an Explanation of Hall's Phenomenon. 263 



between the ends of the plate ; and as regards transverse dis- 

 tortion, the central bridge merely forms a link between the 

 upper and lower portions of the plate, which link shares their 

 movement in the upward direction without impeding it. Thus 

 the reversal of the Hall effect is easily and completely accounted 

 for. 



The i Philosophical Magazine ' for last January contains a 

 translation of a paper by Prof. Bighi, on Hall's Phenomenon. 

 He uses three electrodes instead of four. The current enters 

 or leaves by one. and leaves or enters by the other two. The 

 two partial currents pass through a " Wiedemann's " (pre- 

 sumably a differential) galvanometer, and zero is obtained by 

 the use of resistance- coils. On exciting the electromagnet a 

 galvanometer-deflection occurs, indicating a current in the 

 opposite direction to that named by Hall. 



This also can be explained by strain and Peltier action. 

 Let us suppose, as before, that the plate is of iron, and that 

 the south pole is beneath it. Let a current enter the plate at 

 A, fig. 6, and leave it at the two points D and E by the wires 

 leading to the differential galvanometer Gr. The plate is 

 divided into six strained regions exactly as in Hall's arrange- 

 ment, those which are compressed being marked C, and those 

 which are stretched S. But the neighbourhood of the upper 

 electrode D will, in this case, be heated, because the current 

 is passing from stretched to compressed : the resistance of the 

 region will therefore be increased. Similarly the resistance 

 of the metal in the neighbourhood of the lower electrode F 

 will be diminished. Thus a greater quantity of electricity 

 will pass by E than by D, and the galvanometer-deflection 

 will be in the opposite direction to that which would have 

 occurred if the connexions had been made in the middle of 

 the plate, in the usual manner. Eighi's experiment therefore 

 confirms my explanation of the Hall effect. 



The theory which I have ventured to submit seems to me 

 capable of bearing any test which could reasonably be applied 

 to it. Not only does it completely account for the phenomena 

 described by Hall, but it explains anomalies and enables one 

 to predict results. The chief point of difficulty lies in the 

 strain. In conversation on the subject, I have found that, 

 while it is readily admitted that a strain such as is described 

 might well be produced in a piece of metal a millimetre in 

 thickness, it is by no means so easy to conceive the possibility 

 of a similar strain in a sheet the thickness of which does not 

 exceed a hundredth or a thousandth of a millimetre. But let 

 it be imagined that the thick piece of metal is divided into a 

 hundred or a thousand independent layers, each of which is # 



