148 Mr. F. Merrifield's systematic temperature 



help to show how so great a change as that from single- 

 broodedness with its pupa of 7 months or more to double- 

 broodedness with its summer pupa of 2 or 3 weeks could 

 have been made. Assuming the autumn pupa to have 

 been the original one, it must have had an even stronger 

 disposition than the modern seasonal autumn pupa 

 not to emerge until after a winter had passed over it, 

 and the extreme reluctance of the modern seasonal 

 autumn pupa to anticipate its regular period of emer- 

 gence has been shown. Perhaps the species, or at 

 least the genus, may claim a very high antiquity, and 

 may have lived through several periods of advancing as 

 well as of receding cold ; but I cannot pursue this 

 speculation, for I know too little of such matters to 

 venture beyond the solid ground of fact more than the 

 very few steps which are the necessary incentive to 

 further investigation. 



Note, 6th March, 1890. — I can now complete the 

 history of the iced summer pupae of S. illustraria men- 

 tioned pp. 96—99. I had 12 living on the 28th 

 December, when they had been iced 22 weeks. Taken 

 out at successive intervals of 14 days, 5 of them died, 

 including the 2 taken out at 32 weeks. The survivors 

 show no increase of darkness over those iced 16 or 18 

 weeks, and have altogether a less vigorous appearance 

 than the great majority of those which emerged after 

 20 weeks' icing or less ; but I am doubtful if the longer 

 duration of the icing is the cause of this deterioration, 

 as the larvae of this brood that were late in pupating 

 were decidedly less healthy than the earlier ones. 



