authority which no scholar can dispute, namely that of Quinctilian himself, who says that Latin words terminate in a grave accent, and that invariably; but we learn from the same author in the same page, that the Greek accentuation was different. In comparing the two languages in respect to sweetness of modulation, after giving several instances of particular letters in which the Greek had the advantage, he proceeds to observe, that the Latin accents are less sweet, not only from a certain harshness, but also from their very monotony; their last syllable never having an acute nor a circumflex, but terminating invariably in a grave. For this reason he says, that the Greek language is so much more agreeable than the Latin, that the Latin poets, when they wish a verse to be sweet in sound, ornament it with Greek nouns: "Sed accentus quoque, cum rigore quodam, tum similitudine ipsa minus suaves habemus, quia ultima syllaba nec acuta unquam excitatur, nec flexa circumducitur, sed in gravem vel duas graves cadit semper. Itaque tanto est sermo Græcus Latino jucundior, ut nostri poetæ, quoties dulce carmen esse voluerunt, illorum id nominibus exornent." (xii. 10, 33.) It is difficult to conceive what authority can be set against this passage of Quinctilian, which affords the clearest demonstration that our accentuation of Greek is faulty, for this very reason, that it is the same as that of the Latin; and that it is faulty in this very particular, that

it always makes Greek words barytones. For though Quinctilian does not in so many words predicate that many Greek words are oxytones, that proposition is as clearly implied, in the whole passage taken together, as if it were expressly affirmed. Dr. Gally indeed ventures to assert, that Quinctilian is mistaken in this matter, and that there was not in truth any difference in respect of accents between the Latin and the Greek. Now however specious a modern scholar's reasoning on this subject might have appeared, I should have been very unwilling to trust it on such a subject against a critic and grammarian who constantly heard both Greek and Latin as living languages; and I should have been apt rather to suspect some fallacy in Dr. Gally, though I might not have been able to point out where it lay, than a gross blunder in Quinctilian. But when we come to examine Dr. Gally's reasons, we shall find them built upon two palpable mistakes he says, "This passage hath considerable difficulties. It would not be an easy matter to say what Quinctilian meant by a similitude of accents, if he had proceeded no farther. But he hath explained himself by saying, that the Greeks placed the acute and circumflex upon the last syllable, which the Latins never did, and that upon this account the Latin accents were not so sweet as the Greek. One cannot indeed refuse to Quinctilian the privilege of being his own interpreter. But then as the Latins

had the same number of accents with the Greeks, it cannot easily be conceived how a difference, arising from the mere placing of accents as to one syllable only, could cause a difference in the sweetness of them; and such a difference too as would in this respect give a considerable advantage and superiority to the Greek language; unless it can be proved that the placing of accents on final syllables is more harmonious than the placing them on penultimates and antepenultimates.

"But what is more material, if this point be accurately considered, no such difference between the Latin and Greek accents will be found as Quinctilian suggests. For the circumflex containeth an acute and a grave: therefore, when it is placed upon the last syllable of a Greek word, and resolved into its constituent parts, the pronunciation of this word will end in a grave. And though an accent be placed upon the last syllable of a Greek word, yet this is to take place only when the word is pronounced separately. For in discourse the final acute is always turned into, and pronounced as, a grave. Where then is the real difference, in this respect, between the Latin and Greek accentuation? What foundation does this afford to blame the Latin manner as less harmonious and diversified than the Greek?

"Quinctilian appears still more prejudiced in favour of the Greeks, by what he says at the

close of this passage. For what Latin poets have, in order to make their compositions more harmonious, made use of Greek words, merely because they were accented upon the last syllable?" (Second Dissertation against Greek Accents, p. 36.)

This is a fair specimen of the contradictions into which a correct and elegant scholar is forced by allowing the prejudices of his ear to control his judgement. The difficulties which Dr. Gally finds in the passage are all of his own creation. It is true that the Latins had the same number of accents with the Greeks, but the question here is not as to the number of the accents, but the application of them; and if in one language the accent be admitted in three places, while the other admits it only in two, it seems to be easy to conceive how this could cause a difference in the sweetness of them; and particularly when we learn from a person of taste who had heard both, that he found such a difference. Neither is it necessary to say, that the placing of accents on final syllables is more harmonious than the placing them on penultimates and antepenultimates. If the Greeks had made every word an oxytone, this would have been a monotony still more rigid, and doubtless more inharmonious than that of the Latins. No: the harmony of the Greeks consisted in this,—not that they placed the accent on final syllables, but that they did not exclude it from final syllables, and that by giving

it three places instead of two, they imparted a pleasing variety to the modulation of their language. But Dr. Gally not only doubts Quinctilian's taste, but discredits his testimony. He roundly asserts that no such difference between the Latin and Greek accents will be found as Quinctilian suggests. In the case of a circumflex, it seems that the Greek is the same as the Latin, because a circumflexed syllable ends in a grave. But Dr. Gally here has confounded a grave sound with a grave syllable. It is true that a circumflex contained two sounds, one of which, and probably the latter, was grave; but this grave was so blended (ovvep0apuévn) with the acute, as to produce a peculiar sound, which required a name of its own to describe it, and must have been perfectly distinct from the simple depressed sound of a grave syllable, or Quinctilian would not have said, as he has [p. 91], that the acute and the circumflex are the same. To express the difference at once by an instance: the last syllable of Oe is raised, but the last of Déo is depressed.

As to what Dr. Gally says of the final acute of an oxytone word being in discourse turned into, and pronounced as, a grave, this has already been shown to be a mere misconception, arising from the inclination of the mark. And this passage of Quinctilian, instead of being refuted by Dr. Gally's reasoning, seems to be a strong additional authority for raising all those