It is widely believed that the location of the cochlear excitation maximum, which has been shown by Békésy to depend on sound frequency and move from the cochlear apex to its base as the frequency increases, is a code for subjective pitch. The pitch of a tone is known to be practically independent of sound intensity. If the location does determine the pitch, it too must remain invariant. At the 1990 meeting of the Collegium held in Basel, however, the first author reported compelling indirect evidence indicating that this may not be true. It suggested that, at least in the mid-portion of the cochlea, the most important for speech frequencies, the maximum moves toward the cochlear base as sound intensity is increased. We now have a direct verification of this inference. Recording alternating Hensen's cell potentials at two or three second-turn locations of each of several Mongolian gerbil cochleas, we observed that the maximum response produced by a single tone moved substantially toward the cochlear base as sound intensity increased. For example, an intensity increment of only 10 dB caused the maximum to move by about 0.225 mm. Since Hensen's cells are known to reflect closely the excitation pattern of the outer hair cells, similar to that of the inner hair cells, the discovery makes it impossible for the cochlear excitation maximum to be an adequate code for pitch. We observed, on the other hand, that the apical excitation cut-off did not depend on sound intensity. Every cochlear location investigated had its invariant characteristic cut-off frequency. It is possible, therefore, that the cut-off location provides the place code for pitch. These findings may have profound consequences for our understanding of auditory mechanisms as well as for the technology of cochlear implants.