18/3.] DR. J. E. GRAY ON THE TRIONYCHIDJE. 39 



of specimens in this state only affords a general impression of the 

 alveolar surface. 



The skulls of the Mud-tortoises are so uncommon, and I was so 

 disinclined to take the skulls out of the specimens, that I availed 

 myself of the characters which the preserved heads afforded me : 

 but I have now determined to extract the skulls in the most careful 

 manner from the specimens themselves ; and this has given me a 

 greatly increased knowledge of the species, and of the characters the 

 skulls afford. 



The examination of the skulls of the different specimens has had 

 the effect of putting together specimens that had been considered not 

 only distinct species but distinct genera, and has shown them to be 

 only various ages of the same species, as Dogania and Sarbieria ; 

 and at the same time it has given more important characters for the 

 separation of species and characteristics for the groups to which they 

 belong. 



In all the other groups of Tortoises I have only figured the skulls 

 that belonged to skeletons ; but in this group of Mud-tortoises the 

 number of skulls that I had was so small, that I was induced to 

 figure and try to identify two skulls which Prof. Oldham had given 

 to the Museum with two Asiatic species : and I am sorry to say that 

 now we have been able to examine other skulls, one of them has 

 been proved to have been wrongly identified ; that is to say, that 

 which I figured for Potamochelys proves to be the skull of an 

 Emyda. 



In my former papers I had only the opportunity of examining the 

 skulls of most of the species in the heads of the stuffed specimens, 

 or of those preserved in spirits, and which consequently had the 

 horny coat to the alveolar process of the jaws and palate, which are 

 naturally very different from the bones which they cover. All the 

 skulls described in this paper have been extracted from the specimens, 

 and have had the horny coat of the alveolar surface removed. This 

 explains why they differ from descriptions in former papers ; and 

 those in this paper are to be considered the most correct. 



There is a very great difficulty in comparing animals in spirits with 

 prepared dry specimens. The living animals and the specimens in 

 spirit have the bony disk of the back and sternum covered with a 

 thick skin, which, perhaps with the exception of very old specimens, 

 entirely hides the callosities on the surface of the back and part of 

 the sternum, which are so prominent, and from which we take many 

 characters in the dry preserved and stuffed specimens. The rugosity 

 or callosities seem to cover the lateral bones of the sternum simul- 

 taneously over the whole surface, except the diverging rays by which 

 the bones are united. In the hinder pair of sternal bones the callo- 

 sities form a rounded or oblong spot near the internal side, and 

 gradually enlarge themselves so as to cover the whole surface of the 

 bone, leaving the diverging rays. 



The genera and families have various relations to each other, 

 which I think are well exhibited in the following Table, which shows 

 in one view the alliance of Heptathyra to Cyclanosleus, and Chitru 



