350 



also in Staurotypus, while L. Agassiz (Contributions to the Nat. Hist, of 

 the U. S. A., vol. I, p. 267) states that it disappears in old specimens of 

 other Cinosternidse. With these exceptions the entoplastron occurs as a 

 single median bone in all known species of turtles and tortoises both liv- 

 ing and fossil, save where in some of the latter the fragments are too 

 meagre to permit its presence or absence being positively determined. It 

 is therefore phjiogenetically a very old element in the testudinate skeleton, 

 and was probably, in some form or other, a direct inheritance from the 

 more generalized reptilian stock' from which this order arose. 



It follows, therefore, that we have in the paired entoplastron of the 

 embryo Trionyx, a very primitive character, so primitive, indeed, that it 

 occurs nowhere in the adult of any known species of Testudinata either 

 living or fossil. It is therefore an indication that Trionyx is to be re- 

 garded as more primitive than any other known genus of the order. Were 

 this the only evidence of primitiveness k >wn to occur in Trionyx, one 

 would not, perhaps, be justified in making so broad an assertion. But a 

 considerable amount of corroborative evidence is also at hand. Thus in 

 Trionyx, the atlas is temnospondylous, i. e., its three constituent parts, 

 the neural arch, the centrum, and the intercentrum, are not ankylosed but 

 remain loosely connected, there is no odontoid process on the second ver- 

 tebra, the first centrum being freely movable on the second; the pubic 

 and ischiadic symphyses are broad and are connected with each other by 

 a longitudinal cartilaginous band, which is replaced in other testudinates, 

 except Ghelone, by a broad completely" ossified plate (Gadow). In the 

 young of all tortoises, but in the adult only in the Cheloniche and Tri- 

 onychida?, the plastral plates are separated by large fontanelles (Fig. 1, f). 

 And finally, as reported by Wiedersheim (Vergl. Anat. der Wirbelthiere, 

 5. Auflage, 1902) teeth rudiments also occur in the embryo of Trionyx and 

 nowhere else among the Testudinata. I have not been able so far to cor- 

 roborate this observation, but it is certainly, if correct, a most important 

 argument in favor of the view herein set forth. 



This conclusion regarding the Trionychia is not invalidated by certain 

 secondary specializations, such as the flatness of the body and the webbed 

 feet, all clearly adaptations to an aquatic habitat. However, these adapta- 

 tions do show that Trionyx is in no sense directly ancestral to the other 

 Testudinata ; the Trionychia are to be regarded as an early offshoot of the 

 main stem, which has retained certain of its primitive characters. 



State University of Oklahoma. 

 Norman, Oklahoma. 



