202 BULLETIN 114, UNITED STATES NATIONAL MUSEUM. 



Similarly an aberrant individual from Delaware (Acad. Nat. Sci. 

 Philadelphia, no. 3597) became the type of temporalis Cope. There 

 may be others like this in Delaware (Stone, 1906, 167) but so far 

 this appears to be the only example of temporalis. The head pattern 

 is obsolete; the dorsal blotches are wide, few, and somewhat fused 

 with each other; and the scale formula is 19-17. 



Affinities, — ^Under variation it was shown that triangulum attains 

 its most perfect development in the northern States, that in the 

 Alleghenies and on the Atlantic coast it has suffered more or less 

 reduction in scutellation and pattern, and that the greatest reduction 

 occurs from southern New York to Virginia. The meaning of this 

 extreme condition is not at all clear, and, in attempting an explana- 

 tion, it must be remembered that from the whole peninsula of Dela- 

 ware and eastern Maryland we have but a single specimen, and that 

 all of those from Virginia are from close to the District of Columbia 

 or from the Alleghenies. Some of those from the District have been 

 known under the name of "doliatus/' and have been supposed to be 

 the same thing as what we are here calling syspila. Admittedly some 

 young individuals can not be told from the latter. Others, however, 

 are typically triangulum, and still more are mixtures. Perhaps this 

 is a region of intergradation between a typical triangulum and a 

 coastal variety extending from New Jersey to North Carolina. There 

 seems to be nothing constant. In this connection it may be remarked 

 that more than half of these specimens are very small, and perhaps 

 some of the aberrant ones would have died a natural death very 

 soon if they had not been found and preserved. It is a fact that 

 some of the most aberrant individuals, described and not described, 

 are juveniles. Of the former multistrata Kennicott is an example. 

 If these Washington specimens were fairly constant in pattern and 

 scalation and possessed a definite geographic range, they could be 

 recognized with a name, but such does not appear to be the case. 

 The resemblance to syspila must be regarded as secondary because 

 the scutellation is definitely reduced, and the pattern as well bears 

 every evidence of being capable of derivation by reduction from typi- 

 cal triangulum. Whatever may be the reason it seem.s evident that 

 in the parts of its range where triangulum is farthest removed from 

 syspila it has a reduced scutellation or pattern or both; for example, 

 New England east of the Berkshires, New Jersey and Virginia east 

 of the Alleghenies. 



The fact that the pattern of triangulum seems to be developed in 

 perfection only north of the southern limit of glaciers argues against 

 the assumption of this pattern as primitive. As brought out in the 

 discussion of syspila, triangulum must be considered as at one end 

 of the chain, amaura-syspila-triangulum. If we are to favor evolu- 

 tion or dift'erentiation accom])anying migration, we can not do other- 

 wise than regard triangulum as the latest of this chain. 



