194 G. R. Wieland — Upper Cretaceous Turtles. 



supported ; by the absence of pleuro-marginal fontanelles and 

 the nether articular process of the nuchal, and finally by its 

 mere Chelydra-like pelvis and earlier type of plastron. Doubt- 

 less there are also cranial differences. There is, therefore, 

 between Osteopygis and Chelone, when considered alone, a 

 very distinct structural interim such as might well characterize 

 two subfamilies. Lytoloma, however, is so exactly intermediate 

 that taken with other forms, existing and extinct, there exists 

 a nearly continuous morphological series, passing by simple 

 generic or even lesser stages from the most specialized existing 

 Cheloninse back to Osteopygis. In fact the greatest hiatus 

 remaining unbridged appears to be the ribless ninth or tenth 

 marginal, as seen in Lytoloma and existing forms. I am at a 

 loss to surmise if this condition was due to a slowly effected 

 change, or was suddenly developed after the appearance of 

 pleuro-marginal fontanelles in the early line, giving rise to the 

 two genetic groups mentioned. In either case it must have 

 been, as already hinted, correlated with femoral shortening 

 and the development of strong front flippers, and is scarcely to 

 be considered as of more than generic value. If a separation 

 into two subfamilies were made, it would have to • be based 

 mainly on this feature, as furnishing the only sharp distinction. 



The final conclusion must be that Osteopygis and Propleura 

 can be placed in a separate subfamily, the Proplenrinas, on 

 genetic grounds, but that further discovery may bring them very 

 near if not into the Cheloninge. Moreover, as has been seen, 

 these forms, though not apparently forming a closed series, 

 permit the statement that Osteopygis was no more than gener- 

 ically separated from some strongly web-footed littoral turtle, 

 which was the true ancestor of the existing Cheloninse. 



The facts given in the present and preceding papers on the 

 Upper Cretaceous turtles of New Jersey, as well as in my 

 paper on Toxochelys, indicate the manner in which the marine 

 turtles have been derived from generalized land forms, together 

 with their line of descent. Aside from the carpal and tarsal 

 changes involved and as yet but meagerly illustrated by fossil 

 forms, the most interesting future discoveries will be the 

 ancestral, littoral, and fluviatile Osteopygoid tortoises. 



Measurements of Lytoloma. 

 (A) The dentalium of Lytoloma angusta (Yale speci- 

 men No. 913.) Uncrushed. 

 Width (measured from outer extremities of the coro- 



noids) 7*5 



Distance (on median line) of hook from anterior ends 



of the coronoids „ .. 4*7 



Length of median symphysis 4*2 



Greatest vertical depth of median symphysis 1*1 



Cii) 



