366 SCHAEFFER. 



NH3, H0O2, and water vapor showed no appreciable absorption in this 

 region even in far greater quantities than occur in the atmosphere. 

 Ozone, when very dilute, showed an absorption band of great intensity 

 extending from X2850 to X2320. The limit of transmission was finally 

 restricted to X3160 in presence of greater quantities of the gas. Hart- 

 ley concluded that ozone in the atmosphere is the absorbing factor 

 which accounts for the limit of the solar spectrum, since he found an 

 indication of ozone in free country air. He also presented some 

 evidence to show that the concentration of ozone at higher altitudes is 

 greater than at sea-level. 



In repeating Cornu's experiments on the limit of the solar spectrum, 

 Miethe and Lehmann ^ found no change in the ultra-violet limit when 

 the altitude was varied by 4500 meters. Wigand,* taking special 

 precautions to prevent fog on the photographic plate, took the ap- 

 paratus used by Miethe and Lehmann to an altitude of 9000 meters 

 in a balloon. Confining his attention to the last trace of light action 

 recorded on the plate, he found the limit to be X2S96 and the difference 

 in the limit at this great altitude and at Halle, near sea-level, to be 

 inappreciable. 



In view of these conflicting results, it seemed worth while to investi- 

 gate the absorption of long columns of air near the earth's surface,^ 

 for under these conditions the uncertainties as to the composition of 

 the absorbing layer are largely eliminated. Work was begun in 1916 

 but suffered several unavoidable delays. During the progress of the 

 research Strutt ^ published the results of his investigations on the same 

 subject; the conclusions arrived at in this paper do not materially 

 differ from his, but the author was able to express the magnitude of 

 the effect in quantitative terms and to show that molecular scattering 

 does not explain the observed absorption. 



In the present investigation the spark spectrum of Cd was photo- 

 graphed through columns of air from 160 meters to 8000 meters. By 

 comparing the measured opacities of the plates thus obtained with the 

 opacities produced in the laboratory when the air path was small, the 

 changes in the relative intensities of the principal Cd lines in the ultra- 

 violet were obtained. These changes in intensities give a measure of 

 the absorption due to the column of air under examination. As the 

 results did not point to the presence of ozone as the dominating factor 



3 Miethe and Lehmann, Ber. Preuss. Akad. Wissens., 8, p. 268, 1909. 

 4Wigand, Physik. Zeits., 14, p. 1144, 1913. 



5 Lyman, Monthly Weather Review, 42, 8 August, 1914. 



6 Strutt, Proc. Roy. Soc, London, 94A, p. 260, 1918. 



