We have used the neurofibrillar method of Gros-Schultze to stain the axonless horizontal cells of capybara, agouti, cat, and rabbit retinae. In all of these species, we have found two unusual horizontal cell morphologies: displaced horizontal cells and biplexiform horizontal cells. The displaced horizontal cells have perikarya located in the ganglion cell layer and dendrites branching in the inner plexiform layer. Many dendrites take an ascending trajectory to branch in the outer plexiform layer. The biplexiform horizontal cells are normally placed horizontal cells with descending processes that branch in the inner plexiform layer. Both cell types occur mainly in the retinal periphery, near the ora serrata. They are more numerous in the capybara retina, where they represent as much as 50% of the axonless horizontal cells of the retinal periphery.

The morphology and distribution of normally placed and displaced A horizontal cells were studied in the retina of a diurnal hystricomorph rodent, the agouti Dasyprocta aguti. Cells were labeled with anti-calbindin immunocytochemistry. Dendritic-field size reaches a minimum in the visual streak, of about 9000 μm2, and increases toward the retinal periphery both in the dorsal and ventral regions. There is a dorsoventral asymmetry, with dorsal cells being larger than ventral cells at equal distances from the streak. The peak value for cell density of 281 ± 28 cells/mm2 occurs in the center of the visual streak, decreasing toward the dorsal and ventral retinal periphery, paralleling the increase in dendritic-field size. Along the visual streak, the decline in cell density is less pronounced, remaining between 100–200 cells/mm2 in the temporal and nasal periphery. Displaced horizontal cells are rare and occur in the retinal periphery. They tend to be smaller than normally placed horizontal cells in the ventral region, whilst no systematic difference was observed between the two cell groups in the dorsal region. Mosaic regularity was studied using nearest-neighbor analysis and the Ripley function. When mosaic regularity was determined removing the displaced horizontal cells, there was a slight increase in the conformity ratio, but the bivariate Ripley function indicated some repulsive dependence between the two mosaics. Both results were near the level of significance. A similar analysis performed in the capybara retina, a closely related hystricomorph rodent bearing a higher density of displaced horizontal cells than found in the agouti, suggested spatial independence between the two mosaics, normally placed versus displaced horizontal cells.