CONTROLLING SEX IN BUTTERFLIES. 515 
imagines on the same day. The eggs must have been gathered 
singly, or the larvæ of different ages taken on the same day, or of 
e same age on different days. Ofa given number thus gathered 
I should expect the sexes to be about equally divided, and we in 
reality find that of the one hundred and seventy-one larvæ, partic- 
ularly mentioned, the sexes are almost equally proportioned in 
number, eighty-eight males and eighty females having been ob- 
tained and a few chrysalides (which, as we shall presently see, 
would most likely be females) perishing. 
In Anisota, on the contrary, the eggs are deposited in batches 
and it is more easy to get a number of larvæ of the same age. 
Mrs. Treat’s experience with her thirty-three larvze is quite opposed 
tomine with the same species 
Mrs. Treat does not tell us ihothir she did or did not use any dis- 
cretion as to the size in selecting her intended males and females, 
and this is a very serious omission, as by the criterion of size alone 
among larvæ of the same age, the sexes in many species may be 
Separated with considerable certainty. I regret also that she has 
not specified at what age, and whether always at the same agé, the 
treatment of “feeding up” and “shutting off” was begun, though 
we may infer, from what is said, that it was after the last larval 
molt, 
Mrs. Treat speaks of keeping larvæ eating beyond the period of 
pupating, or rather of preparation for that change, and of ‘“stary- 
ing” them, as though there was hardly any limit to these processes. 
Analyzed, what meaning do these expressions convey? Very 
little. They are deceptive! Most Lepidopterous ‘ih in a state 
of nature would come under the head of “feeding up” as they usu- 
ally have an ample supply of food at command, and eat their fill. 
While, therefore, it is perfectly possible to stunt such larve by fur- 
nishing them with a scant supply of food, and thus to prolong the 
Period and diminish the amount of their development, it is utterly 
impossible, i in the great majority of cases, to get them to eat after 
ey once commence to prepare for the chrysalis state. This is 
my firm conviction after ten years of pretty extensive insect- 
rearing, and I think that most experienced insect-raisers will 
me. If disturbed after preparing to pupate, most larvæ 
will repeatedly renew similar preparations, but if too often frus- 
trated they will either transform without the proper preparation or 
die. They are, doubtless, prompted to forsake their food and 
