Nearly all the stars we know have a place in this series, though there are individual peculiarities, but there is one class which seems to lie outside it. Stars of this class are red and have spectra crossed by dark bands, but they are unlike the Antarian bands and resemble instead those seen in the spectra of comets and of candle-flames. They are due to carbon compounds, so these stars may be called CARBON STARS. Most of them are extremely remote, and all are so faint that among the very brightest is U Hydrae, just visible to the naked eye a little east of Alphard. They are probably aged stars, but no links between them and the other types have yet been discovered to enable us to place them in the series. Strange and interesting discoveries have been made by grouping large numbers of stars into their classes and comparing the average motions, distances, &c., of the groups. It is found that the redder stars are on an average moving more rapidly and in a more random fashion than blue and white stars. Thus, Wolf-Rayet and Orion stars have a low average speed, and both are very much more numerous in and near the Milky Way than elsewhere; Sirian stars are travelling a good deal faster, show a marked tendency to congregate in two streams, and move chiefly parallel with the plane of the Milky Way; solar stars, including our own sun, move more rapidly still and show less preference for the Galaxy and the two streams: Antarians have the most rapid motions of all, but these appear to be haphazard, and the stars are scattered all over the sky in every direction. These facts are very unexpected and very difficult to explain. It looks as if the Milky Way were the birthplace of the stars, and that as they develop they gradually scatter through space; but how are we to explain the fact that speed and direction of movement differ for different types? There seems to be no reason why a cooler star should move more quickly than a hotter one, and none of the theories yet advanced can be considered final. V SOME NEAR NEIGHBOURS If it is remarked that Sirius is fifty millions of millions of miles away from us, it is not at once obvious that he is one of our very near neighbours; but this is equal to 8 light-years, not twice the distance of α Centauri, our next-door neighbour among the stars. Some faint stars in the south must also be counted as very close to us: such are a little star in Cetus, τ Ceti, only 10 light-years away; ε Indi, 11½; and two in the River Eridanus, ε and δ Eridani, 10½ and 18 respectively.