276 THE CANADIAN ENTOMOLOGIST. 



proved by the fact that the second brood as raised by us was almost 

 entirely composed of the form Umbrosa, while it is well known that 

 Fabricii largely predominates in the autumn, which would not be the 

 case if there were no third brood. 



On 26th July, while Mr. Winn's second brood was emerging, he 

 confined a 5 Umbrosa on elm and obtained eggs the same day, which 

 hatched on 30th. Others were caged on 28th and five more on 2nd Aug., 

 on hop, and many eggs were obtained. Some were left on the food plant, 

 but the others were taken on a holiday trip to Metis, Q., the last hatching 

 7th Aug. On Aug. 24th the first chrysalis was formed, and imago emerged 

 4th Sept. and proved to be Fabricii, but at the same time a number of 

 the larvae were just past the third moult. While at Metis the larvae 

 were fed on hop, as elm trees were not found, and when brought back 

 to Montreal were again fed on elm. 



Either from this change of diet or from the colder climate of the 

 lower St. Lawrence, the majority of this brood were greatly retarded and 

 emerged at intervals all through September, and one as late as i8th Oct. 

 Of nineteen individuals seventeen were Fabricii and two Umbrosa. 



One fresh Umbrosa was also seen on i6th Sept., and Fabricii was 

 common on the fine days of the early part of that month. 



This makes the third brood, with a varying preparatory life duration 

 of 40 to 77 days. 



With Mr. Edwards the period of the third brood varied from 31 to 

 probably over 50 days. 



In nature the oviposition of the various broods would doubtless be 

 extended over a longer time and the emergence of the imago similarly 

 spread out, but when a species can go through all its changes in from 31 

 to 36 days it stands to reason that there must be at least three broods in 

 the season in this latitude. 



The third brood must certainly hibernate, and Mr. Winn found that 

 those flying in Se[)tember did not seem inclined to lay eggs, and careful 

 search failed to produce a single one. 



In Can. Ent., X., p. 72, Mr. Edwards states his belief that the 

 scarcity of hibernalois in the spring compared with the abundance of the 

 species in the summer is due to the existence of the species being 

 dependent on the partial fourth brood, which he considers the only one 

 that hibernates, and states that the species does not suffer from parasites 

 to any extent. 



