Early development of visual cortical cells in normal and dark-reared kittens: relationship between orientation selectivity and ocular dominance.

1. 535 units were recorded in the primary visual cortex in twelve normally reared and fifteen dark-reared kittens aged between 8 and 50 days. These results were pooled with a previous study of 582 units recorded from thirty-five kittens reared in similar conditions. 2. These 1117 cells were classified into four functional classes of neurones: (a) visually unresponsive cells, (b) non-specific cells which were sensitive to spots or slits of light moving in any direction, (c) immature cells which were preferentially activated by a rectilinear stimulus but unselective regarding its precise orientation and (d) specific cells that appeared to be as selective for orientation as the simple or complex cells in the adult cat. 3. The results confirm that cells having the same orientation-specific response properties as adult cortical visual neurones are present in the earliest stages of post-natal development, independently of visual experience. However, to maintain and develop these specific cells after the third week of post-natal life, visual experience is necessary. 4. The ocular dominance of visual cells is not constant from the earliest stages of development. A significant increase in binocularly driven neurones occurs with age. 5. Before 3 weeks of age, whatever the rearing conditions, there are more specific cells coding horizontal and vertical orientations than those coding oblique orientations. These 'horizontal and vertical detectors' are preferentially driven by the contralateral eye. 5. After 4 weeks of age, specific neurones are found at all orientations in normally reared kittens. At this stage of development the ocular dominance is independent of orientation preference, of the functional class of neurones considered and of the rearing conditions. The proportion of binocularly driven cells is slightly below adult standard. 7. A hypothesis of differential plasticity is proposed: contralateral, monocular 'horizontal and vertical detectors' are supposed to be stable; they would remain so until they become binocular. Binocular cells, for which competition between two inputs occurs, are the labile units which can be despecified or specified under the control of visual experience.