6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I23 



where the subscripts 19, 20, 22, 24, 26, 28, and 30 refer to the seven 

 spectrum places. On the left-hand side of each equation is the total 

 optical density minus the scattering due to the air molecules, the first 

 being measured and the second known from theory. This difference 

 is equated to the ozone absorption plus the remaining scattering which 

 is described by the term in A"^ and the term ^. These three terms on 

 the right-hand side of the equation contain the three unknowns, x, 

 8, and ^, each with a known coefficient that does not change from day 

 to day. 



4. THE ABSORPTION OF WATER VAPOR AT 0.57^* 



Concerning Place 24, at wavelength 0.570/A near the maximum of 

 the ozone absorption, in addition to the absorption by ozone, there is 

 absorption by water vapor. This is very weak at this wavelength, and 

 correction for it does not appear to be warranted, in spite of the fact 

 that the ozone absorption is itself weak, unless the precipitable water 

 amounts to about 0.5 cm. Above this we believe correction for water 

 absorption is warranted, and it illustrates helpfully that one need not 

 forego the use of a particular wavelength for ozone determination 

 merely because detectable water absorption exists there. In addition 

 to the subtraction of the Rayleigh scattering from the total optical 

 density ( — logT„), a small contribution by water absorption should 

 also be subtracted (if warranted), comprising an additional term on 

 the left-hand side of this one equation (for Place 24). This term is 

 taken here as of the form v;*ppt HoO, where ppt H^O (expressed in 

 cm of liquid water) is the water vapor in the vertical atmospheric 

 path, a quantity regularly evaluated from the observations. 



The water absorption coefficient, rj, at Place 24 must of course be 

 determined. This is an apparent absorption coefficient in the present 

 work, depending on the slit width and resolution of the apparatus, 

 since the absorption arises from the many fine lines of a vibration- 

 rotation band of water, ^° and the fine structure is not resolved. How- 

 ever, the absorption is very weak and the correction for it is usually 

 small. We believe that as an approximation for the present purpose 

 a term of the form ^-ppt HoO suffices. It assumes that this absorption, 

 also, is of the simple exponential form. 



From a careful comparison of holograms at Place 24 on days of low 

 and of high water we estimate that t] has a value for our data of 

 0.0009 cm"^±50%. With this value, the water absorption amounts to 

 0.1% at about 0.5 cmpptHsO. 



10 The combination band 2i'j^-\-2''s- See, for example, Herzberg, G., Infrared 

 and Raman spectra, p. 281, 1945. New York. 



