4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 85 



be unity in regions not influenced by ozone, and somewhat greater 

 than this where the intensity was reduced by the ozone absorption. 

 In view of the exponential decrease in intensity of Hght passing 

 through an absorbing medium, it is the logarithm of the ratio of 

 ordinates which is proportional to the amount of ozone in the path. 

 This will be zero evidently for those cases where the ratio is unity, 

 and positive or negative as the ratio is greater or less than unity. This 

 proportionality is based on the assumption that Beer's law may be 

 used here. There appear to be no causes concerned with the apparatus 

 and structure of the spectrum which could bring apparent false devia- 

 tions from this law, since the spectrum is composed of broad diffuse 

 bands,^ or perhaps more accurately fluctuations in the absorption co- 

 efficient without any apparent discontinuities. The question as to 

 whether the ozone absorption actually obeys Beer's law is an impor- 

 tant one and. so far as the author is aware, has not as yet been 

 satisfactorily answered, although attention has been called to it by 

 Ladenburg." However, this assumption is contained in all previous 

 determinations of ozone such as we are employing here and must be 

 similarly contained in the present work. Actually, instead of a single 

 hologram in these measurements, four consecutive holograms were 

 taken with the cell containing no ozone and four more with the cell 

 containing ozone. The average values of the ordinates at correspond- 

 ing points were compared. 



A difficulty enters, however, because of the time which elapses 

 between the taking of the ozone holograms and the uninfluenced ones 

 which we will call the oxygen holograms, since a very appreciable 

 change in air mass, or amount of atmosphere traversed, occurs be- 

 tween the two as well as possible weather changes. This was unavoid- 

 able with the apparatus at our disposal at the top of Table Mountain 

 because of the nature of the ozone technique, and is not easy to avoid 

 under any circumstances. Under these conditions, the ratio of 

 ordinates outside of ozone absorption would be far from unity and 

 the logarithm far from zero. Standard correction for air mass was. 

 therefore, applied to the average ozone ordinates to bring them to the 

 air mass of the oxygen observations. It is hardly possible, however, 

 to make correction for all differences between the two sets with 

 accuracy sufficient to bring the logarithm of the ratio of ordinates to 

 zero within a quantity small compared to the small ozone effect, and 

 it is not necessary because it is a difference in the logarithm for 

 different wave lengths that is of interest. 



* See footnote 2l> on page 2. 



^ Ladenburg, R., Gerlands Beitr. Geophys., vol. 24, p. 40, 1929. 



