HERPETOLOGY OF JAPAN. 365 



U.S.N.M. mentioned above), as well as one alleged to be from 

 ^' Japan," made similar observations, and also called attention to the 

 lower number of ventrals and higher number of subcaudals in the 

 latter. The numbers of these plates as given by him do not overlap, 

 but from the tables presented by me (pp. 362 and 366) it will be seen 

 that there is considerable overlapping. Wliile it may thus be inexpe- 

 dient to give the Riu Kiu form a binominal appellation, there can 

 be no doubt that it should be recognized nomenclatorially. 



It appears that all the names in the synonymy of D. rufozonatum 

 are based upon Chinese or Formosan specimens with more than 192 

 ventrals and less than 73 subcaudals. The type localities of D. rufo- 

 zonaium, D. r. var. formosana, and 0. striata are known and agree in 

 the scale fornnda. The type of D. cancellatum is from an unknown 

 locality, but the scale formula as well as the number of black bars on 

 back and tail are conclusive. Hence the necessity of giving a new 

 name. 



The number of ventrals and dorsal dark cross-bars being the cliief 

 distinctive characters of this subspecies, as indicated above, it appears 

 superfluous to present a detailed description of a Riu Kiu specimen, 

 but a few remarks on the individual variation within the form pecu- 

 liar to the archipelago may not be out of place. 



Variation. — In the five specimens before me the loreal is in every 

 case excluded from the eye by the preocular. That this relation is not 

 constant, however, is shown by Boettger's statement that in some of 

 the four Miyakoshima specimens examined by liim there is found 

 two preoculars on one side, or on both sides in some, while mostly 

 there is only one. Other variations noticed by him are three post- 

 oculars on one side in one (also found in one of the specimens before 

 me), and only two supralabials entering the eye in another. Ventrals 

 vary between 178 and 195 in nine recorded specimens; subcaudals 

 between 73 and 87. The largest number of dark cross-bars in five 

 specimens is 52, the smallest 41. 



In the youngest specimen before me (Sci. Coll. No. 11) the dark 

 markings on the upper side are nearly solid, though even in this many 

 of the scales in the interior of the blotches have narrow pale edges. 

 In all the other specimens, however, which are much larger, these pale 

 edges are much broader, so that the cross-bars appear lighter brown 

 with dark margins, the paler interior having mostly a streaky 

 appearance. 



The t3^pical D. rufozonatum from the mainland seems to have the 

 cross-bars solid-colored. 



Hahitat. — This race seems to be confined to the southern and middle 

 groups of the Riu Kiu Archipelago. 



Mr. Tashiro has collected specimens in the Yaeyama group, both on 

 Iriomote shima and on Miyakoshima. The Senckenberg Museum 



