صور الصفحة
PDF
النشر الإلكتروني

RESTORATIONS AND EMENDATIONS IN LIVY VI.-X.

(Cf. Classical Quarterly IV. (1910), p. 267; V. p. 1.)

Continued.

BOOK IX.

IX. 6. 12. (The young nobles of Capua describe the bearing of the Romans released from the Caudine Forks after having passed under the yoke.)

iacere indolem illam Romanam ablatosque cum armis animos; non reddere salutem, non salutantibus dare responsum, non hiscere quemquam prae metu potuisse.

So MPFUþOTDLA, but F2O add a punct after salutem, and TDA write the following non with a capital; by this time (pp. 5 sq., 13, sup.) we know what these things suggest.

salutantibus dare responsum means nothing whatever more than salutem reddere, cf. e.g. 7. 5. 4 salute accepta redditaque, so also 3. 26. 9. Hence Madvig puts a comma and the non after salutantibus, which avoids the tautology more in appearance than reality and does not explain why the non was misplaced. The whole clause (non . . . responsum) should be deleted as a gloss, of unimpeachable veracity, to non reddere salutem.

IX. 9. 17. Nihil ergo uobis nec nobiscum est quibus nihil mandastis nec cum Samnitibus cum quibus nihil egistis.

So the clauses stand, rightly in PFUþOT. But in DLA the second member nec cum . . . egistis is completely lost. egistis is completely lost. In M it is omitted, but our friendly his (p. 3 sup.) stands in its place, and far above in the margin hà introduced the omitted words. The corrector in A (A2) following M (as always) has been less careful; while giving the words correctly he indicates that they are to be inserted after ergo uobis. Now if we had only DL and the (uncorrected) A we should have lost half the sentence; if we had only A2 we should have had the two nec- clauses, but in what, as it is, we know to be the wrong order. We commend this simple case, in which every step of the corruption is patent, to any reader who doubts the likelihood of such permutations.

IX. 11. 10. (The Samnite C. Pontius repudiates the surrender of Postumius.)

Ego uero istos quos dedi simulatis nec accipio nec dedi arbitror nec moror quo minus in ciuitatem oblactam sponsione commissa iratis omnibus eluditur numen redeant.

dis quorum

The curious form oblactam is given by MP (obluctam or obuictam O; obluctam TDLA; oblitam Up; obligatam F3; the -g- seems to have come from F himself), and by Harl. 2 (which Zingerle confounds1 with the H. of Books I.-VIII.-an authority five hundred years older) and a late corrector in A: obruptum by dett. aliq. The best conjecture yet offered is that of F3, or Jac. Gronov's obstrictam; but why was either corrupted? We think that conuictam may be right in form as it is in meaning; cf. Cic. de Dom. § 145 hanc ego deuotionem capitis mei, cum ero in meas sedes restitutus, tum denique conuictam et commissam putabo. It is a much commoner construction to attach conuictus to the defaulter than to the pledge which he has taken. as Cicero here does; if our conjecture be right, the combination of conuictus and commissus in both passages may suggest either that it was a standing phrase, or that Livy had Cicero's sentence in his mind.2

But how did conuic- become obluc-? By the use of a symbol whose occurrence in our Nicomachean archetype we have many times observed (see our note on 5. 43. I, and add to its nine examples now further 9. 23. 12; 10. 31. 12; and perhaps 9. 15. 9), and which one of us [in Class. Quart. II. (1908), p. 210 footnote] proposed to call the siglum Floriacense-namely, a vertical instead of a horizontal line above a vowel to denote a nasal. In

minuscule hand (and possibly in majuscule3) có may easily be read as ob-; indeed it was the likeliest interpretation to a scribe unfamiliar with the siglum. O interpreted or rather copied the word as something between obuictam and obluctam, the rest did their best after their several manners. In 10. 46. I F has cō- for ob- ; cf. also 29. 26. 7 and 29. 27. 6 where con- and oc- are twice confused; and 21. 8. 2 obortum in A for coortum in CMD.

IX. 18. 11. Miremur si cum ex hac parte saecula plura numerentur quam ex illa anni, plus in tam longo spatio quam in tredecim annorum aetate fortuna uariauerit? Quin tu hominis cum homine et ducis cum duce fortunam cum fortuna confers? Quot Romanos duces nominem quibus nunquam aduersa fortuna fuit? Paginas in annalibus . . . percurrere licet consulum dictatorumque quorum nec uirtutis nec fortunae ullo die populum Romanum paenituit.

So PFUPOTDLA, except that in the second sentence F has duces instead of ducis and T has fortuna instead of fortunam. Thanks to the occurrence of the word fortuna before uariauerit also, M has unluckily dropped out this sentence save the word confers. The missing words contain just fifty-four letters, and no doubt contained three lines of the uncial archetype; the 3 Where the b commonly appears with a very small upper half.

1 See our Preface to Books I.-V., p. vi foot

note.

2 See Addendum I. on p. 104.

line before them ended with aetate fortuna, and the last of the three lost was clearly fortunam cum fortuna.

The current method of dealing with the confusion is to excise cum fortuna as the addition of a stupid scribe-no doubt a possible hypothesis. So Madvig (anticipated, according to Zingerle, by Benedict). But this is to overlook a difficulty which Livy found quite serious, and which his correctors prudently neglect. Who is the homo, the dux on the Roman side who is to be compared with Alexander? In c. 16. 19 Livy carefully avoids making such a claim for Papirius, though he mentions, in passing, that others did so (destinant, not destinarem). But, by neglecting this, the whole point of Livy's boyish yet still thoughtful deliberatio1 is lost, and that though Livy himself expresses it quite clearly in the sentences immediately preceding and following this comparison and printed above. Granted that no one man was the equal of Alexander; but could the latter alone have met all the company of great commanders whom Rome could have sent against him in any series of years? Quot Romanos duces is Livy's argument; and it is surely absurd in face of this to make him speak as if he had some unique, but quite unnamed, general to set against Alexander.

As usual the MSS. themselves give us the clue. Read duces with F instead of ducis,2 and alter hominis to homines, as Weissenborn suggested long ago. The et might perhaps go too, as being an attempt to introduce some balance into the sentence when once the accusatives had been corrupted into genitives. But it may be defended as coupling homines and duces, the more personal side (character and generalship) in contrast to the fortuna which, though a possession, was personal in a less degree, or at least in another way; indeed it may be that the presence of et here, but not before fortunam, was one of the factors which led to the change to the genitive, if it was a conscious change at all. The contrast between fortuna and uirtus appears in the next sentence, and is a strong reason against restricting the scope of the sentence under discussion to fortuna alone, as is done by the current text.

The sentence then becomes 'Quin tu homines cum homine, duces cum duce, fortunam cum fortuna confers?' But if anyone prefers to keep et before duces we shall not quarrel with him.

quantus maximus and quantus maxime.

IX. 24. 9 decurrit inde quanto maxime poterat cum tumultu clamitans.

' ad arma'

[ocr errors]

IX. 10. 10 quanta maxime poterat ui perculit.

X. 40. 8 quanto maxime posset moto puluere3 ostendere.

XXI. 41. 4 regressus ad nauis quanta maxime potui celeritate tanto maris terrarumque circuitu.

XXIV. 35. 5 quantae maxime possent peditum equitumque copiae in Siciliam traicerentur.

1 On the character of the digression in cc. 18, 19, see App. If in Prof. Anderson's edition. Our own study confirms his conclusion,

2 The confusion of is and -es in any part of

the third declension needs no illustration.

Or as we read motu pulueris se.

In all these five passages Madvig changes the adverb maxime of the MSS. to the adjective maximo, maxima, or maximae, on the authority of a note of Zumpt on Curtius 4. 34. 10 (= Drakenborch's citation 4. 8. 10) quanta maxima celeritate potuit. (We have omitted commas for reasons that will appear below.) Zumpt's note is possibly based on Drakenborch's notes on 7. 9. 8 and 9. 10. 10; in the former Drakenborch says 'uel etiam legi posset (sc. unless the MSS. were against it) quanta (instead of quantum) maxima uoce potuit: cuius locutionis ex Liuio plura exempla uide infra ad lib. IX. cap. x. $10;' then giving some non-Livian passages, and amongst them the passage in Curtius .c. he says ubi uidendi Acidal. et Freinshem. nec non doctiss. Snakenb.1 ad Curtii lib. V. cap. ix. § 1. Adde etiam uiros doctos ad Quinctil. Declam. CCCXXIII. pag. 653.' At the 9. 10. 10 passage Drakenborch, after saying that quanta maxima ui poterat would be possible (as in fact some later MSS. write instead of poterat ui), adds with his usual caution: 'quod non displiceret, si aut plures aut grauioris auctoritatis codices testimonio suo comprobarent. Liuius enim saepius ita locutus est.' These words of wisdom and caution are disregarded by Madvig, and with them the grauioris auctoritatis codices in favour of some codices deteriores (as so often, especially in the third Decade) and Zumpt's note on Curtius! Even the dett. sometimes fail to support him; and even if we grant it to be possible that in 24. 35. 5 Put. may have written maxime by mistake for maximae, it must be pointed out that e for ae is not so frequent in Put. as ae for e; also that Put. has written in this very sentence quantae and copiae correctly.

If we turn now to the examples of the adjectival agreement of maximus with the noun we find two slightly different types :2

(a) 10. 39. 9 (and 10. 41. 8) quanta maxima ui posset (possent).
26. 46. 3 quanto maximo cursu poterant.

27. 43. 12 quantis maximis itineribus poterat ad collegam ducebat [Put., but poterat itineribus Pal. 3 (Luchs' V.) aßyde with Gelenius, and therefore possibly Spirensis].

30. 25. 8 quanto maximo impetu poterant; to which may be added 7. 9. 8 quantum maxima uoce potuit.

(b) 10. 29. 9 ut signo dato in transuersos quanto maximo possent impetu incurrerent.

iubet.

23. 16. 12 quanto maximo possent impetu in hostem erumpere

27. 43. 12 if the (possibly Spirensian) reading of Gelenius be adopted (see above).

24. 35. 5 if Madvig be right in changing quantae maxime ... copiae of Put. to quantae maximae . . . copiae.

28. 1. 6 Silanus quantis maximis poterat itineribusimpediebant autem et asperitates uiarum et angustiae saltibus crebris, ut 2 See Addendum III, on p. 104.

1 See Addendum II. on p. 104.

pleraque Hispaniae sunt, inclusae-tamen non solum nuntios sed etiam famam aduentus sui praegressus, ducibus indidem ex Celtiberia transfugis ad hostem peruenit.

We find then that (1) most often the noun is next to maximus, and therefore clearly within the relative-clause, (2) less often it follows possum, and may be regarded as either within or outside the relative-clause according to 'the balance of the clauses' in the period. The principle is particularly clear in the example from Book XXVIII., thanks to the long parenthesis. Drakenborch, we think, agrees with this view, for he regularly puts the comma after the noun1; from Madvig's commas we can hardly with certainty discover anything except 'the golden rule' that all relative-clauses must be always flanked with commas, so that in his text the noun is always included between this pair of policemen, whether justly or not.

[ocr errors]

Applying this principle of 'the balance of the clauses' to the examples of maxime, we consider that at 9. 10. 10 quanta maxime poterat ui perculit the balance rather attaches ui to perculit; if with Drakenborch and Madvig we threw ui into the relative-clause, we feel that we should be inclined by our examples to write maxima with Madvig. At 9. 24. 9 Madvig (writing maximo) throws the cum tumultu into the relative-clause; Drakenborch punctuates thus: decurrit inde, quanto poterat, cum tumultu, ad arma, etc. Here the order in relation to the possum is nearer type (b) above, but the noun is separated further from the maximus or maxime by the cum; also the main verb decurrit precedes the relative-clause, but part of the main clause follows after cum tumultu: it seems best not to obscure the manuscript evidence here. Not unlike this in the precession of the verb is 21. 41. 4 regressus quanta maxime (Put. is represented by CM) potui celeritate, but here the balance and sense require a pause or punctuation after celeritate, if not also a change of reading (with Madvig) to maxima. Finally at 24. 35. 5 Madvig punctuates thus: quantae maxime (but writes maximae) possent, peditum equitumque copiae in Siciliam traicerentur; here we should consider that either his reading or his punctuation must be wrong, and that as no pause is reasonably possible after copiae, which is the subject of the sentence followed directly by the predicate, we conclude that Put. is right and Madvig wrong in altering that text. We think therefore the evidence shows that (1) where the adverb maxime is used, the verb possum immediately follows it, separating the noun from the relative clause; (2) where the noun is indisputably included in the relative clause, the adjective maximus is used in agreement with it; (3) the adjective is also used when the noun follows possum, but that in the majority of cases it is thrown by the balance of the clauses into connexion with the words that precede it. So far therefore as the text is concerned we make no alterations3 in the readings of the MSS.

1 See Addendum IV. on p. 105.

2 There is an exception in 37. 59. I.

3 See Addendum V. on p. 105.

« السابقةمتابعة »