to

and lengthening the time of the contact on the copper side up four seconds, I found, that by twenty such contacts the electroscope was affected negatively to a certain measurable quantity.

3. Inverting the little pile, I had no sensible effect by these contacts of four seconds on the zinc side, and it was necessary to lengthen the time to eight seconds; but in order to produce on the electroscope a positive effect equal to the negative of the preceding trial, I was obliged to make forty contacts, on account of the dissipation of the effects on the condenser in each interval of time. This experiment shows however, sufficiently, that the zinc side of each group yields, through the paper, to the next group, some of the electric fluid that it takes from the copper with which it is associated.

We have from these experiments all the elementary principles necessary for the motion of the electric fluid in the pile, and they are the following:-1. In each binary group, the zinc plate takes some electric fluid from its associate the copper; the latter in my new pile, being the coppered side of the Dutch gilt paper. 2. In each group also, zinc communicates, through the paper, some of its excess of fluid to the copper of the next group on its side. 3. In each group again the copper takes, through the paper from the zinc of the next group on its side, some of the fluid that it has lost to its associated zinc. The same effects taking place in every group, with the next on both sides, along the whole pile, these effects are successively added to those that the respective next groups have already undergone according to their place; and thus the negative state goes on increasing from one end to the other of the pile, toward what is called the copper extremity; and the positive state is increasing toward the zinc extremity.

These effects may be represented by numbers; though, from the great variations in the quantity at different times, and the imperfection of the electroscopes, these numbers remain undetermined: I shall express them in a pile of eleven groups, indicating by A the zinc side, and by B the copper side. The two following series represent the progress of negative and positive effects above mentioned, which, combining in each successive group from A to B, constitute, according to circumstances, the three different states of the pile.

A
0

A

+10

In the insulated pile, when the divergence of the electroscopes is equal on both sides, positive at A, and negative at B, the state of each successive group is the sum of the corresponding numbers of the above series, as expressed in the 1st of the following Tables. When B is placed in communication with the ground, the losses of all the copper plates being repaired by the latter, all the acquisitions of the zinc plates subsist without diminution, which requires the quantity ten, to be added to each number of Table I, as expressed in Table II. When A communicates with the ground, all the acquisitions of the zinc plates being carried into the latter, the losses of the copper plates remain uncompensated, and the same quantity, ten, is to be substracted from all the numbers of Table I, as expressed in Table III.

Now this synthesis of the above fundamental experiments is the real fact; as will be seen from direct experiments in the following paper, of which I shall give here only the general results.

For these experiments I use a horizontal pile, which I have called column; with a gold leaf electroscope at each extremity; and I have also a detached electroscope, which may be applied and observed at every point of the column. The following are the observed phenomena: 1. When the state of the ambient air is such, that in the insulated column the divergence is equal at both extremities, the middle point in its length is zero as represented in Table I, in which, as well as in the two others, the terms are to be considered only as equidistant points, whatever be the number of the groups. 2 When B communicates with the ground, the first plate only at this extremity is zero, and the positive state is gradually increasing towards A: the middle point is plus, of the same quantity (sensibly) as it was at A in Table I, and the divergence plus is doubled at A; a state represented in Table II. 3. When A communicates with the ground, all the effects are reversed: the first plate only at A is zero, and the negative state is gradually increasing towards B: the middle point Ff

is now minus, of the same quantity (sensibly) as it was at B in the case of Table I, and the quantity minus is doubled at B; a state represented by Table III. I do not know any theory on invisible causes, which more exactly follows the visible effects.

What remains to be considered is the physical cause of these phenomena, all originating in this circumstance, that when zinc and copper are in mutual contact, zinc possesses more electric fluid, and copper less, than in another situation. In my first paper de

livered to the Royal Society, I explained this effect by analogy with the phenomenon of the different capacities of bodies for the fluid cause of heat; but having here entered into an explanation of the nature of the electric fluid, I shall derive analogies from the subject itself.

Most of the electric phenomena manifested in our experiments depend on the distinction which I have established between the density of the electric fluid, consisting in the proportional quantity of electric matter and its expansive power, depending on the quantity of vector. This distinction is particularly manifested by the changes that happen along an insulated conductor of some length, when an electrified body is placed at some distance from one of its extremities. It is known in general, that the extremity of such a conductor next to an electrified body acquires an electric state contrary to that of the body, while its opposite extremity has the same electric state as that body: for instance, suppose the electrified body to be positive, it is commonly said, that the extremity of the conductor next to this body becomes negative, and the opposite extremity positive; but these are vague expressions, and as they give no real idea of the effects, they have occasioned the variety of systems, all unsatisfactory and therefore changing, hitherto made on these phenomena.

The cause of obscurity on this object is the want of that distinction above mentioned, between the expansive power and the density of the electric fluids belonging to its nature, as expansible. Every fluid of this class, when confined within a certain space, has necessarily the same degree of expansive power in every part of this space, since this is attached to the very idea of expansibility; but it is not the same with respect to density. For instance: a mass of air, confined in a certain space, has certainly, in all its parts the same degree of expansive power, whatever change may happen in its partial density. If then a hot body be placed near one side of this mass of air, its density will diminish in this part, and increase in the more remote or if a piece of ice be brought on one side of this mass of air, its density will increase near the ice, and diminish in the remote parts; but the degree of expansive power of this confined air will always be equal in all its parts at the same time, increasing or dimi-* nishing in the whole.

The case is exactly the same with respect to the electric fluid on an insulated conductor: an insulated positive body being placed

near one part of the latter, this part receives some of the vector which forms around the positive body a kind of atmosphere, as the hot body has around it an atmosphere of igneous fluid: and the former produces an increase of the expansive power of the electric fluid on this part of the conductor, as the hot body produces it in the air next to it. A negative body produces also an effect analogous to a piece of ice for as this absorbs a part of the free fire in the parts of the mass of air next to, so the negative body absorbs a part of the vector from the electric fluid on the nearest part of conductor. Now, all these changes in the degree of expansive power in both fluids are attended with inverse changes in their density.

Nobody will doubt of the above statement who has repeated the experiments described in my works; and I may say, that every one of the great number of persons, in whose presence I have made these experiments, with the set of small electric instruments described in my Traité élémentaire sur le Fluide Electro-galvanique, has been convinced of this important distinction between the density and expan~ sive power of the electric fluid. It is not therefore for want of progress in the science itself, that such a variety of systems subsists concerning the electric phenomena; it is for want of attention in most experimental philosophers, by whom, though writing on electricity, my published experiments are never mentioned, not so much as to criticise them, or their conclusions. Among the variety of these experiments, I have here chosen those concerning the modifications which take place on a long insulated conductor; and I come now to the particulars, which will prove all that I have stated above.

Three electroscopes are used in that series of experiments, one of which, a movable one, consists of silver laminæ, suspended on small axes, in order to prevent their motion out of the line of their divergence. Of the two other electroscopes, consisting of small pith balls suspended by silver wires, one is permanently connected with the extremity of the conductor furthest from the electrified body; the other, held up at the top of a high insulating pillar, is kept, by means of a thin wire, in constant communication with the movable electroscope: the latter, though principally destined to move along the conductor, may be removed, from every point, to a distance, in order to try the nature of the divergence that it had at this point, which it does not lose on removal. Lastly an insulated brass disc is the electrified body. The following are the phenomena observed, in which must be recollected what I undertake to prove, namely, that the divergences of the electroscopes are produced by plus or minus of electric matter only, comparatively with the standard, which is the actual proportionate quantity of electric matter possessed by the ambient air; an object connected with the proof of the fundamental proposition, that there is an absolute distinction between the density, and the expansive power of the electric fluid.

1. At the beginning of the experiment, the movable electroscope

is placed near that extremity of the conductor, to which the disc, after having received a spark from a Leyden vial, is to be approximated. At the approach of this positive body within a small distance, the silver laminae of this electroscope diverge as negative; some of their electric matter having receded to the remote parts of the system (by this word I express the conductor and its associated electroscopes); and thus, though the expansive power of the electric fluid has equally increased upon its whole extent, the two remote electroscopes, and in particular that which is connected with the silver laminæ, diverge positively by a certain quantity.

2. When the silver lamina are made to recede from the positive body along the conductor, their negative divergence gradually diminishes; it ceases at a certain distance, and farther than this begins a positive divergence, which increases to a certain maximum; nevertheless the remote electroscopes, and in particular that which is in immediate connection with the lamina, remain positive to the same degree as at first. The effect, therefore, of withdrawing the lamina has been to remove them out of the atmosphere of vector of the positive body, which, by increasing the expansive power of their electric fluid, had made a part of their electric matter to abandon them and it is the electric matter retired from the anterior part of the system which occasions the positive divergence in the remote electroscopes.

3. While the movable electroscope is thus removed from the atmosphere of the positive body, if any part of the system be touched with a small wire held in the hand, the expansive power of its electric fluid is thus placed in equilibrium with that of the ground, and the divergence ceases in all the electroscopes. Which shows, that, in the parts of the system over which the atmosphere of the positive body does not extend, the equilibrium of proportional quantity of electric matter has been also produced. If then the silver lamina be moved toward the positive body, when they arrive within that atmosphere they begin to diverge as negative; and this divergence continuing to increase, it is greater when this electroscope arrives at the end of the conductor, than it had been in the same place at the beginning of the experiment. However, the remote electroscope in communication with this remains without divergence; because the small quantity of electric matter newly withdrawn from the silver laminæ is insensible upon the whole system.

4. Now will come a proof (among many others which may be found in my works), that the divergence in the electroscopes depends only on the proportional quantity of electric matter, or density of the fluid. If, in this state of the system, the silver lamina be touched with a wire held in the hand, though this contact places them in communication with the ground, their divergence continues the same because their electric fluid, by the increase of vector proceeding from the positive body, being in equilibrium of expansive power with that of the ground, no electric matter can ascend to them