[ 387 ] 



on which air at an equal height is fiill more rarefied, the mercury 

 muft fink in proportion to the quantity of air that deferts its 

 ftation. The fmall variations of mercury in the intra-tropical re- 

 gions proceed from the fmall alteration of the quantity of air in- 

 cumbent over thofe trads. The perturbations that take place 

 within the -tropics originate in far higher ftrata of the atmofphere, 

 than thofe do that take place in the more diftant extra-tropical 

 tracts. Thus, Gentil has fhewn, that thunder proceeded from 

 from clouds loooo feet above the furface of the earth at Pondi- 

 cherry, latitude 12*. 2 Gentil, p. 79. But in latitude 46°, Sir 

 George Schuckburgh heard thunder grumbling under him when 

 flanding on mount Saleve, an elevation of only 2831 feet over the 

 furface of the plains. Phil. Tranf. 1777, p. 527. Now as a 

 great part of the weight of the atmofphere refides in the inferior 

 and denfer ftrata, it is evident that it muft be more altered by the 

 perturbations that happen in them, than by thofe that happen in 

 the much loftier. Hurricanes alone afFed the lower ftrata, and 

 hence the barometer links confiderably. Thus, in the hurricane 

 that happened in the ifland of St. Bartholemew on the 2d of Auguft, 

 1792, the barometer fell from 30,18 to 28,03 during its continu- 

 ance, and perhaps ftill lower, for the obferver was obliged to quit 

 the houfe, whofe proftration he apprehended, when at its height. 

 See the circumftantial defcription in XI. Voights'g Phy. Magaz. 

 4 Stuck, p. 74. 



It 



