22 



Mr. 0. J. Lodge. Experiments on 



[June 4, 



and the following readings taken : first, with a thin copper wire, and 

 then with a thick copper wire, round room. The \ megohm liquid 

 resistance could be inserted at either M, N, 0, P, or Q. 



The A knobs used were the small ones of a universal discharger, 

 1*4 cm. diameter, and 2'4 cm. apart all the time (equivalent to 

 1 ; 5 cm. spark-length between flat plates). The E knobs were those 

 of a spark-micrometer, and were 1'96 cm. diameter. 



FIQ. 13. 



This table evidently shows that the main part of the E spark is the 

 rushing of the charge in the N part of the wire back to the discharged 

 A knob. It has two paths, through the wire via 0, and direct across 

 the spark-gap E. Most of it chooses E, except when there is high 

 resistance at N or P. Resistance at interferes but little, and in 

 fact it may help more across E ; and resistance at M must certainly 



