954 



SCIENCE 



[N. S. Vol. XXXVII. No. 964 



introduces the assumption that the absolute 

 temperatures of sun and moon are 5,900° and 

 360°, and calculates by Planck's formula for 

 the distribution of energy in the spectrum 

 that the ratio of the radiations emitted at 

 these temperatures by complete radiators is 

 1 : 0.00316. 



This ratio of the emissive power [meaning by 

 this the radiation from equal areas] of the moon 

 to that of the sun [considering the latter to be 

 a complete radiator but the former to have a 

 relative emissivity of 0.1] vrill then be 0.000316, 

 vphich is 16 times (0.000316 h-0.00002) the re- 

 flected' energy of the sun from the moon." 



But, as we have just seen, the derivation of 

 the fraction 0.00002 is erroneous, the value 

 assumed for the lunar emissivity is improb- 

 able even for a narrow region of the spectrum, 

 and still more so for the entire spectrum, and 

 the argument founded on the supposed lunar 

 " reflected energy " is equally inadmissible, as 

 we shall now see. 



When it is remembered that the maximum 

 deflection in the lunar spectrum (furnished 

 by a rock-salt prism) at about wave-length 

 1 ;u, in the Allegheny measures was usually not 

 over 10 divisions of a millimeter scale, and 

 that the solar radiation at 9 /x is certainly not 

 more than a hundredth part of that at 1 fj., or 

 0.1 mm., the supposition by Coblentz that any 

 appreciable part of the lunar spectrum at 9 /i,, 

 coming indiscriminately from all parts of the 

 lunar surface, can be composed of reflected 

 solar radiation, is seen to be preposterous. 



The question of specular reflection does not 

 enter here. Isothermal charts of the moon™ 

 show an entirely difl^erent distribution of total 

 radiant energy on the moon's apparent disk 

 from that of moonlight, but this distribution 

 of total radiation is not much altered by the 

 small amount of reflected light which it in- 

 cludes, and is altogether appropriate to that of 

 the emission from a heated body having its 

 highest temperature at the subsolar point. 



"Op. cit., p. 315. 



^ Frank W. Very, ' ' Prize Essay on the Dis- 

 tribution of the Moon's Heat and its Variation 

 with the Phase," Utrecht Society of Arts and 

 Sciences, The Hague, 1891. Cf. Pigs. 7 to 14. 



Silicates have an emissive power not very dif- 

 ferent from 0.9 (nine times as great as this 

 author assumes), or a reflecting power seldom 

 much over 0.1. Throughout a considerable 

 part of the region of proper lunar radiation, 

 the reflected solar spectrum must have been 

 smaller than 0.1, perhaps not more than 0.01 

 scale division, and it would have been ab- 

 solutely unrecognizable. The actual deflec- 

 tions which reached upwards of 20 or 30 scale 

 divisions in this part of the lunar spectrum 

 were entirely due to emitted radiation; but 

 the part of the lunar spectrum of wave-length 

 shorter than 4 fi corresponded, both in the 

 form of its energy-curve and in the fraction 

 of its included energy, with the reflected solar 

 radiation. 



The supposed similarity between the reflec- 

 tion-curves obtained by Coblentz for some 

 common silicates, and the lunar spectral 

 energy-curve, a resemblance which is by no 

 means conspicuous, is purely fortuitous. The 

 lunar curve owes its shape to alteration by ab- 

 sorption in passing through the earth's atmos- 

 phere, and not to local abnormal reflection. 

 The resemblance would have been even less 

 approximate if Dr. Coblentz had drawn his 

 theoretical radiation-curve for the tempera- 

 ture which I have indicated for the moon, 

 which is not " 300° abs.," notwithstanding 

 that the reader of another work by the same 

 writer, " Infra-red Reflection Spectra," '^ 

 might infer from a footnote that this tem- 

 perature rests upon my measurements. The 

 curves published in Fig. 90 of the same work 

 and labeled " Reflection from Moon (Lang- 

 ley," can not possibly have the assigned origin, 

 as is evident from the preceding argument. 

 In repeating this figure in the Physical Re- 

 view,'' the designation has been changed from 

 " reflection " to a noncommital " radiation," 

 but the quotations cited show that the idea of 

 reflection persists. 



Dr. Coblentz also infers from observations 



'' William W. Coblentz, ' ' Investigations of 

 Infra-red Spectra," Part 4, Appendix 1, p. 114, 

 Carnegie Institution of Washington, 1906. 



^"Vol. 24, Fig. 3, p. 312. 



