صور الصفحة
النشر الإلكتروني

all this. We cannot judge the effect of παῖσον, εἰ σθένεις, διπλῆν unless we have appreciated the full force of line 14II, ἀλλ' οὐκ ἐκ σέθεν ᾠκτείρεθ ̓ οὗτος.

Electra is ruthless here. But it is tragic that she is so. Sophocles knows that his audience thrills to the tragic situation. He has no need to dwell on the fact that matricide is terrible. The audience knows it. The audience does not simply take the view of the chorus . . . ovd' exw téyew. Precisely because the horror of matricide is so appalling, it is tragic that Electra, thinking only of her brother and her father, does not at this supreme moment feel it. And Orestes, fresh from the killing, thinks first of Electra (1426), as Electra thinks of him.

The tragedy, I submit, is accomplished. The very coolness, the calculated cruelty with which Aegisthus is received and trapped, makes his death 'a parergon.' But Electra, when it comes to the killing, calls for haste. She wants to get the whole episode dismissed as if it were an episode. She wants Aegisthus to be despatched and put out of sight. She wants to forget, not Aegisthus, but all the past.

We know she cannot. After line 1490 she does not speak. But while Orestes makes his final moral, and while the chorus raise their song of triumph, Electra still stands before us a tragic, not, I venture to think, a cheerful spectator.

I venture to suggest that if, at the conclusion, the chorus had uttered a bewildered cry instead of a clear shout of triumph, our own realization of the tragedy would be less poignant. Electra has no qualms at the supreme moment of her tragedy? Well, she is a loving and a lovable person; her instincts are womanly. Is it not tragic that such a woman should be found crying παῖσον, εἰ σθένεις, διπλῆν ? And, after all, does not εἰ σθένειs imply something more than is to be expressed by loud and complicated lyrical lamentations about her feelings at that dreadful moment?




THE fabulist Phaedrus, or Phaeder if that was his real name, was unhappy in his life, and ill-fortune pursued him after death. In addition to accidental corruption and the interpolation which this provoked, his five books of verse have suffered from disruption and partial conversion into prose. Restoration of the original, so far as that original is capable of being restored, is neither simple nor easy. The sources of the text are diverse and different. None of them can be trusted, but none on the other hand may be neglected. A short account of them is requisite in order to make what follows immediately intelligible.

The oldest and most important MS. for the bulk of the extant poems is the Pithoeanus (P), the property of the Marquis de Rosanbo, of which there is a palaeographical transcript edited by Ulysse Robert. Of the readings of a sister MS., destroyed by fire in 1774, there are reports by several witnesses which unhappily do not always agree. A small fraction of Book I. is preserved in the Scheda Petri Danielis (D). In it the verses are separated from each other as in N and V (below). But in P (as in the lost R) they are written continuously.

A number of the fables contained in PR and others not so contained, the Appendix Perottina, are extant in N, a MS. at Naples, now almost wholly illegible from damp, and therefore necessarily supplemented by V, a copy in the Vatican, first described and collated by Cardinal Mai.

Nor is this all. At some time or other after the decline of classical literature there came into being and popular currency a collection or collections of Latin prose fables, which included amongst others a number of the Fables of Phaedrus or portions of such Fables variously modified and transformed. These are the 'Fables of the Medieval Paraphrasts' as they are called.1

These paraphrasts' furnish indications here and there by which corruptions in the direct tradition may be removed or gaps therein detected and supplied. But in addition they give a certain number of fables extant neither in PR nor in the MSS. of the Appendix Perottina, which internal testimony stamps as the work of Phaedrus, and which in some instances may be restored

1 For our knowledge of these we are indebted in the first instance to L. Hervieux, Les fabulistes latins, vol. 2 (1893), and subsequently to G. Thiele,

Der illustrierte lateinische Aesop in der Handschrift des Ademar (1905) and Der lateinische Aesop des Romulus (1910).


without difficulty or improbability to their metrical form. For my present purpose, the improvement of the text of Phaedrus, this is all that need be said. It is admitted on all hands that certain vestiges and remnants of the truth as regards our author are preserved in these collections, for whose character and genesis the reader may consult the two works by G. Thiele, and especially Der lateinische Aesop des Romulus, already referred to.

It cannot be doubted that the text of the Fables, in the source or sources from which the authorities already specified have derived it, had been corrupted by incorporation of glosses and notes, whether marginal or interlinear, and by actual interpolation. Of the former we have examples at I. 22. 7, where after deuores the text of P has the intruding addition 'hic intellige soricem' (the late Latin equivalent of murem) 'esse generis masculini' after III. 1. 52, where 'frige (m) fuisse Aesopum' (a truncated comment) stood in the original of PR; after IV. 23. (24) 2 'expectatio quod ille pareret.' I. 28. 1 'homines humiles' P may be another case or a doublet arising from correction. An instructive example is found at I. 26. 4*, where the genuine marmore has been displaced by glosses patena, scutella, catino in PR and all the paraphrasts save one.1

These glosses are sometimes revealed by irreconcilable divergences in or between our authorities, as in the last cited instance and at I. 15. 1, which PR give as 'In principatu commutando ciuium saepius,' where ciuium is merely a qualification of pauperes in the following line.2 Other instances are I. 2. 28* deus PR, Iupiter or altitonans the 'paraphrasts.' Read Tonans. I. 21. 5* ad eum uenit' P, ' uenit ad eum' D, whereas the paraphrasts present 'uenit ad eum spumans' or the like, while spumans is supported by the reminiscence in Martial XI. 69. I. The ad eum is not required; for uenit without it see Verg. Ecl. X. 19 and 24 in a similar situation. App. XXVI. 1* uenatorem N(?)V, persecutorem or persequentem the paraphrasts; read persequentem or se sequentem.

[ocr errors]

Sometimes the gloss may be detected by its senselessness. So 'sine mercede' IV. 2. 8, which should be gratuito, 'poetae' IV. Epil. (V. 5.) 9*, which is a misunderstanding of cantores and 'partes' in NV IV. 25. (26) 13. Sometimes the gloss has been fused with the genuine reading. So apparently at I. 5. 7*, where 'nominor' is partly a gloss on cluo and partly a corruption of nomine hoc (Bentley). The same may have happened at V. 1. 18, where Menander appears before Demetrius, unguento delibutus, uestitu fluens ueniebat gressu delicato et languido.' This puts the tyrant in a fury, and he applies to the poet the opprobrious term cinaedus (15). But on learning who he is he changes round at once mutatis statim | "homo" inquit "fieri non potest formosior," according to the reading of all our texts. But the proper opposite to cinaedus is not formosus (cinaedi were often and no doubt usually claimed to be formosi) but fortis, or a synonym of fortis, as we

1 See Havet's note ad loc. and Thiele's in d. lat. Aesop, No. XCIII., where the MS. evidence is given in full. Compare my remarks in the current volume of Class. Phil., where also the

passages of Phaedrus, distinguished above and below by an appended asterisk, are discussed.


Saepius (by the way) is to be construed not with commutando but with the following mutant.

[ocr errors]

see from Phaedrus himself, App. VIII. 19, 'fortem uirum,' compare ib. 18 cinaedus habitu sed Mars uiribus.' Phaedrus however would not use fortior, as it failed to give him one of his doubles ententes ['inprobi iocus Phaedri ' (Martial); cf. III. 11. 5]. Hence I conjectured in the note to Dr. Gow's Corpus text that neruosior should be read, and I compared Catullus 67. 27. 'formosior' would have come from an over-written 'fortior.' neruis, if M. Havet's conjecture is right, is used in App. XI. 4 as a synonym of uiribus. One of the most frequent causes of interpolation in MSS. is Loss when detected by a scribe or a reader or reviser; and of Loss a very common species is that arising from the proximity of similar letters or words or of similar groups of these (Homoiographon). There is no lack of instances in the textual tradition of Phaedrus. I subjoin a few illustrations, enclosing omitted letters or words between the symbols <>:

I. 1. 7, 'quer<er>is'; I. 8. 2, 'quoniam <in>dignos'; I. 10. 8, sententiam' P;

II. Prol. 9, 'interpone<re>';

III. Prol. 52, 'Anacharsi<s> Scytha'; III. 7. 3, 'oc<cu>currit '; III. Epil. 8, nostrae <prae>mium' P;

[ocr errors]

IV. 1. 4, ' circum <in>'; IV. 7. 13, ' scele<re>' P;

[ocr errors]

V. Prol. 2, reddi<di>'; V. 7. 25, mo<do>reducto'; V. 9. 4, 'ante hoc noui <tu quam natus es.'1

At III. 15. II sq. P omits all from 'porro' to '<<pro>fecisset.'

An excellent restoration of M. Chauvin at App. III. I sq. is based on this consideration, 'Mercurium hospitio mulieres <uiduae> duae | illiberali et sordido receperant.' The proper framing of the anecdote requires that it should be stated that neither of the women had a husband though one of them had a child. Other passages where this principle has been or may be applied to amend the text are II. 2. 3 sq. ' aetatis mediae quendam mulier non rudis | tegebat annos celans elegantia,' 'te<ne>bat' Prasch; 'tebat' was filled up from the neighbouring celans,2 IV. 18. (19) 5, 'maxim<e> explerent famem' L. Mueller, maxim- was filled up to maximam PR. We may add III. 8. 9 sq.* '<pu>pulum' altered in PR from pulum to the facile 'filium,' IV. 18. (19) 17* <cacatus,' PR '<le>gatos' from the context.

I conclude with a pair of examples which illustrate the infidelity of both PR and NV in respect of such alterations.

In V. 1. 15 sq. the PR tradition has 'quisnam cinaedus ille in conspectu meo | audet uenire ?' and NV offer the same, with however the very obvious alteration of in conspectum meum, which L. Mueller, notwithstanding his wellgrounded distrust of Cardinal Perotti's work, has placed in the text. M. Havet, more critically, accepts the witness of PR, which convicts uenire of corruption.

1 The tu is attested by a paraphrast. Some less probably insert it after 'quam.'

2 Just as at IV. 7. 16 'infecit' was changed to 'interfecit' through the vicinity of' caede.'

For uenire however he takes the uenari of M. Chauvin-a conjecture sufficiently refuted by the inadequacy of the parallel adduced to support it. In IV. 5. 4 'unam formosam et oculis uenantem uiros,' the sense intended is obtained by the addition to the verb of an accusative of the object and an ablative of the means. Neither is present here. The expression appropriate to the author and the subject has been divined by Dr. Gow, to wit '<ce>uere.' When homoiographon had produced its effect by reducing this to uere, uenire was the obvious correction. In V. 3. 9 sqq. on the other hand the case is reversed, and editors have allowed a commonplace emendation in PR to blind them to the merits of the alternative in NV. Phaedrus wrote

sed te, contempti generis animal improbum,

quae delectaris bibere humanum sanguinem,
optem carere uel maiore incommodo.

For carere 'to be rid of' compare Ov. Her. VII. 47 sq. 'exerces pretiosa odia
et constantia magno, | si, dum me careas, est tibi uile mori,' Seneca Suasor. 6. 23
'quod nihil in salutem eius aliud illi quam si caruisset Antonio placuit.' The
sense of carere is excellent and beyond the reach of amateur correction. But
somewhere in the line of descent by which PR have come to us re was lost
after re.
And with care left standing by te both case and sense appeared to
clamour for the necare which has found its way into all our texts.

I. 3. 5 sqq. (Thiele, No. XLV.)

deinde contemnens suos

immiscuit se pauonum formoso gregi.

illi impudenti pennas eripiunt aui

fugantque rostris. male mulcatus gragulus e.q.s.

The unmetrical verse 6 in this fable of the Daw with Borrowed Plumes has been amended in various ways. M. Havet's inmiscet se seems more likely than se immiscuit or miscuit. But to account for the uit, I would suggest immiscet se ut with a comma after gregi. For ut 'when' with the historic present cf. III. 10. 27 ut sentit tonsum, gladio pectus transigit.' The postponement of the conjunction ut, like that of other conjunctions, is common enough in Phaedrus.

[ocr errors]

In 8 the rostris of PR might pass without challenge but for the indications of the paraphrasts. 'Ademar', which generally, and in this place particularly keeps very close to the words of Phaedrus, presents 'illi imprudenti pennas eripiunt aui, effugantque miserum. male mulcatus gragulus' e.q.s. miserum is of course corrupt, but not a corruption of rostris. Of what it is a corruption the other paraphrasts may show. The MSS. included under Thiele's Gallic Recension give in their expanded expression 'illi ignoto et impudenti pennas uniuersi eripiunt, calcibus et morsibus fatigant,' continuing 'et grauiter maleque sauciatus redire timuit miser ad proprium suum genus,' the Recension of W (the Wissemburg MS., now at Wolfenbüttel) has for the words in question 'morsibus autem laceratum atque semiuiuum fugauerunt. male acceptus ille

« السابقةمتابعة »