of the balls to be the same in both cases), that is, very nearly as the square of the quantity of redundant fluid in the vial, the difference being not more than what might very easily be owing to the error of the experiment. So that the experiment agrees very well with the theory.

392] It was found that if the communication was made between the two vials by a piece of metal, the electricity was diminished so suddenly as to set the straws a vibrating, and it was some time before they stopt, for which reason the communication was made by a piece of moist wood, which, though it communicates the electricity of one vial to the other very quickly, did not do it so instantaneously as to make the straws vibrate much.

393] The electricity of the vial was found to waste very slowly, so that it could not be sensibly diminished during the small time spent in communicating the electricity from one vial to the other and reading off the divisions, so that no sensible error could proceed from that cause.

394] I tried the experiment before in the same manner, and with the same electrometers, except that the straws were not gilt, but only moistened with salt. It then seemed as if the force with which the balls of the electrometer were repelled with a given quantity of redundant fluid in the vial was to that with which they were repelled with only half that quantity in the vial as 4 to g.

As I suspected that this small difference from the theory was owing to the straws not conducting sufficiently readily, I gilt the straws, when, as was before shewn, the experiment agreed very well with theory.

It must be observed that if the straws do not conduct sufficiently readily, the balls of the electrometer will not be so strongly electrified and will not separate so much as they ought to do, and in all probability the difference will be greater in the stronger degree of electricity, in which the electricity wastes much faster, than it is in the weaker, and will therefore diminish the degree of separation more in the stronger degree of electricity than in the weaker, and will therefore make the force with which the balls repel with the stronger degree of electricity appear to be less in proportion to that with which they repel with the weaker degree than it ought to be.

AN ACCOUNT OF SOME ATTEMPTS TO IMITATE THE EFFECTS OF THE TORPEDO BY ELECTRICITY. BY THE HON. HENRY CAVENDISH, F.R.S.*

395] Although the proofs brought by Mr Walsh†, that the phenomena of the torpedo are produced by electricity, are such as leave little room for doubt; yet it must be confessed, that there are some circumstances, which at first sight seem scarcely to be reconciled with this supposition. I propose, therefore, to examine whether these circumstances are really incompatible with such an opinion; and to give an account of some attempts to imitate the effects of this animal by electricity.

396] It appears from Mr Walsh's experiments, that the torpedo is not constantly electrical, but hath a power of throwing at pleasure a great quantity of electric fluid from one surface of those parts which he calls the electrical organs to the other; that is, from the upper surface to the lower, or from the lower to the upper, the experiments do not determine which; by which means a shock is produced in the body of a person who makes any part of the circuit which the fluid takes in its motion to restore the equilibrium.

397] One of the principal difficulties attending the supposition, that these phenomena are produced by electricity, is, that a shock may be perceived when the fish is held under water; and

* From the Philosophical Transactions for 1776, Vol. LXVI. Part 1. pp. 196–225. Read Jan. 18, 1775.

↑ [Philosophical Transactions, 1773, pp. 461–477. Of the Electric Property of the Torpedo. In a letter from John Walsh, Esq., F. R. S., to Benjamin Franklin, Esq., LL.D., F.R.S., &c. Read July 1, 1773.]

398] RESISTANCE OF IRON, WATER, ETC. TO ELECTRICITY.

195

in other circumstances, where the electric fluid hath a much readier passage than through the person's body. To explain this, it must be considered, that when a jar is electrified, and any number of different circuits are made between its positive and negative side, some electricity will necessarily pass along each; but a greater quantity will pass through those in which it meets with less resistance, than those in which it meets with more. For instance, let a person take some yards of very fine wire, holding one end in each hand, and let him discharge the jar by touching the outside with one end of the wire, and the inside with the other; he will feel a shock, provided the jar is charged high enough; but less than if he had discharged it without holding the wire in his hands; which shews, that part of the electricity passes through his body, and part through the wire. Some electricians indeed seem to have supposed that the electric fluid passes only along the shortest and readiest circuit; but besides that such a supposition would be quite contrary to what is observed in all other fluids, it does not agree with experience. What seems to have led to this mistake is, that in discharging a jar by a wire held in both hands, as in the above-mentioned experiment, the person will feel no shock, unless either the wire is very long and slender, or the jar is very large and highly charged. The reason of which is, that metals conduct surprisingly better than the human body, or any other substance I am acquainted with; and consequently, unless the wire is very long and slender, the quantity of electricity which will pass through the person's body will bear so small a proportion to the whole, as not to give any sensible shock, unless the jar is very large and highly charged.

398] It appears from some experiments*, of which I propose shortly to lay an account before this Society, that iron wire conducts about 400 million times better than rain or distilled water; that is, the electricity meets with no more resistance in passing through a piece of iron wire 400,000,000 inches long, than through a column of water of the same diameter only one inch long. Sea water, or a solution of one part of salt in 30 of water, conducts 100 times, and a saturated solution of sea salt about 720 times better than rain water.

[Arts. 576, 577, 684, 687.]

399] To apply what hath been here said to the torpedo; suppose the fish by any means to convey in an instant a quantity of electricity through its electric organs, from the lower surface to the upper, so as to make the upper surface contain more than its natural quantity, and the lower less; this fluid will immediately flow back in all directions, part over the moist surface, and part through the substance of its body, supposing it to conduct electricity, as in all probability it does, till the equilibrium is restored and if any person hath at the time one hand on the lower surface of the electric organs, and the other on the upper, part of the fluid will pass through his body. Moreover, if he hath one hand on one surface of an electric organ, and another on any other part of its body, for instance the tail, still some part of the fluid will pass through him, though much less than in the former case; for as part of the fluid, in its way from the upper surface of the organ to the lower, will go through the tail, some of that part will pass through the person's body. Some fluid also will pass through him, even though he does not touch either electric organ, but hath his hands on any two parts of the fishes body whatever, provided one of those parts is nearer to the upper surface of the electric organs than the other.

400] On the same principle, if the torpedo is immersed in water, the fluid will pass through the water in all directions, and that even to great distances from its body, as is represented in Fig. 1, where the full lines represent the section of its body, and

Fig. 1.

the dotted lines the direction of the electric fluid; but it must be observed, that the nearer any part of the water is to the

fishes body, the greater quantity of fluid will pass through it. Moreover, if any person touches the fish in this situation, either with one hand on the upper surface of an electric organ, and the other on the lower, or in any other of those manners in which I supposed it to be touched when out of the water, some fluid will pass through his body; but evidently less than when the animal is held in the air, as a great proportion of the fluid will pass through the water: and even some fluid will pass through him, though he does not touch the fish at all; but only holds his hands in the water, provided one hand is nearer to the upper surface of the electric organs than the other.

401] The second difficulty is, that no one hath ever perceived the shock to be accompanied with any spark or light, or with the least degree of attraction or repulsion. With regard to this, it must be observed, that when a person receives a shock from the torpedo, he must have formed the circuit between its upper and lower surface before it begins to throw the electricity from one side to the other; for otherwise the fluid would be discharged over the surface of the fishes body before the circuit was completed, and consequently the person would receive no shock. The only way, therefore, by which any light or spark could be perceived, must be by making some interruption in the circuit. Now Mr Walsh found, that the shock would never pass through the least sensible space of air, or even through a small brass chain. This circumstance, therefore, does not seem inconsistent with the supposition that the phenomena of the torpedo are owing to electricity; for a large battery will give a considerable shock, though so weakly charged that the electricity will hardly pass through any sensible space of air; and the larger the battery is, the less will this space be. The principle on which this depends will appear from the following experiments.

402] I took several jars of different sizes, and connected them to the same prime conductor, and electrified them in a given degree, as shewn by a very exact electrometer; and then found how near the knobs of an instrument in the nature of Mr Lane's electrometer must be approached, before the jars would discharge themselves. I then electrified the same jars again in the same degree as before, and separated all of them from the conductor except one. It was found, that the distance to which the knobs