1896.] on Electric Research at Low Temperatures. 255 



temperatures. We found some very extraordinary results. Although 

 sensibly agreeing in resistivity at ordinary temperatures, in two cases 

 (see Fig. 10) the resistance curves had a minimum j)oint, and after 

 reaching this at about — 80^ tended upwards again; thus showing 

 that the resistance was increasing as the metal was further cooled. 

 These curves could be repeated as often as necessary with these 

 samples. Another specimen gave a curve with a double bend (see 

 Fig. 10). These results convinced us that it would be necessary to 

 prepare bismuth electrolytically, and with the assistance of Messrs. 

 Hartmann and Braun, of Frankfort, who have made a special study 

 of the preparation of electrolytic bismuth, we were provided with a 

 quantity of the metal which examination showed to be chemically 

 pure. On taking the resistance curve of a sam23le of this electro- 

 lytic bismuth when pressed into uniform wire under great pressure, 

 we found that its behaviour was perfectly normal, and that the resist- 

 ance line tended downwards, as in the case of all other pure metals, to 

 the absolute zero. Also we found that the specific resistance of 

 this last is very much less than that of the chemically prepared 

 samples, and less even than that employed by Matthiessen. Hence 

 pure bismuth is no exception to the law enunciated above. Bismuth 

 is characterised especially by many peculiarities. It has been known 

 for some time that the resistance of a bismuth wire is increased when 

 it is placed in a magnetic field, so that the lines of the field are 

 perpendicular to the direction of the current flow. This is easily 

 shown by means of one of Hartmann and Braun's spirals, manu- 

 factured now purposely for measuring magnetic fields. 



We have, however, discovered that if bismuth is cooled to the 

 temperature of liquid air the effect of any given magnetic field in 

 changing its resistance is increased many times. Thus, for example : 

 A certain bismuth wire we used had a resistance of 1 • 690 ohms at 

 20° C. Placed in a magnetic field of strength 2750 C.G.S. imits so 

 that the wire was transverse to the direction of the field, its resist- 

 ance was increased to 1*792 ohms, or by six per cent. The wire was 

 then cooled in liquid air and its resistance lowered to • 572 ohms. 

 On putting it then into the magnetic field of strength 2750 C.G.S. 

 units its resistance became 2*68 ohms. Hence it had increased 368 

 per cent. This magnetic field can thus actually reverse the effect of 

 the cooling, and cause the bismuth, when cooled and magnetised, to 

 have a greater resistance than when at ordinary temperatures and 

 unmagnetised. We are at present engaged in further unravelling the 

 problems presented by this new discovery with regard to bismuth.* 

 It is certainly very startling to find that a magnetic field which in- 

 creases the resistance only 5 per cent, at ordinary temperatures increases 

 it five times at — 186° C. We have recently discovered a similar, 



* Since the delivery of this discourse we have been able, by the employ- 

 ment of a powerful electro-magnet kindly lent to us by Sir David Salomons, to 

 increase the resistance of bismuth, when cooled in liquid air, more than 150 times, 

 by magnetising it transversely in a field of 22,000 C.G.S. units. 



