Of cosmic causes, the progressive cooling of the earth naturally presents itself, at the first blush, as sufficient to account for the glacial period. But although this has doubtless been an all-important factor in pregeological times, in fashioning our planet from glowing vapour into a habitable earth, it is no longer operative as an immediate cause of vicissitudes of temperature. It is enough to say that if it were, the cooling ought to be progressive, and having once got into a glacial period we never ought to have got out of it. But we clearly have recovered from the paroxysms of cold, both of the first and second great glaciations of the recent period; and according to most geologists, from the immensely earlier ones of the Permian and Carboniferous, and perhaps of the Cambrian ages. As far as it acts at all on surface temperature, the secular cooling of the earth only acts indirectly by causing elevations and depressions of the outer crust, and crumpling it into wrinkles, which originate mountain chains, as the nucleus contracts, and thus affecting geographical conditions. The same objection applies with equal force to the theory that the glacial period was caused by the sun giving out less heat owing to its cooling by radiation. Here also it is obvious that if a glacial period were once established from such a cause it ought never to recover, but progress from bad to worse. We ought also, in this case, to have had a uniform progressive refrigeration from the beginning of geological time down to the present day, which has certainly not been the case. On the contrary, geologists are generally agreed that there are unmistakable traces of at least two glacial periods in the Carboniferous and Permian ages, and the earliest Eocene was certainly cooler than its later stages, as shown by their flora. The conjecture that the sun is a variable star is also negatived by the consideration that in this case there ought to have been periodical variations in the earth's temperature, and hot and cold climates recurring at regular intervals throughout geological time, which has certainly not been the case. Again, the passage of the solar system through cold regions of space has been suggested, but it is a mere conjecture, unsupported by a particle of evidence, and opposed to all we know of the laws of heat, and of the constitution of the universe. It is hard to conceive how hot regions can exist surrounded by cold ones, or _vice versâ_, without walls of a non-conducting medium to separate them, or that the faint heat from the fixed stars can ever have perceptibly affected the temperature of space. And such a theory, if it were possible, would fail to account for the frequent vicissitudes of hot and cold at short intervals within the glacial period, and for the great differences of temperature prevailing in the same latitudes.