386 NATURE 
of some hidden physical causes whose exact mode of operation 
isnot known. We all know, however, that variations are facts 
of nature, and it is not difficult to see that they are the necessary 
consequences of the varying number, amount, proportion, and 
manner of action, of the natural forces acting on different portions 
of living matter. Now, in making variations the starting-point, 
the theory of natural selection may justly be considered to be 
“incomplete,” even as our knowledge of electricity is incom- 
plete because we do not know the real nature of the thing, as 
astronomy is incomplete because we do not know for certain 
how, for instance, the solar system was formed ; but in so doing 
the theory cannot be ‘‘ essentially faulty” or ‘ fundamentally 
erroneous.” It is illogical, not to say childish, to think a theory 
to be erroneous because it cannot render a definite explanation 
of some unquestionable facts of nature on which it is based and 
with which it starts. To prove the theory of natural selection 
to be erroneous, it must be shown that it is never competent to 
originate species. If it ever falls, it will fall quite z7respectively 
of its avowed inability to give definite explanations as to the 
exact mode of occurrence of variations. S. B. Mirra 
19, Keppel Street, Russell Square, London, August 17 
Red Sunsets and Volcanic Eruptions 
Pror. S. NEwcomp’s article on the above subject in NATURE 
of August 12 (p. 340) induces me to send you a brief account 
of the atmospheric phenomena that I observed in Palermo 
during and after the recent eruption of Mount Etna, 
This volcano is 133 kilometres distant from Palermo, but the 
transparency of the air here is so great that it is almost always 
visible from this Observatory. 
At dawn on May 21 the smoke from the eruption appeared 
as a great mass of black vapour, rising from the southern side of 
the volcano. At If am. it had formed into rosy balls of 
vapour, orcumuli. With the theodolite I measured the angular 
height—2° 21’, which gives 8 kilometres of linear altitude. On 
May 24 the smoke had the characteristic form of a pine-tree, 
and a greater height, but at 4 p.m. the upper edge of it was not 
well defined, and I obtained (approximately) the altitude as 
14 kilometres. 
Since May 22 these vapours from Etna have spread over the 
eastern, and more recently over the entire, horizon of Palermo. 
In the early morning of June 3 the fog was so dense that the 
sun was invisible, and the towers of the Matrice, 200 metres 
distant, were only indistinctly visible, which in Palermo is quite 
extraordinary. From May 29 to June 3 Italy has been invaded 
from south to north with mist, which was probably also derived 
from Etna. 
Cinders from the volcano have fallen over Eastern and 
Southern Sicily, and over Calabria, as well as in Palermo, where, 
in the dust gathered on the terraces of the Observatory on May 
27, I detected with the microscope some minute crystalline 
laminz of labradorite, which mineral is characteristic of the 
ejections of Etna. 
The sun rising from the sea behind these mists has been 
purple-red and then reddish-yellow ; at a height of about 30° it 
was neutral gray, but never green or blue. In Nicolosi, too, 
on the side of the volcano, these colours of the sun have not been 
observed. The light of the red sun was so faint that it was 
possible to look at it without inconvenience. 
No corona (like ‘‘ Bishop’s ring’’) was observed around the 
sun or moon. Spectroscopic observations of the red sun gave 
only the ordinary atmospheric absorptions, perhaps somewhat 
intensified. 
In the latter part of May and during the month of June the 
red after-glows appeared almost daily, and were stronger than 
before or since, but they were not so brilliant and prolonged as 
in 1831 and 1883-84, and their colour was not properly rosy, 
but an impure reddish-yellow. 
I believe that the red sun was caused by the finest cinders 
from the volcano, suspended in the air, as the like phenomenon 
is produced by the dust of the Hoherauch in Northern Europe, 
of the Sirocco in Sicily, and of the Kamsinin Africa. The blue 
sun (observed after the eruptions of Ferdinandea and of Krakatdo) 
has not appeared, and the after-glows were not strongly brilliant, 
because the vapours ejected from Etna were not so enormously 
abundant as those ejected from Ferdinandea and Krakatao, which 
are marine volcanoes more directly communicating with the 
water of the sea. 
The observation by M. Janssen, mentioned in NATURE of 
[August 26, 1886 
July 29 (p. 299), of a blood-red coloration of light traversing 
dust, gives a strong confirmation to the preceding explanaticn 
of the red sun. A. Ricco 
Palermo Observatory, August 17 
P.S.—Since July Bishop’s ring 
has not been visible in 
Palermo. 
The Bright Clouds and the Aurora 
On the morning of the 11th inst. I had an opportunity of 
watching the curious cirrus-like clouds as daylight came on. 
The display was striking, though not such a bright one as on 
several former occasions. It first appeared about 2.30 a.m., 
when there were very faint indications of the clouds; it was 
some minutes before I noticed that they were the same brilliant 
kind as has appeared so often this summer. Their apparent 
upper border being irregular, it was uncertain whether they in 
any part reached the limit to which the sun could shine upon 
them, or whether the apparent border was altogether the actual 
edge of the cloud-sheet ; however, it rose higher as the sun 
approached the horizon, but this might be owing to the circum- 
stance that the motion of the clouds was, as usual, from an 
easterly direction. At 3.334 a.m. they were visible as far as a 
Andromedz, though they were very faint west of y. By 3.45% 
a.m. they reached down to within 5° of a Aquila, and were 
rather plain there, and by this time the sheet covered most of 
the sky, though none of it remained visible very low down in 
the east. It was no longer bright in any part. At 3.55% they 
reached down to within 4° of a Aquilz, and were plainest about 
there, but growing fainter. I was still uncertain whether the 
sheet extended beyond the western apparent border, that being 
simply the limit of sunshine, or whether the sheet ended there ; 
but probably the former was now the case. At 4 a.m. they 
were scarcely noticeable, and by 4.11 they had disappeared 
altogether. By this time a faint pink glow had appeared in the 
east. 
The question is, Was the disappearance of the clouds due to 
their having evaporated, and ceased to exist, or to their light 
being overpowered by the brightness of the sky? It appeared 
to me that the latter was the case. It will be well if further 
observers can confirm this supposition or otherwise ; if correct, 
they cannot be considered clouds at all in the ordinary sense, 
the sky being beautifully clear and blue after they had ceased to 
be visible. I could not say at any time that the clouds were 
not perfectly transparent to the stars. The circumstance that 
they have never been described as having been seen by day 
seems confirmatory of the above supposition. 
With respect to Prof. Smyth’s remark about the spectrum (p. 311) 
Ido not gather whether he considers that the auroral line noticed 
by him belonged at all to the clouds or entirely to the aurora ; 
but I think that there can be no doubt that the latter was really 
the case. He does not seem to have detected any aurora at the 
spot where the clouds were seen, but doubtless it was there, 
although overpowered by their brightness. As it is so evident 
that these clouds were illuminated by the sun (this being con- 
firmed by their varied colours depending on their altitude, as 
described by Prof. Smyth), we cannot expect their spectrum to 
be otherwise than solar and atmospheric. I looked at them with 
a miniature spectroscope on the evening of July 12, as well as 
on the morning of the 11th inst., but on neither occasion was 
the spectrum bright enough for me to perceive much. I could 
not see any lines, bright or dark, but the spectrum faded very 
abruptly in passing from green to orange, which no doubt was 
owing to the atmospheric bands near D, especially the ‘flow 
sun band.” 
On July 27 I saw the aurora mentioned by Prof. Smyth at 
Gilsland, in Cumberland, and it was a particularly magnificent 
one there, especially about 10.25 p.m., at which time a part of 
it was lilac—a very unusual colour. The bright clouds were 
also visible that night, but chiefly before the aurora appeared 
and after it vanished ; there appears no reason to suppose there 
is any connection between the two phenomena. 
As regards the dark space beneath the auroral arch, has; the 
theory mentioned by Prof. Smyth ever been proved, that there 
is any true darkness there, and that it is not merely the effect of 
contrast with the aurora? My impression is that it must be 
at least mainly the effect of contrast, though perhaps not 
entirely, and the idea is confirmed by a similar darkness some- 
times appearing by contrast with the brilliant clouds, when no 
