A glass plate, if placed between the two ends of the arms of the discharging rod, is pierced by the discharge: to prevent the electricities uniting by making the tour of the plate, the surface of which is always more or less moist, and by this means slightly conductible, it is necessary that the portion through which we would make the discharge pass, should be surrounded on its two faces by a belt of wax in the form of a ring; a drop of oil, placed in the centre of the ring, facilitates also the success of the experiment. A piece of dry wood, or a stone, may in like manner be split into fragments by a discharge traversing them. A card is also pierced; but if the extremity of the two rods that lead the discharge are not exactly opposite to each other, the hole is made opposite to the point that brings the negative electricity. It is also to be remarked, that the hole presents burrs equally on either side, as if the electric fluid had come from the middle of the card to escape by its two faces at the same time. This very extraordinary effect has greatly engaged the attention of philosophers; it appears to arise, as we shall see, from the electric discharge not actually occurring by a finite movement of translation, but rather by a series of small molecular vibratory

movements.

When the air is traversed by an electric discharge, it undergoes a very marked agitation; and an instantaneous expansion, if the phenomenon occurs in a closed vessel. This may be proved by means of Kinnersley's Thermometer (Fig. 64.), which consists of a glass tube, closed at its two ends, and of a lateral tube, in which a liquid rises that gives the measure of the expansion. The discharge is made between two metal balls, that penetrate into the tube.

With regard to the luminous effects, properly so called, and which are not due to a simple incandescence Fig. 64. of wires, they are generally manifested under the form of a spark, nearly similar in intensity, but greater than that which is drawn from an electrical machine, and susceptible, like the latter, of presenting the very varied appearances, that we shall study hereafter. We shall confine

ourselves to quoting here a beautiful experiment, which consists in passing the discharge of a battery through a metal chain of several yards in length, suspended by silk cords from the ceiling of a room: at the moment of the discharge, this chain is illuminated in its whole length by the effect of the sparks passing from one link to the other.

CHAP. V.

THEORY OF STATIC ELECTRICITY, AND DIVERS FACTS CONNECTED WITH IT: DI-ELECTRIC BODIES.

Properties of the two Electric Fluids.

WE have thus far explained all the phenomena of static electricity, setting out on the supposition that electricity is composed of two extremely subtile and imponderable fluids; that the particles of each of these fluids mutually repel each other, whilst those of one of the fluids attract those of the other. The experimental laws that we have established enable us now to enter with more precision than we have done into the properties of the two fluids. Thus, by attributing to the fluids themselves the property that we have recognised in the bodies that contain them, we are able to admit that the force with which the particles attract and repel each other is in inverse ratio to the square of the distance by which they are separated. We must in like manner lay down as a principle, that bodies, in the natural state, contain an equal quantity of both fluids in each of their particles; and that if they cannot exercise any action on surrounding bodies, it is that at the same distance the attractive power of one of the fluids is equal to the repulsive power of the other. We can demonstrate this principle experimentally by rubbing together two glass plates, the one polished, the other roughened. We bring an electroscope near to them, after having rubbed them, but without separating them, and we observe no effect; we separate them, and then find on the roughened face of one of the plates a strong charge of negative electricity, and on the polished surface of the other a strong dose of positive electricity. These two surfaces had been electrised by their mutual friction; the one had acquired a certain quantity of positive electricity, and the other an equal quantity of negative; but the insulating property prevented the two electricities neu

:

tralising each other, notwithstanding the contact of the two faces the nullity of the exterior action, so long as the contact remains, is therefore the proof that the two electric strata situated at a same distance from any body, if they are equal and of the contrary nature, exercise upon this body two effects that neutralise each other.

Theoretic Explanation of the Distribution of the Electric Fluids on the Surface of Insulated conducting Bodies.

The general fact, established by experiment, that electricity distributes itself entirely on the surface of an electrised conducting body, is a natural consequence of the mutual repulsion of all the particles of the same fluid, and of the facility they possess of moving in the body. When arrived at the surface, they accumulate; and, if they do not quit it, it is because they are retained there by the air or by the insulating supports, which oppose themselves to the tendency they would have to escape by virtue of their mutual repulsion. They therefore form there an excessively thin stratum, which does not sensibly penetrate within the surface of the body; for, however thin the metal envelope of a hollow sphere may be, electricity is never found there on touching it interiorly. With regard to the stratum, its exterior surface is evidently the same as that of the body, but its interior surface varies with the form of the conducting body. In a sphere it is similar to the exterior, because all points have the same electric charge; whilst in non-spherical conductors, upon which, as we have seen, the distribution of the electricity is never uniform, we may suppose, that to the most highly charged elements of the surface corresponds a stratum either of greater thickness or of greater density. As considerations of another kind seem to prove that the electric fluids are incompressible, we prefer to admit the former of these two hypotheses; thus the electric stratum would be terminated in the interior by a surface, whose form would be slightly different from that of the exterior surface, at least when nonspherical bodies are in question. This form may be established à priori, setting out, as we have done, from certain principles

of mechanism, and from the definite and precise properties of the electric fluids. As thus:

The stratum is in equilibrio; for its state does not change as long as the body preserves its electricity, and remains protected from all exterior influence. The figure of the stratum, therefore, is that which results from the equilibrium of the repulsive forces of all the molecules of which it is composed, by supposing them subject to the law of the inverse of the square. This is not all: it must be such that the stratum exercises neither attraction nor repulsion; or, in other terms, no action upon any point placed in the interior of the body. In fact, if it exercised an action, it would decompose the natural electricity of the point upon which it acted; this electricity, developed by induction, would react in its turn upon that of the body, and there would be a change in its electric state, which does not take place, since the introduction of a new body into the interior of an electrised body in no degree changes its electric state. Thus the equilibrium can subsist only as long as the resultant of all the repulsive forces upon an interior point is equal to zero. By means of these conditions we are enabled to determine the relations which ought to exist between the intensities of the repulsive forces in different points of the surface of an electrised conducting body; or, which comes to the same thing, the thicknesses of the electric stratum; and we obtain results perfectly in accordance with those, to which Coulomb arrived by experiment. In this manner we find, for a sphere, that the stratum is terminated interiorly by a spherical surface; for an ellipsoïd, that it is by an ellipsoïd surface, whose greater axis is to its smaller in exactly the same relation as those of the exterior surface; so that the stratum goes on increasing in thickness from the extremities of the small axis, where it is at its minimum, to the extremities of the large axis, where it is at its maximum.

By applying mathematical analysis to the principles of mechanics and physics that we have just laid down, M. Poisson has succeeded in determining, à priori, the thickness of the electric stratum for the different points of the surface of a conducting body of any form; he has done the same for two,