684 Bulletin American Museum of Natural History. [Vol. XXIII, 



or scattered. He also saw the ants eating the fungus and was able to satisfy 

 himself that the different species of Atta will eat the Kohlrabi from one 

 another's colonies but not that of Apierostigma or Cyphomyrmex. He 

 gives the following interesting description of the way in which the leaves 

 are comminuted by the workers. "The manipulation of the pieces of 

 leaves is the same in all the Atta species and the following description holds 

 good uniformly for all of them. The ant first cuts the leaf it has brought 

 in through the middle and then busies itself with only one of the halves, 

 cutting off another piece, and so on. When the piece of leaf which it has 

 retained is sufficiently small so that it can be turned round and round be- 

 tween its fore-feet with the aid of its jaws, it is felt of on all sides and turned 

 in all directions as if the insect wished to get a clear idea of its form. Then 

 an even smaller piece is cut off and this is repeated, till the piece that is 

 retained is hardly longer than the ant's head. The rejected pieces are 

 picked up by other workers and treated in the same manner. Then the 

 ant holds the little piece between its fore-feet with the sharp edge directed 

 towards its mouth and begins to pinch its edges at short intervals around 

 the circumference without ever cutting through the substance. The piece 

 thus manipulated shows fine, radial ridges under a good lens. The surface 

 of the leaf is also abraded with the points of the mandibles, wounded, so to 

 speak, so that it soon becomes soft. Then the ant kneads it with the feet 

 and again inserts her jaws into the pellet thus formed in order to mould it 

 thoroughly. Again and again the jaws close upon the pellet while the feet 

 press it and place it in a new position, and again it is kneaded. This ma- 

 nipulation is carried on with great care and deliberation, and I have several 

 times observed that an ant will spend a quarter of an hour in making such 

 a pellet. When it has become a soft mass, the worker takes it in her jaws 

 and seeks a suitable spot for it in the portion of the garden that is just being 

 built. Once I saw an ant that had found such a spot, actually jab the pellet 

 into the garden with a jerk of her head and a simultaneous opening of her 

 jaws, and then carefully pat it down with her fore-feet. Another time a 

 worker laid her pellet in a breach of a newly erected circular wall, and then 

 shook and pushed it into the depression, like a mason setting the last brick 

 in a fresh layer of mortar. During all of this work, the antennte are contin- 

 ually moving and palpating the pellet just as they are while the ant is feed- 

 ing." Into the new material thus added to the garden the fungus hj^phse 

 grow very rapidly. By afternoon pellets built in during the morning hours 

 have become permeated in all directions with mycelium. Belt supposed 

 that the smallest workers or minims comminute the leaves and build up 

 the fungus gardens. According to Moeller, hoAvever, this is the office of 

 the mediae, as the leaves are too thick to be manipulated by the smallest 



