10 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 85 



shown in Figure 5. Thus the ordinates are the logarithms of the ratio 

 of the intensities before and after passing through the atmospheric 

 ozone. Referring this area so obtained to the results of these present 

 measurements pictured in Figure 4 affords a method of determining 

 the atmospheric ozone based on direct intensity measurements, and 

 yields for this day over Table Mountain, Calif., an ozone value of 

 3.8 mm., while the value given by Dobson for Table Mountain on this 

 same day was 3.42 mm. A somewhat more satisfactory extension of 

 this method of determining the ozone transmission may be employed, 

 whereby the curved base is reduced to a straight line. Instead of 

 plotting the logarithm of the atmospheric transmission coefficients 

 against wave length and determining the difl'erence of the observed 

 points from the smooth curve in the ozone region, the logarithm of 

 the logarithm of the transmission coefficients may be plotted against 

 the logarithm of the wave length, yielding very closely a straight line, 

 except for points in the ozone region. This fortunate circumstance 

 is due to the approximate inverse A° dependence of the logarithm of 

 the transmission coefficients on the wave length. The ozone trans- 

 mission can be determined from this plot in a similar way from the 

 difference of observed points from the straight line. The trans- 

 mission coefficients for March 24 were treated independently in this 

 second way. The area resulting was very closely the same as that 

 shown in Figure 5, yielding 3.9 mm. ozone path. It is believed that the 

 application of this method to data of days which give good trans- 

 mission coefficients affords a satisfactory method for determining 

 atmospheric ozone from direct intensity measurements. 



In order to make a somewhat more extensive comparison of ozone 

 values determined by this method wath those previously known from 

 observations made with the Dobson apparatus, a series of eight days in 

 1928 and 1929, for which values by the Dobson method have been 

 obtained at Table Mountain, was treated according to a somewhat 

 abbreviated form of the above method. For these days there were 

 available the atmospheric transmission coefficients as regularly read 

 and already computed from the " long method " observations at Table 

 Mountain. These values are not given at all points used in the deter- 

 mination of the area defined in this present work, but at the regular 

 spectrum points ordinarily determined in the solar-constant work. 

 These relatively few points are scattered over the spectrum in such 

 a way as to outline, somewhat less accurately to be sure, essentially the 

 same area as that defined in the present work. In particular there were 

 but four ordinates lying in the ozone region, but if these values were 

 sufficiently accurate the area of ozone absorption outlined by them, 

 and lying between the wave lengths specified above, would be suf- 



