The Journal of philology

SS:
ance

iD iginenuiihaGjcanmuinanaap sien anna prt mene
‘
‘
i > oe - -
o ton we get CN OO RI Oe TITTIES TEES The “4 Ae oat rae DS te >
© i - Tite aly J 4 x oo.
: ey a hi Lgdubesitets VANE PL le
bd thedeh piri
“A ar M Pte fe

de Bos .

-

Phat,

at

+A Raat ee

f

i]
.
Pre

we

hee

“gh sy
get ;
Ped io. =
= he ae

Pe

woe
a

— a, io ee

. a «

é
I
i

: % 3

MSL eee ee

crs ad ~

|

roe
ty
=)"

ve

2 he Journal

OF

PHILOLOGY.

EDITED BY

W. ALDIS WRIGHT, M.A.
INGRAM BYWATER, MA.

AND

HENRY JACKSON, M.A.
VOL. X.

London and Cambridge :
MACMILLAN AND CO.
DEIGHTON, BELL AND CO. CAMBRIDGE.

1882

%

CONTENTS.

No KEX.

Archaeological Interpretations. R. Burn

— On certain Engineering Difficulties in Thucydides’ ene of the

Escape from Plataea. Bk. 111.20—4. F. A. Paley .
On the First Seven Verses ofthe Antigone. F. A. Paley

On some Fragments of the New Comedy, and some Passages of
Aeschylus, Theognis, Alcaeus and Ibycus. R. Ellis .

The Homeric Trial-Scene. William Ridgeway
Note on Xenophon, de Vect. iv. 14. H. Hager
Note on Plato, Apol. Socr. p. 26,p,u. H. Hager .

- Notes on Gender, especially in Indo-European Languages. James

Gow ; :
Atakta. I. Bywater... : gates : eke a
Notes on some Passages in the Politics. J. Cook Wilson .
Observations on the Oedipus Coloneus of Sophocles. J. P. Postgate
Old German Glosses from a Bodleian Manuscript. F. Madan

Traces. of Different Dialects in the ee of Homer. A. H.

Sayce . ‘ :
On some Difficulties i in the Platonic poe R. D. AoE Ein
On Plato’s Republic VI 509 p sqq. Henry Jackson
Aesch. Ag, 115—120. A. W. Verrall

PAGE

110
120
132
151

iv : CONTENTS. -

No. XX.
PAGE
Thilo’s Servius. Henry Nettleship. . . . . . . 4W58
Pyrrhus in Italy. Henry Elliot Malden . ; . : é ‘ 172
Biology and Social Science. F. Field. ‘ < ; : ;
Horatiana. A. E. Housman . ; ; aise . ae 187 | |
On a Passage in the Rhetorica ad Herennium. W. Warde Fowler 197
Dissignare. Henry Nettleship Fel ayer see an\enom : 206
The Chronology of the Books of Kings. W. Robertson Smith . - ee
On the Text and Interpretation of certain passages in the Aga-
memnon of Aeschylus. W.W.Goodwin. . . . . 214
_ On the Fragments of Euripides, H. A. J. Munro : (3, 90
Sv} Plato’s Later Theory of Ideas. Henry Jackson .. ne 253
The Simile of the Treacherous Hound in the Agamemnon. A. W.
Verrall : ; ‘ : ‘ ; ; ; ‘ 3 299

Aristotle, Politics 1v (vit) 13 §§ 5—7. 13324 7sqq. Henry Jackson 311

(

THE JOURNAL

OF e

PHILOLOGY.

ARCH AOLOGICAL INTERPRETATIONS.

1. PROPERTIUS Iv (v). 8. 1,

Disce quid Esquilias hac nocte fugarit aquosas.

The epithet aquosas as applied to the Esquiline hill by
Propertius has not been rightly interpreted by commen-
tators. They suppose that the hill was so called from its
springs and marshy ground, and refer to Varro’s statement
(L. L. v. § 49) that its name was derived from the ewsculeta
which grew upon it, as proving that marshy ground occupied
at least a part of the slopes. Another explanation given by
commentators has been that aquosas is a translation of the
Homeric epithets wodv7ida& or midjecoa applied to Mount
Ida in IL vue. 47,: x1. 183, as it 1s in Hor. Od. m4 20. 15
“aquosa raptus ab Ida.” But neither of these explanations is
satisfactory. Varro’s derivation of ‘ Esquilie’ from the escu-
leta which grew there seems improbable, and Mommsen R. H.
Bk. 1. ch. 4, and Corssen Lat. Spr. Vol. 11. p. 1023 are probably
correct in taking the true derivation to be from excoline “the
suburban district,” as iaquilinus from incolo and sescenti from
sexcentt. Nor does it seen likely that Propertius would have
used the Homeric epithet merely as a poetical ornament, with-

Journal of Philology. vou.x. 1

2 THE JOURNAL OF PHILOLOGY.

out any special significance. As he lived himself upon the
Esquiline (Prop. Carm. Iv. 23, 24),.he must have intended to
express some feature of the hill which would be at once
recognized by every Roman resident.

Nor is it difficult for the archeologist to see what this
feature of the hill was. He will at once recollect that almost
all the aqueducts of Rome entered the city on the Esquiline
hill near the Porta Maggiore and the Porta 8. Lorenzo, where
considerable remains of the specus of several. of the more
ancient and important aqueducts still remain, the Marcian,
Tepulan and Julian, with the records of their restoration by
Augustus inscribed upon them. Agrippa and Augustus during
the life of Propertius renovated and distributed the water of
these and other aqueducts. We find Pliny, N. H. xxxvr.
§ 121, stating that Agrippa when edile constructed several
hundreds of cisterns and aque castella, and fifty great public
jets of water.

Straba also, v. 3. 8, says that nearly every house had oi-
doves Kal Kpodvot apGovor, and that rivers of water ran through
the city. The passages in Horace, Ovid, Martial and Juvenal
which describe the abundance of water dripping and bursting
from the cisterns and pipes are well known. (Hor. Ep. 1. 10.
20 ‘aqua tendit rumpere plumbum. Ovid. Met. Iv. 122.
Mart. 11. 47; 1x. 19. Juv. 1. 11.)

This escape of water from the pipes and specus and foun-
tains on the Esquiline is probably referred to by Martial
(Ep. v. 22. 6), where he speaks of the stones on the ascent
from the Subura to the Esquiline as “ Nunquam sicco sordida
saxa gradu,” a passage which is often wrongly understood as
referring generally to mud in the streets. So also in Ep. Iv.
18. 2 Martial, speaking of the arch of the Aqua Virgo over
the Via Flaminia, says “Et madet assiduo lubricus imbre
lapis.”

The puticuli on the Esquiline mentioned by Festus, p. 216
ed. Miiller, and referred to by Hertzberg in his commentary,
were, as Festus says, pits for the deposit of the bodies of criminals
executed on the Esquiline, and have no bearing on the epithet
aquosas, but were named from the putrid filth they contained.

ARCHEOLOGICAL INTERPRETATIONS. 3

Propertius refers to the improvements introduced by Agrippa
when edile in the distribution of the water which came by
the Julian and Marcian aqueducts to the Esquiline, when he
speaks (in the line Carm. 11. 2. 12 “ Non operosa rigat Marcius
antra liquor”) of the pipes by which the water was conveyed
to private houses. Frontinus de aq. 9, in reference to the
same act of Agrippa, says that Agrippa “Compluribus salien-
tibus aquis instruxit urbem.” The greatest portion of the
water supplied by the aqueducts to Rome must therefore have
passed over and through the Esquiline hill. No doubt many
of Agrippa’s fountains and castella were constructed there, and
many pipes passed from the main channels in all directions to
the houses and gardens of Mzcenas and other wealthy courtiers
who occupied the Esquiline after the ground had been cleared
and made as Horace calls it saluber. (Sat. 1. 8. 14 “ Esquiliis
habitare salubribus.”) The warm swimming bath which Dion
Cassius, LY. 7, mentions as one of the public improvements
introduced by Mzecenas was probably in his grounds on the
Esquiline.

2. Propertius Iv (v). 4. 14,
Bellicus ex illo fonte bibebat equus.

Commentators and archzologists have been misled in their
statements as to the probable position of this fons by their not
having recognized the fact that the so-called Carcer Mamer-
tinus was at a very early period a tank or well-house which
probably supplied water to the district between the Capitoline
and Quirinal hills. In this poem in connection with the story
of ‘Tarpeia, Propertius has evidently called up in his imagina-
tion the state of the ground occupied by the two contending
troops of Romans and Sabines. The Sabine warriors under
Tatius he represents as having descended from the Quirinal
hill and posted themselves at the foot of the slope of the
Roman Arx on the Capitoline. Here the poet imagines that
there was a clump of trees watered by a spring which rose on
the side of the Capitoline and flowed down towards the site
of the Forum Romanum. He marks out the spot by the words
in line 13, “ubi nunc est curia septa.” The Curia, it is nearly

1—2

4 THE JOURNAL OF PHILOLOGY.

certain, stood on the slope below the site now occupied by
the church of S. Giuseppe dei Falegnami and the ancient
well-house called the Mamertine Prison, and this is exactly
the position which the Sabines, descending from the Quirinal
and posting themselves in the Forum valley, would occupy
between the two hills.

Mommsen, in his admirable discussion of the topography of
the north end of the Forum, mentions this passage of Pro-
pertius and suggests that the lacus servilius is the fons re-
ferred to (Ann. dell’ Inst. xvi. 302). But the lacus servilius
was on the side of the Capitoline nearest to the Palatine, and
therefore on the opposite side to that from which Tatius was
imagined by Propertius to have approached the Forum. Pro-
pertius must have had the Curia as restored by Augustus in
his mind, and this was certainly near the Comitium and at the
north-west corner of the Forum.

The lower chamber of the building now called the Carcer
is constructed in a conical shape by the gradual projection of
the stones forming the sides. This mode of building is of a
very early date, and is found in the Regulini Galassi tomb
at Csere in Etruria, and in well-houses at Tusculum and at
Fiesole and Cortona. See Dennis, Etruria, Vol. u. pp. 46,
128, 451, and Gell, Topography of Rome, p. 432, where’ repre-
sentations of the tomb and the well-house are given, The
well at Tusculum shews most clearly what the nature of this
Roman watering-place originally was. We have there the
truneated conical dome where the spring water collects, and
near it are some troughs supplied from it with water which
were evidently washing troughs, or troughs from which horses
might be watered. Propertius had a horse’s watering trough
of this kind in his imagination when he wrote the line “Bellicus
ex illo fonte bibebat equus.” In the legend, as it was known to
Propertius, Tarpeia drew the water for her water-jug at the
fountain-head, whence it ran down the slope to the troughs
in the Forum where the horses could drink.

The building .as it now stands was not consecrated as St
Peter’s prison before the 9th century (Hemans, Monuments of
Rome, p. 110), and the name Twllianwm was possibly derived

=

a

i a i el

ARCHHOLOGICAL INTERPRETATIONS. 5

from the ancient structure having been a well-house. Festus
gives an interpretation of the word Tullius as meaning a stream
of water. (Festus, p. 353 ed. Miller.) He quotes from the
Ajax of Ennius an instance of the use of the word tulli
corresponding to the ovpuyyes “arteries” of Soph. Aj. 1412
and the avAds trays, “gush of blood” of Hom. Od. xx1. 18.

Corssen Lat. Spr. 1. p. 171 connects the name Tullius with
the root tol tul, which would seem to point to the meaning
‘“‘a place for drawing water.” Tiburtes tullias occurs in Plin.
N. H. xvi. 16. 26, where tullcas probably refers to the well-
known cascades of the Anio at Tibur.

3. Martial vit. 75. 2,

A tecta Flaminiaque recens.

This passage has been wrongly explained by commentators
as referring to the tecta via outside the Porta Capena. Their
mistake arises from ignorance of the fact (well known to Roman
archeologists) that there were two tecte ww at Rome, as is
rightly stated in Forcellini’s Lexicon s. v. recta via. These
tecte vie were no doubt similar to the Porticoes at Bologna
which lead from the Porta Saragozza to the Madonna di S. Luca,
and from the Porta Maggiore to the church of Gli Scalzi. One
of them was placed between the Porta Capena and the temple
of Mars outside the walls, and was intended to secure a con-
venient passage for the votaries who went to dedicate their
armour at the temple after a successful engagement or a happy
escape. This custom is referred to in Propertius Iv (v). 3. 71,
“Armaque cum tulero Portz votiva Capenz subscribam salvo
grata puella viro.”

A grand procession of the Order of Knights (transvectio
equitum, Liv. 1x. 46; Ov. Fast. vr. 191) was conducted on
the Ides of Quintilis every year from this temple of Mars to
the Capitol to give thanks for the aid of the Dioscuri at the
battle of Regillus, and other religious ceremonies were carried
on just outside the Porta Capena, as for instance those con-
nected with the Lapis Manalis. Thus a considerable traffic
similar to that which passes through the Porta Saragozza at

~

6 THE JOURNAL OF PHILOLOGY.

Bologna, passed along the road between the gate and temple,
and was sheltered by a via tecta.

The other path which was sheltered by a portico was on the
Campus Martius near the Flaminian road between the old
Porta Ratumena and the later Porta Flaminia, now the Porta
del Popolo. This latter is the via tecta to which Martial refers
in vill. 75. 2, as quoted above, where he connects it with the
Flaminian road. A further description of it as the Via For-
nicata ad Campum is given in Liv. XxIL. 36. 8, shewing that
it was a vaulted archway. The Roman house of Martial’s
friend Julius Martialis seems to have been near this via tecta,
for his villa was on the Monte Mario to which the approach
from the Flaminian road would be much nearer than from
the Porta Capena. See Mart. m1. 5. 5, “hune queres prime
in limine tectie,” and Mart. Iv. 64. 1, “Juli jugera pauca

Martialis Longo Janiculi jugo recumbunt”; and 23, “Cum sit.

tam prope Milvius.” Seneca, Apol. 13. 1, speaks of this via
-tecta as near the altar of Dis, where Claudius descended ad
inferos. Claudius was buried in the Mausoleum of Augustus,
which was near the southern end of this wia tecta.
Gallienus (Hist. Aug. Gallien. 18) seems to have designed
a grand extension of this via tecta to the Milvian bridge.

4. Martial rv. 18. 1,

Qua vicina pluit Vipsanis porta columnis.

A mistake has been made in the interpretation of this line
by commentators from their ignorance of the position of the
Roman aqueducts. Schrevelius in his variorum edition of

Martial’s epigrams, followed by Paley and Stone, identifies the -

porta here mentioned with the Porta Capena, because the
Porta Capena is in several passages of the Roman poets called
madida from the Aqua Marcia which passed over it.

But the porta here spoken of was ‘vicina Vipsanis co-
lumnis, and the ‘Porticus Vipsania’ was probably a portion
or a continuation of the Porticus Pole built by Pola, Agrippa’s
sister (Dion Cass. Lv. 8), and perhaps connected with the
Porticus Europ (Mart. 11. 14. 3). This colonnade was in sight

"a So oui s ° ft we SEO ee ee
% ‘ae ? : es eel hae

g \ he ee ob mentee? iy Ui

2 ~ = “~~ * a ves bie
;

ARCHAOLOGICAL INTERPRETATIONS. 7

of Martial’s house on the Quirinal (Mart. Ep. 1. 108. 3), and
we may infer from the name Vipsania that it was a part of
the buildings of Agrippa near his Thermz on the Campus
Martius. Thus the water which dripped from the porta must
have been that of the Aqua Virgo, from which Agrippa sup-
plied his baths.

And it appears that the Aqua Virgo was carried over the
arch of Claudius discovered in 1650, which crossed the Via Lata,
now the Corso, near the Palazzo Sciarra, and was built in
memory of Claudius’ expedition to Britain in A.D, 43. Another
archway was also discovered at the same time by which the
Aqua Virgo was carried over a street parallel to the Via Lata
but nearer to the Pantheon. One of these two archways is
plainly alluded to by Martial, and thus the position of the
Porticus Vipsania is approximately determined. The extent
of this Porticus is shewn by the fact mentioned by Tacitus
(Hist. 1. 31) and Plutarch (Galba 25) that troops could be
quartered in it,

(To be continued.)

R. BURN.

ON CERTAIN ENGINEERING DIFFICULTIES IN THU-
CYDIDES’ ACCOUNT OF THE ESCAPE FROM
PLATAA. Br. m1, 20—4.

A CAREFUL consideration of the account given of this event

by Thucydides, aided by modern research, will, I think, tend

to throw considerable doubts over his accuracy, in this respect
at least, as an historian. J am not aware that such doubts
have been raised in any history of Greece, with the exception
of that by Sir George Cox’. He has put forward, at some
length, in an appendix, the views I propose to bring more fully
before the reader; and in doing so, he has acknowledged his
obligations to me as the author of the doubt. The question is,
I think, both interesting and important; for the character of
a°great historian is impugned, if not for truthfulness, at least
for accuracy.

In the year B.c. 429, two years after the outbreak of the
Peloponnesian war, the Peloponnesians, led by King Archidamus,
marched against Platea, This was done at the desire and the
instance of the Thebans, who were the implacable enemies of the
Platzans. As the Platzans, advised by Athens, had rejected the
offer of remaining neutral, their city was at once invested by the
Lacedzemonian army. And here follows, in Thucydides 11. 75—8,
a most interesting and circumstantial account of the siege,
_ which it is necessary to my purpose to epitomise.

The first operation was to barricade the city all round with
trees cut from the spot, to prevent any further egress of the
citizens, tod pndéva ete éEvévar. The next was, to carry a
mound on an inclined plane to the level of the top of the wall.
This mound was raised only against a portion of the city wall,
the object being simply to effect an entry. For the construc-
tion of it, wooden piles and fascines, stones and earth, were
used; and when the besieged made mines under or holes

1 Hist. of Greece, Vol. 11. App. K. pp. 603—6.

eat i ee dl

i nt Oi

on

THE ESCAPE FROM PLATZA, 9

through the bottom of their own wall, and carried off the loose
soil from the mound into the city, so as to cause a continual
subsidence of the enemy’s earth-work, the Peloponnesians had
recourse to reed-baskets filled with clay. As the mound rose in
height, the Platzans kept adding to the height of their own walls
by a hoarding or frame of timber filled in with brick. For seventy
days and nights the whole Lacedemonian army worked con-
tinuously by relays, cat’ avatravias. The Platzans, fearing they
should not hold out, left off raising the wall in this part, and
worked at a crescent-shaped inner wall, commencing from the
lower or unheightened wall on either side, and curving inwards
into the city, in order that, if their great or outer wall were taken,
el TO péya Teiyos adioKotTo, the secondary or crescent-shaped
wall might still hold out for a time. The Peloponnesians now
raise engines on their mound to batter down the Platzan timber
frame or upper wall. This device seems to have been a batter-
ing-ram, a contrivance well known to the Assyrians in much
earlier times. Mr Layard, in his ‘Nineveh,’ p. 217 of the
smaller edition, describes a bas-relief in which the besieged are
endeavouring to catch the ram by letting down a chain from
the wall. In p. 255 of the same work, he says, “the battering
ram was rolled up to the walls on an inclined plane constructed
of earth, stones, and trees, which appears to have been some-
times paved with bricks or squared stones to facilitate the
ascent of the engine.” He adds, “this mode of besieging is
frequently alluded to in Scripture.”

According to Thucydides, the Platzans adopted very similar
measures to prevent the ram making a breach. They let fall
_ heavy beams suspended by chains, and knocked off the heads of
the rams as-they were going to strike. The Lacedemonian
army next tried fire, hoping, says the historian, that if a wind
arose they could set ablaze the city, which was of no great size,
ovoav ov peyadny, 11. 77. They wished, he says, to induce the
city to surrender without the expense and delay of a long siege,—
to frighten the Platzeans in fact, by a strong measure which
should leave them little hope of mercy if they persisted in
holding out. To this end, they piled up the spaces between
the wall and the crescent-shaped barricade with faggots, which

IO THE JOURNAL OF PHILOLOGY.

_they threw in from their own mound, and even for some little
way over or beyond the crescent, and’ into the interior of the
city. Then, setting fire to the heap with brimstone and tar,
they made, says the historian, such a bonfire as had never been
seen up to that time, unless perhaps in the accidental burning
of a wood. But-he adds, with a slight touch of the marvellous’,
“it is said that a thunderstorm with heavy rain descended at
the time and put out the fire.”

The siege at length was turned into a blockade. Archida-
mus appears to have retired with part of his army; for the
historian says, a considerable part, ~épos te Tov otpatomédov,
was left, and that these, the remnant of the force, proceeded to
wall round the city, assigning the work in different Pants to
different cities or bodies of ane, mepiereey TV TOAD
KUKNO, SLEACMEVOL KATA TOAELS TO yoplov.

This wall, he tells us, had a double ditch or moat, both
inside and out, from which they had dug the bricks to make
the wall: tadpos 5& évtos te Hv kal &EwOev, é& 5 emdwOed-
cavto. The whole of this work was finished, he adds, wép
é£eipyaoto, at the rising of Arcturus, that is, at the autumnal
equinox. Now, as the expedition had commenced in spring
(rod émuytyvopuévou Oépous, chap. 71), and seventy days had been
spent in the fruitless working at the mound, we are met at once
by this startling statement,—that in less than three months an
investing double wall of brick, with a double moat, was raised .
all round the city, and even built of materials, wAiv@ou, that
had to be dug and made (either burnt or sun-dried) on the spot! —

Now, though the historian does say that the city was not
large, ov peyadn, we know certainly that it was by no means
small. For there were 400 fighting men left in it, besides
eighty Athenians, and 110 women to make the bread for the ~
besieged (11. 78). All the rest of the women, with the children,
the old men, and the unserviceable population, aypetov 7A Gos,
meaning probably the slaves, had been sent by the Platzans
for safety to Athens. According to the usual average of fighting
men to a whole population, there could not have been less than
10,000 inhabitants. At the battle of Marathon, some 60 years

1 Compare the story in Herod. 1. 87.

THE ESCAPE FROM PLATA, II

before, we read of a contingent of 1000 Platzans; at the battle
of Platza the éaAiras numbered 600. Now, let us consider if
any engineer, civil or military, could by any possibility perform
such a feat in three months as to surround and inclose so large
a town with a wall and a deep double moat! With the number
of hands engaged, it is manifestly and certainly quite impossible.
Such a feat can only be compared to the stories about devil's
dykes and mounds miraculously raised in a single night. Least
of all were the slow and inert Lacedzmonians likely to execute
so stupendous a work in so brief a time.

We happen to know, from existing remains, the extent of
the city wall of Platza. Colonel Leake tells us, in his Travels
in Northern Greece’, that it is 24 miles round. Of course, the
investing wall with its double moats must have been much
larger, whatever may have been the circuit of the ancient city.

But what puts an end to the question of possibility is the
minute account given of this very wallin the 20th and 21st chap-
ter of the Third Book, where it is to be observed that it is in three
separate places distinctly called the “wall of the enemy,” and “the
wall of the Peloponnesians.” It was actually a double wall, with
an interval of 16 feet between. It was so high that it had to be
scaled by ladders; it had battlements, and at intervals of every
ten battlements, dva déxa érrar&éwrv, there were wide and large
towers, spanning right across to the very outside of each wall,
and therefore about as large as ordinary church towers! He
calls them aupyot peyarou kat icomdatets TH Teixet. There
was an arch or entrance in the middle of each, and they were
roofed above, dvwOev creyavoi. The intervening space, he says,
was used as a series of guard-rooms, tots fvAakiw oiknwara
@xodounto. Who these dvdaxes were, we know from I. 78;
they consisted half of Boeotians, that is, Thebans, and half of
Lacedzemonians, though even the pépos tt, or part of the army
that is said to have made the wall, had retired, aveywpnoar,
and been disbanded.

But still further; the whole of this wall, say, two miles in
circuit, was plaistered internally with rough-cast—in itself a
most gigantic task to perform! How could an invading enemy

1 Vol. 1m. chap, 16,

12 THE JOURNAL OF PHILOLOGY.

possibly do this, standing on scaffolds or ladders, without oppo-
sition; and above all, where did all the plaister come from?
Thucydides says, that in order to calculate the height of the
walls by the courses of brick internally, the Plateeans who in-
tended to escape counted the bricks in a part 7 éruye mpos
opds ovn éEadryrAppévov To Tetyos avta@v (that is, TO Tetyos
TOV ToNeuiwy),—words which Dr Arnold incorrectly renders
“thoroughly white-washed.” He should have translated, “in
a part where the plaistering had not been completed.”

Thus then we have some 500 fighting men cooped up in a
double wall of some two miles in circumference, and guarded
by a handful of troops who had their lodging in the double wall!
This alone is quite incredible; for if an army had come from
Athens to aid the party within, these dvAaxes would have been
in a very unsafe position between the Athenians without and
the Platzeans within, and must have surrendered very shortly if
only by a failure of supplies.

The question now arises, is the whole story about this wall
a fiction, or is it the result of an extraordinary and almost in-
credible blunder on the part of Thucydides in confounding the
ancient city walls with a new wall built on purpose to prevent
the escape of the besieged ?

Although there are very great difficulties in either supposi-,
tion, I shall endeavour to show that the latter is the case. The
evidence of modern travellers will go some way in determining
this. But first 1t may be remarked, that we happen to possess
an ancient stone sculpture, or drawing as it might be called,
found by Sir Charles Fellowes at Pinara in Lycia, and of a date
certainly not less ancient than the siege of Platza, with battle-
ments and towers on the walls precisely as described by
Thucydides. Secondly, walls such as those here spoken of as the
siege-works of the enemy existed commonly till quite late in the
middle ages. If I remember aright, the city of Nuremberg still ©
retains its old walls, said to be crowned with 365 towers. I say
“crowned,” bearing in mind the passage in the Cédipus at
Colonus of Sophocles’, who speaks of auvpyor mév of modw

1 Ver. 15. The MSS. indeed give oréyovow, but it seems probable that

arépovow is the true reading.

a
7
q

“4
|

THE ESCAPE FROM PLATA. 13

_ otépovew, in speaking of the-wall then surrounding Athens,
traces of which, with towers, still remain.

The most complete account of the old walls of Platza is
given in Dodwell’s Classical Tour through Greece, Vol. 1. p. 277
—80. The ruins, he says, of the city stand on a low oblong
rock, the narrow extremities of which face north and south, the
longer sides closely corresponding with the still existing re-
mains, Being enormously strong (even without being double),
they might have afforded, in the flanking towers, a safe position
for the warders or dvAaxes of the beleaguered city. They were
supported too by 300 reserves, who are said (111. 22, 7) to have
been posted outside the outer moat to bring immediate aid, if
necessary, ols érétaxto mapaBonOety et Tu déor.

That Thucydides did not know the locality, and fancied that
the city wall was encircled by another double wall, is further
shown by his using é£eA@ovtes in chap. 22. He there says
distinctly that the.Platzans who remained in the city went out
and made an attack on the Peloponnesian wall on the opposite
side to that on which their friends were escaping. This must
mean that they went out of their own city walls and attacked
the Peloponnesian wall of circumvallation. Moreover, he fancied
that the Spartan force was encamped somewhere between the city
wall and their own double lines; a force over and above the
dvraxes and the 300 reserves. For he says that when an alarm
had been raised by the falling of a tile, in the escape, the be-
sieging army rushed to the wall, dpyncev éri ro tetyos. Nor
do I think it possible to understand these words in any other
we

Viewed under any aspect, the matter is full of difficulties.
It is possible that all the walls now remaining may be later
than the siege of Platza; for Thucydides says the Thebans
“destroyed the whole city to the level from the foundations,”
KabeXovtes avtTny és €dadhos Tacav éx TOV Oeperiwr, IIL. 68, but
he does not specify the walls. The city is said to have been
rebuilt after the peace of Antalcidas in 387, and again de-
stroyed by the Thebans in 374, Again it was restored under
the Macedonian supremacy, and lastly its walls were rebuilt by
Justinian.

14 THE JOURNAL OF PHILOLOGY.

Thus great doubt is thrown on the identification of any of.
the present remains with those existing in the time of Thu-

cydides. If those walls have wholly disappeared and been
replaced by others, his account may be substantially true,
provided we understand it of the city walls and not of the
circumvallation, which, I think I have clearly shown, must be
given up as false in fact, because involving an impossibility.

It is likely that this old city wall was the same, as regards
circuit and foundations, with those walls that now exist, and
which Dodwell attributes to the Macedonian era. The old
foundations he speaks of may be those left by the Thebans
when they razed the town to the ground; and the towers now
standing may have been rebuilt after the fashion of the old
ones,

With regard to the brick dug out of the moats, of which
Thucydides says the walls were built, if there is clay on the
spot, it may be said that sun-dried bricks could have been
made with straw or stubble, like those which are used about
Cambridge made out of the chalk marl, and called “batts.” If
there is no clay, then of course the whole story of the bricks is
a fabrication. It would be very interesting to have this matter
settled by a geological examination of the site, which is said to
be on a rock.

My impression is, that Thucydides was so far led away by
the habit of the Xoyoypadort to compose amusing or sensational

' stories—abundance of which we find interposed in the graver

historical narratives of Herodotus—that he here indulged his
hearers with a very exciting story of a hair-breadth escape, and

' was really more intent on making the story a good one (in

which he has certainly succeeded) than in a careful investiga-
tion of the truth. It is not to be denied that in his preface
(1.21) he denounces the Aoyoypadot who compose stories merely
to please, é7i TO mpocaywydTepov TH axpoaces 7 adnPéoTepor,
but then this preface was written last, and it may express the
feelings of the more matured historian, although some approach
to the romantic and the marvellous had been attempted in the
earlier writing of his history. |

There may have been a desire too to insert an exciting and

THE ESCAPE FROM PLATA. 15

amusing episode to counteract the dullness for which he him-
self apologizes in I. 22, where he expresses a fear that some
may think the non-mythical nature of his history somewhat
uninteresting, és axpoacw lows TO pw) pvOades avTadv atep-
mwéaTepov paveiTar.

It will no doubt be objected that Thucydides, writing so
nearly at the time of the event, and a native of Athens, could

not have been so misinformed about so important a town so

near and so friendly to Athens. But Platza had been utterly
destroyed by the Thebans about B.c. 425; and if Thucydides
wrote or published his history twenty years later, when the
town was in utter ruin, such misconceptions are by no means so
improbable as they may appear.

That certain prisoners of war did escape and get safely to
Athens, no one will deny. But we may be pretty sure their
version of the story would not be derogatory to their own
bravery and cleverness.

There is really no difficulty in supposing that the Spartans
had taken possession of the city wall and cooped up the inhabi-
tants within it till they should surrender themselves through
starvation. 'The motive for this is evident, viz. because in the
event of a peace they would not be compelled to restore any
city that had given itself up, but only those captured by force.
This is distinctly stated in 111. 52: “the Lacedemonian general,
aware of the weak state to which the Platezans had been
reduced by the famine, was unwilling to take them by assault,
orders to spare them having been sent from head quarters, in
order that, in the event of a truce, and a restoration on both
sides of the places taken in the war, Platea might not be ceded
to Athens, as having voluntarily come over.”

The double character of the wall, described as spanned by
towers, may be a mistake resulting from some kind of mound
or barricade raised round the city wall on the outside, and the
wonderful story of the deep moat and the frozen water very
likely arose from that of the ditch out of which the mound had
been dug. But if there ever were moats, there must still be
vestiges of some of them.

F. A. PALEY,

ne eee eS ee oe ee ee ee
7 ™

ON THE FIRST SEVEN VERSES OF THE ANTIGONE.

I sEEM to myself to have found a probable solution of the

great difficulties which beset this passage.
I suppose vs. 2—3 to have stood thus originally;
ap oio@ bts Leds tov am Oidimov Kaxeév
ov« éof dtroiov ov>xi vev Edoaw Tere ;
By way of comment, érs was written above dca, this being
the more usual formula, as we have éts Cocav yapeis in v. 750,
étt Sav Ajax 990, and éts f®cav proya in Bacch. 6. But when
the gloss had crept into the text and so made a verse of seven
feet, ovx éo@ was omitted, and the present-reading, which is
nonsense, and cannot be translated, was the result. In other
words, the actors had to choose between rejecting ov« éo6’,
or étt. }
For many years I have held the opinion, that the three next
verses are an interpolation, and that for three good reasons;

(1) The seven verses of Antigone should correspond numerically —

to the seven of Ismene. (2) The words ozrotov ov—kaxdév are
a mere repetition from the preceding sentence. (3) The double
negative, 67rotov ov—ovx dmrw7a, if defensible in itself, seems
due to the same kind of pedantry which intended ovr’ drys
dtep in v. & to stand for ote ovK adtep arns, i.e. ovte Evy
adtn. None of the proposed corrections of this verse seem to
me in the least probable. It is radically bad, together with
the other two lines. The negative ov repeated seven times in
three verses can hardly be attributed to Sophocles, who could
so easily and naturally have used noOouny for ov« d7rwra, ‘of
which I am not fully aware.’ a

Wwe i et oe eS

i Ste. ~

FIRST SEVEN VERSES OF THE ANTIGONE. 17

But why were these verses interpolated? I fancy I can now
give a plausible reason. It was to represent in another way the
ov« éott which had been wrongfully excluded from v. 3, and
was taken into protection by one school of actors. This is now
developed into ovdév yap éott TOV KaKOV OTroloY OUK OTwTrA.
It is evident, not to say certain, that 7f the poet intended in
v. 3 the syntax I have suggested, dp’ oto Ott ovdév eote THY
Kkakov omotov Levs ov tercz, he could not have immediately
added, ovdév yap éort TAV KaKaY OTrOtoV ovK BTrwTra.

As a matter of Greek grammar, it seems to me necessary
that ov« éo@ dzrotov should have commenced vy. 3.

If my reasoning is right, what a curious “muddle” has been
caused by the interpolation of the little word éru?!

1 In Oed. Tyr. 1401 there isa doubt Phil. 416—8, (I am aware, of course,
between dpa wou wéuvnod’ Sri ol’ Epya oof the interpretation suggested by

Spacas, and péuvncd’ érx. Remarkable Professor Kennedy on Oed. Tyr. 328,

examples of a redundant negative oc- where I still prefer my correction ray’
cur Oed. Tyr. 328, El. 626, Trach.158, is ay elrys.)

F, A. PALEY.

Journal of Philology. vou. x. 2

ON SOME FRAGMENTS OF THE NEW COMEDY, AND
SOME PASSAGES OF AESCHYLUS, THEOGNIS.
ALCAEUS AND IBYCUS.

THE following notes on the 4th volume of Meineke’s Frag-
menta Comicorum Graecorum, containing the fragments of the
New Comedy, may, I hope, be not without interest, if not for
themselves, as illustrating Catullus. I shall begin with exhi-
biting some parallelisms of expression or idea. Such resem-
blances are especially likely to be found in Menander, the one
subject running through all whose dramas like a common breath,
was love (Plut. wept "Epwros ap. Stob. Flor. 63. 34).

Cat. LXXVI. 13, 14

Difficile est longum subito deponere amorem,
Difficile est.

Menand. Kapynédovioz fr. 11. M.

Epyov €k paxpov ypovov
dvolay nyépa pEeTATTHoaL pla.

Menand. Inc. Cxc1IL

épyov éoti, Pavia,
paxpav ovvnberav Bpayet AVTAaL ypove.

The repetition Dafficile est, Difficile est is a natural one and
has its counterpart in a fragm. of Apollodorus of Carystus Ine.
v. M.

YanreTrov TUYN 'oTL TPAypLa, yareTOY adrad Set
avTnv dépew Kata tTpoTrov, wamep optiov.
The combination of past, present, and future in the recurring

formula quot aut fuerunt Aut suat aut alvis erunt in annis is
found several times in the New Comedy.

" ee ee ee

- wee

’

FRAGMENTS OF THE NEW COMEDY, de. 19

Menand. Ine, x11. M.
pOovos

Doicvxov weroinke Kal troinoer Kal Trove.

Philem. Inc. u. M.
dv ovde els AEANOEV OSE EV TrOL@V
a> / 7O\ \ ,
ovd av ToLnowY ovde TrETTOLNKWS TraXAat,

to which may be added from Euripides Troad. 467
Tropatov yap d&a
Tacx Te kal Trémov0a Kate Trelcopat.

With Cat. xxi. fin. compare Menand. Inc, LXxxv.
ovoels éf avtod Ta Kaka ovvopa, Iapdure,
cadas, étépov S doynpovodvtos drperar.

With Cat. vil. 3 Fulsere quondam candid tibi soles, compare
Zenob. vi. 13 Mévavdpos 5é dnow ev Aevxadia tv ayabnv
nwépav AevKnv Kareto Oar.

Menand. Ine. vit.
Ma tv ’AOnvav, dvdpes, eixov’ ovK exw
evpely dpmolay TM yeyovoTs mpayyaTt,
EnTav mpos euavtov ti Tayéws arroAdver.
aTpoBiros; é€v bow ovoTpépeTat, TpocépyeTat,
mpoéraBev, é&éppurev, aiov yiyverat.

The word é&é$pu.ev has a strange look, and can hardly, I
think, be genuine. I believe the real word is é£épéy£ev which
is used intransitively of a subterranean wind bursting forth by
Aristotle Meteorol. 11, 8, p. 366. 31 of the Berlin ed. ”“Hédn yap
aevopos ev TOTALS TLTL yLVOmEVOS OU TpoTEpoy EdnkE, ply Expy-
Eas eis Tov Umép yis TOTOV havepas watrep exvedias €EHrOev o
KLVnTAS AVELOS.

Menand. Ine. vii. M.
Tape? yap nuav ovdé eis ef pr) Séx” 4
évdexa yuvaixas’ dadex’ 7) mAclovs Tivés.
“Av téttapas 8 7 mévTe yeyapnkos Try
+Kkatactpodytis, avupévatos aOXdos
dvuppos odTos émiKanreit év Tols éxel.

e |
20 THE JOURNAL OF PHILOLOGY.

Meineke reads after Tyrwhitt t’yn xatactpodjs tis’ ap-
parently in the sense of ‘dies’ Mr Lancelot Shadwell, in his
‘Commentary on the Gospels of Matthew and of Mark’ 1861,
conjectures Kata otpady tis ‘which alters only half a letter,
and exactly agrees with the context. The poet had just before
said

yapuel yap nuav ovdé els ef pr Séx’ 4
évdexa yuvaixas, dédex 7 mrElovs Tivés.

He then adds, But if any one shall take four or five wives,
and stop there, they call him a miserable old bachelor’ (pp. 58,
59).

Menand. Inc. xv. M.

Plut. ap. Stob. Flor. 63. 34 rév Mevavdpou Spayatav opards
ATAVTOV EV CUVEKTLKOV eOTLY, 0 Epaws, olovy TrEeDUa KoLWOY SLaTrEe-
, A > q ¢ x 5 / 0 , a A) a) \
duxas. dv odv (205 av ovv) wadoTa OvacedTnv Tod Oeod Kat
dpytactiy Tov advdpa cupTepirAapBavopev eis THY Entnow, érel
Kal NEAAANKE TEP TOD TaOous PirocopasTepov. aEvov yap Eeivat
t hese \ let Sa ef ” ef A
Oatpatos dynol To repli Tods ép@vtas, tdaTrep Eat, dua rarer}.
a a \
celta atropet Kal Entel Tpos éEavToV

tive SeSoUAWTAL TOTE ;
bf U fal \ 2, / x
over; pAvapos’ THS yap avTHsS Tavtes av -
” . / \ \ , + ”
npov’ Kpiow yap TO Brerrew tony exer.
3 3 ¢€ i, \ > lal > ,
GXN dovn TLS TOUS Ep@vTAas émayeTaL
a ‘df
cuvovoias; mas ovv eTEepos TavTny ~xov
ovdey wémovOev, AAX anHnrOe xatayedrar,
’ ,
ETepos ATOAWAE; KaLpos €oTLVY n VOTOS
es eee Ye .. s y 5) ,
apuyns’ 0 mAnyels O Felow 5) TITpOCKETAL.

‘Multorum coniecturis uexatum est’ says Cobet of this pas-
sage Nov. Lect. p. 82; if therefore I err in my attempt, it is in
good company. And first it seems pretty clear that the words
aEvov yap Savparos are part of the quotation, and on this
hypothesis they have been variously altered, not very happily so
far as they are known tome. Iwouldread —

ov \ /
akvov yap Oavpatos
TO TOV €pOVTwY, ois TapEeTTLV, & pTroLeEl.

FRAGMENTS OF THE NEW COMEDY, ée. 21

‘Strange enough is what happens to men in love, all that
he (Love) produces in those that are under his spell.’ The
change of nominative to "Epos is easily intelligible, and if any-
where, in Menander.

For e’c@ 51) which Bentley altered to eicBorn, Wyttenbach
to eis 0 Se?, Hermann to eis o6/, a marginal annotator of the
Bodleian copy of Bentley’s edition of the fragments has written
eis Any, i.e. I presume, wuynv. This would make excellent
sense if we interpret xavpos a mortal blow: ‘the disorder (love)
is a mortal blow to the soul, the stricken victim is wounded in
every fibre of it.’

Menand. Inc. xxx. M.

Kal ToUTO OWwv ovderrw@troT evEapnv

éy® TO cofov Tv éunv oikiav

GAA TapéAtTroY oiKETOV Elval OTACLW

évdov Tap avT@® Tpaypa ypnowmdtarov.

The paraphrase of this passage given in the Venetian Schol.

A. on Il, xxi. 389 ovdérote Ovwv evEaunv Gdr(ws coler Oar THY
oixiay ) otacw oixeTév evar év avTH as well as the excerpt
from Plutarch Cat. Mai, 21 quoted by Meineke adel 6é twa
ortdow éyew Tovs Sovrovs eunyavaro Kai Siapopav mpos adXd-
Rous, UTovoay THY Omovolay Kal Sedorxas make it probable that
we should read 76 c@fov tTHv eunv av oikiav,’AXN 1 tapédeTrov"
‘numquam id uoui quod domum meam seruaturum esset, nisi ut
relinquerem etiam (without reserving) ut inter seruos fieret
dissensio.’

Menand. Inc. Lxx1x. M.
Tovs Tov lovov SaTravevTas aroylaTas Biov

TO KAN@S aKkoveLw TAXD TOLEL TAC KAKOS.

For wdow Bentley conj. rewfv, Meineke radu or macyeuv.
Is not mpaocew a simpler emendation?

Apollodorus ’Emidcxalouevos i. M.

From the words of Terence on which this corrupt passage is
quoted by Donatus it seems probable that NAAKEIs or Avadxeus
represents év Néoyn.

22 THE JOURNAL OF PHILOLOGY.

Anaxippus "Eyxadumropevos 1. M.
tov dpOpov év tats yepaolv der BiBrILa
éyovta Kal Sytobvta Ta Kata THY TéxyvnD.
ovdev t+yovdpevovar Siahépw T *>Aorrevdiov.

I explain this of the Aspendian salt with the white and
withered look of which the studious and bookish cook is com-
pared, Plin. xxx1. 73 Sal siccatur in lacu Tarentino aestiuis
solibus...item in Sicilia in lacu qui Cocanicus uocatur et alio
iuata Gelam. horum extremitates tantum inarescunt, sicut in )
Phrygia Cappadocia Aspendi, ubi largius coquitur et usque ad =a
mediwm. I would read then ovdév te yovdpou Siadépw 7 *Ac- :
TeEVOLOV.

Poseidippus "AvaB Xé7revr I. M.
€ \ \ a > b] ’
6 Sé THY YAOTTAV Els aoxXNmovas
b] / ” / a ¢ ,
éemuOvpias +évid Te TOY NOvapLaToV
Kaanros, Katofos, yvavoTiKos, TpoTKaVaTLKOS.

For évud te I would suggest éwiadre or épiadrAe. Od. IX.
288 "AXN’ 6 y aval~as érapows eri Yetpas tadnrev.

Aesch. Supp. 615.
toiavd Erevbe pnaw apd npov eyo
v a e / \ Ud
ava& IleXacyeav, tkeciov Znvos KeTov
peyav Tpopwvav pHnTOT EiooTLY ypovou
Tow Taxydvat, Eevixdov aotixov 0 apa
Aéywv SutrAovv placa Tpod Torews Haver
aunyavou Booknpa mnmovijs médeww.

Conington, quoted by Paley, who translates ‘warning them
that the great wrath of Zeus would never hereafter enrich the
city, thought the idea was of a disease draining the body politic,
exhausting its powers of support, and preventing wt from thriving
or becoming fat. This has always seemed to me the exact reverse
of the poet’s meaning. ‘The citizens are warned against allow-
ing the wrath of Zeus Hikesios, incurred by neglect of the sup-
pliant Danaides, at some future time to make the city wax gross
and «increase the unwholesome humowrs in the body politic. The
same metaphor underlies the words 6ABos ayav mayvvOeéis

FRAGMENTS OF THE NEW COMEDY, ¢€e. 23

‘prosperity swoln to an unhealthy bigness.’ And this is surely
the natural meaning of rpodwvév pnort’, ‘warning them that
the anger of Zeus should not make the city grow fat,’ i.e. against
allowing it to do so,

857. aryetos éyo Babvyaios Babp & as Babp é as yépov.

So Med. I cannot help believing after all that has been
written about d@yevos that it is simply "Apyeios: cf. 274° Apyetar
yévos ’EgevyouerOa, and the king’s incredulity 277 “Amora
pubeic®, & Eévar, wrwvew éwol,”Orws TOS dyiv éoriv ’Apyetov
yévos, and especially 322 where the Chorus having proved their
ancient connexion with Argos conclude with saying eides 8
apmov apxaiov yévos IIpaccors av ws Apyeiov dvatnons oTdXov.

Suppl. 875.
Ot OC Of Of OF A-Up-aciao UTP Oo yaouv hacker
Tepe yauTTa Bpvaters 0 ‘ao ép @ Tas.

I think this mysterious passage contains two Egyptian
allusions (1) to Jszs, who is invoked to witness the wrong of the
Suppliants (2) to the crocodile, called in Egyptian yaprpa
(Herod. 11. 69). We may suppose a number of Egyptians by
this time to have gathered round, perhaps wearing the national
ear- and ankle-rings, and gesticulating violently: it would be
natural to the Suppliants bred in Egypt and remembering the
similar adornment of the sacred crocodiles to compare them
with their pursuers: hence I would read

Ol Of OF OF Ol,

Avpas, "lot. Tpo yas vAACKOL,
mTepl, yaprra, Bpvaters,

b0° épwras.

‘Alas for the outrage, o Isis. I would he might howl beyond
the limits of the land. Unmeasured is thy wantonness, thou
crocodile, that thou askest so many questions. With zrepi
Bpudfas we might perhaps compare Nicander’s zrepsSpuns twice
used in the sense of over-luxuriant. 60° for 65 as in ’OAPia, daca
locate’ tavorBia, bs yAveupwve’ seems so natural that I marvel
it should not have been suggested before.

24 THE JOURNAL OF PHILOLOGY.
Suppl. 987.

kab pnt adédatws Sopixavel popw Savor
AdOorut, yopa S aos aelSwv rrédor.

Weil and Wecklein rightly object to this, the former reading
unr && dérxrtov, the latter (Studien p. 88) supposing a verse
lost after 988. May not the right reading be «al pn’a adédarrous
‘and that I might not for some unlooked for reason be slain,
and the land thus incur a burden of undying guilt. The un-
looked for reason would be some sudden surprise, such as might
easily occur to a stranger.

Suppl. 996 sqq.
vuas 8 érawe py KaTarocyvve pe,
dpav éyovaas tTHvd énlotpertov Bporois.
Tépew OTwpa & evpvrAaxTos ovdapos.
Opes 5& xnpaivovar Kai Bportol, tl pv;
Kal Ky@oara TTEpovVTa Kal TredooTLPH.
KapTopata otalovta Knpvace Kurpis
ely

+karw p akw Avovoav Owopév Hv Epa.

It seems probable that the obelized line contained some par-
ticular application of the remark just made. ‘I warn you to
keep your virginity jealously guarded. Fruits just ripe are a
desirable thing, and attract the attack of beasts and men.
Venus gives notiee when the grape is over-juicy, and who has
not seen the fox watching the vine?’

I suggest then
KadwTéx apevovcav (avOocpnv) épa.
eal

The objection to the former of the two bracketed words
is that it is not known to exist. But that it may have existed
as the original word from which the well-known av@ocpias
(otvos) took its origin, is I think likely, because both Suidas
and the Schol. Arist. Plut. 807 say av@ocpuias was formed aad
torov AvOoculov 7 ws amo eidous adutrérov. It is difficult to
believe that the wine was called av@ocpias. As in the case of
xamrvias, the vine would be likely to have a different form.

FRAGMENTS OF THE NEW COMEDY, ée. 25

Schol. Vesp. 151 xcamvn eidos durrérov Enporarov Kal Spipvtatov
olvov TroLovans, Omolws KaTTV@® TrovovyTa SaKpva. TLves O€ KaTrVlaV
olvov év BeveBévt@ ths “Itadlas yiverOai hac, kab xarvia 7
aumedos. As then the vine seems to have been called xaavy
xaTrvia as well as Kamvios Kamveos KaTVEews, but not Kamvias, so
it might be expected the vine would be called av@dcpn avOdc-
pews or by some other similar form, but not avOocpias. The
fox is frequently mentioned as prowling about and injuring
vines, Theoc. 1. 457 audi Sé viv bv’ adw@rrexes, & pev av’ bpyws
Poth civopéva tav tpwkiuov, Nicand. Al. 185 Ilvorépnv ore
Botpuv écivato Knkas adodmnt. Babr. 1.11 ’Arwrén’ éyOpdv
dpuréxwv te Kal KnTov. With épe ‘I will tell of, introducing
an object of comparison, cf. Ag. 896 Aéyouw’ av dvdpa tovde TOV
oTabuav Kuva, Ywtipa vaos mpotovoy x.7.r. and 838 Eides
Aéyouw av, ev yap é€eTrictapat, ‘Optrias KatorrTpov, eldwrov
oxids, Aoxodvtas civas Kapta mpevpevets eéuot. The language
~ seems closely to resemble Chaeremon fr. 12 Nauck, [loAAnv
omwpav Kumpidos eicopav maphnv "“Axpatot mepkafoveay oivay-
dais ypovov.

Suppl. 350.

AUKOOiWKTOV Ws Sauarty am TéTpAaLs
aniBatous, tv add twicvvos péwuKe
dpafovoa Bothpr moxGovs.

Antigonus Hist. Mirab. 29 Keller tds 8& édadous Aéyeu
TLiKTELW Tapa Tas OdoUs, Pevyovcas Ta Onpia. Heiota yap émuti-
Pecbat Tovs NUKOUS évOaSe' dyew Sé Kal TA Téxva err THY oTab-
pov, €O.Covoas ob Set drropevyeuy" civas O¢ TOTO TéTpaV aTroppaya,
plav éyovoav odov. This description would apply exactly if we
might interpret dawadw not of a young heifer, but a young doe,
following Hesych. dauaris. wooyos Kal Kata TavTos véou.

Suppl. 795, 6.
aT pos-
SetKTos oloppwv KpEewas
yumrlas TéTpa.

Antig. Hist. Mirab. 42 Tumos 6€ Aéyeras tro tier bre
Oo \ Cs \ IO\ ger \ ae / \
ovoeis €0pake veoTTdOv ovdE veoTTElav’ O10 Kat ‘Hpddwpoyv tov

26 THE JOURNAL OF PHILOLOGY.

Bpvcwvos tod cogictod matépa amd Tivos avtovs étépas davat
ys elvat petedpov. tixtew 8 ody év amposBdrors wétpass.
Sept. c. Th. 576.
Kal TOV cov avOis mpocpopav adeddedv
éEurr7tadfwv dvoua, IloAvvelKous Biav,
dis 1 év TeXevTH Tovvow évdaTotvpevos
- KaNEL.

mpos popav Med. mpdcpopov most other MSS.

It does not seem to me at all certain that the reading of 576
is specially corrupt. Hermann indeed says ‘adeddeds alienum
est a trimetris tragicorum. But surely forms like édo/ato
twenty lines above (552) in a similar iambic passage, the recur-
rence of adeAgewy in a lyric passage 974, to say nothing of the
generally epic character of the play, are sufficient to determine
the form as Aeschylean, even if we had not the further plea that
as a gloss adeddov, not adeddeov, might have been expected.
Hermann quotes two scholia which to me are very sugges-
tive, Tov mpos Tov cov Bdvatov adedpdv, and Tov ovK éml gidta
évta got. On this view I would read

Kat TOV cov avOis pos wopov 8 adeAdedv
‘he calls on him, that is your brother, but brother as the doom
of death assigns,’ tov aderdedv pev, mpos wdpov Oé. Polynices
and Eteocles are brothers, but brothers in their doom, and in
the fatality which, fulfilling the curse of their father Oedipus,
sentences them to quarrel, as brothers, for the possession of
their father’s throne, and finally die by each other’s hand. The
whole of the conclusion of the play is a commentary on this
remark, cf. especially 930 of 8 do érerev- Tacav vm’ adradogo-
vous Xepoly Omoomdpotowv. 932 oudcmropor SATA Kal TavdreOpor.
811 Oras aderdais yepoly jnvaipovt dyav. Ovtas 6 Saipwv
Kowvos Hv aupoty dua. The A would easily fall out before A.
Sept. c. Theb. 705.

al f / .
vov 8Te, col TapéoTapev’ erred Saipov
a / \
AnHmaTOS aVTP OTTALA YpoVL ameTAadAaKTOS
, 3 VA /
icwo av €or Oa...A\w TEpwt
, fy a > al
m-vevpate’ vov © ére Eel.

7" =

FRAGMENTS OF THE NEW COMEDY, é&c. p ab,

So Med. ap. Merkel. The metre in 705 requires an alter-
nation of dactyls with cretics, as is proved by their regular
recurrence not only in the corresponding strophe 698—700,
but in the previous strophe and antistrophe 686—688, 692—
694. Hence Saiuwv would seem to be wrong, though the
passages quoted by Paley prove that it would be quite Aeschy-
lean as regards mere expression. Accepting GeAeuwtép@ from
Conington, I would read the rest of the passage thus
vov OTe cot TapéotaKev* émrel Sopuwv
Anmatos av TpoTala ypovia peTar-

Aaktos icws av Eau Oerdewwrépw
mvevpate’ voy O éte Cet.

>

and with ddéuev AnpwaTos tpotra/a ‘a changed wind in the spirit
of the house’ I would compare Cho. 1065 65¢ Tou weraOpors Tots
Baotreious Tpitos ad yeiwov IIvevoas yovias éredéc On.

On the interesting fragments newly published by Weil from
a papyrus of M. Firmin-Didot, the following suggestions have
already appeared in the Academy of May 1, 1880.

Brett. =
Perhaps
Kal Tpets ayavas, TpEls yuVvaLKetous TroVvoUs.
| ee eer ge

A€TITHTA PE ATTIC IHAHE TITS TPHME NHI.

Weil reads Xertn ydp éArmis nO ei Evpod wérev, Blass no
émt Evpod péever. I cannot but think that we have here a sur-
vival of a lost word émiEvpety, and would read émweEvpynpévy
‘close-shaven,’ i.e. a hope reduced to very narrow compass.

EP. Vea:
IANTHKTOTOKAAON TOATA®ON TOCEMNON.

Weil mrapéxe:to To Kadov, tayabov, Td cepvov [ov]. Blass
nrraxto, Perhaps mapnx«to ‘was perverted.’
Theognis 125 Bergk.
’ A xX ’ / > NN / > \ /
ov yap av eideins avépos voov ovdé yuvatKos,
mpi teipnOeins @atrep varoluyiou.

28 THE JOURNAL OF PHILOLOGY.

Joe > / vA Aon: . wv r
ovoé KEV EiKdooals woTrEep TOT &€s @pLov ENor.
/ \ / 3 eet eee A
TodANaKkt yap yvopnv. éEatrataa idéat.

A great variety of conjectures have been expended on és
@p.ov, as if it must be wrong. I venture to suggest that it is
either a different spelling of or a mistake for és avjpsov. The
meaning is: you must test a friend before you admit him to
your confidence, not after doing so. You must not guess at his
character as if you were to come the day after and find you were
mistaken in him. Sen. Epist. 3.2 Zu wero omnia cum amico
delibera, sed de ipso prius. Post amicitiam credendum est, ante
anucitiam wudicandum. Isti uero praepostero officia permiscent,
que contra praecepta Theophrasti, cwm amauerunt, iudicant, et
non amant, cum tudicauerunt. Ethic. Eudem. vit. 2 ov« éote
& dvev tiatews piria BéBatos,n 5é wiatis ovK avev ypdvov. Set
yap Teipav NaPeiv, doTep Neyer Kal O<oyvis Ov yap av eidelns—
_ vmotuylov, odd dvev ypdvou didos. And a little lower ov yap
“3 ” / 2O\ va) ene ¢ / > \ , a
€oTw dvev Treipas ovbé mas nuépas 0 pidos, adda ypovov Oeil.

Theognis 1066, 7 Bergk.
, 3Q/ bd 2 ae /
TOUT@Y OVSEV TOL GAN ETL TEPTTVOTEPOV
avopacw dé yuvacki.
Perhaps Tovtwy ov Ovytots.

Theogn. 1257.
@ Tat, KiWdvVOLoL TOAUTAAGYKTOLOLY Opotos
opynv, aXroTE Tols, aAANOTE Tolar iretv.

Perhaps xiAXovpotot ‘wag-tails, or as it seems also to have
been written xiAAvpotot. Hesych. KidAupos cicomvyis. Theo-
gnis elsewhere makes this bird an emblem of instability. 303
Ov Sel Kykriberv dyabov Biov, adr’ atpemifer.

Alcaeus fr. 86 Bergk.

For Ai yap x d@drobev On dé horxnvobev éupevat the

most probable emendation seems to be, retaining Seidler’s o dé

pn,

Ai yap «° Gddobev ErOns, 6 5é hyn Kynvobev Eupevar

‘if you have come from some other place, then he says he is
from that. An excellent description of a not uncommon cha-
racter. | |

FRAGMENTS OF THE NEW COMEDY, ée. 29

In the next fragm. though the restitution of the whole line
is uncertain, it can hardly be doubted that aBas mpoorocw
is a mistake for @4Bas mpémocw. ©

Ibycus fr. 2 Bergk.
"H pav tpouéw vv (Love) érepydpevor,
v4 / vf > t \ / . ae
@ate hepéluyos immos aeOXoghopos roti yypai taéxwv
avy bxerde Boots és auidrav éBa.

I think déxwy is a corruption of cwxev ‘vigorous’ when old
age is approaching. :

Ibycus fr. 29 Bergk.
"Epioos. ott papyov éyov oropa
avtia Snpw +éviows Kopvocot.

Bergk reads éwot. The variants are évioe, vevoot. Hence I
suspect a less common word, perhaps éveis, as in Bacch. 851
> \ > J \ / :
evels Ehahpav AvoCav.

Bentley in his Epistola ad Ioannem Millium, p. 320, ed.
Dyce, correcting the Hesiodic verses quoted by the Scholiast
on Soph. Trach. 266 Tovs dé pé@? ordotatTny téxeto EavOnv
Toreav *Avtiuyn Kpelovoa, tadaiv yévos AvBoridao sug-

gested from Hyg. fab. 14 Clytius et Iphitus Euryti et Antiopes,

- Pylonis filrae, fila that the right reading was IIvA@vos Nav-

Boridao. That Bentley was right in his reference seems to me

certain: but is not IIvAdovos AvBoridao a more probable ver-

sion of it? Bentley indeed denies the existence of such a form
as Aubolus or Aubolides: but instances are common enough of
such dropping of an initial letter, e.g. Omamertes Mamertes
(Tzetz. ad Lyc. 938), Brimo Obrimo, Briareos Obriareos (Bach-
mann. on Lyc. 698), Candulus Andulus (Lobeck Aglaoph.,
p. 1305). To my ear IIvAwvos has a sound unlike the style of
epic verse,

R. ELLIS.

THE HOMERIC TRIAL-SCENE.

a ¥ 45 , a

Keiro 8 ap év pécooics Svw ypvooio Tadavta
a \ \ a“ / 3.’

TO Sdpev, O¢ peta Totar Sikny iOvvtata eimy.

Mr LAURENCE, in a paper entitled “Judges and Litigants,”
published in the Journal of Philology, Vol. vit. No. 15 (1879),
resting on the great authority of Mr Shilleto, renewed the fight
_ over this sorely contested passage, and attacked the interpreta-
tion given to it by Sir H. Maine and other eminent scholars.
Sir Henry Maine has given an excellent exposition of the
passage on the analogy of the ancient Roman Legis actio sacra-
menti, and in reference to the lines in dispute he says :—* The
point of detail, however, which stamps the picture as the coun-
terpart of the archaic Roman practice is the reward designed
for the judges.” Against this, Mr Laurence urges three points.
He (1) first objects to translating d/env eizreiy as “to pronounce
judgment,” at the same time admitting that “den in Homer is
a very complex word used in many different ways.’ Few
modern scholars will regard it as a valid objection to such an
interpretation of an Homeric phrase that no parallel for such
usage can be adduced from Aétic Greek. (2) He objects that
“af Sir Henry Maine and Mr Gladstone are right there must
have been not one trial but two. The merits of the suit had
first to be adjudged and then the merits of the respective
judgments.” Mr Laurence goes on to say :—“ We do not learn
whether the judges were to be rewarded in proportion to their
knowledge of precedent, the skill which they displayed in mas-
tering and grouping the facts of the case, or the elegance of the
Greek in which their decisions were respectively pronounced.”
This is simply quibbling. To expect the rude judicial methods

ae
——

THE HOMERIC TRIAL-SCENE. 31

of a primitive people to accord with our present notions of
Law, which are the slow growth of ages, is indéed unreason-

able. If an English judge at the present day was to adopt

the method pursued by Solomon in his famous decision, public
opinion would scarcely view his procedure with that unmixed
satisfaction with which his subjects hailed the method em-
ployed by the wise King. (3) Like all those commentators
who have held that the two talents of gold represent the »
mown davdpos atropOipévov, Mr Laurence points triumphantly
to the largeness of the sum. On the other hand, Sir H. Maine,
evidently feeling a certain difficulty in making the sum har-
monize with his own admirable exposition, is forced to resort
to the unsatisfactory explanation that “the largeness of the
Homeric sum compared with the ordinary sacramentum indi-
cates the difference between fluctuating usage and usage con-
solidated into law” (Anc. Law, 375). |

The difficulty has arisen from the commentators all esti-
mating the Homeric talent by the standard of historic times,
without ever enquiring as to whether there was an Homeric
standard or not. And yet such a standard can most pasa
be found from the Homeric poems themselves.

First, in a well-known passage, Hoy, 262270, we find
that when Achilleus offered prizes for the chariot race, two
talents of gold stand only fourth in a list of five prizes, The
first prize consisted of a yuv7) duvpova épya idvia, and a tplzrovs
etroes. Now, though the lady was not merely skilled in fair

works, but was also supplemented by the tpézrovs drdes, we

cannot suppose the first prize to have been of startling value,
especially if we find that in the list of gifts offered to Achilleus
by Agamemnon, I. 122—130, 264—272, seven ladies, all of them
dptpova épya idvia:, and furthermore surpassing in beauty the
tribes of womankind, are mentioned quite casually, and when

we see from X. 164 that 7 rpios né yuvn was the ordinary

prize in chariot races. If two talents of gold were so inferior in
value to a slave woman, surely they would be a very insufficient
“eric” for the life of a man of some importance, such as we are
justified in considering the avjp aropOipvevos to have been.
The third prize offered by Achilleus was a Aé8ys, This same

32 THE JOURNAL OF PHILOLOGY.

vessel is denominated vaudavéwvta rAéByra in 1. 613 of this
same book, which are the words applied in rt. 386 to the laver
_in which the nurse, Eurykleia, is washing the feet of the sup-
posed old vagrant, when she discovers that he is her long-lost
master. Now, a vessel employed for such mean purposes cannot
have a high intrinsic value. Again, we find (@. 393) in the
list of presents to be contributed for Odysseus by the twelve
HynTopes nde wédovtes of the Phaiakes that a talent of gold ~
stands last, for Alkinoos. says,

TOV of Papos Exaatos évmAvVEs NOE YLTOVA

Kal ypugoto TaXavToy évelKaTe TiUNEVTOS.
Furthermore, at 6.129 we read of Polybos, King of Egyptian
Thebes,

ds Meveraw dake dv’ apyupéas acapivOous
Sovovs O€ tpimrodas, Séxa 5é ypucoto TadXarTa,

‘Here again we have ten talents of gold standing last in the
enumeration. We shall be probably justified in making a similar
inference from the position of the words ypdcou * * * déKa
wavta TaXavra in the description of the valuables,which Priam
brought forth from his treasure-chamber for the ransom of the
body of Hektor (0. 232). For though the lines at w. 273 seem
to contradict this view at first sight, where seven talents of gold
are mentioned before several leading items of the list in ©. 282,
the moment our attention is drawn to the epithet evepyys, all
difficulty disappears (ypucod pév ot Sax evepyéos éEmta Ta-
Aavta, cf. v. 202). Wrought gold was excessively valuable at
a time when skilled artificers were not to be found in Hellas,
and when all works of art were imported from Sidon or Egypt.
From an examination of the above passages, I think it will
be seen that the raddavtov of the Homeric poems is by no
means a large sum. (Mr Laurence is uncertain whether the
Homeric talent is a weight or a coin. It cannot surely be the
latter. There is no trace of coined money in Homer, and it was
only at a very late date comparatively that the Greeks coined
gold.) We must be careful not to regard it in the light of the
Attic or Euboic, Aeginetan, or Babylonian talent, which seem to ~
have been confined to silver (cf. Herod. vu. 28, Xen. Hell. «

THE HOMERIC TRIAL-SCENE. 33

3. 5. 1), whereas the Homeric talent is confined altogether to
gold. This alone is quite sufficient to overthrow Dr Schlie-
mann’s theory, that in certain oblong pieces of silver found at
Hissarlik we have specimens of the Homeric talent (Troy and
its Remains, 328). Rathér are we to regard the small talent
called the Sicilian or gold talent, used for weighing gold by the
- Greeks of Magna Graecia, which was equal to six Attic drachms,
as the true representative of the Homeric raXavrTov.

If this be so, the sum of two talents would be too small as
composition for a homicide, but would very fairly represent
the sacramentum, and thus the only weighty objection to Sir
H. Maine’s interpretation of the passage will be removed.

WILLIAM RIDGEWAY.

Journal of Philology. vou. x. 3

NOTE ON XENOPHON, DE VECT. iv. 14.

[Read before the Cambridge Philological Society, April 15th, 1880.]

‘ y
TIaraz pev yap Syov ols pepérncev axnnoapev ore Nuxias
a ’ ‘
mote 6 Nixnpatov éxtnaato év Tois apyupelois yidlous avOpo-
« a lel ol ’ \
mous, ods éxeivos Lwola TH Opant éEcuisOwcer, ef’ @ oBodov
\ 3 > / ” ¢ / > 5 50 \ 8 > @ \ y
Mev ATEAH ExaaTou THs nuépas aTrodidovar, Tov 5 apiOpov iaous
del Trapetyev.

THE conditions on which slaves were let out to work in the
mines are generally said to have been that the lessee was bound
to pay an obol a day for each slave and to restore them to the
owner the same in number’.

This profit would have been extraordinarily great, even
allowing for the high interest usual among the Greeks; for on
Boeckh’s calculation, reckoning 350 days and taking the average

1 Cf. Boeckh, Staatsh. d. Ath.? i.
p. 103: ‘Wenn Bergwerksclaven, an
Pichter vermietet, ihrem Herrn tig-
lich einen Obolos einbringen,...riihrt
_ dieser Ertrag keineswegs allein von den
Sclaven, sondern zugleich von den
damit verpachteten Bergwerken her’
and Ueber d. Laur. Silberbergw. in At-
tica (Kleine Schriften, v. p. 47 foll.):
‘Diese reichen Minner hatten ihre
_Sclaven an Unternehmer verpachtet,
unter der Bedingung, dass der Pichter
ausser der Bekéstigung der Sclaven
von jedem Kopfe tiglich einen Obolos
ohne allen Abzug erlege und die An-
zahl stets vollstandig erhalte und zu-
riickliefere’; and Biichsenschiitz, Be-

sitz und Erwerb im griech. Altert. p.
205 ;—Prof. Mahaffy, Rambles and Stu-
diesin Greece, p.130: ‘ Nicias let out 1000
slaves to Sosias, at an obol a day each—
the lessee being bound to restore them to
him the same in number’; and Wallon,
Hist. de Vesclavage dans l’antiquité,
p. 202: ‘ Ainsi les esclaves loués aux
exploitants de Laurium produisaient
net 1 obole par jour a leurs maitres
ou 360 oboles par an; et encore les
entrepreneurs supportaient-ils les chan-
ces des maladies accidentelles ou de la
fuite, puisqu’ils devaient, 4 l’expiration
du contrat, les rendre tout aussi nom-
breux qu’ils les avaient recus.’

NOTE ON XENOPHON. 35

price of a mining slave at 140 drachmae, the return would be
above 40 per cent.; and yet there would have been no risk for
the owner, since the lessee had to restore the full number of
slaves which he had received. For this reason Boeckh suggests
that the obolus a day for each slave included also payment for
the use of the mines in which they worked.

This involves a gratuitous alteration of the text, all the MSS
having zrapetyev, and secondly this solution of the difficulty
does not appear tenable for the following reasons: (a) we learn
from Andoc. de myster. § 38', that Diocleides had one slave
working in the mines; is it conceivable that a mine and one
slave to work it, could be let out to a lessee ?

(6) Xenophon in proposing that the state might, in imita-
tion of private individuals, procure public slaves and let them
out on hire to work in the mines, does not so much as hint that
the lessees would at the same time become entitled to work
such mines as were not yet in private hands—I think therefore
that we must give up the notion that the high pay of an obolos
a day for each slave included payment for the working of the
mines.

From Xenophon we can derive what I consider the more
probable explanation for this high return. He distinctly says
that such public slaves were to be let out on the same conditions
as those of private individuals (§ 19), and then proceeds to
prove that the speculation is safe (§ 21): ‘when the slaves are
marked with the public mark and when a penalty is fixed for
selling and exporting such slaves, how can any one steal them?’
Does. not this imply that the state, in its capacity as slave-
owner, would have to bear the loss, if the slaves were stolen ?
otherwise, why should it take such precautions, if the lessee had
to restore the slaves the same in number? I cannot suppose
that the public mark was merely intended to prevent inferior
slaves being restored to the state at the end of the contract, as

1 gpm yap elvar pev avdpamrodov oi ért habe, die jemand fiir an ihn vermie-
Aauply, deiv 5¢ KkopitacOa dmrogpopay. thete Sclaven dem LEigenthiimer zu
Meier und Schémann, Att. Proc. p. 533, zahlen hatte ; Andocides evidently uses
have: es ist durch keine Stelle belegt, dzogopa in this sense,

...dass drogopa je die Miethe bezeichnet

36 THE JOURNAL OF PHILOLOGY.

Prof. Mahaffy seems to think was probable in such cases, when
he says (Primer p. 40): ‘the contractor was also obliged to
restore them the same zn number, no regard being had of the -
individual slave’; the Athenians were too shrewd business men
to allow themselves to be cheated in that way. I should there-
fore suggest as a more probable explanation for the high profit
on capital invested in mining slaves, that the lessee paid an
obol a day for each slave for his work, and that it was the owner
who ran all risk for the life and safe keeping of the slaves: this
is confirmed by the fact that Nicias paid no less than a talent
for an overseer in the mines (Xen. Mem. ii. 5, 2; cf. Plut.
Nic. 4)*.

The rate of profit on the purchase money was naturally
high; for the value of a slave decreased with his getting old,
not to speak of the danger of his dying comparatively early as a
result of his exposure to the atmosphere of the mines (which
was notoriously noxious (Xen. Mem. iii. 6, 12), in spite of air-
shafts (uyayoyia)), and of the still greater danger of his
‘running away, for which reason some had to work in chains
(Plut. Nic. et Crass. init.), a danger to which Xenophon aliudes
(§ 25) by a reference to the time of the occupation of Deceleia
by the Lacedaemonians, when as is stated by Thucydides 20000
slaves deserted to the enemy.

The meaning of the above passage seems therefore to be,
that Nicias received 1000 obols a day for slaves let out to work
in the mines, and that by fresh purchases he kept up this
number, either to enable Sosias to carry on mining operations
on a large scale by supplying him regularly for the time of the
contract with the same number of workers, or that he might
himself retain a regular source of income.

1 Perhaps also by the fact that cophr. col. ii. supposing he speaks of
Athenaeus (vi. p. 272p) quotes the mining slaves, no mention is made of
above passage only as far as é¢ @ a stipulation to restore the slaves the
éBoddv éxdorou (€xacrov the text) reXetv same in number,

Ths nudpas, and that in Hyp. pro Ly-

it a
.. ol

NOTE ON PLATO, APOL. SOCR., p. 26, D, E.

a s / } > ee
’AvaEaryopou oles Katnyopety, & pire MéXnTe, Kal ovtw KaTa-
a lel ’ / / s a
povels THvde Kal oles avtovs amreipous ypaypatov eivat, wWoTeE
/ A / /
ovK eidévar btt Ta ’Avakayopov BiBdia Tod Kralopeviov yéewer
a a > 2 ,
ToUTw@Y Tav Aoywv; Kal Oy) Kal of véot TadTa Trap’ éuov pavOa-
cv 7 > , > Ud A A Y n
youow & é€eativ éviote, ci TaVY TOANOD, Spayuns EK THS
OPXNTTpPAas Tplapévats LwKpatous Katayenay etc.

All editors, as far as I know, recognise in these words of
Socrates an allusion to exhibitions at the theatre, at a drachma
admission, of plays whose authors had borrowed the notions of
Anaxagoras. But besides the wording, which appears to me
scarcely to admit of such an interpretation, there remains. the
objection that a drachma was not the price of admission to the
theatre. Boeckh (Staatsh. i. p. 68) expresses a different opinion
regarding the meaning of this passage; according to him, in the
opxnotpa of the theatre, when no performances were going on,
there were book-stalls, where the writings of Anax. might be
had for a drachma at the most, an explanation adopted by
Biichsenschiitz, Besitz und Erwerb im gr. Altert. p. 572. This
seems to me the correct interpretation of the above passage as.
far as the fact goes that trade in books is meant, but I submit to
you a passage from Photius which gives in my opinion a much
more satisfactory explanation of the spot, where the trade was
carried on; he says s.v. opynotpa: o. mpetov éxrAROn ev TH
ayopa and to the same locality points Nicopbon (Mein. fr. com.
ii. 2, p. 852) who mentions the @.GA1o7@Aar amongst the motley
crowd of sellers of figs, leather, spoons, sieves, etc. etc., whom
we can only expect to find in the market place. Eupolis (M. iz.
1, p. 550) speaks of a place od ta BuBAla dua; this also is best
understood as referring to the market place, and Boeckh does
so, only since he has once established the book trade in the
opxnotpa of the theatre, he takes AvPdia here to mean not
‘written books’ but ‘paper.’ Without entering more into
detail, by the aid of Photius’ explanation of dpynotpa and of
these and other well-known passages from the Old Comedy

38 THE JOURNAL OF PHILOLOGY.

writers (such as PuSruaypddov S& rapa Kpative év Xelpwow
M. ii. 1, p. 159) the fact seems to be established that there was
at the time of Socrates’ trial and before that time (Cratinus
+423) a trade in books carried on in the market place at Athens;
nay from Xen. Anab. vii. 5,14 we may even conclude that an

export trade had sprung up’.

1 In the reading aloud of the book-
seller, by whose side Zeno of Citium
sits down (Diog. Laert. vii. 2), Mr
Grote (Plat. i, p. 147) sees ‘a feeble
foreshadowing of the advertisements
and reviews of the present day’. But

from Lucian adv. ind. 2 kal dvayryro-
oxets via wavy emitpéxww POdvovTos Tov
opOaruwov To croua it would appear
that it was the custom of the Greeks
to read aloud, cf. dvayryvwoxew in this
sense in the Acts (8, 30), ete.

H. HAGER.

Mm <a
ui | of

NOTES ON GENDER, ESPECIALLY IN INDO-EUROPEAN
LANGUAGES.

IT is curious that, among the many subjects which philology
offers to the speculative mind, gender has been generally
avoided, or, if not avoided, has been treated in an inadequate
or contemptuous manner. In English books especially, except
Harris’s Hermes’, which was published in 1751, long before the
era of the science of language, I can find but few remarks, and
those of little or no value, on the history of gender. Adam
Smith, Lord Monboddo, Dr Beattie and the late Prof. Key have
alluded to it, but have confined themselves to enlarging, with-
out improving, upon Harris. Prof. Sayce, in recent works, has
devoted some pages to the subject and advocated a view which
will be discussed hereafter; but other writers make no more
than a passing mention of Harris’stheory. In Germany, indeed,
there is a considerable literature on the topic, but the collectors
of facts have not, so far as I know, classified them with a view
to theorising, and the theorists have not cared to study the
whole collection of facts. The truth seems to be that the
explanation of gender-distinctions belongs to a department of
philology, the ‘ Bedeutungslehre’ or Sematology or science of
’ meanings, which, as a whole, is as yet non-existent; and thus,
while the thorough investigation of gender has been neglected
as being only a fragmentary contribution to a work still unde-
signed, a plausible and ancient explanation (such as the
sexual theory adopted by Harris) of the more striking facts is
considered sufficient for present purposes. Furthermore, the

1 Chap. rv. pp. 44 sqq. (3rd Ed.).

40 THE JOURNAL OF PHILOLOGY.

dynamic changes by which the genders of words ought to be
distinguished, are not regular enough or indicated with sufficient
clearness to immediately arrest attention and invite inquiry.
Thus, in Latin and Greek, the distinctions of genders are
marked very imperfectly or not at all in the consonantal declen-
sions, and in the vowel declensions by a method which is imper-
fect, uncertain and not primaeval’, Because, moreover, in the
Indo-European family of languages each tongue has a system,
for the most part peculiar, of assigning genders and marks of
gender, the investigation of the history of gender distinctions
has been considered to belong to the special philology of each
separate language, and while the students of such separate
languages are unable to find within the limits of their subject
an explanation of the confusion, with respect to genders, therein
existing, the whole inquiry has been shelved as idle and unfruit-
ful. Even those writers who have ventured to approach the
difficulty, have carefully warned their readers against their own
conclusions. ‘As all such speculations,” says Farne at the end
of his chapter on gender, “are at best but conjectures, they
should therefore be received with candour rather than scruti-
nised with rigour.” Prof. Key*® in a similar manner apologises
for having touched on the topic at all. The writer of the article
“Grammar ” in the Encyclopaedia Britannica says, “Such specu-
lations are wholly fanciful and the principles upon which they
proceed are overturned by an appeal to the facts:” and Schoe-

by the same inflexions, the grammar
of Indo-European languages would be
more logically consistent and far easier
to learn. Suppose, for instance, in
Latin a stem-suffix -ana to indicate

1 See Schleicher, Compendium, pp.
502, 503. In the vowel declensions
gender, which should properly be
marked in the stem, is, at least to all
appearance, indicated by a modifica-

tion of the inflexional suffixes, Simi-
larly the plural, which also ought to
be distinguished in the stem, is marked
by a new set of inflexions. In verbs,
on the other hand, unnecessary varia-
tions are introduced into the personal
endings. If differences of quality and
quantity and the like were marked
always by the same alterations in the
stem and differences of relation always

the feminine, and a stem suffix as to
indicate the plural, then we might
have e.g. masc, sing. equus, fem.. sing,
equanus ; masc. plur. equasus, fem, pl,
equanasus, all declined alike with
one set of inflexions, instead of four,
The Turk himself is far less “ un-
speakable ” than the Aryan.

* Orig. and Develop. of Language,

pp. 865—379,

ee a eee ey a

ed t-
Le m 1

NOTES ON GENDER. 41

mann* recommends to the student of gender “the shrewd saying
of an old commentator on Aristotle, ov de? wiéov érifnreiv mapa
TOD AOYyou }) Boov émidéyeTat ) TAY TpaypaTav capyvea.” It is
possible, indeed, that the advice of these critics may, after all,
be justified, but I submit, with deference, that 7 tav tpayuatov
cadivera has not yet been properly elucidated. I propose, there-
fore, in this paper to sketch briefly but efficiently enough for
the present purpose the nature and distribution of gender-dis-
tinctions in the languages of the world, and to state, in a very
condensed form, a theory, suggested by these observations, of the
history of such distinctions in the Indo-European languages.
It may well be that the various grammars, upon which of course
I have mainly relied, have occasionally mis-stated rules or
stated them broadly without mentioning exceptions which would
disprove my inferences, but I do not think that any or all of
these and other similar slips can have occurred often enough to
seriously damage the plausibility of my main conclusion. That
this is indeed in part plausible, I may perhaps be permitted to
infer from the fact that much of it is not new and has already
been favourably entertained (for reasons not altogether the same
as mine) by such men as Heyse*, Madvig’, Schleicher* and
Kiihner®.

Before entering upon an inquiry mto the linguistic distribu-
tion of gender-distinetions, with a view to discovering the origin
and history of the phenomenon, it will be necessary to adopt
some definition of gender, if not really, at least prima facie
correct. The difficulty, which occurs in some sciences’®, of fram-
ing a provisional definition which does not beg the very question
of debate, need not here create any embarrassment, for gender
is itself a cause and may be well enough defined by its effects.
The gender of a noun’ is, in fact, first known by its environ-

! Redetheile, p. 72. 135; Mill, Unsettled Questions of

2 System der Sprachwissenschaft, Polit. Econ, Essay v., and Logic um.
p. 418. 225.

3 Kleinere Schriften, No. 1. 7 It is assumed here that gender

4 Compendium Vergl. Gram. ante cit. is an attribute of nouns, and is dis-

5 Greek Grammar, vol. 1. p. 284. tinguished, in other parts of speech,

6 See, for instance, Cairnes, Logical only more clearly to mark their rela-
Method of Polit. Economy, pp. 134, tion to the prominent nouns. Prof.

ess)

42

ment, by the adjective or pronoun or, in some languages, by the
verb by which it is accompanied, and all nouns of a certain
gender have, prima facie, this characteristic alone in common, ~
that they are invariably accompanied by the same distinct
forms of these and other parts of speech. It is true that
this result is merely accidental and that gender-distinctions
might well be indicated in nouns alone, without affecting other
parts of speech; but few languages, if any, seem to have adopted
the latter expedient only, and it may therefore be generally said
that in those tongues wherein separate classes of nouns are not
necessarily followed by separate forms of the adjective, pronoun
or verb (such separate forms not being distinctive merely of
number, case or person) distinctions of gender do not exist: and
the contrary, where such separate forms are found. Proceeding
then upon this principle of discrimination, we find broadly that
Radical languages’ and languages of uncivilised peoples generally
(with some remarkable exceptions to be hereafter mentioned
more in detail) admit no distinctions of gender. Agglutinative
languages” also (with one exception) are genderless, save that
an incipient gender-distinction sometimes appears in the in-
terrogative pronoun; the Incorporative, for the most part,
distinguish animate things from inanimate; and finally, of the
Inflexional, the Semitic family has two genders, called mascu-
line and feminine, and the Aryan three, the masculine, feminine

THE JOURNAL OF PHILOLOGY.

Sayce,
opinion.

however, is of -a contrary

According to his theory,
founded on Bleek’s, certain pronouns
were selected and associated (through
habit, euphony or affinity of sense or
sound) to certain classes of nouns, and
where the pronouns differed, the nouns
were considered to differ in gender.
‘¢ Whereas gender started from trans-
ferring the differences between the
pronouns to the substantives connected
with them, we now transfer the in-
herited differences of substantives to
their representative pronouns” (Comp.
Philol. pp. 254, 257). He seems to
think that this appropriation of pro-
nouns was haphazard chiefly, whereas

I believe the evidence shews that it
was, originally at least, systematic,
and depended on the meanings of. the
nouns. His argument seems to me to
shew only that gender was formally
indicated in pronouns earlier than in
nouns, and this is doubtless true of
many languages.

1 Morrison, Chinese Grammar, pp.
66,67,

2 Caldwell, Dravid. Lang. of 8. India,
p. 171; Gabelentz, Gram. Mandchoue,
p.36sqq.; Kellgren, Finnische Sprache,
p. 75; Castren, Burjitische Spr. p. 7,
§ 32; Kasem-Beg, Tiirkisch-Tatar.
Spr. p. 27.

NOTES ON GENDER. 43

and neuter. The foregoing statement is, however, available for
little without further particulars, and I therefore add here such
details as seem to me of special importance’.

In those languages which admit no distinctions of gender,
the sexes are distinguished either by separate names or by
compound words of which one element is common and the other
means ‘ male’ or ‘female.’ (E.g. Chinese jm = homo: nan jin =
vir: niu jin=femina.)* Occasionally, in the latter case, the
common word is used alone for the male sex and the compound
employed only for the female: (e.g. Setshudna khomo = bull,
khomogari = cow:) or in the compound forms the distinctive
parts are clipt, as in Bullom, where pokan ‘male’ and lakan
‘female’ are often shortened in composition to pok, po, lak, la’.
All these processes, it will be seen, actually survive or have
their analogues in the higher languages. Many of these gender-
less tongues, again, draw, in some way or other, a distinction
between persons and things* which approaches to a distinction
of genders. This appears generally, but not always, in the
interrogative pronoun, which has two forms, answering to our

‘who’ and ‘what’. Thus, in

1 In the following remarks, state-
ments not otherwise authenticated are
derived from Pott's article ‘“Ge-
schlecht,” a splendid repertory of facts,
in Ersch and Gruber’s Encyclopedia,
vol. uxt. Bindseil’s Abhandlungen
tiber Allg. Vergl. Sprachl., No. 2, is
also full of information, and contains
in particular (p. 527) a more precise
statement of gender distribution.

2 In Thibetan, curiously, pho ‘male’
and wo ‘female’ are added also to
pronouns.

3 So also in the Mandingo (Mac-
brair’s Grammar, § 14) and Yoruba
languages (Crowther’s Vocab. p. 5) of
Africa.

4TI have already called this “an
incipient gender-distinction.” I mean
that the distinction drawn is really
between things, and not yet stereo-
typed into one between mere words.

Bullom (Africa), the pairs are

In Latin, on the other hand, such a
word as Somniuwm, even when repre-
senting a divinity (e.g. Ov. M. 11. 588),
would be followed by a neuter relative.
At what stage a gender-distinction
becomes complete is very difficult to
determine. I have taken it, for prac-
tical purposes, to be so when a large
class of nouns, representing not exactly
a certain class of things, has appro-
priated to it special forms of the other
parts of speech. In such a case, I
should guess that the speakers had
learnt to use words (like money) with-
out immediately realising what the
words represent. A good example of
the incipient distinction is seen in
Herero, where ombepo (Kafir umoya) is
of the personal class, when it means
‘spirit,’ but when ‘wind,’ of the im-
personal (Bleek, post cit, p. 45).

44 THE JOURNAL OF PHILOLOGY.

nghd, ngho: in Annamese ai, ndo (though the latter is used
sometimes of persons also): in Mandschu we, av: in Magyar ki,
mi: in Samoyed sele, ma: in Turkish kim or kin, neh.
Dyaks of Borneo draw the same distinction in the pronoun of
the third person, using id of persons, té of things. In Zulu,
words which begin with wm (in Cuan mo, in Herero omu’) have
a different plural when they are names of persons from that
which they have when names of things’, The distinction is
carried furthest in Ashantee*®, which, indeed, may be said to
have two genders. In this language the names of persons and
the pronouns which refer to them begin with o (e.g. oba ‘ boy’),
while, of things, the names begin with a, the pronouns with e.
The class of persons, however, is not very strictly defined: for
instance, opodo ‘pot’, oprar ‘broom’, osekan ‘knife’, and other
similar words have the personal ‘o’, while agya ‘father’, abofra
‘child’, and others are impersonal. On this point more will be

said farther on. Both in Africa and Asia remarkable exceptions:

to the general rule of low linguistic developement appear.
Thus, in Africa, the Hottentots distinguish masculine, feminine

and common genders, have a definite and indefinite form of |

each, and singular, dual and plural numbers (e.g. kowb, kos, kovs
are vir, mulier, homo respectively): and the same gender-dis-
tinctions are added to the verb, as gambi, gamst, gami, virum,
mulierem, hominem interficere*. In Namaqua also, according
to Pott’, who quotes a vocabulary published at Barmen in 1854,
the same rules prevail (e.g. aub, vir: aus, mulier: awi, homo),
and gender is distinguished even in the personal pronouns (e.g.
masc. make, fem. mare, com. mada, ‘nos’)*®. In Asia, the excep-

‘Thess

1 See Bleek, De Nominum Generibus
Ling. Afric, Austr. pp. 15 and 13.

2 Grout, Journal of American Orient.
Soc. rt. 403. A less important but
similar distinction appears in Kechua
(S. America) and Malay: where differ-
ent words for ‘male’ and ‘female’
are used of persons from those used of
animals. See Tschudi, Kechua-Sprache,
1. § 114; Crawfurd, Malay Gram. p.
10.

3 Also, according to Krapf (Outlines

of Kisuaheli Lang. pp. 28—33) in
Kisuaheli and other African languages.

4 Bleek, op. ante cit. pp. 40 and 46.

5 In Ersch and Gruber, txt. 410 a.

6 The masculine and feminine gen-
ders are not, in either of these lan-
guages, confined to names of animals,
It is curious that in Namaqua soris
‘sun’ is feminine, while khéb ‘moon’
is masculine, as are hurib ‘sea’ and
huub ‘earth.’ See Bleek, Comp. Gram.
of 8. Afr. Langg. pp. 112 and 120,

NOTES ON GENDER. 45

tion is furnished by the Dravidian languages of Southern India,
which shew, with respect to gender,a very remarkable develope-
ment. In Tamil nouns are divided into ‘high-caste’ and ‘low-
caste’ or ‘casteless’; in Telugu, into ‘mahat’ and ‘amahat’ or
‘majors’ and ‘minors’. The first class includes only the names
of rational beings, and in this class masculine and feminine
genders are distinguished. All words in the second class are
neuter. “This distinction” says Caldwell’ “appears to have
arisen at a late date”, for in older Tamil and in poetry we find
many words neuter, which, later, and in prose, are masculine
(e.g. Dévu ‘God’ (from Sanskrit) is neuter in old and mase. in
modern Tamil): and according to the same authority, the
suffixes which distinguish the masculine and feminine are only
mutilated pronouns. Of the Incorporative languages, many
have no distinctions of gender, and all discriminate the sexes
by the primitive methods before-mentioned. In most, however,
of the tongues of North’ and Central* and in some of South
America‘ a distinction, more or less clear, is drawn between the
names of animate and those of inanimate things. In many cases
the two classes of nouns are distinguished, as in Zulu, only by
separate forms of the plural; but more generally, it would
seem, the gender of nouns is reflected in the verbs, which, when
transitive, vary with the object, when intransitive, with the
subject’. According to Gallatin, some languages of Central
America further divide the animates into rational and _irra-
tional, but more particulars on this subject are wanting. It
appears plainly here as in Ashantee, that the linguistic distinc-
tion between animates and inanimates seldom accords strictly
with the natural, and that the classes differ in different lan-
guages, Thus the Lenni-Lenape or Delaware Indians assign to
the animate gender everything that lives and grows except

1 Dravidian Lang. of S. India, p. 171.

2 Du Ponceau, Mémoire sur le Sys-
téme Gram. des langues de quelques
nations Ind’‘ennes de ]’Amerique.

3 Gallatin, Trans. Amer. Ethnol.
Soc. 1. p. 12.

4 Tschudi, Kechua-Sprache, 1. § 114
sqq.

5 E.g. wiinnehayoo=he loses an ani-
mate object; winnetow=he loses an
inanimate. Howse’s Grammar of Cree
Language, p. 41. Bindseil quotes (p.
527) from some American language
nolhalla (‘habeo’) nolhatton; newa
(‘video’) nemen, as similar pairs of
animate and inanimate verbal forms.

46 THE JOURNAL OF PHILOLOGY.

annual plants and a few special exceptions (e.g. namessall
‘fish’): and the Kechuas add, to the same class, the names of
rivers, the sea, the sky and the stars, while they treat as inanim- _
ate all living things, such as little plants and animals, which do
not obtrude their vitality on the public eye. According to
Schoolcraft, also’, the name of any inanimate, when PAE
is transferred to ha animate gender.

It remains to speak only of the Inflexional languages, Of
these, the Semitic family (including here the doubtful Koptic’,
Galla and Berber) distinguish two genders, the masculine and
feminine*®. To the latter class belong, generally, the following
groups, viz.: names of women, female animals, countries, towns,
double members of the body and tools, abstractions and collec-
tives. Hebrew, at least, uses the feminine adjective where
Indo-European languages would have the neuter (e.g. “a tongue
speaking great things.” Psalm x1. 4). The names of male
animals belong to the masculine gender, but other words seem
to be distributed between the two genders in a manner of which
no grammar yet published suggests a rationale. To the two
genders of the Semitic family, the Indo-European languages
add a third, the neuter, and confusion now becomes worse con-
founded. The signs of gender and the groups of ‘congeneral’
words differ totally sometimes in the most nearly-related tongues,
and words of closely-allied meaning in the same language often
disagree in gender. Hxcept in names of male and female
animals, few glimpses of uniformity appear. The names of
‘child’ ‘egg’ and fruits, and also diminutives, are almost exclu-
sively neuter in all the languages of the family: ‘earth’ is
feminine: ‘wind’ and the names of winds are masculine, and
abstractions seem to be generally feminine. One example, on
the other hand, will shew the extraordinary diversity of genders
ina simple group. The names of trees are in “Greek, Latin (in

1 Quoted by Pott, E. and G. Lxt1. p.
420.

2 Dillmann, Athiopische Gram. §126,
and Bleek, op. ante cit.

3 Gesenius (ed. Rédiger, trans.
Davies), pp. 289—292 ; Forbes, Arabic
Grammar, pp. 41—43 (almost transla-

ting the chapter in Sacy’s Grammaire) ;
Wright, Arabic Grammar, vol. 1. §§
289---297 ; Caspari, Arabic Gram. pp.
120—125 ; Longfield, Introd. to Chal-
dee, § 32, p. 40; Phillips, Syriac Gram.
§ 16, pp. 38—40; Sayce, Assyrian
Grammar, p. 114.

NOTES ON GENDER. 47

spite here of analogies of form which would tend to make them
masculine), Lithuanian and German, mostly feminine: But, in
all these languages, there are many exceptions to this rule, and
that not merely in the names of wild trees, e.g. épuvéos, oleaster,
where the reason of the distinction may be guessed. ‘Thus,
Greek dévdpov, Gothic triu, Slav. drevo are all neuter, while
Germ. baum is masculine, In Greek the following are mascul-
ine: PtAdOs, AwTOs, HoivuE, KuTTOS, MUPpivos, Képacos and a few
more. In Latin except ficus (which is also feminine) the
anomalies occur mostly in the names of plants, as acanthus,
carduus, etc., all masculine. In Old High German ahorn, ase
and élm were all masculine (as was Old Norse poll pine-tree),
and many names of plants. In Slavonic languages many trees
are masculine, but in Sanskrit, according to Benfey, nearly all
trees are masculine and shrubs feminine, while the two dialects
of Keltic shew this extraordinary difference, that in Gaelic trees
are masculine, in Kymric feminine’. Similarly, sun, moon and
stars, mountains, rivers, towns and countries, collective names
and abstractions, though seldom neuter, are divided in inex-
tricable confusion between the other genders. A further peculi-
arity of these languages is that in modern times they shew a
strong tendency to discard one or more of their ancient genders.
Thus Persian’ has lost them altogether. Bengali*®, English and
Danish* preserve the masculine and feminine only for male and
female animals and persons, and assign all other nouns to the
neuter. Lithuanian’ has lost its neuter, as Keltic’ had long
before, and one Spanish pronoun (ello) is the sole relic of the
same gender in the Romance languages’.
The foregoing statement will perhaps suffice to apprise the
reader at least of the nature of the facts with which a theory of

1 Pott on ‘Metaphern’ in Kuhn’s
Zeitschr. vol. 1. Similarly metals are

6 Zeuss, Gram. Kelt. p. 228,
7 The neuter, in the other Romance

in Greek masculine, in Latin neuter,

2 Chodzko, Persian Grammar,

3 Forbes, Bengali Gram. pp. 18, 19.

4 Grimm, Deutsche Gram. 111. pp.
545, 546.

5 Grimm, ubi sup. p. 548; Schleicher,
Litau. Spr. p. 170.

tongues merged in the masculine, has
become feminine in Wallachian only.
See Alexi, Gram. Daco-Romana, p. 35.
So also, in Lithuanian, the old neuter
adj., used substantively, is now repres-
ented in common parlance by the
feminine. Schleicher ante cit. p. 258,

48 THE JOURNAL OF PHILOLOGY.

the history of gender-distinctions must deal. To profess to
divine, with certainty, the order out of which this monstrous
disorder grew, would be an absurd pretence. Nevertheless,
time has left a few traces of what might have been, and the
method of comparison may still suggest what ought to have
been. Conjecture, with these two aids, must be left to do the
rest.

It will have been observed that in all languages, except the
Inflexional, the division (if any) of genders in nouns agrees
pretty closely with a ‘dichotomy’ of things, reasonable enough |
to satisfy even Plato’s Eleatic stranger, the exceptions not being
so numerous or of such a kind as that either inveterate use or
some common characteristic of the barbarous mind may not
fairly be admitted to account for them. A presumption is thus
raised that the genders of inflexional languages are also the
reflexion of a distinction in things analogous to that between
-animates and inanimates, or rational and irrational creatures.
Nor has this presumption been disputed, so far as I can find, by
any philologer until quite recent times, and then only by Dr
Bleek and Prof. Sayce. Of these writers, the former, impressed
by the fact that in the languages of the Congo-Caffrarian
group and in Koptic the noun-affixes are or seem to be mutil-
ated pronouns, suggests a theory that noun-affixes have, in all
languages, come to be used as pronouns and that this accid-
ental circumstance has given rise to a supposed distinction of
genders. Prof. Sayce is of the same opinion except that he
would reverse the process and say (see supra, p. 2, note) that
pronouns came accidentally to be used as noun-affixes’. It isa

1 I think the following examples
{the first of which is suggested by Dr
Bleek himself) will not misrepresent
these two theories. According to
Bleek, such suffixes as -dom, -ric, -ism
of kingdom, bishopric, idealism would
come to be used as pronouns, and
where these nouns had preceded, we
_ might refer to them as ‘dom of Eng-
land,’ ‘ric of Durham,’ ‘ism of Plato.’
Prof, Sayce may be better illustrated
by another set of examples. Thus, in

a given primitive language, say, there
are pronouns ‘he,’ ‘she,’ ‘it,’ and
others all prima facie equally appli-
cable in reference to any noun. Habit,
however, has made it correct to say
‘he’ of male bears, oaks, tables, cups,
spoons, and a thousand other things;
‘she’ of female bears, elms, chairs,
saucers, forks, &c.; ‘it’ of insects,
vines, &c. In process of time these
pronouns were actually appended to
the nouns, and so we get words con-

NOTES ON GENDER, 49

plain corollary from this theory that there may be as many
genders of nouns as there are distinctive noun-affixes, and Dr
Bleek actually assigns sixteen (and in one case eighteen) genders
to the languages of the Congo-Caffrarian group’, these genders
being in fact forms of nouns compounded with and followed by
distinct pronouns. It must be admitted that distinctions of
gender clearly could have arisen by the methods here suggested,
but such a theory is no less clearly inapplicable to the Indo-
European languages, in which pronominal elements are very
doubtful in noun-suffixes and have certainly nothing to do with
the commonest methods of distinguishing gender. In the
Semitic family, where the theory has more plausibility, it is to
be observed, in the first place, that a very large proportion of
feminines have not the distinctive pronominal element, and,
secondly, though here the nouns included under each gender
are on the whole strangely heterogeneous, yet under each there
occur so many well-defined groups of similar meaning that it
becomes incredible that the gender was only accidentally
ascribed. A fortiori is this incredible of American and other
languages which distinguish such well-marked groups only as
animates and inanimates. Lastly, on examining Dr Bleek’s
sixteen or eighteen so-called genders, we find that seven of
them are forms of the singular and seven of the plural. Thus,
all nouns of any given gender out of these fourteen have at any
rate this in common, that they are of the same number, and
number is obviously, as much as sex or life, predicable, not origin-
ally of words, but of the things which words stand for. It might
therefore be as properly said of these South African languages
that they have two genders (or classes), a singular and plural,
each distinguished in seven different ways: it would then be

structed like he-bears (or bear-he), he-
oak, he-table, &c.; she-bear, she-elm,
she-chair, &c.; it-insect, it-vine, &c. ;
but these groups have really nothing
in common except that they are com-
pounded with and are followed by the
same pronoun respectively. Those
-languages which have two or three
genders only have them merely by

Journal of Philology. vou. x.

reason of the fact that they had origin-
ally only two or three noun-affixes
(ace. to Bleek), or pronouns (acc. to
Sayce). See Bleek, Comp. Gram. of
South African Langg. Pt.11. p. 104, and
Sayce, Comp. Phil., ante cit.

1 This comprises all the languages
of South Africa except the Hottentot.

4

50° THE JOURNAL OF PHILOLOGY.

seen that the basis of the distinction is even here rational, though
the marks of gender (i.e. of number) might have been accident-
ally distributed’, We must then, it would seem, in default of
a better hypothesis, fall back upon the theory, so favoured by
analogy and so long sanctioned by belief, that genders in
inflexional languages also are due to a primitive classification of
things. What was the basis of this classification, is now the
question. Clearly it was not number (as in 8. Africa) for that
is otherwise provided for: neither was it the presence or absence
of life (as in America), or of reason (as in Tamil), for the divi-
sions of gender, even where there are but two, do not coincide.
at all nearly with these distinctions. Some other principle of
division is therefore plainly required, and the almost universal
opinion of all ages, since grammars were first devised, has
declared that differences of gender in inflexional languages were
intended originally to mark differences of sex, real or imaginary,
in the things or ideas of which nouns are the names. This
belief rests apparently on the following grounds. The gender of
nouns is, as it has already been stated, primarily indicated by
the adjectives, pronouns or verbs which accompany them. Now
we can easily conceive, with Bleek, that of these parts of speech
there should be as many forms as there are of nouns, that in
fact there should be as many genders as there are distinctive
noun-suffixes. Yet in reality the number of genders in any
language is at most three. It is a striking coincidence that the
number of possible sex-distinctions is also three, viz. male,
female and no sex. Moreover, the number of cases in which

1 The remainder of Bleek’s genders
have nothing to do with number. The
fifteenth belongs to the infinitive mood
and a few nouns signifying place. The
16th, 17th, and 18th,which may be called
peculiar to the Otyihereré language,
are very rare even in ihat, and belong
also only to nouns signifying places.
These four might, therefore, very well
be called the ‘ locative’ gender. Ob-
serve also that of the first fourteen
forms, two, a singular and a plural,
are devoted to the names of men,

animals and personified things; two
more, also sing. and plur., to the
names of rivers and trees. Another,
a sing., implies very great size, while
its corresponding plural is the gender
of collectives (esp. of liquid substances).
See Bleek, De Nom. Gener., ante cit.
pp. 16 and 17, and Id., Comp. Gram.
of South African Langg. Pt. 11. pp.
104, 123, 253 sqq. These facts do not
seem easily reconcileable with Bleek’s
theory of accidental origin of gender-
marks.

NOTES ON GENDER. 51

=

the name of an animal of a certain sex receives an exceptional
gender are so small that, given a sex-name, its gender is almost
certainly predicable. The difficulty of accounting for the attrib-
ution of sex to sexless things is not, by any means, insuperable ;

and-lastly, the evidence of antiquity, i.e. of the writers who

lived nearest to the time when genders were significant distinc-

tions, is entirely in favour of the sexual theory. Thus the

Sanskrit grammarians speak of ‘masculine’ and ‘feminine’

genders, and the third they call ‘kliva’ or ‘napunsaka’, ‘castra-

ted’. The German ‘geschlecht’ positively means ‘sex’. Our
own word ‘gender’ is the lineal descendant of Priscian’s ‘genus’,
which he thus defines: ‘Genera dicuntur propria quae generare
possunt, quae sunt masculinum et femininum.’ Servius had
previously given a similar definition. In Greek yévos, used by
Plato as distinctly equivalent to ‘sex’’, was first applied to
language, according to Aristotle, by Protagoras, but there is
some doubt whether the latter really meant ‘sex’ by the term,
as Aristotle, on the other hand, certainly did.

Evidence, so many-sided as this, in favour of the sexual
theory is not easily to be overcome. Some philologers, however,
perplexed by the innumerable anomalies in the division of
genders upon the sexual principle, have proposed new bases of
distinction. One of these, which has received much favour, was
suggested by W. Mohr’, who, after premising that language
must indicate the quantity, quality and relativity of each thing
or idea, declares gender to represent a distinction of quality, in
respect of activity, passivity and a mixture of these two. Thus,
the names of all active things are or should be of the first, so-
ealled masculine gender; those of passive things are of the

1 This fact has led Grimm (Deutsche
Gram. 111. 317), who is followed by
most German philologers, into the
mistake of distinguishing ‘natural ”
from ‘grammatical’ gender (called by
Bilderdijk ‘eigen geslacht’ and ‘ge-
slacht uit toepassing,’ 7.e. by applica-
tion. Niederlandsche Spraakl.). Sex
is natural, but gender is, of course,
always grammatical.

2 Sympos. 189 d; Art. Rhet. ur. 5,
Poet. 21. Varro and Quintilian per-
haps did not use ‘genus’ as meaning
‘sex,’ for they speak of ‘genera’ of
verbs. For more information see
Schmidt, Beitrige zur Gesch. der
Gram. § 269 sqq.

3 Dialektik der Sprache, Pt. m1. chap.
1. §§ 54—56.

4.— 2

52 THE JOURNAL OF PHILOLOGY.

second or feminine, while those things which, like tools, are at_

once active and passive, are assigned to the third or neuter
gender’. Criticism upon such a theory can hardly be very
vigorous, but it seems worth observing, in the first place, that
the relative activity and passivity of male and female animals is
shewn only in the sexual function, so that, in their names at
least, sex must after all be. considered the basis of the distine-
tion of genders. Further, in the case of many things other
than animals, it is far easier to conceive them as of a certain
sex than of the corresponding activity or passivity. Thus
abstractions are in the Aryan languages generally of the feminine
(Mohr’s passive) gender, yet surely the virtues and vices, for
instance, were generally, in primitive times, imagined as active
powers existing beyond us and interfering in our daily lives to
prompt us to good or evil, to bless and to scourge. The sexual
theory, pointing to Themis, Eris, the Muses and other similar
names, suggests that these abstractions were personified as
females and hence their names were assigned to the feminine
gender*. To put it shortly, Mohr’s theory is a less safe and
satisfactory explanation than the sexual, because sex is of a
much wider connotation than any other single atfribute, such as

activity or passivity. To get over this difficulty, Bindseil*® very

much enlarges the list of qualities upon which gender will de-
pend. According to his view, the first gender will comprise the
names of those things which are, by comparison, “great, strong,
swift, active, stirring, creative”; the second the names of those
which are “small, weak, passive, receptive, productive;” and the

a

1 Mohr’s theory seems to have been
suggested by a remark of Grimm,
who says (Deutsche Gram. 11. p. 311),
‘The distinction of genders lies as deep
in the nature of nouns and their forms
as that between the active, passive,
and middle in verbs. Both divisions
may be compared in more than one
respect: the active appears, like the
masculine, as the strongest and
oldest form; the passive, like the
feminine, as a form derived from the
first: the middle, like the neuter, as a

blending or combination of active and
passive, male and female, forms.”
This hint, however, is not carried fur-
ther by Grimm himself,

2 Bleek, on the other hand, as might
be expected, is of opinion that genders
influence mythology rather than my-
thology genders. See his Reynard the

Fox in §. Africa, p. xx1., and Origin

of Lang. p. 23.

3 Allg. Vergl. Sprachl. ante cit. pp.
496, 497.

—_

[pee se

NOTES ON GENDER. 53

third those which are “impassive, lifeless, undeveloped’.” It

would naturally be expected, if this theory be true, that the

names of all huge and ferocious animals at least would be of the
first gender, and those of small and timid creatures of the
second; yet this is not the case, and again we seem to be thrown
back on sex alone as the basis of gender in one very large and
important group of nouns. Finally, a happy compromise is
adopted by Madvig’, who, arguing from the fact that the neuter
sign, a final -m or -v in Latin and Greek, is also the accusative
sign in other genders, sees in the neuter the passive gender and
confines the sexual theory to the masculine and feminine, which
together form the active gender. An explanation differing, on
the whole, in little but names from Madvig’s, will be proposed
in this paper, with the further advantage, that it will not press
the sexual theory into accounting for all masculine and feminine
words, and that it, at the same time, removes the difficulty
which prevented Bindseil from admitting the sexual theory at
all. He was unable, he says, to account for the apparently late
origin of the neuter and the apparent “retrogression” of so
many languages which have abandoned the sexual distinction
for that between animates and inanimates. It will be suggested
that neither of these appearances corresponds with fact. The
remark should here be added that probably none of the theorists
just reviewed would deny that some words may come by their
gender simply by the accidental analogy of their form®: but
since such cases would only arise at a somewhat late date in

_the history of language, they may well be neglected in a general

discussion of the origin of gender-distinctions.

1 This classification, again, is taken
from Grimm’s division of the words to
which grammatical, not natural, gen-
der belongs (Deutsche Gram. 111. 359) :
but while Grimm admits that sex is
after all the ground of these distinc-
tions, Bindseil stoutly maintains that

- gex has nothing to do with them.

2 Kleinere Schriften, No. 1. (Ge-
schlecht).
3 Thus abentewer in Germ. is neut.

from adventura fem.; geste in French
is fem. from Lat. gesta, neut. plur. ; diu
mire in Thuringian is fem. from old
diumaere neut. plur. Grimm, Kleinere
Schriften, 1. 85. On the other hand,
meaning sometimes determines gender
in spite of form and custom, e.g. tre-
telgia Goth. is fem. when meaning
‘axe,’ but, when ‘carpenter,’ is masc.
So nullus potestas, den Potestat in old
charters, like modern Italian podesta.

54 THE JOURNAL OF PHILOLOGY.

The statements hitherto made andthe theories cited, with
regard to genders in inflexional languages, apply to those
languages only in the highest stage of their development. ‘This

fact, however, is.almost always forgotten by the theorists, who

generally take it that the differences of sex or other qualities
are the only original ground of distinction of words, the common
primitive principle of division which those nations, who acknow-
ledge less than three genders, have abandoned. Now Indo-
European languages at least, appear to have had a very long

history at which we can only guess, and it is likely that, in this

remote past, the distinctions of gender were not such as we find
them at a later date, and again that the earlier usage did not
wholly disappear on the introduction of the modern. The
inquiry into development, which has disturbed so many sym-

metrical a@ priort theories, may perhaps be advantageously —

applied to the study of genders, though unfortunately, from the
nature of the case, the evidence to be procured can be only
fragmentary and, for the most part, disputable.

In the first place, as, on the one hand, we find in South-—

east Asia radical languages slowly becoming agglutinative, and
agelutinative elsewhere (e.g. in Hungary) becoming inflexional,
so, on the other, the more we examine the structure of words in
the Indo-European languages, the more irresistible becomes the

conviction that these have been at one time agglutinative and ~

still earlier radical. This conclusion, based upon the analogy
here mentioned and upon the successful analyses of many in-
flexions of verbs and of the genitive case’, seems to be admitted

_ by all philologers. But the radical languages now known admit,

as we saw, no distinctions of gender, and it is an obvious question
whether the Indo-European languages did so at the time when
they may be supposed to have been radical. Analogy, at any
rate, would shew that they did not, nor is internal evidence
wanting to support the same inference. The nature of that
evidence may be best learnt from a short extract from Schleicher,
who says, “ Gender is marked by no vocal element in nearly all
cases of consonantal stems and stems in diphthongs and in 7 and

1 Garnett, Philolog. Essays,

|

mt

NOTES ON GENDER. 55

-u. In stems in a, we find the feminine, in nearly all cases, an

increase of this a to d,...so that here most cases of the feminine
are distinguished from the masculine and neuter. But the
circumstance that the increase of the a-ending is not wholly
foreign to the masculine and neuter shews that it is not wholly
devoted to the feminine and that we cannot properly speak of
feminine stems in @ Moreover, in some languages the forms of
the a stem with increased ending are used as masculine, e.g.
Lat. advena, Gr. woditns: Slav. vladyka ‘master’, old Lith.
geradéja ‘benefactor’ etc.: and a stems not increased do duty as
feminines, e.g. Gr. odds: Lat. nwrus, domus, etc: so that this
distinction is not universal and its original application to the
distinction of gender is very doubtful.” The other means of
marking gender, enumerated by the same philologer, are also
almost all certainly not primeval, and his conclusion is that “es
ist deutlich war zu nemen, dass in einer aiteren sprachepoche
der indogermanischen ursprache das genus one bezeichnung
war und erst im laufe der zeit durch secundare hilfsmittel die
genera am nomen gesondert wurden’.” It is worth observing,
also, upon the same subject, that, as in radical and other unde-
veloped languages the sexes are distinguished by wholly distinct
words (even in compounds the words for ‘male’ and ‘female’
are invariably quite different) so also in Indo-European languages
the names of the sexes in human beings (e.g. father, mother,
brother, sister, son, daughter) and in the animals probably most
familiar to a primitive people (e.g. bull, cow, dog, bitch, taurus,
vacca, verres, sus, etc.) are generally derived from different roots,
or are represented by one common term differentiated only by
the context or by the addition of the adjective ‘male’ or ‘female’
(e.g. femina bos, Varro, R. R. 11. 1, 17, etc.). The second ques-
tion which now arises is, When the Indo-European languages first -
assumed the practice of distinguishing genders, upon what
principle was the distinction made? Did they at once proceed
to discriminate the sexes, or was there an intermediate stage ?
The ordinary view appears to be that they first distinguished
masculines from all other words which together belonged to the

1 Vergl. Grammatik, p. 501. Comp. also Pott in Ersch and Gruber, ante
cit. (pub, 1853).

56 THE JOURNAL OF PHILOLOGY.

second gender, and that this latter was ultimately again divided,
new forms being assigned to the neuter. This theory, however,

though it derives some colour from the practice of the Semitic

tongues, is not supported by analogy in others or by the imternal

evidence of the Indo-European themselves. On the other hand

there seems to be a fairly strong case, considering the nature of
the subject, for the argument that the languages of our family
at an early time distinguished animate from inanimate gender.
Analogy here is to be looked for not only in the agglutinative
tongues, but also in the incorporative, for the nations who speak
the languages of these groups are, upon the whole, at much the
same stage of mental development. We have already seen that
in incorporative languages a distinction between animate and
inanimate genders is almost universal, and that in the agglutina-
tive, as also in the tongues, not definitively classed, of many of
the higher savages, the same distinction or something like it is

generally incipient and occasionally complete. Itseemsunlikely ——

that the ancestors of the Aryan race, whose. minds doubtless
progressed, by the same paths as those of other peoples, from
brutal ignorance to civilization, should have disregarded, as too
obvious, a general classification which others have deemed so
striking and so necessary to reproduce in speech. And here
also internal evidence is forthcoming to support analogy. The
distinction of only two forms, animate and inanimate, in the
interrogative pronoun, which, as we saw, was the first inkling of
gender in the agglutinative languages, is retained in nearly all the
descendants of the Aryan stock; e.g. Gr. tis, tu: Lat. guis, quid’:
Ger. wer, was: Polish kto, co (so nikt ‘nobody,’ nic ‘nothing’):
Persian keh, tscheh: so also in Albanian and Keltic®. Yet
Sanskrit has a feminine ka (masc. kas), so that the usage of the
other languages would seem to be a survival from a very ancient
time. Nor has some express recognition of this fundamental
distinction ever wholly died out. Thus in Slavonic languages
of the present day, masculine nouns representing animates have
an accusative in -@ which is not given to masculine names of

1 Roby, Latin Gram. § 380, ‘Quis tv. 1. 6, &¢.
ilaec est mulier?” Plaut. Epidic. 2 Bindseil, p. 513.

NOTES ON GENDER. 57

inanimates*. In Lithuanian, masculine names in u% of animates
have a special form of the vocative wanting to other masculines.
In Persian, diminutives of animates are formed with -ek (as
kenizek ‘puellula’) while those of inanimates are formed with
-tsheh (as mah-tsheh ‘lunula’). In Old High German, neuters
which stood for living things formed a special plural in -zren,
etc. But the great sign of one common early usage in this
respect is this, that the oldest and only universal distinction of
gender-forms lies between the neuter on the one side and the
masculine and feminine combined on the other. In the nomin-
ative and accusative singular the neuter has the bare stem or
the objective ending -m (in pronouns a special ending -d) as
against -s of the other genders, and in the plural again the
neuter has final a where the others have ds. In one case, only,
are the masculine and neuter together against the feminine, viz.
in the feminine genitive sing. of the a-declension we find -s
against the -sya of the others, but this variation may be ex-
plained by the fact of the late origin of the feminine, to be
hereafter considered. In all other cases, where the masculine
and neuter are distinguished from the feminine (e.g. Lat. abl.
novo against nova) the difference is not primeval (e.g. in the last
example both forms were originally novat) and its ultimate
-appearance may be similarly explained by the late origin of the
feminine*, Finally, whereas three common genders are con-
ceivable, viz. a masc.-fem., masc.-neut., and fem.-neut., only
one, the masculine-feminine, is found. This evidence, already
reasonably strong, will be further enforced by facts which it is
convenient to reserve for our last inquiry, namely, that into the
general development of the forms by which genders are indicated.
For the same occasion also, may be suitably kept a more detailed
suggestion of the mode in which the animate gender was ultim-
ately divided. It will be sufficient here to say that there
seems little ground for doubting that this final division was
based upon distinctions of sex in the things heretofore con-
sidered merely as animate, though probably not all such things
were further conceived as endowed with sex.

1 Grimm, Vorrede xxx1x to Wuk’s 2 Schleicher, Vergl. Gram. pp. 502,
Serbische Grammatik. 503.

58 THE JOURNAL OF PHILOLOGY.

One more question, as it has been already hinted, remains
to be answered before a theoretical history of genders can be
considered complete. Of existing forms, appropriated to the
several genders, which are the oldest? When the primitive
Aryan language first admitted distinctions of gender, were new
vocal elements introduced to mark the classes, or were the old
forms kept for one class and new provided for the other or
others? And, in the latter case, to which class were the earlier
forms assigned? Upon this point, philologers seem, for some
~ reason or other, to be entirely agreed. From Bernhardi down
to Prof. Sayce’, all concur in the opinion that the masculine
gender is the oldest, the neuter latest. The grounds of this
belief appear to be, first, that in oblique cases the inflexions of
the neuter are the same as and therefore probably adopted from
those of the masculine, and, secondly, that Semitic languages
have no neuter. Now, unless the neuter be a very late inven-
tion indeed, its inflexions in oblique cases are originally not only
tne same as the masculine but (except in one case, the genitive
sing. of the a-declension) the same as the feminine; and that it
is not a late invention, its universality sufficiently declares.
When, moreover, it is sufficiently considered that there are
good grounds for believing these inflexional suffixes to be mere
adaptations of pronominal roots of definite meaning, it will be
seen that it is impossible to give the priority, in respect of their
use, to one gender rather than another. The argument, in fact,
is only saved from being circular by the introduction of the
analogy of Semitic speech. It would be lawful, perhaps, here
_ to protest against an inference derived from languages whose
early history is, In so many important respects, totally incom-
parable with that of the Indo-European family, but it seems
possible also to aver that the real analogy may be different from
that suggested. The genders of Semitic languages do not seem
to have been very carefully treated, as yet, by any writer, but,
in such information as can be gleaned at least from Hebrew

1 Bernhardi, Sprachlehre, pt. 1. pp. wx. 405 b; Whitney, Language and
141 sqq.; Grimm, Deutsche Gram.1. Study of Language, p. 274; Tylor, Pri-
318; Bindseil, Allgem. Verg. Sprachl. mitive Culture, 1.273; Sayce, Assyrian
p- 496; Pott in Ersch and Gruber, Grammar, p. 119.

NOTES ON GENDER. 59

erammars, I can find nothing to disprove and some facts to
favour a theory that these tongues also, like the Aryan, had at
one time no genders; then divided animates from inanimates,
and finally, very imperfectly, distinguished feminines from
masculines, without, however, inventing any new forms indica-
tive of this final division. Thus Gesenius* states that “the dis-
tinction of the feminine sex is often avoided by early writers,
Hebrew as well as Arabic,” and Kalisch, that “as the great
number of communia proves, usage may in many cases have
long fluctuated. * * * * Certain it is that both genders were
not so strictly distinguished by separate forms.as in later epochs
of the language, but, throughout the books of the Old Testa-
-ment, we find traces of a gradual and struggling development
in that direction....In a very considerable number of instances,
the masculine of the pronouns, suffixes and verbs is, in the 2nd
and 3rd persons, used instead of the feminine, and they occur
in such variety and number that it is scarcely possible to
consider them as anomalies or inaccuracies of expression.”
Similarly, the words mother, concubine, ewe, she-ass and others,
names of countries, of parts of the body, of tools and utensils, of
light, fire and other powers of nature are all construed as
feminine, though generally not feminine in form: while the
feminine form is actually preferred only for lifeless things,
abstract ideas, collectives and adjectives used substantively.
Much the same rules seem to obtain also in Chaldee?, and in
Assyrian many feminines have no distinctive termination’.
Generalizations like these, unanimously adopted by the most
eminent professors, seem to me to afford considerable prima
facie evidence for a theory of genders similar to that already
suggested for the Indo-European tongues*. The peculiarity in

4B. Stade

1 Hebrew Grammar, ed. Roediger,

trans. Davies, pp. 239—242; Kalisch,
Heb. Gram. 11. 107 sqq.; Nordheimer,
Heb. Gram. vol. 1. pp. 216—218;
Wright, Arabic Grammar, vol. 1. §§
289—297. |

2 Longfield, Introd. to Chaldee, § 32,
p. 40.

3 Sayce, Assyrian Gram. p. 119.

(Lehrbuch der Heb..
Gramm. § 312 a) distinctly says, “The
use of the feminine to form abstracts:
is explained by the substitution of the:
opposition of males to females for that
of persons to things, which latter op-
position seems to have preceded, in
Semitic languages, the differentiation
of the two sexes.”

60 THE JOURNAL OF PHILOLOGY.

this case would, of course, be that when the animates were
divided according to sex, the feminines received, by way of dis-

tinction, not a new form but that already appropriated to manim-

ates, so that the result is rather to distinguish masculines
alone from feminines and inanimates combined,—a practice not —
without analogies in other phases of Semitic civilization. To
return, however, to the Indo-European languages: it is clear
that if, in fact, the masculine and feminine represent the old
animate gender and the neuter the old inanimate, the neuter
must be older than one of the other two genders. But the
masculine is clearly older than the feminine, so that the question
of absolute priority lies between the masculine and neuter.
Now we are accustomed to hear it argued that of two languages
or two words that is the older which is the simpler in construc-
tion, i.e. which is.the less abundant in functional suffixes. But
the main distinction between the neuter and masculine is that
the former is deficient in about a third of the chief suffixes
which belong to the latter*. The nominative, accusative and
vocative of neuters have but one form, which in the plural is
much shorter than the masculine; and, in these cases, neuters
singular of the consonantal declension are but a bare stem while
masculines have extra suffixes, s or its equivalent in the nomin-
ative and m in the accusative. It is argued, however, that the
final m of neuters in the vowel-declension is borrowed from the
masculine accusative, while the short neuters in the consonantal
declension are abbreviations of the masculine form. As to the
former supposition, surely it is equally tenable that the mascul-
ime accusative m is the neuter form, chosen for this reason, that
an animate thing, when conceived as passive, is to all present
intents inanimate: and as to the latter, is it conceivable that a
whole nation should consciously and systematically reject a
customary suffix in order to mark a new distinction in things ?
It is an almost invariable rule that new associations of words
are marked by new suffixes or internal modification, and it
seems to me unscientific to assume an exception where another

+ The argument of the text may be _ survival from a very remote antiquity,

reinforced by the analogy of the dual shews the same paucity of forms as
number, which, admittedly a useless the neuter gender.

NOTES ON GENDER. 61

explanation is possible. It is observable also that in Latin,
which has preserved so many primeval usages, undeclined
nouns are always neuter, and undeclined adjectives (e.g. nequam,
frugi, tot) are always neuter in form. Let us recall here also
the statement of Caldwell, already cited (supra, p. 45), regarding
Tamil, that “in older Tamil and poetry we find many words
neuter which later and in prose are masculine, and the suffixes
which distinguish the masculine, feminine and common genders
are only mutilated pronouns.” Internal evidence and analogy
are thus strongly in favour of the priority of neuter forms to
masculine, and it is besides a priort improbable that the Aryan
people, at quite a late date, unanimously invented a new gender
which not so long afterwards many of them began to discard.
With regard to the other two genders, it will be sufficient to
say that the masculine is everywhere admitted to be older than
the feminine. In the vowel-declension, where alone the two
are differentiated, the feminine form with lengthened vowel is
clearly later than the masculine, which also is always used in
words of common gender*. The reader who desires more inform-
ation on this topic is referred to Grimm’s Deutsche Gram-
_ matik, 11. pp. 313—315.

The theory of the history of gender in Indo-European lan-
guages, which I have here tried to sketch’, is briefly as follows.
These languages had at the earliest time no gender-distinctions
but afterwards divided nouns into names of animates and those
of inanimates, keeping the old forms for the latter but adding

ae

1 Thus parens, in old Latin, is mas- 151.
* culine even when it distinctly means 2 My theory may be concisely shewn
mater. Festus, s.v. masculine, Mill, by a diagram, thus:

No gender
1

vr . « 3
animate inanimate
(new, later masc., forms) (old forms, later neuter, retained)
=

mase. feminine
(old anim. forms) (new forms or a few old anim. forms adapted)
For Semitic languages, the history of no old material was available, nor did
the classes would be the same, but of any new suffix suggest itself except the
the forms different. Inanimates, not neuter -t. Hence the identity of
animates, received a new form, suffix neuter and feminine forms.
-t. On the differentiation of feminines,

62 THE JOURNAL OF PHILOLOGY.

new suffixes to the former. Still later, the animate class was
divided into masculine and feminine, and with this change the
old forms of the animate gender were assigned to the masculine,
while new characteristics were invented for, or possibly old
material was specially appropriated to, the feminine’.

It may be said that this theory, even if it be plausible, affords
no explanation of the existing confusion in the distribution of
genders; but it is submitted that each new change in the
division of words would be somewhat imperfectly carried out
and would leave a considerable margin of anomalies. In the
first place, when the distinction between animates and inanim-
ates arose, not all living things would be assigned to the
animate class, and the exceptions will now be found as anomal-
ous neuters. Thus, in Ashantee, as we saw, ‘father’, ‘slave’,
‘child’, ‘maiden’ and other expressly personal names and col-
lectives of animates, as ‘family’, ‘company’, ‘party’ (though
these latter may possibly be plurals), belong to the impersonal
class. Similarly, in Aryan languages, we find wéu, pecus, das
Weib, das Rind, das Schaf and other German collectives,
diminutives and names of children and fruits all neuter. Some-
times, doubtless, inveterate usage of a familiar word would
retain the inanimate form for an animate name, and sometimes
(as in the case of réxvov etc.) animates would be deliberately
assigned to the inanimate class .because of their peculiar rela-
tion, as products, to other animates. On the other hand, very

many words, properly belonging to the inanimates, would be:

assigned to the animate gender. The agent in this case would
be a habit of vague personification, such as that of which Mr
Tylor speaks in the following passage. “Certain high savage
races,” says he, “distinctly hold, and a large proportion of other

savage and barbarian races make a more or less close approach |

1 For instance, many names of
females in common use, but not many
males, happened to have a long vowel
in the stem, or to be formed with a
suffix -na or -ana (e.g. Sk. patni, Gr.
déorowa, Lat. regina, Pol. bogini (god-
dess), Ger. géttin, Eng. vixen), and
these forms would become, as a rule,

but not without exceptions, distinetive
of the feminine. In the Semitic lan-
guages, on the other hand, the anim-
ates had no special characteristic, and
consequently, on the specialisation of
the feminine, there was no old material
to adapt.

NOTES ON GENDER. 63

to, a theory of separable and surviving souls belonging to stocks
and stones, boats, food, clothes, ornaments, which to us are not
merely soulless but lifeless” A long and striking array of
examples is given by this writer, but need not here be cited, for
our business is only with the effect of this habit of personifica-
tion on language. In Ashantee, ‘broom’, ‘knife’, ‘pot’ and
other names of inanimates are assigned to the animate. Among
the North American Indians names of trees, the calumet, the
tomahawk, arrow, kettle, piece of wampum and other objects of
familiar use are all habitually animate, and, according to School-
craft, already quoted, any other inanimate may, by personifica-
tion, change its gender®. It is difficult, of course, to find, at
the present day, equally certain examples in Aryan languages,
but the habit of mind now in question was certainly common to
our forefathers, witness only the ancient practice of trying for
manslaughter a rooftree or other lifeless thing which had accid-
entally killed a passer-by: and that the habit of mind affected
the language is rendered in the highest degree probable from
the fact that from the earliest times we find inanimate objects
receiving proper names. ‘hus we read of the ship Argo, of
_ Thor’s hammer Midlnir, of Arthur’s Excalibur, Sigurd’s Gram,
Rustum’s Brand, the Cid’s Tizona, ete. These analogies will
suffice to shew what influence personification could exercise in
swelling the numbers of the animate class. All words of this
gender were, according to my theory, ultimately assigned either
to the masculine or the feminine, and in this redistribution old
and new influences would combine to produce an infinite and
confusing variety. The names of things, which contained no
suggestion of sex, or (as in words of common gender) where sex
was not material, would retain their old forms and associations
and thus be masculine*®: words in common use, properly femin-
ine, would, especially in the consonantal declensions, not be
transferred to their new gender: other words, properly either

1 Primitive Culture, 1. 477 sqq. ing of themselves in the plural, with a
2 See Tylor, Prim. Culture, 1. 285 quasi-sexless editorial ‘we,’ use the
—303. masculine,

’ 3 Similarly, in Greek, women, speak-

64 THE JOURNAL OF PHILOLOGY.

endseulina or sexless, would be assigned to the fonituine ‘because
they resembled, in form, a certain group of words which, by
virtue of their meaning, were now classed as feminine: by
these and many other similar obstacles a systematic classifica-
tion would be prevented. In selecting words for the feminine
gender, the merest fragment of a sexual characteristic would
suffice, whether such characteristic were original or only added
by association. Thus the Dyaks of Borneo say of a heavy down-
pour “watn arat sa” “a he-rain this”: in Bullom the thumbs
and great toes are called the male fingers, the others female: a
Chilian calls soft wool domo-cal, i.e. female wool.- In the French
navy, it is said’, masculine names were given to line-of-battle
ships (as Le Vengeur), feminine to frigates (as La Belle Poule).
The affection of a sailor for his ship leads him to speak of it as
‘she’, as does every mechanic of his engine. So, according to
Cobbett*, a Hampshire labourer refers to his plough with ‘she’,
‘but to all other tools with ‘he’. The distinction made by the
Englishman between ‘male’ and ‘female’ screws, is in other
languages indicated by a difference of gender. Thus in German
we have haft and heftel, haken and schlinge, ohse and éhre.
These last, according to Grimm’, used to be distinguished as
mannl and werbli simply, like the Italian maschio and femmina,
Arabic zend and zendet. Similarly, in Greek, we find puros
distinguished in gender from pvAy. Could there be a more
remote suggestion of sex than that by which certain rimes and
caesurae have been, even in modern times, styled ‘feminine’?
Yet sexual characteristics, no stronger than those here sug-
_ gested, would in primitive times have sufficed to determine a
word to the masculine or feminine gender, even as, in modern
German, the manly qualities muth and hochmuth are distin-
guished from the womanly demuth and wehmuth. It would
appear, also, as a further cause of confusion, that the separation
of the feminine, though begun earlier, was not completed till
after the scattering of the Aryan race: for except the lengthened
stem-vowel and the suffix -ana no marks of the feminine can be

1 Key, Language, Its Origin, etc., * English Grammar, Letter v. _
chap. on gender. 3 Deutsche Gram. m1. 359. _

NOTES ON GENDER. 65

considered common to all Indo-European languages. The ic of
Lat. victrix, the -ya of Gr. dépovea (fepovt-ya), are peculiar to
those languages, as are other forms to other languages. Lastly,
when more civilised habits of observation had displaced the
older tendency to personify, new words, created to represent
new things or ideas, would receive a gender suggested not by
their meaning but their form: compounds (e.g. with heit, ket
in German) would be assigned to the gender of their last com-
ponent; words with vowel-ending (preferred in Latin and
Greek) would be classed as feminine. At the same time, pho-
netic changes and the irksomeness of remembering distinctions
no longer significant would lead to endless mistakes, many of
which would become stereotyped (as in the case of frons and
crux, and many more which were masculine in Old Latin’), or
even to rearrangements of genders, regardless of form*. Other
similar sources of confusion, needless to enumerate or discuss,
will doubtless occur to the reader.

In the foregoing remarks I have endeavoured merely to
suggest a history of the Indo-European genders which will,
without violence, account both for their existence and for the
few regularities and multitudinous discrepancies of their dis-
tribution. The proofs which have been proposed, though not
absolutely cogent, yet seem to me more than strong enough
to support a hypothesis which can never be very hardly
worked. For unfortunately genders, as we know them, are
generally those of only one dialect, fixed at quite a late date
by the introduction of writing and the growth of a literature.
The want of record and the difficulty of conceiving the mental

1 Festus, s.v. masculino, 151 Mill.
ante cit.

2 Thus, according to Grimm, modern
German shows a tendency to make all
names of towns neuter, notwithstand-
ing that they may be compounded with
terms properly of other genders, e.g.
Bam-berg (m), Magde-burg (f), Lands-
hut (m), Elber-feld (n), Neu-hof (m),

Journal of Philology. vou. x.

&c. are all neuters. Deutsche Gram. 111.
426. The loss of genders, like that of
inflexions in modern languages, is
clearly due to the inability of foreign-
ers, adopting a new language, to
remember meaningless distinctions,
even where similar ones existed in
their native tongue.

66 _ THE JOURNAL OF PHILOLOGY.
attitude of a long-distant age will prevent the possibility of

explaining, with any show of certainty, the gender of a par-
ticular word. The same obstacles lie in the way of any practical
application of a theory as to the origin of language, but the

science of philology manifestly requires the aid of such tributary

speculations’.

1 The substance of this article was
written as long ago as 1876. While it

was printing, Mr Brandreth read a

paper on Gender before the Philolog.
Soc. in London. I am told that his
views on the probable growth of gender-

distinctions agree with mine. I may
here observe also that, in the article
on Grammar in the last edition of the
Encye. Brit., Prof. Sayee expressly
adopts the sexual theory of gender.

JAMES GOW.

4

ATAKTA.

1. ARIsTOTELES Poet. 1, 1447*8 (Vahlen) :—repi rointixns
auTns Te Kal Tv cidav avTHs, nv Tia Svvauw ExacToV TL Eye
KTE. |

The reading ékacrov tt has been introduced by Vabhlen ~
(instead of the common reading éxacrov) under a misapprehen-
sion, namely that the Paris MS. 1741 has &caotots. The MS.
has écaorTov, written thus, écaotoN.

2. Aristot. Poet. 3, 1448* 30:—68id xat avtirovodvtat THs
Te Tpaywdlas Kal THs K@pwdias ot Awpucis’s THS pév yap
Kopmoias ot Meyapels ...... Kal THs Tpaywdias Evior Twv ev
Tlexomrovynoe.

The grounds on which the Dorians claimed the invention
of Comedy are stated at some length by Aristotle, whereas
their claim to the invention of Tragedy is merely recorded as
a fact in a clause of some half-dozen words, as though it were
too far-fetched to deserve consideration. If we insert a & after
Kal ThS Tpayedias so as to make the clause mean ‘and even
Tragedy is claimed by certain of the Peloponnesian Dorians,’
we shall improve the sense as well as the construction of
the passage.

3. Aristot. Poet. 8, 1451° 16 :—pid0os 8 éoriy els ody bomrep
Tivés olovrar édv Tept eva 7 TOANA yap Kal adTEeipa TH Evi
oupBalver €& av éviwv ovdév éotw & ovtws 5é Kai mpakets
évds troAXail eiow €& ov pia ovdepla yivetar pais.

The év/, which has supplanted the old reading yéveu, is
- found in a late (Wolfenbiittel) MS.; but, if I am not mistaken,

5—2

68 THE JOURNAL OF PHILOLOGY..

it has on its side much better authority than this, namely
that of the Arabic version of the Poetics. Until we have
Prof. Sachau’s long-promised edition, the Arabic version is
a sealed book to us; in the interim however we may as a
pis-aller turn to the mediaeval Latin version, which is a
translation of the commentary of Averroes. Now the Latin ~
corresponding to the above passage runs thus :—

Uni etenim rei multa accidunt, et similiter reperiuntur
in una et eadem re actiones multae (f. 5 rect. ed. Ven. 1481).

The Arabic text, therefore, which is older than the oldest
Greek MS. and is based on a still older Syriac version, would
seem to presuppose évé rather than yéver. As for éviwv, I think
it should be bracketed, as due to a marginal év/ (intended as a
correction of yéve) which has found its way into the text
in the wrong place. It is not recognized in the Latin version,
and there is no similar limitation in the next clause (é&
ov pia ovdeula yivetar mpakis) which is in other respects the
counterpart of the clause with which we are dealing.

4. Callimachus Hymn. 5 (Lav. Pall.), 45:—
adpepov vopopopor wéev Barrete, capepov “Apyos
miveT amd Kpavav, wn © amd TO ToTAaLO.

This is the reading in Schneider’s edition; but I cannot
think that his note on the passage (1. p. 340) clears up the
difficulty in the words "Apyos miver’. If we are not prepared
to take miver’ as=criveras, or to regard “Apyos as a voca-—
tive =ye Argives, or to write (as Meineke suggests) er’ or
~wwe., there is still another possibility, viz. to read ricer’.
miere will of course govern ”Apyos; and as the future is prac-.
tically equivalent to an imperative (see Goodwin, Moods and
Tenses, § 25. 1. 5 5), Meineke’s doubts as to the pu) that
follows will thus fall to the ground. The pév after vSpodédpor is
Schneider’s unnecessary correction of the MS. reading, pu).

5. Clemens Alex. Paed. 3. 12, p. 307 Potter :—xal oixérais
\ , an al
HEV xXpHaTEoV ws EavTOIs avOpwrroL yap cio ws HMeEis’ 6 yap
Geos maow Trois édevbépors Kal Trois Sovdous eotiv, av oxoTIs,
”
loos.

ATAKTA. - 69

As the context literally swarms with quotations I suspect
that Clemens has here worked into his text the words of some
dramatic poet, perhaps Euripides. The addition of a syllable
or two gives us the following result : =

6 ap Oeds <tor> waot Tois <1’> énevBpors
Kal Toiot SovrAas éotiv, dv oKxoTrHs, toos.

6. Clemens Alex. Strom. 4. 8, p. 589 :—éwoiws 8€ Kal
codotos 6 wuPaydpevos érroincev kat Waddos 6 Aaxvdov yvo-
pyos, ws pnor Tindeos 0 tepyaunvos év TO Tept THs TaV
hirtocipav avdpeias.

Clemens is enumerating the philosophers (Zeno, Anaxar-
chus, etc.) famous for their fortitude under torture; and we
learn from this passage that their heroism was the subject of a
special work by a certain Timotheus of Pergamum, Who then
is the Paulus mentioned? As Lacydes resigned in B.c. 215,
I do not see how one of his disciples can have had such a
name, and am inclined to suspect that the right name was
Davros or PavrAros. The mistake may be as old as Clemens
himself, since Theodoret who copies him has the same reading.
As far as I know, this Paulus or Phayllus is not mentioned
elsewhere: judging from the dates one may perhaps suppose
the story of his sufferings to have been told in some way or
other as an episode in the history of the tyrant Nabis.

7. Clemens Alex. Strom. 7. 11, p. 871:--raya 8 odtau
Kal tods Oavpatotroiods avdpelouvs dyoovaow eis Tas payalpas
kuBiotavtas €& éumretplas Twos KaKoTexvodVTas éTl AUTTPO TO
pice.

Read émi AvTpe to picO@: comp. Diog. Laert. 10. 3:
ypapmata SidacKewv AvTpad Tivos pc Gapiov.

8. Galenus wepl Siadpopas ohvypov 2. 10, t. 8 p. 631
Kiihn:—ov'te yervnbels "AOnvneww ovte tpadels, adda xOes Kal
TpaTtws nKov €« KidtxKlas.

For mpaétas read wrpe@np.

9. Galenus wept tév rapa thv AéEw codicpatwv 4, t. 14
p. 595 —pia pév [scil. audptBoria], ijv Kownv dvoyafover Tod

70 THE JOURNAL OF PHILOLOGY,

Te elpnmévou Kal Tod Starperod, ola éotiv n avdAnTpis Tais
odaa’ Koln yap avT) Tod Te avAnTpls OvoMaTos Kal TOD
elpnmévou.

This occurs in an enumeration of the eight species of
amphiboly recognized by the Stoics. ATAHTPIZ was a
favourite instance of one kind of ambiguity, the question
being whether in a formula like this, ATAHTPI> ecotca

Snuocia éctw, it was to be read as one word or as two—

‘utrum aula quae ter ceciderit an tibicina si ceciderit debeat
publicari,’ as Quintilian says (Instit. 7. 9. 4; comp. Diog.
Laert. 7. 62). The parallel statement in Theon (Progymn. 4,
Rhett. Gr. t. 2, p. 81 Spengel) is as follows:—acady 5é tv
Eppnvelav tose? Kal 7 Neyouévn augiBoria pds THv SvadeKTLKOV
Tapa TV KoWnY Tov adiatpétou TE Kab Sinpnuevov, ws ev TH

ATAHTPIS zesotca dynuocia éorw' &v pév yap Ti eats TO

vd’ év kal adialpetov, avrAntpis Ectw Tecovca Sypocia, Erepov
dé To Sinpnpévov, avr Tpls tecotca éctw Snuoaia. By the
aid of this parallel it is easy to restore the words of Galen
to something like their original form:—pia pév, jv <tapa tiv>

\ ’ , AM % / \ A ? 4 w .
Kowny ovouagovet Tod Te Suypnuévov Kat TOD ad.atpéTov, ola

éorly 7 ATAHTPIS wecodaa’ Kown yap airy tod Te avAntpis
ovo“aTos Kal Tod Sunpnpévov.

10. Galenus (pseudo-Galenus) «¢ {oov TO Kata yaotpcs
5, t. 19 p. 176:—dgnot dé Anuoxpitos dvOpwrov é& avOpeov
éFeceta Oat Kal xuva é« KUVOS.

The astonishing word é£ecez?oAar is a blunder which has
survived from the editio princeps. As Democritus said éfée-
cvtat avOpwros €& avOpdmov (Stob. fl. 6. 57), it is pretty
obvious that what Galen wrote was éfecovcOa. A similar
correction has still to be made in Clemens Alex. Paed. 2. 10,
p. 227 Potter:—dvOpwmos yap é& avOpdrov éxdvetai te Kal
avooma&trar—where éxpveras is surely a scribe’s mistake for
éexoevetat. B. ten Brink (Philol. 8, p. 415) thought é«veras
due to a slip of memory on the part of Clemens himself.

11. Galenus Defin. medic. 487, t..19 p. 462 —vOovowac-

‘ ’ / > / Louan
pos €ott KaOarep éEictavtai tives emt TaV UVToOULLwpéevoY év

wt
{

ATAKTA. 71

Tois lepois OpavTes ) TUuTaveY ) avrAOY } TULBOX@Y axov-
oayres.

This definition may be made complete by the aid of what
precedes it in the text, but even then it requires some slight
correction. Read:—évOovciacpos éote <dtavolas éxatacis>,
Kka0atep éEictaytai ties UTO THY VTOOUMLLMpévwV <i) Ta> év
Tois iepois opevTes 7) TUuTavev } avr\av 7 KuuBddwv daKov-
oayTes.

12. Hippocrates wept tpodys 14, t. 9 p. 103 Littré:—
_yvrol pbeipavtes Kal brov Kal pépos Kal &wbev Kal evdobev
avuTo“aTor Kal ovK a’TopaTol, nuivy pev avTopatoe aitin &
ovK avTopaTot, aiTtins Oe Ta pev Ojra Ta S Adnra Kré.

Although the reading here given is as old as Galen (v. t. 5
p. 393, and t. 15 p. 300 Kiihn), we may be tolerably sure that
Hippocrates himself wrote not aitin and aitins but in both
eases érey—which is the regular Ionic equivalent of duces
as the antithesis to véu@ or mpos muds. Galen elsewhere
(t. 1 p. 417) shews himself quite familiar with the word.

13. Menander mept émidecxtixdv 2. 1, Rhett. Gr. t. 3
p. 346 Spengel:—zapadofa dé [scil. éyxadma], ofov *AdXKida-
pavtos TO ToD Oavatov éyxdov,) TO THS Llevias 7 Tod IIpwréws
TOU KUVOS.

I give the text as it stands in Walz and Spengel (who rely
mainly on two Laurentian MSS.), but every one must feel
that there is something clumsy and unsatisfactory about it;
one would naturally infer from it that Alcidamas wrote an
éyxa@utov IIpwréws tod xuvds! In stating the theory of
éyx@pta Menander divides them according to the nature of
the subject-matter into (1) évdofa, (2) audidoéa, and (3) ra-

paco£a’.

1 The existing text adds ddoéa to the
list; the addition is to my mind—and
Heeren takes the same view—an ab-
surdity too great to be fathered on
Menander, An ddotov éyxupuov is a
contradiction in terms: if you wish to
laud something déofov, you must by

Encomia of the Gods come under the first head:

an effort of ingenuity make it out to
be not really dd0f0v—in which case the
éyxwucov becomes what Menander in
the passage before us terms rapdadogop.
When I read the explanatory clause,
addoia 6¢ ta mepl Satmovwv Kal KaKod
gavepov, I cannot but think that the

72 _ THE JOURNAL OF PHILOLOGY.

under the second head he gives as instances the Panathenaic
orations of Isocrates and Aristides, one instance being taken
from early, the other from more recent literature. In dealing
with the third class we should expect him to select his exam-
ples on the same principle, but if Spengel’s text is right, this
is not actually the case. When we revert, however, to the
pre-Walzian text, that of the Aldine or of Heeren’s edition, we

find that the above passage once had a very different aspect: — —

mapadoka 5é, olov ’AXxidayavtos T6 TOU Oavatou éyxapior,
4 To THS Ilevias Ilpwréws Tov Kuvos. |

This, as I have ascertained, is the reading also of the Paris
MS. 1741, which is, I believe, older than the two ‘Florentine
MSS. followed by Walz and Spengel: Aristotelian scholars will
remember that it is now-a-days recognized as the critical basis
for our editions of the Rhetoric and Poetics; so that authority
as well as internal probability seems to be on the side of the
reading of the Aldine. If we keep the reading of the Aldine,
we recover the name of a writing of the famous Peregrinus
Proteus. We learn that he wrote a ‘Praise of Poverty.’
Menander would seem to regard it as a well-known book, and
also as one of sufficient importanee to be chosen from among

recent works to be put in comparison with the famous ‘Praise
of Death’ of Alcidamas,

14. Plato de Rep. 3, p. 411 B:—érav 8 éréyov pn avin

> \ a A
ANNA KN, TO META TOUTO On THKEL Kal NEL Bet, Ews av exTHEN
tov Ovpov Kal éextéun Bomrep vevpa ex THS ruyys.

Though the words xal Ae(Per are recognized by Demetrius
mept épunvelas 51 (Rhett. Gr. t. 3 p. 274 Spengel), I am in-
clined to think Ae/Bes a mere gloss on tHKes, the two words
being almost synonymous in sense (comp. Hermogenes zrepi
peOddou Sewortntos 4, Rhett. Gr. t. 2 p. 428). The passage will
read better if we cut it out as an emblema and suppose Plato
to have written, rev ws av éxtHnEn Tov Oupor.

addition must have been workedin by (vir Sap. conviv. p. 153 a), might pos-
a later hand, though I am aware that _sibly be quoted as an argument against
Plutarch’s ri BdaBepwrarov; Saluwy me.

=

ATAKTA. | 73

15. Plato de Rep. 3, p. 411 D:—retOo0t pév Sia ANOyor ovdév
ére xphrat, Bia Sé kal aypiornTs @orep Onplov mpos TavtTa
Svar parreras,

Here Baiter, following K. F. Hermann, brackets dcamparre-
rat, in lieu of which Madvig proposes désatdrrerar—a reading

actually found in one of Schneider’s MSS. Prof. Chandler (Mis-

cellaneous Emendations and Suggestions p. 6) suggests that, if
any change is necessary, we should transpose and read mpocé:a-
mpatteta TavTa. ‘The real difficulty in fact is not so much in

the verb as in the preposition. If we provisionally bracket the

mpos, either as the addition of a scribe, or as representing some
small word not affecting the general construction, we shall have
as the result a form of expression precisely similar to that in
Gorg. 451 D:— pyropix? otca Tov Moyo Ta TavtTa SiaTTpaTTo-
pévov Te Kal KUpovpevov. |

16. Plato de Rep. 5, p. 473 A:—robto pév 8) pn dvayxaké
He, ola TH Oy SindOomev, ToLav’Ta TavTaTact Kal TS Epyo
Seiy yuyvopeva atropaivery.

~ Read either, yiyvouev’ AN dzrodgaivesy, or (as a friend has
suggested to me), T@ Epyo AN yuyvopeva arodaivev. In
favour of the second alternative I may remark that Stobaeus
and also some of the MSS. of the Republic have 67 instead of
deiv, and that the difference between AH and AN is very slight
indeed.

17. Plato de Rep. 6, p. 488¢:—avtods 8 avtT@ de To
vavkdnp@ Tepikeyvobar Seouévous Kal TavTa TovobvTas bTwsS
x / \ , > /
av odiot TO ndadLov emit péeven.

For ait@ we should perhaps read, av.

18. Plato de Rep. 6, p. 492 c:—év 6) T@ ToLvovT@ Tov véor,
TO Aeyomevory, Tiva ole. Kapdiav ioyew; 7 Tolav [av] a’t@ mal-
delay iovwtixny avOéFew KE ;

‘The dy after zrodav has been very rightly excised by Baiter
(following Cobet), The clause, however, which precedes requires
the particle, as the symmetry of the sentence demands a future
or the equivalent of a future in both cases. If Plato did not

74 THE JOURNAL OF PHILOLOGY. -

write &£ev (as Demosthenes does in a parallel instance (p. 842),
tiva olecOe avtnv uynv éev;), the reason, | presume, was
that ée.v would offend the ear, when avOé£e.v was so close to it
in the context. But by the insertion of a single letter we get the
equivalent of a future, if we read, riv’ AN oieu kapdiav toyewv;

19. Plato de Rep. 6, p. 503B:—iv yap SindOopev pvow
Seiv vrdpyew avrois, eis ravTO EvxppiecOas avThs Ta pépy OX-
yaks e0éret, TA MOANA O€ SteaTTaTMEéeVa HveETat.

The parts were surely not dverracpéva but dieorappéeva—a
distinction which we realize more easily when we reflect on the
difference between d:ac7mdy and dvaczreipev, though no doubt
the distinction is not quite so clearly marked in the case of the
perfect passive. But there are passages in Plato in which we ©
find the perfect from dvac7eipev, and in which no one would
‘wish to see the perfect of dé:ao7ay substituted for it:—

Phaedr. 265 H:—eis play idéav cuvopdvta d&yew Ta TOAKAYT
SueoTrappéva.

De Rep. 455 D:—opolws Svectrappévat ai pices ev apudoiv
Toi Cadow.

Legg. 945 C:—ovs...ulav odcav diow Svecrappévny Torra-—
Nov, TONNOLS OVOMacL TpOTayopEvopmeD.

A thing is said to be dseoracpévoy when its unity is lost
through a more or less violent disruption—or, to use Aristote-
lian language, tatta padiocta Siacmatat, & eis TovvavTiov Te
Kal ioxupas EdxeTat Kal Kiveitat (Probl. 5. 39, 885° 8). The
explanation in Hesychius, dvaemacau’ Svacrapata, dvacxylicat,
gives, I think, a fair notion of the ordinary use of the word in
Greek writers. On the other hand, the elements which go to
form a composite whole may be said to pre-exist separately,
Sveotrappéva, before their union, but how can we say they pre-
exist dsveoracpéva? In short there is a distinction between the
two participles corresponding to that which we have in English

between separate (adj.) and separated [compare Cobet ov.
Lectt. p. 412].

If what I have just said is true, there are probably not a few
passages in Greek authors in which déowadtaut, SueomadOar, and

ATAKTA. 7s

Svec7ra>pévos have usurped the place of SéomaPraz, Sead POar
and SvecraPpévos. I am inclined to think that this is the case
in the passages in Aristotle’s De Gen. Anim. in which he dis-
cusses the Empedoclean embryology. The theory of Syngenesis
(I take the term on the authority of Lewes’ Aristotle, p. 353), as
maintained by Empedocles, Hippocrates and others, affirms the
embryo to be the product of the union of male and female yov7):
in order to explain the likeness of the offspring to both the
parents, it seemed necessary to assume that both parents make
the same sort of contribution to the physiological result. The
elements of the body of the offspring accordingly were said to
pre-exist partly in the yov7) of the male and partly in that of
the female; but before their pZés, it is clear that the elements
must exist apart, and when thus existing apart they would in
Greek phraseology be spoken of as dveovrapuéva rather than as
dueomracpuéva. I conjecture therefore that Empedocles must
have written dvéo7raprav in the fragment preserved (in a sadly

mutilated form) by Aristotle (De Gen. Anim. 1. 41, 722°
10 =v. 270 ed. Stein):

pyol yap év TH appevt Kat Onrew olov cipBorov éveivar, Odov
& dm ovdetépov amiévat, “arra Stéotactar pedreov vars,
» ev ev avdpos’.”

The same correction has to be made more than once not
only in the immediate context but also in the second discussion
of the Empedoclean view, in the Fourth Book of the De Gen.
Anim. (4. 9, 764” 3). As far as I know, all trace of the word I
wish to restore has disappeared from the MSS. of Aristotle, but
the conspiracy of the scribes has not been equally successful in
the case of the text of Galen. Galen too has a criticism of Em-
pedocles: now one of his objections to the theory of Syngenesis
is this, that if the parts of the offspring pre-exist separately in
the two parents, we shall want a tertium quid, some formative
principle, to account for the possibility of their union (crept
oméppatos 2, 3, t. 4 p. 617 Kiihn):—dadrXov tivds dpa tpirov
Sejoer Tois Sueotrappévors ev éxatépw Tav oTEpuaTov pépect,

1 The next line was perhaps something to this effect: 1 dé yuvatcelasce
yovais tc xwpls éovea.

76 THE JOURNAL OF PHILOLOGY.

rod ouvtafovtos avta Kat Siaxocpyncovtos. If Galen wrote
this, it seems clear that he read d:éo7rapras in the line of Em-
pedocles; and I cannot believe that in other places in the same
discussion he wrote Sséo7ra>Oas (p. 616) dveorvraxXpévor (p. 616)
or Svecrradpéva (p. 618) in his paraphrases of the lengua of
the quotation.

20. Plato de Rep. 6, p. 503 E:—ép Pe ne TOANOLS yuUp-
vatew Set, cxoTrovvtas, eb Kal TA péyiota pabhnuata SvvaTy
gorar eveyKely, elte Kal amrodeiiacet.

The received explanation of the feminine dvvar7 is, I imagine,
to assume with Schneider that the subject present in Plato’s
mind was @v’ovs—a word which in Baiter’s text is just 23 lines
off, with all sorts of things in the interspace! I have more
respect for Plato than his commentators seem to have, and
prefer to think that he wrote not duvatn éoras but Svvycerac.

21: Plato de Rep. 6, p. 504 B:—rta&v pévrow Sumpoobev —
mpoeipnuevon éTomévas atrobelEes ody 7 ely mpocarat.

Read: éyouévas. Similarly in the passage in Poltt. 271 E:
dca THS ToLavTNS eoTl KaTaKOopHnoEwWS ETO MEVA, We May per-
haps be inclined to give Plato the benefit of the doubt, and —
believe the word as written by him to have been éyépeva.

22. Plato de Rep. 6, p. 511 A:—eixoot 58 ypwpévny [scil.
Thy wuxnv] avtois Tols UO TOV KAT aTretKacOeict Kal éxei-
vous mpos éxelva ws évapyéct SedoEacpmévors Te Kal TeTLNMEVOLS.

This is part of Plato’s statement of his theory as to the
nature of dvavoa and its objects, ra dvavonra, but the ingenu-
ous youths for whose benefit the explanation is given in the
dialogue must have been easily satisfied if they could accept
the explanation as it stands without a protest against its
obscurity. That dSvdvora cannot dispense with the assistance of
eixoves, sensible figures and diagrams, is stated here and else-
where with sufficient clearness: the present passage however
has been supposed to imply that dcavora requires in addition a
second sort of edxoves, here described as avra ta Urb Tay KaTo

atreixacGévta. Accordingly Schneider, and more recently Mr
Henry Sidgwick (Journal of Philology 2, p. 96), think the

ATAKTA. 77

formula ta tid tév Kato atretxacGévra means literally, ‘the
things which are themselves copied by the things below them
in the scale,’ and is thus a description of dsavonra. If dcavonra
are themselves eixdves of something higher (vonra proper, if I
may use the expression), the mind in the stage of Scdvoa may
very well be said to deal with two kinds of e/xoves—which for
the sake of brevity I may distinguish as eixéves aicOnrai, and
etxoves Stavontal. This explanation is not in itself un-Platonic
(see esp. Rep. 7, p. 516 A, the Scholiast on 511 A, and Proclus
in Euclid. Prol. 1, p. 10 Friedlein), but one may doubt whether
it has any relation to the passage before us. Two objections
may be urged against it. (1) If the words were to bear the
interpretation put upon them, one would have expected Plato to
prepare us in some way or other for so important a statement ;
whereas, if this refinement is really in the text, it is introduced
without a word of warning—without a syllable in the context
to suggest it. (2) The context on the other hand supports the
older and more natural interpretation, viz. that avta Ta arrei-
xaobévra denotes the sensible eixoves. The sentence in which
the formula occurs is a mere repetition or summary of what
precedes; and we have in the sentence which immediately
precedes this what I take to be an obvious equivalent of avta
Ta amrecxacbévta :—

7 \ \ a AN , , \ ,
avTa pev TAUTA A WAaTTOVGL TE Kai ypahovaL.

So that avra ra drexacbévta=avtTa a ameckafovor, and
denotes the sensible copies, and not the (intelligible) originals—
TO opotwbév, and not To @ dpuoiwOy. The general sense of the
passage seems to me to demand this (the old) interpretation,
but we cannot get it. out of the text as it now stands, which
must therefore be amended before we can interpret it on the
ordinary principles of philological probability. A very slight
change however will give us the meaning which the logic of the
passage necessitates. Let us for two rév Katw read vroxata,
and for cal éxelvors, kat éxel (or rather xaxe?), so as to make the
passage run thus:—

Ui \ / b] a a a a
eixoot O€ Ypwpuevny avTots Tols UmoKaTw aTretKacOetor KaKet
\ an
mpos exeiva ws evapyéat SedoEacpévols Te Kal TETLUNPEVOLS.

78 THE JOURNAL OF PHILOLOGY.

In regard to troxatw I may refer to Madvig (Advers, 1
p. 27) who has a good deal to say about the corruption of the
cognate word aepuxatw. As soon as vroxatw was turned into
tad TON karo, the further corruption of cai éxe? into Kai éxei-
vous was probably deemed an improvement: it was not seen that
éxel, coming after vaoxatw, meant ‘in the lower sphere,’ the
region of sensible counterfeits in which the common man is said
to live. A very similar instance of «axe? with a supplementary

clause trailing after it is found in another place in the Republic |

(7, p. 532 B):—é« tod Katayelou eis Tov HALOV emavodos, Kal
> an \ \ \ a? \ \ \ \ w-1€ / an ge
éxel mpos pev Ta Coa Te Kal huUTA Kal TO TOV HAiov das ET

advvauia Brérew xté. As I have been led to quote this place -

in the Seventh Book, I may as well say here that the reading

ér aédvvapia, which the editors take from Naegelsbach who.
. conjecturally restored it, is to be found in the paraphrase in

Iamblichus wept ts Kowhs pabnuatixns értotnuns printed in
Villoison’s Anecdota t. 2 p.196. Iamblichus reads érz advvapia,
which shows how the faulty reading éIl’ advvayia originated,

23. Plato de Rep. 10, p. 607 B:—xal péyas év adpovev

/ ee a a Vv A
Keveayoplaict Kal 6 TOV OLacoPav OXAOS KpaTOv.

The prevailing view among Editors is that we have here two
quotations, the second beginning after cal. This is by no
means obvious, but even if the view is true, it seems to me not
impossible that Plato combined the two quotations into one. I
suspect that «at 6 is a mistake for 6 «al, and that the original
_ reading was, 0 cai tov Ala copav dynos Kpatav—rtov Aia being
governed by xpat@v used perhaps in the sense which the verb
has in Aeschylus Choeph. 958, xpatettat ro Ociov. The reading
of the Paris MS. is rev dla codar.

24. Porphyrius de Abstin. 2. 34, p. 104 Nauck:—rois 5é
avtov éxyovots, vontots && Beots nbn, Kal Thy éx TOD NOYyoU
vuv@diav tpoaberéov. |

Porphyry is distinguishing between the honours due to
the supreme deity and those due to his offspring—the mortal
gods of Plato’s Timaeus, Read, therefore, @vntots for vonrois.

4

ATAKTA. | 79

25. Porphyrius ad Marc. 1, p.193 Nauck :—@vyartépwv pév
mévte, Svoiv 8 appévav ovaay pntépa, Tav pev Kal ere vytrior,
Tov dé non els yapou HALKLaVY HAAY époppovyTwr.

If, as Nauck thinks, 78déyv is an intruder, how did it come
to find its way into the text? On the alternative hypothesis
that Axia is the intruder we may by the change of an accent
restore és ydauou 7S8av—which looks like the end of a line of
some bucolic poet. A similar instance of the use of the sub-
stantive 78 I am unable to find, though Oppian has the verb
in a sense approximating to that which we want here: elapse 8€
yAuKvs olatpos avayKains afppoditns Kal yapor n8dacr (Hal. 1.
474), - époppovytwr, at the end of the passage, seems an odd
word to use when one would expect édopuovTwr or oppovTav.

26. Strabo Geogr. 14. 19, p. 658 Cas,:—xal Kal’ nyuds
Nixias 6 Kat tupavynoas Kear, cal ’Apictwv 6 dxpoacapevos
TOU TEPLTATHTLKOD Kal KANPOVOMNCAS Exetvor.

This appears in an enumeration of the eminent natives of
Cos. Aristo the Peripatetic is a well-known personage; but
who is the Aristo whom it was possible to describe, as Strabo
seems to do here, as ‘the pupil and heir of the Peripatetic’? If
we cannot know more about this Aristo, it might be as well to
get rid of him altogether, which we can easily do by the inser-
tion of a single letter, so as to read "Aplatwvo> axpoacdpevos.

27. Timo Sillogr. ap. Diog. Laert. 2, 126 (ed. Cobet) :—
AHpov advactncas sdpvwpévos appoorPouBak.
Anpov is a conjecture, the MSS. having Adyov. I would suggest
dyXov, or Aaor.

I, BYWATER,

°..

NOTES ON SOME PASSAGES IN THE POLITICS.

Pot. IV; 111;

The whole of Pol. bk. 1v, ch. iii (or ch. 3 and ch. 4, §§ 1—

15, 1289 27—-1291" 13) is considered by Susemihl ungenuine,
-and due either to one interpolator, or, because of the parallelism
of the transition passage 1290” 21 (67c pév ovv rodsTelas
melovs K.T.r.) to the transition passage 1291? 14 (67e weév ovv
lot ToNTEtat TAcious x.T.r.) at the end of ch. iii, perhaps to
two interpolators. In the latter case Susemihl would make
the second interpolation begin at the first of these transitions
(1290” 21), and conjectures that the second interpolator, finding
the first interpolation already in the text and supposing it
referred to by the words in 1291? 14, joined on to it the
part which he had himself written by a transition (1290° 21)
imitated from 1291" 14. This seems seareely possible, for the
whole of the first transition passage reads thus:

a \ > e / \ x. A > / y P
OTl “ev OV TOALTELaL TrAELOUS, Kal Sb HY aiTiav, elpnTat
q? \ / a bf , \ / \ \ / ,
du0Te 6€ TAElous TeV elpnuévav, Kal Tives Kal dia TL, Aéyopev
apynv AaBovTes THY Eipnuévnv TpoTEpov.

Thus it would be implied that the subject of the first (supposed)
interpolation—the reason why there is a plurality of Con-
stitutions (cf. 1289 27)—-was done with, and that a new subject
was to follow—the reason why there are more than have been
mentioned (one might suppose democracy and oligarchy in-
tended, which at the end of the first part of the first interpola-
tion are said to form the usual division of Constitutions).

¢ - - ‘
1 tt ee eee

= PL, 6-5 ard iy pct Ce av tt ee 5 1 t os
‘ kanes + : | — 4 r 7
lt = wu wre 4 >

NOTES ON SOME PASSAGES IN THE POLITICS. 381

But instead of this the second (supposed) interpolation
treats over again the main subject of the first, and more fully,

ending with the same statement about the usual twofold divi-

sion of Constitutions into democracy and oligarchy: and the
writer of it therefore could hardly have joined it to the first
interpolation by such an introduction.

‘A more probable account seems to be that two parallel
versions have here been unskilfully put together, not intended
by the author of either to stand in the same context.

The two parallel passages are 1289” 27—1290* 29 and
1290” 21—1291° 13.

Thus (i) 1289> 27—8 = 1290> 24 and 1290° 389.
(ii) 1289” 28—1290* 3 corresponds to 1290 40—
12917 10 and 1291? 33—? 1.
(iii) 1290 5—7 and 11—13 correspond to 1290”
25—38.
(iv) 1290* 13—16 corresponds to 1291 1—13 (esp.
8—13).
Susemihl says (Intr. p. 58, |. 6, edn. of 1879) that the inter-
polator of ch. iii (1289 27—1291” 13) refers (1290* 1 seqq.) to

Pol. bk. vil, and therefore had the ‘original’ order of the books
before him. ‘The reference is found in one version only of the

proposed resolution of ch. iii; the second version instead of the

reference inserts a long passage similar to that part of bk. viz
which the first version refers to. So far therefore it is by no
means certain that the second version is older than the received
order of the books.

The words however at the beginning of the second version,
dpxnv thy eipnuévnv mpotepov (1290.23), may perhaps refer to
bk. vil; but Susemihl thinks bk. Iv, ch. 111 1s intended, 1.e. the
beginning of what seems the other parallel version in ch. iii.

A third parallel version seems to be found in ch, iv init.
1291” 14—30.

The passage 1290* 30—° 20 which intervenes between the
first two versions is obviously parallel to bk 111, 1279” 11 seqq.

There is room for the suspicion that the preceding part of
the book, chs. i and ii, also contains three versions. . The list of

Journal of Philology. vou. x. 6

"A Vala. & The Te lb ee Se eh er en ee ee
’ = = ~ F rey

82 THE JOURNAL OF PHILOLOGY.

contents in ch. i is repeated in the last part of ch. 11. The first
part of ch. ii may be another independent list. For it announces
the subjects thus, Aoumdv mept modutelas OvedOely TIS TO KoWm

\ n / q n
Mpocayopevowéevns avowatt Kal mept TOV GAWY TONTELOV,

oruyapylas te Kal Snuoxpatias Kai Tupavvidos, 1289* 35: and
(1), democracy and oligarchy having been discussed, it is said
(ch. vi or 8 init.) that two subjects are left (Novrcy) mondtrela
and tupavvis; (2) after the account of zoduTe/a it is said that
tupavvis is left (aepl 5€ tupavvisos jv Nowzrov eirety 1295* 1,
cf, reXevtatoy 1293” 25): while (3) after the chapter on tupavyis
follow other subjects peculiar to the lists in ch. 1 and the last
part of ch, ii. The argument is not conclusive, and the unity
of the text could be defended ; so that the evidence for disunity
might not be worth stating were it not for the more obvious
triplicity in chs. iii and iv.

The double enumeration of the kinds of democracy and
oligarchy in chs. iv—v (4—6), is also suspicious. |

IL.

Pol. Ill, x—x1.

Chs. x and xi (or chs. 15 and 16, 1285” 34 seqq. and 1287
1 seqq.) discuss the same subject, the wauBaourela. Out of a
part of ch. x (15), 1286* 26—® 3, and a part of ch. xi (16),
1287" 8—35, Susemihl forms two parallel versions, printing

1286* 26—» 3 + 1287" 8—15 parallel to 1287 16—35. The

remaining parts of these chapters he tries to form into a con-
tinuous context by a number of rearrangements.

It may be however that the two chapters belong almost
wholly to two parallel versions, and that instead of being com-
bined in this way they should be still further resolved.

The beginning of ch. xi (16), 1287 1—8, is closely parallel ~

to a passage near the beginning of ch. x (15) 1286* 2—7, and
a better case for parallel versions can be made out here than in
the part selected by Susemihl.

The matter which follows these is in general of the same
kind in both chapters for some distance, from 1286* 7 to 1286” 8

r
a

NOTES ON SOME PASSAGES IN THE POLITICS. 83

in ch, x (15), and from 1287 8 to 1287” 35 in ch. xi, though
there are additions and differences of arrangement in the one
context as compared with the other.

Thus (i) 1286* 2—7 = 1287" 1—8.

(ii) 1286* 7—24 corresponds in subject to 1287°
18—® 8 + 1287” 16—24.

Compare 1286 7—9 with 1287 18—19 and 1287° 20—1
1286? 11—16 with 1287% 33—® 5
1286* 16—20 with 1287 28—32
1286 21—2 with 1287° 23—4
1286 23—4 with 1287 17—18
1286* 24 with 1287° 17 and 19
1286° 26 with 1287” 15—16.

(iii) 1286 25—” 3 corresponds in subject to 1287"
24—35 and 1287° 8—15. Of these passages, the third disturbs
the context and looks like a parallel version of the second; there
is some ground therefore for placing it as Susemihl does.

From the above it would seem also that the first part of
Susemihl’s second column, 1287 16—26, with the exception of
the first two lines, has its analogue not in his other column but
higher up in ch. x, that is in the first column of the resolution
of the text here proposed.

In ch. xi (16) the passage 1287* 248 interrupts the argu-
ment of the context: it belongs to the same part of the subject
as 1287 16—23, and may be read after éoriv, 1287° 23: but if
it belongs to this place it is hard to see how it could have been
removed.

The repetition at the beginning of ch. x (or ch. 14 fin.) of
the characteristics of the first four kinds of monarchy, before
the introduction of the fifth, may indicate that the discussion of
monarchy was double from the beginning: though the circum-
stance taken by itself would not count for much.

6—-2

84, THE JOURNAL OF PHILOLOGY.

III.
Pol, Vv, i—ii1.

_ Pol. v, i may be divided into three parts from 1301* 25
onwards, viz. 1301? 25—1301" 6, 1301° 6—1301* 25, 1301° 26
—1302? 15. The third of these does not cohere with the
second, but is an abrupt unexplained return to the subject of

the first, which has been already wound up with the words |

a & , ’ .
dpyal péev ob ws elev avTal Kal THYyal TOV OTdcEwY ELoW,

1301°4—5. The first and third passages seem to be duplicates: _

both deduce in the same way political disturbances (and the
existence of different forms of government) from the different
interpretations which contending parties put upon a commonly
accepted principle of right. Compare especially 1301* 26—7

with 1301° 35—6: 1301* 28—9 and 1301* 31—2 respectively ©

with 1301° 37—8 and 1301 88—9 together with 1301 39—40.

Ch. 1i, 1302? 16 seqq. returns again to the causes of political
changes: Anmwréov KaforAov mpOTov Tas apxas Kal Tas aitias
avta@v, words which should be contrasted with the ending of the
first passage, adpyal pév ovv ws eimeiy x.t.r. (quoted above).
A wider classification of these dpyai is given, and under the
first head the main thought of the other two parallel passages
is repeated in a shorter form. There is here then perhaps
another rewriting, seemingly by a later hand, of the introduction
to the book, and with this third beginning seems to cohere the
rest of ch. 11 and ch. iii.

The references in 1301” 36—7, 1302* 24, if genuine, may be

to the third book of the Politics, like the reference in 1301% 28:

thus there would be three parallel references to the third book
of the Politics (111, 9 init., or its probable duplicate 111, 12 init.),
one in each of the supposed parallel passages.

IV.
Pol. vil i—iii and xii—xiii (=13—15).

Susemih] has remarked (Note 712) that the subjects of ch. i
and chs. ii—iii are repeated by chs. xii (13) and xiii (1415).

NOTES ON SOME PASSAGES IN THE POLITICS. 85

Possibly there is even a threefold treatment; for ch. xii (13)
seems like a shorter duplicate of ch. xiii (14—15). In each the
same question is proposed, What is happiness or the chief Good ?
(compare 1332* 7 and 1333* 15—16); and the discussion of it
is followed in each by a transition, in almost the same terms, to
the subject of education (compare 1332? 39—1332° 11 with
1334° 5—11 seqq.). The chief difference is that ch. xii (13,
1332* 7—9) takes the definition of the Good in the general form
given in Nic. Eth. 1, vii (or Eud. Eth. m1, 1), while ch. xiii.
(14—15), like Nic. Eth. x, distinguishes between the life of
moral virtue and the higher life of philosophic contemplation,
the second discussion not being put as a continuation of the first.
The division of Goods into cadd and dvayxaia is made inde-
pendently in both chapters (xii and xiii), compare 1332 10 seqq.
with 1333* 32—-° 3 and 1334 16 seqq.: but in ch. xii there is
no consciousness of what seems implied in ch. xii (15), that
moral virtue does not belong so completely to the xadov as
Oewpla. — |
The nature of the parallelism in the two transition-passages
(1332? 39—1332° 11 and 1334” 5—11, compare especially 1334°
%—8 with 1332 8—11 and 1334> 10—12 with 1332 5—8)
confirms the order of the text in the first of them against
Bocker’s transposition of 13832” 5—-6 (for which see Susem.
vol. i, p. 446, note 2), and makes Broughton’s supposition
(Susem. i, 462, note 2), that the second of them is an interpola-
tion unlikely, especially if the other parallelisms of the two
chapters be taken into account. Yet the beginning of the
second passage—1334° 6, ruyyavopev 57 Sunpnuévoe pcTepov—
may have been altered, unless the reference is to the Ethics.

In the version of ch. xii (13), the language which follows
the words dayev S€ Kal év Tots HOtKots (1332* 7), has more
affinity for the Eudemian than the Nicomachean Ethics. Suse-
mihl says (n. 876) of the distinction in 1332* 10, that ‘it is not
~ in the Ethics (Nic.), but is put here by Aristotle—if he is the
author—to avoid possible misunderstandings. It occurs how-
ever in the Eudemian Ethics 1238 6. 1332* 19 should be
compared with the same Eudemian context, 1238 6—7; and

86 THE JOURNAL OF PHILOLOGY.

1332° 22—3 with Eud. Eth. 12499 12. Compare also (though
not so distinctive) the use of ypjovs in the formula for the
Good in 1332* 9 with the repeated association of ypjow and
évepye(a throughout Eud. Eth. 11, i (= N. Eth. 1, vii, etc.)

¥.

Pol. 1, xili, 1260* 22, caOavep wero Lwxparns. The refer-
ence being to one of Plato’s dialogues, 6 Lwxparys would be
expected. One MS (P* Susem.) has the article, but it has not
been followed in the editions, though apparently right. The
ordinary reading is accounted for by the last syllable of dero.

The article is similarly wanting in one other place, Pol. vut,
vii, 1342° 23, d:0 Kados éritipa@ot Kal TotTo Lwxparer, though
the reference is to the Republic (cf. 1342? 32—3, 6 & év 77
Toditela Lwxpatns), but probably 7@ has been lost after
Tov TO.

Pol. Iv, xiv, 1298? 1,
Sevtepov Se TO wept tas dpyas todto 8 éartilv as Se
Kat Tivey elvat xuplas Kal trolav twa Set yiverOat THY aipecw
auTov.
For éotiv as should perhaps be read éori (r/)vas: one tt
may easily have dropped out.

J. COOK WILSON.

{
—! ied
—_. , a

OBSERVATIONS ON THE OEDIPUS COLONEUS OF
SOPHOCLES.

(Read before the Cambridge Philological Society.)

THE following paper is chiefly occupied with the elucidation
of passages in the Oedipus Coloneus which have been obscured
through inattention to the main argument of the play. I will
begin with the passage that suggested it.

384. Tovs 6€ govs b1rot Beol
MTOVOUS KATOLKTLOVTLY OvK exw pabetp.

The present reading is intolerable, whatever sense we assign
to it, as Professor Madvig has seen. He conjectures, in the
Adversaria, xafoppsotaw, which is hardly near enough to the
MSS. Besides, an examination of other places in the play where
Oedipus speaks of his last resting-place suggests a different
metaphor. The houseless, homeless wanderer finds at last in
the territory of Athens the promised dwelling of which he has
been so long in search.

Thus in vv. 87—93 ®oiBw... | 6s wou Ta TOAN exety’ 67
é&éypn xaxa, | ta’tnyv éreEe wadrav ev ypove paxp@ | éAGovTe

‘ yopav Teppiav... | évTadda Kayrpew Tov taraltwpov Biov |
Képon pev OLKHTaVTA Tos Sedeypevots | ATHY Sé Tots TéprpacwY
of p amndacay. Oedipus is to dwell in the land to work
mischief to those who cast him out, and to be a source of ad-
vantage to those who received him. The plural «épdy has the
same force as in traiyvia, amores, &c. This natural sense is
needlessly obscured by Madvig’s conjecture ofcicavta. Again
in v. 626, 627 odor’ Oidirour épeis | aypetov oikntipa SéEac-
Oat torwv. So in 635, 636 ayo oeBicbeis ot ToT éxBaro
Xap Thy Todde yopa 8 EuTradw Kkatoiki@. Tr ‘act the con-

88 THE JOURNAL OF PHILOLOGY.

trast between the knightly hospitality of the stranger Theseus
and its reward and the very different conduct of Oedipus’ own
kinsmen and its retribution, is the very pivot of the play. The
generous promise in the passage just quoted, ywpqa S éumadu
(€umoXw) KaTovxtd, is in marked opposition to the conduct of
those who first drive the old man from their borders (é«BadXov-
ow, é&eXavvovow); and then, when forced by a divine necessity,
grant him the privilege—the privilege of what?—of lying just
outside their borders (v. 401 @vpacu Keipévov, 784 ovy wv’ és
Sopous ays, GAN ws Wapavarov oikions).
Hence in v. 631

tls Snr ay avdpos evpévercay exBaror

TOLOVE ;
the commentators are wrong in taking éxBanor as = ‘waste,
throw away:’ it means ‘drive from my bounds.’

So in Oedipus’ speech 1348 sqq. in which he invokes upon
his unnatural son a fate like the one in which he has involved
his father. He has driven his father from his country, and
from that country—for the curse is already working—he has
been driven by his brother never to return. To begin with, the
bitter antithesis of v. 1373 has been misunderstood :

Tolyap o 6 daiwa eicopa pev ov ti Tw

ws avrtix’, elmep olde KivodvTaL oxoL

mpos datv Ons ov yap éo@ ows more

Kelvynv épet TUS,
‘to the streets of Thebes: for I know one who shall never call
it native city.’ Oedipus says to his son ‘ You have driven me a
homeless exile from my native city (v. 1357 €@nxas dzronuv), and
with the same measure that you mete shall it be meted out to
you. You shall never see your native town again.’ Mr C. §.
Palmer in a note on this passage has pointed out that ris is
here used for the second personal pronoun, though he has failed
to interpret the general sense’. This use is too well known to
require much illustration. I may however refer to one place,

1 I see that Professor L. Campbell probably taking a hint from Mr Palmer,

in the second edition of his Sophocles _ interprets the passage as above,
which has only just come into my hands,

> 73 ee ee ep 4, ee rr ee ae > ( s I -_
peg Rp vr st Ce my . a a r 4 =

OBSERVATIONS ON THE OEDIPUS COLONEUS. 89

at present corrupt, in which it should be restored. We should
read in Aristophanes Lysistr. 657
el O€ AUTTNGELS TL [E,
TOedE TaYNKT@ TaTa—w Tw (for TH) KoOdprv@ THY yvabov.

‘Tl strike somebody on the face with this untanned buskin.’
The ictus falls on t@ as in Ran. 708.

Again towards the end of the very same speech vv. 1389,
1390 we have the very curious expression

Kal KaX@ TO Taptapou

oTvyvoyv Tatp@ov épeBos Os © aTroLKion.
Here azovxion ‘settle you far from your fatherland’ is to be
understood in the same reference. ‘You have driven me from my
home, but I have found refuge and an abiding dwelling-place
ie on the friendly soil of Athens; your brother has driven you too
from home, but your new resting-place shall be the abhorred
womb of death.’

Finally I think we may find the same keynote struck in the

13 very beginning of the play vv. 25—27.

-

= 8
|

my 2)
>

-

*e

ef.

AN. adn bots 6 TéTOS 7 MAaOw porotca Got;

OI. val, réxvov, eimep oti y éEotenopos.

AN. adn éotl pry oikntos.
The antithesis of é£o0vxnya.pos)(o¢enros which is thrown into
the strongest relief by their close juxtaposition, will not permit
us to suppose that their sense is the same. And the idea of Din-
dorf that é& in é£ovuxnovpos means ‘completely’ is opposed to the
ordinary sense of the word and gives a meaning which can be

- only described as inane. It only remains then to take é£ouxnat-

fos in accordance with the common usage of é£oixéw ‘to live out
of a place or away from it,’ i.e. ‘to emigrate.” Oedipus, thinking
of his weary wanderings and the divine promise that they
should end at last, asks ‘Is this a place where exiles can find a ©
habitation?’ to which Antigone, perhaps misunderstanding the
question, returns ‘Nay, it is inhabited already.’ These two
short speeches, properly interpreted, are seen to cohere with v. 39
where the stranger corrects the idea which underlies the action
of the blind man and his guide, by saying the place is Otros
ovo oixyntos. And perhaps it is not too subtle to suppose that

90 THE JOURNAL OF PHILOLOGY.

this peculiar turn of phrase was intended by Sophocles to be a
foreshadowing and indication of the drama’s main motive, which
otherwise breaks unexpectedly and abruptly upon us in v. 45
with the flat declaration of Oedipus that he will not leave the
holy ground. It will now be obvious what change I propose to

make in v. 384. For cato:xtiodow I would read ckatotKktovoty, —

so that the sense will be ‘I cannot tell where the gods will settle
thee the toiling one” The use of zo of course does not require
support; but one example from this very play is'so apt that I
cannot refrain from quoting it v. 23 éyess dudacKxew On pw Orroe
Kabéorapev;

I take this opportunity of adding two suggestions on other
lines of the play; and one on a fragment.

v. 30 Seipo mpocotelyovta KaEoppmopmevor;

This is explained, I believe, by all the commentators as a
hysteron proteron ‘going and starting,’ a figure here ludicrously
out of place. é£opuepevoy means ‘hastening’ as in Trach. 929
év © TO Keice Sedpo T éEopuwpcOa ‘while we hurry hither and
thither.’ Oedipus with the querulousness of a blind man wants
to know, not only if the stranger is coming towards them, but if
he is making good haste. This is on a piece with his words in
v. 21 and Antigone’s gentle protest, and with the pettishness of
v. 29.

v. 153 GAN ov pav ey y enol

mpoacOnaets Tacd apas.

It is singular the editors can have put up with this so long.
- The middle is imperatively required both by the rules of Greek
writing and Sophocles’ own usage. Compare Antig. 40 zpoc-
Gécbat rréov, O. T. 1460 rpocA7 pépiuvav, and in this very
play v. 404 wpocGécOar ce. O. T. 820, which Professor Campbell
quotes, is decisively against him xal tad’ otis aAXos jv 7H 'yo
“T €“avT@ Tad apas 6 tpoaTieis, as the ém’ éwaut@ makes
all the difference’. Read therefore rpoc@yoeu ‘take to thyself,’
which was changed through being mistaken for the active.

1 In his last edition ProfessorCamp- that it is not easy to see how the ac-
bell supplies 7@ 7ueréoy Syuw. By tion of a foreigner like Oedipus could

ellipses of this kind anything can be bring dpds upon the deme of Colonus.
exnlained. But it must be observed

,

Ae og a ~~
as

MF ow Ah ee he kD) A ad wy FTe a oo tree are.
a s ‘ ms
A Pw PA? * - ‘

OBSERVATIONS ON THE OKDIPUS COLONEUS. 9g!

Fragm. 319. amnke réudiéw ov méras dopov.

This is a passage from the lost play Koryo, quoted by Galen
9. 385 (5.454) in a philological discussion of the meaning to be
assigned to qweuduywdns in a medical dictum of Hippocrates.

This interesting word, which Curtius Gr. Etym. p. 708 con-
nects with ¢ducaw, seems to have meant originally either (A)
‘something blowing’ or (B) ‘something blown out.’ (A) gives
the meaning mvoy ‘blast, which is assigned to it here by Galen.
Under (B) we place the meanings ‘blister, ‘bubble,’ ‘drop’ of
rain or blood, ‘flash of light, (so called from its evanescence);
and then by a very curious metaphor 7éudié 7Alov seems to
mean the ‘sun-bubble, (unless indeed it is taken here also to
be a ‘flash’).

The meaning assigned to wéuduf here by Galen is avon
which agrees with the original meaning of the word from root
gu ‘blow’ Curt. Gr. Etym. /.c. A comparison of the other two
passages quoted by Galen for this sense will shew that he uses

avon with some latitude. In the fragment of the Salmoneus it

refers to the blackening rush of the lightning, in the Prometheus
Bound (rather Unbound) to the dark sweep of the storm. To
either of these it may refer in this passage, or possibly to the
scorching breath of the bulls of Colchis (as the play is the
Kodyor) from whose nostrils issued fire and smoke. Hermann
has suggested ws imvod ceXacdopov. cedacopov is very tempt-
ing, and, with the slight change of v to v, I propose to accept it.
I propose also to keep ov and to find the lost substantive which
agreed with ceXacddpoyv in méudiéiv, which I take to be for
meupré[vE|uv. vs is a good word, used by Eurip. Troad. 396
for ‘coming,’ and by Hippocrates for ‘movement’ (dopa) and ‘a
straight course’ (ev@vwpia). So that the sense is ‘the réudué
sprang away on its dark path.’

J. P. POSTGATE.

OLD GERMAN GLOSSES. FROM A BODLEIAN
MANUSCRIPT.

Tue Manuscript which contains the following glosses is in
the Bodleian Library at Oxford, marked Auct. F. 1. 16, and
was written not later than the early part of the tenth
century.

It contains the text of the Georgics of Virgil from 2. 120,

Servius’s Commentary on the Eclogues and Georgics, the text

of the Aeneid, and Servius on the Aeneid. But between ~

the commentary on the Georgics and the text of the Aeneid
_ occur several pages of excerpts from Isidorus and other authors
and the first sixty-four of the following glosses, the whole
being arranged so as to form a brief commentary on all Virgil.
This is followed by a heading Incipiunt uaria glosemata and
the glosses 65 to 121. The rest of the glosses are either
marginal or interlinear, and appear to be written by two

different hands within a century of the date of the manuscript —

itself. It may be added that the glosses 1 to 121 were not
first. written by the scribe but copied from a book before
him. .

The history of the MS. before the seventeenth century is
unknown. It was one of three lent by Bernard Rottendorph,
a physician of Miinster, to Nicholas Heinsius, who used it

for his editions of Virgil, giving it the name Rottendorphianus

tertius, but forgot to return it to its owner. In 1672, Francis
Junius, author of the Etymologicum Anglicanum, then at
the age of 83, saw and copied the more important of the
textual glosses in Heinsius’s house at Copenhagen. In 1678,
Junius transcript came into the Bodleian, and in 1697

OLD GERMAN G@LOSSES. 93

‘the original MS.: but the connexion between them was
unknown, and when a selection of the following glosses was
printed in Nyerup’s Symbole ad Literaturam Teutonicam.
(Haunie, 1787), it was from transcripts of Junius’s tran-
scripts that they were taken; and all the references to them
in Graff’s Althochdeutscher Sprachschatz are from this source.
At last, in 1877, the real history and significance of the
- long-lost MS. was discovered, and the whole of the glosses have
been carefully transcribed for the present paper, an answer
being thus supplied to the anxious query in Haupt’s Zeitschr.
fiir Deutsches Alterthum, Vol. xv. p. 103 (1872). This list
will be reprinted in the second volume of Steinmayer and
Sievers’s Althochdeutsche Glossen, of which the first volume
was published in 1877.

No. Gloss on

1 Eel. 1. 56 | Hibleis - herba est quam nos dicimus
aduch.

2 59 | Palumbes - columbe sunt - quas dici-
mus menistuba.

3 2. 36 | Cicuta - herba est quam nos dicimus

scherning.

4 50 | Calta - cle.

5 3. 20 | Carecta multitudo herbarum. In palus-
tribus - quas dicimus semithan.

6 5, 39 | Carduus - thistilcarda.

7 » | Paliurus - hagan.

8 7. 32 | Coturno - calciamento uenatricio quod
alii dicunt periscelidas - aut hoson.

9 50 | Fuliginem - quod nos dicimus rwot.

10 8. 74 | Licia . id est quod dicimus harluf . cum
quo ligant mulieres,

11 | Georg. 1. 75 | Vicie - Vuicchun.

12 94 | Rastrum - recho.

13 95 | Crates - egida.

14 139 | Visco - Audis [1 bulis].

15 144 | Cuneus - wuecke,

16 153 | Lappe - cledthe.

Q4 THE JOURNAL OF PHILOLOGY.
No, Gloss on

17 | Georg. 1. 162 | graue robur - id est grendil.

18 164 | Tribula - flegii.

19 » | Lrahe - egida.

20 165 | [V]irgea preterea . id est gart.

21, 22 166 | Crates - hurth - aut egida.

23 172 | Bine aures - que riestra dicimus.

24 173 | Tilia . linda.

25. 264 | Vallos - sunt quos dicimus phale.

26 | Georg. 2. 189 | Filix - farn.

27 374 | Vri - id est animal quod dictum est
urront,

28 389 | Oscilla - scocga.

29 | Georg. 3. 147 | Asilo - bremo.

30 338 | Achalantida id est auis - nathagala.

31 366 | stiria id est 2hilla.

32 543 | Phoce id est animal marinum - quod nos
dicimus elah,

33.| Georg. 4. 63 | Melisphilla - herba quam dicimus bini-
wurt.

34 131 | Papauer - herba quam dicimus maho.

35 271 | Amello - herba - golthblomo.

36 307 | Tigna .- latta.

37 Aen. 1, 123 | Rimis - nwoe - in quibus tabule in
unum coniunguntur,

38 169 | Vnco morsu - quem nos dicimus chram-
pho.

39 323 | Lincis - id est Jos apud nos animal quod
dicimus.

40 435 | Fucos - drenon quod nos dicimus.

41 698 | Sponda - lectum siue beddipret.

42 Aen, 2, 135 | In ulua- hoc ~[in marg. is added “~ est”}
in palustribus locis ubi crescit iuncus
ac papyrus - et quod nos dicimus
suuerdollon.

43 Aen, 3. 428 | Delfinum - mirisuwin.

44 453 | Dispendia - ungifuori.

45, 46 Aen. 4. 131 | Lato uenabula ferro . id est staph - in
se habentem latam hastam quam nos
dicimus ewurspioz.

47 Aen. 5. 177 | Clauum - quod nos dicimus . heléa . in
summitate est.

48 208 | Trudes - furka.

49 Aen. 6. 13 | in tribus locis ubi tres uie in unum con-
ueniunt - que nos dicimus giuwiege.

50 205 | Viscum .« id est budis. :

si

OLD GERMAN GLOSSES. 95

Gloss on
Aen. 6. 209 | Bratiliitea . blech.
Aen. 7. 48 | Picus « auis - speth.
378 | Turbo . in modum factus globi rotun-
dus - quem dicimus doch - buxum .
inde erit factus turbo.
_ 390 | Thirsus - stilherbe.
3) 417 | Rugis - hoc dicimus nos rumphusla.
aS 627 | Aruina mittigarne.
Aen. 8. 278 | Sciphus - parua staupa.
‘s Aen, 9, 170 | Pontis - scalis - aut quod rustici dicunt
. clida.
a 59 476 | Radii - rauua.
Se, 60 | Aen. 11. 64 | Crates - chida.
61 862 | Papilla . summitas mamme id est uwwarte.
62 | Aen. 12. 120 | Verbena - herba quam dicimus hanaf.
:) 63 ' 413 | Caulem comantem id est stipitem cum
foliis - quam dicimus sé.
{ 64 470 | A temone - hoc est in anteriori parte
+ plaustri ubi boues ligantur - apud nos
; | thessalia.
| “FINIVNT GLOSE.” ‘INCIPIVNT VARIA
GLOSEMATA.”
65 | Callum caro et cutis indurata quod nos
dicimus - swwzl.
66 Flocci sunt quos nos in uestimentis thiu-
disce wwuloo dicimus.
if 67 Culcites - bedd.
1 68 Culcitum id est plumatium . beddiuurdi.
a 69 Cauteriola - canteri.
70 Toregma - scaperede.
71 Tornarius - thresisa - [sic, manu ut ui-
detur prima].
72 Maialis - barug.
73 Murica - snegil.
74 Muscus - grimo.
75 Migale - harmo . Allec - alerencia - [?]
76 Gobio - grimpo.
at Esox - lahs.
78 Lucius - hacth.
79 Capito - alund.
80 Timallus - asco.
81 Sardinia - hering.
— 82 Axedones id est humeruli - lienisas.

THE JOURNAL OF PHILOLOGY.

No. Gloss on '
83 Scorellus - amer. Pep J
84 Terebra et teretrum - nawuger. 4
85 Crabro - hornut.
86 Ancipula - fugulclouo. , aie
87 Andela - brandereda. q

88, 89 Arula - fiurpanne uel herd. age

90 Apiastrum - biniwurt. ied

91, -92 Aadsculus - boke - uel ec, [A in A+= Al. 4
mutatum est. | F,

93 Aestuaria - jlod - uel bitalassum - ubi

duo maria conueniunt. -

94 Acinum - hindbirt.

95 Atramentarium - blachorn.

96 Atramentum - blac, ae

97 Fasciola - winning - [? iunning] 77

98 Verriculum . besmo,

99 Villosa + ruge. : e
100 Villa - Lininhruge. : ;
101 Vadimoniam - borg. b
102 Bacinia - beri.
103 Botholicula - stoppo.

104 Bracium - male. 7 ,
105 Bracinarium - bro“hus - [sic, manu ut
uidetur prima]. |
106 Bouellium - fuled.
107 Bradigabo - feldhoppo. :
108 Balista - stafslengrie.
109 Brancia - kian.
110 Burdo - uwrenio. ‘
111 Cincindila - uwocco. |
112 Cratus - bollo - [prima manu “gratus”,
rasura mutatum ].
113 Cerasius - kirsicbom.
114 Cerasium - biz.
115 Clauatum - giburdid.
116 Arnoglossa - wuigbrede.
117 Plebeios psalmos id est seculares psalmos |
id est wwinilieth.
118 Reditus - hembrung.
119 Petulans - wurenise.
120 Pastellus - hunegapl.
121 Pustula - angseta.

OLD GERMAN GLOSSES, — 97

(N.B. Glosses 122—end are interlinear or marginal: those marked with an
asterisk seem to be older than the rest.)

No. Gioss on
arbores wuilnan
122a@ | Georg. 2. 257 | picee tantum [sic, quoad locum: “uuil-
nan” perobscurum est].
ichas
122d taxique nocentes [zic, quoad locum].
123 365 | acie : wuthta.
*124 | Georg. 3. 25 | aulaea : wmbihang.
e luue dilectus
125 72 | dilectus [sic, quoad locum].
*126 82 | Huic lineae e regione sunt extrema
margine hach
_ ¥127 imu| [i dubium est].
*128 Uwe
129 173 | temo : thistle.
130 | f Aureus - wuahsbl ——— |[uidetur “anc.”
sequi|.
131 Gilbus badius - falu -
132 Spadix - dun -
133 Glaucus - glasa [forsitan una litera
adiecta est].
134 Cadius - dlas — [forsitan nil nisi “.”
| adiectum est].
135 180—192 | | Petilus . jitilu —— [forsitan “ fitiluiz”
[? asterisco sig- | J uel “ fitiluoz”
136 | nanda omnia] Scutulatus - appulgre.
137 Guttatus - sprutodi [? sprurodi].
138 Mannus - fiarscutig.
139, 140 Mirteus - dosam uel wuirebrun.
141 Maurus - alsuart.
142 Iumenta - mergeh.
143 | Toctonarii - thrauandi [sic: ? Totto-
narii].
144 _Tottolarii - Telderias [4 Toctolarii].
145 310 | mammis : geclerun [? geelerun, geelcrun].
146 385 | Lappae : chiue.
147 | Georg. 4. 38 | tenuia : thunnd.
148 41 | uisco : mistile,
149 141 | tiliae : lindian.
150 168 | fucos : drenan.
151 243 | Stellio : mol.
152 244 | fucus : brana.
153 245 | crabro : hornut.
*154 395 | phocas : mirikot.
155 506 | cymba : nauis parua alii cuba [? Latinum

Journal of Philology.

uerbum |].

VOL. xX. 7

THE JOURNAL OF PHILOLOGY.

’ Gloss on Servius

No. as “
156 | Ecl, 6. 78 | upupam : wutduhoppe, | ¥e
157. : » | hirundinem : sualan. ae
158 Ecl. 8. 74 | stamen : wuarp. | eS
159 5 | Licium : hewild. : a
160 | Georg. 1. 75 | lupini : jicbane. oe 3
‘161 139 | uisco : mistle,

162 178 | glarea : id est arena - grat. a

4 ve Georg. 2. 389 furcille : gaflie uel furke, [Haec uerba
etlam in marg. reperluntur, manu an-
| tiqua. | ;

165 | Georg. 3. 82 | uicinum : wuasblane [contextus est:— =
“album quod pallori constat esse w-
cinum” |.

Gloss on zs

*166 Aen. 1. 323 | lyncis : losses.

*167 337 | suras : wuathan.

1684 427 | portus : cathoma.

*168 b 435 | fucos : waspe.
169 711 | Pallam : hroe.
170 Aen, 2. 16 | abiete : dennium [? dennuin].
171 55 | foedare : gihonen. 4
mapuldreum acernis ; ‘
172, 173 112 | acernis . mapulder. sic, quoad locum}. ‘
174 147 | amicis : friundlicun.
175 229 | merentem : wuirthiganen.
176a_ 441 | testudine : id est densitate armorum id
| est schildwueri.
1766 492 | ariete : murlraca [? murltaca].
taf Aen. 3. 16 | socii : dswese, :
178,179 217, 218 Jorths effasio [sic quoad locum : sed for- :
i. gesseod est _—tasse uerius “ gesscod”|
180, 181 282 | euasisse : owerrunnen habbien. ae
182 286 | Aere cauo clipeum’: quia ex aere factum
erat - chuculan. [*chuculan” quoad —
locum super “clipeum” est: et est
uelut si ‘“‘cauculan” uel “chuculan” uel
“ehuculan” scriptum esset; corrector,
ipse fortasse scriba, certe ‘ huculan”
scripsit, nisi mero casu paene euanuit
a lias sh | ee
183 545 | antennarum : segelgerd.

‘ >

OLD GERMAN GLOSSES. 99

No. Gloss on

184 561 | rudentem : vel rudente - circulo guber-
naculi - id est - st¢rwwith.
185 649 | corna : curnilbom.
186 671 | aequare : igrundian.
introitum <imitthi
187 688 | ostia saxo saxi [sic, quoad locum: ¢ imu-
thi].
188 Aen, 4. 19 eae : odiosum athrotan.
undar intermissa pinne
189, 190 88 | opera interrupta minaeque nwmana [sic,
quoad locum : t= se ano
191 131 | uenabula : lancee ewrspiat.
192 139 | fibula : fibula spenule.
193, 194 152 | caprae : Capra - reho - nam crapra get
, dicitur.
aether

195 167, 168 terre signum - id est erthbigunga

nimphae [sic, quoad locum].

196 239 | talaria : seridfoos [? serid-, sorid-foos].

*197 245 | tranat : vulotad.

198 250 | mento : chinne.
199 490 | ciet : utthalad.
200 534 | procos : appetitores druhtingas.

*201 | . Aen. 5, 128 | mergis : dukiras.

*202 205 | murice : dwuansten. |

*203 230 | pacisci? teneant? : —rihingian | Vna erat

adiecta litera, “r” dubium est].
204 269 | taeniis : tena - nesta.
205 306 | leuato lucida : gifuriwidemo.
206 332 | titubata : calcata wwankonda.
207 546 | impubis : wnbarldharht [% ex “unbarh-

harht” correctum |.
208 566 | primi : wuassitiluot [? uuasfitiluot].
209 578 | Lustrauere : wmbiridun.
210 630 | hospes : wwerd |e” fortasse dubium est].
211 710 | fortuna : missibure.
212 714 | Pertaesum : odiosum sit athrotan.
213 719 | incensus : giscund.
auernum sine uerno

214 732 | auerna per alta auunni [sic, quoad locum].
215 735 | Elysium : sunnanueld.

*216 745 | acerra : cerra - uas turis - arcula tu-
raria - id est rocfat [In “rocfat” “t”
fortasse dubium est].

217 758 | forum : mahal.
218 811 | periurae : forsworenero.

7—2

TOO THE JOURNAL OF PHILOLOGY.
No. Gloss on
*919 852 | adfixus : tohlinandt.
220 | Aen. 6, 180 | piceae : fiuchtie omnes arbores unde pix
uenit {? fuichtie]. —
221 181 | Fraxineae : eschine.
222 205 | uiscum : mstil.
223 214 | robore : rinda.
224 420 | offam : muhful.
#295 offam : dewwin [? cleuuin, deuiun].
226 682 | recensebat : talde.
227, 228 Aen, 7, 109 | adorea liba : bradine diski.
229 319 | pronuba : makerin.
230 506 | torre : brande.
*231 590 | alga : rietgras, |
*232 626 | lucida tergunt - wegadun [sic, quoad lo-
cum].
233 627 Ane : midgarnt.
234 628 | Signa : gutfanan.
235 690 | pero : striorling [1 streorling]..
*236 796 | picti : pictus wehe.
237 Aen. 8, 178 | acerno : mapuldrin.
*238 276 | ? populus : halebirie,
239 Aen, 9, 87 | picea : picea wurie.
240 134 | iactant : hromiat.
241 222 | statione : wuardu.
242 471 | mouebant id est uidebant - scuddun.
*243 505 | testudine : testudo sceldwwara,
244 537 | tabulas : scindulan.
245 608 | rastris : egithon.
*246 616 | manicas : ermberg.
247 629 | petat : stichit [“‘s” fortasse dubium est],
248 701 | pulmone : lungandian.
#249 705 | falarica uenit : stephstrengiere.
250 723 | fortuna : missiburt.
251 724 | conuerso : togidanemo.
252 | Aen, 10, 58 | Dum : aunt,
253 337 | thoraca : brunge.
254 381 | uelit (in “uellit” correctum) : a terra
losda.
255 382 | costis : rabbun.
256 390 | gemini : itwisan [? “itiusan,” uel = “i.
tuisan,” sc. “id est tuisan”],
257 a 444 | cesserunt : rwmdun.
257 6 538 | uitta : wxanding [Vix dubium est quin
scriba “ uuunding” indicet].
258 542

gradine ; quasi gradatim id est stzlo,

OLD GERMAN GLOSSES. Ss

ont a
= es
(ee ee) Be

No. Gloss on
259 649 | pactos : gimahlida.
260 681 | dedecus : turpitudinis honithia.
impellat stachi
261 682 | exigat ensem [sic, quoad locum].
erexit struuide
262 711 | inhorruit armos [sic, quoad locum :
? “ striuude’’ |.
263 735 | Contulit : angenbrahte.
264 736 | abiectum : nithergiwuorpenen.
265 744 | Viderit : gisehe - et hoc uerbum ironia est.
266 795 | Cedebat : retrahebat thanan/for.
267 818 | neuerat : brordade.
268 891 | Bellatoris equi : wuthherses,
269 892 | calcibus : howuwn.
270 893 | effusum : nithergiwuorpenen.
271 901 | nefas : honithia,
272 | Aen. 11. 73 | laeta : wwillich.
273 149 | reposto : nithergisettemo.
274 320 | plaga : wuald.
275 500 | Desiluit : wmbette.
276 524 | quo : thar.
277 562 | sonuere : hullun.
278 579 | fundam : slengiran.
279 589 | omine : hele. :
280 599 | fremit : thrasida.
281 607 | ardescit : gerode.
282 616 | tormento : ...... torqueo + slingirun.
283 663 | lunatis agmina peltis : in modum lune
Te factus - sinwuuellun. ;
284 671 | Suffuso reuolutus : nitheriwallenemu +
| [? “-eruiall-” ; uulgo “ Suffosso”].
285 688 | Verba : hrom.
286 711 | interrita : wngimelademu.
287 777 | Pictus acu : gibrordade,
288 874 | laxos : wnspannane.
289 890 | Arietat : sétet.
290 | Aen. 12. 7 | toros : toros.- crocon.
291 91 | candentem : gloianden.
292 163 | radii : gerdiun [? gerduin].
293 171 | admouit : adiunxit - todeda.
294 174 | notant : steppodun.
295 215 | lancibus : uasis - baewuwegun |e” prius

dubium est, “x” uix dubium : est
uelut si quis scripsisset “ baxuuegun”
et in “baec-” uel “ baex-uuegun” mu-
tasset|.

¥O2 THE JOURNAL OF PHILOLOGY.
No. © Gloss on
296 234 | deuocet [in “ deuouet” correctum|] : bifal.
297 274 | fibula : hringa.
298 * 300 | Occupat : slog.
299, 300 305 | prima : in furistemo.
301 357 | extorquet : utauuende.. .
302 364 | sternacis : id est sternentis spwrnandies.
303. 404 | Sollicitat : wuegida.
304 412 | Dictamnum : umtewurt.
*305 413 | caulem : stok.
306 419 | panaceam : herbam rauuano [? “ riniuo-
| no,” ‘reniuano”].
307 470 | temone : thisle.
308 520 | conducta : ingumedodera.
309 590 | Discurrunt : tiwarad.
310. 646 | miserum : wnothi.
rumdun
31t 696 | spatiumque dedere _[sic, quoad locum].
312 727 | uergat : nitheruwaga.
313 775 | sequi : skietan.
314 857 | parthus : ungar [sc. Hungarius ?]
Gloss on Servius
on
315 Aen. 2. 229 | Expendisse : id est soluisse - wngebdan.
aut : BB clunis : isben uel ars belli uel posterior
54 a eR
318 pars omnis animalis,
319 Aen. 4. 548 | Vrbanus : alter liber dicit urbane -
Sronisco.
320 Aen. 5. 269 | taenis : nestilun.
321 Aen. 6. 704 | Virgulta : sumerladan,
DvBi0vs.
Gloss on
uidere theathe &
322 | Georg. 3. 25 | scena «.* ut uersis [in marg, “ quemad-
modum”].
323 | Aen. 5. 337 | Euryalus : fanfullistia.
324 | Aen. 10. 23 | Quin intra : netian.

OLD GERMAN GLOSSES. 103

INDEX TO THE TEUTONIC GLOSSES.

N.B.—Numerals printed thus 234 are in the text, 234 and 234 are interlinear
(or marginal)—in the older hand, in the more recent hand.

A
aduch 1

- aec 91

alsuart 141
alund 79

amer 83
angenbrahte 263
angseta 121

-apl 120, cf. sq.
appulgre 136, cf. foreg.
ars 317

asco 80

athrotan 188 212
-auuende 301

B

baexuuegun (? baecuuegun) 295
-bane 160

-barldharht (1 barhharht) 207
barug 72

? baxuuegun 295

bedd 67, cf. sqq.

beddipret 41, cf. foreg.
beddiuuidi 68, cf. bedd

belli 318

-berg 246

-beri 102, cf. biri

besmo 98

-bette 275

bifal 296

-bigunga 195

-bihang 124

biniuurt 33 90

biri 114 (-biri 94), cf. beri, birie
-biridun 209

-birie 238, cf. biri

blac 96, cf. sq.

blachorn 95, cf. foreg.
-bl[anc] 130: -blane 165
blas- 134

blech 51

-blomo 35

boke 91

bollo 112

-bom 113 185

borg 101

bradine 227

-brahte 263

brana 152

brande 230, cf. sq.
branderede 87, cf. foreg.
-brede 116

bremo 29

brothus 105

brordade 267 (-brordade 287)
-brun 140

-brung 118

brunge 253

bulis 50, ? bulis 14, cf. hulis
-burdid 115

104. THE JOURNAL OF PHILOLOGY.

C erthbigunga 195
canteri 69 eschine 221
-carda 6 euurspiat 191, cf. sq.
cathoma 168 a euurspioz 46, cf. foreg
? cauculan 132 :
chinne 198 F. Cf. ph
chrampho 38 -fal 296
chuculan 182 faled 106
cle 4 falu 131
cledthe 16 -fanan 234
? eleuuin 225 fanfullistia (1) 323
clida 58 60 farn 26
cliue 146 -fat 216
-clouo 86 feldhoppo 107
crocon 290 fiarscutig 138
euba 155 ficbane 160
curnilbom 185 fitilu- 135

fiuchtie 220

D fiurpanne 88
deenniun (? deennuin) 170 flegil 18 : k
-danemo 251 flod 93 5 .
-deda 293, -foos 196 a
deuuin (? deuiun) 225 -for 266
diski 228 forsuorenero 218 ih
doch 53 forths 178 |
-dodera 308 friundlicun 174 7
dosan 139 fronisco 319 .

drenan 150, cf. sq.

— drenon 40, cf. foreg,
druhtingas 200
dukiras 201

dun 132

duuansten 202

E
ec 91
egida 13 19 22, cf. sq.
egithon 245, cf. foreg.
? ehuculan 182

elah 32
ermberg 246

fugulclouo 86 : 4
? fuichtie 220 ~
-ful 224
-fuori 44
furistemo 300
-furiuidemo. 205 ; q
furka 48, cf. sq.
furke 164, cf. foreg.

oo Ta ia Bel

G

gaflie 163
-garne 56, cf. sq.
-garni 233, cf.. foreg.

ao ee

ceil
al ee
ne a

OLD GERMAN GLOSSES. 105

gart 20

-gebdan 315

geclerun 145

? geelcrun, ? geelerun 145
_ -gerd 183

gerdiun (? gerduin) 292
gerode 281

gesseod (? gesscod) 179
get 194

gibrordade 287
giburdid 115
-gidanemo 251

-gifuori 44
gifuriuidemo 205
gihonen 171, cf. honithia
gimahlida 259
gimedodera 308
-gimelademu 286
giscund 213

gisehe 265

-gisettemo 273
giuuicge 49
-giuuorpenen 264 270
glasa 133.

gloianden 291
golthblomo 35

-gras 231

grat 162

-gre 136

grendil 17

grimo 74

grimpo 76

-grundian 186

“gunga 195

gutfanan 234

habbien 181

hach- 126
hacth 78

hagan 7

-halad 199 °
halebirie 238
hanaf 62

-hang 124

-harht 207

harluf 10

harmo 75

hele 279

helta 47

hembrung 118
-herbe 54

herd 89

hering 81

-herses 268

heuild 159

hindbiri 94
-hlinandi 219
-honen 171, cf. sq.
honithia 260 271, cf. foreg.
-hoppe 156, cf. sq.
-hoppo 107, cf. foreg.
-horn 95

hornut 85 153
hoson 8

houun 269

hringa 297

hroc 169

hrom 285, cf. sq.
hromiat 240, cf. foreg.
-hruge 100, cf. ruge.
? huculan 182

hulis 14, cf. bulis
hullun 277
hunegapl 120
hurth 21

-hus 105

jaunt 252

106 THE JOURNAL OF PHILOLOGY.

ichas 122 b makerin 229

igrundian 186 malt 104

-ihalad 199 mapulder 173, cf. sqq.

ihilla 31 mapuldreum 172, cf. foreg.
imitthi (? imuthi) 187 mapuldrin 237, cf. mapulder.
in 299 - -melademu 286

ingimedodera 308
inu- 127

isben 316

isuese 177
ituisan 256
-iuallenemu 284
? iunning 97

K
kian 109
_kirsicbom 113
-koi 154

-konda 206,

-ladan 321

lahs 77

latta 36

-lieth 117

linda 24, cf. sq.
lindian 149, cf. foreg.
lininhruge 100
los 39, cf. losses
losda 254

losses 166, cf. los
-luf 10
lungandian 248
lunisas 82

luue 125

mahal 217
-mahlida 259
maho 34

-medodera 308

menistuba 2

mergeh 142

midgarni 233, cf. mittigarne
mirikoi 154

mirisuuin 43

missiburi 211 250

mistil 222, cf. sq.

mistile 148 161, cf. foreg.
mittigarne 56, cf. midgarni.
mol 151 |
muhful 224

murlraca (? murltaca) 176 b

N

nathagala 30

nauuger 84:

nestila 204, cf. sq.

nestilun 320, cf. foreg.

netian (?) 324

nithergisettemo 273

nithergiuuorpenen 264 270

nitheriuallenemu (? nitheruialle-—
nemu) 284

nitheruuaga 311

numana (? undar-) 190

nuoe 37

O
-othi 310
ouerrunnen 180

P
-panne 88

ew >

OLD GERMAN GLOSSES. 107

phali 25
-pret 41

4 -raca 176 b

rauua 59

rauuano 306

recho 12

-rede 70 87

reho 193

? reniuano 306.

ribbun 255

-ridun 209

riestra 23

rietgras 23}

-rihingian (amissa litera prima)
203

rinda 223

? riniuano 306

rocfat 216

-rode 281

ruge 99, cf. hruge

rumdun 257 a, 311

rumphusla 55

-runnen 180

ruot 9

S
scaperede 70.
scelduuara 243, ? cf. schilduueri
scherning 3.
schilduueri 176 a, ? cf. scelduuara
scindulan 244
scocga 28
scridfoos 196,
scuddun 242
-scund 213
-scutig 138
segelgerd 183
-sehe 265

semithai 5

? seridfoos 196

-settemo 273

sinuuellun 283

skietan 313

slengiran 278, cf. sqq.

-slengrie 108, cf. foreg.

slingirun 282, cf. slengiran

slog 298

snegil 73

? soridfoos 196

-Spannane 288

spenule 192

speth 52

“Splat 191, cf. sq.

-spioz 46, cf. foreg.

sprutodi (? sprurodi) 137

spurnandies 302

stachi 261, cf. stichit

stafslengrie 108, cf. staph, slen-
giran, stephstrengiere

staph 45, ef. foreg.

staupa 57

-sten 202

stephstrengiere 249, cf, staf-
slengrie

steppodun 294

stichit 247, cf. stachi

stieruuith 184

stiet 289

stil 63, cf. sq.

stilherbe 54, cf. foreg.

stillo 258

stok 305

stoppo 103

? streorling 235

striorling 235

struuide 262

sualan 157

-suart 141

108 = HE JOURNAL OF PHILOLOGY.

sumerladan 321

sunnanueld 215

-suorenero 218

suuerdollon 42
‘suuil 65

suuin 43

? -taca 176 b

talde 226

telderias 144
thananfor 266
thar 276

theathe (7) 322
thessalia 64, cf. sq.
thisle 307, cf. foreg.
. thistilcarda 6
thrasida 280
thrauandi 143
thre‘lsa 71

thunni 147
tiuarad 309
todeda 293
togidanemo 251
tohlinandi 219
-tuba 2

? tuisan 256

Vu. Cfh uu

-uallenemu 284
-uarad 309

uaspe 168 b
uegadun 232
uehe 236

-ueld 215

? -uiallenemu 284
uinning 97
umbette 275
umbihang 124

umbiridun 209
unbarlharht (? unbarhharht) 207. —
undar (? -numana) 189
ungar 314

ungebdan 315
ungifuori 44 ~
ungimelademu 286
unothi 310
unspannane 288
urrint 27

utauuende 301
utihalad 199

uurenio 110, cf. sq.
uurenise 119, cf. foreg.

uurie 239
-uurt 33 90 304
Ve

vulotad 197

VV. OL a |
uua- 128 gs |
uuaga 311 =
uuahsbl- 130, cf, uuasblanc
uuald 274

uuankonda 206 .
uuardu 241 ;
uuarp 158
uuara 243 Ki
uuarte 61 - ee
uuasblane 165, cf. uuahsbl-
uuassitiluot (? uuasfitiluot) 208
uuathan 167

uuecke 15

uuegida 303

-uuegun 295

-uuellun 283 |
-uuende 301

- uuocco 111 |
-uuorpenen 264 270 rf | x
uuuloo 660 een ee
¢ uuunding 257 b | | ee

a a. ‘i re uuunni 214 Mage 3 ;
es FF MADAN, Bie,

TRACES OF DIFFERENT DIALECTS IN THE
LANGUAGE OF HOMER.

THE article published by Mr Monro under this heading in
the last number of the Journal of Philology seems to require
a reply, more especially as the author is so impressed with the

untenability of my position as to regret that a work “so well —

adapted otherwise” for general readers as Prof. Mahaffy’s
_LHistory of Greek Interature, should have given currency to
what he regards as a mass of misstatements and erroneous

reasoning. I hope to show as briefly as possible that the

statements are not misstatements and that the reasoning is
not erroneous.

I must begin by thanking Mr Monro for the clerical errors
he has pointed out in the delinquent Appendix. Perhaps they
will be excused when I say that the whole system of reference
had to be changed while the Appendix was passing through
the press, and that owing to my absence from England I was
unable personally to superintend it. I must next draw atten-
tion to the fact that my primary purpose was not to determine
whether Homer was an actual individual or a mere abstraction,
whether he was the author of the whole of the Iliad and
Odyssey, and whether he lived in the twelfth or the fifth century
before our era, but to examine the age and character of the
Epic dialect as we now have it. All I was concerned with
showing was that the Homeric dialect is an artificial one, that
it bears traces of having passed through several phases of
existence, and that in its present form it is as late as the fifth
century B.C. Professor Mahaffy, however, was perfectly right
in assuming that I placed the date of “the first origin of the
Iliad and Odyssey as complete poems at or near the opening

—

y La |
4 Oe «
4 t fa ™ —
: «yt =
4 ™ : t a Pees =
sd Sian lee ae li pin aie a eS ee ol,
Pave ea eee bs ose ee na

p - '
=. ee ay

aL EH PRS

a a

Eee tha an da | an

DIALECTS IN THE LANGUAGE OF HOMER, 111

of the seventh century B.c.’ I certainly did so at the time
I wrote the chapter, considering that the new Ionic forms
found in Homer and Herodotus might be as old as that period,
and that the Attic colouring which, in common with Aristarchus,
Cobet and Paley, I find in Homer, was simply evidence that
the poems had undergone a process of manipulation in Attica.
Subsequent study and reflection, however, have brought me
more and more over to Prof. Paley’s view, and I find it
increasingly difficult to believe that the Homeric dialect in its
present form can claim a much greater antiquity than the
Periklean era. Many of the forms which are usually regarded
as archaic rather seem to me, to borrow a term from the art-
critics, archaistic. Of course this does not prove anything as
to the age of the original Iliad and Odyssey, or of the original
Homer, whoever he may have been; if Mr Monro likes, he may
still believe that Homer lived before the Dorian invasion of the
Peloponnesus.

I will now take Mr Monro’s objections and criticisms
seriatim, dealing with each as briefly as possible. I cannot
help remarking, however, that the general impression they
produce upon me is that of a system of apologetics which I
fancied had long since been discredited by critical science.

Mr Monro first objects to my use of the term “ period.”
It is, however, consecrated by custom, and I do not see what

other term I could have chosen to express my meaning. The

Ionic genitives in -ov presuppose older genitives in -oo, and
these again still older genitives in -o1o. If we follow Ahrens,
all three forms are found in Homer. I am surely, therefore,
justified in saying that Homer contains forms belonging to
three different periods in the history of the Ionic dialect. Mr
Monro says that “we cannot assume that all the forms which
are similarly intermediate between two others belong to the
same chronological period.” But I never assumed anything of
the kind ; I was dealing with philology, not with history. It
is sufficient to know that in the Homeric language we have
relics of three different phases of growth of the Ionic dialect ;
those belonging to the first and third phases cannot be older
than the earliest beginnings of Epic poetry or later than the

II2 THE JOURNAL OF PHILOLOGY.

final redaction of the Iliad and Odyssey. So far as I can see,
it matters little whether the relics of the second phase all
belong to exactly the same chronological period or not. They
must fall somewhere between the first and the second periods.

Mr Monro asks what was the middle Ionic form of vnos, veds? -

If an answer is necessary, we may say vyds itself, the older
form which may or may not be preserved in Homer being
ynFés. I do not understand the point of Mr Monro’s other
question: “If vyds is old Ionic, and consequently archaic, how
are we to explain the fact that it is very much commoner than
veos ?” since according to my view the choice of equivalent
words in later Epic poetry was determined partly by the
exigencies of the metre, partly by an affectation of archaism.
We come next to the question of the relation of the lan-
guage of Homer to that of Herodotus. Here as elsewhere, it
- must be understood, I have given but a few examples in sup-
port of my position out of the many which I have collected in
my note-books, and as I have taken care not to select the most
typical or convincing, but the first that came to hand, the
examples are necessarily of unequal strength. At the outset
Mr Monro seems to doubt whether he has not “strangely
misunderstood” me. He certainly has done so.. My point was
not to prove that the New-lonic parts of Homer are Hero-
dotean, but that the language of Herodotus and of certain parts
of the Homeric dialect belong to the same period in the history
of Ionic speech. It was not. necessary, therefore, to discuss
whether rifetot and the other words arraigned by Mr Monro
might possibly “date from the earliest periods of Ionic;” all
I had to show was that they were employed by Herodotus,
and were consequently in use in [onic literature during what
I have termed the New-Ionic period of the language. Their
antiquity must be tested by other evidence, and when so tested,
I venture to think, in spite of Mr Monro, fails to be substan-
tiated in the majority of cases’. Except in the case of Tieioe

1 Thus the Attic éouévy is an older dialects first separated: the omission
form than the Ionic eiuév, and must of the augment is distinctly the mark

therefore have been the form used in of a later time; @vAaxos and pudprupor

Old Ionic when the Attic and Ionic are the products of an analogy which

= CO a es —

DIALECTS IN THE LANGUAGE OF HOMER. 113

and its congeners, the Attic forms are older than the corres-
ponding ones found in Homer and Herodotus, and must ac-
cordingly have been the forms used by Old Ionic when the
Attic dialect branched off from it, while in some instances we
meet with forms due to an analogy which seems first setting in
during the age of Herodotus (as maybe inferred from the
‘small number of examples of it found in that author), or (as in
the case of the augment) with marks of phonetic decay which
are actually more numerous in the pages of Homer than in
those of the historian of Halikarnassos. We must not forget
that when the age of the Epic language is in question, we have
no right to assume that forms found for the first time in
Herodotus and the New Ionic inscriptions existed at a much

seems only just setting in during the
age of Herodotus; the etymologically
incorrect jicav is probably late in spite
_ of the Old Persian -disa; and as the
iteratives in -cxoy are not found in
Attic prose we may gather that they
are subsequent to the separation of
the Attic and Ionic dialects. In fact,
the iterative Preterites are confined to
the language of Homer (and his imita-
tors), Herodotus and the later Epic
writers. We know, therefore, that they
characterised the New Ionic; we do
not know that they existed in the
Ionic dialect in any earlier stage of its
career. Mr Monro is mistaken in saying
that xp7v is “probably not an instance
of lost augment;” xpn no doubt was
originally a substantive, but when an
imperfect was formed from it the
analogy of other augmented imperfects
was necessarily followed. That Hero-
dotus should omit the augment in a
_ ease of this kind is a strong proof that
the omission of the augment is a
mark of linguistic decay, characterising
the New Ionic period of the Ionic dia-
lect. It thus throws an important light
on the omission of the augment in
Homer, and I have accordingly referred

Journal of Philology. vou. x.

to it. The loss of the aspirate in
peradpevos and érdd\uwevos is New
Tonic; so therefore would its loss be
in the Epic aro,

In his article on Homer in the En-
cyclopedia Britannica Mr Monro sup-
ports his assertion that the Homeric
dialect is Old Ionic by the fact that
many more weak (miscalled ‘‘strong’’)
or sécond aorists, as compared with
the number of sigmatic aorists, appear
in Homer than in Attic prose. But he
forgets that both aorists existed in the
Parent-Aryan, and that there was no
reason except custom and analogy why
tenses should have continued to be
formed on the one type more than on
the other. Asa matter of fact the weak
aorist is the imperfect of the weak ver-
bal stem, and several so-called second
aorists in Homer are really imperfects.
If there are more weak aorists in Homer
than in Attic.prose, all we are justified
in inferring is that they suited the
metre better than the sigmatic aorists,
or seemed to have a greater flavour of
antiquity about them. Some of them,
like xtxetv, @orvyov, évéviroy and
qvivarov, are certainly analogic for-
mations, |

8

114 THE JOURNAL OF PHILOLOGY.

earlier date, unless (1) they are also found in Attic, or (2) can

be shewn to have belonged to the Parent-Aryan. We must

also not forget that in a question of this kind three or four
certain forms,—and Mr Monro admits that even the short list
I have given contains as many,—are quite sufficient.

Mr Monro now endeavours to set aside my argument from
the fact that whereas forms like the genitives in -ev and -eus
are monosyllabic in Homer, they are written -eo in New Ionic
inscriptions up to the beginning of the fourth century B.c¢.
In this he has the support of German scholars, who finding in
nine instances that ev has been written eo in Ionic inscriptions
infer that ¢o was pronounced as a diphthong. But the inference
is obviously unjustifiable. Eo could not be pronounced diph-
thongally, and nine instances are not sufficient to upset a
phonetic fact, more especially when we consider that they may
either be the result of a misleading analogy, or indicate a
disyllabic pronunciation of the ordinary ev on the part of the

engraver’. I do not find any German scholar venturing to-

assert that @co- in compounds was pronounced as 4 diphthong
unless it was written @ev-.

Mr Monro then suggests—at least, such I understand to be
his meaning—that words like BAwox, otvyciv, cxabo, kpoaiva,
avexnxve from the post-Homeric x«nxis, or the weak passive
future puyjoecOa:, were derived by the Alexandrine poets from
the language of the archaic period. I doubt whether he will
find many comparative philologists to agree with him.

Passing over Mr Monro’s supposition that Homer is older

than the Dorian migration,—a supposition, however, which:

seems to me not only utterly untenable but also to make the
whole history of Homeric poetry unintelligibte’—I come to his
treatment of the Molisms in Homer. Here I would. recom-
mend a perusal of the careful work of Hinrichs De Homerice
Elocutionis Vestigtis Molicis, where, by the way, Mr Monro

1 The latter alternative is supported their three tribes and already esta-
by Zeonpov for Severum (Corp. Insc, lished in Krete, which presupposes
8423). their previous occupation of the Pe-

? How would he explain Od.’x1x.177, loponnesus and maritime extension ?
where we find the Dorians divided into

as — ' bi oe es ae
es

as —— ee

DIALECTS IN THE LANGUAGE OF HOMER. 115

will find an answer to his question as to the Molic character of
«év. Mr Monro further asks why dues, vupes may not be
considered Old Ionic? I answer: (1) because this is phonetic-
ally impossible, and (2) because we know they were Mbolic.
As to Mr Monro’s idea that honorary epithets like anvpev or
traditional proper names like @epairns may have been intro-
duced into Ionic poetry directly from the spoken Molic dialects
of the day, I can only say that it seems to me in the highest
degree improbable. The poets of a pre-literary age are not
likely to have gone for their honorary epithets to another
dialect unless these epithets had already become fixed and
stereotyped in their dialectic form. And the only conceivable
way in which they could have become so fixed and stereotyped
was by their having been coined and used in Molic poetry. In
a footnote Mr Monro quotes G. Meyer’s Griechische Grammatik ;
but I do not think he has quite rightly understood the latter's
meaning. Meyer does not intend to say that we do not know
the phonetic peculiarities of the several Greek dialects and the
forms which belong to each; a large part of his grammar is
- based on the opposite assumption ; but that the relation of the
dialects one to the other is still in great measure disputable.
Even this assertion, however, is carefully guarded by the next
sentence which Mr Monro does not quote.

I have included odics in the list of Atticisms because
Herodotus certainly uses of/ while the reading o¢éov in his
text is not placed beyond doubt, and ézrecov because the parallel
forms in Homer (déere, oie, tEov, &c.) lead me to consider it to
have been borrowed by Attic literature from the Epic dialect.
@ea is not supported by the genitive Oeawy, since I regard the
latter as archaistic, not archaic. G. Meyer, imagining Gedwy
to be archaic, holds that 0ea was “derived from older non-
Ionic poetry, while in Ionic Oeds seems to have stood for
both genders.” We know, however, that the word was Attic.
Mr Monro ignores altogether the Atticisms brought forward by
- Prof. Paley, the most striking of which I have quoted, and
which, I am now convinced, Prof. Paley is right in regarding as
_ evidences of the Periklean age.

8—2

116 THE JOURNAL OF PHILOLOGY.

T now come to the examples of false analogy. Mr Monro’s
question: “how do we know” that a form produced by false
analogy “is the work of poets or rhapsodists, not of the people
at large?” can only be answered by an examination of the false
forms themselves. When they are modelled or supposed to be
modelled after archaic words and forms which had disappeared
from common use, or after corrupted or misunderstood words
and forms preserved in poetry, we possess the criterion that is
required. The instances I have chosen belong to the language
of literature not of every-day life. Mr Monro’s second question
has been already anticipated on page 518 of my Appendix. I
will now take his criticisms in detail.

(1) Our interpretation of eixm differs; I think mine the ©

more natural. It certainly has the support of similar forms.

(2) Iledevryos implies rehevya, so I do not see the force of
the objection.

(3) Benfey has long since shown that the so-called second -

aorist and the imperfect are originally the same. *“Ewéppadov
is as much the imperfect of rwedpadw as éturrov is of Tutte.

Why could the perfect réfpada not be formed ?

(4) I was wrong in saying that the futures iéjom and
Tuxynow existed in Homer, and am duly penitent. But though
Tuxjow does not exist, éruynoa does, which has the same value
as Tuynow for the purposes of my argument.

(5) The “root” of @now is Oe-, whereas the “root” of
éviomnoe is oem or rather oz, not ove.

Mr Monro goes on to blame me for fitting Wackernagel’s
“ingenious hypothesis” into my “general theory.” But surely
I may be allowed to use whatever grist comes to my mill, I
was guilty, however, of writing a misleading sentence when I
said that “the so-called diectasis...has been proved by Mangold
and Wackernagel to be the result of an affected archaism.” I
meant, and ought to have said, that it has been proved by their
researches to be so.

on ) Ff Ly *

nA ‘e > oe a zea

+) _ - -
TAB ME 8

66 Nts See at

;
;
a

— ee

—

DIALECTS IN THE LANGUAGE OF HOMER. 117

The criticisms in detail with which Mr Monro concludes his
article are relegated to a footnote; I have therefore appended

1 (1) I cannot admit Allen’s expla-
nation of the Lokrian Fér1. G. Meyer
says in his Grammar which Mr Monro
quotes asanauthority: ‘Das griechische

2

Relativum lautet és, d, 6. Die beliebte.

Identificierung desselben mit dem ai.
Relativum yds, yd, ydt scheitert an
der einen Form fFémn, die auf der
lokrischen Inschrift von Oiantheia’a 6
als Neutrum des Pronomens steht.
Vergeblich hat Curtius die Bedeutung
dieses F abzuschwiichen versucht.”

(2) Iam duly thankful to Mr Monro

- for pointing out this clerical blunder.

(3) By way of answer I would refer

to Hinrichs: De homerice elocutionis
Vestigiis Aolicis pp. 62, 63.
(4) If micupes is not Aolic, what is
it? Its phonetic form proves that it is
neither Ionic nor Doric, and at the
same time justifies the usual opinion
of scholars, which pronounces it to be
Holic. As Mr Monro himself says,
‘‘the nearest known form is the Les-
bian wécovpes,” and Lesbian is the
Aiolic dialect nearest akin to the Holic
dialects spoken on the mainland oppo-
site. The Lesbian form, however, is
more archaic than the Homeric.

(5) We can explain the various read-
ing épnpédar’ from éd\ydédar7’, but not
é\n\édar’ from épnpédar’. ’Edndé6ar’ is
further supported by €An\éar’, which,
however, as Mr Monro well knows, is
an inferior reading.

(6) Mr Monro does not say why
Clemm ‘‘can hardly be right.”

(7) Prof. Paley’s explanation seems
to me the only correct one. How does
Mr Monro propose to get rid of the con-
junction 7’?

(8) Buttmann’s explanation will not
stand,

my replies to them in the same form’.

(9) Phonology shows that w\ées can
have no real connection with the com-
parative, as the sense and syntax of the
passages in which it is found require.
The only root to which the word can
be referred is wXe- “to be full,” the
form originating in the supposed
analogy of words like evpées. Prof.
Paley may perhaps be right in thinking
that the New Ionic wiedv for mdéov
gave rise to the false interpretation of
m)ées, wAées being to mXedv as edpées to
evpvy.

(10) I must maintain my correctness
in stating that according to Curtius
the first ¢ in the infinitives in -eew is
historically false. “Ideiy for l6é-ev is the
correct form; the insertion of the jirst
e makes it incorrect,

(11) My explanation of éelcaro seems
to me the less ‘‘violent” one. For
reasons against that of Wackernagel
see G. Meyer, Griechische Grammatik
p. 193. I may add that there is quite
a long list of words in Homer (éeéva,
éevoduevos, éelkoot &c.) in which an
initial vowel, erroneously explained as
‘**prothetic,” has been introduced be-
fore the digamma through the influence
of false analogy.

(12) I was thinking of words like
KdOnuat, KadvrepOe, to which we may
perhaps add words like Jdedééyara.,
devdéxaTO.

(13) A ‘‘fixed place” may still be ‘‘a
choice of three or four’”’ when the choice,
as here, is further limited. What does
Mr Monro mean by ‘‘considering the
metrical form” ?

(14) I gratefully accept Mr Monro’s
corrections. I had already noted them
for a second edition of Prof. Mahaffy’s
work. But for obvious reasons I can-

-

Tige THE JOURNAL OF PHILOLOGY.

Mr Monro’s views of the Epic dialect seem to me to be :

influenced by a previous assumption of the antiquity of the
Iliad and Odyssey, and he is therefore anxious to explain away

whatever appears to militate against this assumption. I do not

think I can be justly accused of being influenced by a counter
hypothesis. My first investigations into the Homeric dialect
were made with a full conviction of its great antiquity, and it is
only little by little that I have been forced by what I believe to
be overwhelming evidence into the position I now oceupy. At

the time I wrote the Appendix to Prof. Mahaffy’s volume I

still thought it possible to maintain that the Homeric language
in its present form belonged in substance to the older phase of
the Ionic dialect. I cannot do so any longer. The marks of
conventionality and modernism are too numerous and inter-
penetrating to be ignored, and I cannot resist the cumulative
force of the “Periklean” Atticisms which Prof. Paley has
brought forward. Much, as I now see, that is usually termed
archaic is rather archaistic, metrical necessity and the affecta-
tion of antiquity largely dominating the choice of words and
forms’, Can anyone read Homer and Apollonius Rhodius
together without prepossessions and prejudice, and then say
that the language used in the two works is separated by a wide

not agree with what he remarks about be rejected for phonetic reasons, and
Swornp. the less said about the roots av, ai and

(15) The evidence is of course far
too long to be given here; but it will
_ be found in the works of the scholars
who have laboured upon these Epics,
In the case of the Nibelungen Lied and
the Kalévala it is so notorious that I
should haveimagined it was well known
even to Greek scholars. If the Edda
is not an Epic, what in the world is it?
The Kalévala seems to me to offer the
closest possible analogy to the Iliad
and Odyssey, especially to the latter,
and I fancy the majority of its readers
will be of my opinion.

1 The Epic afa or ain may be quoted
as an example of this. The attempt
of Curtius to derive it from yata must

the like the better. The word always
comes at the end of a verse, and there-

fore at once suggests the common ~

phrase rarpiéa yatav (and rarpié: yatn),
which I believe formed the model for
the new coinage aia, the y being sup-
posed to be the particle ye (y’). Con-
sequently warpidos ains was formed in
imitation of rarpida yatav and marplée
yatn, and the archaic word, as it was
imagined to be because never heard
in living speech, was introduced into
other passages where it suited the
metre better than yata. Hence we
have xafopwuevos alav, émixtivaras alar,
én’ atav, and vrép alys.

— ae

DIALECTS IN THE LANGUAGE OF HOMER. 119

interval of time? Of course, we are told that Apollonius Rho-
dius was an “imitator,’ but how do we know that Homer was
not one too? If we would rightly understand the Epic dialect
I believe we must regard the language of Homer not as a form
of Old Ionic or as a model for later writers, but in tts present
form as the last embodiment of an artificial dialect whose roots
- go back to the lost poems of ancient Afolis and which was
nurtured and moulded by generation after generation of Ionic
poets through long periods of time. In judging thus of the
present text of Homer, however, I do not pretend to determine
when the Iliad and Odyssey first took shape as independent
poems or whether the Homer who composed them was one or
many.

A. H. SAYCE.

ON SOME DIFFICULTIES IN THE PLATONIC
PSYCHOLOGY.

-
\

IF we compare the teaching of the Phaedo concerning yuy7)
with that of several other Platonic dialogues, two startling
discrepancies seem to be manifest. Grote, in his chapter on
the Phaedo’, has with his accustomed clearness stated them

as follows: ‘In the Phaedon, the soul is noted as the seat of

reason, intellect, the love of wisdom or knowledge exclusively:

all that belongs to passion and appetite is put to account of

the body: this is distinctly contrary to the Philebus, in which

dialogue Sokrates affirms that desire or appetite cannot belong’

to the body, but belong only to the soul....That controul, which
in the Republic is exercised by the rational soul over the
passionate and appetitive souls, is in the Phaedon exercised
(though imperfectly) by the one and only soul over the body.

In the Republic and Timaeus, the soul is a tripartite aggreg-

ate, a community of parts, a compound: in the Phaedon,

Sokrates asserts it to be ARCOM ARESE, making this fact a
point in his argument’ .

Thus the difficulties are, (1) in the Phaedo desires, fears,
&e. are attributed to body, while in the Philebus such

1 Grote’s Plato, vol. m p. 159 (2nd
ed.).

2 A third difficulty started by Grote,
I conceive to be illusory: he says,
‘Again in the Phaedon the soul is pro-
nounced to be essentially uniform and
incapable of change: as such, it is
placed in antithesis with the body,
which is perpetually changing: while

we read on the contrary in the Sym-
posion that soul and body alike are in
a constant and unremitting variation,
neither one nor the other ever continu-
ing in the same condition.’ Butin the
passage to which he refers (Symp.
207 p—208 B) there is no question of
the essential nature of soul.

a
i

“DIFFICULTIES IN THE PLATONIC PSYCHOLOGY. 121

passions are expressly denied to body and attributed to soul;
compare especially Phaedo 66 c,D with Philebus 35 ¢, D.

(2) In the Phaedo the human soul is uniform and incom-
posite; in the Phaedrus, Republic, and Timaeus it appears as
threefold and composite; whence arises the often-discussed
question: does the argument for immortality in the Phaedo
apply to all the three parts of soul, or to the highest only ?

Grote’s summary method of dealing with the difficulty is
characteristic. ‘The difference which I have here noted,’ he
says, ‘shows how Plato modified his doctrine to suit the pur-
pose of each dialogue. The tripartite soul would have been .
found inconvenient where the argument required that soul and
body should be as sharply distinguished as possible :’ and more
in the same strain.

To those who see in Plato’s dialogues only a magnificent
series of dissolving views this short cut to a solution may be
altogether satisfactory ; but if we believe that. they compose an
artistic and coherent whole—in which we may trace develope-
ment, but not contradiction—a paxporépa mepiodos must be
followed. It may indeed seem rash to attempt the conciliation
of discrepancies which so many eminent authorities besides
Grote have failed to reconcile; yet, as Sokrates says, Ta Ne-
yoweva un ovyl wravtl TpoT@ édéyxew Kat un Tpoapiotacbat
Tp av Tavtayhn oKoTav aneimy Tis Tavu wadOaKod éoTiv

. av6pos.

The way to the solution of the first problem clearly lies
through the second, which we will accordingly take first.

It appears to me that some light may be thrown on the
question by a closer examination of this very tripartite division
of avy; as a preliminary to which it may be well to note
briefly how the case stands with regard to Wvy7 in the Pla-
tonic dialogues, excluding the Republic and parts of the Timaeus
and of the Phaedrus.

Wuyn then is the principle of life which vivifies the entire
universe, interpenetrating its whole mass from centre to cir- ©
cumference (Zim. 36 £); she is nature’s upholder and sustainer
(Krat. 400 A); having her motion of herself she is to all things
that move the source and principle of motion (Phaedr. 245 ¢

122 THE JOURNAL OF PHILOLOGY.

cf. Laws 895 B—896 A); she is without birth or death, through
all eternity existing (Phaedr. 245 D,E); she is the guardian of
all that is soulless (246 B); she is divine, deathless, spiritual,
uniform, indissoluble, self-identical, changeless (Phaedo 808);
akin to the ideas and coeternal with them (79 D, 92D); the
only seat of reason (Soph. 249.4); the one cause and means of
communion between the ideal and material worlds (Phileb.
28 cseq. where vots is identified with the aitia tis pikes).
The, human soul is derived from the universal soul (Phileb. 30 A),
differing only in the inferior purity of its substance (Tvm. 41 D
cf, Phileb. 29 B, C.); 1t possesses, as we gather from the Phaedo
and Phaedrus, conscious and individual immortality; it is an
indwelling essence distinct from the body but by some mys-
terious union informing and controlling it; it apprehends
sensible objects by means of the bodily organs (Theaet. 184D) ;
‘and it alone, by virtue of its affinity with the ideas, has the
power of contemplating pure being and absolute truth (Phaedo
66).

In all this we find that vy, whether universal or par-
ticular, is treated as a substance, one and indivisible: and so it
is everywhere in Plato, except in the passages I am coming to
consider, the single principle of life, sensation, and thought. It
is certainly not a little strange, if in three of the Platonic
dialogues there exists a theory of vy totally at variance
with this conception ; which is not one belonging to a particular
period but is constantly occurring throughout the whole series
of Plato’s writings; and this strangeness is the more startling
when we observe that in these three dialogues themselves the
usual conception is also apparent. |

It is in the celebrated allegory of the Phaedrus that we are
first brought face to face with the difficulty. The individual
soul, of god or man, is likened to a car driven by a charioteer
and drawn by two steeds, one noble, spirited, and docile, the
other lustful, vicious, and intractable. The meaning of this
similitude, clear enough from the context, is fully explained in
Republic 434—441, where we have. the soul divided into a
AoyoTiKOv. and an ddoyov eldos, the droyov being subdivided
into Ovpoedées and ériOvyntixov. The soul thus appears com-

—
ON engage

DIFFICULTIES IN THE PLATONIC PSYCHOLOGY. 123

pounded of three distinct parts or kinds, rational, emotional,
appetitive: in describing them Plato uses indifferently the
words eidos, yévos, and wépos—cf. Republic 435 0, 441 0, 444 B—
but no expression of Plato’s, I think, warrants Grote in speak-
ing of three souls’.

But if this is really so, what are the consequences? Let us
turn to the Phaedo. At first the mutually complementary
arguments of dvtamddoc1s and avayynots seemed to carry
~ conviction to us; yet presently we feel that even their com-
bined persuasion has no charm potent enough to dispel our
fear lest the soul that passes forth of the body on a stormy
night be blown asunder and scattered like smoke on the blast :
the child in us can only be soothed with the assurance that the
soul is not compounded of parts, therefore into parts it cannot
be resolved; that it shares the nature of the ideas and there-
fore shares their eternity. But now our soul that was incom-
posite and uniform and like to true being has turned out to be
composite and triform and therefore as unlike true being as
can well be conceived. The downfall of the argument is utter
and ruinous; we are left hopelessly wondering whether all the
parts survive the body, or one, or two; and if more than one,
whether in union or apart; if the ésv@uvuntixdy survives, what
does the soul’s release from the body profit her? finally whether
the argument that has betrayed us on so important a point be
altogether faithless, and our soul die utterly with the body.
It has been maintained that the Phaedo deals with the do-
yioTixoy eldos alone; but there is not one word in the dialogue
which countenances the supposition that Plato is using the
term uy?) in a more restricted sense than elsewhere; nor
does he anywhere show an inclination to confine the title to the
highest eiSos. I cannot but regard this explanation as a forlorn
hope. Apart also from the subject of immortality, we are laid

1 Compare Aristotle Nic. Eth.1 xiii xuprdv kal 7) Kothov, obfev Scapéepet mpos
9, 10. ofov 7d wev ddoyov airfs elvae ro 76 rapov, Here we may remark that
dé Abyov €xov. traira 5é rérepov Siwpicrac Aristotle does not regard the distinc-
Kabdmep Ta To0 osuaros wopia Kal wav 7d ~— tion between adoyov and do-yov éxov as
Lepirrov, 4 7H Adyw So early dxwpicra necessarily implying parts,
mepukora Kabdrep ev rH wepipepela 7d

124 THE JOURNAL OF PHILOLOGY.

open to some inconvenient metaphysical questions: for instance,

what is the essential difference between these parts of wuy7 ?
what the nature of their union? what the common principle

by virtue of which they are all termed wuyn*? I cannot be-

lieve that Plato intended to set his scholars adrift on such a

sea of perplexity as this; still less, as Zeller? would have us
suppose, that it is a question which Plato doubtless never defi-
nitely set before himself; which at least he has done nothing
to answer. |
But in one or other of these conclusions we must, I think,
have acquiesced, had Plato never written the Timacus. The
account given in that dialogue still remains for consideration ;
and this, while at first sight it seems to plunge us even deeper
in perplexity, really in my belief gives us the clue that shall

guide us out of the maze. This passage (Tvmaeus 69 C seq.) has ~

such an important bearing on the question that it may be well
to translate some part of it in full.

‘And the created gods following the creator’s example,
when they had received from him an immortal principle of

soul, went on to frame about it a mortal body, and all this.

body they gave it to ride in: moreover they enclosed within

the walls of the body another kind of soul, even that which is

mortal, having in itself dire inevitable passions—first pleasure,
the strongest allurement of evil, then pains that scare away
good things; rashness also and fear, two thoughtless counsellors;
wrath hard to assuage and hope that lightly leads astray—all
these forcibly they mingled with reasonless sensation and love

that ventures all things, and so they composed the mortal kind _

of soul. Wherefore, in awe of defiling the divine, but so far as
needs must be, they lodge the mortal kind apart from the
divine in another chamber of the body; and between the head and
the breast they constructed an isthmus to sunder them, placing
the neck in the midst that they might be separate. So in the

breast and the chest, as it is called, they confined the mortal

part of the soul. And seeing that one part of it was of higher,
another of lower nature, they built a wall across the hollow of

1 Compare Aristotle de anima 1 vy 24-—-27; ur ix.
2 Gesch. d. gr. Phil. 11 i p. 717, 5rd ed.

PIT Les «,

= OO

> =

DIFFICULTIES IN THE PLATONIC PSYCHOLOGY. 125

the chest, as if they were marking off separate apartments for
women and for men, setting the midriff as a partition between
them. That part of the soul which shares courage and spirit,
since it is warlike, they placed nearer the head between the
midriff and the neck, that it might be within hearing of the
reason and might aid it in forcibly restraining the tribe of
lusts, whenever they would not willingly obey the signal and
word of command issued from the citadel’

I have translated so much verbatim, in order to bring out
clearly the highly figurative character of the passage. In 70 D
Plato describes the position of the éwv@upntixdy as follows:
‘That portion of the soul which lusts after meat and drink
and all things which because of the body’s nature it needs,
they placed between the midriff and the boundary at the navel,
constructing in all this region as it were a manger for sus-
tenance of the body; and herein they chained it like a wild
beast, which nevertheless must be reared in union with the
rest, if a mortal race were to be at all.’

Zeller, interpreting this passage literally (p. 715 seq.),
asserts roundly that these three forces are not different forms
of energy but actually different parts of the soul, to which Plato
shall even assign separate positions in space. I believe that
most of Plato’s commentators, from Aristotle downwards, have got
into trouble by failing to realise that the Timaeus is not only a
profound philosophical speculation but also one of the most
fanciful of fairy tales. Zeller, though he has elsewhere shown
himself fully alive to this, has, I think, been here forgetful of
it, and has thus failed to reach Plato’s meaning. Moreover
while he naturally feels that the passage, as he understands it,
leads to grave difficulties (p. 717), he does not seem adequately
to appreciate the irreconcilable inconsistencies it involves.

For to the difficulties before mentioned as attaching to the
triform nature of yvy7) this passage adds two still more hope-
less perplexities: (1) the three parts, as Zeller says, have
distinct locations in specified regions of the body; all three

1 Compare with this Politicus 309 c. pocauévn Seou@, wera 5é 7O Oefov rod
mparov uev Kara TO Evyyeves TO devyeves fwoyeves abrav ad&s avOpwmrivas.
dv THS Wuxns atrov pépos Oelwy suvap-

126 THE JOURNAL OF PHILOLOGY.

therefore, it would seem, have extension in space; yet Plato
has again and again told us that soul is immaterial: (2) the
two inferior parts are declared to be mortal. This does not
seem to have given Plato’s interpreters much anxiety, but surely
‘a more startling statement could not proceed from his mouth;
indeed in what possible ‘sense of the two words Plato can use
the combination Ovn77) Wuvyx7) Iam utterly at a loss to imagine.
It is, if possible, a more absolute contradiction in terms than
mop uxpov or even aptia tpias. If weyn, that is vital prin-
ciple, can die in any case, what becomes of the final argument
in the Phaedo? what is the end of all Plato’s endeavour to
discover some stable and permanent object of knowledge?
acYoAH yap av tL adrXo POopay pn SéyorTo, ei ye TO aPavatov

aidiov dv dOopav Sé£eTar. Again in the Phaedrus (245 D) it is ~

positively asserted that all soul is immortal, 7dca wuy7 abava-
tos; and this only a few lines before the tripartite nature of
the soul appears in full developement: that is to say Plato first
affirms without limitation that all soul is immortal and imme-
diately afterwards describes it as consisting of a mortal and
immortal part united. .

We are therefore driven to choose between the following
suppositions: (1) Plato has directly contradicted himself on a
point of the gravest importance ; (2) the term wWuyx7 is used by
him in different senses ; (3) the expression OvnTov eidos Wuyis is
to be explained so as to harmonise with Plato’s other state-
ments on the subject. The first must, I think, be dismissed
without ceremony: it is surely incredible that the greatest and
most careful of all original thinkers, on a point which he had so
much at heart and on which he bestowed so much pains, has un-
consciously fallen into so obvious and glaring an inconsistency.
Secondly, if the Ovnrdov eidos be a different substance or sub-
stances from the a@avarov, what is this substance? a question
which to ask is inevitable, to answer, I think, impossible. For
what right has this mortal substance to share the name of yuyy,
whose essence is immortality? we have between spirit and
matter a third substance, sharing, as it seems, the properties of
matter—for it is extended and perishable—but classed under
the same title as spirit. As the origin of emotions and appetites

+#
Mag

és
Ao.
:
i
7 AZ
1.4
7
3

a

DIFFICULTIES IN THE PLATONIC PSYCHOLOGY. 127

it is a source of motion: either then its motion is of itself or
from without: if of itself it cannot be mortal, for the self-moved
is immortal ; if from without, it cannot be Wvuy7, for the essence
of aux) is self-motion. We are thus forced to adopt the only
remaining alternative; which amounts to this. It is not yuy
which is mortal, but certain activities of Wuy7 in certain rela-
tions which are terminable and determined by separation from
matter; we must accept, in fact, the conclusion which Zeller
rejects, that the three parts are not ‘verschiedene Theile’, but
‘verschiedene Thatigkeitsformen’, or modes of operation. To be
more explicit: all soul; as such, is eternal and uniform, nor are
there more kinds of soul than one. But soul when it enters
into union with matter is forced more or less to operate through
matter; and the names given to this combined action of soul
and matter are @uyos and émiOvpia. Therefore Oupoedés and
émvOupntixoy are expressions for soul in certain material rela-
tions; and as the connexion of soul with body is terminable,
Oupoedes and ériOvpntexov, as such, are perishable. But this
does not mean that the vital principle, which in its material
connexion assumes these forms, is perishable, but only that the
relation is temporary : yuy7 exists as yuy7 eternally; as éuOv-
pntexov and Ovyoedés only so long as its connexion with matter
continues.

This view will, I think, be found to agree entirely with
Plato’s whole teaching on the subject. It must be remembered
that the two lower eiéy of soul are found only in conjunction
with matter. Even the gods, to whom all three parts are
attributed, are corporeal: a god is a@avarov tt fdov, Eyov pév
wuxynv éyov S¢ cdua, Tov del 8 ypovoy tadta EvytredvKota
(Phaedr. 246 p)*: only, as their bodies are more ethereal, their
soul is more free to act independently ; consequently in them
reason is supreme. Similarly in the Phaedo, in proportion as
the soul withdraws herself from communion with the body,
pure reason predominates over the earthly and sensual appetites.
All this is perfectly reasonable, if we conceive soul as a single
essence, constrained in certain circumstances to work through

1 Even in their case the connexion _tinuance depends upon the will of the
has no inherent permanence: itscon- creator (Timaeus 41 4).

128 THE JOURNAL OF PHILOLOGY.

matter; but surely Plato did not mean that soul, being when

apart from body a uniform essence, on entrance into a material
abode all at once annexes two inferior substances, being parts of

itself and yet essentially different.

It is well also to bear in mind that although the three edn
are commonly spoken of'as three coordinate parts of soul, the
main division is really twofold, Noyiorexdv and ddoyov as ex-
pressed in the Republic, a@avarov and Ovnrov in the Timaeus ;
this is appropriately represented in the Phaedrus by the driver
and pair of horses. This is important to notice; since, were the -
three divisions coordinate, the view I am here maintaining ©
would involve serious difficulty. As it is, the distinction be-
tween Oupoedées and ériOvuntixoy is simply a classification of
the operations of soul through body.

The general physical application of this theory is perfectly
‘simple. All living things derive their life from a single uniform
principle, that is soul. In the gods, if such there be, soul
possesses the highest state of freedom compatible with material
existence: bodily affections they must have, since they are
corporeal; but, owing to the predominance of spirit over matter,
their affections are entirely controlled by the reason: in man,
since soul is bound in a much closer union with matter, the
power of reason is greatly diminished, while that of the passions —
is proportionately increased ; still the philosopher, whose whole
life is a ‘study of death’, can so far abstract his soul from its
bodily connexion as to attain a considerable degree of intellec-
tual freedom. In the lower animals, as we descend the scale,
the implication of soul with matter becomes more and more
complete, and in proportion as the reasoning power decreases *
the purely animal impulses predominate; till in plants all we
find of life is a mere faculty of growth. But the reason of a
god and the growth of a moss are alike operations of one and
the same vital principle; in one case acting in almost complete
independence of matter, in the other inextricably entangled
with it.

The view that by @vnrov eldos Wuyjs Plato means not
a mortal kind of soul but a terminable mode of soul’s exist- —
ence is thus, I think, shown to be in harmony with his general

DIFFICULTIES IN THE PLATONIC PSYCHOLOGY. 129

teaching and to release us from grave difficulties. And surely
it cannot be argued that in this passage of the Timaeus—one of
the most figurative passages of Plato’s most allegorical dia-
logue—we are compelled to understand every phrase with
verbal literalness. It would be as reasonable to maintain that
_ Plato meant us to accept literally the account of soul’s con-
struction by the Snpsovpyos, involving its composition out of
three elements and its beginning in time; both of which are
directly contradictory of Plato’s theory of soul.

It follows from what has been said that the question
whether the reasoning in the Phaedo refers to all the parts
of the soul or not is quite beside the mark. Plato has ignored
the threefold division, not, as Grote says, because it would
have been inconvenient, but because his argument is entirely
unaffected by it. His demonstration applies of course to soul
as such, not to particular relations of soul. The vital prin-
ciple, of which the @upoedés and éavOupnrixdy are manifesta-
tions in conjunction with matter, exists eternally ; but @upyoer-
dés and éwiPuyntixov themselves are merely temporary modes
of its operation. The whole difficulty vanishes with the
notion that Plato held the existence of more than one kind
of soul.
If this solution of Grote’s second difficulty be accepted, the
explanation of the first is easy. In the Phaedo Plato is dealing
with soul as such, with which bodily appetites, &c. have
nothing to do: these belong to soul in its corporeal relation,
and can only affect it through such relation. Consequently
from this standpoint of the Phaedo Plato is perfectly justified in
attributing such passions to body; because they arise from the
union of soul with body. Any closer investigation of their
nature would have been foreign to his purpose: and Plato
always likes to do one thing at once. In the Philebus on the
other hand we are specially concerned to examine scientifically
into the nature of pleasures and desires; and they are accord-
ingly attributed to soul. But the discrepancy is only apparent.
In the Phaedo they are assigned to body, because they can-
not affect soul except when it is in connexion with matter ;
in the Philebus to soul, because matter as such is insensate.

Journal of Philology. vou. x. 9

130 HE JOURNAL OF PHILOLOGY.

The latter statement is more exactly scientific, but it could
not have been made in the Phaedo without raising irrelevant
issues. In both the meaning is the same; that is, desires and
passions are phenomena belonging to a conjunction of soul
with body: inasmuch as the soul is the seat of these passions,
they are properly assigned to soul; inasmuch as body is a ne-
cessary condition of their existence, they may be said to belong
to body. |

The easy explanation thus obtained of apparently so grave
an inconsistency is, 1 think, another strong piece of evidence
for the solution proposed of our former problem, Indeed,
except on the supposition that soul, notwithstanding the figura-
tive descriptions of it in the Phaedrus, Republic, and Timaeus,
is a simple substance, I do not see how any psychological
theory can be attributed to Plato, which he could conceivably
-have constructed. This view is also, I believe, not without
important application to the theory of ideas; but that is far
beyond the limits of the present inquiry.

I conclude with a brief summary of the preceding argu-
ment :—

The difficulties are two: (1) in the Phaedo ériOupias
are attributed to body, in the Philebus to soul: (2) in the
Phaedo soul is simple, in the Phaedrus, Republic, and Timaeus
it is triform. I have endeavoured to solve both problems in
the following way:

In Timaeus 69 c—72 D, we have a @efov eidos and a Ovntov
eldos of Wuyn: of which Oeiov =royiotixdy, Ovntov = Ovpoe-
Sés and ériOupntixdv. Now ~uy7 as such is a@avarov;
therefore the word @vytov can only refer to a particular rela-
tion of uy?) .and oda, or operation of yvy) through capa.
Ovpoedés therefore and éwidupynrixdv are not different parts of
aruxn but only names for different modes of its action through
cépua: thus Pvpoedes and értOupntixoy are Ovnta, because,
when the conjunction between yuy) and oda ceases, they
cease also. 7

Thus, (1) the apparent discrepancy between the Phaedo
and the Philebus is reconciled. In the one émvOuplas are
ascribed to oda, because arising from the conjunction of

"DIFFICULTIES IN THE PLATONIC PSYCHOLOGY. 131

aby) and odua; in the other they are more accurately as-

cribed to Wuyy, because they are an affection of uy through
oodpa: (2) the argument of the Phaedo is entirely unaffected
by the threefold division, All soul is simple, uniform, and
indestructible; but in connexion with body it assumes certain
phases which are temporary and only exist in relation to body.
Thus though the éewvOuuntixdv and Ovpoesdés as such are not

immortal, because they depend for their continuance upon

body, which is mortal; yet the vital principle, which under such
conditions assumes these forms, is immortal and continues to
exist, though not necessarily in the same mode. For the
modes in which vital force acts under temporary conditions

are transitory, but the acting force itself is changeless and
eternal,

R. D. ARCHER-HIND.

ON PLATO’S REPUBLIC VI 509 p sqq.

Mr H. Srpawicr’s excellent article in the Journal of Phil-

ology 11 96—103 is still, so far as I know, the best statement of ©

the received interpretation of the last pages of Republic vi and of
the difficulties which beset it. But his remarks are offered “ with

a view less to solve the difficulties of the passage, than to de- —

fine them more clearly than has yet been done,” and accordingly
cannot be considered final. In the present paper (which would
hardly have been written but for Mr Sidgwick’s) I propose to
review the passage together with its context and to offer some
suggestions for its interpretation. On a future occasion I hope
to comment upon the metaphysical portion of the Philebus, and
to institute a comparison of the ontologies of the two dialogues
such as Mr Sidgwick’s concluding paragraph seems to invite.

§1 Tae Line.

In the pages preceding those which specially concern me
‘Socrates has illustrated his theory of the supremacy of the avro
ayaov, the origin of Being and of Knowledge, by comparing it
to the sun, which in the visible world’ is the origin of Becoming
and of Light. He now 509D proceeds at Glaucon’s request to
amplify and complete the similitude of the sun (T7v ep) rov
HAvov duowoTnta). Let us suppose, he says, the visible world

1 That opardv isnot tobeconfounded to guard against this misconception.
with alc@nrév, might perhaps be as- It is only in the case of sight, he there
sumed; but as commentators have tells us, not in that of the other senses,
supposed the whole region of sense to that there is a rplrov which will serve
be referred to, it may be well to note as an elxdy of the aird dyabév.
that at 507 c sqq. Plato has endeavoured

Ye. oe

wes HOT Wh iJ PVP CE Par Drie
Ud PC LIS lads si ey “ 1“ . ‘
' / et

PLATO'S REPUBLIC VI. 133

presided over by the sun and the intelligible world presided
over by the av’rd dyafdy to be respectively represented by the
two segments of an unequally divided line, and let us further
divide each of the two segments similarly to the whole line.
The four segments thus obtained may be taken to represent, in
respect of comparative clearness or truth,—

1 Images, i.e. (1) shadows and (2) reflections in water, &c.

2 Things by which images are cast, whether (1) products
of art or (2) products of nature.

3 That which the soul studies (a) descending from hypo-
theses to a conclusion, (b) by the aid of visibles treated as
images. |

4 That which the soul studies (a) ascending from hypo-
theses to a principle which is not hypothetical, (b) by the aid
not of images but of forms’. —

At this point I pause to comment. According to Mr Sidg-

_ wick, who assumes at the outset that “the universe is compared

to a quadripartite line,” “We have (omitting the fourth seg-
ment as of no metaphysical importance) three processes of
apprehension carefully distinguished: and corresponding to the
first and third two sets of objects, material things and ¢ié7. We
naturally expect therefore a set of objects intermediate between
the two corresponding to the intermediate process.” Just such
an intermediate set of objects we have, thinks Mr Sidgwick
with the commentators in general’, in the paOnpwatixa men-

1 Until it is possible to arrive at a
distinct conception of the two sorts of
vonrév, 1 shall sometimes for the sake
of brevity speak of them as ‘the in-
ferior vonrév’ and ‘the superior voyrév’
respectively, but I shall mean by these
terms no more than ‘the object of the
inferior intellectual method’ and ‘the
object of the superior intellectual
method.’

2 «Unter der didvora oder ériorjun
versteht Plato (wie auch Brandis

annimmt) ausschliesslich die mathe-
matische Wissenschaft; er selbst sagt
diess Rep. v1, 510 B f. 511, cf. aus-
driicklich.”” Zeller Gesch. d. gr. Ph.
11 i 537. ‘*Da Plato jedoch das
Mathematische und zwar zunichst die
Zahlen, wie wir sehen werden, fiir We-
senheiten hielt,diein der Mittezwischen
dem sinnlich Wahrnehmbaren und
den Ideen, so hat er unter jenen so-
genannten Wissenschaften oder Kiin-
sten doch wohl lediglich die mathema-

134 THE JOURNAL OF PHILOLOGY.

tioned by Aristotle metaphysics I 6 § 4, 987 b 14, though, as
Mr Sidgwick himself acutely remarks, “the language” [of 510 p]/

“in no way supports this interpolation of intermediate objects.”
So far Mr Sidgwick. The obvious, and, I think, fatal, objection

to this interpretation is that it leaves one of the four segments
unexplained. By way of answer to this objection Professor
Jowett suggests (1) that “Plato had been led by the love of
analogy to make four terms instead of three,” and (2) that “each
lower sphere is the multiplication of the preceding”; in other
words, that the proportionals are a, ar, ar’, ar*, so that when
the superfluous a (the first segment) is omitted, it may still be

true that ar is to ar’ as ar’ is to ar*,i.e. that sensibles are to |

intermediates as intermediates are to ideas.’ But these incon-
sistent suggestions do not dispose of the objection: for (1) three
proportionals would have satisfied Plato’s love of proportion just
as well as four; and (2) the proportionals are not, as Professor
Jowett assumes, and apparently Mr Sidgwick also, a, ar, ar’,
ar*, but, as Whewell has pointed out (Philosophy of Discovery
p. 444) a, ar, ar, ar’, so that when a is omitted, the three
remaining terms do not give the relation supposed by. Professor
Jowett and Mr Sidgwick between sensibles, intermediates, and
ideas. It would seem then that the introduction of the first
segment is unmeaning, and worse than unmeaning, on the
assumption that “the universe is compared to a quadripartite
line”; and it may therefore be worth while to inquire whether
this assumption is necessary or justifiable.

_tischen verstanden und sie fiir die telndenDenkens zureichend.”’ Brandis
ausschliesslichen gehalten, bei denen Gesch. d. gr.- rém. Ph. 111 272.
das hypothetischeVerfahren des vermit-

‘*Zusammenfassend schematisirt Plato Rep. 509 ff. und 533 f. in folgender
Weise :
A. OBJECTE.
Nonrdév yévos (ovcta).
"Tédéar. | Maénuarcxa.

‘Oparov yévos (yéveats).
Zwuara. =| Eixoves.

B, ERkENNTNISSWEISEN.

Noyes. Aoéa.
Nods (oder vonars |

Acdvoww. — Iliorcs. | Kixacla.””

oder émornpn).
Ueberweg Grundriss d, Gesch. d. Ph, 1 182.

a _
’ ee
‘ : A
« 7
. ‘7

F a”
i" Mets
~ "flee
ee
’ |
:
‘
a_i.
5
*'
;

PE ee ne ee oe

ae

PLATO'S REPUBLIC VI. 135

Let us return to the preceding context. When at 506 E
Glaucon challenges Socrates to give at all events a popular
account of the aya@ov, such as that which he has previously
given of the virtues, Socrates professes his inability to do so,
but declares himself willing to explain his notion of it by refer-
ence to its éxyovos, the sun, which (not in the whole sensible
world but) in the visible world is duovdratos éxeivw. Now it is
plain that the sun is “of no metaphysical importance,” except
so far as it illustrates the Platonic Socrates’s notion of the avrd
ayaSov; and, as in what he says about the quadripartite line
Socrates is amplifying and completing tiv mepl tov HrALov
\émoudTnTa, and indeed, when he resumes his discourse, is careful
to repeat that the sun presides over the dparov just as the
ayabov presides over the vonrov, it would appear that both the
lower segments, not one only, are devoid of “metaphysical
‘importance,’ except so far as they help us to understand the
‘relation between the two higher segments. We have then here
what may be called (in imitation of Aristotle) an eixoyv Kar
avanroylav— as images of things are to things imaged, so is the
inferior vontov to the superior voytoy, the popular distinction
‘between images of things and things imaged, duoww0év and 6
'ovwOn 510 B, being used to explain the metaphysical distinc-
‘tion between the respective objects of the inferior and the
‘superior intellectual methods.

Now if the object of the inferior intellectual method is to

_the object of the superior intellectual method as an image or
reflection of a thing is to the thing itself—in other words,

if the inferior vontév may be regarded as an image or reflec-
tion of the superior vontov, it would seem that the objects
of the two sorts of intellectual method are not distinct exist-
ences, but the same existences viewed in the one case in-

‘directly and in the other case directly. Thus as soon as we

‘discard the assumption that “the universe is compared to a
quadripartite line,” and recognize the purely illustrative cha-

racter of the first and second proportionals,—for the four.

segments are only four proportionals geometrically expressed,—
all reference to ‘the intermediates’ of the metaphysics (which
differ from the ideas in that each of them is a moAda@ and

a

“J

136 THE JOURNAL OF PHILOLOGY.

not a év’) disappears, and with it the inconsistency which Mr/

_Sidgwick finds between “the Se urpdia 3 of the theory” “a

“the language of the passage.”

But what are the two objects related to one another as
thing to image? That the superior object is the idea, is indi-
cated at 510 BD 511 B, and is indeed generally acknowledged.
What then is the inferior object, ‘the image or reflection of the |
idea’? In the case of every group of particulars to which we
give the same name, we assume the separate existence of an’
idea in which these particulars participate. This idea is the
whole completed connotation of the name, as it would be
understood by omniscience, hypostasized. Now the general
notion is the connotation of the name as we imperfectly under-,
stand it, not hypostasized. For example, the idea of sulphur
is, hypostasized, the whole sum of the properties, known and.
unknown, which are common to specimens of sulphur: the
general notion of sulphur includes, not hypostasized, so many
of these as are known to us. The general notion is therefore’
not the idea, nor a correct and complete representation of the
idea, but an incorrect and incomplete representation of it.
May we not assume, apart from any indications to be found in
Plato’s account of the methods of investigation, that by ‘the
image of the idea’ he means the general notion *?

The view which I take of the significance of the several
segments of the line finds, I think, some confirmation in the
well known chapter of the Phaedo descriptive of the Platonic
Socrates’s aspirations and failures. Having in the course of

his physical researches learnt to draw a distinction between

‘cause’ and ‘condition, the word ‘cause’ being properly used
only in the sense of ‘final cause,’ Socrates was astonished to
find that Anaxagoras, when he had arrived at the notion of an
intelligent author of all things by whom chaos was reduced to .
order, did not complete his theory by providing an ayafov

1 Hereafter I shall have something 2 The elxoves of 402 a are, I conceive,
to say about 510 c—nz, where the to be understood in the same sense.
mathematical object, e.g. ‘the reflec- Of the gayrdcuara év veacw of 516 4
tion’ of the avré rplywvov, is recognized 53281 shall have something to say
as @ part of the inferior vonrép. presently.

eee

PLATO'S REPUBLIC VI. 137

which the intelligent author of all things should seek’. He had
expected Anaxagoras, when he wished to assign the cause of
anything, to show that ‘it is best. that the thing in question
should be as it is” the cause of each thing being 70 éxdot@
BéxtvcTov, and the cause of all things 70 xowov radow ayabor.
Anaxagoras however had done nothing of the sort, and Socrates
had not succeeded in supplying the deficiency. He found
himself therefore obliged to have recourse to another line of
inquiry, though he had never abandoned his conviction that a
really satisfactory theory of the universe must be teleological.
Now in the passage above summarized, although the phrase-
ology of the theory of ideas is carefully avoided, the supremacy
of the aya@dv is as distinctly asserted as anywhere in the
republic’. In both places it is the origin both of Being and of
Knowledge. The main doctrine of the two passages being thus
the same, the resemblance of the next following sentences in
the Phaedo to the sentences in the republic already considered
can hardly be without significance. The investigation of things
(évrTa), we read at 99 D, having proved a failure, Socrates now
proceeded to study their reality (trav dvtwv tiv adnOeav) in
definitions (Adyoz), just as an astronomer, fearing to be blinded

- if he watches an eclipse of the sun directly, observes its image

reflected in water. It will be seen that we have here an eikadv
Kat avadoyiav—as an image of the sun is to the sun, so are
Aéyot to dvta—which efxdv resembles the avadoyia in the

ri | i) if] Bret ie ‘ Om
neat = menses -
\ Pda, Boe rs ee : sna
, d 4 Stein ee
heat ;

1 It would seem that here at any
rate the dyadov is not to be identified
with voiis, of whose operations it is
the ov evexa. .

2-Mr Sidgwick however, who takes

no notice of the indications of doctrine

contained in the criticism of Anaxa-
goras, is of a different opinion. ‘‘ There”
[sc. Phaedo ¢. xu1x], he says, ‘‘ Plato’s
ontology is obviously in a different
phase, as to ayafov (here placed at
the summit) is ranked indiscriminately
with other ef5n that Socrates supposes
(vrorl@era).” But asin the republic,

-in which dialogue its supremacy is

\ admitted, the idea of good is repeatedly

(476 a 484 pd 507 B 531 c¢ 588 8)
ranked indiscriminately with other
ideas, this can be no reason for de-
nying its supremacy in the Phaedo.
In fact, although as compared with
other ideas the idea of good occupies a
higher position, as compared with par-
ticulars it may be ranked with the
rest. Compare republic v1 509 B, where
the dya6dv is placed éwéxewa THs ovctas,
with vir 518 p 526 z, where it is spoken
of as Tov ovros 76 gavdraroy and ré
eddatmovéorarov Tov OvTes.

138 THE JOURNAL OF PHILOLOGY.

sixth book of the republic and countenances the views above _

taken, both of the relation in which the first and second pro-
portionals stand to the third and fourth, and of the signin
of the third proportional’*.

1

§2 THE CAvE.

That ‘the line’ and ‘the cave’ are intimately connected, is
obvious : indeed it is expressly asserted at 517 B, Tavrnv toivun,
hv & éyo, tHv eixova, @ pire TAavKov, tpocamtéov &tacav
Tois éumpoolev NEyopévots.
whether the results so far obtained in regard to the line accord
with the allegory which opens the seventh book.

If we tabulate the objects successively seen by the prisoners
together with a conjectural interpretation of those objects,
placing the imagery of the allegory on our left, and the signifi-
cation on our right, we have (exclusive of details with which I
am not now concerned)—

Within the Cave dotacrov

Shadows of statuettes of things =particulars as apprehended by the senses.
Statuettes of things =particulars as they are (or become) in them-
selves,

Without the Cave

Reflections of things
Things themselves

vonrov
= objects of the inferior intellectual method.

= objects of the superior intellectual method.

That ‘shadows of statuettes of things’ stand for ‘ particulars as
apprehended by the senses,’ and therefore that ‘ statuettes of
_ things’ stand for ‘particulars as they are in themselves,’ almost
every reader? will take for granted: that ‘reflections of things’

1 Tt would appear that évra generally, dya@dv as well as with the parallel pas-

It will therefore be well to inquire ~

not édvyTws dyTa as opposed to yyvo-
peva, are here contrasted with doyou.
Hence I am careful not to identify the
dvaoyla of this passage with the
avadoyla of the sixth book of the re-
public. The commentators with one
accord assume that yiyvoueva as op-
posed to dvtws. dvra are here con-
trasted with Aéyo. This limitation

seems to me inconsistent with So-

crates’s narrative of his search for the

sages.

2 Professor Jowett however makes
“the shadows” and ‘‘the images”? [i.e.
the statuettes which cast the shadows]
‘‘correspond—the first, to the realm
of fancy and poetry,—the second, to
the world of sense.” This interpreta-
tion seems to be precluded by the
description, 516 o—xz, of the mental
condition of the prisoner, when his
gaze is-turned for the first time from

re Tt.

oe

*

Re ae
mii.
OE a yO

PLATO'S REPUBLIC VI. 139

are equivalent to the inferior vontéy of the sixth book, cannot
be doubted, since ‘the contemplation of reflections of things in
water’ must stand for the mpomatde/a, which again includes the
sciences or arts mentioned in 510 c and 511 c (cf. 533 D) as
employing the inferior intellectual method: and that ‘things
themselves’ stand for the superior vonrov, is certain, since in
both passages the highest object is er assigned to dialec-
tic 511 c 534 D 536 c.

Let us now seek in this tabular statement of the allegory
the four terms of the original dvadoyla. If, as Mr Sidgwick
assumes, “the universe is compared” [in 509 D sq.] “to a
quadripartite line,” the four segments should all be found on
the right of the page, sensibles standing second : if however, as
I have supposed, the first and second terms of the avadoyia
merely illustrate the relation between the third and fourth
terms, we shall expect the first and second terms to occupy the

- third and fourth places on the left of the page, the third and

fourth terms as before occupying the third and fourth places on
its right. Now on Mr Sidgwick’s assumption the old difficulty
méets us again—‘the shadows of statuettes of things’ are
superfluous. On the other hand the requirements of my theory
are perfectly satisfied, reflections (ra év tois téacs pavTacpara)
having been expressly mentioned at 510 4 as a species of ecxoves,
and the ‘things themselves’ (avrd) of the allegory being
obviously capable of identification with the Sa xai_mav_to
gutevtov of 510 A. Thus the avadoyia of the former passage,

‘shadows of statuettes of things’ to
‘statuettes of things,’ as it is not easy
to see why the transition from ‘‘the
realm of fancy and poetry” to ‘the
world of sense” should occasion dzopia,
nor why the former should seem ‘more
real’ than the latter. On the other
hand if the repiaywy? or repiorpopy is
the discovery, under the influence of
such an é@\eyxos as is applied to the
young Theaetetus in the dialogue cagled
by his name, that sensation is not ob-
jectively true, nothing could be more
appropriate than the question ovx ole

avrov amopeiy te av kat nyetoOa Ta
Tore Opwyeva adnOéorepa_ y Ta viv
Beckvipeva 3

It is remarkable that Professor
Jowett, who sees that ‘the reflec-
tions in water” of the seventh book
are ‘the abstractions or universals
of sense, of which the mathemati-
cal sciences furnish the type,” should
have given in the sixth book an en-
tirely different meaning to the third
segment, That the two passages must
not be inconsistently explained, seems
to me certain.

140 THE JOURNAL OF PHILOLOGY.

as I understand it, gives us in a compendious form so much of
the allegory of the seventh book as lies without the cave,
together with its interpretation.

While however the scene within the cave is not represented
in the dvadoyia of the sixth book, this part of the allegory with
its interpretation may be expressed in a similar dvadoyia,
namely, as images (in this case, shadows of statuettes) are to
things (in this case, statuettes) so are particulars as appre-

hended by the senses to particulars as they are in themselves. —

Indeed we have been told at 510 A that eéxdves include rpétov

pev oxids (i.e. the shadows in the cave) émevta ta év Tots Boat

davracpara (i.e. the reflections outside the cave), and that to
the second term of the dvadoyia belong both foa Kal wav Td
dutevtov (i.e. avrad outside the cave) and 70 oxevaeTov yévos
(i.e. the avdpiavtes x.7.X, in the cave) ; so that we are prepared
| images of things i

es egal 1 WED
as the words ‘images’ and ‘things’ are of no “ metaphysical
importance,” may be employed to illustrate the difference
between the two sorts of sensible as well as that between the
two sorts of intelligible’.

for the double occurrence of the ratio

images of things.
a 1
things imaged
Aoyla of the vontéy given us in the sixth book and the ava-
Aoyia of the do€acrdv which I have just constructed, we are
now in a position to frame an dvadoyia in which all four terms,

Moreover as the ratio s common to the ava-

1 TE the ratio reflections of things our former statements [at the end of

things
without the cave is equal to the ratio
shadows of statuettes of things within
statuettes of things
the cave, seems to be intimated at
517 a, Tavrny rtoivuv, qv 8 eye, THv
elxova, © ite TAavxwv, mpoocamréov
"racav Tos éumporbey evyouevois, THY
pev Sv dWews pawouevnv eSpav rH Tov
Secuwrnplov olknce. apouootvra, To dé
rou mupos év arp pws TH TOU Alou
duvaue. “You must combine the
whole of this image [of the cave] with

the sixth book], paralleling the visible
region [outside the cave] with the
prison-house [the inside of the cave],
and the firelight in the prison-house
with the sun.” The ordinary rendering
of adouowdtvra, “comparing,” “liken-
ing,” as if 7 60 dpews pawopuévn edpa
were the interpretation of 7 Tov decuw-
rnplov olknais, and 7 Tov Alou divapyus
the interpretation of 7d Tov mupds év
airy pas, treats parts of the imagery
as parts of the interpretation,

¥ at ,
Ls ee :

a ee

PLATO’S REPUBLIC VI. I4I

_ being derived from the interpretation of the allegory, shall be

of “metaphysical importance.” This third avadoy/a will run
thus: particulars as apprehended by the senses are to particu-
lars as they are (or rather, as they become) in themselves, as
the objects of the inferior intellectual method, i.e. Adyou, are to
the objects of the superior intellectual method, 1.e. ideas.

Thus the original avandoyia, as I interpret it, though not
coextensive with the allegory, is perfectly consistent with it.
It is in fact its foundation, what is most important being set
forth in advance in the sixth book and afterwards repeated
with additions in the seventh. At the end of the sixth book
we are told that the inferior vonrdv is an image of the superior
vontov: at the beginning of the seventh we are told (1) that
the inferior vonrév is an image of the superior vontor, and (2)
that the inferior do£acrép is an image of the superior do£acTov.
It is then in the seventh book, not in the sixth, that we find a
division of the universe, and this division is neither quadripar-
tite nor tripartite, but bipartite’, the two parts being yuyvo-

1 At this point it may be asked—
But where do we find a place in this

scheme for the pa@nuarcxéd mentioned

by Aristotle, which are intermediate
between aic@yrd and eldn, differing
from the aic@nrd because unlike aicOnrd
they are eternal and immutable, and
differing from the eléos which is single
because they are plural? There is no
place for these pa@nuarixd. Plato, as
I understand him, is here concerned,
not with pabyuarced as opposed to
other vonrd, but with pa@nuarixd as
types of vonrd. Hence even if we sup-
pose that when he wrote the republic
Plato had learnt to distinguish in the
case of pabnuarixd (in addition to the
idea and the Aéyos which are single)
two sorts of odd, i.e. the triangles
which the geometer sees and the tri-
angles of which he thinks, the re-
cognition of this distinction is in this
place impossible because it is peculiar
to wa@nuarixa. But I cannot think

that Platowould have taken pwadnuarikd
as types of voyra in general, if when
he wrote the passage before us this
refinement had already suggested it-
self to him. In short, the passage in
the republic, making no distinction
between paénuarixa and other voyra,
recognizes (to take a particular ex-
ample) (1) the (single) idea of triangle
as it is, 510 p, (2) the (single) general
notion of triangle implied in the
geometer’s definition, (3) the plurality
of particular triangles as they are, (4)
the plurality of particular triangles
as they are apprehended by sight.
The padnuarixa mentioned by Aris-
totle, i.e. the plurality of particular
non-sensible triangles which particular
sensible triangles suggest to the mind
of the geometer, would have to be in-
terpolated, if anywhere, between (2)
and (3): but it is by no means clear
that Plato entertained the two doc-
trines simultaneously.

°

142

peva and dvtws dvra, which stand second and fourth respec-
tively in my third dvadoyia, while its first and third terms are
respectively the ysyvoweva of the second term and the dvtas
dvra of the fourth, as they are respectively apprehended by us
in aicOyoes and Adyou'. But though there are only two sorts

THE JOURNAL OF PHILOLOGY. ee

of existence, the allegory of the cave, which has for its declared
purpose the representation of our nature waidelas te mépt kab

aratsevolas, distinguishes four stages in the progress from
First, the uneducated man takes his —
sensations for objective realities 515 B; next, he becomes aware —

ignorance to knowledge.

of their subjectivity 515 p; thirdly, he studies the one andthe a

many in the so-called sciences or arts which compose the mpo-
matoeia 525 A 533 D 536 D; and lastly, he will, it is hoped, £
attain to dialectic, which is a ‘coping stone’ to his former

acquirements 533 ¢ 534 E%,

_.1 “Tn logischem und ontologischem
Betracht aber ist die Idee das Object
des Begriffs. Wie durch die Einzel-
vorstellung das LEinzelobject erkannt
wird, so wird durch den Begriff die
Idee erkannt. Die Idee ist nicht das
den vielen einander gleichartigen Ein-
zelobjecten innewohnende Wesen als
solches, sondern das als in seiner Art
vollkommen, unverinderlich, einheit-
lich und selbststiindig oder an und fiir
sich existirend vorgestellte Wesen der
einander gleichartigen Einzelobjecte
(die in den Umfang des Begriffs fallen,
durch den eben diese Idee gedacht
wird).” | Ueberweg Grundriss der
Gesch. der Ph, 1 125. Apparently
these words do not refer to the. passage
before us, as at p. 182 Ueberweg says
expressly that the four objects re-
cognized in republic 509 ff. 533 f. are
idéar wabnuarika odpara elkoves.

2 In tabulating and discussing the
imagery of the cave I have taken ac-
count only of ‘shadows of statuettes
of things,’ ‘statuettes of things,’ ‘re-
flections of things,’ ‘things them-

selves,’ neglecting the subdivision of

‘things’ into ‘things,’ ‘the moon and
the stars seen at night,’ and ‘the sun.’
All these stand for real existences or 3 ,
ideas. ‘The sun’ is plainly the avrd
aya@ov. ‘The moon and stars’ would |
seem to be the ideas of Sixaov, caddy,
and perhaps all the ideas which in the |
republic bear indifferently abstract and
general names. (Cf. Parmenid. 130 B.) - 3
(The phrase repli trav rod dixalov cKxedy 4
9 ayadparwv ov ai oxial 517 2B, ‘about
particular rights as men conceive them
or as they are,’ implies that a place
must be found in the allegory for the %
avro dikaov at any rate; while the re- .
peated mention of dikacov and cadov in
company with dyafdv prepares us to
find them placed second in the list.)
‘Things’ are the rest of the ideas, in- =
cluding dv@pwro 5164 and f¢@a re kal -
guTd 538248. Plato is careful to say
534 = that the names which he gives to,
the waOjuara év TH WuxD yryvoueva ; = 1%
are not important. This is, I take it, .
an apology for a slight inconsistencyin *
the use of them, ‘conjecture’ (elkacia) ==
and ‘belief’ (wxioris) being assigned at
5ll = to ‘images’ and ‘things,’ and

PLATO'S REPUBLIC VI. 143

§3 THE Two METHODS.

Having thus shown oF tried to show that the allegory of
the cave agrees in all respects, as I conceive it should do,
y with the interpretation of 509 DE proposed in § 1 of this
r paper, I now resume the argument of the sixth book. To

pe, ~ the two segments of the vonTdv, 1.¢. (according to ™Y view)
ry general notions and ideas, two methods of investigation are
, assigned. ‘The one ;3 adopted in the case of To Umd TALS —

ryewpeTpiars TE nay Tals TAVTNS Gdendais TEXVALS B11 B, as

> /
eriaTh pas MeV TOANAKIS qr poo elT 0 pev Sid TO e005, Séovrat Sé

- dvdpmatos adov, évapyerTEpov ev 7) S0&Ns, dpvdpoTepou 8é 7
éricTHuns 533 D5 the other, the method which will presum-
ably be adopted by the dialectician of the future, is scien-
tific in a higher sense 534 cD. Plato's account of them 1S,

briefly, as follows :-—

(1) The arithmetician and the geometer start from hypo-
a theses (e.g. odd and even, the geometrical figures, three kinds
3 of angle),—which, not being justified by ascent to an appx
to the last retain their hypothetical character,—and thence
ie descend by mutual agreement 0 the desired conclusions.
| Furthermore, to aid them in their investigation of TO TETPAYO-*/
yov avTo, %) SuapETpos avTy, and the like, they employ models
and diagrams which belong to the visible world, and are them-
selves imaged in shadows and reflections. ‘Thus the objects of
the arithmetician, the geometer, and the man of science 1
general are intelligibles, but intelligibles investigated by means

of hypotheses, with the aid of visibles.

(2) The dialectician, like the mathematician, starts from
hypotheses, but, unlike the mathematician, does not rest con-
tent with them. Hence, instead of immediately descending to

at 5345 (presumably) to ‘the par- right in my explanation of the phrase
ticular as it is apprehended by the ‘reflections of iSéar,’ Sudv ove is clearly
senses’ and ‘the particular as it is’ the right word to describe the cor-
respectively. Tt will be seen that these responding radnua: of Theaetet. 189 E,
terms are more appropriate in the sophist 263 E.

latter use than in the former. If I am

144 THE JOURNAL OF PHILOLOGY.

conclusions, he uses his hypotheses as stepping-stones by which
to ascend to the principle of all things. Having thus reached
that which is not hypothetical in the apy? tod zavrds, he is
in a position to descend, without recurrence to sensibles, from
idea to idea, and so to the conclusion sought. .

Let us now examine the description here given of the
former of the two methods, bearing in mind that, if my in-
terpretation of ‘the line’ and ‘the cave’ is correct, this method,
though here described with special reference to arithmetic and
geometry, should be applicable whenever an idea is studied in
its reflection, the corresponding Adyos, and that the characteris-
tics here mentioned—(1) use of vobécers which never cease
to be hypothetical, and (2) dependence upon sensible images—
should be characteristics, not of mathematical processes only,
but generally of the processes by which Adyou are investi-
gated.

Let us in the first place endeavour to ascertain what Plato

means when he says that ‘mathematicians suppose (vzori- °

Oevrat) the odd and the even, the geometrical figures, three
kinds of angle, &c, assuming them to be obvious to all
and declining to give any account of them.’ His meaning
must be that the geometer starts from such propositions as
‘There may be such a thing as length without breadth, hence-
forward called a line,’ but does not show, or even attempt to
show, that there is such a thing. If he could prove that there
is such a thing, this which is now a voOects, i.e. an apyn
avatrobetxtos, would become an apyn proper. Now according
to Plato there is in the ideal world an édvtws ov answering to

every abstraction. The geometer’s definition is therefore hypo- |

thetical in the sense that it has not been shown to be a correct
and complete account of the idea. Similarly, I conceive, every
Adyos is a UmroOects so long as it has not been shown to be
a correct and complete account of the appropriate idea. When-
ever a Novos can be shown to be a correct and complete account
of the appropriate idea, it will be no longer an v2réGeoxs, it will
become an apy7. .
Next, what are ‘the visibles used as images’ of which the

mathematician’s models and diagrams are typical? They must .

+c, hoeiatiaad yt ae > a
— a NM
= - 7

PLATO'S REPUBLIC JVI. 145

be, I think, the particulars or ‘many’, from which in virtue of
their participation in the idea we derive that imperfect know-
ledge of the idea which is expressed in the Xdyos. So long as
the man of science has not got a firm footing in the world
of ideas, he cannot get clear of the visibles from which the
Aoyos is obtained. 7

The inferior method then starts from Adyot, which (1) are
hypothetical in the sense that they have not been shown to be
correct and complete accounts of ideas, and (2) for that reason
are still dependent upon the particulars or ‘many’ from which
they were originally derived. It is the method pursued by
Socrates when he wishes to ascertain whether a certain person
or a certain thing is just, and by Plato when he inquires
whether the sophist, the statesman, and the philosopher are’
identical. The appeal is in both cases to a Adyos, and the
Aoyos, though perhaps in the one case a more correct and
complete account of the zoAda than in the other, is in neither
case shown to be a correct and complete account of the &’.
Plato wants something more than this, and accordingly tries to
devise a way of converting Adyou which are vrrofécers into Adyou
which are apyai,—doyou which being obtained through particu-
lars are imperfect representations of ideas into Adyos which are
proved to be perfect representations of ideas,—so as to eliminate
at once both the defects of the inferior method. He conceives
that this may be done, if, instead of descending from a hypo-
thetical or unproved Adyos to a conclusion, we ascend from one
hypothetical or unproved Adyos to another, until at last, after
repeated trials, the scale of hypothetical or unproved )dyoe
converges and culminates in the apyn tod tayvTos, i.e. the
avTo aya0ov, éréxewva THs ovacias, from which the ideas derive

1 It is worth while to note that in
Xenophon’s memorabilia tv 6 § 13 the
word troeors is used for ‘the general
definition’ of Socrates: émi riv brobecw
éraviyyey dv mavra Tov Néyov.. Again in
§ 15 we have sentences which seem to
be echoed in the Phaedo 100 p& 101D
105 B: dmére dé abrés Te TH ASyw SreElor,
5a rev waiora dmoroyoupévwv Eropevero,

Journal of Philology. vow. x.

vouitwy Tavrny [rhv] doparecav elvac
AOyou. Tovyapodv word padora wv eye
olda, OTe Aéyor, To’s akovovTas dMoNo-
youvras mapeixev. pn 6é kal “Ounpov 7B
"Odvoce dvabeivac 7d Goparyn propa
elvat, ws ixavdy avrov dvra bia Taw So-
KovvTwy Tots avOpdmos ayew Tovds dd-
“yous.

10

146 THE JOURNAL OF PHILOLOGY.

their being. If in this way we can pass from unproved general
notions, reflections of ideas, to the good, so that we may now
say, not only that the good causes existences, i.e. ideas, to be
what they are, but also that the good causes existences to be
what we conceive them, we may infer, he thinks, that our Adyor,
hitherto provisional, are adequate representations of dévTas
évta. Having thus bridged the gulf between the lower and

the higher vonrov, between Adyou and e/dy, by showing that
" certain Adyou accurately represent e/Sn, we shall be able to
descend in the line of the eién without recurring to the ‘many’
| particulars from which we originally started.

—_

In other words, Plato conceives that, whenever we can draw
up a scheme of vzrofécecs culminating in the ayaGov, so as to
show that the supposed system of dvtws évta is the best which
intelligence working to an end could devise, we may be sure
that our Adyou, though originally derived from the inspection
of particulars, are accurate representations of ideas. The
moment we pass from viroféces to the dyadv, our Adyor will
thereby receive the attestation which they have hitherto lacked
‘and will be converted from vzrobéceis into dpyai, whence we
may descend to conclusions (reAXevtai) as much more certain
than the tedXevra/ of the geometer as certified apyai are more
certain than uncertified tirobécers. The dyador is therefore the
source of all knowledge, just as the sun is the source of all vision
506 A 508 D sqq. Plato does not indeed pretend either to
have drawn up the required scheme of existences, or to be able
to explain the passage from vaobécess to the ayabov. It is not
for him, but for the trained dialectician of the future, to explore
the paxpotépa repiodos 504 A sqq. 506 E. To use the
imagery of the cave, the Heraclitean Cratylus has released
him from his bonds, turned him round, and convinced him
that he has hitherto seen only the shadows of statuettes ; fur-
ther, Socrates. has dragged him up the steep and: rugged
ascent, and taught him to study the reflections of men and
things, of stars, and of the sun; but he has not learnt to regard
with unreverted eye things themselves,.the moon and the
stars, and the sun in all its splendour. In short, he says of
himself what Cowley said of Bacon:

PLATO’S REPUBLIC VI, 147

‘‘The barren Wilderness he past,
Did on the very Border stand
Of the blest promis’d Land,

~ And from the Mountain’s Top of his exalted Wit,
Saw it himself, and shew’d us it,

But Life did never to one Man allow

Time to discover Worlds, and conquer too.”

But, it may be asked, is not the view here taken of the
method to be adopted in the investigation of ideas inconsistent
with the doctrine of dvaurnots, which in some other dialogues
is prominent, and seems to be referred to in republic 621 A?
If the sight of particulars reminds us of the idea which the
soul has known in a previous existence, is not the idea directly
apprehended ? must not the superior method be accounted
a superfluity? The inconsistency is, I think, only apparent.
We may recollect the idea well enough to say that the particu-
lar falls short of it, and yet be unable to form an adequate
notion of the idea recollected, just as (to use an image of
Plato’s) we may perceive that a portrait of Simmias is un-
satisfactory without being able to paint a more perfect likeness
or even to recal his features precisely and correctly. Thus the

doctrine of avayvnows leaves ample room for a theory of the

investigation of the ideas recollected. In fact, avauvnots
assures us that there are ideas to be known, an assumption
which was made at the outset of the passage before us: but it
does not give us correct and complete knowledge of ideas, still
less does it assure us that we have obtained such knowledge.

At this point it is necessary to refer to the well-known
passage of the Phaedo 100 A sq., though it is not without
hesitation that I offer the following summary of it. Finding
himself unsuccessful in. the attempt to trace things to their
cause, the good, the Platonic Socrates proceeded to deal with
causes as he was in the habit of doing with other matters.
His rule was to assume, i.e. to accept without proof, that Novos
which seemed to him most secure, and to account that true
which agreed with it, that untrue which did not. On this
principle he proposes in the present case to justify his belief in
the indestructibility of the soul by showing that it is in accord
with a Adyos aodadys which is no novelty. What in this case

148 THE JOURNAL OF PHILOLOGY.

is assumed without proof is apparently the correspondence of
the idea in general to Socrates’s conception of it. Now Socrates
conceives things to be caused by participation in the idea. On
the strength of this doadys Adyos he proposes, when asked
the cause of anything, to allege participation in an appropriate

idea, neglecting altogether those inconsistent causes which are
alleged by more ingenious thinkers.

If however his original

vTo0ects is disputed, he will justify it by a higher assumption,

and so proceed until he reaches (xavov te’.

1 ras 6€ oxloes tatras Kal mpog-
Oéces kal Tas dANas Tas Tora’Tas Kop-
Yelas éyms dv xaipew, rapels dmoxpivac bat
Tots ceavTov comwrépos* ad dé dediws av
Td Neyduevov THY cavTOD oKiav Kal TH
dewplay, éxduevos éxelvov Tod aopadots

’ an e ee 3 at
THs vmobécews, ovTws dmoxplyao dy.

[ed dé ris avens THs vmrobécews exorTO,
xalpev égns dv kal ovK daoKpiva.o, ews

dy Ta am’ éxeivns opunOervra oxéyaco, et

got dddAndAos Evudwved 7 dStadwve?.]
éreton dé éxeivns airs bor ce diddvac
Nyov, woattws dv didolns, d\Anv ad
umd0eow vmobéuevos, {Tis TwY dvwber
Bertiorn daivorro, Ews éml te ikavov
€2Oos, dua dé ok dv pipos worrep oi
dvridoyxol mepl te Ths apxns Siareys-
bevos Kal Tov €& Exeivns wpunuévew, elirep
BovdAouw Te Tay é6vTwy edpeiv, 101 c—rz.
The sentence which I have bracketed
seems to me in every way suspicious,
That (1) éyo.ro cannot stand in the

’ sense of ‘oppugnare,’ ‘aggredi,’ espe-
p 2 5) L

cially as éxéuevos has just been used
(100 p 101 p) in its proper meaning,
is remarked by Madvig, who would
therefore read égoiro. But this con-
jecture leaves unanswered the further
objection, that, as the text stands, (2)
there is no antithesis (as there plainly
should be) between ef dé tes adrHs THs
Urobécews éxorro (Or Epouro) and ézerd7)
dé éxelvns airns Séou ce Sddvac Aoyov.
Ast endeavours to meet both objections
by reading ef dé ris GAdns vrodécews
éxo.ro, to which correction it may be

objected that elsewhere throughout the
passage Socrates persistently uses the
word védects only when he is speaking
of his own provisional dpx7%,, and would

hardly use a term so characteristic of

his own doctrine in speaking of his
opponent’s more pretentious principle.
Moreover (3) the injunction to consider

Ta am éxelyns dpunbévta cxévaocOa ef .

got GAAjAas Evuhwve? 7 Stadwvei finds
no countenance in the summary state-
ment add’ otv 5h Ta’Ty ye Wpunoa, Kal
vmobéuevos éExdorore hoyov dv dv Kplyw
éppwuevéotarov elvat, d wev dy wor SoxH
TOUTW ~vughwreiv, TLOnue ws adnO7H dvTa,
kal mept airlas xal mepl twv Gd\dX\wv
amdvrov, d 5 dv un, ws odk ann 100 A,
and involves a violation of the precept
dua dé ovx dv pipots werep oi dvTidoyiKol
mepl te THs apxns Siadreyouevos kal Tar
€& éxelyns wpunuévwv 101 £. Indeed it
is not easy to say what is meant by
‘the mutual agreement or disagree-
ment of 7a amd THs vrobcews dpun-
6évra,’ a phrase which looks very like a
perversion of the concluding words of
the sentence quoted above from 100 a,
The sentence seems to me to be the
work of an interpolator, perhaps the
same who in 72D has added the words
kal rats wév 7’ ayabais duewov elvac rats
dé Kkaxais xdxeov. (See Bonitz Pla-
tonische Studien p. 283.) It will be
observed that in both places phrases
which occur in the context are echoed
in a false sense,

In conclusion he

‘
.
:
oe

. \ ©
wae Ne

es

ee

———— le

- PLATO'S REPUBLIC VI. 149

‘warns us not to confound the study of the apy with the

investigation of its consequences.
As I read it, the whole of the passage summarized is con-
cerned with the ‘inferior method’ of the republic. It has indeed

been thought that the ‘«avoy of 101 D is the avuTroberov of

republic 511 B, and therefore that the concluding sentences
refer to the ‘superior method.’ It would seem however that a
reference to the superior method, which at 99 c Socrates has
renounced as beyond his powers, can find no place here, where
he is describing bis devTepos mods. It remains then to under-

stand by the (cavov any more general rvo@eous which gains the

assent of the opponent. Thus, while the passage in the republic
treats of the use of the inferior method in exposition, the. passage
in the Phaedo (as the language throughout shows) deals with
its application to debate, and accordingly provision is now made

-in case the vroears offered by the one disputant should not
-be approved by the other (€zred2) éxelvns avtis déo4 ce didovat

Aoyov). In such a case we are to take sume more general vzro-

ects, and so on, until we arrive at one which the opponent is

willing to accept. As this vzoGeovs, not having been certified
by the ascent from it to the ayaov, is no more an apyy (in the
strict sense of the word) than the v7o@eccs originally offered
and refused, the reasoning which is based upon it is necessarily
of the inferior kind.

In fine, the two methods of the Phaedo—that which Socra~
tes has abandoned, not because he has lost his belief in it, but
because he does not know how to apply it, and that which he
adopts as the least unsatisfactory substitute—are respectively
identical with the two methods described in the republic, the
method of the dialectician and the method of the so-called
man of science. In both dialogues the superior method is an
unrealized aspiration; though it may perhaps be thought that
the republic expresses a more hopeful mood than the Phaedo,
Both dialogues are themselves examples of the inferior method,
the uao0ecis being the theory of ideas; but it would seem that,
in accordance with the precept da dé ov« av dupous batrep of
avTiNoylKol mepi Te THS apyns Svareyopmevos Kal Tadv €F éxei-
yns wppnucvov, elmep Bovroro TL Tov OvTwy evpety, in the

a

150 THE JOURNAL OF PHILOLOGY.

republic the wtdOects itself is under discussion, while in the
Phaedo Socrates traces tov €& éxeivns Opyunpévor TU.

The passage in the republic upon which I have been com-
menting deserves, I think, more attention than it has received
from recent commentators and historians, if only because it is
unusually precise and dogmatic. But when Plato is precise
and dogmatic, he generally contrives to introduce an element of
obscurity into the exposition. In this instance the indirect
description of the third segment is just such an element, made
more perplexing for us by the apparent parallelism of metaph.
16§ 5. I venture to think however that, if we (1) under-
stand the original avadoy/a, not as a quadripartition of the
universe, but as an ekav Kxat dvadroyiav in which the first
and second terms are introduced only to explain the rela-
tion between the third and fourth, (2) equate the four seg-
ments, not with the whole imagery of the succeeding allegory,
but with the imagery and the interpretation of the more impor-
tant part of it, and (3) take the third segment to stand for
those universals which were the foundation of the Socratie
dialectic, we obtain a consistent interpretation both of ‘the
line’ and of ‘the cave.’ Plato, as I read him, gives us here just
what was wanted to complete the outlines of the theory of ideas,
—a comparison between his own position and his master’s.
First, he shows that the theory of ideas is founded on the
Socratic doctrine of universals, which is incorporated in it, not
- superseded. Secondly, he marks the deficiencies of the Socratic
logic, and of all inquiry conducted on merely Socratic prin-
ciples. Thirdly, while indicating a hope that the theory of
ideas, teleologically interpreted, may one day become the
basis of a new and powerful logic, he admits unreservedly
that his own logic is not at present superior in kind to that
of Socrates.

HENRY JACKSON,

+]
7
d

PR wn» » be
> h * " 4 74, Oy”

AESCH. AG. 115 —120.

a a ¢
olwvav Bactreds Baciredar vedv, 6 KeXatwos, 6 7 éEorw apyas,
, \ ;
ghavévtes ixtap weraOpwr, yepos ex Sopimarrov,
TapmpéTrols ev Epa,
Bockopevot Aayivav Epixvpova héppate yévvay,
/ , ld
BraBévra roc biwv Spopor.

It appears from the commentaries that the concluding
words of this citation are very difficult of interpretation. Her-
mann, indeed, would have us suppose that the AolcOtos Spdpos
of the unhappy hare “ portended the capture of Troy just when
it thought itself safe under the feigned retirement of the
Grecian fleet.”

Sewva pel ody Sewa Tepaler codds oiwvobéras.

I am glad to see that upon this curious exposition Professor
Paley, who cites it, observes a significant silence. The latest
German commentary I have seen, that of Enger, is as far gone
in another direction, enquiring by what interpretation or correc-
tion Aolc@io1 Spdpot may be brought to signify the birth of the
hare’s offspring, of which they were hindered by the devouring
of the mother—an enquiry hard to answer.

The word Aotc@iwy, in which all the difficulty lies, seems to
have caused some misapprehension of the rest of the phrase.
Professor Paley, in his translation of 1864, gives as the literal
rendering “stopped from future courses,” and Donaldson in the
New Cratylus, § 454, “stopped from running any more races.”
Both. these versions do some violence to the sense of NolcAos,
in which respect Prof. Kennedy’s “caught ere its closing race

152 THE JOURNAL OF PHILOLOGY.

was over” is more accurate. But apart from that, we may
doubt whether any one, not pressed with the necessity of getting
last courses into the context, would have translated BraBévra

dpopwv ‘stopped from running’ instead of ‘ hindered in running.’ —

ArXarrew, as Eustathius says in the notice cited by Donaldson
himself, is “ properly to éuzrod(feww tov tpéyovta, to hinder or
impede a runner,” and so it should be turned in the two cases
given from Sophokles, £7. 696, and Ant. 455, ef dé tis Deady
Brarrot, diyot Tav Yo KaKds Tov Kpelacova, though Donaldson,
still haunted by the AoweOiwy of the Agamemnon, writes stop in
both of them, to the manifest detriment of the meaning. Nor
is BrAaBévta Spdpeav in this sense at all alien to the passage
before us. The eagles’ quarry was hindered in running by its
pregnant condition, and as it was the cruelty of its capture
under these circumstances which excited the anger of Artemis,
_ this interpretation at once gives the line force and point, which
it is difficult to find in the mere statement that when caught
the hare could not run any more.

Even so however AotoOiwy is a difficulty. An epithet thus
inserted in a compact phrase like PraBévta Spoporv ought to
be closely coherent with it in meaning. Hvindered in running
is sense, hindered in last running is scarcely sense. It is worth
while therefore to ask whether Aovc Aw is likely to be an error.

No symbols are more frequently confused with others than_

those which compose the first syllable of XotcAi:wv—A with A,
ot with v. Each of these permutations is common enough to
give a palzographic basis of possibility to the correction,

BraBévra Suc bowr Spopor,

that is, literally, hindered in its difficult running; compare Oéw,
600s, SUamA00s, SUTTVOOS.

A. W. VERRALL.

¢
Filan aes
._.

THE JOURNAL

OF

: PHILOLOGY.

ie . THILO’S SERVIUS".

THE second volume of this work contains the Servian com-
mentary on the fourth and fifth books of the Aeneid, with an
elaborate preface in which the editor sets forth at length his
views on the work, the manuscripts on which its text is based,
its authorities, its date, and its general character. The appear-
-ance of this preface, while it makes the volume doubly welcome,
also makes it possible for a reviewer to criticize the edition, for
the first time, as a whole.

There are two recensions of the Servian commentary’, one
of which contains many more notes than the other. These
te notes are sometimes supplementary to those of the shorter
version, sometimes repetitions of them, sometimes inconsistent
with them. The fuller recension is generally known as the
Servius of Daniel, from the fact that the different manuscripts
in which it is contained were first used by Peter Daniel, who
edited it from these manuscripts in 1600. An account of the

1 Servi Grammatici qui ferunturin the fourth edition of Conington’s com-
Vergilium commentarii. Recensuerunt mentary were published before the
Georgius Thilo et Hermannus Hagen, appearance of this volume.

Vol. 1. Fasc. 1. Lipsiae, mpcccuxxxI. 2 In this paper the supposed inter-

The essays on the ancient Vergilian polations in Servius are enclosed in
critics and commentators prefixed to __ brackets.

Journal of Philology. vou.x. 1}

154 THE JOURNAL OF PHILOLOGY.

manuscripts used by Daniel, all or nearly all of which are for-
tunately still in existence, is given both by Thilo and Thomas,
of whose excellent essay on Servius I have spoken elsewhere.
The two scholars are in substantial accord on all points but
one. The additional notes on the first and second Aeneids are

contained in a manuscript now at Cassel. A Fulda manuscript, —

containing additional notes on the same two books, was collated
for Daniel by Scioppius. The readings of this codex as given
by Daniel do not always coincide with those of the Cassel MS.
Thomas, like Schubart before him, doubts whether the Ful-
densis of Daniel is the same as the now surviving Cassellanus.
Thilo maintains their identity in a very interesting and ingeni-
ous argument; but until Thomas (whom Thilo treats in this
matter with scant courtesy) has replied to him, it cannot be
said that the last word has been spoken on the subject.

‘The first question to be decided with regard to the Servian
commentary affects the character of the fuller version. Is the
fuller version the true Servius, while the vulgate (as with
Thomas and Thilo we may call it) is an abridgment? Or is
the vulgate the genuine Servius, while the additional notes are
interpolations? And if interpolations, by whom and when were
they added to the genuine commentary ?

The view that the fuller recension represents the genuine
commentary was maintained by Joseph Scaliger, and has been
recently upheld, though in a different form, by Ribbeck. Mas-
vic, on the other hand, and Ottfried Miiller, contended for the

non-Servian origin of the additional notes, and Thomas and

Thilo agree with them.

In the essays prefixed to Conington’s first volume (ed. 4) I
ventured to express a doubt whether this latter view is correct ;
and a further examination of the evidence has led me to form
the opinion that the additional notes have, on the whole, as
good a right to bear the name of Servius as the vulgate. I
doubt whether either recension of the Servian commentary can
claim to come entirely from the hand of Servius, and to repre-
sent all that he had to say upon his author. But as Servius

was celebrated as a very learned lecturer on Vergil, I suspect.

that the commentaries now bearing his name represent, in a

a
:
4

THILO’S SERVIUS. 155

fuller and a shorter shape respectively, notes which were at
various times given by him in his lectures, and which were
edited without any serious attempt to present a properly
homogeneous whole. ,

It used to be supposed that the additional notes were con-
demned absolutely by the words ‘ut dixit Servius, which were
thought to occur in one of them on Eclogue 9.1. But Thomas
informs us in the supplement to his essay that these words are
not really there. The only important piece of external evidence
which could affect the question is therefore gone, and we are
left entirely to considerations drawn from the character of the
notes themselves.

The chief arguments relied upon by Thomas and Thilo as
shewing that these additional notes did not form part of the
original commentary of Servius, are, so far as I can ascertain,
the following :

(1) The additional notes fall into two classes ; one of which
includes comments which are really supplementary to the vul-
gate, while the other consists of notes which, although they
have been inserted in the text in such a way as to present a
specious appearance of coherence with it, are really out of place,
and interrupt the sequence of ideas. In many cases the addi-
tion is made with the aid of conjunctions such as ergo, nam,
enim, quod, quia and the like, which on examination are found
to be out of place. Thilo notices in particular that the word
sane is used in an irrational way in the additional notes. In
some passages again the additional note has had the effect of
mutilating the text of the vulgate.

(2) The additional notes quote a great variety of opinions
upon disputed points without deciding upon any one in par-
ticular, while the vulgate usually does so only to adopt one in
preference to the others. |

(3) The vulgate, when referring to an opinion previously
expressed, or an observation previously made, always uses the
words ut supra diaimus, while the additional notes speak im-
personally, ut supra dictum est.

(4) Where the manuscripts of the vulgate mention the
names of Donatus and Urbanus, the manuscripts containing

11—2

156 THE JOURNAL OF PHILOLOGY.

the additional notes omit these names. This is however the
case only with Donatus and Urbanus, not with Probus, Asper,
or any other commentator mentioned in the vulgate.

(5) The compiler of the additional scholia assumes that

the commentary on the Eclogues and Georgics preceded that

on the Aeneid, while the vulgate assumes the reverse order.
It may be observed by the way that the commentary of Aelius
Donatus must have followed the same order as that observed in
the additional scholia. :

(6) In some cases the author of the additional scholia
seems to have followed a different text from that followed by
the author of the vulgate. . }

(7) The additional notes containing quotations from Sallust
are probably to be attributed to Asper, others to Probus, others
to Aelius Donatus. Many agree with Vergilian notes in Mac-
robius, but it cannot’ be shewn that they are borrowed from the
Saturnalia.

(S) The character of the vulgate differs from that of the
additional notes. The latter sometimes exhibit a deeper learn-
ing than the vulgate, while at the same time they are often
expressed in worse Latin. The notes on grammar are inferior,
but those on lexicography and interpretation, superior, to those
of the vulgate; and the fables are given, in the additional
notes, in a fuller form.

(9) The question must be answered whether scholars later
than Servius, who seem to have known and used the Servian
commentary, had the vulgate or the fuller version before them.
Little can be made, in this connexion, of Cledonius, Pompeius,
Priscian, the scholia on Lucan.or on Statius: but the first
writer among the mythographi and Isidore in his Origines
evidently borrowed from the shorter Servius. In an immense
number of passages, where there is a verbal correspondence
between the notes of Isidore and those of Servius, Isidore
repeats the note of the vulgate, though he might as easily, had
he had the fuller version before him, have copied from it. In
some cases however it appears as if the compiler of the fuller
commentary had taken his notes from Isidore.

The conclusion which Thilo draws with regard to the com-

r
a
7
;
J
:
:

Py Wee =,
—EE——— Ee ~

~¢ ~ eee FO Se eee =<." =

THILOS SERVIUS. 157

position of the additional scholia is this: that they were
compiled by one writer, who had before him not only the
writings from which extracts were made by Macrobius, but also
the Origines of Isidore; that his date must therefore be later
than that of Isidore (about 570—640), and that from some
slight indications it may be inferred that he was a Christian. |

Before passing on to the more important points involved in
the discussion, I may remark that this last inference is based on
the slightest possible evidence. Thilo appeals to two notes on
A. 4 200 and 301, which he thinks (after Burmann) shew a
Christian tone. The first is as follows: significat sine intermis-
sione fieri sacrificia, ad quem (quae ?), excubare per diem et
noctem necesse sit, ut dicimus quotidie in officio esse ; non ergo
apud quas dii excubant, sed quae diis excubantur. The second
is this: commotis excita sacris: verbo antiquo usum tradunt;
moveri enim sacra dicebantur, cum sollemnibus diebus aperie-
bantur templa instaurandi sacrificil causa; cuius rei Plautus in
Pseudolo meminit, ‘scis tu profecto, mea si commovissem sacra,
quo pacto et quantas soleam turbas dare.’ Hoc vulgo apertiones
appellant.

I wish that Thilo had pointed out explicitly what mark of
Christian authorship he finds in these notes. His other argu-
ment, that the compiler of the additional scholia often speaks of
the customs of the Roman ritual as things of the past, need
prove no more than that his notes were written after 382 A.D.

Let us now proceed to examine the arguments for the non-
Servian origin of the additional scholia in the order in which
(nearly following Thilo) I have stated them.

(1) There can be no doubt that many of these notes
are repetitions of what has been said in the vulgate, and that
many again interrupt and interfere with the coherence of the
vulgate. So much is this the case that Thilo sometimes trans-
poses them; a proceeding which, however tempting, is in my
opinion questionable in a case of this kind. If we are to form
an opinion on the character of a supposed interpolation, it is
important that it should be exhibited to the eye of the reader,
so far as possible, in the form in which the manuscripts present
it. When these additional notes are embedded in the text of

158

THE JOURNAL OF PHILOLOGY.

the vulgate, to take them out of their place and print them sepa-
rately is to assume the point which has to be proved, that they
are essentially heterogeneous to their surrounding. Even where
the sense of the vulgate is unquestionably interfered with by
the interrupting matter, it would, in my opinion, have been safer
to print the text as it appears in the manuscripts, relegating —
conjectural transpositions to a note, than to pursue, as Thilo has
done, the opposite method. Indeed I have found two cases,
and I dare say I might find more, in which I think it doubtful
whether any transposition was required’.

The phenomenon presented by these notes does not differ in
kind from what meets us in the Terentian commentary which
bears the name of Donatus. This work abounds in repetitions ;
a fact which may shew either that its author must have copied,
or dictated to a class, identical notes from two or more older com-
mentaries, or that the commentary is not the work of one
scholar but of two, one of whom subsequently added, without
any regard for symmetry, notes taken from a second work
similar in character to the first. Or again, the same scholar
may have given two or more sets of lectures, the notes of which
partly coincided with and partly differed from each other, and
the two sets of notes may have been carelessly embodied, side
by side, in the commentary bearing his name.

1 A. 1 52 Poetae quidem fingunt
hune regem esse ventorum [Hippotae
sive Iovis sive Neptuni filium. Qui
- cum immineret bellum, quo Tyrrhenus,
Lipari frater, Peloponnesum vastare
proposuisset, missus ab Agamemnone,
ut freta tueretur, pervenit ad Liparum,
qui supra dictas insulas regebat im-
perio, factaque amicitia Cyanam filiam
eius in matrimonium sumpsit et Stron-
gulam insulam in qua maneret acce-
pit. Varro autem dicit hunc insularum
regem fuisse,] ex quarum nebulis et
fumo, &c.

A. 1 145 levat, leves ac navigabiles
facit, ut ‘nostrumque leves quaecumque
laborem.’ [Alibi levat, laxat: ut ‘ at-

que arta levari Vincla iubet Priamus.’
Tridenti autem pro tridente, dativum
pro ablativo. Aperit, ideo quod ha-
renarum congerie impediente prae-
clusae ad navigandum erant. Ceterum
bis idem. Ergo inmisso in eas mari
aptas ad navigandum facit. Sic Sal-
lustius, ‘sed ubi tempore anni mare
classibus patefactum est.’ Temperat,
tranquillum facit. Atque rotis sum-
mas, &c. Bene non moratur in de-
scriptione currus, ut citius liberetur
Aeneas.] At in quinto ubi nullum
periculum est, &c.

This is the order in the Cassel ms,
I am not convinced that any change
is necessary in either case.

THILOS SERVIUS. 159

The fuller version of Servius does not essentially differ in cha-
racter, so far as its repetitions and inconsistencies go, from such
scholia as those of Donatus on Terence. The vulgate of Servius
is indeed on the whole a homogeneous work, which may fairly
be supposed to come from one hand. Yet even the vulgate is
not always consistent with itself, and sometimes gives us notes
which bear the appearance of having been transcribed inde-.
pendently of each other and never harmonized’. Taken by
themselves, these considerations point to the conclusion that
though the fuller version of Servius cannot be called a homo-
geneous work, it has at least as good a right to bear the name
of Servius as the Terentian commentary that of Donatus.
And it must further be observed that, as I hope to shew in a
moment, there are many cases in which the vulgate and the
additional notes are absolutely homogeneous.

The second and third arguments are no doubt of importance
as accentuating the facts sleeaty dwelt upon. It cannot be
denied that there are slight differences of character between
some of the additional notes and those of the vulgate.

(4) Iam unable to see how this fact bears on the question
of the Servian character of the additional notes. Where, in the
vulgate, the names of Donatus and Urbanus are expressly men-
tioned, in the corresponding passages of the fuller version they
are suppressed, and ali, or a similar word, is substituted for
them. This shews that there were at least two recensions of
that part of the commentary which is undoubtedly Servian, but
what has it to do with the character of the supposed inter-
polations ?

(5) This fact again proves no more than that there were
two editions of the Servian commentary, one of which began
with the Eclogues, and the other with the Aeneid. But there
is some probability that this was the case with the vulgate as
well. For in the Harleian manuscript of Servius, my account
of which, written in 18787, has not come under Thilo’s notice,
the Servian memoir of Vergil is prefixed both to the com-

1 See, for instance, p. 5, 1. 9—12 in 2 In the preface to a pamphlet en-
Thilo’s edition; p. 51, 1.3 foll., com- titled ‘ Ancient Lives of Vergil.’ _
pared with p. 76, 1. 17 foll.

160 THE JOURNAL OF PHILOLOGY.

mentary on the Aeneid and (in a shorter form) to that on the
Eclogues. I do not gather from Thilo’s account of his manu-
scripts that this is the case with any other copy of the vulgate ;
but it shews that the commentary on the Eclogues was by some
editors of Servius considered to be at least independent of that
on the Aeneid.

With regard to (6) it must be admitted that the facts
adduced by Thilo make in favour of separating the notes of the
fuller version from those of the vulgate. The same may perhaps
be said of (8), though it might be as reasonably inferred that
so far as the notes on lexicography and interpretation go, the
fuller version represents an older commentary than the vulgate.
No conclusion that seriously affects the question can, so far as
I see, be drawn from (7), for there can be no doubt that notes
of Asper and Probus are embedded in the vulgate as well as in
the additional scholia.

(9) I have not examined the passages which are alleged to
have been borrowed from Servius by the first writer among the
mythographi. But on the ,question of the relation between
Isidore and Servius I am wholly unable to agree with Thilo.
This point is of the utmost importance as bearing on the ques-
tion both of the sources of the vulgate, and of the relation
between the vulgate and the additional notes. Could it be
shewn with certainty that Isidore copied from the vulgate of
Servius, while he was.ignorant of the fuller version, no doubt we
should have a strong argument in favour of supposing the notes
of the latter to be insertions by a later hand. But I think, and will
- endeavour to shew, that Isidore did not copy from the vulgate
of Servius, but that the numerous coincidences between the
vulgate and Isidore are due to community of sources, and also
that a comparison between Isidore and the fuller Servius shews
that many notes in the latter are absolutely homogeneous with
the vulgate, and cannot, therefore, be supposed to be interpo-
lations.

All considerations drawn from external evidence make
strongly against the theory that Isidore borrowed from the vul-
sate of Servius. The Origines of Isidore is a work of reference
arranged under heads on a perfectly intelligible system, and

a

THILOS SERVIUS. 161

bears the plainest marks of having been derived from a work or
works of a similar kind. It is certain that Isidore had access
to the Pratum of Suetonius, and nearly certain that he largely
consulted it; and there is no proof that he did not know the
great work of Verrius Flaccus. At least there is much in
Isidore which must directly or indirectly have come from the
latter. Now it is abundantly plain, and is allowed by Thilo,
that the Pratum of Suetonius was much used by Servius. We
shall therefore be prepared, a priori, to find that Suetonius was
the common authority for many identical notes in Servius and
Isidore. Why indeed should Isidore, with Suetonius or an
abridgment of Suetonius before him, go out of his way to look
for information in Servius? It would be like hunting for a
needle in a bottle of hay. But we can safely leave a@ priori
ground, and give instances of notes taken from Suetonius by
Servius and Isidore alike.

Serv. Ecl. 3 8 Acrqui autem sunt oculorum anguli, secundum
Suetonium Tranquillum in Vitiis Corporalibus.

Isid. 12 1 14 hircus lascivum animal et petulcum...cuius
oculi ob libidinem in transversum aspiciunt, unde et nomen
traxit. Nam hirqui sunt oculorum anguli secundum Suetonium.

Serv. E. 3105 ulna proprie est spatium in quantum utraque
extenditur manus. Dicta ulna dwé tov wrEver, i.e. a brachiis,
unde et AevedrAevos “Hpn dicitur. Licet Suetonius unum
cubitum velit esse tantummodo.

Isid. 11 1 64 ulna secundum quosdam utriusque manus
extensio est, secundum alios cubitus, quod magis verum est,
quia Graece wX<vn cubitus dicitur. |

Serv. A. 7 612 Suetonius in libro De Genere Vestium dicit
tria esse genera trabearum. Unum dis sacratum, quod est tan-
tum de purpura. Aliud regum, quod est purpureum ; habet enim
album aliquid. Tertium augurale, de purpura et cocco mixtum.

Isid. 19 24 8 trabea erat togae species ex purpura et cocco,
qua operti Romanorum reges initio procedebant. Hane primum
Romulus adinvenisse dicitur, ad discretionem regii habitus.

Serv. A. 7 627 secundum Suetonium in libro De Vitiis
Corporalibus arvina est durum pingue, quod est inter cutem et
viscus.

162 THE JOURNAL OF PHILOLOGY.

Isid. 11 1 81 arvina pinguedo cuti adhaerens.

In these cases the reference in Servius proves the Suetonian
origin of the note in Isidore, or makes it highly probable. Had
Isidore been copying from Servius, why should he not have
written out his notes in full and without any variation? But
the very points in which the two writers differ shew in my
opinion that Isidore is abridging the passages in Suetonius from
which Servius is quoting more fully. In the case of the note
on hircus, indeed, the explanation given by Servius of ‘trans-
versa tuentibus hircis’ is quite different from that of Isidore.

Let us now consider some instances where there is a verbal
coincidence between Isidore and the vulgate of Servius.

Servius A. 1 12 urbs dicta ab orbe, quod antiquae civitates
in orbem fiebant, vel ab urvo, parte aratri, quo muri desigen:
bantur.

Isid. 15 2 3 urbs vocata ab orbe, quod antiquae civitates in
orbem fiebant, vel ab urvo parte aratri, quo muri designabantur,
unde est illud ‘optavitque locum regno et concludere sulco.’
Locus enim futurae civitatis sulco designabatur, id est aratro.
Cato: ‘qui urbem, inquit, novam condet, tauro et vacca aret, ubi
araverit, murum faciat, ubi portam vult esse, aratrum sustollat
et portet, et portam vocet.’

If Isidore is here borrowing his first words from Servius, it
is natural to ask how it happens that he does not quote the line
on which Servius is commenting, but a line which does not
occur in Vergil at all; and secondly, what was Isidore’s authority
for the second part of his note, which is so closely connected

with the first that it is natural to suppose that the whole comes

from one source. Was Verrius Flaccus the ultimate authority?
See Fest. 375 s.v. urvat. |

Serv. ib. et eam deleverat Scipio Aemilianus. Quae autem
nunc est postea a Romanis est condita.

Isid. 15 1 30 ex iis profecta Dido in litore Africae urbem
condidit, et Karthadam nominavit, quod Phoenicia lingua ex-
primit; mox sermone verso Karthago est dicta: hane Scipio
delevit. Quae autem nunc est, postea a Romanis condita est.
Karthago autem antea Byrsa, post Tyrus dicta est, deinde
Karthago.

THILOS SERVIUS. 163

In this instance also the words common to Servius and
Isidore occur in Isidore as an integral part of a longer note, and
the supposition that they are taken from the passage in Servius
is unnatural, Nor is there any other note in Servius from
which they could be derived.

Much the same may be said of the following notes :

Serv. A. 1 43 rates, abusive naves: nam proprie rates sunt
conexae invicem trabes.

Isid. 19 1 9 rates et primum et antiquissimum navigil genus
e rudibus tignis asseribusque consertum, ad cuius similitudinem
fabricatae naves ratariae dictae. Nunc iam rates abusive naves:
nam proprie rates sunt conexae invicem trabes.

The ultimate authority for this note may have been Verrius
Flaccus: see Fest. 273 s.v. rates, where much the same informa-
tion is given.

Serv. A. 1 62 foedere, modo lege, alias pace, quae fit inter
dimicantes. Foedus autem dictum vel a fetialibus, id est sacer-:
dotibus per quos fiunt foedera, vel a porca foede, hoc est lapidi-
bus occisa, ut ipse ‘et caesa iungebant foedera porca.’

Isid. 18 1 11 foedus est pax quae fit inter dimicantes, vel a
fide, vel a fetialibus, id est a sacerdotibus dictum. Per ipsos
enim fiebant foedera sicut per saeculares bella’ Alii foedera
putant a porca foede et crudeliter occisa, culus mors optabatur ei
qui a pace resiluisset (?). Vergilius, ‘et caesa iungebant foedera ©
porca.’

Now this note of Isidore bears a much closer resemblance to
a note, compounded partly of the vulgate and partly of a
supposed interpolation, on A. 8 641, where the etymology from
fides is given, and referred to Cicero. So far as it goes therefore,
the note would go to prove that in this case the additional mat-
_ ter in the enlarged Servius is not an interpolation. As to the
authority for the note, it may very well be Suetonius, whose
name is mentioned by Isidore in its near neighbourhood, but
ultimately it comes from Verrius Flaccus; Paul. 84 foedus appel-
latum ab eo quod in paciscendo foedere hostia necaretur. Ver-
gilius: ‘et caesa iungebant foedera porca.’ Vel quia in foedere
interponatur fides.

Serv. A. 1 178. fessus generale est: dicimus enim fessus

164 THE JOURNAL OF PHILOLOGY. |

animo, [id est incertus consilii], ut ‘ter fessus valle resedit, et |

fessus corpore, quod magis est proprium, et fessus rerum a for-
tuna venientium, ut hoc loco. 8 232 ter fessus valle resedit ;
egens consilii. Sallustius; ‘fessus in Pamphyliam se receptat.’
Nam corpore futigatum dicimus, animo vero fessum ; quamyis
haec saepe confundat auctoritas. Here again it seems that the
additional note of the fuller version formed part of the original
comment. Let us now compare Isid. 10 101, who adds some-
thing which is in neither note: fessus quasi fissus, nec iam
integer salute; est autem generale. Dicimus enim fessus
animo, ut ‘ter fessus valle resedit,’ et fessus corpore, quod magis
est proprium, et fessus rerum a casu venientium. Fatigatus,
quasi fato agitatus. |

Serv. A. 1 215 feras dicimus aut quod omni corpore ferun-
tur, aut quod naturali utuntur libertate et pro desiderio suo
. feruntur.

Isid.12 2 2 ferae appellatae eo quod naturali utuntur libertate,
et desiderio suo ferantur. Sunt autem liberae eorum voluntates,
et huc atque illuc vagantur, et quo animus duxerit eo feruntur,

Here it is true that Isidore’s comment. corresponds in gene-
ral drift with the vulgate, to which the fuller version adds a
remark which is not in Isidore: still the wording of the two
notes is so different that it is improbable that one was copied
from the other. The additional note, sane veteres prope omnes
quadrupedes feras dicebant, ut ‘inque feri curvam conpagibus
alvum Contorsit,’ et ‘armentalis equae mammis et lacte ferino,’
should be compared with the Verona scholia on A. 7 489, and
Nonius p. 307.

Sery. A. 4 7 nihil interest, utrum umbram an noctem dicat ;
nox enim umbra terrae est, ut supra (2 251) ‘involvens umbra
magna terramque polumque.’

Isid. 5 31 3 noctem autem fieri dicunt, aut quia longo
itinere lassatur sol, et cum ad ultimum caeli spatium pervenit
elanguescit, ac labefactus efflat suos ignes, aut quia eadem vi

sub terras cogitur, qua super terras pertulit lumen, et sic umbra

terrae noctem facit. Unde Vergilius ‘ruit Oceano Nox’ &ce.
Here surely the agreement between Servius and Isidore is of
the slenderest.

ee ee ee

THILO’S SERVIUS. 165

Serv. A. 4 30 sinus dicimus orbes oculorum, id est palpebras,
quae a palpitatione dictae sunt, nam semper moventur.

Isid. 11 1 39 palpebrae sunt sinus oculorum, a palpitatione
dictae, quia semper moventur. Concurrunt enim invicem, ut
adsiduo motu reficiant obtutum &c.

Here not only does Isidore add something which is not in
Servius, but it is plain that the object of his note is different.
He is defining palpebra, Servius is explaining sinus.

Serv. A. 4 130 cubare exorto, nato Lucifero: nam proprie
iubar Lucifer dicitur, quod iubas lucis effundit ; unde iam quic-
quid splendet iubar dicitur, ut argenti, gmmarum. Est autem
Lucifer interdum Iovis; [nam et antiqui ‘iubar’ quasi ‘iuvar’
dicebant ;] plerumque Veneris stella, unde Veneris dicta est,
ut (8 590) ‘quem Venus ante alios astrorum diligit ignes’
[Alii iubar solem, alii splendorem siderum dicunt].

Isid. 3 70 18 Lucifer dictus eo quod inter omnia sidera plus
lucem ferat; est autem unus e planetis. Hic proprie et iubar
dicitur, eo quod iubas lucis effundat; sed et splendor solis ac
lunae et stellarum wubar vocatur, quod in modum iubae radii
ipsorum extendantur.

Isidore’s note here combines observations which are to be
found in the vulgate and the fuller commentary combined.
There is no ground for supposing that he is borrowing from
Servius, nor need we go far for the common source of the note.
Paulus 104 clearly points to Verrius Flaccus: iubar stella
quam Graeci appellant dwadopov vel Exmepor, hoc est Lucifer,
quod splendor eius diffunditur in modum iubae leonis,

In all these cases, where the words of Isidore and Servius
coincide, Thilo remarks ‘excripsit Isidorus ;’ with what reason
I leave readers to decide. As this is a case where the brick
may be taken as a sample of the house, it is not necessary to
quote any more instances. I will only observe that there are
numberless passages where the correspondence between Isidore
and Servius is only of a general kind, and where Thilo observes
not ‘exscripsit’ but ‘conferatur Isidorus.’ In these passages,
as far as I can see, the only hypothesis which can account for
the correspondence is that of a community of sources. And if
Isidore and Servius used the same sources in one large number

166 THE JOURNAL OF PHILOLOGY.

of instances, it is difficult to see why they should not have done
so in another ; or (to put the same thing from the other side)
if Isidore copied from Servius in one set of instances, why he
should have refrained from doing so in another.

Let us now examine the relation of Isidore, not to the
vulgate, but to the fuller version of Servius.

Thilo himself allows that there is a considerable number of
passages, of which he gives a list on p. xliv, in which Isidore
appears to have copied scholia from the fuller version and —
neglected the notes of the vulgate on the same points: nor is
' he disinclined to concede that in this case a community of au-
thorities is the cause of the correspondence. As there is here
no difference of opinion between us I need not dwell further on
this point. It is more important to consider in detail some
passages in which the vulgate and the fuller version can be
shewn, by a comparison with corresponding notes in Isidore,
to be homogeneous.

The first which I will take is discussed by Thilo p. xli.
Isid. 10 260 sequester dicitur qui certantibus medius intervenit,
qui apud Graecos 6 pécos dicitur, apud quem pignora deponi
solent. Quod vocabulum ab sequendo factum est, quod eius
qui electus sit utraque pars fidem sequatur.

Serv. A. 11 133 pace sequestra, media; nam[que] sequester
est [aut] medius inter duos altercantes, [aut] apud quem aliquid
ad tempus seponitur, [dictum autem a sequendo, quod eius qui
electus sit utraque pars fidem sequitur.] Pacem ergo seques-
tram indutias dicit, i.e., pacem temporalem et mediam inter
bellum praeteritum et futurum.

I agree with Thilo that Isidore is not here borrowing from
the fuller edition of Servius, but that both writers are taking
from a common authority, whom I suspect to be not Lavinius
Luscus de verbis sordidis (Gellius 20 11), but Verrius Flaccus:
Festus 339 sequester is dicitur qui inter aliquos [qui certant
medius], ut inter eos convenerit, [ita tenet depositum ali] quid,
ut ei reddat &c. But the point on which stress should be laid is
that the vulgate and the fuller edition of the Servian note are
here homogeneous, and there can therefore be no question of
interpolation. And so with the following instances (Thilo p. xlii).

THILOS SERVIUS. 107

Serv. A. 1 505 testudine, camera incurva, [id est fornicata]
quae secundum eos qui scripserunt de ratione templorum ideo
sic fit ut simulacro caeli imaginem reddat, quod constat esse
convexum. [Quidam tradunt apud veteres omnia templa in
modum testudinis facta, at vero sequenti aetate divinis simula-
cris positis, nihilominus i in templis factas esse testudines, quod
Varro ait, ut separatum esset, ubi metus esset, ubi religio
administraretur. Bene ergo, cum de templo loqueretur, addidit
ei testudinem. Idem Varro de lingua Latina ad Ciceronem,
‘in aedibus locus patulus relinquebatur sub divo, qui si non erat
relictus et contectus erat, appellabatur testudo.’ Cicero in
Bruto, ‘commentatum in quadam testudine cum servis litteratis
fuisse. Quidamtestudinem locum in parte atrii volunt adversum
venientibus. |

Isid. 15 8 8 gives an abridged version of the two notes com-
bined, again shewing that in the common source from which
both were drawn the two formed part of the same comment.
Testudo est camera templi obliqua, nam in modum testudinis
veteres templorum tecta faciebant, quae ideo sic fiebant ut
caeli imaginem redderent, quod constat esse convexum. Alii
testudinem volunt esse locum in parte atrii adversum venien-
tibus. Compare Nonius 58: testudines sunt loca in aedificiis
camerata, ad similitudinem aquatilium testudinum, quae duris
tergoribus sunt et incurvis. Vergilius Aeneidos lib. I (505) ‘in
foribus divae, media testudine templi. Sisenna Historiarum
lib. Iv. ‘C. Titinius quidam...primo ante testudinem con-
stitit, &e.

serv. A. 8 402 liquido electro, [aut liquefacto aut] puro; et
secundum Plinium in Naturali Historia tria sunt electri genera,
unum ex arboribus, quod sucinum dicitur. Aliud quod natura-
liter invenitur, tertium quod fit de tribus partibus auri et una
argenti; quas partes etiam si naturale resolvas invenies. Unde
errant qui dicunt melius esse naturale. Electri autem natura
probatur veneno, quo recepto et stridorem emittit, et varios ad
similitudinem [arcus caelestis] reddit colores, [Et ad lumina
in convivio clarius auro et argento lucet.]

Isid. 16 24 electrum vocatum, quod ad radium solis clarius

168 THE JOURNAL OF PHILOLOGY.

auro argentoque reluceat. Sol enim a poetis Elector vocatur. —

Defaecatius est enim hoc metallum omnibus metaliis. Huius
’ tria genera: unum quod ex pini arboribus fluit, quod sucinum
dicitur, alterum metallum quod naturaliter invenitur et in
pretio habetur, tertium quod fit de tribus partibus auri et
argenti una. Quas partes etiam si naturale solvas invenies.
Unde nihil interest natum an factum, utrumque enim eiusdem
naturae est. Electrum quod naturale est eiusdem naturae est,
ut in convivio et ad lumina clarius cunctis metallis fulgeat et
venenum probet. Nam si eo infundas venenum, stridorem edit
et colores varios in modum arcus caelestis emittit.

In this instance also it is clear that the vulgate and the
fuller version together make up a homogeneous note, which is
given in another and slightly different form by Isidore. Its

source may either be Pliny, with whose words (37 31, 33 81)

much of it coincides, or some later writer, such as Suetonius,
quoting and enlarging Pliny’s observations.

Serv. A. 1 119 gaza Persicus sermo est, et significat divitias,
[unde Gaza urbs in Palaestina dicitur, quod in ea Cambyses rex
Persarum cum Aegyptiis bellum inferret divitias suas condidit.]
Isid. 15 1 16 Gazam oppidum Palaestinae condiderunt Evaei,
in qua habitaverunt Cappadoces pristinis cultoribus interfectis,
Vocata autem Gaza, eo quod ibi Cambyses rex Persarum thesau-
ros suos posuit, cum bellum Aegyptiis intulisset, Persarum enim
lingua thesaurus gaza nominatur. ;

Serv. A. 1 373 annales: inter historiam et annales hoc
interest; historia est eorum temporum quae vel vidimus vel
videre potuimus, dicta a0 tod icropeiv, id est videre; annales
vero sunt eorum temporum quae aetas nostra non novit; unde
Livius ex annalibus et historia constat. Haec tamen confun-
duntur licenter, ut hoc loco pro historia inquit annales, [Ita
autem annales conficiebantur: tabulam dealbatam quotannis
pontifex maximus habuit, in qua praescriptis consulum nomini-
bus et aliorum magistratuum digna memoratu notare consueve-
rat domi militiaeque terra marique gesta per singulos dies,
Cuius diligentiae annuos commentarios in octoginta libros
veteres rettulerunt, eosque a pontificibus maximis a quibus

sae eS ee

THILOS SERVIUS. 169

fiebant annales maximos appellarunt; unde quidam ideo dictum
ab Aenea annales aiunt, quod et ipse religiosus sit et a poeta
_ tum pontifex inducatur.]

Isid. 1 63 3 annales sunt res singulorum annorum. Quae-
cumque enim digna memoria domi militiaeque, mari ac terrae per
annos in commentariis acta sunt, ab anniversariis gestis annales
nominantur. Historia autem multorum annorum vel temporum
est, cuius diligentia annui commentarii in libris delati sunt.
Inter historiam autem et annales hoc: interest, quod historia est
eorum temporum quae vidimus, annales vero sunt eorum anno-
rum quos aetas nostra non novit. Unde Sallustius ex historia,
Eusebius et Hieronymus ex annalibus et historia constant.

The Servian note is here fuller than that of Isidore. The
substance of the whole came, as Gellius (5 18) tells us, from
Verrius Flaccus.

I could add many more similar instances; but enough has,
I think, been quoted to shew that there are a considerable
number of cases where a note in Isidore closely resembles one
only to be found in the fuller version of Servius. The hypo-
theses at command for explaining this phenomenon are, so far as
I can see, the following: either that Isidore borrowed from the
fuller version of Servius, which must therefore be at least as old
as the sixth century, or the beginning of the seventh; or that
the interpolator borrowed from Isidore; or that these notes
were taken by Isidore and the author (or authors) of the fuller
Servian commentary from the same or similar sources. Thilo
rejects the first hypothesis altogether, and seems inclined to
lean in some cases to the second, in some to thethird. Butthe
second assumes that the author of the additional notes was
later than Isidore, which is the very point in question; and I
therefore am strongly inclined to adopt the third, which Thilo
himself allows to be the most natural in some cases (p. xlv.).
If in some cases, why not in all?

If, as I have endeavoured to shew, Isidore did not borrow
from Servius, but used the same authorities, it follows that the
matter common to both writers can claim a very respectable
antiquity, and authority in proportion; while with regard to

Journal of Philology. vou. x. 12

170 THE JOURNAL OF PHILOLOGY.

those additional notes of Daniel’s Servius which are shewn by = a

a comparison with Isidore to be homogeneous with the vulgate,
it is clear that they cannot be regarded as interpolations. Nor

again is there any reason for suspecting the integrity of those __
which are really supplementary to the vulgate. With regard

to those which are not homogeneous with the vulgate, which —
repeat it, or contradict it, I am unable to see that we are com-

pelled to infer more than this, that they represent a different

recension of the Servian commentary; but that they were not

inserted in it until long after the time of Servius I see no

grounds for believing. When we consider the general cha- |
racter of the fourth century commentaries on Roman authors,
such as that of Donatus on Terence and of the Pseudo-Asconius —
on Cicero, when we reflect that their style and manner is in
the main impersonal, that they bear the clearest marks of being
-compiled and abridged from the numerous works of earlier —
scholars, and that they present the same phenomena of repe~ —
titions and general looseness and carelessness in composition,

we are justified in pausing before we deny to the fuller version
of Servius its right to the name which it has so long borne.
The additional notes are undoubtedly drawn from the same —
sources as those of the vulgate; they are often homogeneous
with them, and their style, though later than that of the Verona
scholia, is on the whole neither earlier nor later than that of
Servius.

Thilo has said but little on the sources of the Servian
commentary. He does not, in my opinion, at all succeed in
shewing that Servius borrowed from Aelius Donatus, The
memoir of Vergil which bears the name of the latter is gene-
rally attributed to Suetonius, and I have endeavoured to shew
in my edition of this work that Servius extracted his shorter
biography from the fuller work of the latter, and was thus able
to add details which in the memoir by Donatus are omitted.
Thilo mentions a number of passages in which notes in the
Servian commentary correspond with notes of Donatus on
Terence. But on examining these I find that in many cases
the Servian note is fuller, and that it is not seldom possible to

THILO’S SERVIUS. 171

point out an older form of the comment in Nonius, or Verrius
Flaccus, or both. Nonius and Verrius, it may be observed,
are hardly mentioned in Thilo’s preface. Yet it is these two
authors above all others who must, in my opinion, be more
thoroughly studied than any others, if we would arrive at sound
conclusions respecting the sources of the Latin commentaries
of the fourth century.

HENRY NETTLESHIP.

12—2

PYRRHUS IN ITALY.

THE war waged in Italy by Pyrrhus of Epeiros against
the Romans has always commanded attention, from the
picturesque incidents interwoven with the story, from the
character of the king, and of the Romans, Fabricius and Curius,
and from a false idea of the importance of the struggle. |

It was important; for it secured to Rome, as the repre-
sentative of Italian unity as opposed to foreign intervention,
the control of the whole Italian peninsula for the first time.
The false importance given to the war has sprung from the
consideration of it as a kind of test action between the Roman
and Macedonian methods of fighting, between legionary and
phalangite. It has been spoken of by one popular historian
as if it first introduced the Italians, as a military people, to
the notice of the Greeks and Macedonians.

Lord Macaulay, in his Lays of Ancient Rome, (Prophecy of am :

Capys, Introduction,) uses the following strong expressions con-
cerning it.

“That barbarian warriors, led by barbarian “chiefs, should _
win a pitched battle against Greek valour guided by Greek ©
science, seemed as incredible as it would now ‘seem that the —
Burmese or Siamese should, in the open plain, put to flight
an equal number of the best English troops.” “The Tarentines
were convinced that their countrymen were irresistible in war.”
paibes “His (¢.e. Pyrrhus’) expedition to Italy was a turning
point in the history of the world”...... “The pilum and the
broadsword had vanquished the Macedonian spear. The legion
had broken the Macedonian phalanx.”

This and similar language is sheer misrepresentation. The
Tarentines must have been singularly unacquainted with their

oo ee ae eee eee OY a i an eee
we F< ao 3 7 - Oe ot, 7

< ‘Ne
ae
: a »
‘

_ PYRRHUS IN ITALY. 173

own history, and with the history of their own times, if they
believed their countrymen irresistible. Ever since the great-
| _ ness of the Italiot cities began to decline, after the destruction
:- of Sybaris in the sixth century B.c., the Italians had been
constantly pressing upon the Greeks, curtailing their territory,
and- beating them in the field. In about 473 B.c. the Taren-
tines themselves had been overthrown by the Messapians in a
pitched battle with heavy loss. The interference of the great
despots of Syracuse, Dionysios the elder and Agathokles, had
been invoked again and again in the struggles of the Greeks
against each other, or against the Italians. Champions too
had come over the Adriatic, from the mainland of Greece and
Epeiros, and had fared but. badly at the hands of the Italians.
On the same day, it was said, as the battle of Chaironeia, in
338 B.c., Archidamos the Spartan had been defeated and slain
by the Lucanians, Alexander of Molossos, a predecessor and
cousin of Pyrrhus himself, an uncle of Alexander the Great,
had since then waged war in Italy with doubtful success. He
complained that he had men to fight against, as contrasted
with the Asiatic foes of his great nephew; and by these men
he was finally overthrown and killed, by Lucanians and Bruttii.
Later still, twenty years only before the arrival of Pyrrhus,
the success of Kleonymos the Spartan, on the same field, was
at best doubtful. That the Greeks should finally be subdued
by the Italians; that the strongest Italian power which had
yet arisen should overthrow the most illustrious Greek adven-
turer who had yet come to Italy, this was precisely what a
careful observer of past history would have predicted. What
we call Rome, was a body of colonies, allies and subjects spread
throughout central Italy, directed by the most uniformly
capable aristocracy that the world has seen. That this coherent,
persistent, patriotic power should overthrow the factious demo-
eracies of the Italiot cities, backed by a military adventurer
however brilliant, assisted by disorganised and broken men
like the Samnites and Lucanians however brave, was almost
certain, The Roman victory was the natural consummation
of a long series of events.

Neither is it correct to describe the contest as one between

174 THE JOURNAL OF PHILOLOGY.
rival systems of fighting. A portion of the legion was then

armed with spears after the manner of the phalanx, though ~

not with the extraordinarily long pikes of the Macedonians
(Polyb. vi. 23). Nor were the troops of Pyrrhus merely a

phalanx supported by cavalry. The great reason for the mili-

tary superiority of the armies of Philip and Alexander had
been the combination of other troops, peltasts especially, care-
fully armed and trained, with the phalanx. Their armies, and
those of their successors, exhibited for the first time, among
Greeks, a proper combination of all arms of the service.
Infantry, cavalry, artillery’, troops heavy and light were there.
The Roman, that is the Italian®, method of fighting, with swords
in comparatively open order, must have been equally well
known to the people of Tarentum and Epeiros. Certainly the
scientific soldier Pyrrhus knew all about it. This is not con-
jecture. Polybios, comparing the Roman and Macedonian
armies which fought at Kynoskephalai in 197 B.C., says:
IIuppos ye unv ov povov Odo, dAXA Kal Svvapecw "Itadwkats
ovyKéxypntat, Tels evarra£ onuatay Kal oreipay parayyiTucny
év Tols pos ‘Pwyalovs ayoow, Polyb. xvit.11. Italian allies,
and among others Italian mercenaries, filled his line of battle,
which exhibited alternately manipuli, armed with swords and
javelins, and clumps of phalangite spears.

Dr Arnold (R. Hist. c. 37) remarks that the account of
Pyrrhus’ first battle, Heraclea, is inconsistent with the suppo-
sition that his troops were mere phalangites. The two armies
drove each other back alternately seven times. That the legions
_ should drive back an unbroken phalanx is incredible; that a
broken phalanx should have returned to the charge is well-nigh
impossible. Yet Dr Arnold does not apply the passage of
Polybios quoted above till he comes to the second battle,
Asculum. As Polybios however speaks of battles, and lower
down of victories, in the plural, as being gained by this forma-
tion, he must allude to both Heraclea and Asculum. At all
events at Beneventum, where Pyrrhus was defeated, the bulk of

1 See Arrian, Anabasis 1v. 4, for Lucanians too, were armed like the
Alexander’s use of field artillery. Romans. Liv. 1x.40. Sall. Catil. 52.
2 The Samnites, and therefore the

PYRRHUS IN ITALY. 175

his army was Italian and mercenary. His Epeirot veterans had
mostly perished, and his ranks were recruited from the mer-
cenaries of Sicily and of Italy. The victory of M’. Curius was a
victory over soldiers armed and equipped like his own, backed
only by a reserve of sartssae, if indeed Pyrrhus had any
phalangites armed in the Macedonian manner, left at all’.
Moreover to rightly understand the course of the campaigns
of Pyrrhus we must look beyond the fields of battle and beyond
Italy. He crossed from Epeiros in the spring of 280 B.c., when
his friend Ptolemy Keraunos was just established on the throne
of Macedonia. Ptolemy supplied him with a portion of his
army, perhaps with all his elephants and cavalry (Justin, XvI1.2),
but the different accounts of his numbers are irreconcilable.
After inflicting a severe defeat on the Romans at Heraclea, and
after causing a general rising of the Samnites and Lucanians,
he offered a peace, which could have been only a truce at the
best, for Rome was not seriously crippled. In this peace we see
: = no advantage for himself. The cause was as follows; in the
7 same year, 280 B.c., the Gauls had invaded Macedonia, killed
Ptolemy Keraunos, threatened Epeiros, and deprived Pyrrhus of
all prospect of support from home or Macedonia.
Perhaps the defeat of the Gauls and their allies by the
Romans in the decisive battle of Sentinum a few years before,
. and this invasion of Macedonia, give between them a more
DS valuable criterion of the relative strength of Italy and Greece
} than the campaigns of Pyrrhus can.
However, the Romans refused peace. Some indecisive opera-
tions, and another doubtful victory near Asculum in 278 B.c.,
left Pyrrhus as far as ever from the position of conqueror or

1 Macaulay’s vigorous poetical op- The lines would I believe be more true
position of sword and sarissa falls of Kynoskephalai, or of Pydna, nay,

therefore to the ground : even of the meeting of the Spanish and
‘‘ Hurrah! for the good weapons, Swiss infantry at Ravenna (Machiavelli,
Which keep the War God’s land ; Art of War, B. u. ¢. 3), than of Bene-
Hurrah ! for Rome’s stout pilum, ventum,
In a stout Roman hand. But see Plutarch’s account of the
Hurrah! for the good broadsword, battle, and above all Dionys. Halicar.
Which through the thick array xix. 12, referred to below, for the com-

Of levelled spears, and serried shields, position of the army of Pyrrhus at
Hews deep its gory way.” Beneventum,

Ee Ee EE Ee Le ee ee ee ee Pie eta See ne pea ene ee, oe, Pd oe we we ee

176 THE JOURNAL OF PHILOLOGY. —

arbiter of Italy. Again he negotiated a truce, and this time
successfully. Not that he now desired to return to Epeiros.
The Gauls had retired from Macedonia, and had been defeated
by the Aetolians in their incursion into Southern Greece. If
we could trust an evidently untrustworthy and marvellous
account of that invasion by Justin, we could place it in the
winter of 279—8B.c. But it anyhow probably occurred be-
fore the autumn of 278 B.c. when Pyrrhus left Italy. He
had been there two years and four months since the spring
of 280 B.C. -
Now freed from anxiety about the Gauls, and finding Ital
an unpromising field for adventure, he went to Sicily. His
wife Lanassa was daughter of Agathokles the late tyrant of
Syracuse, a man of talents and influence equal to his crimes.
Lanassa had borne a son, Alexander, who might fairly hope to
inherit his grandfather’s position, if Pyrrhus could only expel
the Carthaginians. But he failed, on this constant field of
Greek military failure, where “barbarian warriors, led by bar-
barian chiefs,’ had often won pitched battles “over Greek
valour, guided by Greek science.”

He returned to Italy to support his former allies, to a task — —

which his former partial success must have convinced him was
now impossible. So only can we explain the readiness with
which after the single defeat of Beneventum he retired from
the contest. He might still have preserved the independence
of Tarentum by staying. He had an army still; he brought
upwards of 8000 men back to Epeiros, and left a garrison behind
in Tarentum; but empire for himself was plainly hopeless, so
he went. At Beneventum his forces had been Tarentines,
armed no doubt as a phalanx, but more probably after the
Greek than the Macedonian fashion’; Samnites, and - mer-
cenaries from Italy and Sicily, Italians too for the most part,
retaining their national arms; with the remnants of his original
force. Plutarch makes no mention of the special formation of
the phalanx in the battle. Indeed no general in his senses,

1 The Achaeans, for instance, after were only induced by Philopoemen to
this, still kept up the old-fashioned adopt the Macedonian arms,
phalanx with the shorter spear; and

PYRRHUS IN ITALY. 177

much less Pyrrhus, would have attempted a surprise, by a
night march, through woods and mountains, with troops prin-
cipally armed with spears twenty-four feet long.

Nor does Dionysios of Halikarnassus (xIx. 12) mention the
phalanx in connection with the battle. He calls the troops
indeed é7A?rac with whom Pyrrhus attempted the surprise;
but he talks of their being encumbered, not with spears, but
Ouvpeois, the oblong shields of the Italians, described by Poly-
bios (VI. 23) as part of the equipment of the Roman émXira:,
or legionaries. Pyrrhus was certainly here employing Italian
weapons, probably Italian men, and these Dionysios tells us
were his best troops. So much for “Greek valour guided by
Greek science,” as exemplified by the army of Pyrrhus.

Nor was the battle a decisive blow to Greek influence in
Italy. The progressive events of two centuries had shewn that
to be already doomed, so far as it was a political influence.
The only question was, what Italian race should rule the Italiot
Greeks. But Beneventum did confirm the authority of Rome,
at the head of her Latin and Sabine confederacy and subjects,
over the Oscans of southern and central Italy. These had
stooped to use foreign aid ; of the Gauls in the third Samnite
war, of the Greeks in the present war; even as they welcomed
Hannibal, and intrigued with Mithradates later. Rome, with a
safer instinct, had in this very war refused to avail herself of
useful Carthaginian succours to counteract the superiority of
the Tarentine fleet; and she reaped the reward of a con-
sistent national policy by becoming beyond dispute the head of
ltaly. .

HENRY ELLIOT MALDEN.

Kitianps, Hotmwoop,
Surrey.

Se a, RP CN 1h ares ae ae ee 7 Y ean tae i ile eg Look
42-. “, Fe Le Mp gio, Se = J ster: a9

oa P . i.e r 4 Pa fe
: - é iat es Shes *

BIOLOGY AND SOCIAL SCIENCE.

‘In 1802, simultaneously, and apparently independently, a German and
two French naturalists proposed the word Biologie for the whole study of living
matter, The idea was accepted, grew, and we in England have for some time
past used the word Biology. Dr Field of Norwich had written to the Lecturer
to point out, on philological grounds, that the word is a bad one, as fios is

applied only to human life, while {wy is applied to other animal life. Although

he suggests a new term, I think it too late to change our present one.”
Prof. Huxley, Lecture at South Kensington, 1877.

‘One of the most singular things that are shewn in-that museum [of tools

and weapons] is the wonderful tendency of the human mind, when it has once
got into a groove, to stick there. The great object of scientific investigation is
to run counter to that tendency.”

Prof. Huxley, Address at the British Borociation: 1878.

Ovx dpa ravrds dvipos...dvoua Oécba éarlv, dAdAd Tivos dvomaToupyod.
Plat. Crat. p. 388 E.

“THE study of living beings, irrespective of the exact.

nature and position of these, is the province of Biology (Gr.
Bios, life; Xoyos, a discourse). All living beings, however,
may be divided into the two kingdoms of animals and plants
Paes in accordance with which division, Biology splits ny into
_ the kindred sciences of Zoology and Botany’.”

It will be the object of this paper to show that the Greek
word Bios has nothing in common with the subjects of either
of these two kingdoms, with the sole exception of Man; and
with him, not as a living, but asa rational, social and account-
able being.

Bios is thus defined by Aristotle: Bios éotl Noyx) Son.
And so the grammarians, as Ammonius: “ Béos differs from —

1 Introduction to the Study of Biology, by H. Alleyne Nicholson, M.D.,
Edinburgh and London, 1872.

BIOLOGY AND SOCIAL SCIENCE. “179

fon. Béos is used of rational animals, that is, of men only;
fon both of men and of irrational animals, and occasionally of
plants (5 8é wore kat él dutév). Whoever, therefore, uses
Bios of beasts (ért tév Onplwv) speaks improperly (axupo-
Aovet)*.’? In such apparent exceptions to this rule as we shall
presently notice, it will always be found that there is no question
of the principle of life, but only of some adjunct or accident
of it.

The various shades of meaning of the word Plos may be
thus arranged’*,

1. The duration of human life is so called. To this head
belong the phrases diayew, diateretv, SuépyecOar tov Biov;
dia Biov, per totam vitam; tov Biov xatédXvoev, KatéoTpewer,
érenevtncev, é€éduTrev etc.; paxpdB.0s, BpayvBros, odwydB.os
(Job xiv. 1); droBiodv, vivendi finem facere. “O Bios Bpayds
(is the well-known aphorism of the Father of Medicine) 7 &é
Téxvn paxpy. In this sense, Ilépas &rracw dvOpdrrois éott Tod
Biov @avatos*; and one who is near that term may say with

Cicero: Mihi quidem BeBiwras ; viderint juvenes*.

2. Life considered in regard to the feelings, with ee to

1 Ammonius Ilepl ouotwv kal duapbpwv
Aé~ewv, p. 30, ed. Valck. Archbishop
Trench, in his New Testament Syno-
nyms, p. 105, points out that the asser-
tion of Ammonius, that Bios is never,
except incorrectly, applied to beasts, is
inconsistent with Aristotle’s use of the
word in Hist. Anim. 1.1, 15 [13] and
1x. 8 [7], 1, ‘“‘ unless, indeed, he means
to include Aristotle in his censure.”

But dxvpodroyla, although reckoned

amongst the vitia orationis, are only
censurable when used harshly or ex-
travagantly ; in many cases, especially
in poetry, they are beauties. Thus
Virgil’s “ Vir gregis,” and ‘*‘ Hunec ego

‘si potui tantum sperare dolorem,” and

Thomson’s ‘ Softly shaking on the
dimpled pool Prelusive drops,”’ are in-
stances of the judicious use of this
figure. The Greek word dpiecOa is

properly said of dogs and wolves; but
this only gives greater significance to
the Psalmist’s (xxxvil. 8) wpuduny do
orevayyod THs Kapdlas wov. And in the
latter of the two places of Aristotle, he

is comparing the habits (rods Blious) of

animals, in regard to skill and inge-
nuity, with those of the human race,
instancing the manner in which the
swallow builds its nest, mixing straw
with the clay, ete.

2 With our arrangement may be com-
pared that of Tzetzes on Hesiod Op. et
D. 689: Blos & onualver ryv réxvny,
TOV TpdToVv, TOV TapovTa KdaMoY, TOY évOS
éxdoTou xpovoy THs wns, THY Teptovolar,
kal Tas mpos TO (Hv ouvTelvovcas TpoPpds.

3 Demosth. IIep! rov crepdvov, p.
258, 20.

4 Cic. Epist. ad Att. x1v. 21.

180 THE JOURNAL OF PHILOLOGY.

-« . 2 1 : .
happiness and misery (eis evppoovvas te kal Avzras'). In this
~! / / > / 1) / >»: / “4 or
view Bios may be yanerros, émimrovos, oouvynpos, AVIrNpOS ; OT,

on the other hand, evSa/uov, waxdptos, drpdypov, dpépyuvos

etc. To this head also belongs the well-known phrase Bios
aBieros, a life so wretched as to be insupportable. An ancient
philosopher recommends 6ddv péev tiv NecordTny éxréyeo Oat,
Biov && tov advidtatov ; but a better rule of life is that of
Pythagoras, who said é7v yx) Biov aipeioOar tov dpiotov* 78dv
yap avtov 7 cvvnbea moujoce. Animals, being endowed with -
feelings like ourselves, are not excluded from this use of the
word per dxuporoyiav. As we speak of leading the life of a
dog, so the Greeks had a saying, AXayo Biov Shr, to lead the life
of a hare, that is, in continual fear and trembling (Sed:as Kai
TpEMo@v)”.

3. Life considered i its moral aspect, or in regard to the
conduct, is emphatically called Bios. To this sense belong the
epithets ceuvos, Kdcptos, évapeTos, cOppav, YPNTTOS, ETLELKNS,,
and their opposites; and the synonyms tpozros, 70n, mpakis
etc. In Christian writers Bios, even without an epithet, is
often contrasted with miotis or déypata; as St Chrysostom,
in distinguishing between heretics and hypocrites, says: Hapa
yap aipetixots éote ToAAGKIS Kal Biov (good living) evpety, Tapa
dé TovToLs ots eizrov ovdamds*.

4. But in forming an estimate of moral worth, it is abso-
lutely necessary to take into consideration the circumstances in
which any one is placed, as well as his conduct under them.
Both these were within the purview of the great Orator in his
celebrated challenge to his rival: “Draw then the parallel be-
tween your life and mine (tad col xapot BeSiwpéva), Aeschines.
ee You were an usher, I a scholar; you were an initiator,
I was initiated; you danced at the games, I presided over
them; you were a clerk of the Assembly, I a member...... your
measures were all in the enemy’s favour, mine always in the

1 Xenoph. Hiero 1. 2: Ij duadépe 6 ford.
Tupayyikos Te Kal o cdwwrikds Blos els 3 Demosth. ut ante, p. 314, 24. _
evppoovvas Te kal-AUwas avOpwro.s. 4 §. Chrysost. Opp. T. vir. p. 293 B,

? Stob. Flor. T. 1. p. 17, ed. Gais- ed. Ben.

i
|

BIOLOGY AND SOCIAL SCIENCE. 181

country’s’.” Hence the title of Plutarch’s great work, BIOI
TIAPAAAHAOI; and the English word, legitimately formed

from the Greek, Biography.

5. A man’s calling or profession (réxyvn, émitndevpa) is
also indicated by this word, as Bios @aXartTuos, -yewpytxds,
AnoTpLKOS, BovKoALKOS, KTNVOTPOhos, TTPATLWTLKOS, PLAdTopos
etc.; and by an easy transition, he is said dd tis Oardtrns,
yewpylas, Anotelas etc. Tov Blov exew, TovetcOat, Topiver Bar,
to get his living. To this head probably belongs one of the
precepts (v7roOjxa.) of the Seven Sages, Té Bio pr payou,
Do not quarrel with your bread and butter ; as well as a pithy
saying of Metrodorus, preserved by Stobaeus in his collection
Tlept pevdwrias, “Eroiuafovrai tives d1a Biov ta pos tov Biov”.
Every reader of the Greek Testament is familiar with this
use of the word*®. Hence (since animals must live, in this

sense) arises another axvpodoyia, of which an instance is com-

monly quoted from Xenophon’s description of spiders: Aé
parayyes, apayvia rewTa vpnvapevat, Onpwot TA Tpds TOV
Biov*. But a more notable example, and one of common
occurrence, is the word audiBios, amphibious, applied to such
animals as pass their time and get their living in both elements;
quia non tantum terrestria, sed aquatila quoque desiderant
pabula®’. For a similar reason a smaller class of animals, as
owls and bats, are called vu«riRvor.

6. Passing from individuals to communities, the diversities
of the various tribes of the human race, in regard to manners,
customs, and modes of subsistence, are properly styled Bios, and
distinguished by such epithets as #pepos, ayptos, oxnvirns,
vouady.xos, Onpiddns etc. Diodorus Siculus, one of the earliest
cultivators of anthropological science, concludes his description
of the races inhabiting the countries bordering on the Arabian
Gulf with these words: “ Now if any of my readers, by reason
of the strange and marvellous character of the habits of life

1 Demosth. ut ante, p. 315, 5 (Lord 3 See Mark xii. 44, Luke viii. 43, xv.

Brougham’s Translation, p. 180). 12,
2 Stob. Flor. T. 1. pp. 116, 341, ed. 4 Xenoph. Mem. Soer. ur. 11, 6.
Gaisford. 5 Columella vir. 13.

182 THE JOURNAL OF PHILOLOGY.

here recorded (trav avayeypaupévwv Biwv) should disbelieve

what has been narrated, let him only bear in mind the differ-
ence between the temperature of the air of Scythia, and of

the country of the Troglodytes, and he will be no longer

1»)

incredulous’.” But even in civilized life, and in the same
community, manners are continually changing; and in the
increase of luxury and its attendant evils, a philosopher may
sometimes wish to recall the old-fashioned and simple habits
(tov apyatov Kai amapackevov Biov) of ages long gone by.

7. Lastly, human life in its most comprehensive aspect,
genus humanum, the world, is properly expressed by 6 Bios 6
avOperwwos, 6 Kowds Bios, or simply 6 Bios. Thus the great
benefactors of the human race are described as of evepyern-
cavrTes peyada Tov Tov avOpworev Biov’; and the greatest of
them, Hercules, who is lauded by the Historian as having by
his own labours humanized the world (é&nwepwoas tiv oixov-

Hévnv*), is represented as saying of himself and his exploits,

ds Atos pev vids eiut, TooadTa Oé weTévnea "EKKA@AIPON
TON BION*. And to quote only one more instance, St Chry-
sostom, commenting on the text (1 Tim. vi. 1) “Let as many
' servants as are under the yoke count their own masters worthy
of all honour,” observes that if this precept were neglected, the
Greeks would have some reason to say that Christianity was
introduced into the world for the subversion of all things (émi
avatpoTn Tov TavTwev eis Tov Biov eicevnvextat)’. Akin to
this is the patristic use of Bios for secular life, as, mapOévos
atrotazauéevn TH Bip; and Bvwtixol dvOpwrot, as opposed to
the followers of a religious life. |

The obvious result of the foregoing enquiry is to shew that
the term Biology, recently imported into the scientific vocabu-
lary, is a BLUNDER, illustrating the old saying, “A little learning
is a dangerous thing.” The inventor of it®, being in want of a

1 Diod. Sic. 1. 32. ed. Ben.

2 Idem tv. 15. 6 Gottfried Reinhold Treviranus

3 Idem rv. 8, (born 1776, died 1837), in his work

4 Lucian Dial. Deor. xin. 1 (coll. entitled, Biologie, ou Philosophie de la
Vit. Auct. 8). nature vivante, which appeared in 1802

® §. Chrysost. Opp. T. x1. p. 774 A, and following years,

. A

BIOLOGY AND SOCIAL SCIENCE. 183

Greek word expressive of life, had recourse to his dictionary,

- which offered him a choice between two, Bios and fo. Zoology"

being already in use, as a derivative from ov, he was forced to
take up with Biology, a well-sounding word, and not likely to
be too closely scrutinized by the general scientific world. So it
has proved. The philosophic mind,’ which, after all, is but
human, has got into a groove, and seems likely to stick there,
unless some helping hand is extended to it from without. Let
us see what philology can do in this matter constructively, as
well as destructively.

To attempt this with any prospect of success, two things
seem to be required: first, to propose a substitute for the
obnoxious term; and, secondly, to find another and a legitimate
use for it.

1. The wtal principle (ado TovTo @ Copev") in Greek is
neither fo») nor Bios, but 76 Cwrixdy. Thus the author of the
Geoponics says that in the trying (dox:uacla) of eggs care
should be taken not to shake them, for fear of destroying the
vital principle (fva pu) SiapOapy ro év avrois Cwrixov)*®. The
term ZOTICOLOGY is not quite on a par with the similarly
sounding word Toxicology, because rtofixdv, poison, though
originally an adjective (Sappaxov being understood), has by use

passed into a noun; whereas Td Gwttxdv (like To é6paruxoy, the

visual faculty, To ooppavtiKov, TO Noyexdv etc.) becomes a noun
only by the help of the neuter article, which unfits it for enter-
ing into the construction of a compound term. Still, without
appealing to such doubtful compositions as CoE from véos,
Hagiology from a@ysos, Eschatology (!) from écyaros (all three
adjectives), we need go no further than the undeniably legiti-
mate formation apyaioroyla Archaeology, the science which
deals with ancient things (ta adpyaia), as Zoticology is concerned
about things endued with life (rd Gwrixad). But since it is
always a matter of difficulty to obtain currency for a new term,
however legitimately formed, it deserves to be considered

1 There is not the same excuse for not from {@or).

such monstrosities as Bioplasm, Bio- 2 Plato ap. Stob. Flor. T. 1. p. 273,
genesis, etc., which might easily be ed. Gaisford.
changed into Zooplasm, etc. (from (wy, 3 Geopon. xIv. 7, o7. :

184 ' THE JOURNAL OF PHILOLOGY.
whether the opposing claims of science and philology might not,

in this particular case, be reconciled by the simple disuse of the

objectionable word, without the adoption of any substitute for
it. Is a term, which is merely a higher generalization of two
subjects, usually studied distinctively, and represented in Uni-
versities by two or more separate chairs, absolutely necessary, or
even highly convenient, for the promotion of scientific research ?
Would not the Biological, section of the British Association for
the Advancement of Science conduct its affairs as well under

the title of the Zoological and Botanical section, as under its

present more compact and euphonious, but decidedly unphilo-

logical, and, therefore, unscientific designation? I lately read

that the late Arctic expedition had “supplied us with a great
mass of additional knowledge, especially as regards the biology
and physical geography of the newly discovered regions.” A
few years back, instead of biology the writer would have said
the Fauna and Flora of those regions, with no loss of elegance,
and with a decided advantage in point of linguistical propriety.
If those who have occasion to use the word would only remem-
ber Prof. Huxley’s candid acknowledgment that it is a “bad
word,” and only to be tolerated on the unphilosophical plea of
“too late to mend,” surely they would not grudge going a trifle
out of their way, and by avoiding the use of the term them-
selves, help to bring about its gradual discredit and final extinc-
tion in the sense which it has hitherto borne.

2. But what shall we do with the discarded word? I
answer: let the admirers of what is now periphrastically called
Social Science assert its claim to be admitted amongst the
ologies ; and having regard to the wide use of Bios, as defined

in this paper, let them consider whether the objects and uses ~

of that branch of science, which they so zealously and lauda-
bly cultivate, may not be fairly represented by the term
BIOLOGY. 3

“Social Science,” says a former President-of the Association
for the Promotion of that science, “I take to be the acquisition

of such knowledge as shall enable the human community by ~

which the earth is inhabited to reach the highest level of moral

and physical well-being, which is compatible with the original

cetgent tn aie —e rT

ER Es

pl te il . 2 . ~ 1?

BIOLOGY AND SOCIAL SCIENCE. - 185

‘conditions of their existence.” And if I were promulgating a
new science under the name of Biology, I should define its
‘object to be, in the words of Polybius, 7 éravépOwors tod Taév

-avOpaérav Biov', the correction of human life, a definition of

which Lord Dufferin’s account of Social Science is little more
than an expansion. The task is an Herculean one, in both
senses of that epithet; it is an arduous task, and it is the very
one which (as we have seen) Hercules, in his generation, and
according to the requirements of that early stage of society, set
himself to perform, éxxa@atpew tov Biov, to purify the world, by
ridding it, as the great Grecian hero of boars, lions and hydras,
so the associated heroes of Adam Street, Adelphi, of the mon-
strous abuses, the Augean accumulation of social disorders and
derangements, which stand between the human race and the
“highest level of moral and physical well-being,” which it is
capable of attaining. But if the “club and lion’s skin” should
appear to belong to a state of civilization widely different from
that with which modern professors have to deal, coming down a
few ages, we meet with the honoured name of one whom the
cultivators of Social Science would do well to adopt as the
founder of their faith, and the model of a true social philosopher.
SOCRATES, says his biographer, rp@ros wept BIOT died éyOn’.
Common life was his lecture-room, his laboratory, his observa-
tory. The subject of his researches was, as he himself avowed,

"“O > / ’ 3 06 /
TTL TOL EV Heyapotat KQKOV T AYQAUOV TE TETUKTAL,
Whate’er of good or ill our homes enshrine’.

Socrates, writes the great Roman philosopher, primus philo-
sophiam devocavit e caelo, et in urbibus collocavit, et in domos
etiam introduxit, et coegit de vita et moribus, rebusque bonis et
malts quaerere*. If the papers hitherto read at the Social
Science congresses should appear to have turned too much
upon the smaller details (yet not small in their aggregate bear-
ing upon human happiness) of sanitary requirements, mercantile
laws, the treatment of lunacy and the statistics of crime, per-
haps the assumption of a new name, at once more comprehen-

1 Polyb. Histor. 1. 35, 1. 3 Hom. Od. A. 392.
2 Diog. Laert. Vit. Socr. v. 4 Cic. Tuse. v. 4.

Journal of Philology. vou. X. 13

186 THE JOURNAL OF PHILOLOGY.

sive and more elevated than the original designation, may have

the effect of bringing out more conspicuously, and keeping —

more steadily in view, the great aims and landmarks of the
‘Science of Life, and of reducing all its parts into harmonious
proportion with each other and with the whole. At the con-

gress of 1878, the President of the Art section (then for the — |

first time admitted into the programme of the Association)
thought it necessary to apologize for the intrusion, and to

answer the question, “What has Fine Art to do with Social —

Science?”’ But if the enquiry had been, “ What has Fine Art
to do with Biology?” the respondent might have taken up a
higher tone. He might have asked in return, What would
human life be, stripped of all those elegances and refinements,
which exercise a humanizing influence upon all classes of
society, down to the very lowest; which contribute so largely
to the employment of the idler, and the enjoyment of the
busier, part of mankind? What but a Bios aBiwtos, destruc-
tive of the ends, and unworthy of the name, of life? The
eloquent eulogium of Cicero upon one of the polite arts is
equally applicable to them all: Huec studia adolescentiam alunt,
senectutem oblectant ; secundas res ornant, adversis perfugium ac
solatium praebent ; delectant domi, non impediunt forts ; pernoc-
tant nobiscum, peregrinantur, rusticantur’. Let the Biologist of
the future take for his motto, Humani nihil a me alienum puto.

Let the National Biological Association, if such should ever

come into existence, write upon its banners (to wit, the title
page of its annual volume of Transactions)

Quicquid agunt homines, votum, timor, ira, voluptas,
Gaudia, discursus, nostri est farrago libelli’.

1 Cic. pro Archia poeta 7. 2 Juven. Sat. 1. 85.

F. FIELD, M.A., LL.D.
Norwicy, August, 1879.

‘ Be een Lee
i ite Aa ee —— ss = ” i ”

HORATIANA.

Carm. 11 2 1—4.

Nullus argento color est auaris

abdito terris, inimice lamnae

Crispe Sallusti, nisi temperato
splendeat usu.

3

ALIKE Lambinus’ ‘abditae’ and Bentley’s the only rational
elucidation of the MS reading compel the words ‘auaris terris’
to mean the miser’s coffers: now when Horace says carm. III
3 49 sq. ‘aurum inrepertum et sic melius situm, Cum terra celat,
spernere fortior Quam cogere humanos in wsus’ he is to be sure

taking the other side as a poet may, but the parallel does seem

to show that ‘auaris terris’ here must have its natural sense of
the mine, ‘in her own loins She hutcht the all-worshipt ore’ as
Comus says. And is not ‘inimice lamnae, nisi temperato
splendeat usu’ or ‘auaris abditae terris inimice lamnae’ a most
dark and helpless way of saying ‘open-handed Sallust’? And
then how ‘inimice’ and its train of dependants encumber and
overbalance the sentence. If then as seems likely it is in
‘inimice’ the corruption lies, this is what I would suggest:

nullus argento color est auaris

abdito terris, minimusque lamnae,

Crispe Sallusti, nisi temperato
splendeat usu.

‘Silver in the mine has no lustre at all, nay even when coined
it has next to none, without it is burnished by changing hands,’
This at least does away with the obscurity and redresses the

15—-2

188 | THE JOURNAL OF PHILOLOGY.

balance of the sentence. It is chiefly I suppose because Horace
was at no period unread that the corruptions in his Mss seldom
lie on the surface, they present a resemblance at least superficial

to sense and metre: if ‘minimusque’ by two common errors

became ‘inimusce’ the further change to ‘inimice’ was all bu
inevitable. ;

Carm. ur 5 31—40.

Si pugnat extricata densis
cerua plagis, erit ille fortis —
qui perfidis se credidit hostibus,
et Marte Poenos proteret altero
qui lora restrictis lacertis
sensit ners timuitque mortem.
hic unde uitam sumeret inscius
pacem duello miscuit. o pudor!
o magna Carthago probrosis
altior Italiae ruinis.

In this the reading of most MSs and well-nigh all editions
Bentley justly finds fault with the lame climax ‘timuitque
mortem’, and ‘hic’ used where the poet should and might have
used ‘ille’: he might too have added, what sort of writer is
Horace if ‘mortem’ and ‘ uitam’ here have nothing to do with
one another? But there is this deeper fault in the reading,
that it makes Regulus lose the thread of his argument; for
what is he debating ? not what is done and cannot be undone,
the surrender of the army, but its ransom, the matter in hand:
his aim is to fence off the pernicies ueniens in aeuum, the
flagitio additum damnum, the probrosae Italiae ruinae, and
down to v. 36 he is speaking straight to the point; but here
with a full stop at ‘mortem’ he loses his way and drifts off into

mere exclamation about what is past mending and will remain

the same whether he gains his cause or loses-it.

But several good Mss, that of Queen’s College Oxford among

them, have ‘aptius’ for ‘inscius’, and very many more give it
for a varia lectio: Bentley then accepting this, proposed
‘timuitque mortem Hinc, unde uitam sumeret aptius, Pacem et

yg ‘ vs ae i
re meen A MO a =

ree ee
= <a 4
ee ee

HORATIANA. a eS

duello miscuit’, comparing carm. 11 11 38 ‘ne longus tibi
somnus unde Non times detur’. This removes at once the
faults of the passage and saves Horace’s credit as a rhetorician :
‘hinc’ for ‘hic’ is the slightest of changes, carm.117 14 and
21 13 the ss have ‘ hinc’ where ‘hic’ must be right: but his
insertion of ‘et’ has not much likelihood, as he himself tacitly
acknowledges on Iv 4 18.

But can ‘pacem duello miscuit’ in Horace mean ‘con-
founded war with peace’? Horace who five times elsewhere
uses ‘duellum’ uses it with a marked restriction, always of
some single war, never of war in the abstract: the word’s fancied
connexion with ‘duo’ was maybe at the bottom of this: war
as opposed to peace is ‘bellum’ carm. 11 19 28 ‘idem Pacis
eras mediusque belli’ serm. 11 2 111 ‘in pace, ut sapiens, aptarit
idonea bello’ 3 268 ‘in amore haec sunt mala, bellum, Pax
rursum’: if he wants to use ‘duellum’ thus he must use the
plural epist. 1 1 254 ‘tuisque Auspiciis totum confecta duella
per orbem, Claustraque custodem pacis cohibentia Ianum’. I
will suggest then that Horace here too was true to his custom

and wrote ‘pacemque bello miscuit’: ‘u’ and ‘b’ are in his Mss

as in others much confused, carm. 111 23 19 ‘mollibit’ for ‘ mol-
liuit’, 1 20 10 where Munro emends ‘uides’ for ‘bides’ or
‘bibes’, 25 20 Aldus’ ‘Euro’ for ‘Hebro’ is probably right:
‘be’ then might well fall out after ‘ue’, and the senseless

‘pacemquello’ would be readily altered by the change of one

letter to ‘pacem duello’.

Carm. T1i--11 15-20:

Cessit immanis tibi blandienti
ianitor aulae .

Cerberus, quamuis furiale centum

muniant angues caput elus atque

spiritus taeter saniesque manet
ore trilingui.

Perhaps neither ‘eius’ alone nor ‘spiritus manet’ alone
would be intolerable, but surely the pair of them is more than
man can stand: so at least thought Bentley Meineke and

190 THE JOURNAL OF PHILOLOGY.

Haupt. Haupt and Meineke however betake themselves to _

the coward’s remedy of declaring the stanza spurious: Bentley
perceiving that the alteration of ‘eius atque’ into a verb for
‘spiritus’ rids us at one stroke of both inconveniences proposes
“exeatque’: he cites instances of ‘spiritus exit’ but candidly
adds ‘verum hic notandum est, quod in his locis spiritus exit
de iis duntaxat dicatur, qui moribundi animam expirant. Quare

ad cuitandum Ambiguum, utinam Noster scripsisset potius-

exeatque halitus teter’. I propose then ‘effluatque’ a word
which can well be applied to ‘spiritus’ or the like, Ovid met.

VI 233 ‘ne qua leuis efluat aura’, Cic.n. d. 11 39 ‘aer efiuens —

huc et illuc uentos efficit ’. Of all errors ‘i’ for ‘1’ is perhaps
the most frequent, ‘s’ for ‘f’ by no means unusual, and if a
copyist read or wrote ‘essiu atque’ then ‘eius atque’ was not
far off.

Carm. lI 26 1—8.

Vixi puellis nuper idoneus
et militaui non sine gloria:
nunc arma defunctumque bello
barbiton hic paries habebit,
laeuum marinae qui Veneris latus
custodit. hic hic ponite lucida
funalia et uectes et arcus
oppusitis foribus mainaces,

Of all weapons the one which doors and doorkeepers can
best afford to laugh at is an ‘arcus’ in any known sense of the
word: Bentley’s ‘securesque’ however is not likely, no more is
Keller’s ‘et ascias’: indeed it surely is plain enough there is no
keeping ‘et’: you can almost count up the available substan-
tives on your fingers and sec that none of them will do. But is

it a substantive that is wanted? Theocritus cited by Bentley

has qredێxers Kal Napurtrades, and that Horace had this in his
head is likely enough ; but why then when Theocritus mentions
only two sorts of ‘arma’ should he mention three? Surely
hatchets alone or crowbars alone are all that is wanted in
addition to the torches, and his ‘uectes’ do duty for Theocritus’

GATT PONT

a al
¥ 2%,

HORATIANA. I9I

meréxels. Then asto the symmetry of the sentence: ‘funalia’

has an epithet to itself, and that ‘uectes’ should tally with it is

at anyrate as likely as not. What Iam trying to make out is
that here we have a corruption such as Bentley detected in
‘eius atque’, that ‘et arcus’ represents a single word, probably
then an imperative co-ordinate with ‘ponite’: can it be ‘et
uectes sacrate Oppositis foribus minaces’? ‘sacrate’ with the
change of one letter is ‘et arcus’ written backwards: to be sure
I know of no quite parallel corruption, but in Propertius
(Baehrens) m1 5 25 DV give ‘integras’ for ‘ et nigras’ precisely
reversing the first four letters.

Carm. Iv 4 65—68.

Merses profundo, pulchrior euenit ;
luctere, multa proruet integrum
cum laude uictorem, geretque
proelia coniugibus loquenda.

Many seem to have felt the strangeness of ‘merses, euenit’
followed by ‘luctere, proruet geretque’, yet ‘exiet’ is quite out
of the question, and ‘proruit’ and ‘geritque’ are not very
taking. And then the unexampled use of ‘euenire’? The Mss
vary between ‘merses’ ‘mersus’ and ‘mersae’: ‘mersus’
which has most authority is of course impossible and is attri-
buted by Keller to the Mavortian recension: among those
which have ‘mersae’ is Keller’s liber archetypus F (= gy), one
of the mss which preserve for instance the genuine reading

- frumpit’ carm. 11 27 5. I think it then not unlikely Horace

wrote ‘mersae profundo pulchrius euenit’, like ‘male istis
eueniat’ etc.: a copyist misunderstanding the construction
might readily write ‘pulchrior’, compare the corruption of ‘ad
uentum’ to ‘aduentus’ carm. I 23 6. This at all events does
away with both difficulties at once.

Carin 19.13 55:

Nidum ponit Ityn flebiliter gemens

infelix auis et Cecropiae domus

aeternum obprobrium, quod male barbaras
regum est ulta libidines.

192 THE JOURNAL OF PHILOLOGY.

Bentley says ‘ideo aeternum opprobrium quod sive quia
male ulta est mariti libidines’, that is he makes ‘ obprobrsai

nominative, ‘ quod’ =‘ quia’ and refers.‘ ulta est’ to ‘auis’: all

commentators seem to follow him in the main, You can hardly
have demonstration on a point like this; but does not ‘auis et
obprobrium ponit nidum’ make a strange hendiadys? one
would rather expect ‘obprobrium’ to be placed in apposi-
tion. I should be inclined to take ‘ obprobrium’ like ‘Ityn’ as

governed by ‘ gemens’, ‘quod’ = ‘namely that’, and refer ‘ ulta

est’ to Cecropia domus: ‘ Cecropiae domus’ wa then be the
‘auis’ and her sister : ‘lamenting Itys, lamenting too her sister’s
infamy and her own, their dreadful revenge on Tereus’.

Hpod. 1 7—14.

Utrumne iussi persequemur otium
non dulce, ni tecum simul,

an hunc laborem mente laturi, decet
qua ferre non molles uiros ?

feremus et te uel per Alpium iuga
inhospitalem et Caucasum

uel occidentis usque ad ultimum sinum
forti sequemur pectore.

The great awkwardness of ‘laturi’ here =‘ laturi sumus’ has
led Nauck to put a comma after ‘laborem’ and govern it by
‘persequemur’: this however only makes matters worse, as
‘persequemur otium’ means ‘Shall I pursue my present stay-
at-home life’: now it is absurd to make Horace say ‘Shall I
continue to stay at home or continue to go to the wars’.
Another objection, though perhaps not a serious one, I will
mention, which applies alike to both interpretations: they make
Horace ask a question of Maecenas to whom throughout this
poem he is speaking, and then take the words out of Maecenas’

mouth and give the answer in his own person. The punctua-

tion I propose then is this,

utrumne iussi persequemur otium
non dulce, ni tecum simul,

HORATIANA. 193

an hune laborem mente laturi, decet
qua ferre non molles uiros,
feremus, et te uel per Alpium iuga
inhospitalem et Caucasum
‘uel occidentis usque ad ultimum sinum
forti sequemur pectore ?

He then makes Maecenas answer this question by a counter-
question, ‘roges, tuum labore quid iuuem meo Imbellis ac
firmus parum’, and everything runs smoothly. Perhaps it is
not worth much that Porphyrion’s lemma consists of these
words thus written, ‘an hunec laborem mente laturi decet qua
ferre non molles uiros feremus’.

Epod. Ix.

Quando repostum Caecubum ad festas dapes
uictore laetus Caesare
tecum sub alta (sic Ioui gratum) domo,
beate Maecenas, bibam
sonante mixtum tibiis carmen lyra, 5
hae Dorium, illis barbarum,
ut nuper, actus cum freto Neptunius
dux fugit ustis nauibus,
minatus urbi uincla quae detraxerat
seruis amicus perfidis ? 10
Romanus eheu (posteri negabitis)
emancipatus feminae
fert uallum et arma miles et spadonibus
seruire rugosis potest, |
interque signa turpe militaria 15
sol aspicit conopium.
ad hune frementes uerterunt bis mille equos
Galli canentes Caesarem,
hostiliumque nauium portu latent
puppes sinistrorsum citae. | 20
io Triumphe, tu moraris aureos
currus-et intactas boues ?
io Triumphe, nec Iugurthino parem

194 THE JOURNAL OF PHILOLOGY.

bello. reportasti ducem, |

neque, Africani cui super Carthaginem
uirtus sepulerum condidit. ~

terra marique uictus: hostis punico
lugubre mutauit sagum.

aut ille centum nobilem Cretam urbibus

uentis iturus non suis 30°

exercitatas aut petit Syrtes noto
aut fertur incerto mari.

capaciores affer huc, puer, scyphos
et Chia uina aut Lesbia,

uel quod fluentem nauseam coerceat 35.

metire nobis Caecubum.
curam metumque Caesaris rerum iuuat
dulci Lyaeo soluere. .

I am constrained to cite this poem in full, though it now has

but one critical difficulty, because I think I can contribute —

something to its elucidation as a whole. It takes some nerve
to say it, but I am much deceived if all the commentators I
have read are not strangely out in supposing it written after,

not before the battle of Actium: I really think this only wants ~

pointing out to be self-evident. Let us see: vv. 1—6 will
square equally well with either view: they are generally taken

to mean ‘when shall we have a chance of carousing together ©

over this victory of Caesar's’: they may just as well mean
‘when will Caesar win his victory and set us carousing’. On
vv. 7—10 I will only say it seems to me unlikely he would care
to say so much about Sex. Pompeius in the full blaze of Actium,
but I lay no great stress on this. Vv. 11—16 the tense is
generally taken to be historical, if I am right it will be present.
Vv. 17—20 are important: the critical hitch in v. 17 need not
delay us for the present: vv. 17, 18 seemingly refer to the
defection of Amyntas and Deiotarus with their Galatians some
time before the batile: what do vv. 19, 20 refer to? The older

commentators say to Cleopatra’s flight to Alexandria: if that is

so, my theory of course crumbles away, and with it Horace’s
reputation for a decent style: to announce .the defection of

i
Ss

HORATIANA. 195

2000 men out of 100,000, and then in the same breath, as an
afterthought, that the world is lost and won! The lines refer
then to some naval defection or mishap or mismanagement
matching the desertion of the Galli on land: what ‘sinistrorsum
citae’ means perhaps no one will ever know: Bentley suggests
it may be some nautical technicality, and if so we need not be
astonished at our ignorance, seeing that Cicero did not know
the meaning of ‘inhibere remis’. What sort of poet now is
this who with the thunder of Actium in his ears can dwell on
the desertion of a handful of barbarians, and mention the
‘hostilium nauium puppes’ without saying they are burnt to
the water’s edge? To proceed: I suppose it is vv. 21—32 that
have thrown the commentators off the scent: I shall be sur-
prised if any one familiar with the locutions of poetry finds a
‘difficulty here, but if he does I will cite a parallel: ‘The
mother of Sisera looked out at a window, and cried through the
lattice, Why is his chariot so long in coming? why tarry the
wheels of his chariots?’ Set that now against ‘io Triumphe,
tu moraris’ cet., and with ‘terra marique uictus’ cet. compare
‘Her wise ladies answered her, yea she returned answer to
herself, Have they not sped? have they not divided the prey ;
to every man a damsel or two; to Sisera a prey of divers
colours?’ this interrogation being of course in Hebrew poetry
tantamount to the strongest affirmation. Horace too returns
answer to himself, and the answer is not correct in its details:
if these lines are meant as a rejoicing over Actium, then what
is the meaning of ‘terra uictus’? there was no land-fighting at
all, except a cavalry skirmish some days before the battle: a
week or so after the battle Antonius’ main army laid down its
arms without a blow disgusted at the desertion of Canidius,
Is that then what Horace means? but if so this poem must
have been written full a fortnight after the battle, and that is
incompatible with the ignorance vv. 29—382 about Antonius’
flight. Truth to tell the poet is trying like the mother of
Sisera to cheer himself with glowing anticipations, and finding
this unavailing is driven to ‘capaciores scyphos’. The last
lines vv. 33—38 are generally supposed to inaugurate a
earouse over the victory, though Horace takes pains to say that

they +5 sotliing of the kind: : fcuram niphuencle
anxiety, what fear could Horace have for the conquero :
the world at his feet? that Octavianus’ difficulties v we
over with Actium may be true as a matter of o ise ut *

‘in dubioque fuere utrorum ad regna cadendum om
humanis esset terraque marique’. | igs

I now come to the well-known crux in v. 17: : a8, hune’
by far the most Ms authority: perhaps an easy and satisfact
correction would be ‘at nunc’, which Horace as Munro
pointed out probably wrote ‘ad nunc’. ‘Frementes’ »
surely belong to ‘equos’ not ‘Galli’, see carm. Iv 14 23°
mentem mittere equum ” : it is alsin an epitheton solle
If there is anything in what I have been saying above, ‘nu
_will seem quite necessary to mark the change from dark to
first streak of light, ‘

A. E, HOUSMAN,

A A AE AS A ARR RAPE ORES eek Sn
: 7 ¥ - = P

\ .

~ en

ON A PASSAGE IN THE RHETORICA AD
HERENNIUM.

x» This paper was read at a meeting of the Oxford Philological Society,
Nov. 19, 1880. Since then H. Jordan has published the suggestion ‘ quotannis’

_ for ‘ quodam is’ in the Hermes (May, 1881), p. 48; whether or not for the first

time, I am not aware.
W. W. F.

Iv. 54. 68. ‘Breuitas est res ipsis tantummodo verbis ne-
cessariis expedita, hoc modo; Lemnum, praeteriens cepit,
inde Thasi praesidium reliquit, post urbem Lysimachiam sus-
tulit, inde pulsus in Hellespontum statim potitur Abydo.
Item: modo consul quodam is deinde primus erat ciuitatis.
Tum proficiscitur in Asiam, deinde hostis est dictus, post impe-
rator et populi Romani consul factus est’.

THE text here given is that of Kayser’s separate edition
of the Rhetorica (1854), which has not been improved upon
by Baiter and Kayser in their joint edition (1860). Of the
two examples of ‘ breuitas’ given in the passage, the first is very
corrupt, and attempts to restore it must necessarily be doubt-
ful in the absence of any certainty as to the events referred to

in it. The last three words, however (statim potitur Abydo),

may be accepted as certain on the manuscriptal evidence, and
may possibly supply a key, as I shall presently show, to the
meaning of the rest of the example.

The second example, which is less corrupt and more im-
portant, has generally been considered as giving a terminus
ex quo for ascertaining the date of the publication of the

108 THE JOURNAL OF PHILOLOGY.

Rhetorica’; a date which is incidentally of much weight in —

determining the relation between this work and the De Inven-
tione of the youthful Cicero. The person whose public career

it so succinctly sketches has been understood by the editors, —

though not universally by the copyists before them’, to be
L. Cornelius Sulla. The last words of the example have con-
sequently been referred to his second consulship in B.c. 80, and
no allusion to any event of a later date than this has as yet been
discovered in the four books of the treatise.

If then this work was written or even published shortly
after B.c. 80, we might expect to find in it at least occasional
allusion to the stirrimg and crowded events of the years imme-
diately preceding. Strange to say, the last event before that
date which can be clearly shown to be alluded to, is the murder
of Sulpicius in B.c. 88°: and though references are frequent to
occurrences of the Gracchan and Marian periods, these eight
years of terror and civil war are entirely passed over. It is of
course perfectly possible that the author may have had a
reason of his own for this at which we cannot now arrive; but
there are two further difficulties, wanting as yet satisfactory
solution, which have led me to suspect that the book has
been post-dated by several years, and that the ‘consul factus
est’ of the passage quoted above does not refer to Sulla or to
B.C, 80, but to some other person, and to an earlier date.

The first of these difficulties can here be only briefly indi-
cated. It is now generally believed, on what seems satisfactory
evidence, that Cicero in writing his fragment de Inventione
had the Rhetorica ad Herennium before him‘; and on the
supposition that the latter work was published in or after the
year B.C. 80, the date of the former should be at earliest B.c. 78

1 See the Preface to Kayser’s edition,
p-. x1. Kayser is followed on this point
by Teuffel, Rom. Lit. sec. 149, 2. Blass,
Griech. Beredsamkeit, p. 121, and
Prof, Wilkins, Introduction to Cic. de
Or. p. 52.

2 This is shown by the introduction
of the word ‘ tribunus’ into the text of
one MS., as is explained below, and

in the reading ‘ Africam’ for ‘ Asiam,”
which, though undoubtedly wrong,

shows that the copyist was thinking ~

of Marius. 7
3 1.15, 25; iv. 22. 31.
4 Kayser Pref. x1; Spengel, Rhein,
Mus. xviii. p. 495.
R. G. vol. v. p. 230.

,

i |e ae

‘
ee

Cp. Drumann, |

~~)

RHETORICA AD HERENNIUM. 199

or 77, 2.e. at a time when Cicero was close upon thirty years
old*. But Cicero himself, in a well-known passage in the De
Oratore (I. 2. 5), writes of his earlier rhetorical work as com-
posed ‘pueris aut adolescentibus nobis’; and however loosely
he may be expressing himself, it is at least surprising that he
should describe himself as still so youthful at a time when he
had already made his mark as an orator, and was on the point
of entering on his public career as a magistrate’. Kayser has
tried to get rid of this difficulty by supposing that the four
books of the Rhetorica were not all published at once, and that
Cicero may have had access to the first three before the publi-
cation of the fourth, which contains our passage with its alleged
allusion to Sulla’s consulship in p.c. 80. The guess is inge-
nious and far from unreasonable; but it becomes at once
superfluous if it can be shown that the year 80 is wrongly
assigned as a terminus ex quo for the date of the whole
work.

The other difficulty simply consists in the fact that the
second example of ‘breuitas’ in the passage quoted above
cannot be tortured into yielding any allusion to Sulla or to the

year B.c. 80 without severe treatment and a very imperfect

result. If the reader will follow it closely with the aid of
Kayser’s critical apparatus and of an exact knowledge of the

_ chronology of the period, I believe he will find, that the

following propositions, taken together, will place it almost
beyond a doubt that Sulla is not the person meant to be
referred to :—

1. If Sulla be meant, the antithesis of the ‘modo’ and
‘deinde’ at the beginning of the example will be entirely
lost: for the words ‘modo consul’ can only refer to Sulla’s
first consulship in B.C. 88, while ‘deinde primus erat ciuitatis’
must be forced to refer to his mastery of the situation at Rome
that same year, after the death of Sulpicius and flight of
Marius to Africa. He was consul during the whole of the year
88, and ‘primus ciuitatis’ during the latter half of it, and
it can hardly be supposed that a careful writer, framing an

1 He was born Jan. 3, B.c. 106, 2 Pro Quinctio, sp.c. 81; Pro Rose.
Drumann, p. 216 reff. Am. 80; election to questorship, 76.

200 THE JOURNAL OF PHILOLOGY.

example of ‘breuitas’ with great exactness, would point this
relation of part to whole by using the words ‘modo’
‘deinde.’ | 7 |

2. The first sentence of the example runs thus in the best
Mss.: ‘Modo consul quodam is deinde primus erat ciuitatis’.
Now for the obviously corrupt reading ‘quodam is’ no correction
has ever been suggested, so far as I am aware, on the supposi-
tion that Sulla is the person alluded to. A single Ms («) of the
best family has ‘quondam tribunus’, which was corrected by a
later hand to ‘quondam tribunus is’. Even if ‘ tribunus’ were
the true reading and not an obvious interpolation, it could not
of course refer to Sulla, who never was or could have been
tribune.
awkwardness with which it breaks in on the contrast of the
‘modo’ and ‘ deinde’.

3. The next words are ‘Tum proficiscitur in Asiam’, If
Sulla be meant, they are curiously inexact for an author writing
so near the event. Sulla went in B.c. 87, not to Asia but to
Greece; he did not cross the Hellespont’ till early in 84, and
remained in Asia but a very short time. The words which
follow (‘deinde hostis est dictus’) may on the other hand refer
easily enough to the outlawry of Sulla by the Senate before his
return from Asia—probably on his refusing to give oF the
command to Flaccus in B.c. 85.

4. In the concluding words of the example, if Sulla be
meant, the word ‘imperator’ is meaningless, or at least needs a
justification which it has not yet found. It is true that Sulla
was technically ‘imperator’. from the day on which he left
Rome in 87, till the day he resigned his Dictatorship in 79.
But the word, if here used of Sulla, must be meant in some

such special and extended sense as was afterwards given it by —

and we have no evidence whatever that Sulla in
Both Sulla and Cesar

as translating

Augustus ;
this point anticipated the Empire’.

understand yéuovos
‘ Dictatoris’

1 Liv. Epit. 81 to 83: Fischer Zeit-

and

But this reading is put entirely out of court by the ©

tafeln, p. 184.

2 In the inscription on the famous
equestrian statue, which (according to
Appian, B. C, 1. 97) ran, ‘ Kopyy lov
DvAAas Tyéuovos evrvxovs’, we must

and not ‘Imperatoris’,
since the latter would have been
rendered avroxpdtwp by a Greek writer.
See Mommsen in Corp. Inscr, Lat.
i. 163.

ara fasts caanuaiss tosh asimlianaiacienrecncemaaae

ro)

A

1 i OR eG egy pe

RHETORICA AD HERENNIUM. 201

found the Dictatorship sufficient for their purposes, and stu-

diously avoided the invention of titles of doubtful Republican

orthodoxy.

5. The words ‘populi Romani’ which follow are not to be
found in any Ms, but were introduced by the Juntine editors in
1537 as a correction for ‘populorum’ which is given by all the
mss of the best family, as well as by a large majority of the
others hitherto collated. It is plain that these editors forgot
that they were dealing with an example of ‘breuitas’—res
tantummodo verbis necessariis expedita—, or they would have
hesitated before inserting an emendation so flat and superfluous.
On the other hand ‘populorum’ seems to have no possible
meaning, if Sulla be the person alluded to; for ‘populorum
consul’ is sheer nonsense, and if the ‘et’ be misplaced, ‘impe-
rator populorum’ is almost equally so.

These considerations seem to me to put it beyond doubt
that the author of the Rhetorica, who in this fourth book
framed his own examples’, never intended that Sulla should
here be understood as alluded to. A still closer examination of
the text and of the circumstances of the time will be found (if
I am not mistaken) to clear up all these difficulties, and at the
same time to elucidate the real meaning of the example. There
is but one other person whose history it can possibly be meant
to sketch, and it remains to be shown that the chief events in
the life of Marius correspond with it in every particular. These
were
1. Five consulships, in B.c. 107, 104, 103, 102 and 101, all
of which, it should be carefully noted, were spent almost
entirely in the field in Africa and Gaul.

2. Sixth consulship in B.c. 100, spent in Rome; where
Marius, with the help of Saturninus and Glaucia, was omnipot-
ent during the entire year, or at least until the death of
Saturninus, which did not take place till December.

3. Departure for Asia in the year following, on the recall

of Metellus. He remained in Asia for several years in a private

capacity, and on his return served without distinction in the
early campaigns of the Social War.
ET ee: eee doves 2 a la
Journal of Philology. Vou. x. 14

202 THE JOURNAL OF PHILOLOGY.

4, Struggle with Sulla for the Asiatic command in 88,

Saabig in his flight to Africa and proclamation as a pune
enemy’.

5. Return to Italy after the departure of Sulla for Asia in
March 87; during the autumn of that year Marius and the

other generals of the populares are at the head of Italian

armies directed against Octavius and the Sullan government at
Rome.

6. Seventh consulship in January 86, held only for a few

days until his death.
I now present the example as I believe it should be read by
the light of these well-known facts.

‘Item: modo consul quotannis* deinde primus erat ciuitatis.
Tum proficiscitur in Asiam, deinde hostis est dictus, post impe-
rator populorum® et consul factus est’.

It will be found that the six landmarks in the public career
of Marius are here indicated with as complete exactness as could
be expected in an example specially framed to illustrate ‘breui-
tas’; viz. ‘Modo consul quotannis’ gives the succession of
consulships from 104 to 101, ‘ipsis tantummodo verbis neces-
saris’. 2. ‘Deinde primus erat ciuitatis’ gives the exact posi-
tion of Marius in B.c. 100. 3. ‘Tum proficiscitur in Asiam’
gives the voluntary exile of Marius in B.c. 99. 4, ‘ Deinde
hostis est dictus’ gives his expulsion from Italy in 88, after
which he was proclaimed ‘hostis’ by the Senate. His previous
return from Asia and service in the Social War are not indic-
ated, as neither could be called a turning-point in his fortunes.

1 Appian B, C, i. 60.

2 For the constant interchange of
‘quot’ and ‘quod’, see e.g. Corp.
Inser. Lat. vol. i, 1016; Festus p. 178
October equus,........ immolatur quod)
annis); Mr Munro on Hor. Od. ii. 3.
11, in number 18 of this Journal. By
this correction the sentence seems
to me to gain exactly that unusual
and antithetical incisiveness which we
should look for in an example specially

framed to illustrate breuitas.

3 I have here placed the ‘et’ after
instead of before ‘populorum’, At
first the natural correction seemed to
be ‘ post imperator et populorum pro-
consul factus est’; but the omission

in that case of any allusion to the

seventh consulship of Marius would
imply that the book was completed

before that event; and this, as I shall —

show, is very improbable.

tees it

fed

ee es
—> “ ~

ice) w

RHETORICA AD HERENNIUM. 203

5. ‘Post imperator populorum’ will be found to give with
_ exactness the position of Marius in Italy on his return from
Africa in the middle of 87*. He was furnished by the consul
Cinna with the proconsulare imperium and the fasces, doubtless
in order to give him, technically hostis as he still was, a definite
position in the eyes of his soldiers. Secondly, the army he
commanded, like those of Cinna, Carbo, and Sertorius at the
same time, was composed beyond doubt of the Italian populi
still in arms’, together with large numbers of the new Italian
cives who were discontented with the inferior position assigned
them by the Senatorial government in a limited number of
tribes. This campaign in fact, though commonly called the
first Civil War, was in reality only a new phase of the Social or
Marsic War; the new feature being that one party at Rome
was now heading the Italians against the other. If we had the
eighth and ninth decades of Livy, we should no doubt find, as
we may guess from his Epitomist, that throughout the war the
Italians were called ‘populi Italici’ or ‘ populi’ only; and in an
example of ‘breuitas’, framed by a person writing soon after
the war, it would be perfectly natural to term a general at the
head of an Italian army ‘imperator populorum’. 6. ‘Et consul
-factus est’ refers of course to the election (or rather appoint-
ment) of Marius as consul for the seventh time after the occupa-
tion of Rome by the united armies in January 86.

If these arguments are well grounded, it will be apparent
that in the seventh consulship and death of Marius, we have
a more natural and more reliable terminus ex quo for the date
of the Rhetorica, than the second consulship of Sulla. But if
January 86 is the latest point of time that can with any
certainty be considered as alluded to in the treatise, have we

1 This is expressly stated by Plu-
tarch (Marius 41) in a passage too
explicit to be the result of a misap-
prehension, and possibly derived from
Posidonius, who was in Italy a few
months later, and had an interview
with Marius on his death-bed (H.
Peter, Quellen Plutarchs p. 103; Plut,

Mar. 45). A parallel case of the illegal
assumption of ‘imperium’ will be
found in Sall. Catilina 36.

2 Liv. Epit. 80. (Samnium). App.
B. C. i. 67. (Etruria). Cp. Kiene,
Bundesgenossenkrieg p. 298 ; Momm-
sen, R. H. (Eng. Tr.) tii, 317 foll,

14—2

204. | THE JOURNAL OF PHILOLOGY.

any grounds for hazarding a conjecture as to the exact date of ‘
its appearance? I am inclined to think that the year in which —

the work was actually completed was 84, on the evidence of the
other example of breuitas, which immediately precedes the one
I have been discussing :

‘Lemnum praeteriens cepit, inde Thasi praesidium reliquit,
post urbem Lysimachiam sustulit, inde pulsus in Hellespontum
statim potitur Abydo.’

Corrupt as these words are, they will help us, if ‘ Abydo’ is
the true reading; for comparing them with Appian (Mithrid.

56), we may guess that they refer to the movements of Lucullus ;

and the fleet co-operating with Sulla early in 84°. If this be
so, the precision of the writer in detailing the movements of
the Roman admiral, together with the fact that the passage is
almost at the very end of the work, suggest a probable comple-
tion and a possible publication very shortly after the news of
these events reached Rome’.

Whether or not this be the exact date, we are now in any~
We no longer have an entire ©

case rid of our former difficulties.
absence of allusion to any event in the eight years between the
death of Sulpicius and the second consulship of Sulla. We are
no longer surprised that Cicero should have written of the De
Inventione as the work of a mere stripling, for if our reasoning
is correct, it might well have been written when he was no
more than five- and! -twenty, and before he had come under
public notice as the defender of either Quinctius or Roscius of

Ameria. And lastly the conjecture of Kayser as to the delay in
the publication of the fourth book of the Rhetorica may now be —

safely dispensed with.
I may add that it seems to me by no means impossible that
the author of this work, who had made no secret in it of his

1 Lucullus, according to Appian,
seized Abydos in advance of Sulla, in
order to secure the safe passage of the
Hellespont for his chief.

2 Kayser (notes, p. 310) sees in iv.
52. 66 (example of sermocinatio) an
allusion to an outrage at Larinum

adverted to by Cicero (pro Cluent; 8.

25), which must have occurred in 83,
after Sulla’s return to Italy. But this
cannot be proved from a comparison of
the two passages; for in the former

no names are given, and in the latter

no details.

-.
f
F

1

RHETORICA AD HERENNIUM. 205

sympathy with the cause of the populares and Italians’, may
have perished in the Sullan reign of terror which followed close
on its publication. This would account in some degree for the
mystery which has shrouded its authorship, and for the fact
that we hear of no subsequent work by the same hand. It is
at any rate quite fruitless to attempt to identify the author
with any individual known to have been living at a later date,
whether bearing the name of Cornificius or any other.

1 See e.g. iv. 9. 13, 22. 31, 34.46, suggests Italian and Marian associa-

55. 68. The name of Herennius, to tions. Plutarch Marius ch. 5.
whom the book is dedicated, also

W. WARDE FOWLER.

DISSIGNARE.

x

Tus word, which I hope to shew should be carefully dis-_

tinguished from designare, is treated even by the most recent
lexicographers (Lewis and Short, and Georges in his seventh
edition), as an alternative form of that word. It is indeed
agreed on all hands that dissignator is the right word, not desig-

nator, for the official who presided over funeral or other games:

- but Iam not sure whether the verb dissignare has been set in
its right connection with dissignator and dissignatio. I suppose
that dissignare as implied in the words dissignatio and dissig-
nator must mean literally to mark out or arrange in different
directions, and so to order or dispose, while designare means to
mark out in one direction, or to mark out in a single line, and
so plan or design. Dissignare has sometimes, I think, been

corrupted into designare, but I doubt whether the converse has — 4

taken place. Vitruvius constantly uses destgno in its proper
sense, and if Miiller-Striibing’s apparatus criticus may be trusted,
the manuscripts do not give dissignare for designare in a single
instance. Nor is there any confusion between the two words in
the manuscripts of Vergil, who twice uses designare of marking
out the walls of a city. Again, is consul dissignatus ever found
for consul designatus? In two passages, however, of Cicero’s
De Natura Deorum (1 § 26, 3 § 85) manuscripts of good charac-
ter give dissignart and dissignata where designart might be
defended. 1 26 Anaxagoras...primus omnium rerum descrip-
tionem ac modum mentis infinitae vi ac ratione dissignari atque
confici voluit. 385 Ut enim nec domus nec res publica ratione

quadam ac disciplina dissignata videatur si &c. In both pas--

DISSIGNARE. 207

sages designari might perhaps stand, but dissignari, to be
arranged or disposed, seems to give a better sense.

But in the following passages (some of which, from the
Christian writers, I owe to Paucker) there can be little doubt
that dissignare is the right reading: and it will be interesting
to follow its various meanings.

Corpus Inscriptionum Rhenanarum no 161 (Brambach) NI*
DISSIGES7NS NIVEVENS TISVOVIS. Brambach would
read neu discindatis neu violetis opus: Georges quotes neu dis-
_ sigilletis: why not neu dissignetis? do not unseal or tear open:
comp. Augustine De Moribus Manicheorum 13 30, signaculi
dissignator, he who tears off the seal. Metaphorically Augus-
tine uses dissignare for to violate, outrage a custom: C, D, 15
16 2 qui (mos) cum...immoderationem continentiae coerceat,
eum dissignari atque corrumpi merito esse nefarium iudicetur.

Manuscript evidence which is above suspicion also gives us
dissignare and dissignatio in the sense (apparently) of tearing
open or divulging unlawfully what ought to be kept sacred and
secret, so to utter something wrong or unlawful. Plautus Most.
413 (Ritschl) according to B, though the editions give desig-
nata: quae dissignata sunt et facta nequiter: Apuleius M. 8 28
(so F') quasi contra fas sacrae religionis dissignaverit aliquid:
Arnobius 1 63 nec reputandum putavit quid ille dissignasset,
dummodo suis ostenderet se. 7 6 si quid animal caecum atque
in nubibus semper iguorationis incedens dissignaverit, dixerit
qui illorum minueretur auctoritas ; (is diwerit a gloss?) 7 9 abo-
litionem dissignationibus comparari. And Porphyrio on Horace
Epistles 1 5 16 says dissegnat, aperit.

Whether this is the right interpretation of this passage
may, however, be doubted. For Nonius p. 96, according, not to
the editions, but to the excellent Harleian manuscript, of which
my friend Mr J. H. Onions has just made a full collation, has
the following note; dissignare, cum nota et ignominia aliquid
facere. Terentius Adelphis (87) Illa quae ante facta sunt
Omitto; modo quid dissignavit? where the Bembine has desig-
navit, but the other good manuscripts dissignavit, and Donatus,
according to the printed editions, says designare est rem novam
facere in utramque partem...designatores dicti qui ludis fune-

-“‘pribus praesunt.— But we know « on re ainhecipoeee “very
inscriptions that. these functionaries were called dissign e
we need therefore feel no hesitation in n reading dissignarit both
in Terence (as Nonius would have us) and in the note of Dor

might readily come to mean to perform any startling or viol
act, any act which upsets the existing order of things: and
is exactly the sense required in the often-quoted line of Hora
‘Of what miracle is not intoxication capable?’ Operta recludit, —
Spes iubet esse ratas, ad proelia trudit inertem, Sollicitis anime ;
onus eximit, addocet artes. ayes

ie ny er, Se 5 a

a
> n

THE CHRONOLOGY OF THE BOOKS OF KINGS. |

THE enquiries of Néldeke’, Wellhausen? and Krey*® have
established the artificial character of the Hebrew Chronology
from the Exodus to the Return from the Captivity. There are
480 years from the Exodus to the founding of the temple, and
480 according to the list of Judaean kings from the founding of
the temple in Solomon’s fourth year to the Return. 1 note that
the epoch-making year does not reckon as the close of the old
but as the beginning of the new cycle. Thus B.c. 535, the year
of the Return, is the first year of the new theocracy. Further it
appears that where historical data failed the chronological in-
tervals were filled in, as appears most distinctly in the period of
the Judges, by numbers based on a generation of 40 years as_
the unit. The system as a whole is later than the Return,
which is its fixed starting-point, and Wellhausen has shewn
in his edition of Bleek’s Hinleitung that 1 Ki. vi. 1, the key
verse of the system, is late and did not stand in the LXX. It
thus becomes a point of great interest to determine which of |
the detailed dates, especially in the period of the kings, are
traditional, which systematic,

Now in the earlier Judaean reigns the only dates other than
those of accessions to the throne refer to the temple, its plun-
dering by Shishak, the change of the system of temple revenues by
Joash. These dates are not systematic, and doubtless are derived

®
1 Untersuchungen zur Kritik des edition, p. 264 sq.; Geschichte Israels,
Alten Testaments, p. 173 sq. p. 287.
2 Jahrbb. f. Deutsche Theol., 1875, 3 Zeitschr. f. wiss. Theol. 1877, p.
p. 607 sq.; Bleek’s LHinleitung, 4th 404 sq.

‘ ae }

210 THE JOURNAL OF PHILOLOGY.

from the temple records, from which we have several curious
and valuable extracts in the books of Kings. They deserve
therefore our special attention. The change in the revenue
system is dated in the 23rd year of Joash, It was a very im-
portant change, tending towards the centralisation of the
hierarchic system, by bringing funds that formerly belonged to
the whole priestly guild under the immediate control of the
high priest, and it continued in force in the days of Josiah.
According to the present chronology this change took place in
the 161st year from the founding of the temple. It marks the
commencement of the second third of the cycle of 480 years,
Again if we reckon 160 years from this epoch we come to the
year of Hezekiah’s death. The first year of Manasseh, whose .
reign is characterised as the decisive cause of Judah’s rejection,
begins the last third of the great cycle, the period of decline and
captivity.

In the first third the details are filled in by the aid of the
number 40, subject to the condition that 37 (= 40-3), the part
of Solomon’s reign after the temple was founded, and 22
(=20+2) years of Joash are fixed data. This requires one
period of 20 years, which is assigned to Jeroboam and Abijah,
one of 41, which goes to Asa, and one of 40, which is the period
of the influence of the house of Omri on Judah—Jehoshaphat
to Athalia inclusive.

In the second period Hezekiah’s reign was fixed at 29 years
(30—1), by the fact that Sennacherib attacked Judah in his
fourteenth year, and that fifteen years were added to his life
after the sickness which occurred “in these days.” Again Joash
reigning a round 40, 18 years of his life (20 — 2) belonged to the
new era. The other reigns had to supply a 2 and a 1 in the
unit place, an 8 and a 8 for the tens. Accordingly Amaziah and
Azariah give 81 years, Jotham and Ahaz 32.

In the third period the length of Zedekiah’s reign (11 years)
was known; for 2 Kings xxv, 8 is confirmed by Jer. xxxii, I,
2 Kings xxiv. 12 and fixed by the synchronism of Nebuchad-
nezzar. The length of the captivity was also known to be fifty
years (585—536 inclusive), for in Babylonia dates were carefully
kept. Now 160—50 gives 110 years for the reigns from

eee

CHRONOLOGY OF THE BOOKS OF KINGS. 211

Manasseh to Zedekiah inclusive. The length of Josiah’s reign
was also known to be 31 years from Jer. xxv. 1—3. On these
data 11 was the natural factor by which to subdivide the reigns,
and we find Manasseh=5 x11, Amon and Josiah=8 x 11,
Jehoiakim=11. If the last clause of Jer. xxv. 1 is genuine,
the 11 years of Jehoiakim are also confirmed by a synchronism
of his fourth year with Nebuchadnezzar’s first; but the clause
is wanting in the Septuagint, as is also the name of Nebuchad-
nezzar at ver. 9 and xlvi. 26, According to Jer. xlvi. 2 the
fourth year of Jehoiakim was the date of the battle of Car-
chemish when Nabopolasar was still alive. In any case the 55
years of Manasseh are suspicious. They have been challenged
by Wellhausen on independent grounds.

The kingdom of Ephraim also lasted 240 years. Wellhausen
and Krey reckon 242, and then correct the number by two
years with the aid of the Jewish synchronisms. This process is
arbitrary, since we know that these synchronisms are not part of
the original chronology; it is also unnecessary, for the number
of 242 is got by allowing a year for Zechariah and Shallum, who
have no more right to be counted than Jehoahaz of Judah,
who is not reckoned in Jer. xxv. The true sum is 241, and
the epoch making year of Samaria’s fall (the 9th of Hoshea) must
be deducted, as in the case of the Judaean periods. The kingdom
thus lasted 240 full years.

Now the first sure date, not an accession, supplied by the
northern history, is the commencement of the great Syrian wars.
There were two years’ campaign under Ahab, then a year of
rest; and in the following year, the third year from the fore-
going campaign according to Hebrew reckoning, Ahab was
killed. Four years of Ahab’s reign belong to the Syrian wars.
Now Ahab on the present chronology died in the 84th year
from the division of the kingdom. The Syrian wars commence
therefore in the 81st and open the second third of the whole
240 years of the kingdom. Again the 79th year of the Syrian
wars is the last of Joash. Jeroboam II. succeeded and com-
pleted the deliverance begun by his predecessor (2 Kings xiv, 28
with xiii. 19). One year of Jeroboam is thus reckoned to the
Syrian period, and his whole reign is 41 years. The last 80

212 THE JOURNAL OF PHILOLOGY.

years consist of 40 years of glory under Jeroboam and 40 of
decadence to the year of Samaria’s fall exclusive.

In this reckoning it is somewhat disturbing that one year of
Jeroboam II. goes to the Syrian wars. This however is con-

nected with a variation in the tradition. In 2 Kings xiii.
Joash, but in 2 Kings xiv. Jeroboam, appears as the deliverer of —
Israel and the restorer of the lost territory on the Syrian

frontier. To reconcile these statements it was necessary to
- take part of Jeroboam’s reign into the Syrian period; but as
three campaigns of Joash were recorded to have recovered the
lost. cities (xiii. 25) one was enough for his son.

The eighty years’ period for the Syrian wars seems-however

to be older than this adjustment, and to belong to the cycle of |

prophetic narratives from which the fundamental date in Ahab’s
reign is derived. For it is noteworthy that Elisha dies just
before the three campaigns of Joash. But in like manner
Elijah first appears three years before the Syrian wars. The
- 80 years of war would thus on the present chronology correspond
to a 79 years’ ministry of these prophets. The discrepancy of
one year between these periods appears to be connected with
the variation in the tradition as to the close of the Syrian
wars, and naturally suggests the conjecture that an earlier ad-
justment of the chronology gave Jeroboam only 40 years and
assigned the odd year to an earlier king, so as to make the
Syrian wars end with Joash’s three campaigns. On this scheme
we get an eighty years period for Elijah and Elisha. It is not
unlikely that this eighty years’ prophetic period was the basis
of the chronology, since all the numerical data apart from ac-
cessions belong to the prophetic cycle. If so it is also possible
to explain as systematic the numbers given to the individual
kings within the period: We have

Elijah under Ahab 7 years a _ War under Ahab 4

Ahaziah 2 [3] |
Joram eat
40 $ 19 [20] } 183+7=20o0r <¢
Jehu 28 [20] : 5a ud
Jehoahaz 17 ]
Elisha under Joash 13 J | Joash 16

But on the hypothesis one of these kings has to get the year

RS ae ey Ae

A
\

CHRONOLOGY OF THE BOOKS OF KINGS. 213

¢

withdrawn from Jeroboam II. It belongs either to Ahaziah or
to Jehoahaz. I apprehend that Ahaziah is the right person, for
3 and 7 are the usual numbers in the prophetic narratives
(3. years and 7 years of famine, 42=6 x 7 children, 3 years’
peace, &c.), and Joram has 3 x 4, Jehu 7 x4 years. Ahaziah
was afterwards reduced to the normal 2 of the short reigns in
the finished scheme, and a year was given to Jeroboam II.

The construction of the first 80 years is so far on the same
model that 22 years of Jeroboam I. +18 of Ahab= 40 or half
the entire period. But of course we cannot expect to find a
uniform system carried out through all the details of the
Chronology. The conclusion to which the present observations
point is that the existing chronological scheme was obtained
by setting down a small number of dates given in the old
records as fixed points, and filling up the intervals by a system
of interpolation in which 20 and 40 were the main units.
But the details were necessarily subject to given determining
conditions, for it was known in a general way that some kings

had long reigns, and others short. We might have expected

that it would also be thought necessary to preserve those
synchronisms between kings of Judah and Israel, which were
given for example by the deaths of Joram and Ahaziah in the

-revolution under Jehu. The fact that these synchronisms are

not observed, and that the band which finally added the
detailed synchronisms of accessions in the North and South
accomplished his work only by the highly arbitrary mode of
calculation, which Wellhausen has explained in Jahrbb. f. D,
Theol., 1875, confirms the arguments adduced in this paper
to shew that the main lines of the Northern and Southern
Chronology were originally drawn from mutually independent
data.

W. ROBERTSON SMITH.

ON THE TEXT AND INTERPRETATION OF CERTAIN
PASSAGES IN THE AGAMEMNON OF AESCHYLUS.

THESE remarks on the Agamemnon of Aeschylus are sub-
mitted in an honest desire to throw light into some of the dark
corners of this greatest of ancient tragedies. They are made in
the full knowledge of the fact—of which indeed only very
superficial scholars can be ignorant at this day—that there
are still many passages of the Agamemnon which no skill of
scholars has ever been able to clear up, and which will probably
always remain a battle-ground for critics.

There is one source of knowledge to which many will think
it is no longer of any avail to turn for new light on Aeschylus :
I mean the manuscripts. The list of these is easily given, so
far as the Agamemnon is concerned. The Medicean with its a
two copies, all sadly mutilated and containing less than a a
quarter of the Agamemnon; the two Venetian fragments; the q
Florentine and the Farnese, the only two which contain the a |
whole tragedy ;—these are the whole. And it might reasonably %.
be thought that the careful collations of the older scholars had _ a
exhausted the resources of these few manuscripts and left them oe
(to use Bentley’s expression) like “squeezed oranges.” I will |
first give a few examples to show that this is not entirely a
correct. A short inspection of the Codex Venetus (616 in the =
Library of St Mark), containing Agam. 1—45 and 1095 to the =
end, showed that some gleanings yet remained in that fragment. t.
In vs. 1196 this MS. reads plainly 7d ya) 5év, ive. rd wr) eiSévan,
Although this reading is adopted in many modern editions,

j she
ee ee
7 . Seen ie

THE AGAMEMNON OF AESCHYLUS. 215

it is always given as an emendation (see Paley’s and Weil’s
notes). Hermann says: “Omnes [1.e. codices], To pw’ etdévat...
Apertum est aut deesse negationem, aut ineptum esse Adyo.”
Others, as Schneidewin, accept Té pw’ eidévac on the authority of
the MSS., and explain or emend to avoid the inconsistency
which Hermann points out. Again, in vs. 1127 many editors
accept werayxépw in the belief that this is the original reading
of the Medicean, which now has perayxépwe. with v written
over the final «. But the first reading of the MS. was clearly
-wv, which was made -w: by correction and was afterwards
restored by a third hand. Recent editors doubt whether rop-
Gciv or moety is the reading of the Codex Florentinus in
vs. 342 (see Hermann’s and Paley’s notes), and Hermann
- accepts 7ro@eiy partly on the authority of his collation of that
MS., saying “idque ex Flor. mihi enotatum est.” But vodeiv
‘is really found in no MS. at all, the Florentine (like all the
others) having zrop@eiy beyond question. Hermann cites the
Codex Florentinus as authority for the singular reading in
vs. 345, Geois & av aumdaxntos et uoror, where I have copied
the reading of this MS. (I think correctly) Oeots 8 avapmaraxn-
tos. Iam at least confident that there is no breathing or other
mark over the syllable aw. I can hardly believe that Her-
mann’s reading could ever have been adopted into any text had
it not been for this supposed authority. Apart from the sense,
av (belonging to yévotro) would be in an absolutely anomalous
position thus imbedded in the protasis, which could be defended
by none of the ordinary examples of double or triple dv in long
sentences, still less by the formula ove oida dv ei, as in Eurip.
Med. 941, ov« oi8 dv ei weicayus. Besides, the sense of the
MSS. reading, Oeots 8 dvapmdaxntos ef pddor otpatis, but
(even) supposing the army to reach home without offending the
Gods (as suggested in vss. 83838—842), seems best suited to the
thought of the following lines, in which Clytemnestra darkly
hints that a reckoning awaits the victors after their arrival at
Argos, even though they may not incur new wrath of the Gods.
by sacrilegious plundering at Troy.

The passages which I have selected for discussion belong
chiefly to the large class in which it seems to me that the

216) °.% THE JOURNAL OF PHILOLOGY.

readings of the manuscripts have been needlessly called in

question, and my object is therefore in great part a defence of
the manuscript text. In many cases I fear that my attempt
will seem both heretical and abortive to older students of
Aeschylus, who have generally assumed that certain passages
are corrupt, and to whom the emended text has in a measure

become the vulgate.

1. Voss. 105—107. ére ydp OedOev Katarveles med pord-

Trav, adKxav cvudutos aidv. Most recent editors read podmdy
depending on 7rev@o), and add depending on ovudutos, omit-
ting the comma. Hermann reads axa and retains podrrap,
but he takes d\Kd cvpdutos aidy in the sense of the time that

the war has lasted, and puts it in apposition with wew@o to

express id quo niteretur ea fiducia. Other interpretations may

be found in Paley’s and in Weil’s notes. It seems to me that |
the emendations are far more difficult to explain than the

reading of the MSS. as given above. In this reading it is hard
to see what there is in either sense or construction to which
almost all editors have taken exception. The asyndeton and
the chiastic order both suit the sense, and we may translate as
follows: “For still (i.e. after these many years of waiting)

persuasion from the Gods inspires me with song; still even my ©

old age (literally ‘the time that has grown with me’ for ‘the
time that I have lived’) inspires me with strength (to sing).”
The first clause was clearly so understood by the Medicean
scholiast who-says: eiOev yap we n mapa Oedy tictis pédtrELV
Kal reve ott ev TpaEovow of ’Atpetdat bcov amo Tov onpelov.
The meaning of cvpduros aiwy (sc. wos) and the construction of
G\xav With catarveles are indicated by the succeeding scholion:
6 yap ctpduTos pot aidv—6 éott TO ynpas—6iad THY Eis Beovs

\ , \ > \ a, GF 2 > \ /
TeLdw KOATHV fol KAL AAKNY KATAaTTVEL O EOTLV, EL KAL YEPO@V —

SING AF L \ 3% , \. of > / 2A
Elpl, OMS MEM Ta yeyovoTa’ TéTroLla yap OTL Eis TEPas avTa
a&ovow ot Geol. The words dud...7es@o here show a reading
and interpretation of the first clause which we cannot reconcile

with any possible form of the words zei6d podmav. Paley

suggests that this scholiast may have read zev@ot, but his
version would require also Kal a@dKxay or adxav te. In the

Medicean zev6d and podmav have been changed by a later

i ~ Tae «>
er ee ‘ey | see y _—

=e ~ ' a f- ee a

THE AGAMEMNON OF AESCHYLUS. ya ey >

hand to 7e:06 and porrday. Weil gives wordy as the reading
of the first hand; but I feel confident that my own collation is
correct here. Perhaps wes06 may confirm Paley’s suspicion
about esOo?. The use of cvpdutos aiwy (sc. wot) in the sense
of the time (or age) which has grown with me is well illustrated
by Agam. 894: audi cot waOn Opdca TrELw TOU TUVEVSOVTOS
ypdovou, i.e. more accidents than could have occurred during the
time I was sleeping (the time sleeping with me being used for
the time I was sleeping). See also Eumen. 286: ypovos xa-
Oaipes wavta ynpaoKkwy ouov. Hermann quotes also Soph.
Oed. Col. 7: 6 ypdvos Evvdy paxpds, and Oed. Tyr. 1082: ot
ovyyeveis unves. It may be added that in the former clause érz
means even now, after ten years’ waiting for the fulfilment of the
predictions, referring to the omen of the two eagles and the
hare, of which the chorus are about to sing, and the interpreta-
tion of it by Calchas; the faith of the chorus in the Gods and
in the ultimate fulfilment of the predictions stl remains
unshaken. In the second clause érz refers to the chorus sézll
having strength afforded even by their old age, ef cal yépwy
ejut. It is perhaps hardly necessary to say that the whole
passage in question, érs...aiav, is a pure parenthesis, the follow-
ing Omws...méumec being the development of the idea first
expressed by od.0v xparos, etc., in the leading clause.

2. Vss. 249—254. No passage in Aeschylus has been read
and explained in a greater variety of ways than this. Between
Hermann’s to mpoxAvew 8 Hrvow mpoyaipétwo (TO péAXov
being joined with the preceding sentence) and Paley’s 76 péA-
rov 8 érret ov yévorr’ av dows, Tpoxatpéro, there is room for
an infinite amount of conjecture and ingenuity. A few recent
editors, Schneidewin, Weil, and Enger (1874), adopt a reading
which is essentially that of the Farnese MS. in all except the
last verse; but none, I believe, now venture to retain the
reading of the best MSS. through the whole passage. As the
text 1s so much in question, I give (from my own collation) the
exact readings of the three principal MSS. in the first part of
the passage. The following is the text of the Medicean (the
words and colon within the brackets being added by a later
hand in blacker ink): .

Journal. sf Philology. vou.x. TS

218 HE JOURNAL OF PHILOLOGY.

Aika 8é Tots perv trabovow pabeiy
€TLppeTres - :

TO pedrov|’ TO 5&€ mpoKAvew]
ériyévoit av KAVOLS TPOValpETo.

The Oxford fac-simile of this tinea (ed. by Merkel, 1871) |

fails to mark the interpolation in 70 6é mpokrverv, and no one -

(to my knowledge) has noticed that the colon after pédXov
is a part of the interpolation. Indeed, the total absence of
punctuation in the Medicean is an important part of the

record.
The Florentine MS. reads:

dixka 6é Tots pev Tabovew pabeiy,
, \
€mrippémres TO péANOV. TO S€ TpoKAvVEL, Eerrel
’ , , :
yévoiT av KAvOLS, TpoyatpéTo.

The reading of Ven. A (468), so far as it could be de-

ciphered, seemed to agree with that of the Florentine, and it is

so given by Hermann. In 1872 the words between mpoxdrvew

and mpoyaipérw were no longer visible, even in the sunlight.
The reading of the Farnese MS. is as follows:

Aika &€ tots pev rabotow pabeiv
b) / \ f
ETLPPETTEL, TO LENOV

3 U /
ével yévort av Krol, TpoyatpEeTo.

The words ro 5é mpoxdvew had evidently been introduced
into the text before the Florentine and Venetian MSS. were
copied, so that these latter have 7d wédAov joined with paeiv
evrippérret, While to mpoxdAvew takes its place as the object of
kdvols. But this construction of 7d mpoxdvew is as fatal to the

sense as the introduction of 7d 6€‘zpoxAvew into the text at

all is to the metre, which is in perfect agreement with that of
the strophe without these words. It is obvious that the only

construction which the original copyist of the Medicean could.

have had in mind is that which the copyist.of the Farnese MS.
(probably. Triclinius) adopted in his text, either by conjecture or
from some purer source than the interpolated Medicean text. Of
course, émruyévotr in the Medicean is only a slip.of the pen or

the ear for ézrel yévorr’, and we thus have the construction 7d-

P in “ — t > Fi ‘ a 2 ~ y 4 aos ee al ns
TS pee DS TIN 2h) peta ane iain 0 aa

isparg
ate

a
P= ve o

ae
“as

SS rr
= eer |
'

shennan aT
1

THE AGAMEMNON OF AESCHYLUS. 219

pédrXov rel yévourr av Kdvots, which requires only 6 after
pédXov to make both sense and metre complete. Davies objects
to this reading on the ground that éel yévour’ dy is not a
possible construction. But the construction is Té wéAXov KAvois
av érel yévouto, you can hear of the future when i comes, the
assimilating force of «Avors (a force which is especially strong in
poetry) causing what would otherwise be éeidav yévntras to
become ézel yévorro, This is like teOvalnv bre pot pyxére
Tavta wédov (Mimn. I. 2) and ws admondorTo Kal GdXos 6 TIS

—TowavTa ye péCou (Odyss. 1. 47), where assimilation alone makes

the optatives more natural. Indeed, this example is a strong
confirmation of the position on the whole subject of assimilation
and its effect on moods which is maintained in the paper on
“ Shall and Should in Protasis'.” There is the same difficulty
in translating yévorro here in English that is felt in translating
HéXoe or pétor, above; and for the same reason. The position
of av, where a comma might precede, is not objectionable so
long as To wéAXov, which is a part of the clause containing ay,
precedes the particle, and ézei yévour’ is only an inserted clause.
See Arist. Pac. 137: dX, @ wénr’, av pot oitiwv SitrA@v et.
The general principle that dv cannot be the first word in a
clause, even after a cumma, is subject to this limitation, not to
speak of others. :

The Medicean scholiast who wrote against vs. 249 rots pév
meTrovOcow 1 dikn Sidact TO pabeiv evidently had the original
construction in mind. But the following note, dixnv yap dovtes
pavOavovet TO wéAXov, must come from some one who joined To
pérrov with pabeiv in the text. When 7d pédXor is rightly
taken with the following words, it will also be the natural sub-
ject of rpoyaipéta, which To mpoxAvecy could hardly be.

In vs. 253 all MSS. and editions agree in icov 5€ TO mpoaTe-
ve. If the interpolated ro 5é mpoxdveuv is left out of the text,
TO mpoyaipew (sc. TO wéAXov) will be the subject; i.e. for the
future to be dismissed (bid farewell) before it comes is just as
well (tcov) as lamenting it before it comes, for -it will surely

~ come, whichever we do. When, however, 7d 6é mpoxAvety was

added, it was taken as subject here, and the meaning was sup-
1 Journal of Philology, Vol. vu. p. 33 sqq.
15—2

220 THE JOURNAL OF PHILOLOGY.

posed to be hearing the future beforehand is equivalent to be-
wailing it beforehand, on the ground that it must be full of
sorrow. The later scholiast on this verse has this idea when he
sayS: 6 yap TpoyvyvwoKwv TO wéddOV Kal Tpoctevater. Indeed,
it is highly probable that 76 5€ wpoxAvew was first written in
the margin as the subject (understood) of icov éoriv, as it only
adds confusion to all the other constructions.

A greater difficulty comes in the last line. Here there is
little or no dissent among recent editors from the emendations
of Wellauer and Hermann, ropoyv ydp i£e. ctvopOpov avryais, —
for cvvopOov avtais (Med. and Ven.). For ocvvop@ov Flor. and

Farn. have cvvapOpov. The words civopOpov avyais are some-

times understood as referring to the actual rays of the morning.
sun (just about to rise), sometimes to the metaphorical sunlight
which is expected to break upon the darkness of uncertainty in
which the Argives at home have been living. The objections
to suvopOdv avrais are, first, that cvvopOes does not elsewhere
occur, and secondly and chiefly, that avrais cannot be referred
to the distant réyvas without great violence to the sense and
still greater obscurity. But ovvopOds (or perhaps cdvopOos), -
though a dzra£ eipnuévor, is no more so than ovvopOpos, and is,
moreover, amply justified by the compounds dvop@os, upright,
with the cognate verb avop@da, set upright again, and éop00s
with é£op8ow. We have the verb cuvopOow in Arrian (see
Lexicon); and an adjective cuvop@ds, coincident with, would
naturally be expected. Compare cvvduidos, ovppetpos, avv-
ofvs, and other such compounds of ovv. A word thus
analogically formed, and found in the Medicean MS. of <Ae-
schylus, is not open to objection as a dmaé& eipnuévoy, provided —
it suits the sense of the passage. (See also § 6, below.) We

come now to avrais, which cannot be referred to anything

nearer than téyvat Kadyavros in vs. 248. But those terrible
words réyvas 5¢ Kadxavtos ove axpavtot, following the minute
description of the preparations for the sacrifice of Iphigenia,
and taking the place of an account of the sacrifice itself,
suddenly bring before the mind the awful reality which faces
the chorus as they think of the condition of things. These
words give unity to the whole choral song, and show more

THE AGAMEMNON OF AESCHYLUS. 221

plainly than any exact language could have done that the
Argive state now stands on the brink of a new gulf of horrors,
which may well exceed all the ancient horrors of the house of
Pelops. Let us trace the course of thought which runs through
the whole chorus, that we may see more clearly the exact
relation of the verse in question to the whole. The first stasi-
mon and the lyric parodos (from vs, 104) form in subject a
single ode.

The chorus first describe the omen which was seen as the
Argives marched forth to Troy, two eagles devouring a pregnant
hare. This Calchas interpreted as portending the capture and
destruction of Troy by the Argives. But, with an ominous
reserve, he fears only that some divine displeasure may cast a
gloom over the bright prospect; for Artemis is watching with
envious eyes her father’s winged hounds, the two eagles, and
the two sons of Atreus whom they represent, and she “loathes
the eagles’ banquet.” And as Artemis, the friend of all the
beasts of the field, is asking her father Zeus to fulfil what the
prodigy portends, the bad as well as the good, so the prophet
in turn prays Apollo to prevent his sister from detaining the
Argive fleet by any contrary winds, which he fears she may do
in her eagerness for “a new sacrifice, a lawless one, of which no
man can partake, a kindred worker of strife, that fears not
man.” “For,” Calchas adds with double significance at the
close, “child-avenging wrath (i.e. the wrath that avenges a
child’s murder) abides firm, terrible, ever rising afresh, haunting
(directing) the house, treacherous, ever remembering.” To the
Argive chieftains just setting forth for Troy this was terrible
enough, as reminding them of the vengeance that still was due
for the murder of the children of Thyestes; while to the chorus,
who quote it after ten years, it has gained a new and more
terrible meaning through the “new sacrifice” at Aulis. To the
chorus, therefore, and to the audience—who know even more
than the chorus—-these last words of Calchas pronounce the
doom of the guilty race. The vague forebodings of the prophet—
his fear lest some divine power might possibly darken the
prospect, lest Artemis might detain the fleet, lest this detention
might in some way cause “a new sacrifice”—had all been

222 THE JOURNAL OF PHILOLOGY.

-yealized in the fullest sense; a child, the darling daughter of — I

the King of Men, had been adveificaa to the father’s ambition;
and now nothing could save the race of Atreus from the double |
retribution of “ ‘child- -avenging wrath.” In this state of mind,
with the hope of victory thus darkly clouded by the sure a
proach of retributive justice, the ‘chorus again sing, in harmony
with the words of the prophet, aiAwov, aidwov eiré, TO & ev
vixatw (vss. 104—159).

The chorus now Sake the aid of Zeus, the only power
which can relieve them from the load of anxiety which
oppresses them. Ouranos and Kronos, the elder divinities,
are past and gone; but he who calls on the name of Zeus
with willing heart shall gain perfect wisdom. But the law of-
Zeus makes wisdom the result of suffering; the “trickling of

drops of torturing recollection before the heart in sleep” sobers

men often in spite of themselves. And it is on the whole a
gracious boon that this is so (vss. 160—183). |

Then, by a sudden transition, the chorus describe the conflict
in the mind of Agamemnon when he is told that his daughter's
life is demanded by the army as a sacrifice to appease Artemis
and still the opposing winds. He yields to the demand and
to his own eagerness for victory. Then follows the graphic
account of the preparations for the unnatural sacrifice, the
maiden’s prayers and cries to her father for help, the lifting
of the victim “like a kid” upon the altar, her falling robes,
the gags which checked her voice, and then her speechless
appeal to the heroes whom she had often seen as her father’s
guests; Iphigenia lies upon the altar, ready for the sacrificial
knife, “ beautiful as a picture” (vss. 184—246). But here the
chorus suddenly pause, and the last scene is left to be imagined.
They say:

“ But what followed we saw not, and we tell at noe But
[we do say] the prophetic arts of Calchas must bring fulfilment
(i.e. the vague horror of his predictions in vss. 147—155 must
surely be realized). But [it is only by actual experience that
we shall ever know what penalty is to be exacted for the
sacrifice of Iphigenia, for] Justice brings knowledge within the
reach of those [only] who have suffered (wader paios) ; the

THE AGAMEMNON OF AESCHYLUS. 223

future you can hear of when it comes; before that bid it
Farewell, and this is as well as to lament it beforehand ; for
[whatever we do] zt will come out clear and plain in full accord
with these (prophetic arts).”

’ It seems to me that no one can thus take a connected view
of the whole song without feeling that the interpretation here
given to the transmitted text of the last verses is not merely
possible but highly appropriate. There is a special force in
avrais, referring to the solemn words téyvas 58 KadXyavtos ovK
axpavrot with emphasis at the end of a sentence which begins
as parenthetical, but which thus leads the thought at the close
back to the poimt from which it digressed. The gender of
avrais, moreover, makes the reference to réyvas much clearer
in Greek than it can be made in English by our vague “them”
or “these.” Indeed, the ambiguity which we feel here can
hardly be said to extend to the Greek.

The emended reading ovvopOpov avyais, understood literally,
with the rays of the coming morning (orietur cum luce solis
eventus, Hermann), implying that the mystery is to be cleared
up at sunrise, cannot give the correct meaning if Td wéAXov has
been rightly explained above. For “the future” here includes
not merely the question of the capture of Troy (which was
to be decided at once), but also and chiefly the dreadful ques-
tion of the doom impending over the race which had spread
the Thyestean banquet and had sacrificed a royal princess on
the altar of its ambition. This last question, as the chorus
have said, can be decided only after the knowledge of the
future has come through suffering; it is this knowledge that
the chorus will bid farewell, for they have as yet no suspicion
of the immediate doom which awaits Agamemnon: on his
return. The thought furthest from the minds of the chorus
is that the coming dawn is to settle this terrible question.
This interpretation is therefore opposed to the obvious sense
of the preceding words. It is perhaps to avoid this that some
recent editors understand the “rays of dawn” metaphorically,
not of the morrow’s sunrise, but of the future emerging from
the darkness of futurity into the ight of.the present. In this
view we have merely a strong expression for “the future will

224. THE JOURNAL OF PHILOLOGY.

come to light plain and clear.’ As this cannot be called im-
possible, two questions arise: first, whether this interpretation -
is better suited to the whole sense of the passage than the one
proposed above, which adds the idea that the future which is
to come out “clear” must accord with the prophecy of Calchas;
secondly, whether, if this is preferred, it is so superior to the
sense afforded by the manuscript reading that it must be pur-—
chased by introducing into the text two conjectures, one a
atraé eipnuévov. I can hardly doubt what answer will be given
to these questions by unbiassed scholars, especially by those —
who will reconsider their opinions from the beginning on a pas-
sage about which they have already made up their minds.

I have felt that the importance of these verses, which’
determine the final turn of thought in one of the grandest of
lyric songs, and greatly affect the whole impression which the
ode makes, is a sufficient justification of the space given to the
discussion of them.

3. Vss. 931—943. These verses stand thus in the manu-
scripts (not to notice unessential variations) :

KA. xal pny 708 etre pr) Tapa yvopnv époi.

AP. yvopnv pev icbe pn StapCepodvT épé.

KA. nv& Oeots Seicas adv GS Epdew Tad ;

AT. eizrep tis, das y ev 108 e&eirov tédaos.

KA. ti 8 dy Soxet cot Upiapos, ef rad Avucev; (935)

AI. év qrouxirow dv Kapta pow BAvat Soxei.

KA. pn vuv tov avOpereov aloea tis aporyov.

AT, ale ye mévTou dnp0Opous péya oOéver.

KA. o 8 adovgrés Y ov émitnros méNet.

AT. ov Tou yuvaikds eat ipeipew payys. (940)

KA. ois & orPious ye Kal TO vixdoOar mpéret.

AT. 4 Kat od vixnv ryvde Sypwos Thess ;

KA. Oot" xpatos pévTou tapes y éExov poi.

In the interpretation of these much-disputed verses, I differ —
from Paley, where he has expressed his opinion, chiefly in
regard to vs. 933 (906 Paley); but it is impossible to discuss
a single verse of a orvyouvOia by itself. In the speech just

THE AGAMEMNON OF AESCHYLUS. 225

finished, Agamemnon has expressed a decided repugnance to
making himself a mark for divine vengeance, after his great
victory, by walking into his palace upon a path spread with
purple embroideries. He is well aware of his danger, already
hinted at by the chorus: tév modvKTOvav yap ovK doKoTrot
Ocot (vs. 461), and ro 8 vrepxoras eb Kdvew Bapv (v. 469) ;
and his mind cannot be entirely free from anxious recollections
of Aulis and Iphigenia. Clytemnestra, who is still more awake
to the importance of the crisis, is determined that her hus-
band’s last act shall be one of defiance against the Gods. But
it is a time for coaxing and for arguments (especially ad homi-
nem), not for open quarrelling with her husband. She there-
fore says (vs. 931): “Now don’t say you won’t walk on the
embroideries, and so go against my wishes.” I think that py
stands after its verb merely to make trapa yvouny éuol more
prominent and to show that the interference with her pet
plan for the king’s reception is what she has most at heart.
The poet says rapa yreunv éwol (rather than éunv) as he might
have said, apa yvdnv éwol éotiv, it 1s against my wishes or not
to my mind, opposed to cata yvopunv éwol éaotw. In this verse
yvounv means wish, hope (cf. Dem. Ol. 1. § 16, p. 14: adv te pur
kata yvodpnv éx3h); but in the next verse (932) Agamemnon
repeats the word with emphasis, giving it a slightly different
turn by the change in expression. He says: “As to yvopn,
please understand that I shall not let my purpose (yvopunv)
be weakened.” This leaves Clytemnestra where she began;
and she now tries a new style of argument, addressed to his
sense of shame: “Could you possibly have vowed to the Gods
in some time of fear that you would act thus?” The form

of the question implies, with bitter sarcasm: “Surely you,
Agamemnon, could never have had a moment of terror in
which yow could make such a vow!” Agamemnon has already

(vs. 924) said that walking on embroideries is éuol wev ovdapuds
dvev doSov. But he now replies with dignity and apparent
firmness: “If ever a man declared a decision knowing perfectly
what he was about, I have done it now.” Hermann says of
ténxos here: “Sic dictum ut sit pro decreto.” This reply suits
perfectly the meaning which I have given to the preceding

226 THE JOURNAL OF PHILOLOGY. ©

verse, and is not at all open to the objection which Professor
Kennedy (Journal of Philology, vu. 13, p. 17) makes to Mr

Paley’s similar version, that it “is no reply to the previous | |

words of Clytemnestra: it is a mere repetition of his refusal,
‘No, I won’t, in another form, rudely ignoring what his wife
had said.” Mr. Paley had omitted the interrogation-mark at
the end of vs. 933 (906) and translated: “You would have
vowed to the gods to act thus in time of fear, i.e. you are pur-
suing a course more like one in peril than a victor.” But if
we suppose Clytemnestra to have just suggested the possibility
(or rather the impossibility) of Agamemnon’s having been
frightened into a vow that he would act with humility if he
should ever capture Troy, the dignified reply of her husband ’
is Just what would be expected. |

A third argument is now tried. Agamemnon is asked what
Priam would have done if he had gained so glorious a victory ;
and he replies that Priam would undoubtedly have walked on
embroideries. After he has been further asked to disregard
human censure, and has replied that the voice of the people
still has mighty power, Clytemnestra tells him that it is not |
desirable to escape the @@ovos of men, for “ he who is unenvied
is not an enviable man,” i.e. he who escapes @Odvos is not the
object of Gros (is not €rAwrTcs). It seems as if Agamemnon
here decided that he was no match for his wife in “chopping
logic,” and that it would be better on the whole to make no
more ungracious objection to her plan for his reception; and
yet his scruples were by no means overcome, as appears in
vss. 944—949, below. He shows his disposition to yield (as
he had doubtless often yielded before) by saying: “ It is not
like a woman to be so eager for a fight as you are.” The
queen replies, now sure of her point: “It becomes the pros-
perous to submit even to defeat,” i.e. they can afford to yield”
a point like this. Agamemnon rejoins, partly in scorn, but
chiefly in jest: “Is this the kind of victory in a strife which
you hold in honor?” i.e. the victory (vikn) which consists not
in TO wKav but in 7d wxacbar. He speaks as if vixen could be
the equivalent of both 76 vxedy and 70 wixdc Oat, as Tipsy is of
both 70 tywav and 76 Tywao Oat, and asks his wife if she adopts

THE AGAMEMNON OF AESCHYLUS. 227

this principle for herself as well as for him, Professor Kennedy
translates this verse: “ Do you really care for victory in this
dispute?” This requires a change of ryvd_e to tHade, which I
cannot feel is necessary unless some objection can be urged
against the interpretation given above. Nothing now remains
for Clytemnestra but to ask that her husband’s compliance
may be not forced but willing.

I should thus translate the whole passage, following the
reading of the MSS. (as given above) :

CL. And now don’t say this and disappoint my wish (yvwpyv).

AG, My purpose (yvwunv) be sure I shall never weaken.

CL. . Could you ever have vowed to the Gods in any time of fear
that you would act as you now do?

AG. If ever a man declared a decision knowing well what he
was about, I have done it now. |

CL. But what do you think Priam would have done if he had
accomplished what you have? (935)

AG. Iam very sure he would have walked on embroideries.

CL. Now don’t be afraid of the blame of men.

AG. Yet the voice of the people has mighty power.

CL. But the lot of the unenvied man is not enviable.

AG. It surely is not womanly to be (so) eager for a fight. (940)

CL. But it is becoming to the prosperous even to let a victory
be gained over them,

AG. What! is that the kind of victory which you hold in honor
(for yourself as well as for me) ?

CL. Be persuaded (i.e. never mind which kind of victory it is) :
at all events let me prevail (here) by your consent,

4, Vess. 1025—1029:

ef O€ pn TeTaypEeva
pLoipa potpay éx Dewy

5 \ / VA

elpye fn TEV HEpeLy,
mpoplacaca Kapdia
yréooav av tad é&éyet.

Every student of Aeschylus knows how unsatisfactory are

-all the widely divergent opinions of editors on these verses.
Paley’s translation —-“ But if the appointed law of fate did

228 _ THE JOURNAL OF PHILOLOGY.

not hinder fate from getting further assistance from the
gods, my heart outstripping my tongue would pour out these
feelings” — seems to give a literal sense of the words in a
perfectly grammatical construction; and his note on the last
two verses shows, I think, that Schutz’s emendation Kapoiav
yAdooa is not only unnecessary but injurious to the sense.
But can we rest satisfied with this interpretation of the first
three verses? I trust that any suggestion on so obscure a
passage will appear better than none.

I think, first, we must certainly take potpa potpay in a
reciprocal sense, like a\Aos dAdov; and secondly, rréov Pépew —
must mean bear away more than its due, after the analogy of —
mréov éxewv, to have more than is due. IIinéov dépecOas is com--
mon in the sense have an advantage (cf. Soph. Oed. Tyr. 500:
méov 1) yo déperat); and a similar use of the active dépw is
familiar, as in Soph. Oed. Col. 651: ov« otv répa y av ovdéev
) roy hépors. The meaning of the passage will then be: “ But
did not one fate appointed by the Gods (sometimes) hinder
another (fate appointed by the Gods) from securing more than
its due, my heart would outstrip my tongue and pour forth its
present burden.” This seems to point to a doctrine of “ inter-
ference” between two lines of fate, by which either may be
checked or balanced in a course which would, if unhindered,
prove too destructive. The chorus would thus imply that this
last desperate hope is all that they can still see to warrant them
in hiding their feelings longer iad oxdt (vs. 1030). In this
song the gloomy forebodings of the chorus assume a more
definite form. The earlier songs have hinted darkly at coming
disaster; while the description of the sacrifice of Iphigenia,
the allusions to the slaughter at Troy, and the fears of the
consequences of human pride, all disclose grounds for the
gravest apprehension. But these fears are all vague and
general; now, however, after Agamemnon has entered his
palace, timidly zopdvpas 7arév, and Clytemnestra has assured
him in bitter irony that she has at her command the whole
Ocean to supply “ purple” to the royal house, the chorus feel
that a deed of blood is close at hand. They do not yet divine
its nature, least of all do they suspect that Agamemnon was

THE AGAMEMNON OF AESCHYLUS. 229

walking to his death; but there is “murder in the air.” The
general tenor of their song is as follows :

“ Why does this hovering phantom ever flit before my heart,
and why can I not spurn it and restore confidence to my soul ?
I have seen the Argive host set sail for Troy; and now with my
own eyes I have witnessed its return. But still my heart of its
own impulse sings the Fury’s lyreless dirge, and refuses to be
eacouraged by hope. And I know that this feeling within me
is not all in vain, and that it points to some fulfilment of my
forebodings; but yet I pray that my fears may prove groundless
and without result.

“Great prosperity is ever insatiate to extend its limits,
reckless of the close neighbourhood of calamity; and human
fortune as it sails onward often strikes a hidden reef. Yet the
sacrifice of part of the cargo to save the rest may keep the ship
from sinking and the fortunes of the house from falling, and one
plenteous harvest averts all danger of famine. But far otherwise
is it when the life-blood of a man has once fallen to the earth ;
this no incantations can recall. Were this not so, Zeus had never
stopped Aesculapius from raising the dead. My only hope is in
the thought that one line of fate fixed by the Gods may some-
times interfere with another line of fate (also fixed by the
Gods) and so hinder it from securing too much; were this not
so,—had I not this desperate hope to encourage me,—my
heart would outstrip my tongue and pour forth all its burden.
But, as it is, I can only hide my grief in darkness, sore vexed,
and with no hope of ever seeing order come out of this con-
fusion, while my soul is burning within me.”

The passage in question thus supplies an important link in
the chain of thought, and gives the ground on which the chorus
decide to suppress their feelings a little longer. The appearance
of Cassandra now gives a sudden turn to the play, and the
affrighted chorus are for the first time made aware of the real
danger which awaits them.

It may be said that no such doctrine of the interference of
two lines of fate as is here supposed can be found elsewhere in
the Greek religion. Even if this is true, I contend that such a
doctrine appears here by the only interpretation of the language

230 THE JOURNAL OF PHILOLOGY.

which is at once plain and consistent with the context. It can-
not be too clearly understood that the ideas of fate which make -
the Motpa: the superiors of Zeus, and the King of the Gods _
merely a helpless agent in their hands, are not Aeschylean,
The verses of the Prometheus (517, 518): |

XO. Tov’Twv dpa Levs éoruw ag Levert Epos; :
IIP. ovscovy av expat ve THY TETPOMEeVND.

represent only the threats of a defiant rebel against the viele
divine order of the world as this was established under Zeus; __
they refer moreover to a disaster which Zeus did avert by his . |
own free-will. Greek orthodoxy—certainly the orthodoxy of ==
Aeschylus—speaks plainly in the following verse (519), in which , va
the chorus indignantly ask

5) a “ a
Ti yap wérpwtar Znvi, wrAnv ael Kparety ; 1

The doctrine of Prometheus probably represents a more,
ancient and gloomy view of inexorable necessity ruling both
Gods and men, while the later view gave the government of the
world to a wise and beneficent personal ruler, the director of
other subordinate rulers, who had displaced a harsher dominion,
and whose laws were made for the best good of mankind in
general. These laws, however, the laws of nature, though bene-
ficent on the whole, were inexorable and unyielding, often
bringing misery upon the innocent children of a guilty race as
the result of ancestral crime, but still by that very misery
working out the great purpose of Zeus and making men wise
through suffering. This stern, inexorable course of nature’s
laws, which all creeds must recognize, whatever they may
_ choose to call it, seems to be the Fate of Aeschylus, the poipa ©
Teraypéva éx Oedv. The Homeric potpa Oedv or aica Ards
stands in the same general relation to the more primitive
government of the world by special interventions in which an
earlier age believed. The frequent statues of Zeds powparyé-
7s which Pausanias found in different parts of Greece show
an absorption of an ancient idea of independent fate into
the more advanced doctrine of the sovereignty of Zeus. (See
Pausanias i. 40. 4; v. 15. 5; viii. 37. 1; x. 24.4) Now, if

THE AGAMEMNON OF AESCHYLUS. 231

this was the poet’s view of fate, that it was the onward

-march of nature’s laws, the universal laws of the Gods,

how could he have failed to see that the workings of several
such laws, i.e. several lines of fate, may—nay, must often—
interfere with each other, like several mechanical forces, and
produce a result which is different from any of them? In
this view, the chorus simply express a last hope that the line of
fate which seems to them to be leading directly to some new
deed of blood may perchance be met and balanced by some
other line of fate as yet unknown to them, so that the horrors
which they see in prospect may be averted.

5. Vs. 13847: ara xowwodpue? dv tas acdary Bovrv-
pata. This reading of the MSS. was emended by Porson to
Kowwoaiue? av més (interrogative). The emendation now
generally adopted is that of Hermann dy tas (for @ dv mas):
The latter is supported by two passages of Sophocles,—d)v’
dvayndoat Oeovs av pur) OérXwow ovS av eis Svvart’ avynp, Oecd.
Tyr. 281; and dpacov tis éotiv' adv réyns Se wn Hdver péya,

_ Philoct. 574;—in both of which the sense makes &v for &@ dv

(=édv) of the MSS. an almost certain correction. It is, how-
ever, quite as possible that dv wws in the MSS. is a mistake
for nv mas, so that we should read adda Kowwodped’, Hv THs
acpharh Bovrevuarta (sc. 7), but let us take counsel, in case there
shall prove to be any plans for safety, i.e. that we may adopt any
plans for safety which there may be. This is a case of the quite

- common absorption of the apodosis in the protasis, which some-

times gives €ay with the subjunctive the appearance of an indi-
rect question. See Plat. Rep. 1. 358 B: dxovoov Kal éuod,
éav cot TavTa doy, hear me too, in case the same shall please
you, 1.e. that then we may adopt it. Here the construction is
obvious; but in Rep. Iv. 434 A: (dé 67, éav cot dep euol Euvv-
Sox), many think they see an indirect question, though they
cannot tell us what the form of the direct question would be.
The change of 7 to av (= éav) in the MSS. here supposed is
confirmed by three passages of Sophocles,—jv dpacw, Trach.
672; nv...mpocOy, Frag. 323 (Nauck); ovd av tov didacKadov
raBy, Frag. 736,—in all of which the MSS. have dy. The
further question, whether all four passages together do not,

2500 THR JOURNAL | or - PHILOL OG. —
furnish Seon for an exception to the penne de etrin
_ for édv was never used by the tragedians, need not be dis
here. The meaning of the line with the reading 7 WV Tas ag
well with that of the preceding verse, rovpyov eipyao0a
Hot Baciréos oipdypate. ;

sa Vs. 1599: @uwéer, dymlarres & aro oparis ‘pan. :

Weil. An instance of é£epev in this sense is BS in] 1e
erates (Pers. Frag. 2) : ey é is

“

"QO parayas pev eEepav, dvarrvéwv & vaxwwOov,

| W. W. GOODWIN.
July; ITE. 24 sey 3 | S

ON THE FRAGMENTS OF EURIPIDES.

I HAVE taken, as was natural, Nauck’s edition of these
Fragments (Lipsiae 1869) for the basis of the following remarks.
His numbering and arrangement of them have been generally
- followed; by Dindorf for instance in his latest recension of the
Scenic Poets. At the same time I have kept constantly in
view the authors who have preserved these fragments for us, of
whom Stobaeus is by far the most prolific and important.
Editors appear to me to have sometimes unduly neglected the
hints afforded by the theme which Stobaeus is illustrating in this
or that fragment and thus to have missed the probable meaning.

58. Alexander.
= -/ \ AS PS) a 3 /
@ TayKaxicTo Kal TO SodAOY Ov AOYo
eyovTes, GANA TH TUYN KEKTNMEVOL.
‘vs. 2 77 pdoes Iacobsius’ Nauck. In several fragments of this
play slaves in name and by position are contrasted with them
who are nominally free, but are slaves from baseness or self-
indulgence. I would therefore suggest rpudy7 for tuyn. Comp.
fr. 55 xaxov Te maldevp’ Hv ap’ eis evavdpiav “O TrodTos avOpo-
Tow att ayav tpudat.
106. Alope.
Cas \ bd a UY / U
Ope pév avdpav Ttovde yupvdda otodov
atelyovta Gewpov éx Tpdywv Temaupévov.
Ammonius quotes this passage to illustrate tpdyos = dpédpos.
‘vs. 2 oteiyov? é@ov Dindorfius; equidem suspicabar avépav—
atddov Yteiyov? dpapev’ Nauck. I propose
¢ A bS > a 4 / /
op@® pmev avdpov Tovde yupvada oTodov
aTeliyovT aOvpovT, €k TpoXwY TeTALMEVOD.

Journal of Philology. vou. X. 16

234 THE JOURNAL OF PHILOLOGY.

‘I see this troop of gymnasts coming on, disporting themselves, ©

now they have done with their racing’. I would suggest in Ion 52

dudiBoplovs tpdxous Hda7’ dOvpwv (or perhaps tpoyxas, for

Tpoxy occurs, apparently with the same meaning, in a tragic
iambic quoted by Hesychius s. v. wpocavpifovea), in place of
tpodas which scarcely gives any sense,

149. Andromeda.
ovKk eat baTis evTuyns ébu BpoTar,
Ov pn TO Ociov ws TA TOANA ovVOEXeL.

‘vs. 2 Ta TOAN érr@pere? Heimsoethius’ Nauck. Perhaps os

Ta TON avEewv Oédes: this first became moddra EuvvOérer.
Comp. Med. 966 xeivns 6 Saipwv, Kxeiva viv aifer Oeos, Nea
tupavvei. This however can scarcely be what Euripides wrote.
Mr Verrall observes: ‘ xetva for ra éxeivns is a loose expression,
and the whole phrase «xeivns—tupavvet somewhat incoherent...
Nauck would strike out cetva—rupavvet’. I would suggest.
an ig / \ a) wv /

Kalvns 0 Saipwv, Kawa viv av&e. Geos,

véa TUpal'vel.
Comp. xawys viudns earlier. in the play; and 76 wadaa Kat-
vov NelTeTaL KNOEU LATO.

162. Antigone.

avdpos & épévtos eis Kump veaviov

advraktos 7 THpHoLs, Kav yap paddos 7 <

TAAN, eis Epwtas Tas avnp copwTEpos®

Ral > > al / es lal

jv © av wpoohntar Kumpis, nductov raPeiv.
This the Ms. reading is manifestly very corrupt. I will first
write down the verses, corrected in sense and metre:

avdpos © opavtos eis Kurpuw veaviov :

aduNakTos 7 Teipacis, Os Kav haddos 7

wy? > + a oN / 3
TAAN, ELS EPHTAS TAS aYnp TopwTEpOS
jv © av wpoontat Kurpis, ndvorov AaPeiv.

vs. 1. épaévros ‘pereleganter emendavit S. Musgravius’ says Valck-
enaer (Diatr. p. 242); and this has naturally been adopted by the
latest editors of Euripides and Stobaeus. In vs. 2 Nauck’s os
xav for kav yap simply and effectually corrects the metre. But

e

, ‘ »
* ahh i
Crh AES
oe “ es. 6...

Af

despa
’

BY a eS eee ae

Me

ON THE FRAGMENTS OF EURIPIDES. 235

his proposed arpaxros 7 pais for the first part of the verse
is surely rewriting. I conjecture qe/pacis for tnpnows: when
we recollect that in very early times, long before Stobaeus, e
and « were indiscriminately interchanged, III Paovs might easily
pass into THPyois. The word I take in the sense in which
Thucyd. vi 56 uses it, tov & ody ‘Apyddiov arapynbévta thy
meipacw. His attempt at seduction is not guarded against,
because love sharpens his faculties and enables him to carry out
his schemes with greater skill. And now in v. 4 Musgrave’s
nv for nv, before unmeaning, becomes significant, as it refers to
meipacis. All editors have mpoonta:, surely a non-existent
word. Nauck says ‘dv mpoojras Kvzpis nondum emendata’ ;
but I think #v 8 av rpdonras K. gives an excellent sense:
‘But whatever form of seduction Cypris approves of and favours,
it is most sweet to experience’;—and thus it becomes irresis-
tible. Comp. Electr. 622 mpoonkayyy to pnOév, ‘I quite ap-
prove what you say’. Cypris, as she wills, makes love prosperous
or unprosperous.

167. Antigone.

) yap SoKnows TaTpdot Taidas ecixévat

TA TOANG TAVTH yiryveTaL Téxva Trépt.
The conjectures of Nauck, and of Meineke in his Stobaeus,
strike me as violent and improbable. I propose

av yap Soxnow matpaor matdas eixévar’

TA WONG TavTH ylyverar TéEKV ewpepy.

‘Yes, there was an expectation that children are like their

fathers’—and this expectation has produced the effect: ‘it is in
this way that for the most part children become like.’

230. Archelaus.

\ e , / \
Aavacs 6 mevtnkovta Ouyatépwv matyp

NeidAov AuTaY KaAMOTOV ex yalas tOwp
# * * * * *

EXOav és “Apyos Kio’ “Ivayou modw.

‘vs. 2 nondum emendatus’ Nauck. é« ypeléas would suit the
sense and the myth.

16—2

236 “THE JOURNAL OF PUILOLOGY.
250. chsh

ovK tot. Tevias tepoyv aloxiorns Oeod.

picd yap avTws oltives ppovovor pér, = 5

dpovodat & ovdevds ye Xpyuatov Umep.
‘vg, 3 ovdéev ypnudtav v7réptepov Pflugkius’ Nauck: and Meineke
actually adopts this conjecture in his Stobaeus. But the title
of that chapter is Ilevias yoyos, and many passages are- given
to shew that poverty at all hazards is to be avoided; not that
riches can be too highly valued. I propose .

dpovotat 8 ov, Séov ye, Xpnudtav vrep.

‘who are men of thought, but take no thought, tho’ they ought 4
to, for riches’.

264. Archelaus.

Tada oKoTTOUWaL Tas TUYas TOY BpoTaV

ws €0 peTadANaccovcwWw' Os yap av ohary

eis 6pOov éotn Yo Tplv evtvyav mirveL.
‘vs. 1 tas ébnuépwv tvyas O. Hense. vs. 2 ws €d] 6oov Herwer-
denus, és O¢ol O. Hense’ Nauck. The following has occurred
to me |

Tana, oxoToupar Tas BpoTt@v TVYas, bras

det [or, efkn] wetadrAaccovew.

~

286. Autolycus.
ayowivas yap immoiot proivas nvias wréKeL.
‘videtur tetrameter trochaicus restituendus esse’ Nauck. The
words suggest to me fragmentary senarii rather: such as
.. cxowilvas yap [auru«as]
immotot prolvas & nvias wréxer.. *
288. Bellerophon. Se

I cite the three last vss. of this fragment, as the first twelve —
present no difficulties.

95 ] x € A yy >’ \ x Lal)
oimat & av vuds,el Tis apyos @v Oeois

v \ \ \ : , /
evyoto Kal pn yetpl cuAdéyou Bion,

Ta Oela trupyotow al Kaxal Te cuudopai.

Ea.
; )

oe eee

ON THE FRAGMENTS OF EURIPIDES. 237
‘vs. 15 non expedio’ Nauck. I would suggest
Taypel vmoupyety rh KaKN T aovpdopa.

This would give a sense in accordance I think with the rest of
the fragment. Ifc be written for e, the letters in vroupyeuy are
then the same as those in trupyovaw, and if it became vuzroupy-
ov, such an impossible word might readily pass into wupyovow.
Cobet, Collect. critica p. 217, says that there is a ‘magna lacuna’
before this verse. I do not think so; for the indefinite tus can
surely designate any among the vas. |

311. Bellerophon.

enTnoo vTeiKov addov 7) waGdAov GéXou.

‘poetae verba mihi obscura’ Nauck. Plutarch in two different
works quotes this line to illustrate the truth that a man should
not be too yielding and submissive to friends, as Pegasus was to
Bellerophon. The following slight alteration would give I
think a good sense:

érTnoo vUTEiKoV faddoV, 7 WaAAOV OéXoL.

‘Pegasus ever cowered and submitted the more, in whatever
way Bellerophon wished him to do so’. With 4 0édou compare
Soph. Ant. 440 od pwév Kopifors av ceavtov 4 Oéres: Electr.
1429 7 voets érresye vov.

324, Danae.

” \ 3 \ SRE. a > Pe gee
EpwS yap apyov KaTrl Tols apyols edu
prrel KatomtTpa Kal nouns EavOicpara,
gevyer 5é pwoxous. év Oé poe TexunpLov’
ovdels Tpocaitav Biotov npacOn Bpotar,
b] al tee. e A / PRAY A

év Tols & éyovow nBnTis Twépvy’ 66.

‘vs. 5 7AnTHs corruptum’ Nauck. I have thought of
év tots & éyovot 57 Barns wépvy’ Ode.

Comp. Hesychius: pos. tpayos Barns: id. Odpos. Barns
adpodiovactns: id. Batas. 0 katagepyns. Tapavtivor.

238 THE JOURNAL OF PHILOLOGY.
406.. Ino. vss. 2 and 8.

“ xphv yap Tov evtvxobv® brrws TreloTas Exew
yuvaixas, elrep Tpopy Sdmors Traphy. :
‘vs. 3 elzrep év Sopots tpody Pflugkius, an elzrep Sdéuacw Tpody

leg. ?” Nauck. I would simply read elrep kat tpopy 6. a. for
kat is surely wanted.

‘414, Ino.

U A \ / bal
ToLavoe xen yuvaset ™poomonoy édv,
ATI TO Bev Stxatov ov ouynoeTat,
ta 8 aicypa picet kal Kat dpOarpovds é exel.

Nauck registers several quite useless conjectures; but does not
mention in v. 1 Musgrave’s almost certain correction, zpoo7o-
rev: comp. Troad. 264 tiuR@ tétaxtar tpootonety "AyiARéas:
nor in v. 3 the quite certain correction of Dobree, épe? for éyeu:
‘but hates what is base, and will tell you so to your face’.
Comp. Aristoph. Ran. 625 ta cot-Kat’ dpOarpods réyo,
referred to by Dobree: also Rhes. 421 xal péudopai coe nat
Aéyo Kat Oupua cov: Electr. 910 & ¥ elwety 7OcAov Kat dupa
cov. For épet comp. fr. 416 Gotis...rard Kapdias épet. In
frag. 613 we should perhaps read Adyous épels for &yeus.

457. Cresphontes.

ovntépayv S€ tTHVd eyo Sid@pml cor

TANYNV.
For ovntépayr is the reading of all known Mss., not écvwrépav
which all editors follow Valckenaer in adopting: he says indeed |

vaguely ‘e Mss.’, but nothing is known of them. It appears

- from both Plutarch and Aristotle, that this verse, when spoken
by Merope as she was on the point of stabbing her son, sup-
posing him to be the murderer of that son, but discovering her
error in time to save him, always caused a prodigious sensation
among the spectators. I would suggest ; ;

dvatT ap tv &) THVd éyod Sidopmi cor
TAY. |
‘Good speed to this stroke which now I strike’,

ON THE FRAGMENTS OF EURIPIDES. 239
514, Melanippe.

a \
éyo pev ovv ovK old bTwsS oKOTEIY YpEecV
\ a j \ > ,
THVv evyéverav’ TOs yap dvdpelous vow
n /
Kab tovs Sixalovs Tav Kevav Sokacpatan,
x > , , / /
Kav wo. SovrAwY, EVyeveTTEpOUS EYyO.

_ ‘vs, 28q. tas ydp avSpelas dices Kal tas Sixalas Kal Kevds
Heimsoethius. vs. 4 evyeveotépas Heimsoethius’ Nauck. Eight
words are thus altered, with no satisfactory result. I would
simply read tév kev So€acpary: the cause of the corruption
is manifest. ‘Them who are by nature brave and by nature
just, even if the sons of slaves, I term more noble than they
who are only so in empty opinion’, I have sometimes thought
that to read rods dicaiws tav Kev@ SoEacpatt would give a
more pointed antithesis, But the two great civic virtues are
courage which repels external, and justice which prevents inter-
nal dangers. The title of the chapter in Stobaeus which con-
tains our fragment is qrepl evyevelas’ OTL evryevels of KaT apeTHY
Covres K.7.X., and several of the fragments dwell on the virtue
of justice,

530. Meleager.

TO T.. KpaTloToV, Kav yuVi) KpaTL.. 7,

tovT éar apetn’ TO é dvow ov Svadépet.
The passage appears in this mutilated shape in the sole Ms. of
Orion. In vs. 1 70s, and in 2 éstw, are doubtless to be read
with all the editors; and it is highly probable that the Ms.
originally had xpatiorov, as Dindorf reads with Schneidewin.
But this, as Meineke says, gives no sufficient sense. Nauck has
yovn KaKos Tis 4, yovn from Conington, the rest from Gomperz.
I propose

TO TOL KpaTLOTOV, KAV YUVALKOKTLOTOD f,

ToUT éaTw apeTn* TOS dvopw ov TL Siadépe.

On the one hand Aristotle in his Poetics quotes from a Tragic
poet omeipwv Geoxtictav droya: on the other Aeschylus has
yuvatkoynpuTos: yuvatkodiéaxtos also occurs. Conington and
Meineke conjecture ov dvafOepe?. Our fragments shew that

240 THE JOURNAL OF PHILOLOGY. ar

the conduct of women, as might be expected, was el cans 2 252
vassed in this play. | ~ | ;

538. Meleager.
TO pev yap év $@, TO SE KATA GKOTOS KaKoD.

‘verba ro Se—xaxév nondum sunt emendata’ Nauck. ‘Com-. os
paring Here, Fur. 563 tod xdtw oxdrovs, I should say that =
Valckenaer’s 7d 8& xdt@ oxdTos Kaxév was probably right.
Dindorf asserts that the ‘veteres’ did not use To cxétos. But —
other editors and the Mss. of Euripides have the neuter more ©
than once. It is found in Thucydides also more than once,
and is common in Plato. Its occurring in Pindar proves it not
to be a late Attic invention. With reference to this neuter I
would now discuss fragment :

537. Meleager.
adteprvov To das por TOO bro ynv & ddov cKdTOS
ovd eis dvetpov ovd eis avOpdrrovs podciv
éy@® pmev ovv yeyooa THALKHS Gywws
anémtva avto KovToT evyouat Oaveir.

‘I have written down this fragment just as the Ms. gives it, ~
with the exception of Valckenaer’s ryAux)3 Sues for tydlen
Souots. Meineke calls it ‘corruptissimum necdum probabiliter ;
sanatum fragmentum’. But the avvo of v. 4 proves almost to
demonstration that «xdros here is neuter and that Nauck and
Dindorf are wrong in reading 6 6\—oxoros. Of course in vy. 1
Gesner’s rep7rvov is rightly adopted by all editors. But I think
it not improbable that it was preceded by the exclamation 4,
which would explain the dreprvov. This is how I would amend —
the first two vss.

a. TEPT VOD TO pas. sare TO 8 Ure ynv Gdov TKOTOS |
ovd eis Gvetpov Uytés avOpadmrois poneiv.

‘We find @ similarly ‘extra metrum’ in Here. Fur. 629 and
Bacchae 810. oa

t i ng a. 7 f
. 4 - as 4 vi) ne rj d -
ae Se eg ee) tn a ene

oie
oss Sw

554, Ocedipus.

> lal +7 ¢ / / ay
éx TOV aéATToOV » xapis pelSov Bporois,
pavetoa madrov TO TmpoTdoK@pevor.

ON THE FRAGMENTS OF EURIPIDES. 241

‘vs. 2 aut spurius aut corruptus” Nauck. Corrupt, surely not
spurious. I suggest

/ a x \ t
calvovca maddov ) TO TpocdoKapevor.

‘

576. Oenomaus.
I quote the two first lines of this fragment.

&y éott TavtTwv mpe@Tov eidévat TouTt
hépewy TA ovptintovtTa py TadIyKOTOS.

‘vs. 1 tovti] rode B, suspicabar Bpot®’ Nauck. Of course Tdd¢e
of B is a mere attempt to correct the metre. - I would read

n Ud /

év éoTs TaVT@Y TpeToV Eldévat, TO TOL
\ /

dépew Ta cuptintovta pn TadiyKoTos.

is . . . . /
TO Tou is in place in a general maxim: comp. fr. 530 To Tou
KPaTLOTOV K.T.X.

582. Palamedes.

\ A t t ee: ’ ay
Ta THS ye ANOns happak opPwoas povos,
addova kal dwvodvta cvdAdaBds Te Oels
> a > A / bd] 3Q7
éEedpov avOpwrroict ypampar eidévat,

4 ? b]
@OT OU TapovTa TovTlias vTép TAAKOS

’ lal al
TaKel KaT olKoUS TavT érrictacGaL KaXOs,
Tac T aTroOvnoKovTa ypnuaTtwv péTpov
ypavaytas eiteiv, Tov NaBovra SO eidévar.
“A eT a
a8 eis Epw wintrovaw avOpdéras Kaka

f- a > an A /
SérTos Saupe, Kove €& rvpevdH réyeuv.

‘v. 6 admoOvnoxovta suspectum. v. 7 ypawavtTa eltrew Sca-
liger. vs. 8 xaxnv Heathius’ Nauck. No word in the whole
passage is more genuine than dmovycKovta. I propose with
some confidence |

Taiciv T awoOvicKovta YpnuaTov wéTpor,
ypawave ba’, eimeiv.
‘And at his death to tell his children the amount of his riches,

having left in writing how much they are’. With ooa compare
Soph. Aj. 118 dpas, "Odveced, tiv Oedv iayvy, bon.

242 THE JOURNAL OF PHILOLOGY.

_ In v. 8 xaxjv may be the right correction of the corrupt »
xkaxa: I have thought of «cada, for which the context might
readily suggest «axa to a careless scribe.

608. Peliades.

70 8 ésxatov 6) totTo Gavpacrov Sporois

tupavvis, ovx evpols av aOwTEpOD.

dirous te “ropbeiv Kal Kataxtaveiv peor,

mreiotos PdoBos mpocerts pn Spdcwot TA. a
‘ys, 1 écyaras scripserim. vs. 3 fort. woXeus Te 7. Kal pirous
KTaveiv. vs. 4 wAéloTos| ével Pflugkius’ Nauck. These changes
are violent and yet insufficient; nor do other editors help us.
I offer the following reconstruction of the passage :

ti 8, €xyatov 6) TOO 0 Oavpactov Bpotots,
Tupavuls; ovx eUpots av aOdu@TeEpor. | ;
dirous. 7 atwdetv, Kal Kataxtaveiy ypewr,
Tr«ic 0 ols bdBos mpdcecte pr) Spdowal Tt.
With the beginning compare fr. 900, 7 ri todTo 5x Td xpyoTor ;
ovx apxet x7... Euripides is very fond of the word amwOeiv.
The tyrant must drive away his friends, and put to death such
of them as are most dangerous.

anit

Pe

penis 5

620. Peleus.

E,] BA > / A ,
ovk éoTlv avOpdtrotoct TOLOVTO OKOTOS,
5) A / t ” \ r ; a
ov dopa yalas KrELoTOV, évOa Thy dvow '
6 dvaoyerns Kpuyas dv eln aodds.

3 sd ]

Thus the Mss. We have then another instance of the neut. —
oxotos. Nauck, Meineke, Dindorf and others read TOLOUTOS :
but the euphony is thereby somewhat impaired, fond as Euri-

pides is said to be of c. Elmsley indeed denies that tosodro is -
found in the tragic poets. But Hermann and other scholars do - a
not follow him in this; and Aesch. Prom. 801 tovodto pév coe s

ToUTo seems indisputable. If the tragedians use, as they some-
times do use, tavré, why proscribe the good Attic tovodro ? fe =

In vs. 2 Meineke, followed by Nauck and Dindorf, conjec- -”
tures yaya. This suits yalas: but then xrAyorov? that surely

ieee 2 io
‘7

ON THE FRAGMENTS OF EURIPIDES. 243

fits 6dua, not ydua: Thue. 1 17, speaking of locked buildings,
el TU GAXO BeBaiws KXnoTov Hv: Eurip. himself, Andr. 593, has
adxdynota Seépata. The corruption therefore is in yaias. I
propose ov daua tolws KAnordov. Comp. Hesychius toiws.
ioyupas. Karas. axpiBds: he means evidently ovtws icyupds,
which is illustrated by Thucydides’ BeBalws: and such an adverb,
if not necessary to the sentence, at all events much improves it.
It is evident that Hesychius, or his authorities, often have in
view the Attic poets of the best time, and it is not improbable
that the writer was thinking of our passage, if indeed to/ws was
in it.

Hesychius again tes. ovrws. This Homeric word occurs
four times in Aeschylus, twice in the dialogue; once in Sopho-

cles in the dialogue. I would therefore suggest in the Medea,

909,
See \ > \ lal a ,
eiKos yap opyas OfAv ToetcPar yévos
yapous TapewToAwyTL TWS AAAOUS TOCEL:

Tws, Secxtixas, ‘furtively dealing in other marriages as I am
now doing’. Its use is very similar in Aesch. Sept. 618 7) fdv7’
atiywactipa taés o avdpnrarnv puyf tov adrov tévde ticacbat
tporov. I should not think of rejecting dAdolovs merely
because it is not found elsewhere in the Tragedians: Homer
and Pindar prove it to be suited for poetry; Thucydides and
Plato shew it to be good Attic. But I agree with Dindorf and
Verrall that it has here no senses When -7s tws had passed
into -Tos, aAXolovs might well have been written to make good
the metre. The scholia are now a confused jumble; but they
seem to shew traces of both a@Aovs and mapeurrorA@rTe.

To return to our fragment: in v. 3 Sir G. C. Lewis may be
right in proposing xpUypesev av kav 4 codds. But in another
fragment of this chapter of Stobaeus folly is connected with
dvoyévera. See too what is said below on frag. 739. ‘xpuas
av op0ein Enger, fortasse éxBain praeferendum’ Nauck. Perhaps
Kpvrwas av eit’ ein codes.

652. Protesilaus.

TOA ErTrides Yevdovat Kal droyor BpoTous.

a AAS 2 a el oS ee STE MED ars eg td See eae Pn am = .
yee are We eae en i
. . iene BP ae 5, el: a as oe
- Wiles * = Nel Le idee Nan cae ae
\ = mn Se
a ee eS
244 THE JO URNAL OF PHILOLOGY. -

Dindorf, followed by Nauck, «al oyou. Gaisford and Meineke: |
evdovow &royor. But «al ddroyou is_certainly genuine, and
the corruption lies in wedSoucr. I would read

MOAN édrrides YruyoveLr Kadroyor BpoTos. i irae

iyo in a fragment of Sophocles has the sense of ‘cheer’ — - ee
‘refresh’, like the compounds avayiyo, cataiyo, tapayvyo. —
Comp. too Athen. p. 503 c for the use in mramas i and Euri--
pues of wuetnpia for Tos addowders Kal oveKious TOTOUS...€D
ois oti avawpdEat. ?

sr

7

664. Stheneboea. ay oa

dvev TUXNS Yap, domep n Tapowmla, :
moves povobels ovKéeT adyiver Bporous.
vs. 2 ovxér’ addaives Musgrave. Before I knew of this reading,
I conjectured dAdave?: ‘Without luck, as the proverb goes,
labour left to itself will no more make men plump’. This
Homeric and Aeschylean word is in place in a proverb, |

698. Telephus.
WTO, ap piPrara codpatos AaBov pare,
apKTnpia TUYNS.
‘vs, 2 dkxtypia Bernhardy. tvyns] wvyouvs Dobraeus’ Nauck.
But how with such readings would the verse proceed? I pro-
“pose avymnpaT atuyns.
703. Telephus.
This fragment, which has been recovered from Olympiodorus’
comment on Plato’s Gorgias p. 521, is given as follows by
_Nauck, whose arrangement and explanation of the words are
simply ps ces by Dindorf:
* avdpa Muadv Tyredov...
elt €ott Mucos eite xadroOév trofer,
[éx tov] wpocwmov Tyredhos yrwpiferat.
Nauck has a right I think to say ‘certum est quod v. 1 addidi
of0 dvépa ex Ar. Ach. 430’. But he gets his last verse out of
these words of Olympiodorus, wads 67¢ 6 Tyreghos yvwpiferat,
which are surely the writer's prose paraphrase of Euripides’

>

aa
*
> wh
My
_- =
*
Ps.
> 4
. 7 ps
va
aoe
*
“a
‘
=
x
a) 4
ie
:
Pe
- ee
’ i
3
*
+
%
2
|
ae
“sv
m

ON THE FRAGMENTS OF EURIPIDES. 245

words. This reading of Nauck, who adds that £évos & 68¢ at
the end of v. 1 would complete the metre and the sense, seems |
to me to destroy all connexion between the beginning and end
of the fragment, and to make Olympiodorus’ comment unmean-
ing. To me it is clear that Telephus was in rags-.on the stage
and for some time was not recognised. Some one says: ‘I
know (by fame) of Telephus a Mysian’: [this man may be a
Mysian or not]: and then

GN elite Muads elite kddrobev Tober,

mos ovTOS ws @V TrHrEehos yropifeTat ;

or something to the same purport.

739. Temenidae.

hed hed, TO hovat Tatpos evyevods arro

dony ever ppovnow akiopa Te.

Kav yap wévns av TUYYXaVN, KPNOTOS Yyeyos

TLV EXEL TL, avapeTpovpevos Sé THs

TO TOU TaTpos yevvaiov Oder? TpOTY.
‘vs. 2 ever Soxnow afiduatos Meinekius. éyes 7 dvnow
Schmidtius. vs. 5 @dede? tpor@ verba corrupta’ Nauck.
‘ @heret] ov POepet Engerus’ Dindorf. If dpdvnow could bear
_ the meaning of dpovnpa, ‘high spirit’, it might be in place; but
no instance of this seems to be known. For Suppl. 216 is
hardly a case in point. It was said above on fr. 620 that ro
f@pov was sometimes connected with dvoyévera: see fr. 166
from the Antigone; and, what is more to the present point,
comp. fr. 138 of the Adespota 77v evyévecav, Hv Oédns avacKo-
mei, Ev tols Kadkos dpovodow evpnoeas Bpoteév. But would
the word at all harmonise with a&/wua, and the rest of the
fragment? If not, I would propose dOdvncwv daélwpa Te,
‘envy of the bad and esteem of the good’. $@ovyous occurs in
Soph. Trach. 1212.

For the manifestly corrupt ddere? tpo7@ I have thought of
ov iret purov: ‘Ever thinking of his father’s nobility he
loves not sordid ways’. fvzos is found in Homer, in Aristo-
phanes and the Comic Fragments, in Plato, and once in the
-fragments of Aeschylus. Or ov @éXeu purray might be suggested :

246 THE JOURNAL OF PHILOLOGY.

Ton 1118 6 Oeds ov puavOjvar Oérov. In Latin ‘sordes’ is very

common in this sense: Horace ‘O nec paternis obsoleta sordi- —

bus’: Cicero ‘splendetque per sese semper neque alienis um-

quam sordibus obsolescit [virtus]’: ‘non amat profusas epulas, —

sordes et inhumanitatem multo minus’.

773. Phaethon.

, al lal an. > »

Sewov ye, TOls WAovTODGL ToUTO O EuduTor,

a aAQ/ >
oKaloiolW elval’ TL TOTE TOVOE Y alTLOV;
5 a 7 \ “
dp orBos avtots ott TupAas ouvnpeTel,

\ 4 \ J \ A /
TupAras €yovat Tas Ppevas Kal THS TVS.

‘vs. 4 tentabam cal THs Téxyvys sc. pavTixns. Maehly kab

dvotuyeis’ Meineke. I offer

tudras éxovot Tas dpévas KovK EeVoTOXOUS.
But in v. 2 toddé yx’ aitiov is Nauck’s conjecture for todro
taitiov of the best Mss.
| 781. Phaethon.
v. 50 of this long fragment is said to be thus represented in the
Ms.

dravra tadt nOpnoexavTrwrtovaeyel.
Dindorf adopts Hermann’s ‘correction’: tad7’ alOpn|r axarve-
roi @ 6d0i. It would certainly be nearer the Ms. and would I
think agree with the context, if we read

e/ nm 3 40 > An »”
aATTAVTaA TAVUT 7) pnoe E€VAVTA TWS EXEL.

Comp. Soph. Ant. 1284 évavta mpocBdérw vexpov: Euripides
also uses €vayta in the Orestes; as well as the Homeric dpra in

Alc. 898: elotdefv dvta. This word I would introduce into —

Suppl. 322, where dva8néres can scarcely be genuine: dpés,
adBovros os KexepTounuévyn, Tots Keptopwodat yopyov ws avTa
Bréret & watpis: ‘Seest thou how thy country, when flouted
for its reckless policy, grimly looks the flouters in the face ?’
This was written about the time of the peace of Nicias, when
Athens was at the height of its power.

793 (vss. 4 and 5). Philoctetes.
doTis yap avyet Gedy érictacbar Trépi,
ovdév TL WaAXov older, 7) TEiOer réyor.

:

iy

eiitisinhit Dhaba aeemlinn tice een Cs

ON THE FRAGMENTS OF EURIPIDES. 247

Nauck, followed by Dindorf, reads 7 qei@ew X.: ‘quo locus vix
persanatur’ says Meineke. I propose ¢¢ meiOev Xéyov. ‘He
may boast he knows all about the gods; but he knows not a
whit the more for all that, if he persuades men he does by his
words’, Perhaps we should read oiéde, Kev.

794. Philoctetes.

néEw 8 eyo, dv pov SiadpGeipas Sox7

Aoyous vroaras avTos HouKnKévar.

GN é& euod yap Taya palnon Krvor,

6 8 avtos avtov éudaviel cot réyouv.
This fragment ought to be compared with its context in the
Rhetorica ad Alexandrum, ch. 19. There is I believe a corrup-

. tion in every one of the 4 vss. This is proved by the metre of

the last two, and by the sense in the first two. I will write
down the passage as I propose to correct it:

rAéEw 8 éyd, Kav pov SiapOeipar Soxh

Adyous, Udsatas avTos HoiKnKévat.

GAN €& éuod yap Tay’ dvapabynon KrVor,

6 8 avtos avtTov Hudavilé cot Néywv.
The figure mpoxatadynYis has place, when the first of the two
speakers, conjecturing the charges which his adversary is likely
to bring against him, and for himself, anticipates the other by
giving his own version of these charges and so gaining the ear
of his auditors. The avtitpoxaradn is has place, when the
second speaker retorts on the first by exposing the insufficiency
of what he has said against him and in favour of himself:
cadas eidads ote éEehéyEw avtov mpoKxatéraBé prov Tiv Adyov
Kal mpoduéBarev, tv vets «.7.r. Anaximenes quotes these
verses as an example of skilful dytimpoxaradAn ws. ‘1 will
speak, even tho’ he seem to have spoilt my speech, by setting
forth in his own way his own wrong-doing. But that won’t do,
for you shall hear from me over again my pleas; but he in his
speech made it plain enough to you what a man he is’ ;—so I
need say no more of him.

The infin, dsapGeipas seems clearly called for. With tdic-

ras comp. Soph. Aj. 1091 yvopas vrroatncas cogovs. For the

248 . THE JOURNAL OF PHILOLOGY.

neut, vroards could only mean ‘having agreed’ ‘promised 10.
do something’; and not ‘admitting doing it’, which the context —

calls for. And that these two words hs be easily inter- A
changed, would appear from this: Hesychius says rightly:
Udiotas. UToTels. Both Photius and Suidas, following him

or his sources, have: uduotas. vrrooras. Urotibeis. These late

writers seem to have lost the feeling for the difference between - 3 7

the active and neuter parts of the verb. For dvaya0jon comp.
Hesych. dvaydaba. €& apyfs palo. Inv. 4 I write with some

confidence jpupdvefe, because I believe the fut. éudave? to be a

plain interpolation of a copyist who did not understand the past
tense. Otherwise the old correction of Heath, éudavifer cou,
strikes me as better than various later conjectures. The con- |
jectures recorded by Nauck and Dindorf seem to me very —
insufficient. The resolved foot which I have introduced into
the third verse may be objected to in a play which was in the

same trilogy as the Medea. But the word exactly suits the © 3

context; and even in the Medea we find verses like 375 Ooo
Tarépa TE Kab KOpnv Toow T Eon.

801. Phoenix.
/ b] >’ \ / ,
poxOnpov eat avdpt mpecBitTyn Téxva
didaaw baTls oVKEP patos yapmety’ | Fs
déorrowva yap yépovTs vupdi@ yun.
Yap “<P PP? Tey

Valckenaer’s correction of this fragment, even as corrected by

Porson, is very violent; so I think are Nauck’s and Madvig’s

eo

conjectures, as well as insufficient: the latter for instance makes _

réxva the vocative, and for didwow reads fiywous. But even if

Euripides could have used the latter word, it could scarcely 2

have the sense of marriage absolutely; so that you would need

poxOnpa, the neut. predicate being here not in place. All three :

scholars too adopt yauet for the yauety of Mss. It correct the
passage by altering one termination thus: fea

/ > 3 \ / rs
poxOnpov éotiv, avdpt mpec Buty Téxva
66 ¢ bd LQ” e / n

iowa oats ovKed’ palm yapelv.

The corruption is surely natural for a copyist not understanding
the construction. This use of ders is very idiomatic: fr. 362,

‘*

ON THE FRAGMENTS OF EURIPIDES. 249

v. 1 tas yapitas boris edyevas yapiferat,”Hdi0ov év Bpotoicw*
of 5é Spada pév, Xpdv Sé Spdcr, Spdct Svoryevéctepov: (thus
‘Heinrichius’ completes v. 3; if however we read, as I should
prefer to do, dvcyevéorepov Aé€yw with Meineke, or rode, or
Opdciv, Eats Sucyevéotepov, or something of the sort, we
then have a second example of the same idiom): Electr.
815 ék Tév Karov KouTrodat Toiat Deocaros Eivas Tod’, daTts
Tavpov aptapyet xada@s: Fr. Trag. ap. Plut. Mor. p. 33 768° éort
TO (nrwrov dvOparrois, oT~ Tofov pepiuvns eis 0 BovdeTas tréon:
- Thuc. m1 45 dads te advvarov Kal Todds evnOeias, SoTIS
olerat K.T.r.: VI 14 7d Kareds dpEar tobr elvas, ds dv thy
matpioa wabedAjnoyn. Nor is there any tautology: ‘It is a vexa-
tious thing, for any one to offer his children to an old man for
him to marry, when he is no longer of an age to marry’. He
may be wpaios for dying: comp. Alc. 516; Phoen. 968. The
plur. réxva is I think idiomatic: Phoen. 966 ovS av tov avtod
maida tis Soin Ktaveiv. My pw’ evroyelrw Tapa Tis KTElvov
Téxva, where one only is in question. There is a very similar
example in Heracl, 410—414 of réxva thus used.

830. Phrixus.

tis 8 odev et Shv Tov 0 KéxrnTat Oaveiv,

To Chv 88 OvncKkew éotl; mrj\v bus Bpotav

vooovaww of Brérrovtes, of & OAwWAOTES

ovdeyv vorovaw ovdsé KéKTHVTAL KaKd.
‘vs. 2 adv Opws verba corrupta’ Nauck. The sentence calls,
not for a connecting particle, but for an adverb to sustain
the parallelism. I would therefore suggest with some confidence
TAnMOvas Bporay x.t.r. Hesych. TAnpovas. éAcewes: Troad.,
40 réOvnke TANmOvas TloAvEEvn.

839. Chrysippus.
yvdun codds pou Kal xép avdpelav éyew (or, éxor)
Svcpophos einv waddov 7 Kaos Kakos.
Frag. 895 Choeroboscus quotes from Euripides d¢pov dv etnr,
et Tpéhow Ta Tay Tédas, to illustrate the form rpépovv. If this
form then be legitimate, |
Journal of Philology. vou. x. 17

250 THE JOURNAL OF PHILOLOGY.

yvoun coos Tor Kal xép avdpelav exouv"
Suopopdos einv madXov 7) kanos KAKOS
would give a satisfactory meaning and would explain the éyew
and éyou of our authorities. .

853. Incertarum.

SelEas yap aotpev THv évaytiay ddov

Sipous T écwoa Kal Tipavvos tlounp.
‘vs. 2 Sypous] Sowous Bergkius. Opovovs Heimsoethius’. Atreus
is speaking of his astronomical discovery, recorded by Strabo
and others, of the heavens moving in a contrary direction to the
sun and stars; which discovery gained him popularity and
made him king. I would therefore propose dypous tT éonva.

892. Incertarum.
here SAOev, elrep Eat ev ovpav@e
Zevs, p1) Tov avrov dvctvyn Kabeotavat,

‘vs, 2 1) tov écOAdv Heimsoethius’, Perhaps rather pu) Tov
ayvov, OF TOLOUTOV.

986. Incertarum.
Polyb. v 106 ov yap of8 Srrws det wore IeXotrovynoiot...cata
Tov Euvpuriény joav ‘aicl mpacinoxOol (or, wAnoiwoyOol) Twes
Kai ovtrote Havyxot Sopi’, Perhaps
aiel TLWeES
amrAnaTd6poxGo Kovitro? Havyot Sopi.
Athenaeus twice quotes from Timon amAgotolvous tr aputaivas.

1028. Incertarum.
Kpwet tis avtov momoT avOperwv péyar,
Ov éEadeiher mpddacis 1 TUXovG’ SXrov;
Thus without remark all the editors of Euripides and Stobaeus,
among them Valckenaer, Gaisford, Meineke, Nauck and Din-

dorf. Yet surely the future with weéore is strange and unpre-
cedented. A simple correction would be |

A e 4
Kpwet TLS avTov Tas ToT avOpanwv péyar ;

4
q
&
a

ON THE FRAGMENTS OF EURIPIDES. 251

1030 (v. 4). Incertarum.
) tatolvy av0évratot Kowwvh Somov.

For the unmeaning 7raiolv I have thought of 7 amratow avbév-
tavot Kk. 6.: ‘He shares his house with childless murderers’; i.e.
shelters them not from charity, but greed, to inherit their
wealth.

1039. Incertarum.

/ ad a v7 a
veavias yap boTis Ov “Apn otvyy,
Koun movoy Kal odpKes, Epya 8 ovdapod.
€¢ aA b) , ¢ ¢€ /
épas tov evtparefov ws dds Bios,
67 OrABos &wlév tis éott Tpaypator,
GX ovK everts otépavos ovd evavédpia,
el poy TL Kal TOAM@OL KVdUVOU péTA.

‘vs, 4 graviter corruptus’ Nauck; and I have seen no specious

correction of it. The add’ ove éveote x.7.r. of v. 5, as well as
v. 3, seems to shew that v. 4 pomted to the power of wealth
over external things, tho’ it has no inward virtue. I have
thought of

67 ddABos Ew odnp tis eat TpaypaTor,

‘Wealth is a wedge, or most effectual instrument, of things’: it
can remove external obstacles and procure external blessings,
but can do nothing more. o¢yy occurs in Aeschylus and Aris-
tophanes: comp. too Aristot. Mech. 17 8a ti 7d odyvi dvti
pixp@ peydra Bapy Stictatas Kai peyéOn copator, Kal Oris
ioxupa yiryverat; x.7.r.. For the metaphor comp. Tertull. adv.
Marc. I 21 hoc enim cuneo veritatis omnis extruditur haeresis,

1044, Incertarum.

eyo yap &w rEKTpA avTois Kadas exeu

Sixavoy éotiv olow ovyynpacomas.
‘locus nondum emendatus’ Nauck, I suggest

” ep Nees a , a le + mer 3 A ”

éywy aveEw NEKTP, Ay ws Kaas EXEL

Sixavdv éeotiw alow cuyynpacopat.
olot = érret avtois. With v. 1 comp. Hecub. 121 ris pavto-
modouv Baxyns avéyov Aéxtp’ ’Ayapéuvor.

17—2

a

252 THE JOURNAL OF PHILOLOGY.

1046. Incertarum.
poxOotpev Grrws Ojdrv hpovpovvtes yévos”
Aris yap avr px wépuev évdov (or, €vdos)
ti Set puddooey (or, 5) gvdacce) xafapaptdvew mréov;
The end of v. 2 was evidently mutilated in the archetype: for
the right meaning of v. 3 comp. fr. 112, from the Alope, § 17 in
the same chapter of Stobaeus in which our fragment is preserved.
I would suggest
HTS yap avtTn pr TwéepuKe vodv exer, 3
ti Set huraccev; é£apaptdver Tréov.

1052. Incertarum.

tov ody 6 maida cwdpovodvt émrictapat

xpnotois & dutroovt evoeBeiv 7 HoKnKoTa.

Tos ovVv av ék ToLodde GMpmaTos KaKOS

/ 3 v b) \ lel / > x / ,

yévoir dv; ovdels TodTS pw’ av miOot troré.
‘vs. 3 cwpatos suspectum’ Nauck. I read é« rowds adowros
}) KaKos: GowTos is the opposite of v. 1, xaxds of v. 2.

1065. Incertarum.
© yhpas, olay édkrriS Hdovis eyeis, -

\ n 4 \ t Sept le? , A
Kal Tas Tis eis oe BovreT avOpodtrev ponreiv'
AaBov Sé meipay petapércray AapPaver,
ws ovdev éott yetpov ev OvnT@ yéver.

‘vs. 2 &roiyuos avOpdrev Elmsleius. Bovretas Bpordv Meine-
kius. vs. 3 werauérXera Meinekius’ Nauck. I propose

kal mas tis els a EXOLT Gv avOpeTav ponreiv'
AaBayv Sé meipay peTapéreiav av NaBor,

H. A. J. MUNRO.

PLATO’S LATER THEORY OF IDEAS.

I. THE PHIJLEBUS AND ARISTOTLE’'S METAPHYSICS I 6,

§ 1 Does the theory of ideas appear in the republic am
ats final form ?

In a former paper’ (Jowrnal of Philology x 132) I proposed
what I believe to be a new interpretation of the concluding
pages of the sixth book of the republic; I compared this
notable passage with another, not less difficult, in the Phaedo,
which seems to me to represent the same phase of doctrinal
development; and I tried to determine the dogmatic content
of the two passages. In both places Plato, as I read him,
contrasts the ‘general notion, i.e. the connotation of the
name as we imperfectly understand it, not hypostasized, with
the ‘idea,’ i.e. the whole completed connotation of the name,
hypostasized: in both places he marks the insufficiency of any
method which, like that of Socrates, whether in its original
shape or as improved and supplemented by Plato himself, has
nothing better than imperfect uncertified general notions for its
apyai; in both places, but in the Phaedo with less confidence
than in the republic, he aspires to a more perfect method,
which should attain scientific truth by converting imperfect
uncertified Adyos into Adyos proved to be the exact represen-

tations of ideas; finally, in both places, but in the Phaedo with
" especial emphasis, he declares his scheme of a higher logic to
contain a fatal flaw.

1 To my friend Mr R. D. Archer paper and of the paper here referred
Hind I am indebted for much helpful to.
‘ and suggestive criticism both of this

254 THE JOURNAL OF PHILOLOGY.

Although however, not only in the Phaedo, but also in
the republic,—the dialogue which is generally accounted the
most perfect representation of the most characteristic’ phase
of his doctrinal development,—Plato frankly acknowledges his
failure to construct on the basis of the theory of ideas a
logic of scientific discovery, in these same dialogues the theory
itself is confidently maintained, and carefully formulated. In
particular there are two passages, dogmatic in spirit and pre-
cise in expression, to which I would here invite the reader’s
attention. (1) “ Wherever we find a plurality of particulars
called by the same name,” says the Socrates, of the republic,
“we assume a corresponding idea,”: eidos yap mov Tt év Exaatov
ei@Sapev TiDecOat Tept. Exacta Ta Toda ols TavTOV bvopa
érupépomev. X 596A. Thus, when the republic was written,
Plato, building on Socratic foundations, assumed for every
general name a corresponding idea, and consequently recog-
nized, with others, ideas of evil e.g. xaxdv adixov, of manu-
factured articles e.g. xAivn tpamefa, and of relations e.g. d:-
Tracwov Hurov’. (2) “If any one alleges as the reason why
anything is beautiful,” says the Socrates of the Phaedo, “that
it has a fine colour or a fine form or the like, regardless of
such explanations, which only perplex me, with artless and I
dare say foolish simplicity I hold fast the principle, that the

1 In the republic ideas are explicit-
ly recognized, not only of dyadév v
4764, vt 493c¢ 5054 5073, vir 5170
518p 519c 526 53810 534¢ 5385
5404, of xadévy or xddAdos v 476 BC
479 an 4804, v1 493 8 5018 5078, vir
538lc 5388p, of Sixaov or dixaocd-
un V 4764 4794 501 3B, vu 5175
538 E, x 6123, of cwdpocivn dvipela
éXevOepiorys meyadorpérera 1 402 o,
of cd¢ppov vi 501 8, but also of xaxov
and ddixcov v 476 A, of aisxpov v 475 B,
of the évayria of cwdpootvn avdpela
éNevdepidrns peyarompérera 111 402 c, of
kh\lyn and rpamega x 5968; while if
we take note of implications, ideas of
dirddovov and you v 479 B, Of uéya.or
béyeOos omikpov OX opikpdTys Kovgov

Bapé v 479 B, vir 523 & 5244, of raxos
Aerrorns parakdrTys cutKporys VII 523 E,
and of wadyuarixd vi 5118, must be
added to the list,

In the Phaedo 65 pv, 74 Aa—78 E,
100 n—106 p there are ideas of icoy
péya or petgovy or péyePos €Xatrov or
outKporns ™AHGos OAov wepirrov OY meEpiT-

TOTns GpTiov movas Suds Tpids OY Tpla -
meumTas ayabov Kandy Sixaoy dovoy pov-

otxov Oepuoy or Oepyorns Puxpov or Wu-
xporns vylea ioxvs vooos mupetés fwh
Oavaros wux7n mip xiv. See especially
65 D Aéyw 5é wepl rdvrwr, ofov meyéOous
mépt, vytelas, loxvos, Kal Tov dAdwy évl
Noyw amrdvrwy THs ovclas, 6 TuUyxXavEL
Exagrov dy.

r

y

» Aes

PLATO'S LATER THEORY OF IDEAS. 255

thing in question is made beautiful only by the presence,
or the communion, or the intervention however styled and
entitled, of the self-beautiful: mind, I don’t insist upon the
name, but I do insist upon the principle that it is the self-
beautiful by which all beautiful things are made beautiful :”
Grn’ éav tis por réyn Oe 8 Te KaXOY eoTLY OTLODY, } OTL ypapua
evavOes éyov 7) oxiwa 7) GAO 6TLdY TAY ToLoUTwY, TA [EV
ara yalpew éd, Tapartowar ydp év Tois aAXOWs Tact, TOUTO
Sé dmrds Kal atéyvas wal icws e’nOws exw Tap éuavTo,
OTL ovk GAO TL Toles a’TO KadOV 7 1 EKelvoU TOU Kadov

elte tapovala elite Kowwvia, elite brn 8) Kal bTwS TpoTayo-

pevouévn’ ov yap ére totto Sucxyvpifouar, adAN OTe TO
Kako TavTa Ta Kadd yiyveras Kardd. 1000. Thus, when
the Phaedo was written, Plato held that the particular is what
it is by reason of the presence of the idea. In these two
passages taken together we have, I conceive, a dogmatic and
precise statement of views entertained by Plato at one period
of his philosophical development in regard to those eternal, im-
mutable, separate existences, which he postulated under the
name of ideas and conceived to be the proper objects of
knowledge. .

Aristotle however in certain well-known passages of the
metaphysics atfords glimpses of a doctrine widely differing from
that of the republic. He tells us (1) that Plato recognized
ideas ért tav dice: only, to the exclusion of manufactured
articles* and of relations*; (2) that he resolved both ideas and

1 metaph. t 9. 991 b 6 Kat modda
ylyverat Erepa, ofov olxla cal daxrTv-
ALos, Oy ov hapev eldn eivac’ wore
Snrov dre éevdéxera kal radda Kal elvac

Kal ylyvecOu dia roatras airlas olas

kal Ta pnOévra viv.

2 metaph. t 9. 990 b15 érc 62 of
axpiBéorepa. Tav oOywr ol wev TaY pds
Tt. mwootow ldéas, cy ov gaye elvac
Kad’ avrd yévos, of 5é rdv rplrov dv-
Opwrov rAéyovow. This passage has,
I think, been misunderstood. Zeller
in his platonische Studien p. 261 and
Bonitz in his commentary (see be-

low) suppose Aristotle to object that.
certain proofs—here spoken of as
axpiBéorepor, **quibus non solum com-
mune quidpiam praeter singulas res
esse demonstretur, sed idem esse ex-

. emplar, quod singulae res imiten-

tur—” involve consequences which
Plato had not foreseen, some of these
proofs necessitating the recognition of
ideas of relations, and others exposing
him to the objection called the rpiros
dvOpwros. They then proceed to ac-
cuse Aristotle of inaccuracy, inasmuch
as (1) ideas of relations are recognized

ne

256 THE JOURNAL OF PHILOLOGY.

particulars into two elements, rd & and 76 péya Kal Td pxpdy, —
whereof the latter was the origin both of multiplicity and of
evil; and (3) that his system as a whole bore a striking re- —
semblance to that of the Pythagoreans. The editors and his-
torians, taking for granted seemingly that in the republic the
theory of ideas has assumed its final form, either reject
Aristotle’s testimony, or reconcile it with the republic by
strained interpretations. They conjecture, for example, that
Aristotle is mistaken when he makes Plato deny ideas of
relations and of manufactured articles’; or that Aristotle is

in republic v 479 B and Phaedo 74 a sq.,
and (2) the objection called the zptros
Gy@pwiros is stated by Plato himself in
the Parmenides. The very passages
just now cited seem to me to suggest
another interpretation. Aristotle, as
I read him, says—‘ We find Plato in

_ his more precise statements of doctrine

(1) distinctly recognizing ideas of re-
lations, which orthodox Platonism
denies, and (2) urging against his own

theory the objection called the rptros—

dvOpwros. In other words, Plato him-
self by means of the rpiros dv@pwios
dealt a fatal blow to the theory of
ideas as it was conceived in the re-
public and the Phaedo, and when he
denied ideas of relations plainly ad-
mitted the position taken up in those
dialogues to be untenable. Further
criticism of that form of the doctrine
in which an idea is assumed for every
plurality of particulars called by the
same name is therefore hardly neces-
sary.’ That the republic, the Phaedo,
and the Parmenides are oi dxpiBéorepot
Tav doywy in the sense which I have
given to the phrase, seems to me in-
contestable.

To the remark made by Zeller pla-
tonische Studien p, 257 and by Bonitz
commentary p. 112, that when Plato
stated the rplros dv@pwros in the Par-
menides, he must have been convinced

that he could meet it triumphantly, ~
I cordially assent: but I infer, not
that the objection was not valid a-
gainst that form of the theory of ideas
which is criticized in the Parmeni-
des, but that, when Plato wrote that
very important dialogue, he had in
reserve a reformed doctrine, which
was, or seemed to be, safe from attack
on this side.

1 “Pie erstere Bemerkung erlautert
Alexander (z. d. St.) in einer tibrigens
nicht sehr klaren Darstellung, an dem
Begriff der Gleichheit. Um so auf-
fallender wird dadurch aber die Be-
hauptung, dass in der Ideenlehre keine’
Ideen der blossen Verhiiltnisse ange-
nommen werden; denn Platon selbst
wahlt als Beispiel fiir die Darstellung
jener Lehre nicht nur iiberhaupt solche
Verhaltnissbegriffe, sondern ausdriick-
lich den Begriff der Gleichheit. Und
ebenso, wenn behauptet wird, yon
Kunstprodukten, wie ein Ring, ein
Haus u. dgl., gebe es keine Ideen, so
ist dagegen geltend zu machen, dass ~
Platon nach Rep. x, 596f. auch in den
Werken der Kunst nur die Nachah-
mung an und fiir sich seyender Wesen-
heiten erkannte.” Zeller platonische.
Studien p. 261. See however the
Ph. d. Griechen 11 i 587, where Zeller —
accepts Aristotle’s testimony, thinking
apparently that Plato in his later years

_ PLATO'S LAYER THEORY OF IDEAS. — 257

mistaken when he makes Plato deny ideas of relations, and
that, although he is right in saying that Plato did not re-
cognize ideas of manufactured articles, this is not inconsis-
tent with republic 596 B, the mention there of the ‘idea of
bed’ not being serious’. When Aristotle says that Plato took
the elements of the ideas to be the elements of all things,
he is again accused of inaccuracy’. It is indeed admitted
on the strength of Aristotle’s testimony that at some period,
probably towards the end of his life, Plato assimilated his
doctrine to that of the Pythagoreans; but it is alleged that
the theory of numbers was a mere “appendix” to the system’,
and that the Pythagorean development has left few, if any,
traces upon Plato’s writings‘, Now, whereas in these criti-
cisms it is plainly taken for granted that the doctrine referred
to by Aristotle in metaphysics I 6 was substantially identical
with the doctrine indicated in the republic, and that the two
statements ought therefore to be consistent, I hope on some

arbitrarily modified the details of his
teaching at a serious sacrifice of general
consistency. I hope to show that
these modifications of details were
parts of a radical reconstruction of
the system.

1 “ Mensae enim et sellae non vide-
tur ideas ponere Plato, sed illo loco,
ad vulgarem intellectum quam maxime
adaptato (cf. x 597 c), haec exempla
tantummodo adhibere ad illustranda
diversa imitationis genera.” Bonitz
commentary p. 118.

2 Zeller platonische Studien 248 ff.
Ph, d. Griechen u i 628 ff. Bonitz
commentary p. 94.

3 “ Eundem vero ideas ad numeros
retulisse et idearum naturam per nu-
meros expressisse, ex ipsius libris non
possumus colligere, nedum pro certo
affirmare..... Atque hance de numeris
doctrinam, quae in ipsa Platonis philo-
sophia vix alium quam appendicis locum
potest obtinere, ii ex discipulis Pla-
tonis, qui in philosophia magistri ac-

quieverunt, tantopere adamavyerunt, ut
omissa, quod est Platonicae philoso-
phiae caput, idearum doctrina in ex-
quirenda numerorum illorum ratione
prope unice elaborarent, unde intelligi-
tur cur tantum operae iis refutandis
Ar, tribuerit.” Bonitz commentary pp.
539, 540.

4 “die uns durch Aristoteles be-
kannte Umgestaltung der platonischen
Lehre,...von der es in den Schriften
des Philosophen an allen Spuren so
sehr fehlt, dass wir sie spiater, als diese,
zu setzen gendthigt sind.” Zeller
Ph. d. Griechen 11 i 462. “Diese
Verbindung der Einheit und der Viel-
heit in den Ideen driickte Plato auch
so aus, dass er die Ideen als Zahlen
bezeichnete. Doch kann diese Dar-
stellung erst seinen spateren Jahren
angehért haben. In den platonischen
Schriften findet sie sich noch nicht.”
567. See also Brandis Gesch, d. gr.-
rom, Ph. 321,

SS ie ee

258 THE JOURNAL OF PHILOLOGY.

future occasion to shew that Aristotle distinguishes the doc-
trine which we know through the~- republic from the doc-
trine which, when he was a member of the school, was con-
sidered orthodox’, and therefore that it is not to be expected
that the two statements should agree. For the present it is
sufficient to note that in a summary of the speculations of his
predecessors Aristotle attributes to Plato views which are —
certainly not those of the republic, It is possible, no doubt,
that Aristotle has seriously misunderstood or misrepresented
his master: but if evidence can be obtained from the writings
oi Plato himself, proving that after the composition of the
republic he modified the theory of ideas in the direction indi-
cated by Aristotle, it will at any rate be worth while to take
the Aristotelian statement into account.

Now I cannot believe that, if Plato reconstituted his system,
he wholly omitted to put his new views upon record; and
accordingly I think I see in several dialogues, not only proofs

that he was no longer content with the doctrine put forward

in the republic, and signs that he had attempted a recon-
struction, but also hints, and something more than hints, as
to the leading principles of the revised ontology. It is in —
the Parmenides, I think, that Plato most loudly proclaims his
rupture with his former self. In that important work he on
the one hand criticizes the theory of ideas with a severity which

1 It may be worth while to note a
single instance. In metaph. x1 4,
where Aristotle speaks of the theory
of ideas in its original form before it

was combined with the theory of num-

bers (unOév cuvarrovras mpds Thy TOV
dpiluav dicw, aX ws vrédaBov é€
apxfs of mp@roa Tas lidas pioavres elvau
1078 b 10), he recognizes in it two ele-
ments—the Heraclitean fiux and the
Socratic definition—and no more: for
the mention of the Pythagorean school
in 1078 b 21 is clearly parenthetical.
Observing that it was the separate. ex-
istence of the idea which distinguished
it from the Socratic universal, he pro-

ceeds to note, as a consequence of the
parallelism of the two doctrines, that
‘there were ideas of all general names:’
wore cuvéBawev avrois cxeddv TE adt@
Nyy mavrwr idéas elvar THv KaOddov e-
youévwv. 1078b32. (That cxedév must
be taken, not, as by Schwegler with
mavrwv, but with rg atr@ Aéyy, is obvi-
ous.) Thus, whatever Aristotle may
have written elsewhere about a. theory
of ideas which in the case of some
general names did not recognize cor-
responding ideas, he was quite aware
that according to the original doctrine
(as in the republic) every general name |
had its corresponding idea.

PLATO'S LATER THEORY OF IDEAS. 259

I cannot believe to be simulated, and on the other with all the
air of earnest conviction insists that, except on the hypothesis of
the existence of ideas, philosophy is impossible. At first sight
ihe two positions appear to be hopelessly irreconcilable. It will
be found however on examination that the doctrine criticized
is precisely that form of the theory of ideas which is known
to us in the republic, i.e. that form in which eidos te év
Exacrov eid0apev TtiPecOar Tepl Exacta Ta ToAAA ols TavTOV
dvowa éripépowev, and that, when Parmenides asserts, Socrates
assenting, that philosophy is impossible except on the hypo-
thesis of the existence of ideas, he does not postulate an idea
in all cases in which several particulars bear the same name.
Thus the two positions cease to be irreconcilable, if we
suppose that, when the Parmenides was written, Plato had
abandoned that form of the theory of ideas in which every
general name was held to imply a corresponding idea, and was
reconstituting his system on a new basis’. In fact the FPar-
mendes seems to me to lead the way to the later doctrine just
as the Theaetetus had led the way to the earlier doctrine, and
is consequently from my point of view one of the most im-
portant of the Platonic dialogues. But partly because it is
critical rather than expository, partly because it has been for
centuries a battle-ground for controversialists, I find it con-
venient to defer the examination of the Parmenides until I have
looked elsewhere for traces of the later theory of ideas.

In the hope then, both of proving that there was a time
when Plato became dissatisfied with the doctrine of the re-
public, and of obtaining hints which may be combined with
Aristotle’s notice of orthodox Platonism as he knew it in the
Academy, I now propose to examine the ontological part of
the Philebus, reserving for future investigation other obvious
sources of information.

1 Tt will be seen that my hypothesis form ofthe theory of ideas which bears
explains the peculiar position which the closest resemblance to the Socratic
Socrates occupies in the Parmenides. ‘Begriffsphilosophie’ is now under ex-
He acts as respondent because that amination.

260 THER JOURNAL OF PHILOLOGY.

§ 2 The significance of the Philebus.

With a view to the better understanding of remarks here-
after to be made I subjoin an analysis of the opening pages of

the Philebus, giving prominence to those portions of the ar-
gument which especially, concern me, commenting in occasional

footnotes upon certain minor difficulties, and now and then,
when I find myself at variance with Badham, the latest and
best editor’ of the dialogue, justifying my dissent.

11B Whereas Philebus has hitherto argued that pleasure
—yalpew ndovn tépvis—is for all creatures good, Socra-

tes’s contention being that, for all who are capable of

it, wisdom—¢povety voeiy pepvincbar S0Ea 6p0n adnOeis
ANoy:opuoi—is better than pleasure or any thing else,

D Protarchus, who now succeeds to Philebus’s place in the
discussion, undertakes to maintain that it is the &:s or
dvaGeos of pleasure—against Socrates who holds that it

is the &is or 8sa8eo1s of wisdom,—which makes human ~

E life happy. Should it however appear that there is a
third é£s superior to both, it will be necessary to inquire
further whether the €&cs of pleasure or that of wisdom is
the more nearly related to the third or victorious €£:s,
and therefore entitled to take precedence of its rival.

128 Beginning with pleasure, Socrates remarks that plea-
sures are various; for it would be absurd to identify the
pleasure of 6 adxoNactaivewy with that of 6 cwdpovar, or

D_ the pleasure of 6 avontaivey with that of 6 dpovdv. Pro-
tarchus does not see how two pleasures, however different

E their sources, can be unlike another. In this way, re-
plies Socrates. One figure may be unlike another figure,
one colour may be unlike another colour, and similarly

13 A one pleasure may be unlike another pleasure. Perhaps,
answers Protarchus: but what then? Why, rejoins

1 I gladly take this opportunity of rendered to readers of the Philebus
expressing my deep sense of the ser- _ both in his careful revision of the text
vices which this excellent scholar has and in his acute commentary.

‘
3
ig
i
B.
i

PLATO'S LATER THEORY OF IDEAS.

261

Socrates, you take for granted that all pleasures resem-
B_ ble one another in being good, whilst I hold that some
D are good, others bad: if we persist in thus withdrawing
our respective clients from examination, the discussion

14B

1 PoBoduce 5¢ wn Twas Hdovas ndovais
evpjcouev évavrias. Il. “Iows* adda rh
To00’ Huav Braver tov Abyor; Z.“Ore
mpocaryopevers ard dvduoa bvra érépy,
onoomev, dvouarr, Réyers yap ayabd
mira elvar Ta HOéa. Td Mev OUP uy OVX
Hdéa elvac Ta dda Advos ovdels duquc-
Bnret? xaxd 6¢ bv7’ air&y rd wodda Kal
dya0a dé, ws nucis payer, Suws ravra od
mpocaryopevas ayabd avira, duodoywpr
dvopoa elvat TH dye, el Tls ce mpoca-
vaykavo., Tl ovv dn Tavrov év rais Kaxats
omolws Kal év dyabais évdv mrdcas ndovds
dyabdv elvar mpocayopevers; 13 AB. Bad-
ham’s revision of this passage seems to
me to be founded on a misconception of
the argument. ‘‘ If Protarchus asserts,”
he says, ‘‘ that they [i.e. pleasures] are
all alike, and yet must confess that
they are not alike good, he is bound to
mention some other ground of likeness.
Socrates therefore cannot be intro-
duced as asking him for a proof that
they are dya0d, but as wanting to
know, forasmuch as they do not agree
in this respect, in what else they do
agree. But the received text makes
him say: ‘ You know they are not all
good, and you are ready to admit that
they are so far unlike; and yet you
call them ail good :’ which is so absurd
that I have changed éuws into dpolws,
and put aydé’ aira and daddy elva in
brackets.” He further drops mdayra
before od and supplies re before rails
and rats before dya@ais. In my opinion
none of these changes are necessary,
while several are positively destructive
of the true sense. Protarchus has not
acknowledged that ‘‘ pleasures are not
all good,” On the contrary he has as-

necessarily falls to the ground’, Both pleasure and wis-

serted at p that all pleasures, whatever
their origin, are alike. Socrates having
replied that, just as xpayara and ox7-
para though like may also be unlike, so
pleasures though like may also be un-
like, Protarchus signifies his assent by
the word” Icws, but still does not see how
Socrates’s remark affects his inference
that all pleasures in virtue of their like-
ness are good. ‘ Because,’ returns Socra-
tes, ‘ although you have admitted that
pleasures have points of unlikeness as
well as points of likeness, you take for
granted that goodness is one of the
points of likeness. Now as this is pre-
cisely what we deny, you ought to tell
us what the characteristic is, common
to those pleasures which I call good
and those pleasures which I call bad,
on the strength of which you attribute
goodness to both my classes, Other-
wise argument between us is impos-
sible.’

So interpreted the passage is in per-
fect accord with the rest of the discus-
sion begun at 12p and ended by common
consent at 148, On the other hand
Badham’s interpretation assumes that
Protarchus has already consciously sur-
rendered the point for which we find
him still contending at 13 Bo.

The sentence xaxad d¢ dvr’ atrév
rd is then correct as it stands except
so far as concerns the syntax of the last
clause; where, inasmuch as Protar-
chus has already under pressure from
Socrates admitted that pleasures are
diverse ("Iows 13 4), instead of supply-
ing av before dvouoa with Hermann
and Badham, I would alter rpocavay-
kago. into mpocavaryxage,

262

THE JOURNAL OF- PHILOLOGY.

dom must be submitted to examination, if we would
decide whether the one or the other or some third thing

is the good. Protarchus assents, but plainly is not alto- °

gether satisfied.

Socrates therefore proposes, before continuing the
main argument, to inquire into the relations of the One
and the Many, which others, besides Protarchus, find
mysterious and paradoxical. You mean, I suppose, re-
plies Protarchus, the union in the same person of differ-
ent and even opposite qualities, as when the same person
is said to be at once tall and short, heavy and light. No,
I do not, retorts Socrates; nor yet the union in the same
individual of a plurality of parts.
doxes are now generally admitted to deserve no attention,
to have no interest except for children, to present no real

difficulty, nay to be serious hindrances to philosophical

progress. No, the paradox of which I am thinking is
not one of these. The One which is in my thoughts is
not a yuyvopevov te kal amoddvpevov, but the Unity
which we see in Man, Ox, the Beautiful, the Good. These
henads give rise to serious controversy: (1) Is each such
monad really existent? (2) How is it that each such
monad, though incapable of generation, of change, and

_of destruction, nevertheless appears in an indefinite plu-

rality of yeyvdpueva, either (a) being itself divided into as
many parts as there are yuyvoueva, or (b) being repro-
duced as a whole in each yiyvopmevor, so that it exists,

would say, évopw &v cot rovro, without

These familiar para-

In the last sentence of the Ajous
Thompson is, I am sure, right in sub-
stituting évopwv for évév: cf. 34E IIpds
rl wore dpa ravrov BNéWarres ovTw odd
Siapépovra. rad’ évt mpocaryopevouer
évouatt; Meno 72 0, Hippias maior
299 un, Sophist 247 p, Hipparchus 230 p.
Badham rejects this conjecture (1)
because his excision of dyadv elvat, cri-
ticized above, makes it necessary to con-
strue ri ravrov éviv with mpocaryopevers,
(2) because he ‘‘very much doubts
whether a good Greek prose writer

adding some participle.” Cf. however
Thucyd. 195 Giep xal év r@ Tavoavig
évetdov (cited by Liddell and Scott) and
rivales 133 .D Il6repov obv év dirovodpla
Tt ToUTO Lcov évopgs KT X.

A few lines further 13 o, where ovdév
Tirpwoke. appears in the middle of a
string of futures, I suspect that we
should read ovdé rt rpwoer. For ovdév

vt, see Stallbaum on Phaedo 658 and

Schaefer’s Gregor. Corinth. Index s. y,
rts (referred to by Stallbatm),

a

PLATO'S LATER THEORY OF IDEAS. 263

not only in itself, but also simultaneously in a multitude

c of particulars? These are the questions which are really
important: and accordingly we must now give them our
best attention.

The verbal difficulties of the passage which begins ré rye pnv
pot icov Tod cod re Kal éuod Aoyou apéoxes 14 4 and ends pu»)
Kiveiv ev Keiwevov 15 C appear to have occupied the attention of
the commentators to such an extent that they have neglected
its substance. Yet, if I am not mistaken, it contains valuable
information, both as to the relation in which the Philebus stands
to several important dialogues, and as to the general purport of
the succeeding argument. Socrates here recognizes identification
of & and roAXa in three distinct senses: (1) the identification
of the One particular and its Many qualities, (2) the identifica-
tion of the One particular and its Many parts, and (3) the iden-
tification of the One idea and its Many particulars. Of these
three paradoxical identifications, Socrates authoritatively sets
aside the first and the second, pronouncing the first to be ‘stale,’
‘by general admission unworthy of investigation,’ ‘childish,’
‘trifling, nay, ‘a serious hindrance to thought,’ and the second
to be no better than the first. On the other hand he declares
the third to be matter of grave controversy; for how can the
One, if it is eternal and immutable, be distributed amongst an
infinite number of particulars? and a fortiori, if it is separately
existent, how can it exist at once by itself and in an infinite
number of particulars? Now the same two identifications of év
and zoAXa which are here accounted trivial and uninterest-
ing, in the republic vil 523 A—526B are made the bases of
dialectical education. A rule is there provided for distinguishing
those studies which will aid us in our progress towards ovcla
from those which will not do so. Any object of sensation which
simultaneously produces inconsistent sensations needing to be
reconciled by an effort of mind—for example, anything which is
at once in different relations péya and cpixpdv, Kodpov and
Bapv—is, we are told, rapaxAntixov or éyeptixdv THS voncews,

THE JOURNAL OF PHILOLOGY.

264

inasmuch as it obliges the soul to inquire—What are To péya
and Td opixpdv, TO Kodgov and 76 Bap’? Plainly this is the
first identification of the One and the Many. Similarly the
material counter of the practical arithmetician, being at the
same time a unity and an infinitely divisible magnitude, obliges
the soul to inquire—What is ro év? and so stimulates vénous".
Plainly this is the second identification of the One and the
Many. Thus the very'same paradoxes which in the Philebus
are pronounced to be (a) Sednuwevpéva, (b) cvyxeywpnuéva...pr
Seiv Tdv TovovTwy artecOar, (c) radapiadn, (d) paddva, (e) ohd-
Spa Trois Adyous eumddia, are in the republic (a) dwelt upon,
(b) as important studies, (c) to be pursued not only by children
but also by men, (d) who must possess qualifications rarely
found in combination, (e) as the only means by which they can
attain truth. On the other hand the distribution of the idea
amongst particulars, which in the Philebus 14 ¢ is ‘a trouble to
all mankind,’ is in the republic tacitly assumed as if Plato had
never noticed that the third identification involved any diffi-
culty whatsoever. .

Similarly in the Phaedo 102B—103 4 the first identifica-
tion is discussed—in regard to the tallness and the shortness
simultaneously discoverable in Simmias—at a length for which
Socrates thinks it necessary to make a sort of apology 102 D,
whilst the simultaneous existence of avrd Td péyeBos and To év
nuiv péyeOos, of the separately existent idea and the same idea
distributed amongst its particulars—a case of the third identifi-

lel § del re adr@ dua dpaira évav-
tlwpa, wore undev wadrov ev 7 Kal Tov-
vavrlov galvecOa, Tov émixpwodvros dh
déo. dv 7dn Kal dvaykdgor’ dv év aire
Wuxh daropeiy kal fnreiv, Kivodca év éavTq
Thy &vvo.iav, Kal dvepwray, ti moré éorw
aitd 7d &, kal ovTw Trav dyuryav dv ety
kal petacrperrixwv ért rhv Tod dvyTos
Oéav 4 wept 7d év wdOnors. "ANG pévToL,
tpn, To0Td y’ Exec ovX HKiora H Trepl [7d]
avro dyus* dua yap ravrov ws & Te dpw-
pev kal ws dmreipa 7d mAROos. 524 E.
After this recommendation ofthe arith-

metic of the accountant, who works
with counters, Socrates proceeds to
speak of the arithmetic of the mathe-
matician, who uses, not counters, but
abstract numbers, as likewise possess-
ing the required tendency. (Sidgwick
is mistaken when he says that the
arithmetic of the multitude is not
‘* recommended as a part of the pro-
paedeutic of dialectic.” Journal of
Philology 11 99, 100. The two sorts of
arithmetic are both recommended, but
on different grounds.)

PLATO'S LATER THEORY OF IDEAS. 265

cation of €v and woAAd—is assumed without a word of expla-
nation’. Suey.
In the Philebus then (and I may parenthetically remark,
in the Parmenides also) Plato recognizes three cases of the
identification of One and Many. We have (1) the division of
One yuyvopuevov into Many qualities, (2) the division of One
yvyvouevov into Many parts, (3) the division of One dy into
Many yyvoueva. Of these three cases the first and the second
are set aside as trifling, uninstructive, and no longer interesting,
whilst the third is declared to require serious consideration.
Dropping the second, which both in the republic and in the
Philebus, though distinctly recognized, occupies a subordinate
position, we observe that the first and the third have important
bearings upon Plato’s theory of real existence. The fundamen-
tal principle of that theory as represented in the republic and
the Phaedo—‘Particulars are what they are by participation in

1 Tf again we turn to Meno 73 csqq.,
we remark at once a similarity and adis-
similarity to Philebus 12D sqq. Meno’s
inability to regard virtue as a év, and
Protarchus’s inability to regard plea-
sure as @ 7oA\d, have a common origin,
and Socrates in his answer to Meno
takes the same sort of line, and em-
ploys the same examples (cx7juara and
Xpwuara), a8 in his answer to Protar-
chus. On the other hand there is
nothing in the Meno to correspond to
Philebus 148—1580. The ontological
difficulty insisted upon in the latter
has not in the former come to the
surface.

It may be worth while to note in
passing another instance of an echo
with a difference, Having at the end
of the passage above summarised 15 Bc
precisely stated the difficulty to be dis-
cussed, Socrates does not immediately
address himself to his task, but first
explains the method which he intends to
pursue. ‘There is no fairer method,’
he says, ‘than that which, despite my

Journal of Philology. vou. x.

constant devotion to it, has often left
me in the lurch.’ This method is logic
with its processes of ocuvaywyy and
dialpeois, which are next copiously illus-
trated and in the subsequent inquiry
carefully applied.. The phrase which I
have just paraphrased—ov why gor kad-
Niwy 650s od’ dv yévoiro <> Fs eyo
épacris uev elus del, modddKis 5é we 757
diagpvyotca epymov Kal dropov Karéorn-
cev. 16B—echoes the words of the
Phaedrus Totvrwy 5) éywye airés re
épacrys, @& Patdpe, Tav Siaipécewv Kal
cuvaywywv, Ww’ olds Te & Néyew TE Kal
gppoveiy? édy ré Tw’ GdXAov Nyhowma Sv-
varov els &v kat él roddd mepuxde’ dpav,
TouTov duoKkw Kkarémicbe mer’ tyviov Wore
Geoto. 266 B, but with a significant addi-
tion. The method of cwaywyn and
dialpeots is not superseded, but we are
reminded that it is not infallible—a
warning which seems specially appro-
priate when the theory with which it
has hitherto been associated is under-
going a radical reform,

18

i Yee
a

266 THE JOURNAL OF PHILOLOGY. —

separately existent realities called ideas’—assumes both forms
of the paradox: each yuyvopevoy partakes of many évta, and
each éy is distributed amongst many yuyvoueva. In the republic
and the Phaedo however, while the one form of the paradox—
the yuyvopuevov’s participation in many dovta—is persistently
- dwelt upon, the other form of the paradox—the distribution of

each dv amongst many yiyvdueva—passes without remark. —
When then we find that in the Phalebus that form of the para-

dox which in the republic and the Phaedo was dwelt upon is,

not ignored, but deliberately set aside, while that form of the _

paradox which in the republic and the Phaedo passed without
remark becomes prominent, we are bound to suppose that the

Philebus was written after the other two dialogues, and repre- _

sents a later stage of doctrinal development. _
And I think I see in the structure and the style of the
Philebus evidence to confirm the theory that it belongs to a

later period than the republic and the Phaedo. The dogmatic
tone of the protagonist, the subordination of the dramatic
interest, and the frequent occurrence of characteristic hyperbata, —
all point to this conclusion. Against this may be set Zeller’s

argument that “the very question which forms the theme of
the Philebus is in the republic v1 505 B treated as a familiar
one, the two views which in the Philebus are criticized at length
being in the republic disposed of in a few sentences,” Ph. d.

Gr. 11 i 464, and Thompson’s remark that the results of the —

long investigation of pleasure in the Philebus “seem to be

taken for granted” in Phaedrus 2518. For my own part, —

holding that in very many of the dialogues it is not the subject
discussed by the interlocutors, but rather some side-issue arising
from it, to which Plato attaches the greatest importance, I find
no difficulty in supposing that he has here restated on a larger
scale his views about the contemporary controversy, not so
much because he was anxious to justify, or to supplement, what

he had said about it in the republic, as because he thus secures —

an opportunity of marking the changes which had taken place
in his metaphysical doctrine. Indeed I must confess that the

ontology of the Philebus seems to me so certainly later than

that of the republic, that, if there were (what I do not think

.
ss jada F as a oe
ae eee CC CC

ladle eit

PLATO'S LATER THEORY OF IDEAS. 267

there is) clear proof that the main argument of the Philebus
is earlier than the corresponding passage in the republic, I
‘should not scruple to regard the ontological parts of the former
dialogue as interpolations introduced by Plato himself sub-
sequently to the composition of the latter.

Whether I am, or am not, right in thinking that Plato
is here taking a new departure, it is at all events clear that he
proposes for discussion a question of profound importance to
the author of the theory of ideas, and the very precision and
formality of the statement of the difficulty (15 B: see analysis,
above) lead us to expect that some answer will be attempted.
Further, as if to preclude all possible doubt, Plato makes
Protarchus, on behalf of the company, distinctly suggest the
investigation of the difficulty, and Socrates as distinctly accepts
the challenge. Hence, when we find that Plato does not directly
answer the question, we shall not, with Grote, assume that
“he enjoins us to proceed as if no such difficulty existed,”
but shall rather suppose that he has deliberately preferred to
answer it indirectly: for when Plato is obscure, he is so, I am
convinced, intentionally, his aim being to compel the reader to
think for himself.

§ 3 The ontology of the Philebus.

The question—‘ How is it that the separately existent monad
or idea is reproduced in a multitude of particulars?’ having been
raised, and all present except Philebus having agreed that the
discussion of it should not be deferred, Socrates addresses himself
to his task.

15 D How shall we begin? he asks. Thus: the identifica-
tion of the One and the Many, which is necessarily
involved in the use of Adyos, has been made by young

16 c_ people the basis of much fallacious argument. There is
however a way by which we may avoid the confusion so
occasioned. Assuming that all things which are said to
exist are reducible to a One and a Many, and have two

18—2

268

D

TA

22 A

THE JOURNAL OF PHILOLOGY.

elements, Limit (aépas) and Indefinity (dzrecpia), in Z

investigating anything we must first take a genus ((déa),

then divide it into two, three, or perhaps more, species,

next divide each species into subspecies, repeating this

process as often as necessary, and taking care not to
attribute (numerical) indefinity to the multitude of —

species until all the species and subspecies have been
enumerated. This is the method of the dialectician,
as opposed to that of the eristic, who is not careful to
mark the intermediate steps of the division. Similarly,
when we have to begin with the indefinity of particulars,
we must not pass from them to the genus until we have
arranged them in subspecies and species. ,
Philebus, who has already. interrupted 18 A, now for

the second time asks how this bears upon the mattcr
under discussion, i.e. the rivalry of 7d0vn and dpdvncts.

Because, replies Socrates, 750v7 and dpdvyats are each of
them a One. Hence, in order to decide which of them
is to be preferred, we must first enumerate their kinds.

That, says Protarchus after he has restated the

question under discussion, will be your duty, Socrates ;
unless you know some other way of deciding the con-
troversy. Here Socrates remembers to have heard it
said that neither 7d0vn nor dpovycrs, but a third thing,

is the ayaOov: if this is acknowledged, it will no longer —

be necessary to enumerate the kinds of ydovy and
dpovnots. Now on being questioned Protarchus admits
that the life of 7do0v7 is inferior to the life of »dovy and

dpovnois combined; and similarly Socrates admits the |

inferiority of the life of dpdvycis, hinting however that
the human vods, whose claims have thus been disallowed
on an appeal to experience, is not to be confounded

with the true or divine voids. Thus, for the present at

any rate, if not finally, the original question falls to the
ground. But though neither Sev) nor dpovnots is
entitled to the first place, it is possible that one of the
two is more nearly related than the other to that,
whatever it may be, which makes the mixed life

!

bi i as

PLATO'S LATER THEORY OF IDEAS 269

desirable and good: and accordingly Socrates continues
E to assert the superiority of vods to 7do0v7y, at the same
time indicating his suspicion that even the third place is
more than 7jdovn deserves.

If however he is to maintain the claims of vods to
_ the second place, he requires other weapons besides
c those hitherto employed. Now the whole contents of the

universe may be arranged under four heads—(1) dzrevpov

and (2) mépas, which have been already mentioned 16 ¢ as
D dareipia and épas, (3) the two united, and (4) the cause
of their union, The dzrecpoy includes érdo’ dv jyiv’
halvnrat wadrov Te Kal HTTov yuyvdpueva Kal TO opodpa
kal npéwa Sexoueva Kal TO Nav Kal dca ToLvadTa TavTa:
for example, hotter and colder, dryer and wetter,
more and less, quicker and slower, greater and smaller,
all which forthwith cease to be, so soon as quantity and
measure (Td te wocov Kai TO pérpiov) establish them-
selves in the seat of the wadndov te Kal }rrov by which
the ta dzrevpa are characterized 24 0. Next, to mépas
we assign Tad 1) Seyoweva tadra [sc. TO waddov Te Kal
HTTOV, TO opddpa Kal jpéua, TO Nlav, KTr], TOUT@Y Sé Ta
évavtTia TavrTa Seyopmeva, TPOTov pev TO icov Kal icdTyTA,
: peta S€ Td icov TO SiXacwov Kal way 6 Ti wep av Tpos
E apiOpov apiOuos 7} wétpov 9 pos wétpov. Thirdly, when

the wépatos yévva or répas éyovta are combined with

the amelpov yévva or ameipa, certain yevéoess result’;

cuvynydyouev. GN tows xal viv radrdv
Spdcer* Tovtwr dudotépwy cuvaryouévov
Karaparys KaKxewn yevnoera. II. Ilolav

1 In the above summary I have been
careful not to depend on the disputed
passage 25 o—xE &. IIpdcbes 5h Enpd-

Tepov kal vypérepov abrots, kal mdéov Kal
&\arrov, kat OGrrov Kal Bpadirepov, Kal
petfov Kal opxpbrepov, kal orbca é&v TH
apoabev ths Td padddv Te Kal HrTov
Sexouevns érifeuev eis &v dicews. II.
Ts tov daelpov Aéyers; =X. Nal. cup-
plyvu 6é ye els abrhy 7d wera tatra rhv
ad Tod wéparos yéwar, II. Iolav; =.
“Hy kal viv 5% déov tyuds, xaOdrep Tip
Tov Gmelpov cuvyydyouev eis &, ovTw

' kal THY Tod meparoedois suvaryayeiv, ov

kal was Aéyers; Z. Ty rod icov kal b-
mraclov, Kal ordon mater mpds dAAnAa
ravavria Suapdpws Exovra, oiumerpa dé
kal cUupuva évOetoa dpiOudv dmrepydverat.
Tl. Mavddvw* galver yap por Aéyew,
puyviot radra yevéces Twas ad éxdo-
Tuv cuuBalvew. This passage as it
stands abounds in difficulties. Take
first the sentences cupplyvy 5é ye—
curnyayouev. Socrates having men-
tioned the wéparos yévva, and Protar-

270

THE JOURNAL OF PHILOLOGY.

26 A e.g. health, music, fine weather, beauty, strength, and —
Ea variety of excellences discoverable in the soul. Fourthly,

chus having asked ‘What is that?’
the reply is ‘The yévva which, whereas
we ought to have collected the yéwa
of the meparoedés just as we had col-
lected the yévva of the de:pov, we just
now omitted to collect.’ Neglecting
for the moment the parenthetical part
of Socrates’s answer, we find that the
words jv xal viv 59 od cuvnyayoue
contain a positive misstatement, the
yévva in question having been ‘col-
lected’ in the phrase 7a Tovrwy Ta évay-
tia mavTa Sexdueva KTrX 25 A, just as
the dzreipov yévva was collected in the
phrase 6é2éo° dy tiv dalynrac paddov
Te Kal yTTov yryvoueva Kal 7d oPddpa
kal. hpéua Sexoueva kal 7d Alay kal doa
Toatra mavra 24 B, Further, if we
take account of the parenthesis, (2)
the words riv Tod weparoedovs sc.
yévvay seem a strange superfluity in
an answer to the question ‘What do
you mean by the wéparos yévva?’ and
(3) while rHv Tov méparos yévvay and
To meparoedés are intelligible phrases,
THY TOU Teparoeldous yévvay has no autho-
rity elsewhere, and contains a hardly
justifiable redundancy. Next, in the
sentence which follows, (4) the words
Taurov Space. can scarcely mean “ will
do as well,” Then, (5) though the
word xaxelyn, which clearly needs ex-
planation, has intervened, Protarchus
repeats his question about the réparos
yévva, and Socrates gives the answer
which he might as well have given
before. Finally (6) Protarchus’s reply
is strangely abrupt. Of these diffi-
culties the last three disappear if, as
I suggested in a paper read before the
Cambridge Philological Society, Oc-
tober 18, 1877, the words dAN iows Kal
viv tavtov Spdoer* rovTwy duorépwv
TUVAyoMEeva KaTapayys KaKelyn yevnoe-

rat are placed after dmwepyagerar: but
the other three remain untouched. It
now seems to me necessary (1) to—
interchange “Hy xal viv 5h déov nuas,
Kabamep Ti Tov dmelpov cuvnydyouev els
év, olrw Kal THv TOO meparoedovs ouva-
yayelv, ob cuvnydyonev and Ty Tod toov
kal dutdaclov, kal drbon waver mpds
addAnr\a TavavTla Siaddpws exovra, cbu-
metpa 5é kal ctupwva, évdcion dprOudr,
amepydferat, (2) in the former place to
bracket ~xa0daep—meparoedous, and (3)
to substitute cuppiocyouévwr for cvvaryo-
uévwv. We shall then get the follow-
ing sense, ‘S. Next you must com-
bine with it [i.e. the azelpov isis]
the family of the limit. P: What is
that? §. <The family of the equal and
the double, that is to say, anything
which puts an end to the mutual dis- —
sensions of the opposites (cf. 25 a), and
by the introduction of number reduces
them to symmetry and harmony.> But
perhaps it will do the same thing now
(i.e. the appearance of this yévva will
give symmetry and harmony to our -
exposition): by the union of these two
families the third will be brought to
light. P. What do you mean by the
third family? and how is it to be
brought to light? S. I mean the other
family which we wrongly omitted to
collect a little time ago. (Cf. 23 z, where
the three yévy are mentioned, but only
two, mépas and dzepor, are taken in
hand.) P. I understand. You mean,
apparently, that if we add these (i.e.
the éparos yévva), certain generations
are the result.’ Badham anticipates
me so far as to declare transposition
necessary, and (with other alterations)
to place rovrwy—yevynoera after amep-
yagera.e He is clearly wrong in giving
to cuvaryouévwv the meaning of cumpex-

pee eee ee See Se ae eee ee ee a i oe
» EPR ie oo phn i — : *s - ;

PLATO'S LATER THEORY OF IDEAS. 271

to these three kinds—of which the first and second are
constituent elements (€& @v yiyverat), while the third
includes the results of their union (rd yiyvdpueva)—we
must now add the airia trys pikews Kal yevérews. This
is vos or dpovnais, which, as others have already seen,
orders and governs the universe, as well as the individual.
Cc The table of the four yévn being now complete, we
return to the main argument. As intelligence in the
abstract is akin to the airia, while pleasure in the ab-
stract, with its correlative pain, belongs to the dzreupor,
we may safely assume intelligence in the abstract to be
superior to pleasure in the abstract. But in order that
we may adjudicate upon the claims of vovs actualized,
and 7S5evn actualized, to stand next to that third thing
which is admitted to be the av@peérwov ayaov, the
two claimants must be studied in their species.
31 B—55c Pleasures are classified as (1) false, (2) true, the
latter class being subdivided into (a) émuornpats érropevat,
(b) aicOnoecww éropevat.

im

— Oévrwv 26 B: but I think that in this
respect the text should be brought into
conformity with his interpretation.
Vahlen thinks that all that is neces-
sary is to add ef after Spdce:, and to
remove the colon after the latter word:
but his interpretation depends upon a
misconception of the words svrnyd-
youev, ouvayayeiy, ovvayouévwv ; for,
whereas he takes cuvdyerw in the two
former cases to mean ‘enumerate in-
stances,’ and in the third to mean
‘unite’ (cuumryrivat), it is quite clear
from 23 £, 25 a, that cuvdyew here
means to ‘collect under a definition,’
in which sense the réparos -yévva as
well as the dwelpov yévva (but not the
pxrév) has been already collected.
That marginal notes and references
have in several cases been incorporated
in the text of the Philebus, seems to
me certain (see, for example, 30 4,
where the words ra rérrapa ékeiva,
mépas Kai dmeipov Kal kowdv xal should,

I think, be bracketed): and it is easy
to see that, when once transposition
had occurred, a diligent annotator
would be very likely to try his hand
upon a misplaced sentence. For
playful applications of a theory under
examination to the circumstances of
the dialogue, such as that which I
think I see in dAN tows kal viv radrdv
Spacer, compare moppwrépw dé éore Tu
rpireluv, el TL TO EUG VQ bel misrevery
Huds TA vov. 22H. amicrets yap 04, was
h Kadoupévn padnors avauvnols éorw;
"Amore pev eywye, 7 8 6s 6 Zimpulas,
ov, avTd dé TodTo, pn, Séouar abet,
mept ov 6 Nbyos, avauynoOjvat, Phaedo
73 B. ovros ovv cou o Abyos exelvy Ts
tuvdoerac; Ovdauds, tpn 6 Zimmlas.
Kal pv, 7 & Os, mpéme ye elmep Tw
DrAw Adyw Evvwd@ elvar al r@ trepl ap-
povias. pére yap, pn 6 Ziputlas.
Oiros Tolvuv, ép7n, gol od Evywdos. Phaedo
92 c.

272 THE JOURNAL OF PHILOLOGY. —

55 c—59 D Arts and sciences are classified as (1) inexact,
(2) exact. :

59 p—64.A We are ‘now in a position to ask which of the
arts and sciences and which of the pleasures are to
be mixed in order that we may obtain that combination —
of intelligence and pleasure which constitutes the happy
life. All arts and sciences, it is answered, but only =
those pleasures which are true and those which are
necessary, the need of intelligence being unreservedly ad-
mitted by pleasure, whilst intelligence resents the intro-
duction of those pleasures which are false and intense.

64 A—E The ingredients having been determined, we have
next to ascertain what it is which makes this mixed life
desirable and good, in order that we may then, as
proposed 22 D, inquire which of the two ingredients is
the more nearly related to it. It is obvious that, if it is
to be harmonious and real, the combination must possess
HéTpov or peTpLorns, Evppetpia or KddAos, and adjGeva.

65 A—66 A Let us now take these three conditions, into
which the aya@ov of the mixed life has been resolved,
one by one, and consider whether vods or 7do0vy is the
more closely related to each of them. It will be found
that vods is nearer akin than 7dovn to each of the three ‘=
—to arjOea, to pétpov or petpidrys, and to Evppetpla '
or kaos: and as the excellence of the combination
depends upon these three things, we must account vods
victorious over ndovn.

66 A—67°C Finally, it is concluded that the condiGiata of
perfect union and the ingredients of which the mixed
life consists may be placed in the following order of merit

HETPOV, METPLOV, KALpLOD.

TUMLMETPOV, KANOV, TENEOV, [KAVOV.

vods and dpovyats, which represent adnGeva.
ETLOTH LAL, TEXV GU So0Eat opOat.

nooval xabapat aurijs TIS puxis~—

(a) ET LOT MALS [? Kal apetais 63 E} ¢ ETTOMEVaL

(b) aicOnoeow éropevat.

6. [2 7doval dvaryKaiat. |

eS, i date Bigs en

~ ges

en

PLATO'S LATER THEORY OF IDEAS. 273

If the conversation imagined in the Philebus had ever taken
place, and one of the interlocutors had afterwards been asked to
say what the character of the conversation had been, he would
have answered, and would have rightly answered, that the
subject of the discussion was ethical, but that incidentally
something had been said about a metaphysical difficulty in
regard to the theory of ideas. If Plato had been asked what
the subject. of the dialogue was, he would no doubt have
answered “I leave that to your own penetration”; but I am
very much mistaken, if to himself, in his heart of hearts, the
metaphysical element of the treatise was not vastly more im-
portant than the ethical, The very pains which have been
taken to obscure the fact, serve to rouse my suspicions. In the
passage of which I have now to speak, 15 c—31 A, where the
two threads are strangely interlaced, the continuity of the
metaphysical thread, though -never really broken, is never
insisted upon; and partly in consequence of this deliberate
reticence, partly in consequence of the reappearance of the
ethical theme, Grote and others have supposed that the inquiry
into the difficulty stated at 15 B loses itself in the mazes of the
subsequent discussion. The statement of the metaphysical
difficulty is however so precise, and the opening of the investi-
gation is so formal, that we may be very sure that a solution of
the problem, if not explicitly offered, is at any rate implicitly
contained in the succeeding pages.

Now in the passage about genus, species, and particulars—
&y, joAXd, and dmewpa mdn0e, 15 D—18 D, there is, if we
except the words mépas dé cai azreipiav év avtois Evudutov
éyovtev 16 C, nothing which might not be found in an account
of the earlier theory of ideas. There is in it moreover nothing
which could possibly be regarded as an attempt to meet the
difficulty raised at 15 B in regard to the distribution of the idea
among its particulars. It is in fact, as is shown by Socrates’s
meditative questions wo@ev ov tis tavTns apEntat, ToAAHS
ovens Kal ravtolas Trepl Ta audpicBynTovpeva payns; ap évOévde;
15 D, a mere preface to the promised explanation. When
. we find then that under pressure from Philebus 18 AD, Socrates,
ig as soon as he has completed his account of the processes of

274 THE JOURNAL OF PHILOLOGY.

/

cuvaywyn and Siaipeccs, recurs to the main question and settles
it without making any use of these processes, we should, L =
think, infer, not that the attempt to explain the relations of the __
idea and its particulars has-been abandoned, but only that it has
been postponed, the contribution to the main argument serving
the purpose of separating the logical doctrine with which we
are familiar from the metaphysical novelties now to be pre-—
sented to us. It is then only what was to be expected when’
the conversation takes a turn which brings the metaphysical
thread again uppermost, its continuity being marked by means

of a direct reference (Tov Oedv édXéyouév tov TO pev areipov
Seifar taév bvTwv Td Sé mépas; 23) to the statement—made-
incidentally at an early stage of the inquiry and not referred to

in the interval—that every thing which is said to exist, not
only is resolvable into a One and a Many, but also has in itself
Limit and Indefinity (ws €& évds wey Kal éx moAdOv dvT@Y TOV -
del Neyouévwn eivar, wépas S€é Kal arreiplav év avTois EvppuTov
éyovtwy 16C). Hence we enter upon the passage which begins

' BaBatl apa, & Ipoétapye, cvyvod péev Noyou Tod Aowtrod, oxEdOv

5€ ovdé fadiov Tavu Tt vov, 23 B, expecting to find in it a resolu-
tion of the difficulty proposed for consideration in 15B; and as

we have seen that the difficulty now considered so serious was
not felt to be a difficulty at the time when the republic and the
Phaedo were written, we shall not be altogether surprised if the
resolution of the difficulty is effected by a reconstitution of the
earlier doctrine. Perhaps we may further conjecture, on the
strength of the sentence cal ydp 8%) daiveras div ddXns unyavys,

€mi Ta Oevtepela Umrép vod Topevopevoy, olov BéXn éxew €repa
Tov EuTpoa bev Adyou' Extt Oé tows via Kal Tavrd. 23 B, that the
reconstitution will involve additions to the original theory. In
this way we are brought face to face with the question raised by
Sidgwick, Journal of Philology 11 103,—How is the ontology

of the Philebus related to that of the republic? but, whereas he
and others start with the assumptions that the Philebus is
earlier than the republic, and presents substantially the same
doctrine, I hold the Philebus to be the later of the two dia-
logues, and expect to find that in the interval the doctrine has 23s"
been added to, and perhaps otherwise modified. 3

PLATO'S LATER THEORY OF IDEAS. — 275

What then is the ontological doctrine of the Philebus ?
According to the Philebus mavta ta viv dvta év TO TavTi may
be arranged under four heads, as follows:

(1) aetpia ‘indefinity’, or. mrripoe ‘the indefinite ’,—which
regarded as a moAda becomes azrevpa ‘indefinites’—includes
everything which exhibits ro warrov Te Kab HTTOV, TO Thodpa
Kal npéua, TO Aiav, KT: for example, Gepudtepoyv Kal >Wuypd-
tepov, Enporepov Kal vypoTepov, TAéov Kal EXaTTOV, OaTTOV Kal
Bpadvrepov, peifov Kal cpixpotepov, 750vn Kal dv77 (so long as
they have not been actualized by the introduction of a épas
éyov") ;

(2) épas ‘limitation’, or wépas éyov ‘limit’, ‘limitant’,
—which regarded as a mwodAa@ becomes répas éxyovra ‘limits’,
‘limitants’,—includes everything which exhibits tovtwy [sc. Tod
fadXrOv Te Kal HTTOV KTA] Ta évayTia, TPOTOV mév TO ioov Kab
iodTnTa, meTa dé TO loov TO SuTAdcLov Kal Tay 6 Ti TEP AV TPOS
dpiOwov apiOuds }) métrpov % pos pérpov, everything which by
the introduction of numbers reduces the divergent dzrespa to
symmetry and concord ;

(3) psxtov or Kowvov includes ro tTovtwy [sc. amelpov Kal
mépatos|] éxyovov array, all psxtyv Kat yeyevnuévnv ovoiar,
all direipa when bound fast by the trépas: for example, vyiea,
KaNXOS, ioxus, povatKkn, appovia’, @pat, the putxtos Bios, ndovat
actualized, whether arn Geis or Yevdetis, whether good or bad ;

- 1 Here I may notice an apparent in-
consistency which has perplexed some
of the editors. In 27 & 7do0v7 is assigned
by Philebus to the dmeipov, on the
ground that any limitation of it would
prejudice its claim to be regarded as
mavayadov. Socrates demurs to the
reason alleged; but, as is clear from
814, is otherwise content with Phile-
bus’s decision. The same view is taken
in 41p. In 8l1c however 7dovy is
assigned to the wxrévor kowdv. “These
two statements ”’ says Jowett ‘are un-
reconciled.” The two statements are
however perfectly consistent: for the
yjoovn mentioned in 27 © 314 41D is
one member of the dvds, ndovh Kal

Urn, not as yet actualized by the in-
troduction of a répas €yov—in the lan-
guage of 31 A, 7507 avr7—and isthere-
fore rightly assigned to the dreipov;
whilst the jéovy of 31¢ is ndovy actu-
alized—yiovn xara piow yryvouévn—
and therefore belongs to the pexrov.
The same confusion might have arisen
in regard to Oepudv Wuxpdv Krad, if
Plato had not, in order to guard against
it, where he means depuov kal puxpov
not actualized, used the comparatives
Oepudrepov Kal Yuxpérepor, or as in 26 A
added the explanatory words dzepa
évra.

2 >. *Ap’ ovx év pév vooos 9 ToUTWY
6p0H kowavia Thy vyelas puow eyévynoer;

276 THE JOURNAL OF PHILOLOGY.

(4) The airia: THs pikes, which by combining zépas and.
drecpov produces yéveots, is voids, the lord of heaven and earth,

which orders and directs the universe, just as the human voids

orders and directs the individual.

Il. Tlavraract pév odp.
kal Bapel xal raxe? xal Bpade?, dielpos
ovow, dp’ od rabra éyyryvoueva TavTa,
dua wépas Te amrepydcaro Kal povorkny
tiumacay redewrara tuveornoare; II.
KadXtora ye. 2. Kal why dv ye xeuudor
kal mvlyeow éyyevdueva TO pev mond Nay
kal dorecpov aelero, TO 5é Euperpov kal
Gua cimmerpov dreipyacaro. IL. Ti pyy;
>. Ovxotv éx rovrwy Gpal re kal doa
Kanda ravra juiv yéyove, Tav Te arelpwy
kal rav mépas éxdvTwv cummyérTwv 5
IL. Ids 8 of; &. Kal ddda 57) wvpla ém-
Aelarw Aéywr, olov wel” byrelas KddXos Kal

loxiv, kal év Wuxats ad mdpmrodXa Erepa.

Kalmayxada. wBpw yap mov kal sdurra-
cay TavTwv Tovnplay airy KaTiovca 7
Beds, & kahé PidnBe, wépas ovre Hndovev
ovdév ore wAHT Mover évdy év adrots, vouov
kal raiiww mépas éxdvrwy &0ero’ Kal od
ev droxvaica pis abriy, éyw 5é rovvar-
tlov dmwocwoau Néyw. 258. Badham is
no doubt right in his acute conjecture
that the words poverkiy é0uracay in the
sixth line of this extract should be
followed by re and some word signify-
ing a genus of which ovorky is a spe-
cies, but I feel no confidence in his
suggestion that rekewrara (which seems
to me an appropriate adjunct) is a cor-
ruption of re Nevornra. There is a dis-
tinct reference to the passage before us
in 8lo &. Kowov rolyw traxovwper 6 87
Tay TeTrdpwy tpirov édéyouev. II. *O
pera 7d drecpov kal mépas reyes; &v G
kal Uylevay, olua 5é¢ kal dppovlav, érl-
eco; But, whereas Protarchus’s cita-
tion in 31c¢ has all the appearance of
being exact, and certainly ought to be
exact as regards dpuovla, seeing that
the word, having been incidentally in-
troduced, gives Socrates his cue, in

=. ’Ev bé dé?

25Esqq. dpuovla is not mentioned. May

I suggest that the requirements of both
passages would be satisfied if we were
to read in 264 kal povotkjy Evuracay
0 dppovlay reXebrara Eweornoaro? This
conjecture is not as tempting as Bad-
ham’s ingenious re Aedtryra: but

it is conceivable that a scribe who —
had before him ZYMTTACANTEdPMO=-
NIANTEAEMTATA might drop a cou-

ple of words in consequence of the
recurrence of the letters aNTe. It
is worth while to note that Olympio-
dorus mentions vyle.a, dppovla, orot-
xXelwy rags, and wpdv meplodos as the
Tmapadelyuara here adduced.

The editors have not been able to
agree about the goddess mentioned
towards the end of the extract. It
seems to me that we have a clue in
63E Kal rpds Tatras Tas pel? Vyelas Kal
TOU cwppoveiy kat 59 Kal Eyumrdons dpe-
THs Owdca Kabdmwep Oeot dmadol yeyvd-
Mevat avr yn Evvaxodovdovar rdvrn, Tavras
blyvv, where vylea and dpery are to-
gether conceived as one goddess. Soin
the passage before us, tyleva in the body
(with xaos and icxvs, cf. Aristot.
topics 116 b 18) and povorxy in the soul
(with wdymod\Xa erepa kal méyKada, i.e.
the virtues) are together conceived as
one goddess, whom, if pressed for a
name, I should call dpyovia. Plainly
Plato here pythagorizes: ef. Diog.
Laert. vir 33 ry 7° dperqv dpuovlay
elvan kal rhv dvyleay Kal TO d-yabdv a&rav
kal Tov Gedy’ 5d kal Kab? dppoviay cuvec-
tavat Ta 8Xa, The whole passage recals
Symposium 185 z—188 p, where (as here)
vyleca and povorxn are the two most
prominent manifestations of dpuovia,

Ae

PLATO'S LATER THEORY OF IDEAS. 277

Further of these four yévy, the first and second are together
spoken of as é& dv yiyvetas Tavta and as 70 Sovdevov els yéveow
aitia, the third as ra yvyvoyeva and as 70 Trovovpevoy, and the
‘fourth as to wdvta tavta Snpiovpyody and as Td ovody.
Finally, we must not forget the important phrase ws é& évds pev
kab éx modAOv OvTwv TaV del Neyouévav eivat, Tépas dé Kal
ametplav év avtois Eipuputov éyovtwy, which shows that the
processes of suvaywyn and diaipeors find a place in the new
system as well as in that of the republic and the Phaedo.

Now the exposition summarized above, though for the most
part precise and even perspicuous, is in one point difficult and
perplexing. The épas éyov appears to perform a double
function. On the one hand it converts non-existence into
existence: on the other hand it converts what is bad into what
is good. But if the function of the wépas éyor is to convert
what is bad into what is good, so that one mépas éyov in con-
junction with one dzreipov produces vyieva, and another wépas
éyov in conjunction with another dzrepoy produces povorkn,
how is what is bad produced, for example, ‘disease’, ‘ discord’ ?
If ‘disease’ and ‘discord’ belong to the puxtov, how do their
elements differ from the elements of ‘health’ and ‘music’? If

they do not belong to the pwixrov, in what part of the system

are they to find a place? It would seem however that the
latter supposition may be immediately rejected, bad pleasures,
as well as good ones, being unhesitatingly assigned to the
puxtov. We have then to ask ourselves—Under what circum-
stances does the union of zrépas éxyov and dzetpov produce what
is good? Under what circumstances does it produce what is
bad ?

Experience seems to shew that with Plato a gap in an expo-
sition does not necessarily mean a lacuna in the system. The
gap may have been intentionally left to be filled up by the
student. In such cases however Plato usually affords one or

Philebus..is appealed to, not “be- ference between him and Socrates is in-
cause /iis goddess was in question,” sisted upon. Whilst Socrates regards
but because here, as in 278 (q. v.), mépas as the airia rov ev, Philebus re-
where Philebus is again brought into gards it as the airla rot kaxds.

the conversation, the width of the dif-

278 THE JOURNAL OF PHILOLOGY. —

two pregnant hints. Now in 240, TO paddov Te Kat HTTOV hay-—
ing been taken as the characteristic of the sabes the évaytia
of To padrév Te Kal HTTOV are TO Te TOTOY Kal TO per pion, where

7d pérpiov is plainly not identical with 76 mocov, Next,
taking Plato’s example, the Qepydrepov Kat apuypoTepov, i.e.
temperature not yet actualized by the introduction of a limitant, cae
let us observe what will be the effect of introducing first wéerpiov,
secondly rocov generally. The effect of introducing the par-
ticular qoodéy called pwérpiov into Oepudtepov Kal yruyporepov or
Oepudv Kal vrvypdv, dareipa dvra—i.e. temperature not actual-—
ized, regarded as extending in opposite directions from a point
of indifference—is to produce in actuality an equable tempera-
ture which is neither Oepuor nor yvypov. But when any other
mocov is introduced into Oeppotepov Kal yruvyporepor, the effect
is to produce in actuality a temperature diverging more or less
either on the side of Qepuorepov or on that of Wvyporepov from
the equable temperature of the point of indifference. In faci,
while the union of Oepyorepov Kal ~uyporepov with any rocdy —
whatever produces an actual temperature of some sort,—it may
be, yetuov or mviyos,—there is one vrocov which produces an
actual temperature which is neither @epyov nor yuypov, namely
@pa, and inasmuch as this is the one point in the infinitely
extended line which is fixed, all the other actual temperatures
must be measured from it. Thus the one ywsxtov produced by
the union of Oeppotepov Kat ruyporepov with the particular
moaov called pérpsov stands in marked contrast to the many
puxta produced by the union of Gepudrepov Kal ~ruypdorepoy with
other zoca: tt is the one fixed standard, and therefore capable
of being known; theyare the many deviations from the standard,
and, inasmuch as, however nearly they may approximate to the -
_ standard, they can never attain to fixity, are consequently in-
capable of being known: it, as the standard, is perfect ; they,
as deviations from the standard, are necessarily imperfect, though
the more nearly any toody approximates to the pétpiov, the -. aes
more nearly the wxrdv, which results from its union with the |
a7etpov, approaches perfection. The apparently distinct func-
tions of the wépas éyov are then in reality one: for perfection and
existence are identical, and the further anything is from perfec-

Co a. re ie
tnd Tne > ere we

y ' *- een d | * 4 : iv ined ok
a ie ot “¥ as che ho aire ae ea 4
5 te cee * “ss -' c : “ 2 A vhs Ay
matt, O's Lael! os * Ee TT dee ae bs
f - 4 errr 22) tose © ~—* ~ 2
feat. Ss Bye, | Cy twee eat

iF i
wy

PoE 4et =

2° concn att

eve eed oc

~

eee

OPW ROL.

(

PLATO'S LATER THEORY OF IDEAS. — 279

-_ tion, the further it is from existence. Thus when 70 pérpuor,

i.e, the appropriate zrocoyv, is added to a given dzrespov, perfec-
tion and existence are the results. When a 7oady more or less
approximating to the appropriate zrocoy is added, the result
approximates correspondingly to perfection and to existence.
When a zrogor is added which is remote from the appropriate
qoaov, the result is correspondingly remote from perfection and
from existence. For example, perfect health and perfect music
are produced by the union in either case of the appropriate
mooov with the azecpov in question: imperfect health and im-
perfect music are produced by the union in either case of a
mooov, more or less approximating to the appropriate zocor, with
the admretpov in question: disease and discord are produced by
the union in either case of a wooov, remote from the appropriate
mogov, with the dzevpov in question; for even disease and dis-
cord must have something of order or goodness in them, or they

could not be existent’.

1 With the aboye should be com-
pared politicus 283 n—287 a. The pas-
sage being too long to be quoted, I ap-
pend a summary, in which I have

endeavoured as far as may be to pre-

_ serve the turns and expressions of the

original: ‘The art of measurement

‘includes two parts, (1) that which

deals with 76 wéya kal 7d ouixpédy in their
relation to one another, and (2) that
which deals with 7d péya Kal 7d opixpdv
in their relation to 7d pérpiov, and so
is concerned with the bare existence
of becoming (rhv Tijs yevéoews dvaryKkalav
ovctav). If we ignore the existence and
themeasurement of7d wéya kalrd ouixpdv
in their relation to 7d uérpiov, we shall
forthwith work the destruction of the
arts: for the arts regard excess or
defect of 7d wérpioy, not as non-exist-
ent, but as an existence detrimental
to their operations, and guarding
against it accordingly, in so far as
they secure pérpov, make all things
good and beautiful. As surely then as

It would seem then that in the case of

there are arts, so surely 7d péya xal
7d oputxpdy must be measured, not
merely in relation to one another, but
also in relation to 7d mwérptov: if there
is a wérpiov, there are arts; if there are
arts, there is a pérpiov; if either is
not, neither is the other. Hence to
the one part of the art of measurement
we assign all those arts which measure
number, length, depth, width, and
speed in relation to their opposites,
and to the other all those arts which
measure them in relation to pérprov
mpémov Karpov Séov and generally every-
thing which migrates from the extremes
to the middle point. It is of the last-
named part of the art that many of
the couol are thinking, when they say
that the art of measurement is con-
cerned with all things which become,
though from want of familiarity with
the processes of dialectic they have
confounded the two parts.’ —

It will be immediately seen that this
passage presumes the theory which

280

a given ameipoy we must. carefully distinguish | the: wérpvov or
appropriate zroodv from all the other toca which may be united ©
and that, where the combination is imperfect,

with it’:

I have elicited from the Philebus, pre-
cisely that part of the theory which in
the Philebus was left obscure being
here emphasized, whilst what in the
Philebus was expressly stated is here
barely indicated. There are further
some points of detail which seem to
deserve a passing notice: (1) when
Aristotle says metaph. 1 9 § 3 xard
Te yap Tovds Noyous Tovs €k Tay émioTNnUaV
elin tora mdvrwv bow émiorjmal eict,
he may very well be thinking of that
part of the passage before us which is
summed up in the words ows dpa
iyyntéov dpolws ras Téxvas mdoas elvar
kal peifov Te dua kal €\arrov perpetarbar
bh mwpds GAXn\a pdvov add Kal mpos
Thy Tod perplov yéveow. 284D. Plato here
from the existence of arts and sciences
argues the existence of a uérpiov, which
(as we have seen in the Philebus)
combines with the dzeipov to produce
the fixed type or idea. Aristotle re-
plies that in that case there should be
fixed types or ideas of rexvnrd, where-
as in the later development of the
system—with which we are now con-
cerned—in the case of oixia daxriédos
xr fixed types or ideas are not recog-
nized: (2) when Plato quotes certain
xouwot—generally admitted to be the
Pythagoreans—who say that perpyrixh
wept wavr’ éorl ra yryvéueva, and re-
interprets the dictum in the light of
what has been said about approxi-
mations to a standard, he must surely
mean that, whereas the Pythagoreans
say that ‘things are dpiOyol,’ he holds
that ‘the degree of the thing’s ap-
proximation to the standard is deter-
mined by dpiduds.’ Now this is ex-
actly what Aristotle says in his com-
parison of the Pythagorean and Pla-

THE JOURNAL OF PHILOLOGY.

tonic systems; 7d uév obv Td &y Kal
Tous dpiOuovds mapa TA mpdymaTa moujoat, —
kal un aomep of IvOaydpeo, Kal 9 Tav

elday eloaywy) dua Ti év Tots oyo.s
éyévero oxéyw* ol yap mporepor Siadex-
TiKns ov peretxov. metaph. 1 6 § 7:
(3) in view of the last words of the
foregoing -quotation from the meta-
physics, it is almost startling to find
Plato 285 a in like manner attributing
the Pythagoreans’ misinterpretation (as
it seems to him) of their own principle
to their want of familiarity with the
processes of dialectic.

1 The distinction between the pér-

ploy or appropriate woody and other

mood differing from the appropriate
mocov, appears to be indicated in the
precedence expressly assigned to itcov
and icdrynra in 25 A OvKodv Td pu dexo-
peva Tavra, ToUTwy 5é Ta évavtia wavTa
dexdueva, mpwrov pevToicoy kalias-
THTA, META OE TH cov Td StmAdaLOV
kal way 6 rl wep av mpos dpiOuov apiOuos
q mérpov 7 mpos wérpov, Tatra Eduravra
els TO mépas amodoyigomevoe KaN@s ap
Soxotpev Spay Tovro;

The latter part of this sentence

seems to imply that the roca are quan- —

tities measured by reference to the
hérpiov as unit. If so, it may well be
asked—(1) Are we then to assume that
the wood of things are in all cases
exact multiples of the associated pér-
piv? (2) Granted that this sort of
measurement is applicable to that
which is in excess of the pérpiov, how
is it to be applied to that which
is in defect of it? In fact the

numerical expression of divergence ~

from the type involves serious diffi-
culties, of which Aristotle was well
aware.

PLATOS LATER THEORY OF IDEAS. 281

while it is some zogov other than the appropriate mooov which
makes the pixrov in question what it is, the perfect pixror,
which results from the union of the appropriate wooov with
the dzrecpor in question, may be regarded as a type to which the

- imperfect yrxtov approximates’.

1 Tt may perhaps be asked—Does
the new explanation of the ontology of
the Philebus throw any light upon the
fanciful order of merit which concludes
the dialogue? I think it does. In
order to establish the claims of the
human vovs actualized against those
of yd0vy actualized, Socrates proposes
to show that vov's is more nearly related
than 7d0r7 to that which makes the
mixed life desirable and good. What
then is that which makes the mixed
life good for human beings? We can
no longer say,—as we used to do,—
that it is participation in the good:
indeed in the republic itself no attempt
was made in this way ‘to hunt the
good with one idea’ mwa liég To dyaddv
Onpetoat 64 u. It is possible however
that we may be more successful if we
take account of the new theory, that
it is ré péerpiov which makes a thing
good—rrept pérpov Kal ro wérpiov Kal
kalpiov kal mdv@ omdca roalra xpy
voulvew TH aldvov npncOa piow. Now,
that our pixrds Blos may be the avApw-
mov ayabdv, firstly, its ingredients
(which have been found to be émor7-
po. and certain jd5oval) must be good,
i.e. they must severally exhibit jer-
pudrns; secondly, they must be mixed
in proper proportions, i.e. the mixture
must exhibit éuyyperpia; and thirdly,
the result must be a reality, and con-
sequently there must be vovs to act as
airia rns pléews Kal yevéoews. Having
thus ascertained what conditions are
necessary that the puxros Blos may be
(a) a good combination (b) of properly
constituted ingredients, (c) actualized,

Journal of Philology. vou. x.

namely werpiorns évumerpla drnbea, we
are in a position (1) to decide the con-
test between volts and nédovy by com-
paring them with each of the three
requisites in turn, (2) to draw up an
order of merit. This order of merit
will be—

1. wérpiov, which, in union with
(a) émvorjun depos ovoa, and (8) n50v7
daretpos ovca, produces (a) émriorjun, and
(8) 750vy, properly constituted.

2. &dpmerpov, which determines the
proportion in which the émiurjywar and
the selected jdovat shall be mixed.

3. vous which effects the two unions,
and their subsequent mixture or com-
bination.

4. émornuan actualized,

5. selected jdoval actualized.

In fact mérpiov and évumerpoy, the
representatives of mépas, stand first and
second; the airla trys pigews comes
next; then come the mixta which are
here to be combined in a single xpéous.
To complete Plato’s list, the remain-
ing jndoval and the two dmepa, ém-
oTnun and 7dov7 not actualized, might
be added. That the four yévy should
reappear here, is very clearly indicated
at 27 p. The difficulty of the passage
as a whole is perhaps in some measure
due to the fact that we are here ex-
amining a pléis of two mixta.

That the votds which stands third is
the avOpwrwos vows seems clear, since
at 22 c (q.v.), where Socrates aban-
dons the claim made to the first place
by the dv@pwrmivos vols, he reserves the
claim of the Geos vows for further con-
sideration. Indeed, it is plainly stated

19

< 3

282. THE JOURNAL OF PHILOLOGY.

‘The puxrdy then includes two orders of existence

result each from the union of the appropriate qoady with t
dmevpov in question; for example, health, music, harmony, «
ble temperature, beauty, strength, virtue; (b) any thing’

to the appropriate zroodv, with the dzeipoy in question
consequently approaches more or less to the one fixed type. mpl

So far I have endeavoured to develope the doctrine of the
Philebus without reference to the theory of ideas. But we must eae
not forget that the purpose of the exposition is the resolution ofan

x

objection which may be raised against that theory, and that the |
objection is apparently to be met by means of modifications and = |
additions. Our next step then should be to throw the new sys-
tem into a shape in which it may be compared with the oldone.
And with a view to this we must plainly begin by asking our-
selves—In what part of the new system are the ideas to be found? -
Of the three answers which have been given to this question,
none seems to me satisfactory. When Brandis Gesch. d. gr.-rém.
Ph. 1 i 332 and Susemihl genetische Entwickelung d. pl. Ph. .
11 13 identify the ideas with the wépas éyovra, the remark im-_
mediately suggests itself, that in that case the difficulty raised
in 15 B is not renoved—the idea still exists at once by itself, 5m
apart, and distributed amongst a multitude of particulars, |
Zeller’s theory plat. Stud. p. 251 and Ph. d. Griechen i577,
that the airla rHs piews represents the ideas, is open tothe “s
same objection, to say nothing of the difficulty of reconcilingthe
hypothesis with Plato’s statements about the airéa. When -
Schaarschmidt (as I learn from Zeller) asserts that the ideas do |
not appear in this passage and infers the voOela of the dialogue,
I can only say that, though I am satisfied that the ideas are not = |
to be found either in the ametpov, or in the mépas éyov, orin |

the aitréa, I cannot accept his assertion until I have looked for a ; |
them in the puxrdv. a

oe st

a

4
‘ h- ae

33 , that pleasure, being the con- ly be right when, in the sentence ray _
comitant of a yéveais els otclay, affects ev ody vixnrnplav mpds tov Kowdy Bloy
the gods as little as pain itself: their ovx augicBnre mw vrép vod 22 c, he fs
life is, in fact,-a life of serene con- brackets rw.) : ea
templation. (Hence Badham can hard- ei

.

PLATO'S LATER THEORY OF IDEAS. — 283

Now the idea as we knewit in the republic and the Phaedo—
existing at once separately and in a multitude of particulars—is
certainly not to be found in the wixrdv. We have seen however
that there are in it (@) certain fixed types resulting each from
the union of the appropriate zrogdv with the dzrecpov in question,
and (b) side by side with each type a divergent multitude, result-
ing each from the union of a zrocdv, more or less approximating
to the wooov of the type, with the dzecpoy in question. The
fixed types are then just what the Ones spoken of at 158
were supposed to be—povddes tives adnOds odcar, whilst the
relation of the fixed type to the puxtrd congregated about it
presents no difficulty such as that which the relation of the idea
to its participant particulars, as originally conceived, was found
necessarily to involve. May we not conclude that Plato meets
the difficulty formulated at 15 B by modifying his conception of
the idea, and that the fixed types which we have discovered in
the pixtey are the reconstituted ideas’? If so, the idea is still
eternal, immutable ; it is still perfect, separately existent; it is

_ still the proper object of knowledge. It is too, in a stricter
_.sense than ever before, a One: for, whereas according to the

earlier theory it was either divided or multiplied amongst

if particulars, its unity is now never sacrificed, But (1) its relations
“to the particular have undergone a complete transformation.

Whereas in the republic and the Phaedo a particular is what it is
by reason of the presence of the idea, so that the idea is its cause,
in the Phalebus both the idea and the particular come into being
through the conjunction of two elements, an indefinite matter
and a limitant quantity. The indefinite matter is the same for
the idea and for the particular. The limitant quantity of the |

particular differs from, but at the same time more or less ap-

proximates to, the limitant quantity of the idea; and the more
nearly the limitant quantity of the particular approximates to
the limitant quantity of the idea, the more closely the particu-

i T have heard the Master of Trinity ing of the ideas as they appear in the
—from whom in common with many Philebus, but I do not know whether
Cambridge men of my generation I in other respects his interpretation of

_ derived my first genuine interest in the dialogue agrees with my own.

Plato—use the word ‘types’ in speak-

19—2

284 THE JOURNAL OF PHILOLOGY.

lar resembles the idea. Thus the relation of the particular to — (
the idea is now no more than resemblance to a type, the causal

function of the idea as conceived in the republic and the Phaedo
having been transferred to the two elements into which the

particular, in common with the idea itself, has been analyzed.
Further (2) whereas in the republic and the Phaedo, Plato, in —

the attempt to convert the Socratic logic of practical morality

into an ontology, has made himself the slave of general names,

and has assumed, wherever he found a general name, the exist-

ence of an idea, the new conception of the idea as a fixed type, ©

to which particulars approximate, implies an immediate depo-
pulation of the world of real existences. Certainly all general
names which connote divergence from types will cease to have

equivalent ideas—e.g. caxdv aicypdv adixov axorAactov Oeppov
Wuypov ndovn Avmn; and it will not surprise us if we are

further told that rd mpds tu and tad teyvnta have also been
struck off the list’. ~

In fact, the doctrine briefly but precisely declared in the
passages quoted at the outset from the republic 596 A and the
Phaedo 100 c, has now been superseded by a doctrine which
finds expression, as brief, but also, I think, as precise, in two
mutually complementary passages, the one from the Par-
mendes 132 C, the other from the Philebus 27 B: (1) dan, &
Tlappevidn, pdduota euovye katadpatverar Se éyew Ta ev

y a f -

edn taita domep Tapadelypara éotdvar ev TH picet, Ta Se
¥ , rl / \ * ¢ Bef A \
adda TovTols eorKevat Kal elvar omor@pata. (2) Ilpérov peév

/ ov ,

Towvy ameipov Aéyw, SevTepov 5é mépas, erat é« TovTwD
, \ \ / A

TPLTOV MLKTHY Kal yeyevnuevny ovolav’ tiv dé THS wlEews aitlay
\ / /

Kal YEverEws TETAPTOV Néyov apa gr) TAnppeoimy dv te; TL,

Kal mas; Whether the new theory is still incomplete, and.

needs to be supplemented by the identification of an ultimate
mépas with the dyaOdv of the republic and of an ultimate
dmetpov with the yépa of the Timaeus, is a question which
I leave to be considered on another occasion,

? Cf. Aristot, Metaph. 19.990 b 15, 991 b 6, quoted above p, 255.

ad
=
:
ae
-—
ae

ay
» x
y
~~
on |
ae
_
a
"

ut

we
“4
: 4
<a |
i +

A ry
SETS HR

L
ee:
pi 4

=

§ 4

In

PLATOS LATER THEORY-.OF IDEAS. 285

The Aristotelian summary of the later theory of ideas.

this attempt to recover Plato’s later doctrine, I have

thus far depended solely upon the Philebus, my reference to
the politicus, p. 279, being purely illustrative and supple-
mentary. I now propose to start afresh, from Aristotle’s sum-
mary of the Platonic ontology metaph. 1 6. If I can shew
(1) that this vexed passage is consistent with itself, and (2)
that the doctrine described in it is in all respects that of the
Philebus, I may, I think, at any rate claim to have made out
a prima facie case.

The principal points insisted upon in this important chapter

§1-

§ 2

§3

§ 4
§5

§ 6

(part of which I shall presently transcribe) are the following:

Though in the main Plato’s system agrees with that
of the Pythagoreans, there are certain dogmas which he
does not share with them.

These distinguishing features of Plato’s teaching are
(1) the doctrine, derived from the Heraclitean Cratylus, .
of the flux of aic@nra which consequently are not the
objects of knowledge, and (2) the theory of ideas ex-
isting apart from the sensibles which from them derive
their being, which theory was based upon the Socratic
doctrine of ethical universals.

We come now to those parts of Platonism which have
analogues in Pythagoreanism. First, the Platonic theory
of the relation of particulars to the idea (wé0e&s) differs
only in name from the Pythagorean theory of the re-
lation of things to the number (uiunots), and the one
theory is just as incomplete as the other.

Next, Plato distinguishes three sorts of existence, aic@n-
Ta, paOnpatixd, e’dn, and, as ideas are causes of particu-
lars, conceives the elements of the ideas to be the elements
of all things—ro péya cal 70 puxpdv being the An, and
To é&v being the ovaia, as it is by wéOe&ss in ro év that
the ideas are derived from 76 wéya kal To pixpdv. Now

286 THE JOURNAL OF PHILOLOGY. R

in this part of his system Plato agrees with the ‘Py- is

. . . , . :
thagoreans in making 70 év an ovcia, and in taking”
apiOuot to be aitios THs ovotas to particulars : he differ

from them however, when he makes: the dmeipov & S
duality, and calls it 7d péya kat TO puxpor, when he :
makes dpvOwot distinct from things, and when he as-
signs to waOnuatixd an intermediate position between $2k

aicOnra and edn. -

§7 Of the doctrines which distinguish Platonism from - |
Pythagoreanism, two—that of the separate existence =
of ro év kab ot aptOuot and that of the ideas—were

due to Plato’s logical studies, while the resolution of

the devpov into a duality was devised in the hope of —

tracing to the material cause the plurality of particulars.

§ 8 This last device is however a failure, as familiar analogies
seem to shew that plurality originates in form rather-
than in matter. The

§9 Such is Plato’s theory of causes: it is however plain
that he recognizes two causes only, a ti éorw and a

§ 10 material cause, the ideas being aitva rod ti éore to
particulars, and the & an altuov tod ri éors to the
ideas, while the material cause both of ideas and of

particulars is a duality, rd péya ad TO puKpov. —
In the two elements he sees the origin of good and
of evil respectively, as Empedocles and Anaxagoras did,

"Ror fuller statements of the difficulties which have been —

found in this passage and of the attempts which have been
made to elude them, I must refer the reader to Bonitz’s com-
mentary ad loc. and Zeller’s platonische Studien and Ph. d.

Griechen. It will be sufficient here to say that the commen-_ |

tators and historians, assuming the dpcOwoé mentioned in

§§ 6,7 to be identical with the ideas, agree in asking—How can E.
the same indefinite which in conjunction with the idea pro-- ae :
duces particulars, in conjunction with the one produce the
idea? and again — What does Aristotle mean by identifying 5S ae
the one, the formal cause of the idea, with the idea, the

* ea paar,
- wee ae a
> pe yo
: ee
nse 2
= ha P
~ ye
—e

st
at ‘ 1
wt ee >
rg ee SEE ea

aS 5
*

bs

7

wt

ai

PLATO'S LATER THEORY OF IDEAS. 287

formal cause of the particular?. They seem further to agree
in supposing by way of explanation that the one and the
indefinite which produce the idea are not identical with, but
only analogous to, the one (i.e. the idea) and the indefinite
_which produce the particular. They differ however as to the
exact import of Aristotle’s testimony; some of them by a
strained interpretation of his words reading into his assertions
their own explanation; while others, seeing that, if words
have any meaning, he distinctly and deliberately makes Plato
identify the elements of the ideas with the elements of all things,
are driven to the supposition that the pupil “has not quite
rightly apprehended his master’s meaning,”—“er habe Plato’s
Meinung, so weit es sich um die vorliegende Frage handelt, —
nicht ganz richtig aufgefasst,’—a supposition, I may remark,
which is insufficient, if for no other reason, because it does not
explain Aristotle’s failure to perceive that he is attributing to
Plato irreconcilable contradictions. Is there then no other
explanation ?

It is possible that the reader will be startled when I say that
in this summary of orthodox Platonism the dpsQuot which are
formal causes of particulars are not the ideas, It is true that
in § 3 Aristotle represents the relation of the particular to the
apiOucs in the Pythagorean system as identical with the re-
lation of the particular to the idea in the Platonic ; and that
in §§ 6, 7 he recognizes 70 év Kal rods apsOpovs as parts of the
Platonic apparatus; but it by no means follows, because the
Platonic idea is equivalent to the Pythagorean dpiOuds, that
the Platonic dpsOu0s is identical with the Platonic idea. It is
also true that at the end of § 5 é& éxelvwv yap xara wé0cEw Tod
évos Ta €ldn elvar Tods apcOuovs, commentators from Alexander
Aphrodisiensis downwards have assumed rods dpiOuovs to be
identified with ra e/5n; but neither Bonitz’s view, that rovs
dpiOuovs is an apposition, nor Zeller’s suggestion, that ra e/5y
is subject, rovs adpsOwovs predicate, platonische Studien p. 236,
carries conviction, ‘Thinking then that Zeller has taken a step
in the right direction when in his Ph. d. Griechen 11 1. 628 he
expunges ta e/dy, I propose provisionally to retain ta ein,
expunging tTovs apiOuovs, for which words I shall be able to

288 THE JOURNAL OF PHILOLOGY.

find a place in the immediate neighbourhood. It is truetoo 4

that § 9 seems to favour the assumption of the commentators |
and historians, but as in this § Aristotle is by his own ad-
mission, not recording Plato’s doctrines, but commenting
upon them, I think myself entitled to defer the consideration
of it until I have examined the rest of the passage. E |

Aristotle is however so far from assuming the identity o
the apvOuoi with the ideas, that in § 7 he seems to distinguish
TO év kal rods apiOuovs from them. Now ro & is expressly
declared §§ 5, 6 to be the ovaia, ie. the formal element of
the ideas, and in the second sentence of § 6, of apuOuol are
as expressly declared to be aitios tis ovoias, ie. the formal
elements of particulars. Would it not seem then that, when
Aristotle says that Plato conceived the elements of the ideas
to be the elements of all things, he understands by the formal
element 76 év kal of dpiOuot, ro & being the formal element of
the ideas, and ot aps@uot the formal elements of particulars ?
It appears to me that §§ 5, 6 might very well bear this meaning
_as they stand: remembering however that we have already
expunged the words tovs apiOuyovs, which at the end of § 5
are superfluous and ungrammatical, I venture to place them,
first prefixing a xal, after os § ovciavy 76 &. The sequence
of thought in §§ 5—7 will now be as follows:

‘Plato conceived the elements of the ideas to be the elements
of things, the material element being rd wéya Kab 0 ptxpov, the
formal element 7d & xal of dpiOuol: more precisely, the
ideas come into being from 7é péya wad rd puxpdv by peeks
in 76 év, which with Plato as with the Pythagoreans is an
ovaia; while of dpcOyuoé are formal causes of particulars, another
point in which Plato and the Pythagoreans agree: Plato differs
from the Pythagoreans however in making the indefinite a
duality (7d pwéya kal 7O puxpov), in separating the dpiOyoil
from sensibles, and in assigning to the HaOnuatixa an inter-
mediate position. Here the separation of rd ay kai of dptOpmoi
from things, and the introduction of the ideas, are the results of
Plato’s logical studies.’

It will be seen that, so far, a consistent sense has been
obtained, and that the doctrine here attributed to Plato is

PLATO'S LATER THEORY OF IDEAS. 289

exactly that which has been found in the Philebus. In fact
(1) as in the Philebus 76 te wétrpiov Kal 7d Tocoy are formal
elements of all things, rd pétpiov being the formal element
of ideas and ta toad formal elements of particulars, so here

‘70 & Kat of apiOpwol are formal elements of all things, ro &

being the formal element of ideas and of apsOyoi the formal

- elements of particulars: (2) as in the Philebus an dzretpov

called 76 paddXov Kal TO Hrrov, so here an dretpoy called ro
péya Kal TO puxpdv, is the material cause at once of ideas
and of particulars: (3) in precisely the same sense in which
in the Philebus both sections of the pxtov are reduced to
the same é& dv yiyverat, all things are here reduced to the
same orovxeta: (4) as in the Philebus the particular stands to
the idea in the same relation in which a copy stands to its
model,—the resemblance of the one to the other being caused
by the approximation of the vrocéy of the one to the pérpsov
of the other,—so here § 3 the pé@e&s of the particular in the
idea would seem to be in reality wiwnows: (5) as in the Philebus

_ TO Te péTpLov Kal TO Trocév and TO padAXov Kal TO HTTOV, SO

here § 10 70 & xat ot apiOuot and 76 péya Kal TO puixpdy, are
respectively the origin of good and the origin of evil.

Thus five principal dogmas—of which not one is consistent
with the theory of ideas as it is stated in the republic—are com-
mon to orthodox Platonism as described by Aristotle and to the
doctrine adumbrated in the Philebus. The terminology is
not indeed precisely the same, but this will not surprise us,
as Plato would naturally avoid in a written exposition the
technicalities of the school, while Aristotle would as naturally
preserve them. In other respects the agreement is exact.

Two paragraphs however still remain to be explained. The
first begins § 7 with the words 70 dé dvada Troujoas THY évépav
duow Sia TO Tors apiOuors Ew THv TpeTav evpuds é& avTHs
yevvaoOa, woTep EK Twos éxuayetov—and ends with § 8,
Here it is immediately obvious that the words tovs dpsOywovs
éw tév mpdtev are full of difficulty, and that until they have
been explained the meaning of the whole sentence must remain
to some extent uncertain. According to Bonitz (after Alex-
ander) Aristotle says—‘ Plato’s reason for making the material

7i*.*)..

290 THE JOURNAL OF PHILOLOGY.

cause, in name rather than in fact, a duality (“ quod infiniter. a
materiae naturam, verbo quidem magis quam re ac notione
duplicem fecit”), was, that numbers, 1.e. mathematical numbers, ae
with the exception either of primes or of odd numbers, are =
generated by the help of the number two’ (“quia numeri —
quorum ad naturam vel similitudinem ideas suas redegerat, aoe
magnam partem dyadis ope progignuntur, exceptis nimirum =
numeris vel indivisibilibus vel omnino imparibus”). In other ~ 4
words, Bonitz supposes Aristotle to say that, when Plato came —
to name his dzrespov, he preferred the dual title wéya Kal pexpov, —
because some mathematical numbers are generated by the —

number two. How Bonitz connects the sentence in question

with § 8 is not clear. Further, in regard to the rpa@tos apuOpot
here excepted, no agreement has been arrived at. Bonitz

hesitates between primes and odd numbers generally : Trende-

lenburg de ideis et numeris p. 79, Zeller platonische Studien p. -
255, and Schwegler ad loc. suppose the ideal numbers to be -

intended: Brandis Gesch. d. gr.-rém. Ph. 11 313 takes them to

be those ideal numbers which are odd. That these interpreta--
tions are anything but certain seems to be admitted even by —

their authors. Does the new conception of the theory as a
whole throw any light upon these incidental sentences ?

We have seen in the Philebus that each dzrespov is a duality _
in the sense that it extends in opposite directions from a point —

of indifference. It is further plain that in § 8, which clearly
should be read in conjunction with the sentence now under ex-
amination, Aristotle refers to the plurality of particulars. Hence

if the mss had exhibited a hiatus where the troublesome words ~~
Tovs apiGuovs é€m THY TpeTa@Y Now occur, no one would have —

scrupled to paraphrase the passage as follows: ‘ Plato’s reason

for making his material element a duality [i.e. for making it —
extend in opposite directions from a point of indifference] was,

that this hypothesis made it easy to suppose the generation —
from it of <a plurality of particulars>. Familiar analogies —

seem however to shew that the origin of plurality should be

looked for, not as Plato supposes in matter, but rather in form:
for example, one table only can be produced from one piece of ae
matter, whilst the joiner, who impresses form wpon the matter —

a

f we
ae de ean him
es BASE Trait: wh Die Oe
Z te rd ve ae Dee tas
th eae d a. 2 ge Rel

: J ‘
cack Sera

Lah dn ja

|

Le
7

" Ne
y

re

;

. “4 a

- " rh ‘
7 + J ‘ \
> JS AEG . 5 P “¢ -

. a . 5 / area >? i - J Fied
A eee Se A hee eee 2 a tp ba - a
ane ey ideal ih SC 3 hock _ : Pag) 3.0 ane Bi - bili cdl Aa

eo
ete

PLATO'S LATER THEORY OF IDEAS. 201

in question, makes several tables; and in like manner the
analogy which Plato himself has used (kal 51) cal mpocexdoas
mpémet TO pev Sexopevov pytpl, TO 8 bev watpl, tHv 8&e
peta&éd Tovtav diow éxyove, Timaeus 50D) may be effectively
turned against him.’ The sense thus obtained being unexcep-
tionable, the question now suggests itself—Is it possible that
Tovs apiOwovs Ew Tév mpdtwv means ‘the multitude of par-
ticulars’? Here Aristotle comes to our assistance. We read

in the physics 219b 6 dpiOpmos éore Svyds' Kal yap TO apiOpuov-

pevov Kab [7d] apiOunroyv apiOpov réyouer, kal @ apiOpovpev :
whence it would appear that there is nothing to prevent us from
using the word dpvOuo/ on the one hand in the sense of ofs dp.0-
podpev to denote the mood of the Philebus apart from any
parXov Kab Troy, and on the other hand in the sense of ta
adptOunra to denote the mood of the Philebus taken in con-
junction with some padrov Kai ArTov. In this latter sense
however, the idea, being a combination of 7d wétpioy or Td &
with TO padXov Kab TO ATTOY or TO péya Kal TO pLKpOY, is itself
an dptOuos, here called szrp@ros to distinguish it from the dpsApot.

or apuOuntra before mentioned. (In fact the mpa@ros dpiOuds

of Aristotle is the évds of Philebus 15.4.) Thus by tovs dp.6-
povs €&m Tav mpetov Aristotle means apiOunra arising from
the union of a péya cal pixpov with dpiO poi or ots apiOuodpev

' as opposed to apsOunra arising from the union of a péya cal

pxpov with the év. It will be seen that the explanation here
given of the double sense in which the word apsOuds is used,
applies to a considerable group of passages, which might other-
wise have been thought fatal to my interpretation of the press
TO &v Kal of aptOpotl.

It remains to say a word or two about § 9, where Aristotle
from his own point of view briefly comments upon the system
which he has been describing. Here he certainly assigns to
the idea the same position in relation to the particular which
the é holds in relation to the idea. That there is in this
place some confusion, inasmuch as throughout the rest of the
chapter, while the & is the formal cause of the idea, the
idea is, not the formal cause, but the type of the particular,

- cannot, I think, be denied. This confusion may be due to the

292 THE JOURNAL OF PHILOLOGY. —

hasty and careless expression of Aristotle’s dislike of a theory — |

which seemed to assign the function of his own efdos in part

to an inherent formal cause, in part to an external type. It
is possible again that Plato himself did something to create
the confusion, if, as perhaps may be inferred from Aristotle’s
statements in § 3 and § 5, he used the word pé0eEvs—which
had formed a part of the terminology of the earlier system—
to express at once the relation of the particular to the idea
and the relation of the idea to the & Or again it may be

that, though in dealing with the particular Plato discriminated —

the cause, i.e. the adpwOuds or rocov, from the type, i.e. the
idea, in dealing with the idea, he assigned to the év both func-
tions. However this may be, I can see nothing here to lead
us to doubt the general accuracy of the precise statements of
the rest of the chapter, confirmed as they are by the evidence
of one of Plato’s most elaborate dialogues.

By way of conclusion to this section I append the text of
the latter part of the chapter upon which I have been com-

;

menting, together with a translation.

‘

‘ \ / » /
mv S€ méOeEv Tovvopa podvov
pereBakev* ot pev yap TvOa-ydpetor
/ 4 et \ i a
pynoe Ta OvtTa daciv elvar TOV

apOpav, Adrwv 8: web ééeu <tav

§4 cidav>, rovvoya petaBadrdv' rv

/ U4 av \ /
pevron ye meOesiv 7 THY piunow
9 a ” a > A
yrs av ein [Tov €idav] adetcay év

owod fyreiv. | uv 8& mapa ta

Kowa €nreiy. rapa Ta

\

aicOnra Kai ta €1dn ta pabn-
4 a 4 > /

HaTLKa TOV Tpaypatwv eival dyot
, “~

petagy, diadépovra trav pev aig On-

a A bad A ign >
TOV TO aidia Kal axivyta elvan,

col > “
tov 8 e€iddv TO Ta pev wOAN

»” id > ‘
aTTa omota €ivar TO O€ Efd0s aibro

a ¢ , > ‘ > »” \
Vv €V EKAOTOV fLovov. | €7T EL 5 QLTLa TA

” a »
(6 Tots dAXos, TaKelvor oTot-

The only novelty in this doc-
trine of participation was the
term employed: for whereas the
Pythagoreans say that things
exist by imitation of numbers,
Plato changes the term, and says,
by participation in ideas: § 4
but what this participation or
imitation was to be, both Plato
and the Pythagoreans left an
Open question. | Furthermore
Plato asserts the existence of
mathematicals, distinct from sen-
sibles and from ideas, and inter-
mediate between them, differing

from sensibles inasmuch as they “

[sc. the mathematicals] are eter-
nal and immovable, and from
ideas inasmuch as of each mathe-

matical there are many similar

instances, whilst the idea is in

each case one alone. | § 5 Now —

/

;

;
‘ .

. ‘ ‘ prs ; . J 4h .

or ake IOS | a nae Wp Pap _

CAA EA ACE OTNINT ING IAAL T, VOT YRS

a ~

PLATO'S LATER THEORY OF IDEAS.

lal , 27 a »” yy

_xela tavtwv onOn Tdv ovTwv «ivat
a e . > Md ‘

oTolxela, ws peyv ovv vAnv TO

, ‘ \ n > >
péya kal TO pukpov elvat ap-

, Co Gs.— See ,a : ‘ ‘
xX4s, ws 8 ovgiav TO €v <Kal Tovs
°
apiOpmovs>. e& éxeivwv yap Kare.

. ‘ » >
peOcEty Tov Evos ta eidy elvat

~§6 [rovs aptOmovs|* To pevtou ye ev

> + > sinh oh x , ,
ovoiav elvat Kal py Erepov ye TL
4 9 ld lal
dv AéyerOar Ev, TaparAnciws Tots
” ‘\ \
TIvOayopeios eXeye’ Kai TO Tovs
3 AQ , > lal »
aptOmovs airtious elvat Tots adXors
Lad > / e 4 ° , ‘
THs ovoias waattws éxeivors. | TO
de > ‘ a ? , e ea) 5 tO
€ avr Tov ameipov ws Evos Svdda
a \ »” 2 ,
Towjoat, TOS ameipov ek weyadou
\ a a> »¥ \o» €
KQL PLLKpOd, ToT tOLoV. Kal ETL O
X\ \ > ‘\ \ \ >
pev Tovs aptOpovs rapa Ta aicOy-
U4 e P. ©e . 9 ‘
ta, ot 8 apiOpuors eivai pacw avira

\ , \ \ \
TO, TpAyyaTa’ Kat Ta pabymaTiKa

§7 peragd rovtwv ov TiBéacw. | TO pev

> en \ \ > N
ovv TO Ev Kal Tovs aptOpods
\ \ , aA \ \
TAA TA. TPAyLaTa ToLpoat Kal [Ay
4 e /, .Y e A
worep ot IvOaydperot, Kat 7 Tov
20.4 oy \ \ ‘ > a
elddv eicaywyn dia tHv ev Tots
Adyous éyevero oKapv* ot yap mpo-
A > A ‘
Tepot OiaAeKTUKNS OV peETELXOV. TO
\ / A \ | ey 4 ,
dé dvada roujoa tHv Erépay prow
‘ \ ‘ > ‘ 5d a
dia TO Tovs aptOports ew Tav
la > “A > 3 A
TpwTwv evpvds e€€ avTns yev-

lal 4 ”
vaobo. worep EK Tivos expayelov.

§8 Kaito. oupBaiver y evavrins' od

yap evAoyov oUTws. of pev yap éx

Ts vAyns woAAa rowodtow, To 8

293

since the ideas are causes of all
besides, Plato conceived that the
elements of ideas are the ele-
ments of all existences. Thus in
his system the great and the
small are material causes, and
the one and the numbers are
formal causes. From the great
and the small the ideas are de-
veloped by participation in the
one: § 6 where indeed he re-
sembled the Pythagoreans in
making the one an existence and
not a mere predicate of some-
thing else which exists. He also
resembled them in his further
doctrine, that the numbers are
the causes of the existence of all
things other than ideas. | But
the substitution of a duality for
the indefinite taken as a unity,
and the resolution of the indefi-
nite into a great and a small, are
peculiarities of Plato’s, Again,
whereas he makes the numbers
distinct from sensibles, they say
that numbers are the things
themselves: and [whereas he
does,] they do not, assign to ma-
thematicals a position interme-
diate between higher and lower
existences. | § 7 The separation
of the one and the numbers
from things, as opposed to the
Pythagorean doctrine, [which
identifies them,| and the intro-
duction of ideas, had their origin
in Plato’s logical speculations,
his predecessors not having culti-
vated dialectic. His reason for
making the other [i.e. the mate-
rial] element a duality, was, that
[on that hypothesis] the numbers
other than the first [i.e. particu-
lars] were naturally generated
from it, as from a lump of wax.
§ 8 Facts are however against him
—the theory is untenable: for,

Cr

$10 tots 8 etdSeor to Ev.

204 THE JOURNAL

” a , 5 ,
eldos amag yevva povov’ daiverar
al @ / / c y
S ek puds vAns pla tpamela, o de
@ a ‘
70 €ldos émipépwv eis av moAXas
a” e , a ‘ .
Tove. Opoiws 8 exer Kal TO appev
OR eee ae ee ae ae
mpos To OAV" TO ev yap Vio pds
a > , ‘ > x
mAnpodrat oxeias, To 5 appev mroA-
‘ a / a ,
Ae wAnpot. Kairot TavTa pipnpa-
a > a > , by , ,
Ta TOV apxav éxeiveov éeortiv. | TIAa-
\ = \ “ tA
Tov pey ovv mept Tov Cytoupevon
9 ‘ > A
ovtw diipirev’ davepov 8 ék Tar
5 4 7 “ $f ; > \
eipnevwv ott dvow airiawv éotl
, , a A /
POvov KEXpnpevos, TH TE TOV TL
> © ies ee, ee er ee
€OTL Kal TH KaTA THV VANV’ TA yap
¥ a t Aa 3 v a +”
eid Tov Ti éoTL aitia Tots aGAAots,
\ , e
Kat tis 7
A ee , > & XN
vAyn 7 vroKepevyn, KaP As Ta €1dy
pev exit trav aicOyrav to 8 ev év
Tois €ldect AeyeTat, OTL avrn dvas
> ‘ / A \ , os
€OTL, TO WEya Kal TO PLKpOV. ert
de ‘ “A = \ “a a“ ; Peer.
€ THV TOU EU Kal TOU KAaKOS aiTiay
lal , > / e ,
TOUS OTOLXELOLS AmédwKEV ExaTépots
e , 7 +” \
EKATEpaV, woTEep <E>gayev Kal
> ed /
erilnTnoat

procodpuv, oiov "EpmredoxAéa Kat

TOV TpoTepwv

“Avagayopav’.

1 In printing the above extract I

have made three alterations of the
text, which need a word of explana-
tion: (1) I have tried to shew above
p. 287 that though the word dp.dyués
occurs in both systems, the Platonic
equivalent of the Pythagorean dp:0uds
is not dpiOuds but eldos. Hence in § 8
we must, at any rate in thought, sup-

Twas.

f

whereas his school acvime iui /

tude from matter, supposing the
form to generate once for all, we
find that one table is produced —

from one piece of matter, whilst ~_

the one person who impresses the —
form makes many tables, So it

is likewise with the sexes: the —
female is impregnated by a single —
congress, while the male impreg-

nates repeatedly. Now the rela-
tion of the carpenter to the piece
of wood, and the relation of the —
male to the female, are similar to
that of form to matter. [Whence
it would appear that Plato is not
justified in assuming matter to
be the origin of multitude.] | § 9

Such was Plato’s decision of the

OF PHILOLOGY. — ae =

points at issue. Enough has been ~

said to shew that he employs two
causes only, a formal cause and

a material cause, the ideas be- —
ing formal causes of all other
existences, and the one a formal .
It is also

cause of the ideas.
clear what the material substra-
tum is, to which the ideas are

attributed in the case of sensibles, _

and the one in the case of ideas:
it is a duality, the great and the
small. Further, he assigned to

the two elements respectively the

origin of good and the origin of

evil, like certain earlier philoso-

phers whose speculations we have
already noticed,—I mean Empe-—
docles and Anaxagoras.

ply after TAdrwy 52 wedéter the words
Tov eldav, while in § 4 the words ray
eldGv, though appropriate to tiv pé-
OeEv, are not appropriate to ray pl-
Know.
racy I have ventured on a transposi-

tion, though, had the inaccuracy been __
I should have thought little of
it: (2) In § 5, for reasons explained —

single,

In view of the double inaccu-

yr

re ee ; weet: vr c

_ Te? 1" 1s 7 ia .
a, ~> : eee, > . 7 ~*~

_ PLATO'S LATER THEORY OF IDEAS. 295

w

§5 Concluding remarks. —

_ It will now be possible to frame a provisional theory of
Plato’s doctrinal development.

I. Starting from the philosophical scepticism which he
had learnt from the Heraclitean Cratylus, Plato seems for a
time, like his master Socrates, to have found employment for
his intellectual energies in the construction of general notions
(Adyos, vmoVécers), within the domain of ethics. That these
general notions are not knowledge in the strict sense of the
word, Plato was quite aware: but this in no wise troubled him,
as in this stage, like Socrates, he held knowledge properly so
called to be unattainable.

II. Overcome by the craving for knowledge properly so
called, he cast about for some method of extracting it from the
Socratic general notions. In order to this it was necessary
(1) to assume that each general notion represented not only
what is common to a multitude of particulars, but also, in
an imperfect way, an eternal and immutable existence,
separate from particulars, and (2) to devise a method of
converting the imperfect representation of the eternal and
immutable existence into a perfect representation of it. The
theory of ideas as we see it in the republic and the Phaedo
is the assumption above named dogmatically expanded into
the following propositions: (a) wherever we find a plurality
of particulars called by the same name, there is, separate from
them, an eternal and immutable existence, which we call

above p. 288, I have added xa rovs
apOuovs after ws & ovclay ro év, and
bracketed rods dpiOuots after 7a etdn

‘elvac: (3) The last sentence of the

extract—ér. dé ri Tov ed Kai TOD KaKa@s
airlavy. krA—appears to contain a di-
rect reference to 3 § 17. 984b 18 and
4 § 3. 985 a 5 (cited by Bonitz): I have
therefore written épdayev for paper.

Further I have throughout ‘spaced’
the Platonic technicalities in order to
distinguish the Pythagorean dp:0uol
§§ 3, 6 from the Platonic dpiOuol, & in
the ordinary sense of the word §§ 4, 6
from the Platonic é, and eZdos in its
Aristotelian sense of ‘form’ § 8 from
the Platonic ¢i6.

296 THE JOURNAL’ OF PHILOLOGY. ~

an idea; (b) each particular is what it is by reason of the —
presence in it of the idea which bears the same name. In
the republic and the Phaedo Plato further propounds a scheme
for the requisite conversion of that imperfect representation of ©
the idea which the general notion affords into that perfect
representation of it which would constitute knowledge properly —
so called: but he frankly confesses that there is in the scheme
a gap which he has not succeeded in bridging. In this stage —
then Plato tries to attain knowledge properly so called through
ideas, but sees as clearly as any of his critics that the attempt
is unsuccessful. Indeed the theory of ideas, which was to be —
. the basis of the higher logic, is itself open to serious objections : _
(a) if we are to postulate an idea wherever we find a plurality
of particulars called by the same name, the argument com-
monly called the tp/ros adv@pwiros may always be urged against
us, and (8) it is impossible to understand how the idea can be
distributed amongst particulars without sacrificing its unity
and its separate existence.

TIT. In order to meet these objections urged against the
theory of ideas, Plato in the Philebus (and elsewhere) amends
his doctrine. Whilst he still postulates eternal, immutable ex-
istences, separate from particulars, he withdraws the assertions
(a) that, wherever a plurality of particulars is called by the
same name, there is an idea to correspond, (b) that the particu-—
lar is what it is by reason of the presence in it of the idea
which bears the same name. He now regards each idea as an
eternal, immutable type in nature, produced by the union of an
appropriate quantity (€v) with a given matter (uéya cal pixpor),
and the allied particulars as divergences from the type, pro-
duced by the union of a quantity (dpv@u0s), differing more or
less from the appropriate quantity, with the matter in question.
Thus the idea is now a trapadevypa, the particular, in virtue of - _
the approximation of its api0ués to the év of the idea, being —
a ofolwua. In this way Plato provides himself with eternal, _
immutable existences mapa ta aicOnrd to be the objects of
knowledge. In the Philebus he makes no attempt to explain
how the knowledge of them is to be obtained: but I hope

PLATO'S LATER THEORY OF IDEAS. — 297

hereafter to show that, whereas in the period of the republic
and the Phaedo it was proposed to pass through ontology to the
sciences, in the period of the Parmenides and the Philebus it
‘is proposed to pass through the sciences to ontology. It is
possible that the statement of the theory of ideas which is
contained in the Philebus was afterwards modified or supple-
mented, but its exact agreement with Aristotle’s summary in
metaph, I 6 would seem to show that it continued to the last
to be in the main a correct account of the Platonic ontology.

The time has not come for attempting to criticize the theory
which has been unearthed, or to trace the consequences of the
discovery, if such it is. But even in this early stage of the
inquiry it is easy to see, that, if the later theory of ideas was
what I have supposed it, Aristotle’s attack upon Plato assumes
a new aspect, in so far as, form and matter being already pro-
vided for the particular in the shape of apiOuds and péya Kat
putxpov, the paradeigmatic idea with the associated doctrine of
pébckis or plwnows is from the Aristotelian point of view a mere
excrescence. It is easy to see too, that, if, as I conceive, the later
theory is represented in certain of the Platonic writings, we
shall obtain an important criterion for the determination of the
order in which they succeeded one another. Again, it may
perhaps be found that the study of the later dialogues from this
novel point of view throws new light upon the teaching of
Plato’s Pythagorean contemporaries, as well as upon that of his
academic and neoplatonic successors. On some of these sub-
jects I hope to say something hereafter, but my first task must
be to complete the examination of the original authorities. _

In the foregoing pages I have endeavoured to interpret and
to apply two passages, the one in the Philebus, the other in the
metaphysics. The special novelty of my interpretation of the
former consists in the discrimination of the puétpsov and the
moody and the assignation of the ideas to the mwuxrdv. The
special novelty of my interpretation of the latter consists in the
recognition of 7d év xal of aptOuoi as the formal element of all
things, év being the formal element of the idea and apsAuds the

Journal of Philology. vow x. 20

298 THE JOURNAL OF PHILOLOGY.

formal element of the particular. In the course of the inquiry
four important propositions have emerged, (1) that internal evi-
dence proves the Philebus to have been written after the repub-
lic and the Phaedo, (2) that in the first-named dialogue a new
and improved theory of ideas is traced out, (3) that metaph. 1 6
contains a consistent account of orthodox Platonism, (4) that

the doctrine ascribed to Plato in metaph. I 6 is precisely the

doctrine of the Philebus.

As I have found myself throughout in antagonism to two

great scholars whose names are honoured wherever Plato is
studied, it seems fitting that the last words of this paper should
express the admiring gratitude which I feel towards Eduard
Zeller and Hermann Bonitz. If, as I am bold enough to
imagine, I have added something to their results, it is their
writings which have enabled me to do so. In any case ov«
épifopev, AAA Stareydpeba.

HENRY JACKSON.
12 Dec. 1881.

- hr orga Rare aoiate these nes. yoeete

THE SIMILE OF THE TREACHEROUS HOUND IN THE
AGAMEMNON.

Lal a) > / > > /
veov T étapxos Idiov tT avactatns
ovK oidev ola yA@ooa puonTHS KUVOS
AéEaca Kaxtelvaca hadpovovs Siknv
y —f , a /
atns NaGpaiov revEerar Kann TvYD.

Zsch. Ag. 1227—1230.

A WELL-KNOWN scholar, who recently revived not without
profit the discussion of the Sécas EvpBdorarat, observed that the
reappearance of the subject would probably raise a smile. In
inviting the student of Aischylus to yet another consideration
of the prontn xvov I certainly feel the same hesitation, and am
therefore encouraged to the same perseverance.

In the most recent edition’ of the Agamemnon (A. Sidgwick,
1881) the simile is abandoned to the obelus; the following is
the editor’s note—

“1228 ‘Knows not what things the tongue of the vile she-
hound, with long-drawn smiling welcome...shall accomplish by
evil fate.” This is the best sense that can be made out of the
text as it stands; but ofa is a clumsy and unlikely accusa-
tive for ofwy, and daidpovous is a very strange adjective, and
the use of adj. for adv. is harsh with éxreivaca: and we can
scarcely resist the conviction that the text is corrupt. On the
whole Madvig’s alteration (following Tyrwhitt) is the most
probable and is certainly highly ingenious ; he reads:

ovK -oldev ola yA@ooa pLoNnTHS KUVOS
NelEaca Kaxtelvaca phadpov ovs, Sixnv
wv / if A /
Arns NaPpatov, dnketar Kaxh TVyN
1 Written in October, 1881.
20 —2

300 THE JOURNAL OF PHILOLOGY.

....knows not what a tongue of the vile she-hound has licked
(his hand) and stretched out a joyful ear, and now like a
stealthy curse shall bite him by evil chance. The violent
stretches of language, making the tongue (instead of the dog)
stretch out a joyful ear and bite, are hardly too strong for
/Eschylus. Still they are strong, and d5feras for rev&eras is
a considerable alteration; so I have not ventured to put the
conjecture into the text.”

Most readers will share Mr Sidgwick’s objection to the
MSS version, and approve his discretion in refusing admis-
sion to that of Madvig. For myself, indeed, I cannot, with
the utmost deference to:the author of the correction, give to
it even such a qualified approval as Mr Sidgwick gives. The
only acceptable thing in it is the Ae/Eaca of Tyrwhitt, which
indicates the true point of the comparison, namely, that the
glozing welcome of Klytemnestra is a preparation for her
treacherous stab, as a dog will lick the confiding hand which it
purposes to bite. I am little disposed to quarrel with anyone
on the shades by which boldness in language is discriminated
from absurdity, but I think it should be proved by some similar
instances that Auschylus could possibly describe a tongue as
putting out an ear and biting. It is no defence that yAdooa
puonThs Kuvos may stand for cua@y picyntn yAOocay éxovca

or the like. Of course it may in proper places; but poetry

cannot be constructed or analysed like a term in an algebraical
equation, To say

narratur et prisci Catonis

saepe mero caluisse virtus,
though it was Cato and not his virtue that warmed, is sense and
poetry, because the phrase virtus calet suggests no visible image at
all, and consequently cannot suggest an absurd one. But yAdooa
exteivaca paidpov ods dy€erar does suggest a visible image, and
that image is ridiculous. Still there is scarcely a limit to the
vagaries of the imagination, and if the rules of critical evidence
seemed to shew that Aischylus made a dog’s tongue bite, we
could only sigh and acquiesce. But this gem of metaphor is
scarcely worth purchasing at the expense of such an alteration
as tevEeras for diferar. Nay, even datdpov ovs, the charming

MISHTH KYON. 301

simplicity of which has lulled the suspicions of criticism, might
not have been so effective against Agamemnon. I cannot
pretend to an intimate acquaintance with dogs, but according
to my small experience the canine manner of making friendly
overtures is not at all happily described by ‘stretching out a
joyful ear, whether this means (for it is not determined)
‘pricking the ear’ or ‘laying it back, The amicable lick is
familiar to everybody, and it is often accompanied by rubbing
the head against the hand saluted, but the ear—I put this
merely as an enquiry—though the chief organ for expressing
excitement, attention, etc., is in the coaxing mood merely pas-
sive, and the position of it would depend on the breed of the
dog *. |

But while I go with Mr Sidgwick in rejecting this, the

‘best of the attempts to make sense out of 1229, I think

that the verse should not be given over until it has at least
been considered from an entirely different point of view. If
we take the lines as they stand in the MSS’,

ola yAOooa pionThns KUVOS
1229 — AéEaca Kaxteivaca padpovous Siknv
arns NaOpaiov TevEeTar Kaxn TVYN

and consider ab integro how we are to find there the meaning
which Madvig justly expects, we ought surely to make our first
essay upon the assumption that 1230 forms a sentence gram-
matically continuous. A malicious dog, he wishes Alschylus

1 Prof. Kennedy (see recently pub-
lished Transactions of the Cambridge
Philological Society, p. 172) has given
a guarded approval to the reading
gpadpov ots. He has stated most of
the objections to it with a vigour
which leaves nothing to be desired, but
he regards as ‘‘all but decisive” on
the other side the occurrence of ¢at-
Spots woly in a parody of tragedy by
Aristophanes Pax 154-6. I cannot but

’ think that Prof. Kennedy greatly over-

estimates the weight of this fact. It
certainly proves (if proof be required)
that to prick a joyful ear (padpor obs)

is an expression proper to the poetical
style. It might even support the con-
jecture that the words actually oc-
curred in some tragedy, though it does
not prove even as much as that, for it
is not to be assumed that Aristophanes
could not invent a quasi-tragic phrase
without an actual model. For con-
necting the parody with this passage
of the Agamemnon there seems to be
no reason whatever,

2 The difference between kal xrel-
vaca and Canter’s kdxrelvaca is not
worth notice.

EN ee

302 THE JOURNAL OF PHILOLOGY.

to say, by licking the hand under the pretence of affection will
obtain the opportunity to wound it. But this is just what
1230 does say, whether we read

drns raOpalov revEeTar Kaxh TUX,
will accomplish by an evil chance of treacherous hurt, or
drns AaOpalov revEcTar Kaxnv TUYNY,

will find an evil chance of treacherous hurt’. If this were
so, the participial clause must of course begin and end
with 1229. And whatever difficulties may ultimately await
us in bringing the whole into conformity with this hypothesis,
if once it occurs as possible, I think we shall soon discover
small but conspiring indications in its favour. In the first place,
not only is it easy to take drys Aapalov with tvyp, but it is
difficult to explain it satisfactorily in any other way. According
to the usual punctuation, déknv arns AaOpaiou, the actions of a
treacherous dog are illustrated by those of "Ary. If our atten-
tion had not been turned elsewhere, it would probably have
been noticed before’, how oddly the relation of copy and original
are thus inverted. When a quality or immaterial thing such -
as dtn is personified, actions may of course be attributed to
it, and the most natural way of making the conception real and
vivid is to compare the action supposed to that of some material
agent which may serve as a type of it. But to reverse the
process is unreasonable, not to say silly. Our only way of
imagining what a treacherous a@rn would do is to figure to
ourselves what a treacherous human being or treacherous
animal would do. What purpose, then, can be served by saying

1 The case of ofa appears to me, as
it has appeared to Prof. Kennedy and
others, quite defensible. It is rather
cognate than object to revéera:, what
success it shall have; olwv revéerau
would mean what will befall it, a
different thing. Moreover, even a
slightly irregular accusative would be,
as Prof. Kennedy says, not surprising
in such a position. It has also been

suggested (Prof. Paley) that revéerar
is the future not of tvyxdvw but of
Trevxw and is equivalent to owmoerat.
Although we should expect revée:, the
middle is not impossible; but it is —
difficult to dissociate revéerac from
TUX}:

2 Prof. Kennedy does observe in
passing how “strange” is the ‘‘pa-
renthetic simile.”

MISHTH KYON. 303

that the behaviour of a dog or a woman is like that of "Arn or
an arn? Let us consider a parallel case in English. When
the Elizabethan poet writes how
pity, like a naked new-born babe,

Striding the blast, or heaven’s cherubim, horsed

Upon the sightless couriers of the air,

Shall blow the horrid deed in every eye,
the figure, though bold or even (some might think) exaggerated,
has a plain function, giving to the abstract notion of pity the
energy and force of a visible presentation. But when the
poetaster in ‘The Chough and Crow,’ encouraged possibly by
this very passage of Shakspeare, tells us that

The hushed wind wails with feeble moan,

Like infant Charity,

the thing in spite of the most pathetic and quavering prolonga-
tions remains hopelessly ineffective; obviously because the
moaning of infant Charity, a purely imaginary sound, presents
itself to the mind so much less distinctly than the wail of the
wind itself, that the image loses in force by the comparison.
And really the “licking” and “ ear-stretching” of a “treacher-
ous curse” (or “secret plague,” or whatever you please) is very
little better. Moreover, if such a comparison was to be made,
the very last word to express it would be dé«nv. This curious
archaism, which signified properly ‘after the wont of so-and-so,’
is elsewhere reserved to the most picturesque similes only, and
was clearly in Aischylus’ day a highly artificial phrase, the
unusual application of which would have been instantly felt
and reprehended. In modern imitations of the Attic dramatists
this déenv is treated as an arbitrary variety for @o7ep, but this
is not so in the originals. Sophokles and Euripides have but
one example between them, which we shall notice more particu-
larly below; Auschylus, for reasons not difficult to see, is
extremely fond of the phrase; but the reader will perhaps be a
little surprised at the following list of his comparisons—a crow,
an ov, a swan, a swallow, a dog, a hare, a fawn, a kid, a wave,
fire, water, the moon, the Gorgons, a diver, a messenger, a
chartoteer, a gardener, sailors, a child, a bride, a man speaking
a foreign tongue, dreams. The limits of this class. are visible

304 THE JOURNAL OF PHILOLOGY.

upon inspection. All these things are things which | ave a

motion or habit of some kind, a ‘way’, in fact, of their own,
which way is called their Sécn. The case of Ag. 980, whatever
view may be taken of it, is peculiar,

pavtione 8 axédevoTos aicbos doda’

ovd amomticat Sixav

Suvoxpitay overpatov

Odpaos evmibes ier ppevds pirov Opdovov.
Neither the object nor the manner of this comparison are like
the supposed Sinnyv drys AaGpaiov, and it does not therefore

concern us here, but Sccay dverpatwrv is certainly unusual, and

there is something to be said for the editors who reject it (see
Karsten, Dindorf 1869, and others). Beyond Auschylus, there
is I believe but one tragic example, voXeuiwv Sixnv in Eur.
Hek. 1162, to which for its own sake I should like to devote a
word or two, Fresh from the vigorous similes of Auschylus, his

Boos Sixnyv mpos Bwopov evTrorApews trates,
or his .

dverdos étuiev Sixav Sippnrdtov

vTO Ppevos,

the mind can scarcely fail to be displeased by the shadowy |

vagueness of zroXeulwy Sixnv, and to wonder why Euripides
should have gone out of his way to pick up so poor a phrase.
At least this was my own impression, and it was certainly not
weakened when on turning to the passage, I read—Polymnestor
is describing the assault of the Trojan women—

_ Kat’ é« yarnvdv mas Sones tpocdbeypatwv
evO’s AaBotoa dacyav é« mémrwv Tobev
KevTovaL Tratdas, ai dé ToNewioy Sikny
Evvapracaca. tas éuds elyov yépas
Kal Koda.

We scarcely need the poet to tell us that. to hold a man
down by main force while your comrades murder his children is
the act of an enemy, and if Euripides did wish to state the fact,
the commonest vocabulary would have been adequate to his
purpose. The adoption of Sécnv should signify some unusually

A Se padkadidie heat mer, snein AO0) Kook a

MISHTH KYON, 305

bold and /éschylean image, some graphic touch from nature
such as the father of tragedy delights in. Perhaps we can even
recover 1t— ‘
ai Sé moAuTrédwv' Slenv

Evvaprdcacat Tas éuds elyov xépas

Kal KoAa—
they clung to my limbs like devil-fish.

With any reading, therefore, the common punctuation of

NéEaca Kakteivaca ghaidpovous, Sixnv
atns NaOpaiouv

is objectionable in point of sense; and this objection applies
to Madvig’s. reading no less. But the correction gaidpdv ods
raises another objection on the ground of rhythm. I appeal
with some confidence to any one who has tuned his ear by the
sound of Aischylean verse, to say whether he is pleased with

gdadpov ots, Sixnv

for the close of an iambic senarius. A pause, even the slightest,
before the final foot is contrary to the principle of the metre
and extremely rare. It occurs indeed with this very word

dtenv in Ag. 297

vrepOopodca médiov "Acwrod, Sixnv
daidpds cedHvys,

but it is plain—I speak of my own sensations—that the ill
effect of the break is very much increased when it follows
immediately upon a heavy and emphatic monosyllable. I do
not of course mean to say that any line could be suspected
merely on this ground, but when we are endeavouring to fix the
exact place of an admitted corruption, we should best begin by
presuming the rhythm normal; and this, corroborating our
other evidence, will incline us to place the division before drns5
AaPpaiov and seek a construction for Sixnv in the line where it
stands.

This, then, is the new point of view of which I spoke. It

1 Or rovAvrwy if this form of the word is to be restored in tragedy. See
Lexicon 8. v.

306 THE JOURNAL OF PHILOLOGY.

remains to take the corrupt line 1229 and see whether, on this
hypothesis, we can make anything of it. | 3 3 7
réEaca (or Ael~aca) Kaxtelvaca faidpovovs SiKny.

Now there is only one word in this clause which does not offer :
hope of a, reasonable sense, and that is dé«nv. Mr Sidgwick {
indeed remarks that ‘dasdpdvovs is a very strange adjective, :
but the objection, unless it be confined to the construction of |
the word in this particular context, seems groundless. The
transference of daidpos from the glad expression to the glad
feeling which the expression is supposed to indicate is quite
natural; we have parallel forms in dynrdvovs (Plato, Phedr.
270 A), vypdvous of weak virtue (mentioned by Pollux 6. 126 in
a list of epithets applicable to the «/vatdos), and probably others.
It is likely that such formations were unusual in Attic prose,
impossible they can scarcely have been in a dialect to which,
evvovs and xaxovovs were familiar; and it might even be
argued from the context that neither vyydovous nor its com-
panion TeAeovovpyos are of Plato’s own mint; but for poetry
dadpovous is perfectly good. But. dé«nv of course cannot be
right if the line is continuous. And to a reader accustomed to
the habits of copyists, no word could appear more promising as
a lurking-place for corruption. It is an extremely common
word and has a wide range of meanings, with a corresponding
capacity for appearing to give a sense when it really does not.
In fact it is just the sort of word which the half-learned scribe
is apt to fabricate. The next thing, then, will be to consider
what letters were likely to be mistaken for AiKHNn, a question
admitting of a brief and positive answer. AikHn in Alschylus may
represent four different groups of letters, of which two have no
meaning, or none which can apply to the present passage. But
let us try Avynv—

by 3 4 A ie \
ovK oidev ola yACooa pioNTHS KUvOs,
NeiEaca Kaxtevaca pardpdvovy dy»,
b) / A
atns AaOpaiov tev&eras KaKh -TUXN,

he knows not what the tongue of the abominable hound, proffering
the lick of gladness, shall accomplish by an evil chance of treach-
erous hurt. The accusative \yyv is to be taken both with AedEaca

MISHTH KYON. 307

and with éxre(vaca. Thus the language of the metaphor be-
comes perfectly natural and consistent: the tongue does not
bite, but by offering to lick the hand in token of friendly wel-
come, it gains an opportunity for the bite. é«rteivacais modelled
upon éxtelveww yetpa to put out or proffer the hand by way of
greeting, and it is this word which, if my suggestion is right,
was probably the cause of all the mischief. Kassandra’s words
are pointed more especially to the long and elaborate speech
with which Klytemnestra receives Agamemnon upon his arrival
855—913, of which the king says (915)

> / > Pe F > A,
aTovala pév eltras eiKoTws éun
paxpav yap é&éreuwas.

It was therefore very natural to connect the éxre/vaca of
one passage with this é£érewas, though in reality it has not any
resemblance, or, except perhaps in a sort of allusive way, any
reference to it. Thus é«re/vaca in fact produced the false sub-
stitution of Aé~aca for Ne/Eaca which it has served to conceal.
The experience of any student of textual degeneration will
supply him with examples of this species of error by false refer-
ence, as it might be called. ‘Thus aided and prepared, the
descent of AlyHN to AIKHN, already sure, would be precipitated,
and dadpévovy must wander, as forsaken adjectives will and do,
to the only remaining support.

I should like to add a few words on the reading xaxnp...
Tvxnv. Paleographically, it is almost an indifferent alternative
for xaxj...tvyn, the confusion of these terminations being in-
cessant. And though, as the dative will pass, we should of
course not change it, 1 am not sure that the accusative is not
better, and possibly right. The adverbial ofa how, for the common
ws, is illustrated by the analogy of é7rota and dota, both of them
Attic; ofa itself occurs as an adverb in Homer. and there is no
reason for denying it to Aischylus. Upon rev£erar tvynp it is
to be observed that tvy7 in its common use is ‘cognate’ to the in-
transitive tuyyavw to happen, befall, rather than to the transi-
tive to find or get, though the two meanings are very near and
sometimes cross. But it is a characteristic of poetical styles,
particularly of archaic poetry, to use words according to their

308 THE JOURNAL OF PHILOLOGY.

etymological and, so to speak, native force, and not according to

the narrower limitations which for the sake of clearness are

imposed upon them in prose. It is one of the many ways of
attaining that remoteness from vulgar associations, which is in
some degree necessary to diction intended for dignified purposes.
The Greek tragedians themselves furnish examples much more
remarkable than this exceptional treatment of tvyy. For
instance, AaB7n in the Greek of all periods signifies either a
handle or a grip and in the latter sense is a technical term of
the wrestling-ground. But notwithstanding this, Aischylus can
fall back upon its etymological relation to XauPavo, like that
of r¥yn to Tvyxave, and boldly writes (Supp. 935)
7d veixos 8 ov év dpytpov rAaBA
éXucen,

making AaB a poetical equivalent for the prosaic AAs. So
again the various uses of mdaAXevv are for the purposes of ordi-
nary speech strictly parted among the different substantives
derived from the same stem; to md)Xew KAHpous corresponds
adros a lot, from wadAeras Kéap we have rads a throb,
while wadrew Tov paxydpevoy to swing or dash down an adver-
sary, a use which, though it must have been once in vogue, had
before the literary epoch been expelled almost to the last trace
from the verb, took by way of compensation the exclusive
possession of the substantive wadn wrestling. Neither warn
nor its derivative mwadadew have any other sense in prose, or
normally in poetry either. Yet the language remains conscious,
so to speak, that mad is, after all, merely waXos in an older
form and can upon occasion remember it, as we see when Euri-
pides writes (Herakl. 158)

ny 8 és Aoyous Te Kal Ta TOVS’ oikTicpata

Breas memavOns, és wadnv Kcablorarat

Sopos TO mpaypa.

It is obvious that this wad Sopés (for the common pdyn
dopos) is intended to recall the cognate verb in its most familiar
application 7aew dSopv, which does not appear to have formed
any substantive in common use, nor are 7dAos and its associa-
tions of arbitrament and decision out of view. This blurring of

eR Se ne ee ee

MISHTH KYON. 309

the hard lines drawn for practical purposes between kindred
forms of speech is, in fact, the essential condition of poetry and
its chief linguistic difference from prose. One more example
out of many I will mention because it is peculiarly striking.
KoTros ache is connected in form with «omrew to bruise, but the
etymology has left no impression upon its ordinary meaning.
The word itself is almost confined to poetry, but its congeners
classical and post-classical xomiaw, xoTiapds, koTalw, koTréw etc.
have all the same notion as xézos itself, fatigue or weariness.
To strike a coin was Kore voutopa, the stamp not Kormos but
xoupa, to beat the breasts in mourning was xérrew orépva, the
act not «xomos but xoppos. It is quite improbable that «dzos
was ever heard among Athenians in any but the one proper
sense. Yet it was observed by Seidler and is generally agreed
that in Eur. Suppl. 789
T&0E TOL OlOomeV
TANYLATA KpaTos oTépywov Te }KTVTOUsST

the poet must have trusted the ears and intellects of his
audience so far as to write

\
TAHYLATA KpAaTOS oTépvwv TE KOTOUS.

It is a bold stroke, though, and Athens one would think
must have contained critics dull enough or keen enough to ask
what ‘breast-aches’ might be. It is possible also that the true
correction is

TANYWATA Kpatos orépvwv TE TUTrOUS,

but if so the example is equally to the purpose, for tv7ros, like
xo7ros, was differentiated for prose purposes by the different
sense of a mark or stamp. Xenophon has it once as an equiva-
lent (unless indeed it is an error) for xrvaros (Eq. 11,12). I
prefer xczrovs, however, in the Supplices, because correc Oat was
the regular Attic word for the gesture of mourning, not tvz-
tea@at, and an exceptional use of this kind would naturally
follow some perfectly familiar analogy.

In the presence of these facts, and many more of the same
kind, there is no reason for surprise at any use of tU’yn which is
justified by its relation to tuyydvw. There are, as I have

310 THE JOURNAL OF PHILOLOGY.

pointed out elsewhere (note to Eur. Med. 198), other passages:
in which téyn in the common sense of chance would give a poor
effect, and the word is certainly coloured by the associations of
ruyxave to hit, so as to suggest, if not to mean, a stroke, notably
Eur. H. F. 1393 7
mavtes e€oXwodapev
"Hpas mid wrnyévtes GOAvoL TVYD.

In the Aschylean example, the actual presence of the verb in a
grammatical connexion implying similarity of meaning would
make the substantive perfectly clear, and the whole would re-
semble more closely than ever the aiveis...caxdv aivov from
1481—2 of the same play, which Prof. Kennedy quotes in
illustration of it.

If the reader, satisfied in other respects with my interpreta-
tion, is disturbed by the absence of Avy from the Lexicon, I
would suggest to him the following reflexions. (1) Isit seriously
to be supposed that the Greek language was incapable of ex-
pressing a lick? (2) If the word for a lick, was not Avyy or
Aovyy, one or both (cf. wAoxy, TUYN, mTvyn, oTOLBH, oTLBN,
0B, oTixos, oTlyes, etc. etc.)—what was it ?

A. W. VERRALL.

Notre.—Having had occasion to refer to Mr Sidgwick’s edition,
I should like to express my thanks for his courteous and appreciative
remarks (Appendix) upon a former paper of mine in the 9th volume
of this Journal. I think his criticisms partly right and partly wrong,
and hope soon to have an opportunity of discussing them further,

are are. ~- Le

ARISTOTLE, POLITICS IV (VII) 13 §§ 5—7. 1332 a 7 sqq.

Papéev Sé kal Siwpicpeba ev trois HOKols, ei TL THY AOYor
exelvov dpedos, evépyerav elvat Kal ypiow apetis Tedelar,
Kal TavTnv ovK €& Umolécews, GNX aTAds. Aéywo 8 &F Urr0-
Gécews tdvayaia, To 8 dards TO Kadds. olov Ta mepl Tas
Ouxaias mpakes ai Sikarac tiywplar Kal Koddoes am dperhs

/ > > a \ =
MeV eiow, Kal avayKaias Sé Kal TO Kad dvayKalas exoucL'

aiperatepov pev yap pndevos SeicOat tev ToLovT@y prjte Tov
avdpa ponte tiv rod ai 8 él Tas Tyds Kal Tas evTroplas
amas clot KaddcTaL Tpdkes. TO pev yap Erepov KaKod
Twos aipeots éotiv, ai. toadTar S& mpdkews Tovvavtiov, KaTa-
oxeval yap dyabdv eict Kai yevrjoets.

Happiness, we are here told, consists in évépyeva al ypjous
dpeths Terela; but in order that évépyera nal ypnous aperis
Tehela may constitute happiness, it must be teAela daAds, not
tehela €& vrroécews. By way of explaining the distinction,
it is added that rdvayxaia are é& vmobécews Tédeva, TO KANGS
being amA@s tédXevov. For example, ai Sixavas tiymwpiat Kat
koracets, though virtuous, are avayxaiat, and exhibit Td Kadds
avaykaiws ; while ai Sixatat mpdées ai él tas Tyuds Kal Tas
evTropias are amas KaddoTat. In the sequel something is
said about ypjoess which are amA@s orrovdaias Kal Kadai.

As the phrases é& taroécews, dds confessedly need ex-
planation, while the words dvayxaia, cadds are ambiguous,
we naturally look to the example in the hope that it will give
us some assistance. Here however a new difficulty meets us.
If we take the phrase ai éml rds tTids Kal Tas evtropias in
its obvious sense, it seems strange that ‘acts done with a view
to, or in the expectation of, honours and rewards,’ should be
contrasted with ‘the infliction of just vengeance and punish-
ment.’ If again with Susemihl we take the phrase in question
to mean “diejenige Ausiibung der Gerechtigkeit, welche
Anderen Ehrenauszeichnungen zutheilt oder Wohlstand ver-
schafft,” it seems strange that acts of distributive justice should

oe ee FOL nk ga one
‘ cr ie: 3 ? Se

312 THE JOURNAL OF PHILOLOGY. .

be accounted so decidedly superior to acts of corrective justice,
while it may be doubted whether the phrase éml Tas TYpas
kal tas evropias can bear the meaning given to it. It is
to be remarked further that in any case the epithet Simasas is —
superfluous, as tad sept Tas Sedans apaéers are alone under
consideration.

Now in Mic. Eth. ut 8 §§ 1—5 and x 9 §§ 4, 9, 10, acts
done 8 dvayny, Sia Tas Tywoplas, are unfavourably contrasted
with acts done 67 xaddv, Sud tas Toads: see especially III 8
§ 5 Sef & ov Sv dvdyxny avdpeiov elvat add’ bTt Kadov. X Y
§ 4 od yap rebixacwv [sc. of rodnoi] aidot reBapyely adda
doBw, ovd aréyecOat tév Gatrov Sid 7d aioypdov adda bia
Tas tiywplas. § 9 of yap ToAXol dvaynn paddov ) OYE
mevOapyovor Kal Enuiats 7) TS KaA@. May we not infer that
in the passage before us, where ravayxaia are contrasted with
Ta KaNaS, al él Tas Tinas Kal Tas edTroplas mpateas being the
second member of the second pair of correlatives, (1) tavayxata
means, not ov ove dvev TO ev (Berlin Index 797 a 43), but
ta 80 avayeny, and (2) ai 8a ras Tyuwplas Kal KoNdoess should
be substituted for ai Skat. Tiywpia Kal KoAadeers ?

In this way we obtain an excellent sense: just acts which
have for their motive avay«n in the shape of vengeance and
punishment are distinguished from just acts which have for
their motive 7rd «xaddv in the shape of honour and reward,
the former being accounted 飀 vobécews tédevat, perfect in a
qualified way, the latter daA@ds rédXevat, perfect without such
limitation: it is in acts, not of the former, but of the latter
sort—in other words, it is in acts not of obligatory, but of
optional, morality—that happiness is to be found’.

HENRY JACKSON.

? Accepting Postgate’s interpreta- virtuous action enforced by punish-
tion of the phrase xaxov twos alpecis ment is good only by comparison with
(Notes on the Politics of Aristotle, p. vicious action, virtuous acts done in
13), I read the sentence in which it the hope of honour and reward pro-
occurs as a justification of the supe- duce positive good, and so are good
riority assigned to ai érl rds tiuds kal absolutely.’

Tas evmopias mpdtes: ‘for, whereas

CAMBRIDGE: PRINTED BY C. J, CLAY, M.A. AT THE UNIVERSITY PRESS.

1 We

ae

1882.

Vol. 10,

p>
&)
Oo
ol
o
rei
a,
GS
Dy
M4
©)
‘3
S
s
2)
Kr

UNIVERSITY OF TORONTO
LIBRARY

Do not
remove
the card
from this \

Pocket. RQ

mene

Acme Library Card Pocket
Under Pat. ‘f Ref. Index File.”?
Made by LIBRARY BUREAU

St wt. on es os oe
LEO RL oe ei OEE RES PS I IS Le waters
1. AEE Let lnk et SAC pe A ee ty lr a bras eae

ind

atest:

ish ananinD ere amine adajinndo Gade Rianne