eee BARBOUR — ASIATIC AMPHIBIA AND REPTILIA 75 
Mr. Owston’s obtained not only four specimens in Okinawa shima, 
but also three in Amami Oshima, and one in Yaku shima. ‘The 
latter occurrence is so extraordinary, however, that it would be well 
to await corroborative evidence, as some mistake in labelling may 
have occurred.”’ The specimen before me was picked out, the 
only one which Mr. Owston had at the time, from a lot of speci- 
mens which seemed to have full data. It was taken 10 April, 1907, . 
and thus cannot be one of the three examples mentioned above. 
There can now be no doubt whatever as to the species occurring on 
Amami Oshima. Unfortunately no corroborative evidence can be 
offered as to the validity of the Yaku shima record. The formula 
for this example follows: V. 129; anal entire; C. 42; interorbitals, 
9; supralabials, 8. ‘This is precisely the condition in Okinawa spec- 
imens. Variation within this form is very slight. 
Trimeresurus flavoviridis (Hallowell). 
Hallowell, Proc. Acad. Nat. Sci. Phila., 1860, p. 492. 
Stejneger, Herp. of Japan, 1907, pp. 475-478, pl. XX VII. 
A truly formidable serpent, which reaches a length of nearly five 
and a half feet. This specimen was taken at Amami Oshima, 
Riu Kiu Islands, 19 September, 1904, by one of Mr. Owston’s 
men. ‘The interorbitals are 13 in number, and supralabials 9. 
V. 225; anal entire; C. 84. This formula is only approximate, as 
the specimen is slightly mutilated. 
Trimeresurus mucrosquamatus (Cantor). 
Cantor, Proc. Zoél. Soc., 1839, p. 32. 
Stejneger, Herp. of Japan, 1907, pp. 467-470. 
A single specimen from Bankoro, in the highlands of central 
Formosa. It differs slightly from the specimen which Stejneger 
described, in that there are 18 scales in a line between the supra-oc- 
ulars. ‘These latter are extremely narrow. ‘There are also 10 su- 
pralabials, instead of 9. Ventrals 215; anal entire; subcaudals 88. 
