9S Mr. Herapath on the QMnntity of Vapour [Aute, 



same elasticity. Hence the vapour found by Mr. Dalton's 

 calculation always exceeds the quantity actually existing in the 

 air. 



Suppose T, be the tension just found, and p the barometric 

 pressure of the air, then /> — t, is evidently the elastic force due 

 to the dry air ; ;> — t, : r, is the ratio of the volumes of dry air 



and vapour; and - is the absolute volume of vapour in any 



unity of the compound mass of vapour and air, supposing the 

 vapour could exist as an air under the pressure p. And if t 

 denote the tension of vapour corresponding to the temperature 



of the air, the humidity of the air is -, absolute humidity being 1. 



Ify, /', be the Fahr. temperatures of the air and water, then 



the absolute volume of vapour by Mr. Dalton is ^' ..2 y and 



the humidity of the air - .t— rr^* The greater the difference 



between f and /„ that is, the less the vapour in the air, the 

 greater will be the proportional error of Mr. Dalton's method. 

 If/* = 60" and/, = 40°, the error will amount to more than 

 4 per cent. 



It has been justly obseived by M. Biot, that this method of 

 experimenting " may not have in ordinary hands all the sensi- 

 bility which its author is pleased to attribute to it." In such hands 

 as Mr. Dalton's, almost any method will succeed ; but certainly 

 the above process, elegant and simple as it is, requires no ordi- 

 nary care to insure success. I have often thought whether a 

 vertical glass cylinder open at the top only, and successively 

 dipped for short times in water or mercury of different tempera- 

 tures, might not afford more uniform results, since the cold air 

 in the cylinder being heavier than the air above, would keep its 

 contact with the glass better than the air cooled on the outside 

 of the glass vessel; but I have never brought this method to 

 trial. 



M. Biot has also objected to Mr. Dalton's method on the score 

 of its not been apphcable to small portions of gas ; but a very 

 trifling modification of this method, which no doubt has occurred 

 to its author, will easily enable us to apply it to any portion of 

 gas. For example, if a small glass vessel furnished with a stop- 

 cock or two, and filled with the gas at a given temperature / 

 and pressure p as ascertained by a manometer, be immersed in 

 a vessel of clear water which is gradually cooled down, the 

 incipient dulness of the immersed glass might be easily seen, 

 and thence the volume, &c. of the enclosed vapour be deter- 

 mined. Suppose the temperature of the water at this moment 

 }kj\ and the corresponding tension of vapour t, then 



