254 



Professor Fleming 



[June 5, 



shown in the case of mercury. A glass tube a metre in length was 

 formed into a spiral coil and filled with pure mercury, suitable con- 

 nections being provided at the ends. This coil was imbedded in a 

 mass of paraffin wax, and a platinum wire thermometer placed in 

 contact with it. The whole mass was then reduced to the tempera- 

 ture of liquid air, and observations taken of the resistance of the 

 mercury as it heated slowly up after being removed from the liquid 

 air. The curve in Fig. 9 shows the manner in which the resistance 

 increases with great suddenness between —41° and —36° as the metal 

 passes into the liquid condition. The resistance becomes four times 

 greater between — 50° and — 36° in the course of 14° rise of 

 temperature, and whilst in the act of passing through the melting 



-200 



100^ 



-100 O" 



Fig. 10. 

 Eesistivity curves of bismutli in terms of platinum temperature. 



point of the mercury at — 38° '8 C. This chart shows that the resist- 

 ance curve of the mercury in the solid state tends downwards, so as 

 to indicate that its resistivity would completely vanish exactly at the 

 absolute zero of temperature. It is interesting to note also that the 

 portion of the resistance curve belonging to mercury in the solid 

 state is sensibly parallel to that portion of it in the liquid state. 



We carried on a long struggle with bismuth in the endeavour to 

 unravel some of the electrical peculiarities of that metal at low tem- 

 peratures. Chemists are aware of the extreme difficulty of preparing 

 bismuth in a state of perfect chemical purity by purely chemical 

 means. From several different sources we procured bismuth which 

 had been carefully prepared by the reduction of the oxychloride or 

 nitrate after careful re-precipitations. This bismuth was then pressed 

 into wire, and its resistance curves taken down to the lowest attainable 



