they were not fimple founds but fome way compounded, we cannot pronounce.

Χ

In fo great a variety of founds, there must be some very harsh, such as the aspirated confonants and 0; but they are fo mixed with others more sweet and pleafant, that the found is neither too foft and effeminate, nor too rough and auftere, but an admirable compofition of both. In fome words, they join the two rough founds, I mentioned, together, as in the word αχθομαι and χριμφθεις : In which laft the φ and are joined together, and the confonant μ prefixed; which I think does very well by way of variety. And, for the fame reason, they fometimes do as the barbarous languages do very frequently, join vowels together, not as diphthongs, but in different fyllables, and not only different vowels, but the fame vowel, as in that famous line of Homer, which, it is faid, deterred Plato from writing verfes,

Πιονες βοοωσιν ερευγομένης άλος έξω.

And there is nothing more common in Homer, and nothing more beautiful in point of

found than the oo. And, in general, in Homer, and in all the Ionic writers, there is a great deal of gaping of vowels upon one another, both in the fame and in different words. This, I think, if there be not too much of it, fwells the found of the language, and, I must own, pleases me, though it offended the delicate ears of later times: And, particularly Ifocrates has, with what I would call a fophiftical nicety, moft carefully avoided it. In Homer, too, there are like endings, both of verfes and of hemiftics, which I think a beauty alfo, (and fo they are reckoned by the antient critics), if they be not too frequent, which they are not in Homer; for he only uses them when he has a mind to adorn his diction, as in his fimilies, which are the most ornamented part of his poem; nor do I remember that he ever uses them in his narrative or speeches *.

* Homer has followed Ariftotle's rules, as in other things, fo in ftile, which he says ought only to be laboured, and much ornamented, ε Tois agyois MeÇEσIN, that is, in such parts of the poem where there is neither reasoning, character, or sentiments, to be expresfed; and he might have added, where there is no nar

The words in Greek are neither too long, like the words in the barbarous languages, nor too fhort, like the words of fome modern languages, by which the flow of the language is much interrupted, (there being neceffarily, as I have obferved, a ftop more or lefs betwixt the words, fo that the fpeech must be full of breaks), but of a moderate length, with the variety of fome longer and fome shorter. And it is to be observed, that the monofyllables, or very short words, are almost all words that occur very frequently, fuch as prepofitions, conjunctions, and the article; thefe, if they were long words, occurring so of ten, would make the difcourfe cumberfome and tedious.

The terminations of the words in Greek are as various as poffible confiftently with the pleasure of the ear, being very different, not only in different words, but in the fame word, by the variation of genders,

rative; for, as by narrative the business of the poem is carried on, it cannot be faid to be agyos, that is, a part where the action ftands ftill, as it does in the fimilies. See Arift. Poetic. Cap. 24. in fine.

numbers, cases, and tenses *. Many words they conclude with a diphthong, fuch as al, oi, & qu; which makes the pronunciation of fuch words go off with a found that both pleases and fills the ear, the termination being, as I have obferved, the moft ftriking part of the found of a word. But they end no words with a mute confonant, fuch as ß,, ♪, which make a harsh and abrupt conclufion †: Much lefs do they conclude with an afpirated confonant, fuch as 8, with which fo many words in English conclude, but which we should think infufferably harsh, and should say with the French, that it fleaed our ears, if we were not so much accustomed to it •

* Antient Metaph. Vol. iii. p. 220. where I have shown, that, from the fame Greek verb, there may be formed tenfes and participles of different terminations to the number of two thousand.

This is obferved by Aristotle, in his Poetics, Cap.21. where he observes also that they terminated no noun with a short vowel; the reason of which seems to have been, that the voice could not reft upon a short vowel, as on a proper bafis, and therefore the word could not be concluded in fuch a way as to please and fill the ear,

The Greeks have but one little word ending in the mutex, viz. ; but, when a vowel follows, they

That the Greeks might have all the variety poffible in the found of their language, they obferved that certain fyllables were enunciated with a breath much thicker, and more condenfed, as it were, than others; and hence the diftinction of the two spirits, which, as I have observed, we must not confound with the diftinction of loud and low, in the fyllables of our words; a distinction which I am perfuaded was unknown both to the Greeks and Latins, who pronounced all the fyllables of the words upon a level, as the French pronounce their corrupt dialect of the Latin. And my reafon for thinking fo is, that, if there had been accents, such as are in our language and fome other modern

make it end in , ufing inftead of ; And, as the word is fo ufed in Latin, I am perfuaded it was originally only used in Greek in that way. But, afterwards, where a confonant followed, they threw out the ☛, for the fake of the better found, as they frequently did on other occafions.

* See what I have faid upon the fubject of Spirits, Vol. ii. p. 34. where I have fhown that the nice Greek ear perceived a third or middle found betwixt the tenuis and the afpirated.