ON MAGNETOSTRICTION. 29 



It may at first sight appear that the smallness of the tliermal 

 expansion in nickel steel necessarily entails the minuteness of the 

 chanofe of lensfth and of volume, but no connection seems to exist 

 between the magnetostriction and the deformation due to temperature 

 variation, as illustrated in the preceding experiments. 



§ 7. Wiedemann Eflect in Nickel Steel. 



As closely allied to the change of volume and of length, the 

 Wiedemann effect comes into oar consideration as due to magneto- 

 striction. Unfortunately we could not investigate the phenomenon 

 in cobalt for want of material of geometrical shape, suitable for the 

 investigation. The results for iron, steel, and nickel have been 

 already described in our former paper, so that we shall consider only 

 nickel steels, which have not yet been investigated. 



The eft'ect was measured in the usual way by suspending the 

 wire vertically in a magnetizing coil, and by passing an electric cur- 

 rent of known streno;th, the ang-le of torsion due to the combined 

 action of circular and longitudinal magnetizations was measured by 

 the rotation of a fine mirror attached to the end of the wire. The 

 vertical component of the terrestrial magnetic force was compensated 

 for by another coil inserted within the coil. The wire was of such 

 length (21 cm.) that the magnetic field was practically uniform 

 throuofhout. For the different sorts of wires tested, we are indebted 

 to Mr. Ch. Ed. Guillaume. 



The measured angles r of torsion per cm. in seconds of arc are 

 given for different samples of wires in the following table : 



