Another point to be noticed is the arrangement of stars in classes, according to their lustre.  The brightest stars, of which there are about twenty, are said to be of the first magnitude.  Those just inferior to the first magnitude are ranked as the second; and those just lower than the second are estimated as the third; and so on.  The smallest points that your unaided eyes will show you are of about the sixth magnitude.  Then the telescope will reveal stars still fainter and fainter, down to what we term the seventeenth or eighteenth magnitudes, or even lower still.  The number of stars of each magnitude increases very much in the classes of small ones.

Most of the stars are white, but many are of a somewhat ruddy hue.  There are a few telescopic points which are intensely red, some exhibit beautiful golden tints, while others are blue or green.

There are some curious stars which regularly change their brilliancy.  Let me try to illustrate the nature of these variables.  Suppose that you were looking at a street gas-lamp from a very long distance, so that it seemed a little twinkling light; and suppose that some one was preparing to turn the gas-cock up and down.  Or, better still, imagine a little machine which would act regularly so as to keep the light first of all at its full brightness for two days and a half, and then gradually turn it down until in three or four hours it declines to a feeble glimmer.  In this low state the light remains for twenty minutes; then during three or four hours the gas is to be slowly turned on again until it is full.  In this condition the light will remain for two days and a half, and then the same series of changes is to recommence.  This would be a very odd form of gas-lamp.  There would be periods of two days and a half during which it would remain at its full; these would be separated by intervals of about seven hours, when the gradual turning down and turning up again would be in progress.

The imaginary gas-lamp is exactly paralleled by a star Algol, in the constellation of Perseus (Fig. 3), which goes through the series of changes I have indicated.  Ordinarily speaking, it is a bright star of the second magnitude, and, whatever be the cause, the star performs its variations with marvellous uniformity.  In fact, Algol has always arrested the attention of those who observed the heavens, and in early times was looked on as the eye of a demon.  There are many other stars which also change their brilliancy.  Most of them require much longer periods than Algol, and sometimes a new star which nobody has ever seen before will suddenly kindle into brilliancy.  It is now known that the bright star Algol is attended by a dark companion.  This dark star sometimes comes between Algol and the observer and cuts off the light.  Thus it is that the diminution of brightness is produced.

[Illustration:  FIG. 3.  PERSEUS AND ITS NEIGHBORING STARS INCLUDING ALGOL.]