514 M. B. W. Feddersen on the Electric Spark. 



Resistance = 240 millims. reduced length of water column. 



On looking at these two Tables, we are at once struck with the 



fact that for both resistances the relation -r- of the eliminated 



electricity to the original charge increases with the distance of 

 discharge; and this is the case both with the mean value 

 obtained from a number of observations, and for the minimum 

 and maximum value of them. When I made the observations 

 described above, the relations of the intermittent discharge were 

 entirely unknown to me. The result observed was consequently 

 a source of great surprise to me, because it appeared scarcely 

 reconcileable with any theoretical view. Proceeding, namely, from 

 the hypothesis of a continuous discharge, there is no reason 

 why, even when the resistances are great, as well as when the 

 metallic circuit connexion is short, such quotient should not be 

 constant with the most various charges*. Admitting, however, 

 an intermittent discharge, it is possible for a difference in this 

 respect to exist between great and small charges, for by the 

 latter the continuous discharge alone may take place. In what 

 way, however, the intermittent discharge by itself can cause an 



B . . . 



increase in the quotient -v-, is not immediately apparent ; for 



such influence must be caused either by the number of the 

 partial discharges or the value of a single partial discharge, or 



* The circumstance that the quantities of electricity were measured by 

 the distance of discharge, would here only account for a very small diver- 

 gence from the proportionality. 



I 



