i. 1), where the word in each case is said to end in ι, have led

some persons to suppose that Dionysius pronounced οι and αι as real diphthongs of two vowels ending in ι. We know, however, that at this time αι was a single vowel ε prolonged, and that it was only called a diphthong because written with two letters, just as _ea_ in _each_, _great_ are often spoken of as a diphthong, in place of a digraph. We know also that ι subscript was not pronounced, and yet Dionysius speaks of ἀγλαΐᾳ as ending with ι. Consequently there is no need to suppose that οι was a real diphthong either. The language is merely orthographical. As to the amount of pause, we find similar combinations within the same Greek word: οι and ε in οἴεται, ν and δ in ἄνδρα, αι and α in Αἴας; while ν before τ is quite common as in ὄντων, and ν before π, κ becomes μ, γ, as in ἔμπορος, ἐγκρατής. Hence much of this criticism may be fanciful. But it is certain that there is a different feeling respecting the collision of letters which end and begin a word, and those which come together in the same word. Thus in French poetry open vowels are entirely forbidden. It is impossible to say ‘cela ira’ in serious French verse. Yet ‘haïr’ is quite admissible. Hence there may be some foundation for the preceding observations, which, however, like many others in the treatise, ride a theory very hard,” A. J. E. [The observations of the critic, himself, must obviously be accepted with considerable reserve: see, for example, the note on =230= 19 _infra._]

[Page 219]

whose literary sense has been tolerably developed. I will attempt to show by what method such results have been achieved, since it is not by spontaneous accident, but by some kind of artistic design, that this passage has acquired its characteristic form.

The first clause consists of four words—a verb, a connective, and two appellatives. Now the mingling and the amalgamation of the verb and the connective have produced a rhythm which is not without its charm; but the combination of the connective with the appellative has resulted in a junction of considerable roughness. For the words ἐν χορόν are jarring and uneuphonious, since the connective ends with the semivowel ν, while the appellative begins with one of the mutes, χ. These letters by their very nature cannot be blended and compacted, since it is unnatural for the combination νχ to form part of a single syllable; and so, when ν and χ are the boundaries of adjacent syllables, the voice cannot be continuous, but there must necessarily be a pause separating the letters if each of them is uttered with its proper sound. So, then, the first clause is roughened thus by the arrangement of its words. (You must understand me to mean by “clauses” not those into which Aristophanes or any of the other metrists has arranged the odes, but those into which Nature insists on dividing the discourse and into which the disciples of the rhetoricians divide their periods.)

The next clause to this—ἐπί τε κλυτὰν πέμπετε χάριν θεοί—is separated from the former by a considerable interval and includes within itself many dissonant collocations. It begins with one of the vowels, ε, in close proximity to which is another vowel, ι—the letter which came at the end of the preceding

[Page 220]

τὸ πρὸ αὐτοῦ. οὐ συναλείφεται δὲ οὐδὲ ταῦτ’ ἀλλήλοις, οὐδὲ προτάττεται κατὰ μίαν συλλαβὴν τὸ ῑ τοῦ ε̄· σιωπὴ δέ τις μεταξὺ ἀμφοῖν γίνεται, διερείδουσα τῶν μορίων ἑκάτερον καὶ τὴν βάσιν αὐτοῖς ἀποδιδοῦσα ἀσφαλῆ. ἐν δὲ τῇ κατὰ μέρος συνθέσει τοῦ κώλου τοῖς μὲν #ἐπί τε# συνδέσμοις ἀφ’ ὧν 5 ἄρχεται τὸ κῶλον, εἴτε ἄρα πρόθεσιν αὐτῶν δεῖ τὸ ἡγούμενον καλεῖν, τὸ προσηγορικὸν ἐπικείμενον μόριον τὸ #κλυτὰν# ἀντίτυπον πεποίηκε καὶ τραχεῖαν τὴν σύνθεσιν· κατὰ τί ποτε; ὅτι βούλεται μὲν εἶναι βραχεῖα ἡ πρώτη συλλαβὴ τοῦ #κλυτάν#, μακροτέρα δ’ ἐστὶ τῆς βραχείας ἐξ ἀφώνου τε 10 καὶ ἡμιφώνου καὶ φωνήεντος συνεστῶσα. τὸ δὲ μὴ εἰλικρινῶς αὐτῆς βραχὺ καὶ ἅμα τὸ ἐν τῇ κράσει τῶν γραμμάτων δυσεκφόρητον ἀναβολήν τε ποιεῖ καὶ ἐγκοπὴν τῆς ἁρμονίας. εἰ γοῦν τὸ κ̄ τις ἀφέλοι τῆς συλλαβῆς καὶ ποιήσειεν #ἐπί τε λυτάν#, λυθήσεται καὶ τὸ βραδὺ καὶ τὸ τραχὺ τῆς 15 ἁρμονίας. πάλιν τῷ #κλυτὰν# προσηγορικῷ τὸ #πέμπετε# ῥηματικὸν ἐπικείμενον οὐκ ἔχει συνῳδὸν οὐδ’ εὐκέραστον τὸν ἦχον, ἀλλ’ ἀνάγκη στηριχθῆναι τὸ ν̄ καὶ πιεσθέντος ἱκανῶς τοῦ στόματος τότε ἀκουστὸν γενέσθαι τὸ π̄· οὐ γὰρ ὑποτακτικὸν τῷ ν̄ τὸ π̄. τούτου δ’ αἴτιον ὁ τοῦ στόματος 20 σχηματισμὸς οὔτε κατὰ τὸν αὐτὸν τόπον οὔτε τῷ αὐτῷ τρόπῳ τῶν γραμμάτων ἐκφέρων ἑκάτερον· τοῦ μὲν γὰρ ν̄ περὶ τὸν οὐρανὸν γίνεται ὁ ἦχος καὶ τῆς γλώττης ἄκροις τοῖς ὀδοῦσι προσανισταμένης καὶ τοῦ πνεύματος διὰ τῶν ῥωθώνων μεριζομένου, τοῦ δὲ π̄ μύσαντός τε τοῦ στόματος 25

2 προτάττεται] παρ’ οἷς τάττεται F || τις FM: τις ἡ PV 4 ἀσφαλῆι· ἐν δὴ P 5 τοῦ κώλου F: τῶν κώλων PMV || σύνδεσμον F 6 δεῖ] δὴ F 8 κατα τί ποτε· ὅτι F: κατά τι δήποτε PMV 9 μὲν εἶναι] μένειν F 11 καὶ ἡμιφώνου om. P || ἑστῶσα P 13 δυσεκφόρητον F: δυσεκφώνητον E: δυσέκφορον PMV 14 ποιήσει EF 17 τὸν om. EF 18 ἀνάγκηι P 19 τοῦ στόματος τότε E: το̈́ῦτοτε et in margine στομ(ατος) F: τοῦ π̄ τότε M: τότε V: τούτου Ps 20 αἴτιον EF: αἴτιος PMV || στόματος] σχήματος V. 22 ἐκφέρον F || ἑκάτερον F: ἑκάτερον τὸ π̄ καὶ τὸ ν̄ PMV || νῦ FM: om. PV 23 γίνεται F: τε γίνεται PMV || γλώττης F: γλώσσης PMV 24 προἀνισταμένης F, M 25 τε τοῦ στόματος om. F

15. =λυτάν=, =λυθήσεται=: possibly an intentional play on words.

18. Clearly Dionysius does not believe that, in this passage, final ν before initial π was pronounced as μ—κλυτάν as κλυτάμ: though final ν sometimes appears under this form in inscriptions, as also does medial ν in such compounds as συμπόσιον. The literal meaning of the passage seems to be, ‘The ν must be firmly planted [pronounced distinctly, dwelt upon], and κλυτὰν πέμπετε cannot be run together in one word, as κλυταμπέμπετε or the like might be.’

[Page 221]

clause. These letters, again, do not coalesce with one another, nor can ι stand before ε in the same syllable. There is a certain silence between the two letters, which thrusts apart the two elements and gives each a firm position. In the detailed arrangement of the clause the postposition of the appellative part of speech κλυτάν to the connectives ἐπί τε with which the phrase opens (though perhaps the first of these connectives should rather be called a _preposition_) has made the composition dissonant and harsh. For what reason? Because the first syllable of κλυτάν is ostensibly short, but actually longer than the ordinary short, since it is composed of a mute, a semi-vowel, and a vowel. It is the want of unalloyed brevity in it, combined with the difficulty of pronunciation involved in the combination of the letters, that causes retardation and interruption in the harmony. At all events, if you were to remove the κ from the syllable and to make it ἐπί τε λυτάν, there would be an end to both the slowness and the roughness of the arrangement. Further: the verbal form πέμπετε, subjoined to the appellative κλυτάν, does not produce a harmonious or well-tempered sound. The ν must be firmly planted and the π be heard only when the lips have been quite pressed together, for the π cannot be tacked on to the ν. The reason of this is the configuration of the mouth, which does not produce the two letters either at the same spot or in the same way. ν is sounded on the arch of the palate, with the tongue rising towards the edge of the teeth and with the breath passing in separate currents through the nostrils; π with the lips closed, the tongue

[Page 222]

καὶ οὐδὲν τῆς γλώττης συνεργούσης τοῦ τε πνεύματος κατὰ τὴν ἄνοιξιν τῶν χειλῶν τὸν ψόφον λαμβάνοντος ἀθροῦν, ὡς καὶ πρότερον εἴρηταί μοι· ἐν δὲ τῷ μεταλαμβάνειν τὸ στόμα σχηματισμὸν ἕτερον ἐξ ἑτέρου μήτε συγγενῆ μήτε παρόμοιον ἐμπεριλαμβάνεταί τις χρόνος, ἐν ᾧ διίσταται τὸ λεῖόν τε 5 καὶ εὐεπὲς τῆς ἁρμονίας. καὶ ἅμα οὐδ’ ἡ προηγουμένη τοῦ #πέμπετε# συλλαβὴ μαλακὸν ἔχει τὸν ἦχον ἀλλ’ ὑποτραχύνει τὴν ἀκοὴν ἀρχομένη τε ἐξ ἀφώνου καὶ λήγουσα εἰς ἡμίφωνον. τῷ τε #χάριν# τὸ #θεοὶ# παρακείμενον ἀνακόπτει τὸν ἦχον καὶ ποιεῖ διερεισμὸν ἀξιόλογον τῶν μορίων, τοῦ μὲν εἰς ἡμίφωνον 10 λήγοντος τὸ ν̄, τοῦ δὲ ἄφωνον ἔχοντος ἡγούμενον τὸ θ̄· οὐδενὸς δὲ πέφυκε προτάττεσθαι τῶν ἀφώνων τὰ ἡμίφωνα.

τούτοις ἐπιφέρεται τρίτον κῶλον τουτί “#πολύβατον οἵ τ’ ἄστεος ὀμφαλὸν θυόεντα ἐν ταῖς ἱεραῖς Ἀθάναις οἰχνεῖτε#.” ἐνταῦθα τῷ τε #ὀμφαλὸν# εἰς τὸ ν̄ λήγοντι τὸ 15 #θυόεντα# παρακείμενον ἀπὸ τοῦ θ̄ ἀρχόμενον ὁμοίαν ἀποδίδωσιν ἀντιτυπίαν τῇ πρότερον, καὶ τῷ #θυόεντα# εἰς φωνῆεν τὸ ᾱ λήγοντι ζευγνύμενον τὸ “#ἐν ταῖς ἱεραῖς#” ἀπὸ φωνήεντος τοῦ ε̄ λαμβάνον τὴν ἀρχὴν διέσπακε τῷ μεταξὺ χρόνῳ τὸν ἦχον οὐκ ὄντι ὀλίγῳ. τούτοις ἐκεῖνα ἕπεται 20 “#πανδαίδαλόν τ’ εὐκλέ’ ἀγοράν#”· τραχεῖα κἀνταῦθα καὶ ἀντίτυπος ἡ συζυγία· ἡμιφώνῳ γὰρ ἄφωνον συνάπτεται τῷ ν̄ τὸ τ̄ καὶ διαβέβηκεν ἀξιόλογον διάβασιν ὁ μεταξὺ τοῦ τε προσηγορικοῦ τοῦ #πανδαίδαλον# καὶ τῆς συναλοιφῆς τῆς συναπτομένης αὐτῷ χρόνος· μακραὶ μὲν γὰρ ἀμφότεραι, 25 μείζων δὲ οὐκ ὀλίγῳ τῆς μετρίας ἡ συναλείφουσα τὰ δύο συλλαβή, ἐξ ἀφώνου τε καὶ δυεῖν συνεστῶσα φωνηέντων· εἰ

1 γλώττης F: γλώσσης PMV || συνεργούσης] μεριζομένη συνεργούσης F: ἐνεργούσης PV 2 ὡς F: ὡς δὴ PMV 3 δὲ F: δὴ PMV || τὸ στόμα PMV: τὸν F 5 εν ὧι διίσταται P: δι’ οὗ συνίσταται FMV || λεῖόν τε F: λεῖον PMV 6 εὐεπὲς F: εὐπετὲς PV: εὐτελὲς M 7 μακρὸν P 8 ἀρχομένη F: ἄρχουσά PMV 10 ποιεῖ F: ποιεῖ τὸν PMV || διερεισμὸν Us.: ἐρισμὸν P: διορισμὸν FMV 11 τὸ ν̄ Sylburgius: τοῦ ν̄ (νῦ F) FMV: om. P || θῆτα F 14 ἀθάναις F: ἀθήναις PMV 16 θῆτα F 18 ζευγνύμενον F: ἐπεζευγμένον PMV 19 λαμβάνοντος F 20 ἦχον] χρόνον F 21 τραχεῖα κἀνταῦθα om. F 22 συνάπτεται F: συνάπτεται γράμμα PMV 23 διάβασιν FM^1: διάστασιν PVM^2 25 συναπτομένης F: ἐπισυναπτομένης PMV || χρόνος F: om. PMV || μακρὰ et ἀμφότερα F || μὲν γὰρ] μὲν P: γὰρ F: γάρ εἰσιν MV 26 μετρίας F: συμμετρίας PMV || τὰ δύο συλλαβή Us.: τὰς δύο (β̄ P) συλλαβὰς libri 27 δυεῖν FP: δυοῖν MV

2. =ὡς καὶ πρότερον εἴρηταί μοι=: the passages which seem to be meant (=144= 22 and =148= 15) do not exactly tally with the present one.

12. We must supply κατὰ μίαν συλλαβήν, which words are found in =218= 14 and =220= 2 (cp. =230= 4): otherwise we are confronted with such examples to the contrary as ἔνθα and (in this immediate context) μεταλαμβάνειν, ἀρχόμενον, etc.

21. =τ’ εὐ-= are treated as one syllable. So in =218= 22, Dionysius probably intends us to divide as follows:

ᴗ ᴗ – ἐπιτε|κλυτάν,

etc.

23. In Dionysius’ own words, it might be said that the interval between the article ὁ and the noun χρόνος with which it agrees is quite an ‘appreciable gap.’ Cp. Introduction, p. 12 _supra_.

24. =τῆς συναλοιφῆς=: the fused or blended syllable—τ’ εὐ-.

[Page 223]

doing none of the work, and the breath forming a concentrated noise when the lips are opened, as I have said before. While the mouth is taking one after another shapes that are neither akin nor alike, some time is consumed, during which the smoothness and euphony of the arrangement is interrupted. Moreover, the first syllable of πέμπετε has not a soft sound either, but is rather rough to the ear, as it begins with a mute and ends with a semi-vowel. θεοί coming next to χάριν pulls the sound up short and makes an appreciable interval between the words, the one ending with the semi-vowel ν, the other beginning with the mute θ. And it is unnatural for a semi-vowel to stand before any mute.

Next follows this third clause, πολύβατον οἵ τ’ ἄστεος ὀμφαλὸν θυόεντα ἐν ταῖς ἱεραῖς Ἀθάναις οἰχνεῖτε. Here θυόεντα which begins with θ, being placed next to ὀμφαλὸν which ends in ν, produces a dissonance similar to that previously mentioned; and ἐν ταῖς ἱεραῖς which opens with the vowel ε, being linked to θυόεντα which ends with the vowel α, interrupts the voice by the considerable interval of time there is between them. Following these come the words πανδαίδαλόν τ’ εὐκλέ’ ἀγοράν. Here, too, the combination is rough and dissonant. For the mute τ is joined to the semi-vowel ν; and the interval between the appellative πανδαίδαλον and the elided syllable which follows it is quite an appreciable gap; for both syllables are long, but the syllable which unites the two letters ε and υ, consisting as it does of a mute and two vowels, is considerably longer than the average. At any rate, if the τ in the syllable

[Page 224]

γοῦν τις αὐτῆς ἀφέλοι τὸ τ̄ καὶ ποιήσειε #πανδαίδαλον εὐκλέ’ ἀγοράν#, εἰς τὸ δίκαιον ἐλθοῦσα μέτρον εὐεπεστέραν ποιήσει τὴν ἁρμονίαν.

ὅμοια τούτοις ἐστὶ κἀκεῖνα “#ἰοδέτων λάχετε στεφάνων#.” παράκειται γὰρ ἡμίφωνα δύο ἀλλήλοις τὸ ν̄ καὶ τὸ λ̄, φυσικὴν 5 οὐκ ἔχοντα συζυγίαν τῷ μήτε κατὰ τοὺς αὐτοὺς <τόπους μήτε καθ’> ὁμοίους σχηματισμοὺς τοῦ στόματος ἐκφέρεσθαι. καὶ τὰ ἐπὶ τούτοις λεγόμενα μηκύνεταί τε ταῖς συλλαβαῖς καὶ διέστηκε ταῖς ἁρμονίαις ἐπὶ πολύ “#στεφάνων τᾶν τ’ ἐαριδρόπων#”· μακραὶ γὰρ καὶ δεῦρο συγκρούονται συλλαβαὶ τὸ δίκαιον 10 ὑπεραίρουσαι μέτρον, ἥ τε λήγουσα τοῦ #στεφάνων# μορίου δυσὶ περιλαμβάνουσα ἡμιφώνοις φωνῆεν γράμμα φύσει μακρὸν καὶ ἡ συναπτομένη ταύτῃ τρισὶ μηκυνομένη γράμμασιν ἀφώνῳ καὶ φωνήεντι μακρῶς λεγομένῳ καὶ ἡμιφώνῳ· διερεισμός τε οὖν γέγονε τοῖς μήκεσι τῶν συλλαβῶν, καὶ ἀντιτυπία τῇ παραθέσει 15 τῶν γραμμάτων, οὐκ ἔχοντος τοῦ τ̄ συνῳδὸν τῷ ν̄ τὸν ἦχον, ὃ καὶ πρότερον εἴρηκα. παράκειται δὲ καὶ τῷ #ἀοιδᾶν# εἰς τὸ ν̄ λήγοντι ἀπὸ τοῦ δ̄ ἀρχόμενον ἀφώνου τὸ #Διόθεν# τε καὶ τῷ #σὺν ἀγλαΐᾳ# εἰς τὸ ῑ λήγοντι τὸ #ἴδετε πορευθέντ’ ἀοιδᾶν# ἀρχόμενον ἀπὸ τοῦ ῑ. πολλά τις ἂν εὕροι τοιαῦτα 20 ὅλην τὴν ᾠδὴν σκοπῶν.

ἵνα δὲ καὶ περὶ τῶν λοιπῶν εἰπεῖν ἐγγένηταί μοι, Πινδάρου μὲν ἅλις ἔστω, Θουκυδίδου δὲ λαμβανέσθω λέξις ἡ ἐκ τοῦ προοιμίου ἥδε·

Θουκυδίδης Ἀθηναῖος ξυνέγραψε τὸν πόλεμον τῶν 25

1 ἀφέλοι Us. (coll. =220= 14): ἀφέλοιτο libri 2 εὐπετεστέραν PM^1V: εὐεπεστέραν M^2: εὐεπεστάτην F 4 ἰωδέτων M: ὃ δ’ ἐγὼν F || λάχετε στεφάνων PMV: λάχει F 5 γὰρ F: om. PMV 6 αὐτοὺς ὁμοίους F: ὁμοίους PMV: τόπους μήτε καθ’ ins. Usenerus 9 τᾶν τ’] τ’ αὖτ’ P: τ’ αὖ M: ἄν τ’ F: τῶν τ’ V || ἐαριδρόπων F: ἔαριδρέπων PM: ἐἀριδρέπτων V 13 ἡ] μὴ F || μηκυνομένη FM^2: μηκυνθεῖσα PM^1V 14 διερισμός M: διορισμός V 17 ὃ F: ὡς PMV || δὲ] τε F || ἀοιδὰν codd.: λοιβὰν s 18 ἀφώνου FM: ἄφωνον PV || διατεθὲν τε F: διόθεν τέ με PMV 19 πορευθέντα· οἱ δε F: πορευθέντες ἀοιδαν (-δὰν M, -δανὶ V) PMV 20 ἀρχόμενον] ἀρχαῖοι μόνον F 22 μοι F: μοι χρόνος PV: μοι χρόνων M 25 τῶν] τὸν P

1. =ποιήσειε ... ποιήσει=: cp. =220= 14, =256= 23.

6. If Usener’s supplement be not accepted, we might read τῷ μηδὲ κατὰ τοὺς ὁμοίους σχηματισμούς, κτλ.

10. =δεῦρο συγκρούονται=, ‘meet here with a clash,’ as it were.

17. =παράκειται= κτλ.: viz. the ν of ἀοιδᾶν comes next to the δ in διόθεν, and the ι at the end of ἀγλαΐᾳ precedes the ι in ἴδετε.—For ν and δ in juxtaposition cp. English _and_ (where the _d_ is often slurred in pronunciation) and, on the other hand, English _sound_ (where the _d_ is not original).

19. The ι at the end of =ἀγλαΐᾳ= seems, therefore, to have been regarded by Dionysius as a separate letter, and not as an ι ἀνεκφώνητον. Perhaps it was sounded in music; cp. the final _e_ in French. In Dionysius’ time it was not uncommon to omit it even in writing: πολλοὶ γὰρ χωρὶς τοῦ ι γράφουσι τὰς δοτικάς, καὶ ἐκβάλλουσι δὲ τὸ ἔθος φυσικὴν αἰτίαν οὐκ ἔχον (Strabo xiv. 1. 50).

22. =ἐγγένηταί μοι=: cp. _de Lysia_ c. 16 ἵνα δὲ καὶ περὶ τῶν ἰδεῶν ἐγγένηταί μοι τὰ προσήκοντα εἰπεῖν, κτλ.

23. Bircovius compares, with the following passage of Thucydides, the opening of Sallust’s _Bell. Iug._ v. 1: “Bellum scripturus sum, quod populus Romanus cum Iugurtha rege Numidarum gessit, primum quia magnum et atrox variaque victoria fuit, dehinc quia tum primum superbiae nobilitatis obviam itum est; quae contentio divina et humana cuncta permiscuit eoque vecordiae processit ut studiis civilibus bellum atque vastitas Italiae finem faceret.”

24. =τοῦ προοιμίου=: probably the first twenty-three chapters are meant—as far as the word Ἐπίδαμνός ἐστι πόλις κτλ.

25. In the English translation no attempt has been made to reproduce the style of the original Greek. For this purpose the long sentences employed in early English prose-writers are most suitable; e.g. Francis Bacon’s rendering (_Considerations touching a War with Spain_ iii. 516, in _Harleian Miscellany_ v. 84) of Thucyd. i. 23: “The truest cause of this war, though least voiced, I conceive to have been this: that the Athenians being grown great, to the terror of the Lacedemonians, did impose upon them the necessity of a war; but the causes that went abroad in speeches were these,” etc. Thomas Hobbes’ translation of the opening of the History keeps close to the sentence-structure of the original: “Thucydides, an Athenian, wrote the war of the Peloponnesians and the Athenians as they warred against each other, beginning to write as soon as the war was on foot; with expectation it should prove a great one, and most worthy the relation of all that had been before it: conjecturing so much, both from this, that they flourished on both sides in all manner of provision; and also because he saw the rest of Greece siding with the one or the other faction, some then presently and some intending so to do,” etc. Hobbes’ version is well known; but the unpublished translation of Francis Hickes [1566-1631], from which the following extract has been taken by the courtesy of the Librarian of Christ Church, Oxford, is also of much interest: “Thucydides the Athenian hath written the warres of the Peloponnesians and Athenians, with all the manner and fashion of their fight, and tooke in hande to put the same in writinge, as soone as ever the said warres weare begone, for a hope he had, that they would be great, and more worthy of memorie, than all the warres of former tyme have been: conjecturinge so much, because he sawe them both so richlie abound with all provisions thereunto belonginge, and all the rest of the Grecian nations, readie to joyne themselves to the one side or the other; some, presentlie upon their fallinge out, and the rest intendinge to do the like. This, no doubt, was the greatest stirre, that ever was amonge the Grecians, consistinge likewise partly of the Barbarians, and to speake in a word, of many and sundrie nations. As for the acts achieved by them before the tyme of this warre, or former matters yet of more antiquitie, it is impossible to finde out any certaintie, because the tyme is so long past, since they weare performed: but, by these conjectures, which upon due examination of former tymes, I believe to be true, I must thinke they weare of no great moment, either for the course of warre, or any other respect. Now it is most probable, that the country which we now call Grece, had not in old tyme any settled inhabitants, but did often change her dwellers, who weare still easie to be removed from their possessions if they weare urged by any greater forces, for when there was as yet no trade of Merchandise amongst men: no free entercourse of traffique one with another, either by land or sea: none that tilled any more ground, than what would serve to sustaine their present lives: none that had any money in this purse nor any that planted the earth with fruits for they knewe not how soone others would come and bereave them of it, their cities beinge all unwalled and bearing the mind, that they should everie where finde enough to serve their turnes for their dailie sustenance, they weare therefore easie to be driven out of any place; and for that cause, did nether strengthen themselves with great cities, nor warlike furniture for defence.”

[Page 225]

be removed and πανδαίδαλον εὐκλέ’ ἀγοράν be read, the syllable, falling into the normal measure, will make the composition more euphonious.

The words ἰοδέτων λάχετε στεφάνων are open to the same criticism as those already mentioned. For here two semi-vowels, ν and λ, come together, although they do not naturally admit of amalgamation owing to the fact that they are not pronounced with the same configurations of the mouth. The words that follow these have their syllables lengthened and are widely divided from one another in arrangement: στεφάνων τᾶν τ’ ἐαριδρόπων. For here also there is a concurrence of long syllables which exceed the normal measure,—the final syllable of the word στεφάνων which embraces between two semi-vowels a vowel naturally long, and the syllable linked with it, which is lengthened by means of three letters, a mute, a vowel pronounced long, and a semi-vowel. Separation is produced by the lengths of the syllables, and dissonance by the juxtaposition of the letters, since the sound of τ does not accord with that of ν, as I have said before. Next to ἀοιδᾶν, which ends in ν, comes Διόθεν τε, which begins with the mute δ, and next to σὺν ἀγλαΐᾳ, which ends in ι, comes ἴδετε πορευθέντ’ ἀοιδᾶν, which begins with ι. Many such features may be found on a critical examination of the whole ode.

But in order to leave myself time for dealing with what remains, no more of Pindar. From Thucydides let us take this passage of the Introduction:—

“Thucydides, an Athenian, composed this history of the war

[Page 226]

Πελοποννησίων καὶ Ἀθηναίων ὡς ἐπολέμησαν πρὸς ἀλλήλους, ἀρξάμενος εὐθὺς καθισταμένου καὶ ἐλπίσας μέγαν τε ἔσεσθαι καὶ ἀξιολογώτατον τῶν προγεγενημένων, τεκμαιρόμενος ὅτι ἀκμάζοντές τε ᾖσαν ἐς αὐτὸν ἀμφότεροι παρασκευῇ τῇ πάσῃ, καὶ τὸ ἄλλο Ἑλληνικὸν ὁρῶν 5 ξυνιστάμενον πρὸς ἑκατέρους, τὸ μὲν εὐθύς, τὸ δὲ καὶ διανοούμενον. κίνησις γὰρ αὕτη μεγίστη δὴ τοῖς Ἕλλησιν ἐγένετο καὶ μέρει τινὶ τῶν βαρβάρων, ὡς δ’ εἰπεῖν καὶ ἐπὶ πλεῖστον ἀνθρώπων. τὰ γὰρ πρὸ αὐτῶν καὶ τὰ ἔτι παλαιότερα σαφῶς μὲν εὑρεῖν διὰ χρόνου πλῆθος ἀδύνατα 10 ἦν· ἐκ δὲ τεκμηρίων, ὧν ἐπὶ μακρότατον σκοποῦντί μοι πιστεῦσαι ξυμβαίνει, οὐ μεγάλα νομίζω γενέσθαι οὔτε κατὰ τοὺς πολέμους οὔτε ἐς τὰ ἄλλα. φαίνεται γὰρ ἡ νῦν Ἑλλὰς καλουμένη οὐ πάλαι βεβαίως οἰκουμένη, ἀλλὰ μεταναστάσεις τε οὖσαι τὰ πρότερα καὶ ῥᾳδίως ἕκαστοι 15 τὴν ἑαυτῶν ἀπολείποντες βιαζόμενοι ὑπό τινων ἀεὶ πλειόνων. τῆς γὰρ ἐμπορίας οὐκ οὔσης οὐδ’ ἐπιμιγνύντες ἀδεῶς ἀλλήλοις οὔτε κατὰ γῆν οὔτε διὰ θαλάσσης, νεμόμενοί τε τὰ ἑαυτῶν ἕκαστοι ὅσον ἀποζῆν καὶ περιουσίαν χρημάτων οὐκ ἔχοντες οὐδὲ γῆν φυτεύοντες, ἄδηλον 20

1 καὶ] τε καὶ P 4 τε om. EF || ἦσαν libri: sed apud Thucydidem lectio potior ᾖσαν [“ᾖσαν F g Schol. Plat. _Rep._ 449 A Suid. Phot.: ἦσαν cett.”] 6 πρὸς ... διανοούμενον om. P 9 πλεῖστον EF: πλεῖστων sic P: πλείστων MV || καὶ τὰ EFs: καὶ PMV 10 ἐρεῖν P 11 μακρότερον F 13 πολεμίους P || τὰ ἄλλα PMV: τ’ ἄλλα F 16 ἀπολιπόντες F 17 ἐπιμιγνῦντες ἀλλήλοις (om. ἀδεῶς) F 20 οὐδὲ γῆν φυτεύοντες om. F

4. =ᾖσαν=: cp. schol. ad Thucyd. i. 1 ᾖσαν] μετὰ σπουδῆς ἐπορεύοντο.

9. =τά= (before ἔτι) is omitted by the Palatine and the Ambrosian MSS. in _de Thucyd._ c. 20.

[Page 227]

which the Peloponnesians and the Athenians waged against one another. He began as soon as the war broke out, in the expectation that it would be great and memorable above all previous wars. This he inferred from the fact that both parties were entering upon it at the height of their military power, and from noticing that the rest of the Greek races were ranging themselves on this side or on that, or were intending to do so before long. No commotion ever troubled the Greeks so greatly: it affected also a considerable section of the barbarians, and one may even say the greater part of mankind. Events previous to this, and events still more remote, could not be clearly ascertained owing to lapse of time. But from such evidence as I find I can trust however far back I go, I conclude that they were not of great importance either from a military or from any other point of view. It is clear that the country now called Hellas was not securely settled in ancient times, but that there were migrations in former days, various peoples without hesitation leaving their own land when hard pressed by superior numbers of successive invaders. Commerce did not exist, nor did men mix freely with one another on land or by sea. Each tribe aimed at getting a bare living out of the lands it occupied. They had no reserve of capital, nor did they plant the ground with fruit-trees, since it was uncertain, especially as they had

[Page 228]

ὂν ὁπότε τις ἐπελθὼν καὶ ἀτειχίστων ἅμα ὄντων ἄλλος ἀφαιρήσεται, τῆς τε καθ’ ἡμέραν ἀναγκαίου τροφῆς πανταχοῦ ἂν ἡγούμενοι ἐπικρατεῖν οὐ χαλεπῶς ἀνίσταντο.

αὕτη ἡ λέξις ὅτι μὲν οὐκ ἔχει λείας οὐδὲ συνεξεσμένας ἀκριβῶς τὰς ἁρμονίας οὐδ’ ἔστιν εὐεπὴς καὶ μαλακὴ καὶ 5 λεληθότως ὀλισθάνουσα διὰ τῆς ἀκοῆς ἀλλὰ πολὺ τὸ ἀντίτυπον καὶ τραχὺ καὶ στρυφνὸν ἐμφαίνει, καὶ ὅτι πανηγυρικῆς μὲν ἢ θεατρικῆς οὐδὲ κατὰ μικρὸν ἐφάπτεται χάριτος, ἀρχαϊκὸν δέ τι καὶ αὔθαδες ἐπιδείκνυται κάλλος, ὡς πρὸς εἰδότας ὁμοίως τοὺς εὐπαιδεύτους ἅπαντας οὐδὲν δέομαι λέγειν, ἄλλως 10 τε καὶ αὐτοῦ τοῦτό γε τοῦ συγγραφέως ὁμολογήσαντος, ὅτι εἰς μὲν ἀκρόασιν ἧττον ἐπιτερπὴς ἡ γραφή ἐστι, “#κτῆμα δ’ εἰσαεὶ μᾶλλον ἢ ἀγώνισμα εἰς τὸ παραυτίκα ἀκούειν σύγκειται#.” τίνα δ’ ἐστὶ τὰ θεωρήματα οἷς χρησάμενος ὁ ἀνὴρ οὕτως ἀπηνῆ καὶ αὐστηρὰν πεποίηκε τὴν ἁρμονίαν, δι’ 15 ὀλίγων σοι σημανῶ· ῥᾴδιον γὰρ ἔσται μικρὰ μεγάλων εἶναι δείγματα τοῖς μὴ χαλεπῶς ἐπὶ τὴν τοῦ ὁμοίου τε καὶ ἀκολούθου μεταβαίνουσιν θεωρίαν.

3 ἀνίστατο F: ἀπανίσταντο Thucyd. 4 αὕτη EF: αὕτη πάλιν PMV || συνεζευγμένας EV 5 καὶ μαλακὴ EFM: om. PV 6 ὀλισθάνουσα P: ὀλισθαίνουσα FMV 7 καὶ τραχὺ om. EF || στριφνὸν F 11 αὐτοῦ τοῦτό γε PMV: αὐτοῦ τε F: αὐτοῦ E 14 ὁ ἀνὴρ EF: ἀνὴρ PMV 15 ἀπηνῆ M: ἀπεινῆ F: εὐπινῆ PV || διαλόγων F^1 16 σοι σημανῶ PM: σημανῶ EFV || ῥᾴδιον Us.: ῥαιδία F: ῥαῖον P, MV || ἐσται F: ἐστι PMV 18 μεταβαίνουσαι F: μεταβαίνουσι MV

3. For estimates of Thucydides’ style in general cp. not only this passage of Dionysius but also D.H. pp. 131-59, 175-82 (Text and Translation of _Ep. ii. ad Amm._, together with notes and some references to Marcellinus); Croiset _Thucydide: Livres i.-ii._ pp. 102 ff. and _Histoire de la littérature grecque_ iv. pp. 155 ff.; Girard _Essai sur Thucydide_ pp. 210-19; Blass _Att. Bereds._ i. pp. 203-44; Norden _Kunstprosa_ i. pp. 96-101; Jebb in _Hellenica_ pp. 306 ff.

4. This long sentence (Il. 4-14) is, itself, a good example of Greek word-order and the lucidity possible to it.

7. Batteux (pp. 250-3) maintains, in detail, that these comments on the style of Thucydides would also apply to a passage of Bossuet (in the _Oraison funèbre de Henriette Anne d’ Angleterre, duchesse d’Orléans_), which “a tous les caractères d’une composition austère; c’est partout un style robuste, nerveux, âpre même quelquefois, et presque rustique.” The passage is that which describes the abasement of all human grandeur by Death: “La voilà, malgré ce grand cœur, cette princesse si admirée et si chérie; la voilà, telle que la mort nous l’a faite. Encore ce reste tel quel va-t-il disparaître; cette ombre de gloire va s’évanouir, et nous l’allons voir dépouillée même de cette triste décoration. Elle va descendre à ces sombres lieux, à ces demeures souterraines, pour y dormir dans la poussière avec les grands de la terre, comme parle Job; avec ces rois et ces princes anéantis, parmi lesquels à peine peut-on la placer, tant les rangs y sont pressés, tant la mort est prompte à remplir ces places,” etc. Batteux begins his careful and interesting analysis as follows: “Nul choix des sons. _Malgré ce grand cœur_ est dur. _Cette princesse si_ est sifflant: _si admirée et si_; choc de voyelles. _La voilà telle que la mort nous l’a faite_: mots jetés plutôt que placés. _Encore ce reste tel quel va-t-il dis_: pointes de rochers. _De cette triste décoration_ n’est guère plus doux. Et ces trois monosyllables brefs et rocailleux, _comme parle Job_, etc.

9. =αὔθαδες ... κάλλος=: this happy description of Thucydides’ style shows that Dionysius saw in style a mirror of the man (cp. ἀνδρὸς χαρακτὴρ ἐκ λόγου γνωρίζεται, Menand. _Fragm._ 72, and Dionys. H. _Antiqq. Rom._ i. 1 ἐπιεικῶς γὰρ ἅπαντες νομίζουσιν εἰκόνας εἶναι τῆς ἑκάστου ψυχῆς τοὺς λόγους).—The general drift of Dionysius’ phrase is, of course, commendatory: he does not (cp. =120= 8, 9) mean ‘but such beauty as it (Thucydides’ style) displays is archaic and perverse.’

12. These well-known words of Thucydides (i. 22. 4) are quoted also in _de Thucyd._ c. 7.—A scholium on Thucyd. (_l.c._) runs: κτῆμα] κέρδος. κτῆμα, τὴν ἀλήθειαν· ἀγώνισμα, τὸν γλυκὺν λόγον. αἰνίττεται δὲ τὰ μυθικὰ Ἡροδότου. The passage is well elucidated by Lucian, and by Pliny the Younger: (1) Lucian _de conscribenda historica_ c. 42 ὁ δ’ οὖν Θουκυδίδης εὖ μάλα τοῦτ’ ἐνομοθέτησε, καὶ διέκρινεν ἀρετὴν καὶ κακίαν συγγραφικήν, ὁρῶν μάλιστα θαυμαζόμενον τὸν Ἡρόδοτον, ἄχρι τοῦ καὶ Μούσας κληθῆναι αὐτοῦ τὰ βιβλία. κτῆμα γάρ φησι μᾶλλον ἐς ἀεὶ συγγράφειν ἤπερ ἐς τὸ παρὸν ἀγώνισμα, καὶ μὴ τὸ μυθῶδες ἀσπάζεσθαι, ἀλλὰ τὴν ἀλήθειαν τῶν γεγενημένων ἀπολείπειν τοῖς ὕστερον, (2) Pliny _Ep._ v. 8 “nam plurimum refert, ut Thucydides ait, κτῆμα sit an ἀγώνισμα: quorum alterum oratio, alterum historia est.”

13. =εἰσαεί=: Thucydides himself no doubt wrote ἐς αἰεί: see Marcellinus § 52 for αἰεί (rather than ἀεί) as constituting a mark of ἡ ἀρχαία Ἀτθίς in Thucydides.

14. =ὁ ἀνὴρ= (_divisim_) should probably be read: cp. =230= 23.

17. The meaning possibly is, “you can easily proceed with the same line of observation right through work which is consistently of a similar character to this.”

[Page 229]

no fortifications, when some invader would come and rob them of their property. They also thought that they could command the bare necessities of daily life anywhere; and so, for all these reasons, they made no difficulty about giving up their land.”[175]

There is no need for me to say, when all educated people know it as well as I, that this passage is not smooth or nicely finished in its verbal arrangement, and is not euphonious and soft, and does not glide imperceptibly through the ear, but shows many features that are discordant and rough and harsh; that it does not make the slightest approach to attaining the grace appropriate to an oration delivered at a public festival or to a speech on the stage, but is marked by a sort of antique and self-willed beauty. Indeed, the historian himself admits that his narrative is but little calculated to give pleasure when heard: “it has been composed as a possession for all time rather than as an essay to be recited at some particular competition.”[176] I will briefly point out to you the principles by following which the author has made the arrangement so rugged and austere. Small things will readily serve you as samples of great: you can easily go on noting resemblances and making comparisons for yourself.

[Page 230]

αὐτίκα ἐν ἀρχῇ τῷ #Ἀθηναῖος# προσηγορικῷ τὸ #ξυνέγραψε# ῥῆμα ἐφαρμοττόμενον διίστησιν ἀξιολόγως τὴν ἁρμονίαν· οὐ γὰρ προτάττεται τὸ σ̄ τοῦ ξ̄ κατὰ συνεκφορὰν τὴν ἐν μιᾷ συλλαβῇ γινομένην· δεῖ δὲ τοῦ σ̄ σιωπῇ καταληφθέντος τότε ἀκουστὸν γενέσθαι τὸ ξ̄. τοῦτο δὲ τραχύτητα 5 ἐργάζεται καὶ ἀντιτυπίαν τὸ πάθος. ἔπειθ’ αἱ μετὰ τοῦτο γινόμεναι συγκοπαὶ τῶν ἤχων, τοῦ τε ν̄ <καὶ τοῦ π̄> καὶ τοῦ τ̄ καὶ τοῦ π̄ καὶ τοῦ κ̄ τετράκις ἑξῆς ἀλλήλοις παρακειμένων, χαράττουσιν εὖ μάλα τὴν ἀκοὴν καὶ διασαλεύουσιν ἀξιολόγως τὰς ἁρμονίας, ὅταν φῇ “#τὸν πόλεμον τῶν Πελοποννησίων 10 καὶ Ἀθηναίων#”· τούτων γὰρ τῶν μορίων τῆς λέξεως οὐδὲν ὅ τι οὐ καταληφθῆναί τε δεῖ καὶ πιεσθῆναι πρότερον ὑπὸ τοῦ στόματος περὶ τὸ τελευταῖον γράμμα, ἵνα τὸ συναπτόμενον αὐτῷ τρανὴν καὶ καθαρὰν τὴν ἑαυτοῦ λάβῃ δύναμιν. ἔτι πρὸς τούτοις ἡ τῶν φωνηέντων παράθεσις ἡ κατὰ τὴν 15 τελευταίαν τοῦ κώλου τοῦδε γενομένη ἐν τῷ #καὶ Ἀθηναίων# διακέκρουκε τὸ συνεχὲς τῆς ἁρμονίας καὶ διέστακεν πάνυ αἰσθητὸν τὸν μεταξὺ λαβοῦσα χρόνον· ἀκέραστοι γὰρ αἱ φωναὶ τοῦ τε ῑ καὶ τοῦ ᾱ καὶ ἀποκόπτουσαι τὸν ἦχον· τὸ δ’ εὐεπὲς οἱ συνεχεῖς τε καὶ οἱ συλλεαινόμενοι ποιοῦσιν ἦχοι. 20

καὶ αὖθις ἐν τῇ δευτέρᾳ περιόδῳ τὸ προηγούμενον κῶλον τουτί “#ἀρξάμενος εὐθὺς καθισταμένου#” μετρίως ἁρμόσας ὁ ἀνὴρ ὡς ἂν εὔφωνόν τε μάλιστα φαίνοιτο καὶ μαλακόν, τὸ μετὰ τοῦτο πάλιν ἀποτραχύνει καὶ διασπᾷ τοῖς διαχαλάσμασι τῶν ἁρμονιῶν· “#καὶ ἐλπίσας μέγαν τε ἔσεσθαι καὶ 25 ἀξιολογώτατον τῶν προγεγενημένων#.” τρὶς γὰρ ἀλλήλοις ἑξῆς οὐ διὰ μακροῦ παράκειται τὰ φωνήεντα συγκρούσεις ἐργαζόμενα καὶ ἀνακοπὰς καὶ οὐκ ἐῶντα τὴν ἀκρόασιν ἑνὸς κώλου συνεχοῦς λαβεῖν φαντασίαν· ἥ τε περίοδος αὐτῷ λήγουσα εἰς τὸ “#τῶν προγεγενημένων#” οὐκ ἔχει τὴν 30 βάσιν εὔγραμμον καὶ περιφερῆ, ἀλλ’ ἀκόρυφός τις φαίνεται

2 ἐφαμαρτόμεν(ον) F: ἐπαγόμενον E 6 μετὰ τούτων F 7 καὶ τοῦ π̄ (post ν̄) ins. Uptonus 8 παρακειμένων Us.: παρακείμεναι libri 11 οὐδὲν PMV: οὐθὲν EF 12 οὖν F: οὐχὶ EPMV: οὐ <σιωπῇ> Us. 13 ὑπὸ] ἐπὶ P || τελευταῖαν F, MV: om. P 17 διέστακεν P, MV: διέστηκε EF 18 γὰρ EF: τε γὰρ PMV 21 καὶ αὖτις F: αὖθις PMV || τὸ F: om. PMV 24 ἀποτραχύνει PV: ἐπιτραχύνει FM || διαχαλάσμασιν P: ἀπὸχαλασμασι F 26 τρὶς Sauppe: τρία libri 27 ἑξῆς οὐ] ἐξ ἴσου P 29 λαβεῖν φαντασίαν F: φαντασίαν λαμβάνειν PMV

9. Perhaps an effect analogous to that of syncopation in music is meant.

10, 11. Different words, and a different order, seem hardly possible here. If πόλεμον were put after Ἀθηναίων, the juxtaposed letters would be much the same as in the existing arrangement.

16. =τελευταίαν=: it may be that some word like συγκοπήν is to be supplied. Or τελευτὴν may be read: or τελευταῖα.

19. The present passage (lines 15-19) shows, as Blass (_Ancient Greek Pronunciation_ p. 66) remarks, that the educated pronunciation of the Augustan period did not confuse αι with ε.

22-5. Here, again, the author would hardly have much _choice_ in the arrangement of the words in question.

26. =τρίς=: viz. in the words καὶ ἐλπίσας, τε ἔσεσθαι, καὶ ἀξιολογώτατον.

[Page 231]

At the very beginning the verb ξυνέγραψε, being appended to the appellative Ἀθηναῖος, makes an appreciable break in the verbal structure, since σ is never placed before ξ with a view to being pronounced in the same syllable with it. The sound of σ must be sharply arrested by an interval of silence before the ξ is heard; and this circumstance causes roughness and dissonance. Moreover, the interruptions of the voice in what follows, in consequence of the four successive juxtapositions νπ, ντ, νπ, νκ, grate violently upon the ear, and cause a remarkable succession of jolts when he says τὸν πόλεμον τῶν Πελοποννησίων καὶ Ἀθηναίων. Of these words there is not one that must not first be checked by the mouth with a stress on the last letter, in order that the next letter to it may be uttered clearly and purely with its own proper quality. Furthermore, the juxtaposition of vowels which is found at the end of this clause in the words καὶ Ἀθηναίων has broken and made a gap in the continuity of the arrangement, by demanding quite an appreciable interval, since the sounds of ι and α are unmingled and there is an interruption of the voice between them: whereas euphony is caused by sounds which are continuous and smoothly blended.

Again, in the second period the first clause ἀρξάμενος εὐθὺς καθισταμένου has been pretty successfully arranged by the author in the way in which it would produce the most smooth and euphonious effect. But he roughens and dislocates the very next clause by sundering its joints: καὶ ἐλπίσας μέγαν τε ἔσεσθαι καὶ ἀξιολογώτατον τῶν προγεγενημένων. For thrice in close succession vowels are juxtaposed which cause clashings and obstructed utterance, and make it impossible for the ear to take in the impression of one continuous clause; and the period which he ends with the words τῶν προγεγενημένων has no well-defined and rounded close, but seems to be without beginning or

[Page 232]

καὶ ἀκατάστροφος, ὥσπερ μέρος οὖσα τῆς δευτέρας ἀλλ’ οὐχὶ [τῆς πρώτης] τέλος.

τὸ δ’ αὐτὸ πέπονθε καὶ ἡ τρίτη περίοδος· καὶ γὰρ ἐκείνης ἀπερίγραφός ἐστι καὶ ἀνέδραστος ἡ βάσις τελευταῖον ἐχούσης μόριον “#τὸ δὲ καὶ διανοούμενον#”· πολλὰς ἅμα καὶ αὐτὴ 5 περιέχουσα φωνηέντων τε πρὸς φωνήεντα ἀντιτυπίας καὶ ἡμιφώνων πρὸς ἡμίφωνα καὶ ἄφωνα, ἅσπερ ἐργάζεται τὰ μὴ συνῳδὰ τῇ φύσει τραχύτητας. ἵνα δὲ συνελὼν εἴπω, δώδεκά που περιόδων οὐσῶν ἃς παρεθέμην, εἴ τις αὐτὰς συμμέτρως μερίζοι πρὸς τὸ πνεῦμα, κώλων δὲ περιλαμβανομένων ἐν 10 ταύταις οὐκ ἐλαττόνων ἢ τριάκοντα, τὰ μὲν εὐεπῶς συγκείμενα καὶ συνεξεσμένα ταῖς ἁρμονίαις οὐκ ἂν εὕροι τις ἓξ ἢ ἑπτὰ τὰ πάντα κῶλα, φωνηέντων δὲ συμβολὰς ἐν ταῖς δώδεκα περιόδοις ὀλίγου δεῖν τριάκοντα, ἡμιφώνων τε καὶ ἀφώνων ἀντιτύπων καὶ πικρῶν καὶ δυσεκφόρων παραβολάς, ἐξ ὧν αἵ 15 τε ἀνακοπαὶ καὶ τὰ πολλὰ ἐγκαθίσματα τῇ λέξει γέγονε, τοσαύτας τὸ πλῆθος ὥστε ὀλίγου δεῖν καθ’ ἕκαστον αὐτῆς μόριον εἶναί τι τῶν τοιούτων. πολλὴ δὲ καὶ ἡ τῶν κώλων ἀσυμμετρία πρὸς ἄλληλα καὶ ἡ τῶν περιόδων ἀνωμαλία καὶ ἡ τῶν σχημάτων καινότης καὶ τὸ τῆς ἀκολουθίας ὑπεροπτικὸν 20 καὶ τὰ ἄλλα ὅσα χαρακτηρικὰ τῆς ἀκομψεύτου τε καὶ αὐστηρᾶς ἐπελογισάμην ὄντα ἁρμονίας. ἅπαντα γὰρ διεξιέναι πάλιν ἐπὶ τῶν παραδειγμάτων καὶ καταδαπανᾶν εἰς ταῦτα τὸν χρόνον οὐκ ἀναγκαῖον ἡγοῦμαι.

XXIII

ἡ δὲ γλαφυρὰ [καὶ ἀνθηρὰ] σύνθεσις, ἣν δευτέραν ἐτιθέμην

2 τῆς πρώτης uncis inclusit Usenerus 4 ἐχούσης Us.: ἔχουσα libri 7 καὶ ... ἐργάζεται om. F || καὶ ἄφωνα P: om. FMV || ἅσπερ] ἅπερ PMV 8 τραχύτητας F: καὶ τραχύτητας PMV 9 εἴ τις] εἴπερ F 10 δὲ F: δὲ τῶν PMV || περιλαμβανομένων F: ἐμπεριλαμβανομένων PMV 11 ταύταις F: αὐταῖς PMV 12 τις ἑξῆς ἢ πάντα ταῦτα κῶλα F 13 συλλαβὰς F 14 καὶ ἀφώνων καὶ ἀντιτύπων P 17 τοσαύτας Uptonus: τοσαῦτα libri (cf. =160= 20) 20 σχημάτων F: σχηματισμῶν PMV 21 τὰ ἄλλα PMV: τἆλλα F || χαρακτηρικὰ F: χαρακτηριστικὰ PV: χαρακτηριστικὰ καὶ M || ἀκομψεύστου FMV 22 αὐστηρᾶς] ἰσχυρᾶς F || ἀπελογησάμην PM^2: ἐπελογησάμην M^1V || διεξιέναι F: ἐπεξιέναι PMV 25 καὶ ἀνθηρὰ om. P || ἐτιθέμην F: ἐθέμην PMV

1. Dionysius seems to discern three periods in the first sentence of Thucydides, viz. (1) Θουκυδίδης ... ἀλλήλους (2) ἀρξάμενος ... προσγεγενημένων, (3) τεκμαιρόμενος ... διανοούμενον. The general sense here is: ‘as there is no connexion between ἀρξάμενος and τεκμαιρόμενος, we must take the latter as beginning a new period, and yet logically ἀρξάμενος belongs to it.’ If the words τῆς πρώτης are to be retained at all, they might possibly be transported with τῆς δευτέρας: ‘as though it were a part of the first period and not the end of the second.’

4. Usener’s =ἐχούσης= seems likely, though the words καὶ γὰρ ... ἡ βάσις might be regarded as parenthetical and ἔχουσα as in agreement with περίοδος.

18. =πολλὴ δὲ καί= κτλ.: cp. Cic. _Orat._ ix. 32. 33 “itaque numquam est (Thucydides) numeratus orator ... sed, cum mutila quaedam et hiantia locuti sunt, quae vel sine magistro facere potuerunt, germanos se putant esse Thucydidas.”

25. For =ἀνθηρά= cp. n. on =208= 26 _supra_.—The whole chapter should be compared with _de Demosth._ c. 40. In c. 49 of that treatise Dionysius refers expressly to his previously written _de Compositione_: εἰ δέ τις ἀπαιτήσει καὶ ταῦτ’ ἔτι μαθεῖν ὅπῃ ποτ’ ἔχει, τοὺς ὑπομνηματισμοὺς ἡμῶν λαβών, οὓς περὶ τῆς συνθέσεως τῶν ὀνομάτων πεπραγματεύμεθα, πάντα ὅσα ποθεῖ τῶν ἐνθάδε παραλειπομένων εἴσεται (cp. c. 50 _ibid._).

[Page 233]

conclusion, as if it were part of the second period and not its termination.

The third period has the same characteristics. There is a lack of roundness and stability in its foundation, since it has for its concluding portion τὸ δὲ καὶ διανοούμενον. Further, it too contains many clashings of vowel against vowel and of semi-vowels against semi-vowels and mutes—discords produced by things in their very nature inharmonious. To sum up, here are some twelve periods adduced by me—if the breathing-space be taken as the criterion for the division of period from period; and they contain no fewer than thirty clauses. Yet of these not six or seven clauses in all will be found to be euphoniously composed and finished in their structure; while of hiatus between vowels in the twelve periods there are almost thirty instances, together with meetings of semi-vowels and mutes which are dissonant, harsh, and hard to pronounce. It is to this that the stoppages and the many retardations in the passage are due; and so numerous are these concurrences that there is one of the kind in almost every single section of it. There is a great lack of symmetry in the clauses, great unevenness in the periods, much innovation in the figures, disregard of sequence, and all the other marks which I have already noted as characteristic of the unadorned and austere style. I do not consider it necessary to waste our time by going over the whole ground once more with the illustrative passages.