Dr. Tyndall on the Laws of Magnetism. 279 



method which pemiits of its exhibition, the increase being most 

 appreciable where the weights are small. This oui- last mode of 

 experiment actually does. Were the angles at the top of the co- 

 lumns in the last two tables a little smaller, the results would be 

 the same as those in Tables I. II. and III. It seems, therefore, 

 at fii'st sight fair to attribute the extraordinary agreement among 

 the earlier results, in part, to lagging magnetism, and hence to 

 infer that the law is not so precise and simple as they would in- 

 thcate. 



21. Plausible as this appears, I cannot accept it as the true 

 explanation of the above divergences. If they be due to lagging 

 magnetism, it is only necessaiy to give the magnet time to sub- 

 side to cause this to lUsappear. The old method of experiment 

 was again resorted to. A weight of 500 grammes was laid upon 

 the scale-pan, and, commencing at the top of the rheostat, the 

 handle was turned and the cm-rent damped until the ball gave 

 way. This was foimd to occur as the handle of the instrument 

 made its thirty-seventh revolution. A series of four descents were 

 first made, slowly, but without pausing ; then four more, stopping 

 for two minutes at the thirty-fifth revolution, and thus allowing 

 the lagging magnetism, if su.ch existed, to subside. Supposing 

 such to exist, then the deflections exhibited in the last four ex- 

 pei'iments must be greater than those exhibited in the first four ; 

 for the subsidence of the lagging magnetism must be compen- 

 sated by an increase of current, the said increase being indicated 

 by a greater angle. This conclusion, which necessarily flows from 

 the above assumption, is, however, diametrically opposed to fact. 

 The angles in the last four experiments, instead of being greater, 

 are actually less than the others. They are as follows : — 



Weight 500 grammes. 



Pausing 2' at the 

 Without pausing. 35th revolution. 



It is thus proved that the deviations cannot be accounted for 

 on the supposition that tlie magnetism lingers in the core after 

 the current has fallen. Further on an attempt will be made to 

 exhibit their true origin. 



22. From Table V. we derive a notion of the influence of 

 size upon tlic attraction of tlie sphere, l^y selecting the weights 

 common to all three balls, and comparing the corresponding 

 magnetic pf)wers, we learn, that, to support the same weight, the 

 ball No. 2, when in contact, requires a current 2*4 times as 



