678 The Corbino Effect in Iron. 



It lias been shown by Prof. A. W. Smith *, however, that 

 in electrolytic iron the Hall effect shows no such behaviour. 

 This apparent difference in the two effects might, of course, 

 be accounted for by the difference in the character of the 

 iron used in the two experiments, a disk made from Norway 

 iron having- been used for the Corbino effect. But it hardly 

 seemed probable that there should be such a pronounced 

 difference between the two kinds of iron. For this reason I 

 have made further experiments on the Corbino effect, using 

 both a disk cut from the' electrolytic iron kindly furnished 

 by Prof. Smith, and the disk of Norway iron previously 

 used. 



The method used is the same as that employed in the 

 previous experiments. The radial current flowing through 

 the disk is reversed in direction about 20 times a second. 

 The circular current produced in a magnetic field is thus 

 reversed with the same frequency and induces a current in a 

 coil placed parallel to the disk. The coil is connected to a 

 sensitive galvanometer through the secondary of a variable 

 mutual inductance, the primary being in the circuit of 

 the radial disk current. The galvanometer connexions are 

 reversed at the same frequency as tlie radial current, and a 

 steady current flo.\ s through the galvanometer, which can be 

 balanced by varying the mutual inductance so as to make the 

 method a null one. The change in the mutual inductance 

 required for a balance when the magnetic field is leversed in 

 direction is a measure of the circular current produced. 



A new double commutator was constructed which is a 

 decided improvement over the one previously used. The old 

 one was a source of considerable trouble, due largely to the 

 variable thermoelectric forces at the sliding contacts. The 

 new instrument is much more massive than the old one, the 

 sliding contacts are of copper with copper gauze brushes 

 more than twice as large as the old ones, and currents of 

 20-30 amperes can be reversed with not excessive sparking, 

 and with almost negligible thermoelectric forces produced in 

 the galvanometer circuit. 



The results of these last experiments have shown that in 

 nei her kind of iron is there any reversal in sign of the 

 Corbino effect. In the lowest fields used the true Corbino 

 effect is so small that it can just be detected ; detailed results 

 of the experiments are not of any particular value, as the 

 errors inherent in the method are of the same order of 

 magnitude as the effect itself. The approximate value of the 



* Phys. Review, xii. p. 337 (1918). 



