Garden experiments are of little use as tests of the capability of any plant to survive in this country; the study of cornfield weeds is no better. In both cases the cultivation of the land produces a bare place on which a foreign introduction has as good a chance as a native. But could this foreigner survive if the seed were dropped on a natural moor or meadow? In this connexion it is noticeable that great part of the rare British plants occur close to the coast, opposite the part of the continent in which they are found, though they are not maritime species. This is probably due to two different causes, both acting in the same direction. In the first place most of these local plants are obviously late comers, which have not yet had time to spread inland or far. And, secondly, on the coast alone do we find any considerable extent of natural bare land—practically garden land—which does not at the same time consist of poor soil. The tumbled undercliffs of our coast are just the places to give a foreign invader a chance; there only will it find patches of bare good soil, full of small cracks in which a seed is hidden from birds. If the view is correct, that a continuous growth of our flora, and to some extent of our fauna, takes place through transportation to our coasts, from which such species as can fight their way tend more slowly to spread inland, it seems to account for the present curious distribution of species, and this in a way that no continuous land-connexion will do. As we have pointed out in a former chapter, the land-connexion across the North Sea was a wide alluvial plain and swampy delta. What use could dry-soil plants make of such a bridge? It would be no easier for them to cross than so much sea; and migrating mammals could not greatly help in the dispersal, where so many rivers had to be crossed. The aquatic species would be helped by such a connexion, and it is curious to note that several of our most interesting aquatic plants are confined to the eastern counties, which in post-glacial times had direct connexion with the delta of the Rhine, and probably with the Elbe. Aquatic species, however, are not dependent on continuous waterways for their dispersal; they have great facilities for overleaping barriers and reaching isolated river-basins and lakes. Every dew-pond on the downs after a few years’ existence contains aquatic plants and mollusca, and a still larger number of species, including fish, will be found in ancient flooded quarries or prehistoric dykes surrounding some hill-fortification. If an aquatic plant is fairly common on the continent near by, it is almost certain to occur in some isolated pond or river in the part of Britain opposite.