The next protuberance lies imbedded in the moon's border, and has in form the appearance of a short, deeply-articulated worm; its mean position is 69° 17'; its length 46,700 miles, and its greatest height 9900; between that protuberance and the point c on the woodcut, are two flames in the midst of the glow previously described. Midway the diffused light rises to an elevation of 60,500 miles. We now come to a curiously-formed protuberance. Some have compared it to an ear of corn, but in the photographs it appears like an eagle with outspread wings resting on the trunk of a tree which leans towards the north. On one plate where the tree-stump is cut off by the advancing moon, the resemblance to an eagle on the wing is perfect. The form of this object indicates instability,

and impresses one with the idea that it is a great travelling whirl of flame, the direction of whose rotation-as indicated by the position of the wings and the projection of one on the other-is retrograde, or in the same direction as the motion of the hands of a watch. Dr. Mayer, chief of the Burlington Section of the Philadelphia Photographic Expedition, has examined, with care, the successive photographs of it, and he says that although at first he thought that the last impression differed from those preceding in that the wings had become longer and more in a line with each other, yet on subsequent examination he could not really decide that a perceptible motion had taken place during the time of totality. The height of this object is 36,700 miles, and the spread of the

VOL. VII.

D

wings 70,800 miles. The next protuberance extends to between E and I on the woodcut; it is of very irregular outline, and shows portions of its substance detached from the general mass and floating freely above it. The most elevated and bright of these detached flames floats at a height of at least 20,000 miles above the surface of the sun. Beyond I a white nebulous cloud rises to the elevation of 60,500 miles. Next follow two protuberances at K.

We now pass to the western limb of the sun, and meet with the remarkably large and massive protuberance at G on the woodcut. It is shaped like a bird's head, with the beak and under-side of the head resting on the limb of the moon. On a photograph taken at Ottumwa, Iowa, just before the sun came out, this protuberance had the exact appearance of an albatross head with the beak open, holding a rounded mass between the extremity of the jaws. The protuberance at F bears the most striking resemblance to a caterpillar. It extends through an angle of 11°, or 81,800 miles; its maximum elevation, which is at the head of the caterpillar, is 23,000 miles. Out of the head issued two horns; the one nearest the front being the higher of the two, and terminated with a knob or ball from which curves a broken line of light to the border of the moon. The next prominence at H has the shape of a grain of rice slightly constricted in the middle. Between H and A is another protuberance.

Professor Young, who examined these prominences during the totality, has continued his spectroscopic notes of the prominences since the eclipse, and on September the 13th he obtained a view from which the accompanying woodcut is taken. He

describes it as a long, straggling range of protuberances-the sketch giving a very fair idea of the number, form, and arrangement of the immense cloudy mass. The points a and b were very bright.

On September 18th he noticed a remarkable phenomenon, which, although not bearing directly on the eclipse phenomena, is sufficiently rare to make it deserve recording in these pages. Whilst examining the spectrum of a large group of spots near the sun's western limb, his attention was drawn to a peculiar knobbiness of the F line (on the sun's disc, not at the edge), represented by the following cut a, at the point e. In a very few moments a brilliant spot replaced the knobs; not merely interrupting and

FIG. 3.

с с

reversing the dark line, but blazing like a star near the horizon, only with blue instead of red light. It remained for about two minutes, disappearing, unfortunately, whilst the observer was examining the sun's image upon the graduated screen of the slit, in order b to fix its position. It is not known, therefore, whether it disappeared instantaneously or gradually. b gives an idea of this appearance. On returning to the eye-piece, Professor Young saw what is represented at c. On the upper e (more refrangible) edge of r there seemed to hang a little black moat, making a barb, whose point reached nearly to the faint iron line just above F. As given on Angström's atlas, the wave-length of F is 486-07, while that of

the iron line referred to is 485-92 (the units being millionths of a millimètre). This shows an absolute change of 0.15 in the wavelength, or a fraction of its whole amount, represented by the decimal 00030, and would indicate an advancing velocity of about 55.5 miles per second in the mass of hydrogen whose absorption produced this barbed displacement. The barb continued visible for about five minutes, gradually resolving itself into three small lumps, one on the upper and two on the lower line, Fig. 1, d. In about ten minutes more the F line resumed its usual

appearance.

Whilst on the subject of the solar prominences it may not be out of place to refer to some observations by Professor F. Zöllner, who has succeeded in observing them without an eclipse with great sharpness and clearness. From the nature of the method the same protuberance was simultaneously observed in three different colours corresponding to the three homogeneous lines of its spectrum. There is, however, a material difference between the red and blue image on the one hand, and the yellow on the other. The latter is very intense only in close proximity to the edge of the sun's disc, and in this respect corresponds to the other images; while the more delicate details disappear at a greater distance. This difference does not seem to be caused by the greater brightness of the spectrum in that region, but appears to depend on one of the two following hypotheses for an explanation:-either that the rays which give rise to the yellow image emanate from a gas having a greater specific gravity than hydrogen, and therefore existing at a lower level, or that the greater intensities of temperature and pressure nearer the surface of the sun cause hydrogen to emit these rays.

Professor Zöllner's paper, which will appear in the next number of the journal of the Franklin Institute (for early proofs of which the writer has to thank Professor Morton, the editor), is illustrated

with beautifully coloured drawings showing the rapid changes which sometimes occur in the forms of these prominences even in the course of a few hours. Observing one of the most remarkable formations, the Professor says, "I hardly believed my eyes when I noticed in it the tongue-like motion of a flame. This motion was slower, however, compared with the size of the flame than that of high towering flames at great conflagrations. The time required by such a wave in passing from the base to the apex was about two or three seconds."

In comparing the general impression of the protuberances with terrestrial phenomena, the author states that the great majority remind him of the different forms of our clouds and fog. The cumulus type is completely developed in the cases here referred to. Other formations remind us of masses of clouds and fogs floating closely over lowlands and seas, whose upper parts are driven and torn by currents of air, and which present the well-known evervarying forms when viewed from the tops of high mountains.

Professor Zöllner hopes, by using larger prisms and a circular slit in the spectroscope, to be enabled to observe simultaneously all the protuberances on the edge of the sun, in the different parts of the spectrum, just as in a total solar eclipse of long duration.

Returning to the August eclipse, one of the most beautiful observations was on the first contact by means of the spectroscope. Professor Young has been giving much attention to this subject, and had fitted up a very efficient instrument for the purpose. The instrument consisted of a spectroscope with five prisms of 45° each, having faces 2 by 3 inches; the collimator and telescope had apertures of 2 inches, with a focal length of 17. These were connected with a comet-seeker of 4 inches aperture and 30 inches focus, used with an eye-piece, and giving an image of the sun 2 inches in diameter on the slit of the spectroscope. A graduated screen at the slit determined positions of points on the sun's limb, and a wire micrometer measured the positions of spectrum lines. The whole was mounted equatorially with slow-motion screws. During the eclipse he was stationed at Burlington, Iowa, and shortly before the first contact was due, he found that there was a solar prominence located at the spot where first contact must occur (see F in cut on page 33). He therefore fixed his spectroscope with the slit radial to the solar edge at the point, so getting a prominent spectrum whose width was determined by the height of the prominence. Closely watching this, he presently found that it began to narrow steadily, and at the instant that it became a mere line and disappeared he recorded first contact. The moon's approach was perceived full 30" before its actual appulse; the observation was perfectly easy, and the time determined is certainly to be relied on within half a second, and probably much less. The presence of a prominence at the point of contact is not

essential to the success of the method, as there is everywhere on the sun's limb sufficient depth of chromosphere to answer the purpose. From the first photograph showing contact made by the Philadelphia party at the same place, Professor A. M. Mayer, who had charge of that division, calculated the time of actual first contact, and found that it came within two-tenths of a second of the record made by Professor Young.

Professor Young proposes to apply the spectroscope in this manner to observations both of the external and internal contacts at the next transit of Venus.

The partial-phase pictures show the various sun-spots visible at the time (about six in number) with admirable definition, the larger ones being surrounded by a marked fringe of faculæ. They all show a beautiful gradation of shade from the border of the sun inwards. This shading of the source of light is due to the absorption of the peripheral rays which necessarily pass through a greater thickness of the dense solar atmosphere than those which emanate from the central portion of the disc; on a more searching examination of the relative intensities of light of different portions of the solar disc, there may be observed on all of these photographs, close to the limb of the advancing or retreating moon, a bright glow like that of early dawn, which extends from the moon to a distance of about 15". Unless this glow can be accounted for in node and in measure by diffraction, it would appear as if it were due to a lunar atmosphere, although Dr. Mayer, in suggesting this explanation, confesses that he cannot understand how an atmosphere capable of producing such marked effects when projected against the intensely lighted disc of the sun, should have no appreciable refractive effect on small stars when occulted by the moon. We should be more inclined to account for this glow as being the effect of specular reflexion from the surface of the moon grazed by the sun's rays.

To

A party under Professor Pierce devoted themselves exclusively to the recording of that strange phenomenon, the corona. secure any impression from this object, which, notwithstanding its apparent brightness, is remarkably deficient in photographic power, it was necessary to make a very small image and to give a very long exposure.

The telescope was therefore arranged to produce an image in its principal focus simply, and during the totality an exposure of forty seconds was given. By this means a picture was obtained of which the cut on the next page is a very careful copy. From the long exposure, the motion of the moon, and probably also of the light in the corona, there is little sharpness of definition, and the prominences only appear as bright spots. The general shape of the corona is, however, very well given, and the curious appearance of curvature, in some parts, is very manifest. Professor Himes, who