178 Mr. Poulton's notes in 1885 upon 



The object in this case was to ascertain the complete 

 loss of weight during the process of pupation, as well as 

 in the periods immediately after the change. In the 

 following numbers the loss of the skin is included. A 

 contracted larva, evidently very near the change, was 

 selected and weighed, and after an interval of lO^^ hours 

 pupation had taken place, and the pupa and larval skin 

 weighed 70 milligrams less than the weight of the larva ; 

 in a further period of about 13 hours there was a further 

 loss of 57 milligrams, and a total difference of 193 

 milligrams between the larva on Aug. 26th and the pupa 

 with the larval skin on Sept. 3rd. This is larger than 

 the loss in the last pupa (IV.), and a much larger 

 percentage of the pupal weight. The difference approxi- 

 mately represents the loss during the actual process of 

 pupation. The loss in the 24|^ weeks was 50|- milli- 

 grams, thus following the most usual result obtained 

 from the previous pup^e. 



It must not be supposed that all the immense loss 

 indicated in the tables printed above is due to evapora- 

 tion from the surface, although doubtless most of it 

 bears this interi^retation. The active muscular efifort of 

 pupation must mean a large loss of water and carbon 

 dioxide through the tracheal system, but it also must 

 mean a gain of weight in the absorption of oxygen, 

 which is stored up in the body as the oxidised products 

 of nitrogenous metabolism. Such substances fill the 

 digestive tract of the pupa and the imago which is 

 formed within it, and are only discharged on the 

 emergence of the latter. In order to estimate the loss 

 of weight due to surface evaporation alone, and the 

 gain (if any) due to oxidation of the surface, the following 

 experiment might be carried out : — A pupa might be 

 taken immediately after pupation, and the spiracles 

 carefully painted over with daumar varnish or Canada 

 balsam. The pupa should then be placed in a dried 

 and weighed flask, and the weight of the whole ascer- 

 tained. At intervals air could be drawn through by an 

 aspirator, having been thoroughly freed from water and 

 carbon dioxide by passing through U -tubes before 

 entering the flask, while the water and carbon dioxide 

 (if any) could be estimated by U -tubes on the proximal 

 side of the flask. If the loss of the flask is less than 

 the gain in the latter tubes, the difference represents 



