Hull — Initial Velocities of the Electrons. 255 









Table I. 









Loss of Neg. Electric 



ity by A. 









Coulombs per rnin, 



, xlO- 8 







, 



H 2 



Light , 



, C0 2 



Light , 



Pot. of 



Without air 



With aii- 



Without aii- 



With aii- 



B volts 



screen 





screen 



screen 



screen 



— 6 



2 





— 0-072 



-0-27 



— 0-065 



— 4 



— 1-5 





— 0-059 



— 0-23 







-3*5 



— 1-25 















-3-0 



-0-05 





0-002 



— 0-17 





— 25 



2-0 





0-077 



— 0-0.8 



— 0-063 



— 20 



5-7 





0-185 



0*38 



— 0-041 



— 1-5 



11-2 





0-34 



1-12 



0-039 



— 1-0 



20-8 





0-640 



2-59 



0-367 



— 0-5 



30-5 





1-14 



4'95 



1-18 







39-8 





199 



7-00 



2-09 



+ 0-5 



42 5 





2-15 



7-53 



2-26 



+ 1 



42'5 





2-15 



7-62 



2-22 



+ 2 



43-6 





2-12 



7-67 



2-23 



+ 4 



45-0 





2-20 



7-66 



2-31 



+ 6 



44-0 





2-18 



7*63 



2-33 



about 1480, and the limit of the absorption of air as X 1107, 

 then the inverse of these wave-lengths are proportional to the 

 maximum velocities, that is, to the square-roots of the maxi- 

 mum energy given above, which fits Ladenburg's hypothesis. 



This suggested that the results with hydrogen light might 

 be explained by a small amount of light of very short wave- 

 length getting through the air in the screen-cell. Lyman * 

 has shown that the absorption of air is in the form of a band, 

 which, for a column of air 0*91 cm long at atmospheric pressure, 

 extends from about X 1710 to X 1270. The more refrangible 

 limit of the band is so indistinct on a photographic plate 

 that Schumann was unable to detect it at all.f But the 

 photo-electric effect is probably more sensitive than a photo- 

 graphic plate, especially for the shortest wave-lengths, since, as 

 Ladenburg has shown, it increases very rapidly with decrease 

 in wave-length. Moreover the ordinates of curve II in figure 

 2 are magnified twenty-five times. Hence a yery small amount 

 of light between X 1270 and X 1230 might produce enough high- 

 velocity electrons to change the form of the lower part of 

 curve II, and since the shortest wave-length, X 1230, is the same 

 as without the air screen (curve I), the maximum velocity 

 would be the same for both. 



To test this point the discharge tube was disconnected from 

 the fluorite cover of the screen cell C, and closed by a separate 



*T. Lyman, Astrophys. J., xxvii, p. 103, 1908. 



f Schumann, Smithsonian Contributions, No. 1413. 



