28 THE CANADIAN ENTOMOLOGIST. 



trophi of the ^rst form, and the other and larger one of the second form. 

 It had therefore been in its mine within a few hours, more or less, of seven 

 days, and had undergone two moults there. On the next Friday, July 

 23rd, I found that the larva was not feeding, so I unrolled its cone and 

 found in it two more casts, one of which was still fresh. Here were there- 

 fore two moults luithin the week, making four in two weeks from the egg, 

 and the larva was a little more than half grown It eats much more and 

 grows much more rapidly in its last stage. When taken from its roll, as 

 just stated, it was placed upon a fresh leaf, where in a little while it had 

 curled up the edge of the leaf, fastening it all around by a web instead of 

 by the little cords as in its first roll. On Monday morning early, July 

 26th, it was still in its roll, but three hours later it had left the roll and 

 begun its cocoon at the edge of the leaf. This is unusual, as the mine 

 when the larva is free is placed over the midrib. The contraction of the 

 silk curled the edge of the leaf over the cocoon. The pupa was dis- 

 closed early in the morning of the 28th July, about two daj's after the 

 cocoon was begun, but less than a day after the cocoon was finished ; the 

 four previous moults each occupied about twelve hours, as I was able to 

 determine approximately by observing whether or not the larva was feed- 

 ing. There are thus five larval stages. The entire larval life 

 before beginning its cocoon is about seventeen days, within a 

 few hours more or less, giving nineteen days as the entire larval 

 life. This is about the length of larval life in Lithocolletis and Leu- 

 canthiza. That of Phyllocnistis is not known ; that of some species of 

 Nepticula (a genus very far removed from the others above named) is 

 sixteen days in some species, only a week in some others, and probably 

 even less in some others. In the larva whose history I have given above 

 the imago was disclosed Aug. 4th, so that the pupa state lasted just a 

 week, or just twice as long as any single larval state. The same rule 

 holds in Lithocolletis and Leucanthiza. In Phyllocnistis we do not know 

 the length of the larval stages, but the pupa state lasts eight days. I 

 refer only to the summer broods, of course. In some species of other 

 genera, as Nepticula pteliceella, not yet described, the first and second larval 

 stages each last three days, whilst the third (and last one) lasts just twice 

 as long — six days, and has just double the rate of growth, so that it looks 

 as if a regular moult had been skipped. Is the pupa state likewise the 

 equivalent of two larval stages ? 



