﻿EQUIPMENT AND WORK OF AN AERO-PHYSICAL OBSERVATORY. 27 



general use. Disruptive discharges occur when the stress in the atmos- 

 phere between the cloud and ground exceeds a certain value, determined, 

 of course, by conditions of pressure, humidity, and dustiness. As the 

 heavily charged cloud approaches the locality where the electrometer is 

 placed, the needle indicates a steadily increasing strain. When this 

 tension exceeds the dielectric strength of the air the lightning occurs. 

 The dielectric strength of the air, according to Sir William Thomson, 

 under ordinary conditions of temperature and pressure, is about 9,600 

 grains weight per square foot, or about 1.37 pounds. This is equivalent 

 to 656 dynes per square centimetre. The pull, then, which the air ordi- 

 narily can withstand is not over .67 of a gramme per square centimetre. 



In the daytime many flashes of lightning may pass unnoticed, and 

 even in the night-time some may be unobserved ; but, aside from these, 

 there exists a myriad of minor discharges which are all unknown to the 

 eye. Hence the electrometer method, which enables us to take cogni- 

 zance of minor discharges, constitutes a decided advance. Another w^ay 

 of accomplishing the same end would be the employment of properly 

 tuned Hertzian resonators. It is quite possible to time lightning flashes 

 without seeing them, and this we accomplished a few years ago in the tow^er 

 of the Smithsonian Institution. An observer, with watch in hand, was 

 asked to time all flashes which he saw. Meanwhile, in a darkened room, 

 we studied the movements of the needle. The times of lightning were 

 found to correspond with certain disturbances in the potential. The 

 agreement is very close, except that there will always be found to be 

 more disturbances than recorded flashes, indicating, perhaps, that there 

 are discharges which the eye does not see. 



Thunder-storms are not the only atmospheric disturbance in which 

 the electrification of the air varies in a noteworthy manner. We have 

 found most remarkable perturbations of the potential occurring during 

 snowstorms. We give the following record at great length because we 

 think it is the most accurate record as yet available. It shows that a 

 snoiostorm is closely akin to a thunder-storm. 



A tmospheric JSleciri. city 



JleatZiyhtiny 



We have, too, the potential variations during heat lightning, and still 

 more, the relations to auroral displays, all awaiting systematic investi- 

 gation. Is it too much to say that in no other way can research be so 

 profitably pursued as in connection with the electrification of the air ? 



