to sink, with expansion, indicating low pressure in the upper regions, but with a figure indicating a real resistance. It is then this invisible atmosphere, through which the upper portion of the prominence has begun to subside, that would seem to constitute the true solar atmosphere, while the chromosphere should be regarded as consisting of a multitude of smaller prominences, either freshly erupted or the remains of

former eruptions, floating in the lower regions of the solar atmosphere, much as the second prominence of Plate V. is floating (for a brief while) in the upper regions.

The inferences which have been deduced respecting the pressure exerted by the solar atmosphere at and near the level of the photosphere remain unchanged, how

ever, whatever view we adopt on this point. Fig. 72 illustrates the evidence on this point. Here 1 is a part of the spectrum of the Sun's limb, while 2 represents the line F in the spectrum of the chromosphere. The widening of this line close by the Sun's limb may be regarded as unquestionably indicating an increase of pressure, because the researches of Plücker, Hittorf, Huggins, and Frankland have demonstrated that the F-line of hydrogen does actually increase in this way in width when the pressure at which the hydrogen subsists is increased. Temperature, also, has an effect on the hydrogen lines; nor is it quite easy to separate the effects due to pressure from those due to temperature.* But, on the whole, it seems probable that pressure is chiefly in question, while it may be regarded as absolutely certain that temperature alone is insufficient to account for the observed change. Now, whether we regard the glowing hydrogen of the chromosphere as forming a true constituent of the solar atmosphere or as bearing a somewhat similar relation to that atmo

* In a paper by Dr. Zöllner (Über die Temperatur und physische Beschaffenheit der Sonne), an abstract of which, by the present writer, appeared in the English Mechanic for September 1870, it is suggested that means might be devised for distinguishing between the effects of pressure and temperature by causing a discharge of gas to take place at the moment when the induction spark is passed through the hydrogen. It is certainly a problem of the utmost importance thus to distinguish between the two effects which play the chief part in solar phenomena.

The disappearance of one of the magnesium lines at a certain height above the spectrum of the limb (see fig. 71) in which the lines в belong to magnesium, is significant of a very low degree of pressure at the level where the shortest line ceases to be visible.

6

sphere that the aqueous vapour in our own air bears to the permanent constituents, it is yet certain that the pressure of the hydrogen near the solar photosphere is a measure of the atmospheric pressure there. And from the observed width of the F-line near the Sun's limb (see fig. 72) it has been estimated by Wüllner that the pressure on the base of the chromosphere, or at the surface of the photosphere, is below the pressure of the Earth's atmosphere. He has even assigned the limits of pressure at the level of the solar photosphere as lying certainly 'between 50 and 500 millimetres (or between 2 inches and 20) of a mercurial barometer at the Earth's surface.'

It is within the chromosphere and certain of the prominences that spectroscopes of high dispersive power exhibit those signs of cyclonic motions taking

It must not be forgotten, however, that the width of the hydrogenline where it actually reaches the spectrum of the limb is not known. The observed width on which Wüllner founded his researches may be that corresponding to a height of 50, 100, or even 200 miles above the photosphere; and within these 50, or 100, or 200 miles an increase of pressure may take place by which the actual density of the atmosphere close by the photosphere may be enormously increased. There may be an atmosphere including the vapours of iron, sodium, magnesium, &c. (of all the elements, in fine, whose dark lines appear in the solar spectrum) extending, say, 100 miles above the photosphere; and yet no instruments we possess could suffice to reveal any trace of its existence, unless the dark lines in the solar spectrum be thought to demonstrate the fact that such an atmosphere actually does exist. The fact that on some occasions Mr. Lockyer has seen hundreds of the Fraunhofer lines (as bright lines) in the spectrum of the chromosphere renders this far from improbable. The arguments on the strength of which it has been assumed that the absorption to which the dark lines are due takes place below the visible photosphere, appear, to say the least, far from demonstrative.

place in the solar atmosphere the recognition of which has seemed so surprising and inexplicable. The study

of the last part of Chapter III. will have shown the reader that if very rapid motions are taking place, either due to the swift rush of the glowing hydrogen through the solar atmosphere or to the effects of cyclonic motions in the atmosphere itself, by which the glowing hydrogen is borne away, the spectroscope-if

only its dispersive power be sufficient-cannot fail to give unmistakable evidence on the point. The problem is altogether simpler and easier indeed than that which Dr. Huggins had to deal with when he undertook to measure the velocity with which Sirius is winging its flight through space. For there in the solar spectrum are the very lines of hydrogen with which the chromosphere lines are to be compared. If the spectroscopic dispersion suffice, the least experienced observer can

tell as certainly that a solar storm is in progress as the terrestrial observer can tell by the motions of cirrus clouds that the upper regions of our own atmosphere are disturbed. The accompanying illustration (fig. 73), for instance, shows how the F-line in the chromospherespectrum is at times swayed (as it were) from coincidence with the dark F-line of the ordinary solar spectrum. At 1 we see the line deflected towards the violet, showing that the portion of the chromosphere under examination was moving rapidly towards the observer; at 2 we see a deflection both towards the red and towards the violet, indicating that in the same field of view (that is, in the portion of the chromosphere included within the slit) there were masses moving towards as well as from the eye; while, lastly, at 3 we see a deflection towards the red, indicating a rapid motion from the eye. During some observations such as these, Mr. Lockyer has had evidence of motions at the almost inconceivable velocity of 120 miles per second. second. I cannot, however, accept his conclusions as to the distribution of the motions over the Sun's surface, because he seems to take very little account of what certainly is the fact, that the extension of the portion of chromosphere under examination is very great in the direction of the line of sight. Assuming an apparent depth of only ten seconds (which is within the usually observed limits), a tangent-line to the Sun's surface passes through a range of upwards of 60,000 miles; and to speak of the clear recognition of a solar cyclonic storm only 1,500 miles in diameter