No abstract

The integration of nervous activity in sensory systems often begins in the receptor organs themselves. Numerous electrophysiological studies of the vertebrate visual system, in particular, have shown that in the eye the activity of elements in one region of the retina may be affected by illumination of other regions. Both excitatory and inhibitory interactions have been demonstrated. Granit's extensive researches, especially, have shown the importance of these retinal processes in visual physiology. (For reviews, see Granit, 1947, sections I and II, and 1955, chapter 2.) Interaction in the vertebrate eye may be ascribed to the complex organization of the retina, which is indeed a "nervous center."In the lateral eye of the horseshoe crab, Limulus polyphemus (L.), the histological organization is much simpler than in the vertebrate retina. Nevertheless, the sensory elements in this eye do exert an influence upon one another (Hartline, 1949). The interaction is inhibitory: the frequency of the discharge of impulses in a single optic nerve fiber is decreased and may even be stopped by illuminating areas of the eye in the neighborhood of the sensory element from which the fiber arises. The occurrence of a purely inhibitory action in a relatively simple eye is of general interest; the role that may be played by inhibitory interaction in enhancing contrast gives it an importance to visual physiology. It is the purpose of this paper to report in detail our experiments on this inhibitory phenomenon in the eye of Limulus. MaterialThe lateral eye of Limulus is a coarsely faceted compound eye containing, on the average, some 800 ommatidia. In a medium sized adult (25 cm. broad) each eye forms an ellipsoidal bulge on the side of the carapace, about 12 mm. long by 6 mm

In the lateral eye of the horseshoe crab, Limulus, the visual receptor units exert an inhibitory influence mutually upon one another. The discharge of impulses in any one optic nerve fiber, generated in the sensory structure of the particular ommatidium from which that fiber arises, is determined principally by the intensity of the light stimulus to the ommstidium and the state of adaptation of this receptor unit. However, the ability of an ornrn&-tidium to discharge impulses is reduced by illumination of the ommatidia in neighboring regions of the eye: its threshold to light is raised, and the frequency of the discharge that it can maintain in response to steady suprathreshold illumination is decreased. This inhibitory action is exerted reciprocally between any two ommatidla in the eye that are separated by no more than a few millimeters. As a result of inhibitory interaction among neighboring receptors, patterns of optic nerve activity are generated which are not direct copies of the patterns of external stimulation, but are modified by this integrative action that takes place in the eye itself.These basic features of inhibition in the eye of Limulus have been described in detail in a recent paper (Hartline, Wagner, and Ratiiff, 1956). In that paper it was shown that the anatomical basis for the inhibitory interaction is a plexus of nerve fibers lying just back of the layer of ommatidia, connecting them together. Furthermore, a direct experiment demonstrated mutual inhibitory action between two omnmtidia whose respective optic nerve fibers were placed on the recording electrodes together. It is the purpose of the present paper to analyze the inhibitory interaction of receptor units in the eye of Limulus, and to describe quantitative properties of receptor activity that arise as a consequence of this interaction. MethodThe experiments reported in the present study are based on the measurement of the frequency of the discharge of nerve impulses from two receptor units simultane-

Appreciation of the form of the retinal image depends upon a correspondence between the distribution of light on the retina and the distribution of activity among the fibers of the optic nerve. This correspondence may be studied directly by recording the activity in single optic nerve fibers in response to illuminating various parts of the retina. A given optic nerve fiber responds to light only if a particular region of the retina receives illumination. This region is termed the receptive field of that fiber. In a previous paper describing the responses in single optic nerve fibers from the cold-blooded vertebrate eye (Hartline, 1938) it was noted that the receptive fields of the optic nerve fibers are of small but appreciable extent, and that their locations on the retina are fixed. It is the purpose of the present paper to describe further the characteristics of receptive fields, and to discuss some of the spatial factors involved in the excitation of the fibers of the optic nerve. METHOD. The method for recording the activity in single optic nerve fibers from the eyes of cold-blooded vertebrates has been described in the previous paper (lot. cit.). An eye is excised, cut open, and small bundles of optic nerve fibers are dissected from the anterior surface of the exposed retina. The action potentials in these bundles are amplified and recorded with an oscillograph. When such a bundle has been split successfully, until only a single active fiber remains, the retina must be searched with a small spot of light to determine the region supplying that fiber. This search is aided by noting the direction, on the retina, from which the nerve fibers in the small bundle come, and by using large spots of light at first to locate the approximate position of the sensitive region. The optical system employed in these experiments has likewise been described. A spot of light of suitable size is projected upon the exposed retina; the coordinates of its position, referred to an arbitrary point of origin on the retina, are obtained from readings of crossed micrometers which control its location. The micrometer readings are reduced to millimeters on the retina by multiplying them by the magnification of the optical system (0.32 or 0.15). Sharpness of focus of the spot on the retina 690