Retinotopic order is surprisingly good within cell columns in the cat's lateral suprasylvian cortex

The retinotopic map in the striate-recipient region of the cat's lateral suprasylvian cortex (referred to here as the lateral suprasylvian area (LS)) has generally been described as quite disorderly. The disorder is commonly attributed to receptive field scatter within cell columns, reflecting the very large size of receptive fields. However, scatter within columns has never been investigated. In the experiments reported here, we examined the receptive field scatter of cells in columns, and also the scatter of a limited sample of their afferents arising from areas 17 and 18. To measure post-synaptic receptive field scatter, electrode penetrations were made parallel to columns in LS, with the electrode approaching from the medial side, traversing the suprasylvian gyrus and emerging into the suprasylvian sulcus. In all 13 such penetrations, receptive fields were clustered together despite their large size. Their centers were scattered over a region that occupied on average less than 20% of the largest field in the column. In contrast, in columns in areas 17 and 18 receptive field centers reportedly are dispersed over regions about equal to the largest of the fields (Hubel and Wiesel 1962, 1965, 1974). The scatter of afferents' receptive fields was assessed anatomically by measuring the overlap between patches of different anterograde tracers in LS. These patches represented terminal labeling from two adjacent or overlapping tracer injections in area 17. While a large degree of overlap would be predicted if afferents have substantial scatter, we found the overlap to be small unless the two injection sites themselves were highly overlapping. Scatter in afferents' receptive fields was measured more directly by physiological recording. In previous experiments, cells in LS were silenced by the local injection of kainic acid, and responses were recorded from axon terminals arising from areas 17 and 18 (Sherk 1989). We examined the receptive field scatter in three penetrations made approximately normal to the cortical surface. Scatter was modest, much less than predicted by the size of post-synaptic receptive fields. Because the degree of receptive field scatter for postsynaptic cells in LS was similar to that of inputs from areas 17 and 18, the scatter of these inputs might be entirely responsible for that seen postsynaptically. Postsynaptic receptive field scatter, on the other hand, was too small to explain the reported disorder in the map in LS.