While _Entada scandens_ has spread in spite of unfavorable adaptation for seed floating, there are some plants whose seeds always float, and in spite of the sea water retain their power of germination. H. B. Guppy, from whom much of the above data are taken and who has experimented for years on the buoyancy and germinating power of sea-borne seeds, reckons about two hundred species that may have spread by ocean currents. Many of these are nearly world-wide in their distribution within their climatic requirements, and most are confined to the tropics. 2. INVASION, MIGRATION, AND RELICS It is perhaps a natural enough question why such elaborate and effective methods of seed dispersal are necessary and why plants, once they grow in any particular locality are not satisfied to stay there. The answer to this is that individual and racial competition is so great that without means of dispersal, which may be looked on as equipment for seeking a more favorable site, species would often be crowded out. There is no better place to see this than at the edge of a forest and grassland. The presence of the forest tells us at once that what may be described as the forest type of climate must have existed in the past to have produced the woods. In clearing off parts of this forest the openings will usually be grassland at first, but never in the end if nature is allowed to work out the solution. Along the edge of such a forest will be found a host of pioneers pushing out among the grasses, making ready, by conditions of shade, protection from drying winds and other influences, for the seedlings of the forest trees that creep slowly but resistlessly out to capture areas that by right of previous occupancy belong to them. In practically all parts of the world, with a few local exceptions, wherever the forest was the original type there is this ceaseless struggle to reclaim the open places, often or usually peopled by grass. It may be set down as almost a rule that if the open places produce herbs with broader leaves than grasses, the forest will capture the area many years sooner than if grass alone is the temporary tenant. Grass by its exclusive growth, its complete monopolization, so far as low seedlings are concerned, of much light and nearly all available surface water, is singularly well able to take care of itself once it is thoroughly established. But plotted and marked areas of this contact between forest and artificial clearings in it, show that in the end the forest will win, often at the rate of five hundred feet in a hundred years, sometimes much quicker than this.