Prof. E. Wartmann's third Memoir on Induction. 269 



current of Ampere. It goes therefore from north to south 

 by the west around the north pole, or that by which the cur- 

 rent of the battery enters, and from north to south by the east 

 around the south pole, or that which communicates with the 

 zinc of the Bunsen's apparatus (fig, 9). In general it does 

 not appear to be more rapid near the cylinders than at a cer- 

 tain distance. It is especially apparent between them, but 

 it also exists at a tolerable distance from the centres. I have 

 more than once seen the velocity vary in an intermittent man- 

 ner, as if it had to overcome passing obstacles. 



102. The theory of these rotations is easily suggested. 

 When we immerse a bar of iron in a solution of sulphate of 

 copper, the electro-chemical action determines in the liquid 

 an electric current, which proceeds from the peripheric parts 

 to those immediately around the cylinder, in the direction of 

 the prolonged radii of the latter. This may be convincingly 

 shown with a good rheometer, the extremities of which, of 

 thick well-polished platinum wire, are placed alternately, the 

 one near the iron, the other toward the margin of the vessel 

 which contains the solution. This current is analogous to one 

 passed into a metallic ring full of mercury, which would tend, 

 radiating towards the centre, to issue by a conductor placed 

 perpendicularly to the surface. An external horizontal cur- 

 rent, and near the vessel, would, according to its own direc- 

 tion, cause the mercury to rotate to the right or to the left. 

 In our experiment the magnet is substituted for the horizontal 

 current. 



103. A chemical action is therefore necessary, and proceed- 

 ing from currents of a certain intensity, for the double polary 

 rotation to take place. The liquids cited as not producing it 

 in a certain manner, evidently did not fulfill this condition : 

 they did not possess a requisite relation of energy between the 

 electric currents and the magnetization. I attribute the ob- 

 served variations in the force of the electric current, and con- 

 sequently of the velocity of rotation, to variations in the play 

 of the affinities between the iron, all the molecules of which 

 are not equally acted upon, and the liquid, the composition 

 of which changes incessantly. The reduced copper, even 

 upon polished cylinders, is always striated; it is not deposited 

 therefore in a uniform manner and without intermitlence. 

 The temperature, by influencing either the chemical action or 

 the viscosity of the liquid, is perhaps not wholly without in- 

 fluence upon the rapidity of the rotation. Moreover, it will 

 be conceived that this latter becomes visible only when it is 

 communicated to solid matter on which the light is reflected*, 



* It is evident, that in employing very intense chemical actions, this 



