Decembee 11, 1896.] 



SCIENCE. 



861 



ting this sheet (through a gypsum win- 

 dow) with rays from an arc light, the 

 system was deflected. The charged sheet 

 seemed to be driven away from the light 

 rays. This may be explained as due to the 

 reaction from the electrified particles, which, 

 under the influence of the ultraviolet rays, 

 are being driven away from the surface. 

 When the illuminated sheet was charged 

 positively, or grounded, no effect was noticed. 



38. Returning to the apparatus first used 

 (metal disk and parallel wire gauze) an in- 

 sulated plate of gypsum was placed be- 

 tween disk and gauze, and was observed to 

 become negatively charged on the side 

 nearest the Zn plate. If two plates of gyp- 

 sum are used, only that nearest the Zn plate 

 is charged. These results are well explained 

 on the hypothesis of electric convection. 



39. It was found that the charged parti- 

 cles producing the convection discharge 

 travel along the lines of force of the field. 

 (Suggested first by Hallwachs.) To prove 

 this a vertical zinc cylinder was charged by 

 a dry battery and placed near a large verti- 

 cal zinc plate, the latter being grounded. 

 The shape of the lines of force of such a 

 system is known. With the exception of a 

 narrow vertical strip the cylinder was lac- 

 quered, previous experiments having shown 

 that a coating of lacquer prevents all ac- 

 tion from ultra-violet rays. On illumina- 

 tion the convection discharge, if there is 

 one, could therefore only take place from 

 the unlacquered portion of the surface. If 

 the electrified particles traveled along the 

 lines of force it would be possible to predict 

 the position at which they would reach the 

 grounded plate. This was done, and a small 

 insulated piece of zinc placed at the position 

 computed was found to become negatively 

 charged, while if at a different point it was 

 unaffected. 



40. In 1890 Eighi* began experiments 



*Acc. del Lincei 6, p. 81. Ace. di Bologna 10, p. 

 85. Beibl. 14, p. 1167. 



upon ' photo-electric convention,' etc. at 

 low air pressures (going as low as .001 mm.). 

 It was found that at ordinary pressures the 

 electrified particles proceeded along lines of 

 force, as shown earlier. But as the pres- 

 sure was reduced there was a tendency for 

 the paths of the particles to deviate more 

 and more from the lines of force and to 

 become more nearly coincident with the 

 normal to the illuminated surface. In the 

 course of these experiments it was found 

 that the maximum positive surface density 

 acquired by an unelectrified body when il- 

 luminated increases continuously with di- 

 minishing pressure. On the other hand, 

 the rate of dissipation from a negatively 

 charged surface increases to a maximum as 

 the pressure is reduced and then dimin- 

 ishes. Righi concludes that the develop- 

 ment of a positive charge, and the dissipa- 

 tion of a negative one, are different phe- 

 nomena. 



Numerous experiments by Stoletow* on 

 the effect of diminished pressure upon the 

 actino-electric current can merely be re- 

 ferred to. They seem in the main to con- 

 firm Righi's observations. 



41. Observations by Bichat and Blondlotf 

 appear at first to be in confirmation of the 

 hypothesis of Righi that the discharge 

 takes place by convection. These observers 

 used an apparatus like that of Righi, gauze 

 and plate being of the same metal. On il- 

 lumination the plate became positively 

 charged to a potential of 3 to 4 volts. If a 

 draft of air is blown upon the plate, either 

 by fanning or by allowing compressed air 

 to escape against the plate, the potential 

 was increased six or eight times. It would 

 seem that this might be explained by con- 

 vection taking place more readily under 

 the action of the draft. Further experi- 

 ments show, however, that the explanation 

 is not so simple. 



* Journ de Phys. 9, 468. Beibl. 15, 233. 

 t C. R. 107, p. 29, 1888. Beibl. 13, 38. 



