1891.] the Discharge of Ley den Jars. 11 



Effect of High Resistance. 



7. Interpose the capillary liquid tube ( megohm) in the circuit 

 of the thick copper wire, putting it at one or other end of it, and the 

 jar refuses to overflow, although the spark-length A is increased to 

 2-^ inches. 



The spark is quiet, long, and zigzaggy. The resistance has the 

 same effect at either end, but the spark seemed straighter when the 

 resistance was at jar end of long wire. 



To test effect of putting resistance into the middle of a long con- 

 nector, both the thick wires round room (one copper, the other iron) 

 were joined in series and used as connector. Overflow began when 

 A = 0'6 inch. The wires were now disconnected at their far ends, 

 and the capillary tube made to bridge the gap. The jar now refused 

 to overflow, though A was more than trebled in length. (Fizzing 

 stopped it at that point.) 



Contrast between C Path and Overflow. 



8. But when an artificial overflow path is supplied to the coatings 

 (as indicated by the strong line to a C knob in fig. 5) the matter is 

 different. It does not now feel the effect of a long circuit as different 

 from that of a short one. The space at C being O94 inch, a spark 

 jumped there sometimes and sometimes at A = 0'75, with the high 

 resistance interposed in the two long leads ; and just the same hap- 

 pened when the resistance was removed and the long wires directly 

 connected. 



Shorten A to 0'64, and it was unable to select C, but it jumped the 

 lip of the jar instead. It preferred 8 inches of jar-lip to 1 inch 

 between the C knobs. When strong enough it would seem to go 

 at C ; when too weak for that it jumps the edge ; but this is not 

 a clear account of the matter. A better statement is the follow- 

 ing : 



An A spark precipitates an overflow (i.e., over the lip of the jar), 

 but it does not precipitate a C spark. When a spark occurs at C 

 there is quiet at A. The A and C sparks are alternative, not 

 simultaneous. Moreover a C spark does not cause overflow. An 

 A spark can easily occur without the edge of the jar being jumped, 

 but the edge is never jumped without an A spark. (Connexions 

 being as in fig. 4, with the addition of a short C or artificial overflow 

 path, as shown by the thick line in fig. 5.) 



Long Connector in C Circuit. 



9. But now the thick copper spiral above mentioned (6) was 

 arranged to connect one of the knobs with the outer coat of its jar 



