EXPERIMENTAL AND THEORETICAL.

539

tion, would accumulate to the greatest extent around the upper edge of the lining,and consequently would have a greater tendency to strike the wire from some pointnear the top of the lining than from any other part of the jar.

Although there are many jars that will withstand the utmost force of a discharge,nevertheless there can be no doubt but in every jar that is fitted up in the usual waythe fluid strikes the wire from the upper part of the lining in almost every dischargefrom high intensity ; and that those jars which have not broken owe their safety to thestrength of the glass. This conclusion may, I think, he fairly admitted when it isconsidered that many jars will withstand the shock for hundreds of times, althoughone of the balls of the discharging-rod be placed against the coating at every discharge,whilst many others, which are of more feeble glass, have broken at the first dischargeby this unskilful application of the discharging-rod.

If due attention be paid to the nature of the electric jar, and the manner in whichit is usually constructed, it will appear evident that during a discharge, if the wholequantity of fluid from the lining were to flow through the conducting channel pur-posely arranged for it, that portion which occupied the upper part of the lining wouldhave to travel by a very circuitous, and perhaps injudiciously chosen route. It wouldfirst have to descend to the bottom of the jar before it arrived at the axial wire, thencethe whole length of that wire before it arrived at the ball on its top. Moreover,whilst ascending the axial wire, the fluid would have to pass the very point fromwhence it first set out. Now the electric fluid has never been understood to evinceany tendency to move by a circuitous route, and more particularly so when one portionof it has to meet or pass close to another portion in a recurved circuit similar to thatpresented by the lining and axial wire of ajar. Moreover, it appears to me that thegreater part of the axial wire, which is within the jar, is in a negative condition, rela-tively to the lining and upper part of the stem and ball, from the moment the jar ceasesto receive fluid till it is discharged ; and this negation would be increased for amoment by the presentation of the discharging-rod to the ball of the jar, and thus placethat part of the axial wire in very ample condition for the reception of the fluidfrom the most vicinal and intensely-charged point of the lining, which, as alreadyassumed, would be in some part of its upper edge. There can be no just reason todoubt, however, that some part of the charge actually makes its escape by the chainfrom the bottom of the jar ; but the proportion of fluid discharged by this route isprobably very small, and especially in very tall jars, such as are most frequentlyselected for electrical purposes, where the fluid in some cases would have to traversetwo or three feet of lining, chain, and wire, and yet be no nearer the ball on the topof the jar than it was before the approach of the discharging-rod.

Now since the conclusions drawn from this investigation show that the electricfluid in the interior of an intensely charged jar indicates the greatest tendency to

3 v 2