Effect of Electrical Convection. 187 



remained on the condansing-plates, so that the charge taken 

 by the caoutchouc surface was less than that on the surface of 

 the disk. The slight deflexion observed was then really due 

 to the difference in magnetic action of these two surfaces. 

 It is also seen why a deflexion of the magnitude expected 

 was not obtained. 



If Cremieu's conclusions are true, then " open currents " 

 can exist. Cremieu has tested this deduction by the following 

 experiment*: — An ebonite disk, 37 cm. in diajneter and 

 2*5 mm. thick, was gilded (presumably on one side, if one 

 may judge from the diagram given) in 25 sectors, the distance 

 between sectors being 1 cm. This disk turned about a 

 horizontal axis between two fixed ebonite plates. One of 

 these plates (presumably the one facing the side not gilded) 

 carried a sector of tinfoil twice as wide as the sectors on the 

 disk. On the second plate, opposite this sector, was fixed a 

 brush which made contact with the sectors on the disk when 

 immediately under the fixed sector. This brush was also 

 connected to a second brush on the same plate, about 60° 

 ahead of the first brush. According to Cremieu, if the fixed 

 sector is connected to a source of electricity, and the disk set 

 rotating, each sector, as it comes under the fixed sector, will 

 receive a charge by induction ; when this sector makes con- 

 tact with the second brush, the charge on it will be neutral- 

 ized. There will then be a conduction-current flowing in the 

 wire connecting the brushes, which circuit is completed by the 

 charge carried convectively between the two brushes by the 

 rotating disk. An astatic system was suspended with its 

 lower needle near the edge of the disk between the two 

 brushes. The conduction-circuit between the two bruslio 

 contained a galvanometer. The fixed sector was charged to 

 a potential of 100 to 130 C.G.s. electrostatic units, its distance 

 from the rotating disk was *5 cm. When the disk was set 

 rotating the galvanometer indicated a current of from 10 -4 to 

 2 x 10 -4 amperes. The astatic system, however, remained at 

 rest ; though a conduction-current of 10 -4 amperes in a 

 circuit occupying the mean effective position of the charged 

 disk caused a deflexion of 15 mm. on a scale 4 metres 

 distant. Cremieu therefore concludes that the electricity 

 carried convectively between the two brushes is without 

 magnetic effect ; i. e. that the current in the wire connecting 

 the two brushes is an " open current/" 



There is, however, a possible explanation of this experiment 

 which requires no such conclusion. As was stated above, the 

 * C. Ji. cxxxii. p. 1108. 



