H. E. Crampton 
115 
normal. And it also appeared that many of the normal living pupae failed to meta- 
morphose perfectly, whether they were kept in the laboratory or out of doors. The 
opportunity thus offered was taken for an examination into the phenomena of 
elimination for two successive developmental periods. The tiist, or period of 
pupal elimination, extends from the formation of the cocoon and pupation to 
the beginning of the formation of the imago. The second period immediately 
supervenes, and, as the period of pupal-imaginal elimination, is marked by the 
deeply-seated structural and physiological changes of metamorphosis. The re- 
duction of the second period, whicli results in the weeding out of certain of the 
adidt moths, is none the less entirely dependent upon the ability of the pupa 
to accomplish successfully the production of the imago. On this account, as well 
as for the reason that the moth is constructed in strict accordance with the lines 
of the pupa, the examination into selection of this second period may justly be 
conducted with reference to pupal characters. To the two sections in the 
following account which treat separately these periods of elimination, a third 
section is added, which deals briefly with the comparative variability of the pupae 
of the two sexes. 
II. The Material and its Treatment. 
In December, 1899, all the accessible cocoons were obtained from a small area 
of less than one acre in the upper part of the City of New York. The cocoons 
were found principally upon Ailanthus trees, the favourite and original food-plant, 
where clusters of twenty-eight or thirty were sometimes to be found upon a single 
branch. Many were taken from vines and fences near these trees, and still others 
were picked up from the ground whence they had fallen from the trees. It may 
be noted in passing that these last, far from being unfavourably affected, contained 
as many living pupae as those which remained attached to the trees. 
The total number of cocoons was 1090. Upon examination of the contents, 55 
(4*8 °/J were found to contain dead and shrivelled individuals which had failed to 
pupate. Ninety-three contained pupal cases from which the moths, probably of an 
interpolated summer brood, had escaped, or where, in nine cases, the fully-formed 
imago had been unable to emerge from the cocoon. Out of the 942 remaining, 6^3 
had pupated, forming in most cases an entirely normal pupa, but they were dead. 
Thus only .319 "selected" individuals remained to provide for the continuance of 
the species. 
Equal numbers of the dead and surviving pupae were carefully measured, as 
stated below, and the living ones were kept thi'ough the metamorphosis, care being * 
taken to isolate them as this time approached. Only 181 (97 males and 
84 females) produced perfect moths, 75 were imperfect to a slight degree, 38 
were hopelessly malformed and finally 16 failed to metamorphose at all. The 
perfect imagines constituted only 16'6°/„ of the whole number of individuals which 
entered the cocoon, from which we may gain an idea of the severity of the con- 
ditions under which the quiescent pupa exists. 
^5— 2 
