The modes of disappearance of Sun-spots are as various as those of their apparition. The spots rarely close up by a gradual diminution or contraction of their umbra and penumbra. This mode of disappearance belongs exclusively to the spots deprived of faculæ and veils. One of the most common modes of the disappearance of a spot is its invasion by large facular masses, which slowly advance upon its penumbra and umbra and finally cover it entirely. It is a process precisely the reverse of that in which spots are formed by the shifting aside of the faculæ, as above described. In other types, the spots close up by the gradual enlargement of the luminous bridges traversing them, which are slowly transformed into branches of the photosphere, all of the characteristics of which they have acquired. In many cases, the spots covered over by the faculæ continue to exist for some time, hidden under these masses, as is often proved, either by the appearance of small spots on the facular mass left at the place they occupied, or even by the reappearance of the same spot. Apart from the general movement of rotation of the solar surface, some of the spots seem to be endowed with a proper motion of their own, which becomes greater the nearer the spots are to the solar equator. According to the observations of Mr. Carrington, the period of rotation of the Sun, as deduced from the observations of the solar spots during a period of seven years, is 25 days at the equator; while at 50° of heliocentric latitude it is 27 days. But the period of rotation, as derived from the observations of spots occupying the same latitude, is far from being constant, as it varies at different times, with the frequency of the spots and with the solar activity, so that at present the law of these variations is not well known. From the character of the solar envelope, it seems very natural that the rotation should differ in the different zones and at different times, since this envelope is not rigid, but very movable, and governed by forces which are themselves very variable.