The axonal arborizations of lateral geniculate neurons in the striate cortex of the cat

Abstract Horseradish peroxidase (HRP) was injected into the optic radiations of adult cats. With placements close to the lateral geniculate nucleus (LGN), the enzyme diffused retrogradely along the axons of geniculocortical relay neurons, entered their cell bodies, and, after reaction with diaminobenzidine, produced a Golgi‐like staining of entire neurons. When the injections were made close to the visual cortex, the enzyme diffused anterogradely and filled complete axonal arborizations in area 17. In the LGN, examples of type 1 and type 2 relay neurons (Guillery, '66) were reconstructed, and their axon diameters measured. The type 1 neurons (thought to correspond to Y‐cells – LeVay and Ferster, '77) possessed large diameter axons (2–3.3 μm), while the type 2 neurons (thought to be X‐cells) had medium‐sized axons (1–1.7 μm). Both types of neurons gave off axon collaterals to the perigeniculate nucleus. In the cortex, two types of afferent supplied layer IV. One distributed to the upper part of the layer (layer IVab), extending a short distance into layer III. The parent trunks of these axons, measured in the white matter, had diameters matching those of type 1 LGN relay cells. The other type distributed to layer IVc. The diameters of these axons matched those of type 2 LGN relay cells. Most afferents of both types gave off collaterals to layer VI — there were no axons which exclusively innervated this layer. The axons supplying layer IVab had a wide lateral spread in the cortex (up to 2 mm), and the boutons were grouped into two to five clumps, whose size and arrangement were similar to ocular dominance columns. The axons supplying layer IVc had a much more restricted arborization, usually consisting of a single clump of boutons. LGN neurons with very fine axons (less than 1 μm) were found in laminae C1–C3. They probably corresponded to Guillery's type 4 neurons. In the cortex, fine‐diameter axons arborized in the upper half of layer I. These axons sometimes had collaterals in the lower part of layer III and in layer V. Taken together, the arborizations of the cortical afferents observed in the present study account fully for the autoradiographic labelling pattern seen after 3 H‐proline injections into the LGN (LeVay and Gilbert, '76). The identification of type 1 and type 2 neurons as Y‐ and X‐cells is strengthened by the observed difference in their axon diameters, in agreement with the different axonal conduction velocities reported for Y‐ and X‐cells. The presence of cells with very fine axons in the deeper C laminae is consistent with the reported presence of W‐cells (which have slowly conducting axons) in these layers. We conclude that the different classes of geniculate relay neuron have different laminar projections in area 17.