It becomes natural when the words are not an integral part of the work, but only serve to announce the idea on which it is based, and which the music brings out under successive aspects. The same may be said of the use of a melody with many different sets of words. Greek writers regard even the repetition of the melody in a strophe and antistrophe as a concession to the comparative weakness of a chorus. With the Greeks, moreover, the union in one artist of the functions of poet and musician must have tended to a more exquisite adaptation of language and music than can be expected when the work of art is the product of divided labour. In Greece the principle of the interdependence of language, metre, and musical sound was carried very far. The different recognised styles had each certain metrical forms and certain musical scales or keys appropriated to them, in some cases also a certain dialect and vocabulary. These various elements were usually summed up in an ethnical type, one of those which played so large a part in their political history. Such a term as Dorian was not applied to a particular scale at random, but because that scale was distinctive of Dorian music: and Dorian music, again, was one aspect of Dorian temper and institutions, Dorian literature and thought. Whether the Greeks were acquainted with harmony--in the modern sense of the word--is a question that has been much discussed, and may now be regarded as settled[1]. It is clear that the Greeks were acquainted with the phenomena on which harmony depends, viz. the effect produced by sounding certain notes together. It appears also that they made some use of harmony,--and of dissonant as well as consonant intervals,--in instrumental accompaniment ([Greek: krousis]). On the other hand it was unknown in their vocal music, except in the form of bass and treble voices singing the same melody. In the instrumental accompaniment it was only an occasional ornament, not a necessary or regular part of the music. Plato speaks of it in the _Laws_ as something which those who learn music as a branch of liberal education should not attempt[1]. The silence of the technical writers, both as to the use of harmony and as to the tonality of the Greek scale, points in the same direction. Evidently there was no _system_ of harmony,--no notion of the effect of _successive_ harmonies, or of two distinct _parts_ or progressions of notes harmonising with each other. [Footnote 1: On this point I may refer to the somewhat fuller treatment in Smith's _Dictionary of Antiquities_, art. MUSICA (Vol. II, p. 199, ed. 1890-91).]