Of the interior of the sun we can, of course, expect to learn little or nothing. What we observe is the surface-layer, the so-called photosphere, in which the cold of space produces the condensation of the gases into those luminous clouds which we see in our drawings and photographs as "rice grains" or "willow leaves." It has been suggested by Dr. Johnstone Stoney (and afterwards by Professor Hastings, of Baltimore) that these luminous clouds are mainly composed of carbon with those of the related elements silicon and boron, the boiling points of which are much higher than those of other elements which might be considered likely to form the photospheric clouds. The low atomic weight of carbon must also have the effect of giving the molecules of this element a very high velocity, and thereby enabling them to work their way into the upper regions, where the temperature has so fallen that the vapour becomes chilled into cloud. A necessary consequence of the rapid cooling of these clouds, and the consequent radiation of heat on a large scale, would be the formation of what we may perhaps describe as smoke, which settles by degrees through the intervals between the clouds (making these intervals appear darker) until it is again volatilised on reaching a level of greater heat below the clouds. This same smoke is probably the cause of the well-known fact that the solar limb is considerably fainter than the middle of the disc. This seems to arise from the greater absorption caused by the longer distance which a ray of light from a point near the limb has to travel through this layer of smoke before reaching the earth. It is shown that this absorption cannot be attributed to a gaseous atmosphere, since this would have the effect of producing more dark absorption lines in the spectrum. There would thus be a marked difference between the solar spectrum from a part near the middle of the disc and the spectrum from a part near the limb. This, however, we do not find to be the case. With regard to the nature of sun-spots, the idea first suggested by Secchi and Lockyer, that they represent down rushes of cooler vapours into the photosphere (or to its surface), seems on the whole to accord best with the observed phenomena. We have already mentioned that the spots are generally accompanied by faculae and eruptive prominences in their immediate neighbourhood, but whether these eruptions are caused by the downfall of the vapour which makes the photospheric matter "splash up" in the vicinity, or whether the eruptions come first, and by diminishing the upward pressure from below form a "sink," into which overlying cooler vapour descends, are problems as to which opinions are still much divided.