ART. 25 CHINESE AMPHIBIANS AND EEPTILES STEJNEGER 61 



sition of the species here described. In that synopsis the species of 

 this g:enus are divided into those with " I. Head not more than one 

 and three quarter times as long as broad," and those with " II. Head 

 at least nearly twice as long as broad " ; in other words, the ratio 

 between breath and length of head in Group I is 1:1.75, or less; in 

 Group II it is " nearly ''1:2, or more, or if we interpret the " nearly " 

 as '' minus 0.1 " the ratio in Group II may be said to be 1 : 1.85. The 

 table of measurements shows that in the new species the ratio varies 

 between 1 : 1.75 and 1 : 2, averaging in the five specimens 1 : 1.88, or 

 halfway between the two groups with 'a leaning towards Group II. 

 That this is the correct interpretation is evident from an examination 

 of the ratios given by Boulenger himself for three of the species 

 composing Group II, namely, T. smaragdinus^ T . sauteri^ and T. 

 seoclineatus^ in which according to his figures the ratios are respec- 

 tively 1 :1.85 ; 1 :1.93 and 1 :1.98. The Szechwan specimens must 

 therefore be tested both in Groups I 'and II, and as the elongated 

 head points towards the latter, they may be looked for first in that 

 category. Having eight dorsal rows of scales (or plates) and ven- 

 trals in six series they ought to be found under vl, and having four 

 pairs of chin shields they might be suspected of belongiilg to T. 

 sauten^ hitherto only known from Formosa. But this is a very dif- 

 ferent species with sharply keeled venti^als, one inguinal (femoral) 

 jDore, two or three series of keeled scales on the sides above the 

 ventral plates and only 21 lamellae under fourth toe. Tried in 

 Group I, which falls in two Groups A and 5, the latter with four 

 dorsal series, 12 ventrals and three pairs of chin shields, they should 

 be looked for in Group A in spite of the fact that the sj)ecies con- 

 tained in that group are said to have ventrals in eight or 10 series, 

 while our specimens only have six. A is divided in those with "7. 

 four or five pairs of chin shields" and "^. three pair of chin 

 shields." As ours have four pairs the choice is limited to T. anrmr- 

 ens/s\ T . tachydromoides, and T. wolteri. Of these T. aTuurensis has 

 three inguinal pores, and T. tvoUeri one pore, and both have eight 

 ventral series, while ours have two pores and six ventral series. 

 Finally, T. tachydromoides^ from Japan, like our species, has two 

 inguinal pores and four chin shields, but the number of dorsals and 

 ventrals is reversed, namely, six dorsal and eight ventral rows, while 

 T. intermedius has eight dorsal 'and six ventral rows, besides, having 

 much longer head and various peculiarities of its own, such as the 

 numerous granules covering the temples, lack of well-developed 

 " plates " on the flanks adjoining the ventrals ; more numerous la- 

 mellae under the fourth toe; nas'als in contact behind rostral, etc. 



