456 PROCEEDINGS OF THE AMERICAN ACADEMY. 



cent of times (the opportunities being 100 per cent) which ants of these 

 sizes entered into the activities of foraging (Figure 12), tending the 

 young (Figure 13), and building (Figure 14), respectively. The 

 figures below the middle of each class show the number of individuals 

 in that class. There seems to be no marked difference as to the 

 amount of foraging among the various classes, except that the largest 

 ants did not engage much in that activity. It might be objected that 

 the colonies should not have been classed together in this way, since 

 there might be differences in behavior of the classes of different col- 

 onies, depending on the number of ants of each size within the colony, 

 numbers which change with the age and amount of food of the colony. 

 But here the colonies, being all small, were in approximately the 

 same condition, and were kept in the same environment. Moreover, 

 such a combining of colonies would tend not to exaggerate but to ob- 

 literate differences in behavior of the different classes, and yet such 

 differences are more or less evident. 



2. Tending the Young. 



1. Methods. — In endeavoring to ascertain which ants were most 

 active in tending the young, both passages at the ends of the partition 

 in the Fielde nest were left open. The ants were placed in chamber 

 A, and this was left partially darkened until the ants had collected in 

 the darkest part of the chamber. Then the screen was removed, allow- 

 ing the daylight to fall upon the insects. In order to cause ants more 

 quickly to pick up the young, a gentle current of air was pumped into 

 the corner of the nest where the ants were congregated. The pump- 

 ing was done by the following method : A five-gallon glass carboy was 

 used as a compression chamber for the air. Into this the air was 

 pumped by means of a foot-bellows, connected to the carboy by a heavy 

 rubber tube. Another tube leading from the carboy terminated in a 

 piece of glass tubing drawn out to a fine point, which was inserted 

 through a small hole in one of the two roof-panes of the nest. 



Four colonies were used in this experiment. Each colony was ob- 

 served for 50 one-minute periods. During each period a record was 

 made of all ants seen carrying eggs or larvae (there were no pupae). 

 The average number ^ of workers in each colony was then found (Table 

 IV, line 3) and finally the per cent of workers which were active in 

 tending the young (line 4). In a similar manner, line 5 gives the 

 average number of winged queens in the colony, line 7 the average 



' See page 450, foot-note 1. 



