146 BULLETIN 110, UNITED STATES NATIONAL MUSEUM. 



Upper Cretaceous genera, where it covers almost the entire anterior face of the distal 

 end of the tibia. 



Earlier in the present paper I have attempted to show that a complete section 

 across the abdominal curiass of Antrodemus consists of at least seven and probably 

 eight or nine distinct ossifications. 



In this connection it is of interest to know that in the adult Sphenodon a complete 

 section of the abdominal basket is usually composed of three parts — a median and two 

 lateral bones — one at either end, but it has also been shown * that in the embryo 

 Sphenodon each of these three sections of the adult have a compound origin for all 

 of the parts, median and paired alike. This fact explains the presence here and there 

 of transverse joints, as observed by Giinther and as are to be seen in a skeleton of 

 Sphenodon punctatum (No. 29516, U.S.N.M.) now before me. 



Howes and Swinnerton 2 in the paper cited above point out that one of the most 

 distinctive characters of Credner's Palaeohatteria of the Permian is the multiseg- 

 mentated condition of its "abdominal ribs," which, on careful examination of the 

 originals, he has shown to be also the condition in Von Meyer's famous Proterosaurus 

 and other genera to which he refers. In view of the obvious similarity of the embryo 

 Sphenodon, it suggests to the above authorities "that the 'plastron' may be under- 

 going reduction in the living genus. * * * Moreover, the multisegmented con- 

 dition of the 'plastron' segments is nothing short of a Stegocephalian character." 



Applying the above conclusions to the known American Theropods we find 

 the very interesting suggestion that perhaps one of the progressive changes in the 

 skeletal structure of the carnivorous dinosauria of successive geological formations is 

 a reduction in number of the segments going to make up the abdominal ribs, for we 

 find in Antrodemus, from the Morrison formation of the Jurassic, that each section 

 of the curiass is composed of seven or more distinct ossifications, that Gorgosaurus , 

 from the Belly River formation, of the Upper Cretaceous, has four, and that Tyran- 

 nosaurus, of the Lance formation, the latest of the Upper Cretaceous series, have 

 been reduced to two. 



Our knowledge of the abdominal structures of the Theropod dinosauria at 

 this time is too deficient to determine the merit of the above suggestion; but there 

 is, it appears to me, enough evidence to justify one in such an assumption. 



In the pelvis and sacrum the progressive modifications noted at this time are: 

 an increase in the number of coossified vertebrae in the sacrum from three to five, 

 a lengthening of the ilium, and the development of a triangular foot-like enlarge- 

 ment of the distal ends of the pubes. The modifications noted above were largely 

 developed during the early part of the Mesozoic, for the Upper Cretaceous carnivores 

 show no especial change since Morrison time in the above respects. 



In the pubes we also find that the Triassic forms had a closed obturator fora- 

 men, while in the Jurassic, with the exception of Ceratosaurus and possibly Megalo- 

 saurus, all others, including the Upper Cretaceous Theropods, have a notch which 

 probably functioned in the same capacity. 



In this brief review of the observed structural modifications of the Theropod 

 skeleton I feel that only a small number of the changes to be found have been 



i Howes and Swinnerton, Trans. Zool. Soc. London, vol. 16, 1901, pp. 35-36, pi. 2, figs. 8, 9, 10. 

 « Idem., p. 36. 



