South African Tortoises. 409 



Aberdeen. None of the other representatives of the genus have 

 been received from these districts. 



Two of the specimens, one living and the other only a shell, 

 seem at first sight very distinct forms, but the differences can 

 evidently be regarded as expressions of senility. The coloration s 

 a uniform pale yellow in the old shell, and a blackish brown in 

 the living example. The concentric lines on the shields 

 have almost disappeared, and the areolae are scarcely recognizable, 

 while their margins are much swollen or nodulated, giving a very 

 iri'egular appearance to the upper surface of the carapace ; the 

 grooves between the shields are deeper but narrower, and the nuchal 

 is either nearly square or much retluced ; the marginals united 

 with the bridge are rounded, having lost their feeble angularity. 

 The sutures between the plastral shields are deeper and the surface 

 of the plastron is pitted. 



The scales or tubercles on the fore limbs are much shortened, 

 so that thej^ scarcely overlap ; the claws are also much shorter than 

 in ordinary specimens, and the tubercle on the thigh is much 

 reduced. 



In almost all species of tortoises somewhat similar evidences 

 of old age occur, but not often to so marked a degree as in the 

 present form. 



Considering the genus Honiopus as a whole, each species, as 

 shown in the synopsis, has clearly defined, distinctive characteris- 

 tics. Where many individuals are available for compari- 

 son (areolatus, femoral is^ hoiilengeri j certain variations occur 

 among the members of a species, yet these seem in no way sug- 

 gestive of relationship with the other species, except perhaps in 

 the case of the femoral tubercles in areolatus. In this species all 

 stages can be found from tubercles almost as large as those met 

 with in feinoralis to their total absence, and, as stated above, they 

 are wanting in one specimen oihoulengeri. In all the other specific 

 characters there is little or no evidence of intergrading. As will 

 be shown in a later paper this is in marked contrast with 

 the conditions prevailing in the (/eorrietrica-groi\y>, where 

 with a large number of specimens for study specific dis- 

 tinctions are found to be very difficult to establish. If we assume 



