thus become smaller. I will blow a bubble at one end of a glass tube, and leave the tube open at the other end; we shall thus have a small hole formed in one side of the bubble, which, if our theory is correct, will gradually contract and disappear. You now see on the screen images of the ends of two tubes. I have the power of cutting off one tube entirely from access to the other; and I do so now, so that you will, if you please, consider for the purposes of this experiment that tube only as existing at the extremity of which I shall blow the bubble. You now see the image of the soap-bubble, which as long as the tube is closed remains unaltered in size; I open it, and it now at once contracts and disappears. This, then, conclusively proves that the air in the bubble was compressed. I will now go a step farther, and show that the amount of this compression depends on the size of the bubble. If it be large, the air is not so much compressed as if it be small. Let us consider what would happen if I formed bubbles at the two ends of a tube. If they were of the same size, evidently—the one pressing the air in one direction, and the other pressing it in the other with equal force no effect would follow. If, however, one bubble were smaller than the other, and what I have said be true, the small one would compress the air within it, and drive it from left to right (say) with greater force than the other would tend to drive it from right to left: hence the air would flow from the small bubble to the large one; the large one would increase, and the small one diminish. The smaller the bubble, the more the air would be compressed; and thus the current would become greater and greater, until at last we should see the small bubble entirely disappear, the large one having absorbed all the air which it previously contained. I will try to show you this on the screen; first disconnecting the two tubes, I blow at their ends bubbles of unequal size. I will now place them in communication, so that the air can pass from the one to the other. You see, the small bubble contracts and the large one expands, and we thus learn that the pressure of the smaller or more curved bubble upon the air is greater than that of the less curved one.
I now come to the third property of liquids of which I wish to speak; and that is, that the surface of a liquid is generally either more or less viscous than the interior. With reference to the word viscous, you will find a familiar example of two liquids which differ very much in this property of viscosity in treacle and water. Take a vessel of treacle and a vessel of water, pour the liquids out, and note the different way in which they behave; the water flows out smoothly, one part slipping over another, whereas the treacle comes out in a great rolling mass, which seems to stick to the sides of the vessel. Again, put a spoon into a vessel of water, and move it through the liquid, you will find little resistance to its motion, the water seems to flow away to make room for it and closes in again immediately behind. Try the same experiment with the treacle and you will find the resist-
