Colbert, 1946

SEBECUS, REPRESENTATIVE OF APECULIAR SUBORDER OF FOSSILCROCODILIA FROM PATAGONIA

EDWIN- HARRIS- COLBERT

BULLE'TINOF THE~

AMERICAN'MUS'EUM OF-NATURAL -HISTORY

VOLUME 8-7:. ARTICLE -4 NEW YORK: 194-6

06-

---

S---

-

-

-------------E-----iiiSX

|||i---

E|g@-

|-sww|||||E

||asn

i||M.

I|

||I

SEBECUS, REPRESENTATIVE OF A PECULIAR SUBORDEROF FOSSIL CROCODILIA FROM PATAGONIA

SEBECUS, REPRESENTATIVE OF A PECULIARSUBORDER OF FOSSIL CROCODILIA

FROM PATAGONIA

EDWIN HARRIS COLBERTCurator of Fossil Reptiles, Amphibians, and Fishes

PUBLICATIONS OF THE SCARRITT EXPEDITIONS

NUMBER 35

BULLETINOF THE

AMERICAN MUSEUM OF NATURAL HISTORY

VOLUME 87: ARTICLE 4 NEW YORK: 1946

BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY

Volume 87, article 4, pages 217-270, text figures 1-30,plates 11-16, tables 1, 2

Issued November 22, 1946

CONTENTS

INTRODUCTION.'TXrV nr)llTTOlD>DV'Tl-E. ^Dcthtwr^nv,I £1ti VJL; UJ~I1J. DI.N Ir OeJUC

TAXONOMY.

MORPHOLOGY.

Material .

The Skull .General RemarksBones of the Skull

The Mandible ....Bones of the Mandibl

The Dentition ....

The Postcranial Skeleto:VertebraLimb Bones . . .

The Brain ......The Eustachian TubesMyology of the Head of

General RemarksThe Jaw MusclesThe Neck Muscles .

SKULL PROPORTIONS OF St

General Remarks . .

Table 1. Comparative T

Table 2. Comparative I

Sebecus AND OTHER FossI]THE PHYLOGENETIC POSITSebecus AND THE MOOTED

,bcuswaeoND uOF ..OTHER.CRooDILIANS

* . . . . . . . . . . . . . . . . . . . .

* . . . . . . . . . . . . . . . . . . .*

*. .

*. . . ...

le a re nt . . . . . .M ll me *.

, . . . . . . . . . . . . . . . . . . . .

* . . . . . . . . . . . . . . . . . . .*

* . . . . . . . . . . . . . . . . . . . .

f Sebecus .....CASAAYO DINOSA . . .R . . . . . . .

* . . . . . . . . . . . . . . . . . . . .

* . . . . . . . . . . . . . . . . . . . .

ebecus AND OIF OTHER CROCODILIANS. . . . . . . . . . . . . . . . . . . . .

Weasurements (in Millimeters) .. .. .. .

RZatios and Indices . . ...........

L CROCODILES FROM PATAGONIA . . . . ...

7ION OF Sebecus..............CASAMAYOR DINOSAURS . .. .. .. .. .

223

224

.. . . . . 225

.. . . . . 227

227

.. . . . . 228228228

242242

246

248248249

249

.. . . . . 251

.. . . . . 251251253

.. . . . . 256

.. . . . . 259

.. . . . . 259

.. . . . . 260

.. . . . . 261

.. . . . . 262

.. . . . . 263

.. . . . . 267BIBLIOGRAPHY.

221

269

INTRODUCTION

IN 1937 DR. GEORGE GAYLORD SIMPSONdescribed very briefly a remarkable crocodil-ian from the Eocene of Patagonia, which henamed Sebecus icaeorhinus. The originaldescription by Simpson, which was publishedwithout illustrations, was based upon a dis-articulated skull and mandible that had beenfound by himself and Mr. Coleman S. Wil-liams in Patagonia during the course of theFirst Scarritt Expedition of the AmericanMuseum of Natural History. The specimen,together with some supplementary materialsfrom the same horizon, constitutes one of theprincipal discoveries of the Scarritt Expedi-tions, made possible by the generosity of Mr.Horace S. Scarritt.The bones of the skull and mandible were

prepared by Mr. Albert Thomson, formerlychief preparator in the Paleontological Labo-ratory at the American Museum of NaturalHistory. From the disarticulated bones, Mr.Thomson reconstructed the skull and mandi-ble; this reconstruction was cast in plaster byMr. Otto Falkenbach, also of the AmericanMuseum Paleontological Laboratory, andthe plaster cast was then tinted, to show theareas represented by fossil bone as well asthose reconstructed. A dorsal view of thiscast was published by Brown and Schlaikjerin 1940 (pl. 4) and to date this is the onlyillustration of Sebecus in the literature.

It had been Dr. Simpson's intention toprepare a detailed monograph of this mostimportant form, and indeed some preliminarypages of this monograph were written. Owingto the pressure of other matters, however, hewas unable to proceed with the study, andconsequently at the time of his entry into theArmy of the United States he turned over hisnotes to me, with the request that I preparethe manuscript for publication. This I havedone, but not without considerable delay,occasioned by an increase of curatorial dutiesand prior interests in certain other researchproblems. Recently, however, it has beenpossible to take up the study and to carry itthrough to completion.Mr. John C. Germann of the Department

of Geology and Paleontology at the AmericanMuseum of Natural History had prepared aseries of drawings under Dr. Simpson's direc-tion to illustrate the separate bones of theskull and jaw in Sebecus. These, together withadditional drawings by Mr. Germann, con-stitute the figures published in this paper.

I wish at this place to express my great ap-preciation to Dr. Simpson for his generosityin turning over to me this study and for hiskind permission to make free use of his notes,illustrations, and preliminary pages of manu-script in the preparation of this paper.

223

THE OCCURRENCE OF SEBECUS ICAEORHINUS

DR. SIMPSON HAS PREPARED the followingstatement with regard to the occurrence ofthe materials under consideration in thepresent paper:

"All the known specimens are from theCasamayor formation, probably of LowerEocene age, of the Territory of Chubut(central Patagonia) in the Argentine Repub-lic. The principal specimen, A.M.N.H. No.3160, was found in Cafiadon Hondo, whichis a roughly circular drainage basin, withoutpermanent water, tributary to the southeastor right side of the Rio Chico del Chubutimmediately above the crossing usuallycalled Paso Niemann. The exact locality isnot far from the middle of the southeast rimof the basin, the part farthest from the RioChico, on the floor of the very irregular de-pression but near its highest part. Here thechannel of a small intermittent stream had,at the time of our work, cut a nearly verticalbank about 5 feet high in a series of verypure, clear green bentonites interbeddedwith nearly white layers of more ashy bento-nite. This small exposure, which might wellbe obliterated by a swing in the erosionchannel, was discovered by Ingeniero A.Piatnitzky, of the Argentine Federal Petro-leum Administration (Yacimientos Petro-liferos Fiscales), and indicated to us by hiscolleague Dr. Egidio Feruglio.

"In the green bentonites there werenumerous bones of some medium-sized flyingbird,1 also the type material of the meiolaniidturtle Crossochelys corniger (see Simpson,1937a, 1938), fragments of other unidentifi-able turtles, a new amphibian resemblingCeratophrys, and a single mammal jaw, typeof Codna pattersoni. The type of Sebecusicaeorhinus was scattered through the matrixwith these other highly varied remains.

"All the identifiable specimens from thissmall fossil pocket represent hitherto un-known forms and with the exception of

1These are now in the hands of Dr. Alexander Wet-more for description.

Sebecus itself, as noted below, none of theassociated species has ever been found at anyother locality. The internal evidence as toage is extremely vague. The faunule on itsown evidence might belong almost anywherein the later Mesozoic or the Tertiary. InCafnad6n Hondo, however, were found mam-mals of the Notostylops fauna, characteristicof the Casamayor formation, at horizonsapparently both above and below that ofSebecus. The exact succession of strata is herevery confused, since the exposures are dis-continuous and are folded and faulted, but itseems highly probable that this green bento-nite is a peculiar local facies of the Casamayorformation. This is confirmed by the occur-rence of Sebecus at other surely Casamayorlocalities."The principal referred specimen of Sebecus,

A.M.N.H. No. 3159, was found in Cafiad6nVaca, as is Cafiadon Hondo, tributary to theRio Chico and also just above Paso Niemann,but on the other, the northwest, side of theRio Chico Valley. The exact locality is in anembayment known locally and in our fieldnotes as the Oficina del Diablo, on the north-east wall of Cafiad6n Vaca, which here is ascarp bounding the Pampa Pelada, near theupper end of the Canad6n. The specimenwas entirely weathered out, but the condi-tions of erosion were such that it could onlyhave come from beds in which a rich, unified,and typical Notostylops fauna occurs. It ishence certainly of Casamayor age.

"Several isolated teeth almost surely ofSebecus have been found, and always in sureor probable association with Casamayorguide fossils. At least one tooth in theAmeghino Collection (No. 10872 in theMuseo Nacional de Historia Natural ofBuenos Aires) appears to belong to Sebecus.It is recorded as from Notostylops (i.e.,Casamayor) beds west of the Rio Chico,probably in the vicinity of Cafiad6n Vaca.We also have a single tooth, probably of thisgenus, A.M.N.H. No. 3162, found in placein the Casamayor of Cafiadon Vaca."

224

TAXONOMY

ORDER CROCODILIASUBORDER SEBECOSUCHIA SIMPSON, 1937

DIAGNOSIS: Basically crocodilian in cranialstructure and with secondary palate. Internalnares very wide. Whole skull, and expeciallythe facial part, relatively much narrower anddeeper than in other Crocodilia; snout verydeep and with median crest above. Orbitsdirected laterally. Teeth reduced in numberand generally strongly compressed laterally,with serrated edges, the larger teeth withcrowns almost indistinguishable from thoseof some carnivorous dinosaurs. Vertebraefeebly amphicoelous.

FAMILY SEBECIDAE SIMPSON, 1937TYPE: Sebecus.DISTRIBUTION: As for the genus.DIAGNOSIS: Skull compressed and deep,

especially in the facial region. Internal naresanteriorly placed, the front border beingformed by the palatines, the back border bythe pterygoids; palatine tube incipient. Well-developed quadratojugal-surangular articula-tion; quadrate inclined. Supratemporal fe-nestra rather small, broader than long. Fourteeth in the premaxilla, 10 in the maxilla,and 13 in the dentary, none of which aregreatly differentiated as to size. The pre-maxillary and anterior dentary teeth arerounded in cross section, while the otherteeth are laterally compressed, with well-developed serrations on the anterior andposterior cutting edges. There is a shallownotch along the premaxillary-maxillary su-ture for the reception of the fourth dentarytooth.

GENUS SEBECUS1 SIMPSON, 1937TYPE: Sebecus icaeorhinus.DISTRIBUTION: Casamayor formation, Eo-

cene, Patagonia.

1 Egyptian sbk, or sebek, the crocodile god, arbitrarilyLatinized. The usual hieroglyphic writing of the Egyp-tian word for crocodile appears to have been mswk oremsuh, according to one of several arbitrary methods ofmaking the vowelless hieroglyphs pronounceable. Theuse here of the word sebek, for the crocodile god, is in-troduced to vary the general practice of using the wordChampsa in the nomenclature of fossil reptiles of croco-

DIAGNOSIS: Sole known genus of Sebecidae.Therefore at the present time generic andspecific diagnoses cannot be distinguishedfrom the family diagnosis.

Sebecus icaeorhinus2 Simpson, 1937TYPE: A.M.N.H. No. 3160, most of the

bones of the skull and jaw, as listed below.Collected by Coleman S. Williams and G. G.Simpson, March 7, 1931, in Cafiadon Hondo,Chubut.PRINCIPAL REFERRED SPECIMEN: A.M.N.H.

No. 3159, numerous skull and skeletal frag-ments, as listed below. Collected by G. G.Simpson, January 3, 1931, in Cafiadon Vaca,Chubut.A.M.N.H. Nos. 3160 and 3159 differ in

size and in various minor details, and it can-not be absolutely established that the referredspecimen is of the same species as the type,but this is very probable and there is noproper basis for specific separation.HORIZON: Casamayor formation, Eocene.LOCALITIES: Cafiadon Hondo and Cafnadon

Vaca, tributaries to the Rio Chico delChubut, Chubut, Patagonia, Argentina.

DIAGNOSIS: Sole known species of Sebecus.

FAMILY BAURUSUCHIDAR PRICE, 19453TYPE: Baurusuchus.DISTRIBUTION: As for the genus.

dilian relationships or crocodilian appearance. X4j&4Pat isnot a Greek word but was given by Herodotus ("His-tory," Book II) as the Egyptian word for the crocodile.According to the same authority, xpog6aetXos, whencecomes the word for crocodile in most modern languages,properly meant lizard. The crocodile was called d po,68e&-Xos 6 w-or&,os, "river lizard," and eventually the qualifier"river" was dropped. It is an interesting parallelism inlanguages that the word "alligator" also originallymeant only "the lizard," el lagarto.

2 EL,alos, random, not according to plan, and kw6,snout, in allusion to the remarkable departure of thisanimal's snout from the crocodilian plan.

' After the manuscript of the present paper had beencompleted and was ready for press, the author receivedfrom Mr. L. I. Price of the Divisio de Geologia e Min-eralogia, Minist6rio da Agricultura, Rio de Janeiro,Brazil, a paper describing a new and remarkable croco-dilian obviously related to Sebecus. This new crocodilian,considered as representative of a separate family of the

I Sebecosuchia, is inserted here and is discussed brieflyon a subsequent page of the present contribution. How-

225

BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY

DIAGNOSIS: Skull compressed and deep,especially in the facial region. Internal naresposteriorly placed, bounded in front and inback by the pterygoids and laterally by theexpanded ectopterygoids. Well-developedpalatine tube. Quadrate vertical. Supra-temporal fenestra large and longer than it isbroad. Teeth greatly reduced as to number,very strongly differentiated as to size, andlaterally compressed. Three teeth in thepremaxilla, and in the anterior part of themaxilla two large "caniniform" teeth. A"caniniform" tooth in the dentary for thereception of which there is a very deep notchin the anterior part of the maxilla.

ever, it was considered inadvisable to attempt detailedcomparisons between the new genus described by Priceand Sebecus, especially since Price's paper is a prelimi-nary note in which many of the anatomical details ofthe new form are omitted. We will look forward withpleasure to the appearance of Mr. Price's larger study,which is promised at some future date.

GENUS BAURUSUCHUS PRIcE, 1945

TYPE: Baurusuchus pachecoi.DISTRIBUTION: Bauru formation, Creta-

ceous, Sao Paulo, Brazil.DIAGNOSIS: Sole known genus of Bauru-

suchidae. Therefore at the present time gen-eric and specific diagnoses cannot be dis-tinguished from the family diagnosis.

Baurusuchus pachecoi Price, 1945TYPE: D.G.M. [Divisao de Geologia e

Mineralogia] No. 299 R. A skull and mandiblewith the left side partially destroyed byerosion.HORIZON: Baurui formation, Upper Creta-

ceous.LOCALITY: "Twelve leagues (approximately

72 km) Southwest of Vila de Veadinho,municipia of Paulo de Faria, State of SaoPaulo." Brazil.

DIAGNOSIS: Sole known species of Bauru-suchus.

VOL. 87226

MORPHOLOGYMATERIAL

ASIDE FROM ISOLATED TEETH, adding nomorphological data, the material of Sebecusavailable consists of the two specimens al-ready mentioned. In more detail, these in-clude the following parts:A.M.N.H. No. 3160: Left premaxilla, right

and left maxillae, right and left nasals,right and left frontals, right and left parie-tals, right and left lacrimals, right and leftprefrontals, left jugal, left squamosal, leftquadratojugal, right and left palatines, rightand left ectopterygoids, left quadrate, leftexoccipital, left postoptic, basisphenoid, rightand left palpebrals, right and left dentaries,left angular, left surangular, and right articu-lar. Some of these are incomplete (see figures)but they are for the most part well pre-served.A.M.N.H. No. 3159: Left maxilla, left

nasal, left frontal, right and left prefrontals,right and left jugals, left ectopterygoid, rightand left quadrates, left quadratojugal, leftsquamosal, right surangular-all of these areincomplete and most of them are mere frag-ments but so characteristic that they leave nodoubt that the genus and probably thespecies are the same as the type. The follow-ing parts present in this specimen are notentirely duplicated in the type: parts of bothsides probably of fused pterygoids; the lowerpart of the occiput with the ba'ioccipital,essentially complete; part of the supra-occipital; one vertebral centrum; most of theright femur; distal end of the right fibula. Inaddition, there are various fragments, tooincomplete for identification.The material included under No. 3159 was

all weathered out and scattered over a smallarea, with a small amount of extraneouswashed material. It is thus not certain thatthis is a single individual or even that it allbelongs to the same species, but it is highlyprobable. The material considered extraneousis obviously so. There is no material probablyof another crocodilian, or reptile. The variousfragments duplicated in No. 3160 are allclearly of the same genus as the latter and areall from a somewhat smaller individual. Allthe fragments are of such size and character

as to belong to one animal, and there is noduplication. No. 3160 was also disarticulatedand scattered, although in situ, but certainlyrepresents a single animal, from the perfectarticulation and harmony of the variousparts.The skull is thus almost completely known.

The only elements that were present in theliving animal but that are not known in thefossils are the postorbital, prevomer, andprootic. From imperfections of the preservedbones the exact configuration of the externalnares and of the pterygoid region remainsdoubtful. In the lower jaw the coronoid andsplenial are lacking, leaving doubts as to partof the region near the internal mandibularforamen, but otherwise the mandible, too, iswell known. Very little is yet known of thepostcranial skeleton, but that little is of somevalue, especially revealing the nature of atypical vertebra and the essentially croco-dilian modification of the hind limb.

Nothing definitely recognizable as a post-cranial dermal scute was found with eitherspecimen. In view, however, of the scarcity ofpostcranial material in general, this negativefact has little value for inference.These cranial and mandibular elements

were mostly disarticulated as found, and thedisar:iculation has been completed and re-tained except for the most posterior elements,occiput, squamosal, and quadrate. The sepa-rate bones are thus remarkably adapted forstudy. In order to have a representation ofthe skull as a whole, each separate part of No.3160 was cast in plaster, as mentioned above,and a set of these casts was incorporated ina plaster reconstruction of the articulatedskull and of the mandible. The bones arelittle crushed, especially when the fragilenature of some of them and the usuallydestructive effect of a bentonite matrix areconsidered, but there is some slight distortionthat has demanded correction and compro-mise in the reconstruction so that it doubtlessdiffers in certain small details from a freshskull of the species. However, the differencescannot be great except, perhaps, for theexternal nares and pterygoids.

227


THE SKULLGENERAL REMARKS

One is impressed when first viewing theskull of Sebecus by its lack of similarity to thetypical crocodilian skull, for this skull, nar-row and deep, with strongly compressedteeth, appears in a superficial view to havelittle in common with the flattened skull socharacteristic of the Crocodilia. Yet a carefulexamination of the skull in Sebecus will showthat it is crocodilian through and through,and although there are many importantanatomical differences that separate it fromthe skulls of any other known crocodilians,much of its strange appearance is due to thedifferences in proportions between it andother crocodilian skulls.

In discussing the skull and mandible ofthis genus, I propose to take up each ele-ment separately, and after all of the knownbones have been described and compared inthis manner, to treat the skull and jaw as awhole. Thus the details will be presented first,after which a synthesis of the whole will beattempted, and comparisons of the structurein its entirety will be made with other forms.

BONES OF THE SKULLSCULPTURING OF BONY SURFACES: In

Sebecus, as in most of the other members ofthe Crocodilia, the bones on the surface ofthe skull are heavily sculptured. This charac-ter, common to the order, is to be expected,even in an animal seemingly so aberrant as

Sebecus, and needs no particular commentat this place. Mention might be made, how-ever, of the form of the bony rugosities inSebecus, for in this genus the surfaces of thebones are irregularly broken up by thesculpturing, as is the case in the more primi-tive crocodilians such as the mesosuchians.There is in Sebecus none of the rather regularreticulated pattern of ridges and pits that isseen particularly on the skull roof but fre-quently over the entire surface in the eu-suchian skull. In this respect Sebecus retainsa primitive character.

PREMAXILLA: The premaxilla is a rathershort bone that carries the alveoli for fourteeth. Of these, the two middle teeth wouldappear to have been somewhat larger than theanteriormost or the posterior tooth. In palatal

aspect, there are to be seen two pits in thepremaxilla, one placed internally to the frontedge of the alveolus for the enlarged thirdtooth, the other just inside the alveolus forthe fourth tooth. These pits were for thereception of the points of the second andthird dentary teeth. A groove extending fromthe inner side of the premaxilla, between thealveoli of the first and second teeth andaround onto the external surface of the boneindicates that the first dentary tooth bit up-ward between the first and second pre-maxillary teeth. Evidently the two firstpremaxillary teeth were very close to eachother.The premaxillary aperture, or foramen in-

cisivum, was a narrow, elongated opening,extending from just behind the first pre-maxillary tooth to a point opposite themiddle of the third tooth.

Immediately behind the last premaxillarytooth and in front of the suture between thisbone and the maxilla there is a vertical grooveor notch for the accommodation of the en-larged fourth dentary tooth during the closureof the jaws, a character whereby this genusresembles the Crocodilidae among the Eu-suchia. Unfortunately the upper portion ofthe premaxilla is not present in the materialsat hand, but it would seem that the externalnares were rather large, probably with theirrims raised or flared to some extent. Only asmall portion of the edge of the left naris ispreserved. Whether or not the basal bonesextended forward to form a median bridgedividing the nares is a point that cannot besettled upon the basis of the evidence at hand.MAXILLA: The maxilla in Sebecus is note-

worthy, above all, by reason of its greatdepth. Here we see a character whereby thisbone shows great contrast to the maxilla ofother crocodilians, and gives to the skull ofSebecus much of its distinctive appearance.The outer surface of the maxilla in this croco-dilian is essentially vertical, and because ofits height the top of the muzzle, in its poste-rior region, is virtually on a level with thefrontal region, a condition contrasting withthat in the typical crocodilians, in which theexternal surface of the maxilla is largelyhorizontal and there is a decided drop in

VOL. 87228

COLBERT: SEBECUS, A PECULIAR FOSSIL CROCODILE

level from the frontal and interorbital regionto the muzzle. In its posterior portion themaxilla in Sebecus is about twice the depththat it is anteriorly, and this naturally causesthe muzzle to slope down from the preorbitalto the nasal region. As a result of the deepmuzzle in Sebecus the narial passage is verydeep and narrow, whereas in the Eusuchia,for instance, it is broad and shallow.There are some interesting points to be

brought out in connection with the alveolarborder in the maxilla of Sebecus whereby thiscrocodilian may again be contrasted withother members of the order. In the first placethere are but 10 teeth in the maxilla of thisgenus, a low number as compared with mostother crocodilians, either fossil or recent.Moreover, of these teeth, as shown by suchteeth as are preserved (particularly in theleft maxilla) and by the alveoli, the last sixare of almost the same size. Subequal maxil-lary teeth are common in the long-snoutedcrocodilians of the suborders Mesosuchia andEusuchia, but in those genera where no elon-gation of the tooth carrying elements hastaken place, there is a tendency for a strongdifferentiation in size among the maxillaryand dentary teeth. Of the first four maxillaryteeth in Sebecus, the first two are definitelysmaller than the more posterior teeth, whilethe third and fourth teeth in the series arenoticeably larger than the teeth behind them.However, the difference between these en-larged teeth and the teeth that follow them isnot particularly great, as it is in many eusu-chians.

Except for the first two teeth, all of themaxillary teeth are large, with correspond-ingly large alveoli. Indeed, the enlargementof the last five or six teeth has resulted inswellings of the maxilla on its internal surfacefor the accommodation of the tooth bases.These swellings occur above the horizontalpalatine plate of the maxilla. Naturally, thealveoli are elongated and rather narrow, toaccommodate the unique, laterally com-pressed teeth that characterize this croco-dilian genus.

Because of the subequal size of the maxil-lary teeth, the alveolar border in Sebecus iscomparatively even, showing little of thesinuosity so typical of this border in manyeusuchian crocodiles; in fact this border is

essentially straight from the third maxillarytooth to the end of the series. But in frontof the third maxillary tooth (the first of thelarge maxillary teeth) the alveolar bordermakes a sharp bend upward, to meet theventral border of the premaxilla.

Between the last seven alveoli and closeto their inner edges are well-developed pitsin the maxilla for the reception of the pointsof the lower teeth. This is a character fre-quently seen in modern crocodilians. Also,placed medially to the alveoli there is on

Na

A

FIG. 1. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, left premaxilla, X1. A, External lateralview; B, internal view. Na, Border of externalnarial opening; N, notch in premaxilla, just infront of suture, for reception of the fourth dentarytooth; 1, 2, 3, 4, numbers of premaxillary teethand alveoli.

either side a row of small, but quite distinct.vascular canals or foramina, developed inthe palatal plate of the maxilla. In the mod-ern crocodilians these canals communicatewith the alveolar sinus, which is located inthe outer portion of the maxilla above thetooth row and separated from the nasal pas-sage by a vertical bony wall. In the persistingEusuchia this alveolar sinus is very large andwell developed, extending for a greater por-tion of the length of the maxilla and occupy-

2291946


ing a considerable width within the muzzle.It would appear that the alveolar sinus musthave been much constricted in Sebecus, beinghardly more than a narrow canal running thelength of the maxilla immediately internalto the alveolar borders and superior to thevascular canals upon the palatal surface.

that pari passu with the reduction in thenumber of the teeth, there was a generalincrease in the size of these teeth, with theresult that the tooth row occupies almost thesame amount of space, in relation to the totallength of the skull, that it does in many othercrocodilians in which the teeth are more nu-

FIG. 2. Sebecus icaeorhinus Simpson. A.M.N.H. No. 3160, left maxilla, Xi. A,External lateral view; B, internal view. 1-10, Numbers of maxillary teeth andalveoli.

Moreover it would appear that there was noenclosed space for a diverticulum of the nasalsac between the alveolar sinus and the nasalpassage, as is so characteristic of the moderncrocodilians.One other point with regard to the maxilla

should be brought out at this place. This isthe fact that although there is some reductionin the number of maxillary teeth in Sebecus,there has not been much reduction in thelength of the alveolar region. This means

merous but comparatively smaller than theyare in Sebecus.The materials at hand show that the max-

illa was in contact suturally with the nasal,lacrimal, and jugal-in other words, thatthere was no preorbital foramen in this genus.This is of course a typical crocodilian feature,to be expected even in an aberrant form suchas Sebecus.NASAL: Like the maxilla, the nasal in

Sebecus is distinguished from the same ele-

230 VOL. 87


ment in other crocodilians in that its surfaceis for the most part approximately verticalrather than horizontal. The two nasals meetalong a greater portion of their length toform a narrow crest to the muzzle in thiscrocodilian. The anterior portions of bothnasals are missing in the materials at hand,but it seems probable from the configuration

FIG. 3. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, left nasal bone, anterior portion missing,Xi. A, External lateral view; B, internal view.

of the premaxilla and maxilla that the nasalswere somewhat expanded anteriorly andseparated the premaxillae from each otherjust behind the external narial openings. Pos-teriorly these bones are broadened, and eachshows two distinct external surfaces-a ver-tical surface and a horizontal surface almostat right angles to the vertical surface. Thevertical part of the nasal joins the lacrimal,behind the maxillary suture, while the hori-zontal surface terminates in a sutural junc-tion with the prefrontal and frontal.LACRIMAL: The lacrimal bone in Sebecus

has been deepened, correlative with the deep-ening of all of the bones forming the side ofthe skull in this genus, and its external sur-face is vertical, rather than horizontal as inthe typical crocodilians. It has an anteriorextension in the form of a point, boundedabove by the nasal and below by the maxilla.Behind the nasal, this bone articulates withthe prefrontal. Posteroventrally the lacrimalis again pointed, and in this case the upperedge of the point forms a part of the orbitalborder, while below it is suturally joined withthe jugal.

There is a well-defined lacrimal duct withinthe bone, the posterior exit of which is atthe anterior upper corner of the orbit nearthe juncture of the lacrimal and prefrontal,while its anterior opening is on the inner sur-face of the bone near its upper border. Thisduct in Sebecus is not very long, nor is itsexpansion within the bone so great as it is inthe modern crocodiles.PREFRONTAL: This bone in Sebecus shows

the usual crocodilian relationships since itforms the anterodorsal section of the orbitalrim, while it is joined on either side by themaxilla and frontal, respectively, and an-teriorly by the nasal. However, the prefrontalin this fossil differs from the same bone inmany of the modern crocodilians in that it isnot produced anteriorly into a long point,running far forward between the nasal,maxilla, and lacrimal.On the other hand, the prefrontal in Sebe-

cus extends farther back over the orbit thanit does in any of the other crocodilians withthe possible exception of certain metriorhyn-chids. Along much of their length the twoprefrontals are separated from each other bya narrow anterior tongue of the frontal.

Curving down from the posteroventral sur-

FIG. 4. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, left lacrimal, Xi. A, External lateralview; B, internal view. Orb, Position of orbit;Lad, lacrimal duct.

face of each prefrontal in Sebecus is a bonylamina which terminates at a comparativelyshort distance below the cranial roof in an ir-

2311946


regularly broken edge. This is the bony exten-sion homologous with a similar downwardprocess which in the Eusuchia extends downto meet the palatine and pterygoid bones. Thequestion is, Was there a bridge of bone inSebecus similar to that connecting the top andthe bottom of the skull in the modern crocodil-ians? Probably, but it must have been con-siderably different from the same structureseen in the Eusuchia. In the first place, thedistance between the upper surface of thepalatines and the lower surface of the pre-

bones are roughly triangular in shape, andof the two the one on the left side is consid-erably larger than the one on the right side.FRONTAL: This bone is flat and very broad

in its posterior region, where it is boundedbehind by the parietal and on each side bythe postorbital. It forms a small part of theupper rim of the orbit on each side, and inthis region it is perfectly flat and rather thin-not thickened and raised into a prominentrim as in the modem crocodilians. Evidentlythe eye did not protrude above the skull roof

FIG. 5. Sebecus icaeorhinus Simpson. A.M.N.H. No. 3160, frontal,prefrontals, and palpebrals, Xi. A, Dorsal view; B, ventral view. Pal,Palpebral bone; D pro, descending process of prefrontal; Olf, impres-sion of olfactory bulbs on the under surface of the frontal.

frontals is relatively small in the modemforms, whereas it is considerable in the deep-skulled Sebecus. Therefore the bony pillarmust have been rather long in the extinctgenus. However, the dorsally directed lam-inae on the palatines are so thin, it wouldseem probable that the connection betweenthe roof and the palate must have been con-stituted either of very thin bone or in part ofcartilage. In the modem crocodilians thisbony connection serves in part to separatethe nasal sacs from the orbital region, and inpart (where the two plates join each otherventrally along the median line) to form asupport for the free ends of the olfactorybulbs. It is probable that the same functionswere performed by the homologous partitionsin Sebecus, whether they were wholly bonyor in part cartilaginous.

PALPEBRAL: Attached to the orbital rimas formed by the prefrontal there is on eachside a large palpebral bone in Sebecus. The

in Sebecus as it does in the eusuchians, butrather it'occupied a more normal positionand was directed laterally. Anteriorly thefrontal is constricted into a long tongue orprocess, forming somewhat more than halfthe length of the bone, and bounded on eitherside by the large prefrontals.

It might be said here that the interorbitalportion of the frontal is relatively broad.Perhaps this is owing in no small part to thelack of the raised orbital rim which charac-terizes the modem Eusuchia, and of whichso many have the dorsal rims of the orbitsvery close to each other. There is a well-defined median ridge on the back part of thefrontal, and the bone terminates posteriorlyin a very strong transverse suture for articu-lation with the parietal.

Ventrally the frontal shows in its posteriorportion a concave surface for the cerebralportion of the brain,'bounded on either sideby the strong articulations for the postoptics.

VOL. 87232


The endocranial surface takes the form of abroad groove, which is continued forwardwith but little diminution for the accommo-dation of the olfactory stalks. Anteriorly,where this groove enters the region of theprefrontals, it is broadened again, to giveroom for the enlarged olfactory bulbs.

PARIETAL: The parietal, articulating withthe frontal by a very heavy sutural union, isa relatively short bone, much constricted inits middle region by the development of thesupratemporal fenestrae. The upper surfaceof the parietal, which is heavily rugose, isbounded on either side by a strong ridge thatserves to separate the upper surface of thebone from the supratemporal fenestra, andit is to be presumed that this ridge markedthe upper limits of attachment for the strongtemporal muscles. The two ridges, closelyappressed to each other anteriorly by the en-croaching temporal fenestrae, diverge sharplyfrom front to back at an angle of about 75degrees.The upper surface of this bone rises sharply

from front to back, so that in a lateral viewthere is a strong upswing of the cranial roofbehind the frontal-parietal junction. Thismay be contrasted with the more typicalcrocodilian condition in which the back por-tion of the upper frontal and the parietalsurfaces are on the same horizontal plane.Because of this upward growth of the parietalbone, the occipital crest, that is the posteriortransverse boundary of the skull roof formedby the parietal and the squamosals on eitherside, is raised high above the rest of the skull,and this feature serves, together with manyother characters, to distinguish the skull ofSebecus very strongly from any other croco-dilian skulls. The reason for this particulardevelopment in the skull of Sebecus is amatter for conjecture; perhaps it is a corol-lary of the general growth trend in this genus,whereby vertical forces were very strong ascompared with horizontal forces of growth.It did give an expanded area for the origin ofsome of the temporal muscles, which will bedescribed below, and perhaps it may be corre-lated with the development of strong adduc-tor muscles in this region.At the back the parietal is very deep, and

shows sinus cavities which occupied a posi-tion above the back portion of the brain.

There is an extensive sutural union for thearticulation of the supraoccipital, whichlatter bone excluded the parietal from theback margin of the skull. Ventrally the parie-tal shows on either side, as does the frontalanterior to it, a strong articulation for thepostoptic.

POSTORBITAL: This bone is not representedin the materials of Sebecus so far known, butthere is good reason to think that it was notgreatly dissimilar from the same element inthe eusuchian crocodiles. Thus, it undoubt-edly formed a part of the border of the upper

STF

FIG. 6. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, parietal, Xi. A, Dorsal view; B, ven-tral view, STF, Position of supratemporalfenestra.

temporal fenestra, as well as the back portionof the orbital border. As seen from above itwas probably an L-shaped bone, with theangle of the L directed anterolaterally. Fromits under surface there was undoubtedly asubdermal postorbital bar which was directedventrally to meet a similar bar reaching upfrom the jugal.JUGAL: The jugal in Sebecus is a long, com-

pressed bone, which, except for its pro-portions, is generally similar to the sameelement in other crocodilians. The frontpart of the bone is rather deep, and the post-orbital bar arises at a point somewhat infront of the middle of the bone, a conditioncontrasting with that in many Eusuchia, inwhich the postorbital bar is closer to the backend of the jugal bone than to the front border.Anteriorly the jugal forms the lower borderof the orbit and posteriorly the lower borderof the lateral temporal fenestra. It might besaid here that in Sebecus both of these open-ings are directed laterally, whereas in theeusuchians the two openings are pointed inan upward direction. In its anterior regionthe jugal shows an extensive squamous su-

2331946


tural overlap over the back border of themaxilla, while posteriorly there is a longjunction with the quadratojugal.As mentioned above in connection with the

remarks concerning the postorbital, the post-

The quadratojugal is unusual in Sebecus inthat its posterior portion takes part in thearticulation between the skull and thelowerjaw. Consequently, instead of terminatingposteriorly in a point near the posterior ex-

LTF

A

B

FIG. 7. Sebecus icaeorhinus Simpson. A.M.N.H. No. 3160,left jugal, Xi. A, External lateral view; B, internal view. Orb,Position of orbit; LTF, position of lateral tempora fenestra.

orbital bar in Sebecus is subdermal, a char-acterwherebythis genus resembles the eusuch-ians.QUADRATOJUGAL: In Sebecus the quadrato-

A

tremity of the quadrate, as is characteristicof many other crocodilians, the quadrato-jugal in this genus is expanded posteriorlyinto a small condyle. This expanded condyle

B

Cond

FIG. 8. Sebecus icaeorhinus Simpson. A.M.N.H. No. 3160,left quadratojugal, Xi. A, External lateral view; B, internalview. Cond, Condyle forming outer portion of quadrate-quad-ratojugal articulation with the lower jaw.

jugal is an elongated bone, which shows the is continuous with the quadrate articulatingusual eusuchian relationshipsof being bounded surface, and it forms the external part of thealong its entire length by the jugal below and transverse condylar joint upon which thethe quadrate above. Anteriorly this bone is glenoid in the articular bone of the mandiblenotched by the back corner of the lateral rotates.temporal fenestra. QUADRATE: In its general aspects and rela-

VOL. 87234


tionships the quadrate bone in Sebecus showsstrong similarities to the same bone in theeusuchian crocodiles. This is to say that it isan exceedingly complicated bone, not onlyas to its shape but also as to its relationshipswith the various other bones that join it.

Generally speaking, this is an elongatedbone directed posterolaterally, terminatingat its free end in the transverse condylar sur-face for articulation with the lower jaw,bounded below by the quadratojugal, and atits anterior end firmly wedged in between thepostorbital, parietal, exoccipital, postoptic,squamosal, basisphenoid, and pterygoid. Ofthese several bones, the exoccipital andsquamosal overlap the quadrate extensively,thereby hiding a considerable portion of itfrom view externally.Along its back surface there is a longitud-

inal ridge, running from the posterior junc-tion of the exoccipital and squamosal withthe quadrate to the articulating surface. Thisridge is placed about medially upon the bone,and it serves to divide the upper surface ofthe quadrate very distinctly into a medialand an external portion. The ridge terminatesdistally at the articular surface of the quad-rate, where its development has caused thearticulation to assume a shape quite differentfrom the same surface in the modern Eusu-chia. In these latter, for instance, the articu-lation takes the form of a rather elongatedhourglass, as seen in a posteroventral view,expanded at each lateral end and constrictedin the middle. In Sebecus, however, the ar-ticulation is much broader in the middle thanit is at either end, because of the expansion ofthe posterodorsal ridge on the back of thequadrate. In outline the lower border of thearticular surface is essentially straight, whilethe upper border is produced posterodorsallyinto a prominent point in its middle region.The comparison will be made clear by figure9.On the inner side of the upper or back sur-

face of the quadrate there may be seen a

canal, the opening of which is shortly abovethe articulation, running along the length ofthe bone toward the tympanic cavity. Thelower course of this canal has been exposedby a breakage of the bone which originallyroofed it in. According to Miall, "Stanniusfirst pointed out that by means of this canal

[in recent crocodilians] a pneumatic com-munication subsists between the tympaniccavity and the articular piece of the mandi-ble" (Miall, 1878, p. 23).There is also a canal on the ventral side of

the quadrate, seemingly running from thetympanic region to a point about 40 mm.from the articulation, where there is a pos-teriorly directed opening or exit.

In a lateral view two prominent openingsare seen, namely, a very large external audi-tory meatus, in which the tympanic mem-brane was lodged, bounded below by the quad-rate and above by the squamosal, and anenlarged foramen in front of the meatus,leading into the tympanic cavity.

In a posterior view of the quadrate theremay be seen the opening between this boneand the exoccipital which served as an exitfor the facial nerve.SQUAMOSAL: The squamosal bone in Sebe-

cus, as in the eusuchians, forms the postero-external corner of the skull and broadlyoverhangs the ear region, forming a sort ofprotective roof for the latter. From the simi-larities between the squamosal in Sebecus andin the Eusuchia, it would seem probable thatin the extinct genus, as in the existing croco-dilians, there was an epidermal flap attachedto the outer edge of the squamosal, whichserved to close the ear during submergence ofthe animal. In the modern crocodilians thisflap, attached at its upper edge to the squa-mosal, is movable. It can be lowered whenthe animal submerges, so that it effectivelyseals the eardrum from contact with thewater. When the animal raises the head abovethe water this flap is raised, so that the croco-dilian can readily hear airborne vibrations.The squamosal, as seen from above, is

essentially a three-cornered bone,. the dorsalsurface of which is roughly pitted and sculp-tured. Anteriorly this bone forms the pos-terior boundary of the supratemporal fe-nestra, and, as in recent eusuchians, the boneis smooth below the upper rim of the fossa.Unfortunately the squamosal is broken inthe region where it would normally articu-late with the parietal; there is reason tothink, however, that in this genus, as in re-cent eusuchians, there was a foramen for theentrance of the temporal artery into the skullat the junction of the two bones.

2351946


As in the Eusuchia, the squamosal inSebecus is characterized by a vertical de-scending plate beneath its posteriorly di-rected process. On its anterior side this platearticulates with the quadrate behind the ex-ternal auditory meatus, and, as it rises

CX

FIG. 9. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, left quadrate, squamosal, and exoccipital,Xi.7-A, Occipital view; B, internal view; C, ex-ternal lateral view. EAM, External auditorymeatus; FM, foramen magnum; SC, semicircularcanal; Eo, exoccipital; Can, exit of canal inquadrate communicating with tympanic cavity;Opo, opisthotic; Qu, quadrate; Sq, squamosal;Vn, passage for small vein from tympanic cavity;VII, X, XII, foramina for the seventh, tenth, andtwelfth cranial nerves, respectively.

anteromedially to meet the lower surface ofthe horizontal plate of the squamosal, it isdeveloped as a free edge that forms the upperborder of the meatus. Posteriorly this verticalplate is overlapped by the upper edge of thelarge exoccipital.

EXOCCIPITAL: As seen in posterior view,the occiput of Sebecus is dominated by thelarge exoccipitals, so characteristic of thecrocodilians. The paired exoccipitals extendfrom the foramen magnum, of which theyform all but the ventral border, laterally to apoint on either side just beneath the posterioredge of the squamosal. Thus each exoccipitalis a broad bone, and in addition is an elementof considerable depth, since its lower borderis depressed on either side to a level belowthe occipital condyle, while its upper borderextends well up on the occipital surface. Thisbone overlaps rather extensively the quad-rate, thereby hiding a considerable portionof the latter bone, as seen in posterior oroccipital view.

Five foramina are visible in the posteriorview of the exoccipital, as is commonly thecase in the modern Eusuchia. Immediatelylateral to the foramen magnum is the largehypoglossal foramen for the passage of thetwelfth cranial nerve, while lateral to thehypoglossal foramen at a distance of about15 mm. are two foramina, located very closeto each other and contained in a commondepression on the surface of the bone. Ofthese the inner one is small and served forthe passage of a small vein from the tympaniccavity, while the outer one is a large foramen-the exit for the vagus nerve. Below theseforamina is the exit for the internal carotidartery, which latter traversed the exoccipitalvertically from its lower exit, passing fromthis bone into the tympanic cavity. Finally,near the lateral extremity of the exoccipital,between the juncture of this bone with thequadrate, is a large, vertically elongatedforamen which is directed toward the articu-lar surface of the quadrate. This foramenserved for the exit of the seventh or facialnerve, as well as for a branch of the jugularvein and an artery. A long canal traversingthe exoccipital-quadrate suture led from thisforamen into the tympanic cavity and servedfor the passage of the nerve and blood vessels.The exoccipital of Sebecus resembles the

VOL. 87236


same element in the Eusuchia in that it hasa very broad, ventrally directed sutural sur-face for articulation with the basioccipital,and a broken edge of bone shows that therewas also a ventrally directed process, as inthe Eusuchia, also for articulation with thebasioccipital. Between these two articula-tions there is seen in Sebecus a large cavitywithin the bone, located immediately lateraland slightly ventral to the posterior part ofthe brain case. This is the rhomboidal sinusof the lateral Eustachian passage, to be dis-cussed below.As seen from a medial viewpoint, the exoc-

cipital of Sebecus shows the inner wall of theforamen magnum, which formed the lateralenclosure for the medulla oblongata. Thisstrongly concave surface is pierced by threeforamina, as in the Eusuchia. These are thelarge hypoglossal foramen at the back,through which the twelfth nerve passed, asmall foramen for a vein, which joined thevein from the tympanic cavity, mentionedabove, and an elongated slit-like opening atthe front. This last opening is actually lo-cated at the junction of the exoccipital withthe opisthotic, which latter bone in the adultcrocodilians generally is firmly fused withthe exoccipital. The slit, inclined at an angleof about 45 degrees to the floor of the brainbase, is the opening through which passedthe vagus nerve. The opisthotic, in front ofthe slit, is rather well preserved and will bedescribed below.

In this median view of the exoccipital-quadrate-squamosal complex the tympaniccavity, which is exposed, is seen to be largeand even more expanded than it is in therecent crocodilians.

In this region of the skull there is a differ-ence in structure whereby Sebecus may becontrasted with the modern forms with whichit has been compared. In Crocodilus, for in-stance, the bone of the exoccipital behind thetympanic cavity is thickened so that itsanteroposterior diameter is as great as thesame dimension of the cavity in front of it.Internally the bone shows a certain amountof excavation by small sinus cavities, butthese are not extensive. In Sebecus, on theother hand, the back wall of the exoccipitalis more posteriorly placed and its thicknessbehind the tympanic cavity is much less than

in the eusuchians, so that there is a large,hollow space or pocket in this region almostas large as the tympanic cavity itself. More-over, there is only a partial bony wall be-tween the two cavities, and consequently thecavities are open to each other over much oftheir extent. In effect the tympanic cavityhas been greatly enlarged; as a matter offact, it has almost doubled in size.

It is difficult to speculate as to the reasonfor this development in Sebecus. It is to benoted, however, that the external auditorymeatus, and by inference the tympanic mem-brane, is much larger than in recent croco-dilians. Perhaps the enlargement of thetympanic cavity in Sebecus is correlated withthe large meatus, and together they mayindicate an unusualy acute sense of hearingin the extinct genus.

In Sebecus the exoccipital is very firmlyfused to the quadrate by means of a strongand extensive suture. This afforded an almostinseparable union between the two bones.

OPISTHOTIC: The opisthotic is firmly fusedto the exoccipital in the Crocodilia, and gen-erally this bone is with difficulty distinguishedas a distinct element. In the Sebecus skull,however, the opisthotic is plainly seen. Itslateral portion is a squamous plate, attachedfirmly to that thin wall of the exoccipitalwhich partially separates the tympaniccavity proper from the enlarged posteriorcavity described above. Medially, the opis-thotic is thickened and pierced by those por-tions of the semicircular canals traversingthis bone.

SUPRAOCCIPITAL: Although the supraoc-cipital is missing in the Sebecus materials sofar known, the sutures on the exoccipital,squamosal, and parietal give an adequateidea as to the extent and shape of this par-ticular element. This bone extended up fromthe two exoccipitals (which meet each otherabove the foramen magnum) flaring frombottom to top so that the upper edge of thebone was something more than twice thelateral width of its lower edge. The supra-occipital evidently met the squamosal oneither side, whence its upper border extendedforward on the roof of the skull in a wedge-shaped process that met the parietal.Thus the parietal was excluded from the

back border of the skull in Sebecus by the

2371946


forward growth of the superior part of thesupraoccipital, a relationship that sometimesis seen in certain eusuchians.

BASIOCCIPITAL: Our knowledge of thebasioccipital in Sebecus must be gained froma specimen supplementary to the skull andjaw upon which the bulk of this descriptionis based. This bone is partially preserved inA.M.N.H. No. 3159. Here again, as is so fre-quently the case in Sebecus, the resemblancesof the basioccipital are with the same ele-ment in the Eusuchia.At its upper margin the basioccipital car-

ries the rounded occipital condyle, and it ex-

FIG. 10. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3159, basioccipital and exoccipitals, Xi.Occipital view. Bo, Basioccipital; Car, exit ofinternal carotid artery; Eo, exoccipital; X, fora-men for tenth cranial nerve.

tends inward from the condyle to form thefloor of the foramen magnum and the braincase. Below the condyle the bone extendsdownward, its edges being embraced on eitherside by the large exoccipitals. As is commonin the crocodilians, there are on the outeredges of this bone near its ventral terminationlarge expansions in the form of elongatedtubercles which are for the attachment ofcertain neck muscles. Between these basilartubercles, in the base of the basioccipital, isthe vertically directed single opening of themedian Eustachian canal, located betweenthe basioccipital and the basisphenoid. Thedevelopment of the complicated system ofEustachian tubes in Sebecus and in othercrocodilians, of which this median canal is oneelement, will be described in detail below.On the outside of the bone, directly behindthe vertically directed Eustachian canal, is asharp and well-defined vertical ridge.On either side of the median Eustachian

canal, and again in the sutural surface be-tween the basioccipital and the basisphenoid,are the lateral Eustachian canals, runningupward parallel to the median canal. Thesealso will be described below.

BASISPHENOID: The basisphenoid is onlypartially preserved in Sebecus, but such ofthis bone as is present shows the charactersof the element very well. It is quite apparentthat the basisphenoid of Sebecus, like so manyof the other skull bones, shows very closeresemblances to the same bone in the Eu-suchia. On its upper surface is the elongated,transversely concave floor upon which themedulla oblongata rested. On either side thisfloor is pierced by a small foramen whichafforded an exit for the sixth cranial nerve.At the anterior termination of this cranial

floor is a large fossa, directed backward anddown, so that it is beneath the medullar por-tion of the brain. This opening is the sellaturcica that housed the pituitary body of thebrain which in Sebecus, as in large modemcrocodilians, was a lobe of considerable size.In this connection it might be said that apart of the function of the pituitary is thesecretion of the growth hormones, and it isan interesting fact, as has been shown byEdinger, that the large and giant reptilesare characterized by large pituitaries. At the

FIG. 11. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, basisphenoid, Xi. A, Dorsal view; B,right lateral view; C, anterior view. ST, Sellaturcica; Int Car, posterior exits of the internalcarotid arteries from the sella turcica; Med Ob,dorsal surface of the basisphenoid for the accom-modation of the medulla oblongata; VI, foramenfor the sixth cranial nerve.

bottom of the sella turcica are two foraminalexits, directly laterally. These are for thepassage of the carotid arteries, which leavethe basisphenoid dorsolaterally and enter thetympanic region in the prootic bone. In thedevelopment of these openings within the

238 VOL. 87


basisphenoid there is almost complete iden-tity between Sebecus and the Eusuchia.

Ventrally, beneath the sella turcica thebasisphenoid is constricted, and along itsanterior edge it shows a narrow articularsurface, probably for a juncture with thepterygoids.

PALATINE: Both palatines are preserved inthe Sebecus skull, and these show a ratherprimitive stage of development that dif-ferentiates this Eocene crocodilian sharplyfrom the Eusuchia. At the same time, thedevelopment of the palatines, together with

I'n~~~~~~~~~~~~I

FIG. 12. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, palatine Xi. A, Dorsal view; B, ventralview. In, Position of the internal nares; Pt fos,position of the pterygoid fossa. The arrow showsthe course of the nasal passage across the dorsalsurface of the bone.

the conjoined pterygoids, is markedly differ-ent from what is seen in the Mesosuchia, afact that emphasizes once again the distinctposition of Sebecus in the phylogenetic his-tory of the Crocodilia.Each palatine is shaped something like an

arrowhead, the front portion being pointed,the back portion drawn out into a slenderramus or projection. The two palatines meetalong their midline, and laterally each ofthem is bounded in front by the palatineextensions of the maxilla. At the back thepalatine articulates with the pterygoid, andthis articulation is carried up on a lateralvertical plate of the palatine that serves toenclose, in part, the nasal passage. From themanner in which this vertical plate thins to avery delicate edge it would seem apparentthat the upper part of the nasal passage, en-closed within the muzzle of the animal, was

bounded by cartilaginous walls. Posteriorlybetween the two vertical plates of the con-joined palatines, these bones together form asemicircular free edge which constitutes thefront border of the internal nares. This an-terior placement of the narial border is acharacter whereby Sebecus resembles themesosuchians. However, the development ofthe pterygoids, as mentioned above, indi-cates most clearly the fact that, in spite ofcertain resemblances, there are at the sametime strong differences between Sebecus andthe mesosuchians.On each side, lateral to the large semicircu-

lar boundary of the anterior nares, the smoothsurface of the palatine formed by the roundedangle between the horizontal and the verticalplates makes up a part of the inner border ofthe large palatal foramen. The inner bound-ary for the foramen is continued posteriorlyby the pterygoid, while its external boundaryis formed by the maxilla, and its posteriorborder by the ectopterygoid.

ECTOPTERYGOID: The ectopterygoids arelong slender bones, each of which articulatesproximally with the maxilla and the jugalby a broad, sutural surface. The bone thenextends ventroposteriorly to form the frontborder of the large ectopterygoid-pterygoidflange, or wing. That portion of the ecto-pterygoid articulating with the pterygoid istherefore attenuated as a slender, curvedpoint, of which the concave surface is in con-tact with the pterygoid behind it.As is the case with so many of the other

bones in the skull and mandible of Sebecus,the extraordinary downward extension of theectopterygoids and pterygoids in this croco-dile are an expression of the dominance ofvertical growth factors over horizontalgrowth factors during the development of theskull. Yet in spite of the differences in shapeand proportions, the ectopterygoid of Sebecusshows identically the same relationships tothe bones with which it articulates, namely,the maxilla, jugal, and pterygoid, as are tobe seen in the modern Eusuchia. ThereforeMiall's statement concerning this bone in theEusuchia may be applied with equal validityto the same element in Sebecus. "The trans-palatine [ectopterygoid] has two expandedsmooth surfaces, an outer and an inner; thefirst is covered by the mucous membrane of

2391946


the mouth; the other is in contact with themandibular muscles which fill the pterygoidfossa" (Miall, 1878, p. 27).

PTERYGOID: Two fragments of A.M.N.H.No. 3159 have been identified as portions ofthe right and left pterygoids, which bones inlife were very probably joined to form a singleextensive element. The larger fragment con-

A

sion of the pterygoids over at least a part ofthe narial opening, but it is quite possiblethat this roof was in part cartilaginous, sincethe vertical flanges of the palatines withwhich it would have joined are very thin.Indeed, it is quite possible that the roofingof the internal nares was not complete, justas the enclosing of the nasal passage dorsally

B

FIG. 13. Sebecus icaeorhinus Simpson. A.M.N.H. No. 3160, left ectoptery-goid, X. A, External lateral view; B, internal view.

tains the median dorsally directed process,which extended up to meet the basioccipital,and a lateral flange which extended out tojoin the right ectopterygoid. The other frag-ment is a part of the lateral wing of the leftpterygoid. The dorsal process shows no signof a sutural division along the midline, andsimilarly the recent eusuchians show no su-tural division between the pterygoids wherethey join behind the internal nares.The dorsal surface of the lateral pterygoid

wing is smooth, and it was across this smoothsurface that the anterior pterygoid musclesmoved as they were extended or flexed. Theventral surface of the bone is interesting, inthat it bears a low ridge, running from themedian part of the bone laterally and ante-riorly. In front of this ridge the bone is exca-vated into a sort of pocket.Thus it is clear that while the internal

nares of Sebecus were mesosuchian-like inthat they were wide, with the anterior borderformed by the palatines, they were eusuchian-like in that there was also a posterior borderformed by the pterygoids: All of the evidencepoints to the fact that the nares were verylarge in Sebecus. There evidently was a roofabove the nares, formed by a forward exten-

might not have been complete. It is evidentthat the nasal tube, so characteristic of theeusuchians, was in Sebecus only incipient.The pterygoid in Sebecus must have been

a large extensive bone laterally. It articu-lated in front with the palatine and in theback with the quadrate, postoptic, and basi-sphenoid. The long, downward sweep of theectopterygoid shows that the pterygoid wingwas very large indeed, suggesting the devel-opment of strong pterygoid muscles betweenthe skull and the lower jaw.At this place it may be well to discuss the

development of the internal nares and thebones associated with their formation inSebecus as compared to the development ofthe nares and associated bones in the meso-suchians and the eusuchians. As alreadystated, the nares of Sebecus show some pointsof resemblance to those in the mesosuchians,and some to those in the eusuchians. In themesosuchians, the internal nares are com-pletely open behind, while their anteriorborder is formed by the palatines. The ptery-goids are of medium size, with rather smallflanges or wings projecting back behind theectopterygoids. Medially the pterygoids ex-tend far back on either side of the basicra-

VOL. 87240


nium, while between them is a large ventralexposure of the basisphenoid bone. In Sebecusthe anterior border of the nares is still formedby the palatines, but there is now a broadposterior border, formed by the pterygoidbones. The pterygoids have increased greatlyin size, in a large part by the development ofvery large, posteroventrally directed flanges

FIG. 14. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3159, portion of the conjoined pterygoids, Xi.Dorsal view. The heavy dotted line shows thecourse of the low ridge that enclosed posteriorlythe internal nares.

or wings, and, moreover, the ectopterygoidshave been carried back to form the outerborders of these large wings. It would seem,moreover, that because of the enlargementof the pterygoids the basisphenoid has beengreatly constricted in its ventral exposure, sothat very little of this bone is now to be seenon the under surface of the skull. The eusu-chians differ from Sebecus in that the internalnares are much smaller and now are con-

tained entirely within the pterygoids. Thereis a strong resemblance, however, in the largesize of the pterygoids in Sebecus and theeusuchians, in the large pterygoid wings, theouter edges of which are formed by backwardextensions of the ectopterygoids, and in thesmall ventral exposure of the basisphenoid.With regard to this part of the skull, a pro-gressing line might be made up of meso-suchian, Sebecus, and eusuchian, in thatorder. By a closing together of the pterygoidsbeneath the basisphenoid and by an enlarge-ment of the pterygoid wings, the conditionseen in Sebecus might have been derived fromsome basic mesosuchian stock. The eusuchianrepresents an advance beyond Sebecus mainlyby reason of the reduction in size of the naresand their posterior migration to a positionentirely within the pterygoids. Actually, itwould seem probable that the Sebecosuchiaand the Eusuchia developed parallel withone another, rather than that one was suc-cessive to the other. Perhaps the accompany-ing diagram (fig. 15) will make clear thecomparison of bones and structures in thispart of the skull in the three suborders of theCrocodilia.

POSTOPTIC: This bone, commonly knownin the literature as the "alisphenoid," is welldeveloped in Sebecus. Along its anterior anddorsal edges it shows strong articulations forthe frontal, postorbital, and parietal bones.

I

FIG. 15. A comparison of the region of the internal nares in: A, Steneosaurus (Mesosuchia); B, Sebecus(Sebecosuchia); C, Alligator (Eusuchia). Bo, Basioccipital; Ecpt, ectopterygoid; Pal, palatine; Pt, ptery-goid. Compare the intermediate condition of the internal nares in Sebecus between the mesosuchian onthe one hand and the eusuchian on the other.

1946 241


Posteriorly there is a broad articulation forthe quadrate, and presumably to some degreefor the prootic, while posteroventrally is anarticular surface that probably was in con-tact with the pterygoid. Between these lattertwo articulations is a smooth surface thatmarks a part of the anterior border of theforamen ovale for the passage of the fifthcranial nerve. It is not possible to be com-

For ov

FIG. 16. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, left postoptic, Xi. A, Left lateral view;B, dorsal (internal) view; C, ventral view. STF,Position of supratemporal fenestra; Br, surface forthe accommodation of the left cerebral hemis-phere of the brain; For ov V, foramen ovale forthe passage of the fifth cranial nerve; Sn, sinuscavity.

pletely definite about these last articulations,because of the absence of the pterygoid andthe prootic from the preserved materials ofSebecus and because of some destruction ofthe proximal portion of the quadrate. How-

ever, there is every reason to think that therelationships were essentially the same asthose seen in the modern eusuchians.

Medially the postoptic thins to a very at-tenuated edge, which in the specimen un-fortunately has been broken, so that none ofthe original edge is preserved. It is this edgewhich forms a median fissure just above andanterior to the sella turcica, a fissure whichin the recent crocodilians is bridged bycartilage and through which passes the opticnerves. Just below the median fissure for theoptic nerves there is on either side anotherfissure, the upper edge of which is formed bythe postoptic, and through which passes thethird cranial nerve, the orbital artery, and thesixth nerve. On the posterior edge of the bone,between the upper and lower articular sur-faces, serving for articulation with thequadrate, prootic complex, and the pterygoid,respectively, is a smooth surface mentionedabove, which formed the front border of theforamen ovale.The ventral and lateral surfaces of the

postoptic are smooth, the former lookingdown into the orbital region, the latter form-ing a portion of the front edge of the supra-temporal fenestra. In this respect it shouldbe pointed out that the postoptic of Sebecusis more horizontally placed than is the samebone in the modern eusuchians, with theresult that its ventral surface has a muchlower angle than the same surface in therecent crocodilians. The inner surface of thepostoptic is deeply excavated to form theanterolateral wall for the braincase, enclosingthe cerebral hemispheres.

THE MANDIBLE

BONES OF THE MANDIBLE

DENTARY: The dentary in Sebecus is rela-tively deep, as would be expected in an ani-mal in which factors of vertical growth havebeen strong as is the case in this crocodilian.Correlative with this, the dentary in Sebecusis also very thin and compressed, as com-

pared with the same bone in other crocodil-ians.Along the upper edge of the bone are 13

elongated and laterally compressed alveoli

for the lower teeth, which resemble the upperteeth in every respect. These 13 lower teethare opposed to the 14 teeth in the skull.

It will be remembered that the maxillaryteeth of Sebecus are arranged in a compara-tively straight series, that is, the alveolarborder does not show the undulations sotypical of other crocodilian genera. Similarly,most of the dentary teeth, specifically thosefrom the fifth tooth back, are contained in analmost straight alveolar border, to matchthe opposing border of the upper jaw. The

VOL. 87242


4)

4-)

CdW4)

a-

<4)

4'4

0

040

(.)

1946 243


fourth dentary tooth, however, is raised bythe upgrowth of its alveolus above the levelof the other teeth, so that it forms in effecta "canine," biting into the notch between thelast premaxillary and the first maxillarytooth. From the raised alveolus for thefourth dentary tooth the alveolar borderdrops abruptly, so that the first three dentaryteeth are much lower than teeth behindthem, while the jaw is correspondingly shal-low in this region. The first alveolus in thedentary is directed forward on either side,so that the first dentary tooth is partiallyprocumbent. It might be said at this placethat the first three dentary teeth, to judgeby the alveoli, were rounded in cross section,as are the premaxillary teeth with which theyoccluded.

In this region of the lower jaw the externalsurface of the dentary is strongly rugose, incontrast to its more posterior regions whereit becomes externally comparatively smooth.The symphysis, which is rather long, extendsback to a point between the sixth and theseventh dentary teeth. At its posterior endthe dentary is very thin, and it is notchedwhere it forms a part of the anterior borderof the external mandibular foramen.

SPLENIAL: Unfortunately the splenials arenot preserved in the Sebecus material so farobtained from Patagonia, but it is possiblefrom the extent of the sutural surfaces on thedentaries to get a pretty fair idea as to theextent and the shape of these elements inthe genus under consideration. In the firstplace, it may be said that the splenial inSebecus extends forward to participate in thesymphysis, as it does in some of the Eusuchia.Behind the symphysis the splenial was verydeep and thin, covering the inner surface ofthe mandibular ramus from the alveolarborder (or, in the anterior portion, a shortdistance below the alveolar border) to thelower edge of the jaw.CORONOID: The coronoids, like the splenials,

are missing from the materials at hand. It isvery probable that the coronoid in Sebecuswas a comparatively small element at theposterior end of the splenial, as -it is in themodern Eusuchia.ANGULAR: The angular in Sebecus is a long

bone extending from a point approximatelyunder the last dentary tooth to the back of

the lower jaw. It shows in most of its fea-tures relationships that are similar to thoseseen in the eusuchian angular, but there arecertain important differences, which will bedescribed below. Unfortunately only a partof one angular is preserved in the materialsnow available for study, but this fragment,together with sutures on other bones whicharticulated with the angular, gives us fairlyaccurate information as to the shape and therelationships of this bone.The angular forms the lower border of the

mandibular ramus for about half of its length.This relatively great extent of the angular is

FIG. 18. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, left angular, XI. A, External lateralview; B, internal view.

owing to a long anterior process of the bone,extending forward beneath the dentary to apoint below the last dentary tooth. Behindthis process the external surface of theangular, which incidentally is rugose, rises toform a part of the lower border of the ex-ternal mandibular foramen, while behindthis opening the bone still continues to rise,to lap far up on the external surface of theramus. The back portion of the angular isunknown, but from the articular it may beseen that the angular extended back beneaththe surangular and articular to form theventral surface of the postarticular process.On its internal surface the angular shows

certain important differences from the sameelement in many of the Eusuchia. On its

VOL. 87244

BULLETIN AMER. Mus. NAT. HIST.

to

1,:< a

IA

p~~~~~~~~2.

:/

N.

-

Gy'4.- -

C0

-J

4.5

GJ

o 0

4A.

WN^A

4.5 .,.w

q

AC0 ".

CU

75

Q ~

P._

*CUd

Z'0zS

wO. ' .

Ex XCzU

3~8 4

co *^ .

_,2 rE

CU 0

to . _ o

co

* -4 b_=*- -

COcoZ

(A6- r

o o_

os~~~~~~~~~~~~~~

¢d co

VOL. 87, PLATE 11

CLB


-Q j

PoF Ecpt

Pf

La

Pal

Sebecus icaeorhinus Simpson. A.M.N.H. No. 3160, skull, X %. A, Dorsal view; B, palatal view. Bo, basioccipital;Bs, basisphenoid; Ecpt, ectopterygoid; F, frontal; Ju, jugal; La, lacrimal; Mx, maxilla; Na, nasal; Pa, parietal;Pal, palatine; Pf, prefrontal; Pmx, premaxilla; Po, postorbital; Pt, pterygoid; Qj, quadratojugal; Qu, quadrate;Sq, squamosal

vVOL. 87, PLATE 12


Sebecus icaeorhinus Simpson. A.M.N.H. No. 3160, mandible, X M. A, Dorsal view; B, lateralview of left side. Ang, angular; Art, articular; Den, dentary; Sang, surangular; Spl, splenial

VOL. 87, PLATE 13


Ob.Cbl

I Sc7/

PbVI

Car

Ob I CbC

I.-i.

CblI

XII

Sc TcMo

VI

P

MealRs

MepIMep

Mo

Endocranial casts of (A) Sebecus icaeorhinus Simpson and (B) Crocodylus acutus Cuvier, and thebrain (C) of Alligator mississipiensis Daudin (from Hoffmann in Bronn). Left lateral views, X 1.Car, internal carotid artery; Cb, cerebrum; Cbl, cerebellum; Ccb, corpora bigemina; Le, lateralEustachian tube; Me, median Eustachian tube; Mea, anterior branch of median Eustachian tube;Meal, left lateral branch of anterior median Eustachian tube; Mep, posterior branch of medianEustachian tube; Mepl, left lateral branch of posterior median Eustachian tube; Mo, medulla oblon-gata; Ob, olfactory bulb; P, pineal body; Pb, pituitary body; Rs, rhomboidal sinus; Sc, semicircularcanals; Tc, tympanic cavity; I-XII, cranial nerves

A 11

B

Cb

VOL. 87, PLATF, 14


Cb.H .-

Ob

A

Cb

p Ccb

C

-1 X;!- F

IV Vlill XVEndocranial casts of (A) Sebecus icaeorhinus Simpson and (B) Crocodylus acutus Cuvier,

and the brain (C) of Alligator mississipiensis Daudin (from Hoffmann in Bronn). Dorsal views,X 1. Cb, cerebrum; Cbl, cerebellum; Ccb, corpora bigemina; Mo, medulla oblongata; Ob,olfactory bulb; P, pineal body; Sc, semicircular canals; Tc, tympanic cavity; I, II, IV, V, VIII,X, XII, cranial nerves

Cbl Sc1 /111

Mo

1..'.

,,;.

V

CblSc

Tcx Mo

V X

Cbl

VOL. 87, PLATE 15

BULLETIN AMER. Mus. NAT. HIST. VOL. 87, PLATE 16

0Ui)

PI

F.0)'0

.0

4.0.010

I..

0C)

1946

inner edge the aof the articular,runs far forwardAt the very fronthis inner edge (missing) a surfathe articulationcoronoid. Whictended down onit is quite evideramus, formed tcoronoid, did nfar as is the castance, in the 4border of the in]for the most pe

COLBERT: SEBECUS, A PECULIARr FOSSIL CROCODILE 245

Lngular curls around, in front is preserved in the Sebecus materials by ato form a low, free edge that fragment, but again the fragment, plus sup-[on the inner side of the jaw. plementary information afforded by suturesIt end of the preserved part of on other bones, gives a fairly comprehensive(its anteriormost extremity is picture as to the relationships of the wholece can be seen that served for element.of either the splenial or the The surangular occupies a position abovehever bone may have ex- the angular, while anteriorly it laps over onlto the angular at this point, the upper edge of the dentary. Although therent that the inner wall of the is no definite evidence at hand, there cannotby the conjoined splenial and be much doubt that the surangular formed aot extend back relatively so part of the back border of the externalLse in the Eusuchia. For in- mandibular foramen, as it does in theeusuchian jaw the posterior Eusuchia. In the glenoid region the surangu-ner wall of the ramus, formed lar is laterally expanded to form the outerErt by the coronoid, extends portion of the glenoid itself, while externally

the surface is very strongly rugose. Internally

FIG. 19. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, left surangular, Xl. A, Externallateral view; B, internal view.

back to such a degree that it is opposite aboutthe middle part of the external mandibularforamen. Correlative with this backward ex-

tension of the inner wall of the ramus, a smallinternal mandibular foramen is commonlypresent, opposite the front of the externalforamen, and bounded by the angular behindand the splenial in front. In Sebecus, however,the posterior border of the inner wall of theramus is so far forward that it lies consider-ably in front of the external mandibularforamen. With the inner side of the ramus

completely open opposite this foramen, therewas no occasion for an internal foramen to beformed as in the Eusuchia.SURANGULAR: This bone, like the angular,

FIG. 20. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, right articular, Xi. A, Dorsal view; B,internal lateral view.

it has a very strong articular surface for itsjunction with the articular. Behind theglenoid it would appear that the surangularextended back for some distance, finallycoming to a point where it was compressedbetween the angular below and the articularabove.ARTICULAR: So far as may be determined

from the rather broken fragment at hand,the articular in Sebecus is essentially similarto the same bone in the Eusuchia. Thus it wasoccupied by a transverse glenoid surface forarticulation with the condyle on the quadrate.The glenoid is expanded backward in itsmedial-lateral portion, where the articularand angular join, correlative with a similar


expansion in the articular surface of thequadrate. On its external edge the glenoidsurface of the articular terminates abruptly,since the external portion of the glenoid iscarried by the surangular. Naturally, thereis a strong, expanded, sutural surface on theexternal surface of the articular, for junctionwith the surangular.

In front of the glenoid the articular isproduced forward as a point, which joins theangular and surangular, while behind theglenoid the bone is produced back to formthe upper surface of the strong postarticularprocess, for the attachment of the depressormandibulae muscles.

THE DENTITION

The dental formula for Sebecus has alreadybeen discussed in the description of the pre-maxillary, maxillary, and dentary bones.However, it might be well to repeat here thatSebecus is characterized by the comparativelylow number of teeth, for along with thespecialization of the dentition in this croco-dile there evidently was a loss of teeth in theposterior portions of the maxilla and thedentary. Consequently Sebecus has a smallernumber of teeth than is to be seen in any ofthe recent eusuchians, and in most othercrocodilians, for that matter. In this respectit approaches some of the mesosuchians,particularly the Notosuchidae, of which thetype genus, Notosuchus, occurs in the Creta-ceous of Patagonia. In this regard, an extremeof reduction in the dentition is reached by thenotosuchian Libycosuchus, from the Creta-ceous of Egypt, for this genus attained an al-most edentulous condition.Some comparisons of Sebecus with other

crocodilians, with regard to the number ofteeth, may be made as follows:

SebecosuchiaSebecus

MesosuchiaNotosuchusTeleosaurus gladius

EusuchiaOsteolaemusGavialis

PREMAXILLA

4

45

45

Of course one of the remarkable charactersdistinguishing Sebecus from other crocodilians(except Baurusuchus) is the strong lateralcompression of the maxillary and of all butthe first three dentary teeth. The premaxil-lary and the three anteriormost dentaryteeth do not show this compression, butrather are rounded in cross section, as is thecase in other crocodilians.

Ext

%.,

FIG. 21. Sebecus icaeorhinus Simpson. A.M.N.H.No. 3160, seventh left maxillary tooth, X2. A,Cross section of tooth near base; B, externallateral view. Note the serrations on the anteriorand posterior edges of the tooth.

However, the teeth of Sebecus, and simi-

MAXILLA

10

753-55

1322-24

UPPER

14

1158-6017

27-29

DENTARY

13

1058-60

14-1525-26

larly those of Baurusuchus, are further dis-tinguished from the teeth in other crocodil-ians by the fact that they have ridged, ser-rated edges. In the Eusuchia, for instance,there is a sharp ridge on the anterior andthe posterior surfaces of each tooth (laterallyplaced in the front teeth) but these ridgesare smooth, not serrated. In Sebecus, on theother hand, the serrations are invariably

VOL. 87246

-1) t../

int

FIG. 22. Comparison of the skull and jaw in: A, Crocodylus rhombifer Cuvier (Eusuchia) modified fromMook; B, Sebecus icaeorhinus Simpson (Sebecosuchia); C, Baurusuchus pachecoi Price (Sebecosuchia)modified from Price. Left lateral views, not to scale. Ang, Angular; Den, dentary; Ecpt, ectopterygoid;F, frontal; Ju, jugal; La, lacrimal; Mx, maxilla; Na, nasal; Pa, parietal; Pal, palatine; Pf, prefrontal;Pmx, premaxilla; Po, postorbital; Pt, pterygoid; Qj, quadratojugal; Qu, quadrate; Sang, surangular;Sq, squamosal.


present, even on the rounded premaxillaryteeth in which the ridges are very faint. Itwas the discovery of these compressed teethwith serrated edges that led von Huene toexpress the opinion that in Patagonia thetheropod dinosaurs had persisted into theEocene period. Of course, without the evi-dence of the skull, his mistake was a naturalone, for the teeth of Sebecus, by themselves,look for all the world like the teeth of atheropod dinosaur. This problem will beconsidered in more detail on another page.The teeth of Sebecus are so arranged that

those in the upper and lower jaws for themost part alternate, while the upper teethnaturally bite over the lower teeth. As hasbeen noted above, the fourth dentary toothbites up into a notch formed in the skull atthe junctures of the premaxilla and maxilla,while the first dentary tooth, which is par-tially procumbent, bites between the firstand second premaxillary teeth.

This arrangement of the teeth in Sebecusprovides a strongly secant dentition, sincethe edges of the teeth sheared past each otherrather closely. The shear in Sebecus is not tobe compared in perfection of its developmentwith the shear attained in the carnassial teethof the carnivorous mammals, but it is nonethe less significantly evolved in this fossilgenus. Moreover, opposing shearing edgesare repeated many times over along thelength of the jaws in Sebecus, which musthave increased the scissors-like effect of thebite in this animal. Finally, the developmentof the articular surfaces between the quadrateand articular bones and the great downwardextension of the pterygoid-ectopterygoidwings provided a strictly orthal motion tothe jaws, as is the case in the persistingEusuchia.The shape and the development of the

teeth in Sebecus described above indicatethat the eating habits of this extinct croco-dilian may have been somewhat differentfrom what they are in the modern membersof the order. Recent crocodilians are veryaggressive, predatory carnivores. For a ma-jority of them, the bulk of the diet consists offishes, but all crocodilians will seize anddevour any prey that they are able to capture.Consequently mammals and birds, especiallythe former, constitute no small portion of thecrocodilian diet, and in the larger species theadult animals are a real menace to any butthe very largest mammals that may ventureinto crocodilian-infested waters. The modemcrocodilians generally seize their prey andgulp it down whole, but if a crocodilianmanages to catch a large animal, the reptilewill twist over and over rapidly, therebytearing off a limb or a part of the body of theunfortunate victim. Often two crocodilianswill get a hold on the same animal and twistin opposite directions, and by this methodof cooperation they are able to reduce ananimal of some size to pieces that they cancrush and swallow.

Sebecus must have been predaceous, but itwas probably a predator of a kind differentfrom the modern crocodilians. The front teethin this extinct reptile probably were forgrasping, but most of the teeth-the com-pressed teeth of the maxillaries and thedentaries-were obviously cutting teeth.Therefore it would seem probable thatSebecus, after having seized its prey, was ableto cut off large pieces of meat, or limbs, by ascissors-like action of the upper and lowerjaws. In this respect, Sebecus may have hadeating habits similar to those of the theropoddinosaurs, which it so closely paralleled, inso far as the development of the teeth is con-cerned.

THE POSTCRANIAL SKELETONOf the postcranial elements discovered, all

belong to A.M.N.H. No. 3159, and since thiswas a weathered specimen, the bones arerather poorly preserved. Unfortunately onlya few postcranial bones were found; these willbe described below.

VERTEBRAOne centrum of a cervical, or perhaps an

anterior dorsal vertebra, was discovered,which is fortunate, for this bone shows thenature of the articular surfaces, an importantfeature in the classification of the Crocodilia,and of the Reptilia in general.

This vertebra is surprisingly small whencompared with the skull, A.M.N.H. No.3160. The evidence would seem to indicate,however, that No. 3159 is definitely smaller

VOL. 87248


than No. 3160, perhaps on the order of 15 or20 per cent in linear dimensions, and whenthis vertebral centrum is compared with theocciput of No. 3159, of which the completecondyle is preserved, there can be no doubtthat there is an agreement in size betweenthem. Consequently it is logical to assumethat this vertebra belongs to the secondspecimen of Sebecus and is not an intrudedpart of another skeleton.As mentioned above in the diagnosis, the

vertebra is characterized particularly by thefact that both articular surfaces are amphi-coelous. They are not strongly so; in factthey approach a condition that is almostamphiplatyan. Thus a primitive conditionpersists in the vertebrae whereby Sebecusshows definite relationships with the Meso-suchia.

Ventrally the centrum is compressed, witha median ventral ridge, which in its anteriorportion is extended down to form a distincthypapophysis. The presence of a hypapophy-sis on the cervical vertebrae is characteristicof the Crocodilia.On the lateral surface of the centrum there

is, on either side, a facet for the articulationof the capitulum of the rib, while dorsallythere may be seen the sutures for the articula-tion of the neural arch upon the centrum.

LIMB BONESFEMUR: A broken right femur is preserved

among the materials numbered A.M.N.H.No. 3159, and in addition there is a part ofthe distal end of another femur.The femur of Sebecus is on the whole

characteristically crocodilian. It is an elon-gated and rather slender bone, with the headof the femur set at somewhat of an angle tothe shaft. On the ventral surface of the bonein its proximal region there is a faint trochan-teric ridge, not nearly so pronounced as thetrochanter in the modern Eusuchia. Proxi-mally and in its mid-region the shaft of thefemur is laterally compressed, so that in crosssection its greater dimension is vertical, its

lesser dimension horizontal. Distally thebone expands, and terminates in two distinctcondyles, separated from each other by awell-defined fossa.

FIBULA: In addition to the fragments offemora preserved, there is a small singlefragment that has been identified as the

FIG. 23. Sebecus icaeorhinus Simpson. A.M.N.H.NO. 3159, vertebra and limb bones, X'. A,Centrum of cervical vertebra, anterior view; B,the same, left lateral view; C, right femur, an-terior view; D, distal end of fibula, external lateralview.

distal end of the fibula. The small section ofshaft preserved is approximately triangularin cross section, with one flat surface directedmedially towards the tibia, and the other twosurfaces facing anteriorly and posteriorly,respectively. At its termination the bone isexpanded with an articular surface for thecalcaneum or- fibulare. The bone appears tobe less expanded, with the shaft more angularand less rounded than is the same bone in themodern eusuchians.

THE BRAINDescriptions and comparisons of the known

osteological elements in Sebecus having beenpresented, it may be of some value to con-sider so far as possible and as determined

from an interpretation of the skull bonessome points regarding the soft anatomy inthis extinct crocodilian. It is possible to-discuss to a limited extent the brain of

2491946


Sebecus (as interpreted from an endocranialcast), the Eustachian tubes, and the myologyof the head. These subjects will be consideredbelow.With the prefrontals and frontal, the

parietal, one postoptic, one exoccipital, and apartially preserved basisphenoid, it has beenpossible to obtain a partial endocranial castof Sebecus, a difficult task most skilfullycarried out by Mr. Otto Falkenbach of theAmerican Museum Paleontological Labora-tory. This endocranial cast of Sebecus, usedin conjunction with a very fine endocranialcast of a recent Crocodylus (also made by Mr.Falkenbach), affords some opportunity formaking an interpretation of the brain in theextinct genus and for comparing it with thebrain in recent crocodilians.The brain in Sebecus is, generally speaking,

very similar to the brain in Crocodylus. Thereis a slight difference in the flexure of the brainin the two genera, because of the relativelylower position of the medulla oblongata inSebecus, as compared with Crocodylus, whichis in turn a reflection of the relatively deepskull in the extinct genus.

It would appear that the brain of Sebecusshows some difference from that of Crocodylusin the rather high position of the upper sur-face of the cerebellar region, again a reflectionof the deep skull in the fossil form. An exactcomparison is not possible in this respect,however, because of the lack of the supra-occipital bone in Sebecus.

In Sebecus, as in Crocodylus, there is a longolfactory tract running forward from thecerebral hemispheres and terminating ante-riorly in the well-developed olfactory bulbs.In this regard it might be said that the shapeof the olfactory bulbs is more clearly definedon the under surface of the frontal bones inSebecus than is the case in Crocodylus.

In the recent crocodilians the cerebralhemispheres are marked by their shortnessand breadth, the two of them being greaterin combined breadth than they are in length.In Sebecus the cerebral hemispheres are per-haps somewhat narrower and longer thanthey are in Crocodylus or Alligator, and theytaper more gradually into the olfactorystalk than they do in the recent crocodilians.Moreover, the expansion of the temporal lobeof the cerebrum in the extinct genus is lessthan in modern crocodilians, and all in allone gets the impression that the forebrain

was enlarged to a lesser degree in Sebecusthan it is in Crocodylus or Alligator. Naturally,this is what would be expected in a compari-son of an Eocene crocodilian with a recentform, although it must be pointed out thatthe difference is not great, since these animalsare representative of a phylogenetic line inwhich evolutionary change was compara-tively slow.At the base of the cerebral portion of the

brain is the large and well-developed pituitarybody, projecting posteroventrally beneath themid-brain. This part of the brain is housedwithin the sella turcica in the basisphenoidbone, and its form in the Crocodilia is quitedistinct. In Sebecus, it is somewhat elongated,and it tapers posteriorly from its roundedand rather thick anterior region.As may be seen in the basisphenoid bone,

the internal carotid arteries traversed thesella turcica in Sebecus, as they do in theeusuchians. A dissection of a recent crocodil-ian will show that on either side the artery,emerging from the tympanic cavity, entersthe sella turcica from behind by a foramen,and leaves the sella turcica anteriorly by itsenlarged anterior opening.

Behind that portion of the endocranial castrepresenting the cerebrum there is a drop inthe upper profile and a lateral constriction,and this represents the region of the corporabigemina or optic lobes. Evidently this part ofthe brain in Sebecus was very similar in de-velopment to the same region in the moderncrocodilian brain.

Behind the optic lobes the upper profile ofthe endocranial cast of Sebecus rises again, asmentioned above, possibly an indication of anenlarged cerebellum in this genus. This partof the cast then descends very steeply to themedulla oblongata. In the anterior part ofthe medullar region is the large stalk of thecomplex trigeminus nerve, which passed fromthe brain to various parts of the head by wayof the foramen ovale. The other nerves shownin the endocranial cast of Sebecus are thethird or oculomotor, beneath the posteriorpart of the cerebrum, the sixth or abducens,projecting forward from the base of the brainon either side of the pituitary body, and thelarge hypoglossal, at the posterior end of themedulla.The endocranial cast of Sebecus, as com-

pared with that of Crocodylus, is shown inplates 14 and 15.

250 VOL. 87


THE EUSTACHIAN TUBES

The recent Crocodilia are characterized,among other things, by an unusual com-plexity in the development of their Eusta-chian tubes. In fact, the development of thetubes is advanced over anything seen in anyother vertebrate.The exoccipital and quadrate bones, the

basisphenoid, and the basioccipital of thesupplementary specimen of Sebecus all in-dicate that there was a development in thisextinct genus similar to that in the recentcrocodilians. Since the specialization of theEustachian tubes in the recent crocodiles isof very great interest and generally not wellknown, it will be described briefly at thisplace. This description will serve to show inan approximate manner the degree to whichthe Eustachian tubes were specialized inSebecus.

In the Crocodilia there is a ventrallydirected medial opening in the skull, betweenthe occipital condyle and the posterior naresand at the junction of the basioccipital andbasisphenoid bones. This was first shown byOwen in 1850 to be the exit for a medialEustachian tube. On either side of thisforamen there is a very small foramen, alsobetween the basioccipital and the basi-sphenoid, for the lateral Eustachian tubes-seemingly the homologues of the Eustachiantubes in other tetrapods.The two lateral tubes join the median tube

at a point just below the median foramen,and there is a single common opening into thethroat behind the posterior nares.The median Eustachian tube passes up-

ward through the basioccipital bone, and atsome distance up from the foramen it dividesinto an anterior and a posterior branch. Theanterior branch of the median tube continuesforward into the basisphenoid bone, and inturn divides into two lateral branches, andeach of these branches enters an enlargementof the lateral Eustachian tube which Owen

designated as the rhomboidal sinus. Fromthe rhomboidal sinus a single passage con-nects to the tympanic cavity. In a like man-ner, the posterior median branch divides intolateral branches, each of which enters therhomboidal sinus on either side.

This complex system of Eustachian tubesis shown by plate 14.

In the material of Sebecus at hand a portionof the median Eustachian tube can be seenin the basioccipital bone, and in addition theventral openings for the lateral tubes. Also,in the exoccipital is the cavity in which therhomboidal sinus was located. From theseevidences of structure in the extinct form,and in addition from the large size of thetympanic cavity, it is reasonable to supposethat the development of the complex systemof Eustachian tubes was essentially the samein Sebecus as it is in the recent crocodilians,as described above.

It is not in the least surprising to find theEustachian tubes complicated in Sebecus asthey are in the modern crocodilians, becausethe evidence indicates that this unusualspecialization was established very early inthe history of the order. Thus the system ofmedian and lateral Eustachian tubes, pre-sumably with intercommunicating tubesjoining them, is to be seen in the Jurassicteleosaurs and thalattosuchians as describedby Andrews. Whether this specialization ofthe Eustachian system had already takenplace in the Triassic ancestral Protosuchia isnot determinable in the present stage ofpreparation of the unique specimen ofProtosuchus. Be that as it may, the fact is thatthe complex Eustachian system is certainlyan ordinal character for all of the post-Triassic crocodilians, and its presence inSebecus is one more character in the long listof anatomical features that mark this genusas thoroughly crocodilian, in spite of itsaberrant proportions and appearance.

MYOLOGY OF THE HEAD OF SEBECUS

GENERAL REMARKSThe restoration of the muscles in extinct

vertebrates is at best a difficult task, fraughtwith many chances for error. We know thatin modern vertebrates the origins and in-

sertions of muscles are variously adapted, sothat even in closely related genera the rela-tionships of the same muscle may differ con-siderably. Moreover, the adaptations thatcause these differences in the relationships of

1946 251


muscles to bones may not be apparent in thebony structure, so that frequently the osteol-ogy of an animal is not entirely a safe guide asto its myology. Therefore, when we deal withextinct forms, especially those forms that areunlike any persisting related forms, the diffi-culties of interpreting the muscular relation-ships are thereby greatly augmented.

Sebecus is a crocodilian which, despite itsclear crocodilian affinities, is quite differentfrom any persisting member of the order.This fact points to the difficulties to be en-

countered in restoring the musculature of thisextinct animal. On the other hand, in theirdevelopment the individual skull bones,despite differences in proportions, show somany resemblances to the same bones in thepersisting Eusuchia that it is not unreason-able to expect the muscles in the living formsto be fairly accurate guides as to the develop-ment of the same muscles in the extinct genus.Upon this premise an attempt has been madeto restore the musculature of the head inSebecus as described and figured blelow.

In a discussion of the muscles of theCrocodilia, or of any other reptiles for thatmatter, the problem of nomenclature loomslarge. Various names have been applied bydifferent authors to the jaw muscles in thereptiles, with the result that there is con-

siderable confusion and no very great uni-

formity to be found among the several worksthat bear upon this subject. The earlierauthors used names that had been applied tomammalian muscles; subsequently there wasa tendency to apply new names to the musclesin the lower vertebrates-names that wouldexpress functions rather than any assumedhomologies with like muscles in the mammals.Thus, in 1919 Adams, in his review of thephylogeny of the jaw muscles and followingprecedents set by other authors, used theterm adductor mandibulae for the musclesclosing the jaws in the fishes and innervatedby the third branch of the fifth cranial nerve.In the amphibians, reptiles, and birds thesemuscles were termed the capiti-mandibularisand pterygoideus muscles, which in themammals became the masseter, temporalis,pterygoideus, and certain other small muscles.

In 1926 Lakjer, in a very thorough reviewof the jaw musculature of the Sauropsida,proposed a system of nomenclature basedupon functions, and more or less uniformthroughout the group. This system was fol-lowed by Anderson in 1936, in his comparisonof the jaw musculature of the extinct phyto-saur, Machaeroprosopus, with certain modernreptiles, particularly the alligator and Spheno-don.

In the present work both Lakjer's andAdams' terms are used in discussing the jaw

NOMENCLATURE OF JAW MUSCLES IN THE CROCODILIA

EDGE-WORTH, ADAMS, 1919 LAKJER, 1926; ANDERSON, 19361907

Innervated by the fifth cranial nerve

superficialis JsuperficialisCapiti-mandibularis Adductor mandibulae externus medialis

medius profundusTemporal profundus Adductor mandibulae posterior

Capiti-mandibularis (in part) Pseudotemporalisanterior Pterygoideus dorsalis

Pterygoid Pterygoideus Adductor mandibulae internustposterior | Pterygoideus ventralis

1Intramandibularis

Innervated by the seventh cranial nerve

Depressor mandibulae [Depressor mandibulae]

252 VOL. 87


muscles of Sebecus as compared with the samemuscles in modern crocodiles. It is felt thatby doing this there will be less chance for con-fusion than if either system were used alone,with no attempt at correlation with the othersystem of nomenclature. To make the namesof the jaw muscles readily understandable,perhaps it may be well to include a table atthis place which shows the synonymies ofmuscle names in the Crocodilia, as applied byseveral authors.

THE JAW MUSCLES

Adductor mandibulae externus LakjerCapiti-mandibularis superficialis and

medius AdamsAdductor mandibulae externus superficialisALLIGATOR: This outer slip of the temporal

muscle mass arises along the outer edge of thequadrate between the jaw articulation andthe postorbital and, descending as a thinsheet, inserts on the dorsal face of the sur-angular.

SEBECUS: There is every reason to thinkthat the relationships and the developmentof this muscle in Sebecus were essentially thesame as in Alligator. In Sebecus, however,this muscle was proportionately longer thanthe same muscle in the alligator, because ofthe deep skull and jaw of the extinct genus.

Adductor mandibulae externus medialis

ALLIGATOR: According to Lakjer (1926)and to Anderson (1936) the medial portionof the temporal muscle mass has two heads,an anterior one arising from the dorsal por-tion of the medioventral face of the post-orbital and a posterior head arising from theanterior part of the rostral face of the quad-rate, deep to the superficial slip, describedabove. These two heads join to form a thinplate of muscle which is inserted upon thedorsal face of the surangular, underneath thesuperficial slip of the muscle.SEBECUS: It would seem probable that this

muscle slip had the same relationships inSebecus as in the alligator. Again, however,this muscle must have been long, by reasonof the deep skull of the extinct crocodilian.

Adductor mandibulae externus profundusALLIGATOR: Anderson has described this

muscle in detail, showing how in the alligatorit consists of four portions, located deep tothe medialis slip. There is a rostral slip, theorigin of which is on the postoptic behind theeye, and the fibers of this muscle are directedventrally to insert on the intermusculartendon of what Lakjer had termed the in-tramandibularis muscle. Of more importancein the alligator are the middle slips of thismuscle, which originate upon the quadrate,squamosal, and parietal within the supra-temporal fossa and are inserted in part uponthe intermuscular tendon of the intramandi-bularis muscle and in part upon the medialface of the surangular. Finally there is acaudal slip originating upon the quadrate,deep to the medialis slip, and inserting uponthe medial face of the angular and surangular,in front of the articular.SEBECUS: In Sebecus the bones on the

posterior and internal border of the supra-temporal fossa rise abruptly from the level ofthe frontal, as already described. Specificallythese are the parietal and squamosal bones,elevated into a sort of crest at the back of theskull. Moreover, there is a strong crest or edgeformed along the back and inner side of thefossa, where it is bounded by the parietal andsquamosal bones, and there cannot be muchdoubt that this raised rim afforded a partic-ularly strong and expanded origin for theprofundus muscles. In addition, the supra-temporal fossa in Sebecus is relatively largerthan it is in Alligator and many other recenteusuchians. These correlated developmentspoint to the presence of a rather large andstrong profundus muscle mass in Sebecus. Inthis connection it may be recalled that theinner wall of the mandibular ramus, formedby the splenial and the coronoid, extendsback to a much lesser degree in Sebecus thanit does in the modern eusuchians. Thus, theprofundus group of muscles, inserting in thesuprameckelian fossa, would have more roomin which to bulge in Sebecus than is the casein the modern crocodilians, and this againis an indication that these muscles in Sebecusmay have been stronger than the same mus-cles in the Eusuchia.

1946 253


Adductor mandibulae posterior LakjerCapiti-mandibularis profundus Adams

ALLIGATOR: This is an independent musclein the Crocodilia, located entirely behindthe mandibular ramus of the trigeminalnerve. The origin is on the deep portion ofthe quadrate behind the foramen ovale, andthe insertion is on the lower surface of theangular along the posterior part of the lateralmandibular fenestra and on the inner wall of

origin on the postoptic, just in front of theforamen ovale, and it is inserted between thetwo layers of the adductor mandibulae poste-rior, described above. According to Andersonthe functions of these two muscles are closelyrelated.SEBECUS: One can only assume that the

same muscle was present in Sebecus.Pterygoideus

Modern crocodilians are characterized by

Im AmpFIG. 24. Diagrammatic reconstruction of the jaw muscles in Sebecus icaeorhinus. Amem, Adductor

mandibulae externus medialis; Amep-a, adductor mandibulae externus profundus, rostral and middleslips; Amep-b, adductor mandibulae externus profundus, caudal slip; Ames, adductor mandibulaeexternus superficialis; Amp, adductor mandibulae posterior; Dm, depressor mandibulae; Im, intra-mandibularis; Pd, pterygoideus dorsalis.

the angular, medial to this fenestra. Themuscle is split in its lower portion by the in-sertion of the pseudotemporalis.SEBECUS: The origin and insertion of this

muscle in Sebecus must have been essentiallythe same as in the modern crocodiles. Hereagain, however, the muscle was elongated inthe fossil form, and its direction must havebeen more nearly vertical than it is in Al-ligator. In this latter genus the muscle isinclined at an angle of about 45 degrees,whereas in Sebecus the angle was more on theorder of 60 degrees.

Adductor mandibulae internus LakjerPseudotemporalis

Capiti-mandibularis (in part) AdamsALLIGATOR: The pseudotemporalis has its

the very great development of the pterygoidmuscles, the anterior slips of which extendfar forward into the skull, while the posteriorslips are so greatly enlarged as to form swell-ings on the back of the lower jaws. As Ander-son has shown, this muscle is separable intofour heads in the alligator, one of which isanterior in position, and three posterior.

Pterygoideus dorsalis LakjerPterygoideus anterior Adams

ALLIGATOR: The crocodilians are charac-terized by the forward extension of theanterior slip of the pterygoid complex, sothat it is carried far forward within themuzzle of the skull. In the alligator thismuscle arises beneath and in front of theorbit, as far forward as the posterior wall of

VOL. 87254

1 COLBERT: SEBECUS, A PECULIAR FOSSIL CROCODILE

the olfactory capsule, and it is attached overa broad area to the dorsal surface of thepalatal plate of the maxilla, to the edge ofthe palatine and to the upper surface of thepterygoid above the nasal passage. It is in-serted on the medial surface of the angularbelow and in front of the articulation. Thereis also a tendon extending to the intermuscu-lar tendon of the intramandibularis.SEBECUS: The evidence would seem to

indicate that the relationships of the anterior

the postarticular process. In its posteriorextensions this muscle becomes very fleshyand forms a large bulge on the posteroventralportion of the lower jaw. The second part ofthe posterior pterygoid, "B" of Anderson,originates along the posterior margin of thepterygoid bone and is inserted on the ventro-medial faces of the angular and articular. Thethird portion of this muscle group arises fromthe medial posterior edge of the pterygoidbone and from the basisphenoid and is in-

FIG. 25. Diagrammatic reconstruction of the jaw musclesin Sebecus icaeorhinus. Pva, Pterygoideus ventralis, portion"A"; Pvb, Pvc, pterygoideus ventralis, portions "B" andSICOP.

pterygoid muscle in Sebecus were essentiallysimilar to those existing in the moderncrocodilians. The internal surface of themaxilla in this extinct form shows that thecartilaginous posterior wall of the olfactorycapsule met the inner surface of the maxillaalong a line above the space between thefourth and fifth maxillary teeth. This thendetermines the forward limits of the originfor this muscle.

Pterygoideus ventralis LakjerPterygoideus posterior Adams

ALLIGATOR: There are three parts to thisdivision of the pterygoid group in the al-ligator, and these have been designated byAnderson as "A," "B," and "C." Portion A,arising by a tendon from the ventral tip ofthe ectopterygoid, wraps around the poste-rior end of the lower jaw to insert on thelateral faces of the angular and surangular on

serted upon the posterior portion of the post-articular process.SEBECUS: Here again the relationships of

the muscles in Sebecus were probably verysimilar to what they are in the modern al-ligator. Consequently, so far as may be deter-mined, the above description for the alligatormay be taken as indicating essentially theconditions that probably held for Sebecus.

IntramandibularisThis muscle, described by Anderson as a

separate entity, was not differentiated fromthe capiti-mandibularis by Adams (1915). Itis distinguished as a separate muscle byLakjer because it lies medial to the mandib-ular ramus of the trigeminal nerve. Ander-son makes the following remarks concerningthe intramandibularis muscle:

"This muscle may be regarded as anadaptation of the adductors of the mandible

1946 255


to the elongation of the lower jaw. Withelongation of the jaw mechanical advantagemust be gained by increasing the distancebetween the application of the force (muscleinsertion) and the fulcrum (jaw articulation).Hence we find the muscle insertions withinthe primordial canal of the mandible mi-grating forward. As such insertions proceedanteriorly to underlie the dentigrous part ofthe jaw, a tendinous area must developwithin the muscle to ride over the coronoidat the anterior end of the aditus of theprimordial canal. Thus the muscle within thejaw becomes a separate muscle arising fromthis tendon. The insertion upon this tendonof muscles of the internal and external ad-ductor groups permits both to take advantageof the anterior insertion gained by the m.

intramandibularis" (Anderson, 1936, p. 565).ALLIGATOR: This muscle arises from the

tendon that receives the insertions of therostral and middle slips of the adductormandibulae externus profundus, as well as a

slip from the anterior pterygoid. The tendonrides over the coronoid, and the muscle in-serts in the meckelian fossa of the mandible,extending forward beneath the lower teeth.SEBECUS: It is probable that much the

same relationships held for Sebecus that areseen in the modern alligator. Because of theanterior position of the coronoid in the fossilform, the tendinous connection between theintramandibular muscle and the temporalmuscles above it may have been placed some-

what farther forward than it is in Alligator.

Depressor mandibulae

This single muscle opposes all of the mus-

cles previously described. It is the greatdisparity of strength between the powerfuladductor muscles of the jaws and the singledepressor that accounts for the fact that inthe crocodilians the power to open the jawsis relatively slight, even though their closingpower is tremendous.ALLIGATOR: The depressor mandibulae

arises for the most part on the occipital por-

tion of the squamosal and is inserted upon thepostarticular process of the mandible.SEBECUS: Similarly, in Sebecus the de-

pressor mandibulae muscle ran from that partof the squamosal on the back of the skull to

the postarticular process of the mandible. Itis interesting to see that even though thepostarticular process in Sebecus is directedstrongly upward so that it extends high abovethe glenoid (in comparison with the almosthorizontal process in the alligator), the direc-tion of the muscle in the two genera is ap-proximately the same. This is owing to thefact that the skull in Sebecus is so deep thatthe origin of the muscle is comparativelymuch higher than it is in Alligator. In otherwords, both the origin and the insertion of themuscle have been raised in the fossil form,with the net result that the direction andaction of the muscle must have been aboutthe same as they are in the persisting Eu-suchia.

THE NECK MUSCLESIt is to be assumed that the neck muscles

of Sebecus were essentially similar in theirorigins and insertions, relationships andfunctions to the muscles of modern crocodil-ians. Unfortunately, since no cervical verte-brae are completely preserved, it probably isnot feasible to attempt restorations of themuscles connecting the occiput with theanterior vertebrae. However, some attemptto restore the areas of insertions for the mus-cles leading to the occiput may be in order,and this will be described below.

In the modem crocodilians the occiput islow and wide, and the spines of the cervicalvertebrae are correspondingly low. Therefore,one is led to conclude that in Sebecus, with arather high occiput, the spines of the cervicalvertebrae may have been somewhat elon-gated. This in turn leads to the suppositionthat there were relatively increased areas forthe origins and insertions of the neck musclesin the extinct genus, as compared withmodem eusuchians, and therefore relativelystrong muscles in the fossil form. Indeed, onecharacter whereby Sebecus may be comparedwith modem crocodilians is the compara-tively expanded occiput to be seen in theextinct genus.

For instance, as seen in occipital view, theocciput of Sebecus (the area included by theoccipital surfaces of the supraoccipital,squamosal, exoccipital, and basioccipitalbones) has an area that is approximately 30per cent greater than the same area in an

VOL. 87256


alligator with a corresponding occipitalwidth. The figures are as follows:

SebecusAlligator

AREA OF OCCIPUT126 sq. cm.98 sq. cm.

The comparison is shown also by figure 26.A few remarks as to the probable insertions

of muscles on the occiput of Sebecus follow.The occipital muscles of the modern alligator,

FIG. 26. Comparison of the relative areas ofthe occiput in Sebecus (heavy line) and Alligator(light line).

as described by Anderson, are used as aguide to the restoration of the insertions inSebecus, although Anderson's terminology isnot used in this connection.

Depressor mandibulaeThis, the muscle for closing the jaws, has

already been described. In Alligator theorigin for the muscle is on the occipital sur-face of the squamosal, whence it stretches tothe postarticular process of the mandible.The same relationships undoubtedly held inSebecus. Consequently the muscle in Sebecusmay have had a broad, low origin, althoughit is quite possible that in the fossil it wasconfined to the inner part of the squamosal,since the external portion of this bone runsalmost parallel to the course of the muscle.

Rectus capitis posteriorIt is possible that there were two insertions

of this muscle on the occiput of Sebecus, asuperficialis part on the supraoccipital bone,medial to the depressor mandibulae, and aprofundus part on the exoccipital, below thedepressor mandibulae. In Alligator this is abroad muscle inserting on both the exoccipitaland the supraoccipital, below and somewhatmedian to the depressor mandibulae.

Obliquus capitis magnusIn Sebecus this muscle probably was in-

serted for the most part on the exoccipital,as it is in Alligator, in part deep, and in partlateral to the rectus capitis muscle.

Longissimus capitisIt is probable that there were two insertions

on the occiput in Sebecus, a transversaliscapitis part, to the external tip of the ex-occipital, lateral to the rectus capitis posteriorand the obliquus capitis, and a more ventraltransversalis cervicis portion, to the lowerborder of the exoccipital and possibly in partto the basioccipital. Muscles showing theserelationships are to be seen in the alligator.

Rectus capitis anteriorThis muscle, to judge by the alligator, was

inserted in Sebecus by a tendon to the ex-panded tubercles at the suture between thebasioccipital and basisphenoid bones.

Of the above neck muscles inserting uponthe skull, the rectus capitis posterior groupserved for extension of the head, the rectuscapitis anterior and the longissimus capitispars transversalis cervicis for flexion of thehead, and the obliquus capitis magnus andlongissimus capitis pars transversalis capitisfor rotation and abduction of the head.

FIG. 27. Diagrammatic reconstruction of theareas of insertion of the neck muscles and onejaw muscle on the occiput of Sebecus. Dm, De-pressor mandibulae; Lc-a, longissimus capitis,transversalis capitis portion; Lc-b, longissimuscapitis, transversalis cervicis portion; Oc, obliquuscapitis magnus; Rca, rectus capitis anterior;Rcpp, rectus capitis posterior, profundus portion;Rcps, rectus capitis posterior, superficialis portion.

2571946


t_

_

<

In

u

r-4'-n' r

C: c

o

84.

43

03

4. A. aV 0.

'4)bJO 0

ox Q

... 4

0.4

. Q o

- C0 D

C)

I.~ ~ ~~~c

43:o'a cd 0 W

00

0 0)(I)

.0

*d U

8.4 Cd4-0) U)

0

U)

U c

o@Qi

CCw

~ ~ ~

~

bOO0i)

4C3

Cd

.404

0

v Cd 4=

'4-J

U1 ) L).

8rz.U rj ..

258 VOL. 87

SKULL PROPORTIONS OF SEBECUS AND OFOTHER CROCODILIANS

GENERAL REMARKS

Sebecus HAS ALREADY BEEN COMPARED mor-phologically with other crocodilians, and thiscomparison has covered the bones of theskull and jaw, the teeth, the postcranial ele-

Mid-rostral secti6n

Ant.

once distinguish Sebecus, Baurusuchus, andby inference the Sebecosuchia, from all ofthe Protosuchia (known from the singlegenus Protosuchus), Mesosuchia, Thalat-

Lateral aspect

Orbit Post.

Protosuchus

Alligator

Crocodilus

SebecusZNFIG. 29. Diagrammatic comparison of the general shapes of the

skulls of several crocodilians, in mid-rostral sections and lateralaspects. Length of skulls reduced to unity, and other measurements inproportion. This figure gives a comparative diagrammatic impressionof the depth of the skull as related to its width in the mid-rostral re-gion, of the depth of the skull anteriorly, posteriorly, and at the orbit,and the position of the orbit, when the length of the skull is reducedto unity. Protosuchus (Protosuchia), Sebecus (Sebecosuchia), Croco-dylus, and Alligator (Eusuchia).

ments, so far as they are known, and certainsoft structures as interpreted by the osteol-ogy. It has already been emphasized thatSebecus is distinct from all other crocodiliansexcept Baurusuchus because of its high, nar-row skull and jaw, and its laterally com-pressed maxillary and posterior dentaryteeth. These peculiarities of proportion at

tosuchia, and Eusuchia. The proportionaldifferences between the skull and jaws inSebecus and in other crocodilians are broughtout in the accompanying tables of measure-ments and ratios and indices (tables 1 and 2),and they may be still further emphasized bysome simple diagrams.

This is done by reducing the outlines of the259

II

iI

0.. ..,

I

iC... -.

Ir 1-


skull in various crocodilians, as seen in thedorsal, lateral, and occipital views, to simplelines. In this way minor structures that tendto confuse the general picture of skull pro-portions are eliminated and thereby a broadcomparison of the skulls is facilitated.

long, narrow, and deep preorbital part of theskull that makes Sebecus so distinctive.

Sebecus is noteworthy also because of itscomparatively large cranial table, extendingalmost the full width of the skull. Protosuchus,interestingly enough, has a very large, broad

TABLE 1COMPARATIVE MEASUREMENTS (IN MILLIMETERS)

-~~~~~~~~~~~~~~~~~~~~~~~~PooSebecus Croco- Alligator Steneo- Procouicaeo- dylus mississip- saurus richard-

Measurements rhinus niloticus iensis obtusidensA.M.N.H. A.M.N.H. A.M.N.H. (From A.M.N.H.No. 3160 No. 23466 No. 7141 Andrews) No. 3024

A. Length of skull: premaxilla to quadrates 476 414 342 1206 115B. Length of snout: premaxilla to anterior rim of

orbits 291 260 209 720 49C. Breadth of skull: across quadratojugal 190 185 159 402 80D. Breadth of snout: in plane of anterior orbital

rims 125 130 139 204 36E. Breadth of snout: midway between tip and

occiput 87 94 129 112 40F. Breadth of snout: maximum on premaxilla 54 72 89 102 26G. Breadth of cranial table: maximum (poste-

rior) 160 89 90 71H. Breadth of cranial table: minimum (anterior) 140 80 80 56I. Interorbital breadth: minimum 47 40 20 84 30J. Anteroposterior length of orbit 60 51 57 84 21K. Height of orbit (or width) 65 35 45 -L. Height of snout: vertically from alveolus of

last tooth 104 53 36 - 24 (e)M. Height of snout: vertically from alveolus of

last premaxillary tooth 55 25 21 16N. Height of snout: vertically midway between

tip and occiput 101 44 32 230. Total length of tooth row 274 272 214 720P. Length of mandible: maximum 526 451 377 1308 109Q. Length of mandible: to articulation 464 401 337 1170 109R. Depth of ramus: vertically from lower border

to articular 97 65 56 - 15S. Length of tooth row 252 246 188

Dorsally, the skull of Sebecus is noteworthy,of course, because of its lateral compression.For instance, the skull in this genus is notice-ably narrower than is the skull in Crocodylus-a narrow-skulled eusuchian. On the otherhand, the skull of Steneosaurus, a meso-suchian, is proportionately narrower than theskull of Sebecus. But in Steneosaurus the orbitis placed relatively far forward, and the skullis generally low. It is the combination of the

cranial table, but in the eusuchians thecranial table is generally much reduced inrelative size. In many mesosuchians, such asthe one chosen for comparison, the uppertemporal openings are so large that a cranialtable is difficult to define.

In an occipital view, the skull of Sebecus isdistinguished again by its depth and by itsstraight sides. Depth of the skull is also adistinctive character in a lateral view of the

VOL. 87260


Sebecus skull. As may be seen, the depthextends from the back well forward on themuzzle, and is greatest at the front of theorbit. This may be contrasted with the

general crocodilian condition, in which themuzzle is shallow, and the postorbital part ofthe skull is the region of greatest depth.

TABLE 2COMPARATIVE RATIOS AND INDICES

Skull width 100 CSkull length A

Snout length 100 BSkull length A

Snout breadth 100 DSnout length B

Snout height 100 LSnout width D

Snout height (anterior) 100 MSnout height (posterior) L

Interorbital breadthSkull breadth

100 IC

Tooth row 100 0Skull length A

Tooth row 100 SJaw length Q

Skull height 100 LSkull length A

Skull cranial table 100 GSkull width C

Sebecus Crocodylus Alligator Steneosaurus Protosuchus

40

61

43

83

53

25

58

54

22

45

63

50

41

47

22

66

61

13

47

61

67

26

58

13

63

56

10

33

60

28

21

60

70

43

73

67

67

37

21

48

2611946

84 57 89

SEBECUS AND OTHER FOSSIL CROCODILESFROM PATAGONIA

.!6*--tOF THE CROCODILIANS CONTEMPORANEOUSwith Sebecus, there is a single described form,Eocaiman cavernensis Simpson, which likeSebecus is found in the Casamayor formation.This is the common crocodilian in theCasamayor, and, as Simpson has shown, itseems to stand near the ancestry of the SouthAmerican eusuchians, Caiman and Jacare. Ingeneral, the skull bones and the teeth of thisshort-faced alligatoroid are hardly to be con-fused with the laterally compressed bones andteeth of Sebecus. Indeed, all of the differencesthat have been cited above that distinguishSebecus from AUigator also separate it fromEocasiman.Simpson has also described a somewhat

earlier form, Necrosuchus ionensis, fromPatagonian sediments of probable Paleoceneage. This is an eusuchian belonging to theCrocodilidae and as such is readily dis-

tinguishable from Sebecus. Two still oldergenera are the mesosuchians Notosuchus andCynodontosuchus, first described by SmithWoodward in 1896. Notosuchus, a short-faced and almost edentulous mesosuchian,cannot be mistaken for Sebecus, althoughthere is some possibility that the crocodilianbones and the single tooth figured byAmeghino in 1906 as belonging to Notosuchusterrestris actually are referable to Sebecus.Cynodontosuchus is based upon imperfectevidence, but it certainly is distinct fromSebecus.

Other Patagonian crocodiles are Sympto-suchus contortidens, described by Ameghinoupon the basis of insufficient material andtherefore virtually indefinable, and certainrather indeterminate fragments, describedby Kuhn in 1933.

262

THE PHYLOGENETIC POSITION OF SEBECUS

Sebecus AND THE RECENTLY DESCRIBED Bra- From the Protosuchia the later Mesozoiczilian genus Baurusuchus are crocodilians apcodilia originated. It would seem prob-quite distinct from all the other members of Sle that the derivation of more specializedthe Order Crocodilia, and for this reason they forms from this ancestral group showed anare placed in a separate suborder, the Sebeco- early Jurassic division into what might besuchia. Consequently the Order Crocodilia is termed the "central" line of the Mesosuchiahere regarded as being composed of five sub- and the more "lateral" line of the highlyorders, as follows: specialized Thalattosuchia. Both of these

Order Crocodilia groups were marked by a specialization inSuborder Protosuchia the skull whereby the internal nares wereSuborder Mesosuchia carried far back, with a sort of secondarySuborder Thalattosuchia palate formed by the maxillae and palatinesSuborder Sebecosuchia separating the nasal passage from the buc-Suborder Eusuchia cal cavity. On the other hand, the two sub-

orders may be contrasted by the great varietyThe Protosuchia, typified by the single in adaptive radiation among the Mesosuchia

genus Protosuchus, which is in turn known as compared with the rather straight, "ortho-from a unique, but nevertheless well-preserved genetic" trend among the Thalattosuchia forskeleton, already show numerous definite life in marine waters.crocodilian characters, in spite of the early The Mesosuchia show long-snouted andand general position of this group within the short-snouted forms, with consequent multi-phylogeny of the Crocodilia. For instance, plications and reductions in the dentition; inthe skull in Protosuchus is on the whole some the supratemporal fenestrae are veryprimitive and not particularly crocodilian in large while in others they are much reduced;appearance, yet it shows a reduction of the some are large and others quite small. Yetsuperior temporal fenestrae and a develop- in spite of the great variety shown by thement of a cranial table, formed by the post- development of the skull, the dentition, andorbital, squamosal, supraoccipital, parietal, the size of body in the Mesosuchia, there is aand frontal bones, that is remarkable because comparatively uniform trend in the habitusof its similarity to the cranial table in the of the group. These were all shallow-waterhighly specialized eusuchians. In other re- crocodilians, adapted to life along the banksspects, too, Protosuchus shows its crocodilian of inland streams or the shores of largerrelationships, particularly in the elongation bodies of water. Although thoroughly aquaticof the carpal bones and the virtual exclusion the mesosuchians must have spent a greatof the pubis from the acetabulum. deal of time on dry land-but ever in closeOn the basis of our present knowledge, proximity to the water-as do the modern

Protosuchus must serve as the structural crocodiles.ancestor for all of the Crocodilia, and it is As compared with this mode of life, theprobable that this genus approaches rather Thalattosuchia show an early and a veryclosely the actual ancestry of the order. Un- strong trend to adaptations for open, marinefortunately a detailed and comprehensive waters. In these crocodilians the armor andstudy of Protosuchus has not been as yet claws so characteristic of other crocodiliansmade, and our final judgment as to the have been lost, while the bones of the skullposition this genus occupies in the phylo- are secondarily smooth, in contrast to thegenetic history of the Crocodilia must of typically pitted external surfaces in the skullnecessity wait until such a study has been bones of other crocodilians. The thalat-published. However, it seems evident upon tosuchians also show the development ofthe knowledge furnished by Protosuchus that sclerotic plates (a common adaptation inthe Crocodilia, having originated from theco- marine reptiles), a transformation of the feetdont ancestors, had become established as a into paddles, and a downturning of the caudalseparate group of ordinal rank by late vertebrae to form a reversed heterocercal tail.Triassic or early Jurassic times. It is probable that while the mesosuchians

263


and thalattosuchians were following theirseparate trends of evolutionary specializa-tions through the Jurassic period, the sebeco-suchians were at the same time evolvingfrom a protosuchian stem. Thus the speciali-zations of the Mesosuchia, the Sebecosuchia,and the Thalattosuchia in the Jurassic periodmay be regarded as the first broad adaptiveradiation of the Crocodilia.What may be regarded as a second period

of adaptive radiation among the Crocodiliaseemingly took place during Cretaceous andearly Cenozoic times. In this period there wasa continuation of the mesosuchians from thepreceding Jurassic period. Here there appearfor the first time the Sebecosuchia, probablyas a continuation from Jurassic ancestors. Inaddition to these established crocodiliangroups a new line arose, that of the Eusuchia.This line was destined to flourish duringCretaceous and early Cenozoic times and tocontinue into the latest stages of earthhistory, whereas all other lines of crocodilianevolution were to vanish.

It is known that the Eusuchia werederived during Cretaceous times from theMesosuchia through transformations, thecharacters of which are plainly apparent inthe latter suborder. Thus, there was a fur-ther retreat of the internal nares so that inthe Eusuchia they came to be located farback and completely enclosed by the ptery-goid bones. Furthermore, the Eusuchia showa complete withdrawal of the postorbital barto a subdermal position, again a developmentalready under way within the Mesosuchia.Finally, the Eusuchia are marked by theappearance of procoelous vertebrae. Fromthis, and in addition from much other evi-dence which might be cited, it is plain thatthe Mesosuchia evolved gradually into theEusuchia, and that although there werespecializations in morphology from the oneto the other, the trend in habitus was carriedon virtually without interruption. The Eu-suchia, like the Mesosuchia before them, wereshallow-water crocodiles, living in streamsand rivers, or along the shores of lakes andthe sea.From the morphological characters of

Sebecus and Baurusuchus it would seemevident that these crocodiles, and by in-ference the suborder to which they belonged,

were derived from a basic mesosuchian an-cestry. Both Sebecus and Baurusuchus showlarge internal nares bounded anteriorly bythe palatines as in the Mesosuchia, but ex-tending back into the pterygoids and boundedposteriorly by these bones. In Baurusuchusthe expanded ectopterygoids form in part thelateral borders of the internal nares. Thevertebrae in Sebecus are platycoelous. Inmany ways Sebecus shows close resemblancesto the Eusuchia in the basic structure of theskull and jaws; the reason why this genus isso different from the eusuchians is, asidefrom the structure of the internal nares andof the teeth, mainly a matter of proportions.It is therefore evident that the Sebecosuchia,although independently derived from amesosuchian stem, have paralleled the Eu-suchia in many respects. The Sebecosuchia,like the Eusuchia, must have been shallow-water crocodiles, living, so far as our presentevidence indicates, along the banks of riversand streams. Like the Eusuchia, the Sebeco-suchia must have spent a part of their life inthe water and a part of it on dry land, nearthe water. But unlike the Eusuchia or theMesosuchia, the Sebecosuchia were deep-skulled crocodilians, and in this respect theyshow either a retention of primitive charac-ters or a reversal in evolutionary trendswhereby they approximate to some degree thethecodont reptiles that were the ancestors toall of the Crocodilia.Thus the strong divergence of the Sebeco-

suchia from all other crocodiles affords aninteresting contrast to the strong divergenceof the Thalattosuchia from the Mesosuchia.For in the specialization of the Thalat-tosuchia there was a divergence not only inmorphology, but in habitat as well; theadaptations of the two suborders Thalat-tosuchia and Mesosuchia were for two en-tirely different modes of life. As contrastedwith this, the divergence of the Sebecosuchiafrom other crocodilians, although marked bydifferences in structure and by very greatdifferences in proportions, seemingly was notmarked by a particular divergence in habits.It appears that both the Sebecosuchia andthe Eusuchia followed essentially the sametypes of patterns in habits and ecology, andit is very probable that they may have beenfor a time competitors in the struggle for

264 VOL. 87


Mesosuchia, ~ vm

To

Thalatto,suchia

FIG. 30. A phylogeny of the suborders of Crocodilia.

1946 265


existence. If this is so, then it would seemthat the Sebecosuchia were the less welladapted for continuation.The presence of Baurusuchus in the Creta-

ceous and Sebecus in the Eocene of SouthAmerica leads to some interesting specula-tions. Why should these crocodilians, sodifferent from all other members of the order,have evolved? In the case of the thalat-tosuchians we can correlate the specializa-tions in the skull and body with a verydefinite change in habitat; they were marinecrocodiles and were widely spread over largegeographical areas. The sebecosuchians onthe other hand show adaptations that, whiledivergent from those seen in other crocodil-ians, are not to be correlated with anyrecognizable divergence in habitat. It is pos-sible, as suggested on a preceding page ofthis work, that Sebecus may have had habitsof hunting and eating somewhat differentfrom those of the modem crocodilians. Butgranting that such differences did exist-asupposition at best-it is still difficult toaccount for the extreme divergence in theform of the skull, jaws, and teeth of thesebecosuchians from other members of theorder.One is led to wonder if the presence of

Sebecus in the Eocene of Patagonia, andsimilarly of Baurusuchus in the Cretaceousof Brazil, may not represent the geneticisolation of phylogenetic lines in restrictedlocalities. Simpson (1944, pp. 68-69) has

suggested that mutations might be moreeffectively utilized and changes broughtabout more quickly in small, isdUtted popula-tions than is the case in large, widely spreadpopulations. Since the sebecosuchians havebeen found only as rare and isolated fossils,there is the possibility that they representelements in a phylogenetic line that had be-come isolated and was developing withinrelatively small successive populations. Ofcourse the present geographic and geologic re-striction of these crocodilians may be due en-tirely to accidents of collecting, but in view ofrelatively large numbers of eusuchians andsome mesosuchians found in other parts ofthe world, one is led to think that the restric-tion of Sebecus to Patagonia and Baurusuchusto Brazil may represent a real condition thatheld to a large degree for this crocodiliansuborder in Cretaceous and Eocene times. Ifthe Sebecosuchia were restricted to thesouthern part of South America throughouttheir history, and if they were at all timesrather sparsely distributed animals withintheir range, then the key to their widedivergence from other crocodilians may be inthese very facts. Perhaps we see here aphylogenetic line that has resulted fromgenetic isolation and the consequent cumula-tive piling up of successive mutations. If so,there is no particular necessity in trying tocorrelate the structure of the skull of Sebecustoo closely with possible eating or huntinghabits.

VOL. 87266

SEBECUS AND THE MOOTED CASAMAYOR DINOSAURS

IN 1932 SIMPSON DISCUSSED IN SOME DETAILthe long-argued question as to whether or notdinosaurs occur in the Casamayor andwhether, if so, this indicates that the Casa-mayor is Cretaceous, or that dinosaurssurvived into the Tertiary. The conclusionwas that dinosaurs are not known from theCasamayor and that this formation is ofTertiary age. This conclusion is not alteredbut is indeed strengthened by the discoveryof Sebecus. Moreover this discovery permitswhat may be the last word on this subject,at least as far as previous discoveries andclaims are concerned.Without again going into full detail, the

evidence for association of dinosaurs withmammals of Tertiary aspect in Patagoniaconsisted of a series of early discoveries byvarious hands, unskilled for the most part,and another series of somewhat later dis-coveries by Carlos Ameghino. As regards thefirst discoveries, it has been conclusivelyshown that they are not worthy of credence,either because the field data were quite un-reliable or because the interpretation placedon those data was a non sequitur. As regardsthe work of Carlos Ameghino, although hisfield data were questioned by others, Simpsonfound no reason to doubt their essential ac-curacy, but he refused to accept the identi-fication of Ameghino's specimens as dino-saurs. Some of the Ameghino specimens dis-appeared without being figured or exactlydescribed, but those remaining as evidenceappear to be sufficiently typical of all.These specimens are teeth of two sorts. One

sort is indubitably crocodilian. F. Ameghinohimself identified as crocodilian certain teethpractically indistinguishable from these, sothat his failure to correct the earlier identi-fication as dinosaurian can be imputed to anoversight. However, there remain teeth thatare very like those of carnivorous dinosaursand quite unlike those of any previouslyknown crocodilians, typified by No. 10872 inthe Ameghino Collection of the Buenos AiresMuseum. One of the best authorities ondinosaurs, von Huene (1929), agreed withAmeghino that these are dinosaur teeth, butdoubted the horizon record. Simpson con-

firmed the horizon record by subsequentrepetition of the discovery, but suggestedthat the teeth were not, or at least were notproved to be, dinosaurian because (1) theydid not agree exactly with dinosaur teethdespite the close but inconclusive resem-blance, (2) the histology is not characteristi-cally dinosaurian, and (3) dinosaur bones areinvariably more abundant than teeth indinosaur beds, and yet no such bones havebeen found at the localities or horizons fromwhich these teeth came. As shown by Simp-son in 1937, this negative conclusion is con-firmed beyond reasonable doubt by the dis-covery of Sebecus, in the same horizon. Inevery way, the tooth in the Buenos AiresMuseum (M.N.H.N. No. 10872), describedby Ameghino and refigured and discussedby Simpson (1932), corresponds with theflattened teeth of Sebecus. It is like the teethof Sebecus in size, in shape, especially in thecurved anterior edge and the relativelystraight posterior edge, in cross section, and inthe development of serrated cutting edges. Itthus appears that the mystery of the "Eo-cene dinosaurs" of South America is at lastreally solved.

In his paper of 1932 Simpson described aflattened caniniform tooth, which he con-sidered as possibly similar to the tooth de-scribed by Ameghino (M.N.H.N. No. 10872),and which he regarded as mammalian be-cause of its probable but uncertain associa-tion with mammalian cheek teeth. Whetherthe tooth that was associated with the mam-mal Florentinoameghinia mystica, describedby Simpson, should be identified with thisspecies is a question that cannot be decidedupon the basis of present evidence. What isclear is the fact that the tooth earlier de-scribed by Ameghino as of dinosaurianrelationships and now shown to be a tooth ofSebecus is not the same as the flat, cuttingtooth that Simpson provisionally identified asbelonging to Florentinoameghinia.The tooth associated with Florentino-

ameghinia, and figured by Simpson (1932,fig. 6), is considerably larger than the largestteeth in Sebecus. Moreover it is differentlyshaped, for both front and back edges are

267


curved, whereas-in Sebecus the back edge israther straight. In addition, the tooth figuredby Simpson, though crushed, evidently has amuch thinner cross section than any of theflattened teeth of Sebecus. Finally, and thisis a most important point, this tooth under amicroscope shows no trace of serrations alongits edges, whereas every tooth in Sebecus hasdistinctly serrated edges. For these reasons itis concluded there that there is no identitybetween the flattened tooth associated withthe type of Florentinoameghinia and the teethof Sebecus.

To summarize the foregoing paragraphs:1. Flat, dinosaur-like teeth have been

found associated with mammals in SouthAmerica.

2. These teeth and the associated mam-mals are indubitably of Eocene age.

3. These teeth are without much doubt theteeth of Sebecus.

4. The flat tooth found associated with thetype of Florentinoameghinia is different fromthe teeth of Sebecus. Its relationships must beconsidered questionable at the present time.

268 VOL. 87

BIBLIOGRAPHY

ADAMS, LEVERETT ALLEN1919. A memoir on the phylogeny of the jaw

muscles in recent and fossil vertebrates.Ann. New York Acad. Sci., vol. 28, pp.51-166, pls. 1-13.

AMEGHINO, FLORENTINO1906. Les formations sedimentaires du Cr6-

tace sup6rieur et du Tertiaire de Pata-gonie. An. Mus. Nac. de Buenos Aires,vol. 15, pp. 1-568.

ANDERSON, HOWARD T.1936. The jaw musculature of the phytosaur,

Machaeroprosopus. Jour. Morph., vol.59, no. 3, pp. 549-587, 7 pls.

ANDREWS, C. W.1913. A descriptive catalogue of the marine

reptiles of the Oxford clay. London,British Museum (Natural History),xxiv +206 pp., 13 pls.

BENHAM, W. B.1935. A reptilian jaw from Kakanui, South

Island, New Zealand. Trans. Proc. Roy.Soc. New Zealand, vol. 65, pp. 232-238.

BRONN, H. G.1890. Klassen und Ordnungen des Thier-

Reichs. Reptilien. Leipzig, vol. 6, div.3, pp. 443-1399, pls. 49-107.

BROWN, BARNUM, AND ERICH M. SCHLAIKJER1940. The structure and relationships of

Protoceratops. Ann. New York Acad.Sci., vol. 40, art. 3, pp. 133-265.

BYERLY, T. C.1926. The myology of Sphenodon punctatum.

Univ. Iowa Studies in Nat. Hist., vol.11, pp. 3-49.

CAMP, CHARLES L.1930. A study of the phytosaurs with descrip-

tion of new material from westernNorth America. Mem. Univ. Cali-fornia, vol. 10, pp. 1-174, pls. 1-6, 1map.

EDGEWORTH, F. H.1907. The development of the head muscles in

Gallus domesticus, and the morphologyof the head muscles in the Sauropsida.Quart. Jour. Micro. Soc., vol. 51, pp.511-556.

EVANS, FRANCIS GAYNOR1939. The morphology and functional evolu-

tion of the atlas-axis complex from fishto mammals. Ann. New York Acad.Sci., vol. 39, art. 2, pp. 29-104.

FtRBRINGER, M.1876. Zur vergleichenden Anatomie der Schul-

termuskeln. Theil III, Chap. IV, Saurier

und Crocodile. Morph. Jahrb., vol. 1,pp. 636-816, pls. 23-27.

GREGORY, WILLIAM K., AND CHARLES L. CAMP1918. Studies in comparative myology and

osteology. No. III. Bull. Amer. Mus.Nat. Hist., vol. 38, pp. 447-563.

HUENE, F. VON1929. Los saurisquios y ornitisquios del Cre-

ticeo Argentino. An. Mus. La Plata,ser. 2, vol. 3, pp. 1-196.

1933. Ein versuch zur Stammesgeschichte derKrokodile. Centralbl. f. Min., etc.,Jahrg. 1933, div. B, no. 11, pp. 577-585.

LAKJER, TAGE1926. Studien uiber die Trigeminus-Versorgte

Kaumuskulatur der Sauropsiden. Copen-hagen, C. A. Reitzel, 153 pp., 26 pls.

MIALL, L. C.1878. Studies in comparative anatomy. No. 1.

The skull of the crocodile. London,Macmillan and Co., 50 pp.

MoOK, CHARLES C.1921. Skull characters of recent Crocodilia,

with notes on the affinities of the recentgenera. Bull. Amer. Mus. Nat. Hist.,vol. 44, pp. 123-268.

1925. A revision of the Mesozoic Crocodilia ofNorth America. Ibid., vol. 51, pp. 319-432.

1934. The evolution and classification of theCrocodilia. Jour. Geol., vol. 42, no. 3,pp. 295-304.

OWEN, RICHARD1850. On the communications between the

cavity of the tympanum and the palatein the Crocodilia. Phil. Trans. Roy.Soc., London, 1850, pt. 2, pp. 521-527,pls. 40-42.

PARKER, W. K.1883. On the structure and development of

the skull in the Crocodilia. Trans. Zool.Soc., London, vol. 11, pt. 9, pp. 263-310,pls. 62-71.

PRICE, LLEWELLYN IVOR1945. A new reptile from the Cretaceous of

Brazil. Minist. Agricult., Div. Geol.Min., Notas Prelim. e Estudos, no. 25,pp. 1-9.

RATHKE, HEINRICH1868. Untersuchungen tiber die Entwickelung

und den Korperbau der Krokodile.[Edited by Wilhelm von Wittich.]Brunswick, Friedrich Vieweg und Sohn,vii+275 pp., 11 pls.

269


REESE, ALBERT M.1915. The alligator and its allies. New York

and London, G. P. Putnam's Sons, xi

+358 pp.SIMPSON, GEORGE GAYLORD

1932. The supposed association of dinosaurswith mammals of Tertiary type in Pata-gonia. Amer. Mus. Novitates, no. 566,pp. 1-21.

1933. A new crocodilian from the Notostylopsbeds of Patagonia. Ibid., no. 623, pl.1-9.

1937a. New reptiles from the Eocene of SouthAmerica. Ibid., no. 927, pp. 1-3.

1937b. An ancient eusuchian crocodile fromPatagonia. Ibid., no. 965, pp. 1-20.

1938. Crossochelys, Eocene homed turtle fromPatagonia. Bull. Amer. Mus. Nat. Hist.,vol. 74, pp. 221-254.

1940. Review of the mammal-bearing Tertiaryof South America. Proc. Amer. Phil.Soc., vol. 83, no. 5, pp. 649-709.

1944. Tempo and mode in evolution. NewYork, Columbia University Press, xviii+237 pp.

WILDER, H. H.1923. The history of the human body. New

York, Henry Holt and Co., xiv +623 pp.WILLISTON, SAMUEL WENDELL

1925. The osteology of the reptiles. Cam-bridge, Harvard University Press,xiii+300 pp.

WOODWARD, ARTHUR SMITH1896. On two Mesozoic crocodilians from the

red sandstones of the Territory ofNeuquen (Argentine Republic). An.Mus. La Plata, Paleontologia Argentina,vol. 4, pp. 1-20, pls. 1-2.

270 VOL. 87

!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Documents

Colbert, 1946