We might go further, and pass from the general to the particular. If music, by its effect on the organism, creates dispositions, momentary affective situations, the differences in voice, instrument, _timbre_, must produce different and special dispositions, which is indisputable. The tonality of a piece must act in the same way, which is also admitted by many composers. It is true that they are not agreed on the definition and the significance of every tone, and that many amusing discrepancies might be selected from their writings. (So the key of E flat minor, which, for Gevaert, is powerful and majestic, indicates, according to Grétry, an imminent catastrophe.) Here, more than elsewhere, over-precise definition is injurious. Let me add a remark on the therapeutic action of music. We have abundant evidence that this was known in the most ancient times. From the Greek physicians to Leuret, who employed it in his moral treatment of insanity, a long series of cures have been attributed to it. A well-known Russian physiologist, Tarchanoff, has recently lauded and recommended its rational employment in disorders of the nervous system; but it does not act through occult, mysterious, spiritual influences; it acts physically, as a kind of vibratory medicine. The researches of Boudet de Paris, Mortimer Granville, Buccola, Morselli, Vigouroux, furnish proofs of this. Although we might say much more on this subject, the above will be sufficient to show that the most emotional of the arts is also that most intimately dependent on the modifications of the organism. This has seemed to me an argument not to be neglected in favour of the physiological theory of emotion.[76] ----- Footnote 76: Gurney, in a criticism of James’s hypothesis (_Mind_, ix. 425), says: “There is plenty of music from which I have received as much emotion in silent representation” [_i.e._, by purely internal audition, or merely reading the notes] “as when presented by the finest orchestra; but it is with the latter condition that I almost exclusively associate the cutaneous tingling and hair-stirring.” Professor James has, in my opinion, answered this objection (_Psychology_, ii. pp. 469, 470), which I should be inclined to refer to the problem of the “revivability of impressions,” to be examined later on. ----- III.