40 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



irregular surfaces of the glass pane, which produces larger loss when the latter is ontside at a dis- 

 tance from the iubtrument. 



The solar energy which falls on the luoon may be divided into two portions: «, that which is 

 reflected or dilfused ; h, that which is absorbed by the lunar soil and reradiated. We may form 

 some rude « priori notion of the relative value of these from the following considerations : Were 

 the full moon a perfectly diffusive body and reflecting according to the law established by ZoUner's 

 experiments, it should behave nearly as aflat disc would do, and return to us such a portion of the 

 sun's energy as the angular area of its disc bears to that of the hemisphere gyioo? (" )• Hence we 

 may take this fraction to express the ratio of total lunar radiation — i.e.,{a+ b) — in terms of solar radi- 

 ation. The ratio of lunar to solar luminous radiation is here taken to be (roughly) j o orroo — ^"t 

 the ratio of lunar nonluminous to solar nonluminous radiation, owing to the selective absorption 

 of the lunar surface, is probably indefinitely greater. This latter ratio is unknown, but the larger 

 it is the smaller is the portion which we must assign to radiated heat. If, for instance, the per- 

 fectly diflusive moon sends us -^ of the total solar I'adiation, and the ratio of lunar to solar radia- 

 tion within the limits of the solar spectrum be -^j ^ is the proportion which is diffused or reflected 

 to us (rt), and 1—'^ is that which is absorbed and radiated [b). Now, « is a little less than 100,000, 

 and m varies with the degree of selective reflection in the lunar surface. If m be 000,000, one-sixth 

 of the lunar radiation is reflected or diflused solar energy, and five-sixths absorbed and radiated 

 from the soil. If m be 300,000, one-third of the energy is reflected, &c., and somewhere between 

 these two values it seems probable that the ratio sought will lie. The heat sent earthward by the 

 radiation from the lunar soil is almost certainly greater than that reflected or diflused ; but our at- 

 mosphere is, according to what we have been hitherto accustomed to think, comparatively opaque 

 to the first class of heat (that radiated from the soil) and comparatively translucent or diatherman- 

 ous to the.second, so that there seems an a priori probability that the true ratio between a and 6, 

 as it would present itself to an observer outside our atmosphere, will be altered by its absorption, 

 and that actually observed at sea level be something different. It seems certain, at any rate, that 

 the radiation of the lunar soil must be of a quality to which glass is nearly opaque, since the glass 

 which we have employed in our own experiments is nearly opaque to the radiation from a source at 

 lOOoC 



Class 5.* — Heat observations during a lunar eclipsk. 



The only lunar eclipse observed at Allegheny was that of October 4, 1884. The eclipsed 

 moon rose behind clouds, and the first observation, obtained when the penumbra was already pass- 

 ing off, was made while the moon was still partly obscured by haze. Under these circumstances 

 little interest attaches to the observation, which need not be cited in detail. The inference from 

 it, so far as any could be drawn, was that about the same amount of heat was received as was to 

 have been expected had there been no previous eclipse. 



KEVIEW. 



Let us now review our sources of information and weigh the imperfect and sometimes contra- 

 dictory results each has brought us. 



(1) Direct measurement of lunar heat as compared with solar. — Our direct comi)arison indicates 

 that we receive nearly the whole proportion of solar energy from the full moon that we should ex- 

 pect to get from a diflusive disk of the same angular aperture. This heat must in reality be partly 

 diffused and partly radiated, and we do not know (from the present observations) in what propor- 

 tions these two kinds enter. So far as the observation itself is reliable, we may, however, infer that 

 our atmosphere is permeable to most of the lunar heat of either kind, but the method is unfortun- 

 ately subject to such large sources of constant error, that we cannot derive great confidence from 

 the apparent agreement of different observations or even of different observers. It may be said, 

 however, to create a certain presumption that the earth's atmosphere is diathermanous to heat of 

 lower wave-length than has been heretofore snpposed, and of lower wavelemiihthan appears to reach 

 us from the sun. 



" Clnss (i, «i*c ii'fi'fi. 



