349 
Accepting merely for the moment the correctness of this homology, it 
is interesting to note how very rarely a paired interclavicle has been found 
in reptiles. So far as I have been able to discover Parker is the only one 
heretofore to report such an observation, and in his monograph on the 
structure and development of the shoulder girdle and sternum, cited 
above, describes the interclavicle of Anguis as developing from paired ele- 
ments and says: 
“Above the Ganoid Fishes, this is the only instance I can give at 
present of the primordial symmetry of the interclavicle; but a careful 
study of the development of this bone in embryo lizards would, very prob- 
ably, show it to be not at all rare” (p. 99). 
Wxamining the question of the homologies of the plastral elements a 
little more closely, however, one is led to doubt Huxley and Parker’s iden- 
tification of the epiplastra and entoplastron as clavicles and interclavicle 
respectively. In all other reptiles so far as known the clavicle is laid down, 
at least partially, in cartilage and in close connection with the other ele- 
ments of the shoulder girdle. Even in the mammals, while its origin is 
still a matter for further investigation, it is definitely established that a 
portion at least of the clavicle is preformed in cartilage. In Trionya, as 
in other of the Testudinata, the epiplastra, on the contrary, develop en- 
tirely without connection with the shoulder girdle, entirely outside the 
muscular layer of the body wall and within the much thickened dermis. 
They, in company with all the other plastral elements, are wholly without 
a cartilaginous preformation, and develop as direct ossifications in the 
dermal mesenchyme. Without further evidence it is very difficult to ac- 
cept the view that the epiplastra are the testudinate homologues of the 
clavicles and the same arguments hold in regard to the identification of 
the entoplastron with the interclavicle. As is shown in this paper, the 
entoplastron in 7'rionyx is at first a pair of elements, so that there is noth- 
ing to prevent the interpretation of the entire series of plastral bones as 
the homologues of the so-called abdominal ribs so well known in Sphenodon 
and the Crocodilia. 
Recurring to the question of the relative rank of the Trionychia among 
the Testudinata, the paired condition of the entoplastron, as it exists in 
this embryo (Fig. 2, ent) is especially important and instructive. As 
Rathke first pointed out, the entoplastron is wanting in Sphargis, perhaps 
on the whole the most specialized of all the Testudinata. It is reported 
by Stannius (Handbuch der Anatomie der Wirbeltiere, 1854) as absent 
