Megalosaurus, Buckland, 1822

Matthew, W. D., & Brown, B., 1922, The family Deinodontidae, with notice of a new genus from the Cretaceous of Alberta., Bulletin of the American Museum of Natural History 56, pp. 365-385 : 369-371

Megalosaurus — Allosaurus — Ceratosaurus

This group has been very fully described, so far as the American genera are concerned, by Mr. Gilmore in his highly authoritative and admirably written and illustrated memoir on the Carnivorous Dinosaurs in the U. S. National Museum. 1 While we fully agree with most of Mr. Gilmore’s conclusions therein set forth, there are a few points on which, after careful consideration of his evidence and argument, we are unable to adopt his views. Among these differences are the inclusion of the deinodonts in the Megalosauridae , the retention of the Ceratosauridae as a separate family, and the dropping of the well-known name Allosaurus in favor of Antrodemus . The objections to these points of procedure will be stated later.

The characterization of this group as here given rests primarily upon Allosaurus , secondarily upon Ceratosaurus , both known from complete skeletons. Megalosaurus , although discovered nearly a century ago, is still imperfectly known and there are some peculiarities about the known parts that suggest that it is really not closely related to the American genera, and is an early stage in the evolution of the very extraordinary group from the Lower Cretaceous of North Africa that Strömer has recently characterized under the name of Spinosauridae . The reference of the American genera to the Megalosauridae we regard as provisional.

The characters of the group as thus based are:

1.—Moderate to large size and rather massive proportions.

2.—Skull relatively large, teeth large, jaw long and deep. Fenestrae large, arcades moderately heavy. Anterior teeth somewhat U-shaped, both crests tending to become posterior, unreduced in size.

3.—Quadrate loosely united to quadratojugal, rather short, extending strongly backwards as well as downwards.

4.—Considerable movement between frontal and parietal.

5.—Cervicals of moderate length.

6.— Fore limb moderately reduced, powerfully proportioned.

7.—Manus short and spreading with large powerful claws, the outer digits reduced and vestigial to a varying degree.

8.—Ungual phalanges of manus strongly curved, moderately compressed.

9.—Pelvic elements usually separate. Ilium high anteriorly with long and massive peduncle. Ischium with moderately heavy shaft, expanded distally into a massive thickened head.

10.—Tibia shorter than femur, pes rather short, massive or of moderate proportions 11 —Three subequal functional metatarsals, the shafts unreduced and not strongly appressed, of moderate length, the distal ends convex both ways. Digit I reduced to a distal vestige of the metatarsal, appressed against the shaft of me. II with phalanges retroverted. Digit V a proximal splint.

.

12.—Phalanges of pes short and stout, the unguals moderately curved, uncompressed.

13.—Tail long, of many vertebrae, the distal caudals elongate, the prezygapophyses prolonged anteriorly.

Ceratosaurus differs from Allosaurus chiefly in four characters, the median frontal horn, the union of the metatarsals into a single bone, much as in the penguin, the union of the pelvic bones, and the decidedly less reduction of the external digits of the manus. The second and third characters do not involve any very marked difference in the form or proportions of the bones thus united and, while they may be characteristics of the genus and not merely of the individual due to age, yet they do not in our view involve any such wide osteological diversity as would warrant using them in family differentiation. The median horn of Ceratosaurus is a good generic distinction, but it is not at all likely that it would be a family character. The difference in the manus is the one really important distinction and here Ceratosaurus represents a decidedly more primitive stage in a line of specialization apparently somewhat different. As it is a contemporary of Allosaurus , this can be viewed partly as a structurally ancestral stage, partly as a divergent specialization, but it does not appear to us to be comparable to the diversity between Ornitholestes , Struthiomimus , and Gorgosaurus , which represent three distinct divergent lines of specialization in the manus. Taken all together, the differences between Ceratosaurus and Allosaurus do not appear to be more fundamental than those that distinguish Deinodon from Tyrannosaurus , Monoclonius from Triceratops , or Corythosaurus from Trachodon . They are of not more than sub-family value.

The substitution of Antrodemus for Allosaurus does not appear to be warranted at present, although future discovery of topotypes may prove it correct. Antrodemus is based upon an incomplete vertebral centrum, undoubtedly of a large theropod dinosaur from the “Morrison” formation of Middle Park, Colorado. No topotypes are known, and the vertebrate fauna from this vicinity is unknown. Without such topotype evidence we believe it is unsafe to conclude that the Antrodemus type is generically identical with Allosaurus . It may be provisionally so regarded, but a provisional reference is not sufficient ground to change the nomenclature. The type of Allosaurus is hardly more definitely determinable than that of Antrodemus , as Mr. Gilmore points out. But it is supported by adequate topotypes, among which is the fine skeleton which Gilmore describes and figures so fully in his memoir. It seems better, therefore, to retain Allosaurus until adequate topotypes of Antrodemus have been obtained.

It is important to point out that in the differences between Allosaurus and Ceratosaurus we have apparently a guide to the tendencies of specialization in this family. The pes, as one would expect in a massively proportioned animal of this type, tends to consolidate the metatarsals into a single bone, stout and massive, the median element unreduced. This could not possibly lead into such a specialized type as is seen in the deinodonts and ornithomimids, which finds its analogy among rodents, marsupials, and birds in long, slender-footed, and comparatively small animals of leaping or cursorial adaptation. The manus, on the other hand, more specialized in Allosaurus , is a broad, short, spreading affair, not at all reduced in relative size compared with that of Ceratoaurus and having an important prehensile function. The further specialization here would tend apparently to two widely divergent, grasping digits on a moderately large and very powerful limb. The clue to the derivation of the deinodont manus may be found in either the megalosaurid or coelurid type, but only through a marked reduction in size, which would hardly be expected in gigantic predaceous beasts and is not indicated in the megalosaurs. The tail of the megalosaurids is primitive but shows conditions that might lead into the peculiar specialization seen in the coelurids, deinodonts and ornithomimids. Perhaps the most significant differences are seen in the pelvis, which in the deinodonts is very much nearer to the coelurid and ornithomimid type than it is to the megalosaurid construction. The megalosaurs are nearer the sauropoda in their pelvic construction than are the other three families. On the other hand, there are no very obvious characters in the skull of the megalosaurs, which would preclude regarding them as the primitive ancestral group from which the deinodonts were derived.