the pith balls should just separate positively, and what size must be given to it that they should just separate negatively.

I then removed the two plates and suspended the tin cylinder so as to touch the wire Mm, but without touching any other part of the machine, and found what size it was necessary to give to the coating of the sliding plate that the pith balls should separate as before.

By this means the charge of the tin cylinder was found to be to that of the two plates as 1:33 to 1. Therefore the charge of the two plates seems to bear pretty nearly the same proportion to that of the cylinder whether the electricity is of the usual strength or very weak. But if we suppose that the electricity spreads 07 inches on [the] surface of glass with the usual degree of electrification, and that it does not spread sensibly with the weak degree of electrification, then the proportion which the charge of the glass plates bears to that of the cylinder should be less with the usual degree of electrification than with the weak one, and that by about part.

This difference, however, is not more than what might very well proceed from the error of the experiment.

362] On the whole, I am uncertain whether the charge of a glass plate would really bear a rather less proportion to that of a globe or other body when the electricity is strong than when it is weak, provided the electricity was prevented from spreading on the surfaces as it should seem by these experiments, or whether it was not rather owing partly to the error of the experiment, and partly to there not being so much difference in the distance to which the electricity spreads on the surface of the glass according to the different degree in which it is electrified, as I imagined.

If the first of these suppositions is true, I do not know how to reconcile it with the theory, except by supposing that the greater the force with which the plate is electrified the less is the depth to which the electricity penetrates into the glass, or the less is the thickness of the spaces in which we supposed the fluid to be moveable.

Though it seemed natural to expect that the electric fluid should penetrate further into the glass, or that the fluid within the glass should move through a greater space when the glass was

364]

STRONG AND WEAK ELECTRIFICATION.

179

strongly electrified than when weakly, that is, when the force with which the fluid was impelled was great than when it was small, yet it is not strange that it should be otherwise, as it is very possible that the electric fluid may penetrate with great freedom to a certain depth within the glass, and that no ordinary force shall be able to impel it sensibly further, and in like manner it is very possible that the fluid may be able to move with perfect ease in the space ae (Fig. 27) and yet that no ordinary force shall be able to move the fluid at all beyond that space.

But it would be very strange that the fluid should penetrate to a less depth within the glass, or that the fluid within the glass should move through a less space when the glass is strongly electrified than when weakly.

363] The reader perhaps may be tempted from this circumstance to think that the reason of the actual charge of the glass plates so much exceeding their computed charge is not owing to the electric fluid penetrating into the glass, or to any motion of the fluid within the glass, but to some error in the theory. But I think the experiments on the plate of air [Art. 344] form a strong argument in favour of its being owing to the penetration of the electric fluid into, or its motion within the glass, for it appears plainly from these experiments that the electric fluid does not penetrate into the air, and on account of the fluidity of the air it seems very improbable that the electric fluid within the air should be able to move in the manner we supposed it to do within the glass; whereas it appears plainly from Dr Franklin's analysis of the Leyden vial, that the electric fluid does actually penetrate into the glass.

Therefore as this excess of the observed charge above the computed does not take place in the plate of air, where it could not do it consistently with the theory, but does in the glass plate, where it may do so consistently with the theory, I think there seems great reason to think that it is not owing to any defect in the theory, but to some such motion of the electricity as we have supposed.

364] I could not find that there was any difference in the proportion which the charge of a glass plate bore to that of

another body whether they were electrified positively or negatively*.

365] It was said in Art. [331], that there seemed no reason to think that the charge of the plate D, or of any other of those glass plates was sensibly greater than it would be if the electricity was spread uniformly on their surfaces, whereas the charge of most of the plates of air was found very considerably greater than it would be on that supposition. But this is by no means inconsistent, for according to the first way of accounting for the great excess of the real charge of those plates above the computed, namely supposing that the electricity penetrates into the glass to the depth of of its thickness, the increase of its charge on account of the electricity being not spread uniformly, should be not greater than it would be if the glass was only of its real thickness, and the electricity was unable to penetrate into it at all, and therefore should not be greater than it is in a plate of air in which the thickness is of the diameter, and should therefore in all probability be quite imperceptible.

And by Prop. XXXVI. [Art. 170], the increase of charge should hardly be much, if at all, greater according to the second or third way of accounting for this phenomenon.

366] In order to try whether the charge of coated glass is the same when hot as when cold, I made use of the apparatus in Fig, 28, where ABCba represents a short thermometer tube with a ball BCb blown at the end and another smaller ball near the top. This is filled with mercury as high as the bottom of the upper ball, and placed in an iron vessel FGMN filled with mercury as high as FN. Consequently the ball BCb was coated as a Leyden vial, the mercury within it forming the inside coating, and that in the vessel FGMN the outer one.

In trying it, I set the vessel FGMN on the wooden bars of the machine represented in Fig. 20, near the end NP, and dipt a small iron wire bound round the wire Mm into the mercury within the tube, so as to make a communication between the wire Mm and the inside coating, the outside coating, or the mercury in FGMN, being made to communicate with the ground.

[Art. 463.]

† [Art. 556, March 21, 1773. See also Arts. 548, 549, 680.]

would not find that there was any difference in the Tetrated other the charge of glass plate love me proportion to that of another body whether they were electrified pontively of egatively

G

Sorder to try whether the charge of sbated day, is the same when hot as when cold Tomade use of the appar. in fig.28 where ABC6 a reporting, a thart thermom tube with a ball Bll blown at the and another smaller ball near the tape This is filled with & as high

as the

ball & placed in an van vessel FGM & filled with of as heigh as To N consequently the butt BC.b Spated

was charged as a Leyden veal the of with

To face p. 180.