280 
Psyche 
[June-September 
1949), nonadaptive developmental reactions. But the reactions of 
Pierids are the same as those regularly produced by photoperiod, 
and are presumably adaptive. 
Probably most or all individuals of P. protodice and C. eury theme 
can be switched from an estival to a vernal phenotype by some 
degree of chilling at some time in the first 24 hours after pupation. If 
this is true, then temperature can reinforce photoperiod in pro- 
ducing seasonally appropriate adult phenotypes in multivoltine 
species. For diapaused individuals of P. protodice and the closely 
related P. occidentalis Reak., the irreversibly determined vernal 
phenotype may be reinforced by post-diapause chilling, producing 
the characteristically greater phenotypic expression than is seen in 
lab-reared animals. The same is perhaps true for the (non-diapaus- 
ing) late winter pupae of C. eurytheme. The period of post-diapause 
or late-winter chilling is so prolonged in the wild that virtually all 
overwintered animals will be affected. This process will tend to 
smooth over, or conceal, not only the aforementioned genetic 
variance in “sensitive period” but that which characteristically 
occurs for chilling requirement (strength of diapause), as well. Yet 
all of this variance is potentially adaptive in fluctuating environ- 
ments, particularly when the suitability of spring weather for 
butterflies is highly uncertain, and we should expect selection to 
match norms of reaction to environmental uncertainty in popula- 
tions with sufficient genetic integrity. Does it? 
Hoffman (1978) reports that Rocky Mountain Colias philodice 
eriphyle Edw. differs from both C. eurytheme and C. p. philodice 
Godart in having adult phenotype decoupled from photoperiodic 
control (though it may be temperature-sensitive). He attributes this 
to high environmental uncertainty which renders photoperiod a 
poor seasonal predictor. Similar predictions about developmental 
phenology were made by Bradshaw (1973) and Istock (1978). 
Similarly, Shapiro (1973) found phenotype less reliably cued by 
photoperiod in the montane Pieris occidentalis than in P. protodice, 
a lowland species. Crosses of C. p. philodice X C. p. eriphyle would 
be informative as to how simple or complex genetic control of 
photoperiodic coupling can be. Right now everything which is 
known points to a definite evolutionary sequence in the history of 
insect polyphenism: a pre-existing phenotypic response to tempera- 
ture which happens to be adaptive in some environments but not 
others is subject to selection for modifiers which affect threshold of 
