pressure has been shown to do it, thanks to the labors of Humphreys and Mohler at Baltimore. "Anomalous refraction" may do it, as Professor Julius of Utrecht has found out. Finally, changes of density may produce it, as Michelson has discovered. To these causes we may confidently ascribe most of the shiftings in the stellar spectrum, for just such forces must be there at work.
Mr. Monck suggested the idea that new stars are the result of old dark stars rushing through gaseous fields in space and rendered luminous by the encounter. Seeliger revived and developed this idea, which in certain cases is undoubtedly the truth. Probably this occurred to the new star of 1885 which suddenly blazed out almost in the centre of the great nebula in Andromeda. It behaved like a typical nova and in due course faded to indistinguishability. Something like it happened, too, in the nova of 1860, which suddenly flared up in the star cluster 80 Messier, outdoing in lustre the cluster itself, and then, too, faded away.
But just as psychology teaches us that not only do we cry because we are sorrowful, but that we are sorrowful because we cry, so while a nova may be made by a nebula, no less may a nebula be made by a star.
Let us see how this might be brought about and what sign manuals it would present. Suppose that the two bodies actually grazed. Then the disruption would
