A horse was laden with two tubs for carrying a supply of water, and in the bottom of the tubs a valve was fixed. When the horse entered the stream the tubs were partly immersed; the water then exercised its upward pressure, the valve opened, and the tubs slowly filled. When they were nearly full the horse turned round and came out of the water; the pressure had ceased. Thus the action of the water first opened the valve, and then closed it. The particular phenomena observable in the water level in narrow spaces, as of a fine glass tube, or the level of two adjoining waves, capillary phenomena, etc., do not need any special appliance for demonstration, and it is the same with the convexity or concavity of meniscuses. Fig. 65 represents a pretty experiment in connection with these phenomena. I take a glass, which I fill up to the brim, taking care that the meniscus be concave, and near it I place a pile of pennies. I then ask my young friends how many pennies can be thrown into the glass without the water overflowing. Everyone who is not familiar with the experiment will answer that it will only be possible to put in one or two, whereas it is possible to put in a considerable number, even ten or twelve. As the pennies are carefully and slowly dropped in, the surface of the liquid will be seen to become more and more convex, and one is surprised to what an extent this convexity increases before the water overflows. The common _syphon_ may be mentioned here. It consists of a bent tube with limbs of unequal length. We give an illustration of the syphon (fig. 66). The shorter leg being put into the mixture, the air is exhausted from the tube at _o_, the aperture at _g_ being closed with the finger. When the finger is removed the liquid will run out. If the water were equally high in both legs the pressure of the atmosphere would hold the fluid in equilibrium, but one leg being longer, the column of water in it preponderates, and as it falls, the pressure on the water in the vessel keeps up the supply. [Illustration: Fig. 66.—The Syphon.] Apropos of the syphon, we may mention a very simple application of the principle. Cut off a strip of cloth, and arrange it so that one end shall remain in a glass of water while the other hangs down, as in the illustration. In a short time the water from the upper glass will have passed through the cloth-fibres to the lower one (fig. 67). This attribute of porous substances is called _capillarity_, and shows itself by _capillary attraction_ in very fine pores or tubes. The same phenomenon is exhibited in blotting paper, sugar, wood, sand, and lamp-wicks, all of which give familiar instances of capillarity. The cook makes use of this property by using thin paper to absorb grease from the surface of soups.