463 EEPOET— 1886. 



Since contraction occurs when most solids are dissolved in water, it 

 follows that if the dissolved substance could be withdrawn again without 

 changing the temperature of the liquid expansion would occur. Now super- 

 fused substances (and supersaturated solutions) contract as they solidify, 

 with the single exception (so far as I know) of water, which expands when 

 frozen. If it were not for the declaration of Professor Osborne Reynolds 

 that ' dilatancy ' is not a property that can be exhibited by ordinary matter, 

 but only by such hypothetical stuflF as consists of hard, inelastic particles 

 devoid of cohesion and friction, one would be tempted to try and explain 

 superfusion and supersaturation by appeal to a hypothesis of that kind. 

 Certain it is that in a superfused liquid there exists a condition of strain 

 which is overcome by cohesion only when the latter has been considerably 

 increased by lowering the temperature. And relief from this strain may 

 often be obtained in more than one way, as in the crystallisation of 

 dimorphous substances like sulphur, and in the deposition of modified 

 salts (such as Na2S04 7H20, &c.) from solutions. 



The fact that in some cases it is possible to reduce the temperature 

 of small drojas of a superfused liquid, such as sulphur, much below that at 

 which larger masses can be preserved in the liquid state, seems to show 

 that the surface tension in sjDheroidal masses operates against the ten- 

 dency to change of state. Whether it is suflBcient appreciably to retard 

 the change is more than I can say, but Van der Mensbrugghe ^ has shown 

 by mathematical analysis that the potential energy of a free liquid surface 

 increases with the surface, and in this way he explains the very facts to 

 which I have just referred. 



Conclusion. 



Such facts as these lead us to the consideration of what is me&nt 

 by chemical combination. Take, for example, a common metallic salt, 

 such as copper sulphate, CUSO4. Here the law of definite proportions 

 being rigidly observed, and the compound being very difierent in 

 external characters from its ingredients, chemists have no difiiculty in 

 agreeing that this is a case of true chemical combination. But when the 

 salt combines with water of crystallisation great difference of opinion 

 arises as to whether the elements thus superadded are chemically com- 

 bined with those of the salt, or whether a new kind of chemical afBnity 

 is called into play. The difierence is one of degree, not of kind. Blue 

 vitriol is composed of copper sulphate and water united in definite and 

 quite simple proportions, and the chief reason for supposing the water to 

 be combined in a manner difierent fi'om the other ingredients of the salt 

 is that it can be detached by heat or otherwise more easily, and that it 

 occupies a relatively larger volume. But, as already stated, the water* 

 combined with such a salt is attached in difierent degrees of intimacy, 

 the first molecule occupying a smaller volume than the second, and so 

 on, the act of union of the salt with these successive molecules being 

 attended by the loss of successively smaller amounts of energy. 



Taking a step further, suppose this salt dissolved in water, the 

 resulting liquid is produced with many of the attendant phenomena 

 which usually accompany recognised chemical combinations — changes of 



' ' On the Application of Thermodynamics to the Study of the Variations of 

 Potential Energy of Liquid Surfaces,' Pldl. Mag., 1876 [5], ii. p. 450 ; and Fliil. Mag. 

 1877 [5], Iv. p. 40. 



