were on the other side of the curve. The light of the star is gradually blotted out by an enormous quantity of carbon compounds in some form or other, till at last the star gets blood-red (group VI), and finally is lost to human ken. The solar atmosphere at present contains chiefly iron, calcium, and other similar metals, but the hydrogen is disappearing, and there is possibly the slightest trace of carbon, but that trace is so small as to be somewhat doubtful. The composition of the sun's atmosphere at present is, moreover, almost identical with that of a mixture of meteorites driven into vapor by a strong electric current, and, if we except hydrogen, there is scarcely a line of any importance in the spectrum of the one which is not represented in the spectrum of the other. Calcium, aluminium, iron, manganese, and certain lines of nickel and other substances, are present. By means of such experiments as this, the wonderfully close connection between the gases at present existing in the atmosphere of the sun and the gases obtained from the volatilization of meteorites is put before us in the clearest and most convincing manner.
With regard to the fact that carbon comes in and takes the place of highest importance in the atmospheres of these cooling bodies, it is worth while to remark that if, as seems possible, these permanent gaseous compounds of carbon with different substances like oxygen, nitrogen, and hydrogen, and probably hydrogen itself, are kept away from the swarm during its condensation by that form of radiant energy of the center which is evidenced in the case of the sun by its tail-producing action on comets, it is easy to imagine that when that radiant energy is reduced, the carbon compounds will gradually approach the central body, until at length the flickering energy is no longer able to keep these permanent gases away, and then the surroundings of the central body are invaded by these gases in such tremendous quantity that an absorption is produced which first turns the cooler star blood-red, and finally blots it out.
There are several very interesting questions connected with this. Suppose, for instance, that we attempt to discuss the future of that magnificent nebula in Andromeda, the true structure of which Mr. Roberts has recently revealed to us. It is already suspected that the two subsidiary swarms partake of the motion and form a part of the system. Those smaller swarms will naturally condense before the larger ones. Let us imagine ourselves no longer dealing with anything so far away, but with the solar system when it was in that stage. The central sun having this cordon round it can only be formed of those substances which are not repelled by its radiant energy; it will, therefore, be chiefly a mass of metallic vapor. The masses near it for the same reason will be also chiefly of metallic vapors, and their density will be high;
