AM 
to be their last one. This, with other reasons, led me to test the 
mass of the pupæ by weighing. 
TEST OF MASS, OR WEIGHT. 
Over 350 pupe have been weighed in 25 different lots, 12 of 
these lots having been during the early part or the whole of larval 
life, at the higher temperature (60° in 1906, 65°, as afterwards men- 
tioned, in 1910), and 13 at the lower or 54”. The former averaged 
about 30 °/. more in weight than the latter. Nearly all show 
the same gulf of separation in weight between those at the higher, 
and those at the lower temperature; not one of the latter lots 
showed so great a weight in its average individual as any one of the 
former lots. Six of the 25 had been at the respective temperatures 
during the whole larval life, 3 of them at the lower and 3 at the 
higher, the remaining 19 lots during varying portions of the 
earlier life only; in their case, taken as a whole, the difference in 
weight was slightly less than with the 6. Two of the 6 were the 
case before mentioned where the eggs laid by a single parent were 
so numerous that they were sufficient to divide between the two 
temperatures of 54° and 60°, the respective weights here being 
134 and 181 grammes, an excess of about 35 °/o. 
EXPERIMENTS OF IQIO, HIGHER TEMPERATURE RAISED. 
Most of my experiments, especially with S. bilunaria, were 
conducted with the object of imitating as nearly as might be the 
temperatures to which the two phases are in nature subjected 
during their larval stages. For these purposes in 1909 I had placed 
those intended to represent the summer phase larva at the sup- 
posed mean of about 53° or 54° F. (12° C.), afterwards raised to 
about 61° or 62° E. (16°-17° C.); those intended to represent the win- 
ter phase larva at about 61°-62° F. (16°-17° C.), afterwards lowered 
from this to about 58° F. (14°-15° C.). I became satisfied afterwards 
that certainly the higher one of these temperatures was too low to 
represent nature. Moreover, a closer examination of my apparatus 
showed me that what I thought 61°-62 F. was in many cases really 
about 60° or 61° F. Therefore, when subsequent experiments in 
1910 were tried, the higher of the temperatures was raised, 65° F. 
(1819 C.) being the mean then aimed at, perhaps rather too high 
