Hall and Corbino Effects. 325 



Below are the values of c for the bismuth-tin alloy (1 per 

 cent.) ; they must be considered rather unreliable on account 

 of the extremely small effect obtained at these fields. 



The Corbino Effect in a Bismuth-Tin Alloy 

 (al)out 1 per cent, solenoid). 



H. J 



2S/H. 



rXlO 7 . 



380 



0-8 



9-1 



499 



11 



9-5 



618 



1-25 



8-7 



743 



1-6 



9-3 



876 



20 



9-8 





Mean 



... 9-3 



The Effect of the Lack of Symmetry of the Field in the 

 Electromagnet. — As was explained before, in taking measure- 

 ments such as the present ones, the balancing inductance is 

 adjusted so that the galvanometer shows no deflexion. This 

 is done first with the field in the positive direction ; then 

 the field is reversed, the inductance again adjusted for no 

 deflexion, and half the difference between the two readings 

 on the adjustable inductance is taken as a measure of the 

 Corbino current. In every case where the magnet was 

 employed, the two positions of the balancing inductance 

 were asymmetrical with respect to its position for zero 

 field. It was at first supposed that this was due to the lack 

 of exact uniformity of the disks and a consequent circular 

 component of the primary radial current. However, upon 

 experimenting with the solenoid, this lack of symmetry was 

 found to disappear, showing that some other reason must be 

 souo-ht for its presence when the magnet is used. The 

 possibility at once suggests itself that a distortion of the 

 field might be responsible ; it is seen that such a condition 

 could easily arise through the lack of parallelism of the pole- 

 faces, or through a lack of coincidence between the centre 

 of the disk and the centre of the hole traversing one of the 

 pole-pieces, or by a failure to place the disk perpendicular to 

 the flux. In order to test this, a bismuth disk was so 

 arranged as to be movable between the pole-shoes while 

 everything else, including the magnetizing current, remained 

 constant. Approximate settings were then made with the 

 disk in varying positions ; the movements were such as to 

 produce exaggerated representations of changes in position 



