LUNAR RADIATION. 145 



8 to 10 n, we do not know how much is selectively reflected energy from 

 the sun. But the writer maintains that just as soon as we admit the 

 possibility of the moon being selectively reflecting in the region where it 

 has its own proper radiation, then the use of glass 1 as an absorbing screen 

 for testing the quality of that radiation is inadmissible, although it does 

 serve as a rough test of the reflected energy at 0.5 to 3 y.. For this purpose 

 it has been serviceable, and when we consider the great difficulties under 

 which such work must be carried on, and the numerous corrections that 

 must be introduced (which at all times may be larger than the one for 

 selective reflection), the use of a glass screen is not objectionable. 



Furthermore, we have noticed that the silicates reflect as a transparent 

 medium in all regions of the spectrum considered, except from 8 to 10 /*, 

 where it reflects like a metal. This means that the reflected energy in all 

 these regions except 8 to ioju will consist of two parts, viz, that reflected 

 from the outer surface and that due to internal reflection. From 8 to 10 \x 

 there will be but little, if any, energy due to internal reflection. Hence, 

 to use the reflected energy spectrum of the moon from 0.5 to 4 \i as a means 

 of estimating the amount of reflected energy at 8 to 10 p. is not a fair test, 

 and, as used (for want of better data) in the present paper, can only serve as 

 an approximation. As in the case of emission spectra, it ought to be possi- 

 ble to analyze a substance by means of its reflection bands. The constitu- 

 tion of an isolated body like the moon, shining by reflected light, might thus 

 be determined. This will probably never be possible since terrestrial atmos- 

 pheric absorption will interfere with the observations, while a far greater sen- 

 sitiveness in the radiometers will have to be attained than now is possible. 



From whatever standpoint we view the matter, the conclusion is that 

 the lunar surface must be diffusively, selectively reflecting, however small. 

 Hence, in the stream of energy reflected in any direction from the lunar 

 surface the density will be greatest for the wave-lengths of the bands of 

 selective reflection. One would, therefore, expect to detect this difference 

 in all directions, and not simply at the angle of reflection as suggested by 

 Very (Joe. ciL). It is difficult to apply a thorough test which will deter- 

 mine whether, and how much of, the energy from the moon is due to emis- 

 sion, and how much is due to reflection of energy from the sun. A rigid 

 comparison of the energy curves of the sun and the moon in this region 

 would be of great value. The polarization of the radiation from the 

 moon might also be of use in deciding this point. Pfund 2 has shown that 



1 Various observers have compared the total radiation from the moon to that part which 

 is transmitted by glass. Glass being opaque beyond 4 jw. was assumed to absorb the proper 

 radiation from the moon, while the part transmitted by glass was considered to be reflected 

 energy from the sun. Evidently if there is selectively reflected energy of the sun beyond 4 m, 

 then it will be superposed upon the direct radiation of the moon, and the use of a glass screen 

 for testing the lunar radiation is inadmissible. 



2 Pfund: Astrophys. Jour., 24, p. 19, 1906. 



