same whence, it may be inferred, that the number of revolutions of the electrical machine is, under the above condition, a fair measure of the relative quantity of matter. 3-That the explosions of a second jar, charging from the one coating, are proportional to the quantity of matter thrown upon the other. The quantity of matter therefore, may be easily estimated by the number of these explosions. This last method of estimating the quantity of matter, is in most cases, much to be preferred; since it renders the experiment independent of any irregularity in the state of the excitation; which in ordinary circumstances, will frequently happen. The quantity of matter however, may be always accurately estimated, by the revolutions of the machine, if sufficient pains be taken in the construction and use of the apparatus.

14. Having determined these laws, relating to the quantity of matter accumulated, we come now to the consideration of the free action, or intensity, it evinces, under various circumstances.

When a given quantity of matter is disposed on a simple insulated conductor, it endeavours to fly off and discharge itself upon any body which happens to be within the sphere of its action: but if the surface of the insulated conductor on which the matter is accumulated be increased, then the tendency to discharge is lessened. To express this tendency of the accumulation to pass off in any given direction, the term intensity has been employed. Intensity therefore, is but another term for what may be called free action.* Now

*It has been supposed, in Mr. Singer's excellent work on Electricity, that the diminished intensity of an electrified conductor, when its surface is extended, is merely the result of an exposure of the given accumulation, to a greater extent of unelectrified air;—but such an hypothesis, does not appear adequate to explain all the subsequent phænomena, since no extent of a perfectly insulating medium, can controul the free action of an electrical accumulation. It is quite

H

it is possible to controul the free action of any given quantity of electricity; not only by disposing it on a more extended surface, but also by dividing the surface so extended into two equal and similar portions, and placing one of the portions in a freely uninsulated state, parallel, and near to the other as in fig. 19; the two conducting surfaces being separated by some non-conducting medium. By this peculiar disposition of the surface, in the form of an insulated and uninsulated conductor, the electrical accumulation appears to be so determined toward the opposed uninsulated portion, that its tendency to discharge in any other direction is diminished. Suppose such an arrangement made, and let the original surface and the quantity of matter be doubled, then if the two opposed surfaces be sufficiently near, and the intervening medium be strong enough to resist the force of the accumulation, in the direction of the opposed conductor, its free action or tendency to discharge in any other direction than that of the opposed uninsulated portion, will be no greater than in the previous instance, when only half the matter was accumulated.

The given electrical accumulation therefore, is still under the influence of the same increased extent of conducting surface, whether it be in a divided or undivided state; although in the former case one portion of it operate on it at a distance: hence its free action will be more or less controlled in some ratio of the distance of the opposed surfaces, and the greater or

essential for example, in the construction of the Leyden jar, not only that the opposed coating be perfectly uninsulated, but that it be likewise a conductor of electricity. We should rather be induced to believe, that in extending the surface of an electrified conductor, the intensity of the accumulation becomes diminished, in consequence of its becoming diffused over, and retained by, a greater extent of conducting surface: much in the same way if it be allowable to employ a mechanical analogy, that any quantity of fluid stands at a much less altitude in a cylyndrical vessel of six inches diameter, than in a similar vessel of half that diameter; so that by increasing the diameter of the containing vessel, or in other words, the surface of the fluid, we diminish its pressure on any given point.

less resistance of the non-conducting and intervening medium. There is however, this difference in the actual condition of the two arrangements:-when the matter is disposed and retained by means of a continuous surface, the whole of that surface is supposed to be in the same electrical state: whereas, in the latter instance, half the surface only is in a similar state of electricity; the opposed uninsulated portion being in an opposite one. This seems to be a necessary consequence of the arrangement, and is in fact caused by the determination of the charge upon that portion; and without such inductive effect could take place, the accumulation could not to any great extent be determined in that direction.

Although in the previous observations, we have considered the given conducting surface to be extended and divided into two equal and similar parts, yet it is evident, that the whole arrangement is nothing more, than opposing to the original insulated conductor, a second equal and similar conductor; by which means the same result may be obtained. The Leyden jar is in fact an arrangement of this description; since one of its coatings is insulated, and placed near and parallel to the other; the latter being in a freely uninsulated state, and the two coatings separated by an intervening nonconducting medium, The free action of the matter therefore, which we dispose upon the insulated coating, becomes controlled by the opposing uninsulated one.

15. If therefore two conductors (a) (b) (figs. 5 6 7) be separated by a given interval, and be connected with the insulated and uninsulated coatings, as in figs. 7, 14, 15, then the free action may be estimated by means of these conductors; 1-by measuring the interval between them, which it can break through. 2-by measuring its attractive force or tendency to discharge, both of which will increase in some direct ratio of the quantity of matter, and in some inverse ratio of the controlling power of the uninsulated coating; all other things

remaining the same; the latter being continually diminished as the quantity of matter thrown upon the insulated coating increases.

16. The former of these methods will be first considered. Exp. 10. Two similar jars, each containing five square feet of surface, being connected with the discharging electrometer, (fig. 5) and with the positive conductor as represented in fig. 14, and the interval between the balls adjusted to one-tenth of an inch, the discharge took place at the end of two turns and half of the plate. The interval between the balls being now doubled, and the small residuum in the jars neutralized, the discharge passed when five turns were completed. The interval being increased to three-tenths of an inch, or treble the first distance, and the residuum discharged as before, the explosion passed at the end of seven turns of the plate. On increasing the interval to four-tenths of an inch, the discharge took place at the end of ten turns.

Exp, 11. One of the jars in the former experiment being removed, and the interval between the balls adjusted to fourtenths of an inch, the discharge took place with five turns of the plate. The quantity of coated surface being now doubled by adding a second and similar jar, and the interval between the balls diminished to two-tenths of an inch, the discharge again took place at the end of five turns. The quantity of coated surface being again doubled by adding two more equal and similar jars, and the interval between the balls reduced to one-tenth of an inch, or to a quarter part of the first distance, the discharge again occurred at the end of five turns of the plate.

Exp. 12. A single jar being connected as before, and the interval between the balls adjusted to two-tenths of an inch, the discharge occurred at the end of two turns and half. A second similar jar being added, the discharge occurred with

five turns of the plate. A third similar jar being added, seven turns were required to complete the discharge: the interval between the balls remaining in each case the same.

Exp. 13. Two similar jars being connected with the conductor, and the interval between the balls adjusted to twotenths of an inch, the discharge took place with five turns of the plate. One of the jars, that is half the surface, being removed, and the interval between the balls quadrupled, the discharge occurred with ten turns of the plate or double the former.

17. These results as far I could ascertain, did not appear to be influenced by the way in which the surface was disposed; thus whether the extent of surface was made up with separate jars, or otherwise consisted of one large jar, still the same results were obtained, provided the areas of the coatings were accurately adjusted, and that the thickness of the glass did not vary. In repeating the above experiments in this way, four jars were carefully selected, and they were so prepared that the extent of coated surface on the first, was exactly five square feet; on the second and third, each, two and half square feet; on the last, one and quarter square feet. Hence it became easy to vary the surface by changing the jars; or by combining the two equal jars to make up a whole surface equal to the larger one. Under these circumstances, the two equal jars combined, discharged with the same number of turns as the larger one; and similar results to these already described, were obtained by varing the extent of coated surface in a similar way.

18. We may therefore conclude, that the free action of an electrical accumulation as estimated by the interval it can

* It will be necessary to keep in mind, that these results relate only to the free action, and are not to be assimilated with the effects of the explosion on metals under similar conditions.