Of BAROMETRICAL MEASUREMENTS. 129 

 moiflure be either increafed or diminifhed, the expreffion 

 p (i+ m (- -- r)J hg'-j- will no longer be equal to the true 



height, but muft be multiplied into i + in order that it may 

 be equal to z. No.w; this fraction + * reprefents the excefs or 

 defect of the moiflure diflolved in the air above or below its 

 mean quantity ; or, more exactly, it is proportional to the in- 

 creafe or diminution of the elafticity of the air arifing from 



/ H 4- h . \ . b . 



that caufe. When^(i+( - -- ^Jlog. is lefs than the 



true height, the fraction TT muft be affirmative, and indicates an 

 increafe of elafticity, and, confequently, of moifture in the air. 



The contrary happens when p (i+m( -- r)J log. is greatr 

 er than the true height. To determine v, fince * 



z 



/ H+b N . b 

 Or if the error, that is z P ^r-f?( r)) log. = <?, 



then TT = Tj-n : r > or it 



44. To apply the barometer, therefore, for the purpofes of 

 hygrometry, let there be two barometers fixed, the one at the 

 top, and the other at the bottom of a high tower, or hill of 

 moderate elevation, and let them be obferved at the fame in- 

 ftant, together with their correfponding thermometers. If the 

 difference of their heights, computed from thence, be equal 

 precifely to the true difference, then is the moifture difTolved in 

 the air no way different from its mean quantity; but if the 

 difference of the heights fo computed, be greater or lefs than the 

 truth, then *-, as above determined, will give the quantity 

 by which the actual moifture in the air is lefs or greater than 

 the mean quantity. The height at which the one barometer 

 fliould be placed above the other, ought not to be fo fmall that 



R the 



