During the past few years there has been a revival of the controversy as to the relative merits of oxygen and of oxygen-carbon-dioxide mixtures (93% 02+7% C02, or 95% 02+5% C02) in resuscitation from carbon monoxide poisoning (Marriott, 1955; Hill, 1956). This debate has stimulated a renewed interest in the effects of carbon monoxide on man, and has therefore prompted us, in the present paper, to describe and discuss a number of observations, made in 1944 and 1945, on the prolonged effects of carbon monoxide on vision. The data have not hitherto been published in any detail, though they were presented verbally in preliminary form in 1947(Halperin, Niven, McFarland & Roughton, 1947. The data, now to be reported, were obtained in the latter stages of a research project by McFarland, Roughton, Halperin & Niven (1944), the earlier results of which were published in full at the time.In this paper McFarland, Roughton, Halperin & Niven compared the effects on visual discrimination of the anoxia produced (i) by mild degrees of carbon monoxide poisoning (i.e. carbon monoxide anoxia) and (ii) by hypoxic anoxia caused by breathing gas mixtures containing subnormal partial pressures of oxygen. The apparatus used in their tests was the visual discriminometer of Crozier & Holway (1939), which proved especially suited for studying quantitatively the effects of anoxia on normal men (McFarland, Halperin & Niven, 1944). The same procedure was also used for the observations recorded in the present paper and is briefly described later. With this method McFarland, Halperin & Niven found that the 'immediate' effect of a given amount of carbon monoxide in the blood was about the same as that produced by an equivalent amount of oxygen unsaturation in the arterial blood, the effect of the carbon monoxide anoxia being roughly twice as great as in many other physiological tests. They did not, however, examine closely the time course of the effects of carbon monoxide anoxia on visual discrimination. But 584 M. H. HALPERIN AND OTHERS a study of their data from this point of view suggests some significant differences from the effect of hypoxic anoxia. In the present paper it is shown that the difference between the two types of anoxia becomes very manifest when the time course of recovery from each of them is compared. The effect of carbon monoxide is found, indeed, to persist for a remarkably long time, even when compared with its slow rate of disappearance from the blood, whereas the recovery from the visual effects of hypoxic anoxia takes place within a few minutes when 100% oxygen is breathed. Prolonged effects of carbon monoxide have often been observed clinically and specific histological lesions in the central nervous system have also been seen in post-mortem examinations after deaths from carbon monoxide poisoning. The experiments to be described in the present paper, on the effects of breathing oxygen at various pressures during the recovery period, not only support this concept qualitatively, but also give some direct quantitativ...