84 
NATURE 
[Fune 1, 1871 

conflagration has not reached stand in the clearest relief as they 
are seen for probably the last time ; but in a dozen spots, at both 
sides of the bridges, sheets of flame and awful volumes of smoke 
rise to the sky and positively obscure the light of the sun. Jam 
making these notes on the Trocadero. Close and immediately 
opposite to me is the Invalides, with its gilded dome shining 
brightly as ever.” i 
Another as follows :—‘‘ As I drive along the green margin of 
the placid Seine to St. Denis, the spectacle which the capital 
presents is one never to be forgotten. On its white houses the 
sun still smiles ; he will not refuse his beams spite of the deeds 
which they illumine. But up through the sunbeams struggle and 
surge ghastly swart waves and folds and pillars of dense smoke ; 
pot one or two, but I reckon them on my fingers till I lose the 
count.” 
Twenty-four hours later, the change has come. “The rain 
is now falling heavily, has been falling heavily all day, and 
may do something for burning Paris. The sound of artillery has 
died away ;” and from another writer :—‘‘ A heavy smoke hangs 
over Paris and rain is constantly falling.” 
I believe it has often been remarked that rain generally follows 
a heavy cannonading, but in this case there is an almost unex- 
ampled artillery fire and tremendous conflagration at the same 
time, accompanied by a sudden and violent change in the atmo- 
spheric conditions. From where I am writing we noticed a re- 
markable change on Thursday morning, and about 2 P.M., after 
_ intense closeness and oppression, a rain of a tropical character 
set in for twelve hours or more. On many occasions in Queens- 
land, I noticed that in seasons of drought, after extensive grass 
fires, causing intense heat, heavy thunderstorms generally fol- 
lowed. GEORGE PEAKCE SEROCOLD 
Rodborough Lodge, Stroud, May 27 
Alleged Daylight Auroras 
SEVERAL letters having appeared in recent numbers of 
Nature, giving what the writers consider to have been un- 
doubted instances of aurora visible in the daytime, you will, I 
hope, allow me to state the reasons why I still adhere to the 
views expressed in my former communication on this subject. * 
And, first of all, I must beg your correspondent Mr. Jeremiah 
not to think me uncourteous if I dismiss at once, as unworthy of 
serious criticism, the cases which he has dug out of monkish 
chronicles. It is likely enough that some of these old records 
may be imaginative descriptions of nocturnal auroras, and as 
such they are not without interest, but I cannot admit them as 
competent witnesses on a point of nicety. 
A more modern instance adduced by the same correspondent 
will be found at p. 7 of NATURE for May 4, under the title 
“ Aurora Borealis, seen in the daytime at Canonmills.” In this 
case it is difficult to know what relation is intended between the 
title and the account which follows. The account describes the 
clearing off of the clouds ina mass from the north-west, with 
the production of an “ azure arch,” the centre of which ‘‘ reached 
an elevation of 20°.” If I reply to this that the clearing off of 
clouds is not an aurora, even though they clear off in a compact 
body from the north-west, leaving an ‘‘azure arch,” I may be 
met by the rejoinder that nobody said it was ; and yet I strongly 
suspect that the writer had some confused idea that he was 
describing an auroral arch, and I am certain that nine out of ten 
readers, misled by the heading, would take the same view. 
Stripped of the cloud-phenomena, all that remains of the Canon- 
mills aurora is the appearance of some ‘‘very faint perpendicu- 
lar streaks of a sort of milky light,” which could be traced across 
the segment of blue sky, but were “ extremely slight and evane- 
scent.” Considering the probability that the observer regarded 
the cloud-arch as auroral, which it certainly was not, and con- 
sidering how his judgment would be likely to be biassed by that 
idea in the interpretation of ‘‘ extremely slight and evanescent” 
appearances, I think we may fairly regard this testimony as par- 
ticularly weak. 
In NATURE for Dec. 8, 1870, Mr. Cubitt describes and figures 
a double auroral are which he saw in broad daylight on the 25th 
October. It was ‘some 25° above the horizon, and almost due 
east.” In my first letter I expressed a doubt of the correctness 
of this observation on the ground that auroral ares are not seen in 
the east. My criticism has since been challenged on two 
distinct issues. Mr, Jeremiah insists that an auroral arc may 
* NaTurE, vol, iii. p. 126, 

extend towards the east, and that what Mr, Cubitt saw may have 
been the eastern extremity of a northern arc. A reference to 
Mr. Cubitt’s letter and illustration will show at once that if what 
he saw was any part of an arc, it was the apex and not an ex- 
tremity. But another correspondent, Mr. Reeks, in NATURE 
of Dec. 29, 1870, in criticising my remark, makes a statement 
which is more difficult to answer. He affirms positively that in 
Newfoundland he has many times seen the arch nearly due east, 
that is, as he explains, with ‘‘the extremities pointing N.N.W. 
and S.S.E.” I would suggest, however, in reply to this state- 
ment, that in an extensive auroral display there may be fictitious 
arches, produced by the accidental correspondence of streamerson 
either side of the ‘‘cupola.” An arch of this kind may easily 
extend from N.N.W. to S.S.E., spanning the entire heavens. 
It is essentially different from the true auroral arc, which, until 
much stronger evidence to the contrary is adduced, I shall still 
believe to be invariably transverse to the magnetic meridian. 
Obviously, Mr. Cubitt’s arc was not of the kind that Mr. Reeks 
describes. 
I pass on to a record of daylight aurora, which, more than any 
other that I have seen, demands a careful investigation. I refer 
to ‘‘An Account of an Aurora Borealis seen in full Sunshine, by 
the Rev. Henry Ussher, D.D.,” said to be taken from the 
Transactions of the Royal Irish Academy for 1788, and quoted 
by the Rev. T. W. Webb in Nature for May 11. Dr. Ussher’s 
account, it must be admitted, is most particular and complete. 
He describes “‘ whitish rays ascending from every part of the 
horizon, all tending to the pole of the dipping needle, where, at 
their union, they formed a small thin and white canopy similar 
to the luminous one exhibited by an aurora at night.” Nothing 
can be more precise. But is it not also a trifle too wonderful ? 
Surely; if any part of an aurora is to be seen by daylight, it must 
be just one here and there of the most vivid beams. That the 
whole phenomenon should be visible at noon-day in all its com- 
pleteness, just as at night, even to the faint extremities of the 
streamers in the magnetic zenith, is to my mind so entirely in- 
conceivable that not even the authority of a doctor of divinity 
can command my faith in it. I can much more easily believe 
that the sky presented a remarkably symmetrical arrangement of 
radiating cirri, and that the observer, impressed by the recollec- 
tion of the aurora of the previous evening, persuaded himself that 
the “rays coruscated from the horizon to their point of union.” 
The confirmation by ‘‘three different people” is of little yalue 
unless their observations were independent. 
To those who have no clear conception of the difference be- 
tween cirrus and aurora, the foregoing arguments will be meaning- 
less. Some persons write very loosely of ‘‘ luminous cirri,” and I 
have even seen described the transformation of cirrus cloud into 
aurora as it grew dark. I believe that there is no connection 
between the two phenomena beyond an occasional and purely 
accidental similarity of form, and that when the two co-exist, the 
cirrus, instead of being the seat of the aurora or deriving lumi- 
nosity from it, only serves to obscure its brightness, and, if dense 
enough, may appear in the form of dark bands across the auroral 
light, the latter being, as I conceive, at a very much greater 
elevation. 
I adverted in my former letter to the argument that may be 
drawn from the non-yisibility in the day-time of other lights com- 
parable with the aurora, and I will only now add the following 
suggestion. If the auroras that occur in this country are occa- 
sionally visible in daylight, it might be supposed that the much 
grander displays of the Arctic regions would be habitually visible 
in daylight. But is the fact so ? 
Clifton, May 23 GEORGE F, BURDER 

Aurora Australis 
TRAVERSING the Indian Ocean 44° S. 65° E., I observed, 
September 24th, 1870, 4h. till 13h. Greenwich time, a south 
polar light of great intensity and splendour. After my arrival at 
Manado (Celebes) I was just writing a few lines about it for the 
readers of NATURE, with the purpose of knowing whether at the 
same time an aurora, or at least disturbance of the magnetic 
needle, had been observed on the northern hemisphere, when I 
saw in Nature (Nos. 49, 50, and 51, 1870), several interesting 
descriptions of aurora borealis observed September 24 in England, 
&c. I am not aware whether many observations of southern 
polar lights have been recorded, but Iremember that those which 
Cook described in the year 1773 were coincident with aurora 
borealis observed in Friesland, and others observed in 1783 
ata 
