E. M. Cope, Commentary on the Rhetoric of Aristotle

---

On rhythm in Prose.

In the paraphrase of the Introduction I have already given an outline of the contents of this chapter and their connexion, with references and some details, pp. 303—306. And on rhythm in general, and its application to prose, there is an Appendix (C), pp. 379—392; in which is a full account of its original and derived significations in the first part, and of its distinction from μέτρον in the second. The commentary on this chapter will therefore deal principally with the details of the language, allusions, and such particulars as require explanation, which are omitted in the paraphrase.

In the fragments of Isocrates' τέχνη, collected by Benseler in the Teubner series, Vol. II p. 276, we have the following, fragm. 4—cited from Maximus Planudes ad Hermog. and Joannes Siceliotes—ὅλως δὲ ὁ λόγος μὴ λόγος ἔστω: ξηρὸν γάρ: μηδὲ ἔμμετρος: καταφανὲς γάρ: ἀλλὰ μεμίχθω παντὶ ῥυθμῷ, μάλιστα ἰαμβικῷ ἢ τροχαϊκῷ. The first of these precepts is in entire agreement with Aristotle, § 1; the disagreement of the second with the statements of § 4 is equally striking. It seems from what is said of Thrasymachus and the paean in § 4, that the subject of prosaic rhythm was not included in the τέχναι of himself and the succeeding writers on Rhetoric. It does not appear even in the Rhet. ad Alex. Cicero, de Or. III 44. 173, attributes to Isocrates the first introduction of ‘numbers’ into prose composition.

Dionysius de Comp. Verb. c. 25 (p. 197 R.) refers to this chapter of Aristotle in support of his observations on rhythm in prose. His own opinions on the subject are given, pp. 195, 6.

References are made by Cicero to this chapter (§ 4 et seq.), de Or. 1 47. 182, 183, in the course of his dissertation on rhythm, from § 171 foll. The same subject is treated, Orat. c. LXIII 212 seq. The various rhythms heroic, iambic, trochaic, &c. are discussed in c. LXIV, where Aristotle's opinions, as expressed in this chapter, are twice referred to, §§ 215, 218. In § 214 we have, temeritas ex tribus brevibus et longa est, quem (sc. paeanem) Aristoteles ut optimum probat, a quo dissentio. Cicero is referring to this chapter, from which the other references are taken: and as this is not found there, he must be either quoting inexactly, from memory, or perhaps confounding Aristotle's opinion on the point with that of one of the other rhetoricians whom he mentions, § 218. There is likewise an incorrectness in the opinion which he there attributes to Aristotle, that the paean is, aptissimus orationi vel orienti vel mediae: Aristotle says nothing of the ‘middle’ of the sentence.

Compare also, Demetrius περὶ ἑρμηνείας, περὶ μεγαλοπρεποῦς, § 38 seq. (Rhet. Gr. Spengel, III 270—273) who also refers thrice to this chapter of the Rhetoric. Quint. IX 4. 45 seq. There are references to this ch. in §§ 87, 88.

On the abuse of rhythm, which degrades and is incompatible with the sublime, there is a short chapter in Dionysius περὶ ὕψους, c. 41.

---

‘The structure (figure, fashion) of the language (i. e. prose composition) should be neither metrical (run into verse)¹ nor entirely without measure or rhythm; for the one has no power of persuasion, because it is thought to be artificial (supra, c. 2. 4, πεπλασμένως), and at the same time also diverts (the hearers' attention, from the main subject or the proof of the fact): for it makes him attend to the recurrence of the similar cadence. And so (the audience anticipate the answering or recurring cadence) just as the children anticipate the answer to the herald's summons, “Whom does the freedman choose for his attorney? and the answer is, Cleon”’.

ἐπίτροπος one who is charged or entrusted with the management of his case, or of any business as deputy for another; procurator, ἐπιτρόποις Καίσαρος, Plut. Praec. Ger. Reip. c. 17, 813 E, ὡς αὐτὸς μὲν οὐκ ἐπεμελήθη τούτων, ὁ δ᾽ ἐπίτροπος Μιλύας, ‘his man of business, deputy, agent’.

On Cleon's self-assumed functions of public prosecutor and poor man's advocate, see Grote, Hist. Gr. ch. LIV, Vol. VI. p. 667 seq. An example in Arist. Ran. 569, (one of the tavern-keepers says,) ἴθι δὴ κάλεσον τὸν προστάτην Κλέωνά μοι, (and the other) σὺ δ᾽ ἔμοιγ̓, ἐάνπερ ἐπιτύχῃς, Ὑπέρβολον, ἵν̓ αὐτὸν ἐπιτρίψωμεν: from which Mr Grote draws his inferences as to the real nature of Cleon's misrepresented policy. The children, in the illustration, are so accustomed to the invariable reply to the herald's proclamation, for an attorney or deputy to plead some freedman's cause—who by law was not allowed to speak for himself in court— that they have learned to say ‘Cleon’ whenever the question is asked. It has not been noticed that this story is told in the present tense, as if the children were in the habit of doing this in Aristotle's own time. Can it be meant that the custom had been handed down from generation to generation for a century or so after Cleon's death? If so, it is a very remarkable fact.

With the opening words of the chapter, comp. Cic. Orat. LI 172, Is (Aristoteles) igitur versum in oratione vetat esse, numerum iubet. Ib. § 189, of verses unintentionally introduced by the orator in his speech, Inculcamus per imprudentiam...versus; vitiosum genus, et longa animi provisione fugiendum. With ἀπίθανον κ.τ.λ., comp. Ib. LXII 209, Si enim semper utare (these studied arts and tricks of rhetoric), quum satietatem adfert tum quale sit etiam ab imperitis agnoscitur. Detrahit praeterea actionis dolorem, aufert humanum sensum actoris, tollit funditus veritatem et fidem... LXV 220, Multum interest utrum numerosa sit, id est, similis numerorum an plane e numeris constet oratio. Alterum si fit, intolerabile vitium est; alterum nisi fit, dissipata et inculta et fluens est oratio.

---

‘That (composition) which is (entirely) devoid of rhythm (has no measure) is indefinite (or, unlimited), but it ought to be limited, only not by metre (like verse): for the infinite (indefinite, unlimited) is displeasing and (i. e. because it) cannot be known. But everything is defined (or limited) by number; and the number (numerus in both its senses) of the structure of the language (prose composition) is rhythm, of which metres are so many sections’. Here we pass for a moment into Platonic metaphysics. The doctrine of the formless, vague, indefinite, unlimited, infinite of more or less, of degree; into which τὸ μέτριον order, harmony, measure, symmetry, law—the mean—are introduced by the limiting πέρας, the definite principle; coming originally from the Pythagoreans, is adopted and expounded by Plato in the Philebus, 23 E et seq. The principle is applied to the numbers or measures of music and composition, verse and prose, 26 A, ἐν δὲ ὀξεῖ καὶ βαρεῖ (the tones of music) καὶ ταχεῖ καὶ βραδεῖ, ἀπείροις οὖσιν, ἆρ᾽ οὐ ταὐτὰ ἐγγιγνόμενα ταῦτα (τὸ πέρας καὶ τὸ ἄπειρον) ἅμα πέρας τε ἀπειργάσατο καὶ μουσικὴν συμπάσαν τελεώτατα ξυνεστήσατο; From him Aristotle undoubtedly borrowed his conception of rhythm, as he did likewise his grand division of ὕλη, the informis materia, the potential, unenergized matter, the material cause of all things; and λόγος, the formal cause, that which gives form and substance to the brute matter, energizes or realizes it into complete existence, and is the original design, or conception in the mind of the Creator, the ‘what it was to be’, τὸ τί ἦν εἶναι: and also his doctrine of the ‘mean’. With ἄγνωστον τὸ ἄπειρον, compare Anal. Post. A 24, [86 a 5,] ἔστι δ᾽ ᾗ μὲν ἄπειρα οὐκ ἐπιστητά, ᾗ δὲ πεπέρανται ἐπιστητά. Metaph. B 4, 999 a 27, τῶν ἀπείρων πῶς ἐνδέχεται λαβεῖν ἐπιστήμην; κ.τ.λ.

On τὸ ἄρρυθμον ἀπέραντον, compare Cic. Orator, LXVIII 228, Hanc igitur, sive compositionem sive perfectionem sive numerum vocari placet, adhibere necesse est, si ornate velis dicere, non solum, quod ait Aristoteles et Theophrastus, ne infinite feratur ut flumen oratio, seq. On ῥυθμός, μέτρον, ‘measure of time’, Ib. § 227, sonantium omnium quae metiri auribus possumus.

περαίνεται...ἀριθμῷ πάντα] This axiom is doubtless derived ultimately from the Pythagoreans, who traced the laws of the universe in numbers and mathematical symbols. Καὶ πάντα γα μὰν τὰ γιγνωσκόμενα ἀριθμὸν ἔχοντι, οὐ γὰρ οἷόν τε οὐδὲν οὔτε νοηθῆμεν οὔτε γνωσθῆμεν ἄνευ τούτω, ap. Stobaeum, Böckh, Philolaos, p. 58. “The finite in number is the calculable, that which the mind can grasp, and handle; the infinite is the incalculable, that which baffles the mind, that which refuses to reduce itself to law, and hence remains unknowable.” Grant, Essay on Ar. Ethics, p. 202 (Ist ed. [p. 252, 3rd ed.]). Probl XIX 38, ῥυθμῷ δὲ χαίρομεν διὰ τὸ γνώριμον καὶ τεταγμένον ἀριθμὸν ἔχειν, καὶ κινεῖν ἡμᾶς τεταγμένως: οἰκειοτέρα γὰρ ἡ τεταγμένη κίνησις φύσει τῆς ἀτάκτου, ὥστε καὶ κατὰ φύσιν μᾶλλον. This illustrates ἀηδὲς...τὸ ἄπειρον. With ῥυθμὸς...οὗ τὰ μέτρα τμητά, comp. Poet. IV 7, τὰ γὰρ μέτρα ὅτι μόρια τῶν ῥυθμῶν ἐστί, φανερόν: i. e. metres, verses or systems of verses, are definite lengths or sections, into which the indefinite matter of rhythm is as it were cut. Similarly it is said, III 9. 3, that the period and all metres are measured by number.

---

‘From this it may be inferred that the speech (i. e. prose composition) should have rhythm, but not metre; otherwise it will be a poem (verse-composition). Its rhythm however should not be exactly and nicely finished’: (i. e. with exact and systematic accuracy so as to be continuous, and pervade the whole structure of the writing. The description of prose rhythm by Hermogenes, περὶ ἰδεῶν ά, Introd. p. 391, Appendix on ῥυθμός, will serve as a commentary on this and μέχρι τοῦ): ‘and this will be effected if it be only carried up to a certain point (and there stop short; left incomplete and irregular; not finished and systematic, like verse)’.

---

‘Of (the three) rhythms, the heroic (hexameter, epic) is (too) stately (or solemn), and deficient in conversational harmony’. By using the word ‘harmony’, I have left it open whether we are to understand by ἁρμονία ‘harmony’ in its ordinary musical sense—in which case the meaning will be ‘that particular kind of harmony which is adapted to ordinary conversation’, the language of common life, and inferior to that of the heroic rhythm—a somewhat non-natural interpretation; or in the primary, more general sense of the word, ‘an adaptation or fitting of parts into an organized whole’, which with λεκτικῆς will signify ‘deficient in conversational structure’, in an adaptation of parts fitted for conversation (Dionysius uses ἁρμονία as equivalent to λέξις, for style of composition); the iambic is the very language of the vulgar, and therefore of all measures the iambic is most frequently uttered in common speech (or conversation); but it wants (the acquisition of, γενέσθαι) solemnity and dignity and the power (or faculty) of striking. The trochaic is too farcical (has too much of the comic dance about it; reminds one of its indecency and buffoonery²: is totally devoid of all dignity and sobriety, too light and lively): this is shown by the trochaic tetrameters, for the tetrameter is a tripping (running, rolling) measure³.

ὁ...ἡρῶος] The ‘heroic’ measure, also called ‘dactylic’, ‘hexameter’, ‘epic’, including the spondaic and anapaestic, is one of the three kinds of rhythm, its basis, βάσις—corresponding to the ‘feet’ in metre—expressing the ratio of equality 1 : 1. See further on the doctrine and ratios of rhythm, in the Appendix on that subject, Introd. p. 387, foll. where the statements of the following sections are illustrated. The epithet σεμνός has been already applied to it in III 3. 3; Dionysius, de Isocr. Iud. c. 11 (p. 557. 3, Reiske), designates it by the similar epithet μεγαλοπρεπές. Comp. Poet. XXII 9, τὸ ἡρωϊκὸν στασιμώτατον καὶ ὀγκωδέστατον τῶν μέτρων.

σεμνὸς καὶ λεκτικὸς καὶ ἁρμονίας δέομενος is the vulgata lectio. But to say that the heroic or hexameter measure—Homer's verses for instance—are deficient in harmony is absurd in itself, and contradictory to the evidence of our own ears, and all ancient authority: at all events Dionysius was not of that opinion, who says, de Comp. Verb. c. 18 (p. 109, Reiske), the exact opposite; δακτυλικὸς πάνυ ἐστὶ σεμνὸς καὶ εἰς κάλλος ἁρμονίας ἀξιολογώτατος. Victorius, from Demetrius, περὶ ἑρμηνείας § 42, read ὁ μὲν ἡρῷος σεμνὸς καὶ οὐ λογικός, which leaves ἁρμονίας δεόμενος to explain itself as it best may. I have adopted with Tyrwhitt on Poet. IV 19, ἑξάμετρα ὀλιγάκις (λέγομεν) καὶ ἐκβαίνοντες τῆς λεκτικῆς ἁρμονίας, the reading suggested by that passage, which had been already proposed by Vincentius Madius, ad loc., and since approved by Spalding ad Quint. IX 4. 76, and finally adopted by Bekker and Spengel, each in his latest ed.

ἴαμβος.. ἡ λέξις ἡ τῶν πολλῶν] This has been already noticed, III 1. 9, and twice in Poet. XXII. 19. The Latin rhetoricians make the same remark upon their own language. Cic. de Or. III 47. 182, Orat. LVI 189, magnam enim partem ex iambis nostra constat oratio, LVII 192. Quint. IX 4. 76, Illi (trimetri) minus sunt notabiles, quia hoc genus sermoni proximum est.

ἐκστῆσαι] is used here in a much milder sense than its ordinary one, to strike, excite, mettre hors de soi, to displace or remove a man out of his ordinary state of feeling, to a higher one of excitement: whereas in this metaphorical application, it usually implies a much more violent emotion than mere admiration or amusement, as Demosth. c. Mid. 537 ult., ταῦτα κινεῖ, ταῦτα ἐξίστησιν ἀνθρώπους αὐτῶν, ‘drives men besides themselves, drives them mad’. Eur. Bacch. 850, πρῶτα δ᾽ ἔκστησον φρενῶν ἐνεὶς ἐλαφρὰν λύσσαν, equivalent to ἔξω δ᾽ ἐλαύνων τοῦ φρονεῖν, in line 853.

τροχαῖος κορδακικώτερος] Cic. Orat. LVII 193, Trochaeum autem, qui est eodem spatio quo choreus, cordacem appellat (Aristoteles), quia contractio et brevitas dignitatem non habeat. Quint. IX 4. 88, herous, qui est idem dactylus, Aristoteli amplior, iambus humanior (too like the language of vulgar humanity) videatur: trochaeum ut nimis currentem (τροχερόν) damnet, eique cordacis nomen imponat. Harpocr. κορδακισμός: ὁ κόρδαξ κωμικῆς ὀρχήσεως εἶδός ἐστιν, καθάπερ φησὶν Ἀριστόξενος ἐν τῷ περὶ τῆς τραγικῆς ὀρχήσεως. Suidas κορδακίζει: αἰσχρὰ ὀρχεῖται (the rest as Harpocr.). The characteristics of the κόρδαξ, a kind of Comic dance, may be gathered from notices in Theophr. Char. 6, περὶ ἀπονοίας, ‘desperate recklessness’, where it is a mark of this character to dance the cordax sober and without a mask: in Aristophanes, who takes credit to himself, Nub. 540, for never introducing it into his comedies: in Athenaeus, XIV 28, ult. 630 E, who calls it παιγνιώδης, ‘sportive’. Dem. Olynth. II § 18 (of Philip's mode of life), εἰ δέ τις σώφρων ἢ δίκαιος ἄλλως, τὴν καθ᾽ ἡμέραν ἀκρασίαν τοῦ βίου καὶ μέθην καὶ κορδακισμοὺς οὐ δυνάμενος φέρειν κ.τ.λ. It seems therefore to have been accompanied by the grossest indecencies, so that no respectable person could allow himself even to look on the performance of it. See further in Müller, Hist. Gr. Lit. XXVII 7.

This however is not the point of the reference here. But the κόρδαξ was accompanied by verses in the trochaic tetrameter, and these are identified; and all that is implied here by the term is the lightness, the want of gravity and dignity, and the dancing tripping measure, afterwards expressed by τροχερός; as we see also in the passages of Cic. and Quint. This character always belonged to the tetrameter; and hence we are told that the dithyrambs, from which Tragedy took its rise, were originally written in this measure, which was afterwards exchanged for the iambic, the metre nearest to the language of ordinary conversation, when the dialogue had been introduced, and Tragedy assumed a regular form. Τό τε μέτρον (of Tragedy) ἐκ τετραμέτρου ἰαμβεῖον ἐγένετο: τὸ μὲν γὰρ πρῶτον τετραμέτρῳ ἐχρῶντο διὰ τὸ σατυρικὴν καὶ ὀρχηστικωτέραν εἶναι τὴν ποίησιν, λέξεως δὲ γενομένης αὐτὴ ἡ φύσις τὸ οἰκεῖον μέτρον εὗρεν: μάλιστα γὰρ λεκτικὸν τῶν μέτρων τὸ ἰαμβεῖόν ἐστιν (Poet. IV 19). Comp. Rhet. III 1. 9.

These rhythms being set aside, (they are in fact reducible to two, the proportions 1 : 1, and 2 : 1, iambus and trochee, ˘-and -˘ respectively) the third ‘the paean remains, the use of which began with Thrasymachus, though he and his followers couldn't tell what it was (did not know how to define it). The paean⁴ is the third (of the rhythms) and closely connected with the preceding: for it has the ratio of three to two (3/2 : 1, three short, and one long syllable equal to two short), whilst the others have that of one to one (dactyl, spondee, anapaest), and two to one (iambus and trochee), severally. And one and a half (3/2 : 1, the ratio of the paean) is connected with these (two) ratios [‘next to’ both ratios, i. e. the mean between the two extremes, 1 : 1 and 2 : 1], and that is the paean’. On this see Introd. Appendix on ῥυθμός, pp. 387, 8. The paeonic ratio includes also the bacchius and cretic. These three ratios are the βάσεις of the three measures.

---

‘Now all the rest (of the ῥυθμοί) are to be discarded, not only for the reasons already mentioned, but also because they are metrical (too suggestive of the cadence of regular verse): but the paean is to be adopted: for it is the only one of the rhythms named which cannot be made into a regular verse, and therefore (the use of it) is most likely to escape detection’. ἀπὸ μόνου γάρ κ.τ.λ., that is, it is an element of rhythm, not metre. Hermann, Elem. doctr. metr. II 19, de vers. Cret. (near the beginning of the chapter), has a criticism of this passage which he quotes, attributing to the author a misconception of the nature of the paeonic measure, which has caused him to fall into the error of denying it to be a metre⁵. See Cic. Orator, § 194, paean autem minime est aptus ad versum; and the whole section. Also § 218, numerus a quibusdam (Aristotle, no doubt), non pes habetur. ‘At present the one (form of) paean is employed (at the end) as well as at the beginning (of the sentence), but the end ought to be different to the beginning’. Vater proposed to supply τελευτῶντες before καὶ ἀρχόμενοι: but in a writer like Aristotle the supplement or opposite may be very well supposed to be implied in the καί.

---

‘There are two kinds of paean opposed to one another, of which the one is suitable at the beginning (of the sentence or period), as in fact it is employed: and this is the one which begins with the long (syllable), and ends with three short. Δαλογενὲς εἴτε Λυκίαν, “O Delosborn, or if perchance Lycia” (were thy birthplace). The poet, whose alternative is cut short by the inexorable brevity of the quotation, was doubtless going on, as the manner of the ancient poets is, to offer the deity whom he was addressing the choice of the various titles under which he was known and worshipped, expressive of place of birth, special character or office: which was done to avoid the possibility of giving offence by omitting any title of honour of which he might be specially proud. The following specimens of a very frequent custom will suffice to illustrate it. Hor. Carm. Sec. line 14, Lenis Ilithya... sive tu Lucina probas vocari seu Genitalis. Sat. II 6. 20, Matutine pater, seu Iane libentius audis. [We may also compare Horace's enumeration of the favourite haunts of Apollo, qui rore puro Castaliae lavit crines solutos, qui Lyciae tenet dumeta natalemque silvam Delius et Patareus Apollo. Od. III 4. 61.] Ζεύς, ὅστις ποτ᾽ ἐστίν, εἰ τόδ̓ αὐτῷ φίλον κεκλημένῳ, τοῦτό νιν προσεννέπω. Agam. 147. The author of the paean was apparently about to add after Λυκίαν, νέμων or some such word, offering the god the alternative birthplace of Lycia, if he happened to prefer it. The Homeric epithet Λυκηγένης, Il. Δ 101, 119, is usually supposed to denote his Lycian birthplace, Patara, though Müller, Dor. II 6. 8, would “rather understand” by it ‘born of light’. On the epithet Λύκειος, frequently applied to Apollo by the Tragedians, as Aesch. Suppl. 668 (with Paley's note), Sept. c. Theb. 133, Agam. 1228, Soph. Oed. R. 203 (Schneidewin), Electr. 6, &c. &c., see Müller's Dorians, II 6. 8, where the various significations of Apollo's titles are discussed at length; and Donaldson's New Cratylus § 269, on the connexion of λύκος with λευκός and -λύκη. [In G. Curtius' Greek Etymology, § 88 λευκός and ἀμφιλύκη, and § 89 λύκος, no such connexion is suggested.]

Brandis' ‘Anonymus’ [Philologus IV. I] reads “Δαλογενές”, εἶτα, “Λύκιε ἑκάεργε”.

Victorius has noted that this and the following quotation are both commencements of paeans to Apollo, from which the name of the metre is derived: and each of them exemplifies the ‘paean at the beginning’.

‘“Golden-haired Archer son of Zeus”. The other, the opposite to this, in which three short syllables form the beginning, and the long one comes at the end. “After earth and its waters, night obscured (blotted out) ocean”’. In the Greek line there are four pure paeans, all of this construction ˘˘˘-: but Ar. appears to quote it as an exemplification only of this form of paean in the last place of the verse, or rhythm.

ἐξ ἐναντίας]=ἐναντίως, or ἐναντίον, ex opposito. Polit. VIII (V) 11, 1314 a 31, ὁ δ᾽ ἕτερος σχεδὸν ἐξ ἐναντίας ἔχει τοῖς εἰρημένοις τὴν ἐπιμέλειαν. Herod. VII 225, οἱ μὲν ἐξ ἐναντίης ἐπισπόμενοι. Thucyd. IV 33, ἐξ ἐναντίας οὗτοι καθεστήκεσαν, ‘opposite’, opposed to ἐκ πλαγίου. Ep. ad Titum ii. 8, ὁ ἐξ ἐναντίας. ἐξ ἐναντίου is the more usual form. The ellipse to be supplied is according to Bos, Ellips. p. 325 (562, ed. Schäfer), χώρας, corrected to ἀρχῆς by Schäfer ad loc., q. v., where several instances of the omission of that word are produced. But the ellipse of ὁδός, in one or other of its cases, is very much more common than that of χώρα or ἀρχή, in the formation of adverbs and quasi-adverbs in the feminine, genitive, dative and accusative; such as ᾗ τῇ ταύτῃ τῇδε ἐκείνῃ ἄλλῃ et sim.—a large number of instances of these three varieties of the ellipse of ὁδός is collected under that head in the work referred to, pp. 188—192; and at p. 192 init. ἐπ᾽ ἐναντίας φέρεσθαι is rightly inserted among them by Leisner (one of the earlier editors).

‘And this makes a (true and proper) end: for’ (γάρ: the reason of this, that the long syllable is required for the end, may be inferred from the consideration that follows of the incompleteness, &c. of the short syllable) ‘the short syllable by reason of its incompleteness makes (the rhythm appear) mutilated (cut prematurely short)’. Cic. Orator, §§ 214, 215, 218, u. s.

κολοβόν] truncus, de Soph. El. 17, 176 a 40, ὅσα μὴ σαφῶς ἀλλὰ κολοβῶς ἐρωτᾶται, παρὰ τοῦτο συμβαίνει ὁ ἔλεγχος. Poste, ‘elliptical.’ For other examples see the Lexicons.

‘But the (sentence or period) should be broken off (brought abruptly to a close) and the end marked by the long syllable—not (however) by the scribe (or copyist), nor by a marginal annotation (marking the end of the sentence), but by the measure itself’. διά with the accusative, which indicates the cause or motive, (not the medium, channel or means, which is διά with genitive,) here implies that the indication of the end of the sentence should not be due to the scribe or his marks, stops, or what not, but solely to the rhythm: that the end should appear by the abrupt close of that.

παραγραφή, a by-writing, or marginal annotation. That these were occasionally stops appears from our use of the word ‘paragraph’: just as the words that we use for stops, comma, colon, period, originally represented members of the period or the whole period itself. Victorius aptly quotes, Cic. Orat. c. LXVIII § 228 (already referred to), quod ait Aristoteles et Theophrastus, ne infinite feratur ut flumen oratio, quae non aut spiritu pronunciantis aut interductu librarii, sed numero coacta debet insistere. And to the same effect de Orat. III 44. 173, where the librariorum notae are again mentioned. Victorius also cites Isocr. Antid. § 59—to the clerk of the supposed court—ἀρξάμενος ἀπὸ τῆς παραγραφῆς ἀναγνῶθι κ.τ.λ. Ernesti Lex. Tech. Gr. s. v. [In the papyrus of the Funeral Oration of Hyperides, preserved in the British Museum, and edited in fac-simile by Professor Churchill Babington, the approach of the end of a sentence is indicated by a short interlinear dash below the first word of the line in which the sentence is about to close.]

---

‘So this subject, that the composition should be rhythmical, and not altogether without rhythm, and what rhythms, and how constructed, make style rhythmical, is finished and done with’.