By continuing these experiments we arrive at a very important point. Instead of the body, b (Fig. 4), let us take c. As the figure shows, this is a sphere lighter than water, kept in the liquid by a weight. If we present to it the vibrating body, it will be repelled, and we shall obtain the results known by the name of diamagnetism. This curious experiment renders evident the influence of media. As well known, Faraday attributed such effects to the action of the air; and he thought that magnetic motions always resulted from a difference between the attraction exerted by the magnet upon the body under experiment, and the attraction exerted by the air. If the body is more sensitive than the air, there is direct magnetism, but if it is less so, there is diamagnetism. Water between the bodies, in the Bjerknes experiments, plays the same role; it is this which, by its vibration, transmits the motions and determines the phases in the suspended body. If the body is heavier than water its motion is less than that of the liquid, and, consequently, relatively to the vibrating body, it is of like phase; and if it is lighter, the contrary takes place, and the phases are in discordance. These effects may be very well verified by the aid of the little apparatus shown in Fig. 5, and which carries two bars, one of them lighter and the other heavier than water. On presenting to them the vibrating body, one presents its extremity and takes an axial direction, while the other arranges itself crosswise and takes the equatorial direction. These experiments may be varied in different ways that it is scarcely necessary to dwell upon in this place, as they may be seen at the Electrical Exhibition. [Illustration: FIG. 5.] Very curious effects are also obtained with the arrangement shown in Fig. 6. Between the two drums there is introduced a body sustained by a float such as represented at a, Fig. 4. Various results may, then, be obtained according to the combinations adopted. Let us suppose that the phases are alike, and that the interposed body is heavier than water; in this case it is repelled as far as the circumference of the drums, at which point it stops. If the phases are different, the influenced body behaves in the opposite manner and stops at the center. If the body is lighter than water the effects are naturally changed. Placed between two like phases, it is attracted within a certain radius and repelled when it is placed further off; if the phases are unlike, it is always repelled. We may easily assure ourselves that these effects are analogous to those which are produced on bodies placed between the poles of wide and powerful magnets. It is useless to repeat that the analogies are always inverse. [Illustration: FIG. 6.]