If the constitution of the Sun be that which is above inferred, it does not seem difficult to conceive still more specifically the production of these appearances. Everywhere throughout the atmosphere of metallic vapours which clothes the solar surface, there must be ascending and descending currents. The magnitude of these currents must obviously depend on the depth of this atmosphere. If it is shallow, the currents must be small; but if many thousands of miles deep, the currents may be wide enough to render visible to us the places at which they severally impinge on the limit of the atmosphere, and the places whence the descending currents commence. The top of an ascending current will be a space over which the thickness of condensed cloud is the least, and through which the greatest amount of light from beneath penetrates. The clouds perpetually formed at the top of such a current, will be perpetually thrust aside by the uncondensed gases from below them; and, growing while they are thrust aside, will collect in the spaces between the ascending currents, where there will result the greatest degree of opacity. Hence the mottled appearance--hence the "pores," or dark interspaces, separating the light-giving spots.[25] Of the more special appearances which the photosphere presents, let us take first the faculæ. These are ascribed to waves in the photosphere; and the way in which such waves might produce an excess of light has been variously explained in conformity with various hypotheses. What would result from them in a photosphere constituted and conditioned as above supposed? Traversing a canopy of cloud, here thicker and there thinner, a wave would cause a disturbance very unlikely to leave the thin and thick parts without any change in their average permeability to light. There would probably be, at some parts of the wave, extensions in the areas of the light-transmitting clouds, resulting in the passage of more rays from below. Another phenomenon, less common but more striking, appears also to be in harmony with the hypothesis. I refer to those bright spots, of a brilliancy greater than that of the photosphere, which are sometimes observed. In the course of a physical process so vast and so active as that here supposed to be going on in the Sun, we may expect that concurrent causes will occasionally produce ascending currents much hotter than usual, or more voluminous, or both. One of these, on reaching the stratum of luminous and illuminated cloud forming the photosphere, will burst through it, dispersing and dissolving it, and ascending to a greater height before it begins itself to condense: meanwhile allowing to be seen, through its transparent mass, the incandescent molten shell of the sun's body. [The foregoing passages, to most of which I do not commit myself as more than possibilities, I republish chiefly as introductory to the following speculation, which, since it was propounded in 1865, has met with some acceptance.]