THE DlSTRIIiUTION AND CHARACTERS OF REPTH^ES. 243 



within the genus Ayama or Mabuia: Agama atra and A. acu- 

 leata both occur at Kimljerley. hut they do not meet, as the former 

 lives entirely amongst rocks on kopjes, whilst the latter inhabits 

 the open veld or climbs up trees {fide J. H. Power) : exactly 

 similar differences of habit separate the skink Mabuia 

 sulcata from its near ally, j\'I. striata, and they also occur near 

 Kimberley. However, whilst admitting the scantiness of our 

 knowledge with regard to the habits of some of our commonest 

 reptiles and ibatrachians, I think it very improbable in the cases 

 above mentioned, that the species are separated by habits which 

 effect rigid isolation in a topographical sense. Indeed, the most 

 likely explanation of the occurrence together of closely related 

 species in a series of widely separated localities, is that they share 

 a preference for the particular kind of en\'ironment which obtains 

 in those localities. 



Secondly, how can we explain adaptations in terms of the 

 isolation theory? As stated later on, the structural specific charac- 

 ters are not as a rule obviously adaptative, but sometimes they are 

 almost certainly so. How is it possible to suppose that isolation 

 alone, unaided by natural selection or Lamarckian factors can 

 direct the chance variations to which a rounded snout is suscep- 

 tible into a sharp cutting snout, such as is found in various bur- 

 rowing species of reptiles found in the Kalahari? And yet it is 

 such differences that separate the Kalahari species from their 

 allies elsewhere. Again, the lizard genus Eremias comprises many 

 species which inhabit the drier parts of Africa and Asia. Some of 

 them have transparent scales in the lower eyelid ; others have the 

 lower eyelid entirely granular and opaque, which no doubt repre- 

 sents the primitive condition, whilst between the two extremes 

 are other species which collectively show a complete range of 

 intermediate conditions. The character of the lower eyelid is 

 very constant in any one species, and in some cases would alone 

 serve for the identification of the species. Now this transparency 

 of the lower eyelid seems to be an adaptation against the blinding 

 sandstorms of an- arid country, the lizard which is thus provided 

 being enabled to see with his eyelids closed. Similar adaptations 

 are found in other genera, li we accept that explanation, it will 

 be impossilble to explain this typical case of Orthogenesis, the evo- 

 lution along definite lines of the transparent " window-eyelid " 

 from the scaly opaque eyelid, as merely the result of isolation. 

 As a matter of fact, lineocellata which possesses a " window eye- 

 lid " occurs in the same district as nantaquensis, in which the 

 lower eyelid is scaly and only semi-transparent, and indeed the 

 South African species of this genus apparently do not as a whole 

 distribute themselves into (juite distinct areas. Nevertheless, 

 though in such cases topographical or geographical isolation 

 seems to have in no way contributed to the splitting of this genus, 

 yet Eremias has one pair of closely related species, z/i.::., E. lineo- 

 cellata and E. pulchella, in the formation of which geographical 



