NO. 8 WATER-VAPOR TRANSPARENCY FOWLE 33 



tions, taking- as zero the deviation at 1.838 /x, are indicated by the 

 numbers at the bottom of the plot. 



Underneath these three branches of the black-body curve com- 

 puted for a temperature corresponding to the " black-body " tempera- 

 ture (2,200° K.) of the Nernst lamp will be found curves repre- 

 senting the energy spectrum of the lamp through increasing amounts 

 of aqueous vapor. These amounts, as indicated in the table on the 

 plot, range from 0.0035 to 0.012 cm. ppt. H^O when observed 

 through the spectroscope alone; and from 0.028 to .261 when ob- 

 served through the tube. Each curve is the mean of several sets of 

 observations. 



For the curves with 0.0035 to 0.012 cm. ppt. H.,0 there were 7 

 grams carbon dioxide, and in all the others, except the one with 0.254 

 cm. ppt. HoO, there were 83 grams carbon dioxide in a i m. sq. path ; 

 for that one there were 160 grams. For a discussion of its influence 

 on the absorption in the region of figure 9 see the work with the 

 60° prism. 



In the small figure in the upper left-hand corner will be seen a 

 curve (marked a) which shows the percentage change in area of the 

 maximum between 1.25 and 3 /x plotted against the ppt. HoO in cm. as 

 abscissae. In the other small figure the similar function is plotted 

 for the region between 5 and 9 /x. Curve b was read from the large 

 plot and is for a distribution of energy for a body like the Nernst 

 lamp at 2,200° K. Curve c was computed for a distribution of 

 energy for a body at a temperature of about 287° K., which is about 

 the mean temperature of the earth. Table 6 gives more in detail 

 the data from which curves b and c were drawn. It will be noted 

 in the main curves of figure 9 that between 6 and 7 fx the absorption 

 is practicallv complete for ppt. H^O of i mm. or more. 



from a body at 300° K. to the bolometer. Hence the observed intensity of the 

 lamp is really the difference of the deflections in the two cases and should 

 therefore be compared with the black-body curve as drawn. The ratios of the 

 radiation from a body at 300° K. to that from one at 2,200° K., both radiating 

 to absolute zero, are shown in the following table : 



Table 5.— Ratio of Radiation from a Body at 300' K. to that of a Body 



AT 2.200° K. 



Wave-length. I 2/1 ' 6m ' 8m 'iomI2mi4Mi i6> | 18 m 20 fi 

 Ratio iXiO"" .00067 -0031 .0075 .013 .020 I .026 .033 .038 



To a wave-length of about 7 m- the radiation of the shutter is negligible. At 

 6 |x the observed radiation of the lamp needs to be increased i/io of i per cent 

 and at 20 n by 4 per cent in order to represent what it would be if radiating_ 

 to absolute zero. 



