ceivable that a purely psychical cause should directly set in action all the strings concerned. But in the case of mental hearing—which alone we are considering now—the localization of the function appears certain, since a definite injury of the temporal lobe abolishes it; and, on the other hand, we have set forth the reasons which make it impossible for us to admit, or even to conceive, traces of images deposited in any region of the cerebral substance. Hence only one plausible hypothesis remains, namely, that this region occupies with regard to the centre of hearing itself the place that is exactly symmetrical with the organ of sense. It is, in this case, a mental ear.
But then the contradiction we have spoken of disappears. We see, on the one hand, that the auditory image called back by memory must set in motion the same nervous elements as the first perception, and that recollection must thus change gradually into perception. And we see also, on the other hand, that the faculty of recalling to memory complex sounds, such as words, may concern other parts of the nervous substance than does the faculty of perceiving them. This is why in psychic deafness real hearing survives mental hearing. The strings are still there, and to the influence of external sounds they vibrate still; it is the internal keyboard which is lacking.
In other terms, the centres in which the elementary sensations seem to originate may be actu-
