this same rubbed wax would attract them, if they were in a positive state.

9. But you may obtain the most convincing proof, that they are in a negative state, in the following manner. Electrify the other pair of rods positively, as was directed above; and, when the balls of each pair have receded to the same distance from each other, which shows that there is as much positive electricity in one pair as there is negative in the other, bring the ends of the rods in contact, and you will see the balls on each pair of rods come together instantly, one pair of rods restoring to the other the quantity which it had parted with, so that both are restored to their natural state. When both pairs are electrified plus, or both minus, separately and equally, in vain would you bring them in contact; no effect whatever would be produced.

To perform these experiments well, care must be taken to keep the glasses always dry; and the best way to do this is to warm them from time to time, if the air is damp.

TO M. DUBOURG.*

On the Analogy between Magnetism and Electricity.

London, 10 March, 1773.

SIR, As to the magnetism, which seems produced by electricity, my real opinion is, that these two powers of nature have no affinity with each other, and that the

* This letter and the three following it are translated from M. Dubourg's French edition, (Tom. I. pp. 277, 312, 332.) — Editor.

apparent production of magnetism is purely accidental. The matter may be explained thus.

1st. The earth is a great magnet.

2dly. There is a subtile fluid, called the magnetic fluid, which exists in all ferruginous bodies, equally attracted by all their parts, and equally diffused through their whole substance; at least where the equilibrium is not disturbed by a power superior to the attraction of the iron.

3dly. This natural quantity of the magnetic fluid, which is contained in a given piece of iron, may be put in motion so as to be more rarefied in one part and more condensed in another; but it cannot be withdrawn by any force that we are yet made acquainted with, so as to leave the whole in a negative state, at least relatively to its natural quantity; neither can it be introduced so as to put the iron into a positive state, or render it plus. In this respect, therefore, magnetism differs from electricity.

4thly. A piece of soft iron allows the magnetic fluid. which it contains to be put in motion by a moderate force; so that, being placed in a line with the magnetic pole of the earth, it immediately acquires the properties of a magnet, its magnetic fluid being drawn or forced from one extremity to the other; and this effect continues as long as it remains in the same position, one of its extremities becoming positively magnetized, and the other negatively. This temporary magnetism ceases as soon as the iron is turned east and west, the fluid immediately diffusing itself equally through the whole iron, as in its natural state.

5thly. The magnetic fluid in hard iron, or steel, is put in motion with more difficulty, requiring a force greater than the earth to excite it; and, when once it has been forced from one extremity of the steel to the

other, it is not easy for it to return; and thus a bar of steel is converted into a permanent magnet.

6thly. A great heat, by expanding the substance of this steel, and increasing the distance between its particles, affords a passage to the magnetic fluid, which is thus again restored to its proper equilibrium; the bar appearing no longer to possess magnetic virtue.

7thly. A bar of steel, which is not magnetic, being placed in the same position, relatively to the pole of the earth, which the magnetic needle assumes, and in this position being heated and suddenly cooled, becomes a permanent magnet. The reason is, that while the bar was hot, the magnetic fluid which it naturally contained was easily forced from one extremity to the other by the magnetic virtue of the earth; and that the hardness and condensation, produced by the sudden cooling of the bar, retained it in this state without permitting it to resume its original situation.

8thly. The violent vibrations of the particles of a steel bar, when forcibly struck in the same position, separate the particles in such a manner during their vibration, that they permit a portion of the magnetic fluid to pass, influenced by the natural magnetism of the earth; and it is afterwards so forcibly retained by the re-approach of the particles, when the vibration ceases, that the bar becomes a permanent magnet.

9thly. An electric shock passing through a needle in a like position, and dilating it for an instant, renders it, for the same reason, a permanent magnet; that is, not by imparting magnetism to it, but by allowing its proper magnetic fluid to put itself in motion.

10thly. Thus there is not in reality more magnetism in a given piece of steel after it is become magnetic, than existed in it before. The natural quantity is only displaced or repelled. Hence it follows, that a strong

apparatus of magnets may charge millions of bars of steel, without communicating to them any part of its proper magnetism; only putting in motion the magnetism which already existed in these bars.

I am chiefly indebted to that excellent philosopher of Petersburg, Mr. Epinus, for this hypothesis, which appears to me equally ingenious and solid. I say chiefly, because, as it is many years since I read his book, which I have left in America, it may happen, that I may have added to or altered it in some respect; and, if I have misrepresented any thing, the error ought to be charged to my account.

If this hypothesis appears admissible, it will serve as an answer to the greater part of your questions. I have only one remark to add, which is, that, however great the force is of magnetism employed, you can only convert a given portion of steel into a magnet of a force proportioned to its capacity of retaining its magnetic fluid in the new position in which it is placed, without letting it return. Now this power is different in different kinds of steel, but limited in all kinds whatever. B. FRANKLIN.

FROM M. DUBOURG TO B. FRANKLIN.

On the Choice of Glass for the Leyden Experiment.

SIR,

Paris, 25 March, 1773.

If I have rightly understood your principles, the glass to be used in the Leyden experiment ought to combine these two qualities; first, it should be impermeable to the electric fluid; secondly, it should not be impermeable to the action of this fluid; or, to express the same

thing in other words, the electric fluid must not be able to pass from one surface to the other, but its afflux on one of the surfaces of the glass must have the power to excite an afflux on the opposite surface.

Glass generally unites these two qualities, but not every kind of glass. There is even glass that the electric fluid passes through, almost as readily as it enters metals. This is a property natural to some kinds of glass, and accidental to others. It would seem astonishing, that no philosopher had yet thought of seeking out the causes of all these differences, if natural philosophy alone were equal to the task; but there is need of the aid of chemistry, which certainly may throw some light on so interesting a subject.

I would not propose to the chemists to analyze the different kinds of glass, permeable or impermeable to electricity; but to endeavour to imitate them, which would be much easier for them to do.

Pure vitrifiable earth is without doubt the only ingredient in rock crystal, which may be considered as a true natural glass; but art has not yet succeeded in obtaining for us a glass so pure, and there is even very little reason to hope that such perfection can ever be attained.

There is no earth known so vitrifiable as not to require some auxiliary solvent to facilitate its vitrification. Now solvents are distinguished into three principal kinds, which are, saline solvents, metallic solvents, and earthy solvents; for there are different kinds of earths, which, although each singly is refractory, yet serve as mutual solvents, as there are also many kinds of salts, and many kinds of metals, which may be used as solvents for the vitrifiable earths, and which may be combined in different proportions with the same earths. We ought not to be more surprised to find glass more or