so more than a thousand years or so; for, cosmically, the variability of a star is a mere transitory state, as it appears certain that irregularities of surface temperature must ultimately right themselves, although it is singular what a number of phenomena seem to tend in the direction of keeping a star that has been unequally heated by partial impact from having its uniform temperature restored. This matter is fully discussed in a paper in preparation on variable stars. Many doubles are coloured. I shall show in the same paper that in all probability the final state of variability in a star is a metallic absorbing atmosphere producing a coloured star; so that coloured doubles are probably the next youngest pair to the variable binaries. But although the variability of a star is a temporary state, their association with each other is not so. After having once absorbed the nebula their orbit is fixed, nothing but another impact can separate them and that is more likely to make a multiple star of them. The final coalescence of the visible Universe will only weld them into the general mass. It is not wonderful therefore that some 10,000 such pairs exist in the Universe. The fact that there are so few speaks to us in powerful language, telling us that the Universe is not so old as we have pictured it to be—that the first day is scarcely over in proportion to the time before its final coalescence. Without doubt this Universe is quite a new member of the Cosmos, of which it is not improbably as a mere drop in an ocean.

Art. XVI.—On a simple Method of illustrating the Motions of the Earth. By Professor A. W. Bickerton. [Read before the Philosophical Institute of Canterbury, 2nd September, 1880.] Plate IIB. This model is one of the extempore pieces of apparatus that I designed for the purpose of illustrating a course of experimental lectures, which were delivered with the special object of showing that many of the most important of physical phenomena might be illustrated by apparatus at a cost not exceeding a few pounds. The model itself cost less than a shilling, and I made it in about half-an-hour. Since it was made I have found it useful to illustrate so large a number of cosmical phenomena that I thought it of sufficient importance to bring before the Institute. A much larger number of phenomena may be illustrated by its means than by the expensive models usually sold for the purpose. Among these are day and