892 



SCIENCE. 



[N. S. Vol. IV. No. 103. 



wachs, Eighi, etc. The difference in tlie 

 nature of the discharge may be a sufficient 

 explanation of apparent contradictions. 



71. In discussing the various observa- 

 tions upon the phenomenon in question, E. 

 Wiedemann * is inclined to favor the view 

 that the effect is due to the absorption of 

 ultra-violet rays. He points out that the ab- 

 sorption of rays by resonance brings the ab- 

 sorbing molecules temporarily into the same 

 condition of vibration which they would have 

 if heated sufficiently to send out the same 

 kind of rays as those absorbed, e. g., fluo- 

 rescence. The rapid and violent vibrations 

 produced in the surface layers of a body by 

 ultra-violet absorption will produce by col- 

 lisions increased velocity of translation 

 among the molecules, ^. e., a rise of tem- 

 perature. Although the actual tempera- 

 ture rise will be small, yet individual mole- 

 cules may receive very high velocities, and, 

 being close to the surface, may escape. 

 Convection discharge will therefore be ac- 

 celerated by the action of ultra-violet light. 

 ISTote that this explanation fails to give 

 any reason for the selective discharge of 

 negative electricity, unless there is some 

 reason why the particles take on a negative 

 charge by preference. It also fails to ex- 

 plain the fact that the action of ultra-violet 

 rays is much more intense when COj is 

 used instead of air. f 



PHOTO-ELBCTRIO EFFECT OF SUNLIGHT. BE- 

 HAVIOR OF SODIUM AND POTASSIUM. 



72. It will be remembered that all of the 

 early observers failed to detect any action 



* Wied. Ann. 35, p. 257. 



t Wiedemann and Ebert. Wied. Ann. 33, p. 240. 

 This result, first obtained for spark discharges, was 

 found to be also true in the case of a continuous dis- 

 charge. See Stoletow, C. R. 107, p. 91. Beibl. 12, 

 723. Observations by Breisig, in which zinc was 

 used as the metal and a lamp flame as the source of 

 illumination, fail to confirm Stoletow's observations. 

 [ See Beibl. 17, 60, 1891. ] But on the other hand, 

 Breisig's results are contradicted by Cantor. [ Beibl. 

 19, 583. ] 



produced by sunlight, either on the spark 

 discharge or upon the continuous dissipa- 

 tion of negative electricity. In 1889, how- 

 ever, Hoor claims to have detected some 

 slight effect.* During the same year Nodonf 

 found that insulated conductors of carbon 

 or metal became positively charged when 

 illuminated by sunlight. And almost 

 simultaneous with Nodon, Elster and 

 Geitel J observed both the discharge of 

 negative electricity and the development 

 of a positive charge. A freshly polished 

 zinc plate was found to acquire a positive 

 potential of 2.5 volts. Even diffused day- 

 light produced a measurable effect. Mg 

 and Al were found to be more sensitive 

 than Zn. 



73. Based upon this action of sunlight 

 upon the atmosphere and the surface of the 

 earth, Ai-rheniusl developed in the same 

 year an interesting theory to explain atmos- 

 pheric electricity, the aurora, etc. 



74. The first article of Elster and Geitel 

 is merely a brief note, hurriedly published, 

 probably in order to secure priority. Two 

 months later a full description of their ex- 

 periments appear in the Annalen.|| The 

 first results were obtained with zinc. A 

 shallow dish 20 cm. in diameter was set up 

 (insulated) in the open air, and was con- 

 nected with a sensitive electrometer in an 

 adjacent building. The dish was sur- 

 rounded by a metal case, the top of which 

 could be removed by means of a cord run- 

 ning into the laboratory. Above the sur- 



*Exner's Eepertorium, 25, p. 91; Beibl. 13, 731. 

 Eighi, however, (Exner's Eepertorium 25, p. 380. 

 Beibl. 14, 68) contradicts the results of Hoor, and 

 states that he has ' never succeeded in obtaining the 

 slightest trace of an effect from sunlight. ' Elster and 

 Geitel (Wied. Ann. 38, p. 498) were also unable to 

 obtain any action with the substances used by Hoor. 



t C. E. 109, p. 219. Beibl. 13, 976. 



t Wied. Ann. 38, p. 40. Phil. Mag. 28, p. 427. 

 (Abst.) 



§ Meteorologische Zeitsch. 5, pp. 297 and 348. 

 Beibl. 13, 328. 



II Wied. Ann. 38, p. 497. 



