THE CANADIAN ENTOMOLOGIST. 



217 



yet there is some discrepancy in the records of the two Lirvixj which I 

 succeeded in raising from the egg. Since one cannot be positive about- 

 what was not seen, I give the records just as they stand in my notes — 

 records of what was seen. For brevity I use the word "up"' to indicate 

 tliat the larva lias taken its position on the silken mat preparatory to 

 moulting ; the word ''off" to indicate that the skin has been cast; and the 

 word '-fmar' for the last mat to which the chrysalis is to be attached. 

 Where the time is marked with an asterisk it indicates exactitude ; else- 

 where approximation made by halving the time which elapsed between 

 observations. As these were taken at least four times daily (often more), 

 there is no possibility of any approximate record being more than three 

 hours out of the way. 



1905. 



1906. 



It will be seen that the 1905 specimen developed rapidly, completing 

 two moults in eleven days and attaining a length of 5 mm. The ultimate 

 stage was unusually long (more than two weeks) and the larva was three 

 days on its mat before it became a chrysalis. On the other hand, the 

 1906 larva grew to a length of only 4.6 mm. in thirteen days, was only a 

 week in final stage, and became a chrysalis in two days and six hours. 



When nearly full-grown the caterpillar seems to prefer a diet of leaves, 

 and for a day or two refuses the fruit, after which it crawls to the ground, 

 and though several days may pass before it finds a spot suitable for the 

 change to chrysalis, it does not again touch food. The extreme length 

 attained is 17 mm., but the larva decreases rapidly in size during this 

 walking tour, as is the case with iriis. 



PupatioJi. — The caterpillar pupates among the dried leaves and dead 

 grass on the ground, selecting a dark coloured surface in some protected 

 spot. No ''cocoons" were formed by any of those raised by me, although 

 they were offered the same opportunities accorded iriis^ which does form 

 such a shelter. 



(To be continued.) 



