510 RECORD OF SCIENCE FOR 1886. 



"rugose;" the nostrils are "anterior," and the "jaws naked, with sharp 

 cutting edges, the lower curved," and not notched. 



What is especially remarkal)le, however, is the character and form of 

 the members. The arms are "elongate, narrow, compressed, ridged 

 on their rounded anterior jiortion, with narrow plates," and the nails 

 are " free on first two digits onlj^," the rest being " without nails and 

 strongly webbed, the tips flattened," and " the third the longest." The 

 hind legs are " short," and the first two toes have " strong, sharp nails," 

 which nails are alone free, the rest of the toes being " strongly webbed 

 to the tips." 



The tortoise thus distinguished was obtained in the Fly Eiver, and 

 wa»a female of large size, the total horizontal length of the carapace 

 being ^8 inches, and along the curve of the back 19 inches, while the 

 head and neck were about 7 inches long. 



From the paragraphs of the description thus selected, it is not evident 

 why the new tortoise should have been referred to the family of Triony- 

 chidse. Apparently the only reason was because the scutes were not 

 developed, and consequently a rudimentary or tense skin alone invested 

 the bones. The differences manifested from all the TrionychidoB, how- 

 ever, are very great, and in the classification lately proposed or adopted 

 by Professors Cope and Dollo it would belong to a diflerent section of 

 the order. In fact, assumiag that the tortoise is a cryptodire, according 

 to the classifications adopted by those gentlemen it would enter into 

 the group of Clidosterna or Clidoplastra, inastnuch as the plastron 

 unites " with the costal bones of the carapace by suture, with ascending 

 axillary and inguinal buttresses." 



But the assumption that the tortoise is a cryptodire is itself some- 

 what violent, and in view of the geographical distribution of the order 

 to which it belongs it may quite likely prove to be a cryptodire. 



But whatever may be the relations of the new genus, whether to the 

 cryptodirous or pleurodirous tortoises, it has many quite peculiar char- 

 acters. From all known forms it is apparently distinguished by the ab- 

 sence of scuta, the peculiar feet, and other characters. Undoubtedly, 

 therefore, the new genus does not belong to any of the established mod- 

 ern families, and apparently not to any of the extinct ones named, al- 

 though when more is known of Carettochelys, as well as the extinct forms, 

 it may turn out that the Papuan animal is related to one of the fami- 

 lies now regarded as extinct. Meanwhile, it will be better to isolate 

 the genus, and regard it as the type of a peculiar family. This family 

 of Carettochelyidne has a "clidosternal" shell destitute of scuta; the 

 carapace has six vertebral plates separated from each other by inter- 

 vening costals, eight costal plates, the last two of which are connected 

 mesially by the entire borders (no vertebral plates intervening), and 

 ten pairs of marginal plates, as well as a nuchal and a pygal. The 

 plastron is composed of nine plates, completely ossified and leaving no 

 fontanelle, an " interclavicle" or mesosternal being well developed. The 



