the foot of which was immersed in water. They then ran backward and forward along the paper, trying to escape. If a camel's-hair pencil be suspended just over the paper, they pass under it without taking any notice of it; but if it be scented, say with lavender-water, they at once stop when they come near it, showing in the most unmistakable manner that they perceive the odor. This sense appears to reside, though not perhaps exclusively, in the antennæ. I tethered, for instance, a large specimen of Formica ligniperda with a fine thread to a board, and when she was quite quiet I approached a scented camel's-hair pencil slowly to the tip of the antenna, which was at once withdrawn, though the antenna took no notice of a similar pencil, if not scented.
On the other hand, as regards their sense of hearing, the case is very different. Approaching an ant which was standing quietly, I have over and over again made the loudest and most shrill noises I could, using a penny pipe, a dog-whistle, a violin, as well as the most piercing and startling sounds I could produce with my own voice, without effect. At the same time I by no means would infer from this that they are really deaf, though it certainly seems that their range of hearing is very different from ours. We know that certain allied insects produce a noise by rubbing one of their abdominal rings against another. Landois is of opinion that ants also make sounds in the same way, though these sounds are inaudible to us. Our range is, however, after all, very limited, and the universe is probably full of music which we cannot perceive. There are, moreover, in the antennæ of ants certain curious organs which may perhaps be of an auditory character. There are from ten to a dozen in the terminal segment of Lasius flavus, the small meadow ant, and, indeed, in most of the species which I have examined, and one or two in each of the short intermediate segments. These organs consist of three parts: a small, spherical cup opening to the outside, a long, narrow tube, and a hollow body shaped like an elongated clock-weight. They may serve to increase the resonance of sounds, acting, in fact, to use the words of Prof. Tyndall, who was good enough to look at them with me, like microscopic stethoscopes.
The organs of vision are in most ants very complex and conspicuous. There are generally three eyes arranged in a triangle on the top of their heads, and on each side a large compound eye containing sometimes more than two thousand facets between them. Nevertheless, the sight of ants does not seem to be very good. In order to test how far ants are guided by vision, I made the following experiments: I placed a common lead-pencil on a board, fastening it upright, so as to serve as a landmark. At the base I then placed a glass containing food, and then put a L. niger to the food; when she knew her way from the glass to the nest and back again perfectly well, she went quite straight backward and forward, I then took an oppor-
