36 



SMITHSONIAN MISCELLANEOUS COLLECTIONS 



VOL. 68 



the curves taken with and without the tube confirmed the view that 

 <iiffraction with the wide sHts used in this region causes Httle error 

 in the tube work. In table 7 will be found the percentages of absorp- 

 tion at various wave-lengths greater than 9 ju, as indicated from all 

 the experiments. The values are grouped with regard to the quantity 

 of ppt. HoO traversed by the beam, but the results of the years 1914 

 and 1 91 6- 1 7 are kept separate. The lowest line of table 7 comes 

 from observations of Rubens and Aschkinass in which they used an 

 absorbing tube containing 75 cm. of steam at 100° C. This contained 

 .045 grams ppt. HgO. 



Table 7.— Water- Vapor Absorption 9 to 14 fx 



cm. ppt. H2O 



Up to 0.03 



.10 



.15 (I9I4) 



.16 (I9I6-I7) 



.25 (I9I4) 



.045 rubens, steam 



13-5A 







15 

 15 

 21 

 26 

 25 



Note: o indicates no effect detected. 



Although Rubens and Aschkinass' work shown in figure ii was 

 done in a purer spectrum it will be noted from the lines marked j- 

 in the plot, that a wide range of angular deviation was subtended 

 in the spectrum by the widths of the bolometer and the slit. Not 

 even in their work, and still less in the present research, could the 

 line detail which doubtless exists in the water-vapor bands be ex- 

 pected to show. Referring to the table it may be seen that, as in 

 the observation at smaller wave-lengths made with the 60° prism, the 

 same amount of water absorbing (A) in the form of steam at 76 cm. 

 pressure and (B) at the comparatively low pressure of the order of 

 I cm. which prevails under atmospheric conditions, exercises much 

 greater absorption in the form of steam. 



COMPARISON OF NERNST LAMP WITH BLACK-BODY SPECTRUM 



Down to wave-length 10 /x the curve computed from Planck's 

 formula for black-body radiation from a source at 2,200° K. radiat- 

 ing to a black body at 300° K. shows no remarkable departure from 

 the Nernst lamp spectrum, when due allowance is made for the 

 absorption bands. From that point on the greater wave-lengths the 

 observed radiation of the lamp systematically falls below that of 

 the black-body curve. The black body curve as plotted in figure 10 



