About the sun-spots we notice the so-called faculæ—portions which are much brighter than their surroundings. When we carefully examine a strongly magnified image of the sun, we find that it has a granulated appearance (Fig. 18). Langley compares the disk to a grayish-white cloth almost hidden by flakes of snow. The less bright portions are designated “pores,” the brighter portions “granules.” It is generally assumed that the granules correspond to clouds which rise like the clouds of our atmosphere on the top of ascending convection currents. But while the terrestrial clouds are formed of drops of rain or of crystals of ice, the granules consist probably of soot—that is to say, condensed carbon—and of drops of metals, iron, and others. The smallest granule which we are able to discern has a diameter of about 200 km. (130 miles). [Illustration: Fig. 18.—Sun-spot group and granulation of the sun. (Photographed at the Meudon Observatory, near Paris, April 1, 1884)] The faculæ are formed by very large accumulations of clouds which are carried up by strong ascending currents and spread over large areas, as in our cyclones. The spots correspond to descending masses of gas with rising temperatures, which are therefore “dry” and do not carry any clouds, as in terrestrial anticyclones. Through these holes in the walls of solar clouds we peep a little farther into the gigantic masses of gas, and we obtain an idea of the state of affairs in the deeper strata of the sun. The depth of the wall of cloud is, of course, not large compared to the radius of the sun. [Illustration: Fig. 19.—Part of the solar spectrum of January 3, 1872. After Langley. The bright horizontal bands are due to prominences. In the middle (at 208) the hydrogen line F, strongly distorted by violent agitation]