THEORY OF SOLUTION AND CRYSTALLISATION. 2^9 



It seems as though they were capable of separating water 

 or at any rate a weaker solution from a stronger, so that 

 the result is equivalent to the removal of part of the 

 solvent. 



d. Crystallisation occurs when the supersaturated solution is 



cooled to a low temperature ; and 



e. When part of the solvent is got rid of by evaporation. 



/. Contact with oil or grease sometimes appears to cause 

 crystallisation, but the action is limited to one or two 

 cases and the explanation is obscure. 



g. Shaking or electrolysis fails to start crystallisation in super- 

 saturated solutions of sodium acetate, and sulphate. 



THEORETICAL. 



The natural inference from the facts known in connection with 

 the phenomena of solution is, that when a solid dissolves in a 

 liquid it first enters into chemical combination with a portion of 

 that liquid forming a chemical compound, the composition of 

 which under fixed circumstances is perfectly definite. Briefly 

 recapitulated the chief evidence in favor of this assumption is (i) 

 that when a solid comes into contact with a liquid capable of 

 dissolving it, heat is evolved j (2) that changes of colour are often 

 produced; and (3) that water or alcohol is retained by crystals 

 only in definite molecular proportions. We will now endeavour 

 to picture the state in which the substance subsists when solution 

 is complete. Much assistance will be obtained towards the 

 investigation of the problem] if we first briefly review the facts 

 which have been established in the analogous case of certain 

 vapours ; take for example pentachloride of phosphorus. The 

 vapour of this compound, a solid at ordinary temperatures, is not 

 homogeneous, it consists of a mixture of three vapours, viz., 

 pentachloride of phosphorus, PCI5, trichloride of phosphorus. 



