tangled with one another, sometimes neatly truncated, or again spread out with no distinct termination. Their head is generally turned toward the centre of the nucleus. They are not unlike strokes of color laid on with a painting-brush, very white near the head, and gradually less brilliant toward the tail." Fig. 5 represents a group of solar spots, in which this closely-meshed appearance of the ovoid bodies is impressively presented; and in Fig. 6 we see the willow-leaf structure stretching completely across the umbra and forming bridges of light.
The dimensions of solar spots are as variable as their forms. Some are very small, appearing under the highest magnifying power like mere specks; others are of enormous magnitude. Many have been observed which measure from 30,000 to 100,000 miles in diameter, so that, if they are truly chasms and gulfs in the luminous envelope of the sun, our entire globe, as Guillemin remarks, would appear in their depths no larger than a fragment of a rock rolled into the crater of a volcano!
The changes which these bodies undergo are extraordinary. They seem to have an order of development. Proctor says the formation of a spot is usually preceded by the appearance of faculæ. Then a dark point makes its appearance, which increases in size, the penumbral fringe being presently recognized around it, and the distinction between the umbra and the penumbra being well defined. But, when the spot is about to diminish, the edges lose their sharpness as if screened by a luminous veil; capes and promontories jut out, and bridges of light are formed. Lockyer says sometimes changes are noticed even within an hour. Their periods are most variable; the smallest merely appear and disappear, lasting a mere fraction of the time of solar rotation. The larger live frequently during two rotations, and Schwabe saw one which returned eight times, continuing 200 days.
The remarkable changes that sometimes occur are well illustrated in Fig. 7, which represents four drawings of the large spot that appeared October 7, 1865, and which was more than 46,000 square miles in area. The drawings are numbered in the order of date. No. 1 exhibits the oblong, fore-shortened view presented when it first appeared on the edge of the sun. No. 2 represents its aspect three days afterward. No. 3 shows its appearance four days later, and No. 4 is a view of it on October 16th.
Sometimes spots exhibit a rotatory or whirlpool appearance, as if the solar envelope were subject to tremendous tornadoes. Fig. 8 represents a spot seen and drawn by Secchi at Rome, May 5, 1857, which represents a vortex, into which the substance of the photosphere is rushing with an eddying motion. De La Rue took two photographic pictures of the same spot at an interval of two days, and, when these are placed together and looked at through the stereoscope, the spot exhibits a funnel-like concavity with remarkable exactness. There is other evi-
