Prof. E. Wartmann's third Memoir on Lidticlion. 265 



repeat theii* researches, being persuaded that a very intense 

 Ibrce must produce corresponding decisive and considerable 

 effects. Other circumstances have thrown doubt on the con- 

 clusions mentioned by those authors ; account has not been 

 taken of the relations between the surfaces which were the 

 seat of the action to be measured, nor of the differences of 

 temperature, of density, and of composition of the liquids em- 

 ployed. 1 have endeavoured to guard against these causes of 

 error, and to obtain conclusive results by means of numerous 

 trials of sufficient duration. 



95. I procured cylinders of soft iron, taken from the same 

 bar, and about 0^027 in height and 0"''02 in breadth. Four 

 tumblers were so arranged that the plane of their axes co- 

 incided with that of the magnetic meridian and four others 

 arranged in a perpendicular direction. After having poured 

 into each of them a solution of sulphate of copper in layers of 

 equal depth, an iron cylinder was placed in the centre ; and 

 the pole of an iron horseshoe magnet was then set on the 

 upper unsubmerged face. The four magnets, the strongest of 

 which carried about forty kilogrms, were distributed, one N.S., 

 the second S.N., the third E.W,, and the fourth W.E., with 

 relation to the north pole of the terrestrial magnetism ; they 

 were separated, so as not to exert any reciprocal influence. 

 At the end of fifteen hours the deposits of copper on each 

 cylinder presented everywhere the same appearance and the 

 same consistence. The balance proved that they were all 

 nearly of an equal weight; the slight differences found, which 

 scarcely attained to one- or two-thousandths of the total quan- 

 tity of copper reduced, are explained by the inequality of 

 development and cleanness of the surfaces of the eight cylin- 

 ders. The experiment was repeated a great number of times 

 with solutions of copper more or less pure, and more or less 

 concentrated, without the general result varying. 



96. I will further cite the following experiment, which led 

 to the same conclusions. Twelve iron cylinders were distri- 

 buted in pairs, in six distinct vessels containing a solution of 

 iron-alum as neutral as possible. The pairs of cylinders of 

 the second, third, fourth and fifth vessels, were placed in con- 

 tact with the poles of four electro-magnets, the first of which 

 lias been already described (86) ; the second is capable of car- 

 rying sixty kilogrms ; the third at least forty, and the last more 

 than 280. They were in other respects arranged as in the 

 cases before described (95), and their poles were covered with a 

 plate of mica. Lastly, the current of a battery of forty Bunsen's 

 pairs was passed, so as to decompose the liquid by passing 

 across the interval between each couple of cylinders, at the 



