the land began once more to sink ; and brackish-water formations appear, followed later on by purely marine deposits. The Carboniferous shales and Mountain Limestone cover the former land with its imbedded plants. The great extension of the Mountain Limestone over many parts of Europe and America, and the small number of land formations which it covers, show us that this sinking of the land must have been a general phenomenon. The northern hemisphere must therefore most probably have had quite a different aspect at that time than during the Ursa stage. Then the same phenomenon recommenced as in the beginning of the Carboniferous period. In consequence of an extended rise to the west we obtain the continental formations of the Millstone-grit, which afterwards reached their greatest extent and development in the Coal- measures. During this long lapse of time the flora remained, on the whole, the same. Many of the leading species outlived all the mutations, and show that the whole of the land was never under water, even at the time of the formation of the Mountain Limestone ; there always remained enough dry land to support these species of plants, which afterwards extended their range when the land, as in the Millstone-grit, began again to have a greater extension. There can be no doubt that a long period must have elapsed between the beginning of the Ursa stage and the Millstone -grit, and that during these many thousand years the conditions of life of the organic beings must have undergone reiterated changes. It is therefore a very remarkable fact that, in spite of this, so many species lasted through this time and did not undergo any perceptible alteration. The many forms of Calamites radiatus which appear in Bear Island are also found in the lowest member of the Lower Carboniferous and in the roofing-slates of Moravia ; then the species disappears, nor does any similar form of this type (which Schimper has raised into a separate genus, Bornia) extend into the Coal-measures ; the same thing takes place in the species of Knorria and Cardiopteris. These facts tell very decidedly against the continuous and imperceptible transmutation of plant-species. They are the more important, since the plants on Bear Island must clearly have lived under quite different conditions of light than those in the Vosges or in Ireland; for they must have endured a long winter-night. It is indeed remarkable that evergreen trees, such as the Lepidodendra must probably have been, and plants with such large leaves as Cardiopter is frondosa could have withstood such a long winter-night, even if we take into consideration that the Bear-Island flora consisted almost entirely of vascular Cryptogams, which can better dispense with light than the Phanerogams. Moreover the climate of Bear Island must have been as favourable to the growth of plants as that of Ireland or the Vosges, although that island lies 26-1/2° further north ; for the corresponding species are as large and quite as luxuriantly developed, and have even produced more considerable coal-strata than those of lower latitudes in the same period. Warmth, therefore, must at that time have been more equally distributed over the earth, whilst already in the Miocene period a
