which evaporates very rapidly, and therefore above the surface of the liquid there is a small atmosphere of alcohol formed by evaporation from the wine. This evaporation goes on in the drops as well as in the main body of the wine, but more rapidly at the upper surface of a drop than at the lower; the reason being that the lower part is more completely immersed in the little cloud of alcohol which hangs over the wine itself. A drop is thus formed composed in the upper part of water, with comparatively little alcohol, and in the lower part of water with a larger proportion of alcohol. The experiment I have just shown proves that the tension in the upper part must be greater than that in the more alcoholic or lower portion of the drop. Hence you may often see drops of wine actually running up the side of a glass, in obedience to a force exerted in an upward direction, by the greater surface-tension of the portion containing the larger percentage of water. A very important deduction may be made from the fact that the surface of a liquid is in a state of tension. You saw in the first experiment that, as soon as the film on the one side of the thread was broken, that on the other side contracted very rapidly, and took up a form with as small a surface as was possible under the circumstances. If we were to generalize from this particular instance, we should be led to the conclusion that, because the surface of a liquid is in a state of tension, and thus each part of it is tending to contract, therefore it will always assume a shape which will have the smallest possible surface. I will now show an experiment to illustrate this fact. I have here a tube formed of four plane pieces of glass, through which I shall be able to send light, and so show you the image of its contents on the screen. It forms, in fact, a little box of glass, the ends of which are open, one of them being considerably narrower than the other. I now put the larger end of this tube into a mixture of soap and water, and I withdraw it with a film adhering to it. This film has a tendency to assume that shape which has the smallest possible surface; and evidently, by moving up the tube toward the narrow end, its surface can be made smaller than it is at present. You now see on the screen an image of the tube. I move it for a moment in order to form the film. You now see the image of the film, and I think you will observe that it is slowly moving up the tube, and therefore that its surface is becoming smaller and smaller. The experiment might be prolonged until the film burst; but, at all events, you have there sufficient proof that it moves into a position in which its surface is diminished. And, further, inasmuch as the image of the tube is turned upside down on the screen, what appeared, to you to be a motion from above to below was, in fact, a motion from below to above; the film was in reality moving upward, although to you it appeared to be moving down. It was really raising its own weight instead of being pulled down by it. We have thus now established this quality of
