34 



SMITHSONIAN MISCELLANEOUS COLLECTIONS 



VOL. 68 



Taule 6. — Water- Vapor Absorption 5 to 9/x 



DESCRIPTION OF FIGURE 10 



Figure 10 is a continuation for wave-lengths longer than 9 /a of 

 the curves shown in figure 9. The scale of abscissae is much more 

 condensed. The ordinates of the two right-hand branches are magni- 

 fied lo-fold relative to the first or left-hand branch. The extreme 

 right-hand curves represent only the work of 1916-17. As there 

 are so many observations it was thought best to shift the abscissae 

 of the 1 91 6- 1 7 work five deviation units to the right to avoid 

 confusion. 



As in figure 9, so in figure 10, above the first two branches is given 

 the computed black-body curve corresponding to the effective tem- 

 perature of the lamp. This will be presently considered. Just below 

 it is another heavy line. It represents the mean of observations of 

 the last three years mainly through the spectroscope alone and within 

 a range of water vapor from 0.003 to 0.028 cm. ppt. H2O. Within 

 this small range of ppt. H2O no systematic differences were found. 



Observations on a cold day were desired for three purposes : First, 

 to permit the comparison of the curve just described with one through 

 the tube with an amount of vapor of the same order. Generally the 

 tube contained more than lo-fold this amount of water vapor. 

 Second, from a comparison with the mean curve just described to 

 form some estimation of the eflfect of carbon dioxide. It was 

 supposed that on a very cold day the absorption due to the water 

 vapor in the tube would be so nearly the same as with the observations 

 through the spectroscope alone that the differences would be negli- 

 gible and thus the increased effect of the carbon dioxide in the tube 

 would be unmasked. Its effect, since its amount in the tube is prob- 

 ably quite constant from day to day, could be then eliminated from the 

 aqueous vapor effects. Third, if coincidence of the two curves taken 



