in groups of two or three, and animated by special movements. A section of the chain along its axis of revolution gives the forms represented in Fig. 2, A. The superposed forms of the figure, A, are evidently only the successive phases of a single drop as emitted at each pulsation; and all our efforts should be directed to isolating the drops and retarding their descent, so as to make them
Fig. 2.—Forms of a Colored Drop penetrating a Colorless Liquid.
more visible under their different aspects. This is easily done. A variety of colored liquids are at our disposal, and, with a little alcohol, or a trace of glycerine or sugar, we may modify insensibly the density of the liquids, reduce the action of gravity to a minimum, and have at will ascending or descending currents. The last are generally easiest to obtain. Fig. 3 represents a simple apparatus for the experiment, the operation of which is dependent on the familiar principle of the siphon: we have only to raise the level of the colored liquid in the glass to change the rate of flow as we may wish. Certain manual difficulties in handling this apparatus may be obviated by using a system of communicating vessels, such as may be made by taking a common lamp-chimney, corking up the lower part and putting it in communication, by means of a pipe inserted in the cork and an India-rubber tube, with the bottom of a similar vessel or the nozzle of a vertical funnel. A string around the tube, or a pair of nippers, may serve to regulate the flow of the colored liquid, which should be only a little less dense than the other; and by inclosing a bubble of air above this Liquid and covering the. top of the chimney with a membrane, we may vary the pressure without changing the level of the liquid. If we have
