Effect of Electrical Convection. 183 



to the speed of rotation, and for the disk at potentials of from 

 400 to 4000 volts, was proportional to the surface- density of 

 the charge on the disk Above 4000 volts this latter pro- 

 portionality ceased to exist, the deflexion remaining approxi- 

 mately constant up to the highest potential tried — 14,000 volts, 

 (The distance between disks and condensing-plates is not 

 given.) Himstedt explains this by assuming that there is a 

 limit to the amount of charge which a moving body can carry 

 with it. For these high potentials, however* the leakage was 

 considerable; in one minute the potential dropped from 14,000 

 to 40(40 volts. Any theoretical conclusion of the above nature 

 must therefore be accepted with reserve. 



So stood the question as to the existence of a magnetic 

 action due to electric convection in 1897, when Cremieu 

 began his experiments. The first problem* attacked was 

 the reverse of the above experiments, viz., the detection of 

 the mechanical force exerted on a charged body by a rapidly 

 changing magnetic field. This quantity is extremely small, 

 varying as the reciprocal of the square of the ratio of the 

 two systems of electrical units. Cremieu attempted to detect 

 this force by the rotation of a charged aluminium disk sus- 

 pended between the poles of a strong electromagnet* No 

 deflexion of the disk was observed, though the calculated 

 deflexion, as read by means of a telescope and mirror on «. 

 scale 1°1 m. distant, was 100 mm. This is in direct opposition 

 to Lodge's experiment f on the same subject performed in 

 1889, though Lodge's experiment is not altogether satisfactory, 

 as the observed deflexion was only from 2 to 3 mm. Cremieu 

 has published only a short note on this experiment, with no 

 details of the apparatus, so that it is impossible to discuss 

 the work further. 



The negative results obtained from the reverse experiment 

 !<*d Cremieu to investigate once more the magnetic action of 

 n moving electrified body J. A very ingenious method, dis*- 

 tinctly different from any of those previously employed, 

 was used. A solid aluminium disk, 37 cm. in diameter, was 

 mounted on an insulating hub of hard rubber turning on a 

 horizontal axis. Concentric with this disk was placed a coil 

 of 13,000 turns of copper wire '15 mm. in diameter, connected 

 in series with a sensitive galvanometer. The diameter of 

 the coil was 44 em. The coil was inclosed in a brass case. 

 The disk revolved between two thick iron condensing-plates. 

 These plates also served to decrease greatly the reluctance of 



* C. B. exxxi. p. 578 (1900). 



t Phil. Mag. [5] xxvii. p. 469 (1889). 



X C. R. exxx. p. 1544 (1900). 



