374 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



Plate 4 



Moretto's method is defective aud subject to large errors. His con- 

 clusion as to the effect of current density is given apparently as a 

 general property of metals. This is an unwarranted inference from 

 the evidence given in his paper, which does not specify the metals 

 worked upon in obtaining the table reproduced above. 



The writer's work on gold appears to indicate a very small increase 

 in the Hall effect for the weaker values of the piain current, as in a 

 decided majority of the sets of observations the Hall effect comes out 

 a little larger for the weaker values of the main current. The difference 

 is perha[)s within the limits of error. It may possibly be a result of 

 the large stresses in thin gold due to the expansion and contraction 

 of the paraffin and cements. Under the infiiience of the heating due to 

 the strong currents, these stresses might be of a different order of 

 strength from the stresses when a very weak current is flowing. There 

 might therefore be some difference in the properties of the metals in the 

 two cases. 



The writer proposes to continue the investigation, using also some 

 other metals. 



The writer desires to express his gratitude to Pi'ofessor Hall for 

 constant advice and personal aid, and to him and to the other professors 

 of the physical department for the kindly loan of the apparatus for the 

 vvoik. 



