Freezing- Point Carves of Binary Alloys. 161 



metals, one of the two being in each case either silver or copper. It 

 is an extension into temperatures as high as 1100 C., of experiments 

 similar to those at lower temperatures with which we have been 

 occupied for the last seven years. The results of our previous 

 experiments, in which mercury thermometers were used, are pub- 

 lished in the ' Journal of the Chemical Society.' In the work 

 described in this paper the determinations of temperature were made 

 by means of platinum, electrical resistance pyrometers of the 

 Callendar- Griffiths type. 



The paper is divided into four sections. 



Section I contains a short survey of certain points in the theory 

 of concentrated solutions which bear on the interpretation of the 

 experiments. 



Section II is devoted to an account of the experimental method. 



Section III contains the results of the experiments in a tabular 

 form, each table being followed by notes and remarks taken from the 

 experimental note books. 



Section IV contains the results expressed graphically as complete 

 freezing-point curves, together with a discussion and a statement of 

 the conclusions that can be arrived at from a studv of each curve. 



Section I. 



If we plot the percentage composition of an alloy horizontally, and 

 the freezing point vertically we get the freezing-point curve. This, 

 for a pair of metals, would consist of two branches, each starting 

 from the freezing point of a pure metal, and descending until they 

 meet in the eutectic point. Our silver-copper curve gives a fair idea 

 of this case. 



If the metals A and B form a stable compound C, then the theory 

 as developed by Bakhuis, Booseboom, and by Le Chatelier makes it 

 probable that the curve will be divided into the systems A C and 

 C B with two eutectic points, and an intermediate summit at C. This 

 case is well illustrated by a complete freezing-point curve of copper- 

 antimony by Professor Le Chatelier, in which two such summits 

 occur. 



Another not infrequent case is probably that of a compound, which 

 when molten can only exist in a partially dissociated condition. Our 

 silver-antimony curve resembles such a curve. Other points of 

 Section I will be best deferred to the summary of Section IY. 



Section II. 



The alloys, weighing from 200 to 500 grams, were melted in plum- 

 bago (salamander) crucibles, placed in one of Fletcher's blast furnaces. 



