The sun's own light shows under this method of spectroscopic analysis millions of tints, in fact I might say millions of red tints, and so forth, right through the spectral list of colours. But also many thousands of tints are wanting. Imagine a rainbow-coloured ribbon, the colours ranged along its length, so that the ribbon is black at both ends, and that from the black of one end the colour merges into very deep red, and thence by insensible gradations through orange, yellow, green, blue, indigo, and violet, into the black of the other end. Then suppose that tens of thousands of the fine threads which run athwart the ribbon--_i.e._, the short cross threads--are drawn out. Then the ribbon, laid on a dark background showing through the spaces where the threads were drawn out, would represent the solar spectrum. We know then that the light of the sun's glowing mass either wants particular tints originally, or shines through vapours which prevent the free passage of rays of those colours. Both causes might be at work, not one only. At present we are not concerned with this particular point; but I only mention that, in reality, no tints are actually wanting, though some are very much enfeebled.
The sun's light falling on any opaque object is reflected. If the object is white, the light gives exactly the same spectrum, only fainter. Thus, I take a piece of white paper on which the sun's rays are falling, and examine its light with one of Browning's spectroscopes. I get the ordinary solar spectrum. The cold white paper gives me in fact a spectrum which speaks of a heat so intense that the most stubborn metals are not merely melted but vaporized in it. But this heat resides in the sun, not in the paper.
Now, speaking generally, Mars also sends us sunlight, so that when we spread out with the spectroscope the rays coming from this planet, we get the solar spectrum, only of course very much enfeebled. _But_ close examination shows that other tints besides those missing from the solar spectrum are missing from the spectrum of Mars. He reflects to us the sunlight, almost as it reaches him, but he abstracts from it a few tints on his own account.
When we inquire what these tints are, we find that they are tints which are _sometimes_ wanting even from direct sunlight. When the sun sinks very low and looks like a great red ball through the moisture-laden air, his spectrum is not the same exactly as that of the sun shining high in the mid heaven. It shows other gaps than those corresponding to the ordinary myriads of missing tints. Its red colour shows indeed that some thing has happened to the sunlight; but, oddly enough (at first sight at least), the gaps are chiefly in the red part of the spectrum, just what one would expect if the sun's light showed a want instead of an excess of ruddy light. The fact is, however, that the violet, indigo, and blue are weakened altogether, not by the mere abstraction of tints here and there. The red suffers under a few abstractions of tint, but remains on the whole little weakened. Now the same gaps which at such times appear in the spectrum of the sun are found (generally, if not always) in the spectrum of the planet Mars, even when he is shining high in the heavens, so that his light is _not_ at the time absorbed by the denser portions of our air. In fact the gaps have been seen in the spectrum of Mars when the planet has been shining higher in the heavens than the moon, whose spectrum was found on trial (at the time) not to show the same gaps,--as of course it must have done, and even more markedly, if the missing tints had been abstracted by our own air.