REFLECTING POWER OF CRYSTALS 181 



ble results. Drude 5 and Miiller both found that a fresh surface 

 of antimony sulphide deteriorated rapidly after coming into 

 contact with the air. 



The writer thought it to be highly desirable to determine 

 the reflecting power of certain of these crystals by a dired 

 method. If these results should show definitely a difference in 

 the reflecting power with difference in azimuth of the incident 

 plane polarized light, then there would be a larger amount of 

 reliability to be placed in the results that are being obtained 

 in this laboratory by the more exact, bul indirect methods. 



The plan of the experiments was very simple, but the execu- 

 tion has been rather difficult. The first arrangement was an 

 apparatus by which plane polarized light from a monochroma- 

 tor, varying in wave length throughout the visible spectrum, 

 could be thrown upon a linear thermopile, and secondly re- 

 flected upon this thermopile b}^ reflection at nearly normal in- 

 cidence from the crystalline surface. The ratio of the second 

 deflection of the connected galvanometer to the first would give 

 at once the reflecting power. By rotating the crystal in the 

 plane of its surface through 90°, and repeating the above ex- 

 periment, one could get the reflecting power from this azimuth. 

 The results were, however, unsatisfactory because not enough 

 energy could be obtained from the small crystalline areas to 

 give reliable deflections of the galvanometer. 



A second attempt was made in which a crystal of selenium, 

 connected in a Wheatstone's bridge, was substituted for the 

 thermopile. The effect of light on the selenium is to decrease 

 its resistance, and it was thought in view of the fact that a 

 selenium receiver is much more sensitive in the visible spectrum 

 than the thermopile that definite results could be obtained. 

 However, the crystal employed, although more sensitive than 

 the thermopile, proved unfortunately to be more erratic and 

 unsteady in its action. While perseverance would no doubt 

 have led to a more satisfactory crystal receiver, it was de- 

 cided for the time being to abandon this plan of attack and 

 to use a speetrophotometric method. The apparatus was ar 

 ranged in several different ways which need not be gone into 

 here, before what proved to be a satisfactory method was found. 

 The plan is shown in figure 15. Light from a Nernsl glower, G 



5 P. Drude. loc. cit. 

 6 E. C. Miiller, loc. cit. 



