Readings in Aeschylus' Byzantine Triad Young, Douglas Greek, Roman and Byzantine Studies; Spring 1972; 13, 1; ProQuest pg. 5

Readings in Aeschylus' 
Byzantine Triad 

Douglas Young 

I N A recent issue of this journal (GRBS 12 [1971] 303-30) I suggested 
a number of interpretations or conservative emendations of the 
received text of Aeschylus' Choephoroe and Eumenides. In sequel I 

now propose several equally conservative emendations to the para­
dosis of Persae, Septem contra Thebas and Prometheus Vinctus, the so­
called Byzantine triad. l 

I. Persae 

At 97-100 the manuscript paradosis in the as trophic mesode is 
thoroughly intelligible, and may be acceptable metrically if colomet­
rized thus: 

cPLA6cPPWV yap vv--
, ,..... , R ' , 

cawovca TO 7TPWTOV 1TapaY€L fJPOTOV EtC - -vvl- -vvl-vv-
., '(J ,'" ~\(J apKVCTaTa, TO E"V OVK E"CTLV V1TE"P va- - -vvlvv- -Ivv--

100 , '\'/: ,/,. ~ 
TOV al\v!:> aVTa orVYHV. vv--Ivv-

The subject is the oo>"6p:YJTLC 'A1T<XTa (JE"OV named at the start of the 
mesode in 93. "For with a friendly attitude fawning at first she leads a 
mortal aside into places beset with nets, from which it is impossible 
that a human should flee by escaping over." 

Metrically we have a hexachronous rhythm of ionici a minore varied 
by ionici a maiore and a choriamb. Line 97 is an ionic monometer, 98 
and 99 polyschematist ionic trimeters, and 100 an ionic dimeter cata-

1 Editions of the plays are cited by the name of the editor of each. Other works of 
frequent reference are cited as follows: DALE: A. M. Dale, The Lyric Metres of Greek Drama' 
(London 1968). DAWE: R. D. Dawe, The Collation and Investigation of Manuscripts of Aeschylus 
(Cambridge 1964). DENNISTON: J. D. Denniston, The Greek Particles I (Oxford 1954). ITALm: 
G. Italie, Index Aeschyleus' (Leiden 1964). ROSE: H. J. Rose, A Commentary on the Surviving 
Plays of Aeschylus, 2 vols. (Amsterdam 1957-58). SMYTH, GG: H. Weir Smyth, Greek Grammar, 
rev. G. M. Messing (Cambridge [Mass.] 1963). TURYN: A. Turyn, The Manuscript Tradition 
of the Tragedies of Aeschylus (New York 1943). WECKLEIN: N. Wecklein ed., Aeschyli Fabulae, 
II: Appendix coniecturas virorum doctorum minus certas continens (Berlin 1885). 

S 



280 

'" 286 

6 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

lectic. But one can hardly colometrize rigorously what is to be seen 
rather as an ionie 7TVLyOC with no true KwAa. If in threnodie 
anapaests the Greek lyric tongue could cope with four short syllables 
on end when an anapaest follows a dactyl, it would find no trouble in 
99, where an ionicus a maiore is followed by an ionicus a minore. We 
have too few swatches of ionies in extant verse to formulate any <law' 
in terms of which this analysis could be denounced as illicit. Bothe 
printed the transmitted text, with no explanation, but divided the 
lines after 7TPWTOV and apK";cTaTa, which destroys the ionic basis of the 
rhythm. The mesode should be retained where the MSS have it, after 
92, as by Hermann, Mazon, Paley and Wellauer. The rhythm of the 
ionic 7TJ.'tyoc conveys well the feeling of the irresistible onset of the 
'A7T<X-ra O€ou. 

Lines 280-83 and the antistrophe 286-89 can be conservatively 
colometrized as mainly bacchio-paeono-cretic, with some molossi 
as equivalents, as allowed by A. M. Dale, p.lO!: 

" " " f3 ' tV a7TOTfwv oav u-u-I-u- sync. iambic dimeter 
8vcaLC:wfj [Upcatc. v--I--- bacchius+ molossus 
o ' ~, , 

q.OtC wc 7TaJ.'Ta 7TaYKaKWC ---I-u-v- molossus + hypo-
dochmius 

;O€cav. alat. cTpaTov cpOapEJ.'TOc. uuu-I-u-Iu- - 4th paeon+ cretic 
+ bacchius 

cTvyvaL y' 'AOavat O~otc' --u-I-u- sync. iambic dimeter 
f'€f'vfjcOaL Tot 7Tapa. u--I-u- bacchius+ cretic 
· ,\,\' n L8 ' wc 7TO ac €PC wv f'aTav - - -I-u-u- molossus + hypo-

dochmius 
" 'AO ' 8' , I 0 €KTtCaV €VVt ac 'I'J avav povc. -uu-I-u-Iu- - resolved molossus+ 

cretic+ bacchius 

At 286 there is internal correption in 8~otc. At 289 the initial choriamb 
is, in this context, a molossus with its second long resolved. Of course 
283 and 289 can both be termed syncopated iambic trimeters, it 
being allowed that a trimeter can have a choriambie anaclasis in the 
first metron, cf Cho. 1049, cpatOxLTWV€C Kat 7T€7TA€KTaV'I'Jf'EVat. But maybe 
Aeschylus wrote OEccav (283), KT{ccav (289), making initial creties. 
At 289 Italie follows Fraenkel in rejecting Wilamowitz's defence of the 
long iota in EVVtOac. Also the accent is dubious. Maybe we should 
resolve the initial diphthong and print the paradosis as €KTtCaV 



DOUGLAS YOUNG 7 

Evv'ioac ~o' avavopovc, -VVVIVVV-IV- -, first paeon+ fourth paeon+ 
bacchius. 

At 329 MSS generally offer TO£WVO' apxovTwV iJ7T€f.LvrlcfJ1]v 7T'€P£, which 
lacks a short syllable to scan as a trimeter. Perhaps the easiest solu­
tion is to assume a loss by haplography from TOtWV 8' <lxp' > apxovTwV 
v7T'€f.LvrJc81]v 7T'€P£, "Such then are the commanders I recall to mind." 
The 0' apa would be as in 568, TO' 0' apa 7T'PWT0f.L0PO£, cp£ii, A€tcpfJ€VT€C 

7T'pac avaYKav, ~€, ... Lack of caesura is found relatively more often 
in this play than in the later ones, ef Broadhead's edition, p.299. 

In 370 the epicism V7]vdv occurs in M, and in N a second hand writes 
1] above vavd. At 448 also V7]vdv is presented by M ABC II H K Nd 0 
ante corr. P linea Q Y and Ya, as I infer from Dr Dawe's data, p.316. 
These occurrences are both in messenger's speeches, which tend to be 
hospitable to epicisms, so that they may as well be left in and relished 
for their epic flavour. 

At 375 all Dr Dawe's MSS offer in the first metron of the iambic 
trimeter a choriambic anaclasis, O€L7T'VOV €7T'0PCVVOVTO, which is in 
reciprocal support with the generally accepted cpa£OX{TWV€C at Cho. 
1049. Triclinius, no lover of anomalies in metre, added a T' that is 
both needless and rather awkward. 

At 433 €ppWTa£ is offered by M on the line, after an erasure, probably 
from an original €ppWCTa£, and as a ypacp€Ta£ variant in the variorum 
codex P and the Iviron codex I, which seems to be the second best MS 

so far adduced for the triad. If the line runs alaL, KaKWV o~ 7T'€Aayoc 
€ppWTa£ f.L€ya I ll€pca£c ... , the sense would be, HAlas, a great 
sea of evils rages strongly against the Persians ... " The perfect passive 
of PWVVVf.LL, with present sense, can mean simply 'be strong', as at Eur. 
Herael. 636, Y€POVT€C €cf.L€v KOVOaf.Lwc €ppwf.L€fJa. But it also in the 
fifth century may have a more interesting and relevant figurative 
sense, implying emotional energy, 'to be eager, enthusiastic', cf 
LS1 s. v. II.2. Thuc. 2.8.4, €ppwT6 T€ 7T'ac Ka, lo£wT1Jc KaL 7T'6A£C €t T£ 

OvvaLTo KaL A6yCfJ KaL €PYCfJ gVV€7T'LAaf.Lf3av€£v aUTo'c. Lys. 13.31, OVTW 

ccfo6opa €PPWTO ~ f3ovA~ KaK6v T£ €pya'€c8a£. Plato at Symp. 176B plays 
with the two senses: KaL €T£ Evac Olof.La£ Vf.LWV aKovca£ 7T'WC €X€£ 

\ \, ~ lJ , 'A' lJ ,~~./.. ' ,~" \ 7T'pOC TO €ppWCUaL 7T'LV€£V, yauwvoc. - ovoaf.Lwc, .,..avaL, ova aVTOC 

€ppwf.LaL. In later Greek the sense 'to be healthy' prevailed and 
caused the other senses to be forgotten by copyists, as I surmise, so 
that the leetio difficilior of the better MSS, €ppWTa£, was replaced by 
the more conventional verb for a marine context, €Ppwy€V, from 



8 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

PWVVf.kL, meaning <a great sea of evils has burst'. In this substitution 
something may have been due to two common errors, the graphical 
confusion of tau and gamma, and the ear-mistake of mixing aL and €. 

At 474 the singular form cX.7T~PK€C€( v) is offered, in the phrase 
KOVK cX.~PK€C€V I ot)c 7Tp6cfl€ Mapa8c1v f3apf3&pwv cX.7TujA€c€v, by M A I 
K supra N Nd 0 P linea V Y and Va. It is a schema Pindaricum, like 49 
CT€VTaL S' i€pov Tf.kujAov 7T€A&TaL, which is kept by Kirchhoff and Mazon. 
cX.~PK€C€V seems to have been disfavoured by editors since Robortello. 

At 528 the learned have been embarrassed to find an adequate 
interpretation of the reading 7T£CTOLCL, found in I K P Q, and implied 
by the unmetrical7T£cToLc of most MSS, in the lines 

f .... ~\ ", """,~... I 
vf.kac OE XPTJ TTL TOLCO€ TOtC 7TE7TpaYf.k€VOLC 

7T£CTOLct 7T£CTa ~Vf.kcp€P€W f3ovAEvf.kaTa. 

Perhaps the best interpretation might be, "For you must, in view of 
these accomplished facts, bring together loyal counsels with loyal 
counsels," that is, "each contribute his loyal counsel to your joint 
formulation of loyal counsel." But it may be suggestive that the MS 

Ya has the reading 'TTVCTOLC, which it glosses SLSaK'TLKOLC, presumably by 
error for SLSaKToLc. LS] know the adjective 'TTVCT6c only from the 
Etymologicum Magnum and Eustathius. 7TVCTOLct in 528 would give the 
sense, "For you must, in view of the ascertainment of these accom­
plished facts, contribute, bring together, your loyal counsels." At 
Sept. 54 the true reading mJc'TLC has been corrupted to the more 
familiar, and for a Byzantine ear homophonous, 7TtC'TLC, in I supra K 
ante corr. 0 ante corr. Q post corr., and in Stobaeus. Lydia Massa 
Positano, Demetrii Triclinii in Aeschyli Persas Scholia2 [Naples 1963] 
p.47, records the grapplings of Triclinius with the passage in his 
scholia, and at p.102 his gloss, which attaches 7T£CTOLCL to his reading 
Vf.kLV, probably his own conjecture for the vf.kac or ~f.kac of the parado­
sis. His discussion may incorporate some older views. It begins: 
• -, ".I.. \ • A _._\\\ \ \" • I • A' A ~\ 7TLCTWC Wo.r€L"EV €L7T€W W\l\a 7TpOC TO ovol-'a E7TTJyay€ 7TtCTOLC. VO€LTaL O€ 

SL7TAwc, ~ • E7Tl, TOLC 7T€7Tpayl-'€Votc 7T£CTOLC', 1}TOL TOLc cpav€pwc Kat f3€f3atwc 
, .... ....,' ." C', \.... , """" 

Y€Y€VYJI-'EVOLC TqJ CTpaTqJ aTvXTJl-'actv, TJ €7T£ TOLC 7T€7TpaYI-'€VOLC 7T£CTOLC , 

1}TOL TOLC 1Tp6T€POV vcp' vl-'WV YEY€VYJI-'€VOLC 7T£CTWC 1TpOC ~I-'ac EPYOLC 0 Kat 

KP€LTTOV ••• It seems possible that the phrase TOLC cpaVEpWC Kat f3€f3atwc 

Y€Y€VYJf.k€VotC represents an interpretation of the expression I am postu­
lating, TOLC 1T€7TpaYf.k€VOLC 'TTVCTOLCL, Likewise Va's reading 'TTVCTOLC 

and gloss SL8aK'TLKoLC would not be individual efforts of Va's scribe, 



DOUGLAS YOUNG 9 

but derive from older marginal or interlinear variant and gloss 
material. 

In 532, where most MSS lack the long syllable that would make the 
line an anapaestic dimeter, 0 and Y offer c1) ZEV f3aCtAEV, vvv 'TWV 

llEpcwv. Many MSS have the 'TWvas a gloss, and Wilamowitz thought 
Q might have had 'TWV in the erasure of three letters before IIEpcwv. 

Elmsley at one time proposed to read vvv 'TWV llEpcwv, and very 
probably that was genuine paradosis. Though scribes often insert 
articles suo Marte, they equally often omit them pingui Minerva. 

At 649 there is an unusual imperatival employment of the potential 
optative with avo The best text would probably run thus: 

'A"I;' \ 1;" , \"" WWVEVC 0 ava7T'op/TrOC av Et'Y}, 

650 'Ai:SWVEVC, 
.,." A'" OLOV aVaK'Ta ,t"Japnav. 'Y}E. 

"Aidoneus (= Hades) might be upsender, Aidoneus, of the sole 
lord Dareias. Eh-eh!" 

av EL'Y} is the reading clearly intended by M's accentuation, and 
appears as a 'YpacpE'Tat variant among the scholia of the second best 
MS, I, and in P and Q. Smyth, GG §1830, remarks: "The potential 
optative with all may be used, in a sense akin to that of the imperative, 
to express a command, exhortation, or request." Pindar has an in­
stance in the third person, at Isthm. 8.49, where Themis says of Thetis 
AVot KEII XaAtvOV vcp' 7Jpwi: 7T'apfJEIICac. ava7T'olk7T'oC, formed from the verb 
ava7T'EIk7T'w, has the power to govern the accusative phrase at 651, 
otov aVaK'Ta LlapEtav. Cf Cho. 23 Xo~c 7T'P07T'OIk7T'OC. av EL'Y} was appar­
ently approved or conjectured by G. C. W. Schneider, to judge by an 
entry in Wecklein's Appendix. Triclinius also knew the reading, to 

judge by a scholion of his, published by Positano p.54, which runs in 
.,. I;' \ , ,~ \ <t t 'A"I;' \ 1;" t 'A··~ \ " " part: EL'Ta ota IkECOV EpELC 'TO 0 WWVEVC OE, 0 WWVEVC Et'Y}V av 

, , " '\ ~ Of ., "'A ~ Th h . d· f ava7T'0Ik1TOC, av'TL 'TOU Et'Y} 'T'TLKWC. e rat er In Irect type 0 

petition is closer to the expression of wish, EWE ava7T'EIk7T'ot, than to the 
direct petition, ava7T'ElktPov. In framing an appeal to Hades, of all gods, 
a certain gingerly indirection is appropriate. 

The text of 651 can be established only after consideration also of 
the antistrophic verse 656, and a little adjustment is needed in both. 
At 651 the paradosis offers oapEioll otov (or orov) aVaK'Ta oapEtall 

(variously accented). ~E. Editors commonly reject the initial oapEioll 



10 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

as a gloss or marginal variant on the form of the king's name later in 
the line. Glosses have undoubtedly been intruded at some points into 
the text of Aeschylus, even into the relatively sincere M. I pointed out 
some in GRBS 5 (1964) 94f. At 651 it might be correct to print olov 
avaKTa LlapElav. ~€. Accenting LlapElav as from a nominative LlapElac, 
like AlvElac, and taking ~€ as an ululation extra metrum, one 
could analyse the metre as Adonean+ molossus, -vv- x 1- - -, or 
choriamb if the EL diphthong be opened up. At 656 the paradosis 

" ,\ \ . I h t ~I M presents ECKEV E7TEt CTpaTOV unammous y, t en V1TEOWKEt ante corr. 
EO E7TOOWKEt M post corr. and the majority, with E7TEOWKEt in 0 post 
corr. Y Ya, and E7TOOOKEt in K, which is probably the truth or the next 
thing to it. LS] s.v. 7TOOOX€W attest the form 7TOOOK€W, doubtless psilotic 
because Ionic, as a nautical term meaning' guide a ship by means of 
the sheet', which is one of the lower corners of the sailor a 
rope attached to it. With a simple prodelision we arrive at EO '7TOOOKEL. 

Perhaps, indeed, Aeschylus never augmented his imperfect to begin 
with, cf GRBS 12 (1971) 316f. The sense would be, "since he used to 
guide the host well." For the metre, EO '7TOOOKEL would be a choriamb 
answering the molossus or choriamb 651 LlapElav. But €CKEV, E7TEL 

cTpaTov, -vv-vv, could only respond correctly to 651 olov avaKTa 
if one accepts that the final anceps of an Adonean could be 
resolved. 

Now let us reflect that in this play, at 729, we find the word cTpaToc 
as an intruded gloss upon a true reading '\aoc. For the majority 
there have '\aoc 7Tac, but the variant 7Tac cTpaToc appears in the text 
of V N Nd P, on the line. But P elsewhere restores the truth as a 
'YpacpETaL variant. Then let us turn to Pers. 279, where all MSS have in 
their texts the reading cTpaToc oa/LacfJELc ••. but a second hand in Q 
has written '\£(.oc over cTpaToc as a yp( acpETaL} variant. Which word 
would be used to gloss the other? Clearly, once one thinks about it, 
the common prose word cTpaToc is an intruded gloss, and ought to 
be replaced by Q's variant '\EdJC, in 279. Here at 656 I would read 
€CKEV, E7TE~ '\~v EO '7TOOOKEL. ~€. -vv- -I-vv-I- -, Adonean+ chor­
iamb, with the ululation added extra metrum. Italie, s.v. '\aoc ('\EWC) 
rightly gives its first meaning in Aeschylus as 'exerdtus, copiae', as 
also in the Iliad, cf LS] s.v. 1.1. I assume synizesis in '\~v. 

The epode at 672-80 can make good sense and metre with much 
less alteration of the paradosis than the current texts exhibit. The 
most conservative text and a plausible colometry might be as follows: 



675 

679A 

6798 

680 

DOUGLAS YOUNG 

,.... , ..... 
<U<H atat' 
tL 1ToAvK'Aav-rE 4>[AOLCL ()avwv, 
-r[ -raOE ovva-ra ovva-ra 

1TEpL -r~ C~ o LOVjLa ; 
0,' ayo<c> EV O· ap.ap-rLCf 
7T(xcav yav TClVO' 
Eg~cpvv-ra, 

-rpLcK'aAjLo£ va€c 
" " avaEC avaEC. 

transitional spondees 
dactylic tetrameter catalectic 
dochmius 
dochmius 
syncopated iambic dimeter 
transitional spondees 
transitional spondees 
dochmius 
dochmius (in Reizianum form) 

11 

675 Svva-ra ovva-ra Blomfield, Svva-ra Svva-ra M N2 post corr. Svvac-ra Svvac-ra 
vulgo deteriores, Svva-r' &SVva-ra Bothe. 676 -rif, cif, Schlitz alii, -ra ca vel -reX ceX 
codices fere. 677 St' ayoc Tucker, EV S' Young, SW:YOEV M ante corr. StayoEv S' M 
post corr. SLayoLEv o· codices plerique, Sta YOEV 0' I yp. 678 7Tacav yav -ravSE ut 
vid. M ante corr., 7Tacat yat -ratSE M post corr., 7TacaL yat -raSE codd. pler. 
679A Eg~cpVV'Ta£ Young, 1gecpvv-r' at M linea, EgecpOtV'T' at M supra, EgecpOtVO' at 
deteriores plerique. 

The literal sense is: "Alas! Alas! a much lamented by your friends at 
your death, how are these things possible, possible, in regard to your 
(land), twofold? [i.e. disasters by sea and land]. Because of a pollu­
tion and by means of an error three-thole-pinned ships have drained 
off all this land, (ships) that are no ships, no ships [i.e. are wrecks]." 

At 675 the corruption of ovva-ra to QVVaC'Ta would be caused partly 
by the common interchange of T and c-r in minuscules and partly by 
the latent notion that the deceased Dareios being addressed had been a 
QVVac'T'Y}c. For the ellipse of y~ with -r~ c~ cf Smyth, GG § 1027 b. 
Bothe's redivision could yield the sense, "How are these twofold 
impossibilities possible ... ?" At 677 it must be postulated that a round 
sigma had fallen out in the uncial sequence .o.lArOCEN.o.AMAPTIAI 
between other round letters, 0 and E, and through the copyist's 
familiarity with the verb oLayw. The Chorus' references to a pollution 
and an error would have been prompted by the messenger's remarks 
at 354 and 361 on the alas tor and the trick of the Hellene. The EV here 
is instrumental. At 679A Et~4>VV7'aL is the third plural perfect middle 
from itacpvw, cf Gd. 14.95, and Hesychius: Etacpvovcw- Egav7'A~covctv. 

On metrical matters, one may compare the transitional spondees at 
672,678, 679A to those at 930, alvwc alvwc E1TL yow K'EK'AL-raL. There they 
help to form a spondeo-dochmiac clausula for an anapaestic swatch, 



12 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

on which Miss Dale comments (p.54), terming it an "ambiguous 
transition." Later, at p.116 she appears to class the double-spondee 
colon as a form of dochmiac. It is possible that at 678 one should print 
what M apparently had at first, 1Tacav yav 7'eXv8€, unelided, to make a 
dochmius, - - - - x. In the as trophic epode it seems desirable to have 
the metrical cola coinciding so far as possible with the phrases of the 
speech natural to the emotional situation. On the foregoing constitu­
tion of the text the only changes from the best paradosis are addition 
of a sigma in 677 and reinterpretation of epsilon as eta in 679A. 

At 704, where the other MSS present Dareios addressing his spouse as 
7'WV €I'wv AEK7'PWV y€pata gvvvol". €V'}I€vEc yvvat, we find the variant 
8eXI'ap in the Iviron codex, I, which is the next best source for the text 
after M, even though longe secundus. But here I has the better reading. 
Which word could be used to gloss the other? Clearly yvvat is an 
instance of gloss substitution, and I's 8eXI'ap ought to be placed in the 
text. 

At 730 Atossa says, according to M, after an erasure, 1TPOC 7'eX8' WC 

Eovcwv I'Ev aeTV 1Tav K€vav8pta* C7"V€t . .. Most of Dawe's MSS have 
K€Vav8ptav; but consider K€vav8pta I Y, K€vav8ptatc A, obviously 
through dittography of sigma before C7"VEt. Blomfield, on K€VavSpta 

in some recentiores, astutely noted "nempe pro K€VavSpt(/'; and this 
dative of cause should probably be printed: cf. 295, ... K€l C7"V€tC 

KaKo'ic 0l'wc, 

At 806 the reading ... 1TE8tov 'ACW1TOC poa'ic I apSEt c/>{) .. oc. 1TtaCl'a 

Botw7'wV xOovt has superior manuscript support, from M I yp. A H 
Q Y B ante corr. 0 linea P linea ante corr. et supra. The nominative, 
stressing the kindliness of the rivergod Asopos, seems preferable to 
the neuter variant c/>D..ov, which could be taken either with 1TE8lov or 
with 1Ttacl'a. 

At 819 Dareios prophesies about Plataia, according to the usual 
t t O~ ~... .... "I "'/" ~" ex: tV€C V€KPWV OE Kat 7'pt7'OC1TOPCP YEV€' a'f'wva CTJl'avovctv ol'l'acw 
{3p07'Wv. M and A have c7Jl'a,vovctv, properispomenon, and very likely 
we should print a prophetic present, proparoxytone, c7Jl'a{vovctv. 

At 852-57 "'" 858-63 there are some difficult choices both for variants 
and for colometry. Most conservative might be the following: 

'!" ~ 1\ ' 0 ~ 
W 1T01TO','1 I'€yal\ac aya ac 7'€ 1TO-
\ 1 fl. ~, 1 I\tccovol'0V W7'ac E1T€KVpcaI'EV. 
'!'O' • .. €V 0 y€pawc 

dactylic tetrameter 
dactyliC tetrameter 
Adonean 



DOUGLAS YOUNG 

'" "~, \ , , ,..., TTpW'Ta /L€V €VOOKt/LOVC C'TpanaC aTTO-

r/JawoJ.Lf.(}', ~O€ VO/LLC/LaTa mJpjlLva 
860 TTeXV'T' ETTE(JVVOV. 

I ~" , , " , VOC'TOt 0 €K TTOI\€/LWV aTTOVOVC aTTa-
() '" '.l. I €tC EV TTpac-

'I' ,., 
COV'Tac ayov €C 0 tKOVC. 

dactylic tetrameter 
choriamb 
catalectic iambic dimeter 

dactylic tetrameter 
dactylic tetrameter 
Adonean 
dactylic tetrameter 
choriamb 
Pherecratean 

13 

854 y€patoc M I ABO P V A, JI17patoc C K Nd Q Y Va. 858 cX7Tog,aw6ftE8' 
M I linea ABC V Y Ya (cX7TO- etiam alii), cX7TEg,aw6ftE8' I supra 0 2 post corr. 
P yp. A. 860 E7TE8vvov, M post corr. K Q2 ante corr. E7TE1H}vvov M ante corr. 
Q2 yp. reliqui fere. 

The sense seems to be: ceO popoi! Truly a great and good life of civic 
government we obtained when the venerable, omnicompetent, 
unharmful, unfightable king equal to a god, Dareios, ruled the land. 
Firstly we display as proofs glorious expeditions; and towered law­
codes sped over all things. For returns from wars brought men without 
toil and without suffering, in prosperity, to their homes." 

At 854 y€patoc is better attested than YfJpatoc, for what little the 
manuscript evidence is worth on such a point, involving the mere 
interpretation of an original letter E, which, apart from context, 
might mean E, 7], or €t. The main reason for preferring y€patoc is 
metrical, that it makes an Adonean, a colon suited to the dactylic 
context, whereas €v()' 0 YfJpatoc makes the awkward sequence -v- -­

designated by Broadhead, p.293, as an 'ithyphallic syncopated',­
monstrum horrendum informe. 

At 860 an Adonean is made with €TTE()VVOV, which I take to be third­
plural imperfect from *Em-()vvw- a compound of Homer's ()vvw, 

cf II. 2.446 f3aCLA7jEc I (}vvov KptVOV'TEC. At 856 Murray made lc6()~c 
trisyllabic by synizesis. At 862 I scan the manuscripts' €V as Eil, more 
Homerico, in a generally dactylic stanza, to make a choriamb 
transitional to the non-dactylic clausula. 

As for the clausulae, prima fade we have at 857 an iambic dimeter 
catalectic, like the same colon in the epode at 906, and this is answered 
at 863 by a Pherecratean, if we keep the paradosis. It is the sort of 
irregular or anacIastic responsion that Sappho admits when making a 
glyconic equivalent to a choriambic dimeter. Cf P. Maas, Greek Metre, 



14 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

tr. H. Lloyd-Jones (Oxford 1962) p.27. Now at 863 Dawe reports that in 
M the sigma of €C is almost erased, and Porson in fact deleted €c. Per­
haps we might transpose it to make 863 run thus: -coJI'Tac €cayov dtKOVC, 
a choriambic dimeter catalectic, or Aristophanean. Then, if in 857 we 
make the licit internal correption of the €, of Llap€'oc, we see that 
Llap€toc dPX€ xwpac, scanned -vv-Iv- -, could be another Aristo­
phanean. 

At 935-40 '" 944-47 there are problems of responsion, perhaps 
solvable by a new metrical analysis, which naturally must be made in 
relation to the colometry of the whole stanza. Square brackets signify 
deletions required to be made from the paradosis. 
,..... fI~)' I ,.... , , 

'!:!'€. 00 €ywv, 01,01,. aLaKTOC. 
'\ ,- , 

j.t€I\€OC Y€VVIf! Ylf! T€ 7TaTpCtJIf! 
\ .", , 

KaKOV ap €y€voj.tav. 

935 Xo. 7TPO cp86yyov COL V6CTOV ['TC~v] 

anapaestic dim. catalectic 
anapaestic dimeter 
dochmius v~~v­
double molossus 

940 

945 

,... 
'!:!'€. 

XO. 

KaKocpanoa ~oav. KaKoj.t€'\€TOV Mv. 

MapLavovvoiJ 8p"lVYJ'Tfjpoc 7T€j.tt/JW 

[7T€j.tt/JW] 7To,\tJoaKpvv laxav. 
., . . - [ '] ,~ L€T aLaVYJ KaL 7TavovpTOV 
~ , 8 ,~ , ~, ,,,~..,. 

ovc poov avoav. oaLj.twv yap 00 av 
J , " I j.t€TaTp07TOc €7T €j.t0L. 

iicw TOL [KaL] 7TCXVOVPTOV, 

'\ao7Ta8fi T€ C€~t{wv a'\t'TV7Ta T€ ~aP"l 

2 doch. vGUGVv-1 vwwv-
2 dochmii Vv - - - -1- - - - -
dochmius vUuVvv­
anapaestic dim. catalectic 
anapaestic dimeter 
dochmius vVv~v-
double molossus 
dochmiac dimeter -GU-uv-I 

,-...,-... ,-... -vvvvvv-
7T6,\€wc y€vvac 7T€v8rrrfipoc. K'\aygw dochmiac dimeter CV - - - -I 
~. .,. , .,~ 

o av yoov apLoaKpvv. dochmius - ~~v x 

935 1TPO q,8oyyov I ABC 0 VI Y Il Ya ex corr., 7Tpoq,8oyyov M Nd, 
1Tpocq,8oyyov K P Q. COL VOCTOV 'T(XV codices fere, Tav delet Wilamowitz. 940 
1TEJLo/W semel Ya bis ceten, hoc loco semel Wilamowitz. 941 KIX~ delet Passow. 
1T(I.V8VpTOV Blomfield, 7TIXvo8vpTOV codices. 944 KIX~ delet Hartung. 7T&v8vp7'ov 
Blomfield, 7TIXvo8vp7'ov codices. 

The literal sense is: "XERXES: Here am I, Oioi!, lamentable, wretched, 
to my ancestral folk and land I became an evil truly. CHORUS: Instead 
of a speech to you for your return I shall send forth an evil-reporting 
shout, an evil-meditating cry, a Mariandynian mourner's many-teared 
yell. XERXES: Utter a prolonged all-lamenting ill-sounding cry. For 
Fortune here in turn has shifted against me. CHORUS: I shall utter 



DOUGLAS YOUNG 15 

indeed an all-lamenting (cry), paying my tribute (of mourning) for 
the folk-suffered and sea-smitten burdens of the city, mourner of the 
nation. And I shall scream out in turn a very tearful wailing." 

The kommos having begun with anapaests, Xerxes starts his first 
strophe with an anapaestic dimeter catalectic, followed by a full ana­
paestic dimeter, and then a subtly calculated transitional colon, 
933, which may be taken either as a resolved anapaestic monometer 
(i.e. proceleusmatic+ anapaest) or as a dochmius. Miss Dale (p.54) 
appreciated this "ambiguous transition." My approach to the Chorus' 
concluding parts of this strophe and antistrophe is that the transition 
to dochmiacs was exploited by Aeschylus, whereas the learned have 
generally sought to make more anapaestic cola by sundry alterations. 
In 935 WilamowitzdeletedTav, denouncing the placing of a prepositive 
at the end of a colon in catalexis. It could well be a glossing interlinear 
article that has been copied down into the line. That leaves 935 as a 
colon of two molossi, for which one may compare Soph. Trach. 653-
661. Molossi, like variants of the cretic, often occur in dochmiac con­
texts. To make a molossus at 944 one has to accept Blomfield's change 
from the manuscripts' 7Tavo8vpTOV to 7TlXV8VpTOV. This ties up with the 
same change at 941. Normally one must be suspicious of parallel 
changes in two lines, but here there is a plausible explanation. In 941 
Passow deleted the Kat in the sequence alavii Ka, 7Tavo8vp'Tov 8vcOpoov 
avDav. If the adjectives are all three attributive, as they seem to be, it 
would be quite against Greek idiom to connect the first and second but 
not the second and third. Supposing that at 941 and 944 Aeschylus had 
written the form 1TeXv8vp'Tov, a reader might well make a marginal 
note referring to the commoner form of the adjective, thus: Kat 
1TCt.v68vpTOV. A copyist taking this for a correction could well incorpor­
ate it in the text at both the nearby places. 

Line 936 can easily be analysed as anapaestic, with proceleusmatics; 
but much plastic surgery is needed to adapt the paradosis at 945 
to anapaests, whereas it is straightway intelligible as a dochmiac 
dimeter, thus: Aao7TCt.Oij T€ C€f3t,wv aAtTv7Ta T€ f3ap1J, -~-Vv-I 
- VvVu~-. To make anapaests of the rest the nineteenth-century 
metricians had to add an extra KAaytw at 947 and alter 0' a~ to DE; but 
the lines are intelligible as dochmiac if one merely ejects the repeated 
1T€p,zPW at 940, having in this some manuscript support. Miss Dale 
(p.104) suggested that "Possibly Aeschylus himself was the creator of 
dochmiac lyric." The rhythmical unit is likely to have been far older 



16 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

than Aeschylus; but he certainly exploited its multiformity with great 
freedom. On the matter of responsion Miss Dale remarks (p.112) that 
Hin all its diversity the dochmiac is a single type with variants." She 
refers also to the thirty-two species counted in Hthe Protean diversity 
of forms shown by this colarion" (p.105). Especially in a kommos it is 
grossly inept to seek for strict syllabic responsion, and indeed the 
extravagant lack of strict responsion ought to be duly relished. In 
contrast to Xerxes's anapaests the Chorus's dochmiacs are meant to 

sound frantic. 
At 961 M offers Ta:y{l&Tava Amwv, and most of the rest TCtKf1&Tava 

At7TWV, which may be scanned as a resolved hypodochmius, -vvvv-, 

to which responds 973, T&8E c' J'TTav'p0l-'at, GV v GV v-. For the 
resolved hypodochmius cf. Dale p.l05. 

At 967 the paradosis is oloto'i, 'TTOV 8l COt (/>apvovxoc; except that 0 
offers olfour times andp supra has 8~. The corresponding verse is 955, 
oloto'i, {loa Kat 'TT&vr' JK7TE080v. Murray tailors 955 to 967 by taking 
{loa as a monosyllable, which is improbable. Better would be to print 
967 as olo'i olor, 'TToii8' COt (/>apvoiixoc; Thus both lines emerge as doch­
miac dimeters. 

At 974-75", 988-89 the majority of the MSS support this presentation 
of the text: 

974 

975 

1 \ , I 
tW tW I-'0t 

, " ~ , I-'0t Tac wyvywvc KaTtOovrEc 

'" 988 ,vyy& I-'Ot 8ijT' 
, ll- t, ~ , 

989 ayauwv ETapwv V'TTOl-'tI-'VTJCKEtC. 

iambic penthemimer 

anapaestic dimeter 

Concerning the manuscripts' VrrOl-'tl-'vr}CKEtC Wilamowitz remarks 
Hvocem a tragoedia alienam," and Murray terms it Hvocem non tragicam." 
Yet at 329 they both print VrrEI-'vr}c8rJV. Among those keeping Vrrol-'tp.­
vr}CKEtC are Mazon, Broadhead, Chambry, Wecklein, Kirchhoff, Paley, 
Blomfield, Bothe, Wellauer and Pauw. At 975 the initial enclitic I-'Ot 
is in reciprocal support with the same initial enclitic at 1053. Bockh 
defended Pindar's practice of allowing an enclitic to start a new colon, 
at Nem. 4.64; Isth. 8.11. Cf. GRBS 7 (1966) 12. 

At 994 the paradosis has 8&v8r]V, apEtOV T' 'AYX&p1Jv, which can be 
scanned as the required anapaestic dimeter catalectic if one shortens 
by internal correption the Ei of apE'tov, 'martial'. 



DOUGLAS YOUNG 17 

At 1000f the paradosis can be acceptably punctuated thus: 

1001 o1Ttc8€ 8' A, o1TLc8€v 0 y, o1TLc8€V 8' ceteri. 

"I am astonished, astonished. (They are) not around (your) wheel­
driven tents [= tented waggons], but following behind." Line 1000 is 
iambo-dochmiac, uCU uCUI- - - - -, responding to 985, which should 
be printed as eAt71'€C eAt1T€C; W W oextwv (trisyllabic). 1001 scans as 
u-u-ulv-Uvuu-, iambic penthemimer+ dochmius, responding to 
986 n 1 , A \ 1 \ I I .................. , EpCatC ayavoLc KaKa 71'pOKaKa I\EyWV, - -u- - uuuvuv-. 

At 1052-53 there is no sufficient reason to reject the paradosis: 

1\ ~" !e J.LEl\atVa 0 aJ.LJ.L€J.L€W:, €Tat 
I \' 

J.LOt CTOVOEcca 71'l\aya. 

iambic dimeter 
A ristophanean 

The enclitic J.LOt at the start of a colon is supported by the paradosis at 
975. 

At 1060 M should be followed in reading 1T€1TAOV 0' ep€to€ KOA1T'tCl.V 

cXKJ.Lfi X€pwv. Dawe reports: ep€to€ MP supra, epftO' 0, epp€tO' Y, ep€tK€ 
ceteri. LSJ, s.v. ipElow I.3, attest the sense <press hard, attack'. Cf. 
Pind. Ol. 9.32, ijPEtOE IIocELoav, I ijPEtOEV oE J.LtV ••. f/Jo'if3oc. Translate: 
"And attack with the strength of your hands the folded robe on your 
chest." The deteriores' epEtKf probably comes from memory of 537ff: 
71'oAAat 0' a71'aAaLC XEpd Ka/IlJ71'TpaC I KCl.TEPftK0J.LEVCl.t I otaJ.LVOaMOLC oaKpv­

Ct KoA1T'OVC I TEYYovc' .•• 

II. Septem contra Thebas 

Colometrical theories unknown to Aeschylus have caused many of 
the learned to depart unnecessarily from the paradosis at numerous 
passages in Septem. Murray, for example, though expressing in his 
apparatus doubt whether 127-50 respond to 109-26, forced the verses 
into the shapes of a strophe and antistrophe by sundry excisions, 
additions and alterations, although Wilamowitz had seen the whole 
passage to be as trophic, like the Chorus' preceding effusion at 78-107. 
H. J. Rose, in his commentary on 78ff, remarks that the lines "do 
tend to fall into groups of about the same length and approximately 



18 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

the same metrical structure, a phenomenon not uncommon in classi­
cal verse, even when not formally strophic ... " 

Once the passage is conceded to be astrophic, a revised colome try 
can accommodate paradosis readings, impeccable for sense, that have 
long been exiled metricae theoriae gratia. Thus at 114 the manuscripts' 
yap stays if we colometrize thus: 

, "i: ~, "\ 1TCXV7'WC CXp7Jr:, ov OCXLWV CXI\WCLV. 

hypodochmius+ dochmius 
- - - - - I v - - v x 2 dochmii 
UUVV - - - I v - - v x 2 doch-

mii 
iambic trimeter catalectic 

At 120 theo€ of M'sfirst hand, which is the lectiodifficilior, would give 
a Hipponactean, as at 148: 'ApyEtoL 8E 1ToALCP.CX K&.8p.ov. If we prefer 
what may be an intruded gloss, the yap of M2 and most MSS, then we 
can scan ' ApyEtoL yap 1ToAtcp.cx Ku8p.ov, with internal correption of the 
Et:in 'ApyEtoL, as -v-I-v-I v- -, a creticobacchiac trimeter. 

At 122, 8La 01 'TOt YEvDwv t1T1rlWV, we have a dochmius followed by 
an iambic mono meter, vCU-v-l- -v-, as at 143, if we scan upsilon 
long, as at Bur. El. 1214. 

At 125 8opvccoOtC cayatc of the paradosis can stay as an iambic hexa­
syllable, a form of Miss Dale's 'long dochmiac'. Aeschylus uses the 
epic form 8opvccooc at Supp. 182 and 985. Another such hexasyllabic 
long dochmiac is offered by most MSS at 129, PVcl1T'TO"}UC YEVOV. PVcl1TOA,c, 
found in a few weaker witnesses, makes an ordinary dochmius. In 
general, throughout Septem, the paradosis shows Aeschylus exploiting 
what Dale termed «the Protean diversity of form shown by this 
colarion." She suggests that «Possibly Aeschylus himself was the 
creator of dochmiac lyric." However that may be, respect should be 
had for the less usual types evidenced in the paradosis. 

At 131 what exactly is the analysis of iXfJv{JoAcp p.aXa:vq. IlOCEto&'v ? 
Compare 791, p.~ 'TEMcrJ Kap.tfJl1Tovc 'Epwoc, and 854, &AAa yowv, ciJ 
cpD..aL. Ka'T' o~pov. Perhaps we have an Adonean, -vv- -, followed by 
an iambic pentasyllable, v-v--. But maybe one should think of the 
cola as syncopated iambic trimeters, with choriambic anaclasis in 
the first metron. 

At 135 our best witness by far, M, offers E1T£fJVvp.ov K&.8p.ov 1ToA,v 
cp";Aa:gov, an iambic trimeter catalectic, like 117, 1TUV7'WC Ilp7Jgov 8alwv 
aAwcw, except that 135 lacks caesura, as do quite many non-lyric 



DOUGLAS YOUNG 19 

trimeters in Aeschylus. The following phrase K7}8£cat T' Evapywc 
would be an ithyphallic or a syncopated iambic dimeter. The deteriores 
present the order Kaf>p-ov ETrWVVP-0Jl, which can be scanned as a doch­
mius, with epic correption of the -ov, -vv-u-, or by prod elision to 
make Kclf>p-ov 'TrWVVP-0V, - - -U-. With that colometryone makes an 
iambic trimeter cat alec tic out of 7TOAW cpvAasov K1}8£ccd T' lvapywc. 

In 141 the best and the majority of the MSS omit yap, which is dis­
pensable, with explanatory asyndeton; and the colometry of 140-44 
should perhaps proceed as follows: 

\ KJ.. ." , "\ 140 Kat VTrptc. aTE YEJlOVC 7TpOJLaTWp. aIlEVCOJl. 
'8 'c " CE EJI E~ atJLaTOC 

YEy&vaJLEv. AtTatC CE 8EOKAv-ro£c 

--uluuu-Iu- -Iu--
uu - - u - dochmius 
uCU-u-luCUu- dochmius 

+ iambic metron 
-u-I-u-I u- x cretico-bac­

chiac trimeter 

On this view 140 consists of palimbacchius+ 4th paeon+ 2 bacchii. 
At 145-46 one might read thus: 

\ , A' '" c A' -Kat CV, VKE£ ava~. VKELOC YEVOV 
C'TpaTw Satw CT&JlW 'JI aVTac. . . . 

"And you, Lord of Wolves, show yourself wolfish to the hostile host 
amid the groaning of the battle-melee." For my CT&VltJ 'v from the 
manuscripts' CT&VWV cf Agam. 431, Dobree's S&JLltJ 'v for the tradition's 
S&JLWV, with postposition and prodelision. Homer's aVT&' evolved to 
mean 'battle' for Pindar, Nem. 9.35, av KtVSvvov dgE£ae aVTae, where 
Slater's Lexicon renders it 'rush of war'. The dochmiacs are concluded 
by what Miss Dale termed" the iam bo-trochaic pentasy llable," x - u - x, 

which she conjectured to be, like the hypodochmius, an ana clastic 
form of dochmius. 

The as trophic part of the Chorus ends with the following cola: 

CV T" cL AaToYEvE£a Kovpa, 
TOgov EVTVKa~ov. 
"APTEJL£ cptAa. 

Hipponactean 
sync. iambic dimeter (or ithyphallic) 
iambic monometer, with resolution 

At 158 one should divide the paradosis' €TraAgEwv, with G. C. w. 
Schneider and Bothe, to read aKpo{36AwJI S' €7T' aAgEwv AL8ae €PXETa£, 



20 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

"the skirmishers' stone-shower comes against the fortifications." 
Hesychius probably drew from this place his entry ~E"wv' 'TE"LX€WV. 

At 161 the manuscripts' Ka, OU)(JE"V needs only proper division to 

make the passage intelligible, thus: 

Ka, 8t' (=8'ia). o(JE"V 
\ I ., 1\ • I 

7TOI\E"!-'OKpaV'TOV ayvov 'TE"I\OC. E"V !-'axaL-
, I , If '''0 · 1\ 0" 'TE" !-,aKaLp avacc. YKa. 7TpO 7TOI\E"WC 

• I \ ... ~ • I 
E"1T'Ta7TVI\OV E"UOC E"1TLPpVOV. 

"And, 0 divine lady, from whom (is) war-ending holy accomplish­
ment, and in battles a blessed queen, Onka, in defence of the city 
rescue the seven-gated abode." Onka was a Theban by-name of 
Athena, cf Sept. 501, "OYKa IIa'\'\ac. if 'T' uYXt1T'TOALC 7TvAaLa YE"L'TWV. 

Diomedes invokes Athena at Iliad 10.290 as 8'ia (JE"a. The Chorus had 
already invoked Athena, as Pallas, at 130 in their astrophic outburst; 
and at 150 and 154 they have a double invocation also of Artemis, 
another virgin goddess, appropriately to the maidens in the Chorus. 
Line 161 scans - -U-, responding to the resolved iambic metron at 
154, "AP'TE"!-,L cp{).a. 

At 212, in proposing a new emendation, (JE"O'iCLV for the (JE"o'ic of the 
paradosis, I have to take account of the whole colometry of 203-07 as 
well as of 211-15, of which Murray made the proverbial dog's break­
fast. From his text no Bentley or Housman could ever divine what 
Aeschylus wrote. The most conservative colometry seems to be as 
follows: 

l' .1.'\ O'~I I ,,~" I 203 W 'f'lI\OV 'OL7TOV 'TEKOC. E"OE"LC aKOV-
\t I NQ Na 

caca 'TOV ap!-'a'TOK'TV7TOV O'T0fJov O'T0fJov. ...-.... 1"'-"""'-"""'-"" - vv-v- vvvvvuvv 
... I "\ (; <\ I 205 on 'TE" CVpLnE"C E"Kl\aYsav E"I\''TPOXOL. ...-.... "'-""1"'-"" vvv--vv -vv-v-
< ~" I ~ \1 ~, I 
L7T7TLKWV 'T av7TVWV 7T'1}uaI\LWV ULa C'To!-,a. -v-Iv- -II-vv-Iu-vx 

7TVpLyE"VE"'Tfiv xaAwWv. 
..--. 
vvvv-u--

"0 dear offspring of Oedipus, I took fright when I heard the chariot­
rattling clatter, clatter, and the shriek made by the wheel-whirling 
nave-holes. And oh! the equine sleepless rudders in the mouth, the 
fire-born bridle-bits!" Lines 206-07 are a somewhat dithyrambic 
example of the genitive of exclamation. Others take them as depend­
ing on 203-04, aKovcaca ••• t5'To{3ov. In 205 we have two dochmiacs, of 
which the first has two resolutions, one of them in its last element, 
coming close to the form that Miss Dale (p.107) signalled as "curious," 



DOUGLAS YOUNG 21 

u- - -V0. Line 206 may be analysed as syncopated iambic dimeter, 
followed by choriambo-iambic dimeter; or as syncopated iambic 
tetrameter, with anaclasis in its third metron. Its initial cretic+ 
bacchius correspond with two cretics at the start of 214. Line 207 is an 
Aristophanean with initial resolution. 

211 aAX bTl. oaLJLoVwv 7TPOOPOJLoC .ryAeov ap­

xata j3pET'1]. 7T{CVVOC 8€OtC<LV>. vLcpaooc 

OT' DAOaC V€LcpOJLEVaC j3pOJLOC EV 7TvAaLc. 
~, "" 8 ,/,,'Q ' , \' 0'1] TOT TJP '1]V "fJo!-'cr 7TPOC JLaKapwv /\LTac, 

7TOA€WC 
'-" 

~'f' I '\' 21.5 LV V7T€P€XDt€V a/\Kav. 

...-... ...-... I ...-... -uu-uu- u--uuu 

...-... ........, I ...-... uuu--uu -uu-u-

-u-I-u-I-uu-Iu-u-

HBut as a fugitive I ran to the ancient statues of the deities, trusting in 
the gods, when there was a roaring of a murderous snowstorm snow­
ing on the gates. Then indeed I was excited by terror to supplications 
of the blessed ones, that they might extend their protection over the 
city." The assumed loss of -LV in 8€OtCLV would be all the easier before 
the initial letters of VLcPd.ooc. Exact syllabic responsion between strophe 
and antistrophe is not sought by Aeschylus in dochmiac movements, 
so that, for example, a dochmius in dodrans form, -uV-V-, may 
respond to one in hemiepes form, - uu - uu -. In cola of paired dochmii 
he frequently observes diaeresis, but often welds together the pair by 
overlap, as in 205 and 213. These freedoms were further developed 
by Euripides, cf. Dale p.lli. It seems likely that the lack of diaeresis 
and lack of exact strophic responsion contribute to the sense of agita­
tion in the dochmiacs used here by the Chorus. 

In 239 M's reading makes a perfectly acceptable pair of dochmiacs 
by Aeschylean norms: 7ToTa{v~v KMovca 7TaTayov cXJLJLLya. u- -u-I 

wuv-ux. The second iota in 7ToTa{IlLOV is consonantalized. At 288 
Wilamowitz accepted a consonantalized iota in Kapotac of the para­
dosis, which he reaccented to KapoLac. At 781 a dochmiac analysis is 
possible with Kapot~ of most of the MSS, the antistrophe at 788 having a 
dochmiac in hemiepes form. Broadhead, in a note on Pers. 1008 (at 
p.282), defends what he calls "synizesis" of iota in Aeschylus. In Sept. 
it seems possible also at 176 CPLA07TOA~C, 481 ~, 521 LJ~c, 789 O~x€P~' 

826 cW77JP{~ 948 S~c06Twv, Some cases will be discussed in their se-
'-' 

quence. The turning of iota into a glide does not depend on the quality 
of the vowel following, but is connected rather with the consonant 
preceding, and it seems specially common after delta. T. G. Tucker, in 



22 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

Appendix A to his 1908 edition of the play, discusses "consonantizing 
of v and L," and concludes (p.21O): "When we consider the number of 
examples, and also the fact that such pronunciations as T€LP€Ciac, 
• Ap.ch&p€wc, would often assist in normalizing metre, it seems highly 
probable that more account should be taken of such consonantizing 
when we are considering correspondence in lyrics." 

At 291-94 the colometry involving the least departure from the 
tradition seems to be as follows: 

~ I ~. ., I ., opal(ovra 0 wc 'TLC 7'€I(VWV V1T€P J 

8l80LI(a A€Xlwv 
8vc€vV1}7'opac. 

I \ I 1TaV7'pop.oc 1T€/\€Lac. 

u-u-I-u-u- iambic metron+ hypo-
dochmius 

u - vuu- dochmius 
dochmius 
ithyphallic (or sync. iambic dimeter) 

"And as one (fears) a snake in concern for (her) children, I fear (men 
that may be) ill bedfellows of (my) bed, (like) an all-trembling rock­
dove." The accentuation tm€P is found in the MSS that Turyn denotes 
by Ld and Le. The first singular ending of (tJ1TEP )8l80Ll(a is attested 
byMantecorr. (utvid.)K YP supraQ2 supra, and makes the expression 
of fear by the Chorus more immediate than the third singular, for 
which the scholiast supplies as subject Kap8la from 288. Any mother is 
afraid of a snake near her offspring, and not merely mother-birds, 
so that there is no need to think that 1T&vrpop.oe 1TEAELClC in 294 belongs 
closely with we 7'£( in 291. Rather it is attached in apposition to the 
unexpressed subject of 8l80LKa, without any such expression of com­
parison as we1TEp. Cf Theognis 347, £yw 8E I(VWV £1TlfYYlea Xap&8p'Y}v. 

The antistrophe at 308-11 should be cited with its preceding line 
also: 

307 v8wp 7'€ JLpl(atOv. €V- syncopated iambic dimeter 
7'pEcPEC7'a7'ov 1TWP.&7'WV oewv u-u-I-u-u- iamb. metron+ hypodochmius 
t'Y}ew IIocEL- u- -u- dochmius 

310 8m, 0 yaia6xoe -uu-u- dochmius (in dodrans form) 
T'Y}(}vos 7'€ 1Tat8€e; ithyphallic (or sync. iambic dimeter) 

At 307 the form with -7'PEcP- is in MIK 0 and Triclinius, and Wellauer 
preferred it as "magis poeticum." At 310 the ai of yaLa6xoe is shortened 
by internal correption. For the ellipse of 8l80LI(E in the we clause at 
291, cf Smyth, Greek Grammar (1963), §2464: "The verb of the com-



DOUGLAS YOUNG 23 

parative clause is commonly omitted if it is the same as the verb of 
the leading clause." 

At 345ff I would print: 

4 \ ~) , , " 3 5A KOPKOPVYCU 0 av aC-TV, -vv-I v- x Aristophanean 
345B 7TOTL 7TTOALV a' u - u - iambic monometer 
346 0pKeXva TTVpyW'rLC, 

347 A TTPOC av8poc 8' avY]p 

347B 80pt KAlv€Tat. 

-v- - - - troch. dim. brachycatalectic 
v - - v - dochmius 
uu-u- iambic monometer, resolved 

357A 
357B 
358 
359A 
359B 

481A 
481B 

482 
483 

OJ 521A 
521B 
522 

523 

The corresponding cola of the antistrophe at 357ff run thus: 
..- \ ..- \ \ TTaVTooaTTOC OE KapTTOC 

'''- \ xap,aOLC TT€CWV 
'\ A , CtAytJV€t KVp'YJCCtC, 

\ "-'" f) 7TtKPOV 0 OlLILa Ct-

AalL'YJTToAwv 

vv-v-

-u-u--

v--ux 

Aristophanean 
iambic monometer, resolved 
troch. dim. brachycatalectic 
dochmius 
iambic monometer 

HAnd tumults throughout the citadel, and against the city a net like 
a rampart; and man is laid low by man with the spear ... And every 
sort of crop falling on the ground will cause grief as it occurs [i.e. 

the spoiling of stores will grieve the spectator as it meets the eye]; 
and embittered is the gaze of the maidservants." At 346 the scholiast 

., \ f) \..- , <I \ , \,.. Ls:r d h notes: OpKCtVTJ TO 'YJpanKOV OtKTVOV, 0 Kat capyaVTJ KaA€tTat. 'J an t e 
Supplement do not note either noun in this sense. At 347B MSS have 
80p~, and KAlvETat is M' s reading on the line, found also in Turyn's 
Sh on the line. At 358 the future form aAytJv€I:, with short upsilon, is 
inB andH. 

At 481-83 and the antistrophic 521-23 the paradosis is fully intel1i~ 
gible metrically thus: 

u-u--
,..." 

uUu-u--

- CV-v-I-v-v-

v-u-x 
---. 

-uuuv-x 

---. I"""""" -vv-v- -vv-vv-

The only change needed is Brunck's 8alILovoc at 523 for the manu­
scripts' 8allLoCLv, and that is not metrically motivated. 481A and 521A 



24 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

are examples of Miss Dale's 'iambotrochaic pentasyllable' (p.108). 
In 521A we have an example of consonantal iota after a delta; cf the 
remarks on line 239. At 481B we have another consonantal iota, in lciJ, 

'-" 

as at Cho. 466, where Hermann needlessly altered to JJ. At 483 all MSS 

have the elided form {3&~ovc' except for C Il Q post corr. and 
Triclinius, which have unelided {3&~OVCLV. In 523 all have {3po'To'icl, 
which would give a dochmius in hemiepes form responding to a 
hypodochmius, or anaclastic dochmius, in 483. Granted that 
scribes often switch between such forms as -o'ic/-o'iet and -ovC/-OVCLV, 
they still more often copy what they see in their antigraph, and the 
majority variants here may well go back to the poet, and reveal 
again his interest in ringing the changes on his dochmiacs in 
responsion. 

The problem of just what spellings can go back to Aeschylus is 
raised by variants in the best MSS at 497-98, which suggest that the 
text might be Ev(}eoc 8' "ApTI I {3&KXa ••• At 497 M has aPTJ" according 
to Vitelli-Wecklein, and that appears in Naples II F 31 bis, after correc­
tion. apT] is in the second best MS, I, and in A D Rc 0 post. corr. X 
ante corr. Quite many have the accusative apTJV, and the normal dative 
form ape, is in K Q Y N post corr. and a few more. An epic form like 
"ApTI might be acceptable in a messenger's speech. Cf TpolTJv at 
Agam. 577, in the better witness there, F. Professor George P. Goold, 
in a paper on "Homer and the Alphabet" (TAPA 91 [1960] 286), 
argued that the use of the Ionic alphabet for Homer was responsible 
for its universal employment by other literary writers long before 
403 B.C., when it was finally adopted by decree at Athens. I have usu­
ally assumed that a script of Aeschylus submitted to the Archon in 
applying for a chorus would use the pre-Euclidean Attic alphabet 
without an eta. But if Goold is right the dative APHI might have 
been the author's spelling. 

At 498 the unaugmented imperfect {3&KXa appears in M and on the 
line in I, where the present form, found in the rest, is a 'Yp&q,e'TaL 
variant. The unaugmented imperfect goes well in a messenger's 
speech; cf GRBS 12 (1971) 316f for the limit of the use by tragedians of 
past tenses of the indicative without syllabic augment. Like the aorist 
E1T7J>"&>"a~ev at 497 it describes what the messenger saw of Hippo­
medon before he left to report. It is not suggested that Aeschylus 
applied an accent, but that he wrote BAKXA and some copyist later 
added an iota and made the form present. 



DOUGLAS YOUNG 25 

At 562 no letter of the majority need be changed if we write 560-62 
thus, with Eteokles referring to the Sphinx on Parthenopaios' shield: 

T0 ¢EpOIl7"£ p,Ep,I/JETat 
_ "..", t \ 1\ 

7rVKIIOV KpOTTJCfLOV TvyxallOVC V7rO 7rTOl\tJI J 

8 A 8 \ , A' '\ 8 I ,. I EWII EI\OllTWV. "" av J al\1] ~v,atJJ. ~'YW. 

"She [the Sphinx] will be angry with her bearer when she meets with 
a constant rattling close to the city, if the gods will. Zeus, may I 
prove to have spoken truly!" Copyists who interpreted as S' all had 
not read LS1 S.vv. ZEVC and ..dall. 

At 695f the minimal emendation seems to be involved in reading 
,J.. '\ ' ., \ '\ 'A' 'f'LI\OV yap aLcxpa fLOL 7TaTpoc TEI\EOC pa 

gTJPOLC aKAavcToLc ofLfLactJI 7rpOCL~aIlE£J ••• 

"For in its ugliness my own father's accomplishing Curse sits beside 
me with dry un weeping eyes ... " At Cho. 382 we find the two­
termination feminine form TEAELOL. Both TEAEOC and TEAELOC are exem­
plified in Italie's Index. Here T€AEioc could scan by internal correption 
of EL. A scholiast aware of 832, ciJ fL€AaLlIa Kat. TEAEla Y€IIEOC OlSl7rOV T' 

'Apa, could have written in the gloss TEAEla: then that had its final 
alpha illicitly elided, giving rise to the manuscripts' TEAn', variously 
accented; or copyists simply took in TEAEla with some variation in its 
accent. alcxpa is offered by the best MSS M and I and a dozen others, 
and is printed by Tucker, Verrall and Pauw. The more obvious and 
in the context hackneyed €xflpa is attested by K 0 Q and others. At 
696 the paradosis form aKAavcToLc, adding more sibilance than Butler's 
aKAavToLc, is kept by Hermann, Wellauer, Pauw, Paley, Tucker and 
Rose, and in another case by Murray at Eum. 565. 

At 766-68 and the antis trophic 772-74 the colometry and punctua­
tion of the paradosis might be made thus: 

766 T'AELa£ yctp 7raAaLcpaTWII u- -I-u-Iu-u- sync. iamb. trim. 
apat. 

f3apEcCf KaTaMayCf· 
Ta 8' aAoa 7rE>.6JLEV' au 

I 
7rapepXETaL , 

" , ~ ""'" \ I ~, 
'" 772 TLII allopwlI yap TOCOIlO 

'8 I E avp,acall 

flEOt. Kat. gVV€C7"£OL 

7roAewc 7roAvf30TOC T' 
'\ R """ a£wlI ~pOTWII J ••• 

v--Iu-u-

---- ---- ---- I uuuuuuu- u-u-

uu-uuu- --u-----.--... I 

sync. iamb. dim. 
dochmius+ iambic manom­

eter 



780 

,.., 785 

790 

26 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

«For curses of men famed of old are accomplished [or accomplishing] 
with grievous merchandising; and the ruinous deeds that are happen­
ing are not going past." ... «For whom among men did the gods so 
greatly admire, and the city's hearth-sharers, and the populous 
generation of mortals? ... " 

At 766 7Ta>..aup&Twv is in M I B ante corr., and «pat in all MSS. In 767 I 
merely reaccent {1ap£'iat KaTaAAaya2 of the paradosis. In 768 M ante 
corr. and I have 7T£A6p.£v' OV, retained by Hermann, Smyth, Tucker, 
Verrall and Paley. At 772-74 I reproduce the almost unanimous para­
dosis. In 776 the manuscripts' avafY1T~av8pav is metrically acceptable as 
a dochmius and was kept by Wellauer and Pauw, meaning 'man­
upsnatching' . 

At 778-92 the MSS need less adjustment than editors have usually 
bestowed on them, though some small points remain moot. 

• 'S" l4> £7T££ apT PWV u--u- dochmius 
, , 'A '(}A' £Y£V£TO p.£ £oc a twV -- --I uuuuuu u-u- iambic dimeter 

' , , >fA 8 cp A yap.wv. £7T a y£t vc opWV u-u-I--u- iambic dimeter 
, 8' -- dochmius p.a£vop.£vq. Kap ~ -uu---

8'8 ' , J.TI.A£c£v ------ dochmius £ vp.a KaK uuuuuuu-
cp' , A -- -- dochmius (in hemiepes 7TaTpo ovcp X£p£. TWV -uu-uu-

form) 
, 8" , " , KP££CCOT£KVWV a7T op.p.aTWV -uu-I u-u-I u-- anaclastic iamb. trim. 

J.7TAayx(}'Y/· catalectic 
' 8''''''-T£KVO£C <ap > apa£ac --u-uu--

'cpA " cpA £ 'Y/K£v £7T£KOTOVC TpO ac. u-uGUlu-u-

,"" A' " a£a£,7T£Kpoy wccovc apac. --u-I--u-

Kat ccp£ "Sapov6p.cp -- ---uu-uu-
8 ' , A ... -- --I:!;X£P~ 7TOT£ ax£w uu-uuu-

, A 8' , -- --KTTJp.aTa· vvv £ TP£W -uu-uu-

'A' p,t/J' 'E' p.'Y/ TE ECTI Ka £7TOVC p£VVC. -uv-I-v-Iv--

"But after the miserable man became fully conscious of his wretched 
marriage, falling sick through distress, with maddened heart he 
wrought twin evils with his father-slaying hand. For from his children­
surpassing eyes he went astray. And against his children indeed he 
discharged bitter-tongued curses, alas!, enraged at scanty sustenance, 
actually that they should some day be assigned possessions by a trans-



DOUGLAS YOUNG 27 

action apportioning with iron. And now I dread that the leg-bending 
Fury may accomplish (those curses)." 

In 781 KapSt~ of the paradosis is kept by Verrall, making a dochmius 
with consona"ii'tal iota; cf. 288. In 784 we have a hapax compound from 
the verb KpELCCOW+ T{KJlOJl. For the sense cf. 530 piiMov Of of} dfJfW 
7rE7rOLOWC O/-,/-,cX.TWV O' {m{pTEpov, and Catullus 3.5 quem plus illa oculis 
suis amabat. For compound adjectives having as their first part a 
verb-stem cf. Smyth, GG §880. Aeschylean examples include 
< , '" '.1.' ,~ 'Q < 'Q a/-,apTtvOOC, apXEIIE£OC, EPEL'I'LTOLXOC, EXEV[JC, 7rELC£fJPOTOC, PVCLfJW/-,OC, 

C'Tvyavwp, TEAEeclcppwv, cp€pae7rLe, cp()EpeLYEvTJe, c1JAEctoLKoe. For the metre 
of 784, perhaps it is to be interpreted as an anaclastic iambic trimeter 
catalectic, with the anaclasis in its first metron; cf. 131 and my remarks 
on it. Then 791 is a version of the same syncopated in its second metron 
as well as catalectic in its third. 

At 785 the paradosis is T€KVOLe S' apalae, and G. C. W. Schneider 
reaccented to apadic. Aeschylus is punning on the two adjectives 
apaLoe, <thin, scanty', and apa'ioe, <accursed'. Some might feel that 
metrically no change is needed, and the dochmius at 778 could be 
equivalently responded to by T€KVOLe S' apaLae, an iambo-trochaic 
pentasyllable that could be reckoned an anaclastic type of dochmius. 
If this be not so, one might produce an indubitable dochmius at 785 
by a slight supplement, e.g. T€KVOLe S' dxp' > apaLae, where the loss 
would have been by haplography, or T€KVOLC(L> 8' apaLae, where the 
loss would occur through inflectional interchange of a type perhaps 
commoner than haplography. 

At 788, for Ka~ in the sense <actually' cf. Denniston p.321. At 789 
most manuscripts have SLaX€LplaL, but some have the final iota lacking 
or subscript. Triclinius may have hit the mark with his O~XEp{~, 
scanned with its two iotas consonantalized, to make a dochmius ~f 
789, as in the strophe at 782. The word would be a hapax legomenon, 
like its adjective CLoapov6f-'l~' but consider the fifth-century oLaXElpLCLC, 
~ " " , \ 'A 786" f oLaXELpLc/-,OC, EKEX€LpLa, €YXELpLa, 7rOIlVXELpLa. t E7rLKOTOVe 0 most 
MSS and of the scholia is kept by Hermann, Smyth, Tucker, Verrall, 
Wecklein, Wellauer and Pauw. 

Lines 803-21 have been much messed up by editors, among them 
Wilamowitz and Murray, in support of the theory that the end of 
Septem in our MSS was added for a revival after the poet's death, and 
that there are other evidences in the paradosis of double recension. 
If one keeps as closely as possible to the best manuscript tradition this 



28 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

is how the verses appear (the original sequence of lines being retained, 
with numbering to suit, where Murray's numeration is sadly per­
plexing): 

803 XOP. 
,~.., ~ , '\ \ , 

TL 0 ECTL 'TTpa:YJLa VEOKOTOV 'TTOI\EL 'TTI\EOV; 

804 Arr. '\' f3 \ , ~.., 
'TTO/\LC CECWCTaL. aCLI\EWC 0 ofJ.OC'TTOPOL 

805 avSpEc TEf}vacLv £K XEPWV aVToKToVWV. 

806 XOP. , ,~..,. ,/.. -,/.. 'f3 \ , 
TWEC; T£ 0 E £'TTac; 'TTapa'f'povw 'f'o CfJ /\oyov. 

807 Arr. ,/.. - -" O'~' , 't'povovca vvv aKOVCOV. LOL'TTOV TOKOC-

808 XOP. tI,,'\ ' ., \..... .... 
OL yw Tal\awa, JLaVTLc ELfJ.£ TWV KaKWV. 

809 Arr. ,~. , ,/.. \ , , ~, 

OVo afJ.'f'LI\EKTWC fJ.TJv KaTEC'TTOoTJfJ.EVOL -

810 XOP. • ~f} ~ f} f3 ' ~, .,... ,/..' EKEL L KELC OV; apEa 0 ovv ofJ.wc 'f'pacov. 

811 Arr. .. ,~ \,/.. ~ ", ." 
OVTWC aUEI\'f'aLc XEpCW TJvaLpoVT ayav • 

812 XOP. ., .~, '~',/..A<I 
OVTWC 0 oaLfJ.wv KOLVOC 'Iv afJ.'f'0Lv afJ.a; 

813 Arr. , , ~" \ A ~ _ ~ , , 

aVTOC 0 aVal\OL OTJTa OVC'TTOTf.J.OV YEVOC. 

814 - , '~/f} I 
TOLaVTa XaLpELV KaL oaKpVEC aL'TTapa, 

815 1\ \ 1" t~" I 
'TTOI\LV fJ.EV EV 'TTpaCCOvcav, O£ 0 E'TTLCTaTaL, 

816 SLCCW CTpaTTJYW, SLlAaxov C</>VPTJAaT<f:J 

817 n 'f} ~I I , 
",KV '[J CLOTJP<f:J KTTJfJ.aTWV 'TTafJ.'TTTJCLav • 

818 ., I: ~." \ 'f3 '",,/.. ~ (J , 
E~OVC£ 0 "IV I\a WCW EV T~'[J X OVOC, 

819 , ",~ I ,/.. I 
'TTaTpOC KaT EVXac UVC'TTOTfJ.OVC 'f'OpOVfJ.EVOL. 

820 '\ I f3 \ , ~,. I 
'TTOI\LC CECWCTaL. aCLI\EOLV 0 O!-,OC'TT0POW 

821 , .. A' • , '\ \ 1\ ,/.. I 
'TTE'TTWKEV aLfJ.a yaL V'1T al\/\TJI\WV 'f'OVCfJ. 

A literal version might run thus: 

"CHORUS. What novel matter is there for the city besides? 
MESSENGER. The city has been kept safe. But a king's joint-sown males 

have died by self-slaying hands. 
CHOR. Who? What did you say? I am out of my wits through fear of 

your message. 
MESS. Keep your wits about you now and listen. Oedipus' offspring­
CHOR. Oh, how wretched I am! I am a prophetess of evils. 
MEss. Not disputably indeed reduced to the dust-
CHOR. Are they lying (dead) yonder? Grievous,-yet tell it. 
MESS. Thus they were destroyed, with hands too brotherly. 
CHOR. Thus was the deity common to both at once? 
MEss. Yes, and in person he is squandering the ill-fated family. Such 

are the happenings that one has to rejoice at and weep over­
the city indeed faring well, but the rulers, the twain army-



DOUGLAS YOUNG 29 

leaders, have shared out with hammer-forged Scythian iron 
full ownership of possessions. And they shall have whatever 
share of land they get in burial, borne away in accordance with 
their father's ill-destined prayers. The city has been kept safe. 
But earth has drunk the blood of the joint-sown pair of kings, 
amid reciprocal slaughter." 

It is apparent that we have to do with 'Ringcomposition' by Aeschy­
lus, and not with any theoretical double recension by some fourth­
century producer of a revival of the play under the influence of 
Sophocles' Antigone. 

To consider some individual readings: at 803 TTpaYJLa is in M I K 0 
and P on the line, and is kept by Smyth, Tucker, Verrall, Wellauer 
and Pauw, in preference to the majority reading TTpayoc in most of 
the deteriores. TTpayoc is the more poetic form, and occurs in 2 and 861 
of this play, whence some copyist may have taken it. In studying 
scribal errors in Pindar I noted (GRBS 6 [1965] 258) that "The scribes' 
tendency to normalize or trivialize is not uniform, and sometimes 
they slip in a more poetic or dialectally recondite form." 

At 804 f3actMwc appears in M on the line, A, P and 0 post corr. and 
is retained by Tucker. The riddling phrase "a king's joint-sown males" 
(cf. LS] S.v. &v~p I) is typical of the roundabout approach to bad news 
often manifested by messengers in tragedy and accounts for the 
questions in 805 by the initially baffled Chorusleader. Likewise, at 
807, the Messenger chooses a word of some initial ambiguity in TOKOC, 
M's original reading on the line, which is kept by Mazon, Verrall and 
Rose. In 809 the Messenger's word KaT€CTToS7JJL/.VOt has more than one 
level of meaning, including obscene or at least vulgar associations, 
as Wilamowitz noted in his Aischylos Interpretationen (Berlin 1914) 
p.86 n.5. It found favour with Smyth, Mazon, Groeneboom, Tucker, 
Paley, Wecklein, Verrall and Wellauer to leave 809 with the Mes­
senger, as the paradosis has it. 

At 810 again we are indebted to our best witness, M, on the line, 
for the dual KE'iC{}OJl, kept by Smyth, Mazon, Groeneboom, Wecklein 
and Verrall. At 811 the manuscripts' phrase &SEAcpa'ic ••• «yav was 
endorsed by Hermann, Mazon, Wellauer and Pauw, and recalls 
Aeschylus' way of thinking at 871, with the presumed coinage 
(jvcaS€AcpoTaTat. At 821 SVCTTOTJLOVC of the majority has not been uni­
versallyendorsed, and Headlam's SVCTTOTJLWC has found some favour. 



30 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

I note that it occurs in the deteriores Ya, Lc on the line, and probably 
Lh interlinear. If the strict meaning of the adjective be pressed, it is 
Oedipus who was <ill-destined' and not the prayers he voluntarily 
chose to utter, so that if any change is to be made it might well be to 
ovc1TCh/Lov, which would have been altered by the common assimila­
tion of endings. I notice the three-termination form ovc1TC5T/LaC in 
Ld ante. corr. and Le. At 804 and 821 the MSS offer the form clcwcTaL, 
which is kept by Hermann, Smyth, Mazon, Verrall, Paley and Pauw, 
and by LSj. 

Much dissension has arisen about the interpretation of the Chorus' 
question at 827f: 7j TOVC /LOYEPOVC Kal ovcoalfLovac 1 tdTlKvovct KAavcw 
1ToAE/LapxoVc; The interpretation of the letters ATEKNon as <childless' 
is most unlikely, since most people in the fifth century would know 
very well that Polyneikes' son Thersandros became king of Thebes 
when the Epigonoi took the city-and his descendants included 
Theron, an eminent patron of poets,-while Laodamas, son of 
Eteokles, withdrew to Illyria. Professor Lloyd-Jones once suggested 
reading ;KVOVC, and one might consider ~T' (Doric, = -aTE) ;KVOVC, taken 
as nominative singular, with the sense: HOr am I, as one demented, to 
bewail the distressful and ill-fated war-leaders?" On this redivision 
and interpretation verse 828 is an acatalectic dimeter. Alternatively, 
recalling the use of alpha-dOpOLCTLK6v in compound adjectives such as 
u1TA6oc and uOp6oc, one wonders whether Aeschylus did not play with 
the formation uTlKVOVC, perhaps considering the brothers twins, as 
Verrall suggested in his edition (p.x n.7), noting the Hcurious ex­
pression" at 890, o/LOC1TAayxvwv TE 1TAEVpW/LaTwv. More cogently rele­
vantmight be the phrases at 804f and 820, {JanMwc 0' O/L6C1TOPOL 1 avopEc 
and {JacLMow 0' 0fLOC1T6pOLV; but they do not unambiguously mean the 
same as OLovfLoyevr7c, <twin'. Most relevant probably is the utterance 
of the Chorus at 849, though requiring some adjustment of the para­
dosis. 848-50 might best be presented in this form: 

I~' , I~ \ ~ Tao aVTOOTJl\a. 1TPOV1TTOC 
• 1\ \ 1 ayyEl\ov I\OYOC. 

OL1TAatv fLEplfLvaLV oloVfL' u-u-I-GU-GUx 
• 1 avopEa. 
I, • .l.. 1 ~ 1 KaK aVTo,/,ova OL/LOLpa u-uuuu- uu-uuu-,,-...,,-... 1"-'" ,,-... 

1\ " {J TEI\ELCX Ta 1TCXUTJ. 

iambic trimeter 

iambic metron+ doch­

mius 

two dochmii 



DOUGLAS YOUNG 31 

"These things (are) self-evident; the messenger's report (was) fore­
seen [or is manifest]. Twin (were) the manly acts of my twofold 
objects of concern. Evil, kin-slaying, doubly-shared, complete (are) 
the calamities." 

In 849 M's original reading was the genitive dual, 8'1TAaLV f'€plf'vaw, 

kept by Wilamowitz, Mazon, Groeneboom, Verrall and Italie. It was 
later corrupted to the reading of the deteriores, omAa£ fLEP'fLva,. The 
MSS mostly agree in the letter-sequence O£OVfLavOp€a, with or without 
division and variously accented. Of those known to me only Nb 
post corr. seems to have the accentuation I prefer, 8loVfL' avop€a = 
otovfLa aVOp€La, 'twin acts of manhood', such as the heroic age reck­
oned personal slaying of an enemy. For fLEpLfLva in the sense 'object of 
concern' cf. LS] s.v. 2, citing Homer's Hymn to Hermes 160, fL€yaA7]v 

\ ',J. I I I 0 ~'O' \'0' 0 ~ C€ 7Ta77Jp 6I'V'T€VC€ fL€pLJLvav V7]'TOLC av PW7TOLCt Kat a ava'TOtCt €OLCt. 

At 866ff some minutiae involve iotas, where I would write: 

~JLa.C 8e 8lK7] 7TpO'T€POV c/>~fLTJc 
\ ~ 1\ ~ I 0' ., 'E' 'TOV OVCK€l\aOOV VJLVOV ptJIVOC 

, - 'A/~ , taX€W w~ T 

, 0 \ ~ " 1\ €X pOV 7TaLav €7TLfL€I\7T€LV. 

"But it is proper for us first with our songs both to sound forth the 
ill-sounding hymn of the Fury and to chant an odious paean to Hades." 
At 866, for the manuscripts' CPTJJL7]c Hartung already proposed the dative 
plural CPTJfLaLC, but Aeschylus may have had the epicism CPTJfLTlC, from 
which the paradosis has dropped the iota. At 868 an adscript iota in 
al8aL appears in A and K, two of the better independent witnesses in 
the category inferior to M and I, where it may be from continuous 
tradition, and is unlikely to arise by scribal effort in their own period. 
The dative 'At8~ is preferred by Paley, Wellauer, Victorius and Pauw. 
Also at 868 the manuscripts' lax€£v is kept by Hermann, Wilamowitz, 
Smyth, Mazon, Groeneboom and others. 

At 895ff a plausible colometry of the paradosis is attained thus, if 
one restores the words deleted by Elmsley and deletes those added 
by Murray: 

v--I-v- syncopated iambic di-
meter 

\ \ <;:'1 \ I 
1T1\ayuv OOJLOtct Kat cWfLacLv. - -v-I v- -v- iambic metron + dochmius 



933 

'" 947 

32 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

\ I • I '11'£'11'l\ayp.£vove £VV£'11'W 

, -.~, , 
allavoaTlfJ P.£V£L 

, , ,., , 
apaLlfJ T £K '11'aTpoe. 

u-v-I-v-

v--I-v-

v--I-v-

---. v-vv--

syncopated iambic dim­
eter 

syncopated iambic dim­
eter 

syncopated iambic dim­
eter 

dochmius 

"You tell of a penetrating blow against families and persons. Smitten 
I declare them by unutterable and accursed force proceeding from 
their father, through a discordant destiny." The plural 86P.OLC£ is 
relevant because both Eteokles and Polyneikes left children. 

At 915 the problem of responsion is solvable by a transposition. 
The antis trophic colon at 926 consists of two dochmii, the second in 
Reizianum form: 8ve8alp.wv CrpLV & TEKovca '11'pO 1Tacav. - - - v -I v - uv - - . 
In 915 I would read 86p.wv Tove p.aX axa£eca '11'PO'11"P.'11'£L, v- -v-I 
- -uv- -. The subject is Antigone, who has just spoken 911-14. Then 
916f may be construed thus: 8aiK-rTJp (€C'TtV avrijc) y6oc, aVTocT6voc, 
aV'To'11'~p.wlI ••. "Heart-rending (is her) lamentation, self-bemoaning, 
for her own woes ... " The order of words from which I make the 
transposition in 915 is that of M, which offers S6p.wlI p.aA' axac:eea 
Tove '11'po'11"p.'11'n. Nearest is I, on the line, with axane Tove, and an 1] 

over the second a. A marginal variant in I has not proved legible. 
The deteriores get increasingly wilder with axwNxw/axa/aAwv €'11" /€c/Ele 
aV'Tove. But these efforts are tame compared with the wilderness of 
the deterrimi manifested in Wecklein's Appendix and Dawe's 
Repertory. 

The maximum conservatism is achieved at 933f and 947f by this 
colometry: 

• I "'~ , " 8 op.oe'11'OpOL 01]Ta KaL '11'allWI\C: pOL 
S~'Top.ate OU rp{)..atC; •.• 

€xOVCL P.Otpav AaX6V'T£C P.'A~L 

'" '" I 'I o!:,geooTwv ax~v. 

v-v-I-v-Iv-v- sync. iambic trimeter 
rv- -u- dochmius 
v-v-I-v-I-v- sync. iambic trim­

eter 
-v-v- hypodochmius 

"Joint-sown truly and wholly destroyed by unfriendly cuttings up." 
... "They have obtained and possess, poor wretches, a share of heaven­
sent griefs." 

The only departure from the general tradition is that at 947 I go 



DOUGLAS YOUNG 33 

with M in leaving out the ~ or Jj of the deteriores, which is doubtless 
derived from the interlinear exclamation often put in by glossing 
hands to mark a vocative, or what is thought to be such. Lines 933 and 
947 will be syncopated iambic trimeters, the first singly syncopated, 
the second doubly. We see consonantal iotas in 934 OLa'To[-LaLc and 948 
c)Loc06'Twv, and synizesis in 947 fl-€l\eoL and 948 &X€wv. 

Colometrical notes may be in order for 973f and 984f, which seem to 
make sense thus: 

" , ~ I~" /e 973 aXEwv 'T OLWV 'TaO eyyu EV. 

m[Aac 0' aio' &oeAcpa~ &OEAcp@V. 

984 ~ I I~ ,. I 
'" OUC'Tava K1]O€ ofl-wvufl-a. 

olvypa 'TpmaA'TWV 117}fl-aTwv. 

vv-v-Iv-v-

v--Iv-GU--

- -v-Ivv-vx 

v-vvl---v-

iambic dimeter 

bacchius+ dochmius 

iambic dimeter 

2nd paeon+ dochmius 

"And to griefs alone is the situation near."-"And here are sisters 
beside brothers." ... "Unhappy the word that means both kindred 
and mourning."-"Tear-sodden because of thrice-brandished woes." 

At 973 I propose what seems to be a novel division of the paradosis 
'TOLwV. For the general sense perhaps the nearest parallel is Soph. 
Ant. 933, ()ava'Tou 'TOUT' Jyyu'Ta'Tw 'TOV7TOC &cpLK'TaL. Next nearest perhaps 
is Soph. OC 1217, 7ToAAa fl-€V ai fl-aKpa't afl-EpaLKa'TEeEVTOO~ AV7Tac €yyU'TEpW. 
In 973 we would have to allow shortening by internal correption of 
OL in oiwv. In 974 we have synizesis in &oeAcp~v and epic correption at 
the end of &oEAcpa't, for which compare, also in a dochmiac context, 
971, correption of -ou in 7TPOC cplAOU Ecp()LCO. Miss Dale (p.114) cites an 
example from Aristophanes, Thesm. 915, the -w of KVCW in the resolved 
dochmiac pair CPEP€ CE KVCW. a7TaYE fl-' a7Tay' a7Tay' a7TaYE fl-€, which of 
course parodies Euripides. 

At 984 we have a lyric iambic dimeter with an anapaest in the third 
foot. For the OVCTava of Stephanus and Francken we have now manu­
script support, or nearly so, with oVCT7Jva in I Nd F linea Tr. Cf 998, 
i~ iw, ovcTavwv KaKwv avag ... (M K 0 P linea Q yp.) vGU- --I 
-v-v-, dochmius+ hypodochmius. In 985 those who do not believe 
in the existence of second paeons will prefer to call the first ele­
ment a resolved bacchius. 

At 982 the paradosis has &7T(.oA€C€ ofj'Ta, defensible as a dochmius, 
v - Uv - x, answering to 971, 7TpOC cplAOV Ecp()LCO, - GU - v x, each ending 
with a brevis in Longo. 



115 

34 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

III. Prometheus Vinctus 

Conservative colometry is called for at several places in PV where 
editors have transposed or deleted or otherwise altered. Thus at 116 
all Dr Dawe's MSS give us 8~occv7'OC except for Y, and this form, with 
double sigma, fits best with the preceding and following cola, thus: 

"v " Ea Ea. uuu- 4th paeon (first a eor­
repted) 

I , I I '0;:- \ I • 'J. I I I I 4 b h·· 7'LC axw, 'T'LC oOJLCX 7TpOCE7T"rCX JL a'f'~'J'YTJC, u- - u- - u- - u- - ace 11 
8 I "Q ' ,,-, I I I ~OCCV7'OC, 7] I-'p07'~£OC, 7] K~KpaJL~VYJ; u-uu -u- u-- -u-
~ "" LK~7'O 7'~PJLOVLOV ~1T£ 7TCXYOV; vvv- uvuuuv-I ,-..,,-.., 

, '~8 '" I ~ \ 8 1\ 1TOVWV ~JLWV ~wpOC, 7] 'T'L 07] ~I\WV ; 

Line 116 then appears as 2nd paeon+ cretic+ bacchius+ cretic, and 
117 as a 4th paeon (with the initial iota short as an unaugmented 
aorist form), followed by a resolved dochmius. This closes the agi­
tated rhythm of Prometheus' initial utterance, and after a pause he 
goes on with 118, an iambic trimeter of dialogue type. 

The question of Homerisms and Ionisms is raised by the majority 
reading at 138, where all Dawe's manuscripts except K P Il offer 7" 

~l~LCCOJL'vov. Here the tau preceding guarantees the psilotic form. At 
1085 and 1092 the paradosis offers, unmetrically, eAtccovCLv •..• eAtccwv, 

which should be amended to forms in ~lA- rather than the editors' 
~'A -. LS] s. v. EAlccw note that MSS of Hippocrates offer psilotic compound 
forms, Ka7'-~Alcc~w, Ka7'-~£Alga£. At PV 882 one finds 0JLJLa7" ~Aty87]v in 
M I Y 0 ante corr. Tr., and it may be that this should be printed. For a 
general discussion see Broadhead on Pers. pp.250-53. In PV Ionicisms 
include 122 ElcOLXVEVCLV, 831 8WKOC, 677 AipV7]c aKP7Jv 7'E 'the height or 
promontory of Lerna', unjustly altered, with Canter, to A'pV7]c 7'~ 

KP~VYJV, the more so that Lema has many KpfjvaL. 
At 237 there is a small point worth raising about the idiom of the 

article with 7'0£oc8~. Most MSS offer us 7'0 7'0£ 7'o£a£c8~ 7T7JJLOVCX£CL K&JL7T7'oJLaL. 
M has, on the line, 7'W 7'atc 7'0£a£c8~ .•• , with 7'0£ over 7'ate. Apart from 
the problem of euphony, in the succession 7'0 7'0£ 7'0£-, the reading 
TatC 7'oLatc8~ seems more emphatic. There appears to be a gradation, 
from 7'O£OCS~ 'T'LC 'of some such sort', through 7'OLOc8E simpliciter 'of 
such a sort', to <> 7'o£oc8~' of such a sort as this one here'. 

At 425-32 the epode may be most conservatively analysed thus: 



DOUGLAS YOUNG 35 

425 /Lovov OT] 7TpodJ' EV' v--I-v-Iv-v- sync. iamb. trim. 
aAAov EV 7TOVOLC 

~ ", ~, oa/LEvT aKap,avTooETOLC v-vv-vv-

TLTaVa AVWXLC EicLSO/Lav, - -v- -I-vv-vv-

Odw"A­

Tlt.av(}', oc alEv V7TE{pOXOV - v 1-vv -I v-
() ' " I , I C EVOC KpaTaLOv ovpavLOv v-v-v -vv-vv-

TE 7TOAOV 

enoplian (prosodiac) 
iambelegus 

Glyconic 
iambelegus 

430 I f 'Y VWTOLC V7TOCTEva,:>EL. --v-Iv-- iamb. dim. catalectic 
iamb. dimeter 

435 

{3oq. OE 7TOVTLOC KAVOwv 

~VP,7TtTVWV CTEVEL j3a{]vc, 
KEAaLV(Je 0' 

v-v-Iv-v-

-v-Iv-v-Iv-- sync. iamb. trim. 

"A ~ • Q' , --"..... I 2d h .. LOOC V7TOtJpE/LEL /LVXOC vvvvvv- v- - - - oc mn 
yac, 7Tayat 8' 

t I ..... ,-.... ~ 
ayvopvTWV 7TOTap,wv - vv - vv-

, "\ " I CTEVOVCLV al\yoc OLKTpOV. v - v - v - -

dochmius (hemiepes) 
iamb. dim. catalectic 

425 7TpOCO' EV' Zakas, 7Tpoc8EV codices. 432 {3a8vc M H, {3v(J6c ceteri. 

A translation might run: "One other Titan alone did I see previously 
in tribulations, subdued by indefatigably-binding outrages, the god 
Atlas, who always groans under the excessive strong force and the 
heavenly sphere on his back. And at the cry the marine billow, as it 
falls together, groans deep down; and Hades' dark recess of earth 
rumbles below; and the springs of the pure-flowing rivers mourn the 
pitiful suffering." 

Through taking 431 {Joq. as a verb scribes seem to have arrived at the 
noun {Jv(}oc in 432. Aeschylus leaves it vague whether the cry is that of 
Atlas or that of Prometheus; and the same applies to 435 aAyoc. In 
429 TE is appositional or defining, as at Agam. 9f EK Tpolac cpamv I 
aAwcL/L6v TE {3eX~LV: one might translate it "namely." 

An alternative colometry of 433f might be: 

"ALSoc IJ7To{3ptp'EL /LVXoc yae VUVVvvl-v- - trochaic dimeter 
7Tayat {]' ayvopvTwV 7ToTa/Lwv - - -vv-vv- dactylic tetram. catalectic 

Cf Eum. 1042, Aap,7TeXoL TEp7TO/LEVaL Ka{]' &SOv. 

At 553 M's 7TPOLOOUC' is defensible, meaning 'seeing what is before 
my eyes'. Cf. Thuc. 7.44.2, T~V /LEV Ot{;LV TOU cwp,aToc 7TpOOpaV. 

At 617 one can accept M' s 7Tav yap ovv 7TV(}OLO P,OV, the more so in view 
of 520, where TOUT' OVK <Xv OOV 7TV(}OLO appears in MOP V /1. Denniston, 
in what has been called his "magnus de particulis liber" p.425, says that 



36 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

"In A. Pro 520 M's OUK «Xv oov can hardly be right." But at p.424 he 
gives some examples that show oov emphasizing the negative, as 
Xen. Cyr. 3.3.50 OUK «Xv oov Tog&Tac ye. At p.446 he notes that o~v adds 
to yap «the idea of importance or essentiality." At 520 Prometheus is 
saying "That you would not learn by asking." At 617 we have a poten­
tial optative without av, which W. W. Goodwin (Syntax of the Moods 
and Tenses of the Greek Verb2 [Boston 1897] §241) grudgingly admitted 
at Cho. 594, among other instances in Attic poets that he thought 
"mere anomalies, even if we admit that the text is sound." 

At 629 we might repunctuate and accent the paradosis to make: 
p~ pOV1TpOK~8ov pauuov. ~c ~fLO~ yAVKt$. "Do not take heed for me unduly 
far. Thus is my pleasure" (scil. to learn of my future wanderings). For 
wc= OVTWC cf Agam. 930, 1354, Pers. 565, Supp. 622. 

At 667 the epithet 1TVPWTOV is applied to Kepavv&v by M, supported 
by 0 ante corr. and I supra yp. It is used by the comic poet Antiphanes 
and means 'fiery', just as suitable a sense as the majority variant 
1TVPW1TOV, 'fiery-faced, fiery-looking', a commoner word in the sur­
viving literature. 

At 687-95 the astrophic lyric makes good sense and metre thus: 

---- ---- ---­uuuuuuu-
",,, ,,, 

OV1TOT OV1TOT 7JVXOVV - U - U - -

689A glvovc pOAetcOat A&YOVC u-u-I-u-
, , \ , , 

689B €C aKoav eJLav, 
690 ou8' ~8e 8vcOlaTa 

8VCOtCTa 

1T~JLaTa AVJLaTa 8elJLaT' 
',L,.!.. , aJL'f"IKet Kev-

.1. ' .1_.'" Tpcp 'f'vXetv 'f'uXav €JLav. 
" " c.w C.W, 

----uuu-u-

- -u-I u-uv- x 

---- ---- I -uu-uu- u- - --

-----Iu-wu-

Motpa. Motpa. 1Tl4>ptK' -u-u--

695 ECt80vca 1Tpa£W 'lovc. uv-u-Iu--

dochmius 
ithyphallic 
sync. iamb. dimeter 
dochmius 
iamb. metron+ dochmius 

2 dochmii 

2 dochmii 

ithyphallic 
iamb. dim. catalectic 

The sense runs: "Oh! Oh! Keep off! Woe is me! Never, never did I 
think that such strange tales would come to nly hearing, nor that 
such hideous intolerable miseries, pollutions, terrors, with two­
pronged goad would chill my soul. Oh! Fate! Fate! I shiver when I 
see the condition of 10." We have correption of final long vowels in 
687 la la and 692, the first lw. At 688 7JUX&JL7Jv is attested by M and 
others, but is probably an intruded gloss. In 690 M I 0 ante corr. 



901 

903A 

903B 

905 

DOUGLAS YOUNG 37 

B C H rightly omit the Kat before SVCOtCTa. At 695 €ctSovca is in the 
best MSS M and I, and in others. 

At 776 there is a rare particle combination in M B H C, namely with 
the phrase P:T]o€. cavTfjc T' EKfl-afh::'iv. Cf Theognis 1031, and the combina­
tion ovS€ TE or oVo€. ... TE listed by Denniston, who failed to collect 
WT]O€ ••• TE. 

At 872 our best MSS, M and I on the line, give us the dative plural 
KA.€tvOLC, which makes good sense if one punctuates thus: 

C'TT'opac yE fl-~V EK TfjCOE e/>vcETat ()pacvc, 
, c \ ~" , • ~~.. \ TOSOLCL KI\ELVOLC OC 7TOVWV EK TWVO Efl-E 

'\veet. 

"From her seed at any rate there shall be born a bold man, who, with 
his famous bow, shall release me from these troubles." 

At 901-06 the best colometry seems to be as follows: 

, 'S' ~, II ~ A ' r Efl-OL , OT tfl-€V ofl-a oc 0 
~"""-"'I ~...-..... v- vvvvv vvvvvv x 2 dochmii 

yafl-OC, ae/>o{3oc. 
• 0'0 \ 0 \ I ov E La. fl-YJ E KP€LCCOVWV -~v-Iv-v- iamb. dimeter 

() ~" "e/> EWV EPWC a VKTOV v-v-Iv- x iamb. dim. catalectic 
" 0 ' ofl-fl-a 7TpOC paKO£ fl-E. -v-v-x ithyphallic 
"'\ "0 • o 7T(::l'\E-

.............~~ .............. , 
a7TO Efl-OC 0 E Y vvvvvvvvv 

" 
, ,..... ,..... 

2 dochmii fl-oc, a7Topa 7TOpLfl-OC. vvvvvvx 

'8" " ,'1\ , ou EXW T£C av YEVOLfl-av. -v-vl-v-- troch. dimeter 
\ LI" , , t .... Tav LOC yap OVX opw -v-vl-v- troch. dim. catalectic 
~ " e/>' • JI 

-vv-Iv- x A ristophanean fl-YJTLV 07Tf{- VYOLfl- avo 

901 OT· i/LEV Young, OTt /LEV codices. 903B 7TpocSpaKoL Salvinius, 7TpocoapKoL M 

7TpocSepKOt plerique. 

The sense is: "For me, when a marriage is on the same level for going 
on with, it is without terror. I am not alarmed. But may the passion 
of superior gods not look upon me with unescapable eye. This is a 
war that none can war against, source of resourceless evils. Nor can I 
tell what might become of me. For I do not see by what way I might 
escape the intention of Zeus." In 901 my redivision may be acceptable. 
If not so, the easiest alteration is to Arnaldus' aTe /LEV • •. L5j, S.V. 

o/La,\6c II, cite the adverb used with such verbs as f3alvELv, 7Tpo,ivaL, 

KLve'ic{}at, and marriage may be conceived as a side by side progress. 
At 910 M should be supported in omitting T' after {}povwv in the 



38 READINGS IN AESCHYLUS' BYZANTINE TRIAD 

phrase DC ath·ov EK TVpavvtSoc I 8povwv atCTOV EKf3cx>..~t, H(a marriage) 
that shall expel him from the throne of his sovereignty out of sight 
[in utter destruction]." The adjective is proleptic. 

At 926 consideration should be given to the variant KaKWV offered 
by M on the line, K supra, ll. and P yp., giving the verse 7TTatcac SJ T08~ 
7TPOC KaKWV p.a8-r1c~Ta, ... HAnd after stumbling against him [the ad­
versary described at 920ff] through sufferings he shall learn ... " 

At 933 M ante eorr. offers the particle Sal, found also at eho. 900. 

"L Sat 4>0{30LP.'Y}V c[J 8av~Lv o~ p.6pnp.ov; would be an example of Dennis­
ton's category "Emphatic, in a lively or surprised question" (p.263). 
Five lines earlier in the stichomythia the Chorus-leader uses the sole 
Attic example of 8'Y}v. 

At 948 most MSS have !Lv T' €K~ivoc €K7Tt7T"~', which is kept by Her­
mann, Wecklein, Paley, Mazon, Groeneboom and others. But it 
might be better to redivide it into !Lv ,,~ K~tVOC EK7Tt7T"~', for euphony, 
to avoid the succession of two words starting with €K-. 

At 992 the hapax word al8~povcca appears in M supra yp. and Y 
post eorr., supported by the form al8€povca in I supra yp. and Q2 yp. 
Rose dubiously considers it may mean 'sky-ranging'. Perhaps one 
could render it 'heavenly, ethereal'. It is not noticed by Italie, nor by 
LS] and the Supplement. The majority variant a18cx>..ovcca is an epidsm, 
and might be thought stale compared with an epithet stressing the 
non-earthly origin of the fire with which Prometheus expects to be 
assailed. 

UNIVERSITY OF NORTH CAROLINA AT CHAPEL HILL 

December, 1971