NATURE 
THURSDAY, APRIL 20, 1882 
ECLIPSE NOTES 
ca the following notes I propose to discuss certain points 
which in my opinion it is desirable to investigate as 
fully and as carefully as may be during the coming 
eclipse. 
The magnificent volume which astronomers have re- 
ceived from America during the last year, in which are 
garnered all the rich results, or most of them at all events, 
collected during the eclipse of 1878, may really be said to 
have brought to a focus the chief points of study which are 
open to us during eclipses. I shall, therefore, use this 
volume freely in connection with the various branches of 
research. But still there are points of interest which lie 
outside this book, for, since the year 1878, I for one, at 
all events, have been driven to the conclusion that our 
then views of the chemical and physical constitution of 
the solar atmosphere require considerable modification to 
make them accord with the facts. 
I have taken many opportunities of showing that the 
various phenomena observed on the uneclipsed sun are 
more easily explained if we assume our chemical elements 
to be dissociated by the transcendental temperature of 
the sun, than if we hold that their molecular construction 
is the same there as here. 
This question is one, the settlement of which is so 
important if it can be settled, that if an eclipse of the sun 
furnishes us with tests, it is our clear duty not to neglect 
them. I believe that an eclipse does furnish us with two 
or three such tests, and with reference to one of them, as 
I wish in these notes to bring together the various state- 
ments on the subject which have been made, I will begin 
by quoting from a discourse delivered by myself to the 
Astronomical Society last May. (Revised from a report 
in the Odservatory.) 
“The chemical constitution of the heavenly bodies is a 
question which necessitates some amount of attention from 
astronomers. Twenty years ago the observations © of 
Kirchhoff and Stokes enabled us to get glimpses into the 
‘chemical constituents of the Sun. Nine years since, though 
we were in full presence of elements with which we are 
acquainted, other facts had been registered which exercised 
‘the minds of some observers. Kirchhoff’s view was that 
‘the substances with which we are acquainted were demon- 
fexated in the atmosphere of the Sun by an exact matching, 
both as regards wave-lengths and intensity, with the lines 
of certain chemical elements which he employed. Fraun- 
|hofer had earlier noted the coincidence of the bright yellow 
Eee of sodium with the line D. But Kirchhoff showed 
that not only in the case of sodium, but in iron, magnesium, 
‘and cobalt, and several other substances, there were coin- 
cidences which went to show that what was good for 
sodium was good for other bodies. But nine years ago 
pee had not merely the opportunity of comparing these 
bright lines with the spectrum of the Sun’s atmosphere as 
\revealed by Fraunhofer, but we had the opportunity of 
studying the spectra obtained by observing very small 
/portions of the solar atmosphere in regions where we should 
‘expect an exceedingly high temperature, namely, the inner 
‘regions of the solar atmosphere—the regions of spots and 
the regions of prominences. When we began to tabulate 
the lines thickened, the thing began to be very much less 
clear ; of the 460 iron lines recorded by Kirchhoff only three 
were observed in the prominences. Then, when we got 
jindication of a change of refrangibility of the lines due 
VoL. xxv.—No. 651 
58 
to the motion of the solar gases, we found about the 
year 1869 that the thickened lines which indicated iron 
vapour in the spots were not brightened in the promin- 
ences, so that.a great many questions were raised ; and 
when those questions were raised the idea of de- 
composition at a high temperature seemed to arise also. 
I bring before you to-night the results of some purely 
astronomical inquiries lately undertaken by the Solar 
Physics Committee. Of course a great many physical 
inquiries have necessarily entered into the researches. 
But the astronomical inquiries have had this object in 
view, namely—given the fact that a high temperature can 
decompose an elementary body, what happens to the 
spectra of those bodies when we examine the Fraunhofer 
spectrum, the spectrum of spots, and the spectrum of 
prominences? We have had before us the admirable 
work of Professor Young in 1872, but the work’ only 
lasted a month We felt we wanted more facts ; so what 
we have been doing at Kensington during the last two- 
and-a-half years has been to obtain the spectra of 100 
sun-spots—not a perfect record of all lines thickened, but 
results we could compare with Tacchini’s ; because, for 
prominences, we had to depend on Tacchini’s observa- 
tions, observations confined to the brightest lines of the 
prominences. The Committee therefore attempted some- 
thing which was quite modest, and contented themselves 
with observing only the twelve lines most affected in Sun- 
spots. The question was, where to take the lines; and it 
was obviously the wisest course to take them in the most 
visible part of the spectrum; so that for two-and-a-hal. 
years we have been taking the twelve most widened lines 
between F and D. I willonly trouble the Society with one 
set of these observations. At the top of this diagram we 
have carefully chosen among the Fraunhofer lines, the lines 
stated by Angstr6m to be coincident with the bright lines 
of iron ; and we have given these lines of different lengths, 
the length representing the darkness of the Fraunhofer 
lines. In the next horizon we have the actual observations 
of the iron lines given by Angstrém, who used an electric 
arc with thirty or more of Bunsen’s cells. We compared 
the intensities, also represented by length, as given by 
Angstrém and as given inthe sun, You will see a con- 
siderable disparity. Below, we have the lines of Thalén, 
who used a powerful induction-coil, and the lengths of these 
also represent intensities. Comparing the Fraunhofer lines 
and Thalén’s lines, you will see a still further disparity be- 
tween the two spectra. Below, in these 100 horizontal 
strata are all the observations of the spots taken during the 
last two years. Thefirst point which strikes one is the enor- 
mous number of iron lines, both in the solar spectrum and 
in the iron spectrum, which are not affected in spots or 
storms. It is as if on a piano only a few notes are played 
over and over again, always producing a different tune. 
The next point is the inversion of the phenomenon. If you 
examine the lines, you see that every line has been seen 
without the others. That hard fact is one which really is 
very difficult to understand, and what strikes one is the 
marvellous individuality, so to speak, of each of these lines. 
They do not go in battalions, or companies, or corporal’s 
guards, but in single unities. The great importance of get- 
ting these observations wasnotso much for the observations 
themselves as for the comparison it enabled us to make 
with other observations ; and naturally the next thing to do 
was to get a long series of observations of the prominen- 
ces, because the prominences are hotter than the spots. 
The spots are caused by down-currents when the Solar 
atmosphere brings vapour from the coolerregions. They 
are opposed to prominences, which are ejections from the 
heated interior of the Sun. We have arranged here the 
observations of the prominences by Tacchini, since 1872. 
Here we are dealing with one substance—iron—over a 
very limited portion of the Solar spectrum ; and what is 
the result? First of all you will see a very much greater 
1 See vol. xxiv. p. 322, Fig. 35. 
cic 
