Geometric A- Group of S.A. Tortoises. 193 



are present at one stage or another of their elaboration throughout 

 the whole of the geometrica-group. However pronounced a character 

 may be, it is shown to be but a modification of one occurring in a 

 simpler form elsewhere. Transformation has consisted in the elabora- 

 tion of characters traceable throughout. With the exceptions to be 

 mentioned below, there are no new features anywhere introduced ; the 

 differences are simply so many variations of characters common to 

 all, they are quantitative rather than qualitative. Such a homogeneity 

 serves to strengthen the idea of the close relationship of all the 

 members, and of their common genetic source. 



The femoral tubercle and its surrounding supplemental shields 

 may in some ways be regarded as a new character introduced within 

 the limits of the group, though found in some other species of 

 Testudo and other genera of the family Testudinece. The bony 

 process is altogether wanting in some members of the tentoria, 

 geometrica, and trimeni sub-groups, is but feebly developed in other 

 examples of the same sub-groups, and in certain tentoria becomes very 

 large and surrounded by additional shields. In the verreauxii sub- 

 group it is always strongly developed, and usually accompanied by 

 enlarged accessory shields, either all the way round or at only one or 

 a few places. In the ocidifera sub-group the bony tubercle remains, 

 but the supplementary shields are usually wanting. 



It would seem therefore that the tubercle on the back of the 

 thigh is a character in process of either introduction or elimination 

 within the geometrica-'gxovip. On the assumption that tentoria is 

 nearest the original type, the character represents a new introduction, 

 for it may be absent, incipient, or fully developed within the limits 

 of the sub-group. When we reach the verreauxii sub-group it is a 

 permanently fixed structure, appearing in every member, and 

 developed to about the same degree, but the directive pro- 

 cess has extended further and brought about a change in 

 the minute skin shields around the tubercle ; these have been 

 enlarged, and evidently serve the same purpose as the principal 

 spine. We may consider that the femoral tubercle has reached its 

 maximum of possible development consonant with its usefulness, and 

 any further modification along the same lines is possible only by 

 the bringing in of more shields. But the method of carrying out this 

 last transformation is not constant, for the shields in some form a 

 complete circle around the primary, in others there are two or three, 

 and in others only one. 



In ocidifera the tendency towards the elaboration of the femoral 

 shields is not so strong as in tentoria. The principal tubercle is as 

 large as in tentoria, but the accessory shields are either absent or 

 developed to a much less degree. In many ways it has been shown 

 that the tentoria, verreauxii, and oculifera sub-groups represent a 

 gradational series, and the conception is strengthened by the condi- 

 tions of development of the femoral tubercles throughout the series. 

 We have a homogeneous group within which a new structure appears, 

 reaches its climax, and then begins to retrograde. As regards most 

 of the other characters, they have been found to be incipient in 



13 



