APATURA II. 



ceased to feed ami partly clianged color, again began to eat, though tlie leaves 

 ill their glass were thoroughly dry, and some time towards the last of September, 

 passed the third moult. Not anticipating this I discovered the change accident- 

 ally, and was surprised at finding them active, and of a Ijrilliant green color. I 

 have since learned from Mr. Riley's paper, before cited, that he had watched 

 larv* of Chjton when this third autumnal moult was taking place, though his 

 experience agrees with mine that such instances are exceptional. These larvre 

 ao-ain chauii-ed color, but to nothing like the degree of the othui-s. l)ecoming 

 greenish-brown or vinous-brown, and no shrinkage of the body was perceptible. 

 And at intervals through the winter, a few moments in a warm room would 

 arouse them. Unfortunately two of these died before spring, and tlie third was 

 accidentally killed. 



On 9th May following, the shoots of the Hackberry beginning to put forth, I 

 brought the larva? from the cellar. Several were living and awaked soon after 

 exposure. Some began at once to move and eat, and of these part turned 

 green within the next twenty-four hours. Others i-enuxined quiet, and changed 

 gradually to pale vinous and then to green, taking three or four days to make 

 the comnlete chantjce. But others still occupied several davs more. 



On the loth those which had soonest Ijecome active showed the second seg- 

 ment swollen, and that day and the next they passed the third moult, while the 

 most tardy d^d not reach that stage till the first had begun to i)ass the fourth or 

 last moult, which took place on 23d and 24th. By 30th May these were in chrys- 

 alis, and the butterflies appeared on the 10th June and days following. Mean- 

 time some of the larva3 continued to grow and reached a size which seemed 

 enormous for this species. And it turned out that all these large larvoe produced 

 female butterflies, and no female appeared until nearly all the males had 

 emerged. I have not noticed this peculiarity, nor have I been able to distin- 

 guish the sexes by the size of the caterpillar, in any species l)ut the present one. 



After the fourth moult, each larva makes for itself a web on the sm-face of the 

 leaf, and draws the sides together until a sort of case is formed, within which it 

 lives. From tliis it emerges to feed at night, as was evident by leaves at a dis- 

 tance being constantly fed upon during the night. This is the habit in captivity, 

 and is probably but a modification of its habit when free. Mr. liiley states that 

 these larv^ scatter after the third moult, and thereafter hide within leaves drawn 

 around them. 



When preparing for the change to chrysalis, the body of the larva becomes 

 contracted, broadened and rounded in the middle, the head is bent under and 

 the last segment pressed to the leaf; the color of the skin changes to a nearly 

 miiform green, and the shape of the chrysalis gradually becomes ajipareut. Tiu' 



