VISUAL RESPONSE LATENCIES IN STRIATE CORTEX OF THE MACAQUE MONKEY

1. Many lines of evidence suggest that signals relayed by the magnocellular and parvocellular subdivisions of the primate lateral geniculate nucleus (LGN) maintain their segregation in cortical processing. We have examined two response properties of units in the striate cortex of macaque monkeys, latency and transience, with the goal of assessing whether they might be used to infer specific geniculate contributions. Recordings were made from 298 isolated units and 1, 129 multiunit sites in the striate cortex in four monkeys. Excitotoxin lesions that selectively affected one or the other LGN subdivision were made in three animals to demonstrate directly the magnocellular and parvocellular contributions. An additional 435 single units and 551 multiunit sites were recorded after the ablations. 2. Most units in striate cortex had visual response latencies in the range of 30-50 ms under the stimulus conditions used. The earliest neuronal responses in striate cortex differed appreciably between individuals. The shortest latency recorded in the four animals ranged from 20 to 31 ms. Comparable values were obtained from both single unit and multiunit sites. After lesions were made in the magnocellular subdivision of the LGN in two animals, the shortest response latencies were 7 and 10 ms later than before the ablations. A larger lesion in the parvocellular subdivision of another animal produced no such shift. Thus it appears that the first 7-10 ms of cortical activation can be attributed to activation relayed by the magnocellular layers of the LGN. 3. The units with the shortest latencies were all found in layers 4C or 6 and their responses were among the most transient in striate cortex. Furthermore, their responses all showed a pronounced periodicity at a frequency of 50-100 Hz. This periodicity was stimulus locked, and the responses of all short-latency units oscillated in phase. 4. An index of response transience was computed for the units recorded in striate cortex. The distribution of this index was unimodal and gave no suggestion of distinct contributions from the geniculate subdivisions. Magnocellular and the parvocellular lesions affected the overall transience of responses in striate cortex. The changes, however, were very small; extremely transient responses and extremely sustained responses survived both types of lesions. 5. A characteristic profile was observed in the response latencies in superficial layers. Latencies appeared to increase monotonically from layer 4 toward the surface of cortex, with the most superficial neurons not becoming active until 15 ms after responses were observed in layer 4C. This large delay is likely to reflect the anatomic organization of the superficial layers and could have important implications for the information processing that occurs in striate cortex.