THERMODYNAMICS OF ATMOSPHERE VON BEZOLD 277 



At least sleet and hail are certainly observed at elevated stations 

 more frequently than in the low lands. 



Equally do the large drops that not infrequently occur in thunder- 

 storm showers suggest that in such cases we have to do with melted 

 hail or sleet. 'I consider this assumption especially reasonable 

 since I have often had occasion to observe that the occurrence of a 

 heavy fall of hail is announced by the immediately preceding fall 

 of very large drops of rain. 



In such cases I have observed drops of such a size as can only 

 exist momentarily and can indeed only be explained as being melted 

 hailstones. 



I therefore consider it probable that sleet and hail play a greater 

 part in thunderstorms than we have generally assumed, and that 

 their relatively rare occurrence at the earth's surface is to be 

 explained by the fact that they frequently arrive in a melted con- 

 dition. 31 



The above-given presentation of the processes going on within the 

 thunder cloud appears to greatly favor the hypothesis framed by 

 Sohnke as to the electricity of the thunderstorms. 



On the other hand, I would say that for myself at least I do not, 

 on this account, accept the Sohnke theory. For, on the one hand, 

 it is difficult for me to understand how a permanent separation of 

 the positive and negative electricities can be brought about by the 

 mutual friction of the falling sleet or hail, since at temperatures 

 below freezing the water particles must immediately freeze together 

 with ice particles, whereas at temperatures above freezing the sur- 

 faces of the hailstones are already covered with water and hence 

 there can only be friction of water on water. 



On the other hand, even if it be possible to overcome this objec- 

 tion, still I cannot agree with the reasoning by which Sohnke 

 refers even the normal electricity of the atmosphere back to the 

 same source. Especially does it seem to me exceedingly improper 

 to attribute such great importance, even on perfectly clear days, to 



31 In these earlier investigations the processes that occur in thunderstorms 

 are considered only from purely thermodynamic points of view and the nuclei 

 of condensation are only those that have long been recognized. According 

 to recent investigations it is probable that electrons or perhaps also the 

 cathode rays proceeding from the sun play a part in the condensation. More- 

 over in the formation of hail electric processes seem to play an important 

 part. (See W. Trabert: Die Bildung des Hagels. Met. Zeit., 1889, XVI, pp. 

 433-447.) But since all these questions are far from b e i n g settled, therefore 

 I have not attempted to rewrite the whole memoir from these new points of 

 view. (Note added in 1905. W. v. B.) 



