446 Mr. R. S. Willows on the Variation of the 



The amalgam cooled regularly to about 93°, when its tem- 

 perature suddenly rose to 95°, and remained steady for a 

 short time at that point. It then cooled regularly, but more 

 slowly than previously, until it reached the neighbourhood of 

 75°, after which it took a still slower rate of cooling. The 

 reheating at 93° is due to surfusion, and may be got rid of 

 by the usual device of dropping in nuclei. Between 95° and 

 75° there is a gradual separation of solid from the liquid 

 mass, and at 75° this is completed. The critical point near 

 75° is ill-defined. Reference to Table III. shows that these 

 are approximately the temperatures between which the re- 

 sistance varies most rapidly. The curve is very like one 

 given by Roberts-Austen ('Nature,' Oct. 1898, p. 619) for 

 tin metal. 



Above 95° the amalgam is fluid, below 75° it is solid, while 

 between these temperatures it is a mixture of crystals of tin 

 or a compound of tin and mercury in a fluid containing a 

 less percentage of tin. When the amalgam is heated, it 

 is very probable that these changes taking place in the 

 inverse order give rise to the rapid increase in resistance 

 which occurs at these temperatures. The crystals of tin are 

 gradually dissolved iu the mass surrounding them, which 

 contains less tin, and which therefore is fluid at a lower 

 temperature, until when the liquid condition is reached this 

 tin is uniformly diffused throughout. While this solution is 

 proceeding the resistance increases most rapidly. 



That something of this kind takes place in the case of alloys 

 has been shown by various experimenters. Thus Osmond *, 

 from a microscopic examination of steel, concluded that if it 

 is raised to a molten condition and then cooled, we have a 

 separation of crystals of either nearly pure iron or Fe 3 C, 

 according to the percentage of carbon present, and then, when 

 the carbide has reached a certain concentration by the sepa- 

 ration of the constituent in excess, iron and carbide are 

 deposited simultaneously, giving rise to a laminated structure. 



Heycock and Neville f operated more directly by taking 

 photographs of a thin plate of the alloy, to be tested by means 

 of Rontgen rays. The varying opacity of the different metals 

 to these rays enabled them to distinguish crystals of metal 

 surrounded by the eutectic alloy. JS either of these methods 

 is applicable to the tin amalgams mentioned here, for they 

 are too granular and not sufficiently hard to admit of being 

 polished for microscopic examination, and the same property, 

 together with the opaqueness of mercury to Rontgen rays, 



* Comptes Rendus, vol. cxix. p. 329, vol. exxi. p. 684. 

 t Proc. Camb. Phil. Soc. 1898, p. 417. 



