Periodicity, 8fc. iqion the Development of Insects. 67 



by the great augmentation in tlie lossof weigiit, which after that time 

 continued still to increase. In No. IV. that augmentation is not 

 perceptible; the loss of weight increases a little until the 7th of 

 August, but after that time it again decreases. This increase and 

 decrease in the amount of the loss is naturally explained by the 

 rising of the temperature, but it is not so much as it should be, if 

 tlie impulsion were given for the formation of the perfect insect. 

 In No. v., on the contrary, the loss is at all times enormously 

 greater, showing, as I have remarked, the influence of the loss of 

 life. As I commenced my observation of these weights only last 

 year^ and only females came out of the two-year brood pupae in 

 that year, I have not obtained the weights of males of the two- 

 year brood. 



Tlie loss of weight has been found to be greater in the males 

 than in the females, showing a more extensive consumption of 

 matter by the males than by the females. In the last sixty-six 

 days before coming out of the pupa;, the mean loss per cent, on 

 the weights of the pupae (on the sixty-sixth day before they came 

 out) was in the one-year brood of males, 8*48 ; in the one-year 

 brood of females, 7*59 ; and in the two-year brood of females, 7*64. 

 The loss by respiration is, of course, also accompanied by 

 urinary secretions, consisting of the solid and fluid remaining 

 parts of the matter consumed during the pupa state. This secre- 

 tion is accumulated during the pupa state in a vesicular enlarge- 

 ment of the rectum. The perfect insect frees itself by the dis- 

 charge of this mass in leaving the pupa case, and if the pupae are 

 fixed in an upright position before the moths come out, the whole 

 quantity is generally left in the pupa case. It would be too long to 

 explain all the precautions taken to determine exactly the quantity 

 of urinary secretion during the pupa state, and I will give only 

 the ultimate results, showing also the greater amount of matter 

 consumed by the male pupae. The correspondence of the num- 

 bers of the different groups will, as I hope, speak sufficiently in 

 favour of the accuracy of the results. The mean quantity of 

 urinary secretion, during the whole time of the pupa state, has 

 also been expressed by the per-centage on the weights of the 

 jnipag, calculated on the sixty-sixth day before the day the perfect 

 insects emerged, as follows: — 



In males of the one-year brood . . . . 45*89^ 



In females ,, ,, .... 31*76 



In females of the two-year brood . . . . 35*64 



I have also taken the weights of the perfect insects. I will not, 



however, mention all the precautions necessary to obtain accurate 



E 2 



