Colour of Substances and their Magnetic Properties, 215 



use in the arts have furnished either oxides ^ carburets, sulphurets^ 

 or fluorides, which contain particles sujlciently magnetic to move 

 on paper to an ordinary steel magnet passed to and fro under-- 

 neath. 



Iron, nickel, and cobalt, are apart from all the other elementary sub- 

 stances for their ready motion on paper to a magnet. Manganese, pure, 

 and arsenic in the black or supposed partially-oxidized form, approach 

 them; but in so much lower degree, that they class better with metals 

 thatcorae near to them in magnetic attraction when tested on the torsion 

 balance. The distinctive magnetic property of these three metals 

 naturally introduces the question, in an inquiry like the present, Is 

 there anything peculiar in their colour ? To which I think it must be 

 answered, that in their solid form there is not ; for nickel ranks be- 

 tween silver and tin in colour, two metals of very low magnetic pro- 

 perties. But if the colours of the precipitates from the oxides of 

 these metals be looked to, a difference is observable, shewing the 

 more magnetic metal to produce a larger proportion of coloured pre- 

 cipitates; if five of the most important reagents be employed to 

 throw down precipitates from oxide of nickel, four of them are 

 coloured and one white ; while the two oxides of tin precipitated by 

 the same means, give four coloured and six white. Silver produces 

 three coloured and two white substances from its oxide. Referring 

 to a table of precipitates from the oxides of the heavy metals by the 

 reagents in use for testing,* the proportion of coloured to white, 

 taken in the aggregate, is very nearly as 2 to 1 ; for the magnetic 

 metal nickel, the same substances are as 4 to 1 ; and for the non- 

 magnetic metals silver and tin, as 7 to 8 ; thus shewing that a mag- 

 netic metal has a far stronger tendency to form coloured bodies than 

 one which has feeble magnetic properties. Iron and cobalt, in their 

 solid form, are not distinguished by any peculiar colour from less 

 magnetic metals, but in their pyrophoric form they supply the purest 

 blacks found among the metals. The proportion of their coloured 

 precipitates to white ones, is, like nickel, above the average, which 

 has been stated at two coloured to one white ; iron and cobalt give 

 nearly six coloured to one white ; and if these metals be compared 

 with mercury and platina, that with difficulty afford products suffi- 

 ciently magnetic to move on paper, the proportions for the precipi- 

 tates are 6 to 1 for the magnetic against 3 to 1 for the non or 

 feeble magnetic metals. Or if the comparison be made with the five 

 diamagnetic metals, antimony, bismuth, zinc, cadmium, tellurium, 

 the differences widen, for these yield thirteen white precipitates, and 

 only eight coloured; proving that, independently of my experiments, 



he data furnished from a published table of precipitates, shew the 



♦ Published by Maclachlan and Stewart, 1832. This table gives eight re- 

 agents, but three I treat as duplicates of three others, namely, sulphuretted 

 hydrogen and hydrosulphuret of ammonia, caustic potash and ammonia, car- 

 bonates of soda and ammonia. 



