LL2 
NATURE 
[Dec. 8, 1870 

made show that the time of exposure, with ordinary portrait 
collodion, must be nearly four minutes, in order to produce 
“images of a size which would correspond to a picture of the solar 
disc about 2 in. in diameter. This length of exposure demands 
a more perfect clockwork than my instrument possesses, and a 
“more accurate adjustment of the polar axis than it had during 
the experiments, as well as a steadier condition of the atmo- 
sphere, 
Thus far, therefore, I have not been able to produce anything 
which could properly be called a good picture. Negatives have 
been made which show clearly the presence and general form of 
protuberances, but the definition of details is unsatisfactory, This 
amount of success was reached upon September 28, when im- 
pressions were obtained of two protuberances on the S.E. limb 
of the sun, and, slight as this success was in itself, I consider 
it of importance in showing the perfect feasibility of going much 
further with more sensitive chemicals, more delicate adjustments, 
and greater telescopic power. I was aided in the experiments 
by Mr. H. O. Bly, our local photographer, to whom are due my 
warmest acknowledgments for the interest, patience, ingenuity, 
and skill with which he assisted me. 
We worked through the Hydrogen ¥ line (2796 of Kirchhoff’s 
scale) which, though very faint to the eye, was found to be 
decidedly superior to F in actinic power. The photographic 
apparatus employed consisted merely of a wooden tube, about 
6.in. long, attached at one end to the eye-piece of the spectro- 
scope, and at the other carrying a light frame. In this frame 
was placed a small plate-holder, containing for a sensitive-plate 
an ordinary microscope slide, 3 in. by 1. The image of the 
prominence seen through the open slit, is magnified and thrown 
upon this plate by the eye-piece. Fig. 1 shows the instrument 
with this apparatus attached. 
It would be easy to improve this arrangement in many respects, 
and whenever I resume the subject I propose to do so, 
As the equatorial, howeyer, has been dismounted, to be put in 
order for the observation of the December eclipse, further 
attempts in this direction must be postponed until next spring. 
Observations of the Solar Protuberances 
Without prolonging this article with the detail of observations, 
I add a few of the results which have been obtained since 
Sept. 10. 
About forty different prominences have been more or less 
carefully observed ; sixteen have been sketched. Most of them 
fall, naturally enough, into the categories established by Zollner 
and Lockyer, and are fairly represented by figures already 
published in the Fournal of the Franklin Institute. 
A few deserve especial mention, however. Fig. 3 represents 
a small one which was observed upon the I’. limb of the sun, on 
September 14, about 4.30 P.M. From the point marked a, 
which was very brilliant, a small fragment detached itself and 
rose towards A’, enlarging in size and growing fainter as it rose. 
It disappeared (from faintness) in about 12} minutes, at a dis- 
tance of 2’ 30" above the limb of the sun, as determined by the 
time, 8"°5, which was occupied by the intervening space in passing 
over the slit of the spectroscope. Allowing for the obliquity of 
the motion to the parallel of declination, the length of path 
passed over by this cloud was more than 90,000 miles, and the 
velocity above 120 miles per second. 
Fig. 4 represents a prominence observed September 20, at 
4 P.M., on the S.E., limb. (Pos. S., 60° E.) It was a nearly 
vertical stream, made up of spindle-formed filaments, and had 
attained the enormous height of 3 20” or 90,000 miles (deter- 
mined, as in the case above mentioned, by a time-observation, 
corrected for inclination). It was very brilliant near the base, 
and at two or three other points along its length. At 4.30 it 
was nearly gone, only a few faint wisps of cloud remaining. 
Another, observed on September 27, at 4.10 P.M., and situated 
on the W. limb of the sun, is represented in Fig. 5. It was 
formed of separate, well-defined narrow streamers, which ap- 
peared to consist of matter, first violently ejected, and then as 
violently deflected, by some force acting nearly at right angles. 
The altitude of the highest point was 1’ 25”, the length of the 
whole about 3’ 30’. I am unable to see how any mere projection 
from the sun could have produced such a form, and cannot help 
feeling that it indicates a something in which powerful currents 
may exist, even at great elevations above the solar surface ; in 
short, an atmosphere extending far beyond the limits which cal- 
culation would seem to assign as possible. Is it wholly unlikely 
that at such an enormous temperature the law of Mariotte may 


“4 
fail so completely as to destroy the reliability of any computation 
that assumes it as one of the data ? 
Upon the next day the prominence still persisted, but its type 
was wholly changed : it was replaced by one of the mushroom- 
formed masses which are so common, 
Bright Lines 
In the spectra of different protuberances, the following bright 
lines have been observed, the numbers referring to Kirchhoff’s 
scale: C; D, ;-D33Dg5 14745 1515 5 0, 3 bg: 535 6,5 1990); 
2001 ; 2031; F; 2581°5 ; 2796; 4—17 in all. On one occa- 
sion, September 27, the base of a prominence on the N. W. limb, 
close to a spot just leaving the limb, exhibited as many as twelve 
or fifteen short bright lines between E and F, which are not 
included in the above enumeration, as I had not time to identify 
them. It is the only instarice in which I have seen this pheno- 
menon, more than once described by Mr. Lockyer. ; 
I desire to call special attention to 2581°5, the only one of my 
list, by the way, which is not given on Mr. Lockyer’s. This 
line, which was conspicuous at the eclipse of 1869, seems to be 
always present in the spectrum of the chromosphere, and shows 
the form of its upper surface Gr of 4 protuberance nearly as well, 
though of course not so brightly, as the 2796 line. It has no 
corresponding dark line in the ordinary solar spectram, and not 
improbably may be due to the same substance that produces D,: 
The reversal of the sodittm ard magnesium lines is not at all 
uncommon. In some instances these lines were so bright that 
on opening the slit the form of the prominence could be made 
out through them, This was the case with a small hand-shaped 
prominence observed on September 27. Comparing the form 
thus seen through D’ and D, with that given by Dg, it appeared 
that the sodium line was sufficiently developed for observation 
only along the edge and at one or two bright points in the promi- 
nence, most brilliantly neither at its summit nor its base. Fig. 6 
represents the appearance (the slit was perpendicular to the sun’s 
limb). The case was similar with the magnesium lines. 
Spectrum of Solar Spots 
Several spots have been carefully examined at different times ; 
most of them, in their spectra, gave evidence of unusual dis- 
turbances; but by far the most interesting phenomena were 
exhibited by a large group which was first observed near the E. 
limb on September 19. Changes of wave-length were frequent 
in its neighbourhood. 
Figs. 7 and 8 represent the appearances assumed by the F and 
C lines respectively, at the times indicated below each figure, 
during an observation on the afternoon of September 22. The 
point where these changes of wave-length occurred was at the 
western edge of the penumbra, At other times similar changes 
were observed, but not so great or rapidly varying. 
The calcium and titanium lines referred to in my note pub- 
lished in the July number of the Fournal of the Franklin Institute, 
were always conspicously thickened in the nucleus spectrum. 
The C and F lines were reversed in some portion or other of 
the group nearly every time I observed it. On September 22 
the sodium lines were both reversed for several hours, while Dy 
appeared as a dark shade, On September 28, again, at 4 P.M., 
the southern nucleus of the group (which at this time contained 
four large umbre, besides many small ones) reyersed all of the 
following lines, viz.: C; D,; Da; Dg; 14743; 6°5 6); 033 O53 
F ; 2796, and %. All of these were conspicuous, except 1474 } 
TD, and 2, especially so, and the latter (a nickel line) showed 
considerable changes of wave-length, alternate increase and 
diminution, which were not shared by its magnesian neighbours, 
6', b, and b,. : : 
At 4'05 P.M. the brilliance of the F line increased so greatly 
that it occurred to me to widen the slit, and to my great delight 
I saw upon the disc of the sun itself a brilliant cloud in all its 
structure and detail identical with the protuberances around the 
limb. Indeed, there were /wo of them, and there was no diffi- 
culty in tracing ott and delineating their form. Fig. 9 represents 
them as they were from 4°05 to 4:10; Fig. 10 gives the form at 
4°15 to 4°20. They were then considerably fainter than at first. 
During the intervening ten minutes I examined the other lines of 
the spectrum, and found that the form could be distinctly made 
out in all the Aydrogen lines even in #; but that the reversal of 
the other lines, including D3, was confined to the region imme- 
diately over the spot-nucleus, where the smaller but brighter 
cloud terminated abruptly ; or, I might better say, originated. 
The larger one faded out at both ends. When the clock-w ~ 
