Abstract
The third visual area, V3, was initially described using degeneration studies indicating that the region was a mirror image of the retinotopic organization of V2. Though subsequent studies confirmed these retinotopic patterns through tracer connections and microelectrode maps, a major issue remained. Evidence for connections between dorsal V1, representing the upper visual quadrant, and dorsal V3 (V3d) were clearly demonstrated, whereas connections between ventral V1, representing the upper visual quadrant, and ventral V3 (V3v) were either not found, or were considered to be with ventral V2 instead. This apparent connectional asymmetry resulted in several different theories of cortex organization, including those without a ventral V3 (replaced by the ventroposterior area, VP) or a dorsal V3 (replaced by the dorsomedial area, DM). We addressed this issue by using more sensitive tracers to re-examine V1–V3v connections. Ventral and dorsal V1 injections in Old World macaques, 4 species of New World monkeys, and prosimian galagos unequivocally demonstrated connections with V3v and V3d. In addition, the connection patterns confirmed that V3d/v was a mirror image of V2 retinotopy. This V3 is much narrower than initial proposals, about half the width of V2, rather than of equal size. Furthermore, as assessed through extrastriate feedback to V1 (via anterograde tracer injections into V1), the V1–V3d/v connections were second to only V1–V2 connections. An additional finding was that cytochrome architecture revealed dark and light staining bands in V3d/V3v, indicative of modular organization. As V3 was found in all species examined, it likely is a feature common to all primates, including humans. This work was supported by National Eye Institute Grant EY02686.