38 Lacerii(lR\ 



Nucras. As Lac erf a is prol)ahly derived from Nucras, we may imagine 

 some primitive form, no longer in existence, but possessing the generic 

 characters of the former, to have given rise to this genus, characterized 

 bv the lateral denticulation of the digits, as an adaptation to life on a 

 sandy soil. 



In the less specialized forms, constituting the first group of the 

 following classification, there are three series of scales round the 

 digits, as in some specimens of Lacerta {L. agilis, for instance) in 

 which the subdigital lamella3 are divided into two ; the serration of 

 the inner side of the digit, if present, is produced by the antero-outer 

 angle of the dorsal scale, whilst the outer denticulation or fringe, 

 sometimes very strongly develojjed, is formed by a distinct series of 

 scales, much narrower than the upper and lower. In the species of the 

 second group the serration on the inner side of the fingers is produced 

 by a series of scales similar to that of the outer side, there being four 

 series round the digit. This character was first pointed out by Lataste, 

 and Anderson has further dwelt on its importance for the definition of 

 species. 



The group with three series of scales round the fingers is un- 

 doubtedly the more primitive, the nearer to Lacerta. But among its 

 constituents there is not one that can be regarded as the most 

 generalized in all respects, as the survivor of the species from which 

 the others were derived. Looking upon the feeble denticulation of the 

 toes combined with the large size and small number of ventral plates 

 as primitive, we find A. vulgaris and^l. triKtranii to realize the original 

 condition, whilst on the other hand, the disintegration, in these two 

 species, of the first and fourth supraocular shields points to a derivation 

 from a form like A. schreiberi. A combination of the characters of 

 A. tristnnui and A. schreiberi would produce the ideal prototype of 

 Acanthodactylus; and it is interesting to find, in connexion with the 

 question of the original home of the genus, that both these species are 

 Syrian. It is remarkable that the habitat of the nearest ally of 

 A. tristrami, A. vulgaris, should be so remote, at the westernmost 

 extremity of the range of the genus. I have no better explanation to 

 offer for the fact than extinction of the original forms in the intervening 

 area, but it is not without parallel in the distribution of animals and 

 plants* ; other instances are well known to all herpetologists.f 



* For a list of plants, cf. Engler, Versuch einer Eiitwiekeluno-sgeschichte der 

 Pflanzenwelt, i, p. 53 (1879). 



t The genera Clemmys, Blanus, Pelodyles, Molge derjuyini and ilf. as/ieca, 

 Pelobates syriants and P. cultripes ; above all the discontinuous range of 

 Testudo ibera and Vipera lebetina are striking examples. 



