a deoxidizing effect when grown in solutions containing nitrates, with the production of nitrites and ammonia. This double action leads us to inquire into the nature of the physiological processes taking place in either case. His investigations in this field led Stoklasa to assert that there is much in common between the two processes. He believes that the deoxidation of nitrate is due to the action of the bacteria on water, with the liberation of hydrogen and the formation of hydroxyl. Nascent hydrogen is a powerful reducing agent, and would of itself withdraw the oxygen from nitrates to form nitrites, while hydroxyl in contact with ammonia will cause the formation of water and the liberation of nitrogen. Part of this nitrogen is undoubtedly utilized by the bacteria, and the rest is returned to the air. How the inert nitrogen molecule is torn apart and the nascent nitrogen atoms thus formed utilized by the bacteria for their growth is a question that is more difficult of solution. We do know that the amount of nitrogen fixed by B. megaterium is affected by the composition of the nutritive medium. The same is true of denitrification. The molecular structure of some carbohydrates or of organic acids and the arrangements of the atoms in the molecule influence the activity of the bacterial cell and its life processes. I have found, for instance, that the more complex citric acid molecule offers a more favorable source of energy to a denitrifying organism that I have isolated than do either succinic, tartaric or lactic acid. And the laboratory work clearly indicates that the amount of organic substance in the soil, as well as its nature, determines the course of development, and the prevalence of the one or the other of the soil organisms. Certain it is that where the fixation of nitrogen takes place in the soil, it occurs only when its store of nitrogen is very meager. This is analogous to the behavior of the legumes. It has been found that these plants when growing in a soil rich in soluble nitrogen do not to any considerable extent draw upon the atmosphere for that element. It is only when the soil offers little or no nitrogen that the atmospheric treasure house is unlocked for it. All experimental evidence thus far accumulated indicates that there is a struggle between the plant and the bacteria invading its roots, that the latter are so modified by the aggressive activity of their host that they form a fine network of tissue in which the atmospheric nitrogen is captured, as it were. That this assumption is not entirely erroneous is shown by the fact that legumes, inoculated with cultures of B. radicola that are particularly virulent, although they form root tubercles, show nitrogen hunger when there is none in the soil at their disposal, and the microscopic examination reveals bacteria that are not modified as is the case in vigorous plants. In other words, the bacteria resist the encroachment of their host and would not be compelled to furnish it with the nitrogen that it can not get otherwise. In connection with this, the question naturally
