ON THE CHEMICAL COMPOUNDS CONTAINED IN ALLOCS. 



133 



Fig. 1. 



Cone 



per cents. ; the other ordinate is temperature. It is desirable to observe 

 not only the first halt in the cooling, but also any lower ones that occur 

 down to the moment of complete solidification, or even below it. It is 

 also desirable to isolate by filtration the crystals which form at the freezing 

 point, and to analyse them. This would give the composition of the 

 solid and liquid phases which could exist in equilibrium at the observed 

 temperature. Unfortunately, on account of experimental difficulties, 

 isolation of the solid phase has not been carried out in the case of alloys, 

 and a later microscopic study of the wholly solid alloy is a very imperfect 

 substitute for it. 



We now know pretty well the types of freezing-point or equilibrium 

 curves that occur. In the simplest of all cases — that in which the two 

 bodies a and b neither combine chemically nor form mixed crystals — the 

 •complete curve resembles fig. 1. It consists of two 

 branches cutting each other at the eutectic angle. 

 One branch, which starts from the freezing-point of 

 a liquid wholly composed of A, corresponds to the 

 formation of primary crystals of pure A at each 

 freezing-point, the other branch to the formation of 

 primary crystals of pure B. When the liquid, either 

 from its initial composition or through the separation 

 of the primary crystals, reaches the composition of 

 the eutectic intersection, a and b crystallise simul- 

 tajieously but in separate crystals. Thus the solid 

 eutectic alloy is a very minute conglomerate, while 

 all other alloys contain large primary crystals of 

 either a or b embedded in this conglomerate. This 



has been conclusively demonstrated by the exquisite microscopic work of 

 M. Osmond (®) and also by that of M. Charpy (^). 



Curves which approximate to this type have been worked (^° *"** ") 

 out for the pairs Zn.Al, Zn.Sn, Au.Cu, Ag.Cu, and some other 

 metals. Such curves do not indicate the existence of a compound, 

 though it would be too much to say that they disprove the existence of 

 compounds. 



When a compound exists whose melting-point lies in the region above 

 the freezing-point curve of the two metals, it produces a separate branch 

 cutting the other two branches. At points on this intermediate branch 

 the saturated liquid deposits crystals of the compound. The summit 

 of this branch occurs at the concentration corresponding to the formula 

 of the compound. If more than one compound exists there is a branch 

 for each compound, although parts of the branches may be lost by lying 

 below the curves of more stable bodies. 



While the above is the usually accepted view as to the meaning of 

 summits and eutectic angles in a freezing-point curve, two points may be 

 noted. The first is M. Le Chatelier's opinion as to the position of the 

 summit caused by a compound. He thinks ('^) that when a compound 

 partly dissociates on fusion, the summit caused by its presence may not 

 be exactly at the percentage composition corresponding to its formula, 

 and that the formation of mixed cry.stals may have a similar effect on the 

 curve. This is a matter needing further investigation. The other point 

 concerns the position of the eutectic angle. While it is well established 

 that the eutectic alloy is a conglomerate, not a compound, we should be 

 •wrong to ignore the fact that the angle often comes surprisingly near to 



