STUDIES ON THE DURATION OF LIFE 191 



of descent. If, when kept under identical environmental 

 conditions, such lines exhibit widely different average 

 durations of life, and if these differences reappear with 

 constancy in successive generations, it may be justly 

 concluded that the basis of these differences is heredi- 

 tary in nature, since by hypothesis the environment of 

 all the lines is kept the same. In consequence of the 

 environmental equality, whatever differences do appear 

 must be inherently genetic. 



The manner in which these experiments are performed 

 may be of interest. An experiment starts by placing 

 two flies, brother and sister, selected from a stock bottle, 

 together in a half -pint milk bottle. At the bottom of the 

 bottle is a solidified, jelly-like mixture of agar-agar and 

 boiled and pulped banana. On this is sown, as food, some 

 dry yeast. A bit of folded filter paper in the bottle fur- 

 nishes the larvae opportunity to pupate on a dry sur- 

 face. About ten days after the pair of flies have been 

 placed in this bottle, fully developed offspring in the 

 imago stage begin to emerge. The day before these off- 

 spring flies are due to appear, the original parent pair 

 of flies are removed to another bottle precisely like the 

 first, and the female is allowed to lay another batch of 

 eggs over a period of about nine days. In the original 

 bottle there will be offspring flies emerging each day, 

 having developed from the eggs laid by the mother on 

 each of the successive days during which she was in the 

 bottle. Each morning the offspring flies which have 

 emerged during the preceding twenty-four' hours are 

 transferred to a small bottle. This has, just as the 

 larger one, food material at the bottom and like the larger 

 one is closed with a cotton stopper. All of the offspring 

 flies in one of these small bottles are obviously of the 

 same age, because they were born at the same time, 



