Determining the value of v, either by calculation or by means of 
3 
. . . t “Te ~ 
a diagram, we find that, if the proportion — differs much from 
a 
Ti. (07 0 : es 
Es On = (5) = 4.8!) (ef. Suppl. N°. 6, p. 14; Proc. June ’03, 
pe \OT Jer ON 
p. 121) owing to the particular shape of the isothermals near the 
critical point, the difference v’—v, is much larger (of lower order, viz. 
Co 
if ; : a 
3) than the correction term © ye. For that reason and also because 
of the uncertainty about the volumes which belong to a definite 
pressure, again owing to the shape of the isothermals, we need not 
distinguish between ¢ and v’, in other words, we may neglect 
the correction term v ya’). As, however, we intend to determine 
the density of the mixture, we must bear in mind that v is the 
volume occupied by JZ, (1—.«a) + M, x gr, JZ, and M, representing 
the molecular weights of the pure substance and the admixture. Thence 
M, (1—2z) + M‚ z 
follows the density ———— ——, for which, for the same reason 
5 
i 
4 . . . . 
as above, we may put —, i.e. the density of the pure substance itselt 
5 
at the temperature 7” and the pressure p’. 
On the strength of this consideration we may conclude that the 
densities of the two mixtures x, and z,, at a temperature which is 
about the critical temperature of the pure substance 7, may be read 
approximately on a p, @ diagram of the isothermals of the pure 
substance; on the isothermals of the temperatures 7’ = 7; (1—e.r,) 
and 7,’ = Trax) we seek two points for which the pressures 
are p, = pr (ABe) and p,’ = pe (1— 2.) respectively. 
Besides these two layers, however, the tube really contains still 
several others of different composition, because the composition varies 
gradually *). If for different compositions we determine the densities 
1) Further on we shall see that, in the cases known thus far, this condition is. 
satisfied. 
2) The circumstance that we must determine the difference between the v’s for 
two mixtures, does not alter this conclusion in the least. For also the difference 
between v,’ and v,’ is found to be of a lower order i than the first. 
3) What has been said here about accidental impurities, holds also for the 
experiments of Carrerer and Coranpeau (C, R. 108, 1280, 1889) where jodine, 
which had been dissolved in liquid carbon dioxide, was not diffused equally through 
the tube at the temperature at which the meniscus disappeared; it also holds for 
similar experiments of Hagenpacu (Drude’s Ann., 5, 276, 1901), who dissolved 
