48 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
The change in color begins underneath and gradually spreads 
upward, the green passing gradually into yellowish which assumes a 
dirty gray and finally a deep velvety brown. 
They seem to eat little or nothing after this change of color and 
usually pupate in a few days when put into the cage. After moving 
about uneasily for some time, they come to rest. A silken thread is 
spun, attached to the wire of the cage, and passed around the body 
between the sixth and seventh segments. While the caterpillar is 
thus suspended the cuticle hardens gradually, the head drops off, the 
color changes to a mottled gray and brown presenting a rough irregular 
surface strikingly resembling old weatherbeaten boards or bark, the 
usual objects to which they are attached in Nature. I have found the 
chrysalis (pi. 12, fig. 3) thus suspended from the edge of the siding on 
the most exposed wall of an old unpainted house, which it strikingly 
resembled in color. I had passed by it for several months during the 
winter, but had not noticed it till I saw a sparrow fly at it but fail to 
get it. 
These changing colors to suit first the green leaves, then the brown 
withered leaves, and finally the gray, somber shades of weatherbeaten 
support and cloudy winter weather are certainly adaptations in them¬ 
selves marvelous, but doubly so, because of the brief period of time 
(a week or two) in which the changes are brought about, corresponding 
to equally rapid changes in the autumn temperature. Yet the frost 
which kills the leaves and thus causes the deep brown of such leaves, 
is not responsible for the change from green to brown in the cater¬ 
pillar, since the same change occurs when the larva is kept in the 
laboratory at a uniform temperature of about 70°. 
Besides this seasonal correspondence in color between the larval 
stages of this butterfly and the plants on which it lives, there is a deeper 
correspondence between the general life processes of the two. And 
if the change in color from green to deep brown is due in the cater¬ 
pillar to changes in the physiological pro'cesses, it seems not unreason¬ 
able to assume that the transformation in color of autumn foliage 
may likewise be due to internal physiological changes in the plant 
rather than to variations in the external temperature. 
The life phenomena in the young butterfly resemble, broadly speak¬ 
ing, the life phenomena in the willow on which the caterpillar lives 
at this autumnal season of the year, in the following particulars: (1) 
in both, growth ceases; (2) both approach a season of rest when the 
