Number 34

PROCEEDINGS

of the

San Diego Society of Natural History

Founded 1874

Notes on the Anterior Dentition and Skull of Proterixoides (Mammalia:

Insectivora: Dormaaliidae), and a New Dormaaliid Genus from

the Early Uintan (Middle Eocene) of Southern California

Stephen L. Walsh

Department of Paleontology, San Diego Natural History Museum, P. O. Box 1390, San Diego, California 92112

ABSTRACT.—The anterior dentition and the anterior part of a skull of the dormaaliid lipotyphlan Proterixoides davisi have been recovered from late

Uintan (middle Eocene) deposits in San Diego County, California. Proterixoides and other sespedectines are incipiently diprotodont, i. e., they have

moderately enlarged, closely appressed, chisel-shaped first and second lower incisors that probably functioned together with the similarly enlarged

corresponding upper incisors. Sespedectines are also characterized by having relatively small subequal lower third incisors and canines, relatively

small buttonlike third upper incisors, and relatively small double-rooted upper first canines. The distinctive antemolar morphology of Crypholestes,

Sespedectes, and Proterixoides corroborates the morphological integrity of the subfamily Sespedectinae.

Patriolestes novaceki is a new genus and species of dormaaliid from the early Uintan of San Diego County. It is characterized by a relatively large

second lower incisor and lower canine, a usually single-rooted and relatively small third lower premolar, a bulbous trigonid on the fourth lower

premolar not differentiated into separate cusps, the fourth lower premolar slightly longer anteroposteriorly than the first lower molar, lower molars with

erect anteroposteriorly compressed trigonids, first lower molars with trigonids averaging wider than talonids, a relatively large double-rooted upper

canine, single-rooted first and second upper premolars, deciduous and permanent third upper premolars much smaller than the fourth, and relatively

sectorial upper molars. Patriolestes can be excluded from the Sespedectinae but is not readily assignable to either the Dormaaliinae or Scenopaginae.

31 January 1998

Pending collection of more complete material of its presumed relatives, Patriolestes is best classified as Dormaaliidae, incertae sedis.

INTRODUCTION

Insectivores from the Eocene of southern California were first

described by Stock (1935). These early collections were obtained

from the middle member of the Sespe Formation of Ventura County,

representing the later part of the Uintan and the Duchesnean North

American Land Mammal “ages” (NALMAs). See Krishtalka et al.

(1987), Prothero (1996), Walsh (1996), and Lucas (1992) for discus-

sions of the Uintan and Duchesnean NALMAs, and Kelly (1990)

and Kelly et al. (1991) for a revised biostratigraphy of the Sespe

Formation. Stock (1935) named the erinaceomorph genera Proter-

ixoides and Sespedectes, and also tentatively referred a mandible

fragment to the Leptictidae. The Erinaceomorpha (hedgehogs,

gymnures, and their extinct relatives) and Soricomorpha (shrews,

moles, solenodons, and possibly golden moles and their extinct rela-

tives) together comprise the “true” insectivores, or Lipotyphla. See

Novacek (1986a), Butler (1988), and MacPhee and Novacek (1993)

for discussions of the names “Lipotyphla,” “Proteutheria,” and

“Insectivora.”

Novacek (1976) described several new lipotyphlan and “pro-

teutherian” taxa from early and late Uintan localities in San Diego

County. The most common early Uintan lipotyphlan from San Di-

ego is the dormaaliid erinaceomorph Crypholestes vaughni (see

Novacek 1976, 1980, 1985), which is morphologically similar to

the late Uintan and Duchesnean Sespedectes and Proterixoides but

retains a few more primitive dental features. A new, relatively large

” 6

species of Crypholestes and a small species of the scenopagine

dormaaliid Scenopagus were also recorded from the early Uintan

of San Diego by Walsh (1996, table 1), and their description is

planned for a future report. Crypholestes, Sespedectes, and Proterix-

oides are assigned to the dormaaliid subfamily Sespedectinae

(Novacek 1985). New specimens of Proterixoides davisi allow a

revised diagnosis of the Sespedectinae based on the complete ante-

rior dentition and a partial-skull. =

Novacek (1976) assigned Six isolated-teeth from the early Uintan

taxa and sueested that the untlamed genus was most closely related

to Scenopagus and Ankylodon, a conclisian.: #eiterated by Novacek et

al. (1985). The latter two genera were assigned to the dormaaliid

subfamily Scenopaginae by Novacek (1985). Since 1976, a few ad-

ditional specimens of the “erinaceoid genus and species” have been

recognized in UCMP and LACM collections from the Friars Forma-

tion (for abbreviations see Table 1). Several mandibular and maxil-

lary fragments and many isolated teeth have also been collected in

San Diego County by SDSNH personnel. The present material is

sufficient to justify the formal naming of the “erinaceoid genus and

species,” and a more detailed discussion of its phylogenetic position

is now possible.

The lithostratigraphic ranges of several dormaaliid taxa in

southwestern San Diego County are shown in Fig. 1. Novacek’s

(1976) specimens of the “erinaceoid genus and species” were col-

Ne

Lo

Miramar Sandstone Mbr.

Proterixoides davisi

Sespedectes singularis

@e Co

o_o

° ae lower

ovo

ODqVDoCoCo

Cc)

WLLL

STADIUM CONG.| «| P_O

Crypholestes

new sp

upper tongue

pholestes vaug

lower tongue

FRIARS

FORMATION

Patriolestes novaceki

Scenopagus cf. S. priscus

Figure 1. Known lithostratigraphic ranges in southwestern San Diego

County of several erinaceomorph taxa discussed in this paper. Dashed upper

parts of the ranges of Sespedectes singularis and Proterixoides davisi repre-

sent the inferred extent of their biochronozones (see Walsh, in press). Esti-

mated numerical ages of the Friars Formation and Poway Group are based on

information in Walsh (1996), Walsh et al. (1996), and an unpublished late

Duchesnean or early Chadronian mammal assemblage from the upper mem-

ber of the Pomerado Conglomerate (SDSNH Locs. 4041-4042). Ep-lc, lower

conglomerate member of the Pomerado Conglomerate.

Stephen L. Walsh

lected from the Rancho Penasquitos district in the northern part of

the city of San Diego, from strata mapped as the Mission Valley

Formation by Kennedy and Peterson (1975). However, these strata

actually pertain to the upper tongue of the Friars Formation, ac-

cording to the stratigraphic revision of Walsh et al. (1996). Walsh

(1996) discussed the profound faunal differences between the early

Uintan Friars Formation and late Uintan Mission Valley Forma-

tion, and the relationships of these assemblages to those from

Wilson’s (1972) members “B” and “C” of the Santiago Formation

in northwestern San Diego County.

METHODS

Important specimens are often introduced in the form “XX XX/

YYYYY,” where “XXXX” represents the locality number,

“YYYYY” the specimen number. Several Eocene local faunas

and geographic collecting districts of San Diego County are dis-

cussed by Golz and Lillegraven (1977), Walsh (1996), and Walsh

et al. (1996).

Teeth from I1/il to P2/p2 are referred to as antemolars, while

teeth from P3/p3 to M3/m3 are referred to as cheek teeth. Cheek-

tooth terminology largely follows that of Rich (1981). The first upper

and lower premolars of the dormaaliid taxa discussed herein are des-

ignated as “DP1” and “dp1,” following the observation that in eu-

therians these loci are usually not replaced by permanent teeth (But-

ler 1948:463; Ziegler 1971). Thus, for the purpose of assigning iso-

lated antemolars to particular loci, I assume that DP1 and dp1 are not

replaced in the taxa described below (no convincing evidence cur-

TABLE 1. Abbreviations used.

AMNH American Museum of Natural History, New York

AP anteroposterior length

AW anterior width of DP4 and P4-M3

Cc upper canine

c lower canine

CM Carnegie Museum of Natural History, Pittsburgh

CV coefficient of variation

D upper deciduous tooth

d lower deciduous tooth

I upper incisor

i lower incisor

L left

LACM(CIT) Original collections of the California Institute of

Technology, now housed at the Natural History

Museum of Los Angeles County

M upper molar

m lower molar

M arithmetic mean

N sample size

OR observed range of variation

P upper premolar

p lower premolar

PW posterior width of DP4 and P4-M3

R right

SD standard deviation

SDSNH San Diego Society of Natural History

UCMP University of California Museum of Paleontology,

Berkeley

USGS United States Geological Survey

WwW maximum transverse width of upper and lower teeth (for

p4 and teeth anterior to P3/p3 inclusive)

WTAL maximum width of talonid of dp4 and m1-3

WTRI maximum width of trigonid of dp4 and m1-3

YPM Yale Peabody Museum, New Haven, Connecticut

Dentition and Skull of Proterixoides and Patriolestes 3

Tl 12 igs} sp (el DP1 + DP2

W W ° =

AP AP AP AP

il i2

a“ WJ i3 + cl dpl + p2

wW wW

AP AP

AD AP

Figure 2. Generalized occlusal views of upper and lower left antemolars of

Proterixoides davisi showing orientations for measurements used in this paper.

rently suggests otherwise). For the sake of clarity in comparisons

within the Lipotyphla, I use the traditional four-premolar primitive

eutherian tooth formula in preference to the five-premolar terminol-

ogy advocated by McKenna (1975) and Schwartz and Krishtalka

(1976) and further discussed by Novacek (1986b).

Measurements of teeth are in millimeters (mm) and were made

on an Ehrenreich Photo-Optical Industries “Shopscope” to the near-

est 0.01 mm. Orientations and endpoints for the measurements of

antemolars are illustrated in Fig. 2. For 11-P2, I defined the AP axis

as a line that bisects the crown of the tooth into equal lateral and

medial halves as seen in occlusal view. The AP length of 11-P2 was

taken parallel to the AP axis and measured the maximum anteropos-

terior length of the crown as seen in occlusal view. The width of I1-

P2 was taken perpendicular to the AP axis and measured the maxi-

mum transverse width of the crown as seen in occlusal view. The

anteroposterior length of il-2 was measured parallel to the root, from

the base of the enamel on the lateral side of the tooth to the

anterodorsal tip of the crown, as seen in postero-occlusal view. This

dimension is inherently variable owing to occlusal wear, so only es-

sentially unworn i1-2s were included in the measurements for AP.

The width of il-2 measured the maximum transverse width of the

crown as seen in postero-occlusal view. For i3-p3, the AP plane was

defined to be the vertical plane that passes through the apex of the

main anterior cusp and bisects the the area of the crown into two

equal halves as seen in occlusal view. The AP length was then mea-

sured within this plane, but parallel to the maximum elongation of

the crown as seen in lateral view. The width of i3-p3 was taken per-

pendicular to the AP plane and measured the maximum width of the

crown as seen in occlusal view.

Measurements of DP3/P3-M3 and dp4/p4-m3 were taken ac-

cording to the instructions given by Novacek (1976:8-9), with the

following clarifications. The labial endpoint for the width of DP3

was always located at the labial base of the paracone even though the

metastylar lobe sometimes extended even farther labiad. The

anteroposterior lengths of P4s were taken as indicated by Novacek

(1976, fig. 2). The corresponding width measurements for P4 are not

perpendicular to the AP axis but are parallel to a line that passes

through the paracone apex and bisects the lingual lobe into equal

anterior and posterior halves. The anterior width of P4 was measured

along an axis that passes through the apex of the parastyle, while the

posterior width of M3 was measured along an axis that passes through

the apex of the metacone.

Two different terms have been used to designate a tooth that leans

anteriorly relative to the dorsal surface of the mandible. Stock (1935)

used the term “proclivous” to describe the attitude of i2 in

Proterixoides. I use it here to describe lower incisors whose crowns

are elongated essentially parallel to the root (i. e., the angle between

them is 175-180 degrees) and whose roots are implanted at a rela-

tively low angle into the end of the mandible. The term “procum-

bent” describes antemolars whose crowns are elongated anteriorly at

a distinct angle to the axis of the root as seen in lateral view (e.g., i3-

p2 of Proterixoides). Degrees of procumbency may be characterized

as follows: Slightly procumbent, root/crown angle 165-175 degrees.

Moderately procumbent, root/crown angle 150-165 degrees. Highly

procumbent, root/crown angle 120-150 degrees. “Trenchant” de-

scribes antemolars whose crowns are anteroposteriorly elongated,

with a sharp anteroposterior ridge or keel as seen in occlusal view.

“Bulbous” refers to a tooth crown (or part of a crown) that is convex

and broadly rounded, without any prominent cusps or significant

ridges. “Globular” designates upper anterior premolars whose

crowns are essentially spherical or ellipsoidal and situated atop a

single root of smaller diameter. Finally, “transverse” specifies upper

cheek teeth whose labiolingual widths are much greater than their

anteroposterior lengths. Higher-level taxonomy used in this paper

follows that of Novacek (1985).

SYSTEMATICS

Mammalia Linnaeus, 1758

Eutheria Gill, 1872

Insectivora Illiger, 1811

Lipotyphla Haeckel, 1866

Erinaceomorpha Gregory, 1910 (Saban, 1954)

Dormaalioidea Novacek, 1985

Dormaaliidae Quinet, 1964

Sespedectinae Novacek, 1985

Included Genera.--Sespedectes Stock, 1935; Proterixoides

Stock, 1935; Crypholestes (Novacek, 1976; 1980).

Known Distribution.—Early Uintan to Duchesnean of southern

California. Krishtalka and Stucky (1984) reported Crypholestes in

the early Bridgerian of Utah, while Storer (1996) reported Sespedec-

tes in the late Uintan of Saskatchewan.

Emended Diagnosis.—Dormaaliids with moderately enlarged,

proclivous, chisel-shaped i1 and 12, closely appressed to one another.

Lower i3 and cl single-rooted, moderately procumbent, virtually iden-

tical in size and shape, both much smaller than il-2. Lower dp1 and p2

single-rooted, with small, moderately procumbent, triangular, heart-

shaped crowns; p2 slightly larger than dp1, but both distinctly smaller

than cl. I1 moderately enlarged, with curved, transversely compressed

root and oval crown developing a concave lingual wear surface, result-

ing ina shovel-like appearance; I2 subequal in size to I1 but with more

bulbous crown and shorter root. Size and occlusal relationships of up-

per and lower first and second incisors showing an incipiently

diprotodont condition. I3 with single, short, straight, cylindrical root

and buttonlike crown much smaller than I1-2, identical in shape to that

of C1 but about 10% smaller in linear dimensions. C1 double-rooted or

with incompletely fused roots; crown small, buttonlike, anteropos-

teriorly elongate, only slightly longer than P2. DP1 and P2 single-

rooted, with small globular crowns, much smaller than P3. Skull poorly

known, but snout apparently relatively long and narrow; nasals

unfused; postorbital process absent; weak sagittal crest present; palate

not fenestrated; lacrimal foramen situated on anterior dorsal border of

orbit; anteroventral face of zygomatic root distinctly concave. Cheek-

tooth characters diagnostic of the Sespedectinae are discussed by

Novacek (1985).

4 Stephen L. Walsh

Discussion. —The revised diagnosis of the Sespedectinae is based

largely on the anterior dentitions of Proterixoides described below.

That this diagnosis is also applicable to Sespedectes and Crypholestes

is demonstrated by a nearly complete mandible of Sespedectes

singularis (SDSNH 3755/56171), numerous isolated antemolars of

S. singularis from SDSNH Loc. 3564, and numerous isolated

antemolars of Crypholestes vaughni from various SDSNH localities

in the Friars Formation. Once observed in place, the antemolars of

sespedectines are quite distinctive, and isolated teeth can usually be

allocated to specific loci with some confidence.

Proterixoides Stock, 1935

Proterixoides davisi Stock, 1935

Figs. 3, 4, 8c, 12b, 12d

Diagnosis.—Sespedectine much larger than Crypholestes or

Sespedectes, morphological differences between Proterixoides and

Sespedectes are minor; see Novacek (1985:15).

Known Distribution.—Late Uintan to Duchesnean (middle

Eocene) of southern California.

Stratigraphic Units, Localities, and Referred Specimens.—

Santiago Formation, Member C: Jeff’s Discovery local fauna

(SDSNH Locs. 3276 and 3560-3564): Mandible fragments with 12 +

p4: 42710. i3-dp1 + p3-4: 42804. dpl-m3: 48303. p3-4: 49696,

52800. p3-m1: 46627. p3-m2: 42636, 54942. p3-m3: 42709, 42802,

47458, 48309. dp4: 47181. p4-m1: 47682, 47683, 48980. p4-m3:

42639, 42806, 47728, 48401. m1-2: 52682. m1-3: 42805, 47461,

48135, 48728, 49661, 52205. m2-3: 46986, 48448, 48643, 52206.

Palatal fragment with LP2-3 + RC1: 47699. Maxillary fragments

with: P2-3: 42711. P2-M2: 42638. P3-M3: 47657. P4-M1: 48117.

P4-M2: 47180, 47457. M1-2: 54562. M2-3: 48139. 303 isolated

batch-catalogued antemolars and 180 isolated individually cata-

logued cheek teeth are also known from Jeff’s Discovery.

Rancho del Oro local fauna: SDSNH Loc. 3445/58944: Partial

skeleton with partial skull containing LI3-M3 and RI1-2 + P3-4 +

M2-3, associated mandibles containing Lp4-m3 and Rc1l-m3, plus

another mandible fragment from a different individual containing

Lel + p3-m3.

Stadium Conglomerate, upper member: SDSNH Loc. 3536:

12 isolated teeth.

Mission Valley Formation: SDSNH Loc. 3273: 13 isolated

cheek teeth, plus several isolated antemolars. SDSNH Loc. 3426:

two isolated cheek teeth. One isolated cheek tooth each from SDSNH

Locs. 3537, 3627, and 3742. SDSNH Loc. 3870: Mandible fragments

with p3-4: 54434. p3-m1: 54594, 54745. p3-m2: 54430. p3-m3:

54427-54429, 54596. m1-3: 54431: m2-3: 54433. Maxillary frag-

ments with: Cl + P3-M3: 54597. P3-M1: 54436. P3-M2: 54598,

54599. M1: 54435. Numerous additional uncurated jaw fragments

and isolated teeth are also present in the collections from SDSNH

Locs. 3870, 4020, and 4038.

Pomerado Conglomerate, lower member: SDSNH Loc. 3755/

56165: Mandible fragment with il-2 + p3-m3.

Discussion.—At the time of Novacek’s (1985) report on south-

ern California sespedectines, only three published specimens of

Proterixoides davisi were known from San Diego County, all from

UCMP V-72088 (Santiago Formation, Camp San Onofre local

fauna). Golz and Lillegraven (1977) had also recorded Proterixoides

sp. from the latest Uintan or earliest Duchesnean Laguna Riviera

local fauna (Santiago Formation, Member C). Since then, hundreds

of specimens of P. davisi have been collected from Member C, as

well as three different late Uintan stratigraphic units of the Poway

Group in southwestern San Diego County. See Novacek (1985) for a

detailed description of the cheek teeth of this genus. The following

descriptions address only the mandible, skull, and antemolar denti-

tion of Proterixoides. F. S. Szalay and I are currently studying iso-

lated tarsal bones referable to Proterixoides.

Mandible.—As seen in lateral view (Fig. 3B), the mandible of

Proterixoides is similar in general proportions to that of Sespedectes

(see Novacek 1985, fig. 2) and Macrocranion (see Novacek et al.

1985, fig. 1). The mandible is deepest below m2-3, and unlike the

condition in brachyericines (Rich, 1981) and modern erinaceids

(Frost et al., 1991), there is no distinct decrease in depth immediately

below the posterior end of m3. On the basis of 16 measurable speci-

mens of Proterixoides davisi from the Jeff’s Discovery local fauna,

the depth of the mandible below m2 ranges from 4.7 to 7.2 mm,

averaging 5.8 mm. Specimens preserving an intact ventral margin of

the anterior part of the mandible show a strong medially concave

ridge extending from the posteroventral corner of the symphysis to a

level below the posterior end of p4 (e.g., SDSNH 42636, 47458,

49696, 54942). This ridge presumably marks the origin of the

genioglossus muscle and is also present in at least some modern

erinaceids (e.g., Erinaceus, Hemiechinus). The distribution of this

character in other Paleogene erinaceomorphs is uncertain. There are

usually three mental foramina, the most anterior of which occurs

below p2, the most posterior below the middle of p4. As in

Sespedectes (see Novacek, 1985, fig. 2), the posterior margin of the

symphysis ends below p1, the dorsal ridge of the masseteric fossa is

strong, and the coronoid process is high and anteroposteriorly broad

(e.g., SDSNH 58944, mandibles not figured).

Lower antemolar dentition. —The complete anterior dentition of

Proterixoides davisi is now known from several mandibles and a

partial skull. On the basis of these specimens, many isolated

antemolars of P. davisi have been identified in the large screenwashed

sample of microvertebrates from the Jeff’s Discovery local fauna.

Prior to the collection of this new material, the only antemolar of P.

davisi known was the i2 in LACM(CIT) 1676 (Stock 1935, fig. 4).

The il of Proterixoides davisi is proclivous, with a long, robust,

slightly curved, gradually tapering, transversely compressed root

(SDSNH 56165, Figs. 3A-B). The crown is chisel-shaped, trans-

versely narrow, anteroposteriorly elongate, and dorsoventrally com-

pressed at the tip. As seen in anteroventral view, the crown flares out

laterally from the root, again giving the appearance of a chisel. As

seen in posterodorsal view, the medial tip of the crown is taller than

the lateral tip. The anteroventral face of the crown is slightly convex.

The posterodorsal face of the crown is also convex, but it bears a

large, blunt, central ridge that extends from the base of the crown

nearly to the tip. Occlusal wear facets form at the tip of the crown and

along the posterodorsal ridge. No interdental wear facets have been

observed on the medial side of any i1, suggesting that left and right

ils of Proterixoides were not closely appressed to one another. Six-

teen isolated ils are known from SDSNH Loc. 3564 (SDSNH

58945), and 13 deciduous ils also seem to be present (SDSNH

58947). The latter have crowns smaller than the permanent 11, with

thin root walls that are usually broken just below the base of the

crown.

The lower i2s of Proterixoides are similar in size and morphol-

ogy to the ils but are oriented in the mandible in such a way that their

roots are dorsoventrally compressed (Fig. 3). The crown of i2 has a

more rounded tip and a stronger dorsal ridge than il, and also has the

appearance of being given a distinct counterclockwise helical twist

relative to the flat crown of il. Six isolated i2s are known from

SDSNH Loc. 3564 (SDSNH 58946), and 26 deciduous i2s also seem

to be present (SDSNH 58948). Like the putative dils, the di2s have

crowns smaller than the permanent i2, with thin root walls that are

usually broken just below the base of the crown. As seen in occlusal

view, the lateral side of the crown of il is eclipsed by the medial side

of the crown of i2 (Fig. 3A). A distinct interdental wear facet is al-

most always present on the lateral side of the posterodorsal face of

the crown of il and the ventral face of the tip of i2, indicating that

these two teeth were closely appressed.

The i3 and cl of Proterixoides are preserved in place in SDSNH

42804 (Figs. 3C-D). The crowns of both teeth are much smaller than

Dentition and Skull of Proterixoides and Patriolestes 5

Figure 3. Stereophotographs of mandibular specimens of Proterixoides

davisi. (A), Occlusal, and (B), lateral views of SDSNH 56165, L mandible

with il-2 + p3-m3. (C), Occlusal, and (D), lateral views of SDSNH 42804, R

mandible fragment with i3-dp1 + p3-4. Scale bars, 5 mm.

il-2 and are virtually identical in size and morphology. Both are mod-

erately procumbent, nontrenchant, and have the appearance of an

anteroposteriorly elongate rounded triangle in occlusal view. The

crown of i3 is anteroposteriorly slightly more elongate than cl but is

dorsoventrally slightly thinner. Both teeth are single-rooted, with the

root of 13 being smaller in diameter than that of cl (see also the emtpy

alveoli for these teeth in SDSNH 56165; Fig. 3A). Twenty-eight iso-

lated i3s and 25 isolated cls are also known from SDSNH Loc. 3564

(SDSNH 58949 and 58950), but no di3s or dels have been recognized.

Stock (1935) assumed that Proterixoides had a double-rooted p2

and had lost (d)p1. As in Sespedectes and Crypholestes, however, P.

davisi retains single-rooted dp1 and p2 (Novacek 1985). The dp1 is

preserved in SDSNH 42804 (Fig. 3C-D), and dp1-p2 are both pre-

served in SDSNH 48303 (not figured). Twenty-seven isolated dp1s

and 19 isolated p2s are also known from SDSNH Loc. 3564 (SDSNH

58951 and 58952, respectively). Both dpl and p2 are distinctly

smaller than i3 and cl but have similar triangular occlusal outlines

and moderately procumbent crowns with weak dorsal ridges. The

crown and root of p2 are slightly larger than those of dp1. Eight

isolated teeth from SDSNH Loc. 3564 are tentatively identified as

dp2s of P. davisi (SDSNH 58953). They are slightly smaller and dor-

soventrally thinner than the p2 but are relatively more elongated

anteroposteriorly, have stronger dorsal ridges, and have a deeper in-

dentation in the occlusal outline of the posterior end of the tooth.

Skull.- The partial skeleton of Proterixoides davisi represented

by SDSNH 58944 includes the anterior part of a laterally compressed

skull. Both premaxillaries were found “floating” in the matrix and

have been reattached to the skull in their approximately correct ana-

tomical position. On the basis of substantial wear on the permanent

premolars, the individual was a mature adult. At least in part for this

reason, few cranial sutures are detectable.

In dorsal view (Fig. 4A), even allowing for lateral distortion and

fragmentary preservation, the snout seems to be relatively long and

narrow and thus more similar in this respect to extant hylomyine

erinaceids than to erinaceines (Frost et al. 1991). The main skull frag-

ment preserves the transversely thin left nasal, most of which has

broken away from the frontal and been tilted upward so as to appear

as if standing vertically on its lateral edge. The posterior parts of the

unfused left and right nasals are deeply embayed into the anterior

part of the frontal. The rest of the right nasal, most of the right max-

illary, and all of the right lacrimal have been broken away. The ante-

rior portion of the frontal has been pushed ventrally downward, thus

tearing itself away from the left nasal. The metopic suture is barely

discernible. The temporal ridges on the frontal converge posteriorly

at the same anteroposterior level as the constriction. There is no indi-

cation of a postorbital process. A very weak sagittal crest is present at

the posterior end of the skull fragment. Part of this crest presumably

extends to the frontal, but whether any of it is formed by the parietals

cannot be determined, as the suture between these bones cannot be

identified. As in Ankylodon (see Fox 1983, fig. 1), a weak ridge ex-

tends ventrolaterad from the lateral edge of the frontal, just anterior

to the constriction. This ridge quickly swings anteriad to merge with

the dorsal rim of the orbit, but unlike Ankylodon, does not develop

into a strong antorbital crest. As in Brachyerix (see Rich 1981, fig.

8), a knob of bone (lacrimal tubercle) on the most anterior part of the

dorsal rim of the orbit marks the location of the lacrimal foramen.

In left lateral view (Fig. 4B), the skull appears to be relatively

compressed dorsoventrally, more similar to modern hylomyines than

to erinaceines (Frost et al. 1991). The left premaxilla is nearly com-

plete. The ventral margin of the premaxilla is relatively long

anteroposteriorly compared to its dorsoventral height, and the

posterodorsal process appears to have extended posteriad almost to

the level of the infraorbital foramen. Proterixoides is more similar in

these respects to Ankylodon (see Fox 1983, fig. 1) and Brachyerix

(see Rich 1981, fig. 8) than to Macrocranion (see Maier 1979, fig.

6A). The anterior opening for the infraorbital canal occurs immedi-

ately dorsal to P3, as in Sespedectes (see Novacek 1985, fig. 4A),

Ankylodon (see Fox 1983, fig. 1), Brachyerix (see Rich 1981, fig. 8),

and Macrocranion (see Novacek et al. 1985, fig. 1). The dorsoven-

tral and transverse diameters of the infraorbital foramen in SDSNH

58944 are 2.2 and 1.1 mm, respectively. The minimum length of the

infraorbital canal is 5.1 mm.

The maxillary-lacrimal, maxillary-frontal, and frontal-parietal

sutures cannot be identified on the snout or orbital wall. As in various

lipotyphlan taxa, there are-two different dorsoventral levels of the

palate as seen in lateral view. In Proterixoides, the transition between

them is gradual and occurs largely between P2 and P3. This differs

somewhat from the condition in Ankylodon, in which the transition

apparently occurs between P3 and P4 (see Fox 1983, fig. 1). Short

lengths of the anterior part of both zygomatic arches are preserved in

SDSNH 58944. Although a jugal-maxillary suture cannot be dis-

cerned on either arch, it is likely that the region where these sutures

occurred is not preserved (compare with Brachyerix;, Rich and Rich

1971, fig. 1). The anterolateral face of the zygomatic process is

strongly concave, forming a distinct antorbital fossa for the origin of

Stephen L. Walsh

Figure 4. Stereophotographs of SDSNH 58944, partial skull of Proterixoides davisi preserving LI3-M3 and RI1=2 + P3-4 + M2-

3. Scale bar, 5 mm. (A), Dorsal view; (B), left lateral view (apparent concavity of skull roof is due to crushing); (C), ventral view.

Dentition and Skull of Proterixoides and Patriolestes 7

the snout muscles (Novacek 1986a:30-31). As measured behind M2,

the dorsoventral thickness of the left zygomatic process is 3.8 mm,

and its dorsal edge is sharp.

The left orbital wall reveals several structures, although their

identities are generally uncertain because the sutural boundaries be-

tween the frontal, parietal, orbitosphenoid, alisphenoid, and palatine

cannot be located. In the anterior part of the orbit, the dorsal surface

of the maxillary comprises most of the floor and shows the dorsal

ends of the lingual roots of M2 and M3. The posterior end of the

infraorbital canal opens immediately above the floor of the most an-

terior part of the orbit. No postpalatine foramen is evident, but a

small foramen seems to be present on the posterior edge of the max-

illary behind LM3. Immediately dorsomedial to the posterior end of

the infraorbital canal is a small pit at the anterior end of a shallow

groove in the orbital wall. This pit is located in the same place rela-

tive to the infraorbital canal as the pit for the inferior oblique muscles

identified by Novacek (1986a, fig. 9) in the orbit of Leptictis. An-

other, larger pit occurs at the ventral base of the lacrimal foramen. It

also occurs at the anterior end of a groove in the orbital wall. Three

additional, parallel grooves are present in the ventral part of the or-

bital wall. The most ventral of these lies immediately dorsal to the

dorsal surface of the palatine and runs from the lingual root of M3

back to the posterior end of the skull fragment. The middle of these

grooves ends anteriorly in a small pit or foramen (orbitonasal fora-

men?) and ends posteriorly below what is presumably the sphenorbi-

tal and/or suboptic foramen (compare with the orbit of Brachyerix;

Rich and Rich 1971, fig. 15). Another distinct pit immediately antero-

dorsal to the latter foramen may represent the optic foramen. Imme-

diately posterior to the sphenorbital (?) and/or suboptic (?) foramen,

a small bridge of bone may define the anterior and posterior open-

ings of a very short alisphenoid canal. The broken outline of what

may be the anterior rim of the foramen ovale is doubtfully present at

the posterior end of the skull fragment.

In ventral view (Fig. 4C), the palate has been pulled upward

relative to the crowns of LC1-P2, and the left and right cheek teeth

are compressed quite close together. The left maxillary preserves

C1-M3, while the right maxillary preserves alveoli for C1-P2, in-

tact P3, the lingual half of P4, and intact M2-3. The width of the

snout narrows appreciably but gradually from P4 to Cl. The max-

illary root of the zygomatic arch originates lateral to M2, as in

Macrocranion (see Maier 1979), Ankylodon (see Fox 1983), and

many other lipotyphlans. At least part of both palatines contribute

to the hard palate, but the maxillary—palatine sutures cannot be iden-

tified. There is no obvious postpalatine torus, and there does not

appear to be any fenestration of the palate as in Ankylodon (see Fox

1983, fig. 2) or Brachyerix (see Rich 1981, fig. 10). The skull is

preserved back to the posterior part of both palatines, but the

basicranium was broken away prior to burial, and there is no indi-

cation of the palatine-alisphenoid sutures. The internal narial open-

ing is present but distorted.

Upper antemolar dentition. —Combining teeth from the left and

right sides of SDSNH 58944 completes the upper dentition of

Proterixoides (Fig. 4C). 11-2 are present in a fragment of the right

premaxilla and are the largest upper anterior teeth (just as il-2 are

the largest lower anterior teeth). Il has a dorsoventrally elongated

crown, with a relatively flat lingual face and an asymmetrically con-

vex labial face. The flat lingual face becomes concave with wear,

assuming a characteristic shovel-like appearance. The wear facet is

presumably caused by occlusion with the chisel-shaped tip of il. The

large transversely compressed root of I1 curves gently posteriad into

the premaxilla. There is a short diastema of about 1 mm between I1

and 2. I2 is subequal in size to I1 but has a transversely wider, less

dorsoventrally elongate crown whose labial and lingual faces are both

uniformly convex. I2 develops a relatively flat wear facet on the tip

of the crown. The root is robust as in Il but is shorter and curves more

strongly into the premaxilla. Twelve isolated I1s and 21 isolated I2s

of Proterixoides are known from SDSNH Loc. 3564 (SDSNH 58954

and 58955, respectively). In addition, 43 isolated mitten-shaped teeth

with thin root walls probably represent deciduous I1-2s (SDSNH

58958).

The occlusal relationships of the lower and upper first and sec-

ond incisors in Proterixoides and other sespedectines were probably

similar to the relationships documented by Butler (1980:173-175) in

extant tupalids. As discussed by Butler (1980:173) in reference to

Ptilocercus, the arrangement of the anterior incisors in sespedectines

may best be termed “incipiently diprotodont.”

The left premaxilla of SDSNH 58944 is essentially complete. It

contains a crushed alveolus for I1, an intact but empty alveolus for

12, and a moderately worn complete I3 (Fig. 4B-C). The crown of I3

is much smaller than those of I1-2 and differs morphologically from

the anterior incisors in that it is short and buttonlike, very similar to

the Cl preserved in the same specimen but about 10% smaller in

linear dimensions. Twelve isolated I3s from SDSNH Loc. 3564

(SDSNH 58956) show that the crown apex is located near the ante-

rior end of the tooth and sends a weak dorsal ridge to the posterior

base of the crown. The anterior part of the crown is wider than the

posterior part and develops a flat posterodorsally dipping wear facet.

The cylindrical root of I3 is short, straight, and rapidly tapering.

C1 is preserved in place in the partial skull of SDSNH 58944

(Fig. 4C). It has a small buttonlike crown that is anteroposteriorly

slightly longer but distinctly narrower transversely than P2. As in

13, the anterior part of the crown of C1 is taller and wider than the

posterior part. The Cl in SDSNH 58944 misleadingly appears to

have a single anteroposteriorly elongate root. However, 3 of 12

isolated Cls of Proterixoides from SDSNH Loc. 3564 have sepa-

rate anterior and posterior roots (SDSNH 58957). In the remaining

nine Cls, the roots are incompletely fused into a single root, bear-

ing distinct longitudinal constrictions on their labial and lingual

faces. Novacek (1985) described the Cl of Sespedectes stocki as

being single-rooted. However, many isolated Cls of Sespedectes

TABLE 2. Measurements of antemolars in SDSNH 58944, partial skull and associated mandibles of Proterixoides davisi.

Il 12 13 Cl DPI P2

AP W AP WwW WwW AP Ww AP WwW AP WwW

L premaxilla + maxilla — _— = — 1.31 0.97 1.66 1.10 0.98 1.06 1.36 1.37

R premaxilla 1.88 1.34 2.03 1.56 — — — — _ — =

il 12 i3 cl dpl p2

AP WwW AP WwW W AP W AP WwW AP WwW

L mandible 2.34 1.69

R mandible — 2.36 1.67 — — 1.62 1.56

8 Stephen L. Walsh

singularis from SDSNH Loc. 3564 are double-rooted (SDSNH TABLE 3. Standard statistics for teeth of Proterixoides

60036), as are many isolated Cls of Crypholestes vaughni from the davisi from the late Uintan Jeff’s Discovery local fauna,

Friars Formation. Member C of Santiago Formation, San Diego County.

Heavily worn DP1s and P2s are preserved in place in SDSNH

58944 and are much smaller than P3. Nineteen isolated DP1s and 16 Tooth N OR M SD CV

isolated P2s from SDSNH Loc. 3564 (SDSNH 58959 and 58960)

show that both teeth are single-rooted, with globular crowns consist- Il AF 6 1.59-1.79 1.67 0.078 47

5 = ; B : A 9g WwW 6 1.49-1.64 1.58 0.056 3.5

ing of a single apical cusp and a weak posterodorsal ridge. P2 is p AP 1 194-233 212 0.159 15

distinctly larger than DP 1 and has a short postcingulum not present WwW rl 1.46-1.80 161 0.112 70

in DP1. One of the P2s included in SDSNH 58960 may be decidu- B AP 10 1.23-1.44 1.34 0.082 6.1

ous, as it has thin root walls and lacks a postcingulum. The upper Ww 10 0.94-1.08 1.00 0.042 42

anterior premolars of Proterixoides are morphologically very similar Cl AP 8 1.40-1.64 1.55 0.090 5.8

to the much smaller homologous teeth of Sespedectes (see Novacek W 8 1.07-1.26 1.12 0.065 5.8

1985, fig. 3). DP1 AP 16 0.94-1.15 1.08 0.057 5.3

Measurements. —Measurements of antemolars of P. davisi pre- 16 0.93-1.09 0.99 0.055 5.5

served in the associated mandibles and partial skull of SDSNH 58944 mchpes ie ee fe Ree oe

are given in Table 2, while standard statistics for measurements of P3. AP 10 7 87-3. 46 3.10 0.221 72

antemolars and cheek teeth of P davisi from the Jeff’s Discovery WwW ll 265-331 2.96 0.207 70

local fauna are given in Table 3. The latter provides a good indication DP3 AP 4 2.91-3.08 2.98 0.079 26

of the expected size variation in a dormaaliid species based on a large WwW 4 2.31-2.35 2.33 0.023 1.0

sample collected from a single site. P4 AP 12 2.86-3.31 3.09 0.109 35)

Novacek (1985, table 5) presented standard statistics for mea- AW 14 2.71-3.59 3.07 0.242 7.9

surements of cheek teeth of P. davisi from the Sespe Formation. It PW 12 = 3.35-3.95 3.68 = 0.184 5.0

DP4 AP 5 2.98-3.25 3.10 0.127 4.1

AW 5 2.65-3.27 2.93 0.266 9.1

PW 5 2.99-3.51 3.28 0.225 6.7

Mi AP 18 2.80-3.22 2.98 0.119 4.0

will be seen from Table 3 that linear dimensions of the Jeff’s Discov-

ery teeth average between 5% and 15% smaller than the Sespe teeth.

However, the latter sample is stratigraphically and temporally het-

erogeneous, and some tooth positions are represented by just a few AW 19 3.69-4.14 3.89 0.132 3.4

specimens. An analysis of whether more than one species of PW 19 3.92-4.55 4.20 0.189 45

Proterixoides is recognizable in the middle Eocene deposits of south- M2. AP 15 2.54-2.89 2.71 0.109 4.0

ern California is beyond the scope of this paper but certainly feasible. AW 15 3.92-4.61 4.21 0.201 4.8

The phylogenetic position of the Sespedectinae is discussed below PW 15 3.46-4.06 3.77 0.187 5.0

under Evolutionary Relationships. M3 AP 17 —_—‘1.68-2.14 1.92 0.123 64

AW 15 2.86-3.51 3.08 0.196 6.4

PW 18 2.09-2.66 2.27 0.142 6.2

Dormaaliidae, incertae sedis il AP 3 2.70-3.02 2.87 0.160 NG

Patriolestes gen. nov. Ww «| OG LOIS == AS OSS AD

: es DB AD § 3.25-3.64 343 0.158 46

Etymology.—Named for the Friars Formation, in reference to the WwW 5 1.63-1.88 1.75 0.105 6.0

Franciscan friars (Kennedy and Moore 1971:717). Patrio-, father, i3 AP 19 1.84-2.32 2.07 0.110 53

-lestes, robber (Greek). WwW 19 1.32-1.65 1.46 0.090 6.2

Type Species.—Patriolestes novaceki, sp. nov. cl AP 15 2.04-2.24 2.15 0.055 2.5

Diagnosis.—Patriolestes differs from all known Eocene erinaceo- W 20 —1.50-1.86 1.64 0.092 5.6

dpl AP 10 1.33-1.49 1.39 0.053 3.8

WwW 13 1.14-1.33 1.23 0.053 4.3

p2 AP 13 1.47-1.94 1.65 0.151 9.2

WwW 17 1.35-1.60 1.48 0.076 5.1

morph genera in having the following combination of characters (puta-

tive autapomorphies within Dormaaliidae indicated by boldface). No

diastemata in lower anterior dentition; il proclivous, chisel-shaped,

apparently somewhat smaller than i2; i2 proclivous, with anteropos- dp2 AP 7 1.52-1.82 1.70 0.101 59

teriorly elongate, chisel-shaped crown, distinctly larger than i3; 13 mod- Ww 7 1.17-1.22 1.20 0.017 1.4

erately procumbent and trenchant; cl single-rooted, slightly procum- p3 AP 19 2.70-3.07 2.87 0.097 3.4

bent, bulbous, with largest crown of the lower anterior teeth; dp1 single- WwW 19 1.78-2.22 1.94 0.107 5.5

rooted, with moderately procumbent triangular crown, smaller than p2- dp3 AP 12 2.36-2.59 2.50 0.074 3.0

3; p2 single-rooted with moderately procumbent triangular bulbous Ww 14 1.50-1.78 1.63 0.079 4.8

p4 AP 29 2.93-3.75 3.36 0.154 4.6

WwW 29 1.97-2.61 2.24 0.131 5.9

dp4 AP 11 2.99-3.34 3.12 0.125 4.0

WTRI 10 1.49-1.77 1.63 0.082 5.0

crown subequal to that of p3; p3 usually single-rooted, with slightly

procumbent, bulbous crown, somewhat pentagonal in occlusal outline,

and much smaller than p4; p4 slightly longer than m1, double-rooted,

bulbous, without differentiation of trigonid cusps or paracristid, WTAL 11 1.83-2.13 1.98 0.103 52

and with single small cusp at posterolingual corner of short talonid; ml AP 26 2.90-3.34 3.10 0.120 3.9

lower molars with erect anteroposteriorly compressed trigonids and WTRI 25 2.30-2.71 2.44 0.096 3.9

bladelike paraconids; lower and upper molars relatively sectorial; m1 WTAL 25 2.31-2.72 2.51 0.105 4.2

trigonid width averaging slightly greater than talonid width; [1-2 m2 AP 37 2.71-3.13 2.92 0.107 3.7

moderately enlarged, with shovel-like crowns. 13 much smaller than WTRI 38 =. 2.36-2.80 2.57 0.09135

WTAL 37 2.27-2.60 2.41 0.078 7)

m3 AP 28 2.53-3.22 2.94 0.167 Sh7/

WTRI 28 1.75-2.31 2.03 0.128 6.3

WTAL 28 1.51-1.89 1.70. 0.106 6.3

I1-2, with dorsoventrally short, anteroposteriorly elongate crown with

distinct dorsal ridge. C1 double-rooted, with large, bulbous crown,

DP1-P2 single-rooted, with small, globular, slightly anteroposteriorly

elongate crowns with weak dorsal ridges; DP3 triple-rooted, with very

strong metastylar lobe, distinct metacone, but no hypocone; DP3 much

smaller than the triple-rooted, large, molariform DP4; P3 subtriangular

in occlusal outline, without metacone or hypocone, and much smaller

than P4; P4 triple-rooted, large, transverse, with strong posterolabially

Dentition and Skull of Proterixoides and Patriolestes 9

directed metastylar crest but without hypocone or metacone; M1 with

hypocone extending further linguad than protocone. M2 transverse,

usually with protocone extending further linguad than hypocone. M1-

3 generally with strong pre- and post-paraconular and metaconular

wings. See below for detailed character-by-character comparisons of

Patriolestes with other erinaceomorph taxa.

Patriolestes novaceki sp. nov.

Figs. 5-7, 8a, 8b, 9-11

“Erinaceoid-like genus and species” (Novacek 1976:36).

“cf. Proterixoides davisi” (Golz and Lillegraven 1977:53).

“cf. Proterixoides sp.” (Golz and Lillegraven 1977:53).

“Erinaceid-like genus and species” (Novacek et al. 1985:5).

Proterixoides davisi (Novacek 1985:15, in part).

“Dormaaltidae, new genus and species” (Walsh 1996:84).

Etymology.—Species named after Dr. Michael J. Novacek, for

his contributions to vertebrate paleontology.

Type Specimen.—SDSNH 49250, right mandible fragment with

p4-m3.

Type Locality. —SDSNH Loc. 3784, “Stonecrest Square Site 4.”

Friars Formation (undifferentiated), west wall of Murphy Canyon,

San Diego County, California.

Diagnosis.—Morphology as for genus; size as indicated by mea-

surements in Tables 4-6.

Known Distribution.—Early Uintan of San Diego County, Cali-

fornia (Friars Formation and Member B of Santiago Formation).

Localities and Referred Specimens.—See Appendices 1-2.

Mandible.—There are 13 mandible fragments (none preserving

the coronoid or condyle) containing at least one tooth in LACM,

SDSNH, and UCMP collections. The most dentally complete of these

is the type specimen (Fig. 5), but other mandibles preserve more

complete anterior alveoli (Fig. 6). From the few specimens showing

an intact ventral surface of the mandible anterior to p4, there does not

appear to be a distinct ridge for the origin of the genioglossus muscle,

unlike the condition in Proterixoides. The lateral side of the man-

dible consistently has two small mental foramina. In most cases, the

most anterior of these is located below p3, and the posterior foramen

is located below the middle of p4 (e.g., SDSNH 56725). In UCMP

133966, however, the two foramina are situated below p2-3, and

SDSNH 54912 has three small foramina below p2-4. The masseteric

fossa ends anteriorly below the m3 hypoconulid, as in Proterixoides.

The dorsal ridge defining the fossa is much stronger than the ventral

ridge. The mandibular symphysis ends posteriorly below p2. The

dorsal margin of the symphysis is defined by a ridge of variable

strength and is quite pronounced below cl. The observed depth of

the mandible below m2 ranges from 5.1 to 5.4 mm (N = 5).

SDSNH 56725 preserves p3-4 and at least portions of the alveoli

for il-p2 (Fig. 6A). All alveoli are crowded closely together and

show that il-p2 were single-rooted. The alveolus for il is heavily

damaged, but the tooth had a proximally slender root, oriented at a

relatively low angle into the mandible. The i2 alveolus was distinctly

larger in diameter than the il and i3 alveoli. In SDSNH 56725, the

round canine alveolus is much larger than the i3 and dp1 alveoli. In

SDSNH 54912, however, the cl alveolus is only slightly larger than

the adjacent alveoli (Fig. 6D). In all available specimens, the p2 al-

veolus is significantly larger than the dp1 alveolus and is subequal in

size to the p3 alveolus.

Lower Dentition.—Lower antemolars of Patriolestes are not pre-

served in available mandibles, and upper incisors of Patriolestes are

not preserved in the only available premaxillary. However, many iso-

lated antemolars of Patriolestes have been collected from several

SDSNH localities in the Friars Formation, and their allocation to

specific loci is aided by comparison to the dentition of Proterixoides

davisi described above. In addition, the few other lipotyphlan and

“proteutherian” taxa in the Poway fauna whose antemolars might be

confused with those of Patriolestes are relatively rare (Walsh 1996).

The least confidently identified antemolar of Patriolestes is il, in

part because the anterior diameter of the il alveolus is not preserved

on any available mandible. Eight isolated teeth, varying greatly in

size, are tentatively assigned to this locus. Their crowns are similar to

that of the il of Proterixoides in being anteroposteriorly compressed

and chisel-shaped (e.g., SDSNH 58919, Fig. 7). However, most of

these teeth are relatively small compared to the i2s of Patriolestes

described below, whereas the crowns of il-2 in Proterixoides are

subequal in size. Furthermore, the roots of the putative ils of

Patriolestes are cylindrical, whereas the il root of Proterixoides is

more robust and distinctly compressed transversely. Finally, unlike

the condition in Proterixoides, most of these teeth have distinct

interdental wear facets on their medial sides, indicating close

appression of the left and right ils. Suspiciously, most of the putative

ils of Patriolestes seem to be too small to occlude properly with the

relatively large I1s described below. Perhaps only SDSNH 58919 is

an il, and the remaining smaller teeth are dils or do not pertain to

Patriolestes.

Ten isolated teeth from the Friars Formation are identified as i2s

of Patriolestes (e.g., SDSNH 58732, Fig. 7). They are very similar to

the homologous teeth of Proterixoides davisi but are distinctly

smaller. Six of the i2s have an appression facet on the ventral side of

the tip of the crown, caused by interdental wear with il. As in il,

occlusal wear facets are developed on the tip and dorsal ridge of the

crown. Six isolated teeth are tentatively identified as di2s (Appendix

1). They are similar in size and morphology to the permanent 12s but

have sharper dorsal ridges and a small posterodorsal cuspule.

On the basis of the relatively small alveolus for 13 in several man-

dibles, 12 isolated teeth with relatively slender roots are identified as

i3s of Patriolestes (e.g., SDSNH 27628, Fig. 7). They have moder-

ately procumbent, anteroposteriorly elongate, trenchant crowns that

are distinctly smaller than those of i2 or c1. The dorsal keel runs from

the posterolabial corner of the crown forward to the sharp anterior

tip. A very weak posterolabial cuspule is sometimes present at the

posterior end of the keel. The 13s often have an appression facet on

the ventral side of the tip of the crown (caused by interdental wear

with 12) and another appression facet at the posterolingual corner of

the crown (caused by interdental wear with c1). The i3 of Patriolestes

is relatively larger and much more trenchant than that of Prote-

rixoides but appears to be similar in morphology to the i3 of

Macrocranion nitens (see Novacek et al. 1985, figs. 1-2). An addi-

tional 11 isolated teeth are identitifed as di3s of Patriolestes. They

are very similar in size and occlusal outline to the i3s, with much

weaker dorsal ridges but a stronger posterolabial and anterodorsal

cuspule.

In Proterixoides davisi, c1 is only slightly larger than i3, and the

two teeth are very similar in crown morphology. In contrast, the 15

isolated cls of Patriolestes are distinctly larger than the 13s and

slightly larger than the i2s. They have relatively broad, anteroposteri-

orly elongate, slightly procumbent, nontrenchant crowns with a rather

bulbous overall appearance (e.g., SDSNH 58828, Fig. 7). No dels

have been recognized.

Sixteen isolated teeth are identified as dp1s of Patriolestes, and

14 isolated teeth are identified as p2s (e.g., SDSNH 58936 and 58720,

Fig. 7). They are similar to the homologous teeth of Proterixoides

davisi in that they are single-rooted, moderately procumbent, and

have triangular crowns with a weak dorsal ridge and a weak

posterolabial cuspule. A slight indentation in the posterior margin of

the crown makes the occlusal outline of dp1 appear “upside-down

heart-shaped” in postero-occlusal view. Again as in P. davisi, referred

p2s of Patriolestes are slightly larger than dp1s. Nine isolated teeth

are identified as dp2s of Patriolestes. They are similar to p2 in size

and in their roughly triangular slightly to moderately procumbent

crowns. However, the deciduous teeth have more anteroposteriorly

elongate crowns, their posterior margin forms a more oblique angle

with the AP axis, and they usually have a distinct posterior cingulid

and distinct posterolabial cuspule.

Stephen L. Walsh

TABLE 4. Measurements of upper cheek teeth of Patriolestes novaceki from the early Uintan of San Diego County.

Specimen

(DP3)/P34¢

P4

AP

W

AP

AW

PW

M1

M2 M3

AP

AW

PW

AP AW PW AP AW

PW

Friars Fm., undifferentiated

SDSNH 58776

SDSNH 51258

LACM(CIT) 55933

LACM(CIT) 56119

UCMP 101690

SDSNH 51293

SDSNH 49350

SDSNH 41353

(2.45)

(1.54)

4.28

4.23

4.46

4.16

Friars Fm., upper tongue

SDSNH 27011

SDSNH 38150

SDSNH 38122

SDSNH 38149

SDSNH 43963

SDSNH 54857

SDSNH 31767

SDSNH 37998

SDSNH 37999

UCMP 101420

UCMP 133967

UCMP 154719

SDSNH 37616

SDSNH 38004

SDSNH 31765

SDSNH 31766

SDSNH 38005

SDSNH 38006

(2.17)

(2.37)

1.76

1.89

1.93

1.85,

(1.64)

(1.66)

2.07

2.64

3.65

4.44

Friars Fm., cong. tongue

SDSNH 56950

SDSNH 58835

SDSNH 43083

SDSNH 56951

SDSNH 56952

SDSNH 56953

SDSNH 56965

SDSNH 60053

SDSNH 56185

SDSNH 43081

SDSNH 55679

SDSNH 56956

SDSNH 58639

SDSNH 60055

SDSNH 60056

SDSNH 56957

(2.46)

(2.32)

1.99

1.93

1.74

2.01

1.73

1.83

1.82

(1.96)

Friars Fm., lower tongue

SDSNH 39748

SDSNH 58721

SDSNH 39747

SDSNH 39761

SDSNH 46139

SDSNH 39335

SDSNH 39683

SDSNH 46373

SDSNH 45802

SDSNH 39682

SDSNH 37400

SDSNH 39685

SDSNH 46274

(2.29)

(2.12)

1.68

1.72

1.94

(1.72)

(1.58)

1.83

1.73

1.90

3.49

3.94

3.51

NNN

05 ae ||

BAS

4Values in parentheses are for DP3; others are for P3.

Dentition and Skull of Proterixoides and Patriolestes

TABLE 5. Measurements of lower cheek teeth of Patriolestes novaceki from the early Uintan of San Diego County.

(dp3)/p3¢ p4 ml m2 m3

Specimen AP W AP WwW AP WTRI WTAL AP WTRI WTAL AP WTRI WTAL

Friars Fim., undifferentiated

SDSNH 56725 -- 1.85 3.75 2.57

SDSNH 49250 (holotype) — = 3.41 2.37 3.27 2.68 253, 2.93 2.63 2.30 2.80 2.04 1.58

SDSNH 49207 3.26 2.81 2.62

SDSNH 49436 3.33 2.56 2.62

SDSNH 50589 3.07 2.62 2.55

SDSNH 51259 3.12 2.60 2.53 _ — = _— _ _

SDSNH 49611 — — — — 2.73 1.91 1.49

LACM(CIT) 55935 _ — 3.22 2.30 3.20 2.56 2.60 - — — _ =

LACM(CIT) 55936 _ =_ 3.22 2.33 3.07 2.52 Zell

LACM(CIT) 55937 — = — — 3.21 2.71 2.68 2.96 2.66 2.34 _— — 1.55

LACM(CIT) 56189 2.63 — 1.67

Friars Fm., upper tongue

SDSNH 58723 (2.04) (1.40) = = =

SDSNH 58743 (2.03) (1.40)

SDSNH 58744 (2.08) (1.46) — _— — —

SDSNH 58808 (2.13) (1.50)

SDSNH 60026 (2.13) (1.46)

SDSNH 32004 2.04 1.66

SDSNH 32080 1.88 1.60

SDSNH 58747 1.84 1.53 — — — — — —

SDSNH 58748 1.91 1.58 _—

SDSNH 58749 2.08 1.67

SDSNH 58791 1.78 1.66 _— — —_ _— — _—

SDSNH 37609 — — 3.35 2.31

SDSNH 37987 _— — 3.02 2.10

SDSNH 37988 — _ 3.27 2.25 — — — —

SDSNH 42428 — _ 3.40 2.24

SDSNH 45160 — — 3.40 2.42

SDSNH 46436 — — 3.49 2.50 —

SDSNH 31764 2.96 2.24 2.15

SDSNH 37610 3.10 2.31 2:32

SDSNH 37992 = — —_— — 3.15 2.60 2.52 _ — — _— —

SDSNH 45161 3.60 2.85 2.78

SDSNH 56717 _— 3.04 2.57 2.44 _ — _—

SDSNH 37993 = — _— — — = 2.73 2.55 2.42 _— _— —

SDSNH 47865 2.45 _—

SDSNH 37612 2.58 2.05 1.63

SDSNH 37613 — — 1.56

SDSNH 37994 2.50 1.89 1.53

SDSNH 37995 — — — — — — 1.79

SDSNH 46555 — — — _ — — 2.56 1.92 1.55

SDSNH 46556 3.00 2.11 1.65

SDSNH 54858 _— — 3.75 2.65 3.37 2.71 2.95 3.34 2.91 2.82 3.34 2.38 1.96

UCMP 101617 2.82 2.23 2.09 _— _ —

UCMP 133966 —_— _— 3.40 239) 3.24 2.44 2.45 — _— =— _ _— —

Friars Fm., cong. tongue

SDSNH 60042 (2.10) (1.43)

SDSNH 60041 2.06 1.82 _ _ — —

SDSNH 58830 1.95 1.69 _— _— —_— —

SDSNH 58837 1.94 1.67 _— = — —

SDSNH 55498 — — Br12 2.49 — _—

SDSNH 55825 _— — 3.66 2.50 _ — —

SDSNH 60043 — — 3.31 2.38 — — = =

SDSNH 55499 _ — 3.38 2.24 3.20 2.53 2.50 — =

SDSNH 55500 — — 3.40 2.46 3.20 2.59 2.39 — —

SDSNH 54912 3.09" 2.40 2.54 2.91 2.55 2.34 — _ —

SDSNH 54926 3.32 2.56 2.52 _ — —

SDSNH 43276 aa — — _— 2.55 2.35 — = =

SDSNH 56947 3.08 2.47 2.47 3.00 2.54 2.40 2.83 = 1.76

SDSNH 60044 — — — _ 3.18 2.64 2.65

SDSNH 60045 3.24 2.58 2.70 _—

SDSNH 60046 — — — 2.96 2.54 2.30 _— _— —

SDSNH 56948 1.55,

SDSNH 56949 — 2.78 2.00 1.67

SDSNH 60047 _— 2.70 2.22 1.81

(Continued)

TABLE 5 (Continued).

Stephen L. Walsh

(dp3)/p3¢ p4

ml

m2 m3

Specimen AP WwW AP WwW AP

WTRI

WTAL AP WTRI WTAL AP WTRI WTAL

Friars Fm., lower tongue

SDSNH 58935

SDSNH 58971

(1.96)

(1.90)

(1.48)

(1.30) — = —

SDSNH 58927 1.77 1.55 — — —

SDSNH 58972 1.63 1.64 —

SDSNH 45799 = =

SDSNH 45800 _— — =

SDSNH 54995

SDSNH 39680

Nn

Nm hy

of

2.39 18 2.02 —_— — a

NN

2.51 18 1.94 - — —

SDSNH 54996

SDSNH 39681

SDSNH 45801 — =— — = —

SDSNH 46272

Santiago Fm., Member B

SDSNH 36576 —

4Values in parentheses are for dp3; others are for p3.

SDSNH 56725 contains the only permanent erupted p3 preserved

in a mandible (Fig. 6A). It is fully premolariform and much smaller

than the p3 of Proterixoides. Eleven isolated p3s show a bulbous crown

with a rounded pentagonal occlusal outline (e.g., SDSNH 58837, Fig.

7). They lack a dorsal ridge but generally retain a weak posterolabial

cuspule. Although the p3 appears to be single-rooted in SDSNH 56725,

two isolated p3s from SDSNH Loc. 3483 show that a short distance

below the base of the crown, the trunk root bifurcates into two separate

A

but closely appressed anterior and posterior branches (e.g., SDSNH

58749). Of the remaining six isolated p3s with an intact single root, four

have a distinct longitudinal indentation running down the lingual side.

Crown morphologies of the double-rooted and single-rooted morphs

are indistinguishable, being similar to, but slightly larger and dorso-

ventrally thicker than p2.

Eight isolated teeth are identified as dp3s of Patriolestes (e.g.,

SDSNH 60026; Fig. 8A). There is no differentiation of the trigonid,

Figure 5, Stereophotographs of SDSNH 49250, holotype of Patriolestes novaceki, R mandible fragment with p4-m3 + alveoli for i2-p3. (A), Occlusal

view; (B), lateral view. Scale bar, 5 mm.

Dentition and Skull of Proterixoides and Patriolestes 13

Figure 6. Stereophotographs in occlusal view of four right mandible fragments of Patriolestes novaceki, showing anterior alveoli and composite p3-m2.

(A), SDSNH 56725; (B), UCMP 133966; (C), SDSNH 55500; (D), SDSNH 54912. Scale bar, 5 mm.

and only a very weak paracristid. The dp3 differs from p3 in the same

way that dp2 differs from p2 (i.e., the deciduous teeth are more

anteroposteriorly elongate, their posterior margins make a more

oblique angle with the AP axis, and they usually have a distinct pos-

terior cingulid and a distinct posterolabial cuspule). The single root

of dp3 is cylindrical in cross-section. Only SDSNH 60026 preserves

enough of the root to show that a weak longitudinal indentation was

present on the lingual side. The dp3 of Patriolestes differs from the

previously undescribed dp3 of Proterixoides (e.g., SDSNH 46624;

Fig. 8C) in that the latter is larger (both absolutely and relative to

dp4), has a stronger paracristid and postmetacristid, lacks a posterior

cingulid and posterolabial cuspule, and has a posterior margin whose

occlusal outline is essentially perpendicular to the AP axis.

Several isolated p4s show two essentially cylindrical roots of

subequal diameter. The p4 is also present in several mandible frag-

ments (Figs. 5-6) and is unique among known dormaaliids in having

a bulbous trigonid, with no differentiation of metaconid or paraconid

and no paracristid (in this respect, the p4 of Patriolestes is similar to

the p3 of Proterixoides). The p4 is also wide transversely relative to

the p4 in most other dormaaliid taxa. The labial side of the tooth is

uniformly convex in occlusal outline, while the lingual side is slightly

concave. The talonid is relatively short, lacks a cristid obliqua, and

Figure 7. Stereophotographs of composite Ril-p3 of Patriolestes novaceki. (A), Occlusal view; (B), lateral view. il: SDSNH 58919. i2: SDSNH 58732.

i3: SDSNH 27628. cl: SDSNH 58828. dp1: SDSNH 58936. p2: SDSNH 58720. p3: SDSNH 58837. Scale bar, 5 mm.

14 Stephen L. Walsh

Figure 8. Stereophotographs in occlusal view of (A), SDSNH 60026,

isolated Rdp3 of Patriolestes novaceki; (B), SDSNH 45798, isolated Rdp4 of

Patriolestes novaceki; (C), SDSNH 46624, associated Rdp3-4 of Pro-

terixoides davisi. Scale bar, 3 mm.

invariably bears one small anteroposteriorly compressed cusp at the

posterolingual corner of the crown. A short ridgelike postcingulid

usually extends from the apex of the talonid cusp to the posterolabial

corner of the crown.

Two dp4s of Patriolestes are known (e.g., SDSNH 45798, Fig.

8B). They are molariform, anteroposteriorly elongate teeth, with a

distinct metaconid, protoconid, paraconid, cristid obliqua, and

hypoconulid. The talonid basin is completely enclosed. The

protoconid is slightly larger and taller than the metaconid and lies

slightly anterior to the latter cusp. It should be noted that SDSNH

60026 is the largest of the eight known dp3s of Patriolestes novaceki,

while SDSNH 45798 is distinctly smaller than the only other known

dp4. Thus, the teeth illustrated in Figs. 8A-B are probably closer in

size than would be the case in most individuals of this species. The

dp4 of Patriolestes differs from the previously undescribed dp4 of

Proterixoides davisi (e.g., SDSNH 46624, Fig. 8C) in that the latter

has slightly more rounded cusps, a relatively smaller talonid basin, a

weaker cristid obliqua, a relatively narrower trigonid whose labial

face is concave rather than convex, and a much more expanded

paraconid region.

The m1 and 2 of Patriolestes are very similar in size and mor-

phology. In a given mandible, mls are always slightly larger than

m2s (Figs. 5 and 6). In contrast to the condition found in sespedec-

tines and most other dormaaliids (Novacek et al. 1985:4), the ml

trigonid of Patriolestes is usually slightly wider than the talonid, as is

invariably the case in m2. This fact often makes even moderately

worn isolated m1-2s difficult to allocate. Both teeth have erect

anteroposteriorly compressed trigonids with a low bladelike

paracristid that extends from the protoconid apex to the anterolabial

base of the metaconid. The paracristid is weaker and more closely

appressed to the metaconid in m2. The paraconid is never a distinct

cusp and is represented by only a slight swelling at the end of the

paracristid. A short cingulid is present at the anterolabial base of the

teeth. The weak cristid obliqua extends from the hypoconid apex to

the extreme base of protoconid. A weak ridge also extends from the

entoconid apex to the extreme base of the metaconid, closing off the

talonid basin. The hypoconulid is usually a small distinct cusp, cen-

trally located at the posterior magin of the talonid. Both anterior and

posterior roots are anteroposteriorly compressed.

The m3 is usually shorter and always narrower than m1-2, al-

though the trigonid morphology is similar to that of m1-2 (Fig. 5).

The m3 talonid is always narrower than the trigonid. The hypoconulid

is of variable strength but is always stronger than in m1-2, often

extending well posterior to the entoconid. The anterior root is

anteroposteriorly compressed, while the posterior root is anteropos-

teriorly elongate.

Skull. — Very little of the skull of Patriolestes is known. SDSNH

55497 is a palatal fragment consisting of part of the left maxillary

and premaxillary and part of the right maxillary (Fig. 9). As seen in

ventral view, a right-lateral “fault” running through the midline has

sheared the right maxillary about 9 mm anterior to the left maxillary.

At the level of DP1-P3 is a distinct ridge on the medial edge of each

maxillary, possibly not attributable to distortion, since a similar ridge

is present in SDSNH 56185. The left premaxilla fragment preserves

damaged alveoli for 11-3, and the alveolus for I2 retains the dorsal

end of the root. The left maxillary fragment preserves a double-rooted

complete C1, single-rooted complete DP1 and P2, the alveoli for

triple-rooted P3, and the lingual alveolus for a triple-rooted P4. The

right maxillary preserves the broken roots of Cl, the alveolus for

DP1, a complete P2, and the alveoli for P3. There are no diastemata

between any of the upper teeth. However, the left maxillary—premax-

illary suture has pulled apart slightly and filled with matrix, giving

the false impression of a short diastema between C1 and the empty I3

alveolus. The dorsal side of SDSNH 55497 preserves fragments of

the nasals, but their original shape and contact relationships to other

bones cannot be determined.

SDSNH 54857 and 56185 are maxillary fragments of Patriolestes

whose teeth are described further below. They show that the anterior

opening for the infraorbital canal was dorsal to a point between P3

and 4. This position is similar to that seen in Scenopagus (Butler

1972, fig. 5) but perhaps slightly more posterior than in Proterixoides,

Macrocranion, or Ankylodon. The dorsoventral diameter of the

infraorbital foramen is 2.1 and 2.4 mm in SDSNH 54857 and 56185,

respectively, which are about the same as in Proterixoides. The exact

length of the infraorbital canal in SDSNH 54857 and 56185 is not

currently measurable owing to obscuring matrix but appears to be

comparable to that of Proterixoides. As in a variety of lipotyphlan

taxa, the ventral side of the zygomatic root originates lateral to M2.

The anteroventral surface of the zygomatic arch is distinctly concave

in SDSNH 54857 and 56185, as in Proterixoides. As seen in SDSNH

56185, the magnitude and rate of descent of the level of the palate

from M1 to Cl is moderate and gradual. The width of the snout nar-

rows distinctly from P4 to P3, widens slightly at C1, then narrows

again at the level of the premaxillaries.

Upper Dentition.—In SDSNH 55497, the alveolus for I] is dam-

aged but seems to have been at least as large as that for I2. By anal-

ogy with Proterixoides, 17 relatively large isolated incisors are iden-

tified as I1s of Patriolestes (e.g., SDSNH 58973, Fig. 10). They have

transversely compressed slightly curved roots and tall oval crowns

developing lingually concave shovel-shaped wear facets. Unworn

Ils have a weak anterior ridge and a stronger posterior ridge de-

scending from the apex of the crown. Some of these putative I1s are

mitten-shaped in that they possess a small posterior cusp or “thumb”

but seem too robust to be deciduous teeth.

From the broken root in SDSNH 55497, I2 of Patriolestes was

also a relatively large tooth, though slightly smaller than I1, and had

a cylindrical rather than transversely compressed root. Ten isolated

incisors seem to possess this morphology (e.g., SDSNH 56714, Fig.

10). They have slightly curved roots and transversely compressed

crowns similar in morphology to I1. However, the crowns of putative

12 are dorsoventrally and anteroposteriorly shorter than the putative

Ils. Thirteen isolated incisors are tentatively identified as DI1s and/

or DI2s (Appendix 1). They have erect mitten-shaped crowns with

distinct posterior “thumbs” and slightly curved roots that are vari-

ably compressed or cylindrical.

The alveolus for I3 is damaged in SDSNH 55497, but by analogy

with Proterixoides, the 13 of Patriolestes is probably a relatively

small tooth with a short, straight, cylindrical root. The 29 teeth so

identified here have anteroposteriorly elongate, slightly procumbent

Dentition and Skull of Proterixoides and Patriolestes 1S

Figure 9. Stereophotographs in occlusal view of SDSNH 55497, palatal fragment of Patriolestes novaceki with alveoli for LI1-3, complete LC1-P2,

partial alveoli for LP3-4, broken roots of RC1, alveolus for RDP1, complete RP2, and alveolus for RP3. Scale bar, 5 mm.

crowns, which consist of a tall anterior cusp, a central dorsal ridge,

and a very weak posterior cuspule (e.g., SDSNH 58760, Fig. 10).

Twenty-one additional isolated teeth are similar to the putative I3s

but have stronger posterior cuspules and are tentatively identified as

DI3s. Both putative I3 and DI3 develop a steeply dipping planar wear

facet on the lingual face of the crown, presumably caused by occlu-

Figure 10. Stereophotographs of composite RI1-I3 of Patriolestes

novaceki. (A), Occlusal view; (B), lateral view. 11: SDSNH 58973. I2:

SDSNH 56714. 13: SDSNH 58760. Scale bar, 3 mm.

sion with the trenchant crown of i3. The relatively large number of

teeth identified as [3 and DI3 is suspicious. Perhaps the currently

unrecognized DP2 of Patriolestes is morphologically indistinguish-

able from I3 and/or DI3, resulting in an artificially high abundance of

the latter teeth.

The Cl of Patriolestes is preserved in place in SDSNH 55497

(Fig. 9) and SDSNH 56185 (Fig. 11A) and is also represented by 13

isolated teeth. Cl is double-rooted, with a large, bulbous, antero-

posteriorly elongate crown, much larger than I3, DP1, and P2, and

also much larger than Cl in Proterixoides davisi. A weak ridge is

sometimes present on the posterodorsal part of the crown, sometimes

terminating in a weak posterior cuspule. A posteriorly dipping wear

facet (presumably caused by occlusion with c1) develops on the apex

of the crown. Both roots of C1 are cylindrical in cross-section. No

unequivocal DCls have been recognized. SDSNH 58800 has re-

sorbed (?) roots and very thin enamel but does not differ significantly

in size or morphology from undoubted permanent Cls.

DP1 and P2 are preserved in place in SDSNH 55497 (Fig. 9) and

are similar in size and morphology to their homologs in Proterixo-

ides. The DP1 crown is semiglobular, slightly anteroposteriorly elon-

gate, and has a weak ridge running anteroposteriorly down the mid-

line. Isolated DP1s have a slender cylindrical root with no longitudi-

nal indentation on their lingual side (e.g., SDSNH 58820). The P2 is

very similar in crown morphology to DP1 but distinctly larger. The

single cylindrical root of P2 sometimes bears a distinct longitudinal

indentation running down its lingual side (e.g., SDSNH 58823). DP2s

have not been recognized, although as noted above it is possible that

some of the teeth identified as I3s/DI3s are actually DP2s.

The permanent P3 of Patriolestes is preserved in two maxillary

fragments (SDSNH 56185, Fig. 11A; SDSNH 54857; Fig. 11B) and is

also represented by 12 isolated teeth. It is always much smaller than P4

and than the P3 of Proterixoides. All P3s are triple-rooted, but in three

specimens (SDSNH 31996, 56185, and 56965) the lingual and

posterolabial roots are partially fused. Most P3s have an occlusal out-

16 Stephen L. Walsh

Figure 11. Stereophotographs in occlusal view of maxillary fragments of Patriolestes novaceki: (A), SDSNH 56185, R maxillary fragment with C1, alveoli

for DP1-P2, and P3-4. (B), SDSNH 54857, R maxillary fragment with P3-M2. Scale bars, 5 mm.

line approximating a rounded equilateral triangle. The crown is domi-

nated by a large conical paracone and never shows a distinct parastylar

lobe or parastyle. The protocone lobe ranges from virtually absent

(SDSNH 43083) to relatively strong (SDSNH 56951). The protocone

ranges from absent (SDSNH 56953) to weak (SDSNH 46139) to mod-

erately developed (SDSNH 56951). There is no metacone or hypocone.

The metastylar lobe ranges from weak (SDSNH 56952) to strong

(SDSNH 56951). A weak labially concave metastylar crest connects

the apex of the paracone to the metastylar region. A distinct metastyle

is present only on SDSNH 39761.

The permanent P4 is preserved in place in SDSNH 56185 and

SDSNH 54857 (Fig. 11) and is also represented by 12 isolated teeth.

The labial half of the crown is dominated by a large conical paracone,

which sends a sharp metastylar crest sweeping posterolabiad to the

end of a strong metastylar lobe. A metacone and metasyle are absent.

The parastylar lobe is relatively weak but occasionally bears a tiny

parastyle. The protocone is conical, and from its apex a postproto-

crista of variable strength extends posteriad to merge with a weak

posterior cingulum, which in turn extends to the metastylar area.

There is no hypocone. A short posterolingual cingulum is usually

present, but there is never an anterolingual cingulum.

Seven isolated DP3s are known (e.g., SDSNH 56950; Fig. 12A).

Allare triple-rooted, with the posterolabial root being the largest. All

have a very small protocone lobe and protocone, a very small

parastylar lobe, no parastyle, and no hypocone. The anterior half of

the tooth is similar to P3 in that it is dominated by a large conical

paracone. However, the posterior half of DP3 consists of a very large

metastylar lobe, upon which sits a distinct metacone. A cingulum

connects the posterior base of the protocone lobe to the lingual base

of the metastylar lobe. The DP3 of Patriolestes differs from the pre-

viously undescribed DP3 of Proterixoides (e.g., SDSNH 43584, Fig.

12B) in that the latter is much larger relative to DP4, lacks a

metacone, has a transversely very wide parastylar lobe with a distinct

parastyle, and has a much stronger protocone lobe, with a distinct

protocone and hypocone.

A single damaged DP4 of Patriolestes is known (UCMP V-

73138/154720; Fig. 12C). The parastylar lobe, metacone, and most

of the metastylar lobe have been broken away, but enough of the

crown remains to show that the tooth was fully molariform. The

paracone is conical and much larger than in M1-2. A large, posteriad-

jutting metastylar lobe is present. Weak pre- and post-paraconular

wings are present, along with stronger pre- and post-metaconular

wings. A distinct hypocone sends a strong posthypocrista labiad to

the posterior base of the crown. The hypocone extends slightly more

lingually than the protocone. A short anterior cingulum is present at

the anterior base of the protocone. Measurements: AW = 3.57 mm;

PW = 3.72 mm (min.). The DP4 of Patriolestes differs from the pre-

viously undescribed DP4 of Proterixoides (e.g., SDSNH 46638, Fig.

12D) in that the latter is smaller in absolute size, is much less trans-

verse, and has relatively smaller paracones and metacones.

M1 is the largest upper tooth of Patriolestes. The paracone and

metacone are conical, steep-sided, tall, and relatively small in diam-

eter compared to those of Proterixoides. In unworn specimens, the

Figure 12. Stereophotographs in occlusal view of (A), SDSNH 56950,

isolated RDP3 of Patriolestes novaceki; (B), SDSNH 43584, isolated LDP3

of Proterixoides davisi; (C), UCMP 154720, isolated RDP4 of Patriolestes

novaceki; (D), SDSNH 46638, isolated RDP4 of Proterixoides davisi. Scale

bar, 3 mm.

Dentition and Skull of Proterixoides and Patriolestes 17

paracone is slightly taller than the metacone. The parastylar lobe is

weak, with a weak parastyle but no parastylar crest. The metastylar

lobe is well developed, with no metastyle but with a strong metastylar

crest. There is a weak continuous ectocingulum along the labial bor-

der of the tooth. The strong protocone bears strong pre- and

postprotocristae. The preprotocrista branches into a preparaconular

wing (which extends to the anterior base of the paracone) and a

postparaconular wing (which extends to the lingual base of the

paracone). The postprotocrista branches into a premetaconular wing

(which extends to the anterolingual base of the metacone) and a

postmetaconular wing (which extends to the end of the metastylar

lobe as a posterior cingulum). The protocone is much taller than the

hypocone, but the latter always extends further linguad than the

former. A strong posthypocrista extends labiad to the posterior base

of the tooth. A short anterior cingulum is present at the anterior base

of the protocone. The tooth has three roots. The lingual root is some-

what compressed transversely (see empty alveolus in SDSNH 56185;

Fig. 11A), and is much larger than the cylindrical labial roots.

M2 is similar to M1 with the following exceptions. The parastylar

lobe extends labiad much more than the metastylar lobe, and the

resulting ectoflexus is much deeper than in M1. A weak parastylar

crest is present in M2 (absent in M1). The hypocone usually does not

extend lingually past the protocone (one exception is SDSNH 39682).

Both the lingual and labial halves of M2 are anteroposteriorly shorter

than in M1, giving M2 a relatively transverse appearance. The lin-

gual root in M2 is smaller and more cylindrical than in M1.

M3 (not figured) is much smaller than M1-2 and has the triangu-

lar occlusal outline typical of most dormaaliids. The paracone is coni-

cal and much larger and taller than the metacone. A prominent

parastylar lobe extends anterolabiad, and is connected to the base of

the paracone by a weak parastylar crest. There is no parastyle. A

weak paraconule is present, along with a weak preparaconular wing.

The postparaconular wing is sometimes absent (SDSNH 38005) and

weak when present (SDSNH 45162). A weak metaconule on the post-

protocrista is sometimes present (SDSNH 31766), sometimes absent

(SDSNH 38005). There is no premetaconular wing. Weak to moder-

ately strong anterior and posterior cingula are always present but

never complete around the lingual face of the protocone. There is no

hypocone. The lingual root is relatively large and transversely com-

pressed, while the two labial roots are small and cylindrical.

Discussion.—An isolated, fragmentary P4 from the Friars For-

mation (UCMP 101690) was assigned to Proterixoides davisi by

Novacek (1985:16-17). From its size and morphology, however,

UCMP 101690 can now be positively referred to Patriolestes

novaceki. Similarly, the jaw fragments and isolated teeth from LACM

(CIT) Locs. 249, 249-S, and 249-P that were assigned by Golz and

Lillegraven (1977, table 1) to “cf. Proterixoides sp.” and “cf.

Proterixoides davisi” are clearly referable to Patriolestes novaceki.

As aresult of these revised identifications, the genus Proterixoides is

unknown from the Friars Formation and is restricted to the late Uintan

and Duchesnean. Finally, UCMP 101059 was included by Novacek

(1976:36-37) in the original sample of the “erinaceoid-like genus

and species.” This fragmentary M1 has a lingual root that is extremely

compressed anteroposteriorly, unlike the more cylindrical roots of

undoubted upper molars of Patriolestes. lassign UCMP 101059 toa

probable new leptictid genus, which is represented by more com-

plete material from the Friars Formation (e.g., SDSNH 3483/4381 1).

Correlation of Size and Stratigraphic Occurrence

A considerable amount of size variation is evident in the avail-

able sample of Patriolestes. To reveal possible correlations between

size and stratigraphic occurrence, the raw measurements for cheek

teeth are segregated by stratigraphic unit (Tables 4 and 5). Selected

measurements for p4-m3 and P4-M1 are illustrated in the bivariate

plots of Figs. 13 and 14. The few available teeth from the lower

tongue of the Friars Formation (collected from two districts, Santee

and South Creek) usually fall at the small end of the range for each

tooth position. Teeth from the conglomerate tongue of the Friars For-

mation (one district—Scripps Ranch North) consistently fall in the

middle of the range for each tooth position. Teeth from strata consid-

ered by Walsh et al. (1996) to pertain to the upper tongue of the Friars

Formation show the most size variation. As discussed by Walsh et al.

(1996), localities in the upper tongue of the Friars Formation along

State Route 52 (SDSNH Locs. 3611, 3612, 3685, 3883) are roughly

equivalent stratigraphically to localities from the upper part of the

undifferentiated Friars in the lower end of Murphy Canyon (San Di-

ego Mission district; SDSNH Locs. 3784, 3785; LACM(CIT) Locs.

249, 249-S, 249-P). This stratigraphic similarity is reflected by teeth

from these nearby districts tending to be relatively large. Indeed, the

largest examples of p4, m2, m3, M1, and M2 of Patriolestes novaceki

are represented by two specimens from a single locality in the upper

tongue of the Friars in the State Route 52 district (SDSNH 54857 and

54858). Interestingly, however, teeth from the upper tongue of the

Friars Formation in the Rancho Penasquitos and Carmel Mountain

Ranch districts show a wide range in size variation, and a few of

them are even smaller than teeth from the lower tongue of the Friars.

Finally, the single measurable tooth from Member B of the Santiago

Formation falls in the middle of the range of m3s from the Friars

Formation (Fig. 13D).

To use m1 and 2 to illustrate the maximum size variation in the

sample, the longest ml (SDSNH 45161) is 36% longer than the short-

est ml (SDSNH 54995), and the longest m2 (in SDSNH 54858) is

40% longer than the shortest m2 (SDSNH 39680). The holotype speci-

men of Patriolestes novaceki was collected from the undifferentiated

Friars Formation in the San Diego Mission district, which yields rela-

tively large specimens. Given this considerable variation in size, it is

possible that a smaller species of Patriolestes may eventually be rec-

ognizable. However, more specimens should be collected from the

lower tongue of the Friars Formation for the relatively small size of

individuals comprising this “population” to be corroborated. Addi-

tional collections from the upper tongue of the Friars are also needed

to document further the apparent geographic variation in the size dis-

tribution of individuals preserved therein. Finally, as noted above, the

relative size of cl was apparently variable in Patriolestes (e.g.,

SDSNH 56725 vs. 54912), and the possibility that sexual dimorphism

accounts for some of the size variation should not be discounted. For

the present, I regard all specimens as pertaining to a single variable

species evolving through time. Therefore, statistics for the entire

pooled sample of Patriolestes teeth are given in Table 6.

Trigonid/Talonid Proportions of m1-2

Based on the data in Table 5, Fig. 15A plots talonid width vs.

trigonid width for all measurable m1s of Patriolestes. Most points plot

above the line of equal widths, showing that the m1 trigonid is often

slightly wider than the talonid. This condition is unusual for most

Eocene dormaaliids, in which the m1 trigonid is usually slightly nar-

rower than the talonid (Novacek et al. 1985:4). Figure 15B plots m2

talonid width vs. trigonid width and shows that all points fall well

above the line of equal widths. Both figures again show the relatively

small size of teeth fom the lower tongue of the Friars Formation, while

teeth from the upper tongue of the Friars are more variable in size.

Interdental Proportions of Lower Cheek Teeth

Trends in the relative anteroposterior lengths of the lower cheek

teeth of Patriolestes are shown in Table 7, which is based on the data

in Table 5 but incorporates only associated teeth or teeth in man-

dibles. The length of m1 is normalized to 1.00, and the correspond-

ing relative lengths of p4 and m2-3 are indicated. The p4 of

Patriolestes is always slightly longer than m1, and m2 is always

18 Stephen L. Walsh

KEY TO FIGURES 13-15:

x

e

e

fe}

X = Friars Fm. undifferentiated

O = upper tongue of Friars Formation

@ = conglomerate tongue of Friars Formation

0 = lower tongue of Friars Formation

+ = Member B of Santiago Formation

= 2.7 A

fe)

E

= AG

x= x

5 25 ° ©e

= ®

. 2.4

% oO e laa ae

= 2.3 X oo

a oO @

2.2

3.1 3.2 33 3.4 35 3.6 3.7 3.8

p4 ANTEROPOSTERIOR LENGTH (mm)

2.9 B

(e)

— C8 x

E

= PE/ x o)

ANS

= x @ holotype

Q 2.6 xO Ce °,

= x

= e

fa 2.5

z

ro) Oo

9

jag

—_

E

2:6) 2-7) re: By 2-9) 53/0351

mi ANTEROPOSTERIOR LENGTH (mm)

3.2 3.3 3.4 3.5 3.6

2.9

2.8

2.7

holotype x

2.6 x °

(e) Ge50e

m2 TRIGONID WIDTH (mm)

2.4 25 26 2.7 28 2.9 30 3.1 32 33 3.4

m2 ANTEROPOSTERIOR LENGTH (mm)

(o}

E 23

E fo)

zr 2.2 e

as

S 2.1 A, fo)

fas) se) X— holotype

5 e

{e)

0)

a

i

3.2 3.3 3.4

PY PES) BS Fee BS ee) SHO EL

m3 ANTEROPOSTERIOR LENGTH (mm)

Figure 13. Plots of anteroposterior length vs. trigonid width for p4-m3 of Patriolestes novaceki. (A), p4; (B), m2; (C), m1; (D), m3.

slightly shorter than m1. The m3 is usually distinctly shorter than m1

but was equal to m1 in one specimen. For comparative purposes,

mean values for the relative anteroposterior lengths of p4-m3 for

Patriolestes and several other dormaaliid genera are presented in

Table 8, and discussed further below.

TAXONOMIC COMPARISONS

Patriolestes is similar to middle Eocene sespedectines (Sespe-

dectes, Proterixoides, and Crypholestes; Novacek 1976, 1985) in

having a moderately enlarged proclivous chisel-shaped i2, a p4

slightly longer than m1 (Table 8), and relatively small single-rooted

DP1-P2. Sespedectines differ from Patriolestes in having a relatively

larger il, a relatively smaller nontrenchant 13 subequal in size to cl,

a double-rooted relatively large p3, a differentiated p4 trigonid, more

anteroposteriorly elongate m1-2s, a trigonid narrower than the

talonid on m1, a relatively much smaller C1 whose two roots are

sometimes partially fused, a relatively much larger P3, and less

tranverse M1-2.

Patriolestes is similar to the early-to-middle Eocene dormaaliine

Macrocranion (e.g., Bown and Schankler 1982, plate 5; Novacek et

al. 1985, fig. 2; MacPhee et al. 1988) in having a small single-rooted

p2, a very small usually single-rooted p3 (a specimen with a double-

rooted p3 was assigned to M. nitens by Bown 1979, fig. 42d), a rela-

tively simple p4 trigonid, and a very small P3. Macrocranion differs

from Patriolestes in having relatively small lower incisors and ca-

nines, a p4 with a distinct paracristid, a p4 consistently shorter than

ml, a slight progressive increase in anteroposterior length from m1

to m3 (Table 8), longer m1-3 trigonids, an m1 talonid wider than

trigonid (Krishtalka 1976), a single-rooted C1, a double-rooted P2, a

less transverse M1-2, and a relatively larger M3.

Patriolestes is similar to certain European Eocene “amphile-

murids” (e. g., Pholidocercus; von Koenigswald and Storch 1983) in

its reduced single-rooted p2-3, simplified p4 trigonid, and relatively

small P3. Pholidocercus differs from Patriolestes in having i2 smaller

than i3, cl smaller than dp1, dp1-p2 larger than p3, a relatively small

single-rooted C1, a double-rooted P2, and more bunodont molars.

Novacek et al. (1985) considered the Amphilemuridae to be a family

separate from the Dormaaliidae, but Novacek (1985) regarded this

taxon as a tribe of the Dormaaliinae.

Patriolestes is similar to various species of the early-to-middle

Eocene scenopagine Scenopagus in having relatively anteroposteriorly

compressed m1 -2 trigonids, a very small P3, relatively transverse M1 -

2, and relatively sectorial molars (e. g., McKenna and Simpson 1959;

Robinson 1966, plate III; McGrew and Sullivan 1970, fig. 6A;

Krishtalka 1976, figs. 4-6; Bown 1982, plate 2A,B). Interestingly,

Dentition and Skull of Proterixoides and Patriolestes 19

E

£

ac

=

Q

=

a

oO

joa

WwW

=

a

<x

vr

a

“3.0 3.1 3.2 33 34 35 36 3.7 3.8 3.9

P4 ANTEROPOSTERIOR LENGTH (mm)

46 B °

M1 ANTERIOR WIDTH (mm)

3.2 3.3 3.4 3.5

25 26 2.7 28 29 30 3.1

M1 ANTEROPOSTERIOR LENGTH (mm)

Figure 14. Plots of anteroposterior length vs. anterior width for P4-M1 of

Patriolestes novaceki. (A), P4; (B), M1. For definitions of symbols, see Fig-

ure 13.

Robinson (1968:4) reported an enlarged lower incisor in S. priscus

(YPM 15254; assigned to S. curtidens by Krishtalka 1976: 14). If cor-

roborated, this could prove to be a significant similarity shared with

Patriolestes. A complete lower jaw from the Wind River Formation

(Wasatchian) was identified as Scenopagus sp. by Stucky et al. (1990,

fig. 9). This specimen apparently has a double-rooted p2-3, a slightly

enlarged cl, and a small i3, but the nature of il-2 is unclear.

Scenopagus differs from Patriolestes in having a relatively larger

double-rooted p3 with a “paraconid flange,” a differentiated p4 trigonid

(Krishtalka 1976, figs. 4A-B), and a p4 consistently shorter than m1

(Table 8). Some specimens of Scenopagus show a slight progressive

decrease in anteroposterior length from m1 to 3, whereas others do not

(e.g., Krishtalka 1976, fig. 6A vs. fig. 4B).

The upper dentition of Scenopagus anterior to P3 is apparently

unknown. McKenna and Simpson (1959) noted that empty alveoli

TABLE 6. Standard statistics for teeth of Patriolestes

novaceki from the early Uintan of San Diego County.

Tooth N OR M SD CV

Il AP 16 1.53-2.05 1.77 0.172 OY

W 16 1.07-1.48 1.30 0.143 11.0

102 AP 9 1.51-1.96 1.66 0.150 9.0

W 10 1.04-1.45 1.24 0.116 9.4

13 AP 27 1.23-2.04 1.51 0.185 12.2

W 27 0.89-1.43 1.09 0.127 11.7

Cl AP 10 2.60-3.17 2.85 0.197 6.9

WwW 11 1.55=2.11 1.85 0.177 9.6

DP1 AP 8 1.01-1.30 1.15 0.093 8.1

W ©) 0.89-1.15 1.02 0.101 9.9

P2 AP 16 1.49-1.84 1.62 0.103 6.3

W 17 1.19-1.56 1.41 0.107 7.6

P3 AP 14 1.68-2.01 1.84 0.107 5.8

W 14 1.56-2.07 1.87 0.158 8.5

DP3 AP vf 2.12-2.46 Deshi 0.130 5.6

W 6 1.54-1.96 1.68 0.149 8.9

P4 AP 7 3.31-3.81 Shs)! 0.166 4.7

AW 8 2.92-3.42 3.20 0.187 5.8

PW 7 4.16-4.46 4.31 0.121 2.8

Ml AP 15 2.52-3.38 3.02 0.229 7.6

AW 13 3.49-4.61 3.97 0.307 Tell

PW 14 3.89-5.19 4.39 0.386 8.1

M2 AP 3 2.62-2.86 2.73 0.122 4.5

AW 3 4.19-5.52 4.67 0.738 15.8

PW 5 3.80-4.86 4.15 0.411 9.9

M3 = AP 4 1.55-2.12 Nez7/ 0.246 13.9

AW 4 2.77-3.16 3.03 0.175 5.8

PW 8 2.10-2.64 2.31 0.167 7.2

il AP 5 1.39-2.34 1.84 0.352 19.1

W 8 0.84-1.13 1.00 0.120 12.0

i2 AP 7 2.25-2.47 2.39 0.078 3.3

Ww 10 1.00-1.39 1.21 0.129 10.6

i3 AP 11 1.69-2.26 2.02 0.159 7.9

WwW 12 1.04-1.43 1.20 0.122 10.1

cl AP 12 2.08-3.07 2.61 0.302 11.6

W 11 1.39-1.82 1.60 0.153 9.6

dpl AP 15 1.37-1.78 1.61 0.110 6.8

W 16 1.21-1.67 1.39 0.118 8.5

p2 AP 9 1.71-1.92 1.84 0.068 3.7

WwW 12 1.44-1.67 1.56 0.074 4.8

dp2. AP 7 1.85-1.99 1.91 0.053 2.8

W 7 1.31-1.53 1.39 0.081 5.8

p3 AP 11 1.78-2.08 1.92 0.105 5.4

WwW 12 1.53-1.85 1.66 0.096 5.8

dp3 AP 7 1.90-2.13 2.04 0.086 4.2

WwW 7 1.30-1.50 1.43 0.068 4.8

p4 AP 17 3.19-3.75 3.43 0.184 5.4

W 18 2.10-2.65 2.38 0.135 5.7

dp4 AP 2 2.98-3.09 3.04 0.055 1.8

WTRI 2 1.56-1.84 1.70 0.140 8.2

WTAL 2 1.72-1.97 1.85 0.125 6.8

ml AP 23 2.65-3.60 3.16 0.193 6.1

WTRI 23 2.18-2.85 2.54 0.173 6.8

WTAL 23 2.15-2.95 2.52 0.183 13

m2 AP 11 2.39-3.34 2.87 0.259 9.0

WTRI 12 2.18-2.91 2.51 0.214 8.5

WTAL 14 1.94-2.82 2.31 0.216 9.3

m3 AP 14 2.39-3.34 2.72 0.241 8.9

WTRI 13 1.75-2.38 2.01 0.170 8.4

on the type maxilla of S. mcgrewi (AMNH 56035) suggest a double-

rooted P2. If substantiated, this condition in Scenopagus would be a

primitive similarity shared with Macrocranion rather than Patrioles-

tes. However, Novacek (1985, table 7) implied that Scenopagus had

a single-rooted P2. The possibility also exists that the most anterior

preserved alveolus in AMNH 560335 is the posterior alveolus for a

20 Stephen L. Walsh

m1 TRIGONID WIDTH (mm)

21 22 23 24 25 26 2.7 2.8 29 3.0

m1 TALONID WIDTH (mm)

x

holotype — X

m2 TRIGONID WIDTH (mm)

2.2 23 24 25 26 2.7 28 2.9

m2 TALONID WIDTH (mm)

1.9 20 2.1

Figure 15. Plots of talonid width vs. trigonid width for m1-2 of Patriolestes novaceki. (A), m1; (B), m2. For definitions of symbols, see Figure 13.

large double-rooted C1, as in Patriolestes. Stucky et al. (1990:177)

indicated that skull material of Scenopagus is available from the Wind

River Formation, so these questions should eventually be resolved.

Finally, characters of the lower dentition assumed for Scenopagus

are based on species other than the type species S. mcgrewi (see

Krishtalka 1976) and should therefore be viewed with caution.

Krishtalka (1976) tentatively considered S. mcgrewi to be a junior

synonym of S. edenensis, but Novacek et al. (1985) retained the

former as a distinct species.

Patriolestes is similar to the middle Eocene to early Oligocene

lipotyphlan Ankylodon in having relatively sectorial molars, anteri-

orly compressed trigonids, and relatively transverse upper molars.

Ankylodon differs from Patriolestes in having a relatively large

double-rooted p3 and a differentiated p4 trigonid, greatly hypertro-

phied m1-3 entoconids, a relatively large P3, a hypocone on P4,

much stronger M1-2 hypocones, and much stronger metastylar and

parastylar lobes on M1 and M2, respectively (Lillegraven et al. 1981,

figs. 49-51). Given Fox’s (1983) reinterpretation of the dental for-

mula of Ankylodon, this genus also differs from Patriolestes in the

subequal size of i3 and cl, the loss of dpl, the greatly enlarged

fanglike I1 and smaller but still fanglike I2, a relatively small single-

rooted Cl, the loss of DP1, and a fenestrated palate.

Patriolestes is similar to the early-to-middle Eocene erinaceo-

morph TJalpavus in having relatively sectorial lower molars with erect

anteroposteriorly compressed trigonids and bladelike paraconids.

TABLE 7. Relative anteroposterior lengths of associated

p4-m3 of Patriolestes novaceki. Length of m1 normalized

to 1.00.

Specimen p4 ml m2 m3

LACM (CIT) 55935 1.01 1.00 — —

LACM (CIT) 55936 1.05 1.00 _ =

LACM (CIT) 55937 — 1.00 0.92 _

SDSNH 49250 1.04 1.00 0.90 0.86

SDSNH 54858 1.11 1.00 0.99 1.00

SDSNH 54912 _ 1.00 0.94 —

SDSNH 55499 1.06 1.00 —

SDSNH 55500 1.06 1.00 _ _—

SDSNH 56947 — 1.00 0.97 0.92

UCMP 133966 1.05 1.00 = =

Talpavus differs from Patriolestes in having a relatively larger later-

ally compressed double-rooted p3, a well-differentiated p4 trigonid,

relatively narrower lower molars (Robinson 1968; Krishtalka 1976,

figs. 9-11), and a p4 shorter than m1 (Table 8). If the upper molars

referred by Krishtalka and Setoguchi (1977, figs. 2C, 2D) to Talpavus

duplus do indeed pertain to that genus, then Talpavus also differs

from Patriolestes in having slightly more bunodont upper molars

with weaker conular wings. Finally, both known species of Talpavus

are much smaller than Patriolestes novaceki (Krishtalka 1976;

Krishtalka and Setoguchi 1977).

Patriolestes is similar to the early Eocene erinaceomorph

Talpavoides in having erect anteroposteriorly compressed m1-2

trigonids, an m1 trigonid subequal in width to the talonid, and a slight

progressive size reduction from m1 to 3 (Bown and Schankler 1982,

plate 6, figs. 3-4). Talpavoides differs from Patriolestes in having a

relatively short transversely compressed p4 with well-developed

trigonid cusps. The only known species of the genus (T. dartoni) is also

much smaller than Patriolestes novaceki. To my knowledge, no upper

teeth or lower teeth anterior to p4 have been reported for Talpavoides.

Patriolestes resembles some Paleocene and Eocene members of

the Erinaceidae (see Novacek et al. 1985, for generic content) in that

some specimens show a slight progressive reduction in molar size

from m1 to m3. Not all specimens of Patriolestes show this trend,

however. Further, early erinaceids such as Entomolestes, Litolestes,

and the unnamed Tepee Trail Formation taxon (AMNH 88288; cf.

Ocajila of McKenna 1980; Novacek et al. 1985:11) differ from

Patriolestes in having a double-rooted p2, a more anteroposteriorly

elongate p3, and differentiated p4 trigonids (Krishtalka and West

1977; Novacek et al. 1985). Litolestes also differs from Patriolestes

in having small bilobate or trilobate lower incisors (Schwartz and

Krishtalka 1976). Lower incisor crowns are apparently unknown for

Entomolestes. Although the early Uintan Tepee Trail Formation taxon

is similar to Patriolestes in having a slightly enlarged 11-2 and a

moderately enlarged bunodont double-rooted C1, the former differs

from Patriolestes in having double-rooted DP1-P2, a strong P4

hypocone, and much more quadrate upper molars with reduced

parastylar and metastylar lobes.

Patriolestes resembles the aberrant Oligocene erinaceoid

Proterix in having a relatively small single-rooted p2-3, a simplified

p4 trigonid, a large double-rooted bunodont C1, and a small triangu-

lar P3 (Gawne 1968; Bjork 1975). Proterix differs from Patriolestes

in having a greatly enlarged il, more bunodont molars, in lacking

Dentition and Skull of Proterixoides and Patriolestes 21

TABLE 8. Statistics for relative anteroposterior lengths of

associated p4—m3 of several Eocene erinaceomorph taxa.

Species p4 ml m2 m3

Patriolestes novaceki®

N 7 10 5) 3

OR 1.01-L.11 — 0.90-0.99 0.86-1.00

M 1.05 1.00 0.94 0.93

Proterixoides davisi@

N 9 18 16 11

OR 1.01-1.11 — 0.89-1.00 0.91-1.06

M 1.06 1.00 0.95 0.98

Scenopagus edenensis”

N 3 6 4 I

OR 0.91-0.95 _ 1.00-1.09 _

M 0.93 1.00 1.04 1.14

Scenopagus curtidens?

N 4 7 5 1

OR 0.83-0.88 _ 0.95-1.00 —

M 0.84 1.00 0.99 0.89

Scenopagus priscus?

6 9 4 2

OR 0.80-0.93 _— 0.93-1.07 0.67-0.93

M 0.88 1.00 1.02 0.80

Macrocranion nitens©

N 6 9 9 3

OR 0.84-0.95 = 0.95-1.05 1.05-1.05

M 0.91 1.00 1.02 1.05

Talpavus cf. T. nitidus4

N 5 6 4 4

OR 0.71-0.86 o 0.85-1.00 0.86-1.00

M 0.81 1.00 0.89 0.93

Talpavoides dartoni®

N 4 5 2 _

OR 0.77-0.92 — 0.88-0.95 _

M 0.82 1.00 0.92 _—

Entomolestes grangeri!

N 6 9 5 1

OR 0.67-0.80 — 0.87-1.00 —

M 0.72 1.00 0.92 0.87

“This study.

bKrishtalka (1976), appendix A.

“Krishtalka (1976), appendix C.

4K rishtalka (1976), table 6.

Bown and Schankler (1982), table 6.

Krishtalka and West (1977), table 1.

hypoconulids, and DP1, in the loss or aberrant structure of P2, and in

having a large hypocone on P4.

Finally, Patriolestes differs from several Paleogene erinaceo-

morphs of uncertain phylogenetic position (Dormaalius, Diacodon,

Eolestes, Litocherus, and Diacocherus; see Gingerich 1983; Novacek

et al. 1985) in that these taxa, where known, all have a double-rooted

p2 and/or p3, and/or differentiated p4 trigonids, and/or relatively

anteroposteriorly elongate bunodont lower molars, and/or single-

rooted upper canines.

EVOLUTIONARY RELATIONSHIPS

I use the term “evolutionary relationship” to include ancestor-

descendant and cladistic relationships. The higher-level relationships

of early erinaceomorphs and other insectivorans are by no means

settled. At the most basic level, there is no consensus on the identity

of the sister group of the Lipotyphla. Novacek (1986a) and MacPhee

and Novacek (1993) favored a sister-group relationship between

lipotyphlans and leptictids, allying them together in a superorder

Insectivora. In contrast, Stucky and McKenna (1993), closely fol-

lowing McKenna (1975), assigned the Leptictidae and their allies to

a separate superorder Leptictida and regarded the Lipotyphla as a

grandorder within the superorder Preptotheria. Under this view, pos-

sible sister taxa of the Lipotyphla might include the Carnivora and

Tubulidentata (MacPhee and Novacek 1993).

More disagreement exists concerning relationships within the

Erinaceomorpha. The most comprehensive treatments available are

those of Novacek et al. (1985) and Novacek (1985). These authors

recognized a basic dichotomy within Erinaceomorpha correspond-

ing to the superfamilies Erinaceoidea and Dormaalioidea (spelled

Dormaaleoidea by Novacek, though based on Dormaalius). Never-

theless, many of the nodes on Novacek’s (1985) cladogram are de-

fined by relatively minor features of the dentition, and significant

changes to this arrangement may be expected as better specimens of

various poorly known taxa are collected.

In a competing hypothesis of lipotyphlan relationships, Butler

(1988) raised the Scenopaginae (containing Scenopagus and Ankylo-

don) to family rank, regarding it as the sister taxon of the Erinaceidae.

Butler also transferred the Sespedectinae from the Dormaaliidae to

the Scenopagidae and tentatively excluded Macrocranion from the

Lipotyphla, greatly altering the concept of the Dormaaliidae envi-

sioned by Novacek et al. (1985) and Novacek (1985). Another clas-

sification of lipotyphlans was proposed by Stucky and McKenna

(1993), who abandoned the family name Dormaaliidae, used the

name Amphilemuridae for Novacek’s (1985) concept of the

Dormaaliinae, raised the Sespedectinae to family rank, and included

within it Scenopagus and Ankylodon. Neither Butler (1988) nor

Stucky and McKenna (1993) recognized the dormaalioid-erinaceoid

dichotomy fundamental to the classifications of Novacek et al. (1985)

and Novacek (1985).

These divergent opinions stem largely from the relatively incom-

plete materials representing many relevant taxa. Since the cladistic

structure of various potential dormaaliid and erinaceomorph

outgroups is debatable, the polarity of many relevant characters is

also uncertain. Although the reality of the dormaalioid-erinaceoid

dichotomy requires additional corroboration, it seems premature to

attempt a comprehensive revised parsimony analysis of the relation-

ships of these taxa. Nevertheless, the following comments are based

on the general classification and proposed character polarities of

Novacek (1985), with some modifications.

Novacek’s (1985, fig. 9) cladogram shows the Dormaaliidae di-

agnosed by the following putative autapomorphies that may separate

them from early erinaceids: (1) p2 small and procumbent or peglike

and single-rooted; (2) p4 with short bicuspid or unicuspid heel; (3)

P2 peglike and single-rooted; (4) M1-2 hypocones relatively large.

Novacek et al. (1985) and Novacek (1985) noted that this diagnosis

is problematical in that some of these characters also occur in certain

putative Paleogene erinaceids (e. g., a short p4 talonid is present in

Entomolestes). In addition, the upper dentition is unknown for many

of these taxa (e. g., Entomolestes, Leipsanolestes, Eolestes, and

Dartonius). Note that character (3) should not be regarded as an

autapomorphy for the Dormaaliidae because a double-rooted P2 is

retained in Macrocranion (see Novacek et al. 1985, fig. 1),

Pholidocercus (see von Koenigswald and Storch 1983), and possibly

Scenopagus (see McKenna and Simpson 1959). Accordingly, a

double-rooted P2 is included in a list of proposed dormaaliid

plesiomorphies (Fig. 16).

Sespedectinae

The moderately enlarged upper and lower first and second inci-

sors of Proterixoides somewhat resemble the enlarged incisors of

several extant erinaceid genera (e.g., Frost et al. 1991). However,

incisor enlargement is prone to parallelism within the Lipotyphla and

other eutherians (Butler 1980: 173). Thus, the incipiently diprotodont

condition of the Sespedectinae seems to be insufficient evidence for

aclose relationship with the Erinaceidae, and the former subfamily is

Nw

Ne

tentatively retained in the Dormaaliidae. Consistent with this deci-

sion is the apparent lack of incisor enlargement in the Paleogene

erinaceids Entomolestes and Litolestes (as noted, the Tepee Trail

Formation taxon had slightly enlarged il-2, but the upper incisors

are apparently unknown).

Robinson (1966:29) suggested that Scenopagus is a logical an-

cestor for Proterixoides. Although this remains a possibility, the im-

mediate ancestor of both Proterixoides and Sespedectes was prob-

ably Crypholestes (cf. Novacek 1976:24). Crypholestes has some-

what more sectorial molars than Proterixoides or Sespedectes, which

helps to bridge the morphological gap between the latter two genera

and Scenopagus. Nevertheless, derivation of Crypholestes from

Scenopagus requires that the very small P3 of the latter genus is

primitive, capable of evolving into the large P3 of sespedectines (see

further discussion below).

Another possible ancestral morphotype for the Sespedectinae is

represented by the Paleocene erinaceomorph Litocherus (see

Gingerich 1983). As noted by Novacek (1976:24), the upper cheek

teeth of Litocherus notissimus are remarkably similar in morphology

to those of Crypholestes vaughni (see Simpson 1936:24). The p4 of

Litocherus is also similar to that of sespedectines in being slightly

longer than m1, although p3-4 of Litocherus have sharper cusps than

p3-4 of sespedectines (Gingerich 1983, fig. 1). According to Simpson

(1936:25), (D)P1 and P2 of L. notissimus have relatively small,

simple crowns (apparently like sespedectines) but are both double-

SESPEDECTINAE

Sespedectes Proterixoides Crypholestes Patriolestes

8b

7b

2b

6b

10c

Mia eens Fl

Peed Noe ele

10b

other

erinaceomorphs

Stephen L. Walsh

rooted (unlike sespedectines). An important difference between these

taxa is the large recurved single-rooted C1 and large c1 in Litocherus

(Simpson 1936, 1937), as opposed to the small double-rooted C1 and

small cl in sespedectines.

Patriolestes

Patriolestes has all three of the characters that currently diagnose

the Dormaaliidae and is thus assigned to this family. Novacek (1985)

recognized three subfamilies of the Dormaaliidae (Sespedectinae,

Scenopaginae, and Dormaaliinae). Patriolestes can be excluded from

the Sespedectinae on the basis of its enlarged cl and C1, its relatively

small usually single-rooted p3, and its very small P3.

Existing diagnoses of the Scenopaginae and Dormaaliinae are

not entirely satisfactory. The Scenopaginae (containing Scenopagus

and Ankylodon) were diagnosed as having large, distinct P4

hypocones and large and posterolingually situated hypocones on M1-

2 (Novacek 1985, fig. 9). However, the M1-2 hypocones of

Scenopagus are approximately as large, relatively, and posterolin-

gually situated as those of Macrocranion (e. g., Krishtalka 1976, figs.

4C and 7C; Bown and Schankler 1982, plate 5) and do not approach

the hypertrophied condition of Ankylodon (Lillegraven et al. 1981,

fig. 49). The Dormaaliinae were diagnosed by Novacek (1985, fig. 9)

as having the derived characters of a small single-rooted p3 and a

very small peglike or triangular P3. However, Scenopagus also has a

SCENOPAGINAE DORMAALIINAE

cae Ema

Scenopagus Ankylodon Amphilemurini | Macrocranion

* 8d

10c

2b

* Ankylodon autapomorphies:

- DP1/dp1 lost

- 11-2 fanglike

8c - M1-2 hypocones v. large

- v. large meta- and parastylar

lobe on M1-2, respectively

- fenestrated palate

DORMAALIIDAE- putative autapomorphies:

- p2 small and procumbent, or peglike and 1-rooted

- p4 with short bicuspid or unicuspid heel

- M1-2 hypocones relatively large

Dormaaliid plesiomorphies:

Characters 1-10 have state "a".

ODBNOnNSWNH

. 11-2/1-2: a. Relatively small, unspecialized. b. Moderately enlarged.

. Lower and upper molars: a. Relatively sectorial. b. Relatively bunodont.

. ¢1: a. similar in size to i3 and dp1-2. b. Much larger than i3 and dp1-2.

. p3: a. double-rooted, subequal to p4. b. usually single-rooted, much smaller than p4.

. p4 length: a. Slightly less than m1. b. Slightly more than m1.

. p4 trigonid: a. With distinct metaconid, protoconid, and/or paraconid. b. Undifferentiated, bulbous.

. m1 trigonid usually: a. Narrower than m1 talonid. b. Wider than m1 talonid.

. C1: a. Small, double-rooted. b. Large, double-rooted. c. Small, single rooted. d. Large, cliglowcele

. P2: a. Small, double-rooted. b. Small, single-rooted.

10. P3: a. Moderately smaller than P4. b. Much smaller than P4. c. Subequal to P4.

Figure 16. Cladogram showing distribution of selected morphological characters ot various dormaalud taxa. Polarities and topology modified from

Novacek (1985).

Dentition and Skull of Proterixoides and Patriolestes

very small, triangular P3, yet its alleged sister taxon Ankylodon has a

P3 nearly as large as P4 (Lillegraven et al. 1981, fig. 49). Indeed, the

P3 in the holotype specimen of Scenopagus mcgrewi (McKenna and

Simpson 1959) and one specimen of S. curtidens (Bown 1982, plate

2B) are just as small relative to P4 as the P3 in one skull of

Macrocranion nitens (USGS 3676; Novacek et al. 1985, fig. 2).

Novacek (1985, table 7) suggested that a P3 “small to moderate,

triangular, at least 2-rooted,” is primitive for erinaceomorphs (e.g.,

Diacodon; see Novacek 1982) and that the states “very small, peglike

or triangular” and “large with several cusps, similar to P4” are both

derived. I tentatively follow this proposal here, but note that there is

considerable doubt concerning the polarities of the various states of

the character “P3 size relative to P4” (cf. Fox 1983:975).

Patriolestes lacks a hypocone on P4, which might exclude it from

the Scenopaginae. However, the P4 hypocones on the type specimen

of Scenopagus mcgrewi (AMNH 56035) and a referred specimen of S.

edenensis (CM 6433; Krishtalka 1976, fig. 4C) are quite small, so the

absence of this cusp in Patriolestes need not bar a close relationship

between the two genera. Like Scenopagus and Ankylodon, Patriolestes

has relatively sectorial molars, anteroposteriorly compressed trigonids

with bladelike paraconids, and relatively tranverse M1-2s. Although

these characters may be primitive for the Dormaaliidae (Novacek et al.

1985:3-4), they seem to exclude Patriolestes from the Dormaaliinae,

which are characterized by cheek teeth with more bulbous cusps. How-

ever, Patriolestes also has a very small usually single-rooted p3, which

is diagnostic of the Dormaaliinae (Novacek 1985, fig. 9). In view of

this equivocal evidence (Fig. 16), an assignment of Patriolestes to ei-

ther the Scenopaginae or Dormaaliinae is unjustified, and I prefer to

classify it as Dormaaliidae, incertae sedis.

Patriolestes is unique among known dormaaliids in having a rela-

tively large double-rooted bunodont C1, and in this respect it re-

sembles the aberrant Oligocene erinaceoid Proterix (see Matthew

1903; Gawne 1968; Bjork 1975). Novacek (1985:19) implied that

the large double-rooted C1 in Proterix was more “conservative” than

the small single-rooted C1 in sespedectines (as noted above, C1 in

sespedectines is actually two-rooted, with frequent incomplete fu-

sion). The issue of C1 character-state polarity is further complicated

by the presence of a large single-rooted Cl in the Paleocene

erinaceomorph Litocherus (see Simpson 1936, 1937) and the early

Eocene putative dormaaliid Macrocranion (see Novacek et al. 1985,

fig. 1). Although many of the characters in which Patriolestes differs

from Proterix are clearly primitive, the similarity in C1 morphology

shared by the two genera seems to be convergent in view of the pre-

sumed erinaceoid status of Proterix.

Given the limited material of Patriolestes then available, Novacek

(1982) suggested it to be the sister taxon of Ankylodon, on the basis of

the enlarged M1-2 hypocones and enlarged M1 metastylar spurs al-

legedly shared by these taxa. However, the new specimens of

Patriolestes described above generally do not have M1-2 hypocones

or the M1 metastylar lobes significantly enlarged relative to those in at

least some specimens of Macrocranion (e. g., Krishtalka 1976, fig.

7C) and Scenopagus (e. g., Krishtalka 1976, fig. 4C). Although UCMP

101420 does have a relatively large metastylar lobe (Novacek 1976,

fig. 14), this structure is not significantly expanded in most other avail-

able M1s of Patriolestes (e. g., SDSNH 37999 and 54857). The

parastylar lobe of M2 of Patriolestes is also variable in its degree of

expansion (e.g., SDSNH 38004 vs. 54857).

To complicate matters, Fox (1983) described the upper anterior

dentition and portions of the skull of Ankylodon and concluded it is

actually a soricomorph insectivore, rather than an erinaceomorph as

generally supposed (e.g., Lillegraven et al. 1981). With the exception

of the fanglike I1, Novacek et al. (1985) questioned the validity of

the skull characters in Ankylodon cited by Fox to indicate soricomor-

phan relationships. Novacek et al. (1985) and Novacek (1985) there-

fore retained Ankylodon in the Erinaceomorpha and included it with

Scenopagus in the subfamily Scenopaginae. Butler (1988) did not

nN

Ww

comment on Fox’s classification of Ankylodon but supported an an-

cestor-descendant relationship between Ankylodon and Scenopagus.

Since the higher-level position of Ankylodon has not been settled,

I address only the possibility of its being a direct descendant of

Patriolestes. Evolution of Ankylodon from Patriolestes requires sey-

eral character reversals, including a re-enlargement of the small usu-

ally single-rooted p3 into a large double-rooted tooth, a re-differen-

tiation of the bulbous p4 trigonid, a re-enlargement of the small P3,

and a re-reduction of the large double-rooted C1 into a small single-

rooted tooth. Butler’s (1988) derivation of Ankylodon from Scenopa-

gus also requires a re-enlargement of the small P3 in the latter genus.

On the other hand, Patriolestes is a plausible descendant of

Scenopagus in that where known, Scenopagus retains the primitive

condition of the derived character states of Patriolestes (excepting

the small P4 hypocone). Discovery of the anterior dentitions and

upper teeth of Scenopagus, Talpavus, and Talpavoides should clarify

the relationships of these and other dormaaliid taxa.

ACKNOWLEDGMENTS

Collection of the specimens discussed herein was made possible by the

cooperation and financial support of the Buie Corporation, Carmel Mountain

Ranch, Gatlin Development, McMillin Communities, Mission Terrace Asso-

ciates, Pardee Construction, Shea Homes, and the California Department of

Transportation. R. A. Cerutti, B. O. Riney, and R. L. Clark prepared most of

the delicate jaw and skull material. J. D. Archibald provided access to a mea-

suring microscope. I thank J. D. Archibald (SDSU), S. McLeod (LACM), and

J. H. Hutchison and P. Holroyd (UCMP) for the loan of specimens and casts.

I thank T. A. Deméereé for helpful discussions of mammalian cranial anatomy

and him and P. Unitt for editorial improvements. J. A. Lillegraven and M. J.

Novacek reviewed the manuscript and made helpful suggestions. I thank M.

C. McKenna for providing illustrations of the dentition of the Tepee Trail

Formation erinaceid, and McKenna and Lillegraven for access to their manu-

script in progress on the relationships of Ankylodon.

LITERATURE CITED

Bjork, P. R. 1975. Observations on the morphology of the hedgehog ge-

nus Proterix (Insectivora: Erinaceidae). University of Michigan Pa-

pers on Paleontology 12:81-88.

Bown, T. M. 1979. Geology and mammalian paleontology of the Sand

Creek facies, lower Willwood Formation (lower Eocene), Washakie

County, Wyoming. Geological Survey of Wyoming Memoir 2.

Bown, T. M. 1982. Geology, paleontology, and correlation of Eocene

volcaniclastic rocks, southeast Absaroka Range, Hot Springs County,

Wyoming. United States Geological Survey Professional Paper

1201-A.

Bown, T. M., and D. Schankler. 1982. A review of the Proteutheria and

Insectivora of the Willwood Formation (lower Eocene), Bighorn

Basin, Wyoming. United States Geological Survey Bulletin 1523.

Butler, P. M. 1948. On the evolution of the skull and teeth in the

Erinaceidae, with special reference to fossil material in the British

Museum. Proceedings of the Zoological Society of London 118:446-

500.

Butler, P. M. 1972. The problem of insectivore classification. Pp. 253-

265 in K. A. Joysey and T. S. Kemp (eds.). Studies in Vertebrate

Evolution. Oliver and Boyd, Edinburgh.

Butler, P. M. 1980. The tupaiid dentition. Pp. 171-204 in W. P. Luckett

(ed.). Comparative Biology and Evolutionary Relationships of Tree

Shrews. Plenum, New York.

Butler, P. M. 1988. Phylogeny of the insectivores. Pp. 117-141 in M. J.

Benton (ed.). Phylogeny and Classification of the Tetrapods, Vol-

ume 2: Mammals. Systematics Association Special Vol. 35B.

Clarendon Press, Oxford, England.

Fox, R. C. 1983. New evidence on the relationships of the Tertiary

insectivoran Ankylodon (Mammalia). Canadian Journal of Earth

Sciences 20:968-977.

Frost, D. R., W. C. Wozencraft, and R. S. Hoffmann. 1991. Phylogenetic

relationships of hedgehogs and gymnures (Mammalia: Insectivora:

Erinaceidae). Smithsonian Contributions to Zoology 518.

24 Stephen L. Walsh

Gawne, C. E. 1968. The genus Proterix (Insectivora, Erinaceidae) of the

upper Oligocene of North America. American Museum Novitates

2315.

Gingerich, P. D. 1983. New Adapisoricidae, Pentacodontidae, and

Hyopsodontidae (Mammalia, Insectivora and Condylarthra) from

the late Paleocene of Wyoming and Colorado. Contributions from

the Museum of Paleontology, University of Michigan 26:227-255.

Golz, D. J., and J. A. Lillegraven. 1977. Summary of known occurrences

of terrestrial vertebrates from Eocene strata of southern California.

University of Wyoming Contributions to Geology 15:43-65.

Kelly, T. S. 1990. Biostratigraphy of Uintan and Duchesnean land mam-

mal assemblages from the middle member of the Sespe Formation,

Simi Valley, California. Natural History Museum of Los Angeles

County Contributions in Science 419.

Kelly, T. S., E. B. Lander, D. P. Whistler, M. A. Roeder, and R. E.

Reynolds. 1991. Preliminary report on a paleontologic investigation

of the lower and middle members, Sespe Formation, Simi Valley

Landfill, Ventura County, California. PaleoBios 13 (50):1-13.

Kennedy, M. P., and G. W. Moore. 1971. Stratigraphic relations of Up-

per Cretaceous and Eocene formations, San Diego coastal area,

California. American Association of Petroleum Geologists Bulle-

tin 55:709-722.

Kennedy, M. P., and G. L. Peterson. 1975. Geology of the Eastern San

Diego Metropolitan area, California. California Division of Mines

and Geology Bulletin 200-B:43-56.

Koenigswald, W. von, and G. Storch. 1983. Pholidocercus hassiacus, ein

Amphilemuride aus dem Eozan der “Grube Messel” bei Darmstadt

(Mammalia, Lipotyphla). Senckenbergiana Lethaea 64:447-495.

Krishtalka, L. 1976. Early Tertiary Adapisoricidae and Erinaceidae

(Mammalia, Insectivora) of North America. Carnegie Museum of

Natural History Bulletin 1.

Krishtalka, L., and T. Setoguchi. 1977. Paleontology and geology of the

Badwater Creek area, central Wyoming. Part 13. The late Eocene

Insectivora and Dermoptera. Annals of Carnegie Museum 46:71-

99!

Krishtalka, L., and R. K. Stucky. 1984. Middle Eocene marsupials

(Mammalia) from northeastern Utah and the mammalian fauna from

Powder Wash. Annals of Carnegie Museum 53:31-45.

Krishtalka, L., and R. M. West. 1977. Paleontology and geology of the

Bridger Formation, southern Green River Basin, southwestern

Wyoming. Part 2. The Bridgerian insectivore Entomolestes gran-

geri. Milwaukee Public Museum Contributions in Biology and

Geology 14.

Krishtalka, L., R. M. West, C. C. Black, M. R. Dawson, J. J. Flynn, W. D.

Turmbull, R. K. Stucky, M. C. McKenna, T. M. Bown, D. J. Golz, and

J. A. Lillegraven. 1987. Eocene (Wasatchian through Duchesnean)

biochronology of North America. Pp. 77-117 in M. O. Woodburne

(ed.), Cenozoic Mammals of North America, Geochronology and Bio-

stratigraphy. University of California Press, Berkeley.

Lillegraven, J. A.,M. C. McKenna, and L. Krishtalka. 1981. Evolutionary

relationships of middle Eocene and younger species of Centetodon

(Mammalia, Insectivora, Geolabididae), with a description of the

dentition of Ankylodon (Adapisoricidae). University of Wyoming

Publications 45:1-115.

Lucas, S. G. 1992. Redefinition of the Duchesnean land mammal “age,”

late Eocene of western North America. Pp. 88-105 in D. R. Prothero

and W. A. Berggren (eds.), Eocene-Oligocene Climatic and Biotic

Evolution. Princeton University Press, Princeton, N. J.

MacPhee, R. D. E., and M. J. Novacek. 1993. Definition and relationships

of Lipotyphla. Pp. 13-31 in F. S. Szalay, M. J. Novacek, and M. C.

McKenna (eds.). Mammal Phylogeny: Placentals. Springer-Verlag,

New York.

MacPhee, R. D. E., M. J. Novacek, and G. Storch. 1988. Basicranial

morphology of early Tertiary erinaceomorphs and the origin of Pri-

mates. American Museum Novitates 2921.

Maier, W. 1979. Macrocranion tupaiodon, an adapisoricid (?) insectivore

from the Eocene of the “Grube Messel” (Western Germany).

Palaontologische Zeitschrift 53:38-62.

Matthew, W. D. 1903. A fossil hedgehog from the American Oligocene.

American Museum of Natural History Bulletin 19:227-229.

McGrew, P. O., and R. Sullivan, 1970. The stratigraphy and paleontology

of Bridger A. University of Wyoming Contributions to Geology

9:66-85.

McKenna, M. C. 1975. Toward a phylogenetic classification of the

Mammalia. Pp. 21-46 in W. P. Luckett and F. S. Szalay (eds.). Phy-

logeny of the Primates. Plenum, New York.

McKenna, M. C. 1980. Late Cretaceous and early Tertiary vertebrate pa-

leontological reconnaissance, Togwotee Pass area, northwestern

Wyoming. Pp. 321-343 in L. L. Jacobs (ed.). Aspects of Vertebrate

History: Essays in honor of Edwin Harris Colbert. Museum of North-

em Arizona Press, Flagstaff.

McKenna, M. C., and G. G. Simpson. 1959. A new insectivore from the

middle Eocene of Tabernacle Butte, Wyoming. American Museum

Novitates 1952.

Novacek, M. J. 1976. Insectivora and Proteutheria of the later Eocene

(Uintan) of San Diego County, California. Natural History Museum

of Los Angeles County Contributions in Science 283.

Novacek, M. J. 1980. Crypholestes, a new name for the early Tertiary

insectivore (Mammalia) Cryptolestes Novacek 1976. Journal of Pa-

leontology 54:1135.

Novacek, M. J. 1982. Diacodon alticuspis, an erinaceomorph insectivore

from the early Eocene of northern New Mexico. University of Wyo-

ming Contributions to Geology 20:135-149.

Novacek, M. J. 1985. The Sespedectinae, a new subfamily of hedgehog-

like insectivores. American Museum Novitates 2822.

Novacek, M. J. 1986a. The skull of leptictid insectivorans and the higher-

level classification of eutherian mammals. American Museum of

Natural History Bulletin 183:1-112.

Novacek, M. J. 1986b. The primitive eutherian dental formula. Journal of

Vertebrate Paleontology 6:191-196.

Novacek, M. J., T. M. Bown, and D. M. Schankler. 1985. On the classifica-

tion of the early Tertiary Erinaceomorpha (Insectivora: Mammalia).

American Museum Novitates 2813.

Prothero, D. R. 1996. Magnetic stratigraphy and biostratigraphy of the

middle Eocene Uinta Formation, Uinta Basin, Utah. Pp. 3-24 in D.

R. Prothero and R. J. Emry (eds.), The Terrestrial Eocene-Oligocene

Transition in North America. Cambridge University Press, Cam-

bridge, England.

Rich, T. H. V. 1981. Origin and history of the Erinaceinae and Brachyeri-

cinae (Mammalia, Insectivora) in North America. American Mu-

seum of Natural History Bulletin 171:1-116.

Rich, T. H. V., and P. V. Rich. 1971. Brachyerix, a Miocene hedgehog

from western North America, with a description of the tympanic

regions of Paraechinus and Podogymnura. American Museum

Novitates 2477.

Robinson, P. 1966. Fossil Mammalia of the Huerfano Formation, Eocene,

of Colorado. Peabody Museum of Natural History Bulletin 21.

Robinson, P. 1968. Talpavus and Entomolestes (Insectivora, Adapisori-

cidae). American Museum Novitates 2339.

Schwartz, J. H., and L. Krishtalka. 1976. The lower antemolar teeth of

Litolestes ignotus, a late Paleocene erinaceid (Mammalia, Insecti-

vora). Annals of Carnegie Museum 46:1-6.

Simpson, G. G. 1936. A new fauna from the Fort Union of Montana.

American Museum Novitates 873.

Simpson, G. G. 1937. Additions to the upper Paleocene fauna of the Crazy

Mountain field. American Museum Novitates 940.

Stock, C. 1935. Insectivora from the Sespe uppermost Eocene, Califor-

nia. National Academy of Sciences Proceedings 21:214-219.

Storer, J. E. 1996. Eocene-Oligocene faunas of the Cypress Hills Forma-

tion, Saskatchewan. Pp. 240-261 in D. R. Prothero and R. J. Emry

(eds.). The Terrestrial Eocene-Oligocene Transition in North

America. Cambridge University Press, Cambridge, England.

Stucky, R. K., L. Krishtalka, and A. D. Redline. 1990. Geology, verte-

brate fauna, and paleoecology of the Buck Spring Quarries (early

Eocene), Wind River Formation, Wyoming. Pp. 169-186 in T. M.

Bown and T. K. Rose (eds.). Dawn of the Age of Mammals in the

Northern Part of the Rocky Mountain Interior, North America. Geo-

logical Society of America Special Paper 243.

Stucky, R. K., and M. C. McKenna. 1993. Mammalia. Pp. 739-771 in M.

Benton (ed.). The Fossil Record I. Chapman and Hall, London.

Dentition and Skull of Proterixoides and Patriolestes 25

Walsh, S. L. 1996. Middle Eocene mammal faunas of San Diego County,

California, Pp. 75-119 in D. R. Prothero and R. J. Emry (eds.). The

Terrestrial Eocene-Oligocene Transition in North America. Cam-

bridge University Press, Cambridge, England.

Walsh, S. L. In press. Fossil datum and paleobiological event terms,

paleontostratigraphy, chronostratigraphy, and the definition of Land

Mammal “Age” boundaries. Journal of Vertebrate Paleontology.

Walsh, S. L., D. R. Prothero, and D. J. Lundquist. 1996. Stratigraphy and

paleomagnetism of the middle Eocene Friars Formation and Poway

APPENDIX 1.

ALL KNOWN SPECIMENS OF PATRIOLESTES NOVACEKI,

ARRANGED BY ELEMENT.

i1(?): SDSNH 27600, 32487, 58729, 58919, 58980, 58997,

60008, 60024.

i2: SDSNH 56968, 58732, 58788, 58789, 58807, 58826, 58964,

58998, 60013, 60037.

di2(?): SDSNH 58731, 58810, 58812, 58925, 58933, 58965.

i3: SDSNH 27628, 56970, 56971, 58730, 58825, 58921, 58966,

58981, 58983, 58984, 58990, 60019.

di3: SDSNH 37528, 58811, 58824, 58920, 58926, 58932, 58982,

58999, 58797, 58813, 58967.

cl: SDSNH 31922, 31936, 48043, 56967, 58733-58735, 58814,

58827, 58828, 58968, 58969, 60004, 60038, 60039.

dp1: SDSNH 31883, 31913, 31934, 31935, 32000, 56964, 58641,

58736-58739, 58798, 58816, 58936, 58985, 58986, 60040.

p2: SDSNH 31876, 31990, 56966, 58640, 58720, 58740-58742,

58790, 58829, 58836, 58937, 60014. UCMP 154722.

dp2: SDSNH 31983, 32006, 58722, 58745, 58746, 58934, 58970,

58971, 59000.

p3: SDSNH 32004, 32080, 58747-749, 58791, 58830, 58837,

58927, 58972, 60041.

dp3: SDSNH 58723, 58743, 58744, 58792, 58808, 58935, 60026,

60042.

p4: LACM(CIT) 55934. SDSNH 37609, 37987, 37988, 39679,

42428, 45160, 45799, 46436, 47240, 49612, 55825, 60043.

dp4: SDSNH 45798, 56946.

m1: SDSNH 31764, 37610, 37992, 45161, 45800, 46271, 49207,

49436, 50589, 51259, 54926, 54995, 60044, 60045.

m1 or 2: SDSNH 37989, 37991, 39680.

m2: SDSNH 32039, 37611, 37993, 42429, 43230, 51069, 54996,

56717, 60046. UCMP 101617.

m3: LACM(CIT) 56189. SDSNH 37612, 37613, 37994, 37995,

39678, 39681, 45801, 46272, 46555, 46556, 49611, 56724, 56948,

56949, 60047. UCMP 109514, 109536.

mandible fragments with (or associations of):

p3-4: SDSNH 56725.

unerupted p3 + m1-2: SDSNH 54912.

p4: SDSNH 55498.

p4-m1: LACM(CIT) 55935, 55936. UCMP 133966.

p4-m2: SDSNH 55499, 55500.

p4-m3: SDSNH 49250, 54858.

ml or 2: SDSNH 46554.

m2: SDSNH 43276.

Group, southwestern San Diego County, California. Pp. 120-154 in

D. R. Prothero and R. J. Emry (eds.). The Terrestrial Eocene-Oli-

gocene Transition in North America. Cambridge University Press,

Cambridge, England.

Wilson, K. L. 1972. Eocene and related geology of a portion of the San

Luis Rey and Encinitas quadrangles, San Diego County, California.

M. S. thesis, University of California, Riverside.

Ziegler, A. C. 1971. Dental homologies and possible relationships of re-

cent Talpidae. Journal of Mammalogy 52:50-68.

m1-3: LACM(CIT) 55937. SDSNH 56947.

m2-3: SDSNH 47865.

I1: SDSNH 31979, 56972, 58726, 58727, 58757, 58941, 58943,

58973, 58992, 60005, 60015, 60025, 60027, 60048-60050. UCMP

154721.

12: SDSNH 37512, 43251, 56714, 56969, 58796, 58809, 5883 1-

58833, 60017.

DI1-2 (?): 32491, 32494, 58725, 58751, 58753, 58755, 58793,

58799, 58817, 58930, 58938, 60007, 60017.

13: SDSNH 31920, 31929, 37523, 42484, 56973, 58719, 58752,

58760, 58761, 58794, 58805, 58806, 58821, 58834, 58923, 58929,

58939, 58942, 58974-58977, 58991, 58993, 60009, 60010, 60020,

60022, 60051.

DI3 (?): 32500, 37506, 37511, 37530, 58750, 58754, 58756,

58818, 58922, 58928, 58987-58989, 58994-58996, 60001, 60006,

60011, 60021, 60028.

C1: SDSNH 27650, 31998, 43250, 51367, 58758, 56960-56963,

58724, 58800, 58924, 60002.

DP1: SDSNH 27587, 58819, 58820, 60003, 60012, 60018,

60023.

P2: SDSNH 32050, 58762-58766, 58660, 58795, 58801-58804,

58823, 58931, 58979, 60052. UCMP 109546, 110041.

P3: SDSNH 38122, 38149, 39747, 39761, 43083, 43963, 46139,

56951-56953, 56965, 60053.

DP3: SDSNH 27011, 38150, 39748, 56950, 58721, 58823,

58835.

P4: LACM(CIT) 55933, 56119. SDSNH 37614, 43081, 51258,

56954, 56955, 59637, 58638, 60054. UCMP 101690, 106355.

DP4: UCMP 154720.

M1: SDSNH 31767, 36578, 37998-38001, 38003, 39335, 39683,

39684, 43082, 43231, 45802, 46373, 49350, 51070, 51102, 55679,

56718, 56956, 58639, 58768, 60055, 60056, 60057. UCMP 101420,

133967, 154719.

M1 or 2: LACM(CIT) 55945. SDSNH 45163, 60058. UCMP

101654, 101626.

M2: LACM(CIT) 55944. SDSNH 31768, 37615, 37616, 38002,

38004, 39682, 42430, 45552, 46066, 46273, 56957, 58769.

M3: SDSNH 31765, 31766, 36579, 37400, 38005, 38006, 39685,

41353, 45162, 46274, 51103. UCMP 101497, 109796.

palatal fragment with LC1-P2 + RP2: SDSNH 55497.

maxillary fragments with:

Cl + P3-4: SDSNH 56185.

P3-M2: SDSNH 54857.

P4-M1: SDSNH 51293.

26 Stephen L. Walsh

APPENDIX 2.

ALL KNOWN SPECIMENS OF PATRIOLESTES NOVACEKI,

ARRANGED BY STRATIGRAPHIC UNIT, INSTITUTION,

AND LOCALITY.

Friars Formation, undifferentiated. -SDSNH Loc. 3414: 41353.

Loc. 3430: 58942. Loc. 3784: 49207, 49250, 49350, 49436, 49611,

49612, 50589, 51258, 51259, 51293, 56724, 56725. Loc. 3785:

50786, 51069, 51070, 51102, 51103, 58810-58822. Loc. 3832:

51367.

LACM (CIT) Loc. 249: LACM(CIT) 55933, 55934. 55935,

55936, 55937, 55944, 55945, 55949. Loc. 249-S: 56119. Loc. 249-

P: 56189.

UCMP Loc. V-68116: UCMP 101690.

Upper tongue of Friars Formation —SDSNH Loc. 3254:

SDSNH 27011, 27628, 27650, 27600, 27623, 27587. Loc. 3373:

31764-31768, 31876, 31883, 31885, 31913, 31920, 31922, 31924,

31929, 31934-31936, 31964, 31979, 31983, 31990, 31996, 31998,

32000, 32004, 32006, 32039, 32050, 32080, 60013-60016. Loc.

3380: 42428, 42430, 58719, 58720, 60017. Loc. 3391: 32487, 32491,

32494, 32500, 56717, 56718. Loc. 3482: 37609-37616, 58722-

58728, 60018. Loc. 3483: 37987-37989, 37991-37995, 37998-

38006, 38149, 38150, 58729-58770, 58943, 60019-60023, 60028.

Loc. 3484: 58776-58777. Loc. 3591: 43963. Loc. 3611: 45160-

45163, 58788-58796, 60024-60025. Loc. 3612: 45552, 47865,

58797-58806. Loc. 3681: 46554-46556, 58807-58809. Loc. 3685:

46436. Loc. 3771: 47240. Loc. 3883: 54857, 54858, 60026-60027.

UCMP Loc. V-71180: 101654. V-72157: 101420, 106355,

109514, 109536, 109546, 110041. V-72158: 101497, 101617,

101626, 154719. V-73138: 109796, 133966, 133967, 154720-

154722.

Conglomerate tongue of Friars Formation.—SDSNH Loc. 3615:

42484. Loc. 3616: 60004. Loc. 3617: 43081-43083, 43276, 58824-

58837, 60005-60007. Loc. 3620: 43230-43231, 43250-43251. Loc.

3621: 55497-55500, 55825, 56946-56973, 58637-58642, 58659-

58660, 60008-60012. Loc. 3730: 48043. Loc. 3739: 54912, 54926.

Loc. 3824: 56185, 60037-60058.

Lower tongue of Friars Formation. —SDSNH Loc. 3494: 37400.

Loc. 3496: 37506, 37511-37513, 37523, 37528, 37530. Loc. 3503:

39335, 58990-58991. Loc. 3505: 39678-39685, 39747, 39748,

39761, 58964-58989. Loc. 3655: 46066, 46139, 58919-58924, Loc.

3656: 46271-46274, 58925-58931, 58992-58994. Loc. 3657:

46373. Loc. 3658: 45798-45802, 58932-58941, 58995-58996. Loc.

3828: 56714. Loc. 3893: 54995-54996, 58823, 58997-59000,

60001-60003.

Member “B” of Santiago Formation.—SDSNH Loc. 3440:

36576-36579.

n

. ae ia e Tome

: “. Sue _ ek

Qided- Vn ‘ Ay

he iia) Weare

| re nis

asain Gs

j Lea, 4“1F 4406-4

~~

‘ <i

ore 41 MET - thas

vay AA fines” OM Mae

PLL, 1

a oni OR

¥ rm =r

Be |

‘ug cole ae

Hy) Sy rah hes

ul ery)

o

as