[ 319 ] 



XXX. The Magnetic Effect of Electric Convection. 



To the Editors of the Philosophical Magazine. 



Trinity College, Cambridge. 



Gentlemen, Aug. 5, 1901. 



REFERRING to Dr. Cremieu's reply in your August 

 number to my article in your July number, I have to 

 apologize for having (owing to an oversight in reading his 

 short note in the Comptes Rendus) represented Dr. Cre'mieu as 

 using a vertical bar-magnet merely, whereas he actually 

 used a rectangular iron circuit with magnetizing coils on one 

 of its sides. 



I cannot agree, however, with Dr. Cremieu in thinking that 

 my objection to his first experiment falls to ihe ground in 

 consequence of this oversight on my part. 



It is easy to show, using Dr. Cremieu's data, that nearly all 

 the magnetic lines of force in his apparatus must have passed 

 through the air and not round the iron rectangle as he supposes. 

 Consequently my objection to his experiment still holds good ; 

 and in fact this experiment, if it proves anything, proves that 

 the electromotive force due to varying magnetic induction 

 does act on a static charge. I think that if Dr. Cremieu were 

 to repeat this experiment, using merely a bar-magnet instead 

 of his rectangle, he would still obtain no effect ; whereas if he 

 is right and the effect in question does not exist, he ought to 

 obtain a deflexion due to the action of the magnetic field on 

 the current when the disk is charged. 



With regard to the rest of Dr. Cremieu's reply, I do not 

 think that what he says in any way affects my objections. 

 My suggestion that the insulation of his sectors was not good 

 enough, is only a very small part of my criticism of his 

 experiments with rotating disks. 



Since I wrote the article in the July number, Dr. Cremieu 

 has published an account of an experiment which he supposes 

 proves very conclusively that " open currents " can exist. 



In this experiment a disk of ebonite 25 millimetres thick 

 carrying sectors on one side was rotated rapidly. Near the 

 other side of the disk a charged metal sector was placed so as 

 to induce charges on the sectors as they passed by it. The 

 motion of these induced charges constituted the " open w part 

 of the current. The charged sector was 2*5 millimetres distant 

 from the disk, and its potential was about 115 electrostatic 

 units. 



This potential is sufficient to produce a spark 1 centimetre 

 longj so that of course a brush-discharge would occur to the 



