Resistance of certain Amxlcfims with Temperature. 453 



metals yet imcombined. If the latter takes place then, since 

 it was shown in the last section that there is an absorption of 

 heat at this temperature, something equivalent to solution in 

 the excess of mercury must also take place. After 36° the 

 temperature-coefficient is considerably less ; it is usually the 

 case that the temperature-coefficient of an alloy is less than 

 that of pure metals, and hence if zinc was set free at this 

 temperature we should expect the temperature-coefficient to 

 increase. That it decreases points to combination taking 

 place, unless the zinc-mercury compound dissociates into two 

 other compounds, about whose temperature-coefficients we 

 should, of course, know nothing. 



A similar phenomenon occurs at the temperature at which 

 the resistance suddenly decreases, but there is nothing here 

 to show whether combination or dissociation takes place ; 

 whatever it is, the resulting compound has a less specific 

 resistance than the one from which it was formed, and hence 

 the resistance falls. 



As the amalgam is cooled, the same compounds remain 

 until a temperature corresponding to that near R (fig. 3) is 

 reached, when dissociation occurs gradually, accompanied by 

 evolution of heat and increase in resistance. This dissocia- 

 tion, however, is not completed when the initial temperature 

 is reached, unless the substance is cooled very slowly. 



The compound so formed is unstable at ordinary tempera- 

 tures, or else a further gradual dissociation takes place ex- 

 tending over several weeks, until a stable condition is reached. 

 This slow alteration is accompanied by a fall in the resist- 

 ance, and the final result is a compound of zinc and mercury, 

 together with, probably, a mixture of zinc and mercury, the 

 whole being stable at the temperature of the room, but be- 

 coming unstable as a temperature of 36° is approached. 



If the amalgam is heated before this gradual change has 

 had time to take place, there is no instability at 36°, and 

 therefore no sudden change in resistance at that point, as is 

 seen from curve A (fig. 6). 



Since the conclusion of the experiments described above, 

 my attention has been called to a paper by C. E. Guillaume 

 on the Nickel-Steels*. In this paper M. Guillaume advances 

 a theory to explain the changes in the magnetic properties of 

 these alloys which were first noticed by Hopkinson f. On 

 account of the large amount of interest that has been shown 

 of late in the constitution of alloys, and also because the 



* Revue Generate des Sciences, April 1898, p. 282. 

 t Hopkinson, Proc. Roy. Soc. Dec. 12, 1889, Jau. 23, 1890, May 1, 

 1890. 



