phenomenon might be produced, in which the cause of the motion might be visible; and it having succeeded, I have described this experiment from p. 116 to 119 of the same volume, of which it will be sufficient to give here an extract.

15. An accidental observation led me to that experiment, which those who shave themselves and employ soap powder will easily repeat. If the basin used happens to be covered with dust, this swims over the water; but it is very easily swept, by scraping a small flake of soap, taking it on the point of a knife, and endeavouring to let it fall on the middle of the surface of the water with its convexity undermost. As soon as that flake of soap swims on the water, the dust recedes from it, and ascends to the edge of the water on the side of the basin, where it forms a thin fringe.

16. Having taken notice of that motion of the surface of water made visible by the dust, I attempted to imitate the motion of two electrified balls by two discs of soap half an inch in diameter, and about 1-10th of an inch thick, in the centre of which I fixed with gum a very thin thread four inches long, and I suspended these discs to a slip of wood at such a distance that they slightly touched each other. When that little apparatus was prepared, and the two discs made to rest at the same time on the surface of the water, they began to recede from each other, and the motion of the dust on the surface of the water visibly demonstrated that the discs of soap moved both towards the water which had not yet dissolved soap.

17. Lastly, I have demonstrated by a direct experiment, which I shall relate hereafter, that a perfect vacuum does not transmit the electric fluid; which fact will be a peremptory proof of the essential interference of air in the motion of a pair of electrified balls.

18. Another essential part of Volta's system, with which Mr. Donovan is not acquainted, and which when fully understood, removes all the difficulties he has found, is that which Volta has called electric influences. It consists in this effect: when a body positive is brought near one of the extremities of an insulated conductor, it gives more tension to the electric fluid at that end of the conductor, and makes it recede to the further end. This principle removes all the difficulties that Mr. Donovan has found, and in particular that which seems to arise from his experiment related in p. 338, of a pith ball suspended to a glass rod, and an excited glass tube brought under it. The excited tube occasions a slow retreat of electric matter along the thread. While that matter remains together in the ball, the thread, and at the point of suspension, the ball is repelled by the excited tube; but at last the suspended ball loses some electric matter, and being then negative, it is attracted by the excited tube.

19. The author attacks with reason the cause assigned by Franklin to the luminous brush which appears at the extremity of a pointed body fixed to the prime conductor of an electric machine in motion. But Volta has also explained this pheno

menon in a satisfactory manner, as I have explained in pp. 60 and 61. Suppose a conductor positively electrified, and to which is presented another conductor in communication with the ground, but the surface of which is of a certain extent; the particles of air in motion, returning from the electrified conductor, and thus possessing more electric fluid, might transmit some to the other body; but their influence on it producing a diminution of tension in the fluid proceeding from the conductor, at the same time that they increase the tension in the fluid of the body, there is thus a very small disposition of the electric fluid to abandon the particles of air. But if it be a point, the electric fluid belonging to it is in too small a quantity to increase the tension of that belonging to the body: thus, every particle of air which comes in contact with that body discharges its electric fluid, returns instantly, and, as it takes the shorter way, thus is produced a current of air. The same alternate motions takes place in the particles of air, whether the point is negative or positive.

20. The difficulty concerning the impermeability of glass to the electric fluid, which the author opposes to the phenomena of the Leyden vial, as accounted for by Franklin, has also been removed by Volta, by supposing the electric fluid composed of two ingredients, to one of which only glass is impermeable, but the other passes readily through it. If the jar is coated on both sides with tinfoil up to a certain height, it will not receive a charge, unless the outside coating is in communication with the ground.

21. On the explanation of this common phenomenon rests the whole theory of positive and negative electricities, which Volta, having placed on its true foundation, has thereby clearly explained the phenomena of the Leyden vial; and it is his theory which has led me to my systems on the nature of the electric fluid. This fluid, considered in its common phenomena, is composed of two ingredients; one, which does not possess expansibility by itself; the other, which, uuited with it, gives expansibility to the compound. Of this union I have given an example in the aqueous vapour, which is composed of particles of water, a substance not possessing expansibility, and of particles of fire, which gives expansibility to the compound. That analogy, with all its particulars, I have explained in p. 77, sec. I, with this title: "Analogies and Differences of the Electric Fluid, and the Aqueous Vapour." In consequence of this analogy, I have given the name of electric matter to that ingredient of the electric fluid to which glass is not permeable, and deferent fluid to that ingredient which transports the electric matter, and which alone passes through the glass.

22. In this system Mr. Donovan may find the solution of his difficulties concerning the charge and discharge of the Leyden vial. Franklin's system is true so far as he supposes that the vial possesses (nearly) the same quantity of electric fluid, both when it is in its natural state, and when it is charged; but the electric fluid is not distributed in the same manner in both cases. When an electric

machine tends to produce an accumulation of electric fluid on one of the surfaces of the vial, it cannot accumulate except the outside of the vial can part with an equal quantity, its coating being in communication with the ground. But it is not the identical electric fluid which pervades the glass, the electric matter thrown in remains fixed on the inner surface; the deferent fluid alone pervades the glass, and, uniting with the electric matter on the outward surface, forms there an equal quantity of electric fluid, which is transmitted to the ground if it finds a conductor. This is a complete analogy with one of the phenomena of the aqueous vapour mentioned in the above quoted section. The maximum of charge is attained when the quantity of electric matter accumulated on the internal surface is in equilibrium with that transmitted by the electric machine.

23. So far this system might appear only an hypothesis to Mr. Donovan, as he does not know the work to which I refer, where I have given a peremptory demonstration of it, derived from the electric motions, proving by a direct experiment that these motions refer only to the electric matter, without interference of the deferent fluid, except to transport the electric matter.

24. In tome ii, p. 8, I have related an experiment which very much struck M. Cavallo, in presence of whom I made it at Windsor. The apparatus for that experiment is described, with a plate, in my work; it consists of a conductor supported in an horizontal position on an insulating pillar. At one end of this conductor are suspended, by brass wire, two cork-balls, half an inch diameter, going down so as to hang opposite to the centre of two insulated brass discs two inches diameter, brought to about 3 inches of each other, and the balls hung at equal distance from them. I give a spark to each of the discs with a Leyden vial: the deferent fluid of the discs, thus rendered positive, and giving more expansive power to the fluid of the balls, a part of it is carried towards the other extremity of the insulated conductor; and the balls losing thus a part of their electric matter, they diverge as being in a negative state, and are carried on both sides towards the positive discs, which they would strike, were they not prevented by a kind of bridle, placed at the end of the conductor, showing by small motions as it were an avidity to move to

wards them.

25. In that state of the apparatus in which the influence of the discs, being positive, gives more expansive power to the electric fluid of the balls, they are certainly negative comparatively to the ground and the ambient air; but the deferent fluid of the positive discs giving more expansive power to their electric matter, a conductor communicating with the ground may be brought into immediate contact with them without effect: their negative divergence still continues. The same phenomenon is produced by inversely electrifying the horizontal insulated conductor and the discs: when the balls, then positive, are made to communicate with the ground by a brass conductor, they continue to diverge, being under the influence of the negative atmosphere of the discs.

26. These phenomena much interested M. Cavallo, who was a very liberal-minded man. He persisted in his opinions as long as he found reasons to defend them, but was ready to abandon them whenever his arguments appeared to be contradicted by facts, to which he was very attentive. Having made some stay at Windsor, he came often to me and proposed some alterations in my apparatus, which he thought could change the phenomena. I foretold him what would happen by the changes, some of which would change the effect on account of causes which I pointed out, and others would produce the same effect. At last he saw and acknowledged that those phenomena could absolutely not be explained but by admitting the composition of the electric fluid of two ingredients, one of which only, which I had called electric matter, produced electric motions, and the other, which gave expansibility to that matter in proportion to its relative quantity, which I named electric deferent. Thus Volta's theory, applied to all the electric phenomena which were opposed to Franklin's theory, in his own expressions, have cleared the system of positive and negative electricities from the objections of Mr. Donovan, arising from the erroneous principle which he admitted in it, of a natural quantity of electric matter belonging to all terrestrial bodies.

27. I come to another very important object in the science of electricity, that of the effect of friction, which in vol. xxxiii, of Mr. Nicholson's "Philosophical Journal," I have explained in a manner different from that of Mr. Donovan in p. 337. His hypothesis is this: "That during attrition between bodies, one of which must be an electric, the pores of the latter being open will receive the plus quantity, and will give it out again when the pores close."-In this respect I have proved, that there is no other distinction between bodies than that of conductors and non-conductors; the latter of which only can be excited by friction, because the electric fluid, thus set in motion on the surface, moves slowly to make its escape by a conductor offered at a distance. While, if both bodies which exercise friction on each other are good conductors, the equilibrium is incessantly restored; but if one has more disposition than the other to attract the electric fluid thus agitated, with the faculty of transmitting it to its remote parts, when the bodies are separated before the equilibrium is restored between them, one is found positive and the other negative.

28. These effects are distinctly demonstrated by a small electric machine, the figure of which is at the head of my paper in Mr. Nicholson's "Philosophical Journal" for January, 1811, under the title of "Experiments concerning the Electric Machine, showing the Effects of Friction between Bodies."

29. M. Cavallo has given a table containing the results of his experiments of this kind, wherein we find that certain bodies become either positive or negative by friction, according to those by which they are rubbed; but the manner in which his experiments were made did not indicate the effect produced on the rubber itself,

because it was not insulated. I thought it therefore very important to the doctrine of electricity, to have both effects indicated by electrometers. This I obtained by the apparatus described in that paper; in which the bodies rubbed are spindles turned by a winch, and various rubbers made to press on the spindles by proper springs. A small insulated prime conductor is connected by one of its extremities to the spindles, and by the other with a gold-leaf electrometer. The rubbers are insulated, and each of them when applied is made to communicate with a similar gold-leaf electrometer. 30. The constant result of these experiments was, that the quantity of plus on one side, was equal to the quantity of minus on the other. Now the friction being reciprocal, the supposition of the opening of the pores cannot explain that phenomenon, since they ought to be opened equally in the rubbing as in the rubbed bodies.

31. The effect of friction, as I have said above, is to set in motion on the surface of the bodies, and that, if the body which recedes from the point of friction finds in its way a pointed conductor, it transmits to the latter a part of that fluid; which is the effect of the common electric machine: but the same phenomenon extends further; for that effect takes place between bodies of the same kind, if they be non-conductors. This is proved by Exper. 3, p. 9, of the same number of the " Philosophical Journal."

32. A flat piece of the same glass as the spindle, being held at the end of a brass spring, and used as a rubber on the spindle, holding the brass part in my hand near the glass, in order to restore from the ground the electric fluid carried away to the prime conductor by the spindle, the gold-leaf electrometer diverges as positive, though the rubber is of the same substance as the spindle. This is a peremptory demonstration that the effect of friction is not to open the pores for receiving more electric matter which is discharged when the friction ceases, as Mr. Donovan conceives it: since both bodies, which in this case exercise friction on each other, are of the same glass.

33. Experiment 4, p. 10, shows that sealing-wax, used as a rubber over the glass spindle becomes strongly negative, and renders the glass strongly positive. And thus it is directly proved, that the cause of sealing-wax being rendered negative by friction with the glass, is that the latter takes some electric fluid from sealing-wax, which effect could not have been ascertained without the two electrometers.

34. Now, in Exp. 1, a brass rubber applied to the glass cylinder is seen to become negative, and the glass cylinder made positive; but in Exp. 6, having covered a glass cylinder with a thick coating of sealing-wax, producing in fact a sealing-wax cylinder, the same brass rubber was rendered positive. These experiments compared with each other, further demonstrate the real electric effect of friction. But the next experiment will place it beyond all doubt.

35. Exp. 8, is made with what is called India beads, the size and colour of a cherry: they are made, as I have been informed, of an