12. In Euripides, even those dissyllabic words, (alluded to §. 9.) wherever, from its position, the syllable is decisively long or short, exhibit that syllable thrice short to one case of long. Consequently, in certain positions (unictuated) of Iambic or Trochaic verse which indifferently admit either quantity, there can be no reasonable ground for supposing that syllable to be lengthened: of course, therefore, the following lines are thus read:

Med. 226. πι-κρὸς πολίταις ἐστὶν ἀμαθίας ὕπο.

Iph. Α. 891. ἐπὶ τίνος σπουδαστέον μοι μᾶλλον, ἢ τε-κνου πέρι; 13. In cases where the augment falls as in éékλwσev or κεκλῆσθαι, or where, as in πολύχρυσος and ἀπότροποι, the short vowel closes the first part of a composite word, the prolongation of that syllable in Euripides, though not altogether avoided, is yet exceedingly rare. (R. P. ad Orest. 64.)

14. One great cause of the many mistakes about syllabic quantity should seem to be involved in that false position of S. Clarke's, (ad II. B. 537.) that a short vowel preceding any two consonants with which a syllable can be commenced may form a short syllable. Nothing was ever more unluckily asserted, or more pregnant with confusion and error.

15. To the perspicacity and acuteness of Dawes (M. C. pp. 90, 1. 196. 146, 7.) we are indebted for the first clear statement of the principal points in this department of prosody to the deliberate and masterly judgment of Porson (ad Orest. 64. and elsewhere) we owe whatever else is correctly and certainly known.

16. Some little things, however, may serve to show, that an English ear, especially on a sudden appeal, is no very competent judge of Attic correptions, so called.

For instance, in the following lines,

Phœn. 1444. ἐν τῷδε μήτηρ ἡ τάλαινα προσπίτνει,

Alc. 434. ἐπίσταμαί γε, κοὐκ ἄφνω κακὸν τόδε,

it is not from any practice of our own, certainly, that we should pronounce the words προσπίτνει and ἄ-φνω, with precision and facility in that very way.

17. So, too, if akun and eoμev were on a sudden proposed as to the shortening of the first syllable in each, it might seem

NN 2

to an English ear just as improbable in the noun as in the verb ; although in Athenian utterance we know very well the fact was quite otherwise.

That eminently learned and powerful scholar, Toup, (vid. Emendd. Vol. 1. 114, 5. Iv. 441.) stoutly maintained in his day (what is now called) the permissiveness of ou; and actually on that ground suggested the following as an emendation of a passage in Sophocles, for ἐμὲν or ἴμεν:

Elect. 21, 2.......

.......ὡς ἐνταῦθ ̓ ἐσμὲν,

ἵν ̓ οὐκέτ ̓ ὀκνεῖν καιρὸς, ἀλλ ̓ ἔργων ἀκμή. (where άkun of course is right enough, being pronounced ά-Âμŋ.) Since Porson's delicate correction of that error (u. s. p. 441.) no argument has been advanced in its defence. And yet, a priori, why should not ou be permissive, as well as Ou, for instance? "The consonants ou can begin a word: why not commence a separate syllable? How can ou commence a syllable, when notoriously it cannot begin a word ?"

Honesta oratio est.

κμ

18. The plain truth however stands thus: that ku and Oμ, (with xu, pv, Tv,) though never used as initial to any word, yet within the same words are found permissive much too often to admit the shadow of a doubt on that head.

Phœn. 551. Καὶ γὰρ μέτρ ̓ ἀνθρώποισι καὶ μέρη σταθμῶν may be taken for one undisputed example: there is no want of more.

19.

How far in the different pairs of consonants which have been defined as non-permissive (§. 3.) a physical necessity was the obstacle, in some at least, if not in others, might be a question for anatomy rather than for criticism.

PROSODY.

1. Vocalis brevis ante consonantes.

1. Vocalis brevis ante vel tenues, quas vocant, consonantes π, к, т, vel adspiratas 4, x, 0, sequente quavis liquida; uti et ante medias ẞ, y, d, sequente p; syllabam brevem perpetuo

claudit.

2. Vocalis brevis ante consonantes medias ẞ, y, d, sequente quavis liquida præter unicam p, syllabam brevem nunquam terminat, sed sequentium consonarum ope longam semper constituit. Dawes. Misc. Crit. p. 353.

2. Syllabæ in quibus concurrunt consonantes Bλ, yλ, yu, yv, δμ, δν.

Κλύουσα θρήνους, οὐκ ἂν ἐκβάλοι δάκρυ;

Primo Opvos, deinde yλnvous conjicit Musgravius. Nihil opus. Præterea vous metrum vitiaret. Dawesius canonem paullo temerarius, ut solet, statuit, nullam syllabam a poëta scenico corripi posse, in qua concurrant consonantes Bλ, yλ, yu, yv, du, dv. Hæc regula, plerumque vera, nonnunquam ab Æschylo, Sophocle, Aristophane, violatur, ab Euripide credo nunquam.

Porson. ad Hec. v. 298.

3. Παρθένον, ἐμῇ τε μητρὶ παρέδωκεν τρέφειν.

cur N finalem in éπékλwσev, v. 12, et similibus addiderim, nemo nisi qui communi sensu plane careat, requiret. Sed erunt fortasse nonnulli, qui minus necessario hoc factum arbitraturi sint in παρέδωκεν. Tapédwκev. Rationes igitur semel exponam, nunquam posthac

moniturus. Quanquam enim sæpe syllabas natura breves positione producunt Tragici, longe libentius corripiunt, adeo ut tria prope exempla correptarum invenias, ubi unum modo extet productarum. Sed hoc genus licentiæ, in verbis scilicet, cum compositis, qualia Tékvov, Táтρos, ceteris longe frequentius est. Rarius multo syllaba producitur in verbo composito, si in ipsam juncturam cadet, ut in Toλúxpvσos Andr. 2. Eadem parsimonia in augmentis producendis utuntur, ut in éékλwσev sup. 12. kekAñola Sophocl. Elect. 366. Rarior adhuc licentia, ubi præpositio verbo jungitur, ut in άлóтроπоι, Phœn. 595 (600). Sed ubi verbum in brevem vocalem desinit, eamque duæ consonantes excipiunt, quæ brevem manere patiantur, vix credo exempla indubiæ fidei inveniri posse, in quibus syllaba ista producatur. Quod si ea, quæ disputavi, vera sunt, planum est, in fine vocis addendam esse literam, quam addidi. Porson. ad Orest. v. 64.

4. In Anapæsticis συνάφεια.

Nempe dimetri cujuscunque generis continuo carmine per Ovvapeíav decurrunt, usque dum ad versum catalecticum, quo omne systema claudatur, deventum sit. Hanc ovvapeíav in anapasticis locum habere primus docuit, non jam, uti ipse ad Hor. Carm. III. 12. 6. asseverat Cl. Bentleius; sed Terentianus. utique pag. 58. [1. 9.] hæc habet:

Απ' ελάσσονος autem cui nomen indiderunt

In nomine sic est dioμndñs: metron autem

Non versibus istud numero aut pedum coarctant;
Sed continuo carmine, quia pedes gemelli

Urgent brevibus tot numero jugando longas:
Idcirco vocari voluerunt συνάφειαν.

Anapastica fiunt itidem per συνάφειαν.

Is

Dawes. Misc. Crit. pp. 55, 56.

5.

Tragici nunquam ita senarium disponunt, ut pedes tertius et quartus unam vocem efficiant. Porson. ad Hec. 728.

6. Περί ante vocalem.

Tragici nunquam in senarios, trochaicos, aut, puto, anapastos legitimos, περὶ admittunt ante vocalem, sive in eadem, sive in diversis vocibus. Imo ne in melica quidem verbum vel substantivum hujusmodi compositionis intrare sinunt; raro admodum adjectivum vel adverbium.-Huc adde, quod Tragici, si vocem puram a Tepi compositum adhibent, huic vitio per tmesin medentur, ut Bacch. 619. Troad. 561. Porson. ad Med. 284.

7. Τί δὲ πλέον; ἦλθον Αμφιάρεω γε προς βίαν.

Eurip. Supp. v. 158.

Instead of τί δὲ πλέον, Mr Porson (Præf. ad Hec. p. 40) silently reads τί πλεῖον, which reading Mr Gaisford has admitted into the text. It is certain, that in Tragic iambics, a monosyllable which is incapable of beginning a verse, as ἂν, γὰρ, δὲ, μὲν, τε, Tis, is very rarely employed as the second syllable of a tribrach or dactyl. To the best of our knowledge, Eschylus affords no example of this licence, and Sophocles only two:

Οὐδέποτέ γ· οὐδ ̓ ἣν χρῇ με πᾶν παθεῖν κακόν. Phil. 999. Οὐδέποθ' ἑκόντα γ ̓ ὥστε τὴν Τροίαν ἰδεῖν. Ib. 1992. Perhaps, however, in these verses οὐδέποτε is to be considered as one word, as it is commonly represented. In the remains of Euripides, we have observed the following examples:

I. Οὐδὲ πάθος, οὐδὲ συμφορὰ θεήλατος. Οr. 2.

II.

III.

Ξυνδεῖ. Τὸ γὰρ ἴσον, νόμιμον ἀνθρώποις ἔφν. Phan. 548.
Εἰ γὰρ ἐπὶ τέρμα, καὶ τὸ πλέον ἐμῶν κακῶν.

Suppl. 368. IV. Οὐδὲ σε φέρειν γ ̓ ἅπασιν Ελλησιν κακά. Iph. Αul. 308. The common reading is, Οὐδὲ σε φέρειν δεῖ πᾶσιν.

VI.

V. Εἰ δέ τι κόρης σῆς θεσφάτων μέτεστί σοι. Ib. 498. Ἀλλ ̓ οὐχ ὁμοίως ἂν ὁ θεὸς τιμὴν ἔχοι. Bacch. 192. The true reading seems to be,

Ἀλλ ̓ οὐχ ὁμοίαν ὁ θεὸς ἂν τιμὴν ἔχοι.

VII. Ὥστε διὰ τοῦτον τἀγάθ ̓ ἀνθρώπους ἔχειν. Ib. 285.
Perhaps διὰ τοῦτον ὥστε.

VIII. Οὐδέποτ ̓ ἐδόξασ'. Οὐδ ̓ ἐγὼ γὰρ ἤλπισα. Elect. 580.