THE CANADIAN ENTOMOLOGIST. 275 



taken. Most of these were worn, but Mr. Winn saw a fresh specmien, 

 and others were seen and one secured on the 14th. 



These were evidently individuals of the first brood in descent from 

 hibernators or colonists, and assuming that the eggs were laid during the 

 first week of May, would allow about six weeks from egg to imago, which 

 corresponds with the experience of Mr. Edwards with the first brood in 

 West Virginia, which took 37 days — 28th April to 4th June. 



On 14th June Mr. Winn also observed two very much worn Fabricii 

 ovipositing on the young leaves of an elm. This late laying of eggs causes 

 the broods to overlap and makes it almost impossible to tell to what 

 generation any captured specimen belongs. 



From the 15th to the end of June Umbrosa was quite common, but 

 no more Fabricii were seen. On 24th a number of larvse, apparently not 

 more than a day old and quite close to the empty egg-shells, were found, 

 and on 25th about 40 eggs and seven young larvae were found on a bunch 

 of elm leaves plucked at random. These produced the imagos between 

 19th and 29th July and were 31 Umbrosa and two Fabricii, and were 

 doubtless part of the second brood of the season. 



On I St July Mr. Lyman took at Lachine a ^ Umbrosa and confined 

 it over leaves of elm, but no eggs were laid for over a week. 



On 1 2th July the butterfly was found to be dead, but had laid loi 

 eggs, some almost ready to hatch and some just recently laid. 



The eggs began hatching that same evening and others continued to 

 hatch during the 13th and 14th. Some of the earliest to hatch passed 

 first moult on the 15th, the third day from the egg. The first chrysalis 

 was formed on 5th Aug., and the first imago emerged on 13th Aug., 

 giving a pupal period of eight days, a period from hatching of egg to 

 imago of 32 days, and a probable period from oviposition to imago of 35, 

 or, at the outside, 36 days. 



Some, of course, took a few days longer than this, but all had 

 emerged by the 21st August. Of nearly 60 butterflies which emerged, 

 not more than five were Fabricii, all the others being Umbrosa. 



Now it seems clear that the parent butterfly which was taken on ist 

 July, but would not lay till 8th or 9th, must have belonged to the first 

 brood in descent from the hibernators or colonists, whichever the early 

 ones were, and that the brood thus reared represented the second brood, 

 and there would be abundance of time after the. 21st August for a third 

 brood to mature. That such a third brood must exist is practically 



