on Vanessa and other Lepidoptera. 437 
results is very probably owing to the circumstance that 
the pupe were all of them several days old when they 
reached me. 
The experiments now recorded confirm in general the 
conclusions drawn from such as have preceded them, and 
some of which may be briefly enumerated as follows :— 
(1) The effects of temperature are different when applied 
at different periods of the pupal stage. (2) A great range 
of temperature may cause but little difference in appear- 
ance, while a very few degrees near the top or bottom of 
the range the imsect will bear may cause a great 
difference. (3) There may be a great constitutional 
difference in sensitiveness to temperature between two 
seasonal emergences of the same species. (4) This may 
be so even when both pass the pupal period at about the 
same temperature (this is in accordance with Mr. W. H. 
Edwards’ observations above referred to). (5) While 
some kinds of effect seem to be what may be called the 
direct result of temperature, in others, and perhaps the 
most important, temperature appears to operate by 
causing the individual to “throw back” to some 
ancestral form; this last circumstance has been con- 
sidered to explain the reason why a low temperature in 
some species causes darkening of the colours, and in 
other species produces the opposite effect. (6) In these 
cases of “‘reversion,” the kind of effect produced appears 
to depend on the stimulus applied, low temperatures 
producing one class of effects and high temperatures 
a different class of effects. 
The whole subject is one of much complication, and 
calls for further experiments in many directions. The 
direction which mine have taken, following in the lines 
initiated by Weismann and W. H. Hdwards, especially 
if pursued with species belonging to regions where the 
seasonal or other occasional differences of temperature 
are extreme—North America, Siberia, Japan, or the 
vicinity of mountains—will help to trace, and separate 
from the rest, such of the causes of variation as depend, 
directly or indirectly, on temperature. Systematic 
experiments on a number of well-selected species 
belonging to countries where the seasonal difference is 
hygrometric rather than thermometric, would probably 
produce valuable results. The nature of the food-plant, 
