three degrees) or else it may give the spectrum corresponding only to a narrow line of light. In one case the spectrum is produced by the spreading out as it were of multiplied images of the illuminated slit, the illumination of the slit being derived from a wide area of light. In the other the spectrum is formed of multiplied images of the line of light under examination. A spectroscope serving the former purpose may be called (as Professor Yourg, of America, suggests) an

FIG. 34.

S

R

integrating spectroscope; one of the latter kind an analysing spectroscope.

What follows relates to the work of the analysing spectroscope.

It must be remembered that what the analysing spectroscope really does is to give a range of pictures of whatever luminous object or part of an object would be visible through the slit if the spectroscope were removed.

Now, supposing such a portion of the solar photosphere is observed as is shown in the space s s', fig. 34 (s being the solar disc), the images of this portion will

give us the spectrum R V, showing the dark lines due to the absence of certain images, as explained above; and no other portion of the disc produces any effect whatever. It is very essential to remember this. We are in fact analysing under such circumstances the part ss' of the disc, and no other part.

If a spot or a facula be crossed by s s', then the spectrum we get is no longer that of a uniformly, or almost uniformly, bright part of the solar disc. If s s' (fig. 35) represent an enlarged view of the spot and the

FIG. 35.

S

space included by the slit, then this last, seen separately, will be as s s'; and the spectrum will consist of a number of images of s s' ranged side by side, so as to form a strip, as R V in fig. 34. Hence at the top and bottom of this compound spectrum there will be two narrow solar spectra corresponding to the parts s P and s' P'; next to these will be two narrow spectra of the penumbral parts PU and P' U'; and about the middle there will be a narrow spectrum corresponding to the umbral part u u', all these spectra forming one compound spectrum, whose red end is towards the left (assuming the dispersion to be as in the case illustrated

in fig. 34) and its violet end towards the right. The nature of the penumbral and umbral spectra will be stated further on. It is by comparing these spectra with the adjacent solar spectra that the spectroscopist is enabled to form an opinion as to the nature of the spots, and to make inferences as to the general physical constitution of the Sun.

Similar remarks apply to the case where a portion of facula, or pores, or mottlings, or of any other features of the solar disc, falls within the space s s'. All such peculiarities tend to produce peculiarities in the resulting compound spectrum; since the image of the portions s' is repeated along the whole length of the spectrum R V after the fashion already described.

But, having considered these comparatively simple cases, let us deal with the subject which has of late attracted so much attention—the visibility of the spectrum of the prominences when the Sun is not eclipsed. Further on the exact nature of the prominence-spectrum will be considered; but in this place I note only respecting it that it consists of bright lines.

Now, suppose that P P' is a prominence, s s' the edge of the Sun, and ss' the space included by the slit. Then p's', as in the former cases, produces a solar spectrum (which, however, commonly presents certain peculiarities when belonging to the edge of the Sun's disc); the part pp' includes a portion of the prominence, and gives a prominence-spectrum which we may suppose to be represented by the bright lines at C and F and near D. But it will also give a solar

spectrum, for the light of our own illuminated air comes from the space included within the slit s s'; and as our air is illuminated by solar light, it produces (according to rule 4, page 128) a solar spectrum. Also the part s p' will give a solar spectrum due to the illuminated air. Now, the prominence P P' is absolutely obliterated from view by the illuminated air, which extends all round (and over, be it remembered) the place of the Sun. Since then, if we looked at the space s s' alone, with whatever telescopic power, and with whatever contrivances for reducing the glare of light, the portion p p' of the prominence P P' would be

wholly invisible to us, why, it may be asked, should the lines C, F, and the one near D-which are in truth but coloured images of the part p p'-be visible, although the spectrum of the illuminated air falling within s p' is spread over these lines precisely as the illuminated air is itself spread over the prominence? The answer is easy. The whole of the light of the illuminated air within the small space s p' is spread over the large space shaded with cross lines in fig. 36, and is reduced in intrinsic brightness in corresponding proportion. On the other hand, the light of the prominence-matter within pp' is spread only over the three lines shown in the figure (and a few fainter ones),

and is therefore proportionately but very little reduced. Hence, if we only have enough dispersive power, we can make sure of rendering the prominence-lines visible, for we get the same luminosity for them whatever the length of the spectrum, the only effect of an increase of length being to throw the bright lines farther apart; whereas the atmospheric spectrum which forms the background will obviously be so much the fainter as we spread its light over a longer range.

By this plan we get a certain number of images of a

FIG. 37.

portion of a prominence—a mere strip, so to speak; and we can get any number of such portions, and in any direction as compared with the Sun's limb. For example, if s s' (fig. 37) be the Sun's limb, P P' a prominence, we can get from such a strip as s as ss the spectrum R V. And obviously since the length of the bright lines tells us the length of the part p p' in figs. 36 and 37, we can, by combining a number of such parallel strips as s s', learn what is the true shape of the prominence P P'.

L