One might, at first sight, be disposed to consider that the ants with stings must have a great advantage over those with none. In some cases, however, the poison is so strong that it is sufficient for it to touch the foes to place them hors de combat, or at least to render them incapacitated, with every appearance of extreme pain. Such species have the abdomen unusually mobile.
The species of Lasius make up in numbers what they want in strength. Several of them seize an enemy at once, one by each of her legs or antennæ, and, when they have once taken hold, they will suffer themselves to be cut in pieces rather than let go.
Polyergus rufescens, the celebrated slave-making or Amazon ant, has a mode of combat almost peculiar to herself. The jaws are very powerful and pointed. If attacked—if, for instance, another ant seizes her by a leg—she at once takes her enemy's head into her jaws, which generally makes her quit her hold. If she does not, the Polyergus closes her mandibles, so that the points pierce the brain of her enemy, paralyzing the nervous system. The victim falls in convulsions, setting free her terrible foe. In this manner a comparatively small force of Polyergus will fearlessly attack much larger armies of other species, and suffer themselves scarcely any loss.
Much of what has been said as to the powers of communication possessed by bees and ants depends on the fact that, if one of them in the course of her rambles has discovered a supply of food, a number of others soon find their way to the store. This, however, does not necessarily imply any power of describing localities. If the bees or ants merely follow their more fortunate companion, or if they hunt her by scent, the matter is comparatively simple; if, on the contrary, the others have the route described to them, the case becomes very different. To determine this, therefore, I have made a great number of experiments, of which, however, I will here only mention a few. Under ordinary circumstances, if an ant discovers a stock of food, she carries as much as possible away to the nest, and then returns for more, accompanied generally by several friends. On their return, these bring others, and, in this way, a string of ants is soon established. Unless, therefore, various precautions are taken—and this, so far as I know, has never been done in any of the previous observations—the experiment really tells very little.
I therefore made the following arrangement: One of my nests of the small brown garden ant, Lasius niger, was connected with a board, on which I was in the habit of placing a supply of food and water. At a short distance from the board I placed two glasses (b b'), and on b I placed some food. I then connected the glass b with the board a by three slips of paper, c, d, e, and put an ant to the food. She carried off a supply to the nest, returning for more, and so on. Several friends came with her, and I imprisoned them till the experiment was over. When she had passed several times over the
