864 



SCIENCE. 



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



showed themselves about ^ as sensitive as 

 zinc. It appears therefore that the liquids 

 which give a dust reaction are the same 

 that show an increased rapidity of dis- 

 charge of negative electricity. 



54. Further confirmation of the hypothe- 

 sis of convection is given by the observa- 

 tions of Hoor.* His article contains also a 

 discussion of some of the explanations of 

 the phenomena that might be suggested : 

 e. g., (1) Electrostatic action from the arc 

 light ; (2) material particles sent out by the 

 arc ; (3) change in the conductivity of the 

 medium (air) in which the body is placed ; 

 (4) change in the surface of the illuminated 

 conductor by a transformation of light 

 energy into electrical energy; (5) convec- 

 tion, under the influence of illumination. 

 Rather obvious experimental reasons are 

 given for discarding 1, 2,f and 3, and Hoor 

 prefers to adopt the 5th. He is, however, 

 inclined to accept the suggestion of Wiede- 

 mann and Ebert % that the layer of gas con- 

 densed upon the surface plays an impor- 

 tant part in the phenomenon. Particles of 

 gas Irom this layer he thinks become 

 charged, and, under the influence of ultra- 

 violet rays, escape. Several experiments 

 seem to show that any influence which 

 tends to remove the layer of condensed 

 gas diminishes the sensitiveness of the 

 surface. 



55. For example,§ a plate of either glass 

 or gypsum laid upon the surface of a zinc 

 disk prevented the dissipation of a negative 

 charge by illumination. (Convection pre- 

 vented.) The zinc plate remained inactive 

 for a short time after the removal of the 



*Repertoriuin der Physik 25, p. 91, 1889. Beibl. 

 13, 731. 



tHallwachs shows, however [Wied. Ann. 40, p. 

 332] , that charged particles from the arc may cause 

 trouble in some cases and that suitable precautions 

 must be used. 



JThis is practically the view first held by Righi. 



I First paper, already referred to. Wied. Ann. 33, 

 p. 240. 



glass,* but became sensitive again either on 

 resting 5 or 10 minutes, or by repolishing. 

 (Surface gas layer removed by contact with 

 glass, etc., on account of the greater attrac- 

 tion of the latter. A new layer condenses 

 gradually.) (3.) A clean zinc plate was 

 found to lose its sensitiveness after being 

 covered for 3 hours with powdered char- 

 coal. (Condensed gases absorbed by the 

 charcoal.) (4.) A plate is rendered in- 

 active by heating to 55° with a Bunsen 

 burner,f but becomes sensitive again by 

 cooling. (Gas layer dissipated by flame.) 

 (5.) A plate covered with a thin film of 

 glycerine is insensitive. (6.) After a plate 

 has been illuminated for some time its sen- 

 sitiveness diminishes, but increases again 

 after a rest. 



56. Righi at first held the same view as 

 Wiedemann and Ebert and Hoor in regard 

 to the importance of the layer of condensed 

 gas.;}: Later he came to the conclusion 

 that the condensed gases were without 

 influence. § He detected the alteration 

 (roughening) of the surface under the ac- 

 tion of ultra-violet rays independently of 

 Lenard and Wolf, and investigated the 

 shadows and figures produced at some 

 length. 1 1 In some cases the shadows cast 

 by opaque bodies can be better shown by 

 breathing on the surface. 



57. Summing up the results in regard to 

 convection, it appears that almost all ob- 

 servers agree in the belief that the nega- 

 tive charge is removed by convection of 



*Stoletow [C. E. 108, p. 1241. Beibl. 13, 902] 

 contradicts this cibsolutely. He finds no diminution 

 in sensitiveness even after the glass plate has re- 

 mained in place 24 hours, provided that both glass 

 and zinc are clean and dry. 



fThis is also contradicted by Stoletow (1. c. ), who, 

 however, used an air bath instead of a Bunsen burner. 



X Accad. dei Lincei 5, p. 331. Beibl. 13, 566, and 

 14, 68. 



I Atti del Inst. Ven. 7, 1889. Beibl. 13, 976. Also 

 Exner's Eepertorium 25, p. 380. 



II Beibl. 13, 566; 13, 976; 14, 68. 



