of such changes, appropriate to the different styles of music, are further traits having the same derivation.  The slowest movements, largo and adagio, are used where such depressing emotions as grief, or such unexciting emotions as reverence, are to be portrayed; while the more rapid movements, andante, allegro, presto, represent successively increasing degrees of mental vivacity; and do this because they imply that muscular activity which flows from this mental vivacity.  Even the rhythm, which forms a remaining distinction between song and speech, may not improbably have a kindred cause.  Why the actions excited by strong feeling should tend to become rhythmical is not very obvious; but that they do so there are divers evidences.  There is the swaying of the body to and fro under pain or grief, of the leg under impatience or agitation.  Dancing, too, is a rhythmical action natural to elevated emotion.  That under excitement speech acquires a certain rhythm, we may occasionally perceive in the highest efforts of an orator.  In poetry, which is a form of speech used for the better expression of emotional ideas, we have this rhythmical tendency developed.  And when we bear in mind that dancing, poetry, and music are connate—­are originally constituent parts of the same thing, it becomes clear that the measured movement common to them all implies a rhythmical action of the whole system, the vocal apparatus included; and that so the rhythm of music is a more subtle and complex result of this relation between mental and muscular excitement.

But it is time to end this analysis, which possibly we have already carried too far.  It is not to be supposed that the more special peculiarities of musical expression are to be definitely explained.  Though probably they may all in some way conform to the principle that has been worked out, it is obviously impracticable to trace that principle in its more ramified applications.  Nor is it needful to our argument that it should be so traced.  The foregoing facts sufficiently prove that what we regard as the distinctive traits of song, are simply the traits of emotional speech intensified and systematised.  In respect of its general characteristics, we think it has been made clear that vocal music, and by consequence all music, is an idealisation of the natural language of passion.

* * * * *

As far as it goes, the scanty evidence furnished by history confirms this conclusion.  Note first the fact (not properly an historical one, but fitly grouped with such) that the dance-chants of savage tribes are very monotonous; and in virtue of their monotony are much more nearly allied to ordinary speech than are the songs of civilised races.  Joining with this the fact that there are still extant among boatmen and others in the East, ancient chants of a like monotonous character, we may infer that vocal music originally diverged from emotional speech in