Resistance of Metals and Alloys. 291 



Two other coils (a) and (b) of the same wire were also 

 prepared, one of about 2600 centim. in length, and the other 

 about 3700 centim., so as to obtain a mnch greater resistance 

 to measure ; and these coils had their resistances taken in liquid 

 oxygen boiling under normal and under reduced pressures. 

 Hence lower temperature readings were obtained as follows : — 



Iron H. W. (a) 



. Iron H. W. (6) 



i 



B*. 



P- 



pt 



-Rt. 



pt. 



pt. 



7-091 

 4-154 

 3-839 



1220 

 725 



660 



-197-1 



-219 



-222-2 



9-875 

 5-257 

 5-222 



1220 

 649 

 644 



-197-1 

 -222-4 

 -222-7 



Volume-specific resistance at 0° C. . = 9065 C.Gr.S units. 

 Mean temperature-coefficient between 



0° C. and 100° C = -00625. 



It will be seen therefore that Iron H. W. has a much 

 greater conductivity (about 15 per cent.) than Iron A 

 and also a much larger temperature-coefficient. There is 

 reason to believe that Iron H. W. is exceedingly pure. 

 Both these iron lines are delineated on the chart. It will be 

 seen that at the lowest temperature reached, viz. — 222'7pt, 

 the iron has only one thirty-third of its resistance at about 

 -f-200° pt; in other words, we have destroyed 97 per cent, 

 of its resistance. The iron resistance-curves are strongly 

 concave upwards ; that is, the second differential of resistance 

 with respect to temperature is a positive quantity. 



§14. 



VIII. NicJcel 



The only perfectly pure nickel we have been able to obtain 

 is that prepared from nickel carbonyl. Mr. Mond was so good 

 as to prepare for us two pure nickel tubes formed by passing 

 the vapour of nickel carbonyl through a heated glass tube. 

 An attempt was made to draw this nickel down into wire, 

 but it could not be done. We accordingly cut up a portion 

 of a nickel tube into a very fine spiral in the lathe, and em- 

 ployed this spiral of nickel as a resistance-coil. 



The dimensions of this strip were not sufficiently regular 

 to allow the specific resistance to be determined. Pending 

 the preparation of other pure nickel in the form of wire, we 

 have taken the ratios of the resistances of this nickel spiral 

 at various temperatures and converted them into specific 

 resistances by the help of Matthiesen's value at 0° C for that 

 metal as given in Everett's ' Physical Units.' 



