The action of bacteria in causing decay has been said to be in general similar to fermentation--that it is effected by the bacteria in seeking their food. If oxygen be abundant, putrefaction occurs; if it be scant or absent, then fermentation takes place, for the tiny organisms require oxygen, and, if the air fails them, they pull to pieces the organic matters near them to obtain it. In doing this they get the nitrogen into such shape that the plants can use it, and thus digest their food for them. All organic matter contains carbon, hydrogen, and oxygen as a general rule, and to these are often united phosphorus, sulphur, nitrogen, and others, making very complex arrangements, veritable houses of cards, in fact, only held together by the strange power of life. When a leaf falls or a bird dies, some of these combinations are broken, and then the bacteria and other lowly organisms have full sway, for living matter is impregnable to all save a few of them. As oxygen or something else is taken out of the complex molecules, the compound falls to pieces, but as in the kaleidoscope the bits of colored glass tumble into endless varieties of symmetrical figures, so do the atoms fall into new combinations. If the keystone of an arch be removed, the stones fall apart; but atoms, unlike bricks or stones, can not stand alone as a rule; they must be united to something, and so, as soon as old associations are dissolved, new ones are formed. These new ones are those needed by plants, and thus is plant food digested. The term "plant food" has been frequently used, and should now be distinctly explained, for merely stating the chemical elements is not describing the food. When a physician tells a nurse to feed a patient he does not order so much carbon, nitrogen, phosphorus, and the like, but specifies a soup, certain vegetables, and so on, detailing every particular; and the same should be done for vegetable invalids. In medical practice a condition is recognized that is called scurvy. It is not exactly starvation, but is produced by lack of some food materials usually supplied by fresh vegetables. If scurvy appears at sea, no amount of meat, bread, cakes, or pastry will stop it; vegetables, and they only, will stay it. Sometimes a similar condition prevails among crops: some ingredient in a soil is lacking, and the others may be supplied indefinitely without giving the desired relief. To this may be attributed much of the fault found with fertilizers; for if the soil does not need a particular compound it is useless to apply it, and an excellent fertilizer is often blamed for not producing a crop on land already overstocked with it and crying for something else. Let us suppose a field on which cotton has been grown for many successive years until it has become exhausted. Analysis shows that a crop yielding one hundred pounds of lint to the acre removes from the soil: