MAGNETISM AND TWIST IN IRON AND NICKEL. 



505 



vertically in the magnetising helix, and subjected it to the usual treatment. I 

 hoped for an indication, but I got measurable results as follows : — 



had 



Line 

 Current. 



Longitudinal 

 Field. 



Twist ± H. 



Twist ± C. 



1-76 



100 

 73-9 

 47-1 



0-62 

 0-42 

 0-18 



0-29 

 0-18 



o-io 



As in previous cases, the twists are given in thousandths of a radian. They are much 

 smaller than for nickel, and considerably smaller than the maximum twists for iron. 

 This may to some extent be explained in terms of the rectangular section of the rod. At 

 the same time, up to the highest field used, 100 namely, the twists in a cobalt wire might 

 well be expected to be less than the twists in iron or nickel wires, if we accept Maxwell's 

 explanation. For, according to Mr Bid well's measurements, it is not until much higher 

 fields are used that the contraction of cobalt attains a maximum of magnitude somewhat 

 greater than the maximum elongation of iron. The interest of the present experiment 

 lies chiefly in the fact that the direction of the twist in cobalt is the same as in nickel, 

 but opposite to the direction of the twist in iron in low fields. Thus my prediction in 

 Part I. (p. 386) has been partially verified ; and we may now say that the Wiedemann 

 effect in iron is positive in low fields and negative in high ; in cobalt, it is negative in 

 moderate and probably positive in very high fields ; in nickel it is always negative. 



It will be noticed that the current-reversal twist is much smaller than the field- 

 reversal twist ; probably with stronger currents and higher fields we should find cobalt 

 behave very similarly to iron and nickel. 



13. The Behaviour of Nickel Strips. — The satisfactory manner in which the cut 

 cobalt rod behaved induced me to try similar experiments on five nickel strips of varying 

 breadth cut from the same sheet. The sheet was 0'34 mm. in thickness ; and the strips 

 varied in breadth from 1*1 mm. to 3 mm. They were all 26'6 cm. long. The following 

 concise Table gives some of the results obtained. The longitudinal field in all cases was 

 the same, namely, 60. Three different values of current along the strip were taken for 

 each ; and measurements were made of both the field-reversal and current-reversal twists. 

 These form the six columns headed ± H, ± C ; and each row of twists belongs to the 

 strip whose diameter completes the row on the left. The line currents are given as 

 headings to the three pairs of twist columns. The twists are in thousandths of radians. 



Twists of Nickel Strips in Field h = 60. 



Breadth of 

 Strip. 



Line Current. 



34 



21 



1-2 



±H 



±c 



±H 



±c 



±H 



±c 



1-1 

 1-5 

 2-0 

 2-6 

 3-0 



19-5 

 13-1 

 11-5 

 103 

 100 



8-3 

 4-6 

 4-0 

 31 



2-8 



14-4 

 10-1 



8-5 



7-7 

 7-5 



4-3 

 2-6 



i'-8 



1-6 



10-4 

 7-3 

 5-9 

 5-5 



5-6 



2-2 



1-4 



1-1 



•9 



•8 



