The human brain consists of approximately one hundred thousand million cells, arranged in a variety of structures, the largest of which is the familiar neocortex. These cells, or neurons, possess the vital property of excitability, which is dependent upon the differential diffusion characteristics of their bounding membranes. The cells receive and transmit electrical impulses through their numerous tentacle-like extensions, and the signals are passed from one cell to another by the chemical messengers called neurotransmitters, which diffuse across the narrow inter-cell gaps known as synapses. The efficiency of the transmission process is chemically modifiable, and this is believed to imbue the neural network with the ability to learn and remember.
The response to a variety of input patterns has been studied in a vector model assembly of interconnected neurons. The time evolution of the injected signal is followed, attention being paid to both its subsequent topology and phase. The model is realistic in that it includes action potential impulses in the axon regions, statistically distributed synaptic delays, and electronics waves in the dendrites. Of particular interest were the frequency response of the system, and its dependence on the proportions of excitatory and inhibitory synapses. The relevance of the concept of coherence length was also critically examined, in such disparate contexts as association autism and the primary visual processes in the retina. Coherence, and the more general issue of correction, were also considered in connection with memory models, including those of the holographics type. This brief account also includes reference to such unlikely topics as fever and the dreaming state.