296 BULLETIN 58, UNITED STATES NATIONAL MUSEUM. 



two pairs of chin-shields, while the type of A. spinalis has three. 

 But here, again, our specimen No. 30731 from Kiusiu steps in between 

 as having only two, while the specimen in Tokyo and our two spec- 

 imens from Fuji have three like the type." The slight difl'erence in 

 the number of ventrals and subcaudals is completely bridged over 

 by the additional specimens. The coloration, on the other hand, is 

 the only feature which thus far offers a character by which to distin- 

 guish the two forms. A. hraconnieri is described by Boulenger as 

 being uniform l)lackish, the ventrals having a narrow whitish edge. 

 Tlie Japanese specimens all seem to have the median black dorsal 

 line and often a blackish longitudinal median stripe on the underside 

 of the tail, the latter being absent in No. 34047 and very obscure in 

 No. 30731, which also shows trace of a dorso-lateral dusky line. 



It may be that an actual comparison of specimens from Japan and 

 China may demonstrate further differences,'' but it is c[uite clear 

 that the Japanese species is very closely allied to the Ichang speci- 

 mens collected hj Mr. Pratt and described by Boidenger. 



Recently Doctor Wall '^ has come to a similar conclusion as to the 

 probable identity of the two species. He examined tlu^ee Japanese 

 specimens, which all had 23 scale row^s, viz, the two specimens now 

 in the U. S. National Museum alluded to above (Nos. 34046, 34047), 

 and a third one from the same locality. Unfortunately he does not 

 mention the coloration nor the number of chin-shields in the third 

 specimen. It is also impossible to make out which of the scale for- 

 mulas given belong to the the third Japanese specimen.'' 



Whether, on the other hand, the Ichang specimens are identical 

 with Sauvage's originals of A. hraconnieri from eastern Kiangsi is 

 another question. His description is certainh?" very different from 

 Boulenger's. Thus he speaks of the temporals as being numerous; of 

 the prefrontals as small; his specimens have a small lower postocular, 

 and the number of labials he states to be eight; the color, moreover, 

 is a ''brun ardoise uniforme." 



« I may here call attention to the fact that in Boulenger's figure of A. rufescens (Cat. 

 Sn. Brit. Mus., I, 1893, p. 308, pi. xx, fig. 2) there is only indicated two pairs of chin- 

 shields, while the description speaks of three. This figure also shows only five supra- 

 labials as against six in the description. 



b It may be well to point out a difference in the dental formula of the genus and the 

 dentition as I find it in the specimen in the Science College specimen, Tokyo, and in 

 our No. 3404G. In these I count 14-16 maxillary teeth, which are subequal, the first 

 three and the last three slightly smaller, while Boulenger for the genus says: "Max- 

 illary teeth 22 to 25, small, equal." Is this character derived from Ichang specimens, 

 and have we here a possible difference in the two forms? 



e Proc. Zool. Soc. London, 1903, p. 88. 



d I presume that his scale formulas 165-|-44 and lG5-f 48 are meant for our specimens 

 No. 34047 and 34046, respectively. Repeated counts of these specimens give 163+45 

 and 165+49. 



