18 BULLETIN 58, UNITED STATES NATIONAL MUSEUM. 



The adult femah (U.S.N.M. No. 31894; Koclii, Shikoku; May 11, 

 1903, Dr. Hugh M. Smith, collector) is larger, total length being 

 123 mm., and differs from the breeding male in the lesser develop- 

 ment of the parotoid glands as well as the lateral glandidar ridges, 

 absence of the scapvdar gland, and the caudal filament. The body- 

 is also longer, the distance from tip of snout to gular fold being 

 contained more than three times in the distance from gular fold to 

 vent; so that the fingers and toes barely touch when the limbs are 

 pressed against the side, the toes themselves also being shorter 

 than in the male; finally, the vent is quite different, being a nearly 

 cu'cular opening with long radiating papillae forming an oval swelling 

 on a compressed elevated base. 



The nonhreeding male and female (U.S.N.M. Nos. 30735, 30736; 

 Miyazaki, Kiusiu) differ from the breeding ones chiefly in the lesser 

 develo})ment of the glands and of the papillae surrounding the vent, 

 the male also by the absence of the caudal filamentous appendix. 



Color of living specimens. — Dr. Hugh M. Smith has kindly fur- 

 nished the following note on the color of the living specimens collected 

 by him in Kiusiu and Shikoku during the breeding season. The 

 back is black or blackish green; the underside of head, body, legs, 

 and tail is blood red, with dark green mottling or streaks inclined 

 to be synnnetrical. In some specimens the back is lighter green and 

 the sides of the tail may be steel blue. 



Variation. — The chief variation is in the amoimt of dusky or 

 blackish color on the underside. The normal markings seem to consist 

 of a series of irregular spots on each side of the belly sometimes 

 confluent to longitudinal streaks. These markings are often so 

 heavy as to make the sides nearly uniform black, but the center of 

 the belly is usually unspotted. On the other hand, some specimens 

 are nearly devoid of spots. These form Krefft's variety immacu- 

 liventris, but they are not confined to any particular locality and 

 represent only the extreme individual variation. 



A male specimen taken in May, 1898, near Mount Fuji (U.S.N.M., 

 No. 34308), is of a light grayish bro\v^l, almost inispotted underneath 

 and seems to be a sort "of albino. 



Boulenger and Wolterstorff (Zool. Anz., XXX, 1906, p. 562) express 

 the opinion that a more detailed study of B. lyyrrliogaster might 

 show the desirability of breaking it up into several subspecies, and 

 the former mentions several breeding males and females from the 

 island of Iki as being very small, as well as some Kobe specimens 

 as having very weak parotoid glands. Wliether the former island 

 contains a particular dwarfed race it is impossible to say, as no^ 

 measm-ements are given, but I do not believe it will be possible to 

 split the specimens inliabiting the various large islands into geo- 

 gi-aphical forms. The size and prominence of the parotoid glands 

 seem to be more or less dependent on the season. 



