The cat's dorsocaudal claustrum was studied in Golgi preparations, by electron microscopy, and by anterograde and retrograde tracer techniques. It receives a convergent retinotopic projection from several visual cortical ares, including areas 17, 18, 19, 21a and PMLS (posteromedial lateral suprasylvian area). The projection arises from spiny dendrite cells (pyramidal and fusiform) in the middle of cortical layer VI. As shown by a double label experiment, they form a separate population from those projecting to the lateral geniculate nucleus. There are also inputs from the lateral hypothalamus, from the nucleus centralis thalami, and probably from the locus coeruleus, but not from the sensory nuclei of the thalamus. Non-visual cortical areas do not project to the visual claustrum, but many of them are connected to other parts of the nucleus. For example, the splenial (cingulate) gyrus projects to a claustral zone just ventral to the visual area, and regions anterior to the visual area are connected with somatosensory and auditory cortex. The commonest cell type in the claustrum is a large spiny dendrite neuron whose axon leaves the nucleus after giving off local collaterals. Small spine-free cells, with beaded dendrites and a locally arborizing axon, are found also. Electron microscopy of the claustrum after ablation of the visual cortex showed degenerating type 1 axon terminals synapsing on spines and beaded dendrites, suggesting a direct cortical input to both cell types. The visual claustrum projects back to the visual cortex, to the same areas from which it receives an input. The return projection is predominantly ipsilateral, but there is, in addition, a small crossed projection. The claustrocortical axons terminate in all cortical layers but most heavily in layers IV and VI. The majority of the cells in the visual claustrum project to the cortex, and retinotopy is maintained throughout the entire corticoclaustral loop. No subcortical projections from the claustrum could be identified.

The initial aim of the experiments described here was to identify and quantify the cortical and thalamic connections of visual cortical areas located in the vicinity of the suprasylvian sulcus. Inputs to various sites in this region were studied by making small injections of wheat germ agglutinin (conjugated to horseradish peroxidase) at physiologically identified locations. Retrogradely labeled cells were counted in each identifiable area of cortex and in thalamic nuclei. Some injections yielded quantitatively similar distributions of labeled cells, and it is suggested that such evidence provides a useful way of dividing the cortex into areas. By this criterion, a single, relatively large, cortical area was identified that occupied most of the medial bank of the suprasylvian sulcus, all or most of its posterior bank, and a small segment of its lateral bank. It was referred to as the Clare-Bishop area. Because neighboring visual areas were found to lack input from area 17, while the Clare-Bishop area received a strong striate input, its boundaries were investigated by labeling afferents from area 17. Together with the results of retrograde tracer injections, these data suggested that the Clare-Bishop area cuts across several of the visual areas defined physiologically by Tusa et al. ('81). As a consequence, its retinotopic organization must be relatively complex, with duplications of some parts of the visual field. Three other visual areas were tentatively identified on the basis of their distinctive connections. One was situated on the lateral bank of the suprasylvian sulcus and appeared to border the Clare-Bishop area laterally. Another, referred to as area 21, lay adjacent to area 19, and, for part of its length, also appeared to bound the Clare-Bishop area. The third, corresponding approximately to Heath and Jones's ('71) posterior suprasylvian region, lay lateral and anterior to the Clare-Bishop area in the depths of the posterior suprasylvian sulcus.

1. The early visual experience of nine cats was restricted to viewing horizontal or vertical lines inside opaque goggles. 2. When the kittens were 3-4 mo old, extracellular recordings were made in the primary visual cortex. To obtain a representative sample of cortical cells, units were studied at regularly spaced intervals along the course of electrode penetrations traveling oblique to the cortical surface. An automated assessment of preferred orientation using a computer-driven optical display was employed, and during the recording session the experimenters did not know which orientation(s) each animal had viewed in early life. 3. In the cats that viewed horizontal lines with one eye and vertical lines with the other during rearing, two major findings of previous workers (14) were confirmed. First, a majority of units were not selective for orientation. Second, units with preferred orientations near vertical tended to be activated exclusively by the eye that had viewed vertical, and likewise for horizontal. 4. In cats that viewed lines of the same orientation with both eyes during rearing, a substantially smaller proportion of units were selective for orientation; the preferred orientations of these units also tended to match the orientation to which the cats had been exposed. 5. Portions of some electrode penetrations showed an orderly arrangement of cells according to preferred orientation similar to that seen in normal cats, but with regions over which only nonselective cells were found. Many penetrations appeared less orderly. 6. The results are consistent with a role for early visual experience in maintaining the responsiveness and innate selectivity of cortical neurons, although they cannot entirely rule out the possibility that experience may alter or determine the preferred orientation of some cells.

No abstract

1. Extracellular recordings were made from single units in the visual cortices of six kittens deprived of experience with pattern vision by binocular lid suture. 2. Selectivity for stimulus orientation was quantitatively assessed in 98 units; 90 responded selectively to the orientation of a moving bar stimulus, the remainder responding nonselectively or too poorly to classify. Cells in these visually inexperienced kittens were similar in their degree of selectivity for orientation to cells tested in adult cats. However, responses tended to be weaker and somewhat more erratic. 3. About half the cells in this simple responded to both directions of stimulus motion at the optimal orientation. Most of those responding to only one direction of motion were considered orientation rather than direction selective because they responded more strongly or more selectively to a moving bar than to a moving spot. 4. Cells appeared to be organized within the cortex in a pattern similar to that found in adult cats, with cells in one column selective for the same orientation, and adjacent column having similar preferred orientations. 5. It is concluded that selectivity for stimulus orientation in the cat's visual cortex is innately determined.