272 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 51 



thunderstorms (which one may consider as a departure) I will now 

 return to the principal question, i.e., what rdle the supersaturation 

 and subcooling may play in thunderstorms. 



In this investigation I have in mind only the heat thunderstorms, 

 since the observational material available to me for true cyclonic 

 thunderstorms is too scanty and especially since I do not know 

 whether in these latter anyone has observed the peculiar variation of 

 pressure that can be almost invariably recognized in the barograms 

 during heat storms. 



As has already been stated, it is at present still difficult to decide 

 to what extent true supersaturations occur, since as yet we have no 

 sure foundation of experience on this point. 



On the other hand, it seems to me that the above-mentioned 

 observations lately published by Hellmann on the behavior of the 

 thunderstorm rains that pass from the west eastward over Berlin, 

 indicate that supersaturations do play a part in thunderstorms. 

 For the great clouds of dust and smoke that always exist over that 

 city must hinder the formation of the conditions of supersaturation. 



Again the fact demonstrated by me many years ago that buildings 

 within populous cities are much less frequently injured by lightning 

 than those in the surrounding country, may indicate that the sever- 

 ity of the thunderstorms experiences a diminution which may be 

 referred back to similar causes. However, I readily acknowledge 

 that such considerations have only slight force as a demonstra- 

 tion. 



It is otherwise with the peculiar movements and uprisings that 

 the thunder clouds show even when they have not attained to ele- 

 vations at which we may expect subcooling. The shapes of the 

 clouds and especially their changes in appearance do not at all 

 correspond to those that we should expect from currents that are 

 steadily ascending and are accompanied by condensations that are 

 only the result of expansion. As already remarked, by close 

 observation of the clouds we can scarcely avoid the thought that in 

 their interior there are forces in action that cause the peculiar expan- 

 sions and projections of the individual cumulus-heads. One can 

 scarcely suggest any other forces than the heating that must occur 

 within the clouds when there is a sudden release from the super- 

 saturated or subcooled conditions. 



This assumption of the occurrence of supersaturate m also receives 

 an important support when we study the variations of pressure dur- 

 ing thunderstorms in cases where the clouds attain only slight 

 altitudes, as is not infrequently the case in the Riesengebirge, accord- 



