not particular as to the material which they collect and store up for soil, provided that it is a material on which the fungus will grow—orange-peel, certain flowers, etc., being equally acceptable to them. But they are very particular regarding the ventilation of their underground storehouses, on a suitable degree of which the successful growth of the fungus presumably depends. They therefore have numerous holes or ventilating shafts which lead up to the surface from the storehouses or underground gardens, and these they either open or close according to the horticultural requirements as regards temperature and moisture. If the leaves are either too damp or too dry, they will not grow the fungus, and therefore in gathering the leaves the ants are very particular that they should neither be the one nor the other. Thus Bates observed:
If a sudden shower should come on, the ants do not carry the wet pieces into the burrows, but throw them down near the entrances; should the weather clear up again, these pieces are picked up when nearly dried and taken inside; should the rain, however, continue, they get sodden down into the ground, and are left there. On the contrary, in dry and hot weather, when the leaves would get dried up before they could be conveyed to the nest, the ants, when in exposed situations, do not go out at all during the hot hours, but bring in their leafy burdens in the cool of the day and during the night.
Dr. Ellendorf made the experiment of interrupting the advance of a column of these ants, with the interesting result which he thus describes in a letter to Büchner:
Thick dry grass stood on either side of their narrow road, so that they could not pass through it with the load on their heads. I placed a dry branch, nearly a foot in diameter, obliquely across their path, and pressed it down so tightly on the ground that they could not pass underneath. The first comers crawled beneath the branch as far as they could, and then tried to climb over, but failed owing to the weight on their heads. Meanwhile the unloaded ants from the other side came on, and when these succeeded in climbing over the bough there was such a crush that the unladen ants had to clamber over the laden, and the result was a terrible muddle. I now walked along the train, and found that all the ants with their bannerets on their heads were standing still, thickly pressed together, awaiting the word of command from the front. When I turned back to the obstacle, I was astonished to see that the loads had been laid aside by more than a foot's length of the column, one imitating the other. And now work began on both sides of the branch, and in about half an hour a tunnel was made beneath it. Each ant then took up its burden again, and the march was resumed in the most perfect order.
The operations here described show clearly that these ants act upon the principle of the division of labor. In this connection I may also quote an observation of Belt, which shows this fact in perhaps even a stronger light. He says:
Between the old burrows and the new one was a steep slope. Instead of descending this with their burdens, they cast them down on the top of the slope,
