1896.] 



on Electric Research at Low Temperatures. 



249 



81,500 C.G.S. uuits per cubic centimetre at 0° C, but on cooling in 

 liquid air and becoming magnetic it is found to have decreased to 

 about 47,200 C.G.S. units when taken at 0° C. A very pretty way 

 of showing this difference in resistivity is to dip one half of a wire of 

 the 19 per cent, nickel-steel in liquid air, and then take it out, and 

 pass a strong electric current through the wire. The current raises 

 the half which has not been dipped into liquid air to a red heat 

 before the other half is visibly red hot. 



It is, perhaps, more correct to say that this alloy can exist in an 

 infinity of different physical states, because we have found that the 

 lower it is cooled in temperature the lower its resistivity can be made 



-200- -150- -100- 



-50^ -0^ 



Fig. 7 



+ 50^ +100= +150- 



Curve showing the Variation of resistivity of nickel-steel (19-64 percent, nickel) 



when taken through a cycle of temperature from + 150° to -200° 



and back again. 



to be when measured again at ordinary temperatures. On heating up 

 the alloy again to a bright red heat it goes back into the non-magnetic 

 ductile state. 



The chart (Fig. 7) before you shows the manner in which the 

 electrical resistance varies between the limits of —200° C. and 150° C. 

 when the alloy is taken through a cycle of temperature beginning at 

 150° C. in its non-magnetic state. 



The 29 per cent, nickel-steel exhibits the same characteristics in 

 a less marked degree. A close study of this interesting material 

 shows that there is room for much valuable work upon it yet. 



A manganese-steel, brought to notice by Mr. E. A. Hadfield, 

 having about 12 per cent, of manganese in it, is also capable of exist- 

 ing in two states, a magnetic and a practically non-magnetic variety. 



