482 PROCEEDINGS OF THE AMERICAN ACADEMY. 



the number of observations, the second and fourth the average number 

 of soldiers and of minors respectively used in the experiment, and the 

 third and fifth columns the average per cents of soldiers and of minors 

 respectively which took part in this activity. 



It will be noticed that in fighting, the per cents of ants attacking an 

 enemy are fairly large (27.9, 18.7 ; 25.6, G.3) and that the soldiers were 

 much more active than the minors. 



10. Carrying and Being Carried. 



1. Methods. — These activities occurred so rarely that I have not ex- 

 perimented in regard to them. However, when I have seen one ant 

 carrying another, I have recorded the fact, as well as the class to which 

 each belonged, and the number of each class present in the colony at 

 the time of the observation, so that 1 might judge of the proportions, 

 which would be in some degree comparable with those in the other 

 activities. Though these observations were few, perhaps too few to 

 make it safe to draw conclusions from them when they stand by them- 

 selves, nevertheless, in connection with other observations on the same 

 activities, they may have some value. 



2. Observations. — Carrying was observed to take place in only one 

 colony (86) of P. pilifera. In this colony no soldiers were active in 

 either carrying or being carried. The per cent of minors (7.3 per cent) 

 engaged in carrying and being carried was therefore the same, and one 

 table answers for both. 



1 1. Partaking of Different Kinds of Food. 



1. Methods. — According to the accounts of Reichenbach ('96, p. 

 xcv) and Wheeler (: 02, p. 770), it seems probable that different 

 classes of ants have different tastes in regard to various kinds of food 

 and different ways of treating it. Accordingly, while making my other 

 observations on feeding, I have tried to discover whether one class 

 interests itself in certain kinds of food and the other class in other 

 kinds. To this end I recorded the class to which each ant seen at the 

 food belonged, and also the kind of food of which it was partaking. 



2. Observations. — One colony (86) of P. pilifera was observed, 

 and in this colony only one ant (a soldier) ate, its food being a meal- 

 worm. During these observations no ant ate grass-seed. That this 

 species does eat grain, and even grass-seed, I feel sure from AVheeler's 

 (^OS*^, p. 380) observations and from my own observations in the 

 field ; moreover Wheeler states that in the grain-storing species of 

 Pheidole, the soldiers act as seed crushers for the community. 



