136 



REPORT — 1900. 



are especially interesting, as they were the first high-temperature curves 

 of this kind with intermediate summits due to compounds that had been 

 published, and the paper ('■•) in which they appeared marks a new 

 departure in the subject of alloys. 



Aluminium- Antimony (fig. 4). — The maximum point is very much 

 higher than that of either metal ; it is at 1048° C. with 14'66 per cent, of 

 aluminium. The curve is remarkable because it shows that nearly all 

 mixtures melt above the melting-point of either component. 



Nickel-Tin (fig. 4). 



-Here the intermediate summit is at Ni3Sn2 



M. Gautier gives other 

 Fio. 4. curves of great interest, some 



of which were simultaneously 

 studied by Heycock and Ne- 

 ville (^'), but many of them 

 need further work before they 

 can be safely interpreted. 



Perhaps the freezing-point 

 curves investigated by M. 

 Kurnakov throw as clear a 

 light on intermetallic com- 

 pounds as any work that has 

 been done. This work can be 

 readily followed from the two 

 diagrams contained in his paper, 

 and reproduced in figs. 5 and 6. 

 The composition is expressed 

 in atomic percentages. The 

 first diagram (fig. 5) gives the 

 freezing-point curves of amal- 

 gams of sodium and potassium. 

 In the HgNa curve at least six 

 separate branches can be seen, 

 each corresponding to the crys- 

 tallisation of a difierent solid. 

 It would be premature to as- 

 sert that each branch proves 

 the existence of a correspond- 

 ing chemical compound, but 

 there can be no doubt that the 

 summit G proves the existence 

 of the compound NaHg,, a 

 body whose melting-point is 

 much higher than that of either 

 component. Similarly, in the 

 curve for potassium amalgams, besides minor branches there is a well- 

 marked summit at the formula KHga- 



The other figure (fig. 6) gives the freezing-point curves of the mixtures 

 NaBi, NaPb, NaCd. It is unfortunate that so few alloys were examined 

 on the upper part of the NaBi curve, but the existence of the freezing- 

 point M above 700° C. makes a compound very probable. M. Kurnakov 

 seems to have no doubt that it indicates Na-jBi, a body already prepared 

 by Joannis and by Lebeau. There are two summits on the NaPb curve — 

 one at p, which must be very near the formula Na2Pb and one at p' to 

 which he does not at present assign a formula. In the NaCd curve we 



0% 10 20 30 40 SO 60 70 60 90 100 



