Geometrica-Group of S.A. Tortoises. 189 



distinction holds for the few specimens I have examined, but there is 

 little doubt that if more were available some would be found with 

 the tubercle developed to a greater or less degree. Verreauxii was 

 first described by Smith without mentioning the femoral tubercle, and 

 this was taken by Boulenger (1886, p. 541) to imply the absence of 

 such a structure. However, among the many individuals of the 

 verreauxii type, obtained from the same locality as Smith's (sources 

 of the Orange River), I find practically every specimen is provided 

 with an enlarged tubercle, so there can be little doubt that Smith's 

 verreauxii had enlarged femoral tubercles. Siebenrock (1904, p. 313) 

 also states that a form he identifies as verreauxii has very large 

 femoral shields. 



In trimeui the femoral tubercle is altogether absent in some 

 specimens, as stated by Boulenger, yet it is present in others, though 

 never enlarged as much as in tentoria and oculifera. The conditions 

 in tentoria are significant as regards the fluctuating nature of the 

 shield. In some individuals it is wholly absent, in others it is moder- 

 ately developed, while in others again it is strongly developed, and 

 may even be buttressed by smaller ones. The tubercle reaches its 

 maximum in verreauxii, where it is always a very prominent, conical 

 shield, often with a number of much enlarged shields around it, or 

 sometimes with only another of equal or nearly equal size. Also in 

 oculifera the femoral tubercle is always a strong conical projection, 

 but is rarely buttressed by smaller shields. 



We may summarize by saying that the femoral tubercle is either 

 absent or only feebly developed in geometric a and trimeni ; that all 

 stages from its absence to full development occur in tentoria; that 

 it reaches its extreme development in verreauxii, where it is often 

 supported by other smaller shields ; and that it exists as a single, 

 much enlarged shield in oculifera. It is a fluctuating character in 

 the first three groups, but constant in the two latter groups, and there 

 is a direct line of variations from one extreme of its development to 

 the other. 



i. Form of Shields of Carapace and Plastron. 



Much diversity in shape is presented by the individual shields 

 of the carapace, that is, whether flat or elevated in a conical manner ; 

 and by this character the general appearance of the shell is largely 

 determined. The extreme conical condition is afforded by many 

 specimens of tentoria, where each shield is conspicuously elevated, 

 the neural members somewhat more than the costal ; the extreme of 

 flatness is to be found in the oculifera sub-group ; and between these 

 two every intermediate stage can be obtained. The knobbed charac- 

 ter can scarcely be regarded as constant for any of the species ; 

 even some specimens of tentoria have almost perfectly flat shields, 

 though coming from the same locality as others with strongly conical 

 shields, and, on the other hand, some individuals of oculifera have 

 the conical form rather well shown. 



The same can be said of the verreauxii sub-group. Individuals 

 can be selected in which the conical character is almost as marked 



