862 



SCIENCE. 



[N. S. Vol. IV. No. 102. 



42. The gauze was next connected with 

 the pole of a battery (2 volts) , the plate being 

 at first grounded, so that it took a charge 

 by induction, and then connected to the 

 electrometer. When illuminated the plate 

 became negatively charged, i. e., lost elec- 

 tricity, (why?) But on blowing against 

 the plate the deflection of the electrometer 

 was reversed, and the plate became strongly 

 positive. Blowing on the plate does not 

 therefore merely take away the statical 

 charge, regardless of sign. 



43. If the gauze and plate are connected 

 through a galvanometer without any bat- 

 tery, no current results from illumination 

 alone. But when air is blown against the 

 plate (with illumination) a considerable 

 current flows. Without illumination no 

 effect can be noticed on blowing. 



The authors believe that the effects are 

 in part due to the fact that there is a con- 

 tact P. D. between metal and air, the lat- 

 ter being positive.* 



44. Later in the same year Bichatf was 

 led by other observations to believe that the 

 negative discharge was due to convection. 

 An apparatus similar to that of Eighi,J 

 which had been used in 1887 apparently 

 without knowledge of any effect from illumi- 

 nation,§ was found to rotate more strongly 

 under the influence of ultra-violet rays. 



45. The hypothesis of the negative dis- 

 charge being due to convection received the 

 strongest confirmation from observation by 

 Lenard and Wolf.| | Their results, however, 

 contradict E,ighi's view in some particulars. 

 In assuming the existence of convection 

 Eighi was of the opinion that the air parti- 

 cles became charged at the surface of the 

 body and were then repelled.^ This action 

 would continue until a certain positive po- 



*For fuller statement of explanation see article. 



tC. E. 107, p. 557. Beibl. 13, 39. 



tSee I 37. 



§C. E. 104, p. 1786. Beibl. 11, 716. 



II Weid. Ann. 37, p. 443, 1889. 



i C. E. 107, p. 559, 



tential was reached, whereupon the attrac- 

 tion between the positive body and the 

 negative particles would bring about a con- 

 dition of equilibrium. 



46. Lenard and Wolf urged in objection 

 to this view that it is impossible for the 

 particles of a gas to become charged ; dust 

 may receive a charge, but not gaseous mole- 

 cules.* They therefore think that the 

 cherge must be carried away by particles of 

 the body itself, these being shaken loose in 

 some way by the ultra-violet rays. They 

 experimented first with thin gold leaf, silver 

 and copper foil, etc., hoping to detect the 

 loss of particles by some change in the opti- 

 cal transmitting power. After an exposure 

 of some 50 hours to rays from an arc light 

 the surface was found to be roughened at 

 all points not screened from the rays. By 

 interposing obstacles sharp shadows were 

 cast, as shown by the roughening. ISTo 

 change could be noticed, however, by trans- 

 mitted light. Any object which is opaque 

 to ultra-violet rays (e. g., glass) was capable 

 of producing a shadow. 



47. Being convinced by these experi- 

 ments that particles were actually sent off 

 under the influence of ultra-violet rays, the 

 authors next attempted to observe the course 

 of the particles after leaving the body. As 

 a source of light the spark from an induc- 

 tion coil was usually employed. The arc 

 lamp was also used with Zn in place of 

 one carbon. The relative values of these 

 lights in producing the negative discharge 

 are shown by the observations described 

 below : 



A polished zinc disk 8 cm. in diameter 

 was set up 30 cm. from the source of light, 

 connected with a gold leaf electroscope, 

 and charged to 1000 volts. The time re- 

 quired for the potential to fall to 200 volts 

 was found to be 



* In this connection see Nahrwold, Wied. Ann. 31, 

 p. 448 ; also J. J. Thomson, Eecent Eesearches in 

 Electricity and Magnetism, p. 53. 



