TKMI'KKATl ItK OK THE SUIM-'ACE OK THE MOON. 15 



cuvc (iti tlic (Mrtli's surfaco possible I'lit sticli coiiclii.sioiis arc, it must 1)h adrnittiMl, in (',()iitia(li(;- 

 tioii not only to tlie statement alieail.v ((noted fiom Sir -loiin Heiscliel, lint also to what lias been 

 regarded as direct exiierimeiital evidence as to the high temperature of the Innar surface obtained 

 by the Karl of Rosse (that is, if we admit the moon to be absolutely airless, as Sir John Uerschel 

 assumes it; for it is not only possible, but even probab'e, that a gaseous enveloite to the moon, too 

 small to make its presence known to oidinary astiononiical observation, would greatly raise the 

 temjieiature of its surface, and the not impossible existence of sn(!h an envelope mnst be here 

 borne in mind). 



liOrd Kosse's oltservations, in which, as we have already leiiiarked, anything like ((uantitative 

 measurement of the lunar heat has for the first time been attained, we shall jiioceed to examine in 

 some detail. 



AHSTUACI UF IJH!I> ROSSICS VAfKUS US IHE UADIATIOS ill' HEAT FliOM IIIE MOOS', HITH COM- 



MESTS liV rilK I'KESESr II HI TEH. 



[l'ri)f<T(liiij;N IJiiyal .Society, XVi, \>. 4:i(_i (IStJU).] 



'I'lic olijcct of tlic oliseivalioiis ili.scMsscd in tlii> iiapef i.s the ileteniiiiiatioii of wliat inopoi tions the lunar railia- 

 lioii loMtains of — 



1. Heat loiiiing from the interior of the moon, whieh will uot varv with the phase. 



','. Me.it which falls from the snn on the inoon'.s Mirface and is at ouce retlected ref^nlarly and irregularly. 



:i. Heat whieh, fallinj; Iroin the sun on the moon's surface, is ahsculieil, rai.ses the tempirature of the surface, and 

 is afterward radiated as heat of low refranjiibility. 



The apparatus emidoyed was a :5-foot rcHecting telescope, with two small coiideiisiug mirrors and thermopiles. (A 

 tahle of observations made at diltereiit phases of the moou is then given in the original paper.) 



Assuming that the moon is a smooth sphere * without specular rcHection, we may compute, from theory, the form 

 of a curve, represeutiug the amount of heat received from the moou as a function of her phase. This curve approxi- 

 mates to a sinuous form, having a greater curvature at the maximum or at lull moon than at the minimum. The ob- 

 servations given in the table fall tolerably well on this curve, and therefore the increase and diuiinutiou of heat with 

 the varying phase of the moon follow the snuw law as that of light. Tlie heat classilied under head (1) can have no 

 existence in this case. 



We may seek to deterniinc the relative proportions of (2) and (:{) present in the lunar radiation by experiments 

 with thin plate-glass. 



About 80 per cent, of the solar radiation probably pa.sses through glass, from direct observations it was fouiul 

 that only 8 per cent, of the moon's rays was transmitted by the piece used in the experiment. 



From this resitlt, and the generally accepted value of the ratio of sunlight to moonlight, we may deduce the ratio 

 of solar to lunar heat radiation. T« do this Lord Kosse assuniest that all luminous rays are transmitted by glass and 

 all obscure rai/s iire Hojiped. 



We have then (according to him) 



Percentage of luminous rays in lunar radiation 8 per cent. 

 Percentage of luminous rays in solar radiation 80 per cent. 

 KatioJ of solar to lunar luminous rays = 800. 000 : 1. 

 liatio of total solar to total Innar radiation = 80,000 : 1. 



The correctness of the value obtained for this ratio is confirmed liy other considerations: in the lirst place, by 

 direct measurement. § 



The sun's rays were reduced by jia.ssing through a small aperture, and the deviation of the galvanometer con- 

 nected with that previously found for full moon, by using the ileviation produceil by a vessel of hot water as a term 

 of comparison. The ratio thus found was 89,81!) : 1. 



We may also dud a value of this ratio fnun theoretical eonsidcration.s. To do this he makes the Ibllowiug assump- 

 tions the basis of calculation : 



• ZoUner has shown that the lunar surface retleets nearly like a Hat ilisk, but the observations of Lord Kosse, 

 "•iveu in the table, are subject to so great uncertainty, that they would fall eiiually well upon Ziillner's curve. 



tThis assumption is so far from the facts now known as to make the determination of the heat ratio depending 

 u)ion it of little value, for it is now ascertained that in the solar-heat siiectrum fiumed by a glass prism nearly two- 

 thirds of the energy is represented by invisible rays. 



: Lord Ko:<SK assumes the largest known determination (800,000 to 1) as the most probable. There are, however, 

 from eight to ten determinations by observers of repute, and all of theiu smaller: and if we take the mean of these, we 

 have, as elsewhere shown, the approximate ratio 4tlO,000 to 1. The agreement of the value 800,000 with the values 

 obtained by the other methods given, on account of the falsity of the fundamental assumption of the latter, can ouly 

 be regarded as a coincidence. 



v\ To the fj-zif/imeKfa? determinations of tin' solar and lunar-heat intio \yx. make no obji'ction, and they are probably 

 as reliable :is any others. 



