488 



THE AMERICAN NATURALIST [Vol. LV 



mental differences in respect of food liave existed in our 

 experiments have been randomly distributed among the 

 different groups in any given experiment. AVe have had 

 very little trouble at any time with moulds in the cultures, 

 the frequent transfers in the duration of life bottles in 

 an experiment preventing them from getting any start. 



The stock bottles holding reserve stocks of flies have 

 been kept at the varying temperature of the room, but all 

 experimental flies have been kept in electric incubators 

 at 25° C, in which recording thermometers have been 

 placed to insure that no fluctuations of temperature have 

 occurred without our knowledge. All the experiments on 

 duration of life and their results recorded in this paper 

 have been carried through at the constant temperature 

 of 25° C. We have settled on this as a normal for this 

 particular element of the environmental complex. 



During the first year of the experimental work no at- 

 tempt was made to keep the different generations sep- 

 arate in the stock bottles, the process being merely to 

 keep enough bottles (generally 4) of each stock to insure 

 always having pupa^ and newly emerging flies on hand 

 for any matings and experiments to be started. Each 

 week all the flies old and young together from the oldest 

 bottle of each stock were transferred to a fresh bottle. In 

 this way each bottle was kept 4 weeks and tlu-r.' were al- 

 ways on hand bottles witli llie> in nil sta-vs of develop- 

 ment. 



In January, 1921, it was decided that it would be desir- 

 able to keep the generations separate in the stock bottles. 

 All stock bottles were emptied on January 11, and flies in 

 the stock bottles on January 14 were arbitrarily called 

 generation 0. From that time on the procedure has been 

 to empty out all the parent flies from each stock bottle 7 

 days after the bottle was started (when there are usually 

 a large num])er of larvce and some pupae formed). The 

 bottles aic then left for 7 days longer, during which time 

 enough tli.-. em. i-v to start a fresh bottle for the next 

 generation. Several bottles are kept of each stock, as 



