1891.] the Discharge of Ley den Jars. 13 



Effect of Capacity. 



11. The spiral was now shunted out by a couple of Ley den jars in 

 series, i.e., with their knobs touching either end of it and with their 

 outer coats connected. If the jars only touched one end of the wire, 

 they had no effect ; but when they touched both ends, a larger A spark 

 was needed to cause overflow.,, 



With the spiral alone A = O53 



With the capacity shunt A = 076 



Experiments on Large Condenser. 



12. It was not desirable to expose the large condenser 1 to such 

 conditions as would make it want to overflow, because overflow with 

 it would mean bursting ; but one of the pint jars was arranged on it 

 as a safety valve, and it was then connected up to the machine. On 

 now taking machine spark at A, the pint jar might or might not 

 overflow its 4 inches. 



With very short connexions A = 0*5 inch did not overflow it. 



With wires each a yard or so long. . A = 0'4 inch was sufficient. 

 And with spiral of thick copper ... A = 0*3 inch was enough. 



Iron Core Again. 



13. Tried a stout spiral of brass wire (a spiral spring about a foot 

 long and an inch diameter) ; it made the jar overflow fairly easily. 

 Then inserted in the spiral a bundle of fine iron wires wrapped in 

 paraffin paper, but could detect no difference whatever, cf. 10. 



Summary. 



14. The noteworthy circumstance in all these experiments is the 

 remarkable action of a long thick good conductor in causing the jar 

 to overflow, especially if it be insulated, the most powerful con- 

 ductor for this purpose being one with considerable self-induction 

 and capacity but very little resistance. Evidently such a conductor 

 assists the formation of an electric surging, whose accumulated 

 momentum charges the jar momentarily up to bursting point. Re- 

 sistance damps the vibrations down, and short wires have insufficient 

 electric inertia and capacity to get them up. Iron, whether massive 

 or subdivided, shows no effect whatever on the effective inductance of 

 a circuit surrounding it. 



It is also noteworthy how far more readily a jar overflows directly 

 between its coatings over the lip than it does through a pair of 



