l(j MEMOIRS OF THE NATIONAL ACADEMY OK SCIENCES. 



1. The unaiitity of beat leaving tlie moon at auv instant may witlumt mncli cnor be eousiileieil the same as that 

 i'alliug ou it at that iustaut. 



2. The absorptive power oi'onr atmosphere is the same tor Innai- ami solar heat. 



3. As was assumed in a previous formnla, the moou is a smooth sphere, not capable ot'retleetiiig heal regularly. 

 Deducing a t'ormnla for the amount of the moon's ditfuse heat received by the earth and substitutiug in it the 



necessary values of the <iuautitles entering, we tind that the amount is yin'ijjir of that received from the sun : a result 

 which agrees well with the previon.s values. 



The value of the galvanometer deflections was obtained by comparison with a vessel of hot water, which sub- 

 tended the same angle at the thermopile as the large mirror. It was then found (the radia ing power of the moon 

 being supposed equal to that of the lampblack surface and the earth's atmosphere not to iutluencc the result) that a 

 deviation of Oil for full moon (about the average ettect) appears to indicate an alteration of t<'mperatnre through hiW' 

 Kahr. In deducing this result' allowance has been made for the imperfect absorption of the sun's rays by the lunar 

 surface. 



These observations must be regarded as merely preliminary, and the results may be sniiject to revision wlien 

 more accurate measurements are obtained. 



[Proceedings Koyal Society. XIX. page y. (Itf/d)]. 



In the preciding paper it was shown that a large. iiortion of the total lunar radiation cimsists of rays of low re- 

 IVangibility emitted by the heated surface of the moon, which in the time of a complete revolution pa.sses through a 

 range of probably more than ,)00" Fabr. of temperature. 



The ratio of the intensity of solar heat to lunar beat, as deduced from the observations, agreed well witli values 

 gi\en by independent determinations. 



Since the last communication, more accurate measurements have been made, with results substantially the same 

 as those reached in Irtti'.l. 



The glass used was found to transmit 87 |ier cent, of the sun's rays. \'i per cent, of the radiation from the moon, 

 and 1.6 per cent, of that from a body at l;iO^ Fabr. 



Assumiugt then that 92 per cent, of the luminous lays in either moonlight or sunlight is transmitted V>y glass, and 

 l.Gjier leiil. of the obscure rai/s traiixmHted. and tailing 82,60(1 (found by direct experiment) for the heat ratio of solar 

 to lunar rays, the resulting value of the light-ratio is 676:!d0: 1. which agreeing well with the accepted value, 

 show 8 that the heat ratio 82600 : 1 is very nearly correct. 



In these experiments the quautity measured by the tberumpile was the ditlereuce between the radiation from the 

 circle of sky containing the moon's disk and that from a circle of sky of equal diameter not containing the moon's 

 disk : no information in reference to the absolute temperature of either the moon or the sky results from them. 'I'be 

 aiiparent temperature of the sky was found by coniiiarisous with the radiation from blackened vessels of hot watei 

 at different temperatures to be from 17 to '.i'i- Fabr. If the temperature of space be really as low as has been suji- 

 jiosed, this result .seems to indicate considerable opacity of our atniospbere for heat-rays of low refrangibility. 



'I'he observations made to determine tin! dependence of the beating power of the moon, on her allitude, and the 

 law of extinction of her rays in our atmosphere, are not veiy .satisfactory on account of changes in the sky, which 

 jireseiit greiit obstacles to measurements of this character. 



The curve deduced in the first paiier is given together w ith the later observations. As far as can be juilged from 

 so few and imperfect experiments, the maximum of heat seems to be a little after full umon. 



[Proceedings Koyal Society, XXI. p. 241, (187:t).] 



In this )paper is given the result of recent and more careful observations made for the purpose ot determining the- 

 depen<lence of the moon's heating power upou her altitude, the curve obtained being uearly but not <inite the same 

 a* that found by Professor Seidel for the light of Ibe stars, and showing a greater extinction of light than heat. Hy 

 employing the table thus deduced, and introducing a correction for etiect of change of distance of the sun, a move 

 accurate phase curve was deduced, indicating a more rapid increase of the radiant heat on approaching full moon 

 than was giveu by the formula previously employed, but still not so much as Professor ZiiUner gives for the moon's 

 light. At 79'^ zeuith distance, the etiect of atmospheric absorption is about one-tenth of the whole amount. Lord 

 Rosse oliserves that Ibis difference may be due to the fact that Seidel's light observations were made on the stars, not 

 on the moon, and that it hence does not necessarily imply a different law for the extinction of light from that tor 

 beat. 



From a series of simultaneous measurements of the moon's beat and light at intervals during the |)artial eclipse 

 of November 14, 1872, it was f^/uud that the heat and light diminish nearly, if not quite proportionally, the miiiiujiim 

 for both occurring at or very near the middle of the tclipse, when they were reduceil to about half what they w ere 

 before and after contact with the penumbra. 



The probable error of a single set of 10 galvanometer readings at this time is given as about I'J per cent., or 

 about 8.7 per cent, for the jnobable error of a night's observation: but Lord Kos.se concludes that so large constant 

 errors were probably present, tbat any inerea.se in the number of sets was almost powerless to obtaiu a nuire reliable 

 result. 



* It will be seen by reference to a subsequent paper that this result is considered to be erroneous. What is meant 

 by an allowance for imperfect absorption by the lunar surface is n<it <leai, 



tThe assumidions we have italicizeii iit'f^ little mater tlie ti'nt|i than those ot the previous papn and the result 

 entitled to little more weight. 



