156—2 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA. 
elements of the underlying bony framework. The names applied to the individual 
plates are those employed in Gunther’s “ Reptiles of British India, with the 
exception that the term ‘ pygal ’ is substituted for ‘ caudal.’ The constituent scutes 
of the plastron are respectively termed ejDiplastron, xiphiplastron, etc., instead of 
episternum, xiphisternum, etc., as they are often called. The divisions sej)arating 
both the plates and the scutes are termed sutures. 
Difficulty of the subject. — As many specimens of the shells of fossil tortoises 
exhibit only the boundaries of the bony scutes and not those of the epidermal plates, 
and as in recent species as a rule the former are not described and are invisible 
without the removal of the overlying plates, it is in such cases almost impossible to 
determine the affinities of the fossils.^ Owing to this difficulty the writer has 
considered it inadvisable to describe -certain Siwalik specimens in which the 
boundaries of the plates are not visible. 
The great similarity in the shells of many existing emydines, and the doubt 
still existing as to the number of species, likewise renders the specific determination 
of their fossil allies a matter of extreme difficulty, and specific names have therefore 
not been applied in several instances : for the same reason in the majority of cases 
only fairly perfect examples of such forms have been described at all. In the case 
of the larger land-tortoises, where only fragments of the shell are forthcoming, the 
epiplastral element, as the one most frequently preserved, has been selected as the 
one on which to found sjDecific distinctions. It is of course obvious that unless some 
such restriction be observed several species might easily be made from the remains 
of one and the same form. 
Delations of the fossil and existing Indian tortoises. — As far as our present very 
imperfect knowledge admits of generalization it appears that the pleistocene 
tortoises of the Narbadas all belong to existing Indian species. A considerable 
proportion of the medium and smaller sized forms of the pliocene Siwaliks appear 
closely allied to existing species now inhabiting India and the neighbouring regions, 
although in most instances they present more or less well-marked differences, which 
are in many cases regarded as of specific value. It is, however, very frequently a 
matter of extreme difficulty to decide whether the fossil forms should not rather be 
looked upon in the light of varieties of the existing species ; — the course taken really 
depending to a considerable extent on the views of the individual describer as to 
what are specific and what are merely varietal differences. Some of the groups, 
however, _ like that of Glemmys crassicollis and the genus Emyda were much more 
strongly represented in the pliocene than in recent times. With reghrd to the larger 
land-tortoises the case is different ; there being no question as to the specific 
distinctness of' the large number of forms by which this group is represented from 
any existing Indian species. The nearest ally of this group is, however, apparently 
to be found in the existing Oriental Manouria emys, although there are strong 
1 Page 2. 
2 The writer is indebted to Mr. Theobald for numerous specimens of the shells of recent Indian tortoises, which 
have much facilitated his comparisons with their fossil allies. 
