294 



THE POPULAR SCIEXCE ATOXTULY.— SUPPLEMENT. 



portions. The whole region of the sky occupied 

 by the solar corona is unquestionably illuminated 

 by that corona. Some part of the light received 

 from that region must therefore be due to the 

 illumination of our own air. The question 

 arises, then, Where does the true solar corona 

 end, and where does this atmospheric illumina- 

 tion begin ? Or rather (for the atmospheric illu- 

 mination covers the entire heavens, while the solar 

 corona probably has no definite limits), where 

 does the atmospheric illumination begin to over- 

 come the light of the true solar corona ? 



The question is a difficult one. We cannot 

 deal successfully with it, as with the general 

 question of the corona, by mere reasoning, for 

 we have not yet sufficient observational evidence. 

 Still we can form a tolerably definite opinion on 

 one or two point*. 



In the first place, the illumination of the co- 

 ronal region of the sky, by the true solar corona, 

 must be, to all intents and purposes, uniform. If 

 we imagine an observer placed anywhere in the 

 air (say not higher than one hundred miles from 

 the earth), so that he would seem, as seen by the 

 observer of mid-totality, to occupy some point on 

 that region of the sky — in other words, if he 

 were placed anywhere in that inclined, well- 

 shfipcd portion of our air occupied at the mo- 

 ment by the moon's shadow — we know that he 

 would see the whole of the corona, but no part 

 of the sun's true body. Such an observer, placed 

 on one side of that well-shaped shadow-region, 

 would see the moon just touching the sun's edge 

 on one side; an observer on the other side of the 

 shadow-region would see the moon just touching 

 the sun's edge on the other side ; and an observer 

 placed on the axis of the shadow would see the 

 moon centrally concealing the sun. But the 

 quantity of coronal light seen by all three would 

 be appreciably the same. It follows (or rather 

 it is another way of expressing the same thing) 

 that every point in that region of the air — the re- 

 gion lying between the observer of central totali- 

 ty and the coronal region of the sky — is equally 

 illuminated by the solar corona. The variation 

 is, at any rate, very small. Now, there is one 

 part of that region of the sky where the solar 

 corona itself is not shining. I mean the part oc- 

 cupied by the moon's disk. This part of the sky, 

 however, is as brightly illuminated as the rest by 

 the solar corona. We learn, then, how much of 

 the light coming from the coronal region of the 

 sky is due to the illumination of our air by the 

 solar corona ninety millions of miles beyond. 

 This illumination of our air would of itself make 



the coronal region no brighter than the disk of 

 the moon appears as seen during total eclipse. 

 Since the moon appears to ordinary eyesight 

 quite black in such an eclipse, it might seem as 

 though this conclusion were decisive of the whole 

 matter, and that no appreciable illumination of 

 the air is caused by the light of the solar corona. 

 But this conclusion would be incorrect. The 

 moon's body during total solar eclipse is not 

 dark ; it only appears so by contrast with the 

 brilliant light of the inner part of the corona. 

 A certain faint light can be detected when the 

 telescopic field of view is so reduced that the 

 corona is excluded. 



But here another difficulty presents itself. 

 At the time of mid-totality, not only is the part 

 of the sky occupied by the moon's disk illumi- 

 nated by the solar corona, but the moon's disk 

 is itself illuminated by the light of our earth. 

 When the moon is new to us, our earth is full to 

 the moon. Now, at a moderate computation full 

 earthlight on the moon is equal to about sixteen 

 times as much as full moonlight on the earth. It 

 is true the disk of the earth only appears about 

 thirteen times as large at the moon, as the moon 

 appears to us ; but the earth's surface (if we can 

 judge from Mars and Venus) is, on the whole, 

 more reflective than the moon's. On the other 

 hand, we must not forget that the moon's surface 

 is thus feebly reflective, or, to speak plainly, that 

 the moon is of such dark tints on the Avhole; nor 

 must we forget that though at the time of new 

 moon the earth is full to the lunarians, if there 

 are any, at the time of a total solar eclipse the 

 earth seen from the moon shows the moon's 

 shadow. She loses about a thirteenth part of 

 her "full" brightness when thus in eclipse. 

 Still, taking all these considerations into account, 

 the moon's surface at the time of a total solar 

 eclipse must be shining in reality at least ten 

 times as brightly as the surface of a distant hill 

 illuminated by the full moon. We know how 

 white the earth looks when bathed in full moon- 

 light; and we can infer how much more brilliant- 

 ly white the moon's surface must be when bathed 

 in full earthlight. Distance makes no manner of 

 difference in this brightness; though of course it 

 affects the quantity of light sent us, making the 

 moon, in fact, appear so much the smaller as she 

 is farther away. Thus the black body, as we are 

 apt to call it, of the moon, at the time of total 

 solar eclipse, is in reality ten times as bright as a 

 part of a distant hill directly illuminated at mid- 

 night by the full moon. 



We learn, then, ou the one hand, that the 



