May 18, 1871] 

such as electricity, driven off unequally from parts of the sun’s 
surface, and in directions radiating from points either on or 
beneath the surface.* If only a small portion of the sun’s surface, 
such as that covered by sun spots, sent out such streamers, the 
appearance might exactly coincide with that of the corona ; for 
these streamers, when seen projected on to the plane of the sun’s 
limb, might in some places appear to overlap so as to form a 
continuous corona, whereas in others they might appear to be 
separated by gaps. 
So far as the appearance is concerned, it would be the same 
whether the emission consisted of matter or was electricity ; there 
are, however, other indications of its being of the latter kind. 
The action which the sun exerts on terrestrial magnetism 
shows it to be in an electric state, and the observations of Stewart 
and others have established a connection between the variations 
in its electric condition and the changes in the sun-spots and red 
flames, and the observations on the recent eclipse have connected 
the red flames with the brighter parts of the corona. Here then 
we have a distinct and independent reason for assuming that the 
electric condition of the sun’s surface is partial and unequal, and 
for connecting the corona with this electricity. 
As I have already ventured to explain the solar corona, as 
well as comet’s tails and the aurora, to be a kind of electric brush, 
I now offer these remarks on the radial appearance of the corona 
in confirmation of my views. 
Owens College, May 8 OsBORNE REYNOLDS 
A few more Words on Daylight Auroras 
In NATuRE for December 29, 1870, there is a letter from 
Dr. G. F. Burder in reply to a previous correspondent, who 
had sent a description, with an illustration, of a Daylight Aurora 
observed by him, wherein he made the following statement :— 
But auroral arcs, as far as I know, never appear in the 
east, and the conclusion, therefore, is unavoidable, that the 
object observed was nothing more than a remarkably symmetri- 
cal form of cirrus cloud.” 
He then states his convictions that all records of so-called 
daylight auroras are ‘‘ errors of observation.” 
As assertions like these might have an undue influence on the 
minds of those who read my letter on ‘‘ Aurora by Daylight” 
(Nature, May 4, 1871), I am induced to say a few more words 
on this subject, especially to prove the fallacy of such reasoning. 
It is well known that the aurora borealis assumes innumerable 
shapes ; some of the most remarkable were given by me some 
time ago in these pages ;} and that they appear, at times, actually 
in the eas/, but more often in the north, north-west, and stretch- 
ing tothe south. A writer of some excellence in the last cen- 
tury § says :— 
‘*Sometimes the aurora appears like arches, nearly in the 
form of a rainbow, reaching from one point of the horizon to 
another. The arches always cross the meridian at right angles, 
tending to the cast and west point of the compass.” 
The correspondent whom Dr. Burder is so hard upon, most pro- 
bably saw the arc in a position nearer to this, than directly jacing 
him, with his back to the wes¢, and as the illustration sent by 
him shows only a segment of the arc, I am inclined to think that 
the extremities were nearly in the east and west.|| The ‘‘ cirrus 
cloud” hypothesis is simply untenable, when it is known positively 
that on several occasions the aurora was seen against an azure 
background, with no form of cloud in the field of view, as 
for instance that mentioned by me, where a faint arc was seen 
before sunset in the east (possibly N.E.) against a cloudless 
sky.4] His other assertion, from which I must dissent, was : 
“A comparison of the auroral light with the light of other 
objects whose visibility can be more easily measured, tends 
strongly to confirm the view I have advanced.” He then instances 
the invisibility of Donati’s comet by daylight. He might have 
instanced the invisibility of the stars, although they can be seen 
in broad daylight when the observer is placed at the bottom of a 
deep pit ; but this need not be done, for daylight does not always 
mean bright sunshine ; and with diffused light, Venus is often 
seen before the sun has actually gone below the horizon. This, 
so far, may appear mere assertion, but the following will, I hope, 
* We are of opinion that there is still another explanation.—Ed. 
t+ Nature, Dec, 8, 1871. } Ibid, Dec. 29, 1871. 
§ “Compendious System of Astronomy,” by Margaret Bryan. 
3797, P- 132. } ‘ 
|| For evidences of night auroras being seen in the east, consult the letters 
in Nature for 1870, 
§] Nature, May 4, 1871, p. 8. 
London : 
NATURE 


47 

be sufficient to show that the view he holds requires some kind 
of modification. 
“*a.D, 678.—This year the star (called a comet) appeared in 
August and shone like a sunbeam every morning for three 
months” (Anglo-Sax. Chron.).* By every morning I take to 
mean daylight, because in these months the mornings would 
invariably be very light, especially the few moments before the 
comet actually disappeared. 
With regard to all the record of daylight auroras being mere 
“errors of observation,” Iam sure no one will continue to enter- 
tain such an opinion after carefully examining all that has been 
said upon the subject in these pages. As it may be useful to 
those who are interested in this question, I have made a summary 
of all the daylight auroras recorded in this and other publica- 
tions, which do not admit of doubt. 
A.D, 1122. A phenomenon appeared like a great and broad 
fire, and lasted till it was quite light. (Anglo-Sax. Chron.) 
A.D. 1467. A most probable day aurora, described as *‘ horse- 
men and men in armour rushing through the air.”+ (Ingulf, 
Second Cont.) 
A.D, 1788. May 5 at Ir A.M. an auroral display seen, con- 
sisting of ‘‘ whitish rays ascending from every part of horizon.” 
Observed by “three different people.” (Trans. Royal I, 
Academy for 1788, quoted by Rey. T. W. Webb in Nature, 
May 11, 1871. 
A.D. 1827. ‘‘ Aurora Borealis seen in the day-time at Canon- 
mills” at 4.30 P.M. -Described in ‘‘ Jameson’s Journal” and 
AES for May 4, 1871. (Arcana of Science and Art for 
1828.) 
A.D. 1849.—In September an aurora seen, consisting of ‘three 
slightly diverging beams of light on the eastern horizon. One 
might have taken them for beams from asetting sun. . , had 
it not been that they did not emanate exactly from the spot where 
the sun had set ; that they had an evident motion to the southard, 
and that two of them extended to the zenith, and finally down to 
the eastern horizon.” { (Mr. J. Langton,in Narureg, April 27, 
1871.) 
A.D. 1870.—September 4, about 4.30 P.M., an aurora observed 
‘in the form of thin reddish streaks.” (‘*S. B.” in Nature, 
October 13, 1870.) 
A.D. 1870.—October 25, at 4.30 P.M., a brilliant aurora seen 
in the east, and fully described with illustration of it. (NATURE, 
December 8, 1870.) 
A.D. 1870.—December. A frobadle auroral display, which 
was observed a “little before sunset,” and developed as the even- 
ing advanced into a brilliant aurora borealis. The day phe- 
nomenon, however, not sufficiently described to make the record 
trustworthy. (J. Langton, in NATURE, April 27.) 
A.D. 1871.—April 10, about 4.30 P.M., a whitish arc seen, 
almost east, against a cloudless azure sky. On the previous 
night there was a magnificent aurora borealis. § (Mentioned by me 
in NATURE, May 4.) 
These form the whole of the most reliable records, which are 
certainly few, for the period embraced between the earliest and 
present date ; but I am inclined to believe that the occurrence 
of daylight auroras is not so rare as is here shown, but that 
they have been seen and actually recorded in the works which I 
have here and elsewhere quoted, but for the want of the state- 
ment of the time of day or night, one cannot tell to which the 
appearance belongs. Often a display which can be said to have 
been seen in the night, might as easily be said to belong to the 
day, so far as the actual wording of the record goes. It will 
follow from this that the scanty records we have of daylight 
auroras referred to phenomena of extraordinary magnitude and 
magnificence, JOHN JEREMIAH 
The Conservation of Force 
I HAVE been endeavouring to understand what is meant by 
the Conservation of Force; and as it is one of the most interesting 
subjects I have studied, I send you the result of my labours. 
* This is confirmed in Beda, Flor. of Wor. (under A.D. 677) and Chronicum 
Scotorum (under A.D. 673 in error for 677). 
+ See also Pliny, Bk. II. c. lvii. 
t This very singularly explains the following passage in Pliny :—‘‘ Round 
about the sun there was seen an arch when Lu. Opinius and Q. Fabius were 
consuls” (about B.c. 123). This was not an ordinary halo, for he says 
further :—‘“‘and a circle when L. Porcius and M. Acilius were consuls.” 
(Bk. ii. c. xxix.) 
§ It appears curious that the majority of the displays occurred at or about 
4.30 P.M., in the autumn, winter, and spring months. Cases of magnetic 
disturbances during this hour are not rare, accompanying the aurora. It 
may prove of some value to note this. eimas 
