156 Drs. E. Pohl and P. Pringsheim on the 



factor of proportionality was much larger ; hence the ratio 

 of the photo-effect in the two main planes of polarization 

 w^as found to be sometimes as high as 1 : 50. This has been 

 confirmed by different authors. 



6. Finally, Pohl* found that these observations of Elster 

 and Geitel are true only in the visible part of the spectrum, 

 while in the ultraviolet the alkali metals behave like the 

 others, the orientation of the electric vector being 

 immaterial. 



Now we are going to show how the coexistence of these 

 different facts, though they seem to be absolutely incom- 

 patible, may be easily explained by assuming that there is, 

 apart from the normal photoelectric effect, the only one 

 known as yet, still a new selective one. The full report on 

 this subject has been published in the Verhandlungen der 

 Deutschen Physi'kalisclien Gesellschaft, vol. xii.f, and all details 

 may be found there ; the numbers in [ correspond with 

 the divisions of these papers. 



We used a spectroscope with quartz-fluorite lenses, illumi- 

 nated by a mercury arc-lamp ; the energy of the single lines 

 was measured by means of a Rubens thermocouple, and the 

 light was thrown through a quartz polarizer and a window 

 of fused quartz into a tube of the type described by Elster 

 and Geitel, which contained the alkali metal or some alloy. 

 By the aid of this apparatus the photoelectric current could 

 be measured, which was produced by the same energy o£ 

 incident light in the two main planes of polarization and in 

 different parts of the spectrum. 



Table I. and fig. 1 show the results for an alloy of mercury 

 and potassium, containing 2*5 atomic percentages of the 

 alkali [11]. These numbers and the corresponding ones for 

 Na prove that at an incidence of 60° the emission of electrons 

 increases continuously with decreasing wave-length, quite 

 independently of the plane of polarization. The effect is 

 larger if the electric vector swings parallel to the plane of 

 incidence (E ||) than for E J_, in consequence of the greater 

 absorption of light which follows from the optical constants 

 [12]. But for the same energy of absorbed light the piano 

 of polarization has no influence upon the intensity of the 

 emission. Hence for this photoelectric effect of the alloys, 



* E. Pohl, Verh.d. D. Phys. Ges. xi. p. 715 (1909). 



f R. Pohl and P. Pringsheim, Verh. d. D. Phys. Ges. xii. (1910) 

 pp. 215-230, §§1-9; pp. 249-260, §§ 10-17; pp. 682-696, §§ 18-27 ; 

 pp. 697-710, §§ 28-37. 



