ON SOLUBILITY. 433 



solutions of organic compounds sometimes separate Into (wo layers of 

 -jogg different concentration; these cases formed the subject of a 



theoretical discussion by Roozeboom . 141 

 1890. Walker 1 ™ deduced a relation between the solubility of a 



substance in any solvent and its heat of fusion. 



A comparison of the graphs which represent the solubility 

 ' of naphthalene, triphenylmethane, diphenylamine, and phthalic 

 anhydride in carbon bisulphide, in hexane, and in chloroform enabled 

 fitard 200 to conclude that the lower the melting-point of the solvent 

 the greater the solubility of a substance therein at any common 

 temperature. His measurements of the solubility of naphthalene in 

 carbon bisulphide at temperatures approaching the melting-point of this 

 solvent show that at that temperature ( - 115°) the solubility will 

 become zero. He showed the melting-point of the solvent was the 

 inferior limit of solubility. 



Winkler 190 traced an empirical connection between the diminution 

 of the absorption coefficient of a gas brought about by a rise of tempera- 

 1894 * ure anc ^ * ne corres P on( ling decrease in the viscosity coefficient of 

 the solvent Thorp and Rodger 216 criticised this conclusion, and 

 from their discussion of the subject it would appear that for the same 

 gas the diminution in the absorption coefficient for any temperature 

 interval is approximately proportional to the corresponding diminution 

 in the viscosity coefficient of the solvent. 



As indicating the part played by the solvent in conditioning the 



ft dissolution of a substance, mention must be made of Pictet's 230 



observations on the volatility of otherwise non-volatile substances 



when dissolved in volatile liquids. At ordinary pressures borneol is not 



volatile in ether vapour, but when heated with ether to a temperature 



below its melting-point it was found to be completely volatile. 



This and other similar results obtained with borneol and ethyl 

 chloride, and with ether as solvent for guaiacol, iodine, or phenol, 

 appear to support the contention of Hannay and Hogarth that solids 

 dissolve in vapours. 



III. B. — Chemical Nature of Solvent. 



Kremers 17 made numerous solubility measurements with the 

 • object of finding some existing relationship between the solubility, 

 nature of solvent, and the nature of the solute, but no statement of 

 general application was possible. 



The absorption coefficient of carbon dioxide in pure sulphuric 

 " dcid was found by Setschenoff G5 to be the same as in water, but the 

 addition of water caused this value to decrease rapidly to a minimum. 



Bunsen * showed, in the case of alcohol and water, there was 

 apparently no relationship between the solubility of a gas and the 

 nature of the solvent, a conclusion confirmed by Gniewosz and 

 im7 - Walfisz. 130 



Woukolofj li7 adversely criticised the conclusions arrived at 

 1.-89. ky Louguinine, Khanikoff, and Wroblewski, namely, that the 



* Vide Section IV. B, 1855. 



