252 Mr. S. Bidwell's Experiments illustrating 



thirteen different metals which he has tested, three (iron, 

 cobalt, and zinc) are classed as positive ; nine (gold, silver, 

 tin, copper, brass, platinum, nickel, aluminium, and magne- 

 sium) are negative ; and one (lead) shows no signs of 

 rotation. 



I began a repetition of Hall's experiments with the object, 

 if possible, of carrying them further, and especially of estab- 

 lishing in a satisfactory manner the existence of a connexion 

 between the phenomena in question and the magnetic rotation 

 of the plane of polarized light. My earlier experiments were 

 made with iron, of which I happened to possess a suitable 

 specimen in the form of foil *019 millim. thick. The magnet 

 used consisted of a bar of soft iron, 38 millim. in diameter, 

 bent into horseshoe shape ; if pulled out straight its length 

 would be about 50 centim. It was covered with about 600 

 turns of wire, 2 millim. in diameter; the pole-pieces used were 

 30 millim. square, with flat ends, and were generally fixed 

 about 6 millim. apart. The magnet was excited by a battery 

 of either seven Grove's or, more generally, of five 3-quart 

 bichromate cells. 



The arrangement of the apparatus is shown in fig. 1 

 (Plate XI.), where P P' is the strip of metal cemented to a 

 plate of glass, and placed above one pole, S, of the electro- 

 magnet ; M is a section of the yoke of the electromagnet ; 

 B is a battery of one or two cells, generating the current 

 through the strip ; C a commutator to reverse the direction 

 of the current ; B x the electromagnet battery ; (7 a commu- 

 tator to reverse the magnetism ; W W wires from equipoten- 

 tial points in the metallic strip to the galvanometer G ; K a 

 key in the galvanometer circuit. 



The method of proceeding was as follows : — A current was 

 passed from B through the strip P P ; , and sliding-contacts on 

 the opposite edges were adjusted until, when the key K was 

 depressed, no deflection of the galvanometer occurred. The 

 magnet was then excited by the battery B', and on again de- 

 pressing K the galvanometer was deflected. On reversing 

 the direction of the current either from B or B', by means of 

 one of the commutators, an opposite deflection occurred when 

 K was depressed. The galvanometer used was a reflecting 

 instrument, of *2 ohms resistance; it was placed at a distance 

 of 4 metres from the scale, upon which it reflected the image 

 of a very fine wire. The scale, key, and commutators were 

 all within easy reach of the operator. 



With this comparatively small apparatus there was at 

 first some difficulty in producing a sensible Hall effect, and 

 the greatest care had to be exercised to avoid disturbing 



