Article révisé par les pairs
Résumé : In 12 successful experiments on normal human adults, random sequences of equiprobable acoustic clicks or electrical stimuli to the index finger were delivered at regular intervals of 2.5 sec. In alternate runs either the clicks or the finger stimuli were designated as targets to be identified and mentally counted by the subject. The stimulus intensities were carefully adjusted near threshold to provide a feasible, but exacting, task. Scalp-recorded brain potentials were recorded with long time constant (8 sec) on FM tape, edited for removal of sections with artifacts, and averaged. Potentials associated with motor activities were excluded. Large P350 components were elicited by the stimuli designated as targets, but not by identical stimuli when they were non-targets. In many experiments the stimuli were preceded by a transient negative shift of up to 3 μV. The size of P350 to targets correlated with the size of this negative shift. After non-targets a slow transient positive shift could be observed, but no P350 component. The P350 and negative shift were maximal at the vertex and symmetrically distributed on the two sides. The primary components of ERPs to finger stimuli were not modified by the cognitive task. It is hypothesized that both the transient shifts and the P350 are produced by variations in the level of neuromodulation of the cortical circuits by the mesencephalic reticular formation (MRF). These different cortical events reflect corresponding variations in the MRF control by efferent pathways from the prefrontal granular cortex. © 1979.