648 
Proceedings of the Royal Society 
about to set, the lower part of the western horizon became tinged 
with yellow, which deepened to orange as the orb touched the 
horizon. After it had sunk below the horizon, the band of colour 
in the west deepened and widened. The yello\^, which when first 
visible was close to the horizon, after a time rose higher, while the 
colour on the horizon gradually deepened to orange and at last to 
red, the upper limit of the colours at the same time rising higher 
and higher in the heavens. The amount to which the colour 
deepened, and the height to which it rose, varied from evening to 
evening. 
Tor some minutes after the sun had set, nothing very remark- 
able appeared in the sky. Overhead it was blue, and the blue 
gradually changed through white to a dull reddish colour on the 
horizon, in the north, south, and east; in the west the blue 
changed by imperceptible degrees from blue into white, which 
melted into yellow, and the yellow in turn deepened into orange 
and red on the horizon. But on most evenings, within a quarter 
of an hour after the sun had set, a very remarkable reddish 
glow made its appearance high up in the western sky. When 
this aurora-like glow first became visible, its upper edge would be 
about 40° above the horizon, and extended downwards to about 15° 
from the horizon. The colours of this glow were quite different 
from any of those on the horizon^ There the colours varied 
from yellow through orange to red; but this upper glow was a 
very different red from that on the horizon — perhaps crimson is the 
colours to which it most nearly approached. As time progressed, 
this glow gradually descended, till at last it merged into the sunset 
colours on the horizon, and the two bands melted into one. This 
upper glow generally took about 10 or 15 minutes, from the time it 
was distinctly visible, to descend and become absorbed in the 
horizon colours* Owing to the perspective, its upper edge seemed 
to descend more quickly than the lower, which made the glow 
appear to become narrower as it approached the horizon. 
After the sun had been from 20 to 30 minutes under the horizon, 
nothing remarkable was visible ; the upper glow having sunk into 
the horizon colours, the heavens had again much the same appear- 
ance they had before the upper glow began, only somewhat darker. 
But a few minutes after this upper glow had disappeared, a 
