ATMOSPHERIC ATTENUATION OF ULTRA-VIOLET LIGHT. 371 



one-half the intensity of X3404 after passing through a long column of 

 air, the intensity of the line of shorter wave-length is taken as one-half 

 of 0.84. The results obtained in this way are plotted in Figure 1 with 

 the computed curves for scattering. 



Figure 1. The lower curve shows the effect of molecular scattering calcu- 

 lated from Rayleigh's formula for an air path of 8000 meters at N.T.P. — an 

 equivalent to the earth's total atmosphere. The upper curve shows the same 

 for an air path of 2300 meters at N. T. P. The points plotted between these 

 curves were obtained from the measured opacities of the spectral images. 



The decrease in intensity for the shorter wave-lengths, as found in 

 my experiments, is very marked. In view of the results obtained by 

 Fabry and Buisson ^^ on the absorption of ozone and the recent work 

 by Fowler and Strutt ^* on the absorption bands in the ultra-violet 

 spectra of the sun and a number of stars of different spectral types, 

 this deviation from the scattering curve can hardly be attril)uted to 

 ozone absorption. The Cd lines X2748, 2573 and 2313 seem normally 

 to have about the same intensity. X2o73 lies very near the maximum 

 of the great absorption band of ozone in the ultra-violet region, — a 

 path of 1.2 cm. of 1% ozone suffices to absorb it completely, while 

 X2748 and X2313 are only slightly diminished. If atmospheric ozone 

 were the chief factor in the apparent absorption, indicated by a long 



13 Fabrv- and Buisson, Comp. Rend., 156, p. 782, 1913. Journ. de Phy.sique, 

 3, p. 196, 1913. 



14 Fowler and Strutt, Proc. Roy. Soc, London, 93-\, p. 577, 1917. 



