952 



SCIENCE 



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



tiires," " where, however, the application of 

 the theory to Neptune is purely illustrative. 



Mars has a rather rare atmosphere, but a 

 climate of the continental type, giving it 

 warm summers. Quantitative measures of 

 the intensification of the aqueous absorption- 

 bands in the spectrum of Mars" confirm the 

 evidence of melting polar snows, and assure 

 us that the summer temperature of Mars is 

 considerably above the freezing point. 



The earth, since it is nearer to the sun than 

 Mars and has a denser atmosphere, is, on the 

 whole, hotter than Mars. Terrestrial summer 

 climates would be even hotter than they are, 

 if it were not for the tempering efl^ect of its 

 oceans. 



The air modifies surface temperatures both 

 by its absorbent and its convective properties. 

 All parts of the earth's insolated surface are 

 cooled by contact with air in motion. This 

 source of thermal depletion is very much 

 smaller on the moon. On the other hand, the 

 earth's temperature is very much increased by 

 the absorbent action of its atmosphere on tel- 

 luric radiation, an action which is probably 

 very small on the moon, since its visible spec- 

 trum shows not the faintest atmospheric ab- 

 sorption. It is well known that the moon's 

 atmosphere is excessively rarefied, yet a 

 minute amount of some especially absorbent 

 vapor might make a considerable difference in 

 the night temperature, if the absorption-bands 

 were of wave-lengths corresponding to low- 

 temperature radiation. Speetrobolometric ob- 

 servations have not favored the supposition, 

 but are hardly delicate enough to reject it 

 absolutely. Without demanding any exact 

 compensation in these two opposite tendencies, 

 it is sufficient to see that they do oppose each 

 other, and that the final controlling factor is 

 duration of insolation. This is great enough 

 on the moon to permit the formation of a 

 steady subsurface thermal gradient, and the 

 attainment of a maximum temperature de- 



'" Philosophical Magazine (6), Vol. 16, p. 478, 

 September, 1908. 



'^ Frank W. Very, ' ' Measurements of the In- 

 tensification of Aqueous Bands in the Spectrum of 

 Mars," Lowell Observatory Bulletin, No. 36; and 

 "New Measures of Martian Absorption Bauds on 

 Plate Em 3076," Ibid., No. 49. 



pending only on the absorptive coefficient of 

 the surface and the solar constant of radia- 

 tion. The fact to be emphasized is that no 

 estimates of planetary temperatures are pos- 

 sible without considering the nature of the 

 planetary atmospheres and the duration of 

 insolation, and applying a knowledge of the 

 principles of thermal conduction and of the 

 " greenhouse " theory. But the moon is near 

 enough to permit measurements of its radiant 

 emission, in which the only hypothetical ele- 

 ment remaining concerns the explanation of 

 the observed facts. Since Abbot and Fowle, 

 in connection with their objections to my 

 value of the lunar temperature,*^ have re- 

 ferred favorably to the opinions of Dr. W. 

 W. Coblentz, it may be well to point out a 

 few statements by the latter writer which 

 demand reconsideration. Dr. Coblentz in his 

 paper, "Radiation from Selectively Reflecting 

 Bodies," " says : 



The reflecting power of the moon for visible 

 rays, according to Langley, is only 1/500,000 full 

 sunlight. Assuming that at 9 ju. the reflecting 

 power is, on the average, ten times that at 0.5 to 

 4 /i (a low [sic] estimate), the value becomes 

 1/50,000. 



Here, " reflecting power," or albedo, has 

 been confused with the total amount of light 

 reflected by the moon, expressed as a fraction 

 of sunlight. I found that the average reflect- 

 ing power of the moon for solar rays of every 

 wave-length, both visible and invisible, was 

 about 13 per cent." Zollner obtained, for 

 visible rays, a lunar albedo of 17.4 per cent.*' 



^'Annals Smithsonian Observatory, Vol. 2, p. 

 175. 



^Physical Beview, Vol. 24, p. 314, March, 1907. 



" Astropliysical Journal, Vol. 8, p. 275, Decem- 

 ber, 1898. 



" This includes a fraction due to specular reflec- 

 tion which causes the phase-curve for moonlight 

 to differ from that for a diffuse reflector, as well 

 as from the phase-curve for proper lunar radiation 

 emitted from the heated surface (for which see 

 Fig. 15 of my "Prize Essay on the Distribution 

 of the Moon 's Heat and its Variation with the 

 Phase," and compare with Zollner 's curve in his 

 " Photometrische Untersuchungen"). If the defi- 

 nition of albedo is restricted so as to include only 

 diffuse luminous reflection, we have such values as 

 the following: Wollaston, 0.12; Bond, 0.071; 



