Colour of Substances and their Magnetic Properties. 211 



degree in them. The oxalate of nickel is a pure pale-coloured body, 

 magnetic only on the torsion balance ; when heated to redness it 

 passes to a dark colour, and the new compound moves readily on 

 paper to a magnet underneath. 



Manganese. — Proved by Berthier to be a magnetic metal, which 

 an examination of its salts on the torsion balance tends strongly to 

 confirm, for there they resemble those of iron and nickel ; the 

 colourless sulphate of manganese is as much attracted by the magnet 

 as the transparent iron alum, but both are inferior to several dark- 

 coloured compounds of non-magnetic metals. Carbonate of man- 

 ganese is a pale-coloured substance which cannot be shewn to be 

 magnetic without the torsion balance ; by heat it is changed into a 

 brown-tinted oxide which, spread out on paper, moves feebly to a 

 magnet underneath. Arsenic in its pure state is magnetic on the 

 torsion balance, but when pulverised is not disturbed by a magnet. 

 In union with its equivalent of sulphur the ruby-red realgar is pro*- 

 duced. Sulphur has been shewn by Faraday to be diamagnetic, yet 

 in this union with arsenic a compound is given that is far more 

 magnetic than the metal or any other of its products, for realgar 

 when bruised and spread out on paper is moved and can be streaked 

 by a magnet. In most of the foregoing cases given, the dark 

 magnetic substance is an oxide formed or set free by heating ; with 

 this metal oxygen forms the well known white oxide, a diamagnetic, 

 according to Faraday, but which, with the feebler magnetic force 

 from a small steel magnet, proves slightly magnetic ; however this 

 may be, the contrast is striking between the feeble magnetism of the 

 white oxide and the decided magnetism of realgar. It was this con- 

 trast that led me to try to trace the connection between colour and 

 magnetic properties, which, on account of the near relation of the 

 colour of iron to many other non-magnetic metals I felt a great re- 

 luctance to do. 



Silver is a metal possessed of only feeble torsion balance 

 magnetism ; in its dark-coloured sulphuret and oxide the attraction 

 is more decided ; the white chloride is like the pure metal, very 

 feeble in its attraction. Through the influence of light it passes to 

 the dark oxide, gaining magnetic force with the change. Should 

 these experiments be confirmed, this well known action of light on 

 the chloride of silver will serve as a very beautiful illustration, for 

 in it the whole question is resolved ; the light effects a change in the 

 colour ; the alteration of the colour is accompanied by an increase of 

 magnetic attraction. 



Palladium^ feebly magnetic on the torsion balance, furnishes a 

 dark-coloured oxide, which, if well dried, moves to the magnet on 

 paper, while the sulphuret, a similar body, has not sufficient mag- 

 netism to do so. 



Copper. — This metal possesses a degree of magnetism so minute 

 that it is recognised with the utmost difficulty on the torsion balance, 



02 



