[pratt] absorption OF LIGHT 169 



On examining the intensity curve (Fig. 3), it will be noted that 

 for wave lengths in the region of 6500 A. and all above that wave 

 length the instrument is not very sensitive, similarly, for very short 

 wave lengths (from 4000 A. down) it is much less sensitive, so that for 

 the two ends of the spectrum the results will not be as accurate as for 

 the region about 5000 A. in which the rate of drift is fairly rapid. 



The absorption curve (Fig. 4) indicates that we were wrong in 

 assuming that very little light was being absorbed in the red region. 

 Even if the results are not reliable above 6500 A. we see that there is 

 a large percentage of the light being absorbed which increases as the 

 wave length decreases, but does not become complete at any point as 

 we were led to suppose in our first investigation. The points group 

 themselves in a more or less regular curve with an abrupt change in 

 the percentage absorption. This abrupt change is characteristic of 

 all the curves for the particular wave length at which the change 

 occurs. The curve becomes very irregular as the wave length decreases, 

 owing to the rate of drift being very slow, hence accurate or reliable 

 results are difificult to obtain. Throughout the visible part of the 

 spectrum the proportion of light absorbed is a large fraction of the 

 incident light and increases as the wave lengths decreases till about 

 3800 A. is reached. Beyond this point nothing definite can be con- 

 cluded since the glass of which the photo-electric cell is made may be 

 absorbing an appreciable amount of the ultra-violet rays that are 

 emitted by the Nernst lamp and so their action is not recorded by the 

 apparatus. 



This work was done under the direction of Professor E. F. Burton. 



Sec. Ill, Sig. 12 



