Magnetism on a permanent Electric Current. 305 



tive force of a Bunsen cell, or 340 millims. on reversing the 

 connexions with such a cell. The battery of Leyden jars was 

 charged to a potential sufficient to give a spark of 2 or 3 

 millims. The connexions being thus made, the position of the 

 spot of light was observed and the magnet then operated with 

 the purpose of discovering, if possible, any consequent change 

 of position of the spot of light which would indicate an action 

 of the magnet on the lines of static induction in the glass. 

 The observation failed to establish the existence of any such 

 action. The electrometer being in a very sensitive condition, 

 the spot of light was rather unsteady ; so that any very slight 

 effect of the kind looked for would not have been detected, 

 though it is probable that, if a reversal of the magnet had 

 caused a change of four millims. in the position of the spot of 

 light, this effect would have been apparent. 



We may therefore conclude that any change of relative po- 

 tential on the quadrants of the electrometer caused by reversal 

 of the magnet was probably less than -g^ of that caused by 

 reversing the connexions of the electrometer with a Bunsen 

 cell, as mentioned above. If now we estimate the difference 

 of potential between the plugs A and B, connected with the 

 Leyden jars, to have been, as indicated by the length of the 

 spark, equal to that which would be produced by 10,000 Bun- 

 sen cells in series, we may conclude that any difference of 

 potential between the other plugs C and D which was caused 

 by the action of the magnet must have been less than - % ooWo 

 of the difference of potential between A and B. We must 

 remember, however, that any change of potential on C and D 

 had to be extended as well over the comparatively large area 

 of the electrometer quadrants. Professor Rowland has roughly 

 estimated the capacity of the quadrants as twenty times that 

 of the plugs C and D. If, therefore, these plugs had not been 

 attached to the electrometer, any difference of potential between 

 them due to the action of the magnet would have been twenty 

 times as great as in the actual case; so that instead of ^qoVd^ 

 we have 40000 °f ^ ne difference of potential of A and B as the 

 superior limit of the difference of potential of and D which 

 the magnet might possibly have produced if C and D had not 

 been connected with the electrometer. Representing the 

 former difference of potential by E, the latter by E', and the 

 strength of the magnetic field, about 4000 ccntim.-grm.-sec, 



by M, we have for this case of static induction in glass ^ — ^, 

 if not zero, less than ^ooooooO - . . . X 



Turning to the analogous case of current-electricity in the 

 various metals, and representing now by E the difference of 

 potential of two points a centimetre apart in the direction of 



