The myth of visual “depth cues” II. “Occlusion”

It is standard textbook practice to refer to “occlusion” as one of the so-called a depth-cues. However, perceived occluding relationships are a consequence of the perceptual organization of the retinal point stimulation – which contains no occluded surfaces. The perception of “occlusion” always involves amodal completion of areas perceived as partially occluded. Shapes, occlusions, and relative depth relationships are all descriptions of the percept. To treat one aspect of this percept as prior to other aspects is a logical fallacy linked to a failure to distinguish between the percept and the real world, and to a preference for adopting simple, pseudo-explanations of perceptual phenomena instead of tackling the difficult problems entailed in explaining perceptual organization.

Morphology and topography of on- and off-alpha cells in the cat retina

Neurofibrillar staining methods were found to stain all alpha cells of the cat retina completely, that is the perikaryon, the axon and the dendritic branches. The dendrites of the alpha cells in vertical sections were found to be unistratified and to occupy two narrow strata in the outer half of the inner plexiform layer. This difference in branching level could also be observed in whole-mount preparations and it has been demon­strated in the preceding paper (Peichl & Wässle 1981) that it corre­sponds to the physiological on‒off dichotomy. Thus the topographical distribution of on- and off-alpha cells could be studied. They were found to occur in about equal numbers. Both on- and off-alpha cell perikarya form a regular lattice and both lattices are superimposed independently. The dendritic branches of neighbouring alpha cells overlap and each retinal point is covered by the dendritic field of at least one on- and one off-alpha cell. The dendritic trees of on-alpha cells seem to have more small branches and are on the average smaller than those of off-alpha cells. The density of alpha cells was found to peak in the central area whence it continuously decreased towards the retinal periphery.

Topography of horizontal cells in the retina of the domestic cat

Neurofibrillar methods stain a class of horizontal cells in the cat retina which are shown to be identical with the A-type horizontal cell of Golgistaining. Thus all of the A-type cells of a single retina can be observed. On this basis the changes in density and dendritic field size of A-type horizontal cells with respect to retinal eccentricity were measured. The decrease in density from centre to periphery is balanced by a corresponding increase in size of the dendritic field. Consequently each retinal pointindependent of retinal position — is covered by the dendritic fields of three or four A-type horizontal cells. The nuclei and nucleoli of B-type horizontal cells could also be recognized in neurofibrillar-stained material and thus their distribution was determined. The density ratio B-type: A-type is 2.8 + 0.4 and does not vary much from the centre to the periphery of the retina. Each retinal point is also covered by four B-type horizontal cells. Thus a single cone can contact a maximum of eight horizontal cells. The rate of density decrease from centre to periphery is closely similar in cones and horizontal cells but greater in ganglion cells.

Topographic organization and convergence in corticotectal projections from areas 17, 18, and 19 in the cat

1. Electrical stimulation in areas 17, 18 and 19 of the cat's visual cortex activated neurons in the superficial gray layer of the superior colliculus. The threshold for such excitation was lowest when the receptive fields of cells at the stimulus site lay inside the receptive field of the collicular cell under observation. 2. A substantial percentage of the recorded collicular cells received convergent excitatory input from more than one of the stimulated cortical areas. When stimuli could be applied in parts of areas 17, 18, and 19 related retinotopically to a collicular cell, over half of the units were activated from all three cortical areas by low-intensity stimulation. 3. Cells located near one another in the colliculus received excitatory cortical input from regions of areas 17, 18, and 19 which were not identical in size. This may be attributable to the disparate dimensions of collicular dendritic fields, which permit neighboring cells to sample dissimilar regions of the topographically organized cortical input. It is argued that this projection system preserves the functional unity of cortical and tectal cell groups processing information from a given retinal point.