1901.] I NATURAL SCIENCES OF PHILADELPPIIA. 431 



larvie succeediug the first iu each brood correspouded closeiy with 

 the times of the deposit of eggs succeeding the first laid. Various 

 broods, removed to weak alcohol or hot water aud examined under 

 a len8, showed the larva well formed iu the egg at about the seven- 

 teenth day and no earlier. Broods iu which the first larva 

 ajjpeared on the nineteenth day were immersed iu alcohol and 

 examined under a lens, and they always showed earlier, but never 

 later, stages of larval existence ; while broods in which no larva 

 had appeared on the twentieth day showed, when examined in like 

 manner, a larva perfectly formed within the egg membrane. I 

 therefore conclude that the period of incubation varies between 

 seventeen and tweuty-two days, with nineteen days as the common 

 period. The variation in the ])eriod of incubation bears no fixed 

 relation to the size of the future adult. Eggs of different periods 

 of incubation followed to the adult form were found to produce 

 the same sort of worker. 



]My ants have furnished no evidence that they ever devour the 

 eggs, larvie or pupje of their own colony. One worker, isolated in 

 a Petri cell twenty-one days without food, died leaving five eggs 

 intact during the last sixteen days of her starvation. In all the 

 score of Petri cells in which I have for months watched the condi- 

 tion aud counted the numbers of the eggs, no diminution of them 

 could be logically charged to the mature ants, whose skill and 

 diligence in keeping them clean, safe, dry and iu humid darkness, 

 merits high renown. 



The feeding of the larva, which is bent nearly double in the 

 egg, with regurgitated food begins as soon as it straightens itself and 

 protrudes its mouth. When the larvre begin to appear in the egg- 

 pacKet, the workers lift the packet and hold it free aud still, while 

 one of their number holds a translucent white globule of regurgi- 

 tated food to the larval mouth projecting from the surface of the 

 egg-packet. I have repeatedly seeu the workers thus feeding the 

 very young larva?, a single globule of regurgitated food serving 

 for a meal of which four or five larvi^e successively partook. 



When the larva first emerges, its leugth is nearly double that of 

 the egg. When well fed its growth is rapid and iu a day or two 

 its length is three or four times that of the egg. When about 

 two millimeters loug it is usually removed from Ihe egg-packet 

 and laid on the floor, or associated with others of its size iu a 



