Herschel, that the visible surface of the Sun is a luminous envelope, within which there are cloudy envelopes covering a dark central body; and that, when by some disturbance the luminous envelope is broken through, portions of the cloudy envelope and of the dark central body become visible as the penumbra and umbra respectively. This hypothesis, at one time received with favor mainly because it seemed to permit that teleological interpretation which required that the Sun should be habitable, accounted tolerably well for certain of the appearances—more especially the appearance of concavity which the spots have when near the limb of the Sun. But, though Sir John Herschel supported his father's hypothesis, pointing out that cyclonic action would account for local dispersions of the photosphere, there has of late years become more and more manifest the fatal objection that the genesis of light and heat remained unexplained, and that no supposition of auroral discharges did more than remove the difficulty a step back; since, unless light and heat could be perpetually generated out of nothing, there must be a store of force perpetually being expended in producing them.
A counter-hypothesis, following naturally from the hypothesis of nebular origin, is that the mass of the Sun must be incandescent; that its incandescence has been produced, and is maintained, by progressing aggregation of its once widely-diffused matter; and that surrounding its molten surface there is an atmosphere of metallic gases continually rising, condensing to form the visible photosphere, and thence precipitating. What, in this case, are the solar spots? Kirchhoff, proceeding upon the hypothesis just indicated, which had been set forth before he made his discoveries by the aid of the spectroscope, contended that the solar spots are simply clouds, formed of these condensed metallic gases, so large as to be relatively opaque; and he endeavored to account for their changing forms as the Sun's rotation carries them away, in correspondence with this view. But the appearances as known to observers are quite irreconcilable with the belief that the spots are simply drifting clouds. Do these appearances, then, conform to the supposition of M. Faye, that the photosphere encloses matter which is wholly gaseous and non-luminous; and that the spots are produced when occasional up-rushes from the interior burst through the photosphere? This supposition, while it may be held to account for certain traits of the spots, and to be justified by the observed fact that there are up-rushes of gas, presents difficulties not readily disposed of. It does not explain the manifest rotation of many spots; and, indeed, it does not seem really to account for that darkness which constitutes them spots; since a non-luminous gaseous nucleus would be permeable by light from the remoter side of the photosphere, and hence holes through the near side of the photosphere would not look dark.
There is, however, another hypothesis which more nearly reconciles the facts. Assuming the incandescent molten surface, the ascending
