[ 153 ] 



XXI. Intelligence and Miscellaneous Articles. 



ON MAGNETS. BY O. A. S. PIHL*. 

 rPHE author, in his comprehensive treatise, seeks to trace the at- 

 -*- traction of armature and magnet to the distribution of the mag- 

 netism in the two. He first describes his multifarious and careful 

 experiments on the distribution of the magnetic moments in a 

 straight magnet when both poles are free or when iron rods are 

 placed at the ends, or an iron rod laid between two like or unlike 

 poles of magnets, and likewise in these iron rods and in one with 

 its extremities placed between the arms of a horse-shoe magnet. 

 The determinations were made by means of induction-spirals after 

 the method of van Bees. In like manner were the attractions of 

 the magnets for their armatures at various distances determined by 

 forcible separation. It would carry us too far to specify the parti- 

 cular results, which, for the first determinations, fall in with those 

 obtained by Erman, van Rees, Weihrich, aud others, as they also 

 are to be deduced from the putting into position of the molecular 

 magnets (gradually becoming feebler from the operating force on- 

 wards). Beside the graphically represented distribution of the 

 moments, the above-sketched distributions of the free magnetisms 

 are deduced. 



Accordingly, if we imagine the magnets and armatures made up 

 of slices which contain each the calculated amount of free magne- 

 tism, we can calculate the attraction exerted by each slice of the 

 magnet upon each slice of the armature in juxtaposition with it, 

 assuming that it is distributed uniformly over the slices. The sum 

 of all the attractions should give the attraction of magnet and arma- 

 ture. The calculated and the observed values, however, are not at 

 all equal : if the armature be brought nearer to the magnet, the 

 ratio of the observed to the calculated attraction diminishes, until 

 at a distance of about -| of the radius it reaches its minimum and 

 then with lessening distances increases. The deviation, at all 

 events, lies in the well-known more copious distribution of free 

 magnetism at the places more remote from the axis of the magnet- 

 slices, which can be observed by removing them by corrosion &c, 

 but cannot be accurately determined, since herewith the distribution 

 changes. On the approximately true assumption that one portion 

 of the magnetism is distributed uniformly over the cross section, 

 another only on the periphery of the slices, at greater distances 

 results are obtained agreeing better with the observations. 



If the diameters of armature and magnet are unequal, the attrac- 

 tion becomes different, and, correspondingly, the magnetisms distri- 

 buted in it are otherwise arranged, and of another amount, than 

 with an armature of equal thickness, as the curves constructed by 

 the author show. 



On placing two permanent steel magnets in juxtaposition the 

 distributian in the two is likewise altered ; only the direct action is 



* Christiania, 1878 : pp. i-xiv, and 1-135 ; 6 plates. 

 Phil Mag. S. 5. Vol. 9. No. 54. Feb. 1880. M 



