652 
Proceedings of the Royal Society 
One reason why the red light is not visible overhead, is that the 
depth of the red reflecting stratum is not enough to stop out the 
great quantity of blue light coming from the sunlit air higher up ; 
and as this blue light is complementary to the red of the afterglow, 
the two combine, and cause the heavens to appear whitish, both 
colours being destroyed ; but the I^i'col prism, by cutting out the 
polarised light, reveals the red to the eye. By examining the sky 
overhead with the prism, it was always possible to tell, before the 
first afterglow in the west began, whether it Would be bright or not. 
If the red glow was not strong overhead When the polarised light 
was cut out, then the afterglow was sure to be poor. From this 
we see that the red seen in the east, shortly after sunset, rises and 
passes overhead unseen, but again becomes visible in the aurora- 
like glow in the west. 
Another reason why we see this afterglow in the west, though 
we cannot see it with the unassisted eye overhead, is, that when 
viewing it towards the horizon^ we are looking obliquely into the 
red stratum, and are thus receiving light from a greater depth of it 
than when looking upwards ; we are therefore receiving a greater 
amount of red light, and the greater thickness also helps to block 
out the blue light beyondi It is very evident, however, that this 
does not explain some of the peculiarities in the visibility of the 
first glow. If the first glow, as seen in the west, is due to the 
sifted rays of the sun falling on particles large enough to reflect the 
red rays^ thus producing in the atmosphere a Stratum of air full of 
red glowing particles, then we naturally ask ourselves, Why is not 
this red light visible to the unassisted eye when looking towards 
the north, South, and east, as well as towards the west ? If at sunset 
there is formed in the sky a red stratum, we can image it so thin 
as to be invisible when looked through overhead, but we should 
expect it to be equally brilliant in all directions, at the same 
elevation, or if it is more brilliant in one direction than another, we 
should expect the east, and not the west, to be most brilliant at 
sunset. Or to put this difflcUlty in another way. Why is it that 
this brilliant upper glow is visible only when viewed from a certain 
direction ? To an observer placed to the noHh, south, or west of it, 
it is quite invisible. We know, for instance, that a short time 
before this upper glow was visible to us on any evening, there 
