1891.] the Discharge of Ley den Jars. 21 



of the discharge is what makes the difference. The spark through 

 high resistance, instead of being alternating, can be seen to be in- 

 termittent {i.e., multiple), when analysed in a revolving mirror. 



There is no need in these sudden rush experiments for the long 

 leads of fig. 9, though perhaps they add to the length of the sparks. 



22. Sparks thus obtained from the outer coats of jars are con- 

 venient for taking under water, or to water ; and the phenomena thus 

 seen are singular, and sometimes violent. 



Water acts mainly as a dielectric under these circumstances, and, 

 with small electrodes, such as the bared end of a gutta-percha wire, 

 the water between gets burst with extraordinary violence : often 

 breaking the containing glass vessel. 



This arrangement of Leyden jars should be handy for blasting 

 operations, because no specially good insulation of the leads is 

 necessary. 



EXPERIMENTS ON SURGING CIRCUIT PROPER. 



23. Although all the overflow experiments are controlled by 

 electrical surgings, I have been accustomed specially to apply the 

 name " surging circuit " to the case where sparks are obtained not 

 between two distinct parts of a circuit, but between two points on 

 one and the same good conductor, under circumstances when it does 

 not form the alternative path to anywhere, and when it would 

 ordinarily be supposed there was no possible reason for a spark at 

 all. For instance, in fig. 12 the loop of wire round the room is a 

 mere off-shoot or appendage of an otherwise complete and very 

 ordinary arrangement, and yet a spark can occur at E whenever the 

 ordinary discharge occurs at A ; a spark, too, often quite as long, 

 though not so strong, as the main spark at A. 



FIG. 12. 



earth 



The jar is not essential to this experiment ; and, in order to analyse 

 it by inserting resistance at various places, it was modified to fig. 13, 



