Meridional differences in orientation sensitivity in monkey striate cortex.

Recent physiological investigations1-3, 5 have suggested that in humans there is greater sensitivity and discrimination for horizontally and vertically oriented edges and gratings than for diagonal ones. Several investigations have looked for a neuro-physiological counterpart of this effect in striate cortex of the cat and monkey. Two hypotheses have been proposed: (1) Neurons with horizontal and vertical orientation selectivity are more common in visual cortex than are neurons selective for obliques. (2) All orientations are equally represented but cells responding to horizontal and vertical orientations are more selective for orientation. Recent work on the cat by Rose and Blakemore 7 suggests that simple cells with vertical and horizontal axes of orientation are more tightly tuned than those with diagonal axes. A more provocative finding reported recently by Mansfield 6 suggests that in monkey foveal striate cortex, cells sensitive to horizontally and vertically oriented edges are more numerous than cells for other orientations. This finding implies a columnar organization somewhat different from that proposed by Hubel and Wiesel 4 in that it would necessitate more space or columns allocated to meridional and fewer columns to diagonal orientation specificity. In view of these findings, as part of a more general effort s, we investigated a sample of 673 neurons in 226 penetrations perpendicular to the cortical surface and 11 penetrations at 10 ° or 20 ° to the cortical surface from the parafoveal striate cortex in 45 monkeys which were flaxedilized and artificially respired with 30~ Oz-70~ N20. To determine if any meridional variations in cell number or orientation speci-ficity existed, we assessed (1) cell number according to optimal orientation and (2) orientation tuning specificity for different axes. The orientation tuning function was derived from a smoothed curve of average response to presentation of a bar or edge stimulus swept across the receptive field at various orientations in a randomized sequence. The width of tuning in degrees at 71 ~ of maximum response was the measure employed. Fig. 1A shows a plot of the number of cells having optimal responses at various orientations. Only cells between 2 ° and 5 ° from the fovea, primarily representing the lower visual field, are included in this sample. In Fig. 1B the data were pooled adding