be one of the most critical aspects of

going toa restaurant, it is certainly possible

that a patron in a restaurant could sud-

denly become ill and leave without eating.

This unexpected turn of events would not

invalidate the experience as an instance of

the restaurant script. Instead the incident

would probably be encoded and stored asa

restaurant instance, but an atypical one.

The results of Bower confirmed the family

resemblance quality of scripts.'* Appar-

ently most members of a culture share cer-

tain stereotypical expectations about rou-

tine activities, yet allow a certain amount of

deviation, particularly with respect to less

central events. The tendency for subjects to

agree on which events were most important

was reflected in their judgment about how

to segment the continuous script activity

into particular chunks or scenes. Resulting

chunks corresponded roughly to the enter-

ing, ordering, eating, and exiting scenes

postulated by Schank and Abelson. '° These

chunks suggested a hierarchical, as well as

sequential, organization in memory for the

activity.

Script recall. Since there is a high de-

gree of overlap among instances of ascript,

Bower hypothesized’’ that subjects might

confuse explicit actions with implied ac-

tions when recalling a scripted text. They

distinguished among a Script, which refers

to a generic memory structure In a person’s

head; a script-based text which is the pas-

sage about a scripted event that is presented

to the subject; and an instantiated script

which is a composite of the script and the

script-based text that is set up in episodic

memory in response to a particular text en-

counter. They hypothesized that memory

for the surface features of the presented

text would fade over time, forcing subjects

to rely on the generic contributions to the

instantiated script. If this happens, then

unstated script actions would be expected

as intrusions in recall. The findings” of

Bower supported this hypothesis. The per-

cent of script-related intrusions ranged

from 19% to 31%, with the level of intru-

sions higher when two or three similar

scripted texts were read than when only one

text was read on a particular topic. Thus,

subjects who read about a visit to the den-

KNOWLEDGE-BASED COMPREHENSION 91

tist or a visit to the chiropractor (or both) in

addition to reading about a visit to the doc-

tor tended to produce more script-related

intrusions at recall than subjects who only

read the doctor script. The precise reason

for this “‘number of script versions”’ effect

is unclear. Bower et a/. argued that each of

the three versions activated a partial copy

of a superordinate ‘‘Visit to a Health Pro-

fessional” script.'° Activation from the

stated actions spread to actions not stated

in any instance of the presented script. This

accumulated activation of the superordi-

nate action nodes in the health professional

script allowed interfering “‘cross-talk”’ to

occur among the different instantiations of

the same abstract script. This cross-talk, in

turn, led to more intrusions at recall if more

than one script version was presented at

input. Some additional comments on this

effect will be made in the next section on

recognition. However, two other recall re-

sults of Bower will be mentioned briefly be-

fore we turn to recognition.’ These two

findings were:

1. subjects reordered the actions in scripts

when the script actions were presented

out of order

2. subjects remembered goal-relevant

deviations from a script better than

typical script actions’” 7’

The tendency to reorder provided addi-

tional evidence of the sequential properties

of scripts. The order imposed on the scripted

events agreed with the order produced in

the generation task.’° The finding was in-

terpreted as an example of proactive inter-

ference. The tendency to remember goal-

relevant deviations was cited by Graesser,

Gordon, and Sawyer’ as an example of the

“‘von Restorff”’ effect (i.e., better memory

for a surprising event). They contended

that properties of a passage’s represen-

tation can explain this discriminative accu-

racy better than the amount of cognitive re-

sources allocated at acquisition.

Script recognition. The recognition re-

sults’? of Bower et al., as well as those re-

ported by Graesser et al. agree with the

major result of the script recall experiment,

namely that subjects have trouble discrimi-

nating stated script actions from unstated

but script-relevant actions. Two factors

were shown to increase the magnitude of

this effect. First, false alarms to unstated

actions increased if subjects read more than

One text referring to the same script. Sec-

ond, false alarms to unstated script actions

increased for very typical actions.

The original version of script theory’®

could not account for these recognition re-

sults. However, Schank” has revised his

ideas about scripts somewhat to include

various levels of abstraction in memory

(i.e., event, general event, situational, in-

tentional). He now argues that no one

‘“‘dental” script exists in memory as a pre-

compiled chunk. Rather “‘the bare bones of

the dentist script’ (p. 262) can be embel-

lished as needed by combining information

from the input set with relevant pieces of

information from memory to create on

demand a dental script that is applicable in

a particular situation. This “‘construct on

demand” aspect of Schank’s revised theory

will accommodate recognition confusion

between a visit to the doctor and a visit to

the dentist. However, direct tests of the

construct-on-demand notion have not been

made.

Reading time for script-based texts. The

final dependent measure used in the Bower

studies was reading time.'* They measured

reading time for statements that were adja-

cent in the surface structure of the text but

varied in their underlying distance in the

script. The results showed facilitation for

text statements immediately adjacent in the

underlying script. However, statements two

steps apart were not read faster than state-

ments that were three steps apart in the un-

derlying script.* A second finding was that

sentences in the second half of the script

were read faster than sentences in the first

half. Both these findings suggest that some

*This weak separation effect was not found by

Abelson and Reder,”® although Smith” recently ob-

tained a strong effect for separation.

92 FRANK R. YEKOVICH

priming effect occurs as activation spreads

within a script, but the details of this

spreading activation are not clear. Ander-

son? and den Uyl and van Oostendorp’’

have also reported within-script priming

effects. These priming studies provide the

most direct support for the notion that ac-

tivating scripts during comprehension fa-

cilitates processing because the scripts make

it possible to expect and predict aspects of

the subsequent input.

So what have we learned from this sum-

mary of script research and how does it re-

late to our experiment? Two basic points

are relevant. First, the fact that it is hard to

discriminate stated script actions from un-

stated script actions thirty minutes after

presentation,’” *’ implicates the generic

script as a source of interference at least at

retrieval. Whether this confusion between

the generic script (stored) and the pre-

sented script (text) is a result of activation

at encoding, at retrieval, or both is not known.

Thus we decided to use an immediate mem-

ory test to see how quickly the surface fea-

tures of the presented text become indistin-

guishable from the stored features of the

pre-experimental script. Second, within-

script priming effects suggest that at least

some script activation occurs during read-

ing.'* '* '’ Exactly what and how much the

reader activates is less clear. In our experi-

ment, we decided to investigate whether

mention of an action within a script auto-

matically activates the accompanying props

and characters. There are two reasons why

we chose to use script-relevant words rather

than propositions or actions in our study.

First, in future studies we intend to study

scripts using the item recognition paradigm

developed by Ratcliff and McKoon.*™” *!

However, we first needed to be sure that

recognition confusion would replicate at

the level of items (i.e., props and charac-

ters). Second, and more important, we

wanted to explore whether actions are auto-

matically instantiated. The actions (or

predicators) within a script normally as-

sume particular characters who use certain

props to carry out the actions of interest. If

an ordering action is specified in a restau-

rant text, for example, the reader may logi-

cally infer that a meal was ordered from a

waiter who provided a menu. Thus, the ac-

tivation of an action node should cause

some activation, and consequently instan-

tiation, of the associated actors and props.

A similar comprehension effect for verb-

based inferences has been shown using con-.

text-target sentence pairs.”

An Exploratory Study of Scripts in

Immediate Memory

The purpose of the pilot work reported

here was to establish that scripts in memory

are retrieved, activated, and used during

the encoding of “‘scripted”’ texts. Two pri-

mary manipulations occurred in the study.

First, the presentation format of the texts

was varied in two ways. Some texts were

divided into thirds during presentation (re-

ferred to as SEGMENTED), and subjects

received a memory test immediately after

the presentation of each segment. Other

texts were presented in complete form

(TOTAL), and subjects received a memory

test immediately after the text’s completion.

This format variation alters the amount of

information resident in immediate memory.

It is reasonable to assume, especially in the

SEGMENTED condition, that the result-

ing memorial representation is a moderately

accurate depiction of recent encoding.*” *”

The second manipulation concerned the

test that subjects saw for recognition. The

test items were single words that were

either relevant or irrelevant to the active

script. Script-relevant words were essen-

tially props that supported the characters’

actions. They were either mentioned explic-

ity in the text (EXPLICIT) or only implied

(IMPLICIT). For example, in one sentence

from the Restaurant passage, Soon Jack

and Chris were seated (at their table.), the

prepositional phrase at their table was

either mentioned or deleted. Script-irrele-

vant words were nouns that were UNRE-

LATED to the active script. UNRELATED

items served as a control. The purpose of

KNOWLEDGE-BASED COMPREHENSION 93

the item type variable was to measure the

degree of confusion that occurs when a

memory script interacts with a text highly

similar to the script, thereby forming a

composite episodic trace in immediate

memory. We also measured the time re-

quired to make a decision regarding each

word’s explicit mention in the text.

By adopting two assumptions about script

activation and use, several interesting pre-

dictions arise concerning the variables out-

lined above. Assume first that the episodic

trace in immediate memory is a composite

of the text input and the memory script.

This means that during encoding people

will ‘‘fill in’? an incomplete text by using

their stored script. So, if props are implied

by the text, but not mentioned, they will be

filled in and become part of the composite

trace in immediate memory. Next, assume

that when information in a text matches in-

formation in the script, the level of activa-

tion of that information probabilistically

increases for a short time.'? Thus, EX-

PLICIT information from the text will

have a higher level of activation in the epi-

sodic trace than will IMPLICIT informa-

tion from the text. Clear predictions emerge

regarding the recognition of EXPLICIT,

IMPLICIT, and UNRELATER nouns after

presentation of text segments or an entire

text. With respect to recognition accuracy,

subjects should be able to reject easily all ir-

relevant or UNRELATED items. It should

also be easy to verify the presence of EX-

PLICIT items because of the additional ac-

tivation present on those nodes in the epi-

sodic trace. However, the correct rejection

of IMPLICIT items should be difficult

since the props should be present though

weakly activated in the composite trace in

immediate memory.

The time required to judge the presence

or absence of a prop in the original text

should show an analogous pattern of out-

comes. Specifically, the UNRELATED

probes should be judged most quickly be-

cause subjects will test the item for its script

relevance and the check will fail. EXPLICIT

props will be slower than UNRELATEDs

because the items will be judged relevant

and then must be searched for in memory.

IMPLICIT props should be the slowest.

Again, the item will be judged relevant and

memory search will occur. However, the

lower level of activation will create uncer-

tainty in the match operation. As a conse-

quence, subjects should take additional

time to make their judgments.

We tested these ideas in an experiment

performed at Catholic University. Thirty

two college undergraduates listened to au-

diotaped scripted texts and received prop

recognition tests. The texts described ste-

reotypical action sequences of four activi-

ties thought to be common among college

undergraduates. The topics were taking a

date to a restaurant, applying for college,

painting a room, and checking out a book

from the library. Two versions of each text

were composed to counterbalance the EX-

PLICITness and IMPLICITness of the

items; EXPLICIT items in one version

were IMPLICIT in the other version. Ex-

amples from the two versions of the Res-

taurant passage follow:

Version A Then he (Jack) drove over to

Chris’ house to pick her up.

They drove to the restaurant

and Jack pulled into the drive-

way. He helped Chris out, left

the keys, and then they walked

inside. Jack spoke to the hostess

at the desk, and they sat fora

few minutes in the waiting

area. Soon they were seated.

Version B- ... He helped Chris out of

the car and let the va/et park

it. Then they walked inside.

Jack checked at the desk, and

they sat for a few minutes in

the waiting area. Soon they

were seated at their table.

The italicized words represent the EX-

PLICIT test items for each version. The

passages varied from 150-200 words in

length and contained 15-20 sentences. Each

version of each passage was recorded on

94 FRANK R. YEKOVICH

audio tape in each presentation format bya

male speaker.

Each recognition test consisted of 32

words; 8 EXPLICIT, 8 IMPLICIT, and 16

UNRELATED nouns. The order of words

within each list was quasi-random. Every

other word was UNRELATED; interspersed

among the UNRELATED words were

EXPLICIT and IMPLICIT? items.; fhe

purpose of this ordering was to eliminate

possible priming of script relevant infor-

mation. The test items were presented indi-

vidually using a tachistoscope. Decision

time was measured to the nearest .01 sec.

Subjects were told to listen closely to

each passage because they would be tested

for their immediate memory of the text.

They received one practice passage and test

to familiarize them with the task require-

ments. Subsequently, each subject listened

to and was tested on four passages (2

SEGMENTED, 2 TOTAL). Subjects saw

only one version of each topic. The tests

occurred immediately after each break point

in each text.

The major results of the experiment are

summarized in Table 1. The table reports

three measures which we will consider in

turn. We first computed the proportion of

test items recognized correctly. As Table 1

shows, there was considerable variability in

the correct recognition rates among the

three types of probes. The most interesting

aspect of the data is that while subjects

were quite good at identifying EXPLICIT

items and rejecting UNRELATEDs, their

performance was relatively poor on IM-

PLICITs. This trend is significant in both

the TOTAL and SEGMENTED conditions;

however, the difference is more pronounced

for the TOTAL than for the SEGMENTED

(the interaction is significant, F(2, 60) =

7.17, p < .002). Note also that when both

presentation conditions are averaged, the

correct rejection of IMPLICITs is only .53,

no better than chance level performance.

This initial evidence suggests that the epi-

sodic trace in immediate memory contained

information from the generic script as well

as information from the presented text,

even when immediate memory was probed

after each segment.

The search of the episodic trace can be

further understood by considering the

decision time results. Newman Keuls tests

showed that for correct decisions, IM-

PLICIT items took significantly longer

than the other two item types (p < .025),

but EXPLICIT and UNRELATED items

did not differ reliably (.05 < p < .10) de-

spite their 100 msec difference (see row 7 of

the table). This pattern conforms generally

to the predictions outlined earlier. In addi-

tion, these data are consistent with theoret-

ical formulations that describe decision

processes underlying the recognition of

Table 1.—Mean Decision Times and Proportions Correct for the Six Experimental Cells.

Presentation Format/Dependent Measure

TOTAL

Correct Decision Time

Time to respond ““YES”’

Correct Recognition

SEGMENTED

Correct Decision Time

Time to respond ““YES”

Correct Recognition

MEAN

Correct Decision Time

Time to respond ‘“‘YES”

Correct Recognition

Item Type

Explicit Implicit Unrelated

1.62 2.04 1.52

1.62 1.79 —

.85 45 .98

1.54 1.86 1.44

1.54 1.84 —

88 .61 99

1.58 1.95 1.48

1.58 1.81 —

86 33 985

KNOWLEDGE-BASED COMPREHENSION 95

scripted information.’ Initially, subjects

made a relevance judgment and based on

the outcome of this check, either rejected

an item outright or proceeded to search the

active memory trace. Evidence of this op-

eration surfaced generally as the difference

between UNRELATED and script related

(i.e., EXPLICIT and IMPLICIT) decision

times. When a probe was judged to be rele-

vant, the next stage involved searching the

episodic representation for a node that

matched the test item. A match occurred

when a node had a high activation level, i.e.

when it had been mentioned EXPLICITly

in the text. If no match was found, a third

process was invoked to decide on the likeli-

hood of an item’s inclusion in the text.

When the activation was above some criter-

ion level, a false positive response was

made, whereas below-threshold activation

led to correct rejection of a probe. The re-

sponse time differences between EXPLICIT

and IMPLICIT probes mimicked the pat-

tern predicted by this stage account. Thus,

the current data are consistent with deci-

sion models proposed by other script theo-

mats.” *°

Several other interesting aspects of the

data surfaced in subsequent analyses, buta

complete accounting is beyond the primary

aim of the paper. As a consequence, we will

close this section with one final comment

about a third aspect of the data. As noted

earlier, our subjects were prone to false

recognitions of IMPLICIT items; about

50% of their responses were false positives.

We computed the response times for these

implicit errors and compared them with

two aspects of the data for correct decision

times. The first comparison involved times

for IMPLICIT false alarms and correct

EXPLICITs. The rationale for this con-

trast is that since our subjects decided (er-

roneously) that an IMPLICIT probe had

been presented, the level of activation on

those nodes might have been roughly equi-

valent to the activation on EXPLICIT

nodes. If this is true, then response times

for EXPLICITs and IMPLICIT false alarms

should be the same. The rows of the table

labelled ‘“‘Time to Respond YES” summar-

ize the means for this comparison. Statisti-

cal analyses revealed that IMPLICIT items

still required longer than EXPLICIT ones

(F(1, 30) = 28, p< .001). Thus, an as-

sumption of equivalent levels of activation

does not seem warranted.

The second comparison involved testing

IMPLICIT false alarms against IMPLICIT

corrects. Since we obtained chance level

performance on IMPLICIT items, it is

reasonable to assume that our subjects

simply could not decide and consequently

resorted to guessing. Under this assump-

tion, decision times for false alarms should

equal correct decision times. The second

column of the table contains the data of in-

terest. Overall, IMPLICIT false alarms

were faster than IMPLICIT corrects (p <

.05), with most of the effect due to the

TOTAL presentation format (recognition

performance was also poor in this condi-

tion). Generally, it does not appear that

guessing alone can account for the pattern

of IMPLICIT response times.

What factors do contribute to these deci-

sion time differences? Two variables sur-

face as possible independent contributors.

The difference between EXPLICIT cor-

rects and IMPLICIT false alarms is prob-

ably due to differing activation levels. In

immediate memory, EXPLICIT terms may

retain surface traces which lead to elevated

activation levels and consequently, faster

decision times. So, although IMPLICIT

false alarms are activated above some cri-

terion level, they do not enjoy additional

activation from the surface traces of the

text. Yet, they were above threshold enough

to give them a speed advantage over IM-

PLICIT corrects. This is probably because

the test items were not controlled for typi-

cality or importance in the script. Conceiv-

ably, the IMPLICIT false alarms occurred

for terms considered more typical as script

props whereas correct rejections of IM-

PLICITs occurred for less typical props.

Previous work by Graesser~’ has shown

that memory discrimination varies with

typicality. Thus, this factor could produce

activation differences, and consequently

response time differences.

96 FRANK R. YEKOVICH

Using Scripts as Tools in Future Research

The experiment just presented suggests

that scripts, as knowledge structures, con-

tribute to the processing of script-based

texts. To reiterate our earlier statement, we

believe that reading comprehension is truly

an interaction among the knowledge and

goals of the reader, and the information

source itself. The view in and of itself is not

a unique one. Several theorists have pro-

posed interactive models of reading.” * °

Yet in much of this previous work, the

burden of processing has rested with text-

driven variables (e.g., word frequency).

Our ultimate goal is to define the knowl-

edge-based variables that cooperate in the

processing effort. Interestingly, even now

the current text variables being used hint at

the need for this definition. Just and Car-

penter’ noted, for example, that “novelty”

(or familiarity?) was a good predictor of a

word’s gaze duration. Similarly, Graesser,

Hoffman, and Clark’ reported that passage

familiarity was one predictor of text read-

ing time. In both cases, the knowledge-

based concept of familiarity appears to

participate in determining reading rate.

In this paper we have considered famil-

larity in terms of scripts, a knowledge

structure we think has potential utility for

investigating comprehension. Our concern

is not with scripts per se, but rather with

scripts as one definable type of complex

knowledge structure. Thus, scripts are use-

ful as a tool for studying knowledge-based

comprehension. They appear more tract-

able than other schemata (e.g., narratives)

because of their structural and content

properties. However, we still do not know

much about the psychologically important

characteristics of scripts. Little research

has explicated clearly how scripts are repre-

sented in memory. What is the constituent

structure of scripts like? Or, might scripts

be represented by some complex set of

overlapping features? These representational

issues are of immediate concern if we hope

to study the processes that operate on

scripts. For example, without having some

idea about the organization of scripts, we

can not discern exactly what information

people will retrieve and activate to guide

the encoding of a text. Nor can we know

how the retrieval and activation operations

work. Through iterations of studies of

representation and process, we will eventu-

ally be able to characterize both represen-

tational and operational characteristics of

scripts. Once psychologists develop an un-

derstanding of these issues, we will be in a

good position to make progress on larger,

more complex memory schemata. The im-

mediate issue, however, is to get a handle

on script representation and process.

Studying knowledge-based comprehen-

sion requires methodologies that allow for

on-line measures of both memory organi-

zation and mental processes. One purpose

of the experiment reported here was to

adapt a standard decision task paradigm to

the study of scripted knowledge in imme-

diate memory. This methodology is only

one way to get at the real-time processes

used during comprehension. In this last

part of the paper, we turn attention to four

alternative paradigms that appear fruitful

for investigating script-based (and more

generally, knowledge-based) comprehen-

sion. For each paradigm we provide a brief

description and then discuss how it might

be used to study scripts >) maiteer

comprehension.

Paradigms for Studying Scripts in

Comprehension and Memory

Priming in item recognition. Priming in

item recognition has been used in a wide

variety of psychological tasks. Recently,

Ratcliff and McKoon” have adapted the

task for use in studying the organization of

textin memory. Essentially, the task requires

a subject to read a short text and then takea

recognition test on some of the content

words from the text. The recognition test is

designed so that occasionally one test item

from the text (the prime) immediately pre-

cedes another word from the text (the

target). The manipulation involves the re-

KNOWLEDGE-BASED COMPREHENSION 97

lation between the prime and the target

(e.g., whether they are from the same propo-

sition or different propositions). The meas-

ure of interest concerns the degree to which

identification of the target (in terms of

speed and/or accuracy) is facilitated by the

presentation of the prime. Psychological

theory predicts that closeness (e.g., same

proposition) should result in greater facili-

tation than remoteness (different proposi-

tions). Ratcliff and McKoon find reliable

priming effects in a number of tasks. For

instance, their research has supported

Kintsch’s* model of memory for text.

Other researchers’* have extended the prim-

ing task to “‘proposition”’ recognition as

opposed to item recognition.

The major consequence of this paradigm

is that psychologists can expose the organi-

zation of memory representations. So, if a

researcher developed a representational

model of script organization, (s)he chould

test the adequacy of the model using a

priming task. In the context of the experi-

ment reported earlier, suppose that charac-

ters and props are more closely associated

with some scene-relevant actions than with

actions from other scenes. Presumably, a

priming paradigm could reveal these asso-

ciations by showing differential priming ef-

fects. In fact, in the study we interleaved

script-relevant words with UNRELATED

words to avoid just such a possible confound.

A second consequence of priming is that

it gives clues about activation.** The most

logical interpretation of priming is that ac-

tivation spreads across nodes in memory,

and closer nodes have a probabilistic ad-

vantage of reaching some threshold for

firing. Although the exact activation

mechanisms are not worked out yet, the

priming technique is potentially useful for

solving some of these questions as well.

Recognition and other decision time tasks.

Two immediate memory tasks seem useful

for mapping the memory features of scripts.

The first involves judgments regarding

whether or not an item or proposition oc-

curred in the text. Simply by getting the

recognition judgment and the amount of

time required to decide, researchers can in-

vestigate memory representations and

recognition judgment processes. For in-

stance the time required to verify a test

statement about a concept is a monotoni-

cally increasing function of the number of

statements learned about the concept (i.e.,

the fan effect'* **). Such effects describe

characteristics of memory representations.

It is reasonable to assume that models of

script representation could be investigated

with similar tests.

The present experiment used a recogni-

tion time task to study the decision process

used in making recognition judgments. It is

easy to imagine how the present task could

be extended to further investigate opera-

tions within the decision process. For in-

stance, if recognition involves direct access

to any part of the memory trace, then deci-

sion time to script-relevant probes should

not vary as a function of the probe’s cen-

trality or importance to the script.

The second immediate memory task in-

volves plausibility or consistency judgments

rather than recognition ones. In other

words, the more a person knows about a

concept, the faster (s)he is at deciding

whether or not a test probe is thematically

consistent with previously learned infor-

mation. It is this sort of task which reveals

““expertise,”’ 1.e., the elaborated knowledge

structure that produces quick and themati-

cally accurate judgments. By incorporating

a consistency task into a priming para-

digm, psychologists should be able to get at

explicit and implicit information which has

been activated as a result of the activation

of a script.

Reading and comprehension time. One

methodology that is becoming increasingly

popular in cognitive psychology is a meas-

ure of the amount of time required to read

and comprehend statements of a text. The

availability of inexpensive computers has

provided an economical way to present

texts in semi-naturalistic formats and col-

lect reading times (usually in milliseconds)

for each segment of text. Typically, the

segments consist of phrases or sentences,

98 FRANK R. YEKOVICH

and the experimental manipulations involve

alterations in various structural and se-

mantic properties of the reading material.

By measuring reading time for the various

conditions, psychologists infer mental proc-

esses required for comprehension. Of course,

we can study the structure of text this way

also. The Bower et al. experiment discussed

earlier is a good example of this paradigm.

This task is easily incorporated into the

study of scripts. Imagine, as we suggested

earlier, that an active script defines loosely,

both global and local coherence of a text.

Two sentences that might seem disjoint

without a script context might be rendered

coherent within a script context. Such an

effect would appear as decreased reading

time for the script consistent sentences. In

fact a similar result has been reported by

den Uyl and van Oostendorp.”’

Eye movements. The final paradigm

that we want to discuss involves measuring

and recording the eye movements that ac-

company reading. Recent advances in

technology have resulted in computer con-

trolled equipment that monitors the accu-

racy of eye movements to within three

character spaces of print. As a consequence,

extremely fine measures can be made of

where the eye stops (fixates) as it reads

across a line of text. Although this equip-

ment is not available to many researchers,

some theorists have initiated in-depth stud-

ies of the reading comprehension process

as it relates to eye fixations.” ** So far the

majority of this work has focused on the

text characteristics that influence fixations.

However, it is reasonable to assume that

the paradigm will be useful for studying the

contribution that knowledge makes to con-

trol eye fixations. By way of speculation,

suppose that activating a script involves

computing the intersection of meaning

among several concepts (realized as words

in a text). Assume also that an active script

provides a domain of ‘“‘active’’ referents.

Presumably, comprehension of script-rele-

vant props and characters should be facili-

tated as a result of the activation. Thus,

gaze durations for script props and charac-

ters might be shorter than expected (i.e.,

based solely on their text-level character-

ists). Thus, changes in modulation patterns

of fixations could be used to characterize

the knowledge-based contributions to

comprehension.

Summary

In this section we have discussed four al-

ternative paradigms for studying how scripts

affect encoding and memory. Some ap-

proaches appear more useful for studying

representational issues whereas other para-

digms seem more suitable for understand-

ing comprehension processes. When used

cotemporaneously, these paradigms can

provide converging sources of evidence re-

garding the representation and use of scripts

in text comprehension. This, of course, will

bring cognitive psychologists closer to

modeling reading comprehension from a

knowledge-based perspective.

References

1. Brown, A. L., Campione, J. C. and Day, J. D. 1981.

Learning to learn: On training students to learn

from texts. Educational Researcher, 10: 14-21.

2. Graesser, A. C. 1981. Prose comprehension beyond

the word. New York: Springer-Verlag.

3. Just, M. A. and Carpenter, P. A. 1980. A theory of

reading: From eye fixations to comprehension.

Psychological Review, 87: 329-354.

4. Kintsch, W. and van Dijk, T. 1978. Toward a model

of text comprehension and production. Psycho-

logical Review, 85: 363-394.

5. Lesgold, A. M. and Perfetti, C. A. 1978. Interactive

processes in reading comprehension. Discourse

Processes, 1: 323-336.

6. Rumelhart, D. E. 1978. Toward an interactive

model of reading. In S. Dornic (Ed.), Attention

and Performance VI. Hillsdale, N.J.: Erlbaum.

7. Graesser, A. C., Hoffman, N. L. and Clark, L. F.

1980. Structural components of reading time.

Journal of Verbal Learning and Verbal Behavior,

19: 135-151.

8. Vipond, D. 1980. Micro- and macroprocesses in

text comprehension. Journal of Verbal Learning

and Verbal Behavior, 19: 276-296.

9. Kintsch, W. and Vipond, D. 1979. Reading com-

prehension and readability in educational prac-

tice and psychological theory. In L. G. Nillson

(Ed.), Perspectives in Memory Research. Hillsdale,

N.J.: Erlbaum.

10. Miller, J. R. and Kintsch, W. 1980. Readability

rt.

20.

os.

22.

23.

KNOWLEDGE-BASED COMPREHENSION 99

and recall of short prose passages: A theoretical

analysis. Journal of Experimental Psychology:

Human Learning and Memory, 6: 335-354.

Thibadeau, R., Just, M. A. and Carpenter, P. A.

1981. A model of the time course and content of

human reading. Unpublished manuscript, De-

partment of Psychology, Carnegie-Mellon

University.

. Schank, R. and Abelson, R. P. 1975. Scripts, plans,

and knowledge. Tbilisi, Georgia, USSR.

. Anderson, J. R. 1980. Concepts, propositions, and

schemata: What are the cognitive units? Nebraska

Symposium on Motivation, 28: 121-162.

. Schank, R. C. 1980. Language and memory. Cog-

nitive Science, 4: 243-284.

. Bower, G. H., Black, J. B. and Turner, T. J. 1979.

Scripts in memory for text. Cognitive Psychology,

11: 177-220.

. Schank, R. and Abelson, R. P. 1977. Scripts, plans,

goals, and understanding. Hillsdale, N.J.: Erlbaum.

. den Uyl, M. and van Oostendorp, H. 1980. The use

of scripts in text comprehension. Poetics, 9:

275-294.

. Meyer, B. J. F. 1975. The organization of prose and

its effects on memory. Amsterdam: North Holland.

. Thorndyke, P. W. 1977. Cognitive structures in

comprehension and memory of narrative dis-

course. Cognitive Psychology, 9: 77-110.

Yekovich, F. R. and Walker, C. H. 1978. Identify-

ing and using referents in sentence comprehen-

sion. Journal of Verbal Learning and Verbal Behav-

ior, 17: 265-277.

Bartlett, F. C. 1932. Remembering. Cambridge:

Cambridge University Press.

Kintsch, W. 1979. On modeling comprehension.

Educational Psychologist, 14: 3-14.

Rumelhart, D. E. and Ortony, A. 1977. The repre-

sentation of knowledge in memory. In R. C. An-

derson, R. J. Spiro, & W. E. Montague (Eds.),

Fat

26.

28.

oo.

30.

31.

32.

33:

34.

Schooling and the Acquisition of Knowledge. Hills-

dale, N.J.: Erlbaum.

. Thorndyke, P. W. and Yekovich, F. R. 1980. A cri-

tique of schema-based theories of human story

memory. Poetics, 9: 23-50.

Rosch, E. and Mervis, C. B. 1975. Family resem-

blances: Studies in the internal structure of cate-

gories. Cognitive Psychology, 7: 573-605.

Smith, E. E., Shoben, E. J. and Rips, L. J. 1974.

Structure and process in semantic memory: A

featural model for semantic decisions. Psycholog-

ical Review, 81: 214-242.

Graesser, A. C., Gordon, S. E. and Sawyer, J. D.

1979. Recognition memory for typical and atypi-

cal actions in scripted activities: Test of a script

pointer plus tag hypothesis. Journal of Verbal

Learning and Verbal Behavior, 18: 319-332.

Abelson, R. P. and Reder, L. M. 1977. Linear order-

ing in scripts. Technical Report, Yale University.

Smith, E. E. Studying on-line comprehension of

stories. Paper presented at the Cognitive Science

Colloquium, Yale University, May 1981.

Ratcliff, R. and McKoon, G. 1978. Priming in item

recognition: Evidence for the propositional struc-

ture of sentences. Journal of Verbal Learning and

Verbal Behavior, 17: 403-417.

McKoon, G. and Ratcliff, R. 1980. The compre-

hension processes and memory structures involved

in anaphoric reference. Journal of Verbal Learning

and Verbal Behavior, 19: 668-682.

Fletcher, C. R. 1981. Short-term memory proc-

esses in text comprehension. Journal of Verbal

Learning and Verbal Behavior, 20: 564-574.

Ratcliff, R. and McKoon, G. 1981. Does activation

really spread? Psychological Review, 88: 454-462.

Carpenter, P. A. and Daneman, M. 1981. Lexical

retrieval and error recovery in reading: A model

based on eye fixations. Journal of Verbal Learning

and Verbal Behavior, 20: 137-160.

This article is based on a presentation given at the 32nd Annual Georgetown University Round Table on

Language and Linguistics. The paper was written while the first two authors were Alfred P. Sloan Foundation

fellows at Carnegie-Mellon University. We thank Rich Balzer for his help in running the subjects.

Journal of the Washington Academy of Sciences,

Volume 73. Number 3. Pages 100-106, September 1983

An Examination of Selected Companion Plant

Combinations, and How Such Systems

Might Operate

W. W. Cantelo and Ralph E. Webb

USDA, Beltsville Agricultural Research Center, Beltsville, MD 20705

ABSTRACT

Tomatoes, beans, cabbage, and squash were interplanted with garlic, chrysanthemums, or

marigolds. Yield reductions of tomatoes and beans occurred with certain companion plants.

The populations of flea beetles, lady beetles, potato leafhopper, Japanese beetle, squash bug,

whitefly, and hornworms differed significantly on one or more vegetables. These differences

have several possible explanations: 1) companion plant volatiles repelled the pest insect, 2)

companion plant volatiles made it more difficult for insects to detect host plant volatiles, 3)

companion plants interfered with the ability of the insect to detect host plant visually, 4) the

companion plants changed the microclimate making it more suitable for insect pest or insect

pest’s enemies, or 5) the companion plants increased the density of vegetation making it far

more difficult for pest to find the preferred area of the host or predatory insect to find pest

insect. These several possibilities are discussed.

Companion planting, or the growing of

particular combinations of ‘“‘compatible”’

plants, is regularly recommended in the

popular literature as a successful, non-

chemical method to thwart insect attack.” ”

The premise is that the odor of one plant

species repels certain insect pests of another

plant species. Documentation to support

or refute this concept is scarce and some-

times contradictory. Sorensen’ reported

that potato plants were protected from in-

sect attack early in the growing season by

planting of marigolds, green beans, and

garlic. Gessell et al.,* on the other hand,

found that cucumber, potato, and cabbage

were not protected by plantings of radish,

snap beans, or thyme, respectively. Varying

results have been reported by others.” ®

Phytophagous insects commonly locate

100

their host plants by means of allelochemics

volatilized from the plant.’ One or many

chemicals in combination may be respon-

sible for the attraction. Negation of the at-

traction might occur if an insect that re-

quired a certain combination of chemicals

to recognize the presence of its host had

other volatiles interfering with the recog-

nition progress. For example, the Colo-

rado potato beetle may be attracted to a

combination of volatiles from the potato

leaf but if geometrical isomers of some of

the attractant chemicals are included in the

atmosphere the beetle no longer can locate

the potato plant.’ Also, tomato or rag-

weed leaves have been reported to emit

volatiles that will reduce flea beetle landing

on collard leaves."

Tests were conducted to determine the

-_-

COMPANION PLANT COMBINATIONS

effect on insect populations and crop yield

of interplanting tomato, squash, cabbage,

and snap beans with three species of plants

frequently reported to provide protection

to other crops i.e., garlic, marigold, and

chrysanthemum. The significance of these

results and the mechanisms that may be re-

sponsible are discussed.

Plot Design and Sampling

Because of space limitations, the mari-

gold and chrysanthemum plots were in a

different field from the garlic plots, with

each field having 4 plots without any com-

panion plants, thus serving as controls.

Since each companion plant species was

evaluated in 4 plots (4 replications), there

was a total of twenty plots. Each plot was

6 X 6 m, comparable to a small home

garden, and consisted of 4 rows of vegeta-

bles. One row contained 14 cabbage plants

(cv. Stonehead) planted 30 cm apart in the

row separated from a row of companion

plants on either side of it by 45 cm. The next

row contained 6 tomato plants (cv. Camp-

bell-28) planted 60cm apart in the row and

separated by 75cm from a row of compan-

ion plants on either side. The third row

contained 44 snap bush green bean plants

(cv. Harvester) planted 8 cm apart in the

row and separated by 60 cm from a row of

companion plants on either side. The fourth

row contained 4 yellow summer squash

plants (cv. Early Prolific Straight Neck)

planted 90 cm apart in the row and sepa-

rated by 90 cm from a row of companion

plants on either side. Each vegetable row

had one companion plant row in common

with the adjacent vegetable row. Two rows

of companion plants were placed along the

ends of the vegetable rows. A generalized

diagram of a plot is depicted in Figure 1.

The plots without companion plants had

the identical design but with no companion

plants. There were 97 marigolds (cv.

Naughty Marietta) spaced 75 cm apart in

the row, 151 chrysanthemum (several cv.)

spaced 30 cm apart, or 512 garlic plants

spaced 15cm apart used per plot. The plots

101

ea Wt

HHH HH HK KKK KKK KK *K

% % *

* * =~— 14 cabbage —~> * *

x * * *

KEKE KEKHKKKKKKEHKHHKEE

* * *

an **=~-—6 tomato *

o * * * *

OKKKKKKKKKKKKK KK

Ex x x

‘O * * ~———- 44 bean * *

* * *

KEK KK KKK KKK KKKES

KK *

4 4 squash ste of

KKHEKKKHKKKEK HK KK *K

Fig. 1. Generalized plot design. Asterisks repre-

sent companion plants.

were aligned so that adjacent plots were no

nearer than 6 m to each other with plowed

land between plots.

Prior to planting, each plot was fertilized

with 8 kg of 10-10-10 and covered with a

sheet of black polyethylene left in place

through the season to suppress weeds. The

vegetable plants were sprayed with a short

residual (pyrethrum-rotenone) spray im-

mediately after planting to eliminate any

insects that may have been present on

them. All plants were examined for insects

weekly. Twenty-three insect species or

groups were recorded during these counts.

Results

The population size of each insect spe-

cies varied among the test crops and among

companion plants (Table 1). For example,

on cabbage there were 40% fewer flea bee-

tles with chrysanthemum plantings than

with no companion plants; on tomato the

reduction was 39% for chrysanthemum vs.

none; on bean there was a decrease of 59%

for the chrysanthemum planting over no

planting; but squash in the chrysanthemum

plantings hosted 15 flea beetles compared

with 0 on the squash with no companion

plants. The last three reductions were not

W. W. CANTELO AND R. E. WEBB

102

ag as I€ Le 97 ((34) 14sI94\)

ges R19 6b (Ns Iv (speay “ON) PI9IA

9 ¢ v ¢ 9 sdnq yung

9 I I 0 I SoTUO]OD pryde Ppoezijiseledg

Iv 61 ve I¢ (47 sajjaoq Apey

I is € I 0 a]190q ssouvder

8SL 899 979 9€9 1€¢ wiomaseqqes payioduy

¢ Z ra 0 4 3nq uinbajiey

€ “6 I I Z sioddoyssein

L09 79¢ eCL7S Q9OLI€ QSE8E sajjaoq b2]4

9 SI 0 Z € S}9YOUID

Ip bb Ol Z Il satuojoo piydy adeqqe)

ve 6 eg] qesl 901 (34) PIA

0 I € 7 ¢ sdnq UNS

99 vIP eel Gegle Q9€E Jaddoujea] 018104

Ze 98 qegez ®00P Qr01 9]]99q UBIq UBIIXIJ|

El LI q8P qr0g 298 sapiaeq Apey

SZ 91 ZL 78 0s 9]190q ssouerder

001 $9 LZ a4 ep -AYauYyM asnoyusainH

jl 81 qeze q€1 eOP sajjaaq eal uvog

auou "SA dpsed auou ‘Sa WNWaYjUBSAIYS ‘sa pjosueUW yoasuy] doiy

TH va1e—sjurjd uoluedu0D

L# Boie—sjuejd uotedui0g

, Sdunued uoruedwos yy sdoid Ino} WOI} SpjatA pue s}o9suT—'T 2/qe]

103

‘suonBAsasgo A]¥2aM Z] Sulinp pajyou sued paysajul jo saquinu [PIO] .

"BOIB PUB MOI IWPS DY) Ul 19}}9] JUDIAJIP B YIM BIeP WO [IAI] BUapIyUOS %¢ ay 1e luasajjIp ApJUBOIJIUTIS d1e 19313] JUIIAJJIP B AQ PIMO]]OJ B1eq |

r9 se e99 q0¢ q6€ (34) PIA

L 0 91 I L WIOMIINIY OFRWIOT

el | Sc LI ie s3nq yuNs

oF b9 101 ca 40) iaddoyjes] 01e10g

9 9 4 I € so1uojoo piyde poziiseieg

6 ré Ee 67 07 sapjaoq Ape]

c I ¢ C 7 9]199q ssoueder

9 4 99 ®61 qP0I SWIOMUIOH

861 6L1 eC8l 978 GET -AYauYyM asnoyusa1H

SI 61 ref bY €9 s9j199q B2]4

4s £6 £8 L8 89 ejiydosoig

Lv Of 9€ Il LI S}9YIUD

LSY 6LS 8L $9 6£ gj199q OVvI0d OpRi0jOg

9 II Sc LI 0€ sa1uojoo piydy o]eWIO]L.

I€1 aa 66 bbl ZOl (34) PISIA

Il 6 SI 81 L 9j199q Joquinons poddins

vc 67 Q87I Perel qe00s 3nq ysenbs

c 0 99 9€1 79 9]199q ysenbs

VET [LZ BEE O87 9ET 9]199q J9quinond pajyjods

91 4 qv7 bop q8 Jaddoyjea] o1e10g

vc ve I¢ 69 Le sajjooq Ape]

0 0 20) qZ qi 9]199q asoueder

8L €L OSI 6v1 pS AYayM asnoyusaiH

qv Li 4 0 SI | $a}120q BII4

£ L £ I I S}aYOUD

v v g C I so1uojoo piydy ysenbs

104 W. W. CANTELO AND R. E. WEBB

significant at the 5% level. On one or more

vegetables, lady beetles, potato leafhoppers,

Japanese beetles, squash bugs, whiteflies,

and hornworms had numbers differing sig-

nificantly between those crops with and

without companion plants. Importantly,

tomato and bean yields were significantly

higher without companion plants of mari-

gold and chrysanthemum than with those

plants, presumably because the companion

plants competed with crop plants for light,

moisture, and nutrients. Garlic, much

smaller plants with obviously less require-

ment for growth, had no significant effect

on yield, except for the number of cabbage

heads.

Discussion

Test results showed no crops to have

their insect populations drastically reduced

by the companion planting although flea

beetles populations were frequently differ-

ent. Several observers have reported flea

beetle populations to be less on crops with

companion plants. The limited studies on

companion planting including this one, al-

though of considerable interest provide

little insight as to the cause of the popula-

tion differences that occur. The number of

variables is too high to determine which are

responsible.

There are several explanations as to why

planting two or more plant species together

could result in the pest populations on one

species being different than if planted as a

monoculture. First, the companion plant

might produce volatiles that repel the pest

insect causing it to leave the planted area.

This is the usual explanation given in the

popular press, but supporting data are

scarce. Altieri et al.'? reported that the

leafhopper Empoasca kraemeri Ross and

Moore was repelled by weeds planted around

the bean plots. Perrin’ showed that tomato

leaf extracts repelled flea beetles on cotton.

However, we consider it unlikely that ac-

tual repellance is common. A more logical

explanation is that volatiles released by the

companion plant disorient the herbivore

by diluting the concentration of volatiles

from the host plant or scrambling the mes-

sage received by the insects’ chemoreceptors

as suggested by Visser and Avé.'° Thus, the

host plant may be more difficult to per-

ceive. The companion plant could physi-

cally protect the host crop. For example,

clover growing in an oat field reduced ovi-

position by the frit fly on oats’* by covering

the preferred oviposition sites on the oat

plant. One alternative explanation is that

the presence of companion plants provides

a different visual pattern to those insects

that use vision to recognize a plant e.g.,

apple maggot fly.'° The addition of com-

panion plants could provide a wider range

of habitats and microclimates for insects,

including beneficial ones.'® Thus, parasites

and predators would be more apt to find

the food and environment they need than

in a monoculture. On the other hand com-

panion plants could make it more difficult

for the parasite to find the host insects.'’ As

a case in point, the denser vegetation usu-

ally associated with companion planting

reduced the ability of a lady beetle to find

European corn borer eggs and resulted ina

destructive population of borers.'* On the

other hand soybean plots diversified with

corn or weeds had more predators.'’ Some

pest species (soybean thrips) prefer the high

density planting usually present in com-

panion planting while others (potato leaf-

hopper) prefer low density planting. '° Some

pest insects also may find the host and

companion plant combination preferable

for multiplication. Moreover, companion

plants could be a source of inoculum for

plant diseases.

Most workers agree that “‘the greater the

diversity the greater the chances of stabil-

ity,”°° but there are many exceptions where

decreased diversity can decrease loss due to

pests.”’ Therefore, to understand the re-

sults of companion planting and to deter-

mine when and why it is effective it will be

necessary to examine each ecological com-

ponent individually.

Our results confirmed those of Cromar-

tie’ that the different herbivorous insects

COMPANION PLANT COMBINATIONS

respond differently to the density of host

plants and their vegetational background.

Also, users of the companion planting strat-

egy should understand that any benefit

from pest suppression may be offset by

yield loss due to competition for available

resources. We suggest that further efforts

to test or explain companion planting con-

centrate on explaining the causes. This will

be a more productive area than empirical

studies.

Acknowledgments

The initial development of this project

was in cooperation with E. C. Bay, A. L.

Steinhauer, and J. L. Hellman of the Uni-

versity of Maryland. R. Hennigar and A.

Nienstedt, USDA technicians, participated

in the sampling.

References

1. Cox, J. 1972. How to plant a companion garden.

Organic Gardening and Farming, 19(2): 50-55.

2. Foster, C. O. 1976. The Companion planter. Hor-

ticulture 54(7): 6-11.

3. Sorenson, A. J. 1976. Effects of companionate

crops and their aqueous extracts on the popula-

tion dynamics of insects injurious to potato

plants. M.S. thesis. University of Massachusetts,

Amherst, Massachusetts, USA.

4. Gessell, S. G., R. J. Precheur and R. W. Hepler.

1975. Companionate plantings—do they work?

Brooklyn Botanic Garden Record 31(1): 24-28.

5. Buranday, R. P. and R. S. Raros. 1975. Effects of

cabbage-tomato interplanting on the incidence

and oviposition of the diamondback moth, Plutella

xylostella (L.). Philipp. Ent. 2: 369-374.

6. Latheef, M. A. and R. D. Irwin. 1979. Effect of

companionate planting on lepidopteran pests of

cabbage. Canad. Ent. 111: 863-864.

7. Latheef, M. A. and R. D. Irwin. 1980. Effects of

companionate planting on snap bean insects, Epi-

lachna varivestea (sic) and Heliothis zea. Environ.

Ent. 9: 195-198.

8. York, A. P. 1978. Organic insect ‘‘control” in

20.

al.

105

Indiana vegetables. Proceed. Indiana Acad. of

Science 87: 243 (abst.).

. Hawkes, C. and T. H. Coaker. 1976. Behavioural

responses to host plant odours in adult cabbage

root fly, Erioischia brassicae (Bouche), Symp.

Biol. Hung. 16: 85-89.

. Visser, J. H. and D. A. Avé. 1978. General green

leaf volatiles in the olfactory orientation of the

Colorado beetle, Leptinotarsa decemlineata. Ent.

exp. & appl. 24: 538-549.

. Tahvanainen, J. O. and R. B. Root. 1977. The in-

fluence of vegetational diversity on the popula-

tion ecology of a specialized herbivore, Phyllo-

treta cruciferae (Coleoptera: Chrysomelidae).

Oecologia 10: 321-346.

. Altieri, M. A., A. van Schoonhoven and J. D. Doll.

1977. The ecological role of weeds in insect man-

agement systems: A review illustrated by bean

cropping systems. Pest Artic. News Summ. 23: 195.

. Perrin, R. M. 1977. Pest management in multiple

cropping systems. Agro-Ecosystems 3: 93-118.

. Adesiyun, A. A. 1979. Effects of intercrop on frit

fly, Oscinella frit, oviposition and larval survival

on oats. Ent. exp. & appl. 26: 208-218.

. Prokopy, R. J. and E. D. Owens. 1978. Visual gen-

eralist with visual specialist phytophagous insects:

Host selection behavior and application to man-

agement. Ent. exp. & appl. 24: 409-420.

. Mayse, M. A. 1978. Effects of spacing between

rows on soybean arthropod populations. J. Appl.

Ecol. 15: 439-450.

. Monteith, L. G. 1960. Influence of plants other

than the food plants of their host on host-finding

by tachinid parasites. Canad, Entomol. 92:

641-652.

. Risch, S. J., R. Wrubel and D. Andow. 1982. Forag-

ing by a predaceous beetle, Coleomegilla maculata

(Coleoptera: Coccinellidae), in a polyculture: ef-

fects of plant density and diversity. Environ. En-

tomol. 11: 949-950.

. Altieri, M. A. and J. W. Todd. 1981. Some influ-

ences of vegetational diversity on insect commun-

ities of Georgia soybean fields. Protection Ecol. 3:

333-338.

Southwood, T. R. E. and M. J. Way. 1970. Ecologi-

cal background to pest management. Jn Concepts

of pest management. R. L. Rabb and F. E. Guth-

rie (Eds.) pp. 6-28. N. C. State University, Raleigh.

Way, M. J. 1979. Significance of diversity in

agroecosystems. Proceedings: Opening Session

and Plenary Session Symposium. IX Interna-

tional Congr. Pl. Prot., Washington, D.C. 1979.

Cromartie Jr., W. J. 1975. The effect of stand size

and vegetational background on the colonization

of cruciferous plants by herbivorous insects. J.

Appl. Ecology 12: 517-533.

1983 Washington Academy of Sciences Membership Directory

Alphabetical List of Members

M = Member: F = Fellow; E = Emeritus Member; L = Life Member or Fellow

ABATE, FRANKS. (Mr), Apt. 1,5311 Connecticut

Ave., Washington, DC 20015 (M)

ABDULNUR, SUHEIL F. (Dr), 5715 Glenwood

Rd., Bethesda, MD 20817 (F)

ABELSON, PHILIP H. (Dr), 4244 50th St., NW,

Washington, DC 20016 (F)

ABRAHAM, GEORGE (Dr), 3107 Westover Dr.,

SE, Washington, DC 20020 (F)

ABSOLON, KAREL B. (Dr) 11225 Huntover Dr.,

Rockville, MD 20852 (F)

ACHTER, MEYER R. (Dr), 417 5th St., SE, Wash-

ington, DC 20003 (F)

ADAMS, CAROLINE L. (Dr), 242 North Granada

St., Arlington, VA 22203 (E)

ADAMS, ALAYNE A. (Dr), 8436 Rushing Creek

Court, Springfield, VA 22153 (F)

ADLER, VICTOR E. (Mr), 8540 Pineway Court,

Laurel, MD 20707 (F)

AFFRONTI, LEWIS F. (Dr), Microbiology, GWU

Med., 2300 Eye St., NW, Washington, DC 20037

(F)

AHEARN, ARTHUR J. (Dr), 9621 E. Bexhill Dr.,

Kensington, MD 20895 (E)

ALBUS, JAMES S. (Dr), 4515 Saul Road, Kens-

ington, MD 20895 (F)

ALDRIDGE, MARY H. (Dr), 2930 45th St., NW,

Washington, DC 20016 (F)

ALEXANDER, ALLEN L. (Dr), 4216 Sleepy Hol-

low Rd., Annandale, VA 22003 (E)

ALEXANDER, BENJAMIN (Dr), P.O. Box 29586

NE Washington, DC 20017 (F)

ALLEN, J. FRANCES (Dr), Rte. 2, Box 3039,

Front Royal, VA 22630 (F)

ALLEN, ANTON M. (Dr), 11718 Lakeway Dr.,

Manassas, VA 22110 (F)

ALLER, ROBERT O. (Mr), 7714 Glenmore Spring

Way, Bethesda, MD 20817 (F)

ALTMAN, STAN (Dr), 16 Park Ave., New York, NY

10016 (M)

ANDERSON, WENDELL L. (Mr), RR #4, Box

4172, La Plata, MD 20646 (F)

ANDERSON, MYRON S. (Dr), 1433 Manchester

Lane, Washington, DC 20011 (E)

ANDREWS, JOHN S. (Dr), 10314 Naglee Rd.,

Silver Spring, MD 20903 (E)

ANDRUS, EDWARD D. (Mr), 6601 Ivy Hill Dr.,

McLean, VA 22101 (M)

APSTEIN, MAURICE (Dr), 4611 Maple Ave.,

Bethesda, MD 20814 (F)

ARGAUER, ROBERT J. (Dr), 4208 Everett St.,

Kensington, MD 20895 (F)

106

ARONSON, CASPER J. (Mr), 3401 Oberon St.,

Kensington, MD 20895 (E)

ARSEM, COLLINS (Mr), 10821 Admirals Way,

Potomac, MD 20854 (M)

ARVESON, PAUL T. (Mr), 10205 Folk St., Silver

Spring, MD 20902 (F)

ASTIN, ALLEN V. (Dr), 508 Battery Lane,

Bethesda, MD 20814 (E)

AXILROD, BENJAMIN M. (Dr), 9915 Marquette,

West Bethesda, MD 20817 (E)

BAILEY, R. CLIFTON (Dr), 6507 Divine St.,

McLean, VA 22101 (M)

BAKER, ARTHUR A. (Dr), 5201 Westwood Dr.,

Bethesda, MD 20816 (E)

BAKER, LOUIS C. W. (Dr), Georgetown Univ.,

Dept. of Chemistry, Washington, DC 20057 (F)

BALLARD, LOWELL D. (Mr), 7823 Mineral Springs

Rd., Gaithersburg, MD 20877 (F)

BARBOUR, LARRY L. (Mr), R.R. 1, Box 492,

Great Meadows, NJ 07838 (M)

BARTFELD, CHARLES I. (Dr), 6007 Kirby Rd.,

Bethesda, MD 20817 (M)

BARBROW, LOUIS E. (Mr), Apt. 918, 6101 16th

St., NW, Washington, DC 20011 (F)

BARUCH, GERARD P. (Dr), Apt D-2, 560 Wilson

Bridge Dr., Oxon Hill, MD 20745 (M)

BASCI, MEHMET I. (Mr), 4209 Isbell St., Wheaton,

MD 20906 (M)

BAUMAN, ROBERT C. (Mr), 9308 Woodberry St.,

Seabrook, MD 20706 (F)

BEACH, LOUIS A. (Dr), 1200 Waynewood Bivd.,

Alexandria, VA 22308 (F)

BECKER, EDWIN D. (Dr), Inst. Arth. & Metab.

Dis., Bldg. 1, Rm. 118, NIH, Washington, DC

20205 (F)

BECKMANN, ROBERT B. (Dr), 10218 Democ-

racy Lane, Potomac, MD 20854 (F)

BElJ, K. HILDING (Mr), 69 Morningside Dr., La-

conia, NH 03246 (L)

BEKEY, IVAN (Mr), 4624 Quarter Charge Dr., Ann-

andale, VA 22003 (F)

BEKKEDAHL, NORMAN (Dr), 405 N. Ocean

Bivd., Apt. 1001, Pompano, FL 33062 (E)

BELSHEIM, ROBERT (Mr), Apt. 514, 2475 Virginia

Ave., NW, Washington, DC 20037 (F)

BENDER, MAURICE (Dr), 16518 NE 2nd Place,

Bellevue, WA 98008 (E)

BENESCH, WILLIAM M. (Dr), Inst. Molecular

Phys. Univ. of Md., College Park, MD 20742 (F)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

BENJAMIN, CHESTER, R. (Dr), Rm. 6552, South

Bidg., OICD-IOA, Dept. of Agri., Washington,

DC 20250, (F)

BENNETT, JOHN A. (Mr), 7405 Denton Rd., Be-

thesda, MD 20814 (F)

BENNETT, MARTIN T. (Mr), 4660 Kenmore Ave.,

#802, Alexandria, VA 22304 (F)

BENNETT, WILLARD H. (Dr), Physics Dept.,

N.C.S.U., Box 5342, Raleigh, NC 27607 (E)

BENSON, WILLIAM (Dr), 636 Massachusetts

Ave., NE, Washington, DC 20002 (M)

BERENDZEN, RICHARD (Dr), President, Ameri-

can Univ., 3800 Massachusetts Ave., NW, Wash-

ington, DC 20016 (F)

BERGER, STEVEN B. (Mr), Apt. 201, 134 Roberts

Lane, Alexandria, VA 22314 (M)

BERGMANN, OTTO (Dr), Dept. of Physics,

George Washington Univ., Washington, DC

20052 (F)

BERKSON, HAROLD (Dr), 12001 Whippoorwill

Lane, Rockville, MD 20852 (M)

BERMAN, ALAN (Dr), Rosenstiel Sch. M &ASci.,

4600 Rickenbacker Causeway, Miami, FL 33149

(F)

BERNETT, MARIANNE K. (Mrs), Naval Res.

Lab., Code 6170, Washington, DC 20375 (M)

BERNSTEIN, BERNARD (Mr), Apt. 608, 7420

Westlake Terrace, Bethesda, MD 20817 (M)

BESTUL, ALDEN B. (Dr), 9400 Overlea Dr.,

Rockville, MD 20850 (F)

BETTS, ALLEN W. (Mr), 2510 S. Ivanhoe Place,

Denver, CO 80222 (M)

BICKLEY, WILLIAM E. (Dr), 6516 40th Ave., Uni-

versity Park, Hyattsville, MD 20782 (F)

BIRD, HERBERT R. (Prof), 5006 Hammersley

Rd., Madison, WI 53711 (E)

BIRKS, L. S. (Mr), 11908 Ledgerock Court, Po-

tomac, MD 20854 (F)

BISHOP, WILLIAM P. (Dr), 4916 Butterworth PI.,

NW, Washington, DC 20016 (F)

BLANCHARD, DAVID L. (Dr), 407 South Wav-

erly, Dearborn, MI 48124 (F)

BLANK, CHARLES A. (Dr), P.O. Box 1481 Ja-

maica Plain, MA 02134-0025 (M)

BLOCH, CAROLYN C. (Mrs), P.O. Box 740, Silver

Spring, MD 20901 (M)

BLONG, CLAIR K. (Dr), 11011 Tenbrook Dr.,

Silver Spring, MD 20901 (M)

BLUNT, ROBERT F. (Dr), 5411 Moorland Lane,

Bethesda, MD 20814 (F)

BOEK, JEAN K. (Dr), National Graduate Univ.,

1101 N. Highland St., Arlington, VA 22201 (F)

BOGNER, M. SUE (Dr), 9322 Friars Rd., Be-

thesda, MD 20817 (F)

BORIS, JAY P. (Dr), 3516 Duff Drive, Falls Church,

VA 22041 (F)

BOTOL, JOSEPH M. (Dr), 9Inkberry Lane, North

Falmouth, MA 02556 (F)

BOWLES, R. E. (Dr), 2105 Sondra St., Silver

Spring, MD 20904 (F)

BOWMAN, THOMAS E. (Dr), Smith Inst. Invert.

Zoology, NHB, Mail Stop 163, Washington, DC

20560 (F)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

BRADY, ROBERT F. JR. (Dr), 706 Hope Lane,

Gaithersburg, MD 20878 (F)

BRANCATO, E. L. (Dr), 7370 Hallmark Rd., Clarkes-

ville, MD 21029 (F)

BRANDEWIE, DONALD F. (Mr), 6811 Field Mas-

ter Dr., Springfield, VA 22153 (F)

BRAUER, GERHARD M. (Dr), 7609 Maryknoll

Ave., Bethesda, MD 20817 (F)

BREIT, GREGORY (Dr), 73 Allenhurst Rd., Buf-

falo, NY 14214 (E)

BRENNER, ABNER (Dr), 7204 Pomander Lane,

Chevy Chase, MD 20815 (F)

BRICKWEDDE, F. G. (Dr), 104 Davey Lab, Dept.

of Physics, PSU, University Park, PA 16802 (L)

BRIER, GLENN W. (Mr), 1729 N. Harrison St.,

Arlington, VA 22205 (F)

BROADHURST, MARTIN G. (Dr), 116 Ridge Rd.,

Box 163, Washington Grove, MD 20880 (F)

BROMBACHER, W. G. (Dr), 17 Pinerun Com-

munity, Doylestown, PA 18901 (E)

BROWN, THOMAS McP. (Dr), 2465 Army-Navy

Dr., Arlington, VA 22206 (F)

BROWN, ELISE A. B. (Dr), 6811 Nesbitt PI.,

McLean, VA 22101 (F)

BRUCK, STEPHEN D. (Dr), 1113 Pipestem PI.,

Rockville, MD 20854 (F)

BRYAN, MILTON M. (Mr), 3322 Glebe Rd., Arling-

ton, VA 22207 (M)

BURAS, EDMUND M. JR. (Mr), Gillette Res. Inst.,

1413 Research Blvd., Rockville, MD 20850 (F)

BURKE, DEAN (Dr), 4719 44th St., NW, Washing-

ton, DC 20016 (E)

BUTTERMORE, DONALD O. (Mr), 1519 No.

Utah St., Arlington, VA 22207 (F)

CAHNMAN, HUGON. (Mr), 162 Pond Dr., Wash-

ington Twp., NJ 07675 (M)

CALDWELL, FRANK R. (Mr), 4821 47th St., NW,

Washington, DC 20016 (E)

CAMPBELL, LOWELL E. (Mr), 10100 Riggs Rad.,

Adelphi, MD 20783 (F)

CAMPBELL, PAUL G. (Dr), Kaetzel Rd., Gap-

land, MD 21736 (F)

CANNON, E. W. (Dr), 18023 134th Ave., Sun City

West, AZ 85375 (F)

CANTELO, WILLIAM W. (Dr), 11702 Wayneridge

St., Fulton, MD 20759 (F)

CARROLL, WILLIAM R. (Dr), 4802 Broad Brook

Dr., Bethesda, MD 20814 (F)

CARTER, HUGH (Dr), Apt. 205, 2039 New

Hampshire Ave., NW, Washington, DC 20009

(F)

CASH, EDITH K. (Ms), 505 Clubhouse Rd., Bing-

hamton, NY 13903 (E)

CERRONI, MATHEW J. (Mr), 10953 Harpers

Square Ct., Reston, VA 22091 (M)

CHAPLIN, HARVEY R. JR. (Mr), 1561 Forest Villa

Lane, McLean, VA 22101 (F)

107

CHAPLINE, WILLIAM R. (Mr), Apt. #614, The

Park Lane, 230 Glenwood Circle, Monterey,

CA 93940 (F)

CHAPMAN, ROBERT D. (Dr), 10976 Swansfield

Rd., Columbia, MD 21044 (F)

CHEEK, CONRAD H. (Dr), 4334 H St., SE, Wash-

ington, DC 20019 (F)

CHEZEM, CURTIS G. (Dr), 46 Center St., P.O.

Box 396, Nantucket IsI., MA 02554 (F)

CHRISTIANSEN, MERYLN. (Dr), Chairman,

U.S.D.A. Plant Physiology Inst., Beltsville, MD

20705 (F)

CHURCH, LLOYD E. (Dr), 4853 Cordell Ave., Be-

thesda, MD 20814 (F)

CLAIRE, CHARLES N. (Mr), 4403 14th St., NW,

Washington, DC 20011 (F)

CLARK, GEORGE E. JR. (Mr), 4022 N. Stafford

St., Arlington, VA 22207 (F)

CLEVEN, GALE W. (Dr), P.O. Box 138, Babson

Park, FL 33827-0138 (F)

CLIFF, RODGER A. (Dr), Box 15, College Park,

MD 20740 (M)

CLINE, THOMAS L. (Dr), 13708 Sherwood Forest

Dr., Silver Spring, MD 20904 (F)

COATES, JOSEPH F. (Mr), 3738 Kanawha St.,

NW, Washington, DC 20015 (F)

COFFEY, TIMOTHY (Dr), 976 Spencer Rd.,

McLean, VA 22102 (F)

COLE, KENNETH S. (Dr), 2404 Loring St., San

Diego, CA 92109 (E)

COLE, RALPH I. (Mr), 3431 Blair Rd., Falls

Church, VA 22041 (F)

COLWELL, RITA R. (Dr), Dept. of Microbiology,

Univ. of Md., College Park, MD 20742 (F)

COMPTON, W. DALE (Dr), Ford Motor Co., P.O.

Box 1603, Dearborn, MI 48121 (F)

COOK, RICHARD K., 4111 BelPre Rd., Rockville,

MD 20853 (F)

COOPER, KENNETH W. (Dr), 4497 Picacho, Riv-

erside, CA 92507 (F)

CORLISS, EDITH L. (Mrs), 2955 Albemarle St.,

Washington, DC 20008 (F)

CORMACK, JOHN G. (Mr), 10263 Gainsborough

Rd., Potomac, MD 20854 (M)

COSTRELL, LOUIS (Mr), 10614 Cavalier Dr.,

Silver Spring, MD 20901 (E)

COTTERILL, CARL H. (Mr), US Bureau of Mines,

2401 E St., NW, Washington, DC 20241 (F)

COYLE, THOMAS D. (Dr), 11400 Game Preserve

Rd., Gaithersburg, MD 20878 (F)

CRAGOE, CARL S. (Mr), 6206 Singleton PI.,

West Bethesda, MD 20817 (E)

CREVELING, CYRUS R. (Dr), 4516 Amherst

Lane, Bethesda, MD 20814 (F)

CULBERT, DOROTHY K. (Mrs), 109 Calle La

Pena., Sante Fe, NM 87501 (F)

CULLINAN, FRANK P. (Dr), 4402 Beechwood

Rd., Hyattsville, MD 20782 (E)

CULVER, WILLIAM H. (Dr), 2841 Chesapeake St.,

Washington, DC 20008 (M)

CURRAN, HAROLD R. (Dr), 3431 N. Randolph

St., Arlington, VA 22207 (E)

108

CURRIE, CHARLES L., S. J. (Dr), President,

Xavier-Univ., 3800 Victoria Parkway, Cincin-

nati, OH 45207-1096 (F)

CURTIS, ROGER W. (Dr), 6308 Valley Rd., Be-

thesda, MD 20817 (E)

CURTISS, LEON F. (Dr), 1690 Bayshore Dr.,

Englewood, FL 33533 (E)

CUTHILL, JOHN R. (Dr), 12700 River Rd., Po-

tomac, MD 20854 (F)

CUTKOSKY, ROBERT D. (Mr), 19150 Roman

Way, Gaithersburg, MD 20879 (F)

DAVIS, CHARLES M. JR. (Dr), 8458 Portland PI.,

McLean, VA 22102 (M)

DAVIS, MARION M. (Dr), Crosslands, Apt. 100,

Kenneth Square, PA 19348 (L)

DAVIS, R. F. (Dr), 8502 Laverne Dr., Adelphi, MD

20783 (F)

DAVISON, MARGARET C. (Mrs), 2928 26th St.

No., Arlington, VA 22207 (M)

DAVISSON, JAMES W. (Dr), 400 Cedar Ridge

Rd., Oxon Hill, MD 20745 (E)

DAWSON, ROY C. (Dr), 7002 Chansory Lane,

Hyattsville, MD 20782 (E)

DAWSON, VICTOR C. D. (Dr), 9406 Curran Rd.,

Silver Spring, MD 20901 (F)

DEAL, GEORGE E. (Dr), 6245 Park Road,

McLean, VA 22101 (F)

DEBERRY, MARIAN B. (Mrs), 3608 17th St., NE,

Washington, DC 20018 (M)

DEDRICK, ROBERT L. (Dr), 1033 Warner Ave.,

McLean, VA 22101 (F)

DELANEY, WAYNE R. (Mr), 4801 Bradley Blvd.,

Chevy Chase, MD 20815 (M)

DEMUTH, HAL P. (Cdr), 24 S. Washington St.,

Winchester, VA 22601 (F)

DENNIS, BERNARD K. (Mr), 915 Country Club

Dr., Vienna, VA 22180 (F)

DESLATTES, RICHARD D. JR. (Dr), 610 Aster

Blvd., Rockville, MD 20850 (F)

DEUTSCH, STANLEY (Dr), 7109 Laverock Lane,

Bethesda, MD 20817 (F)

DEVIN, CHARLES JR. (Dr), 629 Blossom Dr.,

Rockville, MD 20850 (M)

DEVOE, JAMES R. (Mr), 17708 Parkridge Dr.,

Gaithersburg, MD 20878 (F)

DE WIT, ROLAND (Dr), 11812 Tifton Dr., Rock-

ville, MD 20854 (F)

DICKSON, GEORGE (Mr), 52 Orchard Way, No.,

Rockville, MD 20854 (F)

DIMOCK, DAVID A. (Mr), 4291 Molesworth Terr.,

Mt. Airy, MD 21771 (F)

DIXON, PEGGY ANN (Dr), 9011 Eton Rd., Silver

Spring, MD 20901 (F)

DOCTOR, NORMAN (Mr), 3814 Littleton St.,

Wheaton, MD 20906 (F)

DONALDSON, JOHANNAB. (Mrs), 3020 No. Edi-

son St., Arlington, VA 22207 (F)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

——

‘

ues

DONNERT, HERMANN J. (Dr), RFD #4, Terra

Hts., Manhattan, KS 66502 (F)

DOUGLAS, CHARLES A. (Dr), 7315 Delfield St.,

Chevy Chase, MD 20815 (F)

DOUGLAS, THOMAS B. (Dr), 3031 Sedgewick

St., Washington, DC 20008 (M)

DRAEGER, R. HAROLD (Dr), 1201 No. 4th St.,

Tucson, AZ 85705 (E)

DRECHSLER, CHARLES (Dr), University Park,

6915 Oakridge Rd., Hyattsville, MD 20782 (E)

DUBEY, SATYA D. (Dr), 7712 Groton Rd., West

Bethesda, MD 20817 (E)

DUERKSEN, J. A. (Mr), 3134 Monroe St., NE,

Washington, DC 20018 (E)

DUFFEY, DICK (Dr), Nuclear Engineering Dept.,

Univ. of Md., College Park, MD 20742 (F)

DUNCOMBE, RAYNOR, L. (Dr), 1804 Vance Cir-

cle, Austin, TX 78701 (F)

DUNKUM, WILLIAM W. (Dr), 6602 Prospect St.,

Richmond, VA 23226 (F)

DURIE, EDYTHE G. (Mrs), 5011 Larno Dr., Alex-

andria, VA 22310 (F)

DU PONT, JOHN E. (Mr), P.O. Box 358, Newton

Square, PA 19073 (F)

EDDY, BERNICE E. (Dr), 6722 Selkirk Ct. West

Bethesda, MD 20817 (E)

EDINGER, STANLEY E., #404N, 12000 Old George-

town Rd., Rockville, MD 20852 (M)

EISENBERG, PHILLIP (Mr), 7210 Pindell School

Rd., Laurel, MD 20707 (M)

EISENHART, CHURCHILL (Dr), 9629 Elrod Rd.,

Kensington, MD 20895 (F)

EL-BISI, HAMED M. (Dr), 135 Forest Rd., Millis,

MA 02054 (M)

ELISBERG, F. MARILYN (Mrs), 4008 Queen Mary

Dr., Olney, MD 20832 (F)

ELLINGER, GEORGE A. (Mr), 739 Kelly Dr.,

York, PA 17404 (E)

ELLIOTT, F. E. (Dr), 7507 Grange Hall Dr., Ft.

Washington, MD 20744 (E)

EMERSON, K. C. (Dr), 560 Boulder Dr., Sanibel,

FL 33957 (F)

EWERS, JOHN C. (Mr), 4432 26th Rd., No.,

Arlington, VA 22207 (E)

FARMER, ROBERT F. III (Dr), The Gillette Com-

pany, Gillette Park, 6F-1E, Boston, MA 02106

(F)

FARROW, RICHARD P. (Mr), 5012 Weaver Terr.,

NW, Washington, DC 20016 (F)

FAULKNER, JOSEPH A. (Mr), 1007 Sligo Creek

Pkwy., Takoma Park, MD 20912 (F)

FAUST, GEORGE T. (Dr), P.O. Box 411, Basking

Ridge, NJ 07920 (E)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

FAUST, WILLIAM R. (Dr), 5907 Walnut St., Tem-

ple Hills, MD 20748 (F)

FEARN, JAMES (Dr), 4446 Alabama Ave., SE,

Washington, DC 20019 (F)

FEINGOLD, S. NORMAN (Dr), 9707 Singleton

Dr., Bethesda, MD 20817 (F)

FELSHER, MURRAY (Dr), 20138 Wynnfield Dr.,

Germantown, MD 20874 (M)

FERRELL, RICHARD A. (Dr), Dept. of Physics,

Univ. of Md., College Park, MD 20742 (F)

FILIPESCU, NICOLAE (Dr), 5020 Little Falls Rd.,

Arlington, VA 22207 (F)

FISHER, JOELL. (Dr), 4033 (Olley Lane, Fairfax,

VA 22030 (M)

FLINN, DAVID R. (Dr), 8104 Bernard Dr., Ft.

Washington, MD 20744 (F)

FLORIN, ROLAND E. (Dr), A-313, Bldg. 224,

563.02, Nat'l Bureau of Standards, Washing-

ton, DC 20234 (F)

FLYNN, DANIEL R. (Mr), 15 Dellcastle Court,

Gaithersburg, MD 20879 (M)

FLYNN, JOSEPH H. (Dr), 5309 Iroquois Rd.,

Bethesda, MD 20816 (F)

FOCKLER, HERBERT H. (Mr), 10710 Lorain

Ave., Silver Spring, MD 20901 (M)

FONER, SAMUEL N. (Dr), Applied Physics Lab,

JHU, Johns Hopkins Road, Laurel, MD 20707

(F)

FOOTE, RICHARD H. (Dr), Box 166, Lake of the

Woods, Locust Grove, VA 22508 (F)

FORZIATI, ALPHONSE F. (Dr), 15525 Prince

Frederick Way, Silver Spring, MD 20906 (F)

FORZIATI, FLORENCE H. (Dr), 15525 Prince

Frederick Way, Silver Spring, MD 20906 (F)

FOSTER, AUREL O. (Dr), 4613 Drexell Rd., Col-

lege Park, MD 20740 (E)

FOURNIER, ROBERT O. (Dr), 108 Paloma Rd.,

Portola Valley, CA 94025 (F)

FOWLER, WALTER B. (Mr), 9404 Underwood St.,

Seabrook, MD 20706 (M)

FOWLER, EUGENE (Mr), Int. Atomic Agency,

Wagramerstr. 5, P.O. Box 200, A-1400 Vienna,

Austria (M)

FOX, WILLIAM B. (Dr), 1813 Edgehill Dr., Alex-

andria, VA 22304 (F)

FOX, DAVID W. (Dr), Dir., Mathematical Sci., A.F.

Off. Sci. Research, Bolling AFB, DC 20332 (F)

FRANKLIN-RAMIREZ, LOUISE (Ms), 2501 N.

Florida St., Arlington, VA 22207 (E)

FRANZ, GERALD J. (Dr), Box 695, Bayview, ID

83803 (F)

FREEMAN, ANDREW F. (Mr), 5012 33rd St. No.,

Arlington, VA 22207 (E)

FRENKIEL, FRANCOIS N. (Dr), 4545 Connecti-

cut Ave., NW, Washington, DC 20008 (F)

FRIEDMAN, MOSHE (Dr), 4511 Yuma St., NW,

Washington, DC 20016 (F)

FRIESS, SEYMOUR L. (Dr), 6522 Lone Oak

Court, Bethesda, MD 20817 (F)

FRUSH, HARRIET L. (Dr), Apt. #104, 4912 New

Hampshire Ave., NW, Washington, DC 20011

(F)

109

FULLMER, IRVIN H. (Mr), Rte. 1, Box 400-F4,

Altoona, FL 32702 (E)

FURUKAWA, GEORGE T. (Dr), Bldg. 221, Rm.

128, Nat'l Bureau of Standards, Washington,

DC 20234 (F)

FUSONIE, ALAN E. (Dr), 5611 Victoria Lane,

Sunderland, MD 20689 (F)

GAGE, WILLIAM W. (Dr), 10 Trafalgar St., Ro-

chester, NY 14619 (F)

GALLER, SIDNEY R. (Dr), 6242 Woodcrest Ave.,

Baltimore, MD 21209 (E)

GALTSOFF, PAUL S. (Dr), P.O. Box 684, Fal-

mouth, MA 02540 (E)

GANT, JAMES Q. JR. (Dr), 4349 Klingle St., NW,

Washington, DC 20016 (M)

GARVIN, DAVID (Dr), Apt. #519, 18700 Walker's

Choice Rd., Gaithersburg, MD 20879 (F)

GAUNAURD, GUILLERMO C. (Dr), 4807 Macon

Rd., Rockville, MD 20852 (F)

GENTRY, JAMES W. (Prof) Dept. of Chem. &

Nuc. Engr., University of Maryland, College .

Park, MD 20742 (F)

GHAFFARI, ABOLGHASSEN (Dr), 5420 Golds-

boro Rd., West Bethesda, MD 20817 (L)

GHOSE, RABINDRA N. (Dr), 8167 Mulholland

Terr., Los Angeles, CA 90046 (F)

GIACCHETTI, ATHOS (Dr), Dept. Sci. Aff., OAS,

1889 F St., NW, Washington, DC 20066 (M)

GIBSON, JOHN E. (Dr), Box 96, Gibson, NC

28343 (E)

GINTHER, ROBERT J. (Mr), Code 6570-1, Naval

Research Lab, Washington, DC 20375 (F)

GIST, LEWIS A. (Dr), 1336 Locust Rd., NW,

Washington, DC 20012 (F)

GLASER, HAROLD (Dr), 1346 Bonita, Berkeley,

CA 94709 (F)

GLASER, GERARD R., JR. (Mr), 6114 Montrose

Rd., Cheverly, MD 20785 (M)

GLASGOW, AUGUSTUS R. JR. (Dr), 4116

Hamilton St., Hyattsville, MD 20781 (E)

GLUCKMAN, ALBERT (Mr), 11235 Oakleaf Dr.,

Silver Spring, MD 20901 (F)

GLUCKSTERN, ROBERT L. (Dr), Weschester

Park, #1116, 6100 Weschester Park Dr., Col-

lege Park, MD 20740 (F)

GOFF, JAMES F. (Dr), 3405 34th Place, NW,

Washington, DC 20016 (F)

GOKEL, GEORGE W. (Dr), Dept. of Chemistry,

Univ. of Md., College Park, MD 20742 (F)

GOLDBERG, MICHAEL (Mr), 5823 Potomac Ave.,

NW, Washington, DC 20016 (F)

GOLDSMITH, HERBERT (Dr), 238 Congressional

Lane, Rockville, MD 20852 (M)

GOLDSTEIN, SETH R. (Dr), 5906 Kirby Rd.,

Bethesda, MD 20816 (F)

GOLUMBIC, CALVIN (Dr), 6000 Highboro Dr.,

Bethesda, MD 20817 (M)

GONET, FRANK (Dr), 4007 North Woodstock St.,

Arlington, VA 22207 (E)

110

GOODE, ROBERT J. (Mr), Assoc. Supt. MS&T

Div., Code 6301, Naval Research Lab, Wash-

ington, DC 20375 (F)

GORDON, RUTH E. (Dr), Amer. Type Culture

Coll., 12301 Parklawn Dr., Rockville, MD 20852

(E)

GRAMANN, RICHARD H. (Mr), 1613 Rosemont

Court, McLean, VA 22101 (M)

GRAY, IRVING (Dr), 4620 North Park Ave., Chevy

Chase, MD 20815 (F)

GREENOUGH, M. L. (Mr), Greenough Data

Assoc., 616 Aster Blvd., Rockville, MD 20850

(F)

GREENSPAN, MARTIN (Mr), 12 Granville Dr.,

Silver Spring, MD 20901 (E)

GREER, SANDRA (Dr), 11402 Stonewood Lane,

Rockville, MD 20852 (F)

GRESAMORE, NELSON T. (Prof), 9536 East

Bexhill Drive, Kensington, MD 20895 (F)

GROSS, ROSALIND L. (Dr), 6302 Queens Chapel

Rd., University Park, Hyattsville, MD 20782 (M)

GROSSLING, BERNARDO F. (Dr), 10903 Amer-

herst Ave., #241, Silver Spring, MD 20902 (F)

GRUNTFEST, IRVING (Dr), 1900 South Eads St.,

Arlington, VA 22202 (F)

GURNEY, ASHLEY B. (Dr), 4606 No. 41st St.,

Arlington, VA 22207 (E)

HACSKAYLO, EDWARD (Dr), Agri. Res. Ctr.-

West, USDA, Beltsville, MD 20705 (F)

HAENNI, EDWARD O. (Dr), 7907 Glenbrook Rad.,

Bethesda, MD 20814 (F)

HAINES, KENNETH A. (Mr), 3542 N. Delaware

St., Arlington, VA 22207 (F)

HALL, E. R., 1637 W. 9th St., Lawrence, KS 66044

(E)

HAMER, WALTER J. (Dr), 3028 Dogwood St.,

NW, Washington, DC 20015 (E)

HAMMER, GUY S. II (Mr), 07-N-40, ABCL Med.

Serv., Box 121 Via Dhahran Airport, Jubail In-

dustrial City, S. Arabia (F)

HAND, CADET H., JR. (Prof), Star Route Box

247, Bodega Bay, CA 94923 (F)

HANEL, RUDOLF A. (Dr), 31 Brinkwood Rd.,

Brookeville, MD 20833 (F)

HANIG, JOSEPH P. (Dr), 822 Eden Ct., Alexan-

dria, VA 22308 (F)

HANSEN, LOUIS S. (Dr), Oral Path., Rm. S-524,

OM&D, Univ. of Calif., San Francisco, CA

94143 (F)

HANSEN, MORRIS H. (Mr), Westat Research

Inc., 1650 Research Blvd., Rockville, MD 20850

(F)

HARMISON, LOWELL T. (Dr), 2055 Crossing

Gate Way, Vienna, VA 22180 (F)

HARR, JAMES W. (Mr), 9503 Nordic Dr., Lan-

ham, MD 20706 (M)

HARRINGTON, MARSHALL C. (Dr), 4545 Conn.

Ave., NW, Washington, DC 20008 (E)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

HARRINGTON, FRANCIS D. (Dr), Apt. 204, 4600

Ocean Beach Blvd., Cocoa Beach, FL 32931

(F)

HARRIS, MILTON (Dr), Suite 500, 3300 White-

haven St., NW, Washington, DC 20007 (F)

HARRISON, W. N. (Mr), 3734 Windom PI., NW,

Washington, DC 20016 (E)

HARTLEY, JANET W. (Dr), Nat'l Inst. of Allergy,

Nat'l Inst. of Health, Bethesda, MD 20814 (F)

HARTMANN, GREGORY K. (Dr), 10701 Keswick

St., Garrett Park, MD 20896 (F)

HARTZLER, MARY P. (Ms), 3326 Hartwell Ct.,

Falls Church, VA 22042 (M)

HASKINS, CARYL P. (Dr), Suite 604, 2100 M St.,

NW, Washington, DC 20037 (E)

HASS, GEORG H. (Dr), 7728 Lee Ave., Alexan-

dria, VA 22308 (F)

HAUPTMAN, HERBERT A. (Dr), 73 High St., Buf-

falo, NY 14203 (F)

HAYDEN, GEORGE A. (Dr), 1312 Juniper St.,

NW, Washington, DC 20012 (E)

HEADLEY, ANNE R. (Dr), Watergate West, 2700

Virginia Ave., NW, Washington, DC 20037 (F)

HEIFFER, MELVIN H. (Dr), Whitehall Apt. #701,

4977 Battery Lane, Bethesda, MD 20814 (F)

HEINRICH, KURT F. (Dr), 804 Blossom Dr.,

Rockville, MD 20850 (F)

HENDERSON, EDWARD F. (Dr), 4600 Conn.

Ave., NW, Washington, DC 20008 (E)

HENNEBERRY, THOMAS J. (Dr), 1409 E. North-

shore Dr., Tempe, AZ 85283 (F)

HERBERMAN, RONALD B. (Dr), 8528 Atwell

Rd., Potomac, MD 20854 (F)

HERMACH, FRANCIS L. (Dr), 2415 Eccleston

St., Silver Spring, MD 20902 (F)

HERMAN, ROBERT (Dr), 8434 Antero Dr., Aus-

tin, TX 78759 (F)

HERMAN, LLOYD G. (Dr), 4613 Highland Ave.,

Bethesda, MD 20814 (F)

HERSEY, JOHN B. (Dr), 923 Harriman St., Great

Falls, VA 22066 (M)

HEWSTON, ELIZABETH M. (Ms), Felicity Cove,

Shady Side, MD 20764 (F)

HEYDEN, FRANCIS J., S.J., Manila Observa-

tory, P.O. Box 1231, Manila, Philippines D-404

(E)

HEYER, W.R. (Dr), Amphibian & Rept. Nat. Hist.,

Smithsonian Inst., Washington, DC 20560 (F)

HIBBS, EUTHYMIA (Mrs), 7302 Durbin Terr.,

West Bethesda, MD 20817 (M)

HICKOX, GEORGE H. (Dr), 9310 Allwood Ct.,

Alexandria, VA 22309 (E)

HILDEBRAND, EARL M. (Dr), 11092 Timberline

Dr., Sun City, AZ 85351 (E)

HILL, FREEMAN K. (Dr), 12408 Hall’s Shop

Road, Fulton, MD 20759 (F)

HILLABRANT, WALTER J. (Dr), 421 Butternut

St., NW, Washington, DC 20012 (M)

HILSENRATH, JOSEPH (Mr), 9603 Brunett Ave.,

Silver Spring, MD 20901 (F)

HOBBS, ROBERT B. (Dr), 7715 Old Chester Rad.,

Bethesda, MD 20817 (F)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

HOFFELD, J. TERRELL (Dr), Clinical Invest.

Sec., CIPCB, NIDR, NIH, Washington, DC

20205 (M)

HOFFMAN, C. H. (Dr), 6906 40th Ave., Hyatts-

ville, MD 20782 (E)

HOGE, HAROLD J. (Dr), 5 Rice Spring Lane,

Wayland, MA 01778 (E)

HOLSHOUSER, WILLIAM L. (Mr), Rte. 2, Bx 151,

Banner Elk, NC 28604 (F)

HOLLIES, NORMAN R. S. (Dr), 9823 Singleton

Dr., Bethesda, MD 20817 (F)

HOLLOWAY, FRANK L. (Dr), 14207 Georgia

Ave., Silver Spring, MD 20906 (F)

HONIG, JOHN G. (Dr), 7701 Glenmore Spring

Way, Bethesda, MD 20817 (F)

HOOVER, LARRY A. (Mr), 801 Croydon St.,

Sterling, VA 22170 (M)

HOPP, HENRY (Dr), 6604 Michaels Dr., Bethesda,

MD 20817 (E)

HOPP, THEODORE H. (Mr), Bldg. 220, Room

A127, Nat'l. Bureau of Standards, Washington,

DC 20234 (M)

HOPPS, HOPE E. (Mrs), 1762 Overlook Dr.,

Silver Spring, MD 20903 (F)

HORNSTEIN, IRWIN (Dr), 5920 Bryn Mawr Rd.,

College Park, MD 20740 (E)

HOROWITZ, EMANUEL (Dr), 14100 Northgate

Dr., Silver Spring, MD 20906 (F)

HORTON, BILLY M. (Mr), 14250 Larchmere Blivd.,

Shaker Heights, OH 44120 (F)

HOWARD, JAMES H. (Dr), 3822 Albemarle St.,

NW, Washington, DC 20016 (F)

HOWELL, BARBARA F. (Dr), 13405 Accent Way,

Germantown, MD 20874 (M)

HUANG, KUN-YEN (Dr), 1445 Laurel Hill Rd.,

Vienna, VA 22180 (F)

HUBBARD, DONALD B. (Dr), 3821 Bevan Dr.,

Fairfax, VA 22030 (F)

HUDSON, COLIN M. (Dr), Product Planning,

Deere & Co., John Deere Rd., Moline, IL 61265

(F)

HUGH, RUDOLPH (Dr), Dept. Microbiol., GWU

Med., 2300 | St., NW, Washington, DC 20037

(F)

HUNTER, WILLIAM R. (Mr), 6705 Caneel Ct.,

Springfield, VA 22152 (F)

HUNTER, RICHARD S. (Mr), Hunter Assoc. Lab,

Inc., 11495 Sunset Hills, Reston, VA 22090 (F)

HURDLE, BURTON G. (Mr), 6222 Berkely Rd.,

Alexandria, VA 22307 (F)

HURTT, WOODLAND (Dr), USDA-SEA, P.O. Bx.

1209, Frederick, MD 21701 (M)

HUTTNER, ANDREW W. (Mr), 109 Colburn Dr.,

Manassas Park, VA 22110 (M)

HUTTON, GEORGE L. (Mr), BX 2055, So. US

421, Zionsville, IN 46077 (E)

IRVING, GEORGE W. (Dr), 4836 Langdrum Lane,

Chevy Chase, MD 20815 (F)

111

IRWIN, GEORGE R. (Dr), 7306 Edmonston Rad.,

College Park, MD 20740 (F)

ISBELL, HORACE S. (Dr), 4704 Bladgen Ave.,

NW, Washington, DC 20011 (F)

ISENSTEIN, ROBERT S. (Dr), 11710 Caverly

Ave., Beltsville, MD 20705 (M)

JACKSON, JO-ANNE (Dr), 4412 Independence

St., Rockville, MD 20853 (F)

JACKSON, PATRICIA (Mrs), USDA Plant Stress,

RM. 207, Bldg. L, ARC-West, Beltsville, MD

20705 (M)

JACOBS, WOODROW C. (Dr), 6309 Bradley

Blvd., Bethesda, MD 20817 (E)

JACOX, MARILYN E. (Dr), 10203 Kindly Ct.,

Gaithersburg, MD 20879 (F)

JAROSEWICH, EUGENE, Mineral Sci., MRC 119,

Smithsonian Inst., Washington, DC 20560 (M)

JEN, C. K. (Dr), Johns Hopkins Rd., Laurel, MD

20707 (E)

JENSEN, ARTHUR S. (Dr), Westinghouse D&E

Center, BX 1521, Baltimore, MD 21203 (F)

JOHNSON, DANIEL P. (Dr), P.O. Bx. 359, Folly

Beach, SC 29439 (E)

JOHNSON, KIETH C. (Mr), 4422 Davenport St.,

NW, Washington, DC 20016 (F)

JOHNSON, CHARLES R. (Dr), 117 Evans St.,

Rockville, MD 20850 (F)

JOHNSON, EDGAR M. (Dr), Army Res. Inst. for

Beh., 5001 Eisenhower Ave., Alexandria, VA

22333 (F)

JOHNSON, PHYLLIS T. (Dr), Nat’! Marine Fisher-

ies Serv., Oxford Lab, Oxford, MD 21654 (F)

JONES, HOWARD S., JR. (Dr), 6200 Sligo Mill

Rd., NE, Washington, DC 20010 (F)

JONG, SHUNG-CHANG (Dr), Amer. Type Cult.

Collect., 12301 Parklawn Dr., Rockville, MD

20852 (F)

JORDAN, GARY B. (Dr), 1012 Olmo Ct.,

Jose, CA 95129 (M)

KAISER, HANS E. (Dr), 433 South West Dr.,

Silver Spring, MD 20901 (M)

KAPETANAKOS, C. A. (Dr), 6101 Overlead Rd.,

Bethesda, MD 20816 (F)

KARR, PHILIP R. (Dr), 5507 Calle De Arboles,

Torrance, CA 90505 (E)

KAUFMAN, H. PAUL (Lt Col),

Fedhaven, FL 33854-1135 (F)

KEARNEY, PHILIP C. (Dr), 8416 Shears Ct.,

Laurel, MD 20707 (F)

KEBABIAN, JOHN W. (Dr), 609 Muriel St., Rock-

ville, MD 20852 (F)

KEGELES, GERSON (Dr), 6 Oakwood Drive,

Stafford Springs, CT 07076 (F)

San

P'O. Box- 1135,

112

KEISER, BERNHARD E. (Dr), 2046 Carrhill Rd.,

Vienna, VA 22180 (F)

KESSLER, KARL G. (Dr), 5927 Anniston Rd.,

Bethesda, MD 20817 (F)

KEULEGAN, GARBIS H. (Dr), 215 Buena Vista

Dr., Vicksburg, MS 39180 (M)

KLEBANOFF, PHILIP S. (Mr), 6412 Tone Dr.,

Bethesda, MD 20817 (F)

KLINGSBERG, CYRUS (Dr), Apt. 1105E, 4620 N.

Park Ave., Chevy Chase, MD 20815 (F)

KNOBLOCK, EDWARD C. (Col), 7767 Dollyhyde

Rd., Mt. Airy, MD 21771 (F)

KNOWLTON, KATHRYN (Dr), Apt. #837, 2122

Mass. Ave., NW, Washington, DC 20008 (F)

KNOX, AUTHOR S. (Mr), 2006 Columbia Rd.,

Washington, DC 20009 (M)

KNUTSON, LLOYD V. (Dr), Rm. 001, Bldg. 003,

Agr. Res. Cntr., Beltsville, MD 20705 (F)

KRAMER, CAROLYN M. (Dr), Gillette Co., GATL,

83 Rodgers St., Cambridge, MA 02142 (F)

KROP, STEPHEN (Dr), 7908 Birham Wood Dr.,

McLean, VA 22102 (F)

KRUGER, JEROME (Dr), 619 Warfield Dr., Rock-

ville, MD 20850 (F)

LANDSBERG, HELMUT E. (Dr), 5116 Yorkville

Rd., Temple Hills, MD 20748 (F)

LANG, MARTHA E. C. (Mrs), 3133 Conn. Ave.,

NW, Washington, DC 20008 (F)

LANGFORD, GEORGE S. (Dr), 4606 Hardwick

Rd., College Park, MD 20740 (E)

LANGSTON, JOANN H. (Ms), 14514 Faraday

Dr., Rockville, MD 20853 (F)

LAPHAM, EVAN G. (Mr), 2202S.E. 29th St., Cape

Coral, FL 33904 (E)

LAWSON, ROGER H. (Dr), 4912 Ridgeview Lane,

Bowie, MD 20715 (F)

LE CLERG, ERWIN L. (Dr), 14620 Deerhurst

Terr., Silver Spring, MD 20906 (E)

LEE, RICHARD H. (Dr), R.D. 2, Box 143E, Lewes,

DE 19958 (E)

LEIBOWITZ, LAWRENCE M. (Dr), 9704 Gals-

worth Court, Fairfax, VA 22032 (F)

LEIBOWITZ, JACK, R. (Dr), 12608 Dayan Be

Silver Spring, MD 20904 (F)

LEINER, ALAN L. (Mr), Apt. 804, 580 Arastradero

Rd., Palo Alto, CA 24306 (E)

LEJINS, PETER P. (Dr), 7114 Eversfield Dr.,

College Heights Estates, Hyattsville, MD 20782

(F)

LENTZ, PAUL L. (D

MD 20770 (F)

LESSOFF, HOWARD (Mr), Code 6820, Naval

Res. Lab, Washington, DC 20375 (F)

LEVISON, N. S. (Dr), CTA-Hurst 206, American

Univ., Washington, DC 20016 (M)

LEVY, SAMUEL (Mr), 2279 Preisman Dr., Sche-

nectady, NY 12309 (E)

r), 5 Orange Ct., Greenbelt,

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

LEWIS, MARY J. (Ms), American Fisheries So-

ciety, 5410 Grisvenor Lane, Bethesda, MD

20814 (M)

LIEBLEIN, JULIUS (Dr), 1621 E. Jefferson St.,

Rockville, MD 20852 (E)

LIN, MING-CHANG (Dr), 8897 McNair Dr., Alex-

andria, VA 22309 (F)

LINDSEY, IRVING (Mr), 202 E. Alexandria Ave.,

Alexandria, VA 22302 (E)

LING, LEE (Mr), 1608 Belvoir Dr., Los Altos, CA

94022 (E)

LINK, GARY L. (Mr), 2607 Oakton Glen Dr.,

Vienna, VA 22180 (M)

LINK, CONRAD B. (Dr), Dept. of Horticulture,

Univ. of Md., College Park, MD 20782 (F)

LIST, ROBERT J. (Mr), 1123 Francis Hammond

Pky., Alexandria, VA 22302 (E)

LOCKARD, J. DAVID (Dr), Botany Dept., Univ. of

Md., College Park, MD 20742 (F)

LOEBENSTEIN, WILLIAM V. (Dr), 8501 Sundale

Dr., Silver Spring, MD 20910 (F)

LONG, BETTY JANE (Mrs), 416 Riverbend Rad.,

Ft. Washington, MD 20744 (F)

LORING, BLAKE M. (Dr), Route 2, Box 137, La-

conia, NH 03246 (F)

LUSTIG, ERNEST (Dr), Ges. Biotechnol. Forsch,

Mascheroder Weg 1, D-3300, Braunschwig,

Germany (F)

LYNCH, THOMAS J. (Mrs), 1062 Harriman St.,

Great Falls, VA 22066 (M)

LYONS, JOHN W. (Dr), 7430 Woodville Rd., Mt.

Airy, MD 21771 (F)

MADDEN, JEREMIAH J. (Mr), Goddard Space

Fit. Ctr., Code 403, Greenbelt, MD 20771 (F)

MADDEN, ROBERT P. (Dr), A-251 Physics Bldg.,

Nat'l Bureau of Standards, Washington, DC

20234 (F)

MAENGWIN-DAVIES, G. D. (Dr), 15205 Totten-

ham Terr., Silver Spring, MD 20906 (E)

MAHAN, A. |. (Dr), 1128 Spotswood Dr., Silver

Spring, MD 20904 (E)

MAIENTHAL, MILLARD (Dr), 10116 Bevern Lane,

Potomac, MD 20854 (E)

MALONE, THOMAS B. (Dr), 6633 Kennedy Lane,

Falls Church, VA 22042 (F)

MANDEL, JOHN (Dr), B-356 Chemistry Bldg.,

Nat'l Bureau of Standards, Washington, DC

20234 (F)

MANDERSCHEID, RONALD W. (Dr), 6 Monu-

ment Ct., Rockville, MD 20850 (F)

MANNING, JOHN R. (Mr), 9725 Digging Rd.,

Gaithersburg, MD 20879 (F)

MARCUS, MARVIN (Dr), Univ. of Cal., Santa

Barbara, CA 93106 (F)

MARTIN, EDWARD J. (Dr), 7721 Dew Wood Dr.,

Derwood, MD 20855 (F)

MARTIN, JOHN H. (Dr), 124 NW 7th St., #303,

Corvallis, OR 97330 (E)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

MARTIN, ROBERT H. (Mr), 2257 N. Nottingham

St., Arlington, VA 22205 (E)

MARTIN, ROY E., (Mr), Nat'l Fisheries Institute,

1101 Connecticut Ave., NW, Washington, DC

20036 (M)

MARTON, L. (Dr), Editorial Office, 4515 Linnean

Ave., NW, Washington, DC 20008 (E)

MARVIN, ROBERT S. (Dr), 11700 Stoney Creek

Rd., Potomac, MD 20854 (E)

MASON, HENRY L. (Dr), 7008 Meadow Lane,

Chevy Chase, MD 20815 (F)

MASSEY, JOE T. (Mr), 10111 Parkwood Dr.,

Bethesda, MD 20814 (F)

MATLACK, MARION B. (Dr), 2700 25th St., No.,

Arlington, VA 22207 (E)

MAYOR, JOHN R. (Dr), Asst. Provost Res. Div. of

Human & Comm. Res., College Park, MD

20742 (F)

MCKENZIE, LAWSON M. (Mr), 1902 Erie St.,

Hyattsville, MD 20783 (F)

MEADE, BUFORD K. (Mr), 5516 Bradley Blivd.,

Alexandria, VA 22311 (F)

MEARS, THOMAS W. (Mr), 2809 Hathaway Terr.,

Wheaton, MD 20906 (F)

MEARS, FLORENCE M. (Dr), 8004 Hampden

Lane, Bethesda, MD 20814 (F)

MEBS, RUSSELL W. (Dr), 6620 3rd St., No.,

Arlington, VA 22213 (F)

MELMED, ALLAN J. (Dr), 732 Tiffany Ct., Gai-

thersburg, MD 20878 (F)

MENDLESOHMN, MARK B. (Dr), 2606 Viking Dr.,

Herndon, VA 22071 (F)

MENZER, ROBERT E. (Dr), 7203 Wells Pkwy.,

Hyattsville, MD 20782 (F)

MERRIAM, CARROLL F. (Mr), Colonial Manor

Nursing Hm., 110 College Ave., Waterville, ME

04901 (F)

MESSINA, CARLA G. (Mrs), 9800 Marquette Dr.,

West Bethesda, MD 20817 (F)

MEYERSON, MELVIN R. (Dr), 611 Goldsborough

Dr., Rockville, MD 20850 (F)

MIDDLETON, H. E. (Dr), 3600 Grove Ave., Rich-

mond, VA 23221 (E)

MILLAR, DAVID B. (Dr), 1716 Mark Lane, Rock-

ville, MD 20852 (F)

MILLER, CARL F. (Dr), P.O. Box 127, Gretna, VA

24557 (E)

MILLER, MARGARET D. (Dr), 11632 Deborah

Dr., Potomac, MD 20854 (E)

MILLER, ROMAN R. (Mr), 1232 Pinecrest Circle,

Silver Sping, MD 20910 (F)

MILLER, PAUL R. (Dr), 207 South Pebble Beach,

Sun City Center, FL 33570 (E)

MITCHELL, J. MURRAY JR. (Dr), 1106 Dogwood

Dr., McLean, VA 22101 (F)

MITTLEMAN, DON (Dr), 80 Parkwood Lane,

Oberlin, OH 44074 (F)

MIZELL, LOUIS R. (Mr), 108 Sharon Lane, Green-

lawn, NY 11740 (F)

MOLINO, JOHN A. (Dr), 102 Spring St., Gai-

thersburg, MD 20877 (M)

MOLLARI, O. MARIO (Prof), 4527 45th St., NW,

Washington, DC 20016 (E)

113

MOORE, JAMES G. (Mr), CRS, Library of Con-

gress, Washington, DC 20540 (M)

MOORE, GEORGE A. (Dr), 1108 Agnew Dr.,

Rockville, MD 20851 (E)

MORRIS, J. ANTHONY (Dr), 23-E Ridge Rad.,

Greenbelt, MD 20770 (M)

MORRIS, MARLENE C. (Mrs), 1448 Leegate Rd.,

NW, Washington, DC 20012 (F)

MORRIS, JOSEPH B. (Dr), Chemistry, Howard

Univ., Washington, DC 20059 (F)

MORRISS, DONALD J. (Mr), 102 Baldwin Ct.,

Point Charlotte FL 33950 (E)

MOSTOFI, F. K. (M.D.), Armed Forces Inst. of

Pathology, Washington, DC 20306 (F)

MOUNTAIN, RAYMOND D. (Dr), B216, Physics

Bidg., Nat'l Bureau of Standards, Washington,

DC 20234 (F)

MUELHAUSE, C. O. (Dr), 9105 Seven Locks Rd.,

West Bethesda, MD 20817 (E)

MUESEBECK, CARL F. W. (Mr), 18 North Main

St., Elba, NY 14058 (E)

MULLIGAN, JAMES H. JR. (Dr), 12121 Sky Lane,

Santa Ana, CA 92705 (F)

MUMMA, MICHAEL J. (Dr), 210 Glen Oban Dr.,

Arnold, MD 21012 (F)

MURDAY, JAMES S. (Dr), 7116 Red Horse Tav-

ern Rd., West Springfield, VA 22153 (F)

MURDOCH, W. P. (Dr), Walnut Ridge Stock

Farm, R.D. #2, Gettysburg, PA 17325 (F)

MURPHY, THOMAS J. (Mr), 6521 Divine St.,

McLean, VA 22101 (F)

MURRAY, WILLIAM S. (Dr), 1281 Bartonshire

Way, Rockville, MD 20854 (F)

MURRAY, THOMAS H. (Lt Col Ret), 2915 27th

St., No., Arlington, VA 22207

MYERS, RALPH D. (Dr), 4611 Guilford Rd., Col-

lege Park, MD 20740 (E)

MC BRIDE, GORDON W. (DrM), 3323 Stuyves-

ant PI., Washington, DC 20015 (E)

MC CONNELL, DUDLEY G. (Dr), 926 Clintwood

Dr., Silver Spring, MD 20902 (F)

MC COLLOUGH, JAMES M. (Dr), 6209 Apache

St., Springfield, VA 22150 (M)

MC COLLOUGH, NORMAN B. (Dr), 6 Apple

Blossom Lane, Okemos, MI 48864 (E)

MC CURDY, JOHN DENNIS (Dr), 5531 Green

Dory Lane, Columbia, MD 21044 (F)

MC ELROY, JOHN H. (Dr), 1794 Stonegate Ave.,

Crofton, MD 21114 (F)

MC NESBY, JAMES R. (Dr), 13308 Valley Dr.,

Rockville, MD 20850 (E)

MC PHERSON, ARCHIBALD T. (Dr), 403 Russel

Ave., Apt. 808, Gaithersburg, MD 20877 (L)

NAESER, CHARLES R. (Dr), 6654 Van Winkle

Dr., Falls Church, VA 22044 (E)

NAIDEN, EULAINE (Dr), 6107 Roseland Dr.,

Rockville, MD 20852 (M)

114

NAMIAS, JEROME (Mr), Scripps Inst. Oceanog-

raphy, 2251 Sverdrup Hall, La Jolla, CA 92093

NAUGLE, JOHN E. (Mr), 7211 Rollingwood Dr.,

Chevy Chase, MD 20815 (E)

NEALE, JOSEPH H. (Dr), Biology Dept., Rm.

406, Reiss Sci. Bldg., Georgetown University, .

Washington, DC 20057 (E)

NEF, EVELYN S. (Mrs), 2726 N St. NW, Washing-

ton, DC 20007 (M)

NELSON, R. H. (Mr), Bethany Village, 512 Al-

bright Dr., Mechanicsburg, PA 17055 (E)

NEPOMUCENE, Sr., St. John Villa Julie Resi-

dence, Valley Rd., Stevenson, MD 21153 (E)

NEUBAUER, WERNER G. (Dr), 4603 Quarter

Charge Dr., Annandole, VA 22203 (F)

NEUENDORFFER, J. A. (Dr), 911 Allison St.,

Alexandria, VA 22302 (E)

NEUPERT, WERNER \M. (Dr), NASA/GSFC, Code

682, Greenbelt, MD 20771 (F)

NEUSCHEL, SHERMAN K. (Dr), 7501 Democ-

racy Blvd., Bethesda, MD 20817 (F)

NEWMAN, MORRIS (Dr), 1050 Las Alturas Rd.,

Santa Barbara, CA 93103 (F)

NICKERSON, DOROTHY (Miss), Apt. 450, 4800

Fillmore Ave., Alexandria, VA 22311 (E)

NOFFSINGER, TERRELL L. (Dr), Rte. #1, Box

305, Auburn, KY 42206 (F)

NORRIS, KARL K. (Mr), 11204 Montgomery Rad.,

Beltsville, MD 20705 (F)

O’CONNOR, JAMES V. (Mr), 10108 Haywood

Circle, Silver Spring, MD 20902 (M)

O’HARE, JOHN J. (Dr), 301 G St., SW, Washing-

ton, DC 20024 (F)

O’HERN, ELIZABETH M. (Dr), 633 G St., SW,

Washington, DC 20024 (F)

O’KEEFE, JOHN A. (Dr), NASA/ GSFC, Code

681, Greenbelt, MD 20071 (F)

OBERLE, MARILYN E. (Ms), Apt. 622, 2801

Quebec St., NW, Washington, DC 20008 (M)

OEHSER, PAUL H. (Mr), 9012 Old Dominion Dr.,

McLean, VA 22102 (E)

OGDIN, CAROL A. (Ms), 114 Harvard St., Alex-

andria, VA 22314 (M) :

OKABE, HIDEO (Dr), 6700 Old Stage Rd., Rock-

ville, MD 20852 (F)

OLIPHANT, MALCOLM W (Dr), 1606 Ulupii St.,

Kailua, Hl 96734 (F)

ORDWAY, FRED D., JR. (Dr), 5205 Elsmere Ave.,

Bethesda, MD 20814 (F)

OSER, HANS L. (Dr), 8810 Quiet Stream Ct.,

Potomac, MD 20854 (F)

OTA, HAJIME (Mr), 5708 64th Avenue, East River-

dale, MD 20737 (F)

OUSLEY, GILBERT W. (Dr), NASA, Code 404,

Goddard Space Flight Cntr., Greenbelt, MD

20771 (F)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983 _

i i i a Dre 5s

re .

PAFFENBARGER, GEORGE C. (Dr), ADA Hlth.

Fdn. Res. Unit, Nat'l Bureau of Standards,

Washington, DC 20234 (F)

PAPADOPOULOS, KOSTANTINOS (Dr), 6346

32nd St., NW, Washington, DC 20015 (F)

PARKER, ROBERT L. (Dr), 9728 Digging Rad.,

Gaithersburg, MD 20879 (F)

PARMAN, GEORGE K. (Mr), 8054 Fairfax Rd.,

Alexandria, VA 22308 (F)

PARRY-HILL, JEAN (Ms), 3803 Military Rd., NW,

Washington, DC 20015 (M)

PARSONS, H. MCILVAINE (Dr), Essex Corp., 333

N. Fairfax St., Alexandria, VA 22314 (F)

PELCZAR, MICHAEL J. (Dr), 4318 Clagett Pine-

way, University Park, MD 20782 (F)

PELLERIN, CHARLES J. (Dr), NASA, Code EZ-7,

600 Independence Ave., SW, Washington, DC

20546 (F)

PERKINS, LOUIS R. (Mr), 1234 Massachusetts

Ave., NW, Washington, DC 20005 (M)

PERROS, THEODORE (Dr), 5825 3rd PI., NW,

Washington, DC 20011 (F)

PHAIR, GEORGE (Dr), 14700 River Rd., Potomac,

MD 20854 (F)

PIEPER, GEORGE F. (Dr), 3155 Rolling Rd.,

Edgewater, MD 21037 (F)

PIKL, JOSEF M. (Dr), 211 Dickinson Rd., Glass-

boro, NJ 08028 (E)

PITTMAN, MARGARET (Dr), 3133 Connecticut

Ave., NW, Washington, DC 20008 (E)

PITTS, JOHN A. S. (Dr), 11527 Hearthstone Ct.,

Reston, VA 22091 (M)

PLAIT, ALAN O. (Mr), 5402 Yorkshire St., Spring-

field, VA 22151 (F)

POLACHEK, HARRY (Dr), 11801 Rockville, MD

20852 (F)

PONANDER, HEATHER B. (Mrs), Geology Dept.,

Stanford Univ., Stanford, CA 94305 (M)

PONNAMPERUMA, CYRIL (Dr), Dept. of Chem-

istry, Univ. of Md., College Park, MD 20742 (F)

POOS, FRED W. (Dr), 5100 Fillmore Ave., Alex-

andria, VA 22311 (E)

POST, MILDRED A. (Miss), 8928 Bradmoor Dr.,

West Bethesda, MD 20817 (F)

PRESLEY, JOHN T. (Dr), 3811 Courtney Cir.,

Bryan, TX 77801 (E)

PRESTON, MALCOMS. (Dr), 10 Kilkea Ct., Balti-

more, MD 21235 (M)

PRINCE, J. S. (Dr), 7103 Pinehurst Parkway,

Chevy Chase, MD 20815 (F)

PRINZ, DIANNE K. (Dr), Code 4142 Naval Res.

Lab, Washington, DC 20375 (M)

PRO, MAYNARD J. (Mr), 7904 Falstaff Rd.,

McLean, VA 22101 (F)

PRYOR, C. NICHOLAS (Dr), Bleak House, Atlan-

tic Ave., Newport, RI 02840 (F)

PUGH, MARION S. (Mr), Little Fiddlers Green,

Round Hill, VA 22141 (M)

PURCELL, ROBERT H. (Dr), 17517 White

Grounds Rd., Boyds, MD 20841 (F)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

PYKE, THOMAS N. JR. (Mr), Technology Bldg.

A-247, Nat'l Bureau of Standards, Washington,

DC 20234 (F)

QUANN, JOHN J. (Mr), 7511 Broken Staff, Co-

lumbia, MD 21045 (F)

QUIROZ, RODERICK S. (Mr), 4520 Yuma St.,

NW, Washington, DC 20016 (F)

RABINOW, JACOB (Mr), 6920 Selkirk Dr., Be-

thesda, MD 20817 (F)

RADER, CHARLES A. (Mr), Gillette Res. Inst.,

1413 Research Blvd., Rockville, MD 20850 (F)

RADO, GEORGE T. (Dr), 818 Carrie Court,

McLean, VA 22101 (F)

RAINWATER, IVAN H. (Dr), 2805 Liberty PI.,

Bowie, MD 20715 (E)

RALEIGH, LANI H. (Ms), 8491 Imperial Dr., Lau-

rel, MD 20708 (M)

RAMSAY, MAYNARD J. (Dr), 3806 Viser Ct.,

Bowie, MD 20715 (F)

RANEY, WILLIAM P. (Dr), 5946 Wilton Rd., Alex-

andria, VA 22310 (M)

RANGO, ALBERT (Dr), 127 Southwood Ave.,

Silver Spring, MD 20901 (F)

RANSOM, JAMES R. (Mr), 107 E. Susquehanna

Ave., Towson, MD 21204 (M)

RAUSCH, ROBERT L. (Dr), Div. of Animal Med.,

SB-42, Univ. of Wash., Seattle, WA 98195 (F)

RAVITSKY, CHARLES (Mr), 1505 Drexel St.,

Takoma Park, MD 20912 (E)

RAY, JOSEPH H. (Dr), 2740 Vassar PI., Colum-

bus, OH 43221 (F)

READING, O.S. (Capt), Bellport, 6 N. Howells Pt.

Rd., Suffolk City, NY 11713 (E)

REED, WILLIAM D. (Mr), Apt. 708, 4740 Connec-

ticut Ave., NW, Washington, DC 20008 (F)

REHDER, HARALD A. (Dr), 5620 Ogden Rd., Be-

thesda, MD 20816 (F)

REINER, ALVIN (Mr), 11243 Bybee St., Silver

Spring, MD 20902 (M)

REINHART, FRANK W. (Dr), 9918 Sutherland

Rd., Silver Spring, MD 20901 (F)

REMMERS, GENE M. (Mr), 7322 Craftown Rd.,

Fairfax, VA 22039 (M)

REYNOLDS, ORR E. (Dr), Amer. Physiological

Soc., 9650 Rockville Pike, Bethesda, MD 20814

(F)

REYNOLDS, HORACE N. (Dr), 9223 Woodland

Dr., Silver Spring, MD 20910 (F)

RHODES, IDA (Mrs), 6676 Georgia Ave., NW,

Washington, DC 20012 (E)

RHYNE, JAMES J. (Dr), 20505 Dubois Ct., Gai-

thersburg, MD 20879 (F)

115

RICE, ROBERT L. (Mr), 12041 Winding Creek

Way, Germantown, MD 20874 (M)

RICE, FREDERICK A. H. (Dr), 8005 Carita Ct.,

Bethesda, MD 20817 (F)

RIOCH, DAVID McK. (Dr), 5525 Surrey St., Chevy

Chase, MD 20815 (E)

RITT, PAUL E. (Dr), 36 Sylvan Lane, Weston, MA

02193 (F)

RIVLIN, RONALD S. (Dr), Lehigh Univ. Linder-

man Library, #30, Bethlehem, PA 18015 (F)

ROBBINS, MARY L. (Dr), Tatsuno House, A-23,

2-1-8 Ogikubo Suginami-Ku, Tokyo 167, Japan

(E)

ROBERTS, RICHARD C. (Dr), 5170 Phantom Ct.,

Columbia, MD 21044 (F)

ROBERTS, ELLIOTT B. (Capt), USC & GS, 4500

Wetherill Rd., Bethesda, MD 20816 (E)

ROBERTSON, A. F. (Dr), 4228 Butterworth PI.;

NW, Washington, DC 20016 (F)

ROBERTSON, RANDAL M. (Dr), 1404 Highland

Circle, SE, Blacksburg, VA 24060 (E)

ROCK, GEORGE D. (Mr), 3133 Connecticut

Ave., NW, Washington, DC 20008 (E)

RODNEY, WILLIAM S. (Dr), 8112 Whites Ford

Way, Rockville, MD 20854 (F)

ROLLER, PAUL S. (Dr), Apt. 1011, 1440 N St.,

NW, Washington, DC 20005 (E)

ROSADO, JOHN A. (Mr), 10519 Edgemont Dr.,

Adelphi, MD 20783 (F)

ROSCHER, NINAM. (Dr), 1040 Hunter Ridge Dr.,

Oakton, VA 22124 (F)

ROSE, WILLIAM K. (Dr), 10916 Picasso Lane,

Potomac, MD 20854 (F)

ROSENBLATT, JOAN R. (Dr), 2939 Van Ness St.,

#702, Washington, DC 20008 (F)

ROSENBLATT, DAVID (Prof), 2939 Van Ness St.,

#702, Washington, DC 20008 (F)

ROSENTHAL, SANFORD M. (Dr), 12601 Green-

brier Rd., Potomac, MD 20854 (E)

ROSS, SHERMAN, (Dr), 19715 Greenside Terr.,

Gaithersburg, MD 20879 (F)

ROSS, FRANKLIN J. (Mr), 2816 N. Dinwiddie St.,

Arlington, VA 22207 (F)

ROSSINI, FREDERICK D. (Dr), Apt. T-900, 605

South U.S. Highway #1, Juno Beach, FL 33408

(E)

ROTH, FRANK L. (Mr), Apt. 33, 20 E. 22nd St.,

Roswell, NM 88201 (E)

ROTKIN, ISRAEL (Mr), 11504 Regnid Dr., Whea-

ton, MD 20902 (E)

RUPP, NELSON W. (Dr), 9125 Levelle Dr., Chevy

Chase, MD 20815 (F)

RUSSELL, LOUISE M. (Miss), 9 Sunnyside Rd.,

Silver Spring, MD 20910 (F)

RYERSON, KNOWLES A. (Dean), 15 Arimonte

Dr., Kensington, CA 94707 (E)

SAENZ, ALBERT W. (Dr), Code 6603S, Naval

Res. Lab, Washington, DC 20375 (F)

116

SAILER, REECE |. (Dr), 3847 S.W. Sixth PI.,

Gainesville, FL 32607 (F)

SALISBURY, LLOYD L. (Mr), 10138 Crestwood

Rd., Kensington, MD 20895 (M)

SALLET, DIRSE W. (Dr), 12440 Old Fletchertown

Rd., Bowie, MD 20715 (M)

SANDERSON, JOHN A. (Dr), B-206 Clemson

Downs, Clemson, SC 29631 (E)

SARMIENTO, RAFAEL, 5426 30th St., NW, Wash-

ington, DC 20015 (F)

SASMOR, ROBERT M. (Dr), 4408 North 20th Rad.,

Arlington, VA 22207 (F)

SASS, ARTHUR H. (Mr), RFD 3, Box 423 A, War-

renton, VA 22186 (M)

SAVILLE, THORNDIKE JR. (Mr), 5601 Albia Rd.,

Bethesda, MD 20816 (F)

SCHALK, JAMES M. (Dr), 267 Forest Trail, Isle of

Palms, SC 29451 (F)

SCHECHTER, MILTON S. (Mr), 10909 Hannes

Ct., Silver Spring, MD 20901 (E)

SCHINDLER, ALBERT I. (Dr), Code 6000, Naval

Res. Lab, Washington, DC 20375 (F)

SCHLAIN, DAVID (Dr), P.O. Box 348, College

Park, MD 20740 (F)

SCHMIDT, CLAUDE H. (Dr), 1827 Third St. No.,

Fargo, ND 58102 (F)

SCHNEIDER, SIDNEY, (Mr), 239 North Granada

St., Arlington, VA 22203 (E)

SCHNEPFE, MARIAN M. (Dr), Potomac Towers

Apts. #640, 2001 North Adams St., Arlington,

VA 22201 (E)

SCHOOLEY, JAMES F. (Dr), 13700 Darnestown

Rd., Gaithersburg, MD 20878 (F)

SCHUBAUER, GALEN B. (Dr), 5609 Gloster Rd.,

Bethesda, MD 20816 (F)

SCHULMAN, FRED (Dr), 1115 Markwood Dr.,

Silver Spring, MD 20902 (F)

SCHULMAN, JAMES H. (Dr), 5628 Massachu-

setts Ave., Bethesda, MD 20816 (E)

SCHWARTZ, ANTHONY M. (Dr), 2260 Glenmore

Terr., Rockville, MD 20816 (E)

SCOTT, DAVID B. (DDS), 10448 Wheatridge Dr.,

Sun City, AZ 85373 (F)

SCRIBNER, BOURDON F. (Mr), 123 Peppercorn

Pl., Edgewater, MD 21037 (E)

SEABORG, GLENN T. (Dr), 1154 Glen Rd., La-

fayette, CA 94549 (F)

SEEGER, RAYMOND H. (Dr), 4507 Wetherill Rd.,

Bethesda, MD 20816 (E)

SEITZ, FREDERICK (Dr), Rockefeller Univ., New

York, NY 10021 (F)

SHAFRIN, ELAINE G. (Mrs), Apt N-702, 800 4th

St., SW, Washington, DC 20024 (F)

SHAPIRO, GUSTAVE (Mr), 3704 Munsey St.,

Silver Spring, MD 20906 (F)

SHEAR, RALPH E. (Mr), 1916 Bayberry Rd.,

Edgewood, MD 21040 (M)

SHELTON, EMMA (Dr), 8410 Westmont Terr.,

West Bethesda, MD 20817 (F)

SHEPARD, HAROLD H. (Dr), 2701 South June

St., Arlington, VA 22202 (E)

SHERESHEFSKY, J. LEON (Dr), 9023 Jones Mill

Rd., Chevy Chase, MD 20815 (E)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

SHERLIN, GROVER C. (Mr), 4024 Hamilton St.,

Hyattsville, MD 20781 (L)

SHIER, DOUGLAS R. (Dr), Dept. Mathematical

Science, Clemson University, Clemson, SC

29631 (F)

SHOTLAND, EDWIN (Dr), 418 E. Indian Spring

Dr., Silver Spring, MD 20901 (M)

SHROPSHIRE, W. JR. (Dr), Radiation Biol. Lab.,

12441 Parklawn Dr., Rockville, MD 20852 (F)

SHUBIN, LESTER D. (Mr), 9108 Saranac Ct.,

Fairfax, VA 22032 (F)

SILVER, DAVID M. (Dr), Applied Physics Lab,

11100 Johns Hopkins Rd., Laurel, MD 20707

(M)

SIMHA, ROBERT (Dr), Case-Western Reserve

Univ., University Circle, Cleveland, OH 44106

(F)

SIMMONS, LANSING G. (Mr), 3800 N. Fairfax

Dr., Villa 809, Arlington, VA 22203 (F)

SLACK, LEWIS (Dr), 27 Meadow Bank Rad., Old

Greenwich, CT 06870 (F)

SLAWSKY, MILTON M. (Dr), 8803 Lanier Dr.,

Silver Spring, MD 20910 (E)

SLAWSKY, ZAKA I. (Dr), Apt. #318, 4701 Willard

Ave., Chevy Chase, MD 20815 (F)

SLOCUM, GLENN G (Mr), 4204 Dresden St.,

Kensington, MD 20895 (E)

SMITH, MARCIA S. (Ms), 6015 North 9th St., Ar-

lington, VA 22205 (M)

SMITH, BLANCHARD D., JR. (Mr), 2509 Ryegate

Lane, Alexandria, VA 22308 (F)

SMITH, ELSKE V. P. (Dr), 8 Waterfall Rd., Rich-

mond, VA 23228 (F)

SMITH, FRANCIS A. (Dr), 5023 55th Ave., So., St.

Petersburg, FL 33705 (E)

SMITH, ROBERT C., JR. (Mr), 6151A Edsall Rd.,

Alexandria, VA 22304 (F)

SMITH, FLOYD F. (Dr), 9022 Fairview Rd., Silver

Spring, MD 20910 (E)

SMITH, JACK C. (Dr), 3708 Manor Rd., #3, Chevy

Chase, MD 20815 (F)

SNAVELY, BENJAMIN L. (Dr), 360 Blossom Hill

Dr., Lancaster, PA 17601 (F)

SNAY, HANS G. (Dr), 17613 Treelawn Dr., Ash-

ton, MD 20861 (E)

SNYDER, HERBERT H. (Dr), RFD 1 A-1, Box 7,

Cobden, IL 62920 (F)

SOKOLOVE, FRANK L. (Mr), 3015 Graham Ra.,

Falls Church, VA 22042 (M)

SOLAND, RICHARD M., 5460 Fillmore Ave.,

Alexandria, VA 22311 (F)

SOLOMON, EDWIN M. (Mr), 1121 University

Bivd., West, Silver Spring, MD 20902 (M)

SOMERS, IRA |. (Dr), 1511 Woodacre Dr.,

McLean, VA 22101 (M)

SOMMER, HELMUT (Dr), 9502 Hollins Ct., Be-

thesda, MD 20817 (F)

SORROWS, HOWARD E. (Dr), 8820 Maxwell Dr.,

Potomac, MD 20854 (F)

SPATES, JAMES E., 8609 Irvington Ave., Be-

thesda, MD 20817 (F)

SPECHT, HEINZ (Dr), 311 Oakridge Ave., Sche-

nectady, NY 12306 (E)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

SPENCER, LEWIS V. (Dr), P.O. Box 3206, Gai-

thersburg, MD 20878-0206 (F)

SPERLING, FREDERICK (Dr), 1110 Fiddler Lane,

Silver Spring, MD 20910 (E)

SPIES, JOSEPH R. (Dr), 507 North Monroe St.,

Arlington, VA 22201 (E)

SPILHAUS, A. F. JR. (Dr), 10900 Picasso Lane,

Potomac, MD 20854 (F)

SPRAGUE, G. F. (Dr), Agronomy Dept., Univ. of

Illinois, Urbana, IL 61801 (E)

STAUSS, HENRY E. (Dr), 8005 Washington Ave.,

Alexandria, VA 22308 (F)

STEELE, LENDELL E. (Mr), 7624 Highland St.,

Springfield, VA 22150 (F)

STEERE, RUSSELL L. (Dr), 6207 Carrollton Terr.,

Hyattsville, MD 20781 (F)

STEGUN, IRENE A. (Ms), Nat’l Bureau of Stand-

ards, Washington, DC 20234 (F)

STEINBERG, ALFRED D. (Dr), 8814 Bells Mill

Rd., Potomac, MD 20854 (F)

STEINER, ROBERT F. (Dr), 2609 Turf Valley

Road, Ellicott City, MD 20143 (F)

STEPHENS, ROBERT E. (Dr), 4301 39th St., NW,

Washington, DC 20016 (E)

STERN, KURT H. (Dr), Naval Res. Lab., Code

6130, Washington, DC 20374 (F)

STEWART, T. DALE (M.D.), 1191 Crest Lane,

McLean, VA 22101 (E)

STEWART, KENNETH R. (Mr), Apt. #16, 12907

Crookston Lane, Rockville, MD 20851 (M)

STIEF, LOUIS J. (Dr), Code 691, Goddard Space

Fit. Ctr., Greenbelt, MD 20771 (F)

STIEHLER, ROBERT D. (Dr), 3234 Quesada St.,

NW, Washington, DC 20015 (F)

STILL, JOSEPH W. (Dr), 1408 Edgecliff Lane,

Pasadena, CA 91107 (E)

STIMSON, HAROLD F. (Dr), 2920 Brandywine

St., NW, Washington, DC 20008 (E)

STOETZEL, MANYA B. (Dr), Systematic Ento-

mology Lab. Rm. 6, Bldg. 004, BARC-West,

Beltsville, MD 20705 (F)

STRAUSS, SIMON W. (Dr), 4506 Cedell Place,

Camp Springs, MD 20748 (L)

STRIMPLE, HARRELL L. (Mr), 904 Bowery, lowa

City, [A 52240 (F)

STUART, NEIL W. (Dr), The Lake Shore, Apt.

P-484, 5600 Lake Resort Terr., Chattanooga,

TN 37415 (E)

SULZBACHER, WILLIAM L. (Mr), 8527 Clarkson

Dr., Fulton, MD 20759 (F)

SWEZEY, ROBERT W. (Dr), Clarks Ridge Rad., Rt.

3, Box 142, Leesburg, VA 22075 (F)

SYKES, ALAN O. (Dr), 304 Mashie Dr., Vienna,

VA 22180 (M)

TALBERT, PRESTON T. (Dr), Department of

Chemistry, Howard University, Washington,

DC 20059 (F)

117

TALBOTT, F. LEO (Dr), R.D. #4, Bethlehem, PA

18015 (E)

TASAKI, ICHIJI (Dr), 5604 Alta Vista Rd., Be-

thesda, MD 20817 (F)

TATE, DOUGLAS R. (Mr), 11415 Farmland Dr.,

Rockville, MD 20852 (F)

TAYLOR, JOHN KEENAN (Dr), 12816 Tern Dr.,

Rt. 3, Gaithersburg, MD 20878 (F)

TAYLOR, BARRY N. (Dr), 11908 Tallwood Ct.,

Potomac, MD 20854 (F)

TAYLOR, ALBERT LEE (Mr), 2620 SW 14th Dr.,

Gainesville, FL 32608 (E)

TAYLOR, LAURISTON SALE (Dr), 7407 Denton

Rd., Bethesda, MD 20814 (E)

TEAL, GORDON K. (Dr), 5222 Park Lane, Dallas,

TX 75220 (F)

TERMAN, MAURICE J. (Mr), 616 Poplar Dr.,

Falls Church, VA 22046 (E)

THOMAS, WALTER |. (Dr), 523 Sunset Rd., State

College, PA 16801 (F)

THOMPSON, F. CHRISTIAN (Dr), 4255 S. 35th

St., Arlington, VA 22206 (M)

THURMAN-SWARTZWELDER, ERNESTINE (Dr),

3443 Esplanade Ave., Apt. 325, New Orleans,

LA 70119 (E)

TILDEN, EVELYN B. (Dr), Box 48, 12101 Lomas

Bivd., NE, Albuquerque, NM 87112 (F)

TODD, RUTH (Miss), P.O. Box 687, Vineyard

Haven, MA 01568 (F)

TORRENT, RAUL R., 254 Tous Soto, Baldrich

Haro Rey, PR 00918 (F)

TOUSEY, RICHARD (Dr), 7725 Oxon Hill Rd.,

Oxon Hill, MD 20745 (F)

TOUSIMIS, A. J. (Dr), P.O. Box 2189, Rockville,

MD 20852 (M)

TOWNSEND, LEWIS R. (Dr), 9900 Ashburton

Lane, Bethesda, MD 20817 (M)

TOWNSEND, CHARLES E. (M.D.), 3529 Tilden

St., NW, Washington, DC 20008 (F)

TOWNSEND, MARJORIE R. (Mrs), 3529 Tilden

St., NW, Washington, DC 20008 (F)

TRAUB, ROBERT (Col, USA Ret), 5702 Bradley

Bivd., Bethesda, MD 20814 (F)

TRUEBLOOD, EMILY E. (Dr), 7100 Armat Dr.,

West Bethesda, MD 20817 (E)

TUNELL, GEORGE (Dr), 4625 Via Gennita, Santa

Barbara, CA 93111 (E)

TURNER, JAMES H. (Dr), 11902 Falkirk Dr., Po-

tomac, MD 20854 (F)

UBERALL, HERBERT M. (Dr), Kenwood Apt.

#1417, 5100 River Road, Bethesda, MD 20816

(M)

UHLANER, J. E. (Dr), 4258 Bonavita Dr., Encino,

CA 91436 (F)

UNO, TAD (Dr), 1714 Fairview Ave., McLean, VA

22101 (F)

USDIN, VERA R. (Dr), 10472 Boca Canyon Dr.,

Santa Ana, CA 92705 (F)

118

VAN COTT, HAROLD P. (Dr), 8300 Still Spring

Ct., Bethesda, MD 20817 (F)

VAN DERSAL, WILLIAM R. (Dr), 8101 Greenspring

Ave., Baltimore County, MD 21208 (E)

VAN TUYL, ANDREW H. (Dr), 1000 West Nolcrest

Dr., Silver Spring, MD 20903 (F)

VARADI, PETER F. (Dr), Solarex Ventures Group,

1301 Piccard Dr., Rockville, MD 20850 (F)

VEITCH, FLETCHER P., JR. (Dr)., P.O. Box 513,

Lexington Park, MD 20653 (F)

VETTER, JEROME R. (Mr), 1000 Cannon Rad.,

Silver Spring, MD 20904 (M)

VILA, GEORGE J. (Mr), 5517 Westbard Ave., Be-

thesda, MD 20816 (F)

VINCENT, ROBERT C. (Dr), 2153 North Pierce

St., Arlington, VA 22209 (F)

VINTI, JOHN P. (Dr), M.I.T., Bldg. W 59-216,

Cambridge, MA 02139 (F)

VITULANO, LAWRENCE A. (Dr), G. B. Children’s

Services, 1635 Central Ave., Bridgeport, CT

06610 (M)

VON HIPPEL, ARTHUR (Dr)., 265 Glen Rd., Wes-

ton, MA 02193 (E)

WACHTMAN, JOHN B. JR. (Dr), 802 Rollins Ave.,

Rockville, MD 20852 (F)

WAGMAN, DONALD D. (Mr), 7104 Wilson Lane,

Bethesda, MD 20817 (F)

WAGNER, A. JAMES (Dr), 7007 Beverly Lane,

Springfield, VA 22150 (F)

WALKER, DELORES H. (Mrs), 2521 Branch Ave.,

SE, Washington, DC 20020 (M)

WALKER, EGBERT H. (Dr), Friend’s House, 17330

Quaker Lane, Sandy Spring, MD 20860 (E)

WALTHER, CARL H. (Dr), 1337 27th St., NW,

Washington, DC 20007 (E)

WALTON, WILLIAM W. SR. (Dr), 1705 Edgewater

Pkwy., Silver Spring, MD 20903 (F)

WARING, JOHN A. (Dr), Apt. #1, 1320 S. George

Mason Dr., Arlington, VA 22204 (M)

WATERWORTH, HOWARD E. (Dr), 10001 Frank-

lin Ave., Seabrook, MD 20706 (F)

WATSON, ROBERT B. (Dr), 1176 Wimbledon Dr.,

McLean, VA 22101 (E)

WAYNANT, RONALD W. (Dr), 13101 Claxton Dr.,

Laurel, MD 20708 (F)

WEBB, RALPH E. (Dr), 21-P Ridge Rd., Green-

belt, MD 20770 (F)

WEBER, ROBERT S. (Dr), c/o D. A. Weber, 2273

McLaughlin, Apt. #2, San Jose, CA 95122 (E)

WEBER, EUGENE W. (Dr), 2700 Virginia Ave.,

NW, Washington, DC 20037 (E)

WEIHE, WERNER K. (Dr), 2103 Bassett St., Alex-

andria, VA 22308 (E)

WEINBERG, HAROLD P. (Mr), 11410 Strand Dr.,

Bldg. 1-B #314, Rockville, MD 20252 (F)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

WEINTRAUB, ROBERT L. (Dr), 305 Fleming

Ave., Frederick, MD 21701 (E)

WEISS, ARMAND B. (Dr), 6516 Truman Lane,

Falls Church, VA 22043 (F)

WEISSLER, ALFRED (Dr), 5510 Uppingham St.,

Chevy Chase, MD 20815 (F)

WEISSLER, PEARL G. (Mrs), 5510 Uppingham

St., Chevy Chase, MD 20815 (F)

WELLMAN, FREDERICK L. (Dr), Plant Pathol-

ogy, N.C. State Univ., Raleigh, NC 27607 (E)

WENSCH, GLEN W. (Dr), 2207 Noel Dr., Cham-

paign, IL 61820 (E)

WERKHEISER, ANN (Ms), 1525 Q St., NW, Wash-

ington, DC 20009 (M)

WERTH, MICHAEL W. (Mr), 14 Grafton St.,

Chevy Chase, MD 20815 (M)

WEST, WILLIAM L. (Dr), Rm. 3412, Preclinical

Bldg., Howard Univ., Washington, DC 20059

(F)

WHITE, HOWARD J. JR. (Dr), 8028 West Over-

look Dr., West Bethesda, MD 20817 (F)

WHITE, MARVIN H. (Dr), 11176 Oakenshield Cir.,

Columbia, MD 21044 (F)

WHITELOCK, LELAND D., 2320 Brisbane St.,

Apt. #4, Clearwater, FL 33575 (F)

WHITTEN, CHARLES A. (Mr), 9606 Sutherland

Rd., Silver Spring, MD 20901 (E)

WHITTLER, RUTH G. (Dr), 83 Bay Drive, Bay

Ridge, Annapolis, MD 21403 (E)

WICHERS, EDWARD (Dr), 9601 Kingston Rad.,

Kensington, MD 20895 (E)

WEINER, ALFRED A. (Mr), 607 Janneys Lane,

Alexandria, VA 22302 (F)

WIGGINS, PEPTERF. (Dr), 1016 Harbor Dr., Ann-

apolis, MD 21403 (F)

WILDHACK, W. A. (Mr), 415 North Oxford St.,

Arlington, VA 22203 (F)

WILHELM, PETER G. (Dr), 206 Gretna Green Ct.,

Alexandria, VA 22304 (F)

WILSON, WILLIAM K. (Mr), 1401 Kurtz Rad.,

McLean, VA 22101 (F)

WILSON BRUCE L. (Mr), Apt. 204, 20 No. Leo-

nora Ave., Tucson, AZ 85711 (E)

WISTORT, ROBERT L. (Mr), 11630 35th PI.,

Beltsville, MD 20705 (F)

WOLFF, EDWARD A. (Dr), 1021 Crestview Dr.,

Silver Spring, MD 20903 (F)

J. WASH. ACAD. SCI., VOL. 73, NO. 3, 1983

WOLFSON, ROBERT P. (Mr), 10813 Larkmeade

Lane, Potomac, MD 20854 (M)

WOMACK, MADELYN (Dr), 11511 Highview Ave.,

Silver Spring, MD 20902 (F)

WOOD, LAWRENCE A. (Dr), Nat'l Bureau of

Standards, Washington, DC 20234 (E)

WOOLEY, BERNICE EDDY (Dr), 6722 Selkirk

Ct., West Bethesda, MD 20817 (E)

WORKMAN, WILLIAM G. (Dr), 5221 42nd St.,

NW, Washington, DC 20015 (E)

WULF, OLIVER R. (Dr), 557 Berkeley Ave., San

Marino, CA 91108 (E)

WYATT, DOROTHY K. (Mrs), 7924 Ilvymount

Terr., Potomac, MD 20854 (M)

YAPLEE, BENJAMIN S. (Mr), 8 Crest View Ct.,

Rockville, MD 20854 (F)

YEKOVICH, FRANK R. (Dr), School of Educa-

tion, Catholic University, Washington, DC

20064 (F)

YODER, HATTEN S., JR. (Dr), Geophysical Lab,

2801 Upton St., NW, Washington, DC 20008 (F)

YOLKEN, H. THOMAS (Dr), Perseverance Farm,

15400 Edwards Ferry Rd., Poolesville, MD

20837 (F)

YOUNG, WHARTON M. (Dr), 3230 Park PI., NW,

Washington, DC 20010 (E)

YOUNG, A. THOMAS (Mr), 947 Placid Ct., Ar-

nold, MD 21012 (F)

YOUNG, DAVIDA. JR. (Dr), 612 Buck Jones Rd.,

Raleigh, NC 27606 (E)

ZELENY, LAWRENCE (Dr), 4312 Van Buren St.,

University Park, MD 20782 (E)

ZIEN, TSE-FOU (Dr), Naval Surf. Wpns. Ctr.,

White Oak, Silver Spring, MD 20910 (F)

ZOCH, RICHMOND T. (Mr), Rte. 1, Box 26,

Shelby, AL 35143 (F)

119

aneend

rig’

a &

— q : +

-— on

- ‘

¥ cs

; 2 b

~ ~ é .

3 +

‘

= x

be ~— 5

7 :

- =

ite =

‘ , i’

‘ .

wes

ae Bs

’

DELEGATES TO THE WASHINGTON ACADEMY OF SCIENCES,

REPRESENTING THE LOCAL AFFILIATED SOCIETIES

EE <t WOSHINGION occa esac cccineecebeveaccvevesveraseseaeueweeue James F. Goff

eer ciety OF Washington ..........cccccccccccccncccvcvcccevcvevcene Ruth H. Landman

ECC WHOSTINIDTON «ccs vec ceccas cu ncccccevesccucveveucenesceven William Ronald Heyer

ee IEERIIIIOC ON) sos cc nic lds dich vucesuvinxsdevscuceuveceveutavcubecesusuws Anne Brown

EME CHE OWASTINICCON: . 0.0 os ec cnc ce cece ecccworeccseveueavegaves Margaret Collins

COE ogo ais igs aks OU Kia's vals oe dv WOR echeie cee é ee uecuaueseaneats T. Dale Stewart

ERNE VORSHINICCON ci. sec cece Nasu cccecescesvceueceuetenven’s James V. O'Connor

mnemonic Bustrict Of Columbia... ......00..ccceusccacccavescecvers Charles E. Townsend

TE ars pn cag Shes vgn w VN A's eA We walked Oe eb eke Cadualnn Paul H. Oehser

ee EPENTESIINGLOT: . oi as cian ceie crac avvececdvevnssetessactesepenanediea Conrad B. Link

IE oc uw 5 Bk Gar die kc. 8 ake Rw eM een pla we dR we Oe MOONE wee Boyd W. Post

RTE MOMNEETS . occ cette cca bawnsivineecesesuccssvsvesuccese George Abraham

een eueetrica: and Electronics Engineers ...'........cccccccceccucecvescunene George Abraham

ren WCE AMIGAl ENPINEEIS. 2... 60.0 sees ccc ececccteevessavcusevcusvenne Michael Chi

Tar OCIELY OF WaASHINPION .. 0.2... cc cece cece ce teceneeuccucuacse Robert S. Isenstein

UG WICFODIOIOLY . 1.2.2.0... ce ce ewe e eu cevececcscesusucesee Lloyd G. Herman

EE RICA INMATY ENPINCETS. 0.0.6.0. ceca scecvecscucvacuccestvessaseee H. P. Demuth

EEMOAME ENGINCETS 2c. ce ey ce eee neces cov eadevenscusevueeeus Wallace J. Cohen

i ——eeertmental Biology and Medicine...............c2ccccceeeccsrscecus Cyrus R. Creveling

EE UMENOINECETIS 5g. acs cess cs eee tec eaaeneatedeecnneuawenueneee Charles G. Interrante

emeemnrinenoo: Dental Resedrch |. 2.2.2... 020. c eee n cscs cceeeenceuccuce William R. Cotton

ean institute of Aeronautics and Astronautics..... 200... 6.. cece cease neces Richard P. Hallion

i EETICU ON CIGICEY C6 2c a od sae swe eee sen aces shinemepigebesdneue A. James Wagner

ATES) a Jack R. Plimmer

EEE MINED ICA occ sac coc ec eee ae dele dep ce aeece ewan s ene ealduyems Richard K. Cook

et oS 5 sw og ak 8015 Genie bw eied Wiel BINS & win diel ese loca 0,8 'n'e oO Dick Duffey

SSG STS ee A. D. Berneking

ES eo nee Gia occu esa din) a wows aa ou, Baad a re Edwin R. Fuller, Jr.

eo 8002 65S cv ajerei's a's A A iw nis ane vine ee ih wg ean em EwlO ee Alayne A. Adams

PICO EICTIC® CIID: 5 5. cece i ccd eee eae ss enue cm sees wae aunie® Deborah J. Warner

ERE CPE ae 0s dis 0 fel gv Dea dkdr sonsd, sve ee wa em eee oo ase eal George J. Simonis

eet er Plant Physiologists ............cccccssecccecncdccavcccene Walter Shropshire, Jr.

i aici ReScArcH COUCH... .....25 beer ev ences cewensessasimuenvascde John G. Honig

RCT 95.08 oc Goo ass Sos: 0) sna 9 od ee wads nom Dea ee ain Ca aeeed we Se Jewel B. Barlow

American Institute of Mining, Metallurgical

RMR SRROIOCES 2 ois as a dina coy have ove Coup asp Ram é wsiae he oh hate A aera Garrett R. Hyde

REVERIE (a) Oe ys ig. brah a's vb o> @ nil ugh Mew ola eats we cietv gate «eialal Benson J. Simon

REARS OC CETOA 5.135 dom cin oo 98 acid wat huey, 0 msdn RS Noel e gine a. tp RON Ere Patrick Hayes

OE RISE) hos pts dis bx sve vie wee AEG Oiptan Mind WELNCao wae Miloslav Racheigl, Jr.

_eencton Paint Technical Group ..........ceccsccvccnevescccctccccesnceees Paul G. Campbell

een aytOpathological Society .......00ccceccercvcccccnccceccesavenece Howard E. Waterworth

eee teeneral Systems Researchi .........2.ccsccscceseccvenensasecees Ronald W. Manderscheid

EE I ee ere re rer rrr rrr rte eee rer re Stanley Deutsch

I IPIESIES ACIOUY on cat aie es dimen eeidig a cena dienee eee secesvesuesauedeeeuns nis Irwin M. Alperin

Association for Science, Technology and Innovation .........2e cece cece cence ene neneeees Ralph I. Cole

EINES ORCTOEW Ss diciu le clare lave wie inh + kuca's aipcln.ele ebb ee oases lv aie Ronald W. Manderscheid

Delegates continue in office until new selections are made by the representative societies.

Washington Academy of Sciences

1101 N. Highland St.

Arlington, Va. 22201

Return Requested with Form 3579

PT a i a

iy .

VOLUME 73

Number 4

rnal of the December, 1983

- a“

-wASHINGTON

ACADEMY .,. SCIENCES

ISSN 0043-0439

Issued Quarterly

at Washington, D.C.

CONTENTS

Articles:

PHILIP SZE: Macroalgal Communities in Tidepools on the New England Coast. .

BRIAN E. LACY: Neutral Crest Cell Migration and the Extracellular Matrix.....

MARK ANGELO JOHNSTONE: Bayesian Estimation of Reliability in the Stress-

Strength Context

Sielelh¢ se aes Cee bla s e<8 oe 6 0 6-4 alesis eS a ip) a6 Moa pie eer ea ee: Ce Cm were ee eee

SHERMAN ROSS, ROBERT M. SASMOR, and JOHN P. WHALEN: Research

Fellows of the Washington Consortium of Universities at the U.S. Army Research

Institute

Washington Academy of Sciences

Founded in 1898

EXECUTIVE COMMITTEE

President

Jean K. Boek

President-Elect

Ralph I. Cole

Secretary

Elise A. Brown

Treasurer

Ronald Manderscheid

Managers-at-Large

Jo-Anne Jackson

| Grover C. Sherlin, Office Manager

Bi Alfred Weissler

Hi Howard E. Sorrows

James Wagner

Simon W. Strauss

BOARD OF MANAGERS

All delegates of affiliated

Societies (see Cover 3)

EDITORS

Irving Gray

Joseph Neale

ACADEMY OFFICE

1101 N. Highland St.

Arlington, Va. 22201

Telephone: (703) 276-9254

The Journal

This journal, the official organ of the Washington

Academy of Sciences, publishes historical articles, critical

reviews, and scholarly scientific articles; proceedings of

meetings of the Academy and its Board of Managers; and

other items of interest to Academy members. The Journal

appears four times a year (March, June, September, and

December)—the September issue contains a directory of

the Academy membership.

Subscription Rates

Members, fellows, and patrons in good standing receive

the Journal without charge. Subscriptions are available on

a calendar year basis only, payable in advance. Payment

must be made in U.S. currency at the following rates:

U.S. and Canada... .. $19.00

FPOTIOO 2c Sone Sates 22.00

Single Copy Price ....: 7.50

UHURU Symposium—2 issues............ $15.00

Delivery by Alr .5...4 22; 2.00

Single-copy price for Vol. 66, No. 1 (March, 1976) is $7.50.

Back Issues

Obtainable from the Academy office (address at bottom

of opposite column): Proceedings: Vols. 1-13 (1898-1910)

Index: To Vols. 1-13 of the Proceedings and Vols. 1-40 0f

the Journal Journal: Back issues, volumes, and sets (Vols.

1-62, 1911-1972) and all current issues.

Claims for Missing Numbers

Claims will not be allowed if received more than 60 days —

after date of mailing plus time normally required for —

postal delivery and claim. No claims will be allowed ~

because of failure to notify the Academy of a change in —

address.

Change of Address

Address changes should be sent promptly to the Academy

office. Such notification should show both old and new

addresses and zip number.

Published quarterly in March, June, September, and December of each year by the Washington

Academy of Sciences, 1101 N. Highland St., Arlington, Va. 22201. Second class postage paid at

Arlington, Va. and additional mailing offices.

Journal of the Washington Academy of Sciences,

Volume 73, Number 4, Pages 121-127, December 1983.

Macroalgal Communities in Tidepools

on the New England Coast

Philip Sze

Department of Biology, Georgetown University, Washington, D.C. 20057

ABSTRACT

During the summer of 1981, macroalgal communities in 65 pools were surveyed on Apple-

dore Island in the Gulf of Maine. Abundance of algae was quantitatively described by estimat-

ing coverage, and patterns of species number and diversity were related to wave exposure and

elevation on shore. Pools in wave exposed areas were distinct in composition from pools in

protected areas, but numbers of species and diversity did not differ significantly between the

two types. Semi-exposed pools were intermediate in composition and tended to have a greater

number of species. In pools with high densities of Littorina littorea, the number of algal species

was consistently low. Algal diversity showed an inverse relationship with coverage by Chondrus

crispus, when Chondrus coverage exceeded 20%. Chondrus coverage may provide an estimate

of the successional stage of a pool.

Introduction

Tidepools are a common feature of rocky

shores in the North Atlantic Ocean. Pools

often support rich communities of macro-

phytic algae and associated animal popula-

tions. Most studies of tidepools have been

descriptions of species present, sometimes

relating their distribution to physical con-

ditions."* More recent studies”’ suggest

that biologic interactions also may influ-

ence distributions of algal species in pools.

The objective of the present study was to

describe quantitatively communities of

macroalgae in tidepools at the Isles of Shoals

in northern New England. The study took

advantage of the large number of pools on

one relatively small island. Ina companion

paper,’ I have described the distributions

of individual species in the pools.

121

Materials and Methods

Study site—The Isles of Shoals are located

approximately 10 km off the coast of New

Hampshire in the Gulf of Maine (42°59’N,

70°37’ W). Pools studied were on Appledore

Island (Maine), with an area of 33.6 ha.

Tides at the Shoals are semidiurnal with an

average range of 2.6 m.'' The seawater sur-

rounding the Shoals normally has a salinity

of 28-33°/o0, and summer water tempera-

tures rarely exceed 20°C (higher tempera-

tures may occur in pools). The seawater is

relatively free of pollutants but enriched by

wastes from the bird colonies on the island.

Previous studies at the Shoals have de-

scribed algae in the intertidal region, '* sub-

tidal region,'* '’ high tidepools’ and su-

pralittoral pools.'° Femino and Mathieson’

and Mathieson et al.'’ described seasonal

122 PHILIP SZE

changes in algal composition in tidepools

on the adjacent mainland.

Field methods—Pools were surveyed

during two periods in 1981: 16 June—4 July

and 27 July-10 August (referred to as June

and August periods respectively.

Each pool was characterized in terms of

its exposure to waves and elevation on the

shore. On the basis of exposure, pools were

divided into three categories. Protected

pools were on the west side of the island

facing the mainland. Semi-exposed and ex-

posed pools were on the east side of the is-

land. Exposed pools received directly the

impact of incoming waves, while semi-ex-

posed pools were protected by some ob-

struction. The bottoms of exposed pools

were usually covered by mussels (Mytilus

edulis), and Littorina littorea was absent. In

exposed areas, there was a well-developed

barnacle zone on emergent surfaces, and

little or no Ascophyllum nodosum was

present.

Elevation of pools on the shore was de-

termined relative to the zones of dominant

organisms on emergent surfaces. In pro-

tected and semi-exposed areas, the interti-

dal region was divided into fucoid and red

algal zones. In exposed areas, a barnacle

zone was also present. Approximate posi-

tions of the boundaries between zones on

different sides of the island were deter-

mined relative to reference points derived

from USGS bench marks (Table 1). Pools

very high in the intertidal region (described

by Sze’) and surrounded by bare rocks

were not sampled. All pools studied were

flushed during each tidal cycle.

Only macroalgae greater than 0.5 cm in

length were included in surveys of pool

communities. Diatom growths in excess of

0.5 cm were occasionally encountered in

semi-exposed pools but were not included

in surveys.

Abundance of macroalgae in pools was

determined with a square plexiglas plate

divided into 40.25 X 0.25 m’ quadrats (re-

ferred to as a set of quadrats). Each quad-

rat had 15 randomly placed dots (positions

derived from random numbers), and pres-

ence of algal species under dots was re-

Table 1.—Elevation of Zone Boundaries on Emer-

gent Surfaces in Meters Above Mean Low Water.

Mean Stan. Dev.

West side

top of fucoid zone 2.2 0.2 (n = 7)

topofredalgalzone 0.7 0.3

East side (exposed)

topofbarnaclezone 2.8 0.3 (n = 7)

top of fucoid zone LZ 0.4

topofredalgalzone 0.8 0.5

corded. Coverage for a species in a pool

was determined from the number of dots

with the species divided by the dots for all

algal species. For rarer and understory

species, presence in quadrats was recorded.

Also determined in quadrats were the

numbers of the herbivorous gastropod Lit-

torina littorea. Quadrat measurements were

facilitated by placing wire frames of the

same size directly under quadrats and then

actively searching through the enclosed area.

For most pools, as much of the surface

area was covered by successive placement

of the plexiglas plate. In larger pools where

more than 8 sets of quadrats could fit, sur-

veys were done along the axes of each pool.

In all pools, the plate was always placed so

that all 4 quadrats were entirely within the

poo! without overlapping areas previously

examined. Pools too small to contain one

set of quadrats were not surveyed.

In addition, pools with the following fea-

tures were not sampled: (a) extensive areas

with water depths greater than 0.5 m, (b)

obviously unhealthy algae, growths of fungi

or extensive detritus, (c) loose rocks likely

to scrape macroalgae from rock surfaces.

Data analysis—For each pool, the com-

munity present was described in terms of

number of macroalgal species and diversity

of macroalgae. Diversity was calculated by

the Shannon-Wiener index using natural

logarithms:

H’ = —.), pin.

where pi = coverage of species 1.

Means for species number and diversity

in different pool types were compared

pte

Pineiro

TIDEPOOL COMMUNITIES 123

Table 2.—Number of Each Pool Type Sampled During the Summer of 1981.

Exposed

June Group (16 June-4 July)

barnacle zone 5

fucoid zone 4

red algal zone 2

subtotal 11

August Group (27 July-10 August)

barnacle zone 4

fucoid zone 4

red algal zone 4

subtotal 12

total 23

using Wilcoxon’s Two Sample Rank-Sum

Tests (Mann-Whitney Tests). Tests were

two-tailed with a P= 0.05 significance

level.

Species coverage data were used to

compare pools by the method of polar

ordination.'*

Results

The 65 pools surveyed during the summer

of 1981 were divided into June and August

groups and then subdivided according to

elevation and exposure as shown in Table

2. A list of macroalgal species found in

pools (with taxonomic authorities) and

mean densities of Littorina littorea in dif-

Semi-exposed Protected Total

5 1] 20

4 5 1]

9 16 36

4 7 15

2 4 10

6 1] 29

15 27 65

ferent pool types are given in Sze.'” Gener-

ally, L. littorea was more abundant in pro-

tected pools than semi-exposed pools and

was absent from exposed pools.

The number of species per pool ranged

from 7-22. Semi-exposed pools generally

had more species, and there was no signifi-

cant difference between exposed and pro-

tected pools (Table 3). Species number did

not show a relationship with density of L.

littorea, except that low algal numbers were

consistently found in pools with densities

of L. littorea exceeding 160 animals m~

(Figure 1).

There was no discernible pattern to algal

diversity in the different pools types (Table

4), and diversity was not related to L. littorea

density (data not presented). The impor-

Table 3.—Mean Number of Algal Species in Pool Types (Standard Deviations in Parentheses).

Exposed Semi-exposed Protected

June Group

barnacle zone 9.40 (2.51)

fucoid zone 11.50 (3.79) 17.20 (1.92) 10.45 (2.21)

red zone 10.50 (0.71) 14.50 (3.70) 14.80 (5.97)

August Group

barnacle zone 11.50 (2.89)

fucoid zone 9.75 (1.26) 14.75 (4.27) 13.00 (3.46)

red zone 11.00 (1.83) 13.00 (1.41) 13.75 (2.06)

Significant differences by rank-sum tests:

sex-fuc-J versus ex-bar-J

ex-fuc-J

pr-fuc-J

ex-bar-A

ex-fuc-A

ex-red-A

sex-red J versus pr-fuc-J

pr-fuc-J versus pr-fuc-A

pr-fuc-J versus pr-red-A

pr = protected, sex = semi-exposed, ex = exposed

bar = barnacle zone, fuc = fucoid zone, red = red algal zone

J = June Group, A = August Group

De _____ ee

124 PHILIP SZE

Number of Species

NS o

0 80 160 240

Littorina Density

320 400 480 560

Fig. 1. Number of algal species in relationship to Littorina littorea density (animals m -)in protected O and

semi-exposed pools V.

tance of coverage by Chondrus crispus was

shown by an inverse relationship with di-

versity, when Chondrus coverage exceeded

0.20 (Figure 2).

Ordination of pool data attempted to

show relationships among pools based on

species composition and importance. In

Figure 3, the abscissas are based on two

protected pools and the ordinates on two

exposed pools. Protected pools were dis-

tinct from exposed pools, particularly in

the June groups. The protected pools fell

along a horizontal line with red algal zone

pools tending to occur to the left and fucoid

zone pools to the right. The exposed pools

generally separated from each other along

Table 4.—Mean Algal Diversity in Pool Types (Standard Deviation in Parentheses).

Exposed Semi-exposed Protected

June Group

barnacle zone 1.32 (0.35)

fucoid zone 1.15 (0.32) 1.51 (0.47) 1.58 (0.17)

red zone 1.55 (0.15) 1.64 (0.08) 1.21 (0.39)

August Group 1.33 (0.16)

fucoid zone 1.29 (0.24) 1.08 (0.73) 1.19 (0.30)

red zone 1.44 (0.20) 1.67 (0.13) 1.30 (0.21)

Significant differences by rank-sum tests (abbreviations as in Table 3):

sex-red-J versus ex-fuc-J

ex-bar-A _ pr-fuc-J versus ex-fuc-J

ex-fuc-A

pr-fuc-A

ex-bar-A

pr-fuc-A

TIDEPOOL COMMUNITIES 125

Diversity

Chondrus Coverage

Fig. 2. Algal diversity in relationship to coverage by Chondrus crispus in protected O, semi-exposed Y and

exposed pools A. Dots in symbols indicate pools with greater than 0.20 coverage by Enteromorpha intestinalis, E.

linza and Ulva lactuca; filled symbols indicate pools with greater than a 0.50 coverage by Laminaria saccharina,

L. digitata and Alaria esculenta. A possible non-linear relationship at Chondrus coverages exceeding 0.20 is indi-

cated (line visually drawn).

a vertical line. Points for barnacle zone

pools tended to be higher than red algal

and fucoid zone pools, which intermixed.

Scatter in both horizontal and vertical di-

rections showed the intermediate condition

of semi-exposed pools. Exposed pools in

August (Figure 3B) showed a poorer sepa-

ration by elevation and more horizontal

scatter than in June (Figure 3A).

Discussion

The results of ordination (Figure 3) indi-

cate that exposed pools and protected

pools have relatively distinct communities.

Neither species diversity nor species number

(Tables 3, 4) was significantly lower in ex-

posed pools. Thus increased wave stress

did not appear to simplify community

Structure. Species in exposed pools tended

to be opportunistic species, such as Chor-

daria flagelliformis and Entermorpha linza,

or persistent individuals of species more

widely distributed in winter and spring,

such as Scytosiphon lomentaria, Petalonia

fascia and Spongomorpha spinescens.'° An

intermediate condition occurred in semi-

exposed pools, possibly as a result of in-

termixing of species from the other two

types.

Better separation of pools by elevation in

the June group than August group (Figure

3) suggests that pools ona given area of the

shore tended to become more uniform as

the summer progressed. This reduced vari-

ation may have resulted from an increase of

green algae, such as Ulva lactuca, and fu-

coids in summer and reduction in kelps

(Laminaria saccharina, Alaria esculenta),

with more restricted vertical distributions.

Also under milder ocean conditions in

summer, exposed and semi-exposed pools

may have received waves of similar force.

Species, such as Ulva lactuca and Chaeto-

126 PHILIP SZE

100

XXX —» XXVIII (exposed)

0 20 40 60 80 100

XXII —> XXX

Fig. 3. Ordination of pools based on dissimilarity

coefficients calculated from species coverage. X-axis

is based on two-protected pools, a red algal zone pool

(XXIII) and a high fucoid zone pool (XXX) with dis-

similarity of 0.98. Y-axis is based on two exposed

pools, a red algal zone pool (XXXII) and a barnacle

morpha melagonium, seemed to spread from

semi-exposed to exposed pools as the sum-

mer progressed. '

Lubchenco’ predicted that both algal di-

versity and species numbers would be great-

est at intermediate densities of Littorina lit-

torea (approximately 150 animals m_’).

These patterns were not found in the present

study (Figure 1). At densities exceeding 160

animals m ”, only species resistant to L. lit-

torea were able to thrive.'” The greater spe-

cies numbers were observed in some pools

with low L. littorea densities.

Chondrus crispus is a major species on

emergent surfaces in the lower intertidal

region of New England. Here Chondrus

successfully occupies emergent surfaces to

the exclusion of other algal species, because

of its resistance to grazing and ability to

outcompete other perennial algae.'” ~° If

Chondrus can also monopolize space in

pools, its coverage should progressively

increase and may provide a measure of

the successional development of pool com-

munities. Chondrus was found in all except

2 exposed pools. Except at low Chondrus

oO 20 40 60 80 100

(protected )

zone pools (X XVII) with 0.99 dissimilarity. June (A)

and August (B) groups shown separately. Scales for

axes are dissimilarity X100. Protected pools O, ex-

posed pools A, semi-exposed pools V. Open symbols

are fucoid zone pools, filled symbols red algal zone

pools, barnacle zone pools with dots.

coverage (<0.20), algal diversity was in-

versely related to its abundance. Low

Chondrus coverage (<0.20) and low diver-

sities were associated with Ulva lactuca, En-

teromorpha intestinalis and E. linza or a

canopy of kelps (Figure 2). Abundance of

these green algae, which are colonizing

species, may indicate recent disturbance

that has cleared part of a pool bottom. In

the presence of kelps, the abundance of

Chondrus may have been underestimated

because of its presence in the understory.

Acknowledgments

I wish to thank the Shoals Marine Lab-

oratory for allowing me to use its facilities

and the National Geographic Society for

financial support.

References

1. Femino, R. J. and A. C. Mathieson, 1980. Investi-

gations of New England marine algae. IV. The

ecology and seasonal succession of tide pool algae

10.

at Bald Head Cliff, York, Maine, USA. Bot. Mar.,

23: 319-332.

. Gustavsson, U. 1972. A proposal for a classifica-

tion of marine rockpools on the Swedish west

coast. Bot. Mar., 15: 210-214.

. Johnson, D. S. and A. F. Skutch. 1928. Littoral

vegetation on a headland of Mt. Desert Island,

Maine. II. Tide-pools and the environment and

classification of submersible plant communities.

Ecology, 9: 307-338.

. Lewis, J. R. 1964. The Ecology of Rocky Shores.

English Universities Press, London.

. Dethier, M. N. 1982. Pattern and process in tide-

pool algae: factors influencing seasonality and

distribution. Bot. Mar., 25: 55-66.

. Goss-Custard, S., J. Jones, J. A. Kitching and T. A.

Norton. 1979. Tide pools at Carrigathorna and

Barloge Creek. Phil. Trans. Roy. Soc., London,

B287: 1-44.

. Lubchenco, J. 1978. Plant species diversity in a

marine intertidal community: importance of her-

bivore food preference and algal competitive abili-

ties. Am. Nat., 112: 23-39.

. Lubchenco, J. 1982. Effects of grazers and algal

competitors on fucoid colonization in tide pools.

J. Phycol., 18: 544-550.

. Sze, P. 1980. Aspects of the ecology of macro-

phytic algae in high rockpools at the Isles of Shoals

(USA). Bot. Mar., 23: 313-318.

Sze, P. 1982. Distributions of macroalgae in tide-

pools on the New England coast (USA). Bot.

Mar., 25: 269-276.

LS.

16.

127

Anonymous. 1980. Tide Tables 1981—High and

Low Water Predictions. East Coast of North and

South America, National Ocean Survey, U.S.

Government Printing Office, Washington.

. Kingsbury, J. M. 1976. Transect Study of the Inter-

tidal Biota of Star Island, Isles of Shoals. Shoals

Marine Laboratory, Ithaca.

. Boden, G. T. 1979. The effect of depth on summer

growth of Laminaria saccharina (Phaeophyta,

Laminariales). Phycologia, 18: 405-408.

. Mathieson, A. C. 1979. Vertical distribution and

longevity of subtidal seaweeds in northern New

England, U.S.A. Bot. Mar., 22: 511-520.

Norall, T. L., A. C. Mathieson and J. A. Kilar.

1981. Reproductive ecology of four subtidal red

algae. J. Exp. Mar. Biol. Ecol., 54: 119-136.

Sze, P. 1981. Observations on microalgae in supra-

littoral rockpools in New England (USA). Bot.

Mar., 24: 337-341.

. Mathieson, A. C., E. J. Hehre and N. B. Reynolds.

1981. Investigations of New England marine

algae I: A floristic and descriptive ecological

study of the marine algae at Jaffrey Point, New

Hampshire, U.S.A. Bot. Mar., 24: 521-532.

. Poole, R. W. 1974. An Introduction to Quantitative

Ecology. McGraw-Hill, New York.

. Lubchenco, J. and B. A. Menge. 1978. Community

development and persistence in a low rocky inter-

tidal zone. Ecol. Monogr., 48: 67-94.

. Lubchenco, J. 1980: Algal zonation in the New

England rocky intertidal community: an experi-

mental analysis. Ecology, 61: 333-344.

Journal of the Washington Academy of Sciences,

Volume 73, Number 4, Pages 127-139, December 1983.

Neural Crest Cell Migration

and the Extracellular Matrix

Brian E. Lacy

Department of Anatomy, Georgetown University School of Medicine,

Washington, D.C. 20007

ABSTRACT

The neural crest cell is a neuroectodermal derivative which first appears during late neuru-

lation in the developing embryo. After condensing in the midline, the crest cells migrate exten-

sively throughout the organism giving rise to a wide variety of neuronal and nonneuronal

structures. Although the migratory pathways are well-defined, the factors responsible for

crest cell migration and cytodifferentiation are still largely unknown. It was previously

thought that crest cell migration was a rigid, predetermined event. However, more recent

experiments have shown that the neural crest cells are pluripotential cells which are actively

translocated through the extracellular matrix. Various components of the extracellular matrix

(ECM)—some of which include fibronectin, collagen, and glycosaminoglycans—have been

implicated in regulating the processes of cell-cell adhesion, cell migration, and cell differentia-

tion. These components of the ECM have been found to vary substantially in type, and in

concentration, both spatially and temporally, along the crest cell migratory pathways. Thus,

neural crest cell migration no longer appears to be a simple predetermined process, but rather

a complex interaction between the constantly changing properties of both the environment

and the neural crest cell.

1. Introduction what initially appears to be a population of

homogeneous cells, arise such diverse struc-

tures as melanocytes, odontoblasts, sen-

sory ganglia and sympathetic ganglia’

(see Table 1). This paper will review the

One of the most important processes to

occur in the developing nervous system is

that of cell migration. Both neuronal and

non-neuronal cells migrate extensively

throughout the central nervous system and

the peripheral nervous system during de-

velopment. In an attempt to identify fac-

tors responsible for the migratory behavior

of these cells, researchers have focused ona

number of neural systems which offer the

advantages of being both anatomically dis-

crete and readily accessible to in situ experi-

mentation. One such system which has

been the subject of active investigation in

many laboratories is that of the neural

crest. The neural crest is an interesting

model for the study of both cellular migra-

tion and cytodifferentiation since, from

128

local environments which surround the

neural crest cells and how the components

of these environments might influence both

the migratory behavior of the crest cells

and their eventual differentiation.

2. Development of the Neural Crest

The neural crest is a transient embryonic

structure which arises during late neurula-

tion. In the chick embryo at stage 6 (staged

according to Hamburger and Hamilton‘)

the neural plate is a thickened sheet of ec-

todermal epithelium which overlies the an-

NEURAL CREST CELL MIGRATION AND THE EXTRACELLULAR MATRIX 129

Table 1—Derivatives of the Neural Crest.’ ‘> '*

1) Neuronal Cells

a. sensory neuron contributions to trigeminal (V), geniculate (VII), superior (IX), and jugular

ganglion (X). (other neuronal contributions are via epithelial placodes of the head region).

b. Dorsal Root Ganglia (DRG)

c. Autonomic Ganglia

2) Support Cells

a. Schwann sheath cells of the Peripheral Nervous System (PNS)

b. Glia and satellite cells of dorsal root ganglia and autonomic ganglia

3) Pigment cells of the dermis, epidermis, and internal organs

4) Endocrine cells

a. Adrenal medullary cells

b. Calcitonin producing cells of the thyroid (C-cells)

c. Cells of the carotid body

5) Skeletal and connective tissue

. Bones and cartilage of the head and face

. Dermis of face, jaw, and upper neck

Pia and arachnoid

Corneal stromal fibroblasts

. Ciliary muscles (striated)

= 09 -», o ie) oO p

teriorly located prochordal plate, and the

notochord and somites which are respec-

tively situated in the middle and posterior

regions of the developing embryo. By stage

8 the deep columnar epithelium of the neu-

ral plate has thickened laterally over the

mesoderm and thinned out medially in the

area over the notochord. This differential

growth of the neural ectoderm allows the

lateral margins of the neural plate to rise up

and form the neural folds. Fusion of the

neural folds to form the neural tube begins

late during stage 8 in the region of the fu-

ture mesencephalon and continues both

cranially and caudally such that by the end

of stage 12, neural tube closure is com-

plete.° Neural crest cells may first be identi-

fied shortly after neural fold apposition as

a distinct longitudinal band of cells situ-

ated dorsomedially to the neural tube.°

During this early stage of crest cell appear-

ance, a dorsal-to-ventral layering of cells

occurs, such that the dorsal-most cells lose

contact with the neural tube while the ven-

tral-most cells maintain contact with the

neural tube. The neural crest cell popula-

tion begins in the posterior prosencephalon

and extends as a continuous band through

. Connective tissue components of the thymus

Mesenchyme of thyroid, parathyroid, and salivary glands

. Connective tissue components of large arteries derived from aortic arches

. Some connective tissue components of striated muscle in the facial region

the myelencephalon into the trunk region

where the crest cell population slowly de-

clines in number. Few, if any, crest cells are

formed in the rostral prosencephalon.°

Crest cells do not all emerge from the

neural tube at the same time. Rather, there

is a temporal pattern of emergence which

originates anteriorly and in a wave-like

manner proceeds posteriorly. In an at-

tempt to simplify, and to unify, descrip-

tions of this complex process, Tosney® has

described four phases of crest cell morpho-

genesis: appearance, condensation, early

migration, and advanced migration. In a

single embryo all four phases may be stud-

ied at different axial levels; thus, while crest

cells are in an advanced stage of migration

cranially, crest cells in the trunk region are

just beginning to appear and to condense.

In addition, the phases of crest cell mor-

phogenesis may be studied at the same

axial level solely by comparing identical re-

gions of the embryo at different develop-

mental stages.

As the crest cells emerge from the dorso-

lateral aspect of the neural tube, the basal

lamina becomes discontinuous. However,

the basal lamina of the lateral neural tube is

130 BRIAN E. LACY

continuous throughout all phases of crest

migration as is the basal lamina of the

superficial ectoderm.° Later, when the crest

cell population has been depleted in the

midline, the basal lamina on the superior

aspect of the neural tube reforms.

3. Neural Crest Cell Migration

After their emergence and subsequent

condensation, neural crest cells begin their

migration along precise, predictable, region-

specific pathways.’ ° Crest cell emigration

from a specific axial level extends over a pe-

riod of approximately twenty-four hours in

the avian embryo.’ Unfortunately, migra-

tory crest cells appear almost identical in

appearance to fibroblastic mesenchymal

cells,° and therefore, they must be marked

in some way in order to be identified as they

migrate and as they later condense to form

a varied array of derivatives. Initially, vital

staining’ was used to visualize migratory

crest cells, although this technique was not

very precise at the level of individual cells.

Later, cells were identified via the use of

autoradiography following the incorpora-

tion of tritiated thymidine.’ * More recently,

specific migratory cell populations have

been tracked with xenoplastic grafts of

quail tissue transplanted into chick em-

bryos. The quail cells can be identified by

the fact that their nuclei are heavily hetero-

chromatic during interphase while chick

nucleolar chromatin is more uniformly dis-

tributed and generally very euchromatic.’

The advantage of this technique is that a

‘natural’ marker can be introduced into a

host embryo which, unlike tritiated thymi-

dine, will not fade in intensity with subse-

quent cell divisions.

In the trunk region of the developing

avian embryo, neural crest cells migrate

along two well-defined pathways. In the

first pathway, a dorsolateral stream of cells

migrates just beneath the surface of the

ectoderm, and gives rise to presumptive

melanoblasts.” '° Crest cells in this dorso-

lateral pathway appear to migrate as a

loosely interwoven sheet of cells, rather

than as single cells.° In the second pathway,

crest cells pass ventrally in between the

neural tube and the somites to give rise to

adrenomedullary cells, dorsal root ganglia,

the sympathetic trunk, Schwann cells, the

enteric nervous system, and aortic and

renal plexuses.” '* '' In both cases, the crest

cells migrate by passing through unclut-

tered, cell-free spaces. However, the rate of

migration of crest cells through these spaces

is not equivalent. Rather, the ventral path-

way is occupied much more rapidly than is

the dorsolateral pathway. ”

In the cephalic region of the developing

chick, the migratory pathways of neural

crest cells are just as precisely defined as

those of the trunk, yet the routes are far

more circuitous, due to the cells having to

migrate around such obstacles as the de-

veloping eyes, olfactory placodes and phar-

ynx. The cranial neural crest cells are re-

sponsible for the development of such

diverse structures in the head region asthe

trigeminal sensory ganglion (V), the genic-

ulate ganglion (VII), the root ganglia of IX

and X, the ciliary ganglia, the pia and

arachnoid, and the skeletal and connective

tissues of the upper face and visceral arch

regions.” °

4. Is Crest Cell Migration Predetermined?

Before investigating possible exogenous

factors which might be responsible for the

differential migration of neural crest cells,

one must first firmly establish that the mi-

gratory abilities of cells are not based solely

on its genotype, i.e., that the cells are not

already completely predetermined, with re-

spect to either the precise pathway of mi-

gration or terminal cytodifferentiation, prior

to the onset of migration. One of the first

experiments to test whether neural crest

cells are completely predetermined prior to

migration was performed by Weston and

Butler in 1966.’ In one experiment, neu-

ral crest cells from anterior axial levels where

crest cells had been migrating for some

time, were excised, labeled with tritiated

thymidine, and implanted at more poste-

NEURAL CREST CELL MIGRATION AND THE EXTRACELLULAR MATRIX 131

rior levels where the neural crest had just

condensed (the host neural crest was ex-

cised before the tritiated thymidine labeled

cells were implanted). These ‘‘old”’ cells

which were transplanted into a “‘young”’

environment were able to migrate exten-

sively and contribute to all structures nor-

mally derived from the neural crest at that

level. These results suggested that either

crest cells are initially pluripotent and are

guided by the local environment as they

migrate and later differentiate at their ter-

minal locations, or crest cells are com-

pletely predetermined prior to their migra-

tion, but because they begin their emigration

at random, a population of crest cells is

maintained at all times which can later mi-

grate and differentiate into any crest-de-

rived structure. To distinguish between

these two alternatives, a second experiment

was performed, where newly condensed

crest tissue was excised, labeled with tri-

tiated thymidine and implanted into pro-

gressively older hosts where crest cell mi-

gration had already begun. In this case, a

wide array of neural crest-derived struc-

tures was found to be lacking, showing that

either migration or terminal differentiation

had failed to occur. Although far from

conclusive, and somewhat incomplete, this

experiment is notable principally because it

was one of the first experiments to try to

distinguish between the possible roles of

the changing environment and the inherent

programming of the genome in crest cell

development.

The concept of migratory pathways ac-

tively altering the potential of crest cells

during development persisted, albeit with

little substantive proof, until the mid-1970’s.

In 1974, Le Douarin and Teillet’’ showed

that most crest cells develop pluripoten-

tially and that these potentialities can be al-

tered, both spatially and biochemically, by

the local environment. In the avian em-

bryo, crest cells corresponding to the level

of somites 18 through 24 normally develop

into adrenomedullary cells. These cells syn-

thesize and release catecholamines and can

be identified by incubating the cells in for-

maldehyde vapor. The formaldehyde reacts

with the catecholamines present in the cells

to produce a fluorescent compound which

can be readily detected. Neural crest cells

from these levels normally do not colonize

the gut. “‘Vagal’’ neural crest cells, corre-

sponding to the level of somites | through

7, and lumbosacral crest cells, located pos-

terior to somite level 28, are normally re-

sponsible for producing all the enteric gan-

glia in the developing gut. When Le Douarin

and Teillet transplanted quail ““adrenomed-

ullary”’ crest cells into the “‘vagal’’ region of

chick, the avian gut became colonized by

quail crest cells which were non-adrener-

gic, i.e., they had lost the ability to synthe-

size catecholamines as determined by for-

maldehyde induced fluorescence (FIF). In

the second experiment, quail ‘“‘vagal”’ crest

cells were grafted into chick ‘‘adrenomed-

ullary”’ crest somitic levels. In this case

quail cells were able to migrate to the su-

prarenal glands, and populate the glands

with cells which were able to produce cate-

cholamines as shown via FIF. Thus, the

pattern of migration of crest cells is deter-

mined not by the source of grafted donor

cells, but rather by the axial level of the

graft site in the host.

Although the migratory pathways of

cranial and trunk neural crest cells differ

dramatically, as do the terminal structures

produced, Noden’® showed that trunk crest

cells, when heterotopically transplanted

into the metencephalon, were able to prop-

erly migrate and condense to form neuro-

blasts of the trigeminal ganglia. Thus, the

trunk crest cells were able to migrate in an

environment which they normally never

encounter and to interact productively with

placodal cells which they normally never

do. Bronner and Cohen’! and Le Lievre ef

al.'° took these experiments one step further

by transplanting fully differentiated crest

cells (either melanocytes or dorsal root

ganglion cells) into different axial levels of

the avian embryo. Pigment cells, which are

not normally motile in vitro, migrated ex-

tensively along the ventral pathway when

injected into the lumen of the posterior

somites. The cells were found to be local-

ized in areas normally populated by crest

132 BRIAN E. LACY

cell derivatives.'’ Dorsal root ganglion

(DRG) cells were found to migrate exten-

sively after transplantation and contrib-

uted to the formation of the host’s DRG,

sympathetic trunk and adrenal medulla.”®

If, as many researchers still believed, crest

cells were irreversibly predetermined prior

to their emigration from the neural tube,

then none of the above mentioned events

would have been able to occur.

5. Specificity of Migratory Pathways

It thus appears that there are preferential

migratory pathways specific for each axial

level in the chick which lead the pluripoten-

tial crest cells to their terminal localization.

Furthermore, the local environment, either

along the migratory pathway, or at the

termination of the pathway, may be re-

sponsible for controlling the biochemical

and morphogenetic differentiation of neu-

ral crest cells. And yet, how specific are

these pathways? We have seen that fully

differentiated neural crest derivatives such

as melanocytes’! and DRGcells’” retain an

ability to remigrate along region-specific

pathways in the avian embryo. Yet, is this

migration a specific process where the crest

cell or the crest-derivative interacts with

the components of the migratory pathway?

Or is the process a nonspecific one, where

any cell type with access to the dorsolateral

or ventral stream of migrating crest cells

will be passively, translocated to varying

sites throughout the embryo? Using three

different cell types Bronner-Fraser and

Cohen"’ were able to show that the ventral

migratory pathway is selective for post-

migratory crest cell derivatives when com-

pared to motile cells not of neural crest

origin. When either somitic cells or fibro-

blastic cells, both of which are normally

motile in vivo and in vitro, were injected

into the lumen of somites in the avian em-

bryo they failed to migrate and remained

aggregated within the somitic cavity. If

either somitic cells or fibroblastic cells were

combined with melanocytes and then in-

jected, only the melanocytes would migrate

along the ventral pathway, and the somitic

cells or fibroblastic cells would again remain

associated with the somitic mesenchyme of

the host. An apparent inconsistency in

these results is that of the melanocytes mi-

grating along the ventral pathway, a course

which they never choose during normal de-

velopment. This most likely occurred due

to the ventrally-skewed orientation of the

injection site. This indicates then, that the

ventral migratory pathway appears to be

selective for crest cell derivatives, although

there may be no difference in the ability of

crest-derived cells to migrate along either

the dorsolateral or the ventral pathways. In

addition, when melanocytes were injected

into different somitic levels, it was found

that injections at the most anterior levels

yielded a significantly smaller number of

crest cell derivatives than did injections at

the more posterior levels, indicating that

the normal range of crest cell migration

had been curtailed. Thus, the extent of neu-

ral crest cell migration, or crest cell trans-

location, is limited by the developmental!

age of the embryo.

The results of Bronner-Fraser and Cohen

were confirmed by Erickson et al.’° with

one interesting addition. A transformed fi-

broblastic cell line, Sarcoma 180, when

grafted into different axial levels, was found

to migrate extensively throughout the em-

bryo along both the dorsolateral and ven-

tral migratory pathways. Somewhere in the

transformation of a normal fibroblast toa

tumorogenic fibroblast, properties were

either acquired, or lost, allowing the Sar-

coma 180 cell to become similar to the

normal, highly invasive crest cell. This re-

sult raises the possibility that any invasive

cell type, such as primordial germ cells,

leucocytes, or tumor cells, may also be sim-

ilar to neural crest cells.

To further complicate the issue, retinal

pigment epithelial cells (RPE cells), the

only pigment cell in the body not of neural

crest origin and normally non-motile, when

injected into somites, migrated along the

ventral crest cell pathway.'* Surprisingly,

also found to be migratory were polysty-

rene latex beads of 6.5 to 35 microns in di-

NEURAL CREST CELL MIGRATION AND THE EXTRACELLULAR MATRIX 133

ameter, as well as polystyrene latex beads

coated with bovine serum albumin (BSA).

The time course of translocation, and the

terminal distribution of RPE cells, uncoated

latex beads, and BSA coated latex beads,

were characteristic of normal, migratory

crest cells. However, latex beads coated

with fibronectin were not migratory and

remained near the site of implantation.’

The fact that an inert material has the abil-

ity to “migrate” along the ventral crest cell

pathway as easily as both crest cells and

crest cell derivatives, quickly dispels the

theory that the crest cell pathways interact

with neural crest cells in a precise, specific,

and unique manner. Rather, these experi-

ments elucidated several novel facts, the

first being that although neural crest cell

derivatives are preferentially translocated

along the migratory pathways when com-

pared to somitic cells and fibroblasts, they

are not the only cell type which can migrate

along the crest cell pathways, as evidenced

by RPE cells and Sarcoma 180 cells. Sec-

ondly, the path of migration for crest cells

and crest cell derivatives is not specific—

whether or not they migrate along the dor-

solateral or the ventral pathway probably

depends more on their initial spatial loca-

tion than on any type of cellular specificity.

Lastly, inert materials which are approxi-

mately the same size as neural crest cells

can interact with the extracellular envi-

ronment along the crest migratory path-

way and can distribute themselves in the

same terminal locations as do normal crest

cells. Thus, since the neural crest migratory

pathways are not specific for crest cells, the

local environment surrounding the nearby

cells must play a crucial role in selecting the

cell types which can migrate along the

pathways, as well as determining the final

pattern of crest cell distribution.

6. Components of the Extracellular Matrix

As crest cells emerge from the neural

tube and begin their migration they enter a

cell-free space bounded by basement mem-

branes and filled with an amorphous extra-

cellular matrix (ECM). Previously, it was

thought that the ECM was a relatively inert

material produced by epithelial and con-

nective tissue cells primarily to serve as a

supporting structure. However, it is now

realized that cells which produce ECM

components maintain an ongoing interac-

tion with the ECM, as do other cells in the

area. Before one can intelligently investi-

gate how crest cell migration might be af-

fected by the local environment, the struc-

ture and composition of the ECM along the

neural crest migratory pathway must first

be established. Then it should be deter-

mined which cell types along the migratory

pathways produce components of the ECM.

Finally, an attempt should be made to deter-

mine how specific developmental changes

in the ECM along the migratory pathways

can influence the migration and cytodiffer-

entiation of neural crest cells. To begin, this

section will look at the possible role of the

basal lamina in crest cell migration, and

then will look at three different compo-

nents of the ECM: collagen, glycosamino-

glycans, and fibronectin.

Basal Lamina

The basal lamina, often interchangeably

used with the term basement membrane, is

a moderately electron-dense structure which

underlies all epithelial cells, and surrounds

muscle cells and fat cells. The basal lamina

extends 50 to 100 nm from the cell surface

and comprises three distinct structural sub-

units: the lamina densa, a thick, central,

electron-dense collagenous sheet, the lam-

ina rara externa, which is less electron-

dense and separates the lamina densa from

the cell surface, and the electron-lucent

lamina rara interna, which separates the

lamina densa from the underlying ECM.”

Macromolecular components of the basal

lamina include collagenous proteins, pri-

marily comprised of Type IV collagen; pro-

teoglycans, primarily made up of heparan

sulfate and chondroitin sulfate; fibronec-

tin, and laminin.”

134 BRIAN E. LACY

There appear to be three distinct ways by

which the basal lamina could modulate

neural crest cell migration. The basal lam-

ina of the ectoderm, somites, and neural

tube, which surrounds the cell-free space

through which neural crest cells migrate,

could well serve as a structural barrier to

prevent cell-to-cell contact between migrat-

ing crest cells and other embryonic cells.

First, by inhibiting cellular contact, which

might then lead to adhesion and condensa-

tion, the basal lamina would provide a

finely delimited border for cellular migra-

tion. Secondly, Bancroft and Bellairs”' have

reported that processes of neural crest cells

are often in intimate contact with the basal

lamina which might serve as a substratum

for cell migration. However, Tosney° has

noted that migrating neural crest cells con-

tact fibrils of the ECM as often as they do

the basal lamina, and thus this may be

more of a random process, rather than a

specific process, where precise migrational

clues are provided. Lastly, as mentioned

previously, the basal lamina of the neural

tube becomes discontinuous superiorly as

crest cell emigration begins, while the neigh-

boring basal lamina of the ectoderm and so-

mites remains continuous, as does the basal

lamina on the lateral surfaces of the neural

tube.° Later, at the same time that crest cell

emigration from that particular axial level

has ceased, the basal lamina reforms on the

superior aspect of the neural tube.° By a

precise temporal sequence of production,

degeneration and then production, the basal

lamina could serve as a “‘gate”’ to crest cell

migration. During early neurulation the

basal lamina would prevent the early mi-

gration of neural crest cells. At the time of

crest cell migration, specific enzymes could

degrade the basal lamina, thereby allowing

the neural crest cells to extend their proc-

esses out and away from the neural tube,

and so begin their emigration. Crest cell

emigration could later be stopped by in-

creased synthesis of basal lamina compo-

nents and reformation of the basal lamina.

Possible factors which might be involved in

this process have yet to be identified.

Collagen

The collagens are a family of fibrous pro-

teins which account for greater than twenty-

five percent of the total body protein. Cur-

rently, five major types of distinct collagen

(Types I through V) are recognized, based

on differences in the composition of the a

chains. Since these different collagens are

all closely related, they will be referred to in

a general, collective sense, except in cases

where the knowledge of the specific type of

collagen involved in an experiment is cru-

cial to our understanding and interpreta-

tion of that experiment.

Collagens can be readily identified by

several characteristics: (1) their triple helix

structure, (2) the longitudinal staggering of

collagen fibrils which results in regular

cross-striations every 67 nm in collagen

fibers, (3) the high degree of resistance of

the triple helix to degradation via pro-

teases—the collagenases are the only en-

zymes to efficiently digest the triple helix

portion of the collagen molecule, (4) the

polypeptide a chains have a repeating

structure which can be characterized by

(G—X—Y)n, where G stands for glycine.

Thus, glycine accounts for every third

amino acid residue in the collagen mole-

cule. The variables X and Y can be any

other amino acid although X = proline in

about 10% of the cases and Y = hydroxy-

proline about 10% of the time.

After the prepro-a chains are synthe-

sized on ribosomes associated with the en-

doplasmic reticulum (ER), they enter the

ER where the signal sequence is then

cleaved off. In the lumen of the ER, hy-

droxylation of lysine and proline residues

on the pro-a@ chain occurs. Subsequent to

hydroxylation, glycosylation of the hydroxyl

group on the hydroxylysine may occur.

These covalently attached oligosaccharide

chains are very short, usually being only

two sugar residues in length. After glycosy-

lation, three pro-a chains combine to form

a left-handed, triple helix structure termed

procollagen. After being exocytosed, pro-

collagen proteases on or near the cell sur-

;

; ;

;

NEURAL CREST CELL MIGRATION AND THE EXTRACELLULAR MATRIX 135

face remove the non-helical extension pep-

tides found at both the carboxyl and amino

terminal ends, to form the definitive colla-

gen molecule. At this stage the a chains

each contain approximately 1000 amino

acid residues, and the collagen molecule is

approximately 300 nm long and 1.5 nm in

diameter. The newly formed collagen mole-

cules spontaneously assemble into collagen

fibrils which then covalently cross-link to

each other to increase their tensile strength.

In terms of its distribution, collagen is very

widespread and is found in and around al-

most every tissue in the body.

Currently, there is no direct evidence

linking the formation and deposition of

collagen in the ECM with the control or

regulation of neural crest cell migration.

Although collagen synthesis is usually as-

sociated with mesenchymal cells, the neu-

ral tube, the notochord, and neural crest

cells have all been shown to produce colla-

gen.” > Collagen is first seen in the inner

cell mass of the three day mouse blastocyst

via immunofluorescence.” From stage 12

on, both striated and non-striated fibrils of

various sizes (2 nm to 100 nm) have been

observed in the ECM surrounding the mi-

grating crest cells.” *° Due to their charac-

teristic banding, the striated fibrils are as-

sumed to be small collagen fibrils.

The collagen of the ECM exists as a loose

meshwork of fibrils which increases in vol-

ume at the time of neural crest cell migra-

tion and is usually decorated, or coated,

with fine granules of proteoglycans as iden-

tified by ruthenium-red staining.” *° The

orientation of the collagen fibrils in the

ECM surrounding the crest cells is com-

pletely random; there does not appear to be

any consistent orientation of fibrils along

either the dorsolateral or ventral migratory

pathways in the avian embryo.° Although

it has been shown in vitro that neural crest

cells can adhere to collagen,”’ there has

been no evidence that crest cells selectively

adhere to or attach to collagen fibrils during

migration in vivo. Collagen in the ECM

should be thought of as providing a general

structural framework for the deposition of

fibronectin, glycosaminoglycans, and gly-

coproteins, rather than as a specific sub-

stratum upon which neural crest cells mi-

grate.

Fibronectin

Fibronectin is one of several, large, non-

collagenous glycoproteins (the others being

laminin and chondronectin) which have re-

cently been implicated in the processes of

cell-cell adhesion and cell-substratum ad-

hesion. Fibronectin can be identified via

immunofluorescence in the blood, in the

ECM, and in basement membranes. On the

cell surface, fibronectin appears as an in-

terwoven, fibrillar material which radiates

out from, and around, the periphery of the

cell. Although antigenically indistinguish-

able, two major types of fibronectin have

been identified and characterized biochem-

ically; plasma fibronectin, also called cold

insoluble globulin, and cellular fibronec-

tin, also referred to as the LETS protein

(large, external, transformation-sensitive

protein). Cellular fibronectin predominantly

exists as dimers and polymers of a single

subunit, with the molecular weight of the

subunit being in the range of 220,000 to

250,000 Daltons. The subunits are linked

by disulfide bonds which are all displaced

towards one end of the molecule. Unlike

collagen, fibronectin is a very asymmetrical

molecule; the polypeptide backbone does

not exhibit a regular, repeating secondary

structure.

The carbohydrate component of fibro-

nectin usually exists as five similar oligo-

saccharide units attached covalently to as-

paragine residues along the polypeptide

chain. Functionally, the oligosaccharide

side chains are primarily used to stabilize

the fibronectin molecule, and protect it

against proteolytic attack.** By carefully

digesting specific peptide bonds via differ-

ent proteases, researchers have shown that

the fibronectin molecule contains functional,

as well as structural, binding domains. Fi-

bronectin appears to have separate binding

136

regions for such diverse molecules as colla-

gen,’ heparin, fibrin, and hyaluronic acid.”

The discovery that fibronectin contains

specific binding regions for other compo-

nents of the ECM, and that it promotes the

adhesion of fibroblasts to collagen, quickly

led to the theory that fibronectin could well

act as either a substrate, or a chemoattract-

ant, for neural crest cell migration. Fibro-

nectin is produced during early avian de-

velopment by cells of the neural tube,

notochord, and somites.” Immunocytochem-

istry revealed that fibronectin was localized

in the basal lamina of the dorsal neural

tube, on 5-10 nm non-striated fibrils in the

ECM, and in the interstitial bodies first

described by Low.” Low” had described

sphere-like structures apparently composed

of ground substance which correlated ex-

tremely well, both spatially and tempor-

ally, with the pathways of neural crest cell

migration. The interstitial bodies were seen

to increase in number prior to migration,

remain in high concentration in the regions

where crest cells were just about to enter,

and decrease in concentration in the re-

gions where crest cell migration was declin-

ing. These aggregations of ground sub-

stance are now known to be composed of

fibronectin and a mixture of proteoglycans

and glycosaminoglycans.*” ~*

Ali and Hynes”™ have shown that the ad-

dition of fibronectin to both normal and

transformed cells in culture leads to an in-

crease in the motility and migration of

these cells, as well as to an increase in the

attachment of these cells to their substrata.

In support of this, it has been shown that

cells are able to attach to all types of colla-

gen (Types I through V) and that the per-

centage of binding is increased when fibro-

nectin is added.”’ In addition, studies using

a modified Boyden chamber have shown

that neural crest cells migrate in the direc-

tion of fibronectin, and when given a choice,

preferentially move towards the region of

greater fibronectin concentration.”

It was previously mentioned that poly-

styrene latex beads, Sarcoma 180 cells and

neural crest cells all have two factors in

common: (1) they are all able to freely

BRIAN E. LACY

translocate along crest migratory pathways,

and (2) they do not produce, nor are they

coated with, fibronectin. However, poly-

styrene beads coated with fibronectin; and

somite and fibroblast cells, both of which

produce fibronectin, were all incapable of

migrating along the crest pathways.'” '”

The lack of fibronectin on the surface of

crest cells may very well drive them towards

a source of fibronectin, substantiating the

theory that fibronectin is a chemoattract-

ant for neural crest cells.*’ Crest cell migra-

tion could occur based solely on a concen-

tration gradient of fibronectin along the

migratory pathways. At the terminal loca-

tion, the concentration of fibronectin would

be so great that further migration could not

occur, and cell adhesion and crest cell con-

densation would occur. Data showing that

increased numbers of catecholamine-posi-

tive cells occur in response to exogenous fi-

bronectin supports the theory that a high

concentration of fibronectin would not

only cause condensation of crest cells but

could also induce terminal cytodifferentia-

tion?

Glycosaminoglycans

Glycosaminoglycans are long, unbranched

polysaccharide chains which are charac-

teristic of the extracellular matrix and the

intercellular spaces. Glycosaminoglycans

(GAG or GAGs) are composed of repeat-

ing disaccharide units in which one of the

two sugar residues is always an amino

sugar (either N-acetylglucosamine or N-

acetylgalactosamine). Each subunit con-

tains an acidic carboxyl group, sulfate

group, or both, and thus GAGs are highly

charged, polyanionic compounds. Unlike

proteins, which often assume a globular

shape, long polysaccharide chains are fairly

inflexible and tend to assume an extended,

random coil conformation. Thus, even

though their molecular weight is usually

fairly low (4,000 to 50,000 Daltons), GAGs

occupy a very large volume relative to their

mass. Being hydrophilic, as well as having a

high density of negative charges, explains

why GAGs attract large amounts of water,

NEURAL CREST CELL MIGRATION AND THE EXTRACELLULAR MATRIX 137

forming a highly hydrated gel. The porous

arrangement of the hydrated matrix allows

for the diffusion of water-soluble nutrients,

and the migration or movement of cells.

At present, seven distinct groups of gly-

cosaminoglycans have been isolated, based

on the sugar residues which make up the

disaccharide subunits, the type of linkages

used, and whether or not the subunits are

sulfated. These GAGs are hyaluronate

(which is not sulfated), chondroitin-4-sul-

fate, chondroitin-6-sulfate, dermatan sul-

fate, heparan sulfate, heparin, and keratan

sulfate. In the extracellular matrix, GAGs

are almost never found in the free state as

isolated polysaccharide chains (except for

hyaluronic acid); rather, they are coval-

ently bound to a protein core to form pro-

teoglycans (formerly called mucoproteins).

In a typical cartilage proteoglycan mole-

cule, the core protein will have a molecular

weight of 250,000 Daltons, and will have

approximately eighty chondroitin sulfate

chains covalently attached to it.*° In other

tissues, the size of the protein core will

vary, and the number and types of GAGs

which radiate out from the protein core will

change as well. This section will deal pri-

marily with the structure, and possible

function, of hyaluronic acid in the ECM of

the developing embryo, although the other

GAGs will be frequently referred to as well.

Hyaluronate differs from the other GAGs

in three distinct ways: the disaccharide

subunit structure of hyaluronate (HA) is

invariably made up of D-glucuronic acid

and N-acetyl-D-glucosamine—there are no

other sugars involved; HA consists of a

polysaccharide chain which may be several

thousand residues long and have a molecu-

lar weight of several million Daltons; HA is

usually found as a free polysaccharide

chain in the ECM, rather than being coval-

ently bound to a protein molecule. If hyal-

uronic acid is to play a role in neural crest

cell migration and differentiation, then

local changes in the distribution and con-

centration of HA should occur during de-

velopment. Toole and Trelstad*’ showed

that initially, HA concentration in the cor-

nea is low, but just prior to the rapid inva-

sion of the cornea by mesenchymal cells,

HA production significantly increased. The

high concentration of HA in the primary

stroma of the cornea results in a large in-

flux of water, which causes the tissue to

swell. Mesenchymal cells then migrate

through the porous, gel-like matrix and set-

tle within the cornea. Shortly thereafter,

synthesis of HA declines, and hyaluroni-

dase activity increases, leading to dehydra-

tion and compaction of the corneal stroma.

In 1972, Toole*® was able to show that a

similar temporal pattern of HA synthesis,

followed by cell migration, and concluded

by both a decrease in HA synthesis and an

increase in hyaluronidase activity, occurred

in limb bud and axial chondrogenesis of the

developing chick. In addition, in vitro stud-

ies showed that high levels of HA inhibited

chondrogenesis. These results indicate that

HA may inhibit, or delay, cytodifferentia-

tion of chondroblasts by interacting with

cell surface receptors.’ Although far from

definitive, these experiments show that a

precise temporal pattern of HA concentra-

tion could well play a role in cell migration

and differentiation.

Hyaluronic acid is synthesized and se-

creted by tissues surrounding the migra-

tory pathways of the neural crest cells. The

isolated neural tube and the surface ecto-

derm are both significant sources of HA

during avian embryogenesis.*” *’ In the de-

veloping chick, seventy-five to eighty-five

percent of all GAGs secreted into the cell-

free space bounded by the somites, the sur-

face ectoderm, and the neural tube, is hyal-

uronic acid.*” *° In addition, it was found

that at anterior axial levels, where crest cell

migration had ceased, the concentration of

HA had declined, concomitant with an in-

crease in the concentrations of chondroitin

sulfate and heparan sulfate. However, at

more posterior levels, where crest cell mi-

gration was just beginning, HA concentra-

tion was very high, but chondroitin sulfate

and heparan sulfate levels were very low.*” “*

The polyanionic, hydrophilic nature of

hyaluronate appears to make it ideally

suited as an environment capable of sup-

porting cell migration. Soluble growth fac-

138 BRIAN E. LACY

tors, trophic factors, and nutrients are all

capable of diffusing through the porous,

gel-like HA domain of the ECM, as are

neural crest cells. An initial high concentra-

tion of HA would allow cell migration to

occur, and possibly, delay terminal cytodif-

ferentiation. A decline in HA concentra-

tion brought about by both a decrease in

HA production and an increase in hyalu-

ronidase activity, would make migration

much more difficult by closing off the po-

rous, migratory pathways, and might also

allow crest cells to complete their terminal

differentiation. Most likely, crest cell mor-

phogenesis is not regulated solely by the

environment acting on the crest cell; rather,

it is more of a reciprocal type of regulation,

since it has been shown that neural crest

cells in culture can produce and secrete a

variety of GAGs, including HA.*” *°

7. Conclusion

The migration and differentiation of the

neural crest is a fascinating developmental

process which occurs with extraordinary

precision and predictability. Although this

topic has been extensively studied for the

last fifty years, and described for more than

one hundred years, the basic mechanisms

responsible for crest cell migration and

cytodifferentiation still remain largely un-

solved. Neural crest cell migration encom-

passes an array of topics fundamental to

‘the study of developmental neurobiology.

These include cell migration, cell-cell adhe-

sion, cell-substratum adhesion, cell surface

properties, and the genotypic and pheno-

typic expressions of the neuroectodermally

derived cells. Tireless, persistent research

into each of the above fields is required if

we are ever to grasp the fundamentals of

crest cell migration. Future research in the

field of the neural crest should be oriented

with the following goals in mind. First, all

aspects of the crest cell surface must be thor-

oughly characterized at each stage of de-

velopment. Secondly, all the components

of the ECM must be assayed in vitro over a

wide range of concentrations to determine

both the degree, and the mechanism, of

ECM interaction with the crest cell surface.

Lastly, precise temporal and spatial meas-

urements of all components of the ECM

must be made in vivo. The results of these

studies would provide much needed infor-

mation detailing how the crest cell surface

interacts with the ECM and how neural

crest cells interact with one another. In ad-

dition, the ability of the ECM to regulate

crest cell migration and development, as

well as the crest cell’s ability to respond to

such modulation, could also be ascertained.

Neural crest cell migration occurs so

precisely, and so predictably, that unfortu-

nately, it is often thought of as a simple,

well defined process. As this review has

shown, neural crest cell development and

migration are amazingly complex processes,

in which the basic rules governing these

two processes are still largely unknown.

References Cited

1. Weston, J. A. 1963. A radioautographic analysis

of the migration and localization of trunk neural

crest cells in the chick. Dev. Bio., 6: 279-310.

2. Johnston, M. C. 1966. A radioautographic study

of the migration and fate of cranial neural crest

cells in the chick embryo. Anat. Rec., 156: 143-156.

3. Noden, D. M. 1975. An analysis of the migratory

behavior of avian cephalic neural crest cells. Dev.

Bio., 42: 106-130.

4. Hamburger, V. and Hamilton, H. 1951. A series of

normal stages in the development of the chick

embryo. J. Morphol., 88: 49-92.

5. Bancroft, M. and Bellairs, R. 1975. Differentiation

of the neural plate and neural tube in the young

chick embryo: A study by scanning and transmis-

sion electron microscopy. Anat. Embryol., 147:

309-335.

6. Tosney, K. W. 1978. The early migration of neural

crest cells in the trunk region of the avian embryo:

An electron microscopic study. Dev. Bio., 62:

317-333.

7. Weston, J. A. and Butler, S. L. 1966. Temporal fac-

tors affecting iocalization of neural crest cells in

the chicken embryo. Dev. Bio., 14: 246-266.

8. Detwiler, S. R. 1937. Observations upon the mi-

gration of neural crest cells, and upon the devel-

opment of the spinal ganglia and vertebral arches

in Amblystoma. Am. J. Anat., 61: 63-94.

9. Le Douarin, N. 1973. A biological cell labeling

10.

11.

12.

23.

14.

15.

16.

Pi.

18.

19.

20.

21.

me.

24.

NEURAL CREST CELL MIGRATION AND

technique and its use in experimental embryology.

Dev. Bio., 30: 217-222.

Erickson, C. A., Tosney, K. W. and Weston, J. A.

1980. Analysis of migratory behavior of neural

crest and fibroblastic cells in embryonic tissues.

Dev. Bio., 77: 142-156.

Bronner, M. E. and Cohen, A. M. 1979. Migratory

patterns of cloned neural crest melanocytes in-

jected into host chicken embryos. Proc. Natl.

Acad. Sci. USA, 76: 1843-1847.

Newgreen, D. 1982. The role of the extracellular

matrix in the control of neural crest migration. In:

Extracellular Matrix. S. Hawkes and J. L. Wang,

eds., New York, Academic Press. pp. 141-145.

Le Douarin, N. 1980. Migration and differentia-

tion of neural crest cells. In: Current Topics in

Developmental Biology, Vol. 16. Academic Press,

Inc. New York. pp. 31-85.

Weston, J. A. 1983. Regulation of neural crest cell

migration and differentiation. In: Cell Interac-

tions and Development: Molecular Mechanisms.

John Wiley and Sons, New York. pp. 153-184.

Le Douarin, N. M. and Teillet, M. M. 1974. Exper-

imental analysis of the migration and differentia-

tion of neuroblasts of the autonomic nervous sys-

tem and of neural ectodermal mesenchymal

derivatives, using a biological cell marking tech-

nique. Dev. Bio., 41: 162-184.

Le Lievre, C. S., Schweizer, G. G., Ziller, C. M.

and Le Douarin, N. M. 1980. Restrictions of de-

velopmental capabilities in neural crest cell deriv-

atives as tested by in vivo transplantation experi-

ments. Dev. Bio., 77: 362-378.

Bronner-Fraser, M. and Cohen, A. M. 1980. Anal-

ysis of the neural crest ventral pathway using in-

jected tracer cells. Dev. Bio., 77: 130-141.

Bronner-Fraser, M. 1982. Distribution of latex

beads and retinal pigment epithelial cells along

the ventral neural crest pathway. Dev. Bio., 91:

50-63.

Farquhar, M. G. 1981. The glomerular basement

membrane: A selective macromolecular filter. In:

Cell Biology of the Extracellular Matnx. E. D.

Hay, ed., New York, Plenum Press. pp. 335-378.

Sanes, R. J. 1982. Laminin, fibronectin, and col-

lagen in synaptic and extrasynaptic portions of

muscle fiber membrane. J. Cell Bio., 93: 442-

451.

Bancroft, M. and Bellairs, R. 1976. The neural

crest cells of the trunk region of the chick embryo

studied by SEM and TEM. Zoon., 4: 73-85.

Cohen, A. M. and Hay, E. D. 1971. Secretion of

collagen by embryonic neuroepithelium at the

time of spinal cord-somite interaction. Dev. Bio.,

26: 578-605.

Greenberg, J. H., Foidart, J. M. and Greene, R. M.

1980. Collagen synthesis in cultures of differen-

tiating neural crest cells. Cell Diff., 9: 153-163.

Leivo, I., Vaheri, A., Timpl, R. and Wartiovaara, J.

1980. Appearance and distribution of collagens

and laminin in the early mouse embryo. Dev.

Bio., 76: 100-114.

25.

26.

27.

28.

30.

31.

32.

33.

34.

35.

36.

eye

38.

39.

40.

41.

42.

THE EXTRACELLULAR MATRIX 139

Hay, E. D. 1978. Fine structure of embryonic ma-

trices and their relation to the cell surface in ruthe-

nium red-fixed tissues. Growth, 42: 399-423.

Lofberg, J., Ahlfors, K., and Fallstrom, C. 1980.

Neural crest cell migration in relation to extra-

cellular matrix organization in the embryonic

Axolotl trunk. Dev. Bio., 75: 148-167.

Greenberg, J. H., Seppa, S., Seppa, H. and Hewitt,

A. T. 1981. Role of collagen and fibronectin in

neural crest cell adhesion and migration. Dev.

Bio., 87: 259-266.

Olden, K., Pratt, R. M. and Yamada, K. M. 1979.

The role of carbohydrate in biological function of

the adhesive protein fibronectin. Proc. Natl. Acad.

Sci. USA 76: 3343-3347.

Hahn, L. and Yamada, K. M. 1979. Identification

and isolation of a collagen binding fragment of

the adhesive glycoprotein fibronectin. Proc. Natl.

Acad. Sci. USA 76: 1160-1163.

Yamada, K. M., Kennedy, D. W., Kimata, K. and

Pratt, R. M. 1980. Characterization of fibronectin

interactions with glycosaminoglycans and identi-

fication of active proteolytic fragments. J. Biol.

Chem., 255: 6055-6063.

Sieber-Blum, M., Sieber, F. and Yamada, K. M.

1981. Cellular fibronectin promotes adrenergic

differentiation of quail neural crest cells in vitro.

Exp. Cell Res., 133: 285-295.

Mayer, B. W., Hay, E. D. and Hynes, R. O. 1981.

Immunocytochemical localization of fibronectin

in embryonic chick trunk and area vasculosa.

Dev. Bio., 82: 267-286.

Low, F. N. 1970. Interstitial bodies in the early

chick embryo. Am. J. Anat., 128: 45-56.

Ali, I. V. and Hynes, R. O. 1978. Effects of LETS

glycoprotein on cell motility. Cell, 14: 439-446.

Loring, J., Glimelius, B. and Weston, J. A. 1982.

Extracellular matrix materials influence quail

neural crest cell differentiation in vitro. Dev. Bio.,

90: 165-174.

Hascall, V. C. and Hascall, G. K. 1981. Proteogly-

cans. In: Cell Biology of the Extracellular Matrix.

E. D. Hay, ed., New York, Plenum Press. pp.

39-63.

Toole, B. P. and Trelstad, R. L. 1971. Hyaluronate

production and removal during corneal develop-

ment in the chick. Dey. Bio., 26: 28-35.

Toole, B. P. 1972. Hyaluronate turnover during

chondrogenesis in the developing chick limb and

axial skeleton. Dev. Bio., 29: 321-329.

Toole, B. P., Jackson, G. and Gross, J. 1972. Hy-

aluronate in morphogenesis: Inhibition of chon-

drogenesis in vitro. Proc. Natl. Acad. Sci. USA 69:

1384-1386.

Hay, E. D. and Meier, S. 1974. Glycosaminogly-

can synthesis by embryonic inductors: Neural

tube, notochord, and lens. J. Cell. Bio., 62: 889-898.

Solursh, M., Fisher, M. and Singley, C. T. 1979.

The synthesis of hyaluronic acid by ectoderm dur-

ing early organogenesis in the chick embryo.

Diff., 14: 77-85.

Pratt, R. M., Larsen, M. A. and Johnston, M. C.

1975. Migration of cranial neural crest cells in a

cell-free hyaluronate rich matrix. Dev. Bio., 44:

298-305.

43. Solursh, M. 1976. Glycosaminoglycan synthesis

in the chick gastrula. Dev. Bio., 50: 525-530.

44. Bolender, D. L., Seliger, W. G. and Markwald, R. R.

1980. A histochemical analysis of polyanionic

compounds found in the ECM encountered by

Journal of the Washington Academy of Sciences,

Volume 73, Number 4, Pages 140-150, December 1983.

migrating cephalic neural crest cells. Anat. Rev.,

196: 401-412.

45. Greenberg, J. H. and Pratt, R. M. 1977. Glycosa-

minoglycan and glycoprotein synthesis by cranial

neural crest cells in vitro. Cell Diff., 6: 119-132.

46. Pintar, J. E. 1978. Distribution and synthesis of

glycosaminoglycans during quail neural crest

morphogenesis. Dev. Bio., 67: 444-464.

Bayesian Estimation of Reliability in the

Stress-Strength Context

Mark Angelo Johnstone

United States Military Academy

The assessment of reliability, defined as the probability that actual strength exceeds in use

stress, is considered from a Bayesian viewpoint. Quantal response data generated when testing

is performed at two stress levels, y; and y2, at which mj failures out of nj; tests (i = 1, 2) are

observed. It isassumed that p; < p2, where pjis the unknown probability of failure at yi. From

these observations a joint posterior distribution for p; and p2 is developed assuming inde-

pendence of p, and po. The conditional marginal posterior distributions of p; and p2 given the

condition that p; < p2are then used to determine a posterior distribution for reliability. Nor-

mal distributions are assumed for both the stress and strength distributions. Asa result of this

Bayesian analysis, point and confidence interval estimates for reliability can be determined. A

Monte Carlo simulation approach to this problem is also discussed and the results compared.

In addition, applications of the approach to other situations involving similar quantal re-

sponse data are discussed.

1. Introduction

In recent years the development of large

scale systems for commercial and military

use has challenged engineers and mathema-

ticians with the task of assessing their in-

use performance. Reliability is a measure

of a system’s ability to perform under cer-

tain conditions and concerns itself with sys-

tem failure, system conformity to specifica-

tions, and system life expectancy. Reliability

as it is addressed in this paper is defined as

the probability that system “strength” ex-

ceeds in-use “‘stress.”’

The stress variable, Y, represents the

levels at which the system is utilized. It is

often referred to as the requirement vari-

able because it represents the levels at

which the system is designed to and ex-

pected to perform. Y may take on a con-

stant value when the system is always sub-

jected to one level of stress. It may also be

randomly distributed as in the case where

the system is used at different stress levels.

It can be measured by examining the fre-

quency that the system is used at different

levels or it may be set as a specification re-

quirement. This paper considers the case

where Y follows a specified (known) nor-

mal (Gaussian) distribution.

The random strength variable, X, rep-

resents the critical stress level at which a

particular system fails. For the population

of systems of the same design, the strengths

will follow some probability distribution.

This paper considers the case where X also

140

BAYESIAN ESTIMATION OF RELIABILITY 141

£y (x) » Strength Distribution

f(y)» In-Use Stress Distribution

Fig. 1. Examples of stress and strength distributions.

follows an unknown normal distribution.

Figure | illustrates a typical relationship

which might exist between the random vari-

ables X and Y. In this context reliability

would then be defined as, R = Pr{X > Y}.

In order to make inferences concerning R,

data must be collected to gain information

regarding the distribution of X. There are

many methods or experimental designs ac-

cording to which this data might be collected.

Very often data referred to as quantal re-

sponse data is the only kind of data which

- may be collected in a particular situation.

Quantal response data involves testing a

sample of identical test specimens at a va-

riety of stimulus levels and observing

whether or not a response occurs. For a

given test specimen a response will occur if

the applied stimulus exceeds the critical

level of stimulus associated with the test

specimen. For quantal response data the

test specimens cannot be retested because

they are either destroyed or their properties

are changed by the testing. For the tests

where a response did not occur, there is no

way of determining how much more stimu-

lus would have been necessary to cause a

response. For tests where a response did

occur there is no way of determining by

how much the critical level was exceeded.

In the stress-strength context the stimulus

is the stress applied to the test specimen, the

critical level is the strength of the test speci-

men, and a response is observed when the

test specimen fails—that is, strength is less

than the stress.

There are many test strategies for select-

ing the stimulus or stress levels which will

be applied to the test specimens. For the

purposes of this paper it will be assumed

that Churchman two-stimuli designs’ were

used to generate the data. Such designs in-

volve testing two samples of test speci-

mens-n, at stress level y; and n> at stress

level y2 and observing the number of fail-

ures m; and mz, respectively. Stress levels y;

and y2 are chosen so that:

0 < (m;/n\) < (m2/n2) < 1.

Before proceeding it is appropriate that

an example of the stress-strength reliability

situation be discussed. Consider the case

where the reliability of an anti-tank sabot

round is defined in terms of its ability to

penetrate the armor of a Soviet T-62 tank

at ranges (stresses) defined by a given nor-

mal distribution. In order fora sabot round

to complete its mission of destroying the

tank, it must completely penetrate the

armor. For each round there is a critical

range at which it will no longer completely

penetrate the armor. This critical range

typically depends on many different pene-

tration factors (e.g. muzzle velocity, weight,

impact angle, etc.). The population of

rounds can be considered to have a strength

distribution of these critical ranges which is

normally distributed. Tests are conducted

using the Churchman approach from which

quantal response data is observed. We

might begin our testing by firing a round at

a distance Y away from an armor plate sim-

ilar to the armor used on the T-62. If pene-

tration does not occur (e.g. the round does

not go all the way through the plate), we

142

must decrease the range. To do this the

cannon from which the round is fired is

moved closer to the plate by one increment,

perhaps 50 meters, and fired again. This

procedure is repeated until a round pene-

trates the plate. n; rounds are fired at this

distance y;, and the number of non-pene-

trations, m;, is recorded. If the number of

non-penetrations is greater than 50 per-

cent, the rounds are fired from a distance

two increments closer to the plate (100 me-

ters). n2 rounds are fired, recording the

number of non-penetrations at this dis-

tance. If the number of non-penetrations at

yi had been less than 50 percent, the rounds

would have been fired from a distance one

half increment (25 meters) further away

from the plate and the number of non-

penetrations recorded. The objective is to

achieve two probabilities of non-penetra-

tion at two levels y; and y2 where neither

probability is 100 percent nor zero percent

and where both probabilities differ by at

least twenty percent. This data may then be

used to estimate the parameters of the

strength distribution of critical ranges for

the sabot round. The model upon which

these estimates can be made and the meth-

ods of estimation will be determined in the

next section of this paper.

There are many other real world situa-

tions where quantal response data is gener-

ated and a similar analysis can be per-

formed. When determining critical dosage

Stress Distribution, f(y)

MARK ANGELO JOHNSTONE

levels for biology cultures, a dosage level is

mixed and after a prescribed time period

either a reaction occurs or it does not.

When determining the impact sensitivity of

an explosive round, a weight is dropped at

different heights on a small quantity or

charge of the explosives and either the

charge explodes or it does not. When de-

termining crash survival speeds of auto-

mobiles, the automobile is driven into a

barrier and the impact damage is either

great enough to kill the occupants or it is

not.

2. The Stress-Strength Reliability Model

In this paper we are interested in making

estimates of reliability, R, defined as the

probability that a random variable, X,

where X denotes the strength, exceeds a

random variable Y, where Y denotes stress.

The probability density function (pdf) of X

and Y are defined as n(x; ps, o8) and n(y;

ur, of) where,

] Pik: (ui yp)

ov 27 o

Figure 2 illustrates the relationship be-

tween the distributions for X and Y. It is

important to note that for the purposes of

this paper weand ogare considered as given

n(u; uw, 0”) =

Strength Distribution, f(x)

Fig. 2. Relationship between stress (Y) and strength (X) distributions and Pr(X > Y).

BAYESIAN ESTIMATION OF RELIABILITY 143

values. We may now define R mathemati-

cally as:

R = Pr(X > Y) = Pr(X — Y > 0)

If we define Z = X — Y, then Zis normally

distributed with pdf given by n(z; us — ur,

og + og). Therefore Rcan be calculated as:

R =[ n(z|us — we, os + o«)dz

For Churchman type data, a sample of

size n, is tested at stress level y;, and it is ob-

served that m; devices fail. m, is then the

observed value of a random variable hav-

ing a binomial distribution with probabil-

ity mass function (pmf) given by b(m;; n;,

pi),m; = 0,1,2,.. . n; where p; is defined

as:

Pi =A n(x; ps, o§)dx = Pr(X < y:)

A sample of size n2 is then tested at stress

level y2 where y2 > y:, and it is observed

that m2 devices fail. m2 is then the observed

value of a random variable having a bi-

nomial distribution with pmf given by

b(m2; n2, p2), m2 = 0, 1,2, . . . np where p»

is defined as:

p2= q. n(x; ps, og)dx = Pr(X < yo)

since, y; < y2 and og > 0, it must be the

case that p; < p>.

For this data the classical statistical es-

timate for R is provided by first solving the

following two equations simultaneously

for estimates of us and o8, denoted jis and

68:

Pi =| n(x; us, 08)dx =m,/n,; Egn.2.1

P2 =[

Reliability is then estimated by:

n(x; us, o8)dx = m2/n, Egn.2.2

oS i. n(z| fis — pr, 68 + on)dz

Egn. 2.3

The major limitation of this classical ap-

proach is that it does not provide a measure

of uncertainty in the estimate of R based

upon the data available. That is, there is no

straightforward way of determining confi-

dence limits for R using the classical ap-

proach. The next section of this paper pre-

sents a Bayesian approach which provides

a means of computing confidence limits.

3. Bayesian Analysis of the Model

The purpose of the Bayesian analysis is

to determine a posterior distribution on the

reliability, R, assumed to be a random vari-

able in the Bayesian sense. Our Bayesian

analysis of the model requires that we first

develop posterior distributions on p; and

Pp2 from which we may determine the first

and second moments of p; and p> condi-

tional upon p; being less than p>. Using the

first conditional moments of p; and p2 we

may solve for estimates of the posterior

means of ws and o§ for the strength distri-

bution (assumed normal). The strength dis-

tribution is then compared with the stress

distribution (given) in order to determine a

posterior mean for reliability, R. The pos-

terior variance of R is then estimated using

a Taylor series expansion based on the pos-

terior variances of p; and p2 and their rela-

tionship to R. Using the posterior mean

and variance of R, we then determine the

“beta fit” for the posterior distribution of

R. This distribution is a beta distribution

for reliability R which may be used to de-

termine point and confidence interval es-

timates of R.

The first step in the approach just sum-

marized is to determine the posterior dis-

tribution on p; and p2. Unlike the classical

statistician who determines only point es-

timates on p; and po, the Bayesian is not

willing to restrict p; and p2 to single values.

The Bayesian prefers to express his degrees

of belief on p; and p2 by determining a dis-

tribution for each.” This approach makes

sense because if we conducted ten separate

trials at level Y, ten times, the ten resulting

144 MARK ANGELO JOHNSTONE

Pestimates would most likely differ and be

distributed in some fashion.

In order to begin our Bayesian develop-

ment of posterior distributions for p; and

p2, some definitions and assumptions must

be made. Testing at levels y; and y2 yields

two independent Bernoulli processes with

parameters p; and p> respectively. At level

‘7’? there are nj trials resulting in m; fail-

ures,i = 1, 2. Pjis the probability of failure

at stress level “1’’. The stress applied at

stress level one is less than the stress applied

at stress level two. Therefore, we may con-

clude that the condition p; is less than p>

always holds.

In the case of each Pi, we begin our analy-

sis by stating a prior distribution on each

P;. This prior distribution reflects our prior

knowledge on Pj before testing begins. This

prior knowledge may express “‘engineering

judgement’’*> determined through experi-

ence; it may be the result of the development

and analysis of other parameters; or it may

express our ignorance of Pj.

In order to determine the prior distribu-

tions on p; and p2 we consider p; and p2 as

unknown binomial parameters of a Ber-

noulli process and therefore random vari-

ables in the Bayesian sense. Because the

beta distribution is the conjugate’ to the bi-

nomial distribution, we will use a beta dis-

tribution to represent the prior distribution

for each parameter. The particular prior

distribution we are using is the uniform dis-

tribution which is a form of the beta distri-

bution with parameters A = 1 and B = 1.

Note that we use a uniform distribution to

represent our “‘prior’’ knowledge on p; and

P2. By using a uniform distribution we are

stating that P; has a equal chance of taking

on any value between zero and one.

Once we have determined our priors on

P;, we update our “‘knowledge”’ using the

likelihood on Pi, from which we determine

a posterior on Pj. The procedure for arriv-

ing at our posterior distribution for Pjis the

direct application of Bayes’ theorem. Bayes’

theorem states that the posterior distribu-

tion is proportional to the product of the

prior distribution and likelihood function

divided by the normalizing constant, where

the normalizing constant is the definite in-

tegral of the product of the prior and likeli-

hood function taken over the full range of

the random variable, P;. In equation form

Bayes’ theorem states the following:

f(p)- Lp)

/ f(pi): L(z|@)dé

f’(pi|z) = Eqn. 3.1

where, pi = the probability of failure

at level 1, for i = 1, 2.

= data (a, b), where “‘a”’ is

the number of failures

and “‘b”’ is the number

of successes

f’(pi|z) = the posterior distribu-

tion on pi, fori = 1, 2

L(z|pi) = the measured likelihood

function for each pi, for

i= 1,2

In order to determine our posterior dis-

tribution on pi, we utilize the following ap-

proach. Because quantal response testing is

defined as a Bernoulli process, the likeli-

hood function is defined as follows:

L(z|pi) = p2(1 — pi)” Eqn. 3.2

The prior density function is the conju-

gate function of p;. The conjugate function

can be defined as the normalized likelihood

function:

ty 4 L(@pilz)

“ (6 |z)d0

ig eae

i, 6°(1 — 6)°dé

which reduces to:

f(pi) = pi(1 — pi)’ T(a + b + 2)/

(a+ DO. +)» Eqns

This is a beta distribution with parameters

A =a-+t landB = b + 1 where we define

the probability distribution function of a

BAYESIAN ESTIMATION OF RELIABILITY 145

beta with parameters A and B as:

PUA! t BY. aa:

r(A)r(B) P

X(1-—p)*! 0<p<I

fa(p; A, B) =

By substituting equations 3.2 and 3.3 into

Bayes’ theorem (equation 3.1), we have the

following expression for our posterior dis-

tribution on pi.

f’(pi|z) =

pi” (1—pi)”"” Tao +a’ + bo +b’)

I'(ao + a’)I'(bo + b’)

or equivalently,

f’'(pi|z) = fe(pi; ao + a’, bo + b’)

The (0) subscripts indicate that the pa-

rameter is from the prior distribution and

the (’) subscripts indicate that the value

came from the sample where ‘“‘a’’ is the

number of failures and “‘b”’ is the number

of successes.

The posterior distribution on each pj is

the beta distribution, fg(a’ + 1, b’ + 1).

Note that the parameters ‘“‘a” and “‘b” of

our prior distribution both equal one due

to our assumption of a uniform prior dis-

tribution. Note also that because our prior

distribution and our posterior distribution

are both from the beta family, “fresh” data

from current testing may be used to update

our current knowledge by simply adding it

to our posterior.

At this point it is possible to apply a simu-

lation approach, such as that described by

Mikasa,’ to determine an approximate pos-

terior distribution on R. The beta distribu-

tions on p; and p> are as just described.

That is, pi ~ fg(pi; A = a1 + ao, B= bi +

bo), where ao and boare the values of the pa-

rameters of a uniform prior distribution

witha = land b= 1. Usinga Monte Carlo

simulation approach, a value for each p,

and each p> is sampled from its respective

beta distribution. Using these p; and p>

values, us and og values are calculated by

solving equations 2.1 and 2.2 simultane-

ously. Next an R value is calculated using

equation 2.3. This procedure is repeated a

large number of times (2500- 10,000). The

mean rand variance 6gforall generated R

values are calculated. jig and Gr are then

used to fit a beta distribution for the poste-

rior distribution of R by using the follow-

ing equations:’

a2 a

8 De

BSA arg

A= m

r-~ | Eqn 3.4

rie | ae

B= + ge 2° Ean. 3.5

While this simulation approach provides

an approximation to the posterior distribu-

tion of R, it could be very expensive in

terms of computer time. Its primary weak-

ness is that it is in fact a simulation approx-

imation. In such cases the questions arrise

of whether enough trials have been run,

whether the simulation model is accurate,

et cetera. It is desirable to have an ap-

proach which eliminates the need for simu-

lation, such as developed in this paper.

Continuing with the direct Bayesian

analysis, now that distributions on p; and

p2 have been determined, we must take into

account the condition that p; is less than p>.

Because p; and p> are initially considered

independent, we may define the joint prob-

ability density on p; and p> as follows:

f(pi, p2) = fa(p1):fa(p2) Eqn. 3.6

where f,(pi) is the posterior on p; with pa-

rameters a’ + 1, and b’ + 1. The condi-

tional distribution on p; and p> given that

p1 is less than p2is somewhat different. This

condition may be represented graphically

by considering equation 3.6 defined in the

shaded area in Figure 3. Mathematically

the joint conditional probability density of

pi: and p2 may be written as:

f(pi, P2)

Eqn. 3.7

Pr(pi < p2)

g(pi, p2|pi < p2) =

This function will be used to determine the

first and second moments of p; and p> given

that p; is less than po.

The expected value of the p; (the first

146

moment) given that p; is less than p2 may be

determined as follows:

E(pi|pi < pz)

= pig(Ppi, p2|pi < p2)dpidp2

By substitution of equation 3.7 this is equal

to:

E(pi|pi < p2)

“tpi psy" p2)

dpid Ean. 3.8

=| f Pig: 2a ee

Note that,

Pr(p: < p2) = | f(Pi, p2)dpidp2

Pi<p2

Eqn. 3.9

By substituting equations 3.7 and 3.9 into

equation 3.8, we have the following

relationship:

M11 = F(pi |p: < pz)

14 i y fe(pi)*fe(p2)dpidp»

re l lf p2

oe [I fe(p1)°fs(p2)dpidp2

0/70

The solution of this expected value reduces

to:

Ai + Ap

Ai + B;

=. (A, +B; —i)\(B:+ B2— 1—i)!

i=1 (Bi—i)!(A1 + A2+ Bi + B2—i)!

=. (B, + B2—1—i)"Ai+B;— 1—i)!

i=1 (Ai +B, + A.+ Bo— 1—i)(Bi—i)!

Eqn. 3.10

pair E(prlpr Spa) =

where, A; = the number of failures plus

one at test level one.

B, = the number of successes plus

one at test level one.

A2 = the number of failures plus

one at test level two.

B, = the number of successes plus

One at test level two.

MARK ANGELO JOHNSTONE

1 mA

Fig. 3. Relationship Between p; and po.

By similar development, the second mo-

ment of p; given that p; is less than p> is:

1 p2

w= Eile <p)= ff Pi

fe(p1)*fa(p2)dpidp2

7 foi fa(p1)* fa(p2)dpidp2

which is reducible to:

wor = E(pi|pi < pz)

_ (Ar + Aa + 1I(Ai + Ad)

(A; + B,)(Ai + B, + 1)

(B, + By— 1—i)(A, +B, + 1—i)!

i=1 (Bi —i)!(Ai1 + B, + A2+ Bo + 1—i)!

=. (Bi +B,—1—i)"Ai+B,—i-— 1)!

i=1 (Ai + B, + A2+ Bo— 1—i)\(Bi—i)!

Eqn. 3.11

The first and second moments of p2 are

calculated in a similar manner.

Miz = E(p2| pi < pz) = 4 oi, P2

fa(p1)*fe(p2)dpidp>2

oat ft fo(p1) *fo(p2)dpidp>

2 7 2

wn = Epil <p) = ff p

0 0

fp(p1)°f(p2)dpidp2

fa(pi)* fa(p2)dpidp2

BAYESIAN ESTIMATION OF RELIABILITY 147

Both of which reduce to:

Mi2 = E(p2| pi < p2) = (Ai + Az)

= (Bi + B2+ 1—i)"(Ai+B,:—1-—i)!

i=1 (A, +B,+A2+B2—i)!(Bi —i)!

=. (Bi + B.— 1—i)'(A,+ Bi—i— 1)!

i=1 (A, +B, + A2+ Bo — 1—i)(Bi —i)!

Egn. 3.12

22 = E(p2| pi < pz)

= (A; + Ao + 1)(Ai + A2)

S. (A; +B;— 1—i)\(B, + B.— 1—i)!

i=1 (Ai +B, +A2+B.+ 1—i)\(Bi—i)!

*. (B,+B)—1—i)'(A: + B;—1—i)!

i=1 (Ay +B, + A2+ B2— 1—i)!(B, —i)!

Egn. 3.13

Using the first moments of p; and p2, we

may now solve for estimates of the poste-

rior means, jis and 68, of us and 65 of the

strength distribution which is assumed to

be normal. We use the following equations:

[ n(x; fis, 63)dx = E(p: |p: < p2)

[ n(x; fis, 68)dx = E(p2| pi < pz)

We also know that these two relationships

may be rewritten as:

yi—ps/Gs

[ n(x; 0, 1) = E(pi| pi < p2)

and

y2—ps/6s

| n(x; 0, 1) = E(p:|p: < pz)

From which we arrive at the two equations:

Egn. 3.14

We know the test levels y; and y2 and we

may determine z; and 22 by looking them

up with their associated probabilities (ie,

E(pi|pi < p2) and E(p2|pi<pz2)) in a

normal probability table. We now have two

equations and two unknowns from which

fis, the mean of the strength distribution

and 6, the variance of the strength distri-

bution, may be calculated. We use these

values as estimates of the posterior means

of ws and os.

By comparing the previously calculated

strength distribution with the given stress

distribution, the posterior mean of the reli-

ability, R, is determined. The following re-

lationship is used:

fr = E(R) =/ n[x; ue — fis, of = G8)dx

fis—pe/VJort as 1 3

—w/2

SSeS dw

Loo \/ 27

Egn. 3.16

This yields a posterior mean value for R.

The following equation is a Taylor series

based on the posterior variances of p; and

p2, and the relationships provided by equa-

tions 2.1, 2.2, and 2.3. It approximates the

posterior variance of R.

2 2

Gk = varR p> (=) varpi

i=1 épi

6R 6R

+ 2 — — cov(pi,p2| pi < p2)

dpi dp2

Eqn. 3.17°

In order to use this approximation the fol-

lowing relationships were derived from

equations 2.1, 2.2, and 2.3:

u(us,0s) 1 :

R —_— eo Ue dx

20 V 2r

us = yi — osP(p:)

os = (y2 — yi)/(® '(p2) — &'(pi))

148 MARK ANGELO JOHNSTONE

®'(p.) 1

e€

he V 27

u(us, os) = (us — we)/Vo8 + of

The respective partial derivatives are calcu-

lated separately in the work that follows.

—x"/2

pes

Sig EEN 200

6pi O6us dpi das Spi

—_ oR = REE hss e 1/2llus—us)'/o8+ 08]

dus Ino + os)

7 a 6 2) a a

x | x 4 1

®'(p2) — ® (pi)

_ SR _ ~os(us — ue)(os + o8) 7

dos >

x @~ V2lus—ne)'/08+ oF)

dos (@\p))/2

Toe = ee Ire V

['@2) — © "wor

_ OR _ OR dus , BR bos

6p2 dus dp2 das dp

_ dus _ (2 — yi) V2m el? 7 & ip)

dp2 (®'(p2) — &'(p1))°

boa _ Oy) VR

dp2 (@'(p2) — & (pr)

In order to determine the variance of p; and

P2 the first and second moments of p; and

p2 as defined by equations 3.10 through

3.13 are utilized. The following two equa-

tions define the variance of p; and po.

a 2

var Pi — M2(pi) — M1(p1)

aoe D

Var Pp2 = 2(p2) — 1(p2)

The covariance term is calculated in the

work that follows.

coVv(p1,p2|pi < pz)

= E(p1,p2| pi < p2) — E(pi| pi < pz)

xX E(p2|pi < pz)

Both E(p: | pi < p2)and E(p2| p: < p2)have

already been defined by equations 3. 10and

3.12. E(pi,p2 |p: < pz) is defined as:

1 p2

re [ Pip28(Pi,p2| Pi < p2)dpidp2

0 0

and by substitution is equal to:

fa(pi)*fe(p2)dpidp2

] p>

if f DiD2.—-}-p5 =

is i fa(pi)°fe(p2)dpidp2

Upon integration of this equation we have

the following series:

E(p1,p2|pi < pz)

— (Ai + Az + 1)(A1 + Ad)

(A; + Bi)

2. (A, + B, +i)!(B:+ B2— 1-3)!

i=1 (Bi —i)(Ai + B, + A2+ B. + 1—i)!

=. (B,+ B.—1—i)'(Ai+ Bi— 1—i)!

i=1 (Bi —i)!(Ai+ B, + A2+ Bo — 1—i)!

By substituting these partial derivative,

variance and covariance equations into

equation 3.17, we may approximate the

posterior variance of R. We will fit the pos-

terior variance of R with a beta distribution.

Using jin and Gk the parameters of the

beta fit for the posterior reliability distribu-

tion are determined from:

a2 PI

pe eye

A = Eqn. 3.18

B= -al ieee +) Eqn. 3.19

This beta distribution can then be used

to provide point and interval estimates of

reliability R. For instance with respect to a

squared-error loss function we would point

estimate R by the posterior mean; that is we

would use R = jr. For other loss functions

we would use other characteristics of this

beta distribution. To compute the 100

(1 — a) % lower confidence limit for R, we

BAYESIAN ESTIMATION OF RELIABILITY 149

find the value of R, which satisfies the fol-

lowing equation:

f fs(R; A,B)dR =a_ Eqn. 3.20

In the next section an example is pro-

vided to clarify the approach discussed in

this section.

4. Example

Consider once again the example in

which the reliability for an anti-tank sabot

round fired against a Soviet T-62 tank is de-

fined as the probability that a given round

will penetrate its target. The ranges at

which the round will be fired in battle rep-

resent the stress distribution which is given

as a normal distribution with a mean of

1600 meters and a standard deviation of

100 meters. The strength distribution is a

distribution of ranges at which a given

round just penetrates the target. Reliability

is defined as the probability that the strength

distribution exceeds the in-use stress distri-

bution (see Figure 1.). Churchman two-

stimuli testing produces the following re-

sults. At a range of 1800.00 meters (y1) six

failures (penetrations) out of twenty rounds

tested are observed. At a range of 2000.00

meters (y2) eighteen failures out of twenty

rounds fired are observed.

One method of solving this problem is to

use the Monte Carlo simulation approach

developed in section III. Each pj is distrib-

uted as a beta distribution with parameters

A = aj + aoand B = bj + bo where ao and

bo are both equal to one due to the assump-

tion of a uniform prior distribution and a;

and b; represent the number of failures

(penetrations) at each “i” level. Using

Monte Carlo simulation a p; anda p2 value

is sampled from its respective beta distribu-

tion. Next equations 2.1 and 2.2 are solved

simultaneously to find the mean and var-

iance, fis and 68, of the strength distribu-

tion. Next an R value is calculated using

equation 2.3. Upon repeating this proce-

dure 3000 times, the mean and variance of

R for our given test results are .9349 and

.002693. Using equations 3.4 and 3.5 the

beta parameters for the reliability distribu-

tionare A = 20.189and B = 1.4051. Using

equation 3.20 the 95 percent lower confi-

dence limit is .83318. A program has been

written to perform these calculations.

The direct method of solving this prob-

lem as developed in section III follows. The

first and second moments of p; and p2 are

computed using equations 3.10 through

3.13. For the given test results the first and

second moments of p; are .3182 and .1107

and the first and second moments of p> are

.8636 and .7510 respectively. Next the cor-

responding z values of p; and p2 are found

to be —.4730 and 1.0966. These values are

used to solve equations 3.14 and 3.15 si-

multaneously for the mean and variance of

the strength distribution which upon calcu-

lating are 1860.26 meters and 16,236.11

meters. Equation 3.16 is then used to com-

pare the strength distribution with the

given stress distribution so that a posterior

mean on reliability, R, may be determined.

Using equations 3.16and 3.17 for the given

data, the estimated mean for Ris .9459 and

the approximate variance for Ris .00145044.

Equations 3.18 and 3.19 are then used to

determine beta parameters for the reliabil-

ity distribution. For the given data the beta

parameters for the reliability distribution

are A = 32.3978 and B = 1.8511. We de-

termine a 95 percent lower confidence limit

on R of .8718 by using equation 3.20. A

program has also been written to perform

all of the necessary calculations for this di-

rect approach to reliability in the stress-

strength context. Note that the parameter

estimates for both methods are similar.

Sample data with various stress and strength

means, various test levels, and various

sample sizes were analyzed using both the

simulation approach and the direct ap-

proach. The results using either method for

the many different cases examined were

very similar.

5. Conclusions and Recommendations

In the preceding sections of this article a

Bayesian statistical approach was presented

150 MARK ANGELO JOHNSTONE

for analysis of Churchman two-stimuli

quantal response data to estimate reliabil-

ity in a Stress-strength context. It was

shown how point and confidence interval

estimates of reliability could be obtained.

Classical analysis procedures provided con-

fidence limits which applied only to large

sample situations. The approach presented

herein can therefore be considered to fill a

void. That is, confidence limits on reliabil-

ity can be computed for even extremely

small samples.

By comparison of the direct calculation

approach presented herein with the Monte

Carlo simulation approach, it was found

that the results agreed very well. The direct

caclulation approach therefore eliminates

the need for simulation and its inherent

weaknesses. Study of the simulation results

also verified that a beta fit to the posterior

distribution of reliability was excellent.

Finally, it should be noted that the pos-

terior beta distribution for reliability can

be used as a “prior’’ distribution with re-

spect to future straight forward reliability

tests that is, success-failure testing. In this

way all available information is utilized in

arriving at an estimate of reliability. In

order to fully realize the advantages of this

approach, further research is recommended

in the following areas:

(i.) properties of the point and interval

estimates of reliability resulting from the

approach should be studied.

(ii.) consider application of an analo-

gous approach to data resulting from ex-

perimental designs other than Churchman

1.e., Langlie or Bruceton designs.

(ii1.) develop methods of using the relia-

bility parameters of the stress-strength con-

text so that they may be updated with ac-

tual reliability results of fielded systems.

Endnotes

'The Churchman testing procedure is outlined in

detailin R. J. Christie and L. D. Maxim, An Investiga-

tion of Test to Failure Methodology (Princeton, N.J.:

Mathematica, 1971), pp. III.16-21.

A discussion of the Bayesian definition of parame-

ters as random variables is found in Alfredo H-S Ang,

and Wilson H. Tang, Probability Concepts in Engi-

neering Planning and Design (New York: John Wiley

and Sons Inc., 1975), p. 329.

* By allowing for subjective inputs such as “‘engi-

neering judgement,” all available knowledge may be

incorporated into the prior distribution.

* A list of conjugate distributions is found in How-

ard Raiffa and Robert Schlaifer, Applied Statistical

Decision Theory (Clinton, Mass.: Colonial Press Inc.,

1961), p. 70.

° Raiffa and Schlaifer, p. 216.

°A Bayesian simulation approach is described in

Glenn Mikasa, “‘A Distributed Estimate for Reliabil-

ity in Churchman Test.”’ (Oahu, Hawaii: Quality Eval-

uation Laboratory, Naval Ammunition Depot, 1968).

’ Equations 3.4 and 3.5 were derived by taking the

mean and variance equations for the beta distribution

with limits zero and one and solving the two equations

simultaneously for A and B. The mean and variance

equations for the beta distribution are found in Ang

and Tang, p. 131.

* Anders Hald, Statistical Theory with Engineering

Applications (New York: John Wiley and Sons, 1952),

p. 118.

Journal of the Washington Academy of Sciences,

Volume 73, Number 4, Pages 151-152, December 1983.

Research Fellows of the Washington Consortium of

Universities at the U.S. Army Research Institute

Sherman Ross,’ Robert M. Sasmor,” and John P. Whalen?

Howard University, U.S. Army Research Institute and

Consortium of Universities

Since 1981 a program of mutal collabo-

ration through graduate student involve-

ment has been operating in the Washing-

ton, D.C. area. One element of this program

is the Consortium of Universities in the

Washington Metropolitan Area (The Ameri-

can University, The Catholic University of

America, Georgetown University, Howard

University, University of the District of Co-

lumbia, Gallaudet College, Mount Vernon

College and Trinity College). The other

element is the U.S. Army Research Insti-

tute for the Behavioral and Social Sciences.

The Consortium is a non profit corpora-

tion chartered in 1966 to encourage and

coordinate student exchange, there is joint

academic and resource planning. Students

at the member institution can make use of

the courses, faculty and libraries at other

institutions. The group involves nine insti-

tutions, 75,000 students, 15,000 courses,

almost 6000 faculty members, and com-

bined library resources of over 5 million vol-

umes, including special collections.

Project Director, Previous directors have been

E. M. McGinnies (The American University) for

1981-82, and C. E. Rice (The George Washington

University) for 1982-83.

* Director, Basic Research, U.S. Army Institute for

the Behavioral and Social Sciences.

> Executive Director, Consortium of Universities of

the Washington Metropolitan Area.

151

The Army Research Institute has been

operating since 1972 as the U.S. Army’s

agency for research and development in the

behavioral and social sciences. It has a

large staff and appropriate central and field

facilities for the civilian, psychologists,

statisticians, computer specialists, etc. as

well as military, administrative and sup-

port personnel. The broad function is con-

cerned with the person side. ARI scientists

are involved in people related problem

solving research and development on Army

personnel, as well as research on the future

issues and problems to improve manage-

ability.

Some 175 research psychologists are in-

volved in the ARI Headquarters in Alex-

andria, VA, and about 100 more are in the

Field Unit. The areas of research and de-

velopment include human factors, cognitive

and information processing, perception and

memory, educational psychology, compu-

ter assisted instruction, tests and measure-

ments, test development, performance eval-

uation, individual and group training,

leadership, and personnel and industrial

psychology. Three major laboratories are

operating: Manpower & Personnel Research,

Training Research and Systems Research.

The interaction between these organiza-

tions is through a contract (MDA 903-82-

C-0021) under which graduate students

from the universities participate in and

152 SHERMAN ROSS, ROBERT M. SASMOR AND JOHN P. WHALEN

contribute to selected research projects at

the Institute by technical and analytic sup-

port activities. The program has involved

mainly graduate students in psychology at

the master’s and doctoral levels. These stu-

dents have been designated as Consortium

Research Fellows. Nominations are reviewed

and selected by the Steering Committee

composed of representatives from the par-

ticipating departments. Then a match is

made with an ARI research staff member—

the mentor. The Research Fellow is required

to be in excellent standing in the depart-

mental program, to assure that the activity

will not delay academic progress, to omit

other assistantship appointments and de-

mands and tocarry some basic skills, expe-

riences and knowledge to the particular re-

search effort. The involvement is 20 hours/

week during the academic year, and can

become full time during the summer months.

After appointment, the activities are task

based under the direction of the mentor,

and have involved four types of tasks: (1)

literature review and analysis, (2) planning,

design and development of experimental

studies, (3) report analysis and writing, and

(4) data collection and data analysis. Some

areas of Research Fellow activities are in

automated speech technology, the effects

of training devices on learning, reenlist-

ment decision making, voice actuated com-

puter input, training of senior level leaders,

portable air defense weapons, etc. Policy

guidelines provide for a limit of three years

of the fellowship, and brief progress re-

ports are provided each month to the Proj-

ect Director. An evaluation from the men-

tor will be required after each year of

participation.

In the late summer of 1983, 17 graduate

students in psychology (10 women and 7

men) were in the program. They came from

the following universities: The George

Washington University (6), The American

University (5), Howard University (4), The

Catholic University of America (1), and

Georgetown University (1). There were 14

mentors from the ARI involved with these

students.

The participation of these students has

contributed materially to the ARI research

program. At the same time these Fellows

have received on-the-job research expe-

rience during their university studies as

well as excellent financial support. The

purposes of the program are being met for

the graduate students, for the universities,

and for the Army Research Institute, and

has yielded this progress report.

—

DELEGATES TO THE WASHINGTON ACADEMY OF SCIENCES,

REPRESENTING THE LOCAL AFFILIATED SOCIETIES

TO WUMMIISIOUONY vas cvs as dade ones scan eedstaencdvasvuvacaucsuewne James F. Goff

ar OCICLY Of WASHINGTON... ...0c cc cic cep nccccvacccscececacvessuces Ruth H. Landman

IRIE WWASTHMNGION 24.00 cnc s esc c cee ccvescseccecccuncveceseces William Ronald Heyer

ENE See a bisck ib dw Km ov -a-W KEM a bin CORR Oe Aen CRlee ce we Maen e ee Anne Brown

rr 10 WARTINOTON 5, caver scan tesesewvcnctasseecesdéeccdaeege arts Margaret Collins

IE kt Sk ard v aS Wain’ ve a 'b wb einasd vaen uO alae Ow we ko oe ae T. Dale Stewart

IE WW OSTMMNGTETY, sec cae ns bade cs ubecevsacdsuveeoeupedceuenwes James V. O’Connor

rr tne Pistrict Of Columbia ...........0csccnccssccnccccvccecsecs Charles E. Townsend

EE Em Ps | ae Ee Paul H. Oehser

VE SIITCCOR, oases ves saves cet e sekecebececubauws eva n¥babeu sews Conrad B. Link

SEO ooo ecw nie eu enlace ave te W celian Rie ede When We wee kee wenliws Boyd W. Post

EPRICE CLG on avo ied x's a eed ks VOM OWES Raw aid eed Gabe ema George Abraham

re eerrica! and Electronics Engineers ........00eccccccccccvsccsacceccnce George Abraham

ec ROCCHANCAl ENPINGELS . 6.66.6 ce ke cs dd ee vaveccdsewnwecctaveecean Michael Chi

eermear mecicty Of Washington ... 5.0.0.0... ccc cece ccc cwbececcscceces Robert S. Isenstein

tena NTICTOOIOIORY 2... ccc ce wee cc tee bbe deeecaenaevesna cues Lloyd G. Herman

tien NINIATY ENGINGETS.... 0.5. cee cence swe wewcetecuceteceseceunee H. P. Demuth

MIUCE IVIL ENGIICETS <5. 56 oh 2 anise a von econ a ale Silene ot ns ae eens cares Wallace J. Cohen

ewes experimental Biology and Medicine...............cccrcecccccccccecses Cyrus R. Creveling

eeeeeraon Of Dental Research ...........ccccce cence cu ncatusavedvcne William R. Cotton

eee institute Of Acronautics and Astronautics... ..3.. 5... 0c cece cece een ones Richard P. Hallion

EE PIES A ONEIOEY 5 ooo fa ais a lels oe daisive en wou ee nee os wie a Wielnbe e'e/diaeme es A. James Wagner

OTE ASIII SLOT 25.55 < aie tic a oo Sun diel o ang ww mie ainthe bled aeele sie clowns odes Jack R. Plimmer

EN ooo cc scis Aiavaco sini sale e's B'S wie Wm dp we OR DIOS ee ee Richard K. Cook

ENE Ef. eA we ga iets sb nw wee a code Sen x Pe SIO Mwide Sa weeee GR eee Dick Duffey

I SEDATE MET SSIS = 22 5 oo os was tala 1a, gm lak wx Wadd a, Mca ae e Ridin aa Won 0 asi OE A. D. Berneking

EIEN on So a oie 6 odin ue waa Ve nee abe oe Maes oa ween weret Edwin R. Fuller, Jr.

IE i cA ch eis Was eye.uly hae Ame eie wroaher ele aru. mremie eee nate Alayne A. Adams

ESTENE SE TIC: CUPELEY 20 ic oso 5 hse aie. a: cvew ae. occu Misia @ aka Gace atoll aonkeiin wat Deborah J. Warner

CE rma Cf EMYSICS 1 CACHEIS sb iis csc ae seo e'ne lee pee a alee hed wee Se ee Peggy A. Dixon

nmeery oF Plant Physiologists ...........cccechecseeccaceseveavesens Walter Shropshire, Jr.

eet iGHS MESEATEH COUNEH . 6.5. ccc ese oe cine enna eecsuees ele sue cess we John G. Honig

EE NIE ENC. oe oii cies w eselaliy face's a wie We aoe’ we ROR aoe sw Rat ..... Jewel B. Barlow

American Institute of Mining, Metallurgical

METER MET Fao ecg vi-4V'n, v ayaesv sist pin < WS ol nme Wa eats Ww oe Garrett R. Hyde

ESE IRIE Trice fol ale-a elas cake Mee a Ta eR Sie Kae sw SR a Benson J. Simon

na mneTA TE OW (AIMIEKIEA .(.s < sexu aiecls deeds sea cwiee etwas esute valet dwanec a Patrick Hayes

EE CRNINES rae Se So ea ie a slang Sem eee Reece civ wee « Miloslav Racheigl, Jr.

MET ED ESS ICRAN CE Tg Ss conc eS ain ovo ok bcs ms hates BS neem pee 0 Ceres Paws’s H. N. Reynolds

meee CANsE AIT RECHINCAl CITOUD 6 6. cisinies ns oe a's e's badGue eee sisleckaad sds 0eeminr Paul G. Campbell

EINECCEMNCDICAL SOCIETY cael scien okyaecs aac doe sales Gas exh umes vee Howard E. Waterworth

etGuedcial Sustemas RESCAFCH 1.5.16 ccccccncckeecdsecvndiecwschas’s Ronald W. Manderscheid

EO BE 0 sd On Les Sie vivid Bya(diwmie WA Wy 0c RA- in we igi nt wv 8 20S wom Stanley Deutsch

EI Fo ig fara g gdp 1000S ws Wray pied Vin ole nbd ros apile (wieinle ao tla Wine Irwin M. Alperin

eed AOL Science, Pechnology and Innovation ......6...0c0csccccccnncacoscescnae Ralph I. Cole

EIEN ce srg ig ge tele vs on viccide oe bin nd nied eo ers piles sare Ronald W. Manderscheid

Delegates continue in office until new selections are made by the representative societies.