680 The American Naturalist. [August, 
summer heat, is very dark, almost black, but the spring form corre- 
sponds with our German red-golden phlæas. 
Although to-day I still look upon this view as correct, and a directly 
altering effect of temperature as proved, yet I have gradually been 
convinced, that this is not the sole origin of seasonally dimorphie varia- 
bllity, but that there is also adaptive seasonal dimorphism. We must, I 
believe, distinguish direct and adaptive seasonal dimorphism ; and, I see 
in this distinction an important advance, which, before all, places us in 
position to explain the results of the various experiments undertaken 
by myself and others in a much more satisfactory manner. 
I have already pronounced this view in a lecture delivered at Oxford 
in the beginning of 1894, and I have sought to show that adaptive 
seasonal dimorphism, which I had previously only put forward as possi- 
ble, does actually occur. The example there given for perfect insects 
was, indeed, only a hypothetical one, viz., the case of Vanessa prorsa- 
levana; but for larvæ, at least, I can select an example from Edward’s 
excellent work on the North American butterflies with tolerable cer- 
tainty, viz., that of Lycæna pseudargiolus, which will be more accurately 
iscussed later on. I did not then know what I learnt shortly after- 
wards from an interesting little pamphlet of Dr. G. Brandes, that cases 
of seasonal dimorphism had been known for a long time among tropical 
butterflies, and that among these, at least, one of the seasonal forms 
depend upon the assumption of a special protective coloring. Brandes, 
maintains, with justice, that the view hitherto widely held among us is 
erroneous, according to which seasonal dimorphism was not to be ex- 
pected in trophical countries, since the alternation of seasons is absent 
there. Periods of rain and drought, at least for many tropical coun- 
tries, form such an alternation very sharply. At any rate, Doherty, 
and, somewhat later, de Nicéville, have pointed out, for Indian butter- 
flies, a series of seasonally dimorphic species, not merely by the observa- 
tion of the alternation of the two forms in nature, but by rearing the 
one form from the eggs of the other ; thus among Satyridæe of the genera 
Yphthima, Mycalesis, and Melanitis, and for the species of Junonia, it 
is accepted as proved; and in all these cases the difference between 
the two forms principally consists in the fact that the one form seems 
like a dry leaf on the under side, while the other possesses another 
marking, and at the same time a number of ocelli. 
Without engaging in the controversy as to the biological value of 
these ocelli, I do not for a moment doubt but that the coloring with 
ocelli is also an adaptive form, possibly protective or intimidating color- 
ing. If one of the two forms had no biological significance, it could 
