510 BoEDDiCKER — On Lunar Radiant Heat, 



which, gradually increasing, ended (as far as observed) in reducing the lunar heat 

 to 460 divisions of the galvanometer-scale shortly after S*" 30"". Unfortunately I 

 then stopped observing for 30 minutes, so that the recovery of the curve (which 

 certainly must have taken place during this interruption) was not recorded. The 

 character of the curve depends, therefore, essentially on the observations obtained 

 after the interruption, and it is not unlikely that all these last observed quantities 

 are still to some extent affected by the preceding disturbance, and, consequently, 

 somewhat too small. 



7. A satisfactory explanation of the deficiency of lunar heat after the 

 end of the eclipse, in spite of the rapid fall to almost zero during the first 

 half of it, I have as yet been unable to find. One fact, however, which 

 may have some bearing on the question I may here mention, viz. that the 

 heat- values of 1884 and 1888 corresponding to the last contact with the 

 shadow and to the last contact with the penumbra seem to be invei'sely pro- 

 portional to the times elapsed since the beginning of the two eclipses. We 

 obtain, namely, under the assumption of such a proportionality for the heat at 

 these two epochs in 1884, the figures 38'2 % ^"^^ 83-9 7o) while the actually 

 observed quantities were 41-4 % and 85"2 7o' I^ this proportionality were actually 

 established — which is at present not the case as far as my observational material 

 goes — it would seem to indicate that the amount of lunar heat transmitted by our 

 atmosphere depends in some way on the amount previously absorbed. The facts 

 would perhaps have to be imagined as follows. The heat immediately reflected 

 by the Moon passes almost undiminished through the atmosphere, and thus causes 

 the rapid rise after totality, while the emitted heat is largely absorbed, so much 

 the more the cooler the atmosphere is. Thus this absorbed quantity of heat 

 increases steadily with the progress of the eclipse ; it reaches a maximum towards 

 the end of totality (or, in other words, the total measured becomes a minimum at 

 this epoch) and begins then steadily to decrease again. The heat measured after 

 the end of the eclipse falls thus short of the Full Moon value by the amount of 

 emitted heat which the atmosphere has absorbed, and rises slowly until the 

 atmosphere is, so to speak, saturated, or the maximum of possible absorption has 

 taken place, i. e. until the quantity of heat corresponding to Full Moon has been 

 reached. The total heat measured after an eclipse must thus be inversely 

 proportional to the duration of an eclipse. If the above reasoning holds good, 

 the gradually rising heat-curve of 1884 would be the more probable one. It is 

 well known that the idea of a very considerable absorption of the lunar heat by 

 our atmosphere was familiar to Sir John Herschel, as seen from the following 

 remarks {Outlines of Astronomy, 1873, p. 285): — "Though the surface of the Full 

 Moon exposed to us must necessarily be very much heated, . . . yet we feel no 

 heat from it. . . . No doubt, therefore, its heat (conformably to what is observed 



