Abstract
During the summer of 1940 electroencephalograms were recorded, as part of a program of physiologic and psychologic testing, on approximately 450 aircraft pilots.1One objective was to determine the incidence (if any) of cerebral dysrhythmias. It is well known that voluntary hyperventilation often precipitates both petit mal attacks and their accompanying electrical patterns (Gibbs, Lennox and Gibbs2). Hyperventilation is employed as an adjunct to diagnostic procedure in many hospital electroencephalographic laboratories. We therefore included a three minute period of maximal voluntary hyperventilation as part of our routine test. We immediately confirmed the already familiar observation that many presumably normal persons show more or less prominent delta waves, or at least waves in the 4 to 8 cycle band, as a result of such hyperventilation. It seemed possible that the amount of such slow wave activity might serve as a useful index of "cerebral instability" and have some predictive