Lightning Discharges and of the Aurora Borealis. 349 



photograph. A brief consideration, however, of the laws of 

 electrical oscillations shows, I think, that such writers are 

 mistaken ; for the rate of decay of the amplitude of such 



oscillations is expressed by the well-known factor e 2L . In the 

 case of the brush-discharge R is enormously large. A re- 

 sistance of thirty or forty ohms was sufficient to completely 

 damp the oscillations of the sparks studied by me in the 

 research on the damping produced on iron wires*. In the 

 case of the brush-discharge, although we may be dealing 

 with very small values of self-induction and small values of 

 time, we have, on the other hand, great values of R. I 

 believe, therefore, that the brush-discharge is reduced to the 

 case of one throb, which is analogous to the pilot spark in 

 disruptive discharges. In regard to the Aurora, it may be 

 urged that the resistance of the rarefied air is not enormous. 

 In answer to this it can be said that the phenomenon of the 

 Aurora can be best reproduced by intercalating a tube 

 of rarefied air with some megohms of a water-resistance 

 between the terminals of a suitable transformer. The sup- 

 position that the Aurora is produced by the action of ex- 

 tremely rapid electrical oscillation on molecules of rarefied 

 air is not borne out by the theory of transient currents ; and 

 experiment shows that the phenomenon of the waving and 

 apparent stratification observed at times in the Aurora is due 

 to the redistribution of the lines of force which is produced by 

 suitable earths or conductors in the shape of regions of cloud 

 or moisture. 



The comparatively small resistance of the electric spark in 

 air, noticed by many observers, is due, I believe, to the per- 

 manence of path ; for this path is intensely heated, and is 

 practically a charred hole in the air. When this path no 

 longer becomes such a hole and the heated air rises and is 

 dissipated, the oscillations of the electric spark become 

 rapidly damped, and we have the phenomenon of the brush- 

 discharge— a glow at each of the spark terminals without a 

 disruptive discharge — the lines of force crowding from one 

 terminal seek the other terminal through the air of the room, 

 and in passing through rarefied air the energy along the lines 

 of force is manifested by molecular actions which are appa- 

 rently protean in form. I see, therefore, no evidence for 

 believing in the rapid oscillation of the Aurora. 



Jefferson Physical Laboratory, 

 Cambridge, U.S. 



* Phil. Mao-. Dec. 1891. 



