Sune 19, 1873] 
—, 
Bye an ° 
- NATURE 
149 
THE CORONAL ATMOSPHERE OF THE SUN* 
Il, 
V HEN the subject is a phenomenon se complex as 
that of the corona, it is necessary to bring to bear 
upon it various methods of study. This is why I have 
thought it indispensable to consider the corona from the 
triple standpoint of its aspect, the analysis of its light, 
and its polariscopic manifestations. Let us discuss these 
varied observations. 
Let us first of fall see what can be learned from an 
examination of the corona during the first instants of 
totality. We have seen that the general structure of the 
corona persisted throughout the duration of totality. We 
cannot, then, admit here any effect of diffraction engen- 
dered at the surface of the lunar screen by the rays 
grazing the edges of that screen Let us revert to 
the geometric circumstances of a total eclipse. At the 
moment when totality is produced, the disc of the moon is 
tangent, at one point, to that of the sun, and edges off 
gradually more and more to the opposite point. Diffrac- 
tion will be produced, then, under physical conditions the 
most different, at) various points of the lunar limb, and 
an aureole due to that cause will reveal, by its dissym- 
metry, such a diversity of conditions. Moreover, an 
aureole of this kind will present a continually varying 
aspect during the various phases of totality. Unsym- 
metrical at the outset, it will be modified with the move- 
ment of the moon, and will tend to assume the same 
form all round our satellite, when the disc of the latter 
is equidistant from that of the sun. Finally, from that 
point this aureole will pass through the same phases 
inversely until the reappearance of the sun. 
However, nothing like this was produced at Shoolor. 
The general structure of the corona remained the same 
throughout the continuance of totality.+ 
It is unnecessary to dwell on the hypothesis of an 
aureole produced by a lunar atmosphere. We know now 
that if a gaseous layer exists on the surface of our satel- 
lite, it must be of so small extent that the grand pheno- 
menon of a corona could not be produced by it. 
Our own atmosphere cannot be adduced as the cause 
of the phenomenon, though it is evident that it plays an 
important part in the particular aspects which the corona 
may present at different stations, according to the state 
of the sky at these stations, It acts as a modifying, but 
not as a producing cause. 
Let us pass, meanwhile, to the spectroscopic observa- 
tions. The corona presents the hydrogen lines through- 
out all its visible extent ; in certain parts as far as to 12’ 
or 15’in height. This observation is certain. The pre- 
cision of the spectroscopic scales, the experience we have 
had in such determinations, and the care which was taken 
in the last observation to compare the lines of the corona 
with those of a protuberance, of which they are only a 
prolongation, leaves no doubt as to this point. 
But if the corona presents the hydrogen lines, we must 
ask this testing question—Is this light emitted or re- 
flected? The constitution of the coronal spectrum will 
afford us an answer. 
If the light of the corona is reflected, this light can 
only have a solar origin. It proceeds from the photo- 
sphere and the chromosphere, and its spectrum ought to 
be that of the sun, that is, a luminous ground with ob- 
scure lines. But such is not the constitution of the 
coronal spectrum ; that presents to us the hydrogen lines 
standing in strong relief on the ground ; after the green 
line (1474) this is the most striking manifestation in the 
phenomenon. We must conclude that the coronal medium 
is self-lighted, in great part at least, and that it contains 
* Continued from p. 127. 
+ It is quite evident that this constancy of aspect only agrees with points 
of general structure sufficiently distant from the sun not to be influenced by 
variations of light resulting from the displacement of the moon, relatively 
to the low and very luminous regions of the chromosphere, 
incandescent hydrogen, This first point is conclusively 
established. But is it to be inferred from this that the 
whole of the light of the corona is emitted light? Evi- 
dently not ; and on this point a delicate observation in 
spectrum analysis and polarisation may inform us. In 
fact, the spectrum of the corona presented to me, besides 
these bright lines, many obscure lines of the solar spec- 
trum, the line D,and some in the green. This fact proves 
the presence of reflected solar light. We may ask why 
the principal Fraiinhofer lines are reduced to the line D. 
It should be remarked that the coronal spectrum, not 
being very luminous, is especially perceptible in its central 
part, and that, in this part, the lines C, F, &c., are re- 
placed by the bright lines. In these conditions the line 
D alone remains important; thus it is on it I have 
directed all my attention. As to the finer lines, they were 
much more difficult to discern, a fact very easily explained 
by the very large opening I was obliged to give to the 
slit of the spectroscope. 
The proof of the existence of the Fraiinhofer lines in 
the spectrum of the corona is a work of delicacy ; it was 
not obtained by the other observers. This fact is ex- 
plained partly by the great purity of the sky at Shoolor, 
partly by the power of my instrument. I have no doubt 
that the observation will be confirmed by astronomers 
who work under conditions equally favourable. 
The presence of reflected solar light in the spectrum of 
the corona is of great importance ; it shows the double 
origin of this coronal light ; it explains observations of 
polarisation which appeared irreconcilable ;* but above 
all, it enables us to understand how the solar light form- 
ing in some sort the ground of the spectrum of the corona, 
this spectrum may be considered continuous; and we 
know that hitherto this circumstance has been the great 
obstacle which prevented the corona from being regarded 
as entirely gaseous. The phenomena of polarisation pre- 
sented by the corona are for the most part those of radial 
polarisation, which shows that reflection takes place 
chiefly in the corona, and that that which may be pro- 
duced in our atmosphere is only secondary. Polarisation 
then agrees here with my observation of the Fraiinhofer 
lines ; but in order that the agreement may be complete, 
it is necessary that the polariscopic analysis, like the 
spectral analysis, should show that the light of the corona 
is only partially reflected. This is precisely what hap- 
pened, We have seen, in fact, that near the limb of the 
moon, where the coronal light is brightest, polarisation 
appears less pronounced than ata certain distance. The 
reason is, that in the inferior regions emission is so strong 
that it conceals reflection, and the latter appears, with its 
peculiar characteristics, only in the layers where it is 
able to assume a certain relative importance. 
Thus the two analyses, spectral and polariscopic, fairly 
interpreted, agree as to the double origin of the coronal 
light, and all the observations unite in demonstrating the 
existence of this circumsolar medium. This medium is 
distinguished both by its temperature and by its density 
from the chromosphere, of which the limit, moreover, is 
perfectly distinct, as is shown in all the drawings ofthe 
protuberances and of the chromosphere. There is thus 
a necessity for giving ita name: I propose that of “coro- 
nal envelope” or “coronal atmosphere,” to remind us 
that the luminous phenomena of the corona owe to it 
their origin. ; 
The density of the coronal atmosphere must be exces- 
sively rare. In fact, itis known that the spectrum of the 
chromosphere in its superior parts is that of a hydrogen 
medium successively rarified ; but as the coronal medium, 
according to the indications of the spectrum, ought to be 
even infinitely less dense, we see how rare this medium 
* If we consult the history of eclipses we shall see that observers have 
often obtained contrary results, which has been the means of casting a kind 
of discredit on this kind of observation. But if these observations are con- 
sidered in view of the double nature of the light of the corona, and of the 
effects of our atmosphere, we shall be able to remove most of the difficulties. 
