This principle is, in the Greek language, limited by a restric tive law; viz. that only one of the last three syllables of any word is capable of accent-a law which, in that rigorousness with which it governs the Greek accent, cannot be accounted for (as Mr. G. has attempted) either by the nature of language in general, or by that of primitive languages. (We beg to remind the reader of the accent in the English and German.) We think there is no better mode of explaining it, than by tracing this law to that delicate sense of euphony, peculiar to the Greek mind; for, the accent being on the fourth syllable from the end, renders indeed the pronunciation heavy and clumsy.

In consequence of this restrictive law, the above-mentioned fundamental principle is modified thus: the tone is placed ou the radical syllable, or that which (by determination in compositions) specifies or alters the primitive notion of the word, in so far as the number of the final syllables allows to do so; hence μειλίχιος, and not μείλιχιος.

On this principle, limited only by the quantity of the last syllable, (which however cannot be called an alteration of it, as will soon be seen), the accentuation of the eldest Greek language seems to have been conducted throughout; and the Æolic dialect has preserved this standard. (Etymol. Max. p. 312. 45. Gregor. Corinth. &c.) Mr. G. therefore justly considers the Æolic dialect as the only guide for any investigation respecting Greek accentuation.

Mr. G. now states three causes which have modified this original accentuation: the dialectical developement of the language, the endeavor to distinguish words of like sound, and quantity. The last circumstance (quantity), however, cannot be said to have occasioned a real breach of the stated principle; the deviation is only apparent, and therefore it has exerted its influence also on the Æolic dialect. We turn first to it.

I. Quantity. As to quantity, every thing depends, says Mr. G., on the ultimate. Is it long by nature, the long vowel is reckoned as two short ones (two mora), and the accent cannot take its place on the ante-penultimate, because otherwise, it would be, against the principal rule, on the fourth syllable from the end; v. c. sυGUÓNS EUPUÓTTEES. This is certainly correct; and the method in which Mr. G. has cleared up the seeming exceptions to the rule, is no less deserving of our approbation. But we ask, what is the reason, that in proparoxytonis, (v, c. ǎvOpanos) the long vowel in the penultimate, if the ultimate is short, has not the same value of two short ones (two mora) for accentuation, as it unquestionably has in the metre? Why is it

allowable to accent avoperos, and why is not avogaros equal for accentuation to ἀνθροπος, so that it should be accented ἀνθρωπος ? Mr. G. says: "The quantity of the last syllable of a word is of importance for the accentuation, but never the quantity of the penultimate. We must account for that again by the musical, method which the language took in its developement; for in song the last syllable is of the most distinct and undisturbed. expression." This reason, we must confess, appears to us destitute of any weight; for in song (in the metre), all long syllables were alike equal to two mora-two short ones. In the next place it is false, that the quantity of the penultimate never operated on the situation of the accent: it did, strongly enough, if the ultimate was long, v. c. prop; for in this case, the circumflex can never be placed on the penultimate. For the circumflex (as Mr. G. justly observes) is by no means to be considered as a particular sort of accent, but always as the acute (the accent) on the FIRST of two mora, (two short vowels) which were contracted into one length (one long vowel, equal to two short ones); the second of which had consequently the gravis, i. e. no accent at all: pua is equal to, and contracted from, péèua (hence the figure of the circumflex). Now, as the penultimate, if the ultimate be long, never has the circumflex, but the acute (if it has the tone at all), it was counted for two mora, of which the latter had the acute; prwp is equal to peéroop, and the accent stands on the third syllable from the end. ῥήτωρ would be equal to ῥέτοορ, and the accent would fall on the fourth syllable from the end. Nay, the long penultimate also (if it has the tone at all), the ultimate being short, had for accentuation the power of two more, because then it has constantly the sign, i. e. the accent was on the first mora, because nothing prevented the tone from being brought as near to the radical syllable as possible, v. c. σῶμα is equal to σόμα. Why then in proparoxytonis, was not the long penultima, the ultimate being short, counted, as to accentuation, for two mora; is neither explained by Mr. G., nor by any one else: the reason may lie in some relation of the tone to the music, which is unknown.

11. Contradistinction of like sounding words. The custom of distinguishing words of like sounds, by diversity of tone, which creates deviations from the general laws of accentuation, is found also in modern languages (v. c. in the English and German) whose accentuation is guided by principles; vóμos, a law, voμós, a pasture—próduce, prodúce—überlegen, uberlégen.

1

III. Dialectical developement. The two above-mentioned deviations cannot, strictly speaking, be regarded as such: the former is but apparently so, the latter is perfectly justified on logical principles; for where accent has any logical meaning, as it ought to have, it will always serve to distinguish different senses of homonymous words. But the third circumstance which produced anomalies, seems more striking, particularly in the manner in which Mr. G. has expressed it. He says: "The Greek language divided itself into different dialects; and these dialects, derived from one fountain-head, acquired, in a manner, independency of each other by literary productions. By this means, an apparent approach to that principle, which we have stated to regulate the accent of the derivated languages, operated in the accentuation of the younger dialects; we mean the tendency to place the accent on the later accessions to the radical syllables, exhibiting neither an independent notion (as the roots), nor an essential alteration or specification (as the compounds)." Mr. G. here refers to that section, in which he has noticed the character of French accentuation. But it would be committing an unpardonable insult to the genius of the Greek language, to charge him with the moon-shine of French accentuation.' Mr. G. himself feels that this assertion rests on a sandy foundation: he accordingly corrects himself some pages further. "The deviation from the Eolic accentuation, in the other dialects, is, however, regulated by fixed principles, tending to point out by the accents, certain modifications of the radical notion." With this we perfectly coincide. In the French and other posterior languages, the accent, deprived of any logical power, indicates merely the external transformation of a primitive language into the new shape. It is not so with the Greek: the accent always preserves-thus Mr. G. continues to illustrate his subject—a logical meaning; but the younger dialects, particularly the Ionic and Attic, (which originally were identical,) adopted the custom of marking out by the accent, more particularly, the modifications which the radical notion had acquired by derivation, than this radical notion itself. This pro

1 Mr. Planche indeed says, in his translation of Demosthenes, that he had discovered an astonishing resemblance between the French and the Greek. The Edinburgh Review (No. LXV. 1820) observes, that this discovery was reserved for Mr. Planche; to which we add, we are afraid that it will also die with him. There is no greater difference between Plato and Helvetius, than between the prattle of French verse, and the harmony of Greek poetry.

ceeding however, was conducted by clear analogical views. Either it was intended to notice, by increased energy of the voice, the character of derivation in general, in contradistinction from the root, and this was mostly the case when the derivative syllable was strong and heavy, or had strikingly altered the radical sound (v. c. βασιλεὺς, ἡγεμών, κτιστὴς, μονὴ from μένω); οι it was proposed to point out by a stress of the tone, certain particular kinds of derivative notions, for instance, the notion of the concrete, or of the acting person, or similar derivated notions: these rules of derivated accentuation, (as we may call it) Mr. G. has attempted to discover, and classify; and in many respects he has succeeded. Wherever the dialectical conformation of the language has not followed these rules, Mr. G. justly asserts, that the custom of language must be said to have preserved the Eolic accentuation, i. e. the original principle of accentuation modified by the euphonic laws of the number of three syllables.

It is evident from our remarks, that the successful progress of this system depended almost entirely upon a judicious and philosophical development of the operation which proceeded from the cause mentioned.

III. And in general the success is unquestionable; though it is to be observed on the one hand, that there were many excellent remarks of German and English philologists, of which Mr. G. could avail himself; and on the other hand, that he has not always written with equal precision, distinctness and felicity. For the essential problem and more arduous part of this system, (not to mention the easier parts) is to detect with sagacity the different, and in part very nice analogies, by which this dialectical accentuation was guided; and we do not think that Mr. G. has exhausted the compass of these analogies. Nevertheless the system is sketched out, and the merit undoubtedly belongs to him. We subjoin some further information respecting it.

Mr. G. reviews, according to these principles, the single parts of speech. The verb presents the fewest difficulties. We cannot however, permit it to pass unnoticed, that Mr. G. did not improve the excellent remarks of Buttmann, which explain the accentuation of the aorist so satisfactorily. To the review of the substantive, the essential rules following from the principle of the dialectical accentuation, (which may be termed the Attic accentuation) are premised; however, says Mr. G., these rules have not entirely prevailed, and in many instances, we VOL. XXXIII. Cl. JI. NO. LXV.

B

must account for the accentuation, by recurring to the Eolic principle, which was prevalent over the secondary principle.

The most essential operations of this secondary (Attic) principle are the following:

I. If the derivation of a noun from a verb, or any other part of speech, is very obvious, the Attic dialect, to mark out the derivative element in sound and notion, has mostly the accent attached to it, it is an oxytonon; however, as we have observed, neither this, nor scarcely any other analogy, is altogether prevailing, and many such words retain the Æolic tone. (v. c. all those ending in os, derived from the verb and denoting the action or the effect, Aóyos, στóλos, póos) Mr. G. however has accurately marked the classes of substantives which in general take the Attic accent.

On the contrary, if the substantives are formed as simply out of the root, by the addition of the final vowel, as the verb by the termination, so that they may be regarded as co-ordinate with, and not subordinate to the verb, the Æolic tone is preserved, unless some other influence takes place, v. c. the simple nouns in anda, after the first declension; in -os, after the second declension, v. c. λúpa, diun, wátos. Here also the enumerations are exact.

II. A second analogy is given by the best of the derivated dialects, to point out the notion of concrete nouns, (to which may be added the ideas of frequentation and collectivity) which are consequently, where the analogy prevails, oxytona; the abstract nouns, on the other hand, retain the old accentuation. Though this analogy also did not altogether prevail, yet the reverse will scarcely be found (unless another circumstance interfere); i. e. a class of substantives, oxytonated in order to denote the idea of abstractness. The only instance which Mr. G. adduces, arises from an oversight. He says, "The feminine substantives in us, vos, chiefly abstract nouns, are oxytona; the masculine and feminine substantives implying a concrete notion, bear the Eolic accent." But the analogy mentioned in a, takes place. The former are derivated by strong additions, the latter in the most simple manner; every one may compare the forms, ἰσχὺς, ὀρχηστὺς, κιθαριστὺς, κτιστὸς, &c. with ἴτυς, χέλυς, νέκυς, γένυς, &c.

From these statements it is evident, that in order to trace the principles which governed the accent of the substantives, we must consider first, whether the Æolic or Attic mode prevailed, and then, which of several analogies obtained the preference; as