on the Co rhino Effect. 693 



recently published oy one o£ us* an expression for the 

 magnitude of this effect was obtained. The assumptions 

 made were that the electric current is carried wholly by free 

 electrons, which in the absence of an external electric field 

 pursue, with equal velocities, zigzag paths between collisions 

 with the atoms of the metal; that the collisions are similar 

 to those between hard elastic spheres, and that the effect of 

 an external electric field is to give the electrons a drift in 

 the opposite direction. The effect of the magnetic field is to 

 bend the paths of the electrons so that there is a component 

 velocity at right angles both to the electric and magnetic 

 fields. It was also indicated that this Corbino effect is 

 closely related, if not identical, to the Hall effect. There is, 

 however, this difference. The Hall effect is the production 

 of a transverse difference of potential when an electric 

 current flows through a plate in a magnetic field. To get 

 effects large enough to be measured the plates must be made 

 very thin in order to increase their resistance. The Hall 

 effect is thus proportional to the primary current and the mag- 

 netic force, and inversely proportional to the thickness. The 

 Corbino effect, on the other hand, is the production of a 

 transverse electric current, and is proportional to the primary 

 current and the magnetic force, but independent of the 

 thickness of the plate. The Corbino effect would seem to 

 be the fundamental galvanomagnetic effect, rather than the 

 Hall effect, in that the free transverse boundaries necessary 

 in the latter effect are wholly absent in the Corbino effect. 

 And it is known that these free boundaries introduce very 

 considerable complications into the interpretation of the 

 Hall effect. The fundamental explanation of one effect will, 

 however, explain the other effect. 



This conclusion is supported by the results of the experi- 

 ments to be described, in so far as the sign of the effects is 

 concerned. In all metals tried the Corbino effect has been 

 found to have the same sign as the Hall effect. But there is 

 a marked difference in the relative magnitudes of the effects 

 for different metals. For example, the ratio of the Hall 

 effects in bismuth and copper is 11,000 ; while the ratio of 

 the Corbino effects is only 150. Whether variations of this 

 kind are the result of using different .specimens in measuring 

 the two effects, or whether they arise from an inherent dif- 

 ference — such as boundary conditions — remains to be found 

 out by further experiments that are in progress. 



The method we have employed to measure the Corbino 

 effect consists in determining the current induced in a coil 

 * Phil. Mag. February 1914, p. 244. 



