[Skip to Navigation]
October 1941


Author Affiliations


From the Department of Neurology, Harvard Medical School, and the Neurological Unit, Boston City Hospital.

Arch NeurPsych. 1941;46(4):613-620. doi:10.1001/archneurpsyc.1941.02280220046003

The theoretic as well as the practical advantages of considering the electrical activity of the cortex in terms of a spectrum has been emphasized in earlier papers.1 The most important of these advantages is that a spectrum reduces to an orderly arrangement a tremendous volume of data and thereby makes it comprehensible. A continuous spectrum, however, is necessary unless one is willing to discard a large part of the data.

Even before a technic was available for making this transformation, attempts were made to represent the electroencephalogram as a distribution of energy over a range of frequencies. Travis and Knott2 plotted the mean amplitude of waves of a given duration against the frequency. Gibbs, Gibbs and Lennox3 expressed their results in terms of energy distribution on a scale of frequency, relying for evidence as to whether the energy distribution was predominantly fast or slow on wave counts

Add or change institution