aes
%

Putt Fee:
Vike nae
Foe

wie!
ee.

—

ER 2 aes = GE > Cay = eo 2 WR OE
= SES NG y, - © YAN be io ee E
> CARRS ps on DEY Zz 7, m <on pe z uy m 2
\N INSTITUTION NOILALILSNI NVINOSHLINS S31uvadlt LIBRARIES SMITHSONIAN I
z o- ” ce ” z 2 2) z
z : = < = f= = <
= = z 4 z = Zz
)3 g : 8 Ud 2 g :
= g Ee 2 E = fs
= > > = +
5 3 a 2 o = =
IS S3IuVY q 17 LIBRARIES SMITHSONIAN _ INSTITUTION NOLLMLILSNI WING -
n 3 n 3S “” S 2)
ud n eo : <” Fe wn ul
— = — %, nF A = = 4 —
E 5 EPL > E 5 e
re 5 no % Oo se ro) =
4 2 Ee z = 2 x) ae
\N INSTITUTION NOILONLILSNI NVINOSHLINS S3I1YVYdIT LIBRARIES SMITHSONIAN |
te z ° cx z ; Gr
o 2 o = o ° Y, a
. 2 = » Y = x = iy,
i: D> = ca ASS = > = yj >
2 2 2 AS 5 2 pO ks 2
m 2 m SN 2 m g m
wn ma 7p) _ wn , wn
VS satuvual LIBRARIES SMITHSONIAN PREM LS NOILONLILSNI NVINOSHLIWS 5
2) z ". n z 24) wn
= < = < = =
= z = = = a
Be
B 2 g i Ny Z g
2 E Z E Av" 2 Z.
a : : Fue =
\N INSTITUTION NOLLALILSNI NVINOSHLIWS S3!1YVYEIT LIBRARIES SMITHSONIAN |

=

=

z

tO

7)

er

&

=

w
A w & uw a uw a
Ye =e = a = lead ye a
4; met o So
< S S wWWe S & e
3 = oS SPS 3 Ss )
2, 3 2 -I = 3 z
IS SAIYVNa@IT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS |
z cz = = z f& z
° ow = “oO ° w °

Ke

5 = >. = 5 cad 5
E — Hs _ Vs as ie
Z a z o z ay Zz
\N NOILNLILSNI NVINOSHLINS S3I1YVYEIT LIBRARIES SMITHSONIAN |
=a w a nw Fede n =
< = ES = 2 & = <
z = z = z NS 4 =
\ Oo De Oo [z Tay So SWS Be roy
} 2 oe f o Bd AN o 4
} = 2 S z = «2 S
= 3 ra = a ‘ue =
NS S3IuNVY q 17 LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS |
Ww we Ww Oo WwW a wi
=z ea ot | CU Ree A, + == + a
aie led 7 ~ [aad = a
< z 2p < z <
oc 4 ce Fj sj ee a =
= 35 Ey ae rs} = rs) 2
3 Fe a 2 ai 2 <3
AN INSTITUTION NOILNLILSNI NVINOSHLIWS S3a!uyvuatit LIBRARIES SMITHSONIAN _|
a G = S = = oa) Ae
o = ao = o = 1)
y, wD = na WS . 5 » 5 p>)
“i, > = > VSS ra > = ‘ >
2 F 2 WAY = : = 2
2 nS 2 a z a

]
)
)
)
J

Ys eT Pe SAS le TT eecsenereerseteam ¢

Sains

INSTITL

2.
e
4

; S
One Y

wae FT UY

NOILNLILSNI

SU POC VEIT ee

LIBRARIES

PENSE mE Pe ie I

NOILNLILSNI

ee D

LIBRARIES
‘

SS
NOILALILSNI

Se FP EN PU EN FE Re wh PEPE E SEEN TENE

LIBRARIES SMITHSONIAN

wae tw Vw ll

NOILNLILSNI

SU CPt wee tae FT

LIBRARIES

CUNT Em em EE

NOILNLILSNI

(DD ee ee

S3JIYVYSGIT_LIBRARIES SMITHSONIAN

SMITHSONIAN

SMITHSONIAN

We ‘Lp
ie

NVINOSHLIWS

\
S NS

SMITHSONIAN

NVINOSHLINS

NVINOSHLIWS

NOILNLILSNI

INSTITUTION

SMITHSONIAN

NVINOSHLIWS

S3IYVYUYEITLIBRARIES

NVINOSHLIWS

NVINOSHLIAS

NOILNLILSNI

INSTITUTION

NVINOSHLIWS s3luvyugiy LIBRARIES SMITHSONIAN

INSTITUTION NOILNLILSNI

NVINOSHLINS S3IYV

> a) — [2 Sec
4 3] [o 3} >
Kos Us 5 ¥ Is as
ne > Sass} m
— wn a
SAIMVESIT_LIBRARIES SMITHSONIAN
< See
io SN =
é & =
7) Him ee)
£ \ o
Lm x 2
F = ‘ >
wn =
INSTITUTION NOILNLILSNI NVINOSHLIWS
: z
= no
ow =
x +
ac S
a =
~ 1)
=, = <
SS1UVaGIT LIBRARIES SMITHSONIAN
= Y
a =)
> =
a Ee.
7)
D z
INSTITUTION NOILALILSNI
© ” z
= <
Ng q z
WY aie e ©
KAY 9 =
We
S3IYVYEGIT LIBRARIES SMITHSONIAN

INSTITUTION NOILNILILSNI

S3luYvVudIT LIBRARIES

LIBRARIES _SMITHSONIAN_INSTITUTION NOILNLILSNI

SSIYVUSIT

INSTITUTION NOILNLILSNI

SAIYVYaIT_LIBRARI

ES SMITHSONIAN

NVINOSHLINS S31YVUSIT_ LIBRARIES

2,

SMITHSONIAN

INSTITUTION NOILALILSNI

S3INVUGIT LIBRARIES SMITHSONIAN

NOILNLILSNI

2
°
tal
4
o
nus

4,
FINOSKA

NVINOSHLINS S31uV

=)
EE
K
72)
=

saluvaaii

NOILNLILSNI

INSTITUTION

NVINOSHLIWS
SMITHSONIAN

ASW

NO

LI

“~wsvea

INSTITUTION _ NO

NVINOSHLINS S3IYVYdIT_ LI

INSTITUTION NO

SSIYVYGIT LIBRARIES SMITHSONIAN
NOILNLILSNI

INSTITUTION

(op)

lw

a

<x

[aa

fee}

a
NVINOSHLINS S3/1uvHSIT LI

; = Se

fe) ©
- wm [sz
:
= a \

” m

2 2)
INSTITUTION. NO

eal w

=A

LA a

e o
2m
2%

S

ae
ret

INSTITUTION _ NO

Aa ASONS

q i! wy
2h” a He iy

nes

Peabody Museum of Natural History
Yale University
Bulletin 23

Systematics and Morphology

of American Mosasaurs

by
Dale A. Russell

—————_—

s THCS et
aM! $ HSON7, vy.
Hee 1g 1867

&
Cw es

a

PEABODY MUSEUM OF NATURAL HISTORY
YALE UNIVERSITY
BULLETIN 23

Systematics and Morphology
of American Mosasaurs

(Reptilia, Sauria)

BY
DALE A. RUSSELL

Curator of Fossil Vertebrates
National Museum of Canada

NEW HAVEN, CONNECTICUT
6 November 1967

Bulletins published by the Peabody Museum of Natural History at Yale Uni-
versity are based on research carried out under the auspices of the Museum. The
issues are numbered consecutively as independent monographs and appear at
irregular intervals. Shorter papers are published at frequent intervals in the Pea-
body Museum Postilla series.

PUBLICATIONS CommiTTEE: A. Lee McAlester, Chairman

Alfred W. Crompton, ex officio
Theodore Delevoryas
Willard D. Hartman
Elwyn L. Simons
Keith S. Thomson

Epitor: Jeanne E. Remington

Ass’t Epitor: Nancy A. Ahlstrom

Communications concerning purchase or exchange of publications should be
addressed to the Editor, Peabody Museum of Natural History, Yale University,
New Haven, Connecticut 06520, U.S.A.

Printed in the United States of America

Frontispiece: Restoration of Tylosaurus proriger, by Charles R. Knight under the
direction of Henry Fairfield Osborn (courtesy of the American Museum of
Natural History).

CONTENTS

WASTHOF ET USTRATIONS tis o-5 cr ave-e 16 hrs oe 5 oie cies ce oha cb aiocgeslin ay slate diay Mtoe, ane brere Vv
FABSTRAGTSy (ENGLISH, INUSSIAN, | ERENGE)) | vriccyejcis sei eittee sini cists isla feaieni <= I
INTER OD UW GTTO Niw astote statersl ost orale (oveie ahckeunetaiok halal eich Pe ae eRe Gels boas 4
ELISTOR Ve OES CLASSIFICATION) aeiaritiscle erode: cried oo os rere oie cldin aioe inecle s eioels 8
GENERAT, DESGRIPTION OF MOSASAUR SKULL 22.02 .cse00 ce eee tee ite tes 13
INAUY 271 SUTIN Cate pers coher nee Vaal ree ace RA acre UTS Colca own etay tee Sereveal ou acgl borers BRIS 14
JEW ae Velecll ly OT OT ES By Ce ee ih ee ee Sen Aa eh come Cra are eerie ee aera rap
Oaijoell. wine ose b ooo odooodos oedndots core dona oconccaccmen ad 29

Le Tayo OLCmUIINALS meperee ce Verte ots teua ty fora t sess tayoueaai te tetsto) on owe Ney ekstsrcs= eels) ee ieehe 41
IDjompoliernycorel’ NON, Seep eso doe souAueCRdeaneo CoCo poo moncocnorce 45
VAD WNT, Ga cogkasbdp oo ey eOUnnod nee Sane t ont ads HH arr h alee 45
(QyoRyeheerits WiThES einios 4 ea do ue Uo upon od Mead oian oon clam Scape 46

PATNEE TIONS OW ETB AWarctspe retriever Pastel eatatoreteys 2 aerate rete aera rain te) 49
ROSLETIOTOWEINI AWier write te ele eS tie one Noctis vale Warne euske chai etane ara 51
MiareimalerclemtitiOny scbeptons seca cre sce rersisveroicletors ses est nny sich ver Ste clan snatch lr oe:
FUNCTIONAL ANATOMY OF CRANIAL SKELETON............00c0eceeveees 58
SENSORY! LUNCHEONS Meractersaercrccis else © eistesueuecninia eels sis. oss a.d eersiNeerar 58
Cramiale circulation arrose sean cake eer oie Ghee tevettae oles lateraoiaoe See 59
GATT A GT TESTS seeeed s ereenoist rece noe teres oie Sasha iw cee ass om OG LSC Ow eaves 60

SUMED LOSUVL Vi eeocteretbeg Mere sarare Neste icles A share) sieht oraysvatNyaitey saya sisich coho! Saat a 63
Mani ci tila anOVEMIemes ments stations eel alereranetoree he atel sie seis cw sree ete = 64
GOMCIISIONS creche Perce tea ee Eola ay ale nae ideas 65
GENERAL DESCRIPTION OF POSTCRANIAL SKELETON ..........-.00eeeeeeeee 70
Gervicalevertebraclp er erianc nics srcisiics Sabha temas a ie Guars, sisecunes 70

WD OTSAl BVEKLE DIAC; srsrctsee RA tate ta tas fal oer otra NaTeR Sek nore reien ccc hetenc ete one nee 76
Gaudaleverte Dia eian ses srepeyaycte cterave cape htew or tale eos eae osscc eS RSI, Ses 78
Bectoralgappendave gerry tick teeters nc kr ie ef seiortrae cenit seas 82
Pelvics appendage’, sacdctys cic citegseapecsrspegn anate ts eld ays MeMnla atcha eee 97
FUNCTIONAL ANATOMY OF POSTCRANIAL SKELETON..........-ccecececeees 107
Functional anatomy of vertebral column ...................+: 107
Hunctional anatomy of forelimb, 222% 6.000. 02+. 000s. sence s os 111
Hunctionalvanatomy of hinGlimabiees se acs c1c1c eis.cc) 6) aleyte avs oie «rele e'eiei= 115
SWIMMING ANG MOSASAUTS oii gs.5 serseisisiesns vais oss s Sas sl esl « s.svertigte mere 119
SYSTENCATI CS epeteyctetevteicestoars tisicnae aaa tomar aa Nata. Fal atapini eshte ia Greie aisislevs nishel 12]
Subfamily SMiosasavirinae® t.,,2 aysterebewrersae kets isla stay sale tonanets oroke a egersksrs aha = 125
(QUGIDS TAS, artes clin RIANA ein Gi. o Or Oe RC Ree ee 124

IVEOSOS GLAS rs perc te Ah scene REE TOE Eo sta SCT oT a ais OS REE 131

AIM PRERCDILDIS Meera hist tN ete NN RTA I Pepe anee Grieve ere 141

TEL ODOT ae rae ROS Lee ET OMI Bete See Eons Sass We 142
GLO IL EN See TO ENTS ieee scree cise Sun abe otreie 144

PL OLOSAULT USE Ree a iB RO eee Se eon 145

HOKeI PIE PEN er ay OLs MI OSASAUITIMAC wererstteisy of stere che telepes Stel sterecieieieyers 147

Subfamilyeblioplatecanpimae ater. claie-steie) else ncisis) as-is sieierseeieieys ciate 148

Plabecanpuss <2 cis: ccirpts ted opel cites ieleias Mole hearin ore chee ates epee 148

EECLENOSQUTUS Mo eiaconhe ars Keres Ee Eee 156
PILOPIGLECOT DUS cot s-sierays: sieie ie wislore he ole ses nel rets siceife a ietol oe eke eRe 158
PLOGNALNOGON Vase x zjera close 50x40 es spo )ere a ermine ee eiers at sh oer TSE 162
PlestOtylosaurus, (accra cieisteve.tlolacs-1toke elaine ei oe teen eee ee 167

FV ALUSQUY US. wera ee eee Ree ee ee REE eee 168
Roreion’Plioplatecarpinaelrrr. iy. ec tee 169
Subfamily Wylosaurinae Mass fe crete settee. Toererchsetrertele tetera 170
USUOXWEOS “GB aqecnonodeGuoobgetoogoncoog nano bOdKDHRaOGOCKbOs 170

oats Ah Woreniguelsneoanaaosgancdoooanusccanneocdd0Goddso- 176
Mosasaurs of uncertain taxonomic standing ................... 176
STRATIGRAPHIC AND PALEOECOLOGIC CONSIDERATIONS .........eeeeeeeeees 185
Gulfi@oast iGretaceous=s.ss eee cee eee eee Eee 185

New sijemsey., GretaccOusy ye myeter lets re telete foleretolefefetel-fetsieher cierto etree 186
Niobrara) Formation 3s oi arcs ie scvetoreietens eheteneters ote tare fevolel oe otner eee 186
Pierre: FOrmMatiOm® 4x) 5f1. le tele [ste levovateteg eatols ele heveteleteraveaiest, Glekele este 187
Additional occurrences of American mosasaur.................-- 189
Foreign Occurrences: Of -mOSaSaUTs) cite ieterel olete << roel eater oer 190
Ecological conclusions: ye %ccjsiarele teresoresefotelotniel sj-toisloter> lett ee 19]
EVOLUTIONARY (GONSIDERATIONS \acccnes cites sulin ete ciee ee Gees 193
Mosasaurs compared with platynotan lizards .................... 193
Phylogeny» of *platynotams ete eireterene eleterre cleus ei aes elke eee 199
Phylogeny (Of mosasauss eres eietecieeieielermietete rere crete evel lei) Sere 202
Evolution of mosasaursic. see e oe es eee eieic ae ee eee 204
APPENDIX A—IMEASUREMENTS) ci... 500s 2s ee cis le os sinc trosinieereieie ei eee 208
APPENDIX) B—lLOGCALITIES, OF (PIERRE MOSASAURS! ~~... 245.0% +2 sale cele eee 210
PETTERATUREMGITED oissce ons.a cede susie ecsieislel ous aicis.ciishare. oveteisienafets te rete T et reeatare 212
1 S10) o> Ce A SA Rn ee ee ae ee te nS OSA 000 a 06 225
GHAR TS 7 oi ecctctare Biche to Sie etetore oe EE Le ERR Va ee eile Otero ORO eee 231

PRATES I) 2 Glee cdc o ews bee ee en tad eae Oe eee eee 239

ILLUSTRATIONS

TEXT-FIGURES
1. Diagram of the functional units of a mosasaur skull
2. Mosasaur premaxillae
3. Frontal of Mosasaurus maximus
4. Undersurface of frontals
5. Prefrontals
6. Right palatine of Platecarpus ictericus
7. Orbitosphenoid of Platecarpus
8. Parietal of Mosasaurus maximus
9. Basioccipital-basisphenoid of Platecarpus, lateral view
10. Basioccipital-basisphenoid of Platecarpus, dorsal view
11. Basicranial circulation of Platecarpus
12. Braincase of Clidastes propython, lateral view
13. Braincase of Clidastes propython, medial view
14. Periotic labyrinth of Clidastes propython
15. Prootic of Tylosaurus nepaeolicus
16. Endocranial cast of Platecarpus
17. Occipital view of skull of Platecarpus ictericus
18. Occipital muscle insertions in Platecarpus
19. Supraoccipital of Platecarpus
20. Quadratic suspensorium of Platecarpus
21. Pterygoid of Tylosaurus proriger
22. Pterygoids
23. Stapes and epipterygoid of Platecarpus ictericus
24. Quadrates
25. Quadrate of Platecarpus ictericus
26. Anterior mandibular unit of Clidastes liodontus
27. Anterior tip of dentary of Tylosaurus proriger
28. Splenio-angular articulation of Platecarpus
29. Mandible of Platecarpus ictericus
30. Marginal tooth of Mosasaurus maximus
31. Reconstructed cross section of skull of Platecarpus
32. Reconstructed temporal region of skull of Clidastes liodontus
33. Restored superficial mandibular musculature of Clidastes
34. Restored deep mandibular musculature of Clidastes
35. Kinesis in mosasaurs
36. Streptostyly in mosasaurs
37. Restored skull of Platecarpus ictericus with jaws opened and protracted
38. Restored skull of Platecarpus ictericus with jaws closed and retracted
89. Atlas-axis vertebrae of Clidastes propython
40. Anterior portion of vertebral column of Platecarpus
41. Areas of muscle attachment on anterior portion of vertebral column of
Platecarpus
42. Vertebra of Clidastes showing zygosphene-zygantrum
43. Scapula-coracoid of Clidastes liodontus
44. Scapula-coracoid of Platecarpus
45. Scapula-coracoid of Tylosaurus

ILLUSTRATIONS

. Interclavicles

. Humeri of Plotosaurus and Mosasaurus, flexor aspect
. Humeri of Tylosaurus and Platecarpus, extensor aspect
. Humeri of Mosasaurus and Clidastes, extensor aspect
. Forelimb of Clidastes

. Forelimb of Mosasaurus conodon

. Forelimb of Plotosaurus tuckeri

. Forelimb of Platecarpus

. Forelimb of Ectenosaurus clidastoides

. Forelimb of Tylosaurus proriger

. Pelvic girdle of Mosasaurus conodon

. Pelvic girdle of Platecarpus

. Pelvic girdle of Tylosaurus

. Femur of Mosasaurus hoffmanni

. Hind limb of Clidastes liodontus

. Hind limb of Mosasaurus conodon

. Hind limb of Platecarpus

. Hind limb of Tylosaurus proriger

. Cross section of the trunk of a mosasaur in the pelvic and caudal region
. Scapula-coracoid of Clidastes showing muscle attachments
. Extensor musculature of forelimb of Clidastes

. Flexor musculature of forelimb of Clidastes

. Pelvis of Clidastes showing supporting musculature

. Pelvis of Clidastes showing appendicular musculature
. Hind limb of Clidastes showing arrangement of appendicular muscles
. Muzzle of Clidastes sternbergi

- Dorsal view of skull of Clidastes liodontus

. Muzzle of Clidastes liodontus

. Muzzle of Clidastes propython

. Clidastes parietals

. Clidastes quadrates

. Dentary of Mosasaurus conodon

. Dentary tooth of Mosasaurus conodon

. Muzzle of Mosasaurus tvoensis

. Quadrate of Mosasaurus maximus

. Pelvis girdle of Amphekepubis johnsoni

. Marginal teeth of Liodon sectorius

. Dorsal view of skull of Platecarpus ictericus

. Ventral view of skull of Platecarpus ictericus

. Platecarpus dentaries

. Dorsal view of skull of Ectenosaurus clidastoides

. Quadrate of Plioplatecarpus primaevus

. Mandibular elements of Plioplatecarpus depressus

. Dorsal view of skull of Prognathodon overtont

. Lateral view of skull of Prognathodon overtont

. Quadrate of Prognathodon rapax

. Dorsal view of skull of Tylosaurus proriger

. Tylosaurus parietals

. Tylosaurus quadrates

. Lateral view of skull of Tylosaurus nepaeolicus

. Mosasaur skulls from the Niobrara Chalk

. Phylogeny of platynotans

ILLUSTRATIONS vil

98. Restored skulls of Clidastes liodontus, Opetiosaurus bucchichi and hypotheti-
cal ancestral platynotan
99. Diagrammatic family tree of the mosasaurs

CHARTS

1. Western gulf mosasaurs

. Eastern gulf mosasaurs

. New Jersey mosasaurs

. Mosasaurs of the interior

. Belgian mosasaurs

. French mosasaurs

. English and Swedish mosasaurs

“ID Ot P 9 DO

PLATES

Frontis piece—Life restoration of Tylosaurus proriger

1, Restorations of Clidastes liodontus, Platecarpus ictericus, and Tylosaurus
proriger

2. Restorations of Tylosaurus proriger and Plotosaurus

—

YALE UNIVERSITY PEABODY MUSEUM OF NATURAL HISTORY BULLETIN
No. 23, 240 P., FRONTISPIECE, 2PLS., 99 TEXT-FIGS., 1967

SYSTEMATICS AND MORPHOLOGY
OF AMERICAN MOSASAURS

(Reptilia, Sauria)

By DALE A. RUSSELL

ABSTRACT

Mosasaurs were large, marine platynotan lizards which became abundant
and diversified during the latter half of Cretaceous time, but disappeared at
the close of the period. Three subfamilies are recognized within the Family
Mosasauridae: the Mosasaurinae (including the new tribes, Mosasaurini, Globid-
ensini and Plotosaurini), the Plioplatecarpinae (including the new _ tribes,
Plioplatecarpini and Prognathodontini), and the Tylosaurinae. At present
thirteen genera and thirty species are diagnosed from the North American
Cretaceous, with one new genus (Ectenosaurus) and one new species (Plioplate-
carpus primaevus). Of the older names available, nineteen generic and twenty-
nine specific names have been placed into synonymy, two generic and eighteen
specific names are based on indeterminate material, and an additional generic
and six specific names are only of dubious validity.

The osteology of mosasaurs is described in detail and with reference to the
soft anatomy. Mosasaurs possessed a good sense of sight and a poor sense of
smell. A calcified tympanum, present in all three subfamilies, was probably
useful in transmitting waterborne sound to the middle ear and is not indicative
of deep-diving habits. Streptostylic quadrates permitted anteroposterior move-
ment of the mandibles, which in turn facilitated the underwater swallowing of
prey. Mosasaurs swam by lateral undulations of the body, the flippers and rela-
tively long neck serving as organs of equilibration. They fed on smaller mosa-
saurs, chelonians, fish, ammonites, belemnites, echinoderms and pelecypods, and
for the most part were highly active aquatic carnivores.

Mosasaurs inhabited subtropical epicontinental seas of less than 100 fathoms
(about 180 meters) depth and of variable salinity. Individual forms had a wide
geographic distribution that was little affected by changes in depositional en-
vironments. The range of many species of mosasaurs inhabiting the western edge
of the North Atlantic basin probably extended into western European waters.
Those from the eastern edge of the Pacific basin appear to have belonged to a
distinct zoogeographic province.

Mosasaurs descended from primitive middle Cretaceous varanoids (aigialo-
saurs) possessing many cranial characteristics of mosasaurs but with a_post-
cranial morphology similar to that of the modern Varanus. Ancestral mosasaurs
seem to have been of two basic types; forms with long bodies and short dilated
tails giving rise to the mosasaurines, and forms with short bodies and long
pointed tails giving rise to the plioplatecarpines and tylosaurines. During their
relatively brief geologic existence mosasaurs exhibited a few clearly progressive
trends, such as a tendency to increase in overall size, to telescope the frontals
over the anterior edge of the parietals (consequently suppressing kinesis), to in-
crease the number of pygal vertebrae, and to alter the primitive webbed
paddles into long, hyperphalangic flippers.

1

2. PEABODY MUSEUM BULLETIN 23

AMEPHKAHCKVHE MO3A3ABPBHI *

W93J1b PYCCEJIb
KYPATOP OTJ[EJIA MCKOMAEMbIX MO3BOHOUHBIX,

HAUMOHAJIbHbIM MY3EM KAHAJIbI

ABCTPAKT

Mosa3asppt Obl KpymHbIe MOpcKuve MmMIocKO3yOne aAuepuuE. Onn u306ur0BanHK uM ObIIM
pasHooOpasHbIMM BO BTOPO!t NOOBHHE MeJa, HO Mcue3IM B KOHUWe ero. B cemeiicrBe MO-
3a3aBpOB pa3IMyaloT TPH MOACeMelCcTBa:

MO3Q3QyPHHbI (BKIIOYaT HOBbIe TpHObI MO3a3aypUHH, TIOONAeHCHHU MU MWOTOSaypuHN),
navonAaTeKapnuubl (BKMIOYaA HOBbIe TPuObI WIMOMAAaTeKapNuHH WM MpOrHaTOAOHTHHN),
THJO3Q)pHHbI.

B nactosuree Bpema B Mente CepepHoit AMepuku ompexereHb! TPHHAasWaTb POAOB U TPUA-
WaTb BHLOB, OMMH pO (9KTeHOSaBpb!) HM OAMH BUA (MAMONAaTeKapnyc mpuMaeByc) sABAA-
}OTCA HOBLIMH. M3 yucda UMeIOWINXCA CTApbIX BHA3BaHHil AEBATHAAUMATb POAOBbIX VM ABakWaTb
A€BATh BULOBbIX ObIIM MOMELICHbI B CHHOHHMUKY, ABA POAOBbIX HM BOCe€MHALMATb BUAOBbIX
Ha3BaHHit OCHOBaHbI Ha HEACHOM MaTepHae, OJHO POAOBOe Ha3sBaHHe HM WeCTb BHAOBbIX
BbISBIBAIOT COMHEHHA B CMBbIC€ HX MpHrOHOCTH.

Jlaerca eTanbHOe OmMcaHve OCTeOJOrHH MO3a3aBPOB CO CCbINKAaMM Ha aHaTOMMIO HX
tena. Mosa3apppt oOnaqaM XOpOWO pasBUTbIM 3pe€HHEM MH M0XO pa3sBuTBIM OOOHAHHEM.
Oxamenesuiaa GapabaHHasd MONOCTb, HMeIOlLaACA y BCeX Tpex NOAcemelicTB, NMOBHJHMOMY
Oblia NoNe3Ha AA Mepexaun B CpeAHee yxO TeX 3BYKOB, KOTOPble pacnpOcTpaHAIHCb B Bose,
OHA He YKa3bIBaeT Ha TO, YTO 2KHBOTHOe MOFIO HbIPATh rayOOKo.

CrpentocTHabHble KpakpaTbl NO3BOAAIM LBYIKEHHe BepXHMX YeMOCTeH Biepex HM Hasan,
4TO B CBOW OYepedb OOMer4YaNO MporsaTHIBaHHiO AOOw4u oA Bo,oW. Mosa3aBpbl NaBalM
MOCpeACTBOM BOJHOOOpASHbIX JABHXKEHHM Tela B JJIMHY, MWaBHMKH M OTHOCHTebHO JWIMHHad
wed CIYKWIH Wa paBHoBecus. IIutanucb OHM GOee MENKHMH MO3a3aBpaMu, YepenaxaMH,
pbiOol, AMMOHUTaMH, GeneMHHTAMH, HrAOKOMKHMH, NeeUHNOAMH; OHH Ob OYeHb AKTHB-
HbIMM BOJHBIMH MIOTOAHbIMM 2KHBOTHBIMH,

On Hacetanu cyOTpomuyeckHe SMMKOHTHHEHTAMbHBIe MOPA TayOuHOIO MeHee 4eM CTO
MOPCKHX CaxKeHeH (OKONO CTa BOCbMHAeCATH METPOB) HM HMeIOWIMe pa3siM4HoOe cOsepKanne
coleif. OTyembHble UX CPOPMbI HMeIM OobWON reorpaduyeckuH apean, Ha KOTOPbIt Mao
BAMA pasiH4yna B OKpy2xKaroule cpeze. Apean MHOTHX BHUAOB MO3a3aBPOB, HaceMABLUIUX
3anayHy¥o oKpaHHy Cesepo-AtaauTuyeckoro Gaccelina, JOXOAHA MOBHAMMOMy JO Mopeli
Sanaguon, Esponsr. Ux BuxbI M3 BOCTOYHOM OKpanHbl OaccetHa Tuxoro okeaHa MNOBMAM-
MOMY TIPHHadexanu K Apyroi 300reorpaduyeckol MpOBHHUH.

Mosa3appbl TpOM30WAM OT MPHMHTHBHBIX BapaHOB cpeszHero Mena (auruamo3aBpbl),
ocoOeHHOCTH Yepena NOCNeAHUX ObWIM MOXOKH Ha Te KE Y MO3A3ABPOB, HO OCTabHad MOP-
(bonorna” Oblia Kak yY COBpeMeHHbIX BapaHoB. I]pexzkH MO3a3aBPOB NOBHAMMOMY OBWIM ABYX
OCHOBHBIX THMOB; ObIIM CPOPMbI, HMEBLUIMe JAHHHbIe Tea HM paCclMpeHHbIle XBOCTbI, OT HHX
TIPOH3OLWIM MO3a3aypHHEI; Opin CpOpMbl C KOPOTKHMH TelaMM VM AJMHHbIMH OCTPOKOHeY-
HbIMM XBOCTaMH, — 9TO ObIIM MpeAKM MAMOMMATeKAPNUHOB HW THAOZAYPHHOB.

B Teyenve cBoero KOpOTKOrO NepvOsa XKH3HH B HCTOPHM 3eMJH MO3a3aBpbI NOKa3salH
HECKOJIbKO MpOrpeccHBHbIX TeHJCHUHH, KaK, HapuMep, TeHAeCHUHIO K yBesHYeHHIO OOULero
pa3Mepa, K MOMeLIeHHIO MepesHeli MOMOBUHEI TEMeHH Had MepeAHHM KpaeM 3aHel NoNO-
BUHbI TEMeHH (BCTeEACTBHe YeroO YMeHbIUMaCb MOABHXKHOCTb KOCTeH ugpena), yBernyeHHe
4uCla THPaJIbHbIX MO3BOHKOB H H3Me€HeHHe NPHMHTHBHBIX MepenOHyaTLIX IaBHHKOB B JIMH-
Hple, C OOJbLUIHM YHCIOM cadaHr.

. aren “a: .
* Sistematika i morfologiia amerikanskikh mozazavrov (Reptilia, Sauria)

RUSSELL: AMERICAN MOSASAURS 3

SYSTEMATIQUES ET MORPHOLOGIE
DES MOSASAURES AMERICAINS

(Reptilia, Sauria)
Dave A. RussELL

RESUME

Les mosasaures étaient des hydrosauriens ayant la forme de grands lézards
qui se sont multipliés et diversifiés dans la seconde moitié du Crétacé pour
disparaitre 4 la fin de cette période. La famille des Mosasauridés se subdivise en
trois sous-familles: les mosasaurinés, comprenant les nouvelles tribus des mosa-
saurins, des globidensins et des plotosaurins, les plioplatécarpinés, comprenant
les nouvelles tribus des plioplatécarpins et des prognathodontins, et les tylo-
saurinés. On a jusqu’ici catégorisé 13 genres et 30 espéces du Crétacé en Amérique
du Nord, y compris un nouveau genre (Ectenosaurus) et une nouvelle espece
(Plioplatecarpus primaevus). Parmi les noms déja existants, dix-neuf noms
génériques et vingt-neuf noms spécifiques ont été placés en synonymie, deux
noms génériques et dix-huit noms spécifiques sont basés sur des données in-
suffisantes, et la valeur d’un autre nom générique et de six noms spécifiques est
contestable.

L’ostéologie des mosasaures est décrite en détail et en rapport avec la partie
non-osseuse de leur anatomie. Les mosasaures étaient doués d’une bonne vue,
mais leur odorat était faible. Un tympan calcifié, que l’on a retrouvé chez les
trois sous-familles, devait servir 4 transmettre les sons produits dans l’eau a
Yoreille médiane, et n’indique pas que ces sauriens avaient l’habitude de plonger
en profondeur. Les os carrés streptostyliques permettaient le mouvement antéro-
postérieur des mandibules, ce qui permettait d’avaler facilement des proies sous
l'eau. Les mosasaures nageaient par mouvements ondulants latéraux, se main-
tenant en €quilibre grace a leurs nageoires et 4 leur long cou. Grands carnivores
aquatiques, ils se nourrissaient de mosasaures plus petits qu’eux, de chéloniens,
de poissons, d’ammonites, de bélemnites, d’échinodermes et de pélécypodes.

Les mosasaures habitaient les mers semi-tropicales d’une profondeur de moins
de 100 brasses (environ 180 métres) et d’une salinité variable. Des formes partic-
uliéres de ces animaux suivaient une répartition géographique étendue qui
n’était que peu influencée par les dépéts sédimentaires présents. Plusieurs espéces
de mosasaures habitant la section occidentale du bassin de l’Atlantique Nord
se sont probablement répandues jusque dans les eaux de I’Europe de l'Ouest.
Les mosasaures de la section orientale du bassin du Pacifique semblent avoir
appartenu a une province zoogéographique distincte.

Les mosasaures descendaient de varanidés primitifs (aigialosaures) du Crétacé
moyen qui possédaient plusieurs caractéristiques du crane que l’on retrouve chez
les mosasaures, mais conservaient une morphologie post-cranienne semblable
a celle du Varanus moderne. A l’origne il semble y avoir existé deux types
de base de mosasaures: un type au corps allongé et a la queue courte,
mais déployée, qui a donné naissance aux mosasaurinés, et un type de plus
petite taille, mais 4 la queue longue et effilée, qui a donné naissance aux plio-
platécarpinés et aux tylosaurinés. Au cours de leur présence relativement bréve
dans l'histoire géologique, les mosasaures ont donné quelques signes évidents
d'une évolution: leur taille a eu tendance a s’accroitre, leur frontal s’est recouvert
sur la section antérieure du pariétal (éliminant ainsi toute possibilité de la
cinétique cranienne), le nombre de leurs vertebres a la base de la queue a
augmenté, et leurs nageoires primitives palmées se sont transformées en longues
pattes-nageoires hyperphalangiques.

SYSTEMATICS AND MORPHOLOGY
OF AMERICAN MOSASAURS

(Reptilia, Sauria)

DALE A. RussELL*

INTRODUCTION

Surely as much has been written about mosasaurs as any other group of
fossil reptiles of comparable abundance and diversity. The division of the family
Mosasauridae into three major subfamilies and the close relationship to varanid
lizards has been generally recognized for more than fifty years. Yet on a generic
and specific level their taxonomy has been so confused that even after a
careful survey of the literature it was usually impossible to be sure of the correct
binomial to use in identifying a given specimen. As a consequence modern verte-
brate paleontologists have by and large tended to ignore mosasaurs as a subject
of research.

The present study grew out of an attempt to provide a systematic revision
of the mosasaurs of the United States. The morphology of the mosasaur skull
was studied in some detail in the course of preparing a doctoral dissertation for
the Graduate Faculties of Columbia University. A review of the postcranial
skeleton was largely completed during the tenure of a postdoctoral fellowship
at Yale University. Speculations on the nature of non-ossified organs are to
be found with the descriptions of the associated bones. A dissection of the head
and limbs of Varanus niloticus, together with the assistance of various publica-
tions on the anatomy of lizards, has been a great aid in interpreting the func-
tional significance of joints and areas of muscle attachment in the head and body
of mosasaurs.

Mexican mosasaurs reported on by Mehl and Miilleried have not been seen,
and information included below on Californian mosasaurs is based entirely on
Camp’s monograph. Undescribed mosasaur collections from the Selma Forma-
tion of Alabama in the Chicago Natural History Museum and from various
localities in Canada preserved in the museums of that country are not considered
here. Otherwise mosasaurian material in the collections of some nineteen
North American institutions has been examined. References in the primary
scientific literature (not including bibliographies, encyclopedias, textbooks, etc.)
pertaining to material of American mosasaurs are listed under the proper generic
and specific headings in the section on systematics. Where possible, published
figures of specimens are referred to species recognized in this study. In many
cases, however, when figures do not show diagnostic characters, they are placed
after the generic headings or under the species as identified by the particular
worker. The frontispiece was painted by Charles R. Knight in 1899, under the
direction of H. F. Osborn, and plates I and II have been taken from illustra-
tions in Williston (1898b), Osborn (1899a) and Camp (1942). Miss Linda McKnay
prepared figures 46, 52 and 59, and the remaining illustrations are by the author.

A tooth and jaw fragment from the Cretaceous of New Jersey, recognized by
Mitchill in 1818 as being similar to the then unnamed type of Mosasaurus
hoffmanni, was the first mosasaur specimen described from North America. The
next half-century was characterized by a steadily increasing series of short publica-

*Published with the aid of a National Science Foundation Publication Grant, No. GN-528.

4

RUSSELL: AMERICAN MOSASAURS 5

tions on fragmentary remains from New Jersey and the southern Atlantic and
Gulf states. An important exception was the discovery in about 1830 (see Simpson,
1942, p. 172) of the type of Mosasaurus missouriensis from the Missouri territory,
which was later described by Harlan and Goldfuss. In 1865 Leidy published the
first revision of American mosasaurs with excellent illustrations of the material
from New Jersey. During the latter part of the 1860’s Cope and Marsh described
new, and in some instances relatively well-preserved, specimens from the New
Jersey and Gulf Coast Cretaceous. These new taxa were listed by Cope in 1870
together with a review of all previous work.

Marsh and Cope led field parties into the Niobrara Chalk of western Kansas
in 1870-1871 and soon began describing new mosasaur material. Far too many
species were named; all were inadequately diagnosed; and the resulting con-
fusion has been a serious handicap to subsequent workers. Marsh, however,
correctly diagnosed the three common Niobrara genera in 1872, and Cope in
his 1875, “The Vertebrata of the Cretaceous formations of the West,” republished
his original descriptions of Niobrara mosasaurs, figured his types and published
an up-to-date summary of named mosasaur species from America. Marsh had
collectors working in the Niobrara for seven consecutive field seasons between
1874-1880 and amassed the very large collection of mosasaurs in the Yale museum,
which for the most part remained unstudied. Special mention is due to Prof.
B. F. Mudge, who collected for Marsh during the summers of 1874-1876. ‘The
industry and thoroughness with which he worked the Niobrara Chalk, the
excellence and number of his specimens, and the relative accuracy and complete-
ness of his field journal were outstanding for his time. C. H. Sternberg collected
the Niobrara for Cope in 1875 and 1877, but met with relatively little success
as Yale crews had prospected the outcrops before him.

Most of the mosasaur specimens in the museum of the University of Kansas
were obtained by H. T. Martin, C. H. Sternberg, E. P. West, and S. W. Williston
from the Niobrara Chalk during 1890-1895. These specimens formed the basis
of Williston’s 1898 revision of the Kansas mosasaurs, an excellent work which
was unfortunately limited because Williston was unable to examine Cope’s and
Marsh’s types. Other important papers published during this time were Baur’s
detailed description of the skull of Platecarpus (1892) and Merriam’s doctoral dis-
sertation on Kansas mosasaurs in the collections of the University of Munich
1894).

Bac the 1880’s C. H. Sternberg and his sons have collected many excellent
mosasaur specimens from the Niobrara and sent them to museums all over the
United States and Canada, as well as to China, England, Germany and Sweden.
The writings of the elder Sternberg present a very interesting account of his
collecting experiences in the Niobrara Chalk and give a unique and vivid im-
pression of mosasaurs as living animals. For a more detailed history of early
collecting in the Niobrara see Williston, 1898a, pp. 28-32.

After 1910 the literature has been mainly restricted to the documentation of
sporadic new occurrences, with the description of a few obviously distinct forms.
In 1942 Camp published detailed descriptions and figures of the unusual
mosasaurs from the latest Cretaceous of California, and Zangerl during 1945-
1946 collected many new specimens from the Mooreville Member of the Selma
Chalk in Alabama.

Although the imagination and industry of the individual researcher lies at
the base of progress in our understanding of vertebrate evolution, this progress

6 PEABODY MUSEUM BULLETIN 23

is certainly greatly facilitated by mutual co-operation between workers. The
present study could not have been sustained without the generous assistance of

many vertebrate paleontologists throughout the United States and Canada.
Above all, I am deeply grateful to Dr. Edwin H. Colbert, of the American Mu-
seum of Natural History and Columbia University, for his sponsorship of this
project, for his critical reading of much of the manuscript and for making the
collections of the American Museum available to me for study. His wise guidance
and patient counsel were invaluable and have made being his student a pleasure.
I would especially like to thank Dr. Bobb Schaeffer and Dr. Malcolm McKenna,
both of the American Museum of Natural History and Columbia University,
for their encouragement, their review of much of the manuscript, and for their
many valuable and instructive suggestions on its organization and content. I am
indebted to Dr. Donald Baird, of Princeton University, for many stimulating
conversations and for the interest he has shown in this work. I have also profited
greatly from discussions on the kinetics of the sauropsid skull with my fellow
student at Columbia, Georg Zappler.

I was fortunate to receive a grant from the American Museum of Natural
History to visit several important collections in the United States and Canada
east of the Rocky Mountains. Of the numerous people who have aided me in
the course of these travels I would like to acknowledge the following who freely
gave much of their time to acquaint me with mosasaur specimens in the collec-
tions of their respective institutions: Dr. Theodore Eaton of the University of
Kansas, Dr. William Fox of Williams College, Dr. Nicholas Hotton III and Dr.
C. Lewis Gazin of the United States National Museum, Dr. John H. Ostrom of
Yale University, Dr. Horace Richards of the Academy of Natural Sciences of
Philadelphia, Dr. Alfred S. Romer of Harvard University, Dr. L. S. Russell,
formerly of the National Museum of Canada, now of the Royal Ontario Mu-
seum, Mr. Myrl Walker of Fort Hays Kansas State College, Dr. Robert W. Wilson
of the South Dakota School of Mines and Technology, Dr. Albert Wood of
Amherst College, and Dr. Ranier Zangerl of the Chicago Natural History Mu-
seum. I am particularly grateful to Dr. and Mrs. William A. Clemens of the
University of Kansas for inviting me to stay in their home during the time spent
studying mosasaurs in the extensive Kansas collections.

I would like to express my sincerest thanks to Dr. Elwyn L. Simons of Yale
University for making his home available to me over a two month period in
1963 and again for a month in 1964, enabling me to study the collections there
without financial difficulty. I then entered on the tenure of a National Science
Foundation Postdoctoral Fellowship generously granted to me to complete the
review of the postcranial skeleton based for the most part on the great collection
of mosasaurs assembled at the Peabody Museum of Yale by Prof. Marsh. Discus-
sions of the functional anatomy, particularly that of the head, have been greatly
improved by the detailed constructive criticism of Dr. John H. Ostrom and Dr.
James A. Hopson of Yale University. Dr. Wann Langston, Jr. of the Texas Me-
morial Museum has supplied much information on the skull and skeleton of a
Mosasaurus maximus from the Cretaceous of Texas. Dr. Karl Waage of Yale
University and Dr. J. A. Jeletzky of the Geological Survey of Canada have very
kindly proofread the section on stratigraphy.

Finally I would like to thank my brother and sister, Dr. Donald E. Russell
of the Muséum National d’Histoire Naturelle, Paris, and Mrs. John E. Mawby
and especially my wife, Janice A. Russell, whose assistance and encouragement
during various phases of this work have been very much appreciated.

The following abbreviations of institutional names precede the specimen

RUSSELL: AMERICAN MOSASAURS #]
numbers referred to in the text and identify the place of storage of the specimens:

A.C.—Museum of Amherst College.

A.M.N.H.—American Museum of Natural History.

A.N.S.P.—Academy of Natural Sciences of Philadelphia.

C.1.T.—California Institute of Technology, collections are now in the Los
Angeles County Museum.

C.M.—Carnegie Museum.

C.N.H.M.—Chicago Natural History Museum.

F.H.M.—Museum of Fort Hays Kansas State College.

K.U.—University of Kansas Museum of Natural History.

M.C.Z.—Museum of Comparative Zoology, Harvard University.

M.S.S.C.—Museum of Southern State College, Arkansas (casts of specimens
given to the A.M.N.H. by Dr. J. Chapman).

Museum of Williams College (not abbreviated).

N.J.G.S—New Jersey Geological Survey.

N.J.S.M.—New Jersey State Museum.

N.M.C.—National Museum of Canada.

P.U.—Museum of Princeton University.

S.D.S.M.—Museum of the South Dakota School of Mines and Technology

T.M.M.—Texas Memorial Museum.

U.S.N.M.—United States National Museum.

Y.P.M.—Yale University, Peabody Museum of Natural History.

HISTORY OF CLASSIFICATION
GENERAL RELATIONS OF MOSASAURS

The history of the first mosasaur specimen to attract the attention of the
scientific world is most unusual and has been repeated often. Its discovery
antedates that of the first dinosaur by some 42 years. The fossil was found in 1780
by quarrymen working in a shaft 90 feet below the surface of St. Peters Mount
at Maestricht, Holland, in rocks of latest Cretaceous age. Dr. Hoffman, a
surgeon of Maestricht, was called to examine the remains of what proved to be
a fine skull and vertebral column and proceeded to direct their excavation. How-
ever the specimen was found beneath ground owned by a clergyman named
Goddin who laid claim to it and forced Hoffman to yield the specimen as well
as pay the costs of the resulting lawsuit. In 1795 the troops of the French Re-
public repulsed the Austrians, laid siege to the city of Maestricht and bombarded
Fort St. Peter. The commanding general of the French army ordered his artil-
lerymen to avoid Goddin’s country house where the now-famous fossil was kept.
Goddin guessed the reason behind this apparent act of benevolence and had the
specimen removed and hidden in the city. After Maestricht fell a reward of 600
bottles of wine was offered for its recovery and the following day the specimen
was borne into French hands by a dozen grenadiers and sent to Paris, where it
remains in the collections of the Muséum National d’Histoire Naturelle (see
Faujas Saint Fond, 1799, quoted in Williston, 1898b, pp. 84-85; and Cuvier,
1808).

The first general review of the “grand animal fossile de Maestricht” was that
of Cuvier (1808) in which the remains were redescribed and refigured, and the
speculations of four earlier authors were discussed. Pierre Camper and Van
Marum are stated to have thought the animal was a cetacean, though Faujas
Saint Fond firmly believed that it was a crocodile. Adrien Camper recognized
that the animal was indeed a lizard, showing resemblances to iguanas on one
hand and to varanids on the other, a conclusion which Cuvier (1808, pp. 175-176)
endorsed. Forty-two years after its discovery the fossil was named Mosasaurus by
Conybeare (1822, p. 298). The trivial names belgicus, camperi, giganteus and
hoffmanni have all been applied to this specimen. Camp (1942, pp. 44-45) after
a careful perusal of the literature found that Mosasaurus hoffmanni Mantell
(1829, p. 207) is probably the earliest name available, with the famous Maestricht
fossil as its designated type.

For the next eighty years European workers continued to regard mosasaurs
as large marine lizards closely allied to varanids. Goldfuss (1845, pp. 179, 188)
in describing an American species of Mosasauwrus supported this view. Owen
(1840-1845, p. 258) agreed that the true relations of Mosasaurus lay between the
iguanids and varanids. He noted that the teeth of his newly-named Liodon
resembled those of Mosasaurus more closely than those of any other reptile
known at that time. Later (1849-1884, pp. 29-45) Owen placed Mosasaurus and
Liodon together in the undiagnosed tribe Natantia of the Lacertilia. In 1853
(p. 471) Gervais proposed the family Mosasauridae and included it in the Lacer-
tilia, although provisionally in a distinct suborder along with the Megalosaur-
idae and Iguanodontidae, both of the latter being families of dinosaurs. Within
the Mosasauridae were ranged Mosasaurus, Liodon, Onchosaurus (a batoid shark,
see Romer, 1945, p. 577), Oplosaurus (= Pelorosaurus, a sauropod, see Romer,

8

RUSSELL: AMERICAN MOSASAURS 9

1956, p. 621), Macrosaurus and Geosaurus (a thalattosuchian crocodile, see
Romer, 1956, p. 607). Owen made a detailed comparison of the skulls of chelon-
ians, lizards (Amblyrhynchus, Varanus, etc.), snakes (Python) and mosasaurs
(Clidastes, Mosasaurus, Platecarpus) and concluded (1877, p. 715) in response to
Cope’s (1875, pp. 113-127) assertion that mosasaurs belonged to a separate order
allied to snakes, “. . . that the fossil evidences of the mosasaurians hitherto made
known do not yield a single character peculiar to and characteristic of the
ophidian order.’’ Owen implied that mosasaurs should be classed within the
Lacertilia as a distinct suborder, taking the assignment of the Pinnipedia under
the Carnivora as an analogous example.

Cope (1869b, p. 254) had erected a new order of reptiles, the Pythonomorpha,
solely for the inclusion of the mosasaurs. These animals were postulated to
combine lacertilian and ophidian characters and were diagnosed by Cope (1869b,
p- 254) as follows: “Opisthotic distinct, not, or scarcely in contact with parietal
arc, embraced at one end by the prootic and exoccipital, and supporting squa-
mosal and quadratum.” Later Cope (1875, p. 126) stated that, “As a conclusion,
it may be derived that (mosasaurs) are not nearly related to the Varanidae as
has been supposed, but constitute a distinct order of the streptostylate group
(which also includes the orders Lacertilia and Ophidia); that they are primarily
related to the Lacertilia, secondarily to the Ophidia, and thirdly to the
Sauropterygia; that they present more points of affinity to the serpents than does
any other order; and their nearest point of relationship in the Lacertilia is the
Varanidae or Thecaglossa.” Throughout his life this remained his opinion in
the face of much opposition (see Cope, 1869-1870, 1878, 1895a, 1895b, 1896;
for summaries see Baur, 1892; and Williston, 1898b, pp. 94-99).

Marsh (1872b, p. 448) initially followed Cope in regarding mosasaurs as a
separate order of reptiles. Leidy (1873, p. 271) preferred not to use Cope’s term
“Pythonomorpha” but separated the mosasaurs on an ordinal level from the
lizards. By 1880 (p. 87) Marsh referred the suborder Mosasauria (also abandoning
the term “Pythonomorpha’”’) to the Lacertilia because of the presence of a
sternum, fore and hind limbs, and the absence of palatine teeth. Merriam
(1894, p. 14), after a critical appraisal of Cope’s (1878) list of characters of
pythonomorphs, likewise concluded that they could not be separated from
lizards. The polemic between Cope (1895a, 1895b, 1896) and Baur (1892, 1895,
1896) on mosasaur relationships is particularly hard to follow because of their
differing views on the homologies of the bones of the quadratic suspensorium
(see Camp, 1923, p. 347). Cope’s position was that the element here termed
the supratemporal was actually a separate paroccipital ossification. Irrespective
of the nomenclature involved, precisely the same bony elements bearing almost
exactly the same mutual relations exist in the suspensorium of Varanus, as Baur
(1896, p. 146) pointed out, and Cope’s key distinction for his concept of the
Pythonomorpha cannot be vindicated.

Beginning with the last decade of the nineteenth century speculation on
mosasaur-lizard affinities took on a new dimension. Two groups of amphibious
lizards had been discovered in rocks of early late Cretaceous age in Yugoslavia
and England (Aigialosauridae and Dolichosauridae). These lizards are clearly
varanoid, with well-developed limbs, but also possess many characteristics of
mosasaurs.

Some authors thought the mosasaurs were a lacertilian derivation that
branched off the main phylum far back in Mesozoic time. Osborn (1899a, p.
172) did not mention the aigialosaurs, but stated that mosasaurs could not have
descended from the dolichosaurs because of the greater number of cervical

10 PEABODY MUSEUM BULLETIN 23

vertebrae in the latter group. He concluded (1899a, p. 188), “The Mosasaurs are
a very ancient marine offshoot of the Lacertilia, retaining certain primitive and
generalized Lacertilian characters and presenting a few resemblances in the skull
to the Varanoids; they are very highly specialized throughout for marine pre-
daceous life, and constitute a distinct subdivision of the order Lacertilia.” Fiir-
bringer (1900, pp. 616-622) followed Osborn’s conclusions and separated the
varanids, aigialosaurs and dolichosaurs from the mosasaurs on a subordinal
level. Fejérvary (1918, p. 435) held that it would have been impossible for the
aigialosaurs to acquire all the specializations of mosasaurs by late Cretaceous
time and instead derived mosasaurs from hypothetical Jurassic platynotans.

Another school of thought, represented by Boulenger, Dollo, Gadow, Korn-
huber, Kramberger, and Nopcsa, probably influenced by the analogy drawn
by Owen between the Lacertilia and Mosasauria, and the Fissipeda and Pin-
nipedia, considered the former two groups to be of equal taxonomic rank.
Boulenger (1891, p. 118; 1893, p. 205) regarded dolichosaurs as the common
ancestors of mosasaurs, lizards and snakes. Kramberger (1892, pp. 102-105) was
the first to point out that aigialosaurs were intermediate in morphology
between lizards and mosasaurs and were probably ancestral to both. Dollo (1894,
pp. 252-258; 1903, pp. 137-139; 1904b, pp. 217-222) and Nopcsa (1903a, p.
41; 1903b; 1923, p. 144) agreed in deriving the mosasaurs from the aigialosaurs,
but maintained that varanids and aigialosaurs descended from primitive ter-
restrial varanoids (“Proplatynota’’) that resembled modern varanids in many
respects. Kornhuber (1901, pp. 20-22) believed his Opetiosawrus (an aigialosaur)
was the best available ancestor for mosasaurs and did not distinguish aigialosaurs
from varanids, combining both within the family Varanidae. Nopcsa (1908, p.
61) suggested that snakes descended from the dolichosaurs, a possibility that was
opposed by Janensch (1906, p. 32), Fejérvdry (1918, p. 441), Camp (1923, p.
332), and others. Gadow (1901, p. 489) considered mosasaurs to be the marine
collateral branch of the Lacertilia.

The contemporary view of the relations of mosasaurs finds its roots in
the publications of Baur (1890, p. 262; 1892, p. 21) who insisted that mosasaurs
be considered as true lacertilians and classed them with the varanids in the
superfamily Varanoidea. Williston (1904) found Baur’s arrangement to be some-
what extreme, yet largely substantiated by the discovery of the aigialosaurs.
Williston (1904, p. 48; 1914, pp. 146-148; 1925, pp. 269-273) grouped the
varanids, dolichosaurs and aigialosaurs in the superfamily Platynota, included
the mosasaurs in the superfamily Mosasauria, and listed both superfamilies
under the Lacertilia. He stated (1904, p. 47) that he knew of “‘. .. no more striking
examples of evolution presented in all vertebrate paleontology than that of the
aquatic mosasaurs of the Upper Cretaceous, through the semiaquatic aigialosaurs
of the Lower Cretaceous, from the (hypothetical) terrestrial varanoids of the
lowermost Cretaceous or Upper Jura.” Versluys (1907) noted that an inter-
mandibular joint was already present in varanids and aigialosaurs and was well-
developed in mosasaurs. In his classification of lizards Camp (1923, pp. 321-322)
- . considerfed] the aigialosaurs as derived from true lizards near the
Varanidae and as ancestral to both Dolichosauridae and Mosasauridae . . . and
regard[ed] the presence of the annectant Aigialosauridae as grounds for denying
rank higher than that of a superfamily to the mosasaurs.” ‘This last work also
contains a summary of the history of classification of mosasaurs (1923, pp. 322-
325).

5 1942 (p. 44) Camp reinforced his earlier conclusions on the systematic
position of mosasaurs: “Altogether, the inner ear, skull elements, and cranial

RUSSELL: AMERICAN MOSASAURS 11

foramina of Platecarpus have so many of the peculiarities of Varanus that we
must admit a close relationship of the two forms. It may be correct to say
that the skull differences are much less than those between the fissiped and
pinniped Carnivora, and are also less than those between Varanus and the Iguani-
dae, Agamidae, Amphisbaenidae, and Gekkonidae. Manifestly there are good
morphological and paleontological grounds for considering the mosasaurs to be
derived from early Cretaceous varanoid lizards. Retention of so many close sim-
ilarities in the skull elements, brain case, cranial foramina, inner ear, extra-
columella, and lower jaw, as well as the presence in the lower Cretaceous of the
annectant Aigialosaurinae, should be recognized in the classification. This should
permit the mosasaurs to be included in the suborder Sauria as a superfamily of
the subsection Platynota.”

Camp has been followed by Gilmore (1928), Nopcsa (1928), and Hoffstetter
(1962), but not by McDowell and Bogert (1954, pp. 131-133) who combine
the Mosasauridae, Aigialosauridae, Dolichosauridae and Lanthanotidae in an
“ ‘aigialosaurian’ group” and thus separate them from the Varanidae and
Helodermatidae, also in the superfamily Platynota. Hoflstetter (1955, p. 611)
and Romer (1956, pp. 557-563) do not allow more than a familial distinction
between mosasaurs and other platynotans.

SUBDIVISIONS OF MOSASAURS

The separation of mosasaurs into subordinate groups has had a rather
complicated history. Cope (1869b, p. 258) originally divided his Pythonomorpha
into two families, the Mosasauridae and the Clidastidae. The former were
diagnosed as lacking a zygosphenal articulation on the vertebrae and having
the pterygoids fused along their medial surfaces. The last character was based
on Goldfuss’ (1845, pl. 8) erroneous figure of the palate of Mosasawrus mis-
souriensis in which the pterygoids were drawn as co-ossified. Included genera
were Macrosaurus, Mosasaurus and Platecarpus. ‘The Clidastidae, including only
Clidastes, were characterized by the presence of a vertebral zygosphenal articula-
tion and separate pterygoids. In his 1869-1870 synopsis Cope added Halisaurus,
Liodon, Polygonodon (a fish tooth, not mosasaurian) and Tylosaurus to his
Mosasauridae. The first indication of a modern type of classification is seen in
a semi-popular paper by Marsh (1876, p. 59) where the families Edestosauridae
(based on Edestosaurus = Clidastes) and Tylosauridae were mentioned in pass-
ing but not characterized.

Following Owen and Marsh, in 1884 (p. 653) and 1885b (p. 335) Dollo
classed the mosasaurs as a suborder of the Lacertilia and separated them into
two families. The Plioplatecarpidae were defined as possessing a sacrum, an
interclavicle and a canal through the basioccipital-basisphenoid, all three
characters presumably being absent in the Mosasauridae. Clidastes, Mosasaurus,
Liodon, Platecarpus, Halisaurus and Tylosaurus were placed in the latter family,
while the former was monotypic, including only the genus Plioplatecarpus. In
1890 (pp. 162-163) Dollo divided his Mosasauridae into three groups, based on
the degree of development of a rostrum on the premaxilla, the size of the
suprastapedial process of the quadrate, and whether or not the haemal arches
were fused to the centra of the caudal vertebrae. A “microrhynchous” group
contained Platecarpus and Prognathodon; a “mesorhynchous” group, Clidastes
and Mosasaurus; and a “megarhynchous” group, Tylosaurus and Hainosaurus.
Dollo noted (1890, p. 169) that Plioplatecarpus had the characteristics of the
microrhynchous group. In 1894 (p. 221) Dollo realized that the “sacral”

12 PEABODY MUSEUM BULLETIN 23

vertebrae of his original two specimens of Plioplatecarpus were pathologically
fused “‘lumbars” (pygals), and that an interclavicle was also present in Mosasaurus.
He therefore synonymized his Plioplatecarpidae with the Mosasauridae and
recognized but one family of mosasaurs. Evidently abandoning his own Clida-
stidae, Cope (1889, p. 866; 1891, p. 50; 1898, p. 79) continued to subdivide the
mosasaurs into the two families used earlier by Dollo.

Williston (1895, p. 169; 1897b, p. 102) included Clidastes and Mosasaurus
in the Mosasauridae and named the Tylosauridae for Tylosaurus and Platecarpus,
diagnosing both families on postcranial characters. Soon after, he (Williston,
1897d, pp. 176-177) formally replaced Dollo’s microrhynchous, mesorhynchous and
megarhynchous groups of mosasaurs with the subfamily names Platecarpinae,
Mosasaurinae and ‘Tylosaurinae. These subfamilies were well-defined (1897d, pp.
180-181) and correspond to the three subfamilies used in the present work, after
reducing the rank of Dollo’s older Plioplatecarpidae and substituting it for Wil-
liston’s Platecarpinae. It should be noted that Dollo redefined his Plioplatecar-
pidae again in 1924 (p. 188) for the genus Plioplatecarpus and named the
Globidensidae for Globidens and Compressidens. The latter family name has been
accepted in modified form by Williston (1925, p. 273), Gilmore (1928, p. 8),
Romer (1956, p. 562), and others, but is probably best considered a synonym of
the Mosasaurinae.

The major classification adopted here is that of Camp (1961, p. 522) and
the arrangement of subfamilies is similar to that of Williston. Tribal categories
within the Mosasauridae are new.

Crass REPTILIA

SUPERORDER SQUAMATA

Orver SAURIA (= LACERTILIA)
SusorpDER ANGUIMORPHA
INFRAORDER PLATYNOTA

SUPERFAMILY VARANOIDEA
Famity AIGIALOSAURIDAE
Famity DOLICHOSAURIDAE
Famity HELODERMATIDAE
Famity LANTHANOTIDAE
Famity VARANIDAE

SUPERFAMILY MOSASAUROIDEA
Famity MOSASAURIDAE

SUBFAMILY MOSASAURINAE
TrisE MOSASAURINI
Tribe GLOBIDENSINI
Tre PLOTOSAURINI

SuBFAMILy PLIOPLATECARPINAE (= PLATECARPINAE)
TrinsE PLIOPLATECARPINI
TrinE PROGNATHODONTINI

SUBFAMILY TYLOSAURINAE

GENERAL DESCRIPTION OF MOSASAUR SKULL

The general shape of the skull is quite constant among the different genera
of mosasaurs and suggests a closely-knit group of reptiles related to, but distinctly
separated from, the varanid lizards. Although the range of form within the
Mosasauridae is greater than that in the Varanidae, it is interesting to observe
parallel variations occurring within both families; for example, the convergence
of the margins of the parietal “table’’ at the center of the dorsal surface of the
bone in Platecarpus and Varanus grayi, the unusually broad parietal of Clidastes
sternbergi and Varanus subgenus Tectovaranus, and the powerfully built jaws
and crushing dentitions of Globidens and Varanus subgenus Empagusia (see
Mertens, 1942, pls. 22-26 for figures of the above-mentioned varanids).

The mosasaur skull grossly resembles an elongate, laterally compressed and
bluntly terminated cone, the posteroventral corner of which has been horizontally
transected. The muzzle forms the anterior half or more of the skull and lies in
front of the antorbital wall and the splenio-angular joint of the lower jaw.
Except for the slender bilobate external narial opening and the space between
the ventral margin of the dentary, the muzzle is completely surfaced with bone.
The large orbits are located behind the anteroposterior midpoint of the skull
and open laterally and slightly dorsoanteriorly. The supratemporal fenestra is
located above and behind the orbit and is widely separated from its opposite
by the broad dorsal surface of the parietal. The large infratemporal opening
is separated from the above fenestra by a slender arcade of bone that connects
the dorsoposterior rim of the orbit with the quadratic suspensorium. The
quadrate is streptostylic, articulating with the distal end of the suspensorium
above and the mandible below, and was presumably bound ventroanteriorly
to the posteroventral corner of the jugal by a long ligament. The lateral
surface of the lower jaw narrows posteriorly and terminates in a rounded
retroarticular process which projects behind the occipital condyle. The roof
of the oral cavity is broken by five major fenestrae: two anterior openings
between the vomero-palatines and the maxillae for the internal nares; two
suborbital vacuities bounded by the palatines, pterygoids, ectopterygoids and
jugals; and a slender lenticular opening (incisura piriformis) between the
anterior portions of the pterygoids. Behind the ectopterygoidal processes the
pterygoids are constricted to allow the passage of the temporal musculature to
the lower jaws. The pterygoids then diverge posteriorly to meet the quadrates.
On either side of the supraoccipital the posttemporal fenestrae pierce the
occipital face of the skull between the suspensorial bridge from the parietal
above and the paroccipital process below. The foramen magnum is relatively
smaller than in Varanus.

In Clidastes the skull is slender with relatively long jaws and a shortened
temporal region. The skull is similar but larger in Mosasaurus, where the jaws
are more powerfully constructed. In Plotosaurus (Camp, 1942, fig. 1) the skull is
remarkably ichthyosaurian in lateral outline, with large orbits and dorsally
convex, gently posterior-inclined, alveolar margins. The skull in Platecarpus is
short and deep, with a relatively large temporal region. Prognathodon generally
resembles Mosasaurus in the shape of the head, although the anterior portion of
the muzzle is deeper and more bluntly terminated. Except for its very elongated
rostrum, the skull of Ectenosaurus seems to be close to that of Platecarpus. In

13

14 PEABODY MUSEUM BULLETIN 23

Tylosaurus the head is more perfectly conical than in the above genera, and the
muzzle is proportionally longer than in any of them, with the exclusion of
Ectenosaurus.

The mosasaur skull can be divided into nine structural components, each
primitively articulating with at least one other such unit through a moveable
joint, although later in the phylogeny of mosasaurs some of these articulations
lost their mobility. glinese components are the muzzle unit, the parietal unit,
the occipital unit, pterygoid units, epipterygoid units, stapes units, quadrate
units, the anterior division of the mandible and the posterior divison of the
mandible. In the following section each individual bony element in the mosasaur
head is described under the structural unit of which it is a part. The variations
of that particular element among the different genera are noted. A survey of
the literature on varanids together with a dissection of the head of Varanus
niloticus has greatly facilitated speculation on the interrelationships between
the bones and soft anatomy of the head, for the mosasaur skull is generally
quite comparable to that of varanids. The terminology of musculature is after
Lakjer (1926), that of the cranial nerves is after Watkinson (1906) and Bahl
(1937).

For additional descriptions of the cranial morphology of American mosasaurs
the reader is referred to the works of Camp (1942), Williston (1898b), Huene
(1910, 1911), Merriam (1894), Baur (1892), and Osborn (1899a), listed in order
of decreasing importance. The California material has not been examined and
the information included below relating to these forms has been taken from
Camp’s (1942) monograph.

Muzzie UNIT

The largest of the nine structural units in the mosasaur cranium is the
muzzle unit. This forms the anterior two-thirds of the skull and includes 24
bones (fused premaxillae, maxillae, nasals, fused frontals, lacrymals, prefrontals,
supraciliaries, postorbitofrontals, jugals, squamosals, palatine-vomers, septomaxil-
lae and orbitosphenoids) united together to form a half-cone with the vaulted sur-
face above and two arms (supratemporal arcades) projecting posteriorly to the
quadratic suspensoria. The muzzle unit meets the parietal dorsoposteriorly in
a broad transverse hinge (mesokinetic axis) and the pterygoids (pterygoid
units) ventroposteriorly on either side of the incisura piriformis through flexible
overlapping sheets of bone (hypokinetic axis). The jugal contacts the lateral
face of the ectopterygoid (pterygoid unit) through a flat vertical sliding surface.
On the posterior termination of the supratemporal arcade the expanded end
of the squamosal forms a rotating joint (metakinetic axis) with the lateral face
of the supratemporal (occipital unit), The anterior and posterior portions of
the muzzle unit are then raised or lowered about the mesokinetic axis. The unit
is supported below by two flexible struts (pterygoid units) and posteriorly by
the connection of the supratemporal arcades with the distal end of the quadratic
suspensoria.

PREMAXILLAE

The mosasaur premaxillae differ markedly from those of Varanus because
their anterior rostral surfaces are always smoothly continuous with the rostral
surface of the skull, instead of being depressed posteriorly to form a part of
the floor of the nasal capsule. They possess alveoli for only two teeth instead of
four as in Varanus. Like Varanus, however, the premaxillae are co-ossified with
no indication of a suture and are described below as one bone.

RUSSELL: AMERICAN MOSASAURS 15

Text-fig. 1. Diagram of the functional units of a mosasaur skull. Abbreviations for this and all
cranial figures are as follows.

Kinetic units: Am, anterior mandibular unit; Ba, basal unit; Ep, epipterygoid unit; max,
mesokinetic axis; mtj, metakinetic joint; mtx, metakinetic axis; Mu, muzzle unit; Oc, occipital
segment; Pa, parietal unit; Pm, posterior mandibular unit; Qu, quadrate unit; St, stapes unit.

Cranial structures: a, angular; ah, ampulla of horizontal and anterior vertical semicircular
canals; ala, alar process of basisphenoid; ap, ampulla of posterior vertical semicircular canal;
ar, articular; ba, basilar artery; bab, branch of basilar artery; bo, basioccipital; bs, basisphenoid;
bt, basipterygoid process of basisphenoid; c, coronoid; ca, anterior vertical semicircular canal;
ch, horizontal semicircular canal; che, cerebral hemisphere; cp, posterior vertical semicircular
canal; ct, foramen for corda tympani; d, dentary; ds, dorsum sellae; e, epipterygoid; ec,
ectopterygoid; en, external naris; end, endolymphatic duct; ens, endolymphatic sacculus; f,
frontal or frontal suture; fo, fenestra ovalis; gl, intermandibular articulation; IAM, internal
auditory meatus; ic, internal carotid artery; icb, branch of internal carotid artery; in, internal
naris; inc, incisura piriformis; ios, interorbital septum; j, jugal; jo, aperture for Jacobson’s
organ; 1, lacrymal; la, lagena; m, maxilla or maxillary suture; 0, opisthotic; of, olfactory lobe;
or, orbit; os, orbitosphenoid; ot, otosphenoidal crest of basisphenoid; p, parietal or suture for
parietal; paf, parietal foramen; pl, palatatine; pld, perilymphatic duct; pls, perilymphatic sac-
culus; pm, premaxilla; pof, postorbitofrontal or suture for postorbitofrontal; popr, paroccipital
process of opisthotic; pr, prootic; pra, prearticular; prf, prefrontal or suture for prefrontal;
ps, parasphenoid; pt, pterygoid; q, quadrate; sa, surangular; set, sella turcica; soc, supraoccipital;
sp, splenial; sq, squamosal; st, supratemporal; sta, stapes; tym, calcified tympanum; u, utriculus;
v, vomer; vc, vidian canal; vcl, vena capitis lateralis.

Cranial nerves: I, olfactory nerve; II, optic nerve; V, trigeminal nerve; VI, abducens nerve;
VI, facial nerve; VIII, acoustic nerve, -a, anterior branch, -p, posterior branch; IX, glosso-
pharyngeal nerve; X, vagus nerve; XI, accessory nerve; XII, hypoglossal nerve.

Muscles inserting on cranium: AEMS, Mm. adductor mandibulae externus medialis et super-
ficialis; AEP, M. adductor mandibulae externus profundus, -pp, posterior head, -pq, quadrate
head; AM, M. adductor mandibulae, undivided; AMP, M. adductor mandibulae posterior; B,
bodenaponeurosis; CM, M. cervicomandibularis;s DM, M. depressor mandibulae; ICC, M.
iliocostalis capitis; LAO, M. levator angularis oris; LCap, pars articulo-parietalis of M. longis-
simus capitis; LCtc, pars transversalis capitis of M. longissimus capitis; LCtce, pars transversalis
cervicis of M. longissimus capitis; LPt, M. levator pterygoid; OCM, M. obliquus capitis magnus;
PPt, M. protractor pterygoid; PQ, M. protractor quadrati; Ps, M. pseudotemporalis, -pr, pro-
fundus, -sup, superficialis; Pt, M. pterygoideus, undivided; PtP, M. pterygoideus profundus;
PtS, M. pterygoideus superficialis; RCA, M. rectus capitis anterior; RCP, M. rectus capitis
posterior, -d, dorsal portion, -v, ventral portion; SC, M. spinalis capitis.

16 PEABODY MUSEUM BULLETIN 23

For descriptive purposes the mosasaur premaxilla may be divided into two
major portions, a broad anterior tooth-bearing apex and a slender internarial
bar. A large cylindrical “prow” projects in front of the premaxillary teeth in
Tylosaurus, and a smaller, more conical one is present in Clidastes, Mosasaurus,
Ectenosaurus and to some extent in Plotosaurus. The premaxilla terminates
directly in front of the most anterior tooth in Platecarpus, Plioplatecarpus (Dollo,
1889b, p. 275), Prognathodon and Plesiotylosaurus. The dorsal midline of the

Text-fig. 2. Mosasaur premaxillae. A. Platecarpus ictericus (AMNH 1820, x 14). B. Clidastes
propython (YPM 1319, x 14). C. Tylosaurus proriger (AMNH 1592, x 14).

bone may be smooth (Clidastes, Mosasaurus missouriensis, Ectenosaurus and
Tylosaurus), with a slight dorsal crest (Mosasaurus maximus and Plotosaurus)
or slightly sulcate (Platecarpus and Prognathodon). Paralleling the dorsal mid-
line on either side is an irregular row of foramina in Clidastes, Mosasaurus,
Plotosaurus, Platecarpus and Prognathodon. In Tylosaurus they are distributed
randomly over the sides of the “prow.” These foramina mark the exits of the
ophthalmic ramus of the fifth nerve which in Varanus enters the premaxilla
near the top of the premaxillo-maxillary sutural surface. In lateral trace this
suture varies among the species of mosasaurs, but generally is parabolic in form,
rising steeply from the alveolar margin and sweeping posteriorly to merge with
the outer edge of the internarial bar. Ventrally the roots of the premaxillary
teeth are separated from those of the opposite side by a grooved median ridge
(Clidastes, Platecarpus and also in Tylosaurus, according to Williston, 1898b, p.
104). A nubbin on the center of the ventral surface of the “prow” in Tylosaurus
is connected by weak ridges to the lateral margin of the dental alveoli and
probably marked the anterior limit of the gum line.

The internarial bar arises from a triangular base enclosed between the
anterior wings of the maxilla on the dorsal surface of the skull in two species of
Clidastes, in Mosasaurus, Plotosaurus, Platecarpus, Ectenosaurus and Prog-

RUSSELL: AMERICAN MOSASAURS 187

nathodon. It arises from a broad rectangular base in a similar position in
Tylosaurus and a narrower rectangular one in Clidastes sternbergt. In cross-
section the bar exhibits the shape of a “IT” or an inverted triangle, the
descending keel of which probably supported a cartilaginous internasal septum
posteroventrally, as in Varanus (Bahl, 1937, p. 163). The internarial bar is
laterally constricted between the dilated anterior portion of the external narial
opening and re-expands behind this region just before being wedged into the
median cleft in the anterior end of the frontals in Clidastes, Mosasaurus, Ploto-
saurus, Platecarpus, Ectenosaurus, and Plesiotylosaurus. The internarial bar is
nearly unconstricted and is relatively wide throughout its length in Tylosawrus,
and penetrates the frontal to a point far behind the posterior termination of
the narial opening, unlike in any of the foregoing genera. In Varanus the fused
nasals link the internarial bar of the premaxilla with the frontals, although the
premaxilla and frontals may contact beneath this bone. In mosasaurs the nasals
are vestigial and separated by the premaxillary bar and frontals (Camp, 1942,
pp. 27-28, fig. 14).

MAXILLA

The maxilla is rather constant in form in mosasaurs. Ventrally there is a
longitudinally flattened buttress supporting the tooth bases and housing the
conduit for the maxillary artery and the maxillary division of the fifth nerve.
Out of the lateral edge of this buttress a vertical lamina of bone ascends to form
a squamose suture with the prefrontal posteriorly. It is emarginated above by
the opening for the external nares and re-expands anteriorly to be cut off by
the obliquely descending premaxillary suture. The ventral outline of the buttress
is slightly concave upwards in Platecarpus, Ectenosaurus and Plotosaurus, but is
straight in Clidastes, Globidens, Mosasaurus, Prognathodon and Tylosaurus.
As would be expected the buttress is much wider and heavier posteriorly in
Globidens to support the powerful dentition. There is a medially projecting keel
on its posterointernal edge that contacts the palatine in Clidastes, Platecarpus
and Tylosaurus. Terminal branches of the maxillary nerve emerge through a
row of foramina just above the gum line on the lateral surface of the maxilla
and through smaller foramina scattered over the anterior surface of the bone.

In lateral aspect the suture for the premaxilla rises steeply from the alveolar
margin, then slopes posteriorly at a geometrically decreasing angle with the
horizontal axis of the skull in Clidastes, Globidens and Mosasaurus. In Ploto-
saurus, Platecarpus, Ectenosaurus, Prognathodon and Tylosaurus, the suture
usually rises vertically from the alveolar border, makes an abrupt turn after
reaching a position lateral to the top of the maxillary buttress, and continues
posterodorsally in a straight line. The sutural surface appears triangular in
dorsal outline, with the base of the triangle at the transversely truncated ventral
maxillary buttress and the apex pointing posteriorly along the anterior ascend-
ing lamina. In Platecarpus and Tylosaurus the apex of the triangle coincides with
the posterodorsal termination the suture; its surface is longitudinally ribbed and
grooved and is bounded externally by a large crest. In Clidastes, Globidens and
Mosasaurus the triangular area is small and lies far anteriorly on the suture;
its surface is smooth and continues posteriorly into an even crest, shaped like
an inverted “V”’ in cross-section, that extends back to meet the anterior end of
the narial emargination. The ante-narial portion of the maxilla is never bent
medially to form a floor beneath the nasal capsule, as it is in Varanus.

The external nares are bounded laterally by the maxilla, medially by the
internarial bar of the premaxilla, and posterolaterally by the prefrontal in

18 PEABODY MUSEUM BULLETIN 23

Clidastes, Mosasaurus, Platecarpus, Prognathodon and Plesiotylosaurus, but not
in Plotosaurus, Ectenosaurus and Tylosaurus where the prefrontal is excluded
from the narial margin by a tongue of the maxilla. They terminate posteriorly
in a “V’-shaped notch formed by the frontal and prefrontals meeting at an
acute angle in Clidastes, Mosasaurus, Prognathodon and Plesiotylosawrus, but
not in Platecarpus, Plotosawrus and Tylosaurus where the narial opening ends
in a rounded notch in the anterior margin of the frontal. The lateral swelling
of the internarial bar before it reaches the frontal and a gentle convexity on
the dorsal rim of the narial margin of the maxilla indistinctly divide the narial
opening into a large anterior lobe, presumably occupied by the external naris,
and a smaller posterior one (Clidastes, Mosasaurus, Plotosaurus, Platecarpus,
Ectenosaurus, Prognathodon and Tylosaurus). In Plesiotylosaurus the lobes are
subequal in size. The external narial fenestra is relatively long in Plotosaurus
bennisoni (amounting to 39% of skull length), intermediate in Prognathodon
overtoni, Plotosaurus tuckeri, Mosasaurus missouriensis (Williston, 1898b, pl.
20; Goldfuss, 1845, pl. 6), Platecarpus ictericus and Clidastes liodontus (amount-
ing to 26-34% of skull length) and short in Tylosaurus proriger and Plesiotylo-
saurus crassidens (amounting to 20-24% of skull length).

Behind the premaxillary suture and from a point dorsal to the second
maxillary tooth back to the seventh (eleventh in Clidastes sternbergi) the upper
rim of the maxilla is sharply emarginated from above for the anterior lobe of
the external narial opening. It re-ascends posteriorly in a smooth curve to be
emarginated again, but to a lesser extent, forming the ventral border of the
posterior lobe of the narial fenestra. In some genera Plotosaurus, Ectenosaurus
and Tylosaurus, the maxilla may send a tongue posteromedially over the pre-
frontal to contact the lateral edge of the frontal, but usually this tongue is
small and the maxilla forms a squamose suture directly with the lateral lamina
of the prefrontal. This suture descends in a straight line to the posterior end
of the maxilla. The maxilla overlaps the prefrontal anterolaterally and is in
turn overlapped posteriorly by the spatulate anterior end of the horizontal ramus
of the jugal.

NASAL

The nasal is known in only three specimens of American mosasaurs. The
nasal in the type of Plotosaurus bennisoni “‘. . . is a vestigial bone only 60 mm.
in length which lies at the junction of the premaxillary and frontal under the
broadened posterior fourth of the internarial bar. Its anterior tip loosely
occupies a hollow in the side of the premaxillary bar and its thin posterior
end articulates with the lateral side of the internarial process of the frontal on
a squamous suture. One-third of the way forward is a small, rounded
vertical ridge which may mark an attachment to the outer wall of the nasal
membrane. The loose articulation of the nasals with the internarial bar and
their frequent loss may indicate that the nasals were closely united to the saclike
integuments of the narial chamber.’ (Camp, 1942, pp. 27-28, fig. 14)

The nasals in Tylosaurus “. . . sind schmale langliche Knéchelchen an dem
Medialrand jeder Nasenéffnung. Sie kommen beinahe, aber nicht ganz mit dem
spitzen des Frontale zusammen.” (Huene, 1910, p. 303, fig. 5) Separate nasals
were reported by Wiman (1920, p. 15, fig. 4) in Clidastes sternbergi.

FRONTALS

The frontals are solidly fused along the midline of the skull, forming a flat
triangular plate linking the internarial. process from the premaxilla with the

RUSSELL: AMERICAN MOSASAURS 19

median dorsal axis of the parietal. The frontal is narrow anteriorly and forms
the posteriormost portion of the internarial bar, enclosing the premaxillary shaft
in a tight “V’-shaped notch (Clidastes, Plotosaurus, Platecarpus, Ectenosaurus
and Plesiotylosaurus). In Tylosaurus the internarial bar is formed almost en-
tirely of the premaxilla which penetrates deeply into the anterior end of the
frontal.

The frontal dilates behind the external narial opening, in some instances
forming a rounded posterior boundary for the narial fenestrae (Plotosaurus,
Platecarpus and Tylosaurus), in others the anterolateral margins diverge poste-
riorly in a more regular manner (Clidastes, Plesiotylosaurus). In Mosasaurus the
anterior edges of the frontal parallel the cranial midline behind the pointed

Text-fig. 3. Frontal of Mosasaurus maximus (NJSM 11052, x 34).

posterior ends of the external nares, then in front of the antorbital wall diverge
at a nearly 180° angle for a short distance, and finally turn posteriorly and a
little laterally to continue in a straight line back to the posterior base of the bone.
In Platecarpus and Ectenosaurus the lateral margins of the frontal curve out-
wardly back to a point above the antorbital wall. Here they turn gently medially
and then diverge again, forming a shallow sulcus over the orbits. The frontal
is similar in Clidastes but is relatively narrower and has straighter margins. In
Prognathodon the frontal is shorter and broader than in Platecarpus. Its anterior
margins diverge in straight lines back to the prominence above the antorbital
wall and from here posteriorly parallel the longitudinal axis of the skull, show-
ing no signs of a supraorbital emargination. In Tylosaurus, Plesiotylosaurus and
Plotosaurus the frontal margins are more regular than in the foregoing genera
and not so strongly divergent posteriorly. In the former two genera the margins
are slightly convex; in the latter one they are slightly concave. A median dorsal
ridge is present on the frontal of Clidastes (weakly developed), Globidens (present
but not sharply defined), Mosasaurus, Platecarpus, Ectenosaurus and Prognatho-

20 PEABODY MUSEUM BULLETIN 23

don. It may be either well-developed or absent in Tylosawrus and is absent in
Plotosaurus.

On the undersurface of the frontals a triangular excavation, representing
the suture for the prefrontals, lies anterior to the center of the orbit and lateral
to the impressions of the olfactory tract along the ventral midline. In forms where
the prefrontal and postorbitofrontal are separate (Clidastes and Globidens?)
the posterior boundary of the suture is clearly marked by a laterally directed
buttress. In other mosasaurs (Platecarpus and Tylosawrus) where these two
bones meet beneath the frontal, the posterior boundary of the suture is obscured
by that of the postorbitofrontal. The postorbitofrontal suture proper is located
on the ventral surface of the posterolateral corner of the frontal. It extends
medially about one-half of the distance from the edge of the frontal to its midline
in Clidastes, Platecarpus and Tylosaurus, less than this distance in Globidens,

Text-fig. 4. Undersurface of frontal. A. Platecarpus ictericus (AMNH 1820, xX 14). B. Clidastes
propython (YPM 1368, x 14).

and only one-fourth of this distance in Clidastes sternbergi where it ends against
a buttress that approximately parallels the longitudinal axis of the skull. Both
of the above sutural surfaces are relatively smooth in Clidastes, but are marked

RUSSELL: AMERICAN MOSASAURS 21

with ridges and grooves radiating from the center of the bone in Globidens,
Platecarpus and Tylosaurus. The suture with the parietal is long and trans-
versely oriented (see description of parietal). Ridges and grooves on its surface dip
more steeply as the center of the frontal is approached, becoming vertical in
the middle of the bone. Tongue-like extensions from the frontal overlap the
mesokinetic axis on either side of the cranial midline to cover the anteromedial
surface of the parietal in several genera (see description of parietal).

The outline of the cerebral hemispheres is not clearly defined over the
posterior undersurface of the frontal but, as Camp (1942, p. 40, fig. 24) has
indicated, the hemispheres themselves must have been smaller than in Varanus.
Impressions in the frontal suggest that the cerebral hemispheres partly roofed
the posterior base of the olfactory tract. In front of the hemispheres this tract
expands slightly to form the olfactory bulbs which were sheathed ventrally by
wings from the frontal much as in Varanus. Fibers of the first cranial nerve then
probably spread out anteriorly under the frontal portion of the internarial bar.
In Clidastes and Globidens the olfactory tracts arise on a level with the dorsal
surface of the cerebral hemispheres, but in Platecarpus and Tylosaurus they arise
more ventrally. In Mosasaurus the flanges beneath the olfactory bulbs may
have been absent and the first nerve seems to have been more divergent ante-
riorly than in the above genera.

LACRYMAL

The lacrymal is a small bone located between the prefrontal and maxilla,
dorsal to the jugal on the anteroexternal edge of the orbit. It is rarely preserved
and, when present, is usually badly crushed. The lacrymal in Plotosaurus lies
almost entirely within the ventroanterior wall of the orbit. There is a small
foramen between the prefrontal, maxilla and lacrymal which may be homologous
with the much larger foramen between the lacrymal and prefrontal in Varanus.
Merriam (1894, pp. 28-29) states that in Platecarpus the lacrymal is a four-sided
plate with a small, deep notch appearing on one border.

In Tylosaurus the roughened lenticular outer surface of the lacrymal separates
the maxilla and prefrontal posterolaterally (see Osborn, 1899a, pl. 21; Huene, 1910,
fig. 5; Merriam, 1894, p. 22). The apparently well-exposed lateral surface of the
lacrymal seems to have had a quadrilateral outline in Clidastes. According to
Williston (1898b, p. 109) its outer side is also slightly roughened. The element
figured by Goldfuss (1845, pl. 7) is probably a fragment of the jugal instead of the
lacrymal as stated.

PREFRONTAL

The prefrontal in mosasaurs consists essentially of a flat dorsal plate of
bone that roofs the anterior half of the orbit, is in turn overlain by the frontal
above, and extends anteriorly out onto the face of the rostrum as far as the poste-
rior border of the external nares. The prefrontal is excluded from the narial mar-
gins in Plotosaurus, Ectenosaurus and Tylosaurus by a tongue of the maxilla.
A thin lateral lamina descends vertically from the rostrally exposed portion of
this dorsal plate to be overlapped ventroanteriorly by a thin sheet of bone from
the maxilla. Another heavier lamina that is vertically-convex posteriorly descends
perpendicularly from a point one-third of the way from the posterior end of
the dorsal plate, forms the anterior border of the orbit, meets the lateral plate
at right angles, and contacts a similar rising wall from the palatines below.
The connections with the palatine seem to have been cartilaginous. Thus the
prefrontal may be divided into: 1) a supraorbital plate or process continuing

22 PEABODY MUSEUM BULLETIN 23

Text-fig. 5. Prefrontal. A. Tylosaurus proriger (AMNH 4909, x 14). B. Platecarpus ictericus
(AMNH 1820, x 14).

anteromedially into an exposed diamond-shaped surface; 2) a triangular lateral
plate bounded by the maxilla anteroventrally, by the orbit posteriorly, and by
the dorsal surface of the skull above; 3) a vertical wall of bone descending from
the anterior limit of the supraorbital process,

The supraorbital process of the prefrontal is quite variable in mosasaurs. In
Clidastes and to a lesser extent in Mosasaurus, Prognathodon, Plotosaurus and
Plesiotylosaurus, there is a broad triangular ala projecting laterally from the
supraorbital process. This ala is present only as a small nubbin in Platecarpus
and Plioplatecarpus (Dollo, 1889b, pl. 9, fig. 6) and is entirely absent in Ecteno-
saurus and Tylosaurus. In Clidastes and possibly also Globidens the prefrontal
and postorbitofrontal do not come in contact beneath the frontal. In Prognatho-
don these bones contact, and in Platecarpus and Mosasaurus the postorbito-
frontal underlies a small portion of the posterior tip of the prefrontal. In
Tylosaurus the supraorbital process of the prefrontal receives a tongue from the
postorbitofrontal in a posteriorly deepening groove in its lateral margin. These
two elements are nearly fused in Plotosawrus.

SUPERCILIARY

There is an element infrequently associated with Platecarpus skulls that can
only be identified as the superciliary. It is, like the ectopterygoid, an “L-shaped
bone with a broad short wing possessing an oblique sutural surface on its dorsal
(?) face and a longer and slenderer wing whose lateral edge is beveled off.
There is no sharp, anteriorly projecting point at the anterior apex of the “L” as.
is usually the case in Varanus. What is here taken as the ventral surface is smooth
and longitudinally concave. Both Baur (1892, p. 15) and Camp (1942, p. 32)
have noted a roughened area on the lateral rim of the supraorbital wing of the
prefrontal in Platecarpus (present also in Plioplatecarpus, see Dollo, 1889b, pl.
9 fig. 6) that may have marked the point of attachment of the superciliary bone.

RUSSELL: AMERICAN MOSASAURS 23

Baur (1892, pl. 1 fig. 1) has illustrated the element under consideration and iden-
tified it as the superciliary, while Camp's (1942, fig. 18) superciliary is here
believed to be an orbitosphenoid.

Osborn (1899a, p. 171, pl. 21 upper figure, element marked “x”’) tentatively
identified a small bone as a superciliary in Tylosaurus. It has not been seen in
exceptionally well-preserved skulls of Clidastes and Plotosaurus.

POSTORBITOFRONTAL

The postorbitals and postfrontals are co-ossified in mosasaurs to form a
single four-pronged element, much as in Varanus (Camp, 1942, p. 5, reports
that a trace of a suture remains between these bones in Plotosaurus). The main
body of the postorbitofrontal consists of the dorsally exposed surface of a pro-
cess that extends anteriorly along the lateral edge of the frontal and bounds
the orbit posterodorsally. From this central body another process passes along
the posterior border of the frontal to contact the parietal medially, a third
descends to the jugal, and a final long and slender process fits into a deep
groove in the dorsal edge of the squamosal. The first two processes are connected
underneath the frontal by a thin sheet of bone and are separated from the other
more ventrally located rami by a groove in some genera (Platecarpus and
Tylosaurus) but are smoothly continuous with these rami in Clidastes, Globidens,
Plotosaurus, Mosasaurus and Prognathodon.

The main body of the postorbitofrontal is expanded laterally into a broad
triangular ala in Clidastes, comparable to the one developed on the prefrontal in
this genus. A similar but smaller and heavier ala exists in Globidens. In Mosa-
saurus, Plotosaurus and Prognathodon the ala is not so well developed, and in
Platecarpus and Tylosaurus the anterolateral edge of the postorbitofrontal is
only narrowly exposed. The external apex of the ala is curved ventrally in
Clidastes, Globidens, Mosasaurus, Plotosaurus and Prognathodon to form a
suture with the dorsolateral surface of the jugal. The process to the jugal is
larger in Platecarpus and Tylosaurus and in the former genus is distally ex-
panded, overlapping the jugal anteriorly and externally.

The medial process to the parietal is short in Platecarpus and loosely artic-
ulated with the parietal. In Mosasaurus and especially Clidastes the process is
largely covered by and apparently firmly sutured to a lateral wing from the
parietal. In Clidastes, Globidens, Mosasaurus, Plotosaurus, Platecarpus, Progna-
thodon and Tylosaurus the long process to the squamosal forms the anterolateral
boundary of the supratemporal fenestra and bears a sharply defined keel along
the entire length of its ventral surface which fits into a groove on the squamosal.
This process is exceedingly broad in Globidens, indicating the insertion of a
powerful M. adductor mandibulae externus on the ventral surface of the under-
lying squamosal.

The face of the postorbitofrontal lying beneath the posterolateral corner of
the frontal is smooth in Platecarpus and Tylosaurus. There is a ventral ridge
trending transversely and somewhat posteriorly across this surface in Clidastes
and Globidens which in Mosasaurus is less clearly developed. In Mosasaurus,
Plotosaurus and Platecarpus all four processes of the postorbitofrontal lie in the
same curved plane within the dorsolateral margin of the skull; in Clidastes and
Globidens the broad ala must have projected beyond this plane. In Tylosawrus
the parietal and jugal processes and a third antero-posterior axis formed by the
processes to the prefrontal and squamosal meet each other at right angles. The
prefrontal and parietal processes never enclose a lateral lobe from the parietal
between them in mosasaurs as they do in Varanus.

24 PEABODY MUSEUM BULLETIN 23

JUGAL

The mosasaur jugal in an “L”-shaped bone with a broad ascending ramus
forming the posterior boundary of the orbit and a narrower anterior horizontal
ramus forming its ventral boundary. There is a well-marked posteroventral
tuberosity at the crux of the “L” in Platecarpus, Mosasaurus and Ectenosaurus,
which is weaker in Prognathodon, Tylosaurus and Plotosaurus, and absent in
Clidastes. This tuberosity represents the anterior point of attachment of the
“quadratomaxillary” ligament which in Varanus arises partly on a lateral prom-
inence of the ectopterygoid and partly on the adjacent posteriormost tip of the
maxilla,

The vertical ramus of the jugal in Platecarpus and Plotosaurus is overlapped
dorsally by the postorbitofrontal but does not form an interlocking suture with
this element. Anteriorly, the horizontal ramus becomes laterally compressed and
is loosely applied to the posterior surface of the maxilla in Clidastes, Platecarpus,
Ectenosaurus, Prognathodon and Tylosaurus. Medially the horizontal ramus con-
tacts the ectopterygoid but there is no indication of a suture and the joint was
probably a sliding one.

SQUAMOSAL

The squamosal is a comma-shaped bone in mosasaurs that connects the upper
temporal arcade with the quadratic suspensorium. For the insertion of the post-
orbitofrontal the anterior shaft of the squamosal is deeply grooved dorsally
from its slender anterior termination next to the jugular process of the postorbi-
tofrontal back to a point opposite the posterolateral corner of the supratemporal
fenestra in Clidastes, Mosasaurus missouriensis (Williston, 1898b, pl. 20), Ploto-
saurus tuckeri, Platecarpus, Prognathodon, Plesiotylosaurus and Tylosaurus, and
somewhat behind this point in Mosasaurus maximus, Plotosaurus bennisoni and
Ectenosaurus.

Posteriorly, the squamosal is expanded into an arrowhead-shaped body, whose
median axis is more ventrally inclined than the anterior shaft of the bone. The
suture for the supratemporal covers all of the internal face of the head of the
squamosal and contains a shallow excavation that parallels its median axis antero-
dorsally to the posterior end of the squamosal shaft (Mosasaurus, Platecarpus) or
extends for a short distance over the posteromedial surface of the shaft (Clidastes,
Tylosaurus). A small wing of bone extends from the triangular body of the squa-
mosal dorsomedially to contact the suspensorial ramus of the parietal in
Clidastes, Mosasaurus (where it is very small), Plotosauwrus, Platecarpus, Progna-
thodon, Plesiotylosaurus and Tylosaurus. The squamosal has a horizontal facet
at its ventroposterior margin for articulation with the quadrate in Clidastes,
Platecarpus and Tylosaurus. In Mosasaurus maximus the surface of the facet is
vertical and located on the ventral face of the body of the squamosal.

‘The upper temporal arcade is bowed somewhat dorsally in Platecarpus and
Clidastes but is approximately straight in Mosasaurus, Plotosaurus, Progna-
thodon and Tylosaurus. Its ventral border is sheathed entirely by the squamosal
and it must have been on this bone that the pars superficialis and medialis of
the M. adductor mandibulae externus originated.

PALATINE-VOMER

The palatine is a rectangular plate of bone wedged between the pterygoids
posteriorly, the maxilla laterally, and the vomer anteriorly. The posterodorsal
surface of the palatine slopes up from its posterior termination to form a “J”’-

RUSSELL: AMERICAN MOSASAURS 25

shaped ridge, which surrounds a basin in the superior surface of the bone poste-
riorly and medially and contacts the descending antorbital wall from the pre-
frontal above. In Platecarpus and Tylosaurus there are two small crests which
trend anteromedially over the surface of this basin. The ventral surface of the
palatine is ridged posteromedially to receive an anteromedial process from the
pterygoid. Laterally the palatine expands somewhat to enclose an elongated

Text-fig. 6. Dorsal view of right palatine of Platecarpus ictericus (AMNH 1820 x 4).

elliptical cavity which fits over a longitudinal ridge on the posteromedial side
of the maxilla. The bone thins anteriorly to end in a shallow sulcus forming
the posterior border of the internal naris. In Varanus the palatine foramen lies
just medial to the maxillary suture on the posterior rim of the palatine and
transmits the maxillary branch of nerve V, the inferior orbital artery, and the
maxillary vein into the alveolar canal of the maxilla (Bahl, 1937, p. 161). This
foramen has not been seen in Clidastes, Platecarpus or Tylosaurus; it may have
been destroyed by crushing or the blood vessels and nerve may have passed
through a gap between the palatine and prefrontal in the lateral portion of the
antorbital wall. The palatines are relatively longer in Clidastes than in Plotosaurus
or Tylosaurus and are shortest in Platecarpus. They do not meet along the
midline.

In Plotosaurus the vomerine process of the palatine arises medial to the
posterior border of the internal naris, gently curves medially and then laterally
to complete the boundary of the naris internally and obliquely overlap the vomer
laterodorsally. The vomers cannot surely be distinguished from the vomerine
processes in Clidastes, Platecarpus and Tylosaurus. The vomers are much longer
and more closely appressed along the midline of the skull in all of these genera
than in Varanus. In Platecarpus one anterointernally trending crest crosses the
posterior portion of the undersurface of the vomer and another similar but
smaller crest crosses just behind the aperture for the Jacobson’s organ, very
reminiscent of the ventral appearance of this bone in Varanus. In Tylosaurus a
single heavy crest extends from the lateral edge of the center of the vomer past
the aperture for Jacobson’s organ to contact the premaxilla far anteriorly. There
is a longitudinal crest along the ventrolateral edge of the ventral shaft of the
vomer in Clidastes.

The bony opening for the internal naris lies medial to the posterior half of
the maxilla in mosasaurs. It is wide and triangular in Platecarpus, less wide but
also triangular in Clidastes, teardrop-shaped in Plotosaurus, and with the form
of a compressed teardrop in Tylosaurus. The aperture for Jacobson’s organ lies

26 PEABODY MUSEUM BULLETIN 23

between the maxilla and vomer, making a shallow groove in the side of the latter
element opposite the second maxillary tooth in Platecarpus and the fourth in
Tylosaurus. Immediately anterior to this aperture the vomer is sutured to the
medial wall of the maxilla. The foramen for the vein to the medial palatine sinus
emerges in front of the aperture for Jacobson’s organ, instead of behind it as
in Varanus (Bahl, 1937, p. 163). In Platecarpus the exit for this vein lies wholly
within the vomer; in Tylosaurus it is bounded laterally by the maxilla. The most
anterior portion of the vomer is quite slender and drawn-out in Tylosaurus com-
pared. to Platecarpus.

SEPTOMAXILLA

“In the type skull of Kolposaurus (=Plotosaurus) bennisoni there is a long,
thin, bladelike bone, the septomaxillary, lying beneath the internarial bar and
spreading out anteriorly above the prevomers. Posteriorly it continues to a point
beneath the frontal notches of the external nares. There it gives rise to a pair
of small, hooked, dorsal processes which curve backward toward the borders of
the palatine conchae. The thin dorsal margin of the septomaxillary contains a
longitudinal groove in its posterior half and this marks the midline fusion. The
deepest part of the internarial premaxillary bar lies in the forward continuation
of this groove. Near the anterior limit of the narial chamber, the bone becomes
depressed and enwraps a canal which passes forward and downward, evidently to
continue into the cavity where the Jacobson’s organ was lodged. Parts of the
septomaxillary have evidently been seen in Tylosaurus by Merriam (1894, p. 21,
pl. 1 fig. 3) and by Huene (1910, p. 303, fig. 5)... .”” (Camp, 1942, p. 28)

In a Tylosawrus skull in the Yale museum (YPM 4002) the septomaxillary
bones are clearly not fused, one appearing on the medial side of each narial
opening.

ORBITOSPHENOID

According to Camp (1942, p. 32, fig. 15) the orbitosphenoid in Plotosaurus

. supported the ventrolateral wall of the cerebrum and half enclosed the
fenestra for the optic chiasma which it half encircled above and posteriorly. It
extends obliquely forward and upward. Two small bony points extend upward
from its dorsal curved border. The general shape and position of the element is
much like that of Varanus and Gerrhonotus. The orbitosphenoid of Kolposaurus
(=Plotosaurus) displays another interesting feature which is not found, so far
as known, in recent lizards. From near its anteroventral border, a thin, bony
process extends forward and downward. The tip of this process adjoins its fel-
low from the opposite side. A brownish stain, accompanying granules of bone
in the matrix, runs from these tips obliquely forward and downward. This stain
and the granules seem to indicate the presence of a calcified cartilage in the
interorbital septum. The stain merges ventrally with a thick nodule of true bone
lying on the midline 12 mm. above the parasphenoid. This nodule was evidently
an ossified part of the ventral margin of the interorbital septum.”

Elements associated with several Platecarpus skulls in various museums have
the same general form as the orbitosphenoid in Plotosaurus. The dorsal portion
of the bone is flattened and fan-shaped and curves posteroventrally into a
smaller, vertical column that ends abruptly below in a rounded cap. Similar
bones from Platecarpus have been figured by Williston (1898b, pl. 29 fig. 5) and
Camp (1942, fig. 18). Williston (1902, pp. 250-251) also identified the orbitosphe-
noids in the specimen here referred to Mosasaurus ivoensis, subsequently lost,
and in another specimen of Platecarpus in which the expanded dorsal portion

“ce

RUSSELL: AMERICAN MOSASAURS PH |

Text-fig. 7. Lateral view of left orbitosphenoid of Platecarpus (AMNH 122, x 1).

of each bone was lying directly beneath an elliptical excavation for the cerebral
hemispheres located on either side of the olfactory groove on the posteroventral
face of the frontal.

PARIETAL UNIT

In mosasaurs the parietal unit is formed of only one element, the parietal
itself, but relations with other units in the skull are complicated. Ventrally a
descending flange loosely overlaps the prootics (occipital unit) and epipterygoids
(epipterygoid units). Further back on its posteroventral midline the parietal
articulates with the supraoccipital (occipital unit) by a narrow, interdigitating
hinge (metakinetic joint). The parietal is connected with the frontal (muzzle
unit) by the mesokinetic axis on the dorsal surface of the skull. Posterolaterally
each suspensorial ramus of the parietal slides over the dorsal surface of a long
medially-directed process of the supratemporal (occipital unit). Thus the parietal
unit is rocked up and down about the metakinetic joint through rotation of the
occipital unit below. It partially supports the occipital and epipterygoid units
anterodorsally through the hinge with the muzzle unit.

PARIETAL

The parietals are fused in mosasaurs without a trace of a suture and are here
considered as a unit. The parietal is broad anteriorly forming the anteromedian
margins of the supratemporal fenestrae, narrows posteriorly, and then branches
into two laterally diverging suspensorial rami which close the posterior bound-
aries of the fenestrae. In most mosasaurs the median dorsal surface of the parietal
is in the form of a rectangular “table” with the long axis oriented in the antero-
posterior direction. It is bounded on each side by an overhanging crest, beneath
which the M. pseudotemporalis superficialis presumably originated on the
dorsolateral wall of the parietal. These crests usually extend caudally across to
the posterior edge of the suspensorial rami and may indicate a more extensive
posterior area of origin for the M. pseudotemporalis superficialis than in Varanus.
In a few mosasaurs, Platecarpus, Plioplatecarpus (Dollo, 1889b, pl. 9 fig. 6), and
Ectenosaurus to a degree, these overhanging crests converge posteriorly and
meet in the midline of the parietal, forming a triangular “table” dorsally on
the parietal. In these forms the M. pseudotemporalis superficialis may have been
concentrated in the anterior region of the supratemporal fenestra.

The parietal foramen is always present close to or rarely, as in Platecarpus, is
bordered anteriorly by the frontal and is quite large in some forms, Platecarpus,
Clidastes sternbergi, Plotosaurus and Plioplatecarpus (Dollo, 1889b, pl. 9 fig. 6).

28 PEABODY MUSEUM BULLETIN 23

In Prognathodon the parietal foramen is located in the center of a small rectangu-
lar prominence in the anterior end of the parietal “field.”

The suspensorial rami of the parietal are horizontal in cross-section instead
of vertical as in Varanus and are covered ventrally by long flat extensions from
the supratemporal. In Plotosaurus, Tylosaurus (Huene, 1910, fig. 7) and Progna-
thodon a second posterolateral wing from the parietal covers the ventromedial
termination of the supratemporal, a condition that is incipient in Clidastes
(Camp, 1942, p. 35). In Varanus the suspensorial ramus of the parietal forms the
entire posteromedial surface of the suspensorial arcade, sheathing the supra-
temporal posteriorly and providing a major point of origin for the M. adductor
mandibulae externus profundus distally. If the latter muscle occupied the same
general area of the skull in mosasaurs as in Varanus it must have originated in-
stead largely from the flat ventral surface of the supratemporal. The suspensorial
ramus abuts against a short, anteromedially extending wing from the squamosal
in mosasaurs. As in Varanus, the M. spinalis capitis and dorsal portion of the
M. rectus capitis posterior probably inserted next to the cranial midline be-
tween the diverging bases of the suspensorial rami. The pars articuloparietalis
of the M. longissimus capitis, and fibers of the M. depressor mandibulae origi-
nated further laterally on the posterior surface of the rami.

In Platecarpus the parietal broadly invades the posterior dorsal surface of
the frontal medially and forms a relatively loose suture with that element. In
Clidastes and Tylosaurus this suture was probably less moveable, for the keels
and grooves on the sutural surfaces are finer and more tightly interlocking. The
suture meets the dorsal surface of the skull in the latter two genera in a nearly
straight transverse line, although the parietal may invade the frontal slightly on
either side of the dorsal cranial midline. In Mosasaurus, Plotosaurus, Progna-
thodon, Plesiotylosaurus and to a slight degree in Ectenosaurus the parietal is

Text-fig. 8. Dorsal surface of parietal of Mosasaurus maximus (NJSM 11052, x 1).

overlapped on either side of the midline anteriorly by posteriorly projecting
wings from the frontal. This overlapping must have inhibited movement about
the mesokinetic axis.

The parietal meets the medial wing from the postorbitofrontal loosely just
lateral to the center of the anterior wall of the supratemporal fenestra in Plate-

RUSSELL: AMERICAN MOSASAURS 29

carpus and Tylosaurus. In Clidastes and Mosasaurus the parietal has a much
broader contact with the medial wing from the postorbitofrontal and overlaps
its posterior surface extensively. The suture between these two bones lies in the
center of the anterior border of the supratemporal fenestra in Plotosaurus tuckert;
in Mosasaurus and Plotosaurus bennisoni it is more laterally situated. The suture
is located in the anteroexternal corner of the supratemporal fenestra in
Prognathodon as in Clidastes and the parietal probably also covers much of the
medial wing of the postorbitofrontal. The parietal meets the postorbitofrontal
in the anteromedial corner of the supratemporal fenestra in Plesiotylosaurus.

In Tylosaurus, Clidastes and Platecarpus there are two longitudinally-
grooved, knuckle-like projections that fit onto similar projections on the dorsal
apex of the supraoccipital, forming a sliding joint. In Mosasaurus these projec-
tions are apparently absent and a cartilaginous processus ascendens tecti synotici
presumably fit into a longitudinal sulcus along the ventral surface of the posterior
parietal midline. Though almost absent in Varanus, in all mosasaurs there are
flattened descending alae from the parietal forming part of the medial wall of
the supratemporal fenestra. These alae are crescentric in lateral profile, except
in Plotosaurus and Mosasaurus missouriensis (Goldfuss, 1845, pl. 7) where they
are especially large and fused to the prootic and supraoccipital.

As in Varanus, the cerebral hemispheres of Platecarpus were centered di-
rectly beneath the mesokinetic axis. Ventrally, on each side of the parietal
foramen, a ridge converges posteriorly, nearly meets its opposite across the
midline and then diverges to terminate at the condylar articulation for the
supraoccipital. A shallow sulcus, which must have roofed the optic lobes of
the mesencephalon, lies between this ridge and the medial surface of the brain-
case wall. In Clidastes propython and C. liodontus the region behind the parietal
foramen and between the converging ridges is flat or convex, but in Platecarpus
and Tylosaurus it is concave. In C. sternbergi the ridges extend in a straight line
back to the condylar articulations for the supraoccipital and the region between
them is flat. The two ridges fuse into a single median crest behind the ventral
opening for the parietal foramen in Tylosaurus.

OccIPITAL UNIT

The occipital unit is a key element in the structure of the mosasaur skull.
In it are included the basioccipital, basisphenoid, parasphenoid, prootics, opis-
thotic-exoccipitals, supraoccipital and supratemporals. The entire unit rotates
in a vertical longitudinal plane on the anterior end of the vertebral column.
This motion causes the quadratic suspensoria to turn relative to the posterior
ends of the supratemporal arcades (muzzle unit), the dorsal portion of the
supraoccipital to slide through the metakinetic joint and the anterior ascending
ramus of the prootics to slide between the descending flanges of the parietal unit.
Longitudinal motion is imparted to the pterygoid units ventrally via the
basipterygoid processes of the basisphenoid.

BASIOCCIPITAL

The basioccipital is a stout, triradiate bone in mosasaurs. Anteriorly, there
is a very heavy, ventrolaterally directed basal tuber on each side of the cranial
midline. Posteriorly, the main shaft of the bone expands laterally, then terminates
abruptly in a rounded crescent-shaped occipital condyle whose dorsolateral
corners extend out onto the exoccipitals. The dorsal surface of the basioccipital
is hollowed longitudinally to support the medulla from below. The short, heavy
wall bounding the medullary cavity meets the opisthotic-exoccipital above and the

30 PEABODY MUSEUM BULLETIN 23

prootic on an anteriorly inclined facet, through a flat cartilaginous surface
which is overlapped laterally by a tongue of the opisthotic firmly sutured to
the posterolateral corner of the basal tuber below. Anteriorly the basioccipital
abuts against the basisphenoid to which it is anchored by the interdigitating
contact of the anterior face of the basal tubera with the posteroexternal processes
from the ventral surface of the basisphenoid.

‘The basal tubera are very large in Mosasaurus, large and ventrally connected
by a transverse swelling of bone in Clidastes, intermediately developed in
Tylosaurus, and small and more anteromedially rotated in Platecarpus. ‘The
distal end of the tuber is pitted and ridged and was probably capped by a
meniscus of cartilage. The floor of the medullary cavity is smooth in Clidastes,
but in Platecarpus it is interrupted posteriorly by a large bilobate foramen
through which the basilar artery entered the basioccipital. The tunnel for this
artery curves ventrally and then anteriorly to pass into the basisphenoid. The
floor of the medullary cavity in Prognathodon seems to have been deeply
grooved, but there is no tunnel as such for the basilar artery. In Tylosaurus a
broad sheet of bone from the basisphenoid floors the medullary cavity over the
basioccipital all the way back to the foramen magnum. A slender “‘ear’’ of this
sheet extends laterally out over the anterior portion of the exoccipital suture.
The M. rectus capitus anterior must have inserted on the broad posterior face
of the basal tuber, and the M. ilio-costalis capitis together with the pars trans-
versalis cervicis of the M. longissimus capitis inserted on the roughened distal
surface of the tuber, as they doin Varanus.

BASISPHENOID

The basisphenoid is a square block of bone with a long parasphenoid rostrum
extending forward ventromedially from the dorsum sellae to floor the sella
turcica. Ventrally it is tetraradiate with a basipterygoid process projecting
anteroexternally from each side of the parasphenoid rostrum and a wide, ventrally
convex tongue of bone spread posteroexternally on either side of the midline

vel

Text-fig. 9. Lateral view of basioccipital-basisphenoid of Platecarpus (YPM 4025, x 1).

over the anteroventral face of the basal tuber of the basioccipital. Behind the
dorsum sellae the dorsal surface of the bone is shallowly concave and expanded
dorsolaterally into a heavy alar process which overhangs the lateral wall of the
basisphenoid below and is sutured to the basisphenoid ramus of the prootic
above. The basisphenoid meets the basioccipital through a vertically transverse
cartilaginous suture which is strengthened below by the interdigitating suture
between the ventroposterior tongues of the basisphenoid and the basal tubera.

RUSSELL: AMERICAN MOSASAURS 31

The lateral wall of the basisphenoid is smoothly sulcate inwardly from the
ventral wall of the alar process down over the thin sheet of bone covering the
vidian canal and out across the base of the basipterygoid process. The internal
jugular vein must have passed along this sulcus. The vidian canal transported
the internal carotid artery and the palatine branch of the seventh nerve into
the basisphenoid beneath the posterior end of the alar process and out again
onto the anterolateral surface of the parasphenoid rostrum. A small branch of
the internal carotid artery left the vidian canal and passed through the basi-
sphenoid medially to the posterior end of the sella turcica as in Varanus (see
Bahl, 1937, pp. 145-146). The basipterygoid process is wide and separated ante-
riorly by a small vertical notch from the parasphenoid rostrum. Its elliptical distal
surface trends posterolaterally and was capped by a cartilaginous meniscus con-
necting it to the sides of the notch between the basisphenoid and quadrate
processes of the pterygoid below. Fibers of the M. protractor pterygoid may have
originated on the dorsal surface of the basipterygoid process as they do in
Varanus.

The parasphenoid rostrum is long and laterally compressed, terminating
anteriorly in a vertical transverse surface for attachment of the presphenoid
cartilage. The parasphenoid arises beneath this surface. Two sharp crests extend
longitudinally over the entire dorsal length of the rostrum, enclosing a broad
medial groove between them. Externally, a narrower groove between these crests
and a ridge from the lateral wall of the vidian canal conducts the internal
carotid artery anteriorly into the ventral region of the orbit. The vertical anterior
wall of the alar processes is formed by the dorsum sellae which rises behind the

Text-fig. 10. Dorsal view of basioccipital-basisphenoid of Platecarpus (YPM 4025, x 1).

sella turcica and curves around it somewhat anterolaterally. The foramen for the
sixth nerve enters the anterolateral corner of the dorsal surface of the alar process
and emerges from the dorsolateral corner of the dorsum sellae. Between the
prootic above and the deep sulcus for the internal jugular vein below, the
vertical external surface of the alar process is approximately rectangular in
shape and is roughened, probably for the insertion of part of the M. protractor
pterygoid. The suture for the prootic is transversely horizontal and slopes pos-

32 PEABODY MUSEUM BULLETIN 23

teromedially to the posterior edge of the bone. The center of the suture is car-
tilaginous but this area is bounded internally by a dentate ridge and externally
by a posterolaterally descending dentate sutural area. There is a small “U”-
shaped trough located along the posterior half of the dorsal surface of the basi-
sphenoid.

The foregoing description is based on the basisphenoid of Clidastes. The
basisphenoid of the following genera is described only as it is known to differ
from the condition of the element in Clidastes.

1. Globidens. Only the dorsal half of the basisphenoid is known but this
portion of the bone, although more massive, is practically identical to that of
Clidastes. The dorsum sellae is short and heavy and the sixth nerve emerged
from a foramen near its ventrolateral corner. There is no ‘“U’’-shaped trough
on the posterior half of the dorsal surface of the basisphenoid.

2. Mosasaurus. The basisphenoid is generally similar to that of Clidastes.
However, the parasphenoid rostrum is less elongated, has a broader groove
along its dorsal midline, and is fused anteroventrally to the parasphenoid bone.
The lateral wall of the vidian canal may be incompletely developed.

3. Plotosaurus. Camp’s (1942, fig. 15) illustration of the lateral aspect of the
basisphenoid of Plotosaurus resembles that of Mosasaurus. The vertical external
face of the alar process is deeper and the base of the basipterygoid process
appears to be heavier.

4. Platecarpus. The basisphenoid is superficially like that of Clidastes but
differs greatly from this genus in detail. The vertical external face of the alar
process is larger and projects posteriorly to form the lateral wall of the vidian
canal. ‘The basipterygoid processes are wider and terminate anteriorly behind
the anterior end of the parasphenoid rostrum, The distal cartilage-capped
extremity of the process is larger, dorsoventrally compressed, and more strongly
posteriorly inclined than in Clidastes. The parasphenoid bone is fused to the
anteroventral end of the parasphenoid rostrum. A significant difference from

Text-fig. 11. Reconstructed basicranial circulation of Platecarpus.

the Mosasaurinae and Tylosaurinae lies in the penetration of the basisphenoid
posteriorly by a bilobate tunnel, which branches beneath the sella turcica,
each branch emerging through a smaller canal on either side of the parasphenoid
rostrum below and anterior to the anterior opening of the vidian canal. This
tunnel contained the basilar artery. Behind its anterior forking, two tiny fora-

RUSSELL: AMERICAN MOSASAURS 33

mina rise from the dorsal surface of the tunnel to the dorsal surface of the
bone behind the central portion of the dorsum sellae.

5. Tylosaurus. The basisphenoid is more elongated than in Clidastes. The
basipterygoid processes are similar to but somewhat narrower than those of
Platecarpus. The vertical external face of the alar process is very small and
overhangs a shallow groove between it and the vidian canal below, which
may have contained the internal jugular vein.

Two reconstructions of cranial circulation remain to be discussed. DeVillers
(1943) described in detail the braincase of Plioplatecarpus and his identifications
of the cranial nerve foramina agree with those given here for the very similar
braincase of Platecarpus. However, it appears that the anterior portion of the
basisphenoid of DeVillers’ Plioplatecarpus has been abraded and the basiptery-
goid processes as well as the lateral wall of the vidian canal have been lost.
His “‘gouttiére latérale inférieure du basisphenoide” is believed to be the vidian
canal and to have housed the palatine ramus of the facial nerve and the internal
carotid artery, not the internal jugular vein (vena capitis lateralis) which was
probably more laterally located along the concave external wall of the basi-
sphenoid.

A basisphenoid of Platecarpus in the Yale museum (YPM 4025) exhibits the
following differences from that illustrated by Camp (1942, figs. 19, 20): 1.)
There is no opening for the internal jugular vein on the lateral surface of the
basisphenoid dorsal to the basipterygoid processes. The internal jugular vein
is therefore assumed to have lain along the lateral wall of the braincase instead
of entering the basisphenoid and basioccipital along with the basilar artery.
2.) The channel for the internal carotid artery does not connect with that of
the basilar artery anywhere within the basisphenoid. 3.) The internal carotid
does not bifurcate into two major lateral branches below the sella turcica.
In the Yale specimen, Camp’s foramen for the ‘dorsal branch of the internal
carotid” is the outlet for the anterolateral branch of the basilar artery.

PARASPHENOID

The parasphenoid in Plotosaurus, Platecarpus (Camp, 1942, figs. 15, 19), and
Ectenosaurus is a long, dorsoventrally flattened, bayonet-like process of bone,
fused posteriorly to the center of the anteroventral edge of the basisphenoid. It
extends anteriorly, curves slightly dorsally, and ends in a point along the midline
of the skull ventral to the orbits. The parasphenoid is shorter and straighter in
Tylosaurus.

PROOTIC

As in Varanus the prootic is a triradiate bone in Clidastes. One ramus de-
scends anteriorly to meet the dorsolateral edge of the basisphenoid just behind
the sella turcica, forming a posteriorly inclined suture with a flat cartilaginous
center and internally and externally dentate borders. Another ramus of some-
what greater proportions ascends anteriorly to meet the parietal anterodorsally
in a broad cartilaginous inner, and a narrower digitate outer, suture. The ramus
is bounded dorsoposteriorly by a similar, but longer suture for the supraoccipital.
The exit for the fifth nerve is enclosed within the apex of a ““V’-shaped prootic
incisure between these two rami. A longer and narrower third ramus covers the
lateral wall of the suspensorium, projecting posteriorly, dorsally and externally
over the oposthotic, back to meet the supratemporal in a digitate suture that is
oriented at right angles to the main suspensorial axis. The prootic and opisthotic
meet internally in a flat cartilage-surfaced vertical suture that bisects the otic

34 PEABODY MUSEUM BULLETIN 23

Text-fig. 12. Lateral view of braincase of Clidastes propython (YPM 1368, x 1).

labyrinth transversely. Lateral to this the bones are more firmly interconnected
by longitudinal ridges and grooves lying in a plane coinciding with a vertical
surface drawn through the central axis of the suspensorium.

By analogy with Varanus, the M. protractor pterygoid must have originated
in part on the rectangular surface of the basisphenoid ramus of the prootic, be-
tween the ventral edge of the prootic incisure anteriorly and the otosphenoidal
crest posteriorly. On the ventroanterior portion of the parietal ramus and just
above the dorsal edge of the prootic incisure, there is a small ridge upon which
the M. levator pterygoid may have originated. It is likely that the M. pseudo-
temporalis profundus was attached dorsally to a gentle convexity that parallels
the axis of the parietal ramus, located above this ridge. In Clidastes the otosphe-
noidal crest has developed a thin, posteroventral-projecting ala which overhangs
the channel for the internal jugular vein from above (see Bahl, 1937, p. 151).
The seventh nerve emerged through a single opening from the dorsal wall of
the channel, medial to the central portion of the otosphenoidal crest, and
divided into the palatine and hyomandibular rami after reaching the lateral
surface of the prootic. In Varanus there are separate exits for these two branches
of the facial nerve. The tunnel transmitting the proximal end of the stapes to
the fenestra ovalis passes anteroventrally through the ventral portion of the
lateral prootic-opisthotic suture, above and behind the exit for the seventh nerve.
Behind the internal auditory meatus the ventral border of the suspensorium is
shallowly sulcate and probably housed the internal carotid artery (see Bahl,
1937, fig. 13) and internal jugular vein. Fibers of the pars posterior of the M.
adductor mandibulae externus profundus may have originated on the dorsal
edge of the prootic suspensorial ramus, but there is no indication of this on the
surface of the bone.

The lateral face of the prootic is vertical, though the internal side curves
dorsomedially, giving the bone the outline of an inverted triangle in cross-section.
The labyrinth occupys the upper portion of the triangle. Only the parietal and
basisphenoid rami of the prootic are visible on its longitudinally concave medial
surface. ‘The ventral suture slopes slightly posteriorly on the basisphenoid over

RUSSELL: AMERICAN MOSASAURS 35

the anterior two-thirds of its length, then rises at an angle of 45° to the vertical
posterior end of the bone, and abuts posteroventrally against the basioccipital.
A rounded pit located at the posterodorsal corner of the internal surface of the

Text-fig. 13. Medial view of braincase of Clidastes propython (YPM 1368, x 1).

prootic contains the exits for the undivided stem of the seventh nerve antero-
ventrally and those for the two branches of the eighth nerve dorsally. A slight
prominence descends anteroventrally from the top of the bone in front of this
pit, then curves posteriorly to merge with the inner wall of the prootic. The
anterior vertical semicircular canal intersects the cartilaginous surface of the
supraoccipital suture above and somewhat behind the posterior apex of the
prootic incisure. The posterior edge of this suture is strongly emarginated near
its center for the ascending sinus utriculus. Camp (1942, fig. 19) illustrates a
tiny foramen for the endolymphatic duct lying in front of and medial to this
emargination in Platecarpus.

The anterior part of the vestibular cavity itself is very similar to that of
Varanus (see Bahl, 1937, fig. 10). The posterior branch of the acoustic nerve
opens into the utriculus just above the cochlear crest and anterior to the prootic-
opisthotic suture. The anterior branch of this nerve enters the anterior ampul-
lary recess which contains the horizontal and anterior ampullae. The horizontal
and anterior vertical semicircular canals diverge from their respective ampullae,
the former laterally and the latter anteriorly. The cochlear lagena is divided
into equal halves by the prootic-opisthotic suture; it is conical in form with a
pointed ventral terminus. A slight ridge separates the internal auditory meatus
from the utricular cavity and marks the site of the fenestra ovalis.

The prootic of the following forms is described only as it is known to differ
from the condition of the element in Clidastes.

1. Clidastes sternbergi. This species exhibits the very peculiar feature of hav-
ing the medial entrance of the foramen for the seventh nerve located in the
center of the prootic incisure.

2. Mosasaurus. In lateral aspect, the prootic of Mosasaurus closely resembles
that of Clidastes, one comparatively minor difference being the heavier and
larger otosphenoidal crest in the former genus. Medially the pit enclosing the

36 PEABODY MUSEUM BULLETIN 23

foramina for the seventh and eighth nerves is larger and less well-defined. The
lagena seems to lie wholly within the prootic.

Text-fig. 14. Periotic labyrinth of Clidastes propython (YPM 1368, x 2). A. Anterior view of
opisthotic. B. Posterior view of prootic.

3. Plotosaurus. According to Camp’s (1942, pp. 5, 35, fig. 15) description and
figure of the lateral wall of the braincase, the parietal ramus of the prootic is
extensively overlapped by and firmly sutured to the descending wing of the
parietal, a condition very different from that found in Clidastes.

4. Platecarpus. Both the basisphenoidal and parietal rami of the prootic are
distally expanded and much larger than in Clidastes. The otosphenoidal ala
is usually small and poorly developed. The pit enclosing the foramina for the
seventh and eighth nerves is small and deep and borders directly on the prootic-
opisthotic suture. The medial wall of the prootic is anteroposteriorly shortened
and the prootic incisure contacts the ventroanteriorly descending prominence
lying in front of the above-mentioned pit; the incisure and pit are widely sepa-
rated in Clidastes. The labyrinth is, however, identical to that of Clidastes (see
Camp, 1942, fig. 25).

5. Tylosaurus. The most striking peculiarity of the lateral surface of the
prootic in this genus is the enormous development of the otosphenoidal ala
which broadly connects the basisphenoidal and suspensorial rami and covers
the whole ventroposterior face of the prootic, including even the anterodorsal

PEABODY MUSEUM BULLETIN 23 37

Text-fig. 15. Medial view of prootic of Tylosaurus nepaeolicus (AMNH 124, x 14), showing the
otosphenoidal crest and exit for cranial nerve VII.

portion of that of the opisthotic. The labyrinth is again very similar to that of
Clidastes, except that the anterior vertical semicircular canal is located much
closer to the sinus utriculus on the supraoccipital suture.

OPISTHOTIC-EXOCCIPITAL

The opisthotic and exoccipital are fused in Clidastes, as well as in mosasaurs
in general, and are considered here as one element. This ossification is composed
of a rectangular block anteriorly, lying on the basioccipital behind the prootic,
and a long distally expanded paroccipital process that supports the suspensorial
ramus of the prootic and the supratemporal posteromedially. Laterally the
anterodorsal quarter of the opisthotic-exoccipital block is overlapped by that
portion of the prootic that forms the medial wall of the groove for the internal
jugular vein and covers the proximal end of the internal auditory meatus. Part
of the articular surface of the occipital condyle extends onto the posteroventral
corner of the exoccipital region of the bone.

The exposed lateral surface of the opisthotic-exoccipital is triangular in
outline. The ventral margin of the paroccipital process, narrowly visible under
the prootic and supratemporal, rises less steeply posterodorsally from the
anterior edge of the triangle. The opening for the internal auditory meatus is
large and lens-shaped and situated beneath a shallow sulcus near the base of the
suspensorial process of the prootic. Lying ventroanteriorly to the internal
auditory meatus and separated from it by a rounded bar of bone is another

Text-fig. 16. Endocranial cast of Platecarpus (after Camp, 1942, fig. 24A, x 14, with slight modifi-
cations). X, cartilage between parietal and supraoccipital.

slightly smaller elliptical opening, whose long axis parallels the otosphenoidal
crest. This is the fenestra rotunda, which in life was covered over by a heavy

38 PEABODY MUSEUM BULLETIN 23

membrane that closed off the perilymphatic sac from the exterior (see Bahl,
1937, p. 140). Within the fenestra near its ventral corner is a tiny foramen
marking the exit of the ninth nerve. A rounded crest from the ventral edge
of the paroccipital process continues ventroanteriorly to form the posteroventral
border of the internal auditory meatus. It then descends vertically behind the
fenestra rotunda, forming the posterior boundary of a thin superficial tongue
of bone firmly sutured below to the posterolateral corner of the basal tuber
of the basioccipital. Beneath and behind this crest and paralleling its dorsal
margin are two lenticular foramina, separated from each other by a thin
sheet of bone. The larger dorsal foramen (jugular foramen) transmitted the
vagus and accessory nerves, the posterior cerebral branch of the internal jugular
vein, and the occipital branch of the internal carotid artery to the lateral sur-
face of the braincase. The smaller ventral foramen (condylar foramen) trans-
mitted the hypoglossal nerve to the exterior (see Bahl, 1937, p. 139). In Clidastes
these two foramina are located on the lateral face of the braincase, though in
Varanus the above-described crest is greatly enlarged and the single composite
foramen has been shifted around to the occipital face.

The internal surface of the opisthotic-exoccipital is longitudinally concave,
the cartilaginous sutures bounding it ventrally, anteriorly, and dorsally meeting
at nearly right angles. The jugular foramen, which enclosed the roots of the
tenth and eleventh nerves, is large and crescentic, and extends anteroventrally
across the anterior half of the medial wall. The foramen for the ninth nerve is
situated midway between the ventral end of the jugular foramen and the opis-
thotic-prootic suture. This foramen opens externally into the floor of the perilym-
phatic sac, meeting it at an angle of about 60° in the center of the braincase
wall. Two small anterior foramina and a larger posterior one lie in a horizontal
row beneath the jugular foramen. They conducted the three roots of the twelfth
nerve laterally into the single condylar foramen, which opens on the external
surface of the braincase. The suture for the supraoccipital is divided into two
portions of approximately equal area. Located in the medial half of the flat,
transversely vertical prootic contact is the anterointernal portion consisting of
a smooth “U’-shaped cartilaginous surface that has a deep notch anteriorly for
the sinus utriculus. The posterior vertical semicircular canal crosses this surface
near the center of its lateral margin. The posteroexternal portion is “V’’-shaped,

Text-fig. 17. Occipital view of skull of Platecarpus ictericus (AMNH 1820, X 14).

RUSSELL: AMERICAN MOSASAURS 39

strongly ridged and grooved longitudinally and encloses the cartilaginous sur-
face laterally and to a slight extent posteromedially. Its posterior apex is di-
rected out along a ventrally inclined ridge on the posterior face of the paroc-
cipital process.

In posterior view the paroccipital process appears as an elongated rectangular
bar of bone, supporting the supratemporal at its somewhat dilated distal end.
A ridge crosses the posterolateral corner of the supraoccipital and continues
ventrolaterally over the proximal part of the paroccipital process. The ventral
part of the M. rectus capitis posterior originates from the same region in Varanus,
though it lacks this ridge. Another crest descends ventrointernally from the
center of the lateral rim of the process and turns medially to parallel its ventral
margin for a short distance. The M. obliquus capitis magnus may have inserted

sc
RCPd

Text-fig. 18. Muscle insertions on occipital surface of skull in Platecarpus.

on this crest or in the shallow sulcus lying between it and the dorsal edge of the
process.

The posterior part of the vestibular cavity is also very close to that of Varanus.
The horizontal semicircular canal passes laterally through the opisthotic just
above the fenestra ovalis. Ventral to the posterior end of the horizontal canal,
in the floor of the utriculus, is the ampulla for the posterior vertical semicircular
canal. The perilymphatic duct leaves the lagena beneath its medial rim.

The opisthotic-exoccipital of the following genera is described only as it is
known to differ from the condition of the element in Clidastes.

1. Mosasaurus. ‘The fenestra rotunda is relatively smaller than in Clidastes
and the foramen for the ninth nerve is located in the anterodorsal corner of the
medial wall of the opisthotic instead of near the center of its anterior border.

2. Platecarpus (see Camp, 1942, figs. 19, 25). The jugular and condylar
foramina are fused into a single opening externally. On the internal wall of the
braincase the foramen for the ninth nerve lies in the ventroanterior corner of
the opisthotic and the three foramina for the twelfth nerve parallel the ventral
border of the jugular foramen instead of the horizontal axis of the bone. The
posterior vertical semicircular canal intersects the center of the anterior portion
of the cartilaginous supraoccipital suture much closer to the sinus utriculus
than in Clidastes.

3. Tylosaurus. The relationships between the foramina on the medial wall of

40 PEABODY MUSEUM BULLETIN 23

the braincase are like those of Clidastes, but the foramina are all confined to
the anterodorsal quarter of its surface. Laterally, the jugular and condylar
foramina are fused into a single opening. The labyrinth is like that of Clidastes.

SUPRAOCCIPITAL

In Clidastes the supraoccipital is a rectangular, roof-shaped element. A crest
rises upward anteriorly from the foramen magnum to the parietal and forms the
“ridge” of the roof. From the posterolateral corner of the bone another crest
extends dorsally across the sloping side of the roof to a longitudinally grooved,
knuckle-like process near the midline of the bone. This process articulates with
a similar projection near the posteroventral midline of the parietal. The latter
crest probably provided an area for the insertion of ventral fibers of the M. rectus
capitis posterior. The ventral surface of the supraoccipital is hollowed to fit over
the posterior part of the brain stem, probably originally separating the medulla
oblongata from the cerebellum (see Camp, 1942, fig. 24A and B). The contacts
with the prootic and opisthotic-exoccipital are the mirror images of the sutures
as described above under these elements. The supraoccipital caps the otic capsule
on each side of the braincase and contains the upper portions of the anterior
and posterior semicircular canals and sinus utriculus (see Camp, 1942, fig. 25).

Text-fig. 19. Lateral aspect of supraoccipital of Platecarpus (YPM 1488, x 1).

The anterior edge of the lateral wall of the supraoccipital is truncated by a
transversely vertical surface. Posteriorly the bone forms the dorsal border of the
foramen magnum.

The supraoccipital of Platecarpus is essentially the same as that of Clidastes.

SUPRATEMPORAL

The supratemporal is a rectangular chip of bone supported posteromedially
by the strong distally expanded suspensorial process of the opisthotic and loosely
linked to the prootic anteromedially by a digitate suture. Its external face is
faceted to receive the dorsolateral edge of the quadrate. Two processes arise
from the dorsal margin of the bone; one extends anteriorly becoming sutured
to the internal face of the squamosal and the other curves medially to pass
under the suspensorial ramus of the parietal. In Clidastes, Platecarpus and
Tylosaurus there is a gentle convexity with its long axis trending dorsoanteriorly
over the sutural surface of the first process. This convexity fits into a similar-
shaped concavity in the medial face of the squamosal, forming a structure that
would resist rotation of the squamosal on the supratemporal. The convexity
has developed into a strong crest in Mosasaurus maximus. The second process
is long and flat and covers most of the dorsolateral surface of the posttemporal
fenestra, probably providing a large part of the area of origin of the M. adductor

RUSSELL: AMERICAN MOSASAURS 4]

mandibulae externus profundus. In Plotosaurus, Tylosaurus and Prognathodon
it is overlapped medially by a more ventral tongue from the parietal.

Posteriorly the supratemporal is poorly exposed behind the paroccipital
process and under the suspensorial ramus of the parietal in Clidastes, Mosasaurus,
Platecarpus and Tylosaurus. It seems quite likely that the pars transversalis
capitis of the M. longissimus capitis originated on the vertical notch lying be-
tween the distal end of the paroccipital process and the supratemporal and that
fibers of the pars ar ticulopar ietalis of the M. longissimus capitis originated only
along the posterior rim of the parietal, as they do in Varanus. The M. depressor
mandibulae may have lacked an origin on the supratemporal in mosasaurs, al-
though this is a major site of attachment of this muscle in Varanus.

PTERYGOID UNITs

The pterygoid units are composed of the pterygoids and ectopterygoids.
They link the ventromedial corners of the quadrates (quadrate units), the
basipterygoid processes (occipital unit), and the base of the epipterygoids
(epipterygoid units) with a long bar of bone to brace the jugal and palatines
(muzzle unit) anteriorly. Through these bones longitudinal motion from the
ventral portion of the occipital unit is transmitted to the posteroventral edge
of the muzzle unit.

Text-fig. 20. Left quadratic suspensorium of Platecarpus (AMNH 1820, X 1).

PTERYGOID

The pterygoid in mosasaurs may perhaps be best considered as composed of
a central, anteroposterior body with two large diverging processes, one antero-
laterally to the ectopterygoid and another posterolaterally to the medial edge of
the quadrate. The pterygoids do not meet along the midline of the skull, al-
though the incisura piriformis is narrower than in Varanus.

The main body of the pterygoid is a short, dorsoventrally flattened column,
terminating posteriorly in a tongue-shaped wing (basisphenoid process) which
lies medial to the basipterygoid process of the basisphenoid. On its ventral sur-

PEABODY MUSEUM BULLETIN 23

42

pe

c

EES asians gasapn@npeaaeney

er (AMNH 4909, x 4/).

leg
i)

1

Text-fig. 21. Ventral view of left pterygoid of Tylosaurus pror:

RUSSELL: AMERICAN MOSASAURS 43

face is a row of teeth commonly bent in a gently recurving line extending
from the medical base of the basisphenoid process to the anteroexternal termina-
tion of the bone (Tylosaurus, Platecarpus, Prognathodon and Plotosaurus), but
these teeth may be oriented in a straight anteroposterior line (Clidastes). The
pterygoid teeth are usually smaller, more strongly recurved, and more nearly
circular in horizontal section than those of the mandibles (Clidastes, Mosasaurus,
Plotosaurus, Platecarpus, Ectenosaurus and Tylosaurus). In Prognathodon and
probably also Plesiotylosaurus, however, the pterygoid teeth are not subequal in
size, as in the foregoing genera, but increase anteriorly from a very small diameter
to equal the mandibular teeth in size. The number of pterygoid teeth varies
from a minimum of seven in Prognathodon to a maximum of fifteen in Ploto-
Saurus.

Although the pterygoid must have been firmly buttressed against the palatine,
this contact is very rarely preserved. In Platecarpus the main body of the ptery-
goid abuts anteriorly onto the medial half of the posterior edge of the palatine
and sends a long process anteriorly to cover the ventromedial surface of this bone.
In ventral aspect the suture runs from the external side of the pterygo-palatine
bar anteromedially in a nearly straight line to meet the internal side of the bar
just behind the posterior boundary of the internal naris. Sutural indications on
the ventral surface of the palatine in Tylosaurus and Clidastes suggest that the
pterygo-palatine contact is probably also similar in these two genera. In Ploto-
saurus the suture is the same as in Platecarpus, except that it is more medially
directed and reaches the medial edge of the pterygo-palatine bar far behind the
internal narial opening.

The ectopterygoidal process projects from the main body of the pterygoid
at a right angle in Clidastes, Mosasaurus and Tylosaurus and at an acute angle
anteriorly in Platecarpus. Ventrally the process springs from a broad base in
common with the quadratic ramus in Clidastes, Plotosaurus and Tylosaurus and
from a relatively narrow base, which is more distinctly separated from that of
the quadratic ramus, in Platecarpus and perhaps also in Mosasaurus. It passes
smoothly into the superior surface of the main body of the pterygoid in Plate-
carpus and Tylosaurus, but in Clidastes it curves abruptly anteriorly to form a

Text-fig. 22. Dorsal view of right pterygoid of: A. Platecarpus ictericus (reconstructed after
AMNH 1820, 5811, x 34). B. Clidastes propython (AMNH ND2, x %

Jo] *

44 PEABODY MUSEUM BULLETIN 23

long anteroposterior crest which is in turn smoothly continuous with the medial
face of the pterygoid. In Platecarpus two distinct horizontal facets sweep around
the posterolateral corner of the base of the ectopterygoidal process. Although
less clearly defined, they continue posteriorly onto the lateral and ventrolateral
edge of the quadratic ramus. The dorsal of these facets in Platecarpus is prob-
ably homologous with a lunate sulcus on the posterodorsal face of the ectoptery-
goidal process in Clidastes, Mosasaurus, and Prognathodon and perhaps also
with a shallow groove that extends anterolaterally across the dorsal face of this
process in Tylosaurus. The dorsal facet is here interpreted as the point of origin
of the M. pterygoideus profundus and the ventral facet as that of the M. ptery-
goideus superficialis. These facets are not so clearly apparent on the pterygoid of
Varanus. The ectopterygoidal process in mosasaurs is short and dorsoventrally
flattened. It ends laterally in an expanded termination that is beveled off ven-
troexternally to form a loose, undulating suture with the ectopterygoid in
Clidastes and Platecarpus. The ectopterygoidal process is broader in Tylosaurus.
The distal suture for the ectopterygoid is more anteriorly situated and probably
made a firmer connection with the latter element. In all three of these genera
there is a tuberosity on the ventroposterior portion of the sutural area that
would have prevented the ectopterygoid from slipping backwards.

‘The quadratic ramus arises from the base of the ectopterygoidal process and
the side of the main body of the pterygoid and extends posterolaterally to
contact the medial edge of the shaft of the quadrate. The basipterygoid process
of the basisphenoid is enclosed in a “U’-shaped notch between the quadratic
ramus and the much shorter basisphenoid process. Behind the anterior apex of
this notch a thin wing of bone curves ventrally and medially off the base of
the lateral axial portion of the quadratic ramus and expands posteromedially
to give the ramus a feather-shaped appearance when viewed from below. This
lamina is not present in Varanus and its function here is speculative. Perhaps it
served as an area of attachment for fibers of the M. pterygoideus superficialis
below or was partly invaded by the M. protractor pterygoid above. There is a
groove along the dorsal edge of the axis of the quadratic ramus in Clidastes and
Tylosaurus that probably marked the insertion of the M. protractor pterygoid.
In Platecarpus this groove is absent and the muscle may have inserted somewhat
medially onto a keel along the top of the axis of the ramus. In Platecarpus and
Tylosaurus the quadratic ramus is relatively long and slender; in Clidastes it is
much shorter and broader. There is a slight swelling for articulation with the
quadrate at the posterolateral tip of the ramus in all three genera.

The epipterygoid articulates with the dorsal surface of the pterygoid in a
rounded excavation that is continuous ventrally with, and immediately anterior
to, the notch for the basipterygoid process of the basisphenoid in Clidastes,
Mosasaurus, Platecarpus, Prognathodon and Tylosaurus. This pit is more lat-
erally compressed in Platecarpus and Tylosaurus. In Clidastes its anterior border
is not expanded enough posterolaterally to meet the anterior margin of the M.
pterygoideus profundus facet, as it does in the other four genera. In Varanus the
M. levator pterygoid inserts on the dorsal surface of the pterygoid just medial
to the epipterygoidal pit. No indication of a scar for the muscle has been noted
in the corresponding region in Clidastes and Platecarpus, although it is probably
represented by a shallow anteroposterior sulcus in this position in Tylosaurus.

ECTOPTERYGOID

The ectopterygoid is an “L’’-shaped bone, with one arm of the “L” extend-
ing laterally from the pterygoid to meet the jugal and the other more slender

RUSSELL: AMERICAN MOSASAURS 45

arm following the medial face of the jugal forward to end in a point approaching
or beneath the posterior tip of the maxilla. Postero-laterally and -ventrally its
suture with the ectopterygoid process of the pterygoid is ribbed and grooved;
the face abutting against the jugal is smooth and vertical.

The longitudinal arm of the ectopterygoid is short in Platecarpus and longer
in Tylosaurus (Williston, 1898b, pl. 18) and Plotosawrus where it meets the
posterior rim of the palatine. The jugal contact is smooth in Plotosaurus and
Platecarpus, but the posterior portion of the lateral face of the bone is medially
offset from the anterior portion in Tylosaurus. A notch may be present or absent
in the anteromedian corner of the “L” in Platecarpus. Such a notch is apparently
lacking in Plotosaurus, Prognathodon and Tylosaurus. Only a slight trace of the
transverse process of the bone is present in Plotosaurus.

EPIPTERYGOID UNITS

The epipterygoid units include only the epipterygoids, slender rods of bone
linking the center of the pterygoid units to connective tissue between the descend-
ing flanges of the parietal (parietal unit) and anterior ascending ramus of the
prootics (occipital unit).

EPIPTERYGOID

In a specimen of Platecarpus described by Marsh (YPM 1277, Marsh,

1872b, p. 450) the epipterygoid is a simple dowel of bone with a small, rounded,

ventral termination fitting into a shallow pocket on the dorsal surface of the
pterygoid, anterior to the notch for the basipterygoid process. The bone swells

A

Text-fig. 23. A. Stapes. B. Epipterygoid. Platecarpus ictericus (YPM 1277, x 1). The dorsal end
of the epipterygoid is to the left.

slightly above its ventral articulation, then constricts, and finally re-expands. It
is flattened dorsally, presumably lying on the lateral face of the parietal ramus
of the prootic. In Varanus the epipterygoid both separates and provides areas of
attachment for fibers of the pars anterior of the M. pseudotemporalis profundus
laterally and the pars posterior medially.

Camp (1942, p. 30, fig. 15) describes the epipterygoid of Plotosauwrus as being
slender and more spirally curved than in Platecarpus. Its dorsal end is expanded
into a thin blade, and a narrow groove in front of the anterior border of the
descending process of the parietal evidently supported the posterior edge of this
portion of the bone. Its anteromedian edge was supported in part by an ascending
process of the parietal.

The epipterygoid has been reported in Tylosawrus by Merriam (1894, p. 21)
and Huene (1910, p. 301) and is described by the latter author as “Komprimiert,
leicht S-formig geschweift, oben verschmélert, unten mit verdickter Gelenck-
flache, die in eine vertiefte Stelle des Pterygoides eingreift.”

STAPES UNITS

Properly speaking the stapes is a separate structural unit within the mosasaur
skull, It had no function other than that of sound transmission.

46 PEABODY MUSEUM BULLETIN 23

STAPES

The stapes in the Platecarpus specimen studied by Marsh (YPM no. 1277,
Marsh, 1872b, p. 449; see also Baur, 1892, p. 12; Huene, 1911, p. 50) is a long
delicate rod of bone, expanded at both ends and narrow in the middle, and is
very similar to the stapes of Varanus (Bahl, 1937, fig. 12). In Plotosaurus the
stapes is columnar and much more heavily constructed (Camp, 1942, figs. 7, 23).

“The extracolumella of mosasaurs is known only in the European Plioplate-
carpus (Dollo, 1905); it is ossified and fused with [a calcified] tympanic oper-
culum. . . In this genus the extracolumella is attached to the operculum by a
Y-shaped extremity, the arms of which may be homologized with the pars
superior and pars inferior of Varanus (Versluys, 1858, p. 65 and fig. 56). As
Dollo (1888) has elsewhere shown, there is also an ossified quadratic process of
the extracolumella resting in the ‘suprastapedial fossette’ [= ‘stapedial pit’] of
the quadrate in Plioplatecarpus. This process may be identified with the processus
internus of the extracolumella as figured by Versluys.”

“. ,. there is a point, however, at which mosasaurs may agree more closely
with Varanus than with other known lizards. The joint between the extra-
columella and the stapes is set at a high angle, the stapes lying at about forty-five
degrees to the body axis and the extracolumella at a right angle to it. The same
condition must have obtained in mosasaurs, to allow the extracolumella to
penetrate the narrow suprastapedial fossa.” (Camp, 1942, p. 34)

QUADRATE UNITS

The quadrate units (quadrates) articulate above with the quadratic suspen-
soria of the occipital unit and below with the posterior mandibular unit, the
former joint being the less perfect of the two. The posteromedial corner of the
quadrate was loosely bound by ligaments to the posterior tip of the pterygoid
units.

QUADRATE

The quadrate in mosasaurs is a heavy, powerfully constructed bone and is
the most frequently preserved cranial element. Its main body is a vertical shaft,

A c

Text-fig. 24. Posterior view of quadrate. A. Mosasaurus maximus (YPM 430, X 54g). B. Plate-
carpus ictericus (AMNH 1820, x 56). C. Tylosaurus proriger (reconstructed after AMNH
4909 and YPM 3990, x 56).

RUSSELL: AMERICAN MOSASAURS 47

articulating above with a concavity on the lateral face of the suspensorium
(composed of the supratemporal and squamosal) and ending below in a convex
longitudinal condyle for articulation with the mandible. The large suprastapedial
process arises from the posterodorsal surface of the quadrate shaft, curving
ventrally to enclose the stapes in a notch and then ventroanteriorly towards the
smaller infrastapedial process. A large crescentric tympanic ala projects out from
the anteroexternal edge of the quadrate shaft.

The anterior surface of the quadrate is in the form of an upright rectangle.
Its medial boundary is straight in Mosasaurus, Plotosaurus, Platecarpus, Progna-
thodon and Tylosaurus, but is concave laterally in Clidastes. There is a lenticular
rugosity centered above the midline of this boundary for the origin of the M.
adductor posterior in Clidastes, Platecarpus and Prognathodon. In Tylosaurus
the muscle probably originated on an anterointernally directed facet located
more ventrally on the medial boundary of the quadrate. In Mosasaurus maximus
a small triangular medially directed crest in the same region may have served
as the area of origin for this muscle. The muscle seems to have been about as
well-developed in mosasaurs as it is in Varanus.

A large, shallow, triangular depression, whose base extends across the entire
ventral edge of the anterior face of the quadrate, curving around an anterior
projection of the articulating surface to reach dorsally and somewhat medially
to the horizontal midline of the bone, probably marks the ventral origin of the
quadrate head of the M. adductor mandibulae externus profundus. A slight

Text-fig. 25. Left quadrate of Platecarpus ictericus (AMNH 1820, x 14). A. Lateral aspect. B.
Anterior aspect showing outlines of muscle attachments. C. Medial aspect.

rugosity is present at the dorsal apex of this triangular concavity in Clidastes,
Mosasaurus and Plotosaurus which is lacking in Platecarpus, Plioplatecarpus,
Prognathodon and Tylosaurus. Just beneath and lateral to this apex is a small
foramen which passes through the quadrate between the main vertical shaft and
the tympanic ala in Clidastes, Mosasaurus, Platecarpus and Tylosaurus. On the
upper third of the anterior face of the quadrate in all seven of these genera
there is a smaller shallow depression, in the form of an inverted triangle, located
between the quadrate shaft and the tympanic ala. It is probable that this region
provided an area of origin for dorsal fibers of the M. adductor mandibulae ex-
ternus profundus, since this muscle originates over nearly the entire rostral sur-
face of the quadrate in Varanus. The dorsolateral one-third to one-half of the
anterior face of the quadrate is made up of the tympanic ala proper.

The suprastapedial process is large in all mosasaurs, curving ventrally to

48 PEABODY MUSEUM BULLETIN 23

enclose the stapes dorsally and posteriorly in a rounded notch. The main axis
of the process is more medially directed than that of the shaft of the quadrate.
The suprastapedial process is largest in members of the Plioplatecarpinae, inter-
mediate in Tylosawrus, and smallest in the Mosasaurinae. The sides of the proc-
ess are parallel and somewhat dilated distally in Platecarpus, Halisaurus on-
chognathus (Merriam, 1894, p. 37) and Prognathodon overtoni, are constricted
medially in Clidastes, Mosasaurus and Plotosaurus, and strongly constricted
medially in Prognathodon rapax. The supra- and infrastapedial processes are
separate in Clidastes propython, Mosasaurus, Plotosaurus, Platecarpus, Plio-
platecarpus, Halisaurus onchognathus (Merriam, 1894, p. 37) and Tylosaurus,
but are fused, closing the stapedial notch posteriorly and ventrally in Ecteno-
saurus, Prognathodon and Plesiotylosaurus. Some of the fibers of the M. depres-
sor mandibulae must have originated along the distal surface of the suprastape-
dial process, particularly in a circular depression near its end in Clidastes, Mosa-
saurus, Plotosaurus, Platecarpus, Ectenosaurus and Prognathodon and in a
similarly placed, longitudinally elliptical depression in Tylosaurus. In Plioplate-
carpus a keel between the ventral anterior border of the stapedial notch and the
overhanging suprastapedial process projects posteriorly from the quadrate shaft
and nearly closes off the notch from below. This keel is only incipient in Plate-
carpus and is absent in all the other known genera of mosasaurs.

The infrastapedial process is large and triangular in posterior outline with
a bluntly pointed dorsal termination in Platecarpus, is similar but much larger
in Prognathodon, Plesiotylosaurus and Ectenosaurus and covers most of the
posterior base of the quadrate in these genera. The infrastapedial process is
smaller in Tylosaurus and Clidastes propython where it is more clearly separated
from the ventral rim for the tympanic membrane. It lies opposite and widely
separated from this rim low on the posterior shaft of the quadrate in Tylosaurus
nepaeolicus, close to the rim and higher on the shaft in C. propython, and above
and distinctly separated from the rim in the center of the shaft in T. proriger.
In Mosasaurus and Plotosaurus the infrastapedial process is a small rounded
tuberosity situated in the center of the posterior shaft of the quadrate. It is
absent in C. liodontus and C. sternbergi.

A deep pit for the processus internus of the extracolumella (stapedial pit)
lies in front of and slightly above the top of the stapedial notch on the medial
surface of the quadrate. The stapedial pit is elliptical in Clidastes, Mosasaurus,
Plotosaurus and Platecarpus, reniform in Plioplatecarpus, circular in Progna-
thodon, and rectangular in Tylosaurus. Below the stapedial pit the medial face
of the quadrate shaft possesses a heavy vertical crest dorsally and is flat ventrally
in Platecarpus and Plioplatecarpus, is flat in Tylosaurus, is slightly concave in
Clidastes, has a deep sulcus between the infrastapedial process and the antero-
medial edge of the shaft dorsally and is flat ventrally in Mosasawrus and Ploto-
saurus, and has a shallow vertical sulcus that reaches ventrally almost to the
condyle in Prognathodon. The medial surface of the quadrate is always smooth
and flat just above the articular surface for the lower jaw and must have been
connected ligamentously with the tip of the quadratic ramus of the pterygoid.
In front of this flattened surface and beneath the area of origin of the M. ad-
ductor posterior there is a heavy ridge that descends vertically to the condyle in
Mosasaurus, Plotosaurus, Platecarpus, Prognathodon, Plesiotylosaurus and Tylo-
saurus. In Clidastes it projects below the main part of the quadratomandibular
joint. Possibly this ridge served as a point of attachment for a separate bundle
of fibers derived from the M. protractor pterygoid (an M. protractor quadrati).

RUSSELL: AMERICAN MOSASAURS 49

Fibers of this muscle do insert on the quadrate of Varanus niloticus (Lakjer,
1926, p. 14).

A delicate crest of bone curves up from the distal end of the external sur-
face of the suprastapedial process over the dorsal part of the main shaft, projects
anteroexternally from the shaft a short distance and then descends in a sweeping
circle back to the ventral portion of the quadrate shaft, rising again before
terminating near (Clidastes, Mosasaurus, Plotosaurus, Tylosaurus) or fusing with
the infrastapedial process (Platecarpus, Prognathodon, Ectenosaurus, Plioplate-
carpus (Dollo, 1904, pl. vi). This crest is grooved along its anterolateral rim in
Mosasaurus and Plotosaurus. The crest is supported anteriorly by a wing of
bone (tympanic ala) from the anterolateral corner of the quadrate shaft. This
ala is thin and distinctly separated from the shaft of the quadrate in Mosasaurus,
Plotosaurus, Platecarpus, Plioplatecarpus, Ectenosaurus, Halisaurus onchogna-
thus (Merriam, 1894, p. 37) and Tylosaurus proriger. It is much heavier and
merges imperceptibly with the quadrate shaft in Clidastes and T. nepacolicus.
A calcified tympanum has been observed in all three of the common Niobrara
genera (Clidastes, Platecarpus, Tylosaurus) and in Ectenosaurus from the same
formation.

Dorsally, the quadrate shaft, the suprastapedial process, and the crest of the
tympanic ala form a sharply-defined smoothly continuous surface that looks as
if it had been coated with a thin veneer of cartilage in life. The transverse-
oriented condyle on the lower terminus of the quadrate shaft is convex ventrally
in most mosasaurs (Mosasaurus, Plotosaurus, Platecarpus, Ectenosaurus, Progna-
thodon, Plesiotylosaurus) but is flat in Clidastes.

ANTERIOR LOWER JAW

The anterior segment of the lower jaw is a distinct structural unit formed
by the dentary and splenial. It is supported posteriorly by a well-defined joint
between the splenial and angular on the ventral margin of the jaw and a blade-
like extension of the prearticular into the mandibular foramen dorsal to the
splenio-angular articulation. The anterior ends of these segments were liga-
mentously united.

DENTARY

The dentary in mosasaurs is an anteriorly narrowing girder of bone that ends
rather bluntly a short distance in front of the anteriormost dentary tooth.
Usually the longitudinal axis of the bone is nearly straight, as are its dorsal and
ventral margins. The lateral surface of the dentary is gently convex and bears a
dorsal row of foramina conducting terminal branches of the mandibular divi-
sion of the fifth nerve to the exterior, This dorsal row is joined anteriorly by
a second more ventral row along the anterior third of the dentary. A deep
groove excavated in the medial surface of the dentary opens posteriorly to form
the mandibular channel. The thin anterior blade of the prearticular enters this

Text-fig. 26. Left anterior mandibular unit of Clidastes liodontus (YPM 1335, x 1).

50 PEABODY MUSEUM BULLETIN 23

channel from the posterior segment of the lower jaw and extends anteriorly be-
tween the dentary and the splenial. In Varanus a ligament from the M. angulari
oris slides over the coronoid to insert on the tissue at the posterior end of the
gum line on the dentary. In mosasaurs the dorsal surface of the dentary in this
region is longitudinally grooved and may have marked the insertion of a similar
tendon serving to adduct the anterior segment of the lower jaw. The segment
may have been connected by an elastic ligament to the similarly grooved anterior
tip of the coronoid.

The dentary is relatively powerfully built in Mosasauwrus, Prognathodon and
Plesiotylosaurus, is of average proportions in Plotosaurus, Platecarpus and
Tylosaurus, and is more delicately constructed in Clidastes. The edentulous
anterior terminus of the dentary is long and rectangular in outline in Tylosaurus,
smaller in Clidastes, Mosasaurus missouriensis, M. maximus and Plotosaurus,

Text-fig. 27. Anterior tip of dentary of Tylosaurus proriger (AMNH 4909, x 14).

and is absent or nearly absent in M. ivoensis, Platecarpus, Plioplatecarpus and
Prognathodon. In C. propython, C. liodontus, M. missouriensis, M. tvoensis,
Platecarpus and Tylosaurus the longitudinal axis of the dentary is straight; in
C. conodon, M. maximus, Prognathodon and Plesiotylosaurus the axis is slightly
concave upwards. In Plotosaurus the longitudinal axis of the dentary is convex
upwards. Although unobserved in other mosasaurs, in a single fragment of a
dentary referred to Globidens (SDSM no. 4612) there is a deep notch extending
from the dorsoposterior edge of the bone anteroventrally to a point beneath the
penultimate dentary tooth. There is no bony mandibular symphysis known in
mosasaurs. The anteromedial tips of the dentary are roughened for the attach-
ment of an intermandibular ligament.

SPLENIAL

The main body of the splenial is a long, laterally compressed rod of bone
that narrows, is increasingly overlapped anteriorly by the dentary, and finally
disappears from lateral view behind the posterior third of that bone. A broad
thin ala arises from the center of the medial surface of the splenial and sheaths
the mandibular channel and much of the internal surface of the anterior mandi-
bular segment. ‘The ventral margin of this wing rises off the lower border of the
jaw in front of the anterior end of the main body of the splenial. It ascends
to meet the dorsal margin of the wing, which parallels the gum line, in an acute
angle along the anteromedial one-third of the dentary (into the symphyseal
region in Plesiotylosaurus fide Camp, 1942, p. 19). A foramen at the posteroven-
tral base of the ala transmits the inferior alveolar nerve into the mandibular
channel (see Bahl, 1957, p. 167). The posterior-facing surface for articulation
with the angular is located behind the posterior end of the dentary at the ex-
treme posteroventral corner of the anterior segment of the lower jaw. In outline
the surface resembles a vertical ellipse, is concave, but has a perpendicular keel
extending from near the center of the ellipse to its dorsal margin.

‘This perpendicular keel lies just medial to the center of the articulating sur-

RUSSELL: AMERICAN MOSASAURS 51

Text-fig. 28. A. Lateral aspect of splenio-angular articulation of Platecarpus (YPM 1491, x 14). B.
Anterior view of angular articulation. C. Posterior view of splenial articulation.

face in Platecarpus and Tylosaurus where it is of average dimensions. It is cen-
tered in Clidastes where it is nearly absent, in Mosasaurus missouriensis where
it is small, and in Globidens where it is large. The keel lies lateral to the center
of the articulating surface in M. maximus where it is large and in Prognathodon
where it is small. In posterior outline the articulating surface is circular in M.
ivoensis and Ectenosaurus, elliptical in Clidastes, M. missouriensis, M. maximus,
Platecarpus and Tylosaurus, and laterally compressed in Prognathodon. A
rounded prominence surrounding the ventral portion of the splenial in front
of the splenio-angular joint may have marked the point of insertion of fibers of
the M. cervicomandibularis, which would then have served to abduct the ante-
rior segment of the lower jaw.

PosTERIOR LOWER JAW

The posterior segment of the lower jaw includes the angular, surangular,
coronoid and prearticular-articular. It articulates with the quadrate posteriorly
and with the splenial joint (anterior unit of the lower jaw) anteroventrally. Both
mandibular segments were firmly connected dorsally by a tongue of the pre-
articular bound into the mandibular foramen of the anterior segment.

ANGULAR

The main body of the angular is shaped like a laterally compressed cone,
the anterior face of which is rounded to fit into a cup-shaped excavation in
the posterior end of the splenial. This articulating surface is in the form of a
rounded “V” with a median dorsal sulcus that receives a ridge from the dorso-
medial portion of the articulating surface on the splenial. A short heavy wing
arises laterally from the ventral margin of the posterior-directed apex of the
cone and is overlapped above by the surangular. Another thinner and broader
wing from the angular extends up the medial side of the posterior segment of
the lower jaw and covers much of the ventrointernal face of the prearticular.
The prearticular sits in the groove formed between these two alae. Posteriorly
the bone usually narrows to end in a point beneath the glenoid fossa. ‘There
is a foramen or notch in the anterior base of the medial wing of the angular
for the angular branch of the mandibular nerve.

The angular is very constant in shape among the different forms of mosasaurs.

52 PEABODY MUSEUM BULLETIN 23

Its articular face is nearly circular in Ectenosaurus, though in Clidastes, Mosa-
saurus, Platecarpus, Halisaurus platyspondylus, Prognathodon and Tylosaurus
it is more laterally compressed. The internal wing of the splenial is small and
widely separated from the coronoid in Platecarpus and Tylosaurus. It is larger
in Prognathodon, but is overlapped above by a descending wing of the coronoid.
In Clidastes the internal wing of the angular is very well developed and may
contact the coronoid dorsally. The angular of Plotosaurus spreads posteriorly
over much of the ventromedial surface of the articular. The angular branch of
the mandibular nerve emerged through a foramen in Clidastes, Mosasaurus, and
Halisaurus platyspondylus, and through either a foramen or a notch in Tylo-
saurus and Platecarpus (?).

SURANGULAR

The surangular is essentially an elongated rectangular piece of bone with
a tongue projecting posterodorsally over the lateral surface of the articular to
end in a rounded point, usually near the posteroexternal corner of the glenoid
fossa. The dorsal surface of the tongue is convex from side to side and forms
the lateral boundary of the glenoid fossa. This part of the articulating surface
widens anteriorly, finally curving medially behind a transverse ridge of bone
that marks the anterior limit of the fossa. The lateral exit for the cutaneous

Text-fig. 29. Medial view of mandible of Platecarpus ictericus (AMNH 1821, x %).

branch of the mandibular nerve lies under the anteroexternal corner of the
glenoid fossa, Beneath this foramen a rounded crest of bone descends ventroante-
riorly, curving to parallel the horizontal axis of the surangular anteriorly and fi-

nally merging with the lateral surface of the bone below the posterior boundary
of the coronoid suture. The pars superficialis and medialis of the M. adductor
mandibulae externus insert along an identical crest in Varanus. Above this
crest the surangular becomes laterally compressed and rises to form a vertical
posterolateral buttress for the coronoid. The coronoid suture proper is a
smoothly rounded, horizontal shoulder lying medial to and partly in front of
the coronoid buttress. A branch of the mandibular nerve passes anteroexternally
through the shoulder and emerges from a large foramen on the anterior laterally
exposed portion of the coronoid suture. The meckelian fossa is located under
the posterior half of the coronoid suture and is walled dorsally and laterally
by the surangular and medially by the prearticular.

‘The posterodorsal tongue of the surangular terminates at the posteroexternal
corner of the glenoid fossa in Clidastes, M. missouriensis, Platecarpus, Ecteno-
SQUTUS, Prognathodon and Tylosaurus. In M. maximus and Plotosaurus it ex-
tends behind the fossa along the dorsal rim of the articular and covers the
dorsolateral surface of the bone. The coronoid buttress is deep and very thin,
especially dorsally, in Clidastes, Mosasaurus and Plotosaurus. It is low and has
a bluntly rounded dorsal border in Platecarpus, Halisaurus onchognathus, Ecteno-
saurus and Tylosaurus. The condition of the coronoid buttress is intermediate

RUSSELL: AMERICAN MOSASAURS 53

between these two types in Prognathodon. In Platecarpus and Tylosaurus the
meckelian fossa has shifted anteriorly relative to Varanus and the insertions
for the Mm. adductor posterior and pseudotemporalis profundus were probably
located on a broad shelf of the surangular lying between the glenoid and
meckelian fossae and overhanging the prearticular medially. In Clidastes and
Mosasaurus the meckelian fossa occupies the same position that it does in
Varanus. The M. adductor posterior probably inserted partly on the posterior
floor of the fossa (on the surangular) and partly on the dorsal edge of the
prearticular. Further anteriorly the M. pseudotemporalis profundus probably
inserted entirely on the dorsal rim of the prearticular.

CORONOID

The coronoid is a saddle-shaped bone straddling the longitudinal “shoulder”
on the anterodorsal end of the surangular which it enclosed in a deep ventral
sulcus. The anterior end of the coronoid is pointed and cleft along the midline.
In lateral profile the upper margin of the bone slopes back to a posteriorly
ascending process. Fibers of the Mm. adductor mandibulae externus superficialis
and medialis must have inserted on the ventroposterior rim of a large, posteriorly
overhanging crest that rises along the lateral surface of the process to its dorsal
apex and curves medially to join the vertical posterior wall of the coronoid.
The bodenaponeurosis through which parts of the Mm. adductor mandibulae
externus medialis and profundus and the M. pseudotemporalis superficialis in-
sert on the mandible probably was attached to the posterodorsal edge of the
coronoid, as it is in Varanus. The thin posterior wall of the coronoid is sutured
ventromedially to the coronoid buttress of the surangular.

The posterior process of the coronoid is small in Platecarpus, larger in
Tylosaurus, and large in Clidastes, Mosasaurus, Plotosaurus and Prognathodon.
In Prognathodon and Tylosaurus the anterosuperior margin of the surangular
is horizontal on the lateral face of the posterior wall of the coronoid, in Plate-
carpus it rises posterodorsally, in M. missouriensis it ascends more steeply, and
in Clidastes, M. maximus and Plotosaurus it becomes nearly vertical dorsally.
The laterally and medially descending wings of the coronoid are small in
Clidastes, Platecarpus and Tylosaurus, are intermediately developed in M. mis-
sourtiensis, and are large in M. maximus. In Prognathodon the medial wing
is very large and broad and covers the anterodorsal edge of the angular, although
the lateral wing is only intermediately developed. The lateral wing is small in
Plotosaurus and the whole lateral aspect of the coronoid is very similar to that
of Clidastes. There is a “C’’-shaped excavation on the medial side of the coronoid
process in M. missouriensis, M. maximus and Prognathodon in which the seg-
ment of the bodenaponeurosis connecting the M. pseudotemporalis superficialis
with the coronoid may have inserted.

A coronoid (USNM 4993) possibly belonging to Globidens is very massively
constructed and dorsoventrally compressed. "The coronoid process is small and
the posterior wall of the bone extends in a heavy process back over the antero-
dorsal rim of the surangular.

ARTICULAR-PREARTICULAR

The portion of the glenoid fossa located on the articular is a triangular,
dorsally convex, rounded surface, obliquely cut off anterolaterally by the articular
surface of the surangular, and was capped with cartilage in life. The glenoid
fossa caps a broad buttress, which elevates it from the dorsomedial-sloping retro-
articular process which arises from the posterodorsal and inferior edge of the

54 PEABODY MUSEUM BULLETIN 23

fossa. The retroarticular process is generally a moderately heavy rectangular
bar of bone, although its outline is variable in the different genera, and bears
a foramen near the center of its anterodorsal face for the corda tympani. The
margin of the process lies at right angles to its dorsomedially and anterolaterally
inclined surfaces and is rimmed with a smooth band continuous anteriorly with
the glenoid fossa. The M. depressor mandibulae probably inserted over the
dorsoposterior surface of this band, and fibers of the pars profundus of the M.
pterygoideus inserted on the medial face of the retroarticular process and those
of the pars superficialis on its lateral face. The prearticular portion of the
fused articular-prearticular is a very long thin ribbon of bone that arises from
the ventral margins of the retroarticular process and the glenoid fossa. It extends
anteriorly, first forming the ventral margin of the posterior segment of the
lower jaw, then rises medially from between the surangular above and the
angular below into the mandibular foramen of the anterior segment of the jaw
between the splenial and dentary. In Tylosaurus (Huene, 1910, figs. 8-10) the
prearticular ends in a sharp point beneath the fifth dentary tooth. As the pre-
articular passes along the surangular, its medial border becomes slightly sulcate
to form the lateral wall of the meckelian fossa, but anteriorly it becomes a
vertical sheet of very thin bone.

The retroarticular process is triangular in outline in Clidastes and has a
very massive dorsal rim, probably indicating a strong M. depressor mandibulae.
In an articular questionably referred to Globidens (USNM 4993) the retro-
articular process is very powerfully built, circular in outline, and has a large
tuberosity on its ventral margin beneath the posterior end of the glenoid fossa.
In Mosasaurus the process is rectangular and a scarred surface indicates that the
insertion for the M. depressor mandibulae had spread onto its dorsomedial face.
The process may be similar in Plotosaurus. In Platecarpus the retroarticular
process is circular in outline, while in Tylosaurus it is dorsally rounded and
ventrally flat. In Prognathodon the process is rectangular. The retroarticular
process of Halisaurus onchognathus (see Merriam, 1894, pl. 4 fig. 10) is very
peculiar. It rises vertically immediately behind the mandibular cotylus, then
descends abruptly posteriorly, forming an anteroposteriorly compressed fan in
lateral profile. There is a large tuberosity similar to the one in Globidens at its
ventroposterior border.

MARGINAL DENTITION

In mosasaurs teeth in various stages of replacement entirely fill the alveolar
margins of the upper and lower jaws. They are arranged in a single row in
each mandibular ramus, with the posterior edge of the bony base of one tooth
contacting or nearly contacting the anterior edge of the succeeding tooth base.
The marginal dentition of Mosasaurus is first described in detail and then the
tooth crowns are described for the different genera, since these are often ge-
nerically characteristic.

‘The crowns of the marginal teeth in Mosasaurus are large and divided into
a lingual and buccal surface by a longitudinal carina. The lingual surface of
the tooth is “U-shaped in cross-section; the buccal surface is more nearly flat,
both being divided by vertical ridges into prisms. The tips of the teeth are
more posterointernally inclined than the main axis of the crown. The thin
external enamel layer is underlain by a much thicker layer of dentine which
is hollow and forms a pulp cavity in the center of the crown. It fits into an ex-
cavation in the form of an inverted cone in the bony base of the tooth (see
Leidy, 1865a, pl. 20). The depth of the tooth base exceeds the height of the

RUSSELL: AMERICAN MOSASAURS 55

Text-fig. 30. Marginal tooth of Mosasaurus maximus (NJSM 11052, x 14). Lateral and dorsal
views.

crown by about one and one-half times and is also posteriorly inclined, its cen-
tral axis meeting that of the crown in an acute angle. The grain of the bone
in the tooth base is finer than that of the surrounding mandible and parallels
the longitudinal axis of the tooth base instead of that of the mandible. A tun-
nel in the center of the base communicates with the pulp chamber inside the
tooth crown.

In Mosasaurus maximus the teeth are small on the premaxilla and the ante-
rior tips of the dentaries and increase in size posteriorly to a point near the
center of the jaws. They then gradually decrease in size posteriorly. The flat
buccal surface is located across the anterolateral corner of the anterior teeth.
Here the lingual surface of the tooth is strongly “U”-shaped. Further posteriorly
the buccal surface is rotated to face directly laterally and the teeth become
more compressed from side to side and longitudinally symmetrical. ‘Teeth with
only two prisms on the buccal surface may be restricted to the dentaries and
those with three external prisms to the upper jaws. There are two teeth on each
side of the premaxilla, 13-14 on each maxilla, and 14-15 on each dentary in
Mosasaurus.

The first stage in the process of tooth replacement occurs when a shallow
pit is excavated between the posteromedial side of the tooth base and the sur-
rounding bone of the mandible to house the growing successional tooth. This
cavity soon breaks through to the alveolar surface and continues to enlarge at
the expense of the bony base of the old tooth. The enamel outline of the suc-
cessional tooth is formed in this cavity and eventually only the periphery of
the old crown rests on its former base. It is then easily broken off and the new
tooth rises into the cavity to replace it, at which time dentine is probably
deposited into its enamel cap and the new base is formed.

Edmund (1960, pp. 87-91) has described the pattern of tooth replacement
in mosasaurs. Replacement waves proceed from the front of the jaws posteriorly,
alternatingly affecting even and odd numbered teeth. The method of replace-
ment of an individual tooth is as outlined in the preceding paragraph. Edmund
(1960, p. 89) however noted that the newly-formed bony base of a successional
tooth is not immediately fused to the bone of the surrounding jaw, and teeth
in this stage of development are frequently lost before fossilization occurs. This
accounts for the large number of well-formed isolated teeth of Mosasaurus
maximus that have been collected from the Cretaceous greensands of New
Jersey. Third generation replacement teeth have been observed by Edmund in
a single alveolus.

56 PEABODY MUSEUM BULLETIN 23

In the center of the jaws of Clidastes the teeth are bicarinate, laterally com-
pressed, and covered with a smooth coat of enamel. The lingual and buccal
surfaces of the crown are subequal and separated from one another by a well-
developed anterior and posterior carina. The teeth are broad based and tri-
angular in lateral aspect with slightly posteriorly recurved tips. Anteriorly the
teeth are slender and more strongly recurved. The posterior carina is absent
and the teeth are triangular in cross-section, with the rounded base of the
triangle facing towards the rear. The teeth diminish in size towards the poste-
rior end of the tooth row and become more circular in cross-section and strongly
recurved. The teeth are alike in corresponding regions of the upper and lower
jaws. There are two teeth on each side of the premaxilla, 14-15 on each maxilla,
and 16 on each dentary of Clidastes liodontus; in C. propython there are 18
teeth on each maxilla and 17-18 on each dentary.

The maxillary teeth of Globidens alabamaensis (Gilmore, 1921a, pls. 39-40)
are circular in cross-section and nearly so in lateral profile. They are attached
to relatively constricted bony bases, lack anteroposterior carinae, and come to
an indistinct point distally. Anteriorly the teeth are small in diameter with
anteroventrally inclined bases. The teeth increase in size and the bases become
upright passing back to the tenth tooth, the last one preserved in the type maxilla.
The posteriormost tooth in the dentary referred to Globidens (SDSM 4612)
is in the form of a small nubbin with a well-defined apical point.

In Plotosaurus, “There are twenty teeth in the upper row on each side and
seventeen visible teeth and alveoli in each dentary. There are two small pre-
maxillary teeth. Then follow a series of increasingly longer teeth of which the
tenth and eleventh are the largest. The anterior teeth are compressed and
have delicate anterior and posterior carina, while in the eight smaller hind
teeth, both above and below, the cross-section is more circular, the crowns are
shorter, blunter and more abruptly recurved, and the delicate buccal and
lingual vertical striae are more distinct. There are faint traces of striae on some
of the back teeth. The crowns are set higher . . . than in other known mosasaurs.”
(Camp, 1942, p. 4)

In Platecarpus the crowns of the teeth are long and slender. They are
nearly circular in cross-section at the base, the anterior and posterior carinae
becoming more distinct as the tip of the tooth is approached. The buccal
surface of the crown is strongly faceted and the lingual surface is lined with small
vertical ridges. The teeth are poorly differentiated along the jaw margins, al-
though the anteriormost teeth tend to be smaller and somewhat procumbent
and those from the back of the jaws are relatively shorter and stouter. The teeth
in the upper jaws are slightly larger than those in the dentaries. In both
jaws the teeth are posteromedially recurved. There are two teeth on each side
of the premaxilla, 12 on each maxilla, and 12 (rarely 11) on each dentary in
Platecarpus. Isolated teeth of Plioplatecarpus and Ectenosaurus are very similar
to those of Platecarpus.

The bony tooth bases in “Platecarpus” intermedius are peculiarly swollen
to a degree surpassing that of other known mosasaurs. The only teeth preserved
in entirety are those of the posterior portion of the dentary fragments of the type
specimen (see Leidy, 1873, pl. 34 fig. 2). These teeth resemble those of Clidastes
in lateral outline, but have a much more inflated appearance, with smooth,
gently rounded sides. Replacement teeth in the anterior part of the dentary are
more compressed and possess sharply pointed tips (see Leidy, 1873, pl. 34 fig. 5).

In Prognathodon overtoni the tooth crowns in the center and _ posterior

RUSSELL: AMERICAN MOSASAURS 57

regions of the jaws are large and triangular in lateral outline (see Williston,
1898b, pl. 30 fig. 1). They are bicarinate, the lingual and buccal surfaces being
subequal in size and covered with smooth enamel. The tips of the teeth are only
slightly recurved posteromedially. In cross-section the teeth are oval, with the
long axis in the anteroposterior direction, and somewhat inflated, instead of
compressed. The teeth in the anterior portions of the mandibles are more nearly
circular in cross-section, where they are smaller, more slender, and more re-
curved. The teeth are similar in the upper and lower jaws. There are two teeth
on each side of the premaxilla, 12 on each maxilla, and 14 on each dentary.

In Plesiotylosaurus, the teeth are, “. . . exceptionally large [and] closely
spaced,” (Camp, 1942, p. 18). The mandibular teeth appear to be quite similar
to those of Prognathodon, except that the crowns are higher than in the latter
genus. There are two teeth on each side of the premaxilla, 13 on each maxilla,
and 16-17 on each dentary.

The teeth of Tylosaurus are unusually brittle and are very infrequently
preserved in entirety in specimens collected from the Niobrara chalk. In lateral
profile they are triangular with slightly posteromedially recurved tips. The
teeth resemble those of Mosasaurus in cross-section, with a “U’-shaped lingual
and a flattened buccal surface. The carinae separating these surfaces are less
well-developed than in Mosasaurus, however, and the surfaces themselves are not
faceted, although there is some vertical ridging on both and the teeth have a
somewhat inflated appearance. They are largest in the center of the jaws, be-
coming smaller anteriorly and posteriorly. The buccal surface is rotated for-
ward to a position on the anterolateral edge of the anteriormost teeth. The
teeth are similar in corresponding regions of the upper and lower jaws. There
are two teeth on each side of the premaxilla, 12-13 on each maxilla (12 occur
more frequently in T. nepaeolicus, 13 more frequently in T. proriger), and 13
on each dentary.

FUNCTIONAL ANATOMY OF CRANIAL SKELETON

The osteology of the mosasaur skull is described in the preceding section.
Part of the soft anatomy of the cranium is reconstructed therein on the basis of
evidence preserved with the hard tissues. It now remains for the adaptive mean-
ing of various cranial structures to be commented upon.

SENSORY FUNCTIONS

As Camp (1942, p. 40) has noted, the excavation in the undersurface of the
frontal for the olfactory bulbs is relatively smaller in mosasaurs than in
Varanus. This is to be expected, for the sense of olfaction is usually diminished
in aquatic animals (see Howell, 1930, p. 74).

The orbits are large in all known mosasaurs. Interlocking sclerotic ossifica-
tions are present in Clidastes, Platecarpus, Tylosaurus, and also in Mosasaurus,

Text-fig. 31. Reconstructed cross section of skull of Platecarpus between orbits (x 1/4). X, sclerotic
ring.

Plioplatecarpus and Prognathodon (Dollo, 1889b, pl. 9). The function of sclerotic
plates (Walls, 1942, pp. 275-279) is to maintain the shape of the cornea and to
support the sclera in the region of the origin of Briicke’s muscle which affects
accommodation in the lacertilian eye. It is likely that about 14 such plates are
present in mosasaurs. The curvature of this ring of bone in Clidastes and Plate-
carpus shows that the eyeball nearly filled the orbits. The eyeball must have
been quite dorsoventrally flattened in Clidastes, as is usual in large lizards
(Walls, 1942, p. 623), and less so in other genera. The visual axis of the mosasaur
eye was directed laterally and a little dorsoanteriorly.

Vaughn and Dawson (1956) observed the presence of a calcified tympanic
membrane in Platecarpus and noted that a similar condition of the membrane
had been reported earlier in Plioplatecarpus and Tylosaurus. The membrane
is also calcified in Clidastes and Ectenosaurus, as it may have been in all mosa-
saurs. A calcified membrane therefore occurs in each of the three subfamilies of
mosasaurs and does not in itself indicate that any one of them was better adapted
to diving, as has been thought (e.g., Dollo, 1913, p. 621; 1924, p. 188; Williston,
1925, p. 272). The disc of calcified tissue appears to have a diameter always
smaller than that of the circular crest of bone that extends from the suprastape-
dial process around the tympanic ala to the infrastapedial process below on the
external surface of the quadrate. ‘The disc must have been suspended from
this crest by an outer membranous ring of the tympanum. According to Pum-

58

RUSSELL: AMERICAN MOSASAURS 59

phrey (1950, pp. 11-13) a fish has about the same density as the surrounding
water and is therefore virtually transparent to water-borne sound. The dis-
placement of the sound organ relative to the body of the fish is increased if it
is anchored to a mass of greater density which would not be so greatly affected
by sound vibrations in the surrounding medium, such as the attachment of the
saccular macula to an otolith. Since the body of mosasaurs must have been of
approximately the same density and homogeneity as that of fish, perhaps the
membranous suspension of a dense, calcified portion of the tympanum served a
function analogous to that of an otolith by “picking” vibrations out of the
water and transmitting them to the sensory apparatus of the inner ear.

In mosasaurines and tylosaurines the tympanic ala of the quadrate is not
unusually enlarged over the aigialosaur condition, and the space between the
tympanum and body of the quadrate must have been relatively small. In plio-
platecarpines (Ectenosaurus, Platecarpus, Plioplatecarpus, Prognathodon), how-
ever, the tympanic ala is highly developed and projects further anterolaterally
to enclose a chamber between the tympanum and quadrate shaft. This cham-
ber is strongly reminiscent of the cavity formed in the quadrate and adjoining
squamosal of turtles. According to Wever and Vernon (1956, p. 295) the open-
ing of the tympanic cavity in turtles during hearing experiments had no signifi-
cant affect on sound transmission through the middle ear. The purpose of the
chamber in the quadrate of plioplatecarpines is also unknown; it is perhaps
significant that in Plotosaurus, a mosasaurine, the stapes is a robust column of
bone (Camp, 1942, fig. 23) yet in Platecarpus, a plioplatecarpine, the stapes is
slender as in turtles and Varanus. The otic labyrinth of mosasaurs is practically
identical to that of Varanus.

Except for the calcification in the tympanum, the hearing apparatus of
mosasaurs is not greatly modified over conditions seen in terrestrial reptiles.
There is no indication of such elaborate adaptations as in the cetacean ear, where
the filling of the external auditory meatus with a wax plug having the same
sound conducting properties as water and the complicated pterygoid sinus
system regulating air pressure within the middle ear cavity enable whales to
hear well over a great range of external pressures (Fraser and Purves, 1960). A
partially calcified tympanum may have aided in preventing the tympanum from
being ruptured by rapid changes in external pressure as the animal plunged its
head into the water when submerging.

CRANIAL CIRCULATION

The course of some of the major blood vessels in the mosasaur head has been
outlined above with the descriptions of the basioccipital and basisphenoid bones.
Regarding the cranial circulation in Plioplatecarpus, Devillers (1943, pp. 14-15)
aptly concludes: “. . . s'il existait deux vaisseaux importants [internal carotid
artery, basilar artery] protégés par des canaux, cela ne me parait pas sufhsant
pour enférer une adaptation a la plongée. Ce fait existe chez d’autre Reptiles
non plongeurs (Crocodiliens et Chéloniens) et, d’autre part, il existait aussi
d’importants vaisseaux latéraux nullement protégés [internal jugular veins].
Il n'y a rien la de comparable a ce qu’on rencontre chez les Cétacés .... la
circulation céphalique était certainement moins bien protégée que chez les
Chéloniens et les Crocodiliens et surtout que chez les Cétacés.”

This statement could be applied equally well to the cranial circulation in
Platecarpus and Ectenosaurus. In other known mosasaurs not even the basilar
artery lies in a bony channel.

60 PEABODY MUSEUM BULLETIN 23

CRANIAL KINESIS

In the preceding section the elements of the mosasaur skull are grouped
according to the particular functional unit to which they belong. My paper
(1964) has presented some suggestions as to how these units may have operated
in life, and these are restated below in a somewhat condensed form. It may be
useful to compare cranial movement in mosasaurs with that of Varanus, which
has been described in detail by Frazzetta (1962). Mosasaurs did not all possess
especially kinetic skulls, indeed in some genera (Mosasaurus, Plotosaurus, Prog-
nathodon and Plesiotytosaurus) kineticism was entirely suppressed. However to
the extent that kinetic movement was possible in the skull of a given mosasaur,
it is believed to have operated in the following manner.

The term kinesis applies to the general condition in which elements of the
dermal skull roof and palatoquadrate (maxillary segment) move more or less as
a unit with respect to the braincase (occipital segment). The principal or meta-
kinetic axis of rotation between these two fundamental structural divisions is
fixed on either side of the posterodorsal corner of the skull by the loose con-
tact of the squamosal (maxillary segment) with the supratemporal (occipital
segment). Other sliding planes of contact between the maxillary and occipital
segments are situated between the parietal and supraoccipital above (metakinetic
joint) and the pterygoids and basipterygoid processes of the basisphenoid
below (basal articulation).

The occipital segment is composed of ten elements (prootics, opisthotics,
supratemporals, supraoccipital, parasphenoid, basisphenoid and_basioccipital)
all solidly sutured together into a rigid block. The maxillary segment in mosa-
saurs is, however, divided into a number of subunits (see text-fig. 1).

1. The parietal unit, consisting only of the fused parietals. This unit artic-
ulates through a transversely oriented hinge (mesokinetic axis) with the muzzle
unit anteriorly, and with the occipital segment ventrally through the metakinetic
joint and an overlapping suture with the supratemporal.

2. The quadrate units, articulating dorsally with the suspensorial processes
of the occipital segment, medially through ligaments with the quadratic rami
of the pterygoids, and ventrally with the glenoid fossae of the mandibles. The
ventral ends of the quadrates are free to swing in an anteroposterior plane.

Text-fig. 32. Reconstructed temporal region of Clidastes liodontus (after YPM 1335, x 1).

RUSSELL: AMERICAN MOSASAURS 61

Text-fig. 33. Restored superficial musculature of temporal region of Clidastes liodontus.

3. The basal units, composed of the pterygoid and ectopterygoid on each
side of the posterior roof of the oral cavity. Each unit is bound posteriorly by
muscles to the occipital segment and mandible, and by ligaments to the quadrate
units. It forms a sliding contact with the posteromedial corner of the jugal, and
is attached through an overlapping suture to the palatine (muzzle unit) to form
a flat, flexible sheet of bone through which the hypokinetic axis passes ina
transverse direction.

4. The epipterygoid units, forming a single strut on each side of the anterior
portion of the occipital segment. They are anchored to the basal units below,
and connected ligamentously to the occipital segment and parietal unit above.

5. The muzzle unit is the largest rigid structure in the mosasaur skull and
includes the premaxillae, nasals, septomaxillae, vomers, maxillae, prefrontals,
lacrymals, jugals, postorbitofrontals, squamosals, supraciliares and_palatines.
The unit meets the basal units posteroventrally through the hypokinetic axis
and the parietal unit posterodorsally through the mesokinetic axis.

An understanding of the operation of these numerous structural units will
perhaps be facilitated by a consideration of the movement of the muzzle unit
about the mesokinetic axis. In Platecarpus and Tylosaurus the postorbitofrontals
are solidly sutured to the ventral surface of the frontal, and this seems to be the
case in other genera of mosasaurs as well, although the conditions may not be
so unequivocal as in these two genera. This in effect makes the upper temporal
arcades extensions of the muzzle unit that project behind the mesokinetic axis,
since the postorbitofrontals and squamosals overlap each other in an im-
movable tongue-in-groove junction. As the muzzle unit was rotated upward
about the mesokinetic axis, the upper temporal arcades were depressed, and
vice versa.

The elevation and depression of the posterior ends of the supratemporal
arcades caused the paroccipital processes, and thereby the occipital segment, to
rock up and down about the atlas-occipital joint (see text-fig. 35; and Russell,
1964, pp. 6-8). The squamosals of necessity must have pivoted on the lateral face
of the supratemporals. Adjustment in the vertical relations between the paroc-
cipital process and the suspensorial ramus of the parietal took place through
slippage on the loosely overlapping parietal-supratemporal contact.

62 PEABODY MUSEUM BULLETIN 23

Text-fig. 34, Restored deep musculature of temporal region of Clidastes liodontus.

When the muzzle unit was protracted, or lifted in front of the mesokinetic
axis, the paroccipital processes were rotated ventrally about the atlas-occipital
joint by the descending supratemporal arcades, and the basipterygoidal proc-
esses of the basisphenoid were correspondingly brought forward and upward.
Thus the basal units were elevated and displaced anteriorly by a push from the
basipterygoid processes and a pull from the anteroventral corners of the muz-
zle unit (palatines). The reverse movement, or retraction of the muzzle unit
about the mesokinetic axis, caused a similar reversal in the rotation of the oc-
cipital segment and in the movement of the components of the maxillary seg-
ment (see text-fig. 35).

‘The movements of the kinetic units of a mosasaur skull have been described
by considering the consequences of rotation of the muzzle unit about the
mesokinetic axis. However the most important effectors of kinetic movement
in life are postulated to have been the Mm. rectus capitis posterior and ante-
rior, inserting respectively above and below the occipital condyle. If superficial
muscles, like the M. spinalis capitis above and the Mm. sternohyoideus and
geniohyoideus below, held the maxillary segment and lower jaw fixed relative
to the atlas-occipital articulation, then alternative contraction of the two rectus
capitis muscles would rotate the occipital segment up and down about the atlas
vertebra. The rotation of the occipital segment would then be transmitted
through the links of the maxillary segment to protract and retract the muzzle
unit (see text-fig. 35).

When the head of a mosasaur was at rest a line drawn from the metakinetic
joint to the basal articulation would descend anteroventrally at an angle of
about 45° with respect to the horizontal axis of the skull. The line would descend
less steeply during protraction, when the occipital segment was rotated upward,
and more steeply when it was rotated downward. Thus the metakinetic joint
and basal articulation were brought more closely together vertically in the
protracted state of the muzzle unit than in the retracted state. The same geo-
metric relations also obtain for a line drawn from the mesokinetic to the hypo-
kinetic axis. Assuming little or no vertical slipping on the metakinetic joint and
basal articulation, it will be seen from text-fig. 35 that the vertical separation
between them would directly control the vertical separation between the meso-
kinetic and hypokinetic axes, and thereby directly control the degree of pro-
traction of the muzzle unit. Activation of the constrictor dorsalis muscles would

RUSSELL: AMERICAN MOSASAURS 63

Text-fig. 35. Kinesis in mosasaurs. A. Muzzle unit elevated, anterior mandibular unit depressed. B.
Cranium at rest. C. Muzzle unit depressed, anterior mandibular unit elevated.

merely accentuate the elevation of the muzzle unit in the protracted state by
displacing the hypokinetic axis still further anterodorsally. Note that the quad-
rates need not have been incorporated into the kinetic mechanism of the
mosasaur skull.

STREPTOSTYLY

The term streptostyly is here used to describe the condition in which the
quadrate has lost its contact anteriorly with the lower temporal arcade, and is
only loosely bound medially to the pterygoid. The bone is then flexibly sus-
pended from the paroccipital process dorsally, and the activation of any muscle
attached to it will pivot the element about the quadratic cotylus of the paroc-
cipital process. Movement of the quadrate will not necessarily alter the posi-
tions of the occipital or remainder of the maxillary segments relative to each
other.

64 PEABODY MUSEUM BULLETIN 23

Kauffman and Kesling (1960) have studied a large ammonite conch which
was repeatedly bitten by a mosasaur. The dentary tooth impressions from one
mandible always maintain the same anteroposterior relations to those of the
other mandible, indicating there was no anteroposterior movement between
the lower jaws in the symphyseal region. However, tooth impressions from the
upper and lower jaws did not always align with each other when occluded.
This could only occur if the quadrates were independently moveable. Since both
basal units are fixed to a single rigid muzzle unit, it follows that in order for
the quadrates to have been independently moveable they must have been only
loosely attached to the quadratic ramus of the pterygoids. The single solid point
remaining upon which the quadrate could have pivoted is the cotylus on the
side of the suspensorial process of the occipital segment, which was therefore
probably not a sliding articulation.

Muscles that acted to protract the lower jaw (see text-fig. 36) were the M.

Text-fig. 36. Streptostyly in mosasaurs. A. Mandible protracted. B. Mandible retracted.

pterygoideus (the horizontal component of force transmitted through the man-
dible would pull the base of the quadrate anteriorly) and the M. depressor
mandibulae (rotating the mandible ventrally to counteract the dorsal compo-
nent of the M. pterygoideus). The mandible was retracted by the horizontal
component of force from the contracting jaw adductor muscles. Either the
presence of prey between the jaws or a stabilizing pull from the M. depressor
mandibulae would prevent the upper and lower jaws from entirely closing.
Grooves, paralleling the longitudinal cranial axis of the attacking mosasaur,
cut into the conch of the above-mentioned ammonite bear witness to the force
with which the jaws could be retracted (Kauffman and Kesling, 1960, p. 213).

MANDIBULAR MOVEMENTS

The mosasaur jaw is divided into two halves by a joint near the center of
the ventral margin of each mandible. The articular, angular, surangular and

RUSSELL: AMERICAN MOSASAURS ° 65

coronoid are incorporated into a posterior structural unit, and the splenial and
dentary into an anterior one. Dorsally a thin blade-like process of the prearticular
spans the gap separating the two units to penetrate deeply between the splenial
and dentary into the mandibular foramen of the anterior unit. Ventrally the
presence of vertical keels and grooves on the articular surfaces of the ginglymoid
splenio-angular joint further restricted movement of the anterior unit to a
vertical plane.

Most probably the anterodorsal areas of the coronoid and surangular were
bound to the posterodorsal corner of the dentary by elastic ligaments. As the
lower jaws hit the body of a victim the anterior units of the mandibles would
absorb the shock of impact by rotating down about the splenio-angular joint,
putting the ligaments binding the units to the posterior mandibular units
under tension. These tensed ligaments, possibly aided by slips of the M. adductor
mandibulae externus superficialis, would then act to restore the anterior unit
to its former position.

The ratio between the distances from the center of the glenoid fossa to the
anterior limit of the insertional area for the adductor mandibulae externus
muscles, and the center of the fossa to the anterior tip of the dentary may be
taken as an indication of the speed and power with which the jaws were brought
together. In Platecarpus ictericus, P. coryphaeus and Prognathodon overtoni the
ratio is about 1:4; in Mosasaurus missouriensis, Tylosaurus proriger and T.
nepaeolicus it is about 1:4.5; and in Clidastes propython, C. liodontus and
Plotosaurus bennisoni (Camp, 1942, fig. 1) it is usually about 1:5.2. Platecarpus
and Prognathodon presumably had the most powerful bite, and Clidastes and
Plotosaurus had the most rapid.

CONCLUSIONS

So far as is known all mosasaurs possessed a good sense of sight and balance,
good subaqueous and probably impaired subaerial hearing, and a poor sense of
smell. The presence of a calcified tympanum in all three subfamilies is no indi-
cation that any one of them had an unusual capacity for deep diving. The
cranial blood vessels were no more protected than in ordinary terrestrial reptiles.
In fact, the large eyes and lack of any special pressure-protective devices suggest
a surface-swimming mode of life, although some mosasaurs (Globidens, Com-
pressidens) may have subsisted on shallow water benthonic invertebrates.

Generalized mosasaurs possessed a kinetic skull primarily activated by the
operation of anterior axial muscles rotating the braincase in a vertical plane.
Cranial kinesis was evidently lost in many later forms. The quadrates were
streptostylic and independently moveable in all mosasaurs and acted to protract
and retract the lower jaws. The intramandibular joint moved in a vertical
plane and probably served as a shock absorbing device.

Frazzetta (1962) concluded that kinesis permitted the upper jaws to move
downward and engage the prey simultaneously with the lower jaws. This
lessened the chances of the prey being deflected from the mouth by the lower
jaws. Kinesis was evidently not an essential element in the feeding mechanism
of all mosasaurs, as is shown by its loss in some later forms. Perhaps the inertia
of the bodies of larger prey or the viscosity of the aqueous medium in which
mosasaurs lived impaired the usefulness of the kinetic mechanism inherited
from their smaller, more terrestrial ancestors.

Streptostylic quadrates are, however, found in all mosasaurs and must have
been useful adaptations in aquatic feeding. They permitted the mandibles to

PEABODY MUSEUM BULLETIN 23

66

ayzznut

‘payenoid smef ayy pue ‘pajonpqe sirun xze~nqrpuem so19qUe
ay} ‘pouado smef ayy Surmoys (% x) snot19791 sndunzajn]1q Jo [INYs per0ysay 7g

pue
“*BY-1x9,],

67

AMERICAN MOSASAURS

RUSSELL

ayy

‘

pasop> Apteau s

‘papenar smel ay) ‘pajonppe s}TUN Ie[NqIpueUT AOTIaJUe PUe 9]ZZNUT
mel ayy Sutmoys (Wi X) sna24aj22 sndivzajvjq JO [NYS Prtojsoy “8g

By

1x9],

68 PEABODY MUSEUM BULLETIN 23

be retracted and consequently greatly assisted the mosasaur in forcing prey into
its throat without the aid of gravity, claws or some solid point of leverage. It is
doubtful that the inertial feeding method of lizards described by Gans (1961,
pp. 218-219), could have been very effective in underwater swallowing. If a
mosasaur lifted its head above the surface, however, the inertial method to-
gether with the aid of gravity, would also greatly facilitate the engorgement of
large bodies.

The long slender jaws of Clidastes were probably adapted to rapid biting.
The marginal dentition is trenchant, and alternative protraction and retrac-
tion of the mandibles in this genus (and in Liodon) might have been effective
in sawing a large object into pieces of swallowable size. Dollo (1913, p. 614-620;
see below) discussed in some detail the possible food habits of three genera of
mosasaurs. Because of the similarity of its large, firmly rooted marginal teeth
to those of the killer whale (Orcinus), Mosasaurus itself may have preyed upon
the larger contemporaneous marine reptiles. Its skull was akinetic. In Ploto-
saurus the skull was akinetic, but the long slender jaws suggest a rapid bite.
Remains of small fish were discovered in the visceral region of a specimen of
Plotosaurus tuckeri (Camp, 1942, pp. 11, 20).

In Globidens the jaws are powerfully constructed and the teeth are spherical,
indicating a diet of heavily shelled mollusks (Gilmore, 1912, p. 481). Kineticism
may have been useful in orienting the prey in the mouth. Because oysters thrive
in shallow brackish water (see Priddy, 1954), one might expect to find Globidens
remains associated with a similar sort of paleoenvironment. The long jaws and
laterally flattened teeth of Compressidens fraasi (and “Platecarpus’ inter-
medius?) were best adapted for crushing the thin shells of the abundant Mae-
strichtian echinoids. A broken test of Hemipneustes was once found between
the teeth of one individual of the former genus (see Dollo, 1913).

The jaws are short in Platecarpus and the skull is relatively highly kinetic,
kineticism here probably serving to minimize the chance of deflecting small
prey away from the jaws. The teeth are long, slender, and circular in cross-
section. Williston (1899, pp. 40-41) reported that, “Quantities of [food] remains
were discovered in the abdominal region [of a Platecarpus], matted together
and more or less comminuted. A close examination of these remains discloses
nothing but fish bones, and usually only those of a small size. The largest ob-
served is a vertebra of an Empo [= Cimolichthys], or some allied fish, of about
four feet in length.” As belemnites are very rarely found in the Niobrara Chalk
(personal communication, Jeletzky, 1965), they could not have formed an im-
portant food source for Platecarpus throughout the distribution of the chalk
facies. However Dollo (1913) thought that Plioplatecarpus may have subsisted
on naked. cephalopods, as the teeth are weakly implanted and restricted to the
anterior portion of the jaws in Maestrichtian species of the genus, as in the
delphinid Globicephala (‘‘blackfish”) which eats squid and cuttlefish. Two belem-
nite endoskeletons have been found between the mandibles and cervical vertebrae
of one specimen of Plioplatecarpus (Dollo, 1913).

In Prognathodon the skull is akinetic. The jaws are relatively shorter and
heavier, and the teeth appear to be better adapted for crushing than in any
mosasaur except Globidens. It is perhaps significant that in the ammonite de-
scribed by Kauffman and Kesling the tooth marks of the marginal dentition of
the mosasaur have cross-sections similar to those of the marginal teeth of Prog-
nathodon overtoni. The jaws of Tylosaurus are strongly built, moderately
elongated, and filled with heavy teeth similar to those of Mosasaurus. The skull

RUSSELL: AMERICAN MOSASAURS 69

is kinetic. The elongate premaxillary rostrum could have been used to stun
prey or defend the mosasaur against enemies (sharks). Turtle bones have been
found in the body cavity of Hainosaurus, a Belgian tylosaurine (Dollo, 1887, p.
520).

There is an instance of possible cannibalism reported from Canada (Anony-
mous, 1962, p. 5) where fragments of a small mosasaur were found mixed with
the skeleton of a larger one.

GENERAL DESCRIPTION OF POSTCRANIAL SKELETON

In life mosasaurs must have resembled what they indeed were, very large
marine lizards. They were considerably larger than most extant reptiles, av-
eraging 20-30 feet in length. The chest was more concentrated in the anterior
portion of the body than is usual in lizards, and the flanks were probably slender
(Williston, 1899, p. 40), merging imperceptibly into the fleshy base of the tail
posteriorly. The tail was relatively shorter than in Varanus, but was very flat
and deep, forming a more efficient sculling organ. The legs were much shortened
and the feet were transformed into flippers, the fore and hind flipper being of
approximately equal size. The integument has been preserved in specimens of
Platecarpus (Williston, 1899, p. 41), Tylosawrus (Snow, 1878; Williston, 1898b,
p- 215, pls. 69-70), and Ectenosaurus (FHM 7937), all from the Niobrara Chalk
of western Kansas. The body was probably completely covered with small,
rhomboid scales.

For additional descriptions of the postcranial morphology of American
mosasaurs the reader is referred to the works of Williston (1898b), Merriam
(1894), Martin (1953), Osborn (1899a), Camp (1942), Huene (1911), Wiman
(1920), and Marsh (1880), listed in order of decreasing importance. As with
the cranial skeleton, information relating to Californian material has been taken
from Camp’s (1942) monograph.

CERVICAL VERTEBRAE
ATLAS

The atlas complex in mosasaurs includes the four discrete ossifications typical
of all lizards, centrum (odontoid), intercentrum, and two neural arches. There
is no proatlas. The neural arches are linked ventrally to the intercentrum to
form a ring-like structure between the axis and occipital condyle; the atlas
centrum is functionally distinct, being solidly attached to the axis.

Internally the body of the neural arch is subdivided into several surfaces.
There is an anterior concavity, elliptical in outline and vertical in position, for
articulation with the dorsolateral portion of the occipital condyle. Behind this
a triangular facet with a ventrally directed apex faces medially to contact the
atlas centrum. The posterior border of this facet is in turn bounded by an oblique
surface, inclined both posterolaterally and posterodorsally, for articulation with
the axis centrum. These three surfaces are truncated as they converge ventrally
by the articulation for the atlas intercentrum. All of these surfaces were covered
with cartilage in life. The external side of the atlantal body narrows behind the
dilated portion supporting the condylar articulation and terminates postero-
ventrally in a pointed synapophyseal process. The process is well developed in
Plotosaurus, Platecarpus and Ectenosaurus, varies from large to nearly absent
in Tylosaurus, and is especially large in Clidastes and Mosasaurus. Its large size
relative to the condition in Varanus is suggestive of a high degree of develop-
ment of the M. iliocostalis cervicis in mosasaurs, which probably inserted in
part on its cartilage-capped tip.

The dorsal border of the atlas synapophysis extends anterodorsally to form
the posterior margin of a stout, anteriorly inclined spinous process that arched
over the spinal cord. Its anterior border rises directly above the condylar artic-
ulation (Plotosaurus, Platecarpus, Tylosaurus) or is separated from it by a notch

70

RUSSELL: AMERICAN MOSASAURS TALL

Text-fig. 39. Atlas-axis vertebrae in Clidastes propython (YPM 1100, x 14). A. Anterior view of
atlas. B. Lateral view of atlas-axis. Abbreviations: aa, atlas neural arch; ai, atlas intercentrum;
ac, atlas centrum (odontoid); ax, axis; axi, axis intercentrum; hy, facet for hypapophysis
(intercentrum) of third cervical vertebra.

(Clidastes, Ectenosaurus, Mosasaurus). Spinal nerves I and II exited in front of
and behind the base of the spinous portion of the atlantal arch, respectively.
The spinous process expands somewhat distally and has a slightly ventromedially
incurved dorsoposterior corner. The tips of these processes are usually longi-
tudinally ribbed and contact each other in Clidastes and Plotosaurus, are sepa-
rate in Platecarpus, and widely separate in Tylosaurus and Mosasaurus where
the spinous processes are relatively short. Thus a true atlas neural spine is absent
in mosasaurs.

Midway between the tips of the spinous and synapophyseal processes on the
dorsal margin of the atlas there is a longitudinally grooved region in the center
of which is a slight excavation. This pit probably marks the tendonous insertion
of the M. longissimus cervicis and anterior attachment of a tendonous sheet
that extends dorsoposteriorly through the muscle masses of the Mm. transversalis
capitis and articuloparietalis in Varanus (Nishi, 1916, p. 233). The site of inser-
tion of the M. obliquus capitis inferior is probably located immediately ante-
rior to the above-mentioned pit. According to Nishi (1916, p. 243) a small group
of fibers from the M. obliquus capitis magnus (M. obliquus capitis superior)
arises on the surface of the atlas just lateral to the longissimus tendons. In some
mosasaurs (Plotosaurs, Platecarpus, Ectenosaurus, Tylosaurus) a prominent
posterodorsally directed tuberosity is developed at this point, in others (Clidastes,
Mosasaurus) only a small ridge is present.

Each neural arch meets the atlas intercentrum ventromedially. In the
vicinity of their mutual contact the external surface of both elements is strongly
ribbed in a vertical direction, indicating the attachment of strong ligaments
binding them together (Clidastes, M. conodon, Plotosaurus, Platecarpus, Ecteno-
saurus, Tylosaurus). In M. maximus these ribbings interdigitate lateral to the
cartilaginous contact and form a suture solidly uniting the two ossicles.

The atlas intercentrum is in the shape of a prism bowed slightly upwards
at the ends so that its flat base is transversely convex ventrally. Because the
anterior facet is slightly cupped to receive the condyle from the basioccipital and
the posterior contact with the axis intercentrum is flat, these surfaces intersect
dorsally to form a rounded, shallowly concave edge. The triangular ends of the
bowed prism rise to meet the base of the neural arches. These facets were all
covered with cartilage in life. A small tubercle is present on the posterior mid-
line of the rectangular ventral surface. In Zguwana (Evans, 1939, p. 56, fig. 15)
a portion of the M. longus colli inserts on each side of this tubercle.

72 PEABODY MUSEUM BULLETIN 23

The atlas centrum is quite constant in shape in mosasaurs. In lateral aspect
it resembles a very squat, horizontal cone, the dorsal portion of which has been
flattened to floor the spinal cord in its course between the foramen magnum and
neural canal of the axis. ‘The whole anteroventral and anterolateral area of
the cone was covered in cartilage. Posteriorly the cone abuts against the axis
centrum which it meets in two vertical, posteriorly converging flat surfaces,
one on either side of the midline. ‘The ventral contact with the axis intercentrum
slopes posterodorsally in a transverse direction so that the ridges formed by the
intersection of the three surfaces form an inverted “Y”. The surfaces them-
selves are minutely ridged and pitted, suggesting that the atlas centrum must have
been rigidly fastened onto the anterior face of the axis complex.

AXIS

The axis vertebra is a very powerfully built structure lying at the functional
anterior terminus of the vertebral column and linking it to the cranium. The
dual nature of its function is clearly shown in the form of the axis, which
posteriorly is not greatly different from that of a succeeding cervical vertebra.
Anteriorly it is highly modified to provide origins for muscles acting directly on
the skull and to support the skull indirectly through the atlas vertebra.

The neural spine is shaped like a flat-edged hatchet blade, expanding ante-
riorly and posteriorly above the relatively constricted neural arch to a length
about equal to that of the centrum below. The posterior border is very heavy,
giving the spine a ““T’’-shaped outline in horizontal section and providing a
large area of insertion for powerful semi-spinalis cervicis and interspinalis
muscles. It was capped dorsally by a cartilage-covered surface in life which
marked the anteriormost origin of the M. spinalis capitis. The dorsal edge of
the spine becomes compressed and descends to a small elliptical expansion at
its anterior tip. The superficial portion of the M. rectus capitis posterior may
have arisen in front of the spinalis capitis origin and the deeper portion prob-
ably from the anteriormost tip of the spine. They arise from comparable areas
in Varanus. There is a shallow depression, centrally located on each side of the
neural spine anterior to the broad posterior buttress and above the neural arches,
from which the M. obliquus capitis magnus must have arisen,

The neural arches are fused to the axis centrum and to the centra of all
succeeding vertebrae without a trace of a suture in even the most immature
specimens. Except for the complete or nearly complete absence of an anterior
zygopophysis there is little to distinguish the axis neural arch from that of other
cervical vertebrae. The M. obliquus capitis inferior probably arose from its
dorsolateral face as it does in Varanus. The posterior zygopophyses are normally
developed and face ventrolaterally at an angle of about 45°. The first traces
of a zygantral articulation may occur on the internal side of the posterior
zygopophysis, between the zygopophyseal articulation and a median dorsoventral
ridge on the posterior surface of the neural spine. Zygantra may be well de-
veloped in Clidastes, incipiently developed in Mosasaurus, Plotosaurus and
Platecarpus, and are evidently absent in Plioplatecarpus (Dollo, 1894, p. 235)
and Tylosaurus.

Short epaxial muscles link the posterior edge of the neural spine and neural
arch of the axis to the anterior edge of the neural spine and arch of the suc-
ceeding vertebra. In Varanus these include the Mm. interarticulares, inter-
arcuales, and interspinales. The first muscle originates from the cranial edge of
the anterior zygopophysis of the fourth cervical and the anterior surface of the
posterior zygopophysis of the third cervical and passes forward to insert on the

RUSSELL: AMERICAN MOSASAURS 73

posterior surface of the posterior zygopophysis of the axis. The other two muscles
pass between opposite edges of the neural spine and arch of the axis and third
cervical. These three muscles are all more or less intergrown in Varanus and
bear the same relations to the vertebrae along the entire presacral portion of
the vertebral column. There is no reason to think they were any differently
arranged throughout the zygopophysis-bearing portion of the vertebral column
of mosasaurs.

The cranial face of the axis centrum is expanded to a width about equaling
the anteroposterior length of the centrum. In anterior aspect it has an approxi-
mately reniform outline, with the long convex surface facing downwards. In
horizontal cross section the dorsal portion of the cranial face of the centrum
resembles a ‘“‘W,” the lateral prongs of the ‘““W” pointing posteriorly. The paired
anteromedially inclined planes are square shaped and roughtened to provide
an immovable contact for the atlas centrum; the two posterolateral-sloping semi-
circular surfaces are smooth and articulated with the atlas neural arches. Ven-
tromedially the cranial face is hollowed to form an anteroventral-directed,
roughened excavation receiving the posterodorsal side of the axis intercentrum.

The main body of the axis centrum is sharply constricted behind its anterior
face, but re-expands slightly to terminate in a convex, elliptical articulation for
the centrum of the third cervical vertebra. In posterior aspect this articulation
is nearly circular in outline in Clidastes, Mosasaurus and Tylosaurus, but in
Platecarpus its width exceeds the depth by about one-half. The axis synapophysis
is located on the center of the lateral surface of the centrum and is connected
anteriorly by a horizontal ridge of bone to the facet for articulation with the
atlas neural arch. The lateral edge of this ridge is roughened in Platecarpus
and was rimmed in cartilage in Clidastes, Plotosaurus? and Tylosaurus. The
atlas synapophysis overhangs the anterior portion of the ridge in Clidastes and
M. conodon, and it seems likely that this ridge was also an area of insertion of
the M. iliocostalis cervicis. In an excellently preserved specimen of Platecarpus
in the Tiibingen Museum, Huene (1911, pl. 3) shows a rudimentary cervical rib
articulating with the synapophysis of the axis vertebra. Williston and Case
(1892, pp. 20, 24) also report a cervical rib present on the axis of Clidastes. Dollo
(1894, p. 232) reports its presence on the axis vertebra of M. conodon and
Plioplatecarpus. A large peduncle with a distal facet for articulation with a
hypapophysis is located on the ventral surface of the axis centrum. In Clidastes,
Mosasaurus and Plotosaurus it is almost circular in outline and faces ventro-
posteriorly. In Platecarpus it tends to be more triangular, with an apex pointed
directly anteriorly. It is similar in Tylosaurus but more laterally compressed and
usually bears a transverse sulcus across its center. In the last two genera the
facet is only slightly posteriorly inclined.

The axis intercentrum is a short, hemicylindrical ossification firmly attached
to the anteroventral margin of the axis centrum. The facets for articulation
with the axis centrum behind and the atlas centrum above meet at a high angle
and are similarly roughened. In contrast, the facet for the atlas intercentrum
on the anterior face of the bone is smooth. The ventral surface of the bone is
transversely convex and much wider than long. A large posteroventrally directed,
tongue-shaped tuberosity is present on the ventromedian surface, which prob-
ably represents a major site of insertion of the M. longus colli.

CERVICALS THREE TO SEVEN

The form of vertebrae in the posterior cervical region changes gradually
posteriorly into that typical of thoracic vertebrae, making it difficult to define

74 PEABODY MUSEUM BULLETIN 23

the morphological posterior end of the cervical series. The last cervical vertebra
is usually taken to be the one that precedes the first vertebra with ribs reach-
ing the sternum (Romer, 1956, pp. 227, 241). In Varanus there are, by this defini-

Axis

Text-fig. 40. Anterior portion of vertebral column of Platecarpus (YPM 24900, x ¥).

tion, nine cervical vertebra. However, Williston and Case (1892, p. 20) state that
in Clidastes the first long rib with a possible sternal connection occurs on the
eighth postcranial vertebra (see also Williston, 1898b, p. 139). Osborn (1899a,
p- 172), and Williston (1910, p. 538) and Huene (1911, p. 49), on the basis of
excellent skeletons of Tylosaurus and Platecarpus respectively, unequivocally
give the number of cervical vertebrae as seven. It is highly probable that mosa-
saurs have two less anterior vertebrae lacking a sternal contact than does
Varanus.

The neural spine of the third cervical vertebra is slender in lateral aspect,
with slightly converging anterior and posterior margins. The posterior buttress
of the spine is still greatly enlarged and is almost as wide as that of the axis,
though the anterior edge is very thin. A transversely oriented elliptical surface,
which was capped by cartilage in life, occupies the dorsal tip of the spine. The
neural spine changes its shape posteriorly so that by the time the last cervical is
reached the spine is parallelogram-shaped in lateral aspect and both longer and
higher. Laterally the spine also becomes very compressed, causing the posterior
buttress to vanish almost completely and the dorsal edge of the spine to lose
nearly all trace of a transverse elliptical expansion on the last cervical. It is
thus apparent that few, if any, fibers of the M. spinalis capitis, which arises on
the dorsal tip of the neural spines, can have extended behind the last cervical,
and that the Mm. interspinalis, spinalis and semispinalis cervicis diminished to
normal proportions in the posterior cervical region. It is interesting that these last

RcPd RCPy

Text-fig. 41. Areas of muscle attachment on anterior portion of vertebral column of Platecarpus.
Abbreviations: ICE, M. iliocostalis cervicis; ITR, Mm. intertransversari; LCC, M. longis-
simus cervicocapitis; LCO, M. longus colli; LCOS, M. levator costae; OCI, M. obliquus
capitis inferior; TR, transversospinalis system. For other abbreviations see text-figure 1.

RUSSELL: AMERICAN MOSASAURS 75

muscles show no unusual development in the anterior cervical region of Varanus.

Zygopophyses are present on the neural arches throughout the cervical series
in all mosasaurs. The anterior zygopophyses arise from a position on either side
of the neural canal and project anteriorly, dorsally and laterally, the articular
facets facing dorsomedially with their centers just in advance of the anterior
rim of the centrum. A strong rounded crest curves anterodorsally from the
synapophysis to support the anterior zygopophysis laterally in Clidastes, Mosa-
saurus, Globidens (Gilmore, 1912, fig. 3), Plotosawrus, Platecarpus, Plioplate-
carpus primaevus, Ectenosaurus, Prognathodon and Halisaurus. It is straight
in Tylosaurus and absent in Plioplatecarpus depressus. This ridge probably
served as an area of origin for the M. longissimus cervicocapitis. The poste-
rior zygopophyses arise from the posterolateral corner of the neural spine and
project posteriorly, ventrally and laterally. A small process of bone commonly
arises immediately above the neural canal on either side of the vertebral mid-
line in mosasurs. It may bear a laterally facing facet for articulation with the
lateral wall of an elliptical excavation close to the vertebral midline and above
the neural canal of the preceding vertebra. These accessory articulations lie
median to the zygopophyses and are called respectively, a zygosphene and a
zygantrum. They are relatively highly developed in Clidastes, “‘Platecarpus”
intermedius (Leidy, 1873, p. 282) and Ectenosaurus where they may attain a size
as great as one-sixth that of the zygopophyses. Functional zygosphenes exist on
the cervical vertebra in the type of Globidens alabamaensis (Gilmore, 1912,

Text-fig. 42. Vertebra of Clidastes showing zygosphene (zy) and zygantrum (zg) (restored after
YPM 24916, x 14). A. Anterior view. B. Posterior view.

p- 484). In M. conodon the zygosphenes are rudimentary on the third and fourth
cervical vertebrae, but are functional, though small, on cervicals five through
seven. Rudimentary zygosphenes also occur on the vertebrae of the anterior
portion of the column in M, missouriensis (Williston, 1895, p. 167). Zygosphenes
are present but seldom functional in the cervical vertebrae of Platecarpus and
are evidently very small or absent in the corresponding region of Plotosaurus,
Plioplatecarpus, Prognathodon, Halisaurus and Tylosaurus. The articulating
surfaces of both zygopophyses and zygosphenes-zygantra are inclined at an angle
of about 45° to the horizontal and, in well-preserved specimens, show traces of
growth rings.

The centra of the cervical and all succeeding vertebrae are procoelous. The
interarticular surfaces are very smoothly finished, indicating a high degree of

76 PEABODY MUSEUM BULLETIN 23

mobility, and are vertical, instead of anterodorsally inclined as in Varanus. These
surfaces are dorsoventrally compressed in Halisaurus; horizontally oval in Glo-
bidens (Gilmore, 1912, fig. 3), Platecarpus, Ectenosaurus and Plioplatecarpus;
and more nearly circular in Clidastes, Mosasaurus, Tylosaurus and perhaps also
Prognathodon. The length of the centra increases slightly from the third to
seventh cervical in Platecarpus and Plotosaurus, remains approximately con-
stant in Clidastes and Tylosaurus, and decreases slightly in M. conodon and
Plioplatecarpus (Dollo, 1894, p. 234). The synapophysis increases uniformly and
greatly in depth from a circular tuber on the axis to a narrow vertically rectan-
gular surface on the last cervical. Here it is about equal to the body of the
centrum in height and is dorsolaterally inclined in Clidastes, Globidens (Gil-
more, 1912, fig. 3), Platecarpus, Ectenosaurus and Halisaurus, or of a slightly
lesser height and laterally directed in Mosasaurus, Plioplatecarpus and Tylo-
saurus or ventrolaterally directed in Plotosaurus. A ridge of bone extends from
the ventral edge of the synapophysis directly forward to the rim of the anterior
central articulation. This ridge lies well above the ventral margin of the poste-
rior cervicals in Mosasaurus, Plotosaurus, Ectenosaurus and Tylosaurus, ap-
proaches the ventral margin in Clidastes, Globidens (Gilmore, 1912, fig. 3),
Platecarpus and Plioplatecarpus, and extends well below the ventral margin in
Halisaurus. The ridge together with the posterior edge of the synapophysis very
likely served as the area of origin for the Mm. levator costae which extended
posteriorly to insert on the anterior surfaces of the heads of the cervical ribs
(see Olson, 1936, p. 298). The anterior and posterior surfaces of adjacent syna-
pophyses were probably linked by the Mm. intertransversarii (Olson, 1936, p.
297). All of the cervical vertebrae except the atlas bear ribs.

As on the axis, a stout peduncle is situated on the posteroventral surface of
anterior cervical vertebrae in all mosasaurs. The flattened distal end, similar to
that of the axis in the various genera, articulates with a squat posteroventrally
recurved hypapophysis. The axis hypapophysis, morphologically equivalent to
the intercentrum of the vertebra behind (Osborn, 1899a, p. 175), is the broadest
and most powerful of the series. Succeeding hypapophyses diminish in size, the
smallest and last one usually occurring on the sixth cervical in Clidastes, Plate-
carpus and Tylosaurus, on the sixth or seventh cervical in M. conodon, on the
seventh cervical of Ectenosaurus and Plioplatecarpus (Dollo, 1894, p. 235), on
the second or third dorsal of Plotosaurus and on the fourth dorsal of M. mis-
souriensis (Goldfuss, 1845, p. 191). There are frequently two or three vertebrae
with rudimentary peduncles following the last hypapophysis-bearing vertebra.
The Mm. longus colli and rectus capitis anterior must have been attached to
these processes, as has been pointed out by Camp (1942, p. 5).

DorsAL VERTEBRAE

The number of vertebrae between the last cervical, here considered as the
seventh one postcranially, and the “sacrum” is quite variable in mosasaurs. In
Platecarpus and perhaps also Ectenosaurus only 22, and in Tylosaurus 22 to 23
dorsal vertebrae are present, thus about equaling the number (21-22) found in
Varanus. Clidastes sternbergi has 24 dorsals, other Niobrara species of the genus
have 35. Thirty-eight dorsal vertebrae are known in M. conodon and Camp
(1942, pl. 5) restores the number of dorsal vertebrae in Plotosaurus at 44. Dollo
(1894, p. 237) states there are only 13 vertebrae, which could here be termed
dorsals, in Plioplatecarpus, an unusually low number. The dorsal column may
be divided into an anterior thoracic series bearing long ribs and a posterior

RUSSELL: AMERICAN MOSASAURS 4H

lumbar series with short ribs. In C. liodontus there are between eleven and
twelve dorsal vertebrae with relatively long ribs that may be called thoracic,
although several of the more posterior ones probably lack sternal connections
(Williston and Case, 1892, p. 25; YPM 1333). Clidastes sternbergi has about
14 such vertebrae (Wiman, 1920, pl. 3), and there are between 17 and 18 present
in M. conodon (Dollo, 1894, p. 233; Martin, 1953). Platecarpus has between 12
and 13 dorsal vertebrae with long ribs (Williston, 1910, p. 538; Huene, 1911, pl.
3; Lambe, 1914, pl. 1; Wiman, 1920, pl. 3), Ectenosaurus may have about 11
(FHM 7937), and Tylosaurus does have 11 (Osborn, 1899a, pl. 23; USNM 8898).

The neural spines of the dorsal vertebrae resemble those of the last cervical
very closely, being narrow, parallelogram-shaped and posteriorly inclined, and
nearly as long as the vertebral centra. In the thoracic series the anterior and
posterior edges of the spine converge distally, but in the lumbars this convergence
is almost absent. The neural spines are generally rather uniform in height and
inclination, although in Clidastes, Platecarpus and. Tylosaurus the spines in the
lumbar region are taller and more vertically oriented than in the anterior regions.
In M. conodon the spines are nearly vertical in the thoracic series; those of the
lumbar region slope more strongly posteriorly (Martin, 1953). Camp (1942, pl.
5) restores the vertebral column of Plotosaurus as having the tallest spines of
the presacral series in the anterior thoracic region.

In Varanus well developed zygopophyses of equal strength extend the full
length of the dorsal series. This is never the case in mosasaurs. Zygopophyses are
fully developed in the cervical region of Clidastes, M. conodon, M. missouriensis,
Platecarpus, Ectenosaurus and Tylosaurus but are reduced steadily in size poste-
riorly, so that vertebrae in the posterior lumbar region have zygopophyses of
about two-thirds the dimensions of those in the neck. This reduction is ac-
celerated in Plotosaurus, where the last functional zygopophyses occur on the
seventh dorsal, and in Plioplatecarpus where they occur on the third (Dollo,
1894, p. 230). Not only do the zygopophyses diminish in size posteriorly but they
also approach the midline more closely. Since the zygopophyses mark the plane
of division between the transversospinalis and longissimus systems (Vallois,
1922, p. 511), it follows that the former group must be correspondingly reduced
in cross sectional area, and therefore in strength, in the posterior portion of the
trunk vertebrae in mosasaurs. The zygosphene-zygantrum articulation is well
developed on dorsal vertebrae of Clidastes and Ectenosaurus and shows little
reduction in the posterior region of this portion of the column. In M. conodon,
however, this articulation becomes non-functional in the region between the
third and seventh dorsal vertebra. Zygosphenes are absent in Plotosaurus, Plio-
platecarpus, Prognathodon and Tylosaurus.

The centra of the dorsal vertebrae increase slightly in length up to the mid-
dle of the series in Clidastes, then decrease gradually in length so that the last
dorsal vertebra is about equal to the first in this dimension. A similar tendency
to reduce the length of the dorsal centra begins in the anterior thoracic region
of M. conodon. In Plotosaurus the vertebral centra increase in length at least as
far back as the eighth dorsal. The longest vertebrae are found near the middle
of the lumbar region of Platecarpus and Tylosaurus where the centra exceed
that of the first dorsal by about 1.3 times in length. Further posteriorly they
rapidly decrease in length although the centrum of the last dorsal is still ap-
preciably longer than that of the first. The radius of curvature of the central
articulations becomes greater posteriorly in all mosasaurs. These articulations
increase in area posteriorly in Clidastes, Platecarpus and Tylosaurus, where they

78 PEABODY MUSEUM BULLETIN 23

are about twice as large as the first dorsal in the posterior lumbar region. This
increase in size is not so great in Mosasawrus and Ectenosaurus. The articular
facets of the anterior dorsal series are circular in Clidastes, M. maximus and
Tylosaurus, transversely oval in M. conodon, Plotosaurus, Platecarpus, Plioplate-
carpus, Ectenosaurus and Prognathodon, and dorsoventrally compressed in
Halisaurus. In the posteriormost dorsal region they take on the outline of an
isosceles triangle with rounded apices and one side facing directly ventrally.

The synapophyses originate on the anterodorsal portion of the lateral sur-
face of the thoracic centra in Clidastes, Mosasaurus, Plotosaurus, Prognathodon
and Tylosaurus, and from high in the center of the lateral surface in Platecarpus
and Plioplatecarpus. They are extraordinarily heavy in an anterior thoracic of
Halisaurus where their distal ends extend posterolaterally to almost the same
vertical plane as that occupied by the posterior central articulation. In the
lumbar region they move to a more central position on the anterolateral sur-
face of the centrum, and in Mosasaurus and Plotosaurus at least, they become
progressively more ventrodistally inclined on the last few vertebrae preceding
the “sacrum.” Because of the usual crushed condition of mosasaur material from
the Niobrara Chalk it is difficult to tell how the synapophyses vary in a transverse
direction in the vertebral column of Niobrara specimens. In two specimens of
Clidastes (YPM 1333, 1335) the synapophyses appear to increase slightly in
length from the anterior to the posterior end of the dorsal series. In one
specimen of Platecarpus (YPM 24905) they seem to be longest near the mid-
dle of the dorsal series, and Osborn (1899a, p. 176) states they are of uniform
length in the lumbar series of Tylosaurus. Martin (1953) notes that the syna-
pophyses of M. conodon are nearly twice as long in the posterior thoracic region
as they are at either end of the dorsal series. In Clidastes, M. conodon (Martin
1953), Plotosaurus and Platecarpus the articular facet for the rib on the distal
end of the synapophyses attains its greatest depth in the region of the first thoracic
vertebra, diminishes to about two-thirds this size before the end of the thoracic
series, and then remains constant in depth to the end of the lumbar series.

In Varanus a tendon (Ligamentum tuberculi costae) arises from the lateral
surface of the anterior zygopophysis and passes ventrolaterally to insert on the
proximodorsal surface of the adjacent rib. In mosasaurs there is a roughened
area just medial to the anterodorsal corner of the synapophyseal facet that may
have marked the origin of this tendon. It would then have followed a more
posterolateral course to insert on the anterodorsal edge of the proximal portion
of the rib. The ridge extending from the ventral border of the synapophysis to
the anterior rim of the centrum in the cervical vertebrae is gradually reduced
in size posteriorly in Clidastes, to disappear finally near the beginning of the
lumbar series between the thirteenth and fifteenth vertebra. In M. conodon it is
already absent on the third thoracic, and in Plotosawrus it disappears before
the fifth thoracic vertebra. It is rudimentary on the first two thoracics of Plate-
carpus, but in Tylosaurus the ridge is present at least as far back as on the

twelfth dorsal.

CAUDAL VERTEBRAE
SACRUM

The ilium has lost all contact with the vertebral column in mosasaurs, mak-
ing the identification of the homologue of the first sacral vertebra a matter of
some dispute. The pelvis has been found associated with vertebrae in the ante-
rior region of the pygal series (Williston and Case, 1892, p. 21; Huene, 1910,

RUSSELL: AMERICAN MOSASAURS 79

pl. 1), with vertebrae near its center (Williston, 1904, p. 51; Capps, 1907, p. 350;
J. T. Gregory, 1964, personal communication), and with vertebrae near the
posterior end (Marsh, 1872b, p. 453; Osborn, 1899a, fig. 1). Largely or entirely
on the basis of vertebral associations several authors (Marsh, 1872b, p. 453;
Cope, 1875, pl. 55; Owen, 1879, p. 56) concluded that the pelvis lay in life
beneath the center or posterior end of the pygal series. Dollo (1894, p. 224) also
held that the pelvis occupied this position, in the conviction that the pygals
were true ribless lumbar vertebrae. Other workers (Williston and Case, 1892,
p. 21; Williston, 1897c, p. 109; 1898b, p. 135; Merriam, 1894, p. 12; Osborn,
1899a, p. 177; Huene, 1911, p. 50; Camp, 1942, p. 8) have believed that the ilia
were connected ligamentously to the transverse processes of the first pygal
vertebra and that this vertebra was indeed the true “sacral.” The latter opinion
finds its strongest support in the fact that the transverse process of the first pygal
vertebra is more than twice as long as the synapophysis of the last dorsal. This is
a very sharp distinction, nearly as clear cut as the sudden interruption of the
dorsal series by the expanded transverse process of the first sacral vertebra in
Varanus (see also Osborn, 1899a, p. 177). It should be noted that a well defined
lumbar region is never present in reptiles (Romer, 1956, p. 228).

CAUDAL VERTEBRAE

The caudal vertebral column of mosasaurs may be divided into three regions,
a basal region where vertebrae lack haemal arches, an intermediate region where
they possess haemal arches and transverse processes, and a terminal region where
transverse processes are absent. Vertebrae of the first type have been called
“pygial’’ by Williston and Case (1892, p. 22) or “pygal’ by Williston (1898b,
p- 141). Since the sacral vertebra is almost indistinguishable from the anteriormost
caudal in mosasaurs, it is here considered as the first pygal vertebra. The follow-
ing list shows known caudal vertebral counts in mosasaurs.

Counts of caudal vertebra: (See systematic section for sources of information)

chvs. and
pygals tv. proc.* terminal** total
Clidastes liodontus 7 26 approx. 46 approx. 79
Clidastes sternbergi 4 ——approx. 72 approx. 76
Mosasaurus conodon 8 21 approx. 54 approx. 83
Plotosaurus 30 5 approx. 64 approx. 99
Platecarpus 5 26-31 approx. 65 96-101
Tylosaurus 6-7 33-34 56-78 95-119

* caudals with chevrons and transverse processes
** caudals with chevrons only

The neural spines at the base of the caudal series are identical to those of the
posterior dorsals. They decrease in height posteriorly, but at a proportionally
lesser rate than the centra ‘shorten, with the result that the spines assume a pro-
gressively slenderer outline in lateral view. Near the beginning of the posterior
half of the tail the neural spines become vertical and slightly elongated in
Platecarpus (on postsacrals 44-52) and Tylosaurus (postsacrals 38-40). They are
nearly vertical and more greatly elongated in this region of the tail in C. stern-
bergi (Wiman, 1920, fig. 7, pl. 3). In C. liodontus (Williston, 1898b, pl. 62 fig 3),
M. conodon (Dollo, 1894, pp. 230-231; Martin, 1953) and Plotosaurus tuckeri
(Camp, 1942, pl. 3) the neural spines become vertical slightly behind the center
of the tail. They then slope anteriorly and increase in length until, on postsacrals

80 PEABODY MUSEUM BULLETIN 23

41-42 in Clidastes, 42-44 in Mosasaurus and 42-44 in Plotosaurus, they again
become vertical and reach a maximum in height amounting to a 50% increase
over that of the spines on preceding caudals. The neural spines then resume a
posterior inclination and steadily decrease in height to the posterior extremity
of the caudal series. In this region of the tail the spines are strongly posteriorly
inclined near the base, and become more vertically oriented and slightly ex-
panded distally. In Tylosawrus and M. conodon (Martin, 1953) the change in
slope distally is more abrupt than in Clzdastes, Plotosaurus and Platecarpus.

Functional zygopophyses are never present except on the anteriormost caudal
vertebrae in mosasaurs. In Clidastes, M. missowriensis (Goldfuss, 1845, p. 192),
M. conodon (Dollo, 1894, p. 230; Martin, 1953), Platecarpus, Ectenosaurus and
Tylosaurus zygopophyseal contacts between successive vertebrae are lost anterior
to the first chevron bearing caudal. Anterior zygopophyses in a very rudimentary
form occur throughout the caudal series. In the posterior portion of the tail of
Clidastes and Platecarpus, and to a lesser degree of Tylosawrus, the lateral sur-
face of each neural arch projects in a flat plane to buttress the sides of the blade
of the neural spine. Such transversospinalis mucles as were present were prob-
ably interrupted laterally by these surfaces and thereby concentrated in an in-
terspinal position between them and above the remnants of the anterior zygo-
pophyses.

The central articulations are subtriangular in outline at the base of the
tail in Clidastes, Plotosaurus, Platecarpus and Tylosaurus and tend to become
subhexagonal as the transverse processes rise on the centrum in vertebrae behind
the pygal series. The central facets in the anterior portion of the tail are rela-
tively high in Clidastes, less so in Tylosaurus, and lower still in Platecarpus.
Anteriorly, the height of the facet does not exceed its width in these three genera,
although behind the dilated portion of the tail the facets become more com-
pressed and take on a vertically oval shape. According to Martin (1953) the cen-
tral articulations of an undistorted vertebral column of M. conodon are sub-
triangular at the base of the tail, changing gradually to a vertically oval outline
on the thirteenth postsacral, to a circular outline on the twenty-seventh, and
to a horizontally oval outline from thence to the end of the tail. The posterior-
most caudal centra of M. maximus are unknown, but facets of those from more
anterior caudal regions are similar to those of corresponding vertebrae in M.
conodon (see YPM 430, 470). Except for a somewhat less depressed outline,
central articulations in the tail of Ectenosaurus appear to be similar to those of
Platecarpus. The central facet of a caudal of Prognathodon (Cope, 1875, p. 154),
with transverse processes and facets for haemal arches, is subhexagonal in shape
and equal in vertical and transverse dimensions. The radius of curvature of
central facets decreases as the vertebrae diminish in size posteriorly but never
becomes proportionally equal to that found in the presacral portion of the
vertebral column.

The caudal centra decrease posteriorly in length but not at a uniform rate.
In complete tails of C. liodontus (Williston and Case, 1892, pp. 23-24), M.
conodon (Martin, 1953) and Platecarpus (YPM 1429) the centra in the pygal
region decrease in length at a rate of about 3% per vertebra; further posteriorly
the rate is reduced to about 1% per vertebra. It may fluctuate in the center of the
tail but generally increases, and towards the terminal end the rate steadily in-
creases to a maximum of about 5% per vertebra. In Tylosaurus (AMNH 221,
Osborn’s specimen; USNM 8898) the pattern is similar, although the rate of
reduction of vertebrae in the preserved portion of the posterior half of the tail

RUSSELL: AMERICAN MOSASAURS 81

of the former specimen is rather low (about 1.59%). The pygal vertebrae of
Plotosaurus diminish at a lower rate (less than 1% per vertebra) than do those
of other mosasaurs, but further posteriorly rates are similar. It must be noted,
however, that in no observed case were the rates of reduction regular enough to
permit an estimation of the number of vertebrae missing in an incomplete tail.

The transverse processes of the pygal vertebrae are dorsoventrally compressed,
narrow alae of bone. They are slightly more than twice as long as the syna-
pophyses of the preceding dorsals in Mosasaurus, Platecarpus and Tylosaurus
and are even more greatly elongated in Clidastes, Amphekepubis and Ecteno-
saurus. The transverse processes arise from the anterior portion of the ventro-
lateral edge of the pygal vertebrae and project ventrolaterally at an angle of
about 40° from the horizontal in Clidastes and M. maximus, about 30° in M.
conodon, Platecarpus and Ectenosaurus, and about 20° in Tylosaurus. In
Amphekepubis they arise from the anteromedial portion of the lateral surface of
the centrum and project laterally at an angle of about 10° to the horizontal. It
is thus evident that the epaxial muscles were more powerfully developed at the
base of the tail in mosasaurs than in Varanus, where the transverse processes
project at right angles from the anterior caudal centra. In view of the fact that
remnants of anterior zygopophyses are very small in the caudal region, it seems
likely that the M. longissimus caudae was very much reduced in mosasaurs, as
it is to some extent in Varanus (Ali, 1949, p. 165), and that the transversospinalis
and especially the iliocostalis caudae systems made up the bulk of the epaxial
musculature.

In the anterior three or four pygals of M. conodon (Martin, 1953), Platecarpus
and Tylosawrus, the transverse processes have a distinct posterior inclination
which is lost posteriorly where the processes meet the centra more or less con-
sistently at right angles. On the lateral surface of the posterior pygal centra in
Platecarpus and Tylosaurus, the transverse processes begin a gradual rise which
culminates in a rudimentary process occupying an anteromedian position on
the lateral surface of the last transverse process-bearing vertebra. In M. conodon,
however, the transverse processes begin to rise on the 20th postsacral vertebra,
thirteen vertebrae behind the end of the pygal series (Martin, 1953), and in
Plotosaurus tuckeri they begin to rise on the 29th postsacral (last pygal). In
Varanus the transverse processes are present to the end of the tail; in mosasaurs
they are absent throughout the posterior half of the caudal series. The epaxial
M. iliocostalis caudae and hypaxial equivalent of the thoracic obliquus system,
the M. iliocaudalis (Nishi, 1938, p. 403), must have been powerfully developed
only in the more anterior regions of the tail. The ventral surface of the transverse
processes also served as part of the area of origin of the Mm. caudifemorales.

Chevrons are present along the entire post-pygal region of the tail in
mosasaurs and rudimentary articulations or arches may be present on the last
pygals. In the Plioplatecarpinae (except Prognathodon overtoni and Dollosaurus)
and Tylosaurinae the haemal arches articulate with centrally pitted, slightly
posterior-inclined peduncles which lie on either side of the ventral midline
nearer to the posterior margin of the centrum than to the anterior. In the
Mosasaurinae and in Prognathodon overtoni (Williston, 1898b, p. 192) and
Dollosaurus (lakovlev, 1901, pl. 5 fig. 5) the haemal arches are fused to the
same region of the caudal centra. The chevrons are relatively longer in the
Mosasaurinae than in other mosasaurs.

In Clidastes and Mosasaurus the haemal arch is subcircular in cross section
and slopes ventromedially to meet its fellow from the opposite side. The lateral

82 PEABODY MUSEUM BULLETIN 23

surface of the arch continues straight onto the lateral surface of the spine
with little or no curving, giving the spine a transversely elliptical cross section
at the base, a circular one near the middle, and a saggitally elliptical cross sec-
tion at the apex. The anterior and posterior surface of the spines are longitud-
inally ribbed and grooved, and the distal end is truncated at right angles to
the long axis of the spine and was covered by cartilage in life. The chevrons
are more strongly posteriorly directed at the base than near the apex in both
of these genera and in Plotosaurus. In M. conodon (Martin, 1953) the spines
are vertical in the anterior region of the tail, slope more and more strongly
posteriorly to meet the centrum, at an angle of 30° to the horizontal vertebral
axis on postsacrals 27 to 57, and then resume a progressively more vertical posi-
tion backwards to the end of the tail. In Clidastes the spines generally slope back
at an angle of about 40° to the horizontal vertebral axis and show less regional
variation in inclination. The spines increase in length from the anterior region
of the tail to reach a maximum on postsacrals 36-41 in C. liodontus (Williston and
Case, 1892, pp. 23-24) and on postsacrals 46-48 in M. conodon (Martin, 1953). In
C. sternbergi (Wiman, 1920, pl. 3) they show no pronounced tendency to lengthen
near the middle of the tail, but in all three species they diminish steadily in length
from this point to the end of the tail. In Plotosawrus the haemal spines are well
developed on the first chevron-bearing caudal, where they slope posteriorly at an
angle of about 35° to the horizontal vertebral axis. They increase in length, be-
coming longest in the region of the 42-44th postsacral. Behind this region they
diminish in length to the tip of tail and assume an inclination of 45°.

The base of each haemal arch is somewhat expanded in Platecarpus, Ecteno-
saurus and Tylosaurus and has a smooth proximal surface with a cone in its
center to articulate with the pitted peduncle on the centrum. The haemal
arches are more strongly medially inclined in Platecarpus and especially in
Tylosaurus, so that the spines are relatively shorter and heavier than in the
Mosasaurinae. The chevrons are straight in a transverse plane and point
posteriorly at a rather uniform angle of about 40° throughout the caudal column
of these two genera. In Platecarpus (Huene, 1911, pl. 3; YPM 1429) the
chevrons show little variation in length back to the 50th postsacral, behind
which they steadily diminish in length to the end of the tail. In Tylosawrus
(AMNH 221, USNM 8898) they show a tendency to diminish in length back
to the 36th postsacral, then increase slightly over a span of approximately seven
vertebrae, and diminish again to the end of the tail. The caudal extension of
the rectus system, the M. ischiocaudalis (Nishi, 1938, p. 403; Ali, 1949, p. 165),
must have linked the entire chevron series together, while part of the Mm.
caudifemorales originated on the more anterior haemal arches.

PECTORAL APPENDAGE
SCAPULA

The base of the mosasaur scapula is longitudinally elliptical in ventral
aspect and is separated into a smaller anterointernal facet articulating with the
coracoid and a larger posterolateral one forming the dorsal half of the glenoid
fossa. The coracoid facet faces directly ventrally and has an unevenly pitted and
grooved surface that was covered with cartilage in life. Inside and in front of
it the scapula and coracoid are firmly united in an interdigitating suture in
Clidastes and M. conodon, but not in Platecarpus, Plioplatecarpus and
Tylosaurus. The smoothly concave glenoid fossa is indistinctly subdivided into
lateroventrally and posteroventrally facing surfaces in Clidastes, M. conodon and

RUSSELL: AMERICAN MOSASAURS 83

Ectenosaurus. In Platecarpus and Plioplatecarpus the glenoid articulation is a
convex, ventroposteriorly facing condyle; in Tylosawrus it faces the same direc-
tion but is gently concave.

Immediately dorsal to its base the scapula is slightly constricted in all direc-
tions and in transverse section tapers to an abruptly truncated, moderately thick
dorsal edge. In lateral aspect the scapula is expanded in both directions above
this neck into an anterior-inclined, fan-like blade. The superior border of the
scapula is gently convex in Clidastes, M. conodon, Plotosaurus, Plioplatecarpus
and Ectenosaurus and is strongly convex in Platecarpus, Plesiotylosaurus and
Tylosaurus. In Clidastes, M. conodon, Platecarpus and Ectenosaurus the anterior
edge of the fan is short, straight and probably nearly horizontal; in Plotosaurus
and Plesiotylosaurus it is relatively longer; and in Plioplatecarpus (Dollo, 1882,
pl. 6 fig. 1) its length nearly equals that of the posterior border. The anterior

Text-fig. 43. Scapula-coracoid of Clidastes liodontus (YPM 1335, x 1). Abbreviations: cor, cora-
coid; em, coracoid emargination; fcor, coracoid foramen; glen, glenoid fossa; sc, scapula.

edge is very short and ventrally recurved in Tylosaurus. In Plioplatecarpus
(Dollo, 1882) the long posterior border of the scapular blade projects posterodor-
sally in a direct line. In Platecarpus, Plioplatecarpus primaevus and Ectenosaurus,
and to a lesser extent in Clidastes, M. conodon, Plotosaurus and Tylosaurus, the
anterior half or more of the posterior border is shallowly emarginated. External
fibers of the M. serratus anterior superficialis may have inserted on the straight
edge of the distal portion of the border. In C. sternbergi (Wiman, 1920, fig. 8)
and Plesiotylosaurus the entire posterior border of the scapula is sulcate, and the
M. serratus anterior superficialis may have shifted its insertion dorsally onto the

84 PEABODY MUSEUM BULLETIN 23

suprascapula. The M. scapulohumeralis posterior probably originated in part
on the rounded edge of this emargination and in part above it on the postero-
dorsal edge of the lateral surface of the scapula and adjacent suprascapula.
Assuming the coracoid was approximately horizontal in life, the scapula lay
in a dorsolaterally and anteromedially sloping plane. The blade of the scapula
is flat in longitudinal and vertical directions in Clidastes; slightly concave ex-
ternally and convex internally in a longitudinal direction and flat in a vertical
direction in Platecarpus; convex externally and concave internally in a vertical
direction and flat in a longitudinal direction in Tylosaurus. In a well preserved
scapula of Platecarpus (YPM 24904) two short horizontal crests may be seen on
the lateral surface of the scapula. The stronger one, representing the origin of
the scapular head of the triceps, occurs near the base of the emargination on the
posterior border of the scapula. The other one is situated near the center of the
emargination and marks the attachment of the proximal end of the scapulo-

Text-fig. 44. Scapula-coracoid of Platecarpus (YPM 24904, x 34).

humeral ligament. The M. scapulohumeralis anterior probably arose on the
anterior edge of the scapular neck and blade. Although its outline is unknown,
a cartilaginous suprascapula was certainly present in mosasaurs. Because the
dorsal rim of the scapula is widest posteriorly along its most elevated position, it
may be assumed that the suprascapula was most extensively developed in this
region. Important muscles supporting the pectoral arch (Mm. omohyoideus,
levator scapulae superficialis, anteriorly; M. serratus anterior, internally; M.

RUSSELL: AMERICAN MOSASAURS 85

serratus anterior superficialis, posteriorly) insert on the suprascapula of Varanus.
The scapular portion of the M. deltoides also arises on the lateral surface of the
suprascapula in this genus.

The greatest eerie dimension of the scapula is generally slightly more than
half the length of the humerus in C. sternbergi and Tylosaurus, nearly equal to
it in Clidastes, Platecarpus and Eerenonuriet and is almost twice the length of
the humerus in M. conodon, Plotosaurus, and Plioplatecarpus. The scapula is
much smaller than the coracoid in C, sternbergi and Tylosaurus, slightly smaller
in Clidastes and Ectenosaurus, about equal to it in size in M. conodon and
Platecarpus, and larger than the coracoid in Plotosaurus and Plioplatecarpus.

CORACOID

The scapular facet on the coracoid is essentially the mirror image of the
coracoid facet on the scapula. The glenoid fossa is smoothly surfaced and
slightly concave in Clidastes, M. conodon and Tylosaurus; pitted, grooved and
convex in Platecarpus; and reduced and convex in Plotosaurus and Plioplate-
carpus (Dollo, 1882, pl. 6 fig. 1). Just medial to the head of the coracoid is an
anteriorly inclined neck region which is relatively short and broad in Clidastes;
longer in M. conodon (Martin, 1953), Platecarpus, Ectenosaurus and Tylosaurus;
and long and narrow in Plotosaurus and Plioplatecarpus. The coracoid thins in
a transverse direction and expands anteroposteriorly into a fan-shaped structure
as it approaches the ventral midline of the body. This fan is anteriorly elongated
so that its anterior border exceeds the posterior border greatly in length in
Clidastes, Platecarpus and Plioplatecarpus, and to a lesser extent in M. conodon,
Plotosaurus, Ectenosaurus and Tylosaurus. Some fibers of the M. scapulohumer-
alis anterior probably extended along the anterior border of the coracoid, as
they do in Varanus. The medial border of the coracoid fan encompasses about
one-third of the circumference of a circle in Platecarpus and Tylosaurus; more
than one-third in Clidastes, M. conodon (Martin, 1953), Plotosaurus, Plioplate-
carpus and Ectenosaurus; and nearly one-half of the circumference in C. stern-
bergi (Wiman, 1920, fig. 8). The posterior tip of the coracoid fan expands
broadly behind the glenoid articulation in Clidastes and Plotosaurus; expands
only slightly behind it in M. conodon, Plioplatecarpus and Ectenosaurus; and
does not reach its level in a transverse direction in Platecarpus and Tylosaurus.
The M. coracobrachialis longus probably arose from the ventral surface of this
fan.

As in Varanus, the ventral surfaces of the coracoid and of the extensive carti-
laginous epicoracoid attached to its convex medial edge (see Osborn, 1899a,
fig. 9), supplied a large area of origin for the pectoral limb adductor musculature.
The M. coracobrachialis brevis probably arose from the broad posterior portion
of the coracoid fan; the M. biceps brachii arose medial to it on the epicoracoid.
The M. supracoracoideus must have been attached to the anteromedial region
of the fan and broad anterior portion of the epicoracoid. Its nerve pierces the
anterior region of the neck of the coracoid. Above the center of origin of the
M. supracoracoideus the coracoid may be either thin or emarginate. In C. stern-
bergi (Wiman, 1920, fig. 8) and C. liodontus the coracoid is emarginate; in C.
propython it appears to be entire. The coracoid is entire in two specimens of
M. conodon (SDSM 452; Dollo, 1894, pl. 3) but not in the type (AMNH
1380). It is entire in M. ivoensis (Williston, 1902, p. 249). Goldfuss (1845, p. 195)
states that the type of M. missouriensis may have an emarginate coracoid, though
that of Williston’s (1895, p. 167) specimen is “apparently not emarginate.” The

86 PEABODY MUSEUM BULLETIN 23

Text-fig. 45. Scapula-coracoid of Tylosaurus (after Williston, 1898b, pl. 46, x 1/3).

coracoid of Plotosaurus tuckeri is entire. In the two Niobrara species of Plate-
carpus the coracoid varies from deeply emarginate to lacking an emargination
altogether. The coracoid of Plioplatecarpus (Dollo, 1882, pl. 6 fig. 1) is widely
emarginate, that of Ectenosaurus (FHM 7937) is slightly notched, and that
of Prognathodon overtoni is deeply emarginate (Williston, 1897a, p. 97). It is
usually entire in Tylosaurus, but may be slightly notched. In any case the M.
supracoracoideus must have been powerfully developed in mosasaurs. The M.
subcoracoscapularis arises over nearly the entire medial surface of the scapulo-
coracoid in Varanus, although the M. sternocoracoideus does insert on the ante-
rior rim of the epicoracoid. There is little reason to think these muscles were
differently arranged in mosasaurs.

CLAVICLE

The clavicle is known only in Plotosaurus but articulations on the inter-
clavicle show that it was present in several other forms as well. In Plotosaurus
the bone is reminiscent of a jugal but is less sharply recurved. The M. deltoides
clavicularis probably took its origin from the posteroventral surface.

INTERCLAVICLE

An interclavicle has been described in C. propython (Holland, 1908, fig. 5),
M. conodon (ibid., figs. 1-4; Martin, 1953), Plotosawrus bennisoni (Camp, 1942,
fig. 3), Plotosaurus tuckeri (ibid., fig. 9), Platecarpus (Williston, 1899; Capps,
1907, fig. 1) and in Plioplatecarpus (Dollo, 1885b, fig. 1).

RUSSELL: AMERICAN MOSASAURS 87

Text-fig. 46. Ventral view of interclavicles of mosasaurs. A. Mosasaurus conodon (after Holland,
1908, fig. 2, x 1%). B. Clidastes propython (after Holland, 1908, fig. 5, x 2%). C. Plate-
carpus (after Capps, 1907, fig. 1, x 34). D. Plotosaurus tuckeri (after Camp, 1942, fig. 9,
Xx 14) with right clavicle.

The interclavicle of Clidastes is a thin, flat element that tapers anteriorly to
a slightly expanded termination. The posterior end is deeply notched medially
and rather heavily longitudinally striated.

In Mosasaurus conodon the dorsal surface of the interclavicle is concave
laterally and anteroposteriorly. The thin posterior end is expanded and medially
notched. The bone continues anteriorly in a long blade which is interrupted
near its anterior termination by a short lateral process from either side. ‘The
shaft then constricts and gives off two asymmetrically diverging processes for
articulation with the clavicles. In Plotosaurus the interclavicle is oval in cross
section and a swelling of the shaft occurs in the region of the lateral processes
of M. conodon, but the bone is in other respects very similar to that of M.
conodon. In Platecarpus the interclavicle is a simple flattened rod with a sug-
gestion of a swelling near its anterior termination. Williston (in Holland, 1908,
p- 163) states that facets for articulation with vestigial clavicles may be present.
The interclavicle of Plioplatecarpus is probably similar to that of Platecarpus,
but the swelling in the anterior portion of the shaft appears to be more distinct.
The bone served as a line of origin for the M. pectoralis.

CARTILAGINOUS STRUCTURES IN PECTORAL REGION
Marsh (1880, p. 83, pl. 1 fig. 1) described an excellently preserved sternum of

88 PEABODY MUSEUM BULLETIN 23

Clidastes liodontus (YPM 1335, see also YPM 1333) as follows: “The sternum in
this genus is narrow, and elongate in form, and nearly or quite symmetrical. . .
It is thin, slightly concave above, and convex below. Its anterolateral margins

Text-fig. 47. A. Flexor aspect of humerus of Plotosawrus tuckeri (after Camp, 1942, fig. 10, x 14).
B. Same of humerus of Mosasaurus conodon (reconstructed after USNM 11904 and 18255,
x %o). Abbreviations for text-figs, 47-49: ect, ectepicondyle; eg. ectepicondylar groove;
ent, entepicondyle; ld, scar for M. latissimus dorsi; pc, pectoral crest; pgp, postglenoid pro-
cess.

are short and rounded and have distinct grooves for the coracoids. The costal
margins are much longer and converge posteriorly. Each has facets for five
sternal ribs. . .”’ Much less well preserved, but probably similar sterna also occur
in M. conodon (Martin, 1953), Plotosawrus tuckeri (Camp, 1942, p. 14), Plate-
carpus (Williston, 1899, p. 40), Ectenosaurus clidastoides (FHM 7937) and
Tylosaurus proriger (Osborn, 1899a, fig. 9).

The sternal ribs are flattened, strap-like structures linking the anterior
thoracic ribs to the sternum. As noted above, five such bands of cartilage are
present on each side of the sternum of Clidastes. At least five pairs occur in Plate-
carpus (see Huene, 1911, pl. 3) and Ectenosaurus (FHM 7937), and according to
Osborn (1899a, p. 180) there are ten in Tylosaurus. Only three pair of sternal
ribs occur in Varanus.

The tracheal rings have been reported in two excellent skeletons taken from
the Niobrara Chalk. Osborn (1899a, p. 181, fig. 8) noted that in Tylosawrus “. ..
the trachea extends from below the axis, is unfortunately destroyed as far back
as the fifth rib, and diverges into two bronchi just behind the coracoids.” Similar
tracheal structures were seen by Williston (1899, p. 39, pl. 12) in a specimen of
Platecarpus.

HUMERUS

The humerus of Clidastes liodontus and C. propython is a rectangular bone,
longer than wide, and expanded at both ends. The glenoid articulating surface
has its longest dimension and is convex in a dorsoventral direction. A tuberosity
for the M. deltoideus lies in front of the glenoid condyle and is separated from
it anteroventrally by a deep groove. The crest upon which the M. pectoralis
inserted arises on the deltoid tuberosity and is directed posterolaterally across
the ventral (flexor) surface of the humerus. As in Varanus, the M. supracora-
coideus probably inserted along the posterior face of this crest; the M. cora-
cobrachialis brevis was probably attached to the flat triangular area between the
crest and the glenoid condyle; and the Mm. humeroradialis and brachialis in-
ferior arose from the anterodistal face of the crest. Behind the glenoid condyle
a large process projects medially to form the posterointernal corner of the hu-

RUSSELL: AMERICAN MOSASAURS 89

merus. The M. subcoracoscapularis inserted on its triangular apex and the M.
scapulohumeralis posterior on its dorsal surface.

A distinct, posterolateral-oriented scar is present on the dorsal (extensor)
surface of the humerus slightly distal to the constricted midregion of the bone.
The M. latissimus dorsi must have inserted on this scar, which separated the
origin of the medial and lateral head of the M. triceps humeralis, as in Varanus.
The M. scapulohumeralis anterior probably inserted on the triangular surface
of bone between the latissimus scar and the anterior and posterior corners of
the glenoid condyle. The distal end of the humerus is divided into two flat
vertical facets. The oval radial facet faces anterolaterally and meets the more
compressed ulnar facet behind it at an angle of about 135°. A prominent, ante-
rior-directed, radial tuberosity (ectepicondyle) probably provided the point of
origin for highly tendonous, posterolateral-directed extensor musculature and
for ligaments extending along the leading edge of the paddle. The ulnar tuber-
osity (entepicondyle) is very well developed and projects posteroventrally from
the posterolateral corner of the humerus. It is capped by a triangular, ridged
and grooved surface that probably anchored connective tissue which bound
together the M. coracobrachialis longus, inserted on its medial surface, and the
distal flexor muscles fanned out anterolaterally from its lateral surface.

In Mosasaurus tvoensis (Williston, 1902, pl. 12), M. conodon and M. mis-
souriensis (Williston, 1898b, p. 147) the pectoral crest occupies a more central
position next to the anteroventral corner of the glenoid condyle. The M. supra-
coracoideus must have found a new insertion on the anterior surface of the
humerus between the deltoid and pectoral crests. In M. ivoensis the humerus has
the same proportions as that of Clidastes liodontus and C. propython but the
postglenoid process is shorter and the radial process appears to be absent. The
humerus is relatively shorter and wider in M. missouriensis (Williston, 1898b,
pl. 32) and especially in M. conodon. All the tuberosities are well developed in
these two species. In Plotosaurus the humerus is very short and broad. The
pectoral ridge appears to be an extension of the ventral portion of the glenoid
condyle onto the middle of the humerus. The radial tuberosity is very small. A
peculiar prominence is developed on the dorsolateral surface of the humerus,
possibly for the origin of the M. triceps humeralis.

In Platecarpus the crests and tuberosities on the humerus are not so distinctly
developed as those in the members of the Mosasaurinae. At the proximal end of

‘Text-fig. 48. Extensor aspect of humerus of Tylosaurus proriger (YPM 4002, x 445). B. Same
of humerus of Platecarpus (YPM 24904, x 14).

90 PEABODY MUSEUM BULLETIN 23

the bone, which was covered with a thick layer of cartilage in life, the postglenoid
and deltoid processes are hardly differentiated from the glenoid condyle. The
pectoral crest is situated in the middle of the proximoventral surface of the
humerus, at the ventral termination of the glenoid condyle, and is widely sepa-
rated from the deltoid tuberosity. The distal end of the humerus is much more
expanded than the proximal end. The lateral surfaces of the radial tuberosity,
the radial and ulnar facets, and the ulnar process are smoothly united in a half-
circle and were also finished in cartilage. In this genus and in Tylosaurus,
above the anterior portion of the radial facet is a small but deep groove on the
dorsolateral edge of the humerus that may have contained the radial nerve. The
humerus of Ectenosaurus is similar to that of Platecarpus but is less well ex-
panded distally; in Plioplatecarpus (Dollo, 1882, pl. 6 fig. 1) it is shorter with

A

m £
Text-fig. 49. A. Extensor aspect of humerus of Mosasaurus conodon (USNM 11904, x %). B.
Same of humerus of Clidastes (YPM 24938, x 24).

a proximal end that is convex, instead of straight, in lateral profile. The humerus
of Plesiotylosaurus, Platecarpus and Ectenosaurus has a broad, squarely trun-
cated head. Although it is longer, with the the head more convex in lateral
profile and the distal end less expanded, the humerus of Prognathodon crassartus
(Cope, 1875, pl. 26 figs. 9a-d) shows a striking resemblance to that of Platecarpus.
In Prognathodon overtoni (Williston, 1898b, pl. 30 fig. 6, pl. 62 fig. 1) the head of
the humerus is more convex and the pectoral crest is evidently smaller and
more anteriorly situated than in the former species.

The humerus of Tylosaurus is very long and slender and, compared to most
other forms, only slightly expanded at the ends. The postglenoid process
is small and indistinguishable from the glenoid condyle. All of these surfaces
were finished in cartilage in life. On the distal end of the humerus the radial
tuberosity is absent, the radial and ulnar facets are indistinct and were also
finished in cartilage, and the ulnar tuberosity is very small. The humerus of
Clidastes sternbergi (Wiman, 1920, fig. 8) appears to be quite similar to that of
Tylosaurus, except for the presence of a well developed spherical glenoid condyle.

In Clidastes liodontus and Mosasaurus conodon the humerus exceeds an av-
erage anterior dorsal vertebra in length by about 1.25 times. In Platecarpus,
Ectenosaurus, Prognathodon and Tylosaurus the ratio is approximately 1:1.80;
in Varanus the humerus is about 4.5 times longer than an average anterior dorsal
vertebra.

RADIUS
In Clidastes liodontus and C. propython the proximal end of the radius is

RUSSELL: AMERICAN MOSASAURS 91

thickened to fit onto the oval radial facet of the humerus. A prominence is
developed on its dorsoanterior surface upon which the M. humeroradialis and
part of the M. biceps brachii inserted. Distally, the radial shaft narrows and
then spreads out into a wide fan. A slightly concave, teardrop-shaped facet for
the radiale occupies its posterodistal edge. The thinner anterior border of the
fan probably supported ligaments arising on the radial tuberosity of the humerus
and passing laterally along the leading edge of the paddle. This portion of the
fan is not so well developed in C. sternbergi.

In Mosasaurus twoensis (Williston, 1902, pl. 12 lower figure) the radius re-
sembles that of Clidastes but is broader and the distal end is less dilated. The
radius is heavy in M. missouriensis (Williston, 1898b, pl. 32) and in M. conodon.
Compared to Clidastes the articulations for the humerus and radiale are propor-
tionally much broader and more nearly parallel each other; the anterior flange is
thin but large, extends further proximally over the radial shaft, and gives the
element a more rectangular outline. The radius of Plotosaurus is similar to that
of M. conodon but the anterior flange is much smaller.

In Platecarpus the shaft of the radius is wide and the element is expanded
at both ends, distally more so than proximally. The articular surfaces are
roughened and were doubtless completed in cartilage. The radius of Plioplate-
carpus (Dollo, 1882, pl. 6 fig. 1) is identical to that of Platecarpus in general
form, but is much more heavily proportioned. The proximal end of the radius
of Ectenosaurus is not expanded, but the bone is otherwise also like that of
Platecarpus. In Prognathodon (Cope, 1875, pl. 26, fig. 11) the radius is very
short and broad, having an almost square outline. A similar radius figured by
Cope (1869-1870, fig. 49, p. 205) that may belong to Prognathodon rapax has a
well developed facet on the posterodistal edge of the bone. Perhaps this was for
articulation with the ulna. The radius of Tylosaurus is long and slender and
has a very small anterior flange.

ULNA

The ulna is approximately hour-glass shaped in Clidastes liodontus and
C. propython although the symmetry of the proximal end is broken by an antero-
lateral-inclined humeral articulation and the presence of a relatively well-formed
olecranon process. The M. triceps probably inserted in large part upon this
process, but a scarred area lying in front of it suggests that a portion of the
muscle inserted further anteriorly. The M. brachialis inferior and part of the
M. biceps brachii were probably attached to the ventral surface of the ulnar
head. A large median facet for the ulnare, bordered by a small one anteriorly
for the intermedium and by another posteriorly for the pisiform, occupied the
distal end of the ulna. The M. pronator profundus and part of the M. flexor
palmaris profundus probably arose from the anterior border of the ulna. The
radius and ulna enclose a circular antebrachial foramen in Clidastes, through
which the median nerve passed to the flexor muscles of the forearm. The ulna
is essentially similar in Mosasawrus conodon and M. missouriensis (Williston,
1898b, pl. 32), but its distal end is more expanded in Plotosaurus. In all three of
the latter forms the antebrachial foramen is more elliptical than in Clidastes.

The ulna of Platecarpus has a wider shaft than that of Clidastes and lacks
a facet anterodistally for the intermedium. In Ectenosaurus the distal end is
more greatly expanded than the proximal, in contrast to all of the foregoing
genera. In this form and in Platecarpus the antebrachial foramen is nearly
circular. The ulna of Plioplatecarpus (Dollo, 1882, pl. 6 fig. 1) is evidently rather

92 PEABODY MUSEUM BULLETIN 23

Text-fig. 50. Flexor aspect of forelimb of Clidastes (YPM 24938, x 845, phalanges and metacarpal
five after Marsh, 1880, pl. 1 fig. 1). Abbreviations for text-figs. 50-51, 53-55: h, humerus;
i, intermedium; mc 1-5, metacarpals 1-5; pf, pisiform; r, radius; ra, radiale; ul, ulna; ula,
ulnare; 1-4, distal carpals 1-4.

RUSSELL: AMERICAN MOSASAURS 93

Text-fig. 51. Forelimb of Mosasaurus conodon (after Martin, 1953, x 14).

peculiar because both the olecranon process and posterodistal portion of the
bone are reduced, giving it a bowed appearance. The antebrachial foramen is
circular and very small. In Tylosaurus and Clidastes sternbergi the long slender
ulna is hardly expanded distally, although the olecranon process is well de-
veloped. The antebrachial foramen in both forms is lenticular.

CARPUS

The carpal elements are very similar in Clidastes liodontus, C. propython,
Mosasaurus conodon and M. missouriensis, and their general shape may be easily
seen from the figures. The radiale is especially wide and a well-ossified inter-
medium and ulnare, as well as a second, third and fourth distal carpal, are
present. In both genera the intermedium enters broadly into the antebrachial
foramen and in Clidastes the radiale may also border on it. The carpal elements
are quite heavy where they meet the radius-ulna, but decrease to half this thick-
ness at their metacarpal articulations. A thin, kidney-shaped pisiform contacts

94 PEABODY MUSEUM BULLETIN 23

Text-fig. 52. Extensor aspect of forelimb of Plotosaurus tuckeri (after Camp, 1942, fig. 10, x 4).

the ulna and ulnare. All articulations are abruptly squared off, and the flexor
and extensor surfaces of the elements are nearly identical. The carpus of Ploto-
saurus is similar, but the intermedium is excluded from the antebrachial foramen
by the radius and ulna. The carpal elements of the type skeleton of C. sternbergt
are somewhat scattered and cannot be surely identified.

In Platecarpus, by comparison to the above, the carpus is much less well os-
sified. Usually only the ulnare, intermedium, and third and fourth distal carpal
are present, although occasionally small osslets may be preserved in the positions
of the radiale and second distal carpal. Unlike the condition in the Mosasaurinae
both the intermedium and ulnare enter equally into the distal border of the
antebrachial foramen. The individual carpals are of about uniform thickness.
The edges are irregularly surfaced and were continued in cartilage; the flexor
and extensor surfaces are more heavily pitted than in the Mosasaurinae. The
carpus of Ectenosaurus is generally very similar to that of Platecarpus but is

RUSSELL: AMERICAN MOSASAURS 95

Text-fig. 53. Flexor aspect of forelimb of Platecarpus (YPM 1426, x 14).

more completely ossified, for a small radiale and a first and second distal carpal
are also present. The four larger elements, the ulnare, intermedium, third and
fourth distal carpal are subequal and broader than those of Platecarpus. Only
an ulnare (?) and osslet in the position of the fourth distal carpal are present in
Tylosaurus.

METACARPALS AND PHALANGES

The outlines of the metacarpals and phalanges of the various genera of
mosasaurs are well illustrated in the text-figures. Clawed terminal phalanges are
uniformly absent. In Clidastes the metacarpals and phalanges are flattened and
diminish gradually in thickness distally. The articular surfaces are sharply
truncated and lie at right angles to the longitudinal axes of the bones. There
appears to have been very little cartilage intervening between the successive

96 PEABODY MUSEUM BULLETIN 23

Text-fig. 54. Extensor aspect of forelimb of Ectenosaurus clidastoides (FHM 7937, X 14, from a
photograph, courtesy of Myrl Walker).

phalanges. The phalangeal formula of Clidastes is not known but may have been
approximately 4-5-5-5-3. The fingers are relatively short for a mosasaur and are
broadly spaced. The metacarpals and phalanges of Mosasaurs conodon are short
and wide, but the digits contain a greater number of elements (the phalangeal
formula is about 9-10-10-10-4) so that the forepaddle is relatively longer than
in Clidastes. The digits also appear to be more closely adpressed (see Dollo,
1894, pl. 3). In Plotosaurus the multiplication of phalanges is comparable (at
least 10-10-10-7-2) to that of M. conodon and the digits are likewise adpressed
to form a longipinnate paddle.

Because of the more highly divaricate fifth finger the foreflipper of Plate-
carpus appears to be somewhat broader than that of Clidastes. The metacarpals
and phalanges tend to be more hour-glass shaped in vertical and horizontal
aspect, rather than flattened, and the phalanges sweep strongly to the rear. The
ends of the phalanges are roughened and were finished in cartilage. The distal
portion of the foreflipper of Ectenosauwrus resembles that of Platecarpus more

RUSSELL: AMERICAN MOSASAURS 97

Text-fig. 55. Flexor aspect of forelimb of Tylosaurus proriger (YPM 4002, x 14; phalanges after
Osborn, 1899a, fig. 9).

closely than those of the Mosasaurinae. The phalangeal formula of Platecarpus
is approximately 4-6-7-5-3, that of Ectenosaurus is 4-4-5-5-(?).

The digits of the foreflipper of Tylosawrus are more compressed than those
of Platecarpus but less so than in Mosasaurus conodon and Plotosaurus. The fifth
digit is only slightly divaricate (see Williston, 1898b, pl. 48) and the remaining
digits sweep only slightly to the rear distally. Proximally the phalanges are
longer, slenderer and slightly flatter than those of Platecarpus but the ends
were similarly terminated in cartilage. The more distal phalanges are relatively
quite broad and flat. The phalangeal formula probably varies between 5-7-9-10-11
and 5-8-8-9-9.

PrLvic APPENDAGE
PELVIC GIRDLE

When the elements of a beautifully undistorted inominate girdle of Mosa-
saurus conodon (USNM 11396) are articulated, several interesting features

98 PEABODY MUSEUM BULLETIN 23

Text-fig. 56. Lateral view of pelvic girdle of Mosasaurus conodon (USNM 11396, x about Y4)
with the elements in natural articulation. Abbreviations: il, ilium; is, ischium; ist, ischiadic
tubercle; obt, obturator foramen; pu, pubis; put, pubic tubercle.

are apparent. The beveled edges of their symphyseal shafts indicate that the
ischia and pubes of opposite sides met on the ventral midline of the body with
a relatively minor amount of intervening cartilage and were oriented in a trans-
verse plane at an angle of about 30° to the horizontal. The pubic shaft is
twisted so that its leading edge is rotated ventrally. Because it is longer than
the ischiadic shaft, the acetabulum assumes a posterolateral inclination. The
symphyseal shaft of the ischium was depressed anteriorly in life, and has a
posterodorsally turned ischiadic process. In this specimen the three pelvic ele-
ments were closely united by connective tissue to form an imperforate acetab-
ulum. Facets on the pubis and ischium indicate that the shaft of the ilium was
vertical in a transverse plane and anteriorly inclined at an angle of about 60°
to the horizontal in lateral aspect.

ILIUM

The ilium is a slender, club-shaped bone in mosasaurs. Its ventral end is
expanded and, in Clidastes and Mosasaurus conodon, divided into a ventral facet
for the ischium, an anteroventral one for the pubis, and a broad, concave,
lateral facet for the acetabulum. These facets are very much less distinct in
Amphekepubis (Mehl, 1930, pl. 65 figs. 2-3), Platecarpus and Tylosaurus. In
M. conodon, Amphekepubis (Mehl, 1950) and Platecarpus there is a distinct
scar on the posterior surface of the ventral end, possibly for the attachment of
a portion of the M. iliocaudalis.

RUSSELL: AMERICAN MOSASAURS 99

The ilium is constricted into a long, slightly anterior-inclined shaft im-
mediately above the acetabular facet. The shaft is straight in Tylosawrus, but the
distal end curves more directly upward in Clidastes, M. conodon, Amphekepubis
(Mehl, 1930), and is short and nearly straight in Platecarpus. The shaft is
elliptical in anteroposterior cross section and was probably suspended between
the vertical sheet of the M. supracostalis anteriorly and that of the M. iliocaudalis
posteriorly. The ilioischiadic ligament probably inserted somewhere near the
center of the posterior border, as it does in Varanus. Origins of the Mm.
iliofibularis, iliofemoralis and iliotibialis were probably compressed into the
posterior, central and anterior portions, respectively, of the ventrolateral region
of the shaft. The iliopubic ligament must have inserted somewhere near the
ventral end of the anterior edge. The distal end of the iliac shaft is abruptly
truncated in Clidastes, M. conodon, Platecarpus and Tylosaurus and was prob-
ably attached by ligaments to the transverse process of the first pygal (sacral)
vertebra. In Amphekepubis the ilium seems to have tapered to a point distally.

PUBIS

The pubis is approximately the mirror image of the ischium in shape, al-
though it is longer and more strongly directed anteromedially. In Clidastes and
Mosasaurus conodon the head of the pubis is divided into three subequal facets,
a dorsoposterior one for the ilium, a posterior one for the ischium and a con-
cave acetabular facet laterally. As in the other pelvic elements these facets are
only indistinctly separated in Amphekepubis (Mehl, 1930, pl. 65 fig. 1), Plate-
carpus and Tylosaurus. Just medial to the head, the body of the pubis constricts
in an anterolateral-posteromedial direction and then curves more directly
medially into a long spatulate shaft.

In many mosasaurs a stout process arises from the anterior border of the
pubis near the head and projects anterolaterally at right angles to the shaft.

Text-fig. 57. Lateral view of pelvic girdle of Platecarpus (after Capps, 1907, fig. 2, x 14) with
the elements lying in a flat plane.

100 PEABODY MUSEUM BULLETIN 23

In Clidastes liodontus and C. propython this process is large and rectangular.
In M. conodon it is smaller and triangular; in Amphekepubis (Mehl, 1930) it is
developed into an enormous process that rivals the pubic shaft in length. The
pubic process is located relatively closer to the head in Platecarpus than in the
above forms and is small, triangular and slightly recurved laterally. In C.
sternbergi (Wiman, 1920, fig. 9) and Tylosaurus the process is represented by
only a small ala of bone on the head of the pubis just in front of the iliac facet.
The puboischiadic and iliopubic ligaments, as well as part of the M. supracostalis,
attach to the distal end of the process in Varanus, and probably did also in
mosasaurs. The M. pubotibialis may have arisen on the dorsal portion of the
base of the process, and the M. ambiens probably arose from the dorsal border
of the pubis nearer to the acetabulum.

The obturator foramen pierces the pubis between the forks of the pubic
process and shaft in the center of the acetabular region of the pubis in M.
conodon and Platecarpus; in C. liodontus, C. propython and Amphekepubis
it lies closer to the acetabular rim. In C. sternbergi (Wiman, 1920, fig. 9) and
Tylosaurus, where the process is rudimentary, the foramen lies closer to the
anterior margin of the pubis in this region. The foramen transmits the obturator
nerve which passes laterally to enervate much of the ventral musculature of the
thigh (Romer, 1942, p. 260). The shaft of the pubis expands medially to about
double its width at the base of the pubic process in Clidastes, M. conodon and
Platecarpus, but in Amphekepubis (Mehl, 1930, pl. 65 fig. 1) and Tylosaurus
its medial end is hardly expanded at all. In M. conodon the anterior rim of the
pubic shaft is twisted ventrally. The M. puboischiofemoralis internus probably
arose on the anterodorsal surface of the pubis behind this rim, and the M.
puboischiofemoralis externus may have attached to its ventral surface.

Text-fig. 58. Lateral view of pelvic girdle of Tylosaurus (after Williston, 1898b, pl. 41 fig. 1, x
about 349) with the elements lying in a flat plane.

RUSSELL: AMERICAN MOSASAURS 101

ISCHIUM

The shaft of the ischium is “A’-shaped in ventral aspect, with its base on
the ventral midline of the body and its apex in the constricted portion of the
shaft, just medial to the expanded, anterolaterally inclined head.

In Clidastes liodontus, C. propython and Mosasaurus conodon the head of
the ischium is divided into three distinct facets, a large dorsolateral one for
articulation with the ilium, a small anterior one for the pubis, and a small,
lateral-directed segment of the acetabulum. These facets are less distinct in
Amphekepubis (Mehl, 1930, pl. 65 figs. 4-5) and are absent on the pitted
articular surface of the head in Platecarpus and Tylosaurus.

Medial to the head the shaft of the ischium is vertically elliptical in cross-
section and constricted anteroposteriorly in M. conodon and Amphekepubis
(Mehl, 1930). The shaft is narrow in this region in C. liodontus and C. propython,
broader in C. sternbergi and Tylosaurus, and very broad in Platecarpus. The
shaft thins vertically and expands anteroposteriorly toward the midline of the
body in Clidastes, M. conodon and Amphekepubis (Mehl, 1930), is more ex-
panded in Tylosawrus, and particularly expanded in an anteromedial direction
in Platecarpus. The ventral surface of the medial portion of the ischium is
longitudinally concave in Clidastes, Platecarpus and Tylosaurus, and flat in M.
conodon and Amphekepubis (Mehl, 1930). The M. puboischiofemoralis externus
arose largely from this surface, and the puboischiadic ligament probably arose
from the raphe beneath the ischiadic symphasis.

A low, rounded crest runs the length of the medial portion of the dorsal
surface of the ischium. At the midline of the body it is beveled off in a vertical
plane which meets the long axis of the ischial shaft at an angle of about 60°.
The M. puboischiofemoralis internus probably arose along that portion of the
dorsal surface anterior to the center of this crest; the M. ischiotrochantericus
probably arose behind it. The ischiadic tubercle projects directly posteriorly
from the dorsal surface of the shaft. It is separated from the head of the ischium
by a short neck in Clidastes, M. conodon, Amphekepubis (Mehl, 1930) and
Platecarpus, and by a much longer neck in Tylosawrus. The tubercle is well
developed in mosasaurs and served as the point of origin for the tendon of the
M. ischiocaudalis and as a point of attachment for the puboischiadic-ilioischiadic
ligament.

FEMUR

The femur in Clidastes liodontus and C. propython is a dumbbell-shaped
bone, expanded at both ends and constricted along the central portion of the
shaft. The head of the femur is probably horizontally oval in uncrushed
specimens. A small, internal trochanter lies at the anteroventral corner of the
proximal end of the bone, its cartilage-capped apex separated only by a narrow
groove from that of the femoral head. As in Varanus, the M. puboischiofemoralis
externus probably inserted behind the apex of the trochanter, and the M.
caudifemoralis was probably attached to the more distal region of the trochanteric
crest. The M. puboischiofemoralis internus must have inserted fleshily onto the
smooth proximodorsal surface of the femur, with the M. iliofemoralis inserting
just lateral to it. The M. adductor femoris probably inserted on the medial por-
tion of the smooth ventral surface of the shaft. The distal end of the femur
is abruptly truncated by two smooth vertical facets, anteriorly a larger lateral-
facing one for the tibia and posteriorly a smaller posterolateral-facing one for

102 PEABODY MUSEUM BULLETIN 23

the fibula. A small tuberosity is present above the anterior portion of the tibial
facet which may have served as a point of attachment for ligaments extending
along the leading edge of the hind paddle. The M. femorotibialis probably
originated on the dorsal surface of the distal end of the femur. The femur is
slenderer and provided with a well-developed, spherical head in C. sternbergi
(Wiman, 1920, fig. 9) but is evidently otherwise similar.

Text-fig. 59. Femur of Mosasaurus hoffmanni (after a cast of the type specimen in the Peabody
Museum at Yale, x 1). A. Ventral view. B. Posterior view.

In Amphekepubis (Mehl, 1930, pl. 66 fig. 1, pl. 67 figs. 1-2) the shaft of
the femur is very slender. The internal trochanter is rather large and located
on the proximoventral surface of the femur medial to the head, with the result
that the long axis of the proximal end of the femur lies at right angles to that
of the distal end. The femur is more massive in M. conodon than in Clidastes
and Amphekepubis. A plaster cast of the type of M. hoffmanni (=?M. maximus)
shows that the femur in this species is more expanded in a vertical direction

Text-fig. 60. Hind limb of Clidastes liodontus (after Williston, 1898b, pl. 34, x 24), ventral aspect.
Abbreviations for figs. 60-63: as, astragulus; cm, calcaneum; fe, femur; fi, fibula; mt 1-5,
metatarsals 1-5; t, tibia; 4, fourth distal tarsal.

RUSSELL: AMERICAN MOSASAURS 103

a
eet
-

~ \
r\' ;
il '
\ \
\
et,
Patt of
ols

Text-fig. 61. Hind limb of Mosasaurus conodon (after Martin, 1953, x about ¥).

proximally, but less horizontally expanded distally, than that of M. conodon.
A very large internal trochanter is situated directly beneath the head of the
femur. The proximodorsal surface of the femur is developed into a median
longitudinal crest in M. hoffmanni.

In Platecarpus, Prognathodon and Tylosaurus the articular ends of the
femur are irregularly ridged and grooved, indicating the existence of a thick
cartilage cover in life. The tibial and fibular facets are fused into a continuous
surface distally, although the region of the tibial articulation is always thicker
dorsoventrally. In Platecarpus, and especially in Tylosaurus, the distal end of
the femur is more expanded in a horizontal plane than is the proximal end.
The internal trochanter is of average proportions and situated on the an-
teroventral edge of the proximal portion of the femur in both genera. In
Prognathodon (Cope, 1875, pl. 26 figs. 10a-d) the proximal and distal ends are
about equally expanded in a horizontal plane, and the internal trochanter is

104 PEABODY MUSEUM BULLETIN 23

moderately large and located in the center of the proximoventral surface of
the femur.

In Varanus, Clidastes, M. conodon, Platecarpus and Tylosaurus the femur
is approximately equal to the humerus in length; in Prognathodon (Cope, 1875,
pl. 26 figs. 9-10) it is about 25% shorter.

TIBIA

The tibia in Clidastes (Williston, 1898b, pls. 34, 36) is a flattened, rectangular
element which is somewhat constricted medially. ‘The ventral portion of the
bone adjacent to the femur is longitudinally striated for the insertions of the
Mm. pubotibialis, puboischiotibialis and flexor tibialis. The opposite surface
is similarly roughened for the attachments of the Mm. ambiens, femorotibialis
and iliotibialis. An anterior flange is incipiently developed on the anterior rim
of the tibia in Clidastes (not in C. sternbergi, however, see Wiman, 1920, fig. 9)
but it is emarginated opposite the central portion of the shaft. This structure
probably anchored ligaments supporting the leading edge of the hind paddle.
A clearly defined astragular facet is present posteriorly on the distal articular

Text-fig. 62. Hind limb of Platecarpus (after Capps, 1907, fig. 3, x 4%).

RUSSELL: AMERICAN MOSASAURS 105

Text-fig. 63. Hind limb of Tylosaurus proriger (after Osborn, 1899a, fig. 12, x ¥%).

106 PEABODY MUSEUM BULLETIN 23

surface. The tibia of Amphekepubis (Mehl, 1950, pl. 66 fig. 2) is nearly identical
to that of Clidastes. The tibia of Mosasaurus conodon is also similar, but the
anterior flange is larger, only slightly emarginate, and the astragular facet slopes
more strongly posteromedially.

In vertical aspect the tibia of Platecarpus is rather dumbbell-shaped. The
distal and proximal ends are about equally expanded, but may be distinguished
in that the distal end is slightly wider and thinner. The articular surfaces are
irregular and were finished in cartilage. ‘The anterior flange of the tibia is
characteristically expanded to give the bone a more rectangular appearance
in Tylosaurus, otherwise the tibia is similar to that of Platecarpus.

FIBULA

The fibula of Clidastes (Williston, 1898b, pls. 34, 36) is a slender, bell-shaped
bone. Its thickened and slightly expanded proximal end served as the site of
insertion of the M. iliofibularis dorsally. The M. pronator profundus probably
arose from the anterior surface of the fan-shaped, distal end. The tibia and
fibula enclose a lenticular crural foramen. The fibula is heavier in Mosasaurus
conodon and its proximal end is more greatly expanded. The crural foramen
is proportionally somewhat wider than in Clidastes. The proximal end of the
fibula in Platecarpus is only slightly less expanded than the distal end, giving
the bone an hour-glass shaped outline. The posterodistal flange is more heavily
developed in Tylosaurus and the distal half of the fibula is not so symmetrical
as in Platecarpus. In both these genera the crural foramen is lenticular.

PES

Only three tarsal elements are known to be present in the pes of mosasaurs.
These include a large astragulus, a slightly smaller calcaneum (absent in
Tylosaurus), and a fourth distal tarsal. The astragulus of Clidastes, Mosasaurus
conodon and Amphekepubis (Mehl, 1930, pl. 67 fig. 5) is broad with the
fibular facet located on a short, medial-directed peduncle. The astragulus of
Platecarpus is nearly circular and has a small emargination proximally for the
crural foramen. This emargination is reduced to a slight notch or is altogether
absent in Tylosaurus.

The metatarsals and phalanges of the known hind limbs are quite similar
to corresponding elements of the forelimb, although the fifth matatarsal is
broader and more strongly hooked than the fifth metacarpal. The digits sweep
rather strongly posterodistally in Platecarpus and less so in M. conodon and
Tylosaurus. The fifth digit of both the manus and pes is relatively long in
Tylosaurus. The approximate phalangeal formula of the pes in M. conodon
is 8-9-9-9-4, in Platecarpus is 4-5-5-5-3, and in Tylosaurus is 5-8-8-8-6.

FUNCTIONAL ANATOMY OF POSTCRANIAL SKELETON
FUNCTIONAL ANATOMY OF THE VERTEBRAL COLUMN

The axial musculature of reptiles is generally oriented in an anteroposterior
direction and extends from the occipital surface of the skull to the posterior
tip of the tail. It is divided by the ribs of the presacral vertebrae and the
transverse processes of the caudals into a dorsal, or epaxial, and ventral, or
hypaxial series. The epaxial series is in turn divided by two longitudinal
tendonous septae, extending dorsolaterally from the region of the zygopophyseal
and costal articulations respectively, into proximal transversospinalis, median
longissimus and lateral iliocostalis systems (see Vallois, 1922). The hypaxial
series may be divided into dorsomedial subvertebralis, lateral obliquus and
ventral rectus systems. In general the transversospinalis system acts as a dorsal
flexor, opposing the action of the subvertebralis and rectus muscles, and the
longissimus (in Varanus), iliocostalis and obliquus systems act as lateral flexors.
The compressive forces generated are sustained by the vertebral centra.

The epaxial musculature is most typically developed in the dorsal region of
lizards (see Nishi, 1916; 1938; Vallois, 1922). The transversospinalis system is
composed of three layers. The most superficial layer is that of the M. semispinalis
dorsi, which arises on the anterolateral region of the prezygopophyses and
medial surface of the dorsal intermuscular septem, and passes anterodorsally to
insert tendonously on the posterodorsal edge of the neural spine of a vertebra
located several segments anteriorly. The more medial M. spinalis dorsi arises
tendonously on the anterodorsal edge of the neural spine and inserts on the
dorsal intermuscular septum several segments anteriorly, and on the neural
spine and arch of the second vertebra anteriorly. A deep series of short seg-
mental muscles (Mm. interspinales, interarticulares, interarcuales) link the
neural arches and spines of successive vertebrae.

In the cervical region of mosasaurs the transversospinalis system was prob-
ably as well differentiated as in Varanus and was surely more powerfully
developed. An M. spinalis capitis originated on the transversely expanded distal
ends of the neural spines of cervicals two through seven and, together with the
superficial portion of the M. rectus capitis posterior (origin: anterodorsal edge
of the axis spine), inserted on the posterior midline of the parietal and served
to flex the head dorsally. An M. rectus capitis posterior profundus, originating
on the anterodorsal tip of the axis spine and inserting on the supraoccipital
and adjacent portions of the paroccipital processes, also served as an accessory
dorsal flexor and played an important role in kineticism in the skull. The M.
obliquus capitis magnus arose from the posterodorsal edge of the body of the
atlas arch, the central portion of the blade of the axis spine, and perhaps also
the central portion of the blade of the third cervical, and passed anterolaterally
to insert on the dilated distal end of the relatively elongate paroccipital pro-
cesses. It must have been a powerful lateral flexor.

Behind the atlas vertebra the wide spacing of the zygopophyses and broad
posterior buttress on the neural spines indicate a massive development of
interspinalis, spinalis and semispinalis muscles. The short anteroposterior length
of the neural spines enabled these muscles to flex the cervical vertebrae strongly
in a dorsal direction. The neural spines become laterally compressed and
abruptly increase in anteroposterior length on the anterior dorsal vertebrae.

107

108 PEABODY MUSEUM BULLETIN 23

From this region back to the middle of the caudal series the zygopophyses
gradually approach the vertebral midline. Consequently both the cross _sec-
tional area, and thereby strength, of the transversospinalis (dorsal flexor) system
and the mobility of the vertebrae in a dorsoventral plane is slowly reduced
posteriorly. In the posterior half of the tail the transversospinalis system was
probably represented only by short segmental muscles covered with a veneer of
connective tissue. It may be worthy of note that in Varanus the neural spines
of the posterior cervical and anterior dorsal vertebrae are elongated, probably
to increase the strength and lever arm of transversospinalis muscles flexing the
neck in a dorsal direction and thereby supporting the head. Presumably be-
cause the head and neck were usually supported by water, this elongation is
not found in mosasaurs (with the possible exception of Plotosaurus).

In Varanus, according to Nishi (1916, pp. 230-231), fibers of the M. longissimus
dorsi originate from the lateral surface of the dorsal intermuscular septum and
anterior zygopophyses and extend anterolaterally to insert on the dorsal faschia
and posterodorsal edge of the head of the rib four segments anteriorly. In this
position it would seem that the muscle would most effectively operate as a
lateral flexor and is so considered here. In many reptiles, including lizards (see
Vallois, 1922; Nishi, 1938, p. 393), it inserts on both epaxial intermuscular septae
and acts to stabalize them against the pull of the transversospinalis and iliocostalis
systems (Vallois, 1922, pp. 109, 158). The arrangement of the M. longissimus
dorsi in mosasaurs is considered to be similar to that of varanids solely on the
basis of the close phylogenetic relationship of the two groups, since most of
the attachments of the muscle are by unpreservable connective tissue instead
of bone.

In Varanus (Nishi, 1916, pp. 231-233) and presumably also in mosasaurs
the M. longissimus cervicocapitis originates on the rounded crest of bone that
rises from the synapophysis to the anterior zygopophysis in the anteriormost dor-
sals and post-axis cervicals. In the region of the third cervical vertebra the longis-
simus divides into superficial (Mm. articuloparietalis, transversalis capitis) and
deep (M. transversalis cervicis) portions. In mosasaurs the former muscles passed
forward to insert on the posterior edge of the suspensorial ramus of the parietal
and on the supratemporal, respectively. In Varanus a tendonous sheet passes
anteroventrally through the two muscles to insert on the posterodorsal edge of
the atlas arch. A well marked scar for this tendon is also present on the atlas
arch of mosasaurs. These superficial longissimus muscles flexed the head in a
dorsolateral direction. There is no reason to think they were particularly well
developed in mosasaurs. It is possible that dorsolateral-trending fibers similar
to those of the M. longissimus dorsi inserted on the heads of the cervical ribs,
which are present on all cervical vertebrae except the atlas in mosasaurs.

Zygopophyses are invariably well developed at least as far back as the mid-
dle of the thoracic series in mosasaurs. Virchow (1914) has shown that they
function to prevent transverse rotation of the vertebrae about their centra, and
thus resolve the rotational component of any muscle attaching to the vertebrae
into a vertical and/or horizontal direction. However, the axis vertebra represents
the anterior limit of the zygopophysis-bearing portion of the vertebral column.
The hemispherical surfaces of the odontoid process and occipital condyle hold
the washer of the atlantal arch-intercentrum in place, which in turn firmly binds
the craniovertebral joint together. Since the atlas arch lacks zygopophyses and a
neural spine, there is no hinderance to rotational movement between the skull
and axis vertebra. Virchow (1914, pp. 70-71) notes that rotatory movements on

RUSSELL: AMERICAN MOSASAURS 109

the order of 20° are possible between the skull and atlas vertebra in Varanus,
and that rotation of a somewhat lesser amount can occur between the atlas and
axis. In Varanus the deep portion of the M. longissimus cervicocapitis (the M.
transversalis cervicis) passes anteroventrally from an epaxial position to insert,
together with the M. iliocostalis capitis, beneath the craniovertebral joint onto
the distal end of the basioccipital tuber. It is evident that the mosasaur skull
was also rotated by these muscles, perhaps aided by a small bundle of fibers from
the transversospinalis system, the M. obliquus capitis inferior, which may have
arisen on the dorsolateral surface of the axis neural arch and inserted ventro-
laterally on the posterodorsal tuberosity on the atlas arch.

Zygopophyses are reduced or non-functional in the lumbar region of mosa-
saurs. Here the interspinalis muscles and ligaments must have aided in prevent-
ing transverse rotation of the vertebrae. Because a significant part of the area of
origin of longissimus fibers was probably on the anterior zygopophyses, because
their trend duplicates that of the iliocostalis system in the dorsal region, and
because in the caudal region there are no costal articulations and therefore no
necessity for the existence of a distinct longissimus system, it is assumed that
the M. longissimus dorsi dwindled in size posteriorly in mosasaurs and its caudal
continuation must have been very much reduced.

In Varanus (Nishi, 1916, pp. 223-224) fibers of the M. iliocostalis generally
originate on the fascia covering the lateral surface of the longissimus and extend
anterolaterally to insert on the posterior edge of a rib four or five segments
anteriorly. The M. iliocostalis cervicis was more strongly developed in mosasaurs
than it is in Varanus, as evidenced by the much larger, posterolateral-directed
atlas synapophysis and the greater number of cervical ribs.

Posteriorly the M. iliocostalis dorsi increased in width to fill the epaxial space
left by the slowly diminishing transversospinalis and longissimus systems. In

A B

Text-fig. 64. A. Cross section of the trunk of a mosasaur in the pelvic region. B. Same of the tail
near its base. Abbreviations: CF, Mm. caudifemorales; ICD, M. iliocaudalis; IL, Mm. ilio-
costalis dorsi et iliocostalis caudae; ILC, M. iliocostalis caudae; IN, interspinalis ligament;
IS, M. ischiocaudalis; L, longissimus system; QL, M. quadratus lumborum; RA, M. rectus
abdominus; SE, Mm. supracostalis et intercostalis externus; TR, transversospinalis sys-
tem.

110 PEABODY MUSEUM BULLETIN 23

the lumbar region it was limited laterally by the shortened ribs, but gained in
depth through the increased ventrolateral inclination of the synapophyses. The
iliocostalis system must have been maximally developed at the base of the tail,
where the M. iliocostalis caudae filled nearly the entire space between the neural
spines and the long ventrolateral-inclined transverse processes. Because the ilium
lay distal to the transverse process of the “sacral” vertebra it was largely if not
completely bypassed by the iliocostalis muscles. ‘This is in marked contrast to the
condition in Varanus where the ilium projects into the expaxial region and
interrupts a much slenderer iliocostalis system. The M. iliocostalis caudae
diminished in strength posteriorly in proportion to the shortening and eleva-
tion of the transverse processes and was probably terminated near the middle of
the tail. Vallois (1922, p. 105) notes that the M. iliocostalis is “par excellence”
a lateral flexor of the body.

The subvertebralis system of the hypaxial series was certainly highly de-
veloped beneath the cervical and anterior dorsal vertebrae, but was probably
unimportant elsewhere. The muscles arose on either side of the small peduncle
on the ventral surface of the anterior dorsal centra and increased greatly in
strength anteriorly, passing under and attaching to the large cervical hypa-
pophyses. Many of the deeper fibers (M. longus colli) terminated anteriorly on
the axis intercentrum, although a few reached the atlas intercentrum. The more
superficial portion of the muscle mass (M. rectus capitis anterior) passed ante-
riorly beneath the craniovertebral joint and inserted on the posterior face of
the large basal tubera of the basioccipital. These muscles were powerful ventral
flexors of the head and neck and acted as antagonists of the epaxial transverso-
spinalis system.

Medial derivatives of the obliquus system probably attached to the vertebral
column of mosasaurs much in the same manner as they do in Varanus (see
Nishi, 1916, figs. 9-10; Gadow, 1882, pl. 6 fig. 7). Mm. intertransversarii ran be-
tween the synapophyses of the presacral portion of the column and served as
lateral flexors. They must have been most powerfully developed where the
synapophyses were longest, in the middle and near the posterior end of the
dorsal portion of the column. Further posteriorly an M. quadratus lumborum
probably originated from the undersurface of the short lumbar ribs and syna-
pophyses and passed back to insert on the ventral surface of the transverse
processes of anterior pygal vertebrae. Mm. levator costae were present in the
posterior cervical and anterior dorsal region, where they originated on the ridge
of bone in front of the ventral margin of the synapophyses and inserted on the
head of the succeeding rib. Because of the presence of cervical ribs, more distal
portions of the obliquus system probably retained a tripartite division at least
as far anteriorly as the base of the neck. Posteriorly, sheets of the supracostals,
aided perhaps by the external intercostals, and superficial portions of the M.
iliocaudalis suspended the vertical ilium between them on the lateral surface
of the body. Behind the pelvis the M. iliocaudalis probably occupied most of the
hypaxial space between the transverse processes and haemal spines and acted as
a strong accessory to the M. iliocostalis caudae above in flexing the base of the
tail laterally.

The rectus system passed from the ventral margin of the mandibles (Mm.
sternohyoideus, omohyoideus) to the pectoral girdle and from thence (rectus
abdominus mucles) back to insert on the pubic tubercle and puboischiadic
ligament of the pelvis. In the lumbar region it served both to support the viscera
and hold the anteroventral rim of the pelvis in place. The caudal equivalent

RUSSELL: AMERICAN MOSASAURS 111

of the rectus system (M. ischiocaudalis) originated from the haemal spines and
inserted on the ischiadic tubercle, near the acetabulum. ‘Thus a vertical, crescent-
shaped space was left free between the pelvis and caudal musculature for cloacal
structures. In view of their role in supporting the pelvis it is unlikely that the
bulk of the obliquus system and rectus system was of much importance in flexing
the body either laterally or ventrally.

FUNCTIONAL ANATOMY OF THE FORELIMB

The pectoral skeleton of mosasaurs lies in the anteriormost portion of the
thoracic region and is basically very similar to that of Varanus. The clavicles,
interclavicle and a well-developed sternum unite both halves of the girdle on
the ventral midline. In most cases the scapula seems to have rotated to a more
anterodorsal position than in Varanus, and the coracoid is more extensive
anteromedially. Cartilagenous suprascapulae and epicoracoids were certainly
present, and although their outlines are not surely known, they were probably
not far different from those in Varanus. It seems reasonable to assume that the
pectoral girdle was not so flattened ventrally as in Varanus, so that the glenoid
fossae were elevated somewhat above the ventral level of the body, but below
the lateral midline. The glenoid fossa faces more posteriorly in mosasaurs than
in varanids and has a sharply defined, hemicylindrically concave- rather than
subcircular-articulating surface in the Mosasaurinae, with the axis lying per-
pendicular to the body surface. The anterolateral rim of the glenoid fossa was
probably completed in cartilage to form a similar articulating surface in the
other two subfamilies of mosasaurs.

The head of the humerus covers the central portion of the proximal end of
the bone and is hemicylindrically convex, fitting perfectly into the glenoid fossa.
Thus the long axis of the humerus was held in a plane perpendicular to the
body surface and projected posteroventrolaterally at an angle of about 45°
from the long axis of the body. This bone is very much shorter than in Varanus,
and the distal end has been turned so that the medial and lateral proximal
processes and the entepi- and ectepicondyles lie in the same plane, perpendicular
to the body surface. The shape of the glenoid articulation suggests that the
humerus was more free to move dorsoventrally than anteroposteriorly.

The articulating surfaces between the humerus and radius-ulna are flat
and at right angles to the long axis of the arm, making the elbow relatively
inflexible dorsoventrally. The epipodial elements are very short and broad and
support the enlarged pectoral paddle through an also relatively inflexible, often
imperfectly ossified, carpal region. It is evident that the only freely moveable
joint remaining in the mosasaurian pectoral apparatus was the articulation be-
tween the head of the humerus and the scapula-coracoid. Muscles linking the
forelimb with the body functioned to displace the forelimb as a unit with respect
to the body. A reconstruction (see text-figs. 66-67) of the musculature of the
forelimb of Clidastes is based on the muscle scars evident on the bones and a
dissection of the forelimb of Varanus niloticus. The termniology used is that
of Romer (1944).

Muscles that abducted the humerus, and therefore the forelimb, were: the
M. latissimus dorsi, arising on the fascia over the neural spines of the posterior
cervical and anterior thoracic vertebrae and inserting clearly on a scar located
on the distomedial dorsal surface of the humerus; the Mm. deltoides clavicularis
et scapularis, arising on the clavicle and suprascapula respectively and inserting
together on the dorsal surface of the lateral process of the humerus; the M.

112 PEABODY MUSEUM BULLETIN 23

Text-fig. 65. Scapula-coracoid of Clidastes showing muscle attachments. Abbreviations for text-figs.
65-67: BI, M. biceps; BIN, M. brachialis inferior; CBB, M. coracobrachialis brevis; CBL,
M. coracobrachialis longus; D, M. deltoides, undivided; DSC, M. deltoides scapularis;
EX, extensors (diagrammatic); FCR, M. flexor carpi radialis; FCU, M. flexor carpi ulnaris;
FPP, M. flexor palmaris profundus; FPS, M. flexor palmaris superficialis;s HR, M. humero-
radialis; LD, M. latissimus dorsi; LIG, ligament; LS, M. levator scapulae superficialis
superior; LSH, Ligamentum scapulohumerali laterali; OMO, M. omohyoideus; P, M.
pectoralis; PP, M. pronator profundus; SCA, M. scapulo-humeralis anterior; SCP, M. scapulo-
humeralis posterior; SCS, M. subcoracoscapularis; SPC, M. supracoracoideus; SS, M. serrati
superficialis; TCC, M. triceps caput coracoideus; TCL, M. triceps caput lateralis; TCM, M.
triceps caput medialis; TSC, M. triceps caput scapularis.

scapulohumeralis posterior, arising on the posterodorsal surface of the scapula
and inserting on the dorsal surface of the medial process of the humerus; and
the M. scapulohumeralis anterior, arising on the anterolateral surface of the
scapula and inserting on the proximomedial portion of the dorsal surface of the
humerus.

Muscles that adducted the forelimb were: the M. pectoralis, arising on the
interclavicle and inserting on the pectoral crest; the M. biceps, arising on the

RUSSELL: AMERICAN MOSASAURS 113

Text-fig. 66. Extensor aspect of forelimb of Clidastes showing restored musculature. A. Superficial
muscles. B. Deeper muscles.

median ventral margin of the coracoid and epicoracoid and inserting on the
proximoventral portion of.the radius and ulna; the M. supracoracoideus, arising
on the anteromedial portion of the coracoid and epicoracoid and inserting on
the anteroventral portion of the lateral process and the posterior surface of the
pectoral ridge in Clidastes, probably inserting on the leading edge of the
humerus and the anterior surface of the pectoral ridge in other mosasaurs; and
the Mm. coracobrachialis brevis et longus, arising behind the supracoracoideus
and beneath the biceps on the ventral surface of the coracoid and epicoracoid
and inserting on the proximomedial portion of the ventral surface of the humerus
and ventroposterior surface of the entepicondyle respectively.

114 PEABODY MUSEUM BULLETIN 23

Text-fig. 67. Flexor aspect of forelimb of Clidastes showing restored musculature. A. Superficial
muscles. B. Deeper muscles.

Muscles that probably acted to pull the forepaddle anteriorly were: the Mm.
scapulohumeralis anterior, deltoides clavicularis, and supracoracoideus, with
the M. subcoracoscapularis, arising on the medial surface of the scapulacoracoid
and inserting on the inner surface of the medial process, acting as an antagonist.
It is apparent, however, that the greatest cross-sectional area of muscle fibers lay
above and below the humerus, rather than in front of or behind it. This fact,
together with the shape of the glenoid articulation, indicates that the mosasaur
forepaddle could be most effectively moved in a dorsoventral, rather than ante-
roposterior direction, and was much more efficient as an organ of steering

RUSSELL: AMERICAN MOSASAURS 115

than as an organ of propulsion. This contrasts strongly with the condition seen
in plesiosaurs (see Watson, 1924).

Because the paddle projected posterolaterally from the side of the body,
when it was held in abducted position, the flow of water streaming over it
would tend to force that side of the mosasaur’s body down. If the leading edge
of the paddle were turned down and the trailing edge up, the surface area
of the paddle meeting the current would be greater and the downward deflecting
force would also be greater. Simultaneous contraction of the “flexor” muscula-
ture of the forearm and scapulohumeralis posterior would accomplish this move-
ment. It is significant that the entepicondyle, the site of origin of the flexor
musculature, is extremely well-developed in mosasaurs and projects well below
the ventral surface of the humerus.

Correspondingly, when the paddle was held in the adducted position, the
co-ordinated action of the deltoids and coracobrachialis longus on the humerus
would elevate the leading edge and depress the trailing edge, to bring a greater
area of the paddle into the water stream beneath it.

FUNCTIONAL ANATOMY OF THE HINDLIMB

Although the mosasaur pelvis is clearly derivable from the ordinary lacertilian
type, there is much less resemblance between it and that of Varanus than between
their respective shoulder girdles. In lateral aspect the pelvis of Varanus is about
as wide as it is high, each of the three elements being broad, powerfully-built
bones. The ilium is strongly posterodorsally inclined and solidly united to the
anteriormost of the two sacral vertebrae. In mosasaurs the pelvis is much higher
than wide and the component elements are slender. The ilium has lost its con-
tact with the sacral vertebra and is vertical or slightly anteriorly inclined in
position. Both the pubis and ischium meet on the ventral midline of the body,
although the latter contact was much the firmer of the two. The acetabulae
were held about midway between the ventral border and the lateral midline of
the body, occupying a position similar to that of the glenoid fossae anteriorly.

There is a series of broad, tendonous sheets found in lizards (see Romer,
1922, 1942), which are suspended upon the lateral surface of the three bones of
the pelvis and surround the acetabulum. In Varanus the puboischiadic ligament
arises on the ventral midline along the ischiadic symphysis and spreads ante-
rolaterally to the pubic tubercle and posterolaterally to the ischiadic tubercle.
From the ischiadic tubercle the ilioischiadic ligament continues anterodorsally
to insert on the posterior border of the ilium. A smaller, tendonous band links
the pubic tubercle with the ilium anteriorly. The relations of these ligaments
were probably essentially similar in mosasaurs (see text-fig. 68).

Because of the loss of the sacroiliac contact, the mosasaur pelvis was sup-
ported entirely by the belly and flank musculature. The epaxial muscles which
in Varanus are partly interrupted by the pelvic girdle, probably were freely
continuous. The pelvis must have been held in place anteriorly by the M.
supracostalis, arising on the ribs along the lateral midline of the body and
inserting on the iliopubic ligament and pubic tubercle; and the M. rectus
abdominus, arising on the sternum and inserting on the anterior edge of the
puboischiadic ligament. Posteriorly the pelvis was probably anchored by the
M. caudifemoralis brevis, arising on the undersurface of the transverse proc-
esses and lateral surface of the haemal arches of the anterior caudal vertebrae
and inserting in part on the ilioischiadic ligament, together with the M. ischio-
caudalis, arising on the lateral surface of the haemal spines and inserting on the

116 PEABODY MUSEUM BULLETIN 23

Text-fig. 68. Pelvis of Clidastes showing supporting musculature. Abbreviations for text-figs. 68-70:
A, M. ambiens; AF, M. adductor femoris; CFB, M. caudifemoralis brevis; CFL, M.
caudifemoralis longus; FET, M. femorotibialis; FTE, M. flexor tibialis externus; FTI, M.
flexor tibialis internus; IF, M. iliofemoralis; IFL, M. iliofibularis; IIL, ilioischiadic ligament;
ILG, iliopubic ligament; IS, M. ischiocaudalis; IT, M. iliotibialis; ITR, M. ischiotrocanteri-
cus; PFE, M. puboischiofemoralis externus; PFI, M. puboischiofemoralis internus; PIL,
puboischiadic ligament; PIT, M. puboischiotibialis; PP, M. pronator profundus; PT, M.
pubotibialis; RA, M. rectus abdominus; SE, Mm. supracostalis et intercostalis externus.

ischiadic tubercle, and M. iliocaudalis, arising on the ventral surface of the
transverse processes and inserting on the posterior edge of the ilium deep to the
ilioischiadic ligament.

In size and shape the hind limb is very similar to the anterior one. Joints
between the propodial (femur), epipodial (tibia-fibula), tarsal and podial ele-
ments have similarly lost their mobility. The joint remaining between the limb
and the limb girdle is different from that found in the forelimb. The acetabulum
is a bowl-shaped structure which opens laterally and a little posteriorly to
receive the convexity capping the proximal end of the femur. The hind limb
was held in a horizontal plane at more nearly a right angle to the long axis of
the body than the forelimb. This probably brought the hind paddle into con-
tact with water undisturbed by the passage of the anterior portion of the body
and thereby enabled it to exert greater control over the equilibrium of the
body.

It is quite evident from a dissection of the hind limb that the femur of
Varanus could be powerfully rotated about its long axis. Muscles linking the
femur with the pelvis wrap partially around the proximal half of the femur to
insert either fleshily or by tendons at low angles to the surface of the bone. Since
the femur is commonly held in a horizontal plane at a high angle to the body

RUSSELL: AMERICAN MOSASAURS 117

Text-fig. 69. Pelvis of Clidastes showing appendicular musculature. Diagonally ruled areas are
muscles cut off in the plane of the acetabulum.

surface in lizards, it is at once apparent that, if the crus were held at right angles
to the femur and the ground, rotation of the femur would impart a forward
motion to the body of the lizard. Speed of progression would depend on the
length of the crus and rate in which the femora could be alternately rotated in
the power and recovery strokes. Perhaps extension of the crus during the re-
covery stroke would reduce the rotational inertia of the hind limb and increase
the efficiency of rapid progression.

The large size of the internal trochanter, the vertically elliptical cross-section
of the proximal femoral shaft, and the circular acetabulum, all suggest that
the mosasaur femur could be rotated in a manner similar to that found in
Varanus. The elliptical cross-section of the femoral shaft together with the
large internal trochanter would have increased the lever arm of muscles, M.
puboischiofemoralis internus, arising on the medial surface of the pubis and
ischium and inserting on the proximodorsal surface of the femur, and M. caudi-
femorale, arising on the undersurface of the transverse processes of the anterior
caudal vertebrae and inserting in part on the anterior and posterior surface of
the internal trochanter, depressing the leading edge and elevating the trailing
edge of the hind paddle. These are the muscles that provide the power stroke

118 PEABODY MUSEUM BULLETIN 23

Text-fig. 70. A. Hind limb of Clidastes showing the arrangement of some of the appendicular
muscles. Solid lines indicate muscles that are dorsal in position, broken lines indicate
those lying more ventrally. B. Cross section of the proximal portion of the femur, showing
muscles that rotate the femur about its long axis.

in rotating the femur of Varanus, where in each case they insert by a double
tendon near the proximal end of the femur.

The complementary movement elevating the leading edge and depressing the
trailing edge could have been brought about through the action of the M.
ischiotrochantericus, arising on the posteromedial surface of the ischium and
inserting on the dorsal surface of the femur, the M. adductor femoris, arising
on the lateral edge of the puboischiadic ligament and inserting on the posterior
surface of the femur, possibly aided by the M. iliotibialis, arising on the antero-
lateral surface of the ilium and inserting on the proximodorsal portion of the
tibia. In Varanus muscles providing the power stroke are more highly developed
than those providing the recovery stroke. In mosasaurs, where the functional
distinction between the two movements would have been greatly lessened, the
two sets of muscles were probably about equally developed.

Because of its relatively lesser anteroposterior dimensions, slenderer propor-
tions and vertical-oriented ilium, the mosasaur pelvis would not have served so
well as a site of origin of muscles swinging the hindlimb in an anteroposterior
arc as it does in Varanus. By far the most effective plane of muscular operation
must have lain in a dorsoventral direction. The M. iliotibialis, M. iliofemoralis,
arising on the ventromedial portion of the lateral surface of the ilium and
inserting on the dorsal surface of the femur, and M. iliofibularis, arising on the
posterolateral surface of the ilium and inserting on the proximodorsal portion
of the fibula, served as simple abductors, and the M. puboischiofemoralis ex-
ternus, arising on the lateral surface of the pubis and ischium beneath the
puboischiadic ligament and inserting on the proximoventral portion of the
femur; and M. puboischiotibialis, arising on the lateral surface of the pubois-
chiadic ligament and inserting on the proximoyentral portion of the tibia,
also probably acted as simple adductors. The M. ambiens, arising on the
dorsolateral extremity of the pubis and inserting on the proximoanterior por-
tion of the tibia, together with the M. pubotibialis, arising near the tip of the

RUSSELL: AMERICAN MOSASAURS 119

pubic tubercle and inserting on the proximoposterior portion of the tibia,
would have operated to advance the paddle, with the M. flexor tibialis complex,
arising on the ilioischiadic and puboischiadic ligaments and inserting on the
proximoposterior portion of the tibia, acting as an antagonist.

In summary, neither the fore nor hind paddles of mosasaurs could have been
effectively used as oars, but must have served instead to orient the body in
swimming. The forepaddles were held at a low angle to the body surface and
abduction and adduction accompanied by minor inversionary or eversionary
movements of the flippers would have enabled them to operate like the elevators
of an airplane, with the axis of rotation extending along the leading edge. The
hind paddles were held at a higher angle to the body surface and rotation of
the entire limb about a medial axis possibly played a role as important in
equilibrating processes as abduction and adduction. The elbow-knee and wrist-
ankle joints had lost their special mobility and the paddles became more uni-
formly resilient throughout. In some forms (Plioplatecarpinae, Tylosaurinae)
the paddles were probably very flexible because of the extensive replacement
of bony tissue by fibrocartilage (see Howell, 1930, p. 218, for a description of
the same phenomenon in the flippers of cetaceans). In an excellently preserved
specimen of Platecarpus from the Niobrara Chalk of Kansas, ‘The edge of the
fleshy portion fof the paddle], which is clearly outlined, follows the sweep of
the radial side of the fingers about an inch distant. Between the fourth and fifth
finger the edge of the membrane is concave to the tip of the fifth finger. Thence
the margin extends broadly inward to the side of the body. The paddle evidently
[was] not pedunculated, but was broadly connected with the body.” (Williston,
1899, p. 40.)

SWIMMING IN MOSASAURS

In previous sections the general morphology of the mosasaur body has been
outlined and movements possible for the various skeletel structures, based on
the nature of articular surfaces and the presumed strength and disposition of
activating musculature, have been evaluated. From this it is clear that mosasaurs
swam by lateral undulations of the posterior portion of the body and tail and
that the flippers were primarily steering organs, having little capacity for pro-
pulsive fore and aft movements. The generally anguilliform body shape of
mosasaurs was not so ideally suited for rapid rectilinear locomotion as is the
teardrop-shaped body and lunate tail of some bony fishes, sharks and porpoises.
One gains the impression from a survey of Breder’s (1926, see especially pp.
177, 237, 264-265, 279) study of locomotion in fishes that an anguilliform body
shape is most efficient in delivering propulsive energy at low rates and at
relatively slower speeds. Conversely a teardrop-shaped body and lunate tail
are most effective in utilizing great amounts of propulsive energy at high rates
of progression. Bainbridge (1961) has found that in a rather intricate way the
maximum speed attainable by a pisciform organism in water is related to its
body length. It seems likely that its larger size played an important role in
enabling a mosasaur to overtake smaller, more streamlined prey. Like ecologically
analogous carnivorous marine mammals, mosasaurs were also probably quite
maneuverable. As Howell (1930, p. 207) pointed out, inability to steer (i.e.,
make sharp turns) in pursuing individual food items would mean a rapid death
by starvation for a larger aquatic carnivore.

The paddles of mosasaurs were well suited to stabilize or rotate the body.
It is interesting to note that the cervical region is long, not shortened as it is in

120 PEABODY MUSEUM BULLETIN 23

cetaceans, and was operated by powerful muscles. A simple lateral bending of
the neck and anterior portion of the trunk was probably sufficient to make slow
turning movements. When making abrupt lateral turns at high speeds, how-
ever, it is postulated that the body was first spun on its side by the forepaddles.
The neck was then arched in the direction of the turn while the forepaddles
were abducted with their flexor surfaces facing into the stream. Simultaneously
the hind paddles were rotated like the diving planes of submarines, forcing the
posterior part of the body in a direction opposite to that in which the fore-
quarters were moving. As a result the entire body pivoted about its center of
inertia and the animal continued at right angles to its former path of motion
(for similar turning movements in pinnipeds see Howell, 1930; Ray, 1963).
In diving the neck and anterior portion of the trunk were flexed and the fore-
quarters were depressed by the adducted anterior paddles. An abrupt abduction
of the hind paddles elevated the posterior region of the body. It is important to
remember that in all of the above movements the tail and most of the trunk
were nearly inflexible in a dorsoventral direction. The axis-atlas-occipital joint
allowed the head to rotate in a transverse plane and must have greatly enhanced
the ability of the animal to capture agile prey at close range.

Bodily proportions vary somewhat between different genera of mosasaurs.
This variation probably indicates correspondingly different swimming (and
thereby predatory) habits. Clidastes must have been less agile than some
mosasaurs because of its elongate body and small flippers. The development of a
caudal fin indicates it was a relatively rapid swimmer. It is suggested that
Clidastes was better suited to subsist on rapidly swimming fishes which were
relatively incapable of sudden evasive movements. The body of Plotosaurus
is generally similar to that of Clidastes, although the narrower flippers and
more highly developed caudal fin imply that it was a faster swimmer. In the
two above genera and in Mosasaurus, a generally much larger animal, it would
appear that relative lack of maneuverability might indicate a propensity to
feed on faster prey or prey that could not dodge easily.

In comparison to the above-mentioned mosasaurines, the body of Platecarpus
is shorter, the paddles are larger and the tail is not so appreciably dilated in a
vertical direction. Platecarpus was probably a slower swimmer than these
genera, but must have been able to make more abrupt turns. Like its probable
recent analogue, the seals, Platecarpus may have fed predominantly on smaller,
highly evasive fishes. Tylosaurus is a large, moderately slender form with nar-
row flippers and a tail almost identical to that of Platecarpus. It was probably
generally similar to Mosasaurus in swimming ability, although not so fast. As
Williston (1925, p. 272) has observed, the bones of Tylosaurus are very can-
cellous and were probably impregnated with fat (as is also the case in whales,
see Howell, 1930, p. 42). Presumably this adiposity increased the buoyancy of
the animal.

Bertin (1958, pp. 1889-1891) notes that neritic fish living on the continental
shelves are commonly reddish or silver-grey. Since this was evidently the habitat
of most mosasaurs, it might be expected that they were similarly colored. Shal-
low water forms and young inhabiting coastal areas could have been more
variably colored.

SYSTEMATICS

A major problem faced in the beginning of this study was the proper application of
86 described species names of American mosasaurs to the hundreds of fragmentary to
nearly complete skeletons in the collections of the American Museum and the Pea-
body Museum at Yale. Because of very poor locality information regarding horizons in
the Niobrara Chalk from which specimens were taken and the scattering of usually quite
fragmentary specimens from other formations it was felt that a typological approach to
the determination of species was the only one possible. The variation in individual
elements of the skeleton was studied in detail. When it was found that a certain suite
of morphological variations was always associated together in a single specimen and
never mixed with variations of another suite, it was assumed that the suites were diag-
nostic of different taxa. Appropriate suites of diagnostic characters are catalogued under
each taxonomic group in the following section.

If possible, all specimens including types were relegated to one or another of the
typological groups, which received the name applied to the oldest designated type
specimen. In many instances cranial material in types was not adequate to permit assign-
ment to a species group. Names based on such specimens are listed in the last part of
the systematic section.

Size was not considered in the delineation of species. A table of measurements of
representative specimens may be found in Appendix B. As it turned out most of the
specimens referable to a given species fall within a rather limited size range. Only in
Clidastes and Tylosaurus were linear measurements found that varied as much as a
factor of two between the smallest and largest known individuals. Juvenile specimens of
mosasaurs are rarely found and those that have been collected are usually very incom-
plete. These specimens cannot be separated from larger individuals by preserved diag-
nostic characters; this suggests that they are immature and do not belong to a smaller
species. For example, the anterior limit of the narial emargination in the dorsal
border of the maxilla occupies relatively the same position in large and small specimens
of Platecarpus and Tylosaurus. The surface of the bones is less highly polished and the
bones are not so well ossified as in presumed adult individuals of the same species.* It is
therefore hoped that errors due to the confusion of species characters with growth
allometry have been avoided.

When the characteristics of the various genera were compared, it was found that
certain groups of genera appear to resemble each other more closely than other mem-
bers of the same subfamily. The resemblance is expressed by the introduction of tribal
categories within the Mosasaurinae and Plioplatecarpinae. These tribes appear to rep-
resent discrete radiations of mosasaurian forms below the subfamilial level of organiza-
tion.

A summary of mosasaur species recognized in this work is as follows:

SuBFAMILY MOSASAURINAE
Tre MOSASAURINI

Genus CLIDASTES (Coniacian-late Campanian, U.S.A.).
Clidastes sternbergi, Santonian(?), central U.S.A.
Clidastes liodontus, Coniacian, central U.S.A.
Clidastes propython, Santonian-middle Campanian, central and southern U.S.A.
Clidastes iguanavus, late Campanian, eastern U.S.A.

* Gregory (1952) has suggested that the small lower jaws in the type of Ichthyornis dispar
may belong to a mosasaur hatchling. The arrangement of the elements in the posterior half of
these jaws, shape of the retroarticular process of the articular, high number of teeth in the
dentary, and excessively small size coupled with the highly polished surfaces of the bones ail
preclude the possibility of the jaws belonging to an infant mosasaur.

12]

122 PEABODY MUSEUM BULLETIN 23

Genus MOSASAURUS (Santonian(?)-Maestrichtian, U.S.A.;
Santonian-Maestrichtian, Belgium; Maestrichtian,
Holland; Campanian of France and Sweden).
Mosasaurus lonzeensis, Santonian, Belgium.
Mosasaurus tvoensis, Santonian(?), central U.S.A.; early Campanian, Sweden.
Mosasaurus gaudryt, Campanian, France.
Mosasaurus conodon, late Campanian or early Maestrichtian, central, southern and east-
ern U.S.A.; Maestrichtian, Belgium.
Mosasaurus missouriensis, early Maestrichtian, central U.S.A.
Mosasaurus dekayi, Maestrichtian, eastern U.S.A.
Mosasaurus maximus, Maestrichtian, southern and eastern U.S.A.
Mosasaurus hoffmanni, late Maestrichtian, Belgium and Holland.

Genus AMPHEKEPUBIS
Amphekepubis johnsoni, Santonian(?), northern Mexico.

Genus LIODON (Maestrichtian, U.S.A.; Campanian,

England, Belgium, and Sweden; Campanian-

Maestrichtian, France).
Liodon anceps, Campanian, England and Sweden; Campanian-late Maestrichtian(?),

France.

Liodon compressidens, Campanian, France and Belgium.
Liodon mosasauroides, Maestrichtian, France.
Liodon sectorius, Maestrichtian, eastern U.S.A.

Genus COMPRESSIDENS
Compressidens fraasi, late Maestrichtian, Belgium.
Compressidens belgicus, late Maestrichtian, Belgium.

Trine GLOBIDENSINI

Genus GLOBIDENS (Campanian, U.S.A.; late Campanian,
Belgium; Maestrichtian, Brazil, Jordan, North Africa
and Timor).
Globidens alabamaensis, Campanian, central and southern U.S.A.; late Campanian,
Belgium.

Tre PLOTOSAURINI

Genus PLOTOSAURUS
Plotosaurus bennisont, late Maestrichtian, western U.S.A.
Plotosaurus tuckeri, late Maestrichtian, western U.S.A.

Genus TANIWHASAURUS
Taniwhasaurus owent, Maestrichtian, New Zealand.

SuBFAMILY PLIOPLATECARPINAE
Tre PLIOPLATECARPINI

Genus PLA TECARPUS (Coniacian-Campanian, U.S.A.;
Campanian, France and Sweden).
Platecarpus tympaniticus, Santonian, southern U.S.A.
Platecarpus coryphaeus, Coniacian, central U.S.A.
Platecarpus ictericus, Santonian-early Campanian, central U.S.A.; early Campanian,
France.
Platecarpus somenensis, Campanian, France, Sweden and central U.S.A.
“Platecarpus” intermedius, early(?) Campanian, southern U.S.A.

Genus ECTENOSAURUS
Ectenosaurus clidastoides, Santonian, central U.S.A.

Genus PLIOPLATECARPUS (Campanian-Maestrichtian, U.S.A.;
Maestrichtian, Belgium; Campanian, Sweden).

RUSSELL: AMERICAN MOSASAURS 123

Plioplatecarpus primaevus, Campanian, central U.S.A.
Plioplatecarpus depressus, Maestrichtian, eastern U.S.A.
Plioplatecarpus marshi, late Maestrichtian, Belgium.
Plioplatecarpus houzeaui, late Maestrichtian, Belgium.

TriE PROGNATHODONTINI

Genus PROGNATHODON (Campanian, central U.S.A.; Maestrichtian,
central, southern and eastern U.S.A.; late Maestrichtian,
Belgium).
Prognathodon crassartus, Campanian, central U.S.A.
Prognathodon overtoni, early Maestrichtian, central U.S.A.
Prognathodon solvayt, late Maestrichtian, Belgium.
Prognathodon giganteus, late Maestrichtian, Belgium.
Prognathodon rapax, Maestrichtian, eastern U.S.A.
Prognathodon(?) donicus (see Pravoslavley 1914).

Genus PLESIOTYLOSAURUS
Plesiotylosaurus crassidens, late Maestrichtian, western U.S.A.

Genus DOLLOSAURUS
Dollosaurus lutugini, late Campanian, southern European Russia.

PSUBFAMILY PLIOPLATECARPINAE incertae sedis

Genus HALISAURUS (Santonian(?)-Maestrichtian, U.S.A.)
Halisaurus onchognathus, Santonian(?), central U.S.A.
Halisaurus platyspondylus, Maestrichtian, eastern U.S.A.

SuBFAMILY TYLOSAURINAE

Genus TYLOSAURUS (Coniacian-Campanian, U.S.A.).
Tylosaurus nepaeolicus, Coniacian, central U.S.A.
Tylosaurus proriger, Santonian-Campanian, central U.S.A.; Campanian, southern U.S.A.

Genus HAINOSAURUS (Santonian-Maestrichtian, Belgium).
Hainosaurus lonzeensis, Santonian, Belgium.
Hainosaurus bernardi, late Maestrichtian, Belgium.

In the following material, double asterisk (**) preceding a specimen number indi-
cates a complete or nearly complete skull, a single asterisk (*) a fairly complete skull,
and no asterisk indicates a fragmentary skull. In the case of specimens with exceptionally
complete postcranial material the specimen numbers are italicized.

SUPERORDER SQUAMATA
OrpbER SAURIA
INFRAORDER PLATYNOTA
Famity MOSASAURIDAE Gervais 1853
SuBrFAMILY MOSASAURINAE (Gervais 1853) Williston 1897

Clidastidae Cope, 1869b, p. 258.

Edestosauridae Marsh, 1876, p. 59, nomen nudum.
“mosasauriens mesorhynques” Dollo, 1890, p. 163.
Mosasauridae Williston, 1895, p. 169.
Mosasaurinae Williston, 1897d, p. 177.
Globidensidae Dollo, 1924, p. 188.

Dracnosis. Small rostrum present or absent anterior to premaxillary teeth. Fourteen
or more teeth in dentary and maxilla. Cranial nerves X, XI and XII leave lateral wall

124 PEABODY MUSEUM BULLETIN 23

of opisthotic through two foramina. No canal or groove in floor of basioccipital or
basisphenoid for basilar artery. Suprastapedial process of quadrate distally expanded.
Dorsal edge of surangular thin lamina of bone rising anteriorly to posterior surface of
coronoid.

At least 31, usually 42-45 presacral vertebrae present. Length of presacral series
exceeds that of postsacral, neural spines of posterior caudal vertebrae elongated to form
distinct fin. Haemal arches fused to caudal centra. Appendicular elements with smoothly
finished articular surfaces, tarsus and carpus well ossified.

Discussion. Because the fifth digit of the hind foot was not preserved in any of the
known Belgian specimens of Mosasaurus lemonniert (M. conodon), Dollo (1894) as-
sumed that it was tetradactylate in this species. Apparently relying on Dollo’s assumption
and lacking positive evidence to the contrary, Williston (1895, 1897d, 1898b) diagnosed
the Mosasaurinae as all having tetradactylate hind feet. However, Martin (1953) noted
the presence of a fifth digit in his Pierre specimen of M. conodon and it is in all prob-
ability present in earlier members of the subfamily as well.

Tre MOSASAURINI (Gervais 1853) new

Dracnosis. Twelve or less pygal vertebrae present. Radius and ulna widely separated
by bridge of carpalia on distal border of antebrachial foramen. The skeletal areas in
which these characters occur are as yet unknown in Amphekepubis, Liodon and Com-
pressidens. The known morphology of these genera, however, is so similar to that found
in Clidastes and Mosasaurus that in the absence of information to the contrary they are
included in the Mosasaurini.

Grnus CLIDASTES Cope 1868
(Text-figs. 42, 49B, 50, 65-70)

Clidastes Cope, 1868a, p. 181.
Edestosaurus Marsh, 1871a, p. 447.

Generic Type. Clidastes iguanavus Cope.

ADDITIONAL REFERENCES. Cope, 1869a, p. 233; 1869b, p. 258; 1869-1870, p. 211;
1871f, p. 412; 1872d, p. 266; 1872f, p. 330; 1874, p. 31; 1875, p. 130, pl. 19 fig. 11, pl.
55; 1878, p. 301. Marsh, 1872a, p. 291; 1872b, p. 463, pl. 11 figs. 1-2, pl. 12 fig. 1;
1872d, p. 496; 1897, fig. 23. Leidy, 1873, p. 281, pl. 34 fig. 11. Owen, 1877, pp. 690, 696,
701. Lydekker, 1888, p. 272. Hoffman, 1890, p. 1321. Williston and Case, 1892, p. 16.
Merriam, 1894, pl. 3, figs. 9-10. Williston, 1897d, pp. 179, 181; 1897e, p. 245; 1898b,
p. 195, pl. 27 figs. 2-4, pl. 28 fig. 6, pl. 29 fig. 2, pl. 39 lower figs., pl. 42 figs. 5-6, pl. 47
left fig., pl. 54 figs. 4-5, pl. 57 fig. 3; 1914, pp. 165, 166, fig. 68; 1925, p. 273. Hill, 1901,
p. 328. Camp, 1923, p. 322; 1942, pp. 35, 37-40, fig. 24E. Gilmore, 1928, p. 87. Nopcsa,
1928, p. 177. Lane, 1947, p. 318. McDowell and Bogert, 1954, pp. 132, 138. Romer, 1956,
p- 561, fig. 182F.

DrAcnosis. Premaxilla with or without small rostrum anterior to premaxillary teeth.
Fourteen to eighteen teeth in maxilla. Prefrontal forms small portion of posterolateral
border of external nares, broad triangular ala projects laterally from supraorbital
wing. Prefrontal and postorbitofrontal widely separated above orbits. Lateral margins
of frontal nearly straight and converge anteriorly, median dorsal ridge weak. Ventral
process of postorbitofrontal to jugal confluent with broadly exposed dorsal surface of
postorbitofrontal. No ventroposterior process on jugal. Parietal foramen small, located
entirely within parietal. Margins of dorsal parietal surface parallel one another and
cranial midline to posterior base of diverging suspensorial rami, forming narrow rectan-
gular field medially on parietal. Squamosal sends abbreviated wing medially to contact
ramus from parietal. Otosphenoidal crest on prootic covers exit for cranial nerve VII

RUSSELL: AMERICAN MOSASAURS 125

laterally. Fourteen to sixteen teeth in pterygoid. Suprastapedial process of quadrate
moderately large; tympanic ala very thick. Stapedial pit elliptical in form. Sixteen-18
teeth in dentary. Small projection of dentary anterior to first dentary tooth. Medial
wing from angular contacts or nearly contacts coronoid. Dorsal edge of surangular
very thin lamina of bone rising anteriorly to position high on posterior surface of
coronoid. Retroarticular process of articular triangular in outline with heavy dorsal
crest. Mandibular teeth usually compressed, bicarinate and with smooth enamel sur-
faces.

Vertebral formula (Clidastes liodontus): 42 presacral vertebrae, 7 pygals, 26 caudals
with chevrons and transverse processes, about 46 terminal caudals (see Williston, 1898b,
p. 143; Merriam, 1894, p. 10; and AMNH 192).

Articulating surfaces of cervical and anterior dorsal centra circular. Synapophysis
located high in anterior portion of lateral surface of cervical centra, occupies anterodorsal
portion of lateral surface of dorsal centra. Ventral border of anteroventral extension of
synapophysis well developed on cervicals and anterior dorsals, reaches level of undersur-
face of centrum. Anterior zygopophysis of cervicals and dorsals connected by gently
rounded, posteriorly descending crest to synapophysis. Zygosphene-zygantrum strongly
developed, present at least throughout anterior three-fourths of dorsal series. Anterior
base of atlas neural spine arises behind condylar facet; atlas synapophysis long and
tubular. Hypapophyseal peduncle located posteriorly on ventral surface of cervical
centra, articulation for hypapophysis flat and circular, strongly inclined posteriorly.
Five hypapophysis-bearing cervicals, two or three more with rudimentary peduncles.
Transverse processes of pygal vertebrae relatively long. Neural spines of caudal vertebrae
longest and vertical on postsacrals 41-42 (Williston, 1898b, pl. 72).

Scapula smaller than coracoid. Glenoid articulating surfaces concave on both scapula
and coracoid, and smoothly continuous. Superior border of scapula gently convex,
posterior border slightly emarginated posteromedially. Coracoid expands broadly be-
hind glenoid articulation. Distal and proximal ends of humerus widely expanded, facets
for articulation with other elements and sites of muscle attachment very well dif-
ferentiated. Proximal end of radius slightly expanded, ends in flat oval facet for
articulation with humerus. Shaft of radius narrow. Distal end bears greatly expanded
anterodistal flange, is thickened medially to form distinct facet for articulation with
radiale.

Radiale very large. Intermedium shallow. Ulnare does not enter posteroventral
border of antebrachial foramen. Metacarpal one less than metacarpal two in length,
has moderate anterodistal flange resembling that of radius. Proximal ends of metacar-
pals greatly expanded. Phalangeal formula of manus estimated at 4-5-5-5-3.

Acetabular surfaces of pelvic elements concave, form solid smooth-surfaced bowl.
Obturator foramen located near anterior margin of proximal end of pubis, dorsoanterior
process large and rectangular. Ischiadic tubercle separated from acetabulum by short
neck, ischiadic symphasis very narrow. Distal and proximal end of femur about equally
expanded, internal trochanter small and located anteromedially from head. Fibular
facet on femur well marked, tuberosity present on lateral surface dorsal to tibial facet.
Tibia narrow, cartilage capped area of anterior flange extends from base of tibia nearly
one-half way up shaft (Williston, 1898b, pls. 34, 36). Proximal end of fibula very
slightly expanded, distal end fan-shaped (Williston, 1898b). Cartilage capped area of
posterodistal flange of fibula limited to distal end of bone (Williston, 1898b).

Astragulus reniform, facet for fibula distinctly raised above that for tibia, dorsal
notch between facets large (Williston, 1898b, pls. 34, 36). Calcaneum and fourth
tarsal equal in size (Williston, 1898b). Metatarsal one expanded proximally, somewhat
medially recurved. Metatarsal five hook-shaped, concave medially with wide convex
ala laterally (both statements after Williston, 1898b).

Discussion. Vertebrae of the Niobrara species of Clidastes closely resemble the type
vertebra of C. iguanavus, the generotypic species, and certainly appear to be congeneric
with it. All of the species from the Niobrara that have been referred to Edestosaurus be-
long to C. propython (see p. 128). Except for the cranium, the foregoing diagnosis is

126 PEABODY MUSEUM BULLETIN 23

based exclusively on C. propython and C. liodontus. It does not exactly apply to either
C. iguanavus or C. sternbergi, the differences being noted under the respective headings
of these species.

Clidastes iguanavus Cope 1868
Clidastes iguanavus Cope, 1868a, p. 181.

Type. YPM 1601, from “. . . a marl pit near Swedesboro’, Gloucester Co., N. J.”
(Cope, 1869a, p. 233). Type specimen consists of a vertebra.

DisrriBuTion. Marshalltown Formation, New Jersey.

ADDITIONAL REFERENCES. Cope, 1868b, p. 734; 1869b, p. 258; 1869c, p. 86; 1869-1870,
p- 220, pl. 5 fig. 3; 1872f, p. 330; 1875, p. 266. Marsh, 187la, pp. 447-449. Hoffman,
1890, p. 1322. Williston and Case, 1892, p. 16. Merriam, 1894, p. 36. Williston, 1898b,
p- 196. Miller, 1955, pp. 904, 909.

Discussion. The type of the species is a single vertebra from the anterior thoracic
region showing the general proportions and strong zygosphene-zygantrum articulation
characteristic of Clidastes material from the Niobrara Chalk. It differs from vertebrae of
Niobrara forms in the central articulations which are kidney shaped in outline, with
a stronger emargination dorsally for the spinal cord, and in the relatively stouter pro-
portions of the centrum. The vertebra was collected in sediments of late Campanian
age and is thus from a horizon comparable to that of the Lower Pierre Formation in
the Interior, in which Clidastes remains also occur.

Clidastes sternbergi Wiman 1920
(Text-figs. 71, 75C)

Clidastes sternbergi Wiman, 1920, p. 13, figs. 4-9, pls. 3, 4 figs 5a-b.

Tyre. In the Paleontological Museum in Upsala, Sweden, from “. . . Beaver Creek,
Bilby’s Ranch, in Logan Co., Kansas . . .”” (Wiman, 1920, p. 13), collected by Levi Stern-
berg. Type specimen is a nearly complete skull and skeleton.

DistriBuTion. Smoky Hill Member, Niobrara Formation, Kansas.

ADDITIONAL REFERENCES. Abel, 1922, fig. 270; 1924, fig. 35. Dollo, 1924, p. 198.

Diacnosis. Premaxilla “U’-shaped in horizontal cross-section, no rostrum present
anterior to premaxillary teeth. Premaxillo-maxillary suture rises vertically from ventral
jaw margin, turns abruptly and continues posteriorly in nearly straight line to position
dorsal to eleventh maxillary tooth. Suture smoothly keeled and parallels longitudinal
axis of maxilla. Median dorsal surface of parietal very broad. Parietal foramen

Text-fig. 71. Lateral view of muzzle of Clidastes sternbergi (USNM 3777, X 2).

moderately large, opens ventrally into lenticular excavation in parietal length of which
exceeds that of dorsal opening by about three times. Foramen for cranial nerve VII
enters prootic in center of prootic incisure. Infrastapedial process absent on quadrate.
Vertebral formula: 31 presacral vertebrae, 4 pygals, 72 chevron bearing caudals
(Wiman, 1920, p. 17).
Neural spines of caudal vertebrae longest on postsacrals 30-32, do not become vertical.
Scapula much smaller than coracoid. Superior border of scapula gently convex,

RUSSELL: AMERICAN MOSASAURS 127

posterior border emarginate. Coracoid expands broadly behind glenoid articulation.
Distal and proximal ends of slender humerus only slightly expanded, radial and ulnar
tuberosities absent, well-developed spherical head present. Radius elongate, proximal
end slightly expanded. Shaft of radius narrow. Distal end bears moderately well-de-
veloped anterodistal flange.

At least four ossified elements in carpus. Metacarpal one equal to metacarpal two
in length, anterodistal flange small or absent. Proximal ends of metacarpals, especially
of two and three, not greatly expanded.

Acetabular surfaces of pelvic elements convex, do not form solid, smoothly surfaced
bowl. Obturator foramen located near center of proximal end of pubis, dorsoanterior
process rudimentary. Ischiadic tubercle separated from acetabulum by short neck. Shaft
of femur slender, distal end more expanded than proximal, well-developed spherical
head present. Distinct facets for tibia and fibula distally. Tibia and fibula slender,
slightly expanded at distal and proximal ends.

Discussion. The following cranial peculiarities are common to the type of Clidastes
sternbergi and the only specimen (USNM 3777) here referred to this species; a) the
parietal is unusually short and broad, and b) the surangular and articular meet at
the bottom of a notch on the lateral wall of the lower jaw, below the center of the
quadratomandibular articulating surface. These highly distinctive characters are un-
known in any other described species of Clidastes and necessitate the assignment of
the above specimen to Wiman’s species. However the type differs from this specimen
in that the supra- and infrastapedial processes of the quadrate are broadly fused, and
the dorsal margins of the parietal meet posteriorly on the midline of the skull, forming
a “V”-shaped parietal surface. The quadrate looks pathologic (Wiman, 1920, fig. 5)
and perhaps the converging margins of the parietal are due to faulty restoration, for
Wiman (1920, p. 13) mentions that much of the surface of the type skull has been
covered with a thin veneer of plaster.

Wiman’s type skeleton of C. sternbergi presents some very primitive features. The
humerus and femur retain a fully-developed spherical head. The pro- and epipodial
elements are slenderer than those of any known mosasaurian, approaching their condi-
tion in terrestrial lizards, as Wiman (1920, p. 18, pl. 4) has demonstrated. Only four
pygal vertebrae are present which, assuming none have been lost, is the least number
known for any mosasaur. If the elements identified by Wiman (1920, fig. 4) as nasals
are not the vomers which have been crushed upwards into the external narial open-
ing, the nasals then are very large and certainly present a very primitive appearance.

Dollo (1924, p. 198) suggested that Wiman’s skeleton was a composite, but in view
of the fact that characters separating this specimen from any other described species of
mosasaur occur throughout its skeleton this possibility is unlikely. Clidastes affinities
are shown in the elongation of the caudal spines near the center of the tail, the long
fused haemal arches, and the development of a broad anterodistal flange on the radius.
If the vertebral column is complete, as would seem to be the case, the small number
of thoracic vertebrae together with the small scapula and slender limbs may justify the
establishment of a new genus separate from Clidastes for this species.

Clidastes liodontus Merriam 1894
(Text-figs. 26, 32-34, 43, 60, 72, 73, 75B, 76B, 98A; Plate I fig. 1)

Clidastes liodontus Merriam, 1894, p. 35.

Tyrr. Formerly at Bayerische Staatssammlung fiir Palaontologie but probably
destroyed during the Second World War (personal communication, R. Dehm 1963), from
“|. Niobraraschichten der oberen Kreide von Logan County, Kansas . . .” (Merriam,
1894, p. 3), collected by C. H. Sternberg or G. Baur. Type specimen consists of maxillae,
premaxilla and dentaries.

DistRIBUTION. Smoky Hill Member, Niobrara Formation, Kansas.

128 PEABODY MUSEUM BULLETIN 23

ADDITIONAL REFERENCES. Marsh, 1880, pl. 1 fig. 1. Owen, 1880, pl. 8 fig. 1. Williston
and Case, 1892, pls. 2-3. Williston, 1893a, p. 83, pl. 3; 1895, pl. 17 fig. 1; 1897c, pl. 13
fig. 1; 1898b, p. 204, pls. 10-12, pl. 24 fig. 6, pl. 31 fig. 6, pls. 33-34, pl. 60 figs. 6-7; pl.
62 fig. 3, pl. 71, pl. 72 fig. 1; 1925, fig. 146. Ballou, 1898, figs. 5-6. Osburn, 1906, pl. 8 fig.
26. ?Pompeckj, 1910, p. 136. Wiman, 1920, figs. 8a-b. Gregory, 1951, fig. 6. Romer, 1956,
fig. 65. Edmund, 1960, p. 89, figs. 17, 25a-b. Kauffman and Kesling, 1960, fig. 6c. Rus-
sell, 1964, figs. 2-4.

REFERRED SPECIMENS. AMNH nos. **192, **1548. YPM nos. **1333, 1334, **1335,
3982, *3996, 4001, 24914, and three unnumbered specimens. USNM nos. 6546, **11647,
11719. MCZ nos. 278, 1618, 1621. KU nos. 1039, 1121. FHM no. **10668. PU no. 17249.

Diacnosis. Premaxilla “V’’-shaped in horizontal cross-section, small rostrum present
anterior to premaxillary teeth. Posteroventral portion of root of second premaxillary
tooth not exposed on sutural surface with maxilla. Premaxillo-maxillary suture rises
posteriorly to position varying from dorsal to fourth to dorsal to sixth maxillary tooth
and parallels longitudinal axis of cranium. Fourteen to fifteen teeth in maxilla. Median
dorsal surface of parietal narrow. Parietal foramen small, close to or distinctly separated
from frontal suture. Parietal foramen opens ventrally into brain cavity without broaden-
ing into wide excavation. Anterior border of prootic descends beneath prootic incisure
without forming shelf. Foramen for cranial nerve VII leaves brain cavity through
medial wall of prootic. Infrastapedial process absent on quadrate. Sixteen teeth in
dentary.

Discussion. In his description of the type of C. liodontus, Merriam (1894, p. 35)
noted that, “Die maxillen sind vorn bus ungefahr zum 5. Zahn von der Praemaxillare
shrag abgeschnitten.” The species diagnosed above is the only Niobrara Clidastes to
which this statement could apply. Incomplete locality information suggests that C.
liodontus does not occur in the uppermost levels of the Niobrara chalk.

Clidastes propython Cope 1869
(Text-figs. 2B, 4B, 12-14, 22B, 39, 46B, 74, 75A, 76A)

Clidastes propython Cope, 1869b, p. 258.
?Clidastes cinertarum Cope, 1871a, p. 572.
Edestosaurus dispar Marsh, 1871a, p. 447.
Edestosaurus velox Marsh, 1871a, p. 450.
Clidastes wymant Marsh, 1871a, p. 451.
Clidastes pumilus Marsh, 1871a, p. 452.
Edestosaurus tortor Cope, 1872a, p. 297.
Clidastes vymanti, Cope, 1872b, p. 170.
Edestosaurus stenops Cope, 1872d, p. 268.
Edestosaurus rex Marsh, 1872b, p. 463.
Clidastes medius Merriam, 1894, p. 34.
Clidastes westi Williston and Case, 1892, p. 29.

Type. ANSP 10193, from “. . . The Rotten Limestone, near Uniontown in Alabama,”
(Cope, 1869b, p. 258), collected by Dr. E. R. Showalter. Type specimen consists of a
nearly complete skull and skeleton, except pelvic arch.

DistripuTION. Selma Formation, Alabama; Smoky Hill Member, Niobrara Formation,
Kansas and South Dakota; Lower Pierre Formation, Kansas and South Dakota.

ADDITIONAL REFERENCES. Cope, 1869f, p. 117; 1869-1870, p. 221, figs. 49(5), 50-51, pl.
12 figs. 1-21; 1871b, p. 583; 187lc, p. 132; 187le, pp. 217, 220, figs. 15, 19; 1871£, p.
413; 1872b, p. 170; 1872d, pp. 266, 269; 1872e, p. 141; 1872f, p. 330; 1874, pp. 33, 34;
1875, pp. 131, 133, 137, 265-266, pl. 14 figs. 1-2, pl. 16 figs. 1, 6, pl. 17 figs. 1, 7-8, pl. 18
figs. 1-5, pl. 19 figs. 1-6, pl. 21 figs. 214-17, pl. 36 figs. 3, 4, pl. 37 figs. 1-3, pl. 38 figs. 2-3;
1877, p. 583; 1878, p. 303; 1891, fig. 26; 1898, fig. 27. Marsh, 1871b, p. 104; 1872b, pp.
451, 464, pl. 11 figs. 3-4; 1872d, p. 497; 1897, p. 527. Leidy, 1873, pl. 35 fig. 14. Owen,

RUSSELL: AMERICAN MOSASAURS 129

Text-fig. 73. Lateral view of muzzle of Clidastes liodontus (reconstructed after YPM 1335, x 1).

130 PEABODY MUSEUM BULLETIN 23

1877, fig. 23. Hoffmann, 1890, p. 1322. Williston, 1891, p. 345; 1893b, pp. 110, 111;
1897c, p. 110; 1897e, pp. 245-246; 1898a, p. 29; 1898b, pp. 100-234, 197-205, pl. 23, pl. 24
fig. 7, pl. 27 fig. 1, pl. 28 figs. 1-4, pls. 35-38, pl. 39 upper fig., pl. 53, pl. 60 figs. 4-5,
pl. 61 fig. 4, pl. 64 fig. 1; 1902, pp. 248-253; 1914, fig. 71; 1925, figs. 58, 158b. Williston and
Case, 1892, pp. 17, 28, 30, 31, pls. 4-6. Merriam, 1894, pp. 34-36, pl. 1 fig. 4, pl. 3 figs. 6-7.
Takovlev, 1906, fig. 1. Holland, 1908, p. 162, fig. 5. Sternberg, 1909, p. 135, fig. 6; 1917,
p. 21. ?>Woodward, 1922, p. 6. Camp, 1923, fig. 22. Chaffee, 1939, fig. 1(2). Lane, 1947,
pp. 318, 319, fig. 7. Zangerl, 1948, p. 15. Gregory, 1951, fig. 4b; 1952, fig. 6. McDowell
and Bogert, 1954, fig. 32. Romer, 1956, figs. 108c, 145b. Sevon, 1958, p. 145, fig. Edmund,
1960, p. 89, fig. 25f.

REFERRED SPECIMENS. AMNH nos. 182, R173, 1507, **1513 (C. tortor), 1525, 1546,
1575 (C. stenops), 21579 (C. cineriarum), 1593, *ND2, *5812. YPM nos. *1100 (C. dispar),
1105 (C. velox), 1120 (C. wymani), 1123 (C. pumilus), 1308, 1310 (C. rex), 1314, 1315,
*1316, 1317, 1318, 1319, 1321, 1324, 1328, *1368, 1371 and thirteen unnumbered specimens.
USNM nos. 3765, 3778. MCZ no. 1625. KU nos. **1000, 1005, 1026 (C. westi), 1058, 1108,
1124, 1134. CVHM nos. UR901, **P12856. CM no. **1190.

Diacnosis. Premaxilla “V’-shaped in horizontal cross-section, small rostrum present
anterior to premaxillary teeth. Posteroventral portion of root of second premaxillary
tooth exposed on sutural surface with maxilla. Premaxillo-maxillary suture rises poste-
riorly in gentle curve to terminate at point above seventh maxillary tooth. Premaxillary
suture of maxilla smoothly keeled and parallels longitudinal axis of maxilla. Sixteen-18
teeth in maxilla. Median dorsal surface of parietal moderately broad. Parietal foramen
small, lies close to suture with frontal and opens ventrally into elliptical excavation in

Text-fig. 74. Lateral view of muzzle of Clidastes propython (YPM 1319, x 14).

parietal, length of which exceeds that of dorsal opening by about five times. Anterior
border of prootic forms shelf beneath prootic incisure, then descends abruptly to

Text-fig. 75. Dorsal view of parietal. A. Clidastes propython (AMNH ND2). B. Clidastes liodontus
(YPM 1335). C. Clidastes sternbergi (USNM 3777), all x 14.

RUSSELL: AMERICAN MOSASAURS 131

basisphenoid. Foramen for cranial nerve VII leaves brain cavity through medial wall of
prootic. Infrastapedial process present on quadrate. Seventeen to eighteen teeth in
dentary.

Discusston. The type specimens of each species considered above as a synonym of
C. propython include, with the exception of C. cimeriarum, good cranial material that
would not justify their specific separation from the excellent type of the former species.
C. cineriarum is provisionally placed here because of the large size of the single pterygoid
tooth, the only cranial material preserved with the type. The prefrontal figured with

A B

Text-fig. 76. Posterior view of right quadrate. A. Clidastes propython (YPM 1368, x 14. B.
Clidastes liodontus (YPM 3996, x 14).

the type of C. stenops (Cope, 1875, pl. 14 fig. 2a) has been laterally crushed, and the
unusual condition of this element in C. medius is probably best considered as due to
post-mortem distortion or damage (see also Williston, 1898b, p. 203) .

Genus MOSASAURUS Conybeare 1822
(Text-fig. 59)

Mosasaurus Conybeare, 1822, p. 198.
Batrachiosaurus Harlan, 1839a, p. 24.
Batrachotherium Harlan, 1839b, p. 89.
Macrosaurus Owen, 1849, p. 382.
Drepanodon Leidy, 1856, p. 255.
Lesticodus Leidy, 1861, p. 10.
Baseodon Leidy, 1865, p. 69.
Nectoportheus Cope, 1868a, p. 181.
Pterycollosaurus Dollo, 1882, p. 61.

Generic Type. Mosasaurus hoffmanni Mantell 1829.

ADDITIONAL REFERENCES (North American material). Morton, 1830a, p. 289; 1834, p.
27. Harlan, 1834b, p. 81; 1834e, p. 32; 1835a, p. 285. Bronn, 1838, p. 760. DeKay, 1842,
p. 28. Pictet, 1845, p. 621; 1853, p. 504. Holmes, 1849, p. 197. Gibbes, 1851, p. 3. Emmons,
1858, p. 213, fig. 342 Leidy, 1858b, p. 176; 1860, p. 91; 1865a, p. 30, pl. 7 figs. 1, 215-16, pl.
17 figs. 212-13; 1865b, p. 69; 1873, p. 279. Cope, 1869b, p. 260; 1869-1870, p. 186, ?fig.
49(2); 1871f, p. 401; 1875, p. 269; 1877, p. 567; 1879b, p. 36. Hayden, 1872, p. 91. Marsh,
1877, p. 346. Lydekker, 1888, p. 261. Merriam, 1894, p. 6. Williston, 1895, p. 166; 1897d,
pp- 177, 179, 181; 1898b, pp. 91, 195; 1902, p. 249; 1914, p. 165. Douglass, 1902, p. 212.
Pompeckj, 1910, pp. 130, 131. Gilmore, 1912b, p. 2; 1926, pl. 62 fig. 8; 1928, p. 87. Stern-
berg, 1915, p. 132. Camp, 1923, p. 322. Nopsca, 1928, p. 177. McDowell and Bogert, 1954,
pp- 132, 138. Romer, 1956, p. 561.

Dracnosis. Premaxilla with small rostrum anterior to premaxillary teeth. Thirteen
to fifteen teeth in maxilla. Prefrontal forms small portion of posterolateral border of
external nares, small to large triangular ala projects laterally from supraorbital wing.
Frontal not emarginate above orbits, median dorsal ridge present. Parietal foramen

132 PEABODY MUSEUM BULLETIN 23

moderately large, closely embraced on either side by tongues from frontal. Dorsal parietal
field long and narrow. Ventral process of postorbitofrontal to jugal confluent with well
exposed dorsal surface of postorbitofrontal. Ventroposterior process on jugal large.
Otosphenoidal crest on prootic may cover exit for cranial nerve VII laterally. Seven to
ten teeth in pterygoid. Suprastapedial process of quadrate relatively short, tympanic ala
moderately thick with groove around external margin. Stapedial pit elliptical in form.
Fourteen to seventeen teeth in dentary. Dentary terminates abruptly in front of first
dentary tooth. Median descending wing from coronoid incipient or well developed.
Dorsal edge of surangular rather thin lamina of bone, rising anteriorly to middle of
posterior surface of coronoid. Retroarticular process of articular rectangular in outline.
Mandibular teeth usually prismatic with relatively flat external and rounded internal
surfaces.

Postcranial skeleton much like Clidastes with following exceptions. Zygosphene-
zygantrum limited to anterior portion of vertebral column or absent. Transverse proc-
esses of pygal vertebrae relatively short. Scapula and coracoid subequal in size. Ischium
well expanded at symphysis. Internal trochanter located medially from head of femur.
Phalanges more numerous and of stouter proportions than in Clidastes.

Discussion. Batrachiosaurus, Batrachotherium and Pterycollosaurus were based solely
on the type skull and skeleton of Mosasaurus missouriensis, which is surely congeneric
with Mosasaurus hoffmanni, the type species of the genus. The names Macrosaurus,
Drepanodon, Lesticodus, Baseodon and Nectoportheus were all based on inadequate
material from the Cretaceous of the Atlantic seaboard and are discussed under the genus
Mosasaurus in the section on names of uncertain taxonomic standing.

Williston (1895, p. 168; 1897d, p. 179; 1898b, p. 195) remarked several times on the
close similarity between Mosasaurus and Clidastes. The only means by which he distin-
guished them (1902, p. 250) was in the presence of a zygosphene in the latter genus and
its absence in the former. M. conodon is exactly intermediate in this regard, retaining
the zygosphene in the anteriormost portion of the vertebral column only. The species
of Mosasaurus have almost certainly been derived from Clidastes.

M. ivoensis, M. missouriensis and M. maximus-hoffmanni all have powerfully con-
structed skulls with heavy prismatic teeth numbering 13-14 in the maxillae and 14-15
in the dentaries. M. missouriensis has 10 hypapophysis-bearing anterior vertebrae and
M. hoffmanni has 13. These species seem to form a phylogenetic series. M. conodon,
however, has a slender skull with compressed and, at most, weakly prismatic teeth num-
bering 15 in the maxilla and 17 in the dentary. There are only seven hypapophysis-
bearing anterior vertebrae.

It is quite possible that M. conodon descended from C. propython, and the M. ivoensis-
M. missouriensis-M. maximus phylum were derived at an earlier time from a more primitive
Clidastes stock. A complicating factor exists in the great similarity between the fore-
limbs of M. missouriensis and M. conodon. The pectoral crest is, however, medially
located on the proximoventral surface of the humerus in M. ivoensis just as Williston
(1898b, p. 147) states it is in M. missouriensis; in M. conodon it is nearly as anteriorly
placed as in Clidastes. The forelimb of M. missouriensis might well have been derived
from that of M. ivoensis and came to resemble in parallel fashion that of C. propython
and M. conodon. Whether this is true or not, the genus Mosasaurus may include at least
two closely related lines of large mosasaurines evolving in parallel ways, but derived
from different ancestral species within the Clidastes level of evolution.

Mosasaurus conodon (Cope 1881)
(Text-figs. 46A, 47B, 49A, 51, 56, 61, 77, 78)

Clidastes conodon Cope, 1881, p. 588.
?Mosasaurus lemonnieri Dollo, 1889b, p. 278.

Type. AMNH 1380, from “. . . near Freehold, Monmouth County, New Jersey .. .”
(Cope, 1881, p. 587), collected by Prof. Samuel Lockwood. Type specimen includes a

RUSSELL: AMERICAN MOSASAURS 133

dentary, splenial, angular, coronoid, articular, squamosal, twelve vertebrae, scapula,
coracoid, humerus and pectoral limb elements.

Tyre of Mosasaurus lemonnieri in the Musée Royal d'Histoire Naturelle de Belgique,
from “. . . Mesvin, localité située 4 4 kil. 5 de Mons,” Belgium (Dollo, 1889b, p. 273).
Type specimen includes a skull and anterior cervical vertebrae (ibid., p. 274).

Distripution. ?Taylor Marl, Texas; Marlbrook Marl, Arkansas; Verendrye Member,
Lower Pierre Formation, Virgin Creek Member, Upper Pierre Formation, South Dakota;
Navesink and younger Cretaceous, New Jersey; rCraie brune phosphatée de Ciply,
Belgium.

ADDITIONAL REFERENCES. Martin, 1953, pp. 1-62. Miller, 1955, p. 909.

REFERRED MATERIAL, AMNH nos. 1387, ?1395, 1397. YPM nos. ?379, ?1500, 1510, 1573.
ANSP nos. 8469, 8480, 8501, 8502, 28504, ?8509. USNM nos. 11396, 11904, 18255. SDSM
no. 452.

Diacnosis. Tuberosity present(?) below stapedial pit on lower medial body of
quadrate; suprastapedial and infrastapedial process small in lateral profile. Ventral wing
of coronoid well developed on medial surface of lower jaw. Dentary very slender, as in
Clidastes, dorsal and ventral margins converge gradually anteriorly, element comes to
rounded tip anterior to first dentary tooth. Fifteen teeth in maxilla, seventeen teeth in

Text-fig. 77. Type dentary of Mosasaurus conodon (AMNH 1380, x 1).

dentary and ten teeth in pterygoid. Marginal teeth slenderer than in M. missouriensts,
M. maximus and M. woensis, tips posteriorly recurved. External facets narrow and more
numerous than in M. maximus (Cranial diagnosis based on Dollo, 1889b, pls. 9-10; 1924,
p- 181).

Vertebral formula: 45 presacral vertebrae, 8 pygals, 21 caudals with chevrons and
transverse processes, about 54 terminal caudals.

Articulating surfaces of cervical and anterior dorsal centra wider than deep, smoothly
elliptical in outline; posterior dorsals become circular, then vertically oval. Zygosphene-
zygantrum present on second through thirteenth postcranial vertebrae. Anterior hypa-
pophyseal facets face ventroposteriorly, facing directly ventrally further posteriorly.
Seven hypapophysis-bearing cervicals. Transverse processes of pygal vertebrae relatively
short. Neural spines of postsacral vertebrae vertical on postsacrals 42-44, longest on post-
sacral 36.

Scapula and coracoid subequal in size. Glenoid articulating surfaces concave on
both scapula and coracoid, and smoothly continuous. Superior border of scapula gently
convex, posterior border slightly emarginated posteromedially. Coracoid expands
medially behind glenoid articulation. Distal and proximal ends of humerus widely ex-
panded, facets for articulation with other elements and sites of muscle attachment very
well differentiated. Proximal end of radius expanded, shaft bears greatly expanded ante-
rodistal flange. Distal end of radius thickened to form facet for articulation with radiale.
Radiale very large. Intermedium of average proportions. Ulnare does not enter postero-
ventral border of antebrachial foramen. Metacarpal one equal to metacarpal two in
length, is greatly expanded anteriorly at both ends. Proximal ends of other metacarpals
moderately expanded. Phalangeal formula of manus estimated at 9-10-10-10-4.

Acetabular surfaces of pelvic elements concave, form nearly solid, smoothly surfaced
bowl. Obturator foramen located near center of proximal end of pubis, dorsoanterior
process large and rectangular. Ischiadic tubercle small, separated from acetabulum by
very short neck. Shaft of ischium narrow, well expanded medially at symphysis. Distal
end of femur more expanded than proximal, internal trochanter moderately large and
located medially from head. Fibular facet on femur well marked, sharply set at acute

134 PEABODY MUSEUM BULLETIN 23

angle to tibial facet. Tibia with broad anterior flange, cartilage capped area of which
extends from base of tibia one-half way up shaft. Distal end of fibula more expanded
than proximal, fibular shaft narrow.

Astragulus approximately quadrilateral, sub-reniform; facet for fibula slightly raised
above that for tibia, dorsal notch between facets broad and shallow. Metatarsal one
strongly expanded at ends, especially proximally. Metatarsal five broad; concave medial
border gives it slight hook-shaped appearance. Approximate phalangeal formula of
pes 8-9-9-9-4 (postcranial diagnosis based on descriptions and figures of SDSM 452 given
by Martin, 1953).

Discuss1on. The excellent skeleton preserved in the Museum of the South Dakota
School of Mines and Technology is assigned to M. conodon. It has been described by
Harold Martin (1953, unpublished MA thesis) and placed in a new species close to
M. missouriensis. Unfortunately most of the cranium of this specimen was lost through
erosion so that a direct comparison with extant skulls of M. missouriensis is impossible.

A comparison of the forepaddle of SDSM 452 with that of M. missouriensis
(Williston, 1895, pl. 16 fig. 1) shows few significant differences. The intermedium and
radius narrowly exclude the radiale from bordering on the antebrachial foramen in
Martin’s specimen, while they widely exclude the radiale in M. missouriensis. The
ulnare and fourth carpal are large and subequal in size in the present specimen. Al-
though both elements are small, the fourth carpal is even smaller than the ulnare in
M. missouriensis. Goldfuss (1845, p. 191) states that there are ten hypapophysis-bearing
vertebrae in M. missouriensis, but there are only six in the present specimen. These
differences suggest that SDSM 452 belongs to a species distinct from M. missouriensis.

Postcranial material in the type of M. conodon from the Maestrichtian of New
Jersey is very similar to Martin’s specimen from the Virgin Creek Member of the Pierre
Shale. The articulating surfaces of the cervicals and anterior dorsals are similar, and
those of the terminal caudals are wider than deep in both specimens. The zygosphene-
zygantrum is present on the second through tenth postcranial vertebrae in the New
Jersey type; it is present on the second through thirteenth in SDSM 452. There are five
hypapophysis-bearing cervicals and an additional vertebra with a large peduncle in the
New Jersey type and six hypapophysis-bearing cervicals in SDSM 452. These two speci-
mens are therefore included in the same species. The type of M. conodon contains a
slender Clidastes-like dentary fragment with alveoli for 15 teeth. Probably at least two
more teeth were present in the missing anteriormost portion of the element. The atlas
synapophysis is compressed as in M. maximus, not tubular as in Clidastes. The ventral
border of the anteroventral extension of the synapophysis does not descend to a level
with the undersurface of the centrum in the type specimen.

USNM 18255 from the Verendrye Member of the Pierre Shale is referred to M.
conodon because of the slender Clidastes-like form of the dentary and the similarity of
the humerus to that of the type. USNM 11396 from the Marlbrook Marl of Arkansas
includes a nearly complete pelvis and 35 vertebrae from the more posterior regions of the
column. The pelvis is identical to that of SDSM 452 save that the ischiadic tubercle is
more prominent. There are nine pygal vertebrae as compared with eight in the South
Dakota specimen, and the articular facets of the terminal caudals are wider than deep.
One of the caudals with only a button remaining of the transverse process has an ante-
riorly sloping neural spine. In SDSM 452 the spine also slopes forward on the last
(21st) caudal with transverse processes. USNM 11904, consisting of forearm elements
and posterior dorsal vertebrae, was collected by L. W. Stephenson from Cretaceous strata
one mile west of Paris, Texas. On a National Museum catalogue card Stephenson in-
dicated that the specimen came from the “upper part of Taylor Marl?” According to
the USGS Geological Map of Texas (1937 edition) the Brownstown Marl, not the
Taylor Formation, outcrops west of Paris. This unit is of approximately Santonian age
(Stephenson et al., 1942). It is surprising that a humerus agreeing exactly in morphology
with those of USNM 11596 and 18255, both from the late Campanian, should occur in
such an ancient horizon, and it would seem more natural for Taylor strata to produce
such a specimen.

There are detailed resemblances between the vertebral column of M. conodon and

RUSSELL: AMERICAN MOSASAURS 135

those referred to M. lemonnieri from the Maestrichtian of Belgium (see Dollo, 1894).
SDSM 452 has a vertebral formula of 7 cervicals, 38 dorsals, 8 pygals and about 75
chevron-bearing caudals, as compared with 7 cervicals, 39 dorsals, 12 pygals and about
76 chevron-bearing caudals for M. lemonnieri. In the dorsal series there are 17 thoracics
and 21 lumbars as compared with 18 thoracics and 21 lumbars in M. lemonniert. The
zygopophyses become functionless on the first chevron-bearing caudal of SDSM 452
and on the fourth pygal of M. lemonnieri. In the caudal series the neural spine is
vertical on postsacrals 42-44 in SDSM 452, and in M. lemonnieri it is vertical on post-
sacrals 36-37, the longest neural spine occurring on postsacral 36 in both forms. ‘The
longest haemal spine occurs on postsacral 47 in SDSM 452 and on postsacral 46 in M.
lemonnieri, and the transverse processes become rudimentary in the vicinity of the
eighteenth caudal in both forms. It should be noted that marginal teeth in SDSM 452

Text-fig. 78. Tooth preserved in type dentary of Mosasaurus conodon (AMNH 1380, x 3%).

and in the type of M. conodon are smoothly surfaced, but the teeth of YPM 1573 and
the single tooth preserved in the dentary of USNM 18255 are narrowly faceted, as in
M. lemonnieri.

However Dollo (1894, p. 230) did not observe zygosphene-zygantrum articulations
in the vertebral column of M. lemonnieri, which are functional in the anteriormost
presacral series of M. conodon. Dollo lists 12 pygal vertebrae in M. lemonnieri, in
SDSM 452 there are but eight. The posterior caudal centra of M. lemonnieri are verti-
cally oval but in M. conodon they are horizontally oval.

The zygosphene-zygantrum articulation may be variably developed within a species
of Platecarpus and it is possible that they were either small and overlooked by Dollo in
M. lemonnieri or were indeed absent. They are only present in a small portion of the
column in M. conodon. It is quite possible that Dollo included anterior caudals with
rudimentary haemal arches in the pygal region of M. lemonnieri, thus accounting for
the differing pygal count between the two forms. Except for the more heavily constructed
proximal end of the femur, the appendicular girdles and limbs of the American form are
identical to those figured by Dollo (1894, pls. 3-4) of M. lemonnieri. In summary, there
seems to be little justification for separating M. lemonnieri from M. conodon and for
this reason the cranial portion of the diagnosis is based upon the more complete cranial
material from Belgium.

Mosasaurus ivoensis Persson 1963
(Text-fig. 79)

Clidastes stenops Williston, 1902, pp. 248-250, pl. 12 lower fig.
Mosasaurus hoffmanni ivoensis Persson, 1963, p. 5.

Tyrer. No. R1292 in Swedish Museum of Natural History, Stockholm. Type specimen
is the crown of a tooth (Persson, 1963, p. 5).

136 PEABODY MUSEUM BULLETIN 23

REFERRED SPECIMEN. KU no. *1024, from the Niobrara Chalk, collected by Charles
H. Sternberg.

Diacnosis. Narial emargination begins dorsal to fourth maxillary tooth. Fourteen
teeth in maxilla. Splenial has weak median dorsal keel on articulating surface. Fourteen
teeth in dentary. Dentary relatively heavy with parallel dorsal and ventral margins up
to point beneath third and fourth dentary tooth, margins then converge rapidly ante-
riorly to a rounded tip. Marginal teeth short relative to other species of Mosasaurus and
triangular in lateral outline, prisms numerous (6-7) on external face, as many as eleven
distinct prisms on internal face.

Discussion. The referred specimen includes the right side of a muzzle, with the pre-
maxilla, dentary, splenial, jugal and possibly also a pterygoid and ectopterygoid, al-
though the latter two elements may belong to a Tylosaurus. The atlas and axis vertebrae,

Text-fig. 79. Muzzle referred to Mosasaurus ivoensis (KU 1024, x 14).

and a large part of the left forepaddle are also present. The maxillary teeth are indistin-
guishable from those described by Persson (1959, p. 462) from the lower Campanian
of southern Sweden, and later named M. hoffmanni ivoensis by the same author (Pers-
son, 1963, p. 5). The relationship of the above specimen with Mosasaurus is shown by
the abbreviated rostrum of the premaxilla and the number (14 each in the dentary and
maxilla) and form of the teeth. The articular surface of the splenial is more nearly
circular when viewed from the posterior than that of any other mosasaur known to the
author. The jugal resembles that of Mosasaurus except that the ventroposterior process
is very weakly developed.

A zygantrum is apparently absent on the axis vertebra (Williston, 1902, p. 249).

The foreflipper of this specimen was figured by Williston (1902, pl. 12 lower figure).
The humerus bears evidence of its Mosasaurus affinities by its general proportions,
pectoral crest located away from the anterior border of the bone, and the large ulnar
tuberosity. The post-glenoid process is, however, very small and the radial tuberosity
is apparently absent. The radius resembles that of Clidastes more closely than that of
Mosasaurus, but the shaft is broader and the distal extremity less dilated than in the
former genus. The bones of the carpus are evidently in a state of partial disarray. The
element in the position of a radiale is very small, which together with the shallow rec-
tangular outline of the intermedium give the carpus a strikingly Ectenosaurus-like ap-
pearance quite unlike that typical of Clidastes and Mosasaurus. The first metacarpal is
not nearly so short and broad as in M. hoffmanni, M. maximus, M. missouriensis and
M. lemonnieri and has an even less rectangular outline than in Clidastes. The element
figured by Williston in the position of the ulna is probably a metacarpal.

Mosasaurus missouriensis (Harlan 1834)

Ichthyosaurus missouriensis Harlan, 1834a, p. 440.

RUSSELL: AMERICAN MOSASAURS 137

Ictiosaurus missuriensis, Harlan, 1834d, p. 124.
Batrachiosaurus missouriensis, Harlan, 1839a, p. 24.
Batrachotherium missouriensis, Harlan, 1839b, p. 89.
Mosasaurus maximiliani Goldfuss, 1845, p. 179, pls. 6-8
Mosasaurus neovidii Meyer, 1845, p. 312.

Mosasaurus missouriensis, Leidy, 1858a, p. 90.
Pterycollosaurus maximiliani, Dollo, 1882, p. 61.
Mosasaurus horridus Williston, 1895, p. 166, pls. 14-16.

Type. Geol.-Paliiont. Institut der Rhein. Friedrich-Wilhelm-Universitit cat. no.
GOLDFUSS 1327 (personal communication, H. K. Erben 1964) from “. . . der Gegend
des Big-Bend, einer grossen Krummung des Missouri, zwischen Fort Lookout und Fort
Pierre . . .” (Goldfuss, 1845, p. 175), collected by Maj. Benjamin O'Fallon. The anterior
portion of the rostrum described by Harlan has not been located. Type specimen is a
nearly complete skull, most of the vertebral column and fragments of the pectoral girdle
(Goldfuss 1845).

DistRipuTION. Upper Pierre Formation, South Dakota; Bearpaw Formation, Montana.

ADDITIONAL REFERENCES. Harlan, 1834b, p. 80; 1834c, p. 408, pl. 20 figs. 1-8; 1834e,
p- 31: 1835a, p. 284; 1835b, p. 348, pl. figs. 1-6; 1839c, p. 302; 1842, p. 142. Goldfuss,
1847, p. 123. Owen, 1849, p. 382; 1851, p. 32; 1877, pp. 683, 696, 701, figs. 5, 6, 16, 18.
Gibbes, 1850, p. 77; 1851, p. 8. Pictet, 1853, p. 505. Hayden, 1857, p. 113; 1872, p. 87.
Meek and Hayden, 1857, pp. 117, 119. Emmons, 1858, p. 218. Leidy, 1858c, p. 11; 1860,
p- 92; 1865a, pp. 30, 33, 117; 1865b, p. 69. Cope, 1869b, p. 263; 1869c, p. 86; 1869-1870,
pp- 189, 195; 1871a, p. 571; 1871f, pp. 386, 401; 1875, p. 269; 1877, p. 567; 1879b, p. 36.
Hoffman, 1890, p. 1320. Baur, 1892, p. 9. Merriam, 1894, p. 5. Williston, 1895, p. 165, pl.
16 fig. 1; 1897a, p. 95; 1898b, pp. 103-154, pls. 19-21, 32; 1914, p. 151; 1925, p. 273. Ballou,
1898, fig. 1. Osburn, 1906, pl. 8 fig. 27. Camp, 1942, p. 45. Simpson, 1942, pp. 162, 172.
Kauffman and Kesling, 1960, p. 251, figs. 1, 6a, tables 4-6.

REFERRED SPECIMENS. USNM nos. 4910, *8086. KU no. **1034 (M. horridus)

Dracnosis. No median dorsal crest on premaxilla. Fourteen teeth in maxilla. Small
triangular ala projects laterally from supraorbital wing of prefrontal. Narial emargina-
tion begins dorsal to point between fourth and fifth maxillary tooth. Parietal foramen
large, bounded by two short tongues from frontal. Eight to ten teeth in pterygoid.
Tuberosity below stapedial pit absent on lower medial body of quadrate, suprastapedial
process small in lateral profile, infrastapedial process large. Ventral wings of coronoid
poorly developed on medial and lateral surfaces of lower jaw. Splenial has weak median
dorsal keel on articulating surface. Fifteen teeth in dentary. Dentary slender, dorsal
and ventral margins converge gradually anteriorly, element comes to rounded tip anterior
to first dentary tooth. Marginal teeth longer than in M. iwoensis with posteriorly re-
curved tips, prisms fewer in number (4-6) on external face, as many as eight poorly de-
veloped prisms on internal face.

Discussion. The type of M. missouriensis was one of the first fossil vertebrates to be
described from the western part of this continent. It must have come from the Pierre
Shale and, according to Goldfuss’ (1845, p. 175) description of the matrix, probably from
the concretionary zone in the upper part of the Virgin Creek Member (see Crandell,
1950, p. 2338) exposed along the Big Bend of the Missouri River southeast of Pierre,
South Dakota. A fragment of the anterior end of the snout was named Jchthyosaurus
missouriensis by Harlan in 1834, and the remainder of the skull and skeleton was de-
scribed by Goldfuss in 1845 as M. maximiliani. Both Meyer (1845, p. 313) and Camp
(1942, p. 45) have pointed out that the rostral fragment figured by Harlan would fit
perfectly on the broken anterior end of Goldfuss’ skull, thereby making maximiliani a
junior synonym of missouriensis.

In 1895 (pp. 166-167) Williston described a new species of Mosasaurus, M. horridus,
also from the Pierre Shale of South Dakota. It was distinguished on the basis of the lesser
number of pterygoid teeth (8 versus 10 in M. missouriensis), the greater depth of the
posterior portion of the mandible and the greater size of the skull. It should be noted

138 PEABODY MUSEUM BULLETIN 23

that the number of pterygoid teeth is variable within the Niobrara species of Clidastes,
Platecarpus and Tylosaurus, and may have been also in Mosasaurus. Williston took
his measurements of the posterior jaw of M. missouriensis directly from Goldfuss’ plates,
in which the lower jaws may have been figured in a rotated position, not showing their
full depth. The type of M. horridus is not enough larger (only by about 14) than that of
M. missouriensis to be of taxonomic significance. Therefore the characters given by
Williston are not here regarded as sufficient to separate M. horridus from M. missourien-
sis, and a comparison of Williston’s and Goldfuss’ plates will show the striking similarity
of the two type skulls. M. horridus is here, then, regarded as another junior synonym
of M. missouriensis.

There are rudimentary zygosphene-zygantra in the anterior portion of the vertebral
column (Williston, 1895, p. 167).

Mosasaurus dekayi Bronn 1838.

“saurian .. . resembling famous reptile of Maestricht,” Mitchill, 1818, pp. 384, 431, pl.
3 fig. 4.

Ichthyosaurus, de Blainville, 1827, p. 48.

Mosasaurus, DeKay, 1830, p. 135, pl. 3 figs. 1-2.

Mosasaurus dekayi Bronn, 1838, p. 760.

Mosasaurus major DeKay, 1842, p. 28, pl. 22 figs. 57-58.

Mosasaurus maximiliant, Pictet, 1853, p. 505.

Mosasaurus meirsti Marsh, 1869, p. 395.

Type. Specimen not located, originally from “. . . foot of Neversink Hills, Sandy
Hook, Monmouth County (New Jersey), (DeKay, 1830, p. 135). Type specimen is the
crown of a tooth.

Distripution. Navesink Formation and younger Cretaceous, New Jersey.

ADDITIONAL REFERENCES. Morton, 1830a, p. 289; 1830b, p. 246. Gibbes, 1850, p. 77;
1851, pp. 1, 8, pl. 1 fig. 1. Emmons, 1858, p. 217, fig. 37; 1860, p. 208, fig. 180(1), (5).
Leidy, 1860, p. 92; 1865a, p. 32, figs. 7, 13, 14, 26-28, pl. 10 figs. 7, 12-13. Cope, 1869b,
p- 263; 1869-1870, pp. 185, 188, 193, 199; 1875, pp. 269, 270. Miller, 1955, pp. 905, 909.

REFERRED SPECIMENS. AMNH no. 1402. YPM nos. 443 (M. meirsi), 410, 930, 1582.
ANSP nos. 8468, 8471, 8495, 8511.

Discussion. There is a kind of tooth from the latest Cretaceous of New Jersey that
is more or less symmetrically bicarinate, has its external and internal faces strongly di-
vided into prisms, and is bilaterally compressed relative to most of the teeth of M.
maximus. These teeth apparently correspond to the posterior mandibular dentition of
a different species of Mosasaurus. The type tooth of M. dekayi and M. major was of
this kind, as is the splinter of a similarly strongly prismate tooth named M. meirsi by
Marsh.

A short series of anterior caudal vertebrae have been given the same number (YPM
930) as a tooth of M. dekayt. ‘They are very similar to corresponding vertebrae of M.
maximus, although the haemal arches arise at the base of the transverse processes instead
of further medially, as is the case in the latter species.

Mosasaurus maximus Cope 1869
(Text-figs. 3, 8, 24A, 30, 80)

Mosasaurus in part, Morton, 1834a, p. 27.
Mosasaurus mitchilli in part, Leidy, 1865a (see below)
Mosasaurus maximus Cope, 1869b, p. 262.
Mosasaurus princeps Marsh, 1869, p. 392.
Mosasaurus fulciatus Cope, 1869-1870, pp. 189, 194.
Mosasaurus oarthrus Cope, 1869-1870, pp. 189-196.

RUSSELL: AMERICAN MOSASAURS 139

Tyre. AMNH 1389, from “... Monmouth Co., N. J.” (Cope, 1869-1870, p. 189). Type
specimen includes a quadrate, jaw fragments, and seven vertebrae.

DistripuTIoN. Navesink Formation and younger Cretaceous, New Jersey; Ripley
Formation, Tennessee; Navarro Formation, Texas.

ADDITIONAL REFERENCES. Leidy, 1865a, figs. 1-3, 5-6, 8-9, 12, 15-19, 23, 29-31, 33, pl.
8 fig. 11, pl. 9 figs. 1-3, 5-7, 11, pl. 10 figs. 1-5, 8-10, 16, pl. 11 figs. 6-7, 11-13, pl. 19 fig. 6,
pl. 20 figs. 3-6. Cope, 1869c, p. 86; 1869-1870, pp. 188, 189, 192, figs. 48(1), 49(1), pl. 11
figs. 7-8; 187la, p. 571; 1875, p. 269, pl. 37 figs. 13, 15-17. Lydekker, 1888, p. 264. Baur,
1892, p. 4. Williston, 1897c, p. 110. Whitfield, 1900, p. 25, pl. 4 figs. 1-2, pl. 5 figs. 1-2.
Abel, 1922, fig. 269. Gilmore, 1926, p. 191, pl. 71 figs. 6, 8. Chaffee, 1939, fig. 1(5). Miller,
1955, pp. 905, 909. Edmund, 1960. fig. 25d.

REFERRED SPECIMENS. AMNH nos. 71391, 1392 (M. oarthrus), 21393, 71397, 1398
(M. fulciatus), 1404, 1406, YPM nos. 305, 306, 307, 311, 2414, *430 (M. princeps), 470,
508, 509, 510 690, ?773, 1504, 1527, 1529, 1530, 1574, 1575. ANSP nos. 8462, 8472, 28473,
?8474, 8476, 8478, 8479, 8481, 8482, 8503, 8505, 8507, 8508, 8513, 8514, 8515, 8516,
8517, 8519, 8520, 8524, 8531, 8540, 8546, 8547, 29641. USNM no. 10540. CNHM no.
PR491. NJSM nos. *11052, **11053. TMM **specimen.

Diacnosis. Slight medial dorsal crest on premaxilla. Small triangular ala projects
laterally from supraorbital wing of prefrontal. Narial emargination begins dorsal to
point between fifth and sixth maxillary tooth. Parietal foramen large, bounded by two
long tongues from frontal. Large keel-shaped tuberosity below stapedial pit on lower

Text-fig. 80. Lateral view of left quadrate of Mosasaurus maximus (YPM 430, x 36).

medial body of quadrate, suprastapedial and infrastapedial process very small in lateral
profile. Ventral wings of coronoid well developed on medial and lateral surface of lower
jaw. Splenial has strong median dorsal keel on articulating surface. Fourteen teeth in
dentary. Dentary deep posteriorly, rapidly narrows coming to nearly pointed tip ante-
riorly. Marginal teeth long with posteriorly recurved tips, prisms few in number (2-3)
on external face, absent or nearly absent on internal face.

140 PEABODY MUSEUM BULLETIN 23

Articulating surfaces of cervical and anterior dorsal vertebrae circular. Synapophysis
located high in anterior two-thirds of lateral surface of cervical and anterior dorsa
centra. Ventral border of anteroventral extension of synapophysis weakly developed on
cervicals, extending directly anteriorly, absent on anterior thoracics. Anterior zygopophy-
sis of cervicals and thoracics connected by rounded, posterior-descending crest to
synapophysis. Zygosphene-zygantrum, if present, not large. Anterior base of atlas neural
spine arises behind condylar facet, synapophysis moderately large and compressed. Hypa-
pophyseal peduncle located posteriorly on ventral surface of cervical centra, articulation
for hypapophysis flat and triangular, slightly inclined posteriorly. Transverse processes
of pygal vertebrae relatively short (postcranial diagnosis based on YPM 430, 690 and
1504).

a Mosasaurus maximus is the earliest name that can be applied with cer-
tainty to the commonest species of mosasaur in the New Jersey Cretaceous, M. princeps,
M. oarthrus and M. fulciatus being recognized here as junior synonyms. M. maximus
must be very nearly related to M. hoffmanni from the Maestrichtian of western Europe.
The quadrates of the two species are almost identical (see Dollo, 1889b, pl. 10 figs. 12-13;
1909, pl. 9; Cope, 1875, pl. 37 figs. 14-17), and casts of the famous type Maestricht head
in the collections of the Academy of Natural Sciences of Philadelphia and the Peabody
Museum at Yale show the coronoids, the posterior moiety of the lower jaw and mandibu-
lar teeth to be likewise similar. The dentaries of M. hoffmanni are slenderer than those of
M. maximus (see Cuvier, 1834-1836 Atlas vol. 2, pl. 246 fig. 1; Dollo, 1889b, pl. 9 fig. 1)
and the premaxilla of M. hoffmanni is more acutely pointed (see Dollo, 1882, pl. 4 figs.
1-2).

eee the two species may prove to be identical it may be well to point out certain
skeletal features of M. hoffmanni which are presently unknown in M. maximus. There
are 13 maxillary teeth, 14 dentary teeth and eight pterygoid teeth in the European form
(Dollo, 1924, p. 181). Cuvier (1834-1836, p. 165) states there are twelve hypapophysis-
bearing cervicals in M. hoffmanni. Swinton (1930, p. 47) notes that a specimen of this
species in the British Museum definitely has zygosphenes, but an anterior dorsal of the
same species in the Yale collections (YPM 24933) definitely lacks them. The tail ends in
an expanded fin (Dollo, 1917, p. 17).

As Marsh (1872b, p. 455) has pointed out, the bone figured by Cuvier (1834-1836
Atlas vol. 2, pl. 247 fig. 10) is an ischium. The ischiadic tubercle is very widely sepa-
rated from the acetabulum and points distinctly posteromedially instead of posteriorly,
as in other mosasaurs. The shaft of the ischium is broad and evidently well expanded
medially at the symphysis. A cast of the femur is present with the above-mentioned
plastotypes of M. hoffmanni. The shaft of the femur is short, and the distal end is more
expanded that the proximal end. The internal trochanter is very large and located
medially from the head so that the long axis of the proximal end lies at right angles
to the long axis of the distal end of the femur. The fibular facet is well developed and
sharply set at an acute angle to the tibial facet.

A very fragmentary skull in the Yale museum (YPM 773) differs in some respects
from other large New Jersey mosasaurs. The lateral exit for nerve VII is protected by
a short heavy otosphenoidal crest, which is lacking in M. maximus. When viewed from
the posterior the articulating surface of the splenial in this specimen and in ANSP
9641 is elliptical in general outline, not rectangular as in M. maximus or narrowly
compressed as in Prognathodon. The teeth, coronoid and quadrate are identical to
those of M. maximus.

In two fragmentary parietals (YPM 1504 and 1582) the margins of the dorsal parietal
surface converge slightly at their posterior termination, instead of diverging as in ma-
terial surely referable to M. maximus. Associated elements (quadrates, coronoid, splenial,
angular and teeth) with YPM 1504 are like those of M. maximus, but although a prootic
of YPM 1582 resembles that of M. maximus, the associated teeth are those of the M.
dekayi kind. It is conceivable that M. maximus and M. dekayi are synonyms and that
the teeth of a large species of mosasaur have been unnaturally separated here into two
form species. Additional evidence would be very helpful in understanding tooth varia-

RUSSELL: AMERICAN MOSASAURS 141

tion and the meaning of the converging margins of the posterior surfaces of the above-
mentioned parietals.

The infrastapedial process and a vertical keel beneath the stapedial pit on the medial
shaft of the quadrate are independently variable in development in different specimens
of New Jersey Mosasaurus. These processes are always present, however, and so far as
can be told there is no taxonomic significance to their variation.

A coronoid (USNM 10540) from the Coon Creek Formation of McNairy County,
Tennessee, described by Gilmore in 1926, is referred to M. maximus.

Genus AMPHEKEPUBIS Mehl 1930
Amphekepubis Mehl, 1930, p. 383.

Generic Type. Amphekepubis johnsoni Mehl 1930.

Diacnosis. Transverse processes of pygal vertebrae relatively long.

Acetabular surfaces of pelvic elements concave, form nearly solid, smoothly surfaced
bowl. Obturator foramen located near center of proximal end of pubis; dorsoanterior
process exceedingly long and slender, distally pointed. Ischiadic tubercle small, distinctly
separated from acetabulum by moderately short neck. Shaft of ischium broad, slightly

Text-fig. 81. Lateral view of type pelvic girdle of Amphekepubis johnsoni (after Mehl, 1930, pl.
65, X %).

142 PEABODY MUSEUM BULLETIN 23

dilated distally. Shaft of femur slender, distal end more expanded than proximal; in-
ternal trochanter moderately large and located medially from head so that long axis of
proximal end lies at right angles to long axis of distal end of femur. Fibular facet on
femur well developed, small tuberosity present on lateral surface dorsal to tibial facet.
Tibia narrow, cartilage-capped area of anterior flange extends from base of tibia nearly
one-half way up shaft.

Astragulus approximately quadrilateral, sub-reniform; facet for fibula raised slightly
above that for tibia, dorsal notch between facets broad and shallow. Metatarsal one
expanded proximally, somewhat medially recurved.

Discussion. The known parts of the hind limb are very similar to those of Clidastes.
The genus is surely separate from Clidastes and apparently more primitive in its more
slender femur and in the circular instead of trapezoidal outline of the central articula-
tions in the pygal vertebrae. Camp (1942, p. 25) suggested Amphekepubis might be
synonymous with Mosasaurus and figured (ibid., pl. 5) the type pelvic arch in his recon-
struction of Plotosaurus. In both of these genera however the central articulations of
the pygal vertebrae are triangular in outline. The transverse processes of the pygal
vertebrae of Amphekepubis are much longer than in Mosasaurus and arise out of the
center of the anterior one-half of the centrum, thus resembling those of Clidastes more
closely. The very large dorsoanterior process of the pubis at once distinguishes this
element from those of Clidastes, Platecarpus and Tylosaurus (see Mehl, 1930, p. 393).

Amphekepubis johnsoni Mehl 1930
(Text-fig. 81)

Amphekepubis johnsoni Mehl, 1930, p. 383, figs. 2-3, pls. 64-67.

Tyre. Univ. of Missouri no. 509VP, from the ?San Felipe Formation 40 miles east
and somewhat north of Monterey, Nuevo Leon, Mexico (Mehl, 1930, p. 383). Type
specimen includes pelvic arch and portion of hind limb, and nine postsacral vertebrae.

ADDITIONAL REFERENCES. Camp, 1942, p. 25, pl. 5. McDowell and Bogert, 1954, p. 132.
Romer, 1956, p. 561.

Genus LIODON Owen 1841
Liodon Owen, 1840-1845, p. 261, pl. 72 figs. 1-2.

Generic Type. Liodon anceps Owen.

Dracnosis. Premaxilla with small rostrum present anterior to premaxillary teeth.
Thirteen to fourteen teeth in maxilla. Fourteen to sixteen teeth in dentary. Small pro-
jection of dentary anterior to first dentary tooth. Mandibular teeth become highly com-
pressed and bicarinate posteriorly, enamel surfaces smooth (Gaudry, 1892, pp. 5-9, pls.
1-2).

Discussion. The proper application of the name Liodon is still uncertain. The type
species of the genus, L. anceps Owen, was based on a jaw fragment containing smoothly
surfaced, symmetrically bicarinate teeth. Unfortunately no good cranial material has yet
been described that can be assigned to this species, but, on the basis of the dentition,
Gaudry (1892, p. 4), Depéret and Russo (1925, p. 340) and Persson (1959, p. 465) have
suggested that Hainosaurus Dollo, a tylosaurine, may prove to be congeneric with L.
anceps. Hainosaurus is surely generically distinct from L. compressidens and L. mosa-
sauroides, both from the French Cretaceous, upon which the foregoing diagnosis is based.
If new material demonstrates that L. anceps cannot be included in the same genus as
the latter two species, then these two species together with L. sectorius Cope will have
to be placed in a new genus, within the Mosasaurinae. As it now stands Liodon is
placed in the Mosasaurini because of the great resemblance of the muzzle of the French

RUSSELL: AMERICAN MOSASAURS 143

species to Mosasaurus. No postcranial material of any species of Liodon has ever been
described.

Liodon sectorius Cope 1871
(Text-fig. 82)
Liodon sectorius Cope, 1871d, p. 41.
Tylosaurus sectorius, Merriam, 1894, p. 25.

Tyre. AMNH 1401, “. .. from the marl pits of the Pemberton Marl Co., at Birming-
ham, N.J... .” (Cope, 1871d, p. 43), collected by J. C. Gaskill. Type specimen consists
of fragments of cranial bones including those of a maxilla, dentary, coronoid and suran-
gular, and one fragmentary vertebra.

DistripuTIon. Navesink Formation or younger Cretaceous, New Jersey.

ADDITIONAL REFERENCES. Cope, 1875, p. 271. Miller, 1955, pp. 905, 909.

REFERRED SPECIMENS. ANSP no. 9669, 9670.

Discussion. The type specimen of this species has been badly damaged and some of
its parts have been lost; the original description of its dentition is given here:

“The anterior teeth are less compressed, and have but one, an anterior, cutting edge,
the posterior face being regularly convex. The inner face is much more convex than the
outer, and the flatness of the latter is marked at the apex of the tooth by a short ridge
which bounds it posteriorly. This is a trace of the bounding angle which extends to the

A

Text-fig. 82. A. Anterior marginal tooth. B. Posterior marginal tooth of type of Liodon sectorius
(AMNH 1401, x 34).

basis of the crown in Mosasaurus. The anterior cutting edge is in profile convex; the
posterior outline concave to near the tip. The cutting edge is acute, and beautifully
ribbed on each side, but not properly denticulate. The surface of the tooth is not
facetted, but the outer face exhibits the peculiarity of a longitudinal concavity, or shallow
groove extending from the base to the middle of the crown. . . This description is taken
from the second or third (tooth) from the anterior end of the maxillary bone. The
third from the distal end of the dentary is very similar.

“The crowns become rapidly more compressed as we pass backwards. From a broad
oval section of two crown bases, we reach a flattened oval crown, with the cutting edge
sharp behind as well as before, and minutely ribbed. The crown is not facetted, and
is more convex interiorly than exteriorly. The exterior convexity is chiefly anterior; the
posterior face is slightly concave from the open groove already described as present in
the anterior teeth. In two posterior crowns, one still more elongate in section, the ex-
ternal concavity becomes flatter and includes a great part of the outer face. A tooth still
more posterior presents the peculiarity of the species in the strongest light. The crown

144 PEABODY MUSEUM BULLETIN 23

is still more compressed, directed backwards, and only .25 higher than wide antero-
posteriorly at the base. . .’’ (Cope, 1871d, pp. 41-42)

The teeth of L. sectorius are approximately intermediate in form between those of
L. compressidens and L. mosasauroides, excellently figured by Gaudry (1892, pls. 1-2),
and are more strongly compressed than those of L. anceps (see Owen, 1840-1845, pl. 72 figs.
1-2). It will be very interesting to see additional characters of L. sectorius brought to
light as new material is discovered from the Cretaceous of New Jersey.

Tre GLOBIDENSINI (Dollo 1924) new

Dracnosis. The osteology of Globidens remains very incompletely known. It appears
to have descended from a Clidastes-like ancestor, but because of the highly peculiar
nature of its spherical teeth and massive jaws (see below) it is separated from the Mosa-
saurini into a distinct tribe.

Genus GLOBIDENS Gilmore 1912
Globidens Gilmore, 1912a, p. 479.

Generic Type. Globidens alabamaensis Gilmore.

Diacnosis. Median dorsal crest on massive frontal present. Frontal not emarginate
above orbits. Ventral process of postorbitofrontal to jugal confluent with broadly ex-
posed dorsal surface of postorbitofrontal. Mandibular teeth spherical in form.

Articulating surface of anterior thoracic vertebra wider than deep, subcircular in
outline. Synapophysis located in center of lateral surface of anterior thoracic vertebra.
Ventral border of anteroventral extension of synapophysis well developed on anterior
thoracic, reaches level of undersurface of centrum. Zygosphene-zygantrum present but
not large.

Globidens alabamaensis Gilmore 1912
Globidens alabamaensis Gilmore, 1912a, p. 479, figs. 1-3, pls. 39-40.

Type. USNM 6527, from the Selma Chalk of “. . . Bogue Chitto Prairies west of
Hamberg . . . Perry and Dallas Co; Ala. . .” (Gilmore, 1912a, p. 479), collected by L. C.
Johnson. Type specimen includes a maxilla, frontal, postorbitofrontal, basisphenoid,
splenial and one vertebra.

DistripuTion. Selma Formation, Alabama and Mississippi; Lower Taylor Marl, Taylor
Formation, Texas; Lower Pierre Formation, South Dakota.

ADDITIONAL REFERENCES. Williston, 1914, p. 167, fig. 80. Gilmore, 1921, p. 273 (fig.),
p- 280; 1927, p. 452, Abel, 1922, fig. 266; 1924, fig. 37. Camp, 1923, p. 323. Dollo, 1924,
p- 170. Nopcsa, 1928, p. 177. Zangerl, 1948, p. 15. McDowell and Bogert, 1954, pp. 106,
107, 129, 132. Romer, 1956, p. 562. Edmund, 1960, p. 90, fig. 25c. Kauffman and Kesling,
1960, p. 224, table 3.

REFERRED SPECIMENS. PUSNM no. 4993. SDSM no. 4612.

Discussion. Gilmore’s fragmentary type is the most complete Globidens skull yet
known. The premaxillary suture of the maxilla is smoothly keeled and parallels the
longitudinal axis of the maxilla as in Mosasaurus and in some species of Clidastes, the
emargination for the external nares beginning dorsal to a point between the fifth and
sixth maxillary tooth. The prefrontal and postorbitofrontal seem to have been narrowly
separated by the ventral surface of the frontal above the orbits. The postorbitofrontal
has a widely exposed triangular dorsal surface and a jugular suture similar to those in
Clidastes. There is no canal through the basisphenoid for the basilar artery. Gilmore
(1912a, p. 484) referred G. alabamaensis to the Plioplatecarpinae, and Dollo (1924,

RUSSELL: AMERICAN MOSASAURS 145

p- 188) erected a new family for its inclusion. The characters outlined above indicate
that Globidens is a derivative of, and closely related to Clidastes, although very different
in the massive structure of its cranial elements and the spherical form of its teeth.

A very heavy coronoid and articular (USNM 4993) from the ?Marlbrook Marl, Red
River Valley of Hempstead County, Arkansas, probably belong to Globidens. The
coronoid is depressed and extends a long wing posteriorly above the dorsal edge of the
surangular. The retroarticular process of the articular is very strongly constructed,
nearly perfectly circular in lateral outline, and has a large tuberosity on its ventral
margin beneath the posterior end of the quadrato-mandibular joint.

A fragment of the posterior end of a dentary (SDSM 4612) with a single tooth and
an alveolus for another, from the Lower Pierre Shale of Fall River County, South
Dakota, is probably conspecific with G. alabamaensis. There is a deep notch extending
anteroventrally from the dorsoposterior edge of the dentary to a point beneath the
penultimate dentary tooth, which has not been described in any other species of
mosasaur.

Globidens teeth have been discovered in the Selma Chalk in the vicinity of Saltillo,
Lee County, Mississippi (Gilmore, 1927, p. 452) and in the Lower Taylor Marl Member,
Taylor Formation, near Austin, Texas (Hotton, personal communication 1963). It is
very odd that Globidens has not been reported from the Niobrara Chalk. G. alabamaensis
also occurs in the Craie d’Oburg, near Mons, Belgium (Dollo, 1924, p. 170).

TrinE PLOTOSAURINI new

Diacnosts. More than twelve pygal vertebrae present. Radius and ulna nearly
contact above carpalia on distal border of antebrachial foramen. The structure of the
manus in these forms from the Pacific basin suggests they should be placed in a tribe
distinct from other members of the Mosasaurinae.

Genus PLOTOSAURUS Camp 1951
(Plate II, fig. 2)

Kolposaurus Camp, 1942, p. 2, pls. 5-6 (preoccupied, Skuphos 1893).
Garzasaurus Anderson, 1943, p. 186 nomen nudum.
Plotosaurus Camp, 1951, p. 822.

Generic Type. Plotosaurus bennisoni (Camp 1942).

ADDITIONAL REFERENCES. Hesse and Welles, 1936, p. 157. McDowell and Bogert,
1954, p. 132. Romer, 1956, p. 561.

Diacnosis. Premaxilla with very small rostrum anterior to premaxillary teeth.
Eighteen teeth in maxilla. Prefrontal excluded from posterolateral margin of external
nares by slender process of maxilla,! supraorbital wing with small rectangular ala
fused with postorbitofrontal posteriorly over orbits lateral to external edge of frontal.
Frontal not emarginate above orbits, median dorsal ridge absent. Pineal foramen large,
closely embraced on either side by short tongues from frontal. Median dorsal surface of
parietal short, extensively invaded posteriorly by insertional area for cervical epaxial
musculature. Ventral process of postorbitofrontal to jugal apparently nearly confluent
with well exposed dorsal surface of postorbitofrontal. Ventroposterior process of jugal
nearly absent. Twelve-15 teeth in pterygoid. Suprastapedial process of quadrate relatively
short, tympanic ala moderately thick with groove around external margin. Stapedial

1Camp (1942, fig. 6) figures the prefrontal of a skull referred to Plotosaurus tuckeri as being
narrowly separated from the narial margin by a thin lamina of bone from the maxilla, and the
postorbitofrontal-parietal contact as occurring in the center of the anterior border of the supra-
temporal fenestra. Later (ibid., fig. 13) in an illustration of the same skull the prefrontal is
shown to enter broadly into the narial margin, and the postorbitofrontal-parietal contact to
occur in the anteromedial corner of the supratemporal fenestra.

146 PEABODY MUSEUM BULLETIN 23

pit elliptical in form. Seventeen teeth in dentary. Dentary with relatively long projection
anterior to first dentary tooth. Dorsal edge of surangular very thin lamina of bone rising
anteriorly to position high on posterior surface of coronoid. Mandibular teeth bicarinate,
vertically striated, becoming shorter and blunter posteriorly.

Vertebral formula: 51 presacral vertebrae (number restored), 30 pygals, 5 caudals
with chevrons and transverse processes, 64 terminal caudals (see Camp, 1942, pl. 5).

Articulating surfaces of anterior dorsal vertebrae wider than deep, subcircular in
outline. Synapophysis located in center of lateral surface of cervical vertebrae and
anterodorsal portion of lateral surface of thoracic centra. Ventral border of anteroventral
extension of synapophysis poorly developed on cervicals and anterior dorsals, does not
reach level of undersurface of centrum ventrally. Anterior zygopophysis of cervicals
and anterior thoracics connected by gently rounded, posteriorly descending crest to
synapophysis. No zygosphene-zygantrum. Anterior base of atlas neural spine arises
behind condylar facet. Hypapophyseal peduncle located posteriorly on ventral surface
of cervical centra, articulation for hypapophysis flat and circular, initially inclined
posteriorly, becoming horizontal posteriorly. Neural spines of caudal vertebrae longest
and vertical on postsacrals 42-44. Nine to ten hypapophysis-bearing anterior presacrals.

Scapula larger than coracoid. Glenoid articulating surfaces apparently convex on
both scapula and coracoid, do not form a smoothly continuous surface. Superior border
of scapula gently convex, posterior border slightly emarginated posteriorly. Coracoid ex-
pands broadly behind glenoid articulation. Humerus short and broad. Proximal end
of radius expanded. Shaft of radius wide. Distal end bears moderately expanded
anterodistal flange. Radius and ulna subequal in size. Radius has distinct facet for
articulation with radiale.

Radiale very large. Intermedium of average proportions, has wide articulating
facet for ulna. Ulnare does not enter posteroventral border of antebrachial foramen.
Metacarpals shorter and stouter than in other genera of mosasaurs. Metacarpal one
equal to metacarpal two in length, anterodistal tuberosity present. Proximal ends of
metacarpals moderately expanded. Phalangeal formula of manus at least 10-10-10-7-2.

Discussion. The foregoing diagnosis was taken from descriptions and _ illustrations
in Camp (1942). There are at least 38 presacral vertebrae in Plotosaurus tuckeri (ibid., p.
21).

Plotosaurus bennisoni (Camp 1942)

Kolposaurus bennisoni Camp, 1942, p- 2, figs. 1-3, 13-15, 17, pl. 1
Plotosaurus bennisoni, Camp, 1951, p. 822.

Type. UCMP **32778, 225 feet above base of Garzas Sand, Moreno Formation,
Merced County, California (see Camp, 1942, p. 2 for detailed locality information).
Type specimen includes a complete skull, 18 vertebrae, an interclavicle and four ribs
(Camp, 1942, p. 8.)

DistrripuTion. Moreno Formation, California.

ADDITIONAL REFERENCES. Kuhn, 1952, fig. 15(1). Anderson, 1958, p. 71. Kauffman
and Kesling, 1960, p. 220, table 3, fig. 5.

Diacnosts. Tongues from frontal embracing parietal opening, long and rectangular.
Fifteen teeth in pterygoid. Main body of quadrate broad with moderately large sup-
rastapedial process. Ten hypapophysis-bearing cervicals. Four anterior dorsals bear
zygapophyses. Interclavicle relatively small (Camp, 1942, pp. 4, 11, figs. 1-2).

Plotosaurus tuckeri (Camp 1942)
(Text-figs. 46D, 47A, 52)

Kolposaurus tuckeri Camp, 1942, p. 8, figs. 4-10, 13, 23, pls. 2-3.
Plotosaurus tuckeri, Camp, 1951, p. 822.

Tyre. At UCMP, from the Moreno Formation, Fresno County, California (see

RUSSELL: AMERICAN MOSASAURS 147

Camp, 1942, p. 8, for detailed locality information). Type specimen includes 54 vertebrae
and a few limb fragments (Camp, 1942, p. 8).

DistTRIBUTION. Moreno Formation, California.

ADDITIONAL REFERENCES. Stock, 1939, p. 620, pl. 1 fig. b. Welles, 1943, pp. 127,
128, fig. 2. Romer, 1956, fig. 182b. Anderson, 1958, p. 71. Downs, 1959, p. 17.

REFERRED SPECIMENS. CIT nos. **2750, 2751, 2755 (fide Camp, 1942, p. 11).

Diacnosis. Tongues from frontal embracing parietal opening short and triangular.
Twelve to thirteen teeth in pterygoid. Main body of quadrate narrow with small
suprastapedial process. Nine hypapophysis-bearing cervicals. Eight anterior dorsals bear
zygapophyses. Interclavicle relatively large (Camp, 1942, pp. 11, 13).

FOREIGN GENERA OF MOSASAURINAE

Genus COMPRESSIDENS Dollo 1924
Compressidens Dollo, 1924, p. 176.

Generic Type. Compressidens fraasi (Dollo).

Diacnosts. Small projection of dentary anterior to first dentary tooth. Median den-
tary teeth bilaterally compressed, bicarinate, subrectangular in lateral view and with
pointed apices. Anterior teeth circular in cross-section with strongly recurved pointed
apices (Dollo, 1913, pl. 23; 1924, pl. 5).

Discussion. The genus Compressidens was proposed by Dollo for two species of
small Maestrichtian mosasaurs from Belgium, known only from a single dentary and
several teeth. The slender anteriorly narrowing form of the dentary is typically clidastoid
and the teeth could be easily derived from any of the Niobrara species of Clidastes.
There are at least thirteen teeth in the dentary (Dollo, 1913, p. 610). The genus may
be easily distinguished from Globidens by the compressed nature of the teeth and the
relatively delicate dentary. The teeth are unicuspid in C. fraasi and tricuspid in C. belgi-
cus. Compressidens and Globidens appear to be separate derivations from older clidastoid
stock, but because Compressidens is a less divergent form it is still included in the
Mosasaurini.

Genus TANIWHASAURUS Hector 1874

Taniwhasaurus Hector, 1874, p. 353.

Generic Type. Taniwhasaurus oweni Hector.

Diacnosis. Prefrontal lacks broad triangular ala projecting laterally from supra-
orbital wing. Frontal slightly emarginate above orbits, has prominent median dorsal
ridge. Small projection of dentary anterior to first dentary tooth. Mandibular teeth
only slightly compressed, lack carinae, and have finely striated surfaces.

Zygosphene rudimentary. Postglenoid process absent on proximoposterior end of
humerus. Radius very short and broad. Distal end bears heavy anterodistal flange and
distinct facet for articulation with radiale. Radiale very large. Intermedium of average
proportions, has articulating facet for ulna. Ulnare does not enter posteroventral
border of antebrachial foramen. Distal end of radius and ulna approach each other
medially above carpus.

Discussion. The foregoing diagnosis was taken from descriptions and illustrations
in Hector (1874).

Taniwhasaurus was described from the Maestrichtian of New Zealand (Wellman, 1959,
p- 135). The parietal (Hector, 1874, pl. 31 fig. A) is peculiarly reminiscent of that of
Platecarpus. The dentaries are unique in that the teeth are directed as much labially as
vertically and are separated from the buccal cavity by a median longitudinal parapet
(ibid., fig. B’). Could these cranial remains belong to Plioplatecarpus?

148 PEABODY MUSEUM BULLETIN 23

The forepaddle, however, shows (ibid., pl. 51 fig. C) marked mosasaurine affinities.
Unfortunately the humerus was neither well described nor clearly figured. The radius
is more than twice the size of the ulna, although in Plotosaurus they are subequal,
and has a well developed anterior flange which is reduced in the latter genus. Never-
theless Taniwhasaurus resembles Plotosaurus and differs from Clidastes and Mosasaurus
in that the radius and ulna nearly contact each other distally. It is therefore provisionally
assigned to the Plotosaurini.

SupraMity PLIOPLATECARPINAE (Dollo 1884) Williston 1897

Plioplatecarpidae Dollo, 1884, p. 653.
“mosasauriens microrhynques” Dollo, 1890, p. 163.
Platecarpinae Williston, 1897d, p. 177.

D1acnosis. Small rostrum present or absent anterior to premaxillary teeth. Twelve
or more teeth in dentary and maxilla. Cranial nerves X, XI and XII leave lateral wall
of opisthotic through single foramen. Canal or deep groove in floor of basioccipital and
basisphenoid for basilar artery. Suprastapedial process of quadrate large, bluntly ter-
minated and with parallel sides. Dorsal edge of surangular rounded and longitudinally
horizontal.

‘Twenty-nine or less presacral vertebrae present. Length of presacral series less than
that of postsacral, neural spines of posterior caudal vertebrae at most only slightly
elongated, do not form an appreciable fin. Haemal arches usually unfused to caudal
centra. Appendicular elements lack smoothly finished articular surfaces.

Discussion. Williston (1897d, p. 181; 1898b, p. 177) characterized the Plioplatecarpinae
as having an imperfectly ossified carpus and tarsus. This is true of Platecarpus where
only the intermedium, ulnare, two distal carpals and occasionally a diminutive ossicle
in the position of a radiale occur. But in Ectenosaurus the intermedium and ulnare are
large; there is a well ossified radiale; and four distal carpals are present. This makes a
total of seven elements in contrast to the six in the “fully ossified” carpus of most mosa-
saurines (excluding the pisiform). Further, only four ossified elements have been
found in the carpus of Clidastes sternbergi. However the ulnare broadley enters the
posteroventral margin of the antebrachial foramen in both Platecarpus and Ecteno-
saurus which, with the possible exception of Mosasaurus tvoensis (Williston, 1902, pl. 12
lower figure), is unknown to occur in the Mosasaurinae. Perhaps the arrangement, rath-
er than the degree of ossification of the carpals and tarsals, is more important in showing
relationships.

Trine PLIOPLATECARPINI (Dollo 1884) new

Diacnosis. Canal through basioccipital and basisphenoid for basilar artery. Delicately
proportioned jaws, teeth striated and subcircular in cross-section. Chevrons free. These
are generally mosasaurs of moderate size, with relatively short jaws and long, slender
teeth.

Grnus PLATECARPUS Cope 1869
(Text-figs. 7, 9-11, 16, 18-20, 28, 31, 40, 41, 44, 46C, 48B, 53, 57, 62)

Platecarpus Cope, 1869b, p. 264.
Holcodus, Cope, 1872f, p. 331.
Liodon in part, Cope, 1872f, p. 331.
Lestosaurus Marsh, 1872b, p. 454.
Sironectes Cope, 1874, p. 34.
Holosaurus Marsh, 1880, p. 87.
Generic Type. Platecarpus tympaniticus Cope.
ADDITIONAL REFERENCES. Leidy, 1865a, pl. 8 figs. 18-19; 1873, pl. 36 figs. 4-14. Cope,
1869-1870, p. 199; 1872d, pp. 269, 273; 1872e, p. 141; 1874, p. 35; 1875, pp. 139, 141, pl.

RUSSELL: AMERICAN MOSASAURS 149

26 figs. 2-3. Marsh, 1872a, p. 291; 1872b, pl. 12 fig. 2, pl. 13 figs. 3-4; 1872d, pp. 496, 497;
1880, pl. 1 figs. 2-3; 1897, fig. 23. Owen, 1877, pp. 690, 696, fig. 15; 1880, pl. 8 figs. 2-3.
Lydekker, 1888, p. 269. Hoffmann, 1890, p. 1321. Dollo, 1894, pl. 4 fig. 7. Merriam, 1894,
pp. 25, 38-39, pl. 3 figs. 3-5, 11, pl. 4 figs. 2, 8. Williston, 1891, p. 345; 1895, pl. 17 fig. 3;
1897c, pp. 107, 108; 1897d, pp. 178, 181, 183; 1897e, pp. 244, 245, figs. 1(1), 5(2); 1898b,
pp. 103-134, 177, 192, pl. 22 fig. 2, pl. 24 figs. 1-5, pl. 25 figs. 3-5, pl. 26 figs. 1-4, pl. 27,
pl. 28 figs. 5, 7-8, pl. 29 figs. 3-5, pl. 31 figs. 4-5, pl. 41 fig. 2, pl. 42 figs. 3-4, pl. 43, pl. 45
figs. 1-2, pl. 46 figs. 1-4, pl. 47, pls. 51-52, pl. 54 figs. 2-3, pl. 56 figs. 1-5, pl. 57 figs. 1-2,
5-10, pl. 58 figs. 1-10, pl. 60 fig. 61 fig. 3 pl. 63, pl. 64 figs. 2-3, text fig. 5; 1898c, p .235;
1899, p. 39; 1902, p. 248, pl. 12 upper fig., pl. 13; 1904, p. 48; 1914, p. 166, fig. 73; 1925,
p. 273, figs. 57, 80b, 123, 158a. Osborn, 1899a, figs. 2-3, 6a, 7a. Sternberg, 1899, p. 259;
1905, pp. 126, 127; 1907, pp. 122-123; 1908, p. 113; 1909, pp. 50, 204; 1911, p. 17; 1917,
p- 162; 1918, p. 205. Loomis, 1904, p. 254. Janensch, 1906, p. 30. Osburn, 1906, pl. 8
fig. 22. Capps, 1907, figs. 1-3. Merril, 1907, p. 73. Pompeckj, 1910, pp. 131, 137. Huene,
1911, p. 49, pl. 3; 1919, p. 183, pl. 8 fig. 2. Broom, 1913, p. 508. Dreverman, 1914, p. 43.
Lambe, 1914, pl. 1 lower fig. Wiman, 1920, p. 10, fig. 2, pl. 3 upper fig., pl. 4 fig. 7b. Abel,
1922, figs. 259, 263-264; 1924, fig. 36. Woodward, 1922, p. 6, fig. 2. Camp, 1923, p. 323;
1942, pp. 30-34, figs. 18-20, 24a-b, 25-26. Gilmore, 1928, p..87; 1943, fig. 13a. Nopcsa, 1928,
p- 177. Mehl, 1930, p. 393, figs. 4-5. Lane, 1947, p. 314. Gregory, 1951, figs. 4a, 5; 1952,
figs. 4 right, 9. Romer, 1956, p. 562, figs. 120f, 190d. Vaughn and Dawson, 1956, p. 383.

Diacnosis. Premaxilla with no rostrum anterior to premaxillary teeth. Twelve teeth
in maxilla. Prefrontal forms large portion of posterolateral border of external nares
and has small lateral tuberosity on supraorbital wing, contacts postorbitofrontal over
orbits medial to external edge of frontal. Frontal emarginate above orbits, has prom-
inent median dorsal ridge. Parietal foramen large, close to or bordered anteriorly by
frontal. Margins of dorsal parietal surface meet in front of diverging suspensorial rami,
forming triangular table anteriorly on parietal. Ventral process of postorbitofrontal
to jugal distinct from but nearly in same plane as moderately exposed dorsal surface
of postorbitofrontal. Ventroposterior process on jugal well developed. Squamosal wing
to parietal moderate to large. Otosphenoidal crest on prootic small, may cover exits
for cranial nerves VII and IX laterally. Canal for basilar artery through basioccipital
and basisphenoid not open ventrally on basioccipital. Ten to twelve teeth in pterygoid.
Suprastapedial process not fused to infrastapedial process on quadrate. Tympanic ala
very thin. Stapedial pit elliptical in form. Twelve (rarely eleven) teeth in dentary.
Dentary terminates abruptly in front of first dentary tooth. Angular widely separated
medially from coronoid. Retroarticular process of articular circular in outline.

Vertebral formula: 29 presacral vertebrae, 5 pygals, 26-31 caudals with chevrons
and transverse processes, about 65 terminal caudals (see Merriam, 1894, p. 11; Williston,
1898b, p. 143; 1910, p. 538; Huene, 1911, p. 50 and YPM 1350, 1429 and 24900).

Articulating surfaces of cervical and anterior dorsal vertebral centra wider than
deep, smoothly elliptical in outline. Synapophysis located high in center of lateral sur-
face of cervical centra, occupies anterodorsal portion of lateral surface of anterior dorsal
and anterior portion of posterior dorsal centra. Ventral border of anteroventral exten-
sion of synapophysis well developed on cervicals and anterior dorsals, reaches level of
undersurfaces of centrum. Anterior zygopophysis of cervicals and thoracics connected
by gently rounded, posteriorly descending crest to synapophysis. Zygosphene-zygantrum
present but not large. Anterior base of atlas neural spine arises directly above condylar
facet, atlas synapophysis small and tuberculate. Hypapophyseal peduncle located
posteriorly on ventral surface of cervical centra, articulation for hypapophysis horizontal
and triangular. Five hypapophysis-bearing cervicals, one or rarely two more with
rudimentary peduncles. Transverse processes of pygal vertebrae relatively short. Neural
spines of caudal vertebrae vertical on postsacrals 44-52 (Huene, 1911, pl. 3).

Scapula and coracoid subequal in size. Glenoid articulating surfaces convex on
both elements, do not form smoothly continuous surface. Superior border of scapula
strongly convex, posterior border strongly emarginated posteromedially. Coracoid
does not expand medially to point behind glenoid articulation. Distal end of humerus

PEABODY MUSEUM BULLETIN 23

150

Seem
arr NSE

FART

Text-fig. 83. Dorsal view of skull of Platecarpus ictericus (reconstructed after AMNH 1820, x 3). -

RUSSELL: AMERICAN MOSASAURS 151

AYALA 3
Gig ent

BASES
nt or)

A MIERON IES

Text-fig. 84. Ventral view of skull of Platecarpus ictericus (reconstructed after AMNH 1820,
x %).

152 PEABODY MUSEUM BULLETIN 23

very widely expanded, much more than proximal end. Facets for articulation with
other elements and sites of muscle attachment moderately well differentiated. Proximal
end of radius expanded. Shaft of radius wide. Distal end bears greatly expanded
anterodistal flange, is thickened medially to form indistinct facet for articulation with
intermedium.

Radiale reduced to small ossicle, or absent. Intermedium deep. Ulnare enters
posteroventral border of antebrachial foramen. Metacarpal one less than metacarpal
two in length, has moderate anterodistal tuberosity. Proximal ends of metacarpals
moderately expanded. Phalangeal formula of manus approximately 4-6-7-5-3 plus or
minus one phalanyx on each digit.

Acetabular surfaces of pelvic elements convex, do not form solid, smoothly surfaced
bowl. Obturator foramen located near center of proximal end of pubis, dorsoanterior
process small and triangular. Ischiadic tubercle located close to acetabulum. Shaft of
ischium broad, ischiadic symphysis very wide. Distal end of femur more expanded than
proximal; internal trochanter of average proportions, located anteromedially from head.
No distinct facets on distal end, although femur is thickened in region of tibial articula-
tion. Tibia and fibula difficult to distinguish from radius and ulna respectively. Cartilage
capped area of anterior flange extends from base of tibia one-half way up shaft, similar
area of posterodistal flange of fibula limited to distal end of bone.

Astragulus elliptical, dorsal notch between subequal facets for tibia and fibula
small. Calcaneum slightly smaller than fourth tarsal. Metatarsal one slightly expanded
proximally, slightly medially recurved. Metatarsal five not hook shaped, symmetrically
expanded proximally. Phalangeal formula of pes approximately 4-5-5-5-3 plus or minus
one phalanyx on each digit.

Discussion. The genus Platecarpus was named by Cope on fragmentary, though
generically diagnostic material from Mississippi. Apparently not realizing that it also
occurred in the Niobrara Chalk, he referred Niobrara species of the genus to Holcodus
and Liodon during 1871 and 1872, presumably on dental and vertebral characters
respectively. Marsh (1872b, p. 461) recognized the generic distinctiveness of these
species relative to other genera in the Niobrara and included them in his new genus
Lestosaurus. Cope (1872e, p. 141) immediately pointed out that, though agreeing with
Marsh that the type tooth of Holcodus acutidens was more closely related to Mosasaurus
(meaning M. copeanus = Plioplatecarpus), the species under dispute were very prob-
ably congeneric with Platecarpus tympaniticus. Thereafter Cope used Platecarpus for
these species and Marsh used Lestosaurus.

Platecarpus is indeed a senior synonym of Lestosaurus and as such has been generally
recognized since 1890. Leidy (1865a, p. 118) and Williston (1898b, p. 177) have sus-
pected that Platecarpus might be synonymous with Gibbes’ older genus Holcodus, here
considered a nomen vanum. Skeletal material preserved in the types of Sironectes an-
guliferus Cope and Holosaurus abruptus Marsh clearly shows that these genera are
identical to Platecarpus.

Platecarpus tympaniticus Cope 1869

?Holcodus acutidens, Leidy, 1865a, pp. 35, 38, 40-41, 118, 131, pl. 7 figs. 4-7, pl. 11 fig. 14.
Platecarpus tympaniticus Cope, 1869b, p. 265.

Tyre. ANSP 8484, 8487, 8488, 8491, 8558-9, 8562, all one specimen from a greenish
sandstone (Eutaw Formation) near Columbus, Mississippi (see Leidy, 1865a, p. 38)
collected by Dr. William Spillman. Type specimen includes a surangular, quadrate,
braincase, pterygoid and three cervical vertebrae.

ADDITIONAL REFERENCES. Leidy, 1865b, p. 70. Cope, 1869-1870, p. 200; 1872e, p. 141;
1874, p. 35; 1875, p. 267, pl. 37 fig. 11. Merriam, 1894, p. 30. Williston, 1897d, p. 185;
1898b, p. 179. Chaffee, 1939, fig. 1(4).

Discussion. The type and only specimen of the species, from the Eutaw Formation
of Mississippi, exhibits diagnostic characteristics of the genus Platecarpus, such as the

RUSSELL: AMERICAN MOSASAURS 153

large suprastapedial process and delicate tympanic ala of the quadrate, and general form
of the basioccipital and anterior vertebrae. The preserved cranial material of this speci-
men is identical in form to corresponding elements of the Niobrara species P. ictericus
and P. coryphaeus, and when anterior portions of the skull of P. tympaniticus are dis-
covered in the Eutaw Formation the species will doubtlessly be shown to be the senior
synonym of one of the Niobrara forms. It may be distinguished from Platecarpus cf. P.
somenensis from the Lower Pierre in that the ventroposterior processes of the basi-
sphenoid are separated by a shallow horizontal sulcus only, not by a deep longitudinal
cleft.

Platecarpus coryphaeus (Cope 1872)
(Text-fig. 85A)

Holcodus coryphaeus Cope, 1872d, p. 269.
Lestosaurus gracilis Marsh, 1872b, p. 460.
Platecarpus coryphaeus, Cope, 1874, p. 35.
Holosaurus abruptus Marsh, 1880, p. 87.

Tyrer. AMNH 1566, from “. . . the side of a bluff in a branch of the Fossil Spring
cafion . . . from the yellow chalk,” (Cope, 1872d, p. 272) collected by E. D. Cope. Type
specimen includes braincase, quadrates, frontal, pterygoid and seventeen vertebrae.

DistriBuTIon. Smoky Hill Member, Niobrara Formation, Kansas.

AppITIONAL RrFerENces. Leidy, 1873, pp. 276, 344. Cope, 1875, pp. 142, 267-268, pl.
15 fig. 1, pl. 16 fig. 3, pl. 17 fig. 6, pl. 20 figs. 4-7, pl. 21 figs. 1-2, pl. 36 fig. 6, pl. 37 fig.
9, pl. 55 fig. 8; 1877, p. 584. Owen, 1880, p. 180. Hoffman, 1890, p. 1322. Merriam,
1894, pp. 29, $0, pl. 1 figs. 1-2. Marsh, 1897, p. 527. Williston, 1898b, p. 186. Merril, 1907,
p. 73. Sternberg, 1909, fig. 10. Drevermann, 1914, fig. 8. Abel, 1922, fig. 265. Gilmore,
1928, p. 87. Lane, 1947, pp. 316, 317. McDowell and Bogert, 1954, p. 132. Edmund, 1960,

phe
E REFERRED SPECIMENS. AMNH nos. *126, *127, *202, 1510, *1511, **1512, 1645. YPM
nos. *1264 (P. gracilis), **1350 (P. abruptus), 1427, *1428, ** 3972, 4004 and twenty-two
unnumbered specimens, USNM nos. **3774, **3791, 11654, 11655 and one unnumbered
specimen. MCZ nos. 1606, 1607, *1610, 1614. KU nos. **1007, 1092, 1143, 5016 and two
unnumbered specimens. FHM no. **10902.

Diacnosts. Premaxillo-maxillary suture terminates posteriorly at point above third
maxillary tooth. Pineal foramen large, occasionally reaches frontal anteriorly. Squamosal
sends very abbreviated wing medially to contact ramus from parietal. Ventroposterior
process on jugal moderately large. Otosphenoidal crest on prootic does not cover exits
for cranial nerves VII and IX. Articular surfaces of basipterygoid processes large and
laterally directed. Ventral posterior processes of basisphenoid separated by shallow
emargination. Coronoid descends at steep angle posteriorly to dorsal edge of surangular.
Exits for mandibular ramus of fifth nerve separate into two diverging rows anteriorly
on dentary and reconverge at ventroanterior margin of bone. Mandibular teeth bi-
carinate, vertically striated and medially recurved.

Discussion. Platecarpus coryphaeus is as far as is known almost inseparable from
P. ictericus. Only the position of the external nares and exits for the mandibular
ramus of the fifth nerve can be used to distinguish them. There is no diagnostic cranial
material remaining in the type, although Cope (1875, p. 143) describes a maxilla which
has evidently been lost. This element was figured (ibid., pl. 55 fig. 3) in a reconstruc-
tion of the type cranium as having the greatest depth of the maxilla anteriorly, which
in this species coincides with the posterior termination of the premaxillo-maxillary su-
ture, at a point dorsal to the third maxillary tooth. Therefore the name P. coryphaeus
is tentatively conserved for this species, even though the type specimen may have come
from a level high in the chalk (see discussion on p. 187). P. gracilis and P. (Holosaurus)
abruptus may be recognized as junior synonyms on the basis of the diagnostic cranial
characters preserved in their respective types.

154 PEABODY MUSEUM BULLETIN 23

Text-fig. 85. Lateral view of anterior tip of dentary. A. Platecarpus coryphaeus (AMNH 1511,
x 6). B. Platecarpus ictericus (AMNH 1488, x 64).

This species has not been identified from the Pierre and, although locality data is
thus far insufficient, future work may show that it is absent in the upper part of the
Niobrara as well.

Platecarpus ictericus (Cope 1871)
(Text-figs. 2A, 4A, 5B, 6, 17, 22A, 23, 24B, 25, 29, 37, 38, 83, 84, 85B; Plate II, fig. 2)

Liodon ictericus Cope, 1871a, p. 572.

Holcodus ictericus, Cope, 1872a, p. 298.

Liodon curtirostris Cope, 1872a, p. 298 nomen nudum.
Liodon curtirostris Cope, 1872d, p. 10.

Lestosaurus simus Marsh, 1872b, p. 455.

Lestosaurus felix Marsh, 1872b, p. 457.

Lestosaurus latifrons Marsh, 1872b, p. 458.
Platecarpus ictericus, Cope, 1874, p. 35.

Tyre. AMNH 1559, from “. . . the north bank of the Smoky Hill River, thirty
miles east of Fort Wallace, Kansas,” (Cope, 1871b, p. 580) collected by B. F. Mudge.
Type specimen includes a partial skull, 18 vertebrae and fragments of a pectoral girdle.

DistripuTION. Smoky Hill Member, Niobrara Formation, Kansas; Lower Pierre
Formation, Kansas, Wyoming and South Dakota.

RUSSELL: AMERICAN MOSASAURS 155

ADDITIONAL REFERENCES. Cope, 187lc, p. 152; 1871f, p. 406; 1872b, p. 170; 1872c, p.
176; 1872d, pp. 272, 273; 1872e, p. 141; 1872£, p. 331; 1874, pp. 35, 36; 1875, pp. 144,
150, 267, 268, pl. 14 figs. 3-4, pl. 15 figs. 2-3, pl. 16 figs. 4-5, pl. 17 figs. 2-4, pl. 18 figs. 6-8,
pl. 19 figs. 7-10, pl. 20 figs. 1-3, pl. 21 figs. 7-13, pl. 25 figs. 1-24, pl. 36 figs. 5, 7, pl. 37 figs.
8, 10, pl. 38 fig. 1; 1877, p. 584. Marsh, 1872b, p. 461, pl. 10; 1880, figs. 2-4; 1897, p. 527.
Owen, 1877, p. 687; 1878, fig. 4; 1879, p. 58. Lydekker, 1888, fig. 59. Baur, 1892, pp. 1-22,
pls. 1-2. Merriam, 1894, p. 30. Williston, 1897e, fig. 7; 1898a, p. 29; 1898b, pp. 184, 188,
189, pls. 13-15, pl. 25 figs. 1-2; pl. 44, pl. 72 fig. 2; 1902, p. 253; 1910, p. 537, 1 fig.; 1914,
figs. 69, 72b; 1925, fig. 54, 184. Osburn, 1906, pl. 8 fig. 28. Pompeckj, 1910, p. 137.
Wiman, 1920 figs. 7a-b. Lane, 1947, p. 315. Kauffman and Kesling, 1960, fig. 6b. Russell,
1964, figs. 7-8.

REFERRED SPECIMENS. AMNH nos. 515, **1488, 1528, 1532, 1550, 1563 (P. curtirostris),
1588, 1820, 1821, 2178, **2182, *5811, 6159. YPM nos. *1112, 1138, 1153, *1254, 1255,
**1256 (P. latifrons), *1258 (P. simus), 1267, **1269, 1272, 1273, **1277 (P. felix), *1284,
1286, *1292, 1306, *3690, 3986, *3994, *3998, **4003, and sixty-five unnumbered speci-
mens. USNM no. 18256 and five unnumbered specimens. MCZ nos. 1601, 1602, 1603,
1605, 1608, 1609, 1615, 1624. KU nos. 1026, **1031, 1046, *1052, 1054, **1081, 1091,
1093, 1094, 1097, 1114, 1118, 1135, 1140, #1142, 1196, 5042. FHM nos. **2913, *10760,
**10856. SDSM no. **30139. CNHM nos. **UC600, **P12027. NMC ?no. **8163. PU
no. 12999.

Diacnosts. Premaxillo-maxillary suture terminates posteriorly dorsal to midpoint
between second and third maxillary tooth, where anterior portion of maxilla has
greatest depth. Exits for mandibular ramus of fifth nerve separate into two parallel rows
anteriorly on dentary and terminate at ventroanterior margin of bone. Otherwise as
for P. coryphaeus.

Discussion. The arrangement of exits for the mandibular ramus of the fifth nerve
on a type dentary of P. ictericus, the earliest name available for a Niobrara Platecarpus,
shows it to be conspecific with the species diagnosed above. The type specimens of P.
curtirostris, P. simus, P. felix and P. latifrons are also conspecific.

This has been by far the most frequently collected mosasaur from the Niobrara
Chalk. Circumstantial evidence supporting its specific distinction from P. coryphaeus
lies in the fact that crania belonging to P. ictericus are usually less completely preserved
(see text-fig. 96), and that P. coryphaeus has not yet been identified from the Pierre, al-
though P. ictericus does seem to occur in the Lower Pierre in some abundance.

Platecarpus cf. P. somenensis Thevenin, 1896, p. 907.

DistrisuTIon. Sharon Springs Member, Lower Pierre Formation, South Dakota.

REFERRED SPECIMENS. CNHM nos. PR465, *PR466, **PR467. YPM one unnumbered
specimen.

Diacnosis. Premaxillo-maxillary suture terminates posteriorly above midpoint be-
tween second and third maxillary tooth, anterior portion of maxilla has greatest depth
dorsal to third maxillary tooth. Parietal foramen very large and broadly bordered
anteriorly by frontal. Squamosal sends broad rectangular wing medially to contact ramus
from parietal. Ventroposterior process on jugal very heavy. Small otosphenoidal crest
on prootic covers exits for cranial nerves VII and IX. Articular surfaces of basipterygoid
processes small and anteriorly directed. Ventral posterior processes of basisphenoid
separated by deep median longitudinal cleft. Coronoid descends posteriorly at angle of
45° to dorsal edge of surangular. Exits for mandibular division of fifth nerve separate
into two parallel rows anteriorly on dentary, both of which terminate at ventroanterior
margin of bone. Mandibular teeth heavy, bicarinate, vertically striated and medially
recurved.

Discusston. This form is surely distinct from P. coryphaeus and P. ictericus, although
it is sometimes difficult to distinguish from the latter species due to the poor quality of
bone preservation in the Pierre. The most obvious resemblances between this form and

156 PEABODY MUSEUM BULLETIN 23

P. somenensis, from strata of similar age (middle Campanian) of France, lie in the large-
ness of the teeth and the heaviness of the posteroventral process on the jugal (see
Thevenin, 1896, pl. 30). However the type of P. somenensis is not sufficiently well known
to make the assignment certain. Thevenin (1896, p. 910) says there is no canal through the
basioccipital-basisphenoid, but this statement cannot be relied on for he also states that
it is absent in Niobrara specimens of Platecarpus.

“Platecarpus” intermedius (Leidy 1870)
Clidastes intermedius Leidy, 1870, p. 4.

Type. ANSP 9023-4, 9029, 9092-4, all one specimen, from “. . . Rotten Limestone
(Selma) Upper Cretaceous of Pickens Co., Alabama,” (Cope, 1869-1870, p. 221) collected
by Dr. J. C. Nott. Type specimen includes two dentary fragments and three vertebrae.

ADDITIONAL REFERENCES. Cope, 1871f, p. 412; 1875, p. 267. Leidy, 1873, pp. 281, 344,
pl. 34 figs. 1-2, 4-5, 10. Merriam, 1894, p. 36. Williston, 1898b, p. 197. Zangerl, 1948, p. 15.

Dracnosis. Exits for mandibular division of fifth nerve diverge anteriorly on dentary,
one row to parallel longitudinal axis of dentary, other to descend to ventroanterior
margin of bone. Posterior dentary teeth short and inflated.

Discussion. This species is known only from the type specimen. The general heaviness
of the dentary and its abrupt termination directly in front of the first tooth resemble the
condition of this element in Platecarpus more closely than in Clidastes. The axis vertebra
likewise is very similar to that of Platecarpus, differing only in the more central posi-
tion of the hypapophyseal facet on its ventral margin (see Leidy, 1873, pl. 34 fig. 10).
Leidy (1873, p. 282) notes that the zygosphenes of the two mutilated dorsals are as well
developed as in Clidastes. The generic reference of this specimen to Platecarpus must be
regarded as provisional, pending the discovery of new material. If it truly belongs in
the Plioplatecarpinae, the inflated tooth bases and swollen teeth are unique among
the members of the subfamily and show an interesting parallel to its contemporary
Globidens alabamaensis.

Genus ECTENOSAURUS new genus

EtyMotocy. Ectenes, Gr., drawn-out, in reference to the elongated muzzle; sauros, Gr.,
lizard.

Generic Type. Ectenosaurus clidastoides (Merriam)

Diacnosis. Premaxilla with small rostrum anterior to premaxillary teeth. Seventeen
(?) teeth in maxilla. Prefrontal widely excluded from posterolateral margin of external
nares by wing of maxilla and lacks process on supraorbital wing. Frontal emarginate
above orbits, median dorsal ridge rounded. Parietal foramen large, situated in center of
parietal. Margins of dorsal parietal surface converge, may meet, or before meeting may
flow into gently rounded posterior axis of parietal. Ventral process of postorbitofrontal
to jugal meets poorly exposed dorsal surface of postorbitofrontal at right angle.
Ventroposterior process on jugal well developed. Squamosal wing to parietal moderately
developed. Otosphenoidal crest on prootic does not cover exit for cranial nerve VII
laterally. Canal for basilar artery through basioccipital and basisphenoid. Nine to 11
teeth in pterygoid. Suprastapedial process of quadrate broadly fused distally with infrasta-
pedial process. Tympanic ala very thin. Stapedial pit rectangular in form. 13 teeth in
dentary. Dentary terminates abruptly in front of first dentary tooth. Angular widely
separated medially from coronoid. Mandibular teeth bicarinate, vertically striated and
medially recurved.

Vertebral formula: at least 29 presacral vertebrae (see FHM no. 7937).

Articulating surfaces of cervical and thoracic vertebral centra wider than deep,
smoothly elliptical in outline. Synapophysis located high in anterior portion of lateral

RUSSELL: AMERICAN MOSASAURS 157

Text-fig. 86. Dorsal view of skull of Ectenosaurus clidastoides (reconstructed after FHM
7937, X Y)-

surface of cervical centra, occupies anterodorsal portion of lateral surface of anterior
dorsal centra. Ventral border of anteroventral extension of synapophysis well developed
on cervicals and anterior dorsals, does not reach level of undersurface of centrum.
Anterior zygopophysis of cervicals and thoracics connected by gently rounded, posteriorly

158 PEABODY MUSEUM BULLETIN 23

descending crest to synapophysis. Zygosphene-zygantrum strongly developed. Anterior
base of atlas neural spine arises behind condylar facet, atlas synapophysis small and flat.
Hypapophyseal peduncle located posteriorly on ventral surface of cervical centra, articula-
tion for hypapophysis horizontal and triangular. Six hypapophysis-bearing cervicals, one
more with rudimentary peduncle. Transverse process of pygal vertebrae relatively long.

Scapula smaller than coracoid. Glenoid articulating surface concave on scapula, pro-
bably smoothly continuous with that of coracoid. Superior border of scapula gently
convex, posterior border slightly emarginated posteromedially. Coracoid expands medially
to point behind glenoid articulation. Distal and proximal ends of humerus moderately
expanded, facets for articulation with other elements and sites of muscle attachment
moderately well differentiated. Proximal end of radius not expanded. Shaft of radius
moderately wide. Distal end bears greatly expanded anterodistal flange.

Radiale present. Intermedium of average proportions. Ulnare enters broadly into
posteroventral margin of antebrachial foramen. Metacarpal one equal to metacarpal
two in length, has very slight anterodistal flange. Proximal ends of metacarpals moderately
expanded. Phalangeal formula of manus approximately 4-4-5-5-?.

Discussion. The form of the teeth, shape of the frontal, large suprastapedial process
of the quadrate and presence of a canal for the basilar artery through the basioccipital
show that Ectenosaurus is closely related to Platecarpus. In addition, the narial emargin-
ation begins above and a little behind the third maxillary tooth and the exits for the
mandibular ramus of the fifth nerve diverge anteriorly on the dentary of Ectenosaurus,
greatly increasing the probability of confusing it with Platecarpus coryphaeus (as the
author has done on at least four occasions). However, the elongated snout, exclusion
of the prefrontals from the narial borders, fusion of the supra- and infrastapedial
processes of the quadrate and peculiarities of the forelimbs necessitate the establishment
of a new genus.

Ectenosaurus clidastoides (Merriam 1894)
(Text-figs. 54, 86)

Platecarpus clidastoides Merriam, 1894, p. 30.
Platecarpus oxyrhinus Merriam, 1894, p. 30.

Type. Formerly at Bayerische Staatssammlung fiir Paladontologie but probably
destroyed during the Second World War (personal communication, R. Dehm 1963),
from “. . . Niobraraschichten der oberen Kreide von Logan County, Kansas . . .”” (Mer-
riam, 1894, p. 3) collected by C. H. Sternberg or G. Baur. Type specimen includes a
posterior portion of the skull and several vertebrae.

DistripuTion. Smoky Hill Member, Niobrara Formation, Kansas.

ADDITIONAL REFERENCES. Cope, 1875, pl. 25 figs. 25-26? Williston, 1898b, p. 190. Lane,
1947; p: 317.

REFERRED SPECIMENS. YPM nos. 4671, *4672, **4673, *4674. FHM no. **7937.

Locality of FHM no. 7937; from the Niobrara Chalk on Garrett Ranch, seven or
eight miles northwest of Wakeeney, Trego County, Kansas (personal communication,
George Sternberg 1963).

Discussion. The above is the only species presently referable to the genus. The
morphologic characteristics of a complete skull in the Fort Hays Kansas State College
Museum demonstrate that the posterior portion of the skull of the type of Platecarpus
clidastoides and the type muzzle of P. oxyrhinus actually belong to the same species,
if not the same individual.

Genus PLIOPLATECARPUS Dollo 1882

Plioplatecarpus Dollo, 1882, p. 62.
Phosphorosaurus Dollo 1889c, p. 68.
Oterognathus Dollo, 1889c, p. 69.

RUSSELL: AMERICAN MOSASAURS 159

Generic Tyre. Plioplatecarpus marshi Dollo.

DrAcnosis. Premaxilla with no rostrum anterior to premaxillary teeth (Dollo, 1889b,
p- 275). Prefrontal forms large portion of posterolateral border of external nares, small
lateral tuberosity on supraorbital wing. Prefrontal and postorbitofrontal do not meet
above orbits. Frontal rectangular in outline, not emarginate above orbits and has
median dorsal ridge. Parietal foramen very large, bordered anteriorly by frontal. Mar-
gins of dorsal parietal surface meet in front of diverging suspensorial rami forming tri-
angular table anteriorly on parietal (see Dollo, 1889b, pl. 9 fig. 6). Exit for cranial nerve
VII opens directly on lateral surface of prootic, no suggestion of otosphenoidal crest on
prootic (DeViller, 1943, pl. 2 fig. 1). Canal for basilar artery through basioccipital and
basisphenoid opens ventrally on basioccipital (Dollo, 1885, p. 319). Suprastapedial process
not fused to very large infrastapedial process on quadrate. Tympanic ala moderately
thick, laterally cup-shaped. Stapedial pit reniform in outline (Dollo, 1904, pl. 6). Den-
tary teeth reduced in number. Dentary terminates abruptly in front of first dentary tooth
(Dollo, 1889b, pl. 10 fig. 14). Mandibular teeth bicarinate (?), vertically striated and
medially recurved (Dollo, 1882, p. 64; 1889b, pl. 10 fig. 14).

Vertebral formula: 38 vertebrae anterior to chevron-bearing caudals, 55 caudals with
chevrons and transverse processes, 33 terminal caudals (Dollo, 1894, pp. 233-234).

Articulating surfaces of cervical and anterior dorsal centra wider than deep, smoothly
elliptical in outline. Synapophysis located in center of lateral surface of cervical and
anterior dorsal centra, occupies anterior portion of lateral surface of posterior dorsal
centra. Ventral border of anteroventral extension of synapophysis well to weakly developed
on cervicals and anterior dorsals, directed horizontally forward from ventral edge of
synapophysis on a level with undersurface of centrum. Gently rounded, posteriorly
descending crest connecting anterior zygopophysis with synapophysis present or absent
on cervical and anterior dorsal vertebrae. No zygosphene or zygantrum. Hypapophyseal
peduncle located in center of ventral surface of cervical centra, articulation for
hypapophysis horizontal and circular. Six hypapohysis-bearing cervicals (Dollo, 1894,
p- 235). Neural spines of caudal centra vertical on postsacrals 54-72, decrease uniformly
in length from base to end of tail (Dollo, 1894, pp. 235-236).

Scapula larger than coracoid. Glenoid articulating surfaces convex on both elements,
do not form smoothly continuous surface. Superior border of scapula gently convex,
posterior border may or may not be emarginated. Coracoid expands medially to point
behind glenoid articulation. Distal and proximal ends of humerus moderately expanded,
shaft of humerus short and broad. Facets for articulation with other elements and sites of
muscle attachment poorly differentiated. Proximal end of radius expanded. Shaft of
radius very wide. Distal end bears greatly expanded anterodistal flange (last paragraph
based largely on Dollo, 1882, pl. 6).

Discussion. The diagnosis is based on the fragmentarily known North American
species Plioplatecarpus primaevus and P. depressus, supplemented by Dollo’s descriptions
of the European P. marshi (1882, 1885b, 1904a, 1917) and P. houzeaui (1889b, 1894,
1904a). Only the last species is known from complete material. The other three forms
surely belong to Plioplatecarpus and it will probably be necessary to modify the above
diagnosis somewhat as their osteology becomes better known. If Dollo’s (1894, pp. 233-
234) estimate of the vertebral count is correct, Plioplatecarpus is the only mosasaur to
reduce the number of dorsal vertebrae during the course of its evolution.

Plioplatecarpus primaevus new species
(Text-fig. 87)

Platecarpus cf. P. brachycephalus, Dunkle, in Crandell, 1958, p. 14, fig. 8.
Tyrer. USNM 18254, from NE corner sec. 35, T112N, RSOW, near Pierre, Hughes

County, South Dakota (Crandell, 1958, fig. 8) collected by D. R. Crandell. Type speci-
men includes quadrate, three anterior dorsal vertebrae and both scapulae-coracoids.

160 PEABODY MUSEUM BULLETIN 23

DistR1BUTION. DeGrey Member, Lower Pierre Shale, South Dakota.

Diacnosis. Large prominence projects posteriorly beneath stapedial pit on quadrate,
nearly closes stapedial notch from below. Anterior zygopophysis of cervical and anterior
dorsal vertebrae connected to synapophysis by gently rounded, posteriorly descending

&

Text-fig. 87. Medial view of right quadrate of type of Plioplatecarpus primaevus (USNM
18254, X 34).

crest. Functional zygopophyses probably end near posterior end of dorsal series, or in
pygal region. Posterior border of scapula strongly emarginated posteromedially. Coracoid
foramen small.

Discussion. Plioplatecarpus primaevus is confidently referred to Plioplatecarpus on
the basis of the characteristic shape of its quadrate, the elongate, gently-convex superior
border of the scapula, and the greater size of the scapula when compared to the cora-
coid. It differs widely from P. marshi in the larger size of the posterior prominence
beneath the stapedial pit on the quadrate, its much smaller coracoid foramen and in
the presence of a posteromedial emargination on the posterior border of the scapula (see
Dollo, 1904a, pl. 6; 1882, pl. 6). The quadrate of P. primaevus is very similar to that of
P. houzeaui, but the former species differs in that the zygopophyses are well developed
in the anterior dorsal region while in the latter they disappear on the third dorsal
(Dollo, 1894, p. 235). In P. primaevus the anterior zygopophyses of the anterior dorsal
vertebrae are linked posteriorly to the synapophyses by a strong crest that is completely
lacking in P. depressus. It is seen that, except for the absence of zygosphene-zygantra,
the anterior vertebrae of P. primaevus are nearly identical to those of Platecarpus. It
is likely therefore, that this species had a presacral count closer to 29 vertebrae, as in
Platecarpus, than to the reduced number of 20 found in P. houzeaut. Plioplatecarpus
primaevus must be very closely related to Platecarpus.

Plioplatecarpus depressus (Cope 1869)
(Text-fig. 88)

Mosasaurus depressus Cope, 1869b, pp. 262, 263, nomen nudum.
Mosasaurus depressus Cope, 1869e, p. 16.

Mosasaurus copeanus Marsh, 1869, p. 393.

Halisaurus fraternus Marsh, 1869, p. 397.

?Liodon laticaudus Marsh, 1870, p. 2.

RUSSELL: AMERICAN MOSASAURS 161

Tyre. Specimen not located, originally “. . . from Burlington Co., N. J.,” (Cope,
1869-1870, p. 196) collected by L. T. Germain. Type specimen includes a quadrate,
cranial fragments and dorsal vertebrae (Cope, 1869-1870, p. 196).

DistripuTion. Navesink Formation or younger Cretaceous, New Jersey.

AppDITIONAL REFERENCES. Leidy, 1865a, pl. 7 figs. 2-3, ?8, 79-11. Cope, 1868b, p. 734;
1869-1870, pp. 189, 198, 210, fig. 48(3), ?pl. 5 fig. 4, pl. 11 fig. 6; 1875, pp. 269, 270, 272,
pl. 37 fig. 12. Marsh, 1872b, p. 455. Merriam, 1894, p. 37. Williston, 1898b, p. 207. Hay,
1902, p. 472 (L. laticaudus). Gilmore, 1926, p. 191, pl. 62 fig. 5. Miller, 1955, pp. 905, 909.

REFERRED SPECIMENS. AMNH nos. 1382, 1430, 2304. YPM nos. 312 (P. copeanus),
411 (L. laticaudus), 445 (P. fraternus).

D1acnosis. Anterior zygopophyses of cervical and anterior dorsal vertebrae separated
from synapophyses by longitudinal sulcus, lateral surface of neural arch smooth. Func-
tional zygopophyses probably end near center of dorsal series.

Discussion. The quadrate belonging to the type of M. depressus, figured by Cope
(1869-1870, fig. 48(3), pl. 11 fig. 6), has a large suprastapedial process with parallel sides,
as is the case in all members of the Plioplatecarpinae. It also has a relatively thick
tympanic ala and, in medial aspect, resembles the quadrate of Plioplatecarpus more
closely than that of any other genus of mosasaur. The New Jersey quadrate cannot be
confused with Platecarpus because, lying posterior to the ventral margin of the stapedial
pit, there is a prominence that nearly closes off the stapedial notch from below, which
is lacking in the latter genus.

The blunt termination of the dentary immediately anterior to the first dentary
tooth and the bicarinate, medially recurved and vertically striated teeth in the type speci-
men of M. copeanus all strongly support the inclusion of this species in the Plioplate-

Text-fig. 88. A. Lateral view of tip of left dentary, B. Lateral view of mandibular tooth.
C. Medial aspect of the left splenial. Plioplatecarpus depressus (YPM 312, x 34).

carpinae. The type is not complete enough to determine conclusively whether it be-
longs to Platecarpus, Ectenosaurus or Plioplatecarpus, but because it was collected from
strata of Maestrichtian age, in which the former two genera are unknown to occur, it
is much more probable that it belongs to Plioplatecarpus.

The type of Halisaurus fraternus consists of one cervical and two dorsal vertebrae.
They are referred to Plioplatecarpus because of the depressed, elliptical shape of the
central articulations and lack of zygantra on the cervical vertebrae (see Dollo, 1894, pp.
234-235 for descriptions of the same characters in the vertebral column of Plioplate-
carpus houzeaut).

The types of M. copeanus and H. fraternus were collected from Maestrichtian green-
sands in New Jersey, as probably was the missing type of M. depressus. All three of these
specimens are referable to Plioplatecarpus and in the absence of any evidence to the
contrary they are here assumed to belong to a single species, the oldest name available

162 PEABODY MUSEUM BULLETIN 23

for which being Plioplatecarpus depressus. If this is correct, the American Maestrichtian
P. depressus will resemble the European P. marshi most closely, but may be distinct in
that the known teeth are less medially recurved (see Dollo, 1913, pl. 25 fig. 2).

The type of Liodon laticaudus consists of a single mutilated caudal vertebra with
facets for articulation with the haemal arches. It may belong to P. depressus or perhaps
Prognathodon rapax. Additional caudal material of these two forms will be needed
to assign this vertebra properly.

Trine PROGNATHODONTINI new

Diacnosis. Deep groove in floor of basioccipital and basisphenoid for basilar artery.
Massively proportioned jaws, teeth faceted or smooth and elliptical in cross-section.
Chevrons either free or fused to caudal centra. The following mosasaurs are clearly of
plioplatecarpine derivation but have massive jaws and a heavy dentition much more
suited to crushing prey than that of members of the Plioplatecarpini.

Genus PROGNATHODON Dollo 1889

Prognathodon Dollo, 1889a, p. 214.

Prognathosaurus Dollo, 1889b, p. 293, pl. 9 figs. 4-5, pl. 10 figs. 8-9.
Brachysaurus Williston, 1897a, p. 96, pl. 8 (preoccupied Hallowell 1856).
Brachysaurana Strand, 1926, p. 54.

Ancylocentrum Schmidt, 1927, p. 59.

GENERIC TYPE. Prognathodon solvayi Dollo.

ADDITIONAL REFERENCES. Leidy, 1865a, pl. 3 figs. 1-2. Cope, 1869-1870, fig. 49 (p. 205)?.
McDowell and Bogert, 1954, p. 132. Romer, 1956, p. 562.

Dracnosis. Premaxilla with no rostrum anterior to premaxillary teeth. Twelve teeth
in maxilla. Prefrontal forms large portion of posterolateral border of external nares;
supraorbital wing with heavy triangular ala contacts postorbitofrontal posteriorly over
orbit medial to external edge of frontal. Frontal not emarginate above orbits, median
dorsal ridge present. Parietal foramen small, located anteromedially on small prominence
on parietal, and closely embraced on either side by short tongues from frontal. Margins
of dorsal parietal surface parallel one another and cranial midline to posterior base of
diverging suspensorial rami, forming rectangular field medially on parietal. Ventral
process of postorbitofrontal to jugal indistinctly separated from moderately well ex-
posed dorsal surface of postorbitofrontal. Ventroposterior process on jugal slightly de-
veloped. Squamosal wing to parietal large. Deep groove in floor of basioccipital for basilar
artery. Seven teeth in pterygoid. Suprastapedial process distally fused to infrastapedial
process on quadrate. Tympanic ala thick. Stapedial pit nearly circular in form. Four-
teen teeth in dentary. Dentary terminates abruptly in front of first dentary tooth.
Medial wing from coronoid contacts angular. Retroarticular process of articular rectan-
gular in outline. Mandibular teeth stout, bicarinate.

Vertebral formula: The type of Prognathodon solvayi from the Maestrichtian of
Belgium has at least 38 vertebrae anterior to the chevron-bearing caudal series (Dollo,
1889b, p. 294). Platecarpus has only 34 such vertebrae. ‘The differing counts may be due
to the incorporation of several caudals into the pygal region of Prognathodon.

Articulating surfaces of thoracodorsal vertebrae wider than deep, smoothly elliptical
in outline. Synapophysis located anterodorsally on lateral surface of centrum. Anterior
zygopophysis of thoracodorsals swollen, connected by gently rounded, posteriorly des-
cending crest to synapophysis (see AMNH 1562, type of P. crassarius).

Articulating surfaces of lumbodorsal vertebrae wider than deep, gently convex at base,
with dorsomedially curving lateral borders. Superior border narrow. Synapophysis lo-
cated at dorsoventral midpoint of anterior half of lateral surface of lumbodorsal centra.

RUSSELL: AMERICAN MOSASAURS 163

Text-fig. 89. Dorsal view of skull of Prognathodon overtoni (slightly reconstructed, after SDSM
3393, X 14).

164 PEABODY MUSEUM BULLETIN 23

Small, sharply defined crest of bone rises rather steeply from synapophysis to base of an-
terior zygopophysis (see NJGS 9827, P. rapax).

Zygosphene-zygantrum rudimentary or absent (Cope, 1875, p. 153; Dollo, 1889b, p.
297; Williston, 1897a, p. 96). Chevrons articulating in P. solvayi (Dollo, 1890, footnote
p- 163) and in P. crassartus; fused in P. overtoni (Williston, 1898b, p. 192). Neural
spines of caudal centra decrease uniformly in length from base to end of tail in P.
solvayi (Dollo, 1917, p. 20).

Scapula of moderate proportions, probably smaller than coracoid (Dollo, 1889b, p.
295). Distal and proximal ends of humerus moderately expanded, facets for articulation
with other elements and sites of muscle attachment moderately well differentiated (Cope,
1875, pl. 26 fig. 9; Williston, 1898b, pls. 30 fig. 6, 22 fig. 1). Proximal end of radius
greatly expanded, shaft of radius very wide. Distal end bears greatly expanded anterodistal
flange, is thickened medially to form indistinct facet for articulation with intermedium
(Cope, 1875, pl. 26 fig. 11).

Distal and proximal end of femur about equally expanded, shaft of femur con-
stricted. Internal trochanter moderately large, medial to head. No distinct facets on
distal end, although femur is thickened in region of tibial articulation (Cope, 1875, pl.
26 fig. 10).

Discusston. Although excellent postcranial material of Prognathodon has been col-
lected from Belgium (see Dollo, 1889b, pp. 293-294; 1890, p. 163; 1917, p. 20) none of
it has been described, and of necessity the foregoing diagnosis is based on very incom-
plete material from this continent. The chevrons are articulating in P. solvayi and P.
crassartus although Williston (1898b, p. 192) explicitly states they are fused to the caudal
centra in the type of P. overtoni. For this reason he was reluctant to synonymize his
Brachysaurus with Dollo’s Prognathosaurus (=Prognathodon) in which the chevrons are
free. Chaffee (1939, p. 2) asserts that there is no indication that the chevrons are fused
to the caudals of Ancylocentrum hungerfordi (=Prognathodon rapax) on the mistaken
assumption that the lumbodorsal vertebrae of his specimen were caudals.

Prognathodon crassartus (Cope 1872)

Liodon near L. proriger Cope, 1872b, p. 168.
Liodon crassartus Cope, 1872d, p. 278.
Platecarpus crassartus, Cope, 1874, p. 36.
?Platecarpus crassartus, Williston, 1898b, p. 180.

Tyre. AMNH 1562, from the Pierre Formation near Eagle Tail, Kansas (see Wil-
liston, 1898b, p. 180) collected by B. F. Mudge. Type specimen included “. . . a series
of dorsal, lumbar, and caudal vertebrae, with some bones of the limbs,” (Cope, 1875,
p- 153), only five vertebrae of the type have been located.

DistR1BuTION. Lower Pierre Formation, Kansas.

ADDITIONAL REFERENCES. Cope, 1875, pp. 153, 268, pl. 26 figs. 4-12. Merriam, 1894,
p. 31. Williston, 1898b, p. 180, pl. 45 figs. 3-5. Lane, 1947, p. 314.

Diacnosts. Chevrons free. Pectoral crest of humerus large, medially located.

Discussion. Cope’s Platecarpus crassartus is referred to Prognathodon, as was tenta-
tively suggested by Williston (1898b, p. 180), on the basis of the close similarity of its
humerus to that of P. overtont. These two humeri resemble each other and differ from
those belonging to Platecarpus in the rounded profile of the proximal end in lateral
profile, which in the latter genus is usually more angular; in the less constricted humeral
shaft, and in the lesser degree of expansion of the distal end. The radius of P. crassartus
differs markedly from that of Platecarpus in that it is wider than long and has a
much broader shaft. The internal trochanter is medial to the head of the femur in
position; in Platecarpus it is anteromedial.

RUSSELL: AMERICAN MOSASAURS 165

Prognathodon overtoni (Williston 1897)
(Text-figs. 89, 90)

(undescribed genus) overtoni Williston, 1895a, p. 169.
Brachysaurus overtoni Williston, 1897a, p. 96, pl. 8.
Ancylocentrum overtoni, Schmidt, 1927, p. 59.

Tyre. KU 950, from “... near the top of the Pierre deposits of the Cheyenne River
of South Dakota, and probably a hundred feet or more above (the level) of M. horridus
(synonym of M. missouriensis) . . .”” (Williston, 1897a, p. 95) collected by T. R. Overton.
Type specimen includes a fragmentary skull, 20 vertebrae, humeri and two paddle
bones.

DistrisuTion. Upper Pierre Formation, South Dakota.

ApDDITIONAL REFERENCES. Williston, 1897a, pl. 8; 1897d, pp. 178, 181; 1897e, p. 246;
1898b, pp. 88, 91, 107, 111, 127, 134, 148, 192, pl. 22 fig. 1, pl. 30, pl. 40 figs. 10-11
(undetermined bones), pl. 62 figs. 1-2; 1925, p. 273. Ballou, 1898, p. 217. Osburn, 1906, pl.
8 fig. 24. Gilmore, 1928, p. 87. Nopcsa, 1928, p. 177. Lane, 1947, p. 320. Kauffman and
Kesling, 1960, pp. 222, 230, 231, 234, 235, tables 3, 5, 6.

REFERRED SPECIMEN. SDSM no. 3393. Locality of referred specimen; from the Virgin
Creek Member, Upper Pierre Formation, six miles southwest of Cuny Table, on Mule
Creek, Shannon County, South Dakota (personal communication J. P. Gries, 1963).

Diacnosis. No tuberosity on anteromedian edge of quadrate shaft, suprastapedial
process not constricted dorsally. Chevrons fused to caudal centra. Pectoral crest of
humerus small, anteriorly located.

Discussion. Prognathodon overtoni may be distinguished from its European con-
temporary, P. solvayi, by the smooth enamel surfaces of its teeth (compare Williston,
1898b, pl. 30 fig. 1 with Dollo, 1889b, pl. 9 fig. 4) and the non-procumbant nature of
the anterior teeth in a beautifully preserved skull referred to P. overtoni in the collec-
tions of the South Dakota School of Mines. P. giganteus, from the same area and horizon
as P. solvayi, is defined by Dollo (1904, p. 213) as also possessing smoothly surfaced tooth
enamel and, in the absence of other described characters, could be identical to either
P. overtoni or P. rapax (see below).

Prognathodon rapax (Hay 1902)
(Text-fig. 91)

Macrosaurus laevis in part, Leidy, 1865a, p. 75.

Liodon validus, Cope, 1869-1870, pp. 201, 207, fig. 48(5).

Tylosaurus rapax Hay, 1902, p. 473.

Ancylocentrum hungerfordi Chaffee, 1939, p. 1, figs. 1 (1, 3, 6), fig. 2.

Type. AMNH 1490, “. . . from near Barnesboro, Gloucester Co., N. J.,” (Cope, 1869-
1870, p. 207). Type specimen includes quadrates of two individuals.

DistTRiBUTION. Navesink Formation or younger Cretaceous, New Jersey.

ADDITIONAL REFERENCES. Leidy, 1865a, ?pl. 9 figs. 8-9. Cope, 1869-1870, fig. 49 p. 205;
1870a, p. 272; 1875, p. 271, pl. 37 fig. 4. Miller, 1955, pp. 905, 909.

REFERRED SPECIMENS. AMNH no. 2205? YPM nos. 413, 1597. ANSP no. 8550. NJGS
no. 9827 (P. hungerfordz).

Diacnosis. Large tuberosity on anteromedian edge of guadrate shaft, suprastapedial
process strongly constricted dorsally.

Discussion. Lumbodorsal vertebrae of Prognathodon rapax closely resemble those of
Mosasaurus, but may be distinguished by the depressed form of their centra. Cervical
vertebrae of P. rapax have not yet been recognized, as the cervical in the type of
Ancylocentrum hungerfordi (= P. rapax) belongs to Thoracosaurus, a crocodile. Cope
(1869-1870, fig. 49, p. 205) figures a very broad radius from the late Cretaceous green-

PEABODY MUSEUM BULLETIN 23

166

WSS 401k ‘payon.sysuosax APYSIs) 2071920 uopoyjvuso1g JO [[Nys Jo

(He X ‘gegg
MITA [IILT ‘0G “8Y-7x9,7,

RUSSELL: AMERICAN MOSASAURS 167

sands of New Jersey that may belong to P. rapax. It has a broad longitudinal facet on
its posterodistal margin apparently for articulation with the ulna that is absent in all
other known mosasaurs.

Prognathodon rapax is easily separated from P. overtoni and P. solvayi by the char-
acters of the quadrate, outlined above, in which the latter two species resemble one

A B

Text-fig. 91. A. Lateral view of left quadrate. B. Posterior view of left quadrate. Prognathodon
rapax (reconstructed after the quadrates of NJGS 9827, x %4).

another. P. rapax may be close to P. giganteus but both species are much too incom-
pletely known to hazard speculation on interrelationships.

Genus PLESIOTYLOSAURUS Camp 1942

Plesiotylosaurus Camp, 1942, p. 18.

Generic Type. Plesiotylosaurus crassidens Camp 1942.

Diacnosis. Premaxilla with no rostrum anterior to premaxillary teeth. Thirteen
teeth in maxilla. Prefrontal forms large portion of posterolateral border of external nares;
supraorbital wing with triangular ala, contacts postorbitofrontal dorsally over orbit lateral
to external edge of frontal. Frontal not emarginate above orbits, median dorsal ridge
(?)absent. Parietal foramen small, closely embraced on either side by short tongues from
frontal. Margins of dorsal parietal field diverge posteriorly. Dorsal surface of postorbito-
frontal very well exposed. Squamosal wing to parietal large. At least seven pterygoid
teeth. Suprastapedial process distally fused to infrastapedial process on quadrate.
Tympanic ala thick, quadrate massive. Sixteen to seventeen teeth in dentary. Small
projection of dentary anterior to first dentary tooth. Mandibular teeth large and heavy.

Superior border of scapula strongly convex, posterior border very strongly emargi-
nated. Distal and proximal ends of humerus moderately expanded.

Discussion. The foregoing diagnosis was taken from descriptions and illustrations in
Camp (1942). Although the reviewer has not examined the California material, it is his
opinion that Plesiotylosaurus belongs in the Plioplatecarpinae close to Prognathodon
because of its bluntly terminated premaxilla and the large maxillary and pterygoid
teeth. It is distinguishable from Prognathodon in its long and slender snout, the greater
number of mandibular teeth and in its peculiar quadrate. The head of the humerus
is squarely terminated as in Platecarpus and unlike Prognathodon, but the radial
process is larger than in the former genus.

168 PEABODY MUSEUM BULLETIN 23
Plesiotylosaurus crassidens Camp 1942

Plesiotylosaurus crassidens Camp, 1942, p. 18, figs. 11-13, pl. 4.

Type. CIT no. *2759, 760 feet above base of Moreno Formation, Fresno County,
California (see Camp, 1942, p. 18, for detailed locality information). Type specimen
includes a skull and lower jaws (Camp, 1942, p. 18).

DisrrisuTion. Moreno Formation, California.

ADDITIONAL REFERENCES. Welles, 1943, p. 128, fig. 2. McDowell and Bogert, 1954, p.
132. Romer, 1956, p. 562. Anderson, 1958, p. 71.

REFERRED SPECIMEN. CIT no. 2753 (fide Camp, 1942, p. 19).

?PSUBFAMILY PLIOPLATECARPINAE incertae sedis:
Genus HALISAURUS Marsh 1869

Halisaurus Marsh, 1869, p. 395.
Baptosaurus Marsh, 1870, p. 3.

Generic Type. Halisaurus platyspondylus Marsh.

ADDITIONAL REFERENCES. Cope, 1869-1870, p. 208; 1874, p. 31; 1875, p. 272. Merriam,
1894, pp. 36, 39. Williston, 1897c, p- 107; 1897d, pp. 179, 182; 1898b, p. 207. Gilmore,
1928, p. 87. McDowell and Bogert, 1954, p. 132. Romer, 1956, p. 562.

Diacnosis. Articulating surfaces of cervical and anterior dorsal vertebral centra nearly
twice as wide as deep, sub-rectangular in outline. Synapophyses located in center of
lateral surface of cervical centra, occupies somewhat more posterior position in anterior
thoracics. Ventral border of anteroventral extension of synapophysis weak and _hori-
zontal in anterior cervicals, becomes much enlarged in posterior cervicals and anterior
thoracics, extending far below flattened undersurface of centrum. Anterior zygopophysis
of cervical and anterior thoracics connected by gently rounded, posteriorly descending
crest to synapophysis. No zygosphene or zygantrum. Hypapophyseal peduncle located
posteriorly on ventral surface of cervical centra, articulation for hypapophysis flat and
lenticular, slightly inclined posteriorly.

Discussion. In spite of the great span of time separating the two known species of
Halisaurus, their vertebrae are so similar and so characteristic that there would be no
justification for relegating them to separate genera (see also Merriam, 1894, p. 37). The
large, Platecarpus-like suprastapedial process of H. onchognathus supports the reference
of this genus to the Plioplatecarpinae.

It has been stated (Dollo, 1889b, p. 275; Williston, 1898b, p- 207; etc.) that the
hypapophyses of Halisaurus are fused to the cervical centra. The cervical vertebra of
the type of H. platyspondylus has a long, laterally compressed peduncle that lacks any
articulating facet distally for the hypapophysis. In a larger specimen, surely referable to
H. platyspondylus because of its greatly depressed centra (YPM 412), the peduncle bears
a well-developed facet, as does a cervical in the type of H. onchognathus. The cervical
vertebra of Marsh’s type must have belonged to the posterior cervical or anterior dorsal
region, where the peduncles lose the facet for articulation with the hypapophysis.

Halisaurus onchognathus (Merriam 1894)
Baptosaurus onchognathus Merriam, 1894, p. 37, pl. 4 figs. 9-10.
Type. Formerly at Bayerische Staatssammlung fiir Paldéontologie but probably de-

stroyed during the Second World War (personal communication, R. Dehm 1963), from
“. . . Niobraraschichten der oberen Kreide von Logan County, Kansas . . .” (Merriam,

RUSSELL: AMERICAN MOSASAURS 169

1894, p. 3) collected by C. H. Sternberg. Type specimen includes a surangular, articular,
prefrontal, quadrate and eight vertebrae.

ADDITIONAL REFERENCES. Williston, 1897d, p. 179; 1897e, p. 244; 1898b, p- 208, pl.
25 fig. 6, pl. 41 fig. 3. Dollo, 1924, p. 211.

Discussion. No additional specimens of this highly distinctive form have been seen by
the author in the collections of American museums. A possible specific distinction be-
tween H. platyspondylus and H. onchognathus may be that in the former the hypapophy-
seal peduncle is somewhat dilated distally; in the latter it is tapered (Merriam, 1894,
pl. 4 fig. 9). H. onchognathus is easily distinguished from Ectenosaurus clidastoides by
the fan-shaped retroarticular process of the articular. Merriam (1894, pp. 36-37) de-
scribes the following cranial material with the type of H. onchognathus:

“Das einzige Kieferstiick . . . stimmt bis auf das abweichende Articulare mit dem
von Platecarpus oder Tylosaurus iiberein, interscheidet sich jedoch von diesen dadaurch,
dass der Oberrand des Articulare unmittelbar hinter der Gelenkpfanne, anstatt sich
um etwa 40° nach unten zu biegen, vertical harauf in einem hohen Fortsatz ausgezogen
ist. Dadurch erhalt das Hinterende des Unterkiefers ein hakenartiges Aussehen. Auch
an der unteren hintern Ecke, wo bei Platecarpus und Tylosaurus eine langliche Ver-
dickung verhanden ist, verdickt sich der Rand pl6tzlich um mehr als zwei Mal der Dicke
hinter dieser Stelle und wird dann plotzlich unter dem Hinterende der Gelenkpfanne
diinner . .. Ein stiick des Hinterendes des Praefrontale hat viel Aehnlichkeit mit dem
von Tylosaurus, nur ist auf der Aussenseite desselben keine tiefe Rinne zur Aufnahme
des spitzen Vorderendes des Postfrontoorbitale vorhanden. Es liegt auch ein sehr zer-
quetschtes Quadratum vor, das einem Supracolumellar-Fortsatz, der dem von Platecarpus
sehr ahnlich sieht, besitzt. Der Fliigel des Quadratums ist wie bei Platecarpus und
Tylosaurus in der Mitte sehr diinn und durchbrochen.” The intercentral articulations
of this specimen become more circular posteriorly (ibid., p. 36).

Halisaurus platyspondylus Marsh 1869

Halisaurus platyspondylus Marsh, 1869, p. 395.
Baptosaurus platyspondylus, Cope, 1869-1870, p. 209.

Tyre. YPM 444, from “. .. Hornerstown, New Jersey . . .” (Marsh, 1869, p. 396) col-
lected by J. G. Meirs. Type specimen includes an angular and two vertebrae.

DisTRIBUTION. Navesink Formation or younger Cretaceous, New Jersey.

ADDITIONAL REFERENCES. Cope, 1875, p. 272. Merriam, 1894, p. 37. Williston, 1898b,
p. 207. Miller, 1955, p. 909.

REFERRED SPECIMEN. YPM no. 412.

Discussion. This species was apparently described verbally as Macrosaurus platy-
spondylus by Marsh during the Salem meetings of the American Association for the
Advancement of Science in August, 1869. However a description was not published until
November of that year, when it was made the genotype of Halisaurus (see also Hay,
1902, p. 468).

There is a fragment of an angular preserved with the type vertebrae. Its articular
surface resembles that of the angular of Platecarpus but is more symmetrically heart-
shaped in anterior aspect. In lateral profile the angular appears somewhat inflated, with
a convex anteroventral outline that is gently continuous with the anteroexternal margin
of the articular surface, similar to its condition in Clidastes.

FOREIGN PLIOPLATECARPINAE

Genus DOLLOSAURUS Iakovlev 1901
Dollosaurus Iakovlev, 1901, p. 518.

Generic Type. Dollosaurus lutugini Iakovlev 1901.

170 PEABODY MUSEUM BULLETIN 23

Dracnosis. Premaxilla with no rostrum anterior to premaxillary teeth. Twelve(?)
teeth in maxilla. Seven teeth in pterygoid. Thirteen teeth in dentary. Dentary termi-
nates abruptly in front of first dentary tooth. Mandibular teeth bicarinate, slightly
faceted, and slightly posterolingually recurved.

Articulating surfaces of cervical and anterior dorsal vertebral centra wider than deep,
smoothly elliptical in outline. Synapophysis occupies anterodorsal portion of lateral sur-
face of anterior dorsal vertebra. Zygosphene-zygantrum strongly developed. Chevrons
fused to caudal centra.

Vertebral formula: At least 25 dorsals (see Iakovlev, 1901, pp. 517-518; Dollo, 1924,
pp- 193-197; Tsaregradskii, 1935, pp. 53-54).

Discusston. The only known specimen of Dollosaurus is a partial skull and skeleton
from the upper Campanian of the Donetz Basin, southern European Russia (Iakovlev,
1901, p. 516). The animal is evidently a plioplatecarpine in spite of its co-ossified
chevrons and zygosphenes, as Dollo (1924, p. 197) suspected. The dorsally concave
alveolar margin of the dentary (Iakovlev, 1901, fig. 2), the “mosasaurine” shape of the
coronoid (Tsaregradskii, 1935, p. 53), the extension of the splenial anteriorly into the
symphysial region (Dollo, 1924, p. 196; compare with Camp, 1942, p. 19) and the reduc-
tion in number and increase in size of the anterior pterygoid teeth (Iakovlev, 1901,
pl. 5 fig. 1) all suggest affinities with Prognathodon and Plesiotylosaurus. However the
great enlargement of the first two dentary teeth and large zygosphenes are unique fea-
tures of the genus (Dollo, 1924, p. 196).

SupFAMiILy TYLOSAURINAE (Williston 1895) Williston 1897

Tylosauridae Marsh, 1876, p. 59, nomen nudum.
“mosasaurines megarhynques” Dollo, 1890, p- 163.
Tylosauridae Williston, 1895, p. 169.
Tylosaurinae Williston, 1897d, p. 177.

Diacnosis. Large rostrum present anterior to premaxillary teeth. Twelve or more
teeth in dentary and maxilla. Cranial nerves X, XI and XII leave lateral wall of opisthotic
through single foramen. No canal in basioccipital or basisphenoid for basilar artery.
Suprastapedial process of quadrate moderately large, distally pointed. Dorsal edge of
surangular rounded and longitudinally horizontal.

Twenty-nine or more presacral vertebrae present. Length of presacral series less
than that of postsacral series in Tylosaurus, neural spines of posterior caudal vertebrae
at most only slightly elongated, do not form an appreciable fin. Haemal arches unfused
to caudal centra. Appendicular elements lack smoothly finished articular surfaces.

Discussion. There is little justification for separating the Plioplatecarpinae and
Tylosaurinae on the basis of the postcranial skeleton. The large premaxillary rostrum,
widely expanded otosphenoidal crest and lack of basioccipital canals at once distin-
guish the premaxillae, prootics and basioccipital-basisphenoids from those of members
of the Plioplatecarpinae, but otherwise the cranial skeleton is very similar. The two
subfamilies must be much more closely related to each other than either is to the
Mosasaurinae.

Genus TYLOSAURUS Marsh 1872
(Text-figs. 45, 58)

Liodon in part, Cope, 1869-1870, p- 200.

Rhinosaurus Marsh, 1872b, p. 17 (preoccupied, Fischer and Waldheim 1847).
Rhamphosaurus Cope, 1872e, p- 141 (preoccupied, Fitzinger 1843).

Tylosaurus Marsh, 1872c, p. 47.

Generic Type. Tylosaurus proriger (Cope).

RUSSELL: AMERICAN MOSASAURS 171

ADDITIONAL REFERENCES. Cope, 1871f, p. 401; 1872d, p. 273; 1872f, p. 331; 1874, p. 36;
1875, p. 160; 1878, p. 301. Marsh, 1872a, p. 291; 1872b, pl. 13; 1872d, pp. 496, 497; 1877,
p- 346; 1897, fig. 23. Leidy, 1873, p. 271, pl. 36 figs. 1-3. Owen, 1879, pl. 8 figs. 13-14.
Lydekker, 1888, p. 264. Smith-Woodward, 1889, p. 280. Hoffmann, 1890, p. 1321. Willis-
ton, 1891, p. 345; 1897b, fig. 1, pls. 9-12; 1897d, pp. 177, 180; 1897e, p. 244; 1898b, p.
171, pl. 40 figs. 1-9, pl. 41 figs. 1, 4-5, pl. 42 figs. 1-2, pl. 46, pls. 48-50, pl. 54 fig. 1, pl.
55, pl. 59 figs. 1-8, pls. 65-70; 1914, p. 166, figs. 70, 79; 1925 fig. 148. Williston and Case,
1892, p. 15. Merriam, 1894, pp. 14, 38, 39, pl. 4 figs. 1, 3-6. Whitfield, 1900, pl. 5 fig. 3.
Osburn, 1906, pl. 8 fig. 25. Merril, 1907, p. 80. Huene, 1910, figs. 1-4, pls. 41-42. Camp,
1923, pp. 323, 324; 1942, pp. 28, 32, fig. 23. Gilmore, 1928, p. 87. Nopcsa, 1928, p. 177.
McDowell and Bogert, 1954, pp. 106, 152, 138, fig. 10. Romer, 1956, p. 562, fig. 154c.

Dracnosis. 12 to 13 teeth in maxilla. Supraorbital wing of prefrontal covered dorsally
by frontal. Prefrontal excluded from posterolateral margin of external nares by wing
of maxilla. Frontal not emarginate above orbits, median dorsal ridge prominent to
nearly absent. Parietal foramen moderately large, located entirely within parietal.
Margins of dorsal parietal surface parallel one another and cranial midline to posterior
base of diverging suspensorial rami, forming rectangular field medially on parietal.
Ventral process of postorbitofrontal to jugal meets poorly exposed dorsal surface of
postorbitofrontal at right angle. Ventroposterior process on jugal present. Squamosal
wing to parietal moderately developed. Large otosphenoidal crest on prootic covers
exits for VII and IX laterally. Ten to 11 teeth in pterygoid. Tympanic ala of quadrate
thin to rather thick, stapedial pit rectangular in form. Thirteen teeth in dentary. Broad
projection of dentary anterior to first dentary tooth. Angular widely separated medially
from coronoid. Retroarticular process of articular dorsally rounded, ventrally straight
in outline.

Vertebral formula: 29-30 presacral vertebrae, 6-7 pygals, 33-34 caudals with chevrons
and transverse processes, 56-78 terminal caudals (see Williston, 1898b, p. 143; 1910, p.
539; Huene, 1910; Osborn, 1899a; and YPM 1250 and 24908, USNM 8898).

Articulating surfaces of cervical and anterior dorsal vertebrae nearly circular. Syna-
pophysis located in center of lateral surface of cervical centra, occupies anterodorsal
portion of lateral surface of dorsal centra. Ventral border of anteroventral extension of
synapophysis not strongly developed on cervicals and anterior dorsals, does not reach
level of undersurface of centrum. Anterior zygopophysis of cervicals and dorsals con-
nected by sharp, ram-rod straight crest posteroventrally to synapophysis. Zygosphene-
zygantrum rudimentary. Anterior base of atlas neural arch arises directly above condylar
facet, atlas synapophysis small and flattened or rudimentary. Hypapophyseal peduncle
located posteriorly on ventral surface of cervical centra, articulation for hypapophysis
teardrop-shaped, slightly inclined posteriorly. Five hypapophysis-bearing cervicals, two
or three more with rudimentary peduncles. Transverse process of pygal vertebrae rela-
tively short. Neural spines of caudal vertebrae longest and vertical on postsacrals 38-40.

Scapula much smaller than coracoid. Glenoid articulating surfaces concave on both
scapula and coracoid, nearly form a smoothly continuous surface. Superior border of
scapula strongly convex, posterior border slightly emarginated posteromedially. Coracoid
does not expand medially to point behind glenoid articulation. Distal and proximal ends
of slender humerus only slightly expanded, radial process absent. Facets for articulation
with other elements and sites of muscle attachment not well differentiated. Radius very
elongate, proximal end very slightly expanded. Shaft of radius narrow. Distal end
slightly expanded, has only trace of anterodistal flange.

Ulnare and fourth carpal present, lack articulating surfaces. Metacarpal one equal to
metacarpal two in length, has very slight anterodistal flange. Proximal ends of meta-
carpals, especially of two and three, not greatly expanded. Phalangeal formula of manus
variable, in one (see Williston, 1898b, pl. 48) it is 5-7-9-10-11; in another (see Osborn,
1899a, p. 183) it is 5-8-8-9-9.

Acetabular surfaces of pelvic elements convex, do not form solid, smoothly surfaced
bowl. Obturator foramen located near anterior margin of proximal end of pubis,
dorsoanterior process rudimentary. Ischiadic tubercle widely separated from acetabulum.

172 PEABODY MUSEUM BULLETIN 23

Text-fig. 92. Dorsal view of skull of Tylosaurus proriger (reconstructed after AMNH 4909, x
ca. 14).

Shaft of ischium broad, well expanded medially at symphysis. Distal end of femur much
more expanded than proximal; internal trochanter of average proportions and located
medially from head. No distinct facets on distal end. Cartilage-capped area of anterior
flange of tibia extends two-thirds way up shaft giving bone characteristic broad rectan-
gular outline, similar area of posterodistal flange of fibula extends one-third way up shaft.

Astragulus nearly circular, dorsal notch between facets for tibia and fibula rudimen-
tary or absent. Calcaneum unossified, fourth tarsal small. Metatarsal one expanded

RUSSELL: AMERICAN MOSASAURS 173

proximally. Metatarsal five hook shaped, with strongly concave medial border and
shallowly concave or straight lateral border. Phalangeal formula of pes estimated at
5-8-8-8-6 (Osborn, 1899a, p. 183).

Discussion. The first described species subsequently to be included in the genus
Tylosaurus was Macrosaurus proriger Cope. This species was later referred by Cope
(1869-1870, p. 201) to Liodon Owen, and the latter genus was redefined by him (ibid., p.
200) to include some very incompletely known North American material. Cope (1871b,
p- 576) allied his L. dyspelor with Amphorosteus brumbyi Gibbes, and his L. proriger
with Nectoportheus (see Cope, 1869-1870, p. 208). The single type vertebrae of Macro-
saurus and Nectoportheus belong to the posterior dorsal region of Mosasaurus, and the
type vertebrae of Amphorosteus are indeterminate (see p. 176). These names are thus
not available to replace Tylosaurus.

Merriam (1894, p. 14) demonstrated that the name Liodon should not be applied
to the Niobrara species discussed below, because in Liodon the enamel surfaces of the
teeth are smooth, resembling those of Clidastes; in Tylosaurus the internal face is
striated. It is appropriate to add that the teeth of Liodon anceps (Owen, 1840-1845, pl.
72 fig. 1), the type species of the genus, are symmetrically bicarinate although those of
Tylosaurus are relatively inflated with a flattened labial face and a larger, more rounded
lingual face, and lack strong anterior and posterior carinae. These distinctions are of a
generic order.

Marsh (1872b, p. 17) recognized for the first time the generic identity of Tylosaurus,
clearly separated it from Platecarpus and proposed for it the name Rhinosaurus (pre-
occupied) supported by an excellant generic diagnosis based for the most part on cranial
characters. Cope (1872e, p. 141) adhered to his opinion that “Rhinosawrus” was actually
synonymous with the English Liodon, but proposed the name Rhamphosaurus (pre-
occupied), should the Kansas species prove to belong to a distinct genus. Marsh (1872c,
p. 47) replaced Rhamphosaurus with Tylosaurus, which became accepted by nearly all
workers, with the notable exception of Cope.

Williston (1897d, p. 182) suggested that Lesticodus might be a senior synonym of
Tylosaurus, but the type species of this genus is based on a pterygoid tooth of a
Mosasaurus.

Tylosaurus proriger (Cope 1869)
(Frontispiece; Text-figs. 2C, 5A, 21, 24C, 27, 48A, 55, 63, 92, 93B, 94A;
Plate I, fig. 3; Plate II, fig. 1)

Macrosaurus proriger Cope, 1869d, p. 123.
Macrosaurus pririger Cope, 1869g, p. 122.
Liodon proriger, Cope, 1869-1870, p. 201.
Rhinosaurus proriger, Marsh, 1872b, p. 19.
Rhinosaurus micromus Marsh, 1872b, p. 17.
Tylosaurus (proriger), Marsh, 1872c, p. 47.

Type. Formerly at MCZ, not located, from “. . . vicinity of Monument Rocks, the old
overland stage station, (Kansas) . . .” (Williston, 1898b, p. 174) collected by Messrs.
Connyngham and Minor and sent to L. Agassiz. Type specimen included the anterior
portion of a skull, cranial fragments, and thirteen vertebrae (Cope, 1869-1870, p. 202).

DistTR1BuTION. Smoky Hill Member, Niobrara Formation, Kansas and Colorado; Lower
Pierre Formation, Kansas and South Dakota; Telegraph Creek Formation, Montana.

ADDITIONAL REFERENCES. Cope, 1869-1870, p- 201, pl. 12 figs. 22-24; 1871f, p. 401;
1872a, p. 297; 1872c pp. 173-174; 1872d, pp. 279-280; 1872e, p. 141; 1872f, p. 333; 1874,
pp. 37-38; 1875, pp. 161, 167, 271, fig. 7, pls. 28-30, pl. 31 figs. 1-4, pls. 32-33, pl. 36 figs.
1-2, pl. 37 figs. 5-6; 1879a, p. 132. Marsh, 1872d, p. 497; 1880, p. 85, fig. 1. Leidy, 1873, pp.
274, 343-344, pl. 35 figs. 12-13, pl. 36 fig. 3. Snow, 1878, pp. 54, 57. Merriam, 1894, pp.
23-24, pl. 1 fig. 3, pl. 2, pl. 3 figs. 1-2, 8, pl. 4 fig. 7. Williston, 1895, pl. 17 fig. 2; 1897b,

174 PEABODY MUSEUM BULLETIN 23

p. 102; 1897c, p. 110, pl. 13 fig. 3; 1897d, pl. 20 fig. 1; 1898a, p. 28; 1898b, pp. 102-134,
173, 175, pls. 16-18, pl. 19 fig. 1, pl. 31 figs. 1-3, pl. 60 figs. 1-2, pl. 61 figs. 1-2, pl. 72 fig. 3;
1902, p. 253; 1914, fig. 72c; 1925, p. 273, fig. 54. Osborn, 1899a, p. 169, figs. 1, 7, 8-9, 11-14,
pls. 21-23; 1899b, p. 912. Fiirbringer, 1900, p. 616. Merril, 1907, p. 80. Sternberg, 1909, fig.
8; 1917, pp. 13, 162, fig. 5. Pompeckj, 1910, p. 137. Stromer, 1910, pl. 2. Dreverman, 1914,
p- 43, fig. 7. Lambe, 1914, p. 402. Fejérvary, 1918, fig. 25. Wiman, 1920, pl. 4 figs. 6a-b.
Gilmore, 1921, p. 273. Lane, 1947, pp. 313-314, fig. 5. Romer, 1956, fig. 220. Kauffman
and Kesling, 1960, fig. 6d.
REFERRED MATERIAL. AMNH nos. **221, 1493, 1529, 1535, 1555, 1560, 1585, 1592, 2160,
**4909, **7220. YPM nos. *1268 (T. micromus), 1288, 1302, 1305, *3873, 3977, 3978,
3981, 3984, 3987, 3989, 3990, 3993, *3999, #002 and about twenty-four unnumbered
specimens. ANSP no. 5. USNM nos. 3776, 3794, 3798, 3897, **6086, *8898, *17909. MCZ
no. **1030. KU nos. 1020, **1023, **1032, *1033, 1084, 1115, 1135, 1194, *5033. FHM
no. **4 CNHM nos. UR820, **P15144, UR902. NMC no. 8162. PU no. 12000.
Diacnosis. Anterior tip of premaxilla rectangular in lateral profile, in form of ellipse
with vertical long axis in cross-section. Premaxillo-maxillary suture terminates posteriorly
at point dorsal to or somewhat behind fourth maxillary tooth. Parietal foramen close
to frontal suture. Lateral margin of prefrontal deeply grooved for anterior process of

Text-fig. 93. Dorsal view of parietal. A. Tylosaurus nepaeolicus (restored after AMNH 124 and
YPM 3980, x ca. 4). B. Tylosaurus proriger (slightly reconstructed, after YPM 3990, x
ca, 14).

postorbitofrontal. Posterior surface of parietal moderately to extensively invaded
medially by insertional area for cervical epaxial musculature. Infrastapedial process
on quadrate small. Lateral crest of tympanic ala descends posteriorly nearly to mandibular

RUSSELL: AMERICAN MOSASAURS 175

Text-fig. 94. Lateral view of left quadrate. A. Tylosaurus proriger (reconstructed after YPM 3990
and AMNH 4909, x 54). B. Tylosaurus nepaeolicus (YPM 3992, x 54).

articulating surface of quadrate, then ascends to terminate lateral to and slightly below
infrastapedial process.

Discussion. The type of “Macrosaurus” proriger unfortunately has not been found.
Thus far it has been possible to distinguish only two species of Tylosaurus from the
Niobrara Chalk, one of which may be characterized by a premaxillo-maxillary suture
that extends at least as far back as the fourth maxillary tooth. From Cope’s (1869-1870,
pl. 12 figs. 22-24) figures this would seem to be its condition in the type of T. proriger,
as it surely is in the specimen from Butte Creek that Cope referred to this species
(AMNH 1493; Cope, 1872c, p. 175). One of the points by which Cope (1874, p. 38)
separated the type of T. nepaeolicus from T. proriger was the relatively advanced posi-
tion of its external nares, thus intimating that in the latter species the external nares
were more posteriorly situated and that the premaxillo-maxillary suture was also extended
further posteriorly. It seems safe therefore to assume that the correct name has been
applied to the species diagnosed above.

Tylosaurus nepaeolicus (Cope 1874)
(Text-figs. 15, 93A, 94B, 95)

Liodon nepaeolicus Cope, 1874, p. 37.
Tylosaurus nepaeolicus, Merriam, 1894, p. 24.

Typr. AMNH 1565, from‘. . . the gray shale of the Niobrara Cretaceous, a half
mile south of the Solomon River, Kansas,” (Cope, 1874, p. 38) collected by B. F. Mudge.
Type specimen includes a partial skull and one vertebra.

DistriBuTION. Smoky Hill Member, Niobrara Formation, Kansas.

ADDITIONAL REFERENCES. Cope, 1875, p. 168, pl. 35 figs. 11-15. Owen, 1879, p. 55.
Merriam, 1894, p. 24, pl. 4 fig. 7. Williston, 1898b, p. 176; 1910, p. 539, figs. 5-10. Stern-
berg, 1908, p. 113, fig. 1. Huene, 1910, p. 297, figs. 5-10, pl. 1, pl. 2 figs. 2-3; 1919, p. 183,
pl. 8 fig. 1; 1953, fig. 1. Pompeckj, 1910, pp. 134, 137. Abel, 1922, figs. 260, 267. Renger,
1935, p. 560, fig. 5?

REFERRED SPECIMENS. AMNH nos. **124, 131, **134, 1524, 1561, 2167, 2319. YPM

176 PEABODY MUSEUM BULLETIN 23

nos. 3969, 3970, *3974, 3976, 3979, 3980, 3992, *4000 and about twenty unnumbered
specimens. USNM no. 3894. MCZ nos. 1592, 1604, 1626. FHM no. *1654.

Diacnosts. Anterior tip of premaxilla rounded in lateral profile, nearly circular in
cross-section. Premaxillo-maxillary suture terminates posteriorly above midpoint between
third and fourth maxillary tooth. Parietal foramen widely separated from frontal
suture. Lateral margin of prefrontal rounded, abuts against postorbitofrontal poste-
riorly. Posterior dorsal surface of parietal slightly invaded medially by insertional area for
cervical epaxial musculature. Infrastapedial process on quadrate small. Lateral crest of
tympanic ala ends posteriorly near mandibular articulating surface of quadrate.

Discussion. Tylosaurus nepaeolicus seems generally to be a smaller species than T.
proriger. It does not occur stratigraphically above the Niobrara Chalk, and incomplete
locality information suggests that it may not even occur near the top of this formation.

FOREIGN TYLOSAURINAE

Genus HAINOSAURUS Dollo 1885
Hainosaurus Dollo, 1885, p. 288.

Generic Type. Hainosaurus bernardi Dollo, 1885, p. 288.

Diacnosis. 72 teeth in skull (Tylosaurus has 70-74). Ventroposterior process on jugal
present. Stapedial pit in form of longitudinally compressed ellipse. Broad projection of
dentary anterior to first dentary tooth. Retroarticular process of articular dorsally
rounded, ventrally straight in outline.

Vertebral formula: 52 vertebrae in cervical-dorsal-pygal region (35 are present in
Tylosaurus), at least 33 diapophysis-bearing caudals.

Anterior zygopophysis of cervicals and anterior dorsals not ramrod straight, arise
high on neural arch and project horizontally anteriorly. No zygosphene-zygantrum.
Neural spines of caudal centra decrease uniformly in length from base to end of
tail.

Scapula much smaller than coracoid. Superior border of scapula strongly convex,
posterior border strongly emarginated posteromedially. Distal and proximal ends of
slender humerus moderately expanded, radial process small. Facets for articulation with
other elements and sites of muscle attachment not well differentiated. Radius very
elongate, proximal end slightly expanded. Shaft of radius narrow. Distal end somewhat
expanded, has small anterodistal flange.

Carpus poorly ossified, as in Tylosaurus. Femur longer than humerus (Dollo, 1885;
1889b, pl. 9 fig. 3, pl. 10 figs. 1-3; 1917, p. 19).

Discussion. Hainosaurus has been identified only from the late Cretaceous of Belgium.
Dollo (1885, p. 288) states that the suprastapedial process of the quadrate is absent. It
would seem from his figures (1889b, pls. 9-10) that the type of H. bernardi, although
nearly complete, is badly preserved, and the quadrate may have originally had a
suprastapedial process as large as in Tylosaurus. As it now stands the only good
character separating Hainosaurus from Tylosaurus is the greater number of vertebrae
between the cranium and the chevron bearing caudals in the former genus. One other
species is known, H. lonzeensis (Dollo, 1904a, p. 213; 1909, p. 102), based on a pre
maxilla and vertebra.

MOSASAURS OF UNCERTAIN TAXONOMIC POSITION

Amphorosteus brumbyi Gibbes 1851 nomen vanum.

Amphorosteus brambyi Gibbes, 1850, p. 77 nomen nudum.

177

AMERICAN MOSASAURS

RUSSELL:

521 HNIWY Jae poinmsuoras) snayoavdau sninvsojk{yz Jo [NYS jo

(85% X ‘FET
MIA [e19IeT GG “3-29,

178 PEABODY MUSEUM BULLETIN 23

Amphorosteus brumbyi Gibbes, 1851, p. 11.
Mosasaurus brumbyi, Cope, 1869b, pp. 260, 264.

Type. Specimen not located.

ADDITIONAL REFERENCES. Leidy, 1865a, pp. 52, 118. Cope, 1869-1870, pp. 186, 198;
1871b, p. 576; 1871£, p. 411; 1875, p. 270. Lydekker, 1888, p. 267. Zangerl, 1948, p. 15.

Discussion. Amphorosteus brumbyi is based on two vertebrae from the Cretaceous
of Alabama, figured by Gibbes (1851, pl. 5 figs. 10-16). They are indeterminate.

Clidastes? congrops (Cope 1870) nomen vanum

Liodon congrops Cope, 1869-1870, p. 206.
Tylosaurus congrops, Hay, 1902, p. 472.

Type. Specimen not located, originally from the Selma Chalk of Alabama (Cope,
1869-1870, p. 207) collected by Dr. E. R. Showalter.

ADDITIONAL REFERENCES. Cope, 1870a, p. 271; 1875, p. 272. Merriam, 1894, p. 25.
Zangerl, 1948, p. 15.

Discussion. The type specimen of Clidastes? congrops is a single cervical vertebra,
which has apparently been lost. It would appear from the original description that the
vertebra belongs to Clidastes, as Cope (1875, p. 272) later suggested.

Elliptonodon compressus Emmons 1858 nomen dubium

Elliptonodon compressus Emmons, 1858, p. 222, figs. 41-42.

Type. In Museum of Williams College, from “. . . the miocene near the Cape Fear
River, in Bladen county (North Carolina),’’ (Emmons, 1858, p. 223).

ADDITIONAL REFERENCES. Emmons, 1860, p. 208, fig. 180(2), (5). Cope, 1869b, p. 263;
1869-1870, p. 186. Gilmore, 1928, p. 87. McDowell and Bogert, 1954, p. 132.

Discussion. The type and only specimen is a symmetrically bicarinate, somewhat in-
flated and gently posteriorly recurved tooth that looks mosasaurian and resembles most
closely teeth of Prognathodon. It differs from any species of this genus however in that
the enamel surface of the tooth is strongly vertically striated. The name Elliptonodon
compressus must be regarded as a nomen dubiwm due to the incompleteness of the type
and the uncertainty regarding the original horizon of the reworked tooth.

Gilmore (1926, p. 192, pl. 72 figs. 1-2) described and figured a large fragment of a
mosasaur dentary from the Coon Creek Formation of McNairy County, Tennessee, men-
tioning that the teeth were similar to that of Elliptonodon compressus. The dentary
probably belongs to Prognathodon, and its dorsally concave alveolar border resembles that
of P. solvayi more closely than the more nearly straight border of P. overtont.

Holcodus acutidens Gibbes 1851 nomen vanum

Holcodus columbiensis Gibbes, 1850, p. 77 nomen nudum.
Holcodus acutidens Gibbes, 1851, p. 9, pl. 3 figs. 6-9.
Mosasaurus acutidens, Marsh, 1872b, p. 455.

Lectotyrr. ANSP 8594, from the “Cretaceous of Alabama,” (Gibbes, 1851, p. 9).

AppITIONAL REFERENCES. Leidy, 1865a, pp. 32, 71, 118, 130, fig. 33, pl. 10 fig. 17.
Cope, 1869b, p. 265; 1869-1870, p. 210; 1872d, p. 269; 1875, pp. 141, 270. Marsh, 1869,
p. 394. Merriam, 1894, p. 30. Williston, 1897d, p. 185; 1898b, p. 179. Gilmore, 1928, p.
87. McDowell and Bogert, 1954, p. 132.

RUSSELL: AMERICAN MOSASAURS 179

Discussion. Holcodus acutidens was originally proposed by Gibbes for three teeth,
one from the Cretaceous of Alabama, a second from the Cretaceous of New Jersey, and
a third from the Eocene of Orangeburg, South Carolina. The third tooth has not been
located but if it was indeed from the Eocene it could not have been mosasaurian. The
second tooth was identified as crocodilian by Leidy (1865a, p. 32), leaving the tooth
from Alabama which Williston (1897d, p. 184) treated as a lectotype. This tooth is in-
distinguishable from those of Platecarpus, Ectenosaurus, or Plioplatecarpus and certainly
is plioplatecarpine. Because of the inadequacy of the type material Holcodus acutidens
must be considered a nomen vanum.

Mosasaurus caroliniensis Gibbes 1851 nomen vanum

Mosasaurus caroliniensis Gibbes, 1850, p. 77 nomen nudum,
Mosasaurus caroliniensis Gibbes, 1851, p. 8, pl. 2 figs. 1-3.

Type. Specimen not located.

ADDITIONAL REFERENCES. Leidy, 1860, p. 92; 1865a, p. 32. Cope, 1869b, p. 262; 1869-
1870, p. 193; 1875, p. 269.

Discussion. The type is a mandibular fragment probably of a Mosasaurus. It is from
the Pliocene of Darlington, South Carolina, and was derived from the underlying
Cretaceous (Gibbes, 1851, p. 7).

Mosasaurus couperi Gibbes 1851 nomen vanum

Mosasaurus couperi Gibbes, 1850, p. 77 nomen nudum.
Mosasaurus coupert Gibbes, 1851, p. 7, pl. 2 figs. 4-5.
Mosasaurus cow pert, Cope, 1869b, p. 262.

Type. Specimen not located, originally from “. . . the Cretaceous deposits of the
banks of the Chattahoochie, Georgia, discovered by J. Hamilton Couper, Esq. . .”” (Gibbes,
1851, p. 7).

ADDITIONAL REFERENCES. Leidy, 1860, p. 92; 1865a, p. 32. Cope, 1869-1870, p. 193;
1875, p. 269.

Discussion. The type teeth figured by Gibbes (1851, pl. 2 figs. 4-5) may belong to
the pterygoid dentition of Mosasaurus.

Mosasaurus? crassidens Marsh 1870 nomen dubium
Mosasaurus crassidens Marsh, 1870, p. 2.

Type. At Museum of Williams College, from ‘‘.. . the Cretaceous of North Carolina,”
(Marsh, 1870, p. 2) collected by E. Emmons.

ADDITIONAL REFERENCES. Cope, 1869-1870, p. 198; 1875, p. 270.

Discussion.. The type specimen is a fragment of a right maxilla with a well-marked
dorsal narial emargination and large non-prismate teeth. It could belong to either
Mosasaurus or, more likely, to Prognathodon and new material from the North Carolina
Cretaceous may make it possible to recognize this species.

Mosasaurus impar (Leidy 1856) nomen dubium
Drepanodon impar Leidy, 1856, p. 255.

Lesticodus impar, Leidy, 1861, p. 10.
Mosasaurus impar, Leidy, 1865a, p. 33.

180 PEABODY MUSEUM BULLETIN 23

Tyre. At Museum of Williams College, from “. . . Elizabethtown, Cape Fear, North
Carolina,” (Leidy, 1858, p. 224) collected by E. Emmons.

ADDITIONAL REFERENCES. Leidy, 1857, p. 271; 1858, p. 224, figs. 45, 46. Cope, 1869b,
p- 263; 1869-1870, pp. 185, 205; 1875, p. 270. Merriam, 1894, p. 25.

Discussion. Mosasaurus impar is based on a pterygoid tooth, probably of Mosasaurus.
Pterygoid material of Atlantic Coast mosasaurs is too poorly known to assign this tooth
to any described species.

Mosasaurus laevis (Owen 1849) nomen vanum

Macrosaurus, de la Beche, 1849, p- xliii.
Macrosaurus laevis Owen, 1849, p. 382.
Tylosaurus laevis, Miller, 1955, p. 909:

Type. Specimen not located.

ADDITIONAL REFERENCES. Leidy, 1865a, pp. 74, 118; 1865b, p. 70. Cope, 1868b, p.
734; 1869b, p. 260; 1869-1870, p. 205; 1875, pp. 161, 270. Owen, 1879, p. 24. Merriam,
1894, p. 24. Williston, 1897d, pp. 182, 183; 1925, p. 273. Gilmore, 1928, p. 87. McDowell
and Bogert, 1954, p. 132. Romer, 1956, p. 562.

Discussion. Macrosaurus laevis is based on two vertebrae from the New Jersey
Cretaceous, figured by Owen (1849, pl. 10 figs. 1-6). They belong to the posterior dorsal
region of a species of Mosasaurus.

Mosasaurus? mitchilli (DeKay 1830) nomen vanum

Geosaurus mitchilli DeKay, 1830, p. 140, pl. 3 figs. 3-4.
Geosaurus soemeringt, Bronn, 1838, p. 534.
Mosasaurus mitchilli, Leidy, 1860, p. 92.
Liodon mitchilli, Cope, 1869-1870, p. 205.
Tylosaurus mitchilli, Miller, 1955, p. 909.

Tyre. Specimen not located, originally from “
Hook, (New Jersey), (DeKay, 1830, p. 135).

ADDITIONAL REFERENCES. Morton, 1830b, p. 246; 1834, p. 28, pl. 11 fig. 10. DeKay,
1842, p. 28, pl. 22 figs. 55, 56. Pictet, 1853, p. 506. Leidy, 1860d, p. 92; 1865a, pp. 32,
116; 1865b, p. 69. Cope, 1868b, p. 733; 1869b, pp. 262, 263; 1869c, p. 86; 1869-1870, pp.
185, 205; 1875, p. 270. Merriam, 1894, p. 24.

Discussion. The type tooth of “Geosaurus” mitchilli is indeterminate. DeKay’s
(1830, pl. 3 figs. 3-4) figure looks like that of a mandibular tooth of a mosasaurian.

. . . foot of Neversink hills, Sandy

Mosasaurus occidentalis Morton 1844 nomen nudum
Mosasaurus occidentalis Morton, 1844, p. 133 nomen nudum.

ADDITIONAL REFERENCES. Gibbes, 1850, p. 77. Leidy, 1860, p. 921; 1865a, p. 32. Cope,
1869-1870, p. 193.

Mosasaurus reversus (Leidy 1865) nomen dubium

Baseodon reversus Leidy, 1865a, pp. 69-70, 118, pl. 10 figs. 14-15.
Mosasaurus dekayi, Cope, 1869-1870, pp. 186, 194.

RUSSELL: AMERICAN MOSASAURS 181

Tyre. Specimen not located, originally “. . . from Freehold, Monmouth County, New
Jersey,” (Leidy, 1865a, p. 69) .

ADDITIONAL REFERENCES. Cope, 1868b, p. 734; 1875, p. 269.

Discussion. Baseodon reversus is based on two pterygoid teeth of Mosasaurus, which
are specifically unassignable due to the lack of good pterygoid material from the
Cretaceous of the Atlantic Coast.

Mosasaurus validus (Cope 1868) nomen vanum

Nectoportheus validus Cope, 1868a, p. 181.
Macrosaurus validus, Cope, 1868b, p. 734.
Liodon laevis in part, Cope, 1869-1870, p. 205.
Liodon validus, Cope, 1869-1870, p. 207.
Clidastes antivalidus Cope, 1869-1870, p. v.
Tylosaurus validus, Merriam, 1894, p. 25.
Clidastes validus, Miller, 1955, p. 909.

Type. AMNH 1415, from the Navesink Formation or younger Cretaceous, “. . . near
Medford, N. J... .” (Cope, 1869-1870, p. 206), collected by Charles Braddock.

ADDITIONAL REFERENCES. Cope, 1869b, p. 260; 1869c, p. 86; 1870a, p. 272; 1871b,
p- 584; 1871f, p. 414; 1874, p. 36; 1875, pp. 161, 271. Williston, 1897d, pp. 182, 183;
1898b, p. 173. Hay, 1902, p. 467. Gilmore, 1928, p. 87. McDowell and Bogert, 1954, p.
132. Miller, 1955, p. 905.

Discussion. Cope (1868a, p. 181) proposed the name Nectoportheus validus for a single
vertebra from the “middle green-sand bed” near Medford, New Jersey. It was distinguished
from other known mosasaurs by its “compressed elevated form.” Later (1869b, p. 260)
Cope referred the species to Macrosaurus, again noting that the dorsal vertebrae were
compressed and elongate. In his Synopsis (1869-1870, p. 208) however, describing Liodon
validus, Cope stated, ‘“The posterior dorsals are so much more depressed (italics mine)
than in Liodon laevis, that future discovery may justify the generic separation of the
genus Nectoportheus, which I originally applied to this animal.” It is apparent that
these last dorsals do not resemble the original type of Nectoportheus validus. On a
preceding page (ibid., p. 206) a vertebra “found near Medford, N. J., in the second
green sand bed” was assigned to L. laevis. This vertebra (AMNH 1415) corresponds
both in morphology and in collecting locality to the type of N. validus. Cope (1869-1870,
pl. 5 fig. 5) figured this specimen under the name of Macrosaurus validus and founded
the new species Clidastes antivalidus on it in the explanation to the plates. Hay (1902,
p- 467) correctly referred C. antivalidus to N. validus in the belief that the type specimen
of both species was the same vertebra. Both this species and Macrosaurus laevis are
founded on vertebrae belonging to the posterior dorsal series of a species of Mosasaurus.

The following four nominal species of Platecarpus belong either to P. ictericus or
P. coryphaeus but adequate material is not present in the types to assign them to one of
these species.

Platecarpus mudgei (Cope 1871) nomen vanum

Liodon mudgei Cope, 1871a, p. 572.
Holcodus mudgei, Cope, 1872d, p. 273.
Rhinosaurus mudgei, Marsh, 1872b, p. 463.
Platecarpus mudget, Cope, 1874, p. 36.

182 PEABODY MUSEUM BULLETIN 23

Tyre. AMNH 1501, from “. . . yellow chalk at a point six miles south of Sheridan,
Kansas,” (Cope, 1871b, p. 581).

ADDITIONAL REFERENCES. Cope, 1871c, p. 132; 1871f, p. 405; 1872f, p. 331; 1875,
pp: 157, 268, pl. 16 fig. 2, pl. 17 fig. 5, pl. 37 fig. 7. Merriam, 1894, p. 31. Williston,
1898a, p. 29; 1898b, p. 186. Lane, 1947, p. 317.

Platecarpus affinis (Leidy 1873) nomen vanum
Clidastes affinis Leidy, 1873, pp. 283, 344, pl. 34 figs. 6-7.

Type. USNM 69, from “. . . Cretaceous formation on the Smoky Hill River, Kansas,”
(Leidy, 1873, p. 283).

ADDITIONAL REFERENCES. Cope, 1874, p. 33; 1875, p. 266. Williston and Case, 1892,
pp. 17, 28. Merriam, 1894, p. 35. Williston, 1898b, p. 198. Merril, 1907, p. 66.

Platecarpus planifrons (Cope 1874) nomen vanum

Clidastes planifrons Cope, 1874, p. 31.
Platecarpus planifrons, Williston, 1898b, p. 188.

Type. AMNH 1491, from “Niobrara . . . Kansas,” (Cope, 1875, p. 265) collected by
B. F. Mudge.

ADDITIONAL REFERENCES. Cope, 1875, p. 135, pl. 22 figs. 1-7, pl. 23 figs. 1-15, pl. 35
fig. 16; 1878, p. 299. Williston and Case, 1892, p. 17. Merriam, 1894, p. 35. Lane, 1947,
pests.

Platecarpus anguliferus (Cope 1874) nomen vanum

Sironectes anguliferus Cope, 1874, p. 34.
Platecarpus anguliferus, Williston, 1897c, p. 107.

Type. AMNH 1551, from “. . . gray calcareous shale of Trego county, Kansas,”
(Cope, 1874, p. 34) collected by B. F. Mudge.

ADDITIONAL REFERENCES. Cope, 1875, pp. 139, 267, pl. 23 figs. 16-18, pl. 24 figs. 1-15.
Merriam, 1894, p. 59. Williston, 1891, p. 345; 1897d, p. 181; 1897e, p. 244; 1898b, p. 192.
Gilmore, 1928, p. 87.

The following three nominal species of Platecarpus are based on vertebrae alone.

Platecarpus latispinus (Cope 1872) nomen dubium

Liodon latispinus Cope, 1872b, p. 169.
?Platecarpus latispinus, Cope, 1875, pp. 155, 268, pl. 27 figs. 1-4.
Platecarpus latispinus, Merriam, 1894, p. 31.

Type. AMNH 1536, from “. . . one mile south-west of Sheridan near the ‘Gypsum
Buttes’ ” (Cope, 1872d, p. 277) collected by B. F. Mudge from the Pierre Shale.

ADDITIONAL REFERENCES. Cope, 1872f, p. 332; 1874, p. 38. Williston, 1898b, p. 181.
Loomis, 1915, p. 557.

RUSSELL: AMERICAN MOSASAURS 183

Platecarpus tectulus (Cope 1872) nomen vanum

Holcodus tectulus Cope, 1872d, p. 271.
Platecarpus tectulus, Cope, 1874, p. 36.

Type. AMNH 1570, from “. .. a low bluff . .. on Butte Creek, fourteen miles south
of Fort Wallace,” (Cope, 1872d, p. 272) collected by E. D. Cope, from the Niobrara
Chalk.

ADDITIONAL REFERENCES. Cope, 1875, pp. 159, 269, pl. 21 figs. 3-6, (referred specimen)
pl. 27 figs. 5-10. Merriam, 1894, p. 31. Williston, 1898b, p. 183. Loomis, 1915, p. 557.

Platecarpus glandiferus (Cope 1872) nomen vanum

Mosasaurus near M. depressus, Cope, 1872b, p. 168.

Liodon glandiferus Cope, 1872d, p. 276.

Platecarpus glandiferus, Cope, 1874, p. 36.

?Platecarpus glandiferus, Cope, 1875, p. 268, pl. 26 figs. 13-14.

Type. AMNH 1553 (from Niobrara Chalk), ‘““This species is represented by portions
of two individuals (in each case only a cervical vertebra) from localities twenty-five
miles apart. (One) . . . was found... at the foot of a bluff on the lower part of the
Butte Creek; the second . . . from a point one mile southeast of Sheridan near the North
Fork of the Smoky River,” (Cope, 1872d, p. 276). The first vertebra was collected by
E. D. Cope, the second by B. F. Mudge. The second vertebra has not been located;
Williston (1898b, p. 182) noted that it must have come from the Pierre.

ADDITIONAL REFERENCES. Cope, 1872f, p. 332; 1875, p. 156. Merriam, 1894, p. 31.
Williston, 1898b, p. 182.

Platecarpus brachycephalus Loomis 1915 nomen dubium

Platecarpus brachycephalus Loomis, 1915, p. 556, figs. 1-9.

Tyre. AC 389, from “. . . the head of Mule Creek, 20 miles due west of Edgemont
(South Dakota) .. .” (Loomis, 1915, p. 555), “. . . from DeGrey, Verendrye, or Virgin
Creek (Member of Pierre Formation) . . .” (Kauffman and Kesling, 1960, p. 232).

ADDITIONAL REFERENCES. Crandell, 1950, p. 2345.

Discussion. Loomis (1915, pp. 559, 560) stated as a specific character the presence of
only 11 teeth in the jaws of the type and cotype, but the preservation of cranial ele-
ments is not adequate to maintain this with any degree of certainty, nor for that matter
to identify the skulls on a specific level.

The scapula figured by Loomis (1915, fig. 6) has a constricted neck and gently
rounded dorsal margin reminiscent of Plioplatecarpus. Another scapula identified as a
coracoid (ibid., fig. 7) has a small but sharply defined emargination anterodorsally, but
is otherwise similar to that of Platecarpus. The hind limb (ibid., fig. 9) is very peculiar
and was not. found associated with other material. The femur is unusually broad for
that of a mosasaur, the tibia is hardly expanded at its ends and the element in the posi-
tion of a fibula resembles the ulna of Plioplatecarpus (see Dollo, 1882, pl. 6 fig. 1). It is not
at all certain that the material referred by Loomis to P. brachycephalus all belongs to
the same form as that of the type skull.

Platecarpus? minor (Gibbes 1850) nomen vanum
Mosasaurus minor Gibbes, 1850, p. 77.

Tyrer. Specimen not located.

184 PEABODY MUSEUM BULLETIN 23

ADDITIONAL REFERENCES. Gibbes, 1851, p. 7, pl. 1 figs. 4-5. Leidy, 1860, p. 92; 1865a,
p. 32. Cope, 1869b, p. 264: 1869-1870, pp. 189, 198; 1875, p- 270. Zangerl 1948, p. 15.

Discussion. Mosasaurus minor was based on three vertebrae from the Cretaceous of
Alabama and two indeterminate teeth from unspecified localities in Georgia and Ala-
bama. A vertebra figured by Gibbes (1851, pl. 1 fig. 3) resembles those of the posterior
dorsal region of Platecarpus.

Tylosaurus dyspelor (Cope 1871) nomen vanum

Liodon dyspelor Cope, 1871a, p. 572.
Rhinosaurus dyspelor, Marsh, 1872b, p- 19.
Tylosaurus dyspelor, Leidy, 1873, p. 271, pl. 35 figs. 1-11, pl. 36 fig. 16.

Type. AMNH 1580 and USNM 41 (same individual), “. . . from the yellow beds of
the Niobrara epoch of the Jornada del Muerto, near Fort McRae, New Mexico,” (Cope,
1875, p. 167), collected by Dr. W. B. Lyon.

ADDITIONAL REFERENCES. Cope, 1869-1870, errata and addenda p. ii; 1871b, p. 574;
1871c, p. 132; 1871f, p. 410; 1872b, p. 169; 1872c, p. 280; 1872f, p. 333; 1874, p. 38; 1875,
p- 167, pl. 31 fig. 5. Merril, 1907, pp. 74, 80.

REFERENCES TO Non-TyPICAL MATERIAL. Merriam, 1894, p. 23, pl. 4 figs. 3-4. Williston,
1897c, p. 110; 1898b, p. 175; 1902, p. 253. Sternberg, 1908, p. 111. Huene, 1910, p. 297,
fig. 1. Lambe, 1914, p. 402.

Discussion. Tylosaurus dyspelor cannot be referred to either of the two species of
Niobrara Tylosaurus and must be regarded as a nomen vanum. Because of the large size
of the type it is more likely to be a junior synonym of T. proriger than a senior synonym
of T. nepaeolicus.

Tylosaurus? perlatus (Cope 1870) nomen vanum

Mosasaurus brumbyi, Cope, 1869-1870, p- 198.
Liodon perlatus Cope, 1869-1870, errata and addenda, p. ii
Tylosaurus perlatus, Merriam, 1894, p. 25.

Type. AMNH 2391, from the Selma Chalk of Alabama (Cope, 1870b, p. 497).

ADDITIONAL REFERENCES. Cope, 1871b, p. 576; 1874, p. 37; 1875, pp. 161, 271.
Lydekker, 1888, pp. 267-269. Zangerl, 1948, p. 15.

Discussion. The type specimen of Tylosaurus? perlatus is a single vertebra from the
posterior thoracic region. It is generically indeterminate.

Mosasaurus FROM NEw JERSEY, species uncertain.

Harlan, 1825, p. 235, pl. 14 figs. 2-4; 1834b, p. 81; 1834e, p. 32; 1835a, p. 285; 1835c,
p. 384, pl. figs. 2-4. Morton, 1830a, p. 289; 1834, p. 27, pl. 11 fig. 9. Pictet, 1845, p. 64.
De la Beche, 1849, p. xliii. Owen, 1849, p. 382, pl. 10 fig. 5. Holmes, 1849, p. 197. Meek
and Hayden, 1857, p. 127. Leidy, 1858b, p. 176; 1860, p. 91; 1865a, figs. 4, 20, 21, 25, 32,
pl. 7 figs. 9-14, pl. 9 figs. 8-9, pl. 10 fig. 6, pl. 11 figs. 1, 2, 4, 5, 15. Emmons, 1860, p. 207.
Cope, 1869b, p. 260; 1869c, pl. 2; 1869-1870, figs. 47(2), 49(4); 1895, pl. 31 figs. 2-2a. Owen,
1877, p. 683. Hoffmann, 1890, p. 1321.

UNDETERMINED MOSASAURS FROM THE WESTERN UNITED STATES.

Leidy, 1865a, pl. 2 figs. 15-16, pl. 7 figs. 17-20, pl. 8 figs. 6, 8; 1873, pl. 36 fig. 15. Cope,
1875, pl. 26 fig. 1, pl. 27 fig. 11, pl. 34 figs. 1-42, pl. 35 figs. 1-10. Williston, 1898b, pl.
Sslehow i.

STRATIGRAPHIC
AND PALEOECOLOGIC CONSIDERATIONS

In North America mosasaur remains have been discovered wherever marine
rocks of upper Cretaceous age outcrop and there is reason to believe that the
animals lived in abundance in coastal waters around the entire continent at
that time. The ecology and faunal succession of mosasaurs will now be considered
in the light of information contained in the sediments in which they are found.
All known occurrences of American mosasaurs are listed (see Charts 1—4).

Gur Coast CRETACEOUS
(Charts 1 & 2)

The earliest known occurrence of mosasaurs in the New World is in the
Turonian of Texas, where a continuous sequence of mosasaur-bearing rocks
extends up to the top of the Cretaceous section. However most of the material
consists of isolated specimens taken from localities widely scattered both in time
and space. We have much to learn about the history of mosasaurs in this region.
According to Scott (1940, p. 1193) ammonites that inhabited depths of 100
fathoms or more are not found in the outcropping Texas Cretaceous and the
seas must have been shallow in this area. Foraminifera from the Taylor Forma-
tion in the Brazos River Valley (east central Texas) suggest deposition in normal
marine water not more than a few hundred feet deep, but the shoreline may
have lain far to the northwest (Smith, 1959, pp. 7-8). The Navarro Formation
was deposited under similar conditions, although occasional shallow marine or
beach sands indicate that at times the shoreline moved as far gulfward as the
present area of outcrop.

The mosasaurs studied by Gilmore (1926, identified in the present paper as
Mosasaurus maximus, Plioplatecarpus sp. and Prognathodon? sp.) probably
came from the Coon Creek Formation of McNairy County, Tennessee (see
Pryor, 1960, p. 1481). This unit was deposited in marine waters of inner neritic
depths (about 300 feet) along the growing edge of a delta front, fed by a river
flowing from the Appalachians into the northeast corner of the newly-formed
Mississippi embayment (Pryor, 1960, p. 1496, fig. 17). Oxygen isotope ratios in
mollusk shells from the Coon Creek Formation indicate water temperatures of
68°-84°F and subtropical climates comparable to that of present-day Bermuda
(Lowenstam and Epstein, 1954, p. 229).

During periods of maximum transgression part of the Cretaceous sea over
Mississippi and Alabama may have exceeded 100 fathoms (600 feet) in depth,
but most of the sedimentation occurred on the continental shelf, and the
Cretaceous rocks in northern Mississippi and western Alabama were laid down
along the seaward edge of a highly indented coastline (Mellen, 1958, p. 10;
Conant, 1964). The Eutaw Formation and shales in the Mooreville Member
of the Selma Chalk were apparently deposited in shallow, brackish water
(Priddy, 1954, p. 164), though most of the Selma Chalk was probably laid down
in deeper offshore waters (Pryor, 1960, p. 1481). Platecarpus occurs in the Eutaw
Formation and a large undescribed mosasaur fauna has been collected from the

Mooreville Member (Zangerl, 1948, p. 15).
185

186 PEABODY MUSEUM BULLETIN 23

NEw JERSEY CRETACEOUS
(Chart 3)

In most instances locality information is inadequate to determine exactly
the mosasaur-producing horizons of the New Jersey Cretaceous, but it seems
certain that nearly all of the material has come from the Navesink, Redbank
and Tinton Formations, and equivalent portions of the New Egypt. Benthonic-
planktonic foraminiferal ratios indicate these units were deposited in 100-300
feet of normally saline water (Olsson, 1963). Parasitic sponges on the oysters
further suggest clear, silt-free water and the presence of a nearby river mouth to
supply an abundance of nutrients. The occurrence of interlocked aggregates of
oyster shells show that the generally quiet bottom conditions were occasionally
disturbed by strong currents (Krinsley and Schneck, 1964, p. 278). According to
Lowenstam and Epstein (1954) oxygen isotope ratios taken from belemnite
rostra indicate water temperatures of 67°F, but as belemnite paleotemperatures
usually fall below the mean, the ocean temperatures were probably similar to
those found for the Mississippi embayment (Coon Creek Formation) and the
general climate must have been subtropical.

NIOBRARA FORMATION
(Chart 4)

The Niobrara Formation of western Kansas has yielded more mosasaurs
of more different kinds and generally in a better state of preservation than any
other formation in the world. Niobrara specimens made up by far the greatest
number of those examined in the course of this study.

According to Reeside (1957, pp. 525-528, figs. 15-16) the extent of the Cretace-
ous sea over the midcontinent was at its greatest during the period of time in
which the Niobrara Formation was being deposited. The marine invertebrates
found in the chalk suggest shallow depths; well-bedded deposits and articulated
vertebrate remains suggest relatively quiet waters. Lithologic evidence indicates
a low elevation and slow rate of erosion of the bordering land some 200 miles
to the east of western Kansas, with a land of considerable relief about 600 miles
to the west, and low coastal swamps to the southwest. LeRoy and Schieltz (1958,
p- 2461) note that the foraminiferal fauna of the Smoky Hill Chalk Member
along the Front Range in Colorado implies a “. . . normal, well-circulated, open
sea, shelf environment.”” According to Merriam (1963, p. 67), ““The Fort Hays
Limestone (at the base of the Niobrara Formation) indicates a clear-water
marine environment. A moderate change in conditions, seemingly marked by
increased turbidity of the shallow sea, resulted in deposition of the Smoky Hill
Chalk. Volcanic activity must have been at a maximum at this time in order
to produce the many bentonite beds in the Niobrara.”

Reeside (1957, p. 528) postulated that subsequent to Niobrara deposition
the sea became restricted to a narrow north-south band (Telegraph Creek—
Eagle deposits) in the western portion of its former basin, though Niobrara
deposits to the east were subjected to erosion. Jeletzky (1955), Gill and Cobban
(1961, p. D185) and Scott and Cobban (1964, p. L26) have shown, however, that
the Telegraph Creek Formation and Eagle Sandstone are lateral equivalents of
the upper portion of the Smoky Hill Chalk. Merriam (1963, p. 42; see also
Gill and Cobben, 1961) notes that the Niobrara-Pierre contact in western
Kansas appears to be gradational, and that there was no break in the con-
tinuity of sedimentation. As seen below, vertebrate evidence is suggestive of a

RUSSELL: AMERICAN MOSASAURS 187

shallowing of the sea and/or approach of the eastern shoreline during latest
Niobrara time.

Williston (1897e) separated what is now called the Smoky Hill Member
into two units, the yellowish Hesperornis beds above and white Rudistes beds
below. The smaller pterodactyls, toothed birds, turtles and Clidastes were re-
stricted to the upper unit, and vertebrates of all kinds except plesiosaurs were
relatively far more abundant there. Williston interpreted this as evidence of
shallowing water and approaching shore lines near the top of the formation.
Zangerl (1953, p. 270) also noted that the species of toxochelyid turtles from
the Niobrara are endemic and considered this as evidence of progressive isolation
of the Niobrara Sea from the Gulf Coast.

There are indications of a subdivision of mosasaur assemblages within the
Smoky Hill Chalk of western Kansas, which may correspond approximately
to Williston’s original divisions. During the Yale expedition of 1871 Marsh
and his crew worked along the banks of the Smoky Hill River east of Fort
Wallace, where only the uppermost part of the Niobrara is exposed imme-
diately beneath the Pierre Shale. Marsh’s locality data are good and, of the
specifically determinable specimens collected, ten were of Clidastes propython,
nine were of Platecarpus ictericus, and four were of Tylosaurus proriger. Two
skulls of P. coryphaeus were discovered near Russell Springs about 24 miles
east of Fort Wallace, but except for the first three species no other kind of
mosasaur was found in the chalk west of this point. It is postulated that the
Smoky Hill Chalk may be divisible into an upper C. propython-P. ictericus-T.
proriger zone and a lower zone in which C, liodontus-P. coryphaeus-T. nepaeolicus
occur, perhaps to the exclusion of other forms. Negative evidence supporting
this hypothesis lies in the fact that C. propython, P. ictericus and T. proriger
all occur in the lower Pierre, but the other three species do not. The presence
of the latter three species in the lower Pierre also supports Jeletsky’s (1955, p.
883) contention that there can be little, if any, break in time between the end
of Niobrara and beginning of Pierre deposition.

Of the 36 generically identifiable skulls collected by Marsh in 1871 from
the upper part of the Niobrara about 21% are of Tylosaurus, 30% of Clidastes
and 49% are of Platecarpus. Considering the total number of skulls (398) taken
from the Niobrara by Marsh or his collectors, 25% are of Tylosaurus, 13% are
of Clidastes and 62% are of Platecarpus. These figures show that Clidastes be-
comes relatively more abundant higher in the formation, and support Williston’s
(1897e, p. 245) statement of the fact. This may be taken as evidence that Clidastes
inhabited near-shore waters and that Platecarpus and Tylosaurus ranged more
widely out to sea.

PIERRE FORMATION
(Chart 4)

Following the Eagle regressive phase, the sea expanded to the west over most
of the area it had formerly occupied, at the same time dark shales were being
deposited in the central and eastern portions of the seaway. The western strand
line then moved eastwards, bringing with it littoral sands and continental
sediments (Mesaverde group) out into the central region of the seaway. The
sea again re-advanced, depositing muds far to the west, ultimately to withdraw
to the east where sands (Fox Hills Formation) mark the closing phase of Cretace-
ous marine deposition on the midcontinent (see Reeside, 1957, pp. 530, 540,
modified by Gill and Cobban, 1961). From a study of the foraminifera of the lower

188

PEABODY MUSEUM BULLETIN 23

sninoaolt gaisvall>

=
le)
=
x
a
¢
e
a
5S
$
co

TYLOSAURUS NEPAEOLICUS

nner

2
=
yy
™
S
ey
%
c
ry
(a)
m
2
QO
cS
a

/O
to the species level:

Text-fig. 96. Mosasaur skulls in the Marsh collections at the U. S. National Museum and Peabody
belong to the genus Platecarpus, 25% to Tylosaurus and 13% to Clidastes. Ectenosaurus,

Museum at Yale, collected from the Smoky Hill Member, Niobrara Chalk of western Kansas
Total number of generically recognizable skulls = 398. Sixty-two per cent of these skulls
which makes up 1%

of the Marsh collection, has been omitted. Of the skulls identifiable

In Platecarpus 22.6% are cory phaeus (26 individuals)
77.4%, are ictericus (89 individuals)
In Tylosaurus 52.75% are proriger (19 individuals)
47.25% are nepaeolicus (17 individuals)
72% are

propython (33 individuals)

In Clidastes

26% are liodontus (12 individuals)

2% are sternbergi (1 individual)
The average number of elements collected of the skulls of specifically recognizable specimens
is indicated by the numbers on the vertical radius. Thus in Clidastes propython on the average
about 10.5 elements were collected of each skull. This is presumably an indication of how
many elements were originally preserved.

It may be possible to divide the Smoky Hill Chalk into two zones on the basis of mosasaur
species (see text), specimens from which would be mixed in this chart. Note that the three

species from the lower zone (Clidastes liodontus, Platecarpus coryphaeus, Tylosaurus nepaeolicus)
are apparently more completely preserved on the average than the other species.

RUSSELL: AMERICAN MOSASAURS 189

Pierre (Gregory Member) of South Dakota, Johnson (1959, p. 49) suggested that
the waters of the Pierre sea were shallow, warm, slightly alkaline and of marine
to brackish salinity. Crandell (1958, p. 18) postulated that during Pierre deposi-
tion the eastern shoreline lay 200-300 miles to the east of central South Dakota.
He also noted that at least part of the Pierre in central South Dakota accumulated
in rather deep water. Fine layers of bentonite indicated deposition in quiet
water below the wave base. Mosasaurs known from the lower Pierre include
Clidastes propython, Globidens alabamaensis, Platecarpus ictericus, P. somenensis
and Tylosaurus proriger. Mosasaurus conodon and Plioplatecarpus primaevus
occur in the middle Pierre, and M. missowriensis and Prognathodon overtoni
are known from the upper Pierre (see Appendix B).

The Pierre mosasaurs, although as yet poorly understood, promise to be a
very rich and diversified assemblage. In view of the survival of at least three
Niobrara species into the lower Pierre it would seem that physical changes
in the sea caused by the replacement of chalk by dark shale deposition were of
little consequence to mosasaurs. The time during which the Pierre Formation
was being deposited spans nearly the entire period between the end of Niobrara
sedimentation and the close of the Cretaceous. The Niobrara mosasaur fauna is
grossly very well known and scattered collecting has indicated that a good suc-
cession of mosasaur faunas is present from the base to the top of the Pierre.
Further sampling would doubtless reveal highly interesting changes in the
composition and ultimate fate of the mosasaur community.

ADDITIONAL OCCURRENCES OF AMERICAN MOSASAURS

1. Three vertebrae (Holmes, 1849, p. 197) from the “marl” of Ashley River,
South Carolina. See also the section on systematics for miscellaneous nineteenth
century occurrences in the southern Atlantic states, (p. 176).

2. The type vertebra of Clidastes iguanavus from the Marshalltown Forma-
tion of New Jersey (see Cope, 1869a, p. 233).

3. Mosasaur teeth (Leidy, 1873, pp. 279-280, pl. 34 figs. 18-22) from the
Pierre (?) Formation near Santee, L’Eau qui Court County, Nebraska.

4. Vertebral fragments (Cope, 1877, p. 569) probably from the Judith
River Formation, Judith River region, Montana.

5. A Clidastes skeleton from near Flager, Colorado, and mosasaur bones
from near Canyon City, Colorado (Lee, 1897, p. 614; Williston, 1898b, p. 196).

6. A Clidastes (identified by Lucas in Hill, 1901, p. 328) from the top of the
Eagle Ford Formation, southwest of Waco, Texas.

7. Mosasaurs (Douglass, 1902, p. 212) from the Bearpaw Formation of
Sweetgrass County, Montana.

8. A Platecarpus hind paddle (Williston, 1902, p. 252) probably from the
Niobrara Chalk near Milton, North Dakota (see also Laird, 1951, p. 12).

9. Platecarpus remains (Loomis, 1904, p. 254) from the Pierre redeposited in
the Chadronian of South Dakota.

10. Mosasaurus teeth (Gilmore, 1912b, p. 2) from the Fox Hills Formation
south of Bismark, North Dakota. K. M. Waage has also recently discovered
mosasaur teeth in the Fox Hills Formation of Montana and Wyoming.

11. Mosasaurus? caudal vertebrae (Sternberg, 1915, p. 152) from the Bearpaw
Formation, Judith River region, Montana.

12. M. conodon from the Marlbrook Marl near Columbus, Arkansas.

13. The type of Amphekepubis johnsoni from the ?San Felipe Formation,
north and east of Monterrey, Nuevo Leon, Mexico (Mehl, 1930).

190 PEABODY MUSEUM BULLETIN 23

14. Mosasaur vertebrae from Rayon, Tamaulipas, Mexico (Miillerried, 1931).

15. Tylosaurus skeleton (Renger, 1935) from the lower Selma Formation,
near West Greene, Greene County, Alabama.

16. Platecarpus skeleton (Dowling, 1941) from the lower Selma Formation,
five miles southeast of Eutaw, Greene County, Alabama.

17. Skeletal remains of Plotosaurus bennisoni, P. tuckeri and Plesiotylosaurus
crassidens from the Moreno Formation (Maestrichtian) along the west side of
the San Joaquin Valley, California (Camp, 1942).

18. Mosasaurus? dentary (USNM 3026) from the Tombigbee Sands of
Plymouth Bluff on the Tombigbee River, Lowndes County, Mississippi.

19. Mosasaurus tooth (USNM 12287) from the upper part of the Selma Chalk,
one and one-half miles northeast of Scooba, Kemper County, Mississippi.

20. Globidens? posterior half of a mandible (USNM 22965) from the upper
part of the Austin Chalk, one and one-half miles northeast of Spofford, Kinney
County, Texas.

21. M. maximus skull and skeleton (TMM) from the Navarro Formation
on an east bluff of Onion Creek, about 500 yards north of the bridge on the
Austin-Bastrop highway, Travis County, Texas (personal communication, Langs-
ton, 1963).

22. Tylosaurus proriger quadrate (TMM 40601) from the Taylor Forma-
tion, Texas (personal communication, Langston, 1963).

23. T. proriger skull (USNM 17909) from the Telegraph Creek Formation,
Bighorn County, Montana (see appendix B).

24. Tylosaurus and Clidastes skulls (USNM 18496, 22964) from the Taylor
Formation, one mile east of Culeoka in the Lavon Reservoir area, Collin County,
Texas.

25. Large Platecarpus cranial fragments (Museum of Southern Arkansas
State College) from the Brownstown Formation, northeast of Columbus,
Arkansas, and midway between Prescott and Delight (personal communication,
Chapman, 1963).

26. Platecarpus skull (YPM 1444) possibly from a marine sandstone tongue
of the Mesaverde Formation, along one of the tributaries of Crow Creek,
Wyoming.

27. Platecarpus cf. somenensis (YPM unnumbered) from the Pierre Shale,
near Fort Randell, South Dakota.

28. Mosasaurs have been discovered from the Arctic Coast of the Northwest
Territories, Canada (Russell, 1967).

FOREIGN OCCURRENCES OF MOSASAURS
(Charts 5, 6, & 7)

Outside of the United States, the best known superposed succession of
mosasaur faunas occurs in Belgium. Other countries with a wide vertical range
of mosasaur-producing horizons are France, Sweden and England. These faunas
are listed in stratigraphic order along with those from the different regions of
the United States in the accompanying charts (Charts 1—7).

Fragmentary remains of mosasaurs have been discovered on all of the con-
tinents of the world except Antarctica. In Europe they have also been reported
from Ireland (Swanston, 1886), Germany (Pompeckj, 1910), Italy (Leonardi,
1946) and southern european Russia (Iakovlev, 1901, 1906; Bogoliubov, 1910;
Pravoslavlev, 1914, 1916; Tsaregradskii, 1927, 1935). Mosasaurs are known from
Israel (Raab, 1963) and Jordan (Repelin, 1915, Avnimelech, 1949, Signeux, 1959)

RUSSELL: AMERICAN MOSASAURS 191

in western Asia. In Africa mosasaurs are known to occur in the Union of South
Africa (Broom, 1912), Nigeria (Swinton, 1930), northwestern Africa (Aram-
bourg, 1952), Libya (Quass, 1902), Egypt (Stromer and Weiler, 1930; Zdansky,
1935, Leonardi and Maloroda, 1946) and Angola (Telles Antunes, 1964). They
occur in Venezuela (Pierce and Welles, 1959), Brazil (Price, 1957) and Argentina
(Ameghino, 1918) in South America. In the southwestern Pacific mosasaurs are
known from Timor (E. Huene, 1935), Australia (Lundelius and Warne, 1960)
and New Zealand (Hector, 1874).

In general the material from these localities is too incomplete to be of
much use in understanding the paleozoogeography and evolution of mosasaurs.
It is interesting to note that spherical teeth of the Globidens alabamaensis kind
have been found in the Maestrichtian of Brazil (Price, 1957), North Africa
(Arambourg, 1952; Zdansky, 1935), Jordan (Signeux, 1955) and probably also
Timor (E. Huene, 1935), and from the lowermost Upper Campanian of Belgium
(Dollo, 1924). Thus, the distribution of Globidens very likely coincides with the
tropical “Mediterranean belt” of coral and rudistid reefs (see Berquist and
Cobban, 1957, p. 871).

Tylosaurus or a very closely related form occurs in the Santonian or lower
Campanian of Russia (See Tsaregradskii, 1927, p. 572, pl. 8 fig. 1), and in the
Maestrichtian of New Zealand (Hector, 1874, pl. 30; Wellman, 1959) and Bel-
gium (Hainosaurus bernardi). In North America, however, tylosaurines seem
to disappear after early Campanian time. The osteology of mosasaur species
inhabiting North America east of the Rocky Mountains and western Europe
during the Maestrichtian is generally not well known, but it is quite possible
that the same forms were common to both regions. An excellent skeleton of
Mosasaurus conodon from South Dakota agrees in astonishing detail with other
complete specimens from Belgium. If Funnell’s (1964) observation is correct
that the North Atlantic Ocean may have been 1,500 miles narrower at the close
of Cretaceous time than it is now, it would not be surprising to find that species
of large marine reptiles ranged across it.

It has commonly been assumed that individual genera of mosasaurs did not
have a wide geographic distribution, but it is now evident that this was not the
case. However, an indication that some faunal provinciality existed lies in the
occurrence of apparently endemic mosasaur genera in Californian sediments of
Maestrichtian age. An extensive north-south land isthmus perhaps linked to
South America, or an ecologic barrier of some other kind must have prevented
the dispersal of eastern genera into the California seas. Unfortunately almost
nothing is known of the mosasaurs that inhabited other regions of the Pacific
basin at that time, although Taniwhasaurus from New Zealand does show some
similarity to the Californian Plotosaurus.

EcoLocicaAL CONCLUSIONS

Stratigraphic evidence demonstrates that in the United States during late
Cretaceous time mosasaurs inhabited shallow epicontinental seas, or in some
cases, brackish water, near-shore environments. Oxygen isotope ratios derived
from mollusk shells indicate water temperatures similar to those found in sub-
tropical regions today. In the Niobrara Chalk Platecarpus and Tylosaurus may
have frequented deeper water located farther from the coast more often than did
Clidastes. Mosasaurs in the interior of North America have been found at least
400 miles away from the nearest contemporaneous strand line. Assemblages
from the upper Niobrara and lower Pierre are similar and show that changes

192 PEABODY MUSEUM BULLETIN 23

in deposition between the formations had little effect on them. Similarly
Mosasaurus conodon has been found in shales (South Dakota), greensands (New
Jersey) and chalk (Belgium), and T. proriger has been found in sandstone
(Montana), shales (Kansas), calcareous shales (Colorado) and chalk (Kansas). The
change in mosasaur faunas from the base to the top of the Pierre is attributed to
a gradual evolution of the Atlantic mosasaur population.

The remains of only one obviously immature mosasaur hatchling (YPM
4818) have been seen by the author and a few juvenile specimens have been
collected, all from the Niobrara Chalk. Many excellent adult skeletons are
known in which even the stomach contents and cartilagenous parts are pre-
served, but embryonic young have never been found inside the body cavity.
This is rather compelling negative evidence that mosasaurs were not ovovivipar-
ous. Female mosasaurs may have crawled, probably with great difficulty, out
onto the banks of large rivers in order to lay their eggs in the vicinity of shallow
protected waters offering an abundance of small prey for the newly-hatched
animals, as suggested by Williston (1904, pp. 48-50; 1914, pp. 163-165). An-
other possibility for which there is at present no evidence is that mosasaurs,
like the pinnipeds and some marine turtles (see the recent work of A. Carr and
associates), may have migrated to island or offshore bar rookeries where the
adults were safe from large terrestrial carnivores. In this case, however, it
seems that the young would have been at a more serious disadvantage because
of their greater exposure to marine predators.

EVOLUTIONARY CONSIDERATIONS

Mosasaurs were the largest lizards that have ever lived. Their sudden ap-
pearance early in the fossil record of lizards and truly spectacular though brief
success in the oceans of late Cretaceous time underline their uniqueness in
the evolutionary history of reptiles. Although their derivation and initial radia-
tion are as yet poorly documented, enough information is at hand to obtain a
general notion of the relationships and phylogeny of mosasaurs.

MosASAURS COMPARED WITH PLATYNOTAN LIZARDS

One of the few things that has been nearly universally agreed upon regarding
the classification of mosasaurs is that they resemble the varanids more closely
than any other group of extant lizards, a fact that Camp (1923, 1942) has
demonstrated in detail. It is instructive, however, to point out certain skeletal
features in which mosasaurs consistently differ from living varanids:

1. There are alveoli for only four premaxillary teeth instead of eight.

2. The anterior portion of the maxilla is never bent horizontally to floor
the nasal capsule.

3. The nasals are vestigial and separated from each other by the internarial
bar of the premaxilla instead of being fused along the midline and form-
ing the bridge between the premaxilla and frontals.

. The frontals are always fused in mosasaurs; they are separate in Varanus.

. The lacrymal is relatively much smaller.

. The jugal and postorbitofrontal are in contact behind the orbits, and
the “quadratomaxillary’” ligament was probably attached to the pos-
teroventral corner of the jugal instead of to a tuber located near the
juncture of the jugal, maxilla and ectopterygoid.

7. The parietal and postorbitofrontal meet in a vertical suture on the
anterior margin of the supratemporal fenestra; in Varanus a wide wing
from the parietal extends anterolaterally between the frontal and pos-
teromedial process of the postorbitofrontal.

8. The foramina for nerves X-XII open on the lateral instead of occipital
face of the skull.

9. In cross-section the long axis of the parietal-suspensorial arcade is hori-
zontal instead of vertical.

10. ‘The pterygoids possess a single row of teeth.

1]. The incisura piriformis is relatively quite narrow.

12. The suprastapedial process of the quadrate and the tympanic ala are
very much larger.

13. There is a well-developed, transversly-oriented joint between the splenial
and angular. In Varanus these elements loosely overlap one another with-
out forming articulating facets.

14. The long, posteromedially curving process from the coronoid which
bounds the meckelian fossa internally in Varanus is lacking in mosasaurs.

15. The intercentral articulations of mosasaur vertebrae face directly an-
teroposteriorly; in Varanus they are inclined in an anteroventral-postero-
dorsal direction.

16. There are seven cervical vertebrae in mosasaurs as compared to nine.

193

Doe

194

7

18.

19.

20.

Pale

22.

PEABODY MUSEUM BULLETIN 23

Cervical ribs occur on every vertebra but the atlas, instead of only on the
last 3-4 cervicals.

At least five ribs have cartilaginous sternal connections in mosasaurs;
only three such ribs exist in Varanus.

The presence of long ribs indicates that the thoracic cavity was restricted
to the anterior portion of the dorsal region. Long ribs occur on all but
the last six dorsal vertebrae in Varanus and the thoracic cavity is corre-
spondingly more elongate.

The vertical dorsal tip of the ilium could have been only ligamentously
bound. to a single sacral vertebra; in Varanus the posterodorsally sloping
ilium contacts two sacral vertebrae.

‘There are never less than five pygal vertebrae. In Varanus there is but one.
‘Transverse processes do not occur on vertebrae behind the center of the
caudal series in mosasaurs; in Varanus they occur to the end of the tail.
The limbs have been transformed into paddles with a concomitant
shortening and broadening of the pro- and epipodials, hyperphalangy
and loss of ungual phalanges. The limbs in Varanus are of the normal
lacertilian type.

On the island of Hvar (Lesina) in the Dalmatian archipelago and near
Comeno (Kémen) north of Trieste in Yugoslavia, the remains of small aquatic
lizards, called aigialosaurs, have been discovered in rocks of Cenomanian-Turon-
ian age (Langer, 1961*). So far only two aigialosaur skeletons with associated
skulls have been described, one by Kramberger (1892) of Aigialosaurus dalmaticus
and another by Kornhuber (1901) of Opetiosaurus buccichi. Another excellent
skeleton lacking the skull was named Carsosaurus marchesetti by Kornhuber
(1893). The descriptions and figures of these animals given by Kramberger and
Kornhuber are compared point by point with the list of characters separating
modern varanids from mosasaurs given above.

IK

Upon comparison of Kornhuber’s (1901) plates 2 and 3 showing the
course of the premaxillary-maxillary suture over the top of the rostrum
with the alveoli in the same region on the undersurface of the skull in
Opetiosaurus, it seems probable that four premaxillary teeth were present.
In Opetiosaurus the condition of the anterior portion of the maxilla is
as in mosasaurs.

The nasals are unknown.

The frontals are fused in Opetiosaurus.

. The lacrymal is relatively large in Opetiosaurus and Aigialosaurus.
. The jugal and postorbitofrontal are in contact behind the orbit in

O petiosaurus.

In both Opetiosaurus and Aigialosaurus the parietal-frontal-postorbito-
frontal contact is as in Varanus.

The braincase is unknown.

. The parietal-suspensorial arcade is not well-preserved.

Nopcsa (1905b, p. 121) in a footnote observes, “It cannot be certainly
known whether the pterygoids of the Aigialosaurs bore teeth, but I am
inclined to believe they did, since in Opetiosaurus the crown of a tooth
lying near the hyoid bone seems to differ in size both from the mandibular
and (presumably also) maxillary teeth of this animal.” Camp (1923, p.

*I am indebted to Mr. George Callison of the University of Kansas for bringing this
reference to my attention.

RUSSELL: AMERICAN MOSASAURS 195

$21) states that the pterygoid teeth of aigialosaurs are large, from a
source of information unknown to the present author.

11. The incisura piriformis appears to be relatively as wide in Opetiosaurus
as in Varanus.

12. The suprastapedial process of the quadrate and tympanic ala are as well
developed in Opetiosaurus and Aigialosaurus as in mosasaurs.

13. There is a well-developed, transversly oriented joint between the splenial
and angular in Opetiosaurus.

14. ‘The medial surface of the lower jaw is unknown.

15. The nature of the intercentral articulations is unknown.

16. There are eight cervical vertebrae in Opetiosaurus and all but the first
three bear ribs.

17. Five ribs have cartilaginous sternal connections in Carsosaurus.

18. The last four dorsal vertebrae in Opetiosaurus, at least the last five dorsals
in Aigialosaurus and the last seven dorsals in Carsosaurus bear short ribs
(the last dorsal may lack ribs in Azgzalosaurus and Carsosaurus). This is
true of the last six dorsals in Varanus and at least the last nine dorsals
of mosasaurs, by comparison.

19. The posterodorsally sloping ilium contacts two sacral vertebrae in
Aigialosaurus, Carsosaurus and O petiosaurus.

20. There is only one pygal vertebra in Carsosaurus.

21. Transverse processes do not occur on vertebrae behind the center of the
caudal series in Opetiosaurus.

22. The epipodials are shorter relative to the propodials in Carsosaurus,
Aigialosaurus and Opetiosaurus than in Varanus, but are not broadened.
The manus and pes in Opetiosaurus have the normal reptilian phalangeal
formula, and the terminal phalanges are clawed in Carsosaurus and
O petiosaurus.

It is evident that the two aigialosaur skulls are very similar to those of
mosasaurs, although they do possess several Varanus-like characteristics. There
are some additional noteworthy features of these skulls. The prefrontals and
postorbitofrontals fail to contact in Aigialosaurus and Opetiosaurus. In both
genera there appears to be a triangular ala projecting from the lateral margin
of the supraorbital wing of the prefrontal. In Opetiosaurus the posterodorsal
process of the coronoid is very high and may have been buttressed by a thin
sheet of bone from the dorsal rim of the surangular, and there is a large number
of teeth (at least 17) in the dentary. In Aigialosaurus the suprastapedial process
is very large and apparently fused to the infrastapedial process, although the
bone may have been partially crushed. The appearance of the quadrate in
Opetiosaurus is very similar to that of Clidastes, and in both Aigialosaurus and
Opetiosaurus the tympanic ala is heavy. The latter series of characters is also
found in mosasaurs of the subfamily Mosasaurinae and among these is especially
reminiscent of Clidastes.

The retention of a solid sacrum, clawed digits and a well-developed joint
between the pro- and epipodial elements gives a strongly varanoid appearance
to the post-cranial skeleton of aigialosaurs. Certain features, however, such as
the incipient concentration of thoracic cavity into the anterior portion of the
dorsal region and the lack of transverse processes on the vertebrae of the distal
half of the tail, are highly suggestive of mosasaurian affinities. Thus, although
some portions of the aigialosaur skeleton resemble corresponding elements of

196 PEABODY MUSEUM BULLETIN 23

the mosasaur skeleton more closely than do others, the known aigialosaurs surely
represent a complex of lizards which is admirably suited in morphology to give
rise to the mosasaurs.

Kuhn (1958) has recently described a small varanoid from the Solenhofen
Limestone near Eichstaétt in southern Germany. It is evident from his illustra-
tion of the outline of the skull (ibid., fig. 1) that the maxilla was not bent
medially to floor the nasal capsule, and that the jugal was sutured to the post-
orbitofrontals behind the orbits, both characteristics being typical of aigialosaurs
and mosasaurs. Kuhn concludes that his Proazgialosaurus demonstrates the
existence of an aigialosaurian type of semiaquatic lizard as early as late Jurassic
time.

The dolichosaurs are another group of small aquatic lizards that occur to-
gether with aigialosaurs in the middle Cretaceous strata of Komen and Lesina,
with one form (Dolichosaurus) known from equivalent rocks in England. Cranial
material has been described in three species, Adriosaurus suessi, Pontosaurus
lesinensis and Dolichosaurus longicollis. A head of Adriosaurus (Nopcsa, 1908,
pp. 51-52, pl. 3 figs. 1-2) is small and poorly preserved. The nasal and frontal are
separated by sutures from their opposites along the midline of the skull. The
orbits, external nares and parietal foramen are relatively large. Kornhuber
(1873, pl. 21) shows a large lacrymal in Pontosaurus and a well-developed
splenio-angular joint. The tiny mandibles of Dolichosaurus figured by Owen
(1849-1884, lacertilians pl. 8 fig. 2) appear to lack a splenio-angular joint, al-
though it is difficult to be certain. In general, the relatively short skull, high
pre-sacral vertebral count (at least 34, usually more), long neck (11 or more
cervical vertebrae) and reduced forelimbs of known dolichosaurian skeletons
(see Nopcsa, 1923, pp. 105-107), including those of Mesoleptos zendrini (Nopcsa,
1923, p. 107), Acteosaurus tommasini (Meyer, 1860, pp. 223-231, pl. 24, which is
probably conspecific with the imperfectly preserved type skeleton of Adriosaurus
suessi Seeley, 1881, see also Romer, 1956, p. 562; Nopcsa, 1908, pl. 3), Ponto-
saurus lesinensis (Kornhuber, 1873, pls. 20-21) and Dolichosaurus longicollis
(Owen, 1849-1884, lacertilians pl. 8 fig. 1, pl. 9 fig. 4), exclude them from the
aigialosaurs and the possible ancestory of the mosasaurs.

The fossil record of terrestrial platynotans is very poor. For the present pur-
poses only the skeleton of Saniwa (Eocene of North America) is sufficiently well
known to be compared usefully with those of mosasaurs and living varanids.
For a review of other Cretaceous and Tertiary platynotans see Féjervary (1918),
McDowell and Bogert (1954, p. 53) and Estes (1964, pp. 128-137). The following
information on Saniwa was taken from Gilmore (1928, pp. 53-83, figs. 31-42, pls.
3-10).

1. Alveoli for six premaxillary teeth are present in the premaxilla.

2. The anterior portion of the maxilla is bent horizontally to floor the
nasal capsule.

3. ‘The nasals are unknown.

4. The frontals are unknown.

5. ‘The lacrymal is relatively almost as large as in Varanus.

6. ‘The jugal and postorbitofrontal probably contacted behind the orbits and
there is a slight indication of an angulation on the posteroventral
corner of the jugal, indicating a possible area of attachment for the
“quadratomaxillary” ligament.

7. The parietal-frontal-postfrontal contact is as in Varanus.

8. The foramen for nerves X-XII opens on the occipital face of the skull

RUSSELL: AMERICAN MOSASAURS 197

but the thin crest of bone separating it from the lateral face of the brain-
case is not half so large as in Varanus.
9. ‘The parietal-suspensorial arcade is not known.

10. The pterygoids possess a single row of teeth.

11. The incisura piriformis is relatively narrower than is usual in Varanus
but not so narrow as in mosasaurs.

12. The suprastapedial process of the quadrate is relatively small as in
Varanus.

13. The splenio-angular joint is similar to that of Varanus.

14. There is a posteromedial process of the coronoid to the prearticular
around the meckelian fossa.

15. The intercentral articulations are inclined in an anteroventral-postero-
dorsal direction.

16. The number of true cervical vertebrae is unknown, but the first ribs
occur on the fourth? or fifth vertebra postcranially.

17. The number of dorsal ribs with cartilaginous sternal connections is un-
known.

18. The rib cage is inadequately preserved in known specimens of Saniwa.

19. The posterodorsally sloping ilium contacts two sacral vertebrae.

20. ‘The number of pygal vertebrae is unknown.

21. The posterior portion of the caudal series is unknown in Saniwa.

22. The epipodials and feet are practically unknown, but from the shape
of two preserved phalanges it seems likely that Saniwa possessed a clawed
manus and pes of the normal varanoid type.

Saniwa is clearly more closely related to the varanids than to mosasaurs,
although it does possess a few features that are intermediate between the two
groups. The postorbital-postfrontal ossifications are separate in this genus and
in Telmasaurus (Gilmore, 1943, p. 380), a questionable saniwinine from the
Cretaceous of Mongolia. Teeth are present on the palatines of Saniwa and there
is an unusually high number of marginal teeth on the maxilla (24) and dentary
(22).

Besides the monitors (varanids) there are only two living genera of rather
peculiar terrestrial lizards that have ever been considered as having a special
affinity to the mosasaurs. One of these, Lanthanotus, is a small obscure creature
from the rain forests of northern Borneo. The other, Heloderma, is a larger,
sluggish venomous lizard from the deserts of the southwestern U.S.A. and north-
western Mexico. In Lanthanotus (see McDowell and Bogert, 1954) the twenty-
two characters under discussion are as follows:

—

. There are eight premaxillary teeth.
2. The anterior portion of the maxilla is bent medially to floor the nasal
capsule as in Varanus.
3. The nasals are relatively large, fused, and form a bridge between the pre-
maxilla and frontal on the internarial bar.
4. The frontals are separate.
5. The lacrymals are small.
6. The jugal and postorbitofrontal are in contact behind the orbits.
7. The postfrontal-parietal-frontal contact is as in Varanus.
8. The foramen for nerves X-XII opens on the occipital face of the skull.
9. The suspensorial arcade from the parietal is nearly vertical in cross-
section, as in Varanus.
10. The pterygoids have a single row of teeth.

198

Male

12

“

13:
14.
15s
16.
17.
18.
IG):
20.

ie
22.

PEABODY MUSEUM BULLETIN 23

The incisura piriformis is broad as in Varanus.

The suprastapedial process of the quadrate is small as in Varanus.

The splenio-angular joint is approximately intermediate in develop-

ment between varanids and mosasaurs.

A long posteromedial process from the coronoid curves around the

anterior border of the meckelian fossa to contact the prearticular.

The intercentral articulations are inclined in an anteroventral-poster-
odorsal direction.

There are nine cervical vertebrae in Lanthanotus, with cervical ribs on

all but the first three.

Three of the anteriormost dorsal ribs have cartilaginous sternal con-

nections.

Only four of the posteriormost dorsal vertebrae lack long ribs, and the

thoracic cavity is long and slender.

The posterodorsally sloping ilium contacts two sacral vertebrae.

The number of pygal vertebrae is unknown.

Transverse processes occur to the end of the tail.

Propodial-epipodial ratios are not known in Lanthanotus. The manus

and pes have the phalangeal formula of 2-3-4-4-3 and the terminal

phalanges are clawed.

Lanthanotus also possesses teeth on the palatines. In spite of the loss of the
supratemporal arcade and parietal foramen, and the lesser number of marginal

teeth,

Lanthanotus has more characters in common with the saniwinine

(Saniwa) varanids than with any other group of extinct platynotans. It is be-

lieved

that the evidence supports a varanid ancestry for Lanthanotus more

strongly than a derivation from the Cretaceous dolichosaur-aigialosaur-mosasaur
complex as suggested by McDowell and Bogert (1954, p. 57).
In Heloderma the twenty-two characters are:

ie
Ze

3:

oOo

There are usually eight premaxillary teeth.

The anterior portion of the maxilla is bent medially to floor the nasal
capsule as in Varanus.

The nasals are moderately large, sutured together along the midline
and form a bridge between the premaxilla and frontal on the internarial
bar.

. The frontals are separate.
. The lacrymals are relatively as large as in Varanus.
. The jugal and postorbitofrontal are firmly sutured together behind the

orbits, and there is a prominent angulation on the posteroinferior corner
of the jugal indicating a possible area of attachment for the ‘“‘quadrato-
maxillary” ligament.

. The postfrontal-parietal-frontal contact is as in Varanus.
. The composite foramen for nerves X-XII opens on the occipital face of

the skull, but the thin crest of bone separating it from the lateral face of
the braincase is very small.

. The dorsal surface of the parietal-suspensorial arcade is flattened,

making the arcade triangular in cross-section.

. The pterygoids have a single row of teeth.
. The incisure piriformis is relatively as broad as in Varanus.
. The suprastapedial process of the quadrate is as small as in Varanus, but

the tympanic ala is better developed.

RUSSELL: AMERICAN MOSASAURS 199

13. The splenio-angular joint is similar to that of Varanus.

14. A long posteromedial process from the coronoid curves around the
anterior border of the meckelian fossa to contact the prearticular.

15. The intercentral articulations are inclined in an anteroventral-postero-
dorsal direction.

16. There are eight cervical vertebrae in Heloderma, with cervical ribs on all
but the first three.

17. Four of the anteriormost dorsal ribs have cartilaginous sternal connections.

18. Six of the posteriormost dorsal vertebrae lack long ribs, and the thoracic
cavity is long and slender.

19. The posterodorsally sloping ilium contacts two sacral vertebrae.

20. There is one pygal vertebra.

21. Transverse processes occur to the end of the tail.

22. The epipodials are not reduced in Heloderma. The manus and _ pes
have a normal reptilian formula (2-3-4-5-3), and the terminal phalanges
are clawed.

The premaxillae are not fused posteriorly in Heloderma and teeth are
present on the palatines. The number of teeth in the marginal dentition is com-
parable to that found in Varanus and less than in Lanthanotus and Saniwa.
The parietal foramen and supratemporal arcade are also lost, and Heloderma
appears in some respects to approach the saniwinine (Santwa) varanids more
closely than Lanthanotus.

PHYLOGENY OF PLATYNOTANS

Although the fossil record of lizards is very incomplete, it is believed that
the following conjectures are consistant with our present knowledge of the
skeleton in platynotan lizards. They are partly at variance with the interpreta-
tion presented in the detailed and richly illustrated work of McDowell and
Bogert (1954).

A diagnosis of the hypothetical ancestral platynotan may prove to be:
Four premaxillary teeth. Anterior portion of maxilla curves dorsomedially,
not bent horizontally to floor nasal capsule. Nasals separate; contact premaxilla,
frontals and maxilla; external nares do not reach prefrontals. Frontals separate.
Lacrymal relatively large. Jugal and postorbitofrontal contact behind orbit; pro-
cess on posteroinferior corner of jugal for “quadratomaxillary” ligament. Wide
wing from parietal extends anterolaterally between frontal and posteromedial
process of postfrontal; postfrontal and postorbital separate. Foramina for nerves
X-XII open on lateral face of braincase. Parietal-suspensorial arcade with hori-
zontal long axis in cross section. Single row of teeth on pterygoids and palatines.
Incisura piriformis broad. Suprastapedial process on quadrate small. Splenial
and angular overlap in loose joint in lower jaw. Distinct posteromedial process
from coronoid extending around meckelian fossa to prearticular. Intercentral
vertebral articulations face anteroventrally or posterodorsally. Nine cervical,
20 dorsal, two sacral and one pygal vertebrae. Transverse processes present on
all caudal vertebrae. Three to four cartilaginous ribs articulate with sternum,
long ribs present on all but last four dorsals. Pelvic girdle and limbs of normal
lacertilian type.

Dolichosaurs were probably derived from primitive platynotans, but their
tiny skull, long neck and reduced forelimbs preclude a close relationship with
either the aigialosaur-mosasaur or varanoid lineages. Primitive varanoids (ex-

200 PEABODY MUSEUM BULLETIN 23

Varanines
Helodermatids
Voranines Lanthanotids
Saniwinines
Saniwinines
Helodermatids Lanthanotids
Varanids TERTIARY VARANOIDS
MOSASAURS
Dolichosaurids /
Aigialosaurids
Dolichosaurids
MOSASAURS

Aigialosaurids

A B

ANCESTRAL PLATYNOTANS ANCESTRAL PLATYNOTANS

Text-fig. 97. Phylogeny of platynotans. A. After McDowell and Bogert (1954). B. The present
interpretation.

cluding aigialosaurs) may have been derived from the foregoing hypothetical
ancestral platynotan through several structural modifications: two teeth were
added to the premaxilla, and the anterior portion of the maxilla curved medially
under the nasal capsule. A ridge of bone grew up beneath the paroccipital process
that separated the exit for cranial nerves X-XII from the lateral face of the
braincase off onto the occipital faces. In cross-section the long axis of the
parietal-suspensorial arcade became vertical. In varanines proper the postorbital
arch is broken and the point of attachment of the “quadratomaxillary” ligament
shifts onto a tuber located near the juncture of the jugal, maxilla and ectoptery-
goid. The postorbital and postfrontal are fused as are the two nasals, and the
row of teeth on the palatine and pterygoid has been lost. Four teeth have
been added to the premaxilla making a total of eight present. ‘The Heloder-
matidae and Lanthanotidae are derived separately from saniwinine stock through
the addition of two more premaxillary teeth to the primitive number and the
loss of the supratemporal arcade and parietal foramen. The nasals are fused in
Lanthanotus and the splenio-angular joint is more highly developed. Teeth are
retained on both the palatines and pterygoids. About six more vertebrae have
been added to the dorsal region of Lanthanotus and Heloderma and the limbs
have become slightly reduced.

The skull of aigialosaurs would be modified from (hypothetical) ancestral

RUSSELL: AMERICAN MOSASAURS 201

platynotans in the fusion of the frontals, the development of a large supra-
stapedial process of the quadrate, and the highly developed articulation be-
tween the splenial and angular. As the mosasaurian grade was attained the
nasals were reduced and separated from each other by the internarial bar of the
premaxilla and frontal. The palatine teeth were lost as was the posteromedial
process of the coronoid to the prearticular. The lacrymals were reduced, the
postorbitals and postfrontals were fused, and the anteromedial portion of the
parietal lost its contact posteriorly with the postfrontal. The intercentral artic-

Text-fig. 98. Restored skulls. A. Clidastes liodontus (after YPM 1335). B. Opetiosaurus bucchichi
(after Kornhuber, 1901, pls. 1-2). C. Hypothetical primitive platynotan. All are drawn to
approximately the same length, showing the suggested transition from a primitive varanoid
to the mosasaurian type of skull.

902 PEABODY MUSEUM BULLETIN 23

PLOTOSAURUS
BENNISONI
\ TANIWHASAURUS yi
MOSASAURUS MAXIMUS - PLOTOSAURUS
HOFFMANNI TUCKERI
OSA RUS
CSOSA \ LIODON
MISSOURIENSIS
MOSASAURUS |
CONODON
MOSASAURUS COMPRESSIDENS
IVOENSIS

CLIDASTES PROPYTHON —
Reese)

CLIDASTES

LIODONTUS
GLOBIDENS

CLIDASTES
STERNBERGI

AMPHEKEPUBIS

CLIDASTES~LIKE
ANCESTOR

Text-fig. 99. Diagrammatic family tree of the mosasaurs,

ulations became vertical and the tail was changed into a sculling organ, losing
the transverse processes on the vertebrae of the distal half of the caudal series.
There are indications of an increase in the number of long ribs with cartilaginous
sternal connections and of posterodorsal vertebrae lacking long ribs. Mosasaurs
may have descended from aigialosaurs through primitive forms similar to
Clidastes sternbergi.

PHYLOGENY OF MOSASAURS

As has been shown, the mosasaur skeleton could easily have been derived
from that of an aigialosaur. Typical aigialosaurs are, however, known only
from the Cenomanian-Turonian of the Balkans, and the record of mosasaurs
clearly begins with the succeeding Coniacian stage. Although these aigialosaurs
may or may not have given rise to true mosasaurs, as here considered, it is well
to remember that Proaigialosaurus from the late Jurassic indicates that aigia-
losaurs were surely present during earlier Cretaceous time.

\

|
AIGIALO

RUSSELL: AMERICAN MOSASAURS 203

\ /
PLIOPLATECARPUS PLESIOTYLOSAURUS HALISAURUS
DEPRESSUS f PLATYSPONDYLUS
PROGNATHODON /
\ RAPAX HAINOSAURUS
PLIOPLATECARPUS .
MARSHI DOLLOSAURUS PROGNATHODON
| OVERTONI
PLIOPLATECARPUS ']
ATH
HOUZEAUI PROGNATHODON
CRASSARTUS
\ TYLOSAURUS
PLIOPLATECARP PROR
Pee ‘i us PLATECARPUS LGER
IMAE
vUS SOMENENSIS
\
PLATECARPUS
“PLATECARPUS" ee
T
INTERMEDIUS ERICUS
y | HALISAURUS TYLOSAURUS
ONCHOGNATUS NEPAEOLICUS
SOTERGER PLATECARPUS
SUE CORYPHAEUS

PLATECARPUS=— LIKE
ANCESTOR

SAURIDS

Probably sometime during the Cenomanian two basic types of mosasaurs
separated off from somewhat advanced aigialosaur stock. In one type, represented
by the Plioplatecarpinae and Tylosaurinae, the number of marginal teeth is
slightly reduced, the dorsal edge of the surangular is rounded and the lateral
exits for cranial nerves X-XII fused into a single foramen externally. With the
possible exception of Hainosaurus there are never more than 29 presacral verte-
brae, and this section of the column is less than the caudal series in length. ‘The
tail is not especially dilated distally and the articular surfaces of the limbs were
finished in thick cartilage. These two subfamilies are obviously closely related
but they soon developed divergent characters in the cranium. In one (the Tylo-
saurinae) the otosphenoid crest became very large and a long conical rostrum
appeared in front of the teeth on the premaxilla. In the other (the Plioplate-
carpinae) a large channel or groove for the basilar artery developed in the
basisphenoid-basioccipital, and the suprastapedial process of the quadrate
curved ventrally to reach nearly to the base of the quadrate.

204 PEABODY MUSEUM BULLETIN 23

The tylosaurines manifest relatively little specific diversity and may have
disappeared on this continent before the beginning of Maestrichtian time, al-
though some forms survive into the late Maestrichtian of Europe (Hainosaurus)
and New Zealand. Generalized Plioplatecarpines appear in abundance from
the Santonian through Campanian in the United States and probably also in
Europe. Known genera include Ectenosaurus, Platecarpus and Plioplatecarpus,
the last genus continuing until the late Maestrichtian of North America and
Europe. In the late Campanian and Maestrichtian these forms were joined by
plioplatecarpines of the Prognathodon type (Dollosaurus, Plesiotylosaurus, Prog-
nathodon).

The Mosasaurinae represent derivatives of the other basic stock of mosasaurs,
and Clidastes may be considered as morphologically close to the ancestral type of
the subfamily. The number of marginal teeth in the jaws of members of this
group is greater than in the Plioplatecarpinae- Tylosaurinae. The dorsal edge of
the surangular is a thin ala of bone that rises anterodorsally to buttress the
posterior end of the coronoid. Cranial nerves X-XII consistently exit through
two foramina on the lateral wall of the braincase. There are never less than 31
presacral vertebrae and usually more than 40 are present, since this portion of the
column exceeds the caudal series in length. The neural and haemal spines are
elongated near the center of the tail and support a well-formed caudal fin. The
individual elements of the limbs articulate with a minimum of intervening
cartilage.

Mosasaurus and Liodon are large typical mosasaurines, ranging from the
Santonian to Maestrichtian of North America and Europe. Plotosaurus from
western North America and Taniwhasaurus from New Zealand may be con-
sidered as late, highly advanced Clidastes derivatives. Possibly in the latest
Cenomanian a type of clidastoid acquired shellfish-eating habits and developed
into Globidens, the characteristic spherical teeth of which have been found in
sediments of later Campanian and/or Maestrichtian age on five continents.
Another less divergent shellfish-eating form of more delicate proportions is
known from the Maestrichtian of Europe (Compressidens).

EVOLUTION OF MOSASAURS

In spite of the great numbers of mosasaur specimens collected from the late
Cretaceous deposits of the world and the detail in which their osteology has
been studied, several difficulties are encountered when an attempt is made to
discern evolutionary trends within the group. The early Cretaceous record of
reptilian life is as yet rather sparsely documented and this is especially true for
platynotan lizards. As a consequence new discoveries may greatly modify any
suggestions offered here on the mode of derivation and early radiation of mosa-
saurs. Even the relatively excellent record of late Cretaceous mosasaurs presents
difficulties. Although late Santonian, early Campanian and late Maestrichtian
forms are well known, those from rocks of an intermediate age are not, and even
good faunas from the former two ages have been described only from the United
States and Belgium. As it stands, the well-preserved portion of the mosasaur
record covers only about 20 million years, which allows little time for the de-
velopment of adaptive grades such as have been demonstrated in the teleost
fishes. The following remarks must be considered in the light of these limitations.

The reason behind the sudden appearance and rapid proliferation of
mosasaurs must have lain in the exploitation of an unoccupied or inefficiently
occupied zone. All the evidence indicates that mosasaurs were near-surface

RUSSELL: AMERICAN MOSASAURS 205

carnivores and actively pursued their prey. The remarkable convergence of some
of the later mosasaurs (Plotosaurus, Camp, 1942, pl. 5) with primitive ‘Triassic
ichthyosaurs (Mixosaurus, Huene, 1925, fig. 3) shows that they were an ecological
replacement of the declining ichthyosaurs. Whatever the habits of plesiosaurs
were, there must have been little direct competition with mosasaurs, for mosa-
saurs were able to enter and thrive in a marine environment and both groups
inhabited the same seas until the close of Maestrichtian time. Perhaps plesiosaurs
subsisted on fish living closer to the bottom of shallow marine bodies of water.

Aigialosaurs possess no character that would prevent them from being ances-
tral to the mosasaurs. As in other forms that bridge the morphologic gap be-
tween two categories (DeBeer, 1954), aigialosaurs are not simply intermediate
between varanoids and mosasaurs but present a ‘“‘mosaic’”’ of ancestral (varanoid)
and descendent (mosasauroid) characters (characters possessed by aigialosaurs
are italicized):

VARANOID CHARACTERS MOSASAUROID CHARACTERS

A. external nares located far A. external nares located nearer
anteriorly on rostrum. center of rostrum

B. incipient suprastapedial B. large suprastapedial process
process on quadrate on quadrate

C. splenio-angular joint C. splenio-angular joint well-
imperfect developed

D. two sacral vertebrae present D. sacrum absent

E. tail tapers gradually toa E. tail comes more abruptly to a
sharp point rounded termination in lateral

profile
F. terrestrial limbs F. limbs modified into paddles

Aigialosaurs approach mosasaurs more closely in the skull than in the post-
cranial skeleton. The highly streptostylate quadrates and splenio-angular artic-
ulation apparently gave the mandibles sufficient mobility to force objects into
the throat without the aid of gravity or some solid point of leverage. This
adaptation, which freed aigialosaurs from the need of surfacing or climbing out
upon the land to swallow their prey, was probably critical in enabling them to
become truly pelagic carnivores. As a consequence the body and limbs were later
modified into more effective organs of aquatic locomotion. With this last de-
velopment the mosasaurian level of evolution was attained.

The radiation of mosasaurs into two distinct lineages by Santonian time sug-
gests the presence of two subzones within the general adaptive zone occupied by
the mosasaurs. That the clidastoids were shallow-water forms is strongly sug-
gested by their evolutionary proximity to the aigialosaurs, their abundance in
portions of the Niobrara Chalk that were apparently deposited in shallower
water, and by the fact that shellfish-eating forms (Compressidens, Globidens)
were derived from them and not from the remaining groups of mosasaurs. Plio-
platecarpines and tylosaurines separated at an early point in their history and
may represent an invasion of mosasaurs into the surface regions of deeper wa-
ters. In Platecarpus, a generalized plioplatecarpine, the body appears to have
been very maneuverable and the animal is assumed to have fed on small fish.
The large size and heavy marginal dentition of Tylosaurus suggest that it preyed
upon large fish and marine reptiles.

By the latter half of the Campanian the composition of the mosasaur faunas

206 PEABODY MUSEUM BULLETIN 23

of North America had changed. The tylosaurines were evidently replaced by
ecologically analogous large mosasaurines (Mosasaurus, Liodon), and smaller,
advanced clidastoids (Plotosaurus, Taniwhasaurus) inhabited the Pacific basin.
The genus Platecarpus had given rise to belemnite-eating (Plioplatecarpus) and
possibly ammonite-eating forms (prognathodonts).

There are indications that these later Campanian reptiles represented a
slightly higher adaptive level in the evolution of mosasaurs. There is an average
increase in size over pre-late Campanian forms and suggestions of telescoping in
some of the cranial elements (although not nearly approaching the condition
seen in cetaceans). For example, tongues of the frontal spread over the antero-
dorsal portion of the parietal in Mosasaurus, Plotosaurus, Prognathodon and
Plesiotylosaurus, and the surangular extended over the dorsolateral edge of the
articular in Mosasaurus and Plotosaurus. A high degree of streptostyly is pre-
served in all mosasaurs, as is the splenio-angular articulation, but in most of
the genera characteristic of the late Campanian-Maestrichtian (Mosasaurus,
Plotosaurus, Prognathodon, Plesiotylosaurus) cranial kinesis was limited or lost,
in contrast to earlier forms. Interestingly, there is no marked tendancy for the
external nares to migrate posteriorly in these late forms. There is a widespread
tendency to increase the number of pygal vertebrae (M. conodon, Plotosaurus,
Plioplatecarpus), and in members of the Mosasaurinae the number of presacral
vertebrae is also generally increased with time. The digits of M. conodon and
Plotosaurus tend to be pressed together into a flipper rather than divaricate and
supporting a web, as in earlier mosasaurs (see Camp, 1942, p. 17), and the in-
dividual phalangeal elements are shorter. The humerus is more massively pro-
portioned and the scapula is relatively larger in Mosasaurus, Plotosaurus, Plio-
platecarpus and Plesiotylosaurus.

The present state of our knowledge of the general natural history of the
reptiles which form the subject of this review may be summarized as follows.
Between 100 and 63 million years ago, during the concluding phases of the Age
of Reptiles, a group of giant marine lizards lived in great abundance in the
shallow tropical to subtropical seas then covering vast areas of the North
American continent. Adult animals, depending on the particular species, ranged
in length from six to over fifty feet. Their heads were conical and their bodies
were elongate, attaining a maximum diameter in the chest region. The tail was
a laterally compressed sculling organ. The dorsal portions of their bodies were
probably dark with a reddish or silverish tint and the underparts were lighter.
As in many kinds of vertebrates, the smaller species or fry living close to the
shore may have been marked with brighter color patterns tending to obscure
the outline of the body.

These reptiles were powerful swimmers, guiding their progression by vertical
and rotational movements of the paddles and turning movements of the neck.
They were generally voracious carnivores. The severe wounds occasionally in-
flicted on others of their kind bear witness to a dumb, extremely savage dis-
position. They roamed surficial waters relying on their keen sense of sight and
hearing for locating their prey. Medium-sized fish formed the preponderant
element in the diet of many of these reptiles, but various species were especially
adapted for feeding on shallow water benthonic echinoderms or pelecypods,
small squid-like cephalopods, shelled cephalopods, or larger aquatic vertebrates

RUSSELL: AMERICAN MOSASAURS 207

such as turtles and large fish. An additional articulation between the upper
and lower jaws at the back of the head enabled them to dismember and swallow
prey more effectively under water. Their breeding habits remain conjectural,
although it is probable that the eggs were laid in subaerial nests of some kind.

These reptiles descended from small semi-aquatic lizards of a normal lacerti-
form appearance. Two ancestral stocks quickly radiated into a profusion of dif-
ferent lineages. Generally members of these lineages improved in aquatic
adaptation and increased in size throughout the brief geologic duration of the
group. The initial diversity and abundance of these reptiles were maintained
until they suddenly vanished along with many other groups at the close of the
Age of Reptiles. Their extinction was most likely due to a failure in food sup-
ply, as they had no obvious competitors, but how this was related to the con-
temporaneous causes of extinction of aquatic and terrestrial organisms is un-
known.

APPENDIX A MEASUREMENTS

TABLE, 1
MEASUREMENTS OF REPRESENTATIVE MOSASAUR SKULLS

The specimens within each species of mosasaur are arranged according to size from
the smallest to the largest. The same is true of the species, so that Clidastes sternbergt
possesses on the average the smallest known skull and Tylosaurus proriger the largest.
All measurements are in millimeters, a small ‘a’ preceding the number indicates that
it is approximate. The distance between the first and sixth tooth in a dentition is
measured from the anterior base of the first tooth to the posterior base of the sixth. The
length of the dentary is measured from its anterior tip to the posterior apex of the

alveolar margin of the bone.

A. Length of skull along midline. B. Length of premaxillary rostrum. C. Width of frontal be-
tween orbits. D. Length between first and sixth maxillary tooth. E. Height of quadrate. F. Length
of lower jaw. G. Length of dentary. H. Length between first and sixth dentary tooth.

Clidastes sternbergi
Type (Wiman 1920)
USNM 3777

Clidastes liodontus
USMN 11719
USNM 6546
YPM 1335
YPM 24914
YPM 1333
USNM 11647

Plotosaurus bennisont
Type (Camp 1942)

Clidastes propython
YPM 1368
YPM 1318
YPM 1319
KU 1000

“Platecarpus”’ intermedius
ANSP 9023

Platecarpus coryphaeus
MCZ 1610
YPM 4004
YPM 3972

Platecarpus ictericus
YPM 1256
YPM 1277
YPM 4003
YPM 3690
AMNH 1820
AMNH 1821

A

B

G

D

38

E

428

RUSSELL: AMERICAN MOSASAURS

YPM 1258

Globidens alabamaensis
USNM 6527

Plotosaurus tuckeri
(Camp 1942)

Ectenosaurus clidastoides
FHM 7937

Plactecarpus cf. P. somenensis

CNHM PR 466
CNHM PR 467
CNHM PR 465

Prognathodon overtont
SDSM 3393
KU 950

Prognathodon rapax
NJGS 9827

Mosasaurus ivoensis
KU 1024

Tylosaurus nepaeolicus
YPM 3979
FHM 1654
YPM 3992
YPM 3974
YPM 4000
AMNH 124, 134
YPM 3980
YPM 3976

Mosasaurus missouriensis
Type (Goldfuss 1845
and Harlan 1834c)
KU 1034

Plesiotylosaurus crassidens
(Camp 1942)

Mosasaurus maximus
NJSM 11053
YPM 773

Tylosaurus proriger
AMNH 1585
YPM 1268
USNM 6086
AMNH 4909
YPM 3977
USNM 8898
YPM 4002
FHM 4
KU 1032
YPM 3981

A

588

B

G
a99

135

148

83

D
130

120

E
94

284

210 PEABODY MUSEUM BULLETIN 23

TABLE 2
BODILY PROPORTIONS OF MOSASAURS

HEAD NECK TRUNK TAIL OVERALL LENGTH

Varanus (YPM Osteology 4568)
6.85% 8.65% 24.10% 60.50% 1.94m = 6 1/3 ft.

Clidastes sternbergi (Wiman, 1920, pl. 3)

13.2% *43.2% 43.5% 2.65m = 83/4 ft.
Clidastes liodontus (Williston, 1898b, p. 143)

1231% 6.5% 39.2% 42.2% 3.465m = 11 1/2 ft.
Platecarpus (Huene, 1911, p. 49)

9.8% 5.7% 31.0% 53.5% 5.60m = 18 1/3 ft.
Platecarpus (Lambe, 1914)

10.4% *33.5% 56.1% 6.25m = 20 1/2 ft.
Tylosaurus (Williston, 1898b, p. 143)

13% 5.6% 31.5% 49.9% 6.341m = 20 3/4 ft.
Tylosaurus proriger (Osborn, 1899a, p. 170)

13.8% 6.9% 27.6% 51.6% 8.83m = 29 ft.

* includes combined length of neck and trunk

TABLE 3
ADDITIONAL ESTIMATED LENGTHS OF MOSASAURS

Glidasiess( WAP MG4 813) hy uvenileanrec)tneineteiece Gcincimeicicae ceiseiecieet eee 1.22m) = 45
Mosasaurusiconodon' (Dollo, 1894; p: 248) \2)<< «+12. i-le-s1> oie a\e1-solevelsisielovore) ieletetelele im) =) 23s
Mosasaurus:conodon (Molloy 194: 7inp. 19) eieryetevercustaeccvele ro sietetsisroreneraiensterereyeiete 10m = 32.8 ft.
Mosasaurus hofmanny (Dollon 19 li ype lid) eieteteious |< ercistereicietarsioielotere eiereeicteiente 15m = 49 ft.
Mosasaurus hoffmanni (YPM plastotype) given length of jaw equals 10%

ofibodyilengthy:s;, e.<ces.icreteortrs crated wide eevsraiersre eee eteteceats chats 12.4m = 40.66 ft.
Mosasaurus maximus (NJSM 11053) given length of jaw equals 10%

of body length. yaxct civatecoorse arene erniere e cteieletieielerseorereicietske ereretore 12.4m = 40.66 ft.
Plioplatecarpus houzeaui (Dollo, 1894, p. 250). ........ 2c eee cece ee ee eee 5m = 16.33 ft.
Phoplatecarpus houzeaut (Dollo, 1917, os 19) ye are ewes icioleleverereie o7o cietelvieisyelere 6m = 19.5 ft.
Prognathodon soluays:(Dollo; 191'7,.p) 20) ns sc .c1ereiei ere oe 2 seis @\ ele «1s oisisieleerersier ote 6m = 19.5 ft.
Prognaihodon’ giganteus. (Dollo; 19044) ps 213) yee ers:cretereseicie oe a eialelelele eles ciel 10m = 32.8 ft.
Hammosaurus bernardt (Dollo, 1917, p.19)\seers weitere cieiers s\ele elsleielsielele erseletetotons 17m = 56 ft.

APPENDIX B
LOCALITIES OF PIERRE MOSASAURS

Because scattered collecting indicates that mosasaurs are probably abundant and
diversified in the Pierre Shale, and the formation appears to contain distinctive mosasaur
associations at different levels, a list of mosasaur species and their localities is appended
to encourage future collecting.

RUSSELL: AMERICAN MOSASAURS 211

I. Mosasaurs from the lower Pierre (Sharon Springs, Gregory and Crow Creek Members
and equivalents):
Clidastes propython
AMNH 5812, from bluff near A. Finney’s ranch, Cottonwood Creek, SSW of Edge-
mont, Fall River County, South Dakota. Collected by B. Brown, 1903.
KU 1026, (type of C. westi), from near McAllister, Wallace County, Kansas.

Globidens alabamaensts
SDSM 4612, from 2 miles S of Buffalo Gap, Fall River County, South Dakota, low

in Pierre. Collected by J. D. Bump.

Platecarpus ictericus

AMNH 2182, from N half section 7, T38N, R60W, Niobrara County, Wyoming, near
base of Pierre. Collected by Bobb Schaeffer and Walter Sorenson.

AMNH 5811, from in and near pasture of Mr. J. Doo, N and middle Alkali Creek,
tributaries of Indian Creek, 18 miles SSW of Edgemont, Fall River County, South
Dakota. Collected by B. Brown.

USNM 18256, from Missouri River valley near Yankton, Yankton County, South
Dakota, from the Sharon Springs Member. Collected by Simpson, 1948.

KU 1026 (mixed with the type of Clidastes “westi’’), same locality as given for KU
1026 above.

KU 1081, from Wallace County, Kansas. Collected by G. P. Crazer.

KU 1093, from the Pierre, no locality data.

SDSM 30139, from 26 miles W of Edgemont and 1 mile N of Red Bird, Niobrara
County, Wyoming. Collected by J. D. Bump and H. Martin.

Platecarpus cf. P. somenensts
CNHM nos. PR 465, 466, 467, from W side NW quarter, SW quarter section 29,
T5N, RSE, Custer County, South Dakota, from Sharon Springs Member. Collected
by John Clark.

Platecarpus, unidentified species

AMNH 1570, from 14 miles S of Wallace, Wallace County, Kansas (the specimen
must have come from SE of Wallace as there is no Pierre exposed 14 miles due
S of Wallace).

AC nos. 389, 398 (type and referred specimen of Platecarpus brachycephalus), from
the head of Mule Creek, 20 miles due W of Edgemont, Niobrara County,
Wyoming.

SDSM 4910, from 214 miles E of Provo, Fall River County, South Dakota. Collected
by J. D. Bump and H. Martin.

CNHM nos. PR 215, PR 216, from 8 miles S of Fairburn, Custer County, South
Dakota. Collected by G. Langford, 1946.

Tylosaurus proriger

AMNH 7220, from SW quarter, SE quarter section 36, T29S, R65W near Walsen-
berg, Colorado. One-third mile NW boundary between Huerfano and Las Animas
Counties. From top of Niobrara very near Pierre contact. Collected by E. H. Col-
bert.

USNM__ 17909, from NE quarter, SE quarter section 36, T2S, R33E, Bighorn County,
Montana, from Telegraph Creek Formation. Collected by C. H. Rogers, 1947.

KU 5033, from 21% miles S of Wallace on Pinnell Ranch, Wallace County, Kansas.
Collected by C. D. Bunker, 1911.

CNHM 820, from Pierre Formation, South Dakota. Matrix looks like Sharon
Springs Member.

II. Mosasaurs from the middle Pierre (DeGrey and Verendrye Members and equivalents):

Mosasaurus conodon
USNM 18255, from E center section 28, T6N, R8OW, Hughes County, South

212 PEABODY MUSEUM BULLETIN 23

Dakota, from the DeGrey or Verendrye Member. Collected by D. R. Crandell,
1948.

SDSM 452, from NE quarter, NE quarter section 6, T127N, R64W, spillway of Elm
Lake 10 miles W of Fredrick, Brown County, South Dakota, possibly from the
Virgin Creek Member.

Plioplatecarpus primaevus
USNM 18254, from NE corner section 35, T112N, R80W, near Pierre, Hughes

County, South Dakota, from the DeGrey Member. Collected by D. R. Crandell,
1948.

III. Mosasaurs from the upper Pierre (Virgin Creek, Mobridge and Elk Butte Members
and equivalents):

Mosasaurus missouriensis

The type specimen was collected near the Big Bend of the Missouri River, between
Fort Lookout and Fort Pierre.

USNM 8086, from NW quarter, NW quarter section 33, TI9N, R21E, Feigns County,
35 miles N of Lewistown, Montana, from the Bearpaw Shale.

USNM 4910, from the Pierre Shale of South Dakota. Collected by J. B. Hatcher.

KU 1034 (type of M. horridus), from the Pierre Shale along the Cheyenne River,
Custer County, South Dakota.

Prognathodon overtoni
KU 950, from Pierre Shale along the Cheyenne River, 100 feet above the level of
KU 1034, Custer County, South Dakota.
SDSM 3393, from 6 miles SW Cuny Table, on Mule Creek, Shannon County, South
Dakota, from the Virgin Creek Member.

LITERATURE CITED

Abel, Othenio. 1922. Lebensbilder aus der Tierwelt der Vorzeit. Jena. 643 p.

1924. Die Eroberungsziige der Wirbeltiere in die Meere der Vorzeit. Jena. 121 p.

Adkins, Walter S. 1928. Handbook of Texas Cretaceous fossils. Univ. Texas Bull. 1838.
303 p.

Ali, Syed Muzammil. 1950. Studies on the anatomy of the tail in sauria and rhynchoce-
phalia part 3 Varanus monitor. Indian Acad. Sci. Proc. 30: 155-167.

Ameghino, Florentino. 1918. Nouvelles découvertes paléontologiques dans la Patagonie
australe in Ameghino, F., Obras completas. Buenos Aires. 10: 530-533.

Anderson, F. M. 1943. Synopsis of the later Mesozoic in California. Calif. Dept. Nat. Res.
Bull. Div. Mines, 118: 183-186.

1958. Upper Cretaceous of the Pacific Coast. Geol. Soc. Amer. Mem. 71. 378 p.

Anonymous. 1962. Reptiles of Saskatchewan’s ancient seas. Saskatchewan Mus. Nat. Hist.,
popular ser. 1. 12 p.

Arambourg, Camille. 1952. Les vertébrés fossiles des gisements de phosphates (Maroc-
Algerie-Tunisie). Serv. Géol. Maroc Notes Mém. 92. 372 p.

Avnimelech, M. 1949. On vertebrate remains in Senonian phosphate beds in Trans-
jordan. Ecol. Geol. Helvetiae, 42: 486-490.

Bahl, K. N. 1937. Skull of Varanus monitor. Indian Mus. Rec. 39: 133-174.

Bainbridge, Richard. 1961. Problems of fish locomotion. Zool. Soc. London Symp. 5:
13-32.

Ballou, W. H. 1898. The serpent-like sea saurians. Popular Sci. Monthly 53: 209-225.

Baur, G. H. C. L. 1890. On the characters and systematic position of the large sea-
lizards, Mosasauridae. Science 16 (405): 262.

1892. On the morphology of the skull in the Mosasauridae. Jour. Morph. 7(1):

1-22.

1895. Cope on the temporal part of the skull, and on the systematic position of
the Mosasauridae, a reply. Amer. Nat. 29: 998-1002.

RUSSELL: AMERICAN MOSASAURS 213

1896. The paroccipital of the Squamata and the affinities of the Mosasauridae
once more. A rejoinder to Professor E. D. Cope. Amer. Nat. 30: 143-147.

Bergquist, Harlan R., and William A. Cobban. 1957. Mollusks of the Cretaceous in
Ladd, Harry S., Treatise on Marine ecology and paleoecology. Geol. Soc. Amer.
Mem. 67: 871-884.

Bertin, Léon. 1958. Ecologie in Grassé, Pierre, Traité de Zoologie. 13(3): 1885-1933.

Blainville, H. M. Ducrotay de. 1827. Memoire sur les Belemnites, considerés zoologique-
ment et géologiquement. Paris.

Bogoliubov, Nikolai N. 1910. Sur les restes des mosasauriens trouvés dans le gouvernement
d’Orenbourg. Ezhegodnik Geol. Min. Rossi 12:8-14.

Boulenger, G. A. 1891. Notes on the osteology of Heloderma horridum and H. suspectum,
with remarks on the systematic position of the Helodermatidae and on the vertebrae
of the lacertilia. Zool. Soc. London Proc. (1891) : 109-118.

1893. On some newly-described Jurassic and Cretaceous lizards and rhynchoce-
phalians. Nat. Hist. Ann. 6th ser. 11: 204-210.

Breder, C. M., Jr. 1926. The locomotion of fishes. Zoologica 4: 159-297.

Bronn, H. G. 1838. Lethea geognostica. Stuttgart 2: 545-1346.

Broom, Robert. 1912. On a species of Tylosaurus from the upper Cretaceous of Pondo-
land. South African Mus. Ann. 7: 332-333.

1913. On the squamosal and related bones in mosasaurs and lizards. Amer. Mus.
Nat. Hist. Bull. 32: 507-508.

Camp, Charles L. 1923. Classification of the lizards. Amer. Mus. Nat. Hist. Bull. 48:
289-481.

1942. California mosasaurs. Univ. Calif. Mem. 13. 68 p.

1951. Plotosaurus, a new generic name for Kolposaurus Camp, preoccupied. Jour.
Paleont. 25(6): 822.

Camp, Charles L. and H. J. Allison. 1961. Bibliography of fossil vertebrates 1949-1953.
Geol. Soc. Amer. Mem. 84. 532 p.

Capps, S. R., Jr. 1907. The girdles and hind limb of Holosaurus abruptus Marsh. Jour.
Geol. 15: 350-356.

Chaffee, Robert G. 1939. A New Jersey mosasaur of the subfamily Platecarpinae. Acad.
Nat. Sci. Philadelphia Notulae Nat. 37: 1-5.

Cobban, William A., and John B. Reeside, Jr. 1952. Correlation of the Cretaceous
formations of the western interior of the United States. Geol. Soc. Amer. Bull. 63:
1011-1044.

Conant, Louis C. 1964. General remarks on the pre-Selma Cretaceous strata of western
Alabama in Monroe, W. H. et al., Studies of pre-Selma core samples from the outcrop
area in western Alabama. U. S. Geol. Surv. Bull. 1160: 97-101.

Conybeare, William D. 1822. Jn Parkinson, J., An introduction to the study of fossil
organic remains. First ed. London, p. 298.

Cope, Edward Drinker. 1868a. (Remarks on Clidastes iguanavus, Nectoportheus validus
and Elasmosaurus.) Acad. Nat. Sci. Philadelphia Proc. 20: 181.

1868b. Synopsis of the extinct reptilia found in the Mesozoic and Tertiary

strata of New Jersey in Cook, George H., Geology of New Jersey. Newark, app.

B: 733-742.

1869a. On some Cretaceous reptilia. Acad. Nat. Sci. Philadelphia Proc. 20:233-242.

1869b. On the reptilian orders Pythonomorpha and Streptosauria. Boston Soc.

Nat. Hist. Proc. 12: 250-266.

1869c. The fossil reptiles of New Jersey. Amer. Nat. 3: 84-91.

1869d. (Remarks on Holops brevispinus, Ornithotarsus immanis and Macro-

saurus proriger.) Acad. Nat. Sci. Philadelphia Proc. 21: 123.

1869e. (Remarks on fossil reptiles.) Amer. Phil. Soc. Proc. 11: 16.

1869f. (Remarks on fossil reptiles, Clidastes propython, Polycotylus latipinnus,

Ornithotarsus immanis.) Amer. Phil. Soc. Proc. 11: 117.

1869g. Specimens of extinct reptiles. Nature 1: 121-122.

1869-1870. Synopsis of the extinct Batrachia, Reptilia, and Aves of North Amer-

214 PEABODY MUSEUM BULLETIN 23

ica. Amer. Phil. Soc. Trans., n.s., issued in parts: 1 (1869), p. 1-105; 2 (1870), p. 106-
235; 3 (1870), p. i-vii, 236-252.

1870a. On some reptilia of the Cretaceous formation of the United States. Amer.
Phil. Soc. Proc. 11: 271-274.

1870b. (Statements communicated regarding Liodon perlatus.) Amer. Phil. Soc.
Proc. 11: 497.

1871a. (Verbal communication on pythonomorphs.) Amer. Phil. Soc. Proc. 11:
571-572.

1871b. On some species of Pythonomorpha from the Cretaceous beds of Kansas
and New Mexico. Amer. Phil. Soc. Proc. 11: 574-584.

1871c. (Observations on fossil reptiles from Kansas.) Acad. Nat. Sci. Philadelphia
Proc. 22: 132.

1871d. Supplement to the “Synopsis of the extinct Batrachia and Reptilia of
North America.”” Amer. Phil. Soc. Proc. 12: 41-52.

187le. On the homologies of some of the cranial bones of the reptilia, and on
the systematic arrangement of the class. Amer. Assoc. Ady. Sci. Proc. 19th meeting,
Troy, 1870: 194-247.

1871f. On the fossil reptiles and fishes of the Cretaceous rocks of Kansas. U. S.
Geol. Surv. of Wyoming and portions of contiguous territories, 2nd (4th) ann. rep.
F. V. Hayden, U. S. Geologist, Washington, p. 385-424.

1872a. (Remarks on new fossil reptiles from western Kansas.) Acad. Nat. Sci.
Philadelphia Proc. 23: 297-298.

1872b. Note of some Cretaceous Vertebrata in the State Agricultural College of
Kansas, U.S.A. Amer. Phil. Soc. Proc. 12: 168-170.

1872c. (Letter to Professor Lesley giving an account of a journey in the valley
of the Smoky Hill River, in Kansas.) Amer. Phil. Soc. Proc. 12: 172-174.

1872d. Catalogue of the Pythonomorpha found in the Cretaceous strata of
Kansas. Amer. Phil. Soc. Proc. 12: 264-287.

1872e. (Remarks on discoveries recently made by Prof. O. CG. Marsh.) Acad.
Nat. Sci. Philadelphia Proc. 24: 140-141.

1872f. On the geology and palaeontology of the Cretaceous strata of Kansas.
U. S. Geol. Surv. of Montana and portions of adjacent territories, 5th ann. rep. F. V.
Hayden, U. S. Geologist, Washington, p. 318-349.

1874. Review of the Vertebrata of the Cretaceous period found west of the
Mississippi River. U. S. Geol. Surv. Terr. Bull. 1(2): 3-48.

1875. The Vertebrata of the Cretaceous formations of the West. U. S. Geol. Surv.
Terr. Rep. 2: 1-303.

1877. Report on the geology of the region of the Judith River, Montana, and on
vertebrate fossils obtained on or near the Missouri River. U. S. Geol. Geogr. Surv.
Bull. 3(3): 565-597.

1878. Professor Owen on the Pythonomorpha. U. S. Geol. Geogr. Surv. Bull.
4: 299-311.

1879a. The scales of Liodon. Amer. Nat. 13: 132.

1879b. The relations of the horizons of extinct Vertebrata of Europe and
America. U.S. Geol. Geogr. Surv. Terr. Bull. 5: 33-54.

1881. A new Clidastes from New Jersey. Amer. Nat. 15: 587-588.

1889. Synopsis of the families of Vertebrata. Amer. Nat. 23: 849-877.

——— 1891. Syllabus of lectures on geology and paleontology. Philadelphia. 90 p.
1895a. Baur on the temporal part of the skull, and on the morphology of the
skull in the Mosasauridae. Amer. Nat. 29: 855-859.

1895b. Reply to Dr. Baur’s critique of my paper on the paroccipital bone of the
scaled reptiles and the systematic position of the Pythonomorpha. Amer. Nat. 29:
1003-1005.

1896. Criticism of Dr. Baur’s rejoinder on the homologies of the paroccipital
bone. Amer. Nat. 30: 147-149.

1898. Syllabus of lectures on the Vertebrata. Philadelphia. 135 p.

RUSSELL: AMERICAN MOSASAURS 215

Corroy, Georges. 1927. Sur la présence d’ossements de mosasaure dans le Santonien de
l’Aude. Soc. Géol. France Bull. 4° ser. 27, Compt. Rendus: 135-137.

Cox, L. R. 1962. British Mesozoic fossils. London. 205 p.

Crandell, D. R. 1950. Revision of Pierre Shale of central South Dakota. Amer. Assoc.
Petrol. Geol. Bull. 34(12): 2337-2346.

1958. Geology of the Pierre area, South Dakota. U. S. Geol. Surv. Prof. Paper 307.

83 p.

Cuvier, G. C. F. 1808. Sur le grand animal fossile des carriéres de Maestricht. Mus. Hist.
Nat. Paris Ann. 12: 145-176.

1834-1836. Recherches sur les ossements fossiles. 4° ed. Paris. 10 vols.

Dane, Carle H. 1929. The upper Cretaceous formations of southwestern Arkansas.
Arkansas Geol. Surv. Bull. 1. 215 p.

DeBeer, Gavin R. 1954. Archaeopteryx and evolution. Adv. Sci. 42: 160-170.

DeKay, J. E. 1830. On the remains of extinct reptiles of the genera Mosasaurus and
Geosaurus found in the Secondary formation of New-Jersey. Lyc. Nat. Hist. New York
Ann. 3: 134-141.

1842. Zoology of New York. Pt. 3, reptiles and amphibia. Albany. 98 p.

De La Beche, Henry T. 1849. Anniversary address before the Geological Society of
London. Geol. Soc. London Quart. Jour. 5: xix-cxvi.

Depéret, Charles, and P. Russo. 1925. Les phosphates de Melgou (Maroc) et leur faune
de mosasauriens et de crocodiliens. Soc. Géol. France Bull. 4° ser. 25: 329-346.

Devillers, Charles. 1943. Nerfs craniens et circulation céphalique de Plioplatecarpus
marshi. Ann. Paleont. 30: 47-59.

Dollo, Louis. 1882. Note sur l’ostéologie des Mosasauridae. Mus. Hist. Nat. Belgique
Bull. 1: 55-80.

1884. Le mosasaure. Rev. Questions Sci. I® ser. 16: 648-653.

1885a. Le hainosaure. Rev. Questions Sci. 1° ser. 18: 285-289.

1885b. Notes d’ostéologie erpétologique. Soc. Sci. Bruxelles Ann. 9(2): 309-338.

1885c. Premiére note le hainosaure. Mus. Hist. Nat. Belgique Bull. 4: 25-35.

1887. Le hainosaure et les nouveaux vertébrés fossiles du Musée de Bruxelles.

Rey. Questions Sci. 1° ser. 22: 70-112.

1888. Sur le crane des mosasauriens. Sci. France Belgique Bull. 19:1-11.

1889a. Nouvelle note sur les vertébrés fossiles recémment offerts au Musée de

Bruxelles par M. Alfred Lemonnier. Soc. Belge Géol. Procés-Verbaux 3: 214-215.

1889b. Premiére note sur les mosasauriens de Mesvin. Soc. Belge Géol. Mém.

3: 271-304.

1889c. (Sur quelques mosasauriens nouveaux.) Soc. Sci. Bruxelles Ann. 13: 68-69.

1890. Premiére note sur les mosasauriens de Maestricht. Soc. Belge Géol. Mém.

4: 151-169.

1894. Nouvelle note sur l’ostéologie des mosasauriens. Soc. Belge Géol. Mém. 6:

219-259.

1903. Les ancétres des mosasauriens. Sci. France Belgique Bull. 38: 137-139.

1904a. Les mosasauriens de la Belgique. Soc. Belge Géol. Mém. 18: 207-216.

—— 1904b. L’origine des mosasauriens. Sci. Belge Géol. Mém. 18: 217-222.

1905. Un nouvel opercule tympanique de Plioplatecarpus, mosasaurien plongeur.

Sci. Belge Géol. Mém. 19: 125-131.

1909. The fossil vertebrates of Belgium. New York Acad. Sci., Ann. 19: 99-119.

1913. Globidens fraasi, mosasaurien mylodonte nouveau du Maestrichtien

(Crétacé supérieur) du Limbourg, et l’ethologie de la nutrition chez les mosasauriens.

Arch. Biol. 28: 609-626.

1917. Les vertébrés vivants et fossiles. Guide de Touring Club de Belgique 2:

126-161.

1924. Globidens alabamaensis, mosasaurien américain retrouvé dans le Craie

d'Obourg (Sénonien supérieur) du Hainaut, et les mosasauriens de la Belgique en

général. Arch. Biol. 34: 167-213.

I

216 PEABODY MUSEUM BULLETIN 23

Douglass, Earl. 1902. A Cretaceous and lower Tertiary section in south central Montana.
Amer. Phil. Soc. Proc. 41: 207-224.

Dowling, Herndon G., Jr. 1941. A new mosasaur skeleton from the Cretaceous in
Alabama. Alabama Acad. Sci., Jour. 13: 46-48.

Downs, Theodore. 1959. Prehistoric animals. Los Angeles County Museum. 20 p.

Dreverman, Fritz. 1914. Die Meersaurier im Senckenbergischen Museum. Ber. Sencken-
bergische Naturf. Gesell. 45: 35-48.

Edmund, A. Gordon. 1960. Tooth replacement phenomena in the lower vertebrates.
Roy. Ontario Mus. Life Sci. Div. Contrib. 52. 190 p.

Emmons, Ebenezer. 1858. Agriculture of the eastern counties, together with descriptions
of the fossils of the marl beds. North Carolina Geol. Surv. Rept. Raleigh. 314 p.

1860. Manual of geology, 2nd ed., New York. 297 p.

Estes, Richard. 1964. Fossil vertebrates from the late Cretaceous Lance Formation eastern
Wyoming. Univ. California Publ. Geol. Sci. 49. 180 p.

Evans, Francis G. 1939. The morphology and functional anatomy of the atlas-axis
complex from fish to mamals. New York Acad. Sci. Ann. 39(2): 29-104.

Féjérvary, G. J. De. 1918. Contributions to a monograph on fossil Varanidae and on
Megalanidae. Mus. Natl. Hungarici Ann. 16: 341-467.

Fraser, F. C., and P. E. Purves. 1960. Hearing in cetaceans. Brit. Mus. (Nat. Hist.) Bull.
7(1): 1-140.

Frazzetta, T. H. 1962. A functional consideration of cranial kinesis in lizards. Jour.
Morph. 111: 287-319.

Funnell, B. M. 1964. Studies in North Atlantic geology and palaeontology: 1. Upper
Cretaceous. Geol. Mag. 101(5): 421-434.

Firbringer, Max. 1900. Zur vergleichenden Anatomie des Brustschulterapparates und
der Schultermuskeln. Jenaischen Zeitschr. Naturwiss. 34: 215-718.

Gadow, Hans. 1882. Untersuchungen iiber die Bauchmuskeln der Krokodile, Eidechsen
und Schildkréten. Morph. Jahrb. 7: 57-100.

1901. Amphibia and reptiles. London. 668 p.

Gans, Carl. 1961. The feeding mechanism of snakes and its possible evolution. Amer.
Zool. 1(2): 217-227.

Gaudry, Albert. 1892. Les pythonomorphes de France. Soc. Géol. France (Paléont.)
Mém. 10. 13 p.

Gervais, Paul. 1853. Observations relatives aux reptiles fossiles de France. Acad. Sci.
Paris Compt. Rendus 36: 374-377, 470-474.

Gibbes, Robert W. 1850. On Mosasaurus and other allied genera in the United States.
Amer. Assoc. Adv. Sci. Proc. second meeting, Cambridge, 1849: p. 77.

1851. A memoir on Mosasaurus and three allied genera, Holcodus, Conosaurus,
and Amphorosteus. Smithsonian Inst. Contrib. Knowl. 2(5): 1-13.

Gill, James R., and W. A. Cobban. 1961. Stratigraphy of lower and middle parts of the
Pierre Shale, northern Great Plains. U. S. Geol. Surv. Prof. Paper 424D, art. 352:
D185-D191.

Gilmore, Charles W. 1912a. A new mosasauroid reptile from the Cretaceous of Alabama.
U.S. Natl. Mus. Proc. 41: 479-484.

1912b. (Mosasaurian in Fox Hills Formation.) U. S. Geol. Surv. Folio 181: 2.

1921. An extinct sea-lizard from western Kansas. Sci. Amer. 124: 273, 280.

1926. Description of mosasaurians from the Ripley Formation, Tennessee. U. S.

Geol. Surv. Prof. Paper 137: 191-192.

1927. Note on a second occurrence of the mosasaurian reptile Globidens. Science,

n.s., 66: 452.

1928. Fossil lizards of North America. Natl. Acad. Sci. Mem. 22, third memoir.

169 p.

1943. Fossil lizards of Mongolia. Amer. Mus. Nat. Hist. Bull. 81(4): 361-384.

Goldfuss, August. 1845. Der Schidelbau des Mosasaurus, durch Beschreibung einer neuen
Art dieser Gattung erlautert. Nova Acta Acad. Caes. Leopoldino-Carolinae Germanicae
Nat. Curiosorum 21: 173-200.

RUSSELL: AMERICAN MOSASAURS 217

1847. Der Schiidelbau des Mosasaurus durch Beschreibung einer neuen Art dieser
Gattung erlautert. Neues Jahrb. Geognosie Geol., 1847: 122-125.

Gregory, Joseph T. 1951. Convergent evolution: the jaws of Hesperornis and the mosa-
saurs. Evolution 5(4): 345-354.

1952. The jaws of the Cretaceous toothed birds Ichthyornis and Hesperornis.
Condor 54: 73-88.

Harlan, Richard. 1825. Notice of the Plesiosaurus and other fossil reliquiae from the state
of New Jersey. Acad. Nat. Sci. Philadelphia Jour. 4: 232-236.

1834a. On some new species of fossil saurians found in America. Brit. Assoc.

Adv. Sci. Rep. 3rd mtg., Cambridge, 1833: 440.

1834b. Critical notices of various organic remains hitherto discovered in North

America. Geol. Soc. Pennsylvania Trans. 1(1): 46-112.

1834c. Notice of the discovery of the remains of the Ichthyosaurus in Missouri,

N. A. Amer. Phil. Soc. Trans. n.s., 4: 405-408.

1834d. Announcement of the finding of Ichthyosaurus missouriensis and Basilo-

saurus. Soc. Géol. France Bull. premiére ser. 4: 124.

1834e. Critical notices of various organic remains hitherto discovered in North

America. Edinburgh new Phil. Jour. 17: 342-362, and 18: 28-40.

1835a. Critical notices of various organic remains hitherto discovered in North

America in Harlan, Richard, Medical and physical researches. Philadelphia: 253-313.

1835b. Description of the ichthyosaurian remains recently discovered in the state

of Missouri in Harlan, R. Medical and physical researches. Phila. 344-348.

1835c. Notice of plesiosaurian and other fossil reliquiae from the state of New

Jersey in Harlan, R. Medical and physical researches. Phila: 382-385.

1839a. Notice of the discovery of Basilosaurus and Batrachiosaurus. Geol. Soc.

London Proc. 3: 23-24.

1839b. (Letter regarding Basilosaurus and Batrachotherium.) Soc. Géol. France

Bull. 1° ser. 10: 89-90.

1839c. On the discovery of the Basilosaurus and the Batrachiosaurus. London and

Edinburgh Phil. Mag., 3rd ser. 14: 302.

1842. Notice of two new fossil mammals from Brunswick canal, Georgia; with
observations on some of the fossil quadrupeds of the United States. Amer. Jour. Sci.
43(1): 141-144.

Hay, Oliver P. 1902. Bibliography and catalogue of the fossil vertebrata of North America.
U. S. Geol. Surv. Bull. 179. 868 p.

Hayden, F. V. 1857. Notes explanatory of a map and section illustrating the geological
structure of the country bordering on the Missouri River, from the mouth of the
Platte River to Fort Benton. Acad. Nat. Sci. Philadelphia Proc. 9: 109-116.

1872. Preliminary report of the United States Geological Survey of Wyoming and
portions of contiguous territories. Washington. 188 p.

Hector, James, 1874. On the fossil reptilia of New Zealand. New Zealand Inst. Trans.
Proc. 6: 333-358.

Hesse, C. J., and S. P. Welles. 1936. The first record of a dinosaur from the West Coast.
Science, n.s., 84: 157-158.

Hill, R. T. 1901. Geography and geology of the Black and Grand Prairies, Texas. U. S.
Geol. Surv. Rep. 21(7). 666 p.

Hoffmann, C. K. 1890. Reptilien 2. Eidechsen und Wassereidechsen in Bronn, H. G.,
Klassen und Ordnungen des Thierreichs 6(3): 2017-2089.

Hoffstetter, Robert. 1955. Squamates de type moderne in Piveteau, Jean, Traité de
paléontologie 5: 606-662.

1962. Revue des récentes acquisitions concernant l'histoire et la systématique des
squamates in (Colloques internationaux du centre national de la recherche scienti-
fique) Problems actuel de paléontologie: 243-279.

Hofker, Jan. 1959. Les foraminiféres du Crétacé supérieur de Harmignies, Bassin de
Mons. Soc. Géol. Belgique Ann. 82: 319-325,

1960a. Les foraminiféres du Crétacé supérieur dans le Bassin de Mons. I. Les

218 PEABODY MUSEUM BULLETIN 23

forminiféres de la Craie Phosphatée de Cipley. Soc. Géol. Belgique Ann. 83: 165-180.

1960b. Les foraminiféres du Crétacé supérieur dans le Bassin de Mons. II. Les
foraminiféres du Tuffeau de Saint-Symphorien. Soc. Géol. Belgique Ann. 83: 181-195.

Holland, William J. 1908. An undetermined element in the osteology of the Mosa-
sauridae. Carnegie Mus. Ann. 4: 162-167.

Holmes, F. S. 1849. Notes on the geology of Charleston, S. C. Amer. Jour. Sci., 2nd ser. 7:
187-201.

Howeil, A. Brazier. 1930. Aquatic mammals, their adaptations to life in the water. Spring-
field, Illinois and Baltimore, Maryland. 338 p.

Huene, Erika von. 1935. Mosasaurier-Zahne von Timor. Zentrabl. Min. Geol. Paleont.,
Abt. B, 10: 412-416.

Huene, Friedrich von. 1910. Ein ganzes Tylosaurus-Skelett. Geol. Palaeont. Abhandl.
Jena, neue Folge, 8: 297-314.

1911. Uber einem Platecarpus in Tiibingen. Neues Jahrb. Min. Geol. Palaont. 2:

48-50.

1919. Bilder aus der Paliontologischen Universitatssammlung in Tubingen.

Jahresh. Ver. Naturk. Wurttemberg, 75: 177-184.

1925. Einige Beobachtungen an Mixosaurus cornalianus (Bassani). Zentralbl.

Min. Geol. Palaeont. Abt. B, 1925: 289-295.

1953. Eine Biegungstelle in halber Lange des Unterkiefers als Seltenheit in
Skelettbau und Biologie. Palaont. Zeitschr. 27: 146-148.

Iakovlev, N. N. 1901. Restes d’un mosasaurien trouvé dans le Crétacé supérieur de sud
de la Russie. Izv. Geol. Kom. (U.S.S.R.), 20: 507-520.

1906. Notes sur les mosasauriens. Izv. Geol. Kom. (U.S.S.R.) 24: 135-152.

Janensch, W. 1906. Uber Archaeophis proavus Mass. eine Schlange aus dem Eocan de
Monte Bolca. Beitr. Paleont. Geol. Osterreich-Ungarns, 19: 1-33.

Jeletzky, J. A. 1955. Belemnitella praecursor, probably from the Niobrara of Kansas, and
some stratigraphic implications. Jour. Paleont. 29(5): 876-885.

1962. The allegedly Danian dinosaur-bearing rocks of the globe and the prob-
lem of the Mesozoic-Cenozoic boundary. Jour. Paleont. 36(5): 1005-1018.

Johnson, C. L. 1959. Microfossils of the Gregory Shale Member of the Pierre Formation.
South Dakota Acad. Sci. Proc. 38: 49-52.

Kauffman, Erle G., and Robert V. Kesling. 1960. An upper Cretaceous ammonite bitten
by a mosasaur. Mus. Paleont. Univ. Michigan Contrib. 15(9): 193-248.

Kornhuber, Andreas. 1873. Uber einem neuen fossilen Saurier aus Lesina. Abhandl.
K. K. Geol. Reichanst. 5(4): 77-90.

1893. Carsosaurus marchesetti, ein neuer fossiler Lacertilier aus den Kreide-

schichten des Karstes bei Komen. Abhandl. K. K. Geol. Reichanst. 17(3): 1-15.

1901. Opetiosaurus bucchichi, eine neue fossile Fidechse aus der unteren Kreide
von Lesina in Dalmatien. Abhandl. K. K. Geol. Reichanst. 17(5): 1-24.

Kramberger, Carl Gorjanovic. 1892. Aigialosaurus, eine neue Eidechse aus den Kreide-
schiefern der Insel Lesina, mit Riicksicht auf die bereits beschriebenen Lacertiden von
Comen und Lesina. Glasnik Soc. Hist.-Nat. Croatica 7: 74-106.

Krinsley, D., and M. Schneck. 1964. The palaeoecology of a transition zone across an
upper Cretaceous boundary in New Jersey. Palaeontology 7(2): 266-280.

Kuhn, Emil. 1952. B. Peyer. Die Triasfauna der Tessiner Kalk alpen, 17. Askeptosaurus
italicus Nopcsa. Schweizerische Paleont. Abhandl. 69(2): 1-73.

Kuhn, Oskar. 1958. Ein neuer Lacertilier aus dem frankischen Lithographieschiefer.
Neues Jahrb. Min. Geol. Paldont. (1958): 380-382.

Kulp, J. L. 1961. Geologic time scale. Science 133(3459): 1105-1114.

Laird, W. M. 1951. Geology of the Pembine Hills. North Dakota Geol. Surv. Rep. In-
vest. 4 in North Dakota Outdoors 13 (12): 12-13.

Lakjer, Tage. 1926. Studien iiber die Trigeminus-Versorgte Kaumuskulatur der
Sauropsiden. Copenhagen. 153 p.

Lambe, Lawrence M. 1914. Report of the vertebrate paleontologist. Geol. Sury. Canada
for 1912 Summary Rep.: 397-403.

RUSSELL: AMERICAN MOSASAURS 219

Lane, Henry H. 1947. A survey of the fossil vertebrates of Kansas. Pt. 3, the reptiles.
Kansas Acad. Sci. Trans. 49(3): 289-322.

Langer, Wolfhart. 1961. Uber das Alter der Fischschiefer von Hvar-Lesina (Dalmatien).
Neues Jahrb. Min. Geol. Palaont. (1961): 329-331.

Lee, W. T. 1897. (Mosasaur bones from Colorado.) Amer. Nat. 31: 614.

Leidy, Joseph. 1856. Notices of extinct vertebrated animals discovered by Prof. E, Em-
mons. Acad. Nat. Sci. Philadelphia Proc. 8: 255-256.

1857. Notices of remains of extinct vertebrated animals discovered by Prof.

E. Emmons. Amer. Jour. Sci., 2nd ser., 23(68): 271-272.

1858a. List of extinct Vertebrata, the remains of which have been discovered in

the region of the Missouri River: with remarks on their geologic age. Acad. Nat. Sci.

Philadelphia Proc. 9: 89-91.

1858b. (Remarks on Mosasaurus.) Acad. Nat. Sci. Philadelphia Proc. 9: 176.

1858c. (Remarks on Drepanodon impar.) in Emmons, E., Agriculture of the

eastern counties, together with descriptions of the fossils of the marl beds. North

Carolina Geol. Surv. Rep. Raleigh: 224.

1860. Synonymy of the American Mosasaurus. Acad. Nat. Sci. Philadelphia Proc.

11:91-92.

1861. (Remarks on remains of some extinct Vertebrata in the territory of

Nebraska.) Amer. Phil. Soc. Proc. 7: 10-11.

1865a. Memoir on the extinct reptiles of the Cretaceous formations of the

United States. Smithsonian Inst. Contrib. Knowl. 14(6): 1-165.

1865b. Brief review of a memoir on the Cretaceous reptiles of the United

States. Smithsonian Inst. Ann. Rep. Board Regents for 1864: 66-73.

1870. (Remarks on Poicilopleuron valens, Clidastes intermedius, Leiodon

proriger, Baptemys wyomingensis, and Emys stevensonianus.) Acad. Nat. Sci.

Philadelphia Proc. 22:3-5.

1873. Contributions to the extinct vertebrate fauna of the western territories.
U. S. Geol. Surv. Terr. Rep. 1: 14-358.

Leonardi, Piero. 1946. I mosasauri del Veneto. Pontificia Acad. Sci. Commentations 10:
463-486.
Leonardi, Piero, and Roberto Malaroda. 1946. Prima segnalazione di un mosasauro del
genera Globidens nel Cretaceo dell 'Egitto. Pontificia Acad. Sci. Acta 10: 183-190.
Leriche, Maurice. 1929. Les Poissons du Crétacé marin de la Belgique et du Limbourg
hollandais. Soc. Géol. Belge Bull. 37: 199-299.

1934. Sur le Crétacé supérieur du Hainaut et du Brabant. Soc. Géol. Belgique
Ann. 58: 118-140.

Leroy, L. W., and N. C. Schieltz. 1958. Niobrara-Pierre boundary along the Front
Range, Colorado. Amer. Assoc. Petrol. Geol. Bull. 42(10): 2444-2464.

Loomis, Frederic B. 1904. On some marine fossils in the Titanothere beds. Science, n. s.,
19: 254.

1915. A new mosasaur from the Ft. Pierre. Amer. Jour. Sci., 4th ser. 39(37):
555-566.

Lowenstam, H. A. and. S. Epstein. 1954. Paleotemperatures of the post-Aptian Creta-
ceous as determined by the oxygen isotope method. Jour. Geol. 62(3): 207-248.

Lundelius, Ernest, Jr., and S. St. J. Warne. 1960. Mosasaur remains from the upper
Cretaceous of western Australia. Jour. Paleont. 34(6) : 1215-1217.

Lydekker, Richard A. 1888. Catalogue of the fossil Reptilia and Amphibia in the
British Museum. London, pt. 1. 309 p.

Mantell, Gideon A. 1829. A tabular arrangement of the organic remains of the county
of Sussex. Geol. Soc. London Trans. 2nd ser. 3: 201-216.

Marie, P. 1956. Sur quelques foraminiféres nouveaux du Crétacé supérieur belge. Soc.
Géol. Belgique Ann. 80: 235-257.

Marsh, Othniel Charles. 1869. Notice of some new mosasauroid reptiles from the
greensand of New Jersey. Amer. Jour. Sci., 2nd ser. 48(144): 392-397.

1870. (Remarks on Hadrosaurus minor, Mosasaurus crassidens, Liodon laticaudus,

220 PEABODY MUSEUM BULLETIN 23

Baptosaurus, and Rhinoceros matutinus.) Acad. Nat. Sci. Philadelphia Proc. 22:2-3.
187la. Notice of some new fossil reptiles from the Cretaceous and Tertiary
formations. Amer. Jour. Sci. 3rd ser. 1(6): 447-459.

1871b. (Communication on some new reptiles and fishes from the Cretaceous
and Tertiary.) Acad. Nat. Sci. Philadelphia Proc. 23: 103-105.

1872a. Discovery of the dermal scutes of mosasauroid reptiles. Amer. Jour. Sci.
3rd ser. 3(16): 290-292.

1872b. On the structure of the skull and limbs in mosasauroid reptiles, with
descriptions of new genera and species. Amer. Jour. Sci. 3rd ser. 3(18): 448-464.
1872c. Note on Rhinosaurus. Amer. Jour. Sci. 3rd ser. 4(20): 147.

1872d. New and remarkable fossils. Amer. Nat. 6(8): 495-497.

1876. Recent discoveries of extinct animals. Amer. Jour. Sci. 3rd ser. 12(67): 59-61.
1877. Introduction and succession of vertebrate life in America. Amer. Jour. Sci.
3rd ser. 14(83): 337-378.

1880. New characters of mosasauroid reptiles. Amer. Jour. Sci. 3rd ser. 19(109):
83-87.

1897. Vertebrate fossils of the Denver Basin. U. S. Geol. Surv. Monogr. 27: 473-

ATs

Martin, Harold. 1953. A South Dakota mosasaur. Unpublished Master’s Thesis, South
Dakota School of Mines and Technology. 65 p.

McDowell, Samuel B., Jr., and Charles M. Bogert. 1954. The systematic position of
Lanthanotus and the affinities of the Anguinomorph lizards. Amer. Mus. Nat. Hist.
Bull. 105(1): 1-142.

Meek, F. B., and F. V. Hayden, 1857. Descriptions of new species and genera of fossils,
collected by Dr. F. V. Hayden in Nebraska territory. Acad. Nat. Sci. Philadelphia
Proc. 9: 117-148.

Mehl, Maurice G. 1930. A new genus of mosasaurs from Mexico, and notes on the
pelvic girdle of Platecarpus. Denison Univ. Jour. Sci. Lab. Bull. 24: 383-400.

Mellen, F. F. 1958. Cretaceous shelf sediments of Mississippi. Mississippi Geol. Surv.
Bull. 85: 1-112.

Merriam, Daniel F. 1963. The geologic history of Kansas. State Geol. Surv. Kansas Bull.
162. 317 p.

Merriam, John C. 1894. Ueber die Pythonomorphen der Kansas Kreide. Palaeonto-
graphica 41: 1-39.

Merril, G. P. 1907. Catalogue of types, cotypes, and figured specimens of fossil verte-
brates in the department of geology, U. S. National Museum. U. S. Natl. Mus. Bull.
53(2): 1-81.

Mertens, Robert. 1942. Die Familie der Varane (Varanidae). Zweiter Teil: Der Schadel.
Senckenbergischen Naturf. Gesell. Abhandl. 465: 117-234.

Meyer, Hermann von. 1845. (Letter directed to Professor Bronn.) Neues Jahrb. Min.
Geognosie Geol. 1845: 308-313.

1860. Acteosaurus tommasini aus dem schwarzen Kreide-Schiefer von Comen am
Karste. Palaeontographica 7: 223-231.

Miller, Halsey W. Jr. 1955. A check-list of the Cretaceous and Tertiary vertebrates of
New Jersey. Jour. Paleont. 29(5) : 903-914.

Mitchill, S. L. 1818. Observations on the geology of North America in Cuvier, G. C. F.,
Essay on the theory of the earth. New York, p. 319-431.

Morton, S. G. 1830a. Synopsis of the organic remains of the Ferruginous sand formation
of the United States. Amer. Jour. Sci. 17(2): 274-294.

1830b. Synopsis of the organic remains of the Ferruginous sand formation of the

United States. Amer. Jour. Sci. 18(2): 243-250.

1834. Synopsis of the organic remains of the Cretaceous group of the United

States. Philadelphia. 104 p.

1844. Fossil bones of Mosasaurus. Acad. Nat. Sci. Philadelphia Proc. 2: 132-133.

Miillerried, Frederico K. G. 1931. Un reptil y algunos invertebrados fésiles de Rayén,
Estado de Tamaulipas. Ann. Inst. Biol. Mexico 2: 171-178.

RUSSELL: AMERICAN MOSASAURS pd |

Nishi, Seiho. 1916. Zur vergleichenden Anatomie der eigentlichen (genuinen) Riicken-
muskeln. Morph. Jahrb. 50: 168-318.

1938. Muskeln des Rumpfes im Bolk, Louis, Handbuch der vergleichenden
Anatomie der Wirbeltiere. Berlin 5: 351-446.

Nopcsa, Franz. 1903a. Uber die Varanus-artigen Lacerten Istriens. Beitr. Paleont. Geol.
Osterreich-Ungarns 15: 30-42.

1903b. On the origin of the mosasaurs. Geol. Mag., n. s., decade 4, 10: 119-121.

1908. Zur Kenntnis der fossilen Eidechsen. Beitr. Paleont. Geol. Osterreich-

Ungarns 21: 33-62.

1923. Eidolosaurus und Pachyophis zwei neue Neocom-Reptilien. Palaeonto-

graphica 65: 97-154.

1928. The genera of reptiles. Palaeobiologica 1: 163-188.

Olson, Everett C. 1936. The dorsal axial musculature of certain primitive Permian
tetrapods. Jour. Morph. 59(2): 265-311.

Olsson, R. K. 1963. Latest Cretaceous and earliest Tertiary stratigraphy of New Jersey
coastal plain. Amer. Assoc. Petrol. Geol. Bull. 47(4): 643-665.

Osborn, Henry Fairfield. 1899a. A complete mosasaur skeleton, osseous and cartilaginous.
Amer. Mus. Nat. Hist. Mem. 1 (4): 167-188.

1899b. Upon the structure of Tylosaurus dyspelor, including the cartilaginous
sternum. Science, 2nd ser., 9: 912-913.

Osburn, Raymond C. 1906. Adaptive modifications of the limb skeleton in aquatic
reptiles and mammals. New York Acad. Sci. Ann. 16(3): 447-476.

Ostrom, John H. 1962. On the constrictor dorsalis muscles of Sphenodon. Copeia, 1962
4: 732-735.

Owen, Richard, 1840-1845. Odontography. London, 2 vols.

1849. Notes on remains of fossil reptiles discovered by Prof. Henry Rogers, of

Pennsylvania, U. S., in the Greensand formation of New Jersey. Geol. Soc. London

Quart. Jour. 5: 380-383.

1849-1884. A history of British fossil reptiles. London, 4 vols.

1877. On the rank and affinities in the reptilian class of the Mosasauridae. Geol.

Soc. London Quart. Jour. 33(4): 682-719.

1878. On the affinities of the Mosasauridae, Gervais, as exemplified in the bony

structure of the forelimb. Geol. Soc. London Quart. Jour. 34: 748-753.

1879. On the occurrence in North America of rare extinct vertebrates found

fragmentarily in England. Ann. Mag. Nat. Hist., 5th ser., 4: 53-61.

1880. On the occurrence in North America of rare extinct vertebrates found
fragmentarily in England. Ann Mag. Nat. Hist., 5th ser., 5: 177-181.

Persson, P. O. 1959. Reptiles from the Senonian of Scania. Arkiv. Min. Geol. Stock-
holm 2(5): 431-480.

1963. Studies on Mesozoic marine reptile faunas with particular regard to the
Plesiosauria. Inst. Min. Paleont. Quaternary Geol. Univ. Lund, Publ. 118: 1-15.

Pictet, F. J. 1845.Traité élémentaire de paléontologie. Paris, 2. 407 p.

1853. Traité de paléontologie. 2° ed. Paris, 1. 584 p.

Pierce, G. R. and S. P. Welles. 1959. First record of mosasaur from the Cretaceous of
Santa Barbara de Barinas, Venezuela. Jour. Paleont. 33(5): 966-967.

Pompeckj, Josef F. 1910. Uber einem Fund von Mosasaurier-Resten im Ober-Senon von
Haldem. Jahresber. Niedersachsischen Geol. Ver. 3: 122-140.
Pravoslavlev, P. A. 1914. Restes d’un mosasaurien trouvé dans le Crétacé supérieur du
bassin de la riviére Liski, région du Don. Izvest. donsk. politekh. Inst. 3: 168-183.
1916. Restes d’un Elasmosaurus trouvé dans le Crétacé supérieur de la Province
du Don. Trudy Leningradskee Obshch. Estest. 38(5): 153-332.

Price, Llewellyn I. 1957. A presence de Globidens no Cretacio superior do Brasil. Bol.
Div. Geol. Min. Rio de Janeiro 169: 1-24.

Priddy, R. R. 1954. Recent Mississippi sound sediments compared with some upper
Cretaceous sediments. Gulf Coast Assoc. Geol. Soc. Trans. 4: 159-168.

222 PEABODY MUSEUM BULLETIN 23

Pryor, W. A. 1960. Cretaceous sedimentation in upper Mississippi embayment. Amer.
Assoc. Petrol. Geol. Bull. 44(9): 1473-1504.

Pumphrey, R. G. 1950. Hearing in Physiological mechanisms in animal behaviour.
Symposia Soc. Exp. Biol. 4, Cambridge: 3-18.

Quass, Arthur. 1902. Beitrag zur Kenntnis der Fauna der obersten Kreidebildungen in
der Libyschen Wiste. Palaeontographica 30(2): 153-336.

Raab, M. 1963. Fossil fish and reptiles from late Campanian phosphate deposits of the
Negev region of Israel. Israel Jour. Earth Sci. 12(1): 26-40.

Ray, Carleton. 1963. Locomotion in pinnipeds. Nat. Hist. 72(3) : 10-21.

Reeside, John B., Jr. 1957. Paleoecology of the Cretaceous seas of the western interior
of the United States. Geol. Soc. Amer. Mem. 67, 2: 505-541.

Renger, J. J. 1935. Excavation of Cretaceous reptiles in Alabama. Sci. Monthly 41:
560-565.

Repelin, Joseph. 1915. Découverte d’ossements de grands pythonomorphes dans le
Crétacé supérieur des environs de Jerusalem. Acad. Sci. Paris Compt. Rendus 161:
735-736.

Romer, Alfred S. 1922. The locomotor apparatus of certain primitive and mammal-
like reptiles. Amer. Mus. Nat. Hist. Bull. 46(10): 517-606.

1942. The development of tetrapod limb musculature—the thigh of Lacerta. Jour.

Morph. 71(2): 251-298.

1944. The development of tetrapod limb musculature—the shoulder region of

Lacerta. Jour. Morph. 74(1): 1-41.

1945. Vertebrate Paleontology. 2nd ed. Chicago. 687 p.

1956. Osteology of Reptiles. Chicago. 772 p.

Russell, D. A. 1964. Intracranial mobility in mosasaurs. Peabody Mus., Yale Univ., Postilla
86. 19 p.

1967. Cretaceous vertebrates from the Anderson River N.W.T. Canadian Jour.
Earth Sci. 4: 21-38.

Schmid, Friedrich. 1959. Biostratigraphie du Campanien-Maestrichtien du NE de la
Belgique sur la base de bélemnites. Soc. Géol. Belgique Ann. 82: 239-256.

Schmidt, Karl P. 1927. New reptilian generic names. Copeia 163: 58-59.

Scott, G. 1940. Paleoecological factors controlling distribution and mode of life of
Cretaceous ammonoids in Texas area. Amer. Assoc. Petrol. Geol. Bull. 24(7): 1164-
1203.

Scott, Glenn R., and William A. Cobban. 1964. Stratigraphy of the Niobrara Formation
at Pueblo, Colorado. U. S. Geol. Surv. Prof. Paper 454 L. 30 p.

Seeley, Harry G. 1881. On remains of a small lizard from the Neocomian rocks of
Comén, near Trieste, preserved in the geological museum of the University of
Vienna. Geol. Soc. London Quart. Jour, 37: 52-56.

Sevon, William. 1958. A Niobrara mosasaur of South Dakota. South Dakota Acad. Sci.
Proc. 36: 144-146.

Signeux, J. 1959. Poissons et reptiles marins in Arambourg, C., Contributions 4 la strati-
grapie et a la paléontologie du Crétacé et du Nummulitique de la marge NW de la
péninsula Arabique. Notes Mém. Moyen-Orient, Mus. Natl. Hist. Nat., 1959: 223-228.

Simpson, George G. 1942. The beginnings of vertebrate paleontology in North America.
Amer. Phil. Soc. Proc. 86: 130-188.

Smith, F. E. 1959. Lower Tertiary and Upper Cretaceous of Brazos River Valley, Texas.
Houston Geol. Soc. and Gulf Coast Sect. Soc. Econ. Paleont. Min., Guidebook Ann.
Field Trip, 1959: 1-54.

Smith-Woodward, Arthur. 1889. A synopsis of the vertebrate fossils of the English chalk.
Geol. Assoc. Proc. 10: 273-338.

1906. Note on some portions of mosasaurian jaws obtained by Mr. G. E. Dibley

from the middle chalk of Cuxton, Kent. Geol. Assoc. Proc. 19: 185-187.

1922. A guide to the fossil reptiles, amphibians, and fishes in the department

of geology and palaeontology in the British Museum of Natural History. 10th ed.

London. 122 p.

RUSSELL: AMERICAN MOSASAURS 223

Snow, F. H. 1878. On the dermal covering of a mosasauroid reptile. Kansas Acad. Sci.
Trans. 6: 54-58.

Sornay, J. (ed.) 1957. Crétacé. Lexique stratigraphique international, 1, fasc. 4-a-vi,
403 p.

Stephenson, Lloyd W., Philip B. King, Watson H. Monroe, and Ralph W. Imlay. 1942.
Correlation of the outcropping Cretaceous formations of the Atlantic and Gulf
coastal plain and Trans-Pecos, Texas. Geol. Soc. Amer. Bull. 53: 435-438.

Sternberg, Charles H. 1899. A Kansas mosasaur. Popular Sci. News 33: 259-260.

1905. Protostega gigas and other Cretaceous reptiles and fishes from the Kansas

chalk. Kansas Acad. Sci. Trans. 19: 123-128.

1907. Some animals discovered in the fossil beds of Kansas. Kansas Acad. Sci.

Trans. 20: 122-124.

1908. My expedition to the Kansas chalk for 1907. Kansas Acad. Sci. Trans. 21:

111-114.

1909. The life of a fossil hunter. New York. 286 p.

1911. In the Niobrara and Laramie Cretaceous. Kansas Acad. Sci. Trans. 23:

70-74.

1915. Evidence proving that the Belly River beds of Alberta are equivalent with

the Judith River beds of Montana. Science, n. s., 42: 131-133.

1917. Hunting dinosaurs in the bad lands of the Red Deer River, Alberta, Can-

ada. Lawrence, Kansas. 232 p.

1918. Five years exploration in the fossil beds of Alberta. Kansas Acad. Sci. Trans.
28: 205-211.

Stock, Chester. 1939. Occurrence of Cretaceous reptiles in the Moreno Shales of the
southern Coast Ranges, California. Natl. Acad. Sci. Proc. 25: 617-620.

Strand, Embrik. 1926. Miscellanea nomenclatorica zoologica et palaeontologia. Arch.
Naturgesch. Berlin 92: 30-75.

Stromer, Ernst. 1910. Neue forschungen tiber fossile Lungenatmende Meeresbewohner.
Fortschr. Naturwiss. Forsch. Berlin, 2: 83-114.

Stromer, Ernst, and Wilhelm Weiler. 1930. Beschreibung von Wirbeltier-Resten aus dem
nubischen Sandsteine Oberagyptens und aus dgyptischen Phosphaten. Abhandl.
Bayerischen Akad. Wissenschaften, neue Folge 7: 1-42.

Swanston, W. 1886. Mosasaurus gracilis, Owen, from the Irish chalk. Geol. Mag., n. s.,

decade 3, 3: 134.

Swinton, William E. 1930. On fossil reptilia from Sokoto Province. Geol. Surv. Nigeria
13: 1-56.

Telles Antunes, Miguel. 1964. O Neocretacico e o Cenozéico do littoral de Angola. Junta
de investigagdes do Ultramar, Lisboa. 259 p.

Thevenin, Armand. 1896. Mosasauriens de la Craie Grise de Vaux-Eclusier prés Péronne
(Somme). Soc. Géol. France Bull. 3° ser., 24(2): 900-916.

Tsaregradskii, V. 1927. Les restes de mosasaures du gouvernment de Saratov et du ter-
ritoire d’Ouralsk. Izv. Geol. Komitet U.S.S.R. 45: 563-572.

1935. Description detaillée du mosasaure Dollosaurus lutugini Jak. Soc. Paléont.
Russie Ann. 10: 49-54.

Vallois, Henri. 1922. Les transformations de la musculature de I’épisome chez les verté-
brés. Arch. Morph. 13: 1-538.

Vaughn, Peter P., and Mary R. Dawson. 1956. On the occurrence of calcified tympanic
membranes in the mosasaur Platecarpus. Kansas Acad. Sci. Trans. 59(3): 382-384.
Versluys, J. 1907. Eene bijzonderheid von de onderkaak der Mosasauria in aanleg reeds
bij Varanus aanwezig ist. Verslag. Wetensch. Vergaderingen Nederlandsche Dierh.

Ver., Tijdschr. Nederlandsche Dierh. Ver., 2nd ser., 10: 28-29.

Virchow, Hans. 1914. Mechanik der Wirbelsiule des Varanus varius. Archiv fiir Anat.
1914: 69-89.

Walls, Gordon L. 1942. The vertebrate eye and its adaptive radiation. Michigan, Cran-
brook Institute of Science. 785 p.

224 PEABODY MUSEUM BULLETIN 23

Watkinson, Grace B. 1906. The cranial nerves of Varanus bivittatus. Morph. Jahrb. 35:
450-472.

Watson, D. M. S. 1924. The elasmosaurid shoulder-girdle and forelimb. Zool. Soc. Lon-
don Proc. 1924 (3): 885-917.

Welles, Samuel P. 1943. Elasmosaurid plesiosaurs with description of new material from
California and Colorado. Univ. California Mem. 13: 125-254.

Wellman, H. W. 1959. Divisions of the New Zealand Cretaceous. Roy. Soc. New Zealand
Trans. 87: 99-163.

Wever, Ernest G., and Jack A. Vernon. 1956. Sound transmission in the turtle’s ear.
Natl. Acad. Sci. Proc. 42: 292-299.

Whitfield, R. P. 1900. Note on the principal type specimen of Mosasaurus maximus
Cope. Amer. Mus. Nat. Hist. Bull. 13: 25-29.

Williston, Samuel W. 1891. Kansas mosasaurs. Science 18: 345.

1893a. Kansas mosasaurs. Part 2, restoration of Clidastes. Kansas Univ. Quart. 2:

83-84.

1893b. The Niobrara Cretaceous of western Kansas. Kansas Acad. Sci. Trans. 13:

107-111.

1895. New or little-known extinct vertebrates. Kansas Univ. Quart. 3: 165-176.

1897a. Brachysaurus, a new genus of mosasaurs. Kansas Univ. Quart. 6: 95-98.

1897b. On the extremities of Tylosaurus. Kansas Univ. Quart. 6: 99-102.

1897c. Restoration of Kansas mosasaurs. Kansas Univ. Quart. 6: 107-110.

1897d. Range and distribution of the mosasaurs. Kansas Univ. Quart. 6:

177-189.

1897e. The Kansas Niobrara Cretaceous. Univ. Geol. Surv. Kansas 2: 235-246.

1898a. The Upper Cretaceous of Kansas. Univ. Geol. Surv. Kansas, 4: 28-32.

1898b. Mosasaurs. Univ. Geol. Surv. Kansas, 4: 83-221.

1898c. Editorial notes. Kansas Univ. Quart. 7: 235.

1899. Some additional characters of mosasaurs. Kansas Univ. Quart. 8: 39-41.

1902. Notes on some new or little known extinct reptiles. Kansas Univ. Sci.

Bull. 1: 247-254.

1904. The relationships and habits of the mosasaurs. Jour. Geol. 12: 43-51.

1910. A mounted skeleton of Platecarpus. Jour. Geol. 18: 537-541.

1914. Water reptiles of the past and present. Chicago. 251 p.

1925. The osteology of reptiles. Cambridge. 300 p.

Williston, Samuel W., and E. C. Case. 1892. Kansas mosasaurs. Kansas Univ. Quart. 1:
15-32.

Wiman, Carl. 1920. Some reptiles from the Niobrara group in Kansas. Geol. Inst. Upsala
Bull. 18: 9-18.

Young, K. 1963. Upper Cretaceous ammonites from the Gulf Coast of the United States.
Univ. Texas Bull. 6304. 142 p.

Zangerl, Rainer. 1948. The vertebrate fauna of the Selma Formation of Alabama. Pt. 1.
Introduction. Fieldiana: Geol. Mem. 3(1): 1-16.

1953. The vertebrate fauna of the Selma Formation of Alabama. Pt. 4. The
turtles of the family Toxochelyidae. Fieldiana, Geol. Mem. 3(4): 135-227.

Zdansky, Otto. 1935. The occurrence of mosasaurs in Egypt and Africa in general. Inst.
Egypt Bull. 17: 83-94.

Zurcher, P. E. F., and H. Arnaud. 1887. Compte rendu de l’excursion 4 Meschers et
Talmont. Soc. Géol. France, Bull. 3° ser., 15: 824-833.

I

il

|

INDEX

Italicized numbers refer to pages with text-
figures. Boldface numbers refer to pages where
a particular taxon is characterized. A list of
all the figures of a species or genus may be
found under the heading of the taxon in the
Systematics section,

Abbreviations

cranial structures, 15

institutional names, 7

postcranial structures, 56, 88, 92, 102, 112,

116

Abstracts, 1-3. See also Summary
Acetabulum, 98, 116
Acknowledgements, 5-6
Aigialosaurs, 9-11, 194-196, 200, 205
Aigialosaurus, 194
Ammonite conch, bitten by mosasaur, 64, 68
Amphekepubis, 122, 141
Amphekepubis johnsoni, 122, 141, 142
Amphorosteus, 173, 176
Ancylocentrum, 162
Angular, 41, 51-52, 52, 60, 66, 67, 166, 177
Antebrachial foramen, 91
Anterior mandibular unit, 15, 49, 65
Articular, 52, 53-54, 60, 66, 67, 166, 177
Atlas, 70, 71
Axis, 71, 72, 74

Bahl, K.N., 14, 34
Baptosaurus, 168
Basal articulation, 60
Basal unit, 15, 60. See also Pterygoid unit
Baseodon, 131, 132, 180
Basioccipital, 29-30, 30, 31, 34, 35, 38, 151
Basisphenoid, 30-33, 30, 31, 34, 35
parasphenoid rostrum, 31
Batrachiosaurus, 131, 132, 137
Batrachotherium, 132, 137
Blood vessels. See also Cranial circulation
basilar artery, 30, 30, 31, 32, 32, 33, 35,
37, 59
inferior orbital artery, 25
internal carotid artery, 30, 31, 31, 32, 33,
34, 37, 59
occipital branch, 38
internal jugular vein (vena capitis latera-
lis), 30, 31, 32, 33, 34, 34, 37, 59
posterior cerebral branch, 38
maxillary artery, 17
maxillary vein, 25
vein to median palatine sinus, 26
Bodily proportions, 70, 210
Bogert, C.M., 197-199, 200
Brachysaurana, 162
Brachysaurus, 162

Camp, C.L., 4, 8, 10-11, 14, 18, 26, 33, 39, 40,
46, 70, 145, 193, 206

Carpus, 90, 93-95, 93, 94, 95, 96, 97, 148
Carsosaurus, 194
Cerebellum, 40
Cerebral hemispheres, 20, 21, 26, 27, 37. See
also Olfactory lobe, Olfactory
tract
Cervical ribs, 73-74
Classification
general, 12, 121-123
history of, 8-12
Clavicle, 86, 87
Clidastes, 121, 124, 138, 187, 189, 190, 204, 210
habits, 65, 68, 120, 187, 191
congrops, 178
iguanavus, 121, 124, 125, 126, 189
liodontus, 65, 121, 125, 126, 127, 187, 188,
208, 210
propython, 65, 121, 126, 132, 187, 188, 189,
208, 211
sternbergi, 121, 126, 188, 202, 208, 210
Cloaca, 111
Coloration, 120, 206
Compressidens, 122, 147, 204, 205
habits, 68
belgicus, 122, 147
fraasi, 122, 147
habits, 68
Condylar foramen, 38, 39, 40
Cope, E.D., 5, 9, 143-144
Coracoid, 85-86, 83, 84, 86
coracoid foramen, 85
Coronoid, 52, 53, 60, 66, 67, 166, 177
Cranial circulation 32, 59, 65. See also Blood
vessels
Cranial kinesis, 60-63, 63, 65, 206
Craniovertebral joint, 61, 62, 108, 120
Crural foramen, 106

DeBeer, G.R., 205
Dentary, 49-50, 49, 50, 52, 66, 67, 133, 136,
154, 161, 166, 177
Devillers, C., 59
Dolichosaurs, 9-10, 196, 199
Dollo, L., 11-12, 68-69, 127
Dollosaurus, 123, 169, 204
lutugini, 123
Dorsum sellae, 31, 37, 35
Drepanodon, 131, 132, 179

Ectenosaurus, 122, 156, 188, 204
clidastoides, 122, 158, 209

Ectopterygoid, 44-45, 151

Edestosaurus, 124

Edmund, A.G., 55

Elliptonodon, 178

Endocranial cast, 37

Endolymphatic duct, 35

Endolymphatic sacculus, 16

Epicoracoid, 85, 111

225

226 PEABODY MUSEUM BULLETIN 23

Epipterygoid, 15, 45, 45, 60, 66 Heloderma, 198-199
Epipterygoid unit, 15, 45, 61 Holcodus, 148, 152, 178
External naris, 17-18, 206 Holosaurus, 148, 152
Eye, 58, 65 Huene, F. von, 18, 45
Humerus, 88-90, 88, 89, 90, 92, 93, 94, 95, 96,

Femur, 101-104, 102, 103, 104, 105 97, 206
Fenestra ovalis (internal auditory meatus), 34, | Hypokinetic axis, 14, 61

34, 35, 36, 37, 38
Fenestra rotunda, 37, 39 Ichthyornis, 121
Fibula, 102, 103, 104, 105, 106 Ichthyosaurs, 205
Foramen magnum, 13 lium, 98-99, 98, 99, 100, 116, 117, 141
Frazzetta, T.H., 60, 65 Immaturity, 121, 192

Frontal, 18-21, 19, 20, 58, 129, 150, 157, 163, 172 Incisura piriformis, 13, 41, 157
Inertial feeding, 68

Gadow, H., 110 Integument, 70
Gans, C., 68 Interclavicle, 86-87, 87
Garzasaurus, 145 Internal naris, 25, 43, 151
Geological formations Internasal septum, 17
Austin Chalk, 190, 225 Interorbital septum, 26, 58
Bearpaw Formation, 189, 228 Ischium, 98, 99, 100, 101, 116, 117, 141
Brownstown Formation, 190, 225
Coon Creek Formation, 185, 226 Jacobson’s (vomeronasal) organ, 25, 26, 151
Eagle Formation, 186, 187, 228 Jeletzky, J.A., 186, 187, Charts 1-4
Eagle Ford Formation, 189, 225 Jugal, 24, 58, 60, 66, 67, 129, 150, 151, 157,
Eutaw Formation, 185, 226 166, 177
Fort Hays Limestone, 186 Jugular foramen, 38, 39, 40
Fox Hills Formation, 187, 189, 228 Juvenile mosasaurs, 121, 192, 210
Judith River Formation, 189
Marlbrook Marl, 189, 225 Kauffman, E.G., 64
Marshalltown Formation, 189, 227 Kesling, R.V., 64
Mesaverde Group, 187, 190 Kolposaurus, 145
Moreno Formation, 190
Navarro Formation, 185, 190, 225 Lacrymal, 21, 129, 172
Navesink Formation, 186, 227 Lakjer, T., 14
New Egypt Formation, 186, 227 Lanthanotus, 197-198
Niobrara Formation, 5, 186-187, 188, 189, Lesticodus, 131, 132, 173, 179
192, 228 Lestosaurus, 148, 152
Pierre Formation, 187-189, 190, 191, 192, | Ligaments
210-212, 228 bodenaponeurosis, 53, 61
Redbank Formation, 186, 227 costal, 78
San Felipe Formation, 189 ilioischiadic, 99, 101, 115, 116
Selma Formation (including Mooreville iliopubic, 99, 100, 115, 116
Chalk), 4, 185, 190, 226 puboischiadic, 100, 101, 115, 116
Taylor Formation, 185, 190, 225 quadratomaxillary, 24
Telegraph Creek Formation, 186, 190, 228 scapulohumeral, 84, 112, 113
Tinton Formation, 186, 22 Liodon, 8, 122, 142, 204, 206
Tombigbee Sand, 190, 226 habits, 68
Glenoid (shoulder) articulation, 111 anceps, 122, 142, 144, 173
Globidens, 50, 122, 144, 190, 204, 205 compressidens, 122, 142, 144
habits, 68 laticaudis, 162
alabamaensis, 122, 144, 156, 189, 191, 209, mosasauroides, 122, 142, 144
211 sectorius, 122, 142, 143

Globidensini, 12, 122, 144
Macrosaurus, 131, 132, 180

Hainosaurus, 123, 142, 176, 203, 204 Mandibular symphysis, 50, 64
habits, 69 Marsh, O:G., 5, 9, 92, 187, 188
bernardi, 123, 176, 191, 210 Martin, H., 70, 93, 103, 134
lonzeensis, 123, 176 Maxilla, 17-18, 66, 67, 126, 129, 130, 136, 150,
Halisaurus, 123, 160, 161, 168 5], 157, 1635, 166, Li 2 5a
onchognathus, 123, 168 McDowell, S.B., 197-199, 200
platyspondylus, 123, 168, 169 Measurements, 208-210
Hay, O.P., 181 Meckelian fossa, 52, 53, 54

hearing, 59, 65 Medulla, 29, 30, 40

RUSSELL: AMERICAN MOSASAURS

Merriam, J.C., 128, 169, 173
Mertens, R., 13
Mesokinetic axis, 14, 21, 27, 28, 60, 61

Metacar

pals, 95-97, 92, 93, 94, 95, 96, 97

Metakinetic axis, 14, 40, 60
Metakinetic joint, 27, 29, 60

Metatar

sals, 103, 104, 105, 106

Middle ear, 59

Mosasaurinae, 11, 12, 121, 123, 204, 206
Mosasaurini, 121, 124

Mosasaurs

Am

erican, miscellaneous occurrences, 184,
189, 190

Belgian, 229

compared with aigialosaurs, 194-195
compared with dolichosaurs, 196
compared with Heloderma, 198-199
compared with Lanthanotus, 197-198
compared with Saniwa, 196-197
compared with varanid lizards, 193-194
English, 231

evolution of, 204-206

foreign occurrences of, 190-191
French, 230

Gulf Coast, 185-186

hab
hist

its of, 65-69, 114-120, 191-192, 205
ory of research, 4-5

juvenile, 121, 192, 210
New Jersey, 184, 186

phy

logeny of, 202-204, 202, 203

Swedish, 231

Mosasaurus, 8, 122, 131, 204, 206
habits, 68, 120
caroliniensis, 179

conodon, 97-98, 122, 124, 132, 189, 191,
210, 211
habits, 192
couperi, 179

crassidens, 179
dekayi, 122, 138

gau

dryi, 122

hoffmanni, 4, 8, 122, 131, 132, 140, 210
impar, 179. See also Drepanodon, Lestico-

dus

ivoensis, 122, 132, 135, 209

lem

onnieri, 124, 133, 135

lonzeensis, 122
maximus, 122, 132, 138, 185, 190, 209, 210
missouriensis, 5, 11, 65, 122, 132, 134, 136,

189, 209, 212

reversus, 180. See also Baseodon

vali
Muscles

M.
M.

M.

M.

dus, 181. See also Nectoportheus

adductor mandibulae externus, 23

adductor mandibulae externus media-
JISH E24 OZ OSs OL

adductor mandibulae externus profun-
dus, 28, 34, 39, 40, 47, 47

adductor mandibulae externus super-
ficialis, 24, 52, 53, 61, 65

. adductor mandibulae posterior, 47, 47,

53, 61

M.
M.
M.
M.
M.
M.
M.

227

adductor mandibulae, undivided, 64
adductor femoris, 101, 177, 118, 118
ambiens, 100, 104, 177, 118, 118
articuloparietalis, 28, 39, 41, 71, 107
biceps brachii, 85, 91, 112, 112, 114
brachialis inferior, 88, 91, 114
caudifemoralis, 81, 82, 101, 709, 117
brevis, 115, 116

longus, 117

. cervicomandibularis, 51, 6/
. coracobrachialis

brevis, 85, 88, 112,

V3; 174

. coracobrachialis longus, 85, 89, 112,

113, 113, 114, 115

. deltoides, 85, 86, 88, 111, 112, 113, 114,
115
M. depressor mandibulae, 28, 39, 41, 48,
54, 61, 64
. femorotibialis, 102, 104, 778

SS85

M.

M.

a ieee ee ere 2 es

. flexor carpi radialis, 114

. flexor carpi ulnaris, 1/4

. flexor palmaris profundus, 91, 1/4

. flexor palmaris superficialis, 1/4

. flexor tibialis, 104, 778, 119

. geniohyoideus, 62

. humeroradialis, 88, 91, 114

- iliocaudalis, 81, 98, 99, 709, 110, 116
. iliocostalis capitis, 30, 39, 109

. iliocostalis caudae, 81, 709, 110

. iliocostalis cervicis, 70, 73, 74, 109

. iliocostalis dorsi, 109

. iliofemoralis, 99, 101, 777, 118, 118

. iliofibularis, 99, 106, 117, 118, 118

. iliotibialis, 99, 104, 777, 118, 118

. interarcuales, 72, 73, 107

. intercostalis externus, 109, 110, 116

. interspinalis, 72, 73, 74, 107, 109, 109

. intertransversarii, 74, 76, 110

- ischiocaudalis, 82, 101, 109, 111, 115,

116

. ischiotrochantericus, 101, 777, 118
. latissimus dorsi, 89, 111, 7/3

. levator angularis oris, 50, 61

. levator costae, 74, 76, 110

levator pterygoid, 34, 44, 62

levator scapulae superficialis, 84, 1/2
longissimus caudae, 81

longissimus cervicis, 71

longissimus cervicocapitis, 74, 75, 108

. longissimus dorsi, 108, 109

. longus colli, 71, 73, 74, 76, 110

. obliquus capitis inferior, 71, 72, 74, 109
. obliquus capitis magnus, 39, 39, 71, 72,

74, 107

. obliquus capitis superior, 71

. omohyoideus, 84, 110, 172

. pectoralis, 87, 88, 112, 114

. pronator profundus (pectoral girdle),

91, 114
profundus

106, 118

protractor pterygoid, 31, 34, 44, 62

pronator (pelvic girdle),

228

. protractor quadrati, 47, 48
. pseudotemporalis, 27, 34, 45, 53, 61
pterygoideus, 44, 54, 62, 64
. puboischiofemoralis externus, 100, 101,
117, 118
puboischiofemoralis internus, 100, 101,
WV, Lie 128:
puboischiotibialis, 104, 777, 118
pubotibialis, 100, 104, 717, 118
quadratus lumborum, 109, 110
. rectus abdominus, 109, 110, 115, 116
rectus capitis anterior, 30, 39, 62, 62,
74,76, 110
rectus capitis posterior, 28, 39, 39, 40,
62,162; 0(2;-445 107
. scapulohumeralis anterior, 84, 85, 89,
P12) 2725 113; VIA
. scapulohumeralis posterior, 84, 89, 112,
112, 113, 115
. semispinalis cervicis, 72, 74
. semispinalis dorsi, 107
. Serrati superficialis, 112
serratus anterior, 83, 84, 85
spinalis capitis, 28, 39, 62, 72, 74, 74,
107
spinalis cervicis, 74
. spinalis dorsi, 107
. sternocoracoideus, 86
. sternohyoideus, 62, 110
subcoracoscapularis, 86, 89, 114, 114
. supracoracoideus, 85, 86, 88, 89, 112,
113, 114, 114
supracostalis, 99, 100, 109, 110, 115, 116
transversalis capitis, 39, 41, 71, 107
transversalis cervicis, 30, 39, 108, 109
. transversospinalis caudae, 81
. triceps, 91, 112, 113
. triceps humeralis, 89, 112, 113
. triceps scapularis, 84, 112, 113
Muscle systems
constrictor dorsalis, 62. See also M. levator
pterygoid and M. protractor
pterygoid
epaxial, 107
hypaxial, 107
iliocostalis, 107, 108, 109, 110
longissimus, 77, 107, 109, 109
obliquus, 107, 110-111
rectus, 107, 110-111
subvertebralis, 107, 110
transversospinalis, 74, 77, 80, 107-109, 109
Muzzle unit, 14, 15, 61

Nasal (bone), 17, 18
Nectoportheus, 131, 132, 173, 181
Nerves
abducens (6th cranial nerve), 31, 31, 35,
37
accessory (11th cranial nerve), 34, 35, 37,
38
acoustic (8th cranial nerve), 35, 35, 36,
36, 37
corda tympani, 52, 54

ea eit ee eee eae re eee eae tele

zzzeezz2

PEABODY MUSEUM BULLETIN 23

facial (7th cranial nerve), 34, 34, 35, 35,
36, 37. See also corda tympani
palatine branch, 31, 33, 34
hyomandibular branch, 34
glossopharyngeal (9th cranial nerve), 34,
FIIs Oo, 09
hypoglossal (12th cranial nerve), 34, 35,
37, 38, 39
inferior alveolar nerve, 50
mandibular nerve, angular branch, 52
mandibular nerve, cutaneous branch, 52
olfactory (Ist cranial nerve), 21, 65
optic (2nd cranial nerve), 26, 37
radial nerve, 90
spinal nerves, 71
trigeminal (5th cranial nerve), 33, 34, 35,
3

mandibular branch, 49
maxillary branch, 17, 25
ophthalmic branch, 16
vagus (10th cranial nerve), 34, 35, 37, 38
Nishi, S., 107-110

Obturator foramen, 100

Occipital unit, 15, 29

Olfactory lobe, 20, 37, 58. See also Cerebral
hemispheres; Nerves, olfac-
tory; Olfactory tract

Olfactory tract, 20, 20, 21, 65. See also Cere-
bral hemispheres; Nerves, ol-
factory; Olfactory lobe

Olson, E.C., 76

Opetiosaurus, 194, 201

Opisthotic (opisthotic-exoccipital), 37-40, 34,

Orbit, 13, 58, 58

Orbitosphenoid, 26-27, 27, 37

Osborn, H.F., 9, 14, 70

Oterognathus, 158

Otic labyrinth, 34, 35, 35, 36, 36, 37, 38, 39, 40,
40, 59

Otosphenoidal crest, 34, 34, 35, 36, 37

Paddles, 90, 93, 94, 95, 96, 97, 102, 103, 104,
105
evolution, 206
movement, 119
Palatine, 24-25, 25, 151
Palatine foramen, 25
Parasphenoid, 31, 33, 34
Parietal, 27-29, 28, 38, 60, 66, 67, 129, 130, 150,
157, 163, 172, 174
Parietal foramen, 27, 37
Parietal unit, 15, 27, 60
Paroccipital process, 35, 38, 39, 41
Pelvic girdle, relations of elements, 97-98, 98
Perilymphatic duct, 36, 39
Perilymphatic sacculus, 34, 37, 38
Phalanges, of forelimb, 95-97, 90, 93, 94, 95,
96, 97
Phalanges, of hindlimb, 103, 104, 105, 106
Phosphorosaurus, 158

RUSSELL: AMERICAN MOSASAURS

Platecarpinae, 12, 148
Platecarpus, 122, 135, 138, 148, 185, 187, 189,
190, 204, 206, 210, 211
habits, 65, 68, 120, 187, 191, 205
affinis, 181, 182
anguliferus, 181, 182. See also Sironectes
brachycephalus, 183, 211
coryphaeus, 65, 122, 153, 155, 187, 188, 208
glandiferus, 183
ictericus, 65, 122, 154, 187, 188, 189, 208,
211
intermedius, 122, 156, 208
habits, 68
latispinus, 182
minor, 183
mudgei, 181
planifrons, 181, 182
somenensis, 122, 153, 155, 189, 190, 209,
211
tectulus, 183
tympaniticus, 122, 148, 152
Platynotan lizards
and mosasaurs, 193-202, 200
hypothetical ancestral forms, 199, 201
Plesiosaurs, 205
Plesiotylosaurus, 123, 167, 190, 204
crassidens, 123, 167, 168, 209
Plioplatecarpinae, 11-12, 122, 148, 170, 203,
204, 205
Plioplatecarpini, 122, 148
Plioplatecarpus, 122, 147, 152, 158, 183, 185,
204, 206
habits, 68
depressus, 123, 159, 160
houzeaut, 123, 159, 160, 210
marshi, 123, 159, 160, 162
primaevus, 123, 159, 189, 212
Plotosaurini, 122, 145
Plotosaurus, 122, 142, 145, 148, 191, 204, 205,
206
habits, 65, 68, 120
bennisoni, 65, 122, 146, 190, 208
tuckeri, 68, 145, 146, 190, 209
Polygonodon, 11
Posterior mandibular unit, 15, 51, 65
Postorbitofrontal, 23, 58, 60, 61, 66, 68, 129,
150, 157,169,166; 172; 177
Prearticular, 52, 54
Prefrontal, 21-22, 22, 58, 60, 66, 67, 129, 150,
157, 163,) 166; 172.7177
Premaxilla, 14-17, 16, 66, 67, 126, 129, 130, 136,
10} 1515) 1575, 1635166, 17.2,
7.
Presphenoid cartilage, 31
Proaigialosaurus, 196, 202
Processus ascendens tecti synotici, 29
Prognathodon, 123, 162, 178, 179, 185, 204
habits, 65, 68
crassartus, 123, 164
donicus, 123
giganteus, 123, 165, 167, 210

229

overtoni, 65, 123, 164, 165, 167, 189, 209,
212
habits, 68
rapax, 91, 123, 164, 165, 209
solvayi, 123, 162, 164, 165, 167, 210
Prognathodontini, 123, 162, 204
Prognathosaurus, 162
Prootic, 33-37, 34, 35, 36, 37, 66
Prootic incisure, 33, 34
Pterycollosaurus, 131, 132, 137
Pterygoid, 41-44, 42, 43, 58, 60, 66, 67, 151
Pterygoid unit, 75, 41. See also Basal unit
Pubis, 98, 99-100, 99, 100, 116, 117, 141
Pumphrey, R.G., 59
Pythonomorpha, 9

Quadrate, 41, 46-49, 46, 47, 60, 66, 67, 131, 139,
150, 151, 160, 166, 167, 172,
175

Quadrate unit, 15, 46, 60, 63

Radius, 90-91, 90, 93, 94, 95, 96, 97
Reproduction, 192
Rhamphosaurus, 170

Rhinosaurus, 170

Romer, A.S., 74, 111, 115

Saniwa, 196-197
Scapula, 82-85, 83, 84, 86, 206
Sclerotic ring, 58, 58
Sella turcica, 31, 31, 35
Semicircular canals, 35, 35, 36, 37, 38, 39, 40,
65
Septomaxilla, 26
Sironectes, 148, 152, 182
Skull
general description, 13-14
structural components, 14, 15
Splenial, 49, 50-51, 51, 52, 66, 67, 136, 161, 177
Splenio-angular articulation, 49, 50-52, 51, 64-
65, 205, 206
Squamosal, 24, 38, 41, 60, 66, 67, 129, 150, 157,
163, 166, 172, 177
Stapes, 15, 37, 45-46, 45. See also Tympanum
extracolumella, 46, 48
Sternal ribs, 74, 88
Sternum, 88
Streptostyly, 63, 64, 64, 65, 205, 206
Summary, 206-207. See also Abstracts
Superciliary, 22
Supraoccipital, 38, 40, 40, 150
Suprascapula, 84, 111
Supratemporal, 34, 35, 38, 40-41, 41, 60
Surangular, 52-53, 52, 60, 66, 67, 166, 177
Swimming in mosasaurs, 119-120
Systematics, 121-184
procedure, 121

Taniwhasaurus, 122, 147, 191, 204, 206
oweni, 122, 147

Tarsus, 102, 103, 104, 105, 106

Teeth
marginal dentition, 54-57

250

premaxillary, 14
pterygoid, 42, 43, 151
replacement, 55
Tibia, 104-106, 102, 103, 104, 105
Trachea, 88
Tylosaurinae, 12, 123, 170, 203, 204, 205, 206
Tylosaurus, 123, 138, 170, 187, 190, 191, 210
cancellous nature of bone, 120
habits, 68, 120, 187, 191, 205
dyspelor, 184
nepaeolicus, 65, 123, 175, 187, 188, 209
perlatus, 184
proriger, 65, 123, 170, 173, 187, 188, 189,
190, 208, 209, 210, 211
habits, 192
Tympanum, 49, 58-59, 60, 131

Ulna, 91-93, 92, 93, 94, 95, 96, 97

Vallois, H., 107, 108, 110

Varanids, 8, 10-11, 193-194, 200

Varanus
bodily proportions, 210
elongation of thoracic spines, 108
motion of hind limb, 117-118
muscles of cervical region, 75

PEABODY MUSEUM BULLETIN 23

skull characters, 13

Vertebrae
caudal, description, 79-82
caudal, number, 79
cervical, description, 70-76, 71, 74
cervical, posterior limit, 74, 74
dorsal, description, 76-78, 74, 75
dorsal, number, 76
intercentral articulations, 75-76
lumbar, 77
pygal, 79, 127, 206
sacral, 78-79
thoracic, 76

Vidian canal, 30, 31, 33, 34

Virchow, H., 108

Vomer, 25-26, 151

Vomeronasal organ, See Jacobson’s organ

Walls, G.L., 58

Williston, S.W., 5, 12, 14, 70, 119, 132, 136,
192

Wiman, C., 127

Zygantrum, 75, 75
Zygopophyses, function, 108
Zygosphene, 75, 75

MAESTRICHTIAN

—
=
=
=
a
=
=
oO

SANTONI AN

CONIACIAN

RUSSELL: AMERICAN MOSASAURS 23]
CHART 1
WESTERN GULF MOSASAURS

Correlation after Stephenson et. al. 1942; Young
1963; Jeletzky 1965 personal communication

Time scale after Kulp 1961

NAVARRO Mosasaurus maximus

Mosasaurus conodon Dane 1929

Tylosaurus proriger
Clidastes (USNM 22964)

Globidens alabamaensis

Globidens(?)(USNM 22965)

Brownstown Platecarpus (MSSC)

AUSTIN

Clidastes Hill 1901
mosasaurs Adkins 1928
EAGLE FORD

232 PEABODY MUSEUM BULLETIN 23
CHART 2
EASTERN GULF MOSASAURS

Correlation after Stephenson et. al. 1942; Young
1963; Jeletzky 1965 personal communication

PRAIRIE BLUFF

MAESTRICHTIAN

RIPLEY GROUP Mosasaurus maximus
(INCLUDING COON CREEK) Prognathodon(?)

DEMOPOLIS Mosasaurus (USNM 12287)

Arcola Limestone

Clidastes propython
Globidens STabamaensie an
Platecarpus Dowling 1941

MOOREVILLE "Platecarpus" intermedius(?)
Tylosaurus Renger 1935

—
<=
=
=
a
=
=
oO

Tombigbee Sand mosasaur (USNM 3026)

SANTON | AN

Platecarpus tympaniticus

CONIACIAN

90 TUSCALOOSA

MAESTRICHT IAN

=
—<c
=
=<
a
=
=
Oo

SANTON IAN

CON! ACIAN

RUSSELL: AMERICAN MOSASAURS 233
CHART 3
NEW JERSEY MOSASAURS

Correlation after Stephenson et. al. 1942; Olsson
1963; Jeletzky 1965 personal communication

Mosasaurus conodon
TINTON, REDBANK, Mosasaurus dekay!
NAVESINK AND EQUIVALENT Mosasaurus maximus
NEW EGYPT Liodon sectorius

Plioplatecarpus depressus
Prognathodon rapax
Halisaurus platyspondylus

MT. LAUREL
WENONAH

MARSHALLTOWN Clidastes iguanavus
ENGLISHTOWN

80
WOODBURY

MERCHANTVILLE

MAGOTHY

234

MAESTRICHT IAN

we
<<
=
=
a
=
—
oOo

SANTON I AN

CONIACIAN

PEABODY MUSEUM BULLETIN 23

CHART 4
MOSASAURS OF THE INTERIOR

Correlation after Jeletzky 1955, 1962

HELL CREEK

UPPER PIERRE AND
BEARPAW

LOWER PIERRE

EAGLE SS
80

TELEGRAPH
CREEK

NIOBRARA CHALK

Rudistes beds

BENTON
SHALE

FORT HAYS
LIMESTONE

Hesperornis beds

Mosasaurus Gilmore 1912B

Mosasaurus missouriensis

Prognathodon overtoni

Mosasaurus conodon

Plioplatecarpus primaevus

Clidastes propython
Globidens alabawaennie
Platecarpus ictericus
Platecarpus somenensis
Tylosaurus proriger

Clidastes propython
Clidastes sternbergi

Mosasaurus ivoensis

Platecarpus ictericus
Ectenosaurus clidastoides

Halisaurus o onchognathus _
Tylosaurus proriger

Clidastes liodontus

Platecarpus coryphaeus
Tylosaurus nepaeolicus

MAESTRICHT IAN

=
=<
ro
=
a
=
=<
(&)

SANTONIAN

CONTI ACI AN

RUSSELL: AMERICAN MOSASAURS
CHART 5
BELGIAN MOSASAURS

Correlation after Leriche 1929, 1934; Marie 1956;
Schmid 1959; Hofker 1959. 1960A, 1960B; faunas after
Dollo 1924

TUFFEAU MASSIF SANS SILEX Mosasaurus hoffmanni
CRAIE GROSSIEREA SILEX GRIS Compressidens fraasi

CRAIE BRUN PHOSPHATEEDE CIPLEY

Mosasaurus hoffmanni
Plioplatecarpus marshi
Compressidens fraasi
Mosasaurus conodon
Compressicens belgicus |
Plioplatecarpus houzeaul
Prognathodon solvayl

70 Prognathodon giganteus
Hainosaurus bernardt

RAIE GROSSIEREDE SPIENNES mosasaurs
CRAIE DE NOUVELLES mosasaurs

CRAIE D'OBOURG

RAIE DE TRIVIERES mosasaurs

Liodon compressidens

Globidens alabamaensis

GLAUCONIE DE LONZEE Mosasaurus lonzeensis
Hainosaurus lonzeensis

235

236 PEABODY MUSEUM BULLETIN 23
CHART 6
FRENCH MOSASAURS

Correlation after author's name

Prov. of Charente Infer. "Liodon" Zurcher & Arnaud

Prov. of Basses-Pyrenees Liodon mosasauroides Pence
2

MAESTRICHT IAN

CRAIE DE MEUDON Lerichel934} Liodon anceps Gaudry 1892

CRAIE PHOSPHATEE DE LA Mosasaurus gaudryi Thevenin
PICARDIE Leriche 1934 1896

Platecarpus somenensis
Thevenin 1896

CRAIE DE MICHERY Sornay 957 Liodon compressidens
Gaudry 1892

BOE ov. of Aude Platecarpus ictericus
Corray 1927

=
<
ae
<<
fol
=
<
Oo

SANTON I AN

CONI ACI AN

RUSSELL: AMERICAN MOSASAURS 237

CHART 7
ENGLISH MOSASAURS SWEDISH MOSASAURS
Correlation after Cox 1962 Faunas and correlation after

Persson 1959, 1963

MAESTRICHT IAN

?Liodon anceps from the Liodon anceps

"chalk of Norfolk" fide Plioplatecarpus
Owen 1851

?Platecarpus from the Mosasaurus ivoensis
"chalk of Sussex" fide Platecarpus somenensis
Lydekker 1888

=a
=
—
=
a
=
=
oO

80

SANTONI AN

CONIAC1AN

90
mosasaur Smith Woodward 1905

PLATE I

Figs. 1-3. Restorations of mosasaurs, after Williston (1898a, pl. 72). 1. Clidastes
liodontus, X 1/16. 2. Platecarpus ictericus, X 1/19. 3. Tylosaurus proriger,
x32:

239

PLATE II

Figs. 1, 2. Restorations of mosasaurs. 1. Tylosaurus proriger, x 1/34, after Osborn
(1899a, pl. 23). 2. Plotosaurus, x 1/47, after Camp (1942, pl. 5).

240

yyy
AANKUA LUNAR

ZN NaLDyyyyyy :
SHIN

bl dst

U

hi

iH,

Vir i ae
y | y 5 i

Ly
ALT

rua Waa ve , ueaty

cy : al

ow ha As = o = a
x 5 = 5 calle 5 2
> Es E > i ci
z = 3 2 F é
m z A z iB we ,
ON NOILNLILSNI_ NVINOSHLIWS $3 luvud Mul BRARI ES SMITHSONIAN INST OO
g S = = es = <
= z = z ») aa z :
Gas [o) a5. QO : ; 28 ro) 7
wn ” Ny ” no ” Y
= E z F < E 4
= > = > 3
= Fe 7 Zz a 3

Th LIBRARIES SMITHSONIAN _INSTITUTION NOILALILSNI NVIBOSH ENS saiuvuei

NOILNLILSNI.
LIBRARIES

AICAP Tmt oar

NOILALILSNI NVINOSHLIWS S3!1YVYdIT LIBRARIES SMITHSONIAN _ INSTITUTION

Y SS
«
S
>

INSTITUTION NOILNLILSNI
S3SIYVYEITILIBRARIES
INSTITUTION NOILNLILSNI
S3JIYVYUEIT LIBRARIES

JAIQTWITII TIP VAL

Saiuvaadil

INSTITUTION

117 EaIBRER IES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S31u¥vugit

NVINOSHLIWS
SMITHSONIAN
Z

Lee
NVINOSHLINS
SMITHSONIAN
NVINOSHLIWS

NOILALILSNI_ NVINOSHLIWS sa1yvualy LIBRARIES SMITHSONIAN INSTITUTION

Z
=<
z
3)
22)
Se
E
=
77)
z 3 Z 3
_ + - el — -
ina = a = a % = c
: 2 : : 5 }
m fe) = fe) a ) :
J a = ie a a '
5 ae BRARI ES_ SMITHSONIAN” INSTITUTION CHOLNELLSNEAyIN CSTs Ssaluvud 7
a = ‘ z
w S 4M ° ow ° |
es) S 2 = es] =
> 5 > a > ae =
2 - = F = |
D z o z 2 ie: |
ION NOILNLILSNI NVINOSHLINS S3IYVYEIT LIBRARIES SMITHSONIAN INSTITUTION
n z 2 cs . 2 z “ |
z z = ‘S &s = ZK
5 Zz = Zz \\ Sy isd z KS
x (e) aoe (e) 4 me Oo SAC
22) ” wn ? oe NS B ca WEL
3 f 2 iS z = |
> >
S 5 S a FS a

Th LIBRARIES SMITHSONIAN _ INSTITUTION NOILMLILSNI NVINOSHLIWS saiuyagl7

= ” — ” = ; ”
o a 1a pe a ANG Ww
a « = a a WAS cc |
= < 3; = < =A . < :
= - = : So OE |
fe) = fe) _ fe) = |
= = z a] z eee =)
ION NOLLA.LILSNITNVINOSHLINS — 3 luvud 4 yt LIBRARI ES_ SMITHSONIAN INSTITUTIONS
z |
fo) = ° ea ° =
\ = = s = is |
;) 2 > = > = SD |
I= 3 az res zw Coed aD |

= ee] & D E = Is
| e y > = > a > = 5
a : 5 : 5 ; :
eon 1G z Ys Z 4h z
NVINOSHLINS S3IYVUSIT LIBRARIES INSTITUTION NOILNLILSNI N
z 2 z se Y z a n z
< = s = < gs = <=
. ~ =
5 =. 5 > 5 Wi niece 5
r 6 f o 2S 6 ¢
E 2 E 2 = ae =
> > >
5 2 a 2 a Beye FA
SMITHSONIAN INSTITUTION NOILOLILSNI SJIYVYSIT LIBRARIES SI
” wo i?p) zs nn
= ‘ = = wo uu
= wh fea z =f
<yvZ < < = %
& a oc Ss .
oy 2 = S)
=) J | z

S3IYVYSIT LIBRARIES

NVINOSHLIWS NI
.

SV WSS
x S
SAS

SMITHSONIAN INSTITUTION NOILNLILSNI NYINOSHLINS S31yuyvyugII

SMITHSONIAN INSTITUTION NOILOLILSNI

1

INSTITUTION NOILNLILSNI
INSTITUTION NOILNLILSNI

Saiuvugit
INSTITUTION

saiuvug

LIBRARIES Sh

NVINOSHLINS S31YVYUG!ITLIBRARI

ILSNI NVINOSHLINS S3ZIuNvYaII

za x ” = ” z
zs = < = <
Zins = z = =
5 NW: 9 Z B NS
EF AG 2 - 2 SENN S
ne ° 5 Bon
—NVINOSHLINS Saluvugit LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI_N\
— =. wo
b ws a tu BR wW &
=a = 4 = 4 re =
e < s < {5 < =
a be = 5 co = oc 4
Ss ill 3 a 3 = S
z ; a 2 ey = J <
|_ SMITHSONIAN INSTITUTION NOILNLILSNI! NVINOSHLINS S3ZIYVYSIT LIBRARIES SI
z = = r oz i =
© _ ° = Xe re) = fe)
4 -_ ow <A —_ ao =a
od vices x (= = r= i=
2 YY eS) a > 5 =e] 5
= % = = = = = Ce
7" = F 2 : = Ee
2 is Z m 2 m z
ENRDSHLNS Saluvegty egSMITHSONIAN _ INSTITUTION NOILNLILSNI Ni
Fa z on Zz ars BS 5 |
an n ” 2) n * ‘ Te ”
24 QO % Be QO bn (ole ae
= 2 E Zz E z =
2 2 ee : :

| SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IYVYSIT LIBRARIES SN

” = ” = =
By We! yo ie SG 4 a us
= iy ips : a ae : :
<7 c = 2 WA oc a < ly
« = x 5 S\, « =r -
ao ‘YL - ao = isa) _ [ea]
- Yo {e) - o) - oO -
=) Z 4 z ae ag 20 = ~
NVINOSHLINS ~S31YVY8IT LIBRARIES SMITHSONIAN” INSTITUTION NOILNLILSNI NV
: z
& == wo —_ “4, ow oo ow
5 2 = G Wy, Pe) = Pe)
= = e Wiss Ee =
Ee a ae Ue ae = a

SMITHSONIAN INSTITUTION LIBRARIES

APIA A

3 9088 01 0164