*Expanding Globules.*—I now wish to show, by an experiment, how sensitive a floating globule is to disturbances in the existing tensions, which maintain it at rest. On the screen is projected a globule of dimethyl-aniline, floating tranquilly on the surface of water. I now allow a small drop of quinoline to fall upon it, and immediately it spreads out over the surface, forming a hole in its centre (Fig. 41), after which it gradually resumes its former shape. Sometimes the action is so violent that the globule is split up into several portions, which, however, join together again after a short time. In order to explain this action, we must again refer to the three tensions operating on the globule (Fig. 35). When in equilibrium, A is balanced by the joint pull of B and C; and hence if either of the latter be weakened, A will predominate and stretch the globule. In our experiment it is the interfacial tension, C, which has been diminished in strength, as we may now prove by a second experiment. In this instance I float on the water surface a globule of lubricating oil, with which quinoline does not readily mix, and which does not act so immediately as dimethyl-aniline. On allowing the drop of quinoline to fall into it, no action is observed until the drop has rested on the junction of the oil and water for a short time; but when it has penetrated the interface the oil globule suddenly spreads over the water surface, and with such violence as to detach several portions from the main globule. Merely touching the upper surface of the oil with a rod moistened with quinoline has no effect, and hence the result is due to the weakening of the interfacial tension. A similar effect is obtained when quinoline is dropped into a globule of aniline, and may be obtained with various other liquids.