"Sounds seem to rise and fall; that is a fact. It is difficult to explain it. Some have seen in it a habit derived from the usual notation by which the height of the note corresponds to its height in the score. But the impression is too deep and general to be explained by so superficial and recent a cause. It has been suggested also that high notes are generally produced by small and light bodies, low notes by heavy bodies. But that is not always true. It has been said, again, that high notes in nature are usually produced by highly placed objects, while low notes arise from caves and low placed regions. But the thunder is heard in the sky, and the murmur of a spring or the song of a cricket arise from the earth. In the human voice, again, it is said, the low notes seem to resound in the chest, high notes in the head. All this is unsatisfactory. We cannot explain by such coarse analogies an impression which is very precise, and more sensible (this fact has its importance) for an interval of half a tone than for an interval of an octave. It is probable that the true explanation is to be found in the still little understood connection between the elements of our nervous apparatus. "Nearly all our emotions tend to produce movement. But education renders us economical of our acts. Most of these movements are repressed, especially in the adult and civilized man, as harmful, dangerous, or merely useless. Some are not completed, others are reduced to a faint incitation which externally is scarcely perceptible. Enough remain to constitute all that is expressive in our gestures, physiognomy, and attitudes. Melodic intervals possess in a high degree this property of provoking impulses of movement, which, even when repressed, leave behind internal sensations and motor images. It would be possible to study these facts experimentally if we had at our disposition a human being who, while retaining his sensations and their motor reactions, was by special circumstances rendered entirely spontaneous like a sensitive automaton, whose movements were neither intentionally produced nor intentionally repressed. In this way, melodic intervals in a hypnotized subject might be very instructive."