HERPETOLOGY OF JAPAN. 451 



(average 143.8); 29-46 subcaudals (average 35). The siipralabials 

 are not given in detail, but judging from Boulenger's account all the 

 specimens examined by him had 7. One of my Korean specimens has 

 8 supralabials on one side. It will be seen that these formulas agree 

 tolerabl}^ with those from Japan except that the subcaudals are much 

 fewer. 



Tables III and IV embrace the specimens recorded from the region 

 lying to the north and northwest of those contamed in Table II — i. e., 

 from the Russian Coast Province and the River Amur to eastern Turk- 

 estan. They represent mostly the specimens attributed to A. inter- 

 medins by Strauch and Boulenger. They agree m having an mter- 

 mediate number of subcaudals, viz, 36-49 (average 41) and [32] 37-53 

 (average 43.7) , respectively. They also agree fairly well in the number 

 of ventrals, viz, 151-163 (average 155.8) and [147] 154-179 (average 

 163.5), respectively, but it will be seen that these figures greatly ex- 

 ceed the corresponding ones of Tables I and II. There is, however, 

 one marked difference between Tables III and IV, since it will be found 

 that among the specimens collected east of the Khmgan JMountains 

 in Manchuria (Table III) the great majority (70.6 per centMn a total 

 of 17 specimens) have 21 scale rows, while of those from the Khingan 

 IVIountains and the country to the west, viz, JVIongolia, Transbaikalia, 

 and eastern Turkestan, the greater number (76 per cent, in a total of 

 29 specimens) have 23 rows or more. 



Before proceeding farther a few words may be said regarding 

 Table lb. The two specimens from Yaeyama in the United States 

 National JVIuseum agree structurally and in their scale formulas 

 exactly with the Japanese specimens, but they differ exceedingly 

 from the latter in color. Their ground color (in alcohol) is a "tawTiy 

 olive, "'^ uniform on top of head, but indistinctly varied and mar- 

 bled on back and sides with blackish and with indications of pale 

 cross bands; a band slightly darker than the ground color through 

 eye and temples; lips whitish, as is also the whole underside; each 

 ventral sprinkled with irregular blackish brown dots and spots on 

 the basal half. It seems probable that they represent a permanent 

 local color race, and are so recognized further on. 



The differences between the specimens listed in Tables I-IV may 

 be sATioptically expressed as follows: 



a' Ventrals 151 or less. 



6' Subcaudals 43 or more Tal)les land I« (.Japan and Yaeyama Islands). 



b- Subcaudals 46 or less Table II (SE. China; Korea; Formosa). 



d^ Ventrals 147 or more. 



¥ Scale rows (in 70.6 per cent) 21.. Table III (Coast Province; Amu rland, west to 

 • Khingan Mountains). 



b- Scale rows (in 79 per cent) 23 Table IV (Khingan Mountains to East Turke- 

 stan). 



aRidgway, Nomencl. Colors, 1886, pi. in, No. 17. 



