[10] Also named ξ Toucani. Near β Aquarii there shines with the light of a sixth-magnitude star another “magnificent ball of stars” which has been compared to “a heap of fine sand.” It is named 2 M Aquarii. Over seventy of these tightly packed balls of stars are known, even counting only the brightest, and their distribution is rather curious. A large number (about twenty) occur in the Clouds of Magellan, and more than half of the seventy are in the Milky Way, not scattered evenly along its course, but almost if not entirely confined to its southern part, and chiefly gathered in a great group in its brightest portion, where it passes through Sagittarius, Ophiuchus, and Aquila. Here they are mingled with—or perhaps projected against—numerous stars of the same magnitudes; but many balls are also found outside the Milky Way, widely scattered, and in these parts of the sky there are relatively few of the faint-magnitude stars which compose all the globular clusters. 47 Toucani, for instance, though it is near the small Magellanic Cloud, stands quite apart from it, isolated in a black sky. We do not know the distances of any of these balls of stars. Those which have been examined spectroscopically shine like Canopus—that is, they are of a type intermediate between Sirius and our sun—but the chief light comes, of course, from the brighter stars, and it may be that the fainter stars mingled with them belong to a different type. A remarkable fact lately discovered is that many globular clusters—but not all—contain a large number of variable stars. These vary in light in a period of about a day and have a range of about one magnitude. They are not of the Algol type, nor quite of the usual “short-period” types, and it is not yet clear what is the cause of variation, though it seems probable that “cluster-variables” are double stars. XIV NEBULAE Athwart the False Cross, from δ Velorum to ι Carinae, a line passing on leads to the round white spot which we found to be a star-cluster. A little further in the same direction is a larger curved white patch, bright enough to be visible, once it is familiar, even after the moon has risen. This is the Great Nebula in Argo, the Keyhole Nebula, in which Eta Argūs once blazed out. Even a binocular will divide it into two parts separated by a chasm, and will show the pearly background powdered over with many small stars. But even the most powerful telescopes do not resolve this pale background into stars, as they resolve the star-cluster just mentioned: it remains a pearly mist, the brighter part strangely broken by dark rifts, the fainter, beyond the chasm, a tangled skein of long cloudy streaks reaching out into the darkness and gradually, irregularly, fading away.