36 Proceedings of the Biological Society of Washington. 



keep notes. I did not use a breeding cage, but left the plant entirely 

 open, trusting that my caterpillars would not leave the food plant. Five 

 of the eight were lost when so small that it was difficult to follow their 

 movements, but the other three grew to maturity, passed through all 

 stages and became butterflies. 



The eggs hatched August 13, three days from the time they were laid. 

 The caterpillars hung for the chrysalis August 28 and 29, and the chrys- 

 alids were formed August 29 and 30, after fifteen to sixteen days in the 

 caterpillar stage. The chrysalids showed first change of color on September 

 2 and 3, and the butterflies emerged September 4 and 5, after only seven 

 days in the chrysalis. These periods are all remarkably shorter than 

 those given by Scudder who gives for the egg, six days; for the caterpillar, 

 one month; and for the chrysalis, thirty to thirty-eight days. 



The conditions under which the caterpillars grew were, I believe, very 

 nearly natural ones. The caterpillars were kept on the sill of an open, 

 though screened window, on the south side of the house, where they 

 obtained sunshine during the greater part of the day. When the chrys- 

 alids were formed I moved them, and fastened them to the wall of the 

 room a few feet from the open window. There was no artificial heat 

 in the room, and none in the house except a kitchen stove on the floor be- 

 low, and on the opposite side of the house. The weather, during the 

 time of the egg stage, was unusually hot. I have no notes or memory 

 on weather conditions during the rest of the time. 



At the time my butterflies emerged from their chrysalids, butterflies 

 of this species were still flying out-of-doors, and caterpillars only a few 

 days old were easily found. I believe that the earlier butterflies of the 

 second brood to appear, can and do produce a third brood, which comes 

 early in September; that this brood appears while the second brood is 

 still flying; and that it is through these earlier butterflies that the species 

 is able to exist in southern New England. The differences in the time 

 occupied by the early stages, shown when Scudder's observations are 

 compared with mine, is simply due to weather conditions. Those in- 

 sects that pass through the early stages in mid-August encounter decidedly 

 warm weather, while those that wait until late August and September 

 meet colder conditions and develop more slowly. 



