we have the repulsive power in weight; and by the ball A, the discharge is caused, when the charge has acquired the strength proposed.

Experiments made with a View to determine in what Degree the Charging Capacity of Coated Jars is Increased by Breathing into them before the Charging.

Experiment 1.-Prepare the electrometer in the manner shewn in the plate, with the jar м annexed, which contains about 168 square inches of coating ;* put into в the pin, marked 15; take two inches of watch-pendulum wire, fix to each end a pair of spring tongs, as is represented at G m, hook one end to m, and the other to the wire N, communicating with the outside of the jar; let the uncoated part of the jar be made very clean and dry; and let the prime conductor of an electrical machine, or a wire proceeding from it, touch the wire L; then if machine be put in motion, the jar and electrometer will charge, as will be seen by the rising of the index of k, and when charged high enough, в will be repelled by b, and a will descend and discharge the jar through the wire which was confined in the tongs, and the wire will be fused and run into balls.

Exp. 2.-Put into the tongs eight inches of the same sort of wire as before, hang one pair of tongs to the hook m, and apply the other to the wire which forms the outside communication: take out the pin in B, and put in its stead one marked 30; all the other part of the apparatus remaining as before, and the uncoated part of the jar being previously cleaned and dried; the machine being then put in motion, the jar and the electrometer will charge, as is shown by the rising of the index as before; but as soon as the jar has received a greater quantity of electric fluid than before, a spontaneous explosion will happen without affecting the balls B b, because the discharge will have passed along the uncoated part of the jar from the inside coating to the outside; whence it follows, that while the jar remains in that clean state, it is incapable of receiving a charge high enough to affect the balls, or even a higher charge than it had received in the first experiment. Let the uncoated part of the jar be therefore rendered in a slight degree damp; which is easily done by breathing into the inside through a glass tube; put the machine in motion, and no spontaneous explosion will happen, but the balls B b will repel, as in the first experiment, and the discharge will happen from A to a, and pass through the wire placed in the circuit; and though it was eight inches, it will be fused in the same degree as two inches in the last experiment, namely, the wire seen red hot the whole length, and then fall into balls.

Very different degrees of fusion are caused by electric discharges, which may cause great mistakes, if not well attended to. It is proper to adhere to the degree above mentioned, and particular care

Take out the pin in B, and observe whether the ball в will remain at rest upon b; if not, turn the adjusting screw at c till it just remains upon A.

ought to be taken to lay the wire, intended for fusion, straight, without any bendings or angles in it. The wire used in the two last experiments was that which is commonly called watch-pendulum wire, which is flatted; and as it approaches very near to such a sharp edge as might be supposed to affect the experiment by permitting a dissipation of the electric fluid in its passage, round wires were tried, and the result was the same.

By the last experiment it appears, that breathing into the jar had increased its charging capacity nearly in the same proportion as it had done the batteries: after breathing, it received a charge sufficient to fuse four times the length of wire it did when clean; but by the weight in the electrometer, and also by the greater number of revolutions given before the discharge happened, it might be supposed that the jar had received only a double charge.

The following experiments are intended to shew the lengths of wire which are just fused by various quantities of electric fluid at the same intensity.

Exp. 3-For this purpose a second jar was placed at the wire L, the pin marked 30 was taken out, and 15 put in its place, two inches of the same sort of wire as used in the last experiment was placed in the circuit, every other part of the apparatus remaining unaltered; the machine was then put in motion, till в begun to ascend, while it was stopped, and before a could reach a, one of the jars was pushed from the wire L (to do which there is always sufficient time while the electrometer is in motion), the discharge was effected, and the two inches of wire was just fused.

Exp. 4.-The jar which was pushed away in the last experiment was discharged, and placed at the wire L, as before, and eight inches of the same sort of wire placed in the circuit; the outside coating of the jars either touched each other, or had a metallic communication. All the other parts of the apparatus remained as before, and the machine was put in motion till в began to ascend; the jar was not removed as in the last experiment, but suffered to discharge with the other, and the eight inches of wire was fused in the same degree as the two inches in the last experiments.

It is evident from the position of the apparatus, that the quantity of electric fluid discharged in the last experiment must be double that of the former; yet, in repeating the experiment I had different results, which made me again suspect the edges of the wire, I therefore resolved to take round wire, and of as large a diameter as could be conveniently fused.

Exp. 5, with three jars.-Iron wire of part of an inch in diameter, and six inches in length, was placed in the circuit; three jars were placed so that the balls proceeding from their insides touched the wire L, and their outside coatings touched each other. The machine was turned till в begun to ascend, the discharge was caused, and the whole length of the wire was just run into balls.

Exp. 6, with three jars, one removed.-Two inches of the same sort of wire was placed in the circuit in the same manner as the last,

and the three jars remained; the machine was turned till в begun to ascend, then one of the jars was drawn away, consequently only two discharged, and the wire just run into balls as the last.

Exp. 7, with four jars.-Wire of part of an inch was taken, and four jars placed in contact with the wire L, with their outside coatings in contact with each other, and eight inches of wire was placed in the circuit; the weight in the electrometer remaining as before; the machine was then put in motion till в begun to ascend, the discharge was effected, and the wire was fused and run into balls. The experiment was repeated with the same sort of wire 8 inches long; the discharge was just sufficient to run it into balls.

Repeated with nine inches of the same sort of wire, and the discharge caused it to be red hot the whole length.

Exp. 8, with four jars, two removed.-Two inches of the same sort of wire was placed in the circuit, all the jars remaining as in the last experiment, the machine put in motion till B begun to ascend, then two of the jars were drawn away; the discharge was caused, and the wire was fused and run into balls.

Repeated with the same sort of wire 24 inches long, the discharge caused it to be red hot the whole length.

Exp. 9, with fourteen jars.-Wire of part of an inch diameter was taken, eight inches long, and proceeded according to experiment 7; it was fused and run into balls.

Exp. 10, with fourteen jars, seven removed.-Two inches of the same sort of wire was taken, and proceeded with according to experiment 8; it was fused and run into balls.

The results of the foregoing experiments proves sufficiently, that double quantities of electric fluid, in the form of a discharge, will melt four times the length of wire of a certain diameter: and experiments 5 and 6 prove that when one-third part is added to two, three times the length of wire was fused.

These experiments give reason to apprehend some error in Dr. Van Marum's experiments, because he found his batteries to increase in power only in proportion as the coated surface was increased, viz. that double surface of coated glass only could fuse double lengths of wire of the same diameter.

The doctor might, perhaps, have been led into a mistake in the following manner: first, he may not have charged the batteries to an equal height, as he did not at that time possess an electrometer of sufficient accuracy for that purpose; and, secondly, he may not have been aware of the different degrees of fusion caused by electric discharges, but only judged of the force by the wires being converted into balls; by which great mistakes may happen. For if a wire be taken eighteen inches long, and of such a diameter that when a jar or battery is charged to such a height as just to cause it to run into balls, much shorter lengths of that same sort of wire may be subjected to the same force, and still be only converted into balls by it; even if only seven inches were taken, nothing but balls will appear; the only difference will be, that the balls will be smaller, and dispersed

to a greater distance, which might be easily overlooked. If six inches of the same sort of wire be taken, it will be converted into balls and flocculi, or brown oxide of iron; so that to be accurate in this point, the lowest degree of fusion must be had, which is known when the charge has passed by the wire being seen red-hot the whole length, and afterwards run into balls.

Having now sufficiently proved, by experiment, in what proportion different quantities of electric fluid act upon different lengths of wire, which was required to be known, in order to explain in what proportion the charging capacity of a jar or battery is increased by breathing into it, before the charging begins, I shall proceed in the next place to explain this point.

The opinion that I had at first entertained (though supported by Dr. Van Marum's experiments), that I had found out a method of increasing the charging capacity of batteries to three times their usual force, was not supported by the facts that the usual power of a clean and dry battery, containing seventeen square feet coated surface, namely, that of fusing from eighteen to twenty-two inches of iron wire of part of an inch in diameter, will be increased by breathing into the jar, so as to become capable of fusing sixty inches. If the first-mentioned effect be taken at a mean, it will be twenty, then the increased effect, gained by breathing, will be just, as determined by the wire; and experiments 5 and 6 prove, that in order to produce such an increased effect, an addition of part of the coated surface must be added to the battery, which is about 816 square inches. This would amount to an addition of fifty-four square inches to each jar; or, in other words, if that quantity of coating could be added to each jar, the same effect would be produced as when breathed into. But this would require the coating to be within an inch of the top, which would render the battery unchargeable, at least, to that degree. A battery of fifteen jars constructed in the usual manner, will, therefore, by this treatment, become equivalent in power to twenty-one jars of the same kind, if clean and dry.

To explain the effect of breathing into the jars appears to be a matter of some difficulty. This experiment has been shewn to several electricians, and different opinions have been advanced, most of which seem to imply, that breathing acts as a coating to the uncoated part, which will appear in the sequel to be absurd. Mr. Nicholson's opinion (see Philosophical Journal, vol. ii, page 219) comes much nearer to the truth, though it does not appear to me to be sufficient to amount to the effect produced. I admit, with him, that a spontaneous explosion over the uncoated part is most commonly caused by undulation; but that this undulation is caused by the discharging of different charged zones, willbe difficult to prove, because such zones cannot exist upon clean and dry glass.

When the uncoated part of a Leyden jar is made perfectly clean and dry, and the jar set to the conductor of a machine in action, it will begin to charge, and, while charging, the coated part of the

jar, and the wire which is connected with it, become equally charged, and each endeavours to throw off that surplus of electric fluid which is forcing into them; the coating from its edge upwards, and that part of the wire which is above the coating and within the jar, will endeavour to throw it in all directions, which will cause it to be surrounded by an electric atmosphere, increasing in density as the charge increases. This atmosphere, together with that given out by the coating, fills the whole jar. Part of the electric fluid forced into the coating enters the surface of the glass, but the uncoated part, being clean and dry, both withinside and without, the inside resists the fluids entering its surface, which is kept suspended at a distance, because the natural electric fluid contained on the outside, finds no means of escape. But the action of the machine still continuing, presses it still closer to the surface, and at last overcomes that resisting force, and some of the particles on the outside give way, which causes an undulation in the inside, and the electric fluid closes instantly in upon its inside surface, and forces a greater quantity from the outside. Flashes, or coruscations, are thus caused, which are always seen when a jar is charging in the above mentioned circumstances: the charge still continuing to be made, forces another quantity from another part of the outside of the jar, and causes a second coruscation and undulation, which may be so strong as to cause a spontaneons discharge; or two or three more coruscations and undulations may happen, before the discharge, according to the steadiness or unsteadiness of the action of the machine, the quantity of electric fluid thrown off from the outside at each undulation, and also the degree of dryness and cleanness of the uncoated part of the jar. A discharge sometimes happens without having previously occasioned any perceptible coruscation. This is the case when the first undulation has been so strong, as to cause the whole discharge with the first coruscation, the one being so quickly followed by the other that it is imperceptible.

A jar will sometimes, while it is charging, give a great many small coruscations, quickly succeeding each other, which afterwards cease without having caused a spontaneous explosiou, though the action of the machine be continued. This happens when the uncoated part is nearly clean and dry, but not perfectly so; its surface still containing some conducting particles, but not so connected that the electric fluid can pass from one to the other without leaps, or small coruscations on the outside, which permit the electric fluid to spread gradually over its inside surface, and prevent the undulations from being so strong as to cause a discharge.

After this explanation of the cause of the flashes, or coruscations, which are seen upon the uncoated part of a jar while charging, and also that such coruscations produce undulations, which terminate in a spontaneous explosion; it remains now to explain how a jar is charged when the coruscations are prevented by breathing upon the uncoated part.

When a coated jar is breathed into, and then subjected to the