358 DR. H. GABOW ON EVOLUTIOX [Mai*. 20, 



opposite to C. guttatus and to those specimens of C communis 

 copei and C. c. australis from Oolima, the Isthmus*, and Ouicatlan, 

 which are very similarly coloured. 



In fact C. bocourti is .structurally indistinguishahle from many 

 specimens of C. communis occidentalism and from some of the 

 C. mexiccmus of the Balsas basin. Fundamentally, the evolution 

 of its pattern is that of the former, but when most agreeing in 

 coloration with the cojisi or australis varieties it differs most 

 from these structurally ; or, vice versa, wdien sti-ucturally most 

 like mexicanus it is diametrically opposed to it in coloration. 

 Upon this ambiguity rests the best claim for separate recognition 

 of C. bocourti, which after all happens to be one of the most easily 

 recognised forms of the whole C. gularis group. 



Cnemidophorus mexicanus Peters. 

 (Text-figs. 69 ; 81 A, B, C, D, F ; 82 A-D, &c.) 



Diagnosis. — Large-sized C. gularis in which the original stripes 

 do not develop pale spots, but are bi-oken up by the encroaching 

 black of the fields and by the transversely combining brownish 

 field-spots, resulting eventually in a tiger-barred pattei-n. 



The most extreme develoj^ment is i-eached in Oaxaca ; this 

 variety I distinguish as var. tyjnca. They reach the largest size, 

 the tiger-pattern is most pronounced, but the collar and the 

 covering of the posterior side of the forearm are variable, inclining 

 more towards granules. 



Those of the Balsas Ilivei--basin are distinguished by a strong 

 collar, pre^'alence of scutes on the forearm, and far less pronounced, 

 more incipient tiger-pattern. They seem, moreover, as fits their 

 distribution, to pass into aberrant C. communis occidentalis. 

 These I refer to as C. mexicanus var. balsas. 



It is significant that these Oaxaca specimens exhibit the same 

 trend of variation away from their relations (decreasing collar 

 and more granular arm-scales and tendency to tiger-pattern) as do 

 the repi'esentatives of C. communis co2)ei in the State of Oaxaca 

 in the .shape of C c. australis. 



It is irony of fate that the three type-specimens of C. mexicanus 

 are all immature, and show but little of the tyjjical features. 



Range. — The temperate regions of the States of Oaxaca and 

 Guerrero, descending into the tropics of South Oaxaca and into 

 the tropical portion of the Balsas basin. 



S'u2Jraocidars 4, apparently without exception ; the posterior 

 separated from the parietal- plates by one row of three or four 

 elongated granules. 



Frenocular variable. 



Collar variable. It reaches its largest development in the var. 

 balsas, being composed of veiy large scales, one row of which 



* In my paper Proc. R. S. 1903, p. 118, I had referred to C. bocourti the hirge 

 specimen from San Domingo, now mentioned as C. communis copei, p. 350 ; and the 

 Naj-arete specimens now described on p. 343. 



