330 Intelligence and Miscellaneous Articles. 



V r^ r'-i ) 



m being the intensity of the magnetism ; 

 i tliat of the current ; 

 ds the element of the moveable current nearest to the magnet ; 

 r and r' the radii vectores drawn from this element to the southern 



and northern poles ; 

 and 0' the angles of these radii with the axis of the magnet on 



the side of the south pole. 

 Taking ds with the sign + for the centripetal current, and with 

 the sign — for the centrifugal current, the sign of the force is in all 

 cases the sign of the rotation. This rotation becomes null if 

 cos 6 cos d' 



The neutral line constructed by starting from this equation does 

 not differ notably from that traced experimentally. 



4. The same law is applicable to electro-magnets, with this dif- 

 ference, that the neutral line starts from the margins of the electro- 

 magnet, as though the poles were at the extremities. 



5. Electro-dynamic coils behave like weak electro-magnets, 

 with this peculiarity, that their action is in the opposite direction 

 internally and externally. 



6. From this it results that a hollow coil and a hollow magnet do 

 not produce the same effect except upon currents wliicli arc exterior 

 to them ; their actions upon the interior ' currents are of opposite 

 signs. To explain this difference, the magnets must be assimilated, 

 not to coils, but to bundles of solenoids. 



7. The original ex])eriment of Davy upon the rotation of mercury, 

 the only one of this kind cited in our treatises on physics, may be 

 easily repeated when the mercury is replaced by an aqueous solu- 

 tion. It is explained by the preceding laws. 



8. When, in experiments upon magnets, the vertical magnet is 

 replaced by a horizontal one, we observe other phasnomcna, of which 

 the laws are furnished by the following theoretical formula. 



-mids . y( ,- ), 



y being the vertical distance from the magnet to the level of the 

 liquid ; regarding it as positive when the liquid is above the mag- 

 net, and retaining the jireceding values, the sign -\- of the forco 

 indicates a rotation from left to right, and the sign — an o])positc 

 rotation. The formula is always in agreement with experiment : 

 both show two neutral planes, the one corresponding with y = 0, and 

 the other with ?•=?•'. 



9. We may also submit to magnetic action currents which would 

 move vertically in a liquid. We then find a new series of i3lia?no- 

 mena, which may be easily referred to the two preceding formula;. 



To sum up, the electro-magnetic rotations of liquids are pheno- 

 mena which it is easy to produce, curious to observe, and interesting 

 to study in a theoretical point of view ; they consequently deserve a 

 place amongst the classical experiments in electro-magnetism. — 

 Comptes Rendus, Kn^xxfit 16, 1858, p. 307. 



