296 On the Elastic Force of the Vapor of Mercury, 



The form of the expression just given, leaving out of considera- 

 tion the determination made of B, is arbitrary ; it has been found ap- 

 plicable to the tension of watery vapor, and M. Roche endeavors to 

 derive it from an examination of the forces which may be supposed 

 to act, in the vaporization of liquids* Without discussing the reason- 

 ing, I purpose to test this formula by my experiments on the vapor of 

 mercury. The boiling point of this liquid being at 680° Fahr. 

 tt = 680-f-448~1128.* My observation of the tension at 500° or 

 180° below the boiling point of mercury, or t= —180°, gives A 



3.976* 

 3.976 and the formula becomesf log. i=jT 2 g i / - If this formula 



were applicable to the observations, it should when thus deduced 

 from the medium temperature, give very nearly the result obtained 

 by observation for the two extremes, or 446° Fahr. and 554° Fahr. ; 

 for the first it gives e=0.091 atmospheres, and for the second, e 

 0.316 atmospheres. The coefficient of this formula, therefore, when 

 determined by an observation at a medium temperature, gives results 

 which are too high at low temperatures, and too low towards the 

 upper limit of the series ; in other words the rate of increase of 

 elasticity, for an increase of temperature, is less than that shown by 

 my observations. The formula would vary still more from the truth 

 if the air thermometer were referred to as a standard. The failure 

 of this formula is unfavorable to the theoretical views of M. Roche, 

 and it seems probable that the expression has no special advantage 

 over other empirical formulae which have a single constant to be de- 

 duced from experiment. 



I have further shown in my memoir that the form of the function 

 which I formerly supposed, from theoretical views, to express the 

 laws of the tension of the vapor of water, J namely, log, e 



c(vt+b 2 -&,) in which e and t represent the same quantities as in 

 the formula last given, and a and b are two constants, to be deter- 

 mined by experiment, does not apply to the tension of the vapor of 

 mercury. 



> 



n = 360+266.67 = 626.67, for the Centigrade scale. 

 . T 3.976* 



t Log. e= — . — 



^ 626.67-f-t' 



{ Pavia Philosophical Journal, 1819 



