98 EEPORT— 1891. 



below it. It should be possible, according to the hypothesis, 

 to cool mixtures of ether and alcohol sufficiently to cause 

 separation into two layers, similar to those observed at the 

 ordinary temperature in the case of ether and water ; but I 

 do not know that this has yet been put to the test of experi- 

 ment. 



Alexeeff's experiments appear to me to be of the very 

 highest importance, and to merit the closest attention in any 

 inquiry into the nature of solution. As already stated, they 

 afford the strongest support to the hypothesis which I have 

 been discussing : indeed, had it not been for this support I 

 should hardly have ventured to discuss it at all. They refer to 

 solutions in water, below and above 100°, of phenol, salicylic 

 acid, benzoic acid, phenylate, and aniline, and to solutions in 

 molten sulphur of chlorobenzene, benzene, toluene, aniline, 

 and mustard-oil. All these afford instances of reciprocal 

 partial solution throughout a considerable range of tempera- 

 ture, leading eventually at a definite temperature to infinite 

 solubility. Several of them afford instances also of solid sub- 

 stances with solution melting-points below their true melting- 

 points. 



Alexeeff experimentally determined the temperatures at 

 which different mixtures of the same two liquids are just 

 converted into clear solutions ; or, in other words, he ascer- 

 tained the strengths of the saturated solutions corresponding 

 to different temperatures. For each pair of liquids he found 

 that when a particular strength of mixture is reached the 

 temperature of saturation is lowered by further addition of 

 either ingredient. Thus, a mixture of about 37 parts of aniline 

 to 63 of water requires a temperature of 164*5° to convert it 

 into a homogeneous solution, but one of 21 of aniline to 79 of 

 water assumes this condition at 156°, while one of 74 of aniline 

 to 56 of water does so at 157-5°. He plotted his results in the 

 form of curves, with temperature and percentage strength as 

 the two co-ordinates. The curve for aniline and water is 

 shown in Fig. 1, PI. IV. ; and this may be taken as a fair repre- 

 sentative, the general form of all being similar. It is at once 

 apparent that for every temperature up to a certain limit there 

 are two possible saturated solutions — one of water in aniline, 

 and one of aniline in water. The limiting temperature at 

 which there is but one possible saturated solution, and above 

 which saturation becomes impossible, is called by Alexeeff 

 the MiscJmngs Temperatur. It is what I have called the 

 critical solution-temperature. It is, in the case of aniline 

 and water, about 167°, as nearly as one can judge from 

 the curve without a greater number of experimental points 

 in this portion of it than Alexeeff gives us ; and the cor- 

 responding saturation - strength is about 50 per cent. It 



