RECENT studies of Kelly et al1 and Celesia et al2,3 have shown that it is possible to identify human somatosensory and auditory cortical areas by recording average evoked potentials obtained by monopolar recordings. However, these studies also pointed out the difficulties in recognizing between responses generated under the recording electrode and responses recorded by volume conduction from distant generators. Kelly et al1 emphasized the need to demonstrate polarity reversal to assign significance to monopolar evoked responses. Nevertheless, sensory average evoked potentials obtained directly from their source may prove to be sufficiently specific in their morphology that a set of criteria could be established for their identification and separation from volume conducted events. To test this possibility, a systematic investigation of auditory average evoked responses has been carried out in cats.
The same technique of computer averaging has been extensively applied in recent years to the study