PLESIOSAURUS FROM THE UPPER KEUPER AND RHAETIC
Numerous finds of Plesiosaurus are known from the Rhaetic, particularly frequent from the region of Bristol and Gloucester in England; also from the Rhaetic of Autun, from where SAUVAGE (1) listed two species, namely cervical vertebrae of Plesiosaurus costatus OWEN and dorsal vertebrae of Plesiosaurus bibractensis, a species named by him. A number of vertebrae of Plesiosaurus rugosus OWEN in the Tübingen Collection have been found in the bonebed of Bebenhausen. Termatosaurus albertii QUENST. is also a true Plesiosaurus; there are cervical vertebrae of a probable English species.
In the Museum at Basel are found the largest Plesiosaurus vertebrae that I have seen (see text-fig. 51 and 52). They come from the Zanclodon beds of Schönthal and were found together with Gresslyosaurus ingens RÜTIM., an extraordinarily large Belodon, and remains of Mastodonsaurus. A dorsal centrum is 8.5 cm long and broad, and 8 cm high. One sees the characteristic cross (spinal cord and diapophysis attachment) on the upper side. The articular facets of the centrum are almost entirely flat, for the rest I refer to the figure. Another vertebra from the border region between neck and; back still has most of the upper arch. The centrum is 5.5 cm long, 6-7 cm broad, and just as high. In cross-section it is wedge-shaped below and rather flattened in the middle. The indicated double rib attachment is very broad and strong. The parapophysis lies about at the level of the centrum. The neural canal is large and wide. Very large and strong postzygapophyses are found above it. When further vertebral remains are considered, there is a left pre- and postzygapophysis, which indicates a short but very high upper arch, probably of a cervical vertebra. The form recalls to some extent the also very high cervical vertebrae from the Muschelkalk of Bayreuth (see pl. V, fig 5); only there the zygapophyses are longer and less strong. A small caudal vertebra (one of the very last, very small) is also present from Schönthal. In the same place were found several long, conical, rather curved teeth circular in cros-section and over 5 cm long, which could also probably belong to this Plesiosaurus.
In the Geological Collection at Göttingen University is found the distal end of a humerus of Plesiosaurus from the Rhaetic Keuper of the Little Hagen near Göttingen (see text-fig. 53). The piece preserved indicates quite a slim bone. One longitudinal contour goes at almost a right angle from the end surface, the other is curved toward the midline, this is the posteriorly directed edge. The muscle attachments are strongly marked. The piece is 8 cm long and 12 cm broad. No one doubts that this humerus belongs to a Plesiosaurus and not perhaps to an Ichthyosaurus.
DOLIOVERTEBRA FRITSCHI n. gen. et sp.
(dolium = barrel)
The Geological Collection at Halle am Saale possesses a very fine sacrum consisting of three vertebrae, from the Schaumkalk of Freiburg am Unstrut, a sketch of which I reproduce here. The centra are strangely barrel-shaped, with narrow raised edges and flat articular surfaces. The upper arch is bulky and recalls in this way the pygmy Sauropterygia of the Lower Muschelkalk of Silesia and Saxony. The barrel-shaped centra are most like those of Proneusticosaurus VOLZ (ms.). The spinal process is not high but so broad that they all three meet. The transverse process is short and very thick. Separated by a clear suture, a long, hollow sacral rib, triangular in cross-section, terminates on it. The flat sides of the sacral ribs curve upward and possess a roughly plate-shaped horizontal expansion in the middle. The sacral ribs stick out at a right angle; only the distal ends curve rather downward. The first and third sacral ribs bend toward the second, which is also the shortest, so that a semi-circular contact surface for the ilium results from it. To judge from the form of the vertebra, I suppose the animal was related to Neusticosaurus and Proneusticosaurus.
Only little is known about Chelonia from the Triassic. From the Upper Keuper of Württemberg there are two specimens of Psammochelys keuperina QUENST. (= Proganochelys quenstedti BAUR): armor and dorsal vertebrae as well as the negative of the scapula described. The vertebrae described by H. von MEYER as Chelytherium obscurum from the same bed and locality also seem to belong to Psammochelys. According to BAUR and FRAAS it is a true pleurodire. It is striking, almost astonishing, that such a highly developed type still appears as the oldest known chelonian. It is obvious that this cannot in reality be the oldest chelonian. Moreover, I believe I can now mention cryptodire Chelonia from the Upper Muschelkalk (see below). Without, however, wishing to give an opinion on the difficult question of the position of the Dermatochelidae, I might try to consider more closely the connections of the Chelonia with other reptile orders, on the basis of general chelonian features.
Comment is repeatedly made on the noteworthy resemblance of the external appearance of the skulls of a chelonian and Oudenodon. Both are highly adapted toothless forms, which on its own makes a direct connection very unlikely. According to SEELEY, Oudenodon is certainly found in the oldest of the five divisions of the Karroo. For all that, it is of value to examine carefully the skull of Chelonia and several Anomodontia. The views of the occiput of Euclastes (Eocene), Dicynodon (which agrees almost completely with Oudenodon), Placodus, Nothosaurus, and finally Pareiasaurus are very instructive. Euclastes, Dicynodon, and Placodus have a three-part condyle, the form of the exoccipital is very similar to these and Nothosaurus, above it the opisthotic likewise. The parietals follow directly on the supraoccipital only in Nothosaurus and Chelonia; in Dicynodon and Placodus interparietals intercede. The pterygoid is only visible from behind in Nothosaurus, Placodus, and Pareiasaurus. The skull of Kistecephalus agrees in all essential points with Dicynodon. I have appended here the posterior view of the skull of Pareiasaurus in order to show the similar hollowing out of the parietals here and in Chelonia. Chelonia have a posteriorly open temporal fossa, but it can be entirely covered by the parietal, squamosal, and postfrontal, as is the case in e.g. Euclastes and Rhinochelys. However only a single pair of temporal fossae exists in Anomodontia. As in Dicynodon, Oudenodon, and Ptychognathus, the frontals are comparatively small and only play a small part in the bordering of the orbit. The same is the case in Theriodontia and particularly in many Stegocephalia; I recall only the temnospondylous Actinodon. Chelonia lack a postorbital as do most of Nothosauria (except Cymatosaurus and Eurysaurus) and Dicynodontia (Therochelonia SEELEY), while other Anomodontia (= Therosuchia SEELEY) including Placodus have one. Therefore in Chelonia the postfrontal is developed very large. As in many Anomodontia, however not in Dicynodon, the nasal opening is single, not paired; in Kistecephalus arctatum OWEN (1) both nasal openings have become a single, spectacle-shaped opening by a narrow lateral connection. The palate is also built on one and the same plan in Chelonia and Dicynodontia. If we next compare Euclastes with Dicynodon and Oudenodon, we see in both a short presphenoid, beside it the narrow, curved pterygoid, elongated forward and backward, close in front of the unpaired nasal opening, which is bordered laterally by the palatine and vomer; the premaxilla follows in front as a continuation of the vomer. In other Chelonia, like Chelydone or Emys, all this is much more broadened, by which the physical appearance is changed; also the nasal opening is divided in two by the vomer and an opening has appeared between the palatine and pterygoid; however the general structure remains the same.
Different analogies and homologies may also be found in the skeleton. The more or less wedge-shaped centra in cross-section are still preserved in the anterior half of the tail; in Dicynodontia in the dorsal and caudal region; the structure of the ilium, e.g. in Chelydra and Dicynodon, is similar (so also in Sauropterygia). The proximal end of the chelonian femur is comparable only with the Therochelonia, and not with the Therosuchia, in that a large femoral head sits in the acetabulum and the greater and lesser trochanters project in front like wings and produce a deep pit between themselves and the head. We find a similar situation in Dicynodon (orientalis LYD.). However, in the Therosuchia, like Pareiasaurus, Rhopalodon, Cynosuchus, etc., the greater trochanter is developed shield-like and the femoral head sits like a small wart on the lower end of the greater trochanter, by which a similar structure is obvious upon consideration of the head. Thus the fossa in the connected skeleton of the Therosuchia is to be seen from the dorsal and lateral sides of the animal, in the Therochelonia (and in Hatteria) from the ventral side. The femur of Nothosaurus has a triangular end surface at the proximal end which corresponds to the head and both trochanters; however the great development of these three parts does not occur, they have become rudimentary to some extent by the swimming way of life; therefore it cannot be decided from which of the two types of femur it is derived. The chelonian humerus has an entepicondyloid fossa at the distal end and often an ectepicondyloid foramen (radial nerve canal), which both occur as foramina in Anomodontia and Rhynchocephalia. The structure of the primary shoulder girdle also agrees in its elements (namely scapula, coracoid, and procoracoid) with Anomodontia and Rhynchocephalia (according to FÜRBRINGER, DOLLO, and HUXLEY; contrary to BAUR).
The numerous analogies between Chelonia and Nothosauria are often pointed out. But since both appear simultaneously the descent of one from the other is impossible. It is conceivable rather that both had a joint ancestor. There are so many points of agreement and similarity with Dicynodontia that this connection seems very likely. Chelonia and Hatteria certainly have some points in common, in particular similar ischium, pubis, and ilium, the number of two sacral vertebrae, and the proximal end of the femur; but the similarity with the Dicynodontia outweighs this, however distant. Indeed I will not even maintain a direct descent of Chelonia from Dicynodontia, as already said, but I cannot entirely agree with the arguments put forward by FÜRBRINGER (1) for the other side.
As already stated and proved, I do not believe that Placodontia are primitive Chelonia, but that they also branched from Anomodontia, but that they are now more connected with Deuterosauridae, while Chelonia has nothing to do with Deuterosauridae.
CHELYZOON LATUM n. gen. et sp.
Pl. VII, fig. 2
Among a large number of vertebrae from the main Muschelkalk of Laineck near Bayreuth, in the Palaeontological Museum at Munich, a cervical vertebra attracted my attention and I took it for chelonian. Already the fact that it is procoelous isolates it completely from all known Triassic vertebrae. The centrum, 5.8 cm long, is strongly keeled on the underside, as in many Chelonia, and in the middle and below it is extremely strongly constricted. The height of the centrum is 3 cm. The posterior, roughly heart-shaped articular surface is convex and only 2.5 cm broad; the anterior, 3.5 cm broad, is round and clearly concave. The actual spinal process is unfortunately broken off, yet the part still present seems to indicate a low, ledge-shaped spinal process. The prezygapophyses are placed flat, are strikingly strong and broad; their outer edges stand 4 cm apart. Below them are found even more strikingly, broad projecting parapophyses with downward-directed articular surfaces. Diapopohyses are absent or are fused with the parapophyses. The postzygapophyses are unfortunately not preserved. The points of the parapophyses stand 5 cm apart. In favor of the chelonian nature are the procoelousness, the keeled centrum, and the development of the prezygapophyses and parapophyses.
The cervical vertebrae of Chelydra and Macroclemmys agree strikingly with that described here. There the anterior articular surface is also very broad. The prezygapophyses are developed similarly, and the articular surfaces are bordered behind in exactly the same way. The parapophysis, found right in front, projecting strongly with a downward-directed surface, is present similarly in both. However no articular surfaces for the cervical ribs are present. A similar development of the parapophyses (or ? diapophyses) is generally characteristic of chelonian cervical vertebrae. Therefore it seems that the affiliation of these vertebrae to Chelonia is justified, and I believe it is most likely they are to be ranked with Cryptodira because of their resemblance to Chelydra. This is a fact of great importance for the phylogeny of Chelonia, because hitherto only the pleurodire Psammochelys was known from much younger beds of the Triassic. The Cryptodira, specifically Sphargis-like forms, have been regarded as the most primitive Chelonia by several people. The occurrence of Chelyzoon in the Muschelkalk is indeed no positive evidence for the opinion just mentioned, but it shows that forms from this relationship group were present already in the Middle Triassic, and also that at that time, or rather later in the Keuper, Cryptodira and Pleurodira already existed as separate related suborders. Here it may be enough merely to establish this.
CHELYZOON BLEZINGERI n. sp.
Pl. VII, fig. 1
Another similar cervical vertebra belongs to the Natural History Cabinet, Stuttgart (No 8728). It was collected in the upper Muschelkalk bonebed of Crailsheim by BLEZINGER. This vertebra is the largest and best preserved. It differs from the previous one by its greater length but it is built almost the same. The differences listed, together with the younger age, make another species of the same genus likely.
The centrum, 9 cm long and 3 cm high, makes the earlier interpretation (on the label) as Tanystrophaeus seem understandable. The posterior articular surface is narrow, heart-shaped and convex; unfortunately the anterior is missing. The underside of thc centrum is keeled. In the middle of the side a high edge runs the length from behind to the parapophysis. The parapophysis projects, in chelonian fashion, from the side with a downward-directed surface. The very strong prezygapophysis is found above it with a small, round, clearly bordered articular surface. The postzygapophyses are also quite massive. The articular surfaces are only slightly curved, almost horizontal. The spinal process rises 2 cm above the postzygapophyses; its base is just less than 5 cm long. This vertebra differs from the previous ones by relatively greater length, broader prezygapophyses with smaller articular surfaces and stronger but lower lying side keels.
These two vertebrae, which were found in different horizons of the Muschelkalk in Swabia and Bavaria, indicate a wide distribution of marine Chelonia. It is striking that they have hitherto been so completely ignored, although these vertebrae are very like those of the living species.
ON ARCTOSAURUS OSBORNI ADAMS
I think that the single cervical vertebra of Arctosaurus osborni ADAMS (1) also belongs here. Whether it is also procoelous cannot be definitely seen from the figure, but it is possible. It is keeled below; the far-projecting prezygapophysis bears the characteristic small round articular surface, which is present in Arctosaurus and Chelyzoon on the edge running anteriorly from the postzygapophysis; the diapophyses are unfortunately not preserved; in front below the prezygapophysis is a fracture place, on which the parapophysis clearly sat. The entire vertebra recalls Chelyzoon blezingeri and Chelyzoon latum very much, but it is considerably shorter than these. The separate generic name is justified for the present by this and other small differences. Arctosaurus figured previously among the Dinosauria, but with which it had little more in common than the keeled underside (of the cervical vertebra). In particular such a bordered small, round articular surface on the prezygapophysis, as Arctosaurus and Chelyzoon and the Chelydridae in general possess, never occurs. Also the Dinosauria, with the exception of Tanystrophaeus, never lack the diapophysis.
VERTEBRA FROM THE RHAETIC
Pl. VII, fig. 5
Finally I may mention a vertebra from the Rhaetic of Schlösslesmühle near Steinbronn (Tübingen Collection), whose figure is almost unrecognizable in QUENSTEDT’s “Jura” pl. 2, fig. 3. It is a rather elongated, weakly procoelous vertebra. The anterior articular surface is oval and barely noticeably deepened, while the posterior is clearly convex, very narrow below and broader above. The centrum is noticeably constricted in the middle and sharpened downward, but is truncated again on the outermost edges by a horizontal surface that is broader anteriorly than posteriorly and clearly shows the three parallel keels there. The spinal process curves posteriorly rather flat. The postzygapophyses are broken off as is the anterior part of the prezygapophyses; the neural canal is high, the attachment of the diapophyses is strikingly high and narrow and is found right in the middle of the length of the vertebra. In order to classify such a caudal vertebra, one must consider Dinosauria, Crocodilia, and Chelonia; other reptile groups cannot come into question. Crocodilian caudal vertebrae are usually rounded; in them and the Dinosauria, the root of the diapophysis is horizontal and not extended vertically; but both these points are possible in the Chelonia, which have procoelous articular facets, which is also a feature characteristic of Chelonia. Therefore this vertebra is probably chelonian. Yet I will not attempt to assign it to a particular family.
CHELONIAN REMAINS FROM BUSENDORF
Pl. VII, figs. 3 and 4
In the Geological Collection of Alsace-Lorraine at Strassburg are two pieces from the Upper Muschelkalk of Busendorf, Lorraine (collected by Prof. BENECKE in 1879). One of them most resembles the free rib end of Thalassemys, the other may be best compared with an armor fragment of the same genus. The rib end shows on the upper side wavy longitudinal grooves lying close together. It is 2 cm broad and still has the natural sharpened end. The fragment is 3 cm long. The broken-off edge begins the granulated sculpture that distinguishes the piece of armor. The granulation is very like that of Thalassemys. On the underside, the ribs and armor piece are smooth. The rib has a strong longitudinal keel and is in general flat. Also the armor piece shows an overall gradual thickening in one direction on the posterior side, and thus could well be a costal plate.
FEMUR inc. sed.
Pl. VIII, figs. 1 and 3
In his “Fauna der Vorwelt”, H. von MEYER has figured (pl. 65, figs. 5-7, p. 79) a left femur of curious shape, which comes from the Upper Muschelkalk of Bischmishein near Saarbruck. It belongs to the Strassburg National Collection. Three similar femora have been found in the same beds at Laineck near Bayreuth and are in the regional museum there; one of them is very much slimmer than the others, two are lefts and one is a right femur. The shaft is considerably S-shaped, in the distal half it is quite flat and thickened at the proximal end. The femoral head stands right-angled inward. The greater trochanter rises above the femoral head as a direct continuation of the shaft. The lesser trochanter is merely a reinforcement of the greater trochanter. The muscle scars are very clearly marked, particularly that of Mm. vastus medius and gastrocnemius. In a specimen from Bayreuth the epiphyses are missing, by which the proximal end retained the triangular appearance, as in Nothosaurus.
I know of such femora in no other reptile orders than Chelonia, although they do not fully agree with any known chelonian (1). The lesser and greater trochanters in particular are developed differently. However, I think that this structure of the femur indicates a transitional form between many anomodont types and the true Chelonia. The similarity with the femur of Dicynodon orientalis LYD. is very great, but the proportions of the femoral head and greater trochanter recall also the primitive anomodont forms such as Pareiasaurus. When both the trochanters divide and move away from each other laterally, as a result of greater adaptation to chelonian life, the perfect chelonian femur exists, which could be connected to the limitation of movement by the armor plates.
The cervical vertebrae and the rib and armor fragments indicate the same group of Chelonia, the size is also in agreement; considering the rarity of Chelonia in the Triassic, these remains could perhaps belong to the same genus. It is also suggested that the femora belong to it.
HUMERUS inc. sed.
Pl. V, fig. 6
An extremely peculiar humerus has been found in the area museum, Bayreuth, which comes from the main Muschelkalk there. Although its affiliations have not been determined, I have figured it because it is without doubt of great interest and perhaps thereby others will be put on the right track.
It is 12.5 cm long, the thickened proximal end is smooth, rather flattened, and more strongly curved to one side than the other, as the figure shows best. The shaft is only allusively curved and very thick (2 cm) in an only slightly greater breadth. A plate-like process (?radial process) extends for almost the entire length on the probable lateral side. The median (?) surface of the shaft has an edge on one side; this same side is entirely flat, while the other is greatly curved in cross-section. The distal end shows a clear separation into two condyles. On the lateral (?) side one notices a clear ectepidondyloid foramen. It is probably a right humerus, and then the intercondyloid foramen is on the anterior side. But very little idea can be formed as to the kind of animal which possessed the humerus. However, Pterosauria, Dinosauria, Crocodilia, and Sauropterygia are excluded. The foramen at the distal end is common to Proganosauria, Anomodontia, and Chelonia (also Sauropterygia, but which because of the particular form definitely cannot come into question). No more can be said than that the humerus belonged to a strongly built animal with long limbs. Without in any way being able to prove or advocate the chelonian nature of the animal in question, however I rank the humerus here as an appendix to the Chelonia.
6. PARASUCHIA (AND EUSUCHIA AS APPENDIX)
One of the most interesting groups of Triassic reptiles are the Parasuchia and Pseudosuchia, together with the theropod Dinosauria. Although according to the textbooks one belongs to the Crocodilia, the others forming a separate order, they have surprisingly many joint characteristics.
The skull of Parasuchia and Pseudosuchia is known to have many differences in the proportions and in the position of the openings, which is the reason for the separation of these two groups. The Parasuchia have enormously developed premaxillae that produce the crocodile-like long snout. The nasal openings (at least in Belodon) lie within the nasals and are bordered by no other bones. They are thus shifted abnormally far backward. This striking fact can probably only be viewed as a secondarily acquired characteristic; digging in the mud with its long snout was thus made very easy for the animal. The Pseudosuchia have a short snout and thus also terminally placed narial openings, in the usual way, which stand laterally and are surrounded by the nasal, maxilla, and premaxilla. The antorbital vacuity lies close behind them. The upper temporal fossae are closed behind (observed in Aetosaurus, Erpetosuchus, and Ornithosuchus), while they are open behind in the Parasuchia (Belodon and Mystriosuchus), and as a result of this nearly seem to be missing. Aetosaurus lacks the lower temporal fossa, Erpetosuchus and Ornithosuchus have well-developed upper and lower temporal fossae. Thus one sees that these conditions vary a lot. In armored animals, like the Crocodilia, this agrees in any case with the cervical armor on the one hand and with the different jaw length on the other. Therefore the different structure of the temporal fossae agrees in this (perhaps single) case with the way of life, for the degree of armament and the jaw length are directly dependent on it. The presence of a postorbital below the small postfrontal is common to both groups. In the previously listed characteristics of the skull, the Triassic theropod Dinosauria stand nearer to the Pseudosuchia (especially Erpetosuchus and Ornithosuchus) than the Parasuchia. In the position of the cranial nerve exits they even show surprising similarity with Belodon and the Eusuchia. The palatal roof shows interesting variety in the different forms. It is most closed in Belodon; the choanae lie far anterior, nearly directly below the external nasal opening; otherwise the palate is entirely closed. In Ornithosuchus the choanae are found rather behind the middle of the complete skull length, but the external nasal opening is far anterior so that a long choanal passage, surrounded by the palatine, nasal, and vomer, is necessary. Beside the internal choana opening and anterior to the ectopterygoid lies a rather smaller perforation communicating with the antorbital vacuity. In Erpetosuchus the choanae are placed far anterior (nearly meeting the short premaxilla), thus almost directly below the nasal opening. The palatines are long and narrow as also in Stagonolepis, whose choanae also lie close below the nasal openings. In Erpetosuchus the perforation in front of the ectopterygoid is very small. On the other hand the opening between the ectopterygoid, pterygoid, and jugal is large and present, of course, in all Parasuchia and Eusuchia. With regard to the palate, Aetosaurus seems to stand between Ornithosuchus and Erpetosuchus (only a single imperfect skull is available for this observation). The palate of Erpetosuchus is by far the most like that of the Triassic Theropoda (Zanclodon); it also has long, narrow palatines, a delicate ectopterygoid, and a broad, posteriorly directed pterygoid.
As for the vertebral column, the numbers of vertebrae in the separate regions (at least in several incomplete specimens of the Pseudosuchia in question) seem to amount to 8-9 cervical, 14-15 dorsal; throughout only two sacral vertebrae are present. These numbers agree with those for the later and Recent Eusuchia almost completely. In any case, the number of cervical and sacral vertebrae is characteristic of the Crocodilia. All Theropoda have at least three sacral vertebrae (as e.g. all Triassic) and 13 cervical vertebrae (Zanclodontidae). The vertebrae of Parasuchia are generally also morphologically more crocodile-like than dinosaur-like. The cervical vertebrae of Stagonolepis are shorter than tall; those of normal Theropoda are up to three times longer than tall; they are the longest in the whole body; in Parasuchia they are the shortest, therefore the neural arch is also developed entirely differently. The spinal processes on the dorsal vertebrae in the Parasuchia are long, narrow spines, but in the Pseudosuchia they are broad plates, as also in the Theropoda. In the Triassic crocodile groups the strong supports below the diapophyses are missing, whereas they are present in the Theropoda. The long sacral ribs of the Parasuchia and Pseudosuchia, horizontally expanded at the distal end, can never produce such a firm connection with the ilium as do the three extraordinarily strong and complicated sacral ribs of the Theropoda. This again is a purely crocodilian characteristic of the Parasuchia. In Belodon the spinal processes of the sacral and lumbar vertebrae are greatly expanded above, so that they have there a horizontal, laterally placed surface. The caudal centra are the most elongated. The Parasuchia first acquire broad spinal processes in the tail. The caudal centra lack the groove on the underside that is so characteristic of the Theropoda.