802 EEPORT— 1898. 



throw additional light on these discrepancies by examining the increase of pressuro 

 produced by adding ether or some other volatile liquid to an atmosphere, the 

 pressure of which uad been previously measured. In the course of these experi- 

 ments, he concluded that he had satisfactorily ascertained the cause of the 

 phenomena to be the condensation of the vapours on the vertical sides of the- 

 containing tube under a pressure of vapour less than the saturation pressure 

 corresponding to the temperature. He regarded Dalton's law as being true in 

 fact for a gas-filled space completely surrounded by a sufficiently thick layer of the 

 evaporating liquid. In a vessel with solid vertical walls gravity prevents the 

 formation of a layer of the required thickness on the vertical sides. This hypo- 

 thesis was held to account (or the fact that if the mixture were compressed, as in 

 a Boyle's law tube, the pressure gradually diminished after the initial compression 

 10 a value less than that required by Dalton's law ; in the case of ether about 10 

 jnm. of pressure were missing on this account. 



Such an hypothesis would give a satisfactory explanation of the phenomena if 

 the sides of the glass vessel were gradually dissolved away by the liquid condensed 

 on the sides, as, for example, would be the case if the tube were made of rock salt : 

 hut if the action between the glass sides and the vapour were merely a mechanical 

 one, the supposed continuous evaporation of liquid from the horizontal surface and 

 its return to the stock by condensation on the vertical sides and running down 

 them would seem to contradict the principle of conservation of energy. 



Regnault's explanation is the more difficult to accept if the density of the 

 vapour be considered. Professor J. J. Thomson, in his ' Applicarion of Dynamics 

 to Physics and Chemistry,' has shown that the addition ot air-pressure upon a 

 liquid surface in contact with its saturatf-d vapour ought to result in a slight 

 additional evaporation causing a slightly increased density of vapour, and the 

 experiments of the Author upon air at various degrees of saturation ^ indicated 

 experimentally some slight increase of density beyond that required to correspond 

 to the vacuum pressure, so that any apparent departure from Dalton's law could 

 not be attributed to thp want of mass of vapour in saturated air. 



The point might hi- investigated by ascertaining experimentally the behaviour 

 of a mixture of air and vapour under isothermal conditions in the neighbourhood of 

 the point at which condensation begins to occur. At great rarefaction the mix- 

 ture would behave as a perfect gas, and if the isothermal relation between p and 

 1/y were plotted on a diagram (with a suitable correction if necessary for the 

 deviation of the air frouj Boyle's law) the curve obtained would be a straigrht line 

 through the origin represented by p = k x Ijv. At great concentration if Dalton's 

 law were strictly true the relation would again be represented by a straight line, 

 viz. : — 



p = k' xllv+p,._ 



■where k' is the Boyle's law constant for the contained dry air, and />,, is the 

 saturation pressure ol' the vapour. This line would cut the vertical line ]/i' = 0, in 

 the point distant p^ (i.e the theoretical saturation pressure) from the origin, 



and it would cut the line p = 0'm the point Ijv = — ^,- These two theoretical 



fC 



lines would meet if produced at an obtuse angle in the theoretical saturation 

 point. 



If, on the other hand, at the greater concentration there were a discrepancy from 

 Dalton's law proportional to the total pressure and therefore represented by the 

 relation, 



p = k' X Ijv+Pf —^Pi 



the observations should still show a straight line not quite coinciding in direction 

 with the theoretical line for Dalton's law, but cutting that line in the point on the 



horizontal axis! -?", O) and cutting the vertical axis at a point distant (1— X)/), 



from the origin. 



' Phil. Trans., vol. clxsix. (1888), 'Report on Hygrometric Methods.' 



