1. The present experiments were conducted to study physiological mechanisms in the superior colliculus and their relation to visual spatial attention. We used a cued reaction time task studied in detail previously (Bowman et al. 1993; Posner 1980). Monkeys learned to fixate a spot of Light and release a bar when a target light appeared. Cues on the same side as the target(valid cue) were associated with faster reaction times than those on the opposite side (invalid cue). The difference in reaction times is hypothesized to be a measure of attention. 2. A total of 79 neurons within the superficial layers of the superior colliculi of two monkeys were studied. When the cues and targets were positioned so that both were within the visual receptive field, the cues excited the cells, and this produced a refractoriness to the targets for the following 400 ms. Both the ON and OFF responses to the cue were constant under all conditions. 3. These neurons were also tested with the cue just outside of the visual receptive field. This was done to avoid refractory effects from the cue; there was no significant modulation of the response to the target under these conditions. The visual responses of neurons in the intermediate layers of the superior colliculus also responded equivalently under these conditions. 4. When the activity of cells within the foveal representation was compared during the performance of three tasks, there differential activity. The appearance of the fixation point during the performance of the cued reaction time task led to a strong, transient discharge. There was an equivalent response when the monkey performed a foveal attention task; there was little or no; activation to the fixation point during a simple fixation task where no additional behavioral response was required. There was no change in activity of these cells at later times when the cue shifted the direction of attention. 5. Regions of the colliculus were modified with injections of muscimol, a gamma-aminobutyric acid agonist. The injection produced a slowing of reaction times to all targets in the affected representation. The reaction times to targets in the ipsilateral visual field were not changed. Comparable injections of saline had no effect 6. These data show that the visual cells within the superior colliculus were activated in the performance of attentional task that were independent of eye movements. The responses of cells to the onset of the cue could signal exogenously initiated shifts of attention. However, these cells appear to exhibit no endogenous attentional modulations. In contrast, cells in the foveal representation have attentional modulations and may signal engagement of attention at the fovea.
