Light Action in Certain Crystals of Metallic Selenium. 501 



from the selenium surface in the so-called selenium cell*, 

 and found more than 98 per cent, to be absorbed in all parts 

 of the spectrum. 



In comparing the observations just referred to it must be 

 remembered that in the front side illumination no light 

 reached directly either the crystal or the point of contact 

 when the beam was on the electrode. On the contrary, 

 when illuminated from the back the light fell on the crystal 

 itself in the experiment marked " On contact n in the table 

 No. 1. A glance at fig. 2 will make this point clear. 

 Further, it is interesting to note that as long as any light fell 

 on any part of the crystal there was a marked change in the 

 conductance, even though the illuminated portion of the 

 crystal was not in direct line of conductance. This re- 

 markable property of transference of action will be further 

 verified and discussed later in this paper. 



It may be mentioned that the points of contact between 

 the crystal and the electrodes were illuminated in a separate 

 experiment, in which the electrodes were semi-transparent 

 silver films deposited on glass, and the action at the electrode 

 contacts was found to be slightly less than elsewhere. 



Another crystal was investigated with the apparatus shown 

 in fig. 2, with a slightly modified procedure. The electrodes 

 were moved so as to be 0*1 mm. apart, and on the same side 

 of the acicular crystal, which was a little over Ol mm. thick. 

 Thus the resistance around by the back side of the crystal 

 should be, if the crystal were considered as being nearly 

 isotropic in its conductivity, approximately three times that 

 along the front face of the crystal. Supposing the crystal to 

 be uniformly conducting, we should obviously expect a wide 

 difference in the conductance changes between front and 

 back side illumination, providing the light acted only at a 

 small effective depth, or conversely the two light actions 

 should be the same, providing that the light action spread 

 throughout the crystal without apparent absorption. A 

 beam of width 0*1 mm. was let fall first on the front side 

 and then on the back side, after having passed through the 

 same thickness of olass in each case. An average of five 

 readings on the front side showed an increase in the con- 

 ductance of 29*2±0'8, and similarly the back side illumina- 

 tion produced a change of 30*4 ±0'5. These changes are so 

 nearly identical that we are led to believe either that the 

 light penetrates the crystal without absorption, or that the 



* Work not yet published. 



