What SEELEY believed was the palate is no more than the skull roof from the inside. One sees the large orbits which are directed dorsally as in Procolophon, and not laterally as in Elginia. Between them remains only the narrow 12 mm bar that is formed from the frontals. Also the roof of the broad snout is clearly recognizable. If one knows where to look, one thinks one can recognize the sutures between frontals and nasals. The snout terminates broadly in front. No nasal openings are present up to there; thus they must lie terminally and be placed vertically as in Procolophon. The right maxilla is for the most part preserved, it bears seven blunt, conical teeth that do not differ essentially from those of the lower jaw. In the posterior half, the rather broad attachment for the ectopterygoid is to be seen. On the maxilla, the jugal and quadratojugal terminate backward, both of which however only show the underside. Below the snout are still stuck four intermaxillary teeth which are rather sharper than the others. From WIEDERSHEIM’s evidence (loc. cit.), the dorsal surface of the skull does not seem to have possessed the pitted sculpture of Elginia and Pareiasaurus, thus it also seems to have been like Procolophon with the exception of the spines.
The lower jaw, which is very like that of Pareiasaurus, still bears three teeth on the left and six on the right. They are ca. 4 mm high, conical, and with rather blunt points. The second most anterior tooth of the right mandible is flattened above. In the middle of each lower jaw Pareiasaurus has thick, downward-directed pegs; Procolophon and Sclerosaurus lack these.
The skull of Sclerosaurus is on the whole most similar to that of Procolophon, only exactly twice as large. It is however more strongly flattened, and indeed to the same extent as in Pareiasaurus; Elginia is much higher, and both certainly preserve the same habitus by means of the head spines. In Procolophon the appendages of two spine zones on the skull are noticeable, namely on the squamosal and on the quadrate. The so-called Permian Pareiasauria from North Russia have several rows of small spines on the skull roof; the postorbital region up to the quadrate is lobe-like, drawn deep down, and bears large spines (quadratojugal, jugal, etc.) (1). Of reptiles that show similar bone growths on the skull, but have absolutely nothing to do with Sclerosaurus, we may name the placodonts Placochelys and to a lesser degree Cyamodus, both also from the Triassic, further the Cretaceous Sauropoda, Triceratops and Agathaumas, and the Miocene chelonian Miolania.
The skull of Sclerosaurus is broader than long. The measurements follow:
Length of the midline 8 cm
Breadth from point of one quadrate to the other 10 cm
Ditto, with subtraction of spines 8.5 cm
Approximate height of occiput 3-4 cm
Height of snout 1.5-2 cm
From anterior edge of orbit to snout point 3 cm
Transverse diameter of orbit 12.5 cm
Length of the same over 3 cm
Length of left lower jaw ramus from medial edge of symphysis
to the angular 6 cm
Length of right lower jaw ramus in middle above 1 cm
Length of left lower jaw ramus in posterior third 3 cm
SEELEY has described the skeleton of Sclerosaurus so thoroughly that little more can be added.
In the vertebral column I count 18 presacral, three sacral, and 13 (+ 2 gaps = 15) caudal vertebrae, thus altogether 34 (36), which are clearly visible; but there are still 3-4 indistinct impressions at the end of the tail. I do not understand how WIEDERSHEIM and SEELEY counted 22 presacral vertebrae. A part of the cervical vertebrae is covered by the displaced portion of the skull on one side and by the shoulder girdle on the other. Of the visible vertebrae, at most the four most anterior can be ranked in the neck; therefore perhaps 2-4 short cervical vertebrae are inaccessible. Then there would be 14 dorsal and 6-8 cervical vertebrae. However the border between cervical and dorsal vertebrae cannot be identified with certainty, for the vertebrae appear so similar to each other. Only a slimming of the spinous process and a narrowing of the diapophysis occurs toward the front. Also the ribs cannot as otherwise be used as a criterion because the neck bears long ribs that do not really differ from the dorsal ribs, as e.g. in Palaeohatteria. Also in Pareiasaurus there is no clear separation between cervical and dorsal ribs; SEELEY counts ten cervical and eight dorsal vertebrae, the reverse seems more plausible to me (I have seen the original), but I dare not say anything on it with certainty. In Procolophon (1) there are 17 presacral vertebrae, including gaps for two vertebrae; I do not think many more were there. Sclerosaurus may have had 20-22 presacral vertebrae.
SEELEY has observed that the ventral side of the centra is always rather shorter than the dorsal side and from this concludes on the presence of intercentra. In fact this is also to be observed in the present cast between the 10th and 11th dorsal vertebrae. The centra are literally hourglass-shaped, in that the hollow cones almost entirely meet in the middle. On the ventral side they all have a sharp edge; their cross-section is triangular. The most posterior dorsal vertebrae become broad and thick. The upper arches of the dorsal vertebrae all have a long spinous process, particularly strongly flattened and expanded distally, which recalls Pareiasaurus very much. In the cervical vertebrae it is longer and slimmer than in the dorsals, both in Sclerosaurus and Pareiasaurus. The diapophysis is extraordinarily strong, broad, and long with double strutting on the underside, particularly in the posterior part of the dorsal vertebral column. Also the pre- and postzygapophyses stand very far laterally, which Sclerosaurus has in common Pareiasaurus and even more with Procolophon. In several places one seldom sees more clearly how the zygapophyses naturally fit into each other. A broad horizontal plate extends between the postzygapophyses. The ribs are single-headed with very high, vertically placed proximal ends.
The last dorsal vertebra has a 10 cm long transverse process, and the first sacral vertebra has one 20 cm long that thickens strongly at the distal end. It is only visible from the abdominal side (slab I). The second sacral rib is only imperfectly recognizable. The third has a length of 16 mm, but is much thinner than the first; it is a little thickened at the distal end and curves somewhat anteriorly. Procolophon likewise has three sacral ribs, which are greatly thickened at the distal end; the middle seems to be the weakest. The most anterior sacral rib of Procolophon likewise curves anteriorly, but is not stronger than the following caudal ribs and in this differs very much from the three sacral ribs. Pareiasaurus has four sacral ribs, but the fourth differs from the preceding in that it stands out from the axis of the vertebral column at a right angle, while the three others bend significantly posteriorly. In Sclerosaurus the vertebra following the third sacral vertebra has such a short centrum that it should not be regarded as the fourth sacral. A fragment of the right rib of this first caudal vertebra lies on the right pubis, from which is to be seen not only its considerable length but also its anterior curvature just like the first caudal rib in Procolophon and the fourth sacral rib in Pareiasaurus.
The 13 visible (+ 2 gaps = 15) caudal vertebrae are shorter and become continually shorter especially posteriorly, they decrease rapidly in size. It did not have more than 18 altogether. They are distinguished by the thin transverse processes that project far, as also in Pareiasaurus and Procolophon. 17-18 caudal vertebrae are visible in Procolophon and some still seem to be missing at the end; Pareiasaurus has 18-19 of them.
The shoulder girdle consists of the interclavicle, clavicle, procoracoid, and scapula. The piece of bone on slab I on the left beside the episternum seems to be a broken-off piece of the episternum. (The bone lying in front between the two lower jaw rami is the point of the broken-off right pterygoid piercing through from slab II.) The episternum and also coracoid and procoracoid have the greatest resemblance to these same bones in Procolophon. Both together are much larger than the scapula. The figures are clearer than a long description. These five bones (slab I) still lie in natural association. The right scapula is overturned so that the contact surfaces of the scapula and coracoid are rather displaced. On slab II the scapula is seen from lateral view, the coracoid from medial view. The scapula is very like that of Pareiasaurus, only the acromion here lies further posteriorly. On SEELEY’s illustration the latter does not stand out much. The clavicle articulates on the acromion as a broad, rather curved bone. The extremely broad and strongly built humerus recalls Pareiasaurus very much. For the arm, I refer to SEELEY’s description, only I might reverse the interpretation of the radius and ulna for the following reasons: the radius is round and proportionally thicker at the distal end than the proximal, this latter does not project on the other side of the humerus as SEELEY figures, by which the appearance of undue length (olecranon on SEELEY’s “ulna”) is made greater, but terminates under the distal blade of the humerus. The ulna is a flat, broad bone without an olecranon, and the proximal end is broader than the distal. The oblique joint surface at the articulation of the distal end of the ulna with the radius, characteristic of many reptiles, is visible; also the sharp edge of the articular surface is characteristic of the distal end of the radius. The position of the humerus also seems to me to justify this interpretation of the bones; as can be best seen on slab I, it must be rotated by 90° about the longitudinal axis to arrive at its natural position, and if radius and ulna change over by this, the radius keeps its rightful lateral position, the ulna the medial position.
Two rows of carpals are present. In the proximal row they consist of an ulnare and a radiale; of the four carpals of the distal row C. I lies on Mtc. I and II, C. II on Mtc. II and III, C. III on Mtc. IV, and C. V nearby, presumably on the missing Mtc. V. A pisiform still lies above it. Thus there are seven bones in all. The numbering of the digits of the manus must also be modified in accordance with the position of the radius and ulna. I can only see four digits remaining (cf. SEELEY, loc. cit.); the fifth, in any case present, is not visible. The first digit lies nearest the lower jaw, the fourth furthest away. As one sees on slab I, the first digit has two phalanges; the others each have three.
The pelvic girdle with both hind limbs is especially beautifully preserved. They come into view only on slab II. The right ilium is found in natural position and shows the lateral side; the left is overturned; one sees the attachment place of the sacral ribs. For a more accurate description refer to SEELEY. The left ilium is still in natural contact with the ischium and pubis: little else is to be seen of the latter. The ilium and ischium take a peculiar intermediate position between Pareiasaurus on the one hand and Dicynodon on the other. The femur is certainly built after the type of Pareiasaurus, but much slimmer. The tibia strikingly recalls Eryops, which in this and many other points resembles Pareiasaurus. The fibula is flat, rather broad, and strongly curved, as in Proganosauria and Archegosauria. The tarsals of the second row are not fossilized. The astragalus and calcaneum are represented by a single large, triangular bone with clearly separated articular surfaces for the tibia and fibula, exactly as also in Pareiasaurus. Five digits are present. Metatarsals I and II are compact and short, the others slimmer. There are two phalanges in the first digit, three in the second, three in the third, and four in the fourth; only the metatarsal of the fifth is there.
As for the dermal armor, SEELEY has already “seen two or three rows of small irregular scutes which lie between the ribs” (1). The complete cast has also provided more clarity on this point. Six longitudinal rows of armor plates extend from the sacrum to the beginning of the neck. On both sides of the spinous process, corresponding to the number of vertebrae, is a thick, oval bone plate generally 12 but up to 20 mm long, and ca. 7 mm broad; each of them bears a longitudinal ridge and has pitted sculpture on both sides of it. All these plates are directed obliquely posteriorly and extend from the spinous process always above the postzygapophyses. The next row consists of rather smaller, rounded plates with the same sculpture; they alternate with the former and are thus found between the ribs. The third row of pairs is quite incomplete and consists of small ossifications, which however also show the pitted sculpture. Behind the sacrum the armament abruptly ceases and decreases gradually toward the neck. Also the flattened dorsally and greatly expanded spinous processes of the dorsal vertebrae have shown the sculpture of the armor plates.
Now for the first time, after having discussed the two most important pieces, the statement made above, that Aristodesmus and Sclerosaurus are identical, should be proved. The original of Sclerosaurus comes from the Upper (green) Bunter Sandstone of Warmbach near Rheinfelden, thus also from the region of Basel (found 1856). 11 dorsal vertebrae and 12 (left) ribs as well as parts of the ischium and pubis are preserved on the one (lower) slab. The second slab, fitting on it, shows the dorsal armor from the inside; the eight split vertebrae, whose upper arches and sometimes the neural canal are visible, as well as the pubis, show surprising similarity with the piece from Riehen; the zygapophyses project far laterally; the horizontal plates between the postzygapophyses and the spinous processes are built exactly the same, as are the ribs. The femora are to be seen from the ventral side; thus they form a good supplement to the block from Riehen which only shows the dorsal side of both femora. As might be expected from the structure of the rest of the skeleton, the femur is formed exactly after the type of Rhopalodon and Pareiasaurus, only rather slimmer. The fragments of the pelvis may be identified by their structure. The dorsal armor is preserved better in situ than in the Riehen block. One recognizes (from the ventral side (2)) the elongated large side plates, on which lie one or two rows of smaller plates in the middle and upper region of the back. In about the eighth dorsal vertebra (reckoned from the rear) the median plates divide into three separate plates. Nothing clear is preserved of such median plates in the Riehen block.
Of the armored reptiles, apart from Chelonia and Parasuchia of course, there have been known hitherto in the European Triassic only Placodontia and Rhynchosaurus; to these is added a find from the Wellenkalk of Gogolin (see below), which Dr. W. VOLTZE has just described, and Sclerocephalus from Warmbach and Riehen. Armored Anomodontia are unknown from South Africa; Pareiasaurus (bombidens, baini, etc.) certainly possesses the same thickened spinous processes but no armor. On the other hand, the Permo-Triassic Pareiasauria from the Suchona, which Prof. AMALITZKY has discovered, are strongly armored on the back and sides, and to increase the resemblance with Sclerosaurus further, the skull is also furnished with strong spines and outgrowths.
Whether Sclerosaurus is related to Sphenosaurus H. v. M. from the Bohemian Permian sandstone, and to Parasaurus geimitzi H. v. M. from the Kupferschiefer of Thuringia, as von MEYER suggested (loc. cit., 1859), cannot be established for the present.
Procolophon, Sclerosaurus, and Elginia form a closely related group that is most nearly related to Pareiasauria. That Procolophon is the most primitive representative of Procolophonia arises from the absence of the bizarre head spines. Which of the two families is more primitive cannot easily be said even if the Pareiasauria most resemble Stegocephalia by their skull sculpture.
VERTEBRA OF PLACODUS GIGAS
Pl. V, Fig. 1 and 2.
A piece of rib of Placochelys agrees completely with the ribs found on two vertebrae from Bayreuth and Göttingen, and thus I must hold these are also those of a placodont. Prof. JAEKEL, to whom I sent a drawing of the Göttingen vertebra, writes to me saying that he also agrees with me. The centrum is rather elongated and narrows ventrally in cross-section; constricted in the middle and almost boat-shaped, it recalls the “thecal” thoracic vertebrae of Placochelys. The articular surfaces are only hollowed a little. In the centrum, a separate long canal of 4-5 mm diameter and round cross-section passes under the neural canal. It recalls the deep openings on the articular surfaces of the vertebrae of Pareiasaurus and Tapinocephalus, only they are separated there by a thin septum, which cannot be established here without a longitudinal section. This feature also points very much to the anomodont nature of the Placodontia. The neural canal is proportionately large and deep. Ribs and transverse process are completely ossified so that a suture can no longer be recognized. The ribs are triangular in cross-section with one flat side below. The lower edge thickens proximally to a broad scar, below which the neurocentral suture is still recognizable. The transverse process or the rib bends first rather dorsally, then runs for a bit horizontally and bends in a broad curve again ventrally. The sharper edge of the rib lies on the posterior side, and the rather more rounded one on the anterior side of the dorsal surface. The spinous process is low and narrow but strong and directed posteriorly. It rises no higher than the highest arch of the rib so that thus the dorsal armor shown by JAEKEL lies quite close above it.
Because Placodus gigas certainly chiefly occurs in the German main Muschelkalk, it is likely that both the vertebrae belong to this species.
PSEPHODERMA AND PSEPHOSAURUS
The discovery of JAEKEL’s Placochelys with complete dorsal armor now at last allocates the isolated pieces and armor plates described as Psephoderma and Psephosaurus to their correct place; formerly traces in them were thought to be sometimes nothosaurian, sometimes chelonian. They belong to Placodus. Just as the skull of Placochelys is more strongly sculptured than that of Placodus, so also are the armor plates of the former high and pyramidal, but those of the latter are almost entirely flat. However the outline is exactly the same, and also the central elevation is more or less recognizable. Also the arrangement of the few larger plates between numerous smaller ones is exactly the same in both. In Psephoderma alpinum only small plates of the same size are preserved, but P. anglicum and especially Psephosaurus suevicus correspond completely to Placochelys. It might in fact be a question of different species or even genera of Placodontia.
Psephosaurus suevicus (1) is found at its best in thc Upper Leettenkohle of Hoheneck near Ludwigsburg, and Psephoderma alpinum in the Upper Bavarian Dachsteinkalk. The small plates as from Hoheneck are also found isolated in the Rhaetic of Aust Cliff near Bristol, such as Psephoderma anglicum H. v. M. (2), and in the Zanclodon Bed of Schönthal near Basel (see Pl. V, fig. 3) together with Capitosaurus, Belodon, and Gresslyosaurus. Such plates are also found in the Rhaetic of Lombardy (3).
ANOMOSAURUS gen. nov.
Pl. IV, fig. 3 and 4; Pl. VI, fig. 4; Pl. VII, fig. 6.
Among large suites of nothosaurid bones a number of vertebrae have been found that differ immediately by their very deeply amphicoelous articular surfaces.
These vertebrae differ among themselves in several points, but besides the deeply amphicoelous articular facets all have a long spinous process and hyposphene; they would probably represent different species of one genus.
They are relatively most frequent in the Schaumkalk (Middle Muschelkalk) of Freiburg am Unstrut (province of Saxony). I have seen seven specimens of them in the University Collection at Halle, one from the region of Quedlinburg and eight from Altenberg near Naumberg. Such ones also occur in the Lower Muschelkalk at Gogolin and Naklo in Upper Silesia and at Sulzbad, Upper Alsace. Then four such vertebrae from the Upper Muschelkalk are present, two of which come from Bayreuth, one from Neckarvaihingen, Württemberg, and one from Mühlhausen, Thuringia.
The vertebrae from the Lower Muschelkalk of Gogolin (Dadocrinus horizon, 3 m above the Röth) are so like the better-preserved ones from the Saxon Schaumkalk that I only describe the last one in more detail but draw attention to several smaller differences.
The centrum is rather longer or as long as high. The articular facets converge rather posteriorly and are so deeply amphicoelous that the hollow cones nearly meet. The deepest point of the latter clearly lies considerably below the median longitudinal axis. In most vertebrae the neural canal is very high but narrow, sometimes wider above, sometimes below. Therefore the upper arch ventral to the diapophysis in many vertebrae has a larger diameter than the centrum. Characteristic of the upper arch are the long transverse processes, thickened at the distal end. In most vertebrae the spinous process is high and narrow, its base is moved a little posteriorly; however in some it is short and quite thick; it is probably a question of different species, or sometimes of different sections of the dorsal vertebral column. The pre- and postzygapophyses are moderately oblique to flat and are strongly developed. In some the postzygapophyses have clear longitudinal grooves and a lateral curving of the lateral edge for firmer control of the joint. A triangular pit is found in the middle behind the prezygapophyses anterior to the base of the spinous process. A very strong hyposphene is always present ventral to the postzygapophyses; it reaches deep down and becomes very broad ventrally. Seen in posterior view, it is almost an equilateral triangle with the point directed dorsally. The vertebrae from Freiburg and Querfurt from the Schaumkalk vary in overall height from 6-12 cm; the larger are more frequent. The pieces from Gogolin are only 6-8 cm high. The transverse processes are hollow in both, as one occasionally sees on fracture surfaces. The transverse process is allusively doubly strutted from below, most clearly in the pieces from Gogolin.
Before me lie four considerably larger vertebrae from the Upper Muschelkalk. The quite thick centra are rather longer than high. The edge of the articular surfaces is tilted over considerably toward the sides. The articular surfaces themselves are deeply amphicoelous; both hollow cones meet in the middle and probably communicate by a small hole. Dorsal to the centrum the upper arch is very high. The border between both is recognizable as a clear suture. The posterior opening of the neural canal is about 2.5 cm high and 1.5 cm broad in a vertebral length of 3.2-2.4 cm. The centrum from Neckarvaihingen has a breadth of 3.5 cm posteriorly; but the dorsal surface of the transverse process is 7.8 cm above the ventral, and the spinous process measures a further 6 cm, thus giving a total height of 13.5 cm. In one of the vertebrae from Bayreuth the transverse process is 6.5 cm long. The transverse process is flat dorsally, begins very thick and broad with only a trace of two- or three-fold strutting from below (this is rather different in the four vertebrae), then becomes thinner again with an approximately triangular cross-section, but thickly club-shaped at the distal end. The spinous process, which is preserved only in one vertebra from Bayreuth and that from Neckarvaihingen, projects vertically upward; one is rather thickened at the distal end. Both display a narrow, projecting longitudinal ledge on the dorsal side of the spinous process, as is found in Nothosaurus and Pareiasaurus. At the root of the spinous process and between the postzygapophyses lies a deep, flat-bottomed cavity. The postzygapophyses in the different vertebrae project laterally in different positions corresponding to their position in the skeleton, however not to any extent, and all ascend more or less outward. In the middle ventral to them an extraordinarily strongly developed hyposphene is formed. This projects as far posteriorly as the postzygapophyses and has articular surfaces that clearly converge dorsally, as is the case in Zanclodon and related Triassic Dinosauria in the posterior dorsal vertebrae. The prezygapophyses are not preserved but it is easy to decide their structure from the form of the postzygapophyses.