DEDUCTION FROM THE GASEOUS THEORY. OF SOLUTION. 291 



reach a t('iii])oratuio at which it becomes iiitiiiitely vohitile — a tem- 

 perature above wiiieli the liquid can iu>t possibly exist in the i^reseiu-e 

 of its own vai>or, no matter how great the pressure may be. At this 

 temperature. (Minilibrium of pressure between the liquid and its vapor 

 becomes impossibh^, and above this ])oiut the substance can exist only 

 as a gas. This is the critical temperature. And so it seems to me 

 that if we cany our analogy to its logical conclusion, we may expect 

 for ex'ery substance and its solvent a detinite tem})erature above which 

 equilibrium of osmotic ])ressure between undissolved and dissolved 

 substance is imi)ossible — a temperature above which the substance 

 can not exist in i)resence of its own solution, or in other words a 

 temperature of intinite vsolubility. This may be spoken of as the crit- 

 ical solution temperature. 



But a little consideration shows that in one particular we have been 

 somewhat inexact in the pursuance of our analogy, for we have com- 

 pared the solution of a solid body to the vaporization of a volatile liquid. 

 We can however do better than this, for volatile solid bodies are, 

 not Avanting. It is to these, then, that we must look in the iirst in- 

 stance. Xow, a volatile solid (such as (*am[)hor or iodine) will not 

 reach its critical point without having .Hrst melted at some lower tem- 

 perature, and a .similar change should be exhibited in the solution 

 process. At some delinite temperature, below that of infinite solubil- 

 ity, we may expect the solid to melt. This solution melting point will 

 iu>t be identical with, but lower than, the true melting point of the 

 solid, and for the following reason: Xo case is known, and probably 

 no case exists, of two li(fuids one of which dissolves in the other and yet 

 can not dissolve any of it in reliirn. Therefore there will b(> formed by 

 melting, not the j)ure liquid substance, but a. solution of tlie solvent in 

 the liquid substanc(\ Hence the actual nndting or freezing point 

 must be lower than the true one, in right of the laws of which I have 

 sp(>ken when discussing Haoult's nu'thods in the earlier part of this 

 address. 



From this solution melting-point ui)wards we shall then have to deal 

 with two liquid layers, each containing both substance A and solv<'nt 

 7>, but the one being mostly substance .1 and the other mostly solvent 

 B. These may be spoken of as the A layer and tlie B layer. As temper- 

 ature rises, the proportion of A will decrease in the A layer and increastr 

 in the /Mayer; ami every gram of A will occupy an increasing sola 

 tion volume in the A layer (7> being absorbed there) and a decreasing 

 solution volume in the B layer. At each temperature the osmotic pres- 

 sures of .1 in Uw two layers must be equal. The whole course of affairs, 

 as thus concei\ed, now admits of the closest comparison with the 

 changes wliich acciunpany gradual rise of temperature in the case of a 

 volatile liquid and its saturated vapor. Tlie liquid is like the substanc(i 

 A in the .1 layer; the vapor (which is the same matter in :inother 

 state) is like the same substance A iu the B layer. As temperature 



