Tycho Brahe, struck with the suddenness of the appearance of the star of 1572, and its position on the edge of the Milky-Way, offered a bold hypothesis which is now abandoned. He believed in a creation or at least in the spontaneous incandescence of the nebulous matter of which he supposed the Milky-Way to be composed: when the new star vanished, the place it had occupied was void, or at least Brahe saw, in the absence of nebulosity at that spot, the result of the condensation of the matter the combustion of which had produced an appearance resembling a star. Humboldt justly compares this view, which at all events was an ingenious one, with the views held by W. Herschel as to the transformation of nebulæ into stellar masses. In Tycho Brahe's time it was not known that the light of the Milky-Way results from the aggregation of an indefinite number of stars, or stellar masses, and that it is within this immense agglomeration and in its vicinity that the nebulæ properly so called are rarest.
Besides, it has been proved that the stars known as "new" stars are anything but new. In the spot where the star of Ophiuchus made its appearance in 1848, there had previously been a star, noted by Lalande in Fortin's "Atlas Céleste" as a vanished star. So, too, the new star which appeared in May, 1866, in the Corona Borealis, and which at the start reached the second magnitude, had been already set down in catalogues as a star of the ninth magnitude; it still possesses the same lustre it had before it underwent, during the six months of its apparition, the extraordinary augmentation which attracts to it the attention of astronomers. Hence astronomers no longer believe in the creation or in the destruction of these stars. Before the sudden incandescence which makes them visible, they occupied the same places, and there they still remain after their more or less perfect extinction. It remains to inquire into the physical causes which produce these variations of lustre.
Spectrum analysis has provided the first positive elements for the solution of this problem. As late as the year 1848 this method was as yet unknown; but when in 1866 the variable star of the Northern Crown (Corona Borealis) appeared, spectrum analysis was already so developed as to be profitably applied to the observation of it. The results obtained by Huggins and Miller in these researches were as follows:
- It is worth while to observe the ease with which these knotty questions are disposed of by persons who are wont to invoke supernatural agencies. Here is an instance dating from the seventeenth century, but similar instances may be found in our own time: Father Riccioli, as an explanation of the appearance of secondary stars, suggested the idea that some stars are luminous on one side; and whenever God would "exhibit to men some extraordinary sign, he turns toward them the luminous side (previously turned away from the earth) by causing the star to revolve about suddenly, either by the agency of some intelligent being, or in virtue of some faculty inherent in the star itself; then, by making another similar revolution, it suddenly vanishes, or pales gradually, like the moon in its phases." The explanation offered by the learned Jesuit is both ingenious and convenient. But, unfortunately, astronomers nowadays are not satisfied with it.