466 FOSSIL TURTLES OF NORTH AMERICA. 



the Trionychoidea as one of three suborders of the Testudines, the other two being the Athecae 

 and the Thecophbra. Further consideration of the subject has convinct him that the group 

 is too closely related to the Cryptodira to justify such a wide separation of the two. On the 

 other hand, there are structural differences of too great value to permit the trionychids to be 

 reduced to the status of one or two families under the Cryptodira. It is now believed that their 

 relationships to other turtles will be correctly exprest by making them a superfamily of the 

 Thecophora, the other superfamilies being the Amphichelydia, the Cryptodira, and the 

 Pleurodira. This classification is that adopted by Dr. Boulenger in his Catalogue of Chelon- 

 ians, etc., in the British Museum. 



For a discussion of the structure and relationships of the trionychids the reader is referred 

 to Baur's papers in the Zoologischer Anzeiger, vol. x, 1887, p. 96; vol. xi, 1888, p. 736; vol. XII, 

 1889, pp. 241-243, and American Naturalist, vol. XXIV, 1870, p. 530; Dr. Boulenger's catalog 

 referred to above, and to a paper by the present writer (Bull. Amer. Mus. Nat. Hist., xi, 

 1905, pp. 137 175)- For a discussion of Haeckel's view that the Trionvchoidea are the most 

 primitive of the Thecophora the reader is referred to page 25 of the present work. 



The geographical distribution of the living Trionvchidae is discust on page 35 and is 

 illustrated by fig. 16. 



Family PLASTOMENIDAE Hay. 



Trionychoidea with skull like that of the Trionvchidae. Neck unknown, but probably 

 like that of the Trionychidae. No peripheral bones. Fpiplastra separated from the hyoplastra 

 by the large, crescentic entoplastron. Hvoplastra, hypoplastra, and xiphiplastra closely united, 

 as in the Emydidae. Feet unknown. 



Up to the present time only one genus is known which may be referred to this family 

 the genus Plastomenus Cope. The structure of this, so far as determined, has been learned 

 from the Bridger species. The cervical vertebras and the feet are unknown. 



The Plastomenidae differ from the Trionvchidae, so far as known, principally in the 

 structure of the plastron. In the latter family the ventral armor differs much from that of the 

 Cryptodira. The bones have suffered reduction, so that great fontanels intervene between the 

 bones of the opposite sides and between the entoplastron and the hypoplastra. The ento- 

 plastron is V-shaped and excludes the epiplastra from the hypoplastra. In the Plastomenidae 

 the three hinder pairs of bones join their fellows closely at the midline, as in the Cryptodira, 

 and the hypoplastra are closely sutured to the hyoplastra and to the xiphiplastra. There 

 appears to have been little space between the entoplastron and the hypoplastra. Indeed, the 

 plastron is more completely closed than in the Chelydridae and Cheloniidae. Furthermore, 

 the entoplastron, if we may form our conclusion from Plastomenus cedemius, was broad fore 

 and aft and elongate kidney-shaped. We do not know what were the form and connections of 

 the epiplastra. 



A peculiarity of structure displayed by some species of the family is the great development 

 of the eighth pair of costals. In most species of Trionychidae the eighth costals are greatly 

 reduced and in some cases wholly wanting. In Plastomenus thomasi and some other species 

 this pair ot costals is greatly developt. The character does not, however, appear to be shared 

 by all the species. 



Genus PLASTOMENUS Cope. 



\r tin present time the generic characters of Plastomenus can not be separated from those 

 assigned above to the family. So far as known there is present a preneural bone, and some of 

 the hinder neurals are wanting. 



1 ype: Plastomenus thomasi Cope. 



Professor Cope's first definition of the genus Plastomenus (Proc. Acad. Nat. Sci. Phila. 

 1873, p. 278) is hardly sufficient to distinguish it from other trionychids. For a reason which 

 the present writer does not understand he concluded that there was present "a nuchal marginal 

 which does not exist in Trionyx" but no such bone has been shown to have existed. He was 



