Abstract
1. Single-unit and multiunit activities were recorded at the area 17-18 border of each cortical hemisphere in paralyzed cats anesthetized with nitrous oxide supplemented with halothane. Cross-correlation histograms (CCHs) were computed between 86 pairs of single units and 99 pairs of multiunit activities. Visually evoked peaks in the CCHs were removed by subtracting the shift predictor. 2. Three types of CCH peaks were observed: T peaks with narrow widths (4-28 ms), C peaks with intermediate widths (30-100 ms), and H peaks with large widths (100-1,000 ms). Oscillatory coupling was observed rarely. This tripartite distribution of CCH peaks is similar to that reported in an earlier study on the temporal coupling between areas 17 and 18. Different types of peaks occurred in isolation or in combination. Combination of different peak types was more often observed in multiunit recordings. 3. CCH peaks of all types were usually centered, meaning that units in opposite hemispheres tend to synchronize their discharges. 4. T peaks were observed almost exclusively for units with overlapping receptive fields and preferentially for units with similar optimal orientations. No dependence on receptive field position or optimal orientation was observed for the encounter rate of C and H peaks. 5. A new method, called the peristimulus CCH, was developed to study the time course of the temporal coupling. This showed that H peaks can occur during visual stimulation and that their time course follows that of the visual responses of the coupled neurons. 6. Using one single bar or two simultaneously presented light bars as stimuli, we studied the effect of visual stimulation on the strength of H coupling. This showed that H coupling observed under stimulation with a single moving light bar can be completely abolished, with little change in visual responses, when the stimulus is changed to two noncoherently moving bars. This was related to a strong decrease of the H peaks in the autocorrelograms. 7. These results demonstrate that T, C, and H peaks constitute, together with high-frequency oscillations, universal forms of temporal coupling between neurons located in different cortical areas. The following paper reports on the effects of cortical lesions on the encounter rate and strength of these different types of coupling.