Hull — Initial Velocities of the Electrons. 



257 



The exact values of the maximum energy are hard to 

 determine, but the points of zero deflection of the electrom- 

 eter, where the curves cut the axis of abscissas, can easily be 

 determined to within 0*01 volt, and since it is probable that 

 the same proportion of light is reflected in each experiment, 

 the potentials represented by these points will be nearly pro- 

 portional to the real maxima. These values will therefore be 

 used in testing the proportionality between light-frequency 

 and the velocity of the electrons, and are designated in Table 

 III as maximum energy. 



If the results of Ladenburg hold for this region, the maxi- 

 mum energies and the shortest wave-lengths in the light pro- 

 ducing them should be related according to the equation 



— : — - = -y/Y., : a/Y 2 > where V is the energy in volts. If 



we take as a fixed point of reference the shortest wave-length 

 in the hydrogen spectrum which can get through white fluor- 

 ite, X 1280 as determined by Lyman, to which corresponds 

 the maximum energy, 3*03 volts, the other wave-lengths can 

 then be calculated by the formula 



A = 1230 



v ¥ 



and the calculated values compared with those which are 

 known from spectroscopic experiments. This has been done 

 in Table III. 







Table 



III. 



, Shortest W 



ave-length ^ 







Maximum 



Calculated 



Estimated 



Source 





Energy 



on 



from 



of 





in 



Ladenburg's 



Spectroscopic 



Light 



Screen 



volts 



Theory 



Experiments 



H s ... 



2 mm fluorite 



3-03 



[1230] 



1230 



CO,.. 



2 mm fluorite 



2'42 



1375 



above 1230 



H 2 ... 



l-5 mm quartz 



2 33 



1400 



1450-1500 



H 2 ... 



l-5 mm quartz 











+ l-5 cm air 



1-58 



mo • 



mo 



CO,.. 



l-5 cm air 



1-62 



1680 



mo 



(or a little less) 



In regard to the values of \ in the last column, those for 

 hydrogen, taken from Lyman's results with the vacuum spec- 

 trograph, may be regarded as well fixed, although some of 

 them may be too high for the photo-electric effect, which is 

 probably more sensitive than a photographic plate, especially 

 in the region of very short wave-lengths. This may account 

 for the low value obtained with quartz. 



