Magnetic Pulariti/ in Metallic Bodies. 37 



tiriff horse-shoe; and as the actuating forces in the edge of the 

 disc operate in the smne direction, both needles will be im- 

 pelled in one and the same way ; so that whatever may be their 

 position when deflected, they will constantly appear in the 

 same vertical plane. The arrows in fig 25. show the direc- 

 tion of the aggregate forces in the edge of the disc, when it is 

 rotated in the direction as shown in fig. 16. 



The singular and complicated distribution of the force dis- 

 covered in these rotating discs of copper, led me to undertake 

 some other experiments, by means of which I considered it 

 possible that I might arrive at some simple law, which would 

 disclose the novel and apparently mysterious arrangement; 

 ibr, whether the phsEUomena emanate from magnetic or from 

 electro-magnetic action, there ajipeared to me to be no law 

 yet discovered in either of these branches of research, that 

 would produce a distribution of polarity like that which 1 

 have portrayed in fig, 16. and 17; notwithstanding which, 

 the uniformity of the distribution, which became manifest at 

 every repetition of the experiments, left no doubt as to the im- 

 mutability of some law, to the operation of which the regu- 

 larity of the distribution was entirely owing. 



In this investigation it was necessary to take into consi- 

 deration the various directions which different parts of the re- 

 volving disc assume with regard to the exciting magnet; for, 

 as the })oles are not placed in the centre of motion, it is })laiii 

 that whilst some parts are advancing towards ihem, other 

 parts are receding from their vicinity; — some parts again are 

 crossing the magnet to the riglit, whilst' others are crossing it 

 towards the left; and all these motions in the disc are going 

 on at the same time; so that upon the whole the apparent 

 complexity of the problem put any inquiry concerning it rather 

 in the position of a "forlorn hope", than of anything like cer- 

 tainty of success. 



Considering, however, that as the vicinal regions of the disc 

 must necessarily receive the exciting impressions in a much 

 higher degree than those more remotely situated from the 

 magnetic poles, it might be ex})ected that if any satisfactory 

 conclusions were to be arrived at, those parts of the disc the 

 most powerfully excited were more likely than any other to 

 afford the necessary data. My inc|uiries were therefore more 

 particularly directed to the investigation of that half of the 

 tlisc which is nearest the magnet, the curvilinear direction of 

 which, with regard to the exciting pole, is easily resolved into 

 four rectilinear motions. 



Let m u, fig. 26*, be the constant radius situated between 



• In coriscqueiue of iiii ovrrsiplit, fig. ~fi. will not be foiiiul in ihc j)Iatc; 

 U \\ili be "ivtn iii our next Nunibcr.— Edit. 



