Effect of Electrical Convection. 185 



seems by no means improbable thnt there may have been a 

 brushing over of the charge from the condensing-plates to the 

 caoutchouc surface. The magnetic action of this charge 

 would then just neutralize that of the charge on the disk. 

 This same objection applies to the third series of experiments, 

 in which was used an ebonite disk 2 mm. thick, gilded on 

 both sides in three sectors and covered with caoutchouc; and 

 to the fourth series, in which the iron condensing-plates were 

 covered with sheets of mica, over which was pasted tinfoil in 

 six sectors, covered with a thin layer of caoutchouc. (The 

 object of the third and fourth series was to see if the negative 

 results were due (J) to the slipping of the charge with respect 

 to the disk, or (2) to the movement of the charge induced 

 on the condensing-plates with the inducing charge on the 

 disk.) 



During the past winter Cremieu has continued his investiga- 

 tions*, replacing the coil and galvanometer by an astatic 

 needle, and again with negative results. The rotating appa- 

 ratus consisted of the same ebonite disk mounted in a bronze 

 frame of the same dimensions as the iron frame previously 

 used. To this frame were screwed brass plates, covered with 

 mica, and over the latter was pasted tinfoil in six sectors, 

 connected to the earth. The astatic needle was placed out- 

 side of these, with the lower needle opposite the upper edge 

 of the disk. To this experiment applies the same objection 

 as to those previously performed, viz., the rotating disk 

 probably did not keep its charge. It is interesting to note 

 that in a second series of experiments with this apparatus, 

 in which the tinfoil was mounted on ebonite plates, and these 

 screwed directly to the bronze frame, so that between the 

 needle and disk there was only one metal plate, a deflexion 

 was obtained, in the proper direction and reversible with the 

 sign of the electrification of the disk and the direction of 

 rotation. This Cremieu states was suppressed by the intro- 

 duction of a brass plate between the needle and the 

 condensing -plates. Without a more detailed account of the 

 exact conditions und^r which this effect was obtained than 

 is given by Cremieu, it is difficult to say exacily what this 

 means. 



Believing that the deflexion observed by previous investi- 

 gators might have been due in some way to the condensing- 

 plates, a different arrangement was then employed f. The 

 same ebonite disk, with the three sectors completely insulated 

 from one another, was again used, but the condensing-plates 

 * C. R. cxxxi. p. 797 (1900). 

 + Ibid, cxxxii. p. 327 (1901). 



