on the Polarity of the Biamagnetic Force. 169 



the position of rest being eleven divisions more than when the 

 bars were central. 



As the positions here referred to will be the subject of frequent 

 reference, for the sake of convenience I will call the position of 

 the cylinders sketched in fig. 1, Position 1 ; that sketched in fig. 

 2, Position 2 ; and that sketched in fig. 3, Position 3. The re- 

 sults which we have just described, tabulated with reference to 

 these terms, would then stand thus : — 



I. 



Bismuth cylinders. 



Length 3 inches. Diameter 0*7. 



Position 1. 468 Position 2. 482 Position 3. 493 



In changing therefore from position 1 to position 3, a deflection 

 corresponding to twenty-five divisions of the scale was produced. 

 Wishing to place myself beyond the possibility of illusion as 

 regards the fact of deflection, I repeated the experiment with 

 successive batteries of two, three and four cells. The following 

 are the results : — 



II. 



In all the cases cited we observe the same result. From posi- 

 tion 2 to position 1 the motion is from larger to smaller num- 

 bers; while from position 2 to position 3 the motion is from 

 smaller to larger numbers. 



It may at first sight appear strange that the amount of the 

 deflection did not increase with the battery power ; the reason, 

 in part, is that the magnet, when the current circulated, was held 

 in a position free from the spirals, by forces emanating partly 

 from the latter and partly from a portion of the external circuit. 

 When the current increased, the magnetization of the bismuth in- 

 creased also, but so did the force which held the magnets in their 

 position of equiUbrium. To remove them from this position, a 

 greater amount of force was necessary than when only the resi- 

 dual action of a feeble current held them there. This fact, cou- 

 pled with the circumstance that less heat was developed, and less 

 disturbance caused by air currents, when a feeble battery was 

 used, induced me for some time to experiment with a battery of 

 two cells. Subsequent experience however enabled me to change 

 this for five cells with advantage. 



Notwithstanding the improbability of the argument, still it 

 may be urged that these experiments do not prove beyond a 



