on the Corbino Effect* 695 



the disk, and simultaneously the induced current in the coil. 

 The commutator makes four reversals for each revolution 

 and was run at five revolutions per second. A null method 

 of measuring the circular current in the disk was used. The 

 radial current went through the primary winding of a cali- 

 brated variable mutual inductance. In series with the 

 secondary windings was a moving coil galvanometer and 

 the coils Ci and C 2 shown in the figure. If M is the mutual 

 inductance between the circular current, C, in the disk and 

 the coil current, I the radial current in the disk, and m the 

 reading of the variable mutual inductance when there is no 

 current through the galvanometer, then 



MC=mI, 



which determines the circular current in the disk. 



The variable mutual inductance consisted of a fixed single 

 turn of wire on a slate ring, inside of which a slate disk, with 

 16 turns of wire wound in a groove on its circumference, 

 could be rotated about a diameter of the fixed ring by means 

 of a worm wheel. The position of the rotating coil was 

 read by means of a pointer moving over a divided circle ; 

 over the range employed, 1° rotation corresponded to a 

 change of mutual inductance of 22*3 absolute units. In 

 order to bring the reading within the proper range of the 

 calibrated instrument a second variable mutual inductance 

 was inserted in the circuit, by means of: which an approximate 

 balance could be obtained. 



It was found impossible to make the metal disks absolutely 

 uniform, with the result that there was always a circular 

 component of the current even in the absence of a magnetic 

 field. But this was allowed for in the following manner. 

 Let M' be the mutual inductance between the coils and the 

 disk for the circular component of the current arising from 

 lack of uniformity in the disk. In M' is included the mutual 

 inductance between the various parts of the circuit which 

 could not by any arrangement of the wires be made wholly 

 to vanish, and also the mutual inductance of the variable 

 coils when set so as to give a suitable range to the calibrated 

 standard. Let C be the circular current in the disk in the 

 absence of a magnetic field when the radial current is I. 

 Let C be the circular current produced by a positive magnetic 

 field, and M the mutual inductance between the disk and 

 coils for this current. Let m l and m 2 be the balancing 

 mutual inductances of the variable standard when the mag- 

 netic field is positive and negative respectively. Then for 



