Data on time response in the nervous system are available for animals. Rate-matching data for humam ,,ubject,, with implanted electrodes are available. These data do not tell us with certainty up to what rate people with normal hearing utilize time information directly. In many experiments an attempt to measure time re,,ponse is confused by the presence of spectral clues. Short-term spectral clues exist in stochastic •,½qu½.cc•, of positive and negative pulses and in modulated white noise, both of which have a white longterm •,p½ctrum. These clues are periodicity (with a period equal to the pulse or modulation rate) and •,ymmctry (about frequencies equal to integer multiples of half this rate), differences at different frequencies in' the peak-or rrns-voltage distribution function of the output of a narrow-band filter and, in the case of ,,rochaslit pulse sequences, differences in extreme intensity at different frequencies. It may be that the ability to match at high rates and the perception of tonal quality that subjects with excellent and trained hcari.g, such a•, musicians, exhibit for white-spectrum stochastic pulse sequences or gated noise is due to the re•,idue phenomenon or to spectral clues discussed in this paper. I'ACS numbers' 43.66.Mk, 43.66.Ba, 43.66.Lj INTRODUCTION In animal hearing direct measurements can be made of the firing of neurons in response to sine waves or pulse trains of various frequencies or rates. Histograms of firings show that statistical synchrony of firings with stimuli is preserved in the eighth nerve up to stimulus frequencies of about 4000-5000 Hz (Rupert, Moushegian, and Galambos, 1963; Peake, Kiang, and Goldstein, 1962; Rose etal., 1967). However, it is not clear up to what repetition rate animals make any use of time information. 1 Neurophysiological data on synchrony in firing are not available for human beings. However, electrodes have been implanted in totally deaf patients who exhibit function of the eighth nerve. These give a partial restoration of hearing through electrical stimulation. In such subjects, all neurons are stimulated by the same signal and any ability to distinguish among stimuli must be based on time information.Tests on a patient with six electrodes permanently implanted in the modiolus portion of the eighth nerve by F. Blair Simmons in 1964 showed an ability to discriminate pulse rates up to 300 pulses per second and an ability to discriminate two short pulses from a single pulse when the interval was increased and when the pulses were separated by more than 5 msec. (Simmons etal. , 1965.) Michelson and his associates have reported on the responses of four patients implanted with intracochlear stimulating electrodes. Prior to implant, the hearing of the ear later implanted had deteriorated to total deafness in three of the patients; in the other, the implanted ear had been deaf from birth (Michelson, 1971). The apparent pitch reported by three subjects rose steadily with increasing frequency to about 500 Hz. At higher frequencies the "tones" heard ...