Texas 
Memorial 
Museum 
Bulletin 
33 


CARNIVOROUS 
MAMMALS 
OF 
THE 
LATE 
EOCENE 
AND 
EARLY 
OLIGOCENE 
OF 
TRANS-PECOS 
TEXAS 


Eric 
Paul 
Gustafson 


•

The 
University 
of 
Texas 
at 
Austin 
2400 
Trinity 
• 
Austin, 
TX 
78705 



© 
1986 
by 
Texas 
Memorial 
Museum 
The 
University 
of 
Texas 
at 
Austin 


All 
rights 
reserved 
Published 
December 
15, 
1986 
Printed 
in 
the 
United 
States 
of 
America 


ISSN 
0082-3074-33 


The 
Bulletin 
is 
an 
irregularly 
published 
series 
of 
technical 
monographs 
deriving 
primarily 
from 
research 
done 
on 
Texas 
Memorial 
Museum 
collections 
and 
projects. 



CONTENTS 


LIST 
OF 
FIGURES 
LIST 
OF 
TABLES 
ABSTRACT 
INTRODUCTION 
ACKNOWLEDGEMENTS 
TERMINOLOGY 
GEOGRAPHIC 
SETTING 
PREVIOUS 
WORK 
STRATIGRAPHY 
CLASSIFICATION 
SYSTEMATIC 
PALEONTOLOGY 
ORDER 
CONDYLARTHRA 
ORDER 
CREODONTA 
ORDER 
CARNIVORA 
CORRELATIONS 
SUMMARY 
AND 
CONCLUSIONS 
BIBLIOGRAPHY 


iv 
v 
1 
1 
2 
3 
3 
4 
4 
7 
8 
8 
14 
36 
58 
61 
63 


III 



FIGURES 


1. 
Dental 
terminology 
3 


2. 
Dental 
terminology 
3 


3. 
Map 
ofBrewster 
and 
Presidio 
Counties, 
showing 
fossil 
localities 
5 


4. 
Diagrammatic 
stratigraphic 
section 
of 
the 
Buck 
Hill 
Group 
6 


5. 
Harpagolestes 
cf. 
H. 
uintensismandible 
10 


6. 
Harpagolestes 
sp., 
maxilla 
and 
. 
upper 
dentition 
11 
1

7. 
Hessolestes 
sp., 
mandible 
and 
partial 
P4-M13 


8. 
IProviverra 
major, 
mandible 
15 


9. 
Hemipsalodon 
viejaensis 
n. 
sp., 
skull 
17 


10. 
Hemipsalodon 
viejaensis 
n. 
sp., 
skull 
and 
mandible, 
lateral 
view 
17 


11. 
Hemipsalodon 
viejaensis 
n. 
sp., 
skull, 
posterior 
view 
18 


12. 
Hemipsalodon 
viejaensis 
n. 
sp., 
skull, 
ventral 
view 
21 


13. 
Hemipsalodon 
viejaensis 
n. 
sp., 
interpretation 
of 
bones 
and 
foramina 
of 
orbit 
and 
lateral 
braincase 
21 


14. 
Hemipsalodon 
viejaensis 
n. 
sp., 
stereophotos 
of 
dentition 
22 


15. 
Ischognathus 
savagei, 
anterior 
mandible 
26 


16. 
Hyaenodon 
cf. 
H. 
vetus, 
mandible 
27 


17. 
Hyaenodon 
cf. 
H. 
vetus, 
stereophotos 
of 
skull 
28 


18. 
Hyaenodon 
montanus 
stereophotos 
of 
dentition 
30 


19. 
Hyaenodon 
montanus,
, 
photo 
of 
mandible 
31 


20. 
Hyaenodon 
crucians, 
stereophotos 
of 
skull 
andM 
0 
32 


21. 
Hyaenodon 
raineyi 
n. 
sp., 
skull 
34 


22. 
Hyaenodon 
raineyi 
n. 
sp., 
skull 
and 
mandible 
35 


23. 
Miacis 
cognitus 
n. 
sp., 
skull, 
lateral 
view 
and 
reconstruction 
36 


24. 
Miacis 
cognitus 
n. 
sp., 
skull, 
ventral 
view 
and 
reconstruction 
37 


25. 
Miacis 
cognitus 
n. 
sp., 
skull, 
dorsal 
view 
and 
reconstruction 
38 


26. 
Miacis 
cognitus, 
stereophotos 
of 
dentition 
and 
basicranium 
39 


27. 
Miacis 
cognitus, 
features 
of 
endocast 
40 


28. 
Miacis 
cognitus, 
photo 
of 
skull 
42 


29. 
Miacis 
australis 
n. 
sp., 
dentition 
43 


30. 
Uintacyon 
scotti, 
skull 
and 
mandible 
45 


31. 
Uintacyon 
scotti, 
right 
upper 
dentition 
46 


32. 
Daphoenocyon 
dodgei, 
Daphoenus 
lambei, 
and 
Daphoeninae 
indet 
49 


33. 
Hesperocyon 
wilsoni 
n. 
sp., 
skull 
and 
mandible 
52 


34. 
Hesperocyon 
wilsoni 
n. 
sp., 
upper 
and 
lower 
dentitions 
53 


35. 
?felid 
indet., 
rostrum 
57 


36. 
Stratigraphic 
ranges 
of 
Uintan-Chadronian 
carnivorous 
mammal 
genera 
59 


IV 



TABLES 


1. 
Museum 
abbreviations 
2 


2. 
Anatomical, 
statistical, 
and 
other 
abbreviations 
2 


3. 
Harpagolestes 
measurements, 
lower 
teeth 
9 


4. 
Harpagolestes 
measurements, 
upper 
teeth 
10 


5. 
Hessolestes 
measurements 
14 


6. 
?Proviverra 
measurements 
16 


7. 
Hemipsalodon 
measurements 
18 


8. 
Hyaenodon 
measurements, 
upper 
teeth 
29 


9. 
Hyaenodon 
measurements, 
lower 
teeth 
29 


10. 
Miacis 
measurements 
41 


11. 
Uintacyon 
measurements 
47 


12. 
Daphoeninae, 
measurements 
ofupper 
teeth 
50 


13. 
Daphoeninae, 
measurements 
of 
lower 
teeth 
50 


14. 
Hesperocyon 
measurements, 
upper 
teeth 
54 


15. 
Hesperocyon 
statistics, 
upper 
teeth 
54 


16. 
Hesperocyon 
measurements, 
lower 
teeth 
55 


17. 
Hesperocyon 
statistics, 
lower 
teeth 
55 


V 



CarnivorousMammalsoftheLate 
Eocene 
and 
Early 
Oligocene
of 
Trans-Pecos 
Texas 


By 
Eric 
Paul 
Gustafson 


ABSTRACT 


The 
fossil 
carnivore 
fauna 
of 
the 
late 
Eocene 
(Bridgerian-
Uintan) 
and 
early 
Oligocene 
(Chadronian) 
of 
Trans-
Pecos 
Texas 
includes 
specimens 
from 
stratigraphicallysuperimposed 
faunas 
of 
three 
areas: 
the 
Sierra 
Vieja

(Vieja 
Group), 
the 
badlands 
near 
Agua 
Fria 
Mountain 
(Buck 
Hill 
Group), 
and 
near 
Castolon 
in 
southern 
BigBend 
National 
Park 
(Chisos 
Formation). 
Members 
of 
the 
mammalian 
orders 
Condylarthra, 
Creodonta, 
and 
Carnivora 
are 
included 
in 
this 
study. 
The 
fossils 
occur 
in 
volcaniclastic 
continental 
sediments, 
interbedded 
with 
lava 
flows 
and 
ignimbrites. 
The 
Cenozoic 
rocks 
unconformably 
overlie 
Cretaceous 
sedimentary 
rocks 
in 
all 
three 
areas. 
Radiometric 
dates 
from 
several 
stratigraphic 
levels 
allow 
the 
calibration 
of 
biostratigraphic 
data. 


The 
Chisos 
Formation 
near 
Castolon 
has 
yielded 
one 
specimen, 
a 
partial 
skeleton 
of 
Uintacyon 
scotti. 
The 
specimen 
came 
from 
about 
185 
meters 
below 
the 
Bee 
Mountain 
Basalt 
and 
above 
the 
Alamo 
Creek 
Basalt 
(about 
42-44.5 
million 
years).

In 
the 
Agua 
Fria 
section, 
carnivores 
come 
from 
several 
levels 
between 
the 
post-Cretaceous 
unconformityand 
the 
Mitchell 
Mesa 
Ignimbrite 
(32.28* 
m.y.). 
The 
lowest 
level, 
the 
Whistler 
Squat 
local 
fauna 
(48.6 
m.y.),
includes 
IProviverra 
major, 
Uintacyon 
sp., 
Hessolestes 
sp., 
a 
possible 
limnocyonine, 
and 
Simidectes 
magnus.
This 
assemblage 
shows 
affinities 
to 
both 
Bridgerian 
and 
Uintan 
faunas 
but 
is 
probably 
Uintan. 
Much 
higher, 
in 
the 
Skyline 
gravels, 
Simidectes 
magnus, 
Harpagolestes 
sp., 
and 
Hyaenodon 
cf. 
H. 
vetus 
are 
found. 
This 
association 
is 
not 
seen 
in 
the 
Utah 
sections 
(where 
Simidectes 
appears 
much 
lower 
than 
Hyaenodon), 
but 
a 
similar 
joint 
occurrence 
of 
Simidectes 
and 
Hyaenodon 
is 
seen 
in 
the 
Pearson 
Ranch 
local 
fauna 
of 
the 
SespeFormation, 
California. 


Of 
the 
several 
superimposed 
local 
faunas 
in 
the 
Sierra 
Vieja, 
the 
lowest 
(Candelaria 
local 
fauna) 
includes 
onlyHarpagolestes 
cf. 
H. 
uintensis. 
It 
is 
Uintan 
on 
the 
basis 
of 
the 
fauna 
and 
of 
its 
position 
below 
the 
Buckshot 
Ignimbrite 
(38.6 
m.y.). 
The 
stratigraphically 
higher 
faunas 
(Porvenir, 
Little 
Egypt, 
and 
Airstrip) 
are 
Chadronian, 
and 
include 
a 
new 
species 
of 
Hemipsalodon,
four 
species 
of 
Hyaenodon, 
Ischnognathus 
savagei 
(anenigmatic 
form), 
two 
new 
species 
of 
Miacis, 
daphoenineamphicyonids, 
a 
new 
Hesperocyon, 
and 
an 
unidentified 
felid-like 
form. 


INTRODUCTION 


Skeletal 
remains 
of 
numerous 
animals 
of 
late 
Eocene 
and 
early 
Oligocene 
age 
are 
found 
in 
thick 
sequences 
of 
sedimentary 
rock 
in 
West 
Texas. 
Interbedded 
with 
the 
volcaniclastic 
sediments 
are 
lava 
flows, 
ignimbrites, 
and 
air-fall 
tuffs. 
The 
certain 
relationship 
of 
superimposedmammalian 
faunas 
and 
widespread 
volcanic 
rock 
units 
provides 
a 
combination 
of 
stratigraphic 
control 
with 
radiometric 
dates 
which 
furnishes 
exceptional 
geochronologic 
documentation 
of 
the 
fossil 
faunas 
of 
the 
area. 
The 
study 
presented 
here 
is 
one 
of 
a 
series 
which 
will 
eventually 
include 
all 
of 
the 
fossil 
mammal 
taxa 
of 
the 
area. 


Fossil 
carnivores 
are 
scarce 
in 
comparison 
to 
herbivores 
at 
all 
of 
the 
West 
Texas 
localities. 
Of 
the 
approximately 
50 
specimens 
on 
which 
this 
report 
is 
based, 
onlya 
few 
consist 
of 
more 
than 
a 
few 
teeth 
or 
skull 
fragments. 
However, 
the 
total 
assemblage 
is 
one 
of 
the 
best 
collections 
of 
carnivores 
from 
stratigraphically 
superimposed 
rock 
units 
of 
late 
Eocene 
and 
early 
Oligocene 
age

in 
North 
America. 


The 
time 
period 
spanned 
by 
these 
faunas 
covers 
a 
crucial 
change 
in 
the 
composition 
of 
carnivore 
faunas. 
It 
was 
in 
the 
late 
Eocene 
that 
members 
of 
the 
modem 
Order 
Carnivora 
began 
their 
major 
increase 
in 
diversity, 
while 
members 
of 
other 
orders, 
for 
example 
the 
Order 
Creodonta, 
became 
greatly 
reduced 
in 
diversity 
by 
the 
beginning 
of 
the 
Oligocene.

The 
purpose 
of 
this 
paper 
is 
to 
identify 
and 
describe 
the 
available 
specimens 
of 
carnivorous 
mammals 
from 
the 
Eocene 
and 
Oligocene 
rocks 
of 
Trans-Pecos 
Texas, 
to 
place 
them 
in 
the 
documented 
stratigraphic 
sequence, 
and 
to 
analyze 
their 
biostratigraphic 
and 
evolutionarysignificance. 
With 
these 
data 
and 
previous 
studies, 
it 
is 
possible 
to 
analyze 
the 
creodont-camivore 
faunal 
replacement 
in'more 
detail 
than 
has 
hitherto 
been 
done. 


The 
completion 
of 
the 
final 
draft 
has 
taken 
several 
years. 
Only 
minor 
additions 
have 
been 
made 
since 
the 
last 
revision, 
made 
in 
1981. 


My 
practice 
of 
making 
comparisons 
between 
the 
Texas 
fossils 
and 
similar 
sized 
modem 
species 
raised 
some 
questions 
among 
reviewers. 
These 
comparisons 
are 
not 
intended 
to 
imply 
any 
taxonomic 
affinities, 
but 


simply 
to 
give 
a 
basis 
for 
comparison 
with 
taxa 
which 
are 
familiar 
to 
most 
readers, 
which 
are 
likely 
to 
be 
represented 
by 
skeletal 
material 
in 
most 
modest 
sized 
museum 
collections, 
and 
whose 
soft 
anatomy 
is 
known. 


’ 


The 
K-Ar 
dates 
given 
in 
this 
paper 
are 
not 
corrected 
to 
the 
lUGS 
Subcommission 
on 
Geochronology. 
1 



Table 
I.—Museum 
abbreviations 
used 
in 
the 
text. 
Other 
abbreviations 
are 
given 
in 
figure 
captions. 


AMNH 
American 
Museum 
ofNatural 
History, 
New 
York 


CIT 
California 
Institute 
of 
Technology 
(collections 
now 
at 
LACM)

CM 
Carnegie 
Museum, 
Pittsburgh

FMNH 
Field 
Museum 
ofNatural 
History, 
Chicago

KU 
University 
of 
Kansas 
Museum, 
Lawrence 


LACM 
Natural 
History 
Museum 
of 
Los 
Angeles 
County

NMC 
National 
Museum 
of 
Canada 


OMSI 
Oregon 
Museum 
of 
Science 
and 
Industry

OU 
Stovall 
Museum, 
University 
ofOklahoma, 
Norman 


PU 
Princeton 
University, 
Philadelphia

ROM 
Royal 
Ontario 
Museum, 
Toronto 


TMM 
Texas 
Memorial 
Museum, 
Austin 


USNM 
United 
States 
National 
Museum, 
Washington, 
D.C. 


YPM 
Yale 
Peapody 
Museum, 
New 
Haven 


Table 
2.—Anatomical, 
statistical, 
and 
other 
abbreviations 
used 
in 
the 
text 


Additional 
abbreviations 
are 
given 
in 
figure 
captions. 


a 
measurement 
taken 
at 
inside 
rim 
of 
alveolus 
AP 
maximum 
anteroposterior 
diameter 
C 
canine 
Co. 
Countyd 
deciduous 
tooth 
e 
estimated 
measurement 
Fm. 
Formation 
I 
incisor 


l.f. 
local 
fauna 
M 
molar 
m.y. 
million 
years 
before 
presentN 
number 
of 
specimens 
measured 
n. 
sp. 
new 
speciesOR 
observed 
range 
of 
variation 
P 
premolar 
s 
standard 
deviation 
sp. 
speciesT 
maximum 
transverse 
diameter 
V 
coefficient 
of 
variation 
X 
sample 
mean 
Specimen 
numbers 
cited 
without 
prefixes 
belong 
to 
the 
Vertebrate 
Paleontology 
Laboratory 
of 
the 
Texas 
Memorial 
Museum, 
which 
has 
detailed 
descriptions 
of 
their 
localities 
on 
file. 
All 
TMM 
numbers 
consist 
of 
a 
five 
digit 
locality 
number 
followed 
by 
a 
hyphen 
and 
an 
individual 
specimen 
number. 
In 
specimen 
lists 
numbers 
have 
been 
abbreviated 
by 
giving 
only 
the 
hyphen 
and 
the 
specimen 
number 
for 
second 
and 
subsequent 
numbers 
from 
the 
same 
locality 
(e.g., 
41372-1, 
-2). 


ACKNOWLEDGEMENTS 
This 
project 
was 
suggested 
by 
Dr. 
John 
A. 
Wilson, 
who 
until 
his 
retirement 
was 
its 
supervisor. 
Indeed, 
without 
Dr. 
Wilson’s 
energetic 
and 
patient 
collecting 
over 
many 


years, 
such 
a 
project 
would 
not 
have 
been 
possible. 
I 
also 
benefitted 
from 
his 
direction 
and 
instruction 
in 
the 
field 
in 
West 
Texas 
and 
from 
his 
editing 
of 
the 
various 
drafts 
of 
this 
work. 
Dr. 
Ernest 
L. 
Lundelius, 
Jr., 
has 
been 
most 
patient 
and 
helpful 
in 
finishing 
the 
final 
drafts 
of 
the 
dissertation 
and 
in 
lending 
support 
during 
my 
stay 
at 
The 
University 
of 
Texas. 
I 
am 
indebted 
to 
Dr. 
Don 
Baird 
of 
Princeton 
University, 
Dr. 
Mary 
Dawson 
of 
the 
CarnegieMuseum, 
Pittsburgh, 
Dr. 
C. 
Lewis 
Gazin, 
Dr. 
ClaytonRay, 
and 
Mr. 
Robert 
Purdy 
of 
the 
National 
Museum 
of 
Natural 
History, 
Dr. 
Robert 
Hunt 
of 
the 
University 
of 
Nebraska, 
and 
members 
of 
the 
staff 
of 
the. 
American 
Museum 
of 
Natural 
History, 
New 
York, 
and 
Field 
Museum 
of 
Natural 
History, 
Chicago, 
for 
providingspecimens, 
casts, 
and 
access 
to 
collections. 
Dr. 
James 
Mellett 
gave 
me 
much 
information 
on 
Hyaenodon 
and 
allowed 
me 
to 
use 
his 
casts 
of 
Hyaenodon 
types. 
Dr. 
Leonard 
Radinsky 
and 
Dr. 
Harry 
Jerison 
examined 
the 
brain 
casts 
of 
Miacis 
and 
other 
skulls 
and 
discussed 
their 
ideas 
with 
me. 
Many 
others 
have 
given 
their 
time 
to 
discuss 
various 
aspects 
of 
this 
project. 
The 
Vertebrate 
Paleontology 
Laboratory 
of 
the 
Texas 
Memorial 
Museum 
has 
given 
me 
its 
full 
support 
and 
use 
of 
its 
facilities. 
Parts 
of 
my 
work 
were 
supported 
financially 
by 
the 


Geology 
Foundation 
of 
the 
Department 
of 
GeologicalSciences, 
The 
University 
of 
Texas 
at 
Austin. 


All 
drawings 
not 
the 
work 
of 
the 
author 
were 
made 
byMargaret 
Skeels 
Stevens, 
and 
are 
identified 
by 
her 
initials. 
Mrs. 
Stevens 
also 
collected 
many 
of 
the 
specimens. 


My 
wife 
Jessie 
has 
given 
me 
her 
wholehearted 
supportduring 
all 
of 
my 
work, 
under 
conditions 
which 
have 
sometimes 
been 
very 
difficult. 
For 
this 
support 
which 
no 
one 
else 
could 
give, 
I 
am 
extremely 
grateful.

This 
publication 
is 
a 
contribution 
of 
the 
Vertebrate 
Paleontology 
Laboratory, 
Texas 
Memorial 
Museum, 
The 
University 
ofTexas 
at 
Austin. 


2 



TERMINOLOGY 
GEOGRAPHIC 
SETTING 


Dental 
terminology 
(Fig. 
1, 
2) 
is 
taken 
from 
Maclntyre(1966), 
whose 
terms 
were 
developed 
for 
a 
generalizedmiacid 
dentition, 
but 
which 
also 
apply 
to 
other 
carnivorous 
mammals. 
Although 
names 
of 
major 
cusps

and 
other 
features 
are 
standard 
terms, 
in 
some 
cases 
this 
terminology 
differs 
somewhat 
from 
that 
of 
Van 
Valen 
(1966) 
and 
Szalay 
(1969a), 
and 
terms 
for 
minor 
features 
often 
are 
not 
the 
same. 


Skeletal 
and 
other 
anatomical 
terminology, 
particularly 
names 
of 
foramina 
and 
muscles, 
was 
taken 
from 
Romer 
(1970), 
Miller, 
Christensen 
and 
Evans 
(1964), 
and 
Davis 
(1964). 


The 
described 
specimens 
come 
from 
three 
areas 
within 
Presidio 
and 
Brewster 
Counties, 
Texas 
(Fig. 
3). 
From 
northwest 
to 
southeast, 
these 
are; 
1) 
the 
rugged 
and 
poorly 
accessible 
area 
between 
the 
high 
rim 
rock 
of 
the 
Sierra 
Vieja 
and 
the 
Rio 
Grande 
in 
Presidio 
County, 
2) 
the 
badlands 
north 
and 
west 
of 
Agua 
Fria 
Mountain 
in 
southern 
Brewster 
County, 
and 
3) 
an 
area 
southwest 
of 
the 
Chisos 
Mountains 
in 
Big 
Bend 
National 
Park 
near 
Castolon, 
also 
in 
Brewster 
County. 
Exposures 
are 
alongthe 
major 
tributaries 
of 
the 
Rio 
Grande. 


Figure 
I.—Diagrammatic 
occlusal 
views 
of 
generalized 
miacid 
dentition, 
illustrating 
cusp 
terminology. 
A. 
Left 
upper 
teeth. 
B. 
Rightlower 
teeth. 
From 
Maclntyre 
(1966, 
Fig. 
1). 


Abbreviations: 
ACC 
ML, 
accessory 
metaconule; 
END, 
entoconid; 
HLD, 
hypoconulid; 
HY, 
hypocone; 
HYD, 
hypoconid; 
ME, 
metacone; 
MED, 
metaconid; 
ML, 
metaconule; 
MSL, 
mesoconule; 
MSS, 
mesostyle; 
MTS, 
metastyle; 
PA, 
paracone; 
PAD, 
paraconid; 
PCL, 
paraconule; 
PE, 
pericone; 
PED, 
periconid; 
POST-HYD, 
post-hypoconid; 
PR, 
protocone; 
PRD, 
protoconid; 
PRE-END, 
per-entoconid;
PRE-HYD, 
pre-hypoconid; 
PS, 
parastyle; 
SC, 
stylocone. 


Figure 
2.—Diagrams 
of 
generalized 
miacid 
molars, 
illustrating 
crest 
terminology. 
A. 
Labial 
view 
of 
upper 
molar. 
B. 
Occlusal 
view 
of 
upper 
molar. 
C. 
Occlusal 
view 
of 
lower 
molar. 
D. 
Lingual 
view 
of 
lower 
molar. 
From 
Maclntyre 
(1966, 
Fig. 
2). 


Abbreviations: 
ANT 
LAB 
CNG, 
anterior 
labial 
cingulum; 
ANT 
LNG 
CNG, 
anterior 
lingual 
cingulum; 
CRN 
NCH, 
camassial 
notch; 
CR 
OBL, 
crista 
obliqua; 
LNG 
TLD 
CR, 
lingual 
talonid 
crest; 
POST 
LAB 
CNG, 
posterior 
labial 
cingulum; 
POST 
LNG 
CNG, 
posteriorlingual 
cingulum; 
POST 
TLD 
CR, 
posterior 
talonid 
crest; 
PR 
BAS, 
protocone 
basin; 
PS 
SHF, 
parastyle 
shelf; 
SR 
CNG, 
stylar 
cingulum;
SR 
SHF, 
stylar 
shelf; 
SUL 
OBL, 
sulcus 
obliquue; 
TLD 
BAS, 
talonid 
basin; 
TRD 
SHF, 
trigonid 
shelf. 


3 



PREVIOUS 
WORK 


Of 
the 
three 
fossiliferous 
sections 
mentioned 
in 
this 
report, 
the 
Vieja 
Group 
has 
received 
the 
most 
attention. 
Various 
taxa 
from 
the 
Vieja 
Group 
are 
discussed 
byStovall 
(1948; 
the 
fauna 
including 
some 
carnivores),
McGrew 
(1953; 
Equidae), 
Wilson 
(1966; 
primate), 
Harris 
(1967; 
rhinocerotoid), 
Harris 
and 
Wood 
(1969; 
eomyidrodent), 
Wilson 
(1971a; 
Entelodontidae, 
and 
1971b;
Agriochoeridae 
and 
Merycoidodontidae), 
Forsten 
(1971aand 
b; 
Equidae), 
McGrew 
(1971; 
Equidae), 
Wood 
(1974;
Rodentia), 
Wilson 
(1974; 
Protoceratidae, 
Camelidae, 
and 
Hypertragulidae), 
Novacek 
(1976; 
Insectivora), 
Wilson 
(1977a; 
Brontotheriidae), 
and 
Wilson 
and 
Schiebout 
(1981 
and 
1984; 
Ceratomorpha). 
Wilson 
et 
al. 
(1968)
present 
preliminary 
faunal 
lists 
(including 
carnivores) 
for 
the 
Vieja 
faunas 
together 
with 
stratigraphic 
information 
and 
radiometric 
dates, 
and 
Wilson 
(1977b) 
summarizes 
the 
localities 
and 
their 
faunal 
correlations. 
The 
geochronology 
of 
the 
general 
area 
is 
discussed 
byMcDowell 
(1979) 
and 
Wilson 
(1980).

The 
faunas 
of 
the 
Agua 
Fria 
area 
(Buck 
Hill 
Group)
have 
not 
been 
as 
fully 
described. 
Taxonomic 
information 
is 
available 
from 
Wood 
(1972, 
1973; 
Rodentia), 
Wilson 
(1974; 
Leptoreodon), 
Stevens 
et 
al. 
(1975; 
mentions 
Hyaenodon), 
Wilson 
and 
Szalay 
(1976; 
primate, 
but 
the 


paper 
mentions 
an 
amynodont, 
artiodactyls, 
and 
Hyaenodon), 
Wilson 
(1977a; 
titanotheres). 
Wilson 
and 
Schiebout 
(1981 
and 
1984; 
ceratomorphs), 
and 
West 
(1982; 
marsupicamivores, 
primates, 
taeniodonts,
condylarths, 
bunodont 
artiodactyls, 
and 
Dinocerata).
Gustafson 
(1979) 
described 
Simidectes 
from 
this 
section. 


The 
faunas 
of 
the 
Tertiary 
formations 
of 
Big 
Bend 
National 
Park 
have 
been 
described 
by 
Wilson 
(inMaxwell, 
et 
al., 
1967; 
includes 
Chisos 
Formation 
faunas 
briefly), 
Stevens 
et 
al. 
(1969), 
Stevens 
(1977; 
Miocene 
faunas), 
and 
Schiebout 
(1974; 
Paleocene 
faunas). 
The 
Uintacyon 
from 
the 
Chisos 
Formation 
has 
been 
mentioned 
in 
several 
publications, 
including 
Wilson 
(1965, 
1967, 
1972) 
and 
Maxwell 
(1968). 


STRATIGRAPHY 


With 
the 
exception 
of 
Tomillo 
Flat 
in 
Big 
Bend 
National 
Park 
(Wilson, 
1967; 
Schiebout, 
1974), 
the 
deposition 
of 
early 
Tertiary 
rocks 
in 
southwest 
Texas 
was 
preceded 
by 
a 
lengthy 
hiatus, 
marked 
by 
a 
widespread 
and 
locallyangular 
unconformity 
at 
the 
top 
of 
Comanchean 
or 
Gulfian 
(Cretaceous) 
sediments 
which 
were 
folded 
duringthe 
Laramide 
orogeny. 
In 
many 
places 
the 
first 
Tertiarysediments 
are 
limestone-quartzite 
conglomerates 
(JeffConglomerate 
of 
DeFord, 
1958), 
but 
these 
are 
generallythin 
and 
are 
buried 
under 
very 
thick 
deposits 
of 
tuffaceous 
sediment 
and 
volcanic 
flows 
(basalt, 
rhyolite,
ignimbrites, 
and 
lahars). 
Tertiary 
volcanic 
activity 
seems 
to 
have 
begun 
in 
the 
Big 
Bend 
area 
in 
the 
mid 
to 
late 
Eocene 
(Stevens, 
1969; 
Fig. 
4) 
and 
continued 
intermittently 
through 
much 
of 
the 
Tertiary. 
Majornormal 
faulting 
began 
in 
the 
early 
Miocene 
(Arikareean)
and 
has 
combined 
with 
numerous 
igneous 
intrusions 
to 


make 
the 
geology 
highly 
complex. 


Chisos 
Formation, 
Castolon 
area, 
Big 
Bend 
National 
Park.—The 
geology 
of 
Big 
Bend 
National 
Park 
was 
described 
by 
Maxwell 
et 
al. 
(1967), 
and 
the 
Castolon 


area 
was 
examined 
in 
detail 
by 
Stevens 
(1969) 
The 
only

significant 
carnivore 
specimen, 
the 
skeleton 
of 
Uintacyonscotti, 
was 
found 
in 
the 
lower 
part 
of 
the 
Chisos 
Formation, 
which 
overlies 
Cretaceous 
rocks 
in 
this 
area 
and 
is 
unconformably 
overlain 
by 
the 
South 
Rim 
Formation, 
of 
Oligocene 
age. 
This 
stratigraphic 
section 
was 
given 
by 
Maxwell 
et 
al. 
(1967, 
p. 
132, 
and 
Plate 
X, 
sec. 
36). 
The 
Uintacyon 
skeleton 
was 
found 
in 
a 
tuffaceous 
mudstone 
between 
2p.7 
and 
27.4 
m 
(68 
and 


90 
feet) 
above 
the 
Alamo 
Creek 
Basalt. 
Maxwell 
et 
al. 
(1967) 
give 
K/Ar 
dates 
for 
the 
Alamo 
Creek 
of 
38.7 
m.y., 
40.1 
m.y., 
42.7 
m.y., 
and 
44.3 
m.y.*.
Information 
from 
dates 
and 
faunas 
elsewhere 
suggeststhat 
the 
oldest 
date 
is 
nearest 
correct 
(Wilson, 
pers. 
comm., 
1981).

Buck 
Hill 
Group, 
Devil’s 
Graveyard 
badlands, 
AguaFria 
area.—Detailed 
stratigraphy 
of 
the 
sedimentaryrocks 
and 
tuffs 
of 
the 
Buck 
Hill 
Group 
in 
this 
area 
has 
been 
worked 
out 
by 
Stevens 
et 
al. 
(1975, 
1984). 
The 
names 
Devil’s 
Graveyard 
Formation 
and 
Bandera 
Mesa 
Member 
(Fig. 
4) 
were 
proposed 
by 
Stevens 
et 
al. 
(1984).
They 
abandoned 
the 
name 
Pruett 
Formation 
because 
the 


top 
of 
the 
Pruett 
Formation 
was 
defined 
by 
Goldich 
and 
Elms 
(1949) 
as 
the 
base 
of 
the 
Crossen 
Trachyte, 
which 
does 
not 
extend 
as 
far 
south 
as 
the 
Agua 
Fria 
area. 
Moon 
(1953) 
described 
and 
mapped 
the 
geology 
of 
the 
Agua 
Fria 
quadrangle. 
The 
fossiliferous 
sediments 
lie 
between 
the 
post-Cretaceous 
unconformity 
and 
the 
Mitchell 
Mesa 
Ignimbrite. 
The 
sections 
figured 
here 
(Fig.
4) 
are 
composites 
derived 
from 
detailed 
sections 
measured 
by 
James 
B. 
Stevens 
and 
drawn 
by 
Margaret 
S. 
Stevens. 
Some 
of 
the 
informal 
units 
shown 
in 
Fig. 
4 
have 
been 
mentioned 
by 
Wood 
(1973), 
Wilson 
and 
Szalay(1976), 
and 
Wilson 
(1977). 
K/Ar 
dates 
from 
tuffs 
above 


and 
below 
the 
Whistler 
Squat 
#2 
locality 
place 
its 
age

between 
43 
and 
46 
m.y. 
The 
Whistler 
Squat 
local 
fauna, 
as 


used 
here, 
includes 
faunas 
from 
the 
basal 
Tertiaryconglomerate 
up 
to 
and 
including 
Whistler 
Squat 
Quarryand 
other 
localities 
at 
approximately 
the 
same 
stratigraphic 
level. 
The 
Serendipity 
local 
fauna 
includes 


fossils 
from 
the 
Serendipity 
locality 
up 
to 
and 
includingPurple 
Bench. 
The 
Skyline 
local 
fauna 
is 
derived 
from 
the 
Skyline 
conglomerate 
(Fig. 
4).

Vieja 
Group, 
Sierra 
Vieja.—The 
stratigraphy 
of 
the 
Vieja 
Group 
was 
worked 
out 
by 
R. 
K. 
DeFord 
and 
numerous 
graduate 
students 
and 
was 
summarized 
byDeFord 
(1958). 
The 
stratigraphic 
positions 
of 
faunas 
and 
radiometric 
dates 
of 
the 
igneous 
rocks 
were 
presented 
byWilson 
et 
al. 
(1968) 
and 
by 
Wilson 
(1977b). 
Walton 


(1972, 
1977) 
described 
the 
sedimentary 
petrology 
of 
the 
Vieja 
Group. 
The 
Group 
includes 
the 
followingformations, 
from 
oldest 
to 
youngest;

Radiometric 
dates 
are 
given 
as 


they 
originally 
appear 
in 
publications.

recognize 
that 
these 
dates, 
given 
accurately, 
should 
include 
some1 


indication 
of 
error. 
It 
is 
unfortunate 
that 
these 
often 
are 
not 
available,

and 
where 
error 
figures 
are 
given 
the 
nature 
of 
the 
indicated 
error 


range 
often 
is 
left 
unclear. 
For 
corrected 
dates 
see 
Stevens, 
Stevens, 


and 
Wilson 
(1984). 


4 



P,

Agua

AF,Marfa;
fauna;Mar,
local

Marathon;


Airstrip


A,1.f.;

EgyptLittle

Abbreviations:M,

LE,

localities.locality;


fossil 


Uintacyon


showing


Texas,Formation,
Counties,Chisos


1.f.;

PresidioCh,

3.—Map


and 


Candelaria


C,

Brewster

of 


Alpine;


Al,

area;l.f. 


Figure 
FriaPorvenir


Figure 
4. 
Diagrammatic 
stratigraphic 
section 
of 
the 
Buck 
Hill 
Group, 
Devil’s 
Graveyard-Bandera 
Mesa 
area, 
Brewster 
and 
Presidio 
Counties, 
Texas, 
to 
show 
approximate 
relative 
thickness 
and 
stratigraphic 
position 
of 
fossil 
localities, 
local 
faunas, 
and 
North 
American 
land 
mammal 
ages. 
Radiometric 
dates 
(in 
millions 
of 
years) 
are 
from 
McDowell 
(1979) 
except 
32.3 
m.y., 
which 
is 
an 
unpublished 
date 
from 
Geochron 
Laboratories. 
The 
number 
in 
parentheses 
following 
each 
data 
is 
the 
number 
of 
samples. 
Letters 
refer 
to 
stratigraphicposition 
of 
one 
or 
more 
fossil 
localities. 
Fossil 
localities 
referred 
to 
in 
this 
paper 
are: 
B. 
TMM 
41372, 
Whistler 
Squat 
quarry; 
TMM 
41576, 
southwest 
of 
Wax 
Camp; 
TMM 
41466, 
300 
yards 
east 
of 
Whistler 
Squat 
quarry. 
C. 
TMM 
41745, 
Serendipity. 
D. 
TMM 
41784,

Unio 
Cliff. 
E. 
Titanothere 
Hill. 
F. 
Purple 
Bench. 
G. 
Skyline 
channels; 
TMM 
41535, 
Boot 
Canyon; 
North 
Fork 
of 
Alamo 
Cesario. 


6 



(I) 
JeffConglomerate.—Basal 
unit 
overlying 
Cretaceous 
sediments, 
composed 
of 
channel-shaped 
bodies 
of 
sandstone 
and 
conglomerate, 
7.3 
to 
91 
m 
(24 
to 
300 
feet) 
thick, 
containing 
many 
limestone 
clasts. 
Not 
fossiliferous. 
Where 
the 
overlying 
Gill 
Breccia 
is 
absent, 
rocks 
correlative 
to 
this 
unit 
are 
assigned 
to 
the 
lower 
part 
of 
the 
Colmena 
Formation. 


(2) 
Gill 
Breccia.—Volcanic 
flow 
breccia 
of 
variable 
composition, 
mostly 
trachybasalt 
porphyry, 
up 
to 
110 
m 
(360 
feet) 
thick, 
restricted 
to 
the 
southern 
Sierra 
Vieja 
area. 
A 
K/Ar 
date 
of 
40.0±2.0 
m.y. 
was 
obtained 
on 
a 
sample 
from 
near 
the 
bottom 
of 
the 
section 
(Wilson 
et 
al., 
1968; 
but 
see 
Wilson, 
1984). 
(3) 
Colmena 
Formation.—This 
formation 
was 
defined 
as 
the 
rocks 
between 
the 
Gill 
Breccia 
(or 
the 
top 
of 
the 
Cretaceous 
rocks 
where 
the 
Gill 
is 
absent) 
and 
the 
bottom 
of 
the 
Buckshot 
Ignimbrite. 
It 
is 
composed 
of 
tuffaceous 
sandstone 
and 
volcanic 
mudflow 
conglomerate 
or 
lahars 
in 
the 
southern 
areas, 
and 
limestone-sandstone 
conglomerate 
in 
the 
northern 
area 
and 
is 
up 
to 
137 
m 
(450 
feet) 
thick. 
The 
Candelaria 
local 
fauna 
comes 
from 
the 
lower 
15.2 
m 
(50 
feet)
and 
upper 
7.6 
m 
(25 
feet) 
of 
the 
61 
m 
(200 
foot)
thick 
section 
near 
the 
mouth 
of 
Capote 
Creek 
(Wilson,
1977b). 


(4) 
Buckshot 
Ignimbrite.—A 
fine 
grained 
rhyoliteignimbrite, 
up 
to 
23 
m 
(75 
feet) 
thick, 
extensive, 
and 
useful 
as 
a 
marker 
bed. 
K/Ar 
dates 
on 
a 
sanidine 
concentrate 
of 
34.7, 
35.2, 
and 
38,6 
m.y. 
are 
reportedby 
Wilson 
et 
al. 
(1968); 
the 
first 
two 
are 
probably 
in 
error, 
with 
argon 
loss 
likely 
from 
the 
devitrified 
glassmatrix. 
The 
most 
recent 
dates 
(Wilson, 
1984) 
are 
37.30 
and 
37.0 
m.y. 
(5) 
Chambers 
Formation.—Up 
to 
257 
m 
(844 
feet) 
of 
volcaniclastic 
sandstones, 
siltstones, 
and 
conglomerates 
lie 
conformably 
on 
the 
Buckshot, 
with 
the 
top 
defined 
as 
the 
base 
of 
the 
Bracks 
Rhyolite.
The 
Bracks 
is 
absent 
in 
some 
areas, 
but 
in 
places 
a 
persistent 
red 
siltstone 
just 
below 
it 
can 
be 
used 
to 
mark 
the 
upper 
Chambers 
boundary. 
However, 
in 
the 
southern 
Vieja 
area 
no 
such 
marker 
bed 
is 
available 
and 
the 
formation 
upper 
boundary 
is 
obscure. 
Two 
internal 
marker 
beds, 
a 
white 
ash 
flow 
tuff 
and 
a 
red 
crystal 
tuff, 
can 
be 
used 
for 
control 
within 
the 
Chambers 
(Walton, 
1972). 
(6) 
Bracks 
Rhyolite. 
—The 
Bracks 
Rhyolite 
is 
an 
important 
cliff 
forming 
unit 
up 
to 
91 
m 
(300 
feet)
thick. 
Two 
dates 
given 
for 
it, 
by 
Evemden 
et 
al. 
(1964; 
36.8 
m.y.) 
and 
Wilson 
et 
al. 
(1968; 
36.5±1.2 
m.y.), 
are 
in 
close 
agreement. 
More 
recent 
dates 
(Wilson, 
1984) 
are 
37.40 
and 
37.70 
m.y. 


(7) 
Capote 
Mountain 
Formation. 
—This 
unit 
of 
volcaniclastic 
sediments 
and 
tuff 
lies 
between 
the 
Bracks 
and 
the 
Mitchell 
Mesa 
Ignimbrite. 
With 
a 
maximum 
thickness 
of 
about 
640 
m 
(2100 
feet), 
the 


Capote 
Mountain 
is 
the 
thickest 
formation 
in 
the 
Vieja

Group. 
The 
Airstrip 
local 
fauna 
occurs 
173 
m 
(570

feet) 
above 
the 
Bracks 
Rhyolite. 
The 
Ash 
Spring 
local 


fauna 
is 
found 
40 
km 
(25 
miles) 
to 
the 
northwest 
of 


the 
Airstrip 
local 
fauna 
and 
appears 
to 
be 
somewhat 


younger 
than 
it, 
though 
there 
is 
some 
question 
of 
the 


exact 
stratigraphic 
position 
of 
the 
Ash 
Spring 
local 


fauna 
because 
of 
the 
difficulties 
of 
correlation 
(Wilson 


et 
al. 
1968; 
Wood 
1974; 
Wilson 
1977b). 


(8)Mitchell 
Mesa 
Ignimbrite.—This 
is 
the 
most 
widespread 
single 
unit. 
It 
occurs 
in 
both 
Vieja 
and 
Agua 
Fria 
sections, 
and 
can 
be 
traced 
southeast 
alongthe 
Rio 
Grande 
for 
about 
40 
miles 
downriver 
from 
Presidio. 
In 
the 
Bofecillos 
Mountains, 
between 
Presidio 
and 
Lajitas, 
the 
Mitchell 
Mesa 
closelyoverlies 
the 
Tule 
Mountain 
Trachyandesite 
Member 
of 
the 
Chisos 
Formation, 
which 
can 
be 
traced 
from 
there 
to 
Big 
Bend 
National 
Park. 
There 
it 
overlies 
the 
Mule 
Ear 
Spring 
Tuff 
Member 
in 
the 
Chisos 
section. 
The 
Mitchell 
Mesa 
was 
called 
the 
Brite 
Ignimbrite 
byDeFord 
(1958) 
and 
is 
listed 
under 
that 
name 
byWilson 
et 
al. 
(1968); 
however, 
the 
name 
Mitchell 
Mesa 
was 
proposed 
by 
Goldich 
and 
Elms 
(1949).
K/Ar 
dates 
ranging 
from 
29.7 
m.y. 
to 
33.9 
m.y. 
are 
given 
by 
Wilson 
et 
al. 
(1968), 
but 
further 
work 
byMcDowell 
(1979) 
has 
produced 
18 
dates 
for 
the 
Mitchell 
Mesa 
which 
average 
32.3 
m.y. 
CLASSIFICATION 


All 
of 
the 
taxa 
considered 
in 
this 
study 
have 
been 
placedin 
the 
Order 
Carnivora 
at 
one 
time 
or 
another. 
Reclassification 
in 
recent 
years 
has 
divided 
them 
amongthree 
orders: 
Condylarthra 
(mesonychids), 
Creodonta 
(in 
a 
restricted 
sense 
not 
including 
miacids, 
sometimes 
under 
other 
names 
as 
in 
Van 
Valen, 
1966), 
and 
Carnivora 
sensu 
stricto. 
The 
most 
important 
factor 
governinginclusion 
in 
this 
study 
is 
the 
possession 
of 
characters 
suggesting 
a 
carnivorous 
diet. 
For 
example, 
the 
mesonychids 
apparently 
began 
as 
specialized 
carnivorous 
animals 
(Szalay, 
1969b), 
and 
though 
later 
mesonychidsapparently 
strayed 
from 
the 
ranks 
of 
the 
primarilycarnivorous, 
the 
group 
as 
a 
whole 
certainly 
deserves 
inclusion. 
The 
taxa 
representing 
Creodonta 
and 
Carnivora 
are 
all 
specialized 
in 
familiar 
ways 
as 
efficient 
carnivores, 
especially 
in 
the 
development 
of 
camassial 
teeth. 


The 
three-fold 
classification 
used 
here 
follows 
Romer 
(1966). 
Separation 
of 
Creodonta 
and 
Carnivora 
as 
distinct 
orders 
was 
discussed 
by 
Van 
Valen 
(1965, 
1966).
Although 
I 
do 
not 
use 
his 
term 
“Deltatheridia” 
for 
the 
creodonts, 
this 
arrangement 
is 
natural 
and 
useful. 
Van 
Valen 
(1974) 
has 
also 
abandoned 
the 
name 
“Deltatheridia”, 
and 
would 
prefer 
“Hyaenodonta” 
(VanValen, 
1967). 
The 
name 
Creodonta, 
however, 
is 
not 
confusing 
if 
its 
contents 
are 
specified 
and 
is 
preferablebecause 
it 
is 
an 
older 
and 
well 
established 
name. 


7 



SYSTEMATIC 
PALEONTOLOGY 


Order 
Condylarthra 
Cope, 
1881 


Family 
Mesonychidae 
Cope, 
1875 


Subfamily 
Mesonychinae 
Wortman, 
1901 


Genus 
Harpagolestes 
Wortman, 
1901 


Harpagolestes 
cf. 
H. 
uintensis 
(Scott, 
1888) 


Candelaria 
local 
fauna 


Figure 
5, 
Table 
3 
Type—PUll6s9, 
isolated 
P 
M 
( 
, 
M 
3, 
and 
another

, 


questionably 
associated 
premolar, 
Uinta 
Formation, 
Utah. 
Referred 
material.—TMM 
40498-3, 
right 
ramus 
with 
CPl’ 
P3-4’ 
and 
l6ft 
P4'

’ 


Stratigraphic 
position.—Within 
the 
lower 
15.2 
m 


(50 
feet) 
(lower 
sedimentary 
unit 
of 
Wilson, 
1977b) 
of 


the 
Colmena 
Formation,Vieja 
Group, 
Presidio 
County, 
Texas. 


Age.—Uintan 
(late 
Eocene). 


Description. 
—This 
specimen 
(Fig. 
5) 
falls 
within 
or 
near 
the 
known 
range 
of 
variation 
of 
Harpagolestesuintensis 
in 
all 
measurements 
(Table 
3). 
There 
are 
two 
incisors, 
of 
which 
the 
lateral 
incisor 
(?I3) 
is 
somewhat 
the 
larger. 
The 
canine 
is 
very 
large 
at 
the 
base 
of 
the 
enamel 


but 
tapers 
rapidly 
toward 
the 
tip. 
is 
set 
close 
behind 


the 
canine. 
It 
is 
single 
cusped 
and 
single 
rooted, 
and 
is 
widest 
below 
the 
enamel, 
about 
5 
mm 
above 
the 
alveolar 
rim. 
is 
missing. 
Its 
smaller 
anterior 
alveolus 
is 
set 
partly 
medial 
to 
the 
Pj 
alveolus. 
There 
are 
no 
diastemata 
such 
as 
are 
seen 
in 
some 
specimens 
of 
H. 
uintensis, 
particularly 
AMNH 
1878 
(Szalay 
and 
Gould, 
1966, 
Fig.
4). 
This 
feature 
seems 
tobe 
variable 
in 
Harpagolestes 
as

, 


is 
also 
the 
depth 
of 
the 
mandible. 
The 
anterior 
basal 
cusp 
(paraconid) 
of 
P 
3 
is 
larger 
and 
much 
more 
distinctlycuspate 
than 
in 
the 
type 
(PU 
11659). 
The 
same 
is 
true 
of 
P4; 
however, 
a 
referred 
specimen 
from 
Utah 
(CM 
2961) 
has 
a 
paraconid 
nearly 
as 
large 
as 
the 
Texas 
specimen 
(Peterson, 
1931a, 
plate 
17). 
P 
and 
P 
4 
are 
more 
robust 
than 
those 
of 
the 
type. 
There 
are 
two 
mental 
foramina, 
a 
larger 
foramen 
beneath 
the 
anterior 
root 
of 
and 
a 


.

smaller 
beneath 
the 
anterior 
root 
of 
P 


Relationships. 
—The 
genus 
Harpagolestes 
is 
in 
need 
of 
revision. 
Except 
for 
a 
few 
specimens 
referred 
to 
H. 
uintensis, 
each 
new 
specimen 
has 
been 
given 
a 
new 
name. 
Furthermore, 
an 
attempt 
to 
designate 
AMNH 
1892 
as 
a 
neotype 
for 
H. 
uintensis 
(Szalay 
and 
Gould, 
1966, 


p. 
142) 
is 
not 
valid 
because 
the 
first 
and 
only 
specimen(PU 
11659) 
placed 
in 
the 
species 
by 
Scott 
(automatically 
the 
holotype 
by 
Article 
73a 
of 
the 
International 
Code 
of 
Zoological 
Nomenclature) 
has 
been 
consistently 
accepted 
as 
the 
type 
by 
various 
authors 
(Osborn, 
1895; 
Peterson, 
1919, 
1931a; 
Thorpe, 
1923). 
Some 
of 
the 
species 
described 
later 
(for 
instance 
H. 
leontensis 
Peterson, 
1931a, 
and 
H. 
immanis 
Matthew, 
1909) 
show 
little 
difference 
in 
the 
teeth 
from 
H. 
uintensis 
and 
may 
be 
synonyms. 
However, 
the 
variation 
in 
some 
features, 


particularly 
in 
the 
depth 
of 
the 
mandible 
shown 
by 
specimens 
from 
the 
Uinta 
Formation 
(for 
instance, 
YPM 


10072 
and 
AMNH 
1878) 
is 
more 
extreme 
than 
might 
be 
expected 
in 
a 
single 
species.

Harpagolestes 
uintensis 
is 
probably 
descended 
from 


H. 
macrocephalus 
from 
the 
Bridger 
Formation, 
providing 
that 
the 
loss 
of 
P2 
in 
H. 
macrocephalus 
is 
an 
individual 
variation 
of 
the 
type 
and 
only 
known 


specimen.

The 
ramus 
(40498-3) 
is 
broken 
posterior 
to 
of 
which 
only 
the 
roots 
are 
preserved. 
A 
second 
fragment 
bearing 
what 
appear 
to 
be 
the 
roots 
of 
has 
a 
subsurface 
fit 
and 
the 
grain 
on 
the 
lateral 
side 
of 
the 


more

ramus 
matches 
perfectly. 
The 
posterior 
of 
these 
roots 
of 
?M_, 
projects 
backwards 
at 
an 
angle 
which 
approaches 
the 
horizontal, 
suggesting 
that 
there 
may 
not 
have 
been 
any 
additional 
teeth 
behind 
it. 
In 
most 
mammals 
this 
would 
probably 
be 
the 
case. 
However, 
the 
lower 
margin 
of 
the 
mandible 
of 
Harpagolestes 
uintensis 
(which 
has 
[Peterson. 
1931, 
Fig. 
1]) 
is 
strongly 


arcuate,so 
that 
the 
roots 
of 
the 
posterior 
cheek 
teeth 
are 
not 
parallel 
but 
rather 
are 
radial 
about 
a 
point 
above 
the 
upper 
cheek 
teeth. 
If 
such 
strong 
curvature 
occurred 
in 


this 
specimen, 
there 
could 
well 
have 
been 
adequate 
room 
for 
a 
small 
M?. 


Harpagolestes 
sp. 


Skyline 
conglomerate 


Figure 
6, 
Table 
4 
Referred 
material.—TMM 
41715-8, 
anterior 
partial

l 


,

rostrum, 
with 
alveoli 
for 
right 
canine 
and 
Pcomplete

P2-M2. 


Stratigraphic 
position.—Skyline 
channels, 
base 
of 
Bandera 
Mesa 
Member, 
Devil’s 
Graveyard 
Formation, 
North 
fork 
of 
Alamo 
de 
Cesario, 
Brewster 
Co., 
Texas. 


Age.—late 
Uintan, 
early 
Duchesnean 
subage.

Description. 
—The 
specimen 
(Fig. 
6) 
consists 
of 
the 
maxillary, 
lacrimal, 
jugal, 
and 
palatine 
bones 
of 
the 
rightside, 
apparently 
separated 
from 
surrounding 
bones 
alongthe 
sutures. 
This 
separation 
and 
the 
lack 
of 
heavy 
wear 
on 
the 
dentition 
indicate 
this 
was 
an 
immature 
individual. 
Mesonychids 
commonly 
show 
substantial 
tooth 
wear, 
and 
most 
specimens 
of 
Harpagolestes 
have 
at 
least 
the 
tips 
of 
the 
cusps 
worn 
flat. 
The 
teeth 
in 
TMM 
41715-8 
show 
wear 
facets 
on 
the 
medial 
side 
of 
the 
paracones 
and 
metacones, 
and 
on 
the 
lateral 
sides 
of 
the 
protocones.

The 
side 
of 
the 
rostrum 
is 
relatively 
short 
and 
high,
and 
the 
dorsal 
half 
of 
the 
maxilla 
is 
concave. 
Because 
of 
this, 
the 
dorsal 
surface 
of 
the 
snout, 
capped 
by 
the 
nasals, 
must 
have 
been 
high 
and 
narrow, 
as 
in 
?Harpagolestes 
orientalis 
Szalay 
and 
Gould. 
In 
the 
skulls 
of 
H. 
uintensis, 
H. 
macrocephalus, 
and 
?H. 
orientalis 
the 
anterior 
border 
of 
the 
orbit 
is 
over 
the 
center 
of 
M“. 
In 
the 
Texas 
specimen, 
however, 
the 
orbit 
is 
further 
forward, 
over 
the 
center 
of 
Ml 
. 
As 
in 
other 
mesonychids, 
the 
lacrimal 
spreads 
over 
the 
anterior 
border 
of 
the 
orbit 
onto 
the 
side 
of 
the 
face. 
There 
are 
two 
openings 
for 
the 
lacrimal 
duct 
separated 
by 
a 
bony 
septum. 
Immediately 
dorsal 
to 
these 
on 
the 
anterior 


8 



13143

immanis 


H. 
AMNHtype 
28.2 
19.3 
21.7 
10.8 
30.4 
15.1 
32.8 
16.3 
30.2 
15.8 
2e

leotensis 
11778 


H. 
CM 
type 
20.11.5 
29.4 
15.1 
30.6 
14.0 
97 
68 
brevipes 
13098 


H.YPMtype 
23 
15.5 
32.2 
16 
36 
19 
37a 
52 


uintensis 
1878 


H. 
AMNH16.7 
28.3 
30 
24.3 
103 
73 
uintensis 


H. 
CM 
2113261530 
16.529 
15.5 
10066 
10072

41715-8). 
uintensis 
20.7 
28 
27.5 
31.5 
47

YPM

(TMMH. 


uintensis 
11659 


H. 
type 
31.7 
28.8
HarpagolestesPU 
28.0 
14.2 
14.7 
15.0 


of

dentitionuintensis 
31.5 
16e

left

uppercf. 
40498-3

ofH. 
TMMright 
10.4 
6.9 
11.0 
9.6 
31.8 
22.0 
1.3 
8.4 
23.7a 
29.4 
15.5 
3le 
16.9 
28.7a 
94e 
51

1

H. 
AP 
AP 
AP 
AP 
AP 
AP 
AP 
AP 
depth,M 


TTT 
TTTTT 


1H

3.—Measurements 
2961 


1

2Z 
2) 
2zP44

Table 
II3 
C 
P, 
Pp. 
P-PJawposterior

M 


9 



Table 
4.—Measurements 
of 
upper 
dentition 
of 
Harpagolestes 
(TMM 
41715-8). 


AP 
T 


Canine 
alveolus 
29.8 
17.7P* 
alveolus 
13.9 
9.6 
P2 
crown 
18.9 
10.5 
P 
3 
crown 
23.0 
12.4 
P4 
crown 
23.0 
20.5 
M 
1 
crown 
23.3 
24.3 
M2 
crown 
15.0 
18.0 


Distance 
from 
anterior 
P1 
alveolus 
to 
posterior 
M 
: 
126.5 


Figure 
5.—Harpagolestes 
cf. 
H. 
uintensis, 
TMM 
40498-3, 
Candelaria 
local 
fauna. 
A. 
Left 
P 
lateral 
view. 
B-C. 
Right 
anterior 
ramus 
of 
mandible 
with 
incisors, 
canine, 
P,, 
broken 
P, 
and 
P„, 
dorsal 
(B) 
and 
lateral 
(C) 
views. 
Matrix 
and 
broken 
bone 
or 
tooth 
surfaces 
not

1 
34 
shaded. 


10 



Figure 
6.—Harpagolestes 
sp. 
TMM 
41715-8, 
Skyline 
channels. 
Lateral 
and 
occlusal 
views. 
11 



orbital 
border 
is 
an 
oval 
knob 
of 
bone. 
The 
infraorbital 
foramen 
opens 
above 
the 
posterior 
root 
of 
P 
at 
the 
narrowest 
part 
of 
the 
snout. 
The 
palate 
is 
relatively 
narrow 
and 
is 
anteroposteriorly

shorter 
than 
in 
other 
specimens 
of 
Harpagolestes. 
This 
is 
largely 
because 
of 
the 
anterior 
position 
of 
the 
internal 
nares, 
with 
the 
anterior 
border 
opposite 
the 
posterior 
root 
of 
M 
l 


The. 
dental 
arcade 
is 
strongly 
arcuate 
in 
the 
vertical 
plane. 
This 
is 
a 
common 
feature 
in 
Harpagolestes, 
which 
has 
been 
related 
to 
a 
mechanism 
for 
holding 
bones 
in 
place 
in 
the 
lower 
jaw 
for 
crushing 
by 
the 
upper 
posterior 
premolars 
and 
molars 
(Szalay 
and 
Gould, 
1966, 
p. 
160). 
The 
roots 
of 
the 
anterior 
premolars 
are 
enclosed 
in 
the 
lateral 
margins 
of 
the 
maxillary, 
which 
projects 
ventrally 
beyond 
the 
anterior 
palate. 
P 
1 
likewise 
extends 
ventrally 
well 
beyond 
the 
palate; 
however, 
the 
bony 
borders 
of 
the 
aveoli 
do 
not 
extend 
over 
much 
of 
the 
roots 
of 
this 
tooth, 
leaving 
them 
(at 
least 
in 
this 
individual) 
exposed 
well 
beyond 
the 
alveolus. 
This 
effect 
may 
be 
exaggerated 
by 
the 
crushing 
of 
the 
maxillary 
in 
the 
vicinity 
of 
the 
roots 
of 
P 
4 
and 
consequent 
ventral 
displacement 
of 
P 
. 
The 
alveoli 
of 
M 
1 
and 
M‘ 
are 
recessed 
above 
the 
palate. 
This 
is 
especially 
evident 
in 
M“, 
which 
is 
in 
a 
position 
which 
in 
most 
mammals 
might 
be 
occupied 
by 
an 
unerupted 


posterior 
molar 
and 
which 
would 
be 
too 
high 
to 
allow 
occlusion. 
In 
this 
specimen 
there 
are 
obvious 
wear 
facets 
formed 
by 
abrasion 
of 
M 
against 
the 
posterior 
end 
of 
M“ 
and 
the 
anterior 
end 
of 


. 


The 
dentition 
fits 
the 
description 
of 
?Harpagolestes

orientalis 
given 
by 
Szalay 
and 
Gould 
(1966), 
which 
includes 
the 
only 
other 
Harpagolestes 
specimens 
with 
lightly 
worn 
upper 
teeth. 
The 
canine 
was 
large, 
oval 
in 
cross 
section, 
and 
may 
have 
been 
somewhat 
procumbent. 
P 
1 
has 
a 
single 
root 
and 
probably 
a 
single 
cusp. 
P" 
is 
a 
laterally 
compressed 
conical 
tooth 
with 
two 
roots. 
The 
axis 
of 
one 
of 
the 
roots 
curves 
posterad, 
and 
there 
is 
a 
small 
accessory 
cusp 
on 
the 
posterior 
side. 
This 
tooth 
resembles 
P2 
of 
Hyaenodon. 
P3 
consists 
of 
a 
large 
paracone, 
though 
not 
as 
tall 
as 
that 
on 
P". 
with 
a 
small 
cusp 
(parastyle) 
on 
the 
anteromedial 
comer 
and 
a 
short 
metacone 
with 
a 
laterally 
bulbous 
base. 
P4 
is 
dominated 
by 
the 
large, 
bluntly 
conical 
protocone 
and 
by 
the 
slightly 
taller 
paracone. 
The 
protocone 
is 
separated 
from 
the 
rest 
of 
the 
tooth 
by 
a 
deep 
V-shaped 
valley. 
Appressed 
to 
the 
posterior 
side 
of 
the 
paracone 
is 
a 
poorly 
developed 
metacone. 
The 
buccal 
side 
of 
the 
tooth 
is 
flanked 
anteriorly 
by 
a 
small 
parastyle 
and 
posteriorly 
by 
an 
even 
smaller 
metastyle. 
M 
1 
is 
the 
largest 
of 
the 
cheek 
teeth. 
The 
protocone 
is 
surprisingly 
large. 
It 
is 
tall 
and 
bullet 


a

shaped, 
and 
is 
separated 
from 
the 
rest 
of 
the 
tooth 
by 
valley 
deeper 
than 
that 
on 
P4 
. 
The 
paracone 
is 
nearly 
twice 
the 
size 
of 
the 
metacone, 
though 
these 
two 
cusps 
are 
distinct 
and 
well 
separated. 
There 
was 
strong

a 
parastyle 
(the 
enamel 
has 
been 
chipped 
from 
it) 
and 
a 
small 
metastyle, 
more 
a 
bulge 
than 
a 
cusp. 
M~ 
is 
much 
smaller 
than 
Ml 
. 
The 
protocone 
is 
small 
and 
low. 
and 
is 
connected 
to 
the 
medial 
side 
of 
the 
paracone 
by 
a 
small 


ridge, 
with 
no 
intervening 
valley. 
The 
paracone

dominates 
the 
tooth, 
but 
is 
not 
particularly 
large. 
The 


metacone 
is 
a 
tiny 
cusp 
appressed 
to 
the 
posterior 
side 
of 
the 
paracone 
near 
its 
base. 
This 
M 
is 
more 
reduced 
than 
that 
of 
any 
other 
specimen 
of 
Harpagolestes.

No 
lower 
dentition 
was 
found, 
but 
it 
appears 
from 
the 
wear 
facets 
on 
the 
existing 
teeth 
that 
there 
was 
a 
complete 
series 
of 
seven 
lower 
cheek 
teeth. 


Relationships. 
—On 
the 
basis 
of 
size 
alone 
(Table 
4), 
as 
well 
as 
the 
measurements 
of 
the 
cheek 
teeth 
and 
the 
relatively 
anterior 
position 
of 
the 
orbit 
and 
internal 
nares, 
this 
individual 
is 
much, 
more 
similar 
to 
?Harpagolestesorientalis 
than 
to 
the 
larger 
species 
of 
Harpagolestes. 
refer 
it 
to 
Harpagolestes 
on 
the 
basis 
of 
the 
dental 
formula 
(P4 
if 
my 
interpretations 
are 
correct), 
the 
arcuate 
form, 
of 
the 
tooth 
row, 
and 
the 
small 
size 
of 
the 
metacone 
on 
M 
2 
(which 
separates 
it 
clearly 
from 
Mesonyx). 
The 
only 
other 
very 
late 
Uintan 
(=Duchesnean) 
mesonychid 
is 
Hessolestes. 
which 
apparently 
had 
lost 
or 
was 
in 
the 
process 
of 
losing 
its 
last 
lower 
molar. 


The 
similarity 
of 
this 
specimen 
to 
?Harpagolestes 
orientalis, 
which 
is 
known 
only 
from 
specimens 
found 
in 
the 
late 
Eocene 
Ulan 
Shireh 
deposits 
at 
Chimney 
Butte, 
North 
Mesa, 
Shara 
Murun 
region 
of 
Inner 
Mongolia, 
poses 
an 
interesting 
biogeographical 
problem. 
The 
Texas 
specimen 
was 
found 
at 
a 
stratigraphic 
level 
which 
also 
produced 
the 
earliest 
record 
of 
Hyaenodon 
(ct. 
H. 
vetus) 
in 
Texas. 
Hyaenodon 
is 
commonly 
accepted 
as 
an 
immigrant 
from 
Eurasia. 
Does 
this 
specimen 
indicate 
a 
coincidental 
immigration 
from 
Asia? 
Or, 
as 
Harpagolestes 
is 
best 
known 
from 
earlier 
beds 
(middle 
Eocene) 
in 
North 
America, 
do 
the 
Asian 
specimens 
represent 
a 
migration 
of 
the 
last 
North 
American 
Harpagolestes 
into 
Asia? 
As 
Szalay 
and 
Gould 
(1966)
pointed 
out 
in 
their 
description 
of 
Asian 
mesonychids, 
not 
enough 
is 
known 
of 
Asian 
middle 
Eocene 
mesonychids 
to 
allow 
a 
definitive 
answer 
to 
these 
questions. 


Genus 
Hessolestes 
Peterson, 
1931a 


Hessolestes 
sp. 


Whistler 
Squat 
and 
Serendipity 
local 
faunas 


Figure 
7, 
Table 
5 
Type 
.—.Hessolestes 
ultimas, 
CM 
11763, 
posteriorright 
mandible 
with 
?M 
, 
Lapoint 
Member 
of 
Duchesne 
River 
Formation, 
Utah. 


Referred 
material.—TMM 
41784-1, 
left 
ramus 
with 
?P 
, 
?MV 
41723-1, 
maxillary 
fragment 
with 
partial 
P4 
M 
; 
41576-6, 
partial 
?M" 



Stratigraphic 
position.—Approximately 
12.1 
m 
(40feet) 
above 
the 
base 
of 
the 
Devil’s 
Graveyard 
Formation 
at 
locality 
41576 
southwest 
of 
“Wax 
Camp” 
on 
Alamo 
de 
Cesario 
Creek, 
Brewster 
Co., 
Texas. 
Locality 
41576 
is 
mapped 
in 
Stevens 
et 
al. 
(1984, 
Fig. 
7) 
and 
is 
at 
approximately 
the 
same 
stratigraphic 
level 
as 
the 
Whistler 
Squat 
quarry. 
Localities 
41723 
and 
41784 
are 
in 
the 
titanothere 
channels 
shown 
in 
Stevens 
et 
al. 
(1984, 
Fig.
5E 
and 
D). 
Locality 
41723 
is 
near 
the 
top 
of 
Titanothere 
Hill 
which 
is 
approximately 
200 
feet 
(60.9 
m) 
above 
the 


12 



Figure 
7.—Hessolestes 
sp. 
A. 
CM 
11763, 
?M 
of 
type 
of 
Hessolestes 
ultimus, 
medial 
view. 
B. 
TMM 
41784-1, 
?M 
, 
medial 
view,
reversed. 
C-D. 
TMM 
41723-1, 
P3-M', 
lateral 
(C) 
and 
ventral 
(occlusal) 
(D) 
views, 
anterior 
to 
right, 
matrix 
and 
Broken 
surfaces 
unshaded; 
short 
diagonal 
lines 
in 
lateral 
view 
indicate 
position 
of 
infraorbital 
foramen. 
E. 
TMM 
41784-1, 
ramus 
of 
mandible 
with 
presumed 
P 
and 
M,, 
dorsal 
and 
lateral 
views; 
vertical 
lines 
indicate 
matrix. 


13 



Table 
5.—MeasurementsofdentitionsofHessolestes. 


H. 
ultimus 
CM 
11763 
p4 


AP 


T 


M1 
T 


P
P
3T 
P, 
AP

d 
AP 


M
M
2 
AP 
22.8 


L 
T 
11.5 


Jaw 
depth

between 
P,4 
& 
M,1 


base 
of 
the 
Devil’s 
Graveyard 
Formation. 
Locality 
41784, 
Unio 
Cliff, 
is 
in 
a 
channel 
(one 
of 
the 
Titanothere 
Channels) 
that 
cuts 
through 
the 
Strawberry 
tuff, 
152 
feet 
(46,3 
m) 
above 
the 
base 
of 
the 
Devil’s 
GraveyardFormation. 


Age. 
—Early 
to 
late 
Uintan, 
late 
Eocene. 


Description. 
—These 
specimens 
belong 
to 
a 
mesonychine 
larger 
than 
Mesonyx 
but 
smaller 
than 
the 
North 
American 
species 
of 
Harpagolestes. 
The 
posterior 
part 
of 
the 
mandible 
is 
badly 
crushed 
and 
distorted, 
and 


the 
front 
part 
of 
the 
mandible 
is 
missing. 
The 
central 
partof 
the 
ramus 
is 
much 
shallower 
than 
in 
any 
known 
specimen 
of 
Harpagolestes, 
but 
approximates 
the 
depth 
of 
the 
mandible 
of 
Mesonyx 
obtusidens. 
The 
identification 
of 
the 
teeth 
is 
uncertain, 
but 
by 
comparison 
with 
the 
type 
of 
Hessolestes 
ultimus 
I 
tentativelyconsider 
the 
partial 
premolar 
to 
be 
the 
aveoli 
to 
accommodate 
? 
and 
and 
the 
posterior 
molar 
to 
be

4

M 
. 
The 
symphysis 
probably 
began 
within 
a 
few 
millimeters 
of 
the 
broken 
anterior 
end 
of 
the 
specimen; 
in 
Harpagolestes 
the 
posterior 
end 
of 
the 
symphysis 
is 
opposite 
the 
anterior 
end 
of 
P 
. 
A 
mental 
foramen 
is 
present 
near 
the 
anterior 
root 
of 
the 
partial 
premolar. 
In 
Harpagolestes 
the 
posterior 
mental 
foramen 
is 
below 
the 
anterior 
P„.

3’

The 
preserved 
portion 
of 
shows 
it 
to 
have 
been 
a 
narrow 
tooth. 
There 
is 
a 
diastema 
of 
about 
10 
mm 
between 
the 
alveoli 
of 
P„4 
and 
M,,1 
but 
the 
crowns 
of 
these 
teeth 
may 
have 
extended 
far 
enough 
to 
fill 
this 
gap. 


The 
posterior 
molar 
of 
41784-1 
differs 
from 
the 
presumed 
of 
Hessolestes 
ultimus 
only 
in 
having 
a 
slightly 
smaller 
hypoconid 
and 
in 
being 
slightly 
more 
worn. 
There 
is 
no 
sign 
that 
M 
was 
present, 
though 
there 
is 
room 
for 
a 
small 
alveolus 
beneath 
some 
matrix 
which 
could 
not 
be 
removed 
because 
it 
supports 
the 
remainingtooth. 


The 
upper 
tooth 
fragments 
are 
too 
worn 
and 
broken 
for 
firm 
identification. 
I 
very 
tentatively 
assign 
them 
to 
Hessolestes 
because 
they 
appear 
to 
be 
the 
proper 
size 
to 
match 
the 
Unio 
Cliff 
jaw. 


H. 
sp. 
?H. 
sp. 
TMM 
41784-1 
TMM 
41723-1 
19.6e 


20.0 
19.5e 
10.5e 
20.8a 
24a 


21.8 


11.0 
37 
Relationships.—The 
occurrence 
of 
Hessolestes 
in 
the 
Whistler 
Squat 
and 
Serendipity 
local 
faunas 
is 
unexpectedbecause 
the 
type 
and 
only 
other 
known 
specimen 
is 
from 
the 
considerably 
later 
Duchesne 
River 
Formation 
of 
Utah. 
The 
material 
of 
this 
genus 
is 
tantalizing 
because 
it 
is 
so 
incomplete. 
If 
the 
Texas 
material 
is 
indeed 
Hessolestes 
, 
then 
that 
genus 
was 
a 
contemporary 
of 
Harpagolestesduring 
the 
Uintan. 


Order 
Creodonta 
Cope, 
1875 


Family 
Hyaenodontidae 
Leidy, 
1869 


Subfamily 
Hyaenodontinae 
Leidy, 
1869 


Tribe 
Proviverrini 
(Schlosser, 
1886) 


Genus 
Proviverra 
Rutimeyer 
1862 


?Proviverra 
major 
(Wortman 
1902) 


Whistler 
Squat 
local 
fauna 


Figure 
8, 
Table 
6 


—

Synonyms. 
Sinopa 
major 
Wortman 
1902;
?Proviverra 
major 
(Wortman), 
Van 
Valen, 
1965. 
Type.—YPM 
11878, 
right 
mandible 
with 
P?-M 
,
Bridger 
Formation 
(Church 
Buttes, 
probably 
Bridger 
B 
level), 
Wyoming, 
Middle 
Eocene. 
Referred 
material.—TMM 
41466-9, 
partial 
left 
ramus 
with 
M 
-M; 
TMM 
41576-5, 
right 
M^

Stratigraphic 
2position.—Locality 
41466 
is 
approximately 
300 
yards 
(274.3 
m) 
east 
of 
the 
Whistler 
Squat 
quarry. 
This 
locality 
overlies 
the 
“variegated 
beds” 
which 
in 
turn 
lie 
directly 
on 
marine 
Cretaceous 
with 
no 
basal 
Tertiary 
conglomerate 
present. 
TMM 
41466 
is 
approximately 
30.4 
m 
(100 
feet) 
above 
the 
base 
of 
the 
Devil’s 
Graveyard 
Formation 
and 
at 
approximately 
the 
same 
stratigraphic 
level 
as 
the 
Whistler 
Squat 
quarry. 
The 
strati


14 



Figure 
8.—?Proviverra 
major 
(Wortman). 
TMM 
41466-9, 
Whistler 
Squat 
local 
fauna. 
Mandible 
fragment 
with 
M 
-M 
. 
A. 
Occlusal 
view. 
B. 
External 
view. 
C. 
Internal 
view. 


15 



Table 
6.—Measurements 
of 
lower 
dentitions 
of 
Proviverra 
major. 


TMM 
41466-9 
TMM 
41576-5 
YPM 
11878 
USNM 
112-41 
USNM 
98-41 


Cast 
of 
type, 
Bridger 
“D” 
Bridger 
“C” 


from 
Bridger 
“B” 


M. 
AP 
9.7 
8.8 
9.5 
8.4 
T 
5.5 
4.8 
5.4 
5.2 
M, 
AP 
11.4 
10.3 
9.4 
10.6 
9.1 
2 
T 
6.5 
6.1 
5.5 
5.8 
5.9 
Jaw 
depth 
23.1 
19.8 
18.8 
18.8 
between 
M 
& 
M

0 


graphic 
position 
of 
TMM 
41576 
is 
given 
in 
the 
description 
of 
Hessolestes, 
above. 


Age.—Early 
Uintan. 


Description.-—The 
badly 
fractured 
mandible 
(Fig. 
8)
has 
about 
the 
same 
thickness 
and 
depth 
as 
the 
equivalentbone 
in 
Canis 
latrans 
(Table 
6). 
Two 
mental 
foramina 
are 
present, 
one 
below 
the 
posterior 
alveolus 
of 
P 
the 
other 
below 
the 
alveolus 
of 
P 
. 
, 


The 
alveoli 
of 
P,-?4 
are 
preserved 
but, 
because 
of 
matrix 
fillings 
and 
fracturing, 
are 
not 
clearly 
delineated. 
There 
appears 
to 
have 
been 
no 
diastema 
between 
? 
and

l

P2‘ 


M 
is 
heavily 
worn, 
especially 
on 
the 
trigonid. 
The 
talonid 
is 
slightly 
larger 
in 
area 
than 
the 
trigonid. 
and 
had 


two 
main 
cusps, 
a 
hypoconid 
and 
an 
entoconid. 
The 
talonid 
basin 
is 
triangular 
and 
was 
enclosed 
prior 
to 
the 
heavy 
wear 
on 
the 
hypoconid. 
The 
widest 
part 
of 
the 
is 
across 
the 
talonid. 
A 
distinct 
labial 
cingulum 
is 
present.

Both 
are 
larger 
than 
the 
M 
of 
41466-9. 
The 
trigonid 
is 
farger 
than 
that 
of 
M 
, 
and 
is 
much 
less 
worn. 
Occlusion 
with 
the 
upper 
camassial 
has 
worn 
the 
enamel 
from 
the 
labial 
sides 
of 
the 
protoconid 
and 
paraconid, 
but 
the 
tips 
of 
these 
cusps 
are 
not 
heavily 
worn 
on 
41466-9. 
Another 
less 
distinct 
wear 
facet 
is 
present 
on 
the 
posteriorside 
of 
the 
protoconid, 
extending 
down 
into 
the 
sulcus 
obliquus. 
The 
enamel 
cap 
on 
the 
metaconid 
is 
still 
intact 
on 
41466-9 
but 
worn 
on 
41576-5, 
mostly 
on 
the 
posteriorside. 
The 
talonid 
is 
a 
simple 
enclosed 
basin. 
The 
hypoconid 
and 
entoconid 
are 
only 
noncuspate 
high 
points 
on 
lingual 
and 
labial 
ridges. 
The 
hypoconid 
ridge 
is 
badly 
worn 
on 
both 
specimens. 
There 
is 
no 
hypoconulid. 
The 
greatest 
transverse 
width 
of 
M 
0 
is 
across 
the 
protoconidand 
metaconid. 
A 
labial 
cingulum 
is 
present 
on 
41466-9 
but 
not 
on 
41576-5. 


The 
roots 
of 
M, 
preserved 
on 
41466-9 
suggest 
a 
tooth 
slightly 
smaller 
than 
Mr 
A 
trigonid 
of 
another 
tooth 
from 
this 
locality, 
lacking 
the 
talonid 
or 
any 
definite 
contact 
with 
the 
broken 
roots 
on 
the 
mandible, 
may 
be 


the 
M„ 
of 
this 
individual. 


—

Discussion. 
I 
follow 
Van 
Valen 
(1965) 
in 
referring 
Wortman's 
Sinopa 
major 
to 
Proviverra 
with 
a 
query.
There 
is 
little 
to 
distinguish 
these 
Texas 
specimens 
from 
the 
type 
and 
referred 
specimens 
of 
Proviverra 
major 


from 
the 
Bridger 
Formation 
of 
Wyoming 
(middleEocene). 


TMM 
41466-9 
differs 
from 
the 
type 
(YPM 
11878) 
of 
?P. 
major 
in 
having 
slightly 
larger 
tooth 
size, 
more 
massive 
mandible, 
no 
diastema 
between 
Pj 
and 
and 
weaker 
labial 
cingula 
on 
Mj 
and 
Mv 
All 
of 
these 
differences 
are 
minor 
and 
are 
of 
a 
magnitude 
not 
greaterthan 
can 
easily 
be 
found 
in 
recent 
carnivore 
populations,
such 
as 
Canis 
latrans 
which 
has 
a 
mandible 
of 
similar 
size. 


Tribe 
Hyaenodontini 
(Leidy, 
1869) 


Genus 
Hemipsalodon 
Cope, 
1885 


Type.—Hemipsalodon 
grandis 
Cope, 
1885 


Generic 
characters.—Very 
large 
(skull 
length 
370450 
mm). 
Distinguished 
from 
Hyaenodon 
by 
the 
presence 
of 
M 
3. 
Distinguished 
from 
Pterodon 
(especiallythe 
type, 
P. 
dasyuroides) 
by 
its 
much 
larger 
size, 
by 
the 
presence 
of 
separated 
paracone 
and 
metacone 
on 
M‘ 
and 
a 
larger 
M 
3, 
by 
the 
presence 
of 
a 
distinct 
protocone 
on 
P 
, 
by 
the 
rotation 
of 
upper 
molars 
along 
the 
proximodistal 
axis 
during 
ontogeny, 
and 
by 
the 
posteriorposition 
of 
the 
constriction 
of 
the 
braincase, 
near 
the 
posterior 
roots 
of 
the 
zygomatic 
arches. 


Discussion.—The 
characters 
given 
above 
are 
modified 
from 
Mellett 
(1969). 
Savage 
(1965) 
considered 
Hemipsalodon 
to 
be 
a 
synonym 
of 
Pterodon. 
I 
agreewith 
Mellett 
that 
the 
New 
World 
Hemipsalodon 
lineageis 
generically 
distinct 
from 
the 
Old 
World 
Pterodon 


lineages. 
I 
am 
not 
ready 
to 
accept 
a 
separate 
ancestry 
for 


Hemipsalodon 
and 
Pterodon 
at 
the 
proviverrine 
level 
of 
development, 
as 
was 
Mellett, 
because 
of 
the 
numerous 
similarities 
between 
the 
two 
genera, 
particularly 
in 
the 
overall 
morphology 
of 
the 
dentition 
and 
form 
of 
the 
basicranium. 


Hemipsalodon 
viejaensis 
n. 
sp 


Porvenir 
local 
fauna 


Figs. 
9-14, 
Table 
7 
Type.—TMM 
40263-1, 
skull. 
Referred 
material.—TMM 
40688-50, 
maxillary 


16 



Figure 
9.—Hemipsalodon 
viejaensis 
n. 
sp., 
type, 
TMM 
40263-1, 
Porvenir 
local 
fauna. 
Skull, 
dorsal 
view. 
Abbreviations: 
F, 
frontal; 
J 
jugal;L, 
lacrimal; 
M, 
maxillary; 
MP, 
mastoid 
process; 
N, 
nasal; 
P, 
parietal; 
PM, 
premaxillary 


Figure 
10.—Hemipsalodon 
viejaensis 
n, 
sp.. 
Porvenir 
local 
fauna. 
A. 
TMM 
40263-1, 
type, 
skull, 
lateral 
view 
(zygomatic 
arch 
restored 
from 
right 
side). 
B. 
FMNH 
PM 
110, 
partial 
left 
ramus 
of 
mandible 
with 
alveoli 
and 
anterior 
M 
. 


17 



Table 
7.—Measurements 
of 
specimens 
of 
Hemipsalodon 
viejaensis 


TMM 
40263-1 
FMNH 
PM 
109 
FMNH 
PM 
96 
FMNH 
PM 
111 
FMNH 
PM 
110 
typeright 
left 
a 
AP 
23e 
p1 
p2 
p3 
p4 
M 
1 
M2 
M3 
T 
AP 
T 
AP 
T 
AP 
T 
AP 
T 
AP 
T 
AP 
T 
AP 
16e 
18.9 
10.0 
23.1 
17.0 
24.3 
20.5 
20.2 
17.7 
27.3 
22.0 
10.0 
15.5e 
8.0e 
22.2 
20.2 
21.2 
18.0 
25.9 
22.0 
10.3 
10.2 
T 
23.6 
23.5 
25.1 
/C 
AP 
T 
M*-M3 
p! 
-m3 
P,-M,
Total 
skull 
lengthMaximum 
width 
60.0 
142e 
378 
225 
60.5 
146.0 
30.0 
21.5 
61e 
154a 
across 
zygomasMaximum 
width 
138 
across 
mastoids 
Maximum 
length 
ofjawJaw 
depth 
at 
M 
290a 
50 


Figure 
11 
—Hemipsalodon 
viejaensis 
n. 
sp., 
type, 
TMM 
40263-1, 
Porvenir 
local 
fauna. 
Skull, 
posterior 
view. 


18 



l

fragment 
with 
alveoli 
for 
dP4 
, 
M 
unerupted 
M2. 
FMNH:PM 
109, 
M 
3. 
FMNH:PM 
110,
, 
broken 
mandible 
with 
partial 
and 
alveoli 
of 
all 
postcanine 
teeth. 
FMNH:PM 
111, 
lower 
canine. 
FMNFPPM 
451, 
fragmentof 
anterior 
mandible. 


Stratigraphic 
position. 
—The 
type 
is 
from 
the 
lower 
part 
of 
Chambers 
Formation, 
Vieja 
Group, 
Presidio 
County, 
Texas, 
Porvenir 
local 
fauna, 
TMM 
40688-50 
and 
FMNH 
specimens 
are 
from 
Blue 
Cliff 
Horizon 
(Wilson,

1977b). 


Etymology.—Native 
or 
resident 
oftheSierra 
Vieja. 


Diagnosis.—Smaller 
than 
Hemipsalodon 
grandis,
skull 
narrower. 
Nasals 
as 
in 
H. 
grandis 
but 
narrower, 
snout 
relatively 
slender, 
frontals 
not 
as 
broad 
as 
in 
H. 
grandis 
but 
more 
anteroposteriorly 
elongate. 
Premolars 
less 
crowded, 
relatively 
larger. 
P3 
may 
have 
a 
small 
protocone. 
P4 
has 
a 
large 
protocone. 
Maxillary 
tooth 
rows 
less 
divergent 
posteriorly. 
M 
3 
without 
lingualcingulum. 


Description. 
—More 
specimens 
of 
Hemipsalodon 
are 
known 
from 
the 
lower 
Chambers 
Formation 
than 
from 
any 
other 
local 
area 
in 
North 
America. 
Specimens 
of 
Hemipsalodon 
from 
Saskatchewan, 
Oregon 
and 
Wyoming 
are 
consistently 
larger 
and 
closely 
match 
the 
excellent 
skull 
of 
H. 
grandis 
from 
Clamo, 
Oregon(Mellett, 
1969). 
The 
specimens 
from 
the 
Chambers 
Formation 
are 
also 
consistent 
in 
size 
(Table 
7), 
and 
there 
is 
little 
overlap 
between 
specimens 
of 
H. 
grandis 
and 
H. 
viejaensis. 
The 
type 
skull 
of 
H. 
viejaensis 
is 
an 
adult 
with 
worn 
teeth, 
complete 
except 
for 
sagittal 
crest, 
left 
zygomatic 
arch, 
and 
portions 
of 
the 
teeth, 
and 
is 
almost 
undistorted. 


Because 
of 
the 
excellent 
preservation 
of 
this 
specimenI 
include 
a 
description 
of 
each 
of 
the 
skull 
bones, 
where 
possible. 
Crushing 
and 
cracking 
of 
the 
bones 
in 
the 
orbital 
region 
and 
fusion 
of 
bones 
in 
the 
posterior 
and 
basicranial 
portions 
of 
the 
skull 
make 
delineation 
of 
some 
sutures 
difficult 
or 
impossible, 
but 
in 
general, 
fusion 
along 
sutures 
is 
less 
than 
is 
present 
in 
living 
bears 
of 
similar 
size 
and 
maturity.

In 
dorsal 
view 
(Fig. 
9) 
the 
dominant 
features 
of 
the 
skull 
are: 
the 
massive 
facial 
region, 
including 
a 
powerful 
snout 
expanded 
anteriorly 
for 
the 
large 
canines 
and 
posteriorly 
in 
the 
broad 
antorbital 
region; 
wide 
and 
strongbut 
not 
massive 
zygomatic 
arches; 
small 
braincase 
with 
an 
hourglass 
outline 
in 
dorsal 
view; 
small 
and 
incomplete 
transverse 
occipital 
crests; 
and 
very 
large 
and 
laterally 
projecting 
conjoined 
mastoid 
and 
paroccipital 
processes.
The 
long 
zygomatic 
arches, 
widest 
posteriorly, 
give 
a 
triangular 
shape 
to 
the 
outline 
of 
the 
skull. 
The 
skull 
is 
approximately 
the 
size 
of 
that 
of 
Ursus 
arctos. 
However, 
the 
skull 
of 
Hemipsalodon 
gives 
much 
less 
of 
the 
impression 
of 
great 
strength 
suggested 
by 
the 
grizzlyskull. 
It 
lacks 
the 
heavy 
postorbital 
processes, 
well 
defined 
and 
rugose 
muscle 
attachments 
especially 
for 
the 
temporal, 
pterygoid, 
and 
masseter 
musculature, 
and 
compact 
appearance 
of 
the 
grizzly. 
It 
is 
likewise 
much 
less 
massive 
than 
either 
the 
Clamo 
(OMSI 
619) 
or 
Cypress 
Hills, 
Saskatchewan, 
(NMC 
8791) 
skulls 
of 
H. 


grandis. 
The 
Texas 
skull 
is 
more 
slender 
although 
robust, 
more 
Pterodon-like, 
than 
H. 
grandis.

The 
ventral 
surface 
of 
the 
orbit 
is 
a 
broad 
shelf 
similar 
to 
that 
seen 
in 
Hyaenodon, 
but 
more 
exaggeratedposteriorly 
because 
of 
the 
presence 
of 
M 
3. 
The 
temporal

fossa 
is 
enormous. 
Its 
anterior 
extension 
to 
the 
rudimentary 
postorbital 
process 
lies 
above 
the 
posterior 
part 
of 
M 
2, 
farther 
forward 
than 
is 
seen 
in 
most 
fissipedcarnivores, 
though 
those 
of 
Procyon 
and 
Ailuropodaapproach 
it. 
However, 
because 
of 
the 
anterior 
expansionof 
the 
nasal 
portion 
of 
the 
braincase 
and 
posterior 
nasal 
cavities, 
the 
anterior 
portion 
of 
the 
temporal 
fossa 
is 


convex 
laterally, 
rather 
than 
strongly 
concave 
as 
in 
Hyaenodon. 


The 
shape 
of 
the 
braincase 
is 
unusual. 
In 
Hyaenodonand 
most 
if 
not 
all 
fissiped 
carnivores 
the 
narrowest 
partof 
the 
braincase 
is 
immediately 
behind 
the 
postorbital 
processes, 
in 
the 
frontal 
bones. 
This 
seems 
to 
be 
the 
case 
in 
Pterodon 
as 
well, 
though 
the 
constriction 
is 
further 
posterior 
in 
Pterodon 
than 
in 
Hyaenodon 
and 
seems 
to 
be 
near 
the 
frontal-parietal 
suture. 
In 
both 
species 
of 
Hemipsalodon 
the 
greatest 
constriction 
is 
about 
at 
the 
middle 
of 
the 
parietals, 
adjacent 
to 
the 
posterior 
root 
of 
the 
zygomatic 
arches, 
and 
involves 
the 
dorsal 
portion 
of 
the 
squamosals. 


In 
lateral 
view 
(Fig. 
10) 
the 
skull 
is 
very 
bear-like, 
though 
with 
a 
more 
dorsally 
convex 
snout 
and 
more 
ventrally 
placed 
orbits. 
The 
face 
is 
long 
and 
deep. 
There 
is 
no 
postorbital 
projection 
on 
the 
jugal 
and 
only 
a 
small 
one 
on 
the 
frontal. 


In 
posterior 
view 
(Fig. 
11) 
the 
skull 
is 
dominated 
bythe 
peculiarly 
restricted 
form 
of 
the 
transverse 
occipital 
crests, 
the 
very 
large 
combined 
mastoid-paroccipital 
processes, 
the 
large 
occipital 
condyles 
and 
foramen 
magnum, 
and 
the 
lateral 
extension 
of 
the 
posterior 


zygomatic 
arches. 
The 
distribution 
of 
muscles 
supporting 
the 
head 
must 
have 
been 
considerably 
different 
from 
that 
in 
living 
large 


bears. 
Instead 
of 
attaching 
to 
a 
broad 
surface 
rimmed 
dorsally 
by 
the 
transverse 
occipital 
crests, 
the 
neck 
muscles 
attached 
to 
two 
very 
distinct 
areas, 
with 
apparently 
much 
less 
total 
surface 
area 
than 
on 
a 
grizzlyof 
equivalent 
size. 
Part 
of 
the 
difference 
may 
be 
ascribed 
to 
the 
much 
smaller 
braincase 
in 
Hemipsalodon, 
and 
part 


to 
a 
greater 
differentiation 
between 
the 
muscles 
which 
attach 
just 
below 
the 
external 
occipital 
protuberance, 
and 
which 
raise 
the 
head 
(probably 
especially 
rectus 
capitisdorsalis, 
semispinalis 
capitis, 
spleniusand 
trapezius) 
and 
those 
which 
attach 
laterally, 
to 
, 
turn 
the 
head 
(obliquus 
capitis, 
rectus 
capitis 
lateralis, 
and 
sternomastoideus). 


In 
ventral 
view 
(Fig. 
12) 
the 
palate 
is 
long 
and 
posteriorly 
widened. 
The 
ventrolateral 
surfaces 
of 
the 
maxillaries 
are 
broadly 
visible. 
The 
nasopharyngeal 
fossa 
is 
long, 
narrow, 
and 
tubular. 
The 
palatines 
and 
?pterygoids 
converge 
ventrally 
but 
do 
not 
meet, 
forming 
a 
partial 
floor 
with 
parallel 
sides, 
and 
descend 
slightlyposterad. 
There 
are 
no 
bullae. 
The 
posterior 
root 
of 
the 
zygomatic 
arch 
is 
much 
less 
expanded 
on 
the 
posterior 


19 



side 
than 
is 
that 
of 
Ursus, 
and 
the 
entire 
basicranial 
region 
is 
small. 


Supraoccipital, 
exoccipital, 
and 
basioccipital 
bones 
are 


all 
thoroughly 
fused 
to 
each 
other 
and 
to 
the 
surrounding

parietal, 
basisphenoid 
and 
squamosal. 
The 
peculiarly

shaped 
posterodorsal 
portion 
of 
the 
occipitals 
presumably

makes 
up 
the 
posterior 
surface 
of 
the 
sagittal 
and 


transverse 
occipital 
crests 
(though 
sutures 
are 
absent).

The 
transverse 
occipital 
crests 
are 
incomplete. 
Instead 
of 


continuing 
laterally 
as 
sharp 
ridges 
which 
connect 
to 
the 


dorsal 
surface 
of 
the 
mastoid 
process, 
as 
in 
Hyaenodon

and 
Canis, 
they 
converge 
ventrally, 
forming 
a 
concave,

oval 
posterior 
surface 
above 
the 
foramen 
magnum. 
This 


condition 
is 
also 
seen 
in 
Pterodon 
, 
but 
the 
lateral 


compression 
above 
the 
foramen 
magnum 
is 
relatively 


greater 
in 
Hemipsalodon 
viejaensis. 


The 
postero-lateral 
sutural 
contact 
of 
the 
occipital 
with 
the 
mastoid 
portion 
of 
the 
periotic 
is 
clearly 
visible. 
The 
paroccipital 
process 
projects 
horizontally 
laterad 
with 
a 


distal 
posterad 
hook 
and 
is 
concave 
on 
the 
posterior 
side. 
The 
occipital 
condyles 
are 
shorter 
anteroposteriorly 
than 
in 
Ursus 
, 
and 
the 
foramen 
magnum 
is 
25% 
wider 
and 
dorsoventrally 
slightly 
narrower. 
The 
hypoglossal

foramina 
open 
directly 
ventrad. 
Mediad 
and 
anterad 
from 
the 
hypoglossal 
foramina 
are 
large 
depressed 
areas 
for 
attachment 
of 
the 
rectus 
capitis 
ventralis 
and 
longuscapitis 
muscles. 
The 
occipital 
forms 
the 
posterior 
and 
medial 
walls 
of 
the 
posterior 
lacerate 
foramina. 


The 
mastoid 
portion 
of 
the 
periotic 
is 
large 
and 
forms 
a 
wedge 
between 
the 
squamosal 
and 
the 
paroccipital 
processes 
of 
the 
exoccipital. 
The 
mastoid 
processes 
face 
laterad 
and 
slightly 
ventrad 
and 
form 
the 
widest 
part 
of 
the 
basicranial 
region. 
The 
position 
of 
the 
stylomastoid

foramen 
(for 
the 
facial 
nerve) 
is 
obscured 
by 
matrix, 
but 
a 
wide 
channel 
is 
present 
for 
this 
nerve 
on 
the 
ventromedial 
surface 
of 
the 
mastoid 
process, 
as 
in 
Ursus. 
The 
petrosal 
portion 
of 
the 
periotic 
is 
visible 
only 
as 
a 
hemispherical 
projection 
in 
a 
deep, 
anteroposteriorlyelongated 
fossa 
from 
which 
matrix 
has 
not 
yet 
been 
completely 
removed. 
The 
fossa 
is 
surrounded 
by 
the 


squamosal 
and 
basisphenoid-basioccipital 
bones. 
The 
fenestrae 
rotundum 
and 
ovalis 
are 
not 
visible. 


The 
parietals 
form 
the 
roof 
of 
the 
posterior 
half 
of 
the 
skull, 
and 
appear 
to 
form 
the 
dorsal 
surface 
of 
the 
entire 
braincase. 
Anteriorly, 
the 
interparietal 
suture 
is 
unfused,
and 
the 
dorsal 
edge 
of 
the 
parietals 
forms 
the 
high

sagittal 
crest, 
which 
is 
broken 
for 
most 
of 
its 
length. 
Two 
lateral 
constrictions 
are 
visible 
in 
dorsal 
view. 
Both 
are 
overlapped 
by 
the 
dorsal 
edge 
of 
the 
squamosal. 
The 
swollen 
area 
between 
the 
constrictions 
indicates 
the 
position 
of 
the 
cerebral 
hemispheres. 
The 
peculiarposterior 
position 
of 
the 
narrowest 
part 
of 
the 
braincase 
makes 
interpretation 
of 
the 
form 
of 
the 
brain 
difficult. 
Compared 
to 
Pterodon 
(Piveteau, 
1935), 
the 
cerebrum 
appears 
to 
have 
been 
shorter 
and 
broader, 
the 
cerebellum 
relatively 
smaller, 
and 
the 
olfactory 
lobes 
larger.
However, 
there 
is 
altogether 
too 
much 
space 
in 
the 
postorbital 
portion 
of 
the 
skull 
within 
both 
parietals 
and 
frontals 
for 
even 
greatly 
expanded 
olfactory 
lobes. 
This 


suggests 
that 
the 
nasal 
cavities 
probably 
extend 
back 


under 
the 
anterior 
parietals. 
The 
lateral 
edges 
of 
the 
anterior 
part 
of 
the 
parietals 
extend 
considerably 
ventrad 
just 
anterad 
from 
the 
constriction, 
presumably 
covering 
the 
sides 
of 
the 
olfactory 
lobes. 


Although 
the 
frontals 
cover 
the 
anterodistal 
part 
of 
the 
brain 
in 
most 
mammals, 
this 
does 
not 
seem 
to 
have 
been 
true 
in 
Hemipsalodon, 
unless 
the 
frontal 
extends 
farther 
under 
the 
parietals 
than 
seems 
likely. 
The 
external 
frontal 
crest 
divides 
the 
dorsal 
part 
of 
the 
frontals 
into 
a 
concave 
anterior 
part 
which 
contacts 
the 
nasals 
and 
maxillaries, 
and 
a 
convex 
posterior 
part 
which 
makes 
up 
the 
anterior 
temporal 
fossa. 
The 
frontal 
contacts 
the 
lacrimal 
broadly 
on 
the 
side 
of 
the 
face 
and 
the 
medial 
wall 
of 
the 
orbit, 
then 
further 
posterad 
forms 
long 
sutural 
contacts 
(partlyobscure) 
with 
the 
palatine 
and 
orbitosphenoid-alisphenoidcomplex. 
The 
sutural 
contact 
with 
the 
parietals 
extends 
far 
anterad 
dorsally, 
unlike 
fissiped 
carnivores 
in 
which 
this 
contact 
is 
nearly 
vertical 
as 
seen 
from 
the 
side. 
If 
ethmoid 
foramina 
are 
present, 
they 
must 
be 
small 
and 
hidden 
by 
cracking 
of 
the 
orbital 
wall. 


The 
sphenoid 
complex, 
made 
up 
of 
the 
basisphenoid,
alisphenoid 
and 
orbitosphenoid, 
is 
difficult 
to 
partitionbecause 
of 
fusion 
of 
sutures 
between 
them 
and 
with 
the 
pterygoid 
and 
basioccipital.

There 
are 
three 
major 
foramina 
in 
the 
posterior 
orbital 
wall, 
all 
apparently 
within 
the 
sphenoid 
complex 
(Fig.
13). 
The 
anteriormost 
of 
these 
appears 
to 
be 
the 
opticforamen, 
for 
the 
optic 
nerve 
(II), 
if 
comparison 
to 
Pterodon 
is 
valid 
(Piveteau, 
1935). 
However, 
I 
could 
not 
find 
an 
ethmoid 
foramen 
in 
the 
orbital 
wall 
at 
the 
frontal-sphenoid 
suture, 
nor 
am 
I 
positive 
that 
myreconstruction 
of 
the 
orbital 
sutures 
is 
correct. 
If 
I 
have 
placed 
this 
suture 
too 
far 
forward 
(Fig. 
10), 
then 
the 
anteriormost 
large 
foramen 
might 
be 
the 
ethmoid, 
which 
is 
large 
and 
in 
this 
position 
in 
Ursus 
and 
Ailuropoda. 
I 
interpret 
the 
next 
most 
posterior 
foramen 
to 
be 
the 
orbital 
fissure 
(fente 
sphenoidale 
of 
Piveteau, 
1935) 
for 
111, 
IV,
the 
opthalmic 
branch 
of 
V, 
and 
VI. 
Further 
posterad 
(38 


mm 
behind 
the 
orbital 
fissure) 
is 
the 
foramen 
rotundum. 
which 
is 
somewhat 
larger 
than 
the 
other 
two 
foramina. 
In 
Pterodon 
(unlike 
Hemipsalodon) 
the 
foramen 
rotundum 
and 
orbital 
fissure 
are 
closely 
adjacent 
and 
equal 
in 
size,
with 
the 
orbital 
fissure 
opening 
directly 
upon 
the 
dorsal 
marginal 
ridge 
of 
the 
foramen 
rotundum. 
There 
is 
no 
alisphenoid 
canal 
in 
H. 
viejaensis. 
Instead, 
a 
broad 


groove 
leading 
from 
the 
foramen 
ovale 
around 
the 
posterolateral 
wall 
of 
the 
alisphenoid 
carried 
the 


maxillary 
artery, 
which 
seems 
to 
have 
been 
quite 
large.

The 
contact 
of 
the 
alisphenoid 
and 
basisphenoid 
with 
the 
squamosal 
is 
clearly 
visible 
on 
the 
side 
of 
the 
cranium 
and 
can 
be 
followed 
posterad 
along 
the 
dorsal

border 
of 
the 
foramen 
ovale 
to 
the 
position 
of 
theeustachian 
tube, 
where 
it 
becomes 
completely 
fused. 
The 
contact 
with 
the 
basioccipital 
and 
pterygoid 
bones 
islikewise 
invisible. 
Just 
ventrad 
from 
the 
groove 
of 
themaxillary 
artery, 
apparently 
in 
the 
pterygoid, 
is 
a 
smallforamen, 
probably 
the 
posterior 
pterygoid 
foramen. 
The 
contact 
of 
the 
pterygoid-sphenoid 
complex 
with 
the

palatine 
bones 
is 
very 
well 
marked. 


20 



Figure 
12.—Hemipsalodon 
viejaensis 
n. 
sp., 
type, 
TMM 
40263-1, 
Porvenir 
local 
fauna. 
Skull, 
ventral 
view 


Figure 
13.—Hemipsalodon 
viejaensis 
n. 
sp. 
Interpretation 
of 
bones 
and 
foramina 
of 
orbit 
and 
lateral 
braincase, 
based 
on 
details 
from 
both 
sides 
of 
the 
skull, 
TMM 
40263-1. 


Abbreviations: 
A, 
alisphenoid 
portion 
of 
sphenoid 
complex; 
AZ, 
anterior 
root 
of 
zygoma 
(cut); 
F, 
frontal; 
FM, 
fossa 
for 
inferior 
obliquemuscle; 
FO, 
foramen 
ovale; 
FP, 
posterior 
pterygoid 
foramen; 
FPP, 
posterior 
palatine 
foramen; 
FR, 
foramen 
rotundum; 
lof, 
infraorbital 
foramen; 
J, 
jugal; 
L, 
lacrimal; 
LF, 
lacrimal 
foramen; 
M, 
maxillary; 
Mp, 
mastoid 
process 
of 
periotic; 
O, 
orbitosphenoid 
portion 
of 
sphenoid 
complex; 
Oc, 
occipital; 
Op, 
optic 
foramen; 
Or, 
orbital 
fissure; 
P, 
parietal; 
Pa, 
palatine; 
Pet, 
petrosal 
portion 
of 
periotic; 
PP,
paroccipital 
process; 
PZ, 
posterior 
root 
of 
zygoma 
(cut); 
S, 
squamosal; 
Sp, 
sphenopalatine 
foramen. 


21 



Figure 
14—Hemipsalodon 
viejaensis, 
Porvenir 
local 
fauna. 
A. 
TMM 
40688-50, 
referred. 
Unworn 
M 


showing 
paracone 
and 
metacone 


asdistinctcusps.B.TMM40263-1,type.Stereophotographofrightupperdentition(P2-M3). 
2, 
22 



The 
squamosals 
form 
the 
posterior 
zygomatic 
arch, 
glenoid 
fossa, 
lateral 
wall 
of 
the 
braincase, 
and 
anterior 
and 
dorsal 
portions 
of 
the 
area 
enclosing 
the 
middle 
and 
inner 
ear. 
The 
dorsal 
border 
overlaps 
the 
edges 
of 
the 
parietals. 
The 
posterior 
edge 
is 
saddle 
shaped, 
with 
the 
contact 
with 
the 
occipitals 
and 
mastoids 
on 
the 
posteradfacing 
surface. 
A 
broad 
lateral 
flange 
covers 
the 
anterodorsal 
surface 
of 
the 
mastoid 
process 
and 
makes 
upthe 
posterior 
and 
dorsal 
sides 
of 
the 
external 
auditorymeatus. 
The 
ventral 
contacts 
with 
the 
periotic 
are 


obscure. 


The 
broad 
spoutlike 
dorsal 
depression 
for 
the 
posteriorportion 
of 
the 
temporal 
muscle 
is 
shorter 
anteroposteriorlythan 
in 
Ursus 
and 
lacks 
the 
broad 
and 
well 
defined 
posterior 
shelf 
seen 
in 
Ursus. 
The 
entire 
posterior 
root 
of 
the 
zygomatic 
arch 
is 
narrower 
and 
apparently 
not 
as 
strong 
as 
in 
equivalent 
sized 
Ursus. 
The 
glenoid 
fossa 
is 
large, 
with 
a 
spiral 
shaped 
articulating 
surface. 
The 
condylar 
process 
of 
the 
mandible 
of 
a 
large 
Ursus 
arctos 
fits 
easily 
into 
the 
fossa 
with 
room 
to 
spare. 
The 
postglenoid 
process 
is 
large, 
with 
more 
ventral 
extension 
than 
in 
Ursus. 
The 
anterolateral 
glenoid 
process 
is 
large,
producing 
a 
tubelike 
articulation 
which 
is 
more 
like 
that 
of 
a 
large 
cat 
than 
that 
of 
a 
bear. 
This 
evidently 
allows 
for 
strength 
and 
the 
precise 
occlusion 
necessary 
in 
some 
animals 
with 
specialized 
caraassial 
dentition. 
A 
similarlyspecialized 
glenoid 
fossa 
is 
found 
in 
larger 
mustelids 
(forexample 
Gulo 
and 
Taxidea. 


The 
lacrimal 
is 
a 
large 
bone, 
expanded 
well 
onto 
the 
side 
of 
the 
snout 
as 
in 
many 
primitive 
mammals. 
In 
this 
it 
is 
more 
Didelphis-like 
than 
dog-like. 
The 
facial 
extension 
of 
the 
lacrimal 
is 
more 
pronounced 
than 
in 
Hyaenodonand 
the 
lacrimal 
as 
a 
whole 
is 
larger. 
In 
Hyaenodon 
. 
both 
the 
dorsal 
and 
ventral 
sutures 
are 
on 
the 
anterior 
orbital 
border. 
In 
Hemipsalodon 
viejaensis 
the 
dorsal 
suture 
is 
on 
the 
dorsal 
border 
and 
the 
ventral 
suture 
extends 
across 
the 
anterior 
base 
of 
the 
zygomaticarch 
on 
the 
ventral 
border 
of 
the 
orbit. 
The 
posteriorborder 
of 
the 
lacrimal 
is 
obscured 
by 
fusion 
and 
fractures, 
but 
this 
bone 
seems 
to 
make 
up 
most 
of 
the 
anteromedial 
wall 
of 
the 
orbit 
and 
may 
form 
the 
dorsal 
border 
of 
the 
infraorbital 
foramen. 
A 
rough 
knob 
of 
bone 
at 
the 
anterior 
point 
of 
the 
orbit 
protects 
the 
lacrimal 
fossa, 
a 
small 
opening 
about 
11 
mm 
inside 
the 
orbital 
border. 
Near 
the 
lacrimal-palatine 
contact 
is 
a 
large 
pitwith 
raised 
posterior 
rim 
which 
is 
similar 
in 
position 
and 
character 
to 
the 
fossa 
for 
the 
inferior 
oblique 
muscle 
of 
the 
eye 
in 
Ursus 
and 
Ailuropoda, 
and 
is 
so 
identified 
here. 


The 
nasal 
bones 
are 
long 
and 
broad. 
The 
dorsal 
border 
of 
the 
external 
narial 
opening 
is 
V-shaped, 
with 
the 
point 
posterad. 
Laterally, 
the 
nasals 
have 
a 
short 
contact 
with 
the 
premaxillaries, 
a 
long 
straight 
contact 
with 
the 
dorsal 
border 
of 
the 
maxillaries, 
and 
slightly 
shorter, 
posteriorlyconverging 
sutures 
with 
the 
frontals. 
The 
forehead 
is 
concave, 
with 
the 
posterior 
part 
of 
the 
nasals 
forming 
the 
anterior 
part 
of 
the 
shallow 
concavity. 
The 
nasals 
are 
convex 
between 
the 
maxillaries. 
The 
anterior 
nasals 
have 
a 
peculiar 
V-shaped 
concavity 
partly 
rimmed 
with 
frac


tures, 
which 
appear 
to 
have 
partially 
healed 
in 
one 
area. 
There 
are 
no 
separated 
or 
offset 
fractures 
as 
would 
be 
expected 
if 
the 
damage 
had 
been 
postmortem; 
therefore 
this 
individual 
probably 
had 
been 
struck 
very 
forcefully 
on 
its 
nose, 
breaking 
and 
deforming 
the 
nasals, 
and 
the 
break 
at 
least 
partly 
healed 
before 
death. 


The 
dorsal 
edges 
of 
the 
premaxillary 
bones 
are 
short,
forming 
the 
lateral 
walls 
of 
the 
nasal 
opening 
and 
extending 
to 
a 
point 
above 
Pl 
. 
Ventrally, 
the 
premaxillaries 
form 
the 
alveoli 
for 
the 
incisors, 
the 
anteromedial 


wall 
of 
the 
canine 
alveolus, 
and 
the 
anterior 
and 
medial 
sides 
of 
the 
anterior 
palatine 
foramina. 
The 
premaxillarymaxillary 
sutures 
converge 
at 
an 
acute 
angle 
just 
posteriorto 
those 
foramina. 


The 
maxillary 
bones 
are 
large, 
forming 
most 
of 
the 
side 
of 
the 
snout 
the 
anterior 
base 
of 
the 
zygomatic 
arch,
the 
alveoli 
for 
canines 
and 
cheek 
teeth, 
and 
much 
of 
the 
palate. 
Sutural 
contacts 
with 
the 
premaxillaries, 
nasals, 
lacrimals 
and 
jugal 
bones 
are 
clear 
and 
incompletelyfused 
despite 
the 
maturity 
of 
the 
animal, 
but 
the 
ventral 
contact 
with 
the 
palatines 
is 
fused 
and 
is 
obscure 
in 
places. 
Large 
palatine 
foramina 
are 
present 
between 
the 
third 
premolars. 
A 
double 
excavation 
between 
M" 
and 
M 
3 
allowed 
space 
for 
the 
tips 
of 
the 
paraconid 
and 
protoconid 
of 
The 
lateral 
surface 
of 
the 
maxillary

’ 


2

just 
dorsal 
from 
the 
roots 
of 
M 
1is 
concave. 
The 
anterior 
opening 
of 
the 
infraorbital 
foramen 
is 
vertical 
and 
oval, 
placed 
above 
and 
between 
the 
roots 
ofP3 
The

. 


sides 
of 
the 
snout 
are 
strongly 
pinched 
inward 
just 
in 
front 
of 
the 
infraorbital 
foramen, 
then 
widen 
anteriorly 
to 
allow 
room 
for 
the 
very 
large 
canine 
roots. 


Anteriorly, 
the 
jugal 
bones 
have 
a 
strong 
ventral 
buttress 
contacting 
the 
posterior 
maxillary 
and 
an 
elongated 
dorsal 
process 
which 
wedges 
between 
the 
lacrimal 
and 
maxillary. 
The 
posterior 
ventral 
extension 
tapers 
to 
an 
end 
at 
the 
widest 
point 
of 
the 
zygoma, 
close 
to 
the 
posterior 
end 
of 
the 
arch. 
The 
ventral 
surface 
is 
scarred 
by 
the 
attachment 
of 
powerful 
masseter 
muscles. 
The 
raised 
rim 
of 
the 
scarred 
area 
extends 
forward 
onto 
the 
lateral 
surface 
of 
the 
maxillary, 
but 
most 
of 
the 


attachment 
area 
was 
on 
the 
jugal. 


The 
palatines 
make 
up 
the 
postero-ventral 
portion 
of 
the 
palate 
and 
the 
antero-lateral 
walls 
of 
the 


nasopharyngeal 
fossa. 
Surrounding 
the 
anterior 
end 
of 
the 
nasopharyngeal 
fossa 
and 
marking 
the 
end 
of 
the 
palate 
are 
two 
triangular 
raised 
areas 
which 
formed 
the 
anteromedial 
attachment 
for 
the 
pterygoid 
muscles. 
Dorsally, 
the 
palatines 
form 
a 
lateral 
and 
dorsal 
wall 
for 
the 
tube-like 
nasopharyngeal 
fossa 
and 
the 
ventromedial 
wall 
of 
the 
orbit, 
contacting 
the 
maxillary, 
lacrimal,
frontal 
and 
sphenoid 
bones. 
Three 
foramina 
are 
present 
in 
this 
bone 
on 
the 
ventral 
and 
medial 
posterior 
orbital 


walls: 
two 
small 
posterior 
palatine 
foramina, 
and 
a 
verylarge 
oval 
hole, 
presumably 
the 
sphenopalatine 
foramen 
(foramen 
palatin 
of 
Piveteau, 
1935, 
in 
Pterodon 
in 
which 
it 
is 
relatively 
much 
smaller). 
, 


The 
upper 
dentition 
is 
well 
preserved 
except 
for 
the 
canines 
and 
P 
1 
(Fig. 
14). 
All 
teeth 
show 
wear, 
but 
the 


23 



teeth 
of 
the 
left 
side 
of 
the 
type 
skull 
are 
more 
worn 
than 
those 
of 
the 
right 
side. 
12

I 
and 
I 
are 
similar, 
relatively 
small 
single 
cuspedteeth, 
r 
is 
slightly 
larger 
than 
and 
possesses 
a 
small

I1

lateral 
cingulum. 
T 
is 
much 
larger, 
semicaniniform 
tooth,
which 
wore 
laterally 
against 
the 
medial 
side 
of 
the 
lower 
canine. 
The 
incisors 
are 
more 
tightly 
packed 
together 
and 
more 
anteroposteriorly 
elongated 
than 
in 
Ursus. 
The 
central 
incisor 
cusps 
form 
an 
arc 
which 
is 
concave 


anteriorly. 


The 
crowns 
of 
the 
canines 
are 
not 
preserved. 
The 
roots 
indicate 
that 
these 
teeth 
were 
about 
the 
same 
size 
as 
in 
Ursus 
arctos 
horribilis. 
The 
canines 
in 
H. 
viejaensis 


were 
considerably 
smaller 
than 
in 
H. 
grandis 
. 
and 
mayhave 
been 
smaller 
in 
relative 
as 
well 
as 
absolute 
terms. 


The 
premolars 
increase 
in 
size 
from 
front 
to 
back. 
Only 
the 
left 
P 
1 
is 
preserved 
in 
the 
type 
skull, 
and 
its 
crown 
has 
been 
completely 
worn 
off. 
As 
in 
H. 
grandis,

P 


1 
has 
two 
roots, 
with 
the 
anterior 
root 
turned 
slightly 
to 


the 
medial 
side 
of 
the 
canine. 
Judging 
from 
the 
preserved 
roots 
and 
alveoli, 
P 
1 
in 
H. 
grandis 
was 
almost 
as 
big 
as 
Pz, 
whereas 
in 
H. 
viejaensis 
it 
is 
somewhat 
smaller 
(alveoli 
just 
over 
half 
the 
length 
of 
those 
of 
P2). 
P" 
in 
H. 
viejaensis 
is 
very 
close 
in 
size 
to 
the 
P2 
of 
the 
Cypress 
Hills 
skull 
of 
H. 
grandis, 
and 
has 
the 
anterior 
root 
directly 
posterad 
to 
the 
posterior 
root 
of 
Pl 
, 
rather 
than 
being 
set 
mediad 
as 
in 
H. 
grandis. 
The 
crown 
of 
P" 
is 
relatively 
simple, 
consisting 
of 
a 
single 
main 
cusp(paracone) 
with 
anteromedial 
and 
posterior 
ridges, 
a 
slight 
medial 
and 
a 
large 
posterior 
cingulum. 
P“ 
is 
separated 
from 
P3 
by 
a 
small 
gap. 
P3 
has 
a 
large

posteromedial 
lobe 
which 
if 
cuspate, 
could 
be 
called 
a 
protocone. 
Supporting 
this 
is 
an 
enlargement 
of 
the 
posterior 
root, 
though 
it 
is 
impossible 
to 
tell 
whether 
the 
posterior 
root 
is 
divided. 
A 
distinct 
metastyle 
is 
presentposterad 
from 
the 
much 
larger 
paracone. 
As 
in 
the 
other 
cheek 
teeth, 
the 
enamel 
surface 
is 
finely 
crenulated. 


P4 
is 
relatively 
large. 
It 
is 
about 
the 
same 
length 
as 
P 
3 
but 
much 
broader, 
taller 
and 
more 
massive. 
The 
protocone 
is 
quite 
large 
and 
is 
supported 
by 
a 
separatemedial 
root. 
The 
metastyle 
is 
a 
distinct 
cusp 
but 
much 
smaller 
than 
the 
paracone. 
The 
tooth 
is 
worn 
mainly 
on 
the 
tips 
of 
the 
cusps 
and 
on 
the 
anterior 
surface. 
This 
is 
the 
first 
specimen 
of 
Hemipsalodon 
in 
which 
P 
4 
is 
complete; 
however, 
Mellett 
(1969) 
deduced 
the 
presenceof 
a 
large 
protocone 
on 
the 
Clamo 
specimens 
from 
the 
size 
of 
the 
medial 
root. 
The 
protocone 
of 
P4 
in 
Pterodon 
dasyuroides 
is 
small, 
only 
about 
the 
relative 
size 
of 
the 
protocone 
of 
P 
in 
Hemipsalodon.

M 
1 
is 
badly 
worn 
on 
both 
right 
and 
left. 
The 
protocone 
and 
metastyle 
were 
both 
large, 
and 
the 
tooth 
was 
certainly 
camassial, 
as 
in 
H. 
grandis, 
but 
no 
details 
of 
the 
cusps 
are 
visible. 
M 
1is 
similar 
in 
size 
to 
P4 
and 
a 
weaklydevelopedlabialcingulumispresent, 
ason,
P4. 


IVT2 
can 
be 
more 
completely 
described, 
thanks 
to 
the 
presence 
of 
a 
partial 
unerupted 
M“ 
in 
TMM 
40688-50, 
as 
well 
as 
the 
worn 
teeth 
on 
the 
type. 
It 
is 
similar 
in 
shape 
to 
Ml 
, 
but 
larger. 
The 
protocone 
is 
a 
large 
crescentic 
cusp 
set 
linguad 
and 
anterad 
from 
the 
paracone 
and 


connected 
to 
it 
by 
low 
anterior 
and 
posterior 
crests. 
A 
small 
parastyle 
is 
present. 
The 
paracone 
and 
metacone 
are 
closely 
appressed 
and 
equal 
in 
height, 
but 
separated 
by 
a 
notch. 
The 
metacone-metastyle 
crest 
is 
long 
and 
deeply 
notched, 
forming 
a 
highly 
developed 
blade, 
which 
is 
honed 
to 
a 
sharp 
edge 
on 
the 
type 
specimen. 
A 
distinct 
cingulum 
is 
present 
along 
the 
entire 
labial 
margin 
of 
the 
tooth. 


M'I 
has 
its 
long 
axis 
transverse 
to 
the 
rest 
of 
the 
tooth 
row, 
and 
wears 
mostly 
on 
its 
anterior 
surface 
and 
the 
tipof 
the 
highest 
cusp. 
It 
is 
relatively 
larger 
in 
Hemipsalodon 
than 
in 
Pterodon 
, 
and 
is 
larger 
in 
its 
transverse 
dimension 
than 
M 
1 
is 
in 
the 
anteroposteriordimension. 
The 
protocone 
of 
M 
3 
projects 
much 
further 
mediad 
than 
in 
Pterodon. 
The 
paracone 
and 
metacone 
are 
not 
seen 
as 
distinct 
cusps 
on 
the 
type 
of 
H. 
viejaensis 
or 
on 
an 
isolated 
M 
3 
(FMNH:PM 
109). 
The 
parastyle 
is 
large, 
apparently 
relatively 
larger 
in 
H. 
viejaensis 
than 
in 
H. 
grandis. 
The 
protocone 
is 
crescentic, 
with 
distinct 
paraconules 
and 
metaconules 
on 
the 
type 
specimen,
though 
these 
accessory 
cusps 
are 
small 
on 
FMNH:PM 


109. 
The 
posterior 
crest 
connecting 
the 
protocone 
and 
metaconule 
is 
deeply 
notched 
and 
is 
lower 
than 
the 
anterior 
crest. 
A 
well 
defined 
posterolabial 
cingulum 
is 
present. 
The 
mandible 
(FMNH;PM 
110) 
is 
only 
slightly 
curved 
along 
its 
lower 
margin. 
In 
this 
and 
most 
other 
characters 
it 
is 
similar 
to 
the 
mandible 
of 
Pterodon 
dasyuroides,
though 
much 
larger. 
The 
symphysis 
is 
not 
fused 
in 
either 
adult 
specimen 
of 
H. 
viejaensis 
or 
in 
specimens 
of 
H. 
grandis. 
There 
are 
two 
major 
mental 
foramina, 
the 
anterior 
and 
larger 
being 
beneath 
P 
0 
and 
the 
posteriorbeneath 
a 
point 
between 
P 
and 
P 
These 
are 
essentiallythe 
positions 
of 
these 
foramina 
in 
Pterodon. 
The 
mandible 
is 
much 
shallower 
in 
H. 
viejaensis 
than 
in 
the 


type 
specimen 
of 
H. 
grandis 
(NMC 
6497), 
and 
is 
less 
strongly 
curved 
on 
the 
ventral 
margin. 
The 
coronoid 
process 
is 
not 
preserved, 
but 
judging 
from 
the 
preservedventral 
part 
of 
the 
posterior 
mandible, 
it 
was 
probablybroad 
and 
rounded, 
as 
in 
Pterodon 
and 
some 
other 
creodonts. 
The 
angular 
process 
was 
small 
and 
hooked 
slightly 
dorsad. 
The 
condyloid 
process 
is 
large, 
with 
the 
articulation 
surface 
carried 
farther 
anterodorsally,
especially 
on 
the 
lateral 
side, 
than 
in 
Ursus 
arctos. 
The 
masseteric 
fossa 
is 
broad 
and 
deep, 
though 
there 
is 
no 
sharp 
ventral 
boundary 
ridge 
as 
in 
Ursus 
arctos. 
The 
lateral 
surface 
of 
the 
mandible 
just 
below 
M 
is 
stronglyswollen. 


Alveoli 
are 
preserved 
for 
all 
the 
teeth 
except 
the 
incisors. 
An 
isolated 
canine 
(FMNH:PM 
111) 
is 
of 


similar 
size 
and 
form 
to 
the 
canines 
of 
a 
specimen 
of 
H. 
grandis 
from 
Clamo 
(OMSI 
619), 
though 
a 
little 
too 
large 
for 
the 
two 
available 
mandible 
fragments 
of 
H. 
viejaensis. 
It 
should 
be 
noted 
that 
the 
mandible 
(FMNH;PM 
110) 
of 
H. 
viejaensis 
is 
a 
little 
too 
large 
to 
fit 
the 
skull 
(TMM 
40263-1). 
Therefore, 
the 
canine 
FMNH:PM 
111 
appears 
to 
represent 
the 
largest 
known 
individual 
of 
H. 
viejaensis. 
had 
a 
single 
root, 
the 
alveolus 
for 
which 
is 
larger 
than 
either 
alveolus 
of 
Pv 


24 



The 
alveolus 
was 
inclined 
from 
the 
vertical 
because 
of 
space 
limitations 
caused 
by 
the 
large 
canine 
root. 
The 
premolars 
are 
increased 
in 
size 
posterad 
to 
P4which 
was 


, 


somewhat 
larger 
than 
M 
as 
in 
H. 
grandis. 
Several

, 


fragments 
of 
teeth 
are 
present 
with 
FMNH:PM 
110, 
but 
only 
the 
anterior 
part 
of 
could 
be 
fitted 
to 
the 
broken 
roots. 
The 
anterior 
surface 
of 
the 
paraconid 
of 
M 
0 
is 
bilobed, 
with 
a 
groove 
between 
the 
larger 
anteromedial 
and 
smaller 
anterolateral 
vertical 
ridges. 
The 
protoconidand 
paraconid 
appear 
to 
have 
been 
about 
the 
same 
height,
though 
this 
is 
very 
uncertain 
because 
of 
breakage. 
This 
is 
unlike 
the 
M., 
of 
Pterodon 
dasyuroides 
in 
which 
the 
paraconid 
is 
relatively 
small. 
Wear 
on 
the 
lateral 
face 
of 
M 
0 
suggests 
a 
shearing 
function, 
though 
the 
tips 
of 
the 
paraconid 
and 
protoconid 
had 
been 
rounded 
sufficientlyto 
greatly 
reduce 
the 
efficiency 
of 
shear 
by 
the 
time 
the 
animal 
died. 
M 
was 
much 
larger 
than 
M 
0 
but 
only 
the

? 


roots 
are 
preserved. 


Discussion.—Some 
comments 
on 
the 
validity 
of 
both 
the 
generic 
and 
specific 
names 
used 
for 
Hemipsalodonviejaensis 
are 
necessary. 
The 
differences 
between 
H. 
grandis 
and 
the 
proposed 
new 
species 
are 
largely 
matters 
of 
size 
and 
proportions. 
Because 
of 
this, 
one 
is 
compelled 
to 
consider 
the 
possibility 
of 
sexual 
dimorphism 
as 
being 
the 
cause 
of 
the 
differences. 
I 
have 
rejected 
this 
possibility 
because 
of 
the 
general 
consistency 
in 
size 
shown 
by 
specimens 
referred 
to 
H. 
grandis, 
thoughthere 
probably 
is 
overlap 
in 
some 
dimensions 
between 
small 
individuals 
of 
H. 
grandis 
and 
large 
specimens 
of 


H. 
viejaensis. 
Neither 
sample 
shows 
dimorphism. 
If 
comparisons 
with 
size 
ranges 
in 
living 
bears 
are 
significant, 
it 
is 
notable 
that 
the 
skulls 
of 
H. 
grandis 
are 
about 
the 
size 
of 
skulls 
of 
the 
Kodiak 
bear 
( 
Ursus 
middendorfi) 
and 
the 
skull 
of 
H. 
viejaensis 
is 
similar 
in 
size 
to 
that 
of 
a 
grizzly 
(Ursus 
arctos). 
Those 
who 
would 
lump 
the 
grizzly 
with 
the 
Alaskan 
coastal 
brown 
bears 
might 
consider 
all 
specimens 
of 
Hemipsalodon 
to 
belong 
to 
a 
singlespecies. 


A 
similar 
problem 
of 
possibly 
marginal 
definitive 
criteria 
exists 
for 
the 
genus 
Hemipsalodon. 
I 
consider 
the 
differences 
in 
size, 
tooth 
proportions, 
and 
cranial 
structure 
discussed 
above 
and 
by 
Mellett 
(1969) 
to 
be 
sufficient 
for 
the 
generic 
distinction 
from 
Pterodon. 
Mellett 
emphasized 
the 
rotation 
of 
the 
upper 
tooth 
rows 
in 
Hemipsalodon, 
stating 
that 
the 
uppers 
were 
rotated 
mediad 
and 
the 
lowers 
laterad. 
He 
suggested 
no 
mechanism 
for 
this 
rotation, 
and 
indeed 
I 
can 
see 
no 
evidence 
for 
rotation 
of 
most 
of 
the 
teeth 
on 
the 
available 
specimens.An 
exception 
is 
the 
upper 
molars, 
which 
have 
strongly 
curved 
roots, 
convex 
laterally 
in 
M 
1 
and 
M“,
posteriorly 
in 
M 
3. 
This 
curvature 
is 
strong 
enough 
that 
as 
the 
teeth 
wear 
down 
and 
are 
extruded 
from 
the 
alveoli 
the 
crowns 
will 
naturally 
be 
slightly 
rotated. 
There 
is 
no 
such 
mechanism 
in 
the 
lower 
molars 
since 
the 
roots 
are 
straight. 
However, 
it 
is 
possible 
that 
fairly 
constant 
pressures 
during 
chewing, 
or 
pressure 
when 
the 
jaws 
were 
closed 
and 
the 
teeth 
occluded, 
could 
cause 
a 
slightmigration 
of 
the 
lower 
molars. 
However, 
the 
upper 
molars 
close 
covering 
the 
lateral 
surfaces 
of 
the 
lower 


molars. 
Thus 
if 
occlusal 
pressures 
were 
present, 
one 
would 
expect 
a 
mediad 
rather 
than 
a 
lateral 
migration.
Moreover, 
the 
M 
’s 
of 
Pterodon 
dasyuroides 
show 
a 
distinct 
lateral 
inclination 
(protoconids 
on 
oppositestrongly 
diverging 
dorsally, 
which 
is 
similar 
to 
tne 
feature 
described 
as 
“rotation” 
described 
by 
Mellett 
in 
Hemipsalodon. 
It 
may 
be 
noted 
that 
in 
Pterodon 
dasyuroides 
the 
roots 
of 
the 
upper 
molars 
appear 
to 
be 


straight, 
so 
that 
the 
absence 
of 
molar 
rotation 
in 
Pterodon 
noted 
by 
Mellett 
is 
probably 
correct. 


Discussion 
of 
the 
ancestry 
of 
Hemipsalodon 
is 
speculative, 
as 
morphologically 
convincing 
close 
“ancestors” 
are 
unknown. 
Certainly 
both 
Hemipsalodonand 
Pterodon 
shared 
a 
common 
ancestry, 
presumably 
in 
the 
Proviverrini, 
and 
the 
genus 
Arfia 
from 
the 
earlyEocene 
of 
North 
America 
may 
be 
close 
to 
that 
common 
ancestor. 
In 
view 
of 
the 
generally 
close 
similarities 
among 
many 
European 
and 
North 
American 
mammalian 
faunas 
of 
Wasatchian-Spamacian 
age, 
it 
seems 
possiblethat 
the 
common 
ancestor 
of 
Hemipsalodon 
and 


Pterodon 
existed 
during 
the 
very 
early 
Eocene. 
This 
appears 
to 
be 
the 
view 
adopted 
by 
Mellett 
(1969)
although 
he 
did 
not 
state 
it 
in 
this 
form. 


Another 
possibility, 
essentially 
that 
suggested 
bySchlaikjer 
(1935), 
is 
that 
some 
species 
already 
at 
the 
stage 
of 
development 
of 
early 
Pterodon 
was 
the 
common 
ancestor 
of 
the 
known 
species 
of 
Pterodon 
and 


Hemipsalodon. 
The 
appearance 
of 
Hyaenodon 
in 
upperEocene 
rocks 
just 
before 
the 
appearance 
of 
Hemipsalodon 
in 
both 
North 
America 
and 
Europeindicates 
that 
a 
migration 
route 
was 
available 
at 
this 
time 
(latest 
Eocene 
to 
earliest 
Oligocene). 
A 
closer 
common 
ancestry 
would 
more 
readily 
account 
for 
the 
close 
similarity, 
seen 
especially 
in 
the 
dentition 
and 
posterior 
cranium, 
than 
would 
parallel 
evolution 
over 
a 
period 
of 


15 
million 
years 
through 
the 
Eocene. 


Ischnognathus 
savagei 
Stovall, 
1948 


Porvenir 
local 
fauna 


Figure 
15 
Type 
.—OU 
32-4-SI, 
symphyseal 
portion 
of 
mandible 
lacking 
crowns 
of 
teeth. 
Stratigraphic 
position. 
—Chambers 
Formation, 
4.6 
to 


6.1 
m 
(15 
to 
20) 
feet 
below 
the 
lower 
marker 
bed, 
Vieja
Group.
Age.—Early 
Chadronian, 
early 
Oligocene. 
Description. 
—Stovall 
(1948) 
accurately 
describes 
the 


shape, 
particularly 
the 
deep, 
narrow 
ventral 
groove 
of 
the 
symphyseal 
region, 
which 
is 
apparently 
unique 
(Fig.
15a). 
The 
form 
and 
arrangement 
of 
the 
incisors 
is 
uncertain. 
At 
least 
two 
small 
incisors 
seem 
to 
have 
been 
present, 
with 
the 
medial 
incisor 
alveolus 
on 
the 
dorsal 
side 
of 
the 
lateral 
incisor 
alveolus 
(or 
alveoli). 
On 
the 
left 
side, 
two 
matrix-filled 
alveoli 
appear 
to 
be 
present. 
On 
the 
right 
side, 
the 
matrix 
has 
been 
excavated 
from 
the 
aveoli, 
and 
the 
ventral 
(lateral) 
alveolus 
appears 
to 
have 


25 



Figure 
15.—Ischnognathus 
sagagei 
Stovall, 
type, 
OU 
32-4-51. 
Mandibular 
symphysis. 
A. 
Anterior 
view. 
B. 
Dorsal 
view, 
anterior 
to 


left. 


consisted 
of 
three 
holes. 
The 
area 
is 
sufficiently 
damaged 


to 
make 
interpretation 
of 
these 
holes 
uncertain. 


There 
are 
six 
premolar 
alveoli 
on 
the 
left 
side. 
Stovall 
(1948, 
Fig. 
2c) 
interpreted 
the 
anterior 
two 
to 
have 
been 
for 
a 
two 
rooted 
P 
. 
By 
comparison 
with 
Hyaenodon 
and 
Hemipsalodon, 
I 
would 
interpret 
the 
first 
as 
the 
lodgement 
of 
a 
single 
rooted, 
anteriorly 
inclined 
premolar. 
The 
aveoli 
of 
P^, 
P 
and 
the 
anterior 
side

O,

of 
the 
P4 
alveolus 
are 
present. 
P 
} 
would 
have 
had 
to 
be 
relatively 
small 
and 
crowded 
between 
the 
canine 
and 
P 
. 


Relationships. 
—This 
specimen 
is 
as 
enigmatic 
now 
as 
when 
it 
was 
first 
described. 
The 
suggestion 
by 
Van 
Valen 
(1966; 
quoting 
a 
personal 
communication 
from 
Patterson) 
that 
“new 
material 
shows 
it 
to 
be 
a 
hyaenodontid” 
may 
have 
been 
initiated 
by 
the 
discoveryof 
material 
assignable 
to 
Hemipsalodon 
viejaensis. 
The 
two 
taxa 
are 
almost 
certainly 
distinct, 
since 
the 
jaws 
of 
of 
Hemipsalodon 
do 
not 
show 
the 
peculiar 
depth, 
narrowness 
and 
flanged 
form 
of 
Ischnognathus. 
It 
is 


unlikely 
that 
this 
specimen 
represents 
a 
juvenileHemipsalodon 
because 
the 
two 
rami 
appear 
to 
have 
been 
firmly 
connected 
and 
probably 
partly 
fused, 
whereas 
the 
two 
mandibles 
of 
H. 
viejaensis 
(both 
probably 
from 


adult 
specimens) 
were 
unfused. 
Moreover, 
since 
the 
posterior 
root 
of 
P3 
of 
I. 
savagei 
seems 
to 
have 
been 
broken 
and 
worn 
from 
use, 
this 
may 
have 
been 
a 
late 
adult 
individual. 


Genus 
Hyaenodon 
Laizer 
and 
Parieu, 
1838 


Where 
possible 
I 
have 
followed 
Mellett’s 
1977 
revision 
of 
the 
taxonomy 
of 
North 
American 
Hyaenodon, 
includinghis 
use 
of 
subgenera. 
Specimens 
reported 
here 
have 
generally 
been 
assigned 
to 
previously 
described 
species,
although 
in 
some 
specimens 
characteristics 
which 
Mellett 
used 
for 
diagnosis 
either 
have 
not 
been 
preserved 
or 
are 


contradictory. 
Comparisons 
have 
been 
made 
with 
original 
types 
or 
casts 
of 
the 
types 
of 
the 
species 
in 
question. 


Hyaenodon 
(Neohyaenodon) 
cf. 
H. 
vetus, 
Stock, 
1933a 


Skyline 
local 
fauna 


Figs. 
16-17, 
Tables 
8-9 
Type.—CIT 
1243, 
skull 
from 
“Sespe 
upper 
Eocene,
North 
of 
Simi 
Valley, 
Ventura 
Co., 
California”. 
Referred 
material.—TMM 
41535-1, 
ramus 
with 
and 
M 
; 
41715-4, 
skull 
with 
left 
M*’~ 
lacking 
snout. 


Stratigraphic 
position. 
—Both 
41535 
and 
41715 
are 
from 
the 
Skyline 
channels, 
which 
are 
at 
the 
base 
of 
the 
Bandera 
Mesa 
Member 
approximately 
213 
m 
(700 
feet)
above 
the 
base 
of 
the 
Devil’s 
Graveyard 
Formation,
Brewster 
Co., 
Texas. 


Age.—Late 
Uintan(early 
Duchesnean 
subage).

Description. 
—The 
anterior 
part 
of 
the 
skull 
(Fig. 
17)
with 
most 
of 
the 
dentition 
is 
missing. 
What 
remains 
is 
fractured 
but 
only 
slightly 
distorted 
and 
the 
bone 
is 
well 


preserved. 
The 
frontals 
and 
posterior 
nasals 
are 
broad 
and 
flattened, 
with 
a 
deep 
depression 
immediately 
anterior 
to 
the 
point 
where 
the 
sagittal 
crest 
divides. 
The 
postorbital 
processes 
are 
large, 
and 
the 
small 
conical 
braincase 
tapersrapidly 
to 
a 
strong 
constriction, 
surmounted 
by 
a 
verylarge 
sagittal 
crest 
which 
divides 
anteriorly 
just 
in 
front 
of 
the 
postorbital 
cranial 
constriction. 
The 
largelambdoidal 
crests 
form 
ridges 
that 
make 
up 
the 
dorsal 
surfaces 
of 
the 
mastoid 
processes. 
The 
orbit 
appears 
to 
be 
relatively 
small 
but 
this 
may 
be 
partly 
due 
to 
dorsoventral 
crushing. 
The 
zygomatic 
arch 
is 
slender, 
as 
is 
usual 
in 
Hyaenodon. 
The 
posterior 
point 
of 
the 
M 
2 
extends 
posteriorly 
just 
beyond 
the 
ventral 
orbital 
shelf. 
The 
posterior 
nasal 
opening 
is 
relatively 
larger, 
or 
less 


26 



Figure 
16.—Hyaenodon 
cf. 
H. 
vetus 
Stock, 
TMM 
41535-1, 
Skyline 
channels. 
Right 
ramus 
of 
mandible 
with 
P-Mlateral 
view 


3,

3 


enclosed 
ventrally 
by 
the 
palatine 
bones, 
than 
in 
H. 


crucians. 
Little 
can 
be 
said 
about 
the 
dentition 
except

12that 
M 
and 
M 
are 
smaller 
than 
in 
H. 
montanus 
and 
similar 
in 
shape 
to 
M 
1 
2 
of 
other 
North 
American 
speciesof 
Hyaenodon. 


The 
mandible 
(TMM 
41535-1, 
Fig. 
16) 
is 
moderate 
in 
depth 
but 
is 
considerably 
thickened 
just 
above 
its 
ventral 
margin. 
There 
are 
only 
two 
alveoli 
between 
P 
and 
the 
canine 
alveolus, 
probably 
both 
for 
roots 
of 
P 
. 
There 
apparently 
was 
no 
Pj, 
P 
has 
a 
large 
protoconid, 
a 
small 
paraconid, 
and 
a 
low 
but 
broad 
posterior 
cusp 
in 
the 
position 
of 
a 
hypoconulid. 
There 
are 
no 
cingula. 
M 
is 
heavily 
worn. 
As 
is 
usual 
in 
Hyaenodon, 
it 
carried 
anteroposteriorly 
elongated 
protoconid 
and 
paraconid, 
and 
the 
carnassial 
notch 
is 
centrally 
placed. 


This 
mandible 
is 
from 
an 
older 
individual 
than 
the 
skull. 
The 
worn 
or 
missing 
parts 
of 
the 
dentition 
make 
a 
species 
identification 
even 
less 
certain 
than 
for 
the 
skull. 
Absence 
of 
a 
P 
is 
unusual 
in 
Hyaenodon, 
but 
with 
only 
a 
single 
specimen 
showing 
this 
character, 
it 
is 
probablybest 
to 
interpret 
it 
as 
an 
individual 
anomaly. 


Relationships.—Assignment 
of 
these 
two 
specimens 
to 
Hyaenodon 
cf. 
H. 
vetus 
is 
tentative. 
I 
have 
assumed 
that 
the 
specimens 
represent 
only 
one 
species, 
because 
of 
similarities 
in 
size 
(Tables 
8-9) 
and 
stratigraphic 
position.
Both 
specimens 
are 
slightly 
smaller 
than 
other 
specimensof 
H. 
vetus, 
but 
considerably 
larger 
than 
H. 
venturae 
Mellett 
(1977), 
the 
other 
known 
late 
Eocene, 
“Duchesnean,” 
species. 
There 
are 
several 
primitivecharacters. 
In 
the 
skull 
the 
most 
striking 
is 
the 
lack 
of 
closure 
of 
the 
palatine 
bones 
over 
the 
posterior 
narial 


opening. 
The 
right 
ramus 
was 
separated 
from 
the 
left 
along 
the 
median 
of 
the 
mental 
symphysis. 
Erosion 
of 
the 
bone 
prevents 
knowing 
for 
certain 
that 
an 
unfused 
symphyseal 
suture 
was 
present, 
but 
this 
seems 
likely. 
The 


posterior 
molars 
strongly 
resemble 
those 
of 
H. 
vetus, 
and 
the 
upper 
molars 
are 
almost 
identical 
in 
both 
size 
and 
form 
to 
those 
on 
a 
maxillary 
fragment 
(CM 
11770) 
from 
the 
Duchesne 
River 
Formation 
of 
Utah 
(Peterson, 
1931b, 
Fig. 
2) 
which 
Mellett 
(1977) 
has 
referred 
to 
Hyaenodoncf. 
H. 
vetus. 


Hyaenodon 
(Neohyaenodon) 
montanus 
Douglass, 
1901 


Porvenir 
local 
fauna 


Figs. 
18-19, 
Tables 
8-9 
Type. 
—CM 
764, 
skull 
and 
mandible 
“found 
northeast 
of 
Toston, 
Montana”, 
Chadronian 
age.
Referred 
material.—FMNFFPM 
82, 
ramus 
with 
P

P4 
3

and 
damaged 
P4 
and 
M 
3; 
TMM 
40203-17 
v 
; 
40203334, 
M 
2; 
40206-24, 
40206-31, 
maxilla 
fragment

!

with 
P3-M 
; 
40688-42, 
P~. 


Stratigraphic 
position.—Localities 
for 
FMNFFPM 
82 
and 
40203 
and 
40206 
are 
in 
the 
lower 
30.4 
m 
(100 
feet)
of 
the 
Chambers 
Tuff 
below 
the 
lower 
marker 
bed 
(Wilson 
1977b); 
40688 
is 
between 
0 
and 
26.8 
m 
(0 
and 
88 
feet) 
above 
the 
lower 
marker 
bed 
(Wilson 
1977b).

Age.—Early 
Chadronian. 
Description. 
—A 
partial 
ramus, 
FMNFFPM 
82 
(Fig.
19) 
is 
similar 
in 
size 
and 
proportions 
to 
the 
mandible 
of 


H. 
montanus. 
P3 
is 
much 
lower 
crowned 
than 
in 
the 
typeof 
H. 
montanus. 
? 
4 
is 
slightly 
larger 
than 
in 
the 
type. 
A 
second 
partial 
P 
TMM 
40688-42, 
is 
similar 
to 
that 
of
, 


FMNFFPM 
82 
and 
is 
also 
a 
little 
larger 
than 
that 
of 
the 
type 
of 
H. 
montanus. 
An 
M 
7, 
40206-24 
(Fig. 
18b) 
is 
relatively 
low 
and 
broad, 
with 
a 
strong 
talonid 
cusp(hypoconid). 
It 
is 
very 
close 
in 
form 
to 
that 
of 
the 
typeof 
H. 
montanus. 


27 



Figure 
17.—Hyaenodon 
cf. 
H. 
vetus 
Stock, 
TMM 
41535-4, 
Skyline 
channels. 
Stereophotographs 
of 
partial 
skull. 
A. 
dorsal 
view. 
B. 
Ventral 
view. 
C. 
Left 
lateral 
view. 


28 



40688-17 


6.5

TMM13.4 
41025-3 


le 


H.TMM16.7.6e

40206-31 
sp. 


TMM14.2 
7.5 
13.7 
10.5 
11.6 
8.8 
8.5e 
PM

crucians 
403 


H. 
FMNH10.1 
5.0
40203-34 


8.3

TMM15e 
40206-24

TMM12.8 
5.6 


82

montanus 
40203-17 


H. 
TMM13.6 
9.2 
montanus 
PM 
le

5.7 
6.8 
9a 
16a 
7.
403 
5e 
7e 
H. 
FMNH13.0 
14e 
12a 
29.7 
PM 
left 
12.7a7.6e 
10.12.6.9e 
22e

crucians 
right 
6.3 
vetus 
41315-1

H. 
FMNH10.4 
cf. 
H. 
TMM15a 
la 
10.2 
5.4 
10a 
14.9 
6.6 
21a
vetus 


41715-4 


cf. 
5e 
left 
5.4e 
6.6

8.4 
9.6
TMM40504-294 
1 


H. 
14e 
15.28e 
TMMright 
5.0 
3.2e 
5.6 
6.7 
3.0 
8.1 
3.5 
6.9 
3.5 
5.2 
6.9 
7.8 
2.8e 
12.5e 
left 
3.5e 
6.8 
2.9 
6.9 
3.3 
7.7 
4.2 
7.8 
5.9

40504-294

left 
4.7 
7.6 
3.6 
8.4 
3.4 
7.8 
3.9 
6.5e 
11.6

5.4 
3.9e 
3.4 
7.8 
4.0 
5.9 
8.1 
5.7 
8.1 
5.6
TMMright 
6.9 
3.1 
7.1 
7.2 
14.9 
46.3 
40504-236

Hyaenodon. 
species 
TMMright 
5.2 
7.4 
3.0 
8.3 
3.1 
7.7 
3.3 
5.9 
2.9 
6.5 
3.5 
7.4 
3.6 
1.0 
left 
4.8 
3.8 
6.7 
3.3 
7.3 
3.2 
8.2 
4.1 
7.7 
6.2 
6.9 
5.0 
8.7 
4.7 
14.9 
44.7 
ofnew1

Hyaenodon. 
40504-236 
raineyi7.8 
7.6 
10.3

left

ofspecies 
TMMright 
3.3e 
7.0 
3.2 
7.2 
3.6 
8.0 
3.9 
7.2 
5.9 
7.3 
4.5 
8.8 
4.7 
15.0 
45.3 
dentitions40504-125

dentitionsnewlower(P.)TMMright 
7.7 
7.6 
6.5e 
6.0e 
8.3e 
l.3e 


1

of 
H.

upper 
raineyi8.0 
8.5 
8.0 
9.5 
16.0

left 
M

40504-125

of(P.)at

TMMright 
8.2e 
APT 
AP 
APT 
APT 
APT 
APT 
APT 
APT 
depth

H.
1

923 
41

Tablecp.Pp.P, 
M 
M 
M 
Jaw

1

APTAPTAPT 
APT 
APTAPTAPTAPT 
2 
3 


-M 
-M
2 


12 
1

B.—Measurements 
H. 
9.—MeasurementsH. 
1

Table 
F 
2P 
M 
4dP 
2M 
p'-M

c 
3P4PM 


29 



Figure 
18.—Hyaenodon 
montanus, 
Porvenir 
local 
fauna. 
stereophotographs. 
A. 
TMM 
40203-34, 
M: 
, 
lateral 
view. 
B. 
TMM 
40206-24M,, 
lateral 
view. 
C-D. 
TMM 
40206-31, 
maxillary 
with 
P-M 
1, 
lateral 
and 
occlusal 
views, 
anterior 
to 
top 
of 
page. 


30 



The 
referred 
upper 
teeth 
(Fig. 
18a, 
c, 
d) 
also 
are 
similar 
in 
size 
and 
proportions 
to 
CM 
764. 
P3 
preserved 
on 
40206-31 
appears 
to 
be 
abnormal, 
with 
a 
relativelyshort 
crown 
which 
was 
probably 
damaged 
during 
its 
development. 
There 
is 
a 
posterior 
lingual 
bulge 
on 
the 
crown 
enamel 
and 
a 
large 
metacone, 
as 
in 
H. 
montanus. 
P 
is 
present 
on 
40206-31 
and 
in 
an 
isolated 
specimen,

40203-17. 
Both 
have 
a 
shelf-like 
protocone, 
a 
small 
parastyle, 
large 
paracone, 
and 
well 
developed 
but 
small 
metacone. 
M 
is 
worn 
but 
complete. 
It 
bears 
a 
small 
parastyle, 
almost 
fused 
paracone 
and 
metacone, 
and 
enlarged 
metastyle.There 
is 
no 
protocone 
cusp 
on 
the 
lingual 
root. 
M 
3, 
40203-34,is 
strongly 
convex 
on 
the 
labial 
side, 
has 
a 
distinct 
anterolabial 
ridge, 
and 
no 
protocone 
cusp.

Relationship.—Most 
if 
not 
all 
other 
occurrences 
of 
Hyaenodon 
montanus 
are 
Chadronian, 
from 
Montana 
and 
South 
Dakota. 
Specimens 
referred 
to 
this 
species 
are 
consistent 
in 
size 
and 
in 
the 
relative 
robustness 
of 
the 
teeth. 
If 
the 
identifications 
made 
here 
and 
by 
Clark 
et 
al. 
(1967) 
are 
correct, 
the 
species 
seems 
to 
span 
essentiallyall 
of 
Chadronian 
time. 
It 
may 
have 
been 
derived 
from 
Hyaenodon 
vetusas 
Mellett 
(1977) 
has 
also 
suggested.

, 


Hyaenodon 
(Neohyaenodon) 
cf. 
H. 
horridus 
Leidy, 
1853 


Little 
Egypt 
local 
fauna 


Table 
9 
Referred 
material.—TMM 
41025-3, 
mandible 
fragment 
with 
partly 
erupted 
P 


Stratigraphic 
position. 
—Same 
level 
as 
Reeves 
bonebed, 
upper 
part 
of 
the 
Chambers 
Formation, 
ViejaGroup, 
Presidio 
Co., 
Texas. 


Age. 
—Chadronian, 
probably 
middle 
Chadronian. 


Discussion.—The 
juvenile 
mandiblefragment 
includes 
a 
partial 
unerupted 
lower 
canine 
and 
?4 
as 
well 
as 
the 
which 
is 
exposed 
but 
not 
worn. 
P3 
is 
almost 
identical 
in 
form 
to 
P3 
of 
FMNFFPM 
82 
from 
the 
Porvenir 
local 
fauna, 
but 
it 
is 
about 
the 
size 
of 
H. 
cruentus 
Leidy. 


Mellett 
(1977) 
has 
synonymized 
H. 
cruentus 
with 
Hyaenodon 
horridus, 
stating 
that 
differences 
between 
these 
forms 
are 
likely 
the 
result 
of 
sexual 
dimorphism.
This 
specimen 
is 
lower 
crowned 
than 
Leidy’s 
type 
of 
H. 
cruentus 
(NMNH 
134) 
and 
the 
posterior 
cusp 
is 
slightlylarger, 
but 
it 
is 
closer 
to 
the 
H. 
cruentus 
than 
to 
other 
described 
species. 


Hyaenodon 
(Protohyaenodon) 
crucians 
Leidy, 
1853 


Porvenir 
and 
?Airstrip 
local 
faunas 


Figure 
20, 
Tables 
8-9 
Type.—ANSP 
11046-7, 
skull 
and 
mandible 
“from 
the 
Mauvaises 
Terres 
of 
Nebraska”. 


Referred 
material.—FMNH;PM 
403, 
skull 
with 
M 
1 
-M2, 
M 
2 
and 
part 
of 
P4; 
TMM 
40688-17 
dP4. 
Questionably 
referred; 
TMM 
40502-9, 
mandible 
fragment

with 
?P.2-4

Stratigraphic 
position.—The 
skull, 
FMNH 
PM 
403 
and 
40688 
are 
from 
0 
to 
26.8 
m 
(0 
to 
88 
feet) 
above 
the 
lower 
marker 
bed, 
Chambers 
Tuff; 
40502 
is 
from 
approximately 
173.7 
m 
(570 
feet) 
above 
the 
Bracks 
Rhyolite, 
Capote 
Mountain 
Formation 


Age.—Chadronian 
(early 
Oligocene).

Description. 
—The 
most 
complete 
specimen,
FMNFFPM 
403 
(Fig. 
20), 
is 
a 
badly 
weathered 
and 
broken 
juvenile 
skull. 
What 
is 
preserved 
of 
the 
teeth 
is 
nearly 
identical 
to 
the 
teeth 
of 
Hyaenodon 
crucians. 
M 
in 
particular, 
is 
indistinguishable 
in 
size 
and 
shape 
from, 
that 
of 
the 
type 
of 
H. 
crucians, 
by 
comparison 
to 
a 
cast 
of 
the 
type.

dP4 
has 
a 
large 
paracone 
as 
well 
as 
the 
metacone 
and 
metastyle. 
The 
protocone 
has 
been 
broken 
away. 
The 
paracone 
is 
a 
little 
larger 
than 
that 
shown 
on 
dP4 
of 
H.
crucians 
by 
Scott 
and 
Jepsen 
(1936, 
Fig. 
3) 
and 
the 
tooth 
is 
somewhat 
smaller, 
but 
it 
is 
otherwise 
similar. 


The 
partial 
mandible, 
TMM 
40502-9, 
includes 
3 
premolars. 
They 
are 
probably 
P 
0 
4 
because 
the 
roots 
of 


Figure 
19.—Hyaenodon 
montanus, 
Porvenir 
local 
fauna. 
FMNH 
PM 
82, 
ramus 
of 
mandible 
with 
P 
partial 
P 
and 
M 
, 
right 
lateral

3v 
43

view. 


31 



Figure20.—Hyaenodoncrucians,Porvenirlocalfauna.FMNHPM403,stereophotographsofskullwith 
M1andpartofmandible.A 
Dorsal 
view, 
B. 
Ventral 
view. 
C. 
Lateral 
view. 
D. 
Right 
M 
, 
lateral 
view. 


32 



the 
next 
posterior 
tooth 
are 
crowded 
as 
in 
M 
of 
Hyaenodon. 
This 
specimen 
is 
questionably 
placed 
in 
H. 
crucians 
solely 
on 
the 
basis 
of 
size. 
Hyaenodon 
crucians 
apparently 
was 
a 
wide 
ranging 
and 
long 
lived 
species.
Mellett 
(1977) 
reports 
it 
from 
the 
early 
Chadronian 
of 
Saskatchewan 
to 
the 
middle 
Orellan 
of 
Nebraska, 
and 
the 
present 
report 
extends 
the 
range 
to 
Texas 
in 
the 
early 
Chadronian. 


Hyaenodon 
(Protohyaenodon) 
raineyi 
new 
species 


Airstrip 
local 
fauna 


Figures 
21-22, 
Tables 
8-9 
Type.—TMM 
40504-236, 
damaged 
skull 
and 
mandible 
including 
all 
teeth 
except 
incisors 
and 
P 
. 


Referred 
material.—TMM 
40504-125, 
partial 
skull 
with 
mandible 
in 
occlusion; 
40504-294, 
partial 
skull 
with 
right 
ramus. 


Stratigraphic 
position. 
—Capote 
Mountain 
Formation 
approximately 
174 
m 
(570 
feet) 
above 
Bracks 
Rhyolite.

Age.—Middle 
Chadronian 
(early 
Oligocene).

Etymology.—Named 
for 
R. 
H. 
(Bob) 
Rainey,
Preparator, 
Vertebrate 
Paleontology 
Laboratory, 
Texas 
Memorial 
Museum. 


Diagnosis.—A 
hyaenodont 
intermediate 
in 
size 
between 
H. 
microdon 
and 
H. 
venturae. 
P 
with 
one 
root, 
posterior 
accessory 
cusp 
on 
P3 
very 
small, 
canines 
relatively 
small, 
lingual 
cingula 
on 
upper 
premolars 
weak 
or 
absent, 
skull 
narrows 
abruptly 
behind 
postorbital 
processes, 
brain 
relatively 
large.

Description. 
—Most 
of 
the 
external 
features 
of 
the 
rostrum 
can 
be 
seen 
on 
at 
least 
one 
of 
the 
three 
skulls, 
including 
well 
preserved 
dentitions 
except 
for 
incisors. 
The 
type 
skull 
(40504-236, 
Fig. 
21b,c) 
has 
part 
of 
I 
3 
and 


' 


Jri2i 
2 
t3 


..

alveoli 
for 
I 
I 
and 
I 
were 
somewhat 
smaller 
than 
I 
The 
canines 
are 
relatively 
smaller 
than 
in 
large 
species 
of 
Hyaenodon 
(H. 
horribilisfor 
example), 
slightly 
curved.

, 


and 
somewhat 
procumbent. 
P 
1 
has 
two 
roots 
and 
is 
elongated 
with 
the 
single 
cusp 
relatively 
far 
forward. 
P“ 
is 
tall 
crowned 
and 
bears 
a 
small 
posterior 
accessor/ 
cusp. 
P 
3 
is 
larger, 
broader, 
and 
shorter 
crowned 
than 
P 
as 
is 
usual 
in 
Hyaenodon. 
and 
has 
a 
weakly 
developed, 
lingual 
cingulum. 
P 
4 
has 
three 
roots 
and 
a 
small 
protocone 
that 
is 
only 
a 
shelf 
of 
enamel 
capping 
the 
medial 
root. 
There 
is 
a 
small 
metacone. 
The 
parastyle 
is 
not 
distinctly 
cuspate. 
The 
protocones 
on 
M 
1 
and 
M"2 
are 
tiny 
cusps 
which 
wear 
down 
rapidly. 
Both 
molars 
are 
highly 
effective 
camassials. 


Portions 
of 
the 
lower 
incisors 
are 
visible 
on 
the 
type 
(Fig. 
22c). 
There 
were 
three, 
increasing 
in 
size 
towards 


but 
all 
small 
and 
simple. 
The 
lower 
canine 
is 
sharply 
recurved 
but 
more 
procumbent 
than 
in 
Urocyon. 
P 
is 
missing 
on 
40504-236 
and 
damaged 
on 
both 
sides 
of 
40504-294. 
It 
has 
only 
one 
root, 
which 
is 
inclined 
anteriorly, 
and 
the 
crown 
is 
elongated 
anteroposteriorly.

is 
close 
to 
Pj 
in 
size, 
and 
like 
P 
carries 
its 
low 
single 


, 


} 


cusp 
far 
anterad. 
P 
and 
? 
are 
similar 
to 
each 
other

4 


except 
that 
? 
is 
slightly 
larger. 
Each 
bears 
a 
small

4

hypoconid. 
M 
is 
badly 
worn 
in 
adult 
specimens, 
and 
no 
details 
are 
visible 
except 
that 
a 
small 
but 
distinct 
talonid 
cusp 
is 
present. 
M 
is 
larger, 
and 
like 
Mj 
has 
a 
small

? 


talonid 
cusp. 
The 
metaconid 
is 
absent 
and 
the 
protoconidis 
the 
highest 
cusp. 
M 
is 
the 
most 
specialized 
of 
the 
camassials, 
entirely 
lacking 
a 
talonid 
and 
possessing 
a 


long 
cutting 
edge 
on 
the 
posterior 
side 
of 
the 
paraconidand 
the 
longer 
anterior 
side 
of 
the 
posterior-inclinedprotoconid. 
Like 
the 
upper 
premolars, 
the 
premolars 
are 
a 
little 
more 
widely 
spaced 
on 
40504-294 
than 
on 
40504236 
but 
all 
other 
dental 
characters 
are 
consistent. 


Sutures 
between 
skull 
bones 
are 
clearly 
visible. 
Most 
bones 
are 
similar 
to 
those 
of 
other 
species 
of 
Hyaenodon 
(Scott 
and 
Jepsen, 
1936), 
except 
as 
noted 
below. 
In 
40504-236 
(Fig. 
21a) 
the 
structure 
of 
the 
endocast 
and 
dorsal 
skull 
cavities 
has 
been 
exposed 
by 
erosion. 
The 
brain 
is 
relatively 
large 
for 
Hyaenodon 
(as 
would 
be 
expected 
in 
a 
small 
mammal) 
and 
is 
about 
the 
equivalentin 
complexity 
of 
folding 
to 
its 
contemporary, 
the 
canid 
Hesperocyon 
, 
in 
which 
only 
the 
coronolateral 
and 
suprasylvian 
sulci 
are 
present. 
The 
olfactory 
lobes 
were 
large 
and 
were 
not 
covered 
by 
the 
anterior 
cerebrum. 
The 
space 
available 
for 
the 
turbinals, 
of 
which 
portions 
are 


visible, 
was 
much 
larger 
than 
incanids 
of 
similar 
size. 


The 
shape 
of 
the 
central 
portion 
of 
the 
skull 
as 
seen 
in 
dorsal 
view 
is 
distinctive. 
In 
most 
Hyaenodon 
species 
this 
area 
is 
somewhat 
hourglass 
shaped; 
that 
is, 
the 
narrowest 
part 
of 
the 
skull 
is 
a 
short 
distance 
behind 
the 
postorbital 
processes, 
and 
the 
skull 
tapers 
posterad 
to 
the 
constriction. 
In 
H. 
raineyi 
this 
taper 
is 
very 
abrupt, 
and 
in 
40504-236 
the 
constriction 
is 
almost 
directly 
between 
the 
posterior 
projections 
of 
the 
postorbital 
processes. 
It 
is 
about 
3 
to 
4 
mm 
behind 
the 
postorbital 
processes 
in 
40504-125. 
Behind 
the 
constriction 
the 
lateral 
walls 
of 
the 
braincase 
swell 
more 
abruptly 
than 
in 
other 
Hyaenodon 
species. 
Because 
of 
damage 
to 
the 
dorsal 
surface 
of 
the 
braincase 
in 
all 
three 
skulls, 
the 
size 
and 
form 
of 
the 
sagittal 
crest 
is 
not 
discernible, 
but 
it 
was 
probably 
relatively 
small. 
There 
is 
no 
crest 
on 
the 
dorsal 
posterior 
extension 
of 
the 
frontals 
in 
40504-125, 
unlike 
specimens 
of 
other 
species 
which 
I 
have 
seen. 


Referred 
skull 
40504-125 
is 
unusual 
in 
having 
the 


tympanic 
ring 
preserved 
(Fig. 
22a,b). 
The 
tympanic 
is 
U-shaped, 
and 
broadest 
at 
the 
anteromedial 
comer. 
It 
is 
relatively 
larger 
than 
that 
pictured 
for 
H. 
“cruentus” 
(now 
H. 
horribilis) 
by 
Scott 
and 
Jepsen 
(1936, 
Plate 
8, 
Fig. 
3). 
There 
is 
no 
evidence 
for 
the 
existence 
of 
an 
ossified 
entotympanic 
such 
as 
has 
been 
found 
in 
H. 
crucians 
(van 
der 
Klaauw, 
1931), 
and 
H. 
brevirostris 
(Mellett, 
1977).

Relationships. 
—This 
species 
fits 
readily 
into 
Mellett’s 
subgenus 
Protohyaenodon 
on 
the 
basis 
of 
the 
following 
characters; 
Small 
Hyaenodon 
in 
which 
M“ 
length 
was 
about 
19% 
of 
P'-M~ 
length 
(actual 
figures 
for 
the 
two 
specimens 
in 
which 
this 
can 
be 
determined 
are 
17.4% 
and 
19.4%). 
Pediculate 
glenoids 
not 
developed. 
Anterior 
and 
posterior 
paracristid 
of 
M 
3 
subequal 
in 
length.

Mellett 
(1977, 
p. 
7) 
recognized 
the 
existence 
of 
this 
species 
from 
measurements 
supplied 
to 
him 
by 
Wilson, 


33 



Figure 
21.—Hyaenodon 
raineyi, 
n. 
sp., 
Airstrip 
local 
fauna. 
A-C. 
TMM 
40504-236, 
type 
skull, 
dorsal 
(A), 
lateral 
(B), 
and 
ventral 
(onlypalate 
and 
dentition) 
(C) 
views. 
Matrix 
fillings 
of 
skull 
cavities 
indicated 
by 
coarse 
stippling, 
matrix 
on 
external 
surfaces 
indicated 
byvertical 
lines, 
broken 
bone 
surfaces 
unshaded, 
D. 
TMM 
40504-294, 
diagram 
of 
dorsal 
rostrum 
showing 
sutures. 
Matrix 
indicated 
byvertical 
lines. 


Abbreviations: 
C, 
canine; 
F, 
frontal; 
L, 
lacrimal; 
M, 
maxillary; 
N, 
nasal; 
P, 
premaxillary. 


34 



Figure 
22.—Hyaenodon 
raineyi, 
n.sp., 
Airstrip 
local 
fauna. 
A 
& 
B. 
TMM 
40504-125, 
skull 
with 
jaws 
in 
occlusion, 
lateral 
(A) 
and 
ventral 
(B) 
views. 
C. 
TMM 
40504-236, 
type, 
mandible, 
lateral 
view. 


35 



though 
he 
was 
unable 
to 
describe 
it 
in 
his 
monograph.
Hyaenodon 
(Protohyaenodon) 
raineyi 
n. 
sp. 
clearlybelongs 
in 
the 
group 
of 
species 
which 
Mellett 
derived 
from 
H. 
venturae. 
The 
Chadronian 
diversification 
of 
Hyaenodon 
produced 
species 
with 
a 
wide 
range 
of 
sizes. 
If 
it 
is 
derived 
from 
H. 
venturae, 
Hyaenodon 
raineyiachieved 
its 
distinctiveness 
through 
a 
reduction 
in 
size,
though 
not 
to 
the 
extreme 
seen 
in 
H. 
microdon. 
In 
bodysize 
it 
was 
probably 
about 
the 
bulk 
of 
a 
fox 
(the 
skull 
is 
similar 
in 
size 
to 
Urocyon, 
the 
gray 
fox) 
or 
a 
house 
cat. 
Its 
prey 
preferences 
and 
hunting 
style 
may 
have 
been 
similar 
to 
these. 


Order 
Carnivora 
Bowdich, 
1821 


Family 
Miacidae 
Cope, 
1880 


Subfamily 
Miacinae 
Trouessart, 
1885 


Genus 
Miacis 
Cope, 
1872 


The 
stratigraphic 
range 
of 
Miacis 
is 
broad 
(early 
Eocene 
to 
early 
Oligocene) 
and 
several 
(if 
not 
most) 
species 
are 
based 
on 
poor 
or 
poorly 
described 
material. 
This 
includes 
he 
type 
species, 
M. 
parvivorus 
, 
based 
on 
an 
from 
the 


lower 
Bridger 
Formation, 
Wyoming. 
Matthew 
(1909, 
p362) 
gave 
a 
diagnosis 
for 
Miacis 
: 


“This 
genus 
includes 
the 
more 
typical 
cynoidMiacidae 
with 
camassial 
and 
tubercular 
dentition 
well 
differentiated, 
lower 
molars 
with 
basin 
heels,
slender 
or 
moderately 
deep 
jaws 
with 
the 
symphyseal 
region 
comparatively 
shallow 
and 
looselysutured, 
and 
typically 
a 
full 
eutherian 
dentition, 
but 


M 
3 
vestigial 
or 
absent 
in 
some 
species. 
The 
general 
proportions 
are 
much 
as 
in 
Cynodictis 
[=Hesperocyon], 
but 
the 
braincase 
is 
smaller. 
The 
skeleton 
is 
decidedly 
more 
primitive 
than 
in 
any 
Oligocene 
Carnivora, 
and 
agrees 
very 
closely 
with 
Vulpavus 
and 
Oodectes. 


“The 
teeth 
are 
much 
like 
those 
of 
Cynodictis 
[ 
=Hesperocyon] 
but 
the 
tubercular 
dentition 
is 
less 
completely 
differentiated 
from 
the 
sectorial 
teeth, 
much 
more 
so, 
however 
than 
in 
Oodectes 
and 
Vulpavus, 
the 
tubercular 
molars 
retaining 
to 
a 
certain 
extent 
the 
similarity 
of 
construction 
to 
the 
camassial, 
and 
the 
subordinate 
shearing 
function, 
more 
obviously 
shown 
in 
these 
genera. 
The 
second 
lower 
molar 
is 
less 
elongate 
than 
in 
Didymictis 
and 
Viveravus, 
and 
the 
third 
is 
always 
present 
thoughsmall, 
with 
either 
one 
or 
two 
roots. 
The 
third 


Figure 
23.—Miacis 
cognitus 
n. 
sp., 
type. 
TMM 
40209-200, 
Little 
Egypt 
local 
fauna. 
A. 
Skull, 
left 
lateral 
view. 
Matrix 
indicated 
byvertical 
lines, 
broken 
bone 
unshaded, 
posterior 
endocast 
coarsely 
stippled. 
B. 
Restoration 
of 
skull 
in 
left 
lateral 
view. 
Zygomatic 
archindicated 
by 
dotted 
lines. 


Abbreviations: 
AC, 
alisphenoid 
canal, 
anterior 
and 
posterior 
openings; 
EF, 
ethmoid 
foramen; 
FPP, 
posterior 
palatine 
foramen; 
lof,
infraorbital 
foramen; 
LF, 
lacrimal 
fossa; 
MLF, 
middle 
lacerate 
foramen; 
Op, 
optic 
foramen; 
Or, 
orbital 
fissure; 
PgF, 
postglenoid 
foramen;
Sp, 
sphenopalatine 
foramen. 


36 



upper 
molar 
is 
always 
small, 
sometimes 
vestigial,
and 
perhaps 
absent 
in 
the 
Uinta 
species. 
The 
parastyle 
of 
P4 
is 
small 
or 
minute, 
and 
the 
accessory 
cusps 
of 
the 
premolars 
less 
developedthan 
in 
the 
Viverravinae.” 


In 
this 
paper 
comparisons 
will 
be 
primarily 
to 
the 
type 


species, 
M. 
parvivorus 
and 
to 
Hesperocyon 
gregarius.
No 
Miacis 
basicrania 
have 
been 
described 
in 
the 
literature 
except 
for 
the 
crushed 
skull 
of 
the 
type 
of 
M. 
gracilis. 
am 
fortunate 
to 
have 
a 
referred 
specimen 
of 
M. 
parvivorus 
(USNM 
214706, 
Bridgerian 
level 
B, 
Wyoming) 
in 
which 
parts 
of 
the 
auditory 
region 
as 
well 
as 
an 
excellent 
upper 
and 
lower 
dentition 
are 
preserved. 


Figure 
24.—Miacis 
cognitus 
type, 
TMM 
40209-200, 
skull, 
Little 
Egypt 
local 
fauna. 
A. 
Ventral 
view. 
Matrix 
of 
endocast 
and 
nasal 
cavities 
indicated 
by 
coarse 
stipple, 
other 
matrix 
indicated 
by 
vertical 
lines, 
broken 
bone 
surfaces 
unshaded. 
B. 
Restoration 
of 
skull 
in 


ventral 
view. 
Right 
tympanic 
restored 
from 
bone 
contacts 
and 
comparisonwith 
Nandinia. 


Abbreviations: 
AC, 
alisphenoid 
canal, 
posterior 
opening; 
APF, 
anterior 
palatine 
foramen; 
FO, 
foramen 
ovale; 
FPP, 
posterior 
palatineforamen; 
FR, 
fenestra 
rotundum; 
HF, 
hypoglossal 
foramen; 
J, 
jugal; 
M, 
maxillary; 
MLF, 
middle 
lacerate 
foramen; 
MP, 
mastoid 
portionof 
periotic; 
Pa, 
palatine; 
Pet, 
petrosal 
portion 
of 
periotic; 
PgF, 
postglenoid 
foramen; 
PLF, 
posterior 
lacerate 
foramen; 
PM, 
premaxillary; 
pp, 
paroccipital 
process; 
PT, 
posterior 
articulation 
for 
tympanic; 
S, 
squamosal; 
SF, 
stylomastoid 
foramen; 
T, 
tympanic. 


37 



Miacis 
cognitus 
new 
species 
Type.—TMM 
40209-200, 
uncrushed 
skull 
lackingzygomatic 
arches, 
canines, 
first 
premolars, 
left 
M 
3 
and 
portions 
of 
right 
frontal, 
right 
parietal, 
and 
nasals. 
Little 
Egypt 
local 
fauna 
Stratigraphic 
position.—Above 
upper 
marker 
bed 
in 
upper 
Chambers 
Formation, 
Reeves 
Bone 
Bed 
locality,
Presidio 
County, 
Texas 
(Wilson, 
1977b, 
Fig. 
9). 
Figures 
23-28, 
Table 
10 
Age. 
—Chadronian 
(early 
Oligocene). 


Figure 
25.—Miacis 
cognitus 
type, 
TMM 
40209-200, 
skull, 
Little 
Egypt 
local 
fauna. 
A. 
Dorsal 
view. 
Matrix 
of 
endocast 
and 
nasal 


cavities 
indicated 
by 
coarse 
stipple, 
other 
matrix 
indicated 
by 
vertical 
lines, 
broken 
bone 
surfaces 
unshaded. 
B. 
Restoration 
of 
skull 
in 
dorsal 
view. 


Abbreviations: 
F, 
frontal; 
J. 
jugal; 
L. 
lacrimal; 
LC, 
lambdoidal 
crest; 
M, 
maxillary; 
N, 
nasal; 
P, 
parietal 
(boundaries 
obscure); 
PM,
premaxillary; 
POP, 
postorbital 
process; 
S, 
squamosal; 
SC, 
sagittal 
crest. 


38 



Figure26.—Miaciscognitusn. 
sp.,type,TMM40209-200,LittleEgyptlocalfauna.Stereophotosofskull.A.Dentition,ventralview 


B. 
Basicranial 
region. 
C. 
Enlargement 
of 
ear 
region, 
abbreviations 
as 
in 
Figure 
25. 
39 

Figure 
27.—Miacis 
cognitus 
n. 
sp., 
type, 
outline 
of 
skull 
showing 
position 
and 
features 
of 
endocast. 
Outline 
of 
endocast 
dashed 
where 
uncertain. 


Abbreviations: 
aeg, 
anterior 
ectosylvian 
gyrus; 
an, 
ansiform 
lobule; 
eg, 
coronal 
gyrus; 
Co, 
coronal 
sulcus; 
Fp, 
fissura 
prima; 
o, 
olfactorylobe; 
peg, 
posterior 
ectocylvian 
gyrus; 
pv, 
posterior 
cerebellar 
vermis; 
Rh, 
rhinal 
fissure; 
Ss, 
suprasylvian 
sulcus; 
Sy, 
sylvian 
sulcus 


(poorly 
developed); 
t, 
turbinals; 
tl, 
temporal 
lobe. 


Etymology.—Latin 
cognitus=known, 
referring 
to 
the 
fine 
preservation 
of 
the 
type 
specimen.

Diagnosis.— 
Distinguished 
from 
previously 
described 
species 
of 
Miacis 
by 
presence 
of 
one 
or 
more 
of 
the 
following 
characters: 
size 
of 
skull 
similar 
to 
that 
of 
M. 


'

parvivorus, 
protocone 
of 
P4 
and 
parastyles 
of 
Ml 
relatively 
small, 
M 
3 
present 
but 
very 
small, 
paracone 
and 


’

metacone 
subequal 
in 
size 
on 
M 
12, 
lingual 
cingula,
especially 
posterior 
lingual 
cingula, 
well 
developed 
on 


M 
. 
Compared 
to 
other 
described 
species: 
Miacis 
exilis 
is 
an 
odd, 
tiny 
European 
form 
whose 
M 
1 
has 
a 
largeparastyle, 
large 
central 
paracone, 
and 
greatly 
reduced 
metacone; 
M. 
jepseni 
is 
also 
smaller 
and 
much 
more 
primitive; 
all 
of 
the 
other 
Eocene 
species 
have 
relativelylarge 
paracones 
and 
parastyles 
and 
small 
metacones 
on 
M 
. 
M 
or 
both; 
M. 
gracilis 
had 
a 
M 
1 
with 
paracone
, 


and 
metacone 
of 
about 
equal 
size 
but 
an 
M' 
which 
is 
much 
reduced 
and 
almost 
lacking 
a 
metacone; 
M. 
hargeri, 
M. 
latidens, 
M. 
medius, 
M. 
uintensis, 
M. 
gracilis, 
M. 
australis 
n. 
sp. 
and 
M.? 
macintyri 
are 
all 
larger; 
M. 
washakius, 
M. 
gracilis, 
M. 
jepseni, 
M. 
sylvestris, 
M. 
latidens, 
M. 
hargeri 
all 
either 
have 
smaller 


1-2 
2

hypocones 
on 
M 
or 
lack 
a 
hypocone 
on 
M; 
M. 
hargeri 
and 
M. 
medius 
have 
relatively 
large 
M 
3; 
and 
M. 
sylvestris 
has 
a 
relatively 
larger 
P 
4. 


Description.—The 
skull 
is 
from 
an 
adult 
individual 
in 
which 
most 
of 
the 
sutures 
of 
the 
cranium 
are 
at 
least 
partly 
fused 
and 
the 
molars 
considerably 
worn. 
It 
is 
uncrushed 
and 
undistorted 
except 
minimally 
around 
a 
break 
at 
the 
postorbital 
constriction. 
Only 
the 
nasals 
are 
entirely 
missing; 
this 
circumstance 
allows 
restoration 
of 
other 
missing 
fragments. 
Natural 
removal 
of 
bone 
on 
the 
right 
side 
of 
the 
braincase 
and 
a 
portion 
of 
the 
rear 
of 
the 
skull 
has 
exposed 
a 
partial 
endocast, 
allowing 
descriptionof 
the 
brain 
cavity.

As 
seen 
in 
dorsal 
view 
(Fig. 
25), 
the 
skull 
is 
verysimilar 
in 
size 
(Table 
10) 
and 
proportions 
to 
that 
of 


Genetta, 
and 
is 
within 
the 
size 
range 
of 
Hesperocyon. 
It 
is 
almost 
exactly 
the 
same 
size 
as 
a 
skull 
(USNM214706) 
of 
M. 
parvivorus 
and 
is 
very 
similar 
in 
proportions 
to 
that 
skull. 
Miacis 
gracilis 
is 
slightly 
largerbut 
also 
is 
similar 
in 
proportions. 
The 
snout 
is 
small,
slightly 
narrower 
than 
in 
Hesperocyon 
gregarius. 
The 
nasals 
were 
narrow 
and 
extended 
back 
to 
a 
point 
oppositethe 
anterior 
rim 
of 
the 
orbits, 
very 
much 
as 
in 
Genetta. 
The 
dorsal 
sutures 
between 
the 
frontal 
and 
parietals 
are 
fused 
and 
obscure. 
The 
postorbital 
process 
is 
small 
and 
blunt, 
smaller 
than 
in 
M. 
gracilis 
or 
in 
specimens 
of 
Hesperocyon 
of 
similar 
size. 
Extending 
posteromediallyfrom 
the 
postorbital 
process 
is 
a 
ridge 
which 
defines 
the 
dorsal 
border 
of 
the 
temporal 
fossa. 
The 
ridges 
of 
both 
sides 
form 
a 
lyrate 
prominence, 
but 
join 
to 
form 
a 
short 
sagittal 
crest 
immediately 
posterior 
to 
the 
maximum 
swelling 
of 
the 
braincase. 
The 
lambdoidal 
crests 
are 
prominent, 
again 
similar 
in 
proportions 
to 
those 
of 
Hesperocyon, 
Genetta, 
and 
M. 
gracilis. 


The 
braincase 
is 
slightly 
smaller 
than 
in 
either 
Hesperocyon 
or 
Genetta. 
differing 
from 
Genetta 
mainlyin 
the 
lesser 
anterolateral 
and 
posterodorsal 
swellings 
of 
the 
bone 
covering 
the 
cerebrum. 


In 
lateral 
view 
(Fig. 
23) 
the 
dorsal 
facial 
outline 
is 
relatively 
straight. 
The 
maxillary 
forms 
nearly 
all 
of 
the 
side 
wall 
of 
the 
snout. 
The 
dorsal 
process 
of 
the 
relatively 
small 
premaxillary 
is 
much 
smaller 
than 
in 
such 
canids 
as 
Canis 
latrans 
and 
Vulpes 
macrotismaking 
contact 
with 
the 
nasals 
for 
less 
than 
a 
quarter, 
of 
their 
length. 
The 
premaxillary 
in 
other 
species 
of 
Miacis 
(M.
gracilis, 
M. 
parvivorus, 
M. 
hargeri) 
also 
appears 
to 
be 
small. 


The 
orbit 
is 
relatively 
large, 
as 
in 
many 
small 
carnivores. 
The 
lacrimal 
is 
larger 
than 
in 
Hesperocyon 
or 
Genetta. 
It 
forms 
a 
sharp 
ridge 
on 
the 
anterior 
edge 
of 
the 
orbit, 
behind 
which 
is 
a 
pronounced 
pit, 
presumably 


40 



Table 
10.—Measurements 
of 
specimens 
of 
Miacis. 


M. 
cognitus 
typeTMM 
40209-200 
right 
left 
M. 
australis 
FMNH 
PM 
423 
(type)
right 
left 
TMM 
41211-5 
left 
M. 
parvivorusUSNM 
214706 
right 
left 
a 
AP 
3.5e 
3.5e 
4.4e 
4.1 
p 
1 
p2 
p3 
p4 
M 
1 
T 
AP 
T 
AP 
T 
AP 
T 
AP 
T 
AP 
2.3e 
2.6a 
2.3a 
4.2 
1.9 
4.7 
2.3e 
6.7 
4.4 
5.1 
2.3e 
2.7a 
1.9a 
4.0 
1.9 
4.6 
2.7 
6.9 
4.2 
5.4 
7.9e 
4.9 
5.8e 
2.5e 
2.6a 
1.9a 
4.8a 
5.9 
7.5e 
5.1 
6.7 
3.4 
2.5 
1.4 
3.5 
1.6 
5.0 
2.6 
7.3 
4.8 
5.9 
4.7 
2.5 
7.6 
4.8 
5.8 
M2 
T 
AP 
7.8 
3.8 
7.5 
3.8 
9.0e 
4.5e 
9.7e 
4.8e 
8.7 
3.7 
8.6 
3.7 
M3 
T 
AP 
6.0 
1.6 
6.0 
6.8e 
6.7e 
6.2 
6.2 
T 
Post. 
C-M2 
2.3 
32.4 
32.5 
37.3 
29.3 
/C 
AP 
T 
4.0 
3.3 
P,1 
AP 
T 
2.0 
1.3 
P,2 
AP 
T 
3.2e 
1.3 
P,5 
AP 
T 
6.1 
2.5 
5.9 
4.2 
2.2 
p4 
AP 
T 
6.5 
2.7 
6.7a 
5.6 
2.9 
m,1 
AP 
T 
7.5 
4.3 
7.8 
4.1 
7.1 
4.3 
6.9 
4.1 
M2 
AP 
T 
5.3 
3.5 
4.9a 
5.6 
3.5 
4.5 
3.5 
4.6 
3.7 
M,3 
AP 
T 
2.5a 
2.3a 
1.9a 
2,2 
2.4 
Post. 
C-M,
Jaw 
depth 
at 
M 
Length 
of 
skull 
Width 
at 
mastoids 
84.4 
29.2 
41.8e 
11.0 
41.5a 
11.1 
10.9 
32.1a 
8.9 
9.2 
Width 
ofcondyles 
21.0 


41 



Figure 
28.—Miacis 
cognitus 
n.sp., 
type, 
TMM 
40206-200, 
Little 
Egypt 
local 
fauna. 
Posterior 
view 
showing 
cerebellar 
endocast. 


for 
a 
large 
lacrimal 
gland. 
A 
similar 
structure 
is 
seen 
in 
Bassariscus, 
but 
it 
is 
not 
present 
in 
Genetta 
and 
is 
much 
smaller 
in 
Hesperocyon. 
The 
lacrimal 
forms 
the 
dorsal 
border 
of 
the 
posterior 
opening 
of 
the 
infraorbital 
canal. 
The 
anterolateral 
border 
of 
the 
orbit 
was 
formed 
by 
the 
jugal, 
of 
which 
only 
the 
anteriormost 
part 
is 
preserved.
There 
was 
a 
distinct 
maxillary 
process 
or 
ventral 
projection 
of 
the 
jugal 
overlapping 
the 
lateral 
surface 
of 
the 
maxillary, 
which 
is 
indicated 
by 
an 
articular 
surface 
in 
M. 
cognitus. 
This 
is 
present 
in 
Miacis 
parvivorus, 
all 
canids, 
and 
at 
least 
some 
procyonids 
(Procyon, 
Nasua), 
but 
not 
in 
viverrids, 
felids 
or 
mustelids. 
The 
dorsomedial 


wall 
of 
the 
orbit 
is 
formed 
by 
the 
frontal. 
Two 
small 
foramina 
are 
present 
immediately 
posterior 
to 
the 
postorbital 
process. 
The 
ethmoid 
foramen 
is 
a 
single 
oval 
opening 
at 
the 
contact 
of 
the 
frontal 
with 
the 
orbitosphenoid. 
The 
boundary 
of 
the 
orbitosphenoid 
is 
difficult 
to 
locate 
because 
of 
breakage 
of 
this 
delicate 
bone. 
However, 
it 
appears 
to 
have 
formed 
the 
opticforamen 
and 
the 
medial 
wall 
of 
the 
orbital 
fissure. 
An 
alisphenoid 
canal 
is 
present, 
apparently 
opening 
as 
in 
Hesperocyon 
within 
the 
foramen 
rotundum 
immediatelylateral 
to 
the 
orbital 
fissure. 
The 
close 
association 
of 
the 
orbital 
fissure 
and 
anterior 
opening 
of 
the 
alisphenoidcanal 
is 
similar 
to 
the 
structure 
found 
in 
Nandinia 
as 
well 
as 
in 
Hesperocyon. 
Boundaries 
between 
the 
alisphenoid 
and 
other 
bones 
are 
poorly 
marked 
on 
this 
specimen.

The 
ventromedial 
and 
ventral 
walls 
of 
the 
orbit 
are 
formed 
by 
the 
palatine 
and 
maxillary, 
respectively. 
The 
lateral 
wall 
of 
the 
palatine 
contains 
the 
sphenopalatineand 
posterior 
palatine 
foramina. 
The 
maxillary 
forms 
a 
flat 
floor 
for 
the 
orbit 
and 
encloses 
most 
of 
the 
infraorbital 
foramen, 
which 
is 
nearly 
circular 
at 
its 
anterior 
opening 
above 
P3. 


The 
palate 
is 
composed 
of 
premaxillary, 
maxillary,
and 
palatine 
bones. 
The 
premaxillaries 
contain 
the 
incisor 
alveoli, 
and 
form 
the 
anteromedial 
border 
of 
the 
canine 
alveoli 
and 
the 
anterior 
and 
medial 
borders 
of 
the 
anterior 
palatine 
foramina. 
The 
palatine 
foramina 
are 
slightlysmaller 
than 
in 
Genetta 
and 
are 
concave 
laterally. 
The 
anterior 
border 
of 
the 
palatines 
lies 
between 
the 
protocones 
of 
P4 
The 
contact 
between 
palatines 
and

. 


maxillaries 
is 
irregular, 
and 
the 
posterior 
palatine

foramina 
and 
several 
minor 
foramina 
open 
along 
it. 


The 
basicranial 
region 
is 
well 
preserved 
(Fig. 
26b, 
c). 
The 
basioccipital 
and 
basisphenoid 
are 
broad 
and 
completely 
fused. 
There 
is 
no 
ossified 
bulla 
preserved,
and 
there 
are 
no 
facets 
on 
the 
bones 
on 
the 
medial 
and 
posterior 
sides 
of 
the 
promontorium 
which 
would 
indicate 
the 
presence 
of 
an 
ossified 
caudal 
entotympanic. 
A 
small 
buttonlike 
process 
on 
the 
mastoid 
part 
of 
the 
periotic, 
justanterior 
to 
the 
stylomastoid 
foramen, 
appears 
to 
be 
the 


articulation 
for 
the 
posterior 
limb 
of 
the 
tympanic, 
to 
which 
it 
apparently 
was 
attached 
by 
a 
small 
ligament. 
A 
similar 
articular 
process 
is 
seen 
in 
Miacis 
parvivorus, 
in 
Recent 
Nandinia 
(in 
which 
the 
tympanic 
does 
not 
fuse 
to 
other 
bones), 
and 
in 
juvenile 
carnivores 
before 
the 
tympanic 
fuses 
to 
other 
bones 
(Hunt, 
1974a, 
Fig. 
6 
and 
25). 
The 
articulation 
for 
the 
anterior 
limb 
of 
the 
tympanic, 
against 
the 
posterior 
side 
of 
the 
alisphenoidand 
squamosal, 
is 
less 
well 
defined, 
though 
it 
is 
broader 
and 
better 
defined 
than 
in 
M. 
parvivorus. 
It 
appears 
that 
this 
articular 
area 
was 
relatively 
smaller 
in 
M. 
cognitusthan 
it 
is 
in 
Nandinia, 
and 
that 
the 
tympanic 
may 
have 
been 
less 
expanded 
than 
in 
Nandinia 
though 
more 
expanded 
than 
in 
M. 
parvivorus. 
I 
cannot 
distinguisharticular 
surfaces 
for 
the 
rostral 
entotympanic, 
and 
this 
element, 
like 
the 
caudal 
entotympanic, 
may 
have 
been 
unossified. 


The 
foramina 
of 
the 
skull 
are 
shown 
in 
Figure 
23a 
and 
24a. 
The 
hypoglossal 
foramen 
is 
round 
and 
separatedfrom 
the 
posterior 
lacerate 
foramen. 
The 
posteriorlacerate 
foramen 
is 
larger 
than 
the 
hypoglossal 
and 
is 
continuous 
in 
part 
with 
the 
inferior 
petrosal 
sinus. 
The 
inferior 
petrosal 
sinus 
is 
narrow 
and 
is 
roofed 
dorsally 
byportions 
of 
both 
the 
petrosal 
and 
basioccipital. 
The 
petrosal 
is 
similar 
to 
that 
of 
Miacis 
parvivorus. 
In 
both,
the 
petrosals 
are 
rounded 
ventrally, 
with 
an 
anterior 
projection 
between 
the 
fossa 
for 
the 
tensor 
tympanimuscle 
and 
the 
middle 
lacerate 
foramen. 
The 
fenestra 
rotundum 
faces 
more 
laterally 
than 
in 
Hesperocyon. 
In 


M. 
cognitus 
a 
small 
shelflike 
projection 
is 
present 
on 
the 
petrosal 
posterior 
to 
the 
fenestra 
rotundum. 
This 
shelf 
is 
absent 
in 
M. 
parvivorus 
(in 
which 
the, 
fenestra 
rotundum 
is 
therefore 
closer 
to 
the 
posterior 
margin 
of 
the 
petrosal),
but 
is 
present 
and 
broader 
in 
Hesperocyon 
gregarius. 
In 
M. 
parvivorus 
a 
small 
groove 
on 
the 
ventral 
surface 
of 
the 
promontorium 
may 
indicate 
the 
presence 
of 
a 
promontory 
branch 
of 
the 
carotid 
artery. 
No 
such 
grooveis 
visible 
in 
M. 
cognitus, 
but 
a 
small 
foramen 
visible 
on 
the 
right 
side 
lateral 
to 
the 
middle 
lacerate 
foramen,
situated 
on 
the 
contact 
between 
the 
alisphenoid 
and 
petrosal, 
may 
be 
a 
promontory 
foramen 
and 
may 
indicate 
42 



the 
presence 
of 
a 
promontory 
artery. 
In 
M. 
cognitus 
the 
ventral 
edge 
of 
the 
petrosal 
contacts 
the 
basioccipitalalong 
3 
mm 
of 
its 
edge. 
The 
extent 
of 
the 
dorsal 
contact 
is 
difficult 
to 
determine, 
but 
is 
probably 
substantially 
less 
than 
in 
Hesperocyon. 
The 
stylomastoid 
foramen 
is 
situated 
posterolateral 
to 
the 
fenestra 
rotundum. 
The 


middle 
lacerate 
foramen 
is 
a 
triangular 
opening 
on 
the 
anteromedial 
side 
of 
the 
petrosal. 
On 
the 
medial 
side 
of 
the 
middle 
lacerate 
foramen 
in 
the 
basisphenoid 
is 
a 
looped 
groove 
by 
which 
the 
medial 
branch 
of 
the 
internal 


carotid 
enters 
the 
cranial 
cavity 
(Hunt 
1974a, 
p. 
39). 
The 


postglenoid 
foramen, 
foramen 
ovale, 
and 
the 
posterioropening 
of 
the 
alisphenoid 
canal 
are 
present 
and 
virtuallyidentical 
in 
M. 
cognitus 
and 
M. 
parvivorus.

The 
paroccipital 
and 
mastoid 
processes 
are 
small, 
and 
neither 
shows 
any 
sign 
of 
being 
involved 
in 
the 
formation 
of, 
or 
of 
being 
in 
contact 
with, 
a 
bulla. 


A 
well 
preserved 
endocast 
(Figs. 
27,28) 
is 
exposed 
on 
the 
right 
side 
of 
the 
cranium 
and 
in 
portions 
of 
the 
occipital 
region. 
This 
is 
the 
first 
known 
specimenreferrable 
to 
Miacis 
in 
which 
the 
endocast 
is 
preserved.
The 
brain 
was 
notably 
primitive. 
It 
shows 
(as 
would 
be 
expected 
in 
a 
miacid) 
relatively 
little 
expansion 
of 
the 


neocortex 
or 
complication 
of 
the 
lobes 
and 
fissures. 
The 
endocast 
resembles 
those 
of 
the 
most 
primitive 
canid,
Hesperocyon 
(early 
Oligocene; 
Radinsky, 
1969), 
the 
late 
Eocene 
European 
miacid 
Procynodictis 
(-“Viverravus”)
angustidens 
(illustrated 
in 
Piveteau, 
1962), 
and 
the 
late 
Oligocene 
viverrid 
Herpestides 
antiquus 
(illustrated 
in 
Radinsky, 
1971) 
in 
the 
presence 
of 
only 
two 
well 
developed 
neocortical 
sulci, 
the 
coronolateral 
and 


suprasylvian. 
The 
rhinal 
fissure, 
separating 
neocortex 
from 
paleocortex, 
is 
high 
and 
is 
indistinct 
except 
in 
the 
anteriormost 
portion 
and 
just 
anterad 
from 
the 
widest 
partof 
the 
endocast. 
The 
fissura 
prima. 
an 
importantlandmark, 
can 
only 
be 
located 
approximately, 
and 
the 
posterior 
cerebrum 
is 
covered. 
From 
the 
form 
of 
the 
parietals 
it 
appears 
that 
the 
posterior 
cerebrum 
overlapsthe 
cerebellum 
to 
a 
slightly 
greater 
extent 
than 
in 
P. 
angustidens, 
but 
less 
than 
in 
Hesperocyon 
gregarius.
The 
shape 
of 
the 
olfactory 
bulbs 
can 
be 
approximatedfrom 
swellings 
on 
the 
medial 
orbital 
walls. 
These 
appearto 
be 
smaller 
and 
more 
overlapped 
by 
the 
anterior 
cerebrum 
than 
in 
P. 
angustidens. 
Compared 
to 
P. 
angustidens, 
the 
brain 
of 
M. 
cognitus 
is 
lower 
and 
flatter, 
particularly 
in 
the 
posterior 
half 
of 
the 
cerebrum,
has 
a 
narrower 
gyrus 
between 
the 
posterior 
rhinal 
fissure 
and 
the 
posterior 
suprasylvian 
sulcus, 
and 
a 
more 


posteriorly 
located 
incipient 
sylvian 
notch. 


Only 
the 
upper 
dentition 
is 
known 
(Fig. 
26a). 
The 
incisors 
are 
small, 
simple, 
and 
closely 
packed, 
with 
I 
3 
slightly 
larger 
than 
the 
others. 
The 
canines 
and 
the 
first 


premolars 
are 
missing; 
these 
teeth 
were 
probably 
similar 
to 
those 
of 
M. 
parvivorus. 
The 
anterior 
premolars 
are 
widely 
spaced, 
more 
so 
than 
in 
M. 
parvivorus, 
in 
which 
P" 
and 
P3 
are 
not 
separated 
by 
a 
diastema. 
P2 
is 
a 
simplepremolar 
lacking 
accessory 
cusps. 
P3 
also 
lacks 
accessory 
cusps 
but 
is 
larger 
and 
has 
a 
distinct 
lingual 
cingulum. 
P4 
differs 
from 
that 
of 
M. 
parvivorus 
in 
the 
much 
smaller 
protocone 
and 
slightly 
longer 
metastyle. 
The 
spacebetween 
the 
lingual 
shearing 
surface 
of 
P4 
and 
the 
anterior 
side 
of 
M 
1 
is 
similar 
to 
that 
of 
M. 
parvivorus 


Figure 
29.—Miacis 
australis 
n.sp., 
type. 
FMNH 
PM 
423, 
Porvenir 
local 
fauna. 
A. 
Maxillary 
fragment 
with 
P4-M" 
(P4 
reversed 
from 
right 
side), 
occlusal 
view. 
B, 
Ramus 
of 
mandible 
with 
P 
-M 
occlusal 
and 
outside 
views.

, 


43 



but 
is 
considerably 
smaller 
than 
in 
Hesperocyon. 
The 
M 
1 
of 
M. 
cognitus 
is 
much 
more 
symmetrical 
than 
that 
of 


M. 
parvivorus 
or 
Hesperocyon 
because 
the 
parastyle 
is 
smaller 
than 
in 
either 
of 
these 
taxa. 
The 
metacone 
and 
paracone 
are 
subequal 
in 
size, 
and 
the 
hypocone, 
or 
widest 
part 
of 
the 
lingual 
cingulum, 
is 
situated 
in 
a 
more 
anterior 
and 
lingual 
position. 
The 
hypocone 
is 
larger 
than 
in 
Uintacyon 
and 
Procynodictis 
and 
resembles 
that 
on 
the 
M 
1 
of 
Daphoenus. 
The 
lingual 
cingulum 
does 
not 
extend 
as 
far 
around 
the 
anterior 
side 
of 
the 
tooth 
as 
in 


M. 
parvivorus 
but 
it 
does 
extend 
farther 
than 
in 
some 
specimens 
of 
Hesperocyon. 
The 
distinguishing 
features 
of 
M 
are 
repeated 
in 
M", 
Although 
M“ 
is 
smaller 
than 
M 
1it 
is 
not 
as 
small 
as 
in 
Hesperocyon. 
The 
posteriorlingual 
cingulum 
is 
smaller 
in 
M. 
cognitus 
than 
in 
M. 
parvivorus. 
M 
3 
is 
a 
small 
oval 
tooth 
with 
probably 
two 
roots 
in 
M. 
cognitus 
, 
and 
judging 
from 
the 
alveoli 
was 
smaller 
than 
in 
M. 
parvivorus. 
Relationships. 
—This 
skull 
shows 
a 
great 
manyprimitive 
features. 
It 
is 
perhaps 
more 
similar 
to 
Miacis 
parvivorus 
of 
the 
Bridget 
Formation 
(middle 
Eocene)
than 
to 
anything 
else 
which 
is 
adequately 
known, 
but 
differs 
from 
that 
species 
particularly 
in 
characters 
of 
the 
dentition. 
The 
dentition 
is 
complete 
and 
lacks 
such 
reduction 
as 
is 
seen 
in 
viverrids, 
canids, 
and 
other 
extant 
families. 
The 
teeth 
are 
perhaps 
more 
amphicyonid-like,
particularly 
resembling 
Daphoenus, 
than 
like 
those 
of 
other 
families. 
The 
skull 
lacks 
the 
expansion 
of 
the 
inferior 
petrosal 
sinus 
which 
is 
characteristic 
of 
the 
Amphicyonidae 
(Hunt, 
1974b). 
It 
also 
lacks 
the 
posteriorexpansion 
of 
the 
caudal 
entotympanic 
chamber 
which 
is 
seen 
in 
the 
Viverridae, 
including 
Nandinia 
in 
which 
the 
caudal 
entotympanic 
is 
not 
ossified. 
The 
skull 
proportions

of 
Miacis 
cognitus 
are 
like 
those 
of 
Genetta 
but 
the 
lack 
of 
specializations 
in 
both 
dentition 
and 
bullae 
precludesclassification 
with 
the 
Viverridae, 
despite 
the 
precedentof 
combining 
Miacidae 
with 
Viverridae 
which 
was 
set 
byGregory 
and 
Heilman 
(1939). 
The 
basicranial 
characters 
of 
M. 
cognitus 
are 
more 
similar 
to 
those 
of 
the 
Canidae 
than 
to 
other 
families; 
in 
fact 
the 
bulla 
of 
the 
adult 
M. 
cognitus 
probably 
closely 
resembled 
that 
of 
a 
juvenilecanid 
(Hunt, 
1974a, 
Fig. 
24-26, 
PI. 
7). 
However, 
Miacis 
cognitus 
appears 
too 
late 
to 
have 
been 
a 
generalizedcanid 
ancestor. 
Miacis 
gracilis, 
which 
is 
more 
like 
Hesperocyon 
(the 
first 
undoubted 
canid) 
than 
is 
M. 
cognitus, 
appears 
in 
the 
Uintan 
along 
with 
Procynodictis,
which 
has 
been 
placed 
in 
the 
Canidae 
by 
previousworkers. 
Miacis 
cognitus 
seems 
to 
have 
been 
a 
late 
survivor 
of 
a 
primitive 
line, 
and 
it 
was 
probably 
not 
ancestral 
to 
any 
other 
known 
form. 


Miacis 
australis 
new 
species 


Porvenir 
local 
fauna 


Figure 
29, 
Table 
10 


Type 
.—FMNH:PM 
423, 
incomplete 
anterior 
skull 
including 
partial 
left 
P 
4-M", 
right 
P4 
mandibular 
rami 
with 
partial 
left 
canine 
and 
right 
P,
3-M2 


_ 


Referred 
material.—TMM 
41211-5, 
mandible 
fragment 
with 
M 
. 


Stratigraphic 
position. 
—The 
type 
is 
from 
the 
Chambers 
Formation, 
0 
to 
26.8 
m 
(0 
to 
88 
feet) 
above 
lower 
marker 
bed 
and 
41211 
is 
near 
Adobe 
Springs(Wilson, 
1977b, 
Fig. 
8), 
7.6 
to 
12.8 
m 
(25 
to 
42 
feet)
above 
the 
Buckshot 
Ignimbrite, 
Presidio 
County, 
Texas. 


Age.—Early 
Chadronian 
(early 
Oligocene). 


Etymology. 
—Refers 
to 
the 
southern 
position 
among

U. 
S. 
Chadronian 
fossil 
localities. 
Diagnosis. 
—Protocone 
of 
P4 
large, 
parastyle 
of 
M 
relatively 
small, 
paracone 
and 
metacone 
subequal 
in 
size 
on 
both 
M 
1 
and 
probably 
M 
2, 
lingual 
cingulum 
well 
developed 
on 
both 
M 
1 
and 
M 
2, 
protoconid-paraconidshearing 
blade 
on 
directed 
more 
anteroposteriorlythan 
in 
M. 
parvivorus 
but 
less 
so 
than 
in 
Hesperocyon.

Compared 
to 
other 
species 
of 
Miacis, 
M. 
australis 
is 


larger 
than 
M. 
cognitus 
n. 
sp. 
and 
M. 
parvivorus, 
and 
much 
larger 
than 
M. 
exilis 
and 
M. 
jepseni. 
It 
differs 
fromM. 
gracilis 
in 
having 
a 
relatively 
smaller 
P4, 
a 
largehypocone 
on 
Ml 
a 
relatively 
larger 
M 
0 
and 
a 
more

, 


anteroposteriorly 
oriented 
protoconid-paraconid 
blade 
on 


M 
. 
It 
differs 
from 
all 
other 
species 
except 
M. 
cognitusin 
the 
relatively 
small 
size 
of 
the 
M 
1 
parastyle, 
but 
differs 
from 
that 
species 
in 
the 
relatively 
large 
size 
of 
the 
P 
protocone. 
The 
form 
of 
the 
trigonid 
of 
M 
also 
separates 
M. 
australis 
from 
all 
other 
Miacis 
species 
in 
which 
this 
structure 
is 
known. 
Description. 
—The 
type 
specimen 
consists 
of 
fragments 
of 
the 
rostrum 
of 
a 
skull 
in 
a 
block 
and 
two 
associated 
partial 
mandibular 
rami. 
The 
skull 
has 
separated 
along 
the 
midline 
and 
the 
fragments 
have 
been 
rotated 
180° 
so 
that 
the 
labial 
side 
of 
the 
right 
P4 
rests 
near 
the 
lingual 
side 
of 
the 
left 
M 
l 
. 
All 
of 
the 
upper 
teeth 
and 
most 
of 
the 
lowers 
are 
damaged. 
Little 
can 
be 
determined 
of 
the 
skull 
structure. 
The 
anterior 
opening 
of 
the 
infraorbital 
canal 
is 
large, 
oval, 
and 
located 
over 
P 
3 
. 
The 
maxillary-jugal 
contact 
passes 
within 
1 
mm 
of 
the 
lateral 
border 
of 
the 
infraorbital 
canal. 
The 
jugal 
bears 
an 
anteroventral 
maxillary 
process 
as 
in 
M. 
cognitus, 
M. 
parvivorus 
, 
and 
canids. 
Alveoli 
for 
the 
canine 
and 
P*-P3 
are 
present 
and 
spaced 
essentially 
as 
in 
M. 
parvivorus,
with 
the 
only 
significant 
diastemata 
anterior 
and 
posterior

1

to 
P 


P4 
differs 
from 
that 
of 
M. 
parvivorus 
in 
having 
a 
slightly 
larger 
metastyle 
and 
smaller 
parastyle. 
The 
protocone 
is 
less 
bulbous, 
though 
larger 
than 
in 
M. 
cognitus.
As 
in 
other 
miacids, 
P4 
is 
relatively 
small 
in 
proportion 
to 
the 
molars. 


M 
1 
bears 
a 
large 
lingual 
cingulum 
which 
is 
a 
continuous 
ridge 
without 
a 
distinct 
hypocone 
cusp. 
Such 
a 
cusp 
may 
have 
been 
present 
but 
removed 
by 
wear. 
The 
protocone 
is 
large, 
and 
a 
small 
metaconule 
is 
present. 
The 
anterior 
central 
portion 
of 
the 
tooth 
is 
broken 
away. 
The 
paracone 
is 
only 
slightly 
larger 
than 
the 
metacone. 
The 
stylar 
shelf 
is 
relatively 
narrow, 
and 
the 
parastyle 
is 
smaller 
relative 
to 
the 
size 
of 
the 
tooth 
than 
in 
M. 
parvivorus. 


2

The 
labial 
side 
of 
M 
is 
badly 
damaged, 
but 
it 
appears 


44 



l

to 
be 
similar 
to 
though 
smaller 
than 
M. 
A 
large 
lingualcingulum 
was 
present 
but 
is 
broken 
off. 
The 
protocone 
is 
low 
and 
broad. 
A 
tiny 
paraconule 
is 
present, 
but 
the 
metaconule 
is 
indistinct. 
The 
overall 
shape 
of 
both 
molars 
is 
similar 
to 
those 
of 
M. 
cognitus. 
No 
M 
3 
is 
preserved,
though 
it 
may 
have 
been 
present. 
The 
bone 
posterior 
to 
IVT 
is 
broken 
and 
covered 
by 
very 
hard 
matrix 
so 
that 
the 
presence 
or 
absence 
of 
alveoli 
for 
M 
3 
cannot 
be 
determined. 


The 
mandible 
is 
gently 
curved 
on 
its 
ventral 
border. 
There 
are 
two 
mental 
foramina, 
one 
beneath 
P 
and 
the 
other 
beneath 
anterior 
root 
of 
P 
The 
mandible 
tapersanteriorly. 
The 
canine 
was 
not 
large. 
P, 
is 
simple 
exceptfor 
the 
presence 
of 
a 
very 
small 
metacohid 
on 
the 
lingualside 
of 
the 
posterior 
slope 
of 
the 
protoconid. 
P 
4 
has 
a 
distinct 
hypoconid 
and 
differs 
from 
P 
of 
M. 
parvivorusin 
the 
absence 
of 
any 
paraconid 
and 
in 
the 
relativelynarrow 
posterior 
portion 
of 
the 
tooth. 
M 
resembles 
that 
of 
M. 
gracilis 
Clark 
in 
the 
shape 
of 
the 
trigonid, 
more 


than 
that 
of 
M. 
parvivorus 
, 
because 
the 
paraconidprotoconid 
blade 
is 
turned 
to 
a 
more 
anteroposteriorposition 
in 
the 
former 
species 
and 
the 
paraconid 
is 
distinctly 
larger 
than 
the 
metaconid. 
The 
talonid 
is 
basin 
shaped, 
with 
equal 
sized 
hypoconid 
and 
entoconid. 
The 
hypoconid 
is 
placed 
closer 
to 
the 
labial 
margin 
of 
the 
tooth 
than 
in 
M. 
parvivorus. 
is 
relatively 
large(Table 
10), 
similar 
to 
that 
of 
~M. 
parvivorus 
and 
somewhat 
larger 
than 
that 
of 
M. 
gracilis. 
The 
trigonid 
is 


low 
but 
all 
three 
cusps 
are 
distinct. 
The 
hypoconid 
is 
placed 
close 
to 
the 
labial 
margin 
as 
in 
M( 
. 
M, 
had 
one 
root 
and 
was 
about 
the 
same 
size 
as 
in 
M. 
parvivorus. 
Its 
crown 
is 
not 
preserved.

Comments.—The 
molar 
structure 
of 
M. 
australis 
appears 
to 
be 
similar 
to 
that 
of 
M. 
cognitus 
in 
that 
both 
preserved 
upper 
molars 
in 
both 
species 
are 
substantially 
more 
symmetrical 
about 
a 
mediolateral 
axis 
than 
is 
usual 
in 
miacids, 
because 
the 
parastyle 
is 
small 
and 
the 
metacones 
and 
paracones 
are 
similar 
in 
size 
on 
each 


Figure 
30.—Uintacyon 
scotti, 
TMM 
40165-4, 
Lower 
chisos 
Formation, 
Big 
Bend 
National 
Park. 
A. 
Skull, 
right 
lateral 
view. 
Skull 
isbadly 
crushed 
and 
distorted; 
the 
undistorted 
distance 
from 
P 
4 
to 
the 
glenoid 
fossa 
would 
have 
been 
about 
15% 
shorter 
than 
shown 
here.
Broken 
surfaces 
and 
matrix 
unshaded. 
B. 
Right 
ramus 
of 
mandible, 
associated 
with 
skull. 
Top 
of 
coronoid 
process 
and 
P 
-P 
drawn 
fromleft 
mandible, 
P 
] 
drawn 
from 
isolated 
tooth. 


45 



molar. 
Neither 
has 
an 
anterior 
cingulum. 
Miacis 
australis, 
of 
Chadronian 
age, 
is 
more 
like 
the 
Bridgerianspecies 
of 
Miacis 
than 
some 
Uintan 
miacids, 
for 
exampleTapocyon 
with 
its 
enlarged 
anterior 
M 
and 
Uintacyonscotti 
with 
its 
broadened 
upper 
molars. 
Other 
than 
a 
slight 
reduction 
of 
the 
parastyle, 
the 
upper 
teeth 
are 
little 
more 
advanced 
than 
those 
of 
the 
Bridgerian 
M. 
parvivorus. 
The 
reduction 
in 
size 
of 
the 
M 
2 
which 
is 
seen 
in 
M. 
gracilis 
and 
Procynodictis 
vulpiceps, 
both 
of 
the 
Uinta 
Formation 
of 
Utah, 
is 
not 
apparent 
in 
M. 
australis. 
M 
is 
slightly 
more 
advanced 
than 
in 
M. 
gracilis, 
but 
it 
is 
more 
primitive 
than 
in 
Hesperocyon. 
Miacis 
australis 
and 
M. 
cognitus 
are 
apparently 
relicts 
of 
the 
considerable 
radiation 
of 
miacine 
miacids 
which 
occurred 
in 
Bridgerian 
and 
Uintan 
time. 
In 
the 
Texas 
sections,
Hesperocyon 
apparently 
does 
not 
appear 
until 
after 
the 
miacids 
are 
gone. 
Though 
the 
evidence 
is 
sparse 
and 
partly 
negative, 
this 
could 
well 
be 
interpreted 
as 
a 


replacement 
of 
a 
primitive 
southern 
miacid 
fauna 
by 
more 
advanced 
and 
adaptively 
similar 
canids 
which 
may 
have 
appeared 
earlier 
in 
the 
Cypress 
Hills 
to 
the 
north. 


Genus 
Uintacyon 
Leidy, 
1872 


Uintacyon 
scotti 
(Wortman 
and 
Matthew, 
1899) 


Figs. 
30-31, 
Table 
11 


Synonyms. 
—Prodaphaenus 
scotti 
Wortman 
and 
Matthew, 
1899; 
Uintacyon 
(Miocyon) 
scotti, 
Matthew, 
1909. 


Type.—PU 
11238,P3-M2 
upper 
part 
ofWagonhound

, 


Member, 
(Uinta 
“B”), 
Uinta 
Formation, 
Utah. 
Referred 
Material.—TMM 
40165-4, 
skeleton 
with 


’

M23

skull, 
left 
canine, 
P3-M3, 
right 
P4, 
, 
mandible 
with 
canines, 
left 
P3-M( 
, 
right 
P 
M[( 
all 
or 
parts 
of 
15 
vertebrae 
and 
several 
ribs, 
,all 
or 
parts 
of 
scapula,
humerus, 
radius, 
ulna, 
femur, 
tibia, 
fibula, 
patella,
tarsals, 
metatarsals, 
and 
phalanges. 
Questionably 
referred; 
TMM 
41372-362, 
partial 
M 
1; 
41372-367, 
P4; 
41672-59,

partial 
M,.

Stratigraphic 
position.—Lower 
part 
of 
Chisos 
Formation; 
probably 
Whistler 
Squat 
and 
Serendipity 
local 
faunas. 
The 
partial 
skeleton 
came 
from 
low 
inthe 
Chisos 
Formation, 
about 
20.7 
to 
27.4 
m 
(68 
to 
90 
feet) 
above 


the 
Alamo 
Creek 
Basalt 
(locality 
T 
14 
of 
Plate 
11, 
Maxwell 
et 
al., 
1967). 
The 
M 
1 
and 
P4 
are 
from 
the 
Whistler 
Squat 
local 
fauna, 
lower 
15.2 
m 
(50 
feet) 
of 
the 
Devil’s 
Graveyard 
Formation. 
The 
M 
is 
from 
the 
PurpleBench 
locality, 
approximately 
67 
m 
(220 
feet)
stratigraphically 
higher 
in 
the 
Devil’s 
GraveyardFormation, 
all 
from 
Brewster 
County, 
Texas. 


Age. 
—Uintan 
(late 
Eocene).

Description. 
—The 
skeleton 
(40165-4) 
allows 
the 
first 
description 
of 
associated 
upper 
and 
lower 
dentitions 
of 
this 
species. 
The 
upper 
canine 
is 
straight 
and 
onlyslightly 
compressed. 
The 
posterior, 
medial 
and 
lateral 
surfaces 
are 
flattened, 
forming 
sharp 
ridges 
at 
their 
junctions. 
A 
low 
anteromedial 
ridge 
is 
also 
present. 
The 
alveolus 
for 
P1 
is 
closely 
appressed 
to 
the 
canine 
alveolus. 
It 
is 
slightly 
larger 
than 
either 
of 
the 
alveoli 
of 
P2 
. 
P3 
and 
P4 
are 
like 
those 
of 
the 
type. 
P3 
is 
short 
and 
wide, 
with 
a 
strong 
cingulum 
raised 
at 
the 
rear 
into 
a 
small 
cusp. 
P 
4 
is 
also 
short 
anteroposteriorly, 
with 
a 
large 
protocone, 
central 
paracone, 
relatively 
small 
metastyle, 
and 
narrow 
labial 
cingulum. 
M 
1 
of 
the 
Texas 
specimen 
shows 
some 
minor 
differences 
from 
that 
of 
the 
type. 
The 
parastyle 
is 
smaller 
and 
the 
stylar 
cingulum 
is 
not 
divided 
into 
two 
cusps 
as 
on 
PU 
11238. 
M 
is 
distinctive. 
It 
is 
long 
anteroposteriorly 
and 
is 
low 
crowned. 
The 
metacone 
is 
much 
larger 
than 
in 
U. 
vorax 
from 
the 
Bridgerian, 
and 
the 
tooth 
is 
almost 
completelyrimmed 
by 
a 
broad 
cingulum 
which 
connects 
the 
indistinct 
protocone 
with 
the 
large 
parastyle, 
the 
stylarcingulum, 
and 
the 
posterior 
side 
of 
the 
metacone. 
There 


1 
23

is 
no 
trace 
of 
a 
hypocone 
on 
either 
M 
or 
M 
. 
M 
is 
oval 
and 
small 
but 
functional. 
A 
low 
ridge 
parallels 
the 
rounded 
edges 
of 
the 
tooth, 
and 
the 
only 
discernible 
cuspis 
an 
indistinct 
paracone. 
The 
right 
M 
3 
has 
two 
roots, 
whereas 
the 
left 
has 
three. 


2

An 
isolated 
lower 
incisor 
resembles 
an 
I 
of 
Canis. 
The 
crown 
includes 
a 
main 
medial 
cusp 
and 
a 
small 
lateral 
cusp. 
The 
lower 
canines 
are 
tall 
and 
slightlyrecurved, 
and 
like 
the 
upper 
canine 
have 
a 
sharp 
posteriorridge 
and 
a 
less 
prominent 
anteromedial 
ridge. 
The 
lateral 
side 
is 
longitudinally 
grooved. 
An 
isolated 
P 
] 
is 
a 
simpletooth 
similar 
to, 
though 
smaller 
than, 
its 
equivalent 
in 
Canis 
latrans. 
The 
anterior 
mandible 
of 
?U. 
scotti 
figured 
by 
Wortman 
(1901, 
Fig. 
17), 
if 
it 
belongs 
with 


Figure31.—Uintacyonscotti,TMM40165-4,LowerChisosFormation, 
BigBendNationalPark.Rightupperdentition,occlusalview,positionofalveoliforC-P~showninoutline.Breakagehasseparatedtheteeth. 
M2drawnfromleftside. 


46 



Table 
11.—Measurements 
of 
dentitions 
of 
Uintacyon 
scotti. 


U. 
scotti 
U. 
sp. 
U. 
scotti 
typeTMM 
40165-4 
TMM 
41372-367 
PU 
11238 
right 
left 
c 
AP 
9.5 


T 
7.8 
p3 
AP 
7.0 
7.6 


T 
4.5 
5.1 
p4 
AP 
11.6 
11.8 
11.4 
T 
9.4 
9.2 
10.5 


1

M 
AP 
9.7 
10.9 
T 
13.4 
16.3 
M2 
AP 
9e 
8.8 
8.5 


T 
12e 
11.6 
12.3e 


M3 
AP 
4.8 
4.9 
T 
6.3 
6.5 
C 
AP 
8.8 
8.7 
T 
6.5 
6.8 


P.1 
AP 
3.2a 
4.0 
P
P
P
P
T 
2.5 
2 
AP 
6.2a 
6.2a 
L3 
AP 
6.2a 
5.4 
P
P
j 
T 
3.5 
4 
AP 
8.2a 
8.7 
7.9 
4 
T 
4.6 
4.8 
MAP 
13.4 
12.4 


1 


M
M
m
m
i 
T 
7.1 
7.2 
2 
AP 
10.3a 
10.2a 
3 
AP 
5.3a 
4.5a 
M 
1-M 
27.5a 
27a 


1 
56a 
60a

prm/ 


this 
taxon, 
is 
abnormal 
in 
having 
two 
single 
rooted 
anterior 
premolars. 
As 
nearly 
as 
can 
be 
told 
from 
the 
alveoli, 
was 
about 
the 
same 
size 
as 
P3-P 
3 
is 
small, 
with 
a 
distinct 
hypoconid 
and 
posterior 
lingual 
cingulum. 
is 
also 
relatively 
small, 
with 
the 
posterad-leaning 
main

P4 
cusp 
that 
is 
typical 
of 
Uintacyon.
is 
tall, 
with 
a 
well 
developed 
protoconid-paraconidblade 
which 
is 
oriented 
more 
anteroposteriorly 
than 
in 
U. 
bathygnathus. 
The 
paraconid 
and 
metaconid 
are 
well 
separated 
and 
equal 
in 
size. 
The 
anterior 
labial 
cingulumis 
present 
but 
is 
not 
as 
large 
as 
on 
U. 
bathygnathus.
The 
talonid 
is 
typical 
of 
Uintacyon, 
consisting 
mostly 
of 
a 
large 
centrally 
placed 
hypoconid, 
with 
a 
small 
marginal 
entoconid. 
The 
alveoli 
of 
show 
that, 
like 
M 
the

, 


tooth 
was 
elongated 
anteroposteriorly 
and 
was 
almost 
as 
long 
as 
M 
. 
was 
also 
relatively 
large, 
though 
singlerooted. 


The 
mandible 
is 
almost 
identical 
in 
length 
to 
that 
of 
Canis 
latrans, 
but 
it 
is 
deeper, 
straighter 
along 
the 
ventral 
margin, 
and 
has 
a 
more 
abrupt 
and 
distinct 
chin. 


The 
tooth 
row 
of 
U. 
scotti 
is 
relatively 
short 
(Table 
11), 


and 
the 
coronoid 
process 
and 
masseteric.fossa 
are 
very 


broad. 
The 
masseteric 
fossa 
is 
extremely 
deep 
and 
its 
margins 
are 
sharply 
delineated. 
The 
angular 
process 
is 
large 
and 
is 
set 
lower 
on 
the 
mandible 
than 
in 
Canis. 


The 
skull 
is 
badly 
crushed, 
and 
the 
premaxillaries,
part 
of 
the 
left 
maxillary, 
zygomatic 
arches, 
and 
the 
posterior 
part 
of 
the 
braincase 
are 
missing. 
The 
rostrum 


was 
relatively 
short 
and 
high, 
though 
somewhat 
less 
blunt 
than 
in 
the 
recent 
large 
cats. 
The 
length 
of 
the 
tooth 
row 
relative 
to 
the 
mandible 
length 
is 
more 
similar 
to 
the 
proportions 
found 
in 
Felis 
concolor 
than 
in 
Canis 
latrans, 
both 
of 
which 
have 
a 
condyle-symphysis 
length 
about 
equal 
to 
that 
of 
U. 
scotti. 
Unlike 
the 
cats, 
U. 
scotti 
has 
not 
lost 
the 
posterior 
molars, 
but 
instead 
reduced 
the 
relative 
tooth 
row 
length, 
although 
retaining 


all 
the 
teeth, 
by 
reducing 
the 
relative 
size 
of 
the 
premolars. 
In 
dorsal 
view 
the 
skull 
is 
more 
catlike 
than 
doglike,
with 
a 
short 
snout 
and 
broad 
frontals. 
Small 
postorbital 


47 



processes 
are 
present 
on 
the 
frontals. 
The 
postorbitalcranial 
constriction 
is 
relatively 
more 
posterior 
than 
in 
living 
carnivores, 
mainly 
because 
of 
the 
small 
size 
of 
the 
anterior 
cerebrum. 
A 
strong 
sagittal 
crest 
was 
present.
From 
the 
side 
the 
dorsal 
outline 
of 
the 
skull 
is 
arched 
like 
that 
of 
Felis 
concolor. 
The 
orbit 
is 
relatively 
smaller 
than 
in 
Felis 
concolor. 


The 
internal 
nares 
opened 
between 
the 
last 
molars. 
The 
anteroventral 
rim 
of 
the 
narial 
opening 
is 
edged 
by 
a 


strong 
ridge, 
and 
a 
deep 
groove 
separates 
this 
ridge 
from 
theM 
3. 


The 
cranial 
foramina 
are 
largely 
obscured 
by 
breakage 
or 
missing 
bone. 
It 
is 
clear 
that 
there 
is 
a 
postglenoidforamen, 
situated 
on 
the 
posteromedial 
side 
of 
the 
postglenoid 
process. 
The 
foramen 
ovale 
is 
present 
and 
is 
large. 
I 
cannot 
be 
sure 
whether 
an 
alisphenoid 
canal 
is 


present.

Major 
portions 
of 
the 
postcranial 
skeleton 
are 
present.
These 
bones 
will 
have 
to 
await 
a 
more 
thoroughdescription 
than 
I 
can 
present 
here. 
It 
is 
worthy 
of 
note 
that 
where 
comparable 
the 
bones 
are 
similar 
(except 
for 
larger 
size) 
to 
those 
of 
U. 
bathygnathus.

The 
P 
4 
and 
Whistler 
Squat 
are 
not

M 
1 
from

particularly 
diagnostic. 
They 
unquestionably 
are 


Uintacyon. 
and 
are 
assigned 
to 
U. 
scotti 
on 
the 
basis 
of 
general 
similarity 
to 
the 
Chisos 
Formation 
specimen.
However, 
they 
could 
also 
belong 
to 
U. 
bathygnathus.
The 
partial 
from 
Purple 
Bench, 
although 
resemblingthe 
equivalent 
anterior 
portion 
of 
the 
Uintacyon 
molar 
in 
size 
and 
form, 
lacks 
the 
diagnostic 
talonid 
and 
maybelong 
to 
another 
genus. 


Relationships. 
—The 
most 
probable 
ancestor 
for 
Uintacyon 
scotti 
is 
U. 
bathygnathus 
(Scott, 
1888), 
from 
the 
Bridger 
Formation 
of 
Wyoming. 
Uintacyon 
majorMatthew 
(1909) 
is 
so 
similar 
in 
size 
and 
morphology 
to 
the 
type 
of 
U. 
bathygnathus 
(PU 
10607) 
that 
there 
can 
be 
little 
doubt 
that 
it 
is 
a 
junior 
synonym. 
All 
known 
specimens 
of 
U. 
bathygnathus 
(including 
U. 
major) 
are 
probably 
from 
the 
upper 
Bridger 
(C 
level), 
though 
the 
exact 
locality 
of 
the 
type 
is 
not 
known 
(Gazin, 
1976).
Uintacyon 
scotti 
is 
at 
the 
end 
of 
a 
branch 
of 
the 
Uintacyon 
lineage 
in 
which 
increase 
in 
size 
and 
increase 


in 
the 
crushing 
function 
of 
the 
molars 
were 
dominant 
features. 


Family 
Amphicyonidae 


Subfamily 
Daphoeninae 


Genus 
Daphoenus 
Leidy, 
1853 


Daphoenus 
cf. 
D. 
lambei 
Russell. 
1934 


Porvenir 
local 
fauna 


Figure 
32c, 
d. 
Tables 
12-13 
Type.—NMC 
6513, 
left 
P4 
Cypress 
Hills, 
Saskatchewan. 
, 


Referred 
material.—OU 
32-2-S2, 
maxillary 
fragmentwith 
part 
of 
P3 
and 
P4-M2; 
FMNH:PM 
31, 
mandible 
fragment 
with 
Mv 


Stratigraphic 
position.—Lower 
part 
of 
Chambers 
Formation, 
Presidio 
Co., 
Texas. 


Age. 
—Early 
Chadronian. 


Description.—A 
portion 
of 
a 
maxilla 
(Fig. 
32c)
12

retains 
a 
small 
part 
of 
P3 
a 
damaged 
P4. 
and 
M 
’ 
. 


, 


Nothing 
can 
be 
said 
of 
the 
characters 
of 
P' 
because 
of 
the 
damage. 
P4 
is 
similar 
in 
shape 
to 
that 
of 
Daphoenus 
vetus 
(Brule 
Formation. 
Orellan) 
but 
is 
much 
smaller 
(Table 
12), 
less 
robust, 
and 
higher 
crowned. 
The 
protocone 
is 
large 
and 
conical. 
Strong 
lingual 
and 
anterior 
cingula 
are 
present, 
and 
the 
small 
part 
of 
the 
central 
labial 
side 
preserved 
carries 
a 
distinct 
cingulum. 
I 
can 
find 
no 
characters 
in 
which 
this 
tooth 
differs 
from 
the 
type 
of 
Daphoenus 
lambei 
Russell 
(1934). 
M 
1is 
also 
similar 
to 
that 
of 
D. 
vetus 
, 
though 
smaller. 
The 
stylarshelf 
is 
broad 
and 
the 
stylar 
cingulum 
is 
low. 
The 
parastyle 
is 
less 
obvious 
than 
in 
Hesperocyon 
or 
Miacis 
parvivorus 
, 
but 
is 
still 
distinct. 
The 
paracone 
and 
metacone 
are 
of 
equal 
size. 
The 
paraconule 
and 
metaconule 
are 
distinct. 
The 
protocone 
is 
anteriorlyplaced. 
As 
is 
common 
in 
Daphoenus, 
the 
hypocone 
is 
larger 
than 
the 
protocone, 
rounded 
and 
even 
bulbous, 
and 
set 
at 
the 
lingual 
end 
of 
the 
tooth, 
giving 
it 
a 
laterally 


,

elongated 
appearance. 
M" 
is 
smaller 
than 
M 
l 
with 
a 
narrower 
stylar 
shelf 
which 
is 
broadest 
opposite 
the 
paracone. 
The 
parastyle 
is 
small 
and 
rounded. 
The 
paracone 
is 
low 
and 
also 
rounded. 
There 
is 
a 
very 
small 
paraconule, 
but 
the 
posterior 
part 
of 
the 
tooth 
including 
metacone 
and 
metaconule 
is 
damaged. 
The 
protocone 
is 
an 
indistinct 
point 
on 
a 
low, 
rounded 
ridge. 
The 
hypocone 
and 
attached 
cingulum 
are 
even 
more 
bulbous 
than 
in 
the 
M 
l 
, 
and 
the 
hypocone 
is 
much 
larger 
than 
the 
protocone.

(Fig. 
32d) 
is 
much 
like 
that 
of 
other 
species 
of 
Daphoenus. 
though 
smaller 
(Table 
13). 
The 
trigonid 
is 
short, 
with 
the 
protoconid 
slightly 
larger 
than 
the 
metaconid, 
and 
the 
paraconid 
smaller, 
but 
still 
distinct. 


The 
talonid 
is 
basined, 
with 
the 
central 
depressionsurrounded 
by 
the 
lingual 
and 
posterior 
talonid 
crests. 
The 
hypoconid 
and 
crista 
obliqua 
are 
relatively 
low. 
There 
is 
a 
small 
anterior 
labial 
cingulum. 


Relationships. 
—The 
maxillary 
(OU 
32-2-S2) 


was 
described 
and 
figured 
by 
Stovall 
(1948, 
Fig. 
3), 
who 


considered 
it 
to 
be 
a 
new 
genus 
and 
species 
of 
canid,
although 
he 
did 
not 
name 
it. 
However, 
its 
veryDaphoenus-Yike 
molars 
and 
the 
virtual 
identity 
in 
shape

and 


size 
of 
P4 
to 
the 
type 
of 
D. 
lambei 
Russell 
(1934;
1972, 
Fig. 
12d 
and 
e) 
leave 
little 
doubt 
of 
its 
close 


relationship 
to 
D. 
lambei. 
Russell 
(1972) 
illustrated 
a 
referred 
M 
1 
of 
D. 
lambei, 
which 
looks 
more 
like 
the 
M 
2 
of 
a 
larger 
species 
of 
Daphoenus. 
This 
isolated 
tooth 
(NMC 
9207) 
is 
practically 
an 
enlarged 
duplicate 
of 
the 
M“ 
of 
the 
Vieja 
specimen, 
especially 
in 
its 
narrow 
posterior 
stylar 
shelf, 
indented 
posterior 
margin, 
and 
verylarge 
bulbous 
hypocone. 
Both 
this 
and 
another 
referred 
M 
from 
Cypress 
Hills 
(ROM 
6331) 
are 
close 
in 
size 
to 


48 



the 
M 
1 
of 
OU 
32-2-S2. 
The 
M, 
(FMNH:PM 
31) 
is 
similar 
to 
D. 
vetus 
in 
form, 
but 
smaller 
than 
in 
D. 
hartshomianus 
and 
may 
reasonably 
be 
placed 
in 
Daphoenus 
cf. 
D. 
lambei. 
D. 
hartshomianus 
(Cope),
from 
Orellan 
deposits, 
which 
is 
closest 
in 
size 
to 
D. 
lambei. 
is 
larger 
except 
for 
one 
questionably 
referred 
P4 
from 
Cypress 
Hills. 
Daphoenus 
minimus 
Hough 
(1948) 
was 
even 
smaller 
than 
D. 
lambei. 
Hough’s 
photographs(1948, 
Plate 
86, 
Figs. 
3 
and 
4) 
show 
too 
little 
detail 
to 
be 
sure 
of 
either 
tooth 
or 
auditory 
structure, 
and 
the 
measurements, 
given 
apparently 
under 
the 
name 
“D. 
gracilis” 
(a 
nomen 
nudum), 
are 
very 
close 
to 
those 
of 
Hesperocyon. 


Comments.—R.M. 
Hunt 
believes 
(from 
unpublisheddata) 
that 
only 
two 
Oligocene 
species 
of 
Daphoenusshould 
be 
recognized, 
namely 
D. 
vetus 
and 
D. 
hartshomianus. 
If 
this 
is 
true 
these 
specimens 
would 
either 
have 
to 
be 
placed 
into 
one 
of 
these 
species(presumably 
D. 
hartshomianus 
as 
it 
is 
more 
similar 
in 
size) 
or 
set 
up 
as 
a 
new 
species. 
I 
cannot 
demonstrate 
this 
conclusion 
by 
published 
data. 
The 
Texas 
specimens 
are 
not 
greatly 
different 
in 
size 
from 
specimens 
referred 
to 
Daphoenus 
demilo 
Dawson, 
1980, 
(M 
1 
AP-7.5, 
T= 
12.5, 
vs. 
OU 
32-2-S2 
AP-7.9, 
T= 
12.1). 
However, 


Daphoenus 
demilo 
is 
more 
primitive 
in 
having 
a 
wider 
stylar 
shelf 
and 
smaller 
hypocone. 


Figure 
32.—A-B. 
Daphoenocyon 
dodgei 
(Scott), 
little 
Egypt 
Local 
fauna. 
A. 
TMM 
40209-205, 
P3 
-M', 
occlusal 
view. 
B. 
TMM 
40209204, 
talonid 
of 
M 
and 
complete 
Mocclusal 
view. 
C-D. 
Daphoenus 
cf. 
D. 
lambei, 
Porvenir 
local 
fauna. 
C. 
OU 
32-2-S2, 
maxillary

vfragment 
with 
P4-M", 
occlusal 
view' 
D. 
FMNH 
PM 
31, 
M 
occlusal 
view. 
E, 
Daphoenine, 
genus 
and 
species 
indeterminate, 
TMM 


40688-72,Porvenirlocalfauna. 
Partialramusofmandiblewith,P4-M,,2 
occlusalandouterviews. 


49 



9114

dodgei 
9256 
Daphoenus 
NMC

11.9 
6.2 
7e 
Oe

D.CM 
4.7 
3.2 
7.4 
5.1 
9.5 
6.4 
14.4 
11.0 
10.16.34.2 


dodgei

dodgei 
(type)

Daphoenocyon11422

PU 
6.0 
8.0 
11.0 
5.0 
14.0 
8.0 
23.0

Daphoenocyon40209-205

TMM4.7a 
9.0a 
9.5 
5.3 
14.0 
11.5 
9.7 
16.5 
7.3 
10.8 
38.4 
17.6 
indet.

6331 
Daphoeninae40203-16

lambeiTMM4e12. 
7.0e 


D. 
ROM8.0 
12.3 
indet.

le 
5e

lambei 
9207 
Daphoeninae40688-72 


D. 
NMC7.2 
10.6 
TMM6.5a 
7.9a 
9. 
3.5e 
12.5 
6.7 
7.0e 
4.0 
25.15.8 
type 
lambei

6513 


cf. 
31

lambei

12.0 
7.2 
PM
D.
Daphoeninae. 
NMCD. 


DaphoenusFMNH

of7.8 
5.4 
17e

lambei 


D.
dentitions

upper 
dodgei 
40209-784 


7e

ofcf. 
32-2-S2 
Daphoeninae. 
D.TMM11.6 
5.9 
13.6.7e 
10.0 
6.8 
4.5a 
28a 
24.2 


2e 
of

DaphoenusOU 
8.4e 
11.7.6e 
7.9 
12.1 
5.8 
9.4 
13.6 
dodgei

dentitions

AP 
AP 
AP 
AP 
AP 
AP4 
lower

TTT 
TT 
T 


P2-M 
of 


11 


1 


23

12.—Measurements 
13.—Measurements

Daphoenocyon40209-204

Table 
p 
p 
p 
4pM 
2M 
1P-
M 
TMM7.8e 
9.6 
6.5 
4.8a 
23e 


M.
AP 
AP 
APT 
APT 
APT 
AP 
depth, 


24

Table 
pJM. 
M2Z 
-anterior

PPI 
Wl3MJaw

50 



Genus 
Daphoenocyon 
Hough, 
1948 


Daphoenocyon 
dodgei 
Scott, 
1898 


Little 
Egypt 
local 
fauna 


Figures 
32a, 
b. 
Table 
12-13 


Type.—PUll422, 
Chadron 
Formation 
(“Titanotherium 
beds”). 
Hat 
Creek 
Basin, 
Nebraska. 


Referred 
materials.—TMM 
40209-204, 
ramus 
fragment 
with 
talonid 
of 
M., 
and 
alveolus 
of 


' 


2

40209-205, 
maxilla 
with 
alveoli 
of 
C, 
P 
1, 
crowns 
of 


v 


3

P 
-M“, 
and 
alveolus 
of 
M 
; 
40209-784, 
ramus 
fragmentwithpartialP„, 
rootsofM,,entireM,androotof!VL.

r 


412 
3 
Stratigraphic 
position. 
—Reeves 
Bone 
Bed, 
upper 
part 
of 
Chambers 
Formation, 
Presidio 
Co., 
Texas. 


Age. 
—Chadronian 
(early 
Oligocene). 


Description. 
—The 
skull 
fragment 
shows 
very 
clearlythe 
short 
face 
and 
crowded 
premolars 
characteristic 
of 
Daphoenocyon. 
There 
are 
no 
diastemata. 
P 
1 
had 
one 
root 
slightly 
longer 
than 
either 
of 
the 
roots 
of 
P" 
(Fig. 
32a).
P' 
is 
a 
broad, 
anteroposteriorly 
elongated 
cone 
with 
an 
encircling 
cingulum. 
There 
is 
a 
small 
posterior 
cusp 
and 
a 
small 
cingular 
cusp. 
P4 
is 
a 
wide, 
very 
robust 
tooth 
with 
a 
relatively 
low 
paracone 
and 
short 
metastyle. 
The 
cingulum 
completely 
surrounds 
the 
tooth. 
The 
parastyleforms 
a 
raised, 
enlarged 
anterolabial 
portion 
of 
the 
cingulum. 
The 
protocone 
is 
large. 
M 
1 
is 
shorter 
anteroposteriorly 
than 
that 
of 
Daphoenns 
veins. 
and 
is 
similar 
to 
M 
1 
of 
the 
paratype 
specimen 
of 
D. 
dodgei 
(CM 


9256) 
listed 
by 
Clark 
et 
al. 
(1967). 
There 
is 
a 
large 
but 
low 
parastyle, 
a 
broad 
stylar 
shelf 
slightly 
wider 
than 
in 
CM 
9256, 
and 
a 
low 
and 
rounded 
paracone 
and 
metacone. 
The 
paraconule 
is 
less 
obvious 
than 
in 
CM 
9256. 
but 
the 
metaconule 
is 
more 
prominent. 
The 
protocone 
is 
the 
high 
point 
on 
a 
curved 
V-shaped 
ridge.
The 
hypocone 
is 
larger 
than 
the 
protocone 
and 
is 
separated 
from 
it 
by 
a 
sharp 
groove. 
This 
groove 
is 
less 
distinct 
on 
CM 
9256. 
M" 
is 
essentially 
like 
that 
described 
for 
Daphoenns 
lambeibut 
larger, 
broader,

, 


and 
with 
a 
relatively 
smaller 
hypocone. 
Parts 
of 
alveoli 
for 
two 
roots 
of 
M 
3 
are 
preserved.

The 
mandible 
has 
the 
thick, 
deep 
horizontal 
ramus 
characteristic 
of 
Daphoenocyon. 
The 
only 
premolar 
preserved 
is 
? 
-It 
shows 
the 
same 
outline 
as 
that 
of 
the

4 


type 
(Scott, 
1898, 
Plate 
19, 
Figs. 
6 
and 
7), 
with 
a 
complete 
cingulum. 
The 
protoconid 
is 
broken 
off, 
so 
that 
the 
only 
cusp 
visible 
is 
a 
tiny 
posthypoconid 
just 
within 
the 
posterior 
cingulum. 
A 
portion 
of 
the 
talonid 
of 
M 
is 
preserved 
on 
40209-204 
(Fig. 
32b). 
There 
is 
a 
short 
but 
prominent 
hypoconid 
and 
crista 
obliqua. 
a 
less 
prominentlingual 
talonid 
crest, 
and 
a 
distinct 
posterior 
labial 
cingulum. 
The 
talonid 
basin 
is 
shallow 
and 
opensposteriorly. 
M 
0 
is 
present 
on 
both 
mandible 
fragments.
The 
crown 
is 
"broad 
and 
low, 
with 
rounded 
crests 
and 
cusps 
and 
shallow 
basins. 
The 
protoconid 
and 
paraconid 
are 
strongly 
convex 
anteriorly, 
without 
a 
distinct 
camassial 
notch. 
The 
hypoconid 
and 
crista 
obliqua 
are 


low 
but 
distinct. 
The 
entoconid 
is 
only 
a 
wide 
spot 
on 
the 
lingual 
talonid 
crest. 
A 
small 
but 
prominent 
anterior 
labial 
cingulum 
is 
present. 
was 
small, 
with 
a 
single 


root 
which 
was 
indented 
on 
both 
sides. 


Relationships.—These 
specimens 
show 
the 
genericfeatures 
of 
Daphoenocyon 
Hough 
(1948) 
and 
are 
within 
the 
size 
range 
of 
D. 
dodgei. 
This 
taxon 
has 
had 
a 
fairlysimple 
history. 
It 
was 
proposed 
by 
Hough 
to 
include 
?Daphoenus 
dodgei 
Scott 
(1898), 
then 
later 
redefined 
byClark 
et 
al. 
(1967) 
to 
include 
a 
slightly 
different 
hypodigm. 
The 
genus 
Daphoenocyon 
is, 
however, 
questionably 
distinct 
from 
Brachyrhynchocyon 
Loomis 
(in 
Scott 
and 
Jepsen, 
1936), 
as 
both 
are 
daphoeninecarnivores 
with 
short 
faces 
and 
trenchant 
talonids 
on 
M,.

3 
i•

The 
M 
has 
apparently 
been 
lost 
in 
B. 
intermedins, 
and 
this 
may 
be 
adequate 
reason 
for 
generic 
separation. 
The 
Chadronian 
species 
Daphoenocyon 
minor 
(Clark 
et 
al., 
1967) 
is 
about 
the 
same 
size 
as 
B. 
intermedins, 
a 
Whitneyan 
species. 
Daphoenocyon 
dodgei 
is 
larger 
and 
is 
distinct 
from 
either 
D. 
minor 
or 
B. 
intermedins 
on 
the 
basis 
of 
size. 


Daphoeninae 
genus 
and 
species 
indeterminate 


Porvenir 
local 
fauna 


Figure 
32e, 
Table 
13 


Referred 
Material.—TMM 
40688-72, 
ramus 
with 
roots 
of 
C-P.,, 
parts 
of 
crowns 
of 
P4-Mroots 
of 
M 
;

9?

40203-16, 
damaged 
M 
. 
, 


Stratigraphic 
position.—Between 
0 
and 
26.8 
m 
(0and 
88 
ft) 
above 
the 
lower 
marker 
bed. 
Chambers 
Formation. 


Age.—Early 
Chadronian. 


Description. 
—The 
more 
complete 
specimen, 
4068872, 
is 
a 
shallow 
mandible 
with 
no 
complete 
teeth 
preserved. 
There 
is 
a 
short 
diastema 
between 
each 
pair 
of 
premolars 
exceptparaconid, 
a 
large 
protoconid, 
a 
hypoconid, 
and 
a 
hypoconulid, 
of 
which 
only 
the 
last 
is 
undamaged. 
The 
cusps 
of 
the 
trigonid 
of 
M 
are 
broken 
off. 
The 
trigonidmakes 
up 
more 
than 
70% 
of 
the 
length 
of 
the 
tooth 
and 
is 
relatively 
larger 
than 
in 
Daphoenns 
vetus. 
The 
talonid 
is 
trenchant, 
dominated 
by 
a 
large 
hypoconid 
and 
crista 
obliqua. 
The 
lingual 
talonid 
crest 
with 
its 
small 
entoconid 
encloses 
only 
a 
small 
lateral 
basin. 
is 
relatively 
smaller 
than 
in 
Daphoenns 
or 
Daphoenocyon 
, 
and 
is 
relatively 
only 
slightly 
larger 
than 
in 
Daphoenictis. 
The 
protoconid 
is 
large 
and 
is 
taller 
than 
the 
metaconid, 
which 
in 
turn 
is 
larger 
than 
the 
paraconid. 
There 
is 
a 
distinct 
anterior 
labial 
cingulum. 
The 
talonid 
is 
broken, 
but 
appears 
to 
have 
been 
small 
and 
narrow. 
M, 
had 
two 
roots. 


Another 
M 
, 
TMM 
40203-16, 
is 
a 
little 
larger 
than 
that 
of 
40688-/2, 
and 
probably 
represents 
a 
daphoenine.
The 
talonid 
is 
broken 
off 
and 
its 
characters 
cannot 
be 
ascertained. 


51 



Figure 
33.—Hesperocyon 
wilsoni 
n. 
sp., 
type, 
TMM 
40504-126, 
Airstrip 
local 
fauna. 
A, 
Skull, 
ventral 
view. 
B. 
Skull, 
dorsal 
view. 
C 
Skull 
and 
right 
ramus 
of 
mandible, 
lateral 
view. 


52 



Relationships.—The 
mandible 
40688-72 
differs 
from 
Daphoenus 
and 
Daphoenocyon 
in 
its 
reduced 
talonid 
on 
Mj 
and 
reduced 
Mr 
It 
resembles 
Daphoenictis 
tedfordi 
Hunt 
(1974b) 
in 
the 
enlarged 
trigonid 
and 
trenchant 
talonid 
of 
and 
the 
reduced 
M, 
with 
a 
largeprotoconid. 
This 
specimen 
does 
not 
sfiow 
the 
extreme 
specializations 
of 
Daphoenicitis 
such 
as 
reduction 
of 
the

, 


paraconid 
and 
loss 
of 
the 
metaconid 
of 
M 
. 
The 


characters 
of 
the 
premolars, 
which 
are 
tall 
and 
narrow 
with 
enlarged 
hypoconids 
in 
Daphoenictis, 
are 
not 
known 
in 
40203-16, 
though 
certainly 
a 
large 
hypoconidwas 
present 
on 
P 
This 
specimen 
may 
represent 
a 
new 
species 
related 
toDaphoenictis 
, 
but 
it 
is 
not 
well 
enoughpreserved 
to 
serve 
as 
a 
type. 


Family 
Canidae 


Genus 
Hesperocyon 
Scott, 
1890 


Hesperocyon 
wilsoni 
n. 
sp. 


Airstrip 
local 
fauna 


Figures 
33-34, 
Tables 
14-17 


Type.—TMM 
40504-126, 
skull 
and 
mandible. 


Referred 
material.—TMM 
40504-99, 
mandible 
with 
right 
and 
left 
P.,

Stratigraphic 
position. 
—Capote 
Mountain 
Tuff,
approximately 
174 
m 
(570 
feet) 
above 
Bracks 
Rhyolite,
Presidio 
Co., 
Texas. 


Age.—Chadronian 
(early 
Oligocene).

Etymology. 
—Named 
for 
Dr. 
J. 
A. 
Wilson. 


Diagnosis.—Skull 
length 
within 
the 
range 
of 
H. 
gregarius. 
Measurements 
of 
P3 
(AP), 
P*-M 
2 
p’-P4 
, 
P 
, 


, 


P 
(AP), 
Pj-M3and 
P 
all 
larger 
than 
observed 
rangeof 
H. 
gregarius., 
Molar 
measurements 
and 
morphologywithin 
the 
range 
of 
H. 
gregarius, 
except 
metacone 
on 
M‘ 
greatly 
reduced. 
M 
3 
absent. 


Description. 
—The 
skull 
(Fig. 
33) 
is 
crushed 
laterallyand 
much 
of 
the 
bone 
has 
been 
weathered, 
but 
the 
teeth 
are 
not 
badly 
damaged. 
The 
preorbital 
portion 
of 
the 
skull 
is 
longer 
and 
was 
probably 
relatively 
narrower 
than 
in 
Hesperocyon 
gregarius 
(compared 
to 
several 
specimens 
from 
the 
Brule 
Formation, 
Orellan, 
South 
Dakota), 
although 
the 
narrowness 
of 
the 
snout 
has 
certainly 
been 
exaggerated 
by 
the 
crushing. 
The 
nasals 
are 
slightly 
longer 
than 
in 
H. 
gregarius, 
and 
the 
dorsal 
process 
of 
the 
premaxillary 
is 
distinctly 
broader 
and 
longer. 
The 
postorbital 
constriction 
is 
not 
as 
pronouncedin 
the 
Vieja 
skull 
as 
in 
H. 
gregarius, 
though 
distortion 
of 
the 
bones 
and 
the 
endocast 
may 
be 
partly 
responsible 
for 
this 
difference. 
The 
sagittal 
crest 
was 
lyrate, 
with 
the 
two 
anterior 
branches 
of 
the 
crest 
remaining 
7 
mm 
apart 
as 
far 
posterior 
as 
the 
frontal-parietal 
suture. 
A 
distorted 
partial 
endocast 
shows 
two 
well 
developed 
sulci, 
the 
coronolateral 
and 
suprasylvian. 
Between 
the 
posteriorends 
of 
these 
sulci 
there 
is 
a 
dimple 
or 
incipient 
sulcus,
possibly 
representing 
the 
ectolateral 
sulcus. 
There 
is 
a 
depression 
medial 
to 
the 
posterior 
end 
of 
the 
coronolateral 
for 
which 
I 
cannot 
find 
an 
equivalent 
in 
H. 


Figure 
34. 
—Hesperocyon 
wilsoni 
n. 
sp., 
type, 
TMM 
40504-126, 
Airstrip 
local 
fauna. 
A. 
Upper 
dentition, 
occlusal 
view. 
B. 
Lower 
dentition, 
occlusal 
view. 


53 



Fm

Uinta
gracilis 
type 
Member,


11900,

MiacisCMMyton3.4 
2.2 
5.4 
2.5 
6.3 
3.3 
9.1 
6.4 
6.8 
9.6 
3.3 
5.6 
36.5 
28.5 
9.7 
97e 


Wyoming

Rim,

paterculus 
63943 
gregarius. 
8.99 
5.56 
6.63 
9.95 
6.44 
7.46 
8.05 
7.53 
9.20 
9.45 
6.18 


H.AMNHFlagstaff5.0 
2.2 
5.8 
3.2 
9.1 
6.0 
6.5 
9.5 
4.1 
6.9 
33e 
25e 
10.1 
10.42 
14.88 
13.47 
15.57

Hesperocyon

V57498 
587

Dakotalength2.7-3.1.5-1.3.0-5.1.7-1.5.0-6.0 
2.3-2.8.0-10.0 
4.9-6.0 
5.7-7.8.0-9.2.8-4.0 
4.0-6.26.7-33.3 
20.2-25.5 
76-96 


S.
50351 
Fm., 
skullOR

gregarius 
and 
0.321 
0.144 
0.6799 
0.1 
0.369 
0.252 
0.578 
0.405 
0.517 
0.667 
0.508 
0.850 
2.87 
2.26 
5.46 


H.AMNHBrule4.2 
3.1 
1.7 
5.0 
1.9 
5.7 
2.5 
8.5 
5.2 
6.3 
8.6 
3.4 
5-1 
33.2 
25.3 
9.6 
91 
teethupper 
3.08 
1.6 
4.57 
1.8 
5.57 
2.53 
8.98 
5.43 
6.42 
8.85 
3.77 
5.46 
31.14 
23.88 
88.36 


92 
393 
55

10 
171217121612 
11

Dakota 
S 


X

S. 
statistics

Hesperocyon. 
gregarius 
63850 
forN 
AP 
AP 
AP 
length

Fm., 
AP 
AP 
AP

TTTTTT 


1 


ofH.AMNHBrule4.5 
3.3 
5.1 
6.0 
9.4 
7.0 
4.0 
32.7 
25.2 
10.6 
89 
p 
2p 
3p 
4p 
M2M 
P*-MpLpSkull

1

skulls2 
4

15.—Standard

and 
Table

(type)

dentitions

6e 
Oe

upper 
left 
3.8 
6.5 
2.5 
9.4 
6.0 
6.7 
9.5 
3.4 
5.3 
34.28.9.8

wilsoni 
40504-126

ofH. 
TMMright 
3.7 
3.3 
1.7 
5.0 
2.2 
6.4 
2.5 
9.3 
5.8 
7.0 
10.1 
3.4 
5.4 
34.5 
27.1 
9.7 
94 


AP 
AP 
AP 
AP 
AP 
AP 
AP 
length

TT 
TTTT 
2¦M 
4P 
2-M


1121

14.—Measurements 
of

1

able 
c1234 
M1--Skull

I 


ppppMPPM

54 



Table 
16.—Measurements 
of 
lower 
dentitions 
of 
Hesperocyon. 


H. 
wilsoni 
TMM 
40504-126 
(type) 
right 
left 
H. 
wilsoni 
H. 
gregariusTMM 
40504-99 
right 
left 
H. 
gregariusAMNH 
5297 
Logan 
Co., 
Colorado 
right 
left 
AMNH 
63850 
Brule 
Fm., 
S. 
Dakota 
P2 
AP 
T 
4.9 
1.9 
5.2 
2.4 
5.3e 
4.8 
1.7 
4.9 
P3 
AP 
T 
6.5 
2.4 
6.9 
2.9 
6.7 
5.5 
2.3 
5.5 
P4 
AP 
T 
7.2 
2.8 
7.8 
2.9 
7.6 
5.8 
2.6 
6.1 
2.8 
6.2 
M 
1 
AP 
T 
9.1 
4.7 
9.1 
4.8 
8.8e 
3.8e 
9.7 
4.3 
8.8 
M2 
AP 
4.7 
4.5 
4.7 
T 
2.9 
2.8 
M 
AP 
2.2 
2.3 
J 
T 
1.9 
PfM3P 
1-P4 
40.0 
24.4 
27.Oe 
26.6e 
18.8e 
35.8 
20.2 


Table 
17.—Standard 
statistics 
forlower 
teethofHesperocyongregarius. 


NX 
SORV 


M
M
M
M
P
P
P
P
2 
AP 
7 
4.96 
0.140 
4.8-5.2 
2.82 
T 
2 
1.75 
0.071 
1.7-1.8 
4.04 
3 
AP 
9 
5.51 
0.376 
5.0-6.1 
6.82 
T 
3 
2.13 
0.208 
1.9-2.3 
9.77 
P 
4 
AP 
12 
6.28 
0.414 
5.8-7.0 
6.59 
T 
5 
2.68 
0.277 
2.4-3.1 
10.35 
M 
1 
AP 
13 
9.51 
0.557 
8.6-10.3 
5.86 
T 
6 
4.27 
0.427 
3.8-4.8 
10.02 
2 
AP 
13 
4.93 
0.354 
4.4-5.6 
7.19 
T 
9 
3.21 
0.314 
2.8-3.8 
9.78 
3 
AP 
4 
2.10 
0.245 
1.8-2.3 
11.66 


T 
3 
1.70 
0.265 
1.5-2.0 
15.56 
p2-M3 
3 
35.93 
2.003 
34.0-38.0 
5.58 
P 
2-Pw' 
6 
20.38 
1.288 
18.8-21.3 
6.32

4 


55 



gregarius. 
Allowing 
for 
distortion, 
there 
is 
little 
difference 
between 
the 
endocasts 
of 
H. 
wilsoni 
and 
H. 
gregarius 
(Radinsky, 
1973). 


The 
various 
cranial 
foramina 
are 
of 
similar 
sizes 
and 
in 
the 
same 
positions 
as 
in 
H. 
gregarius, 
as 
well 
as 
can 
be 
seen 
in 
this 
specimen. 
The 
bullae 
are 
apparently 
a 
little 
larger 
than 
in 
H. 
gregarius, 
but 
are 
similar 
in 
form 
and 
are 
firmly 
attached 
to 
the 
skull. 


Only 
right 
I 
1 
and 
left 
I12 
are 
preserved. 
These 
are 
smaller 
than 
in 
H. 
gregarius, 
simple, 
and 
equal 
in 
size. 
I 
3 
was 
larger, 
as 
is 
shown 
by 
its 
root. 
The 
canine 
is 
slender 
like 
that 
of 
H. 
gregarius. 
There 
is 
a 
diastema 
of 
about 
4 
mm 
between 
the 
canine 
and 
P 
l, 
unlike 
H. 
gregarius 
in 
which 
this 
diastema 
is 
less 
than 
2 
mm. 
P 
1 
is 
identical 
to 
that 
of 
H. 
gregarius. 
P“ 
is 
likewise 
verysimilar 
to 
that 
of 
H. 
gregarius, 
a 
simple 
premolar 
without 
accessory 
cusps. 
P3 
is 
longer 
than 
in 
any 
specimen 
of 
H. 
gregarius 
available 
to 
me 
but 
is 
otherwise 
similar. 
In 
addition 
to 
the 
paracone, 
there 
is 
a 
metacone 
and 
a 
small 
metastyle. 
The 
protocone 
of 
P4 
is 
larger 
than 
is 
usual 
on 
this 
tooth 
in 
H. 
gregarius 
and 
the 
metastyle 
appears 
to 
be 
relatively 
slightly 
smaller. 
P4 
is 
about 
as 
close 
in 
appearance 
to 
that 
of 
Miacis 
gracilis 
(Uinta 
Formation,
Utah, 
late 
Eocene) 
as 
it 
is 
to 
Hesperocyon 
gregarius,
except 
that 
like 
H. 
gregarius 
it 
lacks 
a 
parastyle 
and 
labial 
cingulum. 
M 
1 
is 
very 
similar 
in 
H. 
wilsoni, 
H. 
gregarius, 
and 
M. 
gracilis 
in 
having 
a 
large 
parastyle 
and 
stylar 
cingulum, 
paracone 
and 
metacone 
about 
equal 
in 
size, 
small 
but 
distinct 
paraconule 
and 
metaconule, 
largeanteriorly 
placed 
protocone, 
and 
a 
hypocone 
developedfrom 
the 
posterior 
lingual 
cingulum. 
The 
size 
of 
M 
1 
of 


H. 
wilsoni 
is 
close 
to 
the 
top 
of 
the 
range 
in 
H. 
gregarius, 
the 
hypocone 
is 
relatively 
slightly 
smaller, 
and 
the 
protocone 
basin 
is 
larger 
than 
in 
H. 
gregarius. 
The 
relative 
sizes 
of 
the 
protocone 
basins 
and 
hypocone 
are 
closer 
to 
those 
of 
Miacis 
gracilis. 
The 
M“ 
is 
slightlysmaller 
than 
the 
mean 
for 
H. 
gregarius 
(Tables 
14-15),
and 
is 
close 
in 
size 
to 
that 
of 
M. 
gracilis. 
The 
of 
H. 
gregarius 
is 
highly 
variable 
in 
size 
(V= 
15.57) 
but 
fairlyconstant 
in 
the 
presence 
and 
extent 
of 
cusps 
and 
cingula.
Hesperocyon 
wilsoni 
differs 
from 
H. 
gregarius 
in 
the 
relative 
reduction 
of 
the 
metacone, 
which 
is 
always 
a 
distinct 
cusp 
in 
H. 
gregarius, 
usually 
only 
slightlysmaller 
than 
the 
paracone. 
In 
H. 
wilsoni 
the 
metacone 
is 
so 
small 
as 
to 
be 
almost 
nonexistent, 
while 
the 
paraconeremains 
large. 
In 
this 
feature 
it 
is 
similar 
to 
M. 
gracilis,
though 
that 
species 
retains 
a 
distinct 
but 
very 
small 
metacone. 
Hesperocyon 
paterculus 
(Matthew, 
1903)
differs 
from 
H. 
wilsoni 
in 
the 
relatively 
larger 
M 
2 
and 
presence 
of 
M 
3, 
which 
is 
also 
present 
in 
M. 
gracilis 
but 
absent 
in 
H. 
wilsoni 
and 
H. 
gregarius. 


None 
of 
the 
lower 
incisors 
or 
canines 
is 
preserved. 
In 
the 
type 
of 
Hesperocyon 
gregarius 
(Cope), 
AMNH 
5297, 
the 
premolars 
are 
tightly 
packed 
together 
with 
no 
diastemata. 
Only 
small 
spaces 
are 
present 
between 
premolars 
of 
other 
specimens 
of 
this 
species. 
However, 
in 
H. 
wilsoni 
(Fig. 
33-34) 
there 
are 
definite 
diastemata 


between 
the 
canine 
and 
P 
, 
between 
P 
and 
P 
, 
and 
between 
P., 
and 
P 
. 
As 
witn 
the 
skull, 
the 
mandible 
is 


more 
similar 
to 
M. 
gracilis 
than 
H. 
gregarius. 
The 
total 
length 
of 
the 
lower 
premolar 
series 
in 
H. 
wilsoni 
is 
considerably 
greater 
than 
in 
H. 
gregarius 
(Tables 
16-17). 
A 
Student’s 
“t” 
test 
(Simpson 
et 
al., 
1960, 
p. 
176)
performed 
on 
the 
length 
of 
P 
4, 
comparing 
H. 
wilsoni

J

(2 
individuals) 
to 
H. 
gregarius 
(6 
individuals), 
gave 
a 
probability 
that 
the 
two 
samples 
come 
from 
a 
singlepopulation 
of 
less 
than 
1 
per 
cent. 
Pj 
is 
poorly 
preservedbut 
probably 
was 
much 
like 
that 
of 
H. 
gregarius. 
P 
7 
lacks 
the 
small 
paraconid 
of 
this 
tooth 
in 
H. 
gregariusand 
had 
a 
smaller 
hypoconulid. 
This 
tooth 
is 
the 
same 
size 
in 
both 
species. 
The 
of 
H. 
wilsoni 
is 
larger 
than 
that 
of 
H. 
gregarius, 
as 
is 
P 
also. 
P3 
and 
P4 
differ 
from 
each 
other 
only 
in 
size. 
Both 
have 
a 
low, 
rounded 
paraconid, 
a 
tall 
protoconid, 
a 
distinct 
hypoconid, 
and 
a 
low 
hypoconid. 
M 
differs 
from 
the 
type 
of 
H. 
gregariusonly 
in 
having 
a 
taller, 
more 
conical 
protoconid,
protoconid-paraconid 
blade 
aligned 
less 
directlyanteroposteriorly, 
and 
a 
smaller 
entoconid. 
This 
tooth 
is 
approximately 
intermediate 
in 
morphology 
between 
M. 
gracilis 
and 
H. 
gregarius. 
M 
0 
is 
smaller 
than 
the 
average 
in 
H. 
gregarius 
and 
is 
near 
the 
lower 
end 
of 
the 
range 
in 
width 
(T). 
The 
protoconid 
is 
taller 
than 
the 
paraconid 
and 
the 
metaconid. 
The 
talonid 
consists 
of 
the 
talonid 
basin 
surrounded 
by 
a 
ridgelike 
hypoconid 
and 
a 
low 
lingual 
talonid 
crest. 
There 
is 
a 
small 
anterior 
labial 
cingulum. 
is 
a 
much 
reduced 
single 
rooted 
version 
of 
Mr 
The 
trigonid 
and 
talonid 
are 
discernible 
but 
tiny, 
and 
the 
only 
distinct 
cusp 
is 
the 
protoconid. 


Relationships. 
—Hesperocyon 
gregarius, 
(Cope,
1873), 
H. 
paterculus 
(Matthew, 
1903) 
and 
H. 
wilsoni 


n. 
sp. 
all 
have 
been 
reported 
from 
the 
Chadronian 
of 
North 
America. 
Hesperocyon 
paterculus 
appears 
to 
be 
a 
valid 
species, 
based 
on 
specimens 
from 
Pipestone 
Springsand 
Canyon 
Ferry 
Reservoir, 
Montana 
(White, 
1954), 
the 
Toadstool 
Park 
area 
of 
Nebraska, 
and 
Flagstaff 
Rim,
Wyoming. 
Among 
its 
distinguishing 
features 
are 
the 
relatively 
large 
size 
of 
M 
2 
and 
presence 
of 
a 
small 
M 
3. 
Hesperocyon 
gregarius 
is 
reported 
from 
the 
Peanut 
Peak 
member 
of 
the 
Chadron 
Formation, 
South 
Dakota 
(Clark
et 


al., 
1967) 
and 
from 
Cypress 
Hills, 
Saskatchewan 
(Russell, 
1972). 
The 
Uinta 
C 
species 
Miacis 
gracilis 
is 
structurally 
very 
similar 
to 
all 
three 
Chadronian 
species 
of 
Hesperocyon, 
particularly 
to 
H. 
wilsoni. 
Wilson 
et 
al. 


(1968) 
listed 
Miacis 
gracilis 
as 
part 
of 
the 
Airstrip 
local 
fauna 
This 
listing 
apparently 
was 
based 
on 
the 
typespecimen 
of 
H. 
wilsoni 
as 
a 
result 
of 
its 
considerable 


similarity 
to 
M. 
gracilis 
in 
premolar 
proportions 
and 
in 
the 
form 
of 
M“. 
Though 
lacking 
the 
attached 
ossified 
bulla 
of 
Hesperocyon, 
Miacis 
gracilis 
would 
have 
needed 
little 
alteration 
to 
be 
classified 
with 
the 
Canidae, 
and 
mayhave 
been 
ancestral 
to 
Hesperocyon. 
The 
dental 
changesneeded 
would 
have 
been 
minor; 
a 
reduction 
of 
premolarsin 
H. 
paterculus 
and 
H. 
gregarius, 
loss 
of 
the 
tiny 


M 
3in

H. 
gregarius 
and 
H. 
wilsoni, 
increase 
in 
size 
of 
M 
2 
in 
H. 
paterculus, 
and 
small 
changes 
in 
the 
orientation 
of 
the 
protoconid-paraconid 
shearing 
blade 
would 
have 
been 
involved. 
For 
the 
statistical 
summary 
of 
Hesperocyon 
gregarius 


56 



(Tables 
15, 
17) 
I 
have 
examined 
and 
measured 
the 
type 
and 
several 
referred 
specimens 
from 
various 
institutions 
and 
localities, 
all 
of 
which 
are 
from 
the 
Brule 
Formation 
or 
its 
equivalents 
in 
South 
Dakota, 
Wyoming, 
Colorado, 
and 
Nebraska. 
Hesperocyon 
lippincottianus 
is 
here 
treated 
as 
a 
synonym 
of 
H. 
gregarius. 
The 
coefficients 
of 
variation 
(V) 
are 
not 
particularly 
high 
despite 
the 
heterogeneity 
of 
locality. 
The 
AP 
measurements 
of 
lower 
teeth 
all 
have 
a 
V 
well 
within 
the 
limits 
shown 
by

Gingerich 
(1974, 
Text-Fig. 
1) 
to 
be 
normal 
for 
livingpopulations 
of 
a 
single 
species. 
Where 
possible, 
identifications 
were 
made 
or 
checked 
by 
comparison 
to 
the 
type. 
Measurements 
published 
by 
Scott 
and 
Jepsen(1936) 
were 
also 
included 
in 
the 
statistical 
summary. 


Family 
Felidae? 
or 
Nimravidae? 


genus 
and 
species 
indeterminate 


Little 
Egypt 
local 
fauna 


Figure 
35 


Material.—TMM 
40840-43, 
skull 
fragmentsincluding 
partial 
premaxillaries 
and 
maxillaries 
with 
left 
13,I3, 
right 
I 
1 
and 
roots 
ofboth 
canines. 


Stratigraphic 
position. 
—Upper 
part 
of 
Chambers 
Formation 
at 
Chalk 
Gap 
Draw, 
Vieja 
Group, 
Presidio 
County, 
Texas. 


Age.—Chadronian 
(early 
Oligocene). 


Figure 
35.—?Felid, 
genus 
and 
species 
indeterminate,TMM 
40840-43, 
Little 
Egypt 
local 
fauna. 
Crushed 
partial 
rostrum 
with 
incisors 
and 
broken 
canines. 
A. 
Anterior 
view. 
B. 
Left 
lateral 
view. 
C. 
Ventral 
view. 
Matrix 
and 
broken 
bone 
surfaces 
unshaded. 


57 



Description. 
—The 
bones 
are 
not 
individually 
crushed,
but 
have 
become 
separated, 
offset, 
and 
forced 
together.

The 
premaxillaries 
are 
peculiarly 
shaped. 
In 
lateral 
view, 
the 
narrowest 
preserved 
part 
is 
just 
posterior 
to 
I 
. 
The 
ascending 
portion 
is 
tall 
and 
thin, 
becoming 
broader 
dorsoventrally 
as 
it 
ascends. 
The 
portion 
forming 
the 
incisor 
aveoli 
projects 
strongly 
forward, 
but 
is 
narrow 
laterally. 
The 
form 
of 
the 
premaxillary 
symphysis 
is 
not 
obvious 
from 
the 
exposed 
bone 
surfaces. 
If 
the 
bones 
are 
separated 
and 
restored 
to 
a 
position 
in 
which 
the 
anteroposterior 
axes 
of 
the 
canines 
diverge 
slightlyposteriorly, 
the 
anterior 
surfaces 
of 
the 
premaxillaries 
seem 
to 
have 
formed 
an 
angle 
or 
pointed 
arch 
at 
the 
symphysis, 
quite 
unlike 
the 
flattened 
anterior 
surface 
seen 
in 
Recent 
cats. 
There 
is 
a 
deeply 
indented 
diastema 
between 
the 
I 
3 
and 
the 
canine, 
formed 
largely 
by 
the 
ventrolateral 
surface 
of 
the 
premaxillary. 
The 
premaxillary 
apparently 
did 
not 
form 
part 
of 
the 
canine 
alveolus. 


The 
maxillaries 
are 
remarkable 
mainly 
for 
the 
deeply 
concave 
palatal 
surface 
and 
the 
absence 
of 
alveoli 
exceptfor 
the 
laterally 
compressed 
canines. 
The 
raised 
ventral 
edge 
posterior 
to 
the 
canine 
shows 
no 
sign 
of 
anyalveolus 
for 
a 
space 
of 
at 
least 
15 
mm. 


12 
3

I 
and 
I 
are 
spatulate 
and 
smaller 
than 
I 
as 
is 
usual. 
However, 
seen 
from 
the 
front, 
I 
1 
are 
very 
,
broad 
at 
the 
occlusal 
surface, 
apparently 
meeting 
the 
lower 
incisors 
along 
a 
narrow 
flattened 
edge 
rather 
than 
interlocking 
as 
is 
usually 
the 
case 
in 
felids, 
especially 
those 
with 
saber-
like 
canines. 
I 
3 
is 
more 
caniniform 
but 
is 
not 
as 
large 
as 


is 
usual 
in 
felids. 
The 
canines 
were 
laterally 
flattened,
but 
are 
broken 
off 
near 
the 
base 
of 
the 
enamel. 


Relationships. 
—The 
obviously 
flattened 
and 
presumably 
saber-like 
canines 
immediately 
suggest 
some 
relationship 
to 
the 
Felidae 
or 
Nimravidae, 
particularly 
to 
the 
early 
forms 
such 
as 
Dinictis 
and 
Nimravus. 
The 
specimen 
is 
similar 
to 
Dinictis 
squalidens 
(Orellan, 
Brule 
Formation) 
in 
size, 
form 
of 
the 
canines, 
and 
to 
some 
extent 
in 
the 
rather 
spatulate 
incisors. 
It 
differs 
from 
Dinictis 
in 
the 
lack 
of 
anterior 
premolar 
aveoli, 
the 
small 
size 
of 
13,I3 
the 
shape 
of 
the 
ascending 
portion 
of 
the 


, 


1

premaxillary, 
and 
the 
detailed 
shape 
of 
I 
which 
in 
Dinictis 
generally 
are 
pointed. 
These 
differences 
between 
the 
Vieja 
specimen 
and 
Dinictis 
are 
exaggerated 
in 
nimravid 
genera, 
particularly 
in 
Hoplophoneus 
and 
Eusmilus. 
I 
suspect 
that 
this 
specimen 
represents 
either 
a 
new 
taxon 
or 
an 
unknown 
part 
of 
some 
previously 
known 
catlike 
animal. 
In 
view 
of 
this 
uncertainty, 
no 
definite 


taxonomic 
assignment 
is 
possible. 


CORRELATIONS 


Wood 
et 
al. 
(1941) 
recognized 
four 
“provincial 
ages” 
for 
the 
middle 
Eocene 
through 
early 
Oligocene 
mammalian 
faunas 
of 
North 
America, 
from 
earliest 
to 
latest; 
Bridgerian, 
Uintan, 
Duchesnean 
(all 
Eocene), 
and 
Chadronian 
(Oligocene). 
For 
this 
discussion, 
in 
order 
to 
remain 
consistent 
with 
Wilson’s 
later 
papers, 
I 
have 
arbitrarily 
used 
only 
three 
of 
these 
names 
(Bridgerian,
Uintan 
and 
Chadronian) 
as 
advocated 
by 
Wilson 
et 
al. 
(1968). 
Wilson 
(1977b), 
p. 
33-37) 
has 
given 
in 
detail 


58 


the 
reasons 
for 
avoiding 
the 
name 
“Duchesnean”*. 
For 
any 
of 
these 
names 
to 
be 
useful, 
they 
must 
be 
capable 
of 
definition. 
Biostratigraphic 
terminology 
using 
concurrent 
range 
zones 
should 
provide 
the 
best 
definitions 
and 
would 
allow 
for 
further 
subdivision 
of 
the 
“provincial 
ages”; 
this 
subject 
has 
been 
thoroughly 
discussed 
by 
Tedford 
(1970),
Emry 
(1973, 
1981) 
and 
Woodburne 
(1977). 
The 
definition 
of 
the 
Chadronian 
given 
by 
Wood 
et 
al. 
(1941) 
as 
“the 
time 
during 
which 
Mesohippus 
and 
titanotheres 
co-existed" 
describes 
a 
concurrent 
range 
in 
a 
broad 
sense 
and 
sets 
useful 
boundaries. 
Neither 
the 
Duchesnean 
nor 
the 
Uintan 
has 
been 
defined 
in 
such 
terms; 
rather, 
each 
was 
defined 
using 
an 
assemblage 
zone 
concept 
which 
cannot 
provide 
sharp 
boundaries. 
Since 
no 
singlebiostratigraphic 
event 
was 
used 
to 
mark 
each 
boundary, 
it 
has 
fallen 
upon 
later 
workers 
to 
choose 
suitable 
events 
as 
have 
seemed 
most 
appropriate. 
A 
definition 
of 
the 
base 
of 
the 
Uintan 
as 
the 
first 
appearance 
of 
Epihippus(presumably 
the 
time 
at 
which 
Orohippus 
evolved 
into 
Epihippus 
) 
would 
be 
useful 
in 
some 
areas 
and 
has 
much 
precedent. 
Likewise, 
the 
first 
appearance 
of 
Teleodus 
could 
be 
used 
as 
the 
event 
marking 
the 
base 
of 
the 
Duchesnean. 
As 
used 
by 
Wilson, 
the 
Uintan 
includes 
the 
entire 
range 
of 
the 
genus 
Epihippus. 
Of 
course, 
if 
a 
single 
taxon 
is 
to 
be 
used 
as 
a 
boundary 
marker, 
then 
local 
faunas 
lacking 
this 
taxon 
may 
be 
difficult 
to 
correlate. 
The 
lack 
of 
diagnostic 
equid 
specimens 
thus 
makes 
age 
assignments 
of 
such 
faunas 
as 
Whistler 
Squatin 
Texas 
and 
Lapoint 
in 
Utah 
controversial. 
Wilson 
(1971a, 
1974) 
placed 
the 
Lapoint 
fauna 
within 
the 
Chadronian, 
approximately 
equivalent 
to 
the 
Porvenir 
local 
fauna, 
on 
the 
basis 
of 
its 
artiodactyls. 
However, 
the 
presence 
of 
the 
mesonychid 
Hessolestes 
in 
the 
Lapointfauna 
is 
suggestive 
of 
the 
Uintan, 
as 
mesonychids 
have 
never 
been 
found 
(to 
my 
knowledge) 
in 
association 
with 
Mesohippus. 
For 
the 
present 
I 
consider 
Lapoint 
and 
similar 
faunas 
containing 
mesonychids 
to 
be 
of 
Uintan 
and 
Eocene 
age.

Correlations 
using 
first 
appearance 
of 
immigrant 
forms 
of 
carnivorous 
mammals 
may 
shed 
some 
light 
in 
this 
problem. 
Among 
the 
carnivorous 
mammals, 
two 
genera,
Hyaenodon 
and 
Hemipsalodon 
, 
appear 
to 
be 
immigrants.
Hyaenodon 
appears 
in 
the 
Bartonian 
of 
Europe 
(Mellett,

1977), 
probably 
earlier 
than 
its 
first 
appearance 
in 
North 
America, 
and 
Hemipsalodon 
may 
be 
derived 
from 
the 
Eurasian 
Pterodon. 
The 
one 
“Duchesnean” 
record 
of 
Pterodon 
in 
North 
America 
is 
in 
error 
according 
to 
Mellett 
(1977). 
Both 
of 
these 
genera 
appear 
to 
have 
replaced 
native 
forms, 
Hyaenodon 
supplantingLimnocyon 
and 
Oxyaenodon 
, 
while 
Hemipsalodonsupplanted 
the 
mesonychids. 
The 
first 
appearance 
of 
Hyaenodon 
does 
mark 
off 
a 
portion 
of 
the 
late 
Uintan 
as 
that 
term 
is 
used 
here 
which 
coincides 
with 
the 
Duchesnean 
as 
usually 
accepted. 
The 
first 
appearance 
of 
Hemipsalodon 
does 
apparently 
coincide 
with 
the 
beginning 
of 
Chadronian 
time, 
and 
this 
genus 
is 
restricted 
to 
the 
early 
Chadronian. 
Both 
of 
thee 
genera 
could 
easilybe 
placed 
into 
a 
system 
of 
concurrent 
range 
zones, 
and 
can 
provide 
secondary 
criteria 
for 
the 
definition 
of 
provincial 
age 
boundaries. 


However, 
see 
Wilson 
(1984) 
for 
a 
modification 
of 
this 
view. 
Now 
DuchesneodusLucas 
and 
Schoch 
(1982) 



Figure 
36.—Stratigraphic 
ranges 
of 
Uintan 
and 
Chadronian 
carnivore 
genera. 
Information 
taken 
largely 
from 
Black 
and 
Dawson 
(1966),
Clark 
et 
al. 
(1967), 
and 
original 
data. 
Correlations 
of 
faunas 
as 
given 
here 
are 
based 
on 
published 
radiometric 
dates 
where 
possible. 
In 
some 
cases, 
the 
first 
appearance 
of 
an 
apparent 
immigrant 
form 
is 
used, 
as 
in 
the 
correlation 
of 
Lapoint, 
Skyline 
Channels, 
and 
Pearson 
Ranch 
(Hyaenodon), 
and 
of 
Clamo, 
Cypress 
Hills, 
and 
Yoder 
(Hemipsalodon). 
Inclined 
lines 
between 
units 
in 
the 
three 
left-hand 
columns 
indicate 
uncertainty 
in 
placement 
of 
boundaries 
relative 
to 
the 
absolute 
time 
scale. 
The 
double 
wavy 
line 
below 
the 
Aheam 
memberoftheChadronFormationindicates 
anabruptgeographicalshift.Placementsofallfaunasandgenericrangeswithrespecttoeach 
other 
and 
to 
absolute 
dates 
are 
necessarily 
approximate 
and 
subject 
torevision. 


59 



Lower 
Chisos 
Formation.—The 
type 
locality 
for 
Uintacyon 
scotti 
is 
in 
the 
upper 
Uinta 
B 
at 
Kennedy’sHole, 
Utah. 
The 
presence 
of 
this 
species 
suggests 
an 
early 
to 
middle 
Uintan 
age 
for 
this 
part 
of 
the 
lower 
Chisos 
Formation 
Radiometric 
dates 
for 
the 
underlyingAlamo 
Creek 
Basalt 
vary 
widely, 
but 
the 
basalt 
is 
probably 
not 
older 
than 
the 
oldest 
date 
(44.3 
m.y.), 
well 
within 
the 
40-47 
M.Y. 
time 
span 
listed 
for 
the 
Uintan 
byMcKenna 
et 
al. 
(1973). 


Whistler 
Squat 
localfauna 
(Fig. 
36).—Wilson 
and 
Schiebout 
(1981) 
included 
all 
localities 
in 
the 
Agua 
Fria 
section 
from 
the 
basal 
Tertiary 
conglomerate 
up 
throughthe 
Whistler 
Squat 
Quarry 
in 
the 
Whistler 
Squat 
local 
fauna, 
which 
was 
first 
described 
by 
Wood 
(1973). 
The 
lowest 
part 
of 
the 
section 
yielded 
two 
specimens 
of 
?Proviverra 
major, 
which 
otherwise 
is 
known 
only 
from 
the 
Bridger 
Formation 
of 
Wyoming, 
and 
probably 
onlythe 
upper 
Bridger 
(Gazin, 
1976, 
p. 
10). 
A 
suggestedBridgerian 
age 
agrees 
with 
the 
determination 
made 
byWood 
(1973) 
on 
the 
basis 
of 
the 
rodents 
from 
the 
Whistler 
Squat 
local 
fauna. 
The 
presence 
of 
Leptoreodonmarshi 
in 
the 
lower 
15.2 
m 
(50 
feet) 
of 
the 
Agua 
Fria 
section 
(Wilson. 
1974) 
suggests 
a 
post-Bridgerian 
ageassignment, 
as 
the 
type 
material 
of 
this 
species 
comes 
from 
the 
Uinta 
Formation 
of 
Utah. 
The 
radiometric 
dates 
of 
48.6 
to 
42.9 
m.y. 
at 
Whistler 
Squat 
(Wilson, 
1980)
also 
are 
compatible 
with 
a 
Uintan 
age 
assignment. 


Serendipity 
local 
fauna 
(Fig. 
36).—Wilson 
and 
Schiebout 
(1981) 
considered 
the 
Serendipity 
local 
fauna 
to 
include 
material 
from 
several 
localities 
above 
a 
prominent 
marker 
bed 
in 
the 
Agua 
Fria 
area 
informallyknown 
as 
the 
Strawberry 
tuff 
(Fig. 
4). 
The 
Serendipity,
Unio 
Cliff, 
Titanothere 
Hill, 
and 
Purple 
Bench 
localities 
have 
furnished 
fragmentary 
material, 
but 
only 
Simidectes 
and 
Hessolestes 
are 
identified 
with 
any 
certainty. 
The 
Serendipity 
local 
fauna 
is 
believed 
to 
be 
Uintan 
althoughthe 
whole 
fauna, 
particulary 
the 
rodents, 
has 
yet 
to 
be 
studied. 
Other 
than 
this 
occurrence, 
Hessolestes 
is 
known 
only 
from 
one 
specimen 
from 
the 
Lapoint 
Member 
of 
the 
Duchesne 
River 
Formation, 
possibly 
latest 
Uintan, 
and 
Simidectes 
magnus 
(reported 
by 
Gustafson, 
1979) 
is 
known 
from 
both 
the 
Wagonhound 
and 
Myton 
Members 
of 
the 
Uinta 
Formation. 
Both 
Hessolestes 
and 
Simidectes 
magnus 
from 
the 
Serendipity 
local 
fauna 
suggest 
an 
age 
younger 
than 
Bridgerian. 
Other 
carnivore 
specimens 
are 
too 
fragmentary 
for 
use 
in 
correlation. 


Skyline 
channels 
(Fig. 
36). 
—An 
adapid 
primate(Mahgarita 
stevensi 
) 
was 
described 
from 
this 
local 
fauna 
by 
Wilson 
and 
Szalay 
(1976) 
and 
Amynodontipsis 
bodei 
by 
Wilson 
and 
Schiebout 
(1981). 
Hyaenodon 
cf. 
H. 
vetus 
and 
Simidectes 
magnus, 
have 
been 
reported 
from 
the 
Skyline 
channels 
(Gustafson, 
1979). 
Hyaenodon 
vetus 
and 
Amynodontopsis 
bodei 
were 
first 
described 
from 
the 
Pearson 
Ranch 
local 
fauna, 
Sespe 
Formation, 
California, 
considered 
by 
Stock 
(1948) 
to 
be 
of 
late 
Eocene 
age. 
Hyaenodon 
occurs 
there 
with 
Simidectes 
merriami 
Stock, 
in 
the 
only 
other 
known 
association 
of 
Simidectes 
and 
Hyaenodon. 
A 
Hyaenodon 
specimenreferred 
to 
Hyaenodon 
cf. 
H. 
vetus 
by 
Mellett, 
1977) 


which 
is 
very 
similar 
to 
the 
Texas 
Hyaenodon 
cf. 
H. 
vetus 
is 
known 
from 
the 
Lapoint 
Member 
of 
the 
Duchesne 
River 
Formation. 
The 
association 
of 
the 
two 
genera 
Hyaenodon 
and 
Simidectes 
suggests 
a 
late 
Eocene 
(late 
Uintan) 
age. 
The 
presence 
of 
Harpagolestes 
sp. 
in 
this 
stratigraphic 
position 
extends 
the 
range 
of 
the 
genusbut 
does 
not 
contradict 
this 
age 
assignment. 
Wilson 
and 
Schiebout 
(1981) 
report 
a 
K/Ar 
date 
of 
41.7± 
1.6 
m.y.
from 
a 
biotite 
bearing 
ash 
just 
below 
the 
Skylinechannels. 


Candelaria 
local 
fauna 
(Fig. 
36).—The 
largemesonychid 
Harpagolestes 
cf. 
H. 
uintensis 
is 
the 
onlycarnivore 
yet 
found 
in 
the 
Candelaria 
local 
fauna. 
The 
specimen 
suggests 
an 
early 
to 
middle 
Uintan 
age, 
but 
is 
not 
useful 
for 
more 
detailed 
correlation. 
Other 
taxa 
reported 
from 
this 
local 
fauna 
include 
Epihippus 
cf. 
E. 
gracilis 
(Forsten, 
1971a), 
several 
artiodactyls, 
includingLeptoreodon, 
Toromeryx, 
and 
Protoreodon 
(Wilson,
1971b, 
1974), 
a 
brontothere, 
Sthenodectes 
(Wilson,
1977a), 
and 
rodents 
(Wood, 
1974). 
Evidence 
from 
these 
taxa 
agrees 
with 
the 
conclusion 
of 
Wood 
that 
“the 
Colmena 
[Formation 
containing 
the 
Candelaria 
local 
fauna] 
is 
post-Myton 
[local 
fauna, 
Uinta 
Formation, 
Utah] 
in 
age, 
but 
probably 
not 
very 
much 
later.” 
Radiometric 
dates 
place 
this 
fauna 
between 
40± 
m.y.

and 
38.6 
m.y. 


Porvenir 
localfauna 
(Fig. 
36).—The 
Porvenir 
local 
fauna 
is 
the 
richest 
of 
the 
Vieja 
local 
faunas 
in 
number 
and 
variety 
of 
carnivores. 
Of 
the 
Porvenir 
taxa,
Ischnognathus 
savagei 
supplies 
no 
data 
for 
correlation. 
Miacis 
australis 
is 
not 
much 
more 
advanced 
than 
some 
Uintan 
miacids. 
Hemipsalodon 
viejaensis 
is 
more 
primitive 
than 
H. 
grandis 
from 
Cypress 
Hill, 
Saskatchewan, 
Yoder, 
Wyoming 
and 
Clarno, 
Oregon. 
It 
may 
be 
older 
as 
well 
as 
more 
primitive, 
or 
a 
geographicvariant. 
Hyaenodon 
crucians 
has 
been 
reported 
from 
Cypress 
Hills 
(Lambe, 
1908), 
and 
although 
Russell 
(1972) 
assigns 
these 
specimens 
to 
H. 
mustelinus, 
Mellett 
(1977) 
continues 
to 
recognize 
H. 
crucians 
from 
this 
locality. 
Mellett 
also 
recognized 
H. 
crucians 
from 
the 
early 
Chadronian 
of 
Wyoming. 
Most 
specimens 
of 
H. 
crucians 
are 
from 
late 
Chadronian 
or 
Orellan 
deposits.
Hyaenodon 
montanus 
was 
first 
reported 
from 
Toston,
Montana, 
and 
has 
since 
been 
found 
in 
the 
Aheam 
and 
Peanut 
Peak 
members 
of 
the 
Chadron 
Formation 
in 
South 
Dakota 
(Clark 
et 
al., 
1967). 
Mellett 
(1977) 
considers 
H. 
montanus 
to 
be 
a 
primarily 
middle 
Chadronian 
species.
Daphoenus 
lambei 
was 
first 
reported 
from 
Cypress 
Hills 
but 
has 
not 
been 
recognized 
elsewhere 
except 
in 
the 
Porvenir 
local 
fauna 
The 
unidentified 
daphoenine 
looks 
like 
a 
primitive 
Daphoenictis, 
the 
type 
species 
of 
which 
is 
from 
Cypress 
Hills, 
though 
this 
genus 
itself 
is 
not 
identifiable 
here. 
Other 
taxa 
in 
the 
Porvenir 
local 
fauna 
include 
Poabromylus 
kayi, 
the 
type 
of 
which 
is 
from 
the 
Lapoint 
Member 
and 
referred 
specimens 
of 
which 
have 
been 
found 
in 
early 
Chadronian 
localities 
in 
Wyomingand 
South 
Dakota, 
Oromeryx, 
Hidrosotherium, 
Hendryomeryx 
defordi, 
Protoreodon, 
Agriochoerus, 
merycoido


donts, 
a 
primitive 
Mesohippus 
(M. 
texanus), 
Haplohip


60 



pus, 
Brachyhyops 
wyomingensis 
(also 
known 
from 
Lapoint), 
Archeotherium, 
insectivores 
(Novacek, 
1976), 
brontotheres 
including 
Teleodus 
and 
Menodus 
(Wilson, 


1977), 
and 
rodents 
(Wood, 
1974). 


The 
data 
provided 
by 
the 
carnivores 
are 
consistent 
with 
Wood’s 
(1974) 
conclusion 
that 
the 
Porvenir 
local 
fauna 
may 
be 
the 
earliest 
known 
Chadronian 
fauna 
in 
North 
America. 
Radiometric 
dates 
place 
the 
age 
of 
the 
Porvenir 
local 
fauna 
between 
36.5 
and 
38.6 
m.y., 
older 


than 
the 
dates 
given 
for 
the 
Clarno 
local 
fauna 
(36.4 
m.y. 
or 
older) 
or 
the 
White 
River 
Formation 
at 
Flagstaff 
Rim,
Wyoming 
(Entry, 
1973; 
an 
oldest 
date 
of 
35.7 
m.y.). 


Little 
Egypt 
Local 
Fauna 
(Fig. 
36).—Of 
these 
taxa, 
the 
felid? 
rostrum 
gives 
no 
correlative 
data, 
and 
Miacis 
cognitus 
is 
an 
anachronism, 
a 
late 
surviving, 
relativelyunspecialized 
miacid. 
Daphoenocyon 
dodgei 
was 
first 
described 
from 
the 
Chadron 
Formation 
in 
Nebraska, 
and 
is 
common 
at 
Pipestone 
Springs, 
Montana. 


A 
correlation 
of 
the 
Little 
Egypt 
local 
fauna 
with 
Pipestone 
Springs 
is 
reasonable 
on 
the 
basis 
of 
Daphoenocyon. 
This 
agrees 
with 
information 
from 
insectivores 
(Novacek, 
1976), 
rodents 
(Woods, 
1974), 
and 
artiodactyls 
(Wilson, 
1971a). 
Time 
constraints 
placedby 
radiometric 
dates 
are 
the 
same 
as 
for 
the 
Porvenir 
local 
fauna, 
but 
as 
the 
Little 
Egypt 
local 
fauna 
is 
stratigraphically 
higher 
it 
is 
probably 
closer 
to 
the 
36.5 


m.y. 
figure. 
Airstrip 
Local 
Fauna 
(Fig. 
36).—The 
specimenreferred 
to 
Hyaenodon 
crucians 
gives 
no 
usable 
data. 
The 
three 
skulls 
of 
Hyaenodon 
raineyi 
n. 
sp. 
represent 
a 
species 
unknown 
elsewhere. 
Hesperocyon 
wilsoni 
n. 
sp. 
seems 
to 
have 
been 
derived 
from 
the 
Uintan 
Miacis 
gracilis, 
and 
its 
record 
here 
may 
be 
the 
earliest 
yet 
for 
Hesperocyon. 
It 
is, 
however, 
distinct 
from 
other 
Chadronian 
species. 
Radiometric 
dates 
limit 
the 
Airstriplocal 
fauna 
to 
between 
32.3 
and 
37.5 
m.y. 


SUMMARY 
AND 
CONCLUSIONS 


The 
carnivore 
fossils 
of 
the 
three 
west 
Texas 
sections 
are 
arranged 
in 
a 
stratigraphic 
sequence 
(Fig. 
36) 
similar 
to 
that 
recorded 
elsewhere 
in 
North 
America 
(Black 
and 
Dawson, 
1966; 
Clark 
et 
al., 
1967). 
In 
particular, 
the 
mesonychids 
disappear 
about 
at 
the 
end 
of 
the 
Uintan, 
the 
creodonts 
are 
reduced 
to 
three 
genera 
in 
the 
early 
Chadronian 
and 
to 
one 
in 
the 
late 
Chadronian, 
the 
true 


carnivores 
increase 
in 
numbers 
and 
diversity 
in 
the 
early 
Chadronian. 
the 
miacids 
(also 
Order 
Carnivora) 
dwindle 
though 
they 
last 
well 
into 
the 
early 
Chadronian, 
and 
finally 
extant 
families 
appear, 
though 
clearly 
not 
until 
well 
into 
the 
Chadronian. 


The 
quality 
of 
the 
carnivore 
record 
in 
west 
Texas 
is 
still 
relatively 
poor. 
There 
are 
approximately 
20 
genera 
of 
carnivores 
living 
in 
North 
America 
today, 
some 
of 
which 
include 
several 
species. 
In 
any 
one 
area 
of 
the 
continent, 
especially 
in 
middle 
latitudes, 
it 
is 
common 
to 
find 
from 
ten 
to 
twenty 
carnivore 
species, 
usually 
representing 
almost 
as 
many 
genera, 
present 
within 


historic 
times. 
A 
few 
fossil 
faunas, 
for 
instance 
that 
of 
the 
Bridger 
Formation 
(Bridgerian, 
middle 
Eocene) 
of 
Wyoming 
with 
about 
15 
genera 
and 
the 
Brule 
Formation 
(Orellan, 
middle 
Oligocene) 
of 
South 
Dakota 
with 
about 


17 
genera, 
appear 
to 
reflect 
a 
similar 
diversity 
of 
carnivore 
taxa. 
The 
total 
carnivore 
fauna 
of 
the 
Uinta 
Formation 
of 
Utah 
has 
slightly 
fewer 
taxa 
(from 
9 
to 
12 


genera, 
if 
Fig. 
36 
is 
accurate). 
The 
same 
is 
true 
of 
the 
Chadronian 
as 
a 
whole 
(about 
12 
genera). 
However, 
a 
minimum 
in 
recorded 
diversity 
occurs 
at 
about 
39 
to 
40 
m.y., 
near 
the 
Uintan-Chadronian 
boundary, 
at 
which 
time 
only 
five 
carnivore 
taxa 
are 
known 
or 
confidentlyinferred 
to 
have 
been 
present. 
All 
of 
the 
west 
Texas 
localities 
together 
have 
produced 
only 
12 
identifiable 
genera 
from 
a 
time 
span 
of 
about 
10 
million 
years. 
Despite 
the 
relative 
paucity 
of 
the 
fossil 
record, 
the 
stratigraphic 
sequence 
in 
Texas 
is 
relatively 
complete 
and 
can 
serve 
as 
a 
standard 
for 
comparison. 


The 
Uintan-Chadronian 
transition 
produced 
marked 
changes 
in 
the 
carnivorous 
mammal 
faunas 
in 
North 
America. 
It 
is 
becoming 
clear 
that 
these 
changes 
were 
complex. 
Few 
ancestor-descendant 
relationships 
have 
yetbeen 
documented. 
Replacement 
by 
unrelated 
and 
probably 
immigrant 
forms 
seems 
to 
have 
been 
common. 
There 
are 
still 
so 
few 
specimens 
known 
from 
this 
time 
span 
that 
any 
new 
finds 
are 
likely 
to 
produce 
majorchanges 
in 
our 
perception 
of 
taxonomic 
and 
phylogenetic 
patterns. 


Three 
genera 
of 
mesonychids, 
Mesonyx, 
Harpagolestes, 
and 
Hessolestes, 
are 
known 
from 
the 
Uintan. 
The 
latter 
two 
have 
been 
reported 
from 
Texas. 
Hessolestes 
is 
present 
in 
the 
Lapoint 
local 
fauna, 
stratigraphically 
above 
a 
radiometric 
date 
of 
39.3 
m.y. 
(McDowell 
et 
al., 
1973), 
and 
Harpagolestes 
is 
present 
in 
the 
Candelaria 
local 
fauna 
with 
dates 
of 
40 
m.y. 
below 
and 
38.6 
m.y.
above 
and 
in 
the 
Skyline 
local 
fauna. 
If, 
as 
I 
suspect, 
the 
40 
m.y. 
figure 
is 
a 
minimum 
age 
for 
the 
Candelaria 
local 
fauna, 
the 
Candelaria 
Harpagolestes 
may 
be 
somewhat 
older. 
Nevertheless, 
at 
all 
three 
places 
the 
mesonychid 
is 
the 
only 
large 
carnivorous 
mammal 
known 
in 
a 
small 
fauna. 
Thus 
the 
number 
of 
mesonychid 
genera 
remains 
at 
three 
from 
the 
Bridger 
through 
the 
Myton 
Member 
of 
the 
Uinta 
Formation, 
and 
is 
reduced 
to 
two 
(possibly 
one) 
in 
the 
poorly 
known 
faunas 
near 
the 
Uintan-Chadronian 


boundary. 


Six 
genera 
of 
Creodonta 
are 
known 
from 
the 
Bridger(Gazin, 
1976) 
but 
only 
three, 
all 
limnocyonids, 
survive 
well 
into 
the 
Uintan. 
Only 
?Proviverra 
major 
has 
been 
found 
in 
the 
Whistler 
Squat 
local 
fauna, 
apparently 
the 
latest 
occurrence 
of 
this 
genus 
in 
North 
America. 
A 
poorly 
preserved 
toothless 
skull 
from 
Unio 
Cliff, 
also 
part 
of 
the 
Serendipity 
local 
fauna, 
may 
be 
a 
limnocyonid, 
but 
comparisons 
sufficient 
to 
prove 
this 
have 
not 
been 
made. 
The 
evidence 
suggests 
that 
the 
reduction 
of 
native 
creodonts 
began 
early 
in 
the 
Uintan 
and 
was 
complete 
before 
the 
beginning 
of 
the 
Chadronian. 
Both 
Hyaenodon 
and 
Hemipsalodon 
, 
the 
common 
Oligocene 
creodonts, 
appear 
to 
be 
specializedEurasian 
immigrants. 
Hemipsalodon 
may 
have 
taken 
over 


61 



the 
large-carnivore 
niches 
vacated 
by 
the 
mesonychids. 
In 
turn, 
Hemipsalodon 
disappears 
with 
the 
advent 
of 
largespecies 
of 
Hyaenodon.

Among 
the 
Miacidae, 
5 
genera 
are 
listed 
from 
the 
Uintan 
by 
Black 
and 
Dawson 
(1966), 
a 
reduction 
of 
1 
genus 
from 
the 
Bridgerian. 
Many 
of 
these 
are 
small 
animals, 
generally 
poorly 
known, 
and 
the 
existence 
of 
small 
miacids 
in 
the 
Texas 
Uintan 
is 
only 
suggested 
by 
some 
unassignable 
fragments 
from 
the 
Whistler 
Squatlocal 
fauna. 
Uintacyon 
evidently 
was 
common 
in 
the 
early 
Uintan, 
and 
both 
large 
(17. 
scotti) 
and 
small 
{U.
acutus) 
species 
are 
present 
in 
the 
Uinta 
section. 
Uintacyon 
scotti 
is 
present 
in 
the 
Chisos 
section 
and 
a 
species 
of 
similar 
size 
is 
present 
in 
the 
Whistler 
Squatlocal 
fauna. 
The 
genus 
Tapocyon 
seems 
to 
have 
developed 
from 
Uintacyon 
and 
to 
have 
been 
widespread

(occurrences 
in 
California, 
Wyoming, 
and 
Utah) 
in 
partof 
the 
Uintan. 
The 
genus 
Miacis 
was 
also 
present, 
but 
except 
for 
the 
type 
of 
M. 
gracilis, 
a 
possible 
canid 
ancestor, 
the 
genus 
is 
represented 
by 
only 
very 
poor

specimens 
from 
the 
Uintan. 
The 
occurrence 
of 
Miacis 
cognitus 
in 
the 
Little 
Egypt 
local 
fauna 
is 
the 
latest 
record 
of 
a 
miacid 
from 
North 
America. 


Daphoenines 
certainly 
were 
present 
in 
the 
late 
Eocene. 
Daphoenus 
demilo, 
from 
the 
late 
Eocene 
of 
Badwater 
Creek, 
Wyoming, 
has 
been 
described 
by 
Dawson 
(1980).
Eosictis 
affinovi, 
a 
supposed 
late 
Eocene 
felid, 
bears 
a 
canine 
tooth 
almost 
identical 
to 
that 
of 
the 
daphoenine 


Daphoenocyon 
as 
well 
as 
similarly 
reduced 
and 
crowded 
premolars. 
This 
specimen 
may 
represent 
the 
earliest 
occurrence 
of 
Daphoenocyon. 
Three 
genera 
are 
presentin 
the 
Chadronian, 
of 
which 
Daphoenocyon 
is 
the 
most 
common 
(Hunt, 
1974b). 
Daphoenus 
occurs 
earlier 
(atleast 
in 
the 
Texas 
sections), 
and 
becomes 
the 
common 
genus 
in 
the 
Orellan, 
but 
is 
not 
well 
known 
in 
the 
Chadronian. 
A 
third 
genus, 
Daphoenictis 
(Hunt, 
1974b)
is 
distinct 
and 
specialized 
in 
the 
Chadronian. 


The 
earliest 
canid 
in 
the 
Texas 
sections 
is 
Hesperocyon 
wilsoni, 
in 
the 
Airstrip 
local 
fauna 
It 
was 
apparently 
a 
contemporary 
of 
H. 
paterculus. 
Hesperocyon 
gregarius 
has 
been 
reported 
from 
Cypress 
Hills, 
but 
is 
most 
common 
in 
the 
Orella. 
The 
differentiation 
of 
the 
Canidae 
very 
likely 
began 
in 
the 
Chadronian, 
but 
our 
knowledge 
of 
the 
species 
is 
incomplete. 


The 
origins 
of 
the 
Felidae 
and 
Nimravidae 
still 
are 
uncertain. 
The 
earliest 
good 
record 
of 
nimravids 
is 
in 
the 
Chadronian 
White 
River 
Formation 
at 
Flagstaff 
Rim, 
Wyoming 
(Emry, 
1973), 
but 
these 
specimens 
have 
not 
been 
studied. 
The 
fragmentary 
specimen 
from 
the 
Little 
Egypt 
local 
fauna 
may 
be 
earlier, 
but 
I 
am 
not 
certain 
what 
it 
is. 
Eosictis 
from 
the 
Brennan 
Basin 
Member 
of 


the 
Duchesne 
River 
Formation 
has 
a 
canine 
much 
like 
that 
of 
the 
amphicyonid 
Daphoenocyon 
and 
is 
probablynot 
a 
felid. 
Both 
felids 
and 
nimravids 
may 
be 
Eurasian 
immigrants, 
as 
were 
the 
Oligocene 
creodonts. 


62 



REFERENCES 


Black, 
C.C. 
and 
M.R. 
Dawson, 
1966. 
A 
review 
of 
late 
Eocene 
mammalian 
faunas 
from 
North 
America. 
Amer. 
J. 
Sci., 
264:321-349. 


Bowdich, 
T.E., 
1821. 
An 
analysis 
of 
the 
natural 
classifications 
of 
Mammalia 
for 
the 
use 
of 
students 
and 
travellers. 
Paris, 
J. 
Smith, 
115 
p. 


Clark, 
J., 
J. 
Beerbower 
and 
K. 
Kietzke, 
1967. 
Oligocene

sedimentation, 
stratigraphy, 
paleoecology 
and 
paleoclimatology 
in 
the 
Big 
Badlands 
of 
South 
Dakota. 
Fieldiana: 
Geol. 
Mem. 
5, 
158 
p. 


Cope, 
E.D., 
1872. 
Third 
account 
of 
new 
Vertebrata 
from 


the 
Bridger 
Eocene 
of 
Wyoming 
Territory.
Paleontol. 
Bull. 
3; 
reprinted 
in 
Proc. 
Amer. 
Phil. 
Soc., 
12:469-472 
(1873). 


_, 


1873. 
Third 
notice 
of 
extinct 
Vertebrata 
from 
the 
Tertiary 
of 
the 
Plains. 
Palaeontol. 
Bull.,
16:1-8. 


_, 


1875. 
On 
the 
supposed 
Carnivora 
of 
the 
Eocene 
of 
the 
Rocky 
Mountains. 
Paleontol. 
Bull., 
20:1-4. 
_,lBBO. 
On 
the 
genera 
of 
the 
Creodonta. 
Proc. 
Amer. 
Phil. 
Soc., 
19:76-82. 


1881. 
A 
new 
type 
of 
Perissodactyla. 
Amer. 
Nat., 
15:1017-1018. 


Davis, 
D., 
1964. 
The 
giant 
panda: 
A 
study 
of 
evolutionary 
mechanisms. 
Fieldiana: 
Zool. 
Mem. 
3, 
339 
p. 


Dawson, 
M.R., 
1980. 
Paleontology 
and 
Geology 
of 
the 
Badwater 
Creek 
area, 
central 
Wyoming: 
Part 
20, 
The 
late 
Eocene 
Creodonta 
and 
Carnivora. 
Ann. 
Carnegie 
Mus., 
49(4):79-91. 


DeFord, 
R.K., 
1958. 
Tertiary 
formations 
of 
Rim 
Rock 
Country, 
Presidio 
County, 
Trans-Pecos 
Texas. 
Texas 
J. 
Sci., 
10(1), 
37 
p. 


Douglass, 
E., 
1901. 
Fossil 
mammalia 
of 
the 
White 
River 
Beds 
of 
Montana. 
Trans. 
Am. 
Phil. 
Soc., 
20:237-279. 


Emry, 
R.J., 
1973. 
Stratigraphy 
and 
preliminary 
biostratigraphy 
of 
the 
Flagstaff 
Rim 
area, 
Natrona 
County, 
Wyoming. 
Smithson. 
Contrib. 
Paleobiol.,
18, 
43 
p. 


_, 


_, 


1981. 
Additions 
to 
the 
mammalian 
fauna 
of 
the 
type 
Duchesnean, 
with 
comments 
on 
the 
status 
of 
the 
Duchesnean 
“Age”. 
J. 
Paleontol., 
55(3):563


570. 
Evemden, 
J.F., 
D.E. 
Savage, 
G.H. 
Curtis 
and 
G.T. 
James, 
1964. 
Potassium-argon 
dates 
and 
the 
Cenozoic 
mammalian 
chronology 
of 
North 
America. 
Amer. 
J. 
Sci., 
262:145-198. 


Forsten, 
A., 
1971a. 
Epihippus 
from 
the 
Vieja 
Group,
Trans-Pecos 
Texas. 
Pearce-Sellards 
Ser., 
18:1-5. 


1971b. 
Comparisons 
of 
populations 
of 
Mesohippus 
from 
Trans-Pecos 
Texas 
and 
the 
Big 
Badlands, 
South 
Dakota. 
Pearce-Sellards 
Ser., 
18:12-15. 


_, 


Gazin, 
C.L., 
1976. 
Mammalian 
faunal 
zones 
of 
the 
Bridger 
middle 
Eocene. 
Smithson. 
Contrib. 
Paleobiol., 
26, 
25 
p. 


Gill, 
T., 
1872. 
Arrangement 
of 
the 
families 
of 
mammals 
with 
analytical 
tables. 
Smithson. 
Misc. 
Coll., 
11(1), 
98 
p. 


Gingerich, 
P.D., 
1974. 
Size 
variability 
of 
the 
teeth 
in 
living 
mammals 
and 
the 
diagnosis 
of 
closelyrelated 
sympatric 
fossil 
species. 
J. 
Paleontol., 
48(5);895-903.

Goldich, 
S.S. 
and 
A.A. 
Elms, 
1949. 
Stratigraphy 
and 
petrography 
of 
the 
Buck 
Hill 
Quadrangle, 
Texas. 
Bull. 
Geol. 
Soc. 
Amer., 
60:1133-1183. 


Gray, 
J.E., 
1821. 
On 
the 
natural 
arrangement 
of 
vertebrose 
animals. 
London 
Med. 
Reposit.,
15(1):296-310. 


Gregory, 
W.K. 
and 
M. 
Heilman, 
1939. 
On 
the 
evolution 
and 
major 
classification 
of 
the 
civets 
(Viverndae)
and 
allied 
fossil 
and 
Recent 
Carnivora: 
a 
phylogenetic 
study 
of 
the 
skull 
and 
dentition. 
Proc. 
Amer. 
Phil. 
Soc., 
81:309-392. 


Gustafson, 
E.P., 
1977. 
Carnivorous 
mammals 
of 
the 
late 
Eocene 
and 
early 
Oligocene 
of 
Trans-Pecos 
Texas. 
Dissertation, 
Univ. 
Texas 
at 
Austin, 
209 
p.

1979. 
Early 
Tertiary 
vertebrate 
faunas 
Big 
Bend 
area 
Trans-Pecos 
Texas; 
Simidectes 
(Mammalia,
Insectivora). 
Pearce-Sellards 
Ser., 
31, 
9 
p. 


Harris, 
J.M., 
1967. 
Toxotherium 
(Mammalia,
Rhinoceratoidea) 
from 
western 
Jeff 
Davis 
County,
Texas. 
Pearce-Sellards 
Ser., 
9, 
7 
p. 


Harris, 
J.M. 
and 
A.E. 
Wood, 
1969. 
A 
new 
genus 
of 
eomyid 
rodent 
from 
the 
Oligocene 
Ash 
Springlocal 
fauna 
of 
Trans-Pecos 
Texas. 
Pearce-Sellards 
Ser., 
14, 
7 
p.

Hough, 
J.R., 
1948. 
A 
systematic 
revision 
of 
Daphoenusand 
some 
allied 
genera. 
J. 
Paleontol., 
22:573-600. 


_, 


Hunt, 
R.M., 
Jr., 
1974a. 
The 
auditory 
bulla 
in 
Carnivora: 
an 
anatomical 
basis 
for 
reappraisal 
of 
carnivore 
evolution. 
J. 
Morph., 
143:21-76. 
1974b. 
Daphoenictis 
, 
a 
cat-like 
carnivore 
(Mammalia, 
Amphicyonidae) 
from 
the 
Oligoceneof 
North 
America. 
J. 
Paleontol., 
48:1030-1047. 


Laizer, 
L. 
de, 
and 
de 
Parieu, 
1838. 
Description 
et 
determination 
d’une 
machoire 
fossile, 
Hyaenodonleptorhynchus. 
C.R. 
Acad. 
Sci. 
Paris, 
7:442. 


Lambe, 
L.M., 
1908. 
The 
Vertebrata 
of 
the 
Oligocene 
of 
the 
Cypress 
Hills, 
Saskatchewan. 
Canad. 
Dept.
Mines, 
Geol. 
Surv. 
Branch, 
Contrib. 
Canad. 
Paleontol., 
3 
(Quarto), 
65 
p. 


Leidy, 
J., 
1853. 
Not 
titled. 
Proc. 
Acad. 
Nat. 
Sci. 
Phila.,
1853:392-394. 


_, 


1869. 
The 
extinct 
mammalian 
fauna 
of 
Dakota 
and 
Nebraska, 
including 
an 
account 
of 
some 
allied 
forms 
from 
other 
localities, 
together 
with 
a 


_, 


63 



synopsis 
of 
the 
mammalian 
remains 
of 
North 
America. 
J. 
Acad. 
Nat. 
Sci. 
Phila., 
ser. 
2, 
7:1


472. 
_, 


1872. 
Remarks 
on 
fossils 
from 
Wyoming. 
Proc. 
Acad. 
Nat. 
Sci. 
Phila., 
1872:276-277. 
Lucas, 
S.G., 
and 
R.M. 
Schoch, 
1982. 
Duchesneodusa

, 


new 
name 
for 
some 
titanotheres 
(Perissodactyla,
Brontotheriidae) 
from 
the 
late 
Eocene 
of 
Western 
North 
America. 
J. 
Paleontol., 
56(4): 
1018-1023. 


Maclntyre, 
G.T., 
1966. 
The 
Miacidae 
(Mammalia,
Carnivora) 
part 
I. 
The 
systematics 
of 
Ictidopappusand 
Protictis. 
Bull. 
Amer. 
Mus. 
Nat. 
Hist.,
131:115-230. 


Matthew, 
W.D., 
1903. 
The 
fauna 
of 
the 
Titanotherium 
beds, 
at 
Pipestone 
Springs, 
Montana. 
Bull. 
Amer. 
Mus. 
Nat. 
Hist., 
19:197-226. 


1909. 
The 
Carnivora 
and 
Insectivora 
of 
the 


Bridger 
Basin, 
Middle 
Eocene. 
Mem. 
Amer. 


Mus. 
Nat. 
Hist., 
9:291-567. 


Maxwell, 
R.A., 
1968. 
The 
Big 
Bend 
of 
the 
Rio 
Grande. 
Univ. 
Texas 
Bur. 
Econ. 
Geol. 
Guidebook 
7, 
138 
P-

Maxwell. 
R.A., 
J.T. 
Lonsdale, 
R.T. 
Hazzard 
and 
J.A. 
Wilson, 
1967. 
Geology 
of 
Big 
Bend 
National 
Park. 
Univ. 
Tex. 
Bur. 
Econ. 
Geol. 
Publ. 
6711,320 
p. 


McDowell, 
F.W., 
1979. 
Potassium-argon 
dating 
in 
the 
Trans-Pecos 
volcanic 
field, 
In 
Walton, 
A.W., 
and 


_, 


C.D. 
Henry, 
editors, 
Cenozoic 
geology 
of 
the 
Trans-Pecos 
volcanic 
field: 
Texas 
Bur. 
Econ. 
Geol. 
Guidebook 
19, 
10-18. 
McDowell, 
F.W., 
J.A. 
Wilson 
and 
John 
Clark, 
1973. 
K-Ar 
dates 
for 
biotite 
from 
two 
paleontologicallysignificant 
localities: 
Duchesne 
River 
Formation, 
Utah 
and 
Chadron 
Formation, 
South 
Dakota. 


Isochron/West 
7, 
11-12. 


McGrew, 
P.0., 
1953. 
A 
new 
and 
primitive 
EarlyOligocene 
horse 
from 
Trans-Pecos 
Texas. 
Fieldiana; 
Geol., 
10:167-171. 


1971. 
Mesohippus 
from 
the 
Vieja 
Group,
Trans-Pecos 
Texas. 
Pearce-Sellards 
Ser., 
18:6


_, 


11. 
McKenna, 
M.C., 
D.E. 
Russell, 
R.M. 
West, 
C.C. 
Black,
W.D. 
Turnbull, 
M.R. 
Dawson 
and 
J.A. 
Lillegraven, 
1973. 
K/Ar 
recalibration 
of 
Eocene 
North 
American 
land-mammal 
“ages” 
and 
European 
ages. 
Abstr. 
Prog., 
Geol. 
Soc. 
Amer. 
Ann. 
Meeting, 
Dallas, 
Texas. 


Mellett, 
J.S., 
1969. 
A 
skull 
of 
Hemipsalodon(Mammalia, 
Deltatheridia) 
from 
the 
Clamo 
Formation 
of 
Oregon. 
Amer. 
Mus. 
Novit., 
2387, 


19 
p. 


1977. 
Paleobiology 
of 
North 
American 
Hyaenodon 
(Mammalia, 
Creodonta).
Contributions 
to 
Vertebrate 
Evolution, 
vol. 
1, 
S. 
Karger, 
New 
York, 
N.Y., 
134 
p. 


Merriam, 
C.H., 
1918. 
Review 
of 
the 
grizzly 
and 
bigbrown 
bears 
of 
North 
America. 
North 
American 
Fauna, 
41, 
136 
p. 


_, 


Miller, 
M.E., 
G.C. 
Christensen 
and 
H.E. 
Evans, 
1964. 
Anatomy 
of 
the 
dog. 
W.B. 
Saunders 
Co., 
Philadelphia, 
941 
p. 


Moon, 
C.G., 
1953. 
Geology 
of 
Agua 
Fria 
Quadrangle,
Brewster 
County, 
Texas. 
Geol. 
Soc. 
Amer. 
Bull., 
61:151-196. 


Novacek, 
M.J., 
1976. 
Early 
Tertiary 
vertebrate 
faunas, 
Vieja 
Group, 
Trans-Pecos 
Texas: 
Insectivora. 
Pearce-Sellards 
Ser., 
23, 
17 
p. 


Osborn, 
H.F., 
1895. 
Fossil 
mammals 
of 
the 
Uinta 
Basin. 
Bull. 
Amer. 
Mus. 
Nat. 
Hist., 
7:71-105. 


1910. 
The 
Age 
of 
Mammals 
in 
Europe, 
Asia 
and 
North 
America. 
MacMillan 
Co., 
New 
York, 
635 
p. 


Peterson, 
0.A., 
1919. 
Report 
upon 
the 
material 
discovered 
in 
the 
upper 
Eocene 
of 
the 
Uinta 
Basin 
by 
Earl 
Douglass 
in 
the 
years 
1908-1909, 
and 
by 


_, 


O.A. 
Peterson 
in 
1912. 
Ann. 
Carnegie 
Mus., 
12:40-168. 
1931a. 
New 
mesonychids 
from 
the 
Uinta. 
Ann. 
Carnegie 
Mus., 
20:333-339. 


_, 


1931b. 
New 
species 
from 
the 
Oligocene 
of 
the 
Uinta. 
Ann. 
Carnegie 
Mus., 
21:61-78. 


_, 


Piveteau, 
J., 
1935. 
Etudes 
sur 
quelques 
creodontes 
des 
Phosphorites 
de 
Quercy. 
Ann. 
Paleontol., 
24:75


95. 
1962. 
L’encephale 
de 
Viverrarus 
angustidens, 
Miacide 
des 
phosphorites 
du 
Quercy. 
Ann. 
Paleontol., 
48:163-175. 


_, 


Radinsky, 
L.8., 
1969. 
Outlines 
of 
canid 
and 
felid 
brain 
evolution. 
Ann. 
New 
York 
Acad. 
Sci., 
167(1):277-288. 


1971. 
An 
example 
of 
parallelism 
in 
carnivore 
brain 
evolution. 
Evolution, 
25(3):518-522. 


_, 


1973. 
Evolution 
of 
the 
canid 
brain. 
Brain, 
Behavior 
and 
Evolution, 
7:169-202. 


_, 


Romer, 
A.S., 
1966. 
Vertebrate 
Paleontology. 
Univ. 
Chicago 
Press, 
Chicago, 
3rd 
Ed., 
468 
p.

1970. 
The 
vertebrate 
body. 
W.B. 
Saunders 
Co., 
Philadelphia, 
4th 
Ed., 
601 
p. 


Russell, 
L.S., 
1934. 
Revision 
of 
the 
lower 
Oligocenevertebrate 
fauna 
of 
the 
Cypress 
Hills 
Saskatchewan. 
Trans. 
Roy. 
Canad. 
Inst.,

20(I):49-67. 


_, 


_, 


1972. 
Tertiary 
mammals 
of 
Saskatchewan. 
Part 


II: 
The 
Oligocene 
fauna, 
non-ungulate 
orders. 
Life 
Sci. 
Contrib., 
Roy. 
Ont. 
Mus. 
84, 
97 
p. 
Rutimeyer, 
L., 
1862. 
Eocene 
Saugethiere 
aus 
dem 
Gebiet 
des 
schweizerischen 
Jura. 
Allg. 
Schweizerische 
Gesell., 
neue 
Denkschrifte, 
19:1-98. 


Savage, 
R.J.G., 
1965. 
Fossil 
mammals 
of 
Africa; 
19. 
The 
Miocene 
Carnivora 
of 
East 
Africa. 
Bull. 
British 
Mus. 
(Nat. 
Hist.): 
Geol., 
10(8):239-316. 


Schiebout, 
J.A., 
1974. 
Vertebrate 
Paleontology 
and 
Paleoecology 
of 
Paleocene 
Black 
Peaks 
Formation, 
Big 
Bend 
National 
Park, 
Texas. 
Texas 
Mem. 
Mus. 
Bull., 
24, 
88 
p. 


64 



Schlosser, 
M., 
1886. 
Über 
das 
Verhaltnis 
der 
Cope’schenCreodonta 
zu 
den 
übrigen 
Fleischfressem. 
Morph.
Jahrb., 
18:287-294. 


Schlaikjer, 
E.M., 
1935. 
Contributions 
to 
the 
stratigraphy 
and 
paleontology 
of 
th 
Goshen 
Hole 
Area. 
Wyoming. 
111. 
A 
new 
basal 
Oligocene 
formation. 
Bull. 
Mus. 
Comp. 
Zool. 
Harvard, 
76:71-93. 


Scott, 
W.8., 
1888. 
On 
some 
new 
and 
little 
known 
creodonts. 
J. 
Acad. 
Nat. 
Sci. 
Phila., 
9:155-185. 


1890. 
The 
dogs 
of 
the 
American 
Miocene. 
Princeton 
Coll. 
Bull., 
2:37-39. 


_, 


_, 


1898. 
Notes 
on 
the 
Canidae 
of 
the 
White 
River 
Oligocene. 
Trans. 
Amer. 
Phil. 
Soc., 
n.hs, 
19:325-415. 


1945. 
The 
mammalian 
fauna 
of 
the 
Duchesne 


River 
Oligocene. 
Trans. 
Amer. 
Phil. 
Soc., 
n.hs, 


34, 
pt. 
3:209-252. 


Scott, 
W.8., 
and 
Jepsen, 
1936. 
The 
mammalian 
fauna 
of 
the 
White 
River 
Oligocene. 
Part 
I. 
Insectivora 
and 
Carnivora. 
Trans. 
Amer. 
Phil. 
Soc., 
n.hs, 
28(1),
153 
p.

Scott, 
W.B. 
and 
H.F. 
Osbom, 
1889. 
The 
Mammalia 
of 
the 
Uinta 
Formation. 
Trans. 
Amer. 
Phil. 
Soc., 
16:461-572. 


Simpson, 
G.G., 
1931. 
A 
new 
classification 
of 
mammals. 
Bull. 
Amer. 
Mus. 
Nat. 
Hist., 
59:259-293. 


~ 


1945. 
The 
principles 
of 
classification 
and 
a 
classification 
of 
mammals. 
Amer. 
Mus. 
Nat. 
Hist. 
Bull., 
85, 
350 
p. 


Simpson, 
G.G., 
Ann 
Roe 
and 
R.C. 
Lewontin, 
1960. 
Quantitative 
Zoology, 
rev. 
ed., 
Harcourt, 
Brace 
and 
World, 
N.Y., 
Chicago, 
vii 
+ 
440 
p. 


Stevens, 
J.8., 
1969. 
Geology 
of 
the 
Castolon 
area, 
BigBend 
National 
Park, 
Brewster 
County, 
Texas. 
Dissertation, 
Univ. 
Texas. 
Austin, 
129 
p. 


Stevens, 
J.8., 
M.S. 
Stevens 
and 
J.A. 
Wilson, 
1975. 


Stratigraphy 
of 
Pruett 
and 
Duff 
Formations, 
Agua 
Fria 
and 
Tascotal 
Mesa 
Quadrangles, 
Brewster 
County, 
Texas. 
Abstr. 
Prog., 
South-Central 
Section, 
Geol. 
Soc. 
Amer., 
237. 


Stevens, 
J.8., 
M.S. 
Stevens 
and 
J.A. 
Wilson, 
1984. 
Devil’s 
Graveyard 
Formation 
(new), 
Eocene 
and 
Oligocene 
age, 
Trans-Pecos 
Texas. 
Tex. 
Mem. 
Mus. 
Bull. 
32:1-21. 


Stevens, 
M.S., 
J.B. 
Stevens 
and 
M.R. 
Dawson, 
1969. 
New 
early 
Miocene 
formation 
and 
vertebrate 
local 
fauna. 
Big 
Bend 
National 
Park, 
Brewster 
County, 
Texas. 
Pearce-Sellards 
Ser., 
15, 
53 
p.

Stock, 
C., 
1933a. 
Hyaenodontidae 
of 
the 
upper 
Eocene 
of 
California. 
Proc. 
Natl. 
Acad. 
Sci., 
19:434


_, 


440. 
1933b. 
A 
miacid 
from 
the 
Sespe 
upper 
Eocene,
California. 
Proc. 
Natl. 
Acad. 
Sci., 
19:481-486. 


_, 


1948. 
Pushing 
back 
the 
history 
of 
land 
mammals 
in 
western 
North 
America. 
Bull. 
Geol. 
Soc. 
Amer., 
59:327-332. 


Stovall, 
J.W., 
1949. 
Chadron 
vertebrate 
fossils 
from 
below 
the 
Rim 
Rock 
of 
Presidio 
County, 
Texas. 
Amer. 
J. 
Sci., 
246:78-95. 


_, 


Swanson, 
D.A. 
and 
P.T. 
Robinson, 
1968. 
Base 
of 
the 
John 
Day 
Formation 
in 
and 
near 
the 
Horse 
Heaven 
mining 
district, 
north-central 
Oregon.

U.S. 
Geol. 
Surv. 
Prof. 
Paper, 
600-D:DIS4-161. 
Szalay, 
F.S., 
1969a. 
Mixodectidae, 
Microsyopidae, 
and 
the 
Insectivore-Primate 
transition. 
Bull. 
Amer. 
Mus. 
Nat. 
Hist., 
140(4): 
195-330. 


_, 


1969b. 
Origin 
and 
evolution 
of 
function 
of 
the 
mesonychid 
condylarth 
feeding 
mechanism. 
Evolution. 
23:703-720. 


Szalay, 
F.S. 
and 
S.J. 
Gould, 
1966. 
Asiatic 
Mesonychidae 
(Mammalia, 
Condylarthra). 
Bull. 
Amer. 
Mus. 
Nat. 
Hist., 
132(2): 
127-174. 


Tedford, 
R.H., 
1970. 
Principles 
and 
practices 
of 
mammalian 
geochronology 
in 
North 
America. 
Proc. 
North 
American 
PaleontologicalConvention, 
1969:666-703. 


Thorpe, 
M.R., 
1923. 
Notes 
on 
the 
Bridger 
(Eocene)
Carnivora. 
Amer. 
J. 
Sci., 
ser. 
5, 
5:23-39. 


Trouessart, 
E.L., 
1885. 
Catalogue 
des 
mammiferes 
vivants 
et 
fossiles. 
Ordre 
des 
Carnivores. 
Bull. 
Soc. 
Etudes 
Sci. 
Angers, 
suppl. 
to 
v. 
14 
of 
year1884:1-108. 


van 
der 
Klaauw, 
C.J., 
1931. 
The 
auditory 
bulla 
in 
some 
fossil 
mammals. 
Bull. 
Amer. 
Mus. 
Nat. 
Hist., 
52:1-352. 


Van 
Valen, 
Leigh, 
1965. 
Some 
European 
Proviverrini 
(Mammalia, 
Deltatheridia). 
Paleontology, 
8:638


665. 
1966. 
Deltatheridia, 
a 
new 
order 
of 
mammals. 
Bull. 
Amer. 
Mus. 
Nat. 
Hist., 
132:1-143. 


_, 


1967. 
New 
Paleocene 
insectivores 
and 
insectivore 
classification. 
Bull. 
Amer. 
Mus. 
Nat. 
Hist., 
135(5):217-284. 


_, 


1974. 
Deltatheridium 
and 
marsupials. 
Nature, 
248:165-166. 


Walton, 
A.W., 
1972. 
Sedimentary 
petrology 
and 
zeolitic 
diagenesis 
of 
the 
Vieja 
Group 
(Eocene-
Oligocene), 
Presidio 
County, 
Texas. 
Dissertation, 
Univ. 
Texas, 
Austin, 
264 
p. 


_, 


1977. 
Petrology 
of 
volcanic 
sedimentary 
rocks, 
Vieja 
Group. 
Southern 
Rim 
Rock 
Country,
Trans-Pecos 
Texas. 
J. 
Sed. 
Pet., 
47(1);137-157. 


West, 
R.M., 
1982. 
Fossil 
mammals 
from 
the 
lower 
Buck 
Hill 
Group, 
Eocene 
of 
Trans-Pecos 
Texas: 
Marsupicamivora, 
Primates, 
Taeniodonta, 
Condylarthra, 
bunodont 
ArtiodactylU 
and 
Dinocerata. 
Pearce-Sellards 
Ser., 
35, 
20 
p. 


White, 
T.E., 
1954. 
Preliminary 
analysis 
of 
the 
fossil 
vertebrates 
of 
the 
Canyon 
Ferry 
Reservoir 
area. 
Proc. 
U.S. 
Natl. 
Mus., 
103(3326);395-438. 


Wilson, 
J.A., 
1965. 
Cenozoic 
history 
of 
the 
Big 
Bend 
area, 
west 
Texas. 
In 
Geology 
of 
the 
Big 
Bend 
area, 
Texas, 
Field 
trip 
guidebook 
with 
road 
logand 
papers 
on 
natural 
history 
of 
the 
area. 
West 
Texas 
Geol. 
Soc., 
65-51:34-36. 


_, 


1966. 
A 
new 
primate 
from 
the 
earliest 
Oligocene, 
West 
Texas, 
preliminary 
report. 
Folia 
Primatol., 
4:227-248. 


_, 


65 



1967. 
Early 
Tertiary 
mammals. 
In 
Maxwell,
R.A., 
J.T. 
Lonsdale, 
R.T. 
Hazzard 
and 
J.A. 
Wilson: 
Geology 
of 
Big 
Bend 
Nat. 
Park, 
Univ. 
Tex. 
Bur. 
Econ. 
Geol. 
Publ., 
6711:157-169. 


1971a. 
Early 
Tertiary 
vertebrate 
faunas, 
ViejaGroup, 
Trans-Pecos 
Texas; 
Entelodontidae. 
Pearce-Sellards 
Ser., 
17, 
17 
p. 


1971b. 
Early 
Tertiary 
vertebrate 
faunas, 
ViejaGroup, 
Trans-Pecos 
Texas: 
Agriochoeridae 
and 
Merycoidodontidae. 
Texas 
Mem. 
Mus. 
Bull., 
18,83 
p. 


_, 


_, 


1972. 
Vertebrate 
biostratigraphy 
of 
Trans-Pecos 
Texas 
and 
northern 
Mexico 
in 
the 
geologicframework 
of 
the 
Chihuahua 
tectonic 
belt. 
West 
Texas 
Geol, 
Soc., 
R.K. 
DeFord 
Symposium,

157-166. 


_, 


1974. 
Early 
Tertiary 
faunas, 
Vieja 
Group 
and 
Buck 
Hill 
Group, 
Trans-Pecos 
Texas; 
Protoceratidae, 
Camelidae. 
Hypertragulidae. 
Texas 
Mem. 
Mus. 
Bull., 
23, 
34 
p.

1977a. 
Early 
Tertiary 
vertebrate 
faunas, 
BigBend 
area 
Trans-Pecos 
Texas: 
Brontotheriidae. 
Pearce-Sellards 
Ser., 
25, 
17 
p.
1977b. 
Stratigraphic 
Occurrence 
and 


Correlation 
of 
Early 
Tertiary 
Vertebrate 
Faunas, 
Trans-Pecos 
Texas, 
Part 
1: 
Vieja 
Area. 
Texas 
Mem. 
Mus. 
Bull., 
25, 
42 
p. 


_, 


_, 


1980. 
Geochronology 
of 
the 
Trans-Pecos 
volcanic 
field. 
New 
Mexico 
Geol. 
Soc. 
Guidebook, 
31st 
Field 
Conf., 
Trans-Pecos 
Region, 
205-211. 


1984. 
Vertebrate 
faunas 
49 
to 
36 
million 
years 
ago 
and 
additions 
to 
the 
species 
of 
Leptoreodon(Mammalia, 
Artiodactyla) 
found 
in 
Texas. 
J. 
Vert. 
Paleontol., 
4(2); 
199-207. 


_, 


Wilson, 
J.A. 
and 
J.A. 
Schiebout, 
1981. 
Early 
Tertiaryvertebrate 
faunas, 
Trans-Pecos 
Texas; 
Amynodontidae. 
Pearce-Sellards 
Ser., 
33, 
62 
p.

1984. 
Early 
Tertiary 
vertebrate 
faunas, 
Trans-
Pecos 
Texas: 
Ceratomorpha 
less 
Amynodontidae.
Pearce-Sellards 
Ser., 
39, 
47 
p. 


Woodburne, 
M.0., 
1977. 
Definition 
and 
characterization 
in 
mammalian 
chronostratigraphy. 
J. 
Paleontol., 
51:220-234. 


Wortman, 
J.L., 
1901-1902. 
Studies 
ofEocene 
Mammalia 
in 
the 
Marsh 
collection. 
I. 
Carnivora. 
Amer. 
J. 
Sci., 
ser. 
4; 
11:333-348, 
437-450; 
12:143-154, 
196-206, 
281-296, 
377-382, 
421-432; 
13:3946,115-
128, 
197-206, 
433-448; 
14:17-23. 


Wortman, 
J.L. 
and 
W.D. 
Matthew, 
1899. 
The 
ancestryof 
certain 
members 
of 
the 
Canidae, 
the 
Viverridae 
and 
Procyonidae. 
Bull. 
Amer. 
Mus. 
Nat. 
Hist., 
12:109-139.