SOME SUGGESTIONS REGARUING SOLUTIONS. 301 



zatioii of the unstable i)orti(>u oi' tlieenrve .1 /»'. In the j;iis-li(iui(l curve 

 pressure falls with formation of a shower of drops; iu the solvent-sol- 

 vent eurve crystallization euvsues and the solvents separate. The 

 plienonu'na are liowever not completely ana]oi>'ous; the comi)let<' anal- 

 o<iy wouhl be if the tempeiature were so low that the substance in the 

 li(juid-j4as eoui)]e were to sepaiate iu the solid, not in the li(piid, state. 

 This, so far as I am aware, has not been (experimentally realized, but 

 one sees no reason why it should not l)e possible. 



I hav«! sonn> hesitation in offering;' speculations as to the state of 

 matter at the ]>ortion of the continuous curve D E. It may be that it 

 corresponds to a syrui)y or viscous state. Cane su^ar at the moderate 

 temperature dissolves water; indeed it is possible to obtain a solution 

 of 1 per cent of water in molten cane sugar. And such a solution, if 

 quickly cooled, renuiins a syrup. But it can be induced to crystallize 

 by the presence of crystals. Thus, in such a mixture of sugar and 

 Avater a few grains of crystalline sugar cause the whole mass to crystal- 

 lize, and water saturated with sugar, and sugar, separate into two layers. 

 Here agaiis a complete analogy fails us, for it is a soli<l which separates. 

 As Me know nothing of tlie osmotic [)ressure of a syrup, the analogy is 

 a defecti\-e one: but it is piobable that a dilute solution of sugar would 

 pass continuously into a syrup of pure sugar by evaporation of the sol- 

 Acnt, and analogy would lead to i\w supi)osition that the syrup coin- 

 cid(^s with the unstable state of the liquid. I would therefore offer the 

 analogy between the syrujtpy and the super-cooled states as a tentative 

 one; it lacks foundation in both cases. 



()nei)oint icmains to Ixe mentioned. I have for the past nine months, 

 in eon junction with Mr. Ivdgar I'erman, lieen deteiininingtheadiabatic 

 relations for licpiid and gaseous ethei-: the lise of i)ressure and tempera- 

 tuie when volnme is decreased without tins (\scai)e of heat. Ft is ob- 

 vious that similar relations ai'c det ei-minable for solutions, and probably 

 with mu(;li gr(eater facility. IM. Alexi'clf has made sonue nu'asurements 

 which might be utilized for this purpose, but they are far too few in 

 number, and moreover, the necessary data as regards osmotic pressure 

 are wholly wanting, it would be ]K)ssi])le. by a series of dift'erent ex- 

 periments, to ascertain the evolntion (»f heat on increasing concentra- 

 tion, and so to arrive at a knowledge of the specilic heats of the solu- 

 tion at constant osmotic pressure. coires])onding to the idea of specilic 

 heats at constant pressure; and also of specific heats at ccmstant concen- 

 tration, corresponding to specilic heats at constant volume. I do m)t 

 know whether such researches would yield as accurate results as those 

 we are at present carrying out, but they are at least well worthy of at- 

 tention. 



