Oxidations in organic chemistry are usually bivalent. Univalent oxidation, not complicated by secondary associations or condensations, would produce radicals, molecules with an odd number of electrons which were considered as something exceptional. It is shown that this is not true for a great many of reversibly oxidizable‐reducible compounds. There is an intermediate level of oxidation, represented by a radical, which is in equilibrium with the fully oxidized and the fully reduced form. The value of the equilibrium constant varies from substance to substance, and, besides, varies to an enormous extent with pH. There are cases where the equilibrium constant is very large and the radical may exist to a large extent, and others in which the maximum amount of the intermediate radical is only a few per cent of the whole. Usually, in cationic dyestuffs, the intermediate compound, or semiquinone, exists, especially in very acid solutions, and vanishes in alkaline solution. For anionic dyestuffs the opposite is true. It is remarkable that in some dyestuffs occurring in living organisms and acting as respiratory enzymes or vitamins, a very distinct though small percentage of the dye can exist in the semiquinone form even in the physiological pH range around neutrality. The theoretical implications of this phenomenon for biology, and generally for the process of oxidation of organic compounds, is discussed.