482 



rough with pyramidal eminences % The period 

 at which the large arrau tortoise lays it's eggs 



* I would propose to place them provisionally near the 

 matamata of Bruguieres, or testudo fimbriata of Gmelin 

 (Schoepf, tab. 21), which Mr. Dumeril has taken to form his 

 genus chelys. 



Testudo arrau, testa ovali subconvexa, ex griseo nigrescenti, 

 subtus lutea, scutellis disci 5, lateralibus 8, marginalibus 24, 

 omnibus planis (nec mucronato-conicis), pedibus luteis, mento 

 et gutture subtus biappendiculatis. 



Testudo terekay, testa ovali, atro-viridi, scutellis disci 3, 

 lateralibus 10, marginalibus 24, capitis vertice maculis duabus 

 ex rubro flavescentibus notato, gutture lustescenti, appendi- 

 culo spinoso. 



These descriptions are far from being complete, but are the 

 first which have been attempted of two chelonians, so long 

 celebrated from the narratives of the missionaries, and so re- 

 markable fbr the advantages derived from them by the na- 

 tives. Among the animals contained in the collection of the 

 Jardin du Roi, it is observable, that in the testudo fimbriata 

 (with twenty-five marginal scales) the aperture of the anus is 

 placed nearly in the same manner as in the two tortoises of 

 the Qroonoko, of which I have here given the description, 

 and in the tryonix aegyptiac, that is to say, at one fourth 

 from the extremity of the tail. This position of the anus de- 

 serves to fix the attention of zoologists : it, as well as the ex- 

 istence of an elongated proboscis in the matamata, approxi- 

 mates the chelides to the tryonix ; but these genera differ in 

 the number of their claws, and the consistence of their shell, 

 Mr. Geoffroy, guided by other considerations, had already 

 supposed the existence of these relations. (Ann, du Museum, 

 vol. xiv, p. 19.) The anus in the chelonians, the land-tor- 

 toises, and the real emydes, is placed at the base of the tail. 

 I find described in my journal only very young a^rraus, I 



