During the last decade it has been recognized that the process of photosynthesis in green plants is unique, not because it involves a complicated photochemical decomposition of carbon dioxide for which there is no analogy in the organic world, but because it combines in a unique manner a number of processes each of which may be found in other living cells. If we turn from green plants to purple bacteria, for instance, we find that radiant energy is utilized for the reduction of carbon dioxide. These organisms, however, cannot use simply water as a hydrogen donor and hence are not able to liberate free oxygen. For the reduction of carbon dioxide they depend, in addition to light, upon energetically valuable hydrogen donors such as free hydrogen, hydrogen sulphide, or organic acids. The over-all energy balances of these photoreductions are, therefore, much less favourable than that of photosynthesis in plants. If we turn to organisms not sensitive to light, we find that carbon dioxide can be reduced in complete darkness by several species of bacteria and even by some animal tissues. In this case the mechanism is a coupled oxido-reduction in which an excess of hydrogen donors, either of inorganic or organic nature, has to be sacrificed to promote the ‘chemosynthesis’. It is clear that such dark reactions lead not to a gain but to an over-all loss of chemical energy.Recent advances in the field of respiration and fermentation have taught us that despite the infinite variety of metabolic reactions in living cells the principles governing them are few. Accordingly, it is conceivable that the different reactions involving a reduction of carbon dioxide have many important traits in common, and that the study of any one of them may lead to a better understanding of the process of photosynthesis.The present article is a report on the metabolism of certain unicellular chlorophyllous algae (several species of Scenedesmus, Ankistrodesmus, Rhaphidium) that are able to reduce carbon dioxide either in normal photosynthesis with the evolution of oxygen, or in photo-reduction with the absorption of an equivalent amount of hydrogen, or in chemosynthesis with the oxyhydrogen reaction as the driving force. The two latter reactions do not occur under normal aerobic conditions. They can be observed only after a few hours' incubation in hydrogen gas. The anaerobic treatment brings into play a hydrogenase which enables the algae to absorb or to release molecular hydrogen. This metabolic change we call adaptation. The adaptation consists in an enzymatic activation or rearrangement of some of the catalytic systems. It can be inhibited by traces of specific poisons like cyanide and hydroxylamine.Upon illumination in the adapted state, in presence of hydrogen and carbon dioxide, the algae reduce carbon dioxide with twice the volume of hydrogen, exactly akin to some purple bacteria. This we call photoreduction. The results of experiments with flashing light and with specific poisons show that in photosynthesis and photo-reduction the truly photochemical reactions are the same and remain unchanged. Hence, the difference appears to originate from the ways by which the oxidized products of the photochemical reaction are eliminated. In photosynthesis they are decomposed with the liberation of oxygen, in photoreduction they are reduced to water by hydrogen donors.The adapted state of the algae gives way to normal aerobic conditions not only under the influence of air, but also under the influence of higher light intensities. This we call reversion. It seems that reversion occurs whenever some intermediate oxidized products (which we call ‘peroxides’, because they must be the precursors to molecular oxygen) accumulate faster than they are reduced by the hydrogenase system. In absence of carbon dioxide no intermediate oxidized products (or ‘peroxides’) are formed and the adapted state is stable even at high light intensities. If not only carbon dioxide but also hydrogen is absent (i.e. in an atmosphere of pure nitrogen), light causes a release of hydrogen from the adapted algae. The reversion by light in presence of carbon dioxide can be prevented by the action of certain substances like hydroxylamine, or o-phenanthroline, In the presence of such inhibitors the adapted algae continue to metabolize like purple bacteria even in very strong light. The light saturation rate of carbon dioxide reduction with hydrogen remains, however, far smaller than the corresponding rate of photosynthesis.Very small amounts of oxygen prevent adaptation, but up to 10mm. Hg of this gas are tolerated by adapted algae in hydrogen. The reason is that at these low partial pressures the reduction of oxygen to water in the algae proceeds faster than the reversion. Under optimum conditions the formation of water from the elements, the oxyhydrogen reaction, is coupled with a simultaneous reduction of carbon dioxide, so that one-half molecule of carbon dioxide disappears for each molecule of reduced oxygen. The coupling between oxyhydrogen reaction and the reaction of carbon dioxide is such that in absence of carbon dioxide the first reaction remains incomplete; little more than one molecule of hydrogen is absorbed instead of the two necessary to form water. After a partial reduction of the oxygen leading to the ‘peroxide’ level, the reaction apparently continues in unknown directions with internal hydrogen donors. Again a different result is obtained after adding specific poisons. In presence of carbon dioxide and of poisons, the amount of hydrogen absorbed indicates the straightforward formation of water.A critical comparison of the results reported in this article leads to the assumption that the dark reduction of carbon dioxide can be represented very satisfactorily by reversing essential parts of a diagram drawn to describe various partial reactions in photosynthesis and photoreduction.My thanks are due to Drs J. Franck and D. Goddard, for helpful suggestions. I am particularly indebted to Dr E. Rabinowitch for a very thorough criticism of the manuscript.