274 REPORT— 1901. 



In working backwards it must be borne in mind that a diagram such 

 as fig. 3 is not alone sufficient to give complete information about a 

 reciprocal salt pair. Only two of the three values required can be deduced 

 from it ; to obtain the third, either the model must be used, or a table 

 showing the composition of the various saturated solutions, such as that 

 on page 270, must be referred to. 



Thus, assuming the composition of a solution to be that represented at 

 M in diagram, fig. 3, it is obvious that the 'plane' co-ordinate values are 

 21MgCl2 + 1 4K2CI2. On reference to the model or table it is seen, however, 

 that solution M contains, when saturated, 57 molecules of dissolved 

 salt ; therefore the number of other molecules present is 57 — (21 + 14) =22. 

 But it is to be remembered that these consist of two reciprocal salts, and 

 that in constructing the diagram one member of the pair was equated 

 against the other, so that only half the 22 molecules in solution are to be 

 regarded as present as sulphate — in this case MgS04 — and the remaining 

 11 molecules are considered to be molecules of KoClj, and are added to 

 the number of molecules read off from the diagram. The constitution of 

 the solution at m is therefore : 



21MgCl2 + (14 + ll)K2Cl2 + llMgS04. 



Before passing to the consideration of the diagram thus constructed, it 

 is necessary to realise that the points of equilibrium situated on the 

 margin are not all of the same order of stability. In cases in which 

 double salts are formed, the deposition of the double salt necessarily 

 follows, but never attends, that of the less soluble constituent. That this 

 should be the case is obvious when it is borne in mind that, as water is 

 removed, the more soluble constituent— the action of which is more or less 

 impeded by the water — is able to combine with the less soluble to form a 

 further quantity of double salt. The same argument applies to hydrates : 

 as water is removed from the solution the other salts present gradually 

 assert a dehydrating effect. 



The points F, H, l on the diagram are cases of this kind, and therefore 

 they are united by dotted instead of by full lines to the appropriate points 

 within the diagram. In indicating the direction in which crystallisation 

 proceeds arrows are therefore drawn through, and not towards, these points. 



To illustrate the way in which the diagram is read several cases may 

 be taken. 



At B the solution contains only Potassium sulphate. At a point on 

 B E a little to the right of b there is a small amount of chloride present ; on 

 evaporating such a solution change proceeds along the line B E, Potassium 

 sulphate alone separating until the point E is reached, when Potassium 

 chloride will also be deposited. The solution will then dry up without 

 changing its composition. 



Similarly, starting from a point x a little to the right of B, but a little 

 above b e and within the Potassium sulphate field, the track followed will 

 be along a line b s produced, until e m is reached, which then becomes 

 the track. 



It may not be superfluous to add that the track followed from any 

 point X within the diagram is always along a line drawn through x from 

 the point at which the field is saturated with its characteristic salt. 



represented by narrow strips of cardboard of Ibe required length ; and the vertical 

 -^rdinates of the angular points M to R are represented by strips of cardboard fixed 

 the base of the model 



