in the Electrical Resistance of Bismuth, 309 



wire, was no doubt also different. Hence although both these 

 wires present a point of minimum electrical resistance, the 

 temperature at which this occurs is not identically the same. 

 Another point of interest is that the temperature at which a 

 discontinuity was found by us in the thermo-electric line of 

 the pure bismuth, and which was found to be about —80° C, 

 is also approximately the point at which the resistance-tem- 

 perature coefficient of the same bismuth becomes sensibly 

 negative. It is therefore clear that at this temperature a 

 critical point is reached for this particular sample of bismuth, 

 and below this point the bismuth undergoes a physical change 

 which affects its thermo-electric value as well as its resistance- 

 temperature coefficient. These investigations indicate that 

 the electrical behaviour of bismuth in certain states at and 

 below particular temperatures is irregular, and that further 

 study is necessary to arrive at a complete explanation of the 

 facts. It has been pointed out by one of us (see Friday 

 Evening Discourse at the Royal Institution, by Professor 

 Dewar, " On the Scientific Uses of Liquid Air," Jan. 19, 

 1894) that the tensile strength of very many metals is greatly 

 increased by cooling them down to the temperature of liquid 

 oxygen, but that there is a very marked decrease in the tensile 

 strength of bismuth as well as of antimony and some other 

 metals when taken down to the same low temperature. It 

 is interesting to note that bismuth, which, at any rate in 

 certain states, presents the anomaly of a discontinuity in its 

 curve of thermo-electromotive force at a certain low tem- 

 perature, has also at about the same temperature a change of 

 sign of the temperature coefficient, and becomes in addition 

 exceedingly brittle below that temperature. 



In an interesting paper on the Mechanism of Electrical 

 Conduction (see Phil. Mag. ser. 5, 1894, vol. xxxviii, p. 57) 

 Dr. C. V. Burton makes reference to our former experiments 

 on the resistance of pure metals, and states and proves a 

 theorem to the effect that at the absolute zero of temperature 

 every substance must have either infinite specific resistance 

 or infinite conductivity. In the samples of bismuth A and B 

 we have before us a material w 7 hich exhibits, as far as we can 

 judge, a tendency in its specific resistance to continually 

 increase as it is cooled to temperatures lying below about 

 — 80°. In this respect it resembles a non-metal, and the 

 question arises whether this is the proper characteristic 

 behaviour of pure bismuth, or whether in any other physical 

 condition it would show the normal quality of pure metals in 

 decreasing its specific resistance regularly as it is cooled 

 towards the absolute zero. 



