shows more of the prominences on the superior part of the Sun's limb, while Secchi's series shows more of the prominences on the inferior portion.

It will be unnecessary-so closely do the two series resemble each other in all essential respects-to exhibit both; but further on there will be found a copy of one of Secchi's pictures (fig. 82, p. 327), which the reader may compare with figs. 66 and 67, copied from Mr. De La Rue's photographs.

It will be noticed that fig. 66 represents the earliest phase. The Moon is advancing from right to left, and has just hidden the last fine thread of direct sunlight on the left. Thus we see the full height of the prominences on the left, while no prominences are seen at all on the right of the Sun. At the upper and lower part of the Sun's limb the prominences are partly concealed, and necessarily remain so throughout the eclipse. In fig. 67 the Moon has obliterated a large proportion of the prominences on the left, while it has in turn revealed a number of small prominences, a long range, or sierra, and a lofty and massive projection, on the upper right-hand quadrant. Fr. Secchi observed the prominences directly, and with great care, while his assistants managed the photography. Amongst the phenomena he noticed, I may mention the circumstance that the strange prominence seen in both figures on the upper left-hand quadrant, possessed a helicoidal *

*It would be well if this word 'helicoidal' were always employed in this sense, in preference to the word 'spiral,' which might be conveniently restricted to curves lying in a plane. I am not insisting on either term

structure.

In the magnified picture from Mr. De La Rue's photographs, this structure can be clearly recognised.*

Fr. Secchi thus summed up the result of his observations:

1. The prominences are not mere optical illusions; they are real phenomena appertaining to the Sun. Our observations having been made at two places separated a hundred leagues from each other, it is impossible to suppose that shapes so well defined and so exactly identical could be produced by a phenomenon resembling mirage.

2. The prominences are collections of luminous matter of great brilliancy and possessing a remarkable photographic activity. This activity is so great that many of the prominences which are visible in the negatives could not be seen directly, even with powerful instruments, perhaps because they emitted only chemical rays and few or no luminous rays.†

3. There are masses of prominence matter suspended

as more correct, but only on the convenience of a recognised term to express spiral curves not lying in one plane.

Secchi notices the apparent encroachment of this prominence upon the Moon's limb-a peculiarity which he ascribes to the fact that the Moon was moving away from the prominence while the plate was under exposure. Doubtless, this circumstance produced its effect; but the phenomenon is chiefly due, according to the experimental researches of Dr. Curtis, to a process of chemical encroachment taking place during the development of the plate.

The spectroscopic observations made during the American eclipse tend to throw doubt on this conclusion, which, however, is in accordance with the observations made by Mr. De La Rue, and also, be it noticed, with laboratory experiments on the spectrum of hydrogen.

and isolated like clouds in the air. If their form is variable the variations take place so slowly that it is impossible to recognise their effect during an interval of ten minutes. *

4. Besides the prominences, a zone exists of the same material, enveloping the whole of the Sun's globe. The prominences spring from this envelope; they are masses which raise themselves above the general level, and even at times detach themselves from it. Some among them resemble smoke from chimneys or from the craters of volcanoes, which, when arrived at a certain elevation, yields to a current of air, and extends horizontally.

5. The number of prominences is incalculable. When observing the Sun directly, its globe appeared to be encircled with flames; there were so many that it seemed hopeless to attempt to count them.

6. The height of the prominences is very great, especially if we notice that account must be taken of the portion concealed by the Moon. Thus estimated, the largest protuberance visible in 1860 was certainly not less than three minutes in height, which corresponds to about ten times the diameter of the Earth; the others had a height of from one to two minutes.

* Later observations show that this opinion must be modified, and that though many of the large prominences remain unchanged in figure for a considerable interval, yet others change very rapidly.

+ Grant, Swan, and Von Littrow had already recognised this; and Leverrier, from observations made during the same eclipse, had come to the same conclusion. The matter is alluded to further on. (See note in pp. 291, 292.)

We may consider that the prominences were finally placed in their true place in the solar scheme by the observations made in 1860. Doubts still continued to be expressed by a few ; but the majority were satisfied; and henceforth the coloured prominences were very generally regarded as solar appendages.

It may be well to consider briefly the interest of this result before proceeding to inquire into the researches of the last few years, which even surpass in importance those already described.

In the first place, it must be remembered that though the prominences are seen all round the circumference of the solar disc they do not really form a circle. They are the foreshortened projections of objects which may lie-and many of which must liethousands of miles from that circle on the Sun which at the moment forms the apparent boundary of his disc. We know, in fact, that certain prominences are as high as three minutes-that is, extend to some 80,000 miles from the Sun's surface. Now, supposing A B C to represent a part of the Sun's circumference, and a b c three prominences of this height, an observer, viewing the Sun from a point at a great distance away towards the right or left, would only see the extreme tips of the prominences a and c, while he would see the full height of the prominence b. But in order that these two prominences should be thus in appearance sunk below the solar limb, the line a c would need to be about 500,000 miles in length. So that if there were any prominence of so great a height as a, b, or c

along any part of the arc A B C it would appear to rise above the Sun's level. Nor need we modify this conclusion on account of the layer of red matter which has been shown by the observations already described to surround the solar globe, since undoubtedly the highest prominences extend considerably more than three minutes (even) above the uppersurface of this red layer.

But now, having considered a side view, let us conceive the case of a view from above. Suppose a', b', and e' (fig. 68) to represent the same three

Illustrating the distribution of large prominences over the Sun's surface. prominences brought round so as to be viewed from above. Then the actual boundary of the Sun thus viewed would be represented by the circle s B s', such that whereas a' c' represents 500,000 miles, s s' repre sents 850,000 miles. In order, therefore, that a prominence of the full height of 80,000 miles seen from a