[ 205 ) 



XXXVII. On the Fracture of Electrical Jars ly spontaneous 

 Discharges. By Mr. Thomas Howldy. 



To Mr, Tilloch, 



Sir, — JL HE perforation or fracture of Leyden jars, which some- 

 times takes place when they explode spontaneously, and at 

 other times when they are intentionally discharged, is a phee- 

 nomenon which has often obtruded itself upon the notice of 

 electricians; and yet, so far as I can learn, its cause has not 

 been investigated with much care. My attention was forcibly 

 attracted to this subject a few years since, by the following cir- 

 cumstance : — A small electrical battery of nine jars, that had been 

 charged prettv high, exploded spontaneously at the moment 

 when I was going to discharge it. On examining the jars after 

 the explosion, it was found that three of them had been per* 

 forated nearly in the middle between the top and bottom of the 

 coating. The jars of this battery were four inches in diameter 

 and twelve high, having uncoated margins two inches broad. 

 As I was not then aware that similar accidents had occurred to 

 other experimenters, and as I had never before had more than 

 one jar broken by any single explosion of the battery, I was 

 greatly surprised at the fracture of tlsree on this occasion, and 

 could not account for it in any manner satisfactory to myself. 

 The most plausible conjecture that I was able to form concern- 

 ing its cause was, that the communication between the charging 

 wires and the inside coatings of the jars was not sufficiently 

 extensive; so that, when the discharge of the battery was made, 

 the whole charge of each jar, rapidly converging to the single 

 and very limited point of contact between its coating and the 

 charging wire, experienced a proportionate condensation and 

 resistance on entering it, which caused a momentary retarda- 

 tion of its velocity, and consequently a momentary reaction of 

 the charge from that point against the sides of the jars ; which 

 reaction, when the charge had considerable intensity, might be 

 sufficient to burst one or more of tliem. 



I determined, therefore, to make some experiments with the 

 least injured of the fractured jars, in order to ascertain the pro- 

 bability or improbability of this conjecture. With tliis intention 

 I removed the damaged jars from the battery; and finding that 

 one of them was simply perforated without having any lateral 

 cracks extending from the hole, I took from round it nearly an 

 inch in breadth both of the exterior and interior coating, and 

 made each side of the glass perfectly clean. The jar was then 

 placed directly below the ball and wire at the extremity of the 



prime 



