convergence, which causes double images until it is binocularly fixed, or whether by the aid of prisms we bring the retinal images of a distant object to parts of the retinæ which do not correspond. If we look at a distant object fixed with parallel visual axes, under normal circumstances, prisms of 6° to 8° with the base inwards can be overcome, that is to say, as in weak prisms the deviation is equal to about half of the prism, an absolute divergence of the visual axes of 3° to 4° may be produced by which the double images are blended. It is immaterial whether we apply a prism of, say 8° to one eye, or prisms of 4° with the bases inwards to both. The facultative divergence thus attainable remains the same, which speaks for the fact, that this monolateral movement attainable by prisms is also combined with double innervation; and of course in the eye remaining in unmoved fixation, with impulses to innervation which are reciprocally abolished. In the physiological use of our eyes we certainly never have occasion to practise absolute divergence, but we constantly practise the transition from the inward to the outward movement of the eyes, and experiments with prisms teach us, that the innervation of the externi therewith connected may even be carried somewhat beyond the physiological limits of parallelism. Moreover, the extent of the "facultative" divergence attainable by prisms shows a considerable latitude. The case is similar with deviation in height of the visual axes. In looking upwards or downwards the innervation of both eyes is usually precisely the same, but on looking at any point when holding the head obliquely, the difference in height of the eyes then present must be balanced by a corresponding difference in the direction of the visual axes. The same thing happens, if we hold a vertically deviating prism in front of one eye in binocular vision; prisms of 2° to 3° may then be overcome by difference in height of the eyes; rarely is a much greater difference in height of the visual axes attainable. I have seen this particularly in those cases where facultative divergence also was greater than usual.