HERPETOLOGY OF JAPAN. 83 



from which "it may be easily separated by the following characters: 

 No web between -the outer fingers; toes one-third webbed; no dark 

 loreal stripe," while //. arborea japonica has "one-fourth web between 

 third and fourth finger; toes one-half webbed, and an always dis- 

 tinct dark loreal stripe. In the proportion betw^een the length of 

 the tibia to that of the femur there may also be a slight difference. 

 In var. immaculata the tibia appears to be somewhat longer than the 

 femur; in var. japonica, on the contrary, the femur is longer than the 

 tibia." 



I have not seen any specimens of this form from the type locality 

 and therefore can not express any opinion as to its validity, but if 

 the three specimens from Mongolia, which ]\Ir. Witmer Stone some- 

 what doubtfully referred to this form, really belong to it, it would 

 appear that the characters assigned to it are somewhat unstable and 

 unreliable. 



Thanks to the kindness of the authorities of the Philadelphia 

 Academy, I was recently permitted to examine the three specimens 

 alluded to above, namely, Nos. 4943-4945, collected by Doctors Smith 

 and Furness at Tore River, ^longolia. 



No. 4945, whicli is adult, differs from a German specimen, with 

 which I could compare it, by being entirely immaculate, with no 

 trace of canthal or postocular streaks or dusky crossbars on the legs. 

 The foot is somewhat shorter and the webs slightly larger; the heels 

 tlo not meet by a space of 2 mm. when the folded hind legs are placed 

 vertical to the axis of the body, and they do not reach the axilla 

 when the leg is extended along the side of the body. 



No. 4944, also adidt, has relatively longer hind legs, as the heel 

 reaches the posterior border of the eye. A canthal and postocular 

 line can be distinctly traced, also some spots on the upper lip, and 

 apparently faint cross markings on the legs. 



No. 4943, which is quite young, is nearly unspotted like No. 4945, 

 but with trace of a subocular spot on one side and of cross markings 

 on legs. 



The following measurements were taken of the two atlidts to show 

 the difl'erence in the relative length of the hind legs : 



No. 4945. No. 4944. 



mm. 



Snout to vpnt 



Vent to tip of longest toe. 

 Heel to tip of longest toe. 



mm. 

 40 j 39 



49 59 



25 27 



The status of the tree-toad of northern China is consecjuently still 

 doubtful. It seems to be fairly distinct from the Japanese form, 

 however, in the relatively immaculate coloration. 



