J. Trowbridge — Electrical Discharges in Air, 193 



of the sparks, noted down his impression of the apparent direc- 

 tion of each spark, while another observer reversed the poles 

 which charged the rheostatic machine. On comparing the 

 notes of the two observers, it was found that there was no 

 agreement in regard to direction. This result was to be 

 expected from the oscillatory nature of the discharges. It may 

 be that in the case of lightning the eye is forcibly impressed 

 by the greater brightness of the positive terminal in the cloud, 

 and the observer concludes that the flash has a unidirectional 

 movement. 



When oscillating sparks of the nature represented in figures 

 2, 3 and 4 are passed through Crookes tubes of the focus tube 

 pattern, it was found that photographs could be taken on plates 

 exposed to the inclined surface of the platinum, both when it 

 was made the anode and when it formed the cathode. ISTo dif- 

 ference in definition could be noticed. There was, however, a 

 great difference in actinic effect. Under the oscillatory nature of 

 the Leyden jar discharge the electrodes become alternately pos- 

 itive and negative. Possibly some of the want of definition 

 noticed in liontgen photographs, taken even with the aid of 

 electrical machines, may be due to the fact that the oscillatory 

 discharge does not always emanate from the same point on the 

 anode surface. A small anode should therefore give sharper 

 images than one of a large surface. 



What is supposed to be a resistance in the case of the voltaic 

 arc, and in the modifications of this arc seen in discharges from 

 high tension transformers, and in powerful electric sparks, and 

 presumably in lightning discharges, is a polarization which pro- 

 duces a variable difference of potential at the spark terminals. 



The inconstancy of spark potentials has been shown by Jau- 

 mann.** In working with a revolving mirror, it is found that 

 the spark terminals have to be brightened in order to preserve 

 the same spark length. I was interested also to observe the 

 effect of the surrounding medium upon the spark potentials. 

 Platinum terminals in sodium vapor showed the polarizable 

 condition : but there did not appear to be an appreciable 

 change in resistance, apart from this polarization. The same 

 was true when bromine vapor surrounded the spark terminals. 

 In the case of Crookes tubes, it is customary to apply heat if the 

 discharge will not pass through the tube : and some makers 

 provide a connecting receptacle which contains a substance 

 which, on being volatilized, modifies the internal conditions of 

 the tube. This modification is often spoken of as a diminution 

 of resistance of the tube. It should be more properly termed a 

 method of modifying the state of polarization of the electrodes. 



Jefferson Physical Laboratory, Harvard University. 



* Wied. Annalen, lv, p. 656, 1895. 



