Experiment teaches us, that E. is exhibited only on the surfaces of conductors. A brass ball is suspended by a silk thread, and covered with two hemispheres, which can be held by insulating handles, and which exactly fit it. A charge is then communicated to the ball so compounded. When the hemispheres are withdrawn, they are found take away all the E. with them, not the slightest charge being left in the ball. The same fact is exhibited by a hollow ball placed on a glass pillar, with a hole in the top. large enough to admit a proof plane to the inside. When charged, not the faintest evi dence of E. is found on the inner surface, however thin the material of the ball may be. The thinnest metal plate, when under induction, shows opposite electricities on its two. faces. We learn from these and numerous other experiments, that electricity is only found on the outer surfaces of conductors in an envelope of inappreciable thickness. This fact is quite in keeping with Faraday's theory of the action of dielectrics. Within a conducting body we cannot expect E., for the moment it appears in it, the particles. communicate their electricities to each other, and the electric state ceases. In a dielec tric they cannot communicate, and the charge remains. Hence the charge at the• conductor only appears at the junction of a conductor and dielectric.
We are also taught by experiment that the distribution of E. on the surface of insu lated conductors is influenced materially by their form. An electrified ball, for exam ple, exhibits tLe same density all round, for the resistance is sensibly the same on all sides of it. When, however, a conducting body is made to approach near enough to it, the density of the E. is found to be greater on the side on which the approach is made. This is proved by the aid of a proof plane.and an electrometer. When work is done in: drawing away the proof planes from the charged body, its potential, as tested by the electrometer, is proportional to the density of the charge at the point where it touched. The reason of this unequal distribution is obvious, froni the fact that the potential of the ball must be the same at every point. If, therefore, the resistance at one side be less than at another, the density there must be greater to maintain equality of potential: The disturbance of equal distribution here spoken of holds true only for short distances; the disturbing body, for instance, in the case under consideration, has to be brought very near before any inequality in the distribution of E. on the ball becomes manifest. It is to this concentration of E. on the side of the approaching conductor that we owe the electric spark; and it is as we near the striking or sparking distance that this disturb ance is revealed. The concentration or fixing of E. on the side of the thinnest and best dielectric, is particularly illustrated in the condenser (q.v.) and Leyden jar, whose action depends upon it; but in these the dielectric must be very thin to secure decided effect.
When a conductor somewhat in the form of a prolate spheroid is charged, and the elec tric density of the several parts tested by the proof plane, it is found to be least at the thickest part, and to increase towards either end; and the difference is found to be all the greater as each end becomes more and more pointed. It is found likewise that the electric density on a point is so great with a considerable charge as to destroy the dielec tric condition of the air, the particles of which become electrified, and carry by convec tion the charge of the point to surrounding conductors. We therefore learn that E. concentrates on points and projections. A similar reasoning with regard to the relations of potential resistance and consequent density bears here as in the previous case. It may be here remarked that the density of charge at any point regulates the amount of tension at that point on the molecules of the dielectric. The constraint which they expe rience in being charged, and which Faraday calls tension, can only be carried to a cer tain limit. When that is reached, the molecules are forced to be conducting, and the tension ceases.
Electrometers and words are generally taken as synonymous; electroscopes, however, should be applied to the instruments which give evidence of electrical potential without giving the exact measure of it; and electrometers to such as show both. Of late years, immense progress has been made in the construction of deli cate electrometers, chiefly to meet the demands for such in the working or testing of submarine cables. Sir William Thomson's quadrant electrometer and his absolute elec trometer, in point of exactness and delicacy, are a hundred-fold in advance of previous instruments. We shall here, meanwhile, describe the common forms of electric indica tors. The quadrant electrometer consists of a conducting-rod, generally of box-wood or brass, with a graduated semicircle attached above, in the center of which is a pivot for the rotation of a straw carrying a pith-ball at its outer end. It is used for a charge of high potential, such as that of the electric machine. When placed on the prime conductor of the machine, the whole becomes charged with ± E., and the ball is repelled first by the E. of the rod, and then by that of the prime conductor, the height to which it rises being seen on the semicircle. This is not an electrometer in the strict sense of the word, for although it tells us, by the straw rising and falling, when one potential is greater or less than another, it does not tell us by how much, the condi tions of its repulsion being too complicated for simple mathematical expression. It can show us, however, by the indicator standing at the same point, when the electric poten tial of the machine is the same at one time as another.