To vary the scene a little, I propose to close with a slight comparison betwixt singing and pronouncing. In this comparison the five following circumstances relative to articulate sound, must be kept in view. 1st, It is harsh or smooth. 2d, A sound or syllable, is long or short. 3d, It is pronounced high or low. 4th, It is pronounced loud or soft. And, lastly, a number of words in succession constituting a period or member of a period, are pronounced slow or quick. Of these five, the first depending on the component letters, and the second being ascertained by custom, admit not any variety in pronouncing. The three last are arbitrary, depending on the will of the person who pronounces; and it is chiefly in the artful management of these, that just pronunciation consists. With respect to the first circumstance, music has evidently the advantage; for all its notes are agreeable to the ear, which is not always the case of articulate sound. With respect to the second, long and short syllables variously combined, produce a great variety of feet; yet far inferior to the variety which is found in the multiplied combinations of musical notes. With respect to high and low notes, pronunciation is still more inferior to singing. For it is observed by Dionysius of Halicarnassus[96], that in pronouncing, _i.e._ without altering the aperture of the windpipe, the voice is confined within three notes and a half. Singing has a much greater compass. With respect to the two last circumstances, pronunciation equals singing. * * * * * In this discourse, I have mentioned none of the beauties of language, but what arise from words taken in their proper sense. Those beauties that depend on the metaphorical and figurative power of words, are reserved to be treated in chap. 20. SECT. IV. _VERSIFICATION._