To return once again to our island, if by the continued action of the subterranean forces other neighbouring islands were formed, these would generally be stocked by the inhabitants of the first island, or by a few immigrants from the neighbouring mainland; but if considerable obstacles were interposed to any communication between the terrestrial productions of these islands, and their conditions were different (perhaps only by the number of different species on each island), a form transported from one island to another might become altered in the same manner as one from the continent; and we should have several of the islands tenanted by representative races or species, as is so wonderfully the case with the different islands of the Galapagos Archipelago. As the islands become mountainous, if mountain-species were not introduced, as could rarely happen, a greater amount of variation and selection would be requisite to adapt the species, which originally came from the lowlands of the nearest continent, to the mountain-summits than to the lower districts of our islands. For the lowland species from the continent would have first to struggle against other species and other conditions on the coast-land of the island, and so probably become modified by the selection of its best fitted varieties, then to undergo the same process when the land had attained a moderate elevation; and then lastly when it had become Alpine. Hence we can understand why the faunas of insular mountain-summits are, as in the case of Teneriffe, eminently peculiar. Putting on one side the case of a widely extended flora being driven up the mountain-summits, during a change of climate from cold to temperate, we can see why in other cases the floras of mountain-summits (or as I have called them islands in a sea of land) should be tenanted by peculiar species, but related to those of the surrounding lowlands, as are the inhabitants of a real island in the sea to those of the nearest continent{408}. {408} See _Origin_, Ed. i. p. 403, vi. p. 558, where the author speaks of Alpine humming birds, rodents, plants, &c. in S. America, all of strictly American forms. In the MS. the author has added between the lines "As world has been getting hotter, there has been radiation from high-lands,--old view?--curious; I presume Diluvian in origin."