I 



M. B. W. Feddersen on the Electric Spark. 505 



electricity to that which occurs by the so-called dark discharge*. 

 On diminishing the resistance, the bands became gradually 

 fainter, spreading further and further into the dark space, until, 

 by sufficiently small columns of liquid, this entire space appeared 

 filled with a luminous glow (figs. 16 and 17). The appear- 

 ance of these bands was always introduced by a spark, and has 

 nothing M'hatever in common with the glow which may often be 

 observed upon the limb of the spark-micrometer when the flask 

 is charged for a spark of great length. The latter could not, on 

 account of its feebleness, be seen at all in the rotating mirror. 

 The bands mentioned generally ended suddenly in one or more 

 partial sparks. In those rare cases where this did not occur, 

 they gradually diminished in intensity until they entirely dis- 

 appeared. 



Intermittent discharge. — The second kind of discharge con- 

 sists in a series of at first equidistant partial sparks, whose di- 

 stances from one another gradually become greater towards the 

 end of the discharge (fig 19). It is true that the intervals by 

 different discharges are not always, ceteris paribus, the same. 

 AVhen, however, the intervals are of a minimum magnitude, the 

 greatest regularity occurs in the discharge of the sparks. We 

 may therefore M'ith probability assume that such discharge is 

 the normal one for the resistance employed, and that this is 

 merely interrupted, giving place periodically to the continuous 

 discharge when imperfect isolation or other disturbing influences 

 come into play during the discharge. The intervals of time, 

 moreover, which separate the regvUar partial discharges, do not 

 correspond to times of electrical rest ; for continuous discharge 

 takes place during these periods, as is plainly noticed when 

 the resistance is small, for then the intermediate spaces do 

 not appear absolutely dark, but the sparks are only maxima of 

 light. By greater resistance, however, the continuous luminous 

 appearance is too feeble to be recognized between two con- 

 secutive partial sparks. If, moreover, we remember that the 

 first kind of discharge is more easily obtained by small resist- 

 ances, and the second by greater ones, we can understand why 

 the intermittent discharge gradually degenerates into a con- 

 tinuous one as the resistance diminishes, and why, when the 

 metallic circuit connectors are short, the latter exists alone 

 under all circumstances. 



Neither of these species of discharge, moreover, occurs with 

 great constancy; for, generally speaking, an irregular mixture 

 of both is observed ; that is, separate sparks introduce the con- 

 tinuous discharge. The latter varies very much with one and 



* Faraday's ExiJcrimcntttl Researches, par. 1544. 



