Fourth stasimon. Strophe 1556—1567= antistr. 1568—1578. See Metrical Analysis.—

εἰ θέμις ἐστι : a propitiatory address, since Pluto and the other

χθόνιοι θεοί

are stern to human prayers. So Hades is

δίχα παιάνων

( Eur. I. T. 185 ),

ἀμείλιχος ἠδ’ ἀδάμαστος

( Il. 9.158 ). Hor. Carm. 2. 14. 5

Non si tricenis, quotquot eunt dies, Amice, places illacrimabilem Plutona tauris.

τὰν ἀφανῆ θεὸν , Persephone ( 1548 ), an unusual title, perhaps suggested by the literal sense of

Ἅιδης

: cp. Pind. fr. 207

Ταρτάρου πυθμὴν πιέζει σ’ ἀφανοῦς

: Aesch. Th. 859

τὰν ἀνάλιον | πάνδοκον εἰς ἀφανῆ τε χέρσον

(the nether-world).

ἐννυχίων ἄναξ ( Tr. 501

τὸν ἔννυχον Ἅιδαν

), suggested by Il. 20.61

ἄναξ ἐνέρων Ἀϊδωνεύς

. This poetically lengthened form of

Ἅιδης

(trisyllabic only here) occurs also Il. 5.190

Ἀϊδωνῆϊ προϊάψειν

, Hes. Theog. 913 , and oft. in later poets. A stream of the Troad on Mt Ida was called

Ἀϊδωνεύς

from its disappearing into the ground, Paus. 10.12.3 f.

λίσσομαι =

ἐξ ἄντρων

in the antistrophe (1571): but, since the first syll. of

ἄντρων

is

"irrational,"

i.e. a long syllable doing duty for a short, the normal choree

λισσο

is defensible. (See Metr. Analysis.) Dindorf's λίσσωμαι , which he calls a

"modestior subiunctivi usus"

( i.e.

"permit me to pray"

), will certainly not stand. The schol. had in his text δίδου μοι ,—not instead of λίσσομαι , but (as his words show) in addition to it. As the construction of

λίσσομαι

was clear enough,

δίδου μοι

would rather seem to have been a gloss on some other imperat. with

μοι

. Possibly νεῦσόν μοι (cp. Ph. 484

νεῦσον,... πείσθητι

: Pind. P. 1. 71

νεῦσον, Κρονίων

) which may have been current as a v.l. for λίσσομαι .

L gives μήτ’ ἐπιπόνω ¹ μήτ’ ἐπιβαρυαχεῖ. While this verse is certainly corrupt, the antistrophic verse (1572) seems to afford firmer ground in the words φύλακα παρ’ Ἅιδᾳ . These three words, at least, have every appearance of genuineness; and they tally metrically with ἐπὶ βαρυα—χεῖ . That the latter is Doric for is confirmed by the fact that the schol. actually uses the latter form in his note; cp. 1663 . Brunck's ἀδάματον (for MS. ) being certain in 1572, the question then is:—How are the words μήτ’ ἐπιπόνω μήτ’ to be so corrected that they shall metrically answer to ἀδάματο—ν ? The absence of the ι subscript agrees with the hypothesis of an original μητ’ ἐπιπόνως . If, with Wecklein, we regard this as having been a gloss on a genuine ἄπονα (adv. neut. pl., 319), and read ἄπονα μηδ’ ἐπὶ βαρυαχεῖ , an exact correspondence is obtained, without further change in the strophe, and without any change in the antistrophe. The sense is also clear.—In the Appendix other views are given.

ἐπὶ βαρυαχεῖ … μόρῳ : for the prep. (=

"with"

) see on 1554

. This prayer to Pluto needed the preface

εἰ θέμις

(1556), since he

στεναγμοῖς καὶ γόοις πλουτίζεται

( O. T. 30 ). Cp. Aesch. Th. 915

δόμων μάλ’ ἀχὰν ἐς οὖς προπέμπει

δαϊκτὴρ γόος

.

ἐξανύσαι , reach: Ai. 607

ἀνύσειν... Αιδαν

: Ant. 804

τὸν παγκοίτην...θάλαμον

παγκευθῆ , as Hades is

πάνδοκος

(n. 1556),

πολυδέγμων

( H. Hymn. 5. 31),

παγκοίτας

( Ant. 810 ),

πολύκοινος

( Ai. 1193 ).

πλάκα (1577, 1681), a plain: cp. the lugentes campi of Vergil's Inferno ( Aen. 6. 441 ).

The traditional text, , is usu. understood;

In this there are two difficulties. (1) ἱκνουμένων is thus the partic. of the imperf. ,= . But manifestly the partic. ought here to have a pres. sense, When the pres. partic. (or inf.) serves for the impf., there is usually something in the context which prevents too great ambiguity, as is the case in 1587 ( ), and O. T. 835

τοῦ παρόντος

, where see n. Cp. Xen. Anab. 5.8.1

Ξενοφῶντος κατηγορησάν τινες φάσκοντες παίεσθαι

(= ) (= ) . Dem. or. 20 § 119

ταῦτα αὐτοί τε ποιεῖτε...καὶ τοὺς προγόνους ὀργίζεσθε ἐὰν μή τις φῇ ποιεῖν

(= ).

(2) μάταν is strange in the sense as= Another proposed version, seems inadmissible. Nor can the sense be ( temere ). Hence there is ground for suspecting

καὶ μάταν . Buecheler's

πολλῶν γὰρ ἀνταλλαγὰν πημάτων

is brilliant, though the word occurs only in glossaries, and is slightly prosaic, as properly meaning

"barter."

But it leaves the blot ἱκνουμένων . I would suggest ἱκνούμενον , and, for ἂν καὶ μάταν , αὖ ² τέρματ’ ἂν:

" now that he is coming to the goal of many sorrows. "

So the pl. El. 686

δρόμου...τὰ τέρματα

. A doubled ἄν would not be unsuitable here, as expressing earnest hope; but αὖ , which MSS. often confuse with

ἄν

, would well mark the turning-point: and for its combination with πάλιν cp. 1418.—Another possibility would be

πολλῶν γὰρ ἂν ἀλλαγὰν πημάτων ἱκνούμενον

(cp. O. T. 1206

ἀλλαγᾷ βίου

), or

ἱκνουμέναν

as=

"due"

: so oft.

ὁ ἱκνούμενος χρόνος, ἡλικία

, etc.

The MS. σε is possible; but Reiske's σφε has very strong probability. Campbell says, : but would it be so frigid? Changes to, and from, apostrophe are certainly not rare in choral odes; but this would (to my mind) be a somewhat harsh example;—quite different, e.g. , from O. T. 1201 , where see n. on 1197.

αὔξοι ,

"uplift,"

raise to honour: cp. O. T. 1092 (n.), Tr. 116 (n. on 1453 f.).

χθόνιαι θεαί : schol.

Ἐρινύες

. Hardly Demeter and Persephone (683), who would not be thus associated with the fell Cerberus.

σῶμά τ’ : the periphrasis suggests a more vivid image of the dread monster: cp. Tr. 508

φάσμα ταύρου

: Verg. Aen. 6. 289

et forma tricorporis umbrae

(Geryon). Eur. Phoen. 1508

Σφιγγὸς ἀοιδοῦ σῶμα

: Her. Fur. 24

τρισώματον κύνα.

ἀνικάτου is sound, since the long penult. (=

θε

of

θεόν

1556) is an

"irrational"

syllable. Meineke's ἀμαιμάκου is an unexampled form of

ἀμαιμάκετος

. Cp. Tr. 1097

τόν θ’ ὑπὸ χθονὸς

Ἅιδου τρίκρανον σκύλακ’, ἀπρόσμαχον τέρας

. Homer mentions

"the dog of Hades"

only in reference to Eurystheus sending Heracles

ἐξ Ἐρέβευς ἄξοντα κύνα στυγεροῦ Ἀΐδαο

( Il. 8.368 , Od. 11.625 ). The name Cerberus occurs first in Hes. Th. 311 , where he is the offspring of Typhaon and Echidna, and has fifty heads: Horace makes him

centiceps,

Carm. 2. 13. 34 .

Κερβέριοι

was used (at least in comedy) as=

Κιμμέριοι

, but the connection with

ἔρεβος

is doubtful.

It seems clear that the φασὶ after πύλαισι in the MSS. is an interpolated gloss on λόγος ἔχει . If φασὶ were genuine, it must go with εὐνᾶσθαι only, κνυζεῖσθαι depending on λόγος ἔχει : a construction awkward beyond example. Bellermann seeks to avoid this by pointing thus, etc., supplying with φασί : but this is even worse. The long delay of λόγος ἔχει brought in the gloss.—In the MS. πολυξέστοις the long penult.= of (1559). Even if we assume an syllable ( for ), the sense seems fatal. is surely not a good epithet for the Gates of Hell.

πολυξένοις appears certain: cp. Aesch. Suppl. 157

τὸν πολυξενώτατον

Ζῆνα τῶν κεκμηκότων

ἱξόμεσθα σὺν κλάδοις

ἀρτάναις θανοῦσαι

. See above on

παγκευθῆ

(n. 1561 ff.), and cp. Ant. 893 in n. on 1548.

While κνυζᾶσθαι is the form recommended by the analogy of like words for the sounds of animals ( , etc.), κνυζεῖσθαι has L's support, and also seems better just after εὐνᾶσθαι . If right here, it is, however, much the rarer form of the two.

ἐξ ἄντρων : Verg. Aen. 6. 417

Cerberus haec ingens latratu regna trifauci Personat, adverso recubans immanis in antro. — Immania terga resolvit Fusus humi, totoque ingens extenditur antro.

φύλακα . Hes. Th. 767

ἔνθα θεοῦ χθονίου πρόσθεν δόμοι ἠχήεντες …

ἑστᾶσιν· δεινὸς δὲ κύων προπάροιθε φυλάσσει

. He fawns on those who enter:

ἐξελθεῖν δ’ οὐκ αὖτις ἐᾷ πάλιν, ἀλλὰ δοκεύων

ἐσθίει ὅν κε λάβῃσι πυλέων ἔκτοσθεν ἰόντα.

λόγος … ἔχει , transitive, like Pind. P. 1. 96

ἐχθρὰ Φάλαριν κατέχει...φάτις

, rather than intransitive like

ὁ λόγος κατέχει

(

"the report prevails that..."

) Thuc. 1.10 § 2 . Cp. Paus. 9. 3 § 9

μαντεύεσθαι δὲ τὰς νύμφας τὸ ἀρχαῖον αὐτόθι ἔχει λόγος

.

τόν (as relat.) is more probable than ὅν after the vowel: cp. O. T. 199

ἔρχεται· | τόν, ὦ τᾶν πυρφόρων.

Γᾶς παῖ . This cannot mean Pluto, who was the son of Cronus and Rhea; nor Cerberus (usu. called son of Typhaon and Echidna), unless with Nauck we change τόν to δός . Thanatos is not elsewhere thus described, (in Hes. Th. 211 he is the son of

Νύξ

, no father being named,)—but is probably meant here. The invocation in 1578 is certainly addressed to him.

The MSS. have ἐν καθαρῷ βῆναι . as he passes to Hades. must mean strictly, i.e.to pass to ground which he will not traverse, leaving his path clear. ἐν καθαρῷ is thus virtually equivalent to . Madvig's ἐκ καθαροῦ is proleptic:— Cp. Pind. Ol. 6.23

κελεύθῳ τ’ ἐν καθαρᾷ

βάσομεν ὄκχον

: Her. 1.202

³ ῥέει διὰ καθαροῦ

(through an open country, where its course is not checked). So Il. 8.491

ἐν καθαρῷ

, in a clear space. I suspect the text to be unsound, but the data are inadequate for its certain correction. Two views are possible. (1) τόν in 1574 may be corrupt. If ( e.g. ) Hartung's τόδ’ were read, the sense would be:— Such a view best suits the natural sense of . (2) τόν may be sound, while ἐν καθαρῷ may have supplanted something like . Or βῆναι may have come ( e.g. ) from :

πλάκας : see on 1564

.

τὸν αἰένυπνον , Death, the giver of the

ἀτέρμονα νήγρετον ὕπνον

(Moschus 3. 105): in contrast with his brother who

λύει πεδήσας

( Ai. 676 ).