ART. 25 CHINESE AMPHIBIANS AND REPTILES STEJNEGER 95 



snakes, particularly A. halys and A. hlo'mho-ffi^ the latter being the 

 only form occurring in the territory treated of, I expressly stated 

 that " to what extent the turning up of the end of the snout may 

 serve in all instances as a character to separate A. halys I can not 

 say for lack of material, and for that reason I shall at present treat 

 the latter as a good species." 



The doubts as to the specific distinctness of these forms, first 

 hesitatingly expressed by Guenther in 1896, have since been justified 

 by the investigations of Bedriaga, 1912, and of Nikolski, 1916, who 

 have had access to an unsurpassed material of Central and East 

 Asiatic specimens.*'^ Bedriaga,''* particularly, demonstrated the 

 intergradation between A. halys and A. iiiterinedius^ though as a 

 binominalist he treats them nomenclatorially as species, but as I had 

 already (1907) shoAvn the intergradation between A. interniedius 

 and hlomhoffi, Xikolski who on the contrary is a thoroughgoing 

 trinominalist, accepted the nomenclatorial consequences and enum- 

 erated the various forms, including a new one described by him, 

 as A. halys halys^ A. halys caucasicus, A. halys interniedius^ A. 

 halys hrevicaudus, and A. halys hlomho-ffi:^^ Both Bedriaga and 

 Nikolski tried to introduce new criteria for the discrimination of 

 these forms, the former mentioning the width of the rostral at the 

 apex, the latter the relative width of the anterior and posterior 

 nasals, which, when other characters fail, may be of assistance in 

 dubious cases. Bedriaga also described a new species from wes- 

 tern China as A. strauchi and essayed the following key (pp. 732- 

 733): 



a\ Large posterior supralabials ; height of fifth supralabial equals length 

 of free edge of third supralabial ; rostral somewhat turned over 



above; canthus rostralis not marked A. strauchi. 



«^ Small or medium posterior supralabials ; height of fifth supralabial less 

 than the length of the free edge of third supralabial; rostral not 

 turned over onto the upper surface of head ; canthus rostralis dis- 

 tinctly or sharply prominent. 

 &\ Width of upper, strongly narrowed part of rostral, measured at the 

 level of the suture between inteniasals and nasals, equals half the 

 length of suture between anterior nasal and rostral A. halys. 



83 Nikolski, for instance, had 173 specimens of A. intermedins and 48 of A. halys. 



8* Wiss. Res. Przewalski Central-Asien Reis., Zool., vol. 3, sect. 1, pt. 4, 1912. pp. 719- 

 726. 



«6 During the same year, 1916, Dr. J. C. Thompson (Trans. San Diego Soc. Nat. Hist., 

 vol. 2, no. 2, 1916, pp. 61-76) attempted by the statistical method to reduce these va- 

 rious forms to synonyms of A. halps, but by bunching his figures under such geographic 

 headings as Korea, China, mainland specimens, and island specimens, etc., without giving 

 detailed data by individuals, he failed to bring out the significant facts associated with 

 the geographical distribution of the variations observed by him. 



