ATMOSPHERIC ELECTRICITY. 99 



Measurements of the electrification of falling rain or snow, simple as 

 they appear at first sight, are beset with very serious difficulty. We 

 owe the most complete investigation on the point to Messrs. Elster and 

 Geitel (Wiener Sitzungsberichte, Vol. XCIX, 1890). They find no 

 regularity in the electrification, though positive signs slightly prepon- 

 derate with snow and negative signs with rain. 



The approach of a thunderstorm announces itself by characteristic 

 cumuli clouds, and the general atmospheric condition favorable to their 

 formation is felt by many persons of nervous temperament. Many of 

 us are accustomed to hear that "there is thunder in the air." What- 

 ever the special feeling of "thunder" may be due to, it can not be an 

 electrical effect, for electrical instruments delicate enough to detect a 

 small fraction of the normal force, give no indications of the approach 

 of a thunderstorm, and it is only when the cloud has begun to dis- 

 charge rain or hail that strong electrical effects are noticed. During 

 the thunderstorm the electroscope is, of course, much disturbed, and 

 there are frequent and violent reversals of its indications. 1 The fact 

 that no effects are observed at the surface of the earth during the 

 approach of a thundercloud does not prove that there is no electrical 

 separation, for we may imagine two oppositely electric layers at different 

 levels producing a strong electric field between them, but only weak 

 effects outside. That some such things may possibly occur is indicated 

 by observations made in mountain districts, where violent electrical 

 disturbances are observed previous to the formation of clouds (Trabert, 

 Meteorologische Zeitschrift, 1889, p. 342). The cumulus cloud, from 

 which the lightning strikes out, is nearly always associated with a 

 cirrus layer above it, and the flash occurs more frequently upward or 

 sideward between the clouds than down to earth. Under such cir- 

 cumstances it is clear that instruments on the surface of the earth can 

 only very partially indicate the nature and distribution of electrical 

 stress in the neighborhood of the cloud. 



Thunderstorms seem always to be connected with a vortex motion, 

 and meteorologists distinguish two kinds of thunderstorms. The first 

 kind forms in the outlying portions of a large cyclonic system. The 

 storms which occur in winter are mostly of this nature, and the vortex 

 necessary for its formation is of the nature of a secondary disturbance. 

 The thunderstorm which forms in summer, on the other hand, makes 

 its own vortex, and is of a much more local character than that which 

 is produced round a previously established barometric depression. The 

 summer storm is much influenced by the character of a district. There 

 are certain configurations apparently favorable to its formation, as is 

 clearly brought out by the charts which have been made representing 

 their frequency. 



The route traveled over by the storm is affected by mountain ridges, 

 and rivers also seem to offer a peculiar impediment. Many of them 



1 Weber, Elektrotechnische Zeitschrift, Vol. X ; Elster and Geitel, " Ueber einige 

 Ziele und Methoden luftelektriscber Untersuchungen," Wolfenbiittel, 1891. 



