THE SKITLL OF BRACHAUCHENIUS, WITH OBSERVA-TIONS ON THE RELATIONSHIPS OF THE PLESIOSAURS.
By Samuel W. Williston,Of the University of Chicago, Chicago, Illinois.
The type of the genus and species Brachauchenius lucasi Wilhston "-is an excellent specimen in the collection of the United StatesNational Museum from the Benton Cretaceous of Western Kansas.The genus is of unusual interest because of several remarkablecharacters previously unknown among the plesiosaurs, which itpossesses, in particular the union of the palatine bones in the middleline, and the very short neck. A second specimen belonging tothe same genus, from the Eagle Ford Limestone of Texas, alsoforming a part of the collections of the United States NationalMuseum, was kindly submitted to me for study by the authoritiesof that museum, a brief notice concerning which was publishedin Science for June 19, 1903.These two specimens supplement each other, the type specimenshowing the underside of the skull and the connected vertebralcolumn as far as the lumbar region (Plate XXXIV), while thepresent specimen permits a thorough examination of the upper part ofthe skull, and has, also, eighteen of the early vertebrae, and a partof the front paddle. The most careful comparisons fail to discovergeneric differences between the two specimens, nor can I detectspecific differences even. The Texas specimen is partly inclosedin a hard limestone matrix, and it is possible that, when the under-side of the skull shall have been cleaned up, specific differencesmay be apparent, but I do not think so. The specimen is slightlysmaller than the type. The Eagle Ford Limestone is known to be anequivalent of the Benton Cretaceous, and I suspect that the imme-diate horizon in which the specimen occurred will prove to be anexact equivalent of that which ^aelded the Kansas specimen.^? . .?_
?
? ?a Field Col. Mus., Pub. No. 73, Geol. Ser., II, p. 57; Lucas, Smithson. Misc. Col.Quart., I, p. 96.Proceedings U. S. National Museum, Vol. XXXII?No. 1540.
478 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxii.The skull of the Texas specimen is moderately elongated, notnearly so much so as that of Trinacromerum or Polycotylus, butmore so than is the skull of the known species of Elasmosaurus. Thetemporal fossae are unusually large, the zygomatic bars remarkablyslender posteriorly, and the parietal bone is not elevated into a thin,high crest as is the case with the skulls of the genera mentioned,but is low, straight, and obtusely rounded on its upper surface.The teeth are fewer, less elongated than in those genera; they arecoarsely striate. All parts of the skull are present, save the anteriorportion of the premaxill?e.Premaxillse.?Of the premaxillse, the anterior portion has beendestroyed, the two posterior teeth on each side only remaining.Since most other forms of plesiosaurs have six teeth on each premax-illa, it is probable that this number was originally present in thisspecimen, though Andrews gives " but five as the number in Pli-osaurusferox, a related form. Perhaps two-thirds of the dentigerousportion is missing, and I have so restored the outline of the skull(Plate XXXVII). The facial processes are as in the other forms ofl^lesiosaurs described by me?elongate, parallel processes, with a dis-tinct longitudinal striation, terminating by overlapping the frontals orparietal processes a little in advance of the anterior end of the orbits.Their width throughout is nearl)^ uniform. They articulate, on theouter side of the skull, with maxillae, frontals and (?) parietals.Maxillse.?Each maxilla attains its greatest width over the externalnareal opening. It is here separated from the premaxillary processby a slender, pointed projection from the frontal oi* nasal. Betweenthis process and the naris, an elongated tongue-like process extendson the frontal or nasal to a little beyond the posterior end of thatopening. Below, the union with the prefrontal begins a little infront of and below the anterior end of the nareal opening and extendsdownward and backward to the most anterior extremity of thelachrymal. On the right side of the specimen the front part of thissuture is apparent,' but in the middle of the course there has beenan inward bending on the line of the suture. On the left side,however, the maxilla, while a little displaced, has been separatedfrom its contiguous elements, making certain that the infoldinghas been at the sutural junction. The maxilla lies somewhat overthe prefrontal squamately and helps form only the most anteriorpart of the nareal opening. The maxillary suture turns backwardbelow the lachrymal to terminate acutely a little before the posteriorend of the orbit, joining the jugal posteriorly.( ?) Frontals.?The bones wliich I here call the frontals, for reasonsgiven further on, lie at the sides of the parietal prolongations, extend-ing anteriorly as slender projections between the maxillary and pre-
a Andrews, Quart. Journ. Geol. Soc, LIII, 1897, p. 177,
THE SKULL OF BRACHA UCHENIUS? WILLISTON. 479maxillary processes already described. Each bone is overlapped inpart in front by the tongue-like processes of the maxilla, and I can notbe sure whether any part of it reaches the nareal border, though Ithink not. Near the posterior extremity of tliis tongue there is a verydistinct suture, extending backward and a little outward for a shortdistance, thence nearly directly backward to a point above the upper-most part of the orbit; on the left side the bone has been separatedat tliis suture. Posteriorly, on the outer side, the bone joins thepostorbital for a short distance obliquely; the suture then turnsinward to join the parietal transversely. The bone is long, pointedanteriorly, flattened or gently concave above in its middle, united onits inner side in front with the premaxilla, beliind with the rostrum ofthe parietal, posteriorly with the postfrontal and postorbital, on theouter side with the prefrontal. It is of course possible that the bone,as described, is composed of two elements, the most anterior of whichwould be the nasal, but of such division there is no evidence in thisspecimen. (See Plate XXXVII, /r.? rmf.)Prefrontals.?The prefrontals are rather broad, irregularly shapedbones, forming the whole of the antero-superior border of the orbitsand the posterior inferior margins of the nares. The inner border ofeach bone, as already described, joins the frontal throughout.Posteriorly, for a short distance, it joins the postorbital, differing inthis respect from the prefrontals of Trinacromerum. Its orbitalborder is thin and arched, terminating at the extreme front angle ofthe orbit. Below, the bone joins the lachrymal by a short suture run-ning forward and outward in continuation of the line of the orbitalmargin to the maxillary suture, which has been described. Ante-riorly the bone is emarginated by the hind border of the nareal open-ing, the tongue of the maxilla, as described, overlapping it and conceal-ing its extent. The bones are convex and smooth, each pierced by twosmall foramina. On the left side the bone, while not crushed ordistorted, has been separated from the adjoining bones and forcedupward somewhat. Inasmuch as its shape and extent on tliis sideagree quite with those of the opposite side, as determined from thesutures, there can be no doubt of its relations and form. There is noindication on either side of a sutural division.Parietals.?The parietal foramen is an elongated opening, oval inshape and about 40 mm. in length. In front of the foramen, theparietals appear to continue forward as an elongated, narrow rostrum,to disappear under the facial extremity of the premaxillne, divided inthe middle by a distinct suture from a little in front of the foramen.The surface on either side of this mesial suture is plane or concave,presenting a number of distinct longitudinal ridges and grooves, wliichbegin near or on the sides of the foramen. The greatest width of theseprolongations is at the hind end, where they together measure 50 mm.Proc. N. M. vol. xxxii?07 31
480 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxii.Where they disappear beneath the premaxillge they have a combinedwidth of about 35 mm. On each side the parietal turns downwardand sHghtly outward into a thin descending process or wing, formingthe lateral wall of the brain case to a depth of 60 mm. At theanterior inner angle of the temporal vacuity there is a rather strongemargination of this descending wall for the attachment of the epip-terygoid, from the upper margin of which a somewhat zigzag suturalline runs upward and then forward to join the extremity of thesuture between the prefrontals and frontals. These lines appear tobe quite alike on the two sides and since they agree with the suturaldivisions in Trinacromerum oshorni and also with the recognizedsutures in the skull described by Andrews as Pliosaurus ferox, therecan be no question, I think, but that they indicate the divisionsbetween the parietals and postfrontals. Back of the parietal fora-men the parietals show no clear indications of a median suture. Thepart here, for the rather long distance between the temporal vacuities,is obtusely rounded above and nearly horizontal, very unlike the thin,elevated crest of Trinacromerum, Polycotylus, and Elasmosaurus.On the under side the parietals include a deep valley between thelateral wails, a little wider below and meeting in a rounded roofabove, for the brain case. This cavity measures over 50 mm. in itsgreatest width.The arrangement of the bones in the frontal and antorbital regions,as I have described them, whatever may be their interpretation,doubtless obtains in all plesiosaurs, with minor modifications.Wliichsoever interpretation may be finally accepted the arrange-ment and structure are very remarkable and very unlike what isknown in other reptiles. That the bones are nearly or quite as I havedescribed them in this specimen I have no doubt. Andrews, in hisfigures and description of the skull of Pliosaurus ferox, reaches differ-ent conclusions and has different interpretations, but I am confidentthat, if his specimen be studied in the light of the information furnishedby the present one, other conclusions and other interpretations willbe reached. A positive suture has never been detected separatingthe median bones in front of the pineal foramen from the parietals.Owen, it is true, thought he detected such a suture in a species ofPlesiosaurus, and Andrews thought there was one in his Pliosaurusferox specimen, though he adds that the parietals are probablyanchylosed with the ''frontals." I have been unable to distinguishsuch a suture in four well-preserved skulls of as many differentgenera studied by myself. In the specimen of Trinacromerum oshornistudied by me, while the adjacent sutures are all clearly indicated,save such as were obliterated by crushing, the very narrow prolonga-tions in front of the foramen have no trace whatever of a distin-guisliing suture, either on the upper or the under side. I believe them
N... 1540. THE SKULL OF BKACHATT'HENLUS?WLLLTSTOX. 481to be merely exogenous processes from the parietals, produced for-ward to meet the extraordinarily elongated premaxillary processes.If such be really the case, the bones on their outer sides must ofcourse be the frontals, and, as frontals, they occupy their normal rela-tions with the adjacent bones, save only the parietals, articulatingbehind with the postfrontals, in front exteriorly \vdth the prefrontals,anteriorly with the maxillae and premaxillae. If the median bones bereally the anchylosed frontals, then these bones must be the nasals.As such, however, their relations would be most extraordinary, theonly instance in comparative osteology where they articulate with thepostfrontals and postorbitals.Possibly the same causes which have prolonged so far backward thepremaxillaries may have caused a posterior displacement of thenasals. In any event I feel sure that the bones on the outer sides ofthese, the supposed supraorbitals, the ones bordering the orbits andreaching to the nares, are the real prefrontals. As such their positionand relations are not extraordinary. As supraorbitals they are quiteindefensible.If the former interpretation be correct, that the parietals haveexcluded the frontals from contact in the middle line, the nasals arewanting in the plesiosaurs. If the latter interpretation is correct,then all the elements of the normal reptilian skull are present, but thenasals have become abnormal in position and relations. I do notknow how the problem can be settled, unless, indeed, some favorablypreserved specimen may disclose an actual suture in front of theparietal foramen.Lachrymals.?The lachr^anal is an elongate bone forming the loweranterior half of the orbital margin. Its sutural union with the pre-frontal is very evident on each side; the suture between it and themaxilla is perhaps not wholly free from doubt in this specimen,though there can be little possibility of error, the indications of thetwo sides agreeing as they do. The bone joins the jugal behind by anoblique suture; the maxilla in the middle below; and the prefrontalanteriorly, as already described. Inasmuch as these relations seemto be quite the same as those described by Andrews in Pliosaurusferox, I think that the presence of a lachrymal as a distinct bone in theplesiosaurs may be finally set at rest. In the skull of Trinacromerumoshorni, previously described, there is a pointed process of bone whichhas the same relations with prefrontals and maxilla?, but not with thejugal. I could not detect a suture separating it from the maxilla.Neither is it probable that the lachrymal in Elasmosaurus snowi,which must resemble that of Trinacromerum, articulates with thejugal.Postfrontals and postoriitals.?The postfrontals and postorbitals Ibelieve are distinct bones in this specimen. The parieto-post-
482 PROCEEDINGS OF THE NA TIONAL MUSEUM. vol. xxxn.frontal suture I have already described. A suture quite as evidenton each side runs obliquely outward from the hind end of the frontalsor nasals, and then turns downward, about as figured by Owen forPlesiosaurus. The postfrental, as thus defined, joins the parietaland touches the epipterygoid internally, the frontal anteriorly, andthe postorbital exteriorly. The postorbital articulates with the pre-frontal anteriorly, the postfrontal on the inner side, and, by itsanterior angle, the so-called frontal; and the jugal exteriorly. Thetwo bones, seen from behind, present a broad, nearly vertical wall,deflected somewhat anteriorly below, and ending in a thin, sharp,nearly horizontal margin, continued from the epipterygoid notch tothe jugal. The orbital border of the postorbital is thinned, some-what serrated, and concave. The temporal border above is sharpand angular, curving downward to terminate in the thin upper marginof the zygoma. The bone outwardly is massive and strong, endingin a horizontal suture, which is nearly continuous with the lowerborderof the orbit and the upper front border of the zygoma.Jugals.?The jugal differs considerably from that of other formsof plesiosaurs known to me. The sutures distinguisliing it from thepostorbital, lachrymal, and maxilla are very clear, as I have describedand figured them. That uniting it with the squamosal is doubtfid.On the left side the bone has been separated very cleanly from thematrix, and is in a beautifully uncUstorted condition. A little backof the liind border of the postorbital there are, near the middle of thejugal, the orifices of two or three malar canals. These canals arevery characteristic of the plesiosaurs, and usually open near thesquamoso-jugal suture, but there is not the slightest indication ofsuch a suture here. These canals, piercing the jugal, enter the orbitnear its lower posterior corner. In the orbit the jugal turns inwardfor a considerable distance, forming a bowl-like floor posteriorly; itsinner border I can not trace.Describing the zygoma as a whole, it has a somewhat tliickenedupper border in front, thimied below. The arcade chminishes rapidlyin width, chiefly at the expense of the lower part, to beyond its mid-dle, where its width is less than one inch, and the bone is tliin andweak. At this place the arch, on both sides, shows an oblique fi'ac-ture, which may, possibly, represent the squamosal suture, though Iam very doubtful. I have indicated this possible suture by dottedlines in the drawings. The squamosal beliind broadens to a width ofabout two and a half inches where it joins the quadrate, and is thickerhere, the upper border ascending rapidly; the lower border is concave.The zygoma is very remarkable for its attenuation posteriorly, leavinga large open space above the posterior part of the mandible in frontof the articulation. I can conceive of the complete erosion of the barhere, as occurs in some turtles, leaving the temporal vacuity broadly
NO. 1540. THE SKULL OF BRACHAUCHENIUS?WLLLISTOK. 483open on the sides. It would seem that the chief support for the man-dibular muscles must have been on the sides oi the parietals and thestout postfrontals and postorbitals.The Uniits of the squamosal above can not be determined, owingto the erosion of the specimen, as inchcated in the drawing. Theparieto-squamosal arch is, however, quite stout, narrowed anteropos-teriorly near its upper part. The massive quadrates are exposedbelow on the outer side and behind. Further information concerningthe occipital region can not be had until the matrix has been removed
;
the anterior cervical vertebras are crowded into this space.At the bottom of the large temporal vacuities the supraoccipitals,exoccipitals, petrosals, and stapes were found more or less disarticu-lated and separated. The larger part of the exoccipital is seen some-what removed fi-om its relations to the stout supraoccipitals. Itsanterior, cranial surface presents a deep pit and marginal sutural sur-faces, completed by imion with the supraoccipital and petrosal. Theparoccipital process is rather slender, directed downward, outward,and backward in life, with its distal extremity flattened and appar-ently spatulate, for union with the upper end of the quadrate, asdescribed in THnacromerum osborni.Petrosals.?The petrosal is a peculiar bone. That of the left sidehas been wholly freed from its matrix; on the right side it lies withits free, convex, outer side exposed near the front border of the tem-poral vacuity. Exteriorly the bone is nearly evenly and smoothly con-vex, shell-like. The iimer side I have figured in Plate XXXV (pet) ,natural size. Its precise mode of union with its two contiguous bonescan not be determined. Its two cHverging canals doubtless lead intothe supraoccipital and exoccipital sinuses or semicircular canals, asI have found them in Trinacromerum osborni. The greater part ofthe bone is deeply and smoothly excavated for the internal ear, leav-ing a free border for the petrosal part of the large foramen ovale.The excavation is deep and large for the size of the skull, much largerproportionally than in the mosasaurs.Stapes if).?A small and peculiar bone, lying apparently nearly inposition in the matrix on the right side, I can determine only as astapes, a bone hitherto unknown among the plesiosaurs. It is a short,stout bone, a side view of wliich is shown in Plate XXXV {st) nat-ural size, not unlike a human metatarsal, though less slender, withan attenuated, cyhndrical shaft, and an articular expansion at eitherend. TVliat I believe to be its proximal end, from its position in thematrix, presents a hemispherical articular surface, bounded by ashelf-like ridge, as though for articulation in a foramen. The otherextremity is obliquely expanded, concave somewhat from side toside, smooth, and with a partial longitudinal ridge near one side.The extremity of the bone has been broken away.
484 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxii.Where the external ear was located is a puzzle. In all probabilitythere was an external tympanic closed membrane, as in the turtles.The shortness of the stapes, if stapes this bone be?and there is noother place in the skeleton where it can be located?must mean anexternal surface close to the median line of the skull. There are noincUcations of an otic foramen or notch anywhere about the quadratethat I have discovered in this or other specimens of plesiosaurs.Dollo has ventured the opinion, from the thickness of the preservedcartilage about the external ear in certain mosasaurs, that they, orcertain types of them at least, were deep-sea divers, from the resem-blance in the structure to that of the cetaceans. This ear cartilage isvery abundant and very thick in Platecar'pus and Tylosaurus, less soin Clidastes. No calcified cartilage of any kind have I ever observedanywhere in the skeletons of plesiosaurs, so that any inference as tothe habits of the plesiosaurs from its absence about the ear would notbe legitimate.Paddle.?A part of a paddle, evidently an anterior one, shown inPlate XXXVI, was found in this specimen lying closely upon andacross the face of the skull. Not much information as to the genericor specific characters of the form can be deduced from it, but I give,nevertheless, a good photographic view of the specimen. The limbswere evidently not of the slender type of the elasmosaurs, but whetheror not there was a duplication of the epipodials can not be determined.Vertehrse.?The remains of twelve cervical and six dorsal vertebraeare preserved in the limestone matrix back of the skull. They agreein all respects with the vertebrae described by me in the type speci-men of BrachaucJienius lucasi, save in their slightly smaller size.The cervical series is connected, as are also five of the dorsals, whichare curved forward reversed by the side of the cervicals. Doubtlessthe specimen originally was composed of a large part, perhaps thelarger part of the skeleton, though only the single block containingthe skull and vertebrae and the attached paddle was secured by thecollector. At the angle of the vertebral series one or two may havedisappeared, but probably not more. I have every reason to believethat the number of the cervicals is the same as in the type, namely,13. The cervicals measure, in length, beginning with the axis: 25,25, 25, 25, 28, 28, 30, 33, 35, 40, 40 mm. The dorsals preserved: 45,50, 60, 60, 60 mm. The centra of the dorsals are smoothly roundedon the under side, without excavations or vascular foramina, resem-bling dinosaur vertebras so closely that it would be difficult to dis-tinguish their centra if preserved singly. The cervical ribs are single-headed, with not the least indication of division.The total length of the skull, with the missing premaxillary por-tion estimated, is about 0.80 m., the width at the posterior part ofthe orbits 0.35 m. The length of the type specimen is about 0.90 m.;the width proportionally the same.
NO. 1540. THE SKULL OF BRA CHA UCHENIVS? WILLISTON. 485Relationships of Brachauchenius.?The most distinctive charactersof the genus are found in the broadly united palatines, the broadunion of the pterygoids posteriorly, the short, deep-set interptery-goidal vacuities, the ridge-like buttresses of the pterygoids, theremarkably small number of the cervical vertebrae, the absence ofvascular foramina on their under side, the single-head cervicalribs, etc. The dorsal surface of the skull has a remarkable resem-blance to that of Pliosaurus ferox, as figured by Andrews. Andrewsassumed that the palatines in his specimen were separated by thepterygoids throughout, but expressl}'^ says that indications of the pala-tine relations posteriorly are wanting. I have scarcely a doubt butthat they will be found to have the same structure as in Brachau-chenius in better preserved specimens of the genus.Much stress has been placed upon the palatal structure in thereptiles as indications of phylogenetic relationships, but I have neverhad a great deal of faith in the stability of these parts. Here we havethe union or separation of the palatines in thcr same order. Thegeneral shape of the skull, the depressed parietals, and, I am confi-dent, the relations of all the other bones of the upper side of the skull,are all nearly alike in Pliosaurus and Brachauchenius. Furthermore,in the reduced number of the cervical vertebra^ in the two forms, 18or 20 in the older, 13 in the younger, we have a genetic resemblance,I believe, one that strengthens my assumption that the shortenedneck in the later forms is not a primitive character, but a degenerateone, one that has been acquired in more thai one phylum. Indeed,so far as all these characters of the skull go, in the probability that thearrangement of the skull bones will be found essentially alike in thetwo genera, I should hesitate to separate the two types generically,were it not for the cervical ribs, single-headed in Brachauchenius,double-headed in Pliosaurus. The character of the cervical ribs hasbeen considered as of more than generic importance, Seeley evenproposing an ordinal subdivision based upon the divided or undividedneck-ribs. Here, too, I believe that the fusion of the imperfectlydifferentiated heads is a feature common to more than one line ofdescent, and is of no more than generic importance. It is a fact thatall known American Cretaceous plesiosaurs have cervical ribs withundivided heads, and that is probably the case with all Cretaceousforms, as it is also with the known American Jurassic ones. Double-headed ribs are a primitive character confined to the early forms, forthe most part.In conclusion, I would suggest that the family Pliosaurida^ bemaintained, based upon the common characters apparent or probablein Pliosaurus and Brachauchenius.The characters of Brachauchenius, so far as they are now known, Igive as follows:Brachauchenius.?Mesocephalic. Teeth not more than 2U in each
486 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxii.maxilla, strongly ridged and anisodont. Parietals rounded andobtuse above, not elevated into a crest. Temporal vacuities large;zygomatic bars slender posteriorly. Pterygoids broadly united inthe middle behind ; the interpterygoidal vacuities short, at the bottomof a depressed pit. No palatine foramina. Palatines broadly unitedin the middle in front of the pterygoids. Cervical vertebrae 1.3 innumber, smootlily rounded below, without vascular foramina, shal-lowly concave at extremities; cervical vertebrae broader than long;cervical ribs single-headed. Benton Cretaceous of Kansas and Texas.Relationships of the Plesiosaurs.?In the attempt to reach somedefuiite conclusions as to the habits of the plesiosaurs, I gave ^ fiveyears ago the following list of adaptive characters in aquatic, air-breathing vertebrates
:
1. Elongation of the head, with attenuation of the facial region.2. Elongation of trunk and tail, but especially the latter, with pro-gressive weakening of the zygapophysial articulations posteriorly.3. Shortening of the neck.4. The acquirement of a caudal fin.5. The acquirement of sclerotic plates.6. Recession of the external nares.7. Absence of the sacrum and the absence or progressive obsoles-cence of the sternum.8. Greater slenderness and smaller size or loss of the hind limbs.9. H3q^erphalangy and hyperdactyly.10. Smoothness of the skin.11. Sponginess of the bones of the skeleton.12. Increase in number and decrease in size of teeth.The exceptions which the plesiosaurs present to these adaptationsare
:
1. Elongation of the neck, with increase in number of vertebrae.2. Shortening of tail and body, and the flattened, depressed formof the latter.3. The presence of a well-defined sacrum of three vertebrae.4. The somewhat greater slenderness of the hind limbs, but withlittle or lio decrease in effectiveness as propelling organs.In these exceptions the plesiosaurs agree with the marine turtles.In the tail-propelling, aquatic vertebrate the propodial bones areinvariably shortened, as for instance in the Cetacea, Ichthyosauria,Pythonomorpha, and the front legs of the Thalattosuchia, and thelimbs become merely equilibrational organs in direct proportion tothe effectiveness of the tail as a propelling organ. Experiments onfishes show that the loss of the paired fins does not impau* the swim-ming powers of the individual, but does require the constant vibra-tional use of the tail in the preservation of the equilibrium, while the
a Kansas University Science Bulletin, I, 1902, p. 259. .
NO. 1540. THE SKULL OF BRACHAUCHENLVS?WILLLSTON. 487loss of the caudal fin results in the total disability of the animal. Inanimals propelling; themselves wholly or chiefly hj the aid of thelimbs the propodials are not shortened, but are, on the contrary,elongated, as in the plesiosaurs and marine turtles. The reason isobvious: The propodials become elongated handles of oar-like organs,of which the blades are formed by the progressively widened epi-,meso-, and metapodial elements, and the phalanges. The front limbsof the plesiosaurs are always broader and stronger, but not longerthan the hind ones. The front legs of the marine turtles are not onlybroader and stronger, but also longer than the liind ones, though thelatter have by no means lost their eflPectiveness as propelling, or, moreprobably, guiding organs. The connection of the hind limbs of theplesiosaurs with a well-developed sacrum of three vertebrae conclu-sively proves the propelling function of these limbs, if such proof werenot abundantly furnished by the limbs themselves.We have, then, certain marked resemblances in the form and modeof progression between the plesiosaurs and turtles, as contrasted withthe tail-propelling type presented by the ichthyosaurs, mosasaurs,and thalattosuchians; and Fraas uses" tliis resemblance as a supportfor the diphyletic grouping of the reptilia by Osborn into the Synap-sida and Diapsida, the former having the oar-propelling type, thelatter the tail-propelling type. But the argument is fallacious; theresemblances in mode of progression and bodily form no more implya common phyletic origin than do the much more marked resem-blances of the ichthyosaurs and dolphins.It is chiefly because of the external resemblances of form andsimilarity in mode of locomotion in the water that it has been gener-ally and indefinitely assumed, from Buckland's time to the presentthat the plesiosaurs were related to the turtles. How well thishypothesis is sustafeied by the internal structure may be shown bythe following comments
:
In addition to external resemblances and undoubtetl similarity inhabits of life, two other characters have been often cited as evidenceof relationship between these two orders?the epiphysial mode ofossification of the propodials (or rather of the humeri, since there isno evidence yet that the femora have the peculiar ''epiphyses"), andthe fusion of the procoracoid with the scapula. As to the fii-st ofthese assertions, recent careful investigations by R. Moodie con-clusively prove that the turtles do not have true epiphyses, and aswas long ago stated by Dollo, and recently confii-med by Mr. Moodie,the lizards do have, many of them, at least, distinct terminal bonyepiphyses on then- long bones. The mode of ossification of thehumerus of the plesiosaurs is most extraordinary, without knownparallel among reptiles, or mammals either, so far as that is con-
? Jahresheflen d. Vereins f. vaterl. Naturkunde in Wurteinberg, 1905, p. 363.
488 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxii.cerned." The presence of "epiphyses" forming nearly the wholeof the humerus, their apices separated in the middle of the bone byperforating canals extending through the shaft of the bone, is utterlyunlike anything that has been observed in turtles or any other rep-tiles at any stage of then existence. I trust that the myth of epiphy-ses as an evidence of relationship between the turtles and plesiosaursmay not reappear again.
^
As to the structure of the scapula, all students of the plesiosaursare now agreed that the procoracoid does not unite with the scapula,whatever may be the case in the turtles (where it is equally improb-able). The presence of a distinct foramen in the coracoid of manyplesiosaurs, or its deep emargination posteriorly, points, I think, toa normal reptilian manner of development of this bone. The tri-radiate structure of the scapula is simply a parallel character, broughtabout by the same causes which have produced the enormous develop-ment of the coracoids, a structure absolutely lacking in the earlierand simpler nothosaur type, where it would confidently be expectedwere the orders genetically allied.Whatever of resemblance there may be in the form and habits ofthese two orders of animals has been due solely to parallel evolution,to similar aquatic adaptations. In theu- internal structure they arereally remote from each other, and neither could have been derivedfrom the other type, not even in a remotely antecedent stage. Theturtles have a stegocrotaphous skull, unlike all other reptiles save theCotylosauria, Procolophonia, etc. The plesiosaurs have a large tem-poral vacuity, larger indeed than is to be found in any other reptilesof the therocrotaphous (I coin the word) type. Leaving out ofaccount adaptive characters, we have the following most importantdifferences in the structure of the two orders: The turtles lack thelachrymal, postorbital, and transverse bones, all well developed inthe plesiosaurs. They have a distinct opisthotic, wanting in theplesiosaurs, and a large quadratojugal, probably wholly wanting in theplesiosaurs. The plesiosaurs have a large pineal foramen, wholl}"
o See Williston, Field Columbian Mus. Publ. No. 73, p. 73.6 In the pigeon " at four days there are two cones of gradually ossifying cartilage, theapices of which are close together in the middle of the bone, at the point wherethe primary center of ossification occurred, while the bases, quite unossified, form thearticular ends. These two cones are ensheathed by a layer of periosteal bone, whichof course is thickest opposite the ends of the cones, and thins off as the two extremi-ties are approached. . . . These two cones probably represent the so-calledepiphyses of the Plesiosaurus. I have not been able to find that this reptile possessedanything corresponding to true epiphyses." (Parsons, Jour. Anat. and Physiol.,XXXIX, 1905, p. 403.) The figure of the bird humerus, given by Parsons, strikinglyresembles the ossificatory plan in the plesiosaurs, save that the latter in the early stagehas perforating canals through the rudimentary medulla. These observations of Par-sons, seen by me for the fu-st time since the above was in type, effectually dispose ofthe whole matter of turtle relationships in the manner of ossification of the long bones.
NO. 1540. THE SKULL OF BRA CHA VCHENIVS? WILLISTOX. 489wanting in the turtles. The turtles have a single, unpaired, truevomer and no prevomers; the sauropterygians have large prevomersand a small or no true vomer. A large interpterygoid vacuit}^ ispresent in the plesiosaurs, wanting in the turtles. Furthermore, theturtles have still preserved the primitive hypocentral mode of attach-ment of the thoracic ribs, while the single-headed thoracic ril^s of theplesiosaurs are attached higli uj) on the extremities of the diapophy-ses, and this character can not be ascribed to aquatic adaptation, Ithink, since the ichthyosaiu's and mosasaiu's have preserved theirearly pleurocentral attachment of these ribs.And one is welcome to all the resemblances that may be found inthe vertebrae, girdles, and limbs. I repeat, there is only a remoterelationship between the two orders in osteological structure. Theplesiosaurs coidd not have been derived from any ancestors thatmight by the widest stretch of imagination be called Chelonia, orChelonia-like. Nor could the turtles have come from any forbearseven suggesting the sauropterygian structure.I am still strongly of the opinion that the Sauro])terygia werederived from a primitive therocephalian ancestry; while I am fu-mlyof the opinion that the turtles have had a quite independent originfrom some primitive cotylosaiunan, like the Chelydosauria, as Casehas forcefully shown. The turtles occupy a phylum distinctly thenown, no more intimately related to the ])lesiosain's than they are tothe ichthyosaurs or rhynchocephalians. I can not accept the con-tention of McGregor that the Ichthyosauria had a primitively sauro-crotaphous (I need not apologize for the word) type of skull, butwould rather believe that they, too, enjoyed a genealogical line alltheir own from the most primitive ty])e of reptiles, and that theyshould no more be grouped with the dinosaurs and crocodiles thanwith the plesiosaurs and theriodonts.EXPLANATION OF PLATES.Plate XXXIV.Brachauchenms lucasi, type specimen in U. S. National Museum. After Lucas.(Plate from Vol. I, Quarterly, Smithsonian Misc. Col.)Plate XXXV.Skull of Brachanchenius, U. S. National Museum Collection, from Eagle Ford Shalesof Texas. One-fourth natural size. J5.to, exoccipital; Pf/, petrosal; *S/, stapes.Plate XXXVI.Part of front paddle of Brachauchmius. Texas speeimen, one-half natural size.Plate XXXVIT.Restored outline of skull of Brachauchenius, Texas specimen, one-half natural size,^w^;, angular; ^p, epipterygoid; Fr.?, frontal; Va.?, nasal?; /, jugal; ia, lachrymal;max, maxilla; na, external naris; Pa, parietal; Pf, parietal foramen; Pfr, postfrontal;Po, postorbital; Prf, prefrontal; Pm.v, premaxilla; Q, quadrate; Sq, squamosal; Sur,surangular; Tv, temporal vacuity.

U. S. NATIONAL MUSEUM PROCEEDINGS. VOL. XXXII PL XXXIV

U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. XXXII PL. XXXV

U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. XXXII PL. XXXVI
Part of Front Paddle of Brachauchenius.For explanation of plate see page 489.

U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. XXXII PL. XXXVII