I have already indicated the part which the great air-ways played in distribution of _B. diazotans_ over the world; but once it had been planted in the new centres to which the aeroplanes carried it, other factors came into action. From South-western Europe, the North-East Trade Winds bore the bacilli across the Atlantic and spread them upon the seaboard of South America, especially around the mouths of the Amazon. The winds on the coast of North America caught up the germs and drove them eventually to Scandinavia and even further east. New Guinea, Borneo, Sumatra and the other islands of the chain were devastated from the Australian centres. Madagascar was contaminated also, though the point of origin in this case is not definitely known. Probably the ocean currents played their part, as they certainly did in the destruction of Polynesian vegetation. Climate had a considerable influence upon the development of the bacilli, once they were scattered. In the Tropics, they multiplied with even greater rapidity than they had done in the North Temperate Zone. On the Congo and in the Amazonian forests they seem to have undergone a process of reproduction almost inconceivably swift. Those which drifted up into the frigid regions of the North and South, however, appear to have perished almost without a struggle: either on account of the low temperature or the lack of nitrogenous material, they produced very little effect in either of these districts. The sea-plants seem to have been unaffected by them there; and one of the strangest results of this inactivity was the complete change in habits of various fishes, which now sought in the freezing North the feeding and breeding-grounds which suited them best. The herring left the North Sea and the cod quitted the Banks in search of purer water. On the other hand, the great masses of weed in the Sargasso Sea were almost completely destroyed, along with the other accumulations south-east of New Zealand and in the North Pacific. It must not be assumed, however, that wherever the colonies of _B. diazotans_ alighted, devastation followed as a matter of course. For some reason, which has never been made clear, certain areas proved themselves immune from attack; so that they remained like oases of cultivable land amid the surrounding deserts. The areas thus preserved from sterility were not of any great size; usually they amounted only to a few hundred acres in extent, though in isolated cases larger tracts were found unaffected here and there. With the recognition of the world-wide influence of _B. diazotans_, the land became divided into two sections: the food-producing districts and the consuming but non-productive areas. Nowhere was there sufficient grain to make safety a certainty. In America, most of the available food-stuffs were still in or near their places of origin when the panic began to grow.