L890.] during Thunderstorms, and on the Brontometer. 



63 



id character of the thunderstorms of North Italy, considers the rise 

 due to secondaries on the skirts of cyclones. 



Professor Bornstein regards them as due to temperature changes 

 lot reaching to the upper strata of the atmosphere, in which view he 

 has partly the support of Professor Ferrel. 



Professor Klossovsky, of Odessa, says that, " Every storm, whether 

 with or without hail, is accompanied by barometric oscillation." If by 

 this phrase he means oscillations like those observed in other parts 

 of Europe, the phenomena must be different in Southern Russia, for 

 one of the difficulties in London, for example, is that they do not 

 occur with all storms, but only with some. 



In the ' Annuaire ' of the Montsouris Observatory for 1889, M. 

 Descroix, when referring to the subject, gives a barogram for 

 August 15th, 1888, which, with the accompanying notes, is so 

 typical that it is reproduced (reduced to the same scale as the others) 

 in fig. 2. 



FIG. 2. 



Montsouris, Paris, August, 1888. 



It will be noticed that several of the opinions above quoted are 

 contradictory, and it is not known that any one explanation is 

 generally accepted. 



The author was much struck by the remarkable curve given by 

 his Redier barograph on the evening of August 2nd, 1879, and still 

 more so when he found that the curve from a similar instrument at 

 Messrs. Lund and Blockley's, in Pall Mall, 2| miles S. by W. from 

 his own instrument, was nearly identical (figs. 3 and 4), and he has 

 long desired to investigate the subject. At last, in 1886, he definitely 

 decided upon the phenomena which he considered it necessary to 

 record mechanically. The author had read the description of Sir 

 Francis Ronalds's storm clock, and the objects being similar, that 



