i; 



REPORT — 1900. 



shall arrive at certainty in the interpretation of such curves until? this 

 question of mixed crystallisation has been thoroughly studied. M. Le 

 Chatelier some years ago pointed out the importance of this question and 

 studied it. 



The curve for mixtures of silver and gold has been quoted as a type 

 for bodies which form mixed crystals in all proportions. It is a con- 

 tinuous curve joining the 

 Fig. 6. 



l^oL^BlNn^Pl 



Na M 



Atomic percentage. 



points of fusion of the two 

 components, and diflfers from 

 such a curve as that of fig. 1 

 by consisting of one branch 

 only. Mixed crystallisation 

 of two bodies, one or both 

 of which were compound, 

 might be indicated by a 

 continuous curve joining the 

 freezing-points of the two 

 proximate constituents of 

 the crystals. There are 

 probably several such cases 

 in the curves given by M. 

 Gautier. Charpy and Stead 

 have independently studied 

 with the microscope a pe- 

 culiar crystal structure which 

 they conjecture to be due to 

 mixed crystallisation. But 

 no case of isomorphism in 

 alloys lias been worked out 

 in a manner that is con- 

 clusive. Until lately a 

 satisfactory theory of the 

 subject was lacking, but 

 Professor Roozeboom (i°), to 

 whom physical chemists owe 

 so much, has lately investi- 

 gated it, and MM. Van 

 Eyk, Reinders, and Hissink 

 have verified his views 

 in the case of certain mix- 

 tures of two salts. It 

 seems very desirable that 

 students of alloys should 

 begin to work in the light 

 of this theory. An at- 

 tempt will therefore be 



made here to state Roozeboom 's view in the simplest case he gives. 



For this purpose we must look again at fig. 1, the curve for a pair of 

 metals which neither combine chemically nor form mixed crystals. Here 

 the region above the curve corresponds to liquid states, the line of the 

 curve to equilibrium between a liquid and crystals of A for the left branch 

 and B on the right branch ; the region below the curve, but above the 

 angle, to mixtures of solid a or b with varying liquids ; and, finally, all the 



