186 BULLETIN 58, UNITED STATES NATIONAL MUSEUM. 



orcliiiaiy kind. It inay represent a freak or a sport, but in view of the 

 many points in which it differs, this explanation does not appear 

 plausible. That two separate species so closely related should occur 

 in the same locality in Formosa would scarcely be accepted by 

 herpetologists without incontrovertible proof. The question then 

 naturally suggests itself whether by some accident a wrong locality 

 may not have been assigned to it. To the neck of the sj)ecimen a 

 large paper tag is firmly tied bearing the following inscription in 

 Latin letters: ''No. 22, Taipa, Formosa. 7-96," but there are good 

 reasons for believing that this is not the original collector's field label. 

 However, if not from Formosa proper, where could it have come 

 from? The s])ecimen, as we have seen, is not identical with any 

 from the middle or southern groups of the Riukius. On the other 

 hand, that it was part of the collection made by Mr. Tsunasuke Tada 

 seems unquestionable. He collected in Formosa, Botel Tobago, and 

 the Pescadores Islands. Our specimen agrees even less with J. 

 mitsukurii, from Botel Tobago, than with the other two forms. 

 Could it, then, possibly have come from the Pescadores or any other 

 outlying island ? 



The question can not be decided now, and is thus left for future 

 explorers in that region. 



Descri'ption. — Adult male; U.S.N.M. No. 34082; northern Formosa; 

 June 23, 1903; Owston collection. Rostral low, separated from 

 nasal by one scale; nostril round, in a single, somewhat swollen nasal, 

 wliich is in contact with first supralabial; canthus rostralis very 

 sharp, covered with about five scales; superciliary ridge interrupted 

 by a deep notch at the posterior end of the orbit, followed by a 

 single low spine; various isolated spines on occiput and temporals; 

 eight supralabials, eighth very long and narrow; all head scales 

 wrinkled and keeled; back covered by imbricated keeled scales of at 

 least tliree different sizes; a low nuchal and dorsal crest of pointed 

 scales, those of the former slightly higher, seven in number, flattened, 

 conical, the latter triangular, 33 in number; a similar crest on the 

 base of the tail, the three crests being continuous, the spines over 

 the shoulder and between the hind legs, however, bemg lower than 

 the others; scales on flanks not much different from dorsals, except 

 those back of the axilla and near the groin which are much smaller; 

 scales on underside strongly imbricate and keeled, pointed but not 

 mucronate; digits underneath with a double series of spiny keels; 

 tliird and fourth finger nearly equal; fourth toe, without claw, from 

 base of third considerably longer than distance from center of eye to 

 tip of snout; third toe, without claw, from base of fourth, shorter 

 than distance between superciliary edges at center of eye; distance 

 from upper side of knee of hind leg, bent at right angles, to tip of 



