substance of the great central luminary of our system. The following table, drawn up by M. Ångström, exhibits the number of lines belonging to the several elements enumerated, which have been found to correspond with dark lines of the solar spectrum:

It will be noticed that the solar spectrum shows no traces of the existence of the nobler metals, gold and silver, nor of the heavy metals, platinum, lead, and mercury. On the other hand, it is significant that the lines of nitrogen and oxygen are absent, though these gases can scarcely be supposed to be actually wanting. We must remember, in forming an opinion as to the absence of these elements (as also of such elements as carbon, boron, silicon, and sulphur), that while the presence of certain lines in the solar system may prove abundantly that the terrestrial element which has corresponding bright lines, exists in the Sun's substance, it by no means follows with equal certainty that because all the lines of an element are wanting in the solar spectrum, therefore the Sun does not contain those elements. This will appear when we consider the various circumstances which may cause an element really existing in the Sun's substance to afford no trace of its presence. In the first place, the vapour

of

that element may be of a density causing it to lie always at a very low level, and therefore perhaps altogether beneath that level whence proceeds the white light of the Sun-that is, the light which gives the continuous spectrum across which the dark lines lie. Or again, the element may exist in the Sun's substance at such a temperature, or at such a pressure, as to produce-not well defined absorption lines, but -only broad faint bands, which no optical means we possess can render sensible as such. Or, again, the element may be in a condition enabling it to radiate as much light as it absorbs, or else very little more or very little less-so that it either wholly obliterates all signs of its existence, whether in the form of dark lines or bright lines, or else gives lines so little brighter or darker than the surrounding parts of the spectrum that we can detect no trace of their existence. In these, and in yet other ways, elements may really exist (or rather undoubtedly do exist in the Sun) of whose presence we can obtain no trace whatever.

In other chapters of this work the evidence afforded by the spectroscope respecting the condition of various parts or appendages of the Sun, as the spots, faculæ, pores, prominences, corona, and so on, will be considered at length, as also the various theories to which such evidence has given rise. I conclude this chapter by enunciating the general rules of spectral analysis, and by explaining the rationale of certain recent applications of the spectroscope which have deservedly

attracted great interest, but are not perhaps very generally understood.

The general principles of spectroscopic analysis are as follows:

1. An incandescent solid or liquid gives a continuous spectrum.

2. A glowing vapour gives a spectrum of bright lines, each vapour having its own set of bright lines, so that from the appearance of a bright line spectrum we can infer the nature of the vapour or vapours whose light forms the spectrum.

3. An incandescent solid or liquid shining through absorbent vapours gives a rainbow-tinted spectrum crossed by dark lines, these dark lines having the same position as the bright lines belonging to the spectra of the vapours.

4. Light reflected from any opaque body gives the same spectrum as it would have given before reflection.

5. But if the opaque body be surrounded by vapours, the dark lines corresponding to these vapours appear in the spectrum with a distinctness proportioned to the extent to which the light has penetrated these vapours before being reflected. "

6. If the reflecting body is itself luminous, the spectrum belonging to it is superadded to the spectrum belonging to the reflected light.

7. Glowing vapours surrounding an incandescent body will cause bright lines or dark lines to appear in the spectrum according as they are at a higher or lower

temperature than the body; if they are at the same temperature, they will emit' just so much light as to compensate for that which they absorb, in which case there will remain no trace of their presence.

8. The electric spark presents a bright-line spectrum, compounded of the spectra belonging to those vapours between which and of those through which the discharge takes place. According to the nature of these vapours, and of the discharge itself, the relative intensity of the component parts of the spectrum will vary.

It will be seen, as we proceed, that all these principles bear more or less directly on the application of spectroscopic analysis to the interpretation of solar phenomena.

In all branches of spectroscopic research certain difficulties have to be specially dealt with, while certain circumstances avail to help the observer. We are now to consider,-first, the means which are available to the astronomer for the special purpose of advancing our knowledge of solar physics; and, secondly, the peculiar difficulties which he has to overcome.

In the experiment illustrated in fig. 21 we see the action of a single prism on the solar light. But the time has long since passed when the action of a single prism could teach us anything new respecting the Sun. The observer has to work nowadays with a battery of prisms. A portion of light which has been already dispersed by one prism is received on another, and is thus yet more dispersed; a portion of this doubly dispersed light falls again on another prism and emerges yet more

K

dispersed; a portion of this trebly dispersed light falls on a fourth prism, and so on-the number of prisms actually employed depending on the special branch of inquiry which is being pursued, for some require more dispersive power than is necessary for others.

Space compels me to limit my description specially to the action of the prismatic battery, and I therefore neither describe nor illustrate the means adopted for causing a suitably-shaped beam of solar light to fall on

the first face of the first prism. I therefore simply state the fact that, in the study of the solar spectrum, a beam whose cross section is as s s' (fig. 25), the shape of the slit through which light is admitted, falls, as shown in fig. 25, upon the first prism of the battery, or prism 1 in the figure. If a screen were placed to intercept this beam anywhere between s s' and prism 1, there would be seen upon the screen a bright bar of light shaped like s s'. Prism 1 disperses this beam in the manner already described, and the beam falls thus dispersed on