fay eas Slat te 5 aA ONE RUPEES Nant sh aay Toh rae ws sca TNS re. oe seks CONIA TAP ARNEL ads Ne PALI BS mah Ma NoRS ER Oe ey es Capa eand ete ices en Tu R TREAT. a ako or) Pe BL Sead ees ' he, es a yee g ily AD FS PALME, TON yy AA YE rer ea Asan eR RR uate ets” se erh Nie tay 4 SINAN NT ee. ete Santen’ fag Mt MEALS Like . Wade hs temas TOR Ame ty! Macatee Oe nets tants anes seek Anse et eye Saas heme eee ame NS meses eR DERE sb yee agen A fete Rs MEAN! Atanas SR Reae deh NaS , abet ey ee, tm 8- ek ee EMU tars om hee, aac UN ara hy rer er ae Hay ta tytn ke We nate gy SNES ~ ron VRE NS Mab awa ntine = ethene te anide tsbrer ve Peer Be elas AR SI ar SE ease e Pt feet 2 rien ot Bien: ‘Hip ob awd ah Rte arr oe ai ope Pee prints Pe ae Pins tee (ne Pee (posse nate Soy see a eer Cibey Baers Tiga ere apes ae Taken eee eared po er ee Pat ale ree ee Pile ta FO ye aereas ile Ee rem ap Ber arlene - Berean ° Phe pe Say on yale end eG PO REIT RPP Ser aa a Paoane peat Ml : : fate IN, - are ‘ ae rs atahade &. as aay Tae . . : ee 7 “kate hae Vie we Saka Tat VITA te NERA ; Re BAe aie Hemet aan i : Ses ueane Macey Chon Td : OPE RteD Pe Soha ee Sere ty i . ve ‘ RATA TEE i \ ‘ cer eere) Whee sah : “ a aese. Ne detegn na ; : : wagagte UP uN RE MERTENS i mee ee s . : " feo be Meet bie alt . epee aN - 4 see oa ; ab tea ny ; , : ert BPS ove ete sere rece , 8 A go ener a t ‘ ae ° J We ane * Rr es ; oa sats i ae : 3 ae cutbars i : H ‘ ey ere Peer oes a ty 2 Tetum Seew, mak Ny, RAD: a Wine eee. woe yer re at phe ies OR aad pais ies N en rem eS eae The oe ab ae 6 PBC arate are OO ik ade BITC I Te Seat spake Le Oe weees ale the Sasi tt a agerpe Gib gn bet DOr ore po SEIT EES Nill ay pas ep erase! rn ee hiss anc caved sed SK epee nah OPENS SSCL ee eeee pene NEUE DEE 4 Pa eo vey wee regen we walsers yp (0 7" seaeauay ee ae By oie yer tt etc geivegitg eee SE eres aavi d eX my and Relationships so) BLS a Corosaurus alcovensis ¥- nf JUL TG 1991 _(Diapsida: Sauropterygia) and the Triassic Alcova Lim BE ; ae OAS sy, 5 eee rid zs Bae Pw . Anatomy and Relationships of Corosaurus alcovensis (Diapsida: Sauropterygia) and the Triassic Alcova Limestone of Wyoming GLENN WILLIAM STORRS Peabody Museum of Natural History Division of Vertebrate Paleontology Yale University BULLETIN 44¢@9 JULY 1991 PEABODY MUSEUM OF NATURAL HISTORY YALE UNIVERSITY NEW HAVEN, CONNECTICUT 06511 Bulletins published by the Peabody Museum of Natural History, Yale University, are numbered consecutively as independent monographs and appear at irregular intervals. Shorter papers are published at frequent intervals in the Peabody Museum Postilla series. The Peabody Museum Bulletin incorporates the Bulletin of the Bingham Oceano- graphic Collection, which ceased independent publication after Vol. 19, Article 2 (1967). Communications concerning purchase or exchange of publications should be ad- dressed to the Publications Office, Peabody Museum of Natural History, Yale University, 170 Whitney Avenue, P.O. Box 6666, New Haven, CT 06511, U.S.A. © 1986, 1991 by Glenn William Storrs. All rights reserved. Printed in the United States of America ISBN No. 0-912532-03-3 CONTENTS | ea SSD PCO) EG) hy 71 22a Oe a ee a RO nee era MarR rye A eis Sie he le v1 AS et @ ale ASG Si ieee Dee ce oc deci ee ele Be eae a a eo VIL EIS RSORCAB BRE VITATIONS exis oo cet oh ee vite AGKINOWIGEID GE MEEINES: 225). here oe oat OR ee ee eee X11 J) BASS] BREN Cif Ries ge ae ae ret a ore ene ee ree roe | ee = ieee 1 STE AV AV VOIRSID)'S Site ate rs ce eG ee ee een eee rae ae a 2 ZO SAVINGEINEASSUINIG:? 5... 5. Ue Be aso Pe ea ine orks ee oe 2 PING @ DW GaIOIN oe bs oe) Sot Gk Pe ei Se Bae Ge One a, ae pe cee 4 FNIStODy.OlMlnVestiGatiOm maser wai. sf. c. 016 es 400 1404-4 rate eR 13 PIAS KELELOMI icra neta ites fay kc ook be od ee ee a oO Te 13 SIRS rence Pa ey aerate ee pt fst a ca Sts Se ee eee 13 ETNA aes cate cape eee ie, ar eee ee en a 15 NES aU Bogen ae anne ee ORES ER Tr mee St. lg | Nee Oe wee 15 INAS all Sr Ss aes seo at og, 2) Ge he 2, a eae 5 ERE MTOMMGANS a Bene Peele eid od aah ene a ee i aE 15 FRONMA She Sh retire cette ese yen cons Bec ee ee RR ee ee 16 PAICUANS RY og - Pheer One at ese, Ue es 4m, Gan RE Ee 16 ] EOS eo) G1] 2K i A mR RRL eA g Om Nl Pos Dg 16 AOSEOD ICA S eps. ee Ea cates 2 useless & ee em 16 PUA eecin oo ee circ a A as ga, Shc Oho, Gly, Se me Pe Oe 16 SOUAMMIOSANS: £ eiey ate te eae ela ee eee eee 7 Ouadratesr casos eds oi nas Se ee ee ee 17 BRAINCASEH eas A ae hee Fakes Re A eee eee 17 ANAT AF ee ea NERS ace cay be Be sk ee en 18 Miandibleg Giles 2 3.43 G83 oak St eee eae, aes 19 WertepralyColumingn: sc 22.4 £23.23 eee eee. 20 @ervical Vertebrac-and. Ribs... 54-5 eee 20 DorsaleviertebracandJRibs; -..- 2s eee ee 21 Sacral Ventebrac and wR ibs: -a4...2 5 oe ne te 22 @audalaViertebraceand™ Ribsi: --a eee eee ae 23 Gasthaltateetaay ar: 2002 3 4 Mos Marae ee eae eee eee ee a 23 ~mppendicularsskeleton:: <. <4 asih< 2 e e a e e e 24 RECUR UMI anh tt ss Chee ld he eke ence ee a ey Pe aL AR 24 CAN ICIER eee iis aeons autre, en Ace ee Oe Asie oe ee 25 Mntenclaviel Gis ax. tee 3 eee oe on ea Oo ee aN erate (a eee 26 Scapulay gee tc es. ec, 5 Be oe ae ee ae ea ee eg 26 CORACOUd! Coscia ot ht Bs i ee I eh kd tuk Dif INEStOTATI ONE MC fe occas ey eh Ae oe ee anne cee aa hE ras 28 J ELEIKVISR ea ee ae tae Winer eRe Ran Oana Mea een RMeRUnee MOreew fae 29 | 2A] OTe eee Veter mn. NM) oe Se Lee ONE. eee ae See i 29 RS Gh een pt res cake Ee ee ee ne ie Pe A, 30 Ps uit o.oo seas a Slee op pepeare } Sao ee ee ee 31 IREStOratiom 6 oe bg 6 ne G6 ee te ee a | POrelmiby 096 ee ee eee 32. PWM RUS: 5 cars fe dees. LR et Thess Ae 29 eee ee 33 Wn ee Polo ete Oe Lg BATRA ee 35 Radius... 2 ie syne as, 0 eee a5 Gai DUS oo ee ates enka AC oo eee och aed ea ee ee 35 IManuSi=, cute, 2 ina Ser ete eee fen ASU Cone Wee ee er a5 Flurl lnmly-4 Ow rehash Rae 2 Oe os yee S35 QTM OAR gd pte sn eagle See ve 36 Tubiacand: Biloullas 2 cs oc. ga en oe ee See oho ee ee a7, Mansustand Pest. Ps 2.:-5 ast guscoucher s See cues 0 oe ee ee Sif IRestOnationie.turt.< ga’ eect eee oe oe ee ee 37 Discussionae nas bs heel a ee Ee a) ae OBIOUOGNG 4 oor a. get 59 io Braet ach ponatenn ake renee ee ee 41 PMItROGUEHIOME 18 vor: ns take meee ses a ge oA ee ee 41 Ontogenetic-and Individual Variation --- 5245.2. 4....-+-+ 0040" ee 41 PunctionalMorphology *s. i 4240.2 Seer Se oer ee ee 44 Aquatic Locomotion. <4. 88.0. oss oe eae els ee ee ae Fly potheticalsMyolosy 4.2: «224.0 n eta ee ee 45 DISCUSSION, 25.55 2 ones reed Bue Lee ee oe ee 52, Merrestrialocomotiony: 4373 0.082 oe oe Pa a oe 53 Baleoecology . faite te ee ete Sats Rs Oe 54 PLL VEOGENY ANDPEAXONOMY) 2.208 2a ae eae ee ee 56 INGROGUICHIONs! ake CMe saree ee ee Oe eee eee 56 Fistorical’Goncepts of the Sauropteryeia "50 5...--2- 92. 1.2 eee Si Flistorical Classifications of the INothosauria ~. 3... -2 2224.40 =95e8 58 itheGeneravor (Nothosaurs,.atere tc sa snes ce ee ee 59 Sauropteryeian Origins and ‘Relationships ..2.: 2... ..2.-: a2 -enee 74 JNothosaune Relationships, (42 45en see cen as oka AF hm Sec ee eee 83 PhierarchicaliG@lassification.. 2. aii essynosmss eet cise «ae in eke ee 87 OP RRCACITIG ROAR EIN oso sioly tn) ere ee hud ss eee 90 Introduction. 3 5124 26s eas oy I eae es 7 AS eee 90 Regional Setting). = 62. o eee ee oe boo ee On Physical Stratigraphy and Petrographiy.5- 45 a. =. 2. ae eee 92 Distributionvand. Whickness:. 40) S sotews. ~ 2. cir hu 3 See eee 93 PetChOlO Ry easy: ese rs oa cal on ee im ee el ty ti Pte ee 94 INDMeralODyinc tee cok foes os eR a. Sas whe 2S Aaa ae 95 Carbonate: Minerals 45 425.3 sag Back ma OAs ae Ae ee 103 Physical’ Sedimentary Structures= 422 325 5.45554 ee ee 103 Biogenic: Sedimentary Structures: <0. 22 -.ohace 4,5 ee ee 104 BossilvAssemblace sx 5< 2.7 a as 8.3 Pk oe oe eee Pa eee 105 VertebratesPaunae * i523 44.2 5 oes: 2e eee weet ee eee 105 Invertebrate Fauna: 2.02055.) oa Stee eee ee eee 106 Algal Plorai. 42 2 t2cc..0 2s A Se Ree ee ee ee 110 CC COCHEMIS UR Viney 5. 2k Oar Beg pe ey Mer OS Beh ves 110 Carbonvisotopic: Composition 2... 5 5 .r tg aa ee ee 1 Oxy cenlsotopic Composition. seo hs) = ceo ae ae es eee 113 ACES MINE RELATI OM es 64 Soot faces Gye Uk es Vg ars aa Oe Coe 113 ie PALER OBIOGEOGRAPELY. }. 5.1.5 eee Raritan 105 PntROducti ones cathe. is:c 2S Uae ce Oe I SO ean 15 iNothosaurGeographic Distributiong:-. 2 4a ee 4. ae ee See 115 eNothosaurNemiporal Jane. 22... ac eee ere? oer eS Sees GUT Nothosaursbtabitate Diversity. ... =. taeeeeee ee eae ee eeeeetes oe) Sir VEIN AIRE Peter GE ok 2 af, te eee ene ee ae Pe RE Naot ae 124 Pe Tae Ree CRE CBD Tce ices-s ok csc 5 en EN cs a 123 APPENDIX A. COROSAURUS ALCOVENSIS HYPODIGM ......... 132 AEN DIDX Bo NOTHOSAUR GENERIGINANMIES =) 5) 250) es. 134 APPENDIX C. SAUROPTERYGIAN CHARACTER STATES ZANT) EN RG SGN GANG 3 ee ee ee 141 OSOMANANANIBRWN LIST OF FIGURES MiLocality map iv. fly 8D Uo Aah sea ees Se ee ee 6 Geologie Maps -.o25.024 Dee tees fa ae ee ee ee 7 + Holotype ol Corosaurus-alcovensts 25 haa. 9 4 Partial ‘skeleton, FAMINE PR4AS80...5. 22.650... 342 4622 eee 10 i Partial skeleton, EFMINH PRIG69s-ek Saas ee 10 + Partial’skeleton,. YPMi41031 «02... ee nee Ee fl + Reconstructed: skull of Corosaurus ...2)232) ss. eee 138 HReconstructedtskull of Corosaurus 22.2 ..5. 62. 41 e eeee 14 » MandibleofiGorosaurus:...<%. sse oss 2 oe a es ee 19 Axial skeleton of Corosaurus . 22 2 Si ee. oe ee ee On + Pectrum*of Coresaurus.>. o-oo le A ee 25 + Pectrum\ot Corosauriis s s2 28 Boe ae oe oe 2 eee 26 wReconstmucteds pectrumot Corosaurus.: 8. o.oo eee 28 s Reconstructed pectrum of Corosaurus 7. 2 a 2 ee Za) ie PelwiskOimC O7OSaQUiUS rice, Deere oa ee ee eee 30 wReconsigucted pelvis of Corosaurus...2.. 45... 22204 eee ei! ) Reconsinucted pelvis of (Corosaurus .-- 52.0. 25.8 ee 32 S PUIMERUSEOINCOnOSQUTUS Se. =). Al. 6 od ca OL oe 5 eee 33 Pore liimbsOlGorOSQunuSors s502.0. 2A ac ues ed See 34 POLemlime@lc GO7OSAUUSmNe es os, «cz 0 obs ee ee 36 WReconstruction Of (Ce7osaurus < - oon Gt 2 ee ee 38 WMRECONStructionwOk Corosaunus: s...6..400.65 5. Sos oo 38 MVE C ANCTCOUSAUTUS Wein hice 2 stats eh ym ee oes es ee 42 2 rliunmeralkontoseny ol Corosaunus n-) 6 sae ee eee 43 ~ -lumeral musculature of Corosaurus .......2..0. 525-525 eee 46 ~ Pectoral musculature of @orosaumus ...2.......0.0554008 soe 47 wPectoralymusculature ol Cornosaurus ... 0)... 0. 25 ee eee 48 7 Femoralmusculature of Corosaurus 4... ...5. 4.0505 508 ee 5 Rachypleurosaur skulls > 2225. shies + eis oe On Be VARIOUS: NOtLMOSAUL. SUlIS! . 2. apes oe. on) ee ee 68 SNotmosauritorm’skulls' - 2. 520.42 Sees teen ace eee ee 69 sNotnosaur ipallates: =. 2248). . Bis oo stan 2 ow LOS eRe eee 70 + Various nothosaur skulls 225: Ak. ee. 2 oe ee ee 71 INothosauriform “nothosaur’ skulls... 4. .... ...<5 2.4.5). Saeee 72 _ J Nothosaur#oceiputs’ 2 a. . an Seek ks YS eee 1S EVING@ENOSAU ECan {s&s n es See weer eiccs es 2 ae 74 ys Nothosallk Pectidews Winassie' stratioraphy- of Wyomine <2 ices coe-.050 22 eee of PH lielassicipaleogeoeram my icc sass oe Sor Gee ee 92 Distributionsoicthe Alcova, Limestone =... ....%22 0-2 eee DE Aleovakimestone: Member am outcrop. 2.2555. so ee 95 - Aleovatalgalubeddinge Set Goe.i coe.) oS a ee 96 wAlcovarelectricNoes: (4. .c5 ste na eet ee eee Cea 97 -AlcovarXenay dittraction.: . ic. s.92.0uens coe ees ee ee ee 98 eA leovar thineSechions 2 ...2252 sas ee 99 « @onrelationsot thesAlcova.... «aoc .eo-4ocn eee 102 leeintraclastsconglomeratera vaste. ak aves eee 4 ae ne oe aes 104 SO Possillesbiotunrbation:. 2)... Ak ca) ee ee ee oe 105 Ho? Alcovaspelecy pods. .< .. 314) Rais, ea). key, often Meera) AA a te 107 54 Alcovaspelecypods:s, 25.0454 4 2+ 0. SUA Sen eee Aa eer are 107 Soee Al COVA PElECyPOdS: sc .n« st RR) ie Gs gape 108 JOnpAlcovaspelecypods*.9 2s. .titna.) teihaeee heh eerie: 108 DPE NICOVARDELECVIDOUS a.acs 6 a. ae a SEE ee: Ree 109 DOM NI COV AES ASUFOPOG fre) 2 = ...urs)5 8 2 ues se ee 110 Dee NICOVaStLOMATOMUESHIn Aft. Aeaasieess. 0:6 EALIA 4) cme ere oe en: et GUMAICOVassthOMAtOlItES 4 24... ok 42 Saas Ahk eee a ee 1 Gl Nothosaurcdistributiony m4 7580 eal olds We ed es Qe 116 62s Nothosaur stratigraphic ranges. ...2........... . eee 118 VIL LIST OF TABLES fo) Historical classifications of the nothosaurs’.- «gers eS eee 60 2a) Sauroptenyeian Character states: <<% © 20262. 5 ee ee 78 32) Nomenclature of the Chugwater Group ..... --2heee sper 90 42 Environments of the Chugwater Group: .-:-. aes ee Seer 114 LISTS OF ABBREVIATIONS The following are lists of abbreviations that appear in the text and accompanying tables and figures: dent or ecpt ect ect f ANATOMICAL ABBREVIATIONS angular acetabulum M. adductor femoris basioccipital M. brachialis basisphenoid carpus M. coracobrachialis brevis M. coracobrachialis longus M. caudofemoralis clavicle coracoid coracoid extension coronoid process dentary M. deltoideus (undivided) M. deltoideus clavicularis M. deltoideus scapularis dental orientation ectopterygoid ectepicondylar notch ectepicondylar foramen external naris entepicondylar foramen exoccipital epicondyle epipterygoid extensors frontal femur flexors foramen magnum M. femorotibialis foramina subcentralia gastralia glenoid VULL max can n ) obt f op Pp p for pachy humerus hypantrum hypospene hyperphalangy interclavicle M. iliofemoralis iliac blade ilium internal trochanter intercentra ischium M. ischiotrochantericus intertrochanteric fossa jugal lachrymal M. latissimus dorsi primary muscle maxilla maxillary caniniform nasal orbit obturator foramen opisthotic parietal parietal foramen ‘pachyostosis’ palatine parietal foramen pectoral bar pectoral fenestra pectoral rib M. pectoralis postfrontal puboischiadic fenestra M. puboischiofemoralis externus M. puboischiofemoralis internus palatal dentition pelvic canal premaxilla premaxillary caniniform postorbital popliteal space postparietal prefrontal prootic pterygoid posttemporal fenestra pubis posteroventral ridge quadrate quadratojugal retroarticular process sa sac sacr sac V sbcsc Sc sc b sch cr so spa int Sq st stf sub f sup sup Cc surangular sacrum sacral rib sacral vertebrae M. subcoracoscapularis scapula scapular blade M. scapulohumeralis cranialis supraoccipital spatium interosseum squamosal supratemporal supratemporal fenestra suborbital fenestra supinator process M. supracoracoideus supracoracoid foramen M. supinator longus supinator ridge symphysis tooth temporal emargination tarsus thoracic cavity thyroid fenestra tibial condyle vomer zygapophysis zygantrum zygosphene ‘TAXONOMIC ABBREVIATIONS Anarosaurus Ceresiosaurus Chinchenia Corosaurus C'ymatosaurus Dactylosaurus Keichousaurus Kwangsisaurus Lariosaurus Metanothosaurus Neusticosaurus Nothosaurus “Pachypleurosaurus” Paranothosaurus Pistosaurus Proneusticosaurus Rhaeticonia Sanchiaosaurus Se Serpianosaurus Sh Shingyisaurus Si Stmosaurus STRATIGRAPHIC ABBREVIATIONS € Cambrian Fm Formation fms formations J Jurassic K Cretaceous K/J undifferentiated Cretaceous and Jurassic Imst limestone mbr member P Permian pe Precambrian P/C undifferentiated Permo-Carboniferous Q Quarternary ss sandstone T ‘Tertiary MUSEUM ABBREVIATIONS BMNH British Museum (Natural History), London FMNH _ Field Museum of Natural History, Chicago UW University of Wyoming, Laramie YPM Yale Peabody Museum of Natural History, New Haven MISCELLANEOUS ABBREVIATIONS A angstroms cm centimeters Co County Ce) molecular layer spacing distance d difference E east km kilometers r wavelength of incident X-radiation m meters mm millimeters mtn mountain N north PDB Pedee belemnite isotopic standard R range r-| long spacing resistivity r-S short spacing resistivity Sec section SMOW - standard mean ocean water isotopic standard sp self potential x1 at tier 6 ¥Y, angle between incident and diffracted X-radiation W west WY Wyoming ACKNOWLEDGMENTS I am indebted to Dr. John H. Ostrom of Yale University for his patience, guidance, and supervision during the preparation of this manuscript. Drs. Keith S. Thomson of the Academy of Natural Sciences, Philadelphia, Karl M. Waagée of Yale (ret.) and Rainer Zangerl of the Field Museum of Natural History (ret.) also deserve special thanks. I gratefully acknowledge the help of Drs. Jason A. Lillegraven (University of Wyoming), and William Turnbull (Field Museum of Natural History, ret.) who graciously allowed the extended loan and study of specimens in their care. Additional thanks go to Drs. Robert L. Carroll (McGill University), José L. Sanz (Universidad Autonoma de Madrid), Sabine Schmidt (Universitat Tubing- en), Kim A. Waldron (Edinburgh University), Mr. Kenneth Carpenter (Denver Museum of Natural History), Drs. Alick D. Walker and David S. Brown (The University of Newcastle upon Tyne), Alan J. Charig (British Museum (Natural History), ret.), Niall J. Mateer (University of Wyoming), Mark A. Norell (Amer- ican Museum of Natural History), P. Martin Sander (Universitat Bonn), Adolph Seilacher (Yale), Richard A. Thulborn (University of Queensland), Grant R. Woodwell (Mary Washington College), J. David Love (U.S. Geological Survey, ret.), and Michael J. Parrish (Northern Illinois University) for information and assistance which they so generously provided. Logistical support was given throughout the many stages of this project by Messrs. John C. Brunner, Brent H. Breithaupt, Robert Allen, Richard Board- man, William F. Simpson and Miss Sandra Chapman. Special thanks to Robin Storrs. I thank also the Milne family of Freeland, Wyoming, for extended access to their ranch land during the field summer of 1983. Earlier versions of this manuscript were reviewed by Drs. John H. Ostrom, Keith S. Thomson, Karl M. Waagé, Rainer Zangerl, and Robert L. Carroll. Their help is greatly appreciated and remaining errors are my own. Financial support for this project was provided by: Yale University James Dwight Dana and Louis Pirsson fellowships, the Department of Geology and Geophysics, Yale University; the Division of Vertebrate Paleontology, Peabody Museum of Natural History; the Sierra Madre Geological Foundation; and the Woman’s Seamen’s Friend Society of Connecticut. The author would like to dedicate this paper to the memory of his father, Emory Parker Storrs. xUL YALE UNIVERSITY PEABODY MUSEUM OF NATURAL HISTORY BULLETIN No. 44 151 PP., 62 FIGS., 4 TABLES, 1991 ANATOMY AND RELATIONSHIPS OF COROSAURUS ALCOVENSIS (DIAPSIDA: SAUROPTERYGIA) AND THE TRIASSIC ALCOVA LIMESTONE OF WYOMING GLENN WILLIAM STORRS ABSTRACT The ‘Nothosauria,’ a traditional suborder of Triassic marine reptiles, is of interest because of its presumed relationships to both plesiosaurs and primitive diapsid reptiles. ‘Nothosaurs,’ placodonts, and plesiosaurs together form the Order Sau- ropterygia. The single described New World ‘nothosaur’ species, Corosaurus al- covensis Case, 1936, from the Alcova Limestone of central Wyoming, U.S.A., has long been incompletely known. Numerous new specimens supplement the holotype and virtually complete our knowledge of its skeletal anatomy. Corosaurus has been thought of as a traditional ‘nothosaurid’ and, indeed, has several plesiomorphic sauropterygian features. The relatively expanded appendicular girdles of Coro- saurus are only superficially plesiosaur-like. ‘The axial skeleton is generally con- servative. A discussion of sauropterygian taxonomic characters, a review of ‘notho- saur’ genera, and a cladistic phylogenetic analysis using parsimony are presented by which a basal sauropterygian dichotomy is defined resulting in the monophyletic clades Pachypleurosauria and Nothosauriformes (new taxon). Plesiosauria and Placodontia are monophyletic groups within the Nothosauriformes. Consequently, the traditional ‘Nothosauria’ is paraphyletic. Shared derived characters indicate that Corosaurus is a valid genus within the Nothosauriformes. Claudiosaurus Car- roll, 1981 is the closest known sister group to the Sauropterygia, both apparently derived from plesiomorphic diapsid reptiles. Hypothetical musculature reconstructions and functional morphology suggest that Corosaurus and other large ‘nothosaurs’ favored a primarily limb-dominated method of aquatic locomotion partially analogous to that of plesiosaurs, while the small pachypleurosaurs may have relied more heavily upon tail-dominated pro- pulsion. Structural differences in the appendicular skeletons of pachypleurosaurs, ‘nothosaurids,’ and plesiosaurs probably largely reflect the nearshore, possibly amphibious, behavior of the two former groups. Most ‘nothosaurs’ inhabited paralic marine environments within which a range of habitats is increasingly evident. ‘Their plesiomorphic overall similarity is in part functionally mediated. Close examinations of the geology and structural setting of the Alcova Limestone illuminate the paleoecology of Corosaurus and the biogeography of nothosauriforms minus the plesiosaurs and placodonts. Diverse paleontologic, sedimentologic, and geochemical evidences indicate a restricted, hypersaline marine embayment as in the German Muschelkalk. Stratigraphic analysis places the Alcova Limestone Member, Crow Mountain Formation, Chugwater Group, most probably in the uppermost Lower Triassic (Spathian). 2 PEABODY MUSEUM BULLETIN 44 KEYWORDS Alcova, biogeography, Chugwater, Corosaurus, function, morphology, Notho- sauriformes, ‘nothosaurs,’ Pachypleurosauria, paleoecology, Placodontia, Plesio- sauria, review, Sauropterygia, Spathian, stratigraphy, systematics, ‘Triassic, Wy- oming. ZUSAMMENFASSUNG ANATOMIE UND VERWANDTSCHAFTSBEZIEHUNGEN VON COROSAURUS ALCOVENSIS (DIAPSIDE REPTILIEN, SAUROPTERYGIER) UND DER TRIASSISCHE ALCOVA-KALK WYOMINGS Die traditionelle Unterordnung ‘Nothosauria’ umfasst triassische Meeres-Rep- tilien, die als mégliches Bindeglied zwischen primitiven Diapsiden und Plesio- sauriern ein besonderes Interesse verdienen. Zusammen mit den Placodontiern und Plesiosauriern bilden sie die Ordnung Sauropterygia. Der bisher einzige neuweltliche ‘Nothosaurier,’ Corosaurus alcovensis Case, 1936 aus dem Alcova- Kalk von Wyoming, war lange Zeit nur unvollstandig bekannt. Zahlreiche Neu- funde vervollstandigen dieses Bild. Corosaurus wurde bis jetzt als typischer ‘Nothosauride’ betrachtet. Er zeigt in der Tat mehrere morphe Sauropterygier-Merkmale. Dazu gehoren die relativ breiten Schulter- und Beckengirtel, die denen der Plesiosaurier nur oberflachlich ahneln. Auch das Achsenskelett ist konservativ. Eine Ubersicht tiber die taxonomi- schen Merkmale, eine Zusammenstellung bekannter ‘Nothosaurier’-Gattungen, sowie eine kladistische Analyse nach Parsimonie-Kriterien lasst indessen inner- halb der Sauropterygier eine Dichotomie zwischen den monophyletischen Zwei- gen der Pachypleurosauria und der neu aufgestellten Nothosauriformes erkennen. Innerhalb der Nothosauriformes bilden die Plesiosauria und die Placodontia ihrerseits selbstandige, monophyletische Untergruppen. Dagegen sind die ‘Notho- sauria’ im traditionelle Sinn eine paraphyletische Gruppe. Innerhalb der Notho- sauriformes ist Corosaurus durch abgeleitete Merkmale als selbstandige Gattung ausgewiesen. Die nachstverwandte Schwestergruppe zu den Sauropterygia als Ganzem wird durch Claudiosaurus Carroll, 1981 reprasentiert; beide werden von plesiomorphen Diapsiden abgeleitet. Eine Rekonstruktion des Muskelapparates und funktionsmorphologische Merkmale zeigen, dass Corosaurus und andere grosse ‘Nothosaurier’ sich ahnlich wie die Plesiosaurier, im Wasser hauptsachlich mit Hilfe ihrer Extremitaten fortbewegten. Im Gegensatz dazu spielte beim Schwimmen der kleineren Pachy- pleurosaurier der Schwanz eine wesentliche Rolle. Unterschiede im Extremita- tenskelett der Pachypleurosaurier, ‘Nothosauriden’ und Plesiosaurier deuten auf eine méglicherweise amphibische, kiistenbezogene Lebensweise der beiden ersten Gruppen. Die meisten ‘Nothosaurier’ bewohnten paralische Meeresgebiete mit einem breiten Spektrum spezifischer Habitate. Plesiomorphe Ahnlichkeiten in- nerhalb dieser Gruppe konnen also ebenfalls funktionell bedingt sein. Eine genaue geologische Analyse des Alcova-Kalks erganzt das palokologische Bild von Corosaurus und beleuchtet die biogeographischen Ausbreitungs-Moglich- keiten der Nothosaurier im alten Sinn (d. h. unter Ausschluss der Plesiosaurier und Placodontier). Palaontologische, sedimentologische und geochemische Daten COROSAURUS ALCOVENSIS 3 lassen, wie im germanischen Muschelkalk, ein teilweise abgeschlossenes, uber- salzenes Meeresbecken vermuten. Chronologisch wird der Alcova-Kalk (als Unterglied der Crow-Mountain- Formation und der Chugwater-Gruppe) in den oberen Teil der Untertrias (Spa- thium) gestellt. 4 PEABODY MUSEUM BULLETIN 44 1. INTRODUCTION HISTORY OF INVESTIGATION The ‘nothosaurs’ are a grade-level grouping of sauropterygian marine reptiles well represented by skeletal remains in the Middle Triassic rocks of Europe. They were apparently well adapted to a littoral, possibly amphibious, existence and are of special interest because of their presumed evolutionary relationships to both the primitive diapsid terrestrial reptiles which were their probable an- cestors, and the highly specialized, fully aquatic plesiosaurs of the Jurassic and Cretaceous. The various types of ‘nothosaurs’ (traditionally grouped as a sub- order—seemingly artificially) and the plesiosaurs are obviously closely related and together form part of the Order Sauropterygia. The ‘nothosaurs,’ particularly, are in need of in-depth study and the general anatomic characteristics of many individual taxa are still very confused, as are their systematics, evolutionary relationships, and paleobiology. Major studies of ‘nothosaurs’ have been under- taken in the past by such workers as Arthaber (1924), Edinger (1921), v. Huene (1952), Koken (1893), v. Meyer (1847-55), Nopsca (1928b), Peyer (1931, 1932, 1933, 1934, 1939), Seeley (1882), Young (1958, 1959, 1960, 1965a), Zangerl (1935), and others. Recent efforts of note include those of Carroll (1981), Carroll and Gaskill (1985), Kuhn-Schnyder (1987), Mateer (1976), Rieppel (1987, 1989), Sander (1989), Sanz (1976, 1980, 1983a), Schmidt (1986, 1987), Sues (1987), and Tschanz (1989). While plesiosaurs are primarily known from the Jurassic and Cretaceous, ‘nothosaurs’ are presently restricted to the ‘Triassic. The primary focus of the present study is the largely neglected occurrence of the single described North American ‘nothosaur’ species, Corosaurus alcovensis Case, 1936, from the Triassic Alcova Limestone of central Wyoming. The type specimen was collected in fragments from a highway quarry spoil heap near Goose Egg Ranch, Natrona County, by a University of Wyoming geology student in 1935 (Case 1936). This material was supplemented in 1948 by several partial skeletons and other specimens collected by a Field Museum of Natural History expedition under the leadership of R. Zanger] from the type and adjacent localities in the vicinity of Casper, Wyoming. Of this additional sample, only a portion of one individual has been preliminarily described (Zangerl 1963). Other than in the works of Case (1936) and Zangerl (1963), Corosaurus has been discussed in more than just a cursory way only by E. von Huene (1949) and F. von Huene (1948a) but without the benefit of first-hand examination of the fossils. Additional references to Corosaurus have been essentially limited to mention of the taxon’s existence and to speculation about its possible systematic position. No further discoveries or examinations of Corosaurus were made until 1983 when field work of the present study resulted in the collection of numerous new specimens from the Alcova Limestone in the Casper, Wyoming, area, specifically near Freeland Junction. Studies of the Alcova Limestone itself have previously been limited largely to superficial descriptions of the unit and to attempts at stratigraphic correlation. A famous and easily recognizable stratum, the Alcova has usually been discussed in the context of descriptions and interpretations of its enclosing formations within the Chugwater Group (e.g., Bower 1964; Branson and Branson 1941; Burk 1953; High and Picard 1967a, 1969; Hubbell 1956; Love 1948, 1957; Picard 1967, 1978; Picard et al. 1969; Pipiringos 1953, 1968; Tohill and Picard 1966; etc.). Only Carini (1964) has concentrated specifically on the Alcova in a detailed manner. In many such studies, unsupported interpretations of the geologic age COROSAURUS ALCOVENSIS 5 and paleoecology of Corosaurus alcovensis have been used to make claims con- cerning the geology of the Alcova. PURPOSE AND SCOPE E. von Huene (1949) believed Corosaurus to represent a very primitive stage in the transition of terrestrial reptiles to a secondarily aquatic format, while Zangerl (1963), because of certain apparently derived features of Corosaurus, considered it the most aquatically advanced nothosaur known. F. von Huene (1948a, b, c, 1952, 1956) went still further by placing Corosaurus in the Plesiosauria. ‘These conflicting interpretations were the result of an incomplete knowledge of the anatomy of this animal, especially the anatomy of its limb girdles. A character- ization of the morphology of Corosaurus and the completion of an adequate diagnosis of the taxon are clearly needed. Furthermore, peculiarities of the en- vironment of Corosaurus may indicate that its paleobiology differed radically from that of sauropterygians as a whole. This may bear upon possible habitat and behavioral variations within the Sauropterygia. The apparent geographic isolation of Corosaurus is also a reason for interest. Most traditional ‘nothosaurs’ are known from the Old World, particularly Europe and China, where hundreds of specimens have been assigned to several dozen taxa. The paleobiogeography of early sauropterygians, the paleogeography of the Earth during the Triassic, the distribution of ‘nothosaurs’ in time, and the exact age of Corosaurus are correlative questions. Is a place of origin and route of dispersal of sauropterygians suggested by the spatial and temporal evidence or is the problem merely a function of the distribution of marine Triassic exposures? The systematics of the ‘Nothosauria’ are little understood, due in part to problems of preservation, and there is not yet a consensus as to which skeletal characters are significant in establishing detailed relationships for these animals. Rigorous study of Corosaurus may provide insights not only into the systematics of ‘nothosaurs,’ but also into their relationships with plesiosaurs, and into the origins of sauropterygians in general. What are the structural/functional con- straints that may have led to the evolution of the Sauropterygia and to its differ- entiation into separate clades? Can intermediate stages be envisioned? Bearing such questions in mind, Corosaurus is a fossil which is particularly well suited to analysis for several reasons. Firstly, the preservation of Corosaurus material is generally good. Contained within a carbonate precipitate matrix, many of the specimens are uncrushed and three-dimensional; on occasion it has been possible to totally extract bones from the surrounding rock. This presents an unusual opportunity for description and functional study. Additionally, a good combination of articulated partial skeletons and isolated bones provides an excellent basis for comparison with other taxa. Finally, the relatively large size of the animal fa- cilitates its examination and descriptions, and by being fairly abundant in a localized area, insight into the individual and ontogenetic variation of the species is gained. Corosaurus and ‘nothosaurs’ in general are thus of interest, but so is the geologic aspect of their occurrence. This, naturally, bears directly on the question of sauropterygian paleoecology and biostratigraphy. In specific, the stratigraphic and environmental interpretations of the Alcova Limestone have been a matter of debate for some time. It is therefore necessary to characterize the geology of this unit. This is the secondary thrust of this paper. Although it is a widespread stratum, the Alcova is not easily correlated with nearby Triassic rocks of known 6 PEABODY MUSEUM BULLETIN 44 | N | | WYOMING ee Nite 5 km ? Casper [ Ee =, ed ry ron “, Clore, Bessemer Mtn. ==} = Muddy Mtn. Fic. 1. Locality map of Casper-Goose Egg-Freeland Junction area, Natrona County, Wyoming. Known occurrences of Corosaurus alcovensis located within regions marked by squares (approximate). age. The entire Chugwater sequence is poorly fossiliferous and is in part difficult to date. Most workers have assumed a normal marine setting for the Alcova, but Carini (1964) has proposed a desalted lake-sea as the environment of deposition. These problems deserve additional consideration. PROCEDURE The initial descriptive phase of the project required preparation and study of the existing Corosaurus material. Each of the known specimens, the holotype in the collection of the University of Wyoming and a large amount of primarily un- prepared material in the Field Museum of Natural History collection, was ex- amined. Beyond this, as only a partial composite skeleton could yet be recon- structed, field work was conducted in the summer of 1983 in an attempt to acquire additional and complementary fossil specimens. Exposed examples of Corosaurus were found to be not uncommon in the general Casper, Natrona County, Wyoming area (Figs. 1 and 2). The holotype was originally found near Goose Egg in Jackson’s Canyon, approximately 14 km southwest of Casper, along Wyoming State Highway 220, W'%, NE%, Sec 12, T32N, R81W. Other partial specimens were collected by the Field Museum party in the quarry at this locality, and also from steeply dipping outcrops of the Alcova Limestone approximately 5 km northeast of Free- COROSAURUS ALCOVENSIS 7 ai 2 Se oy i] | WYOMING | —EEE Casper |. Fic. 2. Simplified geologic map of Casper-Goose Egg-Freeland Junction area, Natrona County, Wyoming, after Fig. 1. Triassic and undifferentiated Permo-Triassic sediments stippled. Areas of Quaternary surficial deposits marked by dashed outlines. Faults indicated by heavy lines. €= Cam- brian; J = Jurassic; K = Cretaceous; K/J = undifferentiated Cretaceous and Jurassic; P = Permian; p€ = Precambrian; P/C = undifferentiated Permo-Carboniferous; Q = Quaternary; T = Tertiary. land Junction, Sec 2, T31N, R80W. Most of the Yale Peabody Museum specimens were discovered in talus blocks beneath cliffs of the horizontal Alcova Limestone southwest of Muddy Mountain, along Corral Creek, Milne Ranch, sections 27 and 33, T31N, R79W. It is not possible to prospect directly the resistant, cliff- forming ledge of the Alcova here. Examination of talus blocks yielded occasional Corosaurus bones along exposed bedding plane surfaces. It was originally hoped that the carbonate nature of the Alcova would allow ready acid dissolution of the fossil matrix. However, while the limestone is easily dissolved, the bones themselves have been completely permineralized with calcite and are equally subject to destruction by acid. Due to the relatively dense nature of the bones, no satisfactory method of protective impregnation was found by which the fossils could be easily extracted from the matrix through chemical means. Mechanical preparation with hand and power tools was therefore utilized and, although slow and tedious as noted by both Case (1936) and Zangerl (1963), had the advantage of supplying an intimate knowledge of each fossil. Unfortu- nately, earlier crude mechanical preparation had already damaged some speci- mens. Attempts to determine the nature and extent of imbedded examples through X-radiography failed, as they did for Case (1936), because the approximately equal densities of bone and matrix furnishes little detectable contrast. At times, weathered bones were represented partially or only by matrix impressions. In 8 PEABODY MUSEUM BULLETIN 44 such instances, latex or epoxy casts were fashioned directly from these molds. Casts and plasticine models were useful in functional reconstructions when it was impossible to completely extricate a fossil from its matrix. Field work for this study also allowed first-hand knowledge of the Alcova Limestone and of its stratigraphic relationships. Examination of the Alcova’s geology in the field was supplemented by collection of matrix samples, sedimentary and stromatolitic structures, and fossil invertebrates. Laboratory techniques em- ployed in their study are discussed below (Chapters 5 and 6). COROSAURUS ALCOVENSIS 9 2. OSTEQLOGY MATERIAL The holotype of Corosaurus alcovensis Case, 1936 (originally specimen No. 51000 in the geology collection of the University of Wyoming, now catalogued as UW 5485) remains the best and most complete specimen of this animal known. It consists of a semiarticulated, partial skeleton comprising the greater part of the skull, the vertebral column through the proximal caudals, half of the pectoral girdle (pectrum), most of the forelimbs, and various ribs and gastralia. The fossil was collected from a quarry spoil heap (Case 1936) and is contained in numerous limestone blocks, the majority of which can still be pieced together to show the disposition of the type skeleton. The vertebrae lie in a loop, but the other bones are scattered, often overlapping each other or lying partially imbedded in the matrix. Different sections of the blocks have been prepared from different sides, and the relative position of each bone is therefore not initially obvious. A composite drawing has been prepared to indicate the positions of the more important elements of the skeleton (Fig. 3). A number of bone-containing blocks that were collected and catalogued with Fic. 3. Composite drawing of holotype of Corosaurus alcovensis, UW 5485. 1 = skull; 2 = sacrum; 3 = right clavicle; 4 = left scapula; 5 = left humerus; 6 = left ulna; 7 = left radius; 8 = cervical vertebrae; 9 = right manus; 10 = right humerus; 11 = right ulna; 12 = right radius; 13 = mandible. Coracoids and interclavicle overlie cervicals but are here removed for clarity. 10 PEABODY MUSEUM BULLETIN 44 Fic. 4. Partial skeleton of Corosaurus alcovensis; map plan of FMNH PR480. 1 = sacrum; 2 = anterior caudal vertebrae; 3 = left femur; 4 = tip of ?left fibula; 5 = metatarsal; 6 = left ischium; 7 = left tibia; 8 = left ilium?; 9 = midseries caudals; 10 = left pubis; 11 = right pubis; 12 = right ilium; 13 = right ischium; 14 = associated metatarsals and pes; 15 = dorsal rib. the type do not fit into the skeletal puzzle. Most, if not all of this scrappy material probably represents one or more additional individuals. This was suggested by Zangerl (1963) and indeed, a second sacrum is included in the isolated blocks. A large, isolated block of gastralia may or may not pertain to the true type. Even so, all of the additional material is apparently assignable to Corosaurus. Zangerl (1963) was not, however, correct in assuming that parts of the type have been lost since Case’s study. It has been possible to reassemble the type specimen and to relocate and identify all of the elements referred to in the original description, Fic. 5. Partial skeleton of Corosaurus alcovensis, schematic diagram of portion of FMNH PR1369. 1 = pubes; 2 = dorsal rib; 3 = right femur; 4 = right tibia; 5 = right fibula; 6 = caudal vertebrae. Hatched lines denote impressions. COROSAURUS ALCOVENSIS lal Fic. 6. Partial skeleton of Corosaurus alcovensis, map plan of YPM 41031. 1 = left humerus; 2 = dorsal ribs; 3 = left scapula; 4 = dorsal vertebrae; 5 = left ulna; 6 = left radius. Hatched lines denote impressions. although the interpretation of some of these bones has changed. Only those portions which were never collected, such as the block of six middorsal vertebrae (Case 1936, p. 4), are missing. The bulk of the known Corosaurus material is in the collection of the Field Museum of Natural History, Chicago. Zangerl (1963) preliminarily described the largely disarticulated posterior half of a skeleton making up one of these specimens (FMNH PR480). In the undescribed material are the remains of over a dozen additional individuals. Unfortunately, most are preserved only as isolated or associated vertebrae, ribs, and gastralia, and many such specimens, collected from a single locality near Freeland, Wyoming, have been lumped together under one catalogue number (FMNH PR135). Aside from FMNH PR480, the Chicago collection contains four other Corosaurus fossils, of varying quality, which represent significant portions of single individuals and which are very useful in a study of the whole animal. Some of these specimens have, like the holotype, been collected as groups of bone-bearing limestone blocks and have required reassembly prior to study. Map plans of the two most useful Chicago skeletons are given in Figures 4 and 5. The Corosaurus fossils collected for the present study are now housed in the Yale Peabody Museum of Natural History. Two of these specimens [YPM 41030 and 41031 (Fig. 6)] are partial skeletons; each is contained in a single block. ‘The remainder of the Yale collection consists of isolated bones. From a combination of the existing specimens, most of the bones in the skeleton of Corosaurus alcovensis are now known. Only the phalanges of the pes and the interclavicle are poorly represented. It has also not been possible to directly observe the form of the palate. The known ‘nothosaur’ (i.e., plesiomorphic sauropterygian) palates, however, follow a stereotyped pattern and it is reasonable to assume that the present specimen is structurally similar. The conditions of the bones in each of the three collections ranges from very poor to excellent. Some are crushed and fractured, and others are preserved only as matrix impressions or outlines (see, e.g., Figs. 5 and 6). Certain bones are visible only as cross sections exposed along fracture surfaces through the matrix. On the other hand, many specimens are undistorted and exhibit extremely fine anatomical details. 12 PEABODY MUSEUM BULLETIN 44 SYSTEMATIC PALEONTOLOGY DIAPSIDA Osborn, 1903 NEODIAPSIDA Benton, 1985 LEPIDOSAUROMORPHA Benton, 1985 SAUROPTERYGIA Owen, 1860 NOTHOSAURIFORMES, new taxon COROSAURUS Case, 1936 Type species. Corosaurus alcovensis (the genus is presently monotypic). Holotype. Skull and partial skeleton, UW 5485. Referred material. _Numerous specimens in the Field and Yale Peabody museums of natural history (see Appendix A). Horizon and distribution. Alcova (Limestone) Member, Crow Mountain For- mation, Chugwater Group, Triassic System; various localities in general vicinity of Casper, Natrona County, east-central Wyoming, U.S.A. Etymology. Literally, “northwest-quarter reptile of Alcova.” Diagnosis. A plesiomorphic, intermediately-sized ‘nothosaurid’ (‘nothosauri- form’) (following systematics of Chapter 4), averaging perhaps 2 m in length, possessing a generally conservative axial skeleton and limbs with rather derived limb girdles. Supratemporal fenestrae of skull larger than orbits. Antorbital region slightly longer than postorbital area. Nasals, frontals, and postfrontals large. Posterolateral process of frontal present. No observed quadratojugal. Postorbital bar and temporal arch narrow. Skull table high and broad; pineal foramen located at center of parietals. Moderately sized posttemporal fenestrae relative to other ‘nothosaurs’; opisthotics long and pillar-shaped. Rostrum low and unconstricted. Dermal cranial bones pitted. Upper dentition rather uniform; lower teeth dis- tinctly anisodont with procumbent anterior caniniforms. Mandibular symphysis stout, tip of jaw spatulate. Prominent retroarticular process. Forty-one presacral vertebrae; three sacral vertebrae with distally expanded sacral ribs. Neck of intermediate length relative to other sauropterygians, approximately 50% of tho- rax. Neural arches broad, transverse processes long in extending laterally beyond arches; zygosphene/zygantrum articulations present throughout thoracic series. Neural spines rectangular and of medium, uniform height relative to other saur- opterygians. V-shaped caudal chevrons fully ossified and without distal expan- sion. Gastralia composed of a median element and two pairs of laterals. Medial and posterior processes of clavicle form 90° angle and distinct anterolateral corner; posteromedial shelf present at angle. Interclavicle possibly barlike. No horizontal ventral plate on scapula. Coracoids large and subrectangular; no supracoracoid foramen. Anterior border of pubis convex; obturator foramen distinct. Ischia long and distally expanded. Ilium with well-formed acetabulum and blade; anterior and posterior projections on sacral process. Humerus strongly curved with prom- inent entepicondylar foramen and ectepicondylar notch. Femur sigmoidal, ap- proximately 40% longer than humerus; large internal trochanter. Epipodials dorsoventrally compressed; large spatium interosseum. Ulna and radius short; small ‘olecranon process.’ Tibia and fibula long and narrow. Carpus and tarsus poorly ossified; astragalus twice as large as calcaneum. No evidence of hyper- phalangy. COROSAURUS ALCOVENSIS 13) Fic. 7. Reconstructed skull of Corosaurus alcovensis, dorsal aspect, based upon UW 5485. bo = basioccipital; en = external naris; eo = exoccipital; ept = epipterygoid; f = frontal; fm = foramen magnum; j = jugal; mx = maxilla; n = nasal; o = orbit; op = opisthotic; p = parietal; par f = parietal foramen; pf = postfrontal; pmx = premaxilla; po = postorbital; pfr = prefrontal; pt = pterygoid; q = quadrate; so = supraoccipital; sq = squamosal; stf = supratemporal fenestra. DESCRIPTIVE ANATOMY AXIAL SKELETON Much of the new material assigned to Corosaurus represents parts of the axial skeleton. Aside from the type specimen, several additional strings of vertebrae and associated partial skeletons have now been discovered. The vast majority of new specimens consists, however, of disarticulated and often isolated vertebrae, ribs, and gastralia. Skull Only a single skull of Corosaurus is known, that of the holotype (UW 5485). This was generally well described by Case (1936). Nevertheless, careful restudy in light of our presently greater understanding of sauropterygian anatomy has per- mitted the clarification of certain aspects of the cranial morphology of Corosaurus. A new description and reconstruction are thus necessitated. As noted by Case (1936), the skull, while largely complete, has been subjected to a certain amount of distortion due to its position of preservation across the ventral faces of the fourth, fifth, and sixth caudal vertebrae of the type skeleton (Fig. 3). Sedimentary compaction has caused the offset of the right posterolateral corner of the cranium with the resulting disarticulation of some of the component elements and distortion of the margin of the right orbit. The skull roof and braincase are not crushed, however, and seem to present the true appearance of this region. Dissection of the skull along the fractures reported by Case (1936, p. 5) permitted a three-dimensional examination of the posterior cranial region, which is largely imbedded in supporting matrix. Most of Case’s findings here are confirmed. While high, however, the posterior margin of the skull is not so 14 PEABODY MUSEUM BULLETIN 44 Fic. 8. Reconstructed skull of Corosaurus alcovensis, based upon UW 5485 (mandibular suture patterns unknown). A—left lateral aspect; B—posterior aspect. bo = basioccipital; bs = basisphenoid; cp = coronoid process; de = dentary; en = external naris; eo = exoccipital; ept = epipterygoid; f = frontal; fm = foramen magnum; j = jugal; mx = maxilla; n = nasal; o = orbit; op = opisthotic; p = parietal; pf = postfrontal; pmx = premaxilla; po = postorbital; pfr = prefrontal; pro = prootic; pt = pterygoid; ptf = posttemporal fenestra; q = quadrate; ret p = retroarticular process; so = supraoccipital; sq = squamosal. tall as has been reconstructed by Case (1936, fig. 3). The left squamoso-postorbital bar is not preserved, and both quadrate regions are crushed. As for the preorbital surfaces, compression and concomitant fracturing of the rostrum has obscured the bone relationships and nowhere are the sutures as clear as those of the skull table. The size and shape of the external nares are nonetheless obvious and little broadening of the rostrum has occurred. The left side of the skull is generally well preserved throughout its length and allows an accurate reconstruction of the skull’s gross morphology (Figs. 7 and 8). Both the nostril and the eye faced laterally to a slight degree. The nares are relatively smaller and the orbits larger than in Case’s (1936, figs. 2 and 3) reconstruction. The oblique position of the teeth as noted by Case is undoubtedly true for the anterior rostrum, but the left maxilla has certainly been displaced horizontally and the maxillary dentition should be more correctly regarded as vertical in position. This conclusion is borne out by comparison with the largely undisturbed right maxilla and the configuration of the lower jaw of Corosaurus. The skull of the type specimen is nearly 13 cm long, with a low, broad facial region and a narrow, although short (approximately 2.5 cm) prenarial rostrum. The greatest width of the skull, apparently at the squamoso-postorbital suture, is estimated to have been approximately 7.5 cm. The external nares are retracted posteriorly as is typical for many aquatic reptiles, but remain in a position only midway along the snout. There is no premaxillary/maxillary constriction of the rostrum. The supratemporal fenestrae are large (i.e., larger than the orbits). All COROSAURUS ALCOVENSIS 15 elements of the skull were tightly sutured and the cranium was, as in most, if not all, sauropterygians, virtually akinetic. No sclerotic plates, if originally present, have been preserved. Premaxillae. The description provided by Case (1936) for these bones is accurate except that only five, rather than six right premaxillary teeth are present. No alveolus exists to accommodate a sixth tooth. In addition, I find little justification for Case’s suggestion that the anterior teeth are significantly larger than the others. Any indication of variable length seems to be largely a result of the roots of some teeth breaking through their alveolar walls as the bone was pressed down upon them. The premaxillary—-maxillary suture is digitate and lies near the anterior margin of the external naris, which is longitudinally ovate, whereas the premaxillary—nasal suture meets the nares near their midline. Microscopic ex- amination of this region suggests, however, that the suture forms not a straight line between the nares, but actually a posteriorly directed chevron, as shown in Figure 7. Even so, the premaxillae do not extend beyond the posterior margins of the nares as they do in many sauropterygians. The median suture is straight. Maxillae. ‘The presumed extent of these large, roughly triangular bones can be discerned from a comparison of the two sides of the skull. The lateral margin of the maxilla is long and straight, extending beyond the orbit to the excavated cheek where it meets the posteromedial margin in a relatively sharp spur. The marginal dentition is thus continuous to at least the posterior edge of the orbit. The first maxillary tooth is perhaps slightly more robust than its neighbors, but its ap- parently greater length is again largely the result of a broken alveolar wall. The medial edges of the right maxilla can be clearly seen due to the preservational depression of the nasals and the disarticulation and loss of the right prefrontal. The left maxilla is similarly raised relative to the nasals. The maxillary—nasal and the maxillary—prefrontal sutures are now seen to be the rather straight limbs of an obtuse triangle. The maxilla correspondingly forms the lateral margin of the naris, but only the anterolateral border of the orbit. A distinct, pitted sculpturing can be seen on the surface of the right maxilla. There is also the suggestion of a small, circular depression at the center of each maxilla. It is difficult to determine whether or not these depressions are the product of the crushing of the rostrum. If natural, they may represent pits for housing specialized facial glands, although such glands have not been previously reported in sauropterygians. Nasals. "The median cranial suture continues in a straight line between these two elements. Although crushed and fractured, the configuration of the nasal can now be deduced from the shapes of the surrounding bones. Basically wedge- shaped, the nasals are rather large for a sauropterygian and extend from between the nares to between the orbits where they intertongue with the paired frontals. The posterior terminus of each bone is a sharp point defined by clear sutures. The right nasal is slightly longer than the left, complementing the asymmetrical borders of the frontals. There is no great extension of the premaxillae between the nasals. Prefrontals. While Case (1936) was unable to delimit the nature of the pre- frontals, like the nasals their form can be inferred from the adjoining bones. ‘The left prefrontal, while crushed, is present and forms the anteromedial margin of the orbit. Its serrate suture with the frontal is also evident. The right prefrontal has broken away from the rim of the orbit and was not preserved, separating cleanly along its sutures. The free edges of the frontal, nasal, and maxilla are 16 PEABODY MUSEUM BULLETIN 44 now apparent, revealing also the shape of the missing prefrontal. It had a sharp anterior point and a concave posterior edge. There is no indication of a lachrymal bone. Frontals. These bones, lying directly between and entering into the rims of the orbits, were accurately described by Case (1936, p. 7). All of the frontal sutures are irregularly serrate, including the median one as shown in Figure 7. The left frontal is larger than the right and displays a prominent congenital surface rugosity. Throughout the skull, none of the elements of the paired dorsal series are fused. This is a character, however, which may have varied ontogenetically, and fusion may have been exhibited in older individuals of Corosaurus. Variously fused frontals are present among the many known specimens of Alpine pachy- pleurosaurs (Carroll and Gaskill 1985; Rieppel 1989). While both fused frontals and parietals are characteristic of Nothosaurus (see, e.g., Schroeder 1914, Schultze 1970) and Paranothosaurus (Kuhn-Schnyder 1966), there has been no ontogenetic study of these genera, and juveniles may have possessed unfused skull table elements. Rieppel (1989), however, suggests that mere individual variation may control this trait. Phylogenetic analysis (Chapter 4), on the other hand, indicates that some evolutionary significance is possible for this character. Parietals. These flat components of the skull table are relatively wider than those of most ‘nothosaurs’ possessing so-called large supratemporal fenestrae, and the openings are rather well separated. The conspicuous parietal foramen is centrally placed along the serrate median suture. The jagged anterior end of each parietal is bounded by the frontal and postfrontal, and the posterior end by the supraoccipital and squamosal. The long, narrow posterior parietal process over- laps the squamosal and forms most of the medial wall of the supratemporal fenestra. Case’s (1936) so-called postparietal suture to the rear left of the parietal foramen is nothing more than a hairline fracture. Postparietal bones are unknown in traditional sauropterygians, although they have been mistakenly reported (along with tabulars) in Simosaurus (Kuhn-Schnyder 1961, 1962; see Schultze 1970). Postfrontals. Only the left postfrontal is in place. This stout, ridged, rugose bone forms the posteromedial rim of the orbit and much of the anterior wall of the supratemporal fenestra as described by Case (1936). It is triangular in dorsal aspect and meets the postorbital in a squamous articulation. This relationship can be seen on both sides of the skull, although on the right side both bones have been displaced. The postfrontal meets the parietal in the anteromedial wall of the supratemporal fenestra. Postorbitals. Case (1936) could find no postorbitals but small portions of both are actually preserved, and together with the shape of the squamosal, they can be fairly accurately reconstructed. The posterolateral corner of the left orbit exhibits the impression and fragments of the inner surface of the broken post- orbital. This was a pronged element that clearly formed part of the bony spur at the front of the lower temporal emargination, the anterior portion of the lateral wall of the supratemporal fenestra, and the posterior half of the lateral orbital margin. The vertical “flange” referred to by Case (1936, p. 8) which meets the postfrontal in the wall of the supratemporal fenestra, is also undoubtedly part of the postorbital. The thin anterolateral process of the squamosal presumably lay superficial to the posterior projection of the postorbital, a portion of which is apparently preserved on the right side of the skull. Jugals. No jugal can be observed on the distorted right side of the skull, but its position on the left can be estimated from the divergent bone fibers in the cross- COROSAURUS ALCOVENSIS 7; sectional fracture of the ‘‘spur”’ adjoining the anterior edge of the lateral temporal emargination. A line running through this section may represent the suture between the postorbital and the jugal. From this evidence, it appears that the jugal was a sliver of bone between the postorbital and the maxilla, thinning anteriorly, and not reaching the margin of the orbit. This is the same condition observed in Nothosaurus (Schroeder 1914; Schultze 1970). Squamosals. The position of these bones can be seen in Figure 7 and in Case (1936, plate 1, fig. 1), the right squamosal offset to the right, the medial half of the left still articulated with the parietal. Case’s (1936) account of the form of these bones is correct. The right squamosal is particularly useful in displaying the narrow postorbital process, whereas the left squamosal clearly shows the squamous articulations with both the parietal and the quadrate, and the peg- and-socket joint with the paroccipital process of the opisthotic. The parietal process of the squamosal forms the topographically highest part of the skull. Quadrates. The form of the quadrates is greatly disturbed but it appears that most of the posterior surface of each bone was overlain by the squamosal, leaving only the transverse articular surface exposed. This is rather typical of notho- sauriform (following Chapter 4) suspensoria. Anteriorly, the bone forms an ex- panded plate that lies deep to the squamosal, and to which it is broadly sutured. The pterygoids abut against this sutural line, forming a tight brace with the quadrate and the squamosal. I have been unable to locate quadratojugals in the type specimen, in spite of the suggestion by Case (1936) that they may exist. The squamosoquadrate region of each side of the skull is sufficiently broken to preclude a definite conclusion. Although pachypleurosaurs and possibly Simosaurus apparently retain a vestigial quadratojugal (Carroll and Gaskill 1985; Kuhn-Schnyder 1961; Rieppel 1989; Schultze 1970), this bone is lost in most advanced sauropterygians, perhaps as a consequence of the presumed loss of the diapsid lower temporal arch in the transition to the euryapsid condition (Carroll 1981; Kuhn-Schnyder 1962, 1963a, 1967, 1980) and continued phyletic reduction of the temporal arcade. It is therefore quite likely, and I believe probable, that quadratojugals were lacking in Coro- saurus. This question must be considered unresolved, however. Braincase. This region of the skull has been primarily reconstructed from ex- amination of numerous fractures through the posterior portion of the skull. ‘These fractures extend through the braincase and the bones of the occiput and have necessitated a reliance on the use of bone fragments and impressions. As a result, few of the sutural relationships between bones can be accurately determined. Nevertheless, a generalized picture of the posterior neurocranium can be con- structed (Fig. 8). On the occiput, the basioccipital is prominent and exclusively forms the bulbous occipital condyle and the floor of the large, subcircular foramen magnum. ‘The foramen magnum is situated high on the occipital face. Contrary to Case (1936), the occipital condyle is not constricted at its base. The basioccipital is bounded laterally by the opisthotics and separated from them by the only obvious sutures of the occiput. Just medial to the left of these sutures, and within the basioccipital, a fracture has exposed a small cranial nerve passage originating at the posterior end of the braincase and exiting the occiput as a foramen at the side of the occipital condyle. From such foramina any or all of cranial nerves IX through XII left the skull. The opisthotics form long, cylindrical paroccipital processes quite unlike those of other ‘nothosaurs’ in which these bones are known. Each is directed 18 PEABODY MUSEUM BULLETIN 44 posteroventrally from its position adjacent to the basioccipital towards the pos- teromedial edge of the squamosal. Here the braincase is buttressed against the suspensorium in a single peg-and-socket joint. The proximal extremity of the opisthotics cannot be differentiated in the specimen from the highly fractured exoccipitals that are assumed to flank the foramen magnum. The opisthotic and exoccipital are generally fused in sauropterygians (Romer 1956). The supraoc- cipital roofs the foramen magnum and is apparently a triangular shelf of bone set just below and between the posterior fork of the paired parietals. ‘The post- temporal fenestrae are bounded by the squamosal dorsally and laterally, and the opisthotic/exoccipital ventrally and medially. The left fenestra is preserved and appears, largely from its internal aspect, to be not only rather rhomboidal in cross section, but also unusually large for a ‘nothosaur.’ Like the occiput, the anterior portion of the braincase is very poorly preserved, being heavily fractured. Portions of the left side of the braincase have been lost while the right side is unobservable. However, it is known that the proximal end of the opisthotic approaches a spherical, matrix-filled cavity identified as the otic capsule and the position of the prootic bone. Anterior to this, and lying along the sagittal plane of the skull, a small exposed section of the basisphenoid can be seen. It is situated at a point midway between the pterygoids below and the vertical walls of the parietals above, and is anterolaterally bounded by the epipterygoids. The epipterygoid and the basisphenoid are joined at the basipterygoid process which is just visible. Similar processes appear to buttress the basisphenoid against the prootic and the parietal. No stapes is preserved. Palate. The delicate nature of the skull prohibits the removal of matrix from its undersurface, thus the palatal complex remains largely unknown. Only the pterygoid and the epipterygoid can be partially reconstructed. The posterior edge of the palatal ramus of the left pterygoid is clearly exposed and reveals a typical, smoothly concave anterior margin to the subtemporal fossa. However, while anteriorly the palatal ramus of the pterygoid is a broad, flat, horizontal plate of normal configuration, the posterior edge is ventrally deflected in an apparent pterygoid flange. Additionally, from the position of the epipterygoid caudad, the quadrate ramus of the pterygoid is seemingly not horizontally, but rather vertically expanded, an unusual and possibly primitive condition among nothosauriforms. This is evident from the displaced right temporal region of the skull, where the pterygoid is tightly sutured to both the squamosal and the quadrate, effectively closing the posterior end of the subtemporal fossa. Much of the left pterygoid’s quadrate ramus is broken and missing, but its partial impression indicates a divergence of the rami beneath the basisphenoid and otic capsule, a good deal farther forward than is typical for traditional ‘nothosaurs.’ The presence or absence of a true interpterygoid vacuity cannot, however, be established. As Corosaurus is certainly a primitive nothosauriform in its overall morphology as is later to be discussed in this work and as all known ‘nothosaurids’ have a solid palate, such a vacuity is more than likely absent. ‘The data are, however, inconclusive. Case’s (1936, p. 13) “hook-like projection” on the quadrate ramus of the right pterygoid is difficult to interpret and, if not an artifact, may have functioned in connection with the basisphenoid, as he suggested. The left epipterygoid clearly has a broad footplate that rests on the palatal ramus of the pterygoid. The dorsal process of the epipterygoid is tall, narrow, and rounded; the right one showing these characteristics most effectively. Most nothosauriforms have a narrow dorsal process, although that of Nothosaurus is hourglass-shaped (Romer 1956). COROSAURUS ALCOVENSIS 19 Fic. 9. Mandible of Corosaurus alcovensis. A, partial right dentary of FMNH PR1382, anterior to right; B, proximal end of left mandibular ramus, FMNH PR246; C, dentary symphysis, YPM 41043; Inset, isolated tooth from FMNH PR135. MANDIBLE In addition to the partial lower jaw of the type specimen that was described by Case (1936), four new examples of the mandible of Corosaurus have been recovered (FMNH: PR1368, PR246, PR1382; and YPM 41043). From this material, a more exact knowledge of the form of the mandible may be gleaned. FMNH PR1382 consists of portions of the dorsal edges of both rami, the right exhibiting eighteen teeth in place (Fig. 9A), the left only eleven. Specimen No. PR246 shows the internal aspect of the left ramus from the coronoid process to the retroarticular process (Fig. 9B). The remaining two jaws consist primarily of the symphysial region, but only the Yale specimen (Fig. 9C) is well preserved. The jaws were long, slender, and shallow, with the two rami meeting at an average angle of approximately 40°. The type specimen shows an angle of approximately 35° that matches the angle formed by the rostrum. All known jaw specimens are approx- imately equivalent in size and differences between them probably reflect simple individual variation. The articular region of the mandible is elongate. Specimen No. PR246 displays a long (1.5 cm), straight, retroarticular process, a well-formed, transverse articular cotylus corresponding to the articular process of the quadrate, and a distinct coronoid process. The cotylus and retroarticular process lie along the plane of the straight tooth row. Unfortunately, due to the highly fractured nature of the specimen, no bone sutures are evident. The adductor fossa appears troughlike and relatively deep, but an undetermined amount of preservational distortion may have exaggerated this condition. Anteriorly, the lingual surfaces of the mandibular rami of Corosaurus each bear a single, raised, longitudinal ridge, which is easily seen on the type mandible. The labial surface is smoothly rounded and displays a series of longitudinal striae corresponding to the fibers of the bone. Case’s (1936) report of large, anterior mandibular teeth is obviously correct and is reinforced by examination of YPM 41043. The anteriormost teeth are 20 PEABODY MUSEUM BULLETIN 44 exceptionally large, far larger than the premaxillary teeth, and are directed an- terolaterally. In the region where the mandibular teeth oppose the maxilla, how- ever, the teeth rapidly decline in size and point vertically. In YPM 41043, the observed teeth clearly alternate with adjacent vacant alveoli, whereas in FMNH PR1382 the condition of seemingly less predictable positions for unerupted, young, and mature teeth resulting from the zahnreihe replacement mechanism of reptiles, is evident. With the noted exceptions of size and position, all upper and lower teeth of Corosaurus are alike. They are sharp, conical, and bear fine longitudinal striae, but no carinae. The rami of FMNH PR1382 show particularly well how most teeth are medially recurved, as does a fine example of an isolated tooth from FMNH Lot No. PR135 (Fig. 9). The isolate also displays a wide root that is at least equal in length to the crown. Tooth implantation is thecodont. The symphysial region of the mandible of Corosaurus is more robust than the remainder of the jaw. It is slightly spatulate and was strengthened by an internal thickening of the bone. The symphysis itself, however, while strong is not excep- tionally long. A similar symphysial expansion or “scoop” is known in Nothosaurus (see, e.g., Geissler 1895; v. Meyer 1847-55; and Schuster and Bloch 1925). A small lower jaw with an even more exaggerated scoop was described by von Huene (1958) as belonging to Anarosaurus, although this assignment is question- able. VERTEBRAL COLUMN Essentially the entire spinal column of Corosaurus is now represented in the collected fossils as several articulated partial series and numerous isolated ver- tebrae. Only the very distalmost caudals are unknown. Although the preparation resistant nature of the microsparite matrix has allowed few of the vertebrae to be examined in their entirety, examples of each vertebral type are exposed from several different perspectives (e.g., Figs. 3, 4, 5, 6, 10). The form of the complete column is therefore clearly shown. The total vertebral count of Corosaurus ap- proaches 85 or more. The presacral number is 41. The vertebral centra are generally elongate and nearly cylindrical, ranging from deeply amphicoelous to nearly platycoelous, while the neural spines of Corosaurus are of medium height and relatively uniform design throughout the column. Cervical Vertebrae and Ribs Although badly broken, the vertebrae from the neck of the Corosaurus type spec- imen are all at least partially present. They form a twisted, articulated series, the disposition of which was described by Case (1936). Fragments of an additional series and the cross section of an isolated vertebra (both specimens from FMNH Lot No. PR135) augment our knowledge of the neck. The cervical series is here considered to consist of eighteen vertebrae, making the length of the neck in the type specimen approximately 25 cm. The centra are small (averaging 1 cm in length for the type), but gradually increase in size caudad, as do the narrow, subrectangular, neural spines. The length of each centrum is approximately equal to its height, and no dorsal transverse thickening of the neural spines exists. The smooth neural canal is tubular and unconstricted. The cervical ribs are dichocephalous, articulating exclusively with, and in each specimen examined fused to, the centrum (e.g., Fig. 10A and B). The articular COROSAURUS ALCOVENSIS ZA jst, OB . 2 cm E 1 cm Fic. 10. Axial skeletal components of Corosaurus alcovensis. A, oblique transverse section of midseries cervical vertebra from FMNH PR135 showing bicipital ribs; B, ventral view of midseries cervical vertebra with ribs, UW 5485, anterior to top; C, posterior cervical rib from FMNH PR135, anterior to left; D, midseries caudal vertebra, ventral aspect, YPM 41047; E, caudal chevron, YPM 41045; F, typical median gastralium from FMNH PR135;G, typical lateral gastralium from FMNH PR135; H, asymmetrically pronged median gastralium from FMNH PR135; I, doubly pronged median gastralium from FMNH PR135. facets for these ribs are set upon two short parapophyses lying low on the centrum. The facets are longitudinally oriented and set one above the other. The ribs are distinctly pronged, with both an anterior and a posterior projection lying parallel to the body axis (Fig. 10C). The anterior prong is the largest in the anteriormost ribs; the posterior prong dominates caudally. The atlas/axis complex is poorly known in ‘nothosaurs’ but is partly preserved in the type of Corosaurus. The spine of the axis differs from those of the other cervicals in being broad and roughly triangular. Its anterior edge overlaps the posterior zygapophysis of the atlas. Pronged, bicipital ribs are present on the axis. The spine of the atlas is very low. Only the neuropophysis of the atlas seems to be preserved, although the nondescript “‘preatlas” elements of Case (1936, p. 16) may be fragments of the atlas. In any case, the so-called “preatlas” is difficult to evaluate. Dorsal Vertebrae and Ribs The dorsal series is well known through the collection of several strings of ver- tebrae, groups of associated vertebrae, and isolated dorsals which complement the type specimen. I am inclined to accept Case’s (1936, p. 15) estimate of six missing dorsal vertebrae from the type for a total of 41 presacrals. If then, 18 vertebrae can be counted as cervicals, and ignoring the sometimes nebulous category of transitional “pectorals” often used in describing the Plesiosauria, we are left with 23 dorsal vertebrae for Corosaurus. [It should be noted that the distinction between cervical and trunk vertebrae is relatively clear in pachypleurosaurs (Carroll, personal communication, 1988).] The dorsal series of the type of Corosaurus then, including an estimate for those missing vertebrae, measures approximately 50 cm in length. The 17 preserved dorsals of FMNH PR1383 total 36 cm. The thoracic region of the skeleton was thus about twice the length of the neck. The dorsal vertebrae are the largest of the column, and average nearly 2 cm in length in the holotype. Despite their increasingly greater size, the dorsal 22 PEABODY MUSEUM BULLETIN 44 vertebrae are little different from the cervicals except in possessing relatively long (approximately 1 cm), stout, transverse processes. These processes are fully de- veloped on the neural arch by the twentieth vertebra of the column. Vertebra number 19 exhibits a transitional or “‘pectoral” position of the process. The single- headed processes are as long in the anterior dorsals as the spines are tall, and are approximately equal in length to the height of the vertebral centra. They are directed slightly upwards, are thickest distally, and have ovate cross sections. While the neural arch and zygapophyses are broad as in all ‘nothosaurs,’ the transverse processes extend well beyond their lateral margins. This is a seemingly advanced condition in the Sauropterygia. The transverse processes become some- what more robust, but shorter, caudad. The neural spines average 1.5 cm in height from the level of the transverse process, 1.5 cm long, and are subrectangular in outline. They are thickened dorsoposteriorly. Accessory articulations are present on the neural spines of the dorsals as zygo- sphene and zygantrum. These are particularly well shown on vertebrae 29 and 30 of the type specimen where the basal anterior edge of each spine has a projection (zygosphene) which fits into a wedge-shaped cavity (zygantrum) at the base of the preceding spine. As a consequence, the leading and trailing edges of adjacent spines are in close contact. This condition persists throughout the dorsal series. Accessory articulations have been reported in several ‘nothosaurs’ [e.g., Dactylo- saurus (Sues and Carroll 1985), Nothosaurus (Schmidt 1986), Neusticosaurus (Pachypleurosaurus) (Carroll and Gaskill 1985; Zangerl 1935), Serpranosaurus (Rieppel 1989), Simosaurus (v. Huene 1952)] but are unknown in all plesiosaurs save the primitive genus Pistosaurus (Sanz 1983b; Sues 1987). Most, if not all, nonplesiosaur sauropterygians probably possessed such articulations (placodonts exhibit hyposphene/hypantrum articulations (Rieppel 1989)). The large, dorsal zygapophyses of Corosaurus are set close together with flat, essentially horizontal articular faces. The neural canal remains circular in section but is constricted near the origin of the transverse processes. As in all ‘nothosaurs,’ (i.e., plesiomorphic sauropterygians) no nutritive foramina exist in the floor of the canal or on the undersurface of the centrum as they do in plesiosaurs. The dorsal ribs are of normal appearance; curved, long, and slender with a single, expanded head (see Figs. 4 and 5). None are fused to the transverse processes. The longest complete thoracic rib of FMNH PR480, an animal of approximately equal size to the holotype, is 11.5 cm long. Others were no doubt longer. ‘The posteriormost ribs extend almost horizontally, but most were directed laterally and ventrally. As opposed to such forms as Ceresiosaurus, Neusticosaurus and Lariosaurus (Carroll and Gaskill 1985; Mazin 1985; Peyer 1931; Sanz 1976, 1983a; Seeley 1882; Zangerl 1935), there is no outwardly observable sclerotic thickening (‘“‘pachyostosis”) of the dorsal ribs. Sacral Vertebrae and Ribs The sacrum of Corosaurus consists of only three vertebrae. This is the apparently primitive condition for sauropterygians. Three examples of the sacrum, each complete, are known; that of the type, one from a skeleton (FMNH PR480) preliminarily described by Zangerl (1963), and another specimen numbered as part of the type but obviously belonging to a second individual. Each sacrum is approximately 6 cm long and at a maximum, 9.5 cm across. The vertebrae are very similar to the preceding dorsals. They are not coossified in the type but are tightly articulated; the neural spines closely contact each other. The zygapophyses are smaller than those craniad and have medially inclined articular surfaces. ‘The long (3.5 cm) sacral ribs are tightly sutured to short, stout, transverse processes COROSAURUS ALCOVENSIS 23 arising from both the neural arch and the centrum. These ribs are directed ventrolaterally, with great expansion of their distal ends. The iliac articular surfaces are roughly triangular in section and are deeply excavated. The second specimen of UW 5485 is an isolated, yet articulated, sacrum. ‘The vertebrae are tightly joined to each other and to their ribs. In this case, there is a possibility of some fusion of the elements. Nevertheless, the sutures remain obvious. Caudal Vertebrae and Ribs The primary source of information on these vertebrae is Zangerl’s (1963) spec- imen, FMNH PR480 (Fig. 4), although several other caudal specimens are known (see, e.g., Fig. 5). The anterior caudals are present in both the holotype and FMNH PR480; the latter also retains most of the rest of the tail. At least 33 caudals are preserved in FMNH PR480, possibly as many as 36. The actual number is obscured by the overlap of the distorted column and by covering matrix. The distalmost caudals have not been found but it is estimated that a total of about 40 vertebrae formed the long, tapering, unspecialized tail. ‘The tail was perhaps 1.25 times as long as the thorax, possibly 70 cm long in the type. The anterior caudal centra are short and robust, much like those of the sacrum, but posteriorly they lengthen relative to their diameters. As throughout the column, the centra are cylindrical although constricted at their midsection. The under- surfaces of the anterior caudal centra are smoothly concave. The median and posterior caudal vertebrae each bear twin, longitudinal, ventral ridges which stretch from the chevron facets to the anterior edge of the centrum (Fig. 10D). Stout, horizontally oriented ribs are borne by the anterior caudals upon short parapophyses on the centra, to which they are tightly sutured, possibly fused. These ribs are of similar character to the sacral ribs, but are flatter, generally longer, and without the distal expansion. The first caudal rib is directed towards the sacrum, although not involved in the sacroiliac articulation, whereas the remainder point posterolaterally. The third and fourth caudal ribs are the longest; successive ribs gradually decrease in size through about the fourteenth caudal vertebra (see Fig. 3). Several vertebrae posterior to the fourteenth caudal maintain vestigial parapophyses but these probably held no ribs. The subrectangular neural spines of the anterior vertebrae rapidly shorten and give way to low, rounded, swept-back spines that extend well past the posterior margins of the centra (e.g., Figs. 4 and 5). Eventually these are lost, as are the gradually narrowing zygapophyses. Contrary to Case (1936, p. 20), well-developed chevron facets are visible on the posteroventral ends of the caudals of the type specimen, beginning with the fifth caudal vertebra. The fifth and sixth caudals of FMNH PR480 are damaged but the chevrons appear to have begun on vertebra number seven. If so, this individual difference might be ascribed to a sexual variance in the region of the cloaca. The chevrons themselves are poorly known in most ‘nothosaurs,’ but several good examples are now known for Corosaurus. These are slender, delicate chevrons, the two arms of which are joined in a solid V at their distal ends (Fig. 10E). Proximally, the two arms are free and bear prominent, posteromedially inclined, articular heads for their attachment to the centra. The chevrons were not fused to the vertebrae. Gastralia Very many isolated gastral ribs occur with the Corosaurus specimens, along with rock slabs displaying groups of associated gastralia. Zangerl (1963) has suggested that the block of gastralia associated with the type specimen of Corosaurus (Case 24 PEABODY MUSEUM BULLETIN 44 1936, fig. 14) belonged to a separate individual. This is quite possible in light of the aforementioned second sacrum catalogued with the holotype. Sauropterygian ventral baskets are often found as isolated, coherent units, presumably owing to the interlocking nature of their gastralia and their associated sheets of muscle. YPM 41030 also consists primarily of a cluster of gastralia although these are rather randomly oriented. It is therefore impossible to assign Case’s (1936) gas- tralia specimen to his type skeleton with any degree of certainty. In spite of this problem, the block of gastralia indicate well the pattern of arrangement of the ventral armour of Corosaurus. Each gastral segment is composed of a primitive, V-shaped median element which is closely flanked on each side by two imbricating lateral rods (Fig. 10F and G). The lateral elements are straight and doubly pointed; each lies craniad to its medial neighbor. From the size and concentration of the gastralia it is assumed that two rows of these ribs were associated with each vertebral segment between the pectrum and the pelvis. A solitary median gastral element found in the blocks unarguably containing the type skeleton is approximately 13 cm long. This suggests a rather broad body region for the animal. Other isolated gastralia in the Field Museum and Yale collections show that occasionally, the median elements can be pronged on one or both ends (Fig. 10H and I). This is a congenital deformity of no phylogenetic consequence and has been previously reported in Nothosaurus (Koken 1893). One partial median rib amid the Yale material is very large and stout (approximately 1.5 cm thick at its center), giving the first indication that Corosaurus grew much larger than is suggested by the type. All the gastralia are formed of rather dense, heavy bone. APPENDICULAR SKELETON The appendicular skeleton of sauropterygians is highly modified in response to their use in an aqueous medium. The specialized limbs are broad and flattened and often exhibit hyperphalangy. These limbs usually conform to several similar patterns of little taxonomic value. On the other hand, aside from the skull, the limb girdles are perhaps the most taxonomically useful skeletal elements in the Sauropterygia, as long as ontogenetic variations are taken into account. They form massive, platelike assemblies, often possessing significant intergeneric dif- ferences. The appendages and girdles of Corosaurus have, to date, been poorly understood and inadequately discussed. Now, however, new material in conjunction with the old presents us with the opportunity for a nearly complete description of its appendicular skeleton. Included in the specimens of Corosaurus are the probable remains of an interclavicle, three clavicles, two scapulae, four coracoids, three pubes, two ischia, three ilia, at least seven humeri, three radii, five ulnae, six femora, four or five tibiae, two fibulae, and substantial portions of both a fore and a hind “foot.” Pectrum The pectoral girdle of Corosaurus is unique among the ‘nothosaurs’ but has, due to inadequate material, been incorrectly reconstructed in previous studies (Case 1936; E. von Huene 1949; F. von Huene 1948a; and Zangerl 1963). Only the disarticulated pectrum of the type specimen was previously available for study and while its components were correctly identified, they were often misinterpreted. COROSAURUS ALCOVENSIS 745) Fic. 11. Pectoral elements of Corosaurus alcovensis. A, right clavicle, ventral aspect, UW 5485; B, left clavicle, internal aspect, YPM 41037; C, right clavicle, ventral aspect, from FMNH PR135, interclavicular facet at right; D, posterior view of C, lateral edge to right. Inset, probable interclavicle of UW 5485, oblique section, anterodorsal? surface to top. Elements of the shoulder girdle to be found in the holotype are the major portion and impressions of the right coracoid as exposed from the dorsal (or internal) and medial sides, a partial impression of the left coracoid (dorsal surface), the exposed lateral surface of the left scapula, the ventral surface of the right clavicle, and what appear to be two fragments of the interclavicle. New specimens are a virtually complete, matrix-free, right coracoid (YPM 41034), a large cross-sec- tional fragment from an indeterminate coracoid (YPM 41064), a left scapula exposed from its medial side (YPM 41031), the completely exposed dorsal (in- ternal) surface of a left clavicle (YPM 41037), and a nearly complete, matrix- free, right clavicle (from FMNH Lot No. PR135). These additional fossils leave little doubt about the structure of the pectrum. Clavicle. ‘The dermal girdle comprises the clavicles and the interclavicle. The clavicle is an L-shaped bone with a stout, barlike, pointed, medial process and a thin, spatulate, posterolateral process which meet at an angle of nearly 90° to form a sharp anterolateral corner (Fig. 11). In this regard the clavicle is similar to those of most other ‘nothosaurs,’ i.e., plesiomorphic sauropterygians. The con- cave medial edge of the posterolateral process is smoothly rounded and decidedly thickened (tapering caudad). The lateral and anterior edges, toward which the structural fibers of the bone are directed, are rough and unfinished. The lateral edge is thin, the anterior thickened, and the entire posterolateral process is dorsally deflected. The medial bar of the clavicle is thick and dense. It bears an anterodorsal ridge or tuberosity and a posteroventral, interclavicular facet or attachment scar. The interclavicular facet forms a rugose trough oriented along the axis of the medial process. At the juncture of the two clavicular processes a thin, rounded, tablike shelf projects posteromedially. This shelf is broken in the type specimen and in FMNH PR135, but is complete in YPM 41037. The transverse dimension of the type clavicle is 8 cm. The clavicle of Corosaurus is unusual among those of many ‘nothosaurs’ in not being tightly sutured to either the scapula or the opposite clavicle. Rather than 26 PEABODY MUSEUM BULLETIN 44 Fic. 12. Pectoral elements of Corosaurus alcovensis. A, left scapula of UW 5485, lateral aspect; B, left scapula of YPM 41031, internal aspect; C, right coracoid, internal aspect, YPM 41034; C1, silhouette of symphysis, dorsal side to right, C2, silhouette of glenoid, dorsal side to left; D, right coracoid, internal aspect, UW 5485. Impressions denoted by hatched lines. being joined at a broad contact of their medial processes, the clavicles merely met at their tips and were strongly braced by the interclavicle. Interclavicle. "The presumed interclavicle is exposed next to the right coracoid along a fracture in the matrix of the type specimen. This bone is partially obscured by matrix, and while it cannot be considered a fragment of any other element, it is the only pectoral component which is incompletely known. It appears that the interclavicle is a small, triangular bone with a sharp posterior projection (Fig. 11) as it is in certain forms such as Keichousaurus, Neusticosaurus, and Simosaurus (see Chapter 4). What is probably the anterodorsal surface is smoothly concave. However, the interclavicular attachment scars of the clavicles, as noted by Zangerl (1963, p. 118), seemingly indicate the presence of lateral, barlike projections from the interclavicle, but nothing of the sort is visible in the present specimen. Such projections are possibly broken off or hidden by matrix. Scapula. The endochondral portions of the shoulder girdle, the scapula and coracoid, are dense, robust bones. In a general manner, these elements follow the typical ‘nothosaurian’ pattern, yet also present characters peculiar to the genus. The scapula (Fig. 12A and B) is an independent bone which is sutured to neither the clavicle nor the coracoid. It features a prominent, though relatively narrow, dorsal blade which projected posterodorsally to a point above the glenoid. A somewhat similar scapula is illustrated by Young (1965a, fig. 5) for Chinchenza. The anterior edge of the scapula of Corosaurus is smooth and slopes craniad in a sinuous curve. The distal extremity of the blade is anteroposteriorly widened, COROSAURUS ALCOVENSIS On, although this is exaggerated in the somewhat crushed blade of the type specimen. The glenoid area of the scapula is thickened and rugose, with a slight lateral bulge. Its posterior slope parallels that of the dorsal blade. A distinct notch separates the glenoid from the blade. There is no ventral, horizontal expansion of the scapula as is sometimes observed in ‘nothosaurs’ and is ubiquitously present in the plesiosaurs. The anterior corner or expansion of the bone (“acromion” of Romer 1956) is smoothly rounded in profile and slightly concave on its lateral surface. The anterior edge of this expansion is greatly thickened to form a barlike border opposite the glenoid and acts as the attachment surface for the clavicle. The medial surface is smooth and flat and is demarcated from the thick anterior bar by a sharp ridge or escarpment. The ridge is most pronounced at its center, is reduced at its extremities, and joins the anterior edge of the scapular blade near its base. The medial surface of the “acromion” merges with the inner face of the scapular blade, which is slightly offset in relation to the body of the scapula. The type scapula is approximately 6 cm long from anterior tip to the top of the blade. Coracoid. ‘The coracoid is a large, flat, roughly rectangular bone (Fig. 12C and D). It is unique among known ‘nothosaurs’ in that its anteroposterior dimensions are rather uniform, whereas typical ‘nothosaurs’ exhibit a very pronounced central narrowing of the coracoid between expanded lateral and medial ends. The anterior and posterior edges of the coracoid of Corosaurus are only shallowly concave. These edges are smooth and are the thinnest parts of the bone. They can also display a certain amount of individual variation as witnessed by the wavy posterior edge of the type specimen versus the straighter border of the only slightly larger YPM 41034. This variation is not unexpected in light of the latently cartilaginous nature of sauropterygian limb girdles. The limb girdles of all sauropterygians display a striking amount of ontogenetic variation because of the large amounts of cartilage persistently present in juvenile and subadult specimens. Secondarily adapted aquatic tetrapods often have little need to replace cartilage with bone, at a high metabolic cost, when the extra weight of cartilage can be easily neutralized and supported through natural hydrostatic buoyancy. The resulting ontogenetic variation is especially noticeable in the more aquatically specialized plesiosaurs, whose skeletons are often never fully ossified, but should be expected in ‘notho- saurs’ as well. The ventral surface of the coracoid is essentially flat with only a slight concavity of the medial half, whereas the dorsal or internal surface is marked by a thick, rounded, transverse strut. This strut, formed by a thickening of the coracoid midline, particularly in the glenoid and symphysial regions, is characteristic of most sauropterygians and presumably braced the glenoid against internally di- rected forces generated during forelimb movement. The glenoid edge of the bone is rather straight and deeply pitted where it was capped by cartilage. The large, crescent-shaped symphysial surface was also finished in cartilage and forms the thickest part of the coracoid. Its upper surface is convex, its lower concave. The articular surface of the glenoid is not parallel with the symphysis but is directed slightly craniad as in other ‘nothosaurs.’ There is no supracoracoid foramen or notch in the anterolateral corner of the coracoid. Such a notch, sometimes closed by the adjacent scapula, is known in many other primitive sauropterygians (‘nothosaurs’) in which the pectrum has been described. A possible small notch is indicated in the partial impression of the left coracoid in the type specimen, but may be only an individual imperfection as the other preserved coracoids obviously lack a notch. The right coracoid of the 28 PEABODY MUSEUM BULLETIN 44 Fic. 13. Reconstructed pectrum of Corosaurus alcovensis (interclavicle hypothetical). A, dorsal (in- ternal) aspect, anterior to top; B, ventral aspect. cl = clavicle; cor = coracoid; gl = glenoid; icl = interclavicle; pec f = pectoral fenestra; sc = scapula; sc b = scapular blade. type is estimated to have been approximately 7.5 cm in breadth along the transverse strut. A cross section through YPM 41064 indicates that the coracoid was a heavy, ‘“pachyostotic” bone with dense, thickened compacta layers. Restoration. ‘The complete pectrum of Corosaurus is reconstructed in Figures 13 and 14. The gross morphology of this girdle is unquestionably ‘nothosaurian’— that is, plesiomorphic for sauropterygians, albeit unusual. As in other saurop- terygians, and in contrast to the usual reptilian condition, both the scapula and the interclavicle are positioned superficially to the clavicles. The posterolateral portion of each clavicle meets and overlies the corresponding scapula’s antero- medial ridge as in Nothosaurus. The clavicle and scapula were not tightly sutured as was typical for many sauropterygians. Rather, the scapuloclavicular assembly is assumed to have been held together by attendant musculature, cartilage, and ligaments. The clavicles and interclavicle formed a stout transverse bar across the front of the trunk, bracing the anterior part of the pectrum in a manner similar to the coracoid strut. Behind this bar lay a wide opening that was bounded posteriorly by the large coracoids. This pectoral fenestra was relatively shorter (anteroposteriorly) than in other ‘nothosaurs’ due to the great size and unusual shape of the coracoids of Corosaurus. These coracoids, the dominant structures of COROSAURUS ALCOVENSIS 29 Fic. 14. Reconstructed pectrum of Corosaurus alcovensis (interclavicle hypothetical). A, left lateral aspect; B, anterior aspect. cl = clavicle; cor = coracoid; gl = glenoid; icl = interclavicle; pec f = pectoral fenestra; pec r = pectoral rib; sc = scapula; sc b = scapular blade; thor c = thoracic cavity. the pectrum, met in a strong symphysis to counteract the thrust from the forelimbs. As in all sauropterygians, there was little dorsal development of the pectrum. ‘The scapular blades probably held the large ventral basket only loosely against the ribs of the thorax. The glenoids were, of course, largely formed in cartilage and were positioned between the posterior edges of the scapulae and the lateral faces of the coracoids. This same cartilage held the scapulae to the anterolateral corners of the coracoids. Pelvis As noted by Zangerl (1963), the posterior portion of a skeleton of Corosaurus represented by FMNH PR480 contains the articulated right half of a pelvis (Fig. 15A). The left half is also present, although disarticulated, somewhat distorted, and largely buried beneath matrix and other bones. What Zangerl (1963, p. 120) has interpreted as a fibula is probably the crushed left ischium; his possible left ilium (1963, plate 5) is a caudal vertebra. Two additional specimens are what is probably the left pubis as exposed from the ventral side (YPM 41040) and a well-preserved right ilium (FMNH PR243) with exposed lateral and ventral surfaces (Fig. 15B and C). Pubis. ‘The ventral elements of the pelvis are, like the pectrum’s coracoid, large and platelike. Zangerl (1963, p. 118) has noted that the convex anterior border 30 PEABODY MUSEUM BULLETIN 44 Fic. 15. Pelvic elements of Corosaurus alcovensis. A, right side of pelvis of FMNH PR480, internal aspect: 1, pubis, 2, ilium, 3, ischium; B, ?left pubis, YPM 41040, ?ventral aspect, anterior to left; C, right ilium, FMNH PR243, lateral aspect: C1, silhouette of articular facets. of the pubis of Corosaurus is in sharp contrast to the concave front ends of the pubes of all other described ‘nothosaur’ genera. This border is partially obscured in the pubes of FMNH PR480 but is completely visible on YPM 41040 (Fig. 15B). The posterior border of the pubis is concave in normal fashion. The ventral side of the pubis is flat; dorsally it is contoured to accommodate a transverse strut or thickening as was earlier seen in the coracoid. The iliac and ischial facets are located on stout posterolateral prongs, between which is a large obturator notch. The iliac prong and its semicircular facet are directed dorsad. The thin anterior and thickened lateral and medial edges of the pubis were finished in cartilage; thus a certain amount of ontogenetic or individual variation or both can be expected to have existed in its overall shape. The transverse dimension of the right pubis of FMNH PR480 is approximately 6 cm. Ischium. ‘The ischium of Corosaurus is typical of sauropterygians in having a long shaft, an expanded foot, and a greatly thickened symphysial edge (see Fig. 15A). There are few significant differences from the ischia of other ‘nothosaurs.’ Its posteromedial edge is convex and unfinished; the anteromedial margin is broadly concave. The head of the ischium is stout and bears an ovate iliac facet on its dorsolateral surface. As in the pubis, a thickened transverse strut runs from COROSAURUS ALCOVENSIS 31 obt f Fic. 16. Reconstructed pelvis of Corosaurus alcovensis. A, dorsal (internal) aspect, anterior to top; B, ventral aspect. ile b = iliac blade; ilm = ilium; isch = ischium; obt f = obturator foramen; pif = puboischiadic fenestra; pub = pubis. the glenoid region to the symphysis. The greatest length of the right ischium of FMNH PR480 is approximately 8 cm. Ilium. As in the pectrum, the dorsal component of the pelvis of Corosaurus is reduced as in all sauropterygians. The ilium is the smallest element of the pelvis and is of typical ‘nothosaurian’ appearance, although more robust than most (Fig. 15C). It is a low, stout, laterally curving bone retaining a prominent dorsal blade. The blade is flat across its top, with a very small anterior point and a somewhat larger posterior projection or ramus. These projections, particularly the anterior one, are not present in all ‘nothosaurs.’ The distinct pubic and ischial facets of the ventral surface of the ilium correspond in size and shape to the iliac facets of the ventral bones; the ilium sits nearly vertically upon the ventral elements. ‘The anterior surface of the ilium slopes gently forward; the posterior is smoothly concave. The large acetabulum is subcircular and shallowly concave. A low external ridge runs from the top of the acetabulum to the posterior point of the iliac blade. The right ilium of FMNH PR480 is 4 cm long and 3 cm high. Restoration. The pelvis of Corosaurus is restored in Figures 16 and 17. The broad ventral plates meet in a strong, cartilage supported, symphysis. In anterior 52 PEABODY MUSEUM BULLETIN 44 A ilc b Fic. 17. Reconstructed pelvis of Corosaurus alcovensis. A, left lateral aspect; B, anterior aspect. acet = acetabulum; ilc b = iliac blade; ilm = ilium; isch = ischium; obt f = obturator foramen; pif = puboischiadic fenestra; pelv c = pelvic canal; pub = pubis; sac = sacrum. or posterior aspect, the symphysis presented something of a V-shape, rather than the largely horizontal union seen in plesiosaurs. Between the ventral plates of Corosaurus was a rather typical, large, puboischiadic (thyroid) fenestra. The ilia sat upon the laterodorsal corners of the ventral elements and were joined to them by the cartilage of the acetabula. Only small portions of the pubis and the ischium contributed to the rather well-formed acetabulum, however. At the junction of the three pubic bones, a large obturator foramen was formed by closure of the obturator notch of the pubis. The ilia were apparently tightly joined to the ribs of the three sacral vertebrae. Forelimb The type specimen of Corosaurus (UW 5485) preserves both humeri, the left one being free of matrix, both radii, both ulnae, and portions of the right carpus and manus. Supplementing the information available on the forelimb are three left humeri and the distal end of a right humerus in the Yale collection (YPM 41031, 41032, 41033, and 41035, respectively), a crushed right humerus (from FMNH Lot No. PR135), the impressions of both a left radius and ulna (YPM 41031), an indeterminate ulna impression (FMNH PR135), and the proximal end of a right ulna (YPM 41036) which is free of matrix. COROSAURUS ALCOVENSIS 63) 5 cm ent sup \ Fic. 18. Left humerus of Corosaurus alcovensis, based primarily upon UW 5485 and YPM 41033. A, proximal aspect, anterior to top; B, anterior aspect, proximal end up; C, extensor aspect; D, flexor aspect; E, posterior aspect; F, distal aspect, anterior to bottom. ect = ectepicondylar notch; ent = entepicondylar foramen; sup = supinator process; sup r = supinator ridge. Humerus. The humerus is strongly curved caudad and is generally similar to those of other nonplesiosaurian nothosauriforms (Fig. 18). The shaft is relatively short and stout. The proximal head of the humerus is expanded dorsoventrally and is somewhat rectangular in cross section; the distal end is lateromedially expanded with an ovate cross section. Both ends of the bone are unfinished, being originally capped by cartilage. There is a prominent ectepicondylar notch for the passage of the radial nerve and blood vessels and a large entepicondylar foramen allowing supply of the flexor surface of the antebrachium. Just distal to the foramen, the entepicondylar corner of the humerus bears a small process which enlarged the surface area available for the origins of the flexor musculature. ‘There is no demarcation between the ulnar and radial facets. A small, though distinct, supinator process for the origin of the M. supinator longus is situated immediately proximal to the ectepicondylar groove. A sharp, sinuous supinator ridge runs along the anterior edge of the shaft from this process to the anteroventral corner of the proximal articular head. There is essentially no deltopectoral crest, merely a sharp anteroventral edge to the proximal part of the shaft for apparent insertion of the M. pectoralis. A roughened convexity on the dorsoproximal end of the bone probably served as the attachment site for the M. deltoideus. The dorsal surface 34 PEABODY MUSEUM BULLETIN 44 \ \ 5 cm 1 cm Fic. 19. Forelimb elements of Corosaurus alcovensis. A, left humerus, YPM 41032, shaft preserved by matrix impression, Al, proximal aspect; B, left ulna of UW 5485, extensor aspect; C, proximal end of right ulna, YPM 41036, extensor aspect, C1, silhouette of articular head. of the bone is convex; the ventral surface is largely flat. The proximal head of the humerus bears both shallowly concave scars and low, ridged processes for the insertions of a number of additional shoulder muscles, notably the M. scapulo- humeralis cranialis, the M. subcoracoscapularis, the M. coracobrachialis, and the M. latissimus dorsi (Fig. 18). The attachment particulars of the limb musculature are discussed below in the section on functional morphology. The left humerus of the holotype of Corosaurus is approximately 9 cm long, and while from an individual that was certainly a young adult, it and its right counterpart are the smallest humeri represented in the collection. ‘The largest humerus (YPM 41032) is approximately 21 cm long. The ontogenetic implications of this situation are more fully discussed below, although several morphological changes in the largest humerus are obvious (Fig. 19A). The proximal muscle attachment sites have become more pronounced. The proximal head is much more flattened and expanded, and the subscapular process enlarged and distally de- flected. The site of insertion of the M. latissimus dorsi has become an elongate ridge lying just distal to and behind the much enlarged insertion scar of the M. deltoideus. Distally, the supinator process has been lost, but the supinator ridge is more pronounced and is directed ventrally, forming the lower face of the COROSAURUS ALCOVENSIS 35 humerus into a concave scoop. The ectepicondylar notch has been closed laterally and, along with the entepicondylar foramen, lies relatively farther from the distal end of the bone. Intermediate stages to these changes can be seen in the inter- mediately-sized humerus YPM 41033. Ulna. Contrary to Case (1936, p. 23), the left ulna of the type specimen is not crushed, but provides a good indication of the overall shape of this epipodial (Fig. 19B). It is a short, flat bone with expanded proximal and distal ends. The leading or anterior (internal) edge of the ulna is broadly concave, the posterior (external) edge more nearly straight. As opposed to the condition of the humerus, the dorsal surface of the ulna is, at least proximally, flatter than the ventral. The articular extremities of the bone are, as in all the limb elements, unfinished. The rounded, proximal articular surface is tear-shaped (Fig. 19C). The point of the tear drop forms the slight external expansion of the blunt “olecranon process.” The left ulna of UW 5485 is approximately 5.5 cm long. Radius. The radius of Corosaurus is a curved, narrow bone of approximately equal length to (or somewhat shorter than) the ulna (see Fig. 6). The proximal end is slightly enlarged, whereas the distal end is unexpanded in the type. How- ever, the radius impression of the slightly larger individual in YPM 41031 shows both articular ends as possibly expanded. ‘The radius is not much flattened and has a subcircular cross section. The curvature of both epipodial elements resulted in a large spatium interosseum, as noted by Case (1936, p. 23). The right radius of the type specimen is approximately 5.5 cm long. Carpus. Only four carpal bones of the type right forelimb are known. Case (1936) recognized only three of these. The largest element of the four, that which was partially lost during the original preparation (Case 1936, p. 23), is probably either the intermedium or ulnare. The smaller disks, averaging about 3 mm in diameter, are distal carpals. From the small size of the bones, it is apparent that the carpus of Corosaurus was poorly ossified and consisted largely of cartilage (even if one allows for some progressive ossification in older individuals). A conservative but reasonable restoration might place three small distal carpals and two larger (8 mm-—1 cm diameter?) proximal carpals (intermedium and ulnare) in the wrist as in Lariosaurus. Manus. The manus of Corosaurus is known only from the right forelimb of the type specimen (see Fig. 3). Fortunately, this foot remains largely articulated in a natural position. The typical, rod-shaped metacarpals are only slightly flattened, mostly as a result of diagenetic compression. Metacarpal III is the longest at 2 cm, metacarpal I the shortest at 9 mm. The first digit possesses two phalanges, the second, three. The terminal ungual phalanx of each of these digits is a blunt claw. Only a single phalanx remains articulated to metacarpal III. The remaining phalanges of the foot are not properly articulated, but fragments or impressions of nine of these are exposed beneath the manus along a fracture in the matrix. From the total number and position of the phalanges it is probable that no hyperphalangy was present in the manus, and a nearly primitive phalangeal formula is estimated. A reasonable reconstruction has a formula of 2-3-4-5-3. Hindlimb No hindlimb components are preserved with the holotype but numerous specimens have been found more recently. These include the left femur, tibia, and fibula, and a left metatarsal and a partial foot of FMNH PR480; what are presumably the right femur, fibula and the right tibia impression of FMNH PR1369; a femur 36 PEABODY MUSEUM BULLETIN 44 tib c Fic. 20. Right femur of Corosaurus alcovensis, based primarily upon YPM 41038 and YPM 41039. A, proximal aspect, anterior to top; B, extensor aspect, proximal end up; C, posterior aspect; D, anterior aspect; E, flexor aspect; F, distal aspect, anterior to bottom. int = internal trochanter; itr f = intertrochanteric fossa; pop = popliteal space; pvr = posteroventral ridge; tib c = tibial condyle. and one or possibly two tibiae from an individual in FMNH Lot No. PR135, a crushed tibia belonging to FMNH PR1368, and the proximal end of a left (?) femur (YPM 41055). All of this material is imbedded in matrix, but two undis- torted femora in the Yale collection, a left and a right (YPM 41038 and 41039, respectively), have been freed and are especially useful for descriptive purposes. Femur. ‘The femur of Corosaurus, as pointed out by Zangerl (1963, p. 120), was relatively longer than the humerus. This is evident from a comparison of specimens UW 5485 and FMNH PR480 which represent individuals of approximately equal size. The femur of FMNH PR480 is approximately 13.5 cm long. Unlike the humerus, the femur is little modified from the primitive reptilian condition. It is a slender, sigmoid bone with a long, cylindrical shaft (Fig. 20). The bone is expanded at both ends but is nowhere flattened. The extremities have rough, unfinished, articular surfaces. The proximal articular surface is irregularly tri- angular in outline. There is a very large, crestlike, internal trochanter and an only slightly smaller posteroventral ridge. Between these two ridges lies a broad, concave, triangular, intertrochanteric fossa in which lay the powerful M. pubo- ischiofemoralis externus. There is no fourth trochanter. The distal articular face of the femur is roughly semicircular. The two equisized tibial condyles are reduced relative to the primitive condition but remain distinct (Fig. 20F). There is, how- COROSAURUS ALCOVENSIS 37 ever, no intercondylar fossa, although ventrally a shallowly depressed popliteal space exists. No clear fibular facet is present. Tibiaand fibula. ‘The tibia of Corosaurus is a straight, thick bone of unremarkable appearance and apparently nearly circular cross section. The ends are rounded and a little expanded, both ends being as wide as the distal head of the femur (see Figs. 4 and 5). The tibia of FMNH PR480 is approximately 7 cm long. The fibula, on the other hand, is relatively thin and narrow, with a deeply concave internal edge indicating that a large spatium interosseum was present in the hind epipodium as well as in the fore (Fig. 5). The lateral or posterior edge of the bone is largely straight. The proposed fibula (FMNH PR480) of Zangerl (1963, p. 120) is probably the partly crushed and distorted left ischium. The proximal (2) end of the true fibula of FMNH PR480 projects from the matrix next to the femur. This end of the fibula is only 5 mm thick; it is concave on one side, convex on the other (Fig. 4). The rounded proximal head of the fibula is expanded and directed proxomediad. The distal end of the fibula is also flat and expanded, but has a rather squared-off articular face. The fibula was slightly shorter than the tibia. Tarsus and pes. A partial tarsus and pes (FMNH PR480) gives us an idea of the form of the hindfoot in Corosaurus (Fig. 4). The two large, disk-shaped elements (approximately 1.8 and 1 cm in diameter, respectively) are undoubtedly homologous with the fibulare and intermedium of the primitive reptilian tarsus. They have unfinished rims and depressed centers. A third circular bone (5 mm in diameter) is a distal tarsal. Up to five of these distal tarsalia may have been present in the living animal, but two to three is more likely. Like the carpus, the tarsus was obviously poorly ossified. Three metatarsals are preserved, the longest being 3 cm in length; in general they are much longer and stouter than the metacarpals. A fourth metatarsal is exposed near the distal end of the left femur, along with what are possibly several poorly preserved tarsals. The partial pes contains only one small phalanx (1.3 cm long). In keeping with the forelimb, however, a primitive phalangeal formula is assumed (2-3-4-5-4). RESTORATION Reconstructions of the skull and the limb girdles have been presented above. Now the complete description of Corosaurus alcovensis can be summarized and followed with a restoration of the entire skeleton of the animal (Figs. 21 and 22). Although the limb girdles and the skull display features that are unique to this genus, the gross morphology of Corosaurus is generally similar to that of other known notho- sauriforms. The body was narrow and elongate, and from the known lengths of ribs and gastralia, likely to have been broader than high. The tail was long and tapering, at least as long as the trunk, but was relatively shorter than the tails of the much smaller pachypleurosaurs and was not greatly compressed laterally. In fact, although caudal chevrons were present throughout much of the tail of Coro- saurus, the neural spines were rather low from the midtail region caudad. Throughout the vertebral column, the spines were somewhat rectangular and never high. There was, for example, no elongation of the spines in the shoulder region as is seen in some examples of Nothosaurus (Schmidt 1984). As the transitional “pectoral” rib position suggests placement of the anterior edge of the pectrum beneath the nineteenth vertebra of the column, the probable length of the neck has been established. It was long, thin, and serpentine, as in 38 PEABODY MUSEUM BULLETIN 44 NN Vy 3 i Wy YJ, ) a) DY) 10 cm SN, Uy, Hane Vf, 2 MM URS ©) [o} iy oO O = See oo DP pes aU <5 Ss, [J =oe - oa | Fic. 21. Skeletal reconstruction of Corosaurus alcovensis. A, ventral aspect; B, dorsal aspect. all primitive sauropterygians, but was only about half as long as the body. Many ‘nothosaurs’ had longer necks and at least one (Ceresiosaurus) possessed a neck that equaled the trunk in length. The head of Corosaurus was rather small and brevirostrine; it was also generally broader than high. ‘The total estimated length of the type individual, from the tip of the snout to the end of the tail, was approximately 165 cm. It must be emphasized, however, that individuals of Corosaurus could, and did, grow to much larger sizes, as evidenced by isolated elements. The limbs were long and specialized, but without well-formed osseous joints. The robust forelimb was strongly curved. The “feet”? were small and flat, pre- sumably with little or no hyperphalangy. ‘The metapodials were relatively short by ‘nothosaur’ standards, and were unexpanded. The hindlimbs of Corosaurus were at least 40% longer than the forelimbs in the type specimen, although this Mmmm 610 cm Fic. 22. Skeletal reconstruction of Corosaurus alcovensis, left lateral aspect. COROSAURUS ALCOVENSIS 39 is a proportion which may have changed during ontogeny (Zangerl 1963). The crus of Corosaurus was similarly longer than the antebrachium, and the pes longer than the manus. The ventral side of the body of Corosaurus was fitted with both a dense framework of interlocking gastralia, and expanded, platelike girdle as- semblies. DISCUSSION Through comparison of the skeleton of Corosaurus with those of other sauro- pterygians, it appears that many of its morphologic features display the presum- ably “primitive” character state (see Chapter 4), while certain others can be considered as ‘“‘advanced” or derived. Some of the latter serve as autapomorphies that define the taxon. Relative to other nothosauriforms, its axial skeleton has in general retained many apparently conservative traits. The appendages of Coro- saurus are rather unspecialized (although certainly adapted for aquatic use); yet the limb girdles are notably derived. Among the characters of the skull and vertebral column of Corosaurus that are perhaps primitive with respect to other nothosauriforms, are the short brevirostrine skull; large nasals, prefrontals, and postfrontals; relatively wide skull table with unfused skull table elements; posterolateral process of the frontal; intermediately- sized postorbital region; rather small, equisized upper teeth; generally conservative vertebrae; and existence of only three true sacral vertebrae. On the other hand, the slight elongation of the transverse processes of Corosaurus is unlike that of most ‘nothosaurs,’ but is reminiscent of the larger processes of plesiosaurs. Another plesiosaur-like and possibly derived trait is the presence of relatively large post- temporal fenestrae, creating an “open” occipital face. All other ‘nothosaurids’ in which the occiput is known have a “‘closed”’ occiput, that is, very small posttem- poral fenestrae. The relatively high temporal region of Corosaurus is also char- acteristic of plesiosaurs but can be observed in nothosauriforms such as Cyma- tosaurus and Lariosaurus as well. Although of general ‘nothosaur’ configuration, the limb girdles of Corosaurus are nevertheless uniquely derived relative to those of all other known nonple- siosaurian sauropterygians. The greatly expanded coracoids are, as detailed above, relatively larger and more rectangular in outline than any others known, and are without both the supracoracoid foramen and extreme median constriction of those of other genera. The result of these changes is a very massive, platelike pectrum. Even so, it does not greatly resemble those of plesiosaurs. There is no great posterior elaboration of the coracoids as is found in plesiosaurs (including Pis- tosaurus), no medial expansion of the ventral process of the scapula, no longitudinal division of the pectoral fenestra by a scapulocoracoid midline bar, and the dermal elements are well developed to form the anterior strut of the pectrum, whereas such anterior support is accomplished in plesiosaurs (in which the dermal elements of the shoulder girdle are vestigial or even lost) by the large ventral plates of the scapulae. The large pelvis of Corosaurus superficially resembles those of plesiosaurs, especially in the convex anterior border of the pubis. Corosaurus, however, prim- itively retains an obturator foramen that is lacking in plesiosaurs. The ilium of Corosaurus is also plesiomorphic, larger and better formed than that of any ple- siosaur and indeed, that of most nothosauriforms. The plesiosaur ilium articulates only with the ischium; Corosaurus and other ‘nothosaur’ ilia contact both the ischium and the pubis. 40 PEABODY MUSEUM BULLETIN 44 The classification and relationships of the ‘nothosaurs’ are more fully discussed in Chapter 4. However, as the supratemporal fenestrae of the skull of Corosaurus are larger than its orbits, the animal clearly falls into the nothosauriform clade (Chapter 4) as opposed to that containing the much smaller pachypleurosaurs in which the fenestrae are far smaller than the orbits. Additional comparisons be- tween Corosaurus and other sauropterygians can also be found in Chapter 4. COROSAURUS ALCOVENSIS 41 3. PALEOBIOLOGY INTRODUCTION The study and discussion of a fossil taxon should not be limited to the physical description of specimens but should include, where possible, interpretive analysis of its paleobiology. In the case of fossil vertebrates, preserved bones are only partially indicative of the whole-animal biology of the once living organisms. Among topics that may be addressed in a general study are the theoretical re- construction of unpreserved soft tissues, the functional morphology and behavior of the animal during life, the observed natural (biological) variation among in- dividuals, and the paleoecologic interaction of the animal with its environment. To the extent possible, these areas are here examined with respect to the skeletal anatomy of Corosaurus. Many of the following observations and speculations are also applicable to the Sauropterygia as a whole. ONTOGENETIC AND INDIVIDUAL VARIATION It was hoped at the outset of this study that a sufficient amount of new Corosaurus material could be collected to enable a detailed characterization of ontogenetic changes in the ‘nothosaur’ skeleton. While the growth patterns of the Alpine pachypleurosaurs, Neusticosaurus, “Pachypleurosaurus,” and Serpianosaurus have been discussed by Carroll and Gaskill (1985), Rieppel (1989), Sander (1988, 1989), and by Zangerl (1935), a Corosaurus growth series would be particularly valuable because of its closer relationship to advanced sauropterygians. An un- derstanding of nothosauriform ontogeny and variability could lead to a more critical evaluation of the taxonomic validity of certain characters within the group. A discussion of our current understanding of some of these characters follows in Chapter 4. Unfortunately, while many new Corosaurus specimens have been obtained, all those collected are of adult individuals, most represent only small portions of the entire animal or are isolated bones, and there is little correspondence between the elements represented in the sample. Even so, some variation is evident. The type specimen of Corosaurus alcovensis is presumed to represent a young adult individual. The skeleton is relatively large, all bones are well formed, and the sutures are tight though not fused. The texture of the cranial bones is rough in places and the orbits are not disproportionately large. Orbital size exhibits negative allometry in the Vertebrata (Dodson 1975). Juvenile specimens might be expected to be less well ossified, have a relatively larger head to body size ratio, and perhaps have a more abbreviated rostrum. The very young individual of the pachypleurosaur Kezchousaurus in Figure 23 illustrates this point, as do the ju- venile pachypleurosaurs illustrated by Peyer (1932, plate 29; 1944, fig. 39) and the Neusticosaurus embryo shown in Sander (1988, fig. 1; 1989, fig. 33). However, numerous examples of Corosaurus are comparatively larger, and probably onto- genetically older, than the holotype. Neither does the type display any evidence of age related pathology. Thus it apparently was not fully grown (although it must be admitted that absolute size is not always an accurate indication of relative age). Of course, the attainment of osteologic and sexual maturity are rarely coincident (Johnson 1977). Therefore, no presupposition of sexual maturity or immaturity may be made for specimens of Corosaurus as unequivocal size inde- pendent criteria for such determinations are unknown. The most obvious example of size (as the only available indicator of age), and 42 PEABODY MUSEUM BULLETIN 44 Fic. 23. Immature specimen of Keichousaurus in the collection of the Institute of Vertebrate Pale- ontology and Paleoanthropology, Beijing, exhibiting juvenile body proportions. Scale in one-sixteenths of an inch (1.6 mm). Photo courtesy of N. Mateer. probable ontogenetic, variation in Corosaurus is associated with the humeri of six separate individuals. These range from approximately 9 to 21 cm in length (Fig. 24). Volumetrically, the largest known humerus of Corosaurus (YPM 41032) is approximately 2.5 times larger than the humeri of the holotype. Progressive allometric changes occur most notably in the proximal and distal ends of each humerus in the sample and have been described above for the largest example (Chapter 3). Progressively thinner articular cartilages are assumed as described by Haines (1969) for recent reptiles, principally crocodilians and chelonians. The curvature and relative thickness of the humeral shaft remain constant throughout the sample but the ventral surface becomes increasingly “‘scooped.” As the type individual of Corosaurus is estimated to have been approximately 165 cm in total length, and assuming a crude 1:1 humerus length/total length scaling ratio, the large Corosaurus humerus may have belonged to an animal approaching 3.8 m long. This is the size reported by Peyer (1939) for the type of Paranothosaurus and is in the range of some Nothosaurus specimens. At least in crocodilians, however, relative limb size does not remain constant throughout ontogeny but is negatively allometric (Kalin 1955), although relative propodial size increases COROSAURUS ALCOVENSIS 43 cm UW 5485 YPM 41033 YPM 41032 Fic. 24. Partial ontogenetic series of left humerus of Corosaurus alcovensis, based upon UW 5485, YPM 41032, and YPM 41033. (Dodson 1975). Still, it is evident that individuals of Corosaurus occasionally grew to great size. The indeterminate, yet decelerating rate of growth of reptiles in general suggests that YPM 41032 represents a long-lived individual. Few other elements of Corosaurus can be directly compared or exhibit as large a size range. Most known vertebrae and ribs from corresponding areas of the axial skeleton are of similar size. Only one isolated, partial, median gastralium (YPM 41067) is significantly larger than any other. It is approximately 1.5 cm in maximum anteroposterior breadth versus about 8 mm for average specimens. The five relatively complete femora in the present Corosaurus sample are not greatly divergent in size. They range only from approximately 12.5 cm in length in YPM 41038 to approximately 15 cm in YPM 41039. Therefore, little mor- phologic variation is present among them. Distally, the tibial condyles are only slightly more pronounced in the larger specimen, whereas proximally the articular head is somewhat larger and joins the internal trochanter at a greater slope. If a femur specimen relatively as large as the aforementioned humerus were known, greater variation, perhaps extending these trends, might be seen. The known girdle elements of Corosaurus are all from animals of approximately equal size and ontogenetic variations cannot be shown. In light of the persistent cartilage of sauropterygian limb girdles, as discussed in Chapter 2, age variation may have been considerable, at least between juvenile stages. Some slight indi- vidual variability, however, is seen in the edges of the ventral plates of the girdles, particularly in the coracoid (Fig. 12C and D). Size may have been correlated with gender as it is in modern crocodilians where the male is generally larger than a female of equal age, environmental conditions being equal. Perhaps the position of the first haemal arch behind the cloaca (see Chapter 3) was a variable sexual trait. However, gender cannot be determined in any Corosaurus fossil. No pathologic variations are known in Corosaurus. 44 PEABODY MUSEUM BULLETIN 44 FUNCTIONAL MORPHOLOGY The nearly complete skeletal reconstruction of Corosaurus allows consideration of the potential movement and behavior of the animal as it may have operated while alive. The perfect, matrix-free nature and large size of some of the bones is a fortuitous circumstance allowing the three-dimensional study of numerous skeletal relationships. Of particular interest are the articulation and movement of the limbs and the presumed manner of Corosaurus locomotion. The swimming be- havior of sauropterygians has been a matter of conjecture for some time. Plesi- osaurs, for example, with a locomotor construction radically different from that of most vertebrates, have been claimed both as “rowers” utilizing fore-aft paddle strokes (Newman and Tarlo 1967; Tarlo 1957, 1959a; Watson 1924, 1951) and “underwater flyers” with vertical “wing” movement (Frey and Riess 1982; Rob- inson 1975, 1977; Tarsitano and Riess 1982; Taylor 1981). “Flying” is char- acteristic of modern penguins (Clark and Bemis 1979) and sea turtles (Walker 1971, 1974; Zangerl 1953). ““Rowing” is seen in seals (phocids) and sirenians (Webb and Blake 1985). It now seems probable that the power stroke of plesi- osaurs combined elements of the two styles, with both a vertical and a fore-and- aft (drag-based) component, more in the manner of present-day sea lions (otariids). Here the recovery stroke is primarily horizontal yet also provides thrust through lift because of the hydrofoil action of the limb (English 1976; Godfrey 1984). Although the morphology of Corosaurus, and of ‘nothosaurs’ in general, is far less removed from that of their terrestrial ancestors than is that of plesiosaurs, it may shed light on the functional evolution of the latter. Corosaurus is apparently not ancestral to plesiosaurs (see Chapter 4) but its derived appendicular skeleton may be partially analogous to that of the structural predecessor of plesiosaurs, thus perhaps indicative of the particular functional constraints and precursors which led to the successful invasion of a new functional niche. Carroll and Gaskill (1985) have discussed the question of possible functional relationships between ‘nothosaurs’ and plesiosaurs, particularly as they relate to pachypleurosaurs. At the very least, consideration of the functional morphology of Corosaurus will emphasize the differences in locomotion which obviously existed between the various sauropterygian types. Just as plesiosaurs maintained a single locomotor morphology and style throughout their known history (Robinson 1975), the ple- siomorphic ‘nothosaur’ pattern, appears to have remained relatively constant for nonplesiosaurian sauropterygians (the placodonts are excluded from the present discussion). Corosaurus was certainly an aquatic reptile as evidenced by its occurrence in the Alcova Limestone. Beyond this, its orbits and external nares are dorsal in position, the orbits are large, and the nares retracted. These are all adaptations common in secondary swimmers. The limbs, especially the forelimbs, are greatly modified from those of terrestrial vertebrates, as are the limb girdles, and there is a large percentage of persistent cartilage in the appendicular skeleton. Peyer (1934) and Zangerl (1935) list similar suites of aquatic adaptations observed in Lariosaurus and pachypleurosaurs. What is the functional role of these adaptations and can they be related to the adaptations of plesiosaurs? AQUATIC LOCOMOTION As with plesiosaurs, little agreement has been reached concerning the swimming style of the various forms of ‘nothosaur.’ Carroll and Gaskill (1985) have proposed COROSAURUS ALCOVENSIS 45 an axial propulsion mechanism: that they swam without the use of their limbs by lateral undulations of the tail. Sues (1987) and Sues and Carroll (1985) have accepted this interpretation. Kuhn-Schnyder (1987) gives credit to the forelimbs and tail for aqueous locomotion in Lariosaurus. Sanz (1976, 1980) and Schmidt (1984, 1986) have sided in favor of paraxial, limb-dominated propulsion (“sub- aqueous flight” and “rowing,” respectively) as has Storrs (1988a). It seems very likely that the tails of ‘nothosaurs’ were at times employed in swimming, especially in the forms with the longest tails such as the pachypleurosaurs. Long, powerful tails are characteristic of secondarily aquatic undulatory swimmers. The tail functionally extends and amplifies the undulations of the body. Crocodilians, for example, are noted for propulsive lateral undulations of the tail (Manter 1940). A strongly developed epaxial and hypaxial proximal caudal musculature is attested to in both pachypleurosaurs and plesiomorphic nothosauriforms by the broad shelf of their anterior caudal ribs and by the relatively tall proximal caudal neural spines. Medially and distally, the tails of Corosaurus and similar forms are deeper than broad by virtue of the neural arch and chevron configurations. This may be considered a sculling adaptation, but is plesiomorphic and often the case with terrestrial reptiles as well. The tail of Corosaurus is not exceptionally long, deep, or bilaterally compressed relative to those of reptiles in general. Another suggestive structural feature of Corosaurus is its stiffened trunk. The vertebral column craniad of the tail tends to be relatively stiff in caudal propulsors (Hildebrand 1974). The amphicoelous/platycoelous centra, zygosphene/zygan- trum articulations, broad neural arches, and closely spaced, rectangular neural spines of Corosaurus may all have served to limit flexibility between the precaudal vertebrae. This was probably also true of the densely packed gastralia of the ventral basket. However, because of such traits, the ‘nothosaur’ trunk was perhaps stiffer to a degree greater than in typical undulatory swimmers. The base of the tail in caudal propulsors must also be flexible when used for locomotion. The long proximal caudal ribs of ‘nothosaurs’ may have actually reduced flexibility here as well. On the other hand, the limbs of both pachypleurosaurs and traditional ‘notho- saurids’ must have played a major role in aquatic locomotion. The specializations exhibited by the limbs do not involve general reduction of hydrostatic drag as would be expected in animals using primarily their tails for thrust. Neither are they adapted for efficient terrestrial locomotion. Moreover, the specializations observed, especially in the forelimbs, act to increase the functional surfaces (wheth- er drag- or lift-based) of the limbs as is typical of paraxial swimmers. ‘The humeri of ‘nothosaurs’ are always distally expanded and flattened. Ventrally, they are flat or even “scooped.” The epipodials are universally shortened as well as ex- panded and flattened, and wide spatia interossea are present. The ulnae are particularly broad, especially in Keichousaurus (see Chapter 4). The manus and pes are always broad and flat, as in Keichousaurus (Young 1958) and Lariosaurus (see Boulenger 1896; Peyer 1933, 1934; Sanz 1976), or even, as in Cerestosaurus, display slight hyperphalangy (see Kuhn-Schnyder 1964; Peyer 1931, 1944). Fi- nally, the cartilaginous nature of all ‘nothosaur’ limb joints indicates limited in- tralimb flexibility. Hypothetical Myology What was the specific locomotor pattern of ‘nothosaur’ limbs? The limits of movement and the character of the associated musculature must first be deter- mined. As in plesiosaurs, the appendicular girdles of pachypleurosaurs and ‘noth- osaurids’ are largely platelike, with little elaboration outside the horizontal 46 PEABODY MUSEUM BULLETIN 44 -sch cr delt —~sup | sbcsc fat d sbcsc~™ flex Cc D Fic. 25. Left humerus of Corosaurus, based primarily upon UW 5485 and YPM 41033, with inferred points of muscle attachment. A, dorsal (extensor) aspect; B, anterior aspect; C, posterior aspect; D, ventral (flexor) aspect. br = M. brachialis; cbr b = M. coracobrachialis brevis; cbr 1 = M. coracobrachia- lis longus; delt = M. deltoideus (undivided); ext = extensors; flex = flexors; lat d = M. latissimus dorsi; pect = M. pectoralis; sbcsc = M. subcoracoscapularis; sch cr = M. scapulohumeralis cranialis; sup c = M. supracoracoideus; sup | = M. supinator longus; tri h = M. triceps humeralis. plane. The scapulae and ilia are the smallest components of their respective assemblies and the only elements with a significant vertical orientation. The ventral plates and stout median symphyses obviously acted to brace the body cavity against transverse compressive forces generated by the limbs, particularly in the pectrum in those forms with tightly sutured scapuloclavicular assemblies, as discussed by Watson (1924). They were also apparently the points of origin of major locomotor muscles, the positions of which can be crudely estimated. Watson (1924) discussed the possible disposition and function of the pectoral musculature of ‘nothosaurs’ based primarily upon the pectrum of Nothosaurus and a humerus of ‘‘?Conchiosaurus” (BMNH R. 1409) (probably also Nothosau- rus). The muscle insertion scars of the humerus as interpreted by Watson (1924) were duly figured. Other interpretations of ‘nothosaur’ humeral musculature position can be seen in studies by F. von Huene (1944, 1956) and Mazin (1985) [and Sues (1987) for Pistosaurus]. As noted in Chapter 2, the known humeri of Corosaurus, complete with muscle scars, also allow a hypothetical, yet reasonable, COROSAURUS ALCOVENSIS 47 Fic. 26. Reconstructed pectrum and humeri of Corosaurus with hypothetical musculature (ventral aspect). Right half of figure, deep musculature; left half, superficial musculature. cbr b = M. cora- cobrachialis brevis; cbr | = M. coracobrachialis longus; cl = clavicle; cor = coracoid; delt c = M. deltoideus clavicularis; h = humerus; icl = interclavicle; lat d = M. latissimus dorsi; pect = M. pectoralis; sc = scapula; sch cr = M. scapulohumeralis cranialis; sup c = M. supracoracoideus. reconstruction of pectoral muscle insertions (Fig. 25). The positions of presumably homologous muscles are drawn from comparison with modern reptiles and birds (e.g., Howell 1936; Jenkins and Goslow 1983; Romer 1944; Romer and Parsons 1977; Schreiweis 1982). Muscle relationships are relatively standardized in these groups. Nomenclature is largely that of Romer (1922, 1944). For the most part, the present humeral reconstruction differs little from that of Watson (1924). However, the M. latissimus dorsi seems to have been positioned far more dorsally in Corosaurus; more as shown by Watson (1924) for Plestosaurus dolichodeirus. The M. scapulohumeralis cranialis of Corosaurus also lies in a position similar to that of P. dolichodeirus. The large scar assumed by Watson (1924) to represent the insertion of this muscle in “?Conchiosaurus” probably marks the site of origin of the M. brachialis. The insertion of the M. coracobrachia- lis brevis appears to be located somewhat more proximally in Corosaurus than in “?Conchiosaurus.”’ No obvious scar exists for the M. coracobrachialis longus in Corosaurus, but this is presumed to have inserted along the mediodistal ventral face of the humerus. Watson (1924) has shown the M. coracobrachialis longus to have been present in ‘“‘?Conchiosaurus” and a distinct scar for this muscle occupies a similar position in Lariosaurus (Mazin 1985). The insertions of the M. pectoralis and M. deltoideus are located far more distally in Lariosaurus than in Corosaurus. Watson (1924), Tarlo (1957), and Robinson (1975) have further presented hypothetical reconstructions of plesiosaur pectoral musculature with respect to its origins on the shoulder girdle. Based upon the above humeral insertion re- construction, the form of its girdle elements, and the comparative myology of homologous structures in modern reptiles, a similar reconstruction is here at- tempted for the pectrum of Corosaurus (Figs. 26 and 27). A large M. coraco- brachialis, presumably with short and long branches inserting on the ventral face of the humerus, obviously arose deeply from the ventral surface of the coracoid. The expanded nature of this bone relative to that of other known ‘nothosaurs’ provided space for a possibly larger muscle. The anterior portion of the coracoid 48 PEABODY MUSEUM BULLETIN 44 Fic. 27. Reconstructed pectrum and humerus of Corosaurus with hypothetical musculature (lateral aspect, anterior to left). Compare with Fig. 14 A. A, deep musculature; B, superficial musculature. cbr b = M. coracobrachialis brevis; cbr | = M. coracobrachialis longus; cl = clavicle; delt c = M. deltoideus clavicularis; delt s = M. deltoideus scapularis; h = humerus; icl = interclavicle; lat d = M. latissimus dorsi; pect = M. pectoralis; sbcsc = M. subcoracoscapularis; sc b = scapular blade; sch cr = M. scapulohumeralis cranialis; sup c = M. supracoracoideus. was no doubt also the site of origin of part of the M. supracoracoideus. Inserting on the anteroproximal end of the humerus, this muscle also spread over the ventral surface of the scapula and probably the posterior margins of the ventral sides of the clavicle and interclavicle. Depending on the amount of cartilaginous and ligamentary support present, the M. supracoracoideus may have covered much or all of the pectoral fenestra. A prominent M. scapulohumeralis cranialis inserted on the dorsoproximal end of the humerus. Its origin apparently lay along the lower lateral half of the scapula, which is noticeably dished for its reception, and probably reached across part of the ventral surface of the clavicular shelf. ‘The size of this muscle may be reflected in the sharpness of the anterolateral corner of the clavicle. The medial face of the scapula of Corosaurus, except for the anterior part of its dorsal blade, is also cupped, presumably as the origin for the last major deep pectoral muscle, the M. subcoracoscapularis. Its insertion was on the pos- terior proximal end of the humerus. Part of the anterolateral edge of the visceral surface of the coracoid of Corosaurus probably also contributed to the origin of COROSAURUS ALCOVENSIS 49 the M. subcoracoscapularis. Whether or not this muscle was subdivided into two distinct rami (i.e., M. subscapularis and M. subcoracoideus) cannot be determined. The M. deltoideus and the M. pectoralis of Corosaurus presumably lay super- ficial to the deep ventral muscle masses of the pectrum. The M. deltoideus certainly formed two separate branches, the M. deltoideus scapularis and the M. deltoideus clavicularis. The first probably spread from the pronounced deltoid scar of the humerus to the spoon-shaped lateral side of the scapular blade. The second, from the scar to the lateral and anterior edges of the clavicle. ‘The M. pectoralis scar on the humerus is weak; so this muscle may not have been strong. It originates on much or all of the ventral surface of the pectrum in modern reptiles and this configuration was tentatively adopted for plesiosaurs by Robinson (1975, 1977). Watson (1924), however, limited the M. pectoralis origin to the posteriormost section of the coracoid and extended it onto the anterior gastralia, suggestions followed by Tarlo (1957). Neither of these positions can be directly confirmed in Corosaurus. Two alternatives are presented in Figures 26 and 27. Finally, the M. latissimus dorsi of Corosaurus evidently led from the proximodorsal end of the humerus to the anterior thoracic ribs. While the above reconstruction remains hypothetical, it suggests that the major part of forelimb movement in Corosaurus, and probably in most ‘nothosaurs,’ occurred in the horizontal plane. Apart from the seeming predominance of pectoral muscles occupying this plane and the force vectors they would have generated within it, the configuration of the glenoid and proximal head of the humerus of Corosaurus obviously favored horizontal movement. The glenoid articulation sug- gests possible adduction of the limb in the horizontal plane through an arc of perhaps 80°, from approximately 80° to 160° with respect to the longitudinal axis of the body (0° craniad). The strongly ovoid articular head of the humerus was vertically oriented and probably prevented vertical movement through an arc greater than 40° (approximately 20° of movement possible both above and below the horizontal). However, the proximal cartilaginous cap may have affected this figure to a certain extent. The articular configuration also indicates that the forelimb of Corosaurus could not be held in the rotated position suggested for “Pachypleurosaurus” by Carroll and Gaskill (1985) in their reconstruction of that animal. Additionally, propulsion through a primarily up-and-down limb stroke does not appear to have been possible because of the lack of significant skeletal support between the vertebral column and vertical elements of the pectrum (scap- ulae) as discussed by Godfrey (1984) for plesiosaurs. It is suggested that, in Corosaurus at least, forelimb “rowing” was possible whereby the horizontally held limb, beginning essentially perpendicular to the body, was adducted backward against the body together with a small downward component. This power stroke was accompanied by partial downward rotation of the anterior edge of the limb, especially along its distal half, thus providing a blade surface by which drag-based thrust could be created. Because the limb was not completely (perpendicularly) rotated, lift was also generated in the manner of a hydrofoil or “wing.” Rotational feathering occurred at the end of the power stroke so that the limb was abducted in a horizontal attitude, perhaps still providing lift. Both forelimbs probably acted simultaneously because of the stiffened nature of the thorax. This model is analogous to that suggested for plesiosaurs by Godfrey (1984) but because of their structural differences was probably less efficiently applied by ‘nothosaurs.’ It also approaches that of Tarlo (1957) (for pliosaurs). The greater structural and mechanical efficiency of plesiosaur subaqueous pro- pulsion was probably a result of their likely abandonment of terrestrial locomotion, possibly rudimentarily retained in ‘nothosaurs.’ Godfrey (1984) has shown that 50 PEABODY MUSEUM BULLETIN 44 the model of Watson (1924) overemphasizes the importance of a horizontal power stroke in sauropterygian swimming and ignores the possibility of a vertical ele- ment, just as that of Robinson (1975) exaggerates the role of vertical limb motion while suggesting that “rowing” is inefficient. The relationships of drag as a function of surface area, and momentum of mass, as discussed by Godfrey (1984) apply as equally to pachypleurosaurs and ‘nothosaurids’ as to plesiosaurs and otariids. Thrust that is produced in periodic pulses as this model suggests is therefore also presumed to be not inefficient in at least the large ‘nothosaurian’ nothosauriforms. The muscle assumed to have been primarily responsible for limb adduction in Corosaurus and other sauropterygians is the M. coracobrachialis. Its apparently large origin on the posterior expansion of the coracoid, and its relatively distal insertions on the humerus evidently produced a high degree of leverage and a powerful backwards stroke. Resultant stress vectors were directed predominantly towards the coracoid strut. More precise resolutions of forces cannot be calculated from a hypothetical muscle reconstruction. Also likely contributing to adduction were the M. pectoralis, M. subcoracoscapularis, and M. latissimus dorsi, although Sanz (1980) appears to have overstated the importance of the latter. The M. pectoralis also provided the downward movement, or depression, of the humerus as well as the downward rotation of its leading edge. Abduction was accomplished through the M. supracoracoideus, the M. scapulohumeralis cranialis, and the M. deltoideus. ‘The M. deltoideus scapularis appears to have been primarily respon- sible for the elevation and feathering of the humerus. The muscles presumably originating on the humerus and inserting on the antebrachium, rather than fa- cilitating intralimb flexion and extension, may have helped to stiffen the forelimb and adjust the trim of the “‘wing”’ as suggested by Robinson (1975) for plesiosaurs. These muscles include the M. brachialis, M. triceps humeralis, M. supinator longus, and the flexor and extensor groups. The hindlimbs of Corosaurus are far less specialized for aquatic locomotion than are the forelimbs. Indeed, in the plesiomorphic ‘nothosaurs’ in general, flexure of the femoral/epipodial joint is indicated in many articulated skeletons, whereas the forelimb is always essentially straight. The plesiosaur analogy in which both sets of limbs and their girdles are virtually identical, and thus assumed to have operated in a similar manner, does not strictly apply. As in the forelimb, however, the distal elements of the hindlimb (epipodials, metatarsals, and pes) are somewhat flattened and expanded. The large ventral plates of the pelvis anchored powerful musculature operating primarily in the horizontal plane. Did these muscles produce a “rowing” action of the hindlimbs similar to that of the forelimbs? The rather featureless femora of Corosaurus (all of young adults) display few muscle scars, making a reconstruction of muscle origins and insertions extremely difficult. Only the positions of the M. puboischiofemorales internus and externus are well known, although examination of living reptiles suggests the approximate positions of several other muscles. ‘The hypothetical positions of these major thigh muscles are indicated in Figure 28. Terminology is that of Romer (1923). Because the femoral muscle positions are so poorly known, and because of the complexity and variability of reptilian pelvic muscles, a complete reconstruction of the pelvic musculature of Corosaurus has not been attempted. However, certain of its features are known. As the internal trochanter of the femur is quite pro- nounced and the intertrochanteric fossa very large, it is obvious that a powerful M. puboischiofemoralis externus inserted at these points. Its origin would have covered a large part of the broad ventral surfaces of the pubis and ischium. A COROSAURUS ALCOVENSIS Syl f tib Fic. 28. Right femur of Corosaurus, based primarily upon YPM 41038 and YPM 41039, with inferred points of muscle attachment. A, dorsal (extensor) aspect; B, ventral (flexor) aspect. ad f = M. adductor femoris; cf = M. caudofemoralis; f tib = M. femorotibialis; if = M. iliofemoralis; ist = M. ischiotrochantericus; pif e = M. puboischiofemoralis externus; pif i = M. puboischiofemoralis internus. shallow, slightly rugose, depression on the anterodorsal surface of the internal trochanter probably marks the insertion of a somewhat smaller M. puboischio- femoralis internus, originating from the internal sides of the ventral girdle plates. The expanded anterior edge of the pubis of Corosaurus relative to those of other ‘nothosaurs’ may indicate increased leverage for larger M. puboischiofemorales. Although no fourth trochanter is present, the long ribs of the proximal caudal vertebrae indicate the presence of a large and powerful M. caudofemoralis, which probably inserted on the proximal posteroventral surface of the femoral shaft. It is likely that the M. adductor femoris, originating on the ventral face of the ischium, also inserted along much of this surface of the shaft. The wide inner face of the ischium of Corosaurus may have accommodated a well developed M. ischiotrochantericus. This muscle primitively inserts on the dorsoposterior surface of the femoral head in reptiles. The remnant blade of the ilium suggests retention of workable extensors of the thigh and lower leg, such as the M. iliofemoralis and M. quadriceps femoris of typical reptiles. The M. iliofemoralis usually inserts near the M. ischiotrochantericus. Only one branch of the complex M. quadriceps femoris contacts the femur, the M. femorotibialis, which typically has a fleshy origin along much of the dorsal and lateral surfaces of the femur. It seems clear that powerful fore and aft strokes of the hindlimb of Corosaurus were possible as in the forelimb. The M. adductor femoris, M. ischiotrochanteri- cus, and particularly the M. caudofemoralis provided adduction and presumably rotation; the large M. puboischiofemorales abduction and feathering of the limb. However, because of the bowllike nature of the acetabulum and the rather convex head of the femur, a large amount of rotation and polydirectional limb movement is postulated. This, together with the inferred presence of functioning dorsal 52 PEABODY MUSEUM BULLETIN 44 muscles on the limb, indicate that the hindlimb was probably not as restricted in its movements as was apparently the forelimb. These differences may have resulted from a greater role for the hindlimb in either rudimentary terrestrial locomotion or subaqueous directional control. It seems likely that steering was largely con- trolled by the attitude of the hindlimbs as in Alligator (Manter 1940). Discussion To summarize, the major propulsive force for aquatic locomotion in Corosaurus, and probably in most primitive nothosauriforms, was apparently paraxial “row- ing.” The drag-based thrust of limb adduction was quite likely augmented to some extent by hydrostatic lift as the limbs were concurrently depressed, and perhaps also by the hydrofoil action of their feathered return stroke, much as in otariid sea lions. All four limbs may have operated simultaneously with each stroke followed by a short gliding phase. Plesiosaurs are believed by Godfrey (1984) to have swum in a similar manner. On the other hand, the ‘nothosaur’ hindlimb may have been particularly important in steering. The still long tail of many ‘nothosaurs’ may have acted as a counterbalancing rudder, and possibly as an accessory thrust producing organ, initiating quick starts and rapid changes of direction, for example. Increased neck length and flexibility, disadvantageous in undulatory swimmers, were made possible. With reduction of the tail and con- tinued elaboration of the limbs, ‘nothosaur’-like animals would have made ideal functional precursors of plesiosaurs in which there was apparently no undulatory propulsion. Such a change probably coincided with complete abandonment of the land or paralic environments or both. It remains to consider why elongate, secondarily aquatic reptiles should have developed a limb dominated style of subaqueous locomotion rather than an un- dulatory style such as seen in lizards and crocodilians. The basic ingredients for undulatory swimming are already in place in the undulatory walking format of plesiomorphic “sprawlers.” A sprawling stance was undoubtedly present in the terrestrial forebears of the Sauropterygia. Several possibilities come to mind, each perhaps a contributing factor. Initially, undulatory swimming was probably oblig- atory for the immediate ancestors of the Sauropterygia. Limb reduction and developmental restructuring of the girdles as discussed by Carroll and Gaskill (1985), particularly with regard to pachypleurosaurs, may have followed. They have suggested a need for occasional terrestrial forays as a cause for limb reelabora- tion, although an amphibious capability is far from certain. If, however, a bottom dwelling or feeding mode of life was adopted by these animals, increased limb propulsion may have been advantageous in moving the body along, and/or pushing it off from, the substrate. Perhaps this form of behavior would have “‘preadapted” [exapted of Gould and Vrba (1982)] the limbs for aquatic propulsion. More interestingly, the requirement of neutral buoyancy in habitually aquatic animals may have played an important part in the transition to paraxial swim- ming. Organisms that can maintain a static position in the water column without expenditure of energy are at a decided advantange over those which cannot. ‘The tetrapod lung imparts secondarily aquatic vertebrates with a natural positive buoyancy that tends to float these animals to the surface unless counteracted. Darby and Ojakangas (1980) have shown that crocodiles voluntarily ingest stones as a probable hydrostatic compensation mechanism perhaps as, by analogy, did plesiosaurs. The pachyostotic nature of sirenian ribs is a well known buoyancy compensator. While no ‘nothosaur’ has been discovered with gastroliths, their ribs and gastralia are dense and often “‘pachyostotic.” Their limb girdles too are constructed of dense, heavy bone. It may be assumed that the development of a COROSAURUS ALCOVENSIS 53 thick ventral basket of closely-packed gastralia in ‘nothosaurs’ was a hydrostatic adaptation to their aquatic existence as suggested by Nopsca (1923a). This basket is far denser and more solid than that of crocodilians. As suggested above, the numerous gastralia of ‘nothosaurs’ probably severely limited the undulatory ca- pability of the trunk. In fact, in every known specimen of articulated pachypleu- rosaur or nothosauriform skeleton in which the ventral basket is intact, little flexion is exhibited by the largely straight abdomen. With flexibility reduced by such buoyancy compensation, paraxial propulsion would have added importance; would be developed, perhaps, by necessity. Sanz (1980) has suggested the presence of a ventral keel developed from the gastralia for aid in swimming, but this seems unlikely. ‘TERRESTRIAL LOCOMOTION A few words may be said about the ability of Corosaurus to navigate on land. ‘Nothosaurs’ are often assumed to have been amphibious (e.g., Colbert 1955, 1969; Romer 1933, 1945, 1966). Case (1936) pictured Corosaurus as emerging from the water to bask and lay eggs. Peyer (1931) did the same for Ceresiosaurus. While a habitually aquatic existence for all ‘nothosaurs’ is obvious, they are perhaps not specialized to the point of having lost their ability to come ashore. In Corosaurus, the feet do not seem to have been greatly hyperphalangic, if at all. However, sharp terminal claws, useful on land, are unknown in any sauropterygi- an. While reduced intralimb flexibility existed, especially in the forelimbs, some small movement was probably possible. The femur of Corosaurus retains obvious tibial condyles and the hindlimb was probably flexible to a relatively large degree. The strong caudofemoral musculature might have propelled the animal forward on land while the pelvis was elevated by the dorsal extensors of the ilium and femur. The stout sacral ribs and remnant iliac blade indicate a strong sacroiliac articulation that may have supported the posterior half of the body against the downward-acting force of gravity. The scapular blade was not as strongly sup- ported but was securely anchored by soft tissues. Schmidt (1984) has stated that the presence of elongate anterior dorsal neural spines in upper Muschelkalk specimens of Nothosaurus is evidence for terrestrial locomotion in this genus. Exposure to terrestrial gravity conditions might have necessitated such supporting structures for the head and neck. However, other explanations might exist, such as need for supporting, and facilitating rapid movement of, the neck when hunting. Corosaurus has uniformly short neural spines but may not have required as well developed a nuchal ligature because of its smaller head. Large amounts of leverage and increased structural support would have been even less necessary in the smaller pachypleurosaurs. If ‘nothosaurs’ maneuvered on land, their limited limb and thoracic flexibility would have mandated an awkward progression. The sinuous body and alternate limb movements of sprawlers probably were not possible. Carroll and Gaskill (1985) have suggested a crawling or dragging posture based upon symmetrical movements of the forelimbs. The forelimbs of some ‘nothosaurs,’ such as Neus- ticosaurus, are often stronger than the hindlimbs. The forelimbs of Ceresvosaurus are especially robust. The relatively long hindlimbs of Corosaurus, however, may have pushed the body forwards, whether the forelimbs pulled or not. Here too, the elaborate ventral armor might have been useful, along with the expanded ventral girdle plates, in protecting the underbelly as proposed by Carroll and Gaskill (1985), although no analogous armor is present in pinnipeds. In any 54 PEABODY MUSEUM BULLETIN 44 event, if either pachypleurosaurs or ‘nothosaurids’ were amphibious, they were far less at home on land than in the water, and terrestrial forays, if at all possible, were probably quite rare. Two isolated footprints have been interpreted as those of ‘nothosaurs’ and as providing direct evidence of terrestrial locomotion. The first is the ichnogenus Pontopus Nopsca, 1923b. This is the impression of an apparently webbed foot from the Upper Triassic of Cheshire, England, but was probably made by a terrestrial lacertiloid (see Appendix B). Secondly, F. von Huene (1935) described as of possible sauropterygian origin a small unwebbed print (Nothosauripus Kuhn, 1958a) from the pachypleurosaur-rich Ladinian shales of Besano, Italy. However, this print can not be directly linked with a known sauropterygian genus, and extensive quarrying by Peyer in the ‘nothosaur’-rich shales of Tessin failed to produce a single track (Zangerl, personal communication, 1986). The ichnological evidence is thus inconclusive. K. Thiessen (Arizona) reports (personal commu- nication, 1989) possible “swim-tracks” of an undescribed pachypleurosaur-like animal (younginiform?) from the Wupatki Member of the Moenkopi Formation of Arizona (see Chapter 7 for additional information on this occurrence). These are interesting, but rather nebulous “scratch” marks on bedding plane surfaces and do not bear directly on the question of ‘nothosaur’ terrestrial locomotion. No tracks of Corosaurus are known. The question of webbing in ‘nothosaur’ feet is also problematic. It may have been present in some forms while lacking in others. Peyer (1931, 1934) recon- structed Cerestosaurus and Lariosaurus with webbed feet (actually paddlelike fore- limbs in Lariosaurus), and Case (1936) followed suit with Corosaurus. Webbing would certainly have aided aquatic propulsion but would probably not have hindered movement ashore if ‘nothosaurs’ had this capability. PALEOECOLOGY It is assumed (Chapter 6) that Corosaurus was an indigenous element of the Alcova Limestone fauna. As such, the structure of Corosaurus and the paleoenvironment of the Alcova (see also Chapter 6) clearly indicate that this animal was a shallow water, largely nearshore marine form. While perhaps spending time basking along the shores of the Alcova “sea” in crocodile or seal fashion, most of its activity no doubt occurred subaqueously where it assumed a predaceous role. The apparent top carnivore of its ecosystem, as evidenced by its large size (perhaps partly due to a freedom from predators) and the absence of associated carnivores, Corosaurus is thought to have been primarily, if not exclusively, piscivorous. The long, sharp, recurved, conical teeth, particularly the large caniniforms of the anterior dentary, were especially well suited to piercing and gripping struggling prey. The stout retroarticular and coronoid processes of the mandible indicate high leverage of the depressor and levator musculature, respectively. These are coupled with the large supratemporal fenestrae, possibly accommodating strong adductors, so that a powerful bite is postulated. The strong dentaries were well braced by their relatively stout symphysis against the force of such a bite. All ‘nothosaurs’ display an elongate neck that may have made possible sweeping arcs of the head and jaws through which passing fish were intercepted. The longirostrine, anisodont format of certain nothosauriforms, considered ichthyo- and herpetophagous by Sanz (1980), would have increased their ability to seize and hold other vertebrates. Kuhn-Schnyder (1964) has documented the association of the large nothosauriform Ceresiosaurus with seven individuals of the much COROSAURUS ALCOVENSIS 55 smaller pachypleurosaur “Pachypleurosaurus” (= Neusticosaurus). Both he, Peyer (1932), and Sander (1989) also noted the bones of Neusticosaurus in coprolites presumed to belong to Ceresiosaurus. Remains of the palaeoniscoid fish Gyrolepis have been found in a “‘Nothosaurus” coprolite in the German Muschelkalk (Trush- eim 1937), while small, juvenile tooth plates of the placodont Cyamodus have been found in the body cavity of a Lariosaurus specimen from Monte San Giorgio (Kuhn-Schnyder 1987; Tschanz 1989). In the case of Corosaurus, however, in spite of its carnivorous adaptations, fish or other potential vertebrate prey have not yet been discovered in the Alcova. Nor have coprolites of Corosaurus been found that might illuminate its diet. Some pelecypods and gastropods were available but Corosaurus lacks obvious mollusc- ivorous adaptations, such as a crushing dentition. Sanz (1980) has suggested that Szmosaurus, with its short rostrum and spatulate teeth, possibly ate cepha- lopods but this is unsubstantiated. Mateer (1977) has also proposed that “Pachy- pleurosaurus”’ supplemented its diet with cephalopods. However, cephalopods are not known from the Alcova. I believe it still likely that Covosaurus ate fish but, perhaps due to environmental conditions (Chapter 6), their remains and the fecal pellets of Corosaurus have not been preserved. Corosaurus may have lingered underwater, possibly on the shallow bottom, preferring to wait for fish rather than actively pursuing them. The plesiomorphic ‘nothosaurs’ in general are from shallow paralic environments, whereas plesio- saurs, which may have been faster swimmers and chased their food, were open water forms. Sues (1987) has assumed that the earliest known plesiosaur, Pis- tosaurus, was ecologically isolated from contemporaneous littoral ‘nothosaurs’ by inhabiting offshore waters, thus accounting for its rare occurrence. He further considers the pachypleurosaurs to have inhabited lagoonal and shallow marine environments and the ‘nothosaurids’ only shallow marine ones. Nevertheless, the great size of some nothosauriforms, particularly Nothosaurus, rivaled that of Early Jurassic plesiosaurs and may have allowed them to parallel the plesiosaur niche in some cases. The habits of the young of Corosaurus are unknown. Tarlo (1967) suggests that juvenile ‘nothosaurs’ spent more time ashore than did their parents, although numerous immature pachypleurosaurs have been found in association with adults in subtidal marine environments. Tarlo (1967) also believes that the young may have been littoral scavengers, feeding upon the fish remains with which they have sometimes been found (Tarlo 1959c). It’s possible, however, that young ‘notho- saurs’ fed upon insects and other invertebrates as do juvenile crocodilians today (Sanz 1980). ‘Nothosaur’ reproductive function is equally speculative. Robinson (1977) pre- sumed ovoviviparity in plesiosaurs, but there is no direct evidence either way to suggest that either pachypleurosaurs or ‘nothosaurids’ bore live young or laid eggs. The “immature individuals” of Tarlo (1967) represent a separate genus from their supposed “mother”; as noted above, these seem to be the prey of Ceresiosaurus (Kuhn-Schnyder 1964). Elsewhere, as in China (Fig. 23) and the Alps, very young animals (pachypleurosaurs) are known which may be considered either hatchlings or newborns. Recently, Sander (1988) has described a likely embryonic Neusticosaurus specimen from Monte San Giorgio. However, for rea- sons which he discusses, it remains unclear whether this represents an egg without its shell preserved or is an aborted fetus. No gravid ‘nothosaur’ female has ever been found and all presently known evidence is ambiguous. The question of egg- laying capability in primitive sauropterygians is in part connected to the still open question of their amphibious ability. 56 PEABODY MUSEUM BULLETIN 44 4. PHYLOGENY AND TAXONOMY INTRODUCTION Since the description of Corosaurus in 1936 (Case), there has been little success in classifying this animal, and much disagreement among the numerous schemes suggested. The primary factors responsible for these problems have been the inadequate knowledge of the anatomy of Corosaurus and the lack of a sufficient understanding of sauropterygian relationships. The latter problem is itself a result of inadequate or misinterpreted fossil material and descriptions, of convoluted and unclear synonymies, and of poorly applied evolutionary and hence, phylo- genetic, taxonomic, and systematic, theory. These broader difficulties have been addressed to a certain degree in a number of recent works (Carroll 1981; Carroll and Gaskill 1985; Rieppel 1989; Schmidt 1987; Sues 1987; Taylor 1989) and will be further discussed here. Following Peyer’s (1934) classification, Case (1936) was unable to place Co- rosaurus in a more specific category than Nothosauria. The classification as con- structed was unable to accommodate the apparently conflicting characters of Corosaurus and Case (1936) believed that this placed the genus in a position possibly intermediate between Peyer’s (1934) two accepted families, Pachypleu- rosauridae and Nothosauridae. In 1948a, F. von Huene concluded that Coro- saurus was Closely related to Simosaurus, largely on the basis of proportional similarities in their skulls (e.g., their roughly triangular shapes and short snouts). On these rather shaky grounds, he united them in the family Simosauridae and further considered the two genera, largely by virtue of the postcrania of Corosaurus, to be primitive plesiosaurs (F. von Huene 1948a, b, c). Maintaining the view that Corosaurus was a primitive plesiosaur, he later (von Huene 1952, 1956) assigned it to the family Pistosauridae which he placed in the Plesiosauria, while shifting Simosaurus back amongst the ‘nothosaurs’. E. von Huene, however, had in 1949 judged Corosaurus to be a primitive ‘nothosaur.’ Romer originally (1945) classified Corosaurus as a ‘nothosaurid’ but in 1956 (and questionably in 1966), followed F. von Huene’s lead, calling the animal a ‘simosaurid,’ although placing the Simosauridae in the Nothosauria. Tatarinov and Novozhilov (in Orlov 1964) adhered to this scheme, as did Schultze and Wilczewski (1970). Zangerl (1963), who actually studied the available material of Corosaurus, interpreted the fossil as being that of an advanced ‘nothosaur’ but made no familial assignment. Kuhn (1961, 1964a, b) and Young (1965a) both placed Corosaurus in a monotypic family of its own (Corosauridae), albeit without any formal diagnoses (and with some hesitation from Young). Carroll and Gaskill (1985) and Storrs (1986a, b) returned Corosaurus to the Nothosauridae, although Carroll (1987) merely treats the genus as incertae sedis. As only three examinations of the original material of Corosaurus have been made [including within the present study, Storrs (1986a, b, 1990)], such taxonomic confusion as described above is not surprising. Now that our anatomical knowledge of Corosaurus is for the first time nearly complete, it has become possible to make a taxonomic assignment of the genus with a more reasonable degree of certainty. Nevertheless, before this can be done, the problem of sauropterygian relationships must be examined. COROSAURUS ALCOVENSIS 57 HISTORICAL CONCEPTS OF THE SAUROPTERYGIA The term Sauropterygia was coined by Owen (1860) to include both the ‘notho- saurs’ and the plesiosaurs (as well as the placodonts), two obviously related and important groups of Mesozoic marine reptiles. Romer (1956, 1966) reaffirmed this usage as a clear alternative to more recent and ambiguous designations. ‘That the ‘nothosaurs’ and plesiosaurs together form a monophyletic group can be clearly seen from their many shared derived characteristics, in both the skull and the postcranial skeleton (Sues 1987). Indeed, this relationship was recognized from the time of the earliest descriptions of these animals in the first half of the nineteenth century. One of the most obvious of the characters uniting the ‘nothosaurs’ and plesi- osaurs is the configuration of the temporal regions of their skulls. Both groups possess a single supratemporal fenestra on either side of the skull that is bounded medially by the postfrontal and parietal, and laterally by the postorbital and squamosal—a condition that Colbert (1945) described as euryapsid. ‘The upper temporal opening found in the bizarre reptilian order Placodontia has been con- sidered by many to be similarly euryapsid, and most workers have traditionally followed Owen (1860) and allied them with the sauropterygians. With this ad- dition, Williston (1925) placed the Sauropterygia under the resurrected subclass Synaptosauria of Cope (1885), a name reflecting his belief that they were possibly related to synapsids (Romer 1956). In the search for possible ancestors to the Sauropterygia, the desire to find presumed forebears possessing single supratemporal fenestrae has, understand- ably, been strong. Thus, in 1933, Romer expanded the Synaptosauria to include a poorly known assemblage of Permian/Triassic, primarily terrestrial, reptiles known as protorosaurs or araeoscelids (at that time including Araeoscelis, Proto- rosaurus, Tanystropheus, and Trilophosaurus) whose cranial anatomy seemed to be compatible with this desire. Romer (1933) believed that the construction of the temporal regions of the skulls of these animals was sufficiently close to that of ‘nothosaurs’ to warrant their consideration as the sauropterygian parent stock. In 1945, Colbert renamed the Synaptosauria as the Euryapsida to reflect the terminology associated with the clearly defined anapsid, synapsid, and diapsid skull conditions, and later (1969) included the Ichthyosauria, whose upper tem- poral openings had recently been found to be possibly comparable with those of the Sauropterygia (Romer 1968a). The ichthyosaurs may or may not be true euryapsids but, in any case, as extremely derived reptiles are of no great consequence in the present discussion of sauropterygian origins. If at all related, they likely diverged from an ancestral stock very early in the group’s history. Kuhn-Schnyder (1962, 1963a, 1967, 1980), following an early proposal by Jaekel (1910), has hypothesized the descent of sauropterygians from primitive, diapsid, eosuchian grade reptiles through the loss of the lower temporal arch, not the from the presumed solid-cheeked protorosaurs. Romer (1968b) has disputed this view, primarily as a result of his mistaken impression of the nature of the cheek region of the ‘Nothosauria.’ In fact, the cheeks of ‘nothosaurs’ are not solid as Romer (1933, 1945, 1966) had believed, but are of a fundamentally different nature from that ascribed to the protorosaurs (themselves problematical and probably representing several structural types). While often secondarily closed in the Plesiosauria (euryapsid is literally “broad arched’’) (and the placodonts?), the cheek in ‘nothosaurs’ is deeply emarginated, and the postorbital/squamosal arcade is very narrow (see, e.g., Fig. 8A). Carroll (1981) has reviewed the history of this anatomical misconception and demonstrated 58 PEABODY MUSEUM BULLETIN 44 the likelihood of a diapsid ancestry for the Sauropterygia. His description of Claudiosaurus provides a reasonable transitional analog to ‘nothosaurs’ and ples- iosaurs from the ‘Eosuchia’ and he included his new family Claudiosauridae in the Sauropterygia. Whatever their affinities, the protorosaurs are not ancestral to the Sauropterygia and are probably an artificial taxon (Kuhn-Schnyder 1980; Romer 1971). Benton (1985), Rieppel (1989), and Sues (1987) have further discussed the possible relationship of Claudiosaurus to the Sauropterygia. Each has expressed certain reservations regarding the exact placement of Claudiosaurus within the diapsid hierarchy, but all accept the inclusion of the Sauropterygia within Benton’s (1985) Neodiapsida. HISTORICAL CLASSIFICATIONS OF THE ‘NOTHOSAURIA’ The ‘nothosaurs’ themselves have, from the beginning, been frequently recognized as forming a well-defined group whose constituents largely share a single gen- eralized morphology. This recognition, however, is based on overall similarity rather than demonstrated synapomorphy and the group may actually be para- phyletic (Rieppel 1989; Sues 1987; see also discussion below). This circumstance, along with often incomplete anatomical knowledge of most forms, has resulted in a confused classification history of ‘nothosaur’ taxa. Some of the earliest reports of ‘nothosaur’ remains were produced in the mid- nineteenth century from specimens collected in the Triassic of Bavaria. Of these, the first study which combined thorough descriptions with adequate (in this case excellent) illustrations was provided by von Meyer (1847-55). Included in this work were descriptions of Nothosaurus, Pistosaurus, and Simosaurus. These fossils were known to be closely related to the plesiosaurs and were identified as such by von Meyer and contemporary workers. At the same time, similar animals were being reported from the Alpine region of southern Europe (e.g., Lariosaurus Curioni, 1847 and “Pachypleura” Cornalia, 1854). It was not until 1882, however, that the formal designation ‘Nothosauria’ was established as a taxon of subordinal rank. This was done by Seeley (1882) following his description of Neusticosaurus, although students of the ‘nothosaurs’ were already wrestling with the problem of familial associations. While Gervais had proposed the Simosauridae in 1859, the first widely rec- ognized family was the Nothosauridae (Baur 1889, in Zittel 1887-90), in which had been placed all the genera then known. This was quickly followed by the creation of the Lariosauridae Lydekker, 1889. Lydekker (1889) suggested that his Lariosauridae (Lariosaurus and Neusticosaurus) were perhaps transitional between plesiosaurs and the Nothosauridae (then including Nothosaurus, Pisto- saurus, Simosaurus, and “‘Conchiosaurus’’). It was Arthaber (1924), however, who made the first major attempt at an in-depth classification of the ‘Nothosauria,’ including all of the genera then known. He again distinguished the Lariosauridae from the Nothosauridae, but split the latter into two informal groupings. The Lariosauridae here consisted of Lariosaurus, Partanosaurus, and Proneusticosaurus. Neusticosaurus was shifted to the Nothosauridae which now additionally included Anarosaurus, Cymatosaurus, Dactylosaurus, ““Macromerosaurus,” “Pachypleura,” and Phygosaurus. Unfortunately, Arthaber’s (1924) classification contains several in- consistencies that reflect the difficulty of making systematic judgments among subjects that are not completely known and the difficulty of merely relying upon degree of overall similarity. Subsequent workers have been similarly hampered and the resultant classifications are confusingly varied. The most important of COROSAURUS ALCOVENSIS Se) these schemes are outlined chronologically in Table 1. Obviously, there has been little agreement on the nature of ‘nothosaur’ families, and no clear concept of many of the genera. Furthermore, it can also be seen from Table 1 that certain genera (e.g., Corosaurus, Cymatosaurus, Pistosaurus, Rhaeticonia, and Simosaurus) have at times been considered ‘nothosaurs,’ while at others, have been labeled plesiosaurs. THE GENERA OF ‘NOTHOSAURS’ In the present study of ‘nothosaurs,’ the traditionally included genera are first anatomically compared and taxonomically clarified. ‘The phylogenetic relation- ships of the Sauropterygia as a whole, and of the better known ‘nothosaur’ genera, are then examined cladistically using presumably valid (homogenetic) taxonomic characters. Table 1 lists over fifty names of ‘nothosaurs’ (plesiomorphic sauropterygians) that have previously appeared in the literature. Of these, many are now accepted as junior synonyms of other genera (some are actually misspellings). Corosaurus Case, 1936 is undoubtedly a valid genus as per the diagnosis presented in Chapter 2, and a taxon for which no synonyms exist. Other obviously or presumably valid genera (and traditionally accepted as such) include Anarosaurus Dames, 1890; Ceresiosaurus Peyer, 1929; Cymatosaurus Fritsch, 1894; Dactylosaurus Gurich, 1884; Keichousaurus Young, 1958; Lariosaurus Curioni, 1847; Neusticosaurus See- ley, 1882; Nothosaurus Munster, 1834; Pachypleurosaurus Broili, 1927; Paranotho- saurus Peyer, 1939; Proneusticosaurus Volz, 1902; Psilotrachelosaurus Nopsca, 1928b; Serpianosaurus Rieppel, 1989; and Szmosaurus v. Meyer, 1842. Each of these can be clearly identified as ‘nothosaurs’ (sensu lato) and, with the exception of Pachypleurosaurus, apparently generically differentiated through largely un- ambiguous morphologic criteria. For instance, Anarosaurus (Figs. 29C, 32A, and 33B) is a small supratemporal fenestra form in which the femur is significantly longer than the humerus in the apparent adult condition (Carroll and Gaskill 1985). This pronounced situation is unique among adult ‘nothosaurs’ with small temporal openings [juvenile pachypleurosaurs exhibit relatively long femora (Zan- gerl 1935, 1963)]. Anarosaurus may be further distinguished by its relatively robust humerus. Unfortunately, the type specimen was destroyed during World War II, although casts exist. Ceresiosaurus (Figs. 30D, 37C, and 39A) has, among other traits, large temporal fenestrae, a relatively long neck, massive clavicles and humeri, and slight hyperphalangy. These form a suite of characters suitable for generic distinction. While the body [other than the gastralia (Schrammen 1899)] of Cymatosaurus is unknown (Volz also assigned some questionable postcrania to C'ymatosaurus in 1902), the robust, longirostrine skull is obviously distinct from those of all other known ‘nothosaurs’ (Figs. 31B, 32D, and 34(C). It is proportionally similar to the skull of Prstosaurus v. Meyer, 1839 (Figs. 31D, 32F, and 33D) but unlike the latter, its small, splintlike nasals remain in contact with the borders of the external nares and it lacks an interpterygoid fenestra. It is therefore a ‘nothosaur’ in the traditional sense. The poorly known fossil Eurysaurus Frech, 1903 has regularly been equated with C'ymatosaurus, initially as a subgenus of the latter (e.g., Arthaber 1924), and considering the minor proportional differences in their skulls, Eurysaurus is here also viewed as a junior synonym of C'ymatosaurus. The genus Germanosaurus Nopcsa, 1928a was proposed in place of the preoccupied name Eurysaurus (Nopcsa, 1928b). Schultze (1970) has provisionally equated 60 PEABODY MUSEUM BULLETIN 44 TABLE 1. Chronological outline of the major historical classifications of the 'nothosaurs.’ I. ARTHABER (1924) Suborder Nothosauria Family Nothosauridae "Group I" Anarosaurus, Cymatosaurus (Eurysaurus), Dactylosaurus, Nothosaurus (Conchiosaurus), Pistosaurus, Simosaurus, (Lamprosaurus, Opeosaurus) "Group II" Macromerosaurus, Neusticosaurus, Pachypleura, Phygosaurus Family Lariosauridae Lariosaurus, Partanosaurus (?Microleptosaurus), Proneusticosaurus II. WILLISTON (1925) Suborder Nothosauria Family Nothosauridae Anarosaurus, Cymatosaurus, Dactylosaurus, Doliovertebra, Lamprosaurus, Lariosaurus, Microleptosaurus, Neusticosaurus, Nothosaurus, Partanosaurus, Pistosaurus, Proneusticosaurus, Simosaurus Il. NOPCSA (1928a & b) Suborder Nothosauroidea Family Pachypleuridae (Pachypleurosauridae) Subfamily Pachypleurinae (Pachypleurosaurinae) Anarosaurus, Dactylosaurus, Pachypleurosaurus Subfamily Neusticosaurinae Neusticosaurus Subfamily Simosaurinae Proneusticosaurus, Simosaurus Family Nothosauridae Subfamily Lariosaurinae Lariosaurus, Macromerosaurus, Phygosaurus, Psilotrachelosaurus (Philotrachelosaurus), Rhdticonia Subfamily Nothosaurinae Cymatosaurus, Germanosaurus (Eurysaurus), Nothosaurus, ?Pistosaurus IV. ROMER (1933) Suborder Nothosauria Family Pachypleurosauridae Neusticosaurus, Simosaurus Family Nothosauridae Ceresiosaurus, Lariosaurus, Nothosaurus V. PEYER (1934) Suborder Nothosauroidea Family Pachypleurosauridae Anarosaurus, Dactylosaurus, Neusticosaurus, Pachypleurosaurus, Phygosaurus, Psilotrachelosaurus Family Nothosauridae Ceresiosaurus, Cymatosaurus, Germanosaurus, Lariosaurus, ?Microleptosaurus, Nothosaurus, ?Paranothosaurus, Pistosaurus, Proneusticosaurus, ?Rhaticonia, Simosaurus Continued on next page COROSAURUS ALCOVENSIS TABLE 1 -- Continued VI. KUHN (1934) Suborder Nothosauria Family Nothosauridae Anarosaurus, Ceresiosaurus, Cymatosaurus, Dactylosaurus, Diplovertebra [sic] (Dolichovertebra), Germanosaurus (Eurysaurus), Lamprosaurus, Lariosaurus (Macromerosaurus, Macromirosaurus), Microleptosaurus, Neusticosaurus, Nothosaurus (Conchiosaurus, Dracosaurus, Phanerosaurus), Opeosaurus, Pachypleurosaurus (Pachypleura), Partanosaurus, Phygosaurus, Pistosaurus, Proneusticosaurus, Psilotrachelosaurus, Rhaeticonia, Simosaurus VII. ROMER (1945) Suborder Nothosauria Family Nothosauridae Ceresiosaurus, Corosaurus, Cymatosaurus, Doliovertebra, Germanosaurus, Lamprosaurus, Lariosaurus, Macromirosaurus, Microleptosaurus, Nothosaurus, Opeosaurus, Paranothosaurus, Partanosaurus, Proneusticosaurus, Rhaeticonia, Simosaurus Family Pachypleurosauridae Anarosaurus, Dactylosaurus, Neusticosaurus, Pachypleurosaurus, Phygosaurus, Psilotrachelosaurus Suborder Plesiosauria Infraorder Pistosauroidea Family Pistosauridae Pistosaurus VIII. v. HUENE (1948b) Suborder Pachypleurosauridea [sic] Family Pachypleurosauridae Pachypleurosaurus Family Proneusticosauridae Proneusticosaurus Suborder Nothosauridea [sic] Family Lariosauridae Lariosaurus Family Nothosauridae Nothosaurus Suborder Plesiosauroidea Family Cymatosauridae Cymatosaurus Family Pistosauridae Pistosaurus Family Simosauridae Corosaurus, Simosaurus IX. v. HUENE (1952) Suborder Nothosauroidea Family Proneusticosauridae Proneusticosaurus Family Pachypleurosauridae Anarosaurus, Dactylosaurus, Pachypleurosaurus, ?Phygosaurus, ?Psilotrachelosaurus, Simosaurus Continued on next page 61 62 PEABODY MUSEUM BULLETIN 44 TABLE 1 -- Continued Family Nothosauridae Ceresiosaurus, Lariosaurus, Metanothosaurus, Neusticosaurus, Nothosaurus, Paranothosaurus, ?Partanosaurus Suborder Plesiosauroidea Family Cymatosauridae Cymatosaurus, Germanosaurus, Rhaeticonia, Sulmosuchus Family Pistosauridae Corosaurus, Pistosaurus X. SAINT-SEINE (1955) Suborder Nothosauria Family Nothosauridae viype 1s Cymatosaurus, Eurysaurus (Germanosaurus), Nothosaurus, Paranothosaurus “ype, 2. Ceresiosaurus, Lariosaurus, Simosaurus Indet. "others" Proneusticosaurus (Doliovertebra), Macromerosaurus, Parthanosaurus Family Pachypleurosauridae Anarosaurus, Dactylosaurus, Neusticosaurus, Pachypleurosaurus (Pachypleura), Phygosaurus Suborder Plesiosauria Superfamily Pistosauroidea Family Pistosauridae Pistosaurus XI. v. HUENE (1956) Suborder Pachypleurosauroidea Family Pachypleurosauridae Dactylosaurus (Anomosaurus), Pachypleurosaurus, Phygosaurus, Psilotrachelosaurus, Rhaeticonia, Neusticosaurus Family Proneusticosauridae Proneusticosaurus (Dolichovertebra, Lamprosaurus) Suborder Nothosauroidea Family Lariosauridae Lariosaurus (Macromerosaurus) Family Nothosauridae Ceresiosaurus, Metanothosaurus, Microleptosaurus, Nothosaurus, Paranothosaurus, Parthanosaurus Family Simosauridae Anarosaurus, Conchiosaurus, Simosaurus (Opeosaurus) Suborder Plesiosauroidea Family Cymatosauridae Cymatosaurus (Germanosaurus), ?Sulmosaurus [sic] Family Pistosauridae Corosaurus, Pistosaurus Continued on next page COROSAURUS ALCOVENSIS 63 TABLE | -- Continued XII. ROMER (1956) Suborder Nothosauria Family Nothosauridae Ceresiosaurus, Lariosaurus (Macromerosaurus, Macromirosaurus), ?Metanothosaurus, Nothosaurus (Conchiosaurus, Condriosaurus), Dracontosaurus, Dracosaurus, Kolposaurus, Oligolycus, Opeosaurus), Paranothosaurus, ?Parthanosaurus (?Microleptosaurus, Partanosaurus), ?Proneusticosaurus Family Cymatosauridae Cymatosaurus (Eurysaurus, Germanosaurus), Rhaeticonia Family Pachypleurosauridae Neusticosaurus (?Anarosaurus, ?Dactylosaurus, ?Philotrachelosaurus, ?Phygosaurus, ?Psilotrachelosaurus), Pachypleurosaurus (Pachypleura) Family Simosauridae ?Corosaurus, Simosaurus Nothosauria incertae sedis Deirosaurus, Doliovertebra, Lamprosauroides (Lamprosaurus) Suborder Plesiosauria ?Superfamily Pistosauroidea Family Pistosauridae Pistosaurus XII. KUHN (1964a) Suborder Nachangosauria Family Nachangosauridae Nachangosaurus Family Nothosauravidae Nothosauravus Suborder Nothosauria Family Nothosauridae Ceresiosaurus, Keichousaurus, Kwangsisaurus, Metanothosaurus, Microleptosaurus, Micronothosaurus, Nothosaurus, (Dracontosaurus, Dracosaurus, Kolposaurus, Oligolycus), Paranothosaurus, Parthanosaurus, Proneusticosaurus (Dolichovertebra, Doliovertebra) Family Lariosauridae Lariosaurus (Macromerosaurus, Macromirosaurus) Family Cymatosauridae Cymatosaurus (Eurysaurus, Germanosaurus), Rhaeticonia Family Pachypleurosauridae Anarosaurus, Dactylosaurus, Elmosaurus, Neusticosaurus, Pachypleurosaurus (Pachypleura), Phygosaurus, Psilotrachelosaurus (Philotrachelosaurus) Family Simosauridae Simosaurus Family Corosauridae Corosaurus Continued on next page 64 PEABODY MUSEUM BULLETIN 44 TABLE 1 -- Continued Nothosauria incertae sedis ?Charitosaurus, Conchiosaurus, Deirosaurus, Eupodosaurus, Lamprosauroides (Lamprosaurus), 7Namuncurania, ?Ocoyuntaia, Opeosaurus Suborder Plesiosauria Superfamily Pistosauroidea Family Pistosauridae Pistosaurus XIV. TATARINOV and NOVOZHILOV (in Orlov 1964) Suborder Nothosauria Family Lariosauridae Lariosaurus (Macromerosaurus, Macromirosaurus), Neusticosaurus, Nothosauravus, Parthanosaurus (Microcletosaurus [sic], Partanosaurus) Family Pachypleurosauridae Subfamily Pachypleurosaurinae Elmosaurus, Keichousaurus, Pachypleurosaurus (Pachypleura), Phygosaurus, Psilotrachelosaurus, Rhaeticonia Subfamily Proneusticosaurinae ?Lamprosauroides (Lamprosaurus), Proneusticosaurus (Dolichovertebra) Family Simosauridae Anarosaurus, Conchiosaurus (Condriosaurus), Corosaurus, Dactylosaurus (Anomasaurus [sic]), Simosaurus (?Opeosaurus) Family Nothosauridae Ceresiosaurus, ?Deirosaurus, ?7Kwangsisaurus, Metanothosaurus, Nothosaurus (Dracontosaurus, Dracosaurus, Kolposaurus, Oligolycus), Paranothosaurus, Pontopus Suborder Plesiosauria Superfamily Pistosauroidea Family Cymatosauridae Cymatosaurus (Eurysaurus, ?Germanosaurus), ?Sulmosaurus Family Pistosauridae Pistosaurus XV. YOUNG (1965a) Suborder Pachypleurosauroidea Family Pachypleurosauridae Pachypleura, Pachypleurosaurus, Rhaeticonia Family Keichousauridae Keichousaurus Family Simosauridae Anarosaurus, Elmosaurus, Shingyisaurus, Simosaurus Suborder Nothosauroidea Family Nothosauridae Ceresiosaurus, Chinchenia, Kwangsisaurus, Metanothosaurus, Nothosaurus, Paranothosaurus, Sanchiaosaurus. Family Lariosauridae Lariosaurus Continued on next page COROSAURUS ALCOVENSIS 65 TABLE 1 -- Continued Family Cymatosauridae Cymatosaurus, Germanosaurus ?Family Corosauridae Corosaurus XVI. ROMER (1966) Suborder Nothosauria Family Nothosauridae Ceresiosaurus, Deirosaurus, Keichousaurus, ?Kwangsisaurus, Lariosaurus (Macromerosaurus, Macromirosaurus), Metanothosaurus, Micronothosaurus, ?Nachangosaurus, ?Nothosauravus, Nothosaurus (Chondriosaurus [sic], Conchiosaurus, Dracontosaurus, Dracosaurus, Kolposaurus, Menodon, Oligolycus), Paranothosaurus, ?Parthanosaurus (?Microcleptosaurus {sic]), Pontopus, Proneusticosaurus (Dolichovertebra, ?7Lamprosaurus, ?Lamprosciuroides [sic]) Family Pachypleurosauridae Neusticosaurus (Anarosaurus, 7Anomosaurus, Dactylosaurus, ?Philotrachelosaurus, ?Phygosaurus, ?Psilotrachelosaurus), Pachypleurosaurus (Pachypleura) Family Simosauridae ?Corosaurus, ?2EImosaurus, Simosaurus (Opeosaurus) Suborder Plesiosauria ?Superfamily Pistosauria Family Pistosauridae Pistosaurus Family Cymatosauridae Cymatosaurus (?Eurysaurus, Germanosaurus), ?Rhaeticonia, ?Sulmosaurus XVII. CARROLL AND GASKILL (1985) Suborder Nothosauria Family Nothosauridae Ceresiosaurus, Corosaurus, Lariosaurus, Nothosaurus, Paranothosaurus, Simosaurus Family Pachypleurosauridae Anarosaurus, Dactylosaurus, ?Elmosaurus, Keichousaurus, Neusticosaurus, Pachypleurosaurus, ?Phygosaurus, ?Psilotrachelosaurus XVIII. CARROLL (1988) Order Nothosauria Family Cymatosauridae Cymatosaurus (Germanosaurus, Micronothosaurus) Family Nothosauridae Ceresiosaurus, Lariosaurus (Macromerosaurus), Nothosaurus (Conchiosaurus, Dracosaurus, Oligolycus), Paranothosaurus, Proneusticosaurus, ?Rhaeticonia Continued on next page 66 PEABODY MUSEUM BULLETIN 44 TABLE 1 -- Continued Family Pachypleurosauridae Anarosaurus, Dactylosaurus, Keichousaurus, Neusticosaurus, Pachypleurosaurus (Pachypleura), ?Psilotrachelosaurus, Phygosaurus Family Pistosauridae Pistosaurus Family Simosauridae Simosaurus (? Opeosaurus) Nothosauria incertae sedis Corosaurus, Elmosaurus, Kwangsisaurus, Metanothosaurus, Parthanosaurus XIX. TSCHANZ (1989) Order Sauropterygia Pachypleurosauroidea Family Pachypleurosauridae Anarosaurus, Dactylosaurus, Keichousaurus, Neusticosaurus, Pachypleurosaurus, Serpianosaurus Eusauropterygia Family Simosauridae Simosaurus Eusauropterygia, Nothosauria Family Nothosauridae Nothosaurus, Paranothosaurus Family Lariosauridae Ceresiosaurus, Lariosaurus Sauropterygia incertae sedis Corosaurus, Elmosaurus, Kwangsisaurus, Micronothosaurus, Proneusticosaurus, Psilotrachelosaurus, Rhaticonia [sic] Micronothosaurus Haas, 1963 with Cymatosaurus on the basis of their similar postorbital cranial proportions and centrally located pineal foramen. This is a reasonable proposal, one that is provisionally followed here in the absence of more useful fossil material. Dactylosaurus (Figs. 30A and 37E), while similar to Anarosaurus, lacks the relatively elongate femur of the latter, although Carroll and Gaskill (1985) cite the presumably greater degree of skeletal ossification and possible pisiform bone of Dactylosaurus as sufficient distinctions. Sues and Carroll (1985) cite also the relatively gracile nature of Dactylosaurus. Its epipodials are particularly slender. Keichousaurus (Figs. 23, 29D, and 36F) is undoubtedly a typical pachypleurosaur, but possesses an extremely broad ulna as elsewhere observed only in Lariosaurus. Lariosaurus, on the other hand, is a form with large temporal openings (Figs. 31A, 34B, 37B, and 38B). Features comprising the unique character suite of Larwosaurus include its relatively small size, thickened ribs, and five sacral vertebrae with costae of uniform diameter. Macromerosaurus Curioni, 1847 emend. Cornalia, 1854 is a junior synonym of Lariosaurus, apparently having been founded on a juvenile specimen of the latter. The traditional genera Pachypleurosaurus and Neusticosaurus (Figs. 29A and B, 32B, 33A, 36G and H, and 38A) are obviously closely related and following Carroll and Gaskill (1985), are most notably distinguished from other taxa by their common exclusion of the postorbital from the supratemporal opening. ‘They COROSAURUS ALCOVENSIS 67 Fic. 29. Reconstructed skulls of pachypleurosaurs in dorsal aspect. A, Neusticosaurus modified from Carroll and Gaskill (1985), Mateer (1976) and Sander (1989); B, “Pachypleurosaurus” after Carroll and Gaskill (1985); C, Anarosaurus modified from Carroll (1981) after Nopsca (1928b); D, Kezchou- saurus modified from Young (1958) after photographs (courtesy of N. Mateer). Compare with Fig. 7. Scale bars = 0.5 cm. both have small temporal fenestrae, thickened ribs, relatively short femora, and three [variably four (Zanger] 1935)] sacral vertebrae. The sacral ribs are relatively unexpanded at their distal ends. Carroll and Gaskill (1985) distinguished Neus- ticosaurus from Pachypleurosaurus by, among other things, the smaller temporal openings, relatively narrow skull table, smaller humerus to femur length ratio, slightly greater phalangeal formula, relatively broader ribs, and generally smaller size of Neusticosaurus.The holotype of Neusticosaurus (BMNH R53) and nu- merous small, probably juvenile, fossils described by Fraas (1896), are from the Germanic Province of central Europe. Specimens once identified as Pachypleu- rosaurus, on the other hand, are common in the famous shales of Monte San Giorgio, Switzerland, and adjacent localities of the Alpine Triassic. Pachypleu- rosaurus is the name coined by Broili (1927) and used coincidentally by Nopcsa (1928a) to replace the preoccupied Pachypleura Cornalia (1854). Carroll and Gaskill (1985) have attempted to draw an adequate anatomical distinction between these taxa. Their study suggests that many previously described fossils have been erroneously assigned to Pachypleurosaurus (e.g., certain specimens discussed by Kuhn-Schnyder 1959; Mateer 1976; Zangerl 1935; and other authors), while actually representing Neusticosaurus. Rieppel (1989) has accepted the conclusions 68 PEABODY MUSEUM BULLETIN 44 C— Dista- Fic. 30. Reconstructed skulls of various ‘nothosaurs’ in dorsal aspect. A, Dactylosaurus after Nopsca (1928b) and Sues and Carroll (1985); B, Simosaurus after Kuhn-Schnyder (1961); C, Corosaurus; D, Ceresiosaurus after Peyer (1931). Compare with Fig. 7. Scale bars = 1.0 cm. of Carroll and Gaskill (1985), while most recently Sander (1989) has equated all examples of the two genera as specific variants of Neusticosaurus, the name having priority. This interpretation is welcomed and accepted here, although for purposes of clarity, Neusticosaurus and “‘Pachypleurosaurus” are retained as distinct in the present phylogenetic analysis. Serpianosaurus (Fig. 37D), a relatively ple- siomorphic pachypleurosaur from the Grenzbitumen horizon of Monte San Gior- gio, has recently been described by Rieppel (1989) as a close relative of Neusti- cosaurus. Serpianosaurus is a small to intermediate-sized pachypleurosaur with a relatively large skull, straight mandible, and often nonthickened ribs. The history of these taxa and other pachypleurosaurs has been reviewed by Rieppel (1987). A pair of similarly related forms is Nothosaurus and Paranothosaurus (Figs. 31C; 32E; 34D; 35C; 36A, B and E; and 39B). These are large, derived reptiles that are characterized by long, massive skulls with blunt, constricted premaxillae and extremely elongate supratemporal fenestrae. This unique cranial format is readily identifiable, but the skull of Nothosaurus is virtually indistinguishable from that of Paranothosaurus |see, e.g., Kuhn-Schnyder (1966) and Schultze (1970)]. Kuhn-Schnyder (1987), however, indicates that the postfrontal is excluded from the margin of the supratemporal fenestra in Paranothosaurus, in contrast to the COROSAURUS ALCOVENSIS 69 Ce = Db = Fic. 31. Reconstructed skulls of nothosauriform ‘nothosaurs’ and Pistosaurus in dorsal aspect. A, Lariosaurus after Mazin (1985) with modifications according to Kuhn-Schnyder (1987); B, Cymato- saurus after Arthaber (1924), Fritsch (1894), and Schrammen (1899); C, Nothosaurus after Schultze (1970); D, Pistosaurus after von Meyer (1847-55) and Schrammen (1899). Compare with Fig. 7. Scale bars = 1.0 cm. apparent condition in other known ‘nothosaurids.’ The pectra of the two genera (Figs. 36A, B and E; and 39B), however, are radically different. While the type and only specimen of Paranothosaurus is undoubtedly fully grown, its pectrum is very weakly developed and greatly reduced relative to that of Nothosaurus. It differs also by its open coracoid notch and barlike interclavicle. Another apparently distinct genus is also known only from its (incomplete) type specimen. Its skull is not preserved and many details of its remaining anatomy are unclear. Nevertheless, it has generally been treated as a valid taxon. Psvlo- trachelosaurus apparently has a uniquely long and slender coracoid (Fig. 37F) and ulnae which are two-thirds the length of the radii. In his listing of reptilian genera, Nopcsa (1982a) identifies this fossil as ‘“Philotrachelosaurus.” However, as the full description and designation of the fossil as a new genus appears in a separate paper (Nopcsa, 1928b) consistently naming the animal Pszlotrachelosau- rus, the former name is obviously a misprint and the latter should be considered the available name. Several specimens of Proneusticosaurus are known; yet these are no less frag- mentary. Proneusticosaurus 1s seemingly different from all presently known ‘notho- Saurs’ in possessing six sacral vertebrae (Arthaber 1924; Volz 1902). Although 70 PEABODY MUSEUM BULLETIN 44 Fic. 32. Reconstructed skulls of various ‘nothosaurs’ and Pistosaurus in palatal aspect. A, Anarosaurus after Carroll (1981); B, “Pachypleurosaurus” after Carroll and Gaskill (1985); C, Simosaurus after Jaekel (1910) and von Huene (1921); D, Cymatosaurus after Arthaber (1924), Fritsch (1894), and Schrammen (1899); E, Nothosaurus after Schroeder (1914); F, Prstosaurus after von Meyer (1847-55) and Schrammen (1899). ecpt = ectopterygoid; mx = maxilla; pal = palatine; pmx = premaxilla; pt = pterygoid; q = quadrate; v = vomer. Scale bars = 1.0 cm. from a large animal, the ribs of Proneusticosaurus are thickened and the sacral ribs have little or no distal expansion. The large calcaneum and astragalus are subequal in size. The ischia are noteworthy by virtue of their relatively great breadth and length. The thyroid fenestra is rather small. Cranial material of Proneusticosaurus is unknown. Finally, of the aforementioned genera, Simosaurus (Figs. 30B, 32C, 33C, 35A, and 37A) is perhaps the most distinctive. Its large, round, supratemporal fenestrae and moderately long postorbital region combined with the brevirostrine antorbital area make the skull easily recognizable. This skull is large and massive. Whereas most ‘nothosaurs’ seem to have had long, slender, conical teeth, those of Simosaurus are short, squat, and deeply striated. On the basis of such teeth, and by virtue of similar size, locality, and age, von Huene (1952) attributed a partial postcranial skeleton to Sizmosaurus. A second postcranial skeleton (this time with skull) was described in 1959 (v. Huene 1959a). These display other diagnostic features including five sacral ribs, stout limb girdles, and a large, posteriorly projecting interclavicle. COROSAURUS ALCOVENSIS Tal Fic. 33. Reconstructed skulls of various ‘nothosaurs’ and Pistosaurus in lateral aspect. A, ““Pachy- pleurosaurus” modified from Carroll and Gaskill (1985); B, Anarosaurus after Carroll (1981); C, Sitmosaurus after von Huene (1921); D, Prstosaurus after von Meyer (1847-55) and Schrammen (1899). Compare with Fig. 8, A. a = angular; ept = epipterygoid; op = opisthotic; | = lachrymal; qj = quadratojugal; sa = surangular. Scale bars = 1.0 cm. Some additional fossils possibly represent valid ‘nothosaur’ (primitive sauro- pterygian) genera, although they are rather poorly known. Rhaeticonia Broili, 1927 [initially anglicized by Woodward (in Zittel 1932) from the original Rha- ticonia to conform with English nomenclatural practice] was known only from a single, very small (juvenile?) skeleton (destroyed during World War II). It had thick, “‘pachyostotic” ribs and vertebrae, and stout humeri. The skull was possessed of a conspicuously narrow, medium-length rostrum which seems to set this ‘no- thosaur’ apart from most others which are presently known. The nature of the skull roof and temporal fenestrae is unknown. Metanothosaurus Yabe and Shikama, 1948 from Japan is certainly a ‘nothosaur’ in the traditional usage, but the headless, partial vertebral column of the holotype (now lost) is of little diagnostic value. It is a large animal and, being the first ‘nothosaur’ discovered in Asia, was made the type of a new genus, primarily on the basis of its large size, high neural spines, and extremely slender ribs (“‘costae’’). This status is dubious but may be P2. PEABODY MUSEUM BULLETIN 44 Fic. 34. Reconstructed skulls of nothosauriform ‘nothosaurs’ in lateral aspect. A, Corosaurus; B, Lariosaurus modified from Mazin (1985), Kuhn-Schnyder (1987) and Tschanz (1989); C, Cymatosaurus modified from Arthaber (1924) and Schrammen (1899); D, Nothosaurus modified from Carroll (1981). Compare with Fig. 8, A. ept = epipterygoid; op = opisthotic. Scale bars = 1.0 cm. tentatively maintained in the hope that new, more definitive material may be forthcoming. Young (Yang) has more recently described several additional Asian ‘notho- saurs, including Keichousaurus. While Keichousaurus, as noted above, is an animal with a relatively well-understood anatomy, Chinchenia Young, 1965a; Kwang- sisaurus Young, 1959; Sanchiaosaurus Young, 1965a; and Shingyisaurus Young, 1965a are known from fewer, and more fragmentary, specimens. Each of these forms is from the Triassic of China and the morphology of each appears to justify its position as a separate taxon. Young (1958, 1959, 1960, 1965a, 1972, and 1978) has discussed their differences in greater detail but, for our purposes, Chinchenia is noted for its deep and massive lower jaw and mandibular symphysis, and sharply anisodont dentition; Kwangsisaurus for its stout femur and small pes; and Sanchiaosaurus for its large size, relatively long mandibular symphysis, squat teeth, constricted coracoid with very small supracoracoid foramen, broad ischia, and robust propodials. The only skull of Shingysaurus is of a mesorostrine, large fenestra format which is similar in general appearance to that of Simosaurus. It likewise has a short mandibular symphysis and rounded rostrum. Its temporal COROSAURUS ALCOVENSIS T3 Fic. 35. Reconstructed skulls of various ‘nothosaurs’ in occipital aspect. A, Szmosaurus after von Huene (1921), Kuhn-Schnyder (1961) and Schultze (1970); B, Corosaurus; C, Nothosaurus after Koken (1893) and Schroeder (1914). Compare with Fig. 8, B. Scale bars = 1.0 cm. fenestrae are slightly more elongate than are those of Szmosaurus. Though coarsely ridged, the teeth of Shingyisaurus are slender. Partanosaurus Skuphos, 1893a emend. 1893b from the Middle ‘Triassic of Vor- arlberg, western Austria, is a final problematic genus which may be tentatively retained. It is unusual in possessing: tall, ridged, neural spines; ovate vertebral centra; distally expanded dorsal ribs; and a very slender scapular blade. The small Microleptosaurus Skuphos, 1983c, introduced along with a full description of Partanosaurus, is from the same unit and general locale as the latter and may represent merely the juvenile form of Partanosaurus. Unfortunately, it is known only from fragmentary material and thus is of little use in an ontogenetic study. Also proposed in the same reference is Kolposaurus Skuphos, 1893c from the Muschelkalk of Upper Silesia. This, as well as numerous other genera described from fragmentary specimens or only a few, isolated bones is here considered nomen dubium, although often equated with Nothosaurus (e.g., Romer 1966). These dubious names are listed with all other ‘nothosaur’ genera and their present status in Appendix B. 74 PEABODY MUSEUM BULLETIN 44 Fic. 36. Pectoral girdles of various ‘nothosaurs’ in ventral and lateral aspects. A, B, Nothosaurus after Romer (1956) and Carroll and Gaskill (1985); C, D, Corosaurus; E, Paranothosaurus after Peyer (1939); F, Keichousaurus after Young (1958); G, H, “Pachypleurosaurus” after Carroll and Gaskill (1985). Compare with Figs. 13 and 14. Scale bars = 1.0 cm. SAUROPTERYGIAN ORIGINS AND RELATIONSHIPS The probable origin of the Sauropterygia from primitive diapsids has been briefly discussed above. Carroll (1981) described as Claudiosaurus a small, aquatic reptile from the Upper Permian of Madagascar (Piveteau 1955) which he considered to be the earliest and most primitive sauropterygian known. The appendicular specializations and lower temporal emargination (presumed arcade loss) of Clau- diosaurus clearly support this interpretation. The propodials of Claudiosaurus, for example, very closely resemble those of ‘nothosaurs’ and, in fact, share certain derived features with the Sauropterygia such as reduced epicondyles and an ectepicondylar foramen which has been transformed to a notch. Its gross cranial anatomy (apart from the missing lower temporal arch), the existence of an in- terpterygoid vacuity, the presence of palatal dentition, and certain other skeletal characteristics appear quite similar to those of younginiform ‘eosuchians’ (Sub- order Younginiformes Romer, 1945) such as Hovasaurus, Tangasaurus, Thadeo- saurus, and Youngina (Carroll 1981; Currie 1981, 1982; Currie and Carroll 1984) but these are plesiomorphic characteristics. They are, however, united by the COROSAURUS ALCOVENSIS WS Fic. 37. Pectoral girdles of various ‘nothosaurs’ in ventral aspect. A, Simosaurus after von Huene (1952); B, Lariosaurus after Mazin (1985); C, Ceresiosaurus after Kuhn-Schnyder (1963b); D, Ser- pianosaurus after Arthaber (1924), Deeke (1886), and Rieppel (1989); E, Dactylosaurus after Nopsca (1928b) and Sues and Carroll (1985); F, Psilotrachelosaurus after Nopsca (1928b). Compare with Fig. 13. Scale bars = 1.0 cm. derived features of present suborbital and posttemporal fenestrae, reduced lach- rymals, single headed dorsal ribs, and a reduced olecranon process. Further examination of Claudiosaurus and comparison with the Younginiformes has re- vealed numerous derived characteristics that clearly distinguish Claudiosaurus from ‘eosuchians’ (Carroll 1981). These traits include the apparent loss of the subtem- poral arch and concomitant reduction of the quadratojugal and jugal [a reasonable proposal contrary to Rieppel (1989)], the reduction of the suborbital fenestra and interpterygoid vacuity, the loss of the transverse flange of the pterygoid, and an unossified sternum. It is commonly felt that both Claudiosaurus and the Saurop- terygia have likely diverged from a basal diapsid (younginiform?) stock as sug- gested by Kuhn-Schnyder (1962, 1963a, 1967, 1980). Following Sues (1987), the Sauropterygia are mostly closely related to the Lepidosauromorpha of Benton (1985). From this point, it has been possible to utilize the Captorhinomorpha, the 76 PEABODY MUSEUM BULLETIN 44 rr er peop eA tis 3-52 e5 HH YP) Sasso Ny e] Fic. 38. Reconstructed ‘nothosaur’ skeletons in dorsal aspect. A, “Pachypleurosaurus” (= Neustico- saurus) modified from Carroll and Gaskill (1985) and Peyer (1944); B, Lariosaurus after Peyer (1933). Compare with Fig. 21. Scale bar = 5.0 cm. (Most specimens of Neusticosaurus smaller.) primitive diapsid Petrolacosaurus (following Rieppel 1989), the Younginiformes, and Claudiosaurus as primitive outgroups for comparison with the traditional Sauropterygia during rigorous character analysis. It is a happy circumstance that Claudiosaurus and the Younginiformes, as potential structural ‘“‘ancestors”’ and sister-groups to the unknown ancestors of the mainstream Sauropterygia, are Permian in age, whereas no undisputed ‘nothosaur’ or plesiosaur is known from before the Triassic (although by definition the lineage must have been present). Most ‘nothosaurs’ are Middle Triassic in age and the group does not appear to enter the Jurassic. Plesiosaurs are primarily Jurassic and Cretaceous animals. Each of the major groups of the Sauropterygia have likewise been examined for presumably derived characteristics and these are listed in Table 2. Appendix C details discussions of each character and its significance. Figure 40 provides a hypothetical cladogram of sauropterygian relationships which was constructed from these characters and which places each derived suite in perspective. The 84 characters of the data matrix were analyzed with the branch and bound algorithm of PAUP (Phylogenetic Analysis Using Parsimony) for the MacIntosh v.3.0 (Swofford 1989). Six equally parsimonious trees of 150 steps were produced at COROSAURUS ALCOVENSIS i) yo Z C 2 Z Wat Y, ——S Sa ese WY SZ >_ aie > MUL le > =SS Fic. 39. Reconstructed ‘nothosaur’ skeletons in ventral aspect. A, Ceresvosaurus after Kuhn-Schnyder (1964) and Peyer (1931, 1944); B, Paranothosaurus after Peyer (1939). Compare with Fig. 21. Scale bars = 10.0 cm. or ESN: gH Fp oh HN os Fic. 40. Cladogram of hypothetical relationships of the Sauropterygia and outgroups. 78 PEABODY MUSEUM BULLETIN 44 TABLE 2. Data matrix of sauropterygian character states for 21 taxa (including outgroups). Each character is fully discussed in Appendix C. 2)n Captorhinomorpha Petrolacosaurus Younginiformes Claudiosaurus Pachypleurosaurus ‘Neusticosaurus Serpianosaurus Anarosaurus Daciylosaurus Keichousaurus Placodontia Simosaurus Corosaurus Ceresiosaurus Lariosaurus Cymatosaurus Nothosaurus "POF Oi NEON EN Oi Oi OO? ©: ©: Oi © Paranothosaurus Pistosaurus Plesiosauroidea Pliosauroidea BO? Ot OOF Oi OF OE Oe NINE NE et tt tt © an) 0 0 “0 0 0 0 0 a 0 0 0 1 ] 1 1 G 1 1 ] 1 1 1 pa pe eS ee ne ee ee ee et OP OFTOSOFO'O!} 0:0: 0: vngonyo tt et ot 3 OO: Oi ©? ©} ©: Ot A Oi O32 OO: O72 ©: Op ON ENE NY Pm me ee ee me ee ee EN NEN OIOIO pee pum Spm Speak Spm | peek Speed pes peed! pee pees pees ped Speed: ped bees peed S eed? CE CO! * pee ee eee Wn rar Wt er (=) Continued on next page a consistency index of 0.607. This is a high index for analysis of 21 taxa. Figure 40 is a strict consensus of these trees. No a prior: character weighting was assumed, for such cannot be empirically supported. Characters chosen were, however, presumed to have phylogenetic significance. Although the data exhibit less struc- ture at 151 steps, a strict consensus tree still differentiates between the major clades of Figure 40. Claudiosaurus was undoubtedly an aquatic reptile with an origin among ter- restrial forebears as discussed by Carroll (1981, 1988). It would appear to follow that ‘nothosaurs,’ placodonts, and plesiosaurs also represent a return to an aquatic environment, in spite of the doubts expressed by Romer (1974). How are these groups related to Claudiosaurus and to each other? They retain (as primitive features) the derived characters exhibited by Claudiosaurus over the Youngini- formes, but of course, also display unique characters of their own (Table 2). Claudiosaurus, while probably not the ancestor of the Sauropterygia, represents a suitable structural analog for the animal that was. From this transitional grade, and following the results of cladistic analysis, can be postulated the origin of two divergent sauropterygian lineages currently recognized as placodonts + ‘notho- saurids’ + plesiosaurs, and the pachypleurosaurs. These monophyletic sister clades may hereafter be referred to as the Nothosauriformes (new taxon) and the Pachypleurosauria, respectively. These groups largely correspond to those produced by Sues (1987), Rieppel (1989), and Tschanz (1989). Tschanz’s (1989) Eusauropterygia is similar to the Nothosauriformes but excludes the placodonts. The pachypleurosaurs are monophyletic whereas the monophyletic plesiosaurs arose from the paraphyletic ‘nothosaurids.’ The taxon Nothosauria is paraphyletic and therefore no longer tenable. I believe that Tschanz’s (1989) use of a restricted Nothosauria is confusing and ill-advised in light of the strong historical conno- tations of the word. COROSAURUS ALCOVENSIS 79 TABLE 2 -- Continued Corosaurus Ceresiosaurus Nothosaurus Paranothosaurus | 1 Pistosaurus Plesiosauroide Pliosauroidea Continued on next page The time of origin of the nothosauriform and pachypleurosaur phyletic lines is likely to have been the latest Permian to earliest Triassic but is poorly rep- resented by contemporary sauropterygian fossil material. The presence of Claud- losaurus, the relatively plesiomorphic sister taxon to the Pachypleurosauria, in the Upper Permian suggests a Late Permian minimum time of divergence for these two groups. Thus, by definition, the sauropterygian lineage must also have been present by at least the Upper Permian. Once thought to have been ancestral to plesiosaurs (e.g., Seeley 1882; Tarlo 1967), ‘nothosaurs’ in general were later often excluded from this role on the basis of one discrete character, namely their lack of interpterygoid fenestrae (Romer 1966). Plesiosaurs retain the primitively “open” format of ‘eosuchians’ in which the basicranium is largely exposed between the pterygoids. The functional significance, if any, of an open versus a closed palate is not yet understood and the apparent character (evolutionary) reversal from a closed to open palate has not been evaluated. However, the most parsi- monious explanation in this case, considering the numerous nothosauriform syn- apomorphies of Table 2, should be adopted: the ancestral plesiosaur stock probably arose from within the traditional ‘nothosaurids’ and they coincidentally with the Placodontia (Fig. 40). In the present analysis, the placodonts as currently known are confirmed as sauropterygians, for they share the derived features of: a single upper temporal opening; no interpterygoid vacuity (plesiosaurs seemingly reverse this character); loss of supratemporal, postparietal, tabular, and lachrymal; retracted nares; prom- inent retroarticular process; loss of trunk intercentra; minimum of three sacral vertebrae; no sternum; divided scapulocoracoid; scapula superficial to the clavicle; a straight clavicular bar with a pronounced anterolateral corner (reversal in plesiosaurs); and pectoral and thyroid fenestration. Perhaps surprisingly, they sort out in the analysis as nothosauriforms in sharing: large size; a large supra- temporal fenestra; a posterolateral process to the frontal; an elongate jugal that extends caudad from the orbit; a stout mandibular symphysis; platycoelous ver- 80 PEABODY MUSEUM BULLETIN 44 TABLE 2 -- Continued Captorhinomorpha HES IES o Neusticosaurus Serpianosaurus rinie oo ow ClO! CO! OC! C!C!O: ° oicio Flacodontia _ Simosaurus Siolcloicio oo oye ead ped pede ps ek ee OO? Ol OL mM CIOlO VY C!OlO! Vio Hey — Pistosaurus Plesiosauroidea Pliosauroidea VO Sie o:co HEE Eee VEE NEEM INE NEN EN ENE Rt ROO CLO OC! RIMRIO RPI VIR COIO: OO: OO; O:O:O:0;0:0 ee o SS) —_ Continued on next page tebrae; and a strongly curved humerus (reversal in plesiosaurs). The obvious diagnostic features of the placodonts such as a stout coronoid process, palatines that usually meet to separate the pterygoids, crushing palatal and marginal teeth, hyposphene/hypantrum articulations, and occasional dermal armor are all syn- apomorphies for the group that under the present systematic philosophy can potentially be ascribed to evolution following the placodont/‘nothosaurid’ + ples- iosaur divergence. Similarly, the large jugal and quadratojugal may have been reelaborated as an evolutionary reversal in response to function. The Placodontia are presumably monophyletic; however, this group requires substantially more anatomical elucidation. No clear synapomorphies can as yet resolve a basal trichotomy between the Placodontia and the remaining nothosauriforms. The loss of the quadratojugal is significant beyond the Szmosaurus node, but if Simosaurus in reality also lacks a quadratojugal, traditional ‘nothosaurids’ + plesiosaurs might be resolved in the future. Unfortunately, this character is presently equivocal in Corosaurus (although likely to be derived). Nothosauriformes minus the Placodontia also have no quad- rate notch, usually reduced nasals and prefrontals, and largely platycoelous ver- tebral centra. Reversals and convergences in the various lineages are enumerated in Appendix C. The possible functional changes, in light of nothosauriform anatomy, that might have allowed the evolution of plesiosaurs from an animal structurally akin to ‘nothosaurids,’ have been discussed in Chapter 3. Pistosaurus, a Middle Triassic (Upper Muschelkalk) contemporary of the ‘nothosaurids,’ had an open palate, but the postcrania sometimes assigned to this genus are rather primitive in several respects (Carroll and Gaskill 1985; E. von Huene 1949; F. von Huene 1948c; von Meyer, 1847-55; Sanz 1983b; Sues 1987). The body appears to have been relatively long and narrow; accessory vertebral articulations were present; the hu- merus and femur were slender; the epipodials were long; and the ilium was in contact with the pubis in Pistosaurus as in primitive nothosauriforms. Conversely, Pistosaurus had ventral nutritive foramina in the vertebral centra, long transverse COROSAURUS ALCOVENSIS 81 TABLE 2 -- Continued 30 : f shape : Captorhinomorpha 0 0 oe ay ie aon Sete or - an : pa Vans ne ae oe aie ore Claudiosaurus 1 ror or Ho 1 0 Bache leurosauvus eons aan a oni j . ie ss ae ae ai ; i 2? 0 | 0 1 0 1 1 0 (0) 0 0 1 1 ee agg: aa aaa ea no a a Ceresiosaurus ie 0 Oe ez 0 0 1 es De Gg es aaa RUE Ea HE RS aR ea ge — a Cymatosaurus es 0 0 0 Om mil 1 eather : Ses EE ae — ae aera a Paranothosaurus ihc eae lia ie 1 0 ie il aap uciasitnmm aaa) Roa Sige ate seni Be oa aa Plesiosauroidea oy, Ae 0. 0 0 0 4 ] See aaa mae a Boh eee — oe i : ae oe Continued on next page processes, tall neural spines, elongate coracoids, no entepicondylar foramen, and relatively broad epipodials which are synapomorphic for plesiosaurs. Vertebral nutritive foramina (foramina subcentralia) are a uniquely derived character shared by virtually all plesiosaurs (excepting the very unusual Brachauchenius of the North American Cretaceous), and are unknown in pachypleurosaurs, placodonts and ‘nothosaurid’ grade nothosauriforms. While the skull of Pistosaurus retains the nasals, unlike that of advanced plesiosaurs, the nasals no longer contact the borders of the external nares as they do in more plesiomorphic nothosauriforms. On the basis of these characters as well as the open palate, Pistosaurus must be considered a primitive plesiosaur (Fig. 40). Several problematic Permo- Triassic sauropterygian specimens may also occupy a place in the plesiosaur lineage, but these are usually too fragmentary to be of any real taxonomic value. Others may be the bones of ‘nothosaurids,’ pachypleur- osaurs, or even of representatives of the pre-nothosauriform/pachypleurosaur grade. Von Huene (1929) described two amphicoelous dorsal vertebrae and a dorsal rib from the German Upper Keuper (Late Triassic), probably correctly, as those of a plesiosaur. In overall appearance, these remains differ little from those of undoubted Liassic plesiosaurs. Similar vertebral centra with obvious ventral nutrient foramina, apparently from primitive plesiosaurs, are occasionally contained in collections of assorted Triassic material. Two such specimens are in the British Museum (Natural History) and are associated with isolated centra of ‘““Nothosaurus” (BMNH 1103 and 8201). These vertebrae are from the Bavarian Muschelkalk. A third is figured by Sanz (1983b), again as ‘““Nothosaurus,” while a similar isolated sauropterygian centrum from the Ladinian of the Lena Basin of the Soviet Union has been assigned to “‘Nothosaurus(?)” (Lazurkin and Ochev 1968). The genus Elmosaurus v. Huene, 1957 from the Upper Muschelkalk has been commonly referred to the ‘Nothosauria’ (Carroll 1988; Carroll and Gaskill 1985; v. Huene 1957) but is enigmatic. Known only from a single, intermediately-sized, fragmentary skull, E/mosaurus displays a cranial morphology profoundly different 82 PEABODY MUSEUM BULLETIN 44 TABLE 2 -- Continued 35 en 36 : n/en 40 dentition: t sha rostrum nasal diastema: dent or Captorhinomorpha Petrolacosaurus Younginiformes ‘Claudiosaurus i=) Neusticosaurus Serpianosaurus Anarosaurus Dactylosau rus Keichousaurus Placodontia Simosaurus ee ee ee ee a) C10: 0:0: 0:0:0:03:0:0 i=) =" tS Ww Nothosaurus Paranothosaurus | Pistosaurus Plesiosauroidea ~ Pliosauroidea lol Rim pinicicicioiniclcloicicicicicicio Hinlicicicioiciclcicliciclcicicioicicicioio EMO ClOiO: vO y 0 0 0 0 0 0 0 0 1 0 1 0 1 0 af 0 ames 0 a 0 il 0 1 0 1 0 1 1 4 0 1 0 1 0 1 0 1 1 1 ? 4 ? pe ree eee are ere ar a a er) a" SPIiDiDIOIC;O! CLO OI OC} HID CIC O! CO: CHO! CI C!IO CF0:0:0:0:0;0:0;0:— 2: ah i=) A Continued on next page from those of all other known sauropterygians. The skull is of apparent euryapsid configuration, but the supratemporal fenestra is bounded on three sides by the huge squamosal, excluding both the postorbital and the postfrontal from the temporal opening. Unlike the case in other sauropterygians, an extremely large lachrymal is present. The nature of the posterior palate is unknown and without more material the animal must be considered ?Sauropterygia incertae sedis. F. Von Huene (1944) has described an isolated left humerus of a primitive sauropterygian from the Lower Muschelkalk. This curved element is very rem- iniscent of ‘nothosaurs’ but lacks both epicondylar foramina and thus appears to be part of the plesiosaur radiation. Its plesiomorphic curvature may be the result of an ontogenetic or paedomorphic effect. Von Huene (1951) has also described a possible sauropterygian epipodial from a stromatolitic unit of the Lower ‘Triassic (Scythian) Lower Buntsandstein of Germany. This bone is difficult to interpret but has the appearance of a ‘nothosaur’ tibia. Von Huene (1951) identified the bone as that of a pachypleurosaur, but as its characteristics are primitive, it is also incertae sedis. Its stratigraphic position is, nevertheless, interesting. Lastly, Nothosauravus Kuhn, 1958a was named for a single, small, amphicoelous sacral vertebra (Kuhn 1939) from the lower Upper Permian Kupferschiefer. This prob- lematic bone has a ‘nothosaurian’ appearance and may be that of a primitive sauropterygian. The specimen is, however, generically nondiagnostic. The environmental condition(s) and evolutionary mechanism(s) that might have led to the origin of the Sauropterygia, and then more specifically to the respective origins of the pachypleurosaurs and nothosauriforms, including plesiosaurs and placodonts, are unknown. The limited sample sizes and temporal ranges of the animals involved also preclude any knowledge of the tempo(s) and mode(s) of their evolution. Examination of the ontogenetic series provided by Currie (1981) for the tangasaurid Hovasaurus, by Currie and Carroll (1984) for Thadeosaurus, by Carroll (1981) for Claudiosaurus, and by numerous, isolated, juvenile remains of ‘nothosaurs’ and plesiosaurs (e.g., Andrews 1910) suggests, however, that the various forms of sauropterygians might have been derived in part through a process of paedomorphic (heterochronous) development. The juvenile humeri of COROSAURUS ALCOVENSIS 83 TABLE 2 -- Continued 44 Captorhinomorpha Petrolacosaurus | Younginiformes | Claudiosaurus | Pachypleurosaurus | Neusticosaurus | Serpianosaurus es een) ‘Dactylosaurus Keichousaurus Placodontia Simosaurus Corosaurus Ceresiosaurus Lariosaurus Cymatosaurus — Nothosaurus Paranothosaurus pee eee Plesiosauroidea ‘Pliosauroidea LOI O I FRIHIO RI VICIOIOIO OO! O;O! HO! vii Rin Biel HiniBininivinio olololclojcioioio 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 ? 1 i ? 1 1 Rinieieieiwinie sie ee eee ee OOOO Segiticeiistiseseee eee ss ae SVotiovin) Be ecb aoe cee eee Ss HERO ON OLOC!IO'O!O C1O! SC! C!O! C!C!C!O]N oO: a Continued on next page each taxon, for example, are remarkably similar in form, becoming less similar only in later ontogenetic stages (Baer’s law). The humeri of pachypleurosaurs and primitive nothosauriforms, especially, are very much like those of the juveniles of Hovasaurus and Thadeosaurus and in particular lack the prominent condyles of the adult ‘eosuchians.’ Similarly, juvenile pectra of the Jurassic plesiosaur Cryptocleidus resemble the generalized ‘nothosaur’ pattern with a large pectoral fenestra and relatively reduced coracoids (Carroll and Gaskill 1985). Mosaic, variable rates of embryologic and juvenile development, the retention of juvenile traits, and hypertrophic elaboration of characters during ontogeny are probably major sources of morphologic variability and macroevolutionary change in all phyla. The reduced and latent ossification of marine reptile skeletons may increase the susceptibility of these animals to such change. In the Sauropterygia, specifi- cally, morphologic diversity is largely the result of proportional variation. Carroll and Gaskill (1985) have discussed the possible evolutionary reduction and reelaboration of the sauropterygian pectrum in relation to the reversed en- dochondral and dermal pectoral elements of this group. The scapulae and inter- clavicle are superficial to the clavicles in the Nothosauriformes and Pachypleu- rosauria but in few other groups. According to Rieppel (1989) this may also be true for the Placodontia. Such a change could conceivably only occur at an early stage of ontogenetic development prior to the ossification of the pectrum. Additional examples of presumed proportional and structural change during ontogeny can be hypothesized but, because of the likelihood of allopatric speciation and possible evolutionary punctuations (for whatever reason), few transitional fossils are to be expected. I cannot accept Schmidt’s (1987) interpretation of the dermal elements of the sauropterygian pectrum as new endochondral ossifications. ‘NOTHOSAUR’ RELATIONSHIPS By using the discussions presented above, the relationships of the two ‘nothosaur’ groups—one monophyletic, the other paraphyletic—and the specific position of Corosaurus can be examined. It seems clear that the pachypleurosaurs are mono- 84 PEABODY MUSEUM BULLETIN 44 TABLE 2 -- Continued 56 : sternum astralia: armour Captorhinomorpha | Petrolacosaurus Younginiformes Claudiosaurus Pachypleu rosaurus Neusticosaurus Serpianosaurus HiOiRieiCorOio:io io) _ Oo CC ON HH COC Oi S Placodontia ; Simosaurus Corosaurus Ceresiosaurus Lariosaurus Cymatosaurus Sei Sr seaioy Pea Paranothosaurus Pistosaurus Plesiosauroidea Pliosauroidea nF nb en ent ea CE nd bE ek Ee ee EO DIDID:ICIO: VID DIC! O] RID: SIDS! CO! C!O!C!O CID DIDO VI HH ClO Vin RRR BiB OO O:O DIO VINVINIDL COLO OC! VIO! VICI O!O RR Cio! yi: ClO Clo!lO Vidid ClO Lidiclol cio: cicicic!o NENENIENINE NE NENENIENINENININININIENE Oise CO! HiHiB Continued on next page phyletic and distinct from the remaining sauropterygians. Rieppel (1989) contends that the loss of the ectopterygoid characterizes the pachypleurosaurs. This char- acter is, however, difficult to evaluate in most nothosauriform specimens. A ventral quadrate projection for suspension of the tympanum may also be diagnostic (Rieppel 1989). In spite of Rieppel’s (1989) objections, there may be a divergent trend for supratemporal fenestra reduction in the pachypleurosaurs and a general ‘“nachyostotic” thickening of the ribs accompanied by narrowed distal ends of the sacral ribs may also be significant. As in the works of Rieppel (1989), Tschanz (1989), Sues (1987), and Sues and Carroll (1985), the present study suggests that the Pachypleurosauria are the relatively plesiomorphic sister group to the remaining sauropterygians; they certainly exhibit fewer derived characters than do the nothosauriform ‘nothosaurs.’ These differences are discussed in the above noted works and can be found also in Table 2. Under the present hypothesis of relationships the known members of the Pachy- pleurosauria are Anarosaurus, Dactylosaurus, Neusticosaurus, ““Pachypleurosaurus” (probably a specific variant of Neusticosaurus), Keichousaurus, and Serpianosaurus. These genera share the basic pachypleurosaur suite of derived characteristics as discussed above and in Appendix C. Psilotrachelosaurus, as a poorly known taxon, has not been included in the analysis but, if distinct, is most likely a pachypleu- rosaur. The interrelationships of the Pachypleurosauria are known with far less certainty than is its probable composition. Enough anatomical knowledge is avail- able for construction of the clade, but very little for clarification of its internal genealogy. However, with the few additional derived characters gleaned from individual pachypleurosaur anatomies (Table 2), a preliminary, hypothetical cladogram of part of the group is suggested (Fig. 40). Certainly Neusticosaurus and “‘Pachypleurosaurus” (sensu Carroll and Gaskill 1985) form a clade because of their many synapomorphies (Table 2), while Anarosaurus, Keichousaurus, and Dactylosaurus apparently also form a distinct clade. The present analysis suggests a basal trichotomy between these two pachypleurosaur groups and Serpianosaurus. If the impedance matching middle ear of Rieppel (1989) and Sander (1989), and COROSAURUS ALCOVENSIS 85 TABLE 2 -- Continued cleitrum | _ icl : icl size iclcorner * sc/cor Captorhinomorpha Petrolacosaurus | O Younginiformes on ae Pachypleurosaurus a eee ae Serpianosaurus rere aaa Dactylosaurus Keichousaurus Placodontia Simosaurus Corosaurus Ceresiosaurus “Lariosaurus Cymatosaurus “Nothosaurus Paranothosaurus PE mnt Plesiosauroidea Pliosauroidea : scapula SS ee OC:0!O:C! OC: (=) i RPiOivieie o:O ClOPvinix mie: EVE EEO miei vie ei ei Ri OrOiO =) a vm me ee ee ee OT OO mee Re ee ee ee ee ee ROTO: O:O BK HCO ON COCO ONC OOOO ON NN: mS SOLON OO ROOFS SiO: O:0:0:0:;0:0 [oer ee en ee ee ee | oh) Continued on next page the bone ornamentation of Sander (1989) are accepted as synapomorphies for Neusticosaurus, “Pachypleurosaurus,”’ and Serpianosaurus, the trichotomy is re- solved with Serpianosaurus forming the plesiomorphic sister of the neusticosaurs (Fig. 41). Similarly, for the nothosauriform ‘nothosaurs’ Figure 40 presents a hypothetical cladistic hierarchy that includes Ceresiosaurus, Corosaurus, Cymatosaurus, Lario- saurus, Nothosaurus, Paranothosaurus and Simosaurus. Sanchiaosaurus may be pro- visionally placed in this group but is, however, too poorly known for inclusion in the cladogram. The characters forming the basis of relationship are again listed in Table 2. Nothosaurus and Paranothosaurus form a closely related unit, as expected. Gq o C) 5 ow o> Pe AP LPR EG KG » 5 gi SO SO ORE ee RE SG Go) 62: \ 2 4 \S 4 ) 2 ~~ ae Rs RS ee ye & ee x x Paes & s ue Ss au oo ~ aN RS & : ~\ AWE Te \) ND “ Sr oO > SO SF gh ye nw SoS CC Ome Or ar en gee a NS Fic. 41. Revised cladogram of hypothetical sauropterygian and outgroup relationships indicating likely resolutions of analysis trichotomies as discussed in text. 86 PEABODY MUSEUM BULLETIN 44 TABLE 2 -- Continued Captorhinomorpha Petrolacosaurus Younginiformes Claudiosaurus Pachypleurosaurus Neusticosaurus Serpianosaurus Anarosaurus Dactyl miOO!C!O me he ie oidioi~ Placodontia Simosaurus = oo Ol Ri Mivinimieic:c:oio Oi RID DIO DID! CIC! OC} CF O}O:O SiSiDi CG: S}icjo o:o i cee eee er a Cymatosaurus “Nothosaurus | Paranothosaurus | sagen ST a a Plesiosauroidea Ole 22 eo pl, wh eee Pliosauroidea 0,1 — iSiOiviS mivicic: vidio CLO OC! VO ViDiO: CIOIG;C!O A ye he SSS S) NINiPiBioie Hicidlolvio — VOLO Ni DIDI DIDO VICI HCHO CO! CO CID! CO! CO! O;O;C!O = nN _ ° _ _ = Continued on next page A clade united primarily by distally unexpanded sacral ribs, ““pachyostosis,” and no interclavicular posterior process contains Ceresiosaurus and Lariosaurus as the sister group to Nothosaurus + Paranothosaurus. Small size was apparently developed in Lariosaurus independently of the pachypleurosaurs. ‘The two groups form a larger monophyletic clade which is the sister of C'ymatosaurus. These groups plus Cymatosaurus exhibit the greatest number of uniquely derived traits of the known ‘nothosaurids,’ variously including increased sacral vertebral num- ber, elongate temporal region and fenestrae, elongate skull and rostrum, fused frontals and parietals, maxillary caniniform teeth, the loss of a nasal/prefrontal contact, the presence of a rostral constriction, and a subrectangular postfrontal. The relative position of Cymatosaurus is based, of course, almost entirely upon cranial material. Simosaurus seems to represent a separate lineage with an earlier origin and displays its own unique suite of derivations. The unusual autapomorphies of Simosaurus, such as its brevirostrine format and short, striated teeth cannot at this time support any hypothesis of relationship, although if shared by the poorly known Shingyisaurus, may indicate kinship. Simosaurus is apparently linked to the above nothosauriforms through the shared absence of a pterygoid flange, elongate supratemporal fenestrae, a prefrontal which is normally significantly smaller then the postfrontal, and an increased number of sacral vertebrae. The sister to all of these ‘nothosaurids’ is the monophyletic Plesiosauria (con- taining Pistosaurus + typical Jurassic and Cretaceous forms). Future work is required to determine the phylogenetic validity of the traditional plesiosaur (sensu stricto) and pliosaur lineages. Corosaurus, with its several derived postcranial features, nevertheless appears to have had the least recent common descent of all ‘nothosaurids.’ It has retained the presumably primitive characteristics of only three sacral vertebrae, a relatively unlengthened skull with a short temporal region, round supratemporal fenestrae, noncaniniform premaxillary teeth, amphicoelous vertebral centra, elongate femur, and similar traits. It would seem that only in a latter stage in the history of this COROSAURUS ALCOVENSIS 87 TABLE 2 -- Continued foo) fo) (SS) Captorhinomorpha _Petrolacosaurus Younginiformes Claudiosaurus Pachypleurosaurus Neusticosaurus | Serpianosaurus Anarosaurus |. Dactylosaurus OF 037 O72 O2 O21 O20O102 0 Placodontia Simosaurus Corosaurus Ceresiosaurus Lariosaurus — Cymatosaurus Nothosaurus — Paranothosaurus SSR eRe RE Plesiosauroidea Pliosauroidea — minivi Bini aininioloioliocioicioioiocio =) a —_ 1 NO Ree SS meet FOIPOP VIO OO: Of ) =) NENENININENENININE NIN eS Sieve Sie Oy RiP: OOO: vidi O:O: 0:0: 0:0:0:0:0:0:0:0:0 | pO eal pO ean’ at’ el eh ee Ol el Nix Continued on next page lineage did Corosaurus acquire its confusingly derived features or autapomorphies, particularly the expanded coracoid and pubis through which it is slightly remi- niscent of plesiosaurs. We have already seen how such similarities might arise independently in parallel lineages. The expanded coracoid of Corosaurus is, in fact, only superficially like those of plesiosaurs and lacks the exceptional posterior development of the latter. The unresolved phylogenetic position of Corosaurus with regard to the placodonts has been discussed above and hinges on the equivocal presence or absence of a quadratojugal in Corosaurus and possibly Simosaurus. Assuming such a loss, and incorporating the resolutions of the above discussed trichotomies, a reasonable cladogram of sauropterygian relationships is presented in Figure 41. The preliminary phylogenetic hypotheses presented here are, of course, falsi- fiable and likely to be altered by the acquisition of future data. Several additional nothosaurian-grade sauropterygians are so poorly known that they cannot yet be satisfactorily included in any classification scheme. These forms include Chin- chenia, Kwangsisaurus, Metanothosaurus, Partanosaurus, Proneusticosaurus, and Rhaeticonia. All save Rhaeticonia are large forms possibly having some relationship to the Nothosauriformes but lacking conclusive cranial material. Rhaeticonia could be either an unusual pachypleurosaur with a constricted rostrum, a small ‘notho- saurid’ like Lariosaurus, or even the juvenile of some known form (e.g., Cyma- tosaurus?), but neither is its temporal configuration known. Each of these taxa is presently considered Sauropterygia incertae sedis. HIERARCHICAL CLASSIFICATION DIAPSIDA Osborn, 1903 NEODIAPSIDA Benton, 1985 LEPIDOSAUROMORPHA Benton, 1985 SAUROPTERYGIA Owen, 1860 Diagnosis. Small to large lacertiliform aquatic reptiles with derived diapsid (“euryapsid”’) cranial configuration. Deep lateral temporal emargination (reversal 88 PEABODY MUSEUM BULLETIN 44 TABLE 2 -- Continued car/tar : hindlimb: ulna Captorhinomorpha Petrolacosaurus Younginiformes Claudiosaurus Neusticosaurus Serpianosaurus Anarosaurus Dactylosaurus Keichousaurus Placodontia Simosaurus Corosaurus Ceresiosaurus ' Lariosaurus Cymatosaurus ‘Nothosaurus Paranothosaurus Pistosaurus Plesiosauroidea Pliosauroidea ClO vine vine ei eee eS ee ei ei ei OOOO i=) Sn me ee OU Ee ees a SR OPO Orv BS Oro Oo: Oro Oo O:9O O:O:0:0 Ol micicloi vicoinioimivioimio:s S — in Placodontia and Plesiosauria); no lower temporal fenestra or arcade. Jugal reduced; quadratojugal reduced or absent (reversal in Placodontia); nasals reduced or absent, nares retracted; supratemporal, tabular, and postparietal absent; lach- rymal reduced or absent; no interpterygoid vacuity (reversal in Plesiosauria); prominent retroarticular process; cervical region elongate (reversal in Placo- dontia); trunk intercentra absent; three or more sacral vertebrae; sternum absent; scapulocoracoid divided; scapula and interclavicle superficial (ventral) to clavicle; straight clavicular bar with pronounced anterolateral corner (reversal in Plesio- saurla); prominent pectoral and thyroid fenestration; ectepicondylar foramen re- duced to notch or lost. Range. ?Upper Permian—Upper Cretaceous (Maastrichtian). PACHYPLEUROSAURIA Sanz, 1980 Diagnosis. Plesiomorphic sauropterygians with small supratemporal fenestrae (much smaller than orbits); ectopterygoid perhaps lost; quadrate hooked with pronounced otic notch; general “‘pachyostotic” thickening of bones (variably pres- ent in Dactylosaurus); sacral ribs reduced in diameter distally; slight hyperphalan- Sy- Range. Middle ‘Triassic (lower Anisian—upper Ladinian). NOTHOSAURIFORMES, new taxon Diagnosis. Large sauropterygians (reversal in Lariosaurus) with large supratem- poral fenestrae (larger than orbits); frontal with prominent posterolateral process; quadratojugal lost (independently?) in most lineages; no quadrate notch (except in Placodontia); stout mandibular symphysis; normally anisodont dentition with procumbent or caniniform teeth or both; vertebrae tending towards platycoely; humerus strongly curved (reversal in Plesiosauria); modified (flattened) epipo- dials. Range. Lower Triassic (Scythian)—Upper Cretaceous (Maastrichtian). COROSAURUS ALCOVENSIS 89 PLACODONTIA Zittel, 1887-90 Diagnosis. Broad-bodied nothosauriforms with (secondarily?) short cervical re- gion; quadratojugal contacts jugal; no temporal emargination (reelaboration ?); pterygoids separated by palatines; stout coronoid process; crushing palatal and marginal dentition with diastema; hyposphene/hypantrum accessory articula- tions; dermal armor common. Range. ‘Triassic (upper Scythian—upper Rhaetian). PLESIOSAURIA de Blainville, 1835 Diagnosis. Highly transformed nothosauriforms with stout thoracic and elongate cervical regions; interpterygoid vacuity (reversal); nasals reduced or, more com- monly, lost; zygosphene/zygantrum articulations absent; high neural spines; zyg- apophyses narrower than centrum; foramina subcentralia; clavicular arch reduced; occasional pectoral and pelvic longitudinal “bars”; posterior ramus of iliac blade and iliopubic contact lost; no obturator foramen, anterior border of pubis ex- panded; propodials massive and largely straight; entepicondylar and ectepicon- dylar foramina lost; reduced or lost spatium interosseum; epipodials extremely short and flat; no midlimb joint; extreme hyperphalangy. Range. Middle Triassic (upper Anisian)—Upper Cretaceous (Maastrichtian). 90 PEABODY MUSEUM BULLETIN 44 5. STRATIGRAPHY INTRODUCTION The Alcova Limestone of central Wyoming is a unique carbonate unit in an otherwise uninterrupted stratigraphic sequence of Triassic red bed deposits. Be- cause of its striking departure from the lithologies of over- and under-lying units, the Alcova is readily identifiable in the field and has long had formal stratigraphic status. It has also enjoyed protracted importance as a marker unit and datum in outcrop and subsurface stratigraphic and structural studies. However, the age, regional correlation, and paleoenvironmental interpretation of the Alcova Lime- stone have proven difficult to resolve, largely because of its unusual character and position, its limited geographic extent, the stratigraphic and structural complexity of the Triassic System in Wyoming, and the rarity of fossils within, above, and below the Alcova. An excellent overview of these difficulties and of the Wyoming Triassic in general is provided in McKee et al. (1959). The Alcova Limestone was originally defined by Lee (1927) as a member of the Chugwater Formation of Darton (1904) on the basis of outcrops near Alcova, Natrona County, Wyoming, at the southeastern edge of the Wind River Basin. Since that time, the Alcova has been widely noted and discussed (e.g., Burk 1953; High and Picard 1967a, 1969; Hubbell 1956; Kummel 1954; Love 1948, 1957; Picard 1967, 1978; Picard et al. 1969; Pipiringos 1953; ‘Tohill and Picard 1966; etc.), but rarely studied in detail. The Alcova has generally maintained its member status, although Branson and Branson (1941) and Pipiringos (1968) elevated the unit to formational rank within the Chugwater Group. While the Alcova Lime- stone is easily recognized in the field, its slight thickness and local discontinuity make it generally unmappable at a scale of 1:25,000, a definitional requirement of a formation, and member status should be retained for this stratigraphic unit. Its genetic history, discussed below, is also relevant to its member rank. The nomenclatural histories of the surrounding rocks are somewhat more complex. Love (1939) was first to divide the red beds of the Chugwater into subunits, including in part, the Red Peak, Crow Mountain, and Popo Agie members. There was no mention of the Alcova in Love’s field area where it was TABLE 3. Nomenclature and principal rock types of the Chugwater Group. Oldest units at bottom (modified from Picard 1978). STRATIGRAPHIC UNIT PRINCIPAL ROCK TYPES Popo Agie Formation arkosic silts, carbonates Crow Mountain Formation “upper sandstone/siltstone unit” arkosic sandstone, siltstone "basal sandstone unit” arkosic sandstone Alcova Limestone Member carbonates "variegated sandy facies” arkosic sandstone Red Peak Formation arkosic clay, silts, sands COROSAURUS ALCOVENSIS 91 —— IDAHO ——2 -@———————_- WYO MING LANDER CASPER i] i] JURASSIC Deadman is. Ankareh Fm. Crow Mtn./Popo Agile fms. =. mbe ~ othy © SS ie a: Alcova Is. mbr. P<) Red Peak Fm. aes v e ae \s E ot* * > ae Red bed facies : PALEOZOIC ee © \s- c sc v\ f = ve