MEXICAN SNAKES AND CROCODILMNS — SMITH 411 



justifies, I believe, the assumption that Jan's mexicana (Jan, 1863, p. 

 23; supposedly with 55 scale rows) either does not occur in Mexico 

 or else was improperly counted. The low counts given by Andrews 

 (1937, p. 356) are not correct; he cites them as 56 to 62 in four speci- 

 mens, but a recheck of three (Nos. 26990-2) of the same specimens 

 (the fourth could not be found) yielded maximum counts of 73 to 

 75. Since little has been recorded of the variation in this species, 

 table 8 is of interest. 



It appears that C. c. imperator does not reach the large size of its 

 southern relative, c. constrictor. I have never seen a specimen in the 

 field that would exceed perhaps 7 feet in total length, and the largest 

 measured specimen is a dried, stretched skin of about 2,420 mm. in 

 total length (7.9 feet). 



The only apparent geographic correlation of the variation recorded 

 in Mexican Constrictor is an increase in number of ventral scales on 

 Maria Madre Island. The three known specimens have counts higher 

 than any from the rest of Mexico, and accordingly they may be con- 

 sidered to represent a recognizably different race, which may be called 



•CONSTRICTOR CONSTRICTOR SIGMA, new subspecies 



Eolotype. — U.S.N.M. No. 46484, collected by Nelson and Goldman 

 on Maria Madre Island, Tres Marias Islands, May 12, 1897. 



Paratypes. — Two, including U.S.N.M. No. 24672, collected with the 

 preceding, and Calif. Acad. Sci. No. 58681, collected by Joseph R. 

 Slevin at the type locality on May 21, 1925. 



Diagnosis. — Like Constrictor constrictor imperator, but ventrals 

 more numerous, 258 to 259 as compared with the range of 225 to 253 

 in mainland specimens. 



Description of holotype. — Alcoholic skin of an adult female, esti- 

 mated length about 200 cm., tail and head intact. Scale rows 59-77- 

 39; ventrals 259; 55 subcaudals, a few near tip probably missing; 

 five subcaudals divided, others entire; 20-20 supralabials, 23-23 

 infralabials. 



Color as typical of the species, markings rather indistinct. Preoc- 

 ular dark stripe very short, not so long as eye ; postocular dark stripe 

 very narrow for a short distance behind eye (a little less than length 

 of eye), then abruptly expanding; lower edge of stripe distinct, upper 

 poorly defined; lower edge of stripe passes through rictus oris. No 

 other marks on head, except for two large spots in the infralabial 

 region on each side. Belly with rather small, black spots, scattered 

 along edges of ventrals. 



Variation. — Nos. 24672 and 58681 have, respectively, 258 and 258 

 ventrals, 57-77-41 and ?-77-? scale rows, 65 and 66 subcaudals, 22-22 

 and 19-20 supralabials, 24-? and 23-24 infralabials. Both are males. 



