\Original learning and postoperative retention of 12 rhesus monkeys, trained on a visual pattern discrimination before or after bilateral crosshatching 01' undercutting of their inferotemporal cortices, were compared with that of 4 normal Ss. Crosshatching produced no deficit in learning or retention even on a difficult 5-alternative discrimination, while undercutting led to a deficit in both. The results reopen the issue as to whether the various sectors of the posterior intrinsic association cortex contribute to discrimination per formance by acting on information received through transcortical fibers from the respective primary sensory areas, or whether these intrinsic areas act corticofugally to modify activity in the primary sensory pathways. It has been repeatedly demonstrated that partial ablations of posterior intrinsic as sociation cortex result in sensory-modal ity-specific deficits in the performance of discriminations. The particular sensory modality affected is dependent upon the locus of the lesion. For example, visual dis criminations are disturbed by inferotem poral lesions, auditory discriminations by midtemporal lesions (Pribram, 1954), and so on. One conception of the function of these areas is that they each receive input from one primary sensory area and that they per form some operation on it which is crucial for discrimination. Pribram (1958) has sug gested an alternative to this view, namely, that the various posterior intrinsic associa tion areas act via corticofugal, efferent fibers which alter the activity of the pri mary sensory pathways. The inferotemporal cortex in this case would selectively affect the primary visual system. Evidence for this position comes from the finding that circumsection of the primary visual striate cortex or interruption of thalamo-cortical input to the inferotemporal area does not generally disrupt visual discrimination per formance (Chow, 1952, 1954). In addition, Spinelli and Pribram (1966) have recently shown that stimulation of the inferotem