of appalling length, certain exceptional conditions arise by which this dark piece of matter may be so kindled that, for an epoch, long it may be in years but brief indeed when compared with the span of its total existence, the body would glow as a star.
Provided with this conception let us look on the universe with its millions of orbs. These orbs will be found in every state possible to such bodies; but the enormous majority of them must, in accordance with the principles just laid down, be in the dark and invisible state. Out of some millions it may perhaps be concluded that, at any particular moment, a dozen or so might, by accidental circumstances, be in those phases of their several careers in which luminosity is a characteristic. In such cases only will the orbs be visible. The instructed astronomer will, therefore, believe that the non- visible orbs must be hundreds, thousands, or perhaps millions of times more numerous than those which he can see. When we remember that, by our telescopes and on our photographs, we can discern something like one hundred million luminous stars in the sky; when we remember that every one of these is the indication of a wholly exceptional incident in the career of the body from which the light emanates; and when we further believe, as believe we must, that for each one star which we can thus see there must be a stupendous number of invisible masses, then, indeed, we begin to get some notion of the extraordinary multitude in which material orbs are strewn through space. The theory of probabilities declares to us with a certainty, hardly, in my opinion, inferior to that of optical demonstration, that even within the distance which can be penetrated by our telescopes the visible stars cannot form the
