and of the Aurora Borealis. 199 



discharge or mere luminosity at either of the spark terminals, 

 is a far simpler phenomenon than the luminosity produced in 

 rarefied tubes ; for in the latter phenomenon we have the dis- 

 sociation and impact of molecules, and we must consider all 

 the problems of atomic motion in addition to those of the 

 oscillatory nature of electrical waves. It is not my purpose 

 to enter into a consideration of the molecular movements in- 

 volved in oscillatory discharges in vacuum tubes; but having 

 discussed some of the general features of discharges of elec- 

 tricity in air at the ordinary pressure I shall endeavor to trace 

 the connection between such discharges and the phenomenon 

 of the Aurora Borealis. To my mind the luminosity in a 

 vacuum tube held in one hand while the other hand grasps the 

 terminal of a Ruhmkorf coil, closely represents the phenome- 

 non of the northern light, for we have in this case a discharge- 

 of electricity from a higher level to a lower through a rarified 

 medium. Although in this paper I restrict myself to a discus 

 sion of the general relations between discharges of lightning 

 and the phenomenon of the Aurora Borealis, and do not enter 

 into a study of the molecular movements excited by electrical 

 discharges, I am impelled to devote a few words to the sub- 

 ject of the stratified discharge and to show that it has no con- 

 nection with the oscillatory discharge of electricity such as we 

 are considering. The distances between the stratifications do 

 not seem to be changed by modifying the period of forced oscilla- 

 tion given to the transformer over a wide range. I have pro- 

 duced the stratifications by employing an interrupter of a 

 Ruhmkorf coil giving from sixty to one hundred vibrations 

 per second, and by the use of two alternating machines, one 

 giving 300 to 400 alternations per second and the other 900 to 

 1000 alternations per second. The distances between the strati- 

 fications do not seem commensurate with the rate of alterna- 

 tion of the exciter of the transformer. On the other hand 

 the distance between the stratifications is not dependent upon 

 the amount of self induction in the circuit. In one case 

 quadrupling the self induction reduced the distance between 

 the stratifications one-half. This reduction was due not to the 

 increased self induction but to the increased resistance ; for the 

 introduction of a water resistance of some megohms dimin- 

 ished in a similar manner the distance between the stratifica- 

 tions. In short I could not discover any connection between 

 the law t = 27rVLC and the phenomenon of stratification. 



I have said that we can pass by insensible gradations from 

 the condition of the brush discharge to that of the disruptive 

 discharge. By intercalating a non-inductive water resistance 

 and a vacuum tube between the terminals of a suitable trans- 

 former we can exactly imitate the phenomena observed when 



