1902.] on the Electric Conductivity of Iron and Nickel. 37 



It should be mentioned tltat the change of length produced by 

 magnetisation in iron and nickel is too small in moderate fields to 

 account for the observed variation of resistance ; it is, however, possible 

 that this effect may give rise to an important correction in strong 

 fields. 



It is obvious that the method employed in these experiments com- 

 pletely eliminates any change of resistance which is reversed in sign 

 by reversal of the field ; but such an effect has never been observed. 



Experiments in High Fields. 



Experiment I. — Longitudinal Effect in Nickel. 



Another specimen of wire was used having a diameter of 033 mm., 

 and covered with double silk insulation. 



It being desirable to determine the magnetic field and magnetisation 

 under exactly the same conditions as the change of resistance, the 

 apparatus was arranged so as to allow the three measurements to be 

 made with the same nickel coil. A modification of the " isthmus " 

 method was adopted. The magnetic field was produced by means of a 

 large electromagnet provided with conical pole pieces having faces 

 1*7 cm. in diameter. 



The essential part of the apparatus is represented diagrammatically in 

 figs. 5 and 6. The nickel wire was wound in a single layer, in and 

 out the thin glass cylinder C, 1*1 cm. long and 1*1 cm. in diameter, so 

 that the turns of the coil, 88 in number, were parallel to the axis of 

 the cylinder, as shown in fig. 5. The glass being thin, the transverse 

 end elements of the coil were small in comparison with the length of 

 the cylinder. There were three ballistic coils; the innermost was 

 wound on a glass tube D, the second on the brass cylinder B, and the 

 outermost on the brass flange A, which extended over half the 

 cylinder B. This system of coils was attached to one end of a wooden 

 arm, having a pivot at the other extremity, by means of which the 

 coils could be suddenly placed in the proper position between the poles 

 of the magnet. Before every reading the nickel was demagnetised by 

 a separate arrangement. The ballistic deflections as well as the resist- 

 ance change were produced by placing the coils in the field, the effects 

 of residual induction in the nickel being thus avoided. There being 

 no compensation for temperature variations, it was necessary to make 

 the observations as quickly as possible. It was arranged that the 

 ballistic deflections measured the differences of the total induction 

 through the consecutive coils. The values of H and I were then easily 

 calculated from the two observed deflections. 



As usual, it is assumed that the value of the field just outside the 

 nickel is the same as that in the metal. The innermost ballistic coil 



