NO. 8 



WATER-VAPOR TRANSPARENCY — FOWLE 



6l 



what seems a plausible interpolation. Then this central strip, with 

 its wing-like appendage h-lc^, may be assumed to represent the 

 energy curve of an approximately monochromatic source of radia- 

 tion, including the portion of its energy scattered through the 

 spectrum region on its long wave-length side. A symmetrical wing 

 should evidently stretch to its short wave-length side also. 



It is proposed to use the curve just described to obtain the scat- 

 tering taking place at each point in the spectrum from wave-lengths 



Fig. i8. — Field-light curves. 



greater than 4 /* and not directly obtained by the insertion of the 

 quartz plate. This involves the assumption that the contour of this 

 curve representing the scattering is the same for all wave-lengths 

 coming into consideration. Gorton ^ has published curves showing, 

 for great angles of incidence and unpolished, comparatively rough 

 surfaces, the variation of the scattering with the wave-length. These 

 show, as might be expected, less scattering for the longer wave- 

 lengths. The smallest angle of incidence used was 54°. The obser- 



* Physical Review, 7, p. 74, IQ16. 



