No. 6.

On Variable Stars.

The great number of these stars and their peculiar action seems to point to some definite cause at work in the galaxy producing them. Partial impact appears quite competent to explain the chief variety of these phenomena. I have already shown that, taking into consideration the large proper motion of the stars, that nearly all cosmical collisions will be cases of partial impact, and that generally after the collision three bodies will be produced. A middle intensely heated body, formed by the coalescence of those parts of the two colliding bodies which lie in each other's path. (The molar motion of this mass will be generally inconsiderable, and it may form a temporary star, a nebulae, or a solar system, according to the varying conditions of the impact.) Secondly, the two retreating masses from which the middle part has been struck off. These may be either gaseous, liquid, or solid. If either liquid or solid, it appears almost certain that variable stars must be formed. Let us try and imagine one of these gigantic retreating bodies. The high velocity attained at impact will quickly diminish, both from the attraction of the other retreating body, and also from that of the coalesced mass, and in a few years its motion will probably be less than that of its original proper motion. I he work of cutting off the piece, I have elsewhere demonstrated, would scarcely tell in producing a lessened velocity. The original spheroidal mass would have been stupendously distorted during impact. (I have shown elsewhere that if it were as hard as adamant it would scarcely modify the amount of the distortion that would be produced by attraction in such large bodies as we are considering.) After impact a large piece would be cut out of one side, but as soon as the bodies were free from each other, their own stupendous attraction would render them spheroidal again. But, although it has regained its shape, the two sides of the mass must be at extremely different temperatures. There are many causes at work which would produce this result:—1st. The work of sheering; secondly, the enormous amount of molar motion destroyed in the entangled parts of the other retreating body, chiefly, perhaps, its gaseous atmosphere; thirdly, the exposure of its more highly-heated interior; fourthly, by the distortion and restoration of sphericity, which would also be chiefly on that side, Thus is produced a generally slowly revolving body, one side of which might be almost or quite non-luminous, and the other side an actual sun—in other words, a variable star. With gaseous and liquid convection and diffusion of such a tremendous nature as this condition of things would set up, such a state would appear to be a most fugitive one; but this is not really the case; let us take a body say twenty times the size of our sun; let us suppose the fiery lake to be equal to the sun. Let us suppose the body to be liquid with a tremendous atmosphere, what must happen? A stupendous uprusa from the boiling molten mass, convection currents carrying the heated gas over the entire globe. But as the heated gas rises, it does work, is cooled, and the matter is precipitated back again as metallic rain, so that the amount of heat carried over to the other parts of the body by this cause is comparatively email. But there is liquid connection at work, tending to restore equilibrium of temperature? Not necessarily, for convection is produced by difference of density, and this may be due to either chemical composition or to the same element at different temperatures. In such a body it is extremely probable that these two forces are tending to neutralise each other. Several reasons have been urged for supposing the centres of a globe to be composed of the more dense elements; hence when these are exposed in a variable star the two forces will quickly tend to equilibrium. There is thus only conduction and diffusion left. In such a body as we have been discussing, hundreds of years would not be too large a unit in which to talk of the period necessary to attain moderately uniform temperature. Most of the variable stars are of this character. The variable stars which are the most abnormal are Mira Algol and Eta Argus. The first is invisible two-thirds of the year, gradually becomes very bright, stays so a few days, gradually becomes dark again, and disappears. Clearly a revolving black body with a molten lake would explain it exactly. All the peculiarities of Algol is explained by assuming a large dark body with a bright body revolving around it, which occa- page 8 sionally suffers eclipse by passing partly behind. Eta Argus appears to have a rotation of its own, which, produces a variable star, and, in addition to this, has revolving around it another body which comes into contact with it at each revolution, so developing intense heat. But it is useless to go into details, as partial impact is too wide and the number of variables too great to do more than merely glance at its action in these letters.

Note.—Now we have found to many variable stars to be in pairs, we require to know the proper motion of the constituents of those pairs of variables, which are too distant to be orbitially connected. Of course, this theory requires that generally they should be directed outward. When this proper motion is accurately known, we shall be able to ascertain their approximate age; Chinese and other records may then be searched to ascertain if any temporary star occurred at about that data in that part of the heavens. Of course, the angular proper motion across the line of sight is all that is required for this purpose. The time so obtained would have to be reduced by the greater initial velocity after impact. Astronomers might also see if a email star or nebula lies between these pairs of variables; also look for another variable near the isolated variables.