oe Se a a a ne ia Ke oes er ee . Pakteee> ee : Bee z A - eee : ¥ * ae 7 3 = Pe Sa re eee a ee ; J R “ : > 2 3 She : mete pe J — x oe Se it SO or ’ = cee - 7 mee tes es See leg Re Ee '' BERKELEY LIBRARY UNIVERSITY OF cauroria _/ SCIENCES LIBRARY '' '' '' '' ''A MONOGRAPH ON THE FOSSIL REPTILIA OF THE MESOZOIC FORMATIONS. BY. SIR RICHARD OWEN, K.C.B., DCL. 2 he. FOREIGN ASSOCIATE OF THE INSTITUTE OF FRANCE, ETC, ETC, LONDON: PRINTED FOR THE PALHONTOGRAPHICAL SOCIETY. 1874—1889. '' ''PREFACE TO THE REPTILIA OF THE MESOZOIC FORMATIONS. In collecting the materials for a work on the Teeth of Animals, I made my first acquaintance with some previously unknown Feptilia. A genus was defined for which the name Pliosawrus was proposed, and a characteristic tooth was figured in pl. Ixviii of the volume of my ‘ Odontography,’ published in 1840. The sub- sequently discovered characters were described and illustrated in the Monograph on the Reptilia of the Kimmeridge Clay, in the volume of the Palzontographical Society for 1859, issued in 1861, pp. 15, 16." In the same Monograph of the volume for the year 1860, issued in 1863, the huge teeth of the Pliosawrus grandis were described (p. 27), and figured of the natural size in the folding Plate (Pl. XII). Pursuing the researches, so liberally illustrated in the years 1861 and 1863, further results, enriched by the additional species, Pliosaurus trochanterius and Pliosaurus Portlandicus, were given with similar illustrations in the Paleeonto- graphical volume for 1868, issued in 1869, pp. 1—12. The publication of the first part of the Monograph on Scelidosawrus (Volume for 1859) had the usual result. The active and careful observer, JAmMus Harrison, Hsq., to whom I was indebted for the evidences of the Scelidosaur, supplied me with the subjects for a second part of the Monograph, in which an almost entire skeleton of this extinct British Reptile was described and figured in eleven plates of 4to, and of larger (folding) size (Volume for 1860). My friend, encouraged by the publication of a description of the first indication of the extinct reptilian Scelidosawrus, pursued with increased vigour 1 This was associated, in the Pal. Soc. Vol. for 1859, with the Monograph on Scelidosaurus. 87476 ''PREFACE TO THE and zeal his quest of fossil remains in the cliff of Lower Lias at Charmouth, where he first came upon that indication, and he kindly complied with my desire to send the block of matrix indicating an included bone. I successively received from him twelve additional blocks, and superintended the careful operations of our skilled Museum masons, the results being an almost complete skeleton of the extinct Reptile. Other fossil evidences subsequently submitted to me served for a definition of characters of a Crocodilian genus Poikiloplewron,’ and of a Dinosaurian genus Chondrosteosaurus.:| The volume issued by the Paleeontographical Society for the year 1876 contained the Monograph on these extinct Reptilia of the British Wealden and Purbeck Formations. 165, 1b, 4 In which Cuvier describes the tail as “‘trés-courte, trés-gréle, et l'on n’y compte que douze ou treize vertebres.”’—Tom. cit., p. 368. ''4 FOSSIL REPTILIA OF THE The most striking characteristic difference from that species is the vastly superior size of the seemingly allied Flying Dragons from the British Chalk and Wealden. In the restoration of the skull of Pterodactylus compressirostris (‘ Monograph on the Fossil Reptilia of the Cretaceous Formations,’ 4to, Part I, 1851, Palsontographical Society’s vol. for 1851) I ventured to assign to the mandible a length of 14 inches 9 lines (Pl. XX VII, fig. 5). This species was represented by two portions of the upper jaw (ib., Pl. XX VIII, figs. 8, 9, 10) from the Middle Chalk of Kent, the longest portion being 4 inches in length. Of the nearly allied species, represented by three portions of the lower jaw, discovered by Samunn H. Becxuizs, Esq., F.R.S., F.G.S., in the Hastings series of the Wealden Formation, west of St.-Leonard’s-on-Sea, the restoration figured of half the natural size in Plate II, fig. 8, gives a mandible of between 14 and 15 inches in length, and this on the most moderate estimate of the length of the symphysis. In a sketch of a restoration of the jaw, sent to me with the fossils by Mr. Beckles, the length of the symphysis, which he assigns on the basis or analogy of that in Collins’s or Cuvier’s Pterodactylus longirostris gives a total length of 18 inches to the mandible. The parts obtained by Mr. Beckles are of one and the same lower jaw ; and, as an extent of above 2 inches of both rami are maintained by a portion of matrix (PI. II, fig. 8, m) in their natural relative position, the angle of convergence is shown; and this affords a ground for estimating the length of each ramus from the articular surface to the hind part or border of the symphysis at 18 inches, the extent beyond remaining conjectural. The specimen includes a portion of the left ramus, 9 inches 8 lines in length (of which the anterior 7 inches are given in Pl. II, fig. 1), and two portions of the right ramus, of which the dentary part measures 5 inches (ib., fig. 2) in length, the articular part 2 inches (ib., fig. 5). The portion of the left ramus includes the dentary element (ib., fig. 1, and fig. 4, 32) with the anterior part of the splenial element (fig. 4, 31). The dentary includes ten of the hinder sockets (ib., fig. 1, 1, 2, 3, 4,5), of which the five foremost (ib., 6, 7, 8, 9, 10) retain more or less of their teeth. As the number of these which may have been present in the fore part of the Jaw is unknown, I count those which are preserved from the hind end of the series forwards. Prolonging the alveolar border according to a moderate estimate of the symphysis, and supposing the teeth to maintain the same intervals, about eighteen may be assigned to each ramus. ‘The border of the hindmost socket (fig. 1, 1) is not prominent as in the rest, and there is room for doubt whether'the oval vacuity which indicates the hindmost tooth really contained one. There is none, however, with regard to the next socket (ib., 2), for this, like the antecedent ones, rises at its outlet above the level of the surrounding part of the bone. It projects from the outer part of the thick, transversely convex, upper border of the dentary, and the course of the cavity shows that the tooth must have inclined some- what outward as well as forward from the perpendicular. The long diameter of the ''WEALDEN FORMATIONS. | 5 outlet is in the axis of the jaw, and is 13 lines (8m.m.). The short or transverse diameter is 1 line (2 m.m.). The interval between this socket and the one marked 3 is 5 lines (10 m.m.). The prominent outlet of the socket. 3 gives 5 m.m. in long diameter and 3 m.m. in short diameter; these dimensions with that of the interval are repeated to the socket 6, which retains its tooth. The exserted crown of this is 5 lines (10 m.m.) in length; it is conical, acute, gently curved, with the convexity outward and forward. The apex of the next tooth in advance is broken off, but the basal half is better cleared out of the matrix, giving an antero-posterior breadth of its issue from the socket of 5 m.m. The teeth in the sockets 8 and 9 are better preserved, and show well the characters of the mandibular ones in the present species. As in Pterodactylus longirostris, the teeth of Pter. sagittirostris are subsimilar, divided by nearly equal intervals, these being somewhat wider than in Pter. longirostris,' rela- tively shorter than in Péer. crassirostris,’ and more resembling in disposition the indica- tions given by the sockets in the portion of upper jaw of the Cretaceous Pterodactylus compressirostris. The dentary bone supporting the above-numbered teeth is slender and subcompressed ; its depth is given in figs. 1, 2, and 4 (nat. size) ; its thickness is shown in fig. 3. This is the same at both upper and lower borders, which are similarly rounded off ; it is less half way down, owing to the concavity, vertically, of the inner surface of the ramus (ib., fig. 4). The outer surface (fig. 1) is nearly flat; it is traversed lengthwise by a linear impression, which is 5 m.m. below the upper border at the hind end of the pro- portion of the ramus figured in fig. 1, and is 7 to 8 m.m. below the outlets of the sockets of the teeth 7—9. ‘This linear impression does not indicate a suture. The ramus slightly increases in thickness, with a gain of convexity externally and a deeper concavity internally (both being in the vertical direction), at the fractured end (ib., fig. 1,32) nearest the symphysis. At the opposite end the angular element (ib., fig. 4, 30) forms the inwardly prominent lower border; the line between which and the thin flat splenial forms (ib., ib., 31) is clearly sutural. The portion of the right dentary preserved (PI. II, figs. 2, 3) answers to that containing the sockets of the teeth numbered 2—9 in fig. 1. ‘There is the same obscurity or lack of demonstration of a socket or tooth behind the socket 2. The bases of the teeth are preserved in the sockets (numbered 2—6), and partly project from the sockets 2 and 3, but the sockets 7, 8, 9, are vacant. The articular portion of the right ramus (figs. 5, 6, 7) lacks the prominent, backwardly directed, end of the subangular (30). 1 Monograph above cited, Pal. Soc. vol. for year 1851, Pl. XXVII, fig. 1. a My, 1b, figs. 2 and 3. ''6 FOSSIL REPTILIA OF THE The articular concavity (fig. 6, a) is transversely extended, chiefly by the production of its inner wall (ib., 4); its upper boundary is sinuous by a backward production of its mid part ; the upper surface in advance of the cavity is smooth and gently convex across it narrows to the ordinary thickness of the ramus about an inch and a half in advance of the articulation. In this extent it shows no trace of a coronoid rising. ‘lhe inner surface is impressed with a deep longitudinal cavity (ib., fig. 7, ¢). According to the usual proportions of the upper and lower jaws of Pterodactyles, the premaxillary of the present species must have been twice, or nearly twice, the depth or vertical diameter of the portion of that bone of Pterodactylus compressirostris (figured in Pl. XXVIII, fig. 8, of Monogr. cit.). Both upper and lower jaws of Pterodactylus saguttirostris must have been broader, less compressed, than in the Cretaceous Péer. compresstrostris. ‘The value of a symphysis mandibuli, with its natural anterior termination, like that of the Gault species (Plerodactylus Daviesii), is its demonstration of a character determinative of the genus of Pterosaurian. Were it produced into a slender-pointed edentulous style, or ‘rostrum,’ it would lead to a reference of the species to Von Meyer’s genus Rampho- rhynchus and Family ‘ Sudulirostres.’ The opposite extreme is shown by the thick obtusely terminated snout, as if it had been cut short, giving the character of the Ptero- saurian family Truncirostres? The species of this family which have the foremost pair of teeth projecting forward in the upper jaw from the truncate surface at a higher level than the alveolar border form the genus Coloborhyuchus’ B.—Coloborhyuchus clavirostris, Owen (Plate I, figs. 1—4). In two species of these large Pterodactyles from the Cretaceous series, viz. Colobo- rhynchus Cuvieri, from the Middle Chalk of Kent,* and Coloborhynchus Sedgwickii,’ from the Upper Greensand of Cambridge, the anterior pair of teeth of the upper jaw project, as 1 ©Paleeontographica,’ Heft i, 4to, 1846. 2 Mihi (Truncus, cut short). 5 kodoBos, stunted ; puyyos, snout. I have no evidence, and Mr. Seeley gives none, of such departure from the Pterosaurian type of hand as would justify the term Ornithocheirus proposed by Mr. Seeley for Pterodactylus Sedgwicki in his ‘ Ornithosauria,’ 8vo, 1870, p. 112; or the term Ptenodactylus previously proposed by Mr. Seeley for the same Pterodactyle in the ‘ Index to the Fossils, &c., in the Woodwardian Museum,’ 8vo, 1869, p. xvi. * ‘Monograph on the Fossil Reptilia of the Cretaceous Formations,’ Part I (Paleeontographical Society's volume for year 1851), p. 88, Pl. XXVIII, figs. 1—7. ® Ib., ib., 4to, Supplement No. I (Paleontographical Society’s volume for year 1857), p. 2, Pl. I, figs. 1, e—d. ''WEALDEN FORMATIONS. a in the present species, from the fore part or end of the premaxillary, and are directed forward with a slight downward curve. . In a still larger species (Criorhynchus simus'), from the Upper Greensand of Cambridge, the foremost pair of teeth project from the under surface of the fore end of the premaxillary, and are directed downward like the following teeth. The fore end of the premaxillary was fortunately entire, showing a flattened or feebly concave tract corre- sponding to the part bored by the anterior alveoli in Coloborhynchus. Some reserve may be prudently entertained as to whether a pair of teeth so anomalously located as in Coloborhynchus might not be shed without replacement by successors; and the genus Criorhynchus is to be accepted with this reserve, which future discoveries may dissipate. The manifestation by a ‘truncirostral’ Pterodactyle of the Wealden, and by another from the ‘Greensand,’ of the produced and unopposed pair of teeth from the front surface of the muzzle, have dissipated the doubts as to its accidental and individual character which legitimately attached to the first specimen, from the Chalk, in which it was observed. Coloborhynchus clavirostris is, at present, represented by the fore part of the upper jaw of a Pterodactyle (Pl. I, figs. 1—4) from the Wealden, of equal size with Crio- rhynchus simus, {rom the Upper Greensand, but in which the small anterior pair of premaxillary teeth project from the front surface of the bone, and at a greater elevation above the palate and the sockets of the second pair, than in Coloborhynchus Ouviert or Colob. Sedgwickit. . The flattened fore part of the premaxillary (ib., fig. 2) is broader and of less height in Coloborhynchus clavirostris before the narrow upper surface (g) begins to slope backward to the upper contour of the cranium. The anterior median depression (4) is shorter vertically and deeper in Colod. clavirostris, where it is below the alveoli of the teeth ( a, a). The convexities (¢, 7) on each side of this depression are the fore parts of the sockets of the second pair of teeth, not of the first pair, as in Criorhynchus simus (Monog..cit., Pl. I, fig. 3, a). The sides of the fore part of the premaxillary in Coloborhynchus clavirostris converge, with a slight vertical concavity, to the narrow but obtuse upper border of the skull; the same sides also converge as they recede in a slighter degree, but so that the breadth of the upper jaw behind the sixth pairs of teeth (ib., figs. 1 and 4, ¢ ¢) is less than two thirds the breadth behind the second pair of teeth (ib., 4, 4, fig. 4), whence the name clavirostris (‘ club-snout ’) proposed for the present formidable species of Wealden Pterodactyle. The fore part of the bony palate, between the teeth of the second pair (ib., fig. 4, 4, a), is transversely quadrate and flat (ib., fig. 4, 4). Behind this tract the mid third only of the palate retains its level, the two side thirds subsiding (as it seems when looked down upon) into shallow channels, which expand and are continued into the slope rising 1 Ib., ib., 4to, Supplement No. III to < Monograph on the Fossil Reptilia of the Cretaceous Formations,’ PTEROSAURIA, p. 2, Pl. I (Paleontographical Society’s volume for year 1858). ''8 FOSSIL REPTILIA OF THE to the sockets of the fifth (¢) and sixth (J) teeth, leaving the prominent narrow mid tract to represent, as it were, the bony palate; this part has projected below the level shown between the fourth pair of teeth, behind which the thin compact wall is broken away, exposing the widely cellular structure. A similar abrasion affects the upper border of the skull (beyond 3, fig. 1, Pl. I). The first or anterior pair of teeth (ib., a, a) bears the same relations of size to the second (2) and third (¢) pairs as in Criorhynchus simus, and may be homologous with the first pair in that species (Monog. cit., Pl. I, fig. 1, 2) though differimg so much in position and direction. In the present specimen of Coloborhynchus clavirostris the crown of the. first, as of the second, tooth is broken off at the outlet of the socket. The shape of this outlet is a full ellipse (Pl. I, fig. 2, a, a); the long diameter, of 8 m.m., is vertical ; the short diameter, of 64 m.m., is transverse. The size and shape of the five following teeth are shown in fig. 1; for, as is common in Pterodactyles, the sockets open obliquely upon the outer part of the alveolar border, and in the present species with a nearer approach to verticality than is usual (compare Pl. I, fig. 1, with Pl. I, fig. 1, of the Monograph of 1858, in the Pal. Soc. vol. issued in 1861). The present unique evidence of one of the most extraordinary of the extinct order of volant Reptilia was discovered by S. H. Beckles, Esq., F.R.S., in the Hastings Series of the Wealden. § 3. Prrrosavria or THE Kimmuripen Cray. A.—Pterodactylus Manselii, Owen (Plate I, figs. 10, 11, 12, 20, 21). Figures 10 and 11 of Pl. I show front (thenal) and back (anconal) views of a mutilated proximal end of the left humerus of this rather small species of Pterodactyle. The reniform articular surface of the head of the humerus (fig. 12, ¢) is somewhat less extended transversely in proportion to its breadth than in a similarly sized species from the Lias (Pteroductylus Marderi, ib., fig. 9); its anconal convex border has a bolder curve. There is no indication of a pneumatic orifice on this surface, as in Birds. The pectoral process (4, figs. 10 and 11) stand out more abruptly from a less extended base (compare with 6, figs. 7 and 8, Pl. I). - The proximal end of the first phalanx of the fourth or wing-finger, which is the subject of figs. 20, 21, 21*, corresponds in size with the portions of humerus above described, near which they were discovered. ‘The olecranoid process (ib., fig. 21, ¢) led observers of the first discovered specimens of this eminently pterosaurian bone to regard it as an ulna. Upon this process is extended part of both the outer and inner concave articular surfaces, so placed as to resemble the two divisions of the ‘greater sigmoid ''KIMMERIDGE CLAY. 9 cavity ’ in the human ulna, the curve and depth of which surfaces is thus augmented, and therewith the security of the flexible joint on which the chief movements of the bat-like wing take place. ‘he outer surface, shown in fig. 20, is of less extent, in long diameter, than the inner articulation (ib., fig. 21, a); a larger proportion of it is supported by the olecranoid process ; and it is better defined along the margin next the longer concavity (2). Nevertheless, the smoothness of the surface of the ridge, dividing the concave articulations, suggests that they combined to form a single synovial hinge-joint or ‘ginglymus,’ limiting the movements of the bones so articulated to one plane, and combining freedom and extent of motion in that plane with great strength of joint. The summit of the olecranoid process in the present specimen shows a rough flattened surface, not a fracture, suggestive of the contact of a sesamoid, probably lodged in the tendon inserted into the phalanx (ib., fig. 21%). B.—Pterodactylus Pleydellii, Owen (Plate I, figs. 15, 16, 22, 28, 23%). The portion of the fossil skeleton of the small species of Kimmeridgian Pterodactyle here figured is the distal half of the left humerus. It shows the generic obliquity and superiority of size of the articular convexity for the head of the radius (ib., fig. 15, a,) ; that for the ulna has suffered fracture, and part of it is lost with the ulnar tuberous ridge ; but sufficient remains to show its hemispheroid form, and the mere chink dividing it from the radial condyle instead of the groove which is here seen in Birds. The flexor (?) ridge, leading to the broken tuberosity, extends more forward than in Pterodactylus Duncani (ib., fig. 13), and contributes to a deeper concavity above the condyles on the thenal aspect of the distal expansion of the humerus. ‘The transverse ridge behind the condyles is confluent therewith at its extremities, the defining groove not being developed (ib. ib., 16’). The broad shallow canal for the ‘ triceps’ tendon marks the anconal surface of this expansion (ib., fig. 16). ‘To the same species of Pterodactyle may probably belong the proximal end of the smaller example of the first phalanx of the fourth or wing-finger, of which I have given two views in Plate I, figs. 22, 23, and 23*, to contrast with those of the same bone and part of Pterodactylus Manselit. The olecranon process in Pterodactylus Pleydellii is relatively longer and more incurved; its apex is not truncate; it is more compressed; has a smaller and lower posterior tuberosity, and a smaller basal tuberosity. The longer concave articulation is similarly extended upon the anterior angle. From the tuberosity at the corresponding or lower end of the shorter concavity a ridge is continued down the bone, giving a triedral form to the shaft as far as it is preserved in this and the previously described specimen (figs. 20, 21). The bony wall of the shaft is thin and compact, the air-cavity large, and in one specimen occupied by crystallised calcite. b ''fo FOSSIL REPTILIA OF THE - The two narrower sides are concave or flat transversely ; the broader side is gently convex; it shows, in both species (figs. 21, 28), a longitudinal linear impression, which may indicate a confluent rudiment of a fifth digit. To the above-described, well-defined, trochlear or ginglymoid joint were adapted the two obliquely disposed condyles of the distal end of the metacarpal of the fourth or - wing-finger. I have pleasure in contributing this mite of testimony to the unremitting attention to the fossil evidences of Kimmeridgian Vertebrates, discovered from time to time, on his estates by John C. Mansel-Pleydell, Esq., F.G.S., of Longthorns, Blandford, and to the wise liberality by which they have uniformly been deposited in the National Collection, where inferences and conclusions from their study can be tested by Paleeontologists. C.—Pterodactylus, sp. incert. Two specimens of the carpal bone, provisionally referred in my Monograph of 1857 to a Pterosaurian ‘ wnciforme,? are figured in Plate I, figs. 24—27. ‘They were both obtained from the “ Kimmeridge Clay,” at Weymouth, Dorsetshire. The distal surface of the smaller specimen is given in figure 24; they show the larger concavity (a), and the smaller one (2), adapted to the two proximal condyles of the metacarpal of the wing-finger. The thenal border of the bone is the thinnest, and is produced at each end into a short process; the anconal border of the bone is thicker, especially where it supports the smaller and outer articular metacarpal concavity. The proximal surface (ib., fig. 25) is also divided into two principal articulations, but the larger one (-) is subdivided into a concave and a flattened facet. The smaller concave surface (g) is next the outer and thickest end of the bone. The subject of figs. 26 and 27 is the homologous bone, and from the forelimb of the same side, but it shows modifications that plainly bespeak its having come from a distinct species of Pterodactyle. The outer subhemispheric concavity of the proximal surface (ib., fig. 27 g) is relatively larger, as is likewise the flat facet at the inner part of the larger surface (c). The two condylar concavities (a and 4) on the distal facet are more equal than in the larger unciforme. Both bones exemplify the definite, well-marked, or finished character of the articular surfaces which characterise the bones, especially those of the wing, of the volant Reptile. I would still be understood to be guided by considerations, not beyond probability, in referring this well-marked bone to the distal row of the carpal series ; for I have not 1 Paleeontographical volume for year 1857, issued in 1859, plate iv, figs. 5 and 6. ''OOLITIC FORMATIONS. 1] yet had the opportunity of studying a Pterosaurian carpus or tarsus in so well-preserved and undisturbed a condition as would enable me, with certainty, to determine the homologies of its constituent bones. § 4. Prerosavuria or THE Great Oo.its. A.—Pterodactylus Kiddii, Owen (Plate I, fig. 17). The first phalanx of the wing-finger (fig. 17), referable to this species is somewhat stouter, but about one eighth shorter, than that bone in the Pterodactylus suevicus, Quensted,! from the Lithographic Slate of Wirtemberg. It indicates a species with a more powerful, though, perhaps, less elongate, wing. The groove for the flexor tendon of the fourth digit, bounded by the prominent thenal extensions of the two articular grooves, is well marked. The extensor process (ib., c) has a relatively longer basis than in the Kimmeridge specimens. A rough groove or linear depression beginning about an inch beyond the proximal articulation, and extending as far down the fore or thenal surface of the shaft of the bone, indicates the extensive attachment or insertion of that tendon. The shaft is subtriedral, the anconal side being the broadest; it becomes flattened towards the distal end, which expands unequally towards the ulnar side, and affords an oblong, moderately developed, concavo-convex surface for the second phalanx of the wing-finger. This bone, from the Stonesfield Oolite, is slightly crushed. B.—Pterodactylus Duncani, Owen (Plate I, fig. 18). The first phalanx of the wing-finger, referred to the above species, is of the left wing, and is imbedded with the anconal surface exposed in a slab of Stonesfield Slate. It is from a larger Pterodactyle than the preceding. The extensor process is thicker, but springs from a less extended base, relatively to the length of the bone. C.—Pterodactylus Acland, Owen (Plate I, fig. 19). This species is represented by a still larger specimen of the characteristic wing-bone (fig. 19) in Pterosauria. The olecranoid process (c) is shorter in proportion to the breadth 1 “Ueber Pterodactylus suevicus,” 4to, Tibingen, 1855. ''12 FOSSIL REPTILIA OF THE and thickness of the proximal end, and the free termination of the process is more definitely marked by a smooth and shallow groove, over which it seems that the tendon of the “extensor ala” may have glided before its insertion into the strong rough process (¢). The second phalanx of the wing-finger (Plate I, fig. 28) may have belonged to a Pterodactyle of the same species or size as the proximal phalanx of the Plerodactylus Kiddii. On this hypothesis its proportion of length would resemble that in the Ptero- dactylus (Dimorphodon) macronyz (‘Monograph on the Fossil Reptilia of the Liassic For- mations,’ Pterosauria, Paleontological volume for year 1869, Plate XX). ‘The distal end of the present “ Stonesfield ” bone becomes triedral by the rise of a ridge from the thenal aspect, extending longitudinally, and enlarging, to near the outer end of the distal oblong articular surface; this is more convex transversely than is the proximal surface. The longitudinal ridge in question afforded insertion to a strong flexor tendon. § 5. PrerosauRia FROM THE Lias. I have not yet received any evidence of a Pterosaurian from the “Alum Shales” of Whitby, or any other member of the Upper Lias of our North-Eastern Coast, which represents, by the sum of its paleeontological evidence, the “ Posidonomyen-Schiefer ”’ of Bavaria. ‘There, however, in the locality of Banz, have been discovered instructive remains of a Pterosaurian, which Professor Quensted refers to my Lower-Liassic genus under the name of Dimorphodon Banthensis. The specimen about to be described, from the Lower Lias of Lyme Regis, is insufficient to give subgeneric characters, and is provisionally registered under the wider generic name. A.—FPterodactylus Marderi, Owen (Plate I, figs. 7, 8, 9). Of this species is here figured the upper or proximal half of the right humerus (figs. 7 and 8). The head or articular surface (fig. 9) is a narrow, bent, or reniform convexity, with the concave margin toward the thenal side of the bone (fig. 7). The inner and more obtuse end of the articulation, with the tuberosity of that side, is broken away ; the outer, narrower, and, in this species, pointed end is lost upon the ridge or upper border of the “pectoral process” (4). The expanded part of the shaft, beyond the articulation, is concave transversely on the thenal aspect (fig. 7), convex on the opposite or anconal side (fig. 8), which shows, as usual, no trace of the fossa and foramen characterising that part of the humerus in Birds of flight. The antero-posterior thickness ''LIASSIC FORMATIONS. 13 of this part of the bone is less than that of the contracted cylindrical part of the shaft lower down, the section of which is circular. This humerus, besides being smaller than that of Dimorphodon macronyz,' has a more straight and slender shaft, which in transverse section is more nearly cylindrical. B.—Dimorphodon Macronyx, Owen (Plate I, figs. 13, 14). The other Pterosaurian fossil, obtained by Mr. Marder, from the same formation and locality, might well, by its superior size, and more ellipsoid section of the shaft, have formed part of the first long-bone of the wing of the species restored in the ‘ Monograph of Liassic Pterosauria’ of 1870. The articular surfaces of the humerus in both specimens of this Pterosaurian figured in that Monograph (Mon. cit., Plates XVII, XVIII, 53, 53') were too much crushed and mutilated for profitable description. The present specimen shows instructively the distal articulation. The surface for the radius presents one uniform convexity, a, oblong in shape, and obliquely disposed, extending from the lower part of the radial ridge (c), upward, forward, and ulnad ; it is almost wholly developed from the thenal aspect (fig. 13), only the lower border of the convexity being visible from the anconal side (at a, fig. 14). It is longer and more prominent than the ulnar convexity or condyle. This (ib., 6) is subhemispherical ; its diameter equals the shorter diameter of the radial condyle. The intercondylar fissure is a mere cleft ; and tuberous ridges, extending from the condyles, augment the breadth of the distal end of thé humerus. The outer or radial one (c) is produced forward, bounding there, and in part forming the anterior concavity. The inner or ulnar ridge (Z) is more distally placed, projecting to a lower level than the condyle (4) ; it is contmued upwards with a convex curve, but is not produced forward like the radial ridge. Both ridges are connected by a narrower one, extending transversely behind the two condyles, from which it is divided by a fossa (fig. d, c). There is a broad and shallow depression on the back part of the distal end of the humerus fora large “ triceps” tendon : there is no anconal depression. In my description of the articular end of a long-bone of a Pterosaur in the Supplement, No. 1, to the ‘Monograph on Cretaceous Pterosauria” (1857), p. 16, Plate IV, figs. 1, 2, 3, [ remarked that, ‘guided by considerations of size, the fragment might form the opposite end of the bone, indicated by the articular ends (PI. I, figs. 7 and 8), which were referred to the head of the humerus. But I proceeded to remark, ‘I am not acquainted with the precise configuration of the distal end of the humerus in any Ptero- dactyle. From general analogy, however, one should scarcely be prepared to find so 1 “Monograph of Liassic Pterosauria,’ Pal. vol. for year 1869, pl. xviii, figs. 53, 53 a. ''14 _ FOSSIL REPTILIA OF THE LIASSIC FORMATIONS. feeble an indication of divisions into condyles, an absence of a general convexity, and a presence of a well-defined concavity in one condyle, and as well defined a flattened or feebly concave facet in the other condyle, of the distal end of a humerus” (ib., p. 16). The demonstration of the true characters of this end of the humerus, given in Plate I, figs. 18, 14, and d, c, have justified the refusal to regard the articular end of the bone of the large Cretaceous Pterosaur as part of the humerus. Should it prove to be the head of a tibia, what a monstrous flying dragon it would indicate ! There is no part of the skeleton of the Bird that more resembles the answerable bone in a Pterosaur than the humerus. But the following, with other differences pointed out in the previous Monographs, are well marked and, as far as my observation goes, constant. ‘The pectoral process from the radial side of the proximal expansion of the humerus is relatively longer from base to apex, with a broader, more truncate, or less pointed termi- nation in the flying Reptile : it usually forms a low angle in the Bird. At the distal end of the humerus of the Bird the oblong radial condyle is usually more pointed anteriorly ; the ulnar one is more extended transversely, and the inter- condylar cleft is widened to a groove. ‘The outer and inner ridges are not connected by a post-condylar transverse ridge. ‘The olecranial surface is more depressed, and the tricipital tendinal grooves are better marked ; but the transverse expansion of the distal end is less in proportion to the breadth of the shaft of the humerus in the Bird than in the Pterosaur. Other differences in the Pterosaurian humerus, notwithstanding its adaptive develop- ment to flight, showing departure from the avian, and approach to the crocodilian, type are pointed out in detail in my Monograph (Paleontological volume for year 1858, issued in 1861, Supplement, No. ITI, ‘ Cretaceous Pterosauria,’ p. 13, Plate III). ''MONOGRAPH ON THE Genus BOTHRIOSPONDYLUS. § 1. BorHriosPONDYLUS FROM THE KIMMERIDGE CLAY. Species—Bothriospondylus suffossus, Owen (Plates III—V). Tne subjects of the first section of the present Monocrapu might be deemed to have more interest for the Anatomist, by reason of the singular modification of vertebral structure which they exhibit, than for the Palzontologist, as affording evidence of an additional specific or’ generic form to the already known numerous extinct Saurian Reptiles of the Mesozoic formations. The vertebra, for example (Pl. III), which, by the presence of pre-(p) and post-(,’) parapophyses with expanded rough syndesmotic articular surfaces, is a sacral one of the Dinosaurian type, presents so singular a degree of depression, or horizontal flattening, of the centrum, as to suggest artificial and posthumous pressure as its cause ; and it is true ‘that some of the lumbar or dorsal vertebra therewith associated show unmistakable marks of such violence. But, as the side view of the present vertebra, Plate III, fig. 4, shows, at ¢, c’, there is no such evidence of fracture of the peripheral compact layer of the bone with distortion, causing more or less departure from symmetry in the centrum, as accom- panies every instance of crushing out of shape in the present series of vertebrae (compare figs. 1 and 4, e.g., with fig. 5, in Plate V). There is also evidence of a transitional assumption of the depressed form of centrum, in another sacral one (Pl. IV, figs. 4, 5, 6), which, from having the syndesmosal surface on a single parapophysis (?) on each side, was part of a terminal vertebra of the sacral series. Four views (PI. III, figs. 1—4) are given of the vertebral centrum which appears to correspond with that marked 5 in Tab. V of the ‘Monograph on the Fossil Reptilia of ¢ ''16 : FOSSIL REPTILIA OF THE the Wealden and Purbeck,’ Part IV, Paleontographical Society’s Volume for 1856. In the sacrum of the Hyleosaurus there figured the vertebra No. 5 offers the greatest breadth and flattening of the under surface, which is also notable for the absence of the longitudinal ridges, parial or single, marking the under surface of the succeeding or pre- ceding centrums. The under surface of the present sacral (Pl. III, fig. 1) is less accentuated than the Hyleosaurian one compared with it, and the venous canals are relatively smaller than in it: they also issue irregularly, instead of being symmetrically disposed as are the large pair in Hyle@osaurus. The under surface, as shown in the side view (ib., fig. 4, ¢), is feebly undulate lengthwise, the concave curves being mainly due to the expansion of the articular ends (ib., fig. 8). The under surface of the centrum is as moderately convex across, becoming flat near the free portions of the side of the centrum, (ib., figs. 1, 2, 4, ¢) and very slightly concave through the distal expansion of the parapophyses (ib., fig. 1, pp). But the distinctive peculiarity of the present centrum from the known sacral ones of other Dinosaurs is the continuation of the free surface, over the side of the centrum (c’) between the origins of the parapophyses (p, p’) into a long, low and deep cavity (ib., figs. 1 and 4, ff), overarched by the part of the side of the centrum supporting the neurapophyses (ib., figs. 2 and 4, np), which appear to have been confluent therewith, and to have been removed, with the rest of the neural arch, by fracture. This displacement exposes the floor of the neural canal (ib., fig. 2, ~), the breadth of which indicates a sacral enlargement of the myelon, and consequent development of the pair of limbs deriving their nerve-supply therefrom. The issue of a large pair of these nerves is indicated by the continuation of the neural surface outward at 0, 0, behind the broken bases of the neurapophyses (mp) which have not extended so near to the end 4, as to the opposite end, a, of the centrum.} Owing to the abrupt continuation of the lateral surface of the centrum into the depressions, 7; 7, characteristic of the present genus of Dinosaur, the free surface of the side of the centrum presents the form of a smoothly rounded, longitudinally concave, ridge (ib., figs. 2-& 4, ¢’). It may be that the approximation of the roof and floor of the lateral fossee has been increased by pressure. Yet the. horizontal surface, f could hardly have been bent from the vertical side-surface of the centrum, ¢, without some fracture of the compact outer layer of bone; and, further, if the flat form of the centrum had been due to such cause, the seemingly natural undulate configuration of the under surface, with its expansion at the two ends, would not have been unobliterated and unmodified in the degree exhibited by the fossil specimen. The outward production of the fore part of each side of the centrum (fore parapo- physis, p) has a longitudinal extent of an inch and a half, a vertical one at the articular 1 Compare the figure of the sacral vertebra of Iguanodon, Pl. VII, fig. 4, 0, 0, in the ‘Monograph on Reptilia of the Wealden,’ Part II, in the Paleontographical Society’s volume issued for 1854. ''KIMMERIDGE CLAY. i surface of seven to eight lines. ‘The surface is rough and slightly concave; it may have contributed less than one half of the vertical extent of the sacro-iliac joint at this part. The fractured or roughened surface above this parapophysis indicates a corresponding diapophysial production of the neural arch for extension of the joint. Longitudinally the pre-parapophysial surface slightly inclines toward the front articular surface, a, of the centrum. ‘his surface is flat, very rough, and irregular, indicative of having been broken away from a partial confluence with the opposed surface of a contiguous sacral element ; the lower part showing here and there a smoothness as of the original free sur- face of this end of the centrum. Above this surface large unossified vacuities are shown in the cancellous texture of the bone. The vertical diameter of the articular end of the centrum is one inch three lines; the transverse diameter is three inches six lines. The lower margin is not entire, but has been eroded or worn away for an equable extent of about four lines; along the transverse curve it has not been broken off that end of the centrum. The post-parapophyses (?’) are shorter antero-posteriorly, thicker vertically ; and the articular surfaces of this pair converge at a greater angle to the posterior surface, 4, of the centrum (ib., fig. 3) than in the anterior pair. The upper rough or fractured surface (fig. 3, ”, ) may have coalesced with the fore part of the neural arch of the succeeding sacral vertebra, if such arch, as in other Dinosaurs, has crossed the interval between its own centrum and that of the next sacral. A greater extent of the hinder surface of the present centrum (fig. 3, c), at its lower half, shows freedom from anchylosis than on the fore surface. The Reptile indicated by the portion of the vertebra above described is referable by the characters which such fossil shows to the Dinosaurian group. In the Crocodilia the confluent outstanding parts of centrum and neurapophyses, affording attachment to the pelvic arch, are single on each side of the sacral vertebra, and the neural arch retains its normal position in connection with its centrum.! In Megalosaurus the lateral abutments for iliac attachments have diapophysial bases, or spring exclusively from the neural arch.” _—_- Pre- and post-parapophyses are indi- cated in the sacral vertebrae of Jguanodon by the slightly produced or outstanding parts of the side of the centrum articulating with the two displaced neural arches (compare figs. 1 and 2 of Pl. III, with figs. 3 and 4, Tab. VII).2 In the sacral vertebra of the Hylaosaurus, above referred to, the duplex parapophyses have about the same develop- ment as in Bothriospondylus. . Not any of these earlier described Dinosauria have the flattened form and lateral 1 See Tab. IX, fig. 6, sacral vertebra of Crocodilus Hastingsie, ‘Monograph on the Fossil Reptilia of the London Clay, &c.,’ Part II, Paleeontographical Society’s volume for 1849. ? See Tab. I, ‘Monograph of the Fossil Reptilia of the Wealden and Purbeck Formations,’ Part III, Palzeontographical Society’s volume for 1855. 3 Monogr. cit., Part IJ, Paleeontographical Society’s volume for 1854, ''18 FOSSIL REPTILIA OF THE cavities characteristic of the sacral vertebree of the present genus; in which I infer, from the different relative expanse of the neural canal, as shown in the figures of the vertebrae above compared, that the hind limbs were relatively less in Bothriospondylus than in Zyuwanodon. They, probably, came nearer to Crocodilian proportions. A second more mutilated sacral centrum of Bothriospondylus (P\. IV, figs. 4, 5, 6) shows the modification of that marked 4 in the sacrum of Hy/eosaurus, Tab. V, figs. 1 and 2, Monog. cit., in having the parapophysial expansion limited to one (?) on each side of the centrum. _ In the present genus its base occupies the anterior half of the lateral surface, instead of the smaller proportion shown in Hyl@osaurus; it is also more depressed, and the entire centrum is flatter, though not in so great a degree as in the subject of Pl. I above described. Both ends of the present centrum are flat, and show a greater proportion of the smooth unconfluent condition than in the subject of Plate III, fig. 3. The supporting parts of the neural arch forming the roofs of each lateral cavity (figs. 4 and 5, /) are broken off together with the arch itself, and but a small part of the neural surface (ib., figs. 4 and 5, ~) is preserved. This mutilation exposes the whole depth of the lateral excavations (fig. 4, 7,7) of the centrum, undermining, as it were, the base of the neural arch; and these show that the breadth of the centrum beneath that arch is reduced, about midway between the two ends, a and 4, to half an inch, the breadth of the centrum at the fore end, a, being, when entire, 3 inches 3 lines. At the opposite or hinder end the breadth was less, and the height apparently greater, whence it may be inferred that this vertebra was near to the hinder end of the sacrum. The right half of the anterior, flat, smooth but irregularly indented, articular surface of the centrum is nearly entire. Extending, as far as the origin of the pre-parapophysis, p, Which is preserved, and wanting only part of its upper surface, the entire transverse extent can be estimated, as above noted. The under surface of the centrum (ib., fig. 6) is more convex across than in the subject of fig. 1, Pl. III, concomitantly with its greater extent in the present vertebra. The longitudinal contour of the under surface (Pl. IV, fig. 5) is more uniformly concave. The margin of both articular ends is eroded. The aperture of the lateral excavation (ib., fig. 4, ¢’) is 1 inch 5 lines in longitudinal extent; but the cavity is continued 10 lines further above the pre-parapophysis (ib.,y); the depth of the excavation at the middle of the vertebra is 1 inch 3 lines. The smooth compact crust of the centrum passes, without fracture, over the free lateral tract (ib., fig. 5, ¢’). The vertically convex border of the floor of the cavity is somewhat thicker than in first-described sacral vertebra, but similarly shows a natural condition and contour. The upper surface of the floor of the cavity shows a fine crack (outside the letter ¢ in fig. 4) as if the inner half of that floor, with the adjoining part of the centrum (?) supporting the base of the neural arch had been slightly depressed. ''KIMMERIDGE CLAY. , 19 The proportion broken away from the left side of the present vertebra is indicated in outline in figs. 4 and 6. The subject of figs. 1, 2, 3. Pl. IV, transmitted at the same time with the vertebra above described, and from the same locality, I refer, from the superficial characters of the under and one of the terminal surfaces of the centrum, to the same genus and species of Dinosaur, and it probably formed part of the same individual. The flattened surface of the centrum, at a, fig. 2, in the irregular impressions of its otherwise smooth surface closely accords with the one, 4, of the subject of fig. 5, to which it adapts itself, sufficiently closely to suggest that it may have been liga- mentously articulated thereto. ‘The opposite surface (ib., fig. 1 and fig. 2, 4) is not so impressed, is slightly convex and smoother, and indicates a joint with the succeeding vertebra admitting of more movement. I infer, therefore, that the present specimen is the centrum of the last sacral vertebra, and that the end articulating with the first caudal vertebra had resumed more of the usual vertical proportions of the centrum. The para- pophysis (p), with the irregular syndesmosal surface, has a greater extent, both vertically and lengthwise. Above it extends the narrow fractured surface of the broken off base of the neurapophysis. The floor of the neural canal (fig. 1, ») is preserved, which 1s concave lengthwise as well as across, sinking somewhat into the substance of the centrum. Its diameter midway between the two ends is 7 lines. The lateral excavations of the centrum appear to have ceased at this vertebra, and probably were not resumed in the caudal series. It has been fractured and somewhat distorted by posthumous violence: but this has not affected the contour of the under surface of the centrum (ib., fig. 3), or the vertical proportions of this element, any more than in the case of the two previously described sacrals. In four centrums of dorsal or dorso-lumbar vertebrae of Bothriospondylus suffossus, forming part of the same series transmitted from the Kimmeridge Clay of Swindon, the characteristic excavations are conspicuous and with longer apertures than in the sacral vertebrae, where these are interrupted by the broad articular parapophyses. No trace of the latter processes are present in the trunk vertebree of which the type is selected for the subjects of Plate V. The centrum is subcompressed (fig. 2); its sides moderately concave lengthwise (fig. 1), with one end feebly convex, a, the opposite end rather more concave, 4. I regard the latter as the hinder one, and the trunk-vertebree to be, as in Streptospondylus, of the opisthoccelian type. The free surface of the centrum is smooth, save near the articular ends, where there are low longitudinal risings and shallow channels, as shown in fig. 1, Pl. VI. The under surface (ib., fig. 4) is perforated by two or more small vascular (venous) canals near the articular ends. The fore end (ib., fig. 2) has a somewhat irregular surface. The hind one, which has suffered less from compression (ib., fig. 3), shows a similar coarse pitting and rising at the ''20 FOSSIL REPTILIA OF THE central part of its surface, the peripheral part being smoother than that at the middle, which has yielded to pressure, the large cancelli there having been crushed in. The bases of the neurapophyses (PI. V,. zp), commencing about three lines from the anterior end of the centrum, are continued to the posterior end. They have been anchylosed to the centrum and broken away. Posthumous pressure has crushed this specimen laterally and obliquely. Part of the floor of the neural canal is exposed (at _n, n, fig. 5), and is continued outward, at 0, where the spinal nerve has had issue. The narrowness of the tract of the centrum, between the lateral excavations, /, /, giving support to the coextensive parts supporting the neural arch, is a singular characteristic of the present genus, and made it difficult to conceive that a mere plate of bone like that between ¢ and np in fig. 1, Pl. V, would relate to the support of a neural arch. It recalled the structure of that part of the vertebra in the thoracic-abdominal region of a Chelonian, What the character of such arch may have been we have yet to learn, in the present species, from better preserved specimens. Not a fragment recognisable as belonging to such portion of the vertebra could be found among the fossils sent up from the Kimmeridge locality at Swindon. ‘wo rather more crushed and distorted centrums show, nevertheless, an increase of transverse diameter indicative of their having come from a region of the spine near the sacrum. ‘The centrum shows the same opisthoccelian type, the same wide and deep lateral excavations, undermining, as it were, the neural arch, an absence of transverse processes, and the fractured bases of anchylosed neurapophyses. The “Swindon Brick and Tile Company’s Works,” whence, through the kindness of the managing director, James K. Shopland, Esq., the above-described fossils were obtained, are situated on land adjoining the Wilts and Berks Canal. The vertebrae were found, associated with remains of Pliosaurus brachydeirus, in the Kimmeridge clay, at a depth of fifteen feet. The clay here is of a deep black-blue; and a mass of lignite, seemingly derived from a crushed trunk of a tree, and burning like ordinary coal, was here discovered. I was indebted to Richard Jefferies, Esq., of Coate, Swindon, for the first intimation of this discovery, and for a sketch of part of the mandible, 4 feet in length, and of a femur | foot 1 inch in length, of the Pliosaurus ; of which Sauropterygian genus an entire tooth, and parts of others, with fractured vertebral centrums, were exhumed from the same pit as contained the vertebrae of Bothriospondylus. To Mr. Shopland I am indebted for the transmission of the series of specimens here obtained, which included the evidences of the Dinosaurian genus above characterised. This genus, or type of vertebrae, I next proceed to illustrate by larger fossils from other localities and Mesozoic formations. ''FOREST-MARBLE. : ak § 2. Bora#RiosPoNDYLUS FROM THE FOoREST-MARBLE. Species—Bothriospondylus robustus, Ow. Plate VI. Of the two extinct Crocodilians from the Oolitic deposits (probably corresponding with the Oxford Clay) in the vicinity of Honfleur and Havre, of which the jaws and teeth are figured in Pl. VIII of the 2nd Part of the fifth volume of Cuvier’s ‘ Ossemens Fossiles,’ 4to, 1824, one in which the fore-part of the centrum was convex and the hind- part concave, and to which H. v. Meyer afterwards attached the generic name Streptospondylus, was also characterised by lateral depressions (tom. cit., Pl. VIII, fig. 13).! I have figured a vertebra of this type, from the Lias of Whitby, under the name of Streptospondylus Cuviert. Assuming the accuracy of the ascription of this type of vertebra to the associated jaws and teeth at Honfleur, the genus Streptospondylus is a crocodilian one, and the sacrum would accord with the characters of the order. I have long had or known evidences of the Reptilian genus characterised by the deep and elongated lateral depressions of the centrum undermining the piers or plates supporting the neural arch. But not until the acquisition from the Kimmeridge Clay of the above-described elements of sacral vertebra, similarly excavated, were grounds afforded for the determination of the order or group of Saurians to which the Bothrio- spondylus belonged. The earliest indications of the genus were afforded by fragmentary vertebrae or vertebral centrums from the Forest-marble of Wiltshire, of which the least mutilated is figured in Pl. VI. It is from the same or contiguous region of the spine as the subjects of Pl. V. The degree of convexity of the anterior surface (Pl. VI, fig. 1, a) is the same as that in the subject of fig. 1, Pl. V, a The posterior concavity also shows the irregular tract at the centre, 4, as in that of Bothriospondylus suffossus (Plate V, fig. 3). The lateral fossa (Pl. VI, fig. 1, f), with the same relative longitudinal extent, is some- what less deep. The whole centrum is shorter in proportion to its height and breadth than in the subject of Pl. V, fig. 1; it has come not merely from a larger individual of the species, but from another species of the genus, with seemingly a relatively larger, less elongate trunk. This vertebra repeats the textural character of the genus in the unossified proportions of the centrum, causing the large cancelli (Pl. VI, fig. 2) into which spar has infiltrated and crystallised in the fossil. I regard, therefore, the vertebrae from the Wiltshire Oolitic or Mesozoic deposits called ‘ Forest-marble’ as indicative of a species 1 “PDerrigre la facette, qui regoit la téte de la cdte, est une fosse profonde.’’ Cuvier, tom. cit., p. 155. ''92 FOSSIL REPTILIA OF THE distinct from the smaller kind from the Kimmeridge Clay, and propose to call it Bothriospondylus robustus. It is matter of regret that, as yet, no teeth have been recovered from the Swindon locality of the Kimmeridge Clay which are not referable to saurian genera previously known and distinct from Bothriospondylus. But inthe course of my work on ‘ Odonto- graphy’ I received from an esteemed correspondent and ardent collector of fossils, Mr. Channing Pearce, of Bradford, Wilts, a tooth from the Forest-marble near that town, which I figured in the same plate’ with that of Hy/gosaurus,’ discerning in it an extreme modification of the same type of Dinosaurian dentition. I reproduce the figures from the ‘Odontography ’ in Pl. IX of the present Monograph. The generic name of the Forest-marble Saurian so indicated was suggested by the heart-shaped form of the crown of the fossil tooth (Pl. IX, fig. 2). The crown, being 1 inch in length, 8 lines in breadth, and 5 lines in thickness, might well have come from the jaw of a Dinosaur with dorso-lumbar vertebra of the size of that here referred to Bothriospondylus robustus. ‘The crown suddenly expands above the neck of the tooth, thins to an edge along the fore and hind border (ib., fig. 3), and contracts to a point or apex, which is sub-obtuse, being somewhat worn in the specimen figured. The enamel has a peculiar character (of which a magnified view is given in Pl. IX, fig. 5), being raised into thin wavy longitudinal ridges with widish intervals where it was sculptured by minute ruge.’ The fang or root is cylindrical, coated with smooth cement ; the base of this was preserved in the subject of fig.4, Pl. IX. I have not received any specimen of this kind of tooth, nor any vertebra of the type of that figured in Pl. VI, save from the Forest- marble near Bradford ; but Professor Phillips has given a woodcut of a mutilated crown of a similar sized tooth,* the best preserved margin of which swells out as in Cardiodon, which was discovered associated in the ‘Great Oolite’ with bones which will be subse- quently shown to have the characters of Cetiosaurus. I feel the insufficiency of the present grounds for referring these teeth to any otherwise defined species or genus of Dinosaur. But, if such heart-shaped teeth, with other characters of Cardiodon, and especially if 2m siti in the jaw, or in portions of jaw, should be discovered associated with vertebrae of the Bothriospondyloid type, it may then be a question for the taxonomist whether Bothriospondylus should subside as a synonym of Cardiodon ; unless, indeed, modifications of other parts of the skeleton than are now known of Bothriospondylus robustus should be deemed to support a distinct generic name (Marmarospondylus, e.g.). Meanwhile it seems to me more convenient to retain the verte- bral] designation of the genus, as I have next to show that such osseous generic characters were manifested by still larger species from other members of the Mesozoic series. 1 «Qdontography,’ Pl. 75a, figs. 7, a, 6, ¢, d. ? Ib., figs. 6, a, b. 8 ‘Odontography,’ p. 291. * «Geology of Oxford,’ 8vo, 1871, Diagram LXXXV, p. 253, “Tooth of Ceteosaurus.” ''WEALDEN. 23 § 3. BorHRIosPONDYLUS FROM THE WEALDEN. A.—Bothriospondylus elongatus, Ow. Plate VII. Although my opinion of the nature of this fossil, figured in the Monograph on Wealden Dinosaurs, Palesontograph. Soc. Volume for 1854, Tab. X, as “the tympanic bone of an Iguanodon >” was subsequently modified, as is well known to those who sympathised with, and assisted by materials in, the progress of my work on ‘ British Fossil Reptiles,’ the first published opinion (1870) of the vertebral nature of the fossil, "as far as I am aware, was that of H. G. Srenny, Esq., F.G.S., to which I shall presently refer, retaining my conviction of the closer agreement of the vertebra in question with the subjects of Plates III—VI of the present Monograph than with those of any volant animal. The species, Bothriospondylus elongatus, is represented by the centrum of a dorso- lumbar or trunk-vertebra, from Tilgate, exceeding in the proportion of length to depth that of Bothriospondylus robustus (Pl. V1), in a greater degree than this is exceeded by the corresponding vertebra of Bothriospondylus suffossus (Pl. V). The ratio of augmentation of bulk also becomes greater as we pass or ascend from the small Kim- meridgian type to the colossal Wealden form. The length of the present fossil centrum is eight inches, and it must have been more when perfect, for both ends, and especially the fore or subconvex one, have undergone fracture and abrasion. ‘The fractures at the end, a, however, bring to light the unusually large cancelli, some of them admitting the end of the thumb; and this structure, associated with the length and depth of the lateral depression, Pl. VII, 4 givethe grounds for referring the specimen to the genus, or group, Bothriospondylus. The natural surface of the bone is smooth, or finely striate lengthwise, towards the articular ends, as in Bothr. suffossus, Pl. V, fig. 1. As in that vertebra, also, the centrum expands to both ends, but less gradually, the contracted mid-part being relatively longer ; its transverse section is less cylindrical, the lower surface being more flattened, less convex transversely, and the breadth of the middle of the centrum, 3 inches 6 lines, being greater in proportion to the height measured from the margin of the lateral depression to the under surface, which is 2 inches 6 lines. The part of the centrum above the depression, f, becomes, as in the smaller vertebra, very thin; and, as with the portion preserved in Bothriospondylus robustus, the plate inclines outward as it ascends, indicative of a neural arch of greater breadth than the centrum below. d ''24 FOSSIL REPTILIA OF THE It may be that the more fragile, or less robust, character of the expanded arch is connected with the loss of that part of the vertebra in most of the examples of the genus of which I have hitherto had cognizance. In Chelonia the sides of the neural arches of the abdominal vertebrz are represented by thin vertical plates of bone. The excavate modification of the centrum of Bothriospondylus is more commonly met with in that element of the vertebra of Fishes, as is also the character of the large propor- tion of modified chondrine in its substance. In some of the large Scomberoids there is a lower as well as an upper excavation on each side of the centrum. The bases of the parapophyses of the abdominal vertebra of the Haddock (Gadus aglefinus) are expanded and excavated. Pallas observed that saccular productions of the air-bladder were continued into, and, as it were, lined, homologous cavities at the sides of the trunk vertebree of an allied Gadoid (Gadus Navaga). As the lateral fossee disappear in the caudal vertebrae of Bothriospondylus one is tempted to surmise that saccular processes of the lungs, which probably, as in Sauria and Chelonia, lined the mid-part of the roof of the thoracic-abdominal cavity, may have been prolonged into the lateral excavations of the vertebra. But, if so, the lungs must have extended far forward as well as backward, since posterior cervical or anterior dorsal vertebrae show the lateral fossae, as do the sacral vertebree. In certain Pferosauria (Coloborhynchus Sedgwickii, e.g.) extensions of the lungs or air-cells were continued along the sides of the neck, and did penetrate lateral depressions of the centrum answering to those in Bothriospondylus.’ B.—Bothriospondylus magnus, Ow. Plates VIII and IX. That the anterior dorsal vertebra, of which a side view is given of the natural size in Pl. VIII, does not belong to the same species as the previously described centrum from a hinder part of the trunk, may be inferred from the superior proportions of the articular ends to the length of the centrum. On the supposition that the present vertebra formed part of the back-bone of a larger and older individual of the same species as Bothrio- spondylus elongatus it would show a degree of shortness of the centrum in proportion to its breadth and depth, which is unique in my experience of the characters of centrums in the same region of the vertebral column. A vertebra, as in Pl. VII, with a terminal facet eight inches in vertical diameter, 1 Owen, “On the Vertebral Characters of the Order Péerosauria.” ‘Philosophical Transactions,’ MDCCCLIX, p. 114, pl. x, figs. 5, 7 ''WEALDEN. 25 would be nearly sixteen inches in length if it came from the spinal column of a Bothrio- spondylus with an anterior dorsal or cervical vertebra of the vertical and transverse dimensions of the subject of Pl. VIII. I infer, therefore, that this fossil indicates a species with proportions of the vertebre, and probably of the trunk, more like those deducible from the type of Bothriospondylus robustus. But the present remarkable fossil shows a still larger proportion of unossified parts in the substance of the centrum. The side-pit is short, vertically wide, but deep ; the long diameter of the aperture is somewhat more than one third that of the entire centrum: its compact lining layer of bone is entire, not perforated as in the ‘foramina pneumatica’ of birds. The fore part of this element is strongly convex; the hind part answerably concave. The bases of the neural arch extend to within an inch and a half of the hind margin of the centrum ; they rise at the beginning of the convexity of the fore end. This convexity has suffered abra- sion, and the widely cancellous structure is exposed, as shown in Pl. IX. Craving excuse for premising so trite or elementary remarks, the primal basis of the vertebrate skeleton may be converted into sclerine or chondrine, and ossification may begin in either ‘membrane’ or ‘cartilage.’ In some Vertebrates, chiefly 1f not ex- clusively cold-blooded, more or less of the bone may remaing unossified, retaining the antecedent stage, with some slight modification of tissue, to which, as in Selachian vertebree, the term ‘chondrine’ has been specially given. Such partially ossified bones show corresponding cavities, usually filled with matrix, or spar, in the petrified state. But this condition of fossil bones may depend on other osteogenetic changes. After substitution of bone-earth for gristle, or the conversion of the entire cartilaginous mould into bone, the central part may be absorbed, and an oily substance called ‘marrow ’ be deposited in the cavity. Or, the absorption of previous solid bone may go further, the marrow may also be removed, the wall of the bone may be perforated or ‘tapped,’ and air be admitted from a contiguous extension of the lung. But in cases of petrifaction the non-ossified parts of a bone become filled by the same mineral infiltration, whether the cavities originally contained chondrine, marrow, or air. The inconsiderate conclusion that fossil bones with large cavities and thin compact walls must have been those of Pterodactyles or of Birds led to the supposition that certain fossil eggs belonged to one or other of these volant classes, the bones of the unexcluded embryo showing the above hollow or tubular character. Such eggs in a portion of stone from a quarry in the Island of Ascension were submitted to my examina- tion in 1834, and I detected among them the characteristic scapula and coracoid of a Chelonian embryo. To the objection against that determination, based on the hollowness of the larger limb-bones associated therewith, I showed, by dissection of a newly hatched Chelone preserved in spirits in the Hunterian Museum, that the cavity of such bones had. been filled with chondrine, not with air; that the thin outer shell of bone was a transitory ''26 FOSSIL REPTILIA OF THE WEALDEN. embryonal character, and that the femora, humeri, and other bones became massive and solid in the adult turtle." The earlier stage seems to have been permanent in Poihilo- pleuron and Bothriospondylus. H. G. Srertzy, Hsq., F.G.S., the then able “ Assistant to Professor Sepewick in the Woodwardian Museum, conceiving” that the cavities in the osseous tissue of the subjects of Plates VII—IX had been filled with air, affirms them to have been “ constructed after the lightest and airiest plan, such as is only seen in Pterodactyles and Birds ;’” also that “the neck would appear to have been carried erect after the manner of birds ;”* finally, concluding that “our animal is therefore clearly ornithic,’’ he concentrates these ideas and stamps them for currency under the generic name OrnirHopsis, for such supposed stupendous volant vertebrate. With respect to which the judgment of competent paleontologists may be exercised in considering the applicability thereto of the ‘eleventh’ of the * Rules for Zoological and Botanical Nomenclature, authorised by Section p of the British Association at Manchester,” ‘ Reports of the British Association for the Advance- ment of Science,’ Svo, for the year 1842.° ' See ‘Note,’ p. 292, of Lyell’s ‘ Principles of Geology,’ vol. ii, ed. 1835. 2 «On ORNITHOPSIS, a gigantic animal of the Pterodactyle kind, from the Wealden,” in ‘ Annals and Magazine of Natural History,’ vol. v (4th series), 1870, p. 279. © tb. p. 260. = tb., 1b. © Tb: 1. 6 «A name whose meaning is glaringly false may be changed ’’—“‘ when it implies a false proposition which is likely to propagate important errors.” —Op. cit. ''MONOGRAPH ON THE Genus CHTIOSAURUS.' § 1. Crrrtosaurus FRoM THE GREAT OOLITE. Species—Cetiosaurus longus, Ow. (Plate X, and Woodcuts 1—10). Untit a comparatively recent period the generic or family characters of the great extinct Cetiosauroid Reptiles were founded on a few scattered bones of the trunk and limbs. The texture of these fossils mainly differentiated them from the corresponding vertebre and limb-bones of previously determined genera or species of Saurians. No portion of the skull, not one tooth, had been discovered so associated with Cetiosaurian bones, at the date of my “ Reports on British Fossil Reptiles,’’’ as to throw any additional light on the ordinal affinities of the new genus. I had not, then, grounds for dissociating it from the Crocodilian group or order. ‘The grand accession of evidences of the osseous framework of one of the species* added to the original collection by Buckland, preserved in his Museum at Oxford, by his eminent successor, the late lamented Professor Phillips, by whom they have been instructively elucidated in his excellent work on the ‘ Geology of Oxford,” has proportionally advanced the means of determining the ordinal relations and affinities of the genus. The inferences which may be drawn in favour of the Dinosaurian " Gr, kjrevos, cetaceous ; catpos, Lizard ; “ Report on British Fossil Reptiles,” Part ii, in ‘ Reports of the British Association,’ &c., for the year 1841; also ‘ Proceedings of the Geological Society of London’ for June, 1841 (vol, iii, p. 457). * ‘Reports of the British Association for the Advancement of Science’ for the years 1839 and 1841. ® “ Cetiosaurus longus,” Ib., ‘Report’ of 1841, p. 101. * 8vo, 1871. ''28 FOSSIL REPTILIA OF THE characters of the sacrum will be subsequently discussed. But the demonstration of the sacral characters of a more recently discovered Cetiosauroid genus, the subject of another Monograph in the present volume, adds to the grounds for referring the type- species to the great Dinosaurian group of Reptilia. It is characteristic of the accidents that attend the quest and acquisition of the remains of extinct Vertebrates, that skull, jaws, and teeth should have escaped the careful operations to which we are indebted for the present means of restoring both Cetiosaurus longus and Omosaurus armatus. Of the former reptile a single doubtful and mutilated tooth was all that Prof. Phillips could refer with any degree of probability to that species. That the side-pits of saurian vertebree have no essential relation to largely cancellated, pseudo-pneumatic structure of the bones is shown by their presence in the anterior trunk-vertebre of the genus for which the uniformly close though coarse osseous texture, as in the whale tribe, suggested the generic name Cetiosaurus. The first indication of this type of Saurian was, however, afforded by an inspection of a limb-bone, submitted to me by Dr. Buckianp in 1838, when I was engaged in collecting materials for my ‘ Report.’ to the British Association ‘ On the Fossil Reptilia of Great Britain.” Buckland had referred to this fossil in his ‘ Bridgewater Treatise,’ Ist edit., 1836, in the following terms :—‘ There is in the Oxford Museum an ulna from the Great Oolite of Enstone ”’ (Enslow probably meant), “near Woodstock, Oxon., which was examined by Cuvier and pronounced to be cetaceous; and also a portion of a very large rib, apparently of a whale, from the same locality.” This limb-bone I could not match with any then known to me in the Cetaceous order. Yet, save a thin compact outer crust, the osseous structure was, where exposed, like that in the humerus of a Whale or Grampus; there was no medullary cavity. In shape the resemblance, though remote, seemed nearest to that of the outer metatarsal of a Monitor Lizard.’ Shortly after I was able to differentiate certain saurian vertebra from those ascribed to the genera Jyuanodon, Hyleosaurus, Megalosaurus, and Poikilopleuron, not only by superiority of size, but by differences in form, proportions, and structure.? The latter character applied, more especially, to these huge unknown fossil bones in the comparison with Poikilopleuron, in the vertebree of which four-footed reptile ossification is incom- plete and large chondrosal vacuities are left in the substance of the centrum, which, in the fossils, become filled with spar.° 1 See ‘Monograph on a Fossil Dinosaur (Scelidosaurus) of the Lower Lias,’ tab. xi, fig. 3 0; Paleeontographical Society’s volume for 1860. Prof. Phillips, who had obtained, in 1870, from the Great Oolite at Enslow, the three metatarsals of each hind foot of a Cetiosaurus, wherewith he was able to com- pare the above fossil long bone, “incomplete at both extremities,’’ considers the determination of it as a metatarsal of large size to be ‘ probably true.’—‘ Geology of Oxford,’ &c., 8vo, 1871, p. 285. 2 «Proceedings of the Geological Society of London,’ June, 1841, loc. cit. 3 The chief of these cavities, being in the centre of the vertebrae, was termed ‘medullary’ (loc. cit., ''GREAT OOLITE. = SS ca; cons ---99 . tite Daas sme From the similarity of texture of the vertebrae of the new genus of Saurian so indicated to that in the limb-bone from “ Blechingdon,” Enslow, I suggested that it might belong to Cetiosawrus.! The cetaceous hypothesis of the huge Oolitic Vertebrate was thereupon abandoned, and my determination was adopted in the second edition of the ‘ Bridgewater Treatise,’ and also by Lyell, who gives a reduced cut of the fossil in his ‘Manual of Geology,’ ch. xx. In 1848 Dr. Buckland informed me of the discovery of a femur, 4 feet 3 inches in length, which, from the correspondence of its texture with that of the metatarsal from Blechingdon, and also with that of some fragmentary long bones from Blisworth, Northamptonshire, I referred to the genus Cetiosaurus, and to the species from the Great Oolite called Cetiosawrus longus.” More recently (1868—70) a considerable proportion of the skeleton has been discovered in the quarries of the Great Oolite of Enslow Rocks at Kirtlington Station, eight miles north of Oxford, the bones of which more nearly approach in size to the type specimen of Cetiosaurus longus.’ Such of the trunk-vertebree as were sufficiently entire appear to have come from the fore part of that region, and show the opisthoccelian character of those vertebrae as in certain Dinosaurs. I, therefore, visited Oxford for the purpose of studying these remains. In the best preserved anterior dorsal vertebra the parapophysis, short but large in vertical. extent, shows remains of the articular surface for the head of the rib. The diapophysis, supported by a strong buttress-like ridge, is directed upward and outward at an angle of 45° with the neural spine. The distance between the articular surface for the ‘tubercle’ and that for the ‘head’ of the rib is ten inches, which gives the extent of the ‘neck’ of the rib at this fore part of the thorax. The neural spine is strengthened by lateral buttress-like ridges rising from the neural platform; it is of a massive quadrate form and seems to have terminated obtusely. The zygapophyses are supported by buttress-like vertical ridges.* All the characters of this massive vertebra bespeak the great strength of the back-bone of the enormous saurian. ‘The total vertical extent of the above vertebra, which is incomplete at the wider part of the centrum, is 2 feet A inches; the breadth at the diapophyses is 1 foot 6 inches. The vertebra which is the subject of Plate X, from a hinder position of the trunk than p- 459); but I have since had reason to conclude that it was occupied in the living Saurian by unossified chondrine. 1 * Report,’ ut supra, p. 101. 21h, ib, 3 “ Vertebree 8, 9, and 11 inches in diameter,” ‘“‘ monstrous ribs,” “femora upwards of 5 feet in length.” —‘ Atheneum,’ April 2nd, 1870. 4 “On Cetiosaurus from Oolitic Formations,” ‘Proc. Geol. Soc.,’ 1841, l.c., p. 459. Cetiosaurus longus is defined as in the ‘ Report,’ and distinguished from the Cetiosaurus brevis of the Wealden Forma- tions, pp. 101, 102, which will probably prove to be referable to a distinct cetiosauroid genus. ''BC FOSSIL REPTILIA OF THE the above-described, exemplifies the cetiosaurian characters of texture (fig. 2, 7) also of a contracted antero-posterior extent of the neural arch as it rises from the centrum,! and of a partial subsidence of the anterior ball.’ This vertebra has been crushed and fractured ; the right side is pressed obliquely backward for an inch or so beyond the left side, so that the length of the centrum, measured as it has been squeezed out of shape, exaggerates its original or natural longitudinal diameter. This would not exceed, according to my estimate, eight inches. ‘The vertical diameter of the centrum has also been pressed down beyond its original extent. I put the bail or fore part at 63 inches, the cup behind at 7 inches, in height. The neural arch, as in the type-vertebrae of Cetiosawrus longus,® is retained in anchylosed union with the centrum to the extent shown in Plate X, viz., eight inches. A vertically grooved median ridge appears to commence at the back part of the base of the spine. ‘This process is wanting; it probably would have added a foot to the present vertical extent of the vertebra, which is sixteen inches. Minor projecting parts have been equally broken away, and, as usual, lost in the quarrying or extricating operations. Such fractures occur on both sides of the prominent rim of the hinder cup of the centrum (as at p, fig. 2, Pl. X). The singularly naturally compressed upper and middle part of the centrum (ib. /) underlying the neural canal and forming a vertical plate or medial wall of bone, three to four inches in height, and but six lines to eight lines in thickness, has been in part broken away, exposing that canal. The fore and hind outlets of the neural canal are squeezed into a narrow, vertically lengthened, oval shape (ib., fig. 2, n). The neurapophysis rises by two buttress-like columns (nm, ») which converge as they ascend and overarch the lateral depression ¢. ‘The base of the neural arch is co- extensive with the centrum, save in so far as the anterior ball may have projected beyond; but the neurapophysis soon shows, as it rises, the ‘short antero-posterior extent,’ which is among the characteristics of the genus. One advantage of the fractures, which must otherwise have been got by sections, is the demonstration of the 1 In the account, illustrated by woodcuts, given by Phillips in his excellent ‘Geology of Oxford,’ pp. 246—294, a vertebra, supposed to be lumbar, the subject of the diagram Ixxxviii, p. 257, has assigned to it the following admeasurement :—“ Greatest length from front to back (crushed) 4°6 in.” I have found no trunk-vertebree of the Cetiosaurus from the Kirtlington Oolite so short as this. 2 In a former ‘Monograph’ I remark :—‘If, as is very probable, the cervical and anterior dorsal vertebree above described (pp. 22—26), and provisionally referred to Streptospondylus, belong to the same reptile as the succeeding vertebra, here referred to Cetiosaurus, we should then have a gigantic Crocodilian of the peculiar transitional type, as between that order and the Dinosaurian, which is manifested by the “Second Honfleur Gavial” of Cuvier, 7.e., with convexo-concave vertebree at the fore part of the trunk, graduating into plano-subconcave vertebra with elevated and somewhat complex neural arches, at the middle and back part of the trunk, and with vertebrae sub-concave at both ends in the tail.”’— ‘Monogr. on Wealden Reptilia,’ Suppl. II, p. 34, Paleeont. Soc. Vol. for 1857, issued 1859. 8 «Tn all these vertebra the neurapophyses are anchylosed to the centrum,” &c.—‘ Report,’ p. 102. 4 Monogr. cit. (1859), p. 27. ''GREAT OOLITE. 31 cetiosaurian texture of the bone (Pl. X, fig. 2, p), as contrasted with the cancellated structure in Bothriospondylus (Pl. IX). The resemblance of this close but somewhat coarse osseous tissue to that of cetaceous bone, especially in the larger Whales, and which seems to characterise the whole skeleton of the present genus of gigantic saurians, might well excuse the idea that the huge long bone first observed was cetaceous. The unbroken surface of the vertebra has a fine fibroid character; the interrupted lines affecting a longitudinal course on the centrum and a vertical one on the neurapophysis. How far any exposure of the arch at the base of the spine may have formed a part answering to the ‘platform’ in the antecedent vertebra, and as in most Dinosaurs, the broken state of the specimens does not allow of determination. Near the borders of the articular ends of the centrum, which are more or less rubbed away, stronger sculpturing is indicated, as if in relation to ligamentous attachments. The lower border of the lateral depression, f, is more obtuse, less definite, than in Bothriospondylus (P|. VIL) ; the vertical convexity of the side of the centrum changes in Cetiosawrus more gradually into the concavity of the depression. The sternum of Cetiosaurus longus is a transversely elliptical plate with an almost flat, slightly undulate upper or inner surface (fig. 1); 19 inches broad, 15 inches long, 1 inch to 14 inch thick, increasing to 24 é inches at the coracoid articular surfaces, | fe though, probably, the entire expanse of the border here is not preserved. The hind s border shows prominences for the attachment ji @@ “~~ \ of three pairs of sternal ribs, the hindmost A pair in contact, as in Monitor niloticus. é i In this Lizard the sternum has a rhom- if boidal form, with a low median ridge on the “4 i Hier 1. wt Sl \ sat HN Ye S 4 my) KN ENN outer or under surface, a deep hollow on the opposite surface, and considerable thickening of the articulations for the cora- te ogi e} Aver is x coids. Were these bones fully ossified in that ae ee Lizard they would correspond in breadth with eos those of Cetiosaurus ; there are, however, Sternum, Cetiosaurus longus, {5th nat. size. (Phps., two tracts retaining the primitive sclerous oeiaea state, and an antero-medial part which has not gone beyond that of gristle, in the coracoid of the recent saurian. We have, therefore, in Cetiosaurus, as in some other ancient saurians, especially Dinosauria, a degree of lacertian structure combined with a crocodilian advance of vertebral and concomitant cardiac and pulmonic structures. The scapula of Cetiosawrus (fig. 2) is more crocodilian than lacertian in its ''32 FOSSIL REPTILIA OF THE hie. 2: Right. Left. Scapula, Cetiosaurus longus. ysth nat. size. (Phps., diagr. xcix, p. 270.) proportions. It is an elongate plate, expanded at both ends, but most so and most abruptly at the articulations for the coracoid, ¢, ¢, and humerus, 4,4. The more gradual expansion of the base or free extremity is chiefly due to the hinder border, and this describes a concavity, while the fore border is nearly straight. The outer surface (left) is slightly depressed lengthwise behind a longitudinal ridge near to and parallel with ''GREAT OOLITE. 33 Fig, 3 > Yes 4. H\WINS mh | i 4 |! i ! vt lt i i I i i’ Right. Left. Humerus, Cetiosaurus longus, jyth nat. size. (Phps., diagr. c., p. 272.) the anterior border. The inner surface (right) has a longitudinal rise near the middle, which bifurcates to strengthen the humeral and coracoid surfaces, and to add to the thickening of the articular end of the bone. Pamiips notes the modification of structure of the basal three inches of the blade, indicative of coarse or partial ossification ''34 FOSSIL REPTILIA OF THE of an original cartilaginous superscapula, the proportions of which element would thus be more crocodilian than lacertian. The resemblance of the blade-bone of Cetiosaurus to that of Scelidosaurus has already been noted. But the production of the anterior or humeral angle of the articular end is somewhat greater, approaching that in Hy/cosaurus. The length of the scapula of Cetiosaurus longus is 4 feet 6 inches, the breadth of the articular end is 2 feet 2 inches, the least breadth of the body of the bone 10 inches. The humerus of Cetiosaurus is far from exhibiting the outstanding plates and ridges for muscular attachments, such as we see in the larger existing lizards (Hydro- saurus, Monitor), which run swiftly on land; they are even more feebly indicated than in the Crocodiles, but how much of this inferiority may be due to posthumous injury and abrasion in the present huge fossils is questionable. The head of the humerus, fig. 3, a, A, is an elongate, semi-oval, narrow convexity, broadest at the middle, which projects toward the hinder or anconal surface of the bone, as in Lizards and Crocodiles ;! the degree of the projection is shown in the outline of the proximal end of the bone, at c¢, fig. 3, a. The ridge from the radial side of the proximal third of the shaft (fig. 4, 2), answering to the ‘pectoral’ or ‘deltoidal’ one in the Mammals, commences, as in the Monitors, near the head, not, as in the Crocodiles, abruptly at some distance below ; it has suffered abrasion in the Kirt- . lington specimen, yet seems not to have stood out in the same relative degree as in the Monitor, in which, as in the Crocodile, it is bent toward the fore or palmar side of the bone. : The shaft of the humerus in Cetiosaurus is subcompressed, subtrihedral, through an obtusely angular longitudinal low ridge or prominence, on the anconal side (fig. 3), continued from below the head to near the distal end, inclining toward the radial side. ‘There is no trace of the distal ridge from that border of the shaft which, in Monitors, answers to the ‘supinator’ ridge in Mammals (Pl. XVII, fig. 6, ¢). The more prominent of the two distal articular convexities, that, viz., for the head of the radius, is feebly indicated ; the back part of the convexity for the ulna is traceable at the worn distal end of the bone (fig. 4, a’). The pectoral and supinator ridges are still more feebly in- Ulna, Cetiosaurus longus, 3;th nat. size. (Phps., diagr. ciii, p. 275.) } ‘ Supplement (No. IIL) to Monograph of the Fossil Reptilia of the Cretaceous Formations,’ Paleon- tographical Society’s vol. for 1858 (issued 1861), p. 15, tab. iii, fig. 10. ''GREAT OOLITE. 35 dicated in the humerus of a small or young Cetiosaurus, figured by Phillips at p. 273, Diag. ct. The length of the Kirtlington humerus, figs. 3 and 4, is 4 feet 35 inches; extreme breadth of the proximal end 1 foot 8 inches; of the distal end 1 foot, 3 inches; diameters at the middle of the shaft 8 inches by 4 inches. The proportion of the ulna (fig. 5) to the humerus appears to be nearly that in the Monitor. The shaft is more distinctly three-sided, the anconal surface being strengthened by a median longitudinal rising or ridge not present in Monitor. As in this Lizard the palmar concavity excavates the whole of the upper half of that surface of the shaft except at the outer and inner ridged boundaries. The margin toward the radius is concave, the opposite one nearly straight, feebly convex. Both ends of the ulna of the Kirtlington Cetiosaur are wanting ; it measures in this state upwards of 3 feet inlength. In the section, fig. 5, a, the palmar side, a, is 12 inches across ; the facet, 2, of the anconal side is 11 inches; the narrower facet of the same side, c, is 7 inches. No recognisable bones of the fore foot of the Cetiosaurus longus appear as yet to have been discovered ; but the proportions of the known parts of the fore limbs are such as tomake | it more likely that they took their share in a quadrapedal mode of progression than that they were borne aloft, with the trunk, on the hind legs like the folded wings of a bird. The first almost entire femur of Cetiosawrus longus was obtained mainly through the personal care and supervision of Hugh E. Strickland, M.A, then (1848) of Merton College, from one of the divisions or thin bands of the ‘ Great Oolite’ underlying the Cornbrash near Enslow Bridge, north of Oxford. At the request of Dr. Buckland I inspected this specimen at the Geological Museum, in the ‘ Clarendon,’ Oxford, and by the light of fragmentary specimens from other Oolitic localities and correspondence of texture with the vertebral bones, especially those from Chipping Norton, I referred it to the Cetiosaurus medius. This determination was accepted by Mr. Strickland in his exhibition of the specimen to the Ashmolean Society, March 20th, 1848. The length of this femur is 4 feet 3 inches. In 1868 the femur of a larger individual of Cetiosaurus, and in 1870 other bones with vertebree answering to those of Cetiosauwrus longus, were discovered in the same quarries, close to the railway-station for Kirtlington, eight miles north of Oxford. Pro- fessor Phillips having notice of the first discovery took the requisite steps, with his wonted energy, to prosecute the quest and secure for his science the evidences of the monster dragon. The thigh-bone, first come upon, “ was found to be lying on a freshly bared surface of the Great Oolite, nearly in the line of a natural fissure, and covered by the laminated clay and thin oolitic bands which there occupy the place assigned to the Bradford Clay of Wiltshire.”? | 1 “Geology of Oxford,’ p. 247-8. ''36 FOSSIL REPTILIA OF THE Ni i Nh Femur, Cetiosawrus longus, 2th nat. size. (Phps., evili, p. 281.) The femur was 5 feet 4 inches in length. In the course of the quarrying works the opposite femur and many other bones of the same skeleton were brought to light. ''GREAT OOLITE. 37 The majority of these “did not actually touch the Oolite, still less were embedded in it, though single exceptions occurred of each circumstance’’ (p. 251). “The strata covering the solid Oolite were thus noted, March 21st, 1870: « Thin skerry beds of Forest-marble and shaly clay. Pe. Ge «“ Band of white calcareous concretions and clay QO 10 “Blue and greenish clay with white calcareous spots, and selenite 2 7 «“ Brown, yellow, and grey layers, argillaceous, sandy, and oolitic | 4 0) 2 “ Grey and argillaceous bed, with selenite “Grey and greenish bed loosely oolitic, with denetirst es maxillata, Avicula, Astarte . 0 8 “Clay and loosely aggregated oolitic parts, oath selenite and abet ance of carbonized wood, some shells, and most of the bones “ Clay below . ‘ ; : _ 20 6 od oO “ Great Oolite with undulated and waterworn surface. The two lower bands ‘die out’ to the southward, and there some of the bones came in contact with the rock, and others were engaged in it.” Phillips, ut supra, p. 251. The most striking of the remains here discovered was the fellow femur (right) of the one (left) found in the previous year. The anterior surface of the latter (left) is shown in cut 6. It is 5 feet 4 inches in length, the diameter of the middle of the shaft is 1 foot, that across the condyles 1 foot 5 inches. ‘The shaft is naturally sub- compressed, but the flattening has been exaggerated by posthumous pressure to which the closely cancellated texture of the interior of the shaft has yielded, with fracture of parts of the denser outer crust ; but there is no sufficient indication of the head, a, having been pressed so as to project inward, from any original disposition of that promi- nence forward, such as characterises the femur in modern Crocodiles and Lizards. The relation of the head to the shaft of the bone is thus more mammalian than saurian in the gigantic Cetiosaur. But the ‘neck,’ 4, is short, or almost nil; the trochanterian angle, e, not produced above the level of the neck or head. , The trace of any prominence for muscular attachment at the inner part of the shaft, d, is feeble; by no means such as appears in Scelidosaurus or Iguanodon. The distal end expands to the condyles, e, y, but in a minor degree than in the Monitor. The cut, fig. 7, shows the postero-external surface of the right tibia of Cetiosaurus longus. The prominence, a,is that which receives the outer condyle of the femur; the ''38 FOSSIL REPTILIA OF THE border, 2, in the view of the proximal end, gives the contour of the antero-internal part, which is rather flatter than in Monitor; ¢ shows the production above the procnemial ridge at the fore part of the bone; a is the part which was articulated to the distal epiphysis supporting the outer malleolus. The proportions of the tibia to the femur are less than in Monitors or Crocodiles; the length of that (fig. 7) adapted to the femur Tibia, Cetiosaurus longus , sth nat. size. (Phps., part of diagr. cix, p. 282.) (fig. 6) is 3 feet 2 inches; the breadth of the proximal end is 1 foot 5 inches; of the distal end 1 foot. As Professor Phillips remarks, ‘‘the terminal surfaces are strongly marked by the pitted adherence of cartilage, . . . . . which gives the appearance of deficient epiphyses.” ! 1 P, 282-3. He adopts an idea that the convex part of the anterior surface of the distal portion of the shaft of the tibia in the Crocodile is the homologue of the ascending process of the astragalus of ''GREAT OOLITE. 39 In a full-grown Monitor niloticus the distal epiphysis, which affords the articular surface to the astragalus, is unanchylosed, the line of suture closely resembling that in the distal end of the present fossil. Of the foot-bones, “three metatarsals of each foot were secured.” The largest appeared to be the first or innermost, the slenderest the third or outermost of the series. ‘Perhaps there were only three metatarsals, since the specimens we possess exhibit opposite pairs of three and no more” (p. 285). That these bones are homologous with those determined as the second, third, and fourth of the pentadactyle foot in Scelidosaurus and Iguanodon 1 deem more probable than that they answered to the metatarsals of the first, second, and third digits in Crocodilus. If a first or a fifth digit existed in the hind foot of Cetiosaurus, their shortness or rudimental condition may have prevented their recognition. In the description of the osseous characters then known of the largest species of Whale-Lizard, I remarked : “These enormous Cetiosawri may be presumed to have been of aquatic and, most Coracoid, Cetiosaurus longus, ysth nat. size. (Phps., part of diagr. xcviii, p. 268.) Megalosaurus, but “separated from its base and anchylosed to the tibia; while in Megalosaurus the con- nection remains, and the ascending process is not joined by synostosis to the tibia’ (op. cit., p.. 283). Scelidosaurus instructively exemplifies the homology of the distal epiphysis of the tibia in Dinosaurs with that in the Monitor and the Bird, and demonstrates the separate existence of the bone answering to the astragalus, &c., in both Crocodiles and Lizards, but which is not ossified in the tarsus of Birds. (Monogr. cit., p. 16, pls. x and xi.) ''40 FOSSIL REPTILIA OF THE probably, of marine habits, on the evidence of the coarse cancellous tissue of the long bones which show no trace of medullary cavity.”’’ | In reference to their affinities : “In the great expanse of the coracoid [fig. 8] and pubic bones, as compared with the Teleosaur and Crocodiles, the gigantic Saurians in question manifested their closer affinity to the Znaliosauria’ ?— closer, that is, than the Teleosaurs or Crocodiles show; but “their essential adherence to the Crocodilian type is marked by the form of the long bones of the extremities, especially of the metatarsals: and, above all, by the toes being terminated by strong claws.” Here, in 1842, the clawless character of the limbs of Plesio- and Ichthyo-sauri was the dominant idea, to the exclusion of the then novel group of Dinosauria, “characterised by a large sacrum composed of five anchylosed vertebrae of unusual construction,’ &c.° The question to be determined in respect to Cetiosaurus is the admissibility of the genus by the sacral character to the Dinosaurian order. This character, in 1842, I put in the van, relating as it does, physiologically, to terrestrial progression more after the manner of Mammalian quadrupeds than of existing four-footed Saurians, whether Crocodiles or Lizards; an extent of the trunk being thereby transmitted, through a co-extensive ilium, upon hind limbs, the chief bones of which are ‘medullary’ in Dinosauria. Ta a ee : f a at a i ne ee : q mee Ilium, Cetiosaurus longus, 35th nat. size. (Phps., cv, p. 278.) The ilium (fig. 9) of the Cetiosaurus longus, from the Kirtlington quarry, is estimated by Paituips as probably equal to six vertebra. He writes: «The extreme length of one (ilium) is 42, of the other 45 inches, probably equal to six vertebra,’ such sacral vertebrae being estimated each at a little over 7 inches in length. These vertebrae are briefly noticed as follows :—‘ Several bones of this portion are in the collection, but there is great difficulty in so placing them as to acquire a just notion of the structure or to present a satisfactory drawing. In some degree it (the sacrum) must 1‘ Report,’ ut supra, p. 102. = 1b, 10. = th, p. 102. * Op. cit., p.278: ''GREAT OOLITE. A] have approached that of Hyleosaurus.’' Perhaps a nearer one would be the sacrum in Scelidosaurus.” In either comparison the length of the sacrum is not to be estimated by that of the ilium. In Scelidosaurus, e.g., in which the number of sacral vertebree is ‘ four,’ the parts of the ilium anterior and posterior to the sacro-iliac symphysis, or surface of junction with such vertebrae, give to that pelvic bone almost twice the length of the sacrum. ‘The length of this part of the spine in Scelidosaurus is 10 inches, whilst that of the ilium is 18 inches, “a part, apparently a small one, being wanting from both extremities ”’ of the iliac bone. But, on this basis, we may allow to the ilium of 45 inches length a sacrum of 24 inches, or one of four vertebrae, each 6 inches in length. It is not probable that a saurian with iliac bones between 3 and 4 feet in length, and thigh-bones between 5 and 6 feet in length, would have a sacrum reduced to the crocodilian formula of two vertebree. Admitting, then, that more numerous sacrals, such as the Chelonians show, are not the sole and may not be the chief character of Dinosauria, and that the generalisation signified by that term is a passing one, denoting a step in the progress of knowledge of the extinct Reptilia ; and supposing that it should be now limited to saurian genera, combining, with four or more sacrals, the alternating or interlocking arrangement of the autogenous vertebral elements—as in Bothriospondylus, Megalosaurus, Iguanodon, Hylao- saurus—the question to be solved is :—‘ Does such arrangement characterise the sacrum of Cetiosaurus ?’ Have we, in the absence of any certain or definite knowledge of the cranial and dental characters of the genus, grounds for determining its ordinal relations to the Dinosaurs, Crocodiles, Sauropterygians, Ichthyopterygians, Lacertians, &c.? I am disposed to wait for such additional evidence, admitting, meanwhile, the faculty of terrestrial progression in a superior degree to that of the amphibious Crocodiles ; nevertheless, the habitual element of the Cetiosaur may have been, and I believe to have been, the waters of a sea or estuary. And I may here repeat the remark on the initial evidence of the species :—“ The main organ of swimming is shown, by the strength and texture and vertical compression of the caudal vertebra, to have been a broad vertical tail; and the webbed feet, probably, were used only partially, in regulating the course of the swimmer, as in the puny Amblyrhynchus of the Galapagos Islands, the sole known example of a saurian of marine habits at the present period.”® In fact, to the characters of the caudal vertebrae of Cetiosaurus longus known to me at the date of the above-quoted ‘ Report,’ viz.—‘‘ post-zygapophyses represented by hollow pits,” “slight concavity of both articular ends of the centrum, moderate compression of the sides between the expanded ends, which are subcircular,* the under surface concave lengthwise, 1 Op. cit., p. 257-8. 2 «Monograph on a Dinosaur of the Lower Lias, Scelidosaurus.’ Paleeontographical Society’s volume for 1860 (issued 1863), tab. vi, p. 6. 3 * Report,’ &., p. 102. 4 Ib., pp. 101, 102. ''42 FOSSIL REPTILIA OF THE marked by parial articular surfaces, showing the hemal arches to be articulate therewith over the vertebral interspaces” 1—the discovery of the grand proportion of the skeleton of the individual at the Enslow quarries adds a demonstration that the hemal arch in an anterior caudal vertebra (fig. 10) attained a length of 1 foot 2 inches; and that the neural spine ‘probably rose twelve inches above the canal,’” giving a total vertical extent of upwards of a yard to such anterior caudal. The vertebra probably Fie. 10. exceeded in this dimension at the middle of the tail. The modifications of the caudal vertebree in parts of the tail of Cetiosaurus longus, as exemplified by specimens from the Great Oolite described and figured by Phillips (‘ Geology of Oxford,’ 8vo, 1871), are similar to those in the instructively preserved Dinosaur from the Dorsetshire Lias). Compare fe. 2, p. 260 (Cetiosaur), Phps., op. cit. with fig. 1, Pl. VII (Scelidosaur), Ow., Monogr. cit., figs. 1, 2, 3, p- 265 (Cetiosaur), with figs. 1 and 3, Tab. IX (Scelidosaur) ; and figs. 1, 2, 8, p. 266 (Cetiosaur), with figs. 6, 7, 8, Tab. IX (Scelidosaur). The broad subquadrate coracoid, with rounded angles, of the Cetiosaurus longus from the Enslow quarries (fig. 8) repeats the characters of that bone in the type of the species (‘ Report,’ p. 102). In the Oxford giant the bone measures “from the glenoid cavity to the extremity near the scapular margin (incomplete) 18 inches; if complete, probably 20; Be cic canines amu. breadth between scapular and sternal margins, 18°5 inches ; Jath nat. size. (Phps.Ixxxix,p.259.) greatest thickness 5:0.’ (Phillips, op. cit., pp. 270, 271.) The scapula of Cefiosaurus resembles that in Scelido- saurus, with rather less concavity of the anterior border, and rather more concavity of the posterior one (comp. Fig. 2, with Monogr. cit., Tab. III, Scelidosaur). It sur- passes the humerus in length in a minor degree than in Scelidosaurus, and in a still less degree than in Lguanodon. In the characters of the dermo-skeleton Cetiosawrus would seem not to agree with Scelidosaurus. It is very improbable, if there had been such agreement, that not any skin-scutes or spines should be shown in connection with the large proportion of the skeleton of one and the same individual brought to light on the excavated oolite of Enslow Rocks at Kirtlington.’ The same negative evidence in all the various finds of fossil remains on which the genus was based suggested, in 1841, the idea that the tegument of Cefiosaurus might be 1 © Report,’ &c., pp. 101, 102. 2 Phillips, op. cit., p. 259. 3 Phillips, op. cit., diagr. Ixxxiv, p. 250. ''GREAT OOLITE. 43 smooth, or unarmed, as in Cetacea and Enaliosauria. But, as will be shown in the con- cluding subject of the present contribution to the History of British Fossil Reptiles, a new interest will attach itself to the occurrence of an osseous spine, seemingly dermal, in contiguity with the parts of the fore-limb which were wanting, or not discovered, in the Kirtlington example of Cetiosaurus longus. In Scelidosaurus the number of vertebrae between the skull and sacrum is twenty- three or twenty-four; in Jguanodon the same region includes more than seventeen vertebrae: in this genus there are five sacral vertebree: in Scelidosaurus four. In no Dinosaur has the number of caudal vertebrae been so satisfactorily or approximately demonstrated as in Scelidosaurus. Thirty-five of these vertebrae were obtained in con- secutive articular association in the individual which forms the subject of the Monograph above cited. If we allow the Cetiosaur, on this analogy, twenty-four vertebrae between the skull and sacrum, averaging 5 inches each in length, and add an inch for the inter- vertebral connective tissues, we get a total length of trunk at 12 feet. Four sacral vertebrae would add two feet. ‘Taking the number of the caudal vertebra at that shown in Scelidosaurus, and the reduction of length in the ten terminal ones not to be more than is shown in T'ab. IX, fig. 2, of the Monograph quoted, the length of the tail of Cetiosaurus longus may be set down at 17 feet. ‘Thus we get an approximative idea of the length of this Cetiosaur, minus the head, as 31 feet. The fortunate discovery of the skull or lower jaw, or a mandibular ramus, would supply the ground for completing an idea of the size of the whole animal. As the second femur of Cetiosaurus longus in the Enslow locality exceeded in size the first, so it may ultimately prove not to represent the extreme size attained by individuals of the species; and the length of 7 inches shown by the typical caudals would found an estimate of 35 or 36 feet for the length of trunk and tail of Cetiosaurus longus. As evidences of this species have now reached me from four counties—Yorkshire, Northamptonshire, Buckinghamshire, and Oxfordshire—I submit that there is no case, according to the ‘canons of zoological and botanical nomenclature’ adopted by the ‘ British Association for the Advancement of Science,” for suppressing the original name proposed by the discoverer of the species, and substituting one which is in some degree misguiding. I would also plead for a retention of the orthography of the generic name. £ Monogr. cit., p. 11. 2 Report,’ &c., for the year 1842. 5 In framing this name the diphthong in «/recos was dropped, as in ‘ pliocene,’ ‘ miocene,’ &c. '' ''MONOGRAPH OF A ROSSLL DiMOseU-R (OMOSAURUS ARMATUS, Owen) OF THE KIMMERIDGE CLAY. § 1. Inrropuction.—Having been favoured by James K. Shopland, Esq., Resident Director of the Swindon Brick and Tile Company, to whom the British Museum is indebted for the fossils of Bothriospondylus suffosus, described in a previous Monograph,’ with a note * announcing further discovery of larger bones in their Kimmeridge Clay works, followed by a liberal offer on the part of the Company’ of such of these fossils as might be found worthy of being added to the Geological Collection, Mr. William Davies, of that Department, was instructed to inspect the diggings, and, on his ‘ Report’ of the appearances, was autliorised to take the requisite steps to remove and transmit to the 1 Pal. Monog., ‘ Mesozoic Reptilia,’ “ Bothriospondylus,” &c., 4to, 1875, p. 15. _ ‘«Swinpon Brick anp Tire Company, “ Swinpon, Witts ; 23rd May, 1874. *““Drar Sir, “TI last year had the pleasure of sending you some Saurian Remains discovered in this Company’s Kimmeridge Clay Pits, and I beg to inform you that we have just laid open other remains considered to be unusually large and fine, which are left in situ, carefully covered over. “ As exposure to light and air will, I fear, cause the remains to split and crumble, I should suggest your coming or sending some one to inspect them at once; the clay adjoining I will leave unworked until Wednesday next. 1 am, &., “James K. SHopuanp. “ Proressor Owen, British Museum.”’ To this ‘Note’ were added a few lines from E. C. Davey, Esq., F.G.S., of Wantage, Berks, con- firmatory of the discovery. 8 It is due to their enlightened liberality and prompt co-operation in applying to the advance of science whatever, in the course of the works, might aid therein, to subjoin the names of the Directors of the Company :—J. C. Townsenp, Esq., Tuomas K. Suortanp, Esq., Henry Kriynerr, Esq., Ricarp Row ey, Esq. ''46 FOSSIL REPTILIA OF THE British Museum as much of the matrix as gave evidence or promise of containing organic remains. ‘This operation was carried out with Mr. Davies’ experienced skill and judgment. Some tons weight of matrix was transmitted to the British Museum, and occupied, during the remainder of the year, the practised chisel of Mr. Barlow, the mason-sculptor of the Geological Department, under the guidance and supervision of Mr. Waterhouse, Mr. Davies, and myself. The result was the extrication from these masses of the bones of one and the same individual dragon, or Saurian, and these form the subject of the present Monograph. They were found at a depth of ten feet from the surface soil covering the clay deposit, which deposit, where it surrounded the bones, presented unusual density and almost intractable hardness, and was traversed by fissures or cavities occupied by infiltrated spar, presenting in parts a septarian character. This condition of the matrix suggested that it might, in some degree, be due to the decomposition and exudation of the soft parts of the large reptile when buried in the clay sea-bed into which it had sunk; gaseous emanations might give rise to fissures or vacuities in the surrounding tenacious mass, into which the stalagmitic spar might subsequently infiltrate during the long ages of the condensation, petrifaction, and upheaval of the deposit; but cracks and cavities, from whatever cause, do become so occupied, as in the present local accumulation, and have received the name of ‘ septarian doggers.’ In the borings lately carried on at Netherfield, near Battle, Sussex, 660 feet of « Kimmeridge Clay ’ were traversed before the ‘Oxford Clay’ was reached, without inter- position of “Coral Rag’ or ‘ Coralline Oolite.’* This testimony to the time during which Kimmeridgian strata had been accumulated to such vertical extent gives free scope for surmise and speculation as to the long ages durmg which Phosaurs, Cetiosaurs, Bothriospondylian and other enormous reptiles, lived and died in a world of which they seem to have been masters, as far as grades of organic life and power, acting at that epoch, have been determined. ° Other lines of variation and modification of the dragon type, besides the new one about to be defined, probably remain to be determined by ulterior research, and to reward the labour, skill, and science of investigators and collectors of Kimmeridgian remains. Of the Dinosaurian genus and species, for which the name Omosaurus armatus® is proposed, parts of the vertebral column, the pelvis, a femur, and tibia, and almost all the bones of the left fore limb, have been worked out. The scapular arch, sternum, skull 1 See the processes described by him in his instructive ‘Catalogue of Pleistocene Vertebrata in the Collection of Sir Antonio Brady,’ 4to., 1874, p. 71. 2 A thickness or vertical extent of 1050 feet is assigned to the combined ‘upper’ and ‘lower’ divisions of the Kimmeridge Clay, by the Rev. J. F. Blake, M.A., F.G.S., in his instructive memoir on this formation in England, ‘ Quarterly Journal of the Geological Society,’ vol. xxxi, p. 196. 3 *Quos, humerus ; Zavpos, lacertus : suggested by the unusual development of the muscular crests and processes of the arm-bone, perhaps in relation to the formidable weapon with which the fore limb appears to have been armed. ''KIMMERIDGE CLAY. 47 and teeth, and bones of the hind feet, are still desiderata. That not a single tooth should have been met with in any part of the ossiferous matrix is much to be regretted, but one indulges the hope that teeth of Omosawrus may be one day recovered and be found implanted in their jaws. § 2. Cervican Verresra.—A portion of a neural arch and spine (Pl. XI, figs. 1 and 2), with the right prezygapophysis, z, the left postzygapophysis, 2, the roof of the neural canal, », and the entire neural spine, 2s, might belong, from the shortness -of the latter, to a caudal vertebra. But, from the indicated capacity of the neural canal and the aspects of the articular surfaces of the zygapophyses, I infer the specimen to have belonged to a vertebra from the cervical region. The length of the neural arch is 7 inches 6 lines; the height of the neural spine is 3 inches 6 lines; its fore-and-aft breadth, at the middle, is 2 inches; at the free end 3 inches; the thickness, transversely, is 1 inch: this is at the hind border, near the summit ; it slightly decreases toward the base, and the whole spine thins toward the fore part. The summit, which is rugged, gains in extent by being produced backward. The diameter of the neural canal appears to have been 14 inches. ‘The prezygapo- physis, <, projects about half an inch in advance of the base of the diapophysis, ¢, d, which here has an antero-posterior extent of 2 inches 6 lines. ‘The outer border of the prezygapophysis is slightly raised above the base of the diapophysis; the articular surface of the prezygapophysis looks upward and slightly inward; it is not quite flat, but feebly convex. ‘The articular surface of the postzygapophysis, 2',is in the same degree concave. ‘This surface looks downward and a little outward. In the figure of the upper surface of a cervical vertebra of a large Monitor Lizard (Varanus niloticus, Cuv., Pl. XI, fig. 4) I have indicated by dotted lines the course of the fractures which have reduced the corresponding vertebra of the huge Dinosaur to the condition shown in Fig. 2. ‘The relation of the origin of the diapophysis, d, to the prezygapophysis, z, is the same in both the recent and fossil Saurian; but the breadth across the zygapophyses was relatively less to the length of the neural arch in Omosaurus. The fragmentary condition of this solitary evidence of the region of the vertebral column supporting the skull seems to point to some strange violence by which the head of the Omosaur has become severed from the trunk, and its frame-work probably borne to some part of the old sea-bed at a distance from the rest of the body. § 3. Dorsat Vertesrm.—Amongst the characters of the Order Dinosauria is a lofty and buttressed neural arch in a great proportion of the trunk-vertebre. In former Monographs this is illustrated in the Zguanodon (Suppl., No. II, Pal. vol. for 1857, Tab. VII, figs. 4—6); in the Megalosaurus (Pal. vol. for 1854, ‘ Wealden Reptilia,’ Tab. XIX); in the Cetiosaurus (Suppl., No. I, Pal. vol. for 1857, Tabs, VIII, IX). ''AS FOSSIL REPTILIA OF THE __ This character is strongly marked in dorsal vertebra of the present genus, and with modifications which could hardly have been illustrated or made clear without the above- cited figures of the vertebra of previously defined Dinosaurian genera. In these, however, the degree of complexity of the neural platform varies ; it is least marked in the smaller and more crocodilicid genus Scelidosaurus (Pal. vol. for 1860, ‘ Liassic Dinosauria,’ pp. 5—7, Tabs. II—VI). The vertebra of Omosaurus—the subject of Plates XII, fig. 1, and XIJI—has come from the middle of the trunk. This is inferred from the position of the surface, », for the head of the rib, which has risen from the centrum, or base of the neural arch, to near its summit, where, with its diapophysial productions, a, a, the arch expands to a breadth of 14 inches 6 lines ; the breadth (in the same direction, transversely) of the centrum being 5 inches 8 lines. ‘The vertical diameter of the middle of the articular surface of the centrum is 4 inches 9 lines; the height of the vertebra to the base of the neural spine is 11 inches. This spine has been worked out entire only in the above-described cervical and caudal vertebrae ; but there are indications justifying an estimate of its length in the dorsal series, at from 6 to 8 inches. Thus, the dorsal vertebra, affording material for the present description, which has a breadth, as above shown, of one foot two and a half inches, had a height of at least one foot and a half. The fore-and-aft dimension of the centrum (Pl. XIII, fig. 3) is 4 inches. The anterior surface (ib., a), where it varies from flatness, is toward convexity, but in the feeblest degree ; the posterior surface (ib., 4) is very slightly, but more equably, concave. The free surface of the centrum is moderately concave longitudinally ; slightly depressed at 7, beneath the base of the neural arch. The tissue throughout the vertebra is more compact than in Cetiosaurus (Mon. cit., Tab. IX, fig. 2). The neurapophyses (P]. XII, fig. 1, » p) have coalesced with the centrum ; they quickly narrow transversely, above their base, to a thickness of half an inch, more gradually contract in fore-and-aft dimension (Pl. XIII, fig. 2, » p) to two inches and a half. Over- arching the neural canal (Pl. XII, fig. 1, n), they meet and coalesce about one inch and nine lines above the centrum, whence their compact coalesced mass rises above the crown of the arch, expanding to a height of five inches (posteriorly, Pl. XIII, fig. 1) before giving off the neural spine (ib., n s). At three inches above the base the outer surface of the neurapophysis is excavated by a smooth oval cavity (ib., fig. 2, p), 1 inch 9 lines in vertical, 1 inch 6 lines in transverse, diameter, and about 8 lines in depth. ‘To this cavity was adapted the ‘head of the mb:’ for this part there is no parapophysis, or outstanding process. Below the capitular cavity the outer surface of the neurapophysis is divided from the hinder surface by a low obtuse ridge or angle (ib., ib., e); a broader ridge (ib., ib., a), also low and obtuse, rises along the middle of the outer surface of the neurapophysis, and expands to form the lower margin of the costal pit. In advance of this pit the ''KIMMERIDGE CLAY. 49 neurapophysis extends forward to form the prezygapophysis (ib., and Pl. XII, fig. 1, <). The ridge (e), rising to the costal pit, forms or extends its hind border and is thence continued, expanding or thickening, into the ridge which forms the diapophysial buttress, f. The ridge (Pl. XIII, fig. 3, a) does not, in this vertebra, combine with the ridge, e, to form the buttress, as in the Zywanodon (Mon. cit. Tab. VII, fig. 2), but appears as a shorter independent ridge. A median ridge (Pl. XII, fig. 1, r) rises from above the interspace of the prezygapophyses to the neural spine, x s. Another median ridge (Pl. XIU, fig. 1, s) extends along the back of the neural arch and rises to the interspace of the postzygapo- physes, -’, 2’. The chief expanse of the summit of the neural arch in the antero- posterior direction is a zygapophysial one (Pl. XIII, fig. 2, z, «’); im the transverse direction it is a diapophysial expansion (ib., fig. 1, d, d). Each diapophysis is three-sided; the broadest facet is on the upper side, forming with the zygapophyses the neural platform. External to the zygapophyses this surface is 24 inches from before backward; it’ is flat. The postinferior surface (Pl. XIII, fig. 1, f, d) isin that direction concave, most so below the postzygapophyses, /, and growing shallower to the tumid extremity, d, of the transverse process. ‘The least fore-and-aft diameter of this surface of the diapophysis is 2 inches 3 lines, that of the antero-inferior surface is 1 inch 5 lines; this is feebly concave across, and is divided lengthwise for part of its extent by the zygapophysial ridge (Pl. XII, fig. 1, ¢). The free end of the diapophysis is swollen and tuberous; a well-marked facet (Pl. XII, fig. 1, d, and Pl. XIU, fig. 2, a) cuts the lower part obliquely; it is of a rhomboid shape, nearly flat, and is roughened for the ligamentous attachment of the “tubercle of the rib ;’ it measures 24 inches by 1 inch 9 lines. The postzygapophyses (Pl. XIII, fig. 1, -’ <’) are formed by an expansion backward of the neural platform, the pair of processes being indicated by a medial notch; they are more clearly defined by their flat articular surfaces, which are subtriangular in shape, the angles being rounded off ; their longest diameter is 2 inches: they look outward and downward. The prezygapophyses (Pl. XII, fig. 1, z, 2) have been mutilated in the present vertebra, but the extent of their basal origin, 2 inches, may be traced ; they are more distinct productions of the neural platform, which abruptly sinks to the level of their medial borders. The anterior basal ridge (ib., 7) of the neural spine begins at this lower part of the platform, which it divides into a pair of hollows. The spine rises freely from the broader upper level of the platform. Its base here has a fore-and-aft extent of 3 inches 8 lines. The hind border of the spine is rather sharp; the thickest part of the body of the spine is 9 lines; its free termination was probably, from the analogy of a caudal vertebra subsequently to be described, swollen and tuberous. A vertebral centrum and a portion of the neural arch, from the same region of the spinal column, repeat the characters, so far as they are shown, of the less fragmentary ''50 FOSSIL REPTILIA OF THE vertebra above described and figured. T'wo views of the centrum, of half the natural size, are given in Plate XII, figs. 2 and 3. The capacity of the neural canal (fig. 2, ») is: worthy of note; it is rather Mammalian than Saurian, and implies a great development. and vigour of the muscular system. § 4. Lumbar Vertesra.—The last lumbar vertebra (Pl. XIX, 7) appears to be: confluent with the first sacral (ib., 5 1). Its centrum is 3 inches in longitudinal extent ; the side is slightly depressed below the base of the neural arch, from which extends a lumbar rib (ib., 7, p 7) 9 inches in length; this is 1} inches in breadth at three inches distance from its free extremity. This lumbar rib, and also that of the antecedent lumbar vertebra, are straight and extend transversely to the axis of the vertebral column. ‘The distance in a straight line from the hemal surface of the lumbar centrum to the end of the last lumbar rib is. 1 foot 3 inches. § 5. Sacra Verrepra.— These are five im number (ib., s 1—s 5), coalesced together, and seemingly with their pleurapophyses. The antero-posterior extent of the- five sacral centrums is 1 foot 43 inches, each centrum averaging 33 inches in length. After the first they increase in breadth and decrease in the transverse convexity of the: heemal surface, the middle ones showing traces there of a shallow longitudinal hemal channel with thick low convex borders. ‘The interspace between the heads of the third. pair of sacral ribs (ib., p/. 3) is 7 inches, between the fifth pair it is 6 inches. Fractures of the mass of matrix enveloping the pelvis exposed the close cetiosaurian texture of these vertebrze and the shape, in some degree, of the neural canal ina portion of the sacrum. One (fifth) sacral vertebra was thus divided lengthwise through the centrum, neural arch, and spine, and yielded the following dimensions :—Vertical extent 1 foot 5 inches; ib., length of neural spine, 6 inches; antero-posterior diameter of do., 3 inches 6 lines. This spine for a great part of its length was not in contact with the ante- cedent neural spine. The neural canal partially depresses the upper surface of the centrum of each sacral vertebra, probably in relation to venous sinuses rather than to ganglionic: enlargements of the myelon. ‘The vertical diameter of the neural canal where it dips down into the centrum is 2 inches 3 lines; in the ordinary course of the canal, it is 1 inch 2 lines: but, as the fracture affording this view was not exactly along the middle of the vertebra, the canal might gain more depth at that part. The central part of the sacral centrum shows a rather coarser cancellous texture than the rest, or than is seen in any part of the centrum of an anterior caudal vertebra (PE XI, fig. 1). | What appears to be the first sacral rib (Pl. XIX, pl. 1) is slightly dislocated heemad, and probably, at the same time, bent forward obliquely from above downward and backward in a greater degree than natural, the hemal end of the articular surface ''KIMMERIDGE CLAY. 51 projecting a couple of inches in advance of the second sacral rib (ib., pl. 2). The long or vertical diameter of the head or articular end of this rib-plate is 6 inches; at 3 inches of its outward course it expands to a breadth of 7% inches by a convex extension of the fore border, which appears to have articulated like a rib-tubercle with the neural arch, and to have been underlapped by part of the ilium (Pl. XIX, a). Beyond this point the rib- plate, as it approaches the acetabulum, diminishes. in breadth but increases in thickness and seems to develop from its haemal side a broad, transversely convex ridge or buttress (ib., pl. 1) 5 inches long by 23 broad at the distal end, which abuts upon the fore and heemal angle of the acetabulum, e. A process of the antacetabular part of the ilium (ib., «) is continued inward and hemad to articulate with the upper border of this first broad, sacral rib ; an oblong vacuity, 4 inches by 2 inches, intervenes between this process of the ilium and the acetabulum. ‘The second sacral rib (ib., p/. 2) is indicated by the part of the plate posterior to p/. 1. The proximal portion of this seemingly smgle broad and bifid pleurapophysis is applied to the greater part of the sides of the two anterior sacral centrums (ib., s 1, s 2), showing it to be the confluence of two pleurapophyses, the part described as the convex side or buttress being the distal articular end of the anterior of these. On this view the next independent sacral rib would be the third (ib., pl. 3) ; its proximal end is expanded and applied by a similar, but not so great, obliquity to the side of the third sacral centrum (ib., s 8), having a breadth of 3 inches with a thickness of nearly 2 inches, but contracting to a narrow rounded hemal border, retaining above this part a thickness of 1 inch, then expanding to a breadth of 3 inches to abut upon the hemal border of the acetabular part of the ilium, filling the interval between the like extremities of the second and fourth sacral ribs. The direction of the third pair is nearly trans- versely outward. The length of the interspace between the second and third ribs is 6 inches; the fore-and-aft breadth is 34 inches; it narrows towards the acetabulum, where the distal expansions of these ribbed buttresses come into contact and seemingly coalesce with each other, and similarly narrows to their proximal expansions, thus showing an elliptical shape. The head of the fourth sacral rib (ib., pl. 4) is applied to the whole side of the corresponding centrum (s 4), and is 34 inches in fore-and-aft diameter; from this the rib contracts to the form of a subvertical thick plate, and then expands to a breadth of 4 inches applied to, and confluent with, the lower border of the acetabulum and a considerable extent of the medial surface of the ilium rising therefrom. The fifth sacral rib, with the head reduced to 24 inches in fore-and-aft extent, is applied to the side of the last sacral centrum (s5). This rib, contracting at first like the previous ones, then expands as it extends outwards and slightly backwards, chiefly in the vertical direction, to be applied for an extent of 5 inches to the part of the acetabulum to which the ischium is articulated. A considerable part of the right ischium (ib. 63) is retained, dislocated a few inches from the articular facets (ib., 4, c)» h ''FOSSIL REPTILIA OF THE ew 5 and thrust a little mesiad and forward. This bone will be subsequently described showing the proportion of the acetabular cavity contributed by it. Anterior to the pelvis is a dislocated group of eight hinder trunk-vertebree, each retaining more or less of its neural arch and processes. On the right side of the pelvis a complete dorsal vertebra is exposed, measuring 1 foot 5 inches in length and 13 inches in breadth, between the diapophyses. The centrum is 3 inches 9 lines in leugth, 5 inches in breadth, 44 inches in height, to the base of the neural canal ; the hinder outlet of this is pyriform, the apex about 24 inches in vertical, and 13 inches in transverse, diameters. From the floor of the neural canal to the base of the spine is 8 inches; the length of the spine is 5 inches. Beyond this dorsal vertebra is the body of a caudal one, showing a greater degree of concavity of the fore surface of the centrum, which has a breadth of 6 inches. Behind the sacrum is a dislocated group of four caudal vertebra, mainly agreeing in character with the subject of Pls. XIV and XV. § 6. Cavpat Verresra.—tThe vertebra of Omosawrus which has been most perfectly wrought out of the matrix is one from the base of the tail; it was in the same block with the sacrum, not far from the hind part of the pelvis. This anterior caudal vertebra forms the subject of Pls. XIV and XV, of the natural size; and I here subjoin, also, the following admeasurements : In... Linas. Height or vertical extent of the entire vertebra 14 9 Breadth of ditto : P : . 14 6 Length at the zygapophyses, giving extreme length of neural arch. 4 2 Centrum, length, lower surface 2 410 : upper surface 2 5 x breadth, anterior surface 3 8 if posterior surface 6 0 height, anterior surface 4 5 ‘ - posterior surface 4 6 Neural canal, vertical diameter 2 0 . transverse diameter, least J 4, Neural arch, breadth at upper level of centrum 5 8 As i across prezygapophyses 4 5 7 ‘ » postzygapophyses 2 2 Pleurapophysis, length from base to apex 4 0 . depth from tubercle to under surface 2 3 i thickness, extreme, at base . : 6 ‘ i at tubercle : & 0... 19 pe Gh “ below tubercle 0 8 ''KIMMERIDGE CLAY. 53 In. | Lines. Neural spine, length from fore part of base 7 1 a b, hind part of base . 5 6 - fore-and-aft breadth at mid-length iE 8 - transverse breadth, at mid-length i 0 ! at tuberous end 2 5 29 2? A comparison of such of the above admeasurements as have been recorded of trunk- vertebrae shows that the caudal ones become shortened, at least, at the basal part of the tail. As the length of this appendage would depend upon the number of vertebra, and especially of those reduced nearly to the centrum, which might again gain in length, it would be premature, on present evidence, to hazard an opinion on this dimension in Omosaurus armatus. But the size of the outstanding parts for muscular attachments indicates great power in the tail, which would probably be exercised, as in the largest living Saurians, in delivering deadly strokes on land, as well as in cleaving a rapid course through the watery element. The centrum is transversely elliptical, with both upper and under surfaces sloping from before downward and backward from the terminal articular planes, these being vertical. Of them the anterior (Pl. XTV, fig. 1, a) is flat, with a slight convexity toward the periphery and a shallow transverse groove at the centre; the posterior surface (Pl. XV, fig. 1, 4) is more decidedly, though but slightly concave ; the deepest part here, being along a central transverse groove, with a slight upward bend, like that on the opposite surface. A rugged border for the attachment of a capsular ligament projects from two to five les beyond the articular tract. This, though smoother than any part of the free surface of the centrum, has evidently, by its inequalities or sculpturing, related to a syndesmosal jot, as in the Chelone and Mammalia, not to a synovial one as in Crocodilia. Between the fore and hind borders of the centrum the lower surface is antero-posteriorly concave (PI. XV, fig. 2), the concavity narrowing as it approaches the line of confluence of the pleurapophysis (ib., ib., p!). This line begins below, half way between the under and upper surfaces of the centrum, and extends upward, approaching obliquely the fore surface (ib., 2) to overlap and be lost (by anchylosis) in the base of the neurapophysis ; a feeble trace of the primitive separation of this element may be discerned at the hinder outlet of the neural canal (ib., fig. 1, ap?). ‘The pleurapophysial line of confluence is more distinctly traceable ; the base of the pleurapophysis, representing the head of the caudal riblet, is broadest below, and there extends nearer the posterior than the anterior surface of the centrum; but, as it rises, it narrows and leaves a larger proportion of the post- lateral surface of the centrum free. The ‘tubercle’ (¢) of the rib is a well-marked rough prominence at which the upper border of the rib descends at an open angle with the ‘neck ’ to its obtuse apex. The under border of the riblet is gently concave lengthwise. No diapophysis has been developed, in this vertebra, to afford abutment to the tubercle. ''BA FOSSIL REPTILIA OF THE Each neurapophysis at its confluence with the centrum gives a triangular horizontal section (Pl. XIV, fig. 3, mp), the base of the triangle, 1 inch 5 lines, being anterior, the obtuse apex behind. The inner, shorter side, next the neural canal, is parallel with its fellow and the trunk’s axis, the outer side, 2 inches 9 lines in extent, slopes from the broad fore part backward and mesiad to the hind margin of the neural arch. From the upper and anterior forwardly sloping part of each neurapophysis the prezygapophysis (¢) is developed ; it is short, thick, obtuse, with a flat articular surface, looking upward, inward, and slightly forward ; subcircular, an inch in diameter. From the narrower hind part of the neural arch the common base of the pair of postzyg- apophyses (¢’, *') rises, expanding to form their articular surfaces, which look in directions opposite to those in front. ‘The hind surface of the common base of these articular expan- sions has a wide and deep vertical channel.’ The neural spine (as) is subquadrate at its base, with the lateral angles broadly rounded off (Pl. XV, fig. 2, 2s). The line of attachment of the base of the spine rises from before backward (ib., fig. 3). A median anterior ridge (Pl. XIV, fig. 1, ~) strengthens the lower half of that surface, as a similar but thicker ridge (Pl. XV, figs. 1 and 3, s) does the posterior corresponding tract. Where these ridges cease the spine begins to expand into its rough obtuse summit, chiefly transversely, so as to give it an elliptical contour extended in that direction (Pl. XIV, fig. 2). The foremost of the caudal vertebrae remains in the block of matrix with the sacrum. The present I take to be the second of the series. ‘There is no trace of hypapophysis for a heemal arch in either of these caudals (the under surface of the centrum of the second is figured in Pl. XIV, fig. 4). In Scelidosaurus the first or foremost caudal alone is devoid of hamal arch; in the second caudal the lower part of the hind border is touched by the smaller anterior facet on the base of the hemapophysis (‘ Pal. Monogr.,’ vol. for 1860, p. 8, Tab. VII, figs. 1 and 2, ¢ 2). In the few succeeding caudal vertebra, with diminution of general size, the vertical extent and the length of the pleurapophyses decrease in a greater ratio. A larger proportion of the side of the centrum is left free below the rib’s confluence therewith 5 and this free surface of the centrum shows, as in the specimen selected for Pl. XVI, an upper (¢) and a lower (“’) depression. The transverse extent of the centrum decreases without corresponding loss of vertical extent. The hind surface of the centrum (b., fig. 2) becomes more concave, without corresponding increase of convexity of the fore surface. The contour of the hind surface approaches the subhexagonal. The anterior and posterior ridges of the neural spine subside ; the fore ridge is longest retained, but shrinks toward the base of the spine, as at 7, fig. 1. In the subject of this Plate, as in three other caudals extracted from the matrix, the neural spine has been bent to 1 It is possible that a similar facet may have been ligamentously attached to the rough surface extended from the lower margin of the terminal surface. ''KIMMERIDGE CLAY. 55 one side, as shown in Pl. XVI, fig. 2. This distortion I conceive to be due to movements of the matrix after the fossil had been inclosed thereby and become petrified therewith. For, being thus supported at every point by the matrix, during the slow and continuous partial pressure, the spine has yielded and bent without breaking. In one instance the sustaining neural arch has suffered partial fracture at the side (ib., fig. 1), toward which the spine has been bent. A thickening at the outer side of the neurapophysis, feebly indicated in the larger anterior caudals (Pl. XV, fig. 2, 2p), becomes more prominent near the base of the prezy- gapophysis, as at xp, figs. 1 and 2, Pl. XVI, in the succeeding smaller vertebra, in which the hypapophyses are more distinctly marked. These articular protuberances (ib., figs. 1—3, Ay) form a pair at the hind border of the inferior surface of the centrum ; the articular tracts at the fore border of that surface are barely defined, or may be indicated by an extension backward of the rough marginal syndesmosal tract. The caudal vertebra in Pl. XVI is figured a little more than half the natural size. The answerable caudals in the great Monitor Lizard (Varanus niloticus) are given, of the natural size, in figs. 4 and 5. The heemal arch in the caudal vertebra, with a centrum 54 inches in vertical extent, has the same length. The hemapophyses (ib., fig. 2, %) are 23rd inches in length before coalescing to form the spine (ib. ib., 4s), which is 23rd inches in length in the subject of the Plate; it was probably longer when quite entire. But the length of the arch and spine was plainly less in proportion to the vertical extent of the rest of the vertebra than in Cetiosaurus longus. 'The hypapophyses are accordingly relatively smaller, and are limited to a narrower transverse extent of the inferior surface of the centrum (ib., fig. 3, Ay) than in Cetiosaurus, ov in the recent Varanus (Pl. XVI, fig. 4, Ay). In Cetiosaurus brevis (Pal. vol. for 1857, T. X) the hypapophysial facets (4, %) are broader and wider apart than in Cetiosaurus longus. - In Zguanodon the reverse conditions prevail. ‘These surfaces have become confluent, and present a single bilobed facet to the similarly confluent surfaces on the bases of the right and left hamapophyses. Both neural and hemal spines are relatively longer in Tguanodon ; and the neural spine springs from a smaller proportion of the hind part of the neural arch at a much greater distance behind the prezygapophyses’ than in Omosaurus. The caudal vertebre differ less from each other m Omosaurus and Cetio- saurus than they do in either of these genera as compared with Jguanodon. Asin the case of Cefiosaurus longus and other Dinosaurian subjects of previous Mono- graphs, I have selected the best preserved specimen of an average-sized vertebra for the subject of figures of the natural size. As the number of these monstrous dragons increases, the judgment of the Council of the Palaeontographical Society in voting the cost 1 Pal. Monogr. for 1854, tt. viii and ix. See also the young or small kind of Jguanodon, Mon. cit., ,i,h, in which this vertebral character is significantly repeated. ''56 FOSSIL REPTILIA OF THE of such ‘ Plates’ will be appreciated, the requisite comparisons being much facilitated, and accurate results ensured, by such life-size figures. § 7. Humervus.—Of the skull, teeth, or scapular arch of Omosaurus I have not as yet received evidence. The humerus and some other bones of the left fore limb (Pl. XVID) have been relieved from the matrix in a more or less complete state. The humerus (ib., figs. 1—5) is remarkable for its breadth, especially at the proximal half, compared with the length. The articular surfaces at both ends have been more or less abraded. ‘That at the proximal end (figs. 1 and 2, a and fig. 3, a) shows the elongate oval form, with the larger end, ¢, toward the ulnar aspect, narrowing to the beginning of the great radial crest, 6’, 6, as in Crocodilus' (Pal. vol. for 1858, p. 15; ‘Cretaceous Reptiles,’ Tab. III, fig. 12), Varanus, and most existing Saurians ; as in these, also, the head projects somewhat toward the anconal surface (as at a, fig. 2); but the prominent part of the shaft continued therefrom is Jess marked than in Cetiosaurus longus (Pal. vol. for 1875, ‘ Mesozoic Reptilia,’ p. 32, fig. 3). The radial tuberosity (Pl. XVI, figs. 1 and 2, g) is not developed distinctly as such, but, as in Crocodilus (Monogr. cit., Tab. III, fig. 11, 4) and Varanus (Pi. XVII. fig. 6, 2), is the beginning of a plate or crest of bone, answering apparently to both the deltoid and pectoral in Mammals, which plate extends considerably radiad, but with less inflection palmad, than in Crocodilus or Varanus, so that more of its breadth is seen in a direct palmar view, as in fig. 1, than in the Pterodactyle or the above existing Reptiles.. It has a certain forward or palmar bend, and subsides a little below the middle of the shaft. From the proximal beginning, 2, of this great crest, a broad tuberous rising (ib., fig. 2, @) projects anconad, and is continued, narrowing obliquely distad, to terminate or ‘subside at the radial side of the shaft, close to the termination of the crest 3’. The tuberosity and ridge, d, @, might be regarded as ‘deltoidean,’ as distinct from the ‘pectoral’ 4, 4’, save that its position is anconal instead of palmar. ‘There is a rudiment or indication of this ‘anconal ridge’ in the humerus of the Crocodile (Monogr. cit., vol. for 1858, Tab. IIT, fig. 10, 7), and a shorter one in Varanus. In the latter existing Saurian it gives origin to a muscle answering to the external ‘head’ or portion of the ‘triceps extensor cubite’ in Mammals. The ulnar tuberosity extends ulnad and distad as a thick tuberous ridge, which terminates more abruptly than the radial crest, at c, figs. 1 and 2, about seven inches beyond the proximal end. ‘The broad surface of the humerus between the crests is rather concave across on the palmar surface, somewhat more convex on the anconal surface, which is interrupted by the ‘ anconal or tricipital tuberosity and ridge.’ 1 The terms of aspect and position are the same as those defined,'p. 13 of the above-cited Monograph. Mr. Hulke uses the term ‘ ventral’ to signify the ‘anterior’ or ‘palmar’ surface, and ‘ dorsal’ to signify the opposite, ‘ posterior,’ or ‘ anconal’ surface; but he also applies the term ‘ulnar’ as synonymous with ‘ posterior ’ in his description of a humerus referred to the genus Hyleosaurus.—‘ Quart. Journal Geol. Soc.,’ 1874. ''KIMMERIDGE CLAY. of The shaft at its narrowest part presents in section the form given in fig. 5, Pl XVII, being almost flat, palmad, and convex anconad transversely. It soon begins to expand into the distal end of the bone. The crest, ¢, simulates the ‘ supinator’ one in Mammals, and is not perforated, as is the answerable disto-radial crest in some existing Saurians. Such perforation is very small in Varanus (Pl. XVII, fig. 6, e’). There is no indication of this vascular or nervous canal in Omosaurus, and the crest is relatively shorter than in Varanus. The ulnar expansion, /, of the distal end is thick and tuberous. Sufficient of the radial condyle, y, remains to show its Saurian extension palmad, and its convexity in Omosaurus (ib., fig. 4) ; the precise form and extent of the less prominent ulnar condyle or trochlea is not definable. , The texture of the shaft of this humerus, as exposed by the fracture across its middle narrowest part, is compactly dense; there is a small medullary cavity (fig. 5) which seems to have but a short longitudinal extent. A deep anconal depression (ib., fig. 2, ;), marks that aspect of the distal expansion in a greater degree than in any Crocodilian, Lacertian, Dinosaurian, or Pterosaurian humerus that, as yet, has come under my notice ; it gives to this part of the humerus of Omosaurus something of a Mammalian character. The following are admeasurements of the humerus : ' Pe) Ini hes, Length 4 ‘ : 2 9 0 Breadth across radial or pectoral crest 1 6 0 o distal end Oy Bh 0 aA middle of shaft 0 5 6 Girth of | ae 1.0 0 Length of base of radial or pectoral crest 1 A 0 - ulnar crest 0 8 0 The figures of this bone on Pl. XVII are reduced to one fourth of the natural size. Although I should have hesitated to found a genus or generic term on a solitary limb-bone if such distinction had not been supported by the vertebral characters, yet the features were so much more strongly marked in the present than in previously described or figured humerias to have afforded a better excuse for such taxonomic deduction, which ought to rest, and, as a rule, can only safely do so, on characters afforded by associated parts of the skeleton or teeth. Mutilated as are the humeri discovered with unquestionable vertebrae of Cetiosaurus longus in the Geological Museum of Oxford, justifying the conclusion that they belonged to the same individual, they are unmistakably distinct in character from that bone in Omosaurus. Although the radial or pectoral ridge be broken away in the subjects of figures of the ''58 FOSSIL REPTILIA OF THE preceding Monograph,’ its base has a minor relative extent than in Omosaurus ; the shaft beyond that ridge expands more gradually into the distal end; the entire length of the bone—4 feet 4 inches in Cetiosaurus longus—is greater in proportion to the breadth or thickness of the shaft. The more slender character of the humerus is still more marked in that bone which chiefly represents Mantell’s genus Pelorosaurus (Pal. vol. for 1857, Tab. XII), in which the radial or pectoral crest (ib., fig. 2, 2) subsides above the middle of the shaft, encroaching, as in the Crocodile, Varanus, and Pterodactyle, upon the palmar surface of the bone. The humerus of Jywanodon is still less robust in proportion to its length, not to mention its inferior size as compared with associated dorsal vertebra, than in Omosaurus. In Hyleosaurus we find the nearest approach to Ozosaurus in the proportion of the length of the humerus giving attachment to the great tuberous crests from the radial and ulnar sides of its proximal part. But in the Isle of Wight specimens referred, with doubt, to that Dinosaur, the radial crest is more strongly, and, in reference to its Saurian nature, more typically twisted palmad than in the huger Kimmeridgian genus. It shows a tuberous thickening anconad of its distal end, in the place of the ridge, a’, fig. 2,. Pl. XVII, in Omosaurus? § 8. Raprus.—This antibrachial bone in Omosaurus (Pl. XVII, figs. 7—11) has a subcompressed shaft, expanding moderately and almost equally into the two articular ends, as far as their degree of conservation shows; but it is probable that the more mutilated distal end (fig. 10) when entire would give a somewhat greater breadth than the proximal one or ‘head.’ ‘This (ib., fig. 9) is of a narrow subelliptic shape. A small part of the concave articular surface, a, for the radial condyle of the humerus, is preserved. The anconal surface of the shaft (fig. 7) is feebly divided at its distal two thirds into two facets by a low rising, hardly to be called a ridge, beginning at the middle of that surface at its proximal third and inclining as it descends toward the radial border of the distal end. The concavity of both borders, and especially of the ulnar one, narrows transversely the shaft, but this preserves more equably its ancono-palmar thickness (see the section of the middle of the shaft in fig. 11). The lateral facet (fig. 8, 4) at the proximal end for articulation with the ulna is more convex than is usual in Repéilia. The surface (ib., fig. 8, e) for the insertion of the biceps tendon is well defined. The thenal prominence (ib., figs. 8 and 10, f) extending or deepening the cup, y, for the scaphoid, is strongly developed, and is thicker than usual, as far as it is preserved. Its. outer surface is roughened, as if for the ligamentous attachment of some bone, such sur- face extending to the angle, 9 (fig. 8), at the broadest part of the distal end of the radius. 1 On the genus Cetiosaurus, p. 31, fig. 3. * Compare Plate XVII, figs. 1 and 2, with similar figures of the humerus of Hyleosaurus? given. by Mr. Hulke, in the ‘ Quarterly Journal of Geological Society,’ vol. xxx, pl. xxxi (1874). '' KIMMERIDGE CLAY. \o-. 59 § 9. Unwa.—The proximal extension of the articular cup (PI. XVII, fig. 18, ¢) upon an anconal or olecranal production marks this bone as strongly as in Varanus (fig. 15, a) but the excavation (¢) of the shaft below the proximal end is differently situated. It would seem as if the ulnar or outer border of that depression in Varanus (ib., fig. 15) had been moved or extended palmad, in Omosaurus, toward the narrower, palmar, surface of the bone; and to such an extent that part of this excavation comes into view from the ulnar side, as at ¢, fig. 14. This excavation is continued distad for more than half the length of the bone (c,c’). Below this part the shaft assumes a subtriedral form; and its anconal border bends toward that aspect as it approaches the carpus. The articular surface for this segment of the fore limb is wholly destroyed. § 10. Manus.—Of the carpal bones have been extracted a left scaphoid, left cunei- form, and left unciform. Of these three large wrist-bones the scaphoid is the smallest, as in Varanus, not the largest, as in Crocodilus, in which it is connate with the trapezium and trapezoides. The proximal surface for the radius is more uniformly and less boldly convex; the opposite articular surfaces for the trapezium and lunare is more deeply concave. The outer (ulnar) surface is elongate, narrow, and is the smallest on the bone; it seems barely to have touched the cuneiform, which is here, as in Varanus, the largest of the carpals. The free broader radial surface of the scaphoid is flattened and roughened, and seems to have continued, distad, the corresponding surface of the radius itself, which is on the radial side of the distal end of that antibrachial bone (Pl. XVII, fig. 8, g). The length (transverse extent) of the scaphoid is 5 inches; the extreme (ancono- palmar) breadth is 3 inches; the extreme proximo-distal extent (on the rough flat surface) is 1 inch 10 lines. The cuneiform is a massive cuboidal bone, with a proximal surface less concave for the ulna than in Varanus, but with as deep an opposite (distal) concavity for the division of the unciforme which supports the fourth digit. ‘There is an approach to the croco- dilian character of the bone in the increase of the distal part or surface. ‘The transverse extent of the bone there is 4 inches 9 lines; that of the proximal surface being A inches ; the ancono-palmar diameter of the bone is 3 inches 9 lines; the proximo- distal diameter is 3 inches 10 lines. The unciform seems, as in the Crocodile, to have supported both fourth and fifth metacarpals, not to have been divided to afford articulations for these bones on separate portions. Its transverse extent in Omosaurus is 6 inches 4 lines ; the other dimensions closely correspond with those of the cuneiform carpal. The digits of a hind foot are longer, as a general rule, than those of a fore foot in existing Saurian Reptiles, and the same proportion has been demonstrated in the fore and a ''60 FOSSIL REPTILIA OF THE hind feet of some extinct Dinosauria.’ The proportions, at least, of the metatarsals in Hylaosaurus and Scelidosaurus support a belief that those of the metacarpals would be as in the homologous bones of Zyuanodon. Of the five metapodial bones of Omosaurus which have been wrought clear out of the matrix not any show a length as compared with the breadth which exceeds that of the metacarpal of the first digit in the fore-foot of Zyuanodon (T. III, 1 m, Monogr. cit.); and the homologues of the intermediate metacarpals are shorter in proportion to their breadth than in Lyuanodon. I conclude, therefore, that the above metapodials of Omosaurus are metacarpals, that the digits were less unequal in length, and the whole fore-foot was more massive and elephantine in its proportions, in Omosaurus than in Lquanodon. A metacarpal (Pl. XVIII, figs. 3—6) has a flattened proximal surface (ib., fig. 5) of a subtriangular shape, slightly convex near its radial (r) and anconal (a) periphery, slightly concave toward the palmar border (p), which is broken away, the articular surface being continued a short way upon the ulnar (“) side of the shaft for junction with the second metacarpal. The articular surface is pitted with small deepish depressions, as in most great Saurians, where the joint surfaces seem to have been more syndesmosal than synovial. The transverse and ancono-thenal diameters of the proximal surface are equal, each being 3 inches 6 lines ; but, had the ulnar border been entire, the transverse cameter would have somewhat exceeded the other. The short thick shaft of this bone is three-sided ; one side extends obliquely from the ancono-ulnar (fig. 8, ax) angle to the radio-palmar (rp) angle, with a transverse convexity ; the second, or palmar, side (fig. 4, p) is less convex across; the third, or ulnar side, is flat across at the middle part, and somewhat concave near the two expanded ends of the bone. All these surfaces are concave lengthwise, the palmar one least so ; but the proximal half of this (fig. 4, p, p’) has been crushed. The distal articular expansion (fig. 6), almost flat transversely at its’ anconal part (a), begins to be concave at the middle of the distal surface (g), and this deepening to the palmar one (p) divides the joint there into a pair of convex trochlear condyles. The radial (7, fig. 6) of these, when entire, would have been the most prominent of the two. The metacarpal (Pl. XVIII, figs. 1 and 2) which supported the fourth digit has a proximal articular surface of a more definite triangular figure (Plate XIV, fig. 5); the anconal border (a) being the longest, and the angle between the radial (r) and ulnar (uw) borders being rounded off. The articular surface is continued upon both these sides of the shaft, but further for the articulation with the mid-metacarpal than for that with the fifth. * Iguanodon, Pal. Monogr., ‘ Wealden and Purbeck Reptilia,’ vol. for 1856, p. 1, tt. i, ii, iii (hind foot) ; Lguanodon, Pal. Monogr., ‘ Wealden Reptilia,’ vol. for 1871, p. 8, pl. iii (fore foot) ; Hyleosaurus, Pal. Monogr., ‘ Wealden Dinosauria,’ vol. for 1856, p. 18, t. xi (hind foot) ; Scelidosaurus, Pal. Monogr., Liassic Reptilia,’ for 1866, p. 17, tt. x, xi (hind foot). ''KIMMERIDGE CLAY. 61 The anconal surface (Pl. XVIII, fig. 1) of the shaft is almost flat and lies more on the plane of that surface of the entire metacarpus than in the marginal metacarpal above described (fig. 3). The radial and ulnar surfaces of fig. 1 converge palmad to the narrow convex palmar surface which forms the rounded angle of the proximal triangular tract (ib., fig. 6, uw, p). Both radial and ulnar surfaces of the shaft are concave lengthwise and across (ib., fig. 2, r). The transverse concavity of the distal articular surface is feebly indicated, and the bifid character of the joint is scarcely marked, though fractured surfaces suggest that a pair of low palmar prominences may have been broken away ; but the joint is much less trochlear than in the first metacarpal (ib., fig. 6). A metacarpal of similar type to the preceding has suffered too great mutilation of both ends to serve for profitable description ; it is not a corresponding metacarpal of the right fore-foot, but may be either a second or third, though from the slight superiority of length I should judge it to have been the second metacarpal of the same left fore-foot as the subjects of Pl. XVIII belonged to. A metacarpal with a subtriedral shaft, and an oblique twist at its basal half through an extension radiad of the radial angle, upon which angle the flat proximal articular surface has extended for the metacarpal on that side, is evidently a fifth metacarpal bone. The distal surface (Pl. XIV, fig. 6) is oblong and almost flat save where it becomes convex on being continued from the basal upon the radial surface; it is feebly concave trans- versely at its middle half, but this is not continued, deepening, so as to divide the palmar part of the jomt into a pair of trochlear condyles. ‘The length of this metacarpal is 5 inches 9 lines; the breadth of the proximal end is 4 inches ; of the distal end 3 inches 2 lines; the breadth of the middle of the shaft is 2 inches 3 lines. The largest of the proximal phalanges extracted gives a length of 5 inches 5 lines ; with a breadth of the proximal end of 4 inches, and a breadth of the distal end of 3 inches 7 lines. The breadth of the middle of the shaft is 3 inches ; and this seems not to have been more than 1 inch 7 lines in ancono-thenal diameter, but the thenal surface is partially crushed in. The anconal surface is smooth and flat save toward the expanded articular ends. The proximal surface, moderately concave, appears to have been adapted to a distal articular surface of the simple character of the metacarpal last described (Pl. XIV, fig. 6). The distal surface of the phalanx is moderately trochlear, 7.¢., with a feeble transverse concavity along its middle half; itis strongly convex throughout in the opposite (anconothenal) direction. 'The size of this proximal phalanx indicates it to have belonged to one of the larger middle digits. Of the instructive terminal phalanges, the most entire forms the subject of figs. 4 and 5 of Pl. XV. ‘The small proportion of the thin, smooth, punctate, articular surface shows a partial depression at 4, fig. 4; but the bone is so slightly abraded where that thin smooth crust is wanting as to afford a fairly true figure of its general shape, which is almost flat, with a feeble sinuosity. The anconal border (@) is most produced; conse- quently that surface of the phalanx is longest ; but it is little more than half as long as it ''62 FOSSIL REPTILIA OF THE is broad. The thenal surface is made concave lengthwise by the thenal production of the terminal lobes of the distal end (Pl. XV, fig. 5). ‘There is no appearance of these being articular. I regard them as the free termination of a last or ungual phalanx, and to show a modification of that end like the terminal phalanx of the second toe in /gwanodon (Monogr. cit., Pal. vol. for 1871, Pl. IIT, «3, 3). Not any of the fragments of phalanges suggested a structure for supporting a terminal claw, such as exists in Megalosaurus. The fore-foot of Omosaurus, as represented by the bones above described, was a short, broad, massive member, relating chiefly to progressive motion, and suggests the huge species, if not, like Zgywanodon, phytophagous, to have been a mixed feeder. § 11. Intum.—The mass of matrix with the portion of the skeleton of Omosaurus figured in Pl. XIX, reduced to one ninth of the natural size, includes, with the sacrum, both the iliac bones and a large portion of the right ischium. The left ischium and both pubic bones, one of which was almost entire (Pl. XX, figs. 4 and 5), were wrought out of the block in the course of exposing the rest of the pelvis upon which they were lying dislocated. The length of the ilium is 8 feet 5 inches; that of the antacetabular portion is 1 foot 9 inches; that of the postacetabular portion is 9 inches, but the end of this is broken off on both sides; the breadth of the superacetabular portion is 7 inches; the length of the acetabulum is 1 foot 1 inch; the breadth of ditto is 9 inches; the extent of the unwalled part of the cavity is 7 inches. Besides the pelvis and the detached vertebrae above noted the right femur and probably the shaft of the fibula were left in the mass in the relative positions exposed in Pl. XIX, in which the pelvis is seen from the heemal (ventral or lower) aspect. The ilium (ib., 62—62”) is an oblong, broad, and thick bone, anchylosed by a neu- romedial tract, two fect in length, to the expanded ends of the five sacral ribs Ob:, pl. 1). The heemal surface is divided into an acetabular tract (62), an antacetabular production (62’) of greater antero-posterior extent, and a shorter postacetabular production (62”). The lateral or external surface, or superacetabular tract, extends neurad and outward to terminate in a thick rugged convex border (r), which is continued forward, subsiding as a ridge upon the outer or neural surface of the antacetabular prolongation, (62’); the ridge is lost about nine inches from the fore-end of the antacetabular plate, and gives a triedral form to this part of the ischium. The ridge, continued from r, answers to that in the sacrum of the Zywanodon noted in Pal. vol. for 1854, ‘ Wealden Reptilia” Part I, p. 13.’ But the proportions of the antacetabular. and postacetabular productions are reversed in the Kimmeridgian as compared with the Wealden Dinosaur.’ 1 «The outer surface is divided into two facets by a strong longitudinal ridge, for the attachment of some of the powerful muscles of the hind limb.””—P. 13. 2 Compare Pl. IIL (Monogr. cit.), 62’ and 62”, with Pl. XIX of the present Monograph. ''KIMMERIDGE CLAY. 63 ‘ The length of the antacetabular part of the ilium in Scedidosaurus' more resembles that in Omosaurus, but it is narrower and extended more in the axis of the trunk, or is less inclined outward. The corresponding part of the ilium in Cetiosawrus resembles in breadth that of Omosaurus. In this the acetabular cavity (62) is thirteen inches in longi- tudinal, nine inches in transverse extent. Its outer and hinder border subsides at e, and the cavity is continued upon the superacetabular surface of 7, the break in the boundary being somewhat analogous to the cleft in the more developed border of the Mammalian acetabulum for the passage of vessels to the intra-acetabular synovial mass. The lower or hemal part of the cavity is completed by the ischium (ib., 63), which articulates syndesmotically with the surface (8 ¢). ‘There is no surface for the articulation of a pubis with the ilium, the Omosaurus in this respect corresponding with the Crocodilia. In the breadth also of the ilium as compared with the length that bone of Omosaurus comes nearer to the Crocodilian than to the Lacertian type. And, again, in the extent to which the ilium is prolonged in front of the acetabulum the Crocodiles? depart less from the Dinosaurs than do the Lizards. In Lacerta nilotica, e.g., the ilium is prolonged in front of the acetabulum to an extent equalling only that of the acetabular excavation of the same bone. § 12. Iscutum.—This bone (Pl. XIX, 63, and Pl. XX figs. 1—3) offers the structural type of that in Chelonia and certain Lacertilia (Uromastya, e. g., Pl. XX, figs. and 9, 63), in its ‘tuberosity ’ or posterior process (ec); but, in its slenderness or relation of breadth to length, it exceeds that in any Lacertian or other (to me) known forms of existing Reptile. Of the iliac articular end of the right ischium but little is exhibited, the bone (63, Pl. XIX) having been pressed forward and behind the part of the acetabulum from which it has been dislocated. The process (¢) answering to that so marked in Uromastyx, in the more perfect left ischium (Pl. XX, fig. 8), comes off nearer the articular end than in the Lizard. The rest of the bone is simply styliform and straight, having no process crossing, as in Birds, the obturator interspace between ischium and pubis. The smooth concavity on the under or hamal surface of the expanded end, articulating with the ilium, contributes about a fourth part of the cavity for the head of the femur. The end of the process (¢) is rough, thickened, of an elongate subtriedral form, 24 inches by 1 inch; the opposite or fore-end of the expansion has a rough syndesmotic surface for the attachment of a similarly roughened end of the pubis. The breadth of the ischium, including these processes, is 13 inches; from this part the bone quickly contracts to a narrow plate. ‘The hind margin of this plate (ib., fig. 1, e) is moderately thick and rounded, whence the bone thins off to an edge in front (ib.,/). The heemal surface is flat or feebly concave, transversely, and is smooth (Pl. XX, fig. 1). The upper or neural surface is, transversly, rather convex, save where it 1 Monog. cit., p. 15, pl. vi, fig. 1. 2 Cuvier, ‘Ossemens Fossiles,’ 4to, 1824, vol. v, pl. iv, fig. 15, a. ''64 FOSSIL REPTILIA OF THE extends upon the acetabular part (a,d), and here it is rather concave. The body of the bone gradually contracts to a breadth of 23 inches; it then slightly expands to its symphysial end (ib., gy, and fig. 3), which has a breadth of 4 inches, with a thickness of 2 inches. Restoring a part wanting between the preserved body of the ischium and the symphysial end, to the extent indicated by the dotted lines in Pl. XX, fig. 1, the total length of this pelvic bone in Omosaurus would be 2 feet 6 inches. § 13. Pusrs.—This bone (Pl. XX, figs. 4—7) presents the type of the pubis in Lacer- tians (ib., figs. 8 and 9) in the pectineal process (¢), and the perforation (d), but adheres to the Crocodilian type in presenting one articular surface only at the proximal end (@) for the ischium, and (seemingly) contributing no share to the acetabular cavity. A Chelonian character is shown in the length of the bone between the head (2) and the process (¢). The articular end (a) has been better preserved than the corresponding one of the left ischium (ib., fig. 1). It presents a narrow, elongate, synchondrosal, roughish facet, 6 inches in length, 1 inch 7 lines in breadth, with a moderate convexity in the long axis (ib., fig. 6). The posthumous abrasion of the articular surface checks an absolute state- ment as to the precise configuration of this ischio-pubic joint in the recent Omosaur, but the proportion, if any, contributed by the pubis to the acetabulum must have been very small, for no trace of such appears. The pubis as it recedes from this joint gradually narrows to a breadth of 3 inches 4 lines, then more rapidly expands to form the perforated pectineal plate (c). This plate or process becomes, as in Lizards and Tortoises, thickened and tuberous at its free prominent border, which describes a bold convexity before subsiding into the slender continuation of the pubis (ef. The margin of ¢ continued thereto by the dotted line, in figs. 4 and 5, is a fractured one; and the angle of the border (e) to which the dotted line is continued shows also fracture ; the extension of bone along that line is inferential- Proximad of such fracture the anterior border of the pubis is entire and sharp, a continua- tion of that which partly circumscribes the oblique pectineal hole or channel (4). From the pectineal expansion the pubis contracts to a breadth of 2 inches, then expands to its symphysial end (v), which, when entire, must have had a breadth of from 5 to 6 inches. The abraded surface (ib., fig. 7) gives a fuller ellipse than that of the ischium (ib., fig. 3), but, as in that bone, indicates a symphysial junction with the opposite pubis. The hind border of the pubis (/) is rounded and thicker than.the fore border (e). The neural surface (ib., fig. 5) is feebly canaliculate lengthwise in part of its extent, and this character is shown, though still more feebly, in the pubis of Uromastyz (fig. 9, 64). But the accentuation of this surface in the broader half of the pubis of Omosaurus, as shown in fig. 5, is due to crushing and fracture seemingly in relation to the original prominence of the part of the pectineal process (c, fig. 5), which has been pressed to flatness with slight concavity. I conclude from the length of both ischium and pubis that they diverged from each other, viz., from their outer to their inner or symphysial ends, at an angle nearer that in ''KIMMERIDGE CLAY. 65 ‘Crocodilians than in Lacertians. There is no evidence or indication that these heemapo- physes were disposed otherwise than in the rest of the Reptilian class, meeting, each pair, at the medial line, with a space between ischia and pubes, answering to a common and uninterrupted obturatorial vacuity. This space, in Dicynodon, is obliterated by continuous ossification. | The length of the pubis in Omosaurus is 3 feet 6 inches, the extreme breadth is 9 inches ; the least breadth of the pre-pectineal part (4) is 3 inches 6 lines; the extreme thickness of this part is 1 inch 3 lines. § 14. Femur.—To the right of the pelvis lies the femur of the same side, with the hinder surface exposed (Pl. XIX, 65). The head (a) of the bone is at a distance of 1 foot 8 inches from its socket (e) and a little posterior to it. The distal end lies exterior to and a few inches in advance of the right ilium. The terminal articular surfaces of the shaft are, to some extent, worn away, but sufficient remains to show that the chief convexity or head (a) projected some inches within the inner longitudinal border of the shaft, the proximal surface sloping slightly distad to the rough convex angle, representing a trochanter (4), from which a thick rough ridge is continued, gradually subsiding upon the shaft. The breadth of the proximal end of the bone is 1 foot 1 inch; at 1 foot distance from that end the shaft is contracted to a breadth of 8 inches, and at its middle part to one of 6 inches. Notwithstanding the posthumous pressure which has shattered this part of the crust of the femur, one may infer that the shaft was naturally subcompressed from before backward. At three fourths of the distance from the head of the bone the shaft again begins to expand, attaining at the distal end a breadth of 133 inches. ‘There is a distinct oblong protuberance (e) at the inner and back part of the shaft, 1 foot 6 inches beyond the head, corresponding to that more developed prominence which has received the name of ‘third trochanter’ in Jyuwanodon and Scelidosaurus. here is also evidence of a longitudinal ridge (d) continued from the back part of the trochanter, about 9 inches down the shaft, inclinmg toward the middle of the hinder surface. The popliteal cavity (e) is moderately concave, chiefly transversely through the backward production of the outer condyle (g). This is of less breadth posteriorly than the inner condyle (f) but is more convex as well as more prominent. ‘The outward extension of the femur (4) beyond this prominence is somewhat unusual. § 15. Trs1a.—This bone is represented by its proximal end and three fourths of the shaft (Pl. XXI, figs. 83—6). The shaft is more slender in proportion to the head than in Hylao-" ox Scelido-’ saurus, and yields a full subelliptic section (ib., fig. 6). Part of the } Pal. Monogr., ‘Wealden Reptilia,’ vol. for 1856 (Hyleosaurus), p. 17, pl. vii. ? Pal. Monogr., ‘ Liassic Dinosauria,’ vol. for 1861 (Scelidosaurus), p. 16, pl. x. ''66 FOSSIL REPTILIA OF THE articular surface for the ner femoral condyle may be recognised at g, and that for the outer condyle at 4, fig. 3, Pl. XXI. A procnemial plate (c), with a base of 7 inches in extent, projects forward 4 inches beyond the articular part of the head of the bone. As wrought out of the matrix this plate shows a sharper free border than probably was natural ; its obtusely rounded summit, d, has retained its condition as an epiphysis. The diameter of the head of the tibia in the direction of the procnemial prominence (4 ¢ fig. 5) is 11 inches. The preserved longitudinal extent of the tibia is 2 feet. The two: diameters of the fracture (4 fig. 3) are 4 inches 6 lines and 3 inches 6 lines. The indication of a medullary cavity at the fracture (f) are hardly so definite as in fig. 6, and such as it is, the cavity was short; for at the fracture (e) the corresponding central portion of the shaft shows an open osseous tissue with wide chondrosal interspaces. In the obliquely fractured and partly crushed end of the shaft the trace of medullary cavity has disappeared. The osseous tissue of the rest of the shaft is compact. Not- withstanding the degree of crushing, the beginning expansion in the tibio-fibular direction and of contraction or flattening in the rotulo-popliteal direction is unmistak- able, and has led me to conclude that the distal, more flattened end of the bone is that which is wanting in the present specimen. § 16. Dousrrun parts or Hinp Lims.—Exterior to the right femur and overlain by it is the shaft or slender part of a bone, 16 inches in length and 3 inches in breadth; Pre 11. it bears the proportion of a fibula to the tibia above described. No recognisable tarsal, or other bone of the hind-foot, has been detected in the indurated matrix forming the bed of the Omosaur-. But Professor Phillips, in his instructive ‘Geology of Oxford,’ states,' “Three metatarsals in the Oxford Museum, apparently of Megalosaurus, lying in their original apposition, have been obtained from the Kimmeridge Clay of Swindon and seem to indicate a tridactyle foot (diagram Ixviii).”’ I subjoin a copy of the cut of these bones (Fig. 11), deeming it more probable that they belonged to the genus of Dinosaur now known to have left remains in that formation and locality, than to the Megalosaurus, of which no indubitable evidence has yet been obtained from Kimmeridge Clay, either at Swindon or elsewhere. A is an outline of the proximal, B of the distal, ends. These bones exemplify the ‘leptopodal’ character of the Dinosaurian foot, due to the reduction of thickness or breadth by suppression of two of the toes, and a consequent departure from the short, thick, or broad ‘ pachypodal’ character of the pentadactyle hind foot of the existing and extinct terrestrial Cheionia. Metatarsals of Omosaurus ? tVPe 215: ''KIMMERIDGE CLAY. 67 § 17. Dermat Sprnz.—One osseous spine (Pl, XXI, figs. 1 and 2; Pl. XXII, figs. 2 and 3) has been successfully wrought out of the matrix; but though a close search was made for other evidences of a dermo-skeleton none have been found. The spine in question is 1 foot 63 inches in length, and not more of the tip seems to be wanting than might extend this dimension to 1 foot 7 inches, or, at most, 1 foot 8 inches ; the long diameter of its base (PI. XXII, fig. 2) is 5 inches ; the shaft gradually tapers to a point. The spine is rounded and slightly compressed ; the narrower diameter is shown in Plate XXI, fig. 1, the greater breadth in ib., fig. 2. The surface, smoothest toward the base, becomes slightly broken: by fine longitudinal, gvasi fibrous, markings ; and this sculpturing becomes coarser as the spine contracts. At every part may be seen small orifices, apparently vascular; few in number along the basal two thirds, but more frequent near the point. These indicate a periosteum in relation to the supply of a horny sheath, of which we have here the petrified bony core. The texture of the osseous substance is dense (Pl. XXII, fig. 3). The base is obliquely truncate, with a boldly sculptured border, broadly and deeply notched as if for strong ligamentous attachments, the whole basal surface being coarsely roughened ; it is also channelled, seemingly, by two vessels entering the substance of the spine, one, perhaps, an artery, the other a vein (PI. XXII, fig. 2). The spine is traversed by a central medullary or chondrosal canal, in diameter one third that of the smaller - diameter of the spine (ib., fig. 3). The rough imperforate part of the base, like its coarse periphery, suggests adaptation to syndesmotic junction with some other bone. But with what part of the frame? There is a want of symmetry at the obliquely truncate base, which suggests this spine to have been one of a pair. In Scelidosaurus the dermo-neural spines at the neck and fore-part of the back are similarly ‘somewhat unsymmetrical in form,’ showing a parial arrangement along that part of the trunk,’ but they are succeeded by symmetrical dermo-neural spines having a medial position along the rest of the trunk and tail.” The osscous spines, probably dermo-neural, of Hyleosaurus, show a length in propor- tion to the adjacent vertebral centrums somewhat exceeding the present spine of Omosaurus ; they are, likewise, obliquely truncate at the base, and unsymmetrical in shape, but in a greater degree ; and they are much more compressed.? In the Hylzosaurian specimen in the British Museum, which turned the scale in favour of the dermo-neural hypothesis, an irregular angular depression is described and figured at the base ; and this repeats, though single, the pair of depressions or canals above noted, and reputed vascular, in the base of the spine of Omosaurus. The low, obtuse, thick ridge 1 Pal. Monogr., ‘ Scelidosaurus,’ vol. for 1860, p. 25. 2 Ib., p. 22, pl. ix, figs. 1, 3, 5. 3 Pal. Monogr., ‘ Wealden Reptilia,’ vol. for 185 6, pp. 23—26, pl. 1; Dli ix, k ''68 FOSSIL REPTILIA OF THE girting the base of the spine in Aylzosaurus is, however, simple, unnotched; the provision for attachment of the spine, in Omosaurus, betokens a greater power of resistance against displacement. The superior strength of the spine, due to its full elliptical shape in transverse section, suggests its application as a weapon to be wielded for attack rather than as one of a merely defensive palisade of spines. Considering the number of vertebree—dorsal, sacral, caudal—which have been recovered in more or less completeness from the intractable mass of some tons weight, including the rest of the above described recovered parts of the skeleton of the Omosaur, it might reasonably be expected that, had the trunk and tail been defended by dermal spines, as in Scelidosaurus and Hyleosaurus, especially by spines similar in number and arrangement to the dermal ridged scutes in the more Crocodilian Dinosaur of the Lias, more evidences of such appendages to the trunk-skeleton should have been found in the grave of the great Kimmeridgian dragon. But we are, now, not limited to the head, the trunk, or the tail m quest of positions of armour afforded by dermal bones to extinct members of the Reptilian class. In the great Mantellian Zywanodon, or at least in the male of that species, a pair of spines supported by unsymmetrical conical bony cores were wielded for offensive action by the fore-limbs.' | The form and proportions of the Iguanodontal carpal spine, especially in its degree of compression, are more like those of the spine in Omosaurus than are any of the dorsal spines in Hy/@osaurus. ‘True, the conical spine-core in Jguanodon is shorter i proportion to its basal breadth than is the problematical spine in Omosaurus. It is significant of the nature of this one unsymmetrical osseous spine that the bones of one of the fore limbs, the left, and that limb only, should have been preserved, and ina more complete state than any other part or limb of the present remarkable Dinosaurian framework ; the spine in question lay not far from the radius and carpus. Two spines of similar form to that of Omosaurus, but of larger size, were discovered near each other in a pit of Kimmeridge clay at Wootton Bassett, Wiltshire, and formed part of the well-known collection of William Cunnington, Esq., F.G.S., now in the British Museum. Whatever contiguous bones may have been dug out of the same part of the pit were not preserved. These two spines form a pair, and resemble each other as much as would the right and left radius, or the right and left ulna, of the same Dinosaur. They differ from the (carpal?) spines of Omosaurus in having a sharp edge, which in a transverse section, like that of fig. 3, Pl. XXII, would terminate one end of the long diameter of the ellipse. The lethal power of the weapon was augmented by this character of the sword added to that of the pike. The degree of obliquity, the coarse marginal notching, and vascular perforations of the base, are as in Omosaurus ; but the expansion is greater, yielding dimensions of 8 inches and 6} inches in long and short diameters ; there is a slight submedial ridge dividing the basal articular surface into two | Pal. Monogr., ‘Iguanodon,’ vol. for 1871, ple ii, fig. 1, m, «. ''KIMMERIDGE CLAY. 69 shallow channels. The long diameter of the shaft, four inches beyond the least produced part of the base, is 33 inches, being nearly the same as in Omosaurus. The edge of the spine is along the same line as the most produced part of the base. The shaft has a central cavity, as in Omosaurus. Should these prove to be a pair of carpal spines they indicate a species of Dinosaur distinct from Omosaurus armatus, § 18. Lire anp Arrinirres or DINOSAURIA AS ELUCIDATED BY THE KNOWN CHARACTER or Omosaurus.—In the ‘Isis von Oken,? Band xxii, Heft v, 1830, 4to, p. 518, Hermann von Mzyzr proposed the following distribution of Fossil Saurians according to the structure of their hind-limbs :— “ Saurier mit Zehen, welche denen der lebenden am ersten noch entsprechen wurden, und gwar ““q, VIERZEHIGE. “ Rhacheosaurus, H. v. Meyer. “ Geosaurus, Cuvier (?). “Teleosaurus, Geoffroy (°). “ Aeolodon, H. v. Meyer. “ Streptospondylus, “ Metriorhynchus, “‘ Macrospondylus, “* Lepidosaurus, “ Mastodonsaurus, Jaeger (?). HT. v. Meyer (?). ‘‘ 6, FUNFZEHIGE. “ Protorosaurus, H. v. Meyer. “ Saurier mit flossenartigen Gliedmassen. “ Ichthyosaurus, Conybeare. “ Plesiosaurus, Conybeare. “ Mosasaurus, Conybeare. “ Phytosaurus, Jaeger (?). “ Saurocephalus, Harlan (?). ''70 FOSSIL REPTILIA OF THE “ Saurier mit Ghedmassen, ihnlich denen der schweren Landstugethiere. “ Megalosaurus, Buckland. “Tguanodon, Mantell. “ Saurier mit Flughaut. “ Pterodactylus, Cuvier.” The artificiality of these limb-characters has been pointed out, and accepted by the adoption, e.g., of the ordinal distinction of the Jchthyopterygia from the Sauro- pterygia;' also of the Labyrinthodontia, as represented by Mastodonsaurus and Phytosaurus, from the ‘ Vierzehige’ = Crocodilia, Ow.,° &c. Whether any apology be necessary for the substitution of the latter term for a defined ordinal group including half of the representatives of von Meyer’s “ (a) Vierzehige” I leave to the judgment of unbiassed palzontologists, and proceed to cite the more definite ascription of taxonomical value to the groups above defined proposed by von Meyer, in his useful compilation called ‘Paleologica,’ 8vo, 1832. In this work the author prefixes to the class Reptilien (p. 101), as to that of Mammalia (p. 44), his division of such classes into Orders. Those which he adopts for the ‘ Reptilien’ are— “a. Chelonier. B. Saurier. c. Batrachier. p. Ophidier.” This was the latest step in Paleontological ordinal classification with which I had to contrast the ideas of the Reptilian orders acquired during the researches of which the results were condensed in my ‘Reports to the British Association’ of 1840 and 1841. Von Meyer’s subdivision of the Saurian order is based, as in his previous sketch in the ‘Isrs,’ upon the structure of the limbs : «4. Saurier mit Zehen, ahnlich denen andern lebenden Sauriern und zwar I. Vier- zehige. II. Fiinfzehige.” 1 “ On the Orders of Fossil and Recent Reptilia.” From the ‘ Report of the British Association for the Advancement of Science’ for 1859, 8vo, p. 159. # Ap, p. 158. 8 Ib., p. 164; and “ Report on British Fossil Reptiles,” op. cit. for 1841, 8vo, p. 63. ''KIMMERIDGE CLAY. 71 “y. Saurier mit Gliedmassen ahnlich denen der schweren Landsaiigethiere. 1. Megalosaurus, Buckland. 2. Iguanodon, Mantell.” “c, Saurier mit flossartigen Gliedmassen. 1. Ichthyosaurus, Kénig. 2. Plesiosaurus, Conybeare. 3. Mosasaurus, Conybeare. Streptospondylus, H. v. M.” “p, Saurier mit Flughaut. Pterodactylus, Cuwer”’ (Op. cit., p. 201). In the characters of his subordinate group 3, Von Meyer (Ib., p. 210) condenses the descriptions and accepts the determinations, clavicle included, of Buckland and Mantell. There is no sign of his having examined any of the fossils on which these descriptions ‘and determinations were based. He is struck with a resemblance of the metapodial bones of Megalosaurus in Buckland’s plates with those of a hippopotamus; and with the size of one of these bones, “ zweimal so breit als im Elephanten” of the Zywanodon ; and may have deemed their feet, in like manner, to have been tetradactyle or pentadactyle. Such supposed character seems to have suggested to Von Meyer the name Pachypoda, which he subsequently applied to them, the proportions of the entire foot which would support such term being to him unknown. The feet of Dinosaurs are, in fact, characterised by their narrowness or slenderness rather than by their breadth or thickness. The functional toes (hind feet), are, in the typical species of Von Meyer’s Pachypoda reduced to three,’ and do not exceed four (Sceli- dosaurus, e.g.) in any veritable member of the order. But had Von Meyer known the structure of the Dinosaurian foot, and it had been such as to have been truly defined by his ‘family term,’ this term must have given way to the “ Pachypoda” proposed and accepted in 1821 for a similar group of Mollusca ; as the same term, proposed for a family of Coleoptera, in 1840, had, in like obedience to taxonomic rules, sunk to the condition of a synonym under the law of priority, even when not affected by inapplicability of the name to its objects.’ Every specimen accessible in 1840, of Megalosaur, Iguanodon, Hyleosaur, having been examined and compared by me and the structure of the sacrum elucidated by observations on its development in birds,’ vertebral characters, with dental ones, were substituted for those of the Family above cited from the ‘ Isis’ and ‘ Palzologica,’ in the definition of the Order Dinosauria, quoted by Professor Huxley in his paper on this group.* Of this defi- nition the Professor asserts that “every character which is here added to von Meyer’s diagnosis and description of his Pachypoda has failed to stand the test of critical inves- tigation.”°® This statement is not accompanied with any evidence in its support, but is suggestive that I had dealt unjustly with von Meyer in proposing the name and substi- | E.g. Hyleosaurus (‘Monogr. Wealden Reptilia, Pal. vol. for 1856, p. 18, pl. xi); Iguanodon (‘ Monogr. Wealden Reptilia,’ Pal. vol. for 1856, p. 1, pl. i). * Pachypus was given to a genus of Coleoptera in 1821 ; this, in like manner, reduced the Pachypus applied to a genus of mammals in 1839 to a synonym. > * Report on British Fossil Reptiles,’ p. 106, 1841. * “Quarterly Journal of the Geological Society,’ vol. xxvi, p. 32, 1870. ® W,; p38, ''72 FOSSIL REPTILIA OF THE tuting the alleged inaccurate characters of the reptilian group Dinosauria. If I have to. offer, in relation to the main end and aim of my labours, any remark which may seem critical, it will be accompanied by its grounds. Thus, in regard to the characters pro- posed by Professor Huxley for the Order Dinosauria— “1, The dorsal vertebrae have amphiccelous or opisthoccelous centra. They are pro- vided with capitular and tubercular transverse processes, the latter being much the: Jonger”’ (loc. cit., p. 33). | If by ‘amphiccelous’ be meant ‘biconcave,’ as the term ‘amphiccelian’ has been applied to dorsal vertebra of the Ichthyosaurus, no such vertebre exist in the dorsal region of Dinosauria. The term ‘amphiplatyan’ would more truly express the configura- tion of the terminal articular surfaces of the centrum in such dorsal vertebrae as are figured in Pls. XII and XIII of the present ‘ Monograph,’ and in corresponding vertebrae of Lyua- nodon Megalosaurus, Cetiosaurus, Hyleosaurus, Scelidosaurus, Bothriospondylus, figured in previous Monographs on British Fossil Dinosaurs. Not that the flatness of both ends of the centrum is absolute, but the deviation is slight and usually, when in the direction of concavity, confined to the hinder surface. Neither must it be supposed that the dorsal series may be ‘amphiccelous’ in one Dinosaur, or ‘ opisthoccelous’ in another. The centrum in some Dinosaurs, Zapinocephalus, e.g., shows at the middle of its flat articular surface a foramen one sixth the diameter of such surface. It is the base of a small conical cavity, the apex of which meets that of the cone of the opposite side,—a beaded remnant of the notochord appearing to have traversed the vertebral column. In other species examined by me certain cervical vertebrae and a few consecutive dorsal vertebrae are ‘ opisthoccelian,’ 7. e. have the ‘ball’ in front; and the convexity, in certain of these, does not wholly subside until the lumbar region is reached. But whence did Professor Huxley derive his knowledge of the ‘opisthoccelous’ character in ‘ pachypodal Saurians’? If from the original definition of the Dinosaurian group,’ that character, as there limited, seems to have stood the test of time. The discoverers of the Jyuanodon and Megalosaurus believed the ball to be behind, and von Meyer accepted this view of the conformity of the Dinosaurian with the Croco- dilian dorsal centrums. In fact, the way to distinguish the fore from the hind end of a fossil saurian vertebra seems not to have been known to their describers until the test was defined in 1841. ‘This knowledge, howsoever acquired by the writer of the “ Cha- racter 1,” here discussed, is applied by him in error to Dinosauria: in them the ball subsides at the beginning of the dorsal series.” I would further remark, that, as there are: many modifications and characteristics of the so-called ‘ capitular transverse processes’ and ‘tubercular transverse processes,’ in the varied series, including Dinosaurian, of vertebral 1 © Report on Brit. Foss. Reptiles,’ p. 91: ‘‘ Remarks on Mantell’s ‘ Fourth System’ of Vertebree from. the Wealden.” 2 Ib., ib. ''KIMMERIDGE CLAY. 73 structures, the advantage of single substantive terms is exemplified by the convenience’ and helpfulness to precise description which such terms afford, adjectively, in predicating -of ‘ parapophysial’ and ‘ diapophysial’ modifications. And if by ‘capitular portion of the transverse process’ Professor Huxley may mean ‘ parapophysis,’ and by ‘tubercular portion of the transverse process’ ‘ diapophysis,’ ted have then to object that the ‘dorsal vertebrae’ of Omosaurus do not all possess the two kinds of processes. In the subjects of Pls. XIII and XIV the head of the rib is received by a pit, not articulated to a ‘ capitular process.’ ‘The dorsal vertebrae, of which the ribs have not ‘distinct capitula and tubercula,’ have no ‘capitular portions, or transverse pro- cesses,” 7. €., no parapophyses. In reference to Professor Huxley’s “Character No. 2,” I submit that a Saurian with ‘sacral vertebrae reduced to two in number is not a Dinosaurian. «3. The chevron bones are attached intervertebrally and their rami are united at their vertebral ends by a bar of bone.”* ‘This is a character of Zywanodon* and of Scelidosaurus,* but not of Cetiosaurus* nor of Omosaurus,® “Char. 3” is one of a genus, not of the Order Dinosauria. “5, The skull is modelled upon the Lacertian, not on the Crocodilian type.’ For the instances in which the Dinosaurian skull departs from the Lacertian, and approxi- mates to the Crocodilian type, I refer to the Monograph on Scelidosaurus,’ and to that on Jguanodon.® These instances confirm and add to the combination of Croco- dilian with Lacertian characters, propounded, in 1841, as exemplifying the more generalised Saurian type of the extinct order Dinosauria. “6. The teeth are not anchylosed to the jaws, and may be lodged in distinct sockets.” The modifications of the dental system in Dénosauria concur with those of the skull and jaws themselves in exemplifying the mixed or more generalised character of the group.” “7, There is no clavicle.” This is probable from the crocodilian affinities shown in the skull and vertebree ; and the character founded on the bone, so called, in my diagnosis -of Dinosauria, must be suppressed : but I have not yet seen a specimen of a Dinosaur in which the scapular arch was shown in its natural condition and integrity. Before continuing my remarks on some of the Professor’s remaining twelve characters of Dinosauria, 1 would observe, in reference to comments upon the step taken of substituting 1 * Quarterly Journal of the Geological Society,’ vol. xxxi, p. 426. + Tid, vel. xxv, i od. 3 Monogr. ‘ Wealden Reptilia,’ part ii, Pal. vol. for 1854, p. 15, t. viii. (Zguanodon Mantelli) ib. ib., t. i, Iguanodon Fowii (if this be not an immature specimen). * Monogr. ‘ Fossil Dinosaur of the Lias,’ Pal. vol. for 1860, p. 8, t. vil. 5 Phillips, ‘Geol. of Oxford,’ p. 259, fig. 2, 8vo, 1871. © Alle, @. 30, pl: Xvi, 7 Pal. vol. for 1859, pp. 7—I4, pls. iv, v, vi. 8 Pal. vol. for 1873, ‘ Iguanodon,’ pp. 4—12, pl. i, fig. 9; pl. ii, figs. 1—15. 9 *Odontography, pp. 246—254, 269—272, pls. 62a, 70, 70a, 1840. ''7A FOSSIL REPTILIA OF THE that name of the Order for one of a Family which, for reasons above given, could not have stood in Taxonomy, that the further insight into the structure of /ammalia tersely expressed in the names and characters of the Orders in the ‘ Régne Animal’ was gratefully accepted by all single-minded cultivators of Biology, although some of such orders were the same or nearly the same as those defined and otherwise named in the ‘ Systema Nature.’ Cuvier was not deterred from fixing this additional step in the advance of Zoology by the opportunity it might open to an objector for charging him with unfairness or injustice to Linnzeus; nor was Linnzus much moved by like remarks to which he was subjected by the critics of that era in reference to his names for groups of plants more or less similarly defined, before him, by John Ray, and others. To return, however, to my proper task, more especially in reference to the affinities of the Dinosauria. The first clue to the homology of the supposed clavicular bone of the Iguanodon * was given by Professor Lerpy in the ‘ Proceedings of the Academy of Natural Sciences of Philadelphia,’ December 14th, 1858. In the description there given of the fossil remains of a Reptile, which he calls ‘ Hadrosaurus, from the marl of New Jersey, which marl, from the affinity of this Reptile to the Iguanodon, he surmises may be of the Wealden or Green-sand period, Leidy finds, with the ilium, ‘‘a bone which I suspect to be the pubic, but which appears to correspond with that of the Maidstone Jgwanodon, described as the clavicle’ (p. 9). In a subsequent illustrated Monograph,’ Leidy repeats. his homology of the bone in question and notes—‘ an ilium and a supposed pubic bone, imperfect” (p. 71). Of the latter a figure is given (“ Pl. VIII, fig. 13”), and the accomplished Author truly remarks :—“ It bears a general resemblance to that indicated by Professor Owen and Dr. Mantell as the clavicle of the Iguanodon; but appears to me- rather to resemble the pubic bone of the Jywana and Cyclura than the clavicle of the same animals.” * Professor E. D. Cope, Corr. Sec. Academy of the Nat. Sciences, Philadelphia, commu- nicated to the Academy, in 1867, a paper “On the Extinct Reptiles which approached the Birds,” of which an ‘ Abstract’ was given in the ‘ Proceedings of the Academy ’ for December of that year. In this ‘Abstract’ the Professor is reported as stating that “ he was satisfied that the so-called clavicles of Zywanodon and other Dinosauria were pubes, having a position similar to those of Crocodilia.”* There is no reference, therein, to Professor Leidy, nor to the paper by Professor Huxley “ On the Classification of Birds” which was published in the ‘ Proceedings of the Zoological Society,’ 1867, p. 415.° 1 ¢Philos. Trans.,’ p. 138, 1841. 2 «Cretaceous Reptiles of the United States,’ p. 97, pl. viii, fig. 13: in the ‘Smithsonian Contribu-- tions to Knowledge,’ No. 192, vol. xiv, 4to, 1865. © Op, cit., p, 97. 4 «Proceedings of the Academy of Natural Sciences of Philadelphia,’ p. 234, 8vo, 1867. 5 See “Note,” p. 24, in ‘ Quarterly Journal of the Geological Society of London,’ vol. xxvi (1870). ''KIMMERIDGE CLAY. 7 In the lecture “ On the Animals which are most nearly Intermediate between Birds and Reptiles,” delivered by Professor Huxley at the Royal Institution of Great Britain, 7th February, 1868, he states:—‘‘I hold it to be certain that these bones — the so- called ‘clavicles "—belong to the pelvis and not to the shoulder-girdle, and I think it probable that they are ischia; but I do not deny that they may be pubes.” Thanks to the rapidity by which, through science, sea and land can now be traversed, we get the results of research by our American fellow-labourers within a fortnight, usually, after publication. I have no doubt of the legitimacy of Professor Huxley’s assertion—“TI could not possibly have known anything about them when my ‘Lecture’ was delivered ;” but the originality of his views of problematical pelvic bones by no means called for any reflection on postal arrangements between Great Britain and the United States. The impossibility might merely mean an oversight which left the writer ignorant of both ‘Cope’s and Leidy’s anticipations, as appears to have been the case with regard to von Meyer’s paper in the ‘Isis’ of 1830. In the “Further Evidence of the Affinity between the Dinosaurian Reptiles and Birds,” with confirmatory testimony by Professor Phillips, of Oxford,’ Professor Huxley ‘adopts the ischial homology of the bone in question, and illustrates it by a diagram, “Fig. 3, Dinosaur,” p. 27 (tom. cit.), in which the supposed “ ischium”’ is directed from the acetabulum downward and backward, parallel with the pubis, with which it articu- lates by the process (c, figs. 4 and 5, in Plate XX, “ Omosaurus”’), so as to “ interrupt the obturator space,” and define, as in Birds, an anterior part of that space as an “ obturator foramen ”’ (loc. cit.). ‘lo an advocate of the affinity of Dinosaurs to Birds and of the derivation of Birds from Dinosaurs, such determination of the bone in question gave great help, and the consequent diagram has been mainly subservient in gaining suffrages to the idea—I may term it sensational—of the kinship of the Iguanodon with the Cassowary, carried to the inference of a common bipedal mode of progression. The value of the genus Omosawrus, as of every well-determined new Dinosaur, to the Palaontologist desirous, irrespective of foregone conclusions, to lay the basis of lasting views of affinity on fixed homologies, is here great. The bone, Pl. XIX, 63, which com- pletes the acetabulum, shows by the extent and position of its articulation with the ilium, ) from which it has been but slightly dislocated, that it is the ischium. The recovery of the parial bone to the extent shown in Pl. XX, fig. 1, shows that the shaft gives off no process ; also that an extension of the iliac articular end beyond the acetabular surface of the ischium, and behind it, is the sole production, transverse to the axis of the bone, which can be homologised with a non-articular process in the ischia of other Vertebrates. 1 Erroneously so called in my ‘ History of British Fossil Reptiles,’ part v, p. 265, 4to, 1851. 2 «Quarterly Journal,’ &c., tom. cit., p12 L ''716 FOSSIL REPTILIA OF THE The ischia of Omosaurus being thus determined, the homology of the other pair of pelvic bones (PI. XX, figs. 4 and 5), wrought out of the mass of matrix overlying the heemal surface of the sacrum and ilia, was plain. They confirm the opinion of Professor Dinosaur. Crocodile. Dinornis. Pelvic characters. Leidy as to the nature of the bone; and, so far as their dislocated condition indicated their natural direction, it supports the conclusion of Professor Cope that they had “a position similar to those in the Crocodilia,” i. e., directed forward and downward, as shown by Cuvier, in the ‘ Ossemens Fossiles,’ tome v (1824), Pl. IV, fig. 15, a, Pl. V, fig. 6, and as exemplified in my diagram, p. 76, Fig. 13. So much of the homological ground being thus cleared, we may pass to the question of the affinities it brings into view. In birds, as a rule, the pubis is a long simple style without process (fig. 15, ‘ Bird ’) ; the exceptions are chiefly seen in the wingless forms, dpteryz, ©. g., and the Cassowary, in which latter bird the expanded acetabular end of the pubis projects forward beyond ''KIMMERIDGE CLAY. 77 the joint in a pointed form, about six lines in length. The proximal end of the pubis enters into the formation of the acetabulum in all birds. The distal end terminates, in most birds, freely ; in some it is anchylosed to the ischium ; in the Ostrich it joins its fellow to form a symphysis pubis :’ in all it is directed backward and downward. Monitor. Ornithorhynchus. Echidna. Modifications of pubis. The line A A traverses the corresponding part of the bone. In the Monotremes the pubis (fig. 15, Ornithorhynchus) sends off from its fore part, about one third of its length from the acetabular end, a, a low and broad process, c, giving attachment to the outer part of the base of the marsupial bone. It joins its fellow at the expanded distal end, s, and joins at /, the corresponding end of the ischium, thus dividing the obturator interspace into a pair of foramina. As in all mammals the bone is directed downward (hzmad) and a little backward. ' For other modifications, which, however, give no help in the present inquiry, see my ‘ Anatomy of Vertebrates,’ vol. ii, pp. 35, 36. ''78 FOSSIL REPTILIA OF THE In Crocodilia the pubis (fig. 13, p 4), as in birds, is a simple style slightly expanded distally where it articulates with a cartilaginous abdominal sternum,' but it joins not there, directly, either its fellow orthe ischium. It contributes no part to the acetabulum, but is attached at its proximal end to an anteriorly produced part of the same end of the ischium (ib. is). In Chelonia the pubis is remarkable for its breadth, due to its distal expansion ;. proximally it contributes to the acetabulum, articulating there with both illum and ischium, and at or near half way to the distal end, it sends forward a broad and termi- nally thick pectineal process ;? it unites distally with its fellow, and in some species also, as in Monotremes, with the ischium, dividing the obturator space. The average pro- portions and common character of the pubis in Lacertilia are given in Cut, fig. 15, Monitor ; the perforation d marks the closer resemblance to the Dinosaurian pubis. (fig. 12, po). Notwithstanding the difference in the proportions of breadth and length, the pubis in Iguanodon and Omosaurus, in its essential characters, is more like that in the Tortoise than in any bird. But these proportions are among the most variable characters of the bone, and we have not far to seek in the Lacertian order before finding, as in Uromastye, a pubis combining with the pectineal process (Pl. XX, figs. 8 and 9, 4), as slender a body thence continued as in the Dinosauria. Only, in Omosaurus, the proximal end of the bone seems not to contribute any share to the acetabular cavity; and, if this should be the case with other Dinosaurs, those extinct reptiles would combine, in their pelvis, as in some other parts of their skeleton, characters now restricted respectively to the Croco- dilian and Lacertian groups of the class. Thus, the ischium, in Omosaurus, has no other ‘ process’ save the stunted homologue of the proximal extension supporting the pubis in Crocodilia. In Chelonia, as in Uromastyx, there is a distinct posterior process (marked ¢ in figs 8 and 9, Pl. XX); but in certain Lizards (Varanus niloticus, e. g.)* it isreduced to a mere rudiment, and in the Chameleon it ceases to exist. Thus, the Omosaurus resembles the Crocodilia and some Lacertilia in the simplicity of its ischium, and markedly departs from the type of birds in respect to this bone. But it is alleged that the ilium gives evidence of the avian affinity of Dinosaurs which we have now proved to be wanting in the rest of the pelvis. Among the “ points of difference between any existing Reptile and any existing Bird,” the following is put by Professor Huxley in the foreground. ‘1. In the Reptile the ilium is not prolonged in front of the acetabulum.” “In the bird the ilium is greatly prolonged in front of the acetabulum.” 1 «Anat. of Vertebrates,’ vol. i, p. 68, fig. 56, 5. 2 «Anat. of Vertebrates,’ vol. i, fig. 116, 4. 8 Cuvier, ‘ Ossemens Fossiles,’ tom. cit., pl. xvii, fig. 40, c. ''KIMMERIDGE CLAY. 79 “ Now, in all the Dinosauria which I have yet examined, the ilium extends far in front of the acetabulum.” To the first of these averments it needs only an elementary acquaintance with compa- rative osteology to reply, that in all Crocodilian Reptiles the ilium is prolonged in front of the acetabulum, and to an extent nearly equal to that in which it is produced behind the acetabulum. Reference to the well-known figure in the ‘Ossemens Fossiles,’ which I here reproduce (woodcut, fig. 18, i2) exemplifies this fact : Cuvier has been careful to mark with the letter ‘a’ the antacetabular part of the ilium which the advocate of the avian affinities and bipedal progression of the Dinosauria denies to it and to all other Reptiles, Dinosauria excepted. | The true characteristic of the ilium in Dénosauria is the distinction of the super- acetabular (P]. XIX, r) from the antacetabular (ib., 62’) parts of the bone, with the anterior extension and subsidence, in some species, of the former upon the dorsal surface of the latter. As to the proportions of the ant- and post-acetabular extensions of the ilium, they vary in known Dinosauria: the post-acetabular production (Pl. XIX, 62”) is shorter in Omosaurus than in Scelidosaurus, and is shorter in Scelidosawrus than in [guanodon. From the importance assigned by Professor Huxley to iliac characters, in the con- clusion he advocates, a non-anatomical reader might infer not only that no other Reptiles, but that no other warm-blooded Vertebrates save Birds, had the ilium extended, as in Dinosaurs, far in front of the acetabulum. And yet an impartial quest of the affinities of these huge terrestrial /eptilia would impel the seeker, having such end solely in view, so to extend his comparisons. In Mammals “the ilium is prolonged in front of the acetabulum,” which, as in Reptiles, “ is either wholly closed by bone or presents a fontanelle.” In the spiny Monostremes (woodcut, fig. 15, Hchidna) the ilium (g) extends far in front of the acetabulum (4), and furnishes only an arched roof of that cavity, the inner wall of which (¢) remains membranous, as in the Bird. The pubis (a), after extending hamad (forward or downward) to the pectineal process (c), bends there to be continued back- ward, as in Oruithorhynchus. As arule all Mammals resemble Birds in a backward extension of more or less of both pubis and ischium, from their iliac articulations. Thus the character asserted to be peculiar to Dinosauria among Reptiles exists in the Crocodilian order of that cold-blooded class ; and, amongst warm-blooded Vertebrates, it is common to Mammals with Birds. In my ‘ Anatomy of Vertebrates’ I remarked, ‘‘ the transference of the weight of a horizontal trunk upon a single pair of legs necessitates an extensive grasp of the trunk- segments. When the legs require to be pulled far and strongly back, as in diving and cursorial motions, the origins of the requisite muscles are extended far behind the limb’s ’ 1 “Quarterly Journ. Geol. Soc.,’ vol. xxvi, 1870, p. 26. ''80 FOSSIL. REPTILIA OF THE centre of motion, as in the pelvis of Grebes, Loons, Ostriches, and Emus. When the bird slowly stalks, or hops, or climbs, or uses its legs chiefly in grasping and perching, the pelvis is short and broad, especially behind ; its breadth may even exceed its length, as in Cyclarius guanensis.” * The antacetabular part of the ilium in Birds is usually the longest, but its outer surface is not divided or interrupted by the super-acetabular plate and ridge peculiar to Dinosaurs. ‘To the degree in which the pelvis is produced behind the acetabulum (as in woodcut, Fig. 14, 2), such production helps to transmit the weight of the body upon the legs in a relative position thereto more favorable to the support of such weight ; if the pubis were directed forward instead of backward, it would detract from this relation of the pelvis to bipedal progression. Nevertheless, the balance of the parts so carried in the Bird prepon- derates forward; the weight of the body with the head and fore-limbs is greatest in advance of the acetabula. Among the modifications which are associated with the backwardly produced ilia, ischia, and pubes, in relation to the terrestrial progression peculiar to Birds, may first be noted the great extent of the axial trunk-bones welded into one mass where they are grasped by the bones transferring such mass upon the heads of the femora. In no Birds are the sacral vertebree so few as in Dinosauria ; and in those Birds which, from their size and terrestrial habits, are cited to exemplify Dinosaurian affinities, and which best lend themselves to test the question of the locomotion of the great extinct Reptiles, the number of the sacral vertebra is from 18 to 20. The several species of Dinornis had from 17 to 20 sacrals ; 12 is the average number in Watatores, 12 in Gralle and Gallinacea, 11 in Altrices. The highest number of sacral vertebra yet found in Dinosauria is 5: in Dicynodontia it is 6. The Sloths have 6 (Ai) or 8 (Unau) sacral vertebrae. The extinct Megatherioids, from the great share taken by the massive hind limbs in supporting the body while the fore limbs were engaged in disbranching trees, have a correspondingly closer resemblance to Birds in the structure and proportions of their pelvis than any known extinct Reptiles present. The Mylodon had not fewer than 11 anchylosed sacral vertebrae.’ In Birds, the trunk, properly so called, as distinguished from the neck, is singularly short ; its production in advance of the pelvis is reduced to the utmost, consistently with its visceral relations. The number of vertebrae between the neck and pelvis, ¢.¢. of such as bear pairs of moveable ribs, averages 8, and never exceeds 10; and of these anchylosis commonly fetters the major part. Between such vertebrae and the skull the ‘cervicals’ are as exceptional in excess, 1 «Anatomy of Vertebrates,’ vol. ii, p. 37. * They may in an exceptional instance extend to 6, but demonstrative evidence of this excess has not come to my knowledge. 5 «Description of the Skeleton of an Extinct Gigantic Sloth, &c., p. 64, pls. i, x, 4to, 1842. '' ee ett A RY f oe K eASITY J j & UIs (iN KIMMERIDGE CLA on cows arm = numerically ; and this concurs with the exceptional reduction of number in the ‘ dorsals ;’ both being in special physiological relation to bipedal support and progression. The numerous cervicals have peculiar joints, governing the sigmoid flexure and oscillating sway of the long and slender neck; whereby, in walking, both neck and head, in Birds, may be brought more directly over the supporting columns of the hind limbs as these change their position. ‘These limbs, moreover, have their specialties in relation to 81 their peculiar work in the vertebrate series. The femur (Fig. 14 (Dinornis), f) is relatively short; the tibia (¢) relatively long; the fibula (ya), styliform and short, takes no share in the ankle-joint, but co-operates with the tibia in a special manner to extend and strengthen the articulation of the leg with the thigh. The femoral condyles are concomitantly modified to effect the accessory femoro-fibular joint. Nothing of this exists in Dinosaurian or other Reptiles. Still more special is the modification in Birds by which the leg is united with the foot. No break in the column charged with the sustaining function peculiar thereto in the Bird is allowed beyond the absolute necessities of bending movements of such column when subserving loco- motion. The tarsal segment is suppressed; the metatarsal segment (mé) is aggrandised, lengthened out and confluently compacted; the metatarsals of three toes are welded into one bone. ~The joint of the leg with this bone is closely and tenaciously trochlear, strictly limiting the movements of the foot to one plane. The long and slender phalanges stretch forward at right angles to the metatarsus, and diverge to form a suitable base for the columns to which has been assigned such an unique task—so peculiar a work—as is performed by the hind limbs on the feathered class. Certain Dinosaurs wielded carpal spines and some Mammals bore tarsal ones. It would be as germain on that ground to derive Chauna or Palamedea from Iguanodon or Omosaurus, as Platypus from Phasianus. What are the known structures in Jegalosaurus, Iguanodon, and other Dinosauria, which, corresponding with those in Birds, would justify the conclusion or suspicion that the ischium and pubis, besides being long and slender, as they are demonstrated to be in Omosaurus, were directed from their acetabular ends backward parallel to one another ? Tt is certain that the ischium in Jguanodon had not the ‘obturator’ process charac- teristic of the same bone in Birds, and as certain that there must be a mistake about the matter when the same is predicated of the pelvic bone, erroneously called ischium, m the immature or small kind of Iguanodon which has been termed ‘ Hypsilophodon’ in ignorance of the true structure of the mandibular teeth. That the pelvic bones, truly homologous with ischia, were “united in a median ventral symphysis,’* is most probable from the shape and surface of the somewhat ' Huxley, ‘ Quarterly Journal Geol. Soc.,’ vol. xxvi. ''82 FOSSIL REPTILIA OF THE expanded distal extremities of the unquestionable ischia in Omosaurus. But such union does not exist in Birds. If it should be found in all Dizosauria, it is one of the majority of characters in which that order differs from the class of Birds and agrees with its own class, viz. the Reptiles. Of the comparatively few sacral vertebrae in Dinosauria the ‘ costal portions of the trans- verse processes’ (pleurapophyses) abut chiefly against the part of the ilium contributing to the cup to be upborne by the thigh-bone; there are no postacetabular abutments against other parts of the ilia, or against the comparatively broad ischia, as in Birds. In the latter pelvic character we have again to quit the Reptilian class and to indicate the repetition of it in certain bird-like Lissencephalous Mammals.’ The augmentation of number of sacral vertebrae beyond that—two—in Crocodiles and Lizards, whose bellies trail upon the ground or are but little raised therefrom by the out- sprawling fore and hind limbs in running along, relates in Land-tortoises to a more vertical position of the leg, and to the greater weight which the entire hind limb has to sustain in the progression of those Reptiles. In Dinosaurs (woodcut, Fig. 12) the thigh ( /), as well as the leg (¢), were probably less obliquely disposed, in quadrupedal locomotion, than in any existing Reptiles, save, perhaps, the Chameleons. ‘The four or five sacrals, interlocked, as in Birds and Tortoises, by alternating centrums and neural arches, have been recognised as physiologically related to correspondingly developed hind-limbs and a concomitant carriage of their huge elongate trunk, in a way approaching to that in the large gravigrade Mammals.’ It is requisite, in the present test, to determine as nearly as may be the relative length of the pre-pelvic part of the trunk to the pelvis in Dinosaurs. It may be presumed that those who represent the pubic-ischial elements of such pelvis, as being disposed in the avian fashion, intend the inference that, so far, the pelvis of the Dinosaurs related to the same bipedal mode of progression as in Birds, and that the trunk was similarly borne along, prone, upon the single pair of hind- legs.° : If, however, our knowledge of the dinosaurian pelvis being rectified, it should be averred that the trunk of the Iguanodon or Megalosaur might be otherwise carried than in Birds, that it was reared upright and so balanced, as in Man, upon a pair of hind, or in that case lower limbs, it may then be necessary to enter upon a series of comparisons between the dinosaurian and human skeletons mm connection with such upright mode of progression. 1 * Anat. of Vertebrates,’ ii, pp. 397—402, figs. 263, 264, 266—268. 2 ¢Report on Brit. Foss. Reptiles, 1841. 3 « Not a ground-crawler, like the alligator, but moving with free steps chiefly, if not solely, on the hind limbs, and claiming a curious analogy, if not some degree of affinity, with the ostrich.” Phillips, ‘ Geology of Oxford,’ p. 196. Such an idea, if it ever ‘suggested itself’ to my mind, was never expressed, and must have been instantly dismissed through considerations akin to those detailed in the text. ''Rees? KIMMERIDGE CLAY. 83 At present I shall not spend time in analysing the grounds of such view ; but, return- ing to the avian comparison, I may remark that the number of free vertebrae between the sacrum and skull, in Jgwanodon, is 24, of which 7 are cervical, 17 dorso-lumbar; in Megalosaurus present evidence supports an estimate of 23 such free vertebra allowing 7 to the neck ; in the parts of the skeleton of the same individual Hy/eosaurus, in the British Museum, 10 vertebr in natural succession include the hinder cervicals and succeeding dorsals, but the more or less complete vertebrae scattered in the same mass of matrix support an estimate of the vertebral formula not less in number than in Iguanodon ; whilst, as such vertebrae are shorter in proportion to their breadth than in either Tguanodon or Megalo- saurus, there may have been more than 24 between the skull and sacrum. In Sceldo- saurus 16 dorso-lumbar vertebrae are shown in succession in the blocks of lias in which they have been exposed, and 6 at least, if not 7 cervicals, are also evidenced in the same instructive skeleton of one individual Dinosaur.’ The proportion of the skeleton of Cetiosaurus longus in the Oxford Museum and that of the allied Dinosaur (Omosaurus armatus) in the British Museum demonstrate the absence of anchylosis in the dorso-lumbar region of the spine, and of any of the modifications of the hindmost vertebrae which, in Birds, add to the mechanical bracing of the trunk upon the pelvis: they show no lengthened pleurapophyses, having free proximal articulations to anterior sacral vertebrae; but, on the contrary, as in Mammalian quadrupeds, the lumbar ribs are short, coalesced with their vertebra, and project as straight out- standing transverse processes, not opposing the lateral movements of the trunk upon the pelvis, but, with the antecedent vertebra, negativing the notion of any action of muscles, proceeding from the pelvis and thigh-bones to grasp fast a trunk, and uplift it, together with the fore-limbs, neck, and head, clear of the ground, as during the hypothetical bipedal march and course of the huge dinosaurian Reptiles. The ascertained conformity of organisation in known Dinosauria supports the conclusion that a long, bulky, bendible body stretched forward from the pelvis and-hind limbs throughout the order. In Birds the bony ‘ vertebral’ and ‘sternal’ ribs of the few vertebrae of their short dorsal region are spliced together by a mechanism of which no trace has hitherto been discovered in the corresponding more lengthened region of the spine of Dinosauria ; there is a like absence, in these cold-blooded vertebrates, of the anchylosis of centrums, and of ossified tendons or neurapophysial splints—avian structures—which limit, to the essential minimum, any movement between one prepelvic vertebra and another. Every modifica- tion of the Bird’s skeleton concurs to facilitate the carriage of the prone trunk, as one compacted mass, upon the vertical pair of limbs, and not one of these modifications exists in Reptiles recent or extinct. What, then, we next ask, were the arrangements in the neck to diminish the 1 Paleeont. vol. for 1860 (Scelidosaurus), p. 11, tab. i—vi. ''84 FOSSIL REPTILIA OF THE difficulty which the known structure and proportions of the trunk oppose to the bipedal progression of Dinosauria ? Nothing of such exists in the length of the neck, nothing in the number or in the freedom of flexibility in opposite directions of the cervical vertebra ; on the contrary, those vertebrae in Dinosauria which are anterior to the bearers of the long and free ribs are few in number, with the little flexibility allowed by their reciprocal joints checked by the disposition of their short and mostly imbricate ribs. The neck of the Dinosaur was short, straight or nearly so, and strengthened by the overlapping pleurapophyses for the Fig. 16. “ Sp SS = ae Sno Ge v1 = Ga ' Ss dt Mil a = 5 5 — i ih Scelidosaurus. Varanus. Young Dinornis. Ruminant. carriage of a massive head projecting forward almost in a line with the body: never could such head be carried back, by a graceful sigmoid bend of a long neck, so as to be poised above the centre of support afforded exclusively by a hind pair of limbs. Such head, with its powerful jaws and their dense and weighty dental armature, needed the development and structure of a pair of fore-limbs, to sustain it with the fore part of the trunk, and take the required share in bearing along the bulky dinosaurian quad- ruped. Omosaurus adds a pregnant instance of the requisite anterior pair of supports. What the Dinosaur needed for its mode of terrestrial locomotion the Bird has not; and what the Bird possesses for its mode of terrestrial locomotion the land Reptile is devoid of. I have alluded to the modifications, extreme and beautiful they are, of the hind limb- bones of the Bird for the functions concentrated therein ; the suppression, viz., of the tarsal segment; the simplification, unification, consolidation of the segments above and beneath ''KIMMERIDGE CLAY. 85 it; the tibia alone (woodcut, Fig. 16, 4) articulating with the metatarsus, ib., mt, by a finely fashioned, close-fitting, interlocking joint. As in all warm-blooded quadrupeds and the majority of cold-blooded ones, recent and extinct, the articular ends of the tibia are ossified independently of the shaft, are m the condition of epiphyses in the young Bird (Fig. 16, Dénornis, p *), and retain longer that con- dition in the Reptile (Fig. 16, Varanus, p t, and Scelidosaurus, p t). The attachment of the . distal epiphysis with the shaft of the tibia (¢) is made firmer in the biped (Dzzorais, » t) than in the quadruped (Fig. 16, Ruminant, p ¢) ; and the extent of the attachment is greater, is more irregular or interlocking in the warm-blooded quadruped than in the cold-blooded one; it is still greater in the Bird, in which a process ascends upon the front of the diaphysis, closely fitting to a groove there, and clamping, as it were, the articular epiphysis to the main shaft of the leg bone. The bigger the Bird the greater the share of locomotion allotted to the hind pair of limbs in standing, walking, or running, the longer is the clamping process and the later is the period of the coalescence of the epiphysis with the shaft. The Ostrich among existing Cursores, and the Dinornis amongst extinct ones exemplify this relation. In the metatarsus of the Bird the shafts of the ento-, meso-, and ecto-metatarsi are severally ossified from separate centres, but the proximal epiphyses of the three bones are ossified from one centre, and form a single cap of bone where the shafts are still distinct.1 Such cap (Fig. 16, Dinornis, p m) may be arbitrarily homologised with one or more bones of the distal tarsal series in Reptiles (Fig. 16, Sceldo- saurus, },e; i Varanus, b,e) and in Mammals (Fig. 16, Ruminant, 6, n, e). It seems more natural to regard it as answering to the epiphysial cap, covering the ends of the two chief metatarsals, of the Ruminant (ib. ib., pm, iii, iv), and I associate such instances of complex osteogeny of the metatarsus with the high conditions of organisation differentiating the warm-blooded classes, Aves and Mammalia, from the cold-blood Reptilia. In the Ruminant, as in the Bird, the single epiphysis and multiple diaphyses coalesce into one so-called ‘ cannon bone.’ In the Dinosauria the hind limbs are not adapted, as in the Birds, for transference of the entire weight of trunk, neck, head, and fore limbs, from the leg upon the foot by due development and modifications of the main leg-bone, the tibia ; but the fibula is continued to the ankle-joint, and takes a larger share in its formation than is usual in Mammals. Both leg-bones have their distal epiphyses (Fig. 16, pf, p t. Scelidosaurus, Varanus). The tarsal segment is represented, usually by four ossicles :* one, a, answers, by its connections, to the astragalus, naviculare, and entocuneiform bones of the Mammal ; a second, /, repre- sents the calcaneum with the lever process slightly if at all developed; there are, also, a cuboid, 4, and an ectocuneiform, e. The metatarsals, whether they be three or four «Transactions of the Zoological Society of London,’ 4to, vol. iv (1856), p. 149, pl. xlv (Dinornis elephantopus, pullus ; Dinornis crassus, pullus). * Paleont. vol. for 1860, Oolitic Reptilia (Scelédosaurus), tab. xi, figs. 2, 3, 4, Scelidosaurus, Varanus, Crocodilus. ''86 FOSSIL REPTILIA OF THE in number, never coalesce, but retain their primitive distinctness throughout life. The sole ground taken to bridge over this significant difference in the structure of leg and foot in the Bird and Dinosaur is to affirm that the distal epiphysis, pz, of the tibia in the Bird is the homologue of the astragalus in the Mammal and Reptile (Fig. 16, a). “Tf the whole hind-quarters, from the ilium to the toes, of a half-hatched Chicken could be suddenly enlarged, ossified, and fossilised as they are,”* the ilium would be distinguished from that of a Dinosaur by the major number of its sacrovertebral attach- ments and by their greater extent, by the absence of the ridge continued from the super- acetabular plate upon the antacetabular one; the pelvis would be distinguished by the presence in the ischium of an obturator process wanting in the Dinosaur (Fig. 12, is), and by the absence of a pectineal process of the pubis present in the Dinosaur (ib., pd), by the parallelism of the ischium and pubis, and by the backward extension of both bones (compare Figs. 12 and 14). The differences grow and multiply as the comparison proceeds ; as, ¢.g., by the non-extension, in the Chick, of the fibula (Fig. 14, ¢) to the ankle-joint and by the larger and more complex distal epiphysis of its tibia (Fig. 16, | Dinoruis), by the absence of a tarsus, by the backward direction of the innermost or first toe (Fig. 16,4), as con- trasted with the parallel position of that toe with the second toe in the reptilian foot (Fig. 16, Scelidosaurus, Varanus). Vf the entire skeleton of an immature Chick, Ostrich, or Moa were enlarged, whether suddenly or gradually, to the dimensions of that of a Cetiosaur, and were so ossified and fossilised, the characters of the dorsal vertebrae, of the cervical ver- tebre, of the skull, and the absence of an anterior pair of limbs with fore-paws organized to be applied to the soil and take their share in the support and progression of a long and bulky trunk and massive head as in the Dinosauria, would be decisive against the reference of such imaginary gigantic Chick to any known representative of the Dinosaurian order of Reptiles. But, to the Biologist who rejects the principle of adaptation of struc- ture to function, the foregoing facts and conclusions will have no significance. By a modification of the hind-limbs the Bear, and by addition of a longer sacrum to plantigrade feet the Ground-sloth, may assume a crouching bent-kneed attitude and hold the fore-limbs free to grapple with a foe or a tree. - Such is the plasticity of some mammalian structures that, by due training, a Bear, a Dog, or a Monkey may be taught to dance and walk erect for a brief space. It may be doubted whether a cold-blooded, small-brained Reptile could by any training be brought to exemplify the mode of motion conceived in the quotation at p. 92. But that, like the Chlamydosaur with its long-toed, wide-spread, hind feet, the huge Dinosaurs might assume the fighting posture of the Bear, when occasion called them to wield their carpal weapons, is conceivable without commission of physiological or anatomical solecism. 1 Prof. Huxley, ‘ Quarterly Journal Geol. Soc.,’ vol. xxvi, p. 29. = Tb;, loc, cit., p. 30. ''KIMMERIDGE CLAY. : 87 The woodcuts, p.76, Figs. 12,13, 14,' give the pelvis and hind limb of a Moa (Dinornis) and of a Crocodile (Crocodilus) for comparison with the corresponding parts of a Dinosaur (Omosaurus) : the position, proportions, and structure of the foot of which are guaranteed by those of Zgwanodon and Scelidosaurus. In the Crocodile the foot may be applied flat to the ground and the thigh turned out nearly at right angles to the body; but, in some phases of progressive motion, the limb can assume the position delineated: the same may be predicated of the Dinosaurian Reptile. ‘The Bird occasionally rests on the foot, with the metatarsus flat to the ground: but the thigh cannot be turned outward at the angle, which is possible in the Dinosaur and Crocodile. When an accessory trochanter is present in the femur of a Dinosaur (Zguanodon, Scelidosaurus), it projects from the inner border of the shaft, not from the outer one, as in the restoration given in Fig. 3, p. 27, ‘Quart. Journal Geol. Soc.,’ vol. xxvi, 1870. When the question as to the power of predicating homologies both special and general, as in the case of the bones of the vertebrate skeleton, became finally accepted,” the hypothesis of the successive incoming of specific forms or modifi- cations of the vertebrate archetype through the operation of secondary causes was the only one which could adapt itself intelligibly to the facts. In enunciating my conviction that ‘nomogeny,’ ¢. e. natural laws, or secondary causes, had so operated “ in the orderly succession and progression of such organic phenomena,” I laid myself open to comments from opposite quarters. On the one hand, the admitted ignorance of the nature and mode of operation of such secondary cause or causes led to the rebuke by a Successor in the chair of the Hunterian Professorship, to wit, that, as to the secondary origin of species, my ‘ trumpet gave an uncertain sound.’ On the other hand, an able, theological critic blew the following note of alarm :—‘ It is not German naturalists alone who are contributing to diffuse scientific Pantheism. We have in England an anatomist, Richard Owen. ‘To call him an atheist because of his scientific conclusions would be an imper- tinence ; nevertheless, in a lecture on ‘The Nature of Limbs’ which was delivered at the Royal Institution of Great Britain in February last, and has since been published, he brings all his scientific knowledge and demonstrative skill in support of what is called the T'xory or DeveLopmen’, and which has become popularly known by its introduction into the book called the ‘ Vestiges of Creation. ‘This theory of development, as our - 1 The letters have the same signification throughout; i, ilium; a, antacetabular plate; 2, post- acetabular plate; id (in the Dinosaur) marks the superacetabular plate; is, ischium; pd, pubis; jf, femur (of this only the lower part of the bone is given, so as not to conceal parts of the pelvis important in the comparison) ; ¢, tibia; 4 or 2, fibula; as, astragalus; ca, calaneum; cd, cuboides; i, inner or first toe ; av, second toe; ii, third toe; iv, fourth toe; v, rudiment of fifth toe. «Hunterian Lectures,’ Royal College of Surgeons, 1844 ; ‘Reports of the British Association for the Advancement of Science,’ “On the Archetype and Homologies of the Vertebrate Skeleton,” 8vo, 1846 ; and ‘ Discourse on the Nature of Limbs,’ 8vo, 1849. ''88 | FOSSIL REPTILIA OF THE readers may know, assumes that Gop did not imterpose to create one class of creatures: after another as the consequence of each geological revolution ; but that, through the long course of ages, one class of creatures was developed from another. Now, Richard Owen undertakes to demonstrate scientifically (and his demonstration is very rigorous) that the arms and legs of the human race are the later and higher developments of the ruder wings and fins of the vertebrated animals—that is, those which have a true back- bone ; and he shows in the splint bones of the foot of a horse, bones analogous to those of the fingers of the human hand. ‘Therefore he concludes that Gop has not peopled the globe by successive creations, but by the operation of general laws.”? The sole ground for Professor Flower’s depreciatory remark is my acknowledgment of being “ as yet ignorant ’” of the nature or way of operation of such general or secondary laws; and I regret to say that after all that has been advanced since 1849 in the endeavour to elucidate the way in which one species may be transmuted into another, I am still in need of light. Assuming that the ornithic modification of the vertebrate archetype was one of those under which the ‘ vertebrate idea ’ became embodied im the course of progression from “its old Ichthyic vestment,”* two questions present themselves :—Out of what antecedent vertebrate modification was the avian one evolved? How, or under what conditions or secondary influences, was such evolution effected ? The hypothesis of the bipedal locomotion of the Dinosauria, the advocated homology of their os pubis with the ischium of the bird, and the alleged restriction of the avian antacetabular production of the iliac bone to the Dinosauria among Reptiles, have been superadded to the proved fact of a correspondence of structure between the shorter sacrum of the Dinosaurs and the longer sacrum of Birds as grounds for the conclusion that Birds are transmuted Dinosaurs, and that the feathered class made their first step in advance under the low form of Struthiones or Cursores, incapable, as yet, of flight. The kind and amount of modification required to evolve an Ostrich out of an Iguanodon may be appreciated by the osteological comparisons already submitted in the present mono- eraph. To revert only to the structure of the fore-limb. In losing its power of aiding in the quadrupedal progression, and of grasping or otherwise applying the hand, it has as yet, in the hypothetical first form of Birds, gained no other faculty. At best it may help in the swift course of the ostrich by flapping motions similar to those of better birds during their flight ; or the more minute monodactyle hand may just serve to scratch the back of the head, as in the New Zealand Kivi. In their larger extinct relatives, the Moas, it is still doubtful whether more of the framework of a fore-limb existed than the: supporting scapular arch, and that of the simplest character. 1 © Little Lectures on Great Topics,’ 12mo, 1849. 2 “On the Nature of Limbs,’ p. 86. 5 Tid. ''KIMMERIDGE CLAY. 89 In all these gradations of structure in a limb unavailable for flight or any other mode ‘of locomotion we see no approach in the scapula to the Dinosaurian types of that bone ; it retains in all Cursorials the strictly avian sabre-like shape and pointed free extremity, without expansion and truncation there such as obtains in the alleged ancestral Reptilia! ‘The coracoid still further departs from any well-determined Dinosaurian type of the bone, and as closely adheres to that of the Birds of flight, save such decrease of breadth and of relative size as accords with its necessity to bear upon the sternum in the mechanical mode of inspiration peculiar to Birds with Pterodactyles. What could be the conceivable conditions of the life of an Iguanodon or Megalosaur which rendered a fore-limb useless or cumbersome, and concomitantly called for lengthened and strengthened hind-limbs and a more vigorous and exclusive exercise of these in the acts of locomotion? The abettors and acceptors of the exposition of the operation of the secondary mode of origin of species by way of ‘ natural selection’ are amenable to the call for an explanation of such conditions, especially if such mode of origin be hypothetically applied to the kinds of Birds deprived of the power of flight. But such explanation would have to square with the fact that a loss of one pair of limbs had been associated, on the assumption. of the Dinosaurian ancestry, with an advance of the mechanical structure of the organs of circulation, and in the extent and perfection of the lungs, together resulting in the higher temperature, with more numerous and minute coloured discs of the blood. or these conditions of the vital organs characterise alike both winged and wingless Birds, and the resultant unvarying warmth of the body is accompanied by a clothing of down and feathers, the most exquisite and complex of all tegumentary coverings, common to the Kivi and Ostrich with the Eagle and Swift. But there are other hypotheses of the way of operation of secondary genesis of species anterior in date to that of Darwin. The influence, viz., of exercise and of disuse in altering the proportions of parts mooted by Lamarck ;? the hypothesis of ‘ degeneration ’ pro- pounded by Buffon ;* and the effects of congenital changes in parts of the body, mainly depended upon by the author of ‘ Vestiges, in his endeavour to explain the way of operation of the secondary law of the origin of species. The comparative ease is so refreshing, after the labours of induction and dry descrip- tion, in supposing a case, that I may be forgiven for indulging in a suggestion of a possibility of the few still extant wingless or flightless birds having originated, not from any lower cold-blooded vertebrate form, but from higher active volant members of their own warm-blooded feathered class. Consideration of extinct kinds, in the restoration of which I have been occupied, has strengthened the supposition. Here, in yielding to this indulgence, I own to finding more help from the Lamarckian * Compare, for example, the scapula of the Apteryx, ‘Transactions of the Zoological Society,’ vol. ii, pl. xxx, fig. 2, g, and figs. 3 and 4, with Cut, fig. 2, p. 31. * ¢ Philosophie Zoologique,’ 2 vols., tom. i, chaps. iii, vi, vii, 8vo, 1803. ® «Histoire Naturelle, tom. xiv, p. 311, 4to, 1766. ''90 FOSSIL REPTILIA OF THE hypothesis than the Darwinian one, and I am ultimtely lead to propound the Struthionide as exemplifications of Buffon’s belief in the origin of species by way of degeneration ; on other grounds than those on which my anonymous Critic, above cited (p. 87), views the Papuan and Boschisman in relation to an antecedent higher, indeed perfect, form of man. Let us suppose, for example, an island affording abundant subsistence to vegetarian birds, and, happily for them, to be destitute of creatures able or desirous to destroy such birds. If the food was wholly, or chiefly, on the surface the power of traversing such surface would be of as much advantage to the bird as to the herbivorous quadruped. As flight calls for more effort than course; so cursorial progression would ‘be more commonly practised in such a happy island for obtaining the daily food. The advent or proximity of a known element of danger might excite the quicker mode of motion ; the bird would then betake itself by a hurried flight to a safer locality. If, however, certain insular birds had never known a foe, the stimulus to the use of the wings would be wanting in species needing only to traverse the ground in quest of food. In the case of New Zealand, for example, the roots of wide-spread ferns, being rich in farmaceous and amylaceous princi- ples, the habit of scratching them out of the ground would lead to full development of the muscles of the leg and foot. So, such daily habitual exercise of legs and feet by unscared Rasorials would lead in successive generations to strange developments of hind-limbs ; whilst the disuse of the wings during the pre-Maori zons would lead to their atrophy. The Lamarckian hypothesis has, in fact, this advantage over others of like kidney, that physi- ology testifies to the relation of growth to exercise, and of waste to disuse, and so far votes in favour of the conditions evoked by Lamarck as vere cause in transmutation. We recog- nise in the stunted wings of the Dodo, for example, their close conformity, save in size, and in the prominence of their processes for muscular attachments, to the scapula, coracoid,. brachial and antibrachial bones, carpus, metacarpus, &c., of the perfect instrument of flight in truly winged birds, evidences of its affinity ; and such conformity of structure is agree- able with the hypothesis of the origin of the Mauritian species of ground-pigeon through descent or degeneration. The differences which the wing-bones of the Dodo present when compared with their homologues in the Zywanodon is in the same degree adverse to the hypothesis of its evolution from any such reptile, in the direction of ascent and improve- ment. ‘The same course of argument applies to the impennate Awk, the Cassowary, Rhea, Ostrich, &., as to the wingless birds of the Mascarene, Polynesian, or Melanesian Islands. Confidence in the impartial exercise by Biologists of the logical faculty leads to the conclusion that their science will accept the view of the Dodo as a degenerate Dove rather than as an advanced Dinothere. But whence the dove? Are we then, I will not say driven, but rather guided, to the old belief that the winged bird was “created ” in the sense of being .miraculously made, at once, out of dust, agreeably with the alternative hypothesis conceived by my critic? Or, is a belief in a Dove’s coming to be through the ''KIMMERIDGE CLAY. 91 operation of a secondary law still legitimate and germain to our truth-seeking faculties ? Not necessarily relegating an honest inquirer to the bottomless pit of Atheism, if he should happen to ask :—Were there no volant vertebrates of earlier date and lower grade than the “ Fowls of the Air” ? Without knowing or pretending to know the way of operation of the secondary cause, the vast increase of knowledge-stores of biological phenomena makes it as impossible to comprehend them intelligibly in any degree, on the assumption of primary or direct creation of species, as it was impossible for Copernicus to understand and explain the vast accession of astronomical facts, on the belief of the subservient relation of sun to earth, of the posteriority of the creation of the luminary to the light-receiver, and of their respec- tive relations of motion, as received in his day. ‘To the objection, how, on his assumption of the diurnal rotation of the earth, loose things remained on its surface, Copernicus could offer no explanation. Neither has the Biologist been able, as yet, to explain how the Ramphorhynchus became transmuted into the Archeopteryx. It is open, of course, for any one to deny such change. What seems to me to be legitimate, in giving an account of the labours that have resulted in a certain accession to the knowledge of extinct forms of cold-blooded, oviparous, air-breathing Vertebrates, is the indication of the respective vicinity of certain groups of such now much reduced class to the warm-blooded oviparous Vertebrate air-breathers which in our times so greatly prevail in life’s theatre. Every bone in the Bird was antecedently present in the framework of the Pterodactyle ; the resemblance of that portion directly subservient to flight is closer in the naked one to that in the feathered flyer than it is to the fore-limb of the terrestrial or aquatic Reptile. No Dinosaur has the caudal vertebra: reduced as in Birds; many Pterodactyles manifest that significant resemblance. But some Pterodactyles had long tails and all had toothed jaws. A bird of the oolitic period? combined a long tail of many vertebrae with true avian wings, and it may have had teeth in its mandibles. It is certain that a later extinct bird,’ though of an early tertiary period, far back in time beyond the present reign of birds, had tooth- like processes of the alveolar borders of both upper and lower jaws. Fact by fact, as they slowly and successively drop in, testify in favour of the coming in of species by ‘nomogeny, and speak as strongly against ‘thaumatogeny’’ or the multiplication of miracle on the alternative hypothesis of the writer of ‘ Little Lectures on Great Things.’ He and his school invoke a cataclysm to extinguish the Palzeothere, and an inconceivable operation to convert dust into the Hippothere ; yet a slight disproportion of the outer and inner of the three hoofed toes of each foot of these quadrupeds is their main difference. My critic again invokes a cataclysm to extinguish the race of Hippotherian species and again requires the miracle to create the Horse. Yet the loss of the small side-hoofs that dangled behind the main mid-hoof in the Hippothere is the | Archeopteryz, ‘ Philosophical Transactions,’ 1863. * Odontopteryx, ‘ Quarterly Journal of the Geological Society,’ 1873. > «Anatomy of Vertebrates,’ Svo, vol. iii, p. 814. 2 ''92 FOSSIL REPTILIA OF THE chief organic distinction between, Hippotherium and Hippos. Every bone, every tooth, present in the eocene and miocene predecessors of modern Horses is retained in them, with slight changes of shape and proportion. The second and fourth metacarpals which bore hoofed digits of moderate size in eocene days, bore them of diminutive size in miocene days ; and now, when such dangling spurious hoofs are gone, their metacarpal and meta- tarsal suspensories still remain, hidden beneath the skin, and ending in a point where, of old, was a well-turned joint. It has become as impossible to square the hypothesis of “the peopling of the globe during the long reign of life thereon, by successive and special creations ” with the known vital phenomena, as it was impossible to explain the sum of astronomical facts, accumulated in the fourteenth century by the cumbrous machinery of cycles and epicycles, necessitated under the assumption of the globe as the fixed, central, and largest. body of the Universe. Biology seems now to be at the Copernican stage; and if the rejection of the incoming of species by primary creative acts should exercise an influence on the pro- gress of that science akin to that of astronomy after the abandonment of the faith in the earth’s fixity, Biologists may confidently look for as rapid a progress through acceptance of Nomogeny.. What, then, may be the meaning of the reduction of bulk in the fore-limbs of certain. Dinosaurs? Does that reduction indicate a step in the conversion of such Reptiles into Birds? Dowe get an explanation of the small fore-limbs by the picture which Professor Phillips vividly presents to us “of the grand and free march on land chiefly, if not solely, on the hind-limbs *”’ Or, is the fact of the disproportion of size between the arms and legs in the Megalosaur and Iguanodon, susceptible of other than the Oxfordian hypothesis ? As a matter of fact, such disproportion is shown by Crocodilian Reptiles still in existence ; whilst extinct Crocodiles of more aquatic habits and marine sphere of life had the fore-limbs as much reduced in size as in any known Dinosaur.! Of this Zéleosaurian character the physiological explanation which has been advanced is, that the course of such. Crocodile through water, due to the action of the long, laterally flattened tail, would be facilitated, or less impeded, by such: reduction of size of the fore-limbs, those limbs. taking no share in the forward dash of the piscivorous reptile in pursuit of its prey, and if of any use in the water, being limited in, natatory evolutions to assist in a change of direction ; the fore-limbs, in fact, being mainly if not wholly required to help im the: progress of the amphibious beast upon dry land, or to. scratch out the nest in the sand. Actual observation of a swimming Crocodile testifies to the fore-limbs being then laid flat: and. motionless upon, the sides of the chest. All known Dinosaurs have the Crocodilian: swimming organ; the Iguanodon exemplifies: the compressed vertically broadened tail in 1 “Monograph on the Fossil Reptilia of the London Clay,” part ii, in the Volume of the Paleeonto- graphical Society for 1849, p. 24, t. xi. ''KIMMERIDGE CLAY. 93 aneminent degree. And just as such appendage was essential to the proportion of the active life of these huge cold-blooded amphibians which was spent in the watery element, so such far-produced caudal fin must have been a cumbrous impediment to the way of walking upon dry land pictured in the work giving figures of the caudal vertebrae and other bones of the Cetiosaurus longus. In the ratio in which the fore-limbs approach the hind ones in size may be inferred the proportion of time spent by the huge reptile on land, and the importance of the share taken by these limbs in such quadrupedal mode of progression: when the Dinosaur betook itself to water its fore-limbs would be, most probably, disposed as in the Crocodiles. If, then, the hypothesis that the reduced fore-limbs of Dinosauria receive the most intelligible, and therefore acceptable, explanation, admitting the principle of adapta- tion of structures to functions, agreeably with the analogy of such living animals as are most nearly allied to them in organization; the notion that Birds, under their wingless conditions, were derived from Dinosaurs may be safely left to the judgment of whomsoever may be disposed to bring unprepossessed and impartial judgment to the consideration of the hypothesis. '' ''MONOGRAPH ON THE Genus OMOSAURUS. (Continued.) Speciese—OMOSAURUS HASTIGER, Owen. (Plates XXIIT and XXIV). Ir the grounds assigned in the former part of this Monograph’ for the probable homology of the unsymmetrical spine figured in Plates XXI and XXII, which spine was found with the bones of the fore-limb of Omosaurus armatus, should be deemed to warrant such conclusion, a similar one may be provisionally accepted as applicable to the pair of spines of similar size and character discovered in the same division of the Kimmeridge Clay, in the Great Western Railway Cutting at Wootton Bassett, Wiltshire, briefly referred to at p. 68 of that portion of the Monograph. Many large Saurian fossils were collected from the sections of Kimmeridge Clay at that time exposed; but none have reached me save the subjects of the present Mono- graph, which were there obtained by William Cunnington, Hsq., F.G.S., and have passed with the rest of his collection into the possession of the British Museum. The apical portion of each spine has been broken away, but the degree of decrease from the base affords satisfactory grounds for the restoration given in Plate XXIV, the ratio of decrease being less in the present species than in the almost perfect spine of Omosaurus armatus.” The-base of the spine (ib., 4) expands from the body, a (Plate XXIV), more suddenly and in a greater degree in Omosaurus hastiger. It is suboval in form and, as in Omos. 1 Volume of the Paleeontographical Society issued for the year 1875, p. 67. 2 Ib., pl. xxi, figs. 1 and 2. ''96 FOSSIL REPTILIA OF THE armatus, its plane is oblique to the axis of the spine. The long diameter of the base is 9 inches, the short diameter is 7 inches. The articular surface is divided into two unequal facets by a low ridge of the base (Plate XXIII, fig. 1, +, 7) parallel with the long diameter of the base; each facet is feebly convex lengthwise, less feebly concave transversely. The surface for attachment is roughened by low short ridges diverging from the long ridge, », and is irregularly pierced by vascular canals ; the borders are thick and irregularly notched. The body of the spine is continued more directly from one end (Plate XXIV, figs. 1, 2, 3) of the oval base, a, fig. 2, sloping and expanding more gradually to the opposite end of the base, 4, fig. 2. The body of the spine is a full oval in transverse section (ib., fig. 4), pointed at each end, where the two opposite edges, d, e, are cut. The anterior edge (fig. 1, a), begins about 6 inches beyond the anterior produced part of the base; the posterior edge (fig. 3, e) begins about 2 inches from that end of the base. Both edges extend along the preserved portions of each spine, and were probably continued to, or near to, the pointed end. An additional advantage as a lethal or piercing weapon must have been derived from this two-edged structure. In the right spine (fig. 1) the length preserved is 14 inches; in the left spine (fig. 3) the length preserved is 10 inches. Lach spine may be estimated to have been upwards of 20 inches in length when entire. The transverse section taken from the broken end of the left spine (fig. 4) gives 4 inches and 34 inches in the two diameters: the broken end of the better preserved spine gives 3 inches and 23 inches in the two diameters ; the spine approaches to a circular section as it nears the pointed end. The texture of the outer inch is a compact bone susceptible of a high polish ; it becomes finely cancellous within a few lines of the central cavity, the section of which at the part cut, viz. 85 inches from the base of the spine, gives 1 inch 6 lines, and 1 inch 8 lines, in the long and short diameters. The close correspondence of the present fossil in general form, in basal modifications for attachment, and in texture, with the spine, probably left carpal, of Omosawrus armatus, will be obvious on comparison of Plates XXIII and XXIV with Plates XXI and XXII of the former part of this Monograph, treating of that species; and such correspondence may be deemed to support the provisional reference of the carpal (?) spines from the Kimmeridge Clay of Wootton Bassett to the same genus as that from the Kimmeridge Clay of Swindon ; they manifestly indicate a distinct species on the above hypothesis of their nature. The osseous core of the carpal spine in Jywanodon (‘ Wealden Reptilia,’ Sup., No. 4, Pal. vol. for the year 1871, issued in 1872, Plate II, fig. 2) differs chiefly in its relative shortness or speedier diminution from the base to the apex. After a comparison of these fossils with all the examples of carpal and tarsal spines in existing vertebrates, I found the nearest resemblance to the basal expansion, by which '' KIMMERIDGE CLAY. 97 the spine of Omosaurus has been attached, in the tarsal spine of the Platypus (Ornitho- rhynchus paradoxus, Plate XXIII, fig. 2, twice natural size). There was the same proportion of breadth to the body of the spine; the same sudden expansion to form the base; the same medial rising in the long axis of the base, and furrows extending therefrom to the margin. But these radiating furrows are more numerous, and the spine, though it is hollow as in Omosaurus, has that cavity converted by terminal apertures into a canal, and this canal is traversed, as in the poison-fang of certain Ophidian Reptiles, by the duct of a gland. ‘The affinity shown by the Monotrematous Mammals to the Reptilia in certain parts of the skeleton is well known, and is closer in the structure of sternum, coracoids, and clavicles, than in any Bird. ie. '' '' ''be ‘ol Q OONOoP 09S — 23. 24. 26. ne. 28. PLATE I. Mesozoic Prerosauria. . Left side view of fore end of upper jaw of Coloborhynchus clavirostris. . Front view of the same specimen. . Hind view of ditto. . Under or palatal view of ditto. . Left side view of fore end of lower jaw of Pterodactylus Daviesit. . Upper or oral view of the same specimen. . Front or thenal view of proximal portion of right humerus of Pterodactylus Marderi. . Back or anconal view of the same specimen. . Proximal end, with the ‘ head’ or articular surface, of ditto. . Front or thenal view of proximal portion of left humerus of Pterodactylus Manselii. . Back or anconal view of the same specimen. . Proximal end, with the head or articular surface, of do. . Front or thenal view of distal end of right humerus of Pterodactylus (Dimorphodon) MACTOYNL. . Back or anconal view of the same specimen. . Front or thenal view of distal portion of left humerus of Pterodactylus Pleydellii. Back or anconal view of the same specimen. 16’. Articular end of ditto. . Front or thenal view of the proximal phalanx of the fourth or ‘wing-finger ’ of Pterodactylus Kiddit. ' . Back or anconal view of the proximal phalanx of the ‘ wing-finger’ of Pterodactylus Duancani. . Outer side view of the proximal phalanx of the ‘wing-finger’ of Pterodactylus Acland. . Outer side view of the proximal end of proximal phalanx of the ‘ wing-finger’ of Pterodactylus Manselit. . Back or anconal view of the same specimen; 21*. articular surface of ditto. . Outer side view of proximal end of proximal phalanx of the ‘ wing-finger’ of Péero- dactylus Pleydeltit. Back or anconal view of the same specimen. 23’. Articular surface of ditto. Articular surface of carpal bone of Pterodactylus. . Opposite articular surface of the same specimen. Articular surface of homologous carpal bone of another kind of Pterodactyle. Opposite articular surface of the same specimen. Back or anconal view of second phalanx of ‘ wing-finger’ of a Pterodactyle. All the figures of the natural size. The subject of figures 1—4, from the Hastings series of the Wealden, west of St.- Leonard’s-on-Sea, Sussex, is in the possession of its discoverer, Samuel H. Beckles, Esq.., F.R.S., &. Those of the following figures are in the British Museum :—5 and 6 are from the Gault at Folkestone; 7, 8, 9, 13, 14, are from the Lias of Lyme Regis, Dorset- peer ee 11 ie, 1b, 16, 20, 21, 21%, 22, 28, 23°,°24, 25, 26, 27, are from the Kimmeridge Clay, Dorsetshire; 17, 18, 19, 28, are from the Oolitic State, Stonesfield, Oxfordshire. '' SOZOIG REP TILIA 1 4 ME Mmtern Bros Imp. lth. riesbach del e& Cie '' '' ''PLATE (11. Mesozoic PrerosauRIA. Fre. 1. Outer side view of part of left mandibular ramus, with teeth, of Pterodactylus sagittirostris. 2. Outer side view of part of right ramus of the same mandible. e Upper view of the same specimen. . Inner side view of part of the left mandibular ramus of ditto. . Outer side view of articular part of the right mandibular ramus of ditto. . Upper view of the same specimen. . Inner side view of the same specimen. oS mt Oo Ot The above figures are of the natural size. 8. Reduced upper view of the same specimens, with conjectural restoration, in dotted outline, of the form or proportions of the entire mandible of Pterodactylus sagittirostris. This fossil, from the Hastings series of the Wealden, west of St.-Leonard’s-on-Sea, Sussex, is in possession of its discoverer, Samuel H. Beckles, Esq., F.R.S. '' MESOZOIC REPTILIA. Pi 1 L Griesbach del etlith. Mintern Bros. amp. '' '' ''PLATE III. Bothriospondylus suffossus. Fig. 1. Heemal or under view of centrum of sacral vertebra. 2. Neural or upper view of the same. 3. Hind view of the same. 4. Right side view of the same. All the figures are of the natural size. From the Kimmeridge Clay at Swindon, Wilts. In the British Museum. '' Ps, Pe oUAIG REP TILIA M&N.Hanhart im Pp C.L.Griesbach, del & lith DErOSeUS a i) BUIMRIOSPOND YLUS '' '' '' PLATE IV. Bothriospondylus suffossus. Fic. . Hind view of terminal centrum of sacral vertebra. . Right side view of the same. Heemal view of the same. . Neural view of mutilated centrum of sacral vertebra, restored in outline. Right side view of the same. a Heemal view of the same, restored in outline. All the figures are of the natural size. From the Kimmeridge Clay at Swindon, Wilts. In the British Museum. ''eae Mintern Bros imp Gresbach Cc UFF wo LU BOTHRIOSPONDY '' '' ''PLATE V. Bothriospondylus suffossus. Fig. ‘s 1. Right side view of centrum of dorso-lumbar vertebra. wo . Part of fore surface of the same, restored in outline. Hind surface of the same. Part of hemal surface of the same, restored in outline. = & Neural surface of the same. All the figures are of the natural size. From the Kimmeridge Clay at Swindon, Wilts. In the British Museum. '' MESOZOIC REPTILIA C.L Griesbach del &lith M&NHanhart imp BOTHRIOSPONDYLUS SUPPOSsTTa '' '' ''PLATE VI. Bothriospondylus robustus. Fie. 1. Left side view of centrum of dorso-lumbar vertebra. 2. Section of the same, showing chondrosal cancelli. Both figures are of the natural size. From the Forest Marble of Bradford, Wilts. In the British Museum. '' 3 MESOZOIC -REPTILIA . Pu ipring Pr tiie Si re, Ra C.L.Griesbach del & ith. M&NHanhart imp. BOTHRIOSPUNDYIUS ROBUSTUS, '' '' ''PLATE VII. Bothriospondylus elongatus. Right side view of mutilated centrum of dorso-lumbar vertebra, natural size. From the Wealden of Tilgate. In the British Museum. '' eal SaZzOle REP TILIA C.L Griesbach del &hth . M&N.Hanhart Tup '' '' ''PLATE Vill. Bothriospondylus magnus. Right side view of centrum of dorsal vertebra, natural size. From the Wealden of the Isle of Wight. In the British Museum. ''‘SQNOVN SOTACNOdS OIMH.LOG ; a 2 § = a meu! a AY oq S § a: a = § = S "O fi “- eS 5 ® 2 IN MESO '' '' ''PLATE 1X, Lothriospondylus magnus. Fie. . Fore end of mutilated centrum of dorsal vertebra. . Side view of crown of tooth of Cardiodon rugulosus. Fore end of the same tooth. Hind end of the crown and beginning of the fang of another tooth of Cardiodon rugulosus. 5. Magnified view of markings on the surface of the enamel of the same tooth. Hm 0 ~ All the figures are of the natural size. Fig. 1, from the Wealden of the Isle of Wight. In the British Museum. Figs. 2—5 are copied from Owen’s ‘Odontography,’ 4to, 1840—1845, p. 291, Pl. LXXV, a (the subjects were from the Forest Marble at Bradford, Wilts, and formed part of the collection of Channing Pearce, Esq., of that town). ''1X 1D) < a a A, fr) a © OS Mintern Bros. mip . esbach L.Gri C IS MAGNUS. UO LI BOTHRIOSPONDY. '' '' ''ae eet PLATE X. — Cetiosaurus longus. Both figures are of the natural size. Great Oolite at Kirtlington, Oxfordshire. In the Geological Museum, i 7. : ''Mssozorc Reprimia. CETIOSAURUS LONGUS. Mintern Bros. imp '' '' ''PLATE XI. Omosaurus armatus. Fig. 1. Side-view of neural arch and spine of a cervical vertebra. 2. Upper-view of the same. 3. Side-view of a cervical vertebra of Varanus niloticus. 4, Upper-view of the same. All the figures are of the natural size. The subject of figs. 1 and 2 is from the Kimmeridge Clay at Swindon, Wilts. In the British Museum. ''Mesozorc Reprint. F; ele part NOOR OAY Ny AP MOON a “ tena a oe pretty er tim oenenccwne GLGmesbach del et hith Mintern Bros, imp. '' '' ''PLATE XII. Omosaurus armatus. Fie. 1. Front-view of a dorsal vertebra. Upper-view of the centrum of a dorsal vertebra. 3. Side-view of the same. The figures are of half the natural size. From the Kimmeridge Clay of Swindon, Wilts. In the British Museum. '' Pix. Mesozoic Reprinia. Mintern Bros. imp. C.L.Griesbach del. et lith. '' '' ''PLATE XIII. Omosaurus armatus. Fie. 1. Back-view of the neural arch of the dorsal vertebra, ae. 1, PL XL 2. Side-view of the neural arch of the same vertebra. 3. Side-view of the centrum and base of neurapophysis of the same. The figures are of half the natural size. From the Kimmeridge Clay of Swindon, Wilts. In the British Museum. ''Mesozoic Reprinia Pi Xl. HagoA. C.L.Griesbach del et lith. Mintern Bros . imp. '' '' ''PLATE XIV. Omosaurus armatus. Fig. 1. Front-view of an anterior caudal vertebra. 2. Upper-view of end of neural spine. 3. Upper-view of centrum and right pleurapophysis. 4. Under-view of the same. 5. Proximal end of fourth metacarpal. 6. Distal end of the same. All the figures are of the natural size. From the Kimmeridge Clay of Swindon, Wilts. In the British Museum. ''PL XIV. Mesozoic Reptinia. a th del. et hth. sbac Grie C.L C '' LIBRAaS OF Vag UNIVERsrry ~