520 On the Invisible Solar and Lunar Spectrum. 



between them, so that, even taking regions of transmission 

 and absorption together, on the whole here, i. e. above 3% 

 more of dark heat in proportion is, perhaps, transmitted than 

 of light heat. 



We may roughly illustrate* this portion of the sun's spec- 

 trum after absorption by saying that, if it were visible, it 

 would present almost the appearance of diffuse luminous 

 bands on a dark field, somewhat like those seen in the spectra 

 of stars of the fourth type. As we have observed that absorp- 

 tion seems to increase in the same direction in the sun's own 

 atmosphere as in ours, it would be interesting to know if any 

 thing analogous exists in the absolute solar spectrum, i. e. 

 before absorption in the sun's atmosphere, but this we are not 

 as yet able to determine. 



Since, in the part of the spectrum described here for the 

 first time, the same process of aggregation into ever wider 

 and wider regions of absorption is continued till these interme- 

 diate regions of transmission disappear, the whole, to repeat 

 an expression I have used in a earlier memoir, seems to become 

 one continuous cold band, in which, however, we have found 

 a little heat struggling through in the part beyond 11/*. 

 Briefly, then, we may say, that to an eye which could see the 

 whole spectrum, visible and invisible, the luminous part being, 

 as we know, interrupted by occasional dark lines, the lower 

 part to 5^ would appear to consist of alternate bright and 

 dark bands, and the part below 5^ be nearly dark, but with 

 feeble "bright" bands at intervals. 



In conclusion, we may say that these new researches extend 

 the known solar spectrum from three to much over eighteen 

 microns, shown on our plate, and while confirming the previ- 

 ously announced fact that the solar heat which reaches us 

 here is negligible in amount, show from the fact of the exis- 

 tence of any at all, that the anomaly of our being able to 

 perceive lunar heat where we could not formerly detect solar, 

 can be explained consistently with the possible existence in 

 the latter of every wave-length before absorption. 



These investigations into a problem of Solar Physics have 

 also incidentally led us to the prospective means of solution 

 of an important one in Meteorology, for they have opened to 

 observation the hitherto unknown region of the spectrum in 

 which the nocturnal and diurnal radiations, not only from the 

 moon toward the earth, but from the soil of the earth toward 

 space, are to be found and may be hereafter studied in detail. 



* Figure 3. 



