After the star of 1572 the most celebrated is that which appeared in October, 1604, in Serpentarius, and which was observed by two illustrious astronomers, Kepler and Galileo. As happened with the preceding, its light imperceptibly faded. It remained visible for fifteen months, and disappeared without leaving any traces. In 1670 another temporary star, blazing out in the head of the Fox (Vulpecula), showed the singular phenomenon of being extinguished and reviving several times before it completely vanished. We know of _twenty-four_ stars which, during the last two thousand years, have presented a sudden increase of light; have been visible to the naked eye, often brilliant; and have then again become invisible. The last apparitions of this kind happened before our eyes in 1866, 1876, 1885, and 1892, and spectrum analysis enabled us to ascertain, as we have seen, that they were due to veritable combustion--a fire caused by a tremendous expansion of incandescent hydrogen, and to phenomena analogous to those which take place in the solar photosphere. A rather curious fact about these stars is, that they do not blaze out indifferently in any point of the sky, but in rather restricted regions, chiefly in the neighborhood of the Milky Way. Many other instances have been known, beside those referred to, of variable and temporary stars. There are two distinct kinds of double-stars. First, we have those which merely seem to be double, because one lies almost directly behind the other, though widely distant from it. Just as a church-tower, two miles off, may appear to stand close side by side with another church-tower two and a half miles off, though they are in fact separated. Secondly, we have the real systems of two suns belonging to one another; the smaller moving round the larger, or more correctly both traveling round one central point called the center of gravity, the smaller having the quicker rate of motion. [Illustration: LYRA] Alpha Centauri and 61 Cygni have been already described, as examples of true double-stars. In the constellation Lyra a marked instance is to be seen. The brightest star in the Lyre is Vega, and near Vega shines a tiny star, which to people with particularly clear sight has sometimes rather a longish look. If you examine this star through an opera-glass, you find it to consist of two separate stars. But if you get a more powerful telescope, and look again, you will find that _each star_ of the couple actually consists of _two_ stars. The four are not at equal distances. Two points of light seemingly close together are parted by a wide gap from two other points equally close together. These four stars are believed to have a double motion. Each of the separate pairs revolves by itself, the two suns traveling round one center; and in addition to this the two _couples_ of suns probably perform a long journey round another center common to them all.