Skip to main content

Full text of "University of California publications in classical philology"

See other formats


UNIVERSITY OF CALIFORNIA .-PUBLICATIONS 

IN 



CLASSICAL PHILOLOGY 



VOLUME XII ^3 

1933-1944- ^"0 



EDITORS 

W. H. ALEXANDER W. M. GREEN 

J. T. ALLEN L M. LINFORTH 

G. M. CALHOUN H. C. NUTTING 

H. R. W. SMITH 




UNIVERSITY OF CALIFORNIA PRESS 
BERKELEY AND LOS ANGELES 

1944 



Universitt of California Press 

Berkeley and Los Angeles 

California 

Cajibridge University Press 
London, England 






PRINTED IN THE UNITED STATES OF AMERICA 



CONTENTS 

NO. P^GE 

1. Homeric Eepetitions 1 

George M. Calhoun 

2. Aristophanes and the Pnyx 27 

James Turnet Allen 

3. On the Program of the City Dionysia during the Peloponnesian War 35 

James Ttjrnet Allen 

4. Omitted Speech Formulas in Homer 43 

Frederick M. Combellack 

5. Seneca's Epistulae Morales: The Text Emended and Explained (I-LXV) . . 57 

William Hardy Alexander 

6. Manes exite paterni 89 

H. J. EosE 

7. Physiologus Latinus Versio Y 95 

Francis J. Carmody 
S.Seneca's Epistulae Morales: The Text Emended and Explained 

(LXVI-XCII) 135 

William Hardy Alexander 
9. Notes on Some Didymaean Inscriptions 165 

Joseph E. Fontenrose 

10. Seneca's Epistulae Morales : The Text Emended and Explained 
(XCIII-CXXIV) 175 

William Hardy Alexander 

11. Varia Critica 217 

Joseph E. Fontenrose 

12. Seneca's Dialogi III, IV, V Be Ira Lilri Tres: The Text Emended and 
Explained 225 

William Hardy Alexande^i 

13. A Dravidian Etymology of the Sanskrit Proper Name Nala 255 

M. B. Emeneau 

14. High Comedy in the Odyssey 263 

Walter Morris Hart 

15. The Biographical Fashion in Literary Criticism 279 

Harold Cherniss 

16. The Immigrant's Bath 293 

Hermann Frankel 

17. Soul and Sieve in Plato's Gorgias 295 

Ivan M. Linforth 

18. Augustine on the Teaching of History 315 

William M. Green 

19. The Sinduvdra Tree in Sanskrit Literature 333 

M. B. Emeneau 
Index 347 



^ 



A^ 



I r^<S 



.* 7 ^S*-'-^-' 



V 



THIS VOLUME IS DEDICATED 
TO 

JAMES TURNEY ALLEN 



HOMERIC REPETITIONS 

BY 

GEORGE M. CALHOUN 



For more than two thousand years Homeric critics have been study- 
ing, analyzing, evaluating, by a purely visual process, from manu- 
scripts or printed books, poems composed for oral recitation. I 
propose as one test of this criticism to examine the way in which 
speciaHsts, from the Alexandrians on, have dealt with the repetitions. 

When the statistics of repetition are considered, the Ihad and the 
Odyssey seem to be a mere patchwork of hackneyed verses and 
poetic tags. Lines that may fairly be called repetitions make up 
approximately one-third of our text, and the remainder contains 
innumerable groups of words which the poet uses over and over 
again at his good pleasure.^ In the first twenty-five lines of the Iliad 
Parry has found twenty-nine expressions which are repeated else- 
where, usually quite a number of times; in the corresponding lines 
of the Odyssey he has found thirty-four.^ If these be stricken out, 
there is left, literally, almost nothing. Passages which even the 
higher critics admire are found upon analysis to be made up largely 
of repetitions. What shall we say of such art as this? 

We may begin by noting that it was accepted with no misgivings, 
and subjected to no destructive analysis, so far as we know, down 
to the close of the classical period. 'Long' editions, containing 
many more repeated Hnes than our vulgate, seem to have been 
plentiful and as highly esteemed as the shorter texts.^ The Alexan- 

1 The figures usually accepted are those of C. E. Schmidt, Parallel-Homer 
(Gottingen, 1885), p. viii. 

2 Milman Parry, "Studies in the Epic Technique of Oral Verse-making. 
I. Homer and Homeric Style." Harv. Stud. CI. Ph., XLI (1930), 117 ff. 
Schmidt, loc. cit., estimates that the repetitions of not fewer than six morae 
will total about 16,000 lines. 

' The quotations that point to texts longer or shorter than the vulgate are 
presented by T. W. Allen, Homer. The Origins and the Transmission (Oxford, 
1924), pp. 249 ff. The evidence is not sufficient to establish the proportion of 
long to medium or short texts, but shows that the long texts were apparently 
not regarded with suspicion or disfavor. See also P. Cauer, Grundfragen der 
Homerkritik (ed. 3; Leipzig, 1923), pp. 42 f.; Gilbert Murray, The Rise of the 
Greek Epic (ed. 3; Oxford, 1924), pp. 289 ff. On the early 'long' papyri, T. W. 
Allen, Homeri Bias (Ox-ford, 1931), I, pp. 89 ff.; Paul Collart, "Les papyrus 
de rniade," Part I, Rev. de Phil., ser. 3«, VI (1932), pp. 315 ff., especially 
pp. 338 ff.; Part II, ibid. VII (1933), pp. 33-61; cf. infra, pp. 23f. 

[11 



2 University of California Publications in Classical Philology [Vol. 12 

drian scholars of the third and second centuries B.C., however, 
questioned the authenticity of many repeated Unes and passages. 
The criteria were in large part subjective; the critics were in general 
adverse to repetition, and a considerable part of their work was to 
stigmatize repeated lines as spurious, except in the particular places 
where they supposed them to have been originally used.^ It is sig- 
nificant, in my opinion, that this tendency first appears at a time 
when the practice of publicly reciting Homer is disappearing and 
manuscripts are becoming abundant." For it is fairly clear that the 
work of the Alexandrians was performed visually with written texts 
- — a method of Homeric study that has prevailed ever since. 

Modern criticism, in its attitude toward the repetitions, has much 
in common with that of the Alexandrians; it rests largely on sub- 
jective criteria, it is worked out by a visual process, with printed 
texts in place of manuscripts, its tendency is distinctly adverse to 
repetition. In the two last respects, it goes farther than the Alexan- 
drians. Visual study of the text has been supplemented by the use of 
word-indexes, concordances, and tables of parallel passages, with the 
result that much modern work upon Homer stands at two removes 
from the conditions under which the poems were apprehended prior 
to the Alexandrian age. Lastly, the antipathy toward repetition has 
been intensified by the attempts to analyze and dissect the poems in 
the interest of this or that theory of their origin, since analysts have 
assumed that repetition is a criterion by which the spurious may be 
distinguished from the authentic. This assumption completely domi- 
nated Homeric criticism toward the end of the last century, and still 
has its devotees despite the patent absurdity of the actual results. 

Rothe gave essentially the correct solution when he characterized 
the repetitions as "gemeinsames Versgut."" Lang takes the same 

^ The trenchant criticism of Zenodotus and Aristarchus is api)reciatively 
described hv Murray, pp. 283 ff. See also Allen, Origins, pp. 304 ff., Ilias, 1, 
pp. 196 ff.; E. Dreriip, Homerische Poelik (Wiirzburg, 1921), I, pp. 82 flf.; 
Cauer, pp. 51 ff. (where references to the more detailed studies will be found). 
The Alexandrians apparently based their work on the medium texts which 
closely resembled our vulgate (Allen, Origins, p. 310); cf. G. M. Boiling, The 
External Evidence for Interpolation in Homer (Oxford, 1925), Part I. 

^ Cf . Allen, Origins, pp. 325 f . Without as yet accepting Allen's opinion that 
the plus verses are additions l\v rhapsodes, we may observe the significant 
fact that longer versions were in existence during the period of recitation but 
disappeared with increasing use of manuscripts. This accords with my sug- 
gestion that dislike of repetition is a product of visual study of the poems. 

^ C. Rothe, Die Bedeutung der Wicderholungen fur die homerische Frage 
(Berlin, 1890), p. 154. "Wir miissen also annehmen, dass sich der Sanger wie 



1933] Calhoun: Homeric Repetitions 3 

view upon independent grounds/ and the special studies of Scott 
and Shewan reveal the futility of the attempt to dissect the poems 
by comparing repetitions.^ Although those eminent critics who 
think they are able to identify various strata which have overlaid 
'original' poems are appalled by what they regard as sheer nihilism, » 
I am unable to find that any of them has brought a serious argu- 
ment against Rothe's fundamental assumption. They admit the 
general principle he has laid down, acknowledge the need of cir- 
cumspection in dealing with repetitions, and then proceed with 
their attempts to discover the 'original' instances.^" They have 
been able to maintain this attitude because Rothe and others who 
have adopted his view have emphasized the negative arguments 
relating to particular passages and neglected to develop positively 
the potentially fruitful concept of gemeinsames Versgut. Had they 
established more definitely the nature and extent of this Versgut, 
they could have dismissed as instances of epic formula many repe- 
titions which they attempt to trace to the 'original' or at least to 
the better context. Parry's studies of the briefer formulary elements 
in Homer have now given us more definite knowledge of the vast 
store of epic formula upon which the poet drew constantly," and it 
becomes possible to deal with the problem in a more positive way. I 
shall attempt to show that repetitions of a line or more, like the 

das Sprachgut, so auch das Versgut angeeignet hat so mussen wir uns 

jeden Sanger im Besitz eines reichen Schatzes von Versen denken" (p. 158). 
"Unter keinen Umstanden aber lasst sich bei Wiederholungen aus der 
grosseren oder geringeren Geschicklichkeit oder Angemessenheit, mit der 
dieser oder jene Zug eingefiihrt oder begriindet ist, ohne Weiteres auf 
'Echtheit' oder 'Unechtheit,' auf Urspriinglichkeit oder Nachahmung schhes- 
sen" (p. 166). 
•' A. Lang, The World of Homer (London, 1910), pp. 289 ff. 

8 The bewildering results of these attempts, as they relate to theDoloneia, 
are presented by A. Shewan, The Lay of Dolon (London, 1911), pp. 115 ff., 
where references are given to earlier studies; on the general problem, cf . J. A. 
Scott, "Repeated Verses in Homer," A. J. P., XXXII (1911), pp. 313 ff.; 
The Unity of Homer, Berkeley, 1921, p. 264; Drerup, pp. 368 ff. See also 
Shewan, "Repetition in Homer and Tennyson," CI. W., XVI (1923), pp. 
153-158, 162-166. 

9 Cauer, p. 612; cf. infra, p. 22. 

1° Drerup, pp. 369 ff., gives references to the principal studies. P. Chan- 
traine, "Remarques sur I'emploi des formules dans le premier chant de 
riliade," R. E. G., XLV (1932) pp. 121-154, though he infers from Parry's 
results that many repeated lines are formulas, thinks it is possible in other 
cases to discover 'original' contexts; his discussion is based on the theory 
that Book I is primitive and ancient. 

" Milman Parry, L'epith'ete traditionnelle dans Homere (Paris, 1928), espe- 
cially pp. 8 ff.; Epic Technique, especially pp. 117 ff. 



4 University of California Publications in Classical Philology [Vol. 12 

briefer repetitions out of which they develop, are simply so many 
bits of an inexhaustible store of traditional material, used just as 
freely as, for example, the name-epithet formula. I shall maintain 
definitely that the burden of proof rests upon those who hold that a 
repeated passage is not a formula,^^ and that the cases in which we 
may hope to find the 'original' instance of a repetition must be 
relatively few. Nor shall I hesitate to discard conclusions, no matter 
how illustrious the author, when they involve the fundamentally 
wrong assumption that identical fines or passages must be 'copied' 
from one another.^^ 

In considering these problems, we must keep constantly in mind 
two points that critics are prone to forget, first, that the poems were 
orally composed for oral recitation, ^^ and second, that the audiences 
were thoroughly famifiar with what they were going to hear. It is 
possible for anyone to demonstrate by actual experiment that repe- 
titions produce very different effects when listened to in context and 
when hunted out with a concordance, put side bj' side, and con- 
sidered visually. 

Bowra's recent study^^ is a distinct advance in that it makes no 
search after 'original' instances. It treats repetition as an integral 
part of the technique of oral composition, and attempts to explain 
this technique from considerations which arise when a poem is re- 
cited or chanted and Hstened to by an audience. Thus at last we get 
away from a purely visual analysis. ^^ Although Bowra has not freed 
himself from the feeling that repetition is a blemish, ^^ he does treat 

12 On certain types of repetitions which are not formulas, of. infra, pp. 18 ff. 

1' This is an instinctive feeling of the modern critic, induced hv circum- 
stances that did not exist when the poems were composed. Cf . infra, p. 22. 

1^ Allen, Ilias, I, p. 194, is convinced that Homer 'wrote.' This is of course 
possible, but Parry has shown clearly, in my opinion, that the Homeric style 
can onlv be explained bv assuming oral composition {Epic Technique, pp. 138 
ff.; cf. ""Enjambement 'in Homeric Verse," TAPA, LX (1929), pp. 214 f.). 
Homer may have 'written,' but he 'composed' orally, as had his predecessors 
for generations. Cf. F. Jacoby, "Homerisches," Hermes, LXVIII (1933), 
p. 3, n. 1: "die Hias istfiir Horer geschrieben." Modern scholars often do not 
realize that even in the classic period, when manuscripts were fairly numer- 
ous and authors 'wrote,' compositions were still intended primarily for oral 
publication, and were written merely for convenience. Reading did not sup- 
plant listening until later, and no doubt very gradually. 

15 C. M. Bowra, Trndiiion and Design in the Iliad (Oxford, 1930). pp. 87 fT. 

i« The trend of Parry's studies, which Bowra apparently has not seen, is 
away from the purely visual analysis. 

" E.g., p. 95: "an age when poetry was recited could endure repetition more 
easily than we can .... anything in the nature of repetition was less likely 
to be noticed " 



1933] Calhoun: Homeric Be petitions 5 

the repetitions as functional. He believes that recurring lines and 
epithets rest the mind and prepare an atmosphere; recurring pas- 
sages "give one emotional colour here and another there," and 
through reminiscence bring out contrasts; recurring themes give 
"an old story new life in new and different forms. "^* 

With much of this I agree, but it is far from adequate and it in- 
volves several assumptions which I regard as unsound. We may 
doubt whether repetitions of single lines were really effective in 
resting the mind of the audience ; the longer repetitions might have 
done so, but for them Bowra finds another, different function. 
Granting that repetition was sometimes used to point a contrast, ^^ 
this explains only a small part of the instances and cannot be re- 
garded as satisfactorily establishing the technique. In the last 
analysis it rests upon our habitual feeling that identical passages 
are 'copied' from one another. Moreover, since Bowra allows that 
the poems could not often have been recited in their entirety,^'' 
there was always the equal chance that an individual auditor would 
hear the repeated passages in exactly the reverse order from that on 
which the poet, were he aiming at a definite contrast, depended for 
his effect. Finally, Bowra's treatment is involved with the erroneous 
assumption that the audience was not very familiar with what it 
was going to hear.^^ 

One conclusion which is reasonably certain is that an ancient 
audience, assembled to listen to a recitation from Homer, was thor- 
oughly familiar with what it was going to hear, and likely neither 
to lose the thread of the story, if it dozed a moment, nor to fail of 
noticing the repetitions. In Athens, in the classical period, the free 
citizen was saturated with Homer before he was out of school, and 
he heard Homer constanth' thereafter. In the heroic age it was 
apparently the principal occupation of a gentleman, when not en- 
gaged in fighting or lifting cattle, to Hsten to the bard, who sang, 

18 p. 96. 

1^ Bowra thinks (pp. 91 ff.) that the repetition Iliad xv. 263-268 = vi. 506-511 
aims deliberately at a contrast between Paris and Hector. I should prefer to 
say that the poet chose the same simile in both cases because it was his best 
illustration of exuberance of pure animal spirits, and that is what he wanted. 
Repeated similes are not typical repetitions, as Bowra assumes, but are a 
special class, requiring special study. 

20 P. 95. 

2' The hearers were likely to lose the thread of the story (p. 87), and not to 
notice repetitions (p. 95), though at other times they must be prepared to 
recognize (p. 92) in the middle of Book XV a passage repeated from Book VI. 



6 University of California Publications in Classical Philology [Vol. 12 

oftener than not, the tale of Troy.^^ If we wish to understand cor- 
rectly the conditions under which these poems were presented and 
enjoyed, we must assume that the listeners were as famiUar with 
what they were about to hear as is the audience gathered in La 
Scala to hear Trovatore or Cavalleria, or an assemblage in Munich 
or Berhn which listens for perhaps the fiftieth or one hundredth 
time to Die Meistersdnger or to the Fifth Symphony. Here we find, 
I beUeve, an analogy we may use with profit in attempting to 
understand the art of Homer, for here we have an art w^hich is still 
in modern times auditory, rather than visual, in presentation and 
apprehension. This analogy I propose to use, and I shall attempt to 
show that the Homeric repetition aimed at effects which correspond 
in considerable degree to those produced by recurrent phrases and 
figures in modern musical composition. Like all analogies, this must 
be used with caution, and not pressed too far. Since an analysis of 
many repetitions cannot be attempted in a brief paper, I shall dis- 
cuss selected instances which repeated reading of the poems has 
led me to regard as typical, and merely refer to others.^ 

We may begin by recognizing clearly that no effective distinction 
can be made between the epic formula of two words, or of three 
words or half a line, and the formula of one fine, which very- often 
is merely the combination of two or more briefer formulas.^* No 
more can we distinguish between a repetition of one line and of 
two, three, five, or ten hues. We can only say that as the formula 
grows longer it necessarily becomes more complex, with perhaps an 
increased probabiUty that its employment is conscious and deliber- 
ate, and not merely habitual or instinctive. At times a definite pur- 
pose is served even in the repetition of single lines." When we hear 

j3ij 5e 5td TrpojudxwJ' KtKopvdixkvos aWoTi xo^^kV ' 

we see a mighty warrior striding to attack a foe or save a comrade; 
the formula prepares us for the coming scene by recalling others of 

22 Allen, Origins, pp. 139 ff.; Bowra, pp. 27 ff. 

23 Cf . infra, pp. 9 flf. 

24 This is abundantly illustrated in the material collected by Parry, 
L'epithete traditionnelle, pp. 11 ff. 

26 Bowra is wrong, I believe, in thinking that such brief repetitions rest the 
mind of the hearer, but right in saying that they "prepare an atmosphere — 
I should rather put it "induce a mood" or "evoke a remmiscence. Ihe 
examples he gives are not the best that might be found. 

26 Iliad v. 562 and often elsewhere. 



1933] Calhoun: Homeric Repetitions 7 

like sort; when used of one hero it reminds us of others in a Hke 
situation. In the same way 

/3^ be Kar' OuXujuiroto Kapr]vup at^aaa ^^ 

evokes a picture, composite, but startHngly distinct, of urgent God- 
head flashing down from heaven to help or to punish. 

Any reader of Homer can supply many other repeated hnes of 
this sort. Bowra cites a case from Chaucer, where the Une 

AUone, with-outen any companye 

is used in a context of grave poetic beauty and also to describe the 
vie de gargon of the gay Nicholas.^^ It is strange that this did not at 
once recall to his mind Homer's use in widely different contexts 
of the line 

avrap 6 'iypu f]ai.v kvl tppecrl <p6}vr]akp re. 

This is an excellent example for our purpose because it illustrates 
the almost unhmited adaptability of the repeated line, and also 
shows clearly how it grows out of the briefer formula. When the 
heralds of Agamemnon came to take away Briseis, they stood 
shamed before Achilles, and said not one word to him "but he knew 
in his heart, and he spake" {Iliad i. 333). After Zeus had caught 
Hera and Athena setting out to help the Achaeans, in disobedience 
to his command, and had sent Iris to lash them back to Olympus 
with her sharp tongue, he came into his great hall and found them 
in sullen anger; they sat apart and said not one word to him, "but 
he knew in his heart, and he spake"— not graciously as Achilles to 
the heralds, but words of pure Olympian malice {Iliad viii. 446). 
When Apollo hearkened to the prayer of Glaucus, and suddenly, 
miraculously healed his wound and filled his heart with valor, 
"Glaucus knew in his heart, and he rejoiced, that the mighty god 
quickly had hearkened to his prayer" {Iliad xvi. 530 f.). When 
Hector, tricked by Athena in the guise of Deiphobus into fighting 
Achilles, had cast his spear vainly and stood unarmed, and called 
loudly to Deiphobus for a spear, and his brother was not by him at 
all, "Hector knew in his heart, and he spake, 'alas, truly now have 
the gods called me to death, for I said the hero Deiphobus was by 

" Iliad iv. 74 and elsewhere. 

28 P. 93 f. The Knightes Tale, line 1921; The Milleres Tale, line 18. 



8 University of California Publications in Classical Philology [Vol. 12 

me, but he is within the wall, and Athena hath tricked me, and now 
at last is ill death near to me. . . . ' " (Iliad xxii. 296 ff.). The same 
formula, slightly varied, is in the Odyssey in still another context. 
When Nausicaa asks her father for the wagon to carry her clothes 
to the washing-pits, she modestly hides the fact that she is thinking 
of wedlock, "for she was ashamed to speak the words 'fruitful 
marriage,' to her dear father, but he knew all, and answered her" 
(vi. 67). The poet felt the delicate beauty of this gem, and used it 
delicately in chosen settings, no two of them alike. Had he been 
under the necessity of finding a different phrasing each time, he 
could scarcely have expressed his thought with this perfection and 
with so brief simplicity .^^ 

One other example will suffice, I believe, to show how freely and 
how effectively the repeated line is used in Homer. We meet often 

'iv t' apa ol ipv x^^P^ cttos t' 'e<paT' e/c r' bvoixa^tv, 
or 

X€tpt re ^XLV Karepe^ev eiros t' itpar' tK t' ovofxa^ep, 

or more specialized variants, all suggesting emotion and introducing 
an earnest, affectionate, or cordial address. This so-called stock line 
is one of four which give the final touch of poignant beauty to the 
parting of Hector and Andromache. When Hector had tossed the 
babe Astyanax in his arms, and uttered a prayer which the hearer 
knows is not to be fulfilled, "he put his child in the arms of his dear 
wife ;- and she then took him to her fragrant bosom and smiled amid 
her tears; and her husband felt pity when he saw, and he patted her 
with his hand, and spake, and called upon her, 'Ah, my dear, be 

not too grieved for me ' "^° 

These instances, I believe, show not only that the repeated line 
is just as much a part of the poet's traditional material as is the 
half-line, the noun-epithet, or the single epic word, but also — and 
here is the important point — that it is used just as freely and im- 
consciously as the single word. Though it is true that in any par- 

-^ It is almost impossiy)le in discussing a work of art to refrain from express- 
ing one's individual tastes, and I have naturally chosen exami)les which ap- 
peal to me. It will be found, I think, that the case does not rest on these 
personal judgments. 

^'' Iliad vi. 482 ff. These and many other instances of repeated lines are 
incompatible with Chantraine's feeling that lines often repeated must be 
"banal" (pp. 123 ff. and pasfiim). 



1933] Calhoun: Homeric Repetitions 9 

ticular context the poet may choose the Hne consciously to produce 
a particular effect, just as we may consciously choose a word, it is 
equally true that he may use it unconsciously, instinctively, and as 
freely as he would use a single word. The conclusion is inescapable 
that the repeated hne involves no principle of technique and no 
problem of criticism other than is involved in the lesser traditional 
formulas. By epic canons, repetition of entire lines is no more a 
blemish, no less artistic, than repetition of name-epithet formulas, 
for the author composed in lines as readily as does the modern poet 
in words. 

I have already remarked that no sharp division can be made at 
any point in our series of repetitions; from the single word to the 
passage of many hnes they merge into one another by insensible 
gradation. There are any amount of brief epic formulas, such as 
01 5' ore 8-q, avrlKa 8' e^, rj pa Kal aiiireiraXuv, ovk olos afxa rw ye, which 
are expanded into hne-formulas, and these again into longer for- 
mulas. Although the longer passages cannot be sharply distin- 
guished from the repetitions of single lines, they are naturally more 
complex and more hkely to be used with conscious dehberation. 
Some of them in fact are worked out so elaborately and at such 
length that the probability passes into certainty. In these we may 
hope to find elements of technique which are not to be found in the 
repetitions of single lines, and an effect upon the hearer which the 
single line cannot produce. I propose to examine first a very fa- 
miliar formula which recurs constantly throughout the Odyssey. 

Every reader of Homer will recognize instantly the passage which 
begins 

X^pPilSa 5' ajU(^t7roXos Trpoxow eirex^ve (pepovcra 
KaXfj xP'-'o'ttTj. . . . 

These five hnes of great beauty are used six times in the Odyssey, 
always in the description of a meal offered in the reception of an 
honored guest. ^^ They are used only where there is a retinue of 
servants; thus they do not occur in the entertainment of Odysseus 
by Calypso or by Eumaeus, or of Telemachus by Nestor on the 
seashore, or in the humble abode of Laertes, or in the camp before 

31 Once of the disguised Athena (i. 136 ff.), twice of Telemachus at the court 
of Menelaus (iv. 52 ff.; xv. 135 ff.), twice of Odysseus, once in the palace of 
Alcinous (vii. 172 ff.) and again in the abode of Circe (x. 368 ff., not in all 
MSS), once of Theoclymenus in the home of Odysseus (xvii. 91 ff.). 



10 University of California Publications in Classical Philology [Vol. 12 

Troy.^^ The formula is never varied, but is often elaborated by 
appropriate additions or combined with other formulas, such as 
describe the bathing, clothing, and seating of the guest, the carving 
and serving of the meat, the pouring of the wine, or the offering of 
the chine. The auxihary formulas in turn may be given varying 
degrees of elaboration, and the whole passage may be expanded by 
the insertion of speeches, but always with strict attention to the 
situation the poet has in mind. 

One need only read the passages in which this formula occurs to 
understand the technique in its general outlines. What we have is a 
clearly enunciated central figure used with more or less elaboration 
to introduce important scenes ;^^ it is like a fanfare of trumpets 
which may be prolonged and elaborated as the composer may desire. 

Evidently the poet thought that these recurring formulas had a 
function and produced an effect. When we ask what was the effect, 
we are thrown upon our own experience or that of others with whom 
we have read the poems, aloud or at least so as to be audible to 
ourselves. 

When one comes to these beautiful lines, after having read them 
very many times, he feels, it is true, that passive pleasure of relaxa- 
tion associated with famiUarity. What Bowra conceived to be a 
function of the single repeated line may well be one of the functions 
of the longer repetitions. But one feels also an active, positive 
pleasure closely akin to that induced by the enunciation of a fa- 
miliar figure in a favorite opera or orchestral composition. This is in 

^2 Descriptions of entertainment in the camp, where the five lines under 
discussion cannot be used and the effect must be got by other means, show 
the care with which these materials are used. When Achilles entertains the 
envoys from Agamemnon, the meats are varied and honorable and their 
preparation by Achilles' own hand is described in detail; Patroclus and 
Automedon join in the service, and Achilles himself places the portions be- 
fore his guests (ix. 199 fT.). In the last book, the intense scene between Priam 
and Achilles is closed with the entertainment set before Priam by Achilles, 
with the help of Automedon, and the preparation of Priam's bed (xxiv. 621 
ff.). The wounded Machaon in Nestor's lodge is served with KVKeoiv by 
Hecamede, with promise of a warm bath to follow (xi. 618 ff., xiv. 5 ff.). The 
formula under discussion might have been used where Thetis is offered iflvia 
by Charis and Hephaestus, who have mechanical servants in plenty, but in 
this scene the poet seems to be striking the note of simplicity; the actual 
service is not alluded to, but only the offer of refreshment. The mere sugges- 
tion of the formula is found where Priam makes libation before he starts for 
the camp (xxiv. 304). 

'' It is of course obvious that in epic poetry repeated phrases and 'figures' 
cannot be developed as in music, where a brief phrase may provide the full 
content of a long passage, but must be developed by additions of related 
material. 



1933] Calhoun: Homeric Eepetitions 11 

part a pleasurable expectancy, and curiosity as to whether the sub- 
ject will be worked up in the way he remembers and will introduce 
the scene or the orchestral movement with which he associates it. 
Just as the first notes which warn the hearer of a recurring figure in 
music leave him not quite certain of the exact form in which it is to 
be developed this time — unless he has the composition by heart — 
so the words which mark any one of the prefatory formulas by 
which these five lines may be introduced — the bathing of the guest, 
or his seating — leave him uncertain as to how elaborately the 
subject is going to be developed. But when he hears 

Xtpvi/Sa 8' ampiiroXos .... 

he knows what is to come for a few lines, his mind runs ahead of the 
poet, and he feels a distinct pleasure. Then, as the subject is worked 
up more and more elaborately, the pleasurable feehng of expec- 
tancy becomes progressively more intense and he is sure that a 
notable scene is coming.^'' 

With this we maj' compare the use in the Iliad of the formula of 
arming for battle introduced by the familiar line 

KPTifildas fiiv TpojTa irepl Kvi)p.rj(nv edrjKe. 

Although it might easily have been inserted in many places, it is 
in fact reserved for the introduction of four major episodes, (1) the 
combat of Paris and Menelaus which begins the fighting (iii. 328 
ff.), (2) the general engagement which brings the fortunes of the 
Achaeans to their lowest ebb (xi. 16 ff.), (3) the great counter- 
attack led by Patroclus (xvi. 130 ff.), and (4) the final onslaught 
under Achilles (xix. 369 ff.). In (1) we have what may be termed 
the normal form, with no adornment, but preceded by elaborate 
preparation for the duel; in (2) this is expanded by description of 
Agamemnon's breastplate and shield and bj^ a portent; in (3) the 
normal form is combined with harnessing of the horses and an 
elaborate fivefold marshaling of the Myrmidons; in (4) it is ex- 

5* The scenes which our six passages prelude, are, in order, the crucial 
interview of Athena with Telemachus which starts the working of the plot; 
the first colloquy between Telemachus and Menelaus; the reception of 
Odysseus by Alcinous with the definite promise of the hero's return to 
Ithaca; the rescue of Odysseus' comrades from the enchantment of Circe; the 
departure of Telemachus from Sparta, marked by a propitious omen and by 
Helen's prophecy; Telemachus' recital to Penelope, with the solemn assever- 
ation of Theoclymenus that Odysseus has already come to Ithaca. 



12 University of California Publications in Classical Philology [Vol. 12 

panded by such description as has not been anticipated in the 
preceding book, by the harnessing of the horses, and by the prodigy 
of Xanthus' speech to his master. At other points where the formula 
might have been used with entire suitabihty as regards narration, 
it is omitted and a brief phrase takes its place.^^ The arming of the 
heroes is consistently suited to the situation assumed by the poet.^" 
It would be quite wrong to conclude that these recurring formulas 
are used in a purely mechanical way in accordance with some rigid 
system of composition, or that the poet did not seek for variety 
when he was so minded. The repetitions are combined and recom- 
bined in endless variety, as may easily be seen from those groups of 
related formulas which have to do with sacrifices, feasting, meals, 
and libations. The most elaborate account of a sacrifice {Od. iii. 430 
ff.) concludes the visit of Telemachus to Nestor, where the usual 
formula is expanded bj' the gilding of the heifer's horns, the detailed 
allocation of each act to a member of Nestor's family, and the 
bathing of Telemachus. In Iliad i. 447 ff., where the sacrifice follows 
the restoration of Chryseis (cf. infra, p. 16) the prayer of Chryses 
is inserted in the formula; in Iliad ii. 402 ff., where the sacrifice 
precedes the muster of the host, we have the prayer of Agamemnon 
to Zeus. The sacrifice in Iliad vii. 314 ff. which s the finale to the 
duel of Ajax and Hector is marked by the presentation of the chine 
to Ajax. In the sacrifice which is a prelude to the destruction of 
Odysseus' comrades (Od. xii. 356 ff.) the formula is varied to take 
account of special circumstances (lack of barley and wine), and the 
consternation of Odysseus, the wrath of Helius, and the portent all 
contribute to the effect. The description of the sacrifice with which 
the visit of Telemachus to Pylos begins is suited to the precise 



3* E.g., Iliad vii. 206 ff., where the casting of the lot has provided an ade- 
quate prelude to the duel; xiii. 241; except for Paris before the duel, the arm- 
ing of a Trojan hero is matter for few words, e.g., vi. 504; xvii. 192 ff. 

^^ In Iliad X. 254 ff., Odysseus and Diomede require special equipment and 
the formula cannot be used, but the poet gets his effect by describing the 
helmets and giving the history of the boar-tooth helm. Description is used 
in the prelude to Sarpedon's assault upon the wall (xii. 294 ff.) and the shield 
of Ajax is described in vii. 219 ff., where the formula of arming is omitted; cf. 
also the description of Pandarus' bow, Iliad iv. 105 ff., in the prelude to the 
wounding of Menelaus. In Iliad xiv. 9 ff., Nestor evidently has been sitting in 
his body armor, and merely takes spear and the shield of Thrasymedes, 
which permits a bit of description to take the place of the usual formula. In 
Iliad XV. 479 ff., Teucer has already body armor, and takes shield, helm, and 
spear when he lays aside his bow. InOi/.xxii. 120 ff., Odysseus lays aside his 
bow and takes shield, helm, and spears, the only arms available. Cf. the 
exchange of arms in Iliad xiv. 381 ff . On the arming of Athena, cf . infra, p. 14. 



1933J Calhoun : Homeric Sepetitions 13 

moment of arrival (cf. Od. viii. 469 f.; xxiv. 362 ff.; Iliad xviii. 558 
ff.)- A^Tien elaboration is not demanded by the poet's plan, or by his 
mood, the description is reduced to mere narration, and whole 
hecatombs may be offered in a line or two (e.g., Iliad i. 315-318; 
Od. XX. 276 ff.). 

Sacrifices of course imply meals or feasting, and vice versa, but 
meals are often described without explicit allusion to sacrifice.^' 
Some of the more elaborate descriptions have been discussed, since 
they involve the entertainment of a guest,^* but there are others 
notable for the use that is made of formula. Since a feast is to be the 
setting for the slaying of the suitors, the poet assumes a day of 
festival {Od. xx. 156, Trao-tj/ eoprri; cf. xxi. 258), when feasting on a 
grand scale is appropriate. The sacrifice of a hecatomb in the grove 
of Apollo is dismissed with three fines (xx. 276-278) ,'^« but the 
preparation for the fatal banquet in the palace is elaborately de- 
scribed, the lighting of the fire and decking of the banquet hall, the 
arrival of the herdsmen one after another with the victims, the 
slaughtering, the roasting of meats, and mixing of wine, all ex- 
panded by incidents and speeches pertinent to the action. The 
events of the preceding day have likewise as their setting a repast 
in the hall of Odysseus, but here, except for the special entertain- 
ment of Theoclymenus (xvii. 84 ff.; cf. supra, p. 9), only so much 
is said about the preparation and serving as is essential to the nar- 
ration; the decking of the hall is given one fine, needed to motivate 
the presence of Eurycleia (xvii. 32), and the preparation of the 
repast is limited to a brief formula (xvii. 179 ff.). Odysseus' first 
day in Ithaca, with Eumaeus, concludes with an account of the 
evening meal and domestic sacrifice in which the formulas are 
suited to the special circumstances and considerably expanded. 
With this compare the meal extemporized by Eumaeus upon the 
first appearance of his guest (xiv. 72 ff.). A\Tiere no artistic purpose 
is served by elaboration, or where other subjects make a stronger 
appeal to the poet's mood, feasts are treated summarily. Of the 
three banquets in two days in the palace of Alcinous, the first (viii. 
59 ff.) is cut short for the more attractive material of the games; the 

^' Cf . the familiar use of iepevco of killing animals for food. 

'* Supra, pp. 9 ff . In addition to the passages discussed here, cf . the meal 
and sacrifice on the battlefield (Iliad viii. 545 ff.), and the meal in the house 
of Laertes (Od. xxiv. 362 ff.). 

55 With line 279 we revert to the feasting in the palace; after the entrails 
have been served (252 ff.) the Kpi' vireprepa are set on for the real feast. 



14 University of California Publications in Classical Philology [Vol. 12 

second we enter with Odysseus at the moment when (viii. 470) 
"they were now serving the portions and were mixing the wine," 
and all the preparations are left to inference; the terse description 
of the third (xiii. 24 ff.) makes us feel at every line the hero's im- 
patience to be gone. The funeral feast for Patroclus is dismissed 
with five lines {Iliad xxiii. 29 ff.), for a wealth of material is at hand 
in the games; the two lines which tell of the funeral feast for Hector 
are like a rich chord, softly sounded, at the close of a final move- 
ment (xxiv. 802 f.). Often the poet speaks briefly of meals in the 
ordinary course of the narrative, using a line or two of famihar 
formula,^° or even dismissing them with half a line.^^ In the tale of 
wandering, the simple formulas of ordinary fare^^ are occasionally 
varied by the note of abundant good cheer.^^ 

Other formulas of a line or more that are repeated over and over 
again have to do with assemblies and councils, the coming of dawn 
and of night, arising and dressing for the day, preparation of a bed, 
retiring to rest, drawing up an army, fighting, wounding, slaying 
and being slain, harnessing horses and setting forth in chariots, 
putting out to sea in ships and coming in to shore, storm and ship- 
wreck, the coming of gods from heaven, invocations, omens and 
prodigies, mental and emotional states.*^ These formulas tend to 
accumulate where the poet is introducing or concluding important 
scenes or momentous episodes. An example of elaboration and com- 
bination in a passage of moderate length is the introduction to the 
wounding of Ares {Iliad v. 711 ff.); first we have a most elaborate 
picture of Hera and Hebe assembling the chariot and yoking the 
horses,^^ then comes the arming of Athena, worked up with equal 

" E.g., Iliad ii. 399 (cf. vii. 466); vii. 380 (cf. xi. 730); ix. 88; xviii. 314; Od. 
XV. 500; xvi. 2, 453. 

41 E.g., Od. iv. 429 = 574, 786. « ix. 86 = x. 57; xii. 307. 

« ix. 161 ff., 556 ff.; x. 183 ff., 476 ff.; xii. 29 ff. Cf. the year of plenty in 
Circe's palace, x. 467 f. 

** Cf. Lang's list of the matters expressed by recurring formulas (p. 291). 
I give instances of the Homeric use in the following pages: assemblies, p. 16, 
n. 49; preparation of abed and retiring, p. 17, n..54; chariots, p. 14, n.45; ships, 
pp. 16 f. ; gods, p. 15, n. 46. A complete collection of formulas, followed by a 
careful study of their distribution {infra, p. 18) would add much to our 
understanding of the poems. In particular, a thoroughgoing analysis of the 
formulas of fighting, wounding, and slaying would certainlj- bring interesting 
results. 

*^ 720 ff. Cf. the harnessing of the wagon which is to bear Hector's body 
{Iliad xxiv. 265 ff.). The passages which have to do with harnessing and un- 
harnessing, going and coming with chariot or wagon, range in length from a 
few words to many lines, and often are e.\panded by the insertion of varied 



1933] Calhoun: Homeric Repetitions 15 

elaboration, and a stately progress from Olympus to the battlefield, 
which sounds effectively, if ironically, the note of pomp and circum- 
stance. Per contra, the abortive sally in the eighth book is intro- 
duced by the bare skeleton of this passage (382 ff .) ; cruel as the poet 
is at times toward the Olympians, he will not here involve the god- 
desses in too utterly ridiculous an anticlimax, and his trumpets are 
muted. Again, when rapid narration is wanted, the coming of the 
god is left to inference, or briefly stated in one line or two.'*® This 
suggests that we may profitably give some attention to the distri- 
bution of these formulas."*^ 

Every reader of the Iliad has felt the rapid movement of the first 
book, which covers many days and dexterously interweaves three 
or four threads of the story, and the striking change at its close to 
an economy which requires several lengthy books for the action of a 
single day.^^ The handUng of the recurrent formulas corresponds 
definitely to these two opposite types of exposition. In the opening 
episodes the briefer formulas are used occasionally as they are 
needed, but the longer formulas are absent, although there are 

material. The interweaving of the formulas gets an effect of rich diversity 
interspersed with familiar notes. Cf. for example, Iliad iii. 259 fT.; v. 364 ff.; 
viii. 41 ff., 382 ff., 4.34 ff., 440 ff.; xiii. 23 ff.; xvi. 148 ff.; xix. 392 ff.; xxiii. 291 
ff.; xxiv. 576, 690; Od. iii. 478 ff ., 492 f.; iv. 39 ff.; vi. 72 f., 88 ff., 253; vii. 4. ff : 
XV. 145 ff. 

*^ In the comings and goings of gods, the familiar notes of recurring formu- 
las are varied and elaborated as the artistic purpose of the poet requires or 
his mood dictates. That the goddess Hera now requires a chariot {Iliad v. 
720 ff.; viii. 382 ff .), now flies through the air (xiv. 225 ff., 281 ff.; xv. 79 ff.; 
19. 114 f.), now works her will from afar (i. 55; xix. 407), and now sends a 
messenger (i. 194 ff.; ii. 156 ff.; 18. 166 ff.) does not result from the conflicting 
theologies of different periods, but from the poet's varying moods and pur- 
poses. Often the gods are at hand, and nothing is said of their coming from 
heaven (e.g., Iliad iv. 439, 507; v. 312, 353: x. 507 ff., 515 ff.; xxiii. 383 ff., 
where Athena at least is actually present; Od. vii. 19; x. 277, where it must be 
remembered Odysseus is speaking, not the poet; xxii. 205). At times a brief 
formula brings them darting down {0rj 5k kolt OvKvtnvoio Kapr^vuv or ^n 5i Kar' 
'Idaiwi' opticv occur most frequently, e.g., Iliad ii. 167; vii. 19; xi. 196; xvi. 677; 
.xxii. 187; xxiv. 121; Od. xxiv. 488; cf. cSpro be "Ipis, Iliad xxiv. 159 and "Up-n 
5' at^aaa X'nrev, Iliad xix. 114.), but often it is expanded bv description or by 
a simile (e.g., Iliad i. 44 ff.; iv. 74 ff.; xiv. 225 ff.: xv. 79 ff., 169 ff., 237 ff.; 
.x.xiv. 77 ff., 339 ff .; Od. i. 96 ff.; v. 43 ff.; cf. Iliad .xiii. 17 ff. and xxiii. 198 ff.). 
On Homer's use of gods, cf. J. L. Mvres, "The Last Book of the Iliad," 
J. H. S., LII, 1932, pp. 265 f. 

^' The question of te.xt enters here. The 'long' texts contained repetitions 
at points where they are not in the vulgate; on the other hand, some that are 
in the vulgate are reported to have been absent from certain ancient texts 
and have been attacked as interpolations. On the textual problern, cf . infra, 
pp. 23 f . In the present state of our knowledge any study of the distribution 
of formulas must start from the vulgate, or it will follow a vicious circle. 

« See Myres, p. 293. 



16 University of California Publications in Classical Philology [Vol. 12 

quite a number of places in which they might have been used. We 
do not find an accumulation of formulas until we come to the 
restoration of Chryseis. Thus there is no hint of the gathering or the 
dispersal of the assembly which cried out its approval of Chrj'ses' 
appeal, and even the fatal meeting at which the strife befell is 
opened and closed with a few quiet lines (54 ff., 305). But the 
assembly in book two which leads up to the first battle without 
Achilles is introduced with the most elaborate and sonorous de- 
velopment of the subject found in all the poems.^^ The fatal descent 
of Apollo is told in four lines, with scarcely a word that is not essen- 
tial to the action and ^vith the stark simile wktl eou<is, utterly 
unlike the more conventional elaboration of the subject so often 
found elsewhere. ""^ Again, the Achaeans purify the camp and offer 
whole hecatombs to Apollo in four Hnes (314-317), while the sacri- 
fice at Chryse is elaborately described. In fact the first accumulation 
of formulas we meet is used to set off the restoration of Chr\-seis 
and the revocation of the curse. The embarkation has been sand- 
wiched into the narrative, between the assembly and the purifica- 
tion of the camp,^^ and the episode begins with the landing at 
Chrvse,^^ followed almost immediately by the formulas of sacrifice 



" ii. 50 ff. The customary formulas are expanded by the meeting of the 
council, the similes, and description, and the account of the scepter. The 
nocturnal assembly that precedes the embassy (ix. 9 ff.) and that at which 
the reconciliation takes place (xix. 40 ff . ) have brief special introductions. 
The usual formulas precede the assembly of the Ithacans called by Tele- 
machus (Od. ii. 6 ff. ), but the assembly of the Phaeacians at which Odysseus 
is publiclv presented has a more elaborate introduction, with Athena usurp- 
ing thafunction of the heralds iOd. viii. 4 ff. >. The first assembly of the gods 
is opened with two lines (Ilind viii. 2 f.), but that which precedes the theo- 
machy is more elaborately introduced (Iliad xx. 4 ff.). Assemblies of the 
Trojans call for scant ceremony, and no fanfare of trumpets precedes. 

^* Lines 44-47; cf. supra, p. 15, n. 46. 

51 Lines 308 ff. Since this is the only instance in which a ship is actually 
launched in the Iliad, comparable passages must be sought in the Odyssey, 
e.g., ii. 389 ff.; iv. 778 ff.; viii. 48 ff.; xi. 1 ff. The situations differ and the 
formulas vary greatly, but familiar notes are interwoven. Cf. infra, p. 17, n.5o. 

»2 Lines 430 ff. With 432 ff. cf. Od. xv. 495 ff. and xvi. 322 ff. In Od. xvi. 
351 ff., the landing of the suitors is taken up at the instant Amphinomus 
catches sight of them. The landing of Telemachus in Pylos is briefly treated 
(iii. 10 ff. ). probably because the sailing has just been elaborately developed 
and the poet prefers to pass on to the sacrifice and feast. Similarly in lUad 
i. 484 ff. a brief treatment is desirable after 432 ff. and 478 ff. In the tale of 
wanderings, as a rule, excessive repetition is avoided by condensing the for- 
mulas or substituting other shorter ones, but if an introduction is wanted for 
one of the episodes, description and special circumstances may be introduced, 
as in ix. 106 ff.; x. 1 ff., 81 ff., a device used also for the landing of Odysseus 
in Ithaca (xiii. 93 ff.). Cf. infra, p. 17, n. 55. 



1933] Calhoun: Homeric Eepetitions 17 

with invocation, feasting, and worship,*' the coming of night, going 
to rest, and the coming of dawn, ^putting out to sea,*^ and arrival 
at the camp. The atonement and revocation of the curse are im- 
portant, and the poet is not content with a few hnes which would 
be almost lost between the two scenes with Thetis. Where a modern 
would get his effect by dwelling upon the emotions of father, child, 
and onlooker, the epic poet adorns the episode by these more formal 
means.*^ The scene on Olympus concludes with brief formulas of 
feasting, the coming of night, and going to rest. With Agamem- 
non's rising and the gathering of the assembly we enter on the 
changed economy which marks the days of fighting. From here on 

^^ Lines 447 ff.; cf. supra, pp. 12 ff. 

" Lines 475 ff.; for very similar combinations of these motifs, cf. Od. xix. 
424 fT.; iv. 574 ff., etc. Formulas for the coming of morning and evening ex- 
hibit considerable variety, e.g., morning, Iliad i. 477 = xxiv. 788, often in 
Odi/ssey (ii. 1; iii. 404, 491," etc.); Iliad viii. l=xxiv. 695 (cf. xix. 1 f.; ii. 48 f.; 
xi. 1 L = Od. V. 1 L);Od. vi. 48; xv. 495; xv. 56 = xx. 91: Iliad xxiii. 109 (cf. Od. 
x-xiii. 241); vii. 421 ff.; Od. iii. 1 ft'.; x.\iii. 344 fT.; Iliad x.xiii. 226 ff. (cf. vii. 
433 ft".; Od. xiii. 93 ft".); i. 493 = .\xiv. 31 (cf. Od. v. 390; ix. 76; x. 144; Iliad 
xxiv. 12 ff.); evening, Iliad i. 475 = Od. ix. 168, etc.; Od. ii. 388, etc. (cf. Iliad 
vii. 465; Od. vi. 321; viii. 417); Od. i. 421 ff. = xviii. 304 tT.; iii. 329 = v. 225; 
Iliadi. 605 (cf. Od. xiii. 33, 35; xvi. 220 = xxi. 226 = //iW .xxiii. 154); viii. 485 ff.; 
xviii. 239 ff.; Od. xiv. 457 f.; see also Iliad vii. 282. 293; Od. iii. .334 ff . One who 
takes the trouble to go through the poems and study the beginning and end 
of each day will observe an artistic purpose in most of the variations and 
elaborations; these motifs are usually ancillary to others and frequently 
begin or end groups of formulas. The formulas for evening are not allowed to 
intrude into the narrative at the l)eginning of Iliad xxiii and xxiv, and at xxiv. 
351 we have a casual allusion; the start of a new day is always, I believe, 
definitely marked. For the technique of the formulas of going to rest, which 
often include the preparation of the bed, I refer to the following passages: 
Iliad i. 476, 606 ff.; vii. 482; ix. 658 ff., 712 ff.; x.xiii. 58 ff.; x-xiv. 2 ff., 643-676 
(omitted, vii. 380, 432; xviii; x.xiii. 218 ff.); Orf. i. 424ff.; ii.393ff.; iii. 396 ff.; 
iv. 294 ff., 793 f.; v. 226 f., 474 ff.; vii. .335 ff.; .xiii. 17; .xiv. 518 ff.; xv. 494; 
xvi. 481; .xviii. 428; xix. 47 ff.. 600 ff.; .xx. 1 ff.; xxiii. 288 ff.; cf. iii. 490 = xv. 188. 

" Cf. Od. ii. 389 ff., 414 ff.; iv. 780 ff., 842 f.; v. 263 ff. (Odvsseus' raft); 
viii. 50 ff. and xiii. 70 ff.; xv. 221 ff., 284 ff. In the tale of wanderings, very 
brief formulas are used, except in the major episodes (cf. supra, p. 16, n. 52); 
cf. ix. 62 f., 105, 177 ff., 470 ff., 561 ff., 565 f.; x. 28 f., 77 ff., 126 ff.; .xi. 1 ff., 
636 ff.; xii. 144 ff., 401 f.; see also, in the tales of Nestor, Menelaus, and the 
disguised Odysseus, iii. 153 ff.; iv. 577 ff.; xiv. 252 ff. The sailing is some- 
times separated from the launching; in one instance (viii. 50 ff. and xiii. 
70 ff.) the whole tale of wandering intervenes. 

*^ The episode has often been stigmatized as an interpolation of Odyssean 
material (e.g., Wilamowitz, Die Ilins und Homer, Berlin, 1920, pp. 256 f.; 
Cauer, pp. 615 f.; Chantraine, pp. 147 ff.), on grounds that either are subjec- 
tive or rest on a complete failure to understand the nature and use of epic 
formulas. The reason why the formulas of 430-487 occur very often in the 
Odyssey is simple and should l)e obvious to a person of even moderate intelli- 
gence. When we perceive that repetitions are not 'copied' from one another, 
the case against the episode collapses. 



18 University of California Publications in Classical Philology [Vol. 12 

the formulas appear and reappear, and often are developed quite 
elaborately, but always in harmony with the situations. 

These examples appear to justify a few observations on epic 
technique. Formulas may be repeated word for word ad libitum, or 
they may be varied, either to suit the situations or merely for 
variety's sake. Their groupings are constantly varied, so that the 
patterns into which they are interwoven are of endless diversity. 
Those variations which are required by the narrative and its situa- 
tions are seldom if ever neglected. The formulas tend to accumulate 
and to lengthen in the introductions and conclusions to major epi- 
sodes, and to dwindle or disappear in rapid narration, or where 
other matter offers, more attractive to the poet. Obviously we are 
not dealing with mechanical repetitions of the 'primitive' improv- 
isator, which have been left far behind, or with fortuitous results 
of interpolation by copyists, or with a congeries of reciprocal imi- 
tations. We are dealing with an art which ceased to exist in its 
entirety after the conditions in which it developed had ceased to 
exist, an art we can hope to understand and explain only in the 
degree in which we succeed in recreating those conditions. But we 
can see that the poet composes in Hnes and even in longer formulas 
as readily as in words, and chooses according to his mood and his 
artistic purpose, but otherwise with entire freedom, from all that the 
epic tradition has produced and accumulated. He uses the formulas 
at times to build up sonorous preludes, at other times because the 
narrative calls for them, at still others, we may suspect, merely 
because they suit his mood. His technique can be made out in 
general outline. It cannot be reduced to a sj'stem. 

From the recurring formulas and groups of formulas, more or less 
expanded and elaborated with other material, must be distinguished 
the repetitions of specific passages." These are handled in a way 
which depends to some extent on how widely they are separated 
from their originals. Reports of speeches and recapitulations of 
events are usually motivated effectively, and consist of summary 

^' Repetitions of this type, like the formulas, are unquestionably an ancient 
device of oral composition which has developed to an effective technique in 
the Homeric poems. Since formulas are found everywhere, it is not alwavs 
possible to distinguish specific from formulary repetitions; for example, it is 
hard to classify the two tales of shipwreck {Od. xii. 403 ff . and xiv. 301 IT.), and 
the story of Penelope's web (Od. ii. 93ff.,xix.l37flf.,xxiv.l28flf.) is packed with 
formula. An extensive collection of "l^eabsiclitigte oder berechtigte Wieder- 
holungen" has been made by Pfudel (pp. 9 flf.), who sometimes confuses spe- 
cific and formulary, as in Ilidd v. 719 ff. and viii. 381 ff . (p. 35; cf. supra, pp. 14 f ). 



1933] Calhoun: Homeric Bepetitions 19 

with some verbatim repetition ;5* as a rule they are too far distant 
from the originals to make contrast elTective, and such variations as 
occur are the result of dramatic or psychological necessities.^^ On 
the other hand, the repetitions involved in messages, commands, 
and answers to questions tend to follow closely upon the originals, 
and this gives an opportunity for a very adroit and interesting tech- 
nique of unexpected insertion and omission. It may be illustrated 
from Agamemnon's dream {Iliad ii. 5 ff.). Zeus's message is origi- 
nally only five lines, but the Dream preludes it with an admonition 
of equal length, and puts in two hnes of his own at the end; 
Agamemnon in turn quotes the Dream faithfully to within a little 
more than a line of the end, and then omits, rather unexpectedly, 
the concluding words. The effect of these variations upon the 
hearer, if I may judge from my own experience, is quite definite. 
He waits expectantly for the Dream to begin the actual words with 
which he has just heard him charged, and he knows that verbatim 
repetition may begin at any moment. When it does begin, he listens 
with the pleasure of satisfied expectancy, which is sharply inter- 
rupted when the speech does not stop at the moment expected but 
goes on. In the second repetition, the listener is lulled into a secure 
feehng that he is to hear the words of the Dream in full, and this 
again is interrupted when Agamemnon cuts off the quotation a 
little before the end. Another example is the threat Zeus sends to 
Athena and Hera by Iris. Iris, Uke the Dream, prefaces the original 
message by several lines of her own, and then just when we expect 
her to stop, shocks us by a vicious two-line thrust at Athena, her 
own idea; two lines which most critics would strike out because of 
their "unmannerly rudeness."^'' Such variations are naturally most 
effective when the passages are not too widely separated .''^ Ocea- 
ns E.g., Achilles' narrative to Thetis, Iliad i. 365 ff.; the report of Agamem- 
non's offer, ibid. ix. 264 ff. = 122 fT., and of Achilles' rejection, ibid. 684 {T.= 
417 ff.; Telemachus to Penelope, Od. xvii. 124 fT. = iv. 333 ff.; repetitions from 
Circe's instructions, Od. xii. 37 ff., throughout book xii; the recital of Tiresias 
words bv Odysseus, Od. xxiii. 268 ff. = xi. 121 ff.; cf. also Odysseus' two tales 
of the raid, .xiv. 258 ff. and xvii. 427 ff.; and note the inversion in Od. xiv. 323 
ff. and xix. 288 ff. 

53 E.g., Od. xix. 154 f. gives Penelope's point of view, and ii. 108 f. = xxiv. 
144 f . that of the suitors, and the added lines xxiv. 147 ff . are needed for the 
recital. Cf. the politic change in Iliad ix. 300 ff. from 158 ff. 

6» Iliad viii. 399 ff., 413 ff. Cf. Leaf's note to 420-424, Cauer, p. 432. Pfudel 
(p. 13) excuses Iris on the score of the instructions Zeus gave her in 399 f. 

«i Without attempting strict classification or detailed analysis, I refer to 
some typical repetitions; verbatim, or with slight changes, Iliad iv. 195-197 = 



20 University of California Publications in Classical Philology [Vol. 12 

sionally opportunities for these abrupt changes are offered in the 
commonplaces and stock lines. A good instance is the entertainment 
of Odysseus by Circe, where the motifs of bathing, clothing, seat- 
ing, and serving the guest are sounded impressively, and we are all 
ready for the feasting, when suddenly, in the midst of a line, the 
orderly course of events is sharply interrupted by the anxious grief 
of Odysseus for his unfortunate comrades.'^" When, in the words of 
Bowra,"^ "both the ear and the mind slackened some of their effort," 
the poet thus diverted himself by giving his hearers, suddenly, an 
unexpected nudge. He does so often, yet not often enough to let 
the device become wearisome. 

It may be objected that surprise is impossible if the audience is 
thoroughly familiar with what is to follow. Yet the little twists and 
turns in the dream episode still give us a distinct pleasure after we 
have read the lines so often as almost to know them by heart. There 
is also the undoubted fact that we listen to the opera Hansel und 
Gretel with full and perfect knowledge that it will be the bad witch, 
and not Hansel or Gretel, who will be popped into the fiery oven, 
yet the climax never fails to produce its effect. Ancient Athenian 
audiences in the theater knew perfectly well the precise moment at 
which an incognito would be penetrated, but they were none the 
less thrilled by a good recognition scene. The fact seems to be that 
emotions can be taken by surprise even when reason is forewarned ; 
surprise is involved with suspense and expectancy. This leads to 
the further conclusion that the artistic use of suspense is not in- 
hibited or even greatly limited by foreknowledge on the part of the 
audience. Familiar as we maj^ be with the scene in which the dis- 



205-7; XV. 160 ff. = 176 ff.; xi. 187 ff. = 202 ff.; xvi. 454 ff. = 671 ff.=681ff. (two 
last lines omitted); iii. 69 ff. = 90 ff. (two last lines omitted) = 253 ff. (slight 
changes, two lines restored); Od. i. 374 ff. = ii. 139 ff.; iv. 112 = 143 Uwo lines 
added), 724 fT. = 814 fT. (appropriate variations); vi. 57 f. = 69 f.: vi. 313 ff.= 
vii. 75 ff.; xvii. 346 f. and 351 f.; cf. Hind xv. 561 fT. and 661 ff. In Iliad ii. 158 
ff. and 174 ff., the change in 175 from 1.59 may he for metrical convenience, 
but the effect remains. Note in vii. 362 ff . and 389 ff. the herald's shrewd 
editing of Alexander's proposal. In Od. ii. 218 ff., the poet cheats our apjire- 
hensions by suddenly chopping off the verbatim repetition of i. 287 ff . at 223 , 
when we are holding our breath for fear Telemachus will blurt out too much. 
A comparable effect is got in Hind I. 379 ff., when Achilles suddenly inter- 
rupts the verbatim repetition at the precise moment we expect to hear 
Hrj ere, ykpov . . . 

^^Od. X. 373. The formulary line Iliad i. 363 = xvi. 19 is effectively inter- 
rupted in xviii. 74; Pfudel (p. 8) suggests there mav be an intentional comic 
effect in Od. xviii. 355 (cf. Iliad xiv. 141; Od. xxi. 288). 

63 P. 88. 



1933] Calhoun: Homeric Repetitions 21 

guise of Odysseus is penetrated by his aged nurse, we are still, emo- 
tionally, on the anxious seat during the long recital of the hunt on 
Parnassus, which the poet so artfully interposes. Similarly, we may 
know every note in the long finale of the Fifth Symphony, and yet 
cannot hear it in any other state than tense expectancy, exactly as 
the composer intended. It is true we can never have more than once 
the full effect of a first hearing of these passages, when reason, as 
well as our emotions, was taken by surprise, but the effect upon the 
emotions persists. We can never recapture completely what we feel 
when for the first time we round the shoulder of a mountain and 
come suddenly upon an unexpected vista, but we can never round 
that shoulder without being moved. 

In addition to the recurring formulas and the repetitions of mes- 
sages, there are innumerable instances in which words and even 
lines are repeated for no other reason than the sheer sensuous 
beauty of the echo.*-' These need not be studied here in detail be- 
cause they represent a device which is common to the poetry of all 
ages and therefore requires no explanation or justification in the 
Homeric poems.*^^ I omit also any consideration of the repetitions 
in the similes, which constitute a special problem. 

What are the net results of this inquiry? First, without stressing 
the analogy between the Homeric repetitions and recurrent figures 
and phrases in music, I yet maintain positively that sound Homeric 
criticism cannot be based exclusively upon visual study of the text. 
It must be prepared to regard the art of Homer as a purely auditory 
art, whose effect upon the audience, like the effect of a musical com- 
position, was not destroyed or impaired, but rather intensified, with 
rehearing. In this connection I must again stress the fact that dis- 
like of repetition, so far as we know, has been concomitant with 
visual analysis of the poems, and that lines or passages which seem 
'hackneyed' when examined visually by means of a concordance do 
not produce that effect when read aloud in context. Second, I 
believe I have shown conclusively that no line of demarcation can 

" E.g., Iliad, vi. 269 and 279; ix. 437 and 444; xviii. 38 and 49; Od. xiv. 457 
and 475. 

6* For example such echoing of words and phrases as Od. i. 422 f . can be fully 
appreciated bv anv reader of modern poetry. An interesting collection of 
repetitions, mostlv anaphoral, is made by O. M. Johnston, "Repetition of 
Words and Phrases at the Beginning of Consecutive Tercets in Dante's Diftne 
Comedy;' P. M. L. A., XXIX (1914), pp. 537-549. The Iliad and Odyssey 
illustrate well the various effects that can be got by repetitions of this 
general type. 



22 U7iiversity of California Publications in Classical Philology [Vol. 12 

be drawn between the briefer traditional formulas and repetitions 
of a line or more, and that the epic poet composed in lines or groups 
of lines as freely and as unconsciously as the modern poet composes 
in words. Third, the examples I have given of the recurring formu- 
las and groups of formulas estabhsh definitely, in my opinion, that 
repetitions of this type are deliberately employed, in accordance 
with a conscious technique, to produce definite effects. 

These conclusions, if valid, must modify the presuppositions 
upon which the textual criticism of Homer has hitherto rested. 
First, it is no more possible, in the great majority of cases,''* to 
determine which is the 'original' occurrence of a repeated line or 
passage than it is to find the 'original' instance of /Socottis iroTvia "Hprj 
or TToXu/ir/rts 'Odvacrevs. The attempt must be abandoned. With it 
must go the attempt to pick out 'early' and 'late,' 'original' and 
'imitation,' 'genuine' and 'spurious,' by comparing repetitions and 
evaluating them from subjective criteria. Cauer admits that these 
attempts have been carried too far, which he thinks was no more 
than human; they are not, however, to be given up, but only used 
with greater caution.*'^ He recoils from the possibihty "dass man 
deshalb all die Arbeit, die Wolf, Lachmann, Grote, Kirchhoff, 
Wilamowitz und viele andere seit Generationen getan haber, fiir 
verfehlt halten und die Hoffnung, etwas von der Geschichte des 
griechischen Epos zu erkennen, aufgeben solle." With all deference 
to these illustrious men and all gratitude for their arduous labors, 
and those of their followers, it must yet be said, unhesitatingly, 
that their conclusions must be given up in so far as they rest on the 
false premise that repetitions are 'copied' from one another, and 
the hope of writing the history of the Greek epic must be abandoned 
in so far as that hope is based upon these conclusions and upon 
this premise. 

Second, the concept of the 'dispensable' line can no longer be 
used, with the fine freedom that hitherto has prevailed, to include 
any and every line which may be not strictly necessary to the 
construe; henceforth we must take into account the character of 
the passage in which it stands, the effect at which the passage aims, 

®^ A line found only once may he formulary, as Meillet has pointed out (Les 
origines indo-euro-peennes des mklres grccs [Paris, 1923], p. 61; Parry, L'epithete 
traditionnelle, p. 10). Chantraine's proposal of digamma as a test (p. 137) is 
interesting, hut its feasibility doubtful in the present state of our knowledge. 

*' Pp. 611 ff . This position was taken by Pfudel (p. 7) soon after the publica- 
tion of Rothe's conclusions. 



1933 J Calhoun: Homeric Eepetiiions 23 

and even the distinct possibility that the poet felt like repeating the 
line at this point, and did so, without any more definite motive than 
prompts a modern poet to slip in a 'dispensable,' but colorful, word. 

Third, passages containing a large proportion of repeated lines 
can no longer be dismissed without more ado as 'wretched centos,' 
but must be examined in their relation to the economy of the poem. 

Finally, we must question the universal assumption that the 
shortest text is ipso facto the best, the nearest to the 'original,' and 
all else interpolation. In its last analysis, this is simply the sub- 
stitution of subjective criteria for diplomatic evidence; it is the 
familiar method which starts from the postulate that what is 'best' 
in the text is the work of 'Homer' and ends by attributing to 
'Homer' what the critic likes. In all its various disguises, this 
method assumes two periods of good taste, that of Homer and that 
of the critic, separated by dark ages in which the work of the poet 
was plastered over with unseemly accretions. As regards the Ihad, 
it assumes that a great poet composed an Ur-Ilias, a Menis, or an 
Achilleis, which his successors buried under a conglomeration of 
tasteless imitation, or else he composed a crisp, terse version of the 
poem, somewhat shorter than the text of Zenodotus, which was 
interpolated with innumerable pointless repetitions, first by rhap- 
sodes and later by copyists; now, Uving in a second era of austere 
good taste, we have as our task to purge the text of these accre- 
tions in the hope of uncovering something that we may be able to 
admire. The primary weakness of these methods is their complete 
failure to understand or to explain the poems in the form in which 
they were the very foundation of classic Greek literature — which 
after all is the proper function of the Hellenist. They offer us in 
exchange for the Iliad and the Odyssey selections of fragments with 
which to piece together each for himself our "original Ihad" and 
our "Return." Furthermore, they fail of the objectivity ordinarily 
demanded of the scholar in that they do not attempt to establish 
from the actual tradition of the poems what were the canons of 
taste of the society in which they were composed, but assume that 
these canons were substantially ours, and set aside, upon no valid 
historical ground, the earUest and most direct diplomatic evidence 
we possess, the Ptolemaic papyri. The assertion that the problem of 
the long texts is a question of taste — meaning the taste of the critic 
— has been so often reiterated that it has acquired the sanctity of 
the commonplace. It is simply untrue. This is a historical problem 



24 University of California Publications in Classical Philology [Vol. 12 

in the field of literature, the rational approach is from objective 
interpretation of the actual data, and the question of taste enters 
in only to the extent that we are seeking to ascertain from these 
data what were the epic canons of taste. 

In a great majority of cases, the discrepancies between our texts, 
from the shortest that can be estabhshed on diplomatic grounds, 
through the Soilgate' to the Ptolemaic, involve the presence or 
absence of such repetitions as we have been studying. In the 
shortest text that can be made by the ordinar\' processes of textual 
criticism, although certain repetitions here and there will be eUmi- 
nated, the quantity will not be greatly reduced from that of the 
vulgate or the incidence materially altered. The plus verses of the 
Ptolemaic papyri are for the most part hnes or groups of hnes 
already known to be formulary; they come from more frequent 
repetition of single lines or distichs than is found in the correspond- 
ing parts of the \'iilgate and from aggregations of longer formulas at 
points where they are not in the vulgate.^^ So far as our material 
goes, they resemble the repetitions of the vulgate but are more 
abundant. We cannot discard them en bloc and keep any text that 
will be based on the manuscripts. Textual criticism must take 
account of the possibihty that the Ptolemaic papyri, our earhest 
extant manuscripts, are nearest to the 'original' form of the poems, 
and that the discrepancies between short and long texts arise 
mainly from excision and not interpolation.®® If there is any objec- 
tive ground for ruling out this possibihty, it should be brought 
forward at once. 

Many Homeric critics are reluctant to admit that so much in the 
poems is traditional because they feel that in so doing they impugn 
the poet's originaUty.^" Fuller understanding of the nature and use 

8* The most recent and complete list of plus verses known to me is given by 
Collart, pp. 340-342. It must be kept in mind that in many cases it is uncer- 
tain how the fragmentary lines are to be restored. 

" The possibility is unquestionably reinforced by Collart' s neat demon- 
stration (pp. 342 fT.) that there is in fact no evidence for the existence of the 
vulgate prior to 150 B.C., and his rather startling conclusion, Part II, pp. 
36, 52 fT., that the long texts are in fact the pre-Alexandrian ^•ulgate. If the 
long texts are the 'original,' the variations in the manuscripts are not interpo- 
lations, but the result of contamination in texts which have been abbreviated. 

""^ This feeling again is a more or less unconscious intrusion of modem taste 
into our attempt to recreate Homeric canons; it accounts for the apologetic 
tone so frequently adopted in speaking of the 'stock' lines as a sort of struc- 
tural makeshift; "e.g., Chantraine, pp. 123 f. The architectural analogy will 
seem plausible when certain formulas are compared in brief and in expanded 



1933] Calhoun: Homeric Bepetitions 25 

of repetitions will remove this difficulty. The significant fact is that 
the poet composed in lines, or groups of lines, as well as in words; 
this granted, we cannot reproach him for lack of originality, when he 
repeats lines or passages, any more than we can attack a modern poet 
for using the words he finds in his language and not inventing new 
words to express his thought. The reader of Homer who has grasped 
this point will see abundant originality even in those passages 
which are richest in foniiulas. At the same time, all who find in the 
poems the handiwork of a verj' great artist will allow that this 
artist's part in the creation, the adaptation, and the polishing of 
the epic formulas mvist have been in some measure commensurate 
with his greatness, though we cannot say definitely that a particu- 
lar formula is his original creation. If we can but listen with the ears 
of the ancients and forget for a moment our prejudice against repe- 
tition, we see that formulary' lines ought by rights to be, as often 
they are, among the most perfect and most beautiful; they have 
stood the test of sur\ival with many singers and of selection by the 
singer. We have gradually learned that in every part of the text is 
traditional material that can only be the collective work of ages 
and in every part are touches that can only be from the hand of a 
great master. We are beginning to suspect that the two cannot be 
neatly separated. The facts seem to admit the hypothesis of a su- 
premely great poet, working with traditional material, who left the 
Iliad and the Odyssej' substantially in the form in which we have 
them. Shall we strain at this, after we have been compelled to admit 
a long line of poets capable of evohing traditional material of this 
beauty and perfection, and successive generations of hearers who 
demanded such beauty, appreciated it, and stimulated the poets? 

form (e.g., Iliad v. 719 ff. and viii. 381 ff.; iii. 330 ff. and xix. 369 ff.), but grows 
less satisfying as we extend our inquir}- and see that the formulas do not 
compose a framework on which the decoration is hung, but are the decora- 
tion. On the whole, I believe the analogy with musical composition will be 
found the more useful. 



>c? 



ARISTOPHANES AND THE PNYX 

BY 

JAMES TURNEY ALLEN 

The ghost of Welcker's perverse theory with regard to the Pnyx 
was laid long years ago ; since then Chandler's identification of the 
site has gradually won universal acceptance.^ Yet many problems 
remained obscure until they were solved, or most of them were 
solved, by Kourouniotes and Thompson, whose article "The Pnyx 
in Athens"^ inaugurated a new epoch in our understanding of this 
ancient structure. 

As a result of new and extensive, though unhappily still incom- 
plete, excavations they have proved that in the fifth century before 
Christ the Pnyx was comparatively small and sloped downward 
toward the north. At the end of the century, probably about the 
time of the Thirty Tyrants (404/3),^ it was reconstructed, enlarged, 
and made to slope in the opposite direction. For the Pnyx as it 
existed in the fifth century the rock of the hillside had been artifi- 
cially dressed, and formed a shallow cavea with a gradient of about 
16 per cent. At the foot of the slope Kourouniotes and Thompson 
conjecturally restore an earthen fill — mostly level, but in part slop- 
ing to meet the slope of the hill — and along the northern edge a re- 
taining wall about fifty-six meters long and about two meters high. 
The bases for this wall and some of the stones of the wall itself are 
still extant. On this fill they assume that there was a bema of stone, 
and according to their measurements the level part of the fill at the 
center was about seven meters from front to rear and the sloping 
part about three meters more. It is clear, however, from their draw- 
ing (Plate IV, B) that, if the retaining wall had been a meter higher 
than they assume, the earthen fill would have been level through- 
out its extent and would have had a width at the center of about ten 
meters from front to rear. 

1 For bibliography, etc., see Judeich, Topographie von Athen-, 1931, pp. 
391 fif. 

2 Hesperia, VoL I, 1932, pp. 90-217. 

' Cf. Plutarch, Themistocles, xix, 4: to 0rjfia to kv irvvKl ire7roi.T]fj.hov wot' cltto- 
/SXeireiv irpos T-qp daXaacrav vaTtpov oi rpiaKovTa wpos ttjv xwpaj' aTreffTpexpav. 

[27] 



28 University of California Publications in Classical Philology 

In the course of their discussion the authors very properly cite a 
number of passages from the comedies of Aristophanes. In some 
citations, however, they seem to the present writer to have misin- 
terpreted the evidence, especially with reference to the bema and 
the seating arrangements. Let us accordingly examine anew the 
passages which have a bearing on these two items. 



In verses 303-313 of the Knights, Aristophanes through the mouth 
of the chorus bitterly assails the demagogue Cleon as one who has 
befouled the entire city, one who has filled with, his impudence the 
assembly, the courts, and the pubhc treasury. He concludes the 
strophe (v. 313) with the metaphor of a lookout man posted on the 
rocks to watch for a school of tunny — "watching from the rocks the 
tribute, tunny-fashion, shoaling in," as Rogers neatly turns it: 

aTTO Ta)v ireTpccv avcjodev tovs (j)6povs OvvvoaKoirC^v. 

Because the Knights was presented at the Lenaean festival about 
the first of February, 424, this passage was no doubt inspired bj^ the 
activity of Cleon following his great victory at Pylos in August, 425, 
in connection with the revision of the assessment fists.'* But to say, 
as is commonly done, that the words inro T(hv ireTpdv are a reference 
to the Pnyx or even specifically to the bema^ is to blur an essential 
feature of the figure. They do not refer to the Pnyx or to any of its 
parts. Nor do they refer to the Acropohs, as Blaydes alternatively 
suggested.^ Instead they are merely a part of the tunny metaphor.^ 

* See Meritt and West, "The Athenian Assessment of 425 B.C.," Univ. of 
Mich. Studies, Humanistic Series, Vol. XXXIII (1934); Cavaignac, "L' Aug- 
mentation du tribut des allies d'Athenes en 425," Revue des Etudes Grecques, 
1935, pp. 245-249. See also Tod, Greek Hist. Inscr., 1933, pp. 148 ff. 

* The Pnyx: Mitchell, Kock, Blaydes, van Leeuwen, Neil, Coulon-van 
Daele; the bema: Merry. 

Dobree conjectured a7r6 ttjs wtTpas. Miller {Daedalus and Thespis, Vol. I, 
1929, p. 294) apparently thought that the reference was to the stone seats of 
the Prytanes — "they are still there above the bema" — and translated: "from 
the stones above." The interpretation was wrong even in 1929 when the Pn\TC 
was not understood. Since the publication of the article by Kourouniotes and 
Thompson we have known that the stone seats above the bema did not exist 
at the time of Aristophanes. 

* "Allusio fortasse ad Pnycem . . . aut ad Acropolin." 

^ Rogers in his note on v. 313 rightly insists that "the tribute would not be 
descried, or captured from the Pnyx." 



Allen: Aristophanes and the Pnyx 29 

In tunny fishing the watchman stood on the rocky cUff.* 

The same principle of interpretation obtains in many a metaphor 
and simile. Thus in Pindar 01. II. 94 ff . : 

cro(f>6s 6 TToXXcL eldcbs 4>va' 

fiadbvTts be \a^poL 

Tra'y'i\iaaala KopaKes &s aKpavra yapverov 

Atos irpos opvLXO- Gt'iov 

the words Xd|3pot xaTTXwo-ata are a description of ravens and there- 
fore should not be construed with <ol> nadbvTes, as many editors and 
translators take them,^ but rather with KopaKts}'^ Those who have 
merely learned their art, like ravens impetuous in garrulity, chatter 
ineffectually at the divine bird of Zeus. Aelian testifies" that the 
Kopa^ is of all birds -KoKyKKayyoTaTOS re /cat TroXu^wi^oraTOS. So m 
Aeschylus Frat/. 307:12 

OS acTTtvaKTi dvpvos OJS -qvelx^TO 
avavBos 

avavSos is an epithet of Ovyyos}^ And in the fragment of unknown 
authorship which is sometimes, though doubtfully, ascribed to 
Aeschylus i^* 

ovK rjv a.p' ovbtv Trjn' eXevdepav Bclkvov 

xf^vxw 6/xotcos av8p6s cos drtjuta. 

ovtcjos Trkirovda /cat /xe uvp.4>opa.s det 

^adeta /crjXts e/c ^vdSiV avacrrpecpeL 

\vaar]s TrtKpots KevrpoLcnp rjpedLaiJLeuov 

the words k ^vduv belong with avn<l>opas . . . ^adela /cr/Xts, for the 

8 See Aelian De Nat. Anim. xv, 5, Oppian Hal. Ill, 620 fif . For other passages 
both ancient and modern descriptive of tunny fishing see the notes in the 
various editions of Aeschylus {Pers. 424), Herodotus (I, 62), Theocritus 
(III, 26), etc. 

9 Boeckh, Dissen, Schneidewin, Boehmer, Fennell, Seymour, Myers, 
Sandys, etc. 

1" So Mommsen, Metzger {Siegeslieder, p. 166: "Xafipoi . . . ist der Gegensatz 
vonMos"), Gildersleeve, Puech. Schroeder and Bowra print vv. 95 f. with- 
out punctuation marks as in the text above. 

" De Nat. Anim. 11,51. 

»2 Nauck, Trag. Graec. Fragm., Sec. ed. 1889, repub. 1926. 

" Cf. Aesch. Pers. 577 f.: avavSo^v TraiSwv ras aiiiavTov, Hor. Carm. iv, 3, 
19: mutispiscibus. 

1^ See Nauck., op. cit., p. 861. Clement of Alexandria, Miscellanies, Vol. II, 
p. 462, says that the lines were spoken by Ajax before his suicide. 



30 University of California Publications in Classical Philology 

figure is that of blood spurting from a severe wound as in Sophocles 
Philoctetes 783 f . : 

ara^et yap av jjlol (jiOLVLOv t68' eK ^vdov 

Kf]Kiov aljua. 

No misery biteth a free man's soul (i.e., inflicts so deep a wound) as 
doth dishonor. . . . Ever the deep stain of my calamity (welUng up) 
from the depths (of the wound) upsetteth me, agitated (as I am) by 
the piercing goads of frenzy. 

With these passages in mind — their number could easily be mul- 
tiplied^^ — let us turn to Aristophanes Knights 956, where Cleon's 
signet ring is said to have as its device a gull open-mouthed upon a 
rock haranguing the crowds : 

Xdpos K^xw^^ ^^' irerpas brnirj'yopSiv . 

Here the schohast, who of course was unacquainted with the Pnyx 
of the fifth century B.C., explains irerpa as to ^rjp.a to kv -wwd, and 
modern scholars have adopted this interpretation. But in so doing 
they have overlooked the fact that the word Trerpa belongs to the 
figure, the figure of a gull w^th wide-open beak upon a rock. Of 
course it suggests the bema, but it is not a designation of the bema. 
When Aristophanes refers specifically to the bema in the Pnyx he 
calls it \ldos}^ Failure to appreciate this point led Kourouniotes 
and Thompson" to remark that the literary evidence that the bema 
in the days of Aristophanes was (1) \ldos (a block of detached stone) 
or was (2) Trerpa (living rock undetached) "simply cancels out." But 
it does not. In the plays of Aristophanes the bema is \ldos, not irerpa. 
In no passage does either Aristophanes or any of his contemporaries 
designate the bema as irerpa. 

II 

The seating arrangements of the fifth-century Pnyx have been dis- 
cussed in recent years not only by Kourouniotes and Thompson,'^ 

" Perhaps Aesch. Agam. 1472 ff. is another example. But the rendering of 
these lines is uncertain because the text is corrupt. At any rate the figure is 
that of a foul carrion bird perched upon a corpse: kiri aufxaros SLkuv KopaKos ixOpov 

'* Peace 680: roO XlBov rod h irvKvi, Eccl. 87: vir6 TW Xt0w. 

1^ Op. cit., p. 113 and p. 136. See also Dinsmoor, Amer. Jour. ArchaeoL, 
XXXVII (1933), p. 181, and Kourouniotes and Thompson, ibid., p. 652. 

^^ Op. cit., p. lllf. 



Allen: Aristophanes and the Pnyx 31 

but also by Willems in an article which appeared in 1905/^but which 
has been ignored by almost all subsequent writers.^" The conclusion 
reached in both articles is the same : during a session of the Assembly 
the populace sat or squatted on the floor of the auditorium. Kourou- 
niotes and Thompson point out that the rocky part of the floor was 
not cut into the form of seats and that there is no "evidence that 
wooden benches had ever been placed upon it. Had they been," they 
remark, "we should expect to find some trace of cuttings in which 
they might have been bedded; for without such it would have been 
distinctly difficult to arrange the benches on the sloping irregular 
surface." Their view that the people sat on the floor of rock is con- 
firmed, they beheve, by the testimony of Aristophanes in three 
passages. 

The first of these is Knights 754, where Aristophanes speaks of 
Demos as sitting "on this rock" : 

orap 5' CTTt ravrrjcrl /ca^fjrat ttjs irerpas. 

This is interpreted to mean that "the recognized place for the popu- 
lace was the bare rock."^! That this is a misinterpretation, however, 
is shown by the preceding verse. The passage in full is as follows: 

6 7 dp yepcov 
oIkol pev avSpcjv tart Se^iajxaros, 
orav 5' kirl ravT-qcrl KaOrJTaL ttjs Trerpas 
Kkxt]vtv cbawep kpiro8i^cov tcrxciSas. 



19 "Les Atheniens a I'Ecclesie," Bull. Acad. Roy. de Belgique, 1905, pp. 809 
ff. Willems' conclusion was adopted by Fougeres in Daremberg et Saglio, 
Diet, des Ant., s. v. Pnyx; byStarkie, Acharnians (1909), p. 6, and byRennie, 

Many of Willems' arguments are trivial, and even worthless. Typical is his 

treatment of avavrjdav in Eccl. 428: vtavias . . . av€Tvi]bi)a . . . 57ifj.r]yopriffo>v. 

"Une derniere preuve, d'ordre philologique, acheve de mettre le fait hors de 
doute. Celui qui desirait prendre la parole 'sautait en pied' (dvajnjSav). C est 
le terme propre, et il est si caracteristique que pas un helleniste ne s'y trom- 
pera" (p. 813). . , 

But Willems failed to observe that dreTnjSTjcre (428) is contrasted with 
irapeipirvaev (398) and TapijXee (409), and he ignored such passages as Herod. Ill, 
155:<6 Aap6Tos> €K Tov dpSvov avairrjdijaas, and Xen., Hellen. IV, 5, Q--T.'Ayri<rl\a-os 
. . . KaO-qnevos kirl tov irtpl T-qv \lp.v7\v KVKKoTtpovs olKohop.rip.aTOS . . . evdvs re tK ttjs 
ibpas av€Trr]8r)(T€. The word cii'aTrrjSai' no more implies a previous position upon 
the ground than do the words of the poet: "My heart leaps up." Yet Fougeres, 
op. cit., selected this argument of Willems for special approval! 

20 Kourouniotes and Thompson do not mention it, nor does Judeich, Top. 
vonAthen"^, 1931, pp. 391 ff. Miller also {Daedalus and Thespis, 1929, pp. 293 fif.) 
makes no reference to it. 

21 Op. a/., p. 111. 



32 University of California Publications in Classical Philology 

ext ravT-qai rrjs irerpas is contrasted with ot/coi. At home the old fellow- 
is most clever, but here in the Pnyx he gapes hke one kinroSi^iov 
t(rxd5as, whatever that may mean.^^ The phrase em raurrjo-t rrjs irerpas 
no more implies sitting in contact with the rock itself than do the 
words of Aeschylus-^ when he says that Xerxes "occupied a seat . . . 
a lofty eminence" {edpav yap elxe . . . v\pr]\6v oxdov). Surely a line de- 
tached from its context may prove "a staff of a broken reed." 

The second quotation affords better support of their thesis. This 
is Knights 783, ^^ where the Sausage-seller says that Paphlagon, that 
is, Cleon, has no concern for the discomfort of Demos sitting "on 
the rocks": 



eirl rats Trerpais ov 4>povTi^eL aiiKrjpo^s ere Kadi] fxevov oOrcos." 



Thereupon, after the manner of a professional ward heeler, the 
Sausage-seller offers the old man a cushion lest he should gall his 
precious Salaminian rump {'Lva yiii Tpi^ys r-qv ev SaXa/xTvi) . 

Finally, to clinch their argument, both Willems and Kourouniotes 
and Thompson-'' quote the dream of Sosias as related in Wasps 31 ff . : 

eho^'e jxoL irepl irpwrov virvov ev rfj -kvkvI 
eKKKr](n.a^eLv Trpo^ara avyKaOrjiieva, 
^aKTTjplas exoPTa Kal TpL^covLa. 
/caTreira tovtols toTs TvpajSaTOLaL iiovdoKet 
drjixriyope^v (fjoKaiva iravdoKevTpLa 
******** 

kdoKei 8e IJ.OL Geojpos aurrys irX-qcFLOV 

Xaptai KaOrjadat Tr}v Ke4>a\7]v KopaKOS exoiv. 

"Parmi les assistants," remarks Willems, "il a discerne Theoros 
assis par terre, a cote du demagogue. Voila, j'espere, un detail dont 
la precision ne laisse rien a souhaiter." But it is a dream that Sosias 

22 The scholiasts did not know, nor do modern scholars. 

"Pers. 466 f. 

2* Quoted by Willems, p. 813: by K. and T., p. 111. Ribbeck as early as 1864 
{Die Acharner des Aristophanes, p. 193) also cited this passage as evidence 
that "Mann sass . . . zum grossen Theil unmittelbar auf dem Stein des Fels- 
bodens." 

25 rats Trerpais is the reading of the MSS and of Suidas. If sound, it consti- 
tutes an exception to the rule stated by White in his Verse of Greek Comedy, 
§790. Bentley conjectured rijs irkTpas, Brunck TaTcrt Trerpais, Lenting rrjade 
Trerpas, Velsen raiaSe Trerpais. 

26 Willems, p. 812; K. and T., p. 112. 



Allen: Aristophanes and the Pnyx 33 

is relating, and therefore the passage is of questionable validity as 
evidence. "Ground not upon dreams; you know they are ever con- 
trary."" The people are sheep, the speaker is a whale, Theorus has a 
raven's head and sits upon the ground (xafial). The last detail is in 
keeping with the rest. A raven certainly would not be expected to 
occupy a seat. Surely one would not conclude because Alice came 
upon a caterpillar smoking a hookah that caterpillars regularly 
smoke hookahs. 

The alleged Hterary evidence simmers down then to Knights 783, 
which does seem to indicate that some of the people sat on the rock. 
But that seats were not provided for any except the Prytanes passes 
belief, especially in view of Acharnians 25, where Dicaeopolis says 
that the Prytanes will jostle one another for a irpurov ^v\ov. Now a 
TrpoiTov ^v\ov ought to mean a front seat,^^ as it does in Wasps 90, 
where the reference is to a courtroom. Moreover, ^v\ov implies that 
the bench was of wood. But because of the rock-cut seats above the 
scarp behind the bema of the Hadrianic period, which face the 
audience and which were formerly supposed to belong to the Pnyx 
of the fifth century, and because of Ecdesiazusae 87, where the 
women are advised to seek places just below the bema (\ldos) oppo- 
site (KaTavTLKpi)) the Prytanes, scholars have generally assumed^^ 
that the -wpuiTov ^v\ov of Acharnians 25 referred to seats beside or be- 
hind the bema and facing the auditorium. But we now know, thanks 
to Kourouniotes and Thompson, before the date of the Ecdesiazusae 
(393 B.C.) the Pnyx had been turned around. And it is not impos- 
sible that at the same time the seating arrangements for the Pry- 
tanes were altered. At any rate, there is no evidence that in the fifth 
century, before the date of the Thirty TjTants, the seats of the 
Prytanes faced the auditorium. ^° Indeed, the expression irpoJTov ^v'Kov 
impHes the contrary and suggests that there were other ^v\a for at 
least part of the populace. On the terrace at the foot of the rocky 

" Thomas Middleton, The Family of Love, iv, 3. 

28 The scholiast says irpotbpla. For a discussion of this passage see my 
"Notes on Aristophanes," Class. Phil., XXX, 1935, pp. 262 f. The emenda- 
tion proposed in that article, eUoT for elra 6', clears up the syntactical diffi- 
culties with the least alteration of the text. 

29 Most recently Kourouniotes and Thompson (p. 103): "Benches for the 
officials were probably placed on the same terrace behind or to the sides of 
the bema." 

3" Another innovation of about the same date was the introduction by 
Agyrrhius of the eKKKr^cnaaTiKov. See Aristophanes Eccl. 102, 184 ff., 300-310, 
Plut. 329 f.; Aristotle Const. Athen. xli ad fin. 



34 



University of California Publications in Classical Philology 



slope there would have been space for a number of rows of benches.^i 
Perhaps they even extended up the slope. On this point I do not feel 
myself qualified to judge. But if the early Pnyx was as uncomfort- 
able a place as Willems and Kouroundiotes and Thompson make it 
out, one wonders why it was that the Athenians clung so long to the 
use of the Pnyx when the Dionysiac theater with its commodious 
seats provid ed a far more tempting place of assembly. 

J> Especially if, as was suggested above (p. 27), the retaining wall along 
the north had been about three meters high. ^ 



ON THE PROGRAM OF THE CITY DIONYSIA 
DURING THE PELOPONNESIAN WAR 

BY 

JAMES TURNEY ALLEN 

In a study of the City Dionysia a reading of Schneider's Das 
attische Theaterwesen (1835) side by side with Deubner's Attische 
Feste (1932) is instructive as revealing the progress that has been 
achieved in the course of a hundred years in our understanding of 
that important festival. In a general way the program of the City 
Dionysia as this was observed in the fifth century before Christ is 
now fairly clear, though many details still provoke controversy. 

It is usually assumed on the testimony of Aeschines^ that on the 
8th of Elaphebolion was held the Proagon and in and after 420 also 
the Asclepiaea,^ and that the City Dionysia proper commenced on 
the 9th.^ But whether the statue of the god was brought back to the 
theater from the Academy on the 8th or the 9th/ usually referred 
to as the claaywyn,^ and whether the procession {tto^tttj) was con- 
nected with this eiaayooyi] or was a separate event, and occupied only 
a portion of a day or with its attendant ceremonies and festivities 
filled an entire day, and whether one day or two were assigned to 
the choruses of boys and the choruses of men, and whether the kojuos 
followed these choruses or came at some other point in the program 
— these are questions on which agreement has not yet been reached. 
In addition to these the comedies have been the subject of a lively 

^ III, 67: rg oySoy iaraixtvovTov 'E\a4>rj(5oKiC}vos ijltivos, ot ^v tCo AaK\r]Tnu> 17 dvcria 
Kai 6 irpoayuv. 

2 See Korte, Ath. Mitt., XVIII (1893), p. 246, and ibid., XXI (1896), p. 315. 

3 Or, according to some, on the 10th. See Haigh, The Attic Theatre, 3d ed., 
revised by Pickard-Cambridge (1907), p. 7, n. 4; Xavarre, Dionysos (1895), 
p. 6, n. 4; Le Theatre Grec (1925), p. 106; Flickinger, The Greek Theater and Its 
Drama <*' (1936), p. 197, n. 3; and esp. Adams, Trans. Atner. Philol. Assoc., 
XLI (1910), pp. 60 S. Support for this view is found in the scholium on 
Aeschines III, 67. Adams presented a more elaborate argument. But the 
evidence is late and cannot be made to fit the facts, at least for the year 423 
B.C. See below, p. 39 and n. 35. Most authorities ignore it. 

^ Or on the 10th. Seen. 3. 

5 Nilsson's theory that the eicrayoiyr} was a daily removal of the statue from 
the temple to the orchestra, Jahrbuch, XXXI (1916), pp. 338 f ., is refuted by 
Deubner, op. cit., p. 140. 

[351 



36 University of California Publications in Classical Philology 

dispute. The question at issue here is whether the comedies were 
distributed through the afternoons foUo'VN'ing other events or were 
presented on a single day preceding the days devoted to the per- 
formances of tragedies. Strange as it may seem, these two confhct- 
ing views have never been reconciled ; and yet a reconcihation ought 
to be possible. 

The object of the present paper is to suggest a solution of this 
problem. To keep the matter within reasonable bounds the dis- 
cussion will be confined to the last third of the fifth century and 
especially to the period of the Peloponnesian war. In order to clarify 
the problem and to prepare the way for its solution a brief survey 
of the history of the controversy is desirable, the more so as an 
adequate account seems never to have been pubhshed.^ 

First, however, we should note that whereas during the nine- 
teenth century, and indeed until recently, it was the prevailing be- 
hef that in the fifth century before Christ at the City Dionysia and 
also at the Lenaea there were only three contestants in comedy,^ it 
is now known from the didascahc inscription discovered at Rome^ 
that before the war the program of each of the festivals included five 
comedies, as they did again after the war.^ That is to say, at some 
point early in the war period^" the number of comedies at each festi- 
val was reduced from five to three. This fact, which was divined by 
Petersen in 1885,^^ was proved by Korte in 1905.^- But it did not 

^ Dutoit, Zur Festordnung der grossen Dionysien (1898), traced in part the 
history of the controversy to 1898. But he ignored the English and French 
authorities. His own attempt to solve the difficulties is without value. 

" The hjTDotheses attached to the extant fifth-century plays of Aristophanes, 
all of which were presented during the Peloponnesian war, regularly mention 
three poets. But the hypothesis prefixed to the Lys. omits this item. The 
Thesm. has no hypothesis. 

* The inscription is in fragmentary form; see IG XIV 1097, 1098, 1098a. It 
embodied a record of Athenian comedy: the names of the poets, the titles of 
their plays, the date and rating of each play, and the festival at which it was 
presented. 

^ The latter fact has of course long been recognized. 

1" No data are available at present for the years between 429 (?) B.C., when 
Callias won fourth place with his Frogs (see IG XIV 1097, 1. 6), and 425 B.C., 
the date of Aristophanes' Acharnians. 

" Scaenica, Wiener St udien, VII, pp. 183 f. 

1- "Inschriftliches zur Geschichte der attischen Komodie," Rh. Mus., LX, 
p. 427. Korte's interpretation was accepted by Capps, "Roman Fragments of 
Athenian Comic Didascaliae," Class. Philol., I (1906), pp. 201 S. See also 
Lipsius, "Didaskalika," Rh. Mus., LXV (1910), p. 166. 






Allen: On the Program of the City Dionysia 37 

win secure foothold until Dittmer published his invaluable study in 
1923/^ and only recently has begun to receive recognition in the 
books and articles on ancient Greek drama. ^^ It is a fact of cardinal 
importance. 

The view that the comedies were performed in the afternoons has 
as its chief support verses 785 ff. of the Birds of Aristophanes, the 
date of which was 414. The chorus enumerates several situations in 
which wings would have been of service to the spectators. The first 
of these was that of the man who was hungry and was bored by the 
tragedies {ireLvwv toIs xopotcn twv rpayudi^p ijxSero). "If he had had 
wings," says the chorus, "he might have flown away home, had 
lunch, and then have flown back again to us" (/car' ap kuTfKrjadels e^' 
rinas aWts av KareiTTeTo) .^^ Becker therefore concluded as early as 
18401*^ that in the fifth century at the City Dionysia a tragic tetral- 
ogy was followed in the afternoon by a comedy. His statement im- 
phes a three-day program of drama. For it was known that three 
tragic poets competed each year, and it was believed, as we have seen, 
that in comedy there were also three contestants. In this form the 

" The Fragments of Athenian Comic Didascaliae Found in Rome. See the 
favorable reviews by Geissler, Gnomon, I (1925), pp. 175 f., and by Wust, 
Jahresber., CCVII (1926), p. 96. 

" See Geissler, "Chronologie der altattischen Komodie," Philol. Uniers., 
XXX (1925), p. 17; Flickinger, op. cit., 3d ed. (1926), notes on p. 356 and p. 
359; Allen, Stage Antiquities of the Greeks and Romans and Their Influence (1927), 
p. 38; Pickard-Cambridge, Dithyramb Tragedy and Comedy (1927), p. 218; 
Schmid, Gesch. der griech. Literatur, Vol. II (1934), p. 526. Korte, s. v. "Ko- 
modie," in Pauly-Wissowa, XI (1921), cols. 1228 f ., of course refers to his own 
article (see above, n. 12). It is not mentioned by Norwood, Greek Comedy 
(1931), nor bv Deubner, op. cit. (1932). Maidment, "The Later Comic Cho- 
rus," Class. Quart., XXIX (1935), pp. 1 ff., ignores the evidence of the in- 
scription and thereby seriously impairs his own argimient concermng the 
synchoregia. 

15 The phrase «(/>' vnas has been variously interpreted. The scholium (not in 
R or V) savs eiri to dkarpov. This was accepted by Mliller, Griech. Buhnenalter- 
thiimer (1886), p. 322, n. 4: "zu uns ins Theater." Cook, Zeus, Vol. I (1914), p. 
683, n. 3, supposing that Aristophanes was "joking at his own expense"— the 
parabasis consists of 125 lines,— adopted the old emendation rpvywdwv for 
Tpayoiblhv (see below, p. 39) and interpreted k4>' vixas as meaning "be back in 
time'for the next scene on the stage." Both interpretations are unsatisfac- 
tory, especially the latter. It is most unlikely that the expression rols xopoTcrt 
tSiv TpvyoibCiv would be used of the parabasis. Surely the prevailing view is 
right, that i4>' vnds refers to comedy as Miiller himself later acknowledged. 
Das attischeBuhnenwesen (1902), p. 8. _ utt u 

Not worth recording is the explanation proposed by Oehmichen, ' Ueber 
die Anfange der dramatischen Wettkampfe in Athen," Sitzungsber. d. Akad. 
d. Wiss. zu Milnchen (1889), p. 121. 

16 Charikles, Vol. II, p. 286. 



38 University of California Publications in Classical Philology 

theory was vigorously and clearly restated by Sauppe in 1855/^ and 
thereafter adopted by others. ^^ 

Older is the opposite view that the comedies preceded the trage- 
dies, a view based in the first instance upon the law of Evegoros, in 
which so far as concerns the City Dionysia comedies are mentioned 
before tragedies. ^^ Sauppe tried to annihilate this argument by de- 
claring the law of Evegoros spurious f° some more recent scholars, 
by contending that the order in which the events are mentioned in 
the law and also in the didascalic inscriptions represents not the 
order of performance, but that of an ascending scale of importance.^^ 

This theory that comedies preceded tragedies led an anaemic 
existence until in 1885 Lipsius gave it robust vigor partly by stress- 
ing a fact that had been previously observed but not emphasized,^^ 
that in the didascalic inscriptions the order of the items is the same 
as in the law of Evegoros, but more particularly by definitely pro- 
posing a four-day program of dramatic events with comedies on the 
first day.^^ In order to bring into harmony with this theory the 

1^ "DieWahl der Richter in den musischenWettkampfen an denDionysien," 
Ber. ub. d. Verhandl. d. k. sacks. Gesellsch. d. Wiss. zu Leipzig, phil.-hist. 
CZasse, VII (1855), pp. 19 ff. 

18 Among these may be mentioned Mommsen, Heortologie (1864), p. 388; 
Usener, "Nachtrage zur Geschichte des attischen Theater.s," Symbola philo- 
logorum Bonnensimn (1867), pp. 583 ff.; Haigh, op. cit. (1899, 1898, 1907), pp. 
23 f. of the 3d ed.; Navarre, Dionysos (1895), p. 45, and Le Theatre Grec (1925), 
p. 106; Miiller, Das attische Biihnenwesen (1902), p. 8; Flickinger, op. cit. (1918, 
1922, 1926, 1936), pp. 197 f. of all editions; Pickard-Cambridge, op. cit. (1927), 
p. 218. It should be noted that Mommsen later abandoned this view; see 
below, n. 24. 

1^ Demosthenes, Against Midias 10. 

2" Op. cit., p. 20. He followed Westermann, De litis instrumentis quae extant 
in Demosthenis oratione adv. Mid. (1844), pp. 20 ff. Foucart, "Sur I'authenti- 
cite de la loi d'Evegoros," Revue de Philologie, I (1877), pp. 168 ff., defended 
the genuineness of the law, and since then it has usually been accepted as 
valid evidence. The date of the law is unknown. 

21 So Haigh, op. cit.^^'' (1889), p. 34, and repeated in the succeeding edi- 
tions; Pickard-Cambridge, op. cit., p. 217. See also Navarre, Dionysos (1895), 
p. 44: "rien ne prouve que I'ordre, dans lequel les divers actes de la fete sont 
enum^res, soit celui de leur succession." He does not cite Haigh in this con- 
nection. Stahl, "De Evegori lege disputatio," Ind. led. Monast. 1893 aestiv., 
p. 14, reaches the same conclusion by a different course of reasoning. He also 
does not refer to Haigh. 

22 See, e.g., Kohler, "Zur Geschichte des athenischen Theaters," Ath. 
Mitt.,IU (1878), p. 106. 

23 "Bemerkungeniib. die dramatischen Choregie," Ber. ub^ d. Verhandl. d. 
k. sachs. Gesellsch. d. Wiss. zu Leipzig, phil.-hist. Classe, XXXVII (1885), pp. 
416 ff. 



Allen: On the Program of the City Dionysia 39 

testimony of Aristophanes he adopted in Birds 787 the emendation 
of an anonjonous scholar cited by Scaliger and proposed later also 
by Bentley of Tpvycc8o>p for rpaywbihv. Lipsius won many followers.^* 

The most recent discussion of the entire festival is that of Deub- 
ner.2* Bearing in mind that Thucydides states that in the year 
which we know as 423 B.C. the Athenians in public assembly rati- 
fied a truce with Sparta on the 14th of Elaphebolionj^" and that 
therefore the 14th in that year was evidently a business day," and 
believing with Lipsius that four days were assigned to dramatic 
contests, Deubner is compelled to crowd the early events of the pro- 
gram into too brief a period. His scheme is as follows: 9th tto/ztti?, 
choruses of boys, choruses of men, Kcofios; 10th, comedies; 11th, 12th, 
13th, tragedies. Surely not sufiicient allowance is made here for the 
time that the procession, which was evidently an elaborate affair,^^ 
and the ten choruses would require. Deubner's program for the 
9th is too full. 

Essentially the same is the arrangement adopted by Robert, 

who also had in mind the statement of Thucydides regarding the 

14th.2^ Mommsen assigned the ttomxt? to the 9th, the choruses to 

the 10th, the comedies to the 11th, the tragedies to the 12th, 13th, 

and 14th. As this scheme, however, did not harmonize with the 

testimony of Thucydides, he suggested— quite improbably — that 

before 420, the year in which the Asclepiaea was introduced, the 

festival was begun a day earlier, with the Proagon on the 7th.^'' 

2* Mommsen, who had previously adopted Sauppe's theory (see above, n. 
18), deserted to the ranks of Lipsius in his Feste der StadtAthen (1898), pp. 428 
ff . Of the others in this group one mav mention Oehmichen, op. cit. (1889), pp. 
115 ff.; Christ, Gesch. der griech. Litteratur^-^ (1890), pp. 167 f.; Girard, s. v. 
"Dionysos," in Daremberg et Saglio, II (1892), pp. 241 f.; Dutoit, op. cit. 
(1898), p. 32; Robert, GoU. gel. Am. (1899), p. 543, and ibid. (1913), p. 368; 
Kern, s. v. "Dionysos," in Pauly-Wissowa, V (1903), col. 1024; Cook, op. cit. 
(1914), p. 683; Deubner, op. cit. (1932), p. 142; Schmid, op. cit. (1934), p. 526. 

25 Attische Feste (1932), pp. 138 ff. Schmid, op. cit. (1934), p. 526, does not 
discuss the festival as a whole, but he adopts the same program as Deubner 
for the last four days (10th to 13th). Flickinger, op. ci<.<*' (1936), merely 
repeats the statements contained in the third edition (1926). 

26 IV 118, 12. 

" Dutoit, op. cit., p. 39, contended that the meeting of the assembly was 
held very early in the morning on a festival day. His argument is not con- 
vincing. 

28 See Mommsen, Feste (1898), pp. 437 ff.; Pfuhl, De Atheniensium pompis 
sacns (1900), pp. 74 ff. 

29 Gott. gel. Anz. (1899), p. 543. ^o /^^s^e, p. 434. 



40 University of California Publications in Classical Philology 

Over against these and others in this group^^ stand Haigh, Na- 
varre, Flickinger, and Pickard-Cambridge, who are ahke in distrib- 
uting the comedies through the afternoons, in ignoring the testi- 
mony of Thucydides for the year 423, and in prolonging the festival 
so as to include the 14th or the 15th. ^^ 

Between these two groups the tug of war still surges back and 
forth, the reason apparently being that both are right and yet both 
wrong. This paradox may be solved by combining the hypotheses 
prefixed to Aristophanes' plays with the didascalic inscription 
which was discovered at Rome. The former, as we have seen, men- 
tion three contestants during the period of the war; the latter, five 
in the years preceding the war. ^^ Let us assume first that in the fifth 
century in years when five comedies were included in the program 
of the City Dionysia all five were presented on a single day pre- 
ceding the three days devoted to tragedies. Let us further assume 
that during the war the four days of dramatic contests were reduced 
to three, with the comedies in the afternoons followmg the tragic 
tetralogies. If we can accept these assumptions, we need not with 
the followers of Sauppe reject as evidence the didascahc inscrip- 
tions and also the law of Evegoros, though the date of this is un- 
certain,^* nor yet on the other hand with Lipsius and his successors 
the passage in the Birds. In other words, we may accept Sauppe's 
theory as valid for the war period, and that of Lipsius for the years 
antedating the war. The trouble appears to have been that each 
view, though their authors were not aware of this, was correct for 
only a portion of the history of the festival. 

^1 See above, n. 24. 

2^ Haigh, op. citJ^\ p. 24: either 1) six days (10th to 15th), one comedy on 
each of the last three days, or 2) an alternative five-day program (10th to 
14th); Navarre, Dionysos, p. 7, and Le Tkedtre Grec, p. 106: the same as 
Haigh's first arrangement; Flickinger, op. cit.^*\ pp. 197, 200 (see also notes 
on p. 356 and p. 359) : six daj^s (10th to 15th), one comedy in the prewar period 
on each of the last five days; Pickard-Cambridge, op. cit., pp. 217 f.: six days 
(9th to 14th), one comedy each of the last five days. He does not, however, 
distinguish between the war period and that antedating the war. My own 
book Stage Antiquities of the Greeks and Ro77ians (1927) belongs in this group. I 
now believe that the statement concerning the program (p. 38) is incorrect. 

2' A good illustration is found in Callias, who won fourth place at the City 
Dionysia in 440, fourth place apparentlv at the Lenaea in 437 and again in 
429 or later, fifth place at the Lenaea in"434. See IG XIV 1097, and Dittmer, 
op. cit., pp. 39 ff. 

^* It is often assigned to the fourth century; so, e.g., Stahl, op. cit., p. 11. 



Allen: On the Program of the City Dionysia 41 

The reconciliation, then, of the two coniElicting views^^ can be 
effected by assuming that during the war the number of days de- 
voted to drama was reduced from four to three.^^ What may have 
been the reason for such a change? And why were the comedies re- 
duced from five to three? Both are questions of absorbing interest 
and the answers which we propose are very simple. The latter ques- 
tion has been answered hitherto by making reference in a general 
way to the exigencies of war.^^ No one, however, has attempted to 
probe to the bottom of this explanation. The former is of course 
based on an assumption, but it is an assumption that appears to be 
required if we would harmonize all the several bits of evidence. It 
seems very strange that no serious effort has been made since the 
close of the nineteenth century to effect such a harmony. 

The changes may have been adopted for two reasons, the first 
being to reduce the number of festival days in order to gain an extra 
day for business. If we assume that in the years preceding the war 
the procession had been held on the 9th, the choruses on the 10th, 
the comedies on the 11th, and the tragedies on the 12th, 13th, and 
14th, the reduction in the number of the comedies would make it 
possible to shift the tragedies to the 11th, 12th, and 13th, and to 
put the comedies in the afternoons following the tragedies, as the 
passage in the Birds certainly implies. ^^ By this simple rearrange- 
ment the length of the festival would be reduced from six days to 
five, and thus there would be obtained an extra day, the 14th, on 

3* A different solution is ofifered by Flickinger, op. cit.'^^-*\ p. 200, who fa- 
vors a six-day program with a comedy in the prewar period on each of the last 
five days. "This," he says, "would explain why in the inscriptional records 
the comedies preceded the tragedies, though in the chronological sequence 
of the last three days they followed them." He makes no provision, however, 
for the year 423, when a six-day program would have been too long, and when 
moreover the festival must have begun on the 9th, not the 10th (see above, n. 
3) . He does not mention Lipsius' theory. 

^^ Unfortunately the theoricon cannot safely be cited as evidence. Sauppe, 
op. cit., pp. 20 f., found support for his theory in the statement that the 
theoricon was a drachma, since this amount appeared to imply a three-day 
program at the rate of two obols per day. Lipsius, op. cit., p. 417, did not suc- 
ceed in answering this argument. Cleophon's diobelia was evidently not a 
theoricon. For a succinct discussion of this subject see Tod, Greek Historical 
Inscriptions (1933), p. 206. 

"So Korte, s. v. "Komodie," in Pa^lly-Wissowa, XI (1921), col. 1229; 
Geissler, "Chronologie der altattischen Komodie," Phil. Unters., XXX 
(1925), p. 17; Schmid, op. cit. (1934), p. 526. 

^* The Birds was presented at the City Dionysia. 



42 University of California Publications in Classical Philology 

which business of the state could be transacted, as Thucydides testi- 
fies that it was in the year 423. The second reason for the modifica- 
tion of the program may be found in the financial pressure occa- 
sioned by the building of additional warships. By reducing the 
comedies from five to three, not only at the City Dionysia but also 
at the Lenaea, four citizens of wealth would be released from chore- 
gic duties and so be made available for trierarchic service, an im- 
portant consideration at a time when Athens was straining every 
nerve to increase her naval power. 

In other words, the reduction in the number of comedies to- 
gether with the assumed shortening of the spring festival, may be 
interpreted as a war measure, intended partly to increase the effi- 
ciency of business at the very season of the year when the campaigns 
were beginning, partly to aid in building up a stronger na\^. The 
latter motive may have been chiefly responsible for the reduction in 
connection with the Lenaea. Whether the two festivals were modi- 
fied at the same time cannot at present be determined, though this 
would appear to be probable. Who took the initiative in proposing 
the measure or measures which effected these changes? One is 
tempted to suggest the name of Cleon. 



ACKNOWLEDGMENT 

I wish here to express my deep sense of obligation 
and gratitude to Professor George M. Calhoun, 
whose help in the preparation of this paper has 
been constant and invaluable. 

F.M.C. 



^z 



OMITTED SPEECH FORMULAS IN HOMER 

BY 

FREDERICK M. COMBELLACK 

The reader of Homer soon becomes accustomed to the poet's habit 
of marking clearly and explicitly the end of direct quotation, change 
of speaker, or conclusion of a conversation. The formulas used for 
this purpose might be thought of as the "quotation marks" of a re- 
cited poetry. Often referred to as speech formulas, they have formed 
the subject of a number of studies, by far the most important being 
a series of three monographs by Walther Wahmer.^ In these essays 
of Wahmer the regular usage of Homer has been clearly stated and 
abundantly illustrated. There are in Homer, however, a few pas- 
sages in which the regular usage is not followed and the "quotation 
marks" do not occur. It is only natural, since they are so unusual, 
that these passages should occasionally have excited comment; but 
in general it has been customary for editors and writers on Homer or 
on epic speech formulas either merely to cite them, regarding them 
as apparently inexplicable vagaries of an old poetry, or to attribute 
the omission to interpolation or to lacunae in the text. I shall at- 
tempt to explain them by showing that they fall into certain well- 
defined classes, in each of which there is a reason for the omission of 
the formula. 

In order to make clear the exceptional character of these passages 
it might be well first to state as briefly as possible the regular tech- 
nique. The formulas used after speeches in Homer fall conveniently 
into three classes : 

1 V. Berard, Introduction a VOdyssee, Vol. 1, pp. 85-125 (Paris, 1924) ; J. 
Berger, De Iliadis et Odysseae Partibus Recentioribus sive de arte inducendi et 
concludendi sermonis Homerica (Marburg, 1908); R. Schau, De formulis quas 
poetae Graeci in conclusione orationis directae posuerunt (Tilsit, 1890) ; Walther 
Wahmer, Ueber v, <is 0aro, au (Ituv, und verwandte epische Formeln (Program, 
Gymn. Gottingen; Part I, 1893, Part II, 1894, Part 111, 1903); E. WiUe, Auf 
welche Weise stellt Homer eine Verbindung zwischen der direkten Rede einer Per- 
son und dem Folgenden her? (Program, Neustettin, 1885); idem, Wie verfdhrt 
Homer, wenn er nach der Rede einer Person seiner Gedichte die Darstellung des 
Geschehens ivieder aufnehmen will; wie wenn er auf eine Rede eine under efolgen 
lassenwillf (Program, Neustettin, 1886). 

[43] 



44 University of California Publications in Classical Philology 

1. What we may call the regular us <^dro formula.^ This is an ex- 
pression referring to the fact that someone has just spoken, and is 
regularly used in Homer to mark a transition from speech to nar- 
rative; e.g., A 364, 

aj9 dirccv tovs fiev Xiirev avrov, ^rj 8e pier' aXXous. 

2. What we may call the regular dialogue transition. This ex- 
pression contains no ois ^dro formula, but serves to introduce the 
next speaker; e.g., A 544, 

Trjv 8' rjud^eT' eTretra irarrip avdpcbv re decov re. 

3. What we may call the ds ^dro dialogue transition.This contains 
in a single line a &s (J)clto formula and an introduction to the next 
speech;^ e.g., X215, 

ws k(j>aiJir]v, rj 8' avTiK ajid^eTO irorvLa iJir]T7]p. 

Almost all the speeches in Homer are followed by one of these 
three types of formula. None of the instances of omitted speech 
formulas in Homer involves dialogue or the use of dialogue transi- 
tions. In all there is an abrupt change from speech to narrative 
without the regular ojs </)d7o formula. Therefore we shall be con- 
cerned here only with the omission of the first type of formula. It 
should be borne in mind that, although these passages will be dis- 
cussed at some length, they are, after all, few in number (21) and 
plainly exceptions; the regular usage of Homer is consistent and 
clear. In comparison with the hundreds of passages in which the 
regular technique is followed, the passages I am about to discuss 
are really only a handful. 

The largest single group of passages in which the cos (^dro formula 
is omitted where the regular usage of Homer might lead us to expect 
it is that in which the circumstances are unusual with respect to the 
time sequence. A summary of the usual situations after speeches 
will make more obvious the unusual nature of these passages. The 

2 1 use the expression "ojs c^aro formula" as a convenient generic term for any 
expression used immediately after a speech and referring specifically to the fact 
that someone has just spoken. Besides Cos 4>dTo itself, many other expressions are, 
of course, used; for example, cos eiwcov, i) fiiv fip' cis eliroda', &$ ((j>ad', ^ pa, tis ap 
((pccvrjctv, cos ap f<j)ri, ojs tipa (t>covri(Tas, kt\. 

3 This type of transition is used by Homer only in a specific and unusual 
situation; namely, when dialogue is reported at second hand by one of the par- 
ticipants. It is therefore restricted to those parts of the poem where this situa- 
tion occurs, that is, Menelaus' story to Telemachus in 5 and Odysseus' tale of 
wanderings in i-^. 



Combellack: Omitted Speech Formulas in Homer 45 

most common situation after speeches is this : someone speaks and 
an action immediately occurs. This sequence of speech and act ap- 
pears in two forms : 

1. "A" speaks and then, having spoken, performs some action; 
e.g.,"HTOi y' ws dircbv Kar' 'dp' e^ero (A 68), "Qs ^dro Y[.r]\d6r]s, ttotI 8e 
(TKr\irTpov ^a\e yaly (A 245). A special type of this form is that in 
which the "action" is another speech, e.g., *H pa, Kal aiJL4>i.Tr6\oL<np 
evTr\oKanoLaL KeXevae. (f 198). 

2. "A" speaks and then "B" performs some action; e.g., "S2j 
(pdro, ^rj 5' ap' "Oveipos, eirei tov hWov iiKovcre (B 16), "12s t4>aT', 'ApyetOL 
8e fxey' 'iaxop (B 333). Here also is a special type in which the 
"action" is another speech, e.g., "12s e^ar', 'AvtLvoos 5' eirecnv vdKeaae 
crv^doTrju (p 374).^ This type of line is extremely rare. In all these 
situations it will be seen the poet first closes the speech with a ws 
<^dro formula and then proceeds to the action, which invariably 
is thought of as following the speech. 

A less common situation is this: "A" speaks, and while he is 
speaking something takes place which the poet tells us about im- 
mediately after the speech. Here belong many passages in which we 
are told about the effect of the speech; e.g., after Agamemnon's 
angry words in A 173-187 there is the line, "lis (l>aTo- Ur]\etcx}vi. 8' 
axos yeper', kt\. Here obviously we are not to suppose that axos came 
suddenly upon Achilles after Agamemnon had ceased speaking, but 
that this feeling began early in Agamemnon's speech and increased 
as it progressed. We may say, then, that here the action designated 
by the verb following the ois ^dro formula is simultaneous with the 
speech, whereas in the situations outlined in the preceding para- 
graphs it followed the speech.' A clearer example is in r where in the 
preparations for the duel between Paris and Menelaus we are told 
that Hector and Odysseus K\r]povs ep Kvpkri xo-^n-qpei ttolWop iXopres 
(r 316) : there is a typical prayer by a nameless "somebody"; fol- 
lowing the prayer is the line, "12s ap' e(t>ap, iraWep 8e fxeyas KopvdaloXos 

* Here Melanthius has spoken to the suitors and Antinous rebukes the swine- 
herd. This differs from the regular dialogue situation in which "A" speaks to 
"B" and "B" answers him. Consequently there is a ws <t>dTo formula here in- 
stead of a regular dialogue transition. In the summary given above I have, for 
the sake of simplicity and clarity, ignored dialogue, because it is not involved in 
any of the unusual passages which 1 discuss. 

= Cf. the remarks of Wahmer (II, p. 9) on X 333. 



46 University of California Publications in Classical Philology 

"^KTcop. This is perhaps the clearest indication in the poems that 
speech and action were simultaneous. 

In all the instances I have cited it will be seen that a ojs 4>aTo for- 
mula is used. Furthermore it must be noted that although the action 
designated after a speech may sometimes be thought of as following 
the speech and sometimes as taking place simultaneously with it, in 
all these passages the action is closely connected with the speech in 
time and place. If the action follows the speech, it follows it imme- 
diately; if it is simultaneous with it, it is an action performed by the 
speaker or his audience, that is, it occurs not only at the same time 
but also in the same place. When the "action" which is designated 
just after the ws <^dro formula is another speech, the second speech 
is always to be thought of as following the former speech and never 
as simultaneous with it. 

In one situation, however, Homer sometimes after a speech pro- 
ceeds to describe events which are to be thought of as occurring at the 
same time as the speech but in a different place. Rather often he 
reports a conversation of two or more persons and then describes 
the simultaneous action of another person or group; the transition 
is made clear by the line, "12s ol ixiv TotaDra 7rp6s dXXi7Xoi;s ayopevov. A 
t3^pical example of this technique is E 431. At line 343 Aphrodite 
drops the wounded Aeneas, who is taken under the protection of 
Apollo.The poet then tells us of Diomedes' ordering Aphrodite from 
the field, and her arrival on 01>Tnpus, where the conversation re- 
ported in 372-430 occurs. Then comes the transitional line, "tts ol 
fiep Tocavra irpos aWrjXovs a-yopevop, after which we are back on the 
battlefield again and Diomedes is making a charge upon Aeneas and 

Apollo: .,,.,,, o , , «, . ,f 

Aivei.q. tiropovffe porjv ayados ALOnrjdrjs, 

yLyvu}(7KU}p 6 ol avTOS vireipex^ x^^^pc-^ 'AiroWuv. 

By the transitional line, 431, used so often in situations of this kind, 
Homer has made it clear that this is to be thought of as simultaneous 
with the activities of Aphrodite which he has just described. 

With these facts in mind we are in a position to proceed to the 
discussion of the passages in Homer where there is no formula after 
a speech although Homer's regular usage would lead us to expect it. 
I believe that the formula is omitted here because the temporal and 
local relationship between the sp)eech and the immediately follow- 



Combellack: Omitted Speech Formulas in Homer 47 

ing narrative is different from the relationship in passages where a 
speech formula is used. 

First among these unusual situations let us take r 413. Here 
Autolycus, visiting his daughter to see his new grandson Odysseus, 
says that when the boy grows up and comes to his home he will 
give him gifts. There is no speech formula after Autolycus' words 
but the abrupt transition, tuv eveK rj\d' 'Odvaevs, tva ot TopoL ayXaa 
dupa. The action described in this line is one which occurred many 
years after the occasion when the speech was made. I believe that 
the ws <l)aTo formula has been omitted here because when that 
formula is used the action which is described as following the speech 
follows it immediately. There are, however, two passages in which 
Homer uses a &s <i>dTo formula in a situation rather like that in r. In 
I Odysseus reminds Achilles of Peleus' parting words when he sent 
Achilles to fight at Troy (254-258). After the speech of Peleus, 
Odysseus resumes thus: ws eTrereXX' 6 ykpoip, av 8e \r]dtai. Here the 
time referred to in XT7^€ai is long subsequent to that referred to in 
the ws 0dTo formula. The same episode is related by Nestor to 
Patroclus in A 765 ff., but he reports the parting speech of Menoe- 
tius to Patroclus, following it with the same transition as that used 
by Odysseus in I : ws iTrereXX' 6 yepoov, av U Xrjdeai. Again the verbs 
refer to times separated by a long interval. But although there is in 
these passages, as in t, a long interval between the speech and the 
event which is next referred to, the cases are not strictly parallel. 
The speeches in I and A are quoted within other speeches,® and we 
are specifically told before the speech, in each passage, that it was 
made long ago, fifxaTi. tc3 ore a' eK^dirjs'Ayan€iJ,vovLT€nT€;a,nd,most 
important, the present tense of X-qdeai, makes the sequence clear. 
There is no abruptness here, no possibility of confusion, as there 
would be in r if Homer had said, "So he spoke, and Odysseus came 
that he might give him glorious gifts" (e.g., ws (t>aT'- 'OSvaatvs 5' rjXd' 
ha ol TTopoL ayXaa Sccpa). Because in lines of this type elsewhere in 
the poems the action spoken of as following the speech always fol- 

* Cf. Wahmer, 1, p. 27, "Ueberhaupt hat Homer Inreden anders — soil ich 
sagen freier? — behandelt als die Reden, die er in seinen Namen berichtet. Ich 
verweise auf a 40 (plotzIicherUebergang der obliquen in die direkte Rede), ;u 165 
(ungewohnlicher Redeabschluss), 389 (desgl.), p 147 f. (Asyndeton), cr 271 
(ungewohnlicher Abschluss), B 70 (Rede und Abschluss im gleichen Verse), 
330 (ungewohnlicher Abschluss wie o- 271), n 842 (ungewohnlicher Abschluss), 
T 130 (ct)s elirojv nach obliquer Rede), usw." 



48 University of California Publications in Classical Philology 

lows it at once, almost inevitably a listener would at first have had 
that impression here, and the fact that a long interval separates 
speech and act would have been obscured for hun. 

In two other places in the Odyssey the time element seems to have 
influenced the poet in the omission of the &$ (i>aTo formula after a 
speech. The two situations are the same: Philoetius and Eumaeus 
are with Odysseus, who is still in the guise of a beggar. The beggar 
speaks of Odysseus' return, whereupon Philoetius prays that Odys- 
seus may come home. Both times after the prayer instead of a 
regular cis c^aro formula there is the line, ws b' avTws Eujuatos kirev^aro 
[eTrevx^To (p 203] Trdai deotai (u 238 and 203). Here again it seems 
likely that the ws ^Lto formula has been omitted because of the 
time sequence which that formula implies, and that because of its 
omission here we are to think of the prayers of Philoetius and 
Eumaeus as simultaneous rather than consecutive.' Both were evoked 
by the one speech of Odysseus. The absence of the ws c/xiro formula 
here may also be due in part to the fact that the line which is used 
after Philoetius' speech is very similar in form to a line beginning 
with a &s 4>a.To formula, and might be considered a virtual equivalent.* 

Some confirmation of the idea that the ws 4>aT0 formula is omitted 
in these two passages because the poet conceived of the prayers as 
simultaneous and not consecutive is to be found in three other 
places where pairs of speeches are reported. These differ from the 
instances just discussed in that in them both speeches are reported 
and the line with which we are concerned occurs between the two. 
In i3 323 ff . the suitors begin mocking Telemachus and we are given 
two speeches, each attributed to a nameless "somebod^^" The 
speeches are separated by a single transitional line which contains 
no ojs 4)aT0 formula, "AXXos 5' aur'etTreo-Ke vkoiv vtreprfvopiovTcov. There is 

' Note the imperfect in <^ 203. Cf . also such passages as H 430 and K 25 (no 
speeches are involved in these), where ws 8' avrws also introduces action simul- 
taneous with what precedes. Contrast, however, y 64; here there are two prayers 
and the second is introduced by cis 5' aurcos. But here the former prayer is fol- 
lowed by a cis <f)dTo formula, and this and the line which follows show that the 
prayers were consecutive. 

8 Bdrard (Introduction a. VOdy/^sie, pp. 120 ff.) discusses some of these pas- 
sages in which the ojs (t)aTo formula is omitted. His explanation of </> 203 is as 
follows: ". . . la formule, sans ctre entierement conforme a la coutume hom^r- 
ique, s'en rapproche n^anmoins par le wj . . . du debut." v 238-239, according to 
B6rard, have been introduced by "quelque reminiscence de rhapsode ou 
d'6diteur." The true transition is in 240. 



Combellack: Omitted Speech Formulas in Homer 49 

no ojs <l)aTo formula because the two speeches are not thought of as 
consecutive but as merely typical of many such remarks which were 
being made more or less simultaneously. The same situation and 
technique are found in 396^04 and P 414-423. 

Somewhat similar to these are two passages in the Iliad, B 188- 
207 and A 232-250, where a leader (in B Odysseus, in A Agamem- 
non) is urging on his troops, and the poet gives us two speeches 
typical of a number which he made. In B 188-197 there is an ac- 
count of Odysseus' speech to the leaders, and in 198-206 the account 
of his words to the rank and file. Between the speeches there is no 
ws (f)aTo formula, but instead the lines : 

"Ov 5' av drjuov r' iivSpa Uol ^oouura r' k4>ivpoi, 
Tov aKrj-KTpi^ (XaaaffKev bjj.0KKy](ya(rKk re jjLvdco' 

Here again it seems probable that the cos ^aro formula has been 
omitted because the two speeches were merely typical of many 
(note the iterative verbs) which were made on this occasion and did 
not follow one another in regular sequence as a us <l)aTo formula 
would have implied. The same technique is followed in the other 
passage of this type, A 232-250. 

Finally among these passages where there is no speech formula 
and the time sequence is peculiar, let us consider the much disputed 
Helios episode in n 374-390. » The authenticity of this episode has 
been questioned from the days of the Alexandrian scholars, and it 
is usually considered by modern critics to be an interpolation. It is 
perhaps worth noting that, whether the advocates of interpolation 
realize it or not, their theory must assume that the interpolator of 
the Helios episode, either because he had a subtle appreciation of 
the force of the &s (paro formula and realized it could not properly 
be used before his interpolated lines, or because he wanted to make 
it clear to later students of Homer that the passage was an interpo- 
lation, removed a regular &s 4>aTo formula after Odysseus' words 
when he put the Helios episode into the text. The omission of the ds 
^aro formula in the text as it stands can be explained, but there is 
no possible explanation, so far as I can see, if 391 follows imme- 
diately after 373. 

^ Berard's explanation for the omission of the ws 0aro here in ix is the usual 
one — the episode is an interpolation. 



50 University of California Publications in Classical Philology 

In n 371-373 Odysseus cries out to the gods and there is no speech 
formula after his words but only the abrupt statement, ajxea 5' 
'HeXtw 'TTreplovL ayyeXos rj\de. Again, at the end of the episode when 
Zeus has finished speaking there is no speech formula but simply 
Odysseus' apologetic explanation of how he is so familiar with this 
OljTnpian scene, raura 8' eYcbj" riKovcra Ka\v\povs rimofiOLo. I believe 
there is no speech formula at the beginning or end of the Helios 
episode because the episode is simultaneous with the awakening of 
Odysseus and his crying out to the gods, and takes place elsewhere. 
The "march of time" is resumed at 391, where we return to Odys- 
seus and his actions. 

This is the only passage in Homer where the poet describes action 
simultaneous with the actions and words of a single person who is 
himself telling the story. The plural formula, &s ol ixev rotaOra wpos 
a\\r]\ovs ayopevov, which we found applicable in E, and which is 
used in many other passages to make the transition to a simulta- 
neous episode, is of course not available after the single speech of 
Odysseus in n. While the situation in E is rather common in both 
the Iliad and the Odyssey, Homer almost never, so far as I have ob- 
served, uses a speech formula in the singular as a transition to an 
action which, although simultaneous with the speech, takes place 
elsewhere and involves some other person or group. ^^ Perhaps the 
only time he does this is in p 505 : 

17 fikv ap' COS ayopeve fieTO. SuCf^ai yvvai^lv, 
rifiePT] ev daXanco' 6 8' eSetTiTet 5Tos 'OSvaaevs' 

Here, as the nkv and the 8e, and the imperfects show, Odysseus' din- 
ing in the great hall takes place at the same time as Penelope's talk 
with her servingwomen. But this form of expression, too, could not 
be used in n, since besides its general unsuitability it is in the third 

'*• One formula in the singular which is followed bj' an action simultaneous 
with that designated by the formula is fjos 6 raW' upfiaive /card 4>peva Kai Kara Ovuov. 
This sometimes follows direct quotation, but it is an expression of very limited 
usage and quite inappropriate here in n. Usually after it some person (or per- 
sons) approaches the character who has been pondering, e.g., A 193. Cf. A 411, 
P 106, 2 15, 5 120. More rarely someone or something does something to the 
person who is pondering, e.g., e 365. Cf. « 424. In the similar formvila in K 507, 
while Odysseus is pondering, Athena comes to his associate Diomedes. In short, 
the simultaneous action here is one in which the speaker is immediately con- 
cerned, and there is no change of scene. All the situations in which this formula 
occurs are obviously unlike that in n. 



Combellack: Omitted Speech Formulas in Homer 51 

person, and in )u a formula in the first person would be required. It 
seems to me that four possibilities presented themselves to the poet 
in n'. 

1. To follow the most common technique after speeches, that is, 
to put after Odysseus' speech a &s 0dro formula and proceed at once 
to the Helios episode. But this would indicate the usual sequence of 
speech and action and would compel the poet to renounce the idea 
of making the scenes contemporaneous. 

2. To create an original line to fit this unique situation, which 
would make it perfectly clear that the Helios episode was simul- 
taneous with the words and actions of Odysseus. No one who has 
followed Parry's studies will be greatly surprised that he did not 
choose this method, which seems so natural and obvious to us. 

3. To imitate the technique of H 344-345. Here is a speech by 
Nestor which Homer closes with a regular &s 4>a.To formula and then 
tells us of the audience's reaction, 

ds €(t)ad', ol d' apa irauTes eTrfivr]crav ^aaiKfjes' 

Then, in the next line, he begins an account of the gathering of the 
Trojans, which probably is to be thought of as taking place more or 
less at the same time as the gathering of the Greek leaders addressed 
by Nestor. Although, therefore, the words which immediately follow 
a ws (/>dro formula designate an action which either follows the speech 
at once or occurs at the same time and in the same place as the 
speech. Homer may, after making clear this temporal and local re- 
lationship between the speech and the immediately following nar- 
rative, then retrace his steps and describe an event which occurred 
at the same time as the speech but in another locality. So here in /x 
it looks as if it would have been easy to conclude Odysseus' cry with 
a regular ws (t>aTo formula, indicate in the rest of the line the action 
v/hich immediately followed the cry, and then begin the Helios 
scene, which might well be thought of as simultaneous with the cry. 
By imitating the technique of 1 177 and k 274 it would even be pos- 
sible to avoid the comparatively long and difficult &s k4>anriv, e.g., 
&is etTTOJj' a.\p rj\9ov kfi-qv eirl vrja neXaivav. wKea 8' 'HeXtoj ktX. But although 
this seems easy enough to us we cannot be sure that it would seem 
easy to a poet composing in formulas. The technique of H 344 f . is 
extremely rare, and the formulas used there are in the third person 



52 University of California Publications in Classical Philology 

and before the technique can be adapted for use in /i formulas in the 
first person must be found. Speech formulas in the first person are 
comparatively uncommon in Homer and not one of the Homeric 
lines containing them could as it stands be used here in ju. The diffi- 
culties of adapting the unusual technique of H to the unique situa- 
tion in iJL apparently seemed too great to the poet and caused him to 
reject this solution. 

4. The only remaining possibility was to put no speech formula at 
all after the words of Odysseus; and this was the procedure adopted. 
We have seen from the examination of a number of other passages 
that it was the frequent practice of the poet to omit the speech for- 
mula when the time sequence after speeches was unusual, so that in 
adopting this fourth possibility here in n he was actually employing 
a technique used under similar conditions elsewhere. 

Because of the time sequence the poet could not use a ws 0dro 
formula after Zeus' speech at the end of the episode; if he had used 
a singular formula, it would have implied that Odysseus heard 
about the Olympian conversation just after it took place; if he had 
used the plural formula, us ol nep TOLavra kt\., it would have im- 
plied that he heard of it from Calypso while it was taking place; 
actually, of course, he heard of it long afterwards. 

Also, in this passage, another factor may have had some influence 
in moving the poet to omit a ws ^aro formula. It is worth noting that 
the line he does use, raOra 8' eychv fiKovcra Ka\v\povs rivKonoLo (389), is 
very similar to such lines as avrap eirel rb 7' aKovai Yioaabawv 
kvoalxdwv (v 159), in which a verb of hearing instead of a verb of 
saying is used after a speech. This may well have influenced the 
technique here, even though line 389 does not refer in particular to 
the speech which immediately precedes it, but to the episode as a 
whole. For that matter, the common formula cos at fxev Totavra irpbs 
a\\r}\ovs ayopevov also refers to the scene which it closes rather than 
to the speech which immediately precedes it. 

Another factor which has sometimes impelled Homer to omit the 
a)s 4>a.To formula where it might be expected to occur is the desire to 
contrast persons or groups. Usually if there is a contrast after 
speeches one of the two persons contrasted is the speaker. In this 
type of situation Homer regularly uses a speech formula, and in fact 
had a fairly extensive group of formula types from which to choose, 



Combellack: Omitted Speech Formulas in Homer 53 

e.g., A 210-11,0 405-6, 559. But when the poet wishes to contrast per- 
sons other than the character who has just spoken, the speech 
formula becomes undesirable because it inevitably brings the speaker 
prominently on the scene and thereby weakens or destroys the con- 
trast. In the few places, therefore, in which Homer wished to 
contrast the actions of two persons or groups other than the speaker, 
the speech formula is omitted. As it happens, this situation and this 
thoroughly logical treatment of it are found after the very first 
speech of the Iliad. In A 17-21 Chryses beseeches all the Achaeans, 
but especially the two Atreidae. After his words, instead of the 
regular ws (l>aTo formula there are the lines : 

"Evd' aXXoi nev rdi'res eTrev4>r]iJLr]aav 'Axatot 
aiSeiffdai 6' leprja Kai dy\aa Sex^at airoLva' 
dXX' ovK 'ArpetSrj 'Ay a ixefxvovL ^v5ave OvuQi. 

Here the poet wished to emphasize the contrast between the two 
reactions, and the regular ws ^dro formula was not suitable since it 
would have brought Chryses into prominence. 

The remaining few instances of this technique are in the Odyssey. 
Two are almost the same — a 360 and <^ 354.^^ In both these passages 
a bold speech of Telemachus is followed not by a ds (f>aTo formula, 
but by the line, 'H jiev dan^rjaaaa iraXiv oLKOvde ^(^rjKH. In a, 17 fiep is 
contrasted with fxvrjaTrjpes 8k a few lines farther on, and in <^, 17 fxh 
is contrasted with avrap 6 in 359. In a number of manuscripts have 
immediately after Telemachus' speech the line, cos 0dro' plyrjaev 8k 
irepl<i>puv UrjvekoTraa.This was doubtless inserted by some reader who 
did not realize that he had here an example of a rare but quite na- 
tural Homeric usage. 

A similar but slightly different situation occurs in x 200,^^ where 
after Eumaeus' mocking words to Melanthius there is no cis (/>dro, 

1' According to Berard, the text in (j> 354 is wrong and the best remedy is to 
omit 354-355 and substitute the hne, cos ^aro* piyj^aev bk wepi<i>pui' HrjuiXoireia, 
found in some manuscripts. The incorrect transition, he says, was first put in 
at a 360-361 by "I'auteur recent" of that passage, who thought he had the 
same type of transition in a 319. From a it was copied into the parallel passage 
in </> "grdce a cette manie de I'unification dont les editeurs antiques furent sou- 
vent les dupes." 

^^ Berard's comment on this passage is: ". . . ici encore ws au debut du vers 
nous rapproche un peu de la coutume homerique." 



54 University of California Publications in Classical Philology 

but instead the action, or rather the enforced inaction, of Melan- 
thius is contrasted with the action of Eumaeus and Philoetius : 

'Us 6 nev avdi, XeXetXTO, radels oKoui hi Secr^iqj' 
TOJ 5' es revx^a Svvre, dvprjv kiridevTe (j>aeLVT]v, . . . 

Here Homer wished to contrast Melanthius, on the one hand, with 
Eumaeus and Philoetius, on the other; that is, although the speaker 
(Eumaeus) is here involved in the contrast he is not alone con- 
trasted with someone else, and a regular ws (paro formula could not 
be used. 

Still more unusual is tt 336 ff., where Eumaeus and a herald both 
come to tell Penelope that Telemachus is back. The herald speaks 
first and there is no speech formula after his speech. It has been 
omitted, I suspect, because the poet did not wish simply to contrast 
the herald and Eumaeus, but rather to contrast the bare news of 
Telemachus' return announced by the herald openly to Penelope in 
the midst of her handmaids, with the detailed account of Tele- 
machus given by Eumaeus to Penelope alone. The poet may have 
felt that the force of this contrast would have been weakened by a 
ojs <f>aTo formula, since it would have thrown too much emphasis on 
the unimportant person of the herald. 

In a small group of passages, all very similar, no ws (t)a.To formula 
is used after a speech simply because it would be too clumsy. In 
three. Homer inserts a hypothetical speech by a "somebody" with- 
in another speech in such a way that the words of the "somebody" 
form part of a purpose clause. In M 317, for example, Sarpedon 
urges Glaucus to fight among the foremost, 6<j)pa ns dJ5' dirn AvKiuiv 
irvKa doiprjKTauv. After the imaginary speech no ois </>aTo formula is 
used or very well could be used. The same situation and technique 
are found also in H 300 and ^ 575. (The speech formula in H 303, of 
course, refers to the speech of Hector and not to that of the "some- 
body," which just happens to end Hector's speech.) Quite like these 
is Z 476 ff . Here Hector prays for a glorious career for Astyanax and 
says, inter alia, Kal Tore ns ctTrot WaTpos y' 68e toWov aneivo^v', and 
no (is (^aro formula follows this short speech.'^ 

^^ Regularly, when there is no special reason for omitting the (is (paro formula, 
it is used at the end of even these hypothetical remarks which are quoted with- 
in speeches, e.g., A 182. Even when the hypothetical remarks occur at the end of 
the speech within which they are quoted, Homer regularly puts a ws (pdro after 



Combellack: Omitted Speech Formulas in Homer 55 

A third type, similar to those just mentioned in that it also occurs 
in speeches quoted within other speeches, is illustrated by t 288 ff. 
Here Odysseus tells Telemachus what to say to the suitors if they 
object when he puts away the arms, introducing his words with an 
imperative, or, rather, an infinitive used as an imperative — /xaXaKoTs 
eTre€(TaL/Trap<pd(rdai,. For this unusual situation there was apparently 
no formula available for use after the speech. The same instructions 
are repeated in r 5-13. Here they occur at the end of Odysseus' 
speech, and the regular ws 4>aTo formula which follows them refers, 
of course, to Odysseus' speech as a whole, and as in tt there is no 
formula after the speech which he directs Telemachus to make. 

This leaves us with only one passage in Homer where the omission 
of the speech formula calls for special comment, X 498. Andromache, 
picturing the melancholy lot in store for the fatherless Astyanax, 
quotes the insulting words which a child who has both parents hv- 
ing will address to him : 

epp' ovToos' oil aos y€ warrip neTaSalvvTai, rifxlv. 
There is no cos ^dro formula after this hypothetical speech. This 
does not clearly fall into any of the various classes which have been 
considered, and it is difficult to see why the formula is omitted. 
Perhaps we must conclude it is simply an instance of Homer's 
somewhat freer usage in quoted speeches remarked by Wahmer (cf . 
note 8 above). I suspect, however, that this falls into the group of 
passages in which the ws (paro formula is omitted because it would 
be too clumsy. If used here it would interrupt the flow of Andro- 
mache's remarks, and, by throwing too much emphasis on the 
a.ii4>Ld a\T]s, would to that extent detract momentarily from the im- 
portance of Astyanax and thereby weaken the pathos of the passage. 

In the later Greek epic poets there are also a few places where 
the ws 4>aTo does not occur after speeches where we should expect it.^^ 

them and then follows this with another to mark the end of the speech within 
which the hypothetical words occur, e.g., H 91-92, 

ois TTore Tis ipiei' to 5 kfibv KKko^ ov ttot oXetrai. 

COS i(t>ad , ol d dpa iravTfs Aktiv kykvovro aiuwfj. 
^* The outstanding difference between the regular technique of Homer and 
the later epic poets is that Homer uses a cos <t>aTo formula in dialogue only in the 
unusual circumstances mentioned above in note 3, whereas the later epic poets 
seem to have felt that a ws 0dro was necessary after all speeches, whether in 
dialogue or not (cf. Wahmer, I, pp. 23-26). This change in technique appears as 
early as the Hymn to Demeter (74 and 118) and Hesiod's Theogony (664). 



56 University of California Publications in Classical Philology 

It is interesting to note that many of these follow the Homeric 
usage. Nonnus, for instance, has a number of speeches which are in- 
troduced by subordinate clauses, and there is never any us 4>o.to 
after them; also, like Homer, he does not use any ws <t>aTo after 
speeches introduced by imperatives. Speeches of these two types 
are extremely rare in Greek epic poetry outside of Homer and 
Nonnus. I have noted none in either Apollonius or Quintus. The 
Hymn to Apollo, however, has one of each: introduction with im- 
peratival infinitive, 171, introduction in subordinate clause, 167. 

In Apollonius there are two passages in which there is no ws ^dro 
after a speech, but in which there is a contrast between the actions 
of two persons or groups: ii, 262 f., and iii, 740 f. The former of 
these can be justified by Homeric precedent, because the speaker is 
not one of the persons contrasted after the speech; but the latter 
passage is quite unlike the Homeric usage, since the speaker is one 
of the two persons contrasted. The technique here differs further 
from that of Homer in that Apollonius tells us first of the action of 
the person addressed and then returns to the speaker. 

Among other passages for which there are no Homeric parallels 
may be mentioned the following: once in Quintus (iv, 408) and 
rather often in Nonnus there is no ws 0dro after a very short speech. 
Sometimes a speaker in Nonnus introduces into his remarks a hypo- 
thetical objection of his listener, using the expression, 'aXX' epkets, 
and no ws </)dro is ever used after this type of hypothetical speech. 
Some of Nonnus' more unusual speeches have no us 4>aT0 after 
them; for example, the speech which a character on a shield seemed 
to be making, xiii, 221 f. 

The following conclusions can now be drawn : 

1. Homer has a clear and regular technique for the use of the ws 
(paTo formula. 

2. All the apparent exceptions to this regular technique can be 
explained as situations in which the available formulas were inap- 
propriate, and critics would therefore seem to be wrong in regard- 
ing them as evidence of textual corruption such as interpolation, 
lacunae, and so on. 

3. We have here important evidence of the integrity of the textual 
tradition as far back as the times when the technique of speech 
formulas was clearly understood and consistently observed. 



SENECA'S EPISTULAE MORALES 

THE TEXT EMENDED AND 

EXPLAINED (I-LXV) 

BY 
WILLIAM HAKDY ALEXANDEE 

I, 5 (2 : 13) :^ non puto pauperem, cui quantulumcumque superest, 
sat est : tu tamen malo serves tua, et bono tempore incipies. nam 
ut visum est maioribus nostris, sera parsimonia in fundo est. 

Punctuate with a full stop after serves tua; "it is true that a man 
is not poor who finds a sufficiency in what he has, however little it 
may be, but for all that I prefer to have you keep what is (now) 
yours." Similarly et hono tempore incipies goes with what follows : 
"and you will get about the business (of keeping it, i.e., of being 
prudent and economical) at once, because, as our ancestors saw, 
thrift is late which begins at the bottom of the jar." Punctuate 
with a comma or semicolon after incipies; this future indicative 
carries a certain amount of imperative force, quite usual in familiar 
language. 

I am unable to follow Beltrami in his handling of this passage as 
contained in his text (Edizione Nazionale^) and in his critical note 
on it. The tendency of the users of mss. to write headings in the ms. 
margin, and the subsequent transference of these headings into the 
text, is only too well known. 

Ill, 3 (5 : 20) : nam quidam fallere docuerunt, dum timent falli, et 
illi ius peccandi suspicando f ecerunt. 

The competent mss. vary between illius (so also Q) and illi ius; 
this may be treated as a unanimous testimony to illi ius. Yet Bel- 
trami (II), usually so devoted to Q readings, deserts here in favor 
of alii for illi, and Hense (II) gives alii at least a partial blessing. 

^ The numbers in parentheses give the page and line reference in Hense's 
second edition (Leipzig, Teubner, 1914), hereafter referred to as Hense (II). 
Hense (I) appeared in 1899 under the same auspices. 

- Eome, Regiae oflficinae polygraphicae, 1931. This edition will hereafter be 
referred to as Beltrami (II). Beltrami (I) was published as follows: Vol. I, 
Brescia, Apollonia, 1916; Vol. II, Bologna, Zanichelli, 1927. 

[57] 



58 University of California Fuhlications in Classical Philology 

Aliis was the vulgate reading from the sixteenth century on, and 
of course aliis implies also the possibility of alii. Schweighaeuser" 
adopted it in his text, but Fickert* (app. crit. ad loc.) records his 
subsequent regrets: poenituit Sw. non restituisse scripturam, 
quam ad amicum referri bene perspexit. This is entirely correct. 
Ille, referring to one's friend, occurs p. 5 : 15 {ilium), 5 : 16 (illo), 
is made explicit by amico (5 : 19), again by amico (6:2), and re- 
sumes with illo (6:3). There can be no misunderstanding of illi 
under the circumstances, and no emendation of any kind is re- 
quired beyond the obvious illi ius. 

VI, 7 (13 : 19) : interim quoniam diurnam tibi mercedulam debeo, 

quid me hodie apud Hecatonem delectaverit dicam. "quaeris," 

inquit, "quid profecerim? amicus esse mihi coepi." multum 

prof ec'it : numquam erit solus, scito hunc amicum omnibus esse. 

It seems to be generally agreed that the last sentence means : 

"You are to know that he (Hecato) is the friend of all." I confess 

that I do not see the bearing of this on the context, and it appears 

to me a particularly flat finish to the letter to say that Hecato is a 

friend of all, presumably for having discovered the possibility of 

friendship with one's self. 

The sense may rather be paraphrased thus: "Hecato made a 
remarkable advance by realizing that one's nearest friend may be 
one's self ; he will never be lonely. But the experience is not con- 
fined to Hecato ; it is equally open to everyone to enjoy. You are to 
realize that this friend (viz., one's self) belongs to all of us as well 
as to Hecato if we only cultivate him." 

Since my comment on this passage was written, my attention has 
been directed to the note of G. Hess on p. 18 of his edition of Ejjist. 
Mor. Selectae (1890) on the word omnibus: von esse, niclit von 
amicum abhangig. That is the point precisely. 

VIII, 4 (18 :4) : delude ne resistere quidem licet, cum coepit trans- 
versos agere f elicitas, aut saltim rectis aut semel mere : non vertit 
fortuna, sed cernulat et allidit. 
Unless one happens to have examined the various older editions 

^ L. A. Senecae ad Lucilium Ep. Morales, ed. Johann Schweighaeuser (Strass- 
burg, Bipontine, 1809). 

*L. Annaei Senecae Opera, ed. Carolus Eudolfus Fickert (Leipzig, Weid- 
mann, 1842-1845). Vol. I, the Epistles, 1842. 



Alexander: Seneca's Epistulae Morales (I-LXV) 59 

of the Epistles, he could have no idea of the extraordinary vari- 
ations with which this passage has been printed — so extraordinary 
that it is sometimes difficult to believe that it is the same passage 
that is in question in all of them. The fact is, of course, that as it 
stood, it so baffled the editors that they reworked it according to 
their fancy to give what they considered a plausible text. However, 
as Haupt (Opuscula, II, p. 318) pointed out, Opsopaeus long since 
saw the true sense of the locus. I quote in part from his note : 
"Elliptic ws verba erunt intelligenda quasi dicat Seneca aut rectis 
velis cursum tenere oportet (nam quos transversos agit felicitas, 
ii resistere nequeunt) aut semel ruere, hoc est perire."^ No one 
will now question that the clue for aut saltim rectis aut semel ruere 
lies in the famous defiance of Neptune attributed to the Rhodian 
pilot ; the passage in Haupt's Opuscula mentioned above gives all 
the necessary references, which need not be repeated here. 

But even thus the difficulties are not quite ended; what is the con- 
struction of the proverb aid . . . ruere in the sentence ? What syntac- 
tical relation does it bear to the preceding words, and how is it 
formally connected with those words ? (1) It might be thought that 
the first aut connected what follows it with the preceding sentence ; 
ne . . . quidem . . . aut is found, for example, in Vergil, Georgics I, 
126. But the doubled aut in the proverb seems essential to express 
the mutually exclusive character of the alternatives. (2) It is con- 
ceivable that a <neque> or <nec> to follow ne . . . quidem has 
been dropped at some time in the ms. tradition by a scribe who was 
puzzled by <neque> aut or <nec> aut. (3) It is, however, quite 
possible, and seems on the whole the most reasonable conclusion, 
that we are here dealing with a simple case of asyndeton ; ruere is 
in the same construction as resistere, and the general force non 
licet is transmitted to it from the ne . . . quidem licet preceding. 
Translate : "It's not allowed even to stop . . . (it's not allowed) 
'either with sails full set (to make harbor) or to go down once and 
for all'." 

It is in reference to these last words only, this latter hope ex- 
pressed in aut semel ruere, that the following words non vertit . . . 
allidit apply. Your hope is to sink and be done with it, but Fortune 
doesn't oblige you by capsizing you once and for all ; she keeps turn- 

6 Taken from the Variorum edition printed by Jacobus Stoer, 1607 (place of 
publication not given). 



60 University of California Publications in Classical Philology 

iiig you over, stem over stem, and pounding you on the rocks. I 
prefer, with Madvig and Hense (II) in app. crit. ad loc, the read- 
ing of the editio Mentelini evertit as in Cicero's naviculam evertere 
(De Or. I, 38, 174) ; yet I think that vertit can stand in much the 
same sense (Lewis and Short, s.v. I, B, e) . 

The ms. is sound as explained above, and needs no change. 

IX, 4 (20: 23) : sed quae si desunt, non desiderat, non deesse 
mavult. 

The mss. consensus is : sed quae sihi desunt non desiderat non 
deesse mavult. Haase :^ desiderat deesse, followed by a colon; 
si for sihi, as above, rests on a conjecture of Buecheler's; Beltrami 
(I) : quae si sihi desunt.'' 

Read : sed quae s <0 ihi desunt, etc. The ihi refers very clearly 
to the imminuto corpore et amputato of the preceding sentence, 
having the sense it so often and so naturally has of in eo. 

XIII, 14 (38 : 9) : pudet me ibi sic tecum loqui et tam lenibus te 
remediis focillare. 

S. Linde^ defends the ms. reading here ; "plane recte se habet ; 
significat enim 'in eo, in eis rebus.'" I have examined the passages 
which he cites (Plautus, Epidicus,4:, 2, 24 ; Cicero proEosc.Amerin. 
29, 82 ; Sallust, Cat. 5 and Jugurtha 58 ; QuintUian, 7, 1, 6 ; Seneca, 
De Benef. 3, 38, 3), and in all of them without exception the ihi is 
specific in its reference to a definite something which has immedi- 
ately preceded and has been expressed in a perfectly definite word 
or phrase or clause. On the contrary, the reference of ihi in the 
present Senecan passage, assuming the passage to be sound, can 
be only to the whole run of the letter from section 3 to this point. 
While it may be too much to say that ihi cannot have so extended 
a reference, as Beltrami (II) so easily assumes it may, it is certain 
that such a use is not established by Linde's references. I am still 
skeptical of the soundness of ihi in this passage. I am inclined to 
believe that it originated as a marginal gloss thus. Opposite sic 

* L. Annaei Senecae Opera quae supersuni, ed. Fridericus Haase (Leipzig, 
Teubner, 1872). In three vols. ; the Epistles in III. 

'' But Beltrami (II) expresses in his critical note a preference for si illi, and 
prints his conjecture in the body of his text. 

® Adversaria in Latinos Auctores, Lund, 1900. 



Alexander: Seneca's Epistulae Morales (I-LXV) 61 

loqui a user of the ms. made a reference mark of some kind, perhaps 
a hand with index finger pointing to a passage above to which he 
thought sic loqui had special reference, and wrote the word ihi, 
which subsequently became incorporated into the text. 

XIV, 13 fin. (43 : 2) : cum modo per populi levatus manus et 
obrutus sputis exportandus extra forum traheretur. 

The ms. consensus (only slightly qualified in L) is etportandus, 
for which Pincianus suggested the easy correction exportandus. 
But there is a clumsiness about the participle construction levatus 
. . . ohrutus . . . exportandus which creates a doubt of the soundness 
of the correction. Thus 0. Rossbach has given us exportantium, and 
later, with much less warrant, et exprohrationihus. Beltrami has 
taken Rossbach's former conjecture into his text. 

I suggest exportantis. The construction is a parallelism, per 
manus populi levatus over against sputis exportantis ohrutus, with 
the basic phrase populi exportantis divided between per manus 
(=manibus) and sputis. The force of the present participle is, 
as often, conative, "seeking to thrust him out (into exile)." 

XIV, 16 (43 : 18) : denique consilium rerum omnium sapiens, non 
exitum spectat. initia in potestate nostra sunt : de eventu f ortuna 
iudicat, cui de me sententiam non do. 'at aliquid vexationis 
adferet, aliquid adversi.' non damnatur latro, cum occidit *** 

Some years ago* I suggested <nisi> before cum occidit. By 
this judgment (which, by the way, M. Albertini reminds me," had 
also suggested itself to Lipsius) I am still disposed to stand, nor 
does my <n?s/> require the cumbersome <Ctum vero^ which M. 
Prechac" would add to it ; mistakes of omission occur often enough 
for which no formal paleographic cause can be assigned. 

In my view, latro is identified directly with Fortuna, and I would 
remind those considering this passage that one feature of the 
"pointed" style and its high condensation is to state the point in a 
figure and leave the reader to make his own application, in con- 
fidence that it will be the correct application. It makes the termi- 

^ Notes and Emendations to the Epistulae Morales of L. A. Seneca, Edmon- 
ton (Univ. of Alberta), 1932, p. 5. 

1" Bevue des Etudes Latines, XIII (1935), p. 46. 

" Quoted by M. Albertini in the preceding reference from the Supplement 
Critique Bude, V (1933). 



62 University of California PuMications in Classical Philology 

nation of the argument here quite abrupt, and one might very well 
conceive of an explanatory continuation such as this : nee f ortuna 
igitur, nisi cum saeviit. This would clarify the idea, but is it always 
the virtue of the pointed style that it should be so clear as to raise 
no difficulty of interpretation ? This short figurative sentence (with 
<nisi> added) is, I believe, the actual conclusion of the whole 
argument from the quid ergo? at the beginning of paragraph 15 on. 
"Do the prudent thing ; keep out of contact with impossible politi- 
cal conditions. It may not save you from disaster, but it is more 
likely to do so than is participation in matters political. To decide 
to abstain is within your power ; what will come of it Fortune only 
can say. But don't condemn her before she strikes ; after all, often 
she does not." Compare in this sense the whole argument of Ep. 13, 
8, which is very like the present passage. 

There remains to be said a word about the clause cui de me sen- 
tentiam no7i do, which has also been felt to be somewhat enigmatic. 
It is to be interpreted closely in connection with the finality of 
Fortune's decisions and with their autocratic character. In the 
Roman way of life there is constantly the desire to obtain consilium; 
the consul puts a matter of policy before the senate, obtains the 
views (se7itentiae) of the senators, and bases his action on these 
opinions, or on the majority of them. Under an autocratic govern- 
ment one person decides. Fortune is the autocrat in this case, and 
alone and of herself makes the decision on the outcome without 
bothering about our consilium. I do not offer any opinion about 
myself for the good and siifficient reason that I am not asked to do 
so." The exceedingly curt and abrupt sentence may therefore be 
paraphrased thus : "getting things launched lies within our control ; 
how they will turn out Fortune alone determines, and I am not 
invited by her to give her any advice about myself.'"^ 

XVI, 2 (48 : 10) : itaque tibi apud me pluribus verbis aut adfirma- 
tione iam nil opus. 

12 For this interpretation compare the very interesting passage Ep. 14, 4: 
aut consilio meo nihil fortuna permittit. 

13 Dr. B. L. Charney, who has made a close study of brachylogy in Seneca's 
Epistles, thinks it possible that the words non damnatiir lairo, cum occidit 
mean : "It is not only when he has killed someone that the highwayman is con- 
demned." In other words, he is in a state of constant reprobation as a bad 
character ; this is only accentuated when he is guilty of a killing. Thus too with 
Fortune. 



Alexander: Seneca's Epistulae Morales (I-LXV) 63 

The actual ms. consensus is aut adfirmatis nee tarn longis, QL 
reading also multum between tarn and longis. A variety of conjec- 
ture has gathered around this passage ; the text printed above is, 
from adfirmatione on, a reading of Madvig's. 

My own proposal, adopting nego for nee from 0. Rossbach, is 
to read for the three words nee tarn longis the three words nego iam 
loqui, which run fairly close to the mss. 

Nego in the sense "refuse," "forbid," is employed by Seneca, 
Dial. XII, 1, 4 (twice in the same sentence), with a dative of the 
person and an accusative of the thing, the latter being, however, a 
neuter pronoun; also Epistles XXIV, 4, reditum suum Sullae 
negavit, with no restricting circumstance, but Seneca's freedom in 
the use of the infinitive as substantive is notorious, and I see no 
reason why a person who could write (XCIX, 2) sibi lugere sump- 
serunt, and (CXVII, 9) sa^jere non nisi bonus habet, could not also 
write tibi nego loqui: "I deny you speech with me." 

There are two interesting parallels in Silius Italicus : (a) 1, 107, 

si fata negarint 
dedecus id patriae nostra depellere dextra, 
haec tua sit laus, nate, velis (velim: Eeins.) 

where nostra dextra implies a dative nobis ; and (b) I, 519, 

nota arma viri corpusque superbo 
victori spoliare negant 

and here we have the complete construction under discussion, the 
dative of the person, the thing expressed as an infinitive, and the 
verb nego. 

It should be noted finally that with the reading proposed we have 
the much admired double cretic ending, popular in the clausula 
since the days of Cicero at least. 

XVIII, 11 (56 : 24) : quanta est animi magnitudo ad id sua sponte 
descendere, quod ne ad extrema quidem decretis timendum sit. 

Deeretis, the ms. consensus, has proved a stumbling block in sev- 
eral quarters. Muretus proposed redaetis, but that is too general ; 
we need a word definitely pointing to judicial action. The same 
criticism applies to Madvig's deiectis. L. Waltz's secretis would 
from most points of view be a beautiful correction, if, once again. 



64 Vniversiiy of California PiiMications in Classical Philology 

the meaning of the word in any relation falling under judicial 
sentence could be sustained. G. Hess's damnatis gives the meaning 
precisely, but does not account for the error. 

An examination of the article damno in the Thesaurus reveals the 
combination decreto damnare exemplified twice in Livy (42, 43, 9 ; 
42, 44, 6), and once in Pliny the Younger (10, 31, 4). The first ex- 
ample from Livy reads : novus deinde praetor . . . capitalis poenae 
absentis eos decreto damnat. I think it is very easy to see how from 
such a phrase the use of decreti to signify persons judicially con- 
demned and thus parallel with damnati in the same sense might 
readily arise as a compendious expression. Decretis, retained by 
Hense (II) and Beltrami (II), should therefore stand. 

XX, 2 (62 : 3) : Philosophia hoc exigit, ut ad legem suam quisque 
vivat, ne orationi vita dissentiat, ut ipsa inter se vita . . . his sit 
omnium actio f dissertionum color sit. 

It would be a long, involved, and thankless task to record all the 
conjectures that have been made on this passage, yet b}^ no means 
made in vain, as it is my opinion that by combining parts from two 
of these we can obtain a highh^ satisfactorj^ text. One thing gen- 
erally agreed on, with one or two exceptions, is that dissertionum 
has intruded itself as a kind of doublet from dissentiat preceding, 
taking over the -7ium of actio, and therefore may be dismissed from 
the inquiry altogether. 

I would myself accept Madvig's a2it for the ut immediately fol- 
lowing dissentiat, punctuating with a semicolon after the vita 
following inter se, thus making a continuous noun clause from ne 
orationi to inter se vita, with dissentiat for the verb of both its 
subdivisions. Then with Kronenberg," anticipated however by 
Fickert, for the concluding clause read: ut unus sit omnium 
actionum color [sit]. This produces a tripartite subordinate sen- 
tence with three noun clauses dependent on exigit. Actiotium color 
also provides a double cretic finish. 

XXI, 10 (68 : 1) : cum adieris eius hortulos et inscriptum f hortulis 
'hospes, hie bene manebis, hie summum bonum voluptas est,' 
paratus erit istius domicilii custos hospitalis, etc. 

I join Beltrami in preferring huius to Usener's eius; it agrees 

" A. J. Kronenberg, Classical Quarterly, I (1907), p. 205, 



Alexander: Seneca's Epistulae Morales (I-LXV) 65 

better with the recurring his of all the ms. readings at this point, 
which probably derives from a contracted form of huius such as his. 
Inscriptum f hortulis suggests very definitely an error arising 
through the displacement of the proper word by a wrongly re- 
peated hortulos. Probably at some point in the ms. tradition two 
successive lines stood thus : 

adieris his hortulos 
te inscriptu hortabitur 

which represents a common line length in uncial and semiuncial 
bicolumnar mss. In the two hypothetical lines just given, the 
lengths will be almost identical if we regard the final -ur of horta- 
bitur as indicated by abbreviation. The scribe's eye wandered back 
naturally enough from hortahitur to hortulos. The faulty text re- 
sulting from the writing in of the latter word rather than the 
former was subsequently ''improved" by giving syntax: to the 
second hortulos through making it hortulis, and altering te to et to 
attach inscriptum to hortulos as the second object of adieris. 

Read : Cum adieris huius hortulos, te inscriptum hortabitur, etc., 
and punctuate with a semicolon or period at the end of the inscrip- 
tion. Hortabitur will be found to be in perfect alignment with 
paratus erit and the other futures of the next statement. I under- 
stand hortabitur in the sense "will encourage you," "will cheer 
you up." 

XXII, 7 (70: 20) : exspectas forsitan, ut tibi haec dicant : 'turpe 
est cedere oneri. luctare cum officio, quo semel recepisti. non est 
vir fortis ac strenuus qui laborem fugit, nisi crescit illi animus 
ipsa rerum difficultate.' 

The translations, as not infrequently, appear to have glided over 
this passage without facing its difficulties ; I refer to the last sen- 
tence, where the comma after fugit in Hense (II) is, at least for an 
English-speaking reader, disastrous. The nisi clause, far from need- 
ing separation from the fugit, requires to be as closely attached to 
it as possible. The sentence should be translated thus : "He is no 
brave and energetic man who shrinks from hard work unless his 
courage increases through the sheer difficulty of the circum- 
stances." It may be paraphrased more happily as follows : "One 
cannot be accounted a brave and energetic person if he drops his 



66 University of California Publications in Classical Philology 

hard task unless he feels an access of courage coining to him pro- 
portionate to the sheer difficulty of the circumstances (of that 
task)." That this is what Seneca had in mind is made absolutely 
clear by the words that follow the anticipated Stoic reproach set 
out above, namely, dicentur tibi ista, si operae pretium habebit 
perseverantia. 

XXII, 13 fin. (72 : 8) : 'cuius V inquis : Epicuri, adhuc enim alienas 

sarcinas f adoro. 

The obelized adoro has been the subject of much speculation and 
emendation, unnecessarily. Why should it not mean colloquially 
"worship" with the sense of "strongly admire" ? To "admire" other 
people's baggage must have been a common experience in ancient 
times precisely as it is in modern, and the phrase is pleasantly 
whimsical for the occasion. Seneca is still "worshiping" Epicurus' 
baggage and "lifting" a number of good things from it. 

0. Rossbach"^ also defends adoro, comparing its use in Suetonius, 
Nero, 12, 3. Striking also is Martial, Epigr. 4, 49, 9, where Martial, 
who is defending the common sense of his branch of poetry as 
against tragedy and epic, receives the reply : 

ilia tamen laudant omnes, mirantur, adorant, 

where adorant clearly expresses as a climax the ideas already sug- 
gested in lauda^it and mirantur. 

XXIV, 26 (83 : 12) : in quod (sc. fastidium vitae) prolabimur ipsa 
impellente philosophia, dum dicimus : 'quousque eadem ? nempe 
expergiscar, dormiam, esuriam, . . . , algebo, aestuabo.' 

There is an obvious gap of one word here to complete the second 
pair. All the proposals of the modern editors and commentators 
run to verbs with a future first singular terminating in -am, on the 
theory that a strict paleographic explanation must be given of the 
disappearance; thus Gertz : edam (before esuriam), Hense: fa^ti- 
diam, Beltrami : farciar, which would require rather more paleo- 
graphic grace than the others. Rossbach alone gives a -ho future 
with his explehor, which is along the lines of the satidbor of the 
inferiores. I must profess my surprise that nobody, so far as I 
know, has suggested : nauseaho; cf . LXXXIX, 22 : dominus crudus 
^Philol. Wochenschrift, XXXIV (1914), p. 493. 



Alexander: Seneca's Epistulae Morales (I-LXV) 67 

et nauseans. Nauseaho is the perfect opposite to esuriam, and I 
think that there is something to be said not only for three balanced 
pairs of verbs, but also for another balance between two sets of 
verb formations. 

XXVI, 3 (86 : 17) : ire in cogitationem iubet (sc. animus) et dis- 
picere quid ex hac tranquillitate ac modestia morum sapientiae 
debeam, quid aetati, et diligenter excutere, quae non possim 
facere, quae nolim, f prodesse habiturus. f ad qui si nolim quid- 
quid non posse me gaudeo. 

Q varies from the above only by atqui for ad qui; whatever con- 
fusion there is, is evidently of long standing. By combining the 
critical notes of Fickert, Hense (II), and Beltrami (II) in their 
app. crit. at this point, a complete view will be obtained of the 
efforts made to set things right here, very labored efforts many of 
them, let it be frankly said. The general disposition is to assume 
that the passage is exceedingly corrupt, but A. J. Kronenberg," 
with his proinde for ms. prodesse and aique (correl. to proinde) for 
ms. atqui or ad qui, suggested that perhaps things were not in such 
bad shape after all. 

Following his line of thought, I offer, with some confidence, the 
following reading : prodesse habiturus, aeque <ac> sinolim, quid- 
quid non posse me gaudeo. The translation of the entire passage 
then works out thus, very suitably for the whole tenor of the letter : 
"My mind bids me take to reflection and descry what of this repose 
and balance of character I owe to wisdom and what to (advancing) 
years, and painstakingly discriminate between what I cannot do 
and what I do not want to do, though it intends to regard whatever 
I am glad that I cannot do as being of just the same value as if I 
did not want to do it." Seneca has just told us that his mind is tri- 
umphant (exidtat) now that old age has forced bodily desires into 
the background, with a corresponding advantage guaranteed for 
mental and spiritual interests. It argues with him about old age 
(facit controversiani) , says that old age is the mind's prime. Very 
well; let it enjoy its advantage (bono suo utatur). How does the 
mind do it 1 It is represented as reducing Seneca to a sound condi- 
tion of humility by reminding him, or forcing him to remind him- 

18 Classical Quarterly, I (1907), p. 205. 



68 University of California Publications in Classical Philology 

self, that a good deal of his assumed morality is only senile debility ; 
this is very humiliating, but the mind really intends to regard 
(hahiturus) the one and the other as equally advantageous; thus, 
despite its apparent grimness as moralist, it allows that in either 
event the result is the same toward establishing the calm of old age. 
The quidquid non posse me gaudeo clause I take to be slightly 
brachylogical for quidquid non possum et gaudeo me non posse. 

On aeque ac si see Stolz-Schmalz (5th ed.), II, p. 784, top. It is 
apparently always ac si (not atque si) in these comparisons of 
equality ; otherwise <aeque> atque would be a very tempting cor- 
rection. 

XXVI, 8 (88 : 16) : conficienda sunt sacra et huic epistulae viaticum 
dandum est. puta me non dicer e, unde sumpturus sum mutuum : 
scis cuius area utar. 

The construction sumpturus sum is intolerable in an indirect 
question ; one should therefore either frankly read sim with the 
codex Guelferbytanus, Muretus, Schweighaeuser, and Ruhkopf," 
or else indicate by the manner of printing the text, as Haase does, 
that unde . . . mutuum is to be regarded as a direct interrogation, 
thus : Puta me non dicere : 'Unde sumpturus sum mutuum?' Scis 
cuius area utar. Here dicere is quite literal. "Suppose I'm not say- 
ing : 'Where am I going to raise a loan V (You feel no surprise over 
that because) you know on whose cashbox I depend." Lucilius cer- 
tainly knows ; cf . XVI, 7, de alieno liberalis sum (with mention of 
Epicurus immediately following), XVII, 11, ab Epicuro mutuum 
sumam, XVIII, 14, delegabo te ad Epicurum, XIX, 10, ab Epicuro 
versura facienda est, XX, 9, pro me dependet Epicurus, XXIII, 9, 
hie est locus solvendi aeris alieni (with immediate mention of a vox 
of Epicurus as the means). The paraphrase of the sense is there- 
fore as follows : "I must put up passage money for this letter. 
Usually under the circumstances a man would be wondering where 
he could borrow the sum. Now suppose you don't hear me say what 
another might well be saying : 'Where can I get a loan of that 
amount?' You are not surprised because you know from several 
• past experiences whose cashbox I depend on." 

"L Annaei Senecae PhilosopU Opera ovinia, ed. Fridericus Ernestus Ruli- 
kopf (Leipzig, Weidmann, 1797-1811; 5 vols.). Vols. II and III are the 
Epistles. 



Alexander: Seneca's Epistulae Morales (I-LXV) 69 

XXIX, 2 (94 : 12) : 'non possum scire, an ei profuturus sim, quem 
admoneo : illud scio, alicui me prof utiirum, si multos admonuero. 
spargenda manus est. uon potest fieri, ut non aliquando succedat 
multa temptanti.' 

Of spargenda manus est much has been said, chiefly through con- 
centrating on the idea of spargo in the sense "scatter, sow" to the 
exclusion of the less common sense "spread" — ^less common, that is, 
from the literal point of view. P. Thomas had already seen this,^^ 
as his excellent note shows : nihil mutandum. Cf . Quint. Inst. Orat. 
XI (wrongly printed X in the journal) , 3, 118 ut . . . temere sparsa 
mami in proximos oifendat. Spargere manum est manum longe 
lateque extendere more agricolae serentis. It will be found that the 
Quintilian passage confirms the meaning "spread," while, as for 
the agricultural method involved, I am able to submit the following 
kind opinion from Dean C. B. Hutchison of the College of Agricul- 
ture in the Universitj^ of California : "While I cannot pose as an 
expert broadcaster of seed, I have had some personal experience and 
considerable observation. This, together with conversation with 
some of my colleagues, leads me to feel that you would be well 
within the bounds of proper agronomic procedure to interpret the 
passage in the Letters of Seneca to Mhich you refer, by assuming 
that in broadcasting, the hand is flung wide open, with the fingers 
fully distended at the conclusion of the action. Otherwise, it is our 
judgment, based as I say upon experience, that an uneven distri- 
bution of the seed would result." 

I think that this topic may now be eliminated from further dis- 
cussions on the text of the Epistles ; the ms. reading proves, as not 
seldom, to be correct when it is adequately understood. 

XXX, 8 (99 : 9) : dicam enim quid seutiam : puto fortiorem esse 
eum, qui in ipsa morte est quam qui circa mortem. 

This sentence has been criticized by Rossbach ;^^ he has reversed 
the positions of qui . . . est and qui circa mortem, and Beltrami (II) 
prints this inversion in his text. With this change I cannot agree ; 
the text should stand unaltered. 

It will be found by examining the whole context that Seneca's 

^Mnemosyne, n.s., XLIX (1921), pp. 25-26. 

^^ Philologische Wochenschrift, XXXIV (1914), p. 494. 



70 University of California Publications in Classical Philology 

view is that, when it comes to downright courage (fortitudo) , e\en 
ignorant and inexperienced people who are at the point of inevi- 
table death are braver in the popular sense than persons who are 
merely on the confines of death, being stiffened up by the dread 
actuality to face the inevitable heroically. The majority of people 
who are on the confines of death (as indeed, who is not?) are 
anything but courageous about it, with one notable exception, the 
sapiens. And his sustaining quality is not the fortitudo which can 
screw courage up to the sticking point for a moment, but a tena- 
cious spiritual strength, which moves calmly, without any very 
high ups or any very low downs, to the appointed goal. 

Thus the sentence puto fortiorem etc. is no reflection on Bassus ; 
he is the sapiens type who possesses the higher quality of lenta 
animi firmitas, not one who shrinks from death when he is circa 
mortem and then develops a sudden courage, very imposing in its 
quality, when he finds himself confronted with the inevitable in 
ipsa morte. Seneca's view expressed by puto fortiorem etc. is, in 
general, as applied to humanity as a whole, quite true and fully 
in accord with experience, but it is made clear later that Bassus 
stands outside and apart from such a verdict in virtue of possessing 
a tenacious resolution that moves equably forward to the end, need- 
ing no special augmentation at that end above what it is capable 
of at any time. It is the exposition of this quality by Bassus, feren- 
tem de morte sentejitiam, that attracts Seneca so strongly. 

XXXI, 11 (105 : 8) : quaerjendum est, quod non fiat in dies eius, 
quoi non possit obstari. 

The immediate sense connection has been this : "Not even beauty 
and physical prowess can make you happy ; nothing related to these 
stands advancing years." Out of this emerges the condition of suc- 
cess in the quest for happiness ; "we have got to look for something 
which does not in course of time become the possession of that 
which cannot be withstood" — as, for example, the advance of the 
years, which obviously cannot be very successfully resisted. I think, 
however, that eius needs definition, and would suggest, though 
without unduly stressing it, that this greater clarity might be ob- 
tained by reading quoi non possit ohstari <rei>, a form of sentence 
occurring more than once in the Epistles. Cf . XXXVI, 5 : ubertate 



Alexander: Seneca's Epistulae Morales (I-LXV) 71 

eius, quam colit, terrae ; LVIII, 19 : ilia, quae me doeet et instruit, 
faeies. 

XXXII, 3 (106: 15) : perdue te in tutum et subinde considera, 
quam pulchra res sit consummare vitam ante mortem, deinde 
expectare securum reliquam temporis sui partem, nihil sibi, in 
possessione beatae vitae positum, quae beatior non fit, si longior. 

This is a passage which has provided material for ample discus- 
sion ever since the days of the revival of learning. To begin with, it 
may be noted that Q has brought us no variation in the ms. reading. 
Muretus struck out nihil sihi in his Roman edition of 1585. Grono- 
vius"" has a characteristically vigorous note in which he says of nihil 
sihi (a propos of Gruter's surprise that Muretus had struck it out) : 
at non est mirandum, cum omni sensu careat. He himself con- 
jectured inniti sihi, comparing XCII, 2, med., tenet summa (sc. 
beatus), et ne ulli quidem nisi sibi innixus, which Ruhkopf (not. 
crit. ad loc.) admires but holds to be unnecessary. His own expla- 
nation of the nihil sihi, applauded by Beltrami, follows in part : 
beatus homo, qui vitam ante mortem iam consummaverit, airadeia 
ista prorsus praeditus reliquam temporis ipsi a natura destinati 
partem exspectet, nihil amplius sibi praestolatur, fato suo terreno 
quasi consummato, and of course if amplius were included in the 
ms. tradition, there would be no difficulty about the meaning. 

I have, however, these observations to make on the passage. In 
the first place, I do not think that reliquam temporis sui partem is 
necessarily the object of expectare; I view it as not improbably 
being an accusative of duration of time with Beltrami (II), who, 
however, expresses this idea duhitanter, and thus adverbial to 
expectare, which finds its sole object in nihil. Secondly, I do not 
think that positum relates to the heatus homo, but to the nihil, 
positum being used in the sense "dependent on, contingent on," as 
frequently in the best Latin. Hence I would place no comma after 
sihi, nor indeed anywhere at all between deinde and positum, trans- 
lating the whole context thus : "secondly, to anticipate carefree 
during the remaining portion of his appointed time nothing for 
himself which is dependent on the possession of the happy life, 
which does not become happier in event of its being longer." The 

^ Ad L. et M. Annaeos Senecas Notae, Johaim Fridericus GronoTius (Am- 
sterdam, Elzevir, 1658), pp. 225-226. 



72 University of California Publications in Classical Philology 

happy life is complete and cannot be made more complete, if such 
a phrase can be tolerated, not even by indefinite extension ; hence 
there is nothing- dependent on or connected with the happy life 
which the sage can possibly anticipate, no matter how long he may 
happen to live after attaining to it. Tenet smnma ; further he can- 
not go. Possessor of the happy life complete, he is carefree in 
regard to any object of ambition whatever, and has nothing to 
anticipate based on that life, because he has it all already. 

XXXIII, 1 (107 : 13) : inaequalitatem scias esse, ubi quae eminent, 
notabilia sunt. 

QL : uhique (by abbreviation of the -que in both cases). Sum- 
mers^ properly, I think, remarks : "but quae eminent must neces- 
sarily be notabilia." In other words, he finds in quae eminent not 
an antecedent condition for notahilia sunt, but a statement identi- 
cal with it. He therefore regards the relation between these clauses 
as being an example of asyndeton, and conjectures <a/?> quae. 
But there is no need for any form change. In the first place, the 
neuter plural forms quae and qua of the indefinite are equally well 
attested.^ In the next place, quae after uhi is a correct though less 
common use of the short-form indefinite following a relative pro- 
noun or adverb ; cf . LXXXI, 31 : prima causa est, cur quis ingratus 
sit, si satis gratus esse non potuit. The text therefore is to be re- 
tained, but quae interpreted as aliqua, and an asyndeton noted be- 
tween quae eminent and notabilia. 

XXXIII, 5 (108: 13) : quare depone istam spem, posse te sum- 
matim degustare ingenia maximorum virorum : tota tibi inspici- 
enda sunt, tota tractanda. res geritur et per lineamenta sua 
ingenii opus nectitur, ex quo nihil subduci sine ruina potest. 

It is quite possible that more is suspected to lurk in this passage 
than ever was actually there ; it may well be that the phrase res 
geritur, which is hardly satisfactory as it stands, can be completed 
with some fairly commonplace epitliet for res. Summers followed 
this line of thought in suggesting (op. cit., fn. p. 33) magna or 

"Select Letters of Seneca, edited by Walter C. Summers (London, Mac- 
millan, 1913). 

=2 Stolz-Schmalz, 5th cd. (Munich, C. H. Beck'sche Verlag, 1926), p. 287, 
«),^)- 



Alexander : Seneca's Epistulae Morales (I-LXV) 73 

ingens, but these are both somewhat vague for the connection.^ 
I suggest res <iintegra'> geritur, without offering any reason based 
on paleographic grounds for the disappearance of <.i7itegra> 
from the text, but reljdng only on the obvious sense of the passage. 
I should regret being led by paleographieal quiddities into any- 
thing like the Re res geritur of Beltrami (II) . 

The Stoic philosophy, incapable of being stated in flosculi, is a 
single, harmonious whole, one and indivisible, and must be studied 
and appreciated as such ; this idea is emphasized in these two sen- 
tences by tota . . . tota . . . ex quo nihil sulduci sine ruina potest. See 
also, a few lines below, the phrase universa fades. Therefore an ap- 
propriate qualifying word for res is <.integra^, and the phrase 
thus supplemented translates : "it is a subject as a whole which is 
being handled," not one where you can quote at random."* 

XXXVI, 5 (114: 2) : denique nihil illi iam liberi est: spopbndit. 
minus autem turpe est creditori quam spei bonae decoquere. ad 
illud aes alienum solvendum opus est negotianti navigatione 
prospera, agrum colenti ubertate eius, quam colit, terrae, caeli 
favore ; ille quod debet, sola potest voluntate persolvi. 

Illud is found in all the mss. Rossbach would, suhduhitans, alter 
it to aliud. Hense (II) writes in his critical note : secludent illud 
fortasse alii, and Beltrami (II) encloses it in square brackets in 
his text. 

All this implies overlooking the force of the sentence preceding 
that in which illud occurs. In the minus autem turpe sentence two 
kinds of debts are spoken of, (1) an actual debt to a creditor, and 
(2) metaphorically, a moral debt to the expectations one has raised. 
To pay the former (illud) one must have a certain amount of finan- 
cial success, to pay the latter (ille quod debet) requires resolution 
and that alone. Hence illud is perfectly correct in its reference, and 
also marks a desirable contrast. It is to be retained. Gummere has 
observed the value of this in his translation of the Epistles in the 
Loeb series (Vol. I, p. 249) . 

^ Yet in fairness it should be noted that in XLIX, 9, we find : ingens nego- 
tum in manibus est, which establishes a certain plausibility for res <.ingens'> 
geritur. 

" Dr. B. L. Charney suggests accepting Georgii's shifting of et to follow per 
lineamenta sua, and then treating per lineamenta sua as applying dTro kolvov 
to both geritur and nectitiir. 



74 University of California Publications in Classical Philology 

XXXVII, 3 (116: 11): 'quomodo ergo' inquis 'me expediam?' 
effug-ere uon potes necessitates, potes vincere. 'fit via vi.' et hanc 
tibi viam dabit philosopliia. 

The vi after via is a suggestion of Gruter's, based on the familiar 
Vergilian line, Aeneid II, 494, but to my mind it is a most un- 
necessary and unwarranted intrusion. The combination fit via is 
made up of perfectly commonplace words which fit admirably into 
the sense as they stand, whereas the Vergilian reminiscence of 
carnage and bloodshed is most inappropriate. The sense is : 'There 
is a way being made (i.e., all the time), and this way philosophy 
will provide you.' Further, if one is to insist on the minutiae of 
paleography, the reading should be, if any addition is desired, -fit 
vi via. It is worth noting too that the moment -fit via vi is adopted, 
the disposition appears at once to change viam to vim. 

XLI, 7 (126 : 10) : vitem laudamus, si fructu palmites onerat, si 
ipsa pondere ad terram eorum, quae tulit, adminicula deducit. 

It should be realized in the first place, before any emendations 
are attempted, that ipsa refers to the vitis, and not, as appears to 
be usually thought, to the adminicula. The iysa means "thanks to 
its own particular excellence," "by the exercise of its inherent 
natural powers." Cf . immediately following : propria virtus est in 
vite fertilitas ; in homine quoque id laudandum est, quod ipsius est. 
Subsequently (line 19) quod ipsius est is explained by quod pro- 
prium hominis est. 

Next, the order of the words is called by Beltrami (II) (app. crit. 
ad loc.) "profecto asperum." It would be more correct to say 
"highly artificial" in reference to the order, arising as it does from 
the interlocking of pondere eorum quae tulit with ad terram 
adminicula deducit in a somewhat mechanical manner. Apart from 
that the order presents no difficulty. It is only fair to add that Bel- 
trami retains the ms. order, with a reference to Scncquc Prosaieur,'^ 
where Seneca's liberties with word arrangement are discussed. 

XLII, 4 (128: 4) : multorum crudelitas et ambitio et luxuria, ut 
paria pessimis audeat, fortunae favore deficitur. eadem velle, si 
t subaudis, cognosces; da posse quantum volunt. 
Beltrami (II) in his critical note assembles, as usual, the history 
^ A. Bourgery, Senequt Prosateur (Paris, 1922). 



Alexander: Seneca's Epistulae Morales (I-LXV) 75 

of the conjectures on this passage. He might in fairness have rec- 
ognized and reported that Fickert nearly a century ago saw that 
si suhaudis was an interpolation, and dropped it from his text, thus 
anticipating Haase'" and indeed Beltrami himself, who in his criti- 
cal note practically admits the soundness of Haase's suspicion that 
si suhaudis is a grammarian's gloss on the difficult use of the im- 
perative da (with the force si dahis) following. Yet Beltrami (II) 
prints in his text yet another addition to the large number of 
guesses already in existence. In my judgment there should be no 
hesitation in ejecting si suhaudis from the text altogether; the 
sentence remaining after these words has been removed is typically 
Senecan in its epigrammatic brevity. 

XLV, 2 (133 :6) : 'vellem' inquis 'magis consilium mihi quam libros 

dare.' 

In the light of the opening lines of this letter our superior mss. 
tradition is palpably wrong, and so Hense (II) (app. crit. ad loc.) 
writes : expecto nollem, while Beltrami inserts <^non> in the text 
before magis. The inferior mss. interchange the places of consilium 
and lihros, and so also Rossbach ; with this I agree. It seems to me 
a most obvious case of the simple transposition of words of like 
grammatical relation within a phrase, namely, the phrase lihros 
mihi quam consilium. While some scribes copied painfully, almost 
litteratim, undoubtedly there were others in the line of the ms. 
tradition who could take in phrases at a glance and proceed to re- 
produce them as phrases. The great danger, of course, about taking 
in phrases at a glance and reproducing them the same way is pre- 
cisely the possibility of an interchange in the word order, as has 
happened here. 

XLV, 5 (134: 4) : tota illo mente pergendum est, ubi provideri 
debet, ne res nos, non verba, decipiant. 

This sentence can easily be passed over casually, but there are 
difficulties inhering in it, at least in the manner of its expression, 
which should be faced. 

Seneca has just been dealing with the verbal traps of apparently 
sound syllogisms, and deploring the waste of any time on them ; 

^ Vol. 3, preface, p. vii. 



76 University of California Fuhlications in Classical Philology 

life has so. many real traps (1.7 : res fallunt). It is against these 
latter we should especially be put on our guard : we do not need 
to worry about words. Hence the form of the statement ne res nos, 
non verha decipiant, which may be paraphrased thus : "against 
our being deceived by actualities, dismissing the matter of verbal 
traps." 

Then again, since these deceitful situations are constantly aris- 
ing, the sooner we are prepared to meet them the better : hence "we 
must devote our entire mental energy to arriving at the point 
where precautions can be taken against our being deceived." Such, 
at least, would seem the natural way of stating the situation, using 
the phrase "precautions can be taken," but Seneca writes debet, 
not potest. I think the debet arises from an extension into the sub- 
ordinate clause of the sense of obligation in the pergendum of the 
main clause. Or, perhaps, one might say that Seneca assumes that 
if one reaches the point where one can take precautions against 
being deceived, it becomes for him a moral obligation to see that 
he is not deceived. 

The sentence could undoubtedly be written more felicitously, 
but there is no doubt of its actual meaning. 

XLV, 8 (134: 19) : ceterum qui interrogatur, an cornua habeat, 
non est tam stultus, ut f rontem suam temptet, nee rursus tam in- 
eptus aut hebes, ut ne sciat tu illi subtilissima collectione per- 
suaseris. 

All the mss. read nesciat; Buecheler divided the word, with the 
result that ut becomes the conjunction for persuaseris and ne for 
sciat. Beltrami (both edd.) reads nescio (dative of nescius). 

Punctuate, however, with a semicolon after nesciat; regard per- 
suaseris as the future perfect in a main statement, as in Ep. 
XIII, 14 : Catoni gladium adsertorem libertatis extorque : magnam 
partem detraxeris gloriae. Here the future is not to be explained 
as merety colloquial ; the sense is : "it will prove to be the case that 
you have robbed Cato of a large part of his glory." So in the present 
passage, punctuated as noted above : "You will find that you have 
persuaded him (merely) with a tricky syllogism (and not other- 
wise)." Note that the word is persuaseris, not demonstraveris: what 
has happened has been a process of so far winning a man over to 



Alexander: Seneca's Epistulae Morales (I-LXV) 77 

your way of stating the position that he could not detect the logical 
fallacy, not of demonstrating to him that you are right and his 
senses wrong. Exactly the opposite, in fact ; knowing that you have 
fooled him somehow, he really sticks by the evidence of his senses. 
This is the interpretation put on the passage by the Elzevir text 
of 1649, in which the last part reads as follows : nee rursus tam 
ineptus aut hebes, ut won habere se nesciat, quod tu . . . persuaseris. 
The italicized words are foreign to the ms. tradition, but undoubt- 
edly clear up the meaning of the sentence. 

XLVI, 3 (137 : 12) : sine me et inquirere. non est quod verearis : 
verum audies. o te hominem felicem, quod nihil habes, propter 
quod quisquam tibi tam longe mentiatur ! nisi quod iam etiam 
ubi causa sublata est, mentimur consuetudinis causa. 

Translators are, on the whole, elusive here. Admittedly there is 
a certain vagueness about the Latin, but it does not, I think, defy 
comprehension and explanation. 

Seneca is going to study more closely this book which Lucilius 
had produced, and when he has done this, the latter need have no 
fears concerning how he is going to be treated; he will hear the 
truth. Now hearing the truth is not what one always gets from a 
friend with whom one is in constant and close contact, as, for ex- 
ample, Seneca and Lucilius would be at Rome. "Lucky fellow you 
in having nothing to cause anybody to lie to you at this distance" 
(distance enhancing as it does the possibility of frankness, and 
there being no other factor in Lucilius' case, such as overweening 
authority, to interfere with frankness and to cause adulation). Cf. 
Lagrange's rendering : "Que vous etes heureux de n'etre pas assez 
puissant pour interesser personne a vous mentir de si loin !"'" Then 
Seneca continues : "But perhaps by now, even when the cause has 
been removed (as in this instance through the fact that we are no 
longer in close and intimate association with one another) , we keep 
on lying through force of habit." 

XLVII, 5 (138 : 21) : [quod] cum ad cenandum discubuimus, alius 
(sc. servus) sputa detergit, alius reliquias temulentorum toro 
subditus colligit. 

^ (Euvres de Seneque traduites en Frangois (Paris, J. J. Smits, An III de la 
Eepublique). 



78 University of California Publications in Classical Philology 

In the first place, I agree with Beltrami (II) in retaining the 
quod, nor is Summers' contention to be sustained that there is any 
lacuna to be assumed before it. Quod would be the natural introduc- 
tory word for the description of each type of humiliating service, 
and Seneca seems to have thought of thus handling the sentence 
to begin with, but the later descriptions became so prolonged that 
the original sentence scheme was changed, the later alius clauses 
becoming separate sentences. Such is my view, at any rate, with 
Hense and Beltrami ; Summers has made a single enormous sen- 
tence out of it all. 

Next, I do not feel the necessity for Rossbach's < tor a > , adopted 
by both Hense (II) and Beltrami (II), nor of any other word to 
localize siibditus. The type of slave mentioned first sputa detergit, 
"wipes off the slaver" from the faces of the guests (surely not 
"mops up the disgorged food" as in the Loeb translation, Vol. I, 
p. 305), and such a slave would be standing. The next slave per- 
forms his unpleasant duty of cleaning from the floor the vomit 
of the drunken revellers, naturally suhditus, "crouched," as com- 
pared with the standing slave who wipes faces. The parallel pas- 
sage in Dial 2, 15, I : recumbere infra mensam vescique cum servis 
ignominiosa officia sortitis, suggests mensae for the place we are 
now discussing if anything (and so Summers in his note on the 
passage, op. cit., p. 211, foot) , but actually in my judgment nothing 
is required to supplement the ms. reading. 

XL VII, 8 (139 : 13) : cum his cenare non sustinet (sc. dominus) et 
maiestatis suae deminutionem putat ad eandem mensam cum 
servo suo accedere. di melius ! quot ex istis dominos habet ! 

The ms. reading hahent was altered by Haase to /ta&ef and the 
explanation of the reading thus created is, no doubt, that given by 
Summers (op. cit., p. 212) : "istis: 'this class,' the slaves. The ianitor 
could be very obnoxious; see e.g. 84.12 below and cp. 27.5 n." In 
other words, the reference is thought to be to slaves who from their 
servile position really tyrannized over their master, though this 
is entirely out of line with all that has gone before, where all the 
emphasis has been placed on the absolute authority of the master, 
and does not fit in with what follows. The story that is next told 
does not relate to a slave qua slave managing a master, but to a 



Alexander: Seneca's Epistulae Morales (I-LXV) 79 

slave brought by the curious social conditions of the time to a point 
where, not qua slave but qua freedman and a very highly placed 
f reedman at that, he is lording it over his former master. The trans- 
lator of the Epistles in the Loeb series (Vol. I, p. 307) sees this, be- 
cause he translates habet "is he creating," with an obvious feeling 
for something future. 

It appears therefore that hahent really conceals the form JidbeMt, 
which is paleographieally not difficult, and I would also read the 
quot . . . hahehit sentence as a question, translating : "How many of 
those persons will he have as his masters f The di melius goes with 
the preceding words : "What ? I go to table with my slaves ? God 
forbid !" Then comes the question to constitute the ironic contrast 
with the attitude of the annoyed aristocrat. 

Compare also paragraph 12, where the point is again dwelt upon 
and where the future appears; 'at ego' inquis 'nullum habeo 
dominum.' bona aetas est ; f orsitan habebis. 

XLVIII, 7 (144 : 23) : vis scire, quid philosophia promittat generi 
humano? consilium, alium mors vocat. alium paupertas urit, 
alium divitiae vel alienae torquent vel suae, ille malam f ortunam 
horret, hie se felicitati suae subducere cupit. hunc homines male 
habent, ilium di. 

For vocat Castiglioni prefers avocat, H. Georgii"^ vexat. But 
these are only details ; Muretus raised a more fundamental ques- 
tion by suspecting the whole sentence alium mors vocat. It must be 
admitted that it stands conspicuously alone, without proper con- 
trast, as compared with the three parallel sets of ideas following, 
and that with a repeated alium heading up successive short sen- 
tences there is always a chance that one of the sentences may be 
omitted. Compare the loss of omnes spei in XL VII, 17, where we 
are able to prove definitely the omission by comparison of the 
passage with Macrobius' borrowings from it. I feel therefore that 
after vocat there has been lost what appears from every considera- 
tion the natural opposite, namely, <Calium vita degravat^. For 
degravat see XXX, 1, init., where it is associated with aetas. Death 
calls one man, who does not want to go, life continues to press 
heavily on another till it becomes labor and sorrow and still it is 
not cut off. 

"^ Philologus, LXXXIV (1928-29), p. 85. 



80 University of California Publications in Classical Philology 

XLIX, 4 (148 : 1) : modo te prosecutus sum : et tamen hoc 'modo' 
aetatis nostrae bona portio est, cuius brevitatem aliquando de- 
f ecturam cogitemus. 

The mss. all show futuram; employing this and the theory of a 
de lost after -do, Johann Mueller conjectured def ecturam, accepted 
by Hense as above. P. Thomas excises futuram. Beltrami (II) re- 
tains it, but it seems very harsh to interpret cuius . . . futuram as 
meaning quam hrevem aliquando futuram. It appears to me more 
probable that there is a lacuna after hrevitateni, based perhaps on 
the occurrence in the original text of a second abstract noun end- 
ing in -tatem. Such a word might conceivably be <iparcitatem'^ , 
found very rarely in the Latin available to us, but, as it happens, 
once in Seneca, De Clem. I, 22, 2 : civitatis autem mores magis corri- 
git parcitas animadversionum. With <iparcitatem^ supplied as a 
stopgap in our present passage, the translation would run : "and 
yet this 'but recently' is a generous segment of our years, and let 
us reflect that this brevity will some day be scarcity." 

XLIX, 5 (148 : 10) : negat Cicero, si duplicetur sibi aetas, habitu- 
rum se tempus, quo legat lyricos; eodem loco conloco dialeeticos : 
tristius inepti sunt. lUi ex professo lasciviunt, hi agere ipsos 
aliquid existimant. 

Conloco above is the word conjecturaUy supplied by Aemilius 
Hermes; compare Mdth this Rossbach's <.loca'^, accepted by Bel- 
trami. Schweighaeuser : Jiaheo or pono for the same gap, and 
Haase : <,pone^, with which Kronenberg agrees, but with the 
further addition of <^nisi qiiod^ before tristius.^ 

It occurs to me that the missing word may be <.ego^. We have 
just had Cicero's view about the impossibility of his reading the 
lyric poets, even if life were twice as long as it is. What Cicero as 
a man of affairs felt on that score, Seneca feels in regard to the 
logicians as a devotee of practical ethics, and this is well expressed 
by negat Cicero . . . eodem loco <iego^. It is rather interesting to 
observe that the following sentence begins nee ego nego; this looks 
like a reminiscence of the form of the negat Cicero sentence. With 
the reading <Cego^ the construction is loose but easily compre- 
hensible. The negat Cicero clause says in effect : "Cicero put the 
lyricists in the category of what he would not find time to read 

28 Classical Quarterly, 1 (1907), p. 206. 



Alexander: Seneca's Epistulae Morales (I-LXV) 81 

even if life were doubled"; Seneca follows with eodem loco <,ego> 
dialecticos, with a vague verbal implication which does not need to 
be made as specific as haheo or pono. 

The punctuation should be as follows : a period after dialecticos, 
a semicolon after inepti sunt. 

LII, 5 (158 : 1) : puta enim duo aedificia excitata esse, ab imo dis- 
paria, aeque excelsa et magnifica. 

In the reading given above it should be noted at once that ah imo 
is a conjecture of Buecheler's for the ms. consensus (p however is 
still lacking) amho, and on the other hand that disparia is the ms. 
consensus of the superior documents as against paria of the in- 
feriores. 

The correction ah imo where disparia is retained is most unfortu- 
nate ; it emphasizes the fact of some serious difference in situation 
before we have so much as heard that the two structures are equally 
great and magnificent in appearance. Actually I cannot think of a 
more unhappy conjecture ; the foundation differences should be 
revealed only after we have learned of the similarities of the two 
houses in outward appearance. Haupt with his duo duos aedificia 
excitasse, ambo paria, saw this point quite clearly. 

Read: amho, <.iit cre^dis, paria. The second singular verb 
credis follows well after puta, and the <iut cre'>dis puts the reader 
on his guard but without revealing the actual condition of affairs. 

LII, 5 (158: 6) : intuenti ambo quicquid fecit < alter >, alterius 
magna pars et difficilior latet. 

The actual consensus of the mss. of the superior type for intuenti 
amho is invenieho (eb underdotted in Q), while <ialter^ comes 
from one version of the inferiores. But Hense realizes that his text 
is inadequate, lacking the parallel statement to latet, and he sug- 
gests adding after <Calter^ the words <iin aperto est^. Beltrami 
(I) writes: audeo proponere multumque laboris exhaustum est 
dum pervenitur ad solidum inveniendo (scil. inveniendi opere) ; 
quicquid fecit, alterius (scil. agentis) magna pars et difficilior 
latet, but abandons this in the Edizione Nazionale, where he sug- 
gests invenies ihi for invenieho of the mss. 

Read: invenie<is sic am^ho; quicquid fecit <Calterum appa- 



82 Vniversiiy of California Publications in Classical Philology 

re^>, alterius etc. The ms. reading invenieto in its final syllable 
suggests the possibility of amho again. As for 'Calterum apparet^, 
there is no difficulty involved in having alterum (sc. aedificium) 
as the subject of fecit; in fact it might be a more serious difficulty 
to introduce a new meaning for the pronoun by reading the masc. 
sing, alter, referring to the house builder and no longer to the house 
{aedificium) as previously. I see no objection, so far as meaning is 
concerned, to saying "what the one building has achieved is ob- 
vious" ; we speak constantly in all sorts of languages of a house 
rising rapidly, while of course actually it is being raised. 

The type of mistake which I think I detect in this and the im- 
mediately preceding passage discussed arises not from any of the 
usual formal reasons, but from sheer sleepiness or weariness on the 
part of the copyist at some point in the tradition. It is natural 
that this type of mistake should occur twice Avithin a very short 
space, and then not again ; either the weariness passes off, or the 
work is suspended to some other time. 

LIII, 9 (163 : 22) : exercet philosophia regnum suum ; dat tempus 
non accipit. non est res subseciva; ordinaria est, domina est, 
adesse iubet. 

The adesse iuhet is a conjecture of Haase's for the ms. consensus 
adest et iuhet, adopted by Hense and by Beltrami (II). I confess 
to surprise that Beltrami, who shows himself remarkably open to 
arguments based on clausula rhythm, should not have observed 
that est, adest et iuhet gives the double cretic finish so much affected 
by Seneca, while as for Haase's adesse iubet, as Summers remarks 
(op. cit., p. 54, fn.) , the rhythm is against it. Summers himself pro- 
posed adest ut lihet, which preserves the double cretic. 

In a previous brochure^ on the text of the Epistles I myself in- 
troduced a relatively uncommon clausula rhythm by suggesting 
domina et adest et iuhet (fourth paeon plus cretic) . I now feel that 
it is better to stand bj'' the ms. reading ; but I still hold that the 
punctuation should be a comma after suhseciva and a semicolon, 
or even a period, after ordinaria est, thus bringing together domina 
est and adest et iuhet in the same compartment. Suhseciva is thus 
balanced off against ordinaria, and adest et iuhet is the exegesis of 

^ Op. cit. (Univ. of Alberta Publications), p. 9. 



Alexander: Seneca's Epistulae Morales (I-LXV) 83 

doniina est. Domina est, adest et iiihet means : "she is mistress of 
the house ; she is on the spot and issues her orders." This is a suc- 
cinct description of the good housewife, always attending to busi- 
ness, taking charge of things herself. Compare the description of 
the opposite type, Dial. 6, 10, 6, fin. : ut varia et libidinosa manci- 
piorumque suorum neglegens domina. Such is Fortuna, but Phi- 
losophy adest et iubet. 

LIII, 12 (164 : 15) : incredibilis philosophiae vis est ad omnem fort- 
uitam vim retundendam. nullum telum in corpore eius sedet: 
munita est, solida : quaedam defetigat et velut levia tela laxo 
sinu eludit. 

Yeluti levia tela, suspected by Haupt, was by him emended into 
veluti evitata, by Hense (II) in his critical note to veluti e via 
delata. Rossbach defends the words, and Beltrami (II) retains them 
in the text. But Hense's own comment (Supplementum Quirini- 
anum of the 2d edition, pp. iv-v) on Rossbach's defense is very 
shrewd : sed tela si deesset, nemo desideraret. I think we should 
read levia [tela] , regarding tela as a "helping" word, marginal or 
interlinear, for purposes of interpretation, which has succeeded 
in intruding itself into the text. 

LVI, 2 (170 : 22) : praeter istos, quorum, si nihil aliud, rectae voces 
sunt, alipilum cogita tenuem et stridnlam vocem, quo sit notabi- 
lior, subinde exprimentem nee umquam tacentem, nisi dum vellit 
alas et alium pro se clamare cogit. iam biberari varias exclama- 
tiones et botularium et crustularium et omnes popinarum in- 
stitores mercem sua quadam et insignita modulatione vendentis. 

Apparently the text in Hense as given above carries the force of 
cogita clear through to vendentis; if so, the punctuation should 
show a semicolon, and not a period, after cogit. 

The mss. read bihcrari: QLpVJNI and liherarii: Pb {-ri, b), while 
libarii is the conjecture of Caelius Rhodiginus. This is a word in- 
ferred from lihum; it has no other standing. Similarly hiherari is 
an assumed word, read bj' Hense as above and Beltrami (II) , based 
on hiher (Tlies., II, 1954, 61 ff.), since biheres in the sense potiones 
appears in two sets of rules of monastic orders. 

The natural run of the sentence is cogita alipilum . . . et botu- 
larium et crustularium et . . . institores, which is definitely broken 



84 University of California Publications in Classical Philology 

by iam hiherari varias exclamationes, assuming for the moment the 
existence of such a word as hiherarius. I am led, however, to believe 
that the four words in question refer, not to the main current of 
the sentence, but only to what immediately precedes them, namely, 
alium pro se clamare cogit, in the form of a parenthetical comment 
thereon. Further, I believe liherari (deformed also into hiherari) 
to be a gloss which has displaced the controlling verb of exclama- 
tiones, and that this parenthetical clause read : (wam tollunt varias 
exclamationes). Someone noted marginally against this phrase 
the word liherari, "to be let go," as indicating the object of the 
shrieks, and this comment displaced tollunt, perhaps because the 
reference mark to the marginal comment stood nearest tollunt. 
Once this had happened, na^n illustrative would have to change 
into iam continuative. 

LVII, 7 (176 : 12) : nunc me putas de Stoicis dicere, qui existimant 
animam hominis magno pondere extriti permanere non posse et 
statim spargi, quia non fuerit illi exitus liber? ego vero non 
f acio ; qui hoc dicunt, videntur mihi errare. 

Summers^^ seems to propound a sound objection to the text as it 
stands by pointing out that "there is no other authority for the 
belief here assigned to the Stoics : it certainly was not, as the text 
here implies, the regular Stoic dogma." But it may have been ad- 
vanced as a theory by some Stoics, and I think the difficulty, which 
is a genuine one, will be met by reading the extremely plausible de 
<ieis^ Stoicis etc., made still more plausible by the circumstance 
that the reading in Q is dei stoicis. 

LVIII, 32 (185 : 22) : prope est a timente, qui fatum segnis expec- 
tat, sicut ille ultra modum deditus vino est, qui amphoram exiccat 
et f aecem quoque exorbet. de hoc tamen quaeremus, pars summa 
vitae utrum [ea] faex sit an liquidissimum et purissimum quid- 
dam etc. 

Haase rejected ea, and was followed by Hense and Beltrami; 
all three also read faex of the inferiores as against haec of all the 
superiores, including Q. Contrary to these views, I hold with Fick- 
ert that the ms. reading should be retained in both places, as both 
the ea and the haec have clear and definite references in adjacent 

^ Op. cit., p. 242, note to paragraph 7. 



Alexander: Seneca's Epistulae Morales (I-LXV) 85 

nouns, ea resumptively to pars summa vitae and haec to the faex 
of the preceding sentence. Nor could there be any mistake about 
the haec referring to faex in view of the liquidissimum et puris- 
simum following, which establishes the meaning of haec by con- 
trast. In any event, even if the resumptive use of ea is so far doubted 
as to make the excision of that pronoun seem desirable, the haec 
should be retained, if any weight attaches to the superior value of 
the ''harder" reading. 

LXI, 1 (194: 16) : desinamus, quod voluimus, velle. ego certe id 
ago, senex <ne> eadem velim, quae puer volui. 
The ms. consensus (the testimony of p being again available) is 
senex eadem velle. Hense (II), as given above, in which he was 
followed by Beltrami (I), was based on Madvig's <ne> senex 
eadem velim. But Beltrami (II) adopts the emendation of Casti- 
glioni"^ senex eadem <ut desinam> velle. Hense"" has also aban- 
doned the reading of his second edition for senex eadem <.ut 
vitem^ velle. 

Stylistically there appears to me, as to Castiglioni, though Hense 
I.e. supra thinks otherwise, a great desirability for having the 
desino and the volo of the initial sentence appear again in the 
specific application of that sentence to Seneca himself. Schultess 
saw this long since in his Annaeana Studia, and suggested that the 
-dem of eadem really should be read desii; and anyone who cares 
to experiment for himself by writing desii in not too precise mi- 
nuscules will see the force of the suggestion. Desii is one of the 
very few changes of text that Gummere allows himself in Vol. I 
of the Loeb edition ; he, however, does not throw senex back with 
ego certe id ago as Schultess did, but forward with the rest of the 
sentence, thus getting a good contrast between senex and puer. The 
emendation of Schultess coupled with Gummere's punctuation 
solves the problem, in my judgment. 

LXIII, 4 (197 : 20) : id agamus, ut iucunda nobis amissorum fiat 
recordatio. nemo libenter ad id redit, quod non sine tormento 
cogitaturus est. sic et illud fieri necesse est etc. 

Sicut appears in QpLV (with the ut underdotted in V) , while Pb 

32 Studi italiani di filologia classica, n.s., II (1922), p. 225. 
^Bhein. Mus., LXXIV (1925), p. 120. 



86 University of California PuMications in Classical Philology 

show sic. Heiise, as above, is followed by Beltrami (II), but I am 
no more able than Summers^* to see that this effects any improve- 
ment. I do not, however, feel myself convinced of the necessity of his 
scio: at. I suggest that the objector, or interlocutor, who plays so 
large a part in the dramatics of Seneca's prose, be assigned the 
whole sentence "nemo libenter . . . nomen occurrat," at which point 
the answer is properly taken up by sed (line 25). This sentence 
should, however, be punctuated with a comma only after cogita- 
turus est, and a semicolon after siciit illud. The effect is to throw the 
sicut illud back, not forward; the illud is the amissorum 7'ecordatio 
of the preceding sentence, which is taken by the objector as a type 
case (sicut illud) of the tormentum you encounter by returning to 
an unwelcome subject or experience. 

LXV, 15 (207: 6) : 'quid te' inquis 'delectat tempus inter ista 
conterere, quae tibi nullum affectum eripiunt, uullam cupidita- 
tem abiguntf ego quidem f peiora ilia ago ac tracto, quibus 
pacatur animus, et me prius scrutor, deinde hunc muudum. 

Numerous substitutions for peiora have been suggested, all mak- 
ing in the opposite direction, but peiora may be right as it stands, 
provided we make the easy alteration of animus, set for animus et, 
and read placatur with p for pacatur. 

The questioner says: "Why do you waste j^our time on such 
trifles, ineffective from the main viewpoint?" Seneca replies: "So 
far as I am concerned, I handle and treat these inferior means by 
which a soul is appeased, but I scrutinize myself first, and then the 
universe." That means that while he does not overlook the value of 
intellectual speculation for the soothing (nothing more) of the soul, 
he is a good Stoic in regarding self-examination as coming first, 
then contemplation of the cosmic order. "Not even as things now 
stand," he continues, "am I losing time, as you think." "As things 
now stand" {nunc) means "while I am engaged in purely intellec- 
tual speculation," as is proved by the ista enim omnia immediately 
following (line 12). The effect of this intellectual speculation is, 
ultimately, if overspecialization is avoided (si non concidantur nee 
in banc subtilitatem inutilem distrahantur), to bring the mind to 
view the rerum naturae spectaculum, that is to say, the hunc mun- 

3* Classical Quarterly, III (1909), p. 42. 



Alexander: Seneca's Epistulae Morales (I-LXV) 87 

diini of the sentence from which we began this present discussion. 
Thus purely intellectual speculation is good (1) because it soothes 
the spirit, and (2) because, properly conducted, it leads to an ex- 
amination of one's self and to a contemplation of the universe as a 
whole. 

The peiora is undoubtedly ironical, reflecting the implications 
of ista in the interlocutor's question, and not Seneca's point of view. 



n 



MANES EXITE PATERNI 

BY 

H. J. ROSE 

We possess in Ovid (Fasti, V, 421 sqq.) a brilliant and vivid de- 
scription of the ceremonial of the Lemiiria (May 9, 11, 13). I have 
no desire to comment on it at length, but merely to point out a con- 
fusion which seems to have escaped the notice even of Sir J. G. 
Frazer in his magisterial exposition of Ovid's work.^ The poet, as I 
hold, having no really profound knowledge of the older religious 
practices of his country and getting his facts second-hand, probably 
from Verrius Flaccus — who himself, at that date, could not be ex- 
pected to interpret them as a properly trained modern can, — has 
been inadvertently guilty of an inaccurate expression and so of mis- 
leading students of his poem ever since it was published. 

On the first of the days in question, says Ovid, there will take 
place a very old rite, going back to the age when the Romulean year 
was still in use, and consequently there was no February and (he 
implies but does not say) no Parentalia. This remark (lines 423 sq.) 
is clearly not fact but learned theory; Verrius, or whoever started 
the idea going, was puzzled to find in the calendar two periods, 
Parentalia and Lemuria, in which the shades of the dead were 
propitiated. He therefore supposed that both could not be original, 
and rather unhappily assumed that the rite which belonged to the 
newer month was itself the newer, forgetting that in that case he 
was left with as insoluble a difficulty as ever, namely, why the old 
festival, the Lemuria, was not ousted from May when the new one, 
on this supposition, took its place. He seems to have got around 
this as best he could, however, by supposing the ritual to have been 
divided ; May now-a-days, says Ovid, has but a part of the ceremonial 
it once had, -partem prisci nunc quoque moris (428). 

The ceremony itself, however, is described very clearly as re- 
gards its details and its alleged purpose. It is an offering to the Good 
Ones, inferias tacitis manihus dabunt (422). Hence at the conclusion 
of the ceremonial the operator says manes exite paterni (443). It is 

1 Fasti of Ovid (London, Macmillan, 1929), IV, 36 sqq. 

[89] 



90 University of California Publications in Classical Philology 

in one or both of these phrases and their implications that the inac- 
curacy hes. 

If we ask what the operator does, Ovid is precise and informative. 
The ritual is performed by a householder; hence, like the Parentalia, 
this is not a pubHc festival but a private one recognised by the state. 
He rises at midnight and remains barefooted throughout ; he fa la 
fica lest he encounter a spectre, and washes his hands in water from 
a running spring. Then he turns and takes black beans in his 
mouth, apparently nine of them. These he spits out or takes out of 
his mouth and throws down, being careful not to turn his head. At 
each throw he says, "These I let go, with these beans I ransom me 
and mine." His intention is to feed the spectres which are following 
him: umbra putatur colligere (sc, fabas) et nullo terga uidente 
sequi (439-40). This completes the central rite; the ghosts are fed, 
it remains to get rid of them. This he does in three different ways: 
first, by washing his hands again ; second, by making a noise with 
bronze implements of some sort (Temesaea concrepat aera, 441);- 
third, by repeating nine times the formula of dismissal already 
quoted. When all this is accompHshed — that is to say, when the 
ghosts and their uncanny influence have been washed off, scared 
away by a metallic noise and bidden depart, — he may and does 
safely look around, for the spectres are gone and the ceremony over. 

Frazer, with his immense knowledge of folklore, only just misses 
the true explanation of this passage ; that he does miss it at all is due 
to over-much respect for Ovid's authority, for,hke so many patient 
and absolutely honest researchers, he does not always understand 
that a merely literary man, with no pretence to profound or exact 
learning, is often content to say what will sound well without too 
much consideration whether or not it is true. In his commentary he 
accumulates citations to show that the lemures are terrifying ghosts, 
conceived as more or less dangerous, and he quotes instance after 
instance of ceremonies in various lands whose object is to appease 
or get rid of demonic powers of a hke kind. But, misled by Ovid's 

- Temesaeais of course the merest ornament: « Teniarjv /xerd xa^^Wi aT^ 5' 
aWdiva aiSrjpov, a (184). 

3 It is quite indifferent whether we think of the householder as pouring 
water on his hands to wash away a ghost which is holding on to them or 
simply as cleaning them of the spiritual dirt of ghostliness. With such im- 
personal beings, so vaguely conceived, the two are one. 



Rose: Manes Exite Paterni 91 

identification of the bogeys his house-father expels with the kindly 
ghosts of kinsfolk (though even Ovid backs up his or Verrius' false 
etymology of the name, Lemuria quasi Remuria, by attributing its 
origin to ceremonial intended to appease the very terrible spirit of a 
murdered man)/ he searches about for instances of more or less 
good ghosts who are dealt with at the same time as evil ones. That 
such confusion does exist is perfectly true, and the cause of it is not 
far to seek; most people are afraid of any visitant from the other 
world, no matter how harmless and no matter whether they believe 
in his existence or not.^ But my point is that among the ancestors of 
the historical Romans it did not exist, and that their ritual shows 
it did not. 

I need not argue at length to prove that the Romans were an 
orderly and legal-minded people, who would remember the neces- 
sity of treating different kinds of supernaturals in a different way, 
exactly as they remembered to institute different procedures for 
Roman and foreign litigants. Now if we compare the rites of the 
Parentalia and those of the Lemuria, we shall find some essential 
differences between them. At the former, whole families took part, 
as was natural, seeing that they were showing kindness and respect 
to those members of the same families who no longer lived in this 
world but in the other. The graves were bedecked, food was set out 
for the ghosts on the streets, altars were fit, and prayers or other 
formulae of address recited. The whole business ended with the 
Caristia or Cara Cognatio, at which the living kinsfolk united for a 
social meal.^ It was an unchancy time, during which temples were 
shut, no marriages celebrated, and some ill-disposed folk took the 
opportunity of performing evil magic f but this is true of all times 

* Ovid, ibid., 451 sqq. 

=" For instance: During the war of 1914-18, a Belgian refugee was much per- 
turbed by some alleged hauntings in a Welsh country house. He explained 
his feelings very frankly by saying, "You see, my wife believes in such things, 
and so she is not frightened; I don't, and so I am." Cf. Poe's famous analysis 
of the terror caused by a supposed apparition in The Narrative of Arthur 
Gordon Pym. 

« See Wissowa, RKR'^, pp. 232 sqq.; Ovid, op. cit., II, 533 sqq., and Frazer 
there. 

^ Like the old woman in Ovid, ibid., 571 sqq., whose mumbo-jumbo and in- 
vocation of a goddess she and the poet call the Silent One are gravely referred 
to by many moderns, who evidently have never read the lines with a sem- 
blance of understanding, as evidence for a Roman cult of Tacita. The state 



92 University of California Publications in Classical Philology 

when ghosts are about. One takes precautions with a Hve wire, but 
that does not prove that it is attached to a hostile mine. In sharp 
contrast, the ritual of the Lemuria is not a family matter, but is left 
in the hands of the paterfamilias, who here as normally is the house's 
priest and sacral expert. He does nothing out of doors nor at the 
familiar tombs where his ancestors lie and he will in time be ushered 
into the world of the dead, but confines himself to that critical part 
of his own dwelling where it touches the earth; in other words, to 
the floor, the haunt of the homeless, hungry ghosts who are not 
cared for at Parentalia or Feralia and therefore are always on the 
prowl, seeking to do mischief. For we know from Pliny* that this 
was their place and that precautions were taken accordingly. Of 
course originally the atrium of the simple and old-fashioned Roman 
house had no flooring and those in it walked about on nothing more 

was not troubled by the ghostliness of the time, for not only the rest of the 
dies parentales, save those which coincide with public festivals, but even the 
Feralia, are marked F (see the Fasti Antiates; Wissowa's calendar, op. cit., 
p. 570, is wrong for once). Cf. Frazer, op. cit., II, 434, who very nearly gets 
the point. 

* Pliny, N.H. XXVIII, 27, a passage which I have never seen correctly 
interpreted save by myself {Gnomon, XII, 1936, 390, in a review of X. F. M. 
G. Wolters, Notes on Antique Folklore, Amsterdam, H. J. Paris, 1935, a 
monograph in many ways admirable but quite wrong in this respect.) It has 
actually been twisted into supporting the theory that the Lares are ghosts, 
whereas it is the clearest of testimony that they are no such thing. Pliny 
says: cibus etiam e nianu prolapsus reddebatur utique per mensas, uetabantque 
munditiarum causa deflare, et sunt condita auguria quid loquenti cogitantiue id 
acciderit, inter exsecratissima si pontifici accidat dicis causa epulanti. in mensa 
utique id reponi adolerique ad Ldrem piatio est. That is to say, if any food is 
dropped on the floor at meal-times, it is not a good omen, and the least one 
can do is to put it back on the table, pretending that it has not been dropped 
at all by not blowing the dust off it (or is this a precaution against blowing 
lemures all over the room ?). But if the omen is really very bad, most of all if 
a priest of one of the Immortals is thus brought into contact with such unholy 
things, it is proper to perform the full ceremony {piatio, process for making 
something pium. Pliny uses this instead of the much more common piamen, 
properly a piece of such ritual or of material used in it, to avoid all ambiguitj' 
arising from the secondary meanings of piamen; but his efforts have been in 
vain against the unwillingness of those blinded by a false theory to see the 
facts that prove them wrong). The layman, or even the priest at an ordinary 
meal, might be content to put the defiled food once more in contact with the 
holy table and let that good influence overcome the evil the bogeys have 
done; but when it is a formal feast, imdertaken as part of the official and 
clerical duties, dicis causa, such an attack by the powers of mischance must 
be severely dealt with and divine aid invoked. The Lar Familiaris is there 
with his clear and very holy fire; let that destroy the invader from the dark 
world once and for all. 



Rose: Manes Exite Paterni 93 

elaborate than beaten earth. And this, it would appear, was not 
very good or holy earth, for it is quite untrue that the dead were 
ever buried in it, as they have been in some countries.^ No doubt 
there were many other places where a lemur might crawl, looking 
for something to satisfy his hunger and thirst; but those did not 
concern the householder, who was seeking to get his own home rid 
of these spiritual rats, not to clear the whole neighbourhood of its 
supernatural vermin, for that was the state's busmess, or that of 
other patres familiarum. So, fortified by holy washing and a holy 
sign,!" he ventures into the enemy's territory, barefooted so as to be 
in direct contact with it and in the middle of the night when all evil 
things have most power. He has come as a sort of ambassador, 
bringing Danegeld on an inexpensive scale. He feeds the lemures 
beans, which at best are inclined to be ghosts' food,^^ and makes 
them more tasty by putting them first in his own mouth, so that 
they shall have the flavour of man. The lemures perhaps think they 
have devoured him; at all events they are satisfied with this substi- 
tute. So now, as they have no further business about the house, he 
can frighten them with a noise of metal and bid them begone. It 
may well be that he spoke to them pohtely, saying "Good folk, get 
you hence," manes exite; but it is very certain that he did not insult 
his own di parentes by using the adjective paterni in such a context. ^^ 
That is a blunder either of Ovid or of his source. 

8 The statement of Servius on Aen. V, 64; VI, 152, that the dead used to be 
so disposed of is contradicted by all archaeological evidence for Italy. It 
results from the same ghost theory of the Lares (and here of the di pennies 
also), which was quite popular in antiquity. 

" Lafica, because obscene, is of course holy. Evil things hate all allusions 
to the warm and sacred powers of generation, being the Comstocks of the 
other world. The apparent exceptions are either the result of old gods of fer- 
tility having degenerated into demons under the influence of a new religion, 
or else mere perversions, inimical to the normal satisfaction of normal 
instincts. 

11 For examples, see Frazer, op. cit., IV, pp. 39, 44. 

1^ Naturally one always assumes that one's own kin, alive or dead, are 
doing what is magico-religiously right. To hint that they are lemures would 
be the very opposite of alluding to them as "blessed," "sainted," or the like, 
as scores of languages do, when speaking of the departed members of one's 
own or a friendly family. It is no more possible for a Roman of early days 
than it would be for a modern Greek to call his departed grandfather a 
vrykoldkas instead of piously adding "may God pardon him" to the mention 
of his name, with the implication that he died in the peace of the Church and 
is, at worst, on his way to Paradise. 



i1 



PHYSIOLOGUS LATINUS VERSIO Y 



EDITED BY 

FRANCIS J. CARMODY 



INTRODUCTION 

For more than sixteen centuries the Physiologus has been a favor- 
ite text on animal lore, more widely read and copied than any other 
tract on natural history.^ Its obscure origins and its exotic material 
gave it charm not rivaled until the end of the thirteenth century; 
its moralizations in hexaemeral spirit, short and to the point, ap- 
pealed to the religious; and its immense philological implications 
have kept it alive in modern scholarship. 

Briefly, the history of Physiologus studies is as follows. The edi- 
tion of the Syriac translation from a Vatican manuscript in 1795 
stood alone for a half century .^ Almost simultaneously Heider edited 
the abridged version^ lettered L and Cahier" gave composite texts of 
the Latin manuscripts A (Brussels MS 10074) and B (Bern Lat. 
233), and of A and C (Bern Lat. 318), with notes from L and from 
De Bestiis et aliis rebus, ^ and the texts of the French translations by 
Pierre le Picard and Guillaume le Clerc. In 1855 Pitra edited the 
Greek and the Armenian and gathered a number of Latin fragments.^ 
Cahier resumed his work in 1874 with a French translation of the 
Armenian version, and Land edited in 1875 one Arabic manuscript 
(in the Royal Batavian Academy) and a new Syriac translation 
(Brit. Mus. Add. 25878), with Latin translations of each.'^ Lauchert's 
study of the Physiologus in 1889 remains the most convenient refer- 

1 There are well over 250 MSS in Latin, Romance, and Germanic languages 
copied between 1100 and 1400, plus some 50 of De Bestiis and another 50 of the 
Theobald version in 12 chapters (ed. A. Rendell, London, 1928, and Pat. Lot. 
CLXXI [Paris, 1854]). One should also mention the extensive use by the 
encyclopaedists, especially Isidore, Honorius, and Vincent of Beauvais. 

* O. G. Tychsen, Physiologus Syrus, Rostock, 1795. 

2 Archivfiir Kunde osterreichischer Geschichts-Quellen, V (1850), 552-582. 
^ Cahier et Martin, Melanges d' archeologie ... 4 vols., Paris, 1847-56. 

6 Pat. Lat. CLXXVII (Paris, 1879), 15 f. 

^ Spicilegium Solesmense, Vol. Ill (Paris, 1855). 

^ Cahier, Nouveaux melanges, Paris, 1874; Land, Oiia Syriaca, Leyden, 1875. 

[95] 



96 University of California Publications in Classical Philology 

ence, though his edition of the Greek is unsatisfactory.^ Only two 
works of this period made any contribution to knowledge of the 
Latin translations, an edition of one manuscript related to Cahier's 
B, by Mann (Brit. Mus. Reg. 2 C xii),^ and the general analysis of 
Physiologus problems published by Ahrens.'" Since then the numer- 
ous publications on the Physiologus have concerned the versions in 
Romance languages, in Greek, and in the languages of Asia Minor, 
for which a good bibliography can be found in the most recent edi- 
tion of the Greek, edited by Sbordone in 1936.^^ Sbordone, in his 
discussion of the Latin Physiologus, follows Cahier's unfortunate 
association of manuscripts A and B, and, though he claims acquaint- 
ance with other manuscripts of this translation, makes no advance 
in this important indirect tradition. Sbordone did, however, point 
out that another Latin translation was extant in three manuscripts, 
and gave samples of six chapters. It is these three manuscripts which 
are edited for the first time in the present volume. And here I must 
acknowledge my debt to the Editors of our series in Classical Phil- 
ology for their suggestions, and to the European libraries through 
whose kind help I was able to procure the photostats used for the 
present edition. 

Sbordone's critical text of the Greek is imposing, and follows for 
each chapter some dozen manuscripts. His critical method is un- 
sound, and his classification of the Greek manuscripts can be shown 
to be faulty in the light of the indirect tradition in other languages. 
This is not the time or place to present a full criticism of Sbordone's 
text ; suffice it to say that he has successfully divided the 66 manu- 
scripts into four distinct groups: 1° the probable primitive text, rep- 
resented by 15 manuscripts, among which the best examples are 2 
and W; 2" a very different text, represented by 8 manuscripts, of 
which the oldest and most important is H; finally, 3° a Byzantine 
and 4° a pseudo-Basilian text of small interest to the indirect tradi- 
tion in other languages. 

As between the prototype of the groups represented by W and IT 
no chronology can be determined. In both versions one might get 

* Geschichte des Physiologus, Strassburg, 1889. 

9 Franzosische Studien, VI, pt. 2 (Heilbronn, 1888), 37-73. 

" K. Ahrens, Zur Geschichte des sogenannten Physiologus, Ploen, 1885. 

" F. Sbordone, Physiologus, Milan, 1936. 



Carmody: Physiologus Latinus Versio Y 97 

the impression that the language is of the fourth or fifth century, 
but their use by Justinus Martyr (06. 166), Origen (oh. 254/55), Ter- 
tulHan (oh. 230), and Clement of Alexandria proves greater age, 
and the corrupt Greek style may depend on a slavish imitation of 
some Semitic original. The author cannot at present be identified, 
nor the exact region from which he came. The elaborate Ust of 
sources given by Sbordone may be collectively called parallelisms, 
since they show no convincing similarities. The Greek text was pop- 
ular longer than the Latin; the latter was most read during the 
thirteenth and fifteenth centuries, the Greek (to judge by the extant 
manuscripts) during the sixteenth and seventeenth. 

Scholars have long known of the translations in the languages of 
Asia Minor and Africa, but have never used them in studying the 
Greek text. In general it has been accepted that the Ethiopic trans- 
lation is related to the primitive Greek (W,etc.), while the Armenian 
was to be classed with Greek MS IT and Latin Y. Latin b has no 
close relationship with IT, however, and the whole classification 
clearly needs reorganization. Preliminary to preparing the present 
edition of Y, I therefore established a composite text (in English) 
of the Ethiopic (e), Armenian (d), Syriac (s), and Arabic (c) (the 
last two from Land only), then adding the readings of Latin C, U, 
and h; I concluded that the frequent agreement between all these 
versions and Greek n was proof of the authenticity of many read- 
ings in this text which were rejected by Sbordone. The critical text 
y bears out this contention, and is in fact the most complete and 
reliable Latin translation now known. Its publication will provide 
at once a convenient text, better than h, and a necessary element in 
the indirect tradition of the original Physiologus. I venture to repro- 
duce at this point a few samples of the indirect tradition : 

y 1.8 hi qui sursum sunt {as in b) 
2.7 seculi pompa (bCU) 
5.15 peccata nostra portans (b and Gr. WO) 
13.6 per abstinentiam et trib. {U and implied in c) 
17.15 nauis nisi extensum (6) 
32.9 percisus est lapis de monte (11 and ds) 
38.6 infernus (be) 

42.3 semel generat (ds and Greek AEU) 
43.3 conculcauit (fees and Greek WO) 
43.6 dracones in corde (e and Greek AEU, and of. c) 



98 University of California Publications in Classical Philology 

LATIN PHYSIOLOGUS Y 

The text of the Latin translation y has been presented in three 
manuscripts, Y (Monac. Lat. 19417), Y^ (id. 14388), and Y^ (Bern 
Lat. 611), with 13 chapters in A (y chaps. 10-13, 16, 19-20, 22-24, 
26-27, 29) and probably several chapters in C, M, and De Bestiis 
(E and H). Y and Y^ give the same text, whereas Y^ and A have 
been altered and contain certain chapters from h. The choice of 
variants was facilitated by the obvious grouping YY^ as against 
AY^, and throughout by the Greek texts, cdes, and CU. For most 
chapters this procedure was satisfactory, though in several the 
manuscripts were reduced to one through loss of folios, and the sur- 
viving text left unemended for the most part. Each of the three 
complete manuscripts attributes the work to a different author, Y 
to Chrysostomus (as also I), Y^ to St. John of Constantinople, Y^ to 
an orthodox bishop. There are at least forty-five verses from the 
Bible which contain one or more readings from the Old Latin, as 
confirmed in the works of St. Ambrose and his contemporaries (all 
such readings are here given in italics) . The text of y, like that of C 
but in general unUke b, contains a number of transcriptions of Greek 
words, showing haste or ignorance on the part of the translator; 
such are politeuta, polite nomine (for politeuta), apodumata, cerseus, 
dessicano, eutocium, and the names of the animals. The sjnitax of y 
is that of the Patristic writers, perfect subjunctive after cum and si, 
compound tenses for the passive, future of supposition, passive par- 
ticiple for active, amho in. the plural, and consistent imitation of 
Greek word order. The order of chapters in y is in part more authen- 
tic than that of the Greek editions, and is confirmed by IT and by d. 
There are, however, several transpositions toward the end of y, 
where groups of three to six chapters were transposed intact. 

CRITICAL METHOD 

The conditions of editing y were that it belonged to the fourth or 
fifth century, that it represented valuable evidence of the indirect 
tradition of the lost original, and that at present it exists in onlj^ 
three manuscripts. For the most part, the recommendations of the 
Union Acad^mique Internationale^^ were observed ; but since many 
" Emploi des signes critiques, description de I'apparat, Paris, 1938. 



Carmody: Physiologus Latinus Versio Y 99 

of the devices therein defined concern reediting known works, it 
seemed wise to modify them to accord with those of the Societe des 
Anciens Textes Frangais,!^ whose experience concerns texts of vari- 
able language for which no fixed standards have been determined. I 
shall describe the compromise as here effected. 

Sigla.—A desire to preserve the sigla assigned by Cahier and 
others and to add to them, and the further need of ample lettering 
for the several hundred Latin manuscripts known today required 
the use of superscript numerals to distinguish between manuscripts, 
not between scribes. Lower-case letters are reserved for versions or 
texts, such as h (critical text of the MSS BMZ, DHL, B^'^'^'EEm, 
A), I (critical text, unpublished, of LW^''' etc.), and y {YY^'A). 
Sigla for Greek manuscripts are preceded by Gr. 

Text.— The best manuscript, F, has been reproduced in every 
detail of spelling and word order, either in the text or in the appa- 
ratus; full-size photostats were rechecked a number of times; and, 
except conceivably for the utterly chaotic punctuation, Y need not 
again be examined. The retention of the spelUng of F, even when 
inconsistent, reconciles the recommendations of the Societe and 
those of the best practices for editions of patristic texts. ^^ All the var- 
iants of F2 have been similarly recorded (except spelling of common 
words) and rechecked from full-size photostats. F^ was examined 
from microfilms and the principal variants recorded; the original 
pale palimpsest should some day be consulted. A was used from 
Cahier's transcription, and should eventually be checked from the 
manuscript; MS Gud. 131 at Wolfenbijttel is reputed to contain the 
same text as A, and should be seen, but Rouen 638 contains only 
those chapters of A which are related to h. 

Apparatus. — The lemmata are followed by a bracket; they (and 
the text) reproduce the spelHng of F except when F lacks the word 
in question, in w^hich case preference is given respectively to F^ or 
F^ Each chapter is followed by the fist of manuscripts containing 
its text. Omissions {om.) are of the lemmata as given, not of the 
lemmata as parts of longer omissions, as of lines or foUos. A study of 
the habits of the several scribes permitted solution of most of the 
abbreviations, though some (notably & for ex, it, ei, ec, etc.) had to 

" Romania, 52 (1926), 241-256. 

" Plater and White, A Grammar of the Vulgate, Oxford, 1926. 



100 University of California Publications in Classical Philology 

be retained. Emendations (followed by the authority, sometimes 
that of the editor, edit.) were confined to flexional endings and spell- 
ings of Greek words otherwise deceptive. This procedure leaves y 
free from the influence of the other Physiologus texts and therefore 
more useful for a study of the indirect tradition. Since emendations 
were in general avoided, the crux became superfluous. 

Punctuation. — The punctuation is that of the editor, and was 
suggested in large part by the texts in other languages. Parentheses 
are used as an added device for punctuation only. Bible references 
follow each Bible quotation, and are set in brackets: they are en- 
tirely modern, and can be found in no known manuscript, but, set 
into the text, they have seemed more immediately useful and precise. 



Carmody: Physiologus Latinus Versio Y 101 

PHYSIOLOGUS LATINUS 

PAGE 

I. Leo 103 

II. AUTOLOPS 104 

III. PiROBOLi Lapides 104 

IV. Serra Marina 105 

V. Charadrius 105 

VI. Pelicanus 106 

VII. Nycticorax 107 

VIII. Aquila 107 

IX. Phenix 108 

X. Epops 109 

XI. Onager 109 

XII. Uipera 110 

XIII. Serpens 110 

XIV. Formica 112 

XV. Syrena et Onocentaurus 113 

XVI. Herinacius 114 

XVII. Ibis 115 

XVIII. UxJLPis 116 

XIX. Arbor Peridexion • • • 116 

XX. Elephas 117 

XXI. Dorchon [Caprea] 119 

XXII. AcHATis Lapis 120 

XXIII. [Ostrea] Sostoros Lapis et Margarita 120 

XXIV. Adamantinus Lapis 121 

XXV. Onager et Simius 121 

XXVI. [Indicus] Senditicos Lapis 122 

XXVII. Herodius id est Fulica 122 

XXVIII. PSYCOMORA 123 

XXIX. Panther 124 

XXX. Cetus id est Aspisceleon 125 

XXXI. Perdix 125 

XXXII. Uultur 126 

XXXIII. Mirmicoleon 127 

XXXIV. Mustela et Aspis 127 

XXXV. MONOCERAS 128 

XXXVI. Castor 128 

XXXVII. Hyena hoc est Belua 129 

XXXVIII. [Hydrus] Niluus 129 

XXXIX. [Ichneumon] Echinemon 130 

XL. CORNICOLA 130 

XLI. TURTUR 131 

XLII. Hyrundo 131 

XLIII. Ceruxjs 131 

XLIV. Rana 132 

XLV. Saura ID EST Salamandra 132 

XLVI. Magnis Lapis 133 

XLVII. Adamantinus Lapis 133 

XLVIII. COLtJMBAE 133 

XLIX. Saura Eliace hoc est AnguillaSolis 134 



102 University of California Publications in Classical Philology 

SIGLA 
Versio y — 

Y Munich, Lat. 19417, fol. 29^-70^ s. IX (photostats). 

72 Munich, Lat. 14388, fol. 172^-183^, s. IX-X (photostats). 

Y^ Bern, Lat. 611, fol. 116^-138^, s. VIII-IX (microfilms). 

Compendia — 
A Brussels, MS. 10074, s. X ex. (Cahier). 
D Paris, Bibl. Nat., Lat. 2780, s. XIII (Cahier). 
M Alphabetical fragments, s. VIII (Mai, Pitra). 

Versiones — 

C Bern, LaL3/S,s. IX (Cahier). 

U MS. Hamilton 390 {Zeitschr.fiir rom. Phil., 1888). 

b Edit. F. Carmody, Paris, 1939 (Bern, Lat. 233, etc.). 

c Arabus batavus (Lat. transl. by Land, Leyden, 1875). 

d Armenian transl. (French transl. by Cahier, 1874). 

e Ethiopian transl. (German transl. by Hommel, Rom. Forsch., 1890). 

I Latin attrib. to Chrysostomus (edit. Heider, 1850) : 

L Edit. Heider (three MSS). 

L^ Munich, Lat. 19648, s. XV (photostats). 

L3 Munich, Lat. 23787, s. XV (photostats). 

L^ Munich, Lat. I42I6, s. XV (photostats), 
s Syrus leidensis (Lat. transl. by Land, Leyden, 1875). 
n Mosquensis Gr. 432, s. XI (edit. Karnejev, Sbordone). 
S Mosquensis Gr. 298, s. XIII (edit. Sbordone). 
W Vienna, Theol. Gr. 320, s. XIV (edit. Lauchert, Sbordone). 

Editiones — 

E De Bestiis, Book III, Paris, Bibl. Nat., Lat. 11207, s. XIII (microfilms; 

reference to book and chapter). 
E^ Ibid., Brit. Mus., Harley3244, fol. 36-71, s. XIV (photostats). 
E* Ibid., Bodl. Laud. Misc. 247 (edit. James, The Bestiary). 
F Vincent of Beauvais, Speculum Naturale, Nuremberg, 1494. 
H De Bestiis, Books I-II (Migne, Pat. Lat. CLXXVII). 



Carmody: Physiologus Latinus Versio Y 103 

I. INCIPIMUS LOQUI DE LEONE PRIMUM, 
REGE OMNIUM BESTIARUM 

Iacob, benedicens Iiidam filium suum, ait: Catulus leonis luda 
[Gen, 49. 9]. ^Phisiologus dixit, qui sermonum horum scripsit na- 
turas, leo tres naturas habet. Prima natura est: cum ambulat 
olefaciens in monte, uenit ei odor uenatoris; 'et de cauda sua coope- 
rit uestigia sua quocumque ierit, ^ut non sequatur uenator uestigia 
eius, et inueniat cubile ipsius, et capiat eum. — ^Sic et saluator 
noster, spiritalis leo de tribu luda, radix Dauid [Apoc. 5. 5], missus 
a consempiterno patre, ^cooperuit intellegibilia uestigia sua (hoc 
est deitatem suam) a ludeis incredulis: cum angelis angelus, cum 
archangelis archangelus, cum tronis tronus, cum potestate potestas 
[Ascens. Esai. (apocr.)]; Monec discendens descendisset in uterum 
uirginis, ut saluaret quod perierat genus humanum: Et uerbum 
caro factum est, et habitaret in nobis [loh. 1. 14]. «Et hoc, ignorantes 
eum descendentem atque ascendentem, hi qui sursum sunt, dicunt: 
^Quis est iste rex glorie? et angeli deducentes eum responderunt : 
Dominus uirtutum ipse est rex gloriae [Ps. 23. 10]. 

^"Secunda natura leonis est: cum dormierit, uigilant ei oculi, 
aperti enim sunt ei; "in canticis canticorum testatur sponsus dicens: 
Ego dormio, cor meum uigilat [Cant. 5. 2]. — ^^Etenim corporaliter 
dominus meus dormiuit in cruce, deltas uero eius semper in dextera 
patris uigilat [cf. Matt. 26. 64]: ^^Non enim dormit neque dormitat 
qui custodit Israel [Ps. 120. 4]. 

"Tertia eius natura est: cum leena genuerit catulum suum, 
generat eum mortuum; et leena custodit eum tribus diebus; ^Monec 
ueniat pater eius die tertia et, essufflans in faciem eius die tertia, 
suscitat eum. — ^^Sic omnipotens pater omnium tertia die suscitauit 
primogenitum omnis creature [cf. Col. 1. 15] a mortuis. ^^Bene ergo 
Iacob dixit : Catulus leonis luda, quis suscitauit eum? 

Codd.: Y 72 73. 



Inc.: grecis cognomento quoque crisostimus est dictus os aureum quem de 
naturis animalium ordinauit Y tractatus episcopi ortodoxi de natura ani- 
malium Y^. — Tit.: leone rege bistiarum et animalium Y^; regem F^. — 1 
etenim iacob Y^; filiam suam F. — 2 leone tres res naturales F^ et cf. h; ei] 
F* om. F. — 3 ierit] F" gerit F. — 4 sequatur] F & n quator F^ sequens F'. — 
5 noster] F"meus F. — 6aiud. incred.] YY^om. F^ — 6/7 cum ang. ...discen- 



104 University of California Publications in Classical Philology 

dens] Y^ cum YY\ — 7 utero 7". uirg.] add. marie Y^; et hab.] 7" ut hab. 
Y. — 8 ascendentes Y; hii Y. — 9 est rex gloria Y. — 10 leonis] om. F^ — 14 
eius] Y^ enim Y. — 15 suscitauit Y^. — 17 qui Y. 

II. DE AUTOLOPE 

Est animal quod dicitur autolopus, acerrimum nimis, ut nee uena- 
tor possit ei appropinquare. ^Habet autem longa cornua serre 
figuram habentia, ut possit precidere arbores magnas et altas, et ad 
terram deponere. ^Si autem sitierit, uadit ad terribilem Eufraten 
jflumen, et bibit; sunt autem ibi ricine (dicuntur grece), hoc est 
frutices tenues ramos habentes; %t uenit ludens ad illam ricinum 
fruticem, et obligatur in ramis eius. ^Et clamat, uolens fugere, et 
non potest, obligatus est enim ; audiens eum autem uenator, ueniet 
et interficiet eum. 

^O polite nomine, hoc est cuius conuersatio caelestis est [cf. Phil. 
3. 20], ^presumens duobus uis cornibus, detractationibus et uolup- 
tatibus, aduersaria cupiditate et seculi pompa abstine : ^congaudent 
tibi angelorum uirtutes, duo cornua duo testamenta sunt. ^Sed uide 
ne tenearis modicissime huic ricimine, hoc est frutici paruole per 
quam occasione uestitus, et tenearis laqueis eius, et malus uenator 
occidet te (id est diabulus) : ^"Uinum autem et mulieres effugiet sa- 
piens uir [Eccli. 19. 2, Os. 4. 11]. 
Codd.: Y 72 Y\ 



1 adpropinquar& 7^. — 2 magnos et altos 7^. — 3 ricine] infra rinice YY"^. — 
5 autem eum 7^. — 6 cael. est] add. sic et homo 7'. — 7 detractionibus 7; 
pompa] YY^ om. 7^. — 9 uertitus (?) 7^. — 10 uinum] 7 uirum 7^; enim 7'. 

III. DE LAPIDIBUS PIROBOLIS 

Sunt tantum lapides piroboli in partis orientis, hoc est ignei lapides, 
masculus et femina. ^Et quamdiu longe a se sunt, ignis non ardet; 
^si autem appropiauerit feminae masculus, ignis incenditur, et 
incendit omnia. 

*0 generosa epolitetota (hoc est abstinens in omnibus), ^multi 
enim in tartaro sunt propter mulierem, qui in temptationibus 
inciderunt. 

Codd.: 7 7^ YK 



piropolis 7. — 1 partes orientes 7'. — 3 si enim propiauerit ignis incendit 



Carmody: Physiologus Latinus Versio Y 105 

sic et scriptum si autem longius sit ab animam (inuicem?) mortuos est ani- 
mam (?) si autem propiauerit pin scm ad anima (inuicem?) ignitus ambo 
quomodo ait in psalmo [17. 31] eloquia domini igne examinata F'; appropr. Y 
adpropinquauerit F^. — 5 tempnat. Y^ temptatiobus Y; inciderint Y^. 

IV. DE SERRA MARINA 

Bene Physiologus dixit de abstinentibus in omnia, et non in fine 
permanentibus [cf. Matt. 24. 13]. ^Est (inquid) animal in mari quod 
dicitur serra, alas habens longas. ^Et si uiderit naues nauigantes, 
imitatur eas, et exaltat alas suas, et contendit cum nauibus que 
nauigant; ^et cum fecerit stadia triginta aut quadraginta, aut eo 
amplius, laborans in se, alas suas colligit, ^et fluctus eam referunt in 
ueterem suum locum, ubi prius erat. 

^Mare saeculum dictum est, ^naues prophetas et apostolos, qui 
pertransierunt saeculum hoc; *et uirtutes aduersarii, haec autem 
serra, que non permansit cum nauibus transeuntibus, ^horum sunt 
similitudines qui ad tempus sunt abstinentes, et cursu bono cur- 
rentes non permansenint — ^"incipientes opere bono, non perman- 
serunt in fine, propter cupiditatem et superbiam et turpis lucri 
gratiam [cf. Tit. 1. 11], aut fornicationes aut mechias aut odio: in 
quibus fluctibus maris (hoc est contrarie uirtutes) sunt que eos 
deducunt in infernum. 

Codd.: Y Y^ Y^. 



1 de] Y om. Y^. — 2 saerra Y. — 4 collegit Y^. — 5 fructus Y'^. — 7 apostolus 
Y -ulos F2. _ 10 operi Y^; turbi Y"^; mechinas Y; fluctus F^ 

V. DE CHARADRIO 

Est aliud uolatile genus animalis quod dicitur charadrius : in Deu- 
teronomio scriptum est [cf. Deut. 14. 18]. ^Totus albus est, nullam 
partem habens nigram; et interiora eius apudeumata curant his 
quorum oculi caHgant : ^in atriis autem regum inuenitur. '^Et si quis 
est infirmus, ab eodem charadrio cognoscitur si uiuat aut moriatur; 
^et si est infirmitas hominis ad mortem, auertit faciem suam chara- 
drius ab eo, et omnes cognoscunt quia morietur; ^si autem infirmitas 
eius est ad uitam, aspicit charadrius in faciem eius qui infirmatur, 
et infirmus charadrium, ^et absoluit infirmitatem infirmi charadrius; 
^et euolans in aera solis, comburit infirmitatem infirmi, %i spargit 
eam. 



106 University of California Publications in Classical Philology 

^•'Et saluator noster charadrius est, et is qui infirmatur bonam 
personam accipit saluatoris nostri: totus enim albus est saluator, 
non habens maculam aut rugam [cf. Eph. 5. 27]; "dicit enim in 
euangelio ipse dominus quia : Princeps mundi huius ueniet et in me 
nihil inueniet [loh. 14. 30], etenim: ^^Peccatum non fecit, nee inuen- 
tus est dolus in ore eius [I Petr. 2. 22]. ^^Ueniens enim de celo ad 
populum ludeorum, auertit ab eo suam deitatem, dicens: Ecce 
demittetur domus uestra deserta [Luc. 13. 35]; ^%enit ad nos gentes, 
tollens infirmitates nostras [cf. Matt. 8. 17], ^^et peccata nostra 
portans [cf. Esai. 53. 4], exaltatus in ligno crucis: ^^Ascendens in 
altum, captiuam duxit captiuitatem [Ps. 67. 19], dedit dona homini- 
bus [Ps. 67. 19, Eph. 4. 8]. 

^^Sed dicis quia charadrius inmundus est secundum legem, ^^et 
quomodo fert personam saluatoris? Et serpens inmundus est, et 
testatus est ipse dominus de se in euangelio, ^Micens: Sicut Moyses 
exaltauit serpentem in herimo, ita exaltare oportet filium hominis 
[loh. 3. 14] ; ^''et serpens prudentior dictus est, ^^simiUter et de leone, 
et alia multa: ^Muplicia autem sunt creatura, laudabiHa et uitu- 
perabilia. 

Codd.: Y Y^ Y'. 



de charadra Y^. — 1 alius Y. — 2 aputeu. Y^. — 4 caradrio Y; cognoscit Y. 
— 6 est] YY^ om. Y'^; caradrius F. — 7 caradrius Y. — 10 his Y^; accepit F*. — 
1 1 dixit Y. — 12 faecit Y. — 13 iudeos Y^ iudaicum Y add. cum Y^. — 17 carad- 
rius Y. — IQmoysen Y. — 22dublcitia ... creture F*. 

VI. DE PELICANO 

Dauid in psalmo centesimo primo dicit : Similis f actus sum pelicano 
in solitudine [Ps. 101. 7]. Thisiologus dicit de pehcano quia amator 
est filiorum nimis; ^si autem genuerit natos et paruuli increuerint, 
incipiunt percutere parentes suos in f aciem; parentes autem colaphi- 
zantes eos Occident ; Meinde misericordia moti parentes tribus diebus 
flent eos, lugentes quos occiderunt ; et tertia die mater eorum per- 
cutiens costam suam destillat suum sanguinem super eorum corpora 
mortua (id est puUorum) : ^ipse sanguis suscitat eos de morte. 

^Sic et dominus noster increpat per Esaiam dicens : Filios generaui 
et exaltaui, ipsi autem spreuerunt me [Esai. 1.2]. ^Genuit nos factor 
omnis creature: percussimus eum. ^Quomodo percussimus eum? 
quoniam seruiuimus potius creature quam creatori [cf. Rom. 1. 25]. 



Carmody: Physiologus Latinus Versio Y 107 

^Ascendens autem in altitudinem crucis, percusserunt impii latus, 
et aperuerunt eius costam ; et exiit sanguis et aqua [cf. loh. 19. 34] in 
uitam eternam [cf. loh. 6. 55] — ^^sanguis propter quod dixit: 
Accipiens calicem gratias egit [Matt. 26. 27, Luc. 22. 17], aqua 
autem propter baptismum penitentie [Marc. 1. 4, Luc. 3. 3], "qui 
dixit: Me derelinquerunt fontem aque uite gratis, et cetera [Hier. 
2. 13]. i^Bene ergo Phisiologus arguit de pelicano. 
Codd.: Y Y^ YK Versio alt.: C M. 



1 solitudinis YK — 3 sic Y^. — 4 destillauit Y stillat C stillans YK — 6 sic] Y 
ita C om. Y^; noster] CF" meus Y; esaim 7.-9 costam & iit YK — 12 peli- 
gano Y. 

VIL DE NICTICORACE 

Item in eodem psalmo centesimo primo de nicticorace dicit : Factus 
sum sicut nycticorax in domicilio [Ps. 101. 7]. ^Nicticorax tenebras 
amat magis quam lucem; hoc genus uolatile est. 

^Dominus autem noster Ihesus Christus amauit qui in tenebris et 
umbra mortis erant [cf. Esai. 9. 2], populum gentium et populum 
ludeorum, *qui tunc fiUorum adoptionem [cf. Gal. 4. 5] et patrum 
promissionem [cf. Rom. 15. 8] habuerunt. ^De hoc et saluator dixit: 
Ne timueris, pusillus grex, %uoniam conplacuit patri dare nobis 
regnum [Luc. 12. 32]; ^et propheta dicit: Uocabo non populum 
meum, et non dilectam dilectam [Rom. 9. 25, cf. Os. 2. 24]. ^Sed 
dicis (secundum Deuteronomium) quoniam inmunda auis est nycti- 
corax: ^et apostolus ideo dixit: Qui non cognouit peccatum pro nobis 
peccatum fecit [II Cor. 5. 21]; ^^et: Omnibus omnia factus est, ut 
omnis lucri /acere^ [I Cor. 9. 22]. 

Codi.: Y F« Y\ Sim.: C. 



de nicticorace] om. FF^. — 1 in eodem in ps. centissimo F*; noetic. F; 
nyct.] F. — 2 noct. F. — 3 domine F^; que Y\ umbre F^. — 4 promissione F. 
— 6 patre F=; regum F^ — 7 pop. meum] YY'^ plebem meum plebem meum 
F^.— Snoot. F. 

VIII. DE AQUILA 
Dauid in psalmo centesimo secundo: Renouabitur sicut aquilae 
iuuentus tua [Ps. 102. 5]. ^Phisiologus dicit de aquila quoniam cum 
senuerit, grauantur ale eius, et oculi eius caliginant. ^Quid ergo 
facit? querit fontem aque, et euolat in aera solis, ^et incendit alas 



108 University of California Publications in Classical Philology 

SUES et caliginem oculorum suorum; ^et discendit in fontem, et 
tingit se per ter, ^et renouabitur et nouus efficitur. 

^Ergo et til, si iiestitum habes ueterem, et caliginant oculi cordis 
tui, quere spiritalem fontem dominum; ^quid: Me derelinquerunt 
fontem aque uite et cetera [Hier. 2. 13]. ^Et euolans in altitudinem 
solis iustitie [Mai. 4. 2], qui est Christus Ihesus (siciit apostolus 
dicit), et ipse incendet uetus uestimentum tuum diabuli. ^^Prop- 
terea et duo presbiteri illi in Danielo sic audierunt inueterate 
dierum malorum [cf. Dan. 13. 52], et baptizare in sempiterno fonte, 
^^expolians ueterem hominem cum actibus suis, et induens nouum : 
Qui secundum dominum creatus est [Eph. 4. 24], sicut apostolus 
dicit. ^^Propterea et Dauid ait: Renouabitur sicut aquile iuuentus 
tua. 

Codd.: Y Y^ Y\ Trans, alt.: C. 



4 suas] CF^ suam Y. — 9 incendat Y^; tuum] om. Y. — 11 hominem uet. Y'^. 
— 12aquilaCF2. 

IX. DE PHENICE 

QuoNiAM dixit saluator in euangelio: Potestatem habeo ponendi 
animam meam, et potestatem habeo iterum sumendi earn [loh. 
10. 18] : ^et ludei irati sunt in uerbo eius. ^Est enim uolatile genus in 
Indie partibus quod dicitur phenix; que auis per quingentos annos 
intrat in ligna Libani, et implet duas alas suas aromata. ^Et signifi- 
cat sacerdoti Heliopoleos (ciuitas hec que Heliopolis nomen habet) 
^in mense nouorum, hoc est Adar, quod grece dicitur Farmuti uel 
Phamenoth. ^Sacerdos autem cum significatum ei fuerit, intrat et 
implet aram lignis sarmenteis; ^et intrans ilia auis in Heliopolim 
onerata aromatibus, ascendit super aram, %t sibi ipsa ignem incen- 
dit, et seipsam conburit. ^In crastinum enim sacerdos scrutatur 
aram, et inueniet uermem in cineribus. ^"Secundo autem die inuenit 
auiculam pusillam. "Tertio die inueniet aquilam magnam: ^-et 
euolans, salutat sacerdotem, et uadit in antiquum locum suum. 

^^Si ergo hoc uolatile genus banc talem potestatem habet seipsum 
sic occidere et seipsum sic suscitare, "quomodo stulti homines 
irascuntur in uerbo saluatoris : Potestatem habeo (dicentis) ponendi 
animam meam, et potestatem habeo iterum sumendi eam? ^*Phoe- 
nix personam accipit saluatoris: etenim descendens de celestibus 
locis, duas alas suas boni odoris sui plenas deposuit (hoc est optimos 



Carmody: Physiologus Latinus Versio Y 109 

sermones suos) : ^^ut et nos, extendentes manus operationes, remit- 
tamus ei bonum odorem [cf. II Cor. 1. 15] spiritale in operibus 
bonis. i^Bene ergo Phisiologus dicit de phenice. 
Codd.: Y Y^ Y\ 



tit. de ph.] om. 77". _ 2 eius] add. de phenice YYK — 5 noborum YK — 8 
igne 7. — 14 irascantur 7; dicentis] 7 dicentes 7^ discendi 7"; eum 7. — 15 
accepit 7^; optimo 7. 

X. DE EPOPE 

Dicit enim lex: Qui maledixerit patri aut matri, morte morietur 
[Ex. 21. 17]. 2Et qiiomodo sunt patricide aut matricide? ^Est auis 
que dicitur epopus : si uiderit suos parentes senuisse, et eorum oculos 
caliginasse, "filii uellunt pinnas parentum, et elingunt oculos eorum, 
et calefaciunt parentes suos sub alis suis, ^et nutriunt parentes suos, 
uicem eis reddentes, et nutriunt sicut pullos suos, ^et noui fiunt 
parentes ipsorum; ^et quodam modo dicunt parentibus suis: sicut 
uos laborastis nutrientes nos, ^sic et nos similiter uobis facimus. 

^Et quomodo inrationabiles homines non amant parentes suos? 
^°Bene ergo Phisiologus arguit de epopo. 

Codd.: YY^ — AY'. 

tit. de ep.] om. YY^^K — 1 patrem A 7"; aut] et 7^; matrem A 7". — 3 paren- 
tes suos A 7^; et eorum oculus et eorum oculus 7^; caligasse YY^. — i euellunt 
AY'; lingunt AY'; parentos 7; sub alis suis] YY' om. ACY'^.—5 et nut- par- 
suos] 7 om. A 7"; uicem eis red-] YY^ om. A Y'; redentes 7. — 7 et qu- modo] 
YY^om.AY'. — Set nos] YY^ om. AYK— 9 et quo-] AY om. 7^ par- suos] adrf. 
Ex. 21. 17 Y'. 

XI. DE ONAGRO 
In lob dicitur: Quis est qui dimittit onagrum liberum [Hiob 39. 5]. 
^Phisiologus dicit de onagro quoniam est gregis primus in eis; et si 
generauerit greges masculos, pater eorum confringit necessaria 
eorum, ut non f aciant semen. 

Tatriarche semen carnale querebant creare ; apostoli autem spiri- 
tale filios carnales obtinuerunt, ut semen caeleste possiderent, 
^sicut dicit: Laetare sterilis quae non paris, et cetera [Esai. 54. 1]. 
^Uetus testamentum semen annuntiat, nouum autem abstinentiam 
predicat. 

Codd.:YY^ — AYK 



1 quis dimisit A. — 2 phis- ... quoniam] YY"^ masculus A om. Y'; generaue- 



110 University of California Publications in Classical Philology 

rint Y^ genuerit grex AY^; masculus F^; uti ne A ut ne Y^. — 4 dicit] add. 
spiritus sanctus per prophetam A; sterelis Y. 

XII. DE UIPERA 

Johannes dicit ad Pharisees : Generatio uiperarum [Matt. 3. 7, Luc. 
3. 7]. ^Phisiologus dicit de uipera quoniam faciem habet hominis 
masculus, ^femina autem mulieris usque ad umbilicum, ''ab umbilico 
autem usque ad caudam corcodrilli habet figuram; ^porro femina 
non habet secretum locum, id est menbrum pariendi sinum, sed ut 
foramen acus habet. ^Si autem masculus habeat cum femina, 
effundet semen in os femine, ^et si biberit semen eius, precidet 
femina necessaria masculi (hoc est uirilia), et moritur masculus; 
^cum autem creuerint filii eius in utero matris sue, non habens ilia 
sinum unde pariatur, tunc filii adaperiunt latus matris suae, ^et 
exeunt occidentes matrem. 

^''Assimilauit ergo saluator noster Phariseos uipere: sicut hec 
generatio occidit patrem et matrem, sic et hie populus, qui sine deo 
est, patrem suum Ihesum Christum et matrem terrestrem Hierusa- 
lem. ^^Et quomodo fugiunt ab ira uentura? [cf. Luc. 3. 7]: ^-pater 
autem noster Ihesus Christus et mater ecclesia uiuunt in eternum, 
^^ipsi autem in peccato uiuentes mortui sunt. 

Codd.: FF" AC — H{2. 21)L — M. 



1 bene dixit ad phar. AC; uip.] + quis uobis dixit fugere a uentura ira AC 
progenies (Matt.) uip. quis ostendet (Luc.) HL. — 2 dicit] LMYY^^ monet 
AC. — 4 ad] om. Y. — 5 pocro] YY^ om. ACM; secretum ... sinum] FF^s 
uadum etc. AM; ut] CYY' sicut A et M. — 6 et si AC; habeat] YY^ fiat ACY^ 
quando coitum facit HL et masc. uoluerit cognoscere M; effundit F; os eius 
F' OS suum HL caput suum L^. — 7 ebiberit AC deglutiens HL; necessaria] 
ACYY^ caput HL"^ uirilia L; hoc est uir.] YY^ om. ACM. — 8 increuerint Y; 
pariat AC; tunc] YY^ om. ACM. — 9 et exeuntes A et sic foris exeunt M; occ. 
matr.] YY^ Occident patrem et matrem A patrueli et matrueli C patrolae ergo 
sunt et matrolae M mortua matre HL. — 10 bene ergo simil. AC; noster] AF* 
mens F; qui] YY^ om. AC; suum] YY^ om. AC. — 11 et] YY^ om. AC. — 13 in 
pec. uiu.] FF^ peccatores A. 

XIII. DE SERPENTE 

Saluator dicit: Estote prudentes sicut serpentes, et mites sicut 
columbe [IMatt. 10. 16]. ^Serpens tres naturas habet : 'prima eius na- 



Carmody: Physiologus Latinus Versio Y 111 

tiira hec est: cum senuerit, caligant ei oculi; et si uoluerit nouus 
fieri, abstinet et ieiunat qiiadraginta diebus, donee pellis eius re- 
laxetur a came sua; ^et si relaxata fuerit ieiunis pellis eius, uadens 
querit fissuram angustam in petra, ^et intrat in fissuram, et contri- 
bulat se, et cogit, et deponit pellem ueterem, et nouus fit. — ^Et nos, 
per multam abstinentiam et tribulationem, pro Christo deponemus 
ueterem hominem et indumentum eius. ^Sed et tu, quere spiritalem 
petram Christum, et angustam fissuram: ^Augusta enim est porta, 
et cum tribulatione uia, que ducit ad uitam, et pauci sunt qui 
introeunt per eam [Matt. 7. 14]. 

^Secunda eius natura est : cum uenerit aquam bibere ad flumen, 
non adfert secum uenenum quod gestat in capite, sed in fouea sua 
illud deponit. — ^"Debemus ergo et nos, cum in collectam uenere- 
mus, aquam uiuam et sempiternam haurire [cf. loh. 4. 15]; ^^cum 
audimus diuinum sermonem et celestem in ecclesia, non nobis cum 
uenenum inducamus (hoc est terrestres malasque concupiscen- 
tias [cf. Col. 3. 5]). ^^Multi enim insipientes non uenerunt in illam 
spiritalem cenam; alii quinque iuga bouum ementes, alii agrum 
querentes, alii uxorem ducentes [cf. Luc. 14. 16-24]. ^'Sicut in 
euangelio dicit: Reddite ergo omnibus debita, cui timorem timo- 
rem, cui honorem honorem, cui tributum tributum, et cetera [Rom. 
13. 7]. 

^^Tertia eius natura est: si uiderit hominem nudum, timet eum; 
si autem uiderit eum uestitum, insilit in eum. — ^^Sic et nos spiri- 
taliter intellegimus, quoniam cum primus homo pater noster Adam 
nudus fuisset in paradyso, non preualuit serpens insilire in eum; 
^^sed quando tunica indutus est (hoc est mortalitatem carnalis cor- 
poris peccati), tunc exiliuit in eum. ^^Si ergo et tu habes uestem 
mortalem, hoc est ueterem hominem, et uis audire inueterate die- 
rum malorum [Dan. 13. 52], exiliet in te serpens; ^*si autem expolies 
te indumento principum et potestatum seculi huius rectorum et 
spihtus nequitiae in celestibus (sicut apostolus dixit [cf. Eph. 6. 12]), 
*Hunc non potest insilire in te serpens. ^''Bene ergo dixit Physio- 
logus de serpente. 

[^^Quarta quoque natura serpentis: quando uenerit homo et 
uoluerit eum occidere, ^^totum corpus tradit ad penam, caput autem 
suum custodit. ^^Debemus et nos in tempore temptationis totum 
corpus tradere, caput autem custodire, id est Christum non ne- 



112 University of California Publications in Classical Philology 

gantes ; sicut f ecerunt sancti martyres : Omnis enim caput Christus 
est [I Cor. 11.3].] 

Codd.: FF" AM — P — DH{2.2l)L — U^ — EE^H^{Z.53) — F(21.7, §§ 3-5, 
9,12) — CMK 



1 estote] + ergo AL'^; mites] YY'^ simplices AC; sicut] ACY^ ut F. — 2 tres 
naturas] AEH'^Y'^ tres res naturales Y quatuor CM. — 3 ei oculi] YY^ oculi 
eius AC impedimentum habent DHL; abstinet] + se CE + a cibo Pf^ a cibis 
M; diebus] + et quadr. noctibus DLMP; eius] + ieiunis YY^; a carne ... pellis 
eius] YY^. — 4 ieiuniis Y; uadens] DHLYY^; angustam rimam EH^. — 5 tran- 
situm ... expoliatur DHL; et cogit] FF^; et nouus fit] M et iuuenis fit L et 
iuuenescit DH om. AYY^. — 6 -ionem] EYY^ -iones AC; deponimus A -amus 
E. — 7 phissuram F. — 8 angusta ... uia] YY^ id est ang. portam AC{D) + et 
artauiaL; porta] petra F; sunt qui] FF^'pntrant A.— 9uenerit]+ad F^-pergit 
ad bibendum DHL; adf ert] CFMF auf . F^ adferit F^ portat AE; quod . . . capite] 
YY'^ suum AEFY^; in cubili suo CFM. — 10 ueneremus] YY^ uenimus AEL^Y^ 
ueniamus L + ad Fet F^; haurire] loh. haurientesACE'F'aurimus (?) YY^. — 
11 cum audimus] F cum audiuimus F^ audire AEL audiunt L^; non ... hoc] YY^ 
abicere a nobis uenenum hoc (id E) ACE; malasque] YY^ secularesque L et 
malas ACEYK — 13 dixit F". — 14 natura] + sic Y^; timens fugiat DHL; 
exsilit AY\ — 15 quia EE^H; cum] YY^^ om. AE'^HL; Adam] + quamdiu 
AE; nudus fuisset] YY"^ fuit nudus AE nudus fuit F'; insilere F. — 16 quando] 
AYY"^ postquam EHm; tunica] EHH^Y"^ -icis AYY^; induti sunt YY^; car- 
nalis] YY^ om. AEY^; peccati] YY^ om. AEY^. — 17 et tu] YY'^ o homo i/= om. 
AEE^; uis] A om. YY"^; exiliens F^. — 18 expolias F; indumento] A^^-tum YY^. 
— 19 potest] + iam YY^; insilire] YY^ exilire AEH. — 20 YYK — 21 CM^; 
quoque] M^ om. C; occ. eum C. — 22 CM^; ad penam] M^ om. C; suum] AP om. 
C. — 23 C versio altera L". 

XIV. DE FORMICA 

Salomon in prouerbiis dixit : Uade ad f ormicam, o piger, et imitare 
iiias eius [Prou. 6. 6]. ^De' formica dixit Phisiologus tres naturas 
habere. ^Prima eius natura est: cum ordinate ambulauerint, \ma 
queque granum baiulat in ore suo; ^he quae uacue sunt, nihil ha- 
bentes formice, non dicunt habentibus: date nobis granum uestrum 
[IMatt. 25. 8]; ^sed uadunt per uestigia priorum,et ueniunt in locum 
ubi et ille frumentum inuenerunt ; et tollentes et ipse, deportant in 
cubile suum. — ^Et hec quidem dicta sunt de inprudentibus et 
inrationabilibus. ^Ue autem uirginibus illis que petierunt a pruden- 
tibus, dicentes: ^Date nobis oleum de uasis uestris, quoniam extin- 
guntur lampades nostre [IVIatt. 25. 8]; ^"cum essent rationabiles et 
habentes intellectum, hec autem audierunt: "ab eis non possumus, 
ne forte non sufficiat, et nobis et uobis. 



Carmody: Physiologus Latinus Versio Y 113 

^'^Secunda eius natura est: quando recondit grana in cubili suo, 
diiiidit ea in duo, ^^ne forte hiemps conprehendat ea, et infundente 
pluuia germinent grana, et fame pereat. — "Et tu, iierba ueteris 
testamenti diuide ab spiritali uerbo earnalia, ^^ne quando germinat 
littera, te occidat. "Paulus autem apostolus dicit quoniam: ^^Lex 
spiritalis est [Rom. 7. 14]; ^^t iterum dicit: Littera occidit, spiritus 
autem uiuificat [II Cor. 3. 6]; ^^et iterum: Que sunt uiuificantia, 
haec enim sunt duo testamenta [Gal. 4. 24]. ^"ludei autem solam 
litteram aspicientes, fame necati sunt; ^^et prophetarum homicide, 
sed et dei, facti sunt, radentes uirgas ut pariant ones [cf. Gen. 
30. 37 seq.], et carnales circumcisiones, et sabbatismos, et sceno- 
phagias. -^Haec autem omnia spiritalia sunt et intellegibilia [cf. 
Esai. 7. 9]. 

^^Tertia eius natura est: uadit in agrum tempore messis, et ascen- 
dit in spicam, et deponit granum. ^^Priusquam autem ascendat 
spicam, odorat deorsum spice, et ab odore cognoscit si triticum est 
aut ordeum; ^^et si est ordeum, statim refugit ad tritici spicam. — 
^^Hordeum autem esca est pecorum; et in lob dicit: Pro tritico 
procedat mihi hordeum [Hiob 31. 40]. "Et propheta: Fugi de Babi- 
lonia, fugi de terra Chaldeorum [cf. Hier. 50. 8], ^^oc est, fugi 
aliena doctrina aliene glorie, que ordei sunt escae, occidentes animas 
(hec et inimica ueritatis dicta sunt, et sunt). Bene de formica 
dictum est. 

Codd.: Y F2 7M(§3-6, 16-18, 20-22). 



4 una qu tibi gr. Y^; baiolat Y. — 5 haec Y^; et ea quae uana sunt A add. et 
Y^. — 7 de] + animalibus B. — 8 a] Y om. F^. — g lamp. nos. ext. Y^; lampade 
F. — 11 et nobis] nobis YK — 12 cubil Y. — 13 germinant Y. — 14 ueteri Y; 
uerba Y^. — 15 germinat] B om. YY^; occidit Y. — 18 spiritus] + sanctus Y^; 
autem] om. F^. — 21 et] YY^ om. A; homic. proph. A; dei ... uirgas] YY^ 
iacob tradens uirgas A; sabbata AY^; scenosfegias F^ soenophegia A. — 22 
haec autem omnia figuraliter sunt intellegenda A. — 28 sunt et ... dictum] 
om. F^. 

XV. DE SYRENA ET ONOCENTAURO 

SiGNiFiCAUEEAT ante Esaias propheta dicens: Syrene et onocen- 
tauri et demonia et herinacii uenient in Babilonia et saltabunt 
[Esai. 13. 21]. ^Unius cuiusque naturam Phisiologus disseruit, 
dicens de sirenis, quoniam animalia mortifera sunt in mari, clami- 
tantia uocibus aliis; ^etenim dimidiam partem usque ad umbihcum 



114 University of California Publications in Classical Philology 

hominis habent figuram, dimidio autem uolatilis. ^Similiter et ono- 
centauri, a pectore et sursum hominis habet figuram, ^deorsum 
autem asini. 

^Sic et uir duplex corde, indispositus in omnibus uiis suis [lac. 
1. 8]. ^Ita sunt malorum negotiatorum, et actus anime; ^n ecclesia 
quidem colliguntur, absconse autem peccant: ^Habentes quidem 
(et apostolus dixit) speciem pietatis, uirtutem autem eius dene- 
ganies [II Tim. 3. 5]. ^"Et in ecclesia anime quonimdam sicut ones 
sunt; ^^cum autem dimissi fuerint a collecta, fiunt tamquam pecora: 
Et assimilabuntur iumentis insensatis [Ps. 49. 20]. ^^Hi tales siue 
syrene siue onocentauri figuram ostendunt aduersariorum. 

CoM.:YY^Y\ 



tit. et] edit. om. YY^. — 3 dimidio autem] Y^ om.YY^', uolatilis] graec. uola- 
tiles F' mulieris FF*. — 4 pecatore F^. — 6 indisp.] add. est YK — 7 negotao- 
rum F. — 8 in absconso Y\ — 9 dicit F^— 11 autem dimissi] YY^ audis nisi 
F2;p&cora F. — 12hii FF^. 

XVI. DE HERINACIO 

HiRiNACius autem species non habet sicut pile, totus autem spinis 
plenus est. Thisiologus dixit de eo quoniam ascendit in botrum 
uitis, ^et deicit acinas in terram (hoc est uuas); ^et deuoluens se 
super eas, et adherens fructus uitis eius spinis, affert filiis suis, et 
remittit racemum botrui uacuum. 

^Et tu, christiane, abstinens operari in omnibus, ^astitisti enim 
spiritale nine, propter quod afferis in spiritali torculari, et recon- 
deris in atriis dei regis, in sancto tribunali Christi, et uitam etemam 
capies; ^quomodo dimisisti ilium spiritum nequissimum ascendere 
in locum tuum, «et talem tuam abstinentiam dispersit. ^Et aculeis 
mortis te fefellit intantum, ut spoliam tuam aduersariis uirtutibus 
diuideret. ^^luste ergo Phisiologus statuit naturas animalium spiri- 
talibus rebus. 

Codd.:YY^AY\ 



tit. F" om. Y de herinace hycesi YK — 1 speciem Y^; non] om. F*; est] 
AY^ om. Y. — 2 dicit A; asc] add. autem AY^ add. in botro A. — 3 racemos 
A F'; ubas F. — 4 uoluens ... eas adheret spinis A ; af err& Y^ afferet A ; dimit- 
tit A; uotri F" botri F. — 6 spiritali AY^; uini F" uineae A; afferes yl; regis] 
add. quod affertur AF^; sanctum tribunal A. — 10 rebus] A F regibus F^. 



Carmody: Physiologus Latinus Versio Y 115 

XVII. DE HIBICE 

Est animal quod dicitur hibicis; inmundum quidem est, secundum 
legem, pre omnibus uolatilibus [cf. Leu. 11. 17]. '^Natare nescit, sed 
secundum litorem fluminum uel stagnorum depascit: %on potest 
natare in altitudinem, sed ubi inmundi pisciculi demorantur; 
*inueniturque foris ab altissimis locis. 

^Disce nunc spiritaliter natare, ut uenias in intellegibile et spiri- 
talem altum flumen, et ad altitudinem sapientie uirtutis dei [cf. 
Rom. 11. 33]. ^Si uis ascendere in altitudinem, et mysteria domini 
Ihesu Christi discere, disce spiritaliter natare. ^Nisi enim duas 
manus extenderis et feceris typum crucis, non poteris pertransire 
mare; ^et nisi tu uolueris pertransire seculum ad deum per typum 
crucis, omnia scandala non uitabis: ^nescient enim natare neque 
orare scientes, foris depascuntur ab ecclesia. ^"Foris autem a fide 
sunt fornicationes, moechiae, detractationes, cupiditates. "Radix 
enim omnium malorum est cupiditas: etenim typus crucis super 
omnem creaturam est. 

i^Et sol, nisi extenderit radios suos, non poterit fulgere. ^^Et luna, 
si non extenderit sua dua cornua, non lucebit. '^Nam et uolatilia 
celi, nisi extenderint alas suas, non poterint uolare. ^^Nauis quoque, 
nisi extensum fuerit uelum eius cum uelificat, uentis flantibus non 
mouetur ad nauigandum. ^^Sicut et Moyses, extendens manus suas, 
interfecit Amalech [cf. Ex. 17. 11]; Danihel, cum oraret, leones 
euasit; ^^lonas in uentre coeti similiter orando seruatur; Thecla in 
ignem missa est, et in foueis bestiarum, et typus eam saluauit 
crucis. ^^Et Susanna a presbiteris liberatur, ludit sic occidit Olo- 
fernem dextera forti, Hester Aman, et tres pueri in camino ignis: 
^Hali custoditi sunt signo, et alia multi sancti peiora omnium passi 
sunt genera tormentorum. ^oBene ergo Phisiologus de hibice dixit. 

Codd.: YY^ A{m-Q)Y'. 



1 quod] Y que Y^. — 2 incip.: ibit circa labia flum. AY^; nature F^; nescat 
Y; litore F. — 3 natare] om.A.—5 nature F"j ad] Y om. 7"; dei] F" om. Y; 
uinias intellectuales altum fl. altum diuitiae sap. et uirtutes A. — 6 ascen- 
dere] AY^ desc. YYK — 8 non] Y^ om. Y. — 10 mechie Y; detractiones YY^; 
cupiditas Y. — 11 est enim Y; omnis creaturas Y^. — 14 -erit ... -erit Y. — 15 
extentum Y. — 16 extend& 7^; euasit] 7 efacit Y\ — 17 tecla 7; saluabit Y"^. 
— 18 iudit] add. & 7*; olophorne in dex. Y\ — 19 paiora Y\ 



116 University of California Publications in Classical Philology 

XVIII. DE UULPE 

UuLPis dolosum animal est omnimodo, et dolos parat. ^Si esurierit 
et non inuenerit quod manducet, requirit ubi est scyssura terre et 
palearum, ^et proicit se supina, sursum respiciens, et adducit flatos 
suos infra se: exspanditiir omnino. "Et putant uolatilia quoniam 
mortua est ; Mescendunt super earn ut manducent earn, et rapit ea, 
et extenderat, et mala morte uolatilia ipsa moriuntur. 

•^Et diabulus omnino mortuus est, et actus operi eius : ^qui uult 
communicare carnium eius, morietur; ^etenim cames eius sunt 
fornicationes, cupiditates, uoluptates, aduersantes seculi [cf. Matt. 
15. 19]. ^Ex hoc et Herodes similatus est uulpi [cf. Luc. 13. 32]; ^"et 
scriba, audiens autem saluatorem: Uulpes (inquid) foueas habent 
[Matt. 8. 20, Luc. 9. 58]. "Et in canticis canticorum: Capite nobis 
uulpes pusillas, exterminantes uineas [Cant. 2. 15]. ^^Et Dauid in 
psalmo LXII: Partes uulpium erunt [Ps. 63. 9]. Bene ergo Phisyo- 
logus dicit de uulpe. 

Codd.: Y F^. Trans, alt.: B Y^etc. 



1 uulpes y^ — 2 scyssure Y fissura Y^. — 3 suspiciens Y; flatus 7.-5 ex- 
tenterat F; moriunt F.— 10 saluatore F.— 11 uineas / 20. 33 obediens] deest 
fol. Y\ 

XIX. DE ARBORE PERIDEXION ET COLUMBIS 

Est arbor que dicitur peridexion, inuenitur in India. ^Fructus autem 
arboris illius dulcis est totus ualde, et suauis. ^Columbe autem delec- 
tantur in fructu arboris illius: habitant autem in ea pascentes fructu 
eius. "Inimicus est autem dracho columbis, timet autem arborem 
illam et umbram eius, in qua columbe demorantur: ^et non potest 
draco adpropiare columbis neque umbre eius. '^Si enim umbra 
arboris uenerit ad occidentem, fugit draco ad orientem; ^si iterum 
uenerit umbra eius ad orientem, fugit ad occidentem. *Si autem fiet 
ut columba inueniatur foris arborem aut umbram eius, ^et inuenerit 
earn draco, occidit earn. 

i^Arborem patrem omnium dicit, umbram autem in filium Chris- 
tum; "sicut dixit Gabrihel ad Mariam: Ne timeas, Maria, spiritus 
sanctus superueniet in te, et uirtus altissimi obumbrabit tibi [Luc. 
1. 30, 35]. ^Tructum autem celestem sapientiam dixit, columbam 



Carmody: Physiologus Latinus Versio Y 117 

autem in spiritum sanctum dixit; "uide ergo, o homo, ne postquam 
acceperis spiritum sanctum (hoc est spiritalem cohimbam intellegi- 
bilem, de celo descendentem et manentem super te), "ne fias foris a 
deitate, ahenus a patre et fiho et spiritu sancto, et draco te interi- 
mat (hoc est diabukis). ^^Non potest enim draco adpropiare ad 
arborem, neque ad umbram, neque ad fructum arboris: et tu, si 
habeas spiritum sanctum, non potest tibi draco adpropinquare (hoc 
est diabulus). ^^Bene ergo de arbore et fructu et umbra eius diximus. 
Codd.: Y— Y'AEE^*H{3.39). 

1 inc. haec arbor inuen. A; in indiae partibus EY^. — 2 illius] YY^ huius 
AE'^*; totus dulcis est YY\ — 3 delectant AE^; habitant autem] AY^ hab. 
enim 7; pasc] + et de Y; eius] + accipientes 7. — 4 in qua] YY^ ubi AE^*H. 
— 5 et]AE'^Y^ in arbore Y; draco]+ascendere uel Y; adpropriare Y; columbis] 
7-ae y^arbori AE'^^H; eius] AE^^Y^arhoTis F. — 6fugiet 7.-6/7 ad] AE^^Y^ 
in 7.— 7 iterum] A Y^ autem 7 om. E^; f ugiet draco in occ . 7.— 8 fiet] A 7 fiat 7^ 
euenit E^; foris ab arbore uel umbra 7. — 10 pat. omnipotentem dicit A; dixit 
7; autem in] 7 om. AE^Y^; christum] 7 om. AE^Y^. — 11 sanctus] AE^Y^ 
domini 7; superueniet] A ueniet super 7^ adueniet 7; obumbrabit] E^Y^ 
-auit A 7. — 12 autem in] 7 om. AE^Y'; dixit] 7 om. AE^Y\ — 13 ne] + 
quando 7; acceperis] AE^^Y^ accip. 7; intell.] + columbam 7; descend.] 
AE-^Y^ uenientem 7; te] + uide 7. — 14 ne YY^ om. AE""; deitate] YY^ eter- 
nitate AE"^ + et 7; hoc est] AE'^Y^ am. 7. — 15 habeas] + caelestem AY^; 
draco] + draco 7; adpropiare A 7^. — 16 7. 

XX. DE ELEPHANTO 

Est animal qui dicitur elephans. ^In eo non est concupiscentiae 
coitus. AHud autem est tragelaphus. ^Si autem uoluerit fihos facere, 
uadit in orientem prope paradysum ; *est autem ibi arbor que dicitur 
mandracora; ^et uadit ibi cum femina sua, que prior accipit de 
arbore, et tradit uiro suo, et seducit eum, donee manducet. ^Et cum 
manducauerit mascukis, statim femina in utero concipiet. ^Et si 
tempus aduenerit ut pariat, ^exit in stagnum aque; et ipsa aqua 
uenit usque ad ubera eius (hoc est femine), et demittit partum suum 
(hoc est fiUum) ; Mimittit eum in aquis, et natat puUus quia natus 
est; ^"et uenit in femora eius, et de ubera matris sue nutrietur. 
^^Elephans autem custodit eam parientem, propter serpentem, quia 
inimicus est serpens elephanto. ^^Si autem inuenerit serpentem, 
occidit eum et conculcat, donee moriatur. 

^^Hec est natura eius: si autem ceciderit, non potest surgere. 



118 University of California Publications in Classical Philology 

Quomodo cadit, cum in arborem se reclinat? ^^Non enim habet 
coniuncturas geniculomm lit obdormiat, si uellit; ^^uenator autem, 
qui eum uellit uenari, incidit arborem, modico minus, ubi reclinare 
consueseit; ^^ut dum uenerit et sese reclinauerit elephans, cadat 
arbor et animal cum eadem ipsa, et simul cadent. ^^Clamat autem 
elephans: statim inde exit magnus elephans, et non potest eum 
leuare; "^iterum clamant ambo, et ueniunt alii duodecim elephanti, 
et non possunt nee ipsi leuare eum qui cecidit; ^Meinde clamitant 
iterum omnes, et statim uenit pusillus elephans, et mittit os suum 
promuscidem subtus ilium magnum, et eleuat ilium. ^oRabet autem 
pusillus elephans naturaUter, ut, ubi incensum fuerit de capilUs 
eius uel de ossibus, neque draco neque ahquid malum ibi accidit. 
2iMagnus ergo elephans et muher eius personam accipiunt Ade et 
Eue. 22Cum autem essent in uirtute (hoc est placentes domino), 
ante suam preuaricationem, non sciebant coitum, neque intellectum 
mixtionis sue carnis habuerunt; 23quando autem muher mandu- 
cauit de ligno (hoc est intellegibilem mandracoram), deinde dedit 
uiro suo, pregnans malorum facta est; propter quod exierunt de 
paradyso. ^^Quamdiu autem fuerunt in paradyso, non eam cognouit 
Adam: "hoc manifestum est, quia ita scriptum est: Et postquam 
eiecti sunt Adam et Eua de paradyso, ^^tunc cognouit Adam mulie- 
rem suam, et concipiens peperit Cain [cf. Gen. 4. 1] super uitupera- 
biles aquas: "sicut Dauid dixit: Salua me, quoniam introierim 
aquae usque ad animam meam [Ps. 69. 1]. ^sStatim draco subuertit 
eos, et alienos eos fecit a uirtute (hoc est non placendo deo). ^^Et 
clamant uociferantes ad dominum, et uenit magnus elephans (hoc 
est lex), et non eum leuauit : quomodo nee sacerdos eum qui incidit 
in latrones [cf. Luc. 10. 30]. ^^Neque duodecim elephanti eleuaue- 
runt eum (hoc est prophetarum chorus), quomodo nee leuites, qui a 
latronibus erat uulneratus; ^^sed sanctus intellegibihs elephans (hoc 
est dominus Ihesus Christus) : ^^et cum omnium maior sit, omnium 
pusillus factus est: ^^Humiliauit enim se, factus obediens usque ad 
mortem [Phil. 2. 8], ^^ut hominem eleuaret, intellegibihs Samari- 
tanus, qui inposuit nos super iumentum suum (hoc est super suum 
corpus): ^^Ipse (enim) tulit injirmitates nostras, et inhecillitates 
nostras haiolauit [Esai. 53. 4]. ^einterpretatur autem hie Samari- 
tanus hebreice custos; de quo dicit Dauid in psalmo CXIIII: Cus- 
todiens paruulos dominus [Ps. 113. 7]. ^^Ubi enim dominus mens 



Carmody: Physiologus Latinus Versio Y 119 

presens est, neque draco neque mali aliquid appropiare poterit. 
^^Bene ergo Phisiologus de elephanto dixit. 
Codd.: Y Y^ — AE^(Jol. 39'') E*. Fragm.: F{2049). 



2 eo] Y quo AE^*; aliud ... trag-] Y om. AE'^^YK — 3 ad] E^YY^ in AE"^. — 4 
dicitur] Y uocatur AE'^^Y^. — 5 ibi] om. AE^*; tradidit masculo A Y^ dat mas- 
culo E^'^. — 6 et ... masc-] Y om. AE^; concipit AE^^. — 7 cum tem- ut par- 
uenerit A cum uero pariendi tem- uenerit E^'^. — 8 exiit Y; aque] om. AE^*; 
ipsa uenit aqua Y. — 8/11 ad ubera matris elephans autem AE'^*. — II partu- 
rientem AE'^*; propter ser-] om. AE'^'^; draco inimicus est elephanti AE'^'^. — 12 
eam Y. — 13 est] AE^* om. Y; quo- autem ceciderit quando in arb- A autem 
cum in E^*. — 14 ut dorm- si uelit A om. E'^^. — 15 uult A om. E^; modicum ut 
si uen- A modicum ut eleph- cum se incl- E^ et incl- se A ; ibi Y. — 16 animal] Y 
elifans A; cum eo cadens autem clamat AE'''^ add. flens A fortiter E^'^. — 17 
inde] om. AE"^; exiet Y . — 18 alii] om. AE'^*; elephantes AE^*; possunt eum 
leuare AE^. — 19 clamant omnes AE^; ueniet Y; promus-] AE^ om. Y. — 20 
autem] add. et Y; elephans] add. hoc A hanc E^; ut] om. AE'^; ossibus eius 
neque aliquid mali AE^; accedit Y. ■ — 21 ergo] Y om. AE^; eua Y. — 22 uir- 
tute] F carne AE^^Y^; suam] Y ipsorumAE^; neque intelligentiam {add. pec- 
cati EE*) habuerunt (habebant EE^*) AE^*. — 23 mandragoram AE^; dedit] 
add. et Y; malorum] Y om. AE"^; quod] add. et Y. — 24 quamdiu] AE^* quando 
Y; cognouit eam AE^*. — 25 quia ... et] AE^ de his que scriptum sunt quoniam 
Y; postquam ... tunc] Y om. AE^. — 27 me fac deus quoniam intrauerunt 
aquae A om. E"^. — 28 et statim A ; a uirtute] Y ab arce sua AE* ab are sua E^ 
{pro carnef); deo] Y dominum A -no E^. — 29 et ... dominum] Y tunc E"^* om. 
A. — 30 elephantes AE'^; leuita AE^; erat] Y fuerat A. — 31 sanctus] Y om. 
AE^. — 33 usque etc.] Y^ fol. 178''. — 34 hoc ... corpus] YY'' om. AE^*. — 35 
baiolauit] YY^ portauit AEK — 36 custus YY^; CXIIII] YY' om. AE^. — 37 
enim] YY^ autem E^ om. A add. est E^; opprop. Y. — 38 YY^ om. AE-; phisi- 
logus Y. 

XXI. DE DORCHON 

Est animal in monte qui dicitur dorchon grece, caprea latine. ^Amat 
satis excelsos montes, escam autem inuenit in humilia montium. 
^Et uidet de longe omnes qui ueniunt ad eam, %t cognoscit si cum 
dole ueniant, uel cum amicitia. 

^Habet autem caprea sapientiam dei: prophetas amat, hoc est 
montes excelsos, in quibus oculos suos leuauit propheta: Leuaui 
(inquid) oculos meos in montes, unde ueniet auxilium mihi [Ps. 
120. 1]. ^Et Salomon in canticis canticorum dixit de dorchon (hoc 
est de caprea) : Ecce fratruelis mens saliens super montes, transiliens 
supra colles [Cant. 2. 9]. ''Hoc autem dorchon salit super prophetas, 
transiliens super colles (hoc est apostolos). ^Acute autem uisionis 



120 University of California Publications in Classical Philology 

est dorchon, significans quoniam saluator uidet omnia que aguntur; 
Meus autem uocatur propter quod omnia uidet, et de longe, ad eum 
uenientes cum dolo; cognouit autem uenientem ludam osculo eum 
tradentem [cf. Luc. 22. 48]; ^"et iterum scriptum est: Cognouit 
dominus qui sunt eius [II Tim. 2. 19]; "et lohannes dixit: Ecce 
agnus dei, ecce qui aufert peccata mundi [loh. 1. 29]. 
Codd.:YY^Y\ 



3 longe] add. quoniam YV^. — 4 si] Y sic F^. — 5 habet] Y haec Y^; sap. dei] 
edit. om. Y ad Y\ — 6 fratuelis Y.—7 transili& YK — 8 uidit 7^. — 11 auferit 
72. 

XXII. DE ACHATE LAPIDE 

Est lapis qui uocatur achates : artifices autem margaritas inquiren- 
tes per hunc lapidem achaten inueniunt; ^lligant autem achaten 
lapidem f uniculo firmissimo urinatores, et summittunt in mare ; ^et 
ueniens achates lapis usque ad margaritam, stat et non mouetur; 
%tatim urinatores assequentes funiculum inueniunt margaritam. 
Codd.:YY^ — AY^ Em (3.57). 



agate semper Y^. — 1 incip. artif. enim A; qui] Y^ que Y; autem] YY^ enim 
qui A; inquirunt A quaerunt E'^H; hunc] YY^ om. AE'^H; acaten Y agathen 
E^H add. seu achaten H. — 2 autem] YY^ enim AH; ad funiculum AH; for- 
tissimum A. — 3 stat] Y fiat Y^; stat ... mouetur] YY'^ om. AE'^HYK 

XXIII. DE LAPIDE SOSTOROS ET MARGARITA 

QuoMODO autem nascitur margarita pronuntiabo : est lapis in mari 
qui uocatur sostoros; et uenit a mari matutino ante lucanum; ^et 
aperit conchas (id est os suum), et degluttit celestem rorem, et 
radium solis et lune et que sursum sunt siderum; ^et sic nascitur 
margarita de superioribus astris. 

^Sic autem hie nunc achates lapis, ut lohannes ipse enim ostendit 
nobis, intellegibilem margaritam Ihesum Christum dominum nos- 
trum, Micens: Ecce agnus dei, ecce qui aufert peccata mundi [loh. 
1. 29]. ®]y[are autem de saeculo dixit, urinatores autem qui sursum 
ferunt margaritam chorum sanctorum doctorum; ^peccatores autem 
deorsum ferunt earn, propter suam malitiam, quantum aduersus 
ipsos est; ^medio autem conche uel duarum alarum (hoc est inueni- 
tur intellegibiliter), et dominus meus saluator (hoc est in medio 
ueteris nouique testament!) a superioribus habens escam. ^Dixit 



Carmody: Physiologus Latinus Versio Y 121 

autem dominus meus quia: Regnum meum non est de hoc mundo 
[loh. 18. 36], sed a sempiterno patre et sanctarum eius uirtutum. 
^"Bene ergo Phisiologus dixit de achate et margarita. 
Codd.: YY^ — A Y'E^H (3.57). 



tit. post pronuntiabo] Y; et] add. de Y^.—l adnunt. Y^; in mari] YY^ uel piscis 
E^H; sostoros] F^sosturos Fpoterus AE'^. — 2 concas YY^ om. AE^H; id est] 
YY^ om. AEm; rorem] A ros YY^; lune et] YY"^ om. AE^H. — 3 superibus F — 
4 hie] FF^ om. AE'-H; ut] YY"^ om. AE^H; intellegebilem F. — 5 dei ecce] AH 
dei YY"^; aufert] F -it F^ tollit AH; peccatum F. — 6 de s. dixit] YY'^ praesens 
saeculum H add. designat H dixerim A; autem] YY"^ illorum H; qui sursum] F 
deorsum F^. — 7 suam] YY'^ eorum A; ipsius A. — 8 conce YY'^ cone A; hoc est] 
YY^ in his A; intell. ... dominus] YY"^; nouique] F -que noui F^ et noui A add. 
quae F; habens] A F^ habet F. — 9 autem] YY- enim A ; meus] YY"^ om. A ; de] 
A Fin F^jistosaeculo A;etsanct. ... marg.] YY'^om. A. 

XXIV. DE LAPIDE ADAMANTINO 

Est altera natura adamantini lapidis: neque ferrum timet, nee 
odorem fumi recipit; ^si autem inuentus fuerit in domo, neque 
demonium ibi appropiat, neque aliquid mali; ^in domibus regum 
inuenitur. ^Qui autem tenuerit eum, uincit omnem hominem et 
bestiam. 

^Lapis adamantinus est dominus meus: ^si habueris eum, nihil 
tibi mali occurrit. 

Codd.:YY^ — AYK 



1 est alia n. adamantine lapides F^ adamans lapis est qui nee A; nee ... 
domo] AY^ om. YY"^; accipit F^ — 2 aliquod malum A. — 3 domibus] add. 
enim A. — 4 uincet A. — 5 lapis] A om. YY"^^; adamans F^; meus] lesus Chris- 
tus A . — 6 si] add. quis eum A ; habuerit A F^; mali eis eueniet A . 

XXV. DE ALIA NATURA ONAGRI ET SIMII 

Est alia natura onagri (dixit sapiens) quoniam in regalibus domi- 
bus inuenitur : ^in quinta et uigesima Famenoth mensis cognoscunt 
ab onagro quoniam equitas dierum fit: ^si autem clamauerit duo- 
decies, cognoscit rex et palatium quoniam equitas diei fiet (ysemaria 
dixit grece). *Et simius, si septies mingat, ysemeria fit. 

^Onager est diabulus, quoniam nox (hoc est populus ex gentibus) 
aequalis f actus est diei : ^credentium prophetarum, clamauit onager 
(diabulus) . ^Et simius, hoc ipsam diabuli personam accepit : habuit 



122 University of California Publications in Classical Philology 

autem initium, finem autem non habet (hoc est caudam) ; ^n prin- 
cipio autem fuit unus ex archangelis, finis autem eius nee inuenitur. 
^Beneque simius, non habens caudam, sine specie enim est ; et turpe 
infimio, non habentem caudam ; ^°sicut et diabulus, non habet finem 
bonum. ^^Bene ergo Phisiologus dixit. 
Codd.: Y 7^. Trans, alt.: B Y\ 



simi Y. — 2 uicensima Y uincensima F^; famenoht Y. — 4 si] Y om. F^. — 9 
turbe in simio Y^. 

XXVI. DE LAPIDE INDICO 

Est lapis qui uocatur sindicus, banc habens naturam: ^si homo 
aliquis ydropicus fuerit, artifices medicorum inquirunt lapidem. 
^Si autem inuenerint eum, alhgant ydropico horis tribus, et exiet ab 
60 omnis feditas, conbibens in seipsum lapis; %t cum soluerint 
eundem lapidem ydropico, expendunt lapidem cum homine; ^et 
modicus lapis adducit corpus hominis in statera, hoc est in pondere. 
^Si autem dimittatur lapis in sole horis tribus, fedissimam aquam 
quam tulit de corpore ydropici effundit foras, et fit lapis mundus 
iterum sicut erat. 

^Lapis est dominus noster Ihesus Christus. ^Quoniam ydropici 
fuimus, habentes aquas diabuli in corde, descendens ligatus est lapis 
saluatoris, circa cor nostrum caritas eius; ^surgens autem a mortuis, 
omnem intellegibilem infirmitatem anime nostre sustulit: ^"Et in- 
firmitates nostras ipse baiulauit [Matt. 8. 17]. 

Codd.:YY^ — AY\ 



de lapide indico] edit, de lap- senditicos Y de 1- senditichos Y^. — 1 est ... 
uocatur] YY'^ lapis A; sindicus] A senditichos Y -dit- Y"^; habet AY^. — 2 
fuerit art-] YY"^ om. A Y^; medicorum] add. est ut A Y^; inquirant A. — 3 fedi- 
tas] add. atque YY"^. — 4 lapidem] add. ab Y^; expendent Y suspendunt AY^ 
add. appendentes YY^. — 5 et modicus lapis] A modus lapidis YY^. — 6 
dimittitur A; solis Y; fetidissimam A; toUit AY^; effund& F" add. earn F* 
autem F; foris ut sit A. — 8 quoniam] add. autem F; fumus F; eius] add. et 
F.— 9a]AexFF2. 

XXVII. DE HERODION ID EST FULICE 

Est uolatile qui uocatur herodius, de quo memoratus est Dauid, 
dicens: Herodion domus dux est eorum [Ps. 103. 17] (hoc est fulice, 
in psalmo CIII). ^Est autem hoc animal satis prudens pre omnibus 



Carmody: Physiologus Latinus Versio Y 123 

uolatilibus; ^non multos cubiles querens, sed ubi moratur, ibi et 
pascit, et reuertitur ibi et dormit; "neque morticina manducat, 
neque uolat in multis locis : cubile eiiis et esca in uno loco sunt. 

^0 et tu, homo poliuomene, una tibi sit nutrix et sempiterna 
ecclesia catholica, «ut spiritalis esca et celestis panis digestibilis fiat 
in te: ^noli querere multa loca aliene glorie (hoc est hereticorum) . 

Codd.:YY^ — AYK 



1 est ... herodius] A om. YY"^; herozion domus preest eis YY^; de quo ... 
dauid] YY^ om. A; fulicae domus AY^; eorum] A eis YY^; hoc ... CIII] YY^ 
om. A. — 2 autem] add. autem Y; hoc animal] YY^ uolatile AY^; omnibus] 
AY^ multis YY^. — 3 non multa copia A; moratur ubi et dormit AY^. — 4: 
locis] AY om. Y^; uno boca Y^. — 5 o] FF^ ita A; poliu.] Y^ (F) beneuole Y 
om.A.—6 panis] add. uitae A ; in te] tibi A. — 7 quere se Y; quaerere alienam 
gloriam per multa loca hoc est per multa hereticorum A; loco Y. 

XXVIII. DE PSYCOIVIORA 

Amos dixit: Non sum propheta neque fiHus prophete, sed pastor 
caprarius uellicans sycamina [Am. 7. 14]: pastor caprarum, et 
hyrcos pascit. ^Bene Amos personam accepit Christi saluatoris. 
3Quod autem dicitur sicamina uellicans : intellegibile est uerbum et 
spiritale est; ^Zacheus autem ascendit in sycomora [cf. Luc. 19. 4]. 
^Quod autem dixit psycamina uellicans, uel et cauans Sanctis: 
quoniam priusquam scauatus sycaminas. ^Sunt intra ipsum conopes, 
hoc est intra eum in tenebris habitantes, lumen non uidentes, ^in 
seipsos dicentes: In magna regione habitamus, in tenebris autem 
sunt seden+es [cf. IMatt. 4. 16]. ^Cum autem aperietur sycaminus 
ferro, et exierint inde, uident lucorem solis fulgentis, et lune et 
siderum, 'et dicent : Nos in tenebris fuimus sedentes in umbra [cf. 
Esai. 9. 2, IMatt. 4. 16] — ^°de his et Esaias dicens: Ut educat de 
uinculis uinctos, et de domo carceris habitantes in tenebris [Esai. 
42. 7], "uellicans primo die, tertio autem die parabitur fructus 
maturus, et esca fiet hominum. 

i^Compuncta est costa saluatoris mei, hoc est aperta est in lancea 
et ferro; et exiit sanguis et aqua. ^^Tertia autem die surgens a 
mortuis, uidimus intellegibilia luminaria et spiritalia: ^^sicut et 
scyniphas (hoc est culices), aperientes psycamina, uiderunt lumi- 
naria inmortalia. ^^Hedi personam accipiunt penitendum: etenim 
capilli eorum in sacco uertuntur, et: In cinerem penitentiam agunt 



124 University of California Publications in Classical Philology 

in eis [Matt. 11. 21, Luc. 10. 13]. ^^Populus qui sedebat in tenebris 
lumen uidit magnum, et qui erant in umbra mortis lux orta est eis 
[Matt. 4. 16, Esai. 9. 2]. ^^Adaperto autem psycamino, tertio autem 
die esca fiet : ^^percussus autem dominus noster, et tertia die resur- 
gens ex mortuis, esca et uita omnium nostrorum f actus est. 

Codd.:Y'Ymi-W- 

1 pastor cap- et hyrcos] Y om. Y^. — 4 est zacheis 7^ — 6 intra] F" in terra 
Y. — 8 uidef luxorem Y^. — 9 tenebre fugimus Y^. — 10 esaias] add. ceemit Y 
cecunt Y^; uinctas Y^. — 11 maturus / 29.4 animal] deest fol. Y. — 12 com- 
pucta Y^. 

XXIX. DE PANTHERE 

Propheta dicit : Factus sum sicut leo domui luda, et sicut panthera 
domui Effraim [Os. 5. 14]. ^Panther banc naturam habet: omnium 
animalium amicus est, inimicus autem draconi; omnimodo uarius 
est sicut tunica Joseph [cf. Gen. 37. 3], et speciosus. ^Etenim dixit 
Dauid in XLIIII psalmo: Adstitit regina a dextris tuis in uesti- 
mento deaurato operta uarietate [Ps. 44. 10]. ^Panther quietum 
animal est, et mitissimum nimis. Si autem manducauerit, et satia- 
tus fuerit, ilico dormit in fouea, et tertio die surgit a somno (sic et 
saluator noster). ^Panther autem, si surrexerit de somno tertio die, 
exclamat uoce magna, et de uoce eius omnis odor bonus aromatum; 
^et qui longe sunt et qui prope, audientes eius uocem, assequuntur 
bonum odorem uocis eius. 

^Dominus noster et saluator surgens a mortuis omnibus bonus 
odor factus est nobis [cf. II Cor. 2. 15], his qui longe et qui prope, 
pax [cf. Eph. 2. 17]; ^sicu-t apostolus Paulus dixit: Multifarie sa- 
pientie dei [cf. Eph. 3. 10], hoc est uirginitas, abstinentia, miseri- 
cordia, fides, caritas, unianimitas, pax, gaudium, longanimitas [cf. 
Gal. 5. 22]. ^Omni amore uaria est celestis sapientia dei Christi ; bene 
de panthere dictum est, quoniam inimicus est draconi in aqua. 
^"Nihil ergo sine intentione intellectus de uolatilibus et animalibus 
diuine scripturae dixerunt; "quoniam ait apostulus de satanan: 
Non enim eius uersutias ignoramus [II Cor. 2. 11], ambulat omni 
uia non bona [cf. Hier. 1. 23, 6. 16]. 

Codd.:YY^ — AY\ 



de leone et pantera Y"^. — 1/4 propheta ... animal] om. Y. — 1 dixit Y^. — 3 
dicit A ; dauid . . . psalmo] A om. F"; uestiu A F'; saturatus YY^. — 4 surrexit 



Carmody: Physiologus Latinus Versio Y 125 

a som- ^; tertia F". — ^uocem Y; et de uoce eius] AY^ om. YY^. — 5 panter 
Y. — 6 et qui longe] A qui longe YY'^^; uocis eius] YY^ om. AY^. — 8 abstinen- 
tia ... longan-] FF^ om. AY^; unianimitas] Y^ humanitas Y. — 9 sapientiae Y; 
pantera F^pantere F. — 11 ambulat] add. enim F^. 

XXX. DE CETO ID EST ASPIDOCELEON 

Phisiologus autem dixit de ceto quoddam, ^quod est in man, 
nomine aspidoceleon uocatur, magnum nimis, simile insule, ^et plus 
quam harena grauis, figuram habens diabuli. ^Ignorantes autem 
naute, alligant ad eum naues sicut ad insulam, et anchoras et palos 
nauis configunt in eo; ^et accendunt super eum ignem ad coquen- 
dum sibi aliquid; ^si autem excaluerit cetus, urinat, descendens in 
profundum, et demergit omnes naues. — ''Sic et tu, o homo, si sus- 
pendas te et aligas teipsum in spe diabuli, demergit te secum simul 
in gehennam ignis. 

^Aliut naturale habet cetus: si autem esurierit, multum adaperit 
OS suum, et omnis odor bonus per os eius procedit; ^odorantes autem 
pusilli pisciculi, secuntur eius odorem, et conponant se in ore magni 
ceti illius; ^"cum autem impletum fuerit os eius, concludit os suum, 
et gluttit pusillos omnes illos pisciculos, hoc est modicos in fide. — 
"Maiores autem et perfectos pisces non inuenimus adpropiare ad 
cetum: consummati enim sunt perfecti; etenim Paulus apostolus 
dixit: Non enim eius uersutias ignoramus [II Cor. 2. 11]. ^-lob per- 
fectissimus piscis est, Moises et reliqui alii prophete; loseph effugiit 
cetum magnum, principis cocorum mulierem, sicut in Genesis scrip- 
tum est [cf. Gen. 39] ; ^^sicut et Thecla Thamyridum, sicut Susanna 
duos senes Babylonicos iniquos; Hester et ludit effugerunt Arta- 
xersen et Olofernem; tres pueri Nabuchodonosor regem, magnum 
cetum; et Sarra filia Raguelis Nasmodeum (sicut in Tobia). Bene 
ergo Phisiologus dixit de aspidoceleon ceto magno. 

Codd.: Y Y^ YK 



3 graui F; figura F^. — 4 palos nauium F'. — 6 nauis F". — 7 aliges Y^. — 8 
natura F. — 9 pusillis F^; eius secu. F^. — 12 principes F^; effugit F^. — 13 
tecla F; thamyrido YY^; nabuchodosor F. 

XXXI. DE PERDICE 

HiEREMiAS dixit de perdice quoniam: Clamauit perdix, colhgens 
que non peperit [Hier. 17. 11]. ^Perdix aliena oua calefacit laborans 



126 University of California Publications in Classical Philology 

et nutriens; ^si autem creuerint pulli, et uolare coeperint, *unum- 
quemque genus euolans fugit ad parentes proprios, et solam earn 
dimittunt. 

^Sic et diabulus capit genus paruulorum ; ^cum autem uenerint in 
mensuram aetatis, ueniunt ad Christum et ad ecclesiam, ^et fiet ille 
insipiens; %odie, si quis est in malis moribus, crastinum fiet ut sit 
sobrius; ^et fugisti diabulum, hoc est perdicem, et uenies ad 
parentes tuos iustos et prophetas et apostolos. 
Codd.: Y Y^ Y\ 



1 colligit quod Y^. — 4 parentes] + suos Y^. — 5 sicut et F. — 6 uenerunt 
y. — 8 fiet] Y^ fiat F. — 9 apostulos Y^ + bene de perdice saluator YK 

XXXII. DE UULTURE 

Phisiologus dixit de uulture quoniam in excelsis et in altis inueni- 
etur locis; ^et dormit in excelsis petris et in pinnis templonim. ^Si 
autem pregnans fiat, uadit in Indiam, et accipit eutocium lapidem: 
lapis autem est sicut nix similis magnitudine; ^si autem uolueris 
eum agitare, uel alius capis interius eius mouetur, sicut tintinabu- 
lum, cum motus fuerit. *Si, autem, conprehenderit partus uulturem 
ut pariat, sedit super lapidem hunc eutocium (qui dicitur), et parit 
sine dolore. 

^Et tu, homo, si pregnans anima tua fuerit facta, et malis merori- 
bus aduersarii diabuli, accipe intellegibilem lapidem interiorem 
eutocium. ''Exterior autem eius est theothocos; ^JVIaria autem ha- 
buit interius spiritalem lapidem saluatorem nostrum: ^Lapidem 
enim quem reprobauerunt aedificantes, hie factus est in capud 
anguli [Ps. 117. 22]; et: Percisus est lapis de monte sine manibus 
natus [Dan. 2. 34], ^^et inuolutus propter ueteres errores nostros in 
spem bonam. "Sed et crucifixus est propter redemptionem pecca- 
torum nostrorum: tunc emundatur ab anima tua moechie, fornica- 
tiones, ebrietates, et cetera, et tunc celesti serraonem cjToforis; 
i^hoc est quod dixit Esaias: A timore tuo, domine, concepimus, et 
peperimus spiritum salutis tue quem fecisti super terram [Esai. 
26. 18]. ^^Etenim ueterem testamentum habuit in ministerio salua- 
torem ; "absconsus est enim saluator pseudo Hebreis, nobis ergo ap- 
paruit. ^^Bene ergo de uulture et lapide dictum est. 

^^In quo lapide Dauid occidit Golia; *^in duobus lapidibus susti- 



Carmody: Physiologus Latinus Versio Y 127 

nentes Aaron, et hur manus Moysi, et fugaret Amalech. ^^Uultur 
non unam habitationem neque cubilem habet; ^^et nos ueterem 
idolatriam (hoc est idolorum culturam et multorum deorum seque- 
bamur) cubile non habentes, ^^fidem in ecclesia unde et celestis 
paterna gratia apparuit; ^^et tunc saluati sumus per Christum 
Ihesum. 
Codd.: Y Y^ Y\ 



3 accepit Y^; eutocium] edit, eusthochium Y eustochium Y^. — 4 motus Y. 

— 5 eustocium YY^; que Y^. — 6 eutocium] edit, eusthocium Y eustochium Y^. 

— 7 est eius Y^; theosthocus 7. — 11 redemptione Y; aebri. Y; tunc] F^ cum 
Y; sermone Y. — 13 uetere Y. — 19 deorum] Y om. Y^; sequebantur Y^. — 21 

XXXIII. DE MIRIMICOLEON 

In lob Elefas Temaneorum rex dixit de mirmicoleon : Periit eo 
quod non habeat escam [Hiob 4. 11]. Tater autem eius habet 
uultum leonis, et comedit carnes; ^mater uero uultum formice, 
legumina manducat. ^Si autem peperit mirmicoleonta, perit eum, 
duas naturas habentem : uultum leonis, anteriora et posteriora for- 
micae; non potest manducare carnes propter naturam matris, 
neque legumina propter naturam patris; 4n perdito ergo uadit, 
propter quod non habeat escam. 

^Sic est ergo et omnis : Uir duplex corde, indispositus in omnibus 
uiis suis [lac. 1.8]. 'Non oportet ergo gradere duabus uiis, uecorde, 
dupHci, et manibus resolutis, et peccatori ingredienti duabus uiis; 
%icut in sapientia scriptum est: Sit apud uos est, est, non, non 
[Matt. 5. 37]. 

Codd.: Y YK 



de myr. Y^ de mirmice Y. — 1 escam] edit. om. YY^. — 2 carnis Y'. — 4 
perit] Y caparit Y^. — 7 uiis suae corde Y^. — 8 aput Y. 

XXXIV. DE IMUSTELA 

Lex dicit: Non manducabis mustelam, neque ei similem [Leu. 
11. 27, 29]. ^IMustela banc naturam habet: semen masculi in os 
accipit, et, pregnans facta, auribus pariet. ^Si autem per aurem 
dexteram contigerit ut generit, masculus erit; %i uero per sinis- 
tram, femina. ^IMala ergo ex auribus generantur. 

^Sunt ergo et nunc manducantes spiritalem et celestem panem in 



128 University of California Publications in Classical Philology 

ecclesia; ^cum autem dimissi fuerint, proiciunt uerbum ex auditibus 
suis, et fiunt ut dicitur in psalmo LVII: ^Sicut aspidis surde et 
obturantis aures suas, que non exaudiet uocem incantantis, et 
uenefici que incantantur a sapiente [Ps. 57. 6]. 
Codd.:YY^YK 



1 similis ¥. — 2 accepit F. — 3 generet 7.-7 LVIII 7.-8 surdo et 
obdurantes Y; exaudiunt 7'; incantantium 7^; ueneficiis incantantis sapien- 
ter Y\ 

XXXV. DE MONOCERATON 

MoYSES de monoceraton in Deuteronomio dixit, benedicens loseph: 
Primitiuus tauri species eius, cornua unicornui comua eius [Deut. 
33. 17]. ^Monoceras, hoc est unicornis, banc naturam habet: pusil- 
lum animal est, hedo similis, acerrimum nimis, unum cornum habet 
in medio capite. ^Non potest ei uenator appropiare, propter quod 
ualde fortissimum est. '^Quomodo ergo eum uenantur? uirginem 
castam proiciunt ante eum; ^exilit in sinum uirginis, et ilia calefacit 
eum, et nutrit illud animal; ^et tollit in palatium regum. 

^Unum cornum autem habet, propter quod dixit saluator: Ego 
et pater unum sumus [loh. 10. 30]. ^Suscitauit enim nobis cornu 
salutis, in domo Dauid pueri sui [Luc. 1. 69]; ^ueniens de caelo, 
uenit in utero uirginis Mariae : ^"Dilectus sicut filius unicorniorum 
[Ps. 21. 23], sicut Dauid in psalmo. 

Codd.: 7 7^. Trans, alt.: B YK 



2 animalem 7^^; cornu 7.'— 3 adpropinquare 7^— 7 autem habet / S6.12 
demus] deestfol. 7. — 10 unicornuorum 7^. 

XXXVI. DE CASTORE 

Est animal quod dicitur castor, innocentissimum ualde et quietum. 
^Uirilia autem eius in medicinam proficiunt ; ^inuenitur in palatium 
regis. ^Uenator autem currens, ut eum consequatur in montibus, 
^castor autem uidens uenatorem persequentem, se ore suo incidit 
uirilia sua, et dat uenatori proiciens. ®Si autem alius uenator eum 
persecutus postea fuerit, proicit se supinum, et ostendit se uenatori; 
^et uenator uidens se non habere uirilia, discedit ab eo. 

^O et tu, qui uiriliter agis, poUteuta dei, si dederis uenatori que ei 



Carmody: Physiologus Laiinus Versio Y 129 

sunt, amplius non accedit ad te; ^hoc est si habueris concupiscen- 
tiam malam, cupiditatem, moechiam, furtum, excide haec a te, et 
da diabulo. ^°Dixit ergo et apostolus: Reddite omnibus debita, cui 
tributum tributum, cui honorem honorem, et cetera [Rom. 13. 7]. 
"Prius turpitudines peccatorum, que in nobis sunt, proiciamus dia- 
bulo, hoc est opera eius; ^^et sic demus deo queque dei sunt, uota 
orationis, fructum bonorum operum nostrorum. 
Codd.: YK Fragm.: Y. Trans, alt.: B Y\ 



5 persequente Y^. — 6 subinum 7". — 8 accendit Y^. — 10 cui tributu tr. cui 
honore hon. Y^. — 12 deo ... nostrorum] YY^; orationes Y. 

XXXVII. DE YENA HOC EST BELUA 

Lex dixit: Non manducabis beluam neque similem ei [cf. Deut. 
14. 8, Leu. 11. 27]. ^Est arenotelicon, hoc est masculo-femina; ali- 
quando autem masculus fiet, aliquando uero femina: inmundum 
tamen est animal propter quod duas naturas habet; ^et ideo et 
Hieremias dixit: Numquid spelunca beluae hereditas mea mihi 
[Hier. 12. 9]. 

^Sic et omnis uir duplex animo belue comparatur: ad signum 
colligentis ergo aliquam masculorum habent exinde, hoc est ani- 
mum ; ^cum autem dimissa f uerit collecta, muliebrem naturam acci- 
piunt. ®Bene dixit Phisiologus. 

Codd.: Y Y^. 



ei] Y eius Y^. — 2 arenotelicon] edit, corsnotelicon Y^ arsnotelicon Y; 
mascel fi& F^. — 3 spelunca] add. ac Y^. 



XXXVIII. DE NILUO 

Est animal quod dicitur niluus, hoc est in flumine, figuram habens 
canis. ^Inimicus autem est corcodrilli; ^si autem uiderit corcodrillum 
dormientem, et apertum os eius, uadit niluus et unguit se totum 
luto ; *et cum persicauerit lutum, insilit in ore corcodrilli, ^et omnia 
intestina eius et uiscera eicit. 

^Sic est inintellegibilis infernus, rapiens omnem animam et mor- 
tificans : ^celestis autem noster saluator, accipiens terrenum corpus, 
descendit in infernum, Monec raperet educens eos qui antea mortui 



130 University of California Publications in Classical Philology 

erant, secundum sanctas promissiones, ^et ut solueretur fortitude 
et aculeus mortis. 
Codd.: Y Y^ YK 



2 est autem Y^. — 3 inunguit Y^ unguet F^; totum] add. os eius Y^; luta Y^. 

— 5 intestina] YY^ testimonia F^. — 6 est] add. autem Y^. — 7 discendit YY^. 

— 8 antea] Y ante Y^ a meta Y^; promissionis Y^ om. Y^. 

XXXIX. DE ECHINEMON 

Est animal quod dicitur echinemon, inimicum autem draconi, ^Si 
autem inuenerit draconem, uadit contra eum et unguit se luto, ^et 
cooperiet nares suas de cauda sua, celans et inflans se, et sic astat 
contra draconem, donee eum interimat. 

^Sic et saluator noster, accipiens ex terreni corporis substantiam, 
hoc est corpus quod accepit ex Maria, ^stetit donee occideret intel- 
legibilem draconem Pharaonem, ^qui sedet super flumina Egj^pti, 
hoc est diabulum. 

Codd.: Y Y^ Y\ 



1 echnemon Y om. Y^; inimicum] add. est Y^. — 2 inunguit Y^; lutum Y^. — 3 
interimet F*. — 4 externi F^ exterren F^; substantia F. — 5 occiderit F ne- 
cauerit F^ — 6 flumen F^. 

XL. DE CORNICOLA 

HiEREMiAS propheta testatur quoniam: Sedisti sicut cornicola 
deserta [c/. Hier. 3. 2]. ^Phisiologus pronuntiauit quoniam unius 
uiri est ; et si uir eius naoriatur, non fit alio uiro, neque corpus uiri 
alie mulieri commiscetur. 

^Sic ergo nunc deserta cornix sinagoga ludeorum terrestris Hieru- 
salem, que occidit uirum suum celestem, propter quod deserta est, 
non habens uirum Christum. ^Dixit enim apostolus de his qui ex 
gentibus crediderunt: Statui enim uos uni uiro uirginem castam 
exhibere Christo [II Cor. 11. 2]. ^Habemus ergo nos uirum diuinum 
sermonem in animis nostris, ut non appropinquet alienus, et 
inueniamur spelunca latronum. 

Codd.: Y Y^. 



1 hieremie prophete F^ — 2 moriat F; commissetur F. — 3 conix F. — 4 
statuit F*. 



Carmody: Physiologus Latinus Versio Y 131 

XLI. DE TURTURE 

Dixit Salomon : Uox turturis audita est in terra nostra [Cant. 2. 12]. 
^Turtur sedet in desertis, hoc est in solitudinibus secedens; ^non 
amat in medio multorum esse. 

^Et saluator noster demorabatur in monte Oliueti (dixit enim) ; 
'^assumens Ihesus Petmm, lacobum, et lohannem, ascendit in 
montem [cf. Marc. 3. 13]. ^Turtur secedere in silentio amat: sic et 
generosi christofori eligunt in secreto habitare; ^christofori dicuntur 
Christum induti, qui imitantur turturem dominum nostrum Chris- 
tum. 

Codd.: Y Y^ YK 



4 oliueti] add. Matt. S.17 d. — 4/7 dixit enim ... dicuntur Cbristum] YY^ et 
omnis in solitudine sicut turtur etenim qui YK — 7 turture Y. 

XLII. DE HYRUNDINE 

EzECHiAS rex dixit in Esaia propheta: Sicut hyrundo, ita clamabo, 
et sicut cohimba, sic meditabor [Esai. 38. 14, c/. 59. 11]. ^Et in 
Hieremia dicitur: Turtur et hyrundo et cyconia custodierunt 
tempus introitus sui [Hier. 8. 7]. ^De hyrundme locutus est Phisio- 
logus quoniam semel generat, et iam non. 

''Et saluator meus natus est semel in utero; ^baiulatus est semel, 
crucifixus est semel, et surrexit a mortuis semel: Unus deus, una 
fides, unum baptisma [Eph. 4. 5]. 

Codd.: Y Y^ Y\ 



2 introtus Y. — 5 et surr.] Y resurrexit F*. 

XLIII. DE CERUO 

In psalmo XLI dicit : Sicut ceruus desiderat ad fontes aquarum, ita 
desiderat anima mea ad te, deus [Ps. 41. 2]. ^Ceruus inimicus est 
draconi; draco autem fugit a ceruo in fissuras terre; et uadens 
ceruus, et ebibens, implet nasa sua fontem aque, et euomit in fis- 
suram terre, ^et educit draconem, et conculcauit eum, et occidit 
eum. 

*Sic et dominus noster interfecit draconem magnum diabulum ex 
celestibus aquis, ^quibus habebat sapientie inenarrabihs ; non enim 
potest draco baiulare aquam, neque diabulus sermones celestes. ^Si 



132 University of California Publications in Classical Philology 

enim et tu habueris intellegibiles dracones absconsos in corde tuo, 
inuoca Christum ab euangeliis per orationes, et ipse occidet eum: 
^Tu es enim templum dei, et spiritus dei hahitauit in te [I Cor. 3. 16]. 
^Capilli autem cerui, ubi appanierint in domo, uel de ossibus in- 
cenderis, numquam draconem inuenies: ^uestigium dei et timor 
Christi si inueniantur in corde tuo, nullus spiritus inmundus in- 
troibit tibi. 
Codd.: Y Y^ F3. 



1 dixit Y. — 2 fisuras Y; fisuram F. — 4 noster] Y^ meus 7.-9 inueniuntur 
Y. 

XLIV. DE RANA 

Est rana que dicitur cerseus, hoc est desiccano. 'Cersea rana per 
estatem estus non curat; si autem comprehenderit earn pluuia, 
morietur ; ^ahe autem rane aquatice que inter aquas sunt, si uiderint 
radios soHs et calefacte fuerint, baptizant se in fonte. 

^Hoc est generosi abstinentes, desiccano sunt. ^Non pertinet eis 
de quibus patienter in abstinentiis laborantes; si autem pluuia eos 
comprehendat, moriuntur (hoc est seculi cupiditates) ; ^aquatice 
autem rane in eis, que non sufferunt abstinentiam. ^Et si ad diem 
abstinuerint, non baiulantes radium solis intellegibilis, inmittunt se 
iterum in ea ipsa priora desideria. 

Codd.: Y YK 



1 cerseus] edit, censeus YY^. — 2 pluuiam F". — 3 radius Y radium 7*. — 5 
comprehendit Y^; cupitates Y. 

XLV. DE SAURA ID EST SALAMANDRA 

Est mirabile hoc in Danihelo de tribus fratribus: in camino ignis 
missi sunt, ut et hymnum dicerent deo [cf. Dan. 3]. ^Uere non mirum 
iusti enim fuerunt; ^et si mortuos suos suscitauenmt, neque mag- 
num iustos montes transferri in mare [cf. I. Cor. 13. 2]. ■'Phisiologus 
dixit de saura que dicitur salamandra, quoniam si introiuit in ca- 
mino ignis uel fornace balnearum, omnis ignis extinguitur; talera 
naturam habet. ^Quanto melius, qui secundum iustitiam extingue- 
runt uirtutem ignis, obturauerunt ora leonum [Heb. 11. 33]. 
Codd.: Y Y\ 



de alia id est salamandra FF*. — 3 suscitatae simt F*. — 5 obdurauerunt F*. 



Carmody: Physiologus Latinus Versio Y 133 

XLVI. DE LAPIDE MAGNITEN 

Est lapis qui dicitur magnis, suspendit ferrum, si adheserit ferro, 
hoc est si apponatur ferro. ^Si ergo creatura hec facit, quanto melius 
creator omnium et factor suspendit celum a terra, et extendit 
celum sicut pellem. 
Codd.: Y Y^ Y\ 



1 magnus Y; adherens ei si ergo YK — 2 hec creatura Y; quantum Y^; sus- 
pendit ... ext. celum] om. Y^. 

XLVII. DE LAPIDE ADAMANTINO 

Est lapis adamantinus, in orientis partibus inuenietur, ^non in 
diebus solis fulgens, sed in noctibus inuenitur; Adamantinus dicitur 
quia omnia domat, ipsum enim nemo potest domare. 

^Dominus mens et saluator omnes iudicat, et non iudicatur ab 
aliquo. ^Et in Amos propheta dicit: Et uidi lapidem adamantinum 
[Am. 7. 7]. «Et tu, si inuentus fueris in orientis partibus, maxime si 
uocatus fueris iustus, inreprehensibilis et pius, sicut lob [cf. Hiob 
1. 1], innocens, et cetera, a solis ortu. 

Codd.: Y Y^ Y\ 



1 inuenitur Y'^. — 2 non in die tamen Y^; inuenitur adamantinus] Y^ om. 
YYK — 3 enim] YY^ autem Y^; nemo] YY^ non F*. — 4 sic et salu. noster 
omnis iud. Y^. — 5 dicit] add. de illi Y^. 

XLVIII. DE COLUMBIS 

loHANNES euangelista dicit: Uidi celum apertum, et spiritum dei 
sicut columbam uenientem super eum [IMatt. 3. 16, IVIarc. 1. 10, 
loh. 1. 32]. 2De multis columbis locutus est Phisiologus; sunt peg- 
matistes columbe; ^multa enim sunt genera columbarum, et multis 
coloribus et uariis; est niger, cinericius, auri speciem habens, melli- 
cus (hoc est fuscus), icotus, albus, rufus; "si autem omnes columbas 
mittat ad pegmatistes neminem inducet, solus rufus inducet omnes 
et placat. 

^Sic sanguis saluatoris mei induxit omnes in uitam eternam : ^non 
Moyses, quia legem dedit, non Esaias, non quisque prophetarum, 
sed ipse ueniens dei filius dominus Ihesus Christus saluauit nos 
suo sanguine sancto. TIoc ipsum extra ab fornicaria: propter coc- 



134 University of California Publications in Classical Philology 

cinum signum saluata est anima eius: Maria sortita accepit cocci- 
num purpuram [cf. Apoc. 17. 4]. ^Et in canticis canticorum dixit: 
Uestis coccinea labia tua, et loguella tua speciosa [Cant. 4. 3]. ^Et in 
Matheo: Clamidem coccineam induerunt eum [loh. 19. 2]. 

Codd.: Y Y\ 



2 permatice YY"^. — 3 mellinus F^; icotus] Y totus F^; rufus] add. in F. — 4 
pegmanites F pecmanites F^; plagat F. — 5 mei] F nostri F^. — 8 restis YY"^. 

XLIX. DE SAURA ELIACE HOC EST ANGUILLA SOLIS 

Est qui uocatur saura eliace, hoc est anguilla solis. ^Cum senuerit, 
impeditur duobus oculis suis, et excecatur, non uidens solis lumen. 
^Quid faciet? ex bona sua natura inquirit parietem respicientem ad 
orientem, ^et intrat in fissuram parietis, uidens ad orientem, et 
oriente sole aperientur ei oculi, et noua efficitur. 

^Sic et tu, o homo, si ergo ueteris hominis indumentum habes 
[Col. 3. 9, Eph. 4. 22], uide ne quando oculi cordis tui impediti 
fuerint, ^requiras intellegibilem orientem solem dominum lesum 
Christum, cuius nomen uocatur oriens [cf. Zach. 3. 8, 6. 12, Luc. 1. 
78] in propheta Hieremia. ''Et ipse est sol iustitie, sicut apostolus 
dicit [cf. Mai. 4. 2]; et aperiet tibi intellegibiles oculos cordis tui, et 
nouum per ueteris fiet tibi uestimentum. 
Codd.: Y Y^ Y\ 



1 inc.: qui uocatur nouella cum sen. F'. — 3 facit F'; ex] F^^; naturam F. 
— 5 tu homo si uet. F'; habens F^ habis F^; cordis] am. Y^. — 6 requiras] edit. 
nequirat (/) F' quanta YY^; intell.] add. faciunt YY^; oriente sole YY"^; solum 
F^; dominum] add. nostrum F^; hieremie F. — 6-7 oriens ipse adaperiet oculos 
F*. — 7 ueteris] F" ueteri F; fiet] YY^ dabit F'. — expl.: expet liber crisos- 
tomi de naturis animalium F explicit de natura bestiarum sancti iohaimis 
constantinobolitane untistinie liber conscriptus deo gratias amen {add. de 
pigmentis nardi spicatae) F*. 



SENECA'S EPISTULAE MORALES 

THE TEXT EMENDED AND EXPLAINED 

(LXVI-XCII) 

BY 

WILLIAM HARDY ALEXANDER 



LXVI, 16 (214: 20) :^ quia nihil honestum est, quod ab invito, quod 
coactum fit. 
P : aco actum 

QVb : a coacto (in Q as one word) 
P : quodaco ado 

Hense (II) followed Haase in reading with p the word coactum, 
but subsequently, after the discovery of Q, wrote thus: "Es (sc. 
coacto) wird durch die UeberUeferung (auch durch Q) in gleicher 
Weise wie durch die Konzinnitat gesichert. Auch aco actum in p 
lasst das Richtige noch durchblicken."^ But concinnity is a dan- 
gerous argument in Seneca. 

A few words further on, however (215: 5), we find: omne hones- 
tum iniussum incoactumque est, which is merely the positive repe- 
tition of the negative statement in the passage under discussion. 
The possibility is thus suggested that incoactum is an echo of a pre- 
ceding coactum. It is no argument against this theory to point out 
that iniussum is not an echo of ah invito, as there is no repetition of 
the same word, whereas in the employment of incoactum there is. 
The reading quod coactum fit produces the favorite cretic-spondee 
clausula quite as well as a coacto fit does. 

In a number of places, admittedly, p retains the archetype read- 
ing as against QVPb, which are in other lines of descent, and it may 
have done so here; the division of the words in P {quodaco acto) 
seems to have some bearing on that point. The archetype read aco 
actum, and p, obviously the work of a rather ignorant scribe, has 
preserved this reading including the error involved in the a, whereas 

^ The numbers in parentheses give the page and line reference in Hense's 
second edition (Leipzig, Teubner, 1914), hereafter referred to as Hense (II). 
Hense (I) appeared in 1899 under the same auspices. 

* Rheinisches Museum, LXXIV (1925), p. 120. 

[135] 



136 University of California Puilications in Classical Philology 

in the other two hnes of descent^ the text has been improved by 
writing a coacto. It is also quite possible that this correction may 
have been indicated in the archetype and ignored by p. 

LXVI, 21 (216 : 14) : eundem locum habebit apud ilium honesta 
res, sed tristis atque aspera, quem vir bonus pauper aut exul ac 
palhdus. 

I share with Axelson^ a strong feeling of suspicion in regard to 
the extremely peculiar ending of this sentence, and also his surprise 
that "haben sich die Herausgeber nicht storen lassen," inasmuch 
as the concluding words involve an association of exul with, pallidus 
which cannot be paralleled, and is not self-explanatory or self- 
justifying in the absence of parallels. Axelson proposes aut exul 
<aut exilis>. No doubt aut and an adjective have been omitted, 
but it seems to me that the loss can be adequately explained by 
postulating a scribe at some point in the ms. tradition who, after 
writing exul, picked up the ac before pallidus instead of the aut 
followmg exul, and that the way is thereby opened up for the sup- 
plying of any reasonable adjective to fill the gap, not necessarily 
one to be explained by a haplography. The collocations provided by 
Axelson himself, XXVII, 8 : hominem aegrum, palhdum, gracilem, 
and N.Q. IV b, 13, 10: pallentes et aegros, appear to contain as 
good a solution as one is likely to find, namely, <aut aeger> ac 
pallidus. The fact that aeger seems in Seneca to be so often used 
substantively^ in the sense "invalid," would make it a good third 
after pauper and exul; at the same time its status is so fluid that it 
readily slips back into adjective value when combined with pallidus. 
The clausula rhythm (double cretic) is sound. 

LXVII, 5 fin. (227: 10) : deinde etiam [si] tormentorum fortis pa- 
tientia optabilis est. etiamnunc interrogo : nempe fortitudo opta- 
bilis est? atqui pericula contemnit et provocat. ... si fortitudo 
optabilis est, et tormenta patienter ferre optabUe est. hoc enim 
fortitudinis pars est. 

Deinde has been objected to by Fr. Schultess (Annaeana Studia, 

^ The reference is to the stemma (p. 34) in O. Foerster's Handschriftliche 
Untersuchungen zu Senekas Epistulae Morales und Naturales Quaesliones 
{Wurzburger Sludien z. Aliertu7nswiss., Heft 10, 1936). 

* Bertil Axelson, Neue Senecastudien (Lund, 1939), p. 180. 

^ As, for instance, LXXV, 6: non quaerit aeger medicum eloquentem; of. 
also LXXVIII, 20: temperans aeger, and LXXVIII, 22: animi voluptates . . . 
nemo medicus aegro negat. 



Alexander: Seneca's Epistulae Morales (LXVI-XCII) 137 

Hamburg, 1888), but is entirely sound; it introduces the ultimate 
development of the preceding argument, which has been set out 
under the introductory word primum (1.7) and deinde (1.8). The 
deinde of 1. 10 continues logically this series. For a primum . . , 
deinde . . . deinde in Seneca, compare Dial. 9, 6, 1; Epist. LV, 10; 
XCIX, 6. 

I agree with Madvig, Hense, Beltrami in the excision of the si 
following etiam; this clause is a conclusion, not a condition. This 
interpretation is obvious at once from the run of the argument, and 
also from the etiamnunc ("further," "moreover"), which introduces 
a new turn in the discussion and therefore is not strictly dependent 
on anything before it. Haase retains etiamsi only by postulating a 
lacuna after it. 

For the mss. consensus neme, I prefer Niemeyer's nonne^ to 
Haase 's nempe for paleographical reasons. Neme arises from an 
original nonne through some malformation of the o which allows it 
to be thought of as an e. Either nempe or nonne will serve the sense 
of the passage. 

It would also be possible, and is attractive as giving dramatic 
value to the argument, to consider deinde etiam . . . optahilis est as 
representing the question of an interlocutor, and hence to be 
printed with quotation marks and to be followed by a sign of inter- 
rogation. Then Seneca would answer a question with a question, 
and, assuming an afl&rmative reply to the question nempe (or 
nonne) fortitudo optahilis est which he poses, would proceed to apply 
the results of such a concession in dealing with the interlocutor's 
question deinde etiam . . . optahilis est. If this is a correct conception 
of the passage, etiamnunc interrogo rea,lly means "I have a question 
also for you." 

LXVIII, 1 (230: 11) : hoc te facturum Stoicorum etiam si non prae- 
cepto, at exemplo licet scias. sed ex praecepto quoque facies: et 
tibi et cui voles adprobabis. 
p : et tihi con et cum voles; 
QVP: et tihi et cum voles, and thus b with co for cum; 

Cui is Buecheler's emendation for cum. Hense (II) in the app. crit. 
ad loc. suggested hoc for the curious con of p, which I am myself 
inclined to regard only as another example of p's pecuUar habit of 
6 Philologus, LVIII (1899), p. 444. 



138 University of California Publications in Classical Philology 

developing extra letters or even syllables without ascertainable 
cause. Hense's suggestion points, however, to the fact that he inter- 
preted the sentence as meaning "you will demonstrate this" (viz., 
the fact that you are acting ex praecepto as well as exemplo Stoi- 
corum) "both to yourself and to whomsoever you will." But in fact 
it is Seneca who proceeds to demonstrate that Lucilius will so be 
acting, by explaining the Stoic view in regard to participation in 
public life or retirement from it. I suspect therefore that adprobare 
here is used in the sense "approve," "make good," to complete 
which the reflexive pronoun is required. Read in consequence: te 
tibi et cui voles adprobabis, and compare Dial. 9, 7, 5 ; bonos quibus 
se adprobaret; also Dial. 11, 18, 6, and Epist. CII, 29. 

LXVIII, 11 (233: 3) : ille me gratia forensi longe antecedet, ille 
stipendiis militaribus et quaesita per hoc dignitate, ille clientium 
turba: est tanti ab omnibus vinci, dum a me fortuna vincatur. 
t cui in turba par esse non possum, plus habet gratiae. 

There is no occasion to attach oneself to the reading of p, 
namely, cui in turba, in the face of the reading of QVPb, namely, 
cuius turhae. Hense himself^ now accepts cuius turhae, but would 
insert <clientiuin> between the two words. This seems to me quite 
unnecessary. The thought of the phrase ille clientium turba is still 
active from the first half of the preceding sentence. The transla- 
tion obviously is: "he for whose mob (of retainers) I can be no 
match, has more influence than I have" ; gratia returns to the begin- 
ning of the first half of the preceding sentence as turba comes from 
its close. 

But it would seem an improvement to read habeat; it would bring 
the tone of the cuius turbae sentence into accord with the tone of the 
est tanti statement preceding. I now observe that Vivona^ has 
reached the same conclusion: "It's worth while being beaten by 
everybody provided Fortune is beaten by me. Let the man whose 
mob I cannot match have the greater influence." The implied con- 
clusion then is: "What do I care?" 

LXIX, 5 (234 : 22) : vix effici toto saeculo potest, ut vitia tam longa 
licentia t umida subigantur et iugum accipiant, nedum, si tam 
' Rheinisches Museum, LXXIV (1925), p. 121. 
8 Francesco Vivona, Note Critiche alle Epistole di Seneca (Rome, Casa 

Editrice "Ausonia/' 1932), pp. 31-32. 



Alexander: Seneca's Epistulae Morales CLXYI-XCII) 139 

breve tempiis intervallis caedimus. 
p : intervallum discedimus. 
QVPb : intervallum discidimus. 
Erasmus (I) : per intervallum discindimus. 
Erasmus (II) : per iniervalla discindimus. 
Beltrami (II) : intervallis discindimus. 
Madvig : intervallis caedimus. 

The figure in the first half of the sentence is that of active war- 
fare against the vices until they are subdued and compelled to pass 
under the yoke; even with the campaign actively waged, the reduc- 
tion of these enemies might take a whole century. Nedum si, "to say 
nothing of what will be the case if," obviously should introduce us 
to the picture of something that is not vigorous and aggressive war- 
fare, in short, dalliance and delay. That is why one may guess that 
the original text may have been si tarn hreve tempus intra vallum 
desidimus, "if we have idled away within the rampart the time that 
is so short," a good picture of an army that has forgotten what 
active service means, and hence presenting just the opposite con- 
ditions to those which Seneca has said are required for success. 

On the general force of the passage, compare XLVIII, 12: etiam 
si multum superesset aetatis, parce dispensandum erat, ut sufficeret 
necessariis: nunc quae dementia est supervacua discere in tanta 
temporis egestate? and XLIX, 5: eo magis itaque indignor aliquos 
ex hoc tempore, quod sufficere ne ad necessaria quidem potest, 
etiamsi custoditum diligentissime fuerit, in supervacua maiorem 
partem erogare. 

LXX, 5 (236: 15) : si multa occurrunt molesta et tranquilHtatem 
turbantia, emittit se. nee hoc tantum in necessitate ultima facit, 
sed cum primum illi coepit suspecta esse fortuna, diligenter cir- 
cumspicit, numquid t iHo desinendum sit. 

Illo has been variously emended; so far as the meaning is con- 
cerned, the illic, "at that point," of Beltrami (II) gives what is 
required. However, it is rather easier paleographically to read 
<in> illo, and this can be justified on other grounds also. The 
construction desinere in with the ablative is sound Latin for Seneca; 
compare Dial. 6, 2, 1 : scio a praeceptis incipere omnis, qui monere 
aliquem volunt, in exemplis desinere. Again, the illo of <in> illo 
refers synoptically to the point of time or the nature of the situation 



140 University of California Publications in Classical Philology 

set out in the words cum primum illi coepit suspecta esse fortuna 
preceding; translate <in> illo either "at that time" or "in those 
circumstances." Compare LXXXVIII, 18: in illo feras me necesse 
est non per praescriptum euntem, where I regard in illo as antici- 
pating the words non per praescriptum euntem, equivalent to quod 
non per praescriptum eo, which is deflected to agree with me and is 
thus participially expressed. In this passage in illo anticipates 
synoptically, instead of, as in the passage under examination, re- 
suming synoptically. 

LXX, 5 (236 : 20) : non tamquam de magno detrimento timet : nemo 
multum ex stilicidio potest perdere. 

These words occur in a description of the sage's attitude toward 
death. It is obvious from an examination of the translators' work 
here that the phrase ex stilicidio has eluded them for the most part. 
Thus Barker^ says: "no one loses much of something that's already 
escaping drop by drop." But he bolsters his translation with 
"already," having, no doubt, made up his mind that the reference is 
to life disappearing a day at a time, and it is clear further that, if 
Seneca's words mean what Barker makes them mean, Seneca is 
illogical. If there are two hundred gallons of some liquid stored in a 
tank which has a small leak at one point where the Hquid gets out 
drop by drop, you cannot reasonably say that no one could suffer 
much of a loss in respect of the two hundred gallons if a shell sud- 
denly blew the whole tank up because he was "already" losing its 
contents at the rate of a drop every five seconds. Gummere^" ren- 
ders thus: "no man can lose very much when but a driblet re- 
mains," but I submit that his introduction of "remains" is as 
unjustifiable as Barker's insertion of "already." Lagrange" writes: 
"Eh! peut-elle [sc. la perte]etre bien grande,quand unvase ne coule 
que goutte a goutte," and this, I believe, affords the approach to a 
plausible solution of the meaning of the passage. 

No great loss can be suffered by the advent of death, certainly 
nothing to be feared as a loss, because life itself is something held in 
a very small container out of which the liquid is drawn only a 

9 E. P. Barker, Seneca's Letters to Lucilius (Oxford, 1932), Vol. I, p. 239. 
1" Loeb Classical Library translation, Vol. II ad loc. 

" CEuvres de Sinbque le philosophe (Paris, J. J. Smits, An III de la R6- 
publique), Vol. 2, p. 324. 



Alexander : Seneca's Epistnlae Morales (LXVI-XCII) 141 

drop at a time. And the emphasis is on the smallness of the con- 
tainer; its smalhiess is demonstrated by the drop-by-drop process 
through which it yields up its contents. It does not pour, it gives 
up a drop at a time. It seems to me, therefore, that stilicidium here 
means something Hke a modern eyedropper, a small vial of per- 
fume, perhaps; the whole receptacle is so small that if it were 
broken the total loss could not be very great. Thus life at best is 
short, our time is contained in a tiny vessel which is constantly 
(i.e., drop by drop — day by day) yielding up its contents; but even 
if the whole vessel be suddenly smashed, it is so small to begin with 
that the loss cannot be great. 

My guess is, then, that stilicidium is here technically used in the 
sense I have assigned to it, a "dropper," and that the correct 
translation is: "nobody can lose much out of a dropper." On the 
justification for such an assumption consult the interesting material 
provided by Summers in his edition of certain of the Epistles,^'^ 
pages xlii-xlix, and especially xlvi-xlix. 

LXX, 21 (241 : 6) : undique destitutus invenit, quemadmodum et 
mortem sibi deferret et telum. 

Deheret of the mss. is to be retained against various conjectures, 
including Hense's deferret above. So Beltrami both (I) and (II), 
and Axelson, Neue Senecastudien, page 15. The following para- 
phrase will, I think, show what Seneca had in mind and also the 
odd way in which Silver Latin can go about the business of ex- 
pressing involvedly a perfectly straightforward idea. "Deprived 
of any friends with whom he could argue out both the logicality of 
the suicide and the best way of achieving it, as a Roman in a like 
situation would regularly do, he found out for himself how, in the 
first place, he owed death to himself rather than suffer the ignomi- 
nies of the arena to make a Roman holiday. In the second place, he 
discovered that he owed himself the weapon, that is to say, dis- 
covered that the debt he owed himself was such that he could not be 
fastidious about the means by which it was to be paid, but was 
under obligation to himself to use the first available means, how- 
ever repulsive, he might happen on, because he had learned how few 
in number a t best such means would be." 

^^ W. C. Summers, Select Letters of Seneca (London, Macmillan, 1913). 
Note too the aTra^ Xeyofxevov stillarium in Epist. XCVII, 2 fin. 



142 University of California Publications in Classical Philology 

LXX, 28 (242: 21) : eadem ilia ratio monet, ut, si licet, moriaris 
quemadmodum placet: si minus, quemadmodum potes et quicquid 
obvenerit, ad vim ferendam tibi invadas. 

The mss. consensus is scilicet moriaris quemadmodum potes, and 
Fickert and Beltrami (II) retain this, but with grave doubts about 
the soundness of scilicet. The additional words in Hense's text above 
are substantially Schweighaeuser's, except that Hense improves 
the latter's Latinity by writing si minus for si non autem; he also 
reads si licet with Schweighaeuser, thus obtaining a parallel for the 
conjectural <si minus > . Fickert suggests that the supplementary 
words provided by the later mss. <sine dolore; si autem non potes 
(al. potest), fac> arise from scilicet's being read si licet, precisely as 
Schweighaeuser and Hense have done. The fac is added because the 
possibility of carrying over the governance of monet was not ap- 
preciated. 

My own view is that the sentence originally read eadem ilia ratio 
monet ut, quemadmodum potes et quicquid obvenerit, ad vim ferendam 
tibi invadas. To the ut . . . invadas clause scilicet moriaris is a fairly 
obvious gloss, later incorporated into the text. The sentence as I 
leave it is to be interpreted closely with the context ; it gives advice 
to be acted on if the conditions are such as those we have been dis- 
cussing. The same is true of the sentence next following. 

LXXIV, 9 (263 : 9) : vilem praedam magno aliquo incommodo 
luimus aut destituti fallimur. 

The destituti above is Buecheler's completion of the syllable de in 
the ms. tradition; thus we have aut de aut fallimur in V as it origi- 
nally stood, aut de aut fallimus in P, autdefallimur in Q, and aut 
inde fallimur in b. 

I suggest that the correct reading is decepti (for de aut of VP) 
fallimus (with P). The whole sentence thus appears to have a sound 
meaning; "we pay for our cheap prize w^ith some grave incon- 
venience, or else, deceived (in what we have won), we play the 
hypocrite (about it)," that is, put a good face on our misadventure 
in the hope of concealing its actual nature from others. Fallimus 
would then be used absolutely. Decepti fallimus sounds and looks 
like Senecan Latin ; the spondee-cretic clausula is frequent. Axelson 
might call it double cretic with an irrational long in the first foot. 



Alexander: Seneca's Epistulae Morales (LXVI-XCII) 143 

LXXIV, 33 (270: 2) : quemadmodum in corporibus f insignis lan- 
giioris signa praecurrunt, quaedam enim segnitia enervis est et 
sine labore ullo lassitude et oscitatio et horror membra per- 
currens, sic infirmus animus multo ante quam opprimatur malis 
quatitur. praesamit illi et ante tempus cadit. 
P : insignis langore (sed insignis transfossum) 
QV : insignis (g ex corr. V) langorem ( = uorem V) 
b : insigni languore 

The efforts of emendation have all been directed to replacing 
insignis with some other adjective on the theory that corporibus 
must have a modifier just as animus below has infirmus; the list will 
be found in Beltrami (II). My own belief is that insignis is sound 
(note the mss. evidence) and is the modifier of languoris, the read- 
ing of the recentiores. In PQV insignis appears to have been taken 
with corporibus, and thereafter either an ablative langore was de- 
veloped to supply a causal notion with insignis, or an accusative 
languor em to provide an object for praecurrunt. Insignis languoris 
I take to be a phrase the words of which have become inverted in 
transmission, and I therefore read the passage thus: quemadmodum 
in corporibus languoris insignis signa praecurrunt, "just as in bodies 
symptoms of an extraordinary lassitude run ahead (of the disease)." 
Languor insignis is not improbably technical in the medicine of 
Seneca's day to distinguish the extraordinary nervous "collapse" 
from the ordinary languor, the nervous "spell." Notice the extended 
list of grave symptoms of extreme nervous disturbance in the words 
quaedam . . . percurrens. I find no difficulty in the conjunction of 
insignis and signa; I am not at all sure that a Roman would con- 
sciously identify the sign- element of the two words in ordinary 
speech or writing, any more than in using aedifico ("build" in every 
sense) he ever thought about the aedi-. 

Finally, the rhythm of the suggested reading languoris insignis 
signa praecurrunt is a double cretic-spondaic movement, a remark- 
ably effective clausula termination. 

LXXV, 7 (272: 8) : circa verba occupatus es? iamdudum gaude, si 
sufficis rebus : quando quae multa disces? quando, quae didiceris, 
adfiges tibi ita, ut excidere non possint? quando ilia experieris? 

quando quae VP (the testimony of p ceased at LXXI, 6 fin., 7 
init., videatur. Socrates) with quae corrected to q; in V by the 



144 University of California FuUicalions in Classical Philology 

second hand; guandoque b; quando tarn Q, accepted by Hense,^^ and 
read in the text by P. Hauck.^^ This last reading destroys, however, 
the symmetry in the development of the thought to which the other 
mss. definitely point. There appears to be a regular progression de- 
signed. There are many things to be learned (quae multa) ; when will 
you learn them? These things must next be fixed permanently in 
your mind; when will you so deal with the things you have learned 
{quae didiceris)? They must then be tested in actuality; when will 
you try them out? The quae didiceris clause is, I think, definitely 
corroborative of the quae multa reading. 

It is an interesting question whether the quae multa is incomplete 
as representing a defect in the ms. tradition or is simply a brachyl- 
ogy with a sunt unexpressed. If it is the former, we have the sug- 
gestion of H. Georgii:i5 quando quae multa disces, <didicens>? 
More conformable to ordinary paleographic principles would be: 
quando quae inulta Kdiscenda sunt>, disces? There is a certain 
measure of confirmation for this in the form of expression found on 
the last line of this same page (272) : quae erant complectenda, didi- 
cerunt, and it also affords a sound rhythmical clausula (cretic- 
spondaic), but it is no better than that of quando quae multa disces, 
and I must confess myself as leaning strongly to the brachylogical 
explanation of the quae multa. 

LXXyill, 8 (292: 14) : spiritus naturah prohibitus cursu et muta- 
tus in peius vim suam qua viget admonetque nos, perdit. 

I am not satisfied with admonet nos; "reminds us" of what? 
Seneca is speaking in this passage of the dulling to pain of the nerves 
to the point where sensation finally disappears, where, in short, we 
lose sensation of everything in some particular part of the body 
affected by pain. Now when that occurs, the spiritus, "cut off from 
its natural course and altered for the worse, loses its force," its 
characteristic force, the vim qua viget. What is its characteristic 
quality? The admonet 7ios is, as is so often the case with the second 
member of a pair of thoughts joined by -que, probably epexegetic of 
the first. Th e characteristic quality of spiritus is, then, to remind us 
" Phil. Wochenschrift, XXXIV (1914), pp. 605 ff. 

" P. Hauck, L. Annaeus Seneca, Ausgewdhlte Moralische Briefe (Berlin, 
Weidmann, 1910). 

" Philologus, LXXXIV (1928), p. 88. 



I 



Alexandei-: Seneca's Epistulae Morales (IjXYI-XCII) 145 

of something. It would seem to me that that something is itself ; life, 
here represented by spiritus, surely reminds us of nothing more 
insistently than itself. Read therefore admonetque nos <sui> , and 
compare XI, 2: sui etiam robustissimos admonet (sc. natura). The 
character of the clausula rhythm is not altered by the proposed 
addition; it still remains a cretic-trochee {nos sui, perdit). 

LXXVIII, 11 (293: 9) : 'sed molestum est,' inquit, 'carere adsuetis 
voluptatibus, abstinere cibo, sitire, esurire.' haec prima abstinen- 
tia gravia sunt, deinde cupiditas relanguescit ipsis per [se] quae 
cupimus, fatigatis ac deficientibus. 

per se quae is the consensus of the superior mss. except that Q has 
persequere; this, however, looks merely like a development for the 
worse of the same reading. With all due deference to Muretus, the 
per se seems to me effective; these hunger-pangs and thirst-pangs 
cease of themselves without conscious effort on our part to compel 
them to do so. Compare immediately below: desideria ipsa moriun- 
tur, "desires pass away of themselves." We need also per quae to 
make with cupimus a phrase which, associated with ipsis, means 
"the very senses whereby we form desires." I have no doubt that 
the true remedy for the passage is not to drop se,but to add another 
per (represented, be it remembered, in abbreviation by a single let- 
ter, and that a p before a q) to read: ipsis per se <per> quae 
cupimus. The clausula rhythm remains the same as that yielded by 
the omission of se, namely, a spondee followed by a cretic in which 
the second long is resolved. 

LXXIX, 2 (299 : 10) : neutrum autem incredibile est, nee montem, 
qui devoretur cotidie, minui, nee manere eundem, quia non 
ipsum exest, sed in aliqua inferna valle conceptus exaestuat et 
aliis pascitur, in ipso monte non alimentum habet, sed viam. 

The text given above is substantially that of the ms. consensus; 
the quia non ipsum clause presents variations at its close, but they 
all point to exest. There seems to be no doubt that Seneca's mean- 
ing is as follows: "either theory is tenable, that the mountain is 
being eaten away and hence shrinking, or that it is remaining the 
same because it does not supply the material for the fire that blazes 
up through it, but only serves as a flue, the actual material being 
derived from underground sources other than the mountain itself." 
I do not see how under the circumstances we can avoid giving exest 



146 University of California Publications in Classical Philology 

what seems to me the absolutely obligatory subject < ignis > ; it is 
hardly possible that a writer would mention montem specifically 
and pass over < ignis > . I would place it between non and ipsum, 
reading thus quia non <ignis> ipsum exest; this has some small 
paleographic plausibility, and the clausula, cretic-spondee, is sound. 

LXXX, 1 (303 : 24) : nemo cogitationem meam impediet, quae hac 
ipsa fiducia procedit audacius. non crepuit subinde ostium, non 
adlevabitur velum : licebit tuto vadere, quod magis necessarium 
est per se eunti et suam sequenti viam. 

All the mss. show uno vadere; tuto is a conjecture of Hense's. 
Summers proposes uno <actu> , coll. CI, 4. Beltrami (II) renews a 
conjecture of Erasmus', namely, uni, explained as meaning sine 
interpellatorum comitatu. I believe the required adverb is cito. As 
he is not proposing to depend on other people's roads and other 
people's guidance, he must get along quickly; he has to do all his 
own scouting, he has to be prepared for occasional retracings of the 
path, and hence there can be no lingering. Notice in the preceding 
sentence audacius, with which cito would fit as suggesting at least 
one phase of the audacia. Compare also LXXXIX, 1 : rem ... ad 
sapientiam properanti necessariam. It is the occurrence of audacius 
that makes Hense's tuto a most inappropriate suggestion.^^ If uno 
is to be retained, then Summers has the right idea, but his noun 
seems unfortunate. Dunbabin^^ is happier with his uno <tenore> 
coll. Dial. 8, 1, 1: tunc potest vita aequali et uno tenore procedere. 

LXXXI, 14 (310: 18) :-quantum autem existimas interesse, utrum 
aliquis quod derat a se, quod praestabat, sumpserit an beneficium 
acceperit ut daret? 

The ms. tradition through QVPb is essentially one and the same 
with the exception of dederat in Qb for derat. The various emenda- 
tions are detailed in Beltrami (II). I find myself in agreement with 
Madvig (so also Barker^^) on daret for derat. To all emenders alike 

" Axelson, op. cit., p. 191, thinks otherwise; I fail to follow his argument. 
I agree with him, however, that von lan's crepabit (crepauit, crepuit is the 
line of alteration) is to be accepted. 

" Classical Quarterly, XI (1917), p. 182. One of my former seminary stu- 
dents. Dr. B. L. Charney, reached the same conjecture independently 
coll. XXIII, 7: ex placido vitae et continuo tenore unam prementis viam. 

18 Op. cif.. Vol. IT, p. 329. 



Alexander: Seneca's Epistulae Morales CLXYl-XCU) 147 

quod praestabat is a puzzle. I venture the suggestion that the two 
quod clauses have exchanged their verbs, and that we should read 
utrum aliquis quod praestabat, a se quod daret sumpserit. This I 
translate: "whether a particular individual took what he was offer- 
ing (sc. to the proposed beneficiary) from himself to give." This 
appears to me to give some intelligibility to the quod praestabat; 
further, the transposition produces the double cretic clausula quod 
daret sumpserit. 

LXXXII, 2-3 (316: 1-11) [This passage should read thus] : paula- 
tim enim effeminatur animus atque in similitudinem otii sui et 
pigritiae, in qua iacet solvitur. 'quid ergo?' inquis 'non satius est 
vel sic iacere quam in istis officiorum verticibus volutari?' quid 
ergo? viro non vel obrigescere satius est? deinde idem delicati 
timent, cui vitam suam fecere similem, etc., 

Hilgenfeld regards quid ergo? viro . . . satius est? as an interpola- 
tion, a view which Hense (II) rightly rejects although he feels that 
something has fallen out before deinde. Obviously both critics have 
appreciated the grave difficulty of making anything out of the pas- 
sage in its traditional form, but neither has suggested the possi- 
biUty of a confused ms. tradition resulting from the occurrence of 
two quid ergo? sequences. ^^ My view, as exhibited above, is that a 
scribe either in the archetype common to all our mss.,^" or at some 
point in the tradition earlier than that, after writing quid ergo? 
picked up the viro non sentence in returning to his exemplar, and 
later (after conditivum) , discovering his error, wrote the other quid 
ergo? sequence in the desire to copy all that was before him without 
spoiling his manuscript. Any erasure was impossible, as the mistake 
had now assumed grave proportions. 

The passage, with this change, gives this sense: Little by little 
the spirit is weakened and dissolves into a resemblance with that 
ease which it enjoys and that inactivity in which it hes sunk. "What 
of it?" you say. "Would it not be preferable even thus to lie sunk 
than to be continually tossed in billows of obligations like yours?" 
(Note vel sic iacere.) What of it? For one who is a man would it not 

" But see H. Georgii, Philologus, LXXXIV (1928), pp. 88-89, where this 
possibility has been noted; it is employed, however, as the basis for a trans- 
position of sentences somewhat different from that suggested by me. 

2" This seems to have been pretty conclusively established by O. Foerster 
(as cited in note 3, above) . 



148 University of California Publications in Classical Philology 

be preferable even to feel rigor mortis coming on?^^ (Note vel ohri- 
gescere as a retort to vel sic iacere. Ohrigescere suggests death, and 
hence gives the transition to the next point, introduced by deinde.) 
Next these same fops fear that to which they have made their life 
so like. There is a difference between leisure and being laid away. 
Either excessive application or absolute desuetude is deplorable. 
(Utraque looks forward to contractio and torpor, not back to otium 
and conditivum). According to my way of thinking, the man who 
lies sunk among perfumes^^ is just as dead as he who is dragged by 
the (executioner's) hook. Leisure without letters is death; it means 
a man's being buried alive. 

LXXXIII, 19 (328 : 14) : refer Alexandri Macedonis exemplum, qui 
Clitum, carissimum sibi ac fidelissimum, inter epulas transfodit 
et intellecto facinore mori voluit, certe debuit. 

The ms. consensus is deruit. For this Beltrami (II) in his appara- 
tus criticus tentatively suggests the following explanation: an 
retineri potest intransitive adhibitum (sc. in voluntariae mortis 
speciem)? The account in Quintus Curtius (VIII, 2, 4-5) certainly 
suggests that there was a good deal of play-acting about Alexan- 
der's alleged attempt to take his life, but the verb is rare, and 
Beltrami has to read too much into it. 

Viewing the matter in a purely paleographical light, I came to the 
conclusion that decuit was the most probable word for the original 
text and subsequently found that a few of the recentiores exhibited 
that reading. There is this to be said for it, I think: (1) that the in- 
concinnity voluit rfecwti .(personal and impersonal verb) is striking; 
(2) that the appropriateness of Alexander's taking his own life under 
the circumstances is at least as classical an approach as that of his 
moral obligation; (3) that the end rhythm certe decuit (in which is to 
be included the last syllable of voluit) is sound, being a spondee fol- 
lowed by a cretic in which the second long is resolved ; (4) that had 
decuit been the ms. consensus we should have had no lack of explan- 
ations of it. Of course, if decuit is accepted, since mori the object of 
voluit becomes the subject of decuit a semicolon should be placed 
after voluit. 



" Cf. Cic. In Verrem II, 4, 40, § 87. 

^^ There is a double meaning in these words; they may refer to the attend- 
ant graces of a feast or to the trappings of a corpse. 



Alexander : Seneca's Epistulae Morales (LXVI-XCII) 149 

LXXXIV, 1: (331: 6) : studio quare prosint (sc. itinera), indicabo: 
a lectionibus nihil recessi. 

Nihil was added by Buecheler and adopted by Hense as shown 
above. None of the principal mss., however (QVPb), suggests a 
negative; non appears in the later mss., and this form of negation 
Axelson accepts on the usual rhythmic grounds. ^^ Hense comments 
on non: fortasse recte. 

I am afraid, none the less, that Axelson has done nothing to meet 
Summers' argument:^* "Seneca says the travelling is good for his 
studies, which implies something more than that it does not entirely 
interrupt them." With regard to the relation of o lectionibus recessi 
to what follows, pace Axelson, the connection is as follows: "I have 
given up my reading (sc. on my drives). Reading is, however, neces- 
sary to prevent the growth of self-satisfaction and also to point the 
way to further research. It is also a useful relaxation for the brain, 
which gets tired even of study; it builds it up again, but there must 
be study to accompany it."^^ The natural conclusion of this train of 
thought, not actually expressed by Seneca, is: "I add these reflec- 
tions on the place of reading, that you may not think that I have 
dropped lectiones otherwise than on my rides." Note the autem in 
the sentence immediately following recessi; it establishes a clear 
contrast if one stands with the mss. reading and refuses to allow it 
to be spoiled by the addition of a negative. 

LXXXIV, 8 (333 : 7) : puto aliquando ne intellegi quidem posse, si 
magni viri *** nee enim omnibus, quae ex quo velut exemplari 
traxit, formam suam inpressit, ut in unitatem ilia conpetant. 

The asterisks in Hense's text rather throw the reader off the track 
by creating the idea that a lacuna actually exists in the mss., when, 
as a matter of fact, it is created by Hense himself. 

QV : si magni viri nicenii (V : necenii) 
Pb : si magni viri nee enim 

The codex Ottobonianus-Romanus 1579 gives si magni vir ingenii, 
and in accepting this most modern scholars have concurred. 

" Op. cit., pp. 143-144. 
^* Op. cit., p. 98, in. 2. 

2^ I punctuate with a semicolon after fatigatum, and regard et before studio 
as the equivalent of etiam. 



150 University of California Puhlications in Classical Philology 

In what follows, velut has been found troublesome; as Summers 
says:^^ "but exemplar needs no apology." Hence he reads with 
Schweighaeuser voluit; Kronenberg^^ velit. The objection taken to 
velut seems sound unless quo is to be regarded as the short-form 
indefinite substantive, found not infrequently in combination with 
relatives and relative adverbs, and also interrogatives, for example, 
XL, 6: quid enim quis discere . . . velit? The sense would then be: 
"which he has drawn from anyone, as from a pattern"; this pre- 
serves the ms. tradition and meets Summers' objection because 
exemplar is not, under this interpretation, being apologized for, but 
is added as a graphic example of the way in which a man of great 
native talent uses the work of anyone (quo) on whose work he draws. 
Niemeyer^^ proposes ex <ali> quo velut exemplari with the same 
idea in mind but perhaps dubious about the possibility of enclitic 
quo (cf . si quis, si quern) as object of a preposition.^ 

LXXXVI, 8 (347: 20) : itaque quae concursum et admirationem 
habuerant, cum dedicarentur, devitantur et in antiquorum nu- 
merum reiciuntur. 

There is no need for supplying any word here ; et is for etiam, and 
the sentence translates: "Consequently, places once crowded and 
popularly admired at the time they were dedicated are actually 
dated back into the class of antiquities." Antiquorum means, as 
Beltrami (II) explains, antiquitate obsoletorum. Beltrami (II) takes 
the same view of the passage, but would read <ea> after et, un- 
necessarily, I think. 

LXXXVI, 12 (348: 27) : hoc loco dicet aliquis: 'olim liquet mihi 
immundissimos fuisse.' 

The ms. reading is, however, aliquotis (VM), aliquoties (Q), and 
aliquo (Pb), this last really pointing the same way as the others. 
Aliquis is introduced into V by a late hand. 

Beltrami (II) reads <aliquis> aliquotiens. Aliquotie{n)s I be- 
lieve to be right, but I would certainly not attach it, as Beltrami 
does, to fuisse in sense. It belongs with liquet mihi: "sometimes" (i. 

« Op. cit., p. 100, fn. 
" Classical Quarterly, I (1907), p. 207. 
M Philologus, LVIII (1899), pp. 446 ff. 

" Dr. B. L. Charney seems to set all doubt on that point at rest by the 
citation of CVIII, 19 fin. : Si in quo cognatus aliqui spiritus hospitaretur. 



Alexander : Seneca's Epistulae Morales (IjXYI-XCII) 151 

e., in spite of the great ideals about our forefathers which have 
been impressed on us in childhood, and which dominate our think- 
ing, like so much of our early education) "I feel quite certain that 
they were downright filthy." 

No aliquis is necessary. Seneca often gets along comfortably 
without expressed pronoun subjects. Hoc loco dicet will stand by 
itself without any addition. 

LXXXVI, 14 (349: 11) : didici ab Aegialo, diligentissimo patre 
familiae, is enim nunc huius agri possessor est, quamvis vetus 
arbustum posse transferri. hoc nobis senibus discere necessarium 
est, quorum nemo non olivetum alteri ponit: quod vidi illud t 
arborum trimum et quadrimum fastidiendi fructus aut t depo- 
nere. 

This passage is sometimes asserted to be corrupt beyond re- 
covery. Yet a careful consideration of the context in order to deter- 
mine accurately the drift of the thought, combined with a study of 
such attempts at reconstruction as have been made already and a 
determination of the meanings of the words apparently preserved to 
us in the ms. tradition, suggests that the text as we have it is by no 
means to be despaired of. 

As for the context: (1) It brings out that a group of trees (arbus- 
tum) can be moved quamvis vetus— nor need vetus mean "old" 
absolutely, but only relatively to something else. In the present 
case, for instance, an ohve tree at three or four years of age is 
"old" relatively to an oHve branch (i.e., a slip) just planted as the 
start of a new tree. Consult the Encyclopaedia Britannica, s. v. 
Olive, and compare with what is there said the description of the 
second method of planting as given in paragraphs 19 and 20 of this 
present letter. (2) This fact of the possibilities of late transplanta- 
bility is something old men should become acquainted with for 
their own profit; otherwise they are just setting out orchards for 
posterity. With the ordinary method of planting from slips you do 
not get trees into bearing before from five to seven years. (3) 
Obviously, therefore, from the standpoint of an old man, anything 
that would cut down the elapsed time between planting and ade- 
quate bearing would be desirable. So, if one can successfully move 
trees already well advanced in growth, that would materially help. 

As for previous attempts at amendment of the passage, one may 



152 University of California PuMications in Classical Philology 

refer to Gronoviiis' long note^" out of which emerges: quod vidi 
ilium arborum trimum aut (inaccurately for et) quadrimum fas- 
tidienti fructus autumno deponere. I believe he is right in ilium for 
illud, the masculine pronoun referring to Aegialus above, and also 
in his insistence on deponere as meaning "to plant," surely con- 
firmed by depositum in § 17 and deposu.it in § 19. Beltrami, after 
some aberrations in his first edition, adopts in (II) substantially 
the proposal of H. Georgii,^^ namely, quod vidi ilium arborum 
trimum et quadrimum <non> fastidiendi fructus <c>aut<e> 
ponere. The <non> had already been proposed by Madvig. Bel- 
trami (II) makes the single change of cautum for caute. I confess to 
being impressed by neither cautum nor caute on paleographical con- 
siderations, or any other. Summers^^ offers quite an elaborate re- 
construction of the passage, but his accompanying translation ap- 
pears to me to indicate that he has missed the drift. I believe he is 
right, however, in retaining fastidiendi untampered with. I cannot 
agree in his alteration of arborum trimum etc. to arbor em trimam 
etc. The genitive plural forms trimum and quadrimum are relics of 
the older language and are likely enough to survive in agricultural 
phraseology.^' 

My own suggestion is: quod (sc. olivetum) vidi ilium (sc. 
Aegialum) arborum trimum et quadrimum, <non> fastidiendi 
fructus ant<e>, deponere. Here the genitives arborum . . . 
quadrimum depend directly as genitives of description on quod, 
and the <non> fastidiendi fructus as a like genitive on arborum. 
Ant<e> means "before they were moved from the nursery to the 
orchard proper." I hold the <non> to be necessary because the 
whole point of the passage seems to be that an old man can speed 
up the bringing of his olivetum into effective bearing by moving 
trees already relatively well advanced (vetus arbustum), which have 
already borne a good showing of fruit. I translate thus: "I learned 
from Aegialus that a stand of trees can be moved, even though well 

3" Joh. Fred. Gronovius: ad L. & M. Annaeos Senecas Notae: Lugd. Batav. 
Ex Officina Elseviriana, 1649: pp. 220-221. 

31 Philologus, LXXXIV (1928), p. 91. 

'* Op. cit., footnote to p. 104. 

3' One should also record the conjecture of E. P. Barker, op. cit., Vol. II, 
pp. 329-330, viz.: quod vidi illudi: arborem trimam et quadrimam fastidiendi 
fructus aude tu ponere: te quoque etc. 



Alexander: Seneca's Epistulae Morales (LXVI-XCII) 153 

advanced in growth. This is necessary for us old men to learn, 
every one of whom is setting out an olive orchard for somebody 
else : I saw him planting one consisting of trees three and four years 
old, already bearing no mean volume of fruit." I do not feel that 
quod in the connection in which it occurs presents any difficulty in 
being understood as the equivalent of quorum unum. I submit 
finally that this reading conforms to the general demands of the 
thought of the context and at the same time entails a minimum 
amount of alteration of the ms. tradition at this point. One may 
also direct attention to the double cretic clausula rhythm obtained 
through aw^ < e > deponere.^ 

LXXXVII, 1 (351 : 14) : Naufragium, antequam navem conscen- 
derem, feci, quomodo acciderit, non adicio, ne et hoc putes inter 
Stoica paradoxa ponendum, etc.. 

As the text stands, I do not see how one can escape the logic of 
Summers' enquiry :^^ "How can Seneca prevent Lucilius's reckoning 
this surprising statement a paradox unless he explains it?" In con- 
sequence he reads nunc (iic) for non (h). I have tried^*' to meet the 
difficulty by interpreting non as having the force of nonne, as often 
in Seneca colloquially, and treating adicio as a deliberative indica- 
tive (cf. do in LXXIX, 5, and servio in LXXVII, 15), securing thus 
the idea "am I not to add?" and so getting rid of the objectionable 
negative force of the non. If we regard Seneca's subsequent account 
of the camping trip as an explanation of how the "shipwreck" hap- 
pened, we are bound to seek some method of escape from the em- 
barrassment created by the non. The absurdity of standing by the 
ms. reading unless my somewhat forced elucidation above is ac- 
cepted is revealed in the translations, as, for example. Barker's:" 
"How that happened I refrain from adding, for fear you should 
want to register it as an extra Stoic paradox." But surely we then 
come back to Summers' objection that LuciHus is bound to regard 

3^ Dr. B. L. Charney thinks the position of ante in my conjecture difficult. 
My view is that Seneca has placed it abnormally, or at least elsewhere than 
he ordinarily would, to secure his double cretic clausula. 

35 Op. cit., p. 107, fn. 

36 "Notes and Emendations to the Epistulae Morales of L. Annaeus 
Seneca," Publications of the University of Alberta, p. 10. 

s'Op.dL, Vol.11, p. 48. 



154 University of California Publications in Classical Philology 

it as another Stoic paradox unless and until the explanation is 
provided. Read 7iunc with Summers. 

LXXXVII, 3 (351 : 27) : de prandio nihil detrahi potuit : paratum 
fuit non magis hora, nusquam sine caricis, numquam sine 
pugillaribus. 

The consensus of the mss. is non magis or a with P showing hora. 
But an hour seems a long period of preparation, for as simple a lunch 
as that which Seneca hints at, and hence various emendations have 
been proposed. Muretus long since offered sine magiro, sufficiently- 
remote from the ms. tradition, nor do I think that Seneca, with his 
well-known reluctance to employ Greek words w^here sound Latin 
terms were available, would have written magiro for coquo. This 
objection applies also to H. Georgii's non magiri cura, which, how- 
ever, is paleographically much sounder. Hense's non magidis ora, 
"not on the edge of a plate," should, I think, be included in any 
collection of the curiosities of emendation. Summers comes back to 
the domain of possibility with non agminis cura, that is, "without 
the labor of a horde (of slaves)"; he compares Dial. 9, 1, 6-8, where 
the phrase agmen servorum occurs. The word agmen is used in Latin 
"de hominum turba, multitudine (nee refert, utrum in actu eundi 
sint necne)";'^ compare in Seneca's prose Dial. 9, 5, 3; Benef. 6, 33, 
4 ; Epist. XCV, 24. Summers' suggestion associates itself admirably 
with paratum. 

The same remark applies to another suggestion, non agminis 
mora,^^ "without the delay involved in having a troop of slaves 
get it ready"; it comes close to the ms. tradition, and appears to be 
sound Latin with the Silver Latin flavor of epigrammatic expres- 
sion. The rhythm is hypodochmiac (- ^ - ^ -), Axelson's fourth 
type."" 

LXXXVII, 41 (362: 17) : banc satius est suadere re et expugnare 
adfectus, non circumscribere. 
hanc: QV and the manus prior of P 
haec: b and vulg., also Fickert, Haase 
hac, hinc, hoc: variously the recentiores 

3* Thesaurus linguae Latinae, Vol. I, col. 1342. 

s9 Proposed in my Seneca seminar, March 6, 1940, by a graduate student, 
Mr. J. T. S. Morris, now Latin master at Taft (California) High School. 
" Op. cit., p. 26, n. 35. 



Alexander: Seneca's Epistulae Morales (LXVI-XCII) 155 

The question raised at lines 9-10 concerns (1) the value of 
arguing by syllogisms the pros and cons of a law to abolish wealth, 
and (2) the probability of securing certain sweeping results on char- 
acter by the minutiae of logic ; and the question has only to be asked 
to be answered negatively by its own inherent absurdity. This 
method is therefore to be regarded as rejected. The question was 
put forward by the words: his interrogationibus suasuri aut dis- 
suasuri sumus? One could wish that the alternative had been put 
in a somewhat similar way. Actually the alternative must lie in 
hanc satius est suadere, and, as it is a question of an alternative 
method of getting the law de abolendis divitiis passed, it will not do 
to say with Buecheler, Hense, Beltrami that hanc means hanc legem, 
unless, like Hense, inserting a <re> after suadere, you introduce 
the alternative method in some way or other. It was probably the 
reahzation of this difficulty that produced haec of b, the vulgate, 
and the editions of Fickert and Haase, where haec refers to the 
numerous considerations of the preceding sentence, based not on 
syllogisms, but on Roman history and experience. But where the 
ms. authority is so strong for hanc, it will be best to retain it and 
make the reference to paupertatem (as Fickert ad loc. suggests 
some scribes did), which is, after all, the leading word in the pre- 
ceding sentence, though that fact gets rather obscured by the long 
negative presentation of it from divitias autem on. The translation 
will be: "It would be better to endorse poverty, and (thus) attack 
the passions in force and not simply try to outmaneuver them." 
Endorsing poverty outright means abandoning choplogic and di- 
recting a drive straight against such passions as acquisitiveness, 
love of ease, and love of display; hence expugnare adfectus, non cir- 
cumscrihere is all an exegesis of hanc suadere. 

LXXXVIII, 17 (367: 4) : ego quid futurum sit, nescio: quid fieri 
possit, scio. ex hoc nihil desperabo, totum expecto: si quid 
remittitur, boni consulo. 

The word desperabo has been challenged; we have detraho (M. 
Bonnet), adspernabor (Summers), deprecor (Maehly), and depreca- 
bor (Axelson). Exception is therefore taken, it will be noted, to both 
the word and the tense in which it stands. Hense considers the ms. 
reading sound and Beltrami (II) concurs. I wish to express my 
agreement with them. 



156 University of California Publications in Classical Philology 

It is an ordinary experience to find spero with the sense "I 
expect," where no sense of hope is involved, indeed often quite the 
contrary, and certainly by logical considerations despero should 
mean "I fail to expect," just as, when spero means "I hope," 
despero means "I do not hope." In other words, whatever the sense 
of spero, despero is its antithesis. Ex hoc nihil desperabo therefore 
means: "of this (sc. what I know can happen) there is nothing that 
I shall not expect." This is at once explained by totum expecto: "I 
(always) face the whole thing." Note also what follows: "if any 
part of it is remitted (sc. and it probably won't be much), I am 
very thankful." Those who find a difficulty in the passage un- 
doubtedly do so through thinking only of the sense "despair," and 
failing to recognize that desperabo is here essentially a negative 
verb, the negative quality of which is cancelled by the nihil. Cf. 
CIV, 12: alia sperando, alia desperando. 

Nor is there anything wrong with the future tense. Desperabo is 
written with an eye on what his attitude will he when any particular 
combination of mischances arrives and he begins to consider the 
possibilities one by one, but in the totum expecto he is expressing in 
general (hence the present tense) what his feeling is about all fu- 
turities. If the form of statement now paratactically arranged be 
altered to hypotaxis, this comes out at once: "since I alwaj^s antici- 
pate the whole (of what I know can happen), I shall not on any 
particular occasion fail to expect any part of it." 

LXXXVIII, 44 (374: 24) : Parmenides ait ex his, quae videntur, 
nihil esse uno excepto .universe 

The ms. reading is nihil esse universo, and this Diels*^ defends 
with the subsequent approval of Hense.*^ The numerous conjec- 
tures that have been made, some of them decidedly cryptic, are 
recorded by Beltrami (II) in his critical note ad loc, and I should 
be sorry to add to the list if I could convince myself that Seneca 
would be satisfied to render the vTrdpxetv of Diels' suggested original 
by esse alone. 

Vivona^^ is quite right, in my judgment, in maintaining that the 
emphasis h ere must be on the unum; compare § 45 extr., of this 

" Vorsokratiker^, p. 170, 35. 

*2 Rheinisches Museum, LXXIV (1925), p. 124. 

" Op. cit., p. 39. 



Alexander: Seneca's Epistulae Morales (LXVI-XCII) 157 

same Epistle : si Parmenidi, nihil est praeter unum. Hence the unum 
should appear here as well as in the statement of § 45, and univer- 
sum is insufficiently definite. He suggests, attractively enough, uno 
everso: "if the One is upset." Paleographically uno iverso would be 
more acceptable ; that the meaning would be the same may be de- 
duced from Horace, Epod. 10, 5, inverso man, and Odes, 3, 5, 7, 
inversi mores. Forcellini offers kvaarpk^oi as a Greek parallel for 
inverto, and possibly Seneca was actually translating some such 
phrase as rod evo^ ava<TTpa<i)ePTos. 

As for Parmenides' general position see Plato, Theaet. 180 e 3: 
"all things are One, and the One remains at rest in itself, having no 
place in which it moves." Obviously then, uno inverso, there could 
be no existence for the ordinary phenomena of life (haec quae 
videntur), because they are absolutely involved in the maintained 
existence of the One. 

LXXXVIII, 45 (375: 2) : haec omnia in ilium supervacuum stu- 
diorum liberalium gregem coice; illi mihi non profuturam scien- 
tiam tradunt, hi spem omnis scientiae eripiunt. satius est super- 
vacua scire quam nihil. 

The latter sentence, satius . . . nihil, to be agreeable in meaning 
with the context, should be printed as a question; otherwise it is 
really inconsistent with the sentence next following, where Seneca 
indicates that there is nothing to choose between the one group and 
the other, and particularly inconsistent with lines 13-15 below. The 
question would express utter discouragement coupled with con- 
siderable annoyance: "would be it any better to know useless 
things than to know nothing?" Repeatedly in Seneca interrogations 
have no verbal sign. 

This is less drastic than the method of dealing with the passage 
advanced by P. Thomas,^^ who writes as follows: Apage insulsum 
istud emblema Satius . . . nihil, quo tenor ille orationis inter- 
rumpitur: Illi . . . hi . . . illi . . . hi. Adscripserat scilicet margini 
magistellus aliquis indicium suum. I quote Thomas to draw atten- 
tion to the fact that the impossibility of the standard text, at least 
with its present punctuation, has suggested iteself to others than 
myself; I do not accept his exclusion of the whole sentence. 

*^ Mnemosyne, n.s., Vol. XLIX, p. 28. 



158 University of California Publications in Classical Philology 

LXXXIX, 8 fin. (377: 25) : non enim quemadmodum in iis, qui 
aliquid ex distanti loco ferire conantiir, alibi est qui petit, alibi 
quod petitur. nee quemadmodum itinera quae ad urbes perdu- 
cunt, sic viae ad virtutem sunt extra ipsam: ad virtutem venitur 
per ipsam, cohaerent inter se philosophia virtusque. 

The mss. Q and B (now our only first-class authorities since we 
have passed the end of LXXXVIII) read extra ipsam sunt, though 
originally in B sunt came after ad virtute (sic) and was subsequently 
struck out in that place and written in above ipsam. In the Hense 
text above, the additional words (those in italics) are a conjecture 
of Buecheler's; they must be eliminated in considering Beltrami's 
reading below. 

Beltrami (II) holds by the reading adopted by him in (I), 
namely, ipsa for ipsam, the latter being a relatively easy error with 
per ipsam so close by. Extra is to be treated as an adverb with the 
value of a predicate adjective; ipsa refers to itinera. Beltrami is 
undoubtedly right, though I do not profess to understand his ex- 
planation of ipsa; the Latin signifies to me: "nor, as roads which 
conduct to cities, themselves lie outside (sc. the cities)." Our second 
quemadmodum simile is then on an almost complete par in its 
method of statement with the first, and our whole sentence consists 
of two examples of what the progress to virtue does not resemble, 
followed by a statement in the main clause of how that progress is 
actually achieved, namely, ad virtutem venitur per ipsam. Bel- 
trami's punctuation is also right as against Hense's, but I think that 
the last words, cohaerent etc., are best set off as a separate sentence. 

XC, 9 (385: 17) : furcae utrimque suspensae fulciebant casam. 

Summers^^ remarks on suspensae: " 'set up,' but I can find no 
other ex. of such a force: the natural meaning would be 'sup- 
ported' " (§ 8, n.). His note on § 8 reads: "suspendit — 'rear on 
high': it is an architectural term for supporting a building or roof 
on arches or pillars (so 84. 12 magno aggestu suspensa vestibula)." 
Compare also XLI, 3 med., where the same idea is applied to na- 
ture's architecture; the cave's roof is a grand arch to sustain the 
weight of the mountain above (montem suspenderit), and LIX, 12, 
where the idea is transferred to describe a rider astride ; his legs form 

«Op.a<., p. 316. 



Alexander: Seiieca's Epistulae Morales (LXVI-XCII) 159 

the arch which is sustained by the horse (cms suspensiim equo). In 
our present passage the casa depends for its whole support on the 
furcae which carry the ridgepole. ^^ The natural facts of the case 
seem to point to suspensam as read by Lipsius.^^ 

XC, 22 (389: 24) : narrat enim quemadmodum rerum naturam 
imitatus panem coeperit facere. 'receptas' inquit 'in os fruges 
concurrens inter se duritia dentium frangit, et quicquid excidit, 
ad eosdem dentes lingua refertur; tunc vero miscetur, ut facilius 
per fauces lubricas transeat.' 

Tunc vero is the ms. consensus, retained by Hense, as above, and 
Beltrami. Some of the recentiores insert salivae or saliva before 
miscetur, whence Summers, omitting tunc vero, reads salivae <re> 
miscetur. Buecheler's et umore and P. Hauck's tunc umore are both 
governed by a desire to parallel the aqua of page 390, line 10 (tum 
farinam aqua sparsit), as well as by the general feeling that there 
should be a clear statement of that with which the food mixture is 
lubricated, but the manuscript justification is entirely lacking. I 
suggest tunc universum; if a haplography took place between unc 
and uni, as might easily enough have happened in minuscules, we 
come to tunc versu. This lies appreciably nearer the manuscript 
reading than any other suggestion that has been offered. 

In the preceding part of the sentence, we have outlined for us in 
the first place the process of grinding by the teeth of grains of 
wheat taken into the mouth ; there follows a note to the effect that 
whatever is missed in the first operation is carried back by the 
tongue to the teeth, there to be submitted to the same grinding 
process. This might conceivably happen more than once; tunc 
universum miscetur, that is, "then the whole (sc. now completely 
ground to powder) is reduced to paste." As the object of the mixing 
is at once explained to be the securing of a passage of the chewed 
grains down the throat, the addition of the saliva may be taken for 
granted and is apparently not expressed, though possibly some 
regard should be had to the recentiores with their salivae or saliva 
as noted above. But I do not believe the word to be indispensable; 

" Barker, op. cit., Vol. II, p. 83, translates: "forked poles leaning inward 
from either side formed the groundwork of the hut." I do not follow his archi- 
tectural scheme. 

" G. Windhaus, in Programm des Ludwig-Georg' s Gymnasium zu Darmstadt, 
Herbst 1879 (Darmstadt, H. Brill, 1879), p. 4, concurs. 



160 University of California Publications in Classical Philology 

every human being has the experience of the chewing and swallow- 
ing process and hardly needs to be reminded specifically of its 
every detail. 

Universum in the sense of "the whole," that is, the universe, is of 
course a commonplace in Seneca, but in the ordinary, nonphilosoph- 
ical sense also, such as is required to confirm the emendation sug- 
gested, it will be found in the following passages at least and 
possibly in others as well: LXXXIX, 2; XC, 39; XCIV, 4; XCIV, 
21 ; also Dial. 4, 31, 7; N.Q. 1, 14, 4. 

The rhythm in the clausula is the dispondaeus. 

XCII, 2 (404: 18) : haec enim sola non submittit animum, stat 
contra fortunam : in quolibet rerum habitu securos servat. 

QBA: servitus servat; securos servat is found only in Gruter's 
fourth Palatine codex, and this gives no indication of a possible 
source for the error, while, as Beltrami says in the preface to 
Volume II of his first edition (p. xliii), any correction, in order to 
gain consideration, must take servitus into account. Beltrami him- 
self reads interritos, a conjecture founded in turn on Koch's se 
interritam. Some recentiores, followed by Lipsius, Ruhkopf , Fickert, 
and Haase, read se virtus, but the change of subject from ratio to 
virtus seems very difficult; the argument all along has been about 
ratio. 

The error here appears to me to be of that type which arises when 
a transcriber, noting two words at one and the same time and pro- 
ceeding to record them in his copy, applies to the first word the 
dominant syllable of the second, with consequent elimination of the 
first half of the first word. I therefore think that what originally 
stood in the text was a passive participle in -itos followed by servat, 
which became, as just explained, servitos servat, with an obvious 
subsequent correction to servitus. Beltrami's interritos answers the 
specifications of this theory and gives good sense along with good 
clausula rhythm ; my explanation of the nature of the error provides 
him a better argument than the rather overelaborate theory he 
himself offers at the point mentioned above. 

Interritus is a well-attested Senecan word as a description of the 
perfect sage. Beltrami quotes {loc. cit.) three instances from the 
Epistles and two from the Dialogues. 



Alexander: Seneca's Epistulae Morales (LXVI-XCII) 161 

XCII, 4 (405 : 9) : nam si possunt aliquid non honesta conferre ad 
optimum statum, in his erit beata vita sine quibus honesta. 

It should be observed that honesta at the sentence end is Hense's 
conjecture; the QBA reading is non est. This is adequately explained 
by Kronenberg^^ and by Vivona.^^ I cannot agree with Beltrami's 
view that si has the force of etsi and with his insertion of < tamen > 
before in his. The point is this: if there are non honesta which can 
contribute anything to the optimus status (i.e., the heata vita), then 
that state cannot be called perfect unless it has these things. As 
Kronenberg puts it: iam non conferent ("make a contribution"), 
sed erunt condicio sine qua non. Note the import of the sentence im- 
mediately following in the text, and compare in this same letter 
§§ 14, 15, and especially § 16, wliich should really close the debate 
on the point. Sine quibus non est is also, it should be noted, a proper 
rhythmic clausula, cretic with first syllable resolved, plus spondee, 
a well-established Senecan form. 

XCII, 6 (405 : 25) : adicimus rationali inrationale, honesto inhones- 
tum. ad hanc vitam facit titillatio corporis. 

But the actual ms. readings are QB: magna vitam, and A: mag- 
nevitam. The ad hanc in Hense's text above is an emendation of 
Buecheler's. Beltrami in both editions prints his own conjecture in 
the text : humanae vitae tarn, for ad hanc vitam with tam in the sense 
of tantum; this appears highly improbable. 

The magno of QB (and, in effect, A) seems to me too definite to 
be emended away. It looks like the third positive member of a 
series of adjectives, rationali, honesto, magno, and as each of the 
two former is negatived by its obvious opposite, one should find a 
negative for magno also. This last is more difficult to provide than 
for the other adjectives, but Erasmus in his second edition of the 
Epistles (Basel, 1529) suggested pusillum. Schweighaeuser put for- 
ward vile, which had also occurred to me independently. It ex- 
plains the paleographic error as one of skipping, and provides a 
dispondaic clausula. 

Read therefore: magno <vile>. Vitam facit titillatio corporis? 
The interrogation sign I accept from Beltrami's text. This ques- 
ts Classical Quarterly, I (1907), p. 208. 
" Op. cit., pp. 40-41. 



162 University of California Puilications in Classical Philology 

tion is assumed to be answered positively (cf . ergo in the following 
sentence) and thus becomes the basis for two further interrogations, 
each lightly tinged with scorn, with the net result of rejecting 
the whole absurdity. 

Erasmus' other way out was thus expressed: Tamen erit lectio 
tolerabilis si magno recipias pro sapiente, et addas interrogationis 
notulam. That would mean the placing of a full stop after inhones- 
tum, and following it with Magno vitam facit titillatio corporis? I 
cannot, however, find magnus used thus alone in the Epistles; the 
phrase is always magnus vir in the sense of the perfect sage, the 
sapiens. Examples are XXIX, 3; LXIV, 2; LXVI, 3; LXVIII, 8; 
LXX, 25; LXXVIII, 10. ^°Thus Erasmus' second suggestion, attrac- 
tive as it is in avoiding any change in the ms. tradition, cannot, I 
feel, be entertained. 

XCII, 34 (414 : 1) : ita ille divinus animus egressurus hominem, quo 
receptaculum suum conferatur, ignis illud f excludat an terra 
contegat an ferae distrahant, non magis ad se iudicat pertinere 
quam secundas ad editum infantem. 

QBA all agree on excludat. Some of the recentiores show exurat, 
taken up by Rossbach, coll. de Remediis Fortuitorum, 5, 4, who 
writes ex<urat an lapis in>cludat. In this I concur with respect to 
the paleographical argument based on the excludat of the manu- 
scripts and to the parallel presented by the passage just mentioned 
above, except in regard to exurat. That verb is not sustained by de 
Rem. Fort. 5, 4; it should rather be consumat, this verb occurring 
twice in chapter 5, once with ignis and once with flamma. But 
consumat as the verb for the sentence under discussion may have 
been rejected because of the conferatur immediately preceding; 
small things affect us all in the matter of choosing this word rather 
than that. It is also possible, however, that it was rejected because 
Seneca was suddenly impressed with a less usual phrase which had 
occurred to him. 

Barker^^ suggests excKoguat an mare il>ludat, and I think that 

'" There is, however, at CXV, 2 (556:4): magnus ille remissius loquitur et 
securius, where the word vir does not occur in combination with tnagnus. On 
the other hand, the ille makes the expression perhaps even more specific, 
"the great man of our discussions," so that I think that the point against 
unsupported magnus still holds. 

" Op.ci^., Vol.11, p. 330. 



Alexander: Seneca's Epistulae Morales (LXVI-XCII) 163 

excoquat merits consideration. The verb occurs in Dial. 10, 13, 1: 
excoquendi in sole corporis cura, where it evidently means "to burn 
black," with reference to the acquiring of the fashionable vacation 
tan, like the excodam of Terence's Adelphi 5, 3, 63, the parallel 
adjective there being atram. In N.Q. 4, 13, 11: sed cor ipsum 
excoquit luxuria, it seems to mean "to consume (to a crisp)," meta- 
phorically speaking. Seneca may simply be striking out a bold 
phrase and a very graphic one when he says that the divinus animus 
does not care whether fire "turns to a black crisp" its receptacu- 
lum. I confess that Seneca has provided me with more starthng 
experiences. Despite Beltrami's objection therefore to a four- 
limbed description, read tentatively: ignis illud ex<coquat an lapis 
in > cludat. 



fo^ 



NOTES ON SOME DIDYMAEAN INSCRIPTIONS 

BY 

JOSEPH E. FONTENROSE 

T. Ancient Greek Inscriptions in the British Museum, 922; first cen- 
tury A.D. 

In A.J.P. LVII (1936), 56 f., I suggested a restoration of the 
mutilated lines 2-4, filling in the prophet's name as Lucius Malius 
Saturninus, son of Lucius (Malius) Reginus. Later, in July, 1936, 
I saw the stone in the British Museum and made a squeeze of it.^ 
I now propose the following restoration : 

1 € 

2 [Ilp\o(pr]Tris AevKLo[s Outre] 

3 [XX]ios AevKiov ''P[L'y]ei[vov Kv] 

4 [pet]j'a Kp[i]a(pos Tra[vT]]yvp [ckos] 

In each of lines 2 and 3 there is room for five or six letters at the 
end. The usual letter height is between 3 and 4 cm. ; the average 
width is 2.5 cm. The lines of this inscription are of uneven length, 
varying from 12 to 21 letters, if I is counted as a half letter. 

The name that follows AewLos in line 2 can have a maximum 
length of 10 letters, since the JOS of line 3 is obviously the ending 
of it. Following the patronymic AevKlov a rectangular incision, 
Q}4, by 5 cm., was made in the stone, which cuts across lines 3 and 
4. Over the upper left-hand corner the curved top of a letter P or B 
can be seen. This supports my former conjecture ''Pr]yeivov as a 
second name in the patronymic. But a difficulty arises from the 
fact that the space between P and E can be no more than 3 cm. A 
spelling 'Ttyelvov is possible, however, through iotacization, of 
which there are several instances in contemporary and somewhat 
earher Milesian and Didymaean inscriptions.^ IF can easily fit in. 

1 This was in the course of a year that I spent abroad as Fellow of the Amer- 
ican Council of Learned Societies, whose assistance made possible the re- 
search of which these notes are among the minor results. I also wish to 
express my gratitude to Mr. J. N. Pryce, then Keeper of Greek and Roman 
Antiquities in the British Museum, for his kindness in allowing me access to 
the stones of A.G.I.B.M. 922 and 923 and in giving me all necessary assist- 
ance. 

" See Bonde Bondesson, De sonis et formis titulorum Milesiorum Didy- 
maeorumque (Lund, H. Ohlsson, 1936), 20 f. He cites Abh. Akad. Berlin (1911), 
Anh. 1, p. 68, line 23 (dedLirvLKvias), SiudJ.H.S. VI (1885), p. 351, no. 102, line 
11 {(TvyKexi-optadaL); both belong to the late first century B.C. 

[165] 



166 University of California Publications in Classical Philology 

At the beginning of 4, Hirschfeld published the first group of 
visible letters as NAKrS$02, of which he could make nothing. The 
second of these letters appears as A on my squeeze, so that Hirsch- 
f eld's conjecture about it is confirmed. The fourth is P and there is 
room for an I between it and S. This gives us Kpia<pos, manifestly a 
misspelling of Kplaros.^ 

Now the name Lucius Vitellius Crispus, son of Lucius, occurs 
upon an unpublished Didymaean inscription. Since OuireXXtos will fit 
the available space at the end of 2 and beginning of 3, it is probable 
that he is the prophet of lines 2-4. His father is a Lucius Reginus, 
if my restoration is correct, who may or may not be the Lucius 
Malius Reginus of C.I.G. 2885. If he is the same, then Lucius 
Vitellius Crispus must be a brother of Lucius Malius Saturninus. 
This would not be impossible among Greeks of Roman citizenship 
adopting Roman names. In Milet I 3. 176^ the brothers Poseidonius 
and Diodorus share the names Marcus Aurelius Granianus.^ Per- 
haps Crispus took the name Vitellius instead of Malius because he 
was a partisan of the Emperor Vitellius ; if so, the date of this in- 
scription must be near 69 a. d. 

Between ''Piyeivov and Kpla(pos there is room for a word of not 
more than seven letters ending in NA. This ending suggests a 
Roman tribal name, probably Kvpelva. This is found in C.I.G. 2880 
(ca. 150 A. D.), line 3: MdpKou OvXtIov [$Xa]//3tavoO Aa/jLo. vos Kvpeivlq. 
^\a]/^iavds ^tXeas. An arrangment of names like the one suggested 
in my restoration is seen in the list of Milesian stephanephors:^ 
Map/cos Kopvr}\LOs MapKOV vlos/KoWeiva Kairircov (26/27 A.D.).^ 

^ See Bondesson, op. cit., 96 f., 102 f., on interchanges of n and * in the 
fourth, third, and second centuries b.c. But here the spelling may be in- 
fluenced by the tendency toward dissimilation of contiguous spirants, which, 
like iotacization, is characteristic of the change from ancient to modern 
Greek. That is, we may have in Kpio-v^os the third stage of the following de- 
velopment: (1) a(p changes to (ttt; (2) the spelling aip is sometimes retained, 
though pronounced air; (3) original <nr is sometimes spelled <r<p.See H.Pernot, 
Grammaire du Grec moderne (Paris, Garnier, 1930) I, sec. 78, pp. 49-51. But 
this dissimilation took place in only certain dialects for a<p, not in all as for 
ad and ax- See Pernot, ibid., Remarque (p. 51). 

^ Milet = Th. Wiegand, Milet: Ergebnisse der Ausgrabungen (Berlin, Rei- 
mer, 1906 — ). The epigraphical part of Vol. I, fasc. 3, was done by A. Rehm. 

^ The two prophets named Publius Aelius in Abh. Akad. Berl. (1908), Anh. 
1, p. 45, and in Rev. Phil. XXVIII (1904), p. 202, are probably brothers. 

«Mt7en3.128. 

' I should expect vi6s after the father's name in line 3 of our inscription, but 
there is no room for it. It appears to be the rule to express it when patronyrn- 
ics are long, but it is usually omitted otherwise; see Uoaiduvios Alo86tov in 
A.G.I.B.M.92ZC. 



Fontenrose: Notes on Some Didymaean Inscriptions 167 

II. Ancient Greek Inscriptions in the British Museum, 923a; on one 
side of a block of white marble broken off at the top; width 
59.7 cm.; letter height 2.5 cm.; average letter width 2.5 cm. (ca. 
100 A.D.). Squeeze. 

I propose the following text : 

la [ Kal Tovs 7rat5a] 

lb [s] elaTLaaev [h rdls 'Ai'ot7ju] 

ots kicl rinepas d[co8eKa ko] 

I ovarjs TToXXrys rrjs KaKe[ias] 

ranievovTo^v Evtvxov [tov] 
5 'ExepacrroL' Kal Ua/x^iXoi; [tov] 

'AvTLOXOV Kal TeifJi-qdels [utto] 

TOV SriiJLOV TKeovcLKLS e[l] 

KOffL xpi'O'aTs Kal ap8pLd[aLV.] 

In line 1 there are only traces of letters at the foot of the line, much 
as shown in Hirschfeld's drawing. The key to the restoration is 
supplied by eirl finkpas 8 in line 2, which reveals that this in- 
scription is one in which an outgoing prophet, listing his benefac- 
tions during his term of office, reports that he has celebrated one 
of the festivals for a certain number of days. See Rev. Phil. XXIII 
(1899), p. 317, no. 33 = C.I.G. 2883c (Vol. II addenda, p. 1120) = 
Le Bas-Waddington 227 : lines If. : [—iaTLaaas/Tovs] TroXetras eirl rifxe- 
pas ly' . . . ; lines 5-8 : . . . irot-qaas 8e Kal rots iraL/[a]l to. Seiirva tv toTs 
'Kvoi/^yp^otsviro <7r]yLkavktrlrifik/pas8' . . . . Rev. Phil. ^^111 (1899), 
p. 318, no. 34, lines 7-10: [. . . Kal iaTtaaas]/ KaO' ■qp.kpav TOLS L^' (pv\as, 
[8ovs haaTu ij.e]/pi8a a', 6/zotcos ratcrt 8o[vs h to2s 'Awt]/7MoTs 8Lavop.r]v 
. . . Ihid. p. 319, no. 35: . . . 6/zoicos iralcal 8ovs ev toTs 'ApoljixoIs 5ta]/ 
vonrjv. . . . C. I. G. 2869, hnes 6-8: . . . /Kal tovs re Koanovs eM/rcot 
Lepa}L kirl 86:8eKa r]p.epa$/[e:ire]Te\ecr[ev]. . . . C. I. G. 2880, lines 11-13: 
. . . TTOL-qaavTWV deccpias eirl t]/ p,kpas 8eKa Kal p.ovonax'iO-S aTTOTo/novs ewl 
ilfikpas 8eKa8vo. . . . Cf. Milet I 3. 134, lines 15 f., 29 f. 

The dative-plural ending at the beginning of line 2 points to the 
feast of the Anoigmoi, and the phrase h roTs 'AvoLyjj.ots fills the re- 
quired space at the end of line 1 neatly, since there is room for just 
about eleven letters, if I is counted as a half letter. It will be noticed 
that lines 4-8 show 18 to 22 letters a line. The letter traces in line 1 
seem to be daTiaaev, the initial letter of which is the second letter in 
the line. Since three of the inscriptions quoted above reveal that 



168 University of California Publications in Classical Philology 

Milesian boys were either feasted or otherwise well treated at the 
feast of the Anoigmoi, we are fairly safe in supposmg that the 
words immediately preceding daHaaev were [/cat rods Tra28a/s]. 

To complete the restoration of line 2 the question arises, For how 
many days did this prophet carry on the feast? The sources show 
that this was left to the incumbent prophet's discretion. All that 
we have here is A, which may be either the numeral letter or the 
initial letter of 8vo, Ska, or 8co8eKa. Except for 8vo, there are instances 
of all these possibilities in the sources quoted above. There is space 
for about eight letters at the end of line 2. The last two are prob- 
ably KA, forming the word Kai with the I at the beginning of 3, so 
that 8o}8eKa most nearly fits the available space. The manner in 
which a feast was carried on for twelve days is adequately sho\Mi by 
the quotation above from Rev. Phil. XXIII (1899), p. 318, no. 34. 

The sense of line 3 is made clear by J.H.S. VI (1885), p. 353, 
no. 105, lines 5-8; that is, in spite of hard times the incumbent 
prophet undertook the great expenses mentioned in the inscription. 
Hirschfeld suggested KaK[oTaddas] for the end of line 3. But this is 
about four letters too long for the gap, and after the second K my 
squeeze shows what appears to be the foot of an E. I therefore 
propose KaK€[las] for KaKias.^ The meaning "hard times," "distress," 
is unparalleled for this word, but it is not impossible for an abstract 
noun that signifies the quahty of Kanos.^ 

The treasurer Eutychus (lines 4 f.) is mentioned in Milet I 7. 
204a, which is dated in the first century a.d. by Rehm, and prob- 
ably in Abh. Akad. Bed. (1924), Pt. 1, p. 19, no. 11, dated in the 
second century by Wiegand. 

III. Le Bas-Waddington 222; J.H.S. VI (1885), p. 352, no. 102 
(E. A. Gardner, ed.); Melanges Weil (Paris, Fontemoing, 1898), 
pp. 150-153 (B. Haussoullier, ed.); Milet I 3, p. 275 (Part II, 
lines 18-22, only; A. Rehm, ed.). 16/11 b.c. 

In the first half of the nineteenth century both Cockerell and Le 
Bas saw the stone and made copies of it. Cockerell's copy is superior 
to Le Bas's, but neither is very good. The stone has not been found 
by the exc avators, French or German, of DidjTna and Miletus. 

* €1 is found for short i in Milesian inscriptions of the second century a.d.; 
see Bondesson, op. cit., 25 f. 

^ KaKOTTjs has the meaning "bad condition" as well as "bad character," 
"bad quality," etc.; see Hdt. 2. 128, 6. 67. 3. 



Fontenrose: Notes on Some Didymaean Inscriptions 169 

Part I (lines 1-17)^° is a Milesian decree that provides for re- 
newal of the practice of appointing a jSotjtos every year in the cult of 
Zeus Soter, since for a long time no one had undertaken the office. 
The decree concludes with provision for a stele on which the names 
of the annual jSorjjoi shall be inscribed. Part II (lines 18 ff.) is the 
first part of a list of ^oriyoi. 

The text of Part I has been best edited by Adolf Wilhelm in 
Beitrdge ziir Griechischen Inschriftenkunde (Vienna, Holder, 1909), 
pp. 177-179. As supplement to his and Haussoulher's commentaries 
I should say that the ^or]y6s is probably not identical with the /SorjTta 
viKiiaas that appears in some extant inscriptions." The ^orjyds, I 
think, is the chief priest of Zeus Soter, elected annually. The big 
event of his term was Zeus's festival, the jSorjyla, for which he must 
make provision and over which he presided. 

Both copies show very little for Part II. Cockerell's copy indi- 
cates a gap of ten lines after the end of Part I, then the following 
letters at the foot of the stone: 

EniTPOI 

YHAPAI . NOT . . 

TOYnpo^HrorriosnosiAQ 

NIOS 
Le Bas's copy, which was made viery hastily, shows muchless for 
these four lines, but more for the intervening section : 

EniET . I 

AN0OYBOH0OSEPrONEni 

. OTTOrn . . . 2TQN 

. . . HPAKAEIAHN 

IIPO . . I 

QN TOTA 

QS 

. . OT 

Eninp 

. . YP 
. TOY 
NIO 

1" There are 17, not 18, lines in Part I; see Wilhelm (cited in next para- 
graph), p. 179. 
11 See Melanges Weil, 149. 



170 University of California Publications in Classical Philology 

Rehm's restoration of lines 18-20 shows that the essential for- 
mula in each item of this list is the date expressed in terms of the 
stephanephorship of Miletus, followed by the title and the name of 
the /307776s of the year. The first five letters of line 19 show that 
the first ^oiryos of the list served in the stephanephorship of Hera- 
cleides Evanthus' son, 16/15 b.c.^^ Enough letters can be seen in Le 
Bas's copy to make it plain that the first to undertake the office after 
passage of the decree was Epinicus Epinicus' son, who is mentioned 
in the decree (line 12). Lines 18-20 thus provide the kej^ to the 
restoration of lines 21 £f., since we have the stephanephor list for 
this period of Milesian history. The only problem is how to make 
use of the fragmentary and unreliable evidence from the copies of 
Cockerell and Le Bas. Only once after the first item do we know the 
name of the /307776s ; elsewhere we cannot know how much space to 
allow for the names. In Part I there are from 233^ to 283^ letters to 
a line, if I is counted as a half letter. 

I propose the following restoration of the whole of Part II, mak- 
ing use of the stephanephor list and giving as much weight as pos- 
sible to the two copies : 

'Ext o-T[e]^[aj'77(^6poy 'Hpa/cXetSow rod Eu] 

avdov 18077765 Trp[u)]TOv 'E7ri[i'tKos 'Einvl] 
20 [k]ov tov 'H[(^at] (Tr[t]a)i'[os.] 

['Etti aTtipavribpifOv rov deov rod] 

[ned'] 'UpaKXelSrjv [ayroer77S /307776s /cat] 

Trpo[(fr]]T[r]s{6 8elva) tov {Selvos).] 

['Etti (TTe(pavr](f6pov EvKparovs tov Mr]v] 
25 [o5]cb[poi; 18077765 (6 8elva)] tov A . 

['Etti] (T[Te(pavr](f6pov Xa.pp.ov tov 6pacrc«j] 

[?'t5]oL' [/307776s (6 8eLva) tov (5elvos).] 

'Etti 7rpo^[77Te6oj'TOs Alo86tov tov 'A7a] 

[^to]i; Tap[e]8p[evcov rojutas 6]moO [Kat i3o77] 
30 [76s 6] tov irpotpriTOv vlos Hoacdoo 
PIOS. 

18-20: The restoration is Rehm's, except that he has tp'ltov, not 
irpQiTov, for the EPrON of Le Bas's cop3^ But this is the first item of 
the list, so that if tp'ltov is the right reading its interpretation is not 
easy. It can not be the equivalent of rpts. If this item is the record of 

" Milet I 3. 127, line 3. 



Fontenrose: Notes on Some Didymaean Inscriptions 171 

Epinicus' third year as ^o-qyos, or if he is the third to hold the office 
since passage of the decree, we should expect two entries before it. 
It remains possible, but not probable, that Epinicus had served 
twice as j3oriy6s in the period before the office had lapsed into 
desuetude. But it is much more likely, in view of the fact that 
Epinicus was instrumental in reviving the ^orjyia, that he would 
take care to point out that he was the first to hold the office after 
the new law had been passed. It is true that Le Bas's copy shows no 
space between P and T ; but it was very carelessly made, as a com- 
parison of Le Bas's copy of Part I with Wilhelm's text will reveal. 

21-23: The restoration as far as 'EpaK\ei8r]v is Rehm's. He is 
right beyond a doubt in his belief that Le Bas skipped a line here 
(21). The stephanephor list shows that the successor of Heracleides 
was 'AiroWuv Atos, who received the office fairly frequently.^^ The 
letters IIPO, shown at the beginning of 23, make the word Tpo(pr]Tr]s a 
probable restoration. The title could appear at this point only if 'the 
prophet of the year were also serving as jSot^tos. When the prophet 
held another office concurrently, he was very likely to indicate this 
circumstance by using the word aifToerrjs." Following TrpocprjTrjs there 
is room for a maximum of 21 letters, ample for a complete name. 

24-25 : At this point, too, Le Bas must have omitted a line. His 
copy shows fiN at the beginning of what ought to be line 24. Yet 
if this is the first line of an entry, as it should be, it would begin 
with EIII. The stephanephor of this year (14/13) was Eucrates 
Menodorus' son, so that the restoration of the date formula is 
certain. It appears obvious, therefore, that the fiN and TOTA of Le 
Bas's copy are no part of line 24. Notice too that Le Bas's copy 
shows fragments of but eight lines between the end of Part I and 
my line 28, whereas Cockerell's copy indicates a gap of ten lines. 
fiN perhaps represents the QP of MrivoSupov. If so, the Q, could not be 
the first letter of line 25, since that would require 313^ letters in 
line 24, whereas the maximum in lines that are certain is 29 (line 
18). If the Q belongs to the patronymic (if not, it is a mistake for 
something else),^^ it can at best be the third letter of the line. This 

" For the date formula when Apollo was stephanephor see C.I.G. 2855 = 
Michel, Rec. 836, lines 1 f. 

" See, e.g., A.G.I.B.M. 923b, lines 12 f.; Rev. Phil. XX (1896), p. 100, no. 4, 
line 2. 

15 There is, of course, the possibility that Le Bas's i2N represents the HN 
of Mr)vo563pov. H would then probably be the second letter of the line. But this 
would almost certainly push this entry into a third line, and that, I think, 
cannot be done without making a restoration of line 27 impossible. 



172 University of California Publications in Classical Philology 

sort of thing is possible in both copies; notice the differences be- 
tween Le Bas and Cockerell in lines 29-31, where they agree upon 
several letters but do not agree upon their positions in relation to 
the beginning of the line. Following the conclusion of the patro- 
nymic and the title ^orjySs, which must be next, there is room for a 
maximum of seventeen letters, enough for a complete name of 
moderate length. The letters TOTA, shown bj^ Le Bas, are obviously 
the beginning of the patronymic. Al68otos tov 'AyaOiov, the prophet 
of line 28, is a possible restoration. 

26-27 : Le Bas's copy shows 122 at the begiiming of line 27. The 
best that we can do with this is to suppose that the 2 is the initial 
letter of (7Te(pavr}(p6pov, and that the Q, is an odd misreading of 
letters of eiri. The stephanephor of the year (13/12) was Charmes 
Thrasonides' son. For line 27 Le Bas shows OT in just the right place 
to be the genitive ending of the patronjonic. After ^orjyos there is 
room for a complete name in the same line. 

28-31 : Both copies make it clear that this entry is dated by the 
prophet's name instead of the stephanephor's. As Haussoullier 
points out, line 30 tells us the reason : the ^orjyos of the 3^ear is the 
prophet's son. We should expect, however, that the stephanephor 
would be named too. But for this year (12/11) the stephanephor 
was again the god Apollo, so that under the circumstances the 
name of the prophet may have seemed enough to date the entrj^, 
which is a long one in any case. 

Haussoullier restores these lines as follows : 

'EtI irpo[(pr]TevovTOS rod Selves] 

[to]v , [^orfyos 6 a]v[TO-] 

[v] TOV irp0ifr]T0V vios nociSo)- 

VLOS. 

This restoration makes little use of the letters that Cockerell's copy 
show^s for the second line. Also the word di\dsion at the end of the 
line is impossible, I think, for this inscription. And the third line is 
too short. 

In A.G.I.B.M. 923c we find a prophet Poseidonius, who is son of 
Diodotus. This inscription is dated in the first century a.d., though 
it has been impossible to say whether earh' or late in the century.^^ 
C.I.G. 2884, of uncertain date, speaks of a Poseidonius who was 

1^ Bondesson, op. cit., xv. 



Fontenrose: Notes on Some Didymaean Inscriptions 173 

three times prophet. We can tentatively assume that we are deahng 
with the same Poseidonius in the three inscriptions; for if he was 
/307776s as a young man in 12/11 b.c, his third term as prophet could 
be as late as 30 a.d. In Abh. Akad. Berl. (1911), Anh, 1, p. 66, an 
inscription of uncertain date, we find a prophet Diodotus Agathias' 
son. This name just fits line 28 and the beginning of 29, where the T 
shown by Cockerell and Le Bas is in the right place to be the last 
letter of the patronymic in the genitive. 

Now in the preceding entries the title ^o-qybs and the name of the 
incumbent followed immediately upon the date formula. But here 
the name and a descriptive phrase occur in lines 30-31. Besides the 
title 1807776s there is room for at least 20 letters between the end of 
the date formula and the words tov Tpo(priTov vids, and Cockerell's 
copy shows that there was something unusual in this space. I sug- 
gest Trap[e]8p[ev(jOV raixlas oJmoO [/cat 13077/76$]; that is, the ^or]y6s Posei- 
donius also served as cult treasurer during his father's year as 
prophet. For the addition of irapeSpevwv to the title rajutas see Milet 
I 3. 150 = Ditt. Syll. (3d ed.) 633, Pt. II, line 20; C.I.G. 2855 = 
Michel,, 72ec. 836, lines 6 f.i^ 

The concluding letters NIOS, which are the whole of line 31, may 
not have been placed at the extreme left of the line space, but inset 
about three or four letter spaces to the right. Cockerell's copy 
seems to indicate such an arrangement. These letters also make it 
plain that the letters IIOSIASi of Cockerell's copy complete line 30, 
so that at least four letters must be restored before tov Tpo(priTov, or 
the line will be much too short. The tov therefore must be further 
to the right than either Cockerell or Le Bas place it. 

IV. Ditt. O.G.I.S. 213; Rev. Phil XXIV (1900), p. 246 (Haus- 
souUier, ed.) = B. Haussoullier, Milet et Didymeion (Paris, 
Bouillon, 1902), p. 35: lines 36-38; early third century b.c; 
decree of Demodamas in honor of Antiochus Soter. 

Dittenberger reads : 

36 ... eldKoXeiadaL 8e avT6v[eis wpoeSpiav h MiXtjtcol] 

37 ToTs Alouvo-'lols Kai kv At5u/xo[ts rots AtSu/zetots h] 

38 rots kvkXLols aycbaiv. . . . 

Haussoullier had read Kai at the end of line 37, but Dittenberger 
objected that the kvkXlol kyCivts are not a third festival in addition 
^' For opoD see Ditt. O.G.I.S. 494, line 3: Trpo<pTiTT]s dfj.ov Kai a[p]/xi-T'pvTavLs. 



174 University of California Publications in Classical Philology 

to the Dionysia and the Didymeia, but are a part of both. This is 
still unsatisfactory, however; for the athletic contests at Didyma 
are not mentioned, for which Antiochus must certainly have been 
granted prohedria. I therefore suggest : 

. . . Kal h A^8v^io[LS rots ad\riTLKo7s Kai] 
ToTs kvkX'lols aycbcTLV. . . . 

The phrase rots ad\r]TLKo7s kt\. refers to kv AiSujuois only. The Dionysia, 
everyone knew, had choral contests only. The name AtSujueta is 
absent in my restoration ; but it did not become common until after 
200 B.C. as the designation of the contests at Didyma; for it was 
then that the Milesians made them pan-Hellenic; see Ditt. Sijll. 
(3d ed.) 590, the announcement of the new quinquennial games, in 
which the earlier contests are referred to as at ayuves kv AiSvuols. 



SENECA'S EPISTULAE MORALES 

THE TEXT EMENDED AND EXPLAINED 

(XCIII-CXXIV) 

BY 

WILLIAM HARDY ALEXANDER 

XCIV, 3 fin. (418: 23) :i Sic qui se ad totam vitam instruxit, non 
desiderat particulatim admoneri, doctus in totum, non enim 
quomodo cum uxore aut cum filio viveret sed quomodo bene 
viveret; in hoc est et quomodo cum uxore ac liberis vivat. 

Suspicion first fell on this passage when Erasmus in his second 
(1529) edition dropped the word enim. This became the vulgate 
text, but Fickert restored enim, and though Haase placed it in 
square brackets, Hense and Beltrami^ have retained it. The latter, 
however, says of Erasmus' decision non inepte quidem, but thinks 
that a doctus est or didicit may reasonably be understood in order to 
give sense to the contrast imphed m the enim. Castiglioni^ is incHned 
to believe that enim belongs with the in hoc est of the following sen- 
tence, but the disruption of the phrase is difficult to explain. H. 
Georgii^ holds that non enim . . . viveret is an interpolation, while O. 
Rossbach^ takes the same view of the in hoc . . . vivat sentence. The 
cumulative effect of these judgments is to create a natural distrust 
of the text, and I beheve this to be well justified. 

My own view is that Rossbach is right in ejecting the sentence in 
hoc est . . . vivat, but that in hoc est retains in a deformed way the 
verb for the non enim sentence preceding, namely, doctus est. Once 
doctus est is replaced in that sentence, the difficulties connected 
with enim vanish. Doctus est was written doct' e; this became trans- 
muted into hoc est, around which and out of which the dull and 
repetitive sentence in hoc est... vivat was built up and interpolated. 

1 The numbers in parentheses give the page and line reference in Hense's 
second edition (Leipzig, Teubner, 1914), hereafter referred to as Hense (II). 
Hense (I) appeared in 1899 under the same auspices. 

2 Beltrami's Edizione Nazionale (Rome, Regiae Officinae Polygraphicae, 
1931). This edition will hereafter be referred to as Beltrami (II). Beltrami 
(I) was published as follows: Vol. I (Brescia, Apollonia), 1916; Vol. II, 
(Bologna, Zanichelli), 1927. 

3 Gnomon (1927), p. 667. 

* Philologus, LXXXIV (1928-29), p. 93. 

5 Philologische Wochenschrift, XXXIV (1914), p. 496. 

[175] 



176 University of California Publications in Classical Philology 

XCIV, 8 (420 : 10) . . . cum ilium in conspectum suae condicionis 
adduxeris et cognoverit beatam esse vitam non quae secundum 
voluptatem est, sed secundum naturam . . . 

All the mss. read voluntatem until we reach the recentiores, when 
voluptatem, ever since read, appears. I think, however, we might 
well here recall the note of Windhaus^ on the B reading voluntatem : 
Haud scio an recte. Dici videtur : Cognoscet beatam vitam non esse 
secundum voluntatem (beate vivere non eum, qui faciat, quae 
velit) sed secundum naturam (sed eum, qui faciat, quae naturae 
necessitas postulet). The actual occurrence at the end of the im- 
mediately preceding sentence of the phrase quae nobis mundi 
necessitas imperat would certainly give point to a subsequent con- 
trast of voluntas and natura, natura here being equivalent to mundi 
necessitas, "the categorical imperative of the universe." 

XCIV, 21 (423 : 18) : concede per se efficacia praecepta non esse 
ad evertendam pravam anuni persuasionem : sed non ideo nihil 
ne ahis quidem adiecta proficiunt. 

The mss. Q B A show simply: sed non ideo aliis quidem adiecta 
proficiunt. Fickert tries to save this reading by placing a mark of 
interrogation after it, but I suspect that to give it the meaning he 
intends, the non would have to stand immediately before proficiunt. 
Erasmus' <ne> before aliis in his 1529 edition appears to be in- 
tended to introduce the necessary second negative, but it seems to 
me to fail in that effort because, locked away in the combination 
ne quidem, it has no. effect upon the non. Buecheler, retaining 
Erasmus' <ne>, introduces before it <nihil>, thus securing the 
requisite second negative, but Linde^ prefers <nihil> only, finding 
no need of completing a ne quidem phrase. Beltrami in both editions 
follows certain recentiores (including q) in reading < non > before 
proficiunt. The clausula adiecta proficiunt (cretic plus spondee with 
resolution of the first long, or straight spondee if the second i in 
proficiunt is consonantalized) is a more usual type than non pro- 
ficiunt (spondee plus spondee with resolution, or even spondee- 
spondee), and this circumstance, combined with the more delicate 
shade of expression in non ideo <nihil> as against non ideo . , . 

^ Georg Windhaus, "Varietas lectionis ad L. Annaei Senecae Epistulas e 
codice Bambergensi enotata," in Programm des Ludivig-Georg' s Gymnasium 
zu Darmstadt, Herbst 1879 (Darmstadt, H. Brill, 1879). 

^ S. Linde, Adversaria in Latinos Scriptores (Lund, 1901), pp. 28-29. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 177 

<non>, leads me once again to express my concurrence with 
Linde's view.^ Cf. the closing words of par. 24. 

XCIV, 34 (427: 3): praeterea ipsum de malis bonisque indicium 
confirmatur officiorum exsecutione, ad quam praecepta per- 
ducunt. utraque enim inter se consentiunt: nee ilia possunt 
praecedere, ut non haec sequantur, et haec ordinem sequuntur 
suum. unde apparet ilia praecedere. 

That this is a difficult and elusive passage will be seen by con- 
sulting the translations. Gummere in the Loeb translation is gener- 
ally hazy, and clearly wrong in rendering ut non by "unless." 
Barker's "former" and "latter" may be right, but are certainly not 
lucid. On the other hand, the note in the Lemaire edition,^ the 
La Grange translation, i" and also the Catalan rendering by Cardo^i 
recognize the treacherous nature of the Latin here and clarify it 
when translating by using nouns instead of pronouns. This is very 
sane and very helpful. It brings out the fact that utraque refers, not 
to officia and praecepta as might be at first supposed, but to ipsum 
de malis bonisque iudicium and praecepta, though the former, being 
in meaning equivalent to principia, is subsequently referred to by 
ilia. The passage translates then as follows: 

"Secondly, the actual judgment on the Good and the Bad [i.e., 
the abstract principles of moral philosophy] is strengthened by the 
performance of moral duties, and to that performance ethical 
maxims lead. For both the theoretical principles of Good and Evil 
and ethical precepts on the same are in harmony one with the other. 
Theoretical principles cannot take the lead without ethical maxims 
following, and these latter keep to an order of their own. Therefore 
it is clear that general principles furnish the lead." 

But the clarity of the argument at the end of this section is far 
from impressive. It would be much sounder logic if it read as fol- 
lows: "Theoretical principles cannot take the lead without ethical 

8 As already in my Notes and Emendations to the Epistulae Morales of L. 
Annaeus Seneca (Edmonton, Univ. of Alberta Publications, 1932), p. 11. _ 

3 L. Annaei Senecae Opera Omnia, ed. M. N. Bouillet (Paris, N. E. Lemaire, 
1829; 5 vols.). The present passage is in Vol. IV, p. 86. 

1° (Euvres de Seneque le philosophe traduites en Frangais par La Grange, 
(Paris, J. J. Smits et Ci«, An III de la R^publique; 5 vols.). The present pas- 
sage is in Vol. Ill, p. 50. 

i^L.A. Seneca Lletres a Lucili, text revisat i traducci6 del Dr. Carles 
Card6 (Barcelona, Fundacio Bernat Metge; Vol. I, 1928; Vol. II, 1929; Vol. 
Ill, 1930). The present passage occurs in Vol. Ill, p. 99. 



178 University of California Publications in Classical Philology 

maxims following, and ethical maxims follow an arrangement 
< not > their own. The clear conclusion is that general principles'^ 
precede <and establish this order>." Thus if <non> is inserted 
after ordinem, the logic is improved, and at the same time a triple 
ere tic rhythm is produced in the clausula, ordinem <non> 
sequuntur suum. 

XCIV, 41 (428 : 24) : Nee tibi facile dixerim quemadmodum prosit, 
sic ut illud intellegam profuisse. 

It is not entirely clear what the subject of prosit is; formally, 
regard being had to the immediately preceding sentence, it should 
be magnus vir vel tacens, but it may be inferred from the illvd that 
the subject of prosit is very loosely defined indeed, and is something 
like "the experience of meeting a great man even though he never 
utters a word." The drift of the whole thought up to prosit is, then, 
this: "I could not easily tell you how such an experience does you 
good." That difficulty then arises is attested bj^ Baehrens' cum 
illud and Buecheler's licet illud for sic ut illud, each emender de- 
signing a "though." Yet my impression is that the thought con- 
tinues thus: "not so easily as I could reaHze that it has done (people) 
good," which would more normally be expressed by non tarn facile 
quam. But tarn facile is condensed quite naturally into sic, and the 
negative continues its effect through from the main clause in a 
manner entirely characteristic of Seneca; sic at once calls forth its 
correlative ut, thus producing the present difficult but still ex- 
plicable text. It is easy to understand Schweighaeuser's desire for 
an indicative intellego, but I do not myself feel that the potential 
intellegam is entirely out of the way, as may be seen if we put the 
thought inversely: "I could more readily realize that such experi- 
ence has done good to people than I could teU you how it does it," 

XCIV, 59 (433 : 25) : Necessarium itaque admoneri est, habere 
aliquem advocatum bonae mentis et in tanto fremitu tumultuque 
falsorum unam denique audire vocem. 

For the est following admoneri Q B A all show et; q exhibits est 
before admoneri, while the vulgate reading was est admoneri et. Est 
for et is Buecheler's sound correction. 



12 Cf. XCV, 54: in supervacuumpraeceptaiactavimus, nisi illud praecesserit, 
qualem de quacumque re habere debeamus opinionem. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 179 

For the et in tanto of Hense's text, B A have etantanto and Q e 
tanto. I have no doubt that Q has the correct reading, of which B A 
shows a simple dittography. Punctuate therefore with a comma 
after mentis and read e tanto, dropping et. The phrase audire vocem 
e tanto fremitu seems to me very graphic; the voice arises out of the 
din, which imphes that at times it masters the contending voices. 

The three things necessary, admoneri, habere, audire, are set 
down side by side asyndetically. 

XCV, 1 (437 : 21) : Scio te in bonam partem accepturum, si nega- 
vero. eo magis promitto et verbmn pubhcmn perire non patior: 
'postea noh rogare, quod inpetrare nolueris.' interdum enim 
enixe petimus id, quod recusaremus, si quis offerret. haec sive 
levitas est sive vernihtas, punienda est annuendi facihtate. 

For the explanation of this passage it should first be noted that si 
negavero does not mean that Seneca is not going to produce the 
promised letter; as a matter of fact, he does so from paragraph 4 on 
(sed ut omisso principio rem ipsam adgredior). The scio te . . . 
accepturum is a sly poke at Lucilius for professing an eagerness he 
does not really feel. "Just because of that attitude of yours," says 
Seneca, "I guarantee dehvery of the asked-for letter all the more. 
I do not intend to let the proverb be forgotten which says: 'Don't 
be keen to beg hereafter what you'd really hate to get.' Sometimes 
we make a great show of asking for what we would refuse if anyone 
offered it to us." In this translation the words "I guarantee de- 
hvery" are needed to bring out the sense; the "I promise" in Bar- 
ker's translation,^^ for instance, is wholly misleading. Then Seneca 
continues: "This sort of thing, whether it is just trifling or a desire 
to appear smart, must be punished by promptness of assent," which 
is admirably illustrated by the story of the persevering recitator 
which follows. 

As for annuendi, the suggestion of Buecheler and Windhaus, I 
find myself entirely at one here with Axelson^^ m rejecting it. Aside 
from the styhstic improbabihty of annuendi emphasized by Axelson, 
the mutendi of B and the promittendi of Q, together with the occur- 
rence of promitto already in section 1, certainly point to the cor- 
rectness of the Q reading; more particularly as the loss of pro- in B 

" E. P. Barker, Seneca's Letters to Lucilius (Oxford, Clarendon Press, 1932), 
Vol. II, p. 140. 
1* B. Axelson, Neue Senecastudien (Lund, Hakan Ohlsson, 1939), pp. 122-123, 



180 University of California Pullications in Classical Philology 

means the loss of only the one-letter conventional symbol for that 
prefix. Compare also the use of promitto in section 10 and in section 
39; the latter is an especially clear case. Promittendi facilitate is 
merely the repetition in another form of eo magis promitto. 

XCV, 16 (422 : 4) : aut palpitatio [corporum] sine intermissione 
vibrantium. 

Corporum has been excised from the text by most editors, but 
those who thus treat it might reasonably be asked to explain from 
what source it arose. Madvig, for instance, transferred it, some- 
what arbitrarily no doubt, to follow tahesque, but at all events he 
did not feel that the noun simply appeared from nowhere. With its 
epithet vibrantium it appears to be the natural parallel to nervorum 
iacentium preceding, the same kind of ailment as described in 
section 17 following by the words multa memhrorum quassatione. 
What Beltrami thinks to gain by replacing corporum with ipsorum 
I do not understand. 

Windhaus' objection^^ that up to this point we have had only 
parts of the body mentioned, and that therefore reference to the 
whole body is unsymmetrical, might be further enforced by the 
observation that the account continues in the next sentence with 
the mention of additional locahzed diseases; yet even so it is purely 
arbitrary to say that this renders impossible a reference to a general 
paralysis agitans at the point where, on the authority of the manu- 
scripts, it occurs. The Epistles are literature, not an organized text- 
book of pathology. 

XCV, 19 (442 : 25) : Nee mirum, quod inconstans variusque ex 
discordi cibo morbus est et ilia ex contrariis naturae partibus in 
eundem compulsa redundant. 

Apparently Hense and Beltrami regarded eundem as meaning 
"the same individual"; personally I find this difficult, and the diffi- 
culty has been shared by others. Stephanus suggested nouns for 
eundem to tie it with, such as locum or ventrem, the latter sounder 
paleographically. Hense^^ would read compulsa ventrem redundant 
to avoid a dactylic hexameter close, but of course in doing that he 
sacrifices the paleographical argument for the loss of ventrem. 
Beltrami, while continuing to read eundem in his text, offers in his 

" Op. cii., p. 15, n. 2. 

18 Rheinisches Museum, LXXXIV (1925), p. 126. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 181 

critical note on the passage the reading in eanuem compulsa ahum 
redundant; this seems to me rhythmically less good than Hense's 
proposal, but by no means impossible. 

The most obvious correction is eandem (sc. partem) ; the phrase 
eandem (partem) used to describe the stomach is made perfectly 
intelligible by the context, and the partem suggests itself readily 
enough from the preceding contrariis partibus}'' 

XCV, 26 (445 : 2) : veneriae spondylique et ostrea eatenus cir- 
cumcisa, qua eduntur, intervenientibus distinguebantur echinis. 
totam t destructique sine ulhs ossibus mulU constraverant. 

Q: echini tota districtiq; 

B: echiini totam destructique {ini exp. et nis superscr.) 

A: echinis totam destructique 

Two of the emenders, Beltrami (II) and Aem. Hermes, introduce 
the word turdi in their conjectures, but I think this is excluded by 
the obviously repetitive list below at the end of section 28, ostrea, 
echini, spondyli, mulli, where it should recur had it been used 
previously. Echinis appears to be required to provide the noun for 
intervenientibus. For districti or destructi, read with E. P. Barker^^ 
destricti, "skinned." Que I take to be the residue of <ae>que; the 
aeque means "just hke the shellfish," the boned mullets becoming 
the same sort of food as the boneless shellfish. Cf. below, 11. 7-9: 
quantulo autem hoc minus est, testas excerpere atque ossa et dentium 
opera cocum fungi. This makes the second sentence above read: 
totam, destricti, <ae>que sine uUis ossibus, mulh constraverant; 
translate: "the base for the whole had been provided by skinned 
mullets, as boneless as the boneless shellfish." 

XCV, 29 (445 : 18): quomodo ista (sc. genera cibi) perplexa sunt, 
sic ex istis non singulares morbi nascuntur, sed inexplicabiles, 
diversi, multiformes, adversus quos et medicina armare se coepit 
multis generibus, multis observationibus. 

The last four words are decidedly odd, and have given rise to 
considerable conjecture. The vagueness of multis generibus, the 
reading of all the manuscripts, is the chief difficulty felt, but I 
think with Beltrami in the critical note ad loc. in his second edition 



1^ Dr. B. L. Charney comments: "The combination of the somewhat dif- 
ferent senses of pars (naturae partes, corporis pars) seems to be not the 
happiest of jests." I wish I could conscientiously acquit Seneca of all failures 
of good taste, viewed, that is, from our standpoint. 

18 Op. cit., Vol. II, p. 330. 



182 University of California PiMications in Classical Philology 

that the phrase may reasonably be understood to mean multis 
rationibus medendi. Even so, the interpretation of the last two 
words remains cryptic. Barker translates courageously:^^ ''Medicine, 
too, is beginning to arm herself against them on various hnes, the 
result of various experiences noted," but to me this seems to be 
stretching the original to the breaking point, if not beyond. 

It is perhaps possible that we should read, not ohservationihus, 
but ohversationibus, with the s and the v transposed; the meaning of 
this word would be "oppositions, positions taken against." This fits 
in well with the figure introduced by se armare, and clears up the 
cloudy generibus. The verb obversor is so well attested in the lan- 
guage that a derived noun is fairly to be postulated even if not actu- 
ally quotable by book, chapter, and verse. Seneca appears to be the 
only authority for five other -atio words, according to Summers^ 
list.^" Obversatio in the technical sense of "oppositions" in astronomy 
has been proposed by M. Bonnet in LXXXVIII, 26.2i 

XCV, 46 (450 : 25): hoc in omnibus rebus accidet nobis, <nisi> 
eximuntur quae reprendunt animos et detinent et f preconarique 
totos vetant. 

All the emendations which have come to my attention proceed on 
the assumption that pre- conceals an infinitive parallel with conari, 
for example, Beltrami's par ere conarique and Buecheler's ire con- 
arique, approved by Axelson,^^ who directs attention to the sense 
of motion often apparent in conari, "drive forward." I see no 
reason why it should not conceal a finite verb parallel with re- 
prendunt, detinent, vetant, with -que serving as the coupler between 
the missing third finite verb and vetant. Et in polysyndeton with 
-que for the final link is common enough.^^ 

The word which commends itself to me for the particular context 
is prae<gravant>, "weigh heavily on our souls" ; this fits well with 
conari totos vetant following, because if we are bowed down by any 
kind of weight, we cannot put our whole might into an endeavor.^* 
The only Senecan occurrence I can at present locate {Clem. 1, 24, 1) 

19 Ibid., Vol. II, pp. 148-149. 

20 W. C. Summers, Select Letters of Seneca (London, Macmillan, 1913), 
introd., pp. xlviii-xlix. 

21 Woeimm'sArchivIX (1896), p. 131. 

22 Op. cit., pp. 57-58. , 

23 Thesaurus Linguae Latinae, V, 2, fasc. vi, 884. 

24 Compare Epistle to the Hebrews, 12:1: ". . . let us lay aside every weight 
. . . and let us run with patience the race that is set before us." 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 183 

is rather different in its meaning, but Horace (Satires, 2, 2, 78) uses 
the verb with the object animum in the sense required for our 
passage, and he is there giving a parody of a Stoic sermon. Atten- 
tion is also drawn to degravat (XXX, 1), degravant (LXXIV, 18), 
gravatur (LXXII, 6). 

XCV, 51 (452 : 8) : praecipiemus, ut naufrago manum porrigat, 
erranti viam monstret, ciun esuriente panem suum dividat? 
quando omnia quae praestanda ac vitanda sunt, dicam, cum 
possim breviter banc ilh formulam humani officii tradere etc. 

The latter part of this passage has given rise to much difficulty 
and, in consequence, to numerous emendations. Axelson-^ returns 
(1939) to Hermes' replacement of quando by quare, dropping the 
needless at which Hermes had postulated by a haplography out of 
the final syllable of dividat. 

After a careful study of the several remedies proposed, I have 
come to the conclusion that the passage is probably sound as it 
stands. There should, however, be a sign of interrogation after 
dicam, causing the quando . , . dica?n sentence to stand as a parallel 
with the preceding praecipiemus . . .dividatf "When," writes Seneca, 
"shall I tell all the things that ought to be done or that ought to be 
avoided?" The impHcation is that he could never overtake the 
task.26 The thought transition thereafter is "and why should I, 
since I can supply in brief a formula of the duty of man?" It would 
be possible on lines of normal paleographical practice to postulate a 
<quare dicam> to follow dicam, but I doubt the necessity of it: 
the why is more or less implicit in the when, and cum has the full 
sense of praesertim cum. 

XCV, 57-8 (453 : 19) : non contingit tranquillitas nisi immutabile 
certumque iudicium adeptis: ceteri decidunt subinde et repo- 
nuntur et inter missa adpetitaque alternis fluctuantur. causa his 
quae iactationis est? quod nihil hquet incertissimo regimine 
utentibus, fama. 

B : causarisque, but a recent hand has struck out risque and 
written huius above 

A : causarisque with v superscript on the r 

Q : causa usq; 

q: causavisque as in A corr., and so many of the inferiores 

2^0p. ciL, p. 16, fn. 

2^ Hence Windhaus' explanation that quando . . . dicam means nullo tern-pore 
dicam; op. cit., p. 17, n. 9. 



184 University of California Fuhlications in Classical Philology 

Having regard-to Seneca's fondness for the verb quaeris "do you 
ask?" in introducing questions he himself formulates for the inter- 
locutor, one might on the basis of the ms. tradition conjecture that 
the actual archetype reading here was causam quaeris quaeP The 
first quae became lost in the mental confusion that is always likely 
to attack a scribe when identical forms appear in close proximity to 
one another. Note that the quae est is necessary; Seneca does not 
say: "Do you want a reason for the instability?" but "Do you want 
the (quae est) reason for the instability?" 

XCVII, 10 (462 : 11) : et res ipsa etiam sine duce, sine comite pro- 
cedit. non pronum est tantum ad vitia sed praeceps. 

Where Hense's text above reads non pronum est, the reading of 
Q B A is, with minor and irrelevant variations, non praenuntius. 
Beltrami (II) retains his earher conjecture non pronus nutus, intro- 
ducing it into the text. He collects numerous other conjectures in 
(I), Vol. II, page 1 of the introduction. 

Observing from the first that this was a much discussed and 
emended passage, I decided to form my own conclusions before 
exposing myself to the mass of conjectural emendation, and decided 
on non pronum iter, which I then found to be the reading of Erasmus 
(II), except that he added est; this does not seem obhgatory. I 
imagine that the words non pronum iter were written in highly 
abbreviated form, as would be possible for three such words, and 
emerged in the ms. form non praenuntius. 

For the position of tantum in the combination non pronum iter 
tantum see Beltrami's remarks at the place mentioned above. Such 
hyperbaton is not unusual in the style of Seneca. 

XCIX, 10 (471 : 15): 'sed puer decessit.' nondum dico melius agi 
cum CO, qui cito vita defungitur: ad eum transeamus, qui con- 
senuit. 

Cito was added to the text by Gertz, and has received the ap- 
proval of Hense and Beltrami in their editions ; Rossbach suggested 
ita, and Brakman puer, before vita.Thc intention of all these emenders 
is the same — to bring the unsatisfactory general term cum eo, qui 
vita defungitur to something specifically in agreement with the tenor 

" The suggestion of Dr. B. L. Charney, arising from a discussion of the 
possibilities of the passage. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 185 

of the discussion at this point. Of Rossbach's ita I would say that, 
while paleographically neat, it is a feeble recapitulation of the point 
pressed by the use of puer in the objection sed puer decessit. Gertz's 
cito avoids that charge, and may be thought to find some confirma- 
tion in the occurrence of that adverb in the first sentence of para- 
graph 12; paleographically it is only tolerable. 

Brakman's puer, read before vita, suggests no paleographic justi- 
fication, but is in my judgment a natural word for the passage, (1) 
as repeating identically the gravamen of Marullus' complaint, and 
(2) as giving the precise thought balance to sum qui consenuit. It 
might better be read after defungitur, where it may have been 
omitted through the scribe's returning to his exemplar to resume 
his copying after the wrong r. Placed there it gives a verbal balance 
to consenuit of the next sentence. Metrically fungitur puer may be 
regarded as forming a trochee-cretic combination, or a hypodoch- 
mius according to Axelson's fourth terminal scheme. 

XCIX, 20 (474 : 3) : non est itaque, quod lacrimas propter f cir- 
cumstantem adsidentemque aut contineas aut exprimas: nee 
cessant nee fluunt umquam tarn turpiter quam finguntur. 

We may first dispense with the obelus conformably to Hense's 
latest opinion-^ on the soundness of the present participle singular 
in a generic sense, and circumstantem need not be too literally 
applied and thus reduced to an absurdity. 

The exegesis of the nee cessant sentence has presented such diffi- 
culties that such capable scholars as Baehrens have proposed quam 
<si> and O. Rossbach quam <quom> (after Pincianus' quam 
<cum>). Both of these proposals are rejected by Hense and 
Beltrami on the basis of Ruhkopf 's note explaining that, of all tears, 
feigned tears are the most shocking, more so than any mere stop- 
ping or starting of tears. That may be so, but I do not think that 
Ruhkopf or those who accept his explanation have grasped the 
meaning of the passage, partly because of an undoubted error in the 
tradition which heretofore has been overlooked. 

The preceding paragraph (19) tells how tears get to flowing, some 
in spite of ourselves by sheer pressure of a nervous shock upon the 
secretions adjacent to the eye {illis vincimur), others as a sort of 
nervous relaxation (his indulgemus). Both of these cease presum- 
es Hermes, LXII (1927), p. 111. 



186 University of California Publications in Classical Philology 

ably when the psychological stimulus ultimately fails. Thus the 
natural lacrimae fluunt and lacrimae cessant are explained. 

But there is a less natural occurrence of flow and check "because 
of someone who is standing or sitting by you" ; here flow and check 
are made to order, so to speak, and are devoid of the natural causes 
detailed in paragraph 19. This is never justified; all the justifiable 
cases have already been covered. Read therefore utique for itaque; 
paragraph 20 is not a deduction (itaque) from paragraph 19, but a 
complete exclusion (utique) from the permissible tears mentioned in 
19 of any provoked by mere accidental contacts in society. Trans- 
late: "in any case you may not check or release your tears because 
of some bystander or someone sitting near you." Why? The answer 
is: "Tears never cease, and never flow so shockingly as when they 
are feigned." The quam . . . finguntur is a specific limitation of what 
precedes it, and not the introduction, as Ruhkopf would have it, of 
yet another sort of tears, and to make this specific Umitation effec- 
tive we must have either cum (quom) or si before finguntur. Prob- 
ably the latter is paleographically the easier. This whole contention 
is confirmed by the following eant sua sponte. "The stops and starts 
of tears are never so shocking as when they are feigned; let them 
roll spontaneously." Accordingly, read utique for itaque and si he- 
iore finguntur P 

XCIX, 24 (475 : 4) : obUvisci quidem suorum ac memoriam cum 
corporibus efferre et effusissime flere, meminisse parcissime in- 
humani animi est. sic aves, sic ferae suos dihgunt, quarum 
[contria] concitatus [actus] est amor et paene rabidus, sed cum 
amissis totum extinguitur. 

The B A reading following quarum is as given above if all the 
words are retained. The Q reading is contrarius acconcitatus. The 
recentiores, including q and Harleianus 2659, have only concitatus. 
Of the several emendations proposed, Buecheler's quarum antea 
concitatus [actus] appears to me the best. I am not at all sure, how- 
ever, that the text of B A, with the change of contria to contra, is not 
sound. I think Seneca's words here are contra concitatus actus, "act 
stimulated in return," or "act reciprocally stimulated," and that 

29 But Dr. B. L. Charney may be right in asserting that the nee cessant . . . 
finguntur means: "the disgrace lies not so much in checking your tears or 
letting them flow as in feigning them," though to my way of thmking that is 
too general a sentiment to fit in satisfactorily here. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 187 

he is stating that the affection of birds or wild beasts for their off- 
spring is a contra concitatus actus in the sense just explained. By 
that I understand that it is stimulated in the parents in direct pro- 
portion to the affection displayed by their young, and rises almost 
to a frenzy as the young increasingly make augmented demonstra- 
tions. This view, I think, is substantiated by the phrase immedi- 
ately following, "when the young are gone it disappears entirely." 
The reason is that it is a "reciprocally stimulated action" and when 
one of the reciprocal factors is eHminated the whole process ceases 
quite suddenly and abruptly. On concitatus used of animal impulses 
and passions cf. CXXIV, §18. 

XCIX, 25 (475 : 10) : illud nullo modo probo, quod ait Metro- 
dorus: esse ahquam cognatam tristitiae voluptatem, hanc esse 
captandam in eiusmodi tempore. 

B A show hanc ipse, Q hanc ipsam (which is the reading of Eras- 
mus in his second edition) , the inferiores hanc esse ; finally Windhaus, 
gathering everything up, hanc ipsam esse. 

In the light of our principal manuscripts one should certainly 
read with Q hanc ipsam, and I confess to surprise at Beltrami's 
desertion here of his favorite Q. The whole sense thus created is, in 
my judgment, admirable, with ipsam contributing precisely the 
right emphasis: "there is a certain pleasure closely related to pain; 
this is the very thing to be grasped under such circumstances." 
Unfortunately, while the Latin here is obviously a close translation 
of the Greek following, the Greek is in such bad shape that no cer- 
tain argument can be based upon it; but it is equally true that no 
argument can be based on it against hanc ipsam either. A repeated 
esse, after our quotation opening with esse, strikes me as very otiose, 
even if this opening esse has a substantive value which is much 
more than esse connective, and in any event there is no esse in the 
best manuscript tradition, nor does syntax require that there should 
be any. 

C, 8 (479 : 12) : humilia praeterea tibi videri dicis omnia et parum 
erecta; quo vitio carere eum iudico. non sunt enim <humiha ilia 
sed placida et ad animi> tenorem quietum compositumque 
formata, nee depressa sed plana, deest iUis oratorius vigor 
stimuhque, quos quaeris, et subiti ictus sen<ten>tiarum. sed 
totum corpus, videris quam sit comptum, honestum est. non 
habet oratio eius, sed debet dignitatem. 



188 University of California Publications in Classical PMlology 

In the concluding sentence Hense^° finally accepted dahit for 
debet; this suggestion is due to Lipsius, and received Madvig's 
approval. This must mean: "his style does not possess grandeur, 
but it will confer it." Beltrami, curiously enough, still adheres in his 
second edition to Hense's original explanation of debet dignitatem 
which he had accepted in his first, though it is obviously wrong and 
Hense himself so admitted in abandoning it. 

This short sentence is to be interpreted, I think, as an objection 
raised by Lucilius himself. In the Latin passage quoted above there 
is one objection registered in the words humilia . . . et parum erecta. 
There is a second indicated by deest illis . . . sen<ten>tiarum. 
Each of these is answered immediately after being mentioned. In 
the final sentence of the quotation there is put forward a third ob- 
jection, namely, "his style does not possess grandeur, but it ought 
(to possess it)." The reply, as in the previous two instances, is made 
without delay, but it is a more ambitious and comprehensive reply 
than either of the others; it vindicates the dignitas of the style of 
Fabianus by placing him only just below three masters, Cicero, 
Asinius Pollio, and Livy. 

That there was something in this third objection requiring vigor- 
ous defense of Fabianus' style may be deduced from section 5 pre- 
ceding, where Seneca speaks of the words employed by Fabianus as 
being electa and splendida, but adds the quaHfication quamvis 
sumantur e medio. 

CII, 12-13 (488: 11^89: 4) : 'quid ergo'? . . . dissident. 

This appears to be one of those passages in which Seneca has 
written what was in his mind in a sort of literary shorthand. The 
translators reflect in their versions the difficulty that arises from 
such a procedure. Apparently the embarrassment is occasionally 
increased by the irresistible tendency to run off into the temporarily 
attractive smart phrase, as, for example (488 :25) : non est unius 
una sententia. Despite the doubts of O. Rossbach-' and C. Busche,^- 
I believe that the text of that sentence is sound. Seneca has just 
said : "Do you think that the opinion of people in the mass can be a 
unity?" Then, thinking of the ordinary individual, he throws in the 
remark: "There's no such thing as a unity of opinion in the single 

30 Rhein. Mus., LXXIV (1925), p. 126. 

31 Breslauer philologische Abhandlungen, Vol. II, Pt. 3, p. 158. 
" Phil. Wochenschrift, XLV (1925), p. 572. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 189 

individual." No doubt he means "in the single individual who is a 
unit in the great mass," but unfortunately he has just a short while 
previously been speaking of the absolutely dependable point of 
view of a single individual who is also a good man, and has probably 
overlooked the chance of resultant confusion. 

Nor has he made his references in illic (488:21) and hie (488:23) 
any too specific. I beheve that in order to be clear on these one must 
go back to 488 :14 : diversa horum condicio est at illius, where horum 
refers to the muUorum of the preceding sentence, and illius to unius 
/lomms of 488:11. 1 would read the sentence (488:20) adgloriam. . . 
opinio as a question in which Seneca repeats interrogatively the 
statement of his opponents. The answer which follows I interpret 
thus: "In the one instance, namely, the case of the good man, a 
single opinion has the same value as that of all good men, because, 
if the question be put, the opinion of all good men will be a unity; in 
the other instance, namely, that of the mass of men, the judgments 
vary as the individuals are unlike." 

In the sentence which begins with the first line of page 489 it 
would certainly be a technical improvement to read with Rossbach 
illic, which would have exactly the same meaning as the illic (488 :21) 
discussed above, and it would find its contrast here in the apud hos 
phrase just as the illic of 488:21 finds its corresponding member in 
hie (488 :23), and I would myself feel free to make this change in the 
text. 

CII, 17 (489:24): fama vocem utique desiderat, claritas potest 
etiam citra vocem contingere contenta iudicio. 

B A : desiderat claritas potest enim 
Q: desiderat claritas. Potest enim 
q : desiderat no claritas. potest enim 

It seems pretty clear that there is a definite break in the sense 
after claritas, and that q inserts non in order that, with this break 
maintained, the proper sense may be secured. While it is true that 
non is frequently omitted or inserted in the manuscripts for no 
particular reason that can be discerned, I prefer to read with Wind- 
haus^^ desiderat, <non desiderat> claritas, for the reason (1) that 
it explains neatly and simply the source of the confusion here, and 
(2) that it gives us a double cretic clausula. The enim is clarified 

" Op. cit., p. 24. 



190 University of California Publications in Classical Philology 

and thus defended as the genuine ms. tradition even by q's insertion 
of non, but equally so by <non desiderat> , which has the other 
factors in its favor as well. 

CII, 28 (492 : 29) : huic nunc quoque tu, quantum potes, subdue te 
voluptatique, nisi quae necessariis sermque cohaerebit, ahenus 
iam hinc altius ahquid sublimiusque meditare. 

Q B A agree on suhvoluptariquae nisi quae necessariisque cohaerebit. 
Voluptatique is an immediate and certain correction. For the missing 
verb of which sub is the prefix it must be admitted that Seneca uses 
the word subducere frequently enough. Von Jan's subvola is an un- 
usual word, and if it is adopted, Beltrami's hinc for huic at the 
beginning of the passage should be read. In favor of subvola it re- 
mains to be pointed out that it leads to a double cretic clausula, and 
that the word altius occurs in the latter part of the sentence. 

As for the nisi clause I again follow von Jan in rejecting quae 
after nisi. I presume it to have been engendered between an er- 
roneous quae for que (voluptarique) and an erroneous que for quae 
following necessariis in the now thoroughly confused mind of the 
copyist. Further, quae cohaerebit estabhshes a good clausula of the 
cretic spondaic type. 

CII, 30 (493 : 18): quidni non timeat qui mori sperat? is quoque, 
qui animum tamdiu iudicat manere, quamdiu retinetur corporis 
vinculo, solutum statim spargit, ut etiam post mortem utihs esse 
possit. 

Axelson^^ has handled this passage effectively; I concur with his 
endorsement of Buecheler's is quoque (se quoque B A, sed quoq; Q), 
his rejection of Beltrami's si et is quoque, and his acceptance of the 
same editor's spargi, id agit (with q for spargi and Q for id agit). 

It seems to be the case, as Axelson says, that practically every 
commentator supposes that the mention of the nonbeliever in the 
immortality of the soul and of the view that such a nonbeliever takes 
of his duty to posterity in spite of his theologj-, has reference to the 
quidni non timeat. Actually, what happens is that Seneca concludes 
his description of the glory^ waiting to be revealed with the trium- 
phant rhetorical question: "Why should he fear who in death has 
hopef"^^ It is particularly important to get the meaning of the last 

34 Op. cit., pp. 120-121. 

" Compare CXX, 15, si exire non metuit. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 191 

two words correctly; much time has been wasted in trying to 
discover instances of spero with the present infinitive where the 
significance is future. Hence Madvig's aberration qui morti <se 
superfuturuin> sperat. 

Once the idea "who in death has hope" has been mentioned, that 
promptly suggests to Seneca the idea of the nonbeliever in the im- 
mortality of the soul and of his attitudes toward a future in which 
he expects no share. This idea is not necessary to the argument. It is 
not connected logically with quidni non timeat, and is generated 
purely as a side issue by the use of the phrase mori sperat. For all 
logical purposes the letter might just as well have ended with those 
two words. But the hope of immortality in the sense of the survival 
of the personality was not a Roman hope, certainly not in the great 
days of the Republic, and it may possibly have occurred to Seneca 
to terminate his letter, even at the risk of an antichmax, on a note 
which would glorify many distinguished Romans of the past, even 
if no specific reference is made to them. 

CIV, 6 (497 : 7) : ut primum gravitatem urbis excessi et ilium 
odorem cuhnarum fumantium, quae motae quicquid pestiferi 
vaporis f obruent, cum pulvere effundunt, etc. 

The general drift of the words from et ilium on I take to be as 
follows: "that reek of smoking portable kitchens^*^ which, when set 
in motion, pour forth, along with the dust" they raise, whatever 
vile smell they contain." But of course the point is that when they 
are set in motion, that is, by their proprietors' dragging them to 
another stand, they fail any longer to contain the smell, which now 
drifts all over the place. Obviously therefore, since Seneca means 
"whatever vile smell they (ordinarily) contain," the tense will be 
the present, namely, obruuntf^ nor can I feel that there is any seri- 
ous objection to the verb ohruunt if we remember that Seneca hates 
these kitchens and is not Hkely to select a pleasant word to describe 
their functions. A common sense of ohruo is "bury," and the closed 
stove may, at least as an acid joke, be regarded as the cofl&n of the 
pestiferous smells it contains. The clausula of the hypodochmiac 

2^ For portable kitchens compare LXXVIII, 23. 

" Very possibly the pulvis is the soot and smoke from the chimneys, but 
there is no need to change pulvere to fuligine with H. Georgii, Philologus, 
LXXXIV (1928-29), p. 95. 

28 Thus q, and the oldest editors. 



192 University of California Publications in Classical Philology 

type was already correct with obruent, and that in itself is some 
evidence that the simple change to obruunt is all that is required to 
rectify the passage if other conditions are fulfilled. 

CIV, 20 (501 : 3) : tecum sunt quae fugis. te igitur emenda, onera 
tibi detrahe et demenda desideria intra salutarem modum contine. 

Q B A : et emenda where the passage above reads et demenda. 
Vivona^^ points out that such emendations as Hense's demenda and 
Beltrami's eluenda go much too far in regard to meaning, as Seneca 
has not been speaking of abolishing desires, but rather of setting 
bounds to them. He suggests immodica, which is altogether excellent, 
except paleographically. 

I propose detrahe, timenda. This became in the course of trans- 
mission detrahe <e>t imenda, and the last word was subsequently 
"improved" to emenda; to this process the existence in the sentence 
of emenda already, though from an entirely different verb, may 
have contributed. Desires are to be dreaded (timenda) if they get 
away from us; hence the importance of intra salutarem modum 
contine. This, of course, does not suggest their ehmination, but only 
their control. Haase's salutarem may be regarded as a certain con- 
jecture for salutem. 

CIV, 29 (504 : 2) : et hunc licet dicas non minus quam Socraten f 
inservisse dixisse, nisi forte Cn. Pompeimn et Caesarem et 
Crassum putas libertatis socios fuisse. nemo mutatum Catonem 
totiens mutata re pubhca vidit ; eundem se in omni statu praestitit 
etc. 

Q B A : inservisse dixisse. I agree with Rossbach^^ on the correc- 
tion in servis vixisse. The Latinity of in servis as the equivalent of 
inter servos is sound; cf. T.L.L., VII, 1, col. 776, C, under the 
caption "fere i.q. inter." I should, however, accepting this correc- 
tion, regard everything as far as vidit as constituting a single 
thought unit, and hence would remove the full stop aher fuisse and 
replace it with a comma. The hunc licet dicas sentence is a clause 
subordinate to nemo . . . vidit, with the nisi clause thrown in as an 
ironical parenthesis. Translate: "and though you may say that he 
no less than Socrates lived among slaves — unless by any chance 
you think that Pompey, Caesar, and Crassus were all in partnership 

39 P'rancesco Vivona, Note criliche alle Episiole di Seneca (Rome, Casa 
Editrice "Ausonia," 1932), p. 43. 
^» Phil. Wochenschrift, XXXIV (1914), pp. 496 ff. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 193 

with liberty, — (yet) no one ever saw Cato changed, though the form 
of government changed so often." Slaves, or people of servile dispo- 
sition, readily change their attitude with the imposition of this or 
that form of government upon them; Cato, though he lived among 
such people, virtually slaves under the First Triumvirate, never 
altered his views or attitude, any more than Socrates in like cir- 
cumstances under the Thirty. 

CV, 3 (506 : 9) : spem improborum vitabis, si nihil habueris, quod 
cupiditatem ahenam et improbam inritet, si nihil insigne pos- 
sederis. concupiscuntur enim etiam parva, si f nnotarum sunt, f 
sic raro. 

B A : etiam (etam A) pars innotarum sunt 
Q : etiam si parum nota sunt. Sic raro 

Beltrami has, as usual, assembled the numerous conjectures (II, 
app. crit. ad loc). That of Barker,''^ however, has eluded him; this 
is unfortunate, as I feel that Barker has the right idea in regarding 
notarum as being a noun in the sense "brands, types, sorts," a value 
for nota of which Seneca presents several examples.'*^ This is 
apparently a colloquial use and not common in the Latin liter- 
ature available to us; hence ingeniously Barker suggests that sic 
raro is a scholarly commentator's note on the unusual sense of nota. 

From the B A reading Barker deduces the following: si -r 
notarum sunt, that is, si primarum notarum sunt, and it only re- 
mains to be pointed out that the parum of Q may very well be a 
damaged survival of primarum.^^ 

CV, 6 (507 : 14) : habet unusquisque aliquem, cui tantum credat, 
quantum ipsi creditum est. ut garrulitatem suam custodiat et 
contentus sit unius auribus, populum faciet, si quod modo secre- 
tum erat, rumor est. 

Axelson in his Senecastudien'^^ proposed to read propalam for 
populum, but in his Neue Senecastudien,*^ properly impressed with 

" Op. cit., Vol. II, p. 332. 

42 Epist. XV, 3; De Benef. 3, 9, 1; N.Q. 2, 2, 4. 

43 Compare the reading of B at CVIII, 22 (520:3): in I- hherii principatum, 
and of A at the same place: in -v tiberii principatum, also of Q: in tiberii, with 
no sign of anything between in and tiberii, which is also the vulgate reading. 
But Buecheler discerned the truth of the matter, namely, that I or i repre- 
sented the ordinal primuni. 

" B. Axelson, Senecastudien (Lund, Hakan Ohlsson, 1933), p. 111. 
« P. 16, fn. 



194 University of California Publications in Classical Philology 

the indisputable opposition of unius and populum, he has aban- 
doned his earlier proposal. He expresses himself as still unable to 
see the meaning of the passage as it stands. I do not think that he 
helps matters by preferring sic to si. 

Everything seems to me to turn on the word faciei. Barker^^ 
translates the main clause of the sentence thus: "he'll create an 
army of tattlers when once what was a secret becomes a rumor," 
This would be satisfactory if it read "he'll find he has created." 
and I believe that the faciei will carry that value as indicating the 
end result in the future of an action which is going on continu- 
ously in the present. He is making the army of tattlers every day 
even though he hmits himself to one confidant, and at some time 
in the future he will find he has made such an army. 

CVII, 1 (510 : 20) : si amici deciperent — habeant enim . . . turpius 
non sint — omnibus rebus tuis . . . nunc desunt illi, qui et operam 
tuam conterebant et te ahis molestum esse credebant. 

Since my observations on this passage"*^ appear to have been mis- 
understood by the reviewer in the Classical Review,'^^ I take this 
opportunity of restating them, perhaps with greater clarity. 

I read with Pincianus, also with the editio Curionis and the 
editio Gothofredi according to Fickert's evidence, deceperunt. If 
once this form were written as decepermt, it would be no novelty 
of paleography for it to pass into deciperent. 

I understand amici as referring — ironically, of course, as the 
following parenthesis shows — to the slaves who have absconded; 
we have exactly the saine biting usage of the word "friend" in 
English. Despite what seems to the reviewer an absurditj^ I still 
think the reference to XLVII, 1, much to the point; is it not very 
likely that, in the communication which elicited letter CVH as 
response, Lucihus spoke bitterly of Seneca's previous advice to 
make friends out of one's slaves? In any event the servi me re- 
liquerunt of section 5 seem to me practically decisive on the issue; 
those words are Lucilius' complaint, since they are absolutely an 
echo of servi . . . putaverunt in section 1 , which is immediately fol- 
lowed by si amici etc. 

Madvig's emendation errori for Epicurus is masterly.^' The 

■•« Op. cit., Vol. II, p. 214. "' Op. cit., p. 13. 

^s Classical Review, XLVII (1933), p. 78. 

^9 J. N. Madvig, Adversaria (Copenhagen, 1873), Vol. II, pp. 506 ff. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 195 

course of degeneration is thus : errori becomes eprori, eprori becomes 
in one line of tradition (Q and B p.m.) epriori and in the other 
epicuri, subsequently altered to epicurus (A and B s. m.) under the 
influence of the adjacent noster. 

I treat desunt illi (no nunc is required) as the apodosis to de- 
ceperunt. I regard illi as being a brachylogy for illi tantum; this is a 
common occurrence in the Epistles. 

CVIII, 12 (517 : 2) : cum haec atque eius modi audimus, ad confes- 
sionem veritatis adducimur. illi enim, quibus nihil satis est, 
admirantur, adclamant, odium pecuniae indicunt. 

Difficulty has been felt in several quarters in modern times over 
establishing the proper sense connection and logical relationship 
between these two sentences. Hence Windhaus and Gertz with 
etiam for enim, accepted by Hense (II) unless, as he says, the solu- 
tion is a vel or et before illi. Beltrami (II) writes vel into his text. 

It is unnecessary to seek to attain the proper emphasis by adding 
to the text or altering it. The emphasis is adequately given by the 
relative clause quibus nihil satis est with the voice rising to the peak 
on the words nihil satis: ''for those for whom nothing is ever enough, 
admire, shout approval, and declare war on money." They are 
made the extreme case by the description of them in the relative 
clause; hence nothing to correspond with the English "even" (also 
a device to mark the extreme case) is required. The economy of 
Latin in these things tends constantly to be overlooked. 

CVIII, 15 (518 : 2) : inde mihi quaedam permansere, Lucili. magno 
enim in omnia inceptu veneram. deinde ad civitatis vitam re- 
ductus ex bene coeptis pauca servavi. 

Inde means "from the moral diatribes of Attains." Inceptu is the 
Q B A reading; a few of the inferiores and the majority of the older 
editors in a line reaching from Erasmus to Ruhkopf, inclusively, 
read impetu. This has recently been endorsed by Loefstedt,^'' in- 
fluenced by the double cretic clausular rhythm thus produced and 
by the impetu of section 17. However, there is nothing objection- 
able in the spondee-cretic (or molossus-cretic) clausula of the ms. 
text, and the harder reading inceptu should be retained if it can 
be shown to have a sound meaning. 

5» Einar Loefstedt, "Zu Senecas Briefen," Eranos, XIV (1915), p. 163. 



196 University of California Fuhlications in Classical Philology 

The first sentence says: "some things stuck (sc. and are still 
operative)." The next sentence purports to give the reason: "for 
I had entered on all of them with a tremendous initial effort." The 
comparison between the end result and the enthusiasm of the start 
is thus admirably established, and it is also made clear that the end 
result is as important as it is because the initial impulse was so 
earnest and all-embracing. Further, instead of going as far away as 
section 17 to find a refutation of inceptu by the occurrence there of 
impetu, I should prefer to point to coeptis in the immediate context 
as strongly confirmatory of inceptu. ^^ 

CVIII, 32 (523 : 3) : eosdem libros cum grammaticus explicuit, 
primum verba expressa, reapse dici a Cicerone, id est re ipsa, in 
commentarium refert, nee minus sepse, id est se ipse. 

Exprimere is used of words by Cicero in De Oratore, III, 11, 41 : 
nolo exprimi htteras putidius, nolo obscurari neglegentius, which 
another passage matches closely, namely, De Officiis, I, 37, 133: 
Sonus (sc. vocum Catulorum) erat dulcis, litterae neque expressae 
neque oppressae ne aut obscurum esset aut putidius. It is clear that 
in these places exprimere means to pronounce very precisely, so 
much so that it can very easily be regarded as an affectation. 

In Quintihan, however, I, 11, 4, expressa is a term of praise, 
"clearly pronounced"; imprimis vitia si qua sunt oris emendet, ut 
expressa sint verba, ut suis quaeque litterae sonis enuntientur. I 
think we may safely assume that the phrase expressa verba meant 
to Seneca about what it meant to Quintihan, his contemporary. 
What the grammaticus of the passage therefore records about forms 
like reapse (probably enough to be thought of as dissjdlabic, as in 
Plautus)^^ and sepse is that they are not clearly enunciated, and the 
text consequently requires <n> between verba and expressa. 
Windhaus had a like idea in mind when he suggested verba com- 
pressa,^^ but the paleographic justification is wanting.*^ 

^^ Magno inceptu, if one may go far afield into the context for confirmation, 
seems to me to find a rather strong one in the vehementes irtjpetus primos of §23. 

^^ Compare also what Cicero himself has to say on the point, Orator, 44, 150. 

^^Op. cit.,p.3\. 

^^ Dr. B. L. Charney thinks that verba expressa may mean simply "the 
pronunciation of words" without expressa's having the force of "clearly pro- 
nounced." He points out that in the instances cited from Cicero and Quin- 
tilian there is a secondary exegesis of exprimi, expressae, expressa which 
introduces the idea of "clearly," while in the Senecan passage there is not. 
I am inclined however to think that the very mention of the slurred forms 
reapse, sepse, constitutes the secondary exegesis here. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 197 

CVIII, 33 (523 : 10) : deinde E^nnianos colligit versus et in primis 
illos de Africano scriptos: *cui nemo civis neque hostis quibit pro 
factis reddere opis pretium.' ex eo se ait intellegere, opem aput 
antiques non tantum auxilium significasse, sed operam. ait 
[opera] enim Ennius neminem potuisse Scipioni neque civem 
neque hostem reddere operae pretium. 

The conclusion seems inescapable that opem has been lost from 
the second sentence; the position assigned it by Th. Korsch^^ is 
perhaps as good as any. 

As to what follows, though no certainty is possible in a ms. text 
so faulty, I hazard the conjecture that the ait operaenim ineius of 
Q B A really contains the words ait enim ope Ennius with the mean- 
ing as follows: "For by ops Ennius means that no one, whether 
citizen or foeman, could have given Scipio for his deeds anything 
worth bothering about" — the citizen by way of reward, the foeman 
by way of resistance or revenge. The use of the ablative ope I take 
to be a brachylogy for the idea "by employing the word ops." I 
hesitate to reject the opera with the modern editors on the ground 
of mere stupid repetition if any even partly plausible explanation 
of its presence in the ms. text can be offered. 

CIX, 6 (526 : 5) : proderit sapienti sapiens, non scilicet tantum suis 
viribus, sed ipsius quem adiuvat. potest quidem ille etiam reUctus 
sibi exphcare partes suas. nihilominus adiuvat etiam currentem 
hortator. 

In Q and many inferiores there occur immediately after suas the 
words utetur propria velocitate sed. Axelson,^*' while feehng that the 
transition from the quidem in potest quidem ille to nihilominus is 
most direct and effective in the B A text, discounts the likehhood 
that utetur propria velocitate sed is an interpolation, on the single 
ground that an interpolator would hardly have run off into the fu- 
ture utetur with a potest on the one side and an adiuvat on the other. 

Can the four words in question be reasonably explained in the 
context? I am assuming that they will be introduced into it as 
follows: a new sentence after suas begins with utetur propria velo- 
citate; after this a semicolon, while sed belongs with what follows. 
Let us now proceed by the method of translation, with the addition 
of a running commentary. "One wise man will be of service to an- 

5^ According to Hense (II), app. crit. ad loc. I do not know the name other- 
wise. 
^^ Neue Senecastudien, pp. 118-119, with footnote on p. 118. 



198 University of California Publications in Classical Philology 

other wise man, not merely, you may be sure, through his own 
powers alone, but also the powers of the one he is aiding. Wise man 
B can, to be sure, even if left to himself, develop his ow^n role." Of 
course, in that case all the conditions of the development, including 
its speed, will be determined by B alone, who sets things going 
independently. "He will employ his own speed," not being stimu- 
lated by any challenge involved in the speed of any other wise man, 
since he is now ex hypothesi in touch with none; "but none the less 
a backer^^ helps even a person who is already running."*'' The idea is 
that a footrace may already be under way when a certain spectator 
arrives, but if that spectator promptly becomes a backer, he helps, 
even though he was not there to provide any initial support. Even 
so the wise man B if left alone can develop his own role and will 
employ his own speed; yet even at that point if another wise man, 
say A, comes on the scene with encouragement, he is a help. 

CIX, 17-18 (529 : 16) : nondum exerceri vacat ; adhuc medico mihi 
opus est. quid me poscis scientiam inutilem? magna promisisti; 
exige, vide. 

I am no better able after nine years to understand exige, vide 
than I professed myself in my previous notes on the Epistles.^^ 
Since then. Barker's translation has appeared, but his "Search me 
and see!" conveys nothing more to me than Gummere's "Test me, 
watch me."*^° Beltrami (II) adopts for his text the corrections in B, 
exigua video; the testimony of Q unfortunately fails here through 
the loss of the first leaf of the twentieth quire of that manuscript. 

I formerly proposed exigi vide, "notice that they are being de- 
manded," and this is rhythmically sound as a hypodochmius 
(Axelson CI. 4). So far as the meaning is concerned, the correction 
seems to make the words fit in with what has immediately pre- 
ceded. "You have made great promises to me in the past; observe 
that the fulfillment of these great promises is now demanded." Exigo 
is an exceedingly common word in Latin to describe the collecting 
of all sorts of business obligations that have fallen due. 

" Reading, as in Hense's (II) text, hortator with Schweighaeuser. Hortatur 
is an error which arose from taking et to mean "and." 

** Reading et iam with C. Watzinger for etiam. Hense (II) goes the whole 
way with Watzinger in his Addenda et Corrigenda, pp. 633-634, including a 
return to hortatur. The difficulties inherent in this are pointed out by Axelson, 
op. cil., p. 118, n. 4. 

" Op. cit., p. 13. 

8" Loeb Classical Library edition of the Epistles, Vol. Ill, p. 263. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 199 

CXIII, 8 (540 : 16) : si iustitia animal est, si fortitude, si ceterae 
virtutes, utrum desinunt esse animalia, subinde autem rursus 
incipiunt, an semper sunt? 

The si before iustitia is another vetus lectio of Pincianus, the si 
before fortitudo was supplied by Muretus, and the third si is a cor- 
rection made by Muretus out of sic. 

I regard none of these things as necessary; I would read: iustitia 
animal est? fortitudo? sic ceterae virtutes? utrum . . . sunt? This is 
highly dramatic. Three questions are put rapidly with the minimum 
of words; the implied answer to each is "Yes." Then comes the 
clinching question utrum . . . sunt, based on the preceding questions 
and the implied affirmative answers and introducing a dilemma for 
the answerer. It is not necessary to add to the manuscript text in 
any way, or to subtract from it even one letter. 

CXIII, 10 (541 : 4) : quid est, quo colligas iustitiam animal esse? 
'agit' inquit 'aliquid et prodest. quod autem agit et prodest, 
impetum habet: <quod autem impetum habet>, animal est.' 
vermn est, si suum impetum habet; <suum autem non habet >, 
sed animi. 

Both the inserted clauses are missing in QB A; furthermore, Q 

lacks the habet following the words si suum impetum. 

Hense thought the first of these additions superfluous, having 
regard to the development of the argument in section 2 preceding; 
I concur with him in this view. It is not necessary that every step in 
a series should be repeated when it is perfectly obvious, as a result 
of a previous statement, what step it is that is being assumed. 

As for the second case, I think the reading of Q sufficient, even if 
much abbreviated; indeed the condensation is very effective when 
we remember that we are here dealing with the reply, all the ele- 
ments necessary for explaining which are already to be found in the 
preceding statement of claim. I would print the reply thus: verum 
est si suum impetum ; sed animi. 

Compare on the brachylogy section 18 following, where the very 
considerable complement requisite to explain tunc demum sedeo is 
inferred from a preceding sentence. 

CXIII, 11 (541 :8): omne animal, donee moriatur, id est, quod 
coepit: homo, donee moriatur, homo est, equus <equus, canis> 
canis. transire in aliud non potest. 



200 University of California Publications in Classical Philology 

There seems to be no valid reason for repeating equus canis; they 
appear but once in Q B A. If the words occurred doubled in Q but 
not in B A, that would, of course, be another matter. The words in 
question really belong to what follows, and hence do not have to be 
adapted to what precedes. Read: homo, donee moriatur, homo est; 
equus, canis transire in aliud non potest. "A man's a man till he 
dies" is informally paralleled by "a horse, a dog, cannot pass over 
into something else." 

CXIII, 20 (543 : 17) : eo usque res exegit, ut risum tenere non 
possis : prudenter tacere bonum est, < . . . cenare bonum est > : 
ita et tacere et cenare animal est. 

From the mention of cenare in the conclusion, it would seem 
certain that it must have appeared in the premise; hence the in- 
sertion of cenare bonum est, with difference of opinion concerning 
the adverb to be attached to cenare. Hense suggests frugaliter be- 
fore cenare, Beltrami bene before bonum.^^ 

I think it very likely that the sentence read originally prudenter 
tacere, cenare, bonum est, with the adverb applying to both; I am 
confirmed in this view by the fact that in the conclusion tacere and 
cenare are mentioned together without any adverb, which certainly 
seems to indicate that one adverb preceding modifies both. Pru- 
denter cenare is quite as reasonable a conception and expression as 
prudens ambulatio at the end of section 22 following (545:5). 

CXIII, 30 (546 : 7) : o quam magnis homines tenentur erroribus . . . 
ignari, quod sit illud'ingens parque dis regnum. imperare sibi 
maximum imperium est: doceat me etc. 

Where the text above shows parque dis the ms. testimony is B : 
parqueris, A: parque ris, Q: perqueris, altered by a late hand to 
perquiris (the per in both cases being expressed by abbreviation). 
Obviously the reading which both the Q and B tradition found in 
their archetype was parqueris, that is, par quaeris. 

Now to reconstruct what we may call our urarchetype we have 
these items to work on : (1) the parque dis of the inferiores, which is 
undoubtedly a sound phrase parallchng ingens and giving it quality; 

" Hense (II), Addenda et Corrigenda, p. 634, doubts whether bene with 
cenare gives the parallelism apparently required to go with prudenter tacere, 
suspecting that bene cenare means to dine well in anything but an ascetic 
sense. I do not think that Beltrami successfully meets this objection. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 201 

(2) the par queris of the ms. tradition; (3) the circumstance that the 
words illud ingens parque dis regnum may be regarded as not self- 
explanatory, and therefore capable of producing a question to be 
answered in which queris, that is, quaeris, plays a part ; (4) that this 
question, in order to explain the ms. tradition, must be in the order 
quod, quaeris, regnum? and (5) that our urarchetype as recon- 
structed must show probable cause of error. 

From all this we are in a position to suggest the following : quod 
sit illud ingens par < que dis regnum quod > queris regnum imperare 
sihi maximum imperium est. The scribe's eye on returning to his 
exemplar after par was written, caught the que of queris (not of 
que dis), being assisted in that error by the very great similarity of 
quedisregnum and querisregnum. This breaks up thus : quod sit illud 
ingens parque dis regnum. quod, quaeris, regnum? imperare sihi 
maximum imperium est, after which I should certainly place a full 
stop, since the train of thought there concludes. The translation is: 
"what is that great sovereignty which matches that of the gods. 
What sovereignty, you ask? Sovereignty over oneself is the greatest 
of all sovereignties." 

CXIV, 6 (549 : 6) : non statim cum haec legeris, hoc tibi occurret 
. . . hunc esse, qui <in> tribunali, in rostris, in omni pubhco 
coetu sic apparuit, ut palHo velaretur caput exclusis utrimque 
auribus, non ahter quam in mimo fugitivi divitis solent? 

The text is well enough agreed upon by Q B A, divites fugitivi Q, 
fugitivi divites B, fugitivi diutes (or duites) A, and the correction of 
divites to divitis was made by Lipsius.*'^ Beltrami must use Q and its 
inverted order to produce prodeuntes fugitivi, which I think highly 
improbable. 

This passage is misunderstood by the translators. Excludere is the 
word for "shut out (of a house)"; it is so used in the last part of the 
sentence above. The ears sticking out of the shawl when the rest of 
the head is covered make Seneca think of two people who have been 
shut out of a place of shelter. Apparently in the play two slaves 
have run away from a wealthy master; they decide to return, but 
find themselves shut out, not wanted. That is their punishment, no 

^2 Windhaus, op. cit., p. 35, n. 5, questions the integrity of divites, thinking 
it may be the product of a mechanical error (dittography of the tivi of 
fugitivi plus an s generated out of solent). But the appearance of divites in Q 
before fugitivi shatters that theory. Nonetheless the divites does seem in- 
trusive. 



202 University of California Puhlications in Classical Philology 

doubt, not to be admitted within the conJSnes of the comfortable 
home of their master. In any event, they must have been two 
abject and forlorn-looking persons in the play, and where Mae- 
cenas' headgear is concerned the ears irresistibly make Seneca 
think of them. The exclusis with respect to the ears has a meaning 
somewhat different from that of the excludi, which must be assumed 
to go with the solent, but it is a meaning derived from it and easily 
connected with it. 

CXIV, 16 (552 : 18) : non tantum in genere sententiarum vitium 
est, si aut pusillae sunt et pueriles aut improbae et plus ausae 
quam pudore salvo Hcet, si floridae sunt et nimis dulces, si in 
vanum exeunt et sine effectu nihil amphus quam sonant. 

Before si floridae Fickert reads with the editio Gothofredi and 
Schweighaeuser <sed> , and so also Haase; this is also the vulgate 
reading. It would have been well for Hense (II) and Beltrami (II), 
both of whom often take space and time for less important ex- 
planations, to explain that the dropping of < sed > to conform with 
the reading of the mss. is made logical by Windhaus' explanation 
of the passage.^^ He notes that the true contrary of non tantum . . . 
vitium est is sed in hominihus, which, had it been formally expressed, 
would have followed sonant. As a matter of fact it is taken up in- 
formally by aliquis unus; the fault is not only in the sententiae 
themselves, but in some person who is powerful enough in the 
literary world to estabHsh their vogue. 

CXIV, 22 (554 : 3): hoc a magno animi malo oritur: quomodo in 
vino non ante lingua titubat quam mens cessit oneri et inclinata 
vel prodita est : ita ista orationis quid ahud quam inebrietas nuUi 
molesta est, nisi animus labat. 

In the interpretation of this passage it is worth observing care- 
fully that inclinata vel prodita goes with lingua and not, as Barker 
(for instance) assumes, with mens, as might seem more natural from 
the apparent word arrangement. A comma after cessit would bring 
this out. The translation is: "just as in the matter of drinking the 
tongue does not trip until the mind fails beneath its load, and it 
(the tongue) has been put to rout or, if you will, betrayed." It is the 
tongue that is betrayed by the mind when the latter has succumbed 

p. cit., p. 35, n. 8. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 203 

to its alcoholic burden. In view of the foregoing explanation it will 
be seen how weak is the perdita reading of the inf eriores. 

CXV, 6 (557 : 11) : nemo, inquam, non amore eius arderet, si nobis 
illam videre contingeret : nunc enim multa obstrigillant et aciem 
nostram aut splendore nimio repercutiunt aut obscuro retinent. 

Q B A agree in the reading obscure; this, however, is an imperfect 
parallel to splendore nimio, and obviously we require an ablative. 
Muretus' obscuro is best commented on by Hense's remark in his 
app. crit. ad loc, namely: '^obscuro vel m obscuro si verum est, 
substantivi vice fungitur." There is no objection so far as grammar 
or usage is concerned, but it fails metrically to produce a proper 
clausula. 

On the other hand, Pincianus' obscuritate gives us a sound 
clausula rhythm, a cretic followed by a cretic in which the first long 
is resolved. I have no doubt whatever that the obscure of Q B A is 
simply an error in transmission for obscuritate. The final e seems 
definitely to point that way. 

CXVI, 6 (562 : 25) : quod Panaetius de amore quaerenti respondit, 
hoc ego de omnibus adfectibus dico. quantum possumus, nos a 
lubrico recedamus. 

All the mss. read nos. Haase rejected the word; some editors 
have placed the comma after nos to attach it to possumus. Gertz 
felt that nos was intolerable as the subject of recedamus and so con- 
jectured reducamus. This is also the reading of the Codex Velzianus, 
and demonstrates that some medieval Latinist had felt the same 
way about the situation as Gertz. 

There is absolutely no reason why nos should be expressed as the 
subject of recedamus; in that I agree with Gertz, but would hesitate 
to embark on reducamus in order to make 7ios an object. Nor can I 
accept Haase's disposition of the pronoun by the simple process of 
square bracketing. The attempt to throw it with possumus seems 
trifling. I suspect that actually the word was originally non, and 
that since the meaning, which is somewhat epigrammatic, was 
early missed, nos took its place. 

What Seneca really said was "as far as in us lies, let us not be 
(in the position of) struggling back from a slippery spot." Those 
who follow the practice of yielding something to the emotions will 



204 University of California Publications in Classical Philology 

constantly be finding themselves in a slippery spot from which it 
will be a struggle to regain terra firma; let us who are wise not be 
found in such a situation. I view the verb recedamus as being situ- 
ational in value; that is why I translate it by "trying to scramble 
back." 

With the reading I propose, namely, non for nos, and the view I 
take of the function of recedamus, the rest of the sentence accords 
perfectly. "We have a hard enough time maintaining a footing even 
on dry ground; let us not be trying to scramble back from slippery 
soil." 

CXVII, 8 (565 : 10) : si turpitudo malum est, et turpem esse malum 
est, tam mehercules quam, si hppitudo malum est, Hppire quoque 
malum est. hoc ut scias, neutrum esse sine altero potest, qui 
sapit, sapiens est; qui sapiens est, sapit. 

There appears to be some incoherence here centering around the 
hoc ut scias clause. I suggest following the text of Q, namely: neutrum 
esse sine altero potens, and altering the full stop after the last malum 
est to a comma, so that hoc . . . potens becomes the concluding por- 
tion of the preceding sentence. Translate: "if baseness is an evil, 
then being base is also an evil, exactly as if weakness of the eyes is 
an evil, to be weak-eyed is an evil, with the result that you know 
this, that neither item (sc. in these pairs) has effectiveness without 
the other." Esse potens looks hke a possible translation of a dwarov 
elvat, and potens written as potes would stand a good chance of 
developing into the potest of B A, 

CXVII, 23 (570 : 5) : haec nempe sunt fet] elementa, quibus hie 
mundus administratur, aqua, terra, spiritus. omnia ista tam 
causae vivendi sunt quam viae mortis. 

The mss. agree on et elementa. As Beltrami (II) says in his 
critical note on the passage, et ("actually") is sustained by the tam 
. . . mortis section of the next sentence; furthermore, sunt elementa 
is an unfortunate rhythm, while sunt et elementa gives a cretic with 
the second long resolved and a trochee. 

It is surprising in the conclusion of the first sentence given above 
to find only three of the four elements mentioned, and no amount 
of ingeniously applied learning can explain that surprise away. 
Axelson"" places a semicolon after administratur and eliminates the 

" Neue Senecastudien, pp. 214-216, with several footnotes. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 205 

period after spiritus, so that the recapitulatory omnia ista accounts, 
in his view, for the missing ignis. But the word elementa has been 
specifically used, and the elements are to common knowledge four 
in number; why should a person, if he is going to name them at all, 
stop short with three out of the four named? I am not impressed 
with Axelson's idea that to name the fourth would make a school- 
master out of Seneca. 

Windhaus®^ inserted ignis after spiritus and placed a comma after 
the latter; this is common sense and finds some paleographical 
justification. Unfortunately, however, as Axelson observes, this 
leads to the undesirable clausula heroa. If, nevertheless, spiritus 
can be regarded as a genitive, we have an excellent cretic-trochaic 
close. 

This suggests a clue, and in Dial. 4, 19, 1 we find this significant 
passage : nam cimi elementa sint quattuor, ignis, aquae, aeris, terrae, 
potestates pares his sunt. I beheve that the rather intolerable situ- 
ation in the passage under discussion is cleared up if we read aquae, 
terrae, spiritus, ignis. One can see how easily aquae, terrae would 
be altered to aqua, terra, on the supposition that spiritus, ignis 
were also nominatives. By the reading suggested we obtain the 
mention of the four elements without running like Windhaus into 
an unfortunate rhythm, and for the defense of the proposed change 
we have the perfect parallel in the passage quoted above from the 
Dialogi. 

Haec at the beginning of the passage refers presumably to exitum 
immediately preceding, but now conceived under the plural idea 
since four elements are being mentioned as ways out; it is subse- 
quently attracted into the gender of elementa. 

CXVIII, 4 (573 : 22) : quanto hie maiore gaudio fruitur, qui non 
praetoria aut consularia comitia securus intuetur, sed magna ilia, 
in quibus ahi honores anniversarios petunt, ahi perpetuas potes- 
tates, ahi bellorum eventusprosperostriumphosque, alii divitias, 
alii matrimonia ac Hberos, alii salutem suam suorumque ! 

It seems incredible that Seneca, after dismissing the annual 
elections for the praetorship and the consulship as among the minor 
things by indifference to which a man may prove his claim to be a 
philosopher, should subsequently include honores anniversarios 
among the major things {magna ilia) disregard for which will es- 

" Op. cit., p. 37, n. 4. 



206 University of California Publications in Classical Philology 

tablish a yet better claim to the same distinction. Read therefore 
honores <non> anniversarios, "offices not involving annual elec- 
tion," such as a five-year commissionership or a procuratorship. It 
will be noticed that we then have a regular succession in the sen- 
tence determined by lengths of time, praetorian and consular 
elections, offices not subject to annual election, life (perpetuas) 
powers like Caesar's dictatorship. 

CXVIII, 7 (574 : 18) : maiorque pars miratur ex intervallo fallentia, 
et vulgo bona pro magnis sunt. 

Numerous suggestions have been made for rendering the second 
half of this sentence more logical, as the emenders would say, but 
perhaps, as Seneca might say, only painfully obvious; they are 
assembled in the app. crit. ad loc. in Beltrami (II). 

The drift of the meaning is estabhshed by the first half of the 
sentence; "distant pastures look greenest." The second half of the 
sentence is a restatement of that same idea through an odd bra- 
chylogy. The full expression would be: vulgo bona pro magnis 
<magna pro bonis > sunt; "to the mass of mankind good stands 
for big, and big stands for good," that is, the ideas are perfectly 
interchangeable. The manuscript tradition is sound. ^^ 

CXX, 12 (584 : 31) : quicquid inciderat, non tamquam malum 
aspernatus est et in se casu delatum, sed quasi delegatum sibi. 
hoc qualecumque est, inquit, meum est; asperum est, durum est, 
in hoc ipso navemus operam. 

The translators with extraordinary unanimity translate the first 
of the foregoing sentences precisely as if it ended with the verb 
excepit (cf. p. 584:30). Barker's rendering will serve to illustrate the 
point: "he never revolted against anything that hghted on him, 
but accepted it as a charge." Obviously there is no "accepted" in the 
Latin, nor can it be readily inferred, if at all, from the first half of 
the sentence. The editors pay no more attention to it than the 
translators; yet in the preceding sentence, to which this is a per- 
fect parallel in construction, we have numquam . . . excepit, labores 
. . . subit. 

It is to be observed, however, that Erasmus (II) places a colon 

66 Windhaus, op. cit., p. 37, n. 7, writes: "intellegas: magna vulgo bona 
videntur." By this note he shows that he is in accord with my idea of the 
terms bonus and magnus being to the vulgus interchangeable. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 207 

after casu delatum, thus throwing sed quasi delegatum sihi (after 
which he places a comma) forward with hoc qualecumque est etc. This 
makes sed the adversative connective between aspernatus est and 
inquit, not between casu delatum and quasi delegatum, and in 
general pays regard to the logic of the sentence and the require- 
ments of the meaning. I should, however, prefer, adopting Erasmus' 
punctuation, to read quasi delegatum <sit> sibi, where the se- 
quence of the sit is established by its association with inquit. The 
clausula is unaffected rhythmically by the addition. 

In what follows I should hke to read durum est with an interro- 
gation mark following; this seems to me to improve greatly the 
dramatic setting. 

CXX, 18 (586 : 6): vide in quanta caecitate mens nostra sit: hoc 
quod futurum dico, cum maxime fit, et pars eius magna iam 
facta est, nam quod viximus. 

One would like to know what Vahlen could possibly have written 
in his letter to Hense to persuade him that nam quod viximus was 
sound Latin and that therefore Buecheler's iam quoad was un- 
necessary. I do not myself care for the order of words demanded by 
Buecheler's emendation. Some change or addition, however, is re- 
quired. The reading of Beltrami (II), nempe quoad viximus, would 
be better with quoad left as quod. I quite understand that Beltrami 
thinks the preceding magna pars would demand quam, but I re- 
gard quod viximus as being equivalent to id vitae peractum nobis. 

The other method of correction is to add something. Barker" 
suggests <morti addiximus> to follow viximus. Simpler yet I feel 
would be <degimus> ; degere vitam is a commonplace phrase of 
Latin. The chances for the loss of <degi7nus> between viximus 
and erramus are unusually good.^^ 

CXX, 20-21 (586 : 26 to 587 : 10) : maximum indicium est malae 
mentis fluctuatio et inter simulationem virtutum amoremque 
vitiorum adsidua iactatio. . . . multos dixi? prope est, ut omnes 
sint. 

" Op. cit., Vol. II, p. 332. 

«« My seminar student of 1940-41, Mr. Robert Lane, thought that degimus 
lacks in strength for the very reason that it is commonplace, while we really 
need something that emphasizes the notion of life's being hopelessly gone. 
A fellow student, Mr. Sam Abdallah, felt that that need would be met by 
reading quod viximus, <viximus> , with the same sense of finality as Pilate's 
"what I have written, I have written." (Session of April 23, 1941). 



208 University of California Publications in Classical Philology 

In this passage, too long to be repeated in full, there are two 
difficulties, (1) the mysterious is (587:1) which follows iactatio and 
immediately precedes the quotation from Horace, and (2) the im- 
possible multos of multos dixi? (587: 9). 

In spite of Axelson's feeling®^ that no helpful Ught has been shed 
on is, I venture to reaffirm my guess^" "that one explanation may 
solve both difficulties at a stroke. I should not, however, in my 
previous remarks on the passage have attached fit multis to adsidua 
iactatio, which is a perfect clausula and brings its sentence to a neat 
end. I now suggest that there is a lacuna of some such nature as 
<quod quidem ipsum accidit mult> between iactatio and is; "pre- 
cisely this thing (viz., fluctuatio and iactatio) happens to many." 
Some use is thus found for is, and the multos of multos dixif is 
satisfactorily anticipated, as indeed it must be. 

Whatever is thought of this conjecture as a whole, I think that it 
points to a more intelHgent use of is than other emenders have 
found for it. The cause for the lacuna which I postulate I do not 
pretend to suggest; if it did actually break the word multis, it must 
have been due to an accident of some sort. 

CXXI, 20 (593 :9): et tardum est et varium, quod usus docet; 
quicquid natura tradit, et aequale omnibus est et statim. 

Beltrami (II) presents the same text. Windhaus" observes: ad- 
verbium statim praedicati loco esse non potest, and despite the fact 
that the occasional Latin adverb Uke bene or )7iale is employed in 
what looks like the function of a predicate adjective, I cannot con- 
vince myself that that is true of stati7n here. 

There have been, therefore, numerous attempts to replace 
statim, most of them missing the meaning of the passage. There 
have been efforts also to dispose of est and find a verb which 
statim could modify. To all that has been said I have only this to 
add, that the first half of the sentence has a perfect clausula (-um 
quod usus docet), and that hence it would seem to me unhkely 
that in such a balanced sentence the second half would not. If est 
is dropped, a fair clausula can be obtained by the order et ojmiibus 
aequale et statim; the transposition of two words is no impossibility. 
It should be added that if est is dropped we may either read tradit, 

** In his Senecastudien, p. 113. 

'" In my N-otes and Emendations already referred to, p. 14. 

" Op. cit., p. 40, n. 4. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 209 

<tradit>, or assume that the tradit carries through from the sub- 
ordinate clause. If for some reason tradit fell out and was replaced 
by est with a result that was in part ungrammatical, it may be 
noted that the restoring of tradit where est now stands would also 
give a good clausula of the hypodochmiac type (Axelson 4) . 

CXXII, 1 (595 : 1) : officiosior mehorque, si quis ilium exspectat et 
lucem primam f exuit. 

Exuit: B A p q (the testimony of Q being no longer available, but 
the fragment of p becoming serviceable from CXXI, 12 through 
CXXII). ^2 

For a very complete statement of conjectural efforts on exuit, 
Beltrami's critical note ad loc. in his second edition should be 
consulted. It is obvious that several of them, while intrinsically 
excellent, cannot be entertained because of remoteness from the 
manuscript reading. Beltrami's own efforts with ut . . . exuat and 
Albini's" defense of the phrase primam lucem exuit (sc. umbris) 
seem to labor very heavily. 

Axelson^^ beheves Gruter's excipit "richtig oder wenigstens als 
die einzige un Ernst diskutable Verbesserung dieser vielbesprochenen 
Stelle," where excipit has the sense "welcomes." Incidentally, 
Axelson shows by his footnote on page 13 that he does not under- 
stand Summers' <non> before exspectat. What Summers has in 
mind is a man who does not wait for day to break but is up before- 
times "arousing the first light." His <non> exspectat is quite in 
harmony with his later excitat; his industrious individual "does not 
await the day, but gets the dawn out of bed." This is really, in 
spite of Axelson' s cavalier treatment of it, a very bright idea and 
quite in the Silver Latin manner; but it involves too much textual 
change. 

My earlier suggestions^ si quis ilium exspectat et luce prima 
ex<cw6>uit is paleographically quite possible and provides sound 
sense, but is questionable on metrical grounds. I now propose 
luce prima (the ablative is from Pincianus) excubat; this makes a 

" Hense (II), praefatio, p. vi. 

" Albini, Rivista di Filologia, n.s., Vol. VI (1928), p. 546. 

~* Neue Senecastudien, pp. 13-14. 

'5 Op. cit., p. 14, fn. It is a question whether all the labor that has been 
expended on this word would not be sensibly terminated by recognizing the 
paleographic excellence of Windhaus' exilit and its admirable adjustment to 
the context. 



210 University of California Publications in Classical Philology 

double cretic clausula. Once ba was dropped, it is easy to see the 
development of exuit from excut. Translate: "he is a more indus- 
trious and a better man who is awaiting the day and at dawn is on 
the watch." Compare the expression of thought in section 3 fin.: 
circumscribatur nox, et ahquid ex ilia in diem transferatur. The 
only way to narrow the boundaries of night is to transfer some of 
night to the day, that is, to be up before daylight. 

CXXII, 8 (597 : 10) : non vivunt contra naturam qui hieme con- 
cupiscunt rosam fomentoque aquarum calentium et calorum apta 
mutatione bruma hhum, florem vernimi, <ex>primunt? 

Aquarum has no ms. authority, and the plural certainly suggests 
medicinal springs rather than hot-water apparatus. I think it more 
probable that the original text was fomentoque <aque> calentis, 
and that when <aque> (i.e., aquae) dropped out by a simple 
haplography, calentis was "adjusted" to calorum. This word, by the 
way, appears in the mss. with the exception of A, as colorum; calorum 
seems right. 

With Barker^^ I am opposed to tampering with hrumalium, which 
incidentally is defended by clausula ending within its own phrase 
(-one brumahum). Like Barker, I take the phrase to mean "by a 
clever change inducing heat at mid-winter," a sort of commentary 
on f omenta <aquae> calentis. It seems to me sadly unimaginative 
to drag the hly in for the flos vernus when the rose has already been 
mentioned, the most distinguished flower of spring in the Itahan 
climate or in climates comparable with that of Italy. 

There is something wrong with the concluding verb, but the mss. 
apparently point to something like primunt, and it then becomes a 
question of which compound of pre7no will serve. <Ex> primunt 
"force" is very satisfactory in the connection, and has the recom- 
mendation, for what it is worth, of being found in q. It also pro- 
duces a satisfactory clausula, with a molossus (florem vern-) preced- 
ing, as not uncommonly. 

CXXII, 13 (599 : 4) : deinde cum subinde recitasset (sc. Montanus 
lulius) : 'iam sua pastores etc' 

Beltrami (II) reads demum with q and some of the inferiores, 
rejecting deinde "quod cmn sequente subinde moleste consonat." 

" Op. cit., Vol. II, p. 333. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 211 

Axelson" points out that demum never occurs unsupported, that is, 
unattended by another particle, in Seneca. The deinde seems per- 
fectly natural in the narrative; the suhinde appears to be confirmed 
by the rhythm (cretic with second long resolved, followed by 
trochee). One must therefore apparently reconcile himself to the 
jingle deinde . . . suhinde. The suhinde means here: "shortly after- 
ward." 

CXXII, 15 and 16 (600 : 1 to 600 : 12) : 'audio' inquit . . . 'dicetis.' 
The great difficulty with this passage, which because of its 
length cannot be reproduced here, is to determine where the words 
of Pedo Albinovanus cease. Hense's punctuation carries the speech 
clear through from audio to dicetis, and both the inquit following 
excedebat and also the inquit following sordidum vos apparently 
serve for Hense the purpose of reminding us every little while that 
everything down to dicetis is Pedo's. Beltrami (II) brings Pedo's 
continuous speech to an end with noctem; the words itaque . . . 
sordidum are Seneca's, and the third inquit refers only to vos . . . 
dicetis. Summers makes Pedo's narrative end with a halneo exisse. 
At that point an objection is raised in the normal inquit fashion, 
"says one," to the effect that as cocktails have been called for just at 
dawn and dinner must obviously follow, Papinius' banquet over- 
laps his chosen day (which, of course, is night). Smnmers then 
follows the older editors in closing the objector's remark with cena; 
minime becomes Seneca's "no" and valde . . . vivebat the explanation 
of the "no." Pedo's words are not resumed till vos is reached. I hold 
by this punctuation and this arrangement of words. 

Summers^^ points to another difficulty in vos; "the emphatic vos 
seems to require a quin or a vero : the latter is paleographically the 
easier; quin would require dicitis." It is true that the statement 
headed by vos brings out a contrast with the material of dicentihus 
. . sordidum, but I think that this contrast is obtained inf erentially 
and will not require a particle attached to vos to bring it out. Indeed, 
the mere placing of the vos may be considered sufficient. I could 
myself have enjoyed finding <mox> , "next," after vos. 

Next comes credendo, the reading of B and the fragment of p, 

" Neue Senecastudien, p. 3 and n. 7. From the severity of his condemnation 
of the lapses of Senecan editors and critics in regard to what is and is not 
Senecan usage, I assume that Axelson has fully verified this statement. 

^8 Op. cit., p. 150, fn. 



212 University of California Publications in Classical Philology 

dismissed by Fickert with the remark "sub quo quid lateat, videant 
doctiores," but not so hghtly to be got rid of. The mss. inferiores all 
have the easy credo; "and that is why, I presume." Pincianus and 
Summers have tried to find Pedo's name here repeated, but, I 
think, quite without success. Axelson^^ has, in my opinion, put his 
finger on the solution with his ridendo for credendo, which he 
attaches to dicentibus, but I to the subject of ifiquit. Translate: 
"and so, laughingly, to certain parties who were calUng Papinius 
grasping and mean, he remarked : 'You will find yourselves calhng 
him a liver-by-lamps as well.' " I regard the future dicetis as situ- 
ational; not, with Axelson, as a quasi-imperative. 

Finally, in connection with this remark of Pedo's I think that 
Axelson has rightly advised us against following the elaborate pun 
scented there by Pincianus as between Xvxvos and \ixvos. His 
view that Pedo meant that Papinius would be presently described 
as living not simply hy the lights but off the lights, that is, by con- 
suming their oil in skinflint fashion, appeals to my sense of humor. 

CXXIII, 10 (605 : 9) : fluunt dies et irreparabihs vita decurrit : dubi- 
tamus sapere? quid iuvat [et] aetati non semper voluptates re- 
cepturae interim, dum potest, dum poscit, ingerere frugaHtatem? 
eo mortem praecurre et quidquid ilia ablatura est, iam sibi f 
interere. 

The transposition of sapere from its manuscript position after 
iuvat to follow dubitamus, and the bracketing of et as effected by 
Buecheler and Hense, seem wholly unjustified; sapere means "to 
play the sage" in the almost technical sense. The sohtary dubita- 
mus? may seem difficult, but I prefer it to any emendation such as 
the cur dubitamus? in Beltrami (II), which produces a dactyhc 
hexameter end rhythm. 

For the rest I suggest that the concluding words should be iam 
tibi interierit; the meaning I take to be : "therefore steal a march on 
death, and whatever she is (sc. as things now stand) going to take 
away from you will already (sc. when death comes) have been dis- 
posed of— to your profit." The rhythm is sound, a ere tic followed 
by a cretic with first long syllable resolved. 

CXXIII, 12 (605 : 1) : hae voces non aliter fugiendae sunt quam 
illae, quas Ulixes nisi alhgatus praetervehi noluit. idem possunt: 

'9 Neue Senecastudien, pp. 219-220 with the footnotes. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 213 

abducunt a patria, a parentibus, ab amicis, a virtutibus et inter 
spem t vitam misera nisi turpis inludunt. 

B Apr. : et inter spem vitam misera nisi turpis inludunt 
Codex Harl. 2659: inturpem vitam miseros illidunt 
q : as for codex Harl. 2659 except illudunt. 

The patience of Beltrami has built up in his first edition^" the 
amazing collection of conjectures which have been made regarding 
this passage, and in the app. crit. ad loc. in his second edition^^ he 
has added to that list other later attempts. Of these a number have 
to be rejected at once through their failure to recognize any prin- 
ciple of clausula rhythm. 

As we attack this crux, two things emerge which may be regarded 
as definitive toward a solution. First, inter spem as giving a de- 
scription of a person or thing which might ordinarily be expected to 
be presented by a present participle or a clause of some sort occurs 
elsewhere in Seneca. Thus in Dial. 4, 19, 5 we find: color quaHs 
fieri ceteris inter iram solet, where inter iram has the effect of 
irascentibus or dum irascuntur. I therefore regard inter spem as 
meaning "while raising hopes"; this is made clear by the obvious 
contrast with illudunt. The second definitive point is that turpis 
inludunt is a complete and accurate clausula rhythm (cretic plus 
spondee, Axelson's first type), and may therefore be regarded as 
sound. We are thus left with only vitam misera nisi to bring under 
control. From this group of words as supplied by the manuscripts 
we must find a noun for turpis to modify, and all the circumstances 
point to the noun's being in the genitive singular. I find such a noun 
in Haase's situs, a good Senecan word; the -tus has been lost before 
turpis, while the si- has been united with the final m of miseram to 
produce nisi. We thus recover the sentence in this form : inter spem 
vitam miseram situs turpis illudunt. This translates: "while raising 
hope they mock their pitiful hfe of shameless inertia." 

How does that relate itself to the context? In the preceding 
sentence we have had the Sirens mentioned, and the idem possunt 
with which the sentence under discussion opens suggests that 
through it also we shall still be talking of the Sirens. But reference 
to the Homeric narrative^^ shows that yielding to the enchantment 
of the song of the Sirens brought shipwreck and death, not the 

8° Vol. II, praefatio, p. Ixxvi. ^i Vol. II, p. 284. 
«2 Odyssey, XII, 40-46. 



214 University of California Publications in Classical Philology 

things mentioned in this sentence of Seneca. Its phraseology is 
appropriate in connection with the Lotus Eaters; those who eat of 
the lotus lose all desire for kindred, country, and action, ^^ and are 
willing to remain in Lotusland enjoying inglorious ease. Thus 
Seneca has unconsciously slipped from one Homeric parallel to 
another. A paraphrase of the sentence as I have reconstructed it will 
reveal the appropriateness of the emendation to this view of the 
context. "These voices of pleasure (Hke those of the Lotus Eaters) 
turn (men) from their country, from their parents, from their 
friends, from manly deeds, and while raising hopes (in their victims) 
really make mock of their pitiful hfe of shameless inertia." The 
latter part of this sentiment is no doubt more true of the voices of 
pleasure than of the Lotus Eaters, but parallels have a habit of 
getting out of hand.^^ 

CXXIV, 5 (608 : 6) quid si quis vellet non oculis, sed tactu minuta 
discernere? subtihor adhoc acies nulla quam oculorum et in- 
tentior daret bonum malumque dinoscere. 

See Beltrami (I) (Vol. II, introd., pp. Ixxvii sqq.) for an excellent 
restatement, in much clearer Latin than Seneca's, of the whole 
argument here. In general the contention is this: it is as absurd in 
the moral world to remit the decision on the Supreme Good and 
Evil to the lower rather than to the higher faculties of man, as it 
would be in the physical world to leave the judgment on minute 
objects to a lower sense, touch, rather than to a higher one, sight. 

I read and punctuate thus: suhtilior ad hoc {ad and hoc separately 
as in B A) acies nulla quam oculorum; et intentior, daret bonum 
malumque dinosceref This translates thus: "For this purpose (sc. 
minuta discernere) there is no penetration more subtle than that of 
the eyes; even more keenly applied (sc. than it now is in the physi- 
cal world), would it have conferred the power to distinguish good 
from bad?" Or intentior might be translated "even were it keener 
than it is," another view of the protasis for daret. 

This interpretation, without any alteration of the ms. text, gives 
the meaning which Madvig saw the sentence must convey, though 

83 Odyssey, IX, 94-97. 

8< Dr. B. L. Charncy thinks that the word order vitnm rmseram situs turpis 
is not very attractive, and I agree, but the ms. tradition nonetheless points 
to that order. In much the same way I prefer my former suggestion situ 
turpi; but the clausula rhythm stands in the way of that. 



Alexander: Seneca's Epistulae Morales (XCIII-CXXIV) 215 

he made two serious changes to secure this. It takes the words as 
they come, and suggests nothing of the hberior verborum collocatio 
of which Hense (II) speaks in his critical note. 

CXXIV, 22 (613 : 1) : cum ilkim (sc. capillum) vel effuderis more 
Parthorum vel Germanorum modo vinxeris vel, ut Scythae solent, 
sparseris, in quolibet equo densior iactabitur iuba. 

The meaning of sparseris is of some interest. In the preceding 
phrases two coiffures have been described; that of the Parthians, 
who let their hair hang free; and that of the Germans, who tied 
their hair up. Pincianus' vetus lectio was nodo for the ms. modo, 
and this reading at least expresses the fact as set out by Tacitus in 
describing the powerful tribe of the Suevi.^^ Furneaux^^ doubts 
whether this custom was quite so distinctive of this one tribe as 
Tacitus makes it out to be, and recalls that Seneca^^ assigns the 
crinis rufus et in nodum coactus to Germans generally. The words 
of Tacitus' description obliquare crinem nodoque substringere 
appear to indicate that the long hair was combed back to right and 
left and then wound into a queue. This queue was ordinarily formed 
into a knot at the back of the head, but often (§3) piled on top of 
the head. 

It now becomes clear that about all that is left to assign to the 
Scythians is the practice of dividing the hair into several strands as 
against the German custom of a single queue. In all the instances of 
barbarian coiffure mentioned the hair is obviously long because it is 
compared with the mane of a horse, and we have already seen it left 
to hang naturally by the Parthians, and combed back and formed 
into a queue by the Germans. Barker^^ in his translation suggests 
"wear it in ringlets" as a rendering of sparseris, but I do not see 
that this can be correct; the usual effect of ringlets visually is that 
of close matting, not of scattering. If, however. Barker means that 
the Scythians divided the long hair of the head into six or eight 
strands and formed each of these into a large curl, that is another 
matter. That is what I conceive spargo in this connection to mean; 
it refers to the first process of "scattering" the hair by breaking it 

_*^ Germania, chap, xxxviii. I do not say that this compels us to adopt 
Pincianus' reading. 

*^ Henry Furneaux, The Germania of Tacitus (Oxford, Clarendon Press, 
1894), p. 104. 

" Dial. 5, 26, 3. 

88 Op. cit., Vol. II, p. 322. 



216 University of California PuUications in Classical Philology 

up into several sections for curling. It had better therefore be trans- 
lated ''part (into several curls)" to make the process clear; Seneca 
was unwilling to be more specific with sparseris, wishing' to pre- 
serve a close economy in his parallelism of it with effuderis and 
vinxeris. 



i 



VARIA CRITICA 

BY 

JOSEPH E. FONTENROSE 

I. Theocritus, Idylls 1. 13 and 5. 101 

In Idyll 1. 12-14 the shepherd says to the goatherd, according 
to the currently accepted text : 

Afjs TOTL TO-v Nu/i^aj', Xgs, aliroKif TJ^Se Kadi^as, 
cbs TO KaravTes tovto ye(xi\o(pov a'i re /xypuat, 
avpiadev', ... 

hro: SH^AETr 

are (are) : all MSS^ 

The line is repeated in Idyll 5. 101, where all manuscripts appear 
to read cos and are {are). The goatherd cries to his goats, according 
to the accepted text : 

Strr' airo tcLs kot'lvo} rat fxrjKaSes' cSSe veneade, 
cos TO KCiTavTes tovto yeuiXoifov at re juupt/cat. 

Earlier editors, such as Valckenaer and Kiessling, accepted cos 
and are, interpreting cos as the preposition, though recognizing the 
anomaly of its governing an impersonal object. But Ahrens (who 
rejects the line in Idyll 1) and Meineke, having found cos with the 
apparent meaning of 'where' in a Thessalian inscription, interpreted 
it thus in this line and changed aTe to at re, so that the meaning of 
the line becomes 'where yonder sloping hillock and the tamarisks 
are. '2 This reading and interpretation are now generally accepted, 
and the Liddell-Scott-Jones lexicon has a separate cubicle under cos 
where the editors define it as 'where,' citing this line of Theocritus 
in both occurrences, four inscriptions, and also Idyll 5. 103. 

We can justify the interpretation of cos as a relative adverb of 
place by pointing to its kinship to xcos, to the fact that u8e some- 
times means 'here,' and to Latin wf, which seems to mean 'where' in 

1 To designate the manuscripts of Theocritus I use the symbols of U. von 
Wilamowitz-Moellendorf's edition (Oxford, Clarendon Press, 1905). 

- ws, interpreted as 'where,' and al re are accepted by Ahrens, Meineke, A. 
T. A. Fritzsche, Ziegler, Hermann Fritzsche and E. Hiller, Snow, Wilamo- 
witz, Cholmeley, Edmonds, and Legrand. The interpretation of <hs as 'where' 
is accepted by J. Rumpel, Lexicon Theocriteum (Leipzig, Teubner, 1879), s. v. 
ws, p. 318. 

[217] 



218 University of California Publications in Classical Philology 

a few places. Yet cb$ occurs thousands of times in extant Greek 
literature and inscriptions with a well-established range of meaning, 
and it is seldom difficult in any instance for the reader to discern 
whether cos has its comparative, modal, causal, temporal, or final 
force. But for one or two places in Theocritus and for a very few 
occurrences in inscriptions the reader needs the special knowledge 
that cos can mean 'where'; his acquaintance with cos elsewhere can 
hardly have prepared him to see this sense in cos at a glance. Since 
it is always undesirable to grant a special and unusual meaning to 
a word to suit only a very few instances of its occurrence, unless 
we are forced to,^ let us look more closely at the places where cos is 
interpreted as 'where.' 

The Liddell-Scott-Jones lexicon cites four inscriptions in support 
of cos = where. (1) I.G. IX 2. 205,^ the inscription that first led to 
the interpretation of cos as 'where,' is the record of an arbitration 
of a boundary dispute between the Thessalian cities Meliteia and 
Pereia. Lines 3-8 : . . . "Opta fxev djxev tSs x'*'Ptts MeXtrate j ots /cat 
Ilr/peots cos 6 'AKjievs e^^(SaXX€t ev tov EupcoTroi', kol airo rod \ 'A/c/ieos tp rav 
irayav tov Taka'iov, kol airo tov TaXalov kv tclv KoXco | vav, kol airo toLs 
KoXcovas ewl to 'Ep^iatoi' kwl to. KvpvvLa, /cat airo tu)u E[u] | pvvlup /card 

TCOJ' CLKpCOV COS uScop p€t kv TOV EvpCOTTOV, €K TOV El'pCO | TTOV kv TOV 'EXtTT^. . . . 

(2) S.G.D.L 5597 (287-281 b.c); the government of Ephesus de- 
fines the terms on which it will lease certain public lands, but makes 
an exception of certain areas, one exception being indicated as fol- 
lows (lines 7 f.) : /cat tov "AaTvayov ira7ov|cbs 6 areipuv w€pi,<pep(t kvkKui 
TOLVTa k^aipevneda. . . .^ (3) I.G. XII 1. 736; an inscription of Camirus 
(3d century b.c.) f the boundaries of a grant of land for a cemetery 

are defined as follows (lines 2-9): [ 'ejpai'to-Taj' rcot /cotvcot e|5a)/ce 

Soopeav ts Tatp'ia \ tSs 7as ras kv "Po7/ci'co[i] | cos d 656s d ipkpovaa e^ 'A | 71/- 
Xeias ets 'lTnroTda.v,\fj.a.Kos opyvdv t'lKoaL 7r€i'|re, rrXdros bpyvav ScKajt^, 
cos opot KelvTaL. ... (4) S.I.G.^ 685; the arbitration by Magnesia-on- 
Maeander of a boundary dispute between the Cretan cities Itanus 
and Praesus (139 b.c.).^ Lines 63-65: cbs 6 Z^eSajuvos « Kapvpas ks tolv 
8r]pa.8a /cat ire[pav es] tolv CTupavav /cat irepiapireTL^ cos d (TT€(pava /cat 
eWv I coptg. ts AopOavvas eirl tov \aKK0V /cat cos d 6[56s] ttotI peaav^piav ras 

3 See my article "On the Particle ttw in Homer," A. J. P. LXII (1941), 68. 
'' This inscription is also found in Le Bas-Waddington II 1179, S. G.D.I. 
1415, Michel 22, S.I.G. (3d ed.) 546 B. Its date is 216-203 b.c. 
5 On this inscription see Wilamowitz in Ilcrmcs XXXIV (1899), 209-212. 
8 See als.o S.7.G. (3ded.) 1118, B.C.H. IV (1880), 138 f. 
^ See also O. Kern, Ins. Magn. 105, S.G.D. I. 5060. 



Fontenrose: Varia Critica 219 

6800 ras dTwo-as 5t' 'ArpUva /cat es | MoWov. . . . These words are re- 
peated in lines 66 f . 

To these I can add the following two inscriptions. (5) Bernard 
Haussoullier, Etudes sur Vhistoire de Milet et du Didymeion (Paris, 
Bouillon, 1902), p. 78 (253 b.c.) ; record of the purchase of a section 
of royal domain by Queen Laodice, inscribed at Didyma. Lines 
45-47 : . . . airo I 8e TavT-qs irapa tov tov Aios ^wfxov tov ovra kiravw ttjs | 
Bdpecos Kol cos 6 Tcupos kv 5e|ta rr\s bbov. ... (6) S.I.G.^ 826 E, III 7 
(117/116 B.C.); a decision of the Amphictyons on the boundaries 
of the Apollo of Delphi's sacred lands. Line 24: [7r]apd [ttip pL]^av Kip- 
<pov COS v5c)}p pen. . . . 

In these six inscriptions there are ten occurrences of cos in which 
it is interpreted as 'where.' One must notice first of all that it is 
used in every instance in the defining of boundaries. A stream is 
subject of the cbs clause in 1 and 6 cos vdwp pel, 4 cos 6 I,€8appos ; a road is 
subject in 3 and 4 cos d 656s; a ridge is subject in 2 cos 6 (TTe<p(hv TrepLipepei. 
kvkKco, 4 TreptajuTreri^ cos d o-re^dfa. Rivers, roads, and ridges are lines that 
mark a boundary, so that it becomes apparent that 'where' is only 
a makeshift of translation for cbs in each case. The exact meaning 
is 'as the river flows,' 'as the road goes,' 'as the ridge turns.' These 
phrases do not differ essentially from the recurrent cbs v 686s ayei of 
Max Frankel's Inschriften von Pergamon 245 C, lines 20-24, where 
a boundary dispute between Mitylene and Pitane is arbitrated (2d 
century B.C.). The problem of 1 cbs 6 'AKpevs e^iSdXXct is more difficult. 
But a river is subject and we can expect cbs to have about the same 
force as elsewhere in the defining of boundaries. Thus interpreted, 
the clause means literally 'as the Acmeus flows into the Europus,' 
a shorthand expression, it may be, for 'the course of the Acmeus 
as far as its confluence with the Europus.'^ Notice that a line, river 
or road, rather than a point, is the starting place of boundary speci- 
fications in inscriptions 3 and 4 above. 

In the remaining two places, 'where' is not an especially good 
interpretation of cbs in any case. In 3 (line 9) cbs opoi KeLvrai should 
be interpreted 'according as the boundary markers are placed,' like 
the recurring cbs at o-r^Xat Kelvrai of the Pergamene inscription 
(Frankel 245 C), lines 20-25, cited above. In 5 the case is more diffi- 
cult, since the boundary specifications are not clear. But a similar 

» For a topographical study of this inscription see Friedrich Stahlin. "Die 
Grenzen von Meliteia, Pereia, Peumata und Chalai," Athen. Mitt. XXXIX 
(1914), 83-103 and PL VII. 



220 University of California Publications in Classical Philology 

expression in the Pergamene inscription throws some light upon 
this cbs : (line 25) ecos [roO] Taipov tov rpos riji 65cot. The modern Greek 
cjs, a preposition with the accusative meaning 'as far as/ has de- 
veloped from cws. So in the Laodice inscription, I think, we have 
cos = ecos, a conjunction in this case, but leaning toward a preposi- 
tional status: 'until the tomb [is reached] on the right side of the 
road,' that is, 'as far as the tomb.'^ 

Rumpel and Liddell-Scott-Jones also cite Idyll 5. 103 for cbs = 
where. The shepherd cries to his two sheep : 

Tovrei jSoaK-qaelade tot' avroXas cos 6 $dXapos. 

But cbs need not be interpreted as 'where' here; for we can interpret 
'Graze there to the east like Phalarus.' 

There is, then, no good evidence for giving cbs the meaning 
'where' in 1. 13 and 5. 101. But we can hardly understand cbs as a 
preposition here, used as in cbs ^aatXea ; there is no parallel for such 
a usage in ancient Greek literature. Consequently cbs cannot be the 
right reading in this line. 

I suggest therefore that we adopt the reading «, which a good 
manuscript (S) shows for 1. 13,^° and return to are, read by all 
manuscripts in both 1. 13 and 5. 101 (a fact of great importance 
for the rejection of cbs) . The text then reads : 

(1. 12 : . . . Xtjs, atTToXe, Trj8e Kadi^as) 

or (5. 100 : . . . co5e vefieade) 

es TO KO-Tavres tovto yeooXoipov aTe /iupuat 

'Sitting down on yonder sloping hillock where the tamarisks are.' 
'Go to yonder sloping hillock where the tamarisks are and graze 
there.' 

eis is fairly common after verbs of sitting; see, for example, 
Herodotus, 1. 14, . . . tov fiaaLKrjiov Bpbvov es tov TpoKaTL^cov eSuafe. 
The phrase es to KaTavTes is also found in Bion, frag. 12, line 1 
(Stobaeus,FZor. 110. 17): 

AvTap €7cbj' ^aaevnai ejiav 686v es to Karavres 
Trjvo ttotI \J/afiad6v re Kai aiova xl/LdvplaSav 

I have found no parallel for fire (are) in place of simple f}, but 



' See Haussoullier's translation, op. cit. 79, and his map, ibid. 107. 
^^ See Wilamowitz's edition, Praefalio vi, ". . . nee laude sua fraudandi . . . 
HS." 



Fontenrose: Varia Critica 221 

there is nothing surprising about it; evda re, Iva re, 66 1 re, are very 

common. ^^ Probably are was used because airep would have been 

unmetrical, which is used a few lines later in Idyll 1 in the same 

sort of clause : ^ , ^^ 

, , . airep o uwkos 

TTJvos 6 TTOifjievLKds /cat rat 8pves. (1. 22 f.) 

COS is the harder and better-supported reading in the verse under 
discussion, but seems impossible. The mistake of cos for « may have 
come about through the similarity in shape of e to co in cursives or 
uncials; an e sloping too far to the left could be taken for co. The 
evidence shows that the mistake occurred in 5. 101, where cbs is 
found in all the manuscripts; thence it found its way into 1. 13. 
Obviously the pre-Renaissance scholars and scribes, like the earlier 
modern editors, understood cos as the preposition, since they never 
changed are to at re. 

II. Nicolaus of Damascus, 68, sec. 8 (Constantine Porphyrogenitus, 
Excerpta de virtutibus et vitiis e Nic. Dam. 29, p. 347 BW) 

In Nicolaus' version of the famous story of Croesus on the pyre, 
the Sibyl appears upon the scene. Nicolaus contmues, Kat /ler' 
ov TToXv evTovov TL <pdey^afj.kvn |3oa. Then her words are expressed in 
four hexameters. The addenda of the ninth edition of the Liddell- 
Scott-Jones lexicon refer this passage to the metaphorical sense of 
ivTovos, which is given s. v. as "intense, eager, vehement." It makes 
some sense to say that the Sibyl spoke intensely or vehemently, 
but it is hardly a satisfactory sense. 

However, a wholly satisfactory sense is given if we interpret 
IvTovov as equivalent to ev tovco. This phrase is used several times by 
Herodotus in sentences that introduce quoted verses. See Herodo- 
tus, 1. 47. 2, 17 Uvdit] kv e^anerpu! tovw \kyeL raSe. 1. 62. 4, 'An<piKvTOS 
. . . wpoaMV XP9- '^v e^ap-erpco tovui ra.be Xeyuv. 1. 174. 5, rj 8e livdir] a^pi . . . 
xpa kv TpLixerpcii tovco raSe. Nicolaus' evrovov appears to be used in 
the same sort of context for the same purpose. 

All definitions of evrovos given by the lexicon correspond to recog- 
nized meanings of kvTeivoo. The meaning that I wish to give evrovov 
here also corresponds to a recognized meaning of evrelvu}, namely, 
'put into verse.' See Plato, Phaedo 60 d, where Cebes says to Soc- 
rates, T€pl yap TOL Tojv TOLrnjLaTOJV u)v TeTolrjKas evreivas tovs rod AtcrcoTTOu 

" See Iliad A 247, 9 83, I 441; Odyssey a 50, 5 85, e 280, k 417, X 475. 



222 University of California Publications in Classical Philology 

Xoyovs. . . . Plato, Hipparchus 228 d, where Socrates says of Hippar- 
chus, ravra avTos kvrdvas els eXeyelov. . . . Plutarch, Solon 3, varepov 
de Kal yvconas hereive (pCKoabipovs. . . . See also Julian, Or. 6. 188 a. 
Sunilar is Plato, Protagoras 326 b, where evreivo} means 'set to 
music. '^^ 

III. Scholiast on Nicolaus of Damascus, 68, sec. 1 (Constantine 
Porphyrogenitus, Exc. de virt. 29, p. 345 BW) 

On the margin of Codex Turonensis C 980 (11th century) of Con- 
stantine's Excerpta, at the beginning of the extract from Nicolaus 
that contains the story of Croesus on the pjTe, Biittner-Wobst^^ 
reports the following badly written note in a later hand: '0 yap 
Kpolaos ore TpoorevaeLV Kara Kiipov eneXXev epcoTlaavTi tCj 'AiroXXcovt, 
et Tov exdpov xepiecrrai, eiTre' KpoTcros "AXrjv 5ta/3ds fxeyoKriv apx'rjv Kara 
Xifcei" TOVTO rjv ayvoiarov , el're ttju oiKiav tirt tov Ki'poi;. 66ev cnraTLdels 
6 KpoTcros boKwv tov tov avTi.8iK0v apxw kolto. Xvaai, KaraXeKvKe ttjv 
OLKiav apxW) '^'s opdt {vel bpai) iravTcodev {vel KaTwdtvl) . 

Mueller also reports this scholium with corrected text in his first 
note on Nicolaus 68. Besides obviously necessary changes of spell- 
ing he made the significant emendations of irpcorevaeLv Kara to 
cTpaTtvauv KaTOL, and tco 'AxoXXcoi't to tov ' A.iroKKwva. But his changes 
of KaTOL \vaei to KaTarravaei, Kiipov (second occurrence) to exSpov, and 
Kara XDo-at to Ka.Ta\mtiv are unwarranted. 

Yet the first sentence remains awkward ^\dth its abrupt change 
of subject. However, the same scholium appears among Thomas 
Magister's scholia on the Orestes of Euripides (165) in the Codex 
Guelferhytanus, where it has a better text. There we read. Actios 
6 'AttoXXcoj/ Ka\eiTai . . . kol yap toj Kpotcw, oTt crpareiicretv /card 
Kupoi; e^ieXXei', kpwTrjaavTL el tov exOpov Treptecrrat elwe, kt\. The note on 
the Nicolaus extract is obviously part of a larger scholium and was 
carelessly copied into the margin of the existing manuscript. It 
should read, Tw 7dp Kpoicco . . . epoiT-qaavTL tov 'AiroWcjjva, ktX. 

IV. Diodorus Siculus, 4. 70. 4 

The manuscripts read: . . . tCjv Kevravpuv Tavdrjuel o-TpaTevaavTcov 

12 No editor of Nicolaus has any comment on this passage, but Mueller may 
have recognized that evrovov = kv t6vi^; for he translates in F.H .G. Ill, nee 
multo post carmine haec fari coepit. But this may be meant as a free ren- 
dering; notice that the proper force of 0oq. is not brought out. 

" Edition of Constantine Porphyrogenitus, Exc. de virt. (Berlin, Weid- 
mann, 1906), p. 345, n. 19. 



Fontenrose: Varia Critica 223 

€7rt Tovs AairWas Kal ttoWovs avtkbvTUiv tovs vToXeupdevras (pvyetv eis 
^o\6r]v Trjs 'ApKaSlas, tlvcls 5' eis MaXeai' eKTreaovTas hravda KaTOLKrjaaL. 
TOVS 5e Kevravpovs p.€TtoipiadkvTas toXs irpOTeprjixacn Kal opfio^fxevovs eK rrjt 
$oX6rjs Xfi^ecrdai tovs irapLOVTas tojv 'EWr]uo:v Kal ttoXXous tcoj' irepLolKuv 
avaipdv. This text has been retained by editors of Diodorus, though 
they have recognized its difficulties and suspected corruption.!^ For 
the Centaurs appear to be the victors over the Lapiths, contrary 
to the usual story. ^^ Furthermore, though it seems to be the La- 
piths who are driven to Pholoe and Malea, it is the Centaurs who 
use Pholoe as a base for raids upon their neighbors. And the con- 
nection of Centaurs with Pholoe is well attested. ^^ 

Two scholiasts, who obviously used the same handbook as Dio- 
dorus, throw light on this passage. Schohast on Pindar, Pyth. 2. 85, 
. . . TToXXcoj' 5e avaipedevTuv e/c toov KevTavpccv ol -KtpCKeupQkvTts ets 
^oXorjv aireipvyov. Schohast A on Iliad A 263, . . . odev ol AairidaL av<r- 
Tabriv p,ax^cro.iJLevoi dLCOKOvaLV avToiis ds MaXeai' opos Trjs HeXoTTOvvrjaov. 
See also Apollodorus 2. 5. 4. It is apparent from these passages that 
Diodorus' source had the usual tradition. 

But an emendation of twu Kevravpccv . . . avekovTwv, designed to 
make it clear that it was the survivors of the Centaurs that fled to 
Pholoe and Malea, though necessary, is not sufficient because of 
the words p.eTe(jopL(7devTas rots irpoTeprjijLaaL in the next sentence. ^ 
There appears no ground for supposing that this phrase is a cor- 
ruption or later addition. But it obviously cannot refer to a Cen- 
taur victory over the Lapiths; it can refer only to the Centaurs' 
successes after their irruption into the Peloponnesus. Therefore we 
must suppose the omission of a sentence or two between KaTOLKrjcraL 
and TOVS 8e KevTavpovs that refer to the conquest of the Pholoe and 
Malea regions by the Centaurs. ^^ 

" See the latest edition, that of C. H. Oldfather, Vol. Ill (London and 
Cambridge, Mass., L.C.L., 1939). Some editors have changed ^o\6riv to ^eveov 
and TLvas 5' to reXos 5', but these changes do not affect the problem that I 
discuss 

15 A Centaur victory in the Lapith-Centaur war is attested by Lactantius 
Placidus on Statins, Theb. 2. 563. But this scholium seems based upon a mis- 
understanding of Virgil, Aen. 7. 304 f.: Mars perdere gentem/immanem 
Lapithum valuit. See also Statins, Theb. 7. 203-205. See my "Ares and a 
Lapith Defeat," P.A.P.A. LXVII (1936), xxxv f., and s. v. Peirithoos, 
Pauly-Wissowa-Kroll, R.-E. XIX, 117 f . 

'^ See Apollodorus 2. 5. 4. 

" See my articles, v. Peirithoos, loc.cit. (seen. 15 above), 118. 

18 An objection might be made on the ground that tovs 5i Ktvravpovs implies 
a change of subject. But the subject of the immediately preceding sentence m 
the assumed lacuna could have been other than the Centaurs. Or tovs Kev- 
ravpovs may have been inserted after the corruption occurred in an attempt 
to make the passage clearer. 



224 University of California Publications in Classical Philology 

In the first sentence the sense is right if ttoXKovs aveKovrcov is 
emended to ttoXXwi' apaLpedevrcov, but such an error would be hard to 
explain. It seems better to accept the phrase as it is and assume the 
omission of a few words between Aairldas and Kal, words that we 
can restore from the scholia that I have quoted. I therefore suggest: 
. . . Tuv Kepravpcov Travbrjiid aTparevadvTcov ewi tovs Aairldas, {tccv 8e 
KaTrlduv crvcrTaSrjp fxax^o-anevcav) /cat iroWovs avekovroiv, tovs viroKenf^evras 
(pvyelv . . . kvTavda KaroLKrjaaL. { lac. ) tovs be KevTavpovs, kt\. 

V. Scholiast on Pindar, Pyth. 2. 85 

In his edition of the scholia Vetera on Pindar (Leipzig, Teubner, 
1910) A. B. Drachmann keeps the manuscript reading of the fol- 
lowing sentence, which occurs in a narration of the Lapith-Centaur 
brawl at Peirithous' wedding feast: Kal ourws ol KePTavpoL Kepaa- 
dePTes TTJs T<jiV KpaTTipcop KaTO. TTiv eucoxtcij' aTTOTTPolas aladofxepoL ov KaTkcrxov 
iavTovs Ttjs kixcpvTOv fiaplas. August Boeckh omitted KepaadkvTes in his 
edition (Leipzig, Gottlob Weigel, 1819), but Drachmann admits it 
without question. To me, at least, KepaadePTes is unintelligible as a 
modifier of ol KePTavpoL. Certainly the word should modify KpaTrjpwv; 
see Odyssey y 393, cr 423, KprjTfjpa KepaaaaTo. To emend to KepaaBkPTcov 
we must assume a displacement of ttjs from the position imme- 
diately after KkpTavpoL ; the assumption, moreover, accounts for the 
corruption of KepaadePToop to KepaadePTes. I therefore propose: Kal 
ovTCOs ol KePTavpoL ttjs KepaaOePTWP twp KpaTrjpup Kara ttjv evuiXLO-v 
aTTOTTPolas aladbp^epoL, kt\., 'smelling the exhalation from the mixing 
of the wine bowls at the banquet.' 

It is no objection to this emendation that in the next sentence 
the Centaurs are said to have filled themselves with oIpos aKparos. 
Either the phrase is used loosely, or the Centaurs were supposed to 
have laid hold of the wine that had not yet been diluted. 



SENECA'S DIALOGI III, IV, Y 

DE IRA LIBRI TRES 

THE TEXT EMENDED AND EXPLAINED 

BY 
WILLIAM HAEDY ALEXANDER 



III, 1, 1 (46 : 4)' : ceteris enim aliquid quieti plaeidique inest, hie totus 
c'on'citatus et in impetii doloris est, armorum sanguinis suppliciorum 
minime humana fiirens cupiditate, etc. 

The reading here of the hand a, which filled in the page fol. 12'' left 
blank in A, though with imperfect juncture to the page next following, 
is : in impetu doloris est. The reading of L P is : in impetu est doloris. 
Bourgery' adopts the second of these readings, but the rhj-thmical argu- 
ment favors the first, a hypodochmiac clausula (Axelson, type 4).' The 
placing, however, by W. Gemoll of a comma after est in the L P reading 
(1) produces an excellent clausula rhythm, -tatus et in impetu est (Axel- 
son, type 2a), (2) leaves m impetu as a clear parallel to concitatus, (3) 
gives with the four words following est a sort of interior assonance (-is, 
-orum, -is, -orum) of a kind that has been noted elsewhere in Seneca. 
I find doloris intrusive where it is made, as in a, to occur in the general 
description of the emotion of anger as contrasted with the other emo- 
tions ; quieti, placidi, concitatus, in impetu sounds like the proper series. 

Ill, 1, 7 (48: 5) : neque enim ulla vehementior intrat concitatio, quae 

nihil moveat in vultu. 

While the concitatio of Cornelissen and Gertz is an excellent word (cf . 
concitatus, III, 1, 1) , I think it may be assumed as practically a certainty 
that the ms. intra cogitatio is an erroneous transcription of a genuine 
intrat agitatio. Anyone who has worked with Lombard minuscules will 
recall the very great similarity of appearance between co and a. 

As for intrat to replace the intra of a, I am led to accept it with Gertz,* 
Hermes, and Barriera.^ Bourgery on the contrary retains it. Intrat agi- 

^ The page and line reference is to Hermes' edition of the Dialogi (Leipzig, Teub- 
ner, 1905). The consecutive Dialogue numbering III, IV, V is employed rather than 
the numbering of the books De Ira I, II, III. 

= A. Bourgery, Seneqiic, Dialogues, Tome V, Be Ira (Paris, Societe d'Edition "Les 
Belles Lettres," 1922), p. 2. 

^ All such references are based on the classification of Senecan clausulae as given 
on p. 23, fn. 35, of B. Axelson's Neue Senecastudien (Lund, C. W. K. Gleerup, 1934). 

* M. c'. Gertz, L. Annaei Senecae dialogorum libros xii rec. (Copenhagen, in Lil3raria 
Gyldend'aliana, 1886), app. crit. ad loc, p. 52. The reason there advanced is not con- 

5 A. Barrier a, L. Annaei Senecae de Ira ad Novatum libri ires (Turin, Paravia e 
Ciaf, i919),p. 104. 

[225] 



226 University of California Publications in Classical Fhilology 

tatio produces an excellent clausula rhythm (Axelson, type 2a), and 
this fact, other things being equal, secures the preference for it as against 
intra agitatio and intrat concitatio. 

Ill, 2, 2 (48: 19) : alium ira in cubili suo confodit, alium intra sacra 

mensae iura percussit, alium intra leges celebrisque spectaculum fori 

lancinavit, etc. 

The whole description in the third of the alium sentences above is 
genuinely Silver Latin, even more in the manner of Tacitus than of 
Seneca. The translators are not happy in their handling of it; render 
thus : "anger tore yet another to pieces in the very court over which he 
presided in full view of a crowded forum." This corresponds with the 
account given by Valerius Maximus (IX, 7) of the death of the praetor 
Asellio, of which a yet more detailed version occurs in Appian's Civil 
Wars I, 6, 54. 

In 11. 23-24 the reading of a, namely, in cruces membra diffindere, is 
to be retained with Bourgery. Cruces is in the plural because the writer 
is thinking of the two limbs of the cross, already having in mind the verb 
diffindere. That verb is a repulsively powerful word beside which the 
suggested replacements seem feeble, and in any event diffindere is con- 
firmed by the cliff undere of L. 

Ill, 3, 2 (49 : 23-50 : 2) : deinde nemo tam humilis est, qui poenam vel 
summi hominis sperare non possit : ad nocendum potentes sumus. 
The concluding remark must be considered sound in its present form, 
offering as it does a perfect double cretic clausula. Though it seems over- 
condensed for clarity, it must be remembered that this is a natural and 
not infrequently assumed risk attaching to the "pointed" style. Seneca 
may be presumed to be reflecting on the infinite and universal capacity 
of the human race for doing harm, with the implication for him who will 
take it that we are nothing like so resolved or effective for doing good ; 
"we can be sure of our power for doing hurt." It must be with the in- 
tention of bringing out some such idea that a number of the translators 
introduce an "all" or an "always" into their rendering of the sentence. 
This expresses the universality of our power for evil to which reference 
was made above, and may be justified in a translation. Seneca, however, 
left it to inference and did not formally express it. 

Ill, 3, 4 (50: 7) : sed dicendum est feras ira carere et omnia animalia 

praeter hominem. 

Bourgery and Barriera are right in following the ms. reading and 
rejecting Vahlen's animalia." The context makes omnia sufficiently clear 

" However, as Rarricra (p. 107) points out, what is credited by Hermes to Valilen 
as a conjecture had already l)een brought forward a generation earlier by Fickert in 
his note on the passage. In the vast field of accumulated conjectural readings one is 
often apt to claim as his own something which has been long anticipated; but he may 
partially console himself with the corroborative value of his conjecture if not its 
originality. 



Alexander: Seneca's Dialogi III, IV, V, De Ira Lihri Tres 227 

and we should not demand meticulons verbal accuracy provided the 
point at issue is not left obscure. 

Ill, 6, 2 med. (52 : 28) : proximum est, ut modus proficiat; si modus et 

ordo non profieit, subducit aliqua et circumcidit ; si ne adhoc quidem 

respondet, interdicit cibis, etc. 

For adhoc in its variant form adhiic compare section 3 immediately 
following : qua moneat adhuc et exprobret. Section 3 is the application 
in the field of the ruler and judge of the figure of the doctor and his 
principles of treatment in section 2. The parallelism of the two sections 
is to some degree an argument for the soundness of adhoc in section 2. 
Adhuc (adhoc) means in both places "still." 

The subject of respondet is the treatment indicated in the words suh- 
ducit aliqua et circumcidit, and corresponds to our own idiom with the 
verb ''answer." It is not necessary to be more explicit in regard to the 
subject as Madvig, for example, proposed to be by reading hoc for adhoc.'' 

Ill, 6, 4 (53 : 18) : sed ut documentum omnium sint, et quia vivi nolu- 
erunt prodesse, morte certe eorum res publica utatur. 
The mss. ALP (we have now passed from the page supplied by a and 
are back with the Ambrosian in the A hand) read: et qui alicui. The 
quia vivi in Hermes' edition is an emendation of Haupt's. Barriera, app. 
crit. ad loc, proposes qui alioqui. In both of these suggestions an effort 
is made to get a definitely stated contrast for morte certe, but there would 
appear to be no absolute necessity for this ; perhaps Seneca might pref - 
erably so have written, but he appears not to have done so. 

We should with Bourgery retain the ms. reading. No doubt cuiquam 
would conform better to the grammatical rule for the use of the general 
indefinite pronoun in a negative sentence, but qui cuiquam runs into 
a cacophony. The use of the relative clause to precede while the so-called 
antecedent follows in the main sentence is far too good an idiom to reject. 

Ill, 8, 4 (55 : 18) : utrum ergo ita ut nihil faciant eorum quae ira dictat 

an ut aliquid 1 si nihil f aciunt, apparet, etc. 

The ut nihil faciant alternative is at once followed up in the next sen- 
tence by si nihil f aciunt ; there is no specific, verbally marked resumption 
of the other alternative an ut aliquid (faciant) , and Haase' thought that 
this indicated a lacuna here. The best way to test this is to translate 
carefully from si nihil f aciunt on, with comment. 

"If they do nothing of the thing's that anger dictates, it is obvious that 

'Barriera, op. cit., p. 109: sed cum cuiusvis generis verborum omissio frequens 
apud Senecam, turn pronominum frequentissima. Compare chapter 8, section 7, where 
the subject of quievit is left unstated but must surely be ira. There again misplaced 
ingenuity (W. Gemoll's in that instance) would alter quievit to qiiies fuit. 

"" Fr. liaase, L. Annaei Senecae Opera quae siipersunt rec. (Leipzig, Teubner, 1874), 
Vol. 3, praefatio, p. xxiii. 



228 University of California Fuhlications in Classical Philology 

anger is not necessary for the accomplishment of actions, and yet you 
were advocating it [anger] on the ground that it possessed something 
stronger than reason." They do nothing of the things that anger dictates, 
and obviously reason has entirely prevailed ; it cannot be said that such 
persons have shown ira at all. Thus the first alternative is completely 
disposed of and it becomes appropriate to discuss the second. 

But the second involves the idea of "something or other" of what anger 
dictates, and thus anger is not entirely excluded here as in the first alter- 
native discussed above. It thus becomes important in the discussion of 
the second alternative to assess the relative capacities of reason and 
anger. Thus the question: valentior est quam ratio an infirmior (ira), 
is really the equivalent of stating the premise si aliquul faciunt. Let us 
resume the translation. 

"If anger is the more powerful, how can reason impose a limit on it, 
since it is only the weaker things {m this world) that habitually submit 
(to others) and take orders (from them) ? If anger is weaker, reason 
is sufficient without it to achieve things and has no need of the help of 
the weaker agent." This was obviously the conclusion to be expected if 
the premise si aliquid faciunt had been formally expressed, and as it 
is also the answer to the question valentior est (ira) quam ratio an in- 
firmior, we must conclude that it is in the question that we find the 
equivalent of the unexpressed premise which worried Haase. 

As for denique Johann Mueller" seems to describe its function cor- 
rectly as being "um das blose Wortgefecht abzubrechen und zur sach- 
lichen Beweisfuehrung ueberzugehen." He compares its use in X, 7, 3 
and X, 11, 1 "um den Uebergang von der weniger bedeutsamen Eroer- 
terung zu der entscheidenden anzuzeigen." It does not in such cases 
mark the advent of the last of several items in a catalogue, all of which 
have been stated, but a jump to a conclusion with omission of several 
steps. 

Ill, 10, 1 (56 : 30) : quos (sc. impetus) numquam comprimere possit (sc. 

ratio), nisi pares illis similisque opposuerit [ut irae metum, inertiae 

iram, timori cupiditatem] . 

The words enclosed in square brackets by Hermes, after Gertz, are to 
be retained with Bourgery and Barriera. The very fact that Gertz can 
. bring exact logic out of them by a little shifting of the words is in itself 
the best argument that the text is correct and Seneca perhaps slightly 
illogical, as often. Yet not very much so; irae metum is almost imme- 
diately repeated in the form quietus (esse) nisi timet (non potest), and 
inertiae iram by fortis esse nisi irascitur non potest. Timori cupiditatem 
is less exactly reproduced by industrius (esse) nisi cupit (non potest). 

» Johann Mueller, "Kritisclie Studien zu den kleinercn Schriften des Philosophen 
Seneca," in Sitznnfjsberichle der philosophi.'ich-historischen Classe der l-aiscrlichen 
ATcademie der Wis.senscliaften, Vol. 118 (Vienna, 1889), p. 13. 



Alexander: Seneca's Dialogi III, IV, V, De Ira Libri Tres 229 

It must be repeated that the lack of complete identity as between the 
challenged phrases and the following sentences in which the thought of 
the phrases is developed more fully, is a good argument for the soundness 
of the ms. reading in its entirety. Especially so, it seems to me, where 
the less accurately stated side of the general identity comes first and 
subsequently reaches accurately stated form when the bare phrases are 
developed into sentences. 

Ill, 14, 3 (61 : 21) : errantem per agros ignorantia viae melius est ad 

rectum iter admovere quam expellere. 

The reading of L P rectum iter is to be maintained with Hermes, Bar- 
riera, and Castiglioni" against Bourgery's affedatum iter. The affectum 
of A arose in the first place through an over-tall r being mistaken for an 
/; the ad fectum iter was subsequently written affectum iter. 

Ill, 15, 1 (61 : 27) : at corrigi nequeunt nihilque in illis lene aut spei 

bonae capax est. 

In Hermes' edition this should be printed within quotation marks. 
It is an objection of the interlocutor, as Bourgery clearly indicates by 
his method of punctuation."^ Basore gives it to the interlocutor in his 
translation but not in the Latin text.^ 

The final est looks intrusive. Without it we have a satisfactory clausula 
(Axelson, type 4) ; with it retained the clausula conforms to no estab- 
lished type. The omission of the copula in Seneca is a commonplace, 
though the extent of its omission tends to be obscured by persistent 
editorial emendation." 

Ill, 15, 3 (62 : 8) : nil minus quam irasci punientem decet, cum eo magis 
ad emendationem poena proficiat, si iudicio lata est. 
The more than doubtful Latinity of poena lata est has given rise to 
various proposals for emendation. There is also the serious difficulty 
that iudicio lata est introduces the banned rhythm of the caput versus 
heroici, which would be very conspicuous at the sentence's end. 

With regard to the Latinity of the passage, it seems likely enough that 
the idea which was in Seneca's mind is expressed by Koch's si iudicio 
lata est <sententia>, but that in the expression of that idea the parti- 

^^ L. Castiglioni, "Studi intorno a Seneca prosatore e filosofo," Bivista di Filologia, 
LII (1924),p. 381. 

The Bourgery and Barriera references cited hereafter will be understood as being 
in apparatu critico ad locum; page references will not usually be given. 

" Namely, the use of dashes, e.g., — At corrigi . . . capax est. — Admittedly it is not 
always easy to determine with finality where Seneca ends and the interlocutor begins. 

^ J. W. Basore, Seneca, The Moral Essays (Loeb Classical Library, 1928), Vol. I, 
p. 113. 

1^ This point has been well established by Dr. B. L. Charney, Bracliylogy in the 
Epistulae Morales of Seneca (unpublished Ph.D. thesis. University of California, 
1940). 



230 University of California Fuhlications in Classical Philology 

ciple which would have exactly fitted sententia has been applied loosely 
to poena. The punishment and the sentence are, in one sense, substan- 
tially the same thing. 

As for the rhythm, a satisfactory clausula (Axelson, type 1) is ob- 
tained if we write iudici lata est, dropping the o of the mss. (iudicio 
lata in L, iudici olata in A) . 8i iudici lata est means "if it has been fixed 
by a judge," with "judge" understood in the widest sense of the term 
as "a person of judicial temperament," a common enough metaphorical 
sense of the word in Latin. Or it may be that iudici is generic and 
that the phrase signifies "if the penalty has been fixed by the jury" ; 
this may be assumed to be that method of determination of guilt and 
penalty into which anger least enters. Iudicio of the mss. was an easy 
error when once the phrase iudici lata est was misunderstood. 

I am not impressed by the data of some of the inferiores or by the 
appeal to the poenas lege fatorum datas of Seneca's Thyestes, 74. It 
seems a weak reed to lean on ; at best it would be a single instance, and in 
fact the meaning of datas in the Thyestes phrase is certainly "assigned." 
Further, the power there touched on is not judicial but only the blind 
operation of fate. 

Ill, 16, 2 (62 : 21) : 'tu adhuc in prima parte, etc' 

In this extended account (some seventeen lines) of the process of 
making the punishment fit the crime, Hermes, following Pincianus, has 
inserted <tion> before contineheris, thus causing the words ignominia 
<inon'> contineheris to be the statement of the rejection of the penalty 
of ignominia in order to replace it with relegation and exile. It seems 
to me, however, as to Lipsius, that there are six forms of punishment all 
told (obiurgatio privata, obiurgatio publica, ignominia, esilium, career, 
mors) and that Pincianus and Hermes are mistaken in confusing igno- 
minia with obiurgatio. The latter is a reproof in words ; ignominia points 
to definite punishment of a concrete kind, such as reduction in rank. 
Each one of the six forms is introduced by some form of the pronoun 
tu, the te of line 22 serving for both forms of obiurgatio. The punctua- 
tion therefore following contineheris should be a period instead of a 
comma as in Hermes' edition." 

.Ill, 16, 5 (63: 22) : et cum cervicem noxio imperabo praecidi et cum 
parricidas insuam culleo et cum mittam in supplicium militare et cum 
Tarpeio proditorem hostemve publicum imponam, sine ira eo vultu 
animoque ero quo serpentes et animalia venenata percutio. 

Parricidas above is a reading from the inferiores for per (conven- 
tionally abbreviated) indices of A; why it has been thought necessary 
to supplant A's sound words with parricidas I am, like Barriera, at a 

^' Barriera says : ad priorcm interpunctionera redibo, but fails to do so in his Latin 
text, p. 18. 



Alexander : Seneca's Dialogi III, IV, V, Be Ira Lihri Tres 231 

loss to understand. Cum per indices insuam culleo means "when I give 
orders, based on evidence, for sewing up (the criminal) in a sack"; the 
object noun is to be supplied from 7ioxio in the preceding clause. It 
should be noted that if the text of A is followed, there is no object 
actually expressed for mitfam in the next cum clause. If the question is 
raised why only here (that is, in the insuam clause), of the four cases 
dealt with, witnesses are mentioned, the reply is that the punishment 
is so cruel and unusual that Seneca agrees that nothing must be left to 
chance here, that a charge must be confirmed here if anywhere by sound 
evidence. 

To touch on the matter of paleography, the acute remark of Barriera 
(p. 121) deserves reproduction here: ex verbis quae secuntur insuam 
culleo facilius fuit parricidas, quamvis in scriptura corruptum, quam 
per indices. In other words, in view of the well-known form of punish- 
ment for parricides, if the phrase insuam culleo occurs with some as- 
sumed corruption preceding it in the text, parricidas would be the easy 
correction, not per indices. The whole incident throws light on what may 
be perpetrated as a sound text even in the best editions, and constitutes 
a warning against any serious aberration from the ms. tradition except 
under the direst necessity, and even then with plenty of warning to 
readers of what it is proposed to do with the ms. text. 

Ill, 17, 7 (65 : 17) : ergo non paria patiuntur qui paria commiserant, et 
saepe qui minus commisit plus patitur, quia recentiori obiectus est. 

The reading of A is recentior, and this may very well be allowed to 
stand. The point of the sentence is equally good whichever reading is 
adopted, whether one speaks of the anger as being fresher {recentiori) 
or of the person on defense as being presented among the early cases on 
the list {recentior) and thus that much nearer to the original level of 
the anger involved. The reading recentior ohiectus est gives, with con- 
sonantalization of the i in recentior, two perfect cretics, while recentiori 
ohjectus est gives a molossus and a cretic. This is a place where the 
clausula rhythm may be allowed to decide in favor of the original A 
reading. 

Ill, 18, 3 (66 : 7) : Cn. Piso fuit memoria nostra vir a multis vitiis inte- 
ger, sed pravus et cui placebat pro constantia rigor. 

The word pravus has excited all kinds of suspicions but without ac- 
tually furnishing grounds for any. We are informed first that Gnaeus 
Piso was a man free from many (not all) faults. Next we learn that 
there was a kink in his nature, an oddity, a perversion, and consequently 
he is described as pravus. Finally the nature of his quirk is revealed by 
an appended relative sentence after an et. It would certainly have been 
better style to write ut cui placer et in explanation of pravus; Schultess 



232 University of California Puhlications in Classical Philology 

saw this when lie replaced et cut with eo quod. But Seneca merely ap- 
pended the explanation paratactically, and there is nothing to be done 
about it. There is no real reason for supposing the text unsound, and 
probably many so-called difficult passages are to be explained in this 
same way, in other words, by refusing to believe that Seneca must always 
have written logically, grammatically, and in good taste. 

Ill, 20, 3^ (69 : 1) : ita ira muliebre .... 'oderint dum metuant.' 

Barriera and Bourgery rightly regard the words quid ergo? non ali- 
quae voces ah iratis emittuntur quae magno emissae videantur animof 
as being part of Seneca's reply to 'at incidit et in vivos/ while Hermes' 
printing of the passage assigns them to the interlocutor. The preceding 
sentence nam . . . ingenia sunt was merely a preliminary clever retort 
to 'at incidit et in vivos' ; the real reply begins with quid evgo? and runs 
through to 'odevint dum metuant.' Madvig's immo is properly rejected 
by Barriera and Bourgery; it is a conjecture based on Madvig's failure 
to understand the correct assignment of parts in the dialogue. Barriera 
(p. 126) justly emphasizes the subjunctive in quae magno emissae vi- 
deantuv animo; quae . . . videanfuv means "of a kind to seem," a phrase 
entirely appropriate to Seneca's line of argument and point of view, but 
not to those of his interlocutor. 

Hermes' edition should therefore be corrected by ( 1 ) striking out the 
quotation marks around quid evgo? . . . animo; (2) by dropping immo; 
(3) by dropping the interrogation mark after animo and replacing it 
with a comma, (4) by placing an interrogation mark after 'metuant.' 

III, 20, 8 (70: 2) : C. Caesar, qui iratus caelo . . . quod comessatio sua 
fulminibus terreretur — prorsus parum certis — ad pugnam vocavit 
lovem. 

There is no sense of omen, I think, about the parenthesized words as 
Bourgery would make out by his translation; "presages trop peu surs" 
is reading too much into the Latin. The words are literal, and Basore 
is right in rendering them "surely they missed their mark," though I 
think it would be a little livelier to say "to be sure, i\\ey were decidedly 
off the target !" Had they only been better aimed the world would then 
and there have been rid of Caligula. 

IV, 1, 1 (72 : 1) : primus liber, Novate, benigniorem habuit materiam; 
facilis enim in proclivia vitiorum decursus est. 

Several proposals have been made for emending the second half of the 
sentence ; none is necessary. The idea is that while vice has been soundly 
castigated in Book I (Dialogue III), the castigation was rather easy 
(henigniov), as easy as the swift descent (decursus) of an army launch- 
ing its attack from the heights on an enemy occupying the slopes below. 



Alexander: Seneca's Bialogi III, IV, V, Be Ira Libri Tres 233 

The diatribist of Book I has an easier task to perform than the construc- 
tive philosopher who is now opening Book II. Thus the translation is 
as follows : "The first book, Novatus, provided more pleasing material ; 
the downward rush upon the slopes occupied by the vices is easy." This 
will be found to compare favorably with the somewhat vague renderings 
of most of the translators.^^ 

IV, 1, 1 (72 : 3) : quaerimus enim ira utrum iudicio an inpetu incipiat, 
id est utrum sua sponte moveatur an quemadmodum pleraque, quae 
intra nos non insciis nobis oriuntur. 

Hermes inserted non "ne in utroque quaestionis membro idem dicere- 
tur." He appears not to have realized that sponte corresponds with iudi- 
cio, and that therefore his inserted non brings about exactly the result 
he aims to avoid. It would perhaps be pleasant to delete with Gertz the 
sua before sponte, but if there is carelessness, it would seem likely that 
it is chargeable rather to Seneca himself than to any scribe. Barriera 
refuses to admit that there is anything slovenly about the sua, and feels 
that it takes the sua along with the sponte to make the phrase the equiva- 
lent of iudicio. 

TV, 3, 5 (75 : 3) : ergo prima ilia agitatio animi, quam species iniuriae 
incussit, non magis ira est quam ipsa iniuriae species. 

The view of the translators is substantially that of Basore : "therefore 
that primary disturbance of the mind which is excited by the impression 
of injury is no more anger than the impression of injury is itself auger.'"" 
It is very tempting to feel with Stephanus and, three centuries later, 
Karsten, that the sense should be : "that first mental agitation which the 
impression of a wrong (suffered) excites, is no more anger than the 
impression of the wi'ong (suffered) is the wrong," and to make this 
explicit by adding iniuria est; Barriera, holding the same view, reads 
ipsa species iniuriae iniuria (se. est). It is an ingenious parallelism, 
but there is nothing to indicate that Seneca thought of it. Apparently 
both the prima agitatio and the iniuriae species are in his mind related 
to ira as the focusing point, and the statement is that the first is no 
more the actuality of anger than is the second. The clausula is sound 
(Axelson, type la), resolution taking place in the first long of the 
terminal spondee). 

^= The fact is that to prepare an attractive and at the same time sound translation 
is not a thing that can be come by easily or rapidly. My sympathies are ■v^'ith the 
translator ^vho ninst get along -with his task, especially as I am sure that he often 
knows full well that his translation evades the issue or fails to make sense. As a trans- 
lator he simply has not the time to find out by the slow processes of scholarship what 
a sentence really does mean. 

1° Op. cit., Voi. I, p. 173 



234 University of California Publications in Classical Fhilology 

IV, 1, 3 (78 : 4) : et iudex damnaturus quae fecit eligitur et corona pro 

mala causa . . . bona patroni voce corrupta. 

There is certainly a verb missing in the latter half of this sentence. 
Attempts have been made to supply it as follows : (1) at %>ro (Bourgery, 
coll. IV, 5, 5 : proclamat; Hermes : prohat with mala causa carried into 
the accusative); (2) after causa (Gertz: stat; Barriera: est) ; (3) out 
of hona (Karsten: hovat or boat), but bona is essential to establish the 
mala causa, hona voce contrast; (4) at the sentence's end (Schultess: 
corrumpitur) . Of all these Schultess' proposal is the best because (1) 
corrumpitur written corrupif is liable to accidents; (2) corrumpitur 
is the perfect parallel in form and position and assonance to eligitur; 
(3) the resultant clausula rhythm is a sound double cretic. 

IV, 9, 3 (79 : 15) : non descripsit (sc. Ovidius) castra ex imo partu con- 

traria. 

The unanimous ms. reading is ex una parte; the ex uno partu above is 
a conjecture of Petschenig. There is also Koch's ex uno corpore, and 
ex una patria has been suggested by others. 

It is simpler and perhaps also wiser to see whether the ms. reading 
cannot be explained. Bourgery translates "ces camps ou s'affrontent 
ceux qui devraient etre du meme cote." I think, however, that the mean- 
ing is rather "hostile camps arising out of the one party." Normally 
"hostile camps" implies two parties, two factions, two belligerents ; the 
point here is that the hostile camps emerge from a cleavage in the one 
party. This accords with the general sense of the passage, which deals 
as a whole with the perversion of natural relationships and the break- 
down of human confidences. Pars can be used in the singular as well as 
in the plural to express a political party, and Seneca was here influ- 
enced by the considerations of rhythm. When j^artc is used we have a 
double cretic, whereas partihus would be unrhythmical. This is not abso- 
lutely fatal to Petschenig's conjecture, but the double cretic is better 
than the molossus cretic, and this is of weight on top of other arguments. 

IV, 9, 3 (79 : 23) : (non descripsit sc. poeta) . . . regnorum publicorum- 

que exitiorum clandestina consilia. 

The object phrase here is not altogether easy to interpret. The reason 
4s that the genitives attached to consilia are left a little vague and must 
have their significance developed in the sympathetic mind of the reader. 
I suppose that consilium regni means "a plot for setting up a monarchy," 
and consilium puhlici exitii similarly means "a plot for public ruin"; 
this again must indicate "a scheme for the destruction of popular gov- 
ernment." Purtjior, the fact that the two genitives connected by -que are 
placed in the particular oi-dor they occupy is probably designed to sug- 



Alexander : Seneca's Dialogi III, IV, V, De Ira Libri Tres 235 

gest a cause-and-effect relationship. This will produce as a final render- 
ing : "secret schemes and plots for setting up monarchies on the ruins 
of popular governments."" 

IV, 10, 1 (80 : 8) : quid, si illis irasci velis, qui, quod aegrotant senescunt, 

f atigantur ? 

The qui is added by Hermes, who regards aegrotant senescunt as a 
bimembris dissolutio. In his view these two verbs belong to the quod 
clause, while f atigantur is the predicate of the inserted qui. 

In my judgment Vahlen's ullis for illis remains the best solution of 
the difficulty presented by the passage. Up to this point we have had 
mention of specific cases {in tenehris parum vestigia certa ponentibus, 
surdis, pueris) ; now we are going to deal with any persons whomsoever 
(ullis) who through sickness, old age, or exhaustion slow things up and 
so produce anger in the employer. This makes of the three verbs at the 
end of the sentence a trimembris dissolutio. 

IV, 11, 2 (82 : 7) : deinde non ideo quaedam, quia sunt terribiliora, po- 

tiora sunt, nee hoc sapienti adici velim : 'quod ferae, sapientis quoque 

telum est, timeri.' 

The reading of A is : nehoc sapientia did velim, where we should of 
course take Barriera's adjustment ne hoc sapienti adici velim; he makes 
of this an independent sentence, placing a period after potiora sunt. 
For adici cf . IV, 19, 4 non multum sibi adiciens; the sense is "attach." I 
differ from Barriera in the matter of punctuation. I think the ne . . . 
velim should remain attached to the statement preceding even though 
it does not depend immediately upon the idea conveyed by that state- 
ment but rather on an intermediate implied notion such as "and this fact 
prevents me." It can readily be seen that if the ne . . . velim 'quod . . . 
timeri' were omitted, the sentence following would continue the argu- 
ment in perfect sequence on potiora sunt, but Seneca must get in the 
point about the fera and the sapiens before it goes out of his head, even 
if it tangles up the sentence a little. 

I accept the punctuation of Johann Mueller in regard to the words 
quod . . . timeri whereby he makes them a sentence in oratio recta ex- 
planatory of /iOC." 

IV, 11, 4 (82: 26) : non est ergo quare concupiscat quisquam sapiens 
timeri, nee ideo iram magnum quiddam putet, quia formidini est, 
quoniam quidem etiam contemptissima timentur ut venena t et ossa 
pestifera et morsus. 
In the American Journal of Philology, LIV, p. 357, 1 maintained that 

the reference in the words from ut on was to objects used in the practice 

" See Barriera's excellent note ad loc. 

*^ In the Vienna Academy paper already referred to in fn. 9, p. 228. 



236 University of California Publications in Classical PJiilology 

of magic, and proposed to change morsus to versus in the sense of "incan- 
tations" in order to bring all the items into line. As against those who, 
like Haupt, Gertz, and Hermes, see a serpent in all this (hardly con- 
temptissimus!), or like Barriera, who detects even the serpent's tooth, 
I still maintain that the reference is throughout to objects employed in 
magic rites, things ridiculous enough in themselves but rendered terrible 
by their supposed connection with calamitous results ; but I now see that 
the word morsus is sound and thus too the whole text precisely as it 

stands. 

The venena and the ossa pestifera are commonplaces of witches' rites. 
A reference to Horace, Epodes 5, 17-24, which is an account of the col- 
lection of materials for Canidia's ghoulish sacrifice, shows venena first, 
and then (line 23) ossa ah ore rapta ieiunae canis, "bones snatched from 
the jaws of a starveling bitch.'" Bones of that sort are pestiferous in the 
sense explained by Festus (ed. Mueller, p. 210 and 245), "prophetic in 
a deadly way," "pointing to death or exile." As these "starveling bitches" 
prowled around cemeteries, the ossa pestifera would presumably be por- 
tions of the legs or arms of corpses uprooted by the dogs from their 
shallow burial places. 

There remains et morsus, and this, the most puzzling of all, is satis- 
factorily accounted for by reference to Apuleius, Mctamorplioses 2, 
21-22, where it is explained that in Thessaly it is the regular habit of 
witches to gnaw {demorsitant) the faces of the dead to procure one of 
the ingredients they require for their magical proceedings (ea sunt illis 
artis magicae supplementa), and they assume if necessary the likeness 
of birds, dogs, and mice, or even of flies to achieve their purpose. The 
curious practice of hiring watchers to prevent this desecration of the 
corpse is detailed, with the interesting information that if the least por- 
tion of the corpse is filched the watcher must replace it from his own 
flesh. It thus appears that pieces of flesh bitten (morsus) from corpses 
are a regular part of the material of witch-sacrifice; these morsus, like 
the other things, are contemptible enough except in their association 
with witches and witch ceremonies. 

Something remains to be said about the clausula. If elision of the final 
vowel of pestifera takes place, we are faced with the forbidden caput 
versus heroici ( -^^ | — | - 1 1 ) .'" It appears, however that in respect of 
clausula rhythm there is some freedom in the matter of elision and 
hiatus,'" and if hiatus is here tolerated as between pestifera and et. our 
sentence-end is a clausula of the Axelson la type. I think there is some 
justification for admitting the hiatus because venena et ossa pestifera 
provides an unchallengeable clausula, a proper sentence-end, to which 
et morsus is added as an afterthought. It is the element of hesitation 

" H. Keil, Grammatici Latini, Vol. VI (Leipzig:, Toubiier, 1874), 628, 28. 
™ See fn. 3, p. 225. 



Alexander: Seneca's Dialogi III, IV, V, De Ira Libri Tres 237 

before this afterthouglit is grasped and uttered that excuses the hiatus 
in question. 

Alternatively it is possible that the words ossa . . . morsiis constitute 
the last foot of a hexameter plus the first half of a second hexameter or of 
an elegiac pentameter, and that they should be printed to indicate both 
that fact and also that they are a quotation, not subject to ordinary clau- 
sula considerations. 

IV, 12, 1 (83 : 14) : primum potest aliquis non algere, quamvis [ex rerum 
natura] hiemps sit, et non aestuare quamvis menses aestivi sint. 
The phrase ex rerum natura is entirely sound ; I concur in the views 
of Bourgery and Barriera to that effect. The sentence translates thus : 
"to begin with, it may well be that some individual does not feel cold, 
though in the natural order of events [i.e., the run of the seasons] it is 
winter." In contrast to "the natural order of events" we have below the 
special circumstances {loci heneficium, patientia corpom)'which may 
nullify the operation of that natural order. It is a mistake to destroy this 
effective contrast by striking out ex rerum natura. The occurrence of 
de rerum natura in the preceding sentence does not of itself constitute 
any valid objection to the presence of ex rerum natura in this, especially 
as the "speakers" are different for the two sentences; there is no repe- 
tition or near-repetition by the same speaker.'^ 

IV, 12, 2 (83: 18) : deinde verte istud: necesse est prius virtutem ex 

animo tollas quam iracundiam recipias. 

Istud must refer to the interlocutor's remark with which the chapter 
opens : nequitia de rerum natura tollenda est si velis iram tollere ; "wick- 
edness must needs be expelled from the universe if you want to expel 
anger." Now let us, in obedience to the verb verte, change practically all 
the significant words here to their opposites, thus : nequitia to virtus, 
de rerum natura to ex animo, tollere (second occurrence) to recipere, 
and we get the new sentence : virtus ex animo tollenda est si velis iram 
recipere, and that is substantially the sentence which in the text follows 
deinde verte istud. Thus the new sentence in the form given it in the 
text is approximately the upside-down version of the sentence with 
which the chapter started, and this is what Seneca designed when he 
wrote verte istud. Therefore Gertz is wrong in his attempt to exchange 
the positions of virtutem and iracundiam. 

The foregoing is a conclusion based purely on a verbal examination 
of the text ; for a philosophical refutation of Gertz's view the critique 
by Johann Mueller should be consulted." 

=^Cf. too IV, 27, 2: nos enim non causa mundo sumus hiemem aestatemque refe- 
rendi, where mundus is the same general concept as the natura rerum of the passage 
here under discussion. 

-- As previously referred to, fn. 9, p. 228. 



238 University of California PubJicafions in Classical Philology 

IV, 15,1 (86: 13) : quaedam enim non nisi melioribus innascuntur in- 

geniis, sicut valida arbusta laeta quamvis neclecta tellus creat, et alta 

f ecundi soli silva est. 

Alta, tlie reading of L, is certain as against the alia of A. Apart from 
that the question is one of reading valida arhusta et laeta, or, with Gertz, 
dropping the et. Definitely laeta must go with tellus; laeta tellus is 
balanced by f ecundi soli following, and we require both laeta and f ecundi 
in the phrases of the figurative part of the sentence to correspond with 
the meliora ingenia of the main proposition. These meliora ingenia are 
the rich soil of the higher type of character, needing cultivation to pro- 
duce the best results, but even in their natural state capable of develop- 
ing fine growth. 

Bourgery retains the et and translates : "de meme une terre riche, 
quoique negligee, produit une vegetation vivace et luxuriante," using 
the word laeta twice, which has the effect of concealing the difficulty 
about the et. The conjunction must have been inserted at some point in 
the tradition by a scribe who did not realize that laeta belonged with 
tellus. I agree with Gertz in rejecting it. 

Farther on I would place a semicolon after exurgunt; the second sed 
(line 20), which troubles Barriera, is much more tolerable if a sharper 
break is made in the sentence and indicated by appropriate punctuation. 

IV, 19, 1 (88 : 22) : etlocorum itaque et animalium et corporum et morum 
vari'etates mixtura elementorum facit, et proinde aliquo magis iucum- 
bunt ingenia, prout alicuius elementi maior vis abundavit. 

Aliquo is Madvig's emendation for the in aliquos of the ms. tradition ; 
I agree with Barriera in retaining the ms. reading. The interpretation 
appears to turn on the meaning of the verb incumhunt; I regard it as 
meaning "settle upon, fasten upon." Ingeniame^w^ "types of character," 
and the sentence from iiroinde on translates as follows : "consequently, 
upon this individual and on that types of character settle the more in 
proportion as a greater potency of this and that element is markedly 
present." It will be noticed that the in aliquos is responded to directly 
by the alicuius of the jJrout clause. 

IV, 19, 4 (89 : 10) : siccis aetatibus vehemens robustaque est ira, sed sine 

incremento. 

Gertz denies the possibility that siccis aetatibus is sound Latin. Pre- 
sumably according to his view we should have either simply sicciorihus 
to correspond with quihus umidi plus est (line 6), or in the singular 
sicciori aetati, ".the more sober period of life." But I cannot see any 
objection to the ms. reading with the sense "the soberer periods of life." 
since there might very well be visualized several periods of the mature 



Alexander : Seneca's Dialogi III, IV, V, Be Ira Libri Tres 239 

life between forty and sixty. This is evidently the view taken by Basore^ 
as well as myself. 

IV, 19, 5 (89 : 16) : vinum iucendit iras, quia calorem auget ; pro ciiius- 
que natura quidam ebrii effervescunt, quidam saucii. 

Lipsius recognized that his own interpretation of saucii as leviter tacti 
flore Liberi was really inadequate and unjustified,'* and finally postu- 
lated a lacuna after the second quidam. There is a difficulty in the sen- 
tence, as Lipsius saw, though the moderns, for the most part, have failed 
to see it, but the remedy lies elsewhere than in his suggestion and involves 
no change in the text nor any assumption of a gap. 

In the first place pro cuiusque natura belongs not with quidam ehrii, 
etc., but with the preceding words ; "wine kindles angers because it in- 
creases the (bodily) heat in proportion to the natural constitution of 
each." It will be observed that -usque mitura provides an equally good 
clausula. The statement that wine kindles angers because it increases 
heat is a generalizing statement, and it is with a generalizing statement 
that a phrase like pro cuiusque naiura belongs. 

Then the sentence continues with a note on this generalizing state- 
ment: "some people" (not all) "fairly boil when they are drunk, and 
so too some people when they are wounded." The substance of the last 
remark has, of course, nothing to do with the generalizing statement 
about wine kindling anger because it increases heat, but it is added as 
a corroborative phenomenon. "Getting drunk makes some people boil 
(with rage) ; so too (by the way) does getting wounded." Every football 
coach has men on his team whom he is positively glad to see tackled hard 
by the opposing team because he knows that the rough usage will kindle 
a heat of rage in them that will make them formidable ground gainers 
from that time on. In the Latin text no word is required to establish the 
comparison ; that is done by the parallelism of quidam with quidam. The 
clausula here is a molossus cretic. 

IV, 20, 4 (90 : 15) : et quia aliis contra iram, aliis contra tristitiam reme- 
diis utendum est nee dissimillimis tantum ista sed contrariis curanda 
sunt, semper ei f occurremus quod increverit. 

I have not altered my view as set out in the American Journal of 
Philology, LIV, p. 357, regarding the illogical character of the passage 
as it stands. "Since the remedies required for the one are absolutely dif- 
ferent from those required for the other, we shall always apply our 
efforts to resist the one which has grown faster" is logically incomplete 
and requires to complete it the added phrase "by the means appropriate 
to it." I formerly proposed < ut visum, erit > to follow increverit, 

^ Op. cit., Vol. I, p. 206, fn. h : "we may think here of the middle periods of life." 
-^ Quoted by Barriera, op. cit., p. 142. 



240 University of California Puhlications in Classical Philology 

but < ut lonuni visum erit > would be preferable to secure the familiar 
double eretic termination. 

I may say frankly, however, that calling attention to Seneca's illog- 
icality and suggesting how it might be remedied is one thing and 
definitely proposing to make an addition to the text quite another. In 
the present case I am strongly of the opinion that the mss. give us the 
original reading, which, if illogical, is readily understood. The obelus 
should be removed from the text and no change of any kind made. 

IV, 22, 1 (92 : 18) : sed haec ad liberos nostros pertinent ; in nobis quidem 

sors nascendi et educatio nee vitii locum nee iam praecepti habet : 

sequentia ordinanda sunt. 

The 171 nohis part of the sentence I regard as being a somewhat arti- 
ficial arrangement for the following: in nobis quidem sors nascendi 
locum vitii non habet, neque educatio iam praecepti locum. The meaning 
or sors nascendi is ''our lot in being born (a generation before liheri 
nostri) " ; that lot has brought us to maturity and there should be no 
place in maturity for the vice (of anger). Secondly, no longer at our 
age does our process of getting educated permit us to be trained by 
maxim, like our children. That kind of training is a slow business, and 
our time of life does not allow us to await its lingering reactions; all that 
follows for us (in our future education) must be presented not by maxim 
but in logical sequence. 

Sequentia is complete as it stands despite Wesenberg's citation of IV, 
18, 1 ; it means "the residuum of the subject, the topics that follow." 

IV, 23, 3 (93: 18) : dignus fuit qui innocentem haberet, dignus qui 

faceret. 

How should the last three words be completed in sense, by credendo 
or by convalescendo, to use the suggestions of Lipsius' note f "" I should 
certainly choose the former, influenced not a little by the credidit of the 
preceding sentence, and interpret faceret in the sense "prove, demon- 
strate." Alexander deserved to prove Philip innocent by putting full 
confidence in him. I cannot accept the peculiar ethical twist given by 
Barriera to his explanation of the passage. The last clause is distinctly 
the climax. The king deserved to find his physician innocent, — yes, but 
something more than that; he deserved the glory of giving the demon- 
stration of that innocence by placing full confidence in him. 

IV, 28, 4 (97 : 9) : utique bonis ne irascamur . . . minime diis ; non enim 

illorum vi, sed lege mortalitatis patimur quidquid inconimodi aocidit. 

'at morbi doloresque incurrunt.' utique aliquo defungendum est domi- 

cilium putre sortitis. 

= See Barriera, op. cit., p. 145. Lipsius did not, of course, propose that either of his 
suggestions for implementing the sense should be added to the text. 



Alexander : Seneca's Dialogi III, IV, V, De Ira Libri Tres 241 

I agree witli W, Baehrens and Bourgery in finding no sufficient reason 
for inserting vi. 

Aliquo defimgendum is Madvig's emendation for aliquod (aliqiio: L) 
fugiendum of A. Bourgery places Madvig's emendation in his Latin text, 
but in writing in liis translation "il nous f audra deloger d'une f aeon ou 
d'une autre" he appears to be rendering the A text."** 

While not denying the ingenuity of Madvig's suggestion, I believe 
that the L reading, which is of course substantially confirmed by A, is 
to be retained. The philosopher has asserted that the troubles and ills 
of our life are due to the law that governs our mortal existence. The 
interlocutor objects : "Yes, but diseases and sorrows pour in upon us," 
with the implication of their becoming intolerable. Now in Seneca's 
reply the words domicilium putre sortitis do not, as is usually thought, 
refer to all humanity, but only to that part of it which has been unlucky 
enough in the lottery of life to have drawn a domicilium puire. Shall 
they be denied the privilege of being angry at the afflictions their destiny 
brings them? Seneca's statement on that situation is: "Certainly those 
who have happened to obtain by the lottery of life a tumble-down dwell- 
ing, ought to escape from it in some direction or other." It is the old 
familiar recommendation of the Stoic philosophy, and the advice given 
and the figure under which it is given are highly condensed and are 
actually fused together in the short sentence we are discussing. Compare 
on the whole subject of suicide as an escape V, 15, 3^. 

The exegesis is as follows : "In a country like ours, where earthquakes 
happen, people who live in a tumble-down house must, in event of 
trouble, get out of the building, flee from it somewhere or other. So in 
life, for those of us who have been unfortunate enough to be born physi- 
cally handicapped, there must be, if the situation becomes intolerable, 
a way of escape effected in some direction or other, that is, by using this 
means or that. In short, if you are one of the congenitally unfortunate 
and your resultant pains and diseases become intolerable, there is open 
for you the gate of suicide ; walk out." This theme is often repeated in 
Seneca's works. 

IV, 32, 1 (101 : 19) : 'At enim ira habet aliquam voluptatem et dulce est 
dolorem reddere.' minime : non enim ut in beneficiis honestum est 
merita meritis repensare, ita iniurias iniuriis. illic vinci turpe est, hie 
vineere. inhumanum verbum est et quidem pro iusto receptum [ultio 
et] talio. non multum differt . . . nisi ordine : qui dolorem regerit tan- 
tum excusatius peccat. 

Basore seems to me to have formed the correct idea of the meaning 
of this passage when he writes : "in the code of the XII Tables ialio is 

^ This type of error is rather common in Bourgery's -n-ork and must be carefully 
guarded against by anyone using his edition. 



242 University of California Piiblications in Classical Philology 

the Mosaic 'an eye for an eye,' and ultio {dolor em regerere) says Seneca, 
is merely a more excusable form of this savage law.'"" Accordingly I be- 
lieve that the original text must have stood thus : inhumanum . . . ultio, 
et talio<ine^ non multum differt nisi or dine. The meaning is: "an 
uncivilized word, and yet for all that accepted as a just word, is revenge, 
and it does not much differ from the primitive law of retaliation except 
in degree." It is not hard to guess how ultio et talione became ultio et 
talio, but the remedy does not lie in rejecting ultio; both it and talio are 
necessary for the development of the idea. 

IV, 32, 2 (101 : 25) : M. Catonem [ignorans] inbalineo quidam percussit 

inprudens ; quis enim illi sciens f aceret iniuriam ? 

In his second edition (Basel: 1529) Erasmus excised ignorans on 
the ground that it was a gloss on inprudens. This marginal gloss subse- 
quently worked itself into the text. But there seems to be no reason why 
the sense of the passage should not be this : "not knowing Marcus Cato, 
someone struck him in the public bath, carelessly."'"* This turns out to 
be peculiarly appropriate when we read a little later on that, finding 
out whom he had struck, he sought to make amends to Cato ; but it was 
because he was Cato, and the offender would never have bothered his 
head about what had happened had the victim of his anger been anyone 
else than Cato. 

Further, the quia . . . iniuriam sentence immediately following is a 
much better comment on ignorans than on imprudcns ; also the offender's 
reward (in place of punishment) is said (102: 5) to have been coepit 
Catonem nosse, the opposite of ignorans. 

IV, 33, 6 (103 : 13) : contempsisses Homanum patrem si sibi timuisset; 

nunc iram compescuit pietas. dignus fuit cui permitteretur a convivio 

ad ossa fill legenda discedere, etc. 

The ms. readings are contempsissct and permitteret. Contempsisses 
was derived by Hermes from the Bipontine edition and won the approval 
of Gertz. The impossibility of contempsisset is well explained by Johann 
Mueller, and he also expresses correctly the needs of the passage when 
he writes : "Der Satz ist erst daiin zurechtgeriickt und erhiilt erst dann 
eine rechte Spitze . . . wenn er ein Urtheil des Sehriftstellers Seneca 
enthalt.""^ Various ways have been suggested for emending the locus, 
all presuming some lacuna and most of them rather clumsy and improb- 
able. The shift, however, to the second person is very easy paleographi- 
cally, and the third person is readily explained by n preceding siccaret 
and a following timuisset; moreover, the reading of the second person 

" Op. cit., Vol. I, p. 236, fn. 

-^ Or vnth more weight on inprudens: "quite unconcerned (about possible conse- 
quences)." 

=» Mueller, op. cit., pp. lC-17. 



Alexander: Seneca's Dialogi III, IV, V, De Ira Libri Tres 243 

also meets Mueller's requirement (quoted above) by the very simple 
device of enunciating the author's judgment through transferring it to 
the "ideal observer." 

I think that permitteret can probably be retained. The reader is ex- 
pected to have the proper subject in mind; the author, who was hasten- 
ing forward to permisit . . . adulescens, certainly had. Bourgery retains 
permitteret but gives it a translation with an indefinite subject ; this 
I feel is an error. 

IV, 35, 1 (104 : 26) : numquid velit quisquam tam graviter hostem ferire, 
ut relinquat manum in vulnere et se ab ictu revocare non possit ? atqui 
tale ira telum est : vix retrahitur. arma nobis expedita prospicimus, 
gladium commodum et habilem ; non vitabimus impetus animi . . . hos 
graves et onerosos et inrevocabiles ? 

There are three questions of text arising in the last sentence : (1) so 
far as the original text of A can be determined under the superimposed 
A", it appears to have exhibited animo rather than animi; P, however, 
shows d'l, i.e. animi: (2) the A text is graves funerosos, the latter word 
being without parallel in the Latin known to us: (3) the hos before 
graves of A^ appears as his in L and hiis in P. 

Let us first get the general trend of the passage. "Would anyone want 
to strike an enemy so hard as to leave his hand in the wound and be 
unable to withdraw himself from the blow? Yet anger is that kind of 
weapon, scarcely retractable; we (on the contrary) are looking for 
weapons that will be light for us to handle, a convenient and manageable 
sword." Anger, the most violent impetus animi, is, according to this 
description, a kind of weapon which, once launched on its stroke, can 
hardly be withdrawn from the wound it creates ; we who are cultivating 
the philosophic approach to life and its problems will not desire that 
sort of weapon, but something more like the rapier of the fencer or, in 
any event, a light and manageable blade, not a bludgeon-like broad- 
sword. Following naturally upon this one writes : "shall we not avoid 
the passions f {non for nonne as constantly in Silver Latin usage) . I be- 
lieve therefore that animi is the correct reading. 

As for hos graves, it is my view that this is a limiting phrase to the 
general expression impetus animi, "bursts of passion, the kind that are 
grievous." Gertz's istos for hos shows clearly that he accepts the idea of 
a limiting phrase, but the change of pronoun is unfortunate. With the 
hos the philosopher-author admits that he too is human like the rest of 
us ; such bursts of passion are grievous to him too. I therefore read hos 
graves. 

Though the funerosos of A seems substantiated by the funosos of L 
and the funeros of P, it must be remembered that if the error is a suffi- 
ciently early one, it would antedate A, L, and P alike; in that case 



244 University of California Piiblicatio7is in Classical Philology 

funosos and funeros only confirm the erroneous character of funerosos. 
As inrevocabilcs means metaphorically "incapable of being drawn back," 
as in Pliny, N.H. 32, 1, 1, 2 ancorae pondere inrevocabilcs, I cannot help 
thinking that there is much to be said for Koch's onerosos; what has 
been said by Pliny of the anchors would be appropriate here of a clumsy 
and unwieldy weapon, and Pliny says specifically pondere inrevocabilcs. 
I read accordingly onerosos for funerosos, regarding the latter as one of 
those curious "psychological" errors which occur in copying where con- 
siderable sections of two words are alike ; it may even, as here, lead to 
the creation of a "new" word without standing. I would say further here 
that I do not accept Vahlen's et inserted between graves and onerosos; 
the substantive with its closely attached limiting modifier is impetus 
animi hos graves, while its descriptive modifiers are onerosos and inre- 
vocal)ilesJ'° 

The clausula inrevocahiles seems to require a word. In view of the se 
revocare non possit and the vix retrahitur preceding, the word must be 
considered sound. One must suppose that if it falls within the rhythmic 
canon, we have a molossus here plus an iamb, with the middle syllable 
of the molossus resolved; this seems to approximate Axelson's type 4.'^ 
It is unwise, of course, to attempt to force every clausula into rhythmic 
form. 

IV, 36, 6 med. (107: 24) : avaritiam, durissimum malum minimeque 
flexibile, ira calcavit adacta opes suas spargere et domui rebusque in 
unum conlatis inicere ignem. 

I am unable to follow Madvig, Gertz, and Barriera in reading adactam 
against the adacta of A. Hermes' remark, app. crit. ad loc, viz. : "non 
proho (adactam) propter ea quae sequuntur et domui . . . ignem," may 
be expanded as follows. Had the sentence ended with spargere the read- 
ing adactam would be perfectly sound, as it would be proper to speak 
of avarice as having been compelled to scatter her accumulated wealth. 
But when the opes suas spargere is more specifically defined by domui 
rehusque in unum conlatis inicere ignem we recognize the marks of in- 
sanity springing from anger. This swings the balance back to adacta; 
its meaning is not "compelled" (as adactam would have to signify if 
modifying avaritiam which is under the heel of ira and thus subject to 
.her compulsion) but "brought to the situation" of scattering its wealth 
and making a single grand bonfire of the whole thing. Adacta is defi- 
nitely right. 

V, 1, 2 (108 : 9) : consilium pro moribus euiusque capiendum erit ; quos- 
dam enim preces vincunt, quidam insultant instantque summissis, 
quos terr endo placabimus. 

™ The furiosos attributed by Hermes, app. crit. ad loc, to J. Haas is to be found 
at least two centuries earlier in the small Elzevir edition of Seneca (Leyden, 1649). 
^^ Inretrahibiles would have worked out more smoothly. 



Alexander: Seneca's Dialogi III, IV, V, Be Ira Libri Ires 245 

The reading of the mss. is quosdam terrendo, altered by Gertz to quos 
conterrendo, by Hermes to quos terrendo as above. Bourgery clings, 
mistakenly, to the ms. reading, claiming that we have here a tricolou, 
while what we obviously have is (1) a statement of a first method for 
subduing anger, namely, by preces, (2) a statement of the attitude of 
some angry people toward the humble-minded (summissi), and an ex- 
planation of how such angry people are to be dealt with, viz., terrendo 
(or some compound of it) . The second of these is oddly stated; the atti- 
tude of bullies toward the meek and humble is elevated to the position 
of the principal sentence and the method of treating these bullies be- 
comes a relative sentence attached thereto. 

The syllable -dam in the quosdam of the mss. preceding terrendo may 
be simply an error arising out of a recollection of quosdam, quidam 
which have already occurred. Barriera regards it as a misreading of 
a contraction for demum; I am not clear in what sense he would regard 
that adverb as being used. I venture the suggestion that the -dmn is an 
erroneous writing of a de- which became separated from terrendo and 
attached to quos. It seems to me that it is questionable whether you could 
terrify (terreo) the class of people in question, but you might frighten 
them off (deterreo) and thus quiet them by diverting their energies 
elsewhere. For deterreo cf. IV, 36, 3 init., deterritum ab ira. 

V, 2, 2 (109 : 23) : denique cetera singulos corripiunt, his unus adfectus 
est, qui interdum publice coneipitur. numquam populus universus f e- 
minae amore flagravit, nee in pecuniam aut lucrum tota civitas spem 
suam misit ; ambitio viritim singulos occupat : inpotentia una est ma- 
lum publicum. 

Una is Vahlen's correction for oion of A. Barriera on the basis of a 
single ms., the Angelicus, which, according to his report, sometimes uses 
fi for forms of noster, reads nostri; this is highly improbable. 

Inpotentia undoubtedly means here, as often elsewhere, inpotentia 
sui, "anger." AVe have had statements of three things which cannot be 
called mala puhlica; it would now seem appropriate to have a statement 
about the thing which, by way of contrast, is notoriously and definitely 
a public evil. Yet instead of a statement we could conceivably, and very 
effectively too, have a question ; it would stand out vividly after three 
affirmations. Punctuate therefore with a comma after the word inpo- 
tentia, and treat non as signifying nonne, a Silver Latin commonplace. 
Translate : "anger (however), is it not a public mischief?" The answer 
comes immediately : saepe in iram uno agmine itum est.''' 

V, 2, 4 (110 : 4) : totae cum stirpe omni crematae domus et modo eloquio 
favorabili habitus in multo honore iram suae contionis excepit. 
The first sentence totae . . . domus seems certainly to belong with what 

precedes ; hence Matthias has placed a period after domus, removed the 
^- The suggestion is Dr. Charney's. See Preface. 



246 University of California Publications in Classical Philology 

et following it, and placed an et before totae (this last not, however, 
being essential) . AVhen it comes to the punctuation and arrangement of 
an adjoining sentence or so, it is frequently very hard to say how Seneca 
would have read the sentences in question, and Matthias may be right. 

Favorabilis is the ms. reading, favorabili an emendation of Madvig's. 
Barriera retains the form favorabilis, regarding hahitus in multo honore 
as a gloss on it which has found its way into the text. I agree in reading 
favorabilis but not in the view taken of habitus, etc. 

The sentence requires division after habitus; the in multo honore be- 
longs with what follows. Translate : "he who but recently was favorably 
regarded by reason of his oratory, in the midst of his great repute has 
come to feel the anger of an audience that was once his own." It will 
be observed that I take contio in its literal sense, and this agrees with 
the concrete legiones . . . plebs . . . publicum consilium which follow. 

V, 2, 6 (110: 17) : hie barbaris forte ruentibus in bella exitus est . . . 

ruinae mode legionibus incidunt inconpositi, interriti, incauti, peri- 

cula adpetentes sua. 

The word interriti has excited criticism as being inappropriate in a 
series beginning with inconpositi and ending with incauti. Hence inpa- 
rati (Koch), intecti (Gertz), inperiti (Wolters), intermixti, written 
interihti (Barriera), while Bourgery frankly admits that he cannot 
think what word is concealed under interriti. 

I think that interrupti is a reasonable suggestion. A battle line that 
presents an uneven front, usually as a result of attack but conceivably 
from other causes as here, can be described as interrupta, cf. Livy, 40, 
40 and 44, 41, and Caesar B.C. 1, 64. The paleographical possibility, 
especially if the p was first lost, is good.''' 

V, 6, 2 (115 : 3) : quis impetu turbidus et in aliquem mens non quidquid 
. in se venerandi habuit abiecit? 

Venerandi is Gertz's emendation for the ms. reading verecundi; of 
the latter Bourgery writes "vix recipi potest" though he prints it in his 
text unobelized. The reason is that we have here, presumably, a list of 
six effects of anger ; the first, namely, the destruction of the sense of 
restraint, verecundia, has been disposed of and does not need restate- 
.ment. Therefore verecundi must conceal the second effect. 

Since in the first section of this chapter the character of the man not 
affected by anger lias been described bj^ the words modestus et venera- 
bilis et dispositus, there is every reason to suppose that Gertz's venerandi 
is the proper correction. The psychologj^ of the mistake is simple. The 

•■" It may be, of course, and Dr. Charney takes that view, that interriti is simply 
a rhetorical filler,' thrown in for <jood niensure by Seneca and not lofjically related to 
the context. Johann ATueller in the article so often referred to, pp. 19-20, has collected 
some instances of this practice. I cannot accept all his citations, but some are certain. 



Alexander : Seneca's Dialogi III, IV, V, De Ira Libri Tres 247 

scribe lias just previouslj^ copied verecundiam; he arrives at a word 
beginning- with ve- and seen at the first sweeping glance to contain -nd- 
as well, and slips easily into verecundi for venerandi. 

V, 8, 7-8 (117: 32) : difficiles quoque et indomiti natura blandientem 

f erent : nihil asperum territumque palpanti est. quotiens disputatio 

longior et pngnacior erit, in prima resistamus, antequam <robur 

accipiat> : alit se ipsa contentio et demissos altius tenet. 

In the first half-sentence the A L reading is : difficiljs . . . feret, and 

this text prevailed till Gertz's time, when he corrected it by altering the 

first word to difficiles and the last to f erent. Personally I concur with 

Waltz in the view that the vulgate indomita is preferable; a writer who 

had just written the syllable mi might almost unconsciously alter ta to 

ti in the syllable next following. 

If territum is sound, and certainly none of the suggested replacements 
thus far is convincing, Seneca is probably combining here physical re- 
action (asperum) and psychical reaction [territum).^'' 

In the quotiens sentence I see no reason to read with L rohitr accipiat. 
The words do not occur in A, and there is no syntactical difficulty, as 
Gertz admits,^' in the use of the present indicative alit after postquam, 
though in Seneca it is more usual to find the present subjunctive. Fur- 
ther, the sense obtained is good, and the trouble in regard to in prima 
is cleared up when the phrase is brought into close contact with alit se 
ipsa contentio. In prima means in prima contentione, and the sentence 
translates thus : "Whenever a discussion will be [i.e., gives signs of being] 
somewhat protracted and bellicose, let us stop at the first sign of strife 
before it magnifies itself and holds fast those who have plunged in too 
deep." There is no necessity for adopting Klammer's <i>a> after in 
prima. 

V, 12, 2 (120: 31) : aliquid aetas peccantis facit, aliquid fortuna, ut 
f erre ac pati aut humanum sit aut non humile. 

The non is Madvig's ; Bourgery prefers to reach the same conclusion 
by certe hand, presumably on paleographic grounds. Basore and Bar- 
riera stand by the ms. reading in respect of aut humile, and Basore also 
keeps the mss. aut between ferre and imti. I follow Basore in retaining 
the ms. reading throughout. I cannot agree with Barriera that fortuna 
points to a person better off in wealth or resources than oneself; it 
means "situation in life," being in that respect as indefinite in limits 
as is aetas. 

As to aut between fcrre and pati, it must be retained to exhibit a dif- 

^* The suggestion is Dr. Charney's. I suggest hesitatingly tentum, "tensed," for 
territum. 

^ M. C. Gertz, Stvdia Critica in L. A. Senecae Dialogos (Copenhagen, Libraria 
Gyldendalia, 1884), p. 98. 



248 University of California Publications in Classical Philology 

f erence between the two attitudes here set forth. In ferre there is active 
volition, while imti implies acceptance of a mechanical order. It is con- 
siderations relating to the age of the offender or to his station in life 
which make the endurance or, it may be, just the passive toleration, 
either a humane act or a deferential act. Thus if the offender is young 
or poor and you overlook the injury (ferre or pati), you act as a huma- 
nus, but if he is of mature age or otherwise quite able to meet responsi- 
bility and you overlook the injury, you have acted as a hnmilis. In other 
words, in drawing the line between an overlooking of injury that is 
humanum and an overlooking that is humile, considerations of the 
offender's age and station in life are determinative ; there is no absolute. 

This view of the meaning of the passage is supported by what follows. 
"Let us change places with the offender. What is it that now makes us 
angry? His unfair appraisal of us." Evidently, then, the distinction 
between our being Jmmanus and our being humilis in our toleration of 
an injury rests upon a fair appraisal of the circumstances of the doer 
of the injury, circumstances of age and of social and economic station. 

My translation is as follows : "The age of the offender counts for some- 
thing, and so too his station in life, so that to tolerate an injury or to 
submit more passively to it is either an act of humanity or one of defer- 
ence." I fail to see why Basore inserts "merely" in his version ; it is not 
only superfluous but actually damaging to the sense of the passage. 

V, 12, 6 (121: 19) : velut stupens gestum ilium saevituri deformem 
sapienti viro servabat, oblitus iam servi, quia alium quern potius cas- 
tigaret invenerat. itaque abstulit sibi in suos potestatem et ob pecca- 
tum quoddam commotior : *tu,' inquit, 'Speusippe, ser\Tilum istum 
verberibus obiurga ; nam ego irascor.' 

In this passage the phrase oh peccatum quoddam will, I think, appear 
peculiar on close reading. What is the peccatum quoddam to which ref- 
erence is made? Is it the same peccatum we have read about above 
through the whole of §5 ? I think not, because (1) if it were, Speusippus 
would have been mentioned at once in line 17 where the indefinite phrase 
ah amico, qui forte intervenerat now occurs; (2) ohlitus servi in line 20 
suggests that as far as that slave is concerned the incident is over; Plato 
is not going to ask anyone to beat the slave for him in that particular 
case ; (3) the point of such anecdotes is usually reached in the phrase or 
sentence used by the chief character and, with the point thus made, the 
story is over. Such a phrase in the first story is cxigo pocnas ah homine 
iracundo. This is commented on in a single sentence following and then 
begins the second story. 

LTnder the circumstances, I incline to read quondam for the ms. quod- 
dam: the clausula effect remains the same and the sense is clarified, 
while the paleographical situation is excellent (qu6dam>quodam> 



Alexander: Seneca's Dialogi III, IV, V, De Ira Libri Ires 249 

quoddam). It is worth remarking that Basore in his Loeb translation 
ad loc. inserts "once" though translating quoddam as well by "once." 
He evidently had an instinctive feeling toward the correction I have 
suggested. 

V, 13, 1 (122 : 1) : pugna tecum ipse! si vis vincere iram, non potest te 
ilia, incipis vincere si apsconditur, si illi exitus non datur. 
A reads : si vincere iram non potest, te ilia incipit vincere. si apscon- 
ditur, etc. The second t of potest was, however, erased, and P has potes. 
The insertion of vis is due to Gertz, while IMadvig had already corrected 
the ms. incipit to ificipis. The latter change seems necessary', but not the 
former. 

I read the words of A in the following form : si vincere iram non potes 
totam, illam incipis vincere si apsconditur. My totam is derived from 
the residual t te in the ms. reading as it originally stood ; for the point 
of sense raised cf . V, 42, 1 : ex toto removeamus. The t of totam became, 
in my judgment, erroneously attached to potes; o was lost by accident 
or removed by design. Then torn, written td, was changed in the mount- 
ing confusion to te, and that is why incipis became incipit. My reading 
of the passage establishes a sharp contrast in totam . . . incipis; "if you 
cannot conquer anger outright, yet you make a beginning of conquering 
it if it is concealed." Non potes totam is also a good clausula.""^ 

V, 18, 4 (128: 12) : deinde adeo inpatiens fuit differendae voluptatis, 
quam ingentem crudelitas eius sine dilatione poscebat, ut in xysto 
maternorum hortorum, qui porticum a ripa separat, inambulans quos- 
dam ex illis cum matronis atque aliis senatoribus ad lucernam decol- 
laret. 

The proposals that have been made for the transposition of the phrase 
cum . . . senatoribus to this place or that go on the assumption that the 
phrase belongs with inambulans. Before anything of that kind is at- 
tempted it is wiser to ask what the phrase means where it stands, and if 
it has a reasonable meaning there, to let it alone. Its association, as our 
ms. tradition reads, is with decollaret, and the meaning is "decapitated 
them by lantern light while in the company of ladies and other senators." 
It seems to me that we are merely facing here the ever-recurrent problem 
raised by the lack of a present participle for esse in Latin, and that 
further speculation on what to do about the order of this sentence is 
useless. 

V, 21, 2 (130 : 28) : ibi unus ex iis equis, qui trahere regium currum albi 
solebant, abreptus vehementer commovit regem : iuravit itaque se 
amnem ilium regis comitatus auferentem eo redacturum, etc. 
The word comitatus is sound, though it is very unusual in the plural. 

Cf . however Sallust, Catiline 45, 1 : ut in ponte Mulvio per insidias Allo- 

^^ Barriera (p. 166) also finds the heart of the difficulty in the te, which he regards 
as an abbreviation for tempore. The resultant contrast is less effective. 



250 University of California FiMicaiions in Classical Philology 

brogum comitatus deprehendant. Gudeman at that point makes this 
comment: "i.e. the Allobroges and all those who accompanied them, 
hence the plural abstract which is very rare with this word, unless it 
refers to different occasions.'"' It is in Gudeman's last phrase that the 
explanation of the Senecan passage is to be found. Cyrus represents the 
river as carrying away, not an escort, but escorts of a king, thus magnify- 
ing the incident by his manner of speaking of it, a common form of exag- 
geration. Thus a man through whose window a neighbor's small boy has 
thrown a stone, is likely to go to the lad's father and insist that he stop 
his son from throwing stones through his windows. The word commeatus 
which has been proposed misses the point entirely. 

V, 30, 2 (139 : 15) : carissumis enim irascimur, quod minora nobis prae- 
stiterint quam mente conceperimus quamque alii tulerint. 
The three futures perfect of the quod clause are the readings of L. A 
varies in the second of them by reading concepimus, but if we recall that 
the only difference between the two forms is the presence or the absence 
of a crossbar on the upright of the letter jh ^e shaU not be unduly im- 
pressed with A's concepimus, and it was a bad point for Madvig to pro- 
ceed from in changing tulerint to tulerunt. 

My view is that the three futures perfect of the quod clause set up 
a condition which will have been completed to insure the conclusion 
irascimur. This conclusion might have been expressed as a future, but 
of course it continues to be true, and its continuance is indicated by the 
situational present irascimur. The quod with the futures perfect seems 
odd, but si would not so appear, and the only difference is that what si 
would indicate as a condition^ quod identifies as an actual fact prior 
to that of the main verb in time. I am convinced with Hermes that L 
gives us the correct reading here, and indeed, as explained above, A 
itself comes within one crossbar of giving the same text. 

V, 31, 1 (140 : 6) : nulli ad aliena respicienti sua placent : inde diis quoque 
irascimur, quod aliquis nos antecedat, obliti quantum hominum retro 
sit, et panels invidentem quantum sequatur a tergo ingentis invidiae. 
Johann Mueller's emendation ingruentis for ingentis conforms sty- 
listically and rhythmically to all applicable canons, but one is left to 
suspect that it does not convey what Seneca meant. The ingentis is a 
<;ompressed statement; the thought is quantum sequatur a tergo invidiae, 
nam ingens est. A stylistic turn is achieved and a good clausula rhythm 
obtained by the condensation of nam ingens est into a single epithet 
attached to invidiae, however odd the combination may seem at first 
sight to the modern mind. 

V, 32, 2 (141 : 4) : in hoe enim praecipue fallimur : ad ferrum venimus, 
ad eapitalia supplicia. et vineulis, careere, fame vindicamus rem casti- 
"'A. Gudeman, C. SaUusti Crispi Bellum CatiUnae (New York, Applcton, 1904), 

p. 112. 



Alexander : Seneca's Dialogi III, IV, V, De Ira Lihri Ires 251 

gaudani llagris levioribus. 'quo modo' inquis 'nos iubes intueri quam 
omnia per quae laedi videamur, exigua, misera, puerilia sint?' 
I suspect, on the basis of clausula rhythm, that the words flagris and 
levioribus have exchanged places. 

As for the words of the interlocutor quo modo . . . sint, we have to 
choose in interpretation between an ironical explanation ("How have 
you the assurance to bid us discover, etc.") and a belief that the words 
are a genuine question ("By what method do you bid us discover, etc.") . 
Personally I feel that quo modo is too commonplace a phrase to be 
pressed into the service of the ironical explanation, to be made, as by 
Gertz, the equivalent of quo studio. Further, I feel it natural, after all 
the stabs made by Seneca at unpleasant features of Roman life which 
must have been commonplaces of that life as between slaves and masters, 
that the interlocutor should ask : "Well, how do you propose to change 
that point of view which is at present ours?" It is no fatal objection to 
this that Seneca produces a quite inadequate answer; more than one 
of his answers is that. Nor am I so sure that it is entirely inadequate ; 
after all, Seneca calls for a broad outlook and a recognition of the low 
and despicable {humilia et ahiecta) nature of the objects of most human 
activity. Now these adjectives are far from identical with those used in 
the question {exigua, misera, puerilia), and it is quite possible that Sen- 
eca is asserting, if somewhat vaguely, that if one gets to see the pitifully 
commonplace character of most. human aims, he will readily agree that 
most alleged human wrongs, especially in relation to the master class, 
are "small, pitiful, and childish." 

I therefore hold by the idea that quo modo, etc., is a genuine inquiry 
and that intueri means "to get to perceive," "discover," and so too Basore 
in translating the passage. 

V, 33, 2 (141 : 22) : libet intueri fiscos in angulo iacentis : hi sunt prop- 
ter quos oculi clamore exprimantur, fremitu iudiciorum basilicae 
resonent, evocati ex longinquis regionibus indices sedeant iudieaturi 
utrius iustior avaritia sit. 

In the Elzevir Variorum'" the note of the great Lipsius reads : fiscos : 
sportas in quibus pecuniae : quas nota in angulis conclavis alicuius solere 
deponi. Why, one may ask, if the treasury is under lock and key, is there 
any special reason to store the money-bags in corners ? And in any event 
did the eminent scholar notice that Seneca does not say in angulis but 
in angulo f The fact is that his note is not only not helpful but actually 
prejudicial to a correct interpretation of the passage.^" This is demon- 

^* Pul)lislied by Daniel Elzevir at Amsterdam, 1672. 

"■" If I seem to speak with feeling on this point it is for the reason that I was once 
rebuked by Castiglioni, the Italian scholar, for having ventured to expose the intrinsic 
aljsurdity of an emendation of Lipsius. That does not help classical studies ; the great 
must be recognized as great where they are great, but not hailed as so great as to be 
deemed correct where they are wrong. 



252 University of California Publications in Classical Philology 

strated by the work of all the translators, English, French, and German; 
they have all bowed to the fame of Lipsius and have produced render- 
ings of which Basore's will serve as a type : "it is a pleasure, you say, to 
see money-bags lying in the corner.'"" This translation shows, however, a 
realization that the thought here expressed is attributed by the author 
to the interlocutor; this is undoubtedly correct. I think it might be 
printed as an interrogation. 

One of the great advantages connected with the study of an author 
whose works have survived in sufficient bulk is that he will often explain 
himself if allowed to do so. Now in V, 25, 1 we read : aequiore animo 
filium in angulo flevit, qui vidit acerba funera etiam ex regia duci ; this 
involves the trite comparison between the hovel and the palace, the 
cottage and the mansion. In angulo means "in a hovel," and so Basore 
translates it. Again in Dial. XI, 6, 4 (correct iheTliesaurus here, which 
gives XII for XI )'^ occur these words: multa tibi non licent, quae hu- 
millimis et in angulo iacentibus licent ; the phrase in angulo iacentihus 
by its association with humillimis must mean "living obscure lives in a 
hovel," "in holes and corners." In Seneca's Medea (verses 249-250) the 
heroine begs of Creon miseriis angulum et sedem . . . latehrasque viles, 
where the last phrase is sufficiently descriptive of the mean estate im- 
plied in the word angulum. 

In the passage before us (1) in angulo does not refer to a corner of 
any room in a house, treasury or otherwise ; (2) iacentis does not modify 
fiscos adjectively, but is the possessive genitive of the substantive parti- 
ciple attached to it ; (3) ioi angulo iacentis means "of the man who lives 
obscurely in a hole in a wall," that is, "of the man who lives socially 
obscure." We were talking at the end of the last sentence about kings 
plundering ancient cities in their search for gold and being besmeared 
with the slime of covetousness ; what more effective contrast than to pass 
. from a king to the man who lives humbly in a hovel ? Translate : " (we 
have been discussing kings) ; do you want to consider the money-bags 
of the man who lives obscurely in a mean abode?" (Since this man has 
fisci, 1 should say a miser perhaps, a natural type to introduce where 
money is under process of condemnation). "\Vell, what happens to his 
money-bags ? It's over these that men scream till their eyeballs protrude, 
that law courts re-echo with the din of trials, and that jurors are sum- 
moned from distant localities to decide whether the avarice of A or that 
of B establishes a better claim."*^ 

"o Op. cit., Vol. I, p. 335. 

"^ I presume that a collaborator of the T.L.L. -would follow the Ambrosian order. 

■*- Having worked out the solution to this point, I find that Axelson has anticipated 
me, Neue Senecastudien, pp. 8-10. When I reflect how many have combined their 
forces in the previous errors of interpretation here, I hope that two witnesses agreeing 
with one another wUl not be too few to establish the truth. 



Alexander : Seneca's Bialogi III, IV, V, Be Ira Libri Tres 253 

V, 37, 5 (145: 22) : ergo te Ennius, quo noii delectaris, odisset, et Hor- 
tensius, si orationes eius improhat'es, simultates tibi indiceret, et 
Cicero, si derideres carmina eius, inimicus esset. 

Bourgery does not feel the need of any such addition as that implied 
in the italicized words suggested by Haupt. The text as it stands would 
involve using quo non delectaris a second time when the reference is to 
Hortensius, and this is possible; it might have been carried one stage 
farther and have been used where the reference is to Cicero, but in this 
case a particular weakness in Cicero is recalled, and the thought of this 
drives out quo non delectaris, or, to put it more accurately, brings about 
a replacement centering around carmina. I agree with Bourgery and 
therefore reject Haupt's addendum to the text as well as that which I 
have myself previously proposed/' There is no reason to suppose that a 
separate reason had to be provided with each proper name. 

V, 38, 1 (145 : 1) : quanto Cato noster melius ! 

Fickert added Cato, but it is not necessary and cannot even claim to 
be an improvement. We have just had Diogenes cited as a Stoic philoso- 
pher ; so much for the Greeks. But among Romans, and certainly with 
Seneca, there is but one Roman high priest of Stoicism, the younger 
Cato ; his name it is unnecessary to mention, especially as the story that 
follows, undoubtedly a famous one, promptly clears the point up for 
Romans. The noster without the name is a much more affectionate ex- 
pression than that which would be created by the addition of the name. 

V, 41, 3 (148 : 10) : nihil ergo habet in se utile taeter iste et hostilis ad- 
fectus, at omnia ex contrario mala, ferrum et ignes. pudore calcato 
caedibus inquinavit manus, membra liberorum dispersit, nihil vacuum 
reliquit a seelere, non gloriae memor, non infamiae metuens, inemen- 
dabilis, cum ex ira in odium obcalluit. 

This is very peculiar writing, as will be observed if it is carefully trans- 
lated: "this foul and inimical (?) emotion contains within itself there- 
fore nothing profitable, but on the contrary every e^nl thing, fire and 
steel." The idea seems perfectly complete with the words "every evil 
thing" uttered ; what can possibly be the meaning of the added ferrum 
et ignes f Examples of the "evil thing-s" follow in the next sentence; if 
these followed immediately after omnia ex contrario mala, all would be 
clear, but as it is, we have the interposed (as it seems) ferrum et ignes. 

None of this, of course, escaped the watchful eye of Gertz, who saw 
that if a distinct pause were made after mala and an appropriate verb 
found for fe rrum et ignes, everything would be made orderly and logical. 

^^ Op. cif., p. 13. If there is any merit in making a suggestion for Avords to be in- 
serted, I should like to revise my proposal to make it take the form simnl a te sperni 
sensit orationes for the sake of an attractive rhythmic clausula. This does not affect 
in any way the paleographic possibilities. 



254 University of California Publications in Classical Philology 

The problem posed therefore was what verb to provide for the govern- 
ment of ferrnm et ignes in order to obtain parallelism with mquinavit 
manus, membra dispersit, and nihil vacuum reliquit. Gertz suggested 
tractavit, but if a verb is to be supplied I should prefer on metrical 
grounds paravit ; this gives a double cretie with an extra syllable (Axel- 
son, type 3). The phrase is also a standard Latin form of expression. 

There is, of course, still the possibility that the Ambrosian text here 
is actually what Seneca wrote, the ferrum et ignes being perhaps a rhe- 
torical flourish that ends in air rather vaguely after omnia mala. This 
takes a good deal of courage, however, and as the loss of a single word 
is a common enough occurrence once minuscule mss. have advanced to 
a considerable degree of word separation, as the Ambrosian codex, and 
as presumably its immediate predecessor has also done, I incline to mala ; 
ferrum et ignes <,paravit^, pudore calcato, etc. 

V, 42, 2 (148 : 24) : quid iuvat dies, quos in voluptatem honestam inpen- 
dere licet, in dolorem alicuius tormentumque transf erre ? non capiunt 
res istae iacturam nee tempus vacat perdere. 

Bes istae has been found troublesome; thus Karsten" would alter istae 
to istam, understanding res in the sense "the course of things human" 
and istam iacturam to mean "the loss you thus sustain." Basore trans- 
lates res istae by "your fortunes," which to my way of thinking singu- 
larly fails to relate itself to the context; the difficulty has not been 
thought through. Bourgery writes "les affaires humaines ne peuvent 
souffrir de retard," but I do not know how he would justify translating 
res istae as he does. 

The explanation is, as often, the obvious one, that res istae refers to 
dolorem alicuius tormentumque preceding. "These things (the pain you 
inflict in your anger on some opponent, the anguish to which j^ou succeed 
in putting him) do not contain the loss (of time involved in getting 
these sweet revenges)." The dolor and the tormentum are vessels into 
which you pour your time ; they do not hold it, or, as Seneca says, they 
do not hold the loss, that is, the superfluous time you spend in seek- 
ing these revenges over and above the time that honorable pleasure 
would ask. 

^' H. T. Karsten, Mnemosyne, XVII (1889), p. 154. 



A DRAVIDIAN ETYMOLOGY OF THE SANSKRIT 
PROPER NAME NALA 



M. B. EMENEAU 

The group of Sanskrit nouns nadd- (Rgveda nald-), nala-, nadd- "reed" 
has been much discussed historically and no rehearsal of the details of 
their etymologies need be given here.^ It will be sufficient to say that the 
words fall into two quite independent groups. 

In the first category is nadd- with Iranian, Armenian, and Baltic cog- 
nates. However, the word is very problematical. The occurrence of the 
required meaning ("reed") is posited only for the Rgveda (with repeti- 
tion of one of the Rgveda passages in the Atharvaveda) and has been 
much contested. Ludwig seems to have been the originator of this inter- 
pretation in his translation of 1.179.4.- Pischel extended the interpreta- 
tion to 1.32.8^ and later^ to 10.11.2, 2.34.3, and with a secondary mean- 
ing "membrum virile," to 1.179.4 and 8.69.2 (leaving only one Rgveda 
passage, 10.105.4, for the meaning "horse"). Bohtlingk accepted the 
meaning in Nachtrag 3 to the smaller lexicon, as did Oldenberg in a 
journal article,* and Geldner in his translation of 1.32.8.^ Later, ^ Olden- 
berg repudiated his statements and contested Pischel's interpretation, 
WackernageP regarded the translation as uncertain, and Grassmann,^so 
far as I have been able to find, never acknowledged the interpretation. 
Considering the history of the exegesis of the word and the obscurity of 
the passages in which it occurs, no etymology based on the meaning 
"reed" can be of any value. ^° 

' Nor need a full bibliography be given. It will suffice to refer to Alois Walde- 
Julius Pokorny, Vergleichendes Worterbuch der indogermanischen Sprachen, Vol. 2 
(Berlin u. Leipzig, 1927), p. 317 {s.v. nard(h)o-), 329 (s.v. nedo-), 700 {s.v. (s)ner-). 

2 Alfred Ludwig, Der Rigveda . . . ins deulsche iibersetzt, Vol. 2 (Prag, 1876), p. 663. 

3 ZDMG 35 (1881), pp. 717-724. Lanman, A Sanskrit Reader (Cambridge, Mass., 
1884), p. 361, followed Pischel. 

* Richard Pischel u. Karl F. Geldner, Vedische Studien, Vol. 1 (1889), pp. 183-198. 
Whitney did not follow him for 10.11.2 which is repeated as Atharvaveda 18.1.19 
(W. D. Whitney, Atharva-veda Samhitd translated [Cambridge, Mass., Harvard 
Oriental Series 7, 8, 1905]). 

6 ZDMG 39 (1885), p. 66, fn. 1. 

* Der Rigveda. Erster Teil, erster bis vierter Liederkreis (Gottingen u. Leipzig, 
1923). He did not accept it for 1.179.4 and 2.34.3; I have no access to his translation 
of the later books. 

^ Hermarm Oldenberg, Rgveda. Textkritische und exegetische Noten. Erstes bis 
sechstes Buck (K. Gesell. d. Wissen. zu Gottingen, Abhandlungen, phil.-hist. Kl., 
N.F. 11, 5 [1909]). See especially on 1.179.4, also on 1.32.8. 

* Jakob Wackernagel, Altindische Grammatik, Vol. 1, p. 173. 

' Hermarm Grassmann, Worterbuch zum Rig-Veda (Leipzig, 1873) ; Rig-Veda. Iiber- 
setzt (Leipzig, 1876-1877). 

'" Those who have had occasion to deal with the etymology of Sanskrit words on 
a large scale seem, so far, to have safeguarded themselves with respect to this word 

[255] 



256 University of California Publications in Classical Philology 

In the second category is the group nadd-, nala-, composed of words 
whose existence and meaning are entirely certain. Positing an earlier form 
of stem *nardd-, of which nadd- would be a derivative by a very com- 
mon but not universally operating Sanskrit phonetic change," we find 
Iranian cognates and a less certain congener in Greek vapOr]^. The rela- 
tionship between nadd- and nala- must be of the kind seen in the com- 
plete alteration of d and dh in other Vedic texts to I and Ih in the Kanva 
recension of the Vajasaneyisaihhita, in the Sankhayanasrautastitra, and 
in the 20th book of the Atharvaveda, as well as the sporadic cases of 
alternation of d and I found elsewhere in Sanskrit texts. ^^ 

The history of the two words within Sanskrit will be of interest. Nadd- 
appears once in the Rgveda (in the form nala- regular for that text), 
eight times in the Atharvaveda, in the Satapathabrahmana and other 
Vedic texts, in Panini and later grammatical texts, and very rarely in 
other classical Sanskrit texts. Nala- is found in the Mahabharata and in 
later texts, but apparently in no Vedic text except the late compilation 
Gobhilasraddhakalpa. 

When we examine the proper name Nada/Nala, which is a homonj-m 
of the common noun in both its forms, ^^ we find also that its history 
parallels that of the common noun. Nala is found in the iVIahabharata as 
the name of a king of Nisadha and as the name of a monkey-chief, and in 
later texts as the names of these two and of other, mainly derivative, 
characters. The other form, Xada, is found once in a Vedic text, viz. in 
the Satapathabrahmana 2.3.2.1,2, which makes identifications of various 
characters with the various sacrificial fires, and among them of Nada 
Naisadha (or Naisidha ; the manuscripts vary) with the Anvaharyapacana 
or southernmost fire. It is also said that day bj^ day Nada carries "King 
Yama" south, which Weber'* interpreted as meaning that Nada, King 
of Nisadha, was making warlike expeditions in the south and carrj'ing 
death and destruction there. The name Nada was known also to Panini, 
who gave instructions (4.1.99) for the formation from it of Nadaj'ana "a 

by introducing "vermutlich" (Walde-Pokorny, op. cit., Vol. 2, p. 329), "wahr- 
scheinlich" (P. Persson, Beitrdge zur indogermanischen Wortforschung [Uppsala, 
1912], p. 814, fn. 1), a question-mark (C. C. Uhlenbeck, Kurzgefa-sstes etymologisches 
"Worterbuch der altindischen Sprache [Amsterdam, 1898/1899], s.v.), or the like. 

11 See e.g., Wackernagel, op. cit., Vol. 1, §146b. Professor Franklin p]dgerton of 
Yale University writes me that he would consider this a Middle Indie change. He 
says: "Every intervocalic domal stop in Sanskrit (except where an original z was 
lost) is a Prakritisra." 

12 Wackernagel, op. cit.. Vol. 1, §194a (na(id-> nala- is not listed); Maurice Bloom- 
field and Franklin Edgerton, Vedic Variants, Vol. 2, §270. 

1' See e.g., the play on the two nouns, common and proper, in Harsa's Nai§adha- 
carita 1.35. 

" Albrecht Weber, Indische Studien, Vol. 1 (1850), pp. 224-227. 



Emeneau: A Dravidian Etymology of the Sanslcrit proper name Nala 257 

more remote descendant of Nada." The name is not known in classical 
Sanskrit texts. 

What seems to be quite clear is that the Nada Naisadha (or Naisidha) 
of the Satapathabrahmana passage is to be equated with Nala Naisadha 
of the epic ; it does not matter for our purpose that the implications of the 
brahmana passage seem to be absent from the epic,^^ nor is it of much 
importance whether one legendary figure is involved or two.^^ We can, 
without any hesitation, make the linguistic equation Nada = Nala, 
parallel in chronology to that of the common noun. 

According to the few scholars who have expressed themselves on the 
subject the name is derived from the common noun meaning "reed." The 
etymologists in general ignore the proper name. The only scholars who 
can be quoted are the lexicon-makers, who perhaps imply this derivation 
in that they put the proper name and the common noun under one 
rubric," and students of Hindu proper names, who have explicitly made 
the derivation; as Temple in his list of the kings of the solar dynasty of 
Ayodhya,^^ Hilka/^ van Velze,-" and Pargiter.-^ 

Do the principles followed by the Hindus in giving names make it 
plausible that a man should be called by a word meaning "reed"? We can 
best determine this by examining Hilka's general statements on his lists 
of names derived from natural phenomena.-' He determines (I paraphrase 
and rearrange) that such names are derived in four ways. (1) The name 
is, or contains as an element, the name of a natural entity whose qualities 
it is desired that the bearer of the name should show; e.g. the element 
sirhha "lion" in the name of a man of the warrior caste, or a word for 
"lotus" (an exemplar of beauty) as an element in the name of either 
a man or a woman. (2) The name is, or contains as an element, the name 
of a natural entity whose qualities it is thought that the recipient of the 

1^ Weber, loc. cit.^ 

1^ Sayana on the SB passage took its character and that of the epic to be the same. 

1^ So B6htlingk-Roth, Monier-Williams, Apte, etc.; also Lanman, op. cit., p. 180, 
s.v. nala. 

" R. C. Temple, A Dissertation on the Proper Names of Panjdbis (Bombay, 1883), 
p. 58. 

" Alfons Hilka, Beitrdge zur Kenntnis der indischen Namengebung. Die altindischen 
Personennamen (Breslau, Indische Forschungen herausgegeben von Alfred Hille- 
brandt, Heft 3, 1910), p. 116. H. de Willman-Grabowska's article "Les noms de per- 
sonnes dans le ^atapatha Brahmana" in Melanges linguistiques offerls a M. J. 
Vendryes (Paris, Collection linguistique publiee par la Societe de Linguistique de 
Paris, 17, 1925), pp. 373-390, unfortunately confines itself to general conclusions 
with examples and does not treat or list our word. S. Sorensen, An Index to the 
Names in the Mahabharata (London, 1904-1925), gives no etymologies. 

20 J. A. van Velze, Names of Persons in Early Sanscrit Literature (Utrecht, 1938; 
doctoral dissertation), p. 104. 

21 F. E. Pargiter, Ancient Indian Historical Tradition (London, 1922), p. 133. 
^^ Op. cit., p. 113. 



258 University of California Publications in Classical Philology 

name shows; in most cases indistinguishable from those under (1), except 
when we know the circumstances of the giving of the name ; such a name 
would be a "nickname" in origin, but may replace all other names of the 
recipient. (3) The name is derived from the mythology, the theologj'", or 
the religious practices of the sect according to which the family worships. 
This is true of all names with god-names or epithets of gods as elements, 
or of such a name as Sulasadatta, with Sulasa = TulasI, the sacred 
plant of the Visnu-cult.^^ (4) The name commemorates an event; e.g. 
Uddalaka was so named because he was conceived under an udddla-tree.^* 
In addition Hilka mentions : (5) A name is that of something mean or 
contemptible and is apotropaic of evil influences. 

Unless we are told the circumstances in which the name is given, we 
shall frequently be at a loss to classify names met in texts or in the every- 
day life of India. So with our name Nada Nala. Principles (1) and (2) 
can be eliminated at once ; there is no peculiar excellence in the nature of 
the reed that would make it a suitable name. Likewise, principle (5) can 
be eliminated in all probability, for the reed is not in Temple's extensive 
lists of opprobrious names used apotropaically;^^ there seems to be noth- 
ing intrinsically mean about the reed that would make it a suitable name 
of this kind. Nor is the reed associated with any Hindu cult (to the best 
of my knowledge) so that principle (3) might work. We may contrast the 
word munja which denotes another species of rushlike or grasslike plant 
and is used as a name or an element of names, for this is derived from the 
use of the plant in making the brahman's girdle and sacred thread, and 
from the use of its name as an element in names of Visnu and Siva. We 
are left with principle (4) , and this is unsatisfactory since we are not told 
any story of the giving of the name nor is there any indication at all in 
the texts that the name is in any waj'^ extraordinary, a rather good indica- 
tion that it is not commemorative of am^thing. 

Since we find then that the commonly accepted derivation of the proper 
name is unsatisfactory, it is necessary to seek for a better. I venture to 
suggest a Dravidian source. 

The South Dravidian languages — Tamil, Malayajam, Kannada, Te- 
lugu, Coorg, Tuhi, and Toda (but not Kota) — show a base nal- (or 
"reflexes of it), meaning "good." In Tamil, the base, as applied to persons 
or actions of persons, connotes excellence, virtue, propriety, or beauty 
(so for nalam), and similarly in the other three literary languages, 
Malayalam, Kannada, and Telugu, so far as the dictionaries help to 

^^ Op. cit., p. 117. So also Tulsidas, and the like. 
s-i Op. cit., p. 113. The story is in Jataka487. 
" Temple, op. cit., pp. 22-29. 



Emeneau: A Dravidian Etymology of the Sanskrit proper name Nala 259 



establish the meaning exactly; this is true also in Tulu and Coorg.-^ In 
Toda the base is restricted to the meaning "beauty."-^ 

26 The evidence for Coorg is in my field-notes. The application, so far as I dis- 
covered, centers on "excellence, virtue," and does not extend as widely as in Tamil. 

" The Toda form is nas-. Dravidian Z(0>Toda s (a voiceless alveolar sibilant, a 
different phoneme from the voiceless post-dental sibilant) in many other words, as 
in the following table. (The varying forms for Kannada represent in part different 
chronological periods of the language; e.g., forms with p- are earlier than those 
with h-, and the forms with I in the word for "shade" are earlier than the others. 
Square brackets enclosing the forms indicate that they are not cognate or not 
closely cognate with the other forms in the same line; a dash indicates that cognates 
have not been found.) 



Toda 


Kota 


Tamil 


Kannada 


pas 


pal 


pal 


pal; hal, hallu 


"tooth" 








kas 


kal 


kal 


kal, kallu 


"stone" 








-paspf 


pala'v 


palavu 


[palasa, palasu; halasu ] 


"jack-tree" 








nesof 




nilavu 




"moonlight" 








Tpasj 


e-paj* 


palli 


palli; halli 


"small house-lizard" 






es 


el 


ilai 


etc 


'"leaf" 








-pms 


vily 


vil 


bil, billu 


"bow" 








pysj 


pnj 


puli 


puli; hull 


"tiger" 








isj 


eyj 


eli 


Hi 


"rat" 








nu's 


nu-l 


nu-l 


nu'l, nu'lu 


"thread" 








a's 


a-l 


a-lai 




"Toda house" 


"cattle-shed' 


' "apartment, 


hall" (<Skt. sate) 


vo-s 


pa-l 


pa-l 


pa-l, pa-lu; ha-l, ha-lu 


"milk" 








turn's 


to-l 


to-l 


to'l, lo-lu 


"skin" 








ne-s 


nerl 


nilal 


nelfll, nelalu, neral, nera\u 


"shade, shadow' 








mys 


[tnolm] 


muyal 


[mola] 


"hare" 








Tpi-s 


vi'l 


veyil 


bisil, bisila, bisilu 


"heat of sun" 








po's 


va-l 


va-yil 


ba-gil, ba-gila, ba-gilu 


"entrance" 








pe-sj 


ve-j 


ve-li 


be-li 


"wall of pen" 


"fence" 


"fence" 


"fence" 


kmsf 


Ml 


cil, cila 


kela 


"few" 


in Ml gat- 
"be toosmal 
for one's age' 


"few" 


"few" 



* The first element is uncertain. The only possibilities that I can find in my matenal are e;, 
the interrogative "what?", and ep (1) "long sticks used as outside layer of bundle of twigs, 
(2) "welded joint in iron," none of which seems very appropriate; it is possible also that e- in this 
construction is a variant of some word eC, in which C is some consonant other than p. 

(Note -~ continued on next page) 



260 University of California Publications in Classical Philology 

My suggestion is that Nala is a formation with the Sanskrit primary- 
derivative suffix -a-^ from this South Dra vidian base, and that the mean- 
ing is "the good man," or less probably, "the handsome man." 

So far as I have been able to discover, no reflexes of this base have been 
found in the central and northern Dravidian languages (Kui, Kxirukh, 
Gondi, Kolami, Malto, Brahul) ; a negative for any one of these might be 
laid to the paucity of the material available for study (the material for 
Kui and Brahui is fairly adequate), but a negative for all of them is 
probably significant and to be interpreted as meaning that the base is 
really non-existent in them. What the range of the Dravidian languages 
and of this base over the area covered by them was at a time early enough 
to allow a borrowing from Dravidian to appear in the Satapathabrah- 
mana and the epic, is impossible to tell. However, there is nothing im- 
possible in the suggestion of a borrowing from Dravidian so early and in 
an area so far north as is implied. We can find an epic parallel in palllghosa 
"a barbarian village" of Mahabharata 12.326.20,-^ which is to be inter- 
preted as a descriptive compound with the prior member a noun (palll) 
in apposition to the second (ghosa), i.e., "barbarian village-herdsmen's 
station. "^0 The noun palll (also palli) denotes the village of a wild tribe 
and is undoubtedly the Dravidian palli.^^ A fairly long list of words 

2T continued. 

In a few words Dravidian l{l) is represented by Toda I and not by s. One of these 
is almost certainly a borrowing, viz., pe-lj "fence" beside pcsj "wall of pen," since 
the Todas do not have fences in their culture. For others I have no explanation; e.g., 

Toda Kota . Tamil Kannada 



koi kal 


kal 


kal, kalu 


"leg, foot" 






pul pul 


pul 


pul, pullu; hullu 


"thatching-grass" "grass" 


"grass" 


"grass" 


null- nil- 


nil- 


ml- 


"stand" 







(but cf. the noun nius "a standing-place, place": Tamil nilai, Kannada nele; on the 
Toda word, see Emeneau, "Personal Names of the Todas," in American Anthro- 
pologist iO [1938], pp. 211-212, §11). 

28 W. D. Whitney, Sanskrit Grammar (2d ed.; Cambridge, Mass., 1889), §1148 p. 

29 E. W. Hopkins, "The Social and Military Position of the Ruling Caste in 
Ancient India, as represented by the Sanskrit Epic," in/.40S 13 (1889), p. 77. 

30 Whitney, op. cit., §1280 d. 

" Pace E. Liden, Zcitschrift fur Vergleichende Sprachforschung {Kuhns Zeii- 
schrifl) 40 (1905), pp. 260-261, who attempts a connection with the lexical Sanskrit 
padra- "village." The Dravidian words are: Tamil paJIi "hamlet, small village, 
herdsmen's village, temple (especially of Buddhists and Jains), school, etc."; 
Malayalam palli "small village, church, temple, mosque, school"; Kannada pfl//i, 
haili "hamlet, village"; Toda po\j "sacred dairy: matrilineal sib; Badaga house" 
(see Emeneau, "Toda Marriage Regulations and Taboos." in American Anthro- 
pologist 39 [1937], pp. 103-112). The derivation of Sanskrit palll (palli) from Dra- 



Emeneau: A Dravidian Etymology of the Sanskrit proper name Nala 261 

borrowed from Dravidian and first found in Sanskrit in the epic might be 
compiled, without at all strengthening our case that a word found in the 
Vedic literature was borrowed from Dravidian. It should, however, be 
noted that our word is a proper name, referring to a region south, if not 
too far south, of the Ganges- Jumna Doab which is the land of the Yajur- 
veda and the Brahmana-texts. There would be nothing at all strange in 
finding that that southern land, not yet overspread by the culture 
couched in the Indo-Aryan languages, had kings whose names were 
Dravidian. 

We have as yet no compilations and comprehensive analyses of proper 
names from the early Dravidian literatures and inscriptions (at least, I 
know of none in this field where bibliographical aids are non-existent), 
nor of the non-Sanskritic nomenclatures of the present-day Dravidian 
communities, in so far as they do not follow the Sanskritic system. Lack- 
ing such aids to investigation, I can adduce as a proper name derived 
from the base nal- only the Toda Naso'n, "he who is beautiful," which I 
discussed in a previous paper.^- This, however, is enough evidence to 
allow us to guess that some earlier Dravidian community also found the 
base a suitable one from which to derive a proper name. 

The only weighty objection to the etymology lies in the fact that the 
Vedic form is not the well-known later Nala, which the etymology re- 
quires, but Nada. My answer to this would be that we do not really 
know which of the two variant phonemes involved, I or d, corresponds 
more closely to the I which was presumably in the Dravidian word. 
Tamil seemed to find its retroflex / a closer equivalent to the Sanskrit I 
than its alveolar /, when it borrowed this name of the epic in the form 
Nalan.^^ The Indo-Aryan speakers who recorded the name in the 
Satapathabrahmana knew nothing of its Dravidian origin and meaning, 
but, in an effort to make it meaningful within the terms of their own 

vidian palli was recognized by F. Kittel, A Kannada-English Dictionary (Mangalore, 
1894), p. xxxi, and Robert Caldwell, A Comparative Grammar of the Dravidian or 
South-Indian Family of Languages (3d ed.; London, 1913), p. 572 (the 1st edition of 1856 
is not available). 

The lexical Sanskrit palll "a small house-lizard" (with the derived, also only- 
lexical, pallika) is, in spite of the arguments of Liden {loc. cit.) and the citation in 
Walde-Pokorny, Vol. 2, p. 23, not derived from *palla-, Prakritic for *padla- or 
*padra- "(the snake) having feet"; it is undoubtedly Dravidian, cf. Tamil palli 
with the same meaning and the cognates given in fn. 27. This derivation is found 
in Kittel (op. cit., p. xxi; not in Caldwell's 3d edition). 

32 "Personal Names of the Todas," in American Anthropologist 40 (1938), p. 213, 
§14. 

55 Similarly with many, if not most, other Tamil borrowings of Sanskrit words 
containing I. Has this anything to do with the fact that Sanskrit I has become in 
Marathi a retroflex lateral /, not, to be sure, quite identical with the South Dra- 
vidian 11 



262 University of California Publications in Classical Philology 

language, made it accord in phonetics with their common noun nadd- 

"reed." The epic composers, who spoke a somewhat different dialect from 

that of the Brahmana-litefature, either did not inherit the tradition that 

the proper name was homonymous with the common noun, and therefore 

preserved it in a form (presumably) nearer to the original, or, if they did 

inherit the tradition, let the proper name undergo the same phonetic 

change as that seen in the common noun, quite without reference to the 

original Dravidian phonetics.^* 

3* Though the derivation that I propose seems very obvious to me, I camiot find 
that it has ever been suggested before; certainly no treatment of it is recorded by 
Constantin Regamey, "Bibliographie analytique des travaux relatifs aux elements 
anaryens dans la civilisation et les langues de I'lnde" {Bulletin de VEcole Frangaise 
d' Exlreme-Orient 34 [1934], pp. 429-566). I must record here my thanks to the 
Department of Classics of the University of California for aid in acquiring photo- 
stats of pp. xiv-xlv of Kittel's Kannada dictionary. 



HIGH COMEDY IN THE ODYSSEY 

BY 
WALTER MOERIS HART 

Matthew Arnold's praise of Sophocles as one who "saw life steadily 
and saw it whole" is perhaps even more applicable to Homer. For he, 
like Sophocles, was concerned with both gods and men, and on the human 
side went far beyond Sophocles, approaching even something like 
Shakespeare's cloudless, boundless view. Epic tradition required him to 
write of the heroic past, of kings and heroes, but he manifests his love 
for the humble present and for lowly people. His view of them all is not 
without that sympathy and reverence which, in Chesterton's opinion, 
are of supreme importance to true humor .^ 

Looking at life steadily did not prevent him from shifting his point 
of view, seeing it now as tragedy, now as comedy. The Iliad is a tragedy ; 
but even in the Iliad there are flashes — often, indeed, more than flashes — 
of the comic. 

The Odyssey is romance, but it is rich in comedy ; its comic scenes are 
developed by Homer with especial and loving elaboration." 

The comic in the two poems is remarkable for its variety, for its range 
from low to high. It is well to approach it with Shakespeare's comedies 
in mind. Even in the lightest of these there are serious subplots, as, in 
the Comedy of Errors, the story of old Aegeon, condemned as a Syra- 
cusan to die in Ephesus, or, in The Merry Wives of Windsor, the roman- 
tic love affair of Anne Page and Fenton. In the romantic comedies there 
are comic subplots, and the main plots involve the characters in comic 
situations. These main plots are often little Odysseys, concerned with 
adventures by land and sea, largely in Homer's Mediterranean world : 
Sicily (on whose eastern shore one may still see the small islands which 
are the rocks that Polyphemus flung after the ship of the escaping 
Odysseus), Illyria, the eastern shore of the Adriatic, Epidamnum (not 
far from Ithaca, the island home of Odysseus), Verona, Venice, Rome, 
Athens, the famous "sea coast of Bohemia." There are possibilities of 
tragic outcome, averted to achieve a happy ending : threat of death, with 
actual death for minor characters, even for those who, like Antigonus 
the savior of Perdita, have the reader's sympathy. There is separation 

^ G. K. Chesterton, essay on Bret Harte, in Varied Types (New York, 1903), pp. 
179 ff. 

- Indebtedness to The Composition of Homer's Odyssey, by W. J. Woodhouse, 
Oxford, 1930, is gratefully acknowledged. 

The translations used are those of Lang, Leaf, and Myers, for the Iliad; and 
T. E. Shaw, and Butcher and Lang, for the Odyssey. 

[ 263 ] 



264 University of California Puilications in Classical Philology 

of wives and husbands, brothers and sisters. There are disguises, im- 
personations, mistaken identities, with subsequent recognitions and 
reunions. 

There is a wide range of the comic from low to high in both main 
plots and subplots : in the main plots from matters like the discomfiture 
of Falstaff, in The Merry Wives of Wiiidsor, carried in a basket with 
dirty linen and thrown into the Thames like a barrow of butcher's offal, 
to matters like Portia's outwitting of Shylock in The Merchant of 
Venice: in the subplots, from the practical jokes on Malvolio, in Twelfth 
Night, to the play of character and wit in the affair of Benedick and 
Beatrice in Much Ado about Nothing. 

The Odyssey is, obviously, not comedy in the sense of drama : it is not 
"adapted to be acted upon a stage"; "the story" is not "wholly related 
by means of dialogue and action.'" But it is not purely epic ; the narra- 
tive is not continuous, for as Aristotle said : "Homer . , . knows the right 
proportion of epic narrative ; when to narrate, and when to let the char- 
acters speak for themselves With little prelude [he] leaves the stage 

to his personages, men and women, all with characters of their own."* 
The "right proportion," Aristotle implies, is much dialogue and little 
narrative ; Homer brings epic close to drama ; the casual reference to 
the stage is significant — Homer "leaves the stage to his personages." 
Those personages have characters of their own, which appear in what 
they say, rather than in what the poet says about them. Character is 
implied also by gesture, attitude, surroundings, as on the stage. 

The scenes to which I wish now to invite attention approximate the 
drama in form ; in them Homer leaves the stage to his personages and 
allows them to speak for themselves. 

I shall recall first a few instances of low comedy, in part to illustrate 
Homer's range and variety, in part to bring out by contrast the quality 
of the comedy which I venture to characterize as "high." 

I begin with the Polyphemus episode. This is not high comedy. "We are 
invited to laugh chiefly at the pain suffered by the Cyclops ; also at his 
incredible stupidity. Exaggeration, too, plays its part— the mere size of 
Polyphemus, the horrid details of his Gargantuan meals. It is easy to 
find amusement also in the mental sufferings of Odysseus, who, since it 
is he who is telling the story, cannot very well have been eaten, first 
Qr last. 

Aftertimes found this story comic: two Greek comedies, now lost, 
dealt with it ; Euripides wrote, and Shelley translated, a Satyr play 
based upon it ; Lucian, in his Dialogues of the Sea Gods, could see it from 
the comic point of view ; and so could Alfred Noyes in his Forty Singing 

' Oxford English Dictionary, s.v. "Drama." 

* Poetics 1460 A 5, as translated by W. P. Ker, Epic and Romance (London, 1897), 
p. 17. 



Hart : High Cornedy in the Odyssey 265 

Seamen, who roar out the chorus : "Since Ulysses bung-ed his eye up 
with a pine-torch in the dark" (in the Odysseij, the torch is green olive 
wood; Noyes requires a monosyllabic stake). 

As part of the entertainment of Odysseus at the court of the Phaea- 
cians, the minstrel Demodocus sings not only of the Trojan war and the 
great deeds of the heroes, but also a purely comic tale concerning certain 
scandalous deeds of Ares and Aphrodite. "The Lady Goddesses," says 
Demodocus, "remained at home, all of them, quite out of countenance 
. . . but unquenchable was the laughter that arose from the blessed gods 
as they studied the tricky device of Hephaestus." The other gods whis- 
pered to one another, but Apollo loudly asked, "Hermes . . . would you 
not choose even the bondage of those rough chains if so you might sleep 
... by golden Aphrodite ?" And to him the gods' messenger replied, "If 
there were chains without end, thrice as many as are here, and all you 
Gods and all the Goddesses to look on, yet would I be happy beside the 
Golden One." At his saying more laughter arose among the Immortals. 

"This," Homer adds, "was the song the famous minstrel sang; and 
Odysseus listened and was glad at heart, and likewise did the Phaea- 
cians." 

Mere farce, with the lightest possible touch on the characters of those 
concerned, this story closely resembles the thirteenth-century French 
fabliaux, those medieval drummers' tales, which Chaucer retold in Eng- 
lish verse and Boccaccio in Italian prose. Many of these were of Oriental 
origin and it is not inconceivable that Homer heard this one in Smyrna, 
whither commercial travelers of his day had brought it from regions 
farther south and east. "When the veil is over women's faces," said 
George Meredith, "you cannot have society, without wliich the senses 
are barbarous and the Comic Spirit is driven to the gutters of grossness 
to slake its thirst." It is significant that Gods and Goddesses cannot 
laugh in concert. 

My final instance of the lower kind of comedy springs from the ironic 
situation resulting from Odysseus' presence disguised as a beggar in 
his own home, a situation that persists through the latter half of the 
Odyssey, wherein Odysseus is not recognized by the other persons of the 
epic. There are, in consequence, opportunities for ironic incident and 
dialogue, and of these Homer makes effective use. 

One such incident is the encounter of the disguised Odysseus with the 
tramp Irus, who had come on purpose to pick a quarrel. The Wooers 
of Penelope sprang up, laughing, and pressed round the scarecrow 
tramps. 

The appearance of Odysseus, stripped for the fight, struck terror to 
the heart of Irus. The workmen had to hold him and push him forward, 
the flesh of his limbs quaking in panic. Odysseus was puzzling himself 
whether it were better to strike the other so starkly that life would leave 



266 University of California Publications in Classical Philology 

him when he fell, or to tap him gently and just stretch him out. On the 
whole, the gentle way seemed right. So Odysseus only hooked him to the 
neck under the ear and crushed the bones inward, so that blood gushed 
purple from his lips and with a shriek he fell in the dust, biting the 
ground and drumming with his feet. The suitors flung up their hands 
and died of laughing ; but Odysseus took him by the leg and dragged 
him out and propped him against the fence, saying "Sit there and play 
bogy to the dogs and pigs . . ." The suitors still laughed merrily. 

Four hundred years after Homer, Aristophanes amused Athenian 
audiences by the sight of physical suffering. The Roman Plautus fol- 
lowed his example. From him Shakespeare borrowed a play and de- 
lighted the groundlings with Dromio, expecting thanks for bringing a 
rope's end and, mistaken for his brother, getting a beating with it in- 
stead. In both instances there is something more than the mere beating ; 
there is the matter of the mistaken identity. Irus had unwittingly taken 
on one of the greatest of the Greek heroes, the very greatest of crafty 
fighters. Only Telemachus and we know that this ragged beggar is Odys- 
seus. And we, moreover, have had a glimpse of Odj'sseus' mind : we know 
that, lest he be recognized as Odysseus, he has chosen the gentle way, 
merely crushing the bones inward. We feel superior ; and there is a bit 
of thinking in our laughter; it is not the completely thoughtless mirth 
of the suitors. 

Shakespeare's Twelfth Night furnishes a closer parallel than the 
Comedy of Errors. There practical jokers arrange a duel between two 
unwilling combatants. We know, however, that Viola is a girl in dis- 
guise and recognize in her fear a becoming feminine timidity, as, in that 
of Sir Andrew Aguecheek, the baseless terror of the male coward. The 
pain, it is true, turns out to be Avholly mental ; there is nothing of the 
Homeric brutality in the outcome. The ironic situation, however, is the 
same. Our amusement is heightened by our knowledge that neither com- 
batant has any real reason to fear the other. Not profound thought, 
obviously, but thinking of a sort, is necessary for our appreciation. To 
this extent our laughter is thoughtful laughter. 

We may enter the regions of higher comedj' by way of a portrait of a 
character seen from the comic point of view, the portait of Nestor. 

Nestor is not a conspicuously amusing figure. Indeed, had not Shake- 
speare intervened with his porti'ait of Polonius, it would not. perhaps, 
occur to us to find him comic. However, once we have been made aware, 
it is difficult to believe that Homer did not smile as he wrote his long 
speeches. Like Polonius, Nestor is a highly self-satisfied, loquacious, and 
reminiscent, and yet a wise and kind, old man. He appears first near the 
beginning of tlie Iliad, attempting to make peace between Agamemnon 
and Achilles. "He made harangue to them and said : 'Verily all the 
Trojans would be glad were they to hear all this tale of strife between 



Hart : High Comedy in the Odyssey 267 

you twain. Of old daj^s held I converse with better men even than you, 
and never did they make light of me ... I played my part in fight ; and 
with mine enemies could none of men that are now on earth do battle. 
And all laid to heart my counsels and hearkened to my voice.' " 

In the Odyssey, Telemachus, who has set out to learn tidings of his 
father from Menelaus at Sparta, stops, on his way, at Pylos, home of 
Nestor. As in the Iliad, Nestor speaks of his own past greatness. "Never 
a man," he tells Telemachus, "could match with Odysseus in wisdom, for 
he outdid the rest in all manner of craft ; . . . and Odysseus and I," he 
adds, "were always of one mind." He could tell Telemachus nothing of 
the fate of his father, but he insisted upon detaining the impatient youth 
for the long ceremony of a sacrifice to Athene. However, he supplied 
chariot and horses for the journey to Sparta and sent his son Peisistratus 
to act as guide. 

The two youths became good friends. On the return, after Telemachus 
had lingered far too long at Sparta and was now eager to be in Ithaca 
once more, he beseeched Peisistratus : "Do not drive me past my ship, . . . 
but set me down beside her, that the old man's sense of hospitality may 
not have power to keep me chafing in his house. I would speed home- 
ward." 

The son of Nestor pondered if this was a thing he could properly 
accept and perform. Reflection showed it to be best. So he turned his 
team out of the w^ay to the water's edge and transferred to the ship all 
the noble gifts of Menelaus. . . . Then he said urgently to Telemachus : 
"Now get aboard and have your crew mustered before my reaching home 
warns the old man. My heart and head assure me that his wilfulness 
will take no excuse. He will himself come here and hail you; refusing, 
as I say, to go back alone. This will fling him in a rage." 

Thus youth, as youth will, conspires against age. But it is a defensive 
and justifiable conspiracy, implying no lack of respect for the old man 
and his kindly, if too insistent and selfish, hospitality. Telemachus ap- 
pears in a better light than Hamlet : he has nothing to say like "These 
tedious old fools!" 

Homer, it is to be noted, has said nothing. He has left the stage to his 
personages to speak for themselves. We come to know Nestor through 
his own words and through the words of the tw^o young men and their 
action in relation to him. 

"There never will be civilization where comedy is not possible," I quote 
George Meredith once more, "and that comes of some degree of social 
equality of the sexes." Such equality is clearly indicated in both Iliad 
and Odyssey, and notably in the three great scenes which are now to 
engage our attention : Telemachus and Helen ; Odysseus and Nausicaa ; 
and Odysseus and Athene. In the first of these, Telemachus and Helen, 
there is, beside social equality-, marked feminine superiority in intelli- 



268 University of California Puhlications in Classical Philology 

gence and in awareness of mutual relations. It is this awareness that 
enlivens the scene with the quality of comic irony, 

Telemachus and Peisistratus were well received by Menelaus. He 
waved them to his bounty, saying, "Take of our food and be glad : so 
that after you have eaten we may enquire of you who you are . . ." When 
their longings for food and drink had been put away Telemachus leaned 
his head across near the son of Nestor and whispered in his ear, that the 
others might not catch his words, "See what a blaze of polished copper 
and gold and amber and silver and ivory goes through this echoing hall. 
Surely the mansions of Olympian Zeus must be like this one, one great 
glory within of things wonderful beyond all telling." 

Menelaus had overheard his whisper. He opened his mouth to them 
with thrilling words : "Dear children, with Zeus no mortal man can vie. 
His houses and his treasures are from everlasting to everlasting. On 
earth — well, there may be a man as rich as myself, or there may not : 
but it was only after terrible suffering and eight years of adventure in 
foreign parts that I won home from overseas with this my wealth." 

Menelaus goes on, one thing leading naturally to another, to speak of 
the lands that he has seen, of his brother Agamemnon slain on his return, 
of other companions slain before Troy or lost on the homeward voyage. 

"Yet above and beyond all my company do I especially grieve for 
ONE. No man of the Achaeans deserved so greatly or labored so greatly 
as great Odysseus labored and endured . . . Without doubt they mourn 
him too, old Laertes and self-possessed Penelope, and Telemachus, who 
was no more than a child newly-born, left behind by his father in the 
house." 

Thus they reveal themselves : Telemachus, the country boy, young 
and inexperienced, overheard in his naive whispering ; iMenelaus, vain 
and self-satisfied, saved for our appreciation by his admiration for 
Odysseus. 

And now Helen, like a vision of Artemis of the golden distaff, came 
out from her high-coffered, incense-laden room, with her women; of 
whom Adraste carried the graceful reclining-chair for her mistress, 
while Alcippe had her soft woolen carpet, and Phylo a silver basket 
mounted on a carriage with gold-rimmed wheels. It was heaped fuU of 
the smoothest yarn and across it . . . lay the distaff with wool of a wood- 
violet blue. The queen sat down in her long chair which had a stool to 
support her feet. 

Homer shows no less skill in his dramatic use of stage properties and 
stage business than in his use of dialogue. His method just here fore- 
shadows the famous characterizing line in one of Chaucer's Canterbury 
Tales in which it is said of a certain Friar that, "from the bench he drove 
away the cat . .'. and sat himself softlj' down."' 

^ The Somnour's Tale, v. 1775. 



Hart : High Comedy in the Odyssey 269 

Helen, womanlike, sees at once the resemblance of Telemacluis to 
Odysseus. "Surely," she says, "this must be Telemachus, that son he left 
behind him a mere infant in the house, when for the sake of this worth- 
less self of mine all you Achaeans came up breathing savage war against 
the town of Troy." 

"Indeed," says Menelaus, "now I can see the likeness which you limn." 

Peisistratus speaks for Telemachus, who is, he says, slow-spoken, and 
states that he is the son of that man, the one and only. All four have a 
desire to weep. 

But into the wine they were drinking Helen cast a drug which melted 
sorrow and made men forgetful of their pains. Then she recalled a story 
of Odysseus, how he had entered Troy disguised as a slave, — she alone 
knew him, — and he slew many Trojans. 

"My heart laughed," she says, "for now my desire had shifted to get 
back home, and deplored too late the infatuation engendered by Aphro- 
dite to lead me away from my own dear country, abandoning child and 
marriage-ties and a lord not poor in wit or looks." 

Menelaus is not displeased by this flattery, and he too remembers a 
tale about Odysseus, but in this one it is Helen who plays the chief part. 
He remembers how when they lay hidden in the wooden horse she had 
circled about it calling by name upon the leaders of the Greeks, upon 
each in the voice of his absent wife. They raged furiously to leap up and 
call her, but with main strength Odysseus held them back. 

Thus Menelaus unconsciously reveals Helen — her quick intuition, her 
skill as actress and impersonator, her gay irresponsibility : she did not 
tell the Trojans the truth about the wooden horse, yet by her little joke 
exposed the Greeks to death. Had it succeeded, Troy would not have 
fallen. That drug which she cast into the wine was, surely, her tact and 
the personal charm which led all men to forget their sorrows. What better 
way to hearten the son than to tell tales of the prowess of the father ? 
Homer calls it a drug, not, I am sure, because of his masculine ignorance, 
but rather to rationalize for the benefit of his hearers. 

Menelaus urged Telemachus to tarry for eleven days or even for 
twelve. Telemachus could not stay so long, he said ; yet he lingered, not 
for twelve days but for thirty ; so effective, obviously, was Helen's drug. 
At last Athene put him in mind of his return and he begged Menelaus to 
speed him thence. Menelaus made the famous answer: "Cherish the 
stranger in the house and speed him so soon as he has the mind," which 
Pope translated : "AVelcome the coming, speed the parting guest." 

As a farewell gift Menelaus gave Telemachus a double cup and a silver 
mixing bowl; Helen, a garment of her own needleworking that glittered 
like a star. "This," she said, "is to be my gift, dear child, a keepsake from 
Helen's hand for your bride to wear on the day of expectation, your wed- 
ding day ; till then lay it up with your mother in your house." 



270 University of California PuhUcations in Classical Philology 

Thus the experienced woman speaks to the very young man ; "dear 
child," she calls him. And, "let your mother keep the robe, lest you give it 
on impulse to the first pretty girl you meet. For I see that you are sus- 
ceptible ; the way you look at me shows that. I have often seen that look." 
Telemachus, of course, does not read her meaning quite so precisely ; 
but he must feel vaguely that he has been put in his place. He does not 
fail to report to his mother that he has seen "Argive Helen, for whose 
sake the Greeks and Trojans bore much travail by the gods' designs." 

Throughout the scene, Helen must be aware of more than Menelaus 
and Telemachus hear in her words. In speaking of her "worthless self" 
she is indulging in an irony akin to the Socratic, a dissimulation not, as 
with him, of ignorance, but of worthlessness. It was not her way to 
think of herself as worthless. Her view would be that of the old men of 
Troy, who commented as she passed : "Small blame is it that Trojans 
and . . . Achaeans should for such a woman long time suffer hardships; 
marvelously like is she to the immortal goddesses to look upon." And in 
this scene at Sparta she blames Aphrodite for her infatuation ; she her- 
self was not responsible. 

Equality of the sexes is again fundamental in the second of the great 
comedy scenes. This is symmetrical with the first, for now the naive 
young girl falls in love with the mature, worldly-wise, and already mar- 
ried man : Nausicaa with Odysseus. The implications are now clear. 

Odysseus, after his long and tempestuous voyage, after the wreck of 
his raft and his two days' wandering in the swell of the sea, swam ashore 
at last in the land of the Phaeacians. He crept beneath twin olive trees, 
where Athene shed sleep upon his eyes. 

To Nausicaa, beautiful daughter of Alcinous, king of that land, 
Athene appeared in a dream, suggesting that she go a-washing and re- 
minding her that her marriage day was near at hand. She begged of her 
father mules and wagon that she might take the goodly raiment to the 
river mouth. She was ashamed to speak of glad marriage, but he saw 
all and grudged her nothing. When she and her maidens had finished 
treading down the garments in the trenches and had spread them out 
along tlie shore, they fell to playing at ball. The princess, by chance, cast 
the ball into the river, whereat they all raised a piercing cry. Odysseus 
awoke and crept out from under the coppice. The girls fled in terror; 
only Nausicaa stood firm. To her he spoke a sweet and cunning word : 

"I supplicate thee, queen, whether thou art a goddess or a mortal ! 
Have pity on me. . . . Direct me to the city . . . Give me an old garment 
to cast about me . . . And may the gods grant thee all thy heart's desire : 
a husband and a home and a mind at one with his may they give — a good 
gift, for there is nothing higher and nobler than when man and wife 
are of one heart and mind in a house, a grief to their foes, and to their 
friends great joy, but their own hearts know it best." 



Hart: High Comedy in the Odyssey 271 

Doubtless Odysseus, as he speaks these beautiful words concerning 
wedded life, is thinking of Penelope. But doubtless also in thus praying 
the gods to provide for the princess a husband and a home he is seeking 
tactfully to convey to her the idea that he is not a candidate for her 
hand — much as Helen indirectly warns Telemachus. Nausicaa, however, 
does not seem to understand the warning. She gave him the raiment he 
desired, and when he was clad in it Athene made him greater and more 
mighty to behold, and the princess marveled at him and said to her fair- 
tressed maidens : 

''Erewhile this man seemed to me uncomely, but now he is like the 
gods. Would that such an one might be called my husband, that it might 
please him here to abide." Then she called on Odysseus : ''Up now 
stranger, that I may convey thee to the house of my father. But when 
we set foot within the city, do thou linger in the fair grove of Athene. 
I would avoid the ungracious speech of the people, lest some one of the 
baser sort might meet me and say : 'Who is this that goes with Nausicaa, 
this tall and goodly stranger ? Where found she him ? Her husband he 
will be . . . for verily she holds in no regard the Phaeacians here in this 
country, the many men and noble who are her wooers.' " 

Thus she flatters Odysseus, informs him that she is much sought in 
marriage, naively speaks of him as a possible husband. So Desdemona, 
when Othello had told her the story of his life, bade him, if he had a 
friend that loved her, he should but teach him to tell such a tale and that 
would win her. "Upon this hint" Othello spoke. Odysseus could not 
"speak." But he had foreseen the hint ; he had asked the gods to grant 
Nausicaa a husband and a home. 

The delicate irony of the scene is comparable, too, with Viola's wooing 
of the unconscious Orsino, or Rosalind's of the unconscious Orlando. 
And it suggests that charming story of the Princess Royal, who, being 
perhaps of Juliet's age, informed Lord Kitchener that she had chosen 
him for her husband. Whether his tact equaled that of Odysseus is not 
known. 

Odysseus, who appears only as a shipwrecked stranger, is a king at 
home in Ithaca; Nausicaa is a princess ; there is no doubt of social equal- 
ity. But that that is characteristic of Phaeacian society, that indeed the 
woman's position is even more important than the man's. Homer takes 
special pains to make clear. Nausicaa directs Odysseus, if he wishes to 
insure his return to his fatherland, to pass her father by and throw his 
hands about her mother's knees. The King himself had expressly de- 
creed : "So long as I am to have lordship and life amongst you ... be it 
understood that the word of the Queen holds good." And Athene, too, 
tells Odysseus that Arete is worshiped by the people and that "she will 
resolve the disputes of those for whom she has countenance, even when 
the affair is an affair of men." 



272 University of California Publications in Classical Philology 

Odysseus and Nausicaa part at the city gates. Odysseus is received 
with generous hospitality : there are games in his honor, and feasting ; 
the minstrel sings ; gifts are prepared ; yet the hero has not yet made 
himself known. He is bathed and anointed and dressed in a rich robe 
and tunic. "And he went out from the bath-house to join the men at their 
wine-drinking. On the way, by a pillar of the massy roof, stood Nausicaa 
in her god-given beauty, admiring Odysseus with all her eyes : and she 
spoke to him winged words :* 'Farewell, Stranger, and when in your 
native land think of me, sometimes ; for it is chiefly to me that you owe 
. . . your life.' Odysseus answered her saying, 'Nausicaa, ... if Zeus . . . 
the Thunderer wills that I reach home, . . . then and there will I pay vows 
to you as to a Divine One ; forever and ever throughout all my days. For 
you gave me life. Maiden.' " 

Nausicaa, then, knows that he is not for her ; this is a last farewell. 
She does not know that he is the great Odysseus; he has not yet told the 
tale of his wanderings. She in her ignorance and innocence has been 
directing the experienced man of the world, the lover of Circe and 
Calypso, the husband of Penelope, with whom never a man could match 
in wisdom, Odysseus of many devices. Only we and the hero can be aware 
of the irony. 

The motif of the married hero rescued by the maiden who desires 
him for her husband doubtless existed independently before Homer's 
day. It is to be found in many a folktale and not infrequently in the lit- 
erature of art. In all the versions there is an implication of regret that 
the maiden cannot have her way, and some of the more primitive permit 
the hero to have two wives. Musaus, a reteller of folktales in the gaily 
cynical eighteenth-century manner, provides the Pope's dispensation for 
the bigamous marriage, but doubts, he says, whether this is a favor to the 
hero or a punishment for his sins.' In our own time, in Kipling's Without 
Benefit of Clergy, the hero has, it is true, no wife, but is prevented from 
marriage by other reasons. The death of the heroine is the only possible 
way out. But neither marriage nor death would be a proper ending for 
the episode in the Odyssey. Homer's solution, which is no solution, un- 
satisfactory as it may seem, is the only possible conclusion. 

Equality of the sexes is clearly implied in the household at Sparta ; 
in the land of the Phaeacians it is expressly emphasized ; j-et in neither 
case are the individuals precisely on the same level : Helen, the daughter 
of a God, is in wisdom, in insight, in humorous comprehension of a situa- 
tion, superior to Menelaus and to Telemachus ; and while there is little 

* Here I recur to the Butcher and Lang translation. Winged words, as Professor 
Calhoun has shown, are not merely words that flv from one to the other, but words 
spoken with unusual emotion or intensity. "The Art of Formula in Homer," Classical 
Philology, XXX, 215 ff. Shaw's "words came and she addressed him directly" is 
clearly not so good. 

■^ J. K. A. Musaus, Mclcchsala, in VoRsmarchen der Dculschcn, 1782-1786. 



Hart: High Comedy in the Odyssey 273 

to choose between King Alcinous and Queen Arete, Odysseus, most ex- 
perienced of men, is naturally superior in understanding to the naive 
Nausicaa. In these instances only the woman or the man is aware; 
woman and man cannot laugh in concert. 

It is only in the relations of Athene and Odysseus that this ideal of 
high comedy is achieved. Only a goddess can meet Odysseus on equal 
terms ; only an Odysseus can so meet Athene. If the accident of birth 
gives her a certain superiority, it is no more than is needful for her part. 
The exercise of her divine powers serves only to heighten the effect of 
the hero's easy familiarity with her. Homer's three high-comedy scenes 
form a pattern not unlike that in the "Marriage Act" of Chaucer's 
Canterbury Tales: in The Wife of Bath's Tale, the husband leaves an 
important decision to his wife and is thereupon made happy; in The 
Clerk's Tale, Patient Griselda is wholly governed by her husband's will 
and after much suffering finds content ; in The Franklin's Tale, Dorigen 
and Arveragus agree to obey one another and achieve a perfect union. 

The relation of Odysseus and Athene culminates in a delightful 
comedy scene in the thirteenth book of the Odyssey, but in both Odyssey 
and Iliad there is evidence that they have long been interested in one 
another. He has been the object of her especial care.* And with good 
reason, for of all the heroes he most resembles her. She is inaccessible to 
love ; he is not amatory ;* and in their relation there is no passion," She 

^ In the Iliad, Diomedes declares that Pallas Athene loves Mm (X, 245); Aias, 
that she from of old like a mother stands by Odysseus' side (XXIII, 782 f.) ; she 
chooses him as best fitted to recall the retreating Greeks (II, 169). In the Odyssey, 
Nestor tells Telemachus that Athene had singled out his father with loving care ; never 
had Nestor seen such open affection on the part of the gods as was there displayed by 
Pallas, who would stand openly by his side (III, 221 f.). 

Invisibly, or openly in her own person, or in the character of another, the Goddess 
aids and comforts her favorite. She speaks for him in assemblies of the Gods ; fights 
for him in battle ; saves him from drowning ; sheds upon his eyelids restoring sleep ; 
conceals him, at need, in a mist, or transforms him into a beggar, or gilds his head 
and shoulders with nobility; insures for him a hospitable reception by strangers; 
encourages him by her praise; judges him to be more worthy than Aias to receive 
the arms of Achilles; yet is careful not to show her favor when it might offend 
Poseidon. She is mindful also of his wife and son. 

" Odysseus is involved in no quarrel about women, prefers no woman before his 
wedded wife. He is selected to return Chryseis to her father, Chryses. He becomes 
the lover of Circe only on the urgent advice of Hermes, to win ultimate escape. In 
the matter of Calypso he had no choice ; and even to the Goddess herself he ventures 
to say: "I do most surely know how far short of you discreet Penelope falls in 
stature and in comeliness. For she is human ; and you are changeless, immortal, and 
ever young. Yet even so I choose — yea, all my days are consumed in longing — to 
travel home and see the day of my arrival dawn." And doubtless Odysseus is thinking 
of Penelope when he tells Nausicaa of the happiness of married life. He is speaking 
to a young girl, and his words are appropriate to the occasion; but there is no reason 
to doubt his sincerity. 

^» Homer speaks often of Athene's "heart's friendliness" toward Odysseus, of "the 
love she bore him." But it is abundantly clear that her affection was of a purely 
spiritual character. Athene was no Circe or Calypso. Her relation with Odysseus was 
not that of Aphrodite with Anchises; it was, more nearly, a communion of soul 
with soul. 



274 University of California Puhlicaiions in Classical Philology 

is Goddess of Power, in her chief aspect armed and warlike -^ Odysseus' 
valor is worthy of his patroness.^ 

Athene is Goddess of Wisdom; but her wisdom, in Homer, is of a 
practical, humane, shrewdly human sort, and not without humor. She 
has even at times a certain playfulness, a puckish quality that would 
scarcely be at home in the solemn clubs that bear her name in Boston and 
in London. She is not the immobile and austere figure designed by Phid- 
ias to adorn his Parthenon ; she is rather, one imagines, to be visualized as 
the light and graceful Athene Victory of the little temple on the Acropo- 
lis : the Goddess, it seems, in swift movement, has paused for an instant, 
her garments still a-fiutter, and bends to fasten a sandal that has come 
untied. 

However, it is not possible to visualize Athene as in herself she really 
is. The Goddess eludes us, as a goddess should. For she is not content 
with the varieties of her own character; she delights to impersonate 
others, to assume the appearance appropriate to the part, to invent a 
story to go with the character, to enjoy the ironic situations which in- 
evitably arise." 

If in Athene power and wisdom are blended, so they are in Odysseus. 
He is shrewd, reflective, not likely to act from anger or on any impulse. 
He is likely to consider two possible courses of action before committing 
himself to either." Even as he hesitates to follow the first prompting of 

" As represented in the familiar statues, -with helm and spear and shield, or as 
described in Iliad, V, 732 ff. She fights even with Ares, smiting him on the neck with 
a huge stone (XXI, 400 ff.). With her spear she would daunt the bravest Avarriors 
(Odyssey, I, 100). She joins Odysseus in his great last battle Avith the suitors, proving 
his force and fervor by her Avords: "Hoav are your strength and manhood fallen, 
O Odysseus, since . . . you battled Avith the Trojans, . . . and slaughtered them by 

heaps in the deadly struggle Hither, dear heart ; stand bv me and watch my 

work." (XXII, 226 ff.). 

^- Odysseus Avas not named among those most Avorthy to meet Hector in single 
combat; and he held himself inferior to Achilles AA-ith the spear. Yet when left by 
the Avounded Diomedes, standing alone among the Trojans, he kills or wounds or puts 
to flight all Avho venture to attack him. (Iliad, XI, 401 ff.) 

'•■'In Odyssey, I, she appears to Telomachus as Mentes, and at length almost de- 
ceives the reader, so convincing is the corroboratiA-e detail of tier story — the voyage 
and its purpose, the sparkling iron ore, the copper, the position of the ship, the 
grumbling of the impatient creAv. Later, she impersonates Mentor ; then Telemachus, 
mustering his crcAA'. Again as Mentor (III) she accompanies Telemachus to Pylos, 
hears Nestor speak of her loving care for Odysseus, and declines his invitation to 
spend the night in his house. She must, she says, lie by the ship to hearten the crcAV 
and.in the morning push on to the estimable Cauconians, Avho liave long owed her no 
small sum. At Scheria she appears to Nausicaa in the guise of a playmate (VI, 22). 
She met Odysseus at the entrance of the towni in the likeness of a young girl bearing 
a water jar (VII, 20). In the character of the King's herald she summoned the 
Phaeacians to the council (VIII, 8), and later encouraged Odysseus in the weight- 
throAving contest (VIII, 193 ff.). 

" As, Avhcn left alone among his enemies, Avhether to flee or stand and be taken 
(Iliad, XI, 401 ff.) ; after landing in Scheria, Avhcther to remain for the night down 
by the river or climb the slope to the dark Avood (Odyssey, V, 465 ff.) ; Avhen awakened 
by the cries of the maidens, whether to clasp Nausicaa's knees in entreaty or stand 



Hart: High Comedy in the Odyssey 275 

his own mind, he hesitates to believe what others tell him. His skepticism 
is thoroughly characteristic ; it manifests itself in distrust even of the 
Gods themselves/" 

Odysseus, doubtless, would suspect his fellow men, conscious always 
that his own words were often false. He is preeminently an inventor of 
tales. He is in this way a self-conscious artist. He thinks highly of the 
professionals.^* 

He himself dislikes to tell the same story twice. ''It goes against my 
grain," he says, "to repeat a tale already plainly told" (XII, 452 f .) . 

Odysseus has not Athene's power of transformation; only she can 
make of him now a ragged beggar, now a hero, tall and strong, with 
hyacinthine locks. Yet, like the Goddess, he is a skilled impersonator, and 
like her he delights to invent stories appropriate to the occasion on 
which he tells them and to the characters that he assumes. Indeed, in the 
number and elaboration of these tales he far outdoes Athene herself. 
Many are told after his arrival in Ithaca." He had not yet told these 
false tales in the semblance of truth when Athene calls him a "plausible, 
various, cozening wretch" who delights in "crooked and shifty words" 
and in "speaking in character" (XIII, 291 ff.). "It was thanks to me," 
she says, "that you were welcomed by the entire society of the Phaea- 
cians." The implication seems to be that the long tale of adventure with 
which he charmed Alcinous and his court was also a piece of fiction. 
There are no other earlier tales. 

Arete and Alcinous have had their doubts, but in the end choose to 
believe the story, partly because of Odysseus' appearance, partly be- 
cause of the excellent form of the narrative. "We will not be persuaded 
that you are a pretender or a thief, like those many vagrant liars our 
dark earth breeds to flourish and strut behind so thick a mask of false- 
hood that none can pierce it to read their worth" (XI, 363) . Yet Athene, 
who knows Odysseus best, seems to say that he had, precisely, been strut- 
ting behind a mask of falsehood at the Phaeacian court, a mask that she 
herself had constructed for him. We also, then, may take her view and 
add the Phaeacian story (IX-XII) to the list of his achievements in 
fiction. 

off and cajole her with honeyed words (VI, 142 f.) ; whether to kill Irus with a single 
blow or to tap him gently (XVIII, 90) ; whether or not to explore the island of Circe 
(X, 151). 

^° "Surely," he tells Calypso, "something not at all to my advantage lies behind 
this your command that on a raft I launch out over the sea" (V, 173). He fears that 
Ino's command to abandon the raft may be some new snare (V, 356). 

" "It is right," he says, "that bards should receive honor and reverence from every 
man alive, inasmuch as the Muse cherishes the whole guild of singers and teaches 
to each one his rules of song" (VTII, 479 ff.). He lauds Demodocus above all mortal 
men, praising his history of the mishaps of the Achaeans as accurate, complete, 
realistic, convincing (VIII, 486 ff.). 

"He has a tale (XIV) of his own adventures for Eumaeus (and overcomes with 
difficulty the swineherd's skepticism) ; a similar but shorter tale for Antinous 
(XVII) ; another for Penelope (XIX) ; another for Laertes (XXIV). 



276 University of California Puhlications in Classical Philology 

Odysseus, with the gifts of the Phaeacians, was put ashore in his sleep 
on the island of Ithaca. When he awoke he was ignorant of his where- 
abouts, and so he lamented, not knowing what to do with his wealth or 
with himself. Athene drew nigh, seeming a young man, some shepherd 
lad, but gentle and dainty like the sons of kings when they tend sheep. 
She had gathered her fine mantle searflike round her shoulders and 
carried a throwing-spear ; on her lovely feet were sandals. Odysseus went 
forward with a swift greeting, asking in what land he was. A well-known 
land, she answered : "Stranger, the name Ithaca is rumored abroad, 
even to Troy, which is said to be so far from our Achaean coasts." 

Her word made his heart leap. Yet he swallowed back the words that 
were on his lips to make play with his instinctive cunning. He had heard 
of Ithaca, he said, even in his home in Crete, whence he had fled, having 
killed a man . . . 

As he was running on, the Goddess broke into a smile and petted him 
with her hand. She waxed tall ; she turned womanly : she was beauty's 
mistress, dowered with every accomplishment of taste. She spoke to him 
winged words : "Any man, or even any God, who would keep pace with 
your all-around craftiness must needs be a canny dealer and sharp- 
practised. plausible, various, cozening wretch, can you not even in 
your native place let be these crooked and shifty words which so delight 
the recesses of your mind? Enough of such speaking in character be- 
tween us two past-masters of these tricks of trade — you the cunningest 
mortal to wheedle or blandish, and me, famed above other gods for 
knavish wiles. And yet you failed to recognize in me the daughter of 
Zeus, Pallas Athene, your stand-by and protection throughout your 
toils!" 

Fluently Odysseus answered : "Your powers let you assume all forms, 
Goddess, and so hardly may the knowingest man identify you. Yet well 
I know of your partiality towards me . . . But surely I am not in clear- 
shining Ithaca?" 

Said Athene : "Your mind harps on that, and I cannot leave on tenter- 
hooks one so civil, witty, and shrewd. Any other returned wanderer 
would have dashed home to see his children and his wife. Only you choose 
to be sceptical ..." 

She showed him the familiar landmarks, aided him to conceal his 
treg^sure in a cave, plotted with him the doom of the extravagant wooers, 
told him to wait at the house of the swineherd Eumaeus — "until I have 
recalled Telemachus, who went to the house of Menelaus, trying to find 
out if you were still alive." 

He said to the Goddess : "Why did you not tell him so much, out of 
your all-knowing heart? Must he, too, painfully roam the barren 
seas . . . ?" The grey-eyed one replied : "Take it not so much to heart . . . 
He suffers no hardship." 



Hart : High Comedy in the Odyssey 277 

In this scene, Odysseus and Athene manifest themselves clearly, fully, 
and engagingly to one another and to the reader. Their relation, we are 
reminded, had been of long standing. It had become intimate and in- 
formal: on his part, appreciation of her powers, gratitude, respect 
without awe : he now does not hesitate to express doubt of her veracity 
and to reproach her for sending Telemachus on a futile errand. This, 
however, she takes in good part, for she admires her hero, finding in him 
qualities like her own. Her affection shines through the scolding which 
is also praise : "0 plausible, various, cozening wretch . . . !" 

Odysseus manifests his characteristic propensity to dramatize a part 
for himself and to tell a tale. It must be a new tale ; he has said that he 
hates to repeat himself. And this one must be improvised on the spur of 
the moment. But on this occasion, ironically enough, he has an auditor 
who knows the truth, who cannot help smiling at his inventive fluency. 

And Athene, manifestly, had delighted in her impersonation of the 
gentle and dainty shepherd lad, a very Florizel, Prince of Bohemia, com- 
plete with sandals and throwing-spear. The same delight in playacting — 
which leads her to construct, as a dramatist might, the plot for action 
against the suitors — is evident in her transformation or make-up of 
Odysseus at the end of the scene, with the complete costume and all the 
properties of the ragged beggar, realistic and convincing. Odysseus will- 
ingly accepts the part and plays it well. 

In the dialogue one may venture to see a foreshadowing of the merry 
war between Benedick and Beatrice in Much Ado about Nothing, 
wherein skirmishes of wit conceal and yet express mutual interest and 
affection. The resemblance to Odysseus and Athene is not close. "What 
is remarkable is that Homer had so early, under social conditions so 
different, discovered the possibilities of high comedy in such a pair of 
evenly matched and lively intelligences. 

The Goddess's impersonation of the gentle shepherd may, further- 
more, suggest the male disguises of Rosalind and Viola, or, better, of 
Portia, who, if she lacks the divine power, yet has the human cleverness 
to save the life of her husband's friend. If she lacks the omnipotence of 
a Goddess, she is yet the mistress in her own right of a great estate, who 
commands the obedience of all those about her. If not able to transform 
herself, she can yet impersonate with complete success the young and 
learned doctor of Rome, Balthazar, a Daniel come to judgment. Like 
Athene, she delights in the success of the deception : her gay and teasing 
triumph over her discomfited husband makes the delightfully ironic 
closing scene of the play. 

For thus Homer does seem to foreshadow Shakespeare. The greatest 
of epic poets and the greatest of dramatic poets are interested in the 
same kinds of characters, plots, motifs, situations, human relations. Both 
like to leave the stage to their personages, allowing them to speak for 
themselves. 



278 University of California Publications in Classical Philology 

In comedy, Homer foreshadows Shakespeare's range and variety, and 
awakens the most thoughtful laughter when women have social equality 
with men. 

Both delight in the appearances which contrast with reality and in 
pointing up ironic situations by ironic dialogue. The works of both say 
more than they seem to, as Voltaire remarked of his Zadig. 

Consideration of the comic seems to deepen one's impression of the 
essential humanity of Homer, his tolerance, his kindliness. 

If the High Comedy in the Odyssey leads us to think of Shakespeare, 
it leads us to think also of life itself, of the ways of people whom we 
know : the screams of Nausicaa and her girl friends when their ball is 
thrown wild and falls in the river ; Helen, as with women of all time, 
being first to see in Telemachus the resemblance to his father; Athene's 
very human delight in Odysseus' failure to recognize her in the hand- 
some shepherd lad. 

Merely to name these three — Nausicaa, Helen, Athene — is to set them 
forth as individual women, each with a character in her own right. I 
recall George Meredith's dicta : "When the veil is over women's faces, 
you cannot have society, without which the Comic Spirit is driven to 
the gutters of grossness to slake its thirst" ; and, "There never will be 
civilization where comedy is not possible, and that comes of some degree 
of social equality of the sexes." Today it is well to place Hitler's words 
beside Meredith's. "Hitler said : 'The program of our national socialist 
women's movement has only one point. That point is called the child."* 
That has meant [Deuel explains] the almost total banishment of women 
from public office, their withdrawal from professions in favor of cannon- 
foddering and such 'womanly'. vocations as domestic service." 

From Homer to Hitler ; from the beginning of that common culture 
which has drawn the peoples of the West together, to what threatens 
to be its end ! 

■1^ Wallace R. Deuel, People Under Hitler; review in Time, March 9, 1942, p. 84. 



THE BIOGRAPHICAL FASHION IN LITERARY 

CRITICISM 

BY 

HAEOLD CHEBNISS 

Several years ago a professor of English literature was widely ac- 
claimed for having made an important discovery in his field of research ; 
he had found in certain English archives the record of sale of a house 
belonging to John Milton. This may seem like a parody of what I call 
the biographical fashion in literary interpretation rather than like a fair 
example of that fashion ; but I suspect that few professors of Greek 
literature see anything comic in their scholarly debates concerning the 
number of Euripides' wives, the question of Sophocles' indictment of his 
son, lophon, and the reasons for Aeschylus' removal from Athens. None 
of these questions, however, affects the works of these poets or our under- 
standing of them. Neither has the discovery in the English archives 
elucidated a single word or line in the writings of John Milton ; and yet 
that discovery, far from being taken lightly by any of the journals that 
announced it, was treated by all of them as an addition to our knowledge 
of the history of English literature. It is a detail added to our knowledge 
of the life and movements of an English author; and it is this that made 
the discovery impressive. It increases our knowledge not of any of the 
literary productions which make John Milton's name significant, but of 
the man himself. This is the reason why the discovery aroused so much 
interest ; and it is because of the implied attitude toward the relationship 
between the author and his works that the event epitomizes the biograph- 
ical interpretation of literature. 

History itself, political and social, as well as that natural human curi- 
osity which at its basest is scholarly scandal-mongering and at its best 
authentic interest in the human personality, must be distinguished from 
the study of literature. Certainly, the historian cannot with impunity 
refrain from employing literary monuments as instruments in his task of 
reconstructing and interpreting the character and activities of a people 
or period. Since the proper use of such materials requires comprehen- 
sion of the peculiar nature of artistic production and recognition of the 
way in which it differs from documentary and material evidence, the 
historian too must have the capacity to understand and interpret litera- 
ture. Since every truly artistic production is a historical anomaly, how- 
ever, the historian applies his literary comprehension with a purpose 
diametrically opposed to that of the philologist. He is concerned to 
comprehend the individuality of a work of art only in order that he may 
eliminate it and so extract for use as historical evidence those elements 

[279 ] 



280 University of California PuUications in Classical Philology 

which are not the private creation of the author. The student of litera- 
ture, on the other hand, studying the same text is interested in it as a 
separate and unique phenomenon ; his interest in the common elements 
which it contains is in turn a kind of interest of elimination, for he is 
concerned with the manner in which they have been individualized by 
the artistic form. The historian who uses Thucydides as a source for the 
study of the Peloponnesian War approaches the speeches with the prob- 
lem put by Croiset : "The difficulty then is to know in what degree it is 
Pericles or Cleon, in what degree it is Thucydides whom we are hear- 
ing" ;' and his reason for drawing this distinction is his proper desire to 
recapture so far as possible the words of Pericles or Cleon. The atti- 
tude of the student of literature toward these same texts is admirably 
formulated by Mr. G. F. Abbott when he says: "The question of the 
authenticity of the speeches . . . does not enter into the present criticism, 
which is entirely concerned with their literary treatment. In shaping 
them as he thought fit the historian made them his own, no less than 
a dramatist does when he reproduces a historical speech, even though 
he may give, wherever possible, the very words which had been used. 
The authenticity of some of the speeches in Shakespeare's historical 
plays does not affect their significance as specimens of Shakespearean 
art.'" The book of Thucydides purports to be a history; it is at the same 
time a work of literary art, even though its author claimed for it only 
the value of "usefulness" (I, 22). Consequently it presents an object of 
study to three different disciplines at least (for I omit the dubiously 
philosophical discipline of statecraft) : the historian can use it as a 
source book of events, a repository of material from which, along with 
other evidence, to reconstruct a portion of the past ; the student of his- 
toriography may treat it as an example of the recording and interpreta- 
tion of human activity, criticizing the method which it exemplifies and 
testing wherever possible the accuracy of its records and the propriety 
of its interpretations ; and the student of literature will study it not as 
a source of information about the Peloponnesian War or as an example 
of historical method but as a piece of literature, a drama in prose more 
dramatic than Hardy's Dynasts. This does not mean that one should 
approach Thucydides by one of these disciplines to the exclusion of the 
others; it does mean that they should not be confused, that one should 
understand that the study of history is not the study of literature even 
when the same text is the object of both studies and even though the 
historian, in handling that text, employs the methods of philology. Into 
the old and rancorous debate concerning the relationship of philology 
and history it seems to me unnecessary to enter further. The historian 
from his own point of view is justified in considering philology an an- 

1 A. Croiset, Histoire de la liitfraiure grecqve, Tome IV (Paris. 1911), V-^^''- 
" G. F. Abbott, Thucydides, A Study in Historical Reality (Londoa, 1925), p. 184, 
fn. 1. 



Cherniss : The Biographical Fashion in Literary Criticism 281 

ciliary discipline. That does not mean, however, that it is in essence 
ancillary to history or any other study, any more than mathematics is 
deprived of its essential autonomy by the fact that it is ancillary to all 
the natural sciences ; and those philologists who in the desperation of 
self-defense have insisted that philology and history are identical have 
transgressed the most elementary requisite of their own study, the dis- 
tinction of differences of meaning. 

Philology is ultimately the study of literature for its own sake, and 
everything that the philologist studies as a philologist must be directed 
toward this as toward its final cause. His purpose is to comprehend and 
to interpret — that is, to help others comprehend — as works of art the 
individual productions which comprise the literature of which he is a 
student, in the case of the classical philologist the works of Greek and 
Latin writers. This general formula can hardly be challenged in its 
generality. What is given to the philologist is books, the significant ex- 
pression of thought in artistic form ; only the value of this expression 
and this form has created philology as a study necessary to the spiritual 
life of man, and only this value justifies the continued existence of that 
study as an autonomous discipline. The works of literature are not hy- 
potheses which the philologist uses like the rungs of a ladder on which 
to climb to some higher principle; they are the avvirodeTOL apxal them- 
selves, the clear and direct comprehension of which is the final purpose 
of the student of literature. Not the identity and biography of the artist, 
then, but the unique personality of the literary artifact is the proper 
object of his study. When we proceed to specify this general formula, 
however, to designate the practical procedure by which it is to be realized 
in particular cases, all that we have excluded as distinct from the study 
of literature and extraneous to it seems to require acceptance again as 
essential to the appreciation of each and every particular work. 

If a work of literature is the significant expression of thought, to 
appreciate the work is to understand the significance of the thought 
which it expresses ; and how is this possible without intimate knowledge 
of the physical and spiritual environment, of the political, economic, and 
social conditions in which the thought was formulated and to which its 
expression responded ? A work of art is produced at a definite time, in 
a definite place, and for an audience which itself has certain tastes and 
conventions, accepted ways of thinking, and a common store of knowl- 
edge and belief, all of which the artist takes for granted. These are as 
much the material with which he works as is the language in which his 
thought is expressed or the marble from which his statue is carved. 
Must one not, then, in order to understand a literary production, make 
oneself a member of that original audience to which it was addressed, 
learning what they knew, thinking and feeling as they thought and felt, 
and therewith acquiring the ability to slough off one's own environment, 



282 University of California Publications in Classical Philology 

knowledge, and tastes so far as they are at variance with those others ? 
Really to appreciate a comedy of Aristophanes must one not first trans- 
form himself into an Athenian of the generation of Aristophanes ? This 
should be the necessary conclusion of the historical interpretation; and 
Professor "Wilamowitz drew it with almost complete thoroughness in 
the introduction to his edition of the Lysistrata. There, by way of ex- 
plaining why no one had hitherto succeeded in making a satisfactory 
commentary on Aristophanes, he wrote as follows : "One first becomes 
properly aware of it (that is, of what besides mere jollity there is in 
the poetry of Aristophanes) when with a historically trained eye one 
sees how this mad sport affected on that single day the thousands who 
were its audience, a whole people that still constituted a single society 
in which everyone knew everyone else and each felt himself to be a mem- 
ber of the community. One must also know the conventions of this people, 
its deportment, its thought and belief in work and leisure, in sorrow and 
joy . . . And also it must have come to be felt as natural that Dionysus 
can appear as a comic character in his own sanctuary into which his holy 
statue has been brought in solemn procession, that his priest looks on 
and that the belief in the mighty god remains alive in the hearts of the 
people. Finally the poem that was designed for an hour must first be 
understood as that which it pretended to be at the moment, before it is 
considered with a view to its absolute worth. All this, however, was im- 
possible so long as Athens was as good as unknown, so long as the earth 
covered the theater and the innumerable monuments of Attic art and 
Attic life. Today one need hardly say that no one who is not intimate 
with vase-painting can understand Aristophanes from within. . . . Simi- 
larly the events and personalities of Athenian history must have been 
so thoroughly studied that one can enter into the transitory mood of 
each year and so that the forms of public life, which were immediate 
data for the poet and audience, remain ever present for us too without 
special reminder."* 

The Hindus, who excelled in epigram, put this theory more succinctly 
in their proverb : "He who eats beef can never learn Sanskrit." 

Furthermore, the conventions and habits of thought, the political and 
social environment of the artist were themselves the result of long de- 
velopment, were determined by the events, the thoughts, the people that 
preceded them, and cannot be apprehended more than superficially with- 
out knowledge of these conditions whence they derived. Does not the 
rule of Aristotle hold here also : we know a thing only when we know its 
cause? The single literary artifact, too, did not spring (i7r6 5pi'6joi'5' 
dTTo TrerpTjs.The artist has employed, consciously or unconsciously, the 
form and expression of a definite tradition, and even his deviations from 
these and his innovations are intelligible only to those who know the 

* Aristophanes, Lysistrate erklart von U. von Wilamowitz-Moellendorff (Berlin, 
1927), p. 5. 



Cherniss : The Biographical Fashion in Literary Criticism 283 

course from which he has deviated. A long series of modern studies tes- 
tifies to the assumption on the part of Hellenists that only knowledge 
of the origin and development of the form of comedy can explain— or 
rather, as they seem to feel, excuse — the position of the parabasis in 
the plays of Aristophanes/ So, in order to understand the structure and 
essence of Greek tragedy and comedy it is thought to be necessary to 
discover the original form of Greek drama whereby will be exhibited, 
as Kranz has put it, "the truly constitutive element of these artistic 
productions, the character which decisively determined their develop- 
ment.'" The appreciation of a piece of literature, then, requires not 
merely the reconstruction of the environment in which it was produced, 
the ability to transform oneself into a contemporary of the author, but 
also historical knowledge, as complete as may be, of the processes by 
which all the elements of that environment were produced and of the 
origin and development of the artistic form in which the author has 
expressed his thought. 

Yet even this does not suffice. The environment, physical and spiritual, 
the language and the traditional artistic form, developed to the point 
at which the artist found them, were there for innumerable men to use ; 
but the work of art itself was produced by only one among all these men. 
To him alone, the individual author, is due the artifact in its individu- 
ality ; and since it is the product of an individual personality, since as 
effective influences and as significant constituents of the product all the 
elements of environment, convention, and tradition have been distilled 
through this personality which has gradually assimilated them to itself, 
the work of art can be understood and appreciated only as a moment of 
the personality which created it, a photographic exposure of a single, 
irretrievable instant in the organic development of the artist. The neces- 
sary consequence of this conception of literature too was drawn for 
philology by Professor Wilamowitz in the introduction to his study of 
Plato." "The philologist," he there says, "is once for all an interpreter, 
but not interpreter of the words alone. Them he will never completely 
understand if he does not understand the soul from which they come. 
He must be the interpreter of this soul also, for, since the whole art of 
biography is founded on interpretation, biography is, in the true sense 
of the word, the work of the philologist, only raised to a higher power. 
Yet the task stands no higher than to understand how this man has come 
to be, what was his intention, his thought, his effect." And a little later, 
in generalizing his own method in this book, he says : "The biographer 
proceeds from work to work, from interpretation to interpretation, 
always seeking the author behind the book. If a human being stands out 

* See the resume by Harsh, T.A.P.A., Vol. LXV (1934), pp. 178-179. 

^W. Kranz, Stasimon (Berlin, 1933), p.l. 

' U. von Wilamowitz-Moellendorff, Platon (Berlin, 1919), Bd. I, p. 4. 



284 University of California Publications in Classical Philology 

whom we can recognize as such, if the individual features unite them- 
selves into a single portrait which as a unit is credible, the task of the 
philologist is accomplished.'" The study of any literary work, then, must 
be based upon a study of the author's biography. In that biography the 
proper place of this work must be determined ; only as an element in 
the biography can the work be understood. From this point of view, 
however, any and every detail of information concerning the activities, 
experiences, and habits of the author is of importance, since it is from 
the accumulation of these details that his biography must be constructed. 
The slightest external incident may have precipitated the mood which 
caused the poet to write a particular piece or may have been the impetus 
to a development of character by which the nature of all his subsequent 
work was determined. Even the sale of Milton's house, the number and 
character of Euripides' wives may have been important factors in the 
lives of the artists and so in the nature of their literary productions. 
As such details increase one's knowledge of the author himself, so even 
the student of literature, who is interested primarily neither in social, 
cultural, or political history nor in the personal history and psychology 
of this human individual, must address himself to these details, since 
it is only in the light of the author's biography that the author's produc- 
tions can be adequately understood. 

Without questioning the nobility and the psychagogic advisability of 
making the scholarly ideal as lofty and rigorous as possible, and even 
though subscribing to the doctrine that no detail of knowledge is useless 
or negligible, one may still properly wonder whether the requirements 
set up by the theories of literary interpretation previously outlined do 
not justify despair in the strongest and most ambitious spirits. Even a 
Wilamowitz, with all his knowledge of vase painting, Greek institutions, 
and Athenian history, cannot think and feel as did a contemporary of 
Aristophanes and could not do so were the monuments for study at his 
disposal a thousand times as many as they are. Knowing that Dionysus 
was a god to those who still delighted in his ridiculous antics upon the 
stage, and feeling that he is a god even while enjoying him as a clown — 
between these two states is a gulf that no erudition can bridge, a gulf 
not due to time and the absence of material evidence. It is most probable 
that for Professor Wilamowitz the thoughts and feelings of his own con- 
teipporaries in England were equally unassimilable when they, staunch 
supporters of throne and constitution, met with hearty laughter and loud 
applause — and without any consciousness of incongruity — the mockery 
of the House of Peers in lolanthe, the parody of the royal prerogatives 
in the great duet of the Gondoliers, and the satirical lese-majeste in the 
denouement of the Pirates of Penzance. In the case of ancient literature 
there is not adequate material to enable the student to gain even a fairly 
complete theoretical knowledge of the environment in which were pro- 

' Ibid., p. 8. 



Cherniss : The Biographical Fashion in Literary Criticism 285 

duced most of the works with which he is concerned; to believe that one 
can know the events and personages of Athenian history so intimately 
as to be able to detect the transitory mood of each year is to deceive one- 
self by taking for objective truth the tenuous phantoms of historical 
reconstructions. Yet quite apart from the practicality of this require- 
ment, theoretical knowledge, however exact and complete, is not the im- 
mediate perception which only those can have who are themselves part 
of this environment. Furthermore, if the work of art can be properly 
understood only in this environment and from the point of view of the 
original audience, then it is a hindrance rather than a help to know the 
historical processes by which were developed the environmental complex 
and the artistic form, for in that environment and in the emotional and 
mental constitution of that audience the negative characteristics are 
factors as essential as the positive ones. It is certain that few if any of 
the original audience had this kind of knowledge, and it is at least highly 
improbable that the author himself knew the history of his art and was 
conscious of the place of his own work in that development. At any rate, 
as a means of reconstructing the environmental complex this method not 
only suffers all the disabilities already mentioned in connection with the 
recovery of the period directly from its own monuments, but it also has 
to assume the object of its investigation, knowledge of that complex 
itself. When Professor Wilamowitz asserts that a knowledge of vase 
painting is essential for comprehending Aristophanes, he forestalls the 
objection that this vase painting represents principally the life of the 
generations preceding the appearance of Aristophanes by the statement 
that in essentials the life of his generation had altered very little.* This 
comparison itself, however, implies a knowledge of the environment of 
Aristophanes based upon evidence independent of the very vase painting 
which was asserted to be indispensable for understanding that environ- 
ment. The study of origins and development, social, political, and lit- 
erary alike, consists in establishing hypothetical connections among 
events or objects, the evidence for the existence and nature of which does 
not itself rest upon these hypotheses and the knowledge of which, being 
the premises from which the method proceeds, must always be more 
certain than any conclusion which can be drawn concerning their rela- 
tionships. Moreover, as applied to literature itself this method considers 
the work from the point of view of neither artist nor audience. The poem 
or play becomes a single stage in a hypothetical historical development 
in which the author is only an accident or at most an agent of forces 
which transcend him and of which he is unconscious. The justification 
of this kind of literary history, its assumptions, and its limitations, need 
not be discussed here ; it is enough to say that it is not concerned with the 
understanding and appreciation of any single literary production as 
such but with relationships and connections which come to be taken for 

* Aristophanes, Lysistrate, p. 5. 



286 University of California Puhlications in Classical Philology 

literature itself, whereas they are external to the essence of any particu- 
lar piece of literature whatsoever. As an approach to literature it runs 
the risk of becoming the kind of method which has been characterized as 
"consisting in this : to speak not of the object but of the causes, not of the 
essence but of the relationships; not to interpret the works, but to inves- 
tigate the material, the environment, the influences ; a method which 
ever remains in the sphere of the preliminary, the irrelevant, the ex- 
trinsic, and reduces true scholarship to erudition in what is not worth 
knowing.'" 

To approach the literary work by way of the personality of the artist 
seems to be a more reasonable method, for in any artistic production this 
is the factor which is most important, in which all the others are sub- 
sumed, and through which they become operative. To this insight, I take 
it, is due the present-day popularity of the biographical method in the 
study of literature; if the work of a man must be interpreted as a mani- 
festation of his personality, then we must address ourselves to the history 
of that personality, and this is biography. The origin and environment 
of the man, his associates, activities, and experiences, these are the deter- 
minants of the personality that is the author ; surely it is only by the 
discovery and synthesis of these factors that we can restore the author 
as a human being and understand his works as the expression of specific 
moods and thoughts, the response to definite needs and influences. Never- 
theless, even if the validity of this method be granted, its practicality is 
strictly limited. The Iliad and Odyssey we should have to abandon alto- 
gether ; we have no knowledge of their author or even of the approximate 
date of their composition. And how many works of Greek literature 
could we hope to understand, if this were the sole means of understand- 
ing them ? The champion of the biographical method in the interpreta- 
tion of classical literature, Wilamowitz himself, asserted that "for no 
Greek can we write a true biography, a history of the development of 
the individual within his environment.'"" Even where we have a morsel 
of knowledge concerning the author's activity or experience, it is usually 
the merest conjecture by which this has been connected with some ten- 
dency or element of his literary expression, and yet it is this connection 
itself which is the point at issue. So, for example, the fact that Plato left 
Athens for a trip to Syracuse in 367 cannot explain the difference in 
«tyle between the first and second parts of the Theactetus, even if it be 
granted that the composition of this dialogue is correctly dated in the 
year or two immediately preceding his departure. This explanation itself 
rests upon a critical assumption of a nonbiographical nature, namely 
that the second part is in an unfinished state, and requires a further 

■• Quoted from R E. Curtius (Die literarischen Weghereitrr des ncuen FranTcreich) 
by J. Korner, Neue Jahrhiicher f. d. Iclass. AUertum, Bd. XXV (1922), p. 175. 

^^ U. von Wilamowitz-Moellendorff, Euripides, HeraUes (Berlin, 1889), Bd. I, p. 1 ; 
cf . idem, Platon, Bd. I, p. 6. 



Cherniss : The Biographical Fashion in Literary Criticism 287 

hypothesis to explain why, if this is so, Plato failed to finish it after his 
return." Since the connection between Plato's departure and the Theae- 
tetus is unknown, the knowledge of the date of his departure cannot 
facilitate our understanding of the dialogue. 

Worse still, the validity of the method is itself dubious, for it is a 
method which rests upon the tacit assumption that the sum of biographi- 
cal incidents constitutes the personality and that the essential meaning 
of the artistic expression can be identified with the environment and 
influences which have coincided to form a given moment in the person- 
ality of the author. This assumption is identical with Taine's theory 
which would make of the study of literature a kind of psychophysical 
mechanics and against which even his admirer, Zola, protested that as 
soon as the spirit, the individual personality, strikes where and when it 
will, all influences are merely accidents, the results of which one may 
study and explain but which act upon a natural element that is essen- 
tially free and that has not yet been reduced to any law." Since the 
human individual is not a mechanical combination of events and influ- 
ences, no reconstruction, however complete, of the external incidents of 
a man's life can reproduce or reveal the essence of the man himself. There 
is no such thing as an "influence" in the abstract, and any particular 
"influence," of tradition, of environment, or of personal experience, 
exists only in the individual influenced and is determined by his per- 
sonality, which is passive only in grammar and in histories of literature 
but in fact is the active element in a complex in which the brute event 
is the passive material that gains form, significance, and efficacy only 
according to the way in which the particular individual fashions it. 
Even a phrase "borrowed" by one poet from another does not have in 
the new poem the same significance that it had in the old one ; and merely 
to identify the source whence the author "borrowed" it is not even to 
begin to understand what it means in the context of the new production. 
Students of literature have given too little heed to the trenchant remark 
that it is not important what the poem is made out of but what it is made 
into ; and so too the external incidents of the author's life have meaning 
only as they are assimilated to his personality, and literary significance 
only so far as they have been transmuted by that personality into 
artistic form. 

That the neglect of this fact renders the biographical method invalid 
has not gone unrecognized by certain critics who have understood that 
no personality can be comprehended by a compilation of isolated events 
and further that as an operative artistic entity it is manifested only in 
the artistic product. One group, the circle of Stefan George, accepting 

^ U. von Wilamowitz-Moellendorff , Platon, Bd. I, p. 510. 

^E. Zola, Mes Haines, M. H. Taine, Artiste (Paris, BibliothSque Charpentier, 
1902), p. 227. 



288 University of California Publications in Classical Philology 

this axiom and identifying the artist with his work, has sought to recon- 
struct the figure of the author's personality in accordance with the 
"inner form" which they reach admittedly from his work alone but by 
a kind of intuition. "The works," says Gundolf in laying down the pro- 
grammatic principles of this school, "are not the tokens which indicate 
the life [of the artist] but the bodies in which that life is incarnate. . . . 
One has no scientific justification for investigating the life of the great 
artists outside of their art ; nay, there is no possibility of doing so, for 
what is commonly called the life of an artist and more modernly his 
experience has from the very first been assimilated to his art and is the 
same impulse and force as his work."" From members of this circle have 
come a number of studies of ancient authors, notably works on Plato 
by Singer, Friedemann, and Hildebrandt and on Posidonius by Karl 
Reinhardt; and these have exercised a large influence upon scholars in 
this country, most of whom are unaware of the tenets of the cult on which 
these constructions are based. For it is a cult of a semimystical kind, 
the object of whose reverence is the "heroic individual." With the nature 
and beliefs of this circle, however, I am not here concerned, but only with 
the elements of its method. The intuition which discovers in the writings 
of an author the "natural law" and "inner form" of his personality is 
proof against all objections, logical and philological ; but, while one must 
admit that a certain native insight, call it direct intelligence or intuition 
as you please, is required for understanding any text, it is, all the same, 
a vicious circle to intuit the nature of the author's personality from his 
writings and then to interpret those writings in accordance with the 
"inner necessity" of that intuited personality. Moreover, once the in- 
tuition of the individual critic is accepted as the ultimate basis of 
all interpretation, the comprehension of a literary work becomes a com- 
pletely private affair, for the intuition of any one interpreter has no 
more objective validity than that of any other, and each interpreter lays 
himself open to the peritrope, like Protagoras in the Theaetetus. Yet in 
this regard the circle of George differs from most modern literary inter- 
preters only in being conscious of its method, systematic in applying it, 
and outspoken in advocating it. Other scholars, not of this school, tell 
us that Plato's works must be comprehended as expressions of his life 
because he always continued to develop ;" that Sophocles had the natural 
gift of remaining unaffected by anything foreign to his own nature and 
that the instinctive assurance of the characters of his dramas was an 
endowment of his own ego;'" that Euripides felt himself more strongly 
impelled than Sophocles to take a definite attitude toward the actual 
problems of his country just because he took no active part in political 

" F. Gundolf, Goethe (Berlin, 1925) , p. 2. 

" U. von Wilamowitz-Moellendorff, Platan, Bd. I, p. 6. 

^^ Max Pohlenz, Die Griechische Tragodi-e (Leipzig und Berlin, 1930), Bd. I, p. 160, 



Cherniss : The Biographical Fashion in Literary Criticism 289 

life/" and that lie was conscious of his own individuality and so felt it 
his highest duty to himself and his people to cherish that individuality." 
Such statements refer only to the personality which the critic's intuition 
has seen in the author's works ; they have no other authority, and they 
provide no means for understanding the text which is not in the texts 
themselves; they are the results of interpretation and cannot be used as 
the basis of interpretation without falling into a vicious circle. 

To this extent Gundolf's contention is correct, that the artist qua 
artist exists only in his artistic productions ; one need not follow him 
any farther to see that biography is nothing to the student of literature, 
to whom the only thing of significance in the life or personality of an 
author is his actual literary work. The potential poet may be a proper 
subject for psychological investigation, but the concern of the philologist 
is the actual poet and he is identical with his poems. So we are brought 
back to the texts themselves. If, then, we can never appreciate them as 
the original audience did and if, again, the reconstruction of the author's 
biography cannot lead us to understand them, are we to say that true 
understanding of a work of art is impossible ? Consider first, for a mo- 
ment, what reason there could be for studying a work which had no 
meaning except for a single audience in a single spot at a single moment 
in the past ! The historian interested in the nature of that audience might 
use such a work, but only as a means of understanding the audience and 
without concern for the work itself. Only a madman would even wish 
to transmute himself into a member of that audience in order to appre- 
ciate what could have no meaning for men at any other time or place. 
If, again, a work of art should have significance only as a manifestation 
of an individual personality, it would be of interest only to a psycho- 
analyst. No one else would bother to read Sappho's poems, if they were 
only exhibits in the individual history of a neuropathic case; but her 
poems remain unaffected by the question of her personal virtue, which 
for a century scholars have debated with blind ferocity as if the import 
or beauty of her poetry depended upon their conclusions. One among 
these scholars, the most authoritative champion of her ''good name," 
confides in us that the poetess had "an unquenchable intensity of feeling, 
a yearning of which she herself need not have been conscious.'"' If she 
need not have been conscious of her feeling in order to write, still less 
need we have knowledge of her private life in order to understand her 
writing ; and, if she was fully conscious of what she was feeling, then in 
her writing we have her most exact expression of what she intended 
to say. 

When one reads a poem of Sappho's and reads it as a poem, one is 

i» Ibid., p. 168. 
"Jbid.,p. 164. 
1* U. von Wilamowitz-Moellendorff, Sappho und Simonides (Berlin, 1913), p. 78. 



290 University of California Publications in Classical Philology 

interested primarily neither in Sappho of Lesbos nor in the particular 
audience to which her poems were originally addressed ; one has no de- 
sire to transmute himself into a member of her circle or to gather psy- 
chological data concerning the author, but one does expect to find the 
poem itself directly significant. That this expectation need not be futile 
is due solely to the fact that a work of art exists independently of its 
author and of the accidental circumstances of its production, that its 
artistic qualities are entirely contained within itself and are not to be 
explained by anything outside of the work.'* This alone is the reason 
that it can be understood and appreciated ; it is only for this reason that 
it is worth considering at all, for this independent existence makes pos- 
sible the direct approach of each individual to the work and makes this 
direct approach the only possible way of comprehension and apprecia- 
tion. Those who have recognized the impossibility of attaining the point 
of view of the original audience or that of the author himself have tended 
to overemphasize the individualistic or relativistic nature of apprecia- 
tion and to reduce interpretation to the anarchy of subjectivity. That 
appreciation of any work is ultimately an individual experience is cer- 
tainly true ; it is the chief reason why the study of art and literature 
remains ever new and ever interesting. Yet the same factor that enables 
each person to approach the work directly also limits the extent of the 
subjective element in interpretation. The term "universal," so often 
applied to a work of art, means not that that work is not a unique in- 
dividual, but that it has significance for all men as men in all times and 
places, and this, we saw, is a possibility only if the work has independent 
existence. But the basis of this universal significance is a set of ideas, 
emotions, and values which thUs far in the history of the civilized world 
at least have always been recognized as having validity beyond the arbi- 
trary taste of any individual or the customs of any locality. 

This is not a counsel of indolence. It does not mean that anyone, by 
reason of his humanity alone, can understand any literary work that is 
set before him, or that because it is a work of art with which he is con- 
cerned he need not study assiduously to acquire every instrument that 
may help him to comprehend the significance of the text before him. 
It means that one must never forgot that his ultimate purpose is to 
understand and appreciate particular works of literature, that the one 
•means of accomplishing this purpose is intelligent reading, and that all 
other studies are meant to equip one with this means. It is rather a 
counsel of austerity, for it means that one must learn to distinguish the 
essence from the accidents and to eschew the easy mechanical combina- 
tions that explain nothing. 

The insidious danger of the biographical method lies in its assumption 
that the essence is merely a combination of accidents, that literature is 
an automatic by-product of external forces, whence comes its tacit con- 

^* Cf. J. Korner, op. cit., p. 168. 



Cherniss : The Biographical Fashion in Literary Criticism 291 

elusion that no literary work has autonomous significance. Such an atti- 
tude will have fatal consequences for the study of the classics, for all 
justification of that study depends finally upon the value of the literary 
monuments of Greece and Rome, not their value as source books for the 
historian, the antiquarian, or the psychoanalyst, but their value for men 
as human beings. We pride ourselves mightily on our "true historical 
sense," of which, says one famous modern classicist, "Lessing and Gibbon 
had scarcely a notion, for they thought that man in all ages is essentially 
the same."^ Perhaps that is why the writings of Lessing and Gibbon 
can still be read with understanding today while the books of the new 
scholarship are antiquated after a decade. When the Hellenist no longer 
believes in man as man, he may as well shut his books, for he has confessed 
that he can never understand them. 



^ U. von Wilamowitz-Moellendorff, Sappho und Simonides, p. 3. 



'U^% 



THE IMMIGRANT'S BATH 

BY 

HERMANN FRXNKEL 

(Ovid, Her. 2,90; Pindar, Oly. 12, 19; Alcaeus, fr. 77 Diehl = 29 Lobel) 

In Ovid's Phyllis epistle, we find the following distich {Her. 2, 89) : 

Sed neque consului, nee te mea regia tanget 

fessave [-\& Palmer: -q}iQlibri] Bistonia membra lavabis aqua. 

It is the pentameter which we propose to discuss. The context in which it 
occurs concerns us here only so far as it indicates what the strange line 
purports to express. "You will never bathe your weary limbs in Thracian 
waters" must mean: "You will never take up permanent residence in 
Thrace." Evidently, for Ovid's original public the locution had a familiar 
ring; the notion of a bath in the local waters appears to have been a con- 
ventional symbol for an immigrant's settling down in a new country. 

There is a parallel to support our inference. Pindar, in his twelfth 
Olympian, explains that Ergoteles would have spent his life in obscurity 
had he remained in his native city ; but factional strife caused him to leave 
Cnossus, and subsequently he won his victories in the great games as a 
naturalized citizen of Acragas. The ode ends with these lines : 

NDi' 5' '0\vnTr'Lq. <TTe<pavo3(Ta.nevo^ 

Kal 5is eK Hvdoovos 'ladnol t', 'EpyoreXes, 

depfia NvfKpdv Xovrpa /Sao-rdfeis oniXecov Trap' oiKe'iais apovpais. 

The context makes it probable that the bath in the hot springs near 

Acragas symbolizes not only relaxation from the athletic exertions but 

also Ergoteles' immigration to Acragas, where he had acquired the right 

to own landed property, which was the privilege of a full citizen. 

It is easy to understand that a bath was thought to mark the coming 

to rest after the termination of an incident^ or after the accomplishment 

of a strenuous task,^ and especially that a bath in the native water of a 

place may serve to suggest that a migration is completed and that the 

tu-ed wanderer has come to rest in the country where he intends to stay. 

In Homeric times, a traveler, upon arriving at his destination, was offered 

1 Cf., for instance, Iliad 5, 905-906 (see Gott. gelehrte Anz. 1935, 123).^ 
^ Cf. Pindar, Nem. 4, 1—5: 'Aptoros evippo<7vva irovcov KeKpinkvwv larpos, at 5k aotpai Moiaav 
dvyarpes doiSal dkX^av viv aiTTOnevaC ovhk dtpnov iidwp Tocrov y€ naXeaKO. revxn- Ti'ta TOffirov 
eiXoyia <p6pixi.yyL avvdopos etc.; Artemidorus, Onirocr. I, 64 (p. 57, 29 Hercher) : 
TrdXai . . . ixTi avxex^s eXoiiovro oi ardpo:iroi fir]5e elxov roo-aOra PaXaveta, aXX' rj iroXefiov 
KaTaaTpt\pdij.evoL ij fxeyaXov a.TroTravadjj.evoc wovov kXovovro. In the Iliad (23, 39—47), trie idea 
that Achilles did not rest before he had buried Patroclus is expressed by his refusal to 
take a bath and wash off the battle gore before he has given burial to his friend. 

[293] 



294 University of California Publications in Classical Philology 

a bath by his host in order to wash the dust and weariness^ of his journey 
from his Kmbs.* 

The result we have reached seems in turn to offer a solution for the 
puzzle of a fragment from Alcaeus (fr. 77 Diehl = 29 Lobel) : 

*E(3pe, KaWiffTOs TvoTay-ccv Trap A[i»'OJ' 
f^i\7)aO' OS ?] irop(fvpLav daXacrcrav 
QpaiKlias ep]evyofj.ei'os fa yaias 
. iirir . . . . (."^ 

Kai ae iroXkai irapdeviKai ir(\ 
. . . .]Xo)v firipojv avaXaicTL x«p[<''i 
. . . .] a' 6e\yovTai ro.qv cos aXei[ 
Ofj . . V v5up^ 

With the new volume of the Oxyrhynchus Papyri in our hands, we can 
better appreciate the fact that Alcaeus enjoyed watching pretty women 
and mentioning pretty women in his poetry;^ but that alone doss not 
seem to account for a bathing scene so fully and solemnly described. 
Assuming that the bathers are supposed to be Lesbians, we now recog- 
nize the meaning of the picture. Alcaeus is jubilant to see settlers from 
Mytilene established on the lovely site of Aenus. The colonization' is at- 
tested by Strabo VII, fr. 52 p. 331 : IIpos 8e rfj k/3oXfj rod "E^pov, hcFTbyLov ovtos, 
TToXtse Alvos kv Tc3 MeXa^'t koXtto} Keirat, KTia^a MLTv\r]vaio}v Kai Ku/xaicoj', en 8e 
irporepov 'AXonreKOvvrialuv. 

^ With the fessa membra in Ovid's epistle compare Odyss. 10, 363: (She bathed me) 
6<ppa fioL eK Kafxa.Tov Ov^uxfSopov elXero yvloiv. 

* I can testify from personal experience (in San Lorenzo Creek) that the thoughtful 
custom is still being practiced in present-day California. In the Odyssey (19, 317), 
Penelope offers a bath to the beggar in order to show her hospitaUtj'; this leads to his 
recognition by Eurycleia. Thus a bath makes it certain that Odysseus has finally re- 
turned, and again in 23, 153 ff. a bath marks his final reestablishment after his pro- 
tracted wanderings. 

* Except for the fact that Mount Rhodope does not seem to be mentioned, the 
stanza agrees with Schol. Theocr. VII, 112 (p. 106 Wendel): 'AXxaTos iprjaiv In 'E/3pos 
KaXXiaros worafucv, dia Qp^Kijs (6ta Gp^Kjjs scripsi: AioKXrjs libri) 8e Karaffpecrdai avrov 
airb "PoSoTTTjs Kai e^epevy€(T9ai Kara iroXiv Alyoj'. 

8 For the reading of the last hne see Oxyrh. Pap. XVIII, London, 19-11, p. 38, 
no. 2166 (b), 1233.2. 

' Cf. Oxyrh. Pap. XVIII, p. 33, no. 2165, fr. 1. col. ii, 25-29. 



Z'li 



492 212. . . . Kal jjLoi. \eye' rds jjLeu eTidviiias 0?7S ov 
KoKacTTeov, ei /xeXXet rts olov del elvac, echvTa be auras cbs 
fxeylaTas irXripo^cnv aurats aixbdev ye irodev eToifxa^eiv, Kal 

e TOVTO elvaL rrjv aperrjv ; 
KAA, ^rjul Tavra eyu. 

212. OvK dpa opdus \eyovTaL oi firideuos deofxevoi ei'dal- 
lioves elvai. 

KAA. 01 \WoL yap av ovto) ye Kal ol veKpol ehbai- 
IxovecrraroL elev. 

212. 'AXXd jxev brj Kal cos ye av \eyeLs betuos 6 0los. 
oi) yap TOL davp.a^0L}x' av el Ei'ptTrtSTjs a\r]6rj ev Tolabe 
\iyeL, Xiyicv — 

ris 5' olbev, el to i^rjv p.ev ecrrt KardaveZv, 
TO Kardavelv be ^rjv ; 

493 Kal rinels tw ovtl t'crcos Te6vap.ev' ribrj yap tov eycoye Kal 
fiKovcra TU}v ao(f)U)v cos vvu i7//eTs Tedvaixev Kal to fxev 
(xC^jjia ecTTLP rjfXLP arjixa, ttjs be \pvxv^ tovto ev c^ Itcl- 
dvyXai elal Tvyxavet 6v olov avaweWeadat, Kal jieTairliTTeiv 
avca KttTCO, Kal tovto dpa rts jjivdoXoycov Kon^pos o.vr)p, laois 
2i»ceX6s Tis i) 'IraXiKos, Tapdycov tui dvofxan 5td to indavov 
re Kal TeLarLKOv ojvofiaae ttWov, tovs be dvorjTovs diJLvrjTovs, 

h TOiv b' avorjTCxiv tovto ttjs 4^vxv^ ov al eTLdv/jLiat elcrl, 
TO CLKoKacTOv avTOv Kal ov areyavov, cos Terprjuevos e'lrj 
Trldos, bLCL TTjv aw^rjffTLav aireiKaaas. TovvavTiov brj ofros 
(Tol, CO KaXXt/cXets, evbelKvvTai cos tCjv ev "Atbov — to atbes 
brj \eyoov — ovtol d^Xtcbrarot av elev, ol afxvriToi, Kal (popolev 
els TOV TeTprjjxevov tWov vbccp eTepco tolovtco TeTprj/jLevco 
KoaKLVu). TO be koctklvov dpa \eyeL, cos e4>ri 6 irpos ejue 

C Xe'7cor, TTjv \pvxw etfaf 7171' be ^pvxv^ KoaKivu) airfiKaaev 
TTjv TOiv avoi]Ticv cos TeTprjjievrjv, are ov bvvap.evriv areyeLV 
6t' airLaTiav re Kal \r}dr]v. 

— Gorgias, ed. Burnet, 492 d-493 c 



SOUL AND SIEVE IN PLATO'S GORGIAS 

BY 

IVAN M. LINFORTH 

In the course of the argument which follows the extended statement of 
his philosophy Callicles affirms his belief that if a man is to be happy two 
things are required: he must give free rein to his desires, and he must 
have the means of satisfying them to the full. Before discussing this doc- 
trine dialectically Socrates turns aside for a moment to try the effect of 
an appeal to his opponent's feelings, employing the very art of persuasion 
upon which Gorgias and Polus set so high a value. He draws an analogy 
between the man who is constantly struggling to satisfy unlimited desires 
and the unhappy persons in Hades who are forever bound to the hopeless 
task of filling a leaky jar with water carried in a sieve. The appeal fails of 
its purpose, because Callicles remains quite unmoved, and we must sup- 
pose that it was introduced by Plato for its edifying effect upon the 
reader. Like the myths, it reinforces by its unforgettable images the con- 
viction of a truth which is established independently by rational argu- 
ment. 

Socrates employs for his parable three ingredients: first, the notion 
that life is death and death is life ; second, the belief that certain persons 
in the lower world were compelled to fill a leaky jar with a sieve; and 
third, the theory that the desu'es are located in one part of the soul, which 
is easily swayed by persuasion. The first two of these are certainly not 
original with Plato. The first he takes directly from two lines of the 
Polyidos of Euripides (fragm. 638), which he quotes. Euripides said the 
same thing again in the Phrixos (fragm. 833), and before Euripides similar 
ideas had already been expressed by Heraclitus.^ What lay behind 
Heraclitus we cannot say. The idea may previously have been afloat in 
the world; it may have formed part of some formulation of doctrine about 
the fortunes of the soul, possibly in a poem attributed to Orpheus. It is 
rash to say with Dorfier- that Plato is alluding to Orphic doctrine, or 
with Nestle^ that we have an "orphisch-pythagoreische Umkehrung der 
Bedeutung von Leben und Tod." It is enough for the present purpose to 
recognize th at Plato has taken the idea immediately from Euripides. 

r ragm. 62 dddvarot. dvrjroi, OvriToi ddavaroL, ^wvres top tKelvuiv ddvarov, top 5' tKeivcov 

^lov TedveSiTes, and fragm. 88 ravro r' eve ^uv /cat TedvijKos nal to eypr]yop6s Kal to Kadevbov 

/cat veov Kac yrjpaibv' Ta.be yap fxeraireaovTa eKelva ecrrt KaKdlva irakiv neTa-KecrovTa TavTa. 

2 Josef Dorfler, "Die OrphUv in Platons Gorgias," Wiener Studien, Vol. XXXIII 

(1911), pp. 177-212. 

^ Platons Gorgias, erkl. von C. Cron und J. Deuschle, 5*» Auflage neu bearbeitet von 
W. Nestle, Leipzig, 1909. 

[295] 



296 University of California Publications in Classical Philology 

In stating the idea that life is death and death is life Socrates uses the 
phrase o-w/xa o-ryjua, 'the body is a tomb.' It is not likely that this is the first 
appearance of this striking pun. In the Cratylus (400 c) Socrates mentions 
two possible derivations for the word acona. First, some people say that 
it is derived from arjfia, either because it is the tomb of the soul, or be- 
cause the soul 'makes its signals' to the outer world (o-Tj/xatvet) by means 
of it. Second, the most probable derivation is that given by Orpheus and 
his associates (oi d/x^l 'Opcpea) : it comes from o-wf eiv, 'to keep safe,' because 
the soul is 'kept safe' in the body as in a prison until it has paid certain 
penalties. Plato glances at the pun again in the Phaedrus (250 c aarjiiavTOL 
TovTov 6 vvv aooij-a xept^epovres bvoy.a^otxev) . The passage in the Cratylus 
shows clearly that o-wjua o-^jua was not original with Plato, and at the same 
time that he did not find it in an Orphic poem. It also marks the impor- 
tant distinction between the conception of the body as a tomb and the 
conception of it as a prison. The latter conception was frequently referred 
to (e.g., Phaedo 81 d, 82 e, 92 a, 114 b; Tim. 44 b) and, like the similar 
notions that the soul is naUed to the body (Phaedo 83 d) , glued to the 
body (Phaedo 82 e) , sown in the body (Phaedo 83 e) , was capable of 
development into significant doctrine. The former is barren and irra- 
tional. If the body is a tomb, the soul within it must be dead, — and noth- 
ing can be made of that. There is only one text in which o-wmo crrjua occurs 
which can be earlier than Plato. Clement of Alexandria quotes the follow- 
ing words from the Pythagorean Philolaus (fragm. 14) : napTvpeovTai /cat ol 
TTttXatot deoKoyoL re /cat /xai^rtes, ojs 5td rti'as rtjucoptas d 4^vxo. rw crcbyuari 
avve^evKTaL Kal Kadairep tv <rd/iaTt tovtw Tedairrai. This fragment is regarded 
by some scholars as spurious. According to Thompson,"* it "may be only 
Plato in a Doric dress." Erich Frank* and others argue that Philolaus 
wrote no book and that all the fragments attributed to him come from a 
late forgery. Whatever the truth is about this, we learn nothing signifi- 
cant from the fragment. It confuses the two ideas of a tomb and a prison, 
which Plato keeps distinct, and it attributes both to "old theologues and 
seers." This vague and general phrase is less precise than Plato's "Or- 
pheus and his associates" as a term for the authors of the idea about the 
prison, and it is no more precise than Plato's "some people say" for the 
idea about the tomb. Who said o-ci^ua arjua first we cannot know. Once 
said, it was bound to be repeated from mouth to mouth, owing to the 
assonance. It is used by Philo in a passage where he also recalls part of 
one of the fragments of Heraclitus (fragm. G2) quoted above: eO /cat 6 
'Hpd/cXetros Kata tovto Mcoi'aecos aKoXovdrjaas rc3 667^iart <pr](n' Zd'^iev tov 

* Gorgias of Plato, ed. W. II. Thompson, London, 1871. 

* Plato iind die sogenannten Pythagoreer, Halle, 1923. 



Liji forth: Soul and Sieve in Plato's Gorgias 297 

eKelpcov davaTOv, TtdvqKajitv be top eKelvwv ^iov cos pvv fxev ore ^dnev redprjKvlas 
TT]S ^/vxv^, Kai cos av kv arniari too ao^fiarL ePTeTVujSev^evris , ei de airodapoLfxep rijs 
ypvxns fcocTTjs TOP tSiop ^iop (Leg.Alleg. i, 108, cited by Thompson). This is 
doubtless an echo of the Gorgias and the Cratylus. In the Cratylus, Plato 
introduces the pun because his subject is the derivation of words. Why 
he introduces it in the Gorgias we shall see later, when we trace the move- 
ment of thought in the passage. 

That the fable of the water carriers in Hades was known before Plato 
is sufficiently proved by Pausanias' description® of the painting of Polyg- 
notus at Delphi. In one part of the picture, the scene of which is the lower 
world, two women, one younger and one older, are represented as carry- 
ing water in broken sherds (h KareayoaLP oarpaKOLs) ; over their heads is 
an inscription to the effect that they belong to the number of the unini- 
tiated. A few lines later, Pausanias says that there was also a jar in the 
picture, with an old man, a boy, a young woman, and an old woman. All 
but the old woman are carrying water; she seems to have broken her 
pitcher, says Pausanias, but she is emptying into the jar the water that 
remains in the broken part. Pausanias infers that these people, like the 
two women first mentioned, had neglected the rites of Eleusis. He does 
not say that there was anything to suggest that the jar had a hole in it; 
and there is nothing to suggest that the vessels in which the water was 
carried were leaky like a sieve, unless we suppose that the painter in- 
tended to convey this impression by representing the first two women 
with broken sherds and the other old woman with a broken pitcher. To 
judge, therefore, from Pausanias' description, we cannot say that 
Polygnotus' conception of the fable is identical with Plato's, but the 
most essential feature is common to both : uninitiated persons in Hades 
are carrying water to fill a jar. Still earlier than Polygnotus are two black- 
figured vase paintings which clearly represent other versions of the fable. 
In one'' there is a great pithos, partly sunk in the ground, and at the top 
of it are two small persons with wings pouring some liquid from pitchers 
into the pithos; two other similar figures, also with pitchers, are mounting 
the pithos, one on either side, evidently with the same purpose in mind. 
That the s cene of the picture is in Hades is shown by the presence of 

^ Paus. X 31, 9 and 11. Karl Lehmann-Hartleben , "Cyriacus, Aristotle, and 
Teiresias in Samothrace," Hesperia, Vol. XII, No. 2 (1943), pp. li5-134, finds, in 
the painting of Polygnotus at Delphi, in architectural remains on Samothrace, and 
in certain Italian vase paintings and murals, evidence of a possible common tradition 
in the artistic treatment of eschatological subjects and subjects connected with 
mysteries. 

^ An amphora in the old Pinakothek in Munich (Jahn, Vaserisammlung Miinchen, 
no. 153), reproduced in Baumeister,Vol. Ill, p. 1924; Jane E. Harrison, Prolegomena, 
p. 617, fig. 166; W. K. C. Guthrie, Orpheus and Greek Religion, p. 163, fig. 15; A. B. 
Cook, Zeiis, Vol. Ill, fig. 262. 



298 University of California Publications in Classical Philology 

Sisyphus and his rock at the right. In the other vase painting^ there is also 
a great pithos, half sunk in the ground, toward which are running six 
persons, three men and three women, carrying water jars on their heads; 
a seventh is standing on the rim of the pithos and emptying his jar into it. 
At the bottom of the picture is the figure of an ass, which has fallen for- 
ward on his knees, partly concealing the pithos, and at the left of the ass 
the huddled figin-e of a man, seated, with beard and long hah. ^Miatever 
may be thought of the man and the ass,^ it is clear that this painting rep- 
resents, perhaps in a comic manner, the same conception of the water 
carriers that appears in the former painting. In neither is there anything 
to indicate that the jar has a hole in it or that the pitchers are leaky or 
broken — two features which are essential to Plato's parable. It is not 
likely, however, that he invented them. In the Republic (ii 363 d) he says 
that some people hold the doctrine that in Hades unholy and unjust 
persons are compelled to carry water in a sieve (KoaKlvco vdcop avajKa^ovai 
(pkptiv), showing that this feature of the fable was not original with liim- 
self. That the other feature, the pierced jar, was a part of the fable as it 
was generally known may be inferred from what Ischomachus says to his 
wife in the Oeconomicus of Xenophon (vii, 40) : oi'x opas, ol els t6v reTp-quhov 
t'lOov avrKeXv Xe'Yop.tvoi cos olKripovTai, otl /jLaTrjv iroveiv boKovcn. The history of 
the fable, its origin and development; its relation to the storj' of the 
Danaides (the connection first appears either in the Axiochus or in the 
South Italian eschatological vase paintings) ; the question whether it was 
to be found, in one form or another, in an Orphic poem — these are 
matters for conjecture. '° We can be sure that when Plato wrote the 
Gorgias the fable alreadj^ existed, ready to his hand, as he presents it: in 
Hades some or all of the uninitiated are compelled to carry water in a 
sieve to a jar that has a hole in it. 

Whence Plato got the third ingredient for Socrates' parable is not so 
clear. "The part of the soul in which the desires are located is of a sort to 
yield to persuasion and to shift to and fro." There is nothing peculiar in 
this simple statement to distinguish it from the theor}^ concerning the 
parts of the soul which Plato expounds at length in the Republic (iv 435 
■ff.) and the Timaeus: the soul consists of three parts, or faculties (Plato 

® A lekythos in the nuiseuni at Palermo, reproduced in line drawing in Arch. Zeit. 
XXVIII (1871), Taf. 31; in Harrison, op. cit., p. 61S, fig. 167; and in Cook, Zeuf, 
Vol. Ill, fig. 264. Cook also gives a photographic reproduction of the whole lekylhos 
(PI. XXXVI). 

9 Cf. J. E. Harrison, op. cit., p. 618. 

" See J. E. Harrison, op. ciL, pi>. 614-624; A. Dieterich, Nekyia, 2d ed. (1913), p. 70, 
fn. 1; Rohde, P.^ychc, 9th and 10th eds. (1925), Vol. I, pp. 326 ff.; W. K. C. Guthrie, 
op. cit., pp. 161 if.; I. M. Linforth, .4//^ nf Orpheus, 1941, pp. 75-97. The subject of the 
"holed vessel" is exhaustively discussed by A. B. Cook in Zeus, Vol. Ill (1940), pp. 
338-451 (pp. 397-426 deal with "water-carr3'ing in connection with the mysteries"). 



Linforth: Soul and Sieve in Plato's Gorgias 299 

does not insist on any particular term), the XoytaTLKcv, the Qviiikov, and the 
eTLdvjjLrjTLKov. This theory he recurs to frequently and raises to high 
ethical significance. It is inherent in the figure of the charioteer and 
horses in the myth of the Phaedrus, and in the figure of the chimaera-like 
animal with three parts in the ninth book of the Republic. The restless 
and capricious nature of the eindviJL-qTLKov, which is simply stated in the 
Gorgias, is elsewhere described more fully. The headstrong horse of the 
Phaedrus and the manj^-headed monster which forms part of the chimaera 
are lively representations of it. In the Republic (iv 439 d) the krational 
and appetitive element in the soul {akbyiaTov re Koi eTndvfxrjTLKOp) is defined 
as that with which it hungers, thirsts, and "feels the flutter and titillation 
of other desires" (Shorey's translation of irepl rds aXXas kinQv^las eTrroryrat, 
who says that tTTorjTaL is almost technical, as in Sappho's ode, for the 
flutter of desire.) It is, furthermore, the companion of pleasures and satis- 
factions, and the literal meaning of the word for 'satisfactions' {irXi^pcoaewp) 
recalls the "filling" of the jars in the Gorgias. In the Timaeus (69 c) 
the mortal soul, as contrasted with the immortal, contains within itself 
violent and inevitable emotional disturbances {beiva koI avajKala h eavTU) 
■jradrjjjLaTa exov). A little farther on (71 a), the einBvii-qTLKov is said to be 
"beguiled by visions and fancies night and day" {virb 8e eiSwXcoi/ Kal (pav- 
racTixarcov vvktos re Kal jxed' rijiepav judXtora \pvxo-yuyr](roiTo) , and the word 
xf/vxayuyeiadaL is only a more picturesque word for the more colorless 
ava-ireiOeadaL which is used for a special purpose in the Gorgias. 

The question whether the idea of the tripartite soul and, more gener- 
ally, the notion of parts or faculties of the soul were original with Plato 
is more insistent in the Gorgias than elsewhere because Socrates says that 
he has heard, from some person unnamed, about the part of the soul in 
which the desires are located. Three writers of the first century B.C. 
attribute the doctrine to Pythagoras. According to Alexander Polyhistor 
(the source of Diogenes Laertius, viii, 1, 30), the soul of man is divided 
into thi'ee parts: intelligence, reason, passion {vovv Kal tppevas Kal dvp.bv). 
Galen reports {irepl rdv Kad' 'lTnroKpaTr]v Kal IlXdrcoj^a 8oyp,aTUiv, Bk. v, 478, 
p. 458, Mueller) that according to Posidonius the views of Plato and 
Hippocrates concerning the iradr] and dwafxeLs of the soul were shared by 
Pythagoras; "though no work of Pythagoras himself has been preserved 
to us," Galen adds, "he [Posidonius] bases his statement on the writings 
of some of Pythagoras' followers." Cicero {Tusc. iv, 10) says that 
Pythagoras and Plato divided the soul into two parts, the rational and 
the irrational. Similar statements are made at a later date by others, 
lamblichus (ap. Stob. Eel. i, p. 369, Wachsmuth) says: ol U irepl nXdrcora 
/cat 'Apxvras Kal ol Xolttol HvdayopeioL rrjv \pvxw TpLfxeprj a-KO(paivovTaL biai- 



300 University of California PuMications in Classical Philology 

povvres els \o'Yi<xjibv Kal dvnov /cat eindvulav, and in the Placita, iv, 1 (Doxog- 
raphi, p. 389), we find this: Uvdayopas UXarcov Kara p.tv t6v avwrarw \6yov 
Scfxeprj TTjv ypvxw, to p.ev yap ex^LV \oyiK6v, to 5e oXoyov KaTo. dt to irpocrexes 
Kal OLKptjSes TpLfjLeprj' to yap oKoyov biaipovaLV el's re to dvp.uov Kal to eindvfXT]TL- 
Kov. On the authority of such evidence some modern scholars maintain 
that a Pythagorean origin must be recognized for Plato's doctrine of the 
parts of the soul. Others remain unconvinced because it is possible that 
Posidonius and the other ancient authorities may have based their state- 
ments on Pythagorean books which were written after Plato had for- 
mulated his doctrine and which were themselves indebted to him. It is 
unnecessary to review at length the contentions on both sides. ^^ 

But whether the precise doctrine of the threefold nature of the soul 
was first advanced by the Pythagoreans or not, it cannot be denied that 
their speculations, like those of manj^ others, may have been moving in 
that direction. Popular notions of what may be called the faculties of the 
soul were recognized in Greek speech from the earliest times. Such words 
as aTrjdos, Kapbi-q, Krjp, Ov/jlos, (ppeves in Homer impty an unconscious rudi- 
mentary psychology.^- The problem of the senses, the emotions, the appe- 
tites, and thought itself, occupied the attention of the earl}' philosophers. 
Especially were they concerned to determine the location of these facul- 
ties in the several parts of the body. The medical writers in particular 
must have given special consideration to this phase of the matter. All 
this lay behind Plato's formulation of his own doctrine of parts of the 
soul. But what he took from the science of the day he made his owti, and 
we may be sure that it was he who infused into the doctrine the high 
moral significance which alone made it important to him. In the Gorgias, 
the idea that there is a part of the soul in which the desires are located 
may not have been original with Plato, though we cannot assign an 
authority for it with any certainty; but the language in which he de- 
scribes this part of the soul is novel, and in all probability the application 
which he makes of the idea is original with himself, as we shall see. 

The several ingredients which have been discussed are combined to 
produce an edifying parable. The unhappy fortune of the water carriers 
in Hades is transferred to the life of men on earth by the application of 
the simple formula, death is life. Hades means the realm of the invisible 

" Cf. P. Friitiger, Les Mythes de Platon, Paris, 1930, pp. 76-96 (on the Th(?orie des 
parties de I'anic). Karl Kerc'iiyi (Pi/thagora)^ and Orpheus, .\msterdani, 1940) dis- 
cusses at some lengtli the I'ytliagorean doctrine of the soul, but does not undertake 
the study of its relation to the theories of Plato (p. 21). 

'- Plato plays with one of these words, attributing to Homer an allegorical meaning, 

in Thenl. 194 c: orav ixkv 6 KrjpSs tov ku rfj 4'^xd 0o-6vs TC Kal ttoXus Kai Xelos nai /uerpta's 
wpyaanei>os fi, to. iSvTa 8ia. tG)v alaSriaeciiv, iufftj^iaivd/xeva tis tovto to riji ^f/vxvs atap, 8 i>fTj 
"Ofx-qpos alviTTOfievos T'fjv rod Kijpov ofioiorrjTa, Tore n^v kt\. 



Linforth: Soid and Sieve in Plato's Gorgias 301 

(to dtSes), in which the soul performs its functions. The uninitiated 
(diivtjToi,) are the unintelligent {a.v6r]T0L) .The jar which the uninitiated must 
ever try to fill is the part of the soul in which the desires are located, and 
the sieve in which they try to carry water is the soul itself. 

The fundamental equation for the analogy is that between the jar and 
the eTLdviiriTLKov. It is based on the resemblance between the words tLOos 
and TLdavos. The eTLdvfxyjTLKov, being "of a sort to yield to persuasion" 
(olot- avaTrddeadai), is ireiaTiKov, and, if -jreLdTLKov, indavov. Both adjectives, 
though more often active in sense, are here passive, meaning 'easily per- 
suaded,' as the passive avaTeiOeadai shows. Hirzel,'^ followed bj^ Nestle, 
says that the tWos is the symbol of the k-mdviJL-qTLKov in its normal condition, 
responsive to persuasion {irddtLv), obedient to it {TLQavbv), and faithful 
in observing its commands. This, he says, is the state of the tj.envrjij.hoL; 
in the d/xuTjrot it is leaky. This overlooks the fact that the persuasion to 
which the eTndvjji-qTLKov is responsive is not the persuasion of reason, but 
the persuasion of thronging temptations to pleasure which produce rest- 
lessness and disorder (ijeTairiTTeLv avoo kclto)) . It is as much as to say that 
the k-TTiOviJLrjTLKov Is like a jar because it is faithful and obedient, and that 
when the jar is leaky the eindvtJLrjTLKdv fails to hold the injunctions of 
reason; and further that when the jar is sound the eTndviJ.riTi.K6u can be 
comfortably filled with these same injunctions. The persuasion which 
affects the eTndviJ.r]TLK6p is the temptation to pleasure, and the uncontrolled 
desires are like a leaky jar because of their insatiability. The tLOos is sig- 
nificant only as bearing a name which suggests the easy yielding to 
temptation. Doubtless the intelligent man who practiced self-control 
could fill his jar and satisfy his desires, but there is no point to the equa- 
tion of the iridos with the eTrLBvjxrjTiKov except precisely in the case of the 
man who lacks self-control. 

There is difference of opinion about both terms in the equation 
ol ajxvrjTOL — at avbr}Toi. The aporjTOL, says Erich Frank, are the people 
without reason and without philosophical training. Certainly the adjec- 
tive may bear this meaning, but there is no reason to make it so precise 
here. It is a word in wide use without any special implication, chosen 
here for its resemblance to d/zuTjrot, and refers slyly to just such people as 
Callicles. As for the a.nvrjTOL, Stallbaum^'* has this note: "Hos vero (i.e. 
Tovs avor]Tovs) anvr]Tovs vocat respiciens ad vocis originem. Ducitur enim 
a V. iiveLv 'claudere,' unde dicti sunt jxiidTai., quoniam teste Eustathio ad 
Iliad, co', p. 1492 ed. Basil, rots /iuorats avayKr] iiheiv to aToy^a Kal ixij eKipalveiv 

1^ Rudolf Hirzel, "Pythagoreisches in Platons Gorgias," in Commentationes Philolo- 
gicaein honorem Theodori Momm.seni, Berlin, 1877, p. 12, footnote. 
" Platons Gorgias, rec. G. StaUbaum, Gothae, 1861. 



302 University of California Publications in Classical Philology 

a fxeiJLvrjVTai,. Itaque ol ajivriTOL nunc duplici sensu dicuntur." With regard to 
this it may be said that, even if we accept Eustathius' derivation of 
fjLvcTTai, the word chosen by the author of the equation is avbriroi, not 
aiJivr]TOL, which is given, as belonging to the fable of the water carriers in 
Hades. The double meaning of afjLvr]TOL is accepted also by Nestle and by 
Thompson, who remarks, "Socrates makes aiJ.vr]Tovs synonymous with ov 
(TTeyavovs, 'the contrary of watertight,' deriving the word from juvco 
'claudo,' instead of 'initio.' " But to recognize in djuur/rot a forced and un- 
natural derivative from nvo is entkely gratuitous. It lies in a nest of puns, 
but this pun is unnecessary and adds nothing to the thought. Besides, 
when the word aixvrjTOL first appears no mention has yet been made of 
anything that is "the contrary of watertight," and the hearer would 
inevitably accept the word in its proper sense, the sense which it bears 
in the fable of the water carriers : thej^ would understand it to mean 'the 
uninitiated' and nothing else. 

The equation between the sieve and the soul — to koctklvov apa \kyeL . . . 
TTjv \pvxw ^^vai — is stated as plainly as the equation between the jar and 
the part of the soul in which the desu-es are located. And yet Sauppe^" 
says: "Durch das Schwankende und Wechselnde im Gebrauch solcher 
Bilder findet wohl auch Entschuldigimg, dass die Seele erst mit dem tWos 
und dann mit dem k6(7klvov verglichen wird." But there is nothing to find 
an excuse for; the image is clear and consistent. The only thing to be 
noted is that, whereas if Plato were following the fully developed doctrine 
of the parts of the soul as it appears in the Republic and the Timaeus, he 
would have likened the sieve to the XoyiKov rrjs ^l/vxv^, he here likens it to 
the whole soul. As Olympiodorus says (p. 141, 21, Norvin), kogkivov be 
eaTLv ri XojLKrj ypvxv nentynevr] rjj 01X670?. There are two parts to Callicles' 
doctrine, (1) to encourage the desires and allow them to grow to the ut- 
most, and (2) to indulge them and provide for their satisfaction (492 d 
5). The first is the function of the directing soul {to -nyenoviKov, as the 
Stoics would later call it) ; the second is the function of the tirddvurfTiKov, 
which strives for satisfaction. The uncontrolled eTndvnr]TiK6p is like a leaky 
jar because its desires can never be satisfied; the guiding soul is like a 
.sieve, which is also leaky, as Plato takes the trouble to point out, obvious 
though it is, for a reason which is stated in the words 5t' aincTTLav TCKal \rjdriv. 

The meaning of these words is admirably illustrated by a passage in 

the Timaeus (87 a) . Of the evil humors which cause disease in the body 

it IS said: 7rp6s tovs Tpels tottovs hex^evTa ttjs '/'I'x^s, irpos ov av eVaor' avToov 

TrpoaTriTTTy, TroiKiXXei (xev e'iSr] dvaKo\las Kal dvadi'iiias iravTo5aira, xot/ciXXei 8e 

BpaavTryrosTe Kal deikias, en de \r]dr]s ana /cat dvanadelas. After the disorders 

" Platans Gorgias erkl. von Hermann Sauppe, herausgegeben von Alfred Gercke, 
Berlin, 1897. 



Linforth: Soul and Sieve in Plato's Gorgias 303 

characteristic of the kinOviJLrjTLKov Sind the dvniKovjthe disorders of the \oyubv 
are stated last, and they are dullness and forgetfulness. Similarly in Rep. 
IV 436 A it is said : ixavdavoixev erepu), dufMOVfieda 8e dXXw toju ev riijuv, kindvix- 
ovp.ev 8' av TpiTco tlvl kt\. The apprehension and retention of knowledge 
is the essential function of the 'KoyiKov. The twofold process is neatly ex- 
pressed by Herodotus (iiiSO f.) in what he says of the son of Periander: 
TOVTO TO CTTOS 6 iiep Tpea^VTepos avrwv ev ovbevl \6yui eiroLriaaTO . . . are ov vboo 
\afiibv ovK eixkp.vr}To. I^owledge may be available, but if a man refuses to 
accept it, or, once accepting it, later forgets it, his soul is like a sieve, not 
being retentive. Such a soul deals unwisely with the desires which are lo- 
cated in one part of it, because it is ignorant and unintelligent : it is the soul 
of an avb-qros. ainaTia is a somewhat surprising word to express the idea of 
refusal to accept knowledge. Elsewhere it is Trlcrris that Plato disapproves 
of. In Gorgias 454 d he distinguishes between Trto-Tts and eTiaT-nfir] : maTis 
can be true or false, kmaTripri only true; ireLdco can produce either. In 
Timaeus 29 C he says: 6 tl irpds yhecFLV ovaia, tovto wpos irlaTLV a\r]deLa. 
''irlaTLs (in the widest sense)," says Adam (Republic, Vol. ii, p. 158), "is 
the normal attitude of the airaidevTos towards his do^aara in general as 
well as his aladr]Ta in particular." dxtoria, however, in the present passage 
is not the opposite of xio-rts in this sense. It does not express a wholesome 
skepticism and disbelief, but a stubborn rejection of sound doctrine. Sim- 
ilarly, in a fragment of Heraclitus (fragm. 86, dXXa toov p.iv deicov to. toXKo., 
Kad' ^UpaKktLTOv , aiTLCTTLri 8La<pvyyaveL fxij yLyvoiaKeadaC), aincrTia is blamed 
for failure of knowledge. Jowett's translation of 8l' aincrTLav, "owing to a 
want of faith," is misleading, and Thompson's, "by reason of its fickle 
nature," is wrong. The phrase 'lack of conviction' comes very close to 
what Plato intends. But why does he use this word rather than Svafiadeia, 
which is found elsewhere in a similar connection? Stallbaum says : "Lepide 
dixit OLinaTiav, quo alludit ad irldov, undavov, et avaireWeadai." But if Plato 
intended this and the hearers perceived it, it is a superficial and unbe- 
coming prettiness, distracting to the seriousness of the thought. It may 
well be that he avoided the unpleasant word 8vap.adeLa out of courtesy 
to Callicles, who was certainly neither ignorant nor stupid, and chose the 
word dxto-rta to express his stubborn refusal to allow himself to be con- 
vinced by Socrates' arguments and to accept his noble ethical principles. 
The idea that the resemblance of the soul to a sieve is due, in part, to 
\i]dri reminds one of the end of the myth in the Republic (x 621 a), where 
it is said that the souls come into the plain of Lethe and encamp by the 
river 'kixkX-qs, whose water no vessel can hold {ov to v8cop ayyetov ov8ev 
cTkyeiv). Here no vessel can hold the water because of the quality of the 
water; in the Gorgias the vessel of the soul is unretentive and the opposite 



304 University of California Publications in Classical Philology 

of watertight because it has holes in it like a sieve. The likeness between 
the two images suggested by Xrjdri is manifest, but in the Gorgias X-qOr} is 
to be understood as the ordinary mental experience. There is certainly no 
covert allusion to eschatological ideas. In the Republic, on the other 
hand, we cannot be quite so sure that there is not an echo of the Gorgias. 

Of the equations which are used to produce the allegorical interpreta- 
tion of the fable of the water carriers, four are puns: awna afjua, widos 
TTidavos, afjLvriTOL av6r]T0L, "At5r;s aUrjs. Such wordplay, says Dorfler, is char- 
acteristic of Philolaus and Empedocles, who were both indebted to the 
Orphics, and of the Orphics themselves ; theological poetry is rich in such 
wordplay, and Plato shows other instances of it in passages showing 
Orphic influence: therefore Plato is indebted, through Philolaus and 
Empedocles, ultimately to the Orphics. This is a flimsy argument, and 
it receives no support from the other fiimsj'' arguments which Dorfler 
employs to prove Orphic influence in the Gorgias. It is probably true that 
theological poetry was rich in wordplay, and it is true that some theo- 
logical poetry was attributed to Orpheus. There are, indeed, many in- 
stances of wordplay in the Orphic fragments. But the Greeks were always 
quick to note verbal similarities. There are many examples of the prac- 
tice in Plato, and it is idle to try to find an Orphic source for them all. 
There is a wealth of them in the Cratyhis, which, as Shorey says, "paro- 
dies the etymological speculations of his day."'*^ Furthermore, we do not 
know that Philolaus and Empedocles were Plato's sources for the present 
passage, nor do we know with anything like certainty that their sources 
in turn were Orphic poems. We inust, in fact, recognize that the four puns 
do not furnish any evidence for deciding the question where Plato learned 
the allegorical interpretation of the fable. 

Having examined the materials emploj^ed to produce the edifying par- 
able by which Socrates undertakes to persuade Callicles to a more whole- 
some philosophy, we come now to Socrates' own statement concerning 
the source from which he obtained it. He disclaims originality and asserts 
that he had heard it from someone — riS-q yap tov e7w7e Kal fiKovaa tuv 
coipcov. The common guess is that the "someone" referred to is Philolaus,'^ 
the Pythagorean with whom Simmias and Cebes were associated at 
Thebes. There is no more evidence for this than has already been brought 
forward in the examination of the ingredients of the parable, and the 
fact that Socrates could perhaps have talked with Philolaus or at any 
rate with one of his associates. We do not find much assurance in this. 
Nothing can be made of the word aoipihv. It is idle to say that it means 

»« What Plato Said, p. 259. 

" Cf. Erich Frank's skeptical views about Phikilaus, op. cit., pp. 291-302. 



Linforth: Soul and Sieve in Plato's Gorgias 305 

'philosophers' in the professional sense of the term, as Frank maintains 
(p. 298, n. 1), or Orphics, as Dorfler suggests. Any man could be called 
(T0(p6s who interested himself in such things as the shrewd and significant 
interpretation of the fable of the water carriers. This unknown person, 
whoever he was, told Socrates that the fable of the water carriers had 
been composed originally to convey the important truth of the allegory 
and that its author was "some clever person, a Sicilian perhaps or an 
Italian" — tovto apa tls nvOoXoyoov ko/xi/'os avqp, 'iaoos StKeXos tls fj 'ItoXlkos. 
Here there is a little more precision, but still we can only guess. In fact, 
Socrates only guessed, as tacos shows. The name most often thought of is 
that of Empedocles, which was first proposed by Olympiodorus (p. 143 
Norvin, Zt/ceXt/cos" olov 'E^tTreSo/cXrys, UvdayopLOS yap r\v ovtos. virrjpx^v 5e 
'A/cpaTaj/rTj^os" 'AKpayas 8e rrjs St/ceXias). The scholiast repeats Olympio- 
dorus' words and adds : (SikcXos t) 'ItoXikos) eireibr) ttXtj^Iov DueXtas o re 
'KpoTwv Kal TO M-eraTTovTiov, at xoXets ov ol UvdayopLOL 8l€tpl^ov, at ttjs 
'iToXias dalv. Nestle, who thinks that the man who told Socrates the 
story was probably Philolaus, has the curious notion that 'IraXiKos, too, 
refers to Philolaus, who came from Croton or Tarentum. The 2tKeX6s, 
he thinks, may be Empedocles, but he offers another possibility which he 
expresses thus: "... wenn Sizilien nicht bloss wegen seiner Zusammen- 
gehorigkeit mit Italien, wegen des Rufes seiner witzigen Bewohner 
(Timokr. fr. 6 : SiKeXoj KopL^pos dvrjp ttotI rav parep' e(pa) und im Blick auf die 
Sizilianer Gorgias und Polos neben jenes gestellt ist." This latter sug- 
gestion deserves consideration, and we shall return to it later. 

An ingenious explanation has been offered by Erich Frank. He begins 
by assuming that the Kop\f/6s avijp who composed the myth was the author 
of the Orphic Catabasis. As evidence for this he refers to the Republic 
(ii 363 d), the Cratylus (400 bc), and Plutarch (Cim.-Luc. i 2). But the 
fable of the water carriers is not attributed to Orpheus in the Republic, 
and it is not even mentioned in the Cratylus or by Plutarch. We have 
seen that it was common property long before Plato, and though it 
would not be surprising if it was told in an Orphic poem, there is really 
nothing to prove it. For the present passage it makes no difference 
whether the fable originated in an Orphic poem or not. The story as it 
was told by the original author, Frank continues, was concerned only 
with the leaky jars in Hades; the secondary meaning, involving the satis- 
faction of the desires in the soul, though it is represented as consciously 
intended by the original author, could not have been intended by him, 
but must have been the result of subsequent allegorical interpretation. 
When, therefore, the original author is called a Sicilian or an Italian, the 
purpose is to indicate, wittily and ironically, that the secondary author. 



306 University of California Publications in Classical Philology 

the author of the allegorical interpretation, was a Sicilian or an Italian, 
or, in other words, that this interpretation originated in the "school of Si- 
cilians and Italians," which, as we see from Aristotle, was the customary- 
designation in Plato's time for the circle of "so-called Pythagoreans." 
But the author of the allegorical interpretation is the philosopher {to:v 
GO(puv TLs) from whom Socrates heard it. Thus Socrates indicates unmis- 
takably that this philosopher was a Pythagorean. Now Socrates, as 
Plato tells us, never conversed with Philolaus, and since he hardly ever 
stirred from Athens he could not have met a Pythagorean in Italy or 
Sicily. Plato, however, when he wi"ote the Gorgias, had just returned from 
Italy. We must conclude, therefore, that he is attributing to Socrates, by 
an anachronism, an experience of his own. Plato himself must have heard 
Archytas, or some other of the school in Italy, interpret the Orphic Cata- 
basis. So far Frank. 

Now this chain of reasoning has several weak links. We are not bound 
to suppose that the person whom Socrates heard was himself the author 
of the allegorical interpretation; it is absurd to believe that Socrates 
could not have conversed wdth a Pythagorean even though he did not go 
to Italy ; and the date of the Gorgias is not so definitely fixed that we can 
assume it was written immediately after Plato's return from Italy. But 
the weakest point — and the most essential point — in the argument is the 
assertion that when the author of the fable is said to be a Sicilian or an 
Italian, Plato is indicating, "wittily and ironically," that the author of 
the allegorical interpretation, the man from whom Socrates heard it, was 
a Sicilian or an Italian. The wit and irony are too well concealed. The 
man who told Socrates the story says expressly that the author of the 
fable, with its implicit allegorical meaning, was perhaps a Sicilian or an 
Italian : how can we be expected to understand from this that he means 
"perhaps one of ourselves"? He might be understood to mean this if we 
knew that he was himself a Sicilian or an Italian. But this is just what we 
do not know. The purpose of Frank's whole argument is to determine 
who he was, and, depending as it does upon the assumption of an over- 
subtle piece of wit and irony, it must be judged unsuccessful. 
• Whoever the two persons mentioned by Socrates may have been, most 
scholars are convinced that what Socrates "heard" was all Pythagorean 
lore.^^ Olympiodorus, as we have seen, offered the name of Empcdocles 
as a guess, "because he was a P3^thagorean." The scholiast (on 493 a) 

18 Diimmler's conjecture (Ak-arlemika, Giessen, 1889, pj). 87 fT.) that Plato borrowed 
the allegorical interpretation from Antisthenes was accepted with heartj' approval by 
Joel {Der cchte unci der xenopho7itische Sokrates, B. II, 1, Berlin, 1901, p. 219), but it is 
not sui^ported by convincing arguments. The curious notions advanced by Schuster 
(Rh. Mm. XXIX, 1874, 590-632) have been effectually disproved by Hirzel and need 
not be recalled. 



Linforth: Soul and Sieve in Plato's Gorgias 307 

quoted above finds an explanation for the guess that the author of the 
fable was an Italian in the fact that Croton and Metapontum were the 
home of the Pythagoreans. Again, on 493 d, the scholiast compares the 
two fables which Socrates tells, the one now under consideration and the 
one immediately following (493 d ff.), and asserts that the former is 
Pythagorean, the latter original with Socrates (rjv 8e eKeivo fxev rcbv Uvda- 
yopioiv oUetov, tovto 8e Zi>)KpaTOvs, ws aaupkarepov re /cat ir\r]KTLKOOTepov). 
From these statements we must conclude either that their authors were 
guessing or that they actually found the fable of the water carriers in 
Pythagorean books. The certainty with which they speak makes the 
latter the more probable assumption. But the greater part of Pythagorean 
literature was composed after Plato, by men who learned much from 
him. Unless we can find some reasonably sound evidence that Plato him- 
self was drawing from Pythagorean sources, the words of Olympiodorus 
and the scholiast cannot have much weight. Such evidence we have not 
been able to find. Socrates' introductory remarks about o-aj/xa arjiia and 
the parts of the soul, as we have seen, have only the most tenuous con- 
nection with Pythagorean sources, and for the water carriers there is not 
only no Pythagorean authority but no earlier literary authority what- 
ever. To say that the fable 'sounds' Pythagorean is to offer a very slen- 
der argument. There is, in fact, very little gi'ound for confidence in the 
traditional theory, and one may be justified in discarding it, at least for 
the sake of argument, and in facing the problem anew without prejudice. 
In the attempt to discover the sources from which Plato has drawn in 
writing the passage in the Gorgias which is under consideration, and to 
determine who it was from whom Socrates heard the allegorical interpre- 
tation of the fable of the water carriers, and who the clever man was who 
composed the fable as a parable of the soul, it is obviously necessary first 
to trace the history, as far as possible, of the several elements of which 
the whole is composed. This we haA-e already done, and we have found 
occasion to modify some of the more positive conclusions that have been 
reached by others. But this is not enough. There is another aspect of the 
matter which has too often gone unregarded. Not only the ingredients of 
the parable require investigation; the manner in which Plato employs 
them also challenges the curiosity of the reader. The study cannot be 
complete without a careful examination of the literary expression and 
the dramatic structure which give the piece its pecuHar quality. Schleier- 
macher^^ expressed the opinion that it is "das Kunstreichste" in the 
whole of the Gorgias. Let us see what we can learn by turning our eyes in 
this directio n. 

" Quoted by Paul Friedlander, Platon, Vol. II, p. 266, n. 2. 



308 University of California Publications in Classical Philology 

By way of introduction to the allegory, the man from whom Socrates 
professes to have heard it says: "We are dead here and now; the body is 
our tomb, and the part of the soul in which the desires are located is of a 
sort to yield to persuasion and to shift back and forth" {vvv 17/xeTs redvafiev 
Kai TO fxev (xcbfia eatLV fifxlv arjua, ttjs 5e '/'I'X^s tovto ev cS at eindvuiai elal rvyx^'-^^i- 
ov olov avaweWeadaL Kai ixeTaTLTTTeLV avco /cdTCo).We have seen that crw/ia a-yjiia is 
an old pun, which cannot be fathered upon anyone in particular. The 
notion of the parts of the soul, too, though it is most fully developed by 
Plato, is probably rooted in earlier thought on the subject. But what is 
said in the present passage might have been said equalty well even if the 
notion had never been heard of before and even if Plato had not elabo- 
rated it elsewhere. At the beginning of this whole movement in the 
Gorgias, at 491 d, Socrates draws attention to the importance of a 
man's governing himself — eva eKacxTov Xeyw avTov iavrov apxovTa. This im- 
plies two parts in the personality. In the present passage, for the sake of 
the parable, the lower part, the part which should be governed by the 
higher, is called explicitly and very simply ttjs ypvxrj^ tovto ev w ai kindvfiiaL 
elai; as the parable proceeds, the higher part of the personality, which 
should be dominant, is called just 4^vxv- 

We cannot say that this vivid manner of presenting the dual per- 
sonality was freshly devised for its present use, but it is entirety possible 
that it was. What Socrates' informant is supposed to say, however, about 
the kwidvuriTLKov (there is some impropriety in using this term, which be- 
longs to the developed theory, but it is convenient) could never have 
been said by anyone except under the conditions of the present dialogue. 
The word avaireldeadat sounds again the note of 'persuasion' which is 
dominant throughout the earlier part of the conversation. The phrase 
/xeraxtxTctr avui kcltoj is an echo of words used earlier by Socrates to de- 
scribe the conduct of Callicles under the influence of love (481 d) : "You 
have two loves, both named Demus, the Athenian people and the son of 
Pyrilampes, and I have noticed that every time, whatever j^our favorite 
says and however he looks at things, you, with all your cleverness, have 
not the power to contradict him; you just shift back and forth at his 
whim" (apo) Kai kcltco /xera/SaXXo^uej'oi;) . fxeTa^aXXeadai is in effect the same 
as neTairliTTeLv, which is commonly used as the passive of )uera/3dXXetv. The 
effect of temptation upon the soul is exactly the same as the effect of love 
upon Callicles, and the same words recur. Similar language is used in the 
Protagoras (35G d), where it is said that in choosing one's pleasures and 
pains it is better to employ the art of measurement than to judge by 
appearances; the latter, saj^s Socrates, -q^ids eirXava Kai eiroieL ai^co re Kai 
KOLTCO TToWaKLS iJLeTaKafjL^aveLV Tai'TO. Kai neraneKeiv Kai ev Tals Trpd^ecri Kai h 



Linforth: Soul and Sieve in Plato's Gorgias 309 

rats aipeaecTLv tuv jxeyoKwv re /cat anLKpojp. If Socrates really heard from an- 
other something about the nature of the eindunrjTLKov, he certainly recast 
it in language which is significant for his present purpose. 

Next consider the combination of the two statements in the introduc- 
tory remark, that about the body and that about the soul. With all the 
source hunting, no earlier instance of such a combination has been found. 
They belong to two departments of philosophy : accfia arjua, if to anything, 
belongs to eschatology ; the theory of the k-wLdvyi-qTLKov belongs to psychol- 
ogy. They could not, in fact, have been brought together by anyone, 
except by Socrates here and now for his present purpose. In order to 
understand why this is so, it is necessary to trace the thought that leads 
to the combination. 

At 492 D 5 Socrates is ready for a dialetical discussion of the unpor- 
tance of self-control and begins with the familiar phrase nal fioL 'Keye. 
"You say, do you," he asks, "that no check should be put on the desires 
if a man is to be what he ought to be, but that he should let them be as 
great as possible and provide satisfaction for them from some source or 
other? This is what you call virtue?" Callicles agrees. "In that case it is 
wrong to say that those who want nothing are happy." To which 
Callicles retorts: "If they were, stones and corpses (oi vtKpoi) would be 
happiest of all." The word veKpol catches Socrates' attention and suggests 
a thought to him: "But life would be terrible as you put the matter, 
too" — dXXd ixh 8ti /cat ois ye <tv XtYets Setvos 6 jStos. The words aWa pev 8r] in- 
dicate that the statement is something more than a mere retort and prom- 
ise further amplification of the meaning. Immediately he continues: 
"Let me show you how this may be so (yap)." He then advances the idea, 
on the authority of Euripides, that life may be death and death life. If 
one should stop to think what this means instead of reading straight on, 
he would suspect that Socrates' argument was something like this: "Let 
us accept for the moment what is implied by your words and agree that 
the dead cannot be happy. But there is some authority for believing that 
those who we say are alive are really dead. Now those who indulge in 
pleasure without restriction are certainly to be included among those 
who we say are alive: therefore those who indulge in pleasure without 
restriction must be dead themselves, and, if dead, not happy." This 
syllogism is faultless, but its paradoxical major premise makes it useless 
for rational argument. Socrates does not state it or even hint at it. It 
must have been at the back of his mind, but he was wise enough to sup- 
press it. We see immediately in the next half-dozen Imes that he is going 
to use the paradoxical major premise not as the basis of a rational argu- 
ment, but as the text for a little sermon which is not calculated to prove 



310 University of California Piihlications in Classical Philology 

anything, but to make a solemn appeal to the feelmgs. He is going to pre- 
tend that the fable of the water carriers, being concerned with the dead, 
is therefore concerned with those who we are accustomed to saj^ are alive. 
So, having quoted Euripides, he proceeds directly to report what he says 
he has heard from someone else, and we come to the introductory state- 
ment about aSifxa ayjfxa and the kTndviJ.rjTLK6v. 

Now it is possible to say that death is life and life is death in the sense 
that all the misery which we associate with death really belongs to what 
we call life, and that the joy and happiness which we mistakenly think 
belong to life may really be found in death. But if one attempts to de- 
velop this idea circumstantially it proves intractable. Anything can be 
imagined about the state of death, even though we call it life; but we 
cannot think in any new way about life simply by changing its name to 
death. When in the Phaedo Plato says that the true philosopher practices 
dying, he is thinking of the intensification of the life of the soul and the 
subjection of the body, an approximation to the state of death when the 
soul will have been liberated from the body. Xow the striking phrase 
aoifia arjiJia is only a way of saying that the soul will enjoy its life to the 
full only when it is out of the body. The converse implication, that if the 
soul is lodged in a tomb it must itself be dead, is unthinkable. The phrase 
is perfectly irresponsible and exists only for the pun. Very different is the 
idea that the soul is a prisoner in the body (5e5eo-0at), which is attributed 
to Tovs aynpl 'Optpea in the Cratylus. This is capable of expansion into 
edifying doctrine and may be easily accepted as a description of the true 
state of things. No concise and telling phrase, like o-w/Lta afjua, was devised 
for this doctrine, o-w/xa ay]fia, on the other hand, unforgettable as it is, is 
barren of meaning except for a vague pessimism and otherworldliness. 

Why, then, does Plato introduce the phrase into the present passage? 
He uses it as a means of modulating from the thought in Euripides' lines 
to the myth, which under the guise of an account of things in the lower 
world describes the psychology of the pleasure seeker. This life of ours, 
which is really death, must be represented at the same time as an 
existence in which the motions of the soul can be studied. So, having said 
that we are dead here and now, Socrates ostensiblj' goes on to show what 
this means both for body and for soul, pointing the balance neatly by 
nev and 8e(T6 ixev aojfia . . . ttjs 5e ^ux^s). Of the body he says o-wjua crTj/xa, 
which, if it means anything, means just the Euripidean inversion and 
has no meaning for life as we know it. What he says of the soul is true of 
life as we know" it, and none the less true if we choose to call life as we 
know it death. So he has devised a way, based on a paradox, of employing 
a fable the scene of which is in Hades to describe the psychological state 
of the devotee of pleasure. 



Linforth: Soul and Sieve in Plato's Gorgias 311 

All these considerations lead to the conclusion that the whole of what 
Socrates professes to have heard from someone else, the allegorical inter- 
pretation of the fable and the remarks introductory to it, was original 
with Plato himself. The question then presents itself, why, instead of 
having Socrates present it as his own, he makes him report it indirectly 
as if he had heard it all from another. It may be said, in a general way, 
that the attribution of ideas to fictitious authorities is almost a habit of 
the Platonic Socrates.^" But in the present place Plato has a special 
reason for what he does. In the Phaedrus (229), when Socrates and his 
companion have come to their resting place on the bank of the Ilissus, 
Phaedrus remarks that the scene of the old legend about Boreas and 
Orithyia must be somewhere near, and he asks Socrates if he believes the 
tale. "I should not be in any way peculiar if, like the clever people in the 
world, I refused to believe it. If I did, I might try my hand at cleverness 
(aocpL^onevos) and say that the maiden had been hurled down from the 
cliff there by a blast of Boreas and that this was why it eventually came 
to be said that Boreas had abducted her. Now that kind of thing is all 
very well, but it takes great ingenuity and industry. If you once begin, 
you are bound to go on and explain the Hippocentaurs and the Chimaera 
and the Gorgons and all kinds of strange creatures. If you are going to 
refuse to believe in them and resort to a rough-and-ready kind of clever- 
ness (are dypoiKw tlvI (To<pia xp^j^f^^^os) to account for them all, it will take 
a great deal of time. I simply have no time for it. I have not yet succeeded 
in obeying the Delphic injunction which bids me know myself, and it 
would be a funny thing, when I am still in a state of ignorance on that 
subject, to occupy myself with inquiries which do not concern me." Thus 
Socrates makes it clear that he has a low opinion of the prevailing interest 
in allegorical interpretation : it is a petty business, and there are more im- 
portant things to be done. Plato goes out of his way a little to present to 
us this side of Socrates' character and opinions. How could he, then, in 
the Gorgias, when he wishes to offer, an allegorical interpretation of an 
old fable, even though he does it for a truly edifying purpose, put it 
directly into the mouth of Socrates? Dramatic consistency forbids it. So 
he thrusts the responsibility for the interpretation upon a nameless and 

20 See Shorey (op. cit.) on Hipp. Maj. 286 e (p. 473) and on Symp. 201 e (p. 546). 
Diotima, whom we are almost bound to consider a fictitious personage, is the most 
striking instance of the practice. Cf. Friedlander, Platon, Vol. I, p. 173 (speaking of 
Diotima): "gleichsam die hochste Gestaltwerdung jenes mehr oder minder unbe- 
stimmten 'Jemand,' den er so oft spielend in Gesprach und Redekampf als einen 
andern aus sich heraussetzt, um sich ironisch hinter ihm zu verbergen." Thompson 
(ad loc.) suggests that tcov ao(paiv ns may be an imaginary person, because Socrates 
easily evoked imaginary vouchers for his views; but he rejects the suggestion and 
does not develop it. 



312 University of California Publications in Classical Philology 

fictitious person, only designating him as one of the cro(poi, the very word 
which he uses in the Phaedrus for the allegorizers. In the midst of the 
report of what this person is supposed to have said, Plato allows Socrates 
to interrupt it and speak in his own person, pointing out the misery of the 
d/iur/rot who do not control their desires. This is Socratic ethical doctrine. 
But when he continues the tale and declares that the sieve corresponds 
to the soul, he is careful to attribute this statement to the fictitious au- 
thority (ws e(pr] 6 Trpos efxe \kyuv). And when the report is concluded, 
Socrates is made to say, quite in character, that it is all rather absurd 
but still useful to persuade Callicles to a better philosophy. Thus the 
obloquy is made to fall on the unnamed person from whom he pro- 
fesses to have heard it. The deception — if there was any deception — is 
harmless enough, and is justified by the opinion which Socrates expresses 
in the Charmides (161 b), that it is not important to know the author of a 
doctrine, the important thing is to determine whether it is true. In the 
present place Socrates is very sure of the truth of the doctrine and very 
much in earnest about it. 

Finally, who was the Ko/dxpos avrjp to whom is attributed the invention 
of the fable of the water carriers, with its enigmatic significance? It was 
Plato who discovered the enigmatic significance; but since it was cus- 
tomary, when one found an allegorical meaning in a myth or a legend, to 
assert that this allegorical meaning was intended by the original author 
of the myth, the author of the present fable is called koh\{/6s for his skill 
in devising the enigma. As a matter of fact, he was just the author of the 
fable, as anonymous as "the lawgiver" of whom we hear so much when a 
law is under discussion. Why, then, is the guess hazarded that he may 
have been a Sicilian or an Italian? In the first place, Socrates — and 
Plato — thought it was true. He did not know for certain where the fable 
of the leaky jars in Hades had originated, but he must have known, as ive 
know, the importance of chthonian cults and of eschatological specula- 
tions in Italy and Sicily.^' Indeed, as we have already seen, the only evi- 
dence earlier than Plato for the water carriers in the mysteries is found in 
two Italian vase paintings and in the painting of Polygnotus, for which 
•Italian connections have been suggested by Lchmann-Hartleben.-'- There 
is also further evidence, only a little later than Plato, for the water 
carriers in Ital.y : they are represented in the well-known Apulian eschato- 
logical vase paintings, and they appear in a limestone relief from Apulia, 
dated in the second half of the fourth centurj' b.c, now in the Glyptothek 
in Munich.'-"' Plato himself may have seen such things in Sicily and Italy, 

*' Ker6nyi (op. cit.) offers some interesting suggestions concerning the origin and 
continuance of eschatologv in Italv (see esp. p. 10). 

2' See above, p. 297, fn. 6. "ghown in Cook, Zeus, Vol. Ill, p. 424. 



Lin forth : SouJ and Sieve hi Plato's Gorgias 313 

and there is no serious anachronism in his putting the guess into Socrates' 
mouth, even if he had supposed that Socrates would not have been aware 
of the interest in eschatology prevalent in that part of the Greek world. 
He was justified, therefore, in the guess. But the positive motive for 
introducing it was different. As Plato was writing, three things flashed 
into his mind at once. First, the words K0fjL\}/6s avrjp, which he had just 
used, recalled the line of Timocreon, SiKeXos KOji^pos avrjp ttotI rav ixarep' 
e(j>ar^ Second, remembering that Gorgias and Polus were Sicilians, he saw 
that the attribution of the edifying fable, with its clever allegorical mean- 
ing, to a Sicilian would be a slight but graceful compliment to them. 
Third, he did not object to hinting that the persuasive emotional appeal 
of the fable belonged to the art of Gorgias, in which persuasion was the 
chief end. So, struck by the aptness of the thought, he ventured to sug- 
gest that the author of the fable was a Sicilian. But, in the interest of 
truth, because after all Italy and Sicily form a single community, and 
because Italians are as much interested in eschatology as the Sicilians 
are (perhaps even more), he adds as an afterthought ^ 'IraXuos. 

The question which Socrates put to Callicles at the end of his parable 
may now be addressed to the reader: dXXd irorepov ireiOu rl at . . . rj ovbkv, 
dXX' av Kal aXXa xoXXd rotaOra nvdoKoyo) ovbkv tl paWov neradrjaeL ; Whatever 
the answer maj^ be, it will not be denied that the various solutions of the 
problems presented by the passage which have been obtained by the tra- 
ditional method of approach have not given general satisfaction. No one 
of the names proposed for the man with whom Socrates conversed and 
the man who was the author of the fable is supported by more than a 
little circumstantial and farfetched evidence. The new interpretation 
that has be en offered may be thought to be at least worth a hearing. 

2* See p. 305 above. 



AUGUSTINE ON THE TEACHING OF HISTORY 

BY 
WILLIAM M. GEEEN 

Despite the vast amouut written on Augustine's philosophy of history, 
little notice has been given to a well-defined plan for instruction in 
history devised by him and widely adopted in the schools of later days. 
Even the historians of education, while discussing the curriculum of 
ancient and medieval schools, often fail to recognize that any time was 
given to the study of historJ^ In fact, the statement has been made that 
the subject was not admitted into the schools until the modern era/ Cer- 
tainly modern instruction in history, like the modern conception of his- 
tory, is something very different from that of Augustine and medieval 
times, just as this was something very different from that of pagan Rome. 
But in schools of every period some knowledge of past events has been 
taught. This paper is an attempt to discover what sort of "history" was 
taught in Augustine's time, to describe the plan which he developed, ' 
and to notice a few instances of its later use. 

I. The Teaching of History in the Time of Augitstine 

Augustine's early life, from childhood until his conversion at the age of 
thirty-three, belonged to the schoolroom. According to a curriculum 
essentially unchanged since the time of Quintilian, he learned his letters 
under an elementary teacher (ludi magister, litterator) , and then pro- 
ceeded to the grammaticus for the study of literature, chiefly poetry. 
At the age of sixteen he was sent to the school of the rhetor in Carthage, 
where he read Cicero and learned the art of oratory. He there stood at 
the head of his class, and found the work of the school so congenial that 
he chose for himself the profession of teaching. He was first a grammati- 
cus in his native Thagaste, and later a rhetor in Carthage, Rome, and 
Milan. The crowning recognition of his career came in 384, when he was 
asked to deliver a panegyric in praise of the emperor and the consul- 

^See Monroe, Cyclopedia of Education (New York, 1914), III, 284; S. d'lrsay, 
Histoire des universites (Paris, 1933), I, 251 ; H. Rosenberg, Methodilc des Geschiclits- 
unterriclifs (7th ed., Breslau, 1913), p. 134; H. Johnson, An Introduction to the His- 
tory/ of the Social Sciences in the Schools (New York, 1932), pp. 8-16. For an account 
of Augustine's activities as a teacher see F. X. Eggersdorfer, Der heilige Augustinus 
als Pddagoge (Strasshurger theol. Studien, 8'^ Bd., 3. u. 4. Heft, Freiburg, 1907). 
For his s.^Tithesis of ancient and Christian culture see H.-I. Marrou, Saint-Angustin 
et la fin de la culture antique {Bihl. des Ecoles frangaises d'Athenes et de Borne, fasc. 
145, Paris, 1938) ; a good discussion of the treatment of history is found on pp. 417- 
419, 461-467. The best account which I have found of any medieval history teacher is 
that of W. Levison, "Bede as a Historian," in A. H. Thompson, Bede : His Life, Times 
and Writings (Oxford, 1935), pp. 111-151. Hipler, Die christliche GescMchtsauff as- 
sung (Gdrres-Gesellschaft II. Vereinschrift fur 1884, Koln, 1884) is an uncritical 
survey of the subject. 

[ 315 ] 



316 University of California Publications in Classical Philology 

designate for the next year." From tlie second century before Christ, 
when Rome adopted the culture of Greece, until the end of the Empire, 
the core of the school curriculum was literature and rhetoric. Other sub- 
jects were only casually introduced, to explain passages in the authors 
read and to provide information supposed to be useful for the orator. 
There was a rather vague group of "liberal arts" or "disciplines" which 
every educated person was expected to know; these were supposed to be 
taught along with grammar. Quintilian wished his student before taking 
up rhetoric to know something of music, geometry, and the art of the 
actor.^ Varro wrote a treatise on nine disciplines : grammar, dialectic, 
rhetoric, geometry, arithmetic, astronomy, music, medicine, and archi- 
tecture.* Other writers have similar lists, some longer and some shorter. 
Every student of the grammaticus picked up scraps of these disciplines, 
and one who later felt the need might, like Augustine,"" make a private 
study of the books which treated these subjects in a systematic way. 

"History," that is, the knowledge of the past, was in much the same 
case as the liberal arts.^ The grammaticus was expected to supply the 
historical information required for understanding the authors read, but 
to avoid superfluous and recondite matters.' The writings of the histo- 
rians were taught rather by the rhetor, with the double purpose of fur- 
nishing models of prose and a stock of illustrations.^ Quintilian would 
have a student read Livy or some other historian aloud, and then would 
discuss with the class the merits and defects of the style.^ Familiar ex- 
amples of the virtues were drawn from the stories of early Rome.'" Cicero, 
indeed, had insisted on more — that the orator should be accurately in- 
formed in both Roman and foreign history, chronology, geography, and 
the causes which determine the course of events. He complains that 
rhetors neglected these matters, and praises Atticus for compiling a 
needed chronological work for ready reference." At nearly the same 
time Nepos wrote a universal history in three books, and a few years 
later Sallust was provided with a hreviarium rcnim omnium Eoma- 
narum by the distinguished teacher L. Ateius, suniamed "Philologus."" 
Such epitomes of history may well have been used in instructing boys 

- Conf. 6, 6, 9; e. Petil. 3, 25, 30; Marrou, p. 88. Except as otherwise noted, Latin 
authors are cited according to the abbreviations of the Thesaurus Linguae Latinae. 

^ Inst. 1,10-12. 

■' See Dahlmann in Pauly-Wissowa, Suppl. VI, p. 1256. 

" Conf. 4, 16, 30. 

"Ill fact, Vitruvius 1, 1, 3 includes hisloria as one of the seven branches of the 
encyclins discipJina needed by the architect. 

'Quint. Inst. 1, 2, 14; 1, 8, 18-21, concluding with the words "ex quo mihi inter 
virtutes grammatici habebitur aliqua nescire." 

** Inst. 2, 4, 2 ; 2, 5, 1. 

» Inst. 2, 5, 6-20. 

'« Inst. 10, 1, 3, f. ; 12, 2, 29 f. ; 12, 4, 1-2. 

" Ornt. 120: DeOrat. 2, 62-64. 

^=Catull. 1, 3-7; Suet. Gramm. 10. 



Green : Augustine on the Teaching of History 317 

then, as we know they were later. Statins, when congratulating his friend 
Vibius on the birth of a son, remarks that the boy should learn the out- 
line of history which his father had compiled from Sallust and Livy.^ 
In the rhetorical schools, however, attention was devoted not so much 
to the comprehensive view of history as to the stories which afforded 
illustrations for commonplace topics. "Plena exemplorum est historia," 
Cicero declared," and the effective variety of his illustrations was pos- 
sible only for one well informed in both Greek and Roman history. The 
emperor Augustus read widely to gather excerpts of practical value for 
his friends and subordinates.^" Valerius Maximus dedicated to Tiberius 
a classified collection of exempla in order, as he says, "that those wishing 
to find illustrations might escape the labor of prolong'ed research.'"" The 
common use of such examples soon became a subject of jest. When Seneca 
reminds his pupil Lucilius of the fortitude of the great men of legend, 
he is ready for the rejoinder, "And next you'll tell me of Cato, so that 
I may not fear to die !"" 

In the late Empire the antiquarian spirit had grown, and history seems 
to have been emphasized more than before. Ausonius, a distinguished 
teacher of the fourth century, wrote a series of verses on the tw^elve Cae- 
sars, the length of their reigns, the manner of their deaths, and a few 
other facts about each. Another series continues these curious sketches 
of the emperors as far as Heliogabalus.^* Written by a schoolmaster, the 
pedantic scheme savors strongly of his profession. Whether he required 
his pupils to memorize the verses, we do not know. He also wrote a chron- 
ological table of kings and consuls from Romulus to the year 379 a.d., 
when he himself was consul.^" On one occasion he had his scribes prepare 
a copy of the chronicle of Nepos for the instruction of the children of 
Probus ; on another he wrote a liber 'protrepticus, or outline of studies, 
for his own grandson, in w'hich he recommends the reading of Sallust's 
Catiline and Histories."" He believed that the first business of a gram- 
maticus was to know Cicero, Vergil, and Roman history, and he praised 
the rhetor Staphylius for his mastery of IAyj and Herodotus.^' 

II. Augustine's Criticism of Instruction in History 

Augustine possessed the wide acquaintance with history which was ex- 
pected in his profession. In the City of God he displays an ample knowl- 

" Silvae 4, 7, 53-56. 

" Biv. 1, 50 ; cf. niiet. Her. 4, 101. 

'-^ Suet. Aug. 89, 2. 

^" Praef . 1. 

^' Epist. 26, 6. 

■« Auson. 256-284. 

"" Auson. 405. 

-^^ Auson. 322, 61-65. 

-^ Auson. 212, 13 f.; 210,8. 



318 University of Calif omia Puhlicaiioyis in Classical Philology 

edge of the Roman historians, from Ennius to Eutropius.^ He mentions 
history as a subject of instruction in the schools, a part of the business 
of the grammaticus,"" and informs us that pupils were required to memo- 
rize passages from the writings of Sallust."' But after his conversion he 
turned to criticism of the schools and the sort of history there taught. 
For years he had longed for sapientia, the truth for which philosophers 
searched, but instead had gone on teaching oratory to boys, whose only 
aim, he says, was to excel in the arts of talk (lingnosae artes), gain 
riches, and rule over their fellow men.'' In the orator's career adherence 
to the truth was by no means the invariable rule. Certainly neither 
teachers nor students of Augustine's day were more scrupulous than 
Cicero, who told orators that they might lie in recounting events in order 
to make a point more effective.'" Hence, once a Christian, Augustine 
could look back on his former occupation as a professorship in lying 
(cathedra mendacii) .'" He recalls his troubled emotions at the time of 
his greatest honor, when he delivered his panegyric before the emperor, 
and thus makes his confession to God : "Thou didst cause me to realize 
my misery on the day when I was to recite to the emperor my speech of 
praise, in which I was to tell many lies, and by lying win favor, even 
from men who knew the truth." While he was walking through the streets 
of Milan before delivering his speech, his mind tense with anticipation, 
he met a beggar, evidently well filled from his day's gain, jesting and 
making merry. Augustine turned to his companions with a sigh, remark- 
ing that the beggar had already found the satisfaction which he himself 
could not attain in spite of all his efforts.'"' 

"When converted, Augustine was certain that he had found the truth 
for which he had long been seeking. "Ah, Truth, Truth !" he exclaims, 
"how deeply in the marrow of my soul did I search for thee even then 
[i.e., while nine years a INIanichee] , as they kept sounding thy name ever 
and anon in my ears, but only with their lips, and in many huge books."'' 
When he turned tothe_subject_ of history, he now viewe djt m a new and 
eritic^aTljgTrFTm alnnga^ sharpTTistinctioiTBetween trut lfand error^ ""We 
do n^^rf'^Ti"^T^nTTr^'ps'iary " Hp w'ritesJMr) IipTipvp everytmngjnlhe hlstorv 
of tlie^gans, since even thelTTstorians, as VajTO_wnj^,^iffer_among 
themselves on"mahy pomts7 gT:TTareMIy"jy jtudied_eflfor^ 
But if we^sHTwrmayl^eTieve those things which do not contradict the 

' "Seo S. Angus, T/iVSowrres of the First fen^oolcs of Augustine's City of God 
(Princeton, 1906), pp. 9-49. 

"-^ Docir. Christ. 2, 58; Orel. 2, 12, 37. 
- =» Civ. 3, 17. 

''' Conf. 8, 7, 17 ; 1, 9, 14 ; 1 , 18, 29. 

™ Brutus 42 : Concessum est rhetoribus ementiri in historiis ut aliquid dicere possint 
argntius. 

=" Conf. 9, 2, 4. ■ 

^ Conf. 6, 6, 9. 

==» Co«/. 3, 6, 10 ; 4,1, 1. 



Green : Augustine on the Teaching of History 319 

books which we certainly must believe.""LTlLe_jmyths^fJhe gods, the 
leg-enctS^^nierTTesT'^ndTlie vane'^TaTerof aiitiquity had heretoTore been 
taught togetlierm llie[]scEooTs7_ai^^^ 1matter~fof ' poetry or 

oratory, had not been sharply differentiated. Even tlie Kstorian Livy 
thought it unnecessary to"conSFm or~refute the poetic tales of ancient 
times.'' ButJ:he_Christiansjv iewed the myjhs as lies, invented by demons 
to deceive and~damn the^ople^ Accordingly Augustine is continually 
insisting' on the^ dTstmctibn befweeu true hisfery and fable, rumor, opin- 
ion, or conjecture.^"' Ut course Tie~\vas no more capable of rational criti- 
cism than Livy had been. Many prodigies related in the well-known 
works of history he thinks incredible, but if they do no harm to the 
reader it matters not whether they be true or false."^ But the statements 
of historians must now be tested by a new standard of truth, namely, 
their agreement or disagreement with the Bible, fhe Bible therefor e 
became the basic textbook for learning the events of the past, aU ojther 
documents being"fegarded as inferior and unreliable'suppTements. 

However, when Au"gtistinF^^fin:ds statBmenlslnr pagan authors which 
must be rejected by his new test, he strengthens his case, if possible, by 
pointing to differences already found among the historians."' Thus, in 
spite of its admixture of error, pagan history might be used to refute ^ 
pagan errors. The most notable occasion for such refutation came when "" 
Rome was sacked by Alaric, and the pagans cried out that the desertion, 
of the gods was the cause of all the trouble. In reply Augustine wrote 
his City of God, citing at length the pagan historians to demonstrate 
(1) that Rome had suffered greater ills in former times, and (2) that — 
the gods were proved to be demons by their flagrant immorality. For 
the first point he depends chiefly on Livy, for the second on Varro and 
Euhemerus. 

It w^as in the common secular schools that Augustine had studied these 
authors. But after his conversion he was no longer concerned with those 
schools, and when he refers to them it is usually in a disparaging tone. 
He is aware, however, that many young Christians were still following 
the traditional courses, and for them he has a word of advice. They are 
not to pursue any secular studj^ expecting thereby to attain to the happy 
life, but to scrutinize each soberly, reject everything tainted with super- 
stition and loose living', and retain only what is necessary for the busi- 
ness of this life."'' The necessary subjects are enumerated : those dealing- 
with material objects (including the mechanical arts), the science of 
disputation, and the science of number. The interests of Augustine now 

*' Civ. 21,6. 

^^ Liv. 1, praef ., 3-9. 

^= Civ. 3, 31 ; 6, 4 ; 16, 8.9 ; Epist. 143, 12 ; In Evang. loli. 90, 2. 

"^ Civ. 3, 31. 

^* Civ. 18, 40. 

^'^ Boctr. Christ. 2, 39, 58. 



L^ 



320 University of Calif orn ia PuMications in Classical Ph ilology 

lay in a very different kind of education, wliich was to prepare men for 
their part in the city of God rather than in the ephemeral city of this 
world. From his professorship at Milan he retired to the country for a 
few months, and promptly organized a kind of philosophical school for 
his friends, his mother, and his son. Shortly afterward he became a 
priest in Hippo, and there planned the instruction of candidates for 
baptism and of postulants for orders. The method he used with the first 
group is described in the book De Catechizandis Budihus. The studies 
of the second are best set forth in the four books De Doctrina Christiana. 

III. Augustine's Plan of Instruction in History 

History in Catechetical Instruction. — Of the multitude of pagans being 
converted in the time of Augustine many were admitted into the Church 
without adequate instruction in the Christian faith. This problem Augiis- 
tine attacked with characteristic vigor from the beginning of his priest- 
hood, and thereby soon gained a widespread reputation. About the year 
400, Deogratias, a deacon of Carthage, asked him to set forth his method 
so that others might profit by his experience. The reply is a treatise in 
one book, De Catechizandis Budihus, the most important work of its 
kind which survives from Christian antiquity. 

In that work he explains that when candidates for instruction {ad- 
petentes, postulantes) were brought to him he delivered a discourse, 
interspersed with questions, to introduce them to the Christian faith and 
prepare them for further instruction as catechumeni. The discourse 
(narratio) was chiefly an account of God's dealings with men from the 
Creation down to the present."" It was not complete or detailed, but 
stressed the dividing points (articuli) which marked off the epochs of 
history. One goal was kept constantly in mind — to manifest the love of 
God in the mission of Jesus, and thus to arouse reciprocal love in the 
heart of the convert. He explained that the whole of the Old Testament 
foretold by prophecy and symbol the future advent of Christ and the 
formation of the Church, his spiritual body. For example, when Jacob 
first put forth his hand from his mother's womb, then his head, then his 
body, he was a type of Christ, who first sent patriarchs and prophets 
through five ages (like the five fingers of Jacob's hand) before he him- 
self, the head, appeared ; then followed the Church, his body."^' Whenever 
the interest in the narrative lagged, he remarks, it was revived by no- 
ticing such mystical symbols.™ To make his method clear Augustine 
writes out for Deogratias a model discourse of ten chapters, nearly seven 
thousand words.^" After congratulating the candidate on his decision to 

'^ Catech. Evd. 3, 5 : Narratio plena est cum quisque primo cateehizatur ab eo quod 
scriptum est: In principio fecit Deus caelum et terram, usque ad praesentia tempora 

3' Catech. Eud. 3, 6. ^" Cattch. Eud. 16, 24-25, 49. 

=^ Catech. Eud. 13, 18. 



Green : Augustine on the Teaching of History 321 

apply for instruction, he refers to God's rest on the seventh day as a type 
of the Christian's rest, to be reached in the seventh age, after six ages 
of the world's history are finished. Throughout these ages a struggle goes 
on between two cities, that of the wicked and that of the saints, now 
mingled on earth but destined finally to be separated on the day of judg- 
ment. Five of the ages belong to Old Testament history, beginning re- 
spectively with Adam, Noah, Abraham, David, and the Babylonian 
Captivity. These are filled with symbols pointing forward to the sixth 
age, when man is renewed in the likeness of God even as he was first 
created on the sixth day. Augustine explains that even now the ancient 
prophecies of the Church and its victories over persecution and heresy 
are being visibly fulfilled. This leads to a final exhortation to steadfast- 
ness and righteousness. 

Instruction in Bible history, of course, was not a novelty introduced 
by Augustine, for the Church from its beginning took that history as the 
factual basis of its faith. In the book of Acts two sermons are reported*" 
which consist chiefly of a sketch of Old Testament history interpreted 
to support Christian doctrine. Doubtless such instruction was somehow 
conveyed to every generation of converts, and to children in Christian 
homes, but our literary sources afford little indication of the methods 
used. The apologists of the second century g'ave some attention to chro- 
nology, to show that the sacred books of the Bible were older than the 
writings of the Greeks, and there were frequent appeals to Old Testa- 
ment prophecies and symbols as witnesses to the truth fulfilled in Christ. 
An example is found in Irenaeus' treatise On the Proof of the Apostolic 
Preaching, which sketches portions of Old Testament history as an argu- 
ment for the Faith." More important for comparison with Augustine's 
catechesis is that chapter of the Apostolic Constitutions which discusses 
"How Catechumens Should Be Instructed.'"" Along with other items, 
the candidate was to learn the order of events in Creation and providen- 
tial direction (navdaverco briynovpyia's bca^ibpov ra^iv, irpovoias elpfxov), how 
God in every age {Ka9' eKaaT-qv yevtav) has punished the wicked and glori- 
fied the saints. It is clear that systematic instruction in Bible history was 
given to catechumens, exactly as Augustine prescribes. Unfortunately, 
the date and authorship of the Apostolic Constitutions is uncertain, but 
recent opinion seems to accept the conclusion of Funk that they are the 

*" Chaps. 7 and 13. 

^^ The Armenian text, discovered in 1904, with French and English versions, is 
printed in Patrolngia Orientalis, XII (1919), 653-803. P. Drew, Zeiischrift fiir neu- 
testamentliche JVisscnschaft, VIU (1907), 226, called the work "the oldest church 
catechesis," supporting his claim by citing parallels between it and Augustine's cate- 
chetical treatise. The parallels are not close, and the two works when viewed entire are 
quite different in character. O. Bardenhewer, Geschichte der altlcirchlichen Literatur 
(Freiburg, 1913-1924), I, 411^ correctly classifies Irenaeus' treatise as apologetic 
rather tlian catechetical. 

'- 7, 29, in Patrologia Graeca, I, 1037 f. 



322 University of California Puhlicatiojis in Classical Philology 

work of a Syrian Apollinarist of about 400 a.d." It is quite uncertain 
whether Augustine knew of any similar scheme of catechesis. 

Augustine's model discourse was intended only as an introduction to 
the studies of the catechumenate, but of the remaining studies we have 
no account. Along with those already baptized, the catechumens were 
admitted to hear sermons, and were known as audientes. In the sermons 
of Augustine one finds much of Bible history, with frequent mention of 
the six ages." It is probable that candidates also returned for further 
instruction along the lines which had been indicated in the introductory 
discourse." 

TliiS jnost striking feature of that discourse is the scheme of the six 
ag:es,jwli ich goes _hey^T'^1 Bi ble history to otfe r the learner a pattern of 
universal history as the basis of the Christian faith . The genesis of that 
scheme deserves a more detailed investigation than^t has yet received. 
ThjeiLeatas_a_notion^fJ^ig standing that each of the six days of Creation 
represents,^ 4-housan3^:eai:&-jof -workl hist ofyTThe first hint of this is 
found in the Book of JtiTjilees, a Jewish work of the second century be- 
fore our era. Since the Psalmist had said that "a thousand years in Thy 
sight are but as yesterday when it is past," the author infers that Adam's 
death at the age of 930 was a fulfillment of the word of the Lord that he 
should die on the same day on which he ate of the forbidden fruit.'" About 
the beginning' of the Christian era the author of the Book of the Secrets 
of Enoch arrived at the notion of a world week of seven thousand years.*' 
The seven ages, and the seven thousand years, are also mentioned in the 
Jewish Testament of Abraham, in the first century.** In the Talmud, 
Rabbi Kattina of the third century is named as author of the doctrine 
that the world must last six thousand years, then be destroyed and re- 
main destroyed for one thousand, apparently in recurring cycles.*" In 
Christian literature the idea first appears in the Epistle of Barnahas (of 
uncertain date, but before 150 a.d.) It is there explained that the days 
of Creation represent six thousand years of world history, which are to 
be followed by the Lord's return and a thousand years of Sabbath rest 
for the saints." Irenaeus (about 180) states the view more explicitly, 
and in the third century it was made the basis of the first comprehensive 

'^ Bardenhewer, IV, 271. 

" Serm. 125, 4 ; 259, 2 ; hi Evcwfj. loh. 9, 6 ; In Psalm. 92, 1. 
^ *^ On audientes see Serm. 132, 1. G. Bareillo, in Bieiionnaire de Iheolotjie catholique, 
11:2, 1884 f., enlarges on the probable subject matter of subsequent instruction, but 
acknoAvledges an unfortunate lack of documentary evidence. 

'^ Jubilees 4, 30, quoted in K. H. Charles, Book of the Secrets of Enoch (Oxford, 
1896), pp. 45 f. 

'" Charles, loc. cit. 

'^Tcst. Abr. Recensio A, 19; Rec. B, 7, in Tcrts and Studies, 11:2 (Cambridge, 
1892), pp. 101,112. 

'"Sank. 97 a, inL. Goldschmit, Dcr babylonische Talmud, VII (Berlin, 1903), 
422. 

'"Barnab. 15, 4. 



Green : Augustine on the Teaching of History 323 

Christian chronology by Sextus Julius Africanus.'' It was taught by 
Lactantius, and mentioned by Jerome."' 

The teaching of Ambrose on this point is important, as preparing the 
way for Augustine. He frequently gives a symbolic meaning to the Crea- 
tion week, and to God's Sabbath rest, but his explanations are not alto- 
g-ether clear or consistent. In one passage he says that as God rested 
on the Sabbath day, so Jesus has now brought rest to the faithful by his 
resurrection."" Elsewhere he states that the eighth day is the resurrection 
day, representing the age in which Christians now live, since the seven 
ages of the old world have passed away."^ Two passages refer to the 
familiar theory that one day represented a thousand years, but each 
time it is brushed aside in favor of what Ambrose believed a more spir- 
itual interpretation. Once he opposed the notion of six thousand years 
as the world's duration by declaring that more than that number of years 
had already passed."' 

Although Augustine was at first a believer in the millennial Sabbath, 
there is no evidence that he ever held the old view of the six thousand 
years' duration.'" In one passage he mentions that theory along with his 
own plan of the six ages, saying that according to either scheme of reck- 
oning Christ came in the sixth age.'' The text from the Psalms seemed to 
him inadequate as proof for two reasons : (1) the statement that "a thou- 
sand years in Thy sight are but as yesterday when it is past" was inap- 
propriate if the psalmist intended to refer to the period of time (much 
more than a thousand years) which had passed when he wrote ; and (2) 
the following phrase, "as a watch in the night," cannot be adjusted to fit 
the theory."' The theory, furthermore, is in conflict with the statement 
of the Lord that no one, not even the Son, knows the time of the end.'''_ 

But no exegete of that age could well abandon one symbolic interpre- 
tation without finding: another to take its place, and the circumstances of 

^^ Iren. Adv. Eaeres. 5, 28, 3 ; H. Gelzer, Sextus Julhis Africanus und die hysan- 
tinische Clironographie (Leipzig, 1898), pp. 24-26. 

=^ Lact. Inst. 7, 14, 7; Hier. Epist. 140, 8 (on Ps. 89, 4) : Ergo arbitror ex hoe loco 
et ex epistula quae nomine Petri apostoli inscribitur mille annos pro una die solitos 
appellari ; ut scilicet quia mundus in sex diebus f abricatus est, sex milibus annorum 
tantum credatur subsistere. The words arhitror and eredatur mark Jerome's reserve 
on tlie point. 

-"^ Epist. 44, 10. 

=^ Epist. 44, 16. cf. loscph 7, 39, where the septem anni istius mundi are swallowed 
up by the subsequent ages of the lex spiritalis. 

'' in Luc. 7, 7 : Post sex dies . . . de quo possumus dicere, post sex milia annorum — 
mille enim anni in conspectu Dei tamquam dies una — sed plures quam sex milia com- 
putantur anni, et malumus sex dies per symbolum intellegere. Cf . Zn Luc. 7, 22. 

'"•'' Civ. 20, 7 refers to his earlier belief in the millennium. But in an early sermon 
(Serm. 259, 2), where he makes the millennium the seventh age, to be followed by an 
eighth age after the resurrection, he has already adopted the scheme of the six ages 
based on Matt. 1, and the six-thpusand-year theory seems thus excluded. 

=• De Trin. 4, 7. 

'•" In Psalm. 89, 5; cf. Epist. 199, 17 f.: sed aliud est nosse aliquid, aliud suspicari. 

^'' In Psalm. 6, 1. 



324 University of California Puhlicafions in Classical Philology 

Augustine's conversion made it doubly impossible for him. For years lie 
had been deterred from accepting the Catholic faith by the Manichaean 
conviction that the Old Testament was absurd. But Ambrose taught him 
that the apparent absurdities were really symbols of spiritual truth, even 
as St. Paul had said, "the letter killeth, but the spirit giveth life."*" 
Augustine's conversion thus rested upon his acceptance of the alleg'orical 
method. As for the seven days of Genesis, the apostle had already de- 
clared that the Sabbath was a type of the rest which remains for the 
people of God, and hence the preceding six days must represent six 
ages of toil on earth."' The first two ages of the world Augustine found 
to be obvious and familiar to all, ending with Noah and Abraham.^' The 
next three he found marked off in the first chapter of IMatthew, and the 
sixth, of course, must be the age which begins with Christ. These ages 
are of unequal length, with ten generations of men in the first and in the 
second, and fourteen generations, as ]\Iatthew points out, in each of the 
next three. The number in the sixth is undetermined, and may even 
equal the sum of all the others. All this had a symbolic mean^iing (which 
he goes on to explain), and found a parallel in the six ages of a man's 
life, in which old age begins at sixty, and may itself last another sixty 
years.**^ 

The circumstances under which this doctrine is first set forth by 
Augustine seem clearly to mark him as its inventor. At the time of his 
conversion in 387, Augustine had learned the allegorical method from 
Ambrose. But he did not take over any of his interpretations of the seven 
days, which doubtless seemed to him to lack precision and clarity. On his 
return to Africa, in 388, he was immediately involved in controversy 
with the Manichees, who had -been his former associates. Either that 
year, or very soon after, he published his treatise De Genesi contra Mani- 
chaeos, using against them the allegorical method which had brought 
about his own conversion, and inventing the new scheme of history which 
was to dominate the teaching of tbat subject for a thousand years. 

The novelty of the scheme appears in two aspects : as against the casual 
instruction in history given in pagan schools there was opposed a well- 
defined outline of world history based on the Bible; and as against the 
Jewish and Christian notion of six thousand years of history (with no 
natural division into ages) there was opposed the plan of six ages, clearly 
defined by important crises of Bible history. As priest and bishop Augus- 
tine found his pattern useful for instructing beginners in Christian 
doctrine. It not only provided a convenient plan of Bible history, but 
left room for the continuation of history till the end of the world. The 

""Cow/. 5, 14, 24; 6. 5, 6 f. 

"^ c. Adimant. 16, 3: Interpretatiir apostolus sabbatum ad Ilebraeos cum dicit : 
Remanet igitur sab'batisunis populo Dei. Cf. VuJg. Hehr. 4, 9. 

"" Divers. Quaest. 58, 2 : qui articuli sunt evidcntissinii et notissimi. 
«=• Gen. c. Manich. 1, 35-43. 



Green : Aurjustine on the Teaching of History 325 

student was not to think of a sharp break at the close of the apostolic 
age, but was rather shown the continued revelation of God's hand in the 
progress of the Church. The witness of the martja^s, the conversion of 
kingdoms, and the falling away of heretics were all foretold by God, and 
hence are a continuing demonstration of the truth of His word." The 
convert was therefore bidden to look ahead with confidence to the end 
of the age and the future reward of the saints. History and revelation 
merge into one symmetrical pattern, in which the believer has his own 
stable place, fixed in the eternity of God's beneficent purpose. 

History in the Instruction of the Clergy. — Augustine's work included 
the teaching of clergy as well as laity. When he was ordained a presbyter 
at Hippo, he established a monastery which soon became in effect a semi- 
nary for the training of the clergy of that city {monasterium clericorum) . 
Indeed, many of the monks became bishops elsewhere, and established 
similar institutions in their own cities."'' Not all monks were elevated to 
clerical orders, but only those with sufficient continence, necessary in- 
struction, and a suitable perfection of character."' Of the studies pursued 
in the monastery we have no exact account. For the nuns of the same 
city Augustine prescribed a fixed hour for the daily issuance of books, 
and quiet attention to the lesson read at each meal— a practice of Cap- 
padocian origin which was generally adopted in the "West."' Doubtless 
there were similar rules for men, along with a program of instruction 
intended both for private edification and for training preachers. Augus- 
tine's biographer states that he was especially eager to give instruction 
to future teachers."^ 

Of his extant works the De Doctrina Christiana doubtless gives the 
best view of the content of that instruction. He there states that the 
training of the preacher should consist in finding the meaning of Scrip- 
ture and in mastering the method of expounding it to others."" To under- 
stand its meaning he says that one should have a knowledge both of 
words and of the things signified by words, including natural objects, 
numbers, music, and the institutions and history of men.™ Whatever one 
has learned, even as a child in the pagan schools, about the order of past 
events, is useful in interpreting Scripture. Ambrose demonstrated this, 
he says, in refuting the Platonists' claim that Jesus had borrowed his 
teaching from them, by arguing that since Plato was in Egypt in the 
time of Jeremiah he may have got his ideas from the Hebrew prophet."' 

«^ Catech. Bud, 24, 44 f. 

^^Possid. Vita Aug. 5, 11; Aug. Serm. 355, 2. See H. Leclerq, L'Afrique cliretienne 
(Paris, 1904), II, 72 f. 

"" Epist. 60, 1 : sufficiens continentia, instructio necessaria, personae regularis in- 
tegritas. , 

"' Epist. 211, 8.13 ; Cassian. Inst. 4, 17 ; Bened. Beg. 38. 

^ Possid. Vita Aug. 19. "" Doctr. Christ. 2, 10, 15 - 2, 28, 42. 

«=> Doctr. Christ. 1, 1. '^ Doctr. Christ. 2, 28, 42 f . 



326 University of California Publications in Classical Philology 

In citing this case Augustine reveals his special interest in chronology 
which synchronized secular and Bible history. Byi^_theantiquity of 
God's revelation was con£irmed,,,aji4 the Old Testament stories~vvefe~set 
m. thaSoIid iraniework^of unquestioned world events. 
/^ From this mention of Ambrose and Plato, Augustine^goes on to draw 
/ a distinction between historical writings and the facts of history^- "Al- 
thiTiiglr the pa'sriristitutions~oFmen are relate'd in Hstorical narrative, 
history itself is not to be counted among human institutions. For events 
which have already occurred and cannot be undone are to be considered 
as part of the world order, whose founder and ruler is God.'"^ The very 
; word ''history!ljthus receives a new a nd-jjistinctly ChristiajL-meaning. 
For the Roman, it was simply "the_kno wledge of the past," or ^ narra- 
tive of past events." As suchTit was an entirely human affair — what men 
knew, or related, aboutjthe past. Augu sliheeommonly uses the word in_ 
this customary meaning. But above this incomplete and uncertain human 
^^ knowledge is a higher Jiistpria *|)saj^ a planned ordo temporum, which 
^ belongs to divine^jroyidence. The real aim in Augustine's teaching was 
toimpart _an underst anding of that plan of history, conceived by God. 



It has previously been remarked that the study of the historians had 
been a part of the program of liberal arts. These Augustine admitted as 
part of the training of the elergj^ Shortly after his conversion he had 
begun treatises on seven of Varro's disciplinae, hoping, as he says, to 
ascend, and lead others, by these corporeal steps to the incorporeal 
realities.'* Later, when as bishop he wrote the De Doctrina Christiana, 
he still believed these subjects useful to the preacher of the gospel. Al- 
though they were of pagan origin, the Christian was justified in taking 
them for a better use, even as the Israelites had taken from the Egyp- 
tians their silver and gold and garments in order to devote them to the 
Lord.'' But Augustine's interest in the disciplines was so slight that he 
never completed the treatises begini in INHlan. Some years later, when 
a friend requested his book De Musica for the instruction of his sou, 
Augustine found that only the sixth book had been properly revised. 
His duties, he explains, allowed no time for such trivialities. Of the other 
liberal arts, onlyhi story, § aiar as it relates_the tru th, is worth knowing, 
a;j d mthaTstucIyone must beware of errors nTthe^writingsofTneh not 

inspired of God."^"" " 

^ A more detailed view of Augustine's advanced instruction in history 
may be gained from the City of God. After refuting the charges against 

~" Doctr. Christ. 2, 28, 44. Narratione autem historiea cum praetcrita etiani liomi- 
num instituta narrantur, iion inter humana instituta ipsa liistoria nnnieranda est, 
quia iam quae transierunt nee inf ccta fieri possunt, in ordine temporum habcnda sunt, 
quorum est conditor et administrator Dens. 

■^ Beiract. 1, 6. 

"' Docfr. Christ. 2; 40, 60. The figure was invented by Origen, and had a lasting 
popularity; see E. Norden, Die antike Kttnstprosa (Leipzig, 1898), pp. 676-679. 

' ■ Epist. 101. On the revision of De Mvsica see Marrou, pp. 580-683. 



Green : Augustine on the Teaching of History 327 

the Christians, he proceeds in the last twelve books to expound the provi- 
dence of God in world history. Books XI to XIV set forth the origins 
of the city of God and the city of this world, Books XV to XVIII their 
progress, and Books XIX to XXII their destined end. The progress of 
the cities is marked off by the familiar divisions of the six ages. Book 
XV discusses the first age, from Adam to Noah, and Book XVI the 
second and third, with due notice of the articulus reached in Abraham.'' 
At the end of this book the scheme is explained for the first time, with 
a reference to the passage in Matthew on which it is based. A critic 
unfamiliar with Augustine's previous w^riting-s might be annoyed to find 
the explanation thus delayed, but the tacit assumption obviously was 
that the readers already understood the scheme. Book XVII deals with 
the fourth and fifth ages; Book XVIII again reviews the general plan, 
then goes back to recount the affairs of the pagans since the time of 
Abraham, and continues with the sixth age. The last three books deal 
with future things, finally closing with a rapturous description of the 
seventh age, the Sabbath of rest for the people of God. 

In this grand perspective of history much is left incomplete. Only 
the eighteenth book takes up the history of the pagan nations, and even 
there little more is done than to pair off the names of Biblical persons 
with contemporary kings of Assyria, Egypt, Sicyon, and so on. Augus- 
tine at this point admits his dependence on the chronicle of Eusebius, 
whose value Ambrose had already demonstrated to him," but he does not 
attempt to revise or adapt that chronicle to his new scheme of the six 
ages. To do so he would have had to include the period before Abraham, 
omitted by Eusebius. But in the chronology of that period there was a 
wide discrepancy between the Hebrew text of Genesis, lately translated 
in Jerome's Vulgate, and the Septuagint version, on which the older 
Latin Bible was based. Augustine generally preferred the latter, but was 
perplexed about the figures for Methusaleh, which made him outlive the 
flood by fourteen years.'' The problem was left to vex chroniclers until 

the time of Bede. 

IV. Successors of Augustine 

The general acceptance of Augustine's pattern of the six ages by medie- 
val chroniclers is sufficiently well understood. A detailed study of his- 
torical instruction in the Middle Ages — which has never been made — 
would go far beyond the sc/)pe of this paper and would require the scru- 
tiny of a larg'e number of unpublished chronicles. Something may be 

"« Civ. 16, 12. Throughout these books there are numerous digressions which seem 
to violate all the canons of unity and coherence. On this point see Marrou, pp. 59-76. 

■' Civ. 4, 6 : Nam sicut scribunt qui chronicam historiam persecuti sunt, mille 
ducentos et quadraginta annos Though no author is named, the reference is cer- 
tainly to Jerome's translation of Eusebius. See Angus, pp. 51, 154. On Ambrose's use 
of the chronicle see note 71 above. 

™ Civ. 15, 10-13. 



328 University of California PuUicaiions in Classical Philology 

said, however, of the immediate influence of Augustine, and of some 
representative teachers who followed his method in later times. 

The first historian inspired by Augustine was his friend Orosius, who 
wrote at his request seven books of History Against the Pagans, to con- 
tinue the arguments of the City of God. Orosius, to be sure, does not 
introduce the six ages into his treatise. They would have been no more 
appropriate there than in the first half of the City of God, of which 
Orosius' work is a continuation. He does, however, conform to Augus- 
tine's pattern on one important point, insisting that history must be 
viewed as a whole, beginning with Adam. He remarks that 3,184 years 
from Adam to the time of Ninus and Abraham have been omitted or 
ignored by all historiographers.™ That period, he avers, must be dis- 
cussed in some manner, however briefly, to enforce the point that suffer- 
ings and disasters go back to the very beginning, and are a divine 
punishment for the sins of men. His discussion is limited to a single 
chapter on the fall of Adam and the Flood. To the casual reader it seems 
perfunctorily introduced, with little bearing on the main argument ; 
but it is clear evidence of the acceptance of A ugustine's opm ion tjiat 
world history must herea fter be c onsidered injts entirety^withjdrv'iue 
providence~aslts"giiidm g principle . 

AnotheFcohtemporaix Prosper of Aquitaine, continued the work of 
Augustine in two fields, in the refutation of heresy and in chronography. 
In his later works Augustine was much concerned with the errors of 
Pelagius, and Prosper continued the debate with such zeal that he was 
called Augustine's best disciple."" Besides other works, he wrote an epit- 
ome of Jerome's chronicle, extended back to Adam and down to his own 
time.*' It is the first of the many chronicles which continue the work of 
Jerome throughout the IMiddle Ages. Here, as in the works of Orosius, the 
addition of the period before Abraham indicates the influence of Augus- 
tine. The frequent notice of heresies also well conforms to his plan of 
giving instruction in history usque ad praeseutia tempora ecclesiae. ^ 

Little can be said about the curriculum of schools in the fifth and sixth 
centuries. The schools established in Africa by Augustine and his disci- 
ples must have been swept away by the Vandals, and all the western 
provinces were ravaged by similar invasions. But here and there, as 
opportunity offered, new schools were organized.'' Isidore, who became 
bishop of Seville about the year 600, had a part in that task. He wrote 

™ Oros. Eist. 1, 1, 4 f. 

80 Bardenhewer, IV, 539. , ^ ^, 

SI In MGH A A IX, 385 ff. Prosper's conteniporarj Victorius remarks about the 
change made in beginning with Creation (ibid., p. 345). 

*'"0n the decline of the schools see M. Eoger, L'Enscifjnmeni dr.s Iciires class-iques 
d'Ausone a Alciiin (Paris, 1905), pp. 153-155. In the sixth century, when disturbed 
conditions in Italy prevented Cassiodorus from establishing a school at Eome, he 
built a monastery on his o^^•n estate and drew up a plan of study for the monks which 



Green : Augustine on the Teaching of History 329 

a riile for the monks, based on that of Benedict but with more specific 
provision for the education of children who had been given to the mon- 
astery."" He presided at the Council of Toledo in 633, where further 
regulations were made for two types of schools. Young men and boys 
"of the clergy" {in clero) were to be kept together in one room of the 
bishop's house, under the charge of an approved teacher. Those who 
were not amenable to his discipline were to be assigned to monasteries 
and there kept under a stricter restraint." 

The content of instruction given in the schools of the time appears 
in Isidore's most famous work, the Origpies (also called Etymologiae) . 
It is a sort of encyclopedia, which, as a contemporary declared,*'" "has 
collected in brief form almost everything which needs to be known." In 
this compendium of universal knowledge history is represented by a 
world chronicle of the six ages as defined by Augustine.^ This section 
is an epitome of a longer chronicle earlier written by Isidore, based on 
the chronicles of Jerome and Prosper but wdth the notable addition of 
the sex aetates as a framework.^' The idea was evidently suggested by 
the City of God, from which frequent excerpts were made. Like Augus- 
tine, Isidore closes by saying that the remaining time of the sixth age 
is uncertain since Christ has said that it is not ours to know the times 
and seasons. Elsewhere, also, Isidore follows Augustine in commenting 
on the six days of Creation, the six ages, and the perfection of the 
number six.*" 

The dominant place of Isidore in the learning of the seventh century 
is well matched by that of Bede in the eighth. Born in Northumbria in 
672 or 673, he was given at seven to the monastery at Wearmouth. "From 

included the chief works of Augustine and the historical works of Eusebius and 
Jerome and their continuators. Benedict of Nursia took bovs to rear in his monastery, 
where there was a regular schedule of studies. See G. C. Coulton, Five Centuries of 
Beligion (Cambridge, 1922-1936), I, 222 f. 

^= Isid. Beg. Monach. 20, 5 (PL 83, 891). Cf. Vita S. Isidori, PL 82, 36. 

-'^ Cone. Tolet. a. 633 Mansi 10, 626. 

'^^ Braulio in PL 82, 67. The question arises whether the Origines was intended to 
be a sehoolbook or a general reference work. E. Brehaut, An Encyclopedist of the 
Darlc Ages (New York, 1912), pp. 86-88, notices the parallel between Isidore's work 
and the I)istitutes of Cassiodorus, which avowedly sets forth a program of studies, 
concluding that the first eight books of the Origines form an "encyclopedia of educa- 
tion," and the remainder an encyclopedia of all knowledge. While it is true that the 
first books set forth the traditional disciplinae in regular order, beginning with gram- 
mar, the division at the end of Book VIII seems artificial, and the distinction between 
two kinds of "encyclopedias" is entirely modern. For Isidore, "education" and "knowl- 
edge" were synonymous; Orig. 1, 1: Disciplina a discendo nomen accepit, unde et 
seientia dici potest. By discipliiih Isidore means a subject of school instruction, as 
is evident in the following section De septem liberalibits diseiplinis. The obvious in- 
tention of both Cassiodorus and Isidore was to include all topics which were impor- 
tant for a monk or priest to know. 

*» At the close of Book V. 

«' MGH AA XI, 427-448. 

^^^ Lil. Num. 28 f . (PL 83, 184) ; cf . Aug. Civ. 11, 30. 



330 Vniversity of California Publications in Classical Philology 

that time," he writes in his fifty-ninth year, "spending my entire life 
in the walls of the same monastery, I have given all my attention to 
meditating' on the Scriptures, and have always held it pleasant either 
to learn, or teach, or write.'"" His works are almost as encyclopedic in 
their range as those of Isidore, and throughout display the hand of the 
schoolmaster. They include treatises on grammar, natural science, his- 
tory and chronology, and theology. One of his earliest works, written 
in 703, was a textbook for students on the subject of time reckoning and 
chronology."" It is in large part based on the corresponding sections of 
Isidore's Origpies, including a chronicle of the six ages. One change, 
however, was introduced — the length of the first four ages was reckoned 
according to the Hebrew text."' Bede's little chronicle, De Temporihus, 
soon became known outside his own monastery, and its innovations led 
to a charge of heresy. The letter in which he defends himself, written in 
708 to Plegwin, a monk of Hexham, gives valuable information about 
the historical theories then current. (1) Bede assumes that the doctrine 
of the six ages, in the last of which Christ was born, was universally 
taught by the Church. (2) The chronology of Eusebius and Isidore, 
based on the Septuagint, was so firmly established in England that a 
departure from it was branded as heres3\ (3) The common people and 
uninstructed monks still clung to the six-thousand-year theory, and were 
eager to speculate on the approaching time of the end."" 

Bede's short De Temporihus proved inadequate for the needs of his 
disciples, so in 725 he prepared his larger and better-known work De 
Temporum Eatione. In compiling it he di"ew from a wide range of writ- 
ers, including Augstine. In the preface he defends the Hebrew chron- 
ology, but remarks that the figures of the Septuagint are added in order 
that the reader may choose for himself. The work closes with five chap- 
ters on future things, refuting the six-thousand-year theory and the 
expectation of an earthly millennium. The reader is exhorted to reject 
these heresies, cling to the Catholic faith in the six ages, and enter into 
the Sabbath of eternal rest. The whole work, like its predecessor De 
Temporihus, was intended as a manual for students, with further details 

«» Hist. Eccl. 5, 24, ed. C. Plummer (Oxford, 1896). 

^ De Temporihus, in J. A. Giles, The MisceUaneovs Worls of Venerahle Bede (Lon- 
don, 1843), VI, 12.3-138. The purpose of the work is explained in the preface of De 
Temporum Eatione, Giles, VI, 139. 

™ Mommsen, MGII A A XI, 241, states that Bede's calculation of the length of the 
'ages Avas not part of the shorter chronicle, but a later addition, taken from the longer 
work of 725. The letter written to Plegwin in 708, cited below, is a complete refuta- 
tion of this view. 

"- Epist. 3, Giles, I, 144-154. Among other interesting details of this letter is the 
mention of a chronographus heresiarchcs whose work Bede had read when a boy. Like 
Q. Julius Hilarianus, who wrote a chronicle in 397, he fixed the date of Christ's birth 
at 5500 A.M., the end at 6000. But the extant chronicle of Hilarianus cannot, as W. 
Levison (see work cited in note 1 above, pp. 116, 122), supposes, be the book Bdlo 
refers to. The latter contains several striking details not found in the former. 



Green : Augustine on the Teaching of History 331 

to be supplied in the master's oral discussion.^^ The importance Bede 
attached to the six ages is evident throughout his writings. He constantly 
interprets Scripture by allegory in which the numbers five, six, seven, 
eight, and the mention of the Sabbath remind him of the ages of the 
world."' He even made the topic the subject of a long hymn, which Alcuin 
described as "most beautiful.'"" 

One more teacher will be mentioned, coming four centuries after 
Bede. Hugo of St. Victor was born in Saxony about 1086, and became 
a student, teacher, and master of the school at St. Victor's monastery-, 
in Paris. He was widely hailed as a second Augustine, the most learned 
teacher of his age. His course in theology was based on the three meth- 
ods of Scripture interpretation : history, allegory, tropology. These, he 
states, constitute disciplines which must be mastered in the proper order, 
history coming first to serve as a basis for the others."* Hence the very 
first book to be placed in the hands of his students was a syllabus of 
history, De Trihus Maximis Circumstantiis Gestorum, id est Personis 
Locis Temporihus. It begins with a prologue which states the basic im- 
portance of history and the need of a fully memorized outline of the 
subject and gives a brief explanation of the six days and six ages. Then 
follow seventy-four pages of names of persons and places and of chrono- 
logical data, all to be memorized. ]Most of the material is adapted from 
Bede and later chroniclers. In the works of Hugo collected by his abbot, 
Gilduin, the first volume contained his more elementary schoolbooks, 
among which this history book comes first. Copies were rapidly mul- 
tiplied in France, Germany, the Low Countries, and England. Its 
dissemination is perhaps the best indication of the widespread and long- 
continued teaching of Augustine's six ages in medieval schools."-" — 

The change instituted by Augustine in the teaching of history corres- 
ponds to the change in world outlook between ancient and medieval 
times. ^^oiLthe_Romans, historia was that know ledge of the jast whi ch 

«3 Giles, VI, 182, 253. 

«* The number five : GUes, VII, 275 f . ; VIII, 12, 323 ; six : V, 323 ; IX, 24 ; XII, 49 ; 
seven: VIII, 231; eight: VIII, 82; X, 132, 320; XII, 255; the Sabbath: X, 38; 
XI, 6, 185. Bede invents a new conception of the seventh and eighth ages. The seventh 
is the rest of the saints between death and the resurrection, contemporary with all 
the other six ages since the death of Abel. The eighth will follow the Judgment. This 
does not, as Levison, p. 121, says, follow Augustine, who never mentions an eighth 
age except in passages written while still a millenarian (see note 56 above). 

"5 Giles, I, 78-81 ; Plummer, I, cliii. 

^ PL 175, 777-799. On Hugo's course of study see G. Pare, A. Brunet, P. Tremblay, 
La Renaissance du XII^ siecle (Paris, 1933). 

*" A description of this work, with classification of the manuscripts, text of the 
prologue, and summary of the remainder, is published by the present writer in Spec- 
ulum, XVIII (1943), 484-493. This study should obviously be continued further to 
give a complete account of the six ages in medieval instruction. Much can doubtless 
be learned from published works, such as the Speculum Historiale of Vincent of 
Beauvais, and still more from unpublished chronicles, such as the work of Hugo here 
cited. 



332 University of California Publications in Classical Philology 

was preserved in literature and or al traditi on. It might serve to erdiance 
on«'s prideinTiis country orrluilfamflyi prToTfernisli objectjessons. as 
a guide to conduct. The narrative art was often cultivated for its own 
sake, with attention fixed on style and effect. There was little attempt 
to formulate a philosophy of history. From the time of Ennius to that 
of Augustine the patriotic, ethical, and stylistic motives appear domi- 
nant throughout, influencing writers in varj'ing degrees. The earliest 
works arose in the patriotic fervor of the Punic Wars ; Sallust and Livy 
were most concerned about the moral decline of their times ; and Tacitus, 
it would seem, was thinking chiefly of the effect produced \ij his style. 
The same three motives determined the method of teaching history in 
the schools, but the greatest of these was style. 

For^the Christian^ji the other hand, history was the revelation of a 
diyjiie plan th e chief events of which are set forth in th e Bible. At the 
moment when Christianity won its final victory over paganism, it fell 
to the lot of Ajigustine to devise the p atteini_bvwh ich histor y^ould be 
taught from a Christian vie wi)ojntNeither_^atriotism nor narratTve. 
style was JigjDortant, and even th e moral l essons were but one aspect^ 
Grod^ssovereignty. The six ages of history were viewedasGoHY^ork, 
parallel totheworkof the six days of Creation. All mankind was viewed 
in one perspective, where there was neither Greek nor Jew, barbarian, 
Scythian, bond nor free. The propagation of this view was gradually 
undertaken in the Christian schools, and became an essential part of 
the culture of the Middle Ages. 



-ifi 



THE SINDUVARA TREE IN SANSKRIT 
LITERATURE 

BY 

M. B. EMENEAU 

That classical Sanskrit poetical literature (kdvya) is decorated to a high 
degree with comparisons in the form of similes and metaphors is well 
known, and likewise that there is a large traditional stock of such com- 
parisons upon which all poetical writers alike draw. There is little striving 
after new likenesses; the originality of the poet lies in his verbal adapta- 
tions of the old stock. An encyclopedia of this rich stock should be pos- 
sible, and would be an aid to the scholar who occupies himself with the 
interpretation of Sanskrit literature. Whether it will ever be produced in 
the West, it is hard to say. Few Western sch'olars have the necessary close 
acquaintance with the vast amount of kdvya literature that is extant in 
comparatively good printed editions; still fewer have the hardihood to 
attempt a comprehensive assault upon the even vaster masses that lie 
buried in bad printed editions or in manuscripts. Add to this that kdvya 
is essentially a product of cleverness, which tends to show itself in diffi- 
cult verbal involutions, and the possibility that such an encyclopedia will 
ever be produced, except by a committee of scholars, each of whom could 
contribute from his studies in particular texts, becomes slight. 

The interpretation of a passage involving the sinduvdra tree in a com- 
parison has led me to collect, in rather desultory fashion but during the 
reading of a number of j'ears,^ all the passages that I could find invoking 
the sinduvdra. Such a study of them as I make here would be the pre- 
liminary to framing an entry for tLj encyclopedia envisaged above, and 
perhaps also one of the most A^aluable of its end results, for the value of 
the encj^clopedia would resi, in the last analj^sis, in the fact that it would 
allow one to trace each element in the poetical stock-in-trade through its 

^ The works that I have searched, either by careful or hasty reading in the original 
or in translation, or by consulting indexes when they exist, include the compositions of 
A^vaghosa, Bhasa, Kalidasa, and Bana, Subandhu's Vasavadatta, Bhavabhuti's 
Uttararamacarita, Harsadeva's Priyadar^ika, Raja^ekhara's Karpuramanjarl and 
Viddha^alabhaiijilca, the Vetalapancavimsati in Ksemendra's Brhatkathamanjari, the 
Pancat_antra in Edgerton's reconstructed text, and of Jain texis Hemacandra's Trisas- 
tisalakapurusacaritra, books 1-3, Merutunga's Prabandhacintamani, Dhanavala's 
Bhavisattakaha, and the Sanatkumaracaritam in Haribhadra's Xeminathacaritam. 
The Sanskritist will note that I have included Prakrit passages without mentioning 
that they are such. Renderings of passages are taken from the printed translations 
except for the Mahabharata, Ksemendra's Brhatkathamanjari, or where otherwise in- 
dicated. 

[333 ] 



334 University of California Publications in Classical Philology 

vicissitudes in the whole of Indie Uterature. On the basis of such historical 
and comparative studies our judgments of authors' merits and demerits 
could be made firm. The present study also makes certain the botanical 
identification of the plant, which has recently been called in question. 

The plant has in the Sanski'it lexicons various names, sinduvara, 
sinduvdraka, sinduka, sindhuvdra, sindhuvdrikd, sindhuvdrita, sindhuka, 
kapikd, indrasurasa, indranikd, nirgundl, nirgunt{h)l. Of these only 
sind{h)uvdra and sinduvdraka seem to occur in the kdvya literature. The 
plant has been identified as Vitex negundo, Linn., and /or Vitex trifolia, 
Linn. The botanical books are clear that only a botanist can distinguish 
these two species ; the Hindus do not do so. 

In a number of the passages collected, reference to this plant is non- 
distinctive, that is, it might just as well be to various other plants. In two 
epic passages, Mahabharata 3.296.43 (Poona ed.; 3.312.45 Bombay ed.; 
3.17286 Calcutta ed.) and Ramayana 3.79.34 (Gorresio), the plant occurs 
with others in lists of those that adorn the forest. The southern text of 
Subandhu's Vasavadatta (Louis H. Gray, Vdsavadattd, a Sanskrit Ro- 
mance by Suhandhu [ Columbia University Indo-Iranian Series 8 (1913) ] , 
text on p. 186, translation on p. 125; the page reference to Fitzedward 
Hall's ed. is [263]) has: pardgapvnjapinjarasinduvdrarajyaiymnamadhu- 
karamanjusinjitajanitajanamudd, an epithet oi mahdsdgarakacchopdntena 
(p. 187, Fitzedward Hall [266]), "along the great ocean's lagoon . . . pro- 
ducing delight for mankind by the sweet humming of bees glittering from 
the sinduvdra-trees, tawny with masses of pollen." Here the sinduvara 
might be replaced by any plant whose flowers were attractive to bees be- 
cause of their pollen; the reference is nondistinctive. Another nondistinc- 
tive passage, referring to the perfume of the blossoms, is found in the 
Jain text Trisagtisalakapurusacaritra 1.2.998: 

sinduvareria durvarakusumamodasarhpada 
cakre ghranavi§eiie 'va mahamohah -pravasindm 

"The sinduvara caused great stupor to travelers by its blossoms' wealth 
of perfume hard to restrain like poison in the nose."- 

In an epic passage, Mahabharata 3.155.57 (Poona ed.; 3.158.65 Bom- 
bay ed. ; 3.1 1588 Calcutta ed.), we meet the tree for the first time involved 
in a simile. Its use, however, is again nondistinctive, for the comparison 
is with the "darts of the Love-god," which in kdvya are made of any 
flowers, and the passage is, moreover, found in the midst of a long series 

^ I have used the edition issued hy the Srljainadharmaprasarakasabh.^ in Bhavna- 
gar, and the translation of the .A.dlsvaracaritra by Helen M. Johnson in the Gaekwad's 
Oriental Series ol (1931), p. 157. 



Emeneau: The Sinduvara Tree in Sanskrit Literature 335 

of trees, birds, and other objects of beauty seen in the Gandhamadana 
forest. The Hne in the Poona edition is given thus : 

sindhnvdrun atho 'ddumdii manmathasye 'va tomardn. 

The editor (V. S. Sukthankar) has a wavy hne under 'dddman "im- 
petuous, fiery (said of the darts)," indicating that the reading is "less 
than certain." Examination of the variant readings in his critical appa- 
ratus seems to show that, with less than his usual acumen, or more 
probably only by accident, the wrong reading has been adopted. It is 
attested only by the three manuscripts of the Grantha version that were 
available for this passage and by one of the two Telugu manuscripts. All 
other manuscripts read 'ddi'dn "excellent, fine." Since this means agree- 
ment between the all-important Sarada (or Kashmiri) version and the 
Malayalam version, an agreement which by itself is almost decisive, but 
which here is buttressed by agreement of the "K" version, the Bengali 
version, the Devanagar! version in its three groups, and one of the two 
manuscripts of the Telugu version, the reading 'ddrdn must be the correct 
one. The Grantha and Telugu versions are so closely connected that it is 
not surprising to find them, and them only, agreeing on a reading. It is 
almost unthinkable that here thej'- alone preserve a better reading than 
all other versions. If the editor intended any such implication, he has 
gone in the face of his critical principles as stated in The Mahdhhdrata 
l.lxxxvi-xcii and B.xviii, for as between the two words neither seems to 
be a lectio difficiUor (udddma is far le§s common than the banal uddra, but 
hardly a lectio difficiUor) or for any other reason preferable (note in vs. 59 
uddrdn as epithet of tilakdn) ; we can grant him only his meticulous use 
of the wa\^ line. Restoration to the passage of ^ ddrdn yields an effect of 
sound {°vdrdn . . . 'ddrdn . . . tomardn) at which the epic poet was certainly 
aiming; the context has numerous similar figures of sound.^ It might of 
course be argued that the reading 'ddrdn is a secondary change that was 
made by some early redactor or scribe to achieve just this effect. Only 
if it could be demonstrated that 'dddman is the superior reading intrin- 
sically, which I think cannot be done, would such an argument have 
weight against the distribution of the two readings in the manuscripts. 
Another probabty nondi!stinctive use is found in Asvaghosa's Buddha- 
carita 25.16, a passage of which the Sanskrit text is lost, but which can 
be given in E. H. Johnston's English rendering from the Tibetan version 
(Acta Orien talia 15 [1936], p. 246): "Who else is able by his word to 

^ E.g., vs. 52 madhurasvarair madhukarair virutdn kamaldkardn, the repetition of n4 
in vs. 53, the repetition of madhura- in vs. 54, in the same vs. 54 savildsdn ynaddlasdn 
. . . vanaldsakdn, vs. 55 "samkatesu katakesu, vs. 58 karnikdrdn . . . karnapuran, etc. 



336 University of California Puhlications in Classical Philology 

animate men in whom passion is born with mirth, Uke a cloud at the end 
of spring animating the dried up sinduvara plants?" 

A somewhat more distinctive use is found in the two following pas- 
sages. In Rajasekhara's play Karpuramaiijari iv.7/ in conjunction -svith 
some of the many refrigerants prescribed by Hindu literature for use by 
lovesick maidens or during the hot season there is found a necklace of 
sinduvara flowers (so I interpret, though without access to the native 
commentaries; Lanman, rather incredibty, thought of a necklace of 
berries) : sihinaparisare sinduvdrdna hdro "near the breast, a necklace of 
sinduvdra-^owers." The same use of the flowers as a refrigerant is prob- 
ably intended in Bana's Kadambari: avaddtasindhuvdraddmopahdram 
"garlanded with pure sindhuvara flowers."^ It is a pearl-white stone slab, 
which has been washed over with fresh sandal (a refrigerant) and gar- 
landed with the flowers, on which a lovesick j'outh is directed to rest. 

In a large group of passages, the flowers of the sinduvara, which are 
small and white and in long, loose clusters (panicles), are placed in a 
comparison with something else on the basis of their color, occasionally 
also with thought of their abundance in a stringlike cluster. Here the 
figures are based on distinguishing traits of the plant; no other plant 
could have been substituted. 

The earliest occurrence that I have been able to find is in Asvaghosa's 
Buddhacarita 4.49 : 

dirghikmn prdvrtam pasya llrajaih sinduvdrakaih 
pdndurdmsukasarhvltdm saydndrh pramaddm iva 

"See the pond enveloped bj- the sinduvara bushes growing on its banks, 
like a woman lying down and clothed in white silk."^ This can onh' be 
interpreted with reference to the flowers. Was it A^vaghosa who invented 
the comparison? Kavis certainly composed before Asvagho^^a, even if we 
do not possess their works, except for the two epics, and it would be very 
rash to assume that any poetical element that appears in the preserved 
literature first in his works was due to him. We shall see at once, however, 
that all the other passages involving this comparison make i^ quite clear 
that the flowers are the part of the plant involved; perhaps the slight 

* Sten Konow and C. R. Lanman, Kdja-^ekhara's Karpiira-maftjan (Harvard Orien- 
tal Series 4 [ 1901 ] ), text on p. 95, transl. on p. 277. There is a v.l. simdhiivdr". 

* Parab's Nirnayasfigara Press ed. of 1890, p. 390; Peterson's ed. (Bombay Sanskrit 
Series 24 [1889]), p. 200; transl. in C. M. Ridding, The Kadambari of Edna (Oriental 
Translation Fund, N.S. 7 [1896]), p. 159. Whether the commentaries in Parab's edi- 
tion make it clear, I do not know; I have no access to this edition. Peterson and 
Ridding have no comment. 

" E. H. Johnston, 7'/!ei?i/rf(f/mc«n7a (Panjab University Oriental PubHcations 31, 32 
[1935, 19;}G]), text in 31, p. 36, transl. in 32, p. 52. The reading sinthivdrakaih is 
Cowell's emendation for the ms. sinduvdrajaih, and is of the highest probability. 



Emeneau: The Sinduvara Tree in Sanskrit Literature 337 

immaturity in Asvaghosa's use of the comparison is indeed a sign that it 
was being used for the first time by him7 

Our next preserved poet, KaUdasa, uses the comparison, and in its fully 
matured form, in Kumarasambhava 3.53 : 

asokanirbhartsitapadmardgam 

dkrstahemadyutikarnikaram 
muktakald'pikrtasindh uvdram 

vasantapuspdbharanam vahantl 

"(the goddess Uma) wearing as ornaments the spring flowers; the asoka- 
flowers surpassed rubies, the kaniikdra-^owers drew to themselves the 
brightness of gold, the sindhuvdra-^owers were made her strings of 
pearls."^ Here we have a classical simplicity in the use of the comparison, 
which the later kavis, in their striving for a new and clever turn, fail to 
attain. 

In Bana's works the comparison has been found used three times, twice 
in the Kadambari, once in the Harsacarita. One of the Kadambari pas- 
sages is simple enough, but overartificial, for the author cannot be satis- 
fied with this comparison but must add two others of exactly the same 
type ; the phrase occurs, moreover, in one of those long passages in which, 
in the description of one person or thing, figiu-e follows figure until the 
subject is wrung dry, a technique dear to the Indian litterateur oi the de- 
cadence. Our passage describes the fair, white color of a girl in the 
words: kutajakundasindhuvdrakusumacchavihhir ivo 'Udsitdm "decked 
with the hues of kutaja, jasmine, and sinduvara flowers."^ 

The second Kadambari passage is much more complex in its interweav- 
ing of a number of comparisons : timirajaladharasamaydpaga?ndnantaram 
abhinavasitasiiidhuvdrakusumapdndurair arnavdgatair agdhyanta hariisair 
iva kumudasardmsi candrapddaih "the pools of night-lotuses were plunged 
into by the moonbeams, which were like hamsa birds falling on the ocean 
and white as sindhuvdra flowers fresh and pure after the close of the sea- 
son of black clouds (i.e. the rainy season)."^" 

The complexity attained by Bana in the Harsacarita passage is entirely 
typical of the cleverness of kdvya at its most clever : itarasravanena ca vi- 

'' I am assuming that the comparison is not found in the epics, an assumption based 
only on the fact that the only references to the epics found in our dictionaries are 
those discussed above; this may, of course, be a totally unjustified assumption. 

8 The mss. are recorded as varjang between "sindhuvdrarh and "sinduvdram. The 
commentator Narayana Pandita, whose commentary Vivarana was published in 
Trivandrum Sanskrit Series 32, notes that the sinduvdra flowers are compared to 
pearls bahugunatvdt "because of their abundance." 

9 Parab, p. 261; Peterson, p. 129; Ridding, p. 97. 

1° Parab, p. 100; Peterson, p. 48; Ridding, p. 46. To aid in an estimate of the passage 
I have given mj^ own rather literal translation. 



338 University of California Puhlications in Classical Philology 

kasitasitasindhuvdramanjarljusd hasate 'va prakatitavidydmadd "while the 
other ear [of Sarasvati, goddess of leariung], revelUng in a white full- 
blown Sindhuvdra flower [which she wore over it], betokened as with a 
gleaming smile [which is white by poetical convention] the intoxication 
of knowledge."" 

Another example is in a comic passage put in the mouth of the hungry 
vidUsaka in the Karptiramanjari i.l9 : 

phidlukkaram kalamakurasamarh vahanti 

je sinduvaravidava maha vallaha te 
je gdliassa mahisldahino saricchd 

te Mm ca mriddhnviaillo.pasunapunjd 

"The Sinduvara shrubs that bear a quantity of blossoms like to rice- 
pudding — my favorites are they; and also the multitudes of fair jasmine 
blooms, like to strained buffalo-milk. "^^ Rajasekhara had no scruples 
about repeating the figure in almost identical words, for in the Viddhasal- 
abhanjika i.261 he wrote: kalamakurakumaTjdapdndaresum sitasinduvdra- 
phullesum "though the flowers of the white sinduvara are pale as rice- 
pudding and white gourd-melons."^^ In both passages, the abundance of 
small blossoms is undoubtedly referred to, as well as the color. 

Jain literature has yielded two examples of straightforward use of this 
comparison. One is in a canonical passage given by L. D. Barnett in his 
translation of the Antagadadasao (Oriental Translation Fund, N.S. 17 
[1907]), p. 53: "pouring forth tears Uke a rope of gems, or showers of 
rain, or sinduvara-flowers, or a broken pearl-string."^* Hemacandra's 
Trisastisalakapurusacaritra 1.6.404 has in the description of the city 
Satrunjaya: 

sthdne sthane sinduvdraih irlkhandadravapdndubhih 
kftasarvdhgamangalyapunirdvalim ivo 'ccakaih 

"just as if it had a row of auspicious tilakas made on the body by the 
sinduA^ara-t rccs pale as sandal-paste, here and there. "^^ 

" Parab's Nirnayasagara Press ed. of 1897, p. 9; Fuhrer's ed. (Bombaj' Sanskrit 
Series 66 [ 1909 ] ), p. 14; transl. in E. B. Cowell and F. W. Thomas, The Har^a-carita of 
Bdna (Oriental Translation Fund, N.S. S [ 1897 ] ), p. 5. 1 have adapted this translation 
by additions in square brackets. Fuhrer's ed. omits "sita" and has "sinduvdra". 

^^ See note 4 above. 

" Bhaskar Rj\mchandra Arte, The Viddhasdlabhanjikd of Rdjasekhara (1886), p. 23; 
Louis H. Gray, "The Viddha^alabhanjika of Rajsi^ekhara, Now First Translated 
from the Sanskrit and Prakrit," in JAOS 27 (1906), p. 18. 

^* I have no access at present to a good collection of Jain texts, and it hardly seemed 
worth the etTort to trace down the passage with borrowed books. The date would in 
any case l)e very uncertain; the passage may be early. 

1* Johnson's translation, p. 354. The work was composed in the third quarter of the 
twelfth century. 



Emeneau: The Sinduvdra Tree in Sanskrit Literature 339 

The last passage that I have collected uses a different feature of the 
plant, the matted hair on its young shoots, the undersurface of its leaves, 
and the outside of the flowers,'^ in a comparison with the mane of a lion. 
This is in Ksemendra's Brhatkathamanjari, in the Vetalapancavirhsati 
section, story 11, vs. 5: 

kale kesarini smerasinduvdrasatodbhaie 

"the season being like a lion (lit. the maned one), for it was extraordina- 
rily full of the matted hair on opened sinduvdra flowers. "^^ The question 
arises, whether Ksemendra (11th century) invented this comparison. A 
priori reasons could be given both for and against attributing the inven- 
tion to him; I prefer to wait to see whether the comparison is to be found 
in other, earlier writers. 

The problem of whether the sinduvdra is to be identified with Vitex 
negundo and V. trifolia has to be raised. We have seen what distinctive 
traits of the plant are made elements of comparisons, namely, its white, 
abundant flowers in long panicles, and the matted hair on the outside of 
the flowers. All of this agrees with the descriptions in a large number of 
botanical handbooks of India,^^ except for the color of the flowers, which 
in most of the books is given as blue, purphsh, or bluish purple. The 

i« Sir J. D. Hooker, T}ie Flora of British India (1875-1897), 4, p. 58-3, has in the 
description of Vitex negundo and V. trifolia this phrase: "panicles closely white- 
tomentose"; similar terms are found in other books. 

" In my unpublished critical edition. The ms. and printed sources for the text show 
wide variants. What I read is attested completely by ms. G and the old printed text. 
Q, which goes with G but frequently rewrites, has: tamalamCdatlkekasmerinlsasatod- 
bhate. RST have kesarini. For "satodbhate, RST "iatodbhave, JP "madotkate, K (one of 
the printed ed.'s mss.) "madotkataih. On Ksemendra as a poet, see my article in JAOS 
53, pp. 124-143, and the literature cited there. 

1* One would go first to the standard modern book for the whole of India, Sir J. D. 
Hooker, The Flora of British India (1875-1897), and to the older but still useful book, 
William Roxburgh, Flora Indica;or, Description of Indian Plants (2d ed. 1832; the 
1st ed. was not completed and did not contain our species). Among the earliest bo- 
tanical works, containing also identifications with native terms, were Henrik Adriaan 
van Reede (Rheede) tot Drakestein, Hortus Indicus Malabariciis (1678-1703), and 
Sir WUliam Jones's posthumous article, "Botanical Observations on Select Indian 
Plants," in Asiatick Researches 4 (1795), pp. 237-312. Books in modern times on the 
flora of provinces or other restricted areas of India are: Theodore Cooke, The Flora of 
the Presidency of Bojubay (1901-rl904); J. F. Duthie, Flora of the Upper Gangetic Plain 
(1903-1929); J. S. Gamble, Flora of the Presidency of Madras (1905-1936); Sir Henry 
CoUett, Flora Simlensis (1921); H. H. Haines, The Botany of Bihar and Orissa (1921- 
1925). 

I also saw, by accident, Girija Prasanna Majumdar's Vanaspaii: Plants and Plant- 
Life as in Indian Treatises and Traditions (1927). He is a botanist, but not a San- 
skritist, and his book is noteworthy for the attempt, not argued by him and mistaken 
as I think, to identify nirgundl and sindhuvdra as Vitex trifolia, but nirgunthi (v.l. 
for nirgunti) as V. nirgundo (sic!); no one else has ever attempted to separate nir- 
gundl and nirgunti. He disarms criticism, however, by saying in his preface: "I regret 
that from want of time 1 have not been able to verify all the Latin sjTionyms of the 
plants mentioned in the work." He has no statement about the color of the flowers. 



340 University of California Puhlications in Classical Philology 

almost universal unanimity of the books on this point has led Johnson in 
her Trisastisaldkd'purusacariira, Vol. 1, p. 354, fn. 392, to deny the usual 
identification and to find that "some other identification is probably 
necessary." It is quite clear, however, that independently on several 
occasions native Sanskrit scholars of the old style have identified by the 
names nirgundl, sinduvdra, and so on, the plants which were then sys- 
tematized by botanists as Vitex negundo and V. trifolia; as, for example, 
for Van Reede's book and for Jones's article. The solution of the difficulty 
is to be found in Haines's The Botany of Bihar and Orissa, which is the 
last or almost the last published of the treatises referred to in the footnote 
above. He records for Vitex negundo that its flowers are "white or laven- 
der"; for V. trifolia he found no specimens in the area. Apparently the 
earliest botanists missed the white variety, or by error omitted to record 
it, and all later books down to Haines repeated earlier accounts. Hames's 
fresh observation solves the difficulty automatically, without impugning 
the accuracy of the early identification made with pandits' aid. 

We are now in a position to evaluate the nilasindhuvdra in Bana's 
Harsacarita : manorarhhasi goldhgulakapolakdlakdyalomni nllasindhuvdra- 
varne vdjini mahati samdrudham "he was mounted on a gi-eat steed, swift 
as thought, in colour like a dark Sindhuvdra flower, and with a coat as 
black as a goldngula ape's cheeks."^^ This is, of course, the bluish purple 
variety of flower which has been so often put into the botanical treatises, 
but only once, so far as I have discovered, into Sanskrit kdvya literature. 

The native lexicographers had already made the distinction between 
the two flower colors of the plant. For example, Narahari's Rajanighantu 

sinduvdrah svetapufpah sindukah sinduvdrakah 
nilapuspah siiasaho nirgundl nllasindukd 

"the sindhuvdra with white flowers (is called also) sinduka and sinduvd- 
raka, that with dark flowers siiasaha, nirgundl, and nllasindukd."-^ The 
Amarakoga 2.4.2.48-49 does not make the distinction in its entry: 

aiha sindukah 
sinduvdrendrasurasau nirgiin4lndrdnike 'ty api."^^ 

A more serious problem of identification arises through the fact that 



1' Parab, p. 23; Fixhrer, p. 37; Cowell and Thomas, p. IS. Fiihrer has "romni for 
"lomni and "sinduvdra" for "sindhuvdra". 

^ Anandfii^rama Sanskrit Series 33 (1896), p. 150. Different readings recorded are 
sindhukah and sindhuvdrakah. 

21 A. Loiseleur Deslongcham])s, Amarakocha ou vocahulairc d'Ainarasinha (1S39- 
1845). There are no significant differences in two other editions that 1 have been able 
to consult, that published by the Government Central Book Depot at Bombay and 
that published by the Nirnayasagara Press. 



Emcneau: The Siyidiwdra Tree in Sanskrit Literature 341 

some of the lexicographers make nirgundl also equal sephdlikd. Thus 
Rajanighantu 4.83 has : 

sephdlika 'mjd nirgun4i vanajd nilamanjarl 
sukla 'nyd svetasurasd bhutakesi ca kathyate 

"the one sephdlikd which grows in the forest and has dark clusters of 
blossoms is called nirgundl, the other, white, is called svetasurasd and 
hhHtakesl." To make it quite clear that nirguridl has both meanings, the 
Dhanvantarlyanighantu has as item no. 50 in the list of words with two 
meanings: nirgundl sephdlikd nllasindhuvdras cap Amarakosa 2.4.2.51 

sephdlikd tu suvahd nirgun4t nxlikd ca sd 
sitd 'sail Svetasurasd bhutavesy^^ 

This, like the Rajanighantu, gives two varieties for the sephdlikd, one 
(dark-fiowered) called also suvahd, nirgu7j,di, and nllikd, the other, white, 
called svetasurasd and hhutavesV^ The Western lexicographers of Sanskrit 
quote other native lexicons (which are not accessible for me), which ap- 
parently go farther in applying other terms than nirgundl to both the 
sinduvdra and the sephdlikd; for example, the major Petersburg lexicon 
mentions an identification of sephdli with nilasindhuvdra in the Bhava- 
prakasa as quoted in the Sabdakalpadruma in Gildemeister's Biblio- 
thecae sanskritae sive recensus libronim sanskritorum hucusque typis vel 
lapide exscriptorum critici specimen. The most serious result of this con- 
fusion is that the major Petersburg lexicon gives the botanical identifica- 
tion of sephdlikd as Vitex negundo, which is entirely mistaken. It is 
followed in this by the minor Petersburg lexicon, which records also an 
alternative identification as Nyctanthes arbor tristis,"^^ and by Monier- 
Williams, who gives Vitex negundo as the meaning and follows it with the 
remark: "accord, to others 'Nyctanthes Arbor Tristis.' " Richard 
Schmidt, in Nachirdge zum Sanskrit-Worterbuch in kurzerer Fassung von 
Otto Bohtlingk (1924-1928), has under the rubric sephdli (a synonym for 
sephdlikd) this entry: "sephdli ist in der Strope sakhi sa vijito vlndvdd- 
yaih (Dasar. p. 60 ed. Farab) sicher Nyctanthes arbor tristis, da nur 

22 Ananda^rama Sanskrit Series 33, p. 437. 

23 Colebrooke's note on the first line, as rendered by Deslongchamps, is: "Les noms 
sanserifs sont assignes dans le Bengale au nyctanthes arbor tristis. Quelques-uns appli- 
quent les deux derniers termes a une espece difJerente." The last sentence, of course, 
refers to an equation of nirgundl with sinduvdra; nllikd "the dark one" is actually 
applied to many other plants besides these two. 

2* Bhutavesl and bhfdakesi are obviously variant readings of the same word; which is 
correct will not be clear until we have really critical editions. 

25 The entry says that the identification was made in The Materia Medica of the 
Hindus Compiled from Sanskrit Medical Works, by Udoy Chand Dutt (1877); I have 
not seen the book. 



342 University of California Publications in Classical Philology 

dieser Baum zur Situation passt." He goes no further in attempting to 
clear up the confusion ; I shall discuss the verse below. 

For the sephdlikd, as for the sinduvdra, the native lexicographers speak 
of two varieties. The problem this time eludes solution at present. There 
seems no doubt that the white-flowered variety of sephdlikd is Nyctanthes 
arbor tristis, Linn., the night-flowering jasmine, an identification that was 
made by Sir William Jones with the aid of his pandits and that has not 
been departed from since in the treatises on botany. Editors and transla- 
tors of kdvya texts, so far as I have any collectanea for this plant, have all 
accepted the identification, with the exception of the translator of 
Rajasekhara's Viddhasalabhanjika, who accepts the old dictionary syn- 
onymy. What the dark-flowered variety is, we are still unable to say. It is 
of some interest to see how Jones dealt with the problem (p. 244) : "My 
Pandits unanimously assure me, that the plant before us [Nyctanthes arbor 
tristis] is their Seplidlicd . . . but Nilicd must imply a blue colour; and our 
travellers insist, that the Indians give the names of Pdrijdtica or Pdrijdta 
to this useful species of Nyctanthes'?^ on the other hand, I know that 
Pdrijdta is a name given to flowers of a genus totally different ; and there 
may be a variety of this with blueish corols ; for it is expressly declared, 
in the Amarcosh, that, 'when the Sep'hdlica has ivhite flowers, it is named 
Swetasurasd, and Bhutavesi.' " He hazarded a guess that there might be a 
variety with blue flowers; so far this seems not to have been found, and 
he was probably wrong. The solution of the problem must be undertaken 
by someone who combines qualifications as Sanskritist and as botanist, 
and has, besides, opportunity f-or observation. 

A tentative entry for the sephdlikd (sephdli, etc.) in the poetical en- 
cyclopedia is all that I can attempt, since my collection of examples is 
even farther from exhaustive than that for sinduvdra. However, two 
poetical figures depending on characteristics of the sephdlikd emerge verj'" 
clearly and perhaps these exhaust the range for this plant. 

The flowers are white with an orange corolla-tube. This orange tube is 
involved in three occurrences in a comparison with reddish or reddish- 
tawny objects. 

Probably the earliest occurrence is in two passages in Bhasa's Svapna- 
vasavadatta.^^ Both passages are in the same context at the beginning of 

28 This was the identification given by Van lleede. The pdrijdia is usually given as 
Erythrina indica, the coral tree. 

" If, that is to say, the attribution of the play to Bhasa is accepted. The editions 
and translations referred to are: the editio princeps by T. Ganapati Sastri, The 
Svapnavdsavadatin of Bhnsa (Trivandrum Sanskrit Series 15 [1912]), and his edition 
with a Sanskrit commentary, of which I have had access to the 1924 printing; Banarsi 
Das Jain and Madan Gopal Sliastri, Svnpnavasavadaltam of Bhasa (1920); Lakshman 
Sarup, The Vision of Vusavadattd (1925); V. S. Sukthankar, Vdsavadatta, Being a 



Emeneau: The Sinduvdra Tree in Sanskrit Literature 343 

the fourth act. The first is: hhattiddrie, tdni kusumiddni ndma, pavdlan- 
taridehirh via mottidlamhaehim diddni kuswmehim, which is rendered by 
most of the translators somewhat as Sukthankar has done: "They [the 
sephdlikd] have indeed blossomed, princess. They are laden with flowers 
that look like pendents of pearls interspersed [or, interset] with coral." 
Reference here is to the white and orange of the flowers. One translator, 
Baston, would make the comparison neater by translating the participle 
antarita as "monte" ; this, however, seems unwarranted. The second pas- 
sage is: pekkhadu pekkhadu bhaftiddrid addhamanasildvattaehim via se- 
hdlidkusumehi puriam me anjalirii. Three interpretations are in the field 
for the word in the comparison: addhamanasildvattaehim. Sukthankar's 
translation of the whole passage is: ''Behold, princess, behold. My joined 
hands are filled with sephalika blossoms that shine like crystals of 
arsenic." This, Baston's rendering: "comme des tablettes d'arsenic," and 
Jacobi's: "wie mit Zinnoberkiigelchen," are not quite exact, but seem to 
follow Ganapati Sastrl's Sanski'it chdyd, which renders the crucial word 
by ardhamanassildpattakaih, and his Sanskrit commentary, which inter- 
prets it as meaning "by those of which a portion is a plate (or, slab) of 
red arsenic. "^^ Woolner and Sarup give a different interpretation, taking 
the last element of the compound in the simile as "cloth-bandage" and 
rendering the whole word "with their half-way hose of realgar.'' Their 
note makes it clear that by "half-way hose" they intend "puttees"; real- 
gar is red arsenic. Both these interpretations rely on the color combina- 
tion seen in the flower. How the translators justify'' the details of the latter 
comparison, they do not show (Sarup in his earlier edition held the view 
that we presented first). The third interpretation, that of Banarsi Das 
Jain and Madan Gopal Sastri, departs far from the former two. They 
gloss the crucial word as ardhamanasildvartakaih and translate "as if with 
stone pieces each weighing half a maund." This seems most improbable; 

Translation of . . . SvapnavdHavadalta Attributed to Bhasa (1923); Albert Baston, 
Vdsavadattd . . . de Bhasa (Bibliotheque orientale elzevirienne 87 [1914]); A. C. 
Woolner and Lakshman Sarup, Thirteen Trivandrum Plays Attributed to Bhasa (Pan- 
jab University Oriental Publications 13 [1930-1931]); Hermann Jacobi, "Vasava- 
datta. Ein altindisches Schauspiel von Bhasa," in Internationale Monatsschrift fiir 
Wissenschoft, Kunst und Technik 7 (1912-1913), cols. 653-690. 

2* The editio princeps read addhd mana". T. Ganapati Sastri's 1924 edition dismissed 
this as erroneous (prc'unddika). Though we are still left in the dark about the readings 
of the mss., and the reading adopted later is not entirely perspicuous, yet the earher 
reading is impossible, and we must accept addhamana" on Sastri's authority as 
probably grounded on good manuscript evidence. 

Sastri's commentary glosses patta with dkdramsesa "a particular shape," which would 
correspond well enough with "slab, tablet" of our dictionaries or -nith Sukthankar's 
"crystals." Sastri's analysis of the compound is ardham ekadeso manassildpatto yesdm 
tair iti. Earlier in his commentary he interprets more freely: ardhdrunatvdn manassild- 
pattaghatitaikadesair ivety utpreksyate "because of the flowers being partly tawny-red, 
the simile is made: 'as if having portions formed of pattas of red arsenic' " 



344 University of California PuUications in Classical Philology 

mana "maimd" is not attested,^^ nor is there a variaka or vartakd with the 
required meaning, and the comparison seems pointless. In both these 
passages the comparison is in the form of a simple simile. 
Harsadeva used a similar simile in his play Priyadarsika ii.2 : 

vrntaih ksudrapravalasthagitam iva talarh bhati sephalikanam 

"the ground shuies with the stems of the sephalika flowers as if it were 
covered with fragments of coral. "^° And Bana's comparison in the 
Kadambari is also simple and contained in a fairly simple description of 
a parrot: sephalikdkusumanalapinjaram . . . cancwputena "with its beak 
tawny-red as the stalk of a sep/ia/iA-a-flower."^^ 

The fact that the sephdlikd blossoms open at night is involved in 
figures in three passages. Rajasekhara makes use of it twice, once in the 
Karpuramanjarl, once in the Viddhasalabhanjika. In both it is said that 
the moon is the cause of the flowers opening. The former passage (iv.l8^^) 
is : na hu maalanchanam antarena an^no miahkamaniputtaliam pajjhardvedi 
sehdlidkusumukkararii vd karedi "nobody but the moon — j^ou'd better 
believe— can make the moonstone statue to ooze or the gephalika to 
blossom profusely."^^ The other (ii.l9) is : 

na virid candarh sehalide viasanti husumdirh 
"The radiant lord of night alone evokes 
The white ^ephalika's soft loveliness."" 

The writer on the art of poetry, Visvanatha Kaviraja, in his Sahitya- 
darpana 215, illustrates one of his rules with a verse of his own composi- 
tion containing a figure based on the opening of the flowers at night : 

sephdlikam vidalitdm avalokija tanvi 

prandn katharh cid api dhdrayitum prabhiild 

dkarnya samprati rutam cararidijudhdndm 
kiih vd bhavisyati na vedmi lapasvini sd 

"With diffic ulty was that slender sufferer able to retain the vital spirit 

23 The editors were relying on "Panjabi addhman-dhaun 'twenty seers' "; this, and 
other vernacular phenomena, however, are hardly evidence for the existence of such a 
Sanskrit word. No reliance can be placed on the disputed iiira^ Xtyonevop mnna of Kig- 
veda 8.67.2; see now A. A. Macdonell and A. B. Keith, Vedic Index of Xamcs and 
Subjects, s. voc, for a summary of possible interpretations and tlie literature, and, 
most recently, Keith in Woolncr Commemoration Volume (Mehar Chand Lachhman 
Das Sanskrit and Prakrit Series 8 [1940]) , p. 146. 

2" G. K. Nariman, A. V. Williams Jackson, and Charles J. Ogden, PriyadaHikd (Co- 
lumbia University Indo-Iranian Series 10 [ 192'A ]), pp. 22-23. 

'' Parab, p. 52; Peterson, p. 2-5; Ridding, p. 23. My own translation; Ridding rather 
cheapens the comparison by using "j^ink" for "tawny-red.'' 

^2 Konow and Lanman, text on p. lOo, transl. on p. 282. 

" Arte, p. 65; Gray, p. 37. 



Emcneau: The Sinduvdra Tree in Sanskrit Literature 345 

when she beheld at midnight the expanded Nyctanthes, the harbinger of 
spring, but now when she hears the crowing of the cocks, awakening her 
from the visions of the night to the consciousness of desolation, — I know not 
what will become of her."^* 

With no hesitation, I would place here the phrase in Bana's Harsaca- 
rita: sephdlikdsltallkrtanise . . . saratsamaydramhhe "It was the beginning 
of autumn. . . . Then . . . the nights are cool with the qephalikd.'"^^ 

In Dhanika's commentary Avaloka on Dhanaiiijaya's Dasarupa 2.25a, 
he quotes the following verse by an unidentified author : 

sakhi sa vijito vmdvadyaih kaya 'py aparaslriya 

panitam abhavat idbhyarh tatra ksapdlalitam dhruvam . 

katham itarathd sephdllsu skhalatkusumdsv api 

prasaraii nabhomadhye 'pi 'ndaii priyena vilambyate 

"Ma chere, il s'est laisse prendre aux sons d'une vina, aux charmes d'une 
autre femme; I'une et I'autre m'ont voulu disputer le plaisir de cette 
nuit. Autrement, a I'heure ou les fleurs des gephalis se penchent, ou la 
lune monte au milieu du ciel, comment le bien-aime tarderait-il?"^^ Here, 
there is reference to the blossoms of the sephdlikd falling before morning; 
the moonlight is taken to be the cause of this, as it was above of the open- 
ing of the flowers. In two passages in Hemacandra's Trisastisalakapuru- 
sacaritra the same phenomenon is used in figures. The first is 1 .4.779 : 

yusmatpddair drstamdtrair anyajanmakrtdny api 
galanty endrhsi sephalipuspdin 'ndukarair iva 

"Merely by the sight of your feet, sins even though committed in another 
birth, fade away like sephali-flowers from the moon's rays."^^ The second 
passage is 3.6.43: 

nijarh phalam adattvd 'pi galisyaty asubharh mama 
sephdlikdpuspmn iva nisdkarakardhatam 



^* Roer's text in Bibliotheca Indica 9 (1851), p. 79; translation by Pramad^-dasa 
Mitra in Bibliotheca Indica 9 (1875), p. 114 (words in italics are his additions). 

36 Parab, pp. 83-84; Fiihrer, p. 129; Cowell and Thomas, pp. 70-71. Fuluer reads 
"nisi for "nise. The native commentator Sarhkara has on sephdlikd the note: rdtrdv eva 
vikasaii "it blossoms only at night"; his understanding of the passage is the same as 
mine. 

3^ So numbered in Parab's Nirnayasagara ed. of 1897, p. 60; 2.23b in Fitzedward 
Hall's Bibliotheca Indica ed. of 1865, p. 85. Parab has the misprint kaihd for kayd in 
the 1st pdda; both editions misspell sephdllsu in tlie 3d pada. The translation is by Syl- 
vain Levi, in Le Theatre indieii (Bibliotheque de I'Ecole des Hautes Etudes 83 [ 1S90 ] ), 
1.76. The commentary is assigned to the end of the tenth century; the date of the 
unidentified author of the verse is, of course, unknown. 

" Johnson's translation, p. 267. The translator had not yet found the explanation 
when tlus volume was published; she has it in a footnote on the later passage. 



346 University of California Publications in Classical Philology 

"Impurity will melt away from me without even giving its fruit, like the 
blossom of the sephalika struck by moonlight. "^^ This last passage is a 
guarantee that the former two, though they both use the word sephali, 
intend the sephdlikd (which was hardly in doubt), and further makes it 
unlikely that Hemacandra's sephdli is due to thought of sephdll in the 
Avaloka. 

I have omitted considering a few passages which do not mention the 
sephdlikd in a distinctive fashion, for example, Subandhu's Vasavadatta 
in Gra}^, text p. 187, translation p. 126, Fitzedward Hall's edition p. 
[265]. As in the passages mentioning the sinduvdra which we first dis- 
cussed, nothing is gained here by the emploj^ment of the sephdlikd that 
some other plant or tree could not have given equally well to the poetical 
values of the compositions.^^ 

3^ In Johnson's translation of books 2 and 3 in the Gaekwad's Oriental Series 77 
(1937), p. 317. She has "taking" for "giving/' 

^^ The mention of sephalika in Rtusamhara 3.14 (attributed to Kalidasa) probably 
belongs here, though lack of agreement among the mss. makes it uncertain. The pas- 
sage, in one reading, is: sephdlikdkusumagandhamaiiohardni . . . upavandni "pleasure- 
groves, charming Isecause of the perfume of the sephdlikd flowers." This would be 
nondistinctive. 'The alternative reading has "rdga" "red color" for "gandha" "perfume." 
In view of what we have discovered about the color of the flowers, it seems unlikely 
that "rdga" can be anything more than a corruption, mistakenly produced by a scribe 
trying to repair the loss of the aksara ndha. 



INDEX 

(Titles of contributions are in boldface type.) 

Achaeans, 7, 11, 16 

Achilles, 7, 11, 16 

adoro as exaggeration for "greatly admire," 66 

aeger, "invalid" in Seneca, 136 

Aenus settled by Lesbians, 294 

aeque ac (not atque) si, 68 

Aeschines, 35 

aetas in plural "periods of life," 238 

Aetna, theories on altitude of, 145 

Agamemnon, 7, 11, 12, 17, 19 passim 

Ajax, 12 

Alcaeus (fr. 77 Dielil = 29 Lobel), 294 

Alcinous, 13 

Alexander, play-acting of, over death of Clitus, 148 

Alexander, William Hardy, Seneca's Epistulae Morales : The Text Emended 

and Explained (I-LXV), 57-88; (LXVI-XCH), 135-164; (XCm-CXXIV), 

175-216; Seneca's Dialogi m, IV, V De Ira Libri Tres: The Text Emended 

and Explained, 225-254 
Alexandrian critics, 1-2, and scholars, 2 
Allegorical interpretation of Creation week, 322-324 
Allen, James Turney, Aristophanes and the Pnyx, 27-34; On the Program of 

the City Dionysia during the Peloponnesian War, 35-42 
Ambrose: his interpretations of Creation week, 323; taught allegorical method 

to Augustine, 324; demonstrated value of chronology, 325, 327 
amicus, ironically, 194 
Andromache, 8 

anguine, "hole in the wall," "hovel," 251-252 
Anoigmoi, festival at Miletus, 167, 168 
Antiochus Soter honored at Miletus, 173-174 
Apollo, 13, 16 

Apollo Didymaeus, 165-174 
Apollonius, omission of &s 4>aTo in, 56 

Apostolic Constitutions prescribe instruction in Bible historj^, 321 
Ares, 14; and Aphrodite, 265 
* Aristophanes, 37, 39, 40; Knights 303-313/ 28, 956/ 30, 754/ 31, 783/ 32; 

Wasps 31 ff./ 32; Acharnians 25/ 33; Ecclesiaztisae 87/ 33 
Aristophanes and the Pnyx, 27-34 
Aristotle, 264 
Arnold, Matthew, 263 
Astyanax, 8 
Athena, 7, 8, 14, 19 

Athene, 274-276 passim; and Odysseus, 273, 277 
Augustine on the Teaching of History, 315-332 
aid saltim rectis, etc., syntax discussed, 59-60 



Page number follows slanting line. 

[347] 



348 Index 

Bath at completion of task or migration, 293 

Bede: chronicle of, based on six ages, 330; invented new conception of seventh 

and eighth ages, 331 n. 94 
biberarius challenged as possible word, 83 
Bible basic textbook in history, 319 
Biographical Fashion in Literary Criticism, The, 279-292 
Birds, 37, 39, 40, 41 
Boegia, festival at Didyma, 168-173 
Boegos, official at Didyma, 168-173 
Bowra, C. M., 4, 5, 7, 10, 20 
Brachylogy, marked case of, 199 
Briseis, 7 

Calhoun, George M., Homeric Repetitions, 1-26 

Calypso, 9 

Canons: epic, 9; of taste, 23-24 

Caristia or Cara Cognatio, 91 

Carmody, Francis J., ed., Physiologus Latinus Versio Y, 95-134 

Catechetical instruction, history in, 320-322 

Cauer, 22 

Chaucer, 7, 268, 273 

Cherniss, Harold, The Biographical Fashion in Literary Criticism, 279-292 

Chesterton, G. K., 263 

Chronology: value of. recognized by Augustine, 325, 327; discrepancy in, 
before Abraham, 327, 330 

Chryseis, restoration of, 12, 16 

Chryses, 12, 16 

Circe, 20 

*Cicero: De Or. I, 38, 174/ 60, III, 11, 41/ 196; hi Verrem II, 4, 40/ 148 n. 21; 
De Off. I, 37, 133/ 196 

City of God: cites pagan historians, 319; expounds providence in world his- 
tory, 327; assumes understanding of scheme of six ages, 327; argument of, 
continued by Orosius, 328 

Clausula: rhythm invoked, 176, 185, 191, 195, 198, 203, 205, 208, 212, 213; 
discussed, 225, 226 (bis), 229, 230, 231, 233, 234, 236, 240, 249, 251, 254 

Cleon, 42 

Clergy, instruction of, 325, 329 

Cockerell, copy of Boegos inscription made by, 168-173 

Combellack, Frederick M., Omitted Speech Formulas in Homer, 43-56 

comitatus unusual in plural, 249 

conari with sense of motion, "push forward," 182 

contra CO ncitat us, "reciprocally stimulated" (?), 187 

Crispus, Lucius Vitellius, prophet at Didyma, 165, 166 

Critics: Alexandrian, 1-2; Homeric, 1, 24 

Croesus on the pyre, 221, 222 

Curriculum: unchanged from Quintilian to Augustine, 315; literature and 
rhetoric in, 316; Augustine's necessary subjects of, 319; of Cassiodorus, 
328 n. 82; of Isidore, 329; of Bede, 330; of Hugo of St. Victor, 331 



Page number follows slanting line. 



Index 349 

Danaides, 298 
Deiphobus, 7 

denique, sharp transition in series to leading element, 228 
despero. See spero 
destridus, "skinned," 181 
*Dial. Ill, IV, V; Latin authors quoted in discussion of: 

Apuleius, Met. II, 21-22/ 236 

Caesar, B.C. I, 64/ 246 

Festus (ed. Mueller), pp. 210 and 245/ 236 

Horace, Epod. V, 17-24/ 236 

Livy, XL, 40 and 44/ 246 

Pliny, N. H. XXXII, 1, 2/ 244 

Sallust,Caf.XLV, 1/249 

Valerius Maximus, IX, 7/ 226 
scholars quoted in discussion of: 

Axelson. See Clausula 

Baehrens, W., 241 

Barriera, 225-235 passim, 237, 238, 244, 246, 247 

Basore, 229, 233, 241, 247, 248, 249, 251, 252 

Bourgery, 225-229 passim, 232, 234, 237, 238, 241, 243, 246, 247, 253, 254 

Castiglioni, 229, 251 n. 39 

Charney, 229 n. 13, 245 n. 32, 246 n. 33, 247 n. 34 

Erasmus, 242 

Fickert, 253 

Gemoll, W., 225 

Gertz, 225, 233, 234, 237, 238, 242-247 passim, 249, 251, 253 

Gudeman, 250 

Haase, 227 

Haupt, 227, 253 

Hermes, E., 225, 228-230 passim, 232-235 passim, 242, 245, 250 

Karsten, 233, 234, 254 

Keil, 236 n. 19 

Klammer, 247 

Koch, 229, 244, 246 

Lipsius, 230, 239, 240, 251 

Madvig, 227, 232, 238, 241, 246, 247, 249, 250 

Matthias, 245 

Mueller, J., 228, 235, 237, 242, 250 

Petschenig, 234 

Pincianus, 230 

Schultess, 231,234 

Stephanus, 233 

Vahlen, 235, 244, 245 

Waltz, 247 

Wolters, 246 
*Dial. Ill, IV, V text emended: verbally: III, 1, 1/ 225; III, 1, 7/ 225; III, 2, 
2/ 226; III, 15, 1/ 229; III, 15, 3/ 229; IV, 7, 3/ 234; IV, 10, 1/ 235; IV, 
15, 1/ 238; IV, 32, 1/ 241; IV, 35, 1/ 243; V, 1, 2/ 244; V, 2, 6/ 246; V, 6, 

* Page number follows slanting line. 



850 Index 

2/ 246; V, 12, 6/ 248; V, 13, 1/ 249; V, 32, 2/ 251; V, 41, 3/ 253; defended 
with exegesis: III, 3, 2/ 226; III, 3, 4/ 226; III, 6, 2/ 227; III, 8, 4/ 227; 

III, 10, 1/ 228; III, 14, 3 (LP against A), 229; III, 16, 2/ 230; III, 16, 5/ 
230; III, 17, 7/ 231; III, 18, 3/ 231; III, 20, 3-4/ 232; III, 20, 8/ 232; 

IV, 1, 1 (bis), 232, 233; IV, 3, 5/ 233; IV, 9, 3 (bis), 234; IV, 11, 4/ 235; 
IV, 12, 1/ 237; IV, 12, 2/ 237; IV, 19, 1/ 238; IV, 19, 4/ 238; IV, 19, 5/ 239; 
IV, 20, 4/ 239; IV, 22, 1/ 240; IV, 23, 3/ 240; IV, 28, 4/ 240; IV, 32, 2/ 242; 
IV, 33, 6/ 242; IV, 36, 6/ 244; V, 2, 2/ 245; V, 2, 4/ 245; V, 8, 7-8/ 247; V, 
12, 2/ 247; V, 18, 4/ 249; V, 21, 2/ 249; V, 30, 2 (L against A), 250; V, 31, 
1/ 250; V, 33, 2/ 251; V, 37, 5/ 253; V, 38, 1/ 253; V, 42, 2/ 254 

Dialogue transition, regular, defined, 44 

Didascalic inscription found in Rome, 36, 38, 40 

Didyma: inscriptions of, 165-174; prophets at, 165-168, 171-173; treasurers 
at, 167-168, 173; festival of Boegia at, 168-173; Boegos, official at, 168- 
173; Zeus Soter at, 168-173; Poseidonius a prophet and Boegos at, 172-173 

Didymaean inscriptions. See Index locorum 

Didymeia festival, 173-174 

Dionysia, City: conflicting views respecting program of festival of, 35-40 
passim; changes in program of, and reasons for, 41-42; festival of, 173-174 

Dionysia at Miletus, 173-174 

domicilium putre, of the human body, 241 

Dravidian Etymology of the Sanskrit Proper Name Nala, A, 255-262 

Emeneau, M. B.: A Dravidian Etymology of the Sanskrit Proper Name Nala, 

255-262; The Sinduvara Tree in Sanskrit Literature, 333-346 
Ennius (fragm.), 197 

Epic (technique): formulas, 3, 6, 9; canons, 9; tradition, 18 
Epicurean doctrine in conclusions of earlier Senecan epistles, 68 
Epinicus, Epinicus' son: stephanephor at Miletus, 170; Boegos at Didj-ma, 

170-171 
Epistulae Morales, Seneca's, scholars quoted in discussion of: 
Albini, 209 
Albertini, 61 
Axelson, 136, 142, 146 n. 16, 149, 155, 179, 182, 183, 190, 193, 197, 204, 208, 

209,211,212 
Baehrens, 178, 185 
Barker, 140, 146, 152 n. 33, 159 n. 46, 162, 177, 181, 182, 193, 198, 206, 210, 

215 
Beltrami, 57, 60 and n. 7, 61, 64, 66, 67, 69, 71, 73-76 passim, 78, 80-86 
passim, 137, 139, 142, 143, 146, 148, 152, 155, 156, 158-161 passim, 163, 
175, 176, 180 (ter.), 181 (bis), 182, 184, 185, 188, 190 (bis), 193, 195, 198, 
200, 201, 204, 206-210 passim, 212-214 passim. 
Bonnet, M., 155, 182 
Brakman, 184 
Buecheler, 60, 76, 81, 137, 142, 149, 155, 159, 161, 176, 178, 179, 182, 186, 

190, 207, 212 
Busche, 188 
Cardo, 177 
Castiglioni, 79, 85 



Index 351 

Charney, B. L., 62 n. 13, 73 n. 24, 146 n. 17, 150 n. 29, 153 n. 34, 181 n. 17, 

184n. 27, 186n. 29, 196n.54 
Diels, 156 
Dunbabin, 146 

Erasmus, 139, 146, 161, 162, 175, 176, 184, 206 
Fickert, 58, 64, 67, 75, 142, 155, 160, 176, 194, 202, 212 
Foerster, 136 n. 3, 147 n. 20 
Forcellini, 157 

Georgii, H., 79, 144, 147 n. 19, 152, 154, 175, 191 n. 37 
Gertz, 66, 184, 195, 203 
Gronovius, 71, 152 
Gruter, 71, 74, 209 
Gummere, R. M., 85, 140, 177, 198 

Haase, 60, 75, 78, 80, 82, 84, 135, 137, 155, 160, 175, 202, 203, 213 
Hauck, P., 144, 159 
Haupt, 59, 81, 83 
Hense, 57, 60, 64, 66, 67, 73, 75, 78, 80-86 passim, 135, 137, 138, 141, 142, 

144, 146, 147, 149, 154, 155, 156, 158, 159, 161, 175, 180, 200 n. 61, 211, 

212, 215 
Hermes, Aem., 80, 181, 183 
Hess, G., 58, 64 
Hilgenfeld, 147 
Korsch, Th., 197 
Kronenberg, 64, 67, 80, 150, 161 
Lagrange, 140, 177 
Linde, S., 176 

Lipsius, 61, 159, 160, 188, 201 
Loefstedt, E., 195 n. 50 

Madvig, 60, 63, 64, 85, 137, 139, 146, 152, 180, 188, 194, 214 
Maehly, 155 

Morris, J. T. S., 154 n. 39 , 
Mueller, Joh., 80 

Muretus, 63, 68, 71, 79, 145, 154, 199, 203 
Niemeyer, 137, 150 
Opsopaeus, 59 

Pincianus, 61, 185, 194, 199, 212, 215 
Prechac, 61 

Rossbach, O., 61, 63, 66, 69, 73, 75, 78, 80, 83, 162, 175, 184, 185, 188, 192 
Ruhkopf, 68, 71, 160, 185, 186, 195 
Schultess, 85, 136 

Schweighaeuser, 58, 68, 80, 142, 150, 161, 202 
Summers, 72, 78, 82, 86, 141, 146, 149, 150, 152-155 passim, 158, 159, 182, 

209, 211 
Thomas, P., 69, 80, 157 
Usener, 64 
Vahlen, 207 

Vivona, 138, 156, 161, 192 
Von Jan, 146 n. 16, 190 
Waltz, L., 63 



352 Index 

Watzinger, C, 198 n. 58 

Windhaus, 159 n. 47, 176, 179, 180, 187, 189, 195, 196, 202, 205, 208, 209 
n.75 
*Epistulae Morales, text emended: verbally: (I-LXV) IX, 4/ 60; XIV, 13/ 61; 
XIV, 16/ 61; XVI, 2/ 62; XX, 2/ 64; XXI, 10/ 64; XXIV, 26/ 66; XXVI, 
3/ 67; XXXI, 11/ 70; XXXIII, 5/ 72; XLII, 4/ 74; XLVII, 8/ 78; XLVIII, 
7/ 79; XLIX, 4/ 80; XLIX, 5/ 80; LII, 5 (bis)/ 81; LIII, 12/ 83; LVI, 2/ 
83; LVII, 7/84; LXV, 15/ 86; (LXVI-XCII) LXVI, 16/ 135; LXVI, 21/ 
136; LXVII, 5/ 136; LXVIII, 1/ 137; LXVIII, 11/ 138; LXIX, 5/ 138; 
LXX, 5 (bis)/ 139; LXX, 21/141; LXX, 28/ 142; LXXIV, 9/ 142; LXXIV, 
33/ 143; LXXV, 7/ 143; LXXVIII, 11/ 145; LXXIX, 2/ 145; LXXX, 1/ 
146; LXXXI, 14/ 146; LXXXII, 2-3/ 147; LXXXIII. 19/ 148; LXXXIV. 
1/ 149; LXXXIV, 8/ 149; LXXXVI, 8/ 150; LXXXVI, 12/ 150; LXXXVI, 
14/ 151; LXXXVII, 1/ 153; LXXXVII, 3/ 154; LXXXVII, 41/ 154; 
LXXXVIII, 17/ 155; LXXXVIII, 44/ 156; LXXXVIII, 45/ 157; 
LXXXIX, 8/ 158; XC, 9/ 158; XCII, 2/ 160; XCII, 4/ 161; XCII, 6/ 
161; XCII, 34/ 162; (XCIII-CXXIV) XCIV, 3/ 175; XCIV, 21/ 176; 
XCIV, 59/ 178; XCV, 19/ 180; XCV, 26/ 181; XCV, 29/ 181; XCV, 46/ 
182; XCV, 57-58/ 183; XCVII, 10/ 184; XCIX, 10/ 184; XCIX, 20/ 185; 
CII, 17/ 189; CII, 28/ 190; CIV, 6/ 191; CIV, 20/ 192; CV, 3/ 193; CVII, 
1/ 194; CVIII, 32/ 196; CVIII, 33/ 197; CIX, 6/ 197; CIX, 17-18/ 198; 
CXIII, 20/ 200; CXIII, 30/ 200; CXV, 6/ 203; CXVI, 6/ 203; CXVII, 8/ 
204; CXVII, 23/ 204; CXVIII, 4/ 205; CXX, 12/ 206; CXX, 18/ 207; 
CXX, 20-21/ 207; CXXI, 20/ 208; CXXII, 1/ 209; CXXII, 8/ 210; 
CXXII, 15-16/ 211; CXXIII, 10/ 212; CXXIII, 12/ 212; by change of 
punctuation: (I-LXV) I, 5/ 57; XLV, 8/ 76; XLIX, 5/ 80; LXIII, 4/ 85; 
Hense's 1914 text defended : III, 3/ 57; XVIII, 11/ 63; XXVI, 8/ 68; XXIX, 
2/ 69; XXX, 8/ 69; XXXII, 3/ 71; XXXIII, 1/ 72; XXXVI, 5/ 73; ex- 
tended exegesis of difficult passages: VI, 7/ 58; VIII, 4/ 58-60; XIV, 16/ 
61-62; XXII, 7/ 65-66; XXII, 13/ 66; XXX, 8/ 69-70; XXXVII, 3/ 74; 
XLI, 7/ 74-75; XLV, 5/ 75-76; XLVI, 3/ 77; LIII, 9/ 82-83; LVIII, 32/ 
84-85; LXI, 1 (Schultess)/ 85; defended by exegesis of passage: XCIV, 8/ 
176; XCIV, 34/ 177; XCIV, 41/ 178; XCV, 1/ 179; XCV, 16/ 180; XCV. 
51/ 183; XCIX, 24/ 186; XCIX, 25/ 187; C, 8/ 187; CII, 12-13/ 188; CII, 
30/ 190; CV, 6/ 193; CVIII, 12/ 195; CVIII, 15/ 195; CXIII, 8/ 199; 
CXIII, 10/ 199; CXIII, 11/ 199; CXIV, 6/ 201; CXIV, 16/ 202; CXIV, 22/ 
202; CXVIII, 7/ 206; CXXII, 13/ 210; CXXIV, 5/ 214; CXXIV, 22/ 215 

Ethics, relation of practical and theoretical, 177-178 

Eumaeus, 9, 13 

Eurycleia, 13 

Eutychus, son of Eperastus, treasurer at Didyma, 167, 168 

Evegoros, law of, 38, 40 

excludere, play on two meanings, 201-202 

exigere, business sense, "collect" (?), 198 

expnmere, of pronunciation, 196 

Fabianus, ranked in eloquence, 188 
First Triumvirate, slaves under, 193 



Page number follows slanting line. 



Index 353 

Fontenrose, Joseph E., Notes on Some Didymaean Inscriptions, 165-174; 

Varia Critica, 217-224 
Formulas in Homer, 6, 8-13, 18, 25; epic, 6, 9, 25; traditional, 9, 22; name 

epithet, 9; auxiliary, 10; recurring, 10, 12, 15, 18, 21, 22; prefatory, 11; 

expanded, 12; distribution of, 15; accumulation of, 16, 18; development of, 

18; repeated, 18; groups of, 18, 22; variation of, 19 
Frank, Erich, 296, 301, 305 

Frankel, Hermann, The Immigrant's Bath, 293-294 
Frazer, Sir J. G., 89, 90 
funerosus, a "ghost" word; read onerosos, 243-244 

Ghosts, fear of, 91 and n. 5 
*Greek authors quoted in discussion of Seneca, Dial. Ill, 2, 2, Appian, Civil 
Wars I, 6, 54/ 226 
Greek classic literature, foundation of, 23. See also Palaeography 
Green, William M., Augustine on the Teaching of History, 315-332 

Hades, uninitiated in, in Polygnotus painting, 297 

Hairdressing among barbarian tribes, 215 

Hart, Walter Morris, High Comedy in the Odyssey, 263-278 

Haussoullier, Bernard, on Le Bas-Waddington 222, 168-173 

Hebe, 14 

Hector, 7, 8, 12, 14 

Helen, 267-272 passim 

Helios episode, 49-52 

Helius, 12 

Hera, 7, 14, 19 

High Comedy in the Odyssey, 263-278 

Hirzel, Rudolf, 301 n. 13 

History: Augustine on teaching of, 315-332; in Roman education, 316; defini- 
tion of, 316, 326; Augustine's acquaintance with, 317-318; truth and false- 
hood in, 318; Bible a textbook of, 319; in catechetical instruction, 320; in 
instruction of clergy, 325. See also Six ages of world history 

Hitler, 278 

Homer, critics of, 1, 24; method of study of, 2; criticism of, 2, 21, 22; art of, 
6, 21; repeated line used in, 68; textual criticism of, 22; "dispensable" line 
in, 22-23; discrepancies between long and short texts of, 24; originality in, 
25. See also: Iliad; Time sequence 

Homeric Repetitions, 1-26 
*Horace: Epod. 10, 5/ 157; Odes 3, 5, 7/ 157; Sat. U, 2, 78/ 183 

Hugo of St. Victor, history in curriculum of, 331 

Hutchison, C. B., quoted on technique of sowing broadcast, 69 

iacere in angulo, "to live obscurely," "to live meanly," 252 
ibi in sense of "in eo, in eis rebus," closely limited in use, 60 
ignominia, a concrete form of punishment, 230 
Iliad, 1, 11, 12, 15, 23, 25, 39-47, 293 n. 2 
Immigrant's Bath, The, 293-294 



Page number follows slanting line. 



354 Index 

Immortality personal under Roman republic, 245 
impotentia, sc. sui, "anger," 245 
*Index locorum: for Didymaean Inscriptions : 

Ancient Gr. Inscrip. in Brit. Mus., 922/ 165-166; 923a/ 167-168 
Ditt. O.G.I.S., 213/ 173-174 
Le Bas-Waddington 222/ 168-173 
forVariaCritica: 

Constantine Porphyrogenitus, Excerpta de virtutihus et vitiis e Nic. Dam. 

29, p. 347 BW, 221-222; p. 345 BW, 222 
Diodorus Siculus, 4.70.4, 222-224 
Haussoullier, Etudes ... , 78, lines 45-47, 219 
I.G. IX 2.205, lines 3-8, 218, 219; XII 1.736, lines 2-9, 218, 219 
Inschriften von Pergamon, 245 C, lines 20-25, 219-220 
Nicolaus of Damascus, 68, sec. 8, 221 
Scholiast on: Nicolaus of Damascus, 68, sec. 1, 222; Pindar, Pyth. 2.85, 

224 
S.G.D.I. 5597, Unes 7 f., 218, 219 

S.I.G.' 685, lines 63-67, 218-219; 826 E, III 7, line 24, 219 
Theocritus, Idylls 1.13 and 5.101, 217-221; 5.103, 220 
Thomas Magister, schoha on Euripides' Orestes 165, 222 
in illo synopticaUy used, 140 
Index verborum: 

aire, 217 and n. 2, 221 
airep, 221 
's.Te, 220, 221 
are (are), 217 
'AvoLynols, 167 

aVTOtTT]S, 171 

Povyi<^, 169, 171 

Por]yla vi.Kr)cras, 169 

/3o7j76s, 169-173 passim 

eU, 220 

h, 217, 220, 221 

evnlvoi, IIX, 2/12, 

tvTovov, 221, 222 n. 12 

evTOvos, 221 

tv Tovif, 221, 222 n. 12 

2a)s, 220 

pre (^T«), 220 

KaKt[las\, 168 

Kkravrts, 217 

Kp£cr<pos, 166 and n. 3 

kvkXlol Ayuives, 173 

Kvpelva, 166 

6]iJiov, 173 

irapedpevuv, 173 

■jrpot5plav, "173 

■Kpo<fr)Tris, 165, 171 

* Page number follows slanting line. 



Index 355 

'Viyelvov, 165 

rajutas, 173 

wj, 217 and n. 2, 218-221 passim 
Irenaeus, On the Proof of the Apostolic Preaching, 321 
Iris, 7, 19 
Isidore of Seville, chronicle of, based on six ages, 329 

Lafica, 90, 93 n. 10 
Laertes, 9 
Lang, A., 2-3 

Languor and languor insignis perhaps technical terms in medicine, 143 
Lapith-Centaur war, 222-224 and n. 15 
Lares not ghosts, 92 n. 8 
Latin authors. See Seneca 

Le Bas, Philippe, copy of Boegos inscription by, 168-173 
lemures, 90, 93 
Lemuria, 89, 91, 92 

Lenaea, number of contestants in comedy at, 36, 42 

Liberal arts: in Roman education, 316; in Augustine's training of clergy, 326 
Linforth, Ivan M., Soul and Sieve in Plato's Gorgias, 295-314 
Lipsius, 38, 39, 40 
*Livy, 42, 43, 9/ 64; 42, 44, 6/ 64 
Loeb Library translation of Seneca's letter (Gummere), 73, 78, 79, 85 

Magic rites, objects employed in, 236 

magmis vir, modifies noun to denote Sage, 162 

Manes exite patemi, 89-94 

Manichees: profess devotion to truth, 318; believed Old Testament absurd, 

324; Augustine in controversy with, 324 
Manuscript errors due to fatigue, 82 
*Martial, Epigr. 4, 49, 9/ 66 
Menelaus, 11 
Merchant of Venice, 277 
Meredith, George, 265, 267, 278 
Miletus: office of stephanephors at, 166, 170-172; Anoigmoi at, 167, 168; 

Antiochus Soter honored at, 173-174; festivals at, 173-174 
morsiis, "portions of flesh" used in magic, 236 
Much Ado about Nothing, 277 
Mueller, J., 222 and n. 12 

Name-epithet formulas, 9 

Nausicaa, 8, 271; and Odysseus, 272 

nego, "refuse" with dative of person and infinitive of the action, 63 

Nestle, W., 295 n. 3 

Nestor, 9, 12, 266, 267 

Nonnus, omission of &s 4>aTo in, 56 

noto in sense "brand," "type," 193 

Notes on Some Didymaean Inscriptions, 165-174 



Page number follows slanting line. 



356 Index 

Obscenity and holiness, 93 n. 10 

obversatio, "opposition" (astronomical), 182 

Odysseus, 9, 12, 13, 14, 20, 21, 265, 271, 274-276 passim; and Iris, 265; and 

Nausicaa, 270-272 passim; and Athene, 273, 277 
Odyssey, 1, 8, 9, 23, 25 
Olive growing, 151-153 
Olympian malice, 7 
Olympus, 7, 15, 17 

Omitted Speech Fonnulas in Homer, 43-56 

On the Program of the City Dionysia during the Peloponnesian War, 35-42 
Orosius accepts Augustine's view of world history, 328 
w$ 0dTo formula: regular, defined, 44; omitted in: regular dialogue transition, 

44; unusual time sequence, 46-52 passim; desire for contrast, 52-54; 

clumsy situation, 54-55; later Greek epics, 55-56; Nonnus, 56; ApoUonius, 

56; Quintus, 56 
Ovid: Fasti, V, 421 sqq., description of Lemuria, 89, 91 n. 4; Her. 2, 89, 

distich, 293 

Palaeography, Greek, 221 

Parentalia, 89, 90, 91, 92 

Paris, 11 

Parmenides, theory of the One, 157 

Parry, MUman, 1, 3 

pars in singular to express "political party," 234 

Patroclus, 11, 14 

Perithous" wedding feast, 224 

Philolaus, 296, 304 

Physiologus: history of, studies, 95 ff.; Latin, Y, 98, 101; critical method of 
editing Y, 98-100 

Physiologus Latinus Versio Y, -95-134 

Pindar,OLlI, 94ff., 29 
*Plato: Theaet. 180 e 3/ 157; Gorgias, 295 S.; psychology of, 298, 308; distinc- 
tion between wiffris and iwiffTia in, 303; Rep. X 621 a, 303; Phaedrus 229/ 
311 
*Pliny, N.H. XXVIII, 27/ 92 n. 8 
*Phny the Younger, 10, 31, 4/ 64 

Pnyx, Aristoi)hanes and the, 27-34; designation of the bema, 30; seating 
arrangement of, 30 

poenamferre, is it Latin?, 229 

Polygnotus at Delphi, painting of, 297 

Polyphemus, 264 

Portable kitchens in Rome, 191 

Poseidonius, prophet and Boegos at Didyma, 172-173 

positus with sense "dependent on," "contingent on," 71 

primurn . . . deinde . . . deinde in a series, 137 

promitlere, "guarantee delivery," 179 

Pronunciation, corruption of syllables, 196 

Prosper of Aquitaine influenced by Augustine, 328 



Page number follows slanting line. 



Index 357 

prudenter cenare, "to dine with restraint," 200 
Ptolemaic papjTi, 23, 24 
Punishment, six forms of Roman, 230 
Pythagorean doctrines, 296, 299, 306 

*Quintilian, Inst. Orat: XI, 3, 118/ 69; 1, 11, 4/ 196 
*Quintus Curtius: omission of ws 0dro in, 56; VIII, 2, 4-5/ 148 

Reginus, Lucius (MaHus), prophet at Didyma, 165, 166 
*Rehm, A., on Le Bas-Waddington 222/ 168-171 

Repetitions: Homeric, 1, 21; modern criticism adverse to, 2; as traditional 
material, 4, 8, 22; to point contrasts, 5; endless variety of, 12, 18; of specific 
passages, 18; of messages, 21; dislike of in visual analysis, 21; fuller under- 
standing of, 24-25 

Rhetorical "fillers" in Seneca, 246 n. 33, 254 

Romulean year, 89 

Rose, H. J., Manes exite patemi, 89-94 

Rothe, C.,2,3 

Sabbath, millennial, 322, 323, 330 

Saturninus, Lucius Malius, prophet at Didyma, 165, 166 

Schol. Theocr. VII, 112, emended, 294 n. 5 

Scott, J. A., 3 
*Seneca, quoted from works: other than Dial. Ill, IV, V: Dial. X, 7, 3/ 228 
X, 11, 1/ 228; XI, 6, 4/ 252; Medea w. 249-250/ 252; Thyestes, v. 74/ 230 
other than Epistulae Morales: Benef. 3, 9, 1/ 193 n. 42; 6, 33, 4/ 154 
Clem. 1, 22, 2/ 80; 1, 24, 1/ 182; de Rem. Fort. 5, 4/ 162; Dial. 2, 15, 1/ 78 
4, 19, 1/ 205; 4, 19, 5/ 213; 4, 31, 7/ 160; 5, 26, 3/ 215; 6, 2, 1/ 139; 6, 10 
6/ 83; 8, 1, 1/ 146; 9, 1, 678/ 154; 9, 5, 3/ 154; 9, 6, 1/ 137; 9, 7, 5/ 138 
10, 13, 1/ 163; 11, 18, 6/ 138; N.Q. 1, 14, 4/ 160; 2, 2, 4/ 193 n. 42; 4, 13, 
11/ 163; 4 b, 13, 10/ 136. See also Dial. (Ill, IV, V); Epistulae Morales 

Seneca's Dialog! HI, IV, V De Ira Libri Tres : The Text Emended and Ex- 
plained, 225-254 

Seneca's Epistulae Morales: The Text Emended and Explained (I-LXV), 
57-88; (LXVI-XCn), 135-164; (XCHI-CXXIV), 175-216 

Shakespeare, 263, 266, 277, 278 

Shewan, A., 3 
*Silius Italicus: I, 107/ 63; I, 519/ 63 

Sinduvara Tree in Sanskrit Literature, The, 333-346 

Six ages of world history: in Augustine's catechetical instruction, 321; in 
Augustine's sermons, 322; doctrine of, invented by Augustine, 324; in De 
Doctrina Christiana, 325; in Citij of God, 326 f.; in Isidore of Seville, 329; 
in Bede, 330 f . ; in Hugo of St. Victor, 331 

Six thousand years' duration of world, idea of, 322; rejected by Augustine, 
323; popular in time of Bede, 330 

Slaves, runaway, 194, 201 
*Sophocles, Philoctetes 783 f./ 30 

Soul, tripartite, idea of, 298-300 

* Page number follows slanting fine. 



358 Index 

Soul and Sieve in Plato's Gorgias, 295-314 

spaj-gere: in sense of "spread"' as applied to opening of hand, 69; "to dress the 
hair in several braids," 215 

Sparta, truce with, ratified, 39 
*Speech formulas in Homer: types of, 43-44; omitted, 43-56; index of passages 
discussed: A 22/ 53, B 188-207/ 49, A 232-250/ 49, Z 476 ff./ 54, H 300 ff./ 
54, M 317 ff./ 54, P 414-423/ 49, X 498/ 55, ^ 575 ff./ 54, a 360/ 53, /3 323- 
337/ 48, M 374r-390/ 49, tt 288 ff./ 55, w 336 ff./ 54, r 5-13/ 55, r 413/ 
47, V 238/ 48, 203/ 48, 354/ 53, <t> 396-404/ 49, x 200/ 53 

spero and despero, "expect" and "not to expect," 156 

statim, a predicate adjective?, 208 

Stephanephors, office of, 166, 170-172 

stilicidium, probably in sense "dropper," 141 
*Suetonius, A^ero 12, 3/ 66 

Suicide as escape from congenital ills, in Seneca, 241 

Tacita, Roman cult of, 91 n. 7 
*Tacitus, Germania 37/ 215 n. 85 

Teaching of history, Augustine on, 315-332 

Telemachus, 9, 12, 267-270 passim, 272 

"Temesaean," ornamental epithet, 90 n. 2 
*Terence, Adelphi 5, 3, 63/163 

Theoclymenus, 13 

Thessalian witches, watchers against, 236 

Thetis, 17 

Thucydides, testimony of, respecting festival, 39, 40, 42 

Time sequence in Homer: usual, after speeches, 44-46; omission of speech 
formula after unusual, 46-52 

Triumvirate, first, over slave state, 193 

ultio only legalized talio, 241-242 
universum in sense of totus, 160 

Varia Critica, 217-224 
Varro, treatise on nine disciplines, 316, 326 
♦Vergil: Aeneid II, 494/ 74; Georgics I, 126/ 59 
Verrius Flaccus, 89, 91 

Water carriers in Hades, fable of, 295, 297-298 
♦Wilhelm, Adolf, on Le Bas-Waddington 222/ 169 and n. 10 
Witches. See Thessalian witches 
Woodhouse, W. J., 263 n. 2 

Xanthus, 12 

Zeus, 7, 12, 19 

Zeus Soterat Didyma, 169-173 



Page number follows slanting line. 



UNIVERSITY OF CALIFORNIA PUBLICATIONS 

IN 

CLASSICAL PHILOLOGY 



VOLUME XIII 

1944-1950 



EDITORS 
J, T. ALLEN W. M. GREEN 

W. H. ALEXANDER L M. LINFORTH 

L. EDELSTEIN L. A. MAC KAY 

H. R. W. SMITH 



UNIVERSITY OF CALIFORNIA PRESS 

BERKELEY AND LOS ANGELES 

1950 



University of California Press 

Berkeley and Los Angeles 

California 

Cambridge University Press 
London, England 



PRINTED IN THE UNITED STATES OF AMERICA 



W. H. ALEXANDER 
I. M. LINFORTH 



LITTERARUM ET ELOQUENTIAE LUMINIBUS 

HUMANITATIS AUCTORIBUS 

HUNC LIBRUM 

DEDICAMUS 



quae tabulis nostris toties tarn grata fuistis 
nunc et principio, nomina, ferte decus 



■^ 



CONTENTS 

NO. PAGE 

1. Quotations in the Latin Physiologus from Latin Bibles Earlier than the 
Vulgate 1 

Francis J. Carmody 

2. Seneca's Dialogues VI, XI, XII (Consolationes) : The Text Emended 

and Explained 9 

WiLUAM Hardy Alexander 

3. Seneca's Dialogues I, II, VII, VIII, IX, X (Miscellaneous Moral Essays) : 

The Text Emended and Explained 49 

William Hardy Alexander 

4. Philemon, Lot, and Lycaon 93 

Joseph Fontbnrose 

5. The Corybantie Eites in Plato 121 

Ivan M. Linforth 

6. Telestic Madness in Plato, Phaedrus 244de 163 

Ivan M. Linforth 

7. Horace's Odes and Carmen Saeculare: Observations and Interpretations .... 173 

William Hardy Alexander 

8. Seneca's Naturales Qiiaestiones: The Text Emended and Explained 241 

William Hardy Alexander 

9. Propertius IV.7: Prolegomena to an Interpretation 333 

William C. Helmbold 

10. The Strangling Figs in Sanskrit Literature 345 

M. B. Emeneau , 

11. Propertius and the Roman Career 371 

Joseph Fontenrose 

12. Nugae Propertianae. I 389 

W. C. Helmbold 

13. Initium omnis peccati superbia: Augustine on Pride as the First Sin 407 

William M. Green 

14. Notes on Lucretius 433 

L. A. MacKay 



QUOTATIONS IN THE LATIN PHYSIOLOGUS FROM 
LATIN BIBLES EARLIER THAN THE VULGATE 



BY 

FRANCIS J. CARMODY 



Pre- Vulgate and pre-Augustinian literary Church Latin was a flexible 
language, based in large part on classical Latin, but strongly influenced 
in its phonetics by spoken or Vulgar Latin; at every turn it showed the 
effects of Greek syntax; it was, however, in its vocabulary that it de- 
parted most from Roman literary style, since it was faced with a pressing 
need for new words to express many new ideas. ^ Jerome, in pubhshing the 
Vulgate, between 384 and 405, and Augustine, by a tempered enthusiasm 
for both Cicero and the new linguistic medium, arrested any further vital 
evolution in Ecclesiastical Latin. The language of pre- Vulgate Bibles is 
that of the early Christian Church, from the time of Cyprian and Tertul- 
]ian to that of Ambrose and Augustine f study of its official cultural docu- 
ment, the Old Latin Bible, is essential to an understanding of the rise of 
that internationalized variety of Latin which was destined to remain the 
standard means of intercourse throughout Western Europe for well over 
twelve centuries. 

Lack of unity among Old Latin Bible manuscripts is in no sense a proof 
that they represent several independent translations; it illustrates rather 
the impossibility of fully recapturing the oldest forms of what was gradu- 
ally avoided and rejected by churchmen after the appearance of the 
Vulgate.^ The existence of one or more standard pre- Vulgate Latin Bibles 
alone can explain the frequent identity of wording among Bible quota- 
tions in Patristic writings. In these works, as well as in Bible manuscripts, 
disagreement between readings may depend on several factors ; in the 
first place there were variants known principally in certain regions, nu- 
merous enough to permit us to classifj^ various manuscripts as belonging 

1 On all linguistic matters see W. Plater and H. White, A Grammar of the Vulgate 
(Oxford, 1926) ; for Greek see J. Moulton, A Grammar of New Testament Greek, 2 vols. 
(Edinburgh, 1906-1920), and H. Thackeray, A Grammar of the Old Testament in 
Greek According to the Septnagint (Cambridge, 1909). 

- Concerning Augustine see C. Milne, A Reconstruction of the Old-Latin Text or 
Texts of the Gospels Used by St. Augustine (Cambridge, 1926), and F. Burkitt, The 
Old Latin and the Itala (Cambridge, 1896). 

^ For a searching account of the details see S. Berger, Histoire de la Vulgate pen- 
dant les premiers siecles du moyen age (Paris, 1893). The most recent list of Vetus 
Latina manuscripts appears in A. Jiilicher, Itala, das neue Testament in Altlatei- 
nischer Ubei'lieferung, Vol. 2 (Berlin, 19-10), back cover. 

[1] 



2 University of California Publications in Classical Philology 

to versions called Afra, Itala, European, or Irish;* in the second place 
many readings are undoubtedly due to faulty transcription, to faulty 
memory, or even to additions from non-Latin sources; finally, all known 
copies have been corrupted, contaminated, annotated, and interpolated. 
The immensity of the problem of determining the exact text of the Vetus 
Latina, apparent in the mere length of the Bible itself, has kept scholars 
from solving it in the obvious way, by critical editions.^ Nor can we as yet 
use Patristic literature to its full extent : the fact that, far from the great 
libraries, I could find two fourth-century Latin texts, pertinent to the 
subject but never before edited in anything like critical form,^ may sug- 
gest that many other works will some day be edited which will supply 
vital collateral material. 

The unity of the pre-Vulgate Latin Gospels, or even of the Xew Testa- 
ment as a whole, can no longer be questioned; for the Old Testament, 
however, there is no decisive manuscript testimony. It would be natural, 
on the one hand, for authors and translators to recall Gospel quotations 
from memory, while, when confronted with the immense and less familiar 
Old Testament, they would prefer to translate from their sources or on 
occasion would render from memory in faulty fashion. Furthermore, each 
book of the Old Testament may well have existed in one or more separate 
translations.^ In the Physiologus the translators seem often to follow 
their Greek original, as for example where we find [Hiob 9.9 and cf . 37.9, 
38.32, Phj^s. B 27.7]: Qui facit uirgiUam et septcntrionalem et dextnim et 
promptuaria austri, which stands in marked contrast to the Vulgate: Qui 
facit arcturum et oriona et hyadas et interiora austri.^ In a similar example 
[Prov 6.6-8], where B n.1-2 gives: Uade ad formicam, a piger, meditare 
earn; quae cum sit uirihus infirmior, multum per aestatem frumentum sihi 
reponit, and Y 14.1 gives: Uade ad formicam, a piger, et imitare uias eius, 
the Vulgate reads: Uade ad formicam, o piger, et considera vias eius, et 
disce sapientiam; quae cum non habeat ducem nee praeccptorem nee prin- 
cipem, parat in aestate cihum sihi et congregat in messe quod comedat. Old 
Testament readings are all the more difficult to treat because scholars 
usually avoid speaking of them in connection with the Vetus Latina;^ yet 

■• On these see Berger, op. cit.; on the Irish traits see J. Gwynn, Liber Ardmachanus 
(Dublin, 1913), pp. cxxxv-cclviii. 

^ To date, to my knowledge, there is, in critical editions of the Vetus Latina, 
only A. Julicher's Kala (Berlin, 1938-1940), for Mt and Mc alone. 

* Physiologus Latinus vcrsio B (Paris, 1939), and Phi/siologus Latinus vcrsio Y, 
Univ. of Calif. Piibl. in Class. Phil, Vol. 12, No. 7 (Berkeley, 1941), pp. 95-134. 

^ The three Latin versions of Hiob are mentioned by Burkitt, op. cit., pp. 32-34. 
8 Compare the readings in Philastrius, edit, infra. 

* Berger, op: cit., pp. 65-68, speaks constantly of the "ancienne version" of the 
Old Testament as of a known and unified text; Burkitt, op. cit., treats several 
matters concerning Dan and Hiob but devotes most of his efforts to the Gospels. 



Carmody: Quotations in the Latin Physiologus 3 

it should be obvious that a carefully compiled synthetic edition of varia- 
tions of this nature would be worth a dozen speculative monographs. 

What such a synthetic edition would yield is amply illustrated by the 
Physiologus quotation from Hier 17.11. At first sight, the quotations in 
B 25.3 and F 31.1 might be taken for what they obviously are, fair trans- 
lations of the Greek Physiologus, which in turn gives a good text of the 
Septuagint; however, for this passage, we have a fragment of an Old 
Latin Bible, ms St.-Gall 912, 9 c, which in turn is corroborated by several 
Patristic writers. Burkitt realized the importance of this fragment; he 
used the Hexameron of St. Ambrose as confirmation of its readings; 
Physiologus B is at this point the source of the Hexameron, and by that 
fact improves on the latter and eliminates any need to quote it in our 
collation. I shall give the Vulgate text {S), then that of B and Y, then 
reproduce the variants given by Burkitt from Augustine {Aug) and 
Philastrius {Phil),^^ lettering the St.-Gall fragment temporarily as G: 

S Perdix fouit quae non peperit: fecit diuitias, et non in iudicio: in dimidio 
dierum suorum derelinquet eas, et in nouissimo suo erit insipiens. 

B Clamauit perdix et congregauit quae non peperit, faciens sibi diuitias non cum 
iudicio; in dimidio autem dierum eius relinquent eum, et in nouissimis suis 
erit stultus. 

Y Clamauit perdix, colligens que non peperit. 

clamauit perdix] BGY Phil; congregauit (awriyaye, -ev) ] EG colligens Y colligit 
Y^ collegit Phil fouit S; et] B; quae (a)] BY Phil Aug quos G; non peperit] 
BGSY non parturiit Phil '{end of Y and Hexameron) add. multos pariat sibi 
filios G; faciens {iroiibv)] B Aug Phil fecit S adquirens G; sibi diuitias] B Phil 
diuitias suas G (in some mss Jer); non cum iudicio] BGS non cum sapientia 
Phil; in dimidio] BGS in medio B (mss DHL) Phil; autem] B; dierimi eius 
(avTov) ] BG d. suorum S dierum Phil; relinquent (kyKa-, KaraXdxpovcnv) ] B dere- 
linquent G Phil; eum (avrSv)] B (mss DL) G Phil ea B (mss B'-^'^Z) eas B (mss 
B'H) S earn B (ms B); et] add. usque G; in nouissimis suis (eir* ea-xaruv avrov)] 
B Aug in nouissimo suo GS in postremo Phil; erit stultus (eo-rai a(j)pwv)] BG 
erit insipiens S Aug Phil. 

It should be noted that in B, out of 233 Bible quotations (there are 
133 in Y), no less than 55 give the Vulgate text, a fact to be explained 
primarily by the identity of large parts of Jerome's version with earlier 
Latin Bibles, partly by later contaminations within Physiologus manu- 
scripts themselves; this very fact is evidence that the translator worked 
in large part from memory; furthermore, differences between Vulgate 
and apparent pre- Vulgate quotations in the Physiologus are of identical 
nature and extent regardless of their presence in or absence' from the 
Greek Phys iologus or any of its translations in other languages. ^^ 

^^ Burkitt, op. cit., p. 86. The Praefatio of Philastrius resembles the Physiologus 
in its moral simile, but reproduces no certain verbal detail. 
^^ I have listed the principal translations in my edition of Y (1941), p. 102. 



4 University of California PuMicatioyis in Classical Philology 

In the present paper I propose to give only what appear to me to be 
the most interesting Physiologus readings which have a possible or prob- 
able bearing on Vetus Latina studies, first, those I have not found con- 
firmed in other works but which may well be valid details, second, those 
that are attested in Patristic literature. The Bible versions I have used 
are recognized as standard, for the Vulgate several well-known editions, 
for the Vetus Latina the editions of the most important manuscripts. 
For sigla which are not identified here and for further notes on 
date, provenance, and contents, the reader should consult the works of 
Berger, Burkitt, Gwynn, and Jiilicher, already mentioned. The choice of 
Patristic writers, made according to the availability of reliable editions, 
is representative but far from exhaustive. 

VETUS LATINA MANUSCRIPTS 

6 Veronensis, ed. Old Lat. Bibl. Texts, Vol. 6 (Oxford, 1921). European Gospels. 

c Colbertinus (Bibl. Nat. lat. 254), ed. J. Belsheim (Christiania, 1888). Euro- 

pean Gospels with Afra elements. 

e Palatinus {olim Vienna lat. 1185), ed. Tischendorf (Leipzig, 1847). Afra. 

P Corbeiensis (B.N. lat. 17225), ed. J. Belsheim (Christiania, 1887), and 
O.L.B.T. 5 (1907). Gospels. 

g'^ Sangermanensis (B.N. lat. 11553), ed. P. Sabatier (parts) (Rheims, 1743), 
andO.L.B.T. 1 (1883) (Mtonly). 

k Bobiensis (Turin B.N.G. 7. 15), ed. O.L.B.T. 2 (1886) (Mt Mc). Afra. 

q Monacensis {lat. 6224), ed. 0. L. B. T. 3 (1888). European Gospels. 

r^ Usserianus I (Dublin, Trin. Coll. A. 4- 15), ed. T. Abbott, Evangeliorum versio 
Antehieronymiana (Dublin, 1884). European IVIt and loh. 

VULGATE 

D Liber Ardmachanus, ed. J. Gwynn (Dublin, 1913). Conflate New Testament. 
hex Hexaplar, Liber Hiob, by Jerome from the Greek, ed. Patrol. Lat. 29 (Paris, 

1846), pp. 29 se?. 
Jer Vulgate, ed. Pat. Lat. pp. 28-29, quoted when different from S. 

5 Sixtine-Clementine Revision (London: Bagster and Sons, 1894). 

W Nouum Testamentum, ed. J. Wordsworth and H. White, 3 vols. (Oxford. 

1898-1905), quoted when different from S. 
2 Sangallensis {1395), ed. C. Turner, The Oldest Manuscript of the Vulgate 

Gospels (Oxford, 1931). I have found no variants in this text pertinent to the 

present investigation. 

CHURCH FATHERS 

adv Cypriani (?) Aduersus Iiidaeos, ed. in Opera Omnia by G. Haitcl, Cur. SS 

Eccl. Lat. 3 (Vindobonae, 1868). 
A ug Variants from St. Augustine as given by C. Milne, op. cit. 
B Physiologus Latinus versio B (Paris, 1939). 
cath Cypriani de Catholicae ecclesiae unilate, ed. cit. 
(^yp Cyprian as quoted by C. Milne, op. cit. 



Carmody: Quotations in the Latin Physiologus 5 

Epi Sancti Epiphanii episcopi Inter pretaiio Euangeliorum (5 c ?), ed. A. Erikson 
(Lund, 1939). 

epis Cypriani Epistulae, ed. cit. 

exLc Ambrosii Exposiiio Euangelii secundum Lucan, ed. C. Schenkl, C. SS E. L. 32 
(Vindobonae, 1902). 

expl Ambrosii Explanalio psalmorum XII, ed. Petschenig, C. SS E. L. 64 (Vindo- 
bonae, 1919). 

expos Ambrosii Expositio psalmi CXVIII, ed. Petschenig, C. SS E. L. 62 (Vindo- 
bonae, 1913). 

fort Cypriani ad Fortunatum de Exhortatione martyrii, ed. cit. 

Iren Nouum Testamentum Sancti Irenaei, ed. 0. L. B. T. 7 (1923), pp. 40 seq. 

it Itala, ed. Jiilicher, Mt and Mc only. 

Phil Filastrii Diuersarum hereseon liber, ed. F. Marx, C. SS E. L. 38 (Vindobonae, 
1898). 

sing Cypriani (?) de Singularitate clericorum, ed. cit. 

Tert References to Tertullian from G. Aalders, Tertullianus' Citaten nit de Evan, 
en de Oud-Lat. Bijbel (Amsterdam, 1932). 

test Cypriani ad Quirinum Testimoniorum libri tres, ed. cit. 

Y Physiologus Latinus versio Y (Berkeley, 1941). 

zelo Cypriani de Zelo et liuore, ed. cit. 

SUGGESTED VETUS LATINA READINGS 

After elimination of such readings in the Physiologus as seem to represent 
translations from the Greek original, there remains a body of material 
which surely contains at least a few valid Vetus Latina variants. Our 
almost complete ignorance of the exact interrelationship of the several 
Latin versions of the Psalms has led me to reserve for more exhaustive 
study the twenty-two major variants which pertain to that book, all, 
however, fully attested in Gallican and Roman Psalters, in Jerome's 
translation from the Hebrew, and in Patristic writers. 

Gen 4.1 generauit B 33.20 — concepit et peperit S 

27.28 ubertate B 35.12— pinguedine S 

49.9 de germine B 1.1 — de frutice test 1.21 — ad praedam S 

Ex 17.11 superabat B 14.22 — praeualebat /or< 8 — uincebat S 

17.11 cum remitteret B 14.22 — ubi submiserat /or^ 8 — sin remisisset S 

Hiob 6.5 pabulum desiderans B 21.10 — escam quaerens hex — habuerit herbam S 

Prov 30.28 habitans in domibus B 30.3 — moratur in aedibus S 

Cant 4.3 sparcium (spartum ms A) colorem rubicunduni B 31.28 — restis cocci- 
nea Y 48.8 — resticula coccinea expos 18.22 — uitta coccinea S (cf. loh 
19.2) 

5.10 fratruelis B 31.32 — fraternus expos 5.8 — dilectus S 

8.14 conuertere fratruelis B 20.11 — fuge dilecte S 

Esai 1.7 igni crematae B 34.34 — succensae igni S 

53.4 imbecillitates Y 20.35 — infirmitates expos 22.3 exLc 10.105 — dolores S 

63.1 ascendit B 31.30 34.25— uenitS 

Hier 8.7 introitus Y 42.2— aduentus S 



6 University of California Puhlications in Classical Philology 

Dan 2.34 percisus Y 32.9 — abscissus S 

Mt 10.16 mites Y 13.1— simplices S 

Lc 16.16 praedicauerunt B 31.24 praedicatur Aug — euangelizatur S 

2 Tim 3.5 promissionem B 12.13 — formam D — speciem S 

ATTESTED VETUS LATINA READINGS 

I omit many less interesting though no less important examples, such as 
alternance between sic and ita. When W gives numerous examples, I 
merely set a note to that effect in parentheses. 

Ex 17.11 conualescabat -ebat S 14.23 /ori (ms 5) — inualescebat /ori — interfecit 

Y 17.16— superabat S 
Deut 33.17 primitiuos tauri species eius B 16.12 -tiuus t. s. e. Y 35.1 — primogeniti 
tauri pulchritudo eius S 
33.17 imicornis B 16.12 — unicornui Y 35.1 expl 43.18 — rhinocerotis S 
Hiob 6.5 clamauit -abit B 21.10 hex — rugiet *S 

Cant 1.3 vinguentum exinanitum B 23.49 exLc 6.34 expos 1.5 expl 40.15— oleum 
effusum S 
1.4 in cubiculum B 23.51 expos 1.5 — in cellaria S 

2.8 ecce fratruelis meus B 20.6 Y 21.6 — ecce f rater meus exLc 3.27 — uox 
dilecti mei S 

2.8 super montes B 20.6 12 Y 21.6 exLc 3.27 expos 6.5 — in montibus S 
2.15 pusillas exterminantes B 15.19 Y 18.11 exLc 7.31 expos 11.29 — paruulas 

quae demoliuntur S 
8.14 similis capreae B 20.11 exLc 3.27 — assimilare capreae S 
8.14 conuallium B 20.11 exLc 3.27 — aromatum S 
Amos 7.14 eram pastor B 39.1 pastor eram Y 28.1 exLc 8.28 — armentarius ego 

sum S 
Esai 1.2 genui B 6.6 adv 3 zelo (ms D) generaui Y 6.6 zelo 15 — enutriui S 
1.18 ut fenicium B 31.22 sieut phoenicium exLc 7.14 — ut coccinum S 

7.9 intelligibilia B 11.24— intellegitis test 1.5 — permanebitis 5 

7.14 in utero B 35.7 epis 10.4 exLc 2.15 18 7.10 Phil 142— om. 5 {cf. Mt 1.23) 

11.1 exiet B 35.6 /esi 2.11 exLc 2.24 exibit exLc 3.8— egredietur 5 

43.2 te non comburet B 30.10 fort 10 — non ardebit in te S 
53.4 infirmitates F 20.35 cxLc 10.105— languores S 

54.1 laetare Y 11.5 exLc 2.67 <es< 1.20— lauda <S 
Dan 3.18 imaginem quam statuisti B (ms A) 31.18 /<?s< 3.10 — statuam quam 

erexisti S 
Mt 3.7 generatio B 36.1 Y 12.1 exLc 2.73 expZ 37.2 expos 10.11 Am;? rf {sub Mt 
12.34) — progenies S 
4.4 uerbo dei S 22.13 bcg^ dei uerbo Ter< — omni ucrbo quod procedit de 

ore dei S 
7.14 intrant B 37.8 — introeunt Y 13.8 test 3.6 Au^ — inucniunt S 
10.8 daemonia B 35.36 <7U- A;/^ — daemones S 
10.16 astuti B 32.23 exLc 9.34 exp/ 36.10 37.8 Aw^— prudcntes S 
10.37 diligit B 27.19 b^^'g exLc 7.136 expos 15.20 fort 6— amat S 
12.33 de fructufi 32.19 dA;— ex fructu<S 
13.22 uoluptas B 13.6 c (e) — oblectamentum k — fallacia S 



I 



Carmody: Quotations in the Latin Physiologus 7 

19.27 dimisimus B 35.42 de ilusr— reli(n)quimus S (cf. Mt 19.29) 
24.13 permanserint Z? 4.10 c it {W 8 mss) — perseuerauerit *S 
25.41 praeparauit B 2G.18 def Tert Cyp Iren Epi 43— parauit hcg'^ Epi 27— 
paratus est S — praeparatus est W 
Mc 5.9 respondit B 29.8 e it — respondens dicit c^et dicit S 
Lc 1.69 suscitauit B 16.11 test 2.7 — excitauit e Cyp — erexit S 
loh 1.29 aufert Y 21.11 23.5 auferit 7^ auferet Iren test 2.15— tollit S 

3.5 intrare B 8.8 epis 73.21 ari— intrabit Aug Phil 120 148— introire S 

3.5 caelorum B 8.8 Tert Aug Cyp Phil 120 148 e— dei S 

3.14 herimo (eremo) Y 5.19 test 2.20 Aug Cyp — solitudine r^ -nem p — 

deserto S 
14.9 me uidet uidet B 34.21 expl 35.22 43.12 Tert— me uidit uidit cr^- uidit 

me uidit S 
14.18 dimittam B 23.38 dismittam d — relinquam S 

14.30 inuenit B 5.11 17.14 expos 8.6 Aug — inueniet Y 5.11 expl 38.27 exLc 
4.39 5.11/— habet/S 

16.33 gaudete B 23.35 ac Am^ — fidite e test S.Qfort 11 — scitote Aug — estote 
bf^q — confidite S 

19.2 induerunt {acpr^) clamidem coccineam B 31.29 el. co. ind. Y 48.9 — 
uestiraento purpureo induerunt er' — uestem purpuream circumde- 
derunt bcpD — ueste purpurea circumdederunt S {cf. interp. in D for 
Mt 27.28 induerunt tunicam purpuream; cf. Cant 4.3) 

Act 3.6 nomine domini B-35.30 exLc 7.55 — nomine S 

Rom 8.13 opera B 15.14 zelo 14 expos 12.17— facta S 

1 Cor 10.11 propter nos B 11.19 Aug — ad nos Tert — ad correptionem nostram 5 

15.55 contentio B 19.13 34.32 Tert (IF mss e<c.)— uictoria S 

15.55 aculeus B 19.13 34.32 expl 40.27 expos 12.43 Tert Iren- stimulus S 

2 Cor 5.21 cognouit Y 7.9 exLc 2.83 expl 37.34— nouerat S 

11.2 desponsaui B 34.14 g Aug — statui Y 39.5 de — despondi S 
Eph 6.12 seculi huius rectorum Y 13.18 — mundi huius rectores exLc 10.19 expl 
1.38 — huius mimdi aduersus mundi rectores D — aduersus mundi rec- 
tores S 
6.12 et spiritus Y 13.18 — contra spiritalia D — aduersus spiritus epis 58.8 — 
ab spiritalis expl 4.20 — aduersus spiritales Epi 26 44 — contra spiri- 
talia S 
Phil 2.8 se Y 20.33 expos 8.37 20.17 test 2.13 3.39— semet ipsum S 

3.14 contendo B 27.18 {etc. W) — contendit expos 4.26 — tendamus Epi 37 — 
persequor S {cf. Phil 3.13 extendens S) 
Col 1.19 plenitudinem diuinitatis B 35.20 {etc. W) — plenitudinem S 

3.11 iudaeus et graecus B 35.23 sing 14 exLc 4.9 {etc. W) — gentilis et iudaeus 
S 

1 Tim 3.16 in hoc mundo B 34.30 D — in mundo S 

2 Tim 4.8 sed et omnibus B 35.48 epis 10.4 — sed et iis (his W) S 

4.8 qui deligunt presentiam regni eius B 35.48 — qui dilegimt (for deli-, 
dele- ?) aduentum eius D — qui diligunt aduentum eius *S 
Heb 1.1 multifarie 5 31.10D— multifariam>S 

1.3 imago B 35.19 Phil 89 expl 35.22 38.24 43.12 expos 19.38— figura S 

11.34 uirtutem ignis B 30.8 Y 45.5 — impetum ignis S 



SENECA'S DIALOGUES VI, XI, XII 
(CONSOLATIONES) 

The Text Emended and Explained 

BY 

WILLIAM HARDY ALEXANDER 

VI, 1, 2 (151: 9) :^ non est ignotum, qualem te in persona patris tui 
gesseris. 

The codex Ambrosianus (A) reads : impersonam [using the barred p 
for per] , but the oldest editors of the Dialogi corrected this to in persona, 
and this has become the generally accepted reading, though the other is 
found in Lemaire and Fickert. 

The use of in with a personal word in the ablative to convey the sense 
"in the case of" is a common classical idiom, as in Vergil's famous : 
dolus an virtus, quis in hoste requirat ? Thus in patre tuo means "in the 
case of your father," and this would have served in the present passage 
to express the idea. Yet Cicero himself uses in persona with the de- 
pendent genitive in a not unlike sense, as in Pro Flacco 22, 53 : f ateantur 
in Maeandri persona esse expressam speciem civitatis. 

In the present passage I take persona patris iui to be merely a circum- 
locution, somewhat legalistic in character, for paier tuus, as in Pliny, 
N.H. 18, 3, 4, 17, where we read : ipse sua lege damnatus, cum, substituta 
filii persona, amplius quingentorum iugerum possideret. So too in Seneca 
himself, XII, 19, 2 : in mea tamen persona non tantum pro te dolet, 
where in mea persona replaces in me. This may be to avoid the ambiguity 
of the case of me in a phrase which leads off the sentence, 

VI, 1, 2 (151: 17) : non favisti consilio eius, sed dedisti manus victa 
fudistique lacrimas palam et gemitus devorasti quidem, non tamen 
hilari f route texisti. 

In this age of which and in which Seneca wrote, there were many 
people who, losing relatives under the most appalling and distressing 
circumstances, neither shed tears nor uttered sobs, and indeed managed 
to mask their feelings with a smiling countenance. Obviously, Seneca 
would not attribute to Marcia this ultimate hj^pocrisy ; hence he says : 
"you did not cover up your grief with a smiling face." Therefore the 

^ The numbers in parentheses give the page and line reference in Hermes' edition 
(Leipzig, Teubner, 1905). The VI refers to the number assigned in all modern 
editions of the Dialogues to the ad Marciam de Consolatione ; similarly at a later 
stage in this study XI will refer to the ad Polybiuni de Consolatione, and XII to the 
ad Helviam matrem de Consolatione. References to passages in the Dialogues will be 
made by number simply, without the prefix Dial. 

[9] 



10 University of California Piiblications in Classical Philology 

earlier part of the sentence, to correspond with the facts thus far set 
out, should mention a suppression of tears and a swallowing of sobs 
as things which Marcia did achieve, though she could not arrive at the 
stage play of the cheerful countenance. 

However, if fundere is employed, as the commentators tell us, in the 
sense of efftmdere, "to shed," the requirement just mentioned as neces- 
sary for the first part of the sentence is not met. Hence, if possible, 
fundere must be treated in such a way as to get some negative element 
out of it, and that can be done simply by translating it in its military 
sense "to rout" ; "you routed your tears in public" means "in public 
(whatever your practice in private) you shed no tears." This brings 
everything into order, and Basore's translation' is correct with regard 
to fundere, despite the remarks of Gertz in his apparatus criticus ad 
locum." 

Where Basore goes astray is in sajang in the latter part of the sen- 
tence "in spite of your cheerful face." This would make Marcia guilty 
of what we have called the ultimate hj^pocrisy of complaisance, which 
cannot be reconciled with her high, proud, and rather defiant character. 
Translate as follows: "though you routed public tears and crushed 
down sobs within you, yet you did not mask these things with a smiling 
countenance." At a time when people whose sons or fathers or brothers 
or husbands had recently been brutally done to death by the minions of 
the emperor would go about wearing a ghastly smile on their faces as 
if Fortune had just conferred on them her ultimate blessing, Marcia 
found herself incapable of doing this, "and that, too," as Seneca re- 
marks, "in an age when it was an extraordinary act of filial loyalty 
merely to avoid doing anything that was overtly unfilial." 

The quidem belongs to the entire sentence fudisti . . . devorasti, and 
provides the contrasting element to tamen. 

VI, 1, 8 (153: 25) : nam volnerum quoque sanitas facilis est, dum a 
sanguine recentia sunt; tunc et uruntur et in altum revocantur et 
digitos serutantium recipiunt, ubi corrupta in malum ulcus verteruut. 
non possum nunc per obsequium nee moll iter adsequi tarn durum 
dolorem : f rangendus est. 

In the American Journal of Philology, LIX (1933), p. 358, I reg- 
istered a suggested emendation recavantur, "are hollowed out," basing 
it on cavanfur in Pliny, N.H. 7, 16, 15, 70. As, however, revocantur 
appears in Cicero, De Oratore 2, 21, 88 : facilius in vitibus revocantur 
ea quae se nimium profuderunt, in the obvious sense "cut back," no 
good purpose is served in pressing for recavantur. 

2 J. W. Basore, Moral Essays of Seneca (Locb Classical Library, 1932), Vol. II, 
p. 5, with fn. on p. 4. 

^ M. C. Gertz, L. A. Senecae Dialogorum Lib. XII (Copenhagen, F. Hegel and Son, 
1886), p. 1G3. 



Alexander: Seneca's Dialogues VI, XI, XII 11 

Scrutantium, as against scrutantum, is supported by the double 
cretic clausula it helps to produce. As for adsequi, I favor the adoption 
of Triller's adgredi, not only on the basis of the great difficulty of 
properly explaining adsequi here and of the comparison with V, 40, 2, 
but also as representing a sound correction for that type of error whereby 
a word becomes replaced in transcription with another like it in length, 
general appearance, and grammatical form. 

VI, 2, 2 (154: 4) : quosdam ratio ducit, quibusdam nomina clara oppo- 
nenda sunt et auctoritas, quae liberum non relinquat animum ad 
speciosa stupenti. duo tibi ponam ante oculos maxima et sexus et 
saeculi tui exempla. 

R. Waltz* ends the first sentence after animum; thus ad speciosa 
stupe7iti becomes in his text a characterization of Marcia (tibi). In 
this change he is amply sustained by the clausula rhythm since ad 
speciosa stupenti with its hexameter close would be most unfortunately 
employed to achieve a sentence termination; on the other hand, non 
relinquat animum forms a double cretic with resolution of the initial 
long in the second foot. 

Against the change made by Waltz may be urged the fact that the 
use of auctoritates has already been said to have had no effect on Marcia 
(VI, 1, 6: fatigatae . . . auctoritates), while ad speciosa stupenti can 
only mean "standing dumb with amazement before striking things." 
The answer is, of course, that while chapter 1 has reported the utter 
imperviousness of Marcia to any appeal, yet, as the consolatio is at- 
tempted, Seneca must feel, or at least assume, that Marcia will still be 
accessible to some one of the usual methods, only more vigorously applied 
(VI, 1, 8 fin.). As Marcia has always in the past, before her great 
bereavement, expressed profound admiration for striking things, it is 
there with the speciosa that Seneca will renew the attempt, even though 
others have failed. 

Further, I do not think that the auctoritates of VI, 1, 6 are the same 
as the auctoritas of the passage under discussion. The auctoritates of 
VI, 1, 6 are coupled with adlocuiiones, and I understand the sentence 
to mean : "expressions of sympathy have been exhausted, as well as the 
attempts of great men who are your kinsmen to exercise their prestige 
upon your course of action." In the present passage auctoritas is asso- 
ciated with nomina clara; it is not personal prestige, but the prestige 
of great names in contemporary history, the sort of thing for which 
Marcia had probably been in the habit of expressing profound respect. 
This leads at once, and naturally, to a consideration of the cases of 
Octavia and Livia. 

*E. Waltz, Seneque, Dialogues (Paris, Societe d'Edition, "Les Belles Lettres," 
1923),Vol. Ill, Introd., p. 5. 



12 University of California Publications in Classical Philology 

VI, 2, 3 (154: 13) : Octavia (amiserat) Marcellum . . . adulescentem 
animo alacrem, ingenio potentem, sed frugalitatis continentiaeque in 
illis aut annis aut opibus non mediocriter admirandae. 

The sed is not easy ; even Muretus proposed to delete it, and some of 
the inferiores show sed et. However, Gertz is undoubtedly right in say- 
ing that it is precisely because of the sed that the error inpotentem, 
found in A, arose. Somebody thought that there must be a reflection of 
some kind on Marcellus in the phrase before sed in order to establish 
a proper contrast with frugalitatis . . . admirandae, and that supposed 
need was easily met by writing inpotentem. There is, to be sure, a con- 
trast implied in the sed, but it does not need to be worked out in the form 
of a reflection on Marcellus. The translation and exegesis is as follows : 
"Marcellus was keen and ingenious, and such qualities are in youth, if 
one is well-to-do, often employed, it must be admitted, in merely think- 
ing up ways of spending money and having a good time, but he was 
frugal and temperate as well as keen and clever." 

VI, 2, 4 (154: 21) : nullum finem per omne vitae suae tempus flendi 
gemendique fecit nee ullas admisit voces salutare aliquid adf erentis ; 
ne avoeari quidem se passa est, intenta in unam rem et toto animo 
adfixa. talis per omnem vitam fuit, qualis in f unere etc. 

Hermes, under the persuasion of P. Thomas, has joined intenta . . . 
adfixa to the words which precede it. On the contrary, I feel that for 
technical reasons the older punctuation, maintained by Gertz, should 
be adhered to; it closes a statement with passa est and throws intenta . . . 
adfixa with talis . . . fuit. Toto animo adfixa is a bad clausula, while 
quidem se passa est provides us with a sound dispondaeus. This is 
probably a more important consideration in determination of the text 
arrangement than any modern feelings of taste concerning the disposi- 
tion of the several phrases and clauses as between two adjacent sentences. 

VI, 3, 1 (155: 13) : Livia amiserat filium Drusum, magnum futurum 
principem, iam magnum ducem : intraverat penitus Germaniam et 
ibi signa Romana fixerat, ubi vix ullos esse Romanos notum erat. 

The reading of A : signum Bomani fixerunt, was altered by Muretus 
to the text as read above. But the only change really required is that of 
signum into signa and that for the reason that Latin would not speak 
as we would, of planting the flag, but of planting the flags, severally and 
concretel5^'' 

It has been held by some that the whole clause from et ihi to erat is an 
intrusion, but there seems no adequate ground for such a view ; with the 
Romans, as with all imperial people, it was always good form to mention 

^ Thus too C. Favez, L. Annaei Senecae Dialogorum Liber VI (Paris, E. de Boccard, 
1928),adloc. 



Alexander: Seneca's Dialogues VI, XI, XII 13 

the progress of the imperial arms, no matter what you interrupted in 
order to achieve the introduction of the point, even a funeral notice 
as here. 

VI, 3, 2 (155 : 27) : ut primum tamen (Drusum) intulit tumulo, simul 
et ilium et dolorem suum posuit, nee plus doluit quam aut honestum 
erat Caesare aut aequom Tiberio salvo. 

The reading as printed above from Hermes' text is that of Gertz 
(1889) ; in his edition of the Dialogues (1886) he reads: Aut aequom 
al<.tero filio sal^vo. The Ambrosian codex reads: quam aut honestum 
erat Caesarae aut aequo nialvo. 

It appears indispensable for the meaning that after aequom should 
appear the definite mention of a second party, consideration for whom 
would influence Livia in her attitude on the death of Drusus, and that 
second party must be Tiberius; hence the Tiherio of Gertz's later read- 
ing and the Nerone of Schultess. Yet neither of these readings explains 
the origin of the gap, and to do that fairly we must fall back on the 
possibility of a lacuna produced by the scribe in returning to the wrong 
one of two like syllables occurring close to each other in his exemplar. 
This could happen with <,altero s'>alvo, where the copyist's eye picked 
up the second al- in place of the first. This is, to be sure, Gertz's reading 
of 1886 less the <.filio>, which does not seem necessary for the meaning 
since <.altero^ standing alone is clarified by the preceding ilium; nor 
is there any imperative obligation that Caesare should be paralleled by 
a second proper noun. <cAltero s'>alvo is also a good clausula of Axel- 
son's first type.^ <i8^alvo modifies both Caesare and <,altero'>, of 
course. 

VI, 3, 2 (156 : 1) : non desiit denique Drusi sui celebrare nomen, ubique 
ilium sibi privatim publiceque repraesentare, libentissime de illo 
loqui, de illo audire : cum memoria illius vixit : quam nemo potest 
retinere et frequentare, qui illam tristem sibi reddidit. 

Of my previous note'^ on this passage I withdraw that portion relating 
to the association of cum . . . vixit with what precedes ; my present view 
is that it goes with what follows, and I therefore suggest a full stop 
after vixit. 

I also feel less certain about the possibility of dispensing with an 
expressed object for retinere et frequentare. It is at least possible that 
qui and illam have exchanged places ; the Ambrosian is of the eleventh 

«B. Axelson, Neue Senecastudien (Lund, C. W. K. Gleerup, 1939), p. 23, fn. 35, 
where his various clausula tj^pes are explained as he has worked them out for Seneca's 
prose. This work is an issue in Lunds Universitets Arsslcrift, N.F., Avd. 1, Bd. 36, 
Nr. 1. 

''Notes and Emendations to the XII Dialogues of L. Annaeus Seneca (Edmonton, 
Univ. of Alberta Press, 1934), p. 13. 



14 University of California Publications in Classical Philology 

century, with well-developed word separation, and not improbably is 
derived from an exemplar displaying like features. Where once word 
separation has become the established practice, the interchange of words, 
particularly small words in proximity to each other, is as natural a 
form of error as it is with ourselves. 

VI, 3, 4 (156 : 17) : quae enim, malum, amentia est poenas a se infelici- 
tatis exigere et mala sua fnon augere ? 

Several lines of emendation have been attempted, from the ultrasimple 
process of dropping non (so the F ms.) to finding in non an adjective 
{uno, Gertz; novo, Madvig; novis, Hauler), or an adverb {uliro, E. 
Thomas), or a noun (manu, Waltz), or to postulating a small lacuna 
(non <Cniinuere sed^ augere, Koch: non augere <Csed alere'^, myself 
in a previous attack on the passage*) . 

My present opinion is that non of A represents an original hoc, passing 
through the stages lie, nc, no.^ This hoc would refer to the poenas infelici- 
tatis a se exigere, which is regarded as the culmination of all other ills; 
we increase these ills hy this, namely, the exacting from ourselves of 
penalties for the unhappiness we encounter. The infelicitas we cannot 
escape, but we are fools (cf. amentia) to increase the natural and un- 
escapable pain of unhappiness by the unnatural and therefore escapable 
pain of self-torture for something for which we are not responsible. 

VI, 6, 3 (159: 15) : Turpis est na^agii rector, cui gubernacula fluctus 
eripuit, qui fluvitantia vela deseruit, permisit tempestati ratem ; at ille 
vel in naufragio laudandus quern obruit mare elavum tenentem et 
obnixum. 

The mare was added by the second hand ; Erasmus read elavum for 
navem of A. Erasmus' correction has been followed by all editors since 
his time with the exception of Favez, who writes : "je garde la lecon de A ; 
j'y vols une opposition entre permisit tempestati ratem et 7iavem tenen- 
tem: I'un abandonne son navire a la tempete, I'autre le tient ferme."" 

I wish to express my agreement with him in retaining navem and also 
in his point of view. The picture is a triptych in the first part of the 
sentence ; the contrast to it is not obtained by negativing just one panel 
of the threefold picture, as Erasmus' emendation does. But I would go 
furtlier than Favez. We know that the subject of obruit was missing in 
A, for which the second hand supplied, and no doubt correctly, mare. 
I suspect that whatever carried marc out of the text carried something 
more, namely, <Creciam^, and I therefore propose to read quern obruit 
mare rectam navem tenentem. 

* Op. cit., p. l4. 

" Gertz reports in his apparatus criticus, as he goes along, all the forms of abbrevia- 
tion employed in A ; among these is recorded iio for non. 
^" Op. cit., commentaire ad loc. 



Alexander : Seneca's Dialogues VI, XI, XII 15 

Favez understands tenentem in such a sense as found in arcum tenere 
or eastra tenere ; the navem is, so to speak, the position which the pilot 
holds against the assault of the storm. I am led by the famous passage, Ep. 
Mor. LXXXV, 33 : Neptune, numquam hane navem nisi rectam, where 
navem rectam is a translation of opdav rav vadv, to believe that Seneca, 
who is here dealing with comparable material, actually wrote what I 
have suggested. The pilot who holds the ship on a straight course is the 
skillful pilot, while Favez's phrase need not refer to anything more than 
a hrave pilot who holds the fort. But braveness is not enough to merit 
praise in the circumstances ; it requires the possession of skill as well. 

VI, 5, 2 (158: 6) : Cum secessimus et in unum convenimus, facta eius 
dictaque quanto meruit suspectu celebramus ; coram te altum nobis de 
illo silentium est : cares itaque maxima voluptate etc., 

The altum reading above is derived from the inf eriores ; the reading 
of A is aliud. If it were a question as between an alium and an altum 
there would be nothing to be said, but while an original altum could 
easily degenerate into an alium, it is not so easy to account for its passing 
into aliud. 

It seems possible that aliud should be retained, followed by a full 
stop. This causes the next sentence to begin with nobis and its second 
half with cares; there is thus established a contrast between the "we" 
and the "thou" situations. The preceding sentence, according to the 
proposed retention of aliud, will contain exactly the same contrast : "we 
among ourselves are loud in praise of your dead son ; in your presence 
it is another story." Punctuate with a semicolon after celehramus, a 
period or colon after aliud, and a semicolon after silentium est. 

VI, 6, 1 (159 : 3) : tuum illic, Marcia, negotium actum, tibi Areus adsedit. 

It is not in Seneca's style as presented to us by the manuscripts to 
omit the auxiliary verb with the perfect participle passive in the perfect 
tenses. Further, if we supply est (i.e., e) between negotium and actum, 
we get a greatly improved clausula of Axelson's first type. We should 
therefore read, with Waltz, <es'^>. 

VI, 7, 4 (160 : 12) : paupertatem, luctum, ambitionem alius aliter sentit, 
prout ilium consuetudo infecit, et inbecillum inpatientemque reddit 
praesumpta opinio de non timendis terribilis. 

I stand by my conjecture of 1934,"' namely, the insertion of < exilium > 
to follow paupertatem, lucium, the continuation of the lacuna by 
<ieorum^, and the replacement of the seemingly impossible ambitionem 

" Op. cit., p. 14. It is, of course, possible to cut down the size of the assumed lacuna 
by reading only <eorM/?i> without insisting on the analogy of 9, 4. 



16 University of California Publications in Classical Philology 

by admonitionem in the sense "reminder, suggestion." For this noun cf . 
Ep. Mor. XCIV, 25, 32, 33, 36, 44, with its idea of reminding people of 
things they have forgotten or overlooked. The extremely close analogy 
with the phrasing of chapter 9, 4 of the ad Marciam, and the circum- 
stance that in ethical teaching there is bound to be a reminder of life's 
adverse possibilities, unpleasant though the suggestion is to most people, 
combine with the easy paleographic shift from admonitionem through 
amonitionem to amhitionem to make a plausible case for reading <iexi- 
lium, eoriim^ admonitionem.^ 

VI, 9, 4 (162 : 3) : quis umquam nostrum de exilio, de egestate, de luctu 
cogitare ausus est ? 

But A has vestrum, which I think may well be right. Seneca's mind is 
working up to the idea of someone who is going to deliver an admonitio, 
which comes in the next sentence, and undoubtedly he identifies himself 
with this person who is to do the preaching. In that case, quis vestrum, 
"who of you and your friends," is reasonable enough ; the pulpit does 
not always identify itself with the congregation. 

VI, 9, 5 (162 : 12) : ille amisit liberos : et tu amittere potes : ille damnatus 
est: et tua innocentia sub ictu est. error decipit hie, effeminat, dum 
patimur quae numquam pati nos posse providimus. 

Madvig proposed <e^> to follow hie, with paleographical justifica- 
tion for <e^> out of the first syllable of effeminat, and Waltz has 
adopted this addition to the text. 

Apart from the fact that the alleged asyndeton is unpleasantly inter- 
rupted by hie (adverbial: "in this matter"), one should observe that 
with <e#> read before effeminat we obtain a double cretic clausula for 
the termination of the main sentence. This, added to other considera- 
tions based on the sense rather than the form, may be regarded as deter- 
minative for Madvig's reading. 

VI, 11, 1 (164: 18) : putre ipsa fluidumque corpus et causis fmorbos 
repetita sperasti tam inbecilla materia solida et aeterna gestasse ? 

To the conjectures recorded in Hermes' apparatus criticus add (Waltz 
and Favez) causis [morhos] repleta, where causis is to be regarded as 
a technical medical term meaning "predisposing causes for disease," and 
(Apelt) causis morhorum repleta. 

I stand by my previous comments on this passage." There is no sound 
reason to suspect repetita; morho repetita means simply "the victim of 

"For the whole idea of "reminder" cf. XI, 10, 5: in culpa est . . . mortalis animi 
spes avida, quae subinde, quid rerum natura sit, obliviscitur nee umquam sortis suae 
meminit, nisi cum admonetur. 

" Op. cit., p. 15. 



Alexander: Seneca's Dialogues VI, XI, XII 17 

disease time and ag-ain." Morho means disease in general, not any par- 
ticular disease, and while Gertz could write" comfortably from his happy 
experience "re vera rarius morbis ipsis adficimur" (i.e., the actual dis- 
eases as against occasional symptoms of this or that disease), Seneca's 
lot was different : nullum (genus malae valetudinis) mihi ignotum est.'^ 
Cf . also 22, 2 of this Dialogue : potuisse tot morbos ita evadere. 

If morho repetita is sound, as I think it is, an adjective modifying 
causis has dropped out ; as this chapter is crammed with errors on the 
part of the scribe from start to finish, the assumption is modest enough. 
In the light of the phraseology of §4 of this same chapter, <^minimis> 
is at least a reasonable suggestion to fill the gap. 

VI, 11, 2 (164 : 22) : hoc omnis ista quae in foro litigat, spectat in theatris, 
in templis precatur turba dispari gradu vadit : et quae diligis, vene- 
raris et quae despicis unus exaequabit cinis. hoc fvidelicet ilia Pythicis 
oraculis adscripta : NOSCE TE. 

<Spectat> is Gertz's conjecture, shifted by Hermes from after in 
theatris, where Gertz had placed it, to a position after litigat, for paleo- 
graphic reasons. I agree with Vahlen and Hermes that veneraris should 
be retained on the basis of the many other two-limbed asyndeta in 
Seneca's works; it also seems to provide the true contrast for despicis, 
which diligis does not. 

It is tempting to look at the hoc before videlicet as an anaphora with 
the hoc (i.e., hue) that opens the preceding sentence," but the senses of 
the two ''hithers" would be impossibly different, and the idea must be 
abandoned; hoc is here the pronoun. Videlicet is sound in the sense 
"obviously," used affirmatively and with emphasis. Seneca means that 
there has been much discussion about the true significance of the famous 
Delphic dictum, but that "obviously" the command is designed to force 
each one of us to consider how weak and frail a thing he is, as Seneca 
proceeds to demonstrate. 

No proposal which is particularly convincing has been made for filling 
out the rest of the sentence. Vox seems indicated by the ilia and the 
adscripta, and is the word commonly used in Latin to refer to philosophic 
dicta or oracular utterances, as in Cicero, De Orat. 3, 6, 21. 1 hazard the 
suggestion that <vox monet> may have dropped out after NOSCE TE; 
the scribe may well have become excited by the rare event of handling 
capitals. The nosce te and <vox monet> would provide a double cretic 
clausula. The nosce te occurring before vox would be quite in accordance 
with Seneca's manner of placing all modifiers between hie and ille and 
their nouns. 

" Op. cit., p. 412. 

^ Ep. Mor. LIV, 1. 

^« As I formerly did; cf. A.J.P. LIV (1933), p. 359. 



18 University of California Publications in Classical Philology 

VI, 11, 4 (165 : 15) : (quid est homo? inbecillum corpus et fragile) . . . 

precarii spiritus et male haerentis, fqiia parum repentinum audiet ex 

inproviso sonus auribus gravis excutit. 

The obelized passage is thought to constitute one of the most desper- 
ately corrupted places in a Dialogue which is in general notoriously bad. 
Yet is would seem pretty clear that the prevailing idea is this : that the 
human breath is so precarious a thing that a sudden shock will knock 
it out of us. As a matter of fact there is no difficulty from ex inproviso 
to excutit; it is around the fgwo parum repentirium audiet of A (and 
also of F, so that the corruption is probably of long standing) that 
the trouble is concentrated. Waltz recognizes this by writing quem and 
then cavalierly striking out the next three words ; but something must 
lie behind them. 

My conclusions on the passage after long and careful reflection are as 
follows. (1) qua is for q2ia<ndo> through the same peculiar type of 
error operating all through this section of the Dialogue to produce what 
Gertz calls "dimidiata verba." Its meaning will be "since," as, for ex- 
ample, in X, 7, 3 ; XII, 13, 6. (2) Lipsius' 2:>avor repentinus is the best 
available suggestion for pai-um repentinum, and may very well be right. 
There is more than one type of experience that can knock the wind out of 
us, and that is certainly true of a sudden fright. I suggest that <eM/n>, 
referring to spiritus, followed these words, thus bringing the reconstruc- 
tion at this point up to qua<ndo> pavor repentinus <eum>. It may 
have been a reverse influence -from <eum> that started the preceding 
words on their shift into an accusative, this in its turn absorbing the 
<e?tm>and leaving no trace. (3)1 regard audiet as a wrong develop- 
ment from audit', that is, from audit us ; some scribe supposed that audit' 
was a final verb, and its shift to audiet requires no great assumption. (4) 
Finally, I surmise that before audit' an <aut> has dropped out, along 
with the previously postulated <eum>; the <eum> disappeared into 
the final syllable of repentinum, and the <aut> was missed before 
audit'. The insertion of <aut> provides us formally with an alternative 
subject for excutit. 

• A combination of all these ideas produces : qua<ndo> pavor repen- 
tinus <eum aut> auditus ex inproviso sonus, auribus gravis, excutit; 
this I translate in connection with what goes before thus : ''Man is a 
weak and fragile body whose breath is precarious and insecurely seated, 
seeing that a sudden terror, or a sound unexpectedly heard which is 
harsh to the ears, knocks it out." 

VI, 11, 5 (166: 13) : immortalia, aeterna volutat animo et in nepotes 
pronepotesque disponit, cum interim longa eonantem eum mors 
opprimit et hoc, quod senectus vocatur, paucissimorwwi est circuitus 
annorum. 
The italicized letters rum are supplied from the codex Coloniensis of 



Alexander : Seneca's Dialogues VI, XI, XII 19 

Gruter, and the est by Gertz. Even thus, the passage has provoked 
doubts ; thus Havet writes of it as follows : "il y a du tomber une ligne ou 
plusieurs. Le coutexte, en effet, n'explique ni pourquoi il est question de 
la brievete de la vieillesse plutot que celle de la vie, ni pourquoi la vieil- 
lesse est caracterisee uniquement par sa brievete."" But the criticism is 
founded on a misapprehension. Senectus is here used a little loosely to 
mean not specifically the portion of life which is described as old age 
ordinarily, but a life that has lasted long enough to attain to what is 
called old age. This can be made to appear in translation : "while in 
the meantime death overtakes him as he plans far into the future, and 
the thing that is called 'living to a ripe old age' encompasses (at best) 
a very few years." 

VI, 12, 3 (167: 19) : iuvenis cito prudens, cito pius, cito maritus, cito 
omnis officii curiosus, cito sacerdos, omnia tamquam properaws. 

The reading of A is propera; the addition of the -ns to form a present 
participle is the suggestion of Sehultess. Gertz proposed properaret, 
though not in his 1886 edition. Other alterations have been attempted 
also. 

It is not necessary to make the adjective of the last phrase agree with 
iuvenis or to alter the phrase so that iuvenis will appear as its subject ; 
the resumptive omnia, in my judgment, pushes iuvenis out of the way 
and acquires the modifier for itself. The manuscript reading, if omnia 
is scanned as a dissyllable, is a double cretic with the final long of the 
second foot resolved (-w-|-wwv.^)^ and this tends to confirm one in the 
opinion that it had better be let alone. 

VI, 12, 4 (168 : 1) : circumfer per omnem notorum, ignotorum frequen- 
tiam oculos, occurrent tibi passi ubique maiora. 

The A F reading is retained by Favez, who is consequently obliged to 
exclude oculos from the text; but I observe that -rum frequenti{am) 
oculos produces a double cretic clausula, so that oculos seems to have its 
place securely vindicated. 

It would seem as if we had here once again an example of the extraor- 
dinary habit developed by the scribe of A and described as follows by 
Gertz: "non paucis locis miro modo extremae vocum partes praeter 
omnem cousuetudinem omissae sunt, ut verba tantum dimidiata scriber- 
entur."^* If the correct word here is circumfer, as seems probable, the 
scribe read it circuinfer, and dropped the letters -nfer. 

VI, 12, 6 (168 : 16) : L. Sulla filium amisit, nee ea res aut malitiam eius 
et acerrimam virtutem in hostes civesque contudit aut effecit etc. 

This passage failed to satisfy Lipsius, who wrote militiam for mali- 
tiam, or Gertz, who wished to turn malitiam into maleficam; both pro- 

" L. Havet, Manuel de critique verhale (Paris, Hachette, 1911), p. 200, §846. 
^^ M. C. Gertz, Studia Critica in L. Annaei Senecae Dialogos (Copenhagen, F. 
Hegel, 1874), p. 40. 



20 University of California Publications in Classical Philology 

posals are unfortunate. Cicero provides an admirable definition of 
malitia in N.D. 3, 30, 75 : est enim malitia versuta et fallax nocendi ratio, 
and it must be agreed that Sulla displayed that quality conspicuously 
enough in his career. Equally clear was his vigorous courage whether in 
external wars or in civil strife, so that malitiam et acerrimam virtutem 
is really unchallengeable. One would probably expect his malitia to be 
set over against his virtus by means of an aui, but that was simply not 
possible within the scheme of an aut . . . contudit aut effecit sentence. 
There is the further point to be made for malitiam that almost immedi- 
ately following we have : nee odia hominum veritus, quorum malo illae 
nimis secundae res constabant. Malitia, as Cicero says, is a ratio nocendi; 
it works out in such a way that the prosperity (res secundae) of him who 
applies it is realized at the cost of the hurt (malum) of others. Malitia 
is a studied capacity for inflicting malum. 

VI, 15, 2 (171 : 2) : di\Tis Augustus amissis liberis, nepotibus, exhausta 
Caesarum turba adoptione desertam domum fulsit: tulit tamen 
<tam> fortiter quam cuius iam res agebatur cuiusque maxime in- 
tererat de dis neminem queri. 

There seems to be in most quarters a f ogginess about the translation 
and exegesis of the second part of this sentence which is not at all neces- 
sary. When Augustus adopted Tiberius Claudius Nero in a.d. 4, he was 
sixty -seven years old, and hence nearing the time when, having passed 
from among men, he would himself become a god (di^'ns Augustus). 
Now a god cannot allow himself to be considered as affected by fate, nor 
can he admit criticism of the gods in general for what they either do or 
fail to prevent being done, and Augustus acted in that spirit in respect 
of the many bereavements. that fell on him in rapid succession. The 
translation runs: "yet he bore the blows as bravely as if one whose 
personal interest was already at stake, and whose chief concern it was 
that no one should register a complaint against the gods." The iam is 
very important; he was already, in effect, a god, and had a personal 
interest in appearing calm and undisturbed, Olympian-fashion, amid 
his misfortunes." 

VI, 16, 1 (172: 1) : par illis, mihi crede, vigor, par ad honesta, libeat, 
f acultas est. 

Of liheat Gertz (app. crit. ad loc.) writes : "sic nude positum ferri vix 
potest." That seems an overdogmatic assertion. We are dealing with 
Silver Latin and hence with a syntax and style developed in the schools 
of rhetoric, largely on the basis of poetic models. Now poetry is con- 
stantly obliged to study the accommodation of the normal methods of 
logical expression to a prescribed container dictated by the metrical 

^' See Favez's excellent note ad loc. 



Alexander: Seneca's Dialogues VI, XI, XII 21 

form. This may mean dispensing with much on which normal prose has 
a good right to insist, and it is entirely conceivable that Silver Latin 
carries such economies, the economies of poetry, over into its own poeti- 
cally colored prose. That libeat means dum Wheat, libeat modo, liheat 
tantum in this connection seems obvious, and to my way of thinking its 
subordinate force as a proviso is given by assigning to it a lower voice 
level than the main proposition par ad honesta facultas. There, once 
again, is the influence of rhetoric, in that written prose can contemplate 
the syntactic effect of voice levels. 

VI, 17, 2 med. (174: 24) : licebit enim tibi avidissimum maris verticem 
perstringere. 

Perstringere is the emendation of Pincianus on the A reading re- 
stringere. I fail to see any particular justification for restringere,^' and 
venture to suggest that the A reading is an error of some sort for plain 
and simple stringere, used as in Vergil's famous description of a critical 
moment in the boat race, Aen. 5, 163 : 

litus ama, et laevas stringat sine palmula cautes. 

It may also be observed that verticem stringere is a double eretic clau- 
sula ; this throws doubt both on the A reading and on Pincianus' remedy. 

VI, 17, 6 (175 : 25) : dicit omnibus nobis natura . . . nihil certe spopon- 
derunt. 

Despite a certain speciousness in Gertz's assertion^ that this whole 
passage is properly to be placed immediately after the last sentence of 
our present chapter 18, a speciousness emphasized at first glance by 
sustulerunt and sustuleris, I think with Kronenberg" that no trans- 
position is necessary. Chapter 17 opens up with the assumed complaint 
of Marcia that it is a hard thing to lose the child you have reared just 
when he has reached young manhood and is beginning to be a help and 
a source of pride to his parents. The answer is given under a figure : a 
voyage to Sicily would be a wonderful experience, but it would involve 
discomforts too. Parallel to this figure of a voyage to Sicily lies the 
question of assuming the responsibility of motherhood. You have not 
blundered (incidisset) into it, but have accepted it prudens sciensque; 
there is no just cause of complaint. Nature has made it perfectly clear 
to us, each and every one, that a child born is a highly speculative 
venture from any point of view whatever, and if your particular experi- 
ence is a d isappointment or a grief, you have no recourse against her. 

^ I formerly attempted, op. cit., p. 17, to read the words as verticem re ("actually") 
stringere, but the rhythmic ending is against it. 
^ Studia Critica, p. 112. 
'^C.Q.U (1908), p. 38. 



22 University of California Publications in Classical Philology 

VI, 18, 2 (176: 21) : videbis illic innumerabiles Stellas micare, videbis 
mio sidere omnia implere solem, cotidiano cursu diei noctisque spatia 
signantem, annuo aestates hiemesque aequalius quidem dividentem. 

A reads : videre micahis . . . impleri. solem, etc., with which F agrees 
except in having miraberis for micahis. Aequalius quidem is Gertz's 
emendation on A's aequaliusque ; Favez and Castiglioni read aequalius 
[que], and my own suggestion of aequali usque <'yice>^ still merits 
consideration. 

Larisch's micare videhis for videre micahis of A seems at first sight 
like one of those inspired emendations which should command immediate 
acceptance, but as a matter of fact a second videhis coming so soon is 
improbable if one studies the economy of its use throughout this whole 
passage to the end of § 7. It is not used to introduce contrasting clauses, 
as would be the case if it occurred here, but is always continuative, 
accumulating additional facts or objects of view. 

Under videre probably lurks vigere; we must remember that these 
stars are living things — gods, indeed, according to De Benef. IV, 23, 4. 
Micahis is soundly corrected by the miraheris of F ; miraheris in the A 
tradition was first deformed into mirahis and then either by chance or 
design, more likely the former, strange as it may seem, into micahis. The 
uno sidere omnia inpleri of A F is correct, and solem, signantem, dividen- 
tem should all be in the ablative in apposition with u7io sidere; these last 
have all become accusative under the idea that sole was connected with 
the videhis following, its object in fact, and hence solem. Note in A F the 
period after impleri, which would contribute to that view. 

The translation is : "you will behold there countless stars flourishing, 
you will marvel that by a single heavenly body all things are filled, the 
sun, who with his daily round marks the bounds of day and night, and 
with his annual course divides the summers and winters ever in even 
succession." The point is that, while you see so many stars, you cannot 
but remark that it is from a single heavenly body, the sun, that all things 
are filled ; hence the wonder {miraheris). 

There is no need of a comparative adjective or adverb in the latter 
part of the sentence ; in fact, the effect of the presence of such is, in my 
opinion, to set up a comparison never intended by the author. My sug- 
gested aequali usque <ivice^ takes the aequaliusque of A and, by adding 
<ivice^, provides a noun for aequali to modify, thus leaving the adjec- 
. tive in the positive degree. Further, usque <vice> dividentem is as 
good a clausula as any of the others suggested. 

VI, 18, 5 (177: 16) : varii urbium situs et seclusae nationes locorum 
difficultate, quorum aliae se in erectos subtrahunt montes, aliae fripis 
lacu vallibus pavidae circumfunduntur. 

This is apparently a very hard passage, if one concentrates on the 
=» Op. cit., p. 18. 



Alexander: Seneca's Dialogues VI, XI, XII 23 

emendations offered, but actually there is only one small error in the 
manuscript tradition ; the final w of lacuu was lost. 

After mention of cities there comes naturally the exact opposite, the 
mention of noncity types; of these Seneca selects striking examples: 
(1) people who live high up in mountain ranges into which they have 
withdrawn from contact with other tribes ; (2) people who live on (that 
is, near) the shores of lakes, lake dwellers in short, examples of whom 
Seneca had no doubt heard of, and loves as a scientific amateur to 
mention; (3) people who live in valleys — obviously deep, canyon-like 
valleys, because a contrast is made between them and the sort of people 
who live on mouutaintops. The truth is that for exactitude Seneca should 
have said : "tribes cut off by difficulty of terrain, of whom some withdraw 
into perpendicular mountain heights, while others in their fear seek the 
bottoms of deep canyons." He accomplishes the first part of this satis- 
factorily, but then he thinks of the strange lake dwellers, and, no matter 
what happens, they must be worked in as soon as thought of; they 
"spread themselves along the banks of lakes." Then he comes back to 
his original idea with vallihus, which merits an epithet but is left to 
get its description from antithesis to m erectos monies and must share 
its verb with ripis lacuuni. Under this view the sentence translates thus : 
"city sites of all sorts and tribes cut off (from human intercourse) by 
difficulty of terrain, some of whom seek refuge in the perpendicularity 
of mountains, while others scatter along the banks of lakes, the bottoms 
of valleys in their fear." Ripis lacuum, vallihus is Seneca's bimembris 
dissolutio ; see Hermes' critical note on III, 6, 11. 

VI, 18, 6 (177 : 22) : sparsae tot per vastum insulae, quae interventu suo 
maria distinguunt. 

I wish to associate myself heartily with Axelson"* in his protest against 
the arbitrary way in which paria of the mss. has been transformed by all 
editors except Fickert into the tame maria, and that too in the face of the 
fact that the scribe of the Ambrosianus has rarely if ever confused m 
and p, an intrinsically unlikely happening at best. As Axelson says: 
"worin lage das Anstossige des Gedankens, dass die Inseln 'der Ein- 
formigkeit Abwechslung verleihen'f Paria is the sea, the traits of 
which are always the same except where islands diversify them, but how 
much more brilliant is paria than maria, especially in association with 
a verb like distinguunt! Translate : "islands scattered through the vasty 
deep which by their occurrence differentiate things (otherwise) alike." 

VI, 18, 6 (177: 23) : quid lapidum gemmarumque fulgor et inter rapi- 
dorum torrentium aurum harenas interfluens etc. ? 

The reading of A is harenis; the difficulty in adopting it is that no 
object is pr ovided for the preposition inter, some noun which would at 
^ Op. cit., p. 17. 



24 University of California Publications in Classical Philology 

the same time control the genitive rapidorum torrentium. Hence the 
common reading of the inferiores harenas aurum, with transposition of 
the two words. Gertz adopts the harenas but leaves its position unaltered, 
an arrangement which I cannot myself accept at all. Various other 
proposals have been made ; it is noteworthy that F introduces cursum 
between rapidorum and torrentium. 

I incline to believe that the difficulty lies with inter, in which I detect 
a truncated interim in the sense commonly given to it in Silver Latin of 
interdum ("sometimes") . Rapidorum torrentium harenis is a very strik- 
ing phrase; "torrents swift with sands" means "torrents the swiftness of 
which is revealed by the way they bear sands along with them." This 
phrase and aurum interfluens are then naturally interlocked, without 
distortion of any kind. Interim is soundly inserted because the presence 
of gold in the sands of swift-running torrents is by no means a universal 
phenomenon, and indeed in the Mediterranean area uncommon enough 
to warrant a guarded statement. 

VI, 18, 8 (178 : 19) : respondebis velle te vivere? — quidni? immo, puto, 

ad id non accedes, ex quo tibi aliquid decuti doles — : vive ergo ut 

convenit. 

The text is sound though something less than lucid, and the punctua- 
tion supplied by Hermes does not help. I stand in the main by the 
interpretation offered by myself ten years since."* With one or two 
amendments and additions, it may be restated as follows. 

After the long account of life's difficulties and rewards running 
through chapters 17 and 18, the narrator asks : "will your answer be that 
(understanding that you must go through the d