144 Royal Society : — 



current generated in it by its movement. It was noticed that the 

 plates were 'polarized by the currents so created. As a consequence 

 of this, the deflection of the needle soon fell to zero ; and on the 

 liquid flow being stopped, a polarization current in the opposite 

 direction was obtained. This proved that in experiments on induc- 

 tion in liquids, in order to obtain any constant current, non-polar- 

 izable electrodes must be used. 



2. Production of induced current in a mass of liquid rotating over 

 a magnetic pole. — In this case radial currents should be produced. 

 They were obtained as follows : — Flat porous cells were placed 

 round the circumference of a large basin, and in the centre a cylin- 

 drical one. These were filled with a solution of cupric sulphate, 

 and contained copper plates. The basin was filled with dilute sul- 

 phuric acid. The centre copper plate was connected with one 

 pole of the galvanometer, and the circumferential ones with the 

 other. The whole was placed over the pole of the electromagnet. 

 On exciting the magnet and rotating the dilute acid, a constant 

 current w^as obtained, flowing from centre to circumference or the 

 reverse according to the direction of rotation. With platinum 

 electrodes the effect cannot be obtained, but with non-polarizable 

 electrodes it is easily produced. Mercury was likewise tried with 

 still better results. 



3. Production of induced current in a liquid at rest in a variable 

 magnetic field. — If a flexible tube filled with conducting liquid is 

 wound round an electromagnet, and into the ends electrodes placed 

 so as to include a galvanometer in the circuit, then induced currents 

 are obtained whenever the strength of the magnet varies. This is 

 the case examined by Faraday. His experiment was repeated 

 by MM. Logeman and Van Breda (Phil. Mag. [lY.] vol. viii. p. 465), 

 who noticed that the electrodes were left polarized after the induced 

 current had passed. These experiments were repeated with more 

 powerful apparatus, using a soft-iron wire core within the inducing 

 helix instead of solid iron, and employing a condenser in the 

 primary circuit. Very strong induced currents were obtained, and 

 correspondingly great polarization of the electrodes placed in the 

 ends of the coil of acid. It was hoped that the currents might 

 produce visible inductive electrolysis ; but even this improved 

 arrangement did not yield that result. Other saline solutions 

 were tried with similar results. 



Lastly, the phenomenon observed by Arago, of the retardation 

 in the vibrations of a magnetic needle oscillated near the surface 

 of liquids, is examined. Evidence is brought forward to show that 

 this is not, as in the case of solid plates, due to induced currents 

 created in the liquid : — (1) because the retardation is, cMeris paribus^ 

 not proportional to the conductivity of the liquid but dependent 

 on its volatility ; (2) because it takes place equally when a Ught 

 brass needle, oscillated by torsion, is substituted for the magnet, 

 provided the needle is light and the period of oscillation not very 

 small. Thus a magnetic needle which required 4 min. 20 sec. 

 to suffer a decrement of 25° in the semi-arc of vibration when in 



