Nov. 30, 1882] 
NATURE 
Moe 
phenomena existing beyond the visible surface of the sun. 
It was then that was solved the enigma so long pondered 
over regarding the nature of those roseate protuberances 
which surround in such a singular way the limb of the 
eclipsed sun. 
Spectral analysis taught us that they were immense 
appendages belonging to the sun, and formed almost 
exclusively of incandescent hydrogen gas. Almost im- 
mediately, the method suggested by this same eclipse, 
and which allows of a daily study of those phenomena, 
revealed the relations of those protuberances to the solar 
globe. It was perceived that the protuberances are 
merely jets, expansions of a layer of gas and vapours, 
8” to 12” in thickness, where the hydrogen preponderates, 
and which is at a very high temperature, by reason of its 
contact with the surface of the sun. This low atmosphere 
is the seat of frequent eruptions of vapours coming from 
the solar globe, and among which one chiefly observes 
sodium, magnesium, and calcium. We may even sup- 
pose that, in the lowest part of this chromosphere, as it 
has been called, most of the vapours, which in the solar 
spectrum, produce the dark lines it presents, exist in the | 
state of high incandescence. 
The eclipse of 1869, which was visible in America, 
allowed indeed of the important observation (always con- 
firmed since) of the reversal ot the solar spectrum at the 
extreme border of the disc, that is to say, at the points 
where the photosphere is immediately in contact with 
the chromosphere ; a phenomenon which does not signify 
that the photosphere itself may not contain the same 
vapours and concur in the production of the solar spectral 
lines. 
Thus the discovery of a new solar envelope, the recog- 
nised nature of the protuberances, and the knowledge of 
their relation to the sun; lastly, the conquest of a method 
for the daily study of those phenomena ; such were the 
fruits of spectrum analysis applied to the study of this 
long eclipse of 1868. 
But a total eclipse presents other manifestations com- 
pletely unexplained up to the time of which we speak. 
There is seen, beyond the protuberances and the chromo- 
spheric ring, a magnificent aureole, or luminous corona, 
of soft brightness and silvery tint, which may reach as 
far as an entre radius of the dark limb of the moon. 
The study of this beautiful phenomenon, by methods 
which had given such magnificent results, was imme- 
diately undertaken, and occupied the astronomers during 
the eclipses ot 1869, 1870, and 1871. 
But the aureole or corona, though constituting a 
brilliant phenomenon, has in reality but weak luminous 
power. Hence the difficulty of obtaining its spectrum 
with its true characters. “hus the astronomers differed 
at first as to the real nature of the phenomenon. In 1871, 
and bythe use of an extremely luminous instrument, it 
was definitively proved that the spectrum of the corona con- 
tains the bright lines of hydrogen, and the green line 
called 1474 of Kirchhoff’s maps, an observation which 
demonstrates that the corona is a real object constituted 
of luminous gases forming a third envelope round the 
solar globe. 
If indeed the phenomenon of the corona were a simple 
phenomenon of reflection or of diffraction, the coronal 
spectrum would merely be a weakened solar spectrum. 
On the other hand, the characters of the solar spectrum 
are here quite subordinate, and the spectrum is that of 
protuberantial gases and of matter still unknown, indicated 
by the line 1474.1 
The subsequent eclipses of 1875 and 1878, and that 
* One of us has expressed the idea (Notice du Bureau des Longitudes, 
1879) that the coronal atmosphere whicn is in dependence on the chromo- 
sphere and photosphere must present a much more agitated appearance at 
the epoch uf maximum or spots and protuberances, since the jets of hydrogen 
which then penetrate it are much more numerous and rich Ulterior Ubserva- 
tions, and especially those which have been made during the last eclipse at 
the moment when the sular eruptions were abundant, have confirmed this 
previsiun. 
recently observed in Egypt, have yielded confirmation of 
these results. 
But if the constitution of the sun is being thus rapidly 
unveiled, there still remain great problems to be solved, 
both as to this last solar envelope, and as to the region 
near it. 
First of all, have the immense appendices which the 
corona has presented during some eclipses, an objective 
reality, and are they a dependance of this immense 
coronal atmosphere, or might they rather be streams of 
meteorites circulating round the sun (as one of the 
members of the Bureau has suggested) ? 
We do not forget the zodiacal light, the relations of 
which to those dependances of the sun remains to be 
determined. 
But these problems are not the only ones we have now 
to attack, during the occultations of the solar globe. Do 
the regions with which we are occupied contain one or 
several planets, which the illumination of our atmosphere, 
so bright in the neighbourhood of the sun, may have 
always concealed from us? Leverrier long studied this 
question, and his analytical researches led him to suppose 
their existence. 
On the other hand, several observers have alleged that 
they have observed transits of round and dark bodies in 
front of the sun; but these observations are far from 
being certain. The surface of the sun is often the seat 
of small, very round spots, which appear and disappear in 
a time often short enough to simulate the passage of 
round bodies before that star. 
The question is of capital importance; hence it at 
present justly engages the thoughts of all astronomers. 
May the analysis of Leverrier enrich the solar world 
towards its central regions, as it has done with such a 
magnificent s.ccess in the most distant regions ? 
We have but two means of solving the problem, whose 
solution is more particularly incumbent on French astro- 
nomy; the attentive study of the solar surface, or the 
examination of the circumsolar region when an eclipse 
renders their exploration possible to us. This last means 
seems the most efficacious, but on the condition that the 
occultation is long enough to allow of a minute exploration 
of all the regions where the smail star may be met with. 
This gives a capital importance to the eclipse of May 6 
next, one of the longest of the century. 
We will now examine the circumstances of this great 
eclipse, and the means that it would be well to employ 
for observation of it. 
The total eclipse of May 6 next will have a duration of 
6 minutes at the point where the phase is maximum 
(5m. 59s.) ; a time triple that of ordinary eclipses. 
The central line is wholly comprised in the South 
Pacific Ocean, and we can only hope to observe it in the 
islands of that ocean. 
After an attentive study of the question, it has appeared 
to us that two islands would do about equally well for 
observation ; those of Flint and Caroline. 
Flint Island (lat. 11° 30’ S., and long. 151° 48’ W. of 
Greenwich) is the nearest to the central line. Calculation 
gives for the duration of totality in this island 5m. 33s. 
Caroline Island is 150° 6’ W., and 9° 50’ S.; the duration 
of the totality there will{be 5m, 20s.; that is, only 13 
seconds less than in Flint Island. 
It will be seen that the astronomical conditions of the 
phenomenon are extremely favourable in these islands, 
and it is to these stations we should propose to the 
Bureau to send an expedition. 
This expedition should start from Paris, go to New 
York, traverse the American Continent by the railway to 
San Francisco, and there find a steamer (of a French 
service about to be established), which should carry it to 
the Marquesas Islands. There a man-of-war of the 
French station should take it up, and deposit one portion 
at Caroline Island, the other at Flint Island. This ship 
