Manuscript Problems in Euripides' "Hecuba" Matthiessen, Kjeld Greek, Roman and Byzantine Studies; Winter 1969; 10, 4; ProQuest pg. 293

Manuscript Problems in 
Euripides' Hecuba 

Kjeld M atthiessen 

A NEW CRITICAL EDITION of Euripides' Hecuba, on which I began 
work in 1962, is now almost completed. Since much time will 
be required in press, however, a preliminary report may be 

useful to scholars at work in similar fields-particularly future editors 
of Orestes and Phoenissae-and may provoke constructive criticism in 
time for revisions to be made. 

One must begin with a survey of Euripidean manuscript problems 
in general. The nineteen plays preserved under the name of Euripides1 

can be divided into three groups, according to their different kinds of 
transmission: 

1. The so-called 'Byzantine triad' (Hecuba, Orestes, Phoenissae). 
2. The ten plays equipped with scholia, except the triad: Medea, 

Hippolytus, Andromache, Alcestis, Rhesus, Troades, Bacchae. The Bacchae 
has no scholia in the extant manuscripts, but there are sufficient 
indications that it still had scholia in mediaeval times.2 

3. The nine plays without annotations or so-called 'alphabetical 
plays', because their titles begin with E, H, I, and K. They look like the 
remnants of an alphabetical collection of Euripidean plays:3 Helena, 
Electra, Heraclidae, Hercules Furens, Supplices, Iphigenia Aulidensis, 
Iphigenia Taurica, Ion, Cyclops. 
In a way similar to the extant plays, the extant manuscripts also can 
be divided into three classes: 

1. Mss containing the plays of all three groups: L P. 

1 The fragment of Danae which is preserved only in P (fo1. 147V-148r ) will be entirely dis­
regarded in the present study. The origin of this fragment is a difficult problem of its own. 

2 See Hermes 93 (1965) 148-58. A different but less convincing explanation of how Rhes. 

Tro. Bacch. entered the manuscript tradition is given by A. Tuilier, Recherches critiques sur la 
tradition du texte d'Euripide (Etudes et Commentaires 68, Paris 1968) 116-22. 

3 See B. Snell, "Zwei Topfe mit Euripides-Papyri," Hermes 70 (1935) 119-20. For a different 
theory see Tuilier, op.cit. (supra n.2). 

293 



294 MANUSCRIPT PROBLEMS IN EURIPIDES' HECUBA 

2. Mss containing the plays of groups 1 (triad) and 2 (other plays 
with scholia): M H B A V, and about twenty less important Mss. 

3. Mss containing the triad only: more than 250 Mss.' 
Each of the three groups of plays raises its particular difficulties. In 

discussing these difficulties I begin with the third group, namely the 
plays without scholia, or <alphabetical plays'. All of them are trans­
mitted in two Mss only, Land P. Therefore only one manuscript 
problem must be solved, the relation between Land P.5 The textual 
problems, however, are rather difficult, for L is a good witness, but it 
contains a considerable number of errors; P, if it has independent 
value at all, offers very few variants that are worth discussing. So in 
most cases we are unable to eliminate the errors of L by manuscript 
evidence, and emendation becomes necessary. Papyri and ancient 
quotations give little help because they are not numerous. 

The editor of a play of the second group (plays with scholia, exclud­
ing the triad) is better provided with Mss. In most cases there are more 
than two, and they are representatives of different families. On the 
other hand there are not very many of them, so that manuscript work 
is neither too complicated nor too time-consuming. The establish-

"We do not have complete lists of the Mss of Eur. The first attempt to compose such a 
list was made by Chr. D. Beck (Euripidis Tragoediae, Fragmenta, Epistulae . .. , vol. III cur. 
e. D. Beckius [Leipzig 1788] ix-xiii: HRecensio codicum vel collatorum vel nondum 
collatorum"). His list contains 164 items, but he notes that it is far from being complete. 
In his time many libraries were still inaccessible. Not until 150 years later was the next 
attempt made, by J. A. Spranger ("A Preliminary Skeleton List of the Manuscripts of 
Euripides," CQ 33 [1939] 98-107). His list contains 274 items. A. Turyn gives a "General List 
of the Manuscripts of Euripides," containing 268 items (The Byzantine Manuscript Tradition 
of Euripides [Illinois Studies in Language and Literature 43, Urbana 1957] 3-9). Turyn dis­
covered a great number in addition to Spranger, but on the other hand he does not 
mention most of the Mss that are later than 1500. He has good reasons for omitting them, 
for in 1503 the Aldine appeared, the first printed edition of all EUripidean tragedies (except 
El.). A substantial number of the later Mss depend more or less on the Aldine, or on other 
editions that in turn depend on the Aldine. Besides, I can say from my own experience that 
the quality of the text declines senSibly in the Mss written in the last decades of the XV 
century. Putting together Turyn's and Spranger's lists and aVOiding any double-counting, 
we get the sum of 316 Mss. In the meantime some additional Mss have been found. 
Probably our knowledge is still incomplete, but for the important age before 1500 Turyn's 
list is a reliable instrument. 

6 Because L is the older and, at least for the 'alphabetical' plays, the far better Ms, there 
are only two possibilities: P is either a gemellus or a copy ofL. For a couple of years it seemed 
as if G. Zuntz (An Inquiry into the Transmission of the Plays of Euripides [Cambridge 1965] 
1-15) had found the right solution and had proved that P is indeed a copy of L. But now 
that serious objections have been put forward by Tuilier, op.cit. (supra n.2) 196-209, a 
modification ofZuntz's theory might become necessary. 



K JELD MA TTl-llESSEN 295 

ment of a stemma codicum seems to be not beyond the reach of our 
knowledge. It has to be admitted that there is a fair amount of con­
tamination.6 But one should at least try to establish a stemma before 
resigning. The quality of the text that can be established by an editor 
relying on manuscript evidence alone is far better than in the 'alpha­
betical' plays. In comparing the edition of the 'alphabetical' Helena by 
Karin Alt (Leipzig 1964) and that of the annotated Hippolytus by 
W. S. Barrett (Oxford 1964), I find that within the same number of 
lines Alt has to put an emendation into the text twice as often as 
Barrett does. 

As for the triad (Hecuba, Orestes, Phoenissae), however, manuscript 
problems are really discouraging. For each of the three plays the 
number of extant Mss ranges between 200 and 300. Most of 
them, of course, are worthless, but it is impossible to know before­
hand which is worthless and which is not. Every editor of a triad play 
has to find his own way of handling such an excessive number of Mss. 
Before describing my solution, I should like to discuss that of A. 
Kirchhoff, the most influential editor of Euripides in modern times. 
As a solution it is both simple and convenient, but it is dangerous, too. 
One needs to know that Kirchhoff began his work on Euripides with 
his edition of Medea (Berlin 1852), that is to say, with an edition of a 
play of the second group. He used all available witnesses, and here 
began to form his opinion about the quality and the relationship of 
the different Mss. His next step was an edition of the Troades (Berlin 
1852). But for the Troades there are only few Mss, so in editing this 
play he could not learn anything by which he could get a broader 
view of manuscript problems outside of Medea. After editing two 
plays he felt confident enough to generalize his observations and to 

edit the whole of the corpus according to the same rules that he found 
useful in Medea. 7 What this procedure means for Hecuba and the other 
triad plays is obvious. For Medea, since it is not a triad play, Mss of the 
third class do not exist. So Kirchhoff considered it justified to eliminate 
all Mss of the third class, which means all Mss that contain only the 
triad. This eliminatio recentiorum is a common feature in all editorial 

6 See V. di Benedetto, La tradizione manoscritta euripidea (IIpoaywlI€S, Studi 7, Pad ova 
1965) 53-122. 

7 EuripidiS Tragoediae ex rec. A. Kirchhoffii, I (Berlin 1855) xii: "Textum poetae ... 
constitui ... ea ratione, cuius exemplum proposui in editione Medeae." 



296 MANUSCRIPT PROBLEMS IN EURIPIDES' HECUBA 

work of that time. It is moreover a legitimate part of the method of 
K. Lachmann. One has to remember that Lachmann developed his 
method in his edition of Lucretius.S And on the Latin side, the 
humanistic Mss (written not earlier than 1400) are indeed heavily 
interpolated and consequently very unreliable. In order to establish a 
text that is superior to the textus receptus of the early printed editions, 
one must begin with the elimination of all later Mss and must go back 
to the oldest available witnesses and use them as basis for the new text. 

Kirchhoff called the entire third class of Mss Hrecentiores" or 
HByzantini"; and he presumed that all of them were derived from a 
selection made at a very late date and that they do not represent 
genuine tradition.s Besides MBA V L P he uses for his edition eight 
minor MSS.I0 All of them belong to the second class, because they are 
not exclusively triad Mss but contain at least part of the seven plays 
of the second group. The editors after Kirchhoff tried to reduce the 
number of necessary Mss as much as possible. Prinz and Murray 
quoted the minor Mss only sporadically. Wilamowitz tried to dismiss 
even B, and Barrett disregarded P even in a play with scholia.ll The 
final point seems to be reached by the editors of Euripides in the 
Collection Bude. Here only MBA V L P are mentioned individually 
in the Conspectus Siglorum and in the apparatus; the old and very 
interesting Jerusalem Palimpsest H is disregarded entirely,12 and the 
minor Mss are quoted collectively as "rec." wherever Murray quotes 
one of them. 

It must be admitted that Kirchhoff's and Murray's system works 
quite well in so far as in the triad no part of the Euripidean text is 
missing in MBA V L P. There is nothing like the striking case of 

8 Compare S. Timpanaro, La genesi del mewdo di Lachmann (Florence 1963). 
8 Kirchhoff, op.cit. (supra n.7) vii: "Posteriore mox aetate exarescentibus paullatim 

doctrinae rivulis magister quidam Byzantinus sylloges amplioris compendium confecit 
trium fabularum ... quam varietatis farraginem equidem COtam abiiciendam statui, 
quippe cuius nullus usus esset futurus sanae mentis critico." For a similar judgement see 
G. Murray ed., Euripides, Fabulae I (Oxford 1902) viii. 

10 The Florentine part ofP was used by Kirchhoff only in his 2nd ed. (Berlin 1867-68). 
11 U. von Wilamowitz-Moellendorff, Einleitung in die griechische Tragodie2 (Berlin 1906) 

207 n.I71: "Paris. B, den man jedoch eigentlich auch nur in der Alkestis notig hat"; W. S. 
Barrett ed., Euripides, Hippolyws (Oxford 1964) 73, 429. It is worth mentioning, however, 
that Barrett for the first time since Kirchhoff used some Mss outside of MBA V L P in a 
systematic manner: H 0 and his group C 0 E (Vaticanus 910, Laurentianus 31, 15, Athous 
lberorum 209). 

11 See now S. G. Daitz, The Jerusalem Palimpsest of Euripides, a Facsimile Edition with Com­
mentary (Berlin 1970). 



KJELD MATTHIES SEN 297 

Oedipus Tyrannus 800, which for Sophocles destroys the whole concept 
of a codex unicusP But in the three plays I have counted 36 cases where 
(mostly in minor matters, but not always so) MBA V L P are all 
wrong, and where Murray relies on the Byzantini (sometimes called 
novicii) alone. One might suppose that in all these cases the superior 
variant is no genuine tradition but brilliant Byzantine emendation.14 
But there are too many cc emendations" of this kind, so that one begins 
to doubt whether this assumption is altogether right.15 

Therefore it seems inevitable for an editor of a triad play to use also 
the third class ofMss in one way or another. So I decided to enter that 
labyrinth, which I should not have dared to do without the help of 
Turyn's book. Like Ariadne's clue it prevented me from going astray. 
Turyn divides the whole bulk of triad Mss into two groups: first, the 
Mss influenced by the Byzantine grammarians of the Palaeologean 
period (the "Byzantini"), and second, those that are free from such 
influence (the "veteres"). We know these grammarians from their 
scholia and introductions to several Greek authors. Their names are 
Maximus Planudes, Manuel Moschopulus, Thomas Magister and 
Demetrius Triclinius; their activity falls between 1260 and 1340. In 
some Mss certain chains of scholia are attributed by certain signs to 
one or another of them. The key Ms for the triad of Euripides is T, 
written partially by the hand of Demetrius Triclinius himself. Here 
we find a clear distinction between Moschopulean, Thoman and 
Triclinian scholia. Another key Ms is Y, equipped with Planudean and 
Moschopulean scholia.16 

13 E. Schwartz, Seholia in Euripidem I (Berlin 1887), uses M B V and C (Taurinensis B.IV.13) 
as basic Mss, but often a minor Ms is used to supply a superior variant or to fill a gap in 
MCBV. 

14 So does for instance V. di Benedetto ed., Euripides, Orestes (Florence 1965) x: "Nei 
pochi casi in cui R S Sa [Turyn's main representatives of the reeentiores] presentano una 
lezione esatta contro il resto della tradizione ci si trova di frome a congetture, per 10 pili di 
origine moscopulea." 

15 Hee. 80, 116, 128,274,535,605,1025,1078,1195; Or. 160,293,349,373,378,782,960,999, 
1027,1266,1278,1289,1300,1387,1507,1653; Phoen. 604,608,619,636,878, 1235, 1403, 1485, 
1527, 1690, 1758. 

16 My manuscript symbols are the same as Murray's and Turyn's. New symbols are: 
Aa Ambrosianus C 44 sup., saee. XIV 
G Ambrosianus L 39 sup., ca. A.D. 1320 
Ga Athous Vatopedianus 36, Gnomologium saec. XII 
K Laurentianus cony. suppr. 66, saee. XIV ineuntis 
Le Leidensis Vossianus Q 33, ca. A.D. 1500 
Ms Mosquensis 508, saec. XV 



298 MANUSCRIPT PROBLEMS IN EURIPIDES' HECUBA 

Turyn's procedure in classifying the Mss is as follows. Since we 
know, for instance, the Moschopulean scholia from T, we must try to 
find Mss which are exclusively equipped with scholia of that type. 
Such Mss are representative of the edition of Moschopulus. They have 
in common not only introductions and scholia, but also certain 
features in the text itself of the three plays. Turyn calls them HMoscho­
pulean interpolations" and takes them for deliberate changes made 
by Moschopulus himself. (My impression, however, is that they are, 
in most cases, rather ordinary errors.) Using these common features 
of the text as evidence for Moschopulean character, one might identify 
as Moschopulean (or, in a similar way, as Thoman, or Triclinian) even 
those Mss that by chance have no scholia. 

In this way it becomes possible to divide the approximately 200 Mss 
of the triad that are not codices descripti into five classes,17 three classes 
of "Byzantini" and two classes of "veteres" : 

(1) 114 Moschopulean Mss (for instance X Xa Xb). 
(2) 24 Thoman Mss (for instance Z Zb Zc Zm Zu). 
(3) 21 Triclinian Mss (for instance T). 
(4) About 20 "veteres recentiores," which means Mss that are free of 

the characteristics of the "Byzantini" and of certain common errors of 
MBA V (dividing errors of MBA V against the "recentiores"), but 
have in common certain errors of their own (for instance R S Sa). 

(5) About 20 Hveteres vetustiores," Mss that share the common errors 
of MBA V, but do not share the common errors of R S Sa nor the 
common features of the "Byzantini" (for instance G K). 
So far Turyn's divisions and subdivisions of manuscripts. I had to give 
them in full detail in order to show the starting point of my own studies. 

The first part of my studies consisted entirely of collations. I col­
lated, partly in the original, partly in microfilm, the text and the 
argument of Hecuba in MBA V L P, in H, in Ga, in F and 0, in all 

Pa Parisinus 2801, saee. XIV 
Pb Parisinus 2810, saee. XV exeuntis 
Pr Remensis 1306, saee. XIII exeuntis 
Rf Laurentianus 32, 33, saee. XIII exeuntis 
Rw Vindobonensis phil. 119, saee. XIII exeuntis 
Vb Vatieanus 53, saee. XV exeuntis 

17 Two less important classes do not appear in our classification: the five Mss with 
Planudean and Moschopulean scholia (e.g. Y) are counted as Moschopulean, and the 27 
representatives of the late Byzantine dyad (for instance J U) are disregarded entirely. The 
contribution of both classes to the text of Hee. is next to nothing. 



KJELD MA TTHIESSEN 299 

"veteres vetustiores" and "recentiores," and in some representatives of 
the three classes of" Byzantini." The first of my results was to find that 
Turyn is often wrong in dating Mss. My impression is that in particular 
he tends to date Mss on paper too late.lS Sometimes the watermarks 
demand an earlier date, sometimes the size of the paper is of an 
earlier type, sometimes the character of the handwriting gives an 
earlier impression, sometimes the internal evidence by which Turyn 
was induced to date the Ms turns out to be at least doubtful. Perhaps 
one might think that a change in dating of about fifty years does not 
make much difference. That is true in most cases, but sometimes it 
really does matter. A good instance is F (Marcianus 468), dated by 
Turyn to the early XIV century. It belongs to the second class of Mss 
because in addition to the triad it contains argumentum Medeae and 
Medea 1-41. It seems to offer a good version of the triad, but according 
to Turyn its value is diminished by the fact that there are some 
Moschopulean, Thoman, even Triclinian interpolations.l9 The hand­
writing, however, suggests an earlier date of about 1290.20 The conse­
quence is that the so-called Triclinian interpolations can no longer be 
regarded as such, for we know that Triclinius was not occupied with 
the text of the triad before 1305. Therefore the assumed Triclinian 
interpolations in F turn out to be traditional matter, that is, either 
mistakes that afterwards also the scribe of T adopted from the Ms he 
copied, or variants that afterwards Triclinius also took from Mss 
similar to F and inserted into his own. The same can be said about the 
presumed Thoman interpolations in F, and Moschopulean influence 
is at least doubtful. Whenever an observation of this kind can be made 
several times, the whole pattern begins to change. 

This is what has happened in our case. Turyn dated all Mss of the 
third class not earlier than 1300. Consequently, for him all agreements 
between "veteres vetustiores" and "recentiores" on the one hand and 
"Byzantini" on the other prove an influence of the Byzantine gram­
marians on these Mss. I found that at least half a dozen "vetustiores" 

18 Some dates are corrected already by J. Irigoin, RevPhil 32 (19S8) 323, J. Martin, REG 73 
(1960) 286, and Zuntz, op.cit. (supra n5) 161 nn.I-2, 162 n.2. 

19 For instance Hec. SO XEpaS (Moschopulean), S3 aKrr"~" (Thoman), 922 EflI3€{3wTa (Tri· 
clinian). The first and third cases are a limine suspicious because the "interpolation" is 
actually the superior variant adopted in our editions. For the second case see Zuntz, op.cit. 
(supra n5) 156. 

20 See J. Irigoin, REG 67 (1954) 510; H. Erbse, in Geschichte der TextUberlieferung der antiken 
und mittelalterlichen Literatur I (Zurich 1961) 275; Zuntz, op.cit. (supra n5) 162 n.2. 



300 MANUSCRIPT PROBLEMS IN EURIPIDES' HECUBA 

and Hrecentiores" should be dated towards the end of the XllI century. 
Consequently many so-called Byzantine interpolations appear in Mss 
that are earlier than the grammarians themselves, so they can no 
longer be regarded as interpolations and turn out to be widespread 
errors or variants.21 The trend in this direction seems to be strong 
enough to raise the question whether there were any Byzantine 
interpolations at all; but there remains a hard core of fifteen cases in 
the text of Hecuba where it is highly probable that a Byzantine gram­
marian altered the text deliberately. 

Another consequence of the new dating of some Mss of the third 
class (or Hrecentiores" in Murray's terminology) is that some of them 
turn out to be as old as A V L P (the youngest "basic manuscripts"), 
or even older. We therefore should get used to the idea that there are 
only four really old manuscripts, M B Hand Ga, all of them written 
before 1204, the year of the conquest of Constantinople by the knights 
of the Fourth Crusade. Then follows a gap, and all the rest of our Mss 
are written after 1261, when the capital was recaptured by the Greeks. 
My collation shows clearly that it is no longer possible to maintain 
Turyn's division between Hveteres vetustiores" and Hrecentiores."22 There 
is only one rather homogeneous group of veteres, and the only pos­
sible subdivision lies between the old witnesses M B H Ga and the 
rest. The difference between these two groups has no stemmatic 
character, but concerns age and quality only. 

There is, in the text of the triad, an incredibly large number of 
contaminations, and, to quote P. Maas, HGegen die Kontamination ist 
kein Kraut gewachsen."23 In consequence, it is impossible to form a 
stemma for the triad. The only thing that can be made out is a 
number of manuscript groups, but many of them cannot be divided 
clearly from one another. We should therefore admit that the text of 
the triad of Euripides is not transmitted in a closed tradition, but in an 
entirely open one like those of the Iliad and Odyssey, or the New 
Testament. 24 

The later Mss differ considerably in quality. When establishing the 
text, we have above all to rely on the four old witnesses, M B H Ga. 

21 Some observations of Zuntz, op.cit. (supra n.5) 151-74, point in the same direction. 
22 My observations confirm the opinion of di Benedetto, op.cit. (supra n.6) 81-92. 
23 Textkritik3 (Leipzig 1957) 31. 
24 R. D. Dawe, The Collation and Investigation of Manuscripts of Aeschylus (Cambridge 1964). 

comes to similar conclusions for the transmission of Aeschylus. 



KJELD MATTHIES SEN 301 

But besides, especially where M B H Ga are wrong or disagree, some 
later Mss turn out to be very helpful: 

(1) K and 0 
(2) T25 

(3) G (a Ms probably related to K, but later) 
(4) P (L is rather disappointing in Hecuba), F and X Xa xb (the best 

representatives of the Moschopulean class) 
R S Sa and some Thoman Mss are also useful, but to a smaller degree. 
All later Mss mentioned above were written between 1280 and 1350. 
Later on the quality begins to decline, although there are still some 
quite good Mss. 

It is an interesting question why about 1300 some Mss are of such 
good quality. R. Browning suggests that there were historical 
reasons :26 "It is likely that when Michael VIII recaptured Constan­
tinople in 1261 books from Nicaea and elsewhere were brought to the 
capital ... A little later the gradual loss of most of Asia Minor to the 
Turks is likely to have led to a fresh influx of books to Constantinople, 
as refugees from these regions brought their possessions with them. 
These might include the contents of monastic libraries." But this is 
only a partial explanation. It explains only the availability of good 
Mss in the capital. Something in addition was necessary-intelligent 
men who could make the right use of them. We have to suppose that 
there were several such men, not only the well-known scholars 
Moschopulus, Thomas and Triclinius, but many other attentive scribes 
and correctors whose names we do not know. The traces of their work 
can be found in the "minor" Mss of the period about 1300. 

Now I should like to give some instances of what may be expected 
from a collation of the later Mss. 

In Hec. 332-33 the chorus comments on the pitiless discourse of 
Odysseus: 

,~ \ '" ~\' \',/..' , \ cuat· TO OOVI\OV wS" KaKOV 7TE<pVK aEt 

ToAfLfl 0' a fL~ xp~) Tn {J{q. VtKwfLEVOV. 

The majority ofMss have7TE¢>vKEvaL, and schol. V tries to expound it: 
¢>EU ¢>EU, wS" KaKOV TO 7TE¢>VKEvaL Ttva SouAov. Tn {J{q, yap a SouAoS' 

25 The quality of T is of course due partly to Triclinius' corrections. But here I take into 
consideration only those cases where Triclinius probably took the superior variant from 
Mss, and not the three cases where he emended (Hec. 164 oalfLwv £C17" E1Tapwyo" 911 Ka1Tvou 
om., 932 £, om.). 

26 R. Browning, "Recentiores non Deteriores," BICSLondon 7 (1960) 12. 



302 MANUSCRIPT PROBLEMS IN EURIPIDES' HECUBA 

Kpa/,OlJ/L€VOS '1oApli- 7Tpa7"7"€tV & /L~ XP-r}. Beyond doubt superior is the 
variant 7TECPVK' CUL, which appears not only in the Byzantini (XYP XaYP 

xbYP ZYP ZaYP Zb Zc Zm Zu T), but also in several later Mss (PS G KYP 
LYP OYP P), some of them, as for instance F, probably pre­
Moschopulean. Moreover it appears in Stobaeus. Therefore we should 
believe Triclinius, who writes in the margin of T: TO 7TECPVK' a€£ EV 'ltV, 

ALav 7TaAaujJ '1WV av'itypacpwv €Vp7]'1aL 27 How difficult it was to find the 
right solution by emendation can be seen from fruitless efforts in the 
margin of later Mss to emend the text (7TEcpVKE 'it schol. T,7TEcpVK€V av 

and 7TECPVK' E'it ZbYP, 7TECPVKE 7TWS MsYP PbYP). Even modern editors had 
their difficulties with this line. Many of them preferred to follow 
Porson in emending '10A/Lij, in v.333 to '10A/Lav instead of accepting 
7TEcpVK' ad. Only in 1867 was 7TEcpVK' a€L definitely adopted by Kirchhoff. 

In vv. 739-40 Agamemnon addresses Hecuba: 

, I....' " \ '1L /LOL 7TpoaW7Tl{J VW'IOV €YKI\Lvaaa aov 
'" I \ fJ \ "" '\ I " fJ' <I", OVPTl, TO Kpav €V 0 OV l\€yHS ; 'itS €a OO€ ; 

Nearly all Mss read 7TpaxfJEv, and so does B3. Only in B ante correctionem 
we find KpafJEv (for KpavfJEv, with a common slip). Murray and Meridier 
esteemed this variant so little that they did not even record it. But 
whenever KpavfJ€LS appears in tragic discourse in concurrence with 
7TpaxfJds, it deserves to be preferred. Compared with the flat 7TpaxfJ€LS, 

it is the exquisite, poetic term. Besides, it is more often used by 
Euripides than 7TpaXfJ€LS. While 7TpaXfJEv is a possible gloss for KpavfJEv, 

the reverse can be excluded. Meanwhile we know that KpavfJEv is 
better attested in our line than Murray and Meridier thought. It is a 
variant in G, it appears (corrected from KpafJEv) in K, KpafJEv is found 
also in POxy. 876 from the V century. We learn from this case that 
sometimes even isolated variants in later Mss can be traced back into 
antiquity. 

In vv. 918-22 the chorus sings of the last moments before the sack 
of Troy: 

I 'fJ\I " 7ToaLS €V al\a/LOLS €KH'IO, 
C. \ "'" \ '\ o;va'lov 0 €7TL 7Taaaal\l{J, 

I "fJ' f .... tf \ vaV'Iav OVK€ opWV 0IUI\OV 

T pOLav 'fAuxS' E/L{3€{3W'Ia. 

27 So already Zuntz, op.cit. (supra n.5) 154-55. Turyn, op.cit. (supra nA) 107, mentions 
1T£CPVl(' a€L in his list of Moschopulean interpolations. 



KJELD MATTHIESSEN 303 

€fA-{3€{3WTa is necessary for metrical reasons. Forming an interesting 
word responsion it corresponds to EfA-{3cxnvow in the strophe. But 
nearly all Mss have EfA-{3€{3aurra. The same is found in T ante correction em 
in a page that is written by Triclinius himself (fo1. 34T ). He corrects it 
to EfA-f343wTa and writes in the margin: €/-'f3€f3wTa XP-1 yp&1J€LVI OVK 

€fA-{J€{JawTa' oVX OVTW yap €X€L 7TpO~ TO fA-€TpOV op()W~J eX)"A' €K€{VW~. SO 
one might think that EjJ.,f3€/3wTa is a Triclinian emendation.28 But it 
also appears in the later Mss Aa F G K, among which F and K are 
beyond doubt pre-Triclinian. Therefore it is more probable that here, 
after discussing two transmitted variants, Triclinius, who was excel­
lent in metricis, preferred the one because of its soundness. 

In vv. 1054-55 Hecuba wants to avoid the blinded Polymestor, who 
enters the stage. She steps aside saying: 

"\" ~'" , I a/\/\ €K7TOOWV a7T€tfA-t Ka7ToaTYJaOfA-at 

8vfA-cfJ 'Eovn 8pTJKt 8vaf'-axwT(xnp. 

The majority of Mss read pEovn, but 'EoVTt is found in the text of U 
and Zb, and as a variant in Le Pr Zm. U and Le are late and unreliable 
witnesses, but Zb and Zm are good representatives of the Thoman 
class, and Pr, a pre-Thoman Ms probably written before 1300, shares 
some interesting variants with Band 0 (e.g. in the same line, Hec. 
10550vof'-€v€aT(XTqJ oyp PrYP, oVoy€v€aTaTqJ BYP). The case is discussed by 
Zuntz, who prefers ,Eovn.29 For 8vf'-cfJ 'Eovn he gives many parallels, 
of which the most striking is Soph. OC 434 (e'€t 8vf'-oc;). The abundance 
of datives in v.1055, confusing at first glance, has a good parallel in 
Soph. Trach. 445-46 (7(»fA-cfJ T' cX:VOPL TilO€ Til voaqJ I AYJcP8Evn f'-€f'-7TTO~ 
€Lf'-t).30 In order to avoid the double dative, Ruhnken emends to 8vf'-oV 

'Eovn, Hartung to BVf'-o,Eovn. But the line is sound and needs no 
emendation. 

In vv. 1192-94 Hecuba ends her invective against pettifogging 
orators: 

A.. , , 'I' " t '1:'" Q ' oo,/-,ot f'-€V ovv Eta o£ Tao YJKpt{.JWKOT€C; J 

, \ \' ,~, ~"\ 't A.. I a/\/\ OV uvvav-ra£ uLa T€I\OVC; ELVaL OO,/-,OL, 
~ 1:'" '\ ' " 'c '\ c' KCXKWC; 0 a7TWI\OvT' ovnc; €~ YJI\V~ € 7TW. 

28 So does Turyn, op.cit. (supra n.4) 190. 
29 Zuntz, op.cit. (supra n.5) 156-57. 
30 €K'1TOOWV is construed either with the genitive or with the dative (cf vv.52-53 YEprXLij 

o' €K'ITOOc1v Xwp~uo/Lat I 'EKa/Jll; the same is valid for arfo{uTau(}at (cf Oem. 8.37 arfo'UTaTE 
oii>'ov OTt ati-r0). 8v/L0 is either causal ('boiling with rage') or local ('in his mind'). 



304 MANUSCRIPT PROBLEMS IN EURIPIDES' HECUBA 

Most Mss offer the unmetrical a7TwAOVTO Kouns. The same can be read 
in T ante correctionem, but Triclinius corrects it to a7TC~AovT' ouns. L has 
a7TWAOVTO KOUTLS, but in the margin we find the sign ... , a common 
signum corruptelae. The metrically correct a7TwAovT' ouns31 is attested 
by Ga GOP Vb. Again the question arises whether Triclinius emended 
or took the superior variant from a Ms. GOP and Vb (a late relative 
of P) may be influenced by Triclinius, but the same cannot be said 
about the Gnomologium Ga that was written about 1200. So it seems 
more probable that Triclinius took a7TwAOVT' ouns from a Ms than 
that he emended.32 Very interesting indeed are the reactions of the 
modern editors in the present case. The Aldine and the other editions 
of the XVI and XVII centuries kept a7TwAovTO Kouns; only King (1726) 

preferred a7TWAOVT' ouns, the variant of his favorite Codex Regius.33 

Brunck, who used A as his basic Ms, returned to a7TwAOVTO Kouns, the 
variant of A, and emended it to oAoVTO Kouns; he was followed by 
Hermann (OAOLVT' elV Kouns) and Hartung (€7T'Y}fJpOV Kouns). Kirchhoff 
in his first edition kept the metrically intolerable a7TwAovTO Kouns, 
because it is transmitted unanimously by his basic Mss MBA V L. In 
the apparatus he mentioned a7TwAovT' ouns, but he considered it a 
correction made by the scribe of the Codex Regius. Only in his second 
edition of 1867, when he used the Florentine part of P for the first 
time, did he definitely accept a7TWAOVT' ouns. Such a case gives us the 
opportunity to observe how the XIX-century editors tended to trust 
their own emendatory powers more than the testimony of Mss, in 
particular whenever these Mss did not belong to the time-honoured 
choice of "basic" witnesses. I hope that the prepared edition will 
contribute to the final demolition (at least with regard to the triad) 
of the highly esteemed, but nevertheless shaky, doctrine that there is 
nulla salus extra ecclesiam and no good variant outside of MBA V L P. 

Finally, a few words about the shape of the edition. The text will 
be based on the old Mss M B H Ga, the later Mss A F G K LOP Pa Pr 
R Rf R w S Sa V (Va), some good representatives of the Byzantini eX Xa 
Xb, Z Zb Zc Zm Zu, T), and on five papyri (POxy. 876 and 877 [Pack2 

389, 390], OBerol. 12319 [Pack2 1567], PCantab. [Fitzwilliam Museum] 

31 For the asyndeton see Kiihner-Gerth3 II.2 p.345. 
32 Turyn, op.cit. (supra nA) 190, considers &7TWAovr' oirns in Ga an excellent correction 

made by the scribe that by chance is identical with Triclinius' emendation. The probability 
of such a coincidence, however, is extremely small. 

33 London Royal Society 7, probably a Triclinian Ms. 



KJELD MATTHIESSEN 305 

inv. 2 [Pack2 1571J, PHamb. 118b col. i [Pack2 434 supp!.]. All of these 
thirty-two witnesses will be quoted wherever they attest the text and 
wherever at least two of them agree in error. Errors of single witnesses 
will be mentioned only when interesting from palaeographical or 
other points of view. Thirty-two later Mss are quoted only where they 
give a good or at least an interesting variant. The Apparatus Testi­
moniorum will be based on the indications ofR. Prinz's edition (Leipzig 
1883), but enlarged by more than thirty new items. The text of these 
ancient or Byzantine quotations will be mentioned in the apparatus 
in every case where they disagree with the Mss or where the Mss 
disagree with one another. The edition will be published in the series 
Wissenschaftliche Kommentare zu griechischen und lateinischen Schrift­
stellern (Winter, Heidelberg). The first volume, containing a general 
introduction and the text, is scheduled to appear in 1970 or 1971; the 
second volume, with interpretation and commentary, will appear 
some years later.34 

UNIVERSTTAT MUNSTER 

October, 1969 

34 I am grateful to Bernard M. W. Knox, Stephen G. Daitz, Alexander Kleinlogel, and to 

the other Junior Fellows of the Center for Hellenic Studies in 1968-69 for their valuable 
suggestions to me. Horst D. Blume helped me to reformulate some crucial passages.