189 l.J the Discharge of Ley den Jars. 23 



have this effect. Resistance at Q prevents any sudden effect of the 

 A spark 011 the long circuit, and therefore never calls out, a spark at 

 B at all : the charged wire discharges leisurely through resistance at 

 Q, and accordingly (there being no jar) the spark at A is quiet. 



The fact in the table not immediately intelligible is the extra 

 length of E spark caused by insertion of resistance at P ; or, to a less 

 extent, at 0. It would appear to indicate the effect of surgings in 

 the conductor, which accumulate a momentary opposite charge on one 

 of the knobs before the one partitioned off by high resistance has had 

 time appreciably to discharge. 



EXPERIMENTS ON RECOIL KICK. 

 Early Observations (1 and 2 March, 1888). 



24. Although the overflow experiments are evidence of the 

 momentum of a reflected pulse, an idea which is intended to be 

 conveyed by the term "recoil kick," yet I have been accustomed to 

 apply the term specially to cases where reflexion takes place at the 

 free end of a long wire, constituting an appendage or lateral exten- 

 sion, without forming any necessary part, of a discharging circuit. 

 Usually a pair of similar wires were employed, and their ends were 

 brought near enough for the momentum of the recoiling pulse, when 

 spitting off from each wire, to bridge the interval and thereby cause 

 a regular spark. If the wires were too far separate, a momentary 

 brush leaped from each end and subsided again. Occasionally these 

 brushes extended over a considerable length of wire, giving them a 

 peculiar luminous appearance at each discharge. The fact that the 

 brush was an up-rush and subsidence was shown by the similar 

 appearance of both the wires, and by the fact that if a spark from 

 either wire was tat en into a jar the jar was not found to be charged 

 by it. 



25. The brush or sparking out from the wires, which I call the 

 recoil kick, is most marked at certain places on those wires; and 

 usually at the distant ends. This was what called attention to the 

 effect (see above 3), I find that Mr. A. P. Chattock obtained the 

 first direct evidence of it, in some experiments with my apparatus 

 which he made in my absence on March 1, 1888. The plan of the 

 particular connexions used by him (fig. 14) has no importance, but it 

 sufficed to show how much more readily a long spark could be 

 obtained at the far end of long wires than at the near end ; and 

 Mr. Chattock was quite clear about the effect being due to reflected 

 electric pulses or stationary waves in the wires, and was prepared to 

 look for evidence of nodes and loops if the wires had been long 

 enough. 



In fig. 14, W is a high liquid resistance, the two long wires are the 



