4 BULLETIN 110, UNITED STATES NATIONAL MUSEUM. 



from the same locality as specimen No. 4734 to be described here, and possibly from 

 the same quarry. The fore and hind limbs (No. 4734 and 4735 U.S.N.M.) figured 

 so widely by Marsh do not pertain to the type specimen, although such reference 1 

 has been made regarding them. 



In compliance with my request Prof. R. S. Lull, of the Yale University Museum, 

 very kindly forwarded plaster copies of all of the bones comprising the type speci- 

 men of Allosaurus fragilis Marsh. They consist of a well-preserved tooth, two 

 dorsal centra and the proximal phalanx of digit III of the right hind foot. The 

 vertebral centra are considerably damaged (figs. 2 and 3, pi. 2), reproduced here 

 from a photograph made of the casts, and both pertain to the posterior dorsal 

 region, notwithstanding Marsh's earlier determination that the smaller one of the 

 two belonged to the caudal series. • These casts have been carefully compared with 

 the homologous parts of a very complete skeleton, No. 4734, U.S.N.M., from the 

 same locality as the type, a specimen which Marsh had previously identified as 

 Allosaurus fragilis, and of which he published an illustration of the articulated 

 pelvis, hind limb and foot 2 (pi. 13). A considerable part of the tail of this specimen 

 is missing, and it can not, therefore, be directly compared with the genotype of 

 Antrodemus. Compared with No. 8367, U.S.N.M., consisting of a well-preserved 

 though not a complete backbone and other skeletal parts, shows these two speci- 

 mens as unquestionably belonging to the same genus, and this latter specimen is 

 the link in the chain that enables me for the first time to correctly determine the 

 true affinities of the caudal centrum on which the genus Antrodemus is founded. 

 That all of these specimens belong to the same genus there now seems to be but 

 little doubt. 



The fragmentary nature of the type specimens of Antrodemus and Allosaurus 

 renders the proper disposition of these genera a most perplexing problem. In this 

 connection the following nomenclatural propositions present themselves: 



1. Regard Allosaurus fragilis to be a synonym of Antrodemus valens, because 

 of the very close resemblances pointed out in the preceding paragraphs. 



2. Regard the type of Antrodemus valens as an indeterminate specimen, and 

 continue the use of Allosaurus fragilis, characterizing it on the splendidly preserved 

 topotype No. 4734, U.S.N.M., to which Marsh has applied his name, a name that 

 has become well established both in this country and abroad. 



3. Regard both the types of Allosaurus fragilis and Antrodemus valens as being 

 indeterminate and create a new genus and species based on an adequate specimen. 



On account of the close resemblances, down almost to the minutest details, 

 between the type of Antrodemus valens, fragmentary though it is, and the corre- 

 sponding bone in an adequate specimen shown to be the same as the topotype of 

 Allosaurus fragilis, the first proposition set forth above seems to me the logical one 

 to adopt in the present case. 



It is anticipated that some paleontologists will contend that there is no justi- 

 fication for superseding the long-established and well-known name Allosaurus by 

 the earlier and little known Antrodemus, and that it would be better to either dis- 

 regard the law of priority, or else consider the type of Antrodemus as being an 



1 Hay, O. P., Proc. TJ. S. Nat. Mus., vol. 35, 1903, p. 355. 2 Dinosaurs of North America, 1896, pi. 11, fig. 2. 



