HERPETOLOGY OF JAPAN. 



419 



necessary to use a binocular microscope magnifying 45 diameters, 

 and to make sure of the presence of grooves on the anterior face of 

 the maxillary teeth of Disteira gracilis it was even necessary to raise 

 the power to 65 diameters. From my experience it is then plain that 

 there is no real difference between the "grooved" and the "solid" 

 maxillary teeth in these snakes, that it is merely a question of degree. 



Under these circumstances there seems to be no other course than 

 to unite the two supposed genera (and species!), the name Disteira 

 being available for the thus consolidated genus. The term Hydrophis 

 must disappear from our nomenclature, as Latreille, in 1802, framed 

 it as a substitute pure and simple for Hydrus, while Daudin, in 1803, 

 substituting Pelamis for the latter reemployed the former for the 

 bulk of the Hydrid snakes. 



Within the enlarged genus the limits and characters of the species 

 •offer some of the most difficult problems to the ophiologist. These 

 snakes, in many respects so excessively specialized, often show a 

 correspondingly excessive amount of individual variation, and as 

 the series available for study is insignificant except in the case of a 

 few species the result must of course be in great doubt. 



If, for instance, we take the latest monographic account of these 

 snakes, namely, Boulenger's treatment of them in the Catalogue of 

 the Snakes in British Museum, by far the ablest treatise on the genus, 

 based upon the largest material in any collection, we find that of 22 

 species recognized by him in the genus Hydrophis he had 77 specimens, 

 or an average of 3.5 specimens to each species.'* If we analyze this 

 material for the variability of the number of ventrals we come to the 

 result that the range of the ventrals, that is, the difference between 

 the greatest and smallest number of ventrals — of the specimens of 

 each species in British Museum is greater in proportion as the series 

 of specimens examined increases. This is startlingly shown in the 

 following; table: 



The only exception is the series of eight //. cserulescens in which 

 the range of variability of the ventrals is unusually low. These speci- 

 mens however, cover a relatively small geographic range. 



« In no instance did he have more than I'lxe specimens from one locality. 



