210 Richard Adie, Esq., on Che Connection between the 



particles in motion; and in several instances the powder can be 

 streaked, occasioned by the finer particles being little disturbed, 

 while those of a certain size follow for a time the magnet, and thus 

 produce this appearance. I at 6rst expected that the test of motion 

 on paper would belong to a comparatively small number of bodies 

 attracted by the magnet, but, since by varying the methods of pre- 

 paration, I have succeeded in obtaining magnetic compounds from 

 bodies I had no hope of doing so, the number appears to admit of a 

 considerable increase. Bismuth, zinc, antimony, copper, all furnish 

 compounds to move on paper to a magnet, and yet these metals are 

 the most remarkable for their diamagnetic or non-magnetic pro- 

 perties. Among bodies repelled by the magnet, I have met no in- 

 stance of a sufficiently intense force which would move the substance 

 on paper by the kind of magnet described, so that this test must be 

 understood to apply only to bodies attracted by the magnet. 



Iron. — The compounds of this metal form a series that afford 

 evidence of the existence of a connection between the colour and the 

 magnetic properties of bodies. In the colour of the solid metal itself 

 there is nothing to distinguish it from many metals of feeble mag- 

 netic powers, but in the finely subdivided pyrophoric state it has 

 no equal for a black among the elementaries, save the non-metallic 

 substance carbon. The fluorides of iron are pale-coloured powders 

 possessed of feeble magnetic attraction, insufficient to make their 

 particles move on paper ; when heated with borax, they pass to a 

 dark brown substance, containing oxides and fluorides of iron, which 

 are strongly magnetic. The lactate of iron is a pale-coloured feebly- 

 magnetic body; after ignition, the dark-coloured mixed oxides are 

 left, which are well known for their magnetism. 



The ter-sulphate of iron, and the alums formed with it, furnish 

 three puce light-coloured bodies, quite magnetic on the torsion balance, 

 but not enough so to move Avhen pulverised on paper ; by heat the 

 dark oxides are produced, that move readily to the magnet. But 

 it is to the ferro-cyanide of potassium that I wish to refer for the 

 best proof we have of a connection between the colour and the 

 magnetic properties. If it were a solitary case the change would at 

 once be set down to the decomposition of the cyanogen and the 

 formation of a carburet of iron. In the sequel it will be seen that 

 facts of a similar kind occur under such varied circumstances, that 

 this explanation will not suffice. The ferro-cyanide of potassium 

 is a translucent lemon-yellow salt, possessed of no attraction for the 

 magnet ; when heated moderately, it loses water and assumes an 

 opaque white hue, but is still unattracted by the magnet ; when the 

 heating is continued till the colour darkens, then the degree of dark- 

 ness becomes an index of the magnetic force, until the colour reaches 

 black, when the altered salt has all the characters of an iron body. 



Nickel. --"The salts of this metal are generally coloured, and the 

 magnetic properties of their base are seen reduced in a greater or less 



