JuLy 24, 1902 | 
NATURE 
295 
brightest at gh. 1om., 45 or 50m. after sunset, when the sun 
would be 53° below the horizon, its real height above the 
earth, making proper allowance for its low altitude of 20° and 
for the sunlight’s refraction, was about 13 miles; while its low 
angular distance from the sun probably denoted a not very 
excessive fineness of the sunlit haze material. 
On June 21, at 8h. 53-56m., or about 30m. after sunset, a 
wide expanse of pink glow was seen in the upper parts of three 
broad streamers radiating not far from vertically upwards, to 
altitudes of about 30°, from the sun’s place. The streamers 
sank in altitude to about 20° by gh., and subsided gradually by 
about gh. 5m. into the summit of a ruddy yellow light belt 
extending about 90° along the horizon in the sunset quarter to 
altitudes of $°-15°. This bright tract grew duller yellow and 
orange at its base and borders, until about 9b. 20-25m., when 
it had nearly faded out. At 30-35m. after sunset, when the 
streamers’ crests at an altitude of 20-30° glowed visibly with 
rosy light, the sun was 33° or 4° below the horizon here, and the 
height of the red coronal glow produced there by the sun’s 
parting rays must have been about 64-8} miles above the 
earth, 
For nearly half an hour after sunset on June 27, the clear 
north-western sky showed only weak dull shades of yellow, but 
a long low belt of this at the horizon attained some intensity at 
Sh. 45-50m., when a short arched band of level cloud streaks 
grew pink at first and then bright crimson on their lower edges, 
and from $h. 50m. to gh. presented there a splendid interlacing 
network of red stripes about 35° long and 4°high. At the latter 
hour pink streamers radiating from the hidden sun as centre and 
springing chiefly from the northern upper portion of the yellow 
light belt began to be visible, two immensely long and very 
narrow straight ones leaving it at about N.N.W., altitude 8°, 
with slopes of 15°-20° from horizontal, and reaching out, 3° or 
4° apart, through 53° or 60° to somewhat beyond N.N.E., ending 
at altitudes there of 15°-20°. The upper and stronger one was 
pink towards its end, but where they joined the light belt 
most brightly, and passed through north at altitudes of 
about 10-12", they partook of the light belt’s reddish 
yellow colour; above them some much shorter and weaker 
pinkish streamers soon appeared, and at about gh. 15m. 
two rather tall wide patches of faint rosy pink were formed 
at an altitude of 25° or 30° above the sunset place, by two 
nearly vertical wide streamers, and remained visible with 
pink colour for some minutes. While travelling here by 
train from London during its appearance, I could not note the 
early stages of this glow’s commencement, and my whole view 
of it was very partial; but from the pink tint’s visibility from 
gh. to gh. 15m. atan altitude of about 20-30", at about 40 or 50m. 
after sunset here, when the sun was 4}°-6° below the horizon, 
the heights of the pink crests of the radiating streamers would 
seem to have been, not very accurately, about 124-20 miles 
above the earth. 
A most complete view, however, of the successive features of 
the purple glow was obtained here on the evening of Saturday, 
June 28, when the concluded height of its appearance so sur- 
prised me by its unexpected lowness as to lead me to examine 
also the foregoing observations with a view to a general com- 
parison together of the glow’s real heights that would be found 
to be derived from my notes of it on different evenings. The 
sky was then streaked with cirrostratus cloud-seams ruling it 
with fan-like convergence towards about the sunset quarter ; 
but except in that direction those fleecy stripes dispersed by 
about 8h. 4om., and the nearly clear north-western sky half-way 
to the zenith was pale yellow, passing above about that altitude 
into greyish white, and beyond the zenith into blue. At 
Sh. 50-53m., the yellow sky-zone’s colour having deepened and 
the grey-white tract above having descended to about altitude 
40-45 °, the latter space grew rapidly rose-pink and round its 
centre a nearly circular field about 40° in diameter displayed 
pale pink oleander-flower or almond-blossom colour. Ending 
upwards, under this, in light straw-yellow, lay the wide-arched 
summit of a pretty strong horizon glow about 20” high over the 
sun’s place and about 30° long in span to either side of it, of 
ochre-tinted yellow. The rosé-pink coloured space sank 
gradually, or died out from above, between 8h. 55m. and 
gh., replaced from behind by greyish and blue sky and invading 
the pale straw-yellow summit of the arched horizon-glow, which 
together with that whole glow, by about gh. 5m., grew orange- 
red throughout. This litharge-red, and a little while later fan- 
tail-looking, glow contracted slowly downwards until 9h. 15m., 
NO. 1708, VOL. 66] 
when as it was growing dim and inconspicuous I ceased to 
watch it. But in the last rom, its upper border had in its usual 
way, when beginning to grow red, broken up into bright radial 
streamers crossing what remained visible of the cirrus streaks at 
such appreciable angles as to show them to be true solar light- 
beams quite unconnected with the wind-imparted radiation of 
those cloud-streaks from a near neighbouring but different focal 
centre. Where a long and well-defined straight radial streamer 
shooting up obliquely southwards crossed some of those faint 
cloud-streaks’ strongest ripples, their gauzy cloud materials 
certainly did not add to its brightness; but at the same time, 
they diminished the streamer’s light so much less than that of 
the grey sky immediately adjoining it, that they could hardly 
be said to have very distinctly screened and darkened it. 
As the sun’s parting illumination of the sky with rosy colour, 
from altitude about 50-60°, downwards, in this sunset, occurred 
(at about 8h. 55m.) not much more than 30m. after sunset here, 
when the sun was 34° below the horizon, it would follow that 
this red illumination by direct sunbeams, of microscopically fine 
haze matter took place at about 74 miles above the earth, or 
apparently not far from about the probable real heights of the 
simultaneously noticed cirrostratus cloud-streaks. 
The chief features of these recent after-glows having been just 
the same as those which were generally noticed during the 
gradual subsidence of the volcanic after-glow appearances in 
1883-4, since it was then pointed out by some investigators of 
their real heights that some white cloud-wisps looking phos- 
phorescently bright long after dark, and even sometimes, near 
the horizon in the north, throughout the night, must have been 
floating far above the ordinary height limits of rain and snow 
clouds produced by aqueous vapour, I was led by the com- 
parative lowness of these few new height conclusions to consult 
the original accounts given in NATURE by many good observers 
of the sky-glows in 1883-4, to recall more exactly than I could 
certainly remember what real heights had then been actually 
assigned to them. In such letters as Mr. F. A. R. Russell’s, at 
Haslemere (NATURE, vol. xxix. p. 55), describing the evening 
sky-glow on November 9, 1883, as having twice pervaded the 
sky with rosy red, beginning from overhead, first at 5h. and 
again at 5h. 4om. (or at 42m. and at 1th. 22m. after sunset at 
4h. 18m.), and as having gradually settled down into the 
greenish-yellow glare at the horizon in about 20-25m., it is 
quite evident that much loftier heights of the pink glow were 
then indicated than those which have just now again been 
essayed to be determined. The sun, at these two glows’ com 
mencements, would be about 53° and 114° below the Haslemere 
horizon, and the corresponding vertical heights over Haslemere 
of its earth-grazing beams would, allowing about 4° for their 
downward deflection by refraction, be about 17 and 75 miles 
above the earth. 
In letters from Mr. J. E. Clark at York and from Mr. J. LI. 
Bozward at Worcester (z/zd., pp. 130-131), the sky-glows 
from November 27 to December 4, 1883, were similarly 
described, in general, as usually attaining their strongest and 
brightest redness about one hour after the time of sunset, with 
durations afterwards of the fiery-looking dying-out phase of the 
glow for nearly an hour longer. Although these observations 
were not made with certain enough discrimination of the exact 
times of the rose-red tints’ commencements to afford very 
definite determinations of their real heights, yet in their records 
of about one hour after sunset, at which the whole height and 
width of the sky assumed an especially imposing kind of red 
magnificence, they were for the most part pretty perfectly ac- 
cordant. At an hour after sunset on December 1, the sun would 
have sunk about 7° below the horizons of York and: Worcester, 
and the height above the earth of fine dust-haze beginning to 
shine then overhead with red illumination would be about 
25 miles. But determinations of the pink glow’s real height 
by the method which has here been used, and of the warrant- 
ableness of which the roseate displays’ frequent collection into 
tufted heads of real sheaf-like sunbeam radiations is itself a 
sufficient proof to afford us {ull assurance, were in fact actually 
obtained on November 25-26 <nd 29, 1883, and were communis 
cated in letters (¢déd., p. 103 ard p. 130) by Annie Ley, 
at Ashby Parva, Leicestershire, and by R. von Helmholtz, 
in Berlin, who concluded it to have been at upwards of 13 and 
at about 4o (? more nearly 30) miles high respectively. The 
intervals alter sunset when the wide red glow began in these two 
latter cases were about 50-60m. and a little more than one hour, 
resembling the generality of other observers’ records, in those 
