516 CONTROLLING SEX IN BUTTERFLIES. 
prepare for the transformation by the changes already taking 
place in the system, and in the great majority of cases the mandib- 
ulate is already giving way to the haustellate mouth, and has be- 
come impotent to perform its wonted labor. Larvee can neither 
be forced nor stuffed beyond a certain limit, and this limit is at- 
tained by every well fed larva in a state of nature and in the viva- 
rium, so that if Mrs. Treat’s theory had any real foundation almost 
all insects that were not ‘‘starved” ought to be females. A high 
temperature will cause rapid development, but it does not cause & 
greater aggregate amount of feeding. 
But to my own experiments: Of the six insects chosen, the 
sexes in some differ in the most remarkable manner, while all 
show sufficient disparity to render mistakes in separating the 
sexes impossible. They are, also, all common in this section, 50 
that others will have no difficulty in verifying my facts. Except 
in the case of Thyridopteryx 1 made no attempt to ‘‘feed up ;” my 
efforts all being in the direction of “starving,” or, as Mrs. Treat 
would put it, of producing males. Neither Rave I relied entirely 
on my own observation ; for, being necessarily absent from home, 
at intervals, the experiments, with explicit directions, were at such 
times left in charge of Mr. Otto Lugger and Miss Mary E. Murt- 
feldt, both well practised in rearing Lepidoptera. I would also 
premise that the stunting process began from the time of hatching; 
and that it was carried so far that, of the less hardy species, many 
died under the treatment. It was, also, especially enforced to- 
wards larval maturity. The species chosen were, 1. Thyridoptery® 
ephemereformis (Haw.); 2. Orgyia leucostigma (Sm. and Abb.) ; 
3. Clisiocampa Americana (Harr.) ; 4. Hyperchiria To (Fabr.) ; 5. 
Hemileuca Maia (Drury) ; 6. Anisota rubicunda (Fabr.). 
1. Thyridopteryx ephemereeformis.—Two lots: lot 1 consisting 
at first of between thirty and forty individuals, and abundantly 
and constantly nourished; lot 2, of thirty individuals and very 
poorly nourished or ‘‘starved.” From lot 1, twenty-eight cocoons 
were obtained, of which fifteen were males and thirteen females, 
all of them attaining the imago state. From lot 2, eighteen 
cocoons were obtained, which produced twelve males and six fe- 
to perfect and dying in T 
