Polliblemethods of sensory appreciation of the natureofsounds are briefly surveyed in relation to the restriction imposed by the rate at which the nervour system can respond. It is shown that because of this restriction information must in general be lost unless peripheral frequency analysis occurs and unless the peripheral analyzer conforms to certain further conditions which are enumerated. One of these conditions is that the selectivity of the resonant elements shall be proportional to frequency. Experimental evidence is submitted that the selectivity of the resonant elements even in the upper half of the auditory spectrum of the human ear is in fact roughly proportional to frequency and is very much higher than has generally been supposed. Previous theories of hearing are considered, and it is shown that only the resonance hypothesis of Helmholtz interpreted in accordance with the considerations enumerated in the first part of this paper is consistent with observation. In particular the experimental data which have been supposed to be evidence of high damping of the cochlear resonators are re-examined. It is shown that they are either irrelevant or that, correctly interpreted, they are evidence for the contrary view. Finally, an attempt is made to summarize some important properties of the ear in a diagram which illustrates that the ear is a perfect analyzer up to a frequency of 1 kc./sec. Above that frequency it is imperfect, not because of inadequate selectivity, but because perfection would require an impracticable number of resonant elements and nerve cells.