The organization of corticocortical projections from area 17 to area 18 of the cat's visual cortex.

Retrogradely transported tracers were injected into area 18 of the visual cortex of the adult cat to study the organization of corticocortical projections from area 17 to area 18. All injections, whether very small or relatively large, and irrespective of their exact location in area 18, produced a discontinuous, clustered distribution of labelled cells, mainly in layers II, III and upper IV, in a topographically related region of area 17. The mean centre-centre distance between neighbouring patches was about 750 microns. We conclude that the overall population of cells projecting to area 18 is genuinely distributed in a patchy fashion and that they provide an efficient spatial sample of information from area 17. Comparison of the dimensions of each injection site and of the retrogradely labelled territory suggested that each region in area 18 receives a convergent input from a zone in area 17 whose visual field representation is about 0.8 M-1 deg larger in all directions (where M is the magnification factor in millimetres per degree at the termination site in area 18). Pairs of injection were made in area 18 by placing small volumes of two fluorescent tracers, fast blue and diamidino yellow, side-by-side in either a rostrocaudal or a mediolateral plane, with different distances between them. When the boundaries of the dense central cores of two injection sites were separated, at their closest points, by about 1.6 mm, the two corresponding distributions of labelled cells in area 17 were just non-overlapping, suggesting that each group of cells in area 17 sends a divergent projection in innervate a zone about 0.8 mm larger in all directions in area 18. More closely spaced injections led to overlap of the distributions of labelling by the two dyes, with shared clusters containing a mixture of labelled cells. The proportion of double-labelled cells in these shared clusters never exceeded 4.4% (but was 70% after sequential injection of the two dyes at a single point). We conclude that, although each cluster of cells sends a divergent projection to area 18, the majority of individual axons terminate more discretely, perhaps providing specific inter-connections between functionally corresponding 'columns' in the two areas.