On a null method of measuring the gyro-magnetic ratio

Abstract

It was first shown by O. W. Richardson that if the magnetic properties of ferro-magnetic substances are due to the rotation of the electron orbits of the atom, then a change M in the magnetic moment of a specimen of such material should be accompanied by a change in the angular momentum U given by U/M = 2 m / e , where m and e are the mass and charge of an electron respectively. Several experimenters have shown that the value of the ratio is approximately one-half that expected on the above theoretical grounds. Recently work in this University has indicated that within the limits of experimental error the value of the ratio is exactly one-half for iron and nickel. The present work was undertaken with the object of checking these results, and further, of determining the ratio for the Heusler alloys. Owing to the time lag which accompanies the magnetisation of these alloys, the only method-which has so far proved suitable is that of Barnett, in which the change of magnetic moment is produced by rotation. In an advance note§ it is stated that for certain ferro-magnetic substances, including Heusler alloys, the ratio is one-half, but the results are “within 10 per cent, or less.”