MAGNETISM. 



47 



of an inch. He then arranged a 

 number of equal steel bars (A, B, C, D, 

 E) in a straight line, and made one extre- 

 mity of the conjoined magnets slide at 

 right angles over the line of steel bars. 

 He did not, however, limit himself to one 

 direction, but moved them backwards and 

 forwards the whole length of the united 

 surfaces of the bars ; repeating the ope- 

 ration on each side until he had obtained 

 as great an effect as possible. 



In order to equalize as much as pos- 

 sible the magnetic power of the two ends 

 of each bar, it is expedient to commence 

 each operation by laying the conjoined 

 magnets on the middle of the line of bars, 

 and to pass the magnets over each half 

 of the line an equal number of times ; 

 at the conclusion of which, the magnets 

 being brought again to the middle, 

 they should be raised perpendicularly, 

 so as not to disturb the lateral effects 

 which had been produced. Mr. Mitchell 

 found that the steel bars B, C, D, which 

 were intermediate in the series, acquired 

 by this process a very great degree of 

 magnetic power. Those which formed 

 the extreme bars of the series A, D, were 

 much less impregnated ; but by remov- 

 ing them from this situation, and trans- 

 ferring them to the middle of the series, 

 and then repeating the same operations, 

 they quickly acquired the same degree 

 of magnetism as the rest. 



(189.) The process above described, 

 which soon acquired much celebrity, 

 was called the method by double touch ; 

 and it is asserted by its inventor, that 

 two magnets will impart more magnetic 

 power to a bar of their own size, when 

 employed in this peculiar mode, than a 

 single magnet of five times the strength 

 of the former, when applied after the 

 manner of the single touch. The opera- 

 tion of the two poles of the conjoined 

 parcels of magnets on those portions of 

 the bars over which they pass will readily 

 be understood from what has been said 

 with respect to the methods of Knight 

 and Duhamel. They act by the sum of 

 their inductive powers on those parts 

 of the bar that are situated between 

 them, but with the difference of those 

 powers on all those parts which lie be- 

 yond them ; and the former is therefore 

 always greatly more efficient than the 

 latter. The superiority is the more con- 

 siderable in the present case, inas- 

 much as the magnets are nearer to each 

 other, and therefore act with much 

 greater power when they co-operate, but 

 are nearly inefficient when _they oppose 



one another. The latter of these forces, 

 therefore, will never have sufficient 

 energy to destroy, or even much di- 

 minish, the effect which had been pro- 

 duced by the former; and thus the 

 magnetism of each portion receives con- 

 tinual accessions of strength every time 

 the magnets are made to pass over it. 

 The long line of bars operates in a man- 

 ner similar to the pieces of soft iron at 

 the extremities of those in the paral- 

 lelogram of Duhamel that is, the ex- 

 ternal bars act as armatures to those 

 which lie between them ; and hence may 

 be understood why these intermediate 

 bars receive the strongest impregnation. 



(190.) The different processes for 

 communicating magnetism which we 

 have now described, comprise all those 

 methods that are essentially different in 

 their principle ; all others which have 

 been proposed may be regarded as va- 

 rieties merely in the combinations of 

 which these principles are susceptible. 

 We shall only, therefore, notice those 

 which have been most in repute. 



(191.) Mr. Canton published, in 1751, 

 a method which he considered as su- 

 perior to any of those previously em- 

 ployed. He placed the bars intended to 

 be magnetized so as to form a parallelo- 

 gram with connecting bars or arma- 

 tures of soft iron, as in the method of 

 Duhamel. He then had recourse to the 

 method of double touch as prescribed 

 by Mitchell; after which he separated 

 the two bundles of magnets, and in- 

 clining them to the bars in contrary di- 

 rections, as Duhamel had done, he com- 

 pleted the operation by making them 

 slide from the middle towards the extre- 

 mities. The combination of these two 

 processes was considered by Canton as 

 an improvement upon the method of 

 Mitchell. There is, however, great rea- 

 son to think, as Coulomb and Biot 

 have remarked, that these successive 

 operations are quite superfluous, and 

 that the bars are left at the end of them 

 precisely in the same state as if only the 

 last had been employed. 



$ 8. JEpinuiiS Method. 



(192.) j^Epinus introduced modifica- 

 tions into the process of the double 

 touch, of greater importance and much 

 more judiciously conceived. He first 

 formed the parallelogram of steel bars 

 in the manner of Duhamel ; but in place 

 of the auxiliary cross bars of soft iron, 

 he connected the ends of the steel bars by 

 means of other steel bars which hadpre- 



