I record here an observation which, at first sight, may appear paradoxical; but which seems, however, to be of considerable importance, as it shows unmistakably that the solar light is mainly, if not entirely, generated on its surface, or at least very near to it. On May 26, 1878, I observed a large group of Sun-spots at a little distance from the east limb of the Sun. The spot nearest to the limb was partly covered over on its eastern and western sides by bright and massive faculæ which concealed about two-thirds of the whole spot, only a narrow opening, running from north to south, being left across the middle of the spot. Owing to the rotundity of the Sun, the penumbra of this spot, although partly covered by the faculæ, could, however, be seen on its eastern side, since the sight of the observer could there penetrate sidewise under the faculæ. Upon that part of the penumbra appeared a strong shadow, representing perfectly the outline of the facular mass situated above it. The phenomenon was so apparent that no error of observation was possible, and a good drawing of it was secured. If this faculæ had been as bright beneath as it was above, it is evident that no shadow could have been produced; hence the light of these faculæ must have been mainly generated on or very near their exterior surfaces. This, with the well-proved fact that the bright faculæ lose their light in falling into the interior of the Sun, seems to suggest the idea that the bright light emitted by the faculæ, and very probably all the solar light, can be generated only on its surface; the presence of the coronal atmosphere being perhaps necessary to produce it. Several times before this observation, I had suspected that some faculæ were casting a shadow, but as this seemed so improbable, my attention was not awakened until the phenomenon became so prominent that it could not escape notice. With due attention, some glimpses of the phenomenon can frequently be observed through the openings of some of the faculæ projecting over the penumbra of Sun-spots. It is very seldom that the structure of the penumbra is seen through such openings, which usually appear as dark as the umbra of the large spots, although they do not penetrate through the photosphere like the latter. It is only when the rents in the faculæ are numerous and quite large, that the penumbral structure is recognized through them. Since these superficial rents in the faculæ do not extend through the photosphere, and appear black, it seems evident that the penumbra seen through them cannot be as bright as it is when no faculæ are projected upon it, and therefore that the faculæ intercept light from the exterior surface, which would otherwise reach the penumbra.