162—8 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA. 
in the Siwalik fossil and T. ponderosa the impression in front of the occipital condyle 
is wide, shallow, and horse-slioe-shaped. It is unfortunate tliat the characteristic 
prefrontal of the fossil is broken ; hut enough remains to indicate that the nasal 
opening was probably higher than wide, and that the front margin of the premaxillm 
projected so far in advance of the prefrontals as to cause the nasal opening to slope 
downwards as in the Aldabra tortoises ; the upper surface of the premaxillse being 
highly inclined as in those forms, instead of nearly horizontal as in the other recent 
gigantic forms. The orbit is jjlaced relatively low down in the fossil, and there is a 
rudiment of a frontal convexity ; although this makes no approach to that obtaining 
in the' Aldabra forms. The postorbital bar of the latter is much narrower than in 
the Siwalik specimen. In the general contour of tlie palatal aspect the fossil is very 
like that of T. ponderosa^^ and narrower than T. elephantinad The following table 
shows the dimensions of the fossil and the male crania of T. ponderosa and T. 
elephantina figured by Dr. GiintheP : — 
T. ponderosa. T. elephantina. Fossil. 
Length from hinder border of quadrate to extremity of premaxilla 3'0 41 6'8 
Interval between external surfaces of the quadrates . . . 2'8 3-9 6'3 
The cranium of T. elephantina measures 5 '5 inches from the occipital condyle to the 
premaxilla, and the corresponding length of the fossil would be 9T inches. As 
another male cranium of T. elephantina has a basal length of 6 inches, the fossil 
indicates an animal just one-and-a-half times the size of the largest males of the 
living species ; and as the carapace of the largest male of that species measures 
4 feet 1 inch in a straight line,^ the calculated length of the carapace of the individual 
to which the fossil cranium belonged would be 6 feet ; — or identical with the size 
calculated for the smaller (? female) form of the present species from the size of the 
smaller e|)iplastron. There is therefore no reason as regards size why the present 
specimen should not have belonged to a female of the species under consideration, 
although it cannot be affirmed that it may not have belonged to the species next 
described. It is probably too large to have belonged to either of the Punjab species 
of large tortoises described below, whose remains are not recorded from the Siwalik 
Hills. The calculated basal length of the cranium of the larger individuals of the 
present species would be twice the size of the cranium of T. elephantina^ or 12 inches. 
Affinities . — The foregoing observations indicate that the present species was a 
land-tortoise distinguished from all living forms except the Oriental Manouria emys 
by the non-union of the pygal plates. Assuming the right of that species to generic 
distinction,'" it will be evident that the present form cannot be referred to Testudo ; 
and since the relations of the pectoral plates of the males of the Siwalik species to 
one another are unknown, and the structure of the epiplastron is so very different 
from that of Manouria e^nys, it seems inadvisable to refer the fossil species to that 
1 “ Gigantic Land-Tortoises,” pis. VIII., IX. 2 Ibid, pi. IX. 
3 Ibid, pi. VIII. 4 Ibid, p. 25. 
5 The generic characters of Manouria are the separation of the pectoral plates of the male, the non-union of the pygals, 
and the partially webbed character of the hind feet ; the latter character indicating subaquatic habits. 
