Faraday to Macedonio Melloni   31 July 1854¹

Royal Institutions, London, 31 july 1854.

My dear Melloni,

I have three letters of yours unanswered the two last of the dates of the 1st and the 12th instant². I have been unable to answer them before, because the beginning of the month I was in the country, and I returned from it only to place myself under the surgeon’s hands: and since then have been unable to write or work: I am now getting better but am still in his care.

When I received your last letter, I had no need to repeat your experiments; for they were all perfectly familiar to me, as far as I could judge by your description, and are necessary consequences of the theory of static induction, which I published in the Philosophical Transactions 17 years ago. Still I should have sent your letter to me to the Royal Society as you requested, if it had been open: but its sittings were closed on the 15 june, a month before your letter was written, so that I was unable to fulfil your desires in that respect.

Your difficulties present no difficulties to me, neither do I remember clearly the error or illusion into which we have all fallen and which you say is continued³ in all the books. The fact is I have interpreted induction according to my own views in the last 18 years and have not carefully analyzed the words of recent treatises, but I do not remember that the statement of Pouillet⁴ or De La Rive contradict my notions. As my views though given at length in the Philosoph. Trans. in the years 1837 and 1838 in the series XI, XII, XIII and XIV⁵ of my Experimental Researches, have not been published either in French or Italian, they probably have never come under your notice: I endeavoured to send English copies of them at the time, through the Royal Society, but very likely they never reached you: so I am about to give you a brief summary of them, referring to the members⁶ of the paragraphs in the Exp. Researches, that you may, if you desire it, look at the original matter.

But first let me remind you, that I do not as yet know the nature of your Electrometer and therefore may have misunderstood your statement altogether; though as regards the simple results with (fig. 1), shaded or not.⁷ I do not think that likely, for they are precisely the same as my own old ones.

I profess to know nothing of the existence of either one or two electric fluids, or of the nature of the electric power, exerted either in a P, or an N direction (1298 note⁸ 1667⁹), but I do not think that our ignorance of the essential nature of the electric force offers any difficulties in the consideration of the nature of induction, conduction, etc., provided we do not travel beyond facts and the laws which govern them.

In induction, it is only the surface of the conductor which is finally affected, and not its internal parts: and that whether the conductor be insulated or connected with the earth (1220, 1221, 1295, 1301¹⁰). If it be uninsulated, only that part of its surface is finally affected, on which lines of induction force, proceeding from the ex[c]ited or inductric body A, abut and terminate: if it be insulated, then the parts of the surface, from which an exact equivalents of new lines of force originate and proceed outwards from the compound system (fig. 2) towards neighbouring conductors, are affected also: the first surface is B, the second surface C and the neighbouring parts, and between them there is a part or zone of various from¹¹ according to circumstances, in the neutral or normal conductor.

The induction is limited by the induction surfaces (1231, 1297¹², 1361, 1372¹³, 1483¹⁴, etc.). The one which is primarily charged I have distinguished as the inductric, the other as the inductors.

The lines of induction force, used merely as representations of the disposition of the electric force, are described (1231, 1304¹⁵, 1441, 1450¹⁶): they commence at one and terminate at the other, of the inducting surfaces. If the inductric body A be an ex[c]ited insulator, as a rod of shell lac, they then commence at the ex[c]ited particles.

Induction is not sustained through the body of the thinnest conductor. Theoretically, it occurs at the first instant: but conductors discharges its state within, and it is the surfaces only of the conductor which remains finally affected. An uninsulated gold leaf in a frame or ring, may have its opposite sides raised by induction to the highest opposite or like states, without the slightest interference of one side with the other.

The lines of induction force across the dielectric or insulating medium may be curved (1215, 1219, 1221, 1224, 1230¹⁷, 1374, 1449¹⁸, 1614¹⁹. See also figures 7, 8, 9, 10, 11, 12). In experiments in open spaces they are almost universally curved.

Induction lines have a lateral relation (1224, 1295²⁰, 1449²¹) quite analogous to that of the lines of current force, when the piles of a voltaic battery are dipped apart from each other into a large mass of water or dilute solutions; and which have been so well illustrated by Nobili²² in his metallochromic results²³.

When a charged sphere A (fig. 3) is in the center of a much larger sphere of conducting matter, the lines of force proceed as radii from A to every part of the outer sphere: the sum of force on the surface of A and the sum of contrary force on the inner surface of the surrounding sphere, are exactly equal to each other.

When an insulated conductor B, C, is introduced, then a certain amount of the lines of force from A terminate on B, produces an opposite state there, but an equal amount of force or of lines, originate about C and terminate on D; or in other words the lines of force which would have been passed across the space BC, if the conductor BC were away, have, through the conduction of the particles of BC, been replaced by the equivalents of contrary forces at the respective surfaces of B and C; at the same time, as the resistance or tension set up in induction (1368, 1370²⁴, etc.), is removed as regards the space BC, by the conduction; so, more electricity must induce from A towards B than in other directions, as towards E; and more inductions action is induced on D than elsewhere as at E. But though D differs thus from the other parts of the inner surface of the surrounding chamber, and equal to the amount of force existed²⁵ in all directions by A. I need not refer particularly to irregular and mixed cases, as the walls of a room (1434²⁶) or more complicated results (1337²⁷, 1566²⁸, 1679²⁹, etc.): the principles are the same and the amount of action always definite.

If another conductor, either insulated or connected with the earth, be approached to the former conductor, then lines of induction force are transferred to it from the former induction bodies (1225³⁰, 1449³¹). Thus, if F be such a body, then electric force or lines, which before proceeded towards B or towards H, are transferred to F; and if F be insulated, with the same development of contrary force as before. If F be uninsulated and large, the part behind it, at H, may even receive no charge, but be in the natural state: and the sum of power upon the inner surface of the envellope H, D, E will be less than the sum of contrary power on A, by so much as is disposed of on the surface of an uninsulated F.

I know no distinction between free and dissimulated electricity (1684³²). Both are cases of induction and change by induction. If the electricity between A and E is not dissimulated, neither is that between A and B, or between A and F. It might as well be said that the lightning which falls upon and kills a man is free to him, but is dissimulated to one who is a few miles distant: the difference is merely one of circumstance not of principle. Whether the inductric body induces upon me or upon another person, makes no difference in the action.

If the sum of power be 20, as much as 15 may be towards me and only 5 to him: but the land of action is alike in both cases and the sum of the power remains the same.

All charge is induction: all induction is charge (1177, 1178, 130³³, 1682³⁴). A short but imperfect summary is contained in the paragraphs from 1667 to 1684³⁵. You will find also in the Philosophical Magazine, 1843, vol.XXII, p. 200³⁶, in vol.II of the 8th Edition [sic] of my Experimental Researches, p. 279³⁷, some demonstrative experiments on static induction. I have been accustomed for years past to shew very many forms of experiments founded on these principles; I will describe one which seems to embrace many of the results you describe to me. A conductor A is sustained in a charged state by connexion with a (fig. 4) Leyden- jar: an insulated brass plate is brought near it, and then the surfaces examined by a Coulombs proof plane or carrier³⁸ (which I hold to be unexceptionable when properly employed) and the surface of b seems to be contrary to A (fig. 5), whilst the surface of c is like that of a. The two plates close together and connected (fig. 6), replace the one plate a the surfaces b and c found like the former surfaces. Then the plates are opened about an inch being still connected by a brass rod on (fig. 7) which they hang, b and c are found as before; but within at d there is no electricity of either kind on either surface.

All the time there is electricity like that of A to be found at the edges of the plates, provided irregular lateral inductions be prevented. Then the plates are more opened out, still the electricity at (fig. 8) b and c is the same, but that at the edges of the near plate b, begins to lose its A character and either be neutral or assume that of b. Being more opened out, the electricity of b creeps round the edge of the plate (fig. 9), approaching nearer to the middle at the plate b is smaller, or the sign of A larger, or its power more intense; and the electricity of c also becomes more extended over the surface a round the edges of that plate, according to its size and position in relation to surrounding bodies. If an uninsulated conductor be brought towards any part of the compound system of plates a rod, as at 1, 2, 3, 4, it immediately renders these places electric like A, either causing or exalting their state, whilst it acquires through the induction, a state the contrary of A or like that of b. If ex[c]ited shell lac be brought near 1, 2, 3, 4 it immediately exerts its induction action upon the plates a, in addition to the action of A. All these and a thousand others are the simplest possible results of the theory.

You speak of screening the “pendules accouple” from the action of A whilst you examine them by an ex[c]ited electric (rod?). But in that case the pendules give no indication of the state of the part to which they are attached. They do not receive their final state by conduction from the part they are fixed to, but only by induction as a part of the conducting mass BC: if they are expand to the inductive force of A they will acquire the opposite state; if they are perfectly screened they will be neutral, or if they are so near so expand to surrounding conductors as to be in a position to carry on the forces, they will assume the c state, which is the same as the state of A.

I have often shewn my audience this condition of the pendulous balls³⁹, by placing the cylinder and its balls well insulated, in different positions as respects the inductric ball A: thus (fig. 10), when in this position, B and its balls will be in the reverse state to A, and c and its ball in the same state as A: but then (fig. 11) held it by an insulating handle thus and through B still acquires a state contrary to A, yet the balls attached to it by conducting matters shew the other state, or that of A and c.

Again, if the ball A and cylinder BC (fig. 12) retain their position, it is very easy to have the balls hanging to B, in the state of B, or by approaching an uninsulated ball or screen, either at 1 or 2 or 3, to make them assume the contrary state on that of C, or by adjustment of distance to be perfectly indifferent.

Trusting to the truth of the principle I have described I am accustomed to use wire gauze instead of a continuous metal plate, for screens and other apparatus, which I want my audience to see through: and I have plates like those described above constructed of such wire gauze. Through the openings are three or four times the diameter of the wire, yet no electricity of induction can pass through them, and a plate of such wire gauze is as impervious to conduction as a plate of metal.

I make a cylinder of such wire gauze part of the conductor of the machine, but the most delicate tests inside gave no indication of Electricity there. On the lecture table, I threw a net of common twine over my gold leaf electrometers connecting it well with the ground, and they are rendered perfectly safe from charge of the machine, which by induction would destroy them at once if not so guarded.

All of which, illustrates the powerful effect of screens in static inductive action. I have published no account of these things because they are simple consequences of my theory: but De La Rive who happened to see them once in the lecture room here, gave a brief account of them in the Geneva Journal⁴⁰.

And now, my dear friend, I will relieve you from a tiresome lecture. You speak of the over universal in books, and therefore I may say, that Daniell in his Introduction to Chemical Philosophy⁴¹ adopts my views and therefore as I suppose is not in error.

That Harris I believe agrees with me; that as far as I am aware De La Rive does not put forth any error of the kind you refer to⁴² in his recent work on Electricity⁴³ - or Pouillet - except that he speaks of dissimulated electricity⁴⁴ but by that means electricity of induction. But I will not tease you more.

Ever My dear Friend | Your affectionate servant | M. Faraday

A M | Macedoine Melloni