phase, by temperature applied at a suitable part of the larval stage, 
which is the stage on which I lay special stress (1). 
S. bilunaria, SAME TEMPERATURE DURING 
WHOLE LARVAL LIFE. 
That the same takes place with S. bilunaria, I have abundant 
evidence. In experimenting on this species 1t had occurred to me 
that the whole normal difference of mean temperatures in larval 
life between the two phases being not more than 4° or 5° F. 
(say 2° to 3° C.), it was not probable that mere difference of 
temperature of so small amount would be enough to account for 
the two phases, and I tried some experiments with the special view 
of ascertaining this, by subjecting larva to the respective temper- 
atures during their whole life. A large brood of S. bilunaria from 
the same parents was divided from the time of hatching, one part 
being placed during the whole of larval life at the mean temperature 
of 54° F. {12° C.), the other at the mean temperature of about 60° F. 
(15"-16% C.) till pupation, both being then brought into outdoor 
temperature. Those kept at the lower temperature emerged in an 
average time of 118 days, their range being 67 to 190 days, those 
at the higher temperature in 184 days, their range being 79 to 
208 days (2). Similar experiments with other broods had similar 
differential results: and establish that « spring » temperature 
applied to the larva during its whole life tends to shorten the 
pupal period as compared with a « cool summer » temperature. 
(1) I think the facts recorded by Prof. WEISMANN as to his experiments on 
the July and August larva of 4. /evana accord with my views so far as to show 
that it was only when the Insects were placed at the high temperature in their 
larval stage that there was any considerable conversion from the winter (/evana) 
phase to the summer (frorsa) phase. To establish an exact parallelism with 
nature, or one as nearly approaching to it as the inherent difficulties will permit, 
experiments applying high temperatures, but lower than 80% F, (27% C.), on 
young July and August larva appear desirable. 
(2) As to the difference in weight or mass see later. 
