1913-] N. ANNANDAI.E : Tortoises of Chota NagpuY. 65 



so far as published statements go, that a valid difference existed 

 in the skull, viz. the presence or absence of a temporal arch. Two 

 factors are concerned in the formation of this arch, the ossifica- 

 tion of the quadrato-jugal and the production of a backwardl}^- 

 directed (postorbital) process of the orbital ring in the formation 

 of which both the postfrontal and the jugal bones take a part. 

 If the arch is complete it is formed by a dovetailing of the for- 

 ward process of the quadrato-jugal and the backward process of 

 these two bones. 



In the three species of Geocmyda {s.s.) {i.e. of Heosemys, 

 Stejneger) it is probable that the temporal arch is alwa3^s absent, 

 the quadrato-jugal being absent or vestigial and the postorbital pro- 

 cess of the postfrontal and jugal short and blunt In one skull of 

 Blyth's Geoemyda tricarinata (pi. vi, fig. 6h) from Assam the same con- 

 dition occurs. This skull was removed from the head of a specimen 

 preserved in spirit and only a very thin, almost membranous car- 

 tilage was found in the place of the quadrato-jugal. In two other 

 skulls of the same species, including that of the type from Chota 

 Nagpur (pi. vi, fig. 6a), there is a short postorbital process and 

 the quadrato-jugal, although distincth^ ossified, is ver}^ delicate and 

 slender and barely meets it. In many skulls of the typical form 

 and of both the races thermalis and edeniana of " Nicoria " trijuga 

 the arch, although quite complete, is extremely delicate, being at 

 any rate no thicker than the lower posterior part of the orbital 

 ring. But in some skulls of these races the arch is as stout as any 

 part of that ring : apparently it is always complete. It is also 

 complete in the type-skull of N . tripiga coronata, but its condition 

 much more closely resembles that found in the two skulls of 

 G. tricarinata in which it is ossified. In the two skulls of the new 

 species Geoemyda indopeninsiilaris as yet examined the postorbital 

 process is well developed; but there is no quadrato-jugal. There 

 is thus a distinct arch, but it is incomplete. 



These facts leave, in my opinion, no course but to amalga 

 mate the genera Nicoria and Geoemyda under the latter name ; for 

 I can find no real generic difference between the shells of species 

 assigned to one group or the other. 



A more difficult question is, Should Blyth's genus Chaibassia 

 be recognized as distinct ? The single species included in this 

 genus exhibits considerable variation in skull-characters, but 

 differs from all other Oriental species that can be assigned to 

 Geoemyda (except the recently described G. sylvatica) in the ves- 

 tigial nature of the digital webs and in the approximation of 

 the hind feet to the type characteristic of the genus Testudo and 

 its allies: it and G. sylvatica are true land-tortoises, whereas the 

 other Asiatic species are amphibious. The American forms, 

 however, most closely related to the Oriental " Nicoriae^' show 

 great specific variation in respect to the webbing of the finger? 

 and toes. Until it has been possible to compare large series of 

 specimens from the two continents, I think, therefore, that it is as 

 well not to revive Chaibassia as a distinct genas. 



