194 Roger Verity 
naked eye the chief difference between /yllus and pamphilus, 
far as I have found out, consists in a combination of various fe 
tures, each of which is produced singly in extreme individu 
variations of the other group, but never, on any account, combin: 
in that way, except in the particular races I will deal with, flyiı 
in an zone of transition between their habitats. The next p6i 
to observe is that, although they overlap in single features, th 
have two perfectly distinct centers of oscillation in individu 
variation, which are similar, at a very much higher degree, tot 
difference between the centers of seasonal polymorphism: in pl 
the dry heat characters are pushed much further than they ev 
are in pamphilus and the description one can give of it amoun 
on the whole, precisely to that of these characters; in C. pampl 
lIus the same thing occurs in connection with the damp cold ch 
racters. Ina similar way dry cold probably accounts for the geneı 
appearance of the I generation of /pllus and various degrees 
damp heat for that of the II generations of C. pamphilus in | 
various local races. This is the principal reason which makes r 
think they can scarcely be two species and that it is more pr 
bable they are even a rather low grade of exerges, because mo 
of their features seem due to surroundings and few to heredita 
factors. On the other hand, it seems also sure that some her 
ditary factors do come in, or one would meet with Zyllus as « 
individual variation in southern races of pamphilus living in su 
roundings apparently much more suitable, in some cases, to produ 
it than are some of those where /pllus exists alone. On the co 
trary, they exclude each other and they inhabit two different are 
only occuring together in the intermediate zone. It will be a ve 
interesting subject to work out how they stand to each other the 
The few materials I have been able to collect are suggestive th 
they blend and that transitional forms are frequent. That they a 
due to intercrossing of two different strains and not to simf 
resemblance seems to be shown by the fact noted above that the 
transitions do not occur elsewhere. I have series of hundreds 
specimens from Peninsular Italy, for instance, and I have examin 
hundreds of others, without finding a single /pllus. Instead in t 
Asturias Romei has found the two together; in a little series fro 
Seva, in northern Catalonia, consisting of perfect pamphilus, I ha 
one specimen identical to the Spanish I gen. of /pllus and anoth 
exactly intermediate. A large series of the I gen. of Palermo, whe 
the II is a /pllus of the most perfect description, consists in eve 
grade of transition from individuals similar to the I gen. /yllid 
of Sardinia to others similar to erniaustralis of Italy at high altitud. 
