72 Royal Society. 



that two boilies are at tliii same time in the magnetic fielil, and that 

 one displaces the otlier, he considers the result as a dilferential effect 

 of tlieir ditierence in conducting power. 



If a free portion of space be considered with lines of equal mag- 

 netic force passing across it, they will be straight and parallel lines. 

 If a sphere of paramagneiic matter be placed in sucii a space, tiiey 

 will gather upon and in the sphere, being no longer parallel in their 

 course, nor of equal intensity in every part; or if a sphere of dia- 

 magnetic matter replace the former sphere, the lines of force will 

 open out where the sphere is, being again no longer parallel in direc- 

 tion nor uniform in force. When the field of magnetic force is formed 

 between the opposite flat ends of two large magnetic poles, then 

 these are aHected,and the globes also, and there are mutual actions; 

 a paramagnetic body, if a little elongated, points axially and tends 

 to go to the iron walls of the field, whilst a similar diamagnetic 

 body points equatorially, and tends to go to the middle of the field. 

 Paramagnetic bodies repel each other, and so also do diamagnetic 

 bodies ; but one of each class being taken, they attract one another. 



The convergence of the lines of force upon the opposite sides of 

 the paramagnetic sphere, and the corresponding divergence of them 

 on the opposite sides of the diamagnetic sphere, the author expresses 

 by the term comhictlon polarity. 'I'iiis polarity he carefully di- 

 stinguishes from that which depends upon the reversion of the di- 

 riction of the power ; the latter he considers as a property of the 

 particles of magnetic matter ; the former as dependent rather upon 

 the action of the mass: the latter is an absolute inversion of the di- 

 rection of the power, the former only a divergence or deflection of it. 



Applying the idea of conduction to niagnecrystallic bodies, he 

 concluded that the niagnecrystallic axis would coincide with the 

 direction of better conduction, and thence concluded, that, if a 

 symmetric crystal of bismuth were carefully examined in difl'erent 

 directions, it would be found to be less diamagnetic when its magne- 

 crystallic axis was perpendicular to the axis of magnetic force of the 

 lield in which it was to be submitted to experiment, than when it 

 was parallel to the magnetic axis. By means of the differential 

 torsion balance described in the former paper, he was able to make 

 the trial, and found the results were as anticipated. With calcareous 

 spar and his present balance he was not able to establish any dif- 

 ference, but concludes that it will prove most diamagnetic when the 

 optic axis of the crystal and the magnetic axis of the field are 

 parallel. 



Advancing to the consideration of atmospheric magnetism, the 

 author first refers to the earth as a source of magnetic power from 

 which emanate lines of magnetic force passing into space accord- 

 ing to a particular but recognized distribution, and in obedience 

 to the general laws which govern the distribution of power about a 

 given irregular magnet. In pure space the magnetic power is con- 

 sidered as transmitted onwards with a certain degree of facility 

 which is constant, but may be increased or diminished by the presence 

 of paramagnetic or di;imagnetic matter witiiin that space. The atmo- 

 sphere is a portion of such matter, andean af!'ect the magnetic lines 

 which pass from the earth into space, and aHccts them differently 



