The third great cause of warmer climates in the past is the larger proportion of carbon dioxide, or carbonic-acid gas, in the atmosphere in early geological times, as proved by the immense amount of carbon now sealed up in limestone and coal, and which must at one time have been in the air. It has been shown that a very small additional quantity of this substance would so obstruct radiation of heat from the earth as to act almost like a glass roof. If, however, the quantity of carbonic acid, great at first, was slowly and regularly removed, even if, as suggested by Hunt, small additional supplies were gradually added from space, this cause could have affected only the very oldest floras. But it is known that some comets and meteorites contain carbonaceous matter, and this allows us to suppose that accessions of carbon may have been communicated at irregular intervals. If so, there may have been cycles of greater and less abundance of this substance, and an atmosphere rich in carbon dioxide might at one and the same time afford warmth and abundance of food to plants. It thus appears that the causes of ancient vicissitudes of climate are somewhat complex, and when two or more of them happened to coincide very extreme changes might result, having most important bearings on the distribution of plants. This may help us to deal with the peculiarities of the great Glacial age, which may have been rendered exceptionally severe by the combination of several of the causes of refrigeration. We must not suppose, however, that the views of those extreme glacialists who suppose continental ice-caps reaching half way to the equator are borne out by facts. In truth, the ice accumulating round the pole must have been surrounded by water, and there must have been tree-clad islands in the midst of the icy seas, even in the time of greatest refrigeration. This is proved by the fact that, in the Leda clay of eastern Canada, which belongs to the time of greatest submergence, and whose fossil shells show sea-water almost at the freezing-point, there are leaves of poplars and other plants which must have been drifted from neighbouring shores. Similar remains occur in clays of like origin in the basin of the great lakes and in the West. These have been called "interglacial," but there is no evidence to prove that they are not truly glacial. Thus, while we need not suppose that plants existed within the Arctic circle in the Glacial age, we have evidence that those of the cold temperate and sub-arctic zones continued to exist pretty far north. At the same time the warm temperate flora would be driven to the south, except where sustained in insular spots warmed by the equatorial currents. It would return northward on the re-elevation of the land and the renewal of warmth.