UNIVERSITY OF CALIFORNIA PUBLICATIONS

CLASSICAL PHILOLOGY

EDWARD BULL CLAPP WILLIAM AUGUSTUS MERRILL
HERBERT CHESTER NUTTING
EDITORS

BERKELEY
THE UNIVERSITY PRESS
1904-08

PA

If.

Cited as Univ. Calif. Publ. Class.

Phil.

CONTENTS.
Number 1.—Hiatus
in

Greek Melic Poetry, by Edward Bull Clapp.

Pages

1-34.

Number

'2.

— Studies
II.

tus.

I. Concessive Si-Clauses in Plautus. in the Si-Clause. Subjunctive Protasis and Indicative Apodosis in PlauBy Eerbert C. Nut tiny. Pages 35-94.

Number

3.

— The

Whence ami Whither
by Benj.
I

of the

Modem

Science of Language,

do Wheeler.

Pages 95-109.

NUMBER

4.

— On the Relation — The

of Horace to Lucretius, by William A. Merrill.

Pages 111-129.

NUMBER
NUMBER

;".

Priests of Asklepios, a New Method of Dating Athenian Archons, by William Scott Ferguson. Pages 131-17.'!.

(5.

— Horace's
— Some

Alcaic Strophe, by

Leon Josiah Richardson.

Pages

175-201.

NUMBER

7.

by Henry Washington Prescott.

Phases of the Relation of Thought to Verse in Plautus, Pages 205-262.

Index.— Pages

263-270.

UNIVERSITY OF CALIFORNIA PUBLICATIONS

CLASSICAL PHILOLOGY
Vol.
1,

No.

1,

pp. 1-34

June

I,

1904

HIATUS IN GREEK MELIC POETRY
BY

EDWARD

B.

CLAIM'

It is the

prevailing usage of older Greek poetry to elide mosl
,

1 short vowels, and to shorten most long vowels or diphthongs

when they ocenr
ately, in the

at

the end of words and are followed immediverse,

same

by words beginning with

a

vowel.

The

negleet to elide or to shorten, under such circumstances, constitutes hiatus.

But

in defining the
is

scope of the present investiga-

tion the term hiatus

loosely used to include all eases where, in
a

the ordinary

lowed

word beginning with a vowel, no matter what the explanation of the phenomenon may be.
in the

modern texts, same verse by

word ending with

a

vowel

is

fol-

a

Hiatus in

Homer
2

has been made the subject of exhaustive
3

and the essential features Homeric usage are well known to scholars. Hiatus in Pindar has been touched upon by Hermann " and Boeckh, 8 and the various editors, and Hartel gives some statistics. The observaof
1

study by Ends, Hartel, and Grulich,

4

tions

of

Tycho Mommsen

in his

Supplement
for
is

8

are

still

worth
in

reading,

but the fullest collections

Pindar are
Pindar.
1

found

August Heimer's Studia Pindarica." which
for
its

specially valuable
"

careful study of the
11

digamma

in

Schone's De

Dialecto Bacchylidea
1

is

useful for Bacchylides, and the traces

at

is

II.

a, e and o, are elided with complete freedom, and on the ether hand -De Digamma Homerico l. 35 tl. often elided, and sometimes even oi, l .in Stud. 2 and 3. De Quodam Hiatns Genere, Halle 1876. 5 0pusc. I, 247 ff.
.

But only

6

s 7 Horn. Stud. » Lund, 1884. '"The p. 165 tv. 3, 8 ff. based on the much improved texl of Schroeder, and is independent of Heimer's, though the hitter's results, wherever they cover the same ground as mine, have been compared, to insure completeness. In regard

In his edition

1, 2,

101

ff.

present

investigation

is

to the

digamma,

in particular,
is

1

have been aide

to

add bul

little,

besides a few
181
ff.

new

references, to what

offered by Heimer.

" Leipziger Studien

19,

University of California Publication}

Class. Phil.

of the

digamma
ci1

in

AJcmai

Alcaeus,
mill

and Sappho,
recently

have been

critically treated In

by Cleuim,
£

more

by Solmsen
I

ing

i

be odes and

gments

of

Pindar

have followed

l\ the anmbering, and usually

the text, of Schroeder.
I

For Bacfollowed
of
Eiller-

chylides,

Lnclnding
of

the
foi

fragments,
the

have

usually
thai

the texl

Kenyon,

Melic

Fragments
hiatus

Crusius, and for Timotheus, thai of Wilamowitz.
h should be noted,
al

the outset, thai

is

far

less

freIf

quenl

in

Pindar and the other melic poets thau
a

in

Homer.

we take as

basis of

comparison the
in

last

six

books of the

Iliad.

which are approximately equaJ

extenl

to the

surviving odes

ami fragments of Pindar, we
real or apparenl hiatus in

find

uo

Less

than 2000 instances of
in
tl

the
loo.
at

Homeric books, while
In Attic tragedy,
all.''

Pindar
ther

the

number

is

less

than

on

hand, hiatus scarcely e\i>ts
;i

so that the melic poets

occupy

careful finish of Euripides.

middle ground between the freedom of Homeric poetry and the Certain facts as to the nature and
of
this difference,
a

circumstances

in

detail,

will

appear

in

the

course of the discussion, bul

general

view

may

he obtained

from

the following table,

which shows the progressive disapto the tragedians,

pearance of hiatus, from

Homer

on the basis of
in

number

of instances which are found, on the average,

inn

consecutive verses.
In

loo

Verses of

vol.

i.]

Clapp.—Hiatus
in

in

Greek Melic Poetry.

3

comparatively rare
in

the melic poets, and practically disappears
in

tragedy.

In

discussing hiatus

the melic
hiatus,

poets,

we

shall

Consider,

first,

the eases of apparent

next

the eases of

hiatus after a long vowel or diphthong, and finally those which

occur after

a

short vowel.

I.— Apparent Hiatus.
.More than one-fifth of
all

the instances of hiatus

in

Homer

are only "apparent," or due to the influence of an obsolete eon-

sonant.'

Gottfried

Bermann denied

2

the existence of this phe-

nomenon in Pindar, but few scholars would now agree with the The pronoun oS, o!, e occurs 58 great master in this opinion.
times
in

Pindar, and

1!»

times

in

the other melic poets.

In

75

places out
satisfactory

of the 77 the influence of the

digamma
a

is

the only
a

explanation

of hiatus, or of the lengthening of

syllable consisting of a short vowel followed by

single conso-

nant

at

the

end of the

preceding
the

word.

One
and

ease

proves
3

nothing,
{air

either

for or against

digamma,
illicit

Corinna
is

em), the only instance
3

of neglect of the

digamma,
hiatus
is

proin

bably corrupt.

In view of the fact that
these

not,
if

general, of frequent occurence in

poets,

and

that,

we

admit the influence of the
is

digamma

in places

where the evidence
that

fairly conclusive, the

residuum of unexplained cases of hiatus

becomes almost a vanishing quantity, we can hardly doubt this consonant was felt by the melic writers.

The following
with the places

list

includes the

digammated words

in

Pindar,
is

in

which the influence of the digamma
of hiatus, but in a few.
to

felt.

Most of these are cases
asterisk, the

marked by an

consonant helps
T
<>,

make

position.

1

4:;i

c-iscs in
1.

Hi-

inn' in

every cine verses.

-

Opusc,

217.

latter

See p. 12. So closely, in fact, is the digamma bound t<. this pronoun that the seems scarcely able 1" live without it. In Attic, wlieretlie digamma i- entirely On the digamma in this lost, the pronoun itself leads bu1 a precarious existence. word, see the discussion by Dryoff, in Kuhn's Zeitschrift, 32, 87 H'.
3

University of California Publications.
DIG \mm
\ti:i>

WORDS

IN'

PlND

Vol.

l.l

Clapp.— Hiatus
DlGAMM
\TKI>

in

Greek Melic Poetry.
IN

WORDS

PlNDAR.
D1GAMMA

',

Ol, €

().

1.

23;

1.

65;

2.
!>.

42*;

6.

20; 6. 65;
1<».

7.

89; 7. 91;
•_'!):

L5; 9. tl7;
L3.

87;
:

L3.
7(1:

i:i.

-7

;

65; L3. 71
1".
I.

13.

i:;.

HI
:;.

:

11.

20;
'-':!:

1.

7; 2. 42;
1.

2.
4.

83;
7;i:

63; 4.
L89;
.">.

::7;
1.

48;
4.

I.

I.

197;

243;
9.

264;
;».

4.

287;
84;
;

117: 0.
L09;
1.
!».

:i<i:

56;
1.

82;
1.

!'.

9.
1.

L20; N.

14:
7)7:

L6 b

58;

61; 3.39;
7.

3.

4. 59;
10.

5.

34; 6. 23;
10
31
:

22; 7.
I.

40;

29;
6.

10.

79;

4.

64; 5. 62;
214.
P.
4.
1.

12; 6.

19; 8. 57; Fr.

36 (emendation.
inser.
-

Cf.

flv

in

a

Metapontine and fivavrf
2.

Collitz
in

L643,

iavr$

Leg. Gort.

40).

OS. €05

['.

6.

37;

1.

4. 36.

OIK—

P. 7. 5;

8.

51; N.

6. 25.

Total
be noticed that the cases where the

129

185

li

will

digamma
in

helps to

make

position are very few in
it

number

(4).

comparison with
neglect
its

those in which

prevents hiatus (125).
is

The

of the difailure
to

gamma, on the other hand, make position (133 times),
elision a

seen most often in

less often in its failure to prevent
at the
is

(49 times).

digammated word, and once

Twice crasis takes place a long vowel

beginning of

shortened under

similar circumstances.

To

the words in the above

list

we may add several others,
in

mostly proper names, which probably had the digamma. bul
regard to which the evidence
are:
1.

is

not entirely conclusive.

These

'Iojako?.

This occurs twice

in

Homer, both times with
1

the
five
1

digamma
times:

possible but not required.

In

Pindar

it

appears

In X 256, at verse-end, evpvxbpy 'Iaw\K$.

[n

B

712 ivKTiiUmiv \au\ic6v.

(i

University of California Publications.
77
L88
:;i ;,»

L<

I".
I',

|.

k\«tSs 'IwAkoC
es Si
icai

neutral.
,

i.

'IaiAxdV

required.
[uired.

N. X.
I.

:;.

os

'IwXkov

I

|.

Xarpuiv 'lawXicov
eodd.
f/j"TM

neutral.
bu1
-

s.

in

uhoXkov,

-

is

required,

and

<£u(riV (for <£avTi)
I

isun-Pindaric.

Most edd. adopl the
which
uncer.

Bothe

<£<itis-

'1o)\kov (Schroeder 'IooXkov)

does nol admil
tain, bul

,\

The etymology
th<

ol

the

word

is

Schroeder suggests
qo!

rod svelk (sulcus
in

2.

'Io'Ahck.
it

This
t

word does
Lines:
l

appear

Bomer.

In

Pindar
0.
I

occurs seven
98 79
.

9.
9.
]

avT<3 'IoXaov

I

.

7tot€ kuI 'IdAaov

(_----

I'.

1

GO
:;7

8ia</>ep« 'IoXaov

N.
1.
[.

:;.

TeAa/naw 'IoAa
i;

1.

16
:;ii

'IoXaov (-

- -

— ._-~~ — —
)
)

-

f probable. f probable.
)

p probable.
neutral.

f probable.
neutral.

5. 7.

t7T7roo-oas
rj

'IoAaos

I.

9

d/x(f>'

'IoXaov
calls attention
to
1
1 1 * -

Heimer
name,

f impossible. fad that, in a Bo<
Ik-

like this,

the

digamma would

likely to surv>>-

vive longer than in other dialects.

For Corinthian,

may

add, the

digamma
tin

is

proved by /rioAu Collitz 3133.

For the derivation of
p. 4).

word

cf.

/tioVAoko^ (see above,

3.

'lerfl/Lio?.
it

This word

is

of
I<

uncertain

origin.

Curt ins
4

connected
'Io-0/io?

with
its

-

i

-

to go.

does not appear in Homer.
in

and
i>

compounds occur 21 times
in

Pindar.
1

In IS

places/:

impossible, and six places are neutral
three places:

.

Put the ad-

mission off removes hiatus
I. I.

1.9
1.
:V2

dAiepKt'u

'Io-#/xoi'.

Uoo-etSdwvL 'Io-0/aw (on'- of the two occurrences
hiatus after

in

Pindar of

-

t

of the dative singular)

Frag. 122. 10
Cf. also,

Xe£ovvn

'lo-0fiov.

Bacch.
4.

2.

7

ui'xe'n 'lo-$ixov.

'la'Afcro?.

This word appears once

in

Homer,

in
is

I'.

656

AivSov
sible,

'IrjXvo-ov re

(-~~- -~),
in
2

where the digamma
1.

impos-

as

is

the case, also

Timocreon

7 es irarpih'' 'IciXvaov.
kIoo\k6».
300.

iSoChrist. Schroeder S^IaoXriv.
i

So Christ. Schroeder
<r

-Onm.l'See behrw

But
9.

cf. fodfuov

[necklace] »1 beginning of verse,

(p. 7

1

en 0.

112.

vol.1.]

(

'/a />/>.

— // iat us

in

Greek Melic Poetry.

7

Bui the marked hiatus in the single occurrence of the word

in

Pindar
<

>.

7.

74

-re

\d\vaov
in

points strongly
5.

the opposite direction.

1

'ISato?, "\8as.
in this

There

is

no evidence

in

Homer

for the
to p:
the

digamma
O.
5.

name, and bu1 one place
Here

18

peovra 'iScuov.

in
is

Pindar points
called

for, so thai in

quantity, as well as the aiatus, indicates an error

the

eodd.. the

firsl

syllable of 'ISuiov being long.
(so Schroeder),

Bu1 some

codd. read peovr' 'iScuov
hiatus,

which avoids
\\

and substitutes - - for -

-

.

a license

hich

is

perhaps admissible.
G.

"IXa?, 'iXm'Sa?, "Wiov, ^l\o?.
2

and Curtius
consonant.
0.
7.
9.

thought that the
Suiti 'lAtaSa (see
">.

o in

In Homer. "iXto? has f Otkev; was due to the same
:

In Pindar but one passage supports the p

112

on

I.

1.

:;2

above,

]>.

(i).

Ill

0.

11

iroTa/Jiov

re'Tlavov

(r.l. "(.laviv)

we arc conmenCf.

fronted

by an almost

connected with the

unknown proper name, name of an oriental fish-god
(67)

possibly to be
'Cldvvrjs,

tioned in a fragment

of the historian Apollodorus.

Dagon.
(pdpavos,

The hiatus has caused the digamma
3

to be suspected
.

Bui cercf. the Cretan city "Oaf o<?, i.e. pdpagos) tainty seems impossible, and the suggestion of Horn, rov'Ttavov,
is

probably the besl solution of the
8.

difficulty.

In N.
is

5.

32

rov he opydv kvi^ov aliruvoi \6yoi
restored.

the diqoI
five

gamma
appear
places p

probably to be

The word dpyd does
in
is

in
is

Homer.' hut occurs nine times
inadmissible, and once the word

Pindar.
at

In

the beginning

of a verse.

But

in

I.

6.

14 (toiulctlv opyaU) p
{roiaiai popyafc)
,

may
in

be restored
P.
4.

by

a

very slight change

and

141

dep-iao-apLevovs

popyds

may
is

be read without any change.
(>.

Beside

these nine places there
lrThe

the corrupt passage P.

50,

where the

quantity of the penult of 'IdXi-cros varies. In In B 656 it is long. Pind. 0. 7. 74 it occurs in Doric rhythm where we should expect a long syllable, but where, in each of the other four epodes, a trochee takes the place of a spondee, making this v apparently short. In Timocreon 1. 7 it is probably long. Pape

do poetical use of the word but 0. 7. 71. Ovid (Met. 7. 365) scans it as short, and so many Latin and English dictionaries, as well as Harper's Classical Diction2 Grundzuge 3 Heimer, ary. 'One- in 74 ]>. 76, Curtius, Grundziige 575. Bacchylides, at the beginning of a verse.
cites
.".
.

University of California
t

Publications.

clasb.phil.

best

eodd.

read
ig

rltf

T£k.4\v)(dov i

opyah

iraxravi^

though

-

required.

The

verse

1ms

consequently

been

emended
ma. and

in

manj ways, and
Curtius
1

yields no evidence as to the

digam(to

Bui

connected opyd with
logy

the
is

rool

varg

be

eager, press forward), and this etj
l>.\

accepted by Knd>.

Schroeder, the latter comparing the Bomeric Av/cov/070?

Av/cofopyos.

We

above secondary

have thus L3 instances of hiatus before the words in the list, most of which are most satisfactorily exIn

plained as due to the influence of the digamma.

several of
1

these, to be sure, the hiatus occurs after a dactylic thesis,
tion which, in
hint us.

a posi-

Homer,
even
if

is

considered bj

many
us
If,

scholars to justify
for the older

Bui

we accept

this explanation
to justify
4

poet, the evidence is too slight

in

extending the

application

of

the

principle to Pindar.

then,

we

include

these 13 places
a total of

among

the instances of apparent hiatus,

1T2 cases to be classed under this head.

we have The number

of instances of apparent hiatus in
If

T-fl

is

more than 400.
in

we examine

in detail the

usage of

Homer

contrast with

thai of Pindar,

we

find

no

difficulty in

understanding this great

disparity.

1.

More than 30
at all.
et'/ceXo?,

of

Homer's digammated words do not occur
,

in

Pindar

These are «8o? (abos) aXt?, apaios, apvos, eaw'?,
el\ap
J

eSaw'?, e#&),
eXu&) 7 eX&J/o,
crto?,

elXvoo, eYovca), 'E/ca/3?;,

e/cvpos,
ex?/?,
?>o-v/r,

e'Xo?,
eVeo-

evvvpa

(but
cf.

cf.

tad as P. 4. 253),
9.

eppco,

ev/crjXos

(but

€Ka\o$ 0.

58),

tJkkttos ,
tVe'r;,

qpa,

ripiov,

ihpoa), T/capto?,

IvSaWofiai, *Ip«,

itv?,

«o?/,

tW>;,

ov\os.

2.

A aumber
in

of

words

which

are

digammated

in

Homer

appear
these
e#vo<?,
eXeti>,
in

Pindar with no trace of an

initial

consonant.

Among
y

are

ayvvpt, dXia/copai, ciaTV, eap, eSvov, eepcra, €0€ipa
(et'Xo)),

eUoi (yield), elXeco

eipyw, e/m?, e/cacrTos, e\8op.ai,

eX<Wa>,

ef, ipvco, ta%(e)ft>, I'epai
cj.
2

(hasten),

k
9.

(but with f

P. 4.

253 by a
Sec

of Kayser), «£t-, ohos, 6^r.
oit. p.
.

iindziige L85.
1

0p.

142.

:;

So before 'IAXoos

in

O.

98;

P. 9. 79;

P.

1.

60;

I.

1.

16.

'

belo-w

p. 28.

vol. ij

Clapp. -Hiatus

in

Greek Melic Poetry.

9

3.

Even

in the

case of the 30 words or stems which
practice
is

show the
than
is

influence of the

digamma, Pindar's
while
iii

far

from uniform.

In tact he neglects the

he respects
felt

it,'

digamma in Eomerthe
it

these words more often
influence of the
is
2

digamma
Only
in

almost six times as often as
ol,

ueglected.

the

case of
in

mentioned above,

is

Pindar's usage overwhelmingly

favor of the

digamma.

8

We
lowing

pass

now
of

to the other melic writers,

including the

fol-

poets

whose works we have considerable portions

remaining.
Aleman
Alcaeus
250
160

w.

Simonides
Bacehylides

570
L350

w.

Sappho
Anaereon
If

240 250

Timotheus

275

we add

to these the

fragments of the less-known poets, we

obtain an amount of material not

much

less

than the extant

poems and fragments of Pindar. These poets differ widely in date and in dialect, from the Laconized Lydian (?) Aleman, of was th«- 7th century, to the Ionian Timotheus, whose Persians

Nor are all the fragments soon after the year 400.' included in our examination melic in character, since the fragments, especially of Anaereon and Simonides. include many
written

epigrams and elegiac verses. We shall take this element into consideration whenever any conclusions of importance seem to
be affected by
it.

shows the instances of the observance of the The first digamma, arranged according to the words or steins to which that consonant may be ascribed, together with the number of places in which the digamma is neglected in the use of the same
table

words.

The second

table includes the

same instances, arranged

according to the poets in
'

which the phenomenon appears.
eit.
:;.

See above,

p. 5.

-

See Hartel, op.

71.

3

See above,

p. 5.

'

So Wilam-

owitz, Perser p. 63.

LO

University of California Publicationt

f

ayvvfMi

Bapph.

•_'.

9.

Terp. 2; Sim.
inii s.

2»;
it;:;.

Alcm.
:;.

5.

6;

11:

69;

avSai

>'<

Alcm.
Alcae.

L8.

'-':

61.

L.

33

(see

Dryoff,
103.)
.".7.

in

Kuhn't

Zeitschrift 32,

tap
£17T0V

Alcm.

19.

3;

Sim.

Alcae. I9;8apph.23. 2; Bacch.9. 72

(KUTL
eXciv [?]

Alcm. H.

I*;

Bacch. Pr.

1.

7.

Vol.

l.

Clapp.

— Hiatus

hi

Greek Melic Poetry.
Lmes
he shortening of

1

I

failure to prevenl elision. Hour

t

in

i

a

vowel
in

or diphthong

.-it

the end of the previous word,

and three times

the occurrence of crasis.

L2

University of California

Publications.

[ci«abb.p

[f

we compare

these tables with the results of our observafacts are al

tions in Pindar, several

once apparent.
Pindar,
is

The
its

third

personal pronoun,
in

in

these poets as in
is

the one

word
pos-

which the digamma
05

consistently recognized.

Even

sessive derivative
initial
1,,

in

which Pindar more often ignores the
in

consonant, shows no exception
out
of
19

the other tnelic poets.
in

fact,

places

in

which these words appear
is

the

vmters under consideration there

bul

a

single instance of the

digamma ignored
Coriima
•">

\u>pav t air

eto? iracrav

wvvpr)vev
to
a<f>

where Rieister corrects

when we remember thai p in this pronoun is invariably observed in Pindar. AJcman. Aleaeus, Sappho, Stesichorus, Simonides, and Bacchylides, and once by Corinna herself, and
But
thai there
is

not, in all our extant melie poetry, a single instance

of the
fail

digamma ignored
conclude that

except

the present

one.
to a

we can hardly
corruption of
p.

to

inr

e«s here points

another kind, and

calls for

an emendation which shall restore

Nr\i come the words ava%, avctaaa), and the stems 18- (ot8a, which elSov), ipy-, (ip&-), ctt-(«V-) and aS-(aStfr, avhdvm)
, ,

show 23 instances

of p respected to 58 instances of f
in

neglected.

2

A few words appear with p
other melic poets, and
ri<-<

Pindar but without p

in the

versa, as

shown

in

the following table

vol.1.]

Glapp.— Hiatus
of

in

Greek Melic Poetry.
the above table

13

The Qumber

instances

in

is

probably too

small to allow any Lmportanl conclusions to be
[f

drawn from them.
pods, sepaas mighl

we

consider, now, the usage of the differenl
in

rately, as presented

the second table,
falls

we

find,

be

expected, thai

the

digamma

from century to century.
in

more and more into oblivion Terpander and AJcman are almost as
a

consistent as Ilonicrin their observance of this consonant, and

the case of
a

Alcman we have
adequate
a

sufficient

number

of verses to

afford

fairly

basis

for observation.

Aicaeus and

Sappho show
of the

decided falling
it.

often as they respect

off, and ignore the digamma as The Ionian Anacreon shows no trace

digamma

in

hiatus.

Simonides, the older contemporary
is

of Pindar, and originator of the epinician ode,

far less inclined

than Pindar to remember the digamma, which must be attributed
to Ins ('can birth,

and

also, perhaps, to the fact

thai

the

poems

of Simonides which
elegies.
It is

we possess are
if

for the most pari

epigrams or

not unlikely that
lind
is

of this poet

we should

in

we had complete epinician odes them many more traces of the
by what we see
of

digamma.
same

This opinion

supported

the

usage of Bacchylides.
dialect as his

observance of the
nearer to his greal
for this fact
ferent
in

The younger poet, though reared in the famous uncle, is much more consistent in his digamma, and stands in this respeel much
1

rival Pindar. It seems difficull any other way than by attributing it

to to

accounl
the dif-

poetical character of the extant

poems

of

Bacchylides,

which are epinician odes, with few epigrams or elegies. In the fifth century the digamma practically disappears from melic poetry, so far as our scanty fragments afford us evidence. Ii is
probable,
Pratinas.

however, that
Diagoras,
still

if

we possessed extensive remains of

Melanippides, Philoxenus, and the others.

we should
we

see sporadic traces of the influence of the almost

forgotten consonant.
find, in a

Even at the opening of the fourth century fragment of Timotheus, ctfcXea fe'pya, though the Persians adds nothing to our list. In Attic tragedy itself the

U'host of the

dmainma
"

walks."'

1

For installers of

false

digamma"
5<?

in Bi

hylides, see belo\<

,

p. 33.

-

Cf. Soph. Trach. 650 a

fot (pi\a 5dfj.ap.

See also Elec.

L96.

14

University of California Publications.

[class.phil.

Thedired
poets
is

effecl of

dialed upon the use of the digamma bythese

ii"t

so conspicuous as

we should expect.
all

This

is

probably

owing

to the

fad thai the language of

of them, notwithstand-

ing their differenl places of birth or of residence,

modified by

poetic

tradition.

is more or less The two Ceans, Simonides and

Bacchylides, used the
to us in

common
it

Lyric
is

forms which arc familiar
1

Boeotian Pindar, and
a

only when we reach Timotheus

thai

we find
is

language

free
list.

from Aeolic and Doric influence.

Of the older poets in our
and
his
it

A.nacreon writes the puresl Conic,

significanl thai in the 240 verses

which we have from

pen there appears to be no instance of hiatus before a digam-

maled stem.

II.

— Hiatus
a

after a

diphthong or long vowel.

Hiatus after

diphthong or long vowel, usually with the
is

metrical value of a short syllable,

very frequent in the melic
it

poets as well as in

Homer.

In

Homer, indeed,

occurs on every

page, and almost
license
this
is

in

every verse."

Pindar avails himself of this
in

much more sparingly than Homer, but even

Pindar

by far the most frequent variety of hiatus, occurring no

less than 212 times in the extant odes and fragments, or an The following average of almost six times to each 100 verses.

table records the instances in Pindar of hiatus after each diph-

thong and long vowel, omitting, of course, those cases which For convenhave already been noticed under apparent hiatus.
ience of reference the cases where the natural long quantity
is

retained are noted in a separate column.

Yet see Wilainowitz, Perser,

p. 39.

More exactly, aboul once

in

four verses.

Vol.

1.]

Clapp

.

Hiatus

in

Greek Melic Poetry.

|)II'IITIK)\(is

AN

I.

LiiMi \o\VKI.S IN

II

I

ATI

S.

II.

I'|N|,\I;.

METRICAL!/!

SHOE!

o.

1.

31

:

4.

23; 6. 92;

7.

7;

7.

55;
s.

7.

63; 8. 17: 8.47;8.
9.

69;

69;

14:

9.

23;
1

9.
1.

59; 9.82;

lit.

15; 10. 62;

19; 13. 7: 13. 84; 13. 107; P.
1.
1
:

I.

94;

I.

100;

'J.

51

:

3.90;4. 164;4. 174;
s. 9.

4.

194;

4.27)4: 4. 2712: S. 28; s. 56;

57 b

;

9. 22:

9.

37; 9. 40;

63; 9. 64; 9. 88; 9. 113;
17: 10. 22;
1.
lit.

10.

69; 11.9;
1
;

X.
3.

17:

1. :i2:
(il
;

2.

2.

3;
(i.

54; 3.
6.

4.

75; 5. 7:
7.

49;

54;

(J.

66;

101

:

10.

31; 1(1.47; 10.77: 11.2; 11.
7; 11. 23; I. 1. 2; 1. 48; 1.

57: 5.5; 5. 18;
8.

7. :!2: 8. 5:
7(i.

59; Fras- 1.2;
127.
1: 127.

1; 76.

1;

1:

143.

1:

1(19.

2; 169. 7; 199. 3.

79
74; 4.

-mi

(verbs)

O.

8.

53; 13. 99; P.

2.

273; 4.293:8. 93;
37;

9.

49:

9.
;

56; 9. 59; 12. 29; N. 3. 71
5.
7.
1(5:

11.

13;

I.

4.

68;
13:;.

Frag.
2.

123.

3:

131.

3;

is
2.

-fiat

(verbs)

O.

2.

92; G. 86; 8. 86; P.
4.

4;

N.

35; 5. 16; 9. 29: Frag.
9
1

107. 19: 123. 7.
-vrat.
Intiii.

(verbs)

X.

7. 20.

Hi

University oj California Publications.

j.phh.

I

>

I

l-U

<

TOWELS

IN

HIATUS.

'/.

PlNDAR.— Cont'd.

vol.i.]

Clapp. —Hiatus

in

Greek Melic Poetry.
hiatus,

17

Diphthongs and Long vowels

in

a.

Pindar.
NM.Ti;l.
\;

Cont'd.

METRICALLY SHORT

Dat. sing.

o.

7.

43; 37; 94;

s. 9;

8.
;

16;

13.
1. [.

30
16

o.

10.

25;
15;
I.

X.
1.

6.

22;
I.

13.
t.

P. 4. LM
6.

X.

10.
til

16;

26;
til
:

8.

23;
8

1. 8

1.

11: 5.

6.

Total -w

14

-a

18

University of California Publications.

class

ph

The following table shows the usage
Imt melic poets, excepl

of

Bacchylides and th

Pindar:
in

Diphthongs

ro

Long vowels

hiatus.

&.

Other Melic
\ll

poets.

!

N

SHOE

I'

I

811

ALLY LONG

Baeeh.
Hi:
i
1
.

3. 9.

18;

5.

3]

:

8. 2;

9

79;
11.

LO.

44:
13.

11.

24

66;
1

113;

30; 13
15.

188;
is.

1.

23; 15. 57;
is.

62

38;
1

53;
1.

19.

16;

Frag

60.
1:

:

Eum.
27.
2; 2.
1 :

2;

Alcm. 32
79.
1

Aleae.

66.
1
:

2;

Sapph.
sich.
90.
2:;.
:::
1
:

105. 3;
:

Ste

1.

1

23.

1

:

Anac
19. 2

96.
69.
!>4.

Simon.
3;

2: 80.

89.

::

91.

7:

3; 97. 2: lis. 5
1
:

120. 2:

124.

135. 9; 142

3; Tim..,.. 3.
2.

7.:

Prax.
28;

2.

3

3;

Philox.
1.

2.

2.

32

Telest.

7: Erin. 4. 2; 5

50

—tui (verbs)

Baeeh.

3.

S7
!H.

:

Hi. 6; Hi. 8
1
:

[

.']

:

Sapph.
119. 5;
1
:

90. 4: 95.

I'

:

Anae. Simon.

(is.

1
;

17. 1;

144.
3.
1

'J:

Timoth.

29.

Erin.

12

-/mi (verbs)

Baeeh.

5.

1!'.".:

Simon. 95. ID: Adesp.
1

1

:

Sapph. 2. 16; Timoth. Pers.
56.
1

5

\

-.a

i

\

erbs

I

Baeeh.

in.

:;:5

[.']

1

—at

(

inlin.

i

Aleae. 62.
Si

1

:

Sapph.
S

103 k;
3

mi m.

17)4.

at

i

1.

pi.

,)

Sapph.
Sapph.

26.

1

:

Timoe.

:;.

5

n-aAai

in::

,-

1

voi.. l.l

Glapp.

— Hitttits
in

in

Greek Melic Poetry.

1!)

Diphthongs and Long vowels

biatus

6.

Otheb melic

poets.
vl.1.1

Cont'd.
l.osi;

METRICAL!^ SHORT

MITI;|.

Nom.

pi.

Baceh. 17. 96; Sapph. 25. Aiuc. 97. 2; Simon. 09.
85. 4; 92.
33.
7
'J;

1

:

Baeeh.

11.

L20

[.']

in

92. 2;
1

Adesp.
9

[¥]; 85.
17.

Baech.
7i'.

115;

Baceh. Prag.
1 :

Sapph. 103

i>

[.']

3; Terp. 2.
1

Simon.

3.

3;
ti.

Hi.
;

1

1

:

165.1
5. 7:

:

1

Erin.

Melanip. Adesp. 1.
9

1

Baeeh.

1

1.

104; 11. lis

2

Total
-€l

-oi

20

University of California Publications.

'

i)

Long vowels

in

hiatus,

b.

Otheh

mi lic

poets.— Cont'd.

METRICALLY LONG

I

>:il

.

Sing.

Sapph. 103
SinH.ii.

k

:

Anac.
149.

102.
1

1

:

Simon.
\

22.

1

L26. 3;

Alcm.

1:

Total -a

4

-v

v ol.i.]

Glapp.— Hiatus

in

Greek Melic Poeti

.'/

Diphthongs and Long vowels in hiatus, in Pindar, Bacchylides, and the otheb welio poets.

University of California

Publications.

class phh

Diphthongs and Long vowels in hiatus, in Pindar, Bacch \NI> THE OTHER MEMO POETS.— Cont'd.

Vol..

l.i

Clapp.

— Hiatus

in

Ghreek Melic Poetry.

23

Diphthongs and Long towels in hiatus, in Pindar, Bacohylides, ami THE OTHEB MELIC >ETS.— Co nt'il.
l'<

Nom.

sing.

Doric gen.

Voe. sing.

Total

-V

Voe. sing.

Total
—O)

->/

24

University of California Publications.

[clasb.phil.

Among

the

cases of

oi

the

pronouns are mosl frequent.
the leading place, as
are practically
all
is

the

and the dat. sing, of pi. Under -a tin- word iirei has the The cases of -ov case in Bomer.

nom.

genitives singular, while those of
all

-<>>

are

all

datives singular, as arc
It

bul one of the cases of -a.
that

is

clearly the
in

rale

diphthongs and Long vowels are
of exceptions being bul 28, or

shortened
.ml.v

hiatus, the

number

one-fourteenth of the whole number.

This

fact

is

striking,
in

for in

Eomer diphthongs and
number
li\

Long vowels retain their quantity
1

hiatus about twice as often, proportionately, or in one seventh of
the whole
If

of instances.
first

we

our attention

on the diphthongs, we sec that
vowel with or v, form a group the conception occurs
i

the

first

five,

consisting of a short

group by themselves.
no
less

Within

this

than 289 times, while the natural quantity

is

retained but
List

nine times.

Comparing these

five

diphthongs with the whole

we see

thai five-sixths of the cases

where the rule

is

observed,

and only one-third of the exceptions, occur within this group. So far as the diphthongs are concerned, we find here a strong
i

confirmation of the opinion of Grulich,

2

that both the toleration

of hiatus, and the correction, are to be explained by assuming a

change of the vowel
initial

t

to the corresponding semi-vowel before the
3

Though jot had perished vowel of the following word. long before Pindar wrote, and probably before the time of any
felt.

of the poets
Its

under consideration, yet its influence could still be was sound would naturally emerge whenever an
t

spoken immediately before another vowel. This at once obviates the hiatus, and the remaining vowel of the diphthong, beine,- left
1>\

itself,

shows

its

natural quantity, which in the eases we are
is

discussing (at,
In the
in

ot, ei)

short
a

diphthongs consisting of
list

short vowel with

/

we

find

our whole

of melic poets but four exceptions to the law of

shortening.
Pind.
I.

These are:
8.
"><'>

doiSui ZXlttov

(

-

-

)

.

Tins has been emended by
(

Eermann
3

(7" eXtirov)

and Schroeder

-n

XCttov).

The

latter

emendation should probably he accepted.
2

WSee Grulich, op.
the correption of

cit. 20.

Op.

cit. 21.

Curtius, Studien

[.

2,

279 «., regards

-<*', -01,

as semi-elision.

vol.

i.i

Glapp.

— Hiat us
accepl

in

Greek

M<li<- Poetry.

i!-»

Baech.

11. 1120

-n-poyovoi iaaa/xivoL

(

- -

-

)

is

COmipt.
a parallel
bil

Kenyon and Smyth
1

the

ej.

of

Palmer eao-av e/W, Smyth
which
is

excusing the hiatus by Pind. 0.
ease.

6. 82,

hardly

This reading introduces an entirely unwarranted
history,

of

personal

and the
is

<\j.

of

IJIass

Trpoydvcou eaaap.ev(ov 1

approved by Wilamowitz,
Sappho 103 b
Corinna
4. i

to be preferred.
SoKe'ei fxoL opdvu),

ij/av-qv 8'

ov

and
,

iw«

fjpwiDv dperas

xvp <0°-° <ov are

i'"»

uncertain,

both

in

text

and rhythm,

to afford

any basis for discussion.

Grrulich extended the application of this

same principle
I).

to the
J

diphthong
was
in

—ot/.

Bui here, as pointed out by F.
since the -ov in question
is

Allen.

he

error,

always the so-called

from contraction.
final

"spurious" -ov (usually the ending of the gen. sing.), arising It was never a true diphthong, and hence its
vowel could scarcely have been changed to the correspondProfessor Allen himself suggested that the
fre-

ing semi-vowel.

quent occurrence of hiatus after this genitive ending must
tracted ending -oo.
it

he

explained as arising from an earlier habit of eliding the uneon-

Hiatus

one.- established

after this ending,
to the

would be easy and natural
the older form

to transfer

it

ending -ou,

when
for all

had passed out of use.

This would account
in

our instances of hiatus after -ov used as short (so irov

Pind. P. 4. 87), since in each of them
genitive ending of the

we have

to

do with the

-o- declension.
hiatus after -ov where the diphthong
little difficulty.
is

The
dy-yiara

five cases

of

treated as long offer

In Pind. X.

!•.

55 (tkottov

Pind.

I.

1.

was emended by Ahrens to ctkottoV dy^icna^ and in 10 '1 oXdov ivappco^ai the same emendation was made

by

Mommsen.

Both these emendations are generally accepted,
O. 13. 35).

since Pindar, unlike limner, does not hesitate to elide the final

vowel of -oio

(rf.

The same emendation
e'|

is

easily

made
and

in Pind.

Frag. 177. 4 aiviyfia irapdevov
84.

ciypidv yVaOwv,
ve(j>o<;

in

Simon.
69. 11

2

Kvdveov Oavdrov d/xepefSaXovro
0X1709, this,
in

(cf.

Simon.

Siotol

too,

in

an elegy).
is

The one
1.

remaining instance of -ov long
1

hiatus
2

Pratinas

L5

fjv

See below, p. 27. tions, Papers of the Am. School al Athens, I. 121.
7\u><r<rp a.K6vas
(

^-

— -).

Greek Versification

in inscrip-

•_><;

University of California Publications,

olass.phil

,',,,,,.

;,,,

a,,,

Sefta,

where the sense-pause after IBov makes the
those instances
of

hiatus objectionable.

W,.

have thus Ear considered
after a
v.

all

hiatus

which occur
lowed by
tify
i

or

diphthong consisting of a short vowel foland al the same time accounted for the quanThese number 298 oul of a total of 380 To these we long vowel or diphthong.
5,

of the syllable.
oi

hiatus after

a

may now add
a

the 12 (a

g
a

1,

r«>

6)

occurrences of hiatus after
i.

diphthong consisting of
is

long vowel followed by

where the
<

diphthong

treated as long.

Eere the transformation of

into

remaining long vowel a semi-vowel obviates the hiatus, and the Adding these 12 instances, we retains its natural quantity.
have
in
all

310
for.

<-ascs

und.-r

this

head which are satisfactorily
are

accounted

The 70 which remain

more troublesome.

We

shall first take
is

the diphthong

treated as short.

up the instances of hiatus after -co where These number 28, and are
14

distributed as follows:
Pindar Baechylides
2
1

Sappho
Simonides'

11

28

Here the law of Grulieh would account for the hiatus only, but
n,>t

for the curtailihenl of the quantity, since after the develop-

ment
.

of a semi-vowel from the

t,

a long vowel (to)
at

is left.

But
Gru-

(

V eii here

we ;uv
all

in

.1

altogether

loss.

As remarked above,'
-o

these are
lich

cases of the dat. sing, of the
8

declension.

himself suggested

that the original locative ending in -oi

may have had au
r

,,,

hiatus,
in

and

influence in bringing about the correption of that this locative F. I). Allen has pointed out
'

ending,

The certain dialects, did regular duty as a dative. the two cases was easj as we see from the tact that confusion of Sanskrit the Sanskrit locative ending in-/, and no1 the
.

it

was

dative in -e, which became the standard dative ending

in

Greek.

which was specially Simonides, Boeotian, could have affected such poets as Sappho, ma\ he doubtful, hut in the case of Simonides, and Baechylides,

How

far this dative (locative)

ending

oi,

i

Only in epigram or elegy.

z

p. 17.

3

Op.

cit. p. 44.

'

Op.

cit. p. 121.

vol.1.1

Olapp.

—Hiatus in

Greek Melic Poetry.

21

at

Least, we may take refuge in the facl thai the farther he was removed from Pindar's Boeotian, the nearer he stood to Homeric influence, and in Eomer he could find precedenl enough for the
<o.

correption of

Pindar, however, rejoiced
1

in

declaring his

independence of Eomer, ami

in

the Pindaric instances we prefer

to see the influence of the traditions of the poet's native speech.

Next come 14 occurrences of hiatus after
reption.

a

and

17,

with cor-

These appeal- as follows:
7
1
1

Pindar

(datives only
(dative) (dative)

1

Sappho Anacreon
Simonides

5

(4 datives, 1 subjv.

3d sing.)

Eere agaiD the Pindaric instances stand in a class by themselves. For the Boeotian dialed an older form in -at, of the dative ending of the
a

declension,

is

abundantly proved, and
influence

i1

isdifficull

to escape the conclusion that the existence of these two dative endings, in
to
at

and

in

a,

had

its

upon Pindar, and helped

shape his treatment of the dative ending in hiatus, viz., that 2 a is treated seven times as short and three times as long.

But the instances
(5),

in

Sappho

(

1

>

.

Anacreon

(

1

)

.

and Simonides

cannot

lie

explained by referring them to the peculiarities

Eere we are forced to fall hack upon of the Boeotian dialed. the influence of analogy,— the analogy of Homeric usage, which

was pervasive and powerful in all of the earlj Greek poetry. The following table shows the comparative frequency of m and
i
2

See especially X.
Since Grulich

7.

20

ff.

made no use

makes much

of tin- dative in-ot,

of the dative in -at in explaining hiatus, while he 1 add a Dumber of references, though the facts

are of course familiar to oiosl scholars.

See Meister, Gr. Dial.

1882,

pp.238

f.

'-'71:

Brugmann's Vergl. Oram. (Eng. Trans.) Gust. Meyer, Gr. Gram. 1886, p. 341; ["In Greek we find -ai in place of -at in the dative, a- we find Vol. :;. ri'. 147 f. at is certain for Boeotian, and so -ot, the locative ending, in place of -at.
. . .

it

was doubtless found

in

tin-

other dialect- which had -oi instead of -wi.»

Brug-

then speaks of the confusion which arose between the locative and the dative. and adds "After this, both classes of stems moved on side by side in tie- same direction. In one group of dialects, as in Ionic-Attic, (f and a absorbed -01 and -at forms in the declensions, so thai these survived only in adverbs and in certain fossil -ot and -at gained the day."] (e.g. oIkoi, Uty.iar, ei^s while elsewhere, as in Boeotian, p. p 98. and cf. Pindar's See also Brugmann's Gr. Oram 1900, p. 226, in discussing In Kuhner-Blass 1, p. 371, the dative in -at i- given place, though dative in 193 f.i Blass follows Grulich in taking accounl only of the hiatus,

mann

i

x^^l^^v

(p.

-ot.

and ignoring the dative

in -at.

28

University of California Publications.

LOlasb

Phil

ii

hiatus, used as

long or as shori respectively,

in

Eomer,

Pindar, Simonides, Bacchylides, and the other melic writers:

vol.

l.]

Clapp.

— Hiatus

in

Greek Melie

"Poetry.

29

one (Find.

I.

7. 8)

;m anacrusis.
in

Ten instances are
the thesis of the
thesis of
a a

in

dactylo-

epitritic verse, of

which four arc
in

dactyl,

two

in

the thesis of a

spondee, two

trochee,

two
the
in
It

(Pind.

I.

1.

16,

Bacch.

11.

120)

in

anacrusis.

Three instances

are in dactylic hexameters, of which
third foot

two are

in the thesis of

and one
1.

in

the bhesis of the second foot.
is

Finally,
in

Pratinas

15,

the third syllable of a cretic

long

hiatus.

thus appears that out of 25 instances of the kind we are discussing, only 12 are
in in

the thesis of a dactyl, the remaining cases
in
at

being found

almost every possible metrical position, even
a

the unaccented part of

foot,

or in anacrusis.

The
is

efforl

explanation, then, on the basis of metrical position,
successful than
that

aol

more
in

which seeks the excuse for hiatus

the

nature of the vowels on diphthongs concerned.

In either case

we must be content
But
in

to find a

residuum of phenomena which can
|

be explained only by analogy, or as instances of

tic license.

general
1

the present

writer

inclines
in

to

the opinion of
in

Mommsen

"Tota de hiatu quaestio non
It

numerorum sed
difficult to

vocabulorum natura vcrtitur."

would not be

show
upon

ground for the

belief that far too
in

much

stress has been laid

the effort of verse-position

mitigating hiatus even in Homer.

It

may

be profitable, here, to recapitulate the results of the

preceding discussion of hiatus after a diphthong or long vowel:

Short, 254 instances, to be explained as

by Grulieh and Hartel.

Long, 4 instances, to be emended.

"_'.

After

ev. ou.
:>.">

Short,

instances, mostly genitives, to be explained as due to

the "],icr ending -oo elided.

Long, 5 instances, of which

4 arc genitives to

be

amended

to

—ol.

3.

After «, a,
Short,
\-

77.

instances, mostly datives, probably influenced by old

dative endings in -01 and —at.

Long, 12 instance-,
Supplement

to

be explained as by Grulieh.

p.

1

T

University of California

Publications.

i<

I.

Alt. r a
Short,
-']

instances.
instances.

Long.

7

Total

Short,

Long,
well

It

is

known
in

thai the dactylic fool

is

the natural
tin-

home

of

conception
foot,

in

biatus.

Tin'

following table shows
in

kind of

and place
in

the foot,

which these shortened syllables

appear

the Melic ports:

vol.i.]

Clapp—Hiatus

in

Greek Melic Poetry
dactylic

31

This necessity for the rapid swing of
still

tli<'

movemenl

is

further

shown by the
in

faci

thai

almost

three-fourths of

all

our instances occur
acceleration
of the
to

the third syllable of the dactyl, where the
its

has gained

full

headway.
is

It

is

onlj

in

the ruse

diphthong
of

at thai

there

any approach to
the dacty]
in

equalitj
this

(68

1-1)

between the two shorts of
-ai,

respect.

The correption
times),
set mi is

especially in

the conjunction ical (130
it

to

have been so well established thai
in

could occur

almost as easily

the second syllable of

a

dactyl as in the third.
in

Bui with the other vowels and diphthongs
parity
is

our

list

the dis-

over-whelming (23 times

in

the second syllable of the
a difference

dactyl, 126 times in the third).

Such

can hardly be

accidental.

The few instances
trochee,

of correption
us,

in

a

tribrach,
license

or resolved
is

need

not

detain

since

the

generally

admitted to be allowable under such circumstances.
noted
P.
11.
in
!).

The cases
But
in

cretics

are

more or
fool
is

less

uncertain, especially Pind.
a

where the
it

perhaps
doubtful

tribraeh.

the

trochee

proper

is

very
at
all.

whether

the

curtailmenl
are
as

should be
follows:

admitted

The apparent

instances

Pind. O. 14.

1

Katpuriwv vSdrwv Xa^oiaai cure
ebpav
(

vat'ere

kuAAiVwAov

— — — — — — ~ — — — — — — — —

).

Here the trochee (-olaai) seems to be proved by the corresponding syllables
notoriously
(Xaxolcrav)
in

the

antistrophe

(-T/crt'-)

.

But
injure
it

this

ode

is

corrupt,

and
hiatus,

the

slight

emendation
not

of

Boeckh
sense.

avoids

and

does

the

Schroeder's rat re for aire seems less good, sine - - for - ~.
Pind. P.
8.

introduces

96

0"/aas

ovap

avOpuy—oi.
).

aXX' orav

.

.

(

-

-

=

Here the reading avdpwrros. found

in Pint.
l>\

Cons. Apoll.
editors.

(>.

is

far

more

poetic,

and

is

generally adopted
43.

modern

Bacch.

10. 33

ami

These two cases are peculiar.

They occur

in in

a

short

ode,
is

in

which there are but two

triads,

and the verse

question

the

32

University of California

Publications.

[ci

fifth

of the strophe (antistrophe), and
in
•">•
t

hence this metrical series

occurs

he

poem four
vtavrai
(

t

imes,
)

(...)
i'xtitu

w

15.

yry

SiKus

(kiltl

avdtviv £uv
ap.cpi

33.

»< r« 11

eAAav av vefiovrcu
ol 8'

t

Yivfioi-

43.

ttoikiXov to$ov Tiraivei-

eV

epyoi

In

15 the
art

thai

marked hiatus after e/cari. together with the fad makes a trochee where we should expect spondee,
;i

subjects the text to well-deserved, suspicion.
far better, with Blass
jt

Consequently

it

is
;is

and Jebb,

to alter the division of verses

appears
the

in

the papyrus,

and end the verse with
of
(in

wart,.

This

avoids

hiatus,
at

secures the succession

regular dactylorelieves us of

epitrites,

and

the

same time

.'{.'J

and 43)

two

of our eases of eorreption in a trochaic foot.
Bacch.
16.

20

Kopa t

o/3/jf./x.oSepxet

a£,vya

This

series
in

of

quantities

occurs only once

elsewhere

in

the

poem,
this

the mutilated verse 8,

which closes -tcu
the fact

fjovwv.

In

case

Mr. Kenyon's only reason for making the syllable
short
is

before

the hiatus
in

(apparently)

that

hiatus

occurs
in a

both verses.
is

But we have already seen that eorreption
the retention of the long
of Blass
(

trochee

much more unusual than

quantity

in hiatus.

Hence the scansion

- —

)

is

to be preferred.

III.— Hiatus after a short vowel.
Hiatus after
poetry.
a

short vowel

is

very
1
,

The only instances which appears
digamma.

Pindar have already been noticed
a

suspicion of the

A

in Greek melic modern texts of as affording room for at least small number of cases in
in

uncommon

Bacchylides and the melic fragments, must
Bacch.
2. 7

now
<i.

be mentioned.

a.v\ivi

'laOpov
p.

For puTOfios, see above,
Bacch.
5.

75

i^eLXcro iov
(f>peva tuvOeis

17. 13]

See above pp.

5

fit.

vol.

1.1

Clapp.

— Hiatus in

Greek Melic Poetry.

33

These seem

to

be instances of "false

digamma."

We

have no

evidence thai cither uk (arrow), or

laivco,

was ever digammated,

but both words suggesl to the ear the familiar lov (violet), which

has the

digamma

in

Eomer, Alcman, Ibycus, Simonides, Pindar,
1

and three times
Baech.
3
(>4

in

Bacchylides.

/xey(iivr]Te 'Iepa>v

This

is

a

difficull case,

but

is

eased somewhat U\
is

the fad thai
1>.\

the final
ictus.

vowel of fieyaivrjTe

apparently lengthened
36,

the

Cf.

w.

4, 8,

18,

22,

32,

46,

50,

60,
is

78,

88,

of the

same poem, where the corresponding
an.l

syllable

long.

Yv. 74

92 are mutilated.
Baech!
16. 5

dv^e/mdevrt "E/ipuj

Sapph.

103. 2
'>

ovkIti

d^w

1

Sim. 22.

8ei'/i.aTi

yjpnrev

22. 8

vvktI aXafjLirel

Philox.

2. 2

^XP

L

°v

There are

in

Homer

a

few instances of hiatus, not otherwise
t,

explained, after the vowel

and on these has been based
In

a

law

permitting hiatus after this vowel, though the evidence for such
a

law

is

not
t

entireh
in
is

adequate.

the examples
in a

before

as,

however, the
tion,

most eases occurs
vers seldom elided.
'

word, or

in a

termina-

where
4

it

and hence the hiatus may

be justified.
Timoth. Pers. lis
(f)epufxe$

a

ov

.

.

.

Here the text
the sense

is

not

quite certain,

bul

the

marked pause

in

makes

tic hiatus unobjectionable.

'See above,
in

his odes in

which Pindar N supposed scholiasts and editors ..li
of this.

Did Bacchylides, as a ('can. learn his digamma for use p. 10. Pindaric style, and have we here an instance of tin- diSaKTai dperai Gossiping t" scorn, as contrasted with hi- own tpvdf
1

>.

-'.

Mi

ff.,

<

>.

'.'.

I""

il..

N.

3.

il

i..

could have

made much

-Smyth suspects
3
4

the

digamma

here.

So especially
For Baech.

in the
!.">

dative singular.

Sec pp.

6

t'.

lit.

eVari dvtieaiv see above

p. 32.

:;i

University

oj

California

Publication*.

Conclusion.
In
,-i

number

of places when- hiatus apparently occurs, the two
to
!>«•

vowels or
(synizesis).

diphthongs are

pronounced
a

as

one
deal,

syllable

On

this point

editors differ

good

bu1

the

following instances seem reasonabl} certain.
Pind. O.
13.
7
!»!»

Tafiiai av8pa.cn

(Schr. rafii')

13.

oq dfitftorepiadev
(Ithi et (().

P.

11.
5.
1.

55 50
It

Hermann

i

Alcm.
Sapph.

V 0V X
uipdvoy

aWtpos
e/u<u

1.17
n'M.
1

KWTTL

Keicrem ovSeirora

84.

3

eyw ovSe

Anae.

32.

6

pq

avafi-qvai

67
90.
1

<£<A«0) Ot''T£

(ptXitD
ju.*/

o's

Sim.

:!.

5
1

ou

57.

kAvto, eapos
aw/ w/jttv
r/

Tin ocr.
i

1

12

Ariphr.

1.

6

el

Where
union
is

a

vowel has already been elided from the
a

first

word,

the two words arc to

certain degree
effected

united into one.
1>\

This

not so complete as thai

crasis, bul
fell
dtjei'

seems to
the elided

have been

sufficienl to prevent hiatus

being
2. 41

when

word
uases,

still

ended

in a

vowel, as
1
,

in

0.

'EpiviK.

These

which are not infrequent

require no discussion.

Some

86 in melie poetry,

somewhat more frequenl

in

Homer.

UNIVERSITY OF CALIFORNIA

PUBLICATIONS

CLASSICAL PHILOLOGY
Vol.
1,

No.

2, pp.

35-94

January 31, 1905

STUDIES IN THE SI-CLAUSE.
II.

C.

BY NUTTING.

I.— CONCESSIVE SI-CLAUSES IN PLAUTUS.
In general the hypotaetic concessive period
as a

may

be defined

complex sentence which brings together clauses of such a
in the conclusion

nature that the assertion

might naturally seem

to the hearer to be incompatible
to in the concessive clause;
t

with the state of affairs referred

.</.,

Rud. 1353
Si

ft'.;

immune

mihi ilium reddiderit viduluni.
hodie debeo triobolnni.
si

Xon ego

illic

Among
its

the concessive periods of Plautus introduced by
is

and

compounds there
differentiate this
I

a large
.

and striking

class distinguished

Prom the others by the gradi so

to speak, of the concessive clause.

To

group from what
suggest the

may

be styled the simple

(or normal) type

name "intensive."
it

The simple
is

type of concessive clause (as distinguished from the intensive
is

characterized by the fad that
the situation
thai
as.

goes no further than
a state

de-

manded by

it

simply recognizes
in

of affairs
to the

(real or supposed)

has

some way been suggested

mind of the speaker;
Men. 746
Si
IV.
;

for instance.

me

derides, at pol ilium

mm

potes,

Patrem meum.
Ps. 290
ft'.

Egon

patri subrupere possim quicquam, tarn canto seni
si

.

Atque adeo.

facere possim. pietas prohibet.

36

University

<>/'

California Publications.

[Class.Phil.

hi neither of these passages does the concessive clause exceed the

demands
to

of the situation.

In

the

firsl

case the speaker refers

au obvious fad wheu she says Si nn derides; for Menaechmus
in
a

has been treating her
thf other the phrase
bility.

manner anything

bu1
a

respectful.

In

si

facen possim takes up

supposed

possi-

The intensive concessive clause ou the other hand purposely
exaggerates the state of affairs suggested
Asin. 403
IT.:

to the

speaker,

e.g.,

LI.

Atque hercle ipsum adeo contuor: quassanti capite
cedit.
irato, vapulabit.

in

Quisque obviarn huic occesserit

ME. Siquidem
cedit,
Si
I

hercle

Aeacidinis

minis animisqui

expletus

iratus tetigerit, iratus vapulabit.

Tn this passage the mercator might have been content to confine

himself to the reported

fact,

thus producing a simple concessive

"Though he conies on in anger, he will get a beating if he touches me." But this is too lame an expression for his emphatic
period
in
I.

and he

flies to

the extreme of the improbable or impossible
filled

— though
I

Leonida comes on (not merely angry but)

with

In

boldness and couragt of Achilles, he will get a beating.
is

An .

nlher ease of the same kind
Tri. L184
IV.
:

('II.

Quamquam

tibi

suscensui,

Miseria una

(i.e.,

one wife) nni quidem hominisl adfatim.

CA. Immo huic parumst.

Nam
is

si

pro peccatis centum ducal uxoris, parumst.
line.

Here centum very obviously caps una of the preceding
this gratuitous exaggeration that
is

It

the characteristic feature

of the intensive type.

Concessive periods belonging to this cate-

gory are generally easily recognized when once the peculiarity of the type has been noted, though of course occasionally sentences are met with which are hard to classify.

stylistic

The intensive concessive period and the syntactical point of
in

is

interesting from both the
It

view.

is a

form of sp
in

-It

common

dialogue,

its

most distinctive use being

emphatic

Vol. i]

Nutting.

Studies in

tht

Si-clause.

37

rejoinder.

Willi Plautus

it

is

a

distind mannerism; aboul onein

third of

;ill

the concessive clauses

Ins plays

introduced by

si

and

its

compounds belong

to this class.

The present paper deals
In

particularly

with the syntactical aspects of the case.

the

pages immediately following, the concessive periods introduced
by si and each of its compounds are presented separately., the sentences Palling within the several groups being examined with
reference to the distinction just

made

of simple and intensive.
in

Some

points of minor interest are noted

passing, bu1 the

more

important

questions suggested by the syntactical

form of the
a1

intensive concessive period are reserved for discussion
after
all

the end,

the material has been presented.
SI.

It is quite impossible to

determine the exad number of conIn

cessive si-clauses

in

Plautus.

many

cases the

oature of a

clause depends
is

upon the point of view of the speaker, and there
by which to settle the question definitely.
I

no objective

test

Ex-

cluding the more doubtful examples,

still

find 88 s*-clauses that

seem

to deserve the
3

of Kriege,

mime concessive. This exceeds who puts the number at 66.

the estimate

A.— Simple.
Amph. 908;
Si dixi, nilo
('apt.

mains

es

neque ego

esse arbitror.

12;
est, est

Si

iidii

nbi sedeas locus
ff.

nbi ambules.

Cist. 27 si

idem istud nos faciamus
vivimus

si

idem imitemur,

ita

tamen vix

(

'nm invidia summa.
.Mil.

63]

:

Si albicapillus hie videtur, ne

utiquam ab ingeniost

setiex.

Most. 42

ft'.:

Xon omnes possunt
Si in oles.

olere

unguenta exotica.

Rud.

14(10;

Xon

hercle istoc

me

intervortes,

si

aliam praedam perdidi.
et

])c cimnti.-itis

eoncessivis

apud Plautum

Terentium. Halle, L884,

p.

•).

University of California Publications.

[Class. Phil.

Tri.

185

ff.;

Semper

in hoc facito, Lesbonice, eogites,

M

optumum
Tri.
.".<

esse tute uti sis
u1

optumus.
sis

si id

nequeas, saltern
17

optimis

proxumus.

ff.;

Scd

si

haec res graviter cecidil stultitia mea,

Philto, esl ager

sub urbe hie nobis.

True. 854
Blitea
Si alia
e1

ff.;

luteasl meretrix nisi

quae

sapil in vino
si1

ad rem suam

membra

vino madeant, cor

saltern sobrium.
17!'.

For other
365, 887. 1013

cases see
ff.,

Asm. 603
ff.,

ff.,

933, Aul. 254, Bacch.
ff.,

1193
ff.,

Capt. 223

683

ff.,

742

ff.,

Cas. 298,
ff..

314

ff.,

Cist. 67,

152

Ep. 599, Men. 670, 746, Merc. 636, 819
fit'..

Mil. 298,

306

ff.,

747. Most. 914, Poen. 51, 374. Ps. 290
ff.,

Rud.

159,
ff.,

HU4. 1075, 1353

St.
18.

43

it. Tri.

85

ft'..

465, 607, True, 66

615, 833, S77.
I'

Total.

these simple concessive periods are a mere optional Son form of expression for a thought that might have been conveyed by two coordinate clauses joined by an adversative conjunction.

Such

a

case

is

.Mil.

631 (quoted above in full)

;

there the speaker,
in the fol-

had he so elected, might have expressed his thought
low in g form
:

"He

Looks gray, but in spirit he
striking, and. at
a
first

is

by no means

old.*'

A more

sight,

apparently unwarranted

use of the form of

hypothetical concessive period appears in

passages like
True. 613
ff.;

STR. Verbum unum adde
offatim offigam.

istoc:

iam hercle ego

te

hie

hac

CV. Tange modo: iam ego
distruncabo.
Si in !n
I, lit

te

hi.-

agnum faciam
ego
in

e1

medium

ad legionem bellator dues,
line the
(il

.-n

culina clueo.
is

this lasi

form of the
in

lirst

clause
at

easily justified,

the
t

words

ego

CUlina Clueo, taken
a

their face value.

,l<

complete the meaning of

concessive period.
a

There

is.

it

ls

true, an antithesis

between the two clauses; bu1
re

genuine con-

cessive period

involves something

than mere antithesis

Vol.

l]

Nutting.

— Studies in the Si-clause.
is

39

there

is

an incompatibility between the subjecl matter of the two
surprised
at

clauses such thai the hearer

the statemenl

in

the

conclusion; for the state of affairs here
rally

ntioned would natuin

seem

to be

precluded by thai referred to
lirst

the concessive

clause: as in the typical sentence

quoted,

Rud. 1353

ft'.:

Si niaxunie milii ilium reddideril

vidulum,

Non ego
In

i

1

1

it-

hodie debeo triobolum

the sentence

under discussion, as

it

stiinds.

this

element of

incompatibility appears to be Lacking; whatever the amounl of

warlike fame possessed by Stratophanes, there
ever surprising in the claim of

is

nothing whata

Cuamus

that

lie

is

famous per-

former
If

in the kitchen.

we must take

the words at

<

<i<>

in

culina clueo

at

then- bare

face value, the probable explanation of a sentence of this sort
is

that the line between simple antithesis
is

and
;

antithesis with

incompatibility
occasionally

not always sharply
that clauses
in the

drawn
a

in this

way

it

might

happen

which were merely antithetical
hypotactic concessive senin

would he strung along
tence.
<>n the other

form of
it

hand,

is

quite possible thai
is

the con-

clusion of a sentence like True. 615 the speaker

not expressing
all

himself fully, and that the underlying thought contains

the

elements of
ing
in this

a

genuine concessive period.

For

instance, the mean-

particular case might be

"Though you

are famed for
I

valor in the army, (you need not try to frighten me. for)
a

am

famous performer
passage quoted,

in the

kitchen."

In the line that precedes

th."

Cumamus

has shown that his performances
to

in
till

the

kitchm include the handling of knives, thus helping us
(if

out what
in
is

this interpretation

be correct) he leaves unexis

pressed
one,

615.

This second explanation

a

very attractive

and

the more justified because such abbreviation in verbal
rarity in Language generally.
2

expression as

is here assumed is no With True, bio may be compared

Bacch. 364

ft

.

Si ero reprehensus,
si
illi

macto ego ilium infortunio:
at

sunt vireae ruri,

mihi tergum domist.

'American Journal of Philology, XXIV. p. 294. Cf. Lindskog, De enutiatus apud Plautum et Terentium eondicionalibus, Lundae, 1895, p. 103 ff.

in

University of California Publications.
Bacch. 885

[Class.Phil.

If.;

E1 ego te
Si tibisl

e1

ille

Quid ilium morte territas? mactamus inforl iinio.
a1

tnachaera,
I:
isti

aobis veruinasl domi.

Rud. LO]
si

in proreta

navi's, ego

gubernator

ero.

B.— Intensive.
The
this
iikisI

striking thing aboul
is

the examples thai

fall

under

heading

that, in

more than half of the

eases, the intensive

force centers

around some other won! (or phrase) than the

verb.

As

in

the following;
Asin. 413
II'.:

LI.
LI-!.

Bic me moratust.

Siquidem hercle nunc
is

summum

lovem

te dicas detinuisse

Atque

preeator adsiet,
ff.;

malam rem

effugies

numquam.

Aul. 98

Profecto

in

aedes meas

me absente neminem
tibi

Volo intromitti.
Si

Atque etiam hoe praedico
veniat, ne intromiseris.

Bona Fortuna
Aul. 555
IV.

Quos
Is

si

Argus

servel qui oculeus tutus

t'uit.

Quern quondam Ioni Euno custodem addidit,

numquam
si

servet.

Baech. 128;

Qui

decern habeas Linguas,

mutum

esse addecet.

Baceh. 697;

Quern

si

orem
;

at

mihi

nil

eredat, id nun ausil credere.

Men. 75]

[dem hercle dicam,
Mil. 803
IV.-.

si

avom
si

vis

adducere.

Non

potuil

reperire,

ipsi

Lepidioris duas ad hanc rem

SoU quaerendas dares quam ego.
IV..

other similar cases are Amph. L048
Bacch. 1045
IV..

Asin. 318
IV..

IV..

405

ft'..

1102
IV..

IV..

('as.

93

IV..

Cist. 3
BE.,

Cure
IV..

211, Men.
St.

238

ft'..

.Mere.

838
ff.,

Mil. 188, Most. 115
rr.
:

912
ff.

Rud. 1361,
2.".

287, Tri. 884

962, lis;,

cf.

True. 527

Total.

cases.

Vol. i]

Nutting.

Studies

in thi

Si-clause.

41

Tn these sentences the fad

thai

the intensive

force centers

about some other word or words than the verb affords an interesting illustration of the general principle thai i1 is no1 always
the verb thai
clause.
is

the essential

and characteristic feature of

a

si-

As

a

matter of

fact, in

some of the above examples

all

other elements of the concessive clause are so unimportanl that,

without

loss

to the sense,

they could drop away, leaving the

phrase about which the intensive force centers to be incorporated
in the

conclusion.

<

.</..

Aul. 100;
Si

Bona Fortuna
lines

veniat, ne intromiseris.
in

In the during

which precede

this passage the

speaker has been

giving general directions that no visitor be admitted to the house
his absence.

He would

therefore have been perfectly well

understood had he said simply, Xe
seris,
i.e.,

Bonam Fortunam

intromi-

"Don't

let

even Good Fortune in." 3

In this connection, as also showing the importance of the role played
in

the concessive clause by the words about

which the

intensive force centers, should be mentioned sentences such as

Amph.
Quin
sie

1051

ft'.:

Neque me luppiter neque
faciam
:

di

omnes

id

prohibebunt,

si

volent,

uti constitui.

.Most. 351

Nee Suhis nobis
It will be seen
at

saluti

iam

esse,

si

cupiat, potest.
all

once that each of these sentences contains

the elements that go to
like

make up an intensive concessive period those under discussion. But the elements are different ly

arranged here
the

— the

st-clause

comes

late in the sentence, leaving
1

words about which the

intensive

force centers in
si

a

natural

emphatic position. 4

As

the sentences stand,

volent
it

and

si

cupiat are not only not of the intensive type, but

may

even be

3

ology,

This matter is more fully discussed XXI, p. 260 ff.

in

the

American Journal of
Capt. 529,
Cas.
324.

Phil-

Asin. 153

'Other examples may be found at ff., 237, 894 ff.

Aul.

rill.

Cf.

42

University of California Publications.
8

[Class.Phil.

questioned whetherthej are concessive
to

a1

all.

Yei
(e.g.)

we have only
Most. 351
in

rearrange the elements thai go to make up
,-i

such

way

thai

the

word aboul which the intensive force cen-

ters shall

fall

within the si-clause, to produce an intensive con-

cessive period exactly like those with which the discussion started 8i Sal us nobis saluti esst
see

cupiat,
in

etc'"'

It

is

therefore easy to

how important

a

factor

the concessive clauses of the type

under discussion are the words aboul the intensive force centers. 7 The remaiuine concessive sentences of the intensive type are

Amph. 450
L060
87,
if..

IV..

Bacch. 1004, Cure. 3
IV..

IV.,

449

IV.,

Ep. 610
IV..

if..

Men.
Ps.

Merc. 694
ft'..

.Most.

229

IV..

241, Pers. 40

282

BE.,

265

ff., 7!)l>

True. 315

ff.; cf.

Merc. 595

ff.

Total, 15 cases.

Here the
hut
it

intensive force tends to £Ta vitate toward the verb. oters exclusively
a1

seldon

thai
e.g.,

point

;

more often

it

is

diffused throughout the whole clause;

Amph. 450 (V.: ME. Quo te agis?
Atque nine
Ps.

SO.

Domum.

ME. Quadrigas

si

nunc

inscendas lovis
Eugias,
ff.;

ita

vi\ poteris effugere infortunium.

264

PS. Putin
Atqut

ut

seme! modo, Ballio, hue
istoc pretio:

cum

lucro respieias?

BA. Respiciam
in

nam
ui

si

sacruficem

summo

Iovi

manibus exta teneam

poriciam, interea loci

Si lucri
Tliis last
is

ipud dot ur, potius rem divinam deseram.
a

very striking case.

Ballio has

up
;it

to tins

time de-

clined to parley on the plea of business.

But

the magic word

pretium he
stop even
if

is

ready not only

to

forego business, but

he would

he were sacrificing—and thai too to mighty Jove.

and

at the very critical point of the sacrifice; each of these speci-

fications contributes to the intensive force.

'Kriege (I.e.) includes such sentences withoul commenl as concessive. But it may 1"' noted thai Plautus never uses the (distinctively concessive) compounds of si, e.g., etiam si or tametsi in such a case, though he 'Ices employ these comj nds when the sentence is so arranged thai the words
:il. mt which the intensive force renters fall within the limits of the subordinate clause. 'Such a case occurs in Ter. A del. Till ff. •In this connection it should perhaps be further noted thai in a few intensive concessive periods the emphatic words or a substitute appear also in the conclusion; e.g.. St. i>s7; si re.r obstabil ohviam, re gem ipsum prius

pervortito.

Vol. l]

Nutting.

— Studies in the Si-clause.
ETSI.
A.
26 cases.

43

— Simple.
under
this

In the examples that
the sentence
is

fall

heading the nature of
it

generally so evidenl thai

will

be sufficient

to

quote only the etai-clauses, omitting the conclusions.
Aul. 42]
;

etsi taceas.
;

Bacch. 1160

etsi

.

.

.

prope scire puto me.

Bacch. 1191;
Capt. 543
IV.
;

etsist dedecori.
etsi

ego domi Liber

I'm,

Tu

.

.

.

servitutem

servisti.

Capt. 744; Capt. 842;
Cas. 958
:

etsi alitor ut
etsi nil scio

dicam meres.

quod gaudeam.

etsi

malum

merui.

Mil. 407: etsi vidi.
.Mil.

532;
(i(i(i;

etsi east. etsi etsi

.Most.

procul abest. 8

.Most.

854;

non metuendast.
properas.

Pers. 272; etsi

Pers. 601

ft'.;

etsi

mi hi Dixit

.

.

.

Pers. 655; etsi res sunt fractae.

Poen. 1084;
Ps. 1113
;

etsi hie habitabit.

etsi abest.

Rud. 1044; Rud. 1350
Tri.
:

etsi ignotust.

etsi

tu fidem servaveris.

Tri. 383; etsi advorsatus tibi fui.

474

;

etsi votet.

Tri. 5:27: etsi scelestus est.

Tri. 593

ft'.;

etsi

admodum

Tu ambiguosl

.

.

.

Tri. 600; etsi odi

hane domum.

True. 815

:

etsi

tu taeeas.

B.

— Intensive.
to

There remain bu1 two cases
belong
to the

come under
is

this
i.e.,

head: both
the verb
is

second type of Lntensives described,

the center of intensity or else the intensity

distributed through-

out the elause.
s

In the edition of Goetz and Schoel] this line

is

placed between 609

am

cm.

44

University of California Publications.
Capt. 854

1

Phil.

I

IV.

HE. Nee

nil

hodie nee multo plus tu hie edes, Qe frustra
victi
ut

sis-

:

Proin in tui cottidiani

ventrem ad me adferas.
tute cupias facere

ERG. Quin
Vid. 106

ita

faciam,

sumptum,

etsi

ego vetem.
IV.:

tnalo
I't

semper piscetur,

etsi sit

hunc adligari ad horiam tempestas maxima.
t

In passing, the

exceeding brevity of the

tsi-clsmse

may

be aoted;

20 of the 26 clauses do not exceed four words each.

TAMETSI.
under

16 cases.

A. — Simple.

For the cases

thai

fall

this

heading the material may

be presented in the same

way

as for etsi.
.
. .

Amph.
Anijili.

21

IV.

:

tametsi

Scibat.

977; tametsi praesens aon ades.

Aul. 768; tarn etsi 9 fur mihi's.
(

!apt. 321

;

tametsi tmicus sum.
tain etsi

Chic 259;

non novi.

Cure. 504; tarn

etsi nil fecit.

Mil. 744: tarn etsi

Pers.

:><!l'

-.

tarn etsi id
etsi

dominus non invitus patitur. futurum non est.
in

Poen. 342: tarn

abstruso

sitast.

Poen. 1201; tametsi snnins servae.
l's.

244: tametsi occupatu's.
:

Ps. 471
St. 41
St.
:

tarn etsi
etsi's

til>

;

suscenseo.

1am
-.

maior.
.

205

tain etsi hercle

.

.

iudico.

B.— Intensive.
Men. 92;

Numquam

hercle effugiet, tarn

etsi

capital fecerit.

Tri. 679;

Facilesl inventn

:

datur

ignis, tarn etsi

ab inimico petas.

Tametsi is here written aa one word or two, according to the reading of the Goetz-Schoell edition.

Vol.1]

Nutting.

Studies intfu Si-clause.

45

Though
the smaller

the

number of intensive cases is the same as for etsi. sum total renders the proportion Larger. There is
firsl

also the further difference that these cases are of the type

described

the intensive force centers aboul

some other word or
examples do not

words

Hum

the verb.

Though

not so striking, the brevity of the

hum

tsi-clause also deserves notice; 10 of the 16

exceed Four words.

ETIAM
Ep. 518
11'.:

SI.

-1

cases.

10

iniiiio

etiam

si

alterum

Tantum perdundumst, perdam potius quam sinam

Me inpune
Ps. 626

irrisum
IV.:

esse.

PS. Milii hercle vero,
Ballionis euro,

<iui res

rationesque

eri

HA.

Si

argentum aecepto et quoi debel dato. quidem hercle etiam supremi promptas thensauros

lovis

Tibi libellam argenti

numquam

credam.
first

Both these cases are intensives of the
force centers elsewhere than
the resolution
si
.
.

type

—the
el

intensive

around the verb.
is

In the second case
. .

.

.

etiam
in

precisely parallel to

.

kcu

and
first

''If

.

.

even ;"

translating the sentence the last

named

phrase mighl be used to advantage.

In general, intensives of the
Latin)
.

type
.

(however introduced
.
.

in
.

can be rendered by

even;" in this way we even" and "If have something more than stress of voice to mark the center of

"Though

.

intensive force.
In view of the \rvy restricted
si.
it is

and clearly defined use of
a

<

Ham

inexact,

when dealing with

the Language of Plautus, to

make

the phrase
is

si=

<

Horn

si a

substitute for saying that
a

given

si-clause
II

concessive.

Sonnenschein makes such

note on

ml. 1400:

N'on hercle istoc

me

intervortes,

si

aliam praedam perdidi.
the eisi-clause, as will be
si,

The

real parallel to this si-clause

is

at

Cas. 806 also shows the combination etiam
festly corrupt.

bul

the passage

is

mani-

Hi

University of California Publications.

!lass.
I

I'llll..

once evidenl
thai

if

it

be

compared with the examples quoted under
is

heading; the parallelism
in

complete, even to the

number

of

words

the clause.

TAMEN
('.is.
<

SI."

_*

cases.

795;
si

c

>ui

amat, tamen hercle
St.

esurit,

nullum

esurit.

27
si

ff.;

Tamen
I

Eaciet,
id

minume

irasci

>ece1

:

oeque

immerito eveniet.

Both of these are simple concessive periods.

TAMEN
.Mil.

ETSI.

12
'2

cases.

1209

IV.:

Postremo tamen
Etsi istuc inihi acerbumst, quia ero te

carendumst optumo,

Saltern id volup est
.Most.
'I'll.

quom

.

.

.

1167;

Verberibus,

Lutum, caedere pendens.

TR. Tamen

etsi

pudel

.'

These two cases are also simple concessive periods.
It

now remains

to consider
;i

two general syntactical peculiariFirst as to introductory particle:

ties

brought to Light by

division of concessive clauses according

as they are simple or intensive.

the usage of Plautus can be conveniently
in-

examined

in the follow-

summary.

Vol. l]

Nutting.—Studies

in tin

Si-clause.

47

intensive.

The overwhelming preponderance of
lies in

si

in

sentences

of the intensive type presents an interesting problem.

Appartwo

ently the key to the situation

the facl that, from the subjecdis-

tive point of view, concessive clauses in general fall into
tinct categories;

by the use of such
is

a

clause the speaker
it

may
tht

(a)
safa

concede thai
of argument,

a

thing
the

really true, or (b) concede

for

<>r

like.

As

in

the following examples:

(a)

Cas. 957

ft'.:

vapulo hercle ego invitus tamen,
Etsi

malum

merui.
(b)

Bacch. 1004:

Nam
Qui
In the

ego aon laturus sum,

si

iubeas

maxume.
esse addecet.
tin

Bacch. 128;
si

decern habeas Linguas,

mutum

tifsl

of these passages etsi nullum
;

nn

is

scarcely more
it

than

a

statement of fact
so

without altering the sense
it

could he
place

made such by
But

rearranging the sentence as to give
fall

firs!

in the cases thai

stale of affairs:

under (b) there is a totally different the concessive clause is a mere supposition, and.
pure conditional clause; 13 for
in in

as such,

is

closely allied to the
is

both the speaker
fact.

equally Lacking

assurance of realization in
distinctive function
it

Therefore

if si

a

word whose

is

to

introduce pure conditional clauses
as
a


is

is

also to

do duty anywhere
most freely used

concessive particle, clearly

it

in concessive clauses of this
it

second variety that we should expect to find

— and such
of
si

in fact is

the case.

'Idle

overwhelming preponderance
is

in

sentences of the intensive type

but an illustration of

the workings of this general principle; for in
sive clause by
its

them the concesits

very nature
it

is a

mere supposition

essential

characteristic being that

far exceeds the facts of the case, often
<.<}..

flying to the extreme of the improbable or the impossible;

Asin. 414: Siquidem hercle
(let itlllisse.

nunc

summum

[ovem

te dieas

Aid. 100: Si

Bona Fortuna
Quern
si

veniat.
ut mihi nil credat.

Bacch.

ti!'7:
:

orem

.Men. Tol
i

si

avom

vis

adducere.
p.

i.

American Journal of Philology, XXIV,

279

ff.

University of California Publications.

V

Looked

a1

from

this point of view, the large use of si in sentences

of the intensive type ceases to be surprising.

The fad
Imii

thai

si

introduces

18

of the 90 simple concessive

periods does oo1 perhaps seem to
it

call so

loudly for explanation,
illus-

may
of

be ooted

in

passing thai this ratio completes the
principle above noted with

tration of the genera]
Hi,. ,,sc
si.

reference to

A simple concessive period may be of either of the
and
(b).

varieties above designated as (a)
tically all thai
a

Si introduces pracits

are mere suppositions, and has found
a

way

to

considerable extenl into clauses thai admil
its

fact,

leaving the

larger share of these latter however for

more distinctively
again
classi-

concessive compounds.
In the following table the concessive clauses are
fied, this

time with reference to the

o

I

of the verb.

The

totals

differ slightly
en1

from those of the other table because, for the presit

purpose,

was necessary

to

exclude doubtful forms, such,

for instance, as those in -am.

vol.

l]

Nutting.

Studies in the Si-clause.

4!)

(68:17) for the simple concessive periods begin to appear; for, we have here to do both with clauses thai admil and with those thai are mere suppositions. The (verj a fad
as above noted,

numerous) cases

thai admil a

fad counl solidly on the indicative

mere suppositions contribute a reasonable number to each member of the proportion. Under these circumstances a heavy preponderance of the indicative is just
side of the ratio, whereas the
the thing to be expected in the totals.
15

The

ratio of indicative to subjunctive

i

12:32)

in

the intens-

Of course, we should ive periods cannol be explained so simply. expect to n both moods fairly well represented, for (as already
I

i

(

I

shown) the intensive concessive clause is by its very nature a mere supposition, and would therefore in general follow the rules
for

mood

in

puce conditions.

But

this

is

not a

full

explanation

of the ratio 12: 32; for in conditional sentences Plant us uses the indicative on the average

much more

frequently than he does

the subjunctive.

The intensive concessive clause however is something more than a mere colorless supposition— it is generally a very wild and improbable one. Apparently it is this peculiarity
that turns the scale so heavily in favor of the subjunctive."

14 am speaking here only of the language of Plautus, and in particular of the concessive clauses introduced by si and its compounds. Such a statement would not of course apply to a developed construction like the subjunctive cwm-clause in concessive periods of Cicero's time.
I

and tametei-clausea '•"In this connection it may be noted that the etei almost always concede a fact. The conventional ride for mood with these particles quite disregards this basis of explanation for the use of the indicative.
10

This point

is

further considered in the following paper.

See

p.

88

ff.

University of California Publications.

[Class. Phil.

STUDIES IN THE SI-CLAUSE.
II.

SUBJUNCTIVE PROTASIS WITH [NDICATIVE APODOSIS IN PLAUTUS.
1

In this
tive

apodosis"
I

paper the phrase "subjunctive protasis with indicais used in the broad sense in which ii is commonly
that
is.

undersl

as including all sentences
si,

whose subordinate

clause chances to be introduced by

irrespective of the

exad

nature of
is

1

1

-

underlying thought.

The argument throughout

based

..ii

junctive

.-Hid

sentences which employ undoubted Tonus of the subindicative; those containing forms in -am, -ar, etc.,

could only bring an element of uncertainty into the discussion, without them. 2 ;iikI the material fortunately is abundant

l.-PURE CONDITIONAL SENTENCES.
As
this
a

preliminary to the detailed study of the sentences of

-roup, attention

may properly be

called to the

undeveloped

state of the Language in the time of Plautus.

somewhal With

regard to this two points are of interest for the present discussion.

In the tive

first

place, the uses of the subjunctive
in

and the
at

indicaa later

were not

general so carefully differentiated as
ita

period.

For example,

mi di amabunt and

ita

nn amabit lup-

piter are used freely alongside of ita mi
i

di ament.

Again, take

he deliberative question

:

'See the Classical Review, Vol. XVII, p. 449 IV.. for a critique of the Blase on this f Ldiie, Lebreton and Lodge and the later theory of Blase's earlier view will he found in De modorum temporumque Bubjecl
,,,i

Philoenuntiatis eondicionalibua Latinis permutatione, Dissertationes (MS) IV. <'f. Liiiigon. Beit rage zur logicae Arovnteratenscs, V,»l. -X. \>. !»4 The subjecl is treated indi43 IV. Kritik und Erklarung des Plautus, p. Terentium eondicio e1 rectly by Lindskog, De enuntiatis apud Plautum dea eondi nalibus Lundae L895, and by 0. Brugmann, ttber den Gebraueh There are also manj alteren Latinitat, Leipzig, L887. cionalen Ni in der have received Blase a since this was written other scattered references, Theil, Mainz. L904, ,,„,.,. Studien und Kritiken zur lateinischen Syntax, at several points. pari of which touches the following discussion the latter The following cases also have little value for the presenl discussion on an because the subiunctive of the si-clause may be due to dependence 1193 ff., Ps. 1033 ff., infinitive or the like; A.mph. 675, A.,l. 228, 320, Bacch.
i„
1
I

b.

L12.

vol.1]

Nutting.

Studies in the Si-clause.

51

Ter. Phor. 736-37:

so. qui

est

eius pater.
.'

CH. Quid ago? CH. Adeo, maneo, dum

.

.

.

co-

gnoseo

Ter. And. (539: Sed quid agam? adeamnt ad eum

et

.

.

.

expostulem? 3

Finally mighl be cited eases of remarkable variation of
conditional sentences;
Ps.
t

mood

in

.'/..-

1070

!'!'.:

Roga me viginti minas,
si
ill.-

bodie

ilia sit

potitus muliere

sive earn tuo gnato hodie, ut promisit, ddbit.

Amph. 703

ft'.:

Bacchae bacchanti

si

velis advorsarier,

ex insana insaniorem fades, feriet saepius:
si
4 obsequare, una resolvas plaga.

It is possible that the

very considerable middle ground afforded

by forms belonging
(e.g.,

to

both the subjunctive and the indicative
-eris, ete.)

those in -am, -ar,
tlie

tended to delay a sharp

differ-

entiation between

uses of the two

mood

systems.

In tlie second place, in Plautus' day grammatical conceptions were neither so symmetrical nor so clearly defined as at a Later This is shown in an interesting way in such contrary to time.

fad sentences as the following:
Aul. 523-24:

Gompi Hart m ego ilium, ni metuam ne desinat memorare mores mulierum: nunc sic sinam.
Bacch. 635
PI. Si
niilii

sit.

pollicear.

MX.
si

Scio,

dans:

novi.

Poen. 1251-52:

primum,
St.

id fieri possit,
id

ne indigna indignis dei darent,
510-11:
te

ego evenire vellem.

Vod m ego
3

ad me ad

ceiiatn. t'raler tnos nisi dixisSi

I

mihi
715.
Laeerel
I

So also quid <i!i<>. and quid agam? in Ter. Phor. 447 and Eee. 'With these might be compared fist, 683 ff. (si nemo praeteriit,
esset,
si

and Bud. 744 (iam tanta

vivit).

52

University of California Publications.

[Class.

Pra-

te

apud
\

se

cenaturum

esse hodie,

quom me ad

se

ad cenam

ocat.

True. B30:

Nam

vinuiii

si

fabulari possit, se defenderet*
as here occur
to

Surd combinations of forms
stand, forjusl
ing
ii„.

a1

this

time the contrary
i

we can readily underFad idea was discard-

presenl

(and perfed
a

subjunctive, 6nding in the imper-

Ped and pluperfed
expression; bu1 thai
limits of a
for

more distinctive and satisfactory form of the two forms should be mingled within the
hick of keen

single sentence betrays a

appreciation

symmetrica] sentence structure.

These two characteristics of early Latin distinctly favored
the frequenl occurrence of subjunctive protasis with indicative apodosis.

For the failure

to

differentiate clearly between

the

and indicative forms in general must have affected also the choice of mood in the clauses of conditional sentences -in some cases, so far as meaning is concerned, there was
use of subjunctive

doubtless

Little to choose between the two moods; and to a writer whose ideas of symmetrical sentence structure were somewhat

undeveloped the pairing of different moods

in the clauses of a
a

conditional sentence probably did not appear to be such
ing irregularity as
it

strik-

seems when viewed from a later standpoint.

The many examples however in which Plautus uses the same mood in both clauses show clearly that he had a fairly strong
conception of this procedure as the norm.

This fact

is

by no

means

lost sight
it
it.

of in the following discussion, hut on the other

hand
given

is
7

not there accorded the undue prominence sometimes

So many cases falling under the head of the pure conditional
sentence have forms of posse
Cf. lack of
Mil.
8

in

apodosis that

I

venture to treat
A somewhal similar
is

B

i:;t

711-12, Cist. :; IV. and Cure, 226 IV. symmetry in the matter of sequence of tenses cf. Asin. 589-90 and Capt. 28.

Capt.

noted by

Bris on

;'

This statement may no1 be pul aside with the remark that the Language For the colloquial style, as well as others, was ..I' I'lautus is colloquial. profoundly affected by the developmenl of the language up to the time of cicero; sec Leluvion, Ktmles sur la Langue et la Grammaire de Ciceron Further, some may be surprised to learn from the [ntrod. p. x IV. espc. xv. tahlcs of Lebreton and Blase thai there are more cases of the form si sii est lint) iii Cicero's orations than in his letters; see 1,. p. .",(il and cf. p.
349.
Si e

Langen,

1.

c.

p.

50

fin.

Vol.1]

Nutting.

Studies in

tin

Si-clause.

53

them separately.
first

under

A,

while

Being somewhal simpler, they are presented the remaining sentences appear later

under B.

A.— Posse
The material
Palling

(Potis) in Apodosis.
this

under

heading may be subdivided

on the basis of tense.
(a).

Sentences of the form

si sit

potest.

This group furnishes examples of three different types.
1.

Unconditioned

ability.

Cure. 268-69:

Siquidem incubare velint qui periuraverint, locus non praeberi potis est in Capitolio.
Mil. 763-64:

Haud eentesumam
partem dixi atque, otium
In the
first

rei si sit.

possum expromere.
any way de-

of these examples the inability of the Capitoline to
all

provide accommodation for

perjurers

is

not

in

pendent on their wish to find-a resting place within its limits; and in the second the speaker's fund of information is a fad
uninfluenced by the truth or falsity of the condition. This state affairs makes it possible to provide a very satisfactory explanation of the form of the sentences.
the apodosis he
to

,.f

may

realize that the ability of

For as the speaker comes to which he menus and
say

speak
is

is

not dependent on the fultillmeiit of the condition,
I

he

therefore free to state that ability as unconditioned.
to

\'vov

do
in

so.
a

because

in so

doing he

is

using

a

form of expression
be done under
implies
the

which

way

includes and implies whal

r<>nl<l

the supposed

circumstances

— that

is.

includes and

Logically exact apodosis.

Though there

is

no absolutely certain
;i

case, still a

survey of

the material leaves
carries this process
all

strong impression thai sometimes Plautus
step further

a

and ventures

to substitute an

inclusive statement of unconditioned ability where the Logical
is

apodosis

would rather than could.
308:

Such an example may be

Cist.

Adhinnire equolam possum ego ham

.

si

detur sola

soli.

.-,1

University of California Publications.
be the true explanation of the sentence,
in

[Class. Phil.

If this

a close

parallel

is

afforded by the following case

which, after an indicative con-

dition, the speakers substitute for an assertion of

whal they
Latter

will
in a

do

a

statemenl of

wli.it

they

an wont

to

do

— the

way including and implying
Poen. 516-17
:

the former

si aec recte dicis nobis dives

de

summo

Loco,

divitem audacter solemus mactare infortunio.

2.

Conditioned

ability.

Cure. 246-47:
Potin coniecturam facere,
line
si

oarrem

tibi

nocte quod ego soniniavi dormicns?

Jn this case the ability to make a guess seems clearly dependenl on being provided with the necessary data, and at the same time the phrasing of the sentence shows that the speaker had the siclause in

ram

faceri

mind when he ottered the apodosis; for Potin coniectutaken alone is manifestly incomplete. Here then it
in*

seems that the speaker can have

mind only conditioned

ability.

and the use of

tlie

indicative cannot therefore be justified in the

same way

as in the sentences treated

under the preceding head-

The explanation which suggests itself most readily is the modality of the verb, and if we were dealing with a later But writer there would be little more to say on the subject. since in Plautus (as will soon appear) it is not always a modal
in- heading.

verb that

is

used

in the

apodosis of sentences like the one under
to recognize here also a further cir-

discussion,

we ought perhaps

cumstance which favored the use of the indicative, namely, the

somewhat undeveloped
undeveloped
tendency to
state,
fail
it

state of the Language

at

this time.

This

will

be remembered, betrays itself in the

to distinguish sharply
in

between the use of sub-

junctive and indicative forms, and
t'ical

the tolerance of
of affairs

unsymmet-

sentence structure.

Such

a stale a still

makes the use of

the indicative of the

modal verb

more simple matter. How
of poss<
in

easy

it

was

for Plautus to use that

mood

we may perhaps

judge fairly from the following passages,
one
ii

which he shifts from

(I

to the oilier

:

Vol. 1]

Nutting.

— Studies

in the Si-clause.

.">:>

Asin S7S

I'!'.:

PA.
cuiii

I'nssis

si

forte

accubantem tuom vinim eonspexeris
si

corona

amplexum amicam,
ft'.:

videas cognoscere?

AJRT. Possum ecastor.

Merc. 517

LY. Sed quid
possin tu,
si

ais,

Pasicompsa

.'

lissus venerit,

subtemen temie aere?

PA. Possum.
3.

Anacoluthon.

Rud. 566:
Vel ego

amare utramvis possum
;i

si

probe adpotus siem.
is

When
in

such

sentence as this

is

a

true index of what

passing

mind of the speaker, he enunciates the first clause as a comThen it flashes through his mind that plete stal.Mii. >!it of fact. the act or state in question is subject to a condition of which he
the

has not previously thought, and this he adds, rather lamely at
limes, allowing the hearer to correct the preceding statement of
fact just as his

own thought has been

corrected.

Syntactically

the effect

is

the same when, as seems to be the case here, the

speaker has his whole sentence planned from the beginning, but
purposely deceives the hearer by his enunciation of the
clause that he
surprise.
first

may

raise a

laugh by bringing
is

in

the second as a

In either case the si-clause

really part of another

sentence,

and uses the mood required by the laws of conditional

sentences generally, without reference to the
in
8 the clause which precedes.

mood
sit

of the verb

The remaining examples of the form
follows:

si

— potest

are as

Asin. 164:

Solus

si

ductem, referre gratiam
3

numquam

potes.

Aul. 557

ft .:

praeterea tibicinam

quae mi interbibere

sola,

si

vino scatat,

Corinthiensem fontem Pirenam potest.
* Tn the example under discussion the flexibility of meaning due to the But in the modality of posse tends to make the anaeoluthon loss harsh. ur\t main division (B) where the aon-modal verbs appear, eases will be found in which there is no such mitigating circumstance.

56
Baceh.

University of California Publications.
t79 SO:

[Class.Phil.

XiiIIh paeto res

mandata

potest agi, nisi identidem
a

manusf

feral
:

ad papillas, labra

labris

nusquam auferal

.'

Must. 351

Nee Salus nobis
Poen. 353
Sei sapias,
:

saluti

iam

ess.',

si

eupiat, potest.

curam hanc
:

facere

compendi
si

potes.

Poen. 864
ilium
in

perdanl facere possum,
IV.

velim.

Tri. 85

si

id iKui

feceris,

atque

id

tamen mini

lubeal suspicarier,

qui in id prohibere
All these cases
specified.

me

potes ae suspicer?

may

be broughl under the three headings above

Different

persons however mighl
a

hold diverse views
;

as to the

heading under which

given case should be broughl

but this fact has no bearing on the present discussion,

my aim

to

being simply to single ou1 the various distinguishable types and show what explanations of the phenomenon of subjunctive
protasis with indicative apodosis are suited to the peculiarities

may however say that anacoluthon is a basis of explanation to be sparingly used; for a speaker usually has his whole sentence in mind before the first word is uttered— even when in
of each.
I

the course of his thought the condition does not
clearesl
like

come

first,

The

cases of anacoluthon are deliberately planned surprises

Rud. 566.
(b).

Sentences of the form
IV.:

si

sit

(esset)

potuit.

Cure. 226

Adferre argentum credo.

Nam

si

mm

feral.

tormento non retineri potuit ferreo quin reciperet se hue esum ;id pr;cscpciu suam. Most. 462:

Quo modo

pultare potui,

si

non tangerem

.'

These are both cases of the second type, the (in)ability of the apodosis being clearly felt as dependent on the truth of the pro-

The explanation would therefore he again the modality of tasis. the verb and the undeveloped state of the language. In Cure. 226 ff. the disp.iritv between ferat and potuit is specially striking.
See
<

llassicaJ

l.v\

iew,

I.

e.

p.

l.v_'.

Vol. l]

Nutting.

— Studies

in lh<

Si-clause.

.">,

B.— Other Verbs
(a).
1.

in

Apodosis.
si

Sentences of the form

sit

est.

Unconditioned act or

state.

Merc. 430:

At ego

si

vt'liin. i;im
IV.:

dantur septem

e1

viginti minae.

Rud. 1020

Numqui minus
si

venial

aune dominus quoiust, ego qui inspectavi procul
.'

te

hunc habere, fur sum quam tu
liist

In the

of these examples dantur seems to

mean "I am

of-

fered" 10

a fact in

no way dependent on the willingness to accepl
in

the price, and in the second the participation
is

the guilty secret
not.

real

whether the owner of the property appears or

The

process which produces these sentences seems to be the same as
that described in the discussion of the corresponding cases with

posse in apodosis, namely, that the speaker substitutes for the
logical apodosis

an unconditioned statement which

in

a

cludes and implies that apodosis; thus dantur includes

way in"I might

have" and sum.' "would I be considered?" The difference between these two cases and those with posse in apodosis is that here
the statement of the fact
is

not so closely parallel to what

is

in-

cluded
or
a

and

implied (there the logical

apodosis was "could"
is

"would," and the statement "can"), and hence the usage
little

harsher.
2.

Conditioned act or

state.

Amph.
si

891-92:
illnd fieri

Faciundumst mi

quod

illaec postulat.

me

illam

amantem ad

sese

studeam

recipere.

Cas. 528-29

AL. Attatae, eaedundus tu homo's: nimias LY. Quid me amare refert, nisi sim doctus
In the
first

delicias facis.

ac dicaculus?

of these cases the need for action seems dependent
at

on the truth of the condition;

any

rate to bring the

example
felt
it

under

this

heading we must assume that the speaker so

as he began the sentence.

The second
I

case

is

clear enough,

tot-

As datur

in

Cic.

ad Att.

i.

18.

58

University of California Publications.

[Class. Phil.

obviously Quid mt

aman

refert

is

mean1
II'

to

apply

to the con-

trary to fad state of affairs supposed.

Amph. 891-92 belongs
in

here

il

is

quite like the corresponding examples with possi
the form of the sentence
tinis

apo-

l'>sis.

.-Hid

therefore to be explained

in

the

same way. namely on
first

ground of the modality of the
In

expression and the undeveloped state of the language.

the

second case the

part of this explanation
us.-

is

excluded, and we
the result of the

'in "nly say thai the

of the indicative

is

crude grammatical feeling of the writer.

3.
.Mil.

Anacoluthon.

68f>86:

Nam

bona uxor suave ductus/

si

sit

usquam gentium,

ubi ea possit inveniri.

This example

corresponds

exactly to the case of

anacoluthon

noted

among

the cases with posse in apodosis, excepting that the

effect is not here softened

this

by the presence of a modal verb. Under heading there are however some sentences which, if so inter-

preted, call for a

more elaborate analysis:

e.g.,

Poen. 550

Omnia
In
a

istaec

scimus iam nos,

si

hi spectatores sciant.

simple ease of anacoluthon

like Mil. (585-86 above,

the added

si'-clause corrects

the preceding statement of fact, warning the
is

hearer that the state of affairs there mentioned
condition after
all.

subjed
a

to a

Bui

if

Poen. 550 be regarded as

case of

anacoluthon, the statement of fact with which the sentence begins is in no way affected by the addition of the x/'-dause. Rather it
is

the inferenci

ment, namely "yon

which the hearer might draw from thai not tell us" which is corrected. 11
i I

state-

Other cases of the form

si

sit—esl are as follows:

Amph. 336: Non edepol mule
Capt. 206:

ubi lerranuii sim scio, shjuis roget.

scimus nos

il., si

nun

officium

quod
the

est, si

solutos sinat.
which

"This analysis supplies
(1. e.

p.

18)

to rejecl

the the line.

link,

failure

to

find

led

Lam

Vol. 1]

\ lilting.
Capt. 259-60:

—Studies in the

Si-clans,

.

59

Neque
aeque

p«>l tibi

nos, quia nos servas, at
si

quomsi

vi1 to

vortere

te nobis,

abeamus
sit

nine,

si

£ua1 occasio.

Capt. 850:

Sds bene

esse,

si

unde.

Capt. 906:

Nam

si

alia

memorem quae ad

ventris

vietum

eonducunt,

morast.

Cure. 299: Reete hie monstrat,
si

imperare

possit.

Men. 760:
quas
si

autumem omnes, nimis longus
sanus
si

sermost.

Merc. 497
.1/-

liust,

sis.

Mere. 692-93:

Parumne
ni

est

malai

rei

quod amat Demipho,
siel
.'

sumptuosus insuper etiam
Mil. 1263:

Noii edepol in ilium

magis amas quam ego,

tnea, si

per te liceat.

Poen. 921

:

mine

si

eadem
opus

hie iterum iterem, inscitiast.

Ps. 740:

Quid?
St.

si

sit

ul

dulce promat indidem, ecquid habetf

171-72:
si

Nunc

ridiculum hominem quaeral quispiam,

venalis ego

sum cum ornamentis omnibus.
se abalienarier,

Tri. 557-58:

Quin

hie

quidem cupit ilium ab

siquem reperire
litre
;iii-aiu.

pussit. quoi "s sublinat.

t'spccinlly

when the apodosis precedes,
However,
i.e.,

it

is

difficult
a

to say with

certainty under which of the three heads
classified.
first

given

example should be
seem clearly of the
in the apodosis
is

Ps. 740

and

St.

171-72

type,

the state of affairs referred to

felt as in

no way dependent on the truth of the
si

condition.
(b).

Sentences of the form

sit— erit.

Asin. 699:

Vehes po] hodie me,

si

quidem hoc argentum

ferre speres.

(jo

University of California Publications.
31]

[Class.Phil.

AiiI.

:

Famem
Irasa
/<

hercle

utendam,

si

roges,

numquam

ddbit.

Cure. 186:
.si te

edentem
:

hie a eibo abigat.

Merc. 650-51
si
ilii

amare

forte occipias atque

it*

m

eius

si1

inopia,

iam inde porro aufugiesf
Mil. 571
:

Ne

in hercle,
.Most.

si

te di

ament, Linguam com primt

s.

56-57:

Ita te forabuni patibulatum per vdas
fstimulis,
si

hue revenial senex.

Poen. 729:
Si pultem, qoii recludt
I

Poen. L085:

Quin mea quoque
Tri. 26-27:

isle

huh*

Lit,

siquid

me

fuat.

Coneasl'nialxi pro
invitus, ni id

commerita noxia,
invitet ut faciam fides. 12
last

me

For the purposes of the present discussion the

example

cited

may

be ignored because the apodosis

is

really invitus rather than
is

Concastigabo.

The most striking thing about the group
sentences of the second
a state

the

prevalence of

type,

i.e.,

sentences in

which the apodosis refers to

of affairs felt as conditioned.

Cine. 186, Merc. 650-51 and Poen. 729 (as here punctuated 13 ) are
clear cases.

So apparently Asin. 699, Most. 56-57 and Poen.
first

1085, unless the

be a case of anacoluthon.

In the sentences

of other forms thus far dealt with the explanations for examples

of the second type have been the modality of the verb of the

apodosis and the undeveloped state of the language.
ever none of the verbs are modal, and

Here howwe are again forced back

(as in the case of Cas. 528-29) to the other line of explanation.
I'.nt

in this

caieirory
is

t

he easy tolerance of the iinsym metrical senreadily understood.

tence structure
the apodosis

much more

For the verb of
in

refers to the

futun

a

time realm

which the
leasl

bounds of indicative and subjunctive meaning were perhaps
13

The manuscripl reading would add Bacch. 1172 to this list. The more difficult punctuation is si pultem, non recludet? i.e., "Whal
knock and he does aot open?"

if

I

vol.

]

i

Nutting.

Studies in the Si-clause. Plautus perhaps
in
tliis

(il

clearly set in early Latin.
use the futures
tlic
(if

fell

it

qo barsher to
to so

ordinary verbs
II"

way than

employ

presents of modal verbs.

so.

we can

readily understand
in tins

the prevalence of sentences of the second type

category.
:!11

Of the two
of the
lirst

cases of this Eorm not yel treated, Anl.

seems

type, the action of the apodosis being independent

of the truth of the protasis.
parallel
tive

The other ease
Torre.

(Mil. 571)

has no

among

the sentences thus far treated, the future indica-

having something of imperative

The whole passage

is

PE. Xc tu

hercle,

si

te di anient,

linguam comprimes:

posthac etiam illud quod sceis aesciveris
aec videris

quod

videris.

SC.

Bern m< mones.

The

line here

between indicative and subjunctive was not very

clearly defined, as

we may

see by

comparing

line

293 of the same

play:

Veniin

etiam

tu

istam.

si

te di

anient, temere

ban

tollas

faludam.
(c.)

Sentence
IV.

of the

form

si

fuerim

erit.

Cas. 335

Sed tandem

si

tu [uppiter sis emortuos,

quom ad

deos minoris redierit
nii
I

regnum tuom,

pus mihi subvi
is

tergo aut capiti ant cruribus?

This also

an example of the second type.
Sentences
of the

(d.)

form

si

esset (fuisset)

fuit.

l't

Amph. !>47-48: quae apud Legionem
Bacch. 818-19:

vota vovi,

si

domum

redissem salvos, ea ceo exsolvam omnia.

llunc

si

ullus deus amaret, plus annis decern,

plus iam viginti
('as.

mortuom
si

esse oportuit.

440-41:

Volui Chalinum,

domi

esset,

mittere

tecum obsonatum.
Mil.

175-76:

Quid propius

fuit
.'

quam

u1

perirem,

si

elocutus essem ero

62

University of California Publications.
1356 57:
sententia esset,
alii

[Class. Phil.

.Mil.
e1
si

ita

tibi

servire malui

multo quam
Vide
iii

liberl us esse.

Pers. 594 95:
sis.

ego

ille

doctus leno

paem

in

foveam decidi,

hie adesses.
Ps.

285:
si

Wuit occasio,
Ps.

vellet,

iam pridem argentuni

u1 daret.

1241-42:

At ego iam intus proinam viginti minas

quas [no nasi,
St.

si

effecisset.

563:
voluit,
si

Senex quidem
Tri. 566:

posset, indipisci de cibo.

Licitumst,

si

velles.

True 140:
Si

rem servassem,

fuit

iibi

aegotiosus essem.

Obviously some of these sentences belong to the categories above
described.
tirst

Mil.

1356-57 and True. 140 are most clearly of the

type, and Bacch. 818-19
in

and
Tri.

Mil. 475-76 of the second, with

the modal verb oporten
dosis precedes in Ps. 285
cult.

the former.

The

fact thai the apodiffi-

and

566 makes exact analysis

The other
St.

eases of this group have peculiarities; Cas. 440in

41

and

563 (with forms of velh

apodosis) are hard to deal
its infin-

with because one scarcely knows whether to treat velh or
itive as

the apodosis proper.

Amph. 947-48 and

Ps.

1241-42 are

simply abridged; in the latter case, for instance, viginti minus quas promisi means of course "twenty minae which I promised to give," and it is in this idea of giving that the si-clause finds its
logical apodosis."

The one remaining case
of the whole group.
It

i

Pers. 594 95

I

is

the most interesting
in

is

one of the rare examples 15

Plautus

of the contrary to fad type of sentence which tells what was on
die poini of

of

happening bu1 which did an intervening circumstance. Were
to count this
!"•

not
it

come
for

to pass because

not

paem we mighl
form

perhaps he inclined
IB

another example of the second
:i

Cf. Tri. 835
I's.

ff.,

which

may

so punctuated as to

parallel.

u cr.

199.

Vol.1]

Nutting.

Studies in

tin

Si-clause.

63

type

decidi would then be
;

a

mere piece of exaggeration.

Bu1

paene disqualifies

is

clause for being the apodosis of ni hie ades-

ses; for the realization in fact of that condition

actual falling
his leno paent

in.

qoI almost

falling

in

foveam

<l<ri<li is

would have meanl The phrase ego ilh doctherefore worded without refin.
-;t

erence
of fad

i<»

the addition of the ^'-clause
is

the end,

and
to

as

;i

matter
i

it

in

itself

;i

complete and precise statemenl
In other

ding

no further qualification.

words we seem
is

have to do

with

;i

ease of anacoluthon, but this

different

from any examof the fol-

ples of the

phenomenon yet taken up. lowing sentences will make this point
Vel ego
(

A comparison
clear.

amare atramvis possum
istaec
. .

si

probe adpotus siem.

>

i

«•

r

1

1

1

i

i

scimus iam aos
in

si

hi spectatores sciant.

Ego
In the

.

paene

foveam deeidi

ni tu adesses.

first

of these examples the speaker corrects the opening
S4-clause, Letting the hearer
is
is

remark by the use of the
In the second the

know
a

that

the state of affairs there asserted
tion.
.s/'-clause

after

all

subjeel to
;i

condi-

added as

necessary check

on tin hearer's unconditioned inference from the statemenl
nia
istaec

Omn<
it

scimus iam

nos,

namely "you need not enumerate
is

them."
in

In the last
<h<i<li
fall

example neither of these things
to be

true;

/><i<

foveam

and the obvious inference

drawn from
is

("I did not

in") are both facts subject to no condition, and
a

neither therefore needs

corrective m'-clause;

and such
fall

not

the function of ni hie adesses. Rather, this contrary to fact phrase
is

used to imply the reason

why

the speaker did not
it

into the

trap.

Withoul making any elaborate analysis
is

is

clear that this

implication
is

the chief function of the clause; for the speaker

obviously

using the words to express his obligation to the

hearer for his presence (and advices, representing them as the
cause of his escape.
rect
Iii

other words.

111

hie adesses does not cor-

the preceding statement or the unconditioned inference from
••!

u
tile

did not
liy

fall

iii"

.

but

it

further extends the thoughl of

sentence
11

assigning the cause for the thing to be inferred.
to treal sentences of this sort

is

customary

as the result

t<\'

ellipsis, but

the above analysis suggests another possible line of
In

explanation.

Plautus there are

many

regularly formed con-

CI

University of California Publications.

[Class.Phil.

trary

to

fad

conditional sentences whose chief

function
;

is

to

assigns reason for an existing or pasl Btate of affairs
Mil. 1262:

e.g.,

MI. Nun

vrideo.

Obisl

'.'

AC

Videres

pol,

si

amares. 18
in the

In this passage Videres takes

cognizant of the fad stated

preceding speech (Non video), and the st-clause assigns the reason for thai fact, i.e., that the firsi speaker is qo1 really in love
In the

sentence under discussion (paem
a1

in

foveam

decidi, ni hie

adesses),

the end of the

firsi

clause the speaker

may become
decidis-

conscious thai Ins words take cognizance of the fad thai he did
not
fall

in. just

as

would have been the case had he said
in
is

sem, and this perhaps tempted him to use,
cause of his not falling
in,

the form which
a

acknowledging the generally employed
is

only when the fact for which

reason

is

assigned
in Mil.

implied by a

contrary to fact subjunctive apodosis, as
(e).

1262 above.

Sentences
ff.:

of the

form

si

esset

fuerat (erat.)

Baceh. 563

tibi non erat meretricum aliarum Athenis copia, quibuscum haberes rem. nisi cum ilia quam ego mandassem

Quid?

tibi.

occiperes lute etiam
Mil. 52-53:

amare

.

.

.

.'

Quid
ni

in

Cappadocia, ubi tu quingentos simul,
i'oret.

hebes machaera
St.

uno

ictu occiderasf

512-13:

E1

magis par fuerat me vobis dare cenam advenientibus,
ilium promittere, nisi nollem
ei

quam me ad

advorsarier.

The interesting example of this group is Mil. 52-53, showing as it does the same sort of JM-clause as appears in Pers. 594-95, which has just been discussed at length. The explanation here however
is

much
of

easier,

for the

///'-clause

precedes. 17 and the action re-

ferred to in the ;ipodosis ohvioiisly depends on the coming to
p;iss
tin'

condition thai was not realized.
is

This therefore
to he

is

but

another example of the second type, and
18

explained partly

Vol.
17

The other i-;isis are enumerated XXII, p. 310 IV.

in

the American Journal of Philology,
a

This precludes treating the sentence as

rase of anacoluthon.

Vol.

i|

Nutting.

Studies intJu Si-clause.

65

in

the

same way

as others of thai class, partly on the

ground of

the spirit of exaggeration thai pervades the passage in which the
18 sentence occurs.

The other two

cases in this group seem also
olL'-l:'.

to be examples
in apodosis.

«»!'

the second type; St.

lias

;i

modal verb

In

summing up

the results of this study with reference to
it

pure conditiona] sentences,

may

be

with the assumption that Plautus had
of the

remembered thai we began a fairly dear conception

same mood
is

in

both clauses as the norm.

The problem

in

hand
not

therefore to discover the reasons
to that

why some sentences do

conform

norm.

Four such reasons have been enumentioned
in

merated.
1.

The

fact that the state of affairs

the apodosis

is

often in no

way dependent on

the truth of the protasis; the
l><

indicative statement

includes and implies what would

in the

supposed
2.

case.

The modal meaning of certain

verbs, notably posse.

The union of a complete sentence and a part of another by anacoluthon. The form of each membei of the expression is determined by the thought it is to convey, irrespective of the form of the other member. 4. The somewhat undeveloped state of the language in Plautus' day. as shown (a) in irregular sentence structure and (b) This method of explain the not very precise use of mood forms.
8.

nation finds
the

its

most sweeping application
the
fact

in cases

referring to
indicative

future:

for there

that the realms of

and

subjunctive

meanings were not
in

carefully

differentiated
still

tended to make the lack of symmetry
less

sentence structure

noticeable to Plautus than

it

would otherwise have been.

Aside from sentences referring- to the future there are very few
such sentences as this it should !» remembered also Latin was in the mi. 1st of the process of adopting the use of the secondary tenses for the expression of the contrary to t'a.t idea. In Greek it was the indicativi that was chosen when a similar shift of tense was made in that language, ami it is possible that we should recognize in early Latin some sporadic and unorganized impulses to develop in thai way Cf. Men. L95 (si amabas), rather than toward the use of the subjunctive. An interesting Ps. 286 (si amabas) and perhaps Rud. 379 (si amabat). also displayed in Cas. 811 (si equos esses, esses indomabilis) and variety is With regard Mil. 1111-1° (tu quidem ad equas fuisti scitus admissarius). to the case under discussion Bris seems to lay too much stress on the demands of the metre; cf. Ins note Lbid. 131.
:

"

With regard

to

that at this time

66

University of California Publications.

[Ct^as. Phil.

cases for which this
erally
ii

is

the only possible line of explanation.

Gen-

is

to be

combined with others, as

for instance with 2

above.

One

or two combinations with factors nol here enumerin

ated were mentioned

the discussion of individual

c;is.-s.

,:

'

II.-CONCESSIVE SENTENCES.
(a).

Sentences of the form

si

sit

est.

Asm. 318-19: quidem omnes coniurati crueiamenta conferant, habeo opinor familiarem tergum, ae quaeram foris.
si
As,,,.

933:
nil sit, tui

Pol

si

alind

me, uxor. pud(

I

Bacch. 128:

Qui

si

decern habeas linguas,

mutum

esse addecet.

Bacch. 1045-46:
Si pins

perdundum
tr.

sit.

periisse suaviusi

quam
Quin
ana

illud flagitium volgo dispalescere.
:

Cas. 314
si

no!

is

filiusque etiam tuos,

vobis invitis atque
libella liber

amborum
fieri.

ingratiis

possum

Cist.

27

ff.:
si

Si

idem istud nos faciamus,
vivimus

idem imitemur,

ita

tamen vix

emu
Ibi

invidia

summa.
:

Merc. 841
(|iii(l( j ni si

regnum detur, oon
:

cupitasi civitas.

Pers. 40-41

Quin
quod

si
t

egomel lotus veneam, vix recipi potis

est

n

me

rogas.

I's.

291:
si

Atque adeo,
St.

facere possim, pietas prohibet.

13 IT.:
illi

E1

si

improbi

sinl

.

.

.

nostrum officium meminisse
"The
jussive

d(

a

t.

force of the future indicative (Mil. 571) and geration which pervades the passage in which Mil. 53 occurs.

the exag

Vol. l]

Nutting.

— Studies

in tht

Si-clause.

(i<

Tri.

1186:
pro peccatis centum ducal uxoris, parumst.

Nam

si

True. 877:

Pactum eupio: nam uefacere The sentences of
single variation
this

si

velim, non est locus.
illustrate the

group well
its

tendency of the
is
IV.

concessive si-clause to precede

conclusion: here there

no1

a

from the

rule.

Excepting
is

in

Cist.

27

,-iml

possibly in .Merc. 841

the si-clause

a

mere supposition, and
that thai
ideal

takes the subjunctive
is

mood

for the

same reason

m

I

employed

in

pure conditional sentences of the

and consi-

trary to fact types.
clause.

Having begun

his sentences with such a

Plautus nevertheless does not hesitate to complete them
a
it

with an indicative conclusion, and such
justification.

course
will be

is

not without

For

in the

above examples

found that the

conclusion refers regularly to a state of affairs actually existing

and which would continue

to exist despite the

coming

to pass of

what

is

supposed

in the si-clause.
at
it

Both of these things the

speaker cannot express
in

one and the same time, though perhaps
possihle to lollow a middle course by

some cases he finds

using the indicative

when

the verb chances to be modal.

But

with other verbs at any rate he must make a choice; by the use
of the indicative he can assert tin existing statt of affairs, allow-

ing the hearer to gather that the same state would continue under
the adverse circumstances supposed,

employing the subjunctive he can confine himself
In

and on the other hand by to what would
it

in'i

despite those circumstances, leaving

to the hearer to
is

infer the actually existing state of affairs.

Either mood

there-

fore justified by the nature of the situation and the underlying

thought.

The indicative

is

the

more vigorous and comprehensive
as

form of expression
a

whereas the use of the subjunctive appeals to

mind trained
!onsequen1
:

to

grammatical niceties

producing

a

more sym-

me1 rical sentence structure.
(

ly in

(

Jicero the

subjunctive

is

the normal and reg-

ular usage

(

.</..

p. Sulla 13.38

Ne

si

arguerel quidem turn denique
rt
l

...

id

mihi crimino-

siim vidi

in-.

68

University of California Publications.
as here the supposition

[Class.Phil.

When

is

contrary

to

i'.n-i

.

the choice of

the secondary

tenses of the subjunctive in the conclusion

makes
is

the speaker use the form of unreality oi something which

as

a

matter of fad true.

Nevertheless
in

in

the orations alone there are

some seventy cases
with
itself.-"
p.

which

a

si-clause containing the imperfecl
its

or pluperfect subjunctive forces

conclusion to agree
resisl

in

mood
e.g.,

Even modal rerbs seldom
:

the pressure;

Arch. 7.17
ipsi

Quodsi

haec oeque attingere oeque sensu nostro gustare
ea mirari

]x >ssciii us.

tamen

deberemus. 21

A

ease where, instead of allowing the si-clause to force the use

of the

imperfecl subjunctive
just given
is
I

in

the conclusion

(as in the

two

examples

Cicero chooses to simply assert the existing
e.g.,

state of affairs,

generally counted noteworthy:

Lael. 27. 104:
si
illis

plane orbatus essem,

magnum tamen

adfert mihi aetas

ipsa solacium. 22

Plautus' usage

is

in

sharp contrast to

this, as at

once appears

when we compare
forms
sit

si sit

those concessive clauses in which he uses the
si sit

est

and

erit

with those in which the form
beinir Cist, 27
ff.

si

sit

appears.

Omitting for the time

and

.Merc.
fies

841 which (one or both) have a peculiarity which disquali-

for participation in this comparison, there have been cited
si sit

above ten cases of the form
four of the form
si sit

of the indicative in

Over against these fourteen cases the conclusion, even by including two passerit.
is

— est; below

there will be given

ages

in

which the
si sit

text

corrupt, there are but live examples-'' of
is.

the form

-sit. that

live

examples

in

which the influence

See Amer. Jour. Phil., Vol. XXI. p. 270 (V. So also, oporteret (in Verr. II. I. 27, 70; II. i'. ti. L5 and 40, 99; N. 4, 51, ill. de prov. eons. I. 35); deberem (de prov. cons. 20, 47; deberetis (p. Tull. 15, 36); debereni (in Yen-. II. :i. id. 91); possem (in Pis. 33, 81); posses (in Caeeil. L3, 43 and l!». 62; in Verr. II. :;. 72, L69) posset (in Verr. II. 3, L3, 32); licerei (p. Mil. l'7. 72), etc The idiomatic imperfecl indicative of a modal expression referring to the present occurs de Imp. Pomp. 17, 50.
l
1

Cf. j>. Sulla 30, 83, and the preceding note fin. There are three other cases of this form, bul thej are excluded here because the subjunctive of the conclusion can he otherwise explained— characteristic (Bacch. L79), dependenl on ut (Tri. 487), jussive (True.

vol. ij

Nutting.

— Studies

in tJu

si-da us,

69

of the subjunctive si-clause

was

strong

enough
Less

to

move

the

speaker to choose the more symmetrica] bu1
of conclusion.
resist

vigorous form

And whereas
Surely

in

Cicero even modal verbs seldom
jusl

leveling, in these

examples
it*

menti<

levels a

modal verb.

we

need< d

Plautus nowhere any additional evidence
I

to prove Plautus"

freedom from the
it

thrall of

hard and

fast

gram
are as

matical conceptions, we have

here.
si
sit sit

The
follows
:

live

cases

in

which he uses ihe form

Aul. 555

IT.:

Quos
is

si

Argus
s<

servet, qui oculeus totus fuit,
rr<
I

numquam
si

Bacch. 697:

Quern

orem

at mihi nil credat, id

mm

ausit credere.

Tri. 885
si ante

ff.:

lucemf
sit

ire

occipias

a

meo primo nomine.

concubium

noctis

priusquam ad postremum perveneris.

True. 315-16:
Si ecastor hie

homo

senapi victitet, nmi censeam

tarn esse tristem posse.

True. 527-28:

vSih plane ex medio mari
saviiim petere

tuom

iubeas, petere ban pigeat, mel
still

meum.

Few
a

as these cases are, they

suggesl one of the ways in which

a concessive si-clause containing the subjunctive tended to exert
Levelling influence on
is

its

conclusion.

In the

lirst

passage cited

Euclio

much

distressed Tor fear the cooks will steal something,
is

self

and the thoughl he wishes to convey should undertake to watch them,
from pilfering.

that

though Argus him-

still
1

they could not he kepi
a

Had

the conclusion

phrased

in

this

way
the

the verb would doubtless have been

in

the indicative. bu1

emphatic Argus of the si-clause has tempted Plautus to resume the emphasis in the conclusion with is, and he has thereby committed himself to
junctive
is

a
:

periphrasis
for

in

which anything hut the sub-

dit'liciilt

how can

the clause he

made
is

a state

nt

describing the existing state of affairs when Argus
of discourse of affairs,
a

the subjed

personage who has no connection with that state
is

and who

after

all

only

a

figmenl of the imaeinal ion

.'

7(>

University of California Publications.
for the speaker to do

[Class.Phil.

The only thing
good

left

is

to accepl

the other

alternative and state whal
a

would

fa

despite the selection of so
to infer the existing

guardian, allowing the hearer

state

of affairs."

The second passage above
emphatic
nil

cited

is

of precisely the

same
find

sort, the

of the s*-clause being echoed by id of

the conclusion; having begun with tins
it

word the speaker would
a

difficull

to

complete the clause as

statemenl of fad de-

scriptive of the existing state of affairs.
cases have oo resumptive

The remaining three
the
si-

word

in

their conclusions; bu1

clauses each contain an emphatic

word or phrase which won!.

have allowed of resumption (anU lucem, Tri. 885, senapi, True.

medio mari, True. 527), and the speaker may have fell something of resumptive force even though he did not definitely
315, ex

express
as to
lit

it.

At any rate the conclusion in each case
;i

is

worded

so

such

resumptive word or phrase, and not as
if

it

probably

would have been

the speaker had planned for an indicative

clause descriptive of the existing state of affairs.

In cases like these last three where the si-clause contains an

emphatic element that might

be.

but as

a

matter of

fact

is

not,

resumed
in

in

the conclusion, Plautus' usage probably varies.
is

Thus

Bacch. 128 though the verb

modal he has perhaps chosen
esse addecet.

to assert in the conclusion the existing state of affairs:

Qui

si

decern habeas linguas,

mutum

Had

he allowed himself a resumptive phrase,

we wonder whether
In Eng-

'ven the modal verb would have resisted the pressure.
lish at

any rate we have no option
it

— we cannot say "Though you
fitting that

had ten tongues, with the ten
the ten tongues do not exist.

is

you be

silent.*" for
it

We

must say "with ten tongues

woidd

I"

fitting, etc.*'

C\\ Tri. list;.
si

Before leaving this group of sentences of the form
a

sit—est,

word should

lie

added with reference

} to Cist. L 7

\)\

and Merc

[n terms of the preceding paper such an example is .-in intensivt con cessive sentence. Euclio is nol eontenl with any reasonable concession such .•is "Though we watch them," bu1 in his desire for emphasis lie flies to the must extreme of suppositions, "Though Argus Bhould watch them." Such
e

:essive clauses are a

mannerism with Plautus.
is

When

the elemenl

which
(here

renders

the
tin-

Argus), speaker

supposition extreme periodic nature of the

something other than the verb

e

to

resume the emphasis

in

sessive sentence naturally inclines the the conclusion bj a pronoun or the like

(here is), thus introducing into thai clause an elemenl which is as little suited as the word resumed to I"' a factor in a description of the existing -taie of affairs.

Vol. i]

Nutting.

— Studies

in II"

Si-clause.

<1

841.

These are what might be called general

concessive sen-

tences, 28 differing

Prom the others
is
ii

in

thai

the si-clause neither

refers to the future nor
clearly in Cist. 27
at
ff.)

contrary to

fact, bu1

rather (mosl

deals with something which dues happen

leasi

occasionally.

Such

a

si-clause

is

quite analogous to

a

genera] "condition,"' where the same use of the subjunctive occurs, notably

when
a

the subject of the verb

is

the indefinite second
in

singular.

Such

subjunctive si-clause,
its

even

the
It

strictesl

Latin, exercises little leveling force on
this reason thai these

conclusion.
in

was

for

two cases were excluded
form

the comparison
in

made

to

determine the ratio of indicative to subjunctive
t

the

conclusions of concessive clauses of

lie

si sit

:

their inclusion

would have increased
of indical ive cases.
(b).

a

little,

and perhaps unfairly, the number

Sentences of the form

si

sit

erit

(futurus est).

Am ph.
Quadrigas

450-51
si

:

nunc inscendas
ita

[ovis

atque hinc fugias,

vix poteris effugere infortunium.

Asm. 414-15:
Siquidem herele nunc
atque
is

summum

[ovem

te dicas detinuisse

preeator adsiet,

malam rem
si

effugies

uumquam.

Bacch. 1004:

Nam
Si
ita

ego nun laturus sum.
:

iubeas

maxume.

Ep. 610-11

undecim deos praeter

sese

secum adducal [uppiter
runt eximere Bpidicum.

non omnes ex cruciatu

\><>h

In this group the conclusion refers to something that will not
take place and would
(still
i

not take place despite the
in

coming

to

pass of the state of affairs supposed
able to express
all

the si-clause.

Nb1 being

this definitely in a single clause, the speaker
\

may

either assert that the thine
it

u

question will not lake place

or that

would

not.

(even) in the case supposed.
in

The

firsl

of

these alternatives seems to he chosen

the second and

third

examples.
conclusion,

In the other two cases, despite the indicative of the
it

appears as though the speaker intended

to accepl

the second alternative, giving expression to what would
'>

come

to

r.

Jour. Phil., Yel.

XXIV,

p.

300

72

University of California Publications.

[Class.Phil.

pass.

For

in

both sentences the emphatic elements of the
ita

si-

clause are echoed by

("even so"), which seems
in

to restrid the

conclusion to the supposed case; and
tion
is

Ep. 610-11 such restric-

further indicated by the carrying over of the emphatic
t<

subjed of discourse from the si-clause
result

the conclusion,

nnd the

thus produced on the phrasing there; 20 for otherwise the

clause would naturally have taken the form "nevertheless Epi-

dieus cannot he saved," as
Asin. 414-1")
si

in

the very similar ease

in

quidem hercle nunc
is

summum
why the (Amph.
in

[ovem

te dicas detinuisse

atque

precator adsiet,
is

malam rem

effugies
is

aumquam
used
in the

The question
answer
cases
is

therefore

indicative

two

sentences under discussion
is

450-51

and Ep. 610-11).

The
point

to be

found partly

the fact that the verb in both

posse, partly in the reference to the future

where indicative and subjunctive are

least clearly distinguished. 27
si

The only concessive sentences of the form
which those of
this

sit

— the —
five

sit

with

group may he compared are the

in the discussion of the

form

si sit

quoted

<st.

(c).

Sentence of the form

si

sit

fuit.

Bud. 159:
si

non moneas, nosmet meminimus.
its

By

virtue of

meaning

this sentence might

with those of the form
Sentences

si sit

have been treated
is

est.

Its

explanation

the same.

(d).

of the

form

si

esset (fuisset)

fuit.

Cure. 449 ft:

Quia enim
non pot m

in

cavea

si

forenl

conclusi itidem ut pulli gallinacei
ita

re

uuo anno circumirier.

Ale re. 595-9(»:

iam
20

Sed vtamen demsi prodagrosis pedibus a portu redisse pot ml

essei

Eutychus,

See the discussion above of concessive sentences of the form

si sit

sit.

My
<-•'

colleague Prof. Prescotl calls attention also to the minatory force

Anipli. 450-51.

vol.

l]

Nutting.--Studies

in tin

Si-clause.

13

Merc. 694-95:

fDecem nimium

si

ad cenam voeassel

summos

viros

opsonavit.

Mil. 803-04:

Non

potuit reperire,
a<l

si

ipsi Soli

quaerendas dares,

lepidiores duas
Ps. 792-9:5

li;mc

rem

<|u;un ego.

Nam

ego

si

iuratus peiorem

hominem quaererem
28 quern duco ducere.

eoqum,

rion

potui

quam hunc

In this group the conclusion refers to a present or past state of
affairs

which would be (would have been) unchanged despite the
to pass of the

coming

thing supposed. In Merc. 694-95 the speaker

seems clearly to choose the alternative of asserting the past state
of affairs.
fore,

The other four

cases contain the verb posse,

and

there-

though

in the indicative,

may
this

conceivably refer to wbat
in

would be or would have been:

seems to be the ease

Mil.

803-04, tor the emphatic Soli of the si-clause provides a subject of

discourse for the conclusion, thus dominating the phrasing of
that

member

of the sentence

and

restricting
:

it

to the

supposed
449
ff.

case (see the discussion above of Ep. 610-11)
looks in the

ita of Cure.

same

direction.

The

exact

meaning of the remaining

two eases

is

not clear.
1

With
form
st

the sentences of
esset

his

group may he compared two of the
ff.

(fuisset)-fuisset, Men. 238
is

and Most. 241-42.

The

first

of these

an interesting illustration of the resumption
restricting effect.
si

of emphasis and
(e).

its

Sentence of the form

fuisset

futurus erat.

Cist. lo2-o:5:

quod
ego
(

si

tacuisset. tanien

rti in

dictw

a*

This case
its

is

interesting as being apparently the only example of
it

kind in Plautus, though of course
later.

is

of a type
I

common

As a conclusion of quod si tacuissi Plautus' usage elsewhere would lead us to expect either a statement of the fact of the case ( 'I shall tell") or an announcement that this state
enough
:

of affairs would he undisturbed even under the supposed circumCf. the corrupt Capt. 417 1s.

74

University of California Publications.

[Class. Phil.

stances

"1 shun].

I

have told").
firsl

Following his usual procedure

he leans toward the

of these alternatives, bn1 substitutes "I

was prepared

to tell"

for

"I

shall tell."
is

Though

rare in conin

cessive sentences, such substitution

very frequenl

Plautus

generally: everywhere we find expressions of ability, willingness,
readiness, habil

and the

like

substituted for assertions thai some-

thing will be broughl to pass. 29
this case
is

the

The really noteworthy thing in tense—which however is n question thai belongs
contrary to fad construction rather than to

to the history of the
a

discussion of the concessive sentence.
It is

perhaps
in in

hardly

necessary to

broughl ou1
again thai

the

preceding discussion.

sum up whal lias been I may however say

concessive sentences of the kind treated in this
to a state of affairs actually

paper the conclusion regularly refers
existenl

posed. 30

and which would be undisturbed even in the case supThe speaker must in general choose which of the two
firsl

things he will state, the

naturally calling for the indicative

and the second for the subjunctive. In the case of modal verbs it is hard at times to determine which course a speaker meant to
follow

and

it

is

possible that occasionally in such

examples he

did not

make

a

conscious choice.
is

To

assert the existing state of affairs

unquestionably the

more vigorous and comprehensive form of expression, and it is not strange that i1 was a favorite with Plautus. though the subjunctive was the rule later, even in the case of modal verbs. The few examples in which Plautus uses the subjunctive would seem
to

show

that he

wis moved

in that direction, at least in

pari,

by

the fact that an emphatic element of the .^/-clause echoed in the

conclusion tends to commit the speaker to a turn of phrase unsuited to he
his
;i

description of the existing state of affairs.
a
17.

After

time doubtless
'See
Poen. 516
Vol.

much more important and sweeping
which has
1

influei

n

already discussed, ami Amer. Jour.

Phil.,

XXIV,

p.

294.
I

"This is ;in csscut i;i and fundamental characteristic of the concessive Occasionally there appears a pure conditional sentence which lias very similar accidental characteristic, namely that the apodosis refers to an action or state "f affairs which would occur in the supposed case, but whose happening as matter of fad is not dependent on the truth of the condition. Such conditional sentences provide examples of the firsl type discussed, ami the explanation of indicative apodosis there i^ very similar to thai nt' indicate e e fusion here.
periods.
a
,-i

Vol. l]

Nutting.

— Studies
a

in thi

Si-clause.

<5

was exerted by the growing appreciation of grammatical symmetry which demanded
tences.
III.

subjunctive conclusion for

a

subjunc-

tive concessive si-clause, on the analogy of pure conditional sen-

— SI

IN

OBJECT CLAUSES.
is

This not altogether satisfactory heading
scribe such si-clauses as complete the
fact

designed to dea

meaning of
a

statemenl of

a function very differenl

from thai of

si-clause in a con-

ditional period or concessive sentence,
also peculiar in

These object clauses are
period the
si-

position: for in the conditional

clause

may

either precede or follow, in the concessive sentence
it

it

almost always precedes, but here

regularly follows.

(a).

Sentences

of the

form

si

sit

est.

A.— Dependent
("as.

on Verbs of Expectation and Waiting.

540:
si

Quae iam dudum.
('as.

arcessatur. ornata exspectat domi.

542:
se arcessas,

Intus

ilia te, si

mam

I

Poen. 12:

Iam dudum
Tri. 98:

expecto,

si

tuom

offieium scias.

Expecto. siquid dicas. Tri. 14S:
Ausculto, siquid dicas.
In the
first

two of these sentences the
for.

si-clause tells the
si
is

1

1

1

i i

expected or waited

and the meaning of

conditional.

approaching somewhat that of dum, bu1 conveying
that the thing in question will

less

assurance

ultimately happen.

The third
is

example
not

is

obviously different.
for the other to

There the speaker

of course

waiting
to

know

his

business,

and we are
whether

tempted

render "I have Long been waiting (to see)

you know- your business,." making expectan the poinl of support for an indirect question; as for instance in
Cic. in
i

Ye rr.

II. 1. 59.

154:

xpi ctemus quid dicant ex Sicilia testes?

76

University of California Publications.
true thai
full

[Class.Phtl.

It

is

E.

Becker 83

is

probably righl
in

in

denying thai

si

ever has

interrogative force

the writings of Plautus.

But

the passage
;iik1

in

question
in

is

from

a

prologue probably of later date,
nol

therefore

our analysis we are
to the original
w;is

restricted hy the I'lau(Tri. 98
si

line rule.
difficuH to

In the
s;i\

two remaining cases

whether

speaker
;n

and 148) it is was purely con-

ditional, or

whether there

some

im

i

\t u re

of interrogative
feel.

shading.

This latter we perhaps are too prone to

B.— Dependent
('apt. 100-01:

on Verbs of Action and

Effort.

Homines captivos comrm
Cist.

rcatur,

si

nieal
filium.

aliquem invenire suom qui mutet
183-84:

lubei ilium

eundem
ff.

persequi, siqua queal

reperire quae sustulerit.
Cist.

184

:

ei rei

nunc sua
dot,
si

operam usque assiduo servos
nieret

possiel

rieem

illani

invenire.

Tri. 531-32:

Em
si

istuc oportet opseri

mores malos
interfieri.

in

opserendo possint

It

is

noteworthy that
<{uire

in this

group the verb of the

si-clause
is

is

always

or posse.

The thought of these clauses

akin to

the purpose idea, but with a large

admixture of doubt as to the

A purpose clause with a parenthetical "if possible" or the like would in most cases provide a fair
attainment of the goal.
rendering for the thought;
e.g.

(Capt. 100-01). "lie

is

buying

up

prisoners, thai

if

possible he
In Cist.
i.

may
184

light
ff.

on one who can be

exchanged for

his son."'
<

the si-clause appears to

be ;m expansion of

i

r<

Though
it

the interpretation of such sentences

is

not difficult,
si.

is

hard

to

determine
feel

in a

given case the precise shading of
force,

We
the

can readily

something of conditional
to

as

though

word were chosen

convey uncertainty with regard to the
Ai the
[,

attainmenl of the purpose.
1

same time the English mind

Student unci's Stuilia, Vol.

p.

195.

Vol.

l]

Nutting.

— Studies

in thi

si-clause.

'77

is

not slow here too to find the suggestion of interrogative

meangoing

ing.

For.

in

colloquial speech, with just such a virtual purpose

idea to express

we

freely use the interrogative; e.g.,
if
I

"I

am

to the city (to sec)

can secure some tickets,"

i.e.,

"to secure

some tickets if T can." The interrogative shading is most obtrusive when the action of the main clause is a simuesied experiment as in Tri. 531-32; there we may assume thai other means of suppressing vicious practices have been tried, and the speaker now jocosely suggests that it would he well to make the experiment of
planting them
in that

fatal field (to see?)
off.

if

they

too, as well as

other things, will he killed
(b).

Sentences of other forms.

The remaining cases of object si-clauses containing the subjunctive and dependent on indicative forms are so few and scattering that they can be best presented under this general head.
1

A.— Dependent
A sin. 528-29:

on Verbs of Expectation and Waiting.

An
te

te id exspectart

oportet, siquis promittat tibi
si

facturum divitem,
Poen. 1391-92:

moriatur mater sua

.'

lam pridem equidem
et

istas scivi esse liberas

exspectabam siqui eas assererei manu.
Ps. 1148:
des, porrexi

Iamdudum,
Vid. 68:

si

manum.

Hie

ostalxt

atque observabo, siquem amicum conspicer.
are

These sentences
si si/

manifestly

like

those cited of the

form

est.

B.— Dependent
('apt. 27-28:

on Verbs of Action and Effort.

Coepil captivos commercari hie Aleos,

siquem reperire posset, qui niutet suom
82

(sc.

lilium

.

tains the

Another example is probably to ambiguous form opperiar.

l"'

found

in

True. 692-93, but

it

eon

University of California Publications.
Merc. 622

[Class. Phil.

IV.

Quin percontatu's hominis quae
i|iii

facias

ford

illam emissel
.'

:

eo

si

pacto possel indagarier

mulier

Mil. 1207-08:

Nam
impetrare,

si

possem alio modi)

m

abirel nee te abduceret,

operam

dedi.

Tri. 119-20:
ei

rei

operam dan

te fuerat

aliquanto aequius

siqui

probiorem facere posses.

Yid. 56-57:

Ibo
sen

e1

quaeram, siquem possim socioruin nanciscier
qui advocatus adsiet.

quem norim

Cf. Most. 837-38:

At tu isto ad von optuere,
conspicari,
si

quoniam corniceni nequis

volturios forte possis contui.

Amph.
Mercurium
.Mil.

880-81:
iussi

me

continuo

<<>ns<<iui,

siquid vellem jmperare.

1158!

PA. Date modo operam.
V( in' in lis.

AC. Id nos ad

te.

siquid

velles,

In this group there
the si-clause that

is the same virtual purpose idea underlying was found in sentences of the form si sit—est; is

and. as there, the verb of the si-clause
the exceptions being the
of velle. 33
is

regularly posst (quin

),

last
is

two cases

cited,

which contain forms
si

Ilei-e

too

it

impossible to decide to what extenl

interrogative.

In Mil. 1207-08
it

clause precedes) makes
in
;1
.s-/.

difficult

however the unusual order (sito feel any interrogative fore

In Tri. 119-20 ei rei again anticipates the si-clause.

Before attempting to solve the problem of subjunctive "protasis" with indicative "apodosis" for sentei
s

containing sub-

junctive objed clauses,

it

will he

necessary to consider also those

There would be further exceptions if we should include Aul. 620-21 (perscrutabor, si inveniam) una Pers. it (quaeram, si<|uN credat) these are excluded because of the presence of forms in -am. Cf. also Merc. 941, si. i.-,i 52, the corrupl Cas. 806 and doubtful Amph. 621.
;

f. Blase de mod. temp, permut. p. 22 (78). Lindskog uithuut advancing any satisfactory evidence, is very decided proval of Blase 'a position.
1

(1.

c.

i>.

in

his

73). disap-

Vol. l]

Nutting.—Studies

in tht

Si-clause.

7!)

cases in which an indicative objed clause
tive

is

used.

The subjunc-

examples were subdivided according to the nature of the verb of the main clause; (A depending on verbs of expectation and waiting, (B) depending on verbs of action and effort. A
)

similar plan will be followed here; bu1

necessary to add
ing.

(('

)

is is lacking, and .1 —depending on verbs of seeing and knowi1

We

therefore begin with

B.— Dependent
.lust as in
is

on Verbs of Action and

Effort.
.s/'-clause

the case of the subjunctive the verb of the

here also regularly posst

Bacch. 1151:

Ego ad nunc iratum adgrediar,
inlicere hue.
Cist. (551-52:

si

possumus nos hosce

intro

Ibo,

persequar iam ilium intro.
si

ut

haec ex

me

sciat

eadem.

possum
:

t

rampiillum facere ex irato mihi.

Cure. 701

Animum

advortiU hoc.

si

possum hoc

inter vos componere.

Men. 417-18:
adsentdbor, quicquid dicet, mulieri,
si

possum hospitium nancisci. Men. 1048-49:
ibo intro ad

Nunc
sei

banc meretricem,
ut

quamquam

suscenset mihi.

possum exorare
Rud. 890-91

pallam reddat.

Verum tamen
Tri. 921
:

ibo, ei

advocatus

at siera,

siqua mea opera citius

— addici

potest.

Quod ad exemplumst?
Tri. 958-59:

coniectura

si

reperire possumus.

Enim
si

vero ego nunc sycophantae huic sycophantan volo,

hune

possum

illo

mille

nummum

Philippum

circum-

ducere. 35
In
tliis

group belong also
find a parallel in
I.

which
'

a few conventionalized si vis clauses two subjunctive examples already cited

Km-

St. 740-41,

329 is doubtful in text and meaning. Cf. also Poen. 1063-64 and whirh should perhaps come under this beading.

30

University of California Publications.

[Class.Phil.

(sivellem, Ampli. 880-81,

and

si velles, Mil.

L158

.

All the cases
lisl

iveu perhaps do qo1 contain

objed clauses, bu1 the
qo chance of excluding

is

made complete

so

there

may

be

whal

should be included.
Aul.
l'h!):

Redeo ad

te,

Megadore, siquid me

vis.

Capt. 618:

Do

tilii

operant, Aristophontes, siquid est quod

me
Pers.

velis.

.Men. 566:
till
:

Em

hie abiit,

si

vis persequi vestigiis.

Adduco

hanc, siquid vis ex hac percontarier.

Poen. 207-08:
Poen.

Em

amores
itast,

tuns,

si

vis spectare.
si

1047-48: si

tesseram conferre

vis

hospi-

talem, eccam attuli.
Tri. 516-17:

ST. Philto,
this

te volo.

PH. Siquid

vis.

Stasime.
corre-

As the sentences of
S]

group are compared with the
s«-clause,
it

ling

examples with subjunctive
one looks
in

must be conIt
si-

fessed thai

vaiu
if

for

a

difference of meaning.
is

may

be noted however that
is

a past tense

to be used in the

clause the subjunctive
tive cases jusl

the

mood chosen; 36
tense.

for

all

the indica-

cited

employ the present

C— Dependent
Bacch. 529:
ibo ut
('as.

on Verbs of Seeing and Knowing.

visam hue ad eum,
591
:

si

fortest domi.

Viso

hue

amatoi'

si

a

fioro rediit

domum.

Men. 142:
Tarn sciii, n. siquid titubatumst, ubi reliquias videro.

Mer. 155-56:

Quin iam priusquam sum eloqutus
Pers. 825:

scis, si

mentiri volo.

Vidi

vi'i'ii.

si

tibi satis placet.
p.

"Lindskog
cially with
I

(1.

c.

69)

reference a verb in sons the subjunctive. aotices one exception explaining it away bj facere, ui addicatur. is an exception to the Vid. 56-57, where the
ire,

to posst

the

first

So

this distinction. Further tie adds (espequirt that when there is a reference to person takes the indicative and in other perBu1 Lindskog himself Lindsay, Capt. 28 note.
J

makes and

to the Lattei

pari

of

tin-

rule

(immdv

Ku<l.

s'.»<>

<t]

),

saying bhal mea opera addici potest is equal to possum This is qo1 altogether satisfying, especially as there other pari of the rule which he does doI aoti.ee, aamely first person subjunctive possim is used.

vol.1]

Nutting.

—Studies in

thi

8i-clause.

81

Tri. 748:
Yi<li
si

hoc utibile magis atque

in

rem deputas.

Tri. 763:

Sed

vidi

consilium

si

placet.

37

In this group the nature of the verb of the main clause suggests

most strongly interrogative force for lit.") holds thai even here the word
i

si.

Becker however
no1

(I.e.

p.

is

fully

interrogative.
in

For, he says, an undoubtedly interrogative

word

some of the

above cases would demand the subjunctive, according to Plautus' usage; here only the indicative
is

found.

Treating only those cases which contain undoubted indicative

and subjunctive forms,

Plautus'

osage

in

object clauses

may
1. is

he thus presented in tabular form.

After verbs of expectation and waiting the subjunctive

used.
2.

After verbs of action and effort the mood varies.
(a) In the present tenses both

moods of

posst

are used;

quirt

stands in the subjunctive, velle in the indicative.
(b)

In past tenses the subjunctive of posst

and

vellt

is

employed.
"..

After verbs of seeing and knowing the indicative
the help of this outline
it is

is

used.

With
in
it

possible

by

a process of exclu-

sion to arrive at the probable cause of the use of the subjunctive object clauses.

For

it

may

be remembered that

in

such clauses

was generally found to be true that the force of

si

was waver-

ing between conditional and interrogative.

The

table just given
to the condi-

shows that the use of the subjunctive must he due
tional

force of the

word

i.e.,

that this
it

mood
in

was

chosen

in

accordance with the rule that called for
sentences.

regular conditional

For

in

group 3 (after verbs of seeing and knowing),
is

where the interrogative shading
of the si-clause
is

most pronounced, the mood

always indicative.
1

The weaker interrogative

coloring of

si

in

groups

and

'_'

cannol therefore have been the

factor that caused the frequent use of the subjunctive there.
Cist. 6Si; is

doubtful

in

meaning.

University of California Publications.

[Class-Phil.

IV.-THE INDEFINITE SECOND SINGULAR.
Bacch. 440-41
Ai
:

nunc priusquam septuennis
Capt. 202:

est,

si

attingas

eum mami,

extemplo puer paedagogo tabula disrumpil caput.
mala animo
-2-2}
:

In re

si

bono utan

.

adiuvat.

Capt.

Nam
Quia

doli

qod
:

doli sunt, nisi astu colas.

Cas. 721

quod
fiT4:

tetigere,

ilico

rapiunt:

si

eas

ereptum,

ilico

scindunt.

Ep.

Quaque

tangit,
10:5

omne amburit.
si(|ui<l

Si ashs. aestu ealefacit.

Mm.

Standumst

in

l<'<-to.

de

summo

petas.

Mil. 673:

Nam

in

mala uxore atque inimico siquid sumas, sumptus
sobrie aut frugaliter

est.

Pers. 449-50:

Siquam rem accures
solet ilia rectc

sum manus
id

succedere.

Poen. 635-36:

Malo

si.|iiid

bene facias,

beneficium interit.

Bono

siquid male facias, aetatem expetit.

Poen. 812-13:

Siquid bene facias, levior plumast gratia.
Siquid peccatumst, plumbeas iras gerunt.
Tri. 349:

De magnis
Non
lil>i

divitiis siquid
."»:

demos, plus

lit

an minus?

Tri. 414-1

illud apparere,

si

sumas, potest,

nisi in

Lmmortale rere

esse

argentum

til>i.

Tri. 1053: Si

mage exigere
True. 4U1-62:

<><<i/>i<is.

duarum rerum

exoritur optio.

Nullaui rem oportel dolose adgrediri
nisi

astute adcurateque

'

xst

quart
si
sit

.

To

these sentences of the form
si

est

apparently should be

added one of the form

sit—erit:

Vol.1]

Nutting.

—Studies in the Si-clause.
veils

83

Amph. 703

ff.

Bacchae bacchanti

si

advorsarier,

ex insana insaniorem facies, feriet saepius.
Si obsequare, una resolvas plaga.

Though forms
the

in

-eris

are strictly speaking of uncertain mood,

two

following cases
:

may

be

a1

leasl

enumerated

in

this

conned

ion

Poen. 212-13:

Nam

nullae magis its duae plus negoti
si

hahent, forte
Tri. 1051:

oca

/» ris

exornare.

Siquoi

mutuom quid
final

dederis,

fit

pro proprio perditum. 38

A

full

and

uaturally start with the subjunctive of the si-clause.

explanation of the form of these sentences would But unforis still

tunately the aature of this subjunctive
tainty,

a

matter of uncer-

and the material

a1

hand

is

far too scanty to form the

basis of

conclusion
the

any adequate conclusion on that point. To reach such a it may be necessary to compass the wide field in which
of the coneomitanl
relation

phenomenon

between indefinite

Howsecond singular and subjunctive mood manifests itself ever, thai there is a cause and effect relation involved cannot
I

think be for

momenl doubted, the upholders of the other view For so sweeping is the tendency of a verb notwithstanding.
a

whose subjed

is

the indefinite second singular to go into the sub
a

junctive that Plautus offers hut
si (S (

single

example of the form
si sit

— es

t

to

compare with the fourteen above of the form

—est:

Asm. 241-42:
Portitorum simillumae sunt ianuae lenoniae: si mm est quod des, aedes non patent. si <nlfi rs, turn patent
;

Again
is

a

full) is suggestive

comparison of Poen. 812-13 and 635-36 (given above in in the second of these passages an alternative
;

afforded by siquid hern fori, is and siquid mah facias, while in the Other exactly the same thOUghl findS expression in tile clauses
siquid hern facias and siquid peccatumst.
If the indefinite sec-

ond singular has nothing
Tri. 347-48 has
;<

to

do with the use of the subjunctive
in

hortatory subjunctive

acidosis:

'•!'.

Aul.

> lls

l.

g4

University of California Publications.

[Class.Phd..

it

is

hard

to

accounl for the choice of moods here.
the variation in

Though
in

no1

in si-clauses,

mood

is

quite as striking

the two

following cases
Mil. !»47:

Volup

est,

quod agas

si

id

procedil lepide atque ex sententia.

Poen. 1192:
It volup es1 horaini,

mea

soror,

si

quod

agit

duel victoria. 89

Accepting as

a

fad ool ye1 satisfactorily explained the subis

junctive of the si-clause when the subjed

the indefinite second

singular, the problem of subjunctive "protasis" with indicative

"apodosis" for the sentences under discussion

is

to

determine

why
are
it

ilif

subjunctive si-clause does oo1
a

level

its

conclusion.

One
There

Looks in vain for
is

clear case of such levelling in Plautus.

true sentences like the following:

Cist.

33:

Eas

si

adeas, abitum

quam aditum

malis. 40

Bui the subjed of the verb of the conclusion seems always to be
;is

here the indefinite second singular, and the cases therefore
n<>

give

proof of the workings of a levelling force; for such a

conclusion

may

take the subjunctive on
in
ff.:

its

own

merits, as

is

shown by examples
Bacch. 913

which an indicative si-clause precedes:

Lippi
si si

illic

oculi servos est simillimus:

non
est,

est,

noUs esse aeque desideres;

abstinere quin attingas non queas.
ti

('apt. 11

IT.:

Liber captivos avis ferae consimilis seme] fugiendi
satis est
si

est

:

datasl occasio,
postilla possis prendere.
in
is

numquam

The reason why

the subjunctive si-clausi
not
level
its

the sentences
to be

under discussion does

conclusion

found

in

the nature of the underlying thought.

The

si-clause refers to an
a1

action which the speaker assumes does happen.
ally,

least

occasion</////."

and
The
I

si
shifl

is

therefore practically
in

a

synonym

of ubi or

passage Bacch. 4^t! IV. and in Tri. ill L5 may have been caused by passing from the definite to the indefinite sec singular and \ ice versa. '" other eases are: Amph. 705 and Tri. 1053-54. Asin. 120-2] is similar lnii s hortatory force. See again Amer. Jour. Phil., Vol. XXIV, p. 300 IV.
in

mood

the

long

1

1

;

i

L

vol.1]

Nutting.

Studies in thi Si-clause.

85

The conclusion has

to

do with

a

second ad or state
in

which

is

broughl aboul by thai referred to

the si-clause.

This second
a1

ad

or state

is

accordingly also one thai does actually occur

and the indicative of the conclusion is simply a recognition of that fact. The si-clause serves to define the circumstances of the occurence, just as an ubi- or (////(-clause tnighl do, and the
limes,

mood
clause.
In
si

of

its

verb seems to exercise aboul as

Little

influence on

thai of the conclusion as

would

thai of a subjunctive ubi- or

cum-

conclusion
in
:

may

be mentioned two sentences of the form
n /-clause is

sit— est class name

which the subject of the verb of the

a

Bacch. 447-4S:
llocine hie pacto potest

inhibere imperium magister,

si

ipsus primus vapulet?

True. 234:

Xuuae
[f

sunt, nisi

modo quom

dederit, dare iam lubeai denuo.

the Context Of these passages he
first

examined
class

it

will

he

found

that in the

example

>n<i<)is(< r is a

name "the

master."'

and in the second the subject of discourse is amator "the Lover." These sentences, especially tlie latter, suggest the query whether
the indefiniteness that lucks in
a

class

name

is

no1
If
so.

akin to the
the modal

indefiniteness of the genera] second
4 peculiarity of occasional
-

person.

cases

like

these mighl

be explained

on that analogy.
V.

-LOOSELY ATTACHED CLAUSES.
(a).

The

si

scias type.

Merc. 298-99:

Immo
oculeis

si s<i<ts.

quoque etiam plus iam video quam
magis senex,
si

prius.

Merc. 44o

Multo hercle
I

ille

\u scias.
a

venture to bring these two sentences under
is

special heading
is

because the si-clause

an idiomatic phrase which

capable of

fund toning alone
ind

;

i

..'/..

'Tlie indicative Merc. 744.

is

more common.

See Aul.

'.'AT.

Cure.

142,

Men. 576

86

University of California Publications.
Cas. 668:
I

[Class.Phil.

iiiniu si

si

ms
:

dicta

quae

dixit

hodie.

Cure. 32]
I

mmo
I's.

si

scias reliquiae

quae

sinl

74!):
u1

PS.

Probus homostj
scias.

praedicare

te

audio.

CH. [mmo

si

Cf. Bacch. 698:

[mmo
Ps.

si

audias quae dicta dixit
(b).

tne

advorsum
tvpe.

tibi.

43

The

si

modo

DDT:

Id ago, si taceas Tri. 1187:
I

modo.

)icis. si

facias

modo. 44
set
is
;i

These sentences likewise have been
clause containing the subjunctive

apart because the

si

modo

half independent sentence

element, almosl an expression of wish; cf.:

Capt. 996:

Quod male
(

feci,

erucior:

modo

si

infectum

fieri

possiet.

"as.

742-43

LY. Quid nunc? quam tnox recreas me?
<

>!,.

(

Vim modo
976:
ilia

si sit

cocta.

Ps.

Nam

mea sunt cognomenta

:

nomen

si

memoret modo. 45

VI.— MIRARI (MIRUM) IN APODOSIS.
Cuvr. 265:
Nil est
Ps.
Xc(\
si

mirandum, melius
433
IV.:

si

nil sit tibi.

sint ea vera, nt
.'

mine

nios est, inaxiiiiie.

quid
si

mirum fecit quid novom, adulescens homo amat, si amicam liberal
.'

f. Asin. 744. The use of these and similar phrases in regular conditional sentences (Bacch. 678, Ep. 451-52, Mil. L429, Tri. 538) may perhaps throw some lighl on the two sentences above in which they are loosely
<

attached.
.

Rud. 680, and possibly 552.
I

.

ist.

734.

Vol. l]

Nutting.-

-Studies in the Si-clause.

s;

The number of
,-i

cases in this category

is

too small to justify here
!!< regularly

complete exposition of Plautus' usage.
in

employs

the indicative

the st-clause,

and these are
I

bu1

two scattering

46 cases that have strayed across the line.

am
si

therefore contenl

to

have merely quoted them here: they would be naturally treated
genera] discussion of the idiom mirari

in a

rather than

in

one

of "subjunctive protasis with indicative apodosis."

With regard
say thai
I

to this

paper

in

general

1

perhaps need hardly

do no1 share the hope which seems to characterize
be found which

most of the Inter work on this subjeel thai some sweeping explanation
tive

may

is

valid

for all cases of

protasis with

indicative apodosis."
felt

"subjuncOnly <m the assump-

tion that

Plautus

this

form as

a

linguistic unity could

we

rightly hope to find

any such general explanation; and that he
improbable.

did so

feel
at

it

is.

in

view of the wide variation of the underlying

thought,

least ver\

The

division into conditional

sentences, concessive sentences, etc. seems to

me fundamental,

and

I

have therefore

in

each of these groups based the explana-

tion of the
I

form on the nature of the thought to he conveyed. would here take up one more topic which lias been postto the

poned

cud of the discussion

in

order that
matters

it

might not
inserted
in

distract
its

attention
|

from more important

if

logical place.

refer to the old problem of the difference in

question
tional

meaning of suppositions of the forms si sit and si erit. This is raised especially by what was said of the pure condisentence,

namely
likely
t

that

Plautus'

failure

to

differentiate
in

sharply between the uses of the two mood systems

general
inter-

would he most

to
t

betray

itself
in
[

iii

the somewhat

changeable value of

hese

wo forms
Thai
si sit
lie

>a rl icii la r.

the time realm
a

of both being the future.
extent between the use of
I

did differentiate to
si erit
is

certain

and

unquestionable, and

would suggesl that the differentiation was partly on an objecpartly on
a

tive,

subjective basis,
in

i.e.,

that

Plautus tends

to use

the subjunctive

the two following cases:

"Lindskog (I. e. p. 65 seems nut fco recognize Ps. 433 tl'. .-i< belonging in this category, thus leaving Cure. 265 as tin' only example of the use of the subjunctive. To remove this exception to the rule he suggests that will: B wo read fit fur sit.
l

University of California Publications.
(a)
th)

[Class.Phil.

When When

there

is

actually

Less likelihood

of fulfillment.

the speaker aims to giv< am impression of less like-

lihood of fulfillment.

The

firsl

of these cases

is

mosl strikingly illustrated by concessi
sit

sive sentences of the

form
in

erit.

In a1

least

three of the

four examples found
is

Plautus the supposition of the si-clause

extremely improbable.

Amph. 450: Quadrigas
detinuisse.

si

aunc inscendas

Eovis.

Asin. 414: Siquidem hercle nunc

summum

[ovem

te dicas

Bacch. L004:

si

iubeas maxime.
sese

Ep. 610:

Si

undecim deos praeter

secum adducal

[uppiter.

The peculiarity of these subjunctive cases may be brought out into relief by contrasting the corresponding indicative examples.

of the si-clause
si
<

Counting as concessive one sentence in which the function is somewhat complicated, Plautus uses the form
i

rit

rit

twice:
ff.:
T l

Amph. 1048
si

bi

quemque hominem aspexero
videbo, obtruncabo in aedibus.

aneillam, sen servom, sive uxorem, sive adulterum,

sen patrem, sive
('apt.
Si

avom

683-84:
ille

ego hie peribo, ast

ut dixit non redit,

at eril In

mi hoc factum mortuo memorabile.
si est

force, there are doubtless
si

is a form that often has future some concessive sentences of the form est— <nl which should be added to the two of the form si erit

view of the fact that

erit
firsl

before making
1

a

comparison with the subjunctive cases
list,

cited.

give the complete
in

leaving

it

to the reader to

choose those sentences

which the form

si est

seems to him

to

have future meaning.
lie

clear that the
is

Whatever the sentences chosen it will still sunjunctive tends to be used when the supposiimprobable, which
is

tion

extre

ly

the point

1

am trying

to

illustrate.

Asin. 405-06:

Siquidem hercle Aeacidinis minis animisque expletus
si

ced.it,

nied iratus tetigerit, iratus vapulabit.

Vol. i]

Nutting.

— Studies in the Si-clause.

8!)

Men. 1060-63
Si voltis

:

per oeulos iurare, nilo hercle ea causa magis

facietis u1 ego hine hodie abstulerim pallam.
.Most

229-30:
\renibi1

Siquidem hercle vendundusi pater,

multo potius

quam

te

.

.

.

sinam egere.

Hud. 1014:
Sei lu proreta
isti

navi's, ego gubernator ero.

An

illustration of Plautus'

tendency to use the subjunctive

when the speaker chooses
of fulfillment
sit f

to give ;m impression of unlikelihood
.

is afforded by phrases of the form Quid si Not including' the corrupt Cas. 806, there are nineteen questions of this sort in Plautus. For the present purpose they may be subdivided according to person and number.
.
.

(a).

First

person plural.

Cas. 357-58:

Quid

si

propius attollamus signa eamusque obviam

.'

Sequere.

Cure. 303

Quid Quid

si

adeamust

heus, Curculio, te volo.

Cure. 351
si

abeamus, decumbamus? inquit.

Consilium placet.

Most. 393:

DEL. Quid
1

si

igitur

abeam us hine nos? TR. Non hoc

longe,

>elphium.

Poen. 330
.\(i.

Quid

si

adeamusf
ff.:

MI. Adeas.

Poen. 707

Quid

si

evocemus hue foras Agorastoclem
testis sit sibi cert issimus
.'

Ut ipsus

Ileus tu, qui furem captas, egredere ocius.

Poen. 1KiL -(i3:
)

Quid

si

eamus

illis

obviam

.'

AG. At ae

inter vias

praeterbitamus metuo.
Poen. 1249:

HAN. Quid

si

>

loquamurf

AG. Censeo,

hercle, patrue.

90

University

of

California Publications.
general to our

[Class.Phil.

These questions correspond
thus and so"
sion thai
lus
a

in

"Suppose we do

form which leaves with the hearer the impresis

wish or judgmenl

being consulted, and thai the
is.

coming

to

pass of the thing suggested

from the speaker's
in

point of view, anything bu1 assured.

Bui while

some of the

above cases the bearer shows by his expression of approval or
disapproval thai be feels himself consulted,
58,
in

others (('as. 357-

Cure. 303, and

Poen. 707

Bf.)

the speaker really does not
all,

defer to his wish or judgmenl

at

bu1 without
is.

a

pause pro-

ceeds to do the thing suggested.

Thai
a

in

certain eases the
still

speaker even though he fully expects
uses in a

thing to he done,

somewhat perfunctory way
to consult

a

subjunctive phrase which

appears

the wish or

judgment of the hearer.

(b). First person singular.

('apt.

612:
ais? quid
si

HE. Quid
Cist.

<t<l<am

hunc insanum?

TYN. Nugas:

Ludificabitur.

821:

Quid

si

adeam atque

<i/>/>< lit

m

.'

Mali damnique inleeebra,

salve.

Cure. 145
I'll.

Quid

si

adeam ad

fores atque

occentemf

PA.

Si lubet,

net pie

veto neque iubeo.

Ep. 543:

Quid

si

adeam
si

Pers. 724:

To. Quid

admoneamf

VI.

Tempus

est.

Poen. 728:

AG. Quid

si

recenti re aedis
:

pultem?

ADV.

Censeo.

Rud. 535

CU. Quid si aliquo ad ludos me pro manduco locemf LA. Quapropter
.'

True. 6:

Quid

si

de vostro quippiam oremf

abnuont.
in

With these may very properly be enumerated the single case which the perfecl subjund ive is used:

voi, i|

Nutting.

Studies in the Si-clause.

91

Capt. 599:

BEG.

fHercle quid
Sapias magis.

si

hunc comprehend] iusserimf

TYM

In scvcrnl of these eases the
thai he
is

answer shows
1<»

thai

the hearer feels

consulted with regard
in this

the speaker's action.
is

There-

fore the quest ion
ential address.

mi tuber also
in

properly

a

form of defer-

'That

however

some of the eases the speaker judgmenl
is

did not really

mean
such
in

to defer to the hearer's

rendered

probable
tion
is

by

an

example

as
is

Cist.

321,

where the ques-

spoken

soliloquy and

practically an

announcement

of the speaker's intention

a1

any

rate he at once proceeds to do
is

the thing mentioned.

Whenever

this

true

it

provides another

illustration of the use of the subjunctive in vive the

appearance
he state-

of deferring to the hearer's judgment.

Three cases remain which must be added

to

make

1

ment complete

:

(c).

Third person singular.

Bacch. 781-32:
.MX. Quid scribam
PI.
.'

('11.

Salutem tuo patri verbis

tuis.

Quid

si

potius

morbum, mortem scribat?
illi

id erit rectius.

Merc. 41!):

Quid

si

igitur reddatur
Kili

unde emptast?

True.

Sed quid ego hie clamo? quid
In the
first
first

si

me
is

iubeat intro mittier?
clearly analogous to the

of these sentences scribat

person use

— the

action proposed

is

put

forward as

a

mere

suggestion, here not by the actor himself but by another for him
as
it

were.

In the second case the verb
first

is

passive and

1

he action

devolves upon the

person:

in
in

properly be classed with those
is

meaning the sentence would which the subject of the verb
it

the

first

person.
to

to
its
t

have uothing
likeness of

seems The third example is unique, and do with the idiom under discussion aside from
its

form:
si

force
7

is

akin to thai of indicative ques-

ions of similar

rud

lire.'

This completes the discussion of the difference of meaning
of the forms
17
I

si sit

and

si erit,
Ps.

and the paper migh1 be closed
I

a1

740 because it seems to have do exad parallel either among the subjunctive or 'lie indicative cases. have accepted the punctuation Quid? si opus sit ut dulce promat indidem, ecquid pure conditional sentence. habet? and have treated the case ,-is
(Uiit

from the eumeration

:i

92

University of California Publications.

[Class.Pheu

this point.

Bu1 having given the material

in

full

for questions
to

of the form

Quid

si

.

.

.

sitf

I

oughl perhaps
si
. .

add
.

for the

sake of comparison those of the form Quid

est

(erit)f

Because of strongly Idiomatic use such
Little

a

comparison throws very

dired
in
is

lighl

on

1

1 1 * -

question

last

under discussion (the dif
si sit and si erit), and the matter seems to

ference

general between the meaning of
in

hut

it

interesting

and

I'm-

itself,

be nowhere fully treated. 48
tive
is

The characteristic

force of the indica-

seen most clearly

in

the following examples:

Asm. 536-38:
CL. Non volo ted amare qui dant, quoia amentur
gratia.

PH. Quid
EV. Potin

si

hie

animus occupatust, mater? quid faciam?
sis

Mere. 890:
ut

animo

tranquillo?

CHA. Quid

si

mi animus

fluctuat?

IV is. 612-13:

DO. Enim volo te adesse. TO. Ilau possum, quin huic operam dem hospiti quoi erus iussil Quid si hie non volt me una adesse.'
I'oen. 72.1-22:

AG. Quid mine milii auctores estis? AG. Quid si animus esse non sinitf
Rud. 1085-86:
et

ADV.

IT frugi

sis.

crepundia.

(JR.

TK. Nil peto nisi cistulam Quid si ea sunt aurea?

Rud. 1138-39:

Quid

si

ista

ant superstitiosa au1 hariolasi atque omnia
.'

quidquid

inerit vera diet

I

Tri. 1059-60:

ST. Quid
Questions

si

ego

OH. Te volo. me te velle nolo?
polite
;i

like these ;ire not

and deferential phrases.
protest

On

the contrary they verge toward
Lindskog
(1.

againsl the expressed

11 '! (it' the subjunc ff.) gives incomplete lists, p. Mer. int. Poen. L249, .-mil True. 766; et' the indicative cases, Ani|.h. 7(H. Bacch. 35, Ps. 286, Rud. 1086 and L138 (two of employ the perfeel tense which he does ao1 treal al all); of cases of Regarding the subjunctive am), Rud. 1-7 and \:\\-. ambiguous form he says (p. 109) "Rei oatura lit. ut semper praesens coniunctivi asurpe >. Brugmann (I. c. p. 27) touches on tur;" Inn apt. 599 has iussenm. Cf. Bris on Capt. 613 and this subject, but with very incomplete material. Sonnenschein on Kml. 472.

e.

I

tive

c.-ises

he miiits

(

1

«

<

Vol. l]

Nutting.

Studies in the Si-clause.
person
addressed.
In

93
translation

desire

or advice of the

we

instinctively recognize this Ead
tive particle

by beginning with an adversa.

"But whal

it'

.

.

V

The lone

of the question

may

be even insolent, as
a

in

the

lasl

case cited."

Other examples of
feeling of protest

similar nature bu1

with the speaker's
so clearly

or hesitation perhaps not

marked

are

Amph. 391-92:
so. Tune
so. Quid
si

li.lei

credo?

ME. Meae.

fallesf
iV.
:

Asin. 193
Si

mihi dantur duo talenta argenti numerata
tibi

in

manum,

hanc

noctem honoris causa
si

gratiis

dono dabo.

AR. Quid

uon est?
aon fexcruciem, alterum tantura

Bacch. 1184-85:
XI. Quern quidem ego
auri non meream.
nt

BA. Quid tandem
Cas. 269
ft'.:

si

dimidium auri

redditurt,

CLE. Quid
nt earn
illi
I

si

ego impetro atque exoro

a

vilico,

causa men

permittat?

LY. Quid

si

ego autem ab armigero

impt rn
nt earn
illi

permittal

.'

.Mere. 907-08:

CHA. Opta
KV. Quid
si

ergo oh istunc auntium quidvis

tibi.

optabol
ff.:

Most. 580

TR. Reddet: uunc abi. DA. Quid ego hue recursem autoperam sumam ant conteram Quid si hie manebo potius ad meridiem? 50
is

.'

In these cases the characteristic force of the indicative question
least

clear in Bacch.

1184-85, which shades off toward the

Lindskog's definition
usurpatur,
.

cum

(p. L07) semis to me Ear too vague "indicativus quia quaerit, quid futurum sit. si quod in protasi contineatur

sven< Tit.

Some would
and
if so

belongs probably also the somewhat complicated Amph. 849 tV. include Bp. 599; but the si-clause seems here to be com the sentence should be punctuated Quidl si servo aliter visumst, aon poteras novisse, obsecrol Aul. 776 has been emended to provide still

Here

another case.

94

University of California Publications.
.Must.

[Class.Phil.

meaning of the subjunctive sentences.
At

580
. .

ff.

is

compli-

cated by the interjection of the words Quid
a
lii-si

.

conteram?
no1

reading the exacl Eorce of Cas. 269
ii

ff

may

be evi-

dent.

Bu1

will be ooticed thai

the verbs of line 269 are impt

tran

and exoran
in

(no1

peto or the like': this assumption
thai

of

success

the appeal
in a

inclines one to believe

the question

was spoken
betrayed
.-ill

taunting and exasperating lone

Lysidamus has
succeed in
indic-

too clearly his intention with reference to the mar-

riage of Casina,

and

his wife retorts,
to give her

"But whal

if

1

inducing the steward
ative has
th<'
still

up

.'"

So interpreted the

oormal and characteristic meaning above described.
remain two cases of the form Quid
si

There

— est?

Bacch. 35

BA. Quid
ide
:

si

hoc potis
licet.

est

u1

tu taeeas, ego eloqar?

so. Lep-

Men. 844:

MA. Quid
The meaning
like thai
a

est

.'

Quid agimus? SE. Quid

si

ego hue servos citof

of the second of these examples seems precisely

of the subjunctive cases.
is

Bacch. 35, coming just after

Lacuna,
to

partially devoid of context; but the

meaning here too

seems

51 approach closely that of the subjunctive question.

It will be

remembered that

all

the subjunctive cases exceptI

ing Capt. 599 (iusserim) use the present tense.

have therefore

compared
.

them
.'

with indicative cases of the forms
. . .

Quid

si

.

like

and Quid si There are time realm.
.

<

si

eritf as these
also a

have to do with a

few indicative cases which

employ other tenses: they are Amph. 701, Asin. 720, Ps. 286, 514,
and Rud. 721. 52
Ldndskog (1. c. p. Ill), having omitted from his enumeration Bacch. To avoid naturally finds the only exception to the rule in Men. sit. The line the exception he suggests thai cito is adverb rather than verb. use of the indicative and subjunctive was doubtless not absobetween the It may be remembered thai of the subjunctive cases lutely bard and East. True. 766 approaches close to indicative meaning; this case also was
omitted by
L.

35,

Aside from these there are six of Asin. L05 is doubtful. examples which contain ambiguous forms in am, namely Amph. 313, Merc Most. L093, Rud. l_7t. and L311 IV.; nil excepting the lasl have 564, 578, subjunctive force. Two cases have rerbs terminating in erii (Cas. 345 and with the form faxis (Mil. 1117); these three have 172 ll.) .-nid

The

texl

indicative force.

UNIVERSITY OF CALIFORNIA PUBLICATIONS

CLASSICAL PHILOLOGY
Vol.
1,

No.

3, pp.

95-109

May

19,

1905

THE WHENCE ANT) WHITHER OE THE MODERN SCIENCE OE LANGUAGE.
1

BY

BENJ. IDE

WHEELER.

It

cannol be the purpose of this brief paper to presenl even
a

in

outline
it

history of the science of language in the century
se1

past;
dii'eci

can undertake only to
its

forth the chief motives

and

ions of

development.
this

A hundred years ago
in

year Priedrich von Schlegel was
the

Paris studying Persian and the mysterious, new-found San;

skrit

Franz Bopp was
;it

a
;

thirteen-year old student

in

gymnain

sium

Aschaffenburg

Jacob

Grimm was studying
were

law

the

University of Marburg.

And

yet these three

to he the

men

who should
oni

find the paths
its

by which the study of human speech
in a

mighl escape from

age-long wanderings

wilderness with-

track or cairn or cine, and issue Forth upon oriented highas a veritable science.

ways

Schlegel the Romanticist,

who had peered
Ueber
di(

into Sanskrit literain

ture in the interest of the fantastic

humanism modish
a

his day.

happened

to

demonstrate

in

SpracJu uh<I Weisheii der
genetic relationship

Inder, 1808, beyond cavil the existence of

between the chief members of what we now know as the [ndo-

European family of languages.
this

Bopp2 found
in

a

way

to

utilize

demonstrated

fact

in a

<ptesi
set

which, though now recognized
operation the mechanism of
the

as mostly vain, incidentally

comparative grammar.
national

Grimm, 2 under

promptings of
the
and
1816.

a

enthusiasm, soughl
;ii

after the sonn-cs of
Louis

German
Sciences

'Address delivered October, 1904.

the

St.

Congress of

Arts

In-!

work:

Conjugationssystem
Vol.
1

der
L819).

Sanskritsprache,

Ventschi

Grammatik,

96

University of California Publicationsand, finding
in

[Class. Phil.

national presenl
forth

life,

in

language as

in lore

the roots of the

deep planted

the past, laid the foundations

and

sel

method of historical grammar. The grafting of comparative grammar upon the stock of historical grammar gave it
ilir

century.

wider range and yielded the scientific grammar of the nineteenth The method of comparative grammar is merely auxiliary to historical

grammar;
its

it

establishes determinations of fact

far

behind the point of earliest record and enables historical
to

grammar
It

push

lines of descent in the

form of 'dotted

lines'

far back into the unwritten past.

European scholars

was the discovery of Sanskrit to the attention and use of a1 the close of the eighteenth century thai

a

gave occasion to an effective use of the comparative method and consequenl establishment of a veritable comparative grammar.

But in two oilier distinct ways it exercised a notable influence upon the study of language. First, it offered to observation a language whose structure yielded itself readily to analysis in
terms
<>f

the adaptation of

its

formal mechanism to the expres-

sion of modifications of thought,

and thus gave an encouragethe
interest

ment

t<»

a

dissection

of

words

in

of tracing the national

principles of their

formation.

Second, the Hindoo

grammar
iiccuracy

itself

presented to Western scholars an illustration of
in collecting, codifying,

and completeness

and report-

ing the facts of a language, especially such as related to phonology, inflexion,

of a complete revolution

and word-formation, that involved the necessity in the whole attitude of grammatical
of Panini

procedure.

The discovery

and the Praticakhyas meant
path so clear
a

far more to the science of language than the discovery of the

Vedas.

The grammar of the Creeks had marked
a
all

a

and established

tradition so strong, guaranteed in

prestige so

high, thai the linguistics of the West through

the generations

faithfully abode iu the way.

The grammatical categories once
a

taughl and established became the irrefragable moulds of grammatical

thought, and constituted
if

system so complete
in

in

its

enslaving power that

any man ever suspected himself

bond-

age he w;is ye1 unable to identify his bonds.

The Greeks had ;iddresscd themselves
in

to

linguistic reflexion

connection with their study of the content and the forms of

Vol-

i|

Wheeler.

Tht

Modern Sciend

of Language.

!*7

thoughl

:

grammar

arose as the

handmaiden of philosophy.
it

They
as
a

assumed, without consciously and expressly formulating
doctrine, thai

language

is

the

inseparable shadow of thought,

and therefore proceeded without more ado to find in its structure and parts replicas of the substances and moulds of thought. They soughl among the facts of Language for illustrations of
theories:
it

did not occur to them to colled the facts
their

and organize
uses

them

to yield

own

doctrine.

Two

distinct

practical

finally

brought the chief materials of rules and principles to formulation in the guise of a system of descriptive grammar;
the interpretation of

first,

Homer and

the establishment of a
to aliens,

correct

text: second, the teaching of
a

Greek

and the

establishment of
uses
tical

standard by which to teach.

These practical

came

in

however rather

as fortunate opportunities for prac-

application of an established discipline than as the motives
creation.
a

to

its

With the Hindoos

it

was the direct reverse.

They had

sacred language and sacred texts rescued from ear-

lier days by means of oral tradition. The meaning of the texts had grown hazy, but the word was holy, and even though it remained but an empty shell to human understanding, it was

pleasing to the gods and had served
generations to bring gods and

its

purpose through the

men

into accord,

served; likewise the language of ritual and

and must be precomment thereon,

which, as the possession of a limited

class, required not only to be protected from overwhelming beneath the floods of the vernacular but demanded to be extended to the use of wider circles in

the dominant castes.

Sanskrit had already become

a

moribund

or semi-artificial Language, before

grammar
language
its

laid hold

continue and extend

it.

But from the outstart the
master or
guide.

upon it to Hindoo gramit.

marian

sat

humbly

at the feet of
its

to Learn of

and

never assumed to be

Inasmuch
a

as the

Language had existed and been perpetuated primarily as
of the living voice and not of ink
to reach the ears rather than the eyes of the divine,
in
;i

thing

and paper, and had been used
i1

followed

that

measure remotely true of no other grammatical endeavor the Hindoo grammar was compelled to devote itself to the
niceties of phonetic discrimination represented in the alpha-

most exactingly accurate report upon the sounds of the Language.

The

98

University of California Publications.
the refinements of observation involved

[Class.

Pro,

be1

itself,

in

the reports

nn accenl and the phenomenon of pluU; the formulation of the
principles of sentence phonetics
in

the rules of sandhi; the obser-

vations

i

ii

the physiology of speech scattered through the Prdtiall

cdkhyas are

brilliant

illustrations

of

the

Hindoo's direcl

approach
national

to

the real substance of living speech.

None

of the

systems of grammar, the Chinese, the Egyptian, the

Assyrian, the Greek, or the Arabic had anything to show remotely

comparable

to this;
all

and up

to the

beginning of the nineteenth

century, despite

the long endeavors

expended on Greek and

Hebrew and
si.

Latin, nothing remotely like

the Western world.

it had been known to The Greek grammarians had really never

.fined the barriers of written

Language; they were mostly con-

cerned with establishing and teaching literary forms of the language,
liven

when they

dealt with the dialects, they had the

standardized literary types thereof before their eyes rather than

When the grammars of and of Wilkins (1808) opened the knowledge of Sanskrit to European scholars, it involved nothing short of a grammatical revelation, and prepared
the spoken forms ringing in their ears.

Colebrooke

(1805), of Carey

(1806),

the

way

for an ultimate remodeling of language-study nothing

short of a revolution.

Though

these

Hindoo

lessons in accurate

phonetics as the basis of sure knowledge and safe procedure had
their immediate

and unmistakable influence upon the
its

scientific

work of the
service in

first

half-century, their 1 full acceptance tarried until

the second half was well on

way.

Even Jakob Grimm, whose
phonology musl be

promoting the

historical study of
still

rated with the highest, was

so blind to the necessity of pho-

netics as to express the view that historical

grammar could
in his Dit

be

excused from

much

attention to the "bunte wirrwar mundart-

licher lautverhaltnisse. "

and though von Raumer
not

Aspiforth

ration
in
all

und

dit

Lautverschiebung (1837) had
a

only

set

clearness the theoretical necessity of
in

phonetic basis, hut

given practical illustration thereof

the material with which he

was dealing,
Geschichtt
rarely
l

it

still

was possible
found who

as late as 1868 for Scherer in his

der deutschen Spracht
is

justly to deplore thai

"only

is

a philologist

willing to enter upon phonetic
BE

Cf. H. Oertel, Lectures on the Study of Language, pp. 30

(1901).

Vol. L]

Wheeler.-

Tin

Modem Science

of Language.
1

99

discussion."

and of Merke]
kind, to
est,
briii"it

The phonetic treatises of Briicke 2 185f5 and 1866) failed, though
I

(1S4!)

and

lS(i(i)

excellenl of their

the subject within the range of philological interin

and

remained for Eduard Sievers

bis Grundziigi

der

Lautphysiologii

(1876) and Grundzugi der Phonetik (1881) by

stating phonetics more in terms of phonology to bridge the gap

and establish phonetics as a constituent and fundamental porThe radical chaime of diai-aHer tion of the science of language.
assumed by the science
as
in

the last quarter of the century
this

is

due

much
But

to the

consummation of

union as to any one influ-

ence.

grammarians had developed in their processes of identifying the roots of words a scientific phonology In some of its applithat was all but an historical phonology.
it

was
to

not phonetics alone that the Indian

were able

leach to the West; they

cations

it

was

that

already, for in explaining the relations to
a

each oilier of various forms of
ferent words, even

given root as employed in difto serve

though the explanation was intended

the purposes of

word analysis and not of sound-theory, the gramThe recognition of guna and vrddhi.
brillianl

marians virtually formulated in repeated instances whal we now

know

as

"phonetic laws."

which antedates Panini, must rank as one of the most

inductive discoveries in the history of linguistic science.

The

theory involved became the basis of the treatmenl of the Indo-

European vocalism.
of Schleicher in his in the

The
and

first

thorough-going formulation, that
(1861), was conceived entirely
to the

Compendium
it

Hindoo

sense,

was

opportunity which this

formulation offered of overseeing the material and the problems
involved that we owe the brilliant series of investigations by
<

reorg

Curtius (Spaltung des a-Lautes, 1864),
1875), Osthoff

(N-Declination,

1876),

Amelung3 (1871, 1873. Brugmann (Nasalis sonans,

'E. Briicke, Untersuehungen uber die Lautbildung und daa uaturliche System der Sprachlaute (1849); Grundziige der Physiologie and Systematic

der Sprachlaute (1856).
"C. L. Alerkel. Anatomie and Physiologie des mensehlichen Stimm-und Sprachorgans (1856); Physiologie der mensehlichen Sprache (1806). A. Amelung: Die Bildung der Tempusstamme durcb Voealsteigerung im Dentschen, Berlin, 1871. Erwiderung. KZ. XXI t. 363 IV. completed July, 1873, published 1874, after the author's death. Der Orsprung der deutschen a-Vocale, Baupt's Zeitschr. XVIII, 161 ff. (1875).

inn

University of California Publications.

[Class. Phu,.

L876; OeschichU der stammabstufenden Declination, 1876), Collitz

(JJeber
.

<lit

Annahnu mehrerer grundsprachlichen a-Laute,
and unerringly
to to the definite

L878

Joh. Sclmiidt (Zweiarischi a-Laute, 1879), which led up

step by step steadily

proof that the

[ndo European vocalism was

be understood in terms of the

Greek rather than the Sanskrit.

These

articles,

written

in

the

period of intensesl creative activity the science has known, represent
tins.
tific

in

the cases of four of the scholars mentioned,
Collitz, the

viz.,

Cur-

Amelung, Brugmann,
life

masterpieces of the sciena

of each.

Though dealing with
to

single problem, they

combined both through the

results they achieved

and the method
to

and outlook they embodied
article,

give character and direction

the science of the next quarter-century.

Karl Verner's famous (KZ.
E.

Eim
97
ft',

Ausnahnu

der

ersten

Lautverschiebung,

XX III.
among

July, 1875), which proved of great importance
T.

other things in establishing a connection between

ablaut and accent, belongs to this period; and

Brugmann's article, Nasalis sonans, which served more than any other work to clear the way for the now prevailing view of ablaut, was influenced by Verner's article, which was by a few months its predecessor.

Both

articles,

it is

worthy of noting, were distinctly
it

influ-

by Brugmann's, through a suggestion of Osthoff's, by Sievers, whose Lautphysiologu had just appeared within the same The full effect upon Western science of the introduction year.
chiefly

enced by the new phonetic: Verner's,
Briicke,

would appear,

of the Indian attitude
to

toward language study appears therefore
last

have been realized only with the
.More

quarter of the century.

prompt than the response of European science to the teachings of Hindoo phonetics and phonology had been the acceptance of the Hindoo procedure in word analysis, especially
with relation to suffixes and inflexional endings.
of study of Greet

and Latin had yielded no clue
as

to

The centuries any classifi-

cation or assorting of this material according to
tion.

meaning or funcdomini custos
des Etythat

The medieval explanation of dominicus
as good as any.
I

was

Besnier
it
t

in

his essay,
a

l.<i

sdena

mologies

1694

i.

counted

he

mark of

sound etymologisl

he restrict his attention to the roots of words, for to bother with the other parts would be "'useless

and ludicrous.*'

And when

vol.

i|

Wheeler.-

Tin

Modern Science of Language.
II.

101

Home
just

Tooke

in

the Diversions of Purley,

429 (1786-1805),

before the sunrise, wrote the startling words: "All those

common

terminations

in

any language
(i.e., ly, a us,
;

.

.

.

arc themselves

separate words with distinct meanings," and
tives with such
in truth, all

I

II.

454)
.

:

" Adjecare.

term inations

fnl,

sonn

ish, etc.)
a

compound words*' and when
I

he flung oul like

chal-

lenge the analysis of Latin iho.

shall go,' as three letters con-

taining three words,

viz.

i,

'go,'

(fiovXo/xai)' will,' o( ((jo)

'

1.*

no

one seems to have been near enough to the need of such instruction to

know whether or
fruit,

Q01

he

was

to be

taken seriously

;

for the

woids bore no
trine

and only years afterward, when Bopp's docas antiquarian
later, in the full light of

had been recognized, were they disinterred
Eleven years
his

curiosities.

the Sanskril

grammar, Bopp published
had been Found.
that
lie

Conjugationssystem, and the clue

To be

sure.

Bopp was misguided
of
a
it

in his belief
a

could identify each element

word-ending with

significant word,

and assign

to

a

distinct meaning, hut he

found the key

to

an analysis having definite historical

had value and

permitting the identification of such entities as mode-sign, tensesign,
trine,

personal-endings,

etc.

The erroneous portion

of his doc-

based upon his conception of the Indo-European as an

agglutinative type of speech, dragged itself as an encumbrance

through the
still

first

half-century of

t

he science, and. though gasping,

lived in the second edition of Curtius'

Verbum

I

1877
its

I.

This.

along with

many

other mechanical monstrosities of

kind,

was

gradually banished from the linguistic arena by the saner views
of the life-habits of Language which had their rise from linguistic

psychology as
as well as
to

a

study

id'

the relations of language to the hearing

speaking individual and the relations of the individual

the speech

community, and which asserted themselves with
the seventies. the beginning devoted
in

full

power

in

Bopp had from

himself id language-

study, not as an end

itself, but,
1

as

and sponsor YYindiscluiiann.

ns well

we know from his teacher as infer from the direction
"in
this

and

spirit of his work, he

hoped

to be able

way

to pene-

trate into the mysteries of the
'Introduction (1816).
to

human mind and

learn something
j».

Bopp's Conjugationssystem der Sanskritsprache,

Lv,

L02

University of California Publications.

[Class.Phil.

of

its

nature and
<

its

laws."

He was

therefore unmistakably of

the schoo] of the

Ireeks, qo1 of the

Hindoos; for the Greek gram'why,'

marian

in

facing language asks the question

grammar

being to him philosophy, whereas the Hindoo asks the question 'what,' grammar being to him a science after the manner of what

we
f or

call

the 'natural sciences.'

There

is

indeed bu1 slighl reason

the

common

practice of dating the beginning of the

modern
merely

science of language with Bopp, aside from the one simple resull
of his activity, which must
in

strict

logic be treated as

incidental thereto, namely, thai he gave a practical illustration of
the possibility of applying the comparative

the scope and enriching the results of historical

method for widening grammar.

As Bopp had tried to use the comparative method in determining the true and original meanings of the formative elements, 1 (1802so did his later contemporary, August Friedrich Pott
to use it in finding out the original meaning The search for the etymology or real meaning of words had been a favorite and mostly bootless exercise of all European grammarians from the Greek philosophers down, having its original animus and more or less confessedly its continuing power in

L887

undertake

of words.

the broadly

conviction, that words

human, though barely on occasion half-formulated and their values belong by some mysteriIn the instinct to begin his task

ous

tie

naturally to each other.
still

Pott

was

with the traditions of the Greeks and the Grecoin

Europeans, but

developing

il

he was guided into

by two forces thai had arisen since the century opened.
the guidance of the comparative
laries of

new paths Under

method, whereby the vocabu-

demonstrably cognate languages now assumed a deterto each other, he

minate relation

came unavoidably

to the r

ig-

nition of certain
dift'c rent

normal correspondences of sounds between the

tongues.

On

the other hand, in almost entire indepenin

dence

hereof,

Jakob Grimm

the

pursuit

of

his

historical

method had formulated the regularities of the mutation of consonants in the Teutonic dialects and had set them forth in a
second edition of the
L822.
K.
firsl

volume of
a

his

grammar, appearing

in

In all this
f.

was contained

strong encouragemenl as well
i'

Pott:
.

Btymologisehe Porschun^cn.

vols.

I.cm.in..

ls:;:;-:;r,

:

l'iuI

1859 76.

Vol.

L]

Wheeler—Thi Modern Sciena
to

of Language.

103

as

warning

apply these new definite

tests to
's

every ety

Logical

and therewith arose under H was a of a seieiitifie etymology.
postulate,

Potl
first

hands the beginnings

promise of deliverance

from

;i

long wilderness
positivistic

oi*

caprice.

The
finally

attitude which had been

gradually infused

into Language-study

under the influence of the Hindoo grammar
in

reached

its

extremest expression

the works of August

Schleicher (1821-1868).

The

science

of

Language

he

treated
isolated
to

under the guise of

a

natural science.

Language became

from the speaking individual or the speaking community

an

extent unparalleled in any of his predecessors or successors, and

was viewed

as an

organism having
itself.

a

Life

of

its

own and laws

of

growth or decline within
ventured
reconstruct

Following the analogies of the

natural sciences and trusting to the inferred laws of growth, he
to

from the scattered data of the cognate
visible

[ndo-European Languages the

form of the mother speech.
a

His confidence

in the

character of Language as
i/.er

natural growth

made

hiin the firsl

ureal systemat

and organizer of the mate-

rials of

vergleichenden Gramm'atik,

Indo-European comparative grammar (Compendium der as confidence in the unerring 1 S
»'>1
;

uniformity of the action of the laws of sound made Karl Brug-

mann
It

the second

Grundriss

der

vergleichenden

Grammdtik,
the

1886-1892).
is

not by accident

that

the

first

one

to voice outrighl

dogma
tion
in

of the absoluteness
a

(Ausnahmslosigkeit) of the Laws of
Leskien (Dit
Declina-

sound was

pupil of Schleicher, A.ugus1

Slavisch-litauischen
this

The use of

dogma

as a

und Germanischen xxviii. 1876 norm and test in the hands of a sig-

nally active and gifted body of scholars

who followed

the Leader-

ship of Leskien and were Srlnih

known under

the title of the Leipziger
to
it

or the Junggrammatiker,

and the adherence

in

practice of

many

others

who did

not accept the theory involved.


in

a use

discovery

which was undoubtedly greatly stimulated by Verner's (1875) thai a greal body of supposed exceptions to
in reality

Grimm's law were

obedient to law. gave to the science

the two following decades, along with

abundance of
a

results.

an objectivity of attitude and procedure and
ture that

firmness of struc-

may

fairly

ho said to represent

the

consummation of

104

University of California Publications.

[Class. Phil.

tli.it

positivisl

tendency which we have songhl
of

to identify

with

the

Influence
its

Bindoo grammar.

This

movement, however,

derived

impulse by no means exclusively through Schleicher.
its

A new stream had meanwhile blended
to the

waters with the current.
relations of language

The psychology of Language asa study of the
speaking individual, that
is
is,

of the conditions under which
rela-

language

received, retained,

and reproduced, and of the

tions of the individual to his speech
into play preeminently

community, had been brought

through the labors of Heymann Steinthal,

who. though as
felt

a

psychologist a follower of Herhart, must be

to

represent in general as a linguist the attitude toward
first

language study

established

by Wilhelm

v.

Humboldt.

Wil-

liam D. Whitney shows in his writings on general linguistics the
influence of Steinthal, as well as good schooling in the

grammar

of the Hindoos

and much good common sense. His lectures on Languagi and tk( Study of Language (1867) and the Life and Growth of Language (1875) helped chase many a goblin from
1

the sky.

Scherer's Gcschichte der deutschen
its

combined more than any book of
lines of endeavor,

Sprache (1868), day the influences of new

and

especially gave hearing to the

in the psychology as well as the physiology of speech.

new work To this

period

(1865-1880), under the influence of the combination of

the psychological with the physiological point of view, belongs
the establishment of scientific

common
it

sense in the treatment of

language.

By

virtue of this, as

were, binocular vision, lan-

guage was thrown up into relief, isolated, and objectivised as it had never been before. Old half-mystical notions, such as the
belief in
;i

period of upbuilding in language and a period of

decay,

all

savoring of Hegel, and the consequent fallacy that
a

ancient languages display

keener speech consciousness than the

modern,

—speedily

ferred itself

The center of interest transfrom ancient and written types of speech to the
faded away.

modern and

living.

Men came

to see that vivisection rather

than

II. Steinthal: Der Ursprung der Sprache, im Zusammenhang mit den Fragen alles Wissens, 1851; Characteristik der hauptsachlichsten Typen des SpracKbaues, 1860; Einleitung in die Psychologic und Sprachiri.s.sin.sclidft, Issl; i,. ch. der Sprochcit, bti den (iricclun und h'dnurn, s ls!»i)-«t1. Also editor with Lazarus of the Zeitschrifi fur Vblker/isi/t'lmlnfiii mid Simicliirissciiscluift, from 1859.

letsten

i

«'i"..

vol. i|

Wheeler.

— The
a

Modern Sciena

of Language.

1<»">

morbid anatomy must supply the method and
research.

spiril of linguistic

under which cognate languages may be supposed to have differentiated out of a mother speech, and conceived in terms of the observed
idea affecting the conditions

The germs of

new

relations

of dialects to

languages, were infused by .Johannes
rfrr

Schmidt's Verwancllschaftsrcrhiilhtissc
<>!•

indogcrnum. Sprach-

1872).

The

rigid

formulas of Schleicher's

Stammbaum

melted

away before Schmidt's Wellenthqorie and its line of successors down to the destructive theories of Kretschmer's Einleitung in
<lii

Geschichli

der gricch. Sprarlic (lSOti).

Herein as

in

many

another movement of the period we trace the results of applying
the lessons of living languages to the understanding of the old.

A remarkable document
ing
in

thoroughly indicative of what was movrxiw'nnigt n, Vol.

the spirit of the times was the Introduction to Osthoff and
T

Brugmann's Morphologisclu Unit
the gospel of the period, and
its

(1878).

But
der

theology for that matter, was

most effectively set forth in
Sprachgeschichte
influence
(1st edit.,

Hermann
than

Paul's Principien

1880), a work that has had more

upon the

science

any

since

Jakob

Grimm's

Deutsche Grammatik.

Paul was the

real successor of Steinthal.

He
ical

also represented the strictest sect of the positivists in histor-

grammar.
his

tendencies,

As a consequence of the union in Paul of the two work acquires its high significance. Be estabfrom Schleicher's treatment of language sciit to be beyond peradven
set

lished the reaction

ence as

a

natural science: he showed

ture one of the social sciences, and

forth the

Life

conditions of

language as a socio-historical product.

The work of the period dominated by Paul and the ueo-grammarians, as well as the theories of method proclaimed, show, however, thai the
scientific

two factors just referred

to

had not reached

in

the

thought and practice of the day a perfect blending.
title

A

well-known hook of Osthoff's bears the

Das physiologischt

und psychologischi Moment in der sprachlichen Formenbildung (1879). The title is symptomatic of the times. The physiological and the psychological were treated as two rival interests
vying for the control of language.
phonetic laws,
to be in case
if
it

were not

a

Whal did not conform to the phenomenon of mixture, was
This dualism could

explained

possible as

due

to analogy.

L06

University of California Publications.

[Class.Phbl.

be expected to be bul

a

temporary device
in
it

like the Betting

up of
close

Satan over againsl God,
sin.

order to accounl for the existence of
has proved
itself

A temporary device
first

to be.

The
is

of the

century of the modern science of language

tending

toward

a

unitary conception of the various forms of historical
language.

change

in

The process by which the language

of the

individual adjusts itself to the

community speech
to the

differs in kind

qo

wliii

from thai by which dialed yields

standard

lan-

guage of the larger
language differ

community.

The

process

by which the

products of form-association or analogy establish themselves in
in

do whit in kind from that by which new pro-

nunciations of words,
acceptance.

i.e., new sounds make their way to general The process by which loan-elements from an alien

tongue adjust themselves
processes mentioned.

to use in a given

language differs psy-

chologically and fundamentally no whit from either of the four

In fad they

all.

all

five,

are

phenomena
a

of 'mixture in language.' 3

The

process, furthermore, by which

sound-change
guage

in

one word tends to spread from word to word
lan-

and displace the old throutjhout the entire vocabulary of the
is

also

;i

process of 'mixture,' 2

and depends for

its

mo-

mentum
same
is

in last

analysis upon a proportionate analogy after the
a suffix

essential

model as that by which an added sound or

carried by analogy from

word

to word.

All the

movements

of historical change in language respond to the social motive:

they

all

represent in some form the absorption of the individual

into the

community mass.

It

has therewith become evident that
in

there
ically

is

nothing physiological

language that

is

not

psycholog-

conditioned and controlled.

So then

it

appears that the

'Sec <). Bremer. Dents-he Phonetik, Vorworl X ft. (1893); B. I. Wheeler, Causes of Uniformity in Phonetic Change; Transae. Amer. Philol. ^.ssoc,

xxiii. i iv. mm A point of view involving
)'.
(

speech-mixture

the recognition of a mor< recondite form of sn-eesteil liv C. I. Ascoli ( Sprachwissenschaftwhereby the initiation of pho and ultimately the differentia tion of dialects and even of languages may assume relation to languages of the substratum, as they may be termed, i.e., prior and disuse, languages of or tribes who have through the fate of conquesl or assimilation beeri Notably has this point of view absorbed into another speech community. Keen urged by II. Ilirt (Indog. Forsehungen, V. 36 11'.. L894), and by Wechssler (Giebt e& Lautgesetse, pp. 99 ff.) With this poinl of view the science of language will have largely to deal, we are persuaded, in the sec bnd century of its existence.
is

that

first

liche Briefe, pp. 17 IV., 1881-86; trsl. L887), aetic and syntactical changes in language,

l

I

Vol.

i|

Wheeler.-

Tin

Modern Sciena

of Language.

1<>7

modern science of Language has fairly shaken itself free again from the natural sciences and from such influences of their method and analogies
his period

as

were intruded upon
a

i1

by Schleicher and

(1860-80), and after
definitely oriented

century of groping and experiitself as a social science

ment has

and found

dealing with an institution which represents more intimately and
exactly than any other the total life of

man

in

the historically

determined society of men.
of the nineteenth century establishes
tant
frontier.

Within the history of the science of language the beginning beyond doubt a most impor-

To appreciate how sharp

is

the contrast between

hither and yonder
the

we have only to turn to any pari or phase of work yonder,—the derivation of Latin from Greek, or mayHebrew
to

hap, to be most utterly scientific, from the Aeolic dialect of Greek,
the sage libration of the claims of Dutch as againsl
be the original language of

mankind, the bondage

to the

forms of
point of

Greek and Latin grammar as well as
view of the philosophical
nation of

to the traditional

grammar
the

of the Greeks, the subprdina-

grammar

to logic, the hopeless
in

etymologies and form
of

analyses

culminating

phantasies

Hemsterhuis

and

Valckenaeer, the hick of any guiding clue for the explanation of
how-

sound or form came

to be

what

it

is.

and the curse of arid

sterility that rested
all

upon every

effort. All the

ways were blind and
is

the

toil
a

was

vain.

On

the hither side, however, there
in the mass.
it

every.'

where

new leaven working
if

What was
not

that leaven

To identify
review.
I

possible what

was has been the purpose of
it

this

think we have seen

was

the influence of the
influenei
itself the

natural sciences, certainly not directly: wherever that

found direct application

it

led astray.

It

was
that

not

in

discovery of the comparative method,
auxiliary to
a

for

proved bu1 an

greater.

If a

founder must be proclaimed for the

model n science of language, thai
(

founder was clearly Jakob
<>ne

rrimm, not Franz Bopp.

The leaven
was found
in

in

question was comprised of two elements.

the establishment of historical

grammar,
in

for this

furnished the long-needed clue; the other was found

the disatti-

covery of Hindoo grammar, for this disclosed the fruitful

tude for linguistic observation.

Historical

grammar furnished

L08

University of California Publications.

[Class. Phil.

the missing clue, because

it

represented the form of Language as

created, what
Init

it

is.

qo1

by the thoughl struggling for expression,
it.

by historical conditions antecedenl to

Hindoo grammar
its

furnished the method of observation because by
instind
it

fundamental

asked the question how

in

a

given Language does one
a

say

a

given thing, rather than
it

why

docs

given form

embody

the

thoughl

does.

The germinal
of

forces which have

made

this

century of the
in

sci-

ence of Language are do1 without their parallels

the century

American national

Life

we are me1

to celebrate today.

Jakob

Grimm was

of the school of the Romanticists and he gained his

conception of historical

grammar from

his

ardor to derive the

institutions of his people direct from their sourees in the national
Life.

The acquaintance of European scholars with the grammar

of

arose from a counter-spirit in the world of the day whereby an expansion of intercourse and rule was bringing to the
in

India

wine-press fruits plucked

many

various

fields

of national

life.

Thus did the
After

spirit of national
a science,

particularism reconcile

itself,

in

the experience of
like sort

with the fruits of national expansion.
its

has the American nation in

development

for

the century following

upon the

typical event of 1803

combined
of

the widening of peaceful interchange and

common standards

order with strong insistence upon the right of separate communities
in

things pertaining separately to them to determine their

lives out of the sources thereof.

Therein has the nation given
its

fulfillment to the prophetic hope of
ist.

great democratic imperial-

Thomas

Jefferson,

3

"I am persuaded no constitution was ever

before so well calculated as ours for extensive empire and self-

government.

"

The
the
first.

linguistic science of the second century will build

plateau

leveled

by the varied

toils

upon and experiences of the

speech

More than ever those who are to read the lessons of human will gain their power through intimate sympathetic

acquaintance with the historically conceived material of the individual Language. Bu1 though the wide rangings of the comparative

method have

for the time abated

somewhat of

their interest

Letter to Mr. Madison. L809.

Vol. l]

Wheeler.-

The Modern Scienct of Language.
will

109

and

their yield,

it

remain thai he who would hnvc
an offprinl

l.-iruvst

vision must gain perspective by frequent resort to the extra-mura]

lookouts.

Language

is

of

human

life,

and

to

the

studenl of

human speech nothing

Linguistic can be ever foreign.

UNIVERSITY OF CALIFORNIA PUBLICATIONS

CLASSICAL PHILOLOGY
Vol.
1,

No.

4, pp.

1

1

1-129

October 27, 1905

ON THE INFLUENCE OF LUCRETIUS ON HORACE
BY

William

A.

Merrill.

The purpose of
a

Ibis inquiry

is

the examination of

Horace for
it

evidence of lmcretian influence.

In a general

way

has been

commonplace of literary criticism that the one was indebted to the other, and the scholiasts and editors have cited many The editors of Lucretius have also pointed parallel passages. out in Horace similiarity in thought and expression, and the subject has been treated in special monographs by Goebel,
Reisacker and Weingartner.
Reisacker's
it little

program
to

(Breslau,

1873)

I

have seen and have found in

my

purpose.
/'.

The

other two (Goebel: Horaz mid Lukrez, Zeitschr.

d.

oesterr.

Gymn.

8

(1857), 421-427; Weingartner:

1

De Horatio
in

Lucretii

imitatore, Halle,

1874)

I

have not been able to procure, bul
fancy there
is little

from criticisms of them
special inquiry.
Sat.
I.

I

them

for this

and
1

Beginning with the Satires, Horace's earliest work, examining them in their present order without
1

find in regard to the exact dates of their composition, 1113 cetera de "enere hoc, a Lucretian phrase occurring
in 3,

4S1 and elsewhere.
<>f

Then
is

in

22 praeterea occurs as
in
|

a

word
2.".

transition that

Erequenl

Lucretius, and

in

at

puerisolim dant crustula blandi
prima,
a

doctores, elementa
'

vclint nt diseere
this

reminiscence of Lucr.

1.

r-^>

sed

paper was written Weingartner 's dissertation was found in 'After The canons adopted by him for determining influDiss. Phil. Hal. II. 1 sq. ence appear to me to be too lax.

112

University of California Publications.
veluti pun-is absinthia taetra
tiii-

[Class.Phil.

medentes

eum dare
mellis

conandulci

prius

oras

pocula
Plato,

circum

contingunl
e,

flavoque liquore.

Laws 659
in

says thai the sick are

given

wholesome food
3

pleasanl

meal

and

drink-.

tm1

Quintilian

quotes and comments on

Lucretius; Jerome 2

mentions the honey, and Ausonius 3 the

Seneca4

the

Elder

mentions the

wormwood also: wormwood only, and
to unpleasant

Pliny 5 the
food

Younger reduces the allusion
with caressing tones.
into
I

urged on

Later,

Sir

Philip

Sidney 6 turns the wormwood
tinues the tradition.
Literary influence

rhubarb and Tasso 7 conis

Here

think

a

genuine ease of
so far as the evi1,

from Lucretius down;
vertit arato

dence shows.

— 28

and Lucr.
compared

211, vertentes
|

vomere have oo connection.- 50 quid referat intra naturae
fines

viventi

may

be

with
t>)<?

Epicurus'
tt\ov-

Kvpiai Ao^ai 15 (Diog. Laert.
to<? zeal
et<?

X

144) 6

(f>vaeco<;

wptarat

teal

eviropiaros iaTiv 6 Se

twv

/cevebv

So^cov

aireipov eKiriTrrei.

Horace was not dependent entirely on Lucretius
Ins
facil

for
id

knowledge

of

Epicureanism- In
est, cf.

f>4

quatenus
2,

— Lucr.
thai

3,

424 quatenus

218 and

927: the
use of

fact

Horace and Ovid follow L.

in the causal

quatenus shows merely their agreement
of the language.-In 68 Tantalus
a

in a

development

labris sitiens fugientia

captat

!

flumina-

— L.

3,

981 nee miser impendens mag-

aum
used.

timet acre saxum, differenl

forms of the myths are

70 saccis
is a

indormis inhians
in


fit

L. 1.

36 inhians
a

in te,

dea. visus
se

mere agreemenl

the use of

word.

-

98 ne

penuria victus


3,

L.

~>.

1007 penuria deinde cibi beut raro qui se vixisse

longs to every day language- 117

beatum

dicat

et

exacto contentus tempore vita

cedal ut

conviva satur
1

— L.

938 cur oon

ut

plenus vitae con viva

3, 1, 4.

In.

Etuf
17.
<;,

[,

§

463.

Ep.
'

Suas.
I.

]6.

8,

L2.
p.

1

1

defense of Poe1 ry,
I, iii.

7

Ger. Lib.

Vol.

1.]

Merrill— On
recedis,

Influenct of Lucretius

on Horace.

113

and 959 ante
rerum.

quam

satur ac plenus possis disis

cedere

The conception
to

traced
to Job,

back to

Bion
paro-

through Teles

Ps-Aristotle,
It

and

and

is

died by Babrius.
in

occurs

in

Cicero and Plutarch, and
It

La

Fontaine and Chenier.

is

formulated by Epiit

curus himself.
sources, perhaps

Probably Horace got

from Epicurean
121

from
3,

L.

And

finally,

verbum aon

amplius
is a

addam

— L.

941 cur amplius addere quaeris,

mere coincidence.
In the second sal
ire.

2
gal

verse 8 praeclaram ingrato strin•"..

mains ingluvie rem

— L.

1003 deinde animi injrratani
in

natnram pascere semper merely agree
sententia dia Catonis
tentia ponit

sentiment- 32

— L. —

3,

371 Democriti

— sancta sen-

from Lucilius, Tacitus and Homer. It is a paraphrase that does not belong to any one in particular .- 57 (amator) qui patrium mimae donal L. 4, 1129 et bene parta patrum fundumque Laremque

may

be paralleled

fiunt

anademata, mitrae agree
!

in describing the extrava-

L. 3, 939 ante gance of the lover- 104 ante quam may be paralleled from Aetna and Manilius and

quam
urs

but once in each of them.
Lucretius'
metrical

This

may

be a case where

technique had

some influence, for

there are undoubted imitations of L. in the Aetna and
in

que

Manilius- 119 namque parabilem amo venerem facilemL. 4, 1071 volvivagaque vagus Venere ante recentia

cures

is

an agreement

in a

prescription- 133 denique as

the third
rence,

member of a but who would
14 toga,
5,

series is a

common Lueretian
Erigus
sit

occur.'

say that

it is

solely Lueretian

3

3,

quae defendere

quamvis crassa
quae defendere
that
[

queat

— L.

1429

dum

plebeia tamen

possit:

here rhythm leads

me

to believe
4,

there

is is

imitation.
a

26 cernis acutum.

— L.
wry

802 acute
is

cernere

chance agreement.

Prom 38

to

52

the well

known

passage where

Horace describes the blindness of lovers
turning the
defects of their Loved similar of

and parents
lovers in

in

ones into virtues.
4.

Lucretius has something

1155-1169.

Plato mentions the principle in
it.

Rep. 474: Theocritus, Ovid, .Martial. Moliere allude to

1U
Am\

University of California Publications.

[Class Phil.

one who has witnessed
to its

the

phenomenon can bear
phraseology
in

testimony
in

occurrence, and we shall no1 be justified
unless
the
leads
in

inferring

imitation

thai direction.
15

Delectanl 40 agrees with

deliciis

L156.

male parvusagreement.

1162 parvula; there seems no other
56 sincerum
vas6,

verbal
efficere

17

vitium

vas
of

may
I

possibly be influenced by L. on accounl

Epist.

2 54.

66

communi sensu
in

— L.

1,

422 communis
in

sensus

is

an agreemenl

sound bu1
in

nol

sense.

The
comis

sketch of

human
L.-98

progress
utilitas,

98-112 has

much
et

in

mon

with

iusti

prope mater
occur
8.

aequi
L.

Epicurean and does not expressly
Xpeia,
serunt

in

whose
prorep-

account agrees with thai of Diodorus I
usns.

Diodorus says

became

man's
terris

teacher.-99

primis animalia

— L. —

cum

5,
|

821 quare etiam

atque etiam maternum nomen adepta terra tenet merito.

quoniam

genus

prope certo

humanum atque tempore fudit-100 mutum et turpe
ipsa
ereavit

animal
pecus,

glandem atque cubilia propter 5, 939 glandiferas inter curabant corpore quereus-101 unguibus et pugnis, dein
fustibus, atque ita porro
|

pugnabant armis


|

5,

1283 anna

antiqua
item

manus ungues dentesque fuerunt et lapides et silvarum fragmina rami- 103 donee verba quibus
uominaque invenere
e1

voces sensusque notarent,
si

5,

1057

genus humanum, cui vox
ponere leges

lingua

vigeret,

pro vario

sensn varia res voce notaret- 105 oppida coeperunt munire
et

5,

1108 condere coeperunt urbis arcemque
illi

locare.-108 ignotis perierunt mortibus

5,

326 cur
incertam
iungebal

supera

bellum

Thebanum

etc.- 109
5,

venerem
in silvis

rapientes more Eerarum

962 Venus

corpora amantum.
conciliatrix
viri

11" viribus editior caedebal

5,

963

enim

vel

mutua quamque cupido
libido- 111
5,

vel violenta

vis

atque

bnpensa

iura

inventa

metu

iniusti

Pateare necessesl

1144

iuraque

constituere,

1151

nut us maeiilat
is

poenarum praemia
formula

vitae.

Pateare
evi-

necessesl

a

Lucretian

and clinches the

dence thai Borace was uol only familiar with Epicurean
doctrine hul

had read

Lucretius' description- 112 tern-

Vol.i.]

Merrill.— On Influena
pora

<>/'

Lucretius on Horace.

115

si

Eastosque velis evolvere nmiMli

5,

li'Tti

tempora

rerum.
4
4.

76 Locus

~ conclusus

4,

458 eonclusoque
1282

Loco

is
I*
1

a
""

inert'

coincidence Like avel 87, and also the syntax of
4,

insuevit-hoc me with Lucr.
degere and thai
1005 quaeque
5
--

insuescal

<te>
Lucr.
1.

of

106 vitiorum quaeque and

seminiorum.
73 vaga - flamma

In the

fifth, Line

1..

6,

152 flamma
the close of

vagetur

is

a

mere chance agreement, bu1

a1

the satire, K>1


5,

namque dens

didici
se

securum agere aevum
necessesl

L. 2, G4ii

omnis enim per

divom aatura
etc.,

immortali aevo

summa cum

pace fruatur

and

!-.

82

nam
a

bene qui didicere deos securum agere aevom,

is

almost

quotation from L.
6,
^)H.

The Lucretian passage occurs
si

again in

and Horace 102 nee
j

quid miri

facial

natura, deos id

tristes ex alto caeli demittere tecto cor6,

respond in thought to L.
caeloque tuentur
to the

50 cetera quae

fieri

in

terris

mortales: they attribute their ignorance

gods who. of course, can not be angry, but will
a

bring about
here
is

disturbed

mental state

in

man.
of

Horace

jesting

and

is

speaking

Lightly

Epicurean

principles.
6
In

the sixth satire, line 3 olim qui magnis Legionibus

imperitarint
tarunt, L.
I

L.

3,

1028 magnis qui gentibus imperi-

is

undoubtedly following Ennius.
I,,

Horace

is,

think, following

here.

18 Longe Longeque remotos
is

3.

69

longe

Longeque
It'

remosse

noteworthy.

I;i

the

8

eighth, Line

commune sepulcrum corresponds
The thoughl
In 46 displosa
--

to L. 5,

259

commune sepulcrum.
is

variously
sonat

ex-

pressed
vesica 9 turn

a

trivial
6.
li'ii

one.

quantum
soni-

— L.
a
is

vesieula

saepe

ita

da1

magnum

is

chance agreement.

membra movere
movere
10
a

— Tn mollius — L.
L. 3,

the ninth satire. 24 quis
4,

789

mollia

membra
atque
is

reference to dancing merely.

34

siimil

adoleveril aetas


a

449 adolevil viribus aetas: here
4!)

another national idiom. -In the tenth.

haerenti capiti
o capiti petere

cum multa laude coronamind

-L.

1.

929

>ronam

is

commonplace.

116

University of California Publications.
In the second
satire
is

[Class.Phu,.

Sat. II.

book of the satires, line 17 of the
I-.

first

has Scipiadam
in

3,

L034 Scipiadas; this usage
accessit

conventional

the

hexameter- 25

fervor

eapiti

numerusqne Lncernis
Lumina flammis
lupus,
is


is

L. 4,
a

450 bina lucernarum
physiological allusion.

florentia

merely

52 dente
--illis
i

cornu

laurus petit

5,

1034 conma
a word for
calls

rat

ns

petit.

Here

agreement
4,

in

"butt. "-77 inlidere dentem
for no remark.

1080 dentis inlidunl

2

In

the

second satire
leniet

17
2,

emu
17

sale

panis

stomachum bene
lat

|

latrantem

nil

aliud sibi

naturam

rare;

the expressive

metaphor was known

to

and
ahest

Ennius-28, the hiatus
shows metrical
license

num
and

adest

3,

1082
a

Homer dum
certain

testifies

to

agreement of Horace's
tic

satirical

hexameter with the didacredeuntibus annis;
sufficiently trite.-

and undeveloped Lucretian- 83 diem festum rediens

advexerit annus

1,

311 multis

solis
is

the metaphor of the returning year

88 tarda senectus
of age calls for

1,

414 tarda - senectus; this quality

little

originality- 104 cur improbe carae


in

3,

1026 fuit improbe rebus.
in
is

The convenient

dactylic

word
and

found in Virgil and Persius also. without significance. The syntactical agreement
the fifth foot
is

105 emetiris aeervo

2,

703 egigni corpore belongs

to

historical syntax.
3

In the third satire occur 49 palantes error

certo

de

tramite pellit
vitae,

2,

10 errare atque viam palantis qnaerere

and

6,

27 viam monstranl tramite parvo.

The way
is

of

life,

l'roin

which the ignoranl and the wicked stray,
.-irises

a conception thai

from primitive theologizing and
to

needs not

to be referred

any particular writer
to

Yri

the strange

had

word palantes lends me Lucretius in mind here. 95

think thai

Horace
decus,

virtus,

Eama,

divina

humanaque

pulchris

|

divitiis

parent

5,

1114
hono-

aurum-quod
rem.

facile et validis e1

pulchris dempsil

This melancholy truth of the supremacy of riches
to

comes home
tius.
I

every one as

it

did to Horace and Lucreis

do not know that

II.

altogether indebted to

Vol.i.]

Merrill.

— On

Influence of Lucretius on Horace.

Ill

I,,

for seeing whal

all

musl

have seen.
is.
I

Bui

the

refer

ence to the beauty of riches
Lucretius.

think, a reminiscence of

point also

to

The monosyllabic use of quoad m 91 may -141 I., who lias it in 5, 1213 and elsewhere.

splendida

bilis

— L.

6,

L187 spendidus

humor

is

a com-

mon
dit

medical allusion.- 153 ni cibus atque

ingens acce-

artus

L. stomaeho fultura ruenti reducere is another- 191
re could
still

4,

867 cibus ut suffulcial
1,

228 reducal merely

shows that
cur Aiax

be long

in this

compound- 193
is

- putrescit
etc.

3,

871

aut

putescat

due

to

common
100.

mortality.- 199 tu

cum pro

vitula statuis

dulcem
1,

Aulide natam

has no verbal connection with

84-

Horace could have learned the story of Iphigenia's sacrifice from other sources, yet from the way it is used
by him
I

think there
3,

is

a

Lucretian reminiscence.
a little to

Emevi-

probe in 200 (L.
dence.- 269
errore vagaris

1025) adds
sorte

cumulative
1052

rluitantia
is

laboret
in a

3,

Huitans

an agreement

common metaphor.-

283 surpite

2,

314 surpere

is

an ineleganl syncopation
582 memori mente
metrical substitute
remotos.

which survived from earlier Latin.
4

In
there
for

4.
is

90 memori

- pectore
in

— L.
ut

2,
a

an agreement 94
a

the use of

niemoria.-In
is

foutes

adire

atque

haurire queam
I'miiis

parody on
'Phis

L. 1,

928 integros accedere

atque haurire.

sentiment of L. had

many

admirers.
6

In

6,

1

modus

agri

non

ita

magnus

L

2,

L172 agri

multo modus is a chance agreement. 59 perditur has caused more discussion than L. 2, 831 disperditur. Both are reflections of homely usage.- 61 nunc somno e1 inert
i

bus

horis

|

ducere

sollicitae

iucunda

oblivda

vitac
is

3,

1066 in

somnum

gravis atque oblivia quaeril there

only

a metrical agreement in the use of oblivia.
vestiiria

101 ponit locution.

3.

4 pono - vestiuia

is

a

common
villain

7

In
3,

7,

28

Romae

rus optas; absentem rusticus urbem
praecipitanter
be paralleled
is

1063 currit agens mannos ad

--properans urbem petit atque revisit

may

from other moralizing.

The discontent

human.- 49

lis

University of California Vvtilications.
turgentis verbera caudae
is

[Class Phil.

4.

1034 turgida semine multo

a

physiological

agreement.

Tn 81 the metrical imperperfundit
trite.

itas
I.

again occurs.

90 foribusque repulsum
is

1177 exchisus amator: the thoughl
in
<

In

105

enim
from
8

the third place, as in L.

1,

680.

may

be paralleled

licero also.
8, a 3,

In
merely
dent -

51

inulas

--

amaras
bitter
:

2,

430

inulae
tibi

there

is

mention of
2

a

herb- 75

di--commoda
in

commoda
i

vitae

commoda was common
2.

the

popular philosophy
in General
*^-v

Reid on Cic. Acad.

231).
is

general conclusion from the Satires
al

a)

Horace

was an Epicurean
he

thai

stage of his development; (&)

was familiar with Epicurean principles some of which
;

be had gained from Lucretius1
tion

(c) there is direct imitais

of Lucretius in
also

his

work; (d) there
(e)

a

metrical

influence

from Lucretius;

there are so
in small

many
mai-

places where

Horace and Lucretius agree
Lucretian.

lers thai are also tive effect

round in other authors, that the cumulais

on the reader

Epodes.
L'

I

now

pass to the Epodes.

In

the second

epode line 7 superba
50;
3,

civium

potenis

tiorum Limina


1

L. 2,

1027 rerumque potentes

a

chance
decidere
Libel

agreement- 13
falcibu

falce
is

ramos amputans
agricultural
ilice,
[


in
is

5,

936

ramos

an

allusion- 23
tenaci

iacere

modo sub antiqua
2,

modo

gramine
usta

— L.
1,

29 prostrati

in

gramine molli

a pic-

nic agreement, so to say.- 41 pernsta solihns
|

5,

251 persiccel

solibus seems idiomatic, as also 46 distenta

4

uhera

259

niannis lerit

uberibus-distentis .- 4,

14

el

Appiam

3,

1063 curril

agens mannos: the word
L.;
6.

mannus
6

is

qo1

found before
his

probably these ponies
6 arnica
vis

were imported aboul
6.

limeI

pastoribus

1222

lida

cannm

vis:

think thai

neither Horace
(5,

nor Lucretius was indebted to Theocritus Theocritus
firsi

106) unless

introduced dogs into
is

Italy.

The para-

phrase with vis

xcvy

common

in

L.

bu1

not

unknown

'I'miht, Epicurea, [ndex s. \. Hordtvus, shows thai sources than Lm-rrfms I'm- Kpicuiv.-ui duel rinc.

Eorace had other

Vol.

i.|

Merrill.

On

Influenci

of Lucretius on Horace.

119

9

before and after him.

9,

1

has repostum, an agreemenl

with the old epic style thai

20 citae
11

permitted this syncopation.
is

4.

576 voce eiemus
a

an agreemenl

in

the use
2

of a

word

in

meaning

later

percussum
sit--

gravi

uncommon.

11.

amore
incusin

1.

923

percussil

thyrso-e1

amorem:
note- 13,

tliis

seems idiomatic, as Bentley shows
flumina

13

his

11
is

Scamandri

6,

1114
fifth

flumina
fool

Nili: flumina

a

convenient dactyl for the
a

and
EL

the use of the plural had become

poetic license thai
in

thought
14
14,

permissible here as elsewhere
|

other metres
[lion

13

non pulchrior ignis accendil obsessam

1.

474 ignis
16
love
4.
is

— clara acceridisset ~ certamen belli: common enough- 16, 31 tigres subsidere
:

ignis
cervis

of

1198 equae maribus subsidere possunl
is

this use of sub-

sidere

very rare;

it

was probably
desilil
a

a veterinary

termL-

48

levis

crepante lympha

pede

5,

272 liquido

pede
.~>4

detuli!

undas:

this

seems

bold reminiscence of

aquosus Euros arva radat imbribus
radentia
in

5,

256 imbribus

et ripas

flumina
II.

rodunt: the proximity of the
L. leads

two passages
17
that there
is

both

and

me

to the conclusion
17.

also Lucretian influence here.- In
is

66 the

reference to Tantalus
P
in general.

not significant.
\

In general, for the Epodes

find in

only one of them
in

any

real

evidence of Lucretian influence, namely

the

Kith, one of the earliest

written and contemporary with

the earliest

sat ires.

Carminal.
1
et

I

now

pass to the Odes.
et

The

first

parallel

is I.

1,

20

praesidium

dulce decus

meum — 2,
3,

643 virtute velinl

patriam defendere terrain and

897 non poteris faetis

Hon ntibus
ing
as
2
'21

esse,

tuisque

praesidium.

Bere there
is

is

noth-

common

except the thoughl which
- .">.

sufficiently trite,

is

catulis lidelibns

Sl!4

canum
et

corda.-2, 9 the prodigy piscium

lido cum pectore summa genus haesil

ulmo
3
3,

-3,

785 pisces vivere

in
5,

arvis have o

nnection.
late ter-

22 Oceano dissociabili
distinct

203 mare quod

rarum
seems

oras: this notion of the estranging ocean

Lucretian.

The plural
In

vada

in

24

1.

200

is

without significance.

30 nova febrium

terris incubuil

L20

University of California Publications.
eohors

[Class.Phil.

6,

11-1:!

incubit

morbus
In

tandem

populo

Pandionis omni, the verb and the metaphor are too eom4

mon

to

admil of imitation.

the

fourth ode the menin

tion of

Favonius

1.

11.

is

unimportant, and

7
5,

iam
T:;7

Cytherea ehoros ducil
il

Venus imminente Luna
in

ver

e1

Venus,
In

etc.,

have nothing

common
undique

but Venus'

7

coming.
fronti

the seventh ode. Line 7

decerptam
oubila

praeponere olivam
little
4.

1,

92S novos dcccrpcre no res
1.">

have
caelo

in

common; and
sibi

obscuro detergel

378 nigrasque
parturit

abluil

significant- 16

imbres

umbras

is

oo more

6,

259 gravidam temIn

pestatem atque procellis have a
8

common metaphor-

the 8th, line 10 gestat armis
cassa

|

bracchia

3,

1049 geris

formidine

mentem,

the

verbs

are

synonyms of
Troiae
|

habere,

an idiomatic use- 14, lacrimosa
is

funera

11

5,

326 funera Troiae
|

trivial.

In the 11th, verse 2

nee

Babylonios temptaris

lonica

numeros 5, 727 ut BabyChaldaeum doctrina belong to the common con5,

sciousness.vreseo

oppositis debilitat pumicibus

mare

sale saxa

peresa and

1,

305 fluctifrago

1,

326

in litore
I

can have no relation of influence.
aetas

Line 7 fugerit invida

3,

915 iam fuerit

is

an agreement in the use of

a tense.

12

In the 12th occurs the Latin word for echo
recinit iocosa
|

13

nomen imago 4. 571 imagine verbi.-In the 13th, line 12. inpressit memorem dente labris notam — 4, Hi)!) inspirant pressantes dentibus ora may be paralleled

line 3

16

from the

erotic poets.- In 16, 8

aera

geminanl Corybantes
is

2,

636 pulsarent aeribus aera

merely

a

referIn

ence to the ceremonies in honor of the
22
22,

17 pigris -"campis
suit

Magna Mater.
in

5,

746 pigrumque rigorem, and
|

21

curru niminm propinqui

solis

negata
24

terra

domibus
casus
24.

5,

204 fervidus ardor assiduusque
|

geli

mortalibus
lugubres

aufert

arc
4,

mere

commonplaces.- In

- cantus
2

2

548 lugubri voce querelam

have

no significance.

26

In

26,

protervis

- ventis

6,

111

petulantibus
1.

amis have no connection; and

6 fontibus

integris—

927

Vol.i.]

Merrill.-

On Influena
f ontis
is

of Lucretius on Horace.

L21

integros
L.

--

qo1
is

traced to any source earlier than

Probably there

Lucretian influence here, and Hie
in

28

thought occurs repeatedly

later writers.- In 28, 2
in se cohibet:
;is

men-

sorem cohibent
of cohibeo
lius
is

2,

1031 quaeque
(

this use
in

found

in

!icero
~>

;is

well

else\*

here

Lucre-

and Borace.

Line

aerias tempi

;iss<>

ilnmos animoipie

rotundum

percurrisse

polum

rituro-

174
is

omne
;i

im-

mensum

peragravil

mente animoque: here
et

distinct

reminiscence- 7 occidil
j

Pelopis genitor
is


a

3,

1027 reges

rerumque potentes occiderunt
918 leti«vias
is

probably

reminiscence

of L. as well of the stock consolations.- 16 via leti


noteworthy.

IS
1!»

avidum mare

2,

1,

1031
3,

use of an epitheton otiosum.
71
a

densentur fimera
1.

caedem caede accumulantes and denseri
Lucretian word.
it

656

etc.:

There

is

much
its

in

Ibis puzzling
is

ode

that sets
it

apart from the others;

date

unknown but
agreement

must be one of
is

bis earliest poems, hence the
1

with L.

not strange.

have no doubl thai there was

Lucretian influence on the ode.
:{1

In 31, 8 mordet

- amnis

5,

256 flumina rodunt
a

is

3-1

conventional- 34
insanientis

is

interesting as

palinode.

Verse 2

dum

sapientiae
|

5,

Id nunc appellatur sa-

pientia

;

5 Diespiter,

igni corusco nubila dividens
|

plerum|

que, per
nulla

purum tonantes sereno - mittuntur
caelo iacit
et

egil

equos

6,

247

nam
6.

caelo

<

fnlmina >, and
also

400 cur

oumquam
siinimis

undique puro,
insignem
it

12

valet

ima

nnitare

attenuat
is.

dens,

obscura
5,

promens, commonplace though

agrees with

L127
thai

fuimine
in

summa

vaporanl

plerumque.

It is natural

withdrawing from Epicureanism reminiscences from his old authorities
the

there
I'm-

should

be

that

insanieiis

sapieiitia.

Carm.
1

II.

In

si

ml book of the Odes,

in

the 17th line

<>f

the

first

ode-minaci
is

murmure cornuum
a

1,

276 minaci

murmure ventus
3

and 3D inpia proelia
significance.- In

mere agreemenl in onomatopoeia; 5. 381 pio Qequiquam-bello has no
third

the the

ode

which

is

Epicurean
corre-

throughout,

in

first

line

aequam - mentem

122

University of California Publications.

[Class.Phil.

sponds to aequo animo
the

5,

1119; and
like
2,

in

12 the invitation to
In

picnic

is

something
far

30

sq.

this

ode

it

is

remarkable
6

how

Horace

differs

from

Lucretius
1
1

in

describing Epicurean ideals.
ridel
1.

In the sixth, line
is

angulus
in

8 ridenl

aequora
21

a

chance agreemenl
postulanl
oot

use of
7

a

word; and

beatae
I

arces

the

2,
7,

8 is

sapientum templa serena
fessum militia latus

should
1.

press- In

deponeis

257 fessae pecudes pin8,

gui--corpora
taciturna
9

deponunl
signa

a

ooctis
9,

commonplace; and

10
is

4,

460 severa

another.- In
vexal
is

3 vexant

-

procellae

silentia
1,

noetis

275 venti vis -

10

idiomatic.
e1

In 10, 9 saepius mentis agitatur ingens

pinus

celsae graviores casu

fulgura montes

5,

1127

quoniam ceu fulmine summa vaporant, 6, 421 altaque cur plerumque petil loca; a commonplace which was proverLine 18 tacentem suscitat tnusam bial. 2,413 musaea
|

mele - expergef acta figurant may go back
11
snl)

to

;i

common
temere
et

source but have no mutual connection.- In 11, 13 cur non
alta vel platano vel
2,

hac
|

pinn iacentes

sic

rosa

30

sq.

have only the picnic motif in
vitet,

commonsatis

In

13

13, 13

quid quisque
horas

est in

numquam homini
in

cautum

3,

1085 posteraqne
is a

dubiost

fortunam

15

quam

vchal aetas

commonplace.- In

15, 11 sive reges

sive inopes

erimus coloni

3,

1035 ossa dabit terrae pro-

inde ac fannil infimus esset; 15 per autumnos uocentem
|

corporibus metuemus Austrum

5,

220 cur anni temet

pora morbos apportant; 18 Cocytos errans

— Sisyphus —
tellus
e1

Danai genus
21

3.

992 Tityos-Sisyphus
e1

etc.;

linquenda

domus

placens

uxor

3,

894

non

domus
all

accipiel te laeta,

aeque uxor
2

optima-these arc

16

monplaces- In
-navigia

16,

prensus Aegaeo

com-

6,

42!)

deprensa
sea.

probably belong to the language of the

Line 9 aon enim gazae

2,

37

nil

Qostro

in

corpore gazae.
is a

both with reference to dislodging menial terrors,

remi-

niscence of Epicurean doctrine- 13 vivitur parvo bene

5,

1118 divitiae grandes
Cortes
niti

— sunl
19

vivere parce; 17 quid brevi

iaculamur aevo

multa

—-3,

62 nodes atque dies
se

praestante labore;

patriae quis exnl

quoque

Vol.i.
i

Merrill.—On Influena of Lucretius on Horace.
fugil

123

3,

1068
fit.

hoc

se

quisque

modo
net-

fugitat,

quern

scilicet,

ul

effugere baud potis est; 22 cura aec turrelinquit
tela.

oias

equitum

2,

49

metuunl

sonitus

armorum
influence
in
t

nee fera

In this
is

Epicurean ode the agreeso striking thai
is

ment with Lucretian doctrine
is

a

dired

probable.

The ode

also

one of the

earliest

ime.

18

The beginning of 18-non ebur Deque aureum
renidel in

mea
fulget

domo lacunar
renidet

2, '21

aec

domus argento

auroque
20

thought

is a

have no immediate connection-, the commonplace and renideo is frequent. In 20,

21 alisint inani funere nacniae
arc mutually interpretative.

3,

955 compesce querelas

Carm.
1

III.
|

In the

first

ode of Book III, line 10 hie generosior
petitor

descendat in
is

Campum

2,

11 contendere nobili-

tate

a mere reference to the advantage of noble birth;

and
runi

41 quodsi

dolentem aec Phrygius lapis nee purpura|

- delenit usns
|

2,

34 nee calidae eitius decedunt
si

corpore febres,
2
iacteris
is

textilibus

in

picturis ostroque rubenti
is

another commonplace, as

2.

l'!»

saepe Diesexani-

piter

|

neclectus incesto addidit integrum

— 2,1104
3,

3

mat indignos inque merentis.-In
et sir

melius situni

3,

49 aurum inrepertum

5,

1113 aurumque repertum has no
taeter

11

significance- In
odore, there
is

11,

19 spiritus

581

taetro

doubt about the genuineness of the EToraIn 17, 12 aquae nisi
t'allit

17

tian passage.

augur annosa

27

eornix and

l'7.

H) imbrium divina avis
|

inminentum

^5,

1084 cornicum at saecla vetusta

corvorumque greges ubi
7

28

aquam dicuntur
-sapientiae

are merely proverbial .— In 28, 4 munitae
2.

reminds one of
is

templa and

probably

a

reminiscence
to.

munita -- sapientum famous of that

prooemium, here jestingly alluded
in general.

In

1llt

'

tirs1

three books of the
a

maturity of his powers as
to

lyric

poet,

Odes Horace is in the and has attained
at

independence of thoughl and expression, while
is

the

same time he

free
it

from the tradition of the dactylic
qo surprise
to find so Little thai

hexameter; hence

is

can

be said confidently to betray Lucretian influence.

Add

1

124

University of California Publications.
also thai he
it

[Class.Phil

was following Greek models

a1

this time,

and
for

is

not

to be

wondered

;it

thai

the only odes where one
influence
.-ire

may
Book

state
I.

with

confidence

Lucretian
for

the 26th, 28th and 34th;
III

Book
.-ill

H

the 16th;

and for Book

the 28th.

These are

exceptional for

one reason or another, and both Epicurean and Lnicretian
influence
a1

thai

stage of his developmenl

were at their
hook of

Lowest point.
Epistles
I.

Wxt
the
last

in

time, roughly speaking, comes the
;iikI

first

Epistles,

the

first

one wns probably composed

of

all.

In tlie

4lM

line

is

vides,

quae maxima credis esse mala,
|

exiguum
temptus

censum
et

turpemque

repulsam,
|

quanto devites

animi capitisque Lahore

3,

65 turpis enim ferme. con-

acris egestas-quasi

iam

leti

portas cunetarier

ante; these

Roman

evils are dwelt

upon hy Lucretius with

such force that undoubtedly Horace has him in mind.- In
52 vilins argentum est auro, virtutibus

aurum

5.

1113

aurumque repertum, quod
dempsil bonorem there
tihi
is

facile

et

validis et

pulehris

again agreement; and in 65 Lsne

melius snadt-1. qui rem facias, rem
is

5,

1113 posterius

res inventasl

also

reminiscent- In 82 idem eadem pos-

sunl

differ in

horam durare probantes, with what follows, does not thought from 3, 1058 quid sibi quisque velil
et

aescire

quaerere semper

commutare locum,

etc.

This

introductory epistle was composed when Lucretian influence over Horace had revived, and when also his philosophical opinion

was returning
in

to its

early position;

a1

a

time when,
1:!.

in spite of his protestation

of liberty in verse

he says

nunc

Aristippi furtim praecepta relabor.- In

2

the seeond epistle. Line 31 ad strepitum cithara(

I.

582

quorum (faunorum) -strepitu
to he the
first
it

is

noticeable, as L. seems

to

use Crepitus of a musical sound;

and

Horace has

also in C. 4, 3, 18

and Ep.

1.

14.

26.-40
relic-

sapere aude;
tis
|

incipe-

3,

K»T1

iam rebus quisque

naturam primum studeat cognoscere rerum are the same injunctions practically, and 47 oon aeris acervus et
aiu-i

aegroto domini deduxit corpore fehres

2.

34 nee

Vol.i.]

Merrill.— On Influend of Lucretius on Horace.
ealidae citius decedunl corpore febres
is

125

similar.

cerum
legil

nisi

vas

quodcumque infundis
v;is

acescit

54 sin17
intel-

6,

ibi

vitiinn

efficere

ipsiini
is

onmiaque
fincin

illius

vitio

corrumpier
had

intus.

This simile

ultimately Platonic, bu1
ii.

li.M-onic trite.-

56 eertnni voto pete
to

25 finem

statuit

cuppedinis would seem

3

3,

19 grex avium
4.


a

show reminiscence- In
is

5,

1085 corvorum greges

no1 signifi-

4

cant.- In
is

Hi

cum

ridere voles. Epicuri de grege

porcum

noteworthy as

jesting sign of the poet

's

return to his

earlier philosophy 6
is

and

to Lucretius.

In

ii.

1

nil

admirari

5,

83

si

tamen interea mirantur
ft

pure Epicurean. -In 3 hunc solem
in

stellas, etc., corre-

spond

thoughl

t<»

5,

1l'<>4

nam cum suspicimus magni

caelestia

mundi, etc-4, formidine nulla
11

5,

1218 formi-

dine divom;

improvisa species exterret utrumque

2,

1040 novitate exterritus ipsa- 24 quidquid sub terra
in

est

aprieum proferel aetas

3.

847

si

materiem nostram
5,
1

collegerit aetas agree in the use of aetas, as also

!•">[

sic

unumquicquid paulatim protrahit aetas
ire

in

medium- 27

tamen
sis

restat,

Numa quo

devenil

et

Ancus
is

3,

1025
one

Lumina

oculis etiam

bonu' Ancus

reliquil

proverbial

from Ennius.

The same thoughl occurs

in C. 4. 7. 14.

of the later odes.
7

In

7.

S opella-1,

1114 opella, the form

is

quoted only

once from Lucretius and Horace; and 24 dignum praestabo

me etiam pro laude merentis 5, 1 quis potis est dignum - carmen condere pro rerum maiestate; 7ii mannis arvum caelumque Sabinum 1063 currit agens mannos ad villain praecipitanter 84 vineta crepat
|


;

•».

mera
8

2.

1170

et
is

crepat

arc agreements

in

vocabulary.
8,

In this epistle

latent Lucretian

influence.- In

12

Romae Tibur amem
10

ventosus. Tibure
'>.

Roman
1060
95]
sq.

is

another
In 10, 7

form of the
nnisco

of1

repeated thoughl of

circumlita

viridi stillantia

aemusquemusco would not he
saxa
In

5,

saxa,

super

significant except for
I
1

11

the rarity of the word musco.
cul
e

11.
1

'

terra

spectare furentem


e

Neptunum

pro-

2,

suave mari magno

turbantibus

aequora

mentis,

terra

magnum

alterius

126

University of California Publications.
Bpectare laborem; the thoughl

[Class.Phil.

may have been

familiar to
i1

Sophocles and Menander, bu1 Eorace probably go1
Lucretius
since

from
27
3, u1

21sq.

is

Epicurean,

particularly

caelum, hod
L068
lit.

;iniiiiiiin

mutant, qui trans mare currunt

hoc se quisque

modo
est,
si

Eugitat,
etc.,

quern

scilicet,

effugere haud potis
es1
etc.,

and 29 quod
deficil

petis,
3,

hie

est,

Ulubris,

animus
ut copia

te

uon

aequus

939,

962,

aequo animo.
maior-- possil in
5,

L2

12, 2 nnii est

979 non erat

ut -posset 13

is

an agreemenl
13

a Grecism
es1

which L

has

more than once.
snni peragravit

dum

peregre

animus sine corpore
--

velox of Democritus-

172 vivida vis animi

omne inmen15 sublimia

moenia mundi of Epicurus.
matter discussed by
ferl
e1

cures

1,

127 superis de rebus habenda.-16 quae mare
a

compescant causae,
14
14. 8 istuc

L.

in

6,

608.

tra

amat -- rumpere claus—2, 264 prorumpere - quam mens avet ipsa- 12 stulmens animusque
is

tus uterque locum inmeritum causatur inique

the of1
se

repeated thought of
effugil
has,

3,

1059; and 13 animus, qui

qod

umquam,

of

3,

1068; and 14 tacita prece rura pete3,

nunc arbem-optas, of

1067.-22 incutiunt - desid;

erium

1,

19 incutiens-amorem

Lucretian.-35 eena brevis iuvat
in

et

and 26 strepitum, are prope rivum somnus
etc.,

herba

2,

30 propter aquae rivum,

which has been

compared
18

before.

This epistle was unquestionably written
influence.
18, 9 villus est

under Epicurean and Lucretian

medium vitiorum
trasque
metrical

e1

utrimque reductum
emissum-verhuin

— 5,839
to
5,

interu-

nee

utrum, utrimque remotum seem
-

have

a

likeness.

71

1044 sonitus
superesl

emittere
aevi

Linguae

seems

idiomatic- 108
superest,
5,

quod

3,

904 aevi
|

quod

206 quod superesl

arvi: here there
live with

may

be Lucretian influence as the geni

quod superesl is not common, and the phrase comes Later in >vid and Silius.
<

pi

in

19,

l'1

Libera

per

vacuum
|>ede

posui
3.

vestigia

prin-

eeps,

imii

aliena
2, 6,

meo

pressa
1.

4 pono

vestigia

Sat.
trita solo.

101)

926 peragro Loca oullius ante.

This seems to be an imitation, and 44 poetica

Vol.l

i

Merrill.— On Influena of Lucretius on Horace.

127

mellaeven
if

1,

947 musaeo melle
is

is

also probably reminiscent,

the source

ultimately Greek, as the adjectives

imply.

mineral'.
at1
it

ln seveD of the

twenty epistles of Book

I

there

is.

then,

Lucretian influence, and throughout the book the pud's

ude

to

Epicureanism

is

friendly.

Carm. Saec.
Epist. II.
1

in

the

Carmen Saeculare

there

is

nothing noteworthy.
1.

in the

second book of the Epistles,

8 agros adsig-

mint

5,

1110 agros divisere
portent;]

is

withoul significance- 11
5,

QOtaque

fatali

lahore snheuil

37 sunl

por13

tenta perempta have Hercules in

U ri1

enim
is

fulgore
not

suo

common merely; and
(329)

4,

304

splendor -- acer

adurit

remarkable .- 102 hoc paces habuere bonae

ventique secundi
tnasque secundas
2


is

5,

1230 ventorum pavidus paces ani-

a

chance agreement of words.

In Ep.
in spite

2,

32 claims ob id

1,

639 clarus ob obscuram

:

of
I

Laehmann's dictum
the

thai

Horace got

this

from

Lucretius
before

prefer to wail until the Thesaurus reaches ob

admitting

amantqm

1,

(>41

indebtedness- In 58 mirantur admirantur amantque seems unim-

portant.- 125

Cyclopa
et

movetur

3,

569

moventur-

motus; 135 rupem

puteum
;

vitare
redit

patentem

praeeipitesque locos vitare
se redeunt,

138

ad sese

4,

509

4.

1023 ad

all fail to

show any

filiation- 151 proficiente
is

nihil curarier

2.

39 gazae

proficiunl

an agreement
175

in

vocabulary
nnlli

merely.-- 159

mancipal

usus,

datur nsus

perpetuus

3,

971

vitaque mancipio nulli datur
rt is

omnibus nsn are commonplaces- 207 card
dine et ira

formi-

3,

1045 indignabere obire: here Horace unin

questionably has Epicurean doctrine
it*

mind, ye1
213

I

doubl
si

the

Lucretian

passage influenced him.
'!.

vivere

recte nescis, decede peritis-

938 cur uon-u| conviva
is

recedis, 9(i2 f

magnis concede: necessest
'

also Epicurean.

P
in general.

,n

tll(

'

s,

'
,

,,,1(,

book of the do qo1

Kpisiles

there
in

is

strong

Epicurean influence and some agreement
with
Lucretius, yet
t

expression

I

find

any

real

evidence of

Lnerel ian

radii ion.

128

University of California Publications.
M

[Class.Phil.

Carmina IV.
3

|

the
of
;

fourth

book of Odes the 3rd ode 1ms three
:

cases

verbal

agreemenl

I

clarabil
;

pugilem

3,

36

claranda
4
iiiiiiuin

L8 strepitum
1.


x

4,

582 strepitu

and 22 praeteret.

318 praeterque meantum.
•">.

In

lines 13, 24,

63
;,

1.

II:
5,

In'.!:

1.

seem

1<>

be mere verbal agree-

ments.
saecla

2!'

condil

quisque
7.

diem

3.

1090

condere

7

is

idiomatic- In
aestas,
e1 e1

9

frigora

mitescunl

Zephyris

vrv proteril

interitura, simul

pomifer autumnus

Eruges effuderit,
v.t
r\

nox

bruma

recurrit iners


|

5,

737

it

Venus,

Veneris

praenuntius
etc.,
;

ante

pennatus

graditur, Zephyri vestigia propter,
in

have a similarity
;is

thought, but

iid1

much
e1

of expression
|

is

also the ease

with 14 nos, ubi decidimus
dives.
e1

Ajqcus,

pulvis

quo pius Aeneas, quo Tullus umbra sumus 3, 1025 lumina

sis oculis

etiam bonus Ancu' reliquit.
writ en in
1

Probably

this ode
if

would have been
bad never been
9
a

much

the same form

there

Lucretius.
|

In
1037

9,

5

si

priores .Maeonius tenet
potitus,

sedes

Homerus

3,

Bomerus sceptra

and 25 vixere fortes ante

bellum

5, 326 cur supera Agamemnona r-carenl quia vale saero Thebanum et tunera Troiae non-ceeinerepoetae?
|

This ode
thoughl

is

one of the

latest

and

ripest and, although the

lias

much

in
il

with Lucretius, yet

common with Epicureanism and seems to me that Horace is inde-

pendent
11

in

bis treatment.

In 11,
both

6 ridet

argento domus

flammae trepidant rotantes
13
1
I

3,

21 aether- ridel

1
;

6,

202 rotantque
as
is

- flammam
20 oras

are

withoul
2.

significance,

also

13,

puerat

314 surpere, and

14. 6 inlustrant


5,

sur3,

2

Lnlustrans

commoda
<i<l.

vitae.

28

minitatur agris

386

amnes-minantur omnia diluviare are both commonplaces;
see Bentley
?„
loc.

Eor the Latter.
I

V ar in general.

?i^J

'

In the fourth book of the Odes

find no evidence of
his

Lucretian influence.
;ill ,l

Horace had attained

majority,

even

if all

the odes of this book are no1 his Latest proa

ductions, yet taken as

whole, the odes of the

last

book

show
LOth

little

indebtedness to any definite predecessor: the
is

(0 crudelis adhuc)

of course an exception and

is

Vol.i. i

Merrill.— On Inftuena of Lucretius on Horace.
probably an early study, and

129

I

would

qo1 excepl the Mel-

pomene
Ars Poetica.

ode. the 3rd.

Finally

there

remains the

Ars Poetica. 49

monstrare recentibus abdita rerum
verbis-cum
sil

indiciis

1,

138 multa novis

agendum

is

a

reminiscence - 61
has

cadunt
cance

prima
signifi-

4.

376

primaque

dispereunl

-70

cecidere

cadentque

pressed.- Ill

interprete

lingua

— —

qo

3, 6,

969

can

not

be

1149

interpres--

lingua is a coincidence, and may be paralleled in thought from Cicero.- 173 laudator temporis acti se puero 2,
|

1167 laudat fortunas saepe parentis

is

a commonplace.

359 dormitat Bomerus
test

3,

1037 Homerus-sopitu' quie
rabidos
a
3,

have

no

connection- 393
is

leones

4,

712

rabidi

leones

an agreement in

Finally 467 idem facit occidenti
quiete
in general.
'

standing epithet.

1038 eadem aliis-

is a

syntactical agreement.
to be

in the

There seems, then, Ars Poetica.
rl
'

but one conscious reminiscence

General
Conclusions.

-

,.

, '

'"'

llll;l

lvs,llts

u

,.

,,

.

oi

this

examination
Life

,.

may

,

be sum-

when Borace wrote his Satires, Lucretian influence was strong upon him; during his more mature years, as shown by his Odes, direct Luin early

marized as follows:

cretian influence

is

for the most part absent.

In the

first

book of the Epistles the influence of Lucretius again
vives, but

re-

afterwards

in

the second hook of the Epistles,
in

the fourth hook of the Odes, and
practically non-existent.

the Ars Poetica.

it

is

to
3,

Tim parts of Lucretius' poem that were most familial' Horace were the several prooemia. the hymn to Death, 830 sq., and the social epic in 5, 782 sq., thai is. the more
References to the purely
di-

Poetical parts of the work.

dactic pails are

infrequent.

PREFACE.

All
(lest

luii

a

few copies of the
the

first

edition of this pamphlel were
fire.

roved

iii

San Francisco

Hence
it
I

a

reprinting has

become accessary; and
ditions

in the course of

have made such adaltering
the

and corrections
In so doing

as
I

were possible without

pagination.

have been aided by the kindly reviews
Wochenschrifi lor

of Dr. Kirchner in the Berliner philologischi

1906, pages 980
I.

ft'.,

and Professor Capps

in Classical Philology,

pact's 4:>s

iX.

In addition, the article
l>y

by Dr. Kirchner, referred
Dr. Sundwall, mentioned
I

to

on page 146, and the pamphlel

on page 165, have Keen found useful.

have also entered archon

names

in

the opening table in accordance with later conclusions
in

reached by M. Colin and M. Iioussel

the Hull'

Ha

<h

corres-

pondent
and
by

kelUniqiu

for

1906 pp.

2191 and
in

1907

pp.

33
II

ff.,

me

in

articles published

Classical Philology,
text

:{

and Klin VII
very slightly.

l\

Speaking generally, the

has been altered

W.

S.

V.

UNIVERSITY OF CALIFORNIA PUBLICATIONS

CLASSICAL PHILOLOGY
Vol.
I,

No. 5, pp. 131-173

April 14,

1906

[Reprinted September 10, 1907]

THE PRIESTS OF ASKLEPIOS
A \K\Y

METHOD OF DATING ATHENIAN AECHONS
WILLIAM SCOTT FERGUSON.
I.

The substance

of this investigation

can be presented

besl

as

a

commentary on
Year
B.C.

the following table

353

2

L32

University of California Publications.

.

Phil.

B

1

Vol.

i
|

Ferguson.
Aivi,,,,,

Tlu

Priests of Asklepios.

133

].'';'.''

Demeof
Secretary
Secretary

Tribe of

Pi

Priest of \-i

273 272

2
1

Kekropia
1

1'W]ixok\9i(s) E[*rea«>s]f
I

1

Lppothontis

lemetrias

Avaavla[s Me]

Xt(reivs)

271/0

Pytharatoa

Aiantia
\nt Lochia
\

Erechtheia
Aigeia

1hIkv8os 'Avay(vpd<rios)
NtKO/tiax0S
.'

270/9 269/8
268
7

.'

ni

Lgonia

Pandionia
Leontia

AvK[ofxr)d]r)s
Krjdev)

K[o]p0v-

Philokratea

Melite

Demetrias
Erechtheia
Aigeia

.... -ovvi(evs)

2G7/6
266
265
5
(

Akamantis
Oincis
'Apx">A?? s Aa.Kid5(i)s)
Av<TiK[\]rjs -i'7ro\^r(tios)

Peithidemoa

Pandionia
Diognetoa
Leonl
is

Kekropia
Eippothontia
Aiantia

264 3

[IT/jo]/v\t}s

IIfip[at(ei/s)]

263/2
262/1
],";"'

Akamantis
(

Ai'/ceas Yafx.vo{v<jios)
*

Antipatroa
Archon
I

>ineis

Antiochia
Antiochia

<f>i\fas Ei't€cu(os)

Oincis
>ciiii-

I

KaWido-qs Aiyi\(uvs)
Priest of Asklepios

of

Secretary

Tribe of Secretary and of Priesl

261/0

Arrheneidea

Antigonia

0e6£ei'os llepyacr (^0ev)

Demetrias

0e65wpo(s) MeXiT(elfc)
[.
.
.

259/8 258/7
257 6

Erechtheia

.

os]

Evwvv(nevs)
'

Aigeis

[4>i'\i]7r7ros

\o}vi(8r)s)

Pandionia

AvTOK\TJs"Oade(v)
<Pt\oKpdTT7S 'KKa\rj(0tv)

256/5

Kleomaehos

Kett<

255/4
254 3

itis L Akamantis

IIpa^trArjs [Tijjtdpxoi;
Elpe<rlSr)S

(

)incis

Kr-qcritividris

253 2 252
1

Kekropia
Diogeiton
Olbios
Eitea

BotffKOS <t>\v(ew)

Eippothontia
Aiantia

251/0
250/9

Antiochia

249/8 248/7 247/6
246/5
Lysiades

Antigonia Demetrias
Erechtheia
Plotheia

-3«ire]r[«]

t

«i,

Kallimedea

Aigeia

245/4

GlauMppoa
Thersiloehos

Myrrhinus
Phrearrhoi

Pandionia
Leontia

244/3
243/2
242
1

Akamantis
(

(ineis

241/0 240/9

Kekropia

Hippothontis
Charikles

239/8
238/7

Rhamnus
I

Aiantia
Antiochia

Lysias

237/6
236/5

Kim- >n

I

Antigonia

235/4
234/3
233
2

Ekphantos Lvsanias

Hippotomadai

Demetriaa
Erechtheia
Aigeia

Pandionia

232/1
231/0

Diomedon
Jason

I. in. mis Akamantis

134

Vol.

i
|

L36

Vol.

l
j

Ferguson.
u ''
Argeios
Eerakleitos

Tin

Priests of Asklepios
leme of l'i of Serap

L37

I

b!£
96/5
95
i

Priest of VsWepii

94/3 93/2
92/1

Demochares
Diokles?

91/0
90/9

89/8
88/7
Fear
B.C.

Medeios Medeios Medeios
'Avapxla

Heme
I'hilanthes

of

Secretary

Trii f Priesl of Asklepios

Priesl of Asklepios

87/6

Ereehtheis

86/5

— ophantes

Aigeis

85/4 84/3
83/2 82/1
81/0 80/9 79/8

Pandionis
Leontis
Ptolemais

Akamantis
(

>ineis

Kekropis
Hippothontis
Aiantis Antinchis
Attalia

78/7
77/6
76/5

75/4 74/3

Ereehtheis

Aigeis

73/2
72/1

Pandionis
Leontis

71/0

Ptolemais

70/9

Akamantis
Oineis

69/8

68/7
67/6 66/5

Kekropis
Hippothontis
Aiantis

65/4
64/3 63/2
62/1
Aristaios

Antiochis
Attalis

Ereehtheis

— WKpaTTj? zapairtuvos
Kr)<pi<rieus

Aigeis

Qe68wpos XapiSr/nov iy yivppit>ovTTtjs

61/0
60/9 59/8

Theophemos Herodes
Leukios

Pandionis
Leontis

Ptolemais

58/7 57/6 56/5
55
I

Kalliphon
Diokles
Koi'ntos

Akamantis
Oineis

Aristos

Kekropis Hippothontis
Aiantis

54/3

Zenon

138
Vear
B.C.

University of California Publications.

[Clas

I'iiii..

vol.

i]

Ferguson.

The Priests of AsHepios.

139

ur-cd upon us by the fad that, when, in the second half of the second century B.C., the priests of Serapis and the secretaries
both followed the
called
official

order,

in this

case too the

same

tribe

was

upon each year

for the

two

officials.

The dating of the priests of Asklepios of 1 Gil 836 need no1 now detain us long. The tribe Pandionis is fixed for the secretaryship
in

22]

B.C. by the coincidence of the archon Thrasy139, 4 .-

phon and the Olympiad
this point,

Working back and forward from
It

we

3 tnusl construct, as Kirchner saw, the scheme of

tribal rotation for the third century.

then appears thai there
lisl

are only two
priests to 253

possibilities

— one

to

ascribe the
il

of

fourteen

2—24]
is

0,

the other to date

in

265 4
list

253
I'd.")

2

B.C.
-

The choice
2
it

not difficult.

For by locating the

in

4
in

l. :>.",

,

results that

the two priests

from Antioehis

fall

262

1

B.C.
It

That they belong to the same year
is

may

he taken for

granted.

analogous
in

to

what we find upon considering the

-ructions made

the board of

Amphictyons

in

-77 6
for

ff.,

4

and

is

in

accord with the practice repeatedly attested

the
fl

5 election of suffecti to the priests of Serapis.

Twice

in

319 3

and
in

in

296 5

7

a similar substitution of

magistrates took place

the middle of the year.
reelected.
in
I'Dti

On

each occasion the archon-eponymos
at
si

was
crals

The sane was done with one
5,

least of the gen-

Phaidros of Sphettos being

rattans twice

in

Xikias' archonship. 8

We

have Ion- since concluded/' from

evi-

dence which until recently was perhaps inadequate, 10 thai
year which ended the Chremonidean
like the revolutionists in

in the

War. Antigonos Gonatas,
Moreover, we
1

319 8 and 296 5 B.C., substituted for
set

the old magistrates

a

new

congenial to himself.
in

have lately learned that this war was ended
2

262

B.C.; for

Dittenberger
Gott. gt
I.

:

Sylloge*,

-•"<>.

11.

12

II'.

.I-:..

L900, pp. 433

ff.

•Dittenberger:
5

Sylloge*, 86; cf. Classical Review,

XV.
[1 5

L901, pp. 38

ff.

Sec above,
I'-!

p.

L36.
Beth-epos
I

Apollodoros

see

[

GD
I

Add. 299 b;
[15 299 e.

cf.

299

e.

7

For Nikias forego* G II 331, 1. 21. 'Droysen: Gesch.
I

li

II

299;

G

d.

Eellenismits,
\".

III.

I

,

p. 246;

Beloch
p.

:

Griech. Gesch.

III. 2, section

172.
in

Hegesandros

Athenaeus,

I

1>>7

t'.

;

cf.

below,

154.

140

University of California Publications.

[Class.Phil.

Athens surrendered
the

in

Antipatros' archonship; Antipatros was
Hol-

immediate predecessor of Arrheneides, and Arrheneides
30]

lowed Klearchos
lis.
I">\

by an interval of 39 years and three reckoning

exclusivt

Arrheneides thus

falls

into

261

".

For these reasons we musl date the
265
I

list

of fourteen priests

in

253 2 B.C.

The dating of
air
is
•_'">•'!

Mill

Add. 373 b

is

qoI

so easy.
is

The

limits
it

2

and 23d
for

29. and, since the priest

from Kekropis,
(i

possible

Lysiades to occupy either 247
is

or 235

4.

Bis

rival

for either of these positions

Lysanias, the successor of

Ekphantos.
same, since
a
I

One
(I

is

tempted

to

regard these two archons as the

TI

Add. Nov. 373b gives us only Ly[si]a[de]s

very easy misreading for

Ly [sa] n [ia] s. But
;

the temptation

to identify

them musl be

resisted

for the secretaries are different,

one being 'Apiaropa^o^ 'Apiaro
vos Evcovvpevs.

-, the

other JLvfirjXos

'E/ATreSia)-

and Lysanias.
the

Hence a place must be found for both Lysiades The decision comes from considering the predeHis name occupies eight spaces
1

cessor of Lysanias.

-

—precisely
Lysiades
to

number required

for the archon of 236/5 B.C.

therefore belongs to 247/6 and the priest from

Xypete
if

24S
list
ff.

7.

Since QeoSoopos MeXtrevs would occupy this year,
fourteen were assigned to 253/2
ff..

the

of
is

its

location in 265/4

thus

made doubly

sure.

Now we
official

can proceed farther.

But

first let

us remark that the

order of the priests was not broken by the Chremonidean
that

War, and

by

a

curious coincidence a priest from Antigonis

—the

tribe established by the

Athenians

in

honor of Antigonos

Gonatas' grandfather and namesake
nexl stopping place
is

— was

due for 261/0.
b.

Our

I

G
The

II

Add. Nov. 567

This precious

stone yields us a priest of Asklepios from the tribe Hippothontis.

and the archon
klepios
locates

Isaios.

official

order of the priests of As['

this

priest,

<-Pv\em Xaipiou

EXeva]
in

ivios ,

in

288

7.

and thus

settles a

much debated problem

favor of the

|

" See below, pp. 153 ff. Kmihk: Festschrift /'. Otto Hirschfeld, p. 317, has settle! this point. An additional reason for dating the archon in question, Ekphantos, in 236/5 en in Klio, VI 1. pp. 213 ft'. restoration is undoubtedly right. The decree is one of the tribe
|

Eippothontis.

vol.

1

Ferguson.

Th<

Priests of Asklepios.

Ill

virw originally

proposed by me

accepted by Kirchner."

and rashly The usefulness of the

(so

they

said)

official

order of
firsl

the secretaries' tribes as a canon in dating the archons of the

third of the third century B.C.

depended upon the maintenance

of the archon [saios in 288 7 B.C.
for [saios by the
official

And

since 288/7

is

demanded

order of the priests' tribes, when we work

backward From 262 1 and by the official order of the secretaries' 293/2, there is do tribes, when we work forward from 303/2

Longer any room for discussion as to the location of this archon,

and very

little

for difference of opinion

upon the

archon-lisl be-

tween 293 2 and 271/0 B.C.
Finally
cratic

we observe
in

that

niton the reestablishment of

demo-

government

307/6 B.C. the tribe from which the priest

of Asklepios

was chosen was Erechtheis

—the

first in

the

official

order prim- to the creation in that year of Antigonis and Demet
rias.
2.
15

Let us leave the priests of Asklepios

at this

point and turn

to the prytany-secretaries.

Here

too the official order,

which had

continued without an interruption from 353 2 B.C., was thrown
aside with the establishment of the aristocratic government
:!l'l'

in
till

1.

but, unlike that of the priests,

was not reestablished

three years after 307/6, in 304/3.

In the

summer

of this year

Demetrius Poliorcetes, at the

command
in
it

of his father, Antigonos,
a

abandoned the

siege of

Rhodes

order

second time
to

to rescue

Athens from Kassander, and

was doubtless

commemorate

his

victorious entry into the city thai his father's tribe, Antigonis the
first

in

the

official

order—was
tribe,

given the privilege of possessIn the

ing the secretaryship for the year then commencing.

year

303/2. however, his
si

own

Demetrias, was passed by and the

cretaryship was given to Erechtheis.
to find.
It

The reason

for this

is

not
'•^'>.

hard

was seemingly
in

in

the early part of the year

while Poliorcetes was absent
14 15

16 that the Strathe Peloponnesus,

Gbtt. gel.

Am.,

1900, pp. 436

ff.;

Prosopographia Attica,

II.

p.

636.

took place some nine weeks prior to the beas did that of the archon ginning of the official year (1 G II A.I.I. 489b) and the other ordinary magistrates ill 416). Ajitigonis and Demetrias began to exisl presumably on the first day of the official year. U Studies, VIII. p. 1.

The

election of the priesl

Beitr. alt. Gesch., V.

,..

174, n. 3.

Ill'

University of California Publications.

[Class. Phil.

inkles

Demetrius government was overthrown

a1

Athens, on the

issue of subservience to the

Macedonian prince, and Demochares
affairs into their
re-

and the democratic opponents of Demetrius took

own hands. 17
instated

To be

sure, the deposed

government was soon

and Demochares was Forced
beginning of the
official

into exile. bu1 the elections

and

i

lie

year,

we may assume, came
it

in

i

be

interval

and Stratokles did

oo1

think

worth while
lo1

to take the

secretaryship from the person
hold
it.

whom

the

had designated to

A dislocation of the two systems thus occurred.
this

And
a

I

take

opportunity

to

remark

thai there

was probably
that

small group

of annual single officers lined up with both the priests and the
secretaries.

The

dislocation

was such
in

when Antigonos Go-

natas came to reconstruct the state

262/] Oineis had the secre-

taryship and Antiochis the priestship.

doubl
261

awkward and
we

senseless.

It

was
in

The displacement was no not perpetuated, and in
at

find Antigonis, the

first

the official order, and
it

the

same time the
living

tribe of
in

which the conqueror was, as
break in the

were, the

eponymos,
is

possession of both the offices.
first

Thus
to reject
ciple.

explained the

official

order of the
scholars

secretaries' tribes
all

a

break which has been used by
it

many

archon-lists constructed with

as the guiding prinfor
a

The second break concerns us next

;

between the

last

years of the third century and 188/7 B.C.

breach occurred by
I believe

which some seven tribes were omitted.
explained
all aloni:'

This

can now be
is.

also.

The disturbing event

in this interval
t

as has

been affirmed, the abrogation of the
id' a

ribes

Ant igonis and

Demetrias and the introduction

new

tribe, Attalis.

We

have
in

abundance
200 B.C.

<>\'

literary

evidence that

Attalis was established

Moreover, V. von Schoeffer has recently shown us that.

between the disappearance of Antigonis and Demetrias and the
creation of Attalis.
a

short period interv< oed during which there

in Athens. 18 Antigonis and Demetrias Plutarch: Demetrius, XXIV. In this way the omission of Dem ean be explained, whether the official order began, for some unknown

were but
lT

e leven

tribes

303
••

with Ah, litis in 30G 5, or, .-is assui > above, with Antigonis in 304 ::. mochares was doubtless opp< sed ion cr the twi _ they had net yel become securely established. Pauly-Wissowa V. 1. p. and pp. 38 IV. [Cf. also Tod {At nual
1 I

:

:;•_'

B School </t Minns, 1902-3, pp. L73 conclusion independently of v. Schoeffer.]

ff.),

who has reached

the

same

Vol.1]

Ferguson.

Th< Priests of Asklepios.

143

wen' therefore abolished
occurrences
in

in

203

in all

likelihood.
this

The

importan.1

Athenian history during
into

year were the Acar-

nanian-Macedonian raid
between

Attica,

tl

utbreak of hostilities
has

Athens and
in

Philip,

and.

what

been

insufficiently

emphasized

this

connection,

the assurances of aid

given

to

Athens by Ptolemy of Egypl

with whose couri the city had been
''
1

on the most friendly terms for over twenty years.

The
strated

official

order of the secretaries' tribes
greater pari

is

amply demon-

for the

of the second century.

We

work

back, according to Kirchner's method, from the fixed points to

the uncertain period at

its start.

Roman
1

consul-lists or
7.

the

The system is attached to the olympiads in 110/9, 111* 1. 125 I.
to 201/0, the interesting fad
in

10

39, 168

and.
in

when continued
this
is

is

revealed that

year the tribe

the secretaryship was

Ptolemais.
tion

Everything

now
201.

clear.
to

The outbreak of indignathrow aside Antigonis and
for tribal distri-

which caused the Athenians
in

Demetrias 20 took place
bution of
ing anew
offices
in
2]

The machinery

was thereby thrown

out of working,

and

in start-

20]

the Athenians acted as they did in
off the tribal

261/0 and

gave the honor of Leading
of which too
the
tin-

procession to Ptolemais,

eponymos was the ruling king of Egypt, benefactor from whom at thai moment the Athenians confiliving

dently expected aid against Macedon.

For

less

than
visit

a

year there
in

were eleven tribes

in

Athens.

Then came

the

of Attalos

p.

at the battle

"I. ivy: XXXI. «). 1; ,•/'. Niese: Gesch. d. griech. u. maJced. Staaten, II. 580 and pp. 589 f. Niese doubtless dates the collapse of Egyptian power of Paneion two years It eame in arly. 198 B
|

t

44; <•/'. Dion Chrys: XXXVII, in relates how in the year 200, after the creation of Attalis (XXXI, L5j ef. Polybius: XVI, 25) the Athenians cul from the stones all memorials of the Macedonian rulers an. otherwise indulged their indignation againsl Philip. Either this occurrence is misplaced by Livy, or the tribes Antigonis an. Demetrias the Macedonian institutions at that time casl aside. It is qi sible thai upon the tiist violation of Athens' neutrality by the Macedonians and Acarnanians (Livy: XXXI. L4) the discarded thes tril.es. I'ulyliins says nothing of their abrogation where be describes in detail the circumstances under which Attalis w; -. G II 991 shows is an. Demetrias were oon-existem for some time before the creation .>(' Attalis.
I
I ' I

XXXI,

I

a Of course the disbanding of Antigonis
pla.-e in
,

an.!

Demetrias may have taken
I

the Iii that year 201 <>. rv an. other cas innual officials for the latter pari of 201 only were taken from the tribe Ptolemais.

Ill

University of California Publications.

[Class.]

200,

and the creation of the new
in

tribe Attalis.
0,

The
22

official

order,

which thus started afresh
tion
til!

20]

continued withoul interrupIt

the constitutional changes of 103 2 B.C.
ly

was then
There

abandoned, apparenl
3.

forever.
to

We

must now reverl

our priests of Asklepios.

are

many

of them belonging to the period from the fourth cento the

tury B.C.

second century A..D.
It

to

whom

qo year can be

assigned with any certainty.
of these.-'

will be sufficienl to

append
165

a

list

The

priests

who

are dated exactly between 229 and

88 B.C. are four in number.

They

fall

in

215

4.

4.

138/7,

and 126/5 B.C.,

and came from the denies Oinoe, Pergase, Phlya, and Bestiaia. These demes belonged at this time to Ptolemais, Erechtheis, Ptolemais, and Aigeis respectively, since the maintenance of the
official

the secretaries

demands

order for the priests concurrent with that for for these years priests from Hippothontis,
it is

Ptolemais, Kekropis, and Kekropis,

clear that the

two systems
whole or

were

no1 kept together at this time.

Nor do

the intervals between
;i

the priests allow us to insert these officials either as
paii's

in

upon any orderly scheme of tribal sequences. In other words, the official order was disregarded in the selection of the priesls of Asklepios during the time when it was maintained most
rigidly for the prytany-secretaries,
for chronological purposes,
is

and for

this reason our loss.

not a very threat one.
first
:>1

Proceeding down into the

century B.C.. we have evidence

from the years 63

2,

62

1.

and

that the priests of Asklepios
in

were again succeeding one another
tribes.

the

official

order of their

There can he no doubl as
at

to these dales, or as to the

main-

tenance of the sequence

this time: for the
1

arrangement of the
6
is

whole group of archons between 62
by the combination of
group, Berodes,
180,
a
•.-'
is
I

and 47

demonstrated

(I

III

1015 and 1014. and one of the

fixed

in

60 59 by his synchronism with 01.
at

In order to
alt.

determine the point
1

which the regular

I'„ itr.

Gesch., IV. pp.
17i'
fV.

IV.

below, pp.
'1

IORUS

I.

4, s:iys:
is

6r)<rav 6\vfj.inddos ttjs
'

ro&rov «aTocTT^s kcli
Ii

\0iivr)cnu 'llpwoov.
cl

bj
B

(Julius Caesar) al vpQrai irpd|e« &rereX&rKara rb irpwrov eros ew confirmatory thai Theophemos, the predecessor of Kastor (in Eusebius I. p. L83, 8, p. 295, 33 Schoene)
8'
6-y8oT}KO<TT7Js

to

i;i

kk:

I'.A.

7092.

Vol.1]

Ferguson.

The Priests of Asklepios.

L45

succession was resu
63 2 as
a

d

fixed point,

and

we have again to work backward from is It this time we have qo1 far to go.
in

obvious thai

whal
a1

happened before
this time.

262

1

and 201

B.C.

happened again

In 88 B.C. the

Anthenian democrats

Looking for the coming of Mithradates the Great, overturned the

pro-Koman aristocratic governmenl which had existed from 103 2 They put themselves into the hands of two military B.C. on.
leaders
so-called tyrants,
6.

and offered

a

desperate resistance to

Sulla in 87

When

the

Roman

pro-consul captured the city the

25 the preceding year aristocrats were restored,

was marked on the
307 6 and
At

list

of the eponymi as avapxfa, and the offices were reassigned.
priest
0,

The
26]

of

Asklepios for 87 6 was taken, as
first

in

from the

tribe in the official order
in
87',

Erechtheis.
6,

what time the scheme of sequences, begun
1

ceased to exist,

cannot

at

present determine. 26

III.

We
1.

must now return and take up

a

number

of points in detail.
is

The

list

of secretaries to the treasury-board of Athena
t

given to bring out the fad

hat

1

heir official order does not concur

with that of the prytany-secretaries and priests.
the three sets began cannot he determined.
'2.

At what time-.

HaTai/cos

(P. A. 11G77)

was

priest shortly before
rare,
it

843/2

(archon

Pythodotos).

The name being

is

perhaps adI

missible to identify him with Mutcu/cos 'EXevaii'Hx;
col.
I.

<i

II

834

b,

50 (329/8, P.A. 11679).

Aucn'tfeo? TpiKopvaio<i

(P.A.

!)4(>7)

appeal's in
in

I

(i

TI 7<i7

1.

19,

and

14o!».
list

lie

must have been

priest

Til
I

3,
<;

it
II

this
7ti»i:

fragmenl
hut thai

—a
is

of donations to Asklepios
t

followed
''•

impossible, for

he

lisl

for 334
7(iti.

is

certainly formed no part of
iu

and 7h7 Hence 7n7 must precede 7<i<;.
extant
in 7nti its.
'If.

which case
is

it

should probably he joined with Add. Nov. 766
''>

b.

Lysitheos
B.
tr.

therefore assigned to 344
Gesch., IV.
p.

B.C.

alt.

17.
firsl

A hurried survey of the data for the nothing conclusive en this point.

two centuries

A.l>.

re

1

lii

University of California Publications.

[Class.Phil.

[For

Qovyevvs,

see

Sundwall, Klio:

Beiheft

fFp. 75 and
for

Kirchner, Rhein. Mus.
in

LX1

p.

349.]
in

ErriKio/s 'AXotevs appears
the following connection:

the

A.sklepios-lis1
\II

341/0

M.vvviov
I

ravras
7(>6
II.

e<j)7]

6 iepe(v^)
;

EvriKtcijs '\Xat(eis) 7raXaia<; elvai
y\i')j<T(ipt Ti)
:

(J

7

{'.

cf.

1.

•'!:

A, iWeiirei

|- |- |-,

TavTa? 8eZ[v]
affirmation
in

e'(£?7

airohovvcu

\i<>-

x\ea Mvppi(vovaiov).

Such an
339

could

have

been

made by no one
IloXu|ew9,
pographia."

excepl the priest
in
8,

charge for the year.
to be

priest

seems
to

missing

in

the Proso-

Be

is

possibly
T

be

identified

with

UoXv^evof

Uo\vK P d{Tovi) Sowtew
For
Teto-ta?, priesl in
(

G

II

864 (I'.A. 12066).

338 7c/.TetoYa?Ke<£a\?70«/(P.A, 13478).

TeXecr/a? <J>Xt»(ew)
list

I*.

A. 13520)
11.

is

mentioned as

priest in tin-

of donations
in

I

G

II

766

66 and 67.

No

priest

appears

else-

where

this or the similar lists except the priest of Asklepios.

If the donations arrived with about the
(11.

same frequency
(11.

in

338/7

29

ff.)

and 337

'6

as in

340/39 and 339/8

8 ff.).

we should

expect the priest for 336/5 to appear anywhere in the neighbor-

hood of
3.
in

1.

66.

<J>tXo^a'/o?;?

"Oadev and

'Ovjjraip

MeXtrefc are mentioned

the catalogue of donations published in I

G

II 835.

This

list

follows that of which

part— for

the years
ft',

341/0—336/5 ff.— is
336/5.
it is

extant

in

1

G

II

766.

Since lines 8

of 766 belong to 340/39,
in

and

lines 66

and 67 deal with dedications

clear that
0.

line 11!).

with which the catalogue ended, reached to about 332
t

We

can therefore place the beginning of II 835 at about that
in
it

ime.

Consequently the seventeen priests mentioned

belong in the

main after 330
lished before

I5.('.

Hence

I

(I

II

835 cannot have been pub-

313 2 B.C. It undoubtedly was set up much later. The dating of Onetor and Philochares and of the other fifteen
iii

priests
in

this

group has been carefully investigated by Kirchner
zur attischen Epigraphik. 28

an article entitled Beitragi

The
in

conclusion
321

reached
3L'o

is

that

the two
i

named above
,s:i5.
I.

held office

"and
i

19: I'hilippos
in

1

<;.

II

7m

in

319

8,

and the
f.

other fun tern as indicated
"Similarly
'EiriKpdrvi (I

the table given above on pp. 131
1.

G

II

835

61

1

and HwMwkos

{ibid.

1.

50) have been

omitted. Rhein. Mus. L XI

(1906), pp. 344

ff.

Vol.1]

Ferguson.-

The Priests of Asklepios.
beginning
265
of

117

Beyond the
(which
is
is

fully

extanl

i

I

i

Mill
Mill

836
835

written od the back of the stone on which
lie

inscribed)
to

the years of ten priests, who as ex-officials made
in
Lm'li
.">

donations

Asklepios
a1

and

L'ti:;

2.

Seven of them ap2.

pear together

the end of the catalogue for 263
a
lo1

At this poinl

the commissioners placed in the inventory

of miscellaneous
ol /epet?

items

— the
at
it

weight of gold on hand, the a pyvpcofAara ok

ixpwvro

etc.

Perhaps an assortment of cult-furniture which

had heen contributed by the priests themselves was put ou1 of
service
etc.

this time, a censer, a ladel,

several rjSvrroTLa, a bowl,
lisl

Or

may

be that the commissioners chose this poinl to

the articles of the

permanent

outfit

which were donated by priests
fell
;

who
by

held office prior to the year in which Athens
in this

for

all

bu1

three items

part of the inventory concern objects donated

priests, while of the others

one was apparently the property

of Asklepios himself,

and

a second

came

as a gift

from the

8i)fxo<i

of Alliens.
It is

is

not

important to decide whether this nest of dedications
in

the result of something done
,

262/1 or of the cataloguing
first

in

L^!L

1.

In

any case

its

insertion just before the
in a decisive

year of the

.Macedonian regime checks

way our chronology of the
listed in

whole period.

The
dent

first

dedication bythe

8i)p,o<;

was
to
I

263

-: the next
a

was made
;

in

256

5,

and

this

seems

have established

prece-

for in the
is

two following years
l

for

which alone
a

a

complete

Catalogue

extant

the state likewise
set
iii

made

gift

to the shrine.

Was

the precedent

the year in which Alliens regained her
.">.

freedom.'

That

is

given by Eusebius as 256

not.

;is

is

usually

affirmed, 28 as 255

4.

Macedonian money appears among the dedications
time
in

for the

first

what we have determined

to he
(1.

L'til

o.

Tims during
In

the

priestship of
yoveioi'

Theoxenos of Pergase
le.

45) a [rer/oa^/ioy ^Aiti]

was given as an offering
1 1 1

Euagion.

256 5

four

436. [Hieronymus and Syncellus och, Griech. Gesch. p. This is the year of Abraham 256 5. (524, L2) assign it to ol. mi. L 'I'll-Armenian version puts it in the year of Abraham 1761, which is L761. still Zohrab's reading of the 255 t. equated, however, with 01. L31, 2 01. 131, 1 256 5. [rmenia puts it in the year of Abraham 1760 cf. Jacoby, dpollodors Ch Sehone, Eusebius, M. pp. 120
•_'.

= =

=

=

t'.

:

n.

l.|

148

University of California Publications.

[Class.Phil.

rerpax^" 'Avrtyoveia were dedicated 86), and some more in 254 3 (1. 93).
After
a

(I.

80),

in

255

I

three

(1.

period during which
in

Athens had
229,

Los1

her righl

of

coinage the mint was reopened
coins began.
30

and the new styh of Attic
of suspension to be
It
is

Head determines
Thai
is

the period

from 322

to 229.

assuredly wrong.
1

impossible to
in

believe thai

between 307 and 262

Athens coined no money

her
261

own name.
it

The

firsl

appearance of 'Amiyoveia rerpaxp-a
il

in

tells clearly

enough when

was

thai the old styh of Attic
J.
P.

cuius was abandoned.
i

Head, 33 following
This view
as
is

Six, 32 attributes

he

7-

rpaxp-a' Avrtyoveia of our catalogue toAntigonos, the father
is

of Demetrius Poliorcetes.
1

based on

a false

dating of

i;

II

836.

The coins belong,
a

now

clear, to

Antigonos GoIn
in

aatas,

and can probably be

identified

by the numismatists.

any case we have again

decisive check

upon our chronology

the fact thai these rerpaxpa 'Avrtyoveia
in

make
in

their

ftrs1

appear-

the priestship which

we have dated

the year after the

fall

of Alliens.

Before assigning to precise years the ten priests who made
dedications as ex-officials,
it

will be well to Look to the limits of the

inscription in which they occur.

Its

beginning

is

extant in

a

fragmentary condition, and. judging from the normal number of
lines required for a year,
it

appears that

at

least five

annual cata-

logues preceded thai

from 265/4.

How many

more there were
year required
it

depends upon the extent of the lacuna between fragments nh and
d.

Since, as will be seen in a

moment, the

earliest

by the tribes of the ten priests involved
able that the
list

is

275
It

4.

is

conceivin

began

at

about thai time.

ended
it

232

1

B.C.

Hence about 45 years were included,

since

is

to be sup-

posed thai the inscription on the fronl of the stone was equally
Long,
it

is

probable thai the two sides contained

a

continuous

narrative, and together listed the dedications from aboul 330 to
toria A
1

ummorum
201.
di

p. 3 L6.

Op.

eit.

p.

Numismatiqm

L882,

p.

27.

I

serial or to

any other of the numismatic journals. wngsb. d. Berl. Akad. for L896, pp. L089
of these points.
|

have nol had access to this notice thai Koehler in tl'.. has already taken issue
I
|

with

Bead on mosl

Vol.1]

Ferguson.-

Tin

Priests of Asklepios.

140

232

1

B.C., the juncture being a1

aboul 276 5 B.C.
the

And
1
<
;

this
n :'>~>

proves to
wras
sel

bend an approximate bu1
in

exad date;
officials

for

II

up

the arehonship of

E[u- and the
In

secretaryship of

K\«7[eV?/<?].
begin
in

Between 332

I

and 274/3

whose names
3
is
it

this

way

are possible only twice.

277 6 or 284

laios?

was archon.
to

Mis secretary

is

unknown.
is

Bence there

nothing
either.
earliest

exclude his year.
5,
II

Bu1 there

nothing to

commend

Whereas 276
priest
in
I

the year which
836,

preceded that of the
a

<J

and the year of

most important

change of governmenl
secretary has not been

in

Athens, 33 has for archon Eubulos.
hitherto.

His

known

We may
in

therefore safely

conclude thai
case,

1

<i

II

835 was inscribed
a

276

5.

That being the
;

we have found

reason For

its

peculiar arrangemenl
in

for

this

catalogue differs from the rest

that the donations are
in in

grouped, not under the names of the priests

whose years they
the shrine.

were made. bu1 according

to their location

This

was natural,

if

the articles were listed in 276 5 B.C.

The many
order

changes of government and the abandonmenl of the

official

during the preceding
to

fifty

years made

it

a1

thai time impossible

arrange the dedications chronologically.
Since
it

is

certain thai the ex-priests belong to the period im\\.(
'..
I

mediately preceding 265 4
into which their

have assigned them to the years
[n the process
1

demes distribute them,
273
2,

wo

restora-

tions have been made, one rash, the other probable.

[Tt]/zo«:\f/( ?
in the tribes

'E- belongs
viz.:

to

272

1.

270 69, or 267

6,

and

which are involved by these years only four denies begin with

'E,

Erchia and Erikeia (Aigeis 270 69), Eiresidai (Akamantis

267/6),

and Eitea
(1*.

(Antigonis
last

273/2).

Because of
has been
less

Tifio/c\fj<s

EtVeato?

A.
of

L3733) the

possibility

preferred.

The
[It

ease

Avaavia[s

— ]\t

leaves

for

guesswork.

has been restored with Me]Xt(Teu?) by Sundwall, op. cit.,p. 78,
{Berl. Phil.

and defended by Kirchner
.

Woch., 1906, pp
and below pp.

985

f.)

alt.

Gesch.

V

pp. 167

f.,

1.70,

17:;.

It will be observed that the change from the financial board ol bcl rjj StoiK^irei to the single officer, which was made in 276 5, was accompani* >1 bythe transfer, in p of certain of the duties mensi for inscribing documents, t<> the ra/xias tCov GTpaTiwTiKuv. the military treasurer in this connection is still I *'< II 835; ef. Lar.

feld

II

2,

p. 722.

L50

University of California Publications.

[Class. Phil.

againsl

the objections
1.

made by me, below

p.

169.

Lysanias

is

then assigned to 272
4.
It

.\ few remarks may oow be made on the archon list. should oo longer be doubted thai Philippos belongs in 293 2 Dionysius B.C., and thai oo oame is lacking in the list given by between Philippos and Diokles 200/89) for of Balicarnassus
i

;

place must be found for

Eimon and Charinos.
eirl 8e

The reference
34

in

the letters of Epicurus

Xapt[vov kuI eVt] Aioti>[ov]


to

makes
locate
(

it

practically impossible, as Kolbe rightly remarked,

lharinos after Diokles.
I

rather than to the will have to be assigned to 292 year (or to the preceding) because of the connection befollowing tween the policy of Phaidros in this archonship and the situation

Kimon

1

inevitably arising out of the revolt in

Boeotia

in

292

1

B.C. 34a

Phaidros was doubtless moderate

in

his politics.

He

held the

generalship between 301 and 296/5, and in 296/5 under both the He continued to serve under the dearistocracy and Lachares.

after the moderates again

mocracy of 295/4-276/5, and was publicly commended in 275 4 assumed control. Moderate eounsels were much needed in Athens in the year
which followed thai of Philippos, for the extreme oligarchs were then hack from exile" and many reasons urged the city to join in
the unfortunate Boeotian rebellion (292/1).

That Athens mainit

tained peace, freedom, and a liberal government was.
,l,

seems.

lr

to the influence of

Phaidros

in

Kimon's year (292/1).
is

The
full.

passage

from

which

we learn
Brj/xov

this

worth quoting

in

XeiporovrjOeU 8e virb rov
rbv rov
koivy)?
e7ri

eVt ra orrXa

arpan^os

rov iviav-

Kip-covo? cipxovros BiereXeaev aycovi&pevos vrrep 777s
fcal

o-(UT?;/3ta9,

rrepiardvrwv

rel

rroXet

Kaipcov SvatcoXwv

hcefyvXa^ev ttjv eipr)vrjv r>)i

X^P ai i arro^aiv6ixevo<i ael
teal

ra

icpdrio-ra,

Kal rov alrov

i/c

t>}?

X^pa?

rovs

aWov?

Kaprrovs curios iyevero
c.

elo-fcopuo-dr)vai,

avp(3ovXevaa<; tow Stjpcoi ovvrekeGai. (erasure of

38

XXX, p. L03. [See, however, my article on the Death of Classical Philology 11. where it is shown thai Charinos, Philsucceeded one another in 293/2, 292 I. and 291 0.] ippos, and Kii
'

Aih. Mitt.
in

Menander

i

Kolbe:
n.
1.

loc. cit.,

pp.

L03,

L08; Beloch.: Griech. Gesch.

HI,

l.

p.

234,

35

Dion. Hal.

:

1><

Dinarcho

LX

=

p.

651;

cf. Beitr. alt.

Gesch.

V, p. 161.

vol.1]

Ferguson.— Thi

Priests of Asklepios.

L51

letters in

which there was some reference

to

Demetrius)

xa\

rrjv

ttoXiv e\ev6tpav kcu Bt/poKparovpe'viju avTovo/xou irapehuticev /an roil?
vo/aovs

Kvpiovi
is

roU
in

pe6'

kavrov

(erasure of

c.

71

Letters).

Thai

to say,

spile of the crisis peace

was maintained, bu1
in

contributions of money were accessary to gather

the harvest.

The penalty

for indiscretion
life

would have been the destruction of
and property, and
a

the legal safeguards of

rabid oligarchy

upheld by the drawn sword of Macedon.
II'
!

(i

II

310

is

;i

con-eel reproduction of Hie stone,
in

ii

seems
as

impossible to restore the archon-name found

line 24 excepl

'OXvpirioS^pov.
earlier

In

that case the decree which occupies the

part

of
to

the

stone precedes 301

B.C.;
lie eit

toe
t

the rapias
rap,ia<i
r.
36

was instructed

pay the cost.

This can

her he
rap,ias

rov

Stjpov or the rap-ias rcov arpaTtcoTi/cwv.

The
first

8.

was
the

abolished in 301: the rapias

r. a.

was

entrusted

with

payment for the inscribing of documents on the change oi government in °.7(! 5. Aiaxpwv Upo^evov^ to whom the decree in
question

renders

praise,

290/89 or the following year.

was given Hence
is

Athenian
it

citizenship

in

is
1

much
<! II

the mosl likely

thing that
elpijvi]*;

the rapia^

t.

8.

meant.

310 was passed
is

Se yevop-evi]*;.
to.

The end

of the "four years' war'"

prob-

ably referred

Aur^/o&w, the leader perhaps of a pro-Athenian
his

party

in in

Delphi, was accordingly lauded by Stratokles and
c.
3»).'}

friends

-

and by the same governmenl upon
the time the

its

restoin
'

ration in 294

•'!

(Olympiodoros).
at

For befriending Athenians
Aetolians seized the shrine,

Delphi, probably

he was finally given the citizenship in 290/89.
all

Aischron was

in

likelihood the most
in

prominent man among the out and out
dating the

democrats

Delphi.

We

shall

have

to

reconcile ourselves after
exile,

all

to

return of

Demochares from

the revolt of

Athens from
in

Demetrius Poliorcetes, and the storming of the Museion
I

289

(i

II 331.
p.
I

See above,

19,

a.

33.

I,i ly-Wissowa: tV, p. 2568; Jahrb., L897, p. L87. Pomtow concludes thai Aischron was qoI a Delphian because his uame is wanting in the Delphian inscriptions. The same argument would convicl Lachares of (evlat in

Athens.
tion
«

»\r M7r jo5w]poi'
4::»;

1902 p.

[Kirchner [Berl. Phil. Woeh. 1906 p. 985 objects to the restoraon Oie ground of lack of space, Bnd refers to // where he lias suggested 'ArriirdT]pov (262 LB.i

152

University of California Publications.

[Class. Phil.

and Plutarch
rates the revoll of

will have to be corrected where he narAthens after the expulsion of Demetrios from

Macedon, 88 the only alternative being thai sunn' fallacy exists in 40 our calculation of the limits of Demetrius' reign. G II 331 must have been Urios musl precede Eubulos; Eor
I

passed

in

the year immediately after Eubulos

no1

a

IVw years

later, as

Kolbe assumes.

Certainly n

le

who

accepts Beloch's
I

very plausible dating of Eubulos in 276 5 should attribute
to any year bu1

G

II

33] observation thai the year of this documenl was the second of an 41 This being so, a Olympiad thai Eubulos is assigned to the first.
single officer inl
ttj

thai

of his successor, since

it

is

upon the

Siouajaei appears in 275 4. whereas in Urios'
existed.

year the board
It

still

Hence Tries belongs

to

285

L 42

is

likely thai both Telokles and

Laios? precede

Eubulos;

for after

Epicurus' correspondence.

Eubulos no archon-name, unless it be these, appears in Thai would seem to have ceased with

the infirmities of the philosopher's Latter days.
* The agonothetes,
ffK€6a<rev re?

The only
(Tri]derov
8r)[xov

possibilicare-

AJjulvrpi

«u

Philippides of Kephale, in 288 T, re? K6pe]t [VpJwTos inr6nvr},ua tt,s tov

dyuva

[iXevdeplas]

B( itr. alt.

Gesch.
in

Ill

2,
is

p.

V, pp. 176 ff. pp. 93 IT.) criticism of Beloch's eonclusioii 80) thai Demetrius Poliorcetes was expelled from

Gesch.
it.

(Griech.

Macedon

l'ss

in

B.C. the early
i !

qo1

fatal.
7

The

summer

of 288.

n delivered attack on Demetrius may have Eis abdication was probably made on his deI

for Asi:i. in 288

or later.

41

This Kolbe seems to have overlooked.
l

I,, G II 325, which Kolbe in contradiction to Koehler, who after seeing the stone (115 325 decided for Arrheneides, Locates in Kallimedes' archonol iirl ship appears, as between 295/4 and 276/5, the board of officers erftitled 5p.219 ttj Sioirfaei. [nUAdd. Nov. 373 b (248/7), 11305(245 li. E<p. APX we find on the ether hand 6 brl t% 5u>ucf)<rei. Again and II 334 (232 (235 6 brl t% 5«hkij after 229 B.C. oi e-n-i ttj owixricrei recurs, and before 201 ceeds. It is easy to understand thai in -J'-'!* the democracy reverted to the practice of the earlier democrats of 295 4-276 5 B.C., and then dropped it when the firsl zeal of the restoration wore away and the advantages ol cue Bu1 h..v, explain the responsible administrator prevailed over sentiment. v, e of the college in the middle of the centurj stone has t-o]i5[s, otherwise the easiesl way would assurance thai the It is, however, possible that a college was reapa misreading. pointed when the Chremonidean War began, and thai it remained in charge
. i 1
I

.'

[n that e to have a i ree hand in 256. A.rrheneides and the diss, of Diogenes on Karystos, p. 341) which yield rbv C matter is Bui tllf W tf. TOl'SfVl T T) StXHK^ffei. fTTt T7)S 0L0tK7)<T, [Kirchner (Berl. Phil. Woch. L906, p. 987) suggests thai the tin. G II 325 be restored ['Eiri Qpaffv^SoU |, and thatThrasy archon name in notice, by the way, B.C. e located in one of the years after 221 his hist thai an exceptional letter of Epicurus "as written in the period of

case n 325 should Laertius (Wilamo
1

d

to

!

1

1

<

<

1

I

I

I

i

in 271

l

1

vol.

1

Ferguson.-

The Priests of Asklepios.
6,

153

ities

are 284 3 ami l'77
e

and there

is

no means of deciding

Which Of these
5.

es jo eaell.

Beloch has assigned Antipatros to 263 2 and Arrheneides
1.

to

262

The determining passages are
Kai

as follows: 43

A7roAAo[8o)]po? Se to ku[6y)i[ri6rj(Tt r
)

p^rjaOai

i/v 7rdA.1v

[tV

'Av-

Ti7r]aT/oou t[oi']

tt/jo

Appcveio [ov

K(u tppovpa.[v eis] to Movtrelov [totc
£ " r/},\^['

u

^""'J

'AvTiydvoi; [kou rag
/cat

apX<is [avrjipijard^ at

7rav er[t

fSovkev [eir

>

ec/>]

ei<r#,u

KaOairep iv r v[ t ] ^fpu^ovarji ra
€7rtcrToA.7}t
7rep<.

'Ai'tic/)oji/[tos
[t]
,

Aeyera
Z?;vwv
(7
.

yivera. [i /3etoe

/3ia)K<os 6

.

.

•a rcui/ p kui

ctojv.
€77'

a7ro

KXedp)(OV yap
8^i'.
icfi

'Ap/j]ej/e[i-

ov

(rrjp

[etw^J rjvat [re-

TeAei,'T?/K'€i'a [t]

Z^vwva,

«r?/

eVtu' iwea

Kti.

[t]

Tpuixo [vtu

Kat /x^ves rptis.

[yeyoi'tVut
avdrjv
'

KAe-]

€7r'

ap^oi'[ros]
[

ApurrtHJiavovs Ku
tr^oAI/i' 8tu
€7r'

1

T>yi'

[KaTa-]
i/-

o-^ar

erv^
v.

rpuiK [o]

tu kui [e]

aTrrjWayr) [8'
d<roi'0?

ctt'

ap)(OVTOs
/a

1-]
/.

€t[w]v ra

[dAlOT

p]

.

The sequence
established.

of Antipatros

and Arrheneides
in

is

thus clearly

Klearchos was archoa

30]
0.

B.C.

Thirty-nine

years bring as to the beginning of 261 carry as as well into the year 261
Beloch's 44 calculation
For
.-i

Three months can

as into that of EOearchos.

is

nut the only one possible.

more exacl presentation of Cronert's reading of these papyrus-

II 2, pp. 424, 39, 472 f. ii The text here given indicate the varying degrees of certaintj of particular l<

"Griech. Gesch.
schi

Ill 2,

p.

424.
p.

Literatnrseit. 1907,

[Kolbe's emphatic affirmation thai 934) should ool lead any one astray.]

if

i-<

.'.
]

t

University of California Publications.

[Class. Phil.

The public tomb was decreed
fifth

to

Zeno

in

the latter pari of the

month, Maimakterion, 45
is

a1

the requesl of King Antigonos. 40
his

This being the case, h

probable thai

death occurred two
<>.

months

earlier, in the third

month

of

-til

Zeno's successor,

EOeanthes, was head of the Stoa for

upward

of thirty-one years.
<>

By

inclusivt

reckonings this brings as to 23]

for bis
is

death and

for the

archon Jason.

The

Lack

which eVt

calls for
to

explained

by the three months of Arrheneides' year given
calculation
is

Zeno.

The

then verified by the equation 331
(

<>

(Aristophanes)

minus

i':!1

Jason)

= about

100.

"We need not concern our-

selves bere with other computations as to the lives

and headships

of

Zeno and EQeanthes, since

it

is

through the one which he

himself gives thai the years of Philodemus' archons musl be arrived
at.
it

Since

was
47

not

till

262
is

1

B.C., and, if the distribution of the

dedications to Asklepios
to Kalliades

any criterion

— 3%

lines to Phileas, 4

in the late fall of 262 at the earliest, that
it

Athens

came
ochos

into the

hands of Antigonos,
at the

is

apparenl thai the surren-

der of the city took place
II

time that the young king AntiL'Hii

came
_<>1
I,

to the
48

throne of the Seleucids (between July

and July

and declared war upon Ptolemy Philadelphus 49 and chief hope of the Athenians.
last

—the champion
city could
i"

This new en-

tanglemenl destroyed the

prospect of Egyptian aid, and the

do nothing but yield.
sister.

The marriage

of Antigonos' heir
alli-

Antiochos'

Stratonike, was the consummation of the
its

ance which cost Athens
siege

independence.

The war.

in

which the

and

fall

of Athens formed bu1
in

an episode, continued for

some time after 262, and resulted
of Philadelphus. 50

the downfall of the sea-power

Antigonos, we observe,

is

said Ta?

apxas

[avrnpr)<j6^\ai kcu

wav

ev\y\ /3ov\ev[eiv'.]
I

[e<£]
I,

elaOat.

hogenes Laertius: VI

10- 12.

"Diogenes Laertius: Vil,
pp. lis. 344.
I

15; cf.

Wilamowitz

:

Antigonos

v.

Karystos,
pp.

(i

1265

f.)
|;i
.

[Lehmani Haupl (Berl Phil. Woch. 1906, 836, II. 36 tr. dates the fall of Athens in the spring or summer of 261.]
It
\'
:

Tin

II

mis,

of S< leucilS

I.

p.

His.

B
'

a:
ii
:

Grii ch. Gesch.

[Ill,
2,

p.

615.
ff.

BelO(

[III, p. 618; III

pp. 428

Vol.1]

Ferguson.

-Tin

Priests of Asklepios.

L55

6.

The archons between 26]
a

and 230 29 form

a

group by

themselves and deserve

special study.

Leaving ou1 of accounl
1,

51 Sosistratos and Philoneos,

who belong before 262
ff.,

and

Philos1

tratos,

52 has, Antimachos, and Phanostratos, wh.Hn Kolbe

be-

lieve rightly, assigned to

209 8

there remain for the thirty-one
:

years involved twenty-two archon-names
oil. ios.

Kle

achos, Diogeiton,

53 Thersilochos, ChariLysiades, Kallimedes, Glaukippos,

kles. Lysias,

biades, Hagnias, Lykeas, Pheidostratos, Philippides,

Kimon, Ekphantos, Lysanias, Diomedon, Jason, AlciTheophemos,
bios? and, as
first

Thymochares
des.

a

possible twenty-third. Aristeiin the table to a

Of

these the

fourteen are assigned
a

definite year.

These assignments require
h;is

word
'2'-V2

of justification.
1.

Diomedon
a

heen dated by Kirchner in

for ohvioiis
5

and adequate reasons.

That leaves only 244
i.e.,

::

and 256

open

in

secretary from Leontis,

to Thersilochos
a clear
all

and Kleomachos.
a

Kallimedes precedes Thersilochos by
sion between
L'oti

year; hence

deci11

5

and 244
first
it

'3

involves

three archons.

is

hard
effort

to
54

make.
to

Bu1

should he remarked thai

Kolbe's

carry Kallimedes and Thersilochos hack to 290 89 and
ill-advised.

288
a

7

was most

The decrees of these archons contain
is

formula of allegiance to Macedon which
in

found only between
after

276/5 and 230/21) B.C., 58 and
exhibits the

the second place one of them
not appealtill

form

yivofiai,

which does

2»>1

0.

and then only
a

in unofficial

documents. 56

Vivofiai

demands

as late

year as possible for Thersilochos.

The contents

of the docu-

ments of Thersilochos" year demand that Macedon be on friendly terms with hoth the Boeotian Leamie and Athens: for each of
these accepts arbitrators for
51

a

dispute from the .Macedonian de-

Philoneos cannot be located in _(!."> 4: for AwrticXijs ^i'ttoX^ttios, priest of Asklepios in 265 4 and d/coj'TioTTys in Philoneos' archonship, cannot have held these two offices in tie' same year. An additional ami conclusive argument may now lie cit. pp. 76 il'. added to these given by Kolbe. There is mi ether place in the third century B.C. for the three archons whether Autimachos was the first or middle one of lie doubtless occupied the middle place. the three, The reason for dating Glaukippos in 245 4 rather than in 257 6 or The prosopograph2 is the similarity of content in II 305 and II 325. \. p. 34, also favor 245 t. ical data given in Cornell Si
.

•_'.",;:

<>

i

' Ath. Milt.

XXX,
1 1

L905, pp.
-2

!»8 ff.
f.
t'..

See
56

I.u.-kki.i-:

pp. 684

See Meistekhans: Grammatil der attiscJten Inschriften* pp. 177

n.

147v

156

Univi rsity of California Publications.

[Class. Phil.

pendency, Lamia. The condition thus imposed was nol fulfilled between the revolt of Alexander, Krateros' son. in c. 252 and the There is no unlikekritos a1 Chaeronea in 2 15." defeat of Abj
lihood thai
l'II
3.
it

was mel

in

256 5:

it

was admirably

fulfilled

in

After the battle of Chaeronea the Boeotian League en-

tered into sympolity with Aetolia, and thus

came over
During

to the

Mace-

donian camp. 08
been
in

It

58 and had had sympathized with Alexander,

alliance with

Achaea up

to

245.

this

time

its

relations with

Athens were undoubtedly strained, and upon the
in

change of policy

245

it

is

natural to find disputes referred to

a

Macedonian dependency
donian tutelage, 60 and
extanl

for arbitral ion.

Between 262 and 256 Athens was very completely under Maceis less natural to find a group of decrees
it

from

this period

than from 246

ff.

Hence

for these va-

rious reasons 246 5

and 244

3 should he assigned to

Kallimedes

and Thersilochos, and 256

5 to

EQeomachos.

In a decree of Kallimedes' year (I
tos 6 TraT7][p

G
.

II 306)
.]

we read

arp'ja-

] fiaai\e'(0<; Ar]fx[7]Tpiov

"As

far as one

may

judge,

i1

is

here said thai the father of the person eulogized in

the decree did some services to Athens during the reign of De-

metrius Poliorcetes."

This interpretation,

made
p.

in

Cornell Stu.-is

dies X. p. 31, Kolhe (Ath. Mill..
;j(

XXX.

1!><>5.

100) regards
it

kiinsti

It

and

a

Verlegenheitsauskunft.

To me

seems most

natural as well as correct.
I (
;

Did Kolbe forgel such documents as
father of
devoted.'

1

1

331,

in

which the earlier pari of the decree enumerates the

services rendered to Athens by the (grandfather and'

the benefactor to

whom
is

the body of the psephisma
in

is

Ahout forty years had elapsed
to be a king.

246 5

sine,.

Demel

rius

had

That

also natural.

well have been a hoy of 10 or L5 in 290 B.C. while his father
still

The benefactor of 246 could was
Kallimedes and

in
7.

the prime of

life.

The appearance
a

in

the archonships of

Glaukippos of
:..

cull

of Zeus Soter in Athens as well as in the
Ill
1.

.en

:

Gru

tli.

Gesch.

p.
-

642.

,l.

Staaten
p.

II.

p.

250.

ii

:

[III, 639;
Phili ehoros.

Niese:

I

I.

249;
pp.

ef.

however Beli
f.

p.

138.

[das:

J

I

I

2,

135

Vol.1]

Ferguson.

The Priests of Asklepios.
Eor

1"><

Peiraieus should be
stones, nn

noted™

the

finding
i

in

Athens of the
Kal-

which were written

lull

305
62
I,

Glaukippos), 325

Limedes, according to Kolbe),326 (same time as 325 ,616
of third century
clearly enough.
i,

middle

and 1387
!i

I

dateless
thai

demonstrates

this poinl

is

significanl

all

these stones belong to
of worship
thai
is

the period 262 229.

The same duplication
still

demL902

onstrable

in

this

period Eor

another

cull

of Bendis.
in

an interesting inscription published by Wilhelm

we

Learn

thai

in

Polystratos'

archonship

(with which

Lykeas

from an unpublished documenl must be closely associated branch cull of this goddess had recently been established among
the Thraeians in the city,

and

thai
a

by formal resolution the old
friendly attitude toward
ii

organization agreed to assume

kol vvv ol

?)i[pi]~\iAevoi ev tool

darei KaraaKeudaaadaL tepov oiovrai

SeivoUeicos 8uucel[<r0]aiirpbs aWrfXovs.

Provision

is

made

Eor co-

operation between the two societies
the

in

the Tro^trr) Erom Athens to
a

Peiraieus which

formed

so characteristic
in

Eeature of the

The iirLfiekwrai vide sponges, basins, and wreaths
Bendis worship.

the Peiraieus were to pro-

Eor the
in

members

of both clubs

upon the arrival of the procession
Furthermore,
nut
a
it

the harbor-town.

is

to be

observed thai between 260 59 and 229

single person
in

65 or Phaleron apErom either the Peiraieus

pears

any capacity whatsoever
is

in

the Athenian documents.

One
the

tempted

to believe that the Peiraieus
rest

and

its

environs

were taken away Erom the
military

of Athens in 256

and pu1 under

governmenl of the Athenian strategos, "tyrant,"

Herakleitos, Asklepiades' son. of

Athmonon. 66

Bu1 on close ex-

amination this view

is

proved untenable; Eor the Athenian archon

was eponymos

in

the Peiraieus in Polystratos' year,

and the Thra-

eians there resident claim certain exclusive rights on the strength
adt .III" n istence of a separate cull in Athens. Judeich p. 524, stoutly maintains it.

ira

AlU rthum

II.

p.

L45, denies

I

Mommsen:
:

Dit

FesU dei Stac
Ithen
p. 302,

TopOi
(c.
IV.

with .Men:!'

-

enience of "

I

G
\
.

ill
|

L67

134 A..D.

i

is

disputed.

p.

127

\\ ii.i.Kl.M

:

loe.

[...
certain.

n«p]aie?

in

1

<;

II

330

Kimon

II

237 6)
kolI

is

quite un-

'" KaOeffTT)Kws vwo rod (ia<Ti\tw$ (TTpar^bs (iri rod lltipaUm 591 b. raTTO.uevwv nera rod llttpattus. I G
1 1
•"•

tQiv

&\\uv

tCjv

/

nivi rsity of

California Publications.

[Class. Phil.

of

the

Athenian
in

Laws. 67

Moreover,
68

in

239
in

8

(Charikles)
I

the
"
1

ekklesia me1
:;;:;,.

the Peiraieus,

and again

230 29?

G

1

the senate was convened there.
still.

The
is

Peiraieus was thus

more than an ordinary deme

The
greatly

solution of these various problems

undoubtedly

this:

thai the citizen population in the Peiraieus
in

had diminished very
political
Los1

creed than thai dominanl
stant touch

numbers, and had perhaps accepted another It moreover had in the city.
with the city
It

con-

of the long walls.

through the destruction or delapidation was still possible for processions to go from

the one town to the other,

and

for the

populace or the senate to

proceed to the harbor when
Bu1
all

local
in

business

made

thai expedient.

this could be

done only

time of peace, and there was no

longer the unbroken intercourse between the two places which daily wormade" it possible for men resident in Athens to render ship to deities resident only in the Peiraieus.
8.
it

is

clear that in
is

Since the ekklesia met in the Peiraieus in Charikles' year 239/8 the war with Aratos, which in Plutarch's
described for us as a series of disconnected incidents, A similar state of peace is presupposed for

narrative

had not yet begun.
09 of I the time

G

II 5 373c,

i.e.,

for Skirophorion of

229—after

after a good the withdrawal of the Macedonian garrison, and That understanding had been reached with the Achaean League.

Ptolemais did not yet exist
,w certain thai this tribe

is

no objection to this date: for

it

is

was created in the course of 225 4 or m at the bein -2-24 3 or in 223 2: for while it was not in existence it already received officers under ginning of Xiketes" archonship.
Menekrates.

By
Tor

far the most likely year in this interval

is

2:24,

ad

so

much

the

reason
p.

urged by

Kirchner and Zhebelev

,,,,//.

g e l An:.. 1900,

450), that
a

the

archon-eponymos

\'^v

224 3 was taken from Aphidna.
of the mention of

deme
in

of Ptolemais, bu1 because

King Ptolemy

connection with the gymna{'Ecfy.'Apxis

siarch for 224/3.
p
1:
;

Qnfortunately the document

1897,

i

s

badly damaged, bu1 the conjecture
V. L902, pp. L27
IV.

obvious that we

«>

Oesterr. JahreshefU

Vv-

1901,

]>.

52.
3,

Tl.r possibilities are 254

242

1.

and 230

2!

vol.1]

Ferguson.

The Priests of AsUepios.

L59

have

i"

do with the donation of the gymnasi

named from

its

founder the Ptolemaion
: nistischer Z<il. "

du

erstl

grosst

Baustiftung aus

Iwlle-

The establishmenl

of the tribe

was Alliens'
Ptolemy
in

way

of rendering thanks for the gift.

The

interesl of

Athens was doubtless due

in pari a1 Leasl to

the good understand-

The ing reached by Antigonos Doson and the Achaean League. same evenl forced Alliens to secure the good will of Ptolemy. 73 The formula and usages which prove Koll>e*s Location of Kallimedes and Thersilochos in 290 S!) and 2SS 7 to he wrong, prove
with equa] cogency his dating of Lysias and Kimon
II to

be right.

The

72 has the notice of sacrifices chief inscription of these years

offered for the

Macedonian rulers

—King Demetrios II and Queen
yiuofiai.
I

[Phthia]

an<l also the late

form

Now
in

that there
it is

is

ab-

solutely no
that

room for Lysias before Kiinon

292/1,

certain

this pair belongs in

238/7 and 237/6.

As already

pointed

out, their

immediate successors were Ekphantos and Lysanias. 73

prior to 239 to capture the

The attempts which Aratos and the Achaean League had made Peiraieus were renewed upon the
a

death of Antigonos Gonatas, and

had not yet come

to

an end in 236/5. 74

war broke out in 238/7 which The Athenians are cenrejoicing over the

sured by Plutarch for indecently

reported

death of their distinguished
7.

adversary,

and indeed
in

Athenian

troops joined the Macedonian garrisons
try.

protecting the coun-

The struggle was one in which, according to Aratos' usual tactics, his enemies had more to fear from surprises night atJudeich Topographs von Athen p. 315, d. 27. Beloch Grii ch. Gesch. Ill 2, p. 61. "I G II 614b; ef. Kolbe: Festschrift/. Otto Hirschfelti
:

ro
71

:

r,

p. 314.

T:!

See above
a

p.

140.

M For
!

description of this struggle sec
pp. 315
1

Kolbe

in

Festschrift fur Otto

Hirschfeld
1

if.

G 5 614b is the only documenl relating to garrisons in Eleusis in which a detachment of foreign mercenaries appears. The others belong between 318 7 and 276 5? and after 229. The nationality of the merei of 238 IV. is worth noticing. So far as the extant names permil a .in it seems thai there were no Celts among them. They are mainly Greeks. One is designated 'AxcuAs a deserter or traitor. One of the soldiers' decrees ('E#. 'Apx. L896, p. 33) found at Eleusis begins as follows: 'EireiSI) Avrl[yovo]s 6 [/3a]<ri\et>s d<f>iKo/xepot. Unfortunately nothing further is extant. The orator, however, was Afietvoickijs Ta\i\\or KvSadrivateOs. name appears in The same G II 1024 I. 9 a list which belongs before 307. The probabilities, given by the name -connections, are
'
'
I

160

University of California Publications.

[Class. Phil.

tacks,
to a

ambuscades,

etc.,

than from drawn battles,

H seldom came
one occaIn

regular campaign, bu1 the destruction of the crops had cona1

stantly i" be expected by the Athenians, and on
sion Aratos
it

Leasl

marched even

into the

suburbs of Athens.

236 5

is

said of the general
07r&>[9€«
rrj<;

eV
0i

'EXevalvos ,
crt']

Aristophanes, eTre/ieX^
'

(\

tcai

x™P a<i

TOt A167

ciacf)a\eia<;
1.

elaevex^oyaiv.

The

situation had oo1 essentially changed in 232
a

On

the

Lasl

of Elaphebolion of this year

subscription was started to pro"

vide

ili«'

Tafitas tg>v o-TpQTLooTiKMv willi t'uinls.
is

The purpose

of

the contribution

stated to be [Zva
ol
e/c

Kara rov K~\aTakonrov x?° vov
yr]^.

rov iptavTOV <rvvic[ofu*daxriv

K]apirol per aacpaXelas.
in

The

inference

to

be

made

is

thai

the

earlier

pari

of

any rate acthe year the harvesting had been molested or at companied by danger. A study of the provenience of the subscribers will,

them came from Erchia,

show where the exposed crops lay: 9 of 5 from Paiania, 5 from Sphettos, 3 from Phlya and Halai, 3 from Phlya, and 3 from Eephisia. Of these eastern side of the Athenian plain, placed a1 Kephisia Lay in the
1

believe,

the entrance of the valley

The others were
tuted
i,,

in

the hearl of the Mesogeia

which led into the Mesogeia proper. itself, and consti-

its

most importanl villages.
portion of the
list

The only other deme, which
furnished more than two sub-

the extanl

scribers,
nidai,
a

was Phyle.

Prom

the six city

demes

Melite,

Skambo-

Kerameikos, Kollytos, Kydathenaion, Kolonos, there came There were none for Phaleron or total of only four <>v five.

Peiraieus,

and only one each

fr

Eleusis and Sunion.

It

is

true

of the men in the eata that it belongs before 320, and a grandson of one while the ami ol logue, HaipiivSrts Bparuvldov AWaXt&p, was ephebe in 283/2 'Kircyivov* Ku5a0i?Mu«Js, eame according to the grandson of another, "Eiriyirr)s that the Kirchner in c. 268 B.C. There is. therefore, no unlikelihood and Ameinokles of II L024 and of'E^.'Apx. 1896,. p. 33 are grandfather And a1 whal latter documenl is 276 5 266 ... grandson -if the time of the Athenian other time could a king Antigonoa come in contact with and 240 39 mercenaries would have Between 262 Ai asociated with Athenians and Eleusinians in the decree. was king between 306 and 301, bul was never uear Athens in terms with Athens Antigonos Doson was no1 on Buch thai interval. On the other hand eisil ( W ha1 else does i<puc6fuvos mean?) possible. visitor o1 Athens our tradition represents Antigonos Gonatas as a frequenl of Zeno), i.e., 266 5 B.C. (outbreak ot (death between 276 5 and 261
;

I

:

1

monidean War
I

|.

HI 5
..

61 n.

11.

66f.

"]

ll

334.

vol. i]

Ferguson.— Tht
list

Priests of Asklepios.

161

thai the
qo1

as

we have

it

is

fragmentary, bu1 the demes

in

it

are

arranged on any principle, nor are the individuals from one
listed together.
a

deme
thai

We

have no reason, therefore,

to

suppose

differenl

proportional distribution would resull

from an

analysis of the entire catalogue.

The explanation of
which could
for the must
itself,
still

these facts

would seem
Those

to be that the
in

crops

be preserved and harvested
in

232

1

B.C. lay

pari

the Mesogeia.

in

the Athenian plain

part.

we may suppose, were already destroyed in whole or in And whal was true in 232 1 was. we may safely assume.
greal

true

in a

many

instances

in

tl

ourse of the third century.
to the

While the

rest of Attica

was exposed

ravages of war. from

7s from the soldiers of Alexander. Krapirates on the coast places.

teros' son. in

e.

252

if.,

and of Aratos, and the Achaean League
Antigonos Gonatas' reign and the whole

during the

latter pari of
s.

of Demetrios II

by Pentelikon and Ilyniettos and by the flanking position which Athens and the
the Mesogeia proper, protected

Peiraieus assumed to an invader of the trans-Hymettos region,

enjoyed practical immunity from devastation, and came in consequence to be politically the most important section of Attica.

The kolvov twv Meaojeicov meets us
first

in the inscriptions for the
last

time

in

one of Olbios' archonship and for the

time

in a

contemporary document.
Olbios must necessarily occupy the year 251/0; for 239/8. the

only other place between 261 and 229 open to an archon whose
secretary was from Aiantis, must reason
for this
is

lie

assigned to Charikles.

The

as

follows:

Aristokreon, the nephew of the

philosopher Chrysippos,

who

is

commended

for various services to

Athens
to 229. 79

in

Charikles' year, cannot possibly have been old enough

for such a distinction in 251/0, yet the decree

was passed prior
beyond
be-

The constitution
our ken.
It

of the kolvov

lies

for the mosl

pari

undoubtedly embraced men from denies which
in
'"
>

longed to different tribes and trittyes

the EQeisthenian system.
, 1
'

Curiously enough the chief
-I
:
'

officer

(

apx (0V

'

'"'

""'.v

tw " OCCa-

<;

[15 591b.
"'•">.

Wh.kf.i.m: 'E#. 'Apx- 1901, pp. 52,

'

162

University of California, Publications.

(
!

|v

-•

Fu-

sions

i>ii

which he
city

is

suburban or

demes.

known came Erom Bate and Kydathenaion, The patron deity of the kolvov was
is

Herakles, and, since two of the three stones which have inscriptions relating to the association were Eound in Diomeia,
thai the
ii

clear

temple in which the decrees of the kolvov are said
up,

to

have

been

sel

was the Eamous one of Berakles was primarily

in

Kynosarges.
<>f

Although the kolvov tmv MeaoyeLwv,
cities",
(

like thai

the "four
its

TerpaTroXis

)

a

religious federation,

creation or revival in about 250 cannol have lacked
significance.

some

political

The onion of the demesmen

ii

involved must have

given them increased influence in the ekklesia.
prosperity came into relief

Their materia]

now

that the shipping

and trade of

the Peiraieus and Athens had diminished, and the weakening of

Athens' predominance
aratisl tendencies

in

Attica must have strengthened the sepIntent
in a

always

mountainous country.

The

Mesogeia was exposed to spoliation because Athens was of necessity involved in all of

Macedon's wars.

It

could not escape them
It got

by making the state join Macedon's enemies.

no help from
lie

Athens' walls, nor did the recovery of sea-power
range of
its

within the

ambition.

And now
doubt.

for the

first

time since the days

of Kleisthelies the
this there
in

.Meso'_;ei;i

determined the policy of Athens.

Of

can he

little

The three most

influential families

the state before and after 229 came.

Dromeas-Diokles from

Erchia,

Mikion-Eurykleides from Kephisia, Zenon-Asklepiades
all

from
geia.

Phyle

from denies located

in

the Kleist heiiian

.\t'-n-

These were the men who foiled Aratos of his hope of bringAlliens into the

ing

Achaean League, and carried through the

policy of strict

neutrality which gave the country respite from

wars and devastation for nearly thirty years.
9.

From Kimon 's year we

possess a
It

lisi
is

of ephebes. 80

It

contill

tained from twenty to thirty names.

the last of the kind
!'».('.

we reach the second sition of the names
loneos),
significanl
kles).
I

half of the second century
in

The
:!:'.^

dispo-

the

list

is

like that

of
all

Mill
I

(Phiin
a

and 324 (Polyeuktos, 275
point
t

4),

ami

three differ

Erom the catalogue of 283
list

2

G

II

316, Mene-

n
II

lie

earlier

the

deme

is

used to segregate the names

M

IG

330.

Vol. l]

Ferguson.- -Tin Priests of Asklepios.

L63

into groups; in the Later ones the tribe alone
tion.

performs

this funcin

Since the

number of names
them

is

aboul equally small

each

— explicable only on
ii-~>

case; the classification of

in aboul

L50 de
it

-groups

is

absurd

the supposition thai

is

the survival of an

idea, sensible in the nol
1 >>
»

very distanl past.
is

In
it

305/4

(]

G

1I 5

»

the

same
EI 5,

sjjstem

employed, and

meets as again
i1

in

334 3 (I G
to a

563b), bu1 in eaeh of these instances

is

applied

much
In
.'534

larger

number

oi names.

3 the ephebe systesq described

by Aristotle 83 was

in

existence.

All the

young men

in their eighteenth

and nineteenth
of

years were obliged to serve as ephebes.
turity, they
citizens,

Upon
till

attaining legal malist

were entered by the demarches

in the official

and became thereby attached

their sixtieth year for

ephebe. military, and judicial service to the archon-eponymo.s for
the year of their registration.

They were put
the most
all

as ephebes

under

the supervision of state

officials,

important of

whom

were one Icosmetes, chosen from
rdstai,

the citizens, and ten sqphro-

taken from thirty reputable and qualified citizens nomi-

nated by the tribes.

The
eirl

list

for 334/3 contained ol e [<pi]8oi]
1

(t>}<?

Ke/c/307r/So?) oi

KT7]aiK\eo(u)<; ap^ovTOS tvypafyevres*

In

it

there were from

forty-four to fifty

names. 83

There were therefore aboul
as

500

ephebes enrolled under the archon Ktesikles, and

many more

are to be added for the archon of the preceding year-, so thai the

young men of Athens in their eighteenth and nineteenth years numbered about 1,000. 84 Since there were only 33 in 283 2,
it

is

clear that the compulsory service has already
too.

become voluntary.
most

The term.
change
to the

was seemingly reduced

to

one year, and the sophroa
Il

nistai exist

no longer.

We

have

to

do with

importanl

in the life

and institutions of Attica.

was equivalent
a

abandonment

of universal conscription as
in

national sys-

tem of defense, and
51

the case of Athens that

meant the con-

Ath. Pol.

41'.

~-

The

technical term for registration with the demarchs; cf. Aristotle:
<;

Inc. cit.

of «Fotjcart: B.C. E., XIII, p. 263, thinks thai col. more than 22 names. Col. II had 22. 'So Girard: article "E^t//3oi in Daremberg >> Saglio.
I

I

II

5

563b had

164

University of California Publications.

[Class. Phil.

IVssinn

of the city's inability

to

proted

herself with

her

own

resources.
to train her
it

When

in

an age of war Athens renounced the efforl
in

young men

the highly technical profession of arms,
as
a

was over with her days

free-acting political agent.

The time of
vestigating.
Its

this confession of

impotence

is

surely worth in-

determination rests largely with the ephebe-lisl

[GII 5
for

>

251b.

This

is like

that of 334 3 in giving us the ephebes
is a

one year only.

What we have
iirl

catalogue of

tow

icfryjfiovs

tou]? ivy[pa<f)evTa<;

Kopotfiov ap^ovro^.
if

Moreover two sons

of Ergokles from the

same deme,

the restorations are correct,
if

which

is

doubtful, and two sons of Kephisokles of Kikynna,
in
it.

the restorations again are correct, appear
in
is

Unless these arc
it

both cases twins, or mere coincidences, or false restorations,

imperative for us to assume that
in 283, 2
IV..

in

305/4 the term of

office

was

already reduced, as

to one year.

The number of ephebes must next be ascertained.
is

The stone
is

so

badly damaged thai an approximation

is all

that

possible.

Eighteen names are extant in part or in whole from the tribe
Erechtheis and fourteen from the tribe Akamantis.
are lost?
It is

How many
is

known

88 and denies that the population of the tribes

quite evenly represented in the surviving names, and that it is the big denies that have the largesl representation in Kirchner's

Prosopographia Attica.

A comparison

of the relative strength of

the demes in the P. A. and in the prytany-lists will show this to

be the case.

Enonymon had

eleven ephebes

in

305 4:

it

has 208

names out of
ibis tribe.
a total

a total of 929 for the tribe Erechtheis in the P. A.
total of slightly

That suggests a

over

fifty

for the ephebe-list of
1

In the case of Akamantis. Thorikos had
-\

ephebe

to

representation of 129: Kerameikos had

to

14:!-.

Kephale

had over 5
11

to 44S.

The four have over to 120; and Kikynna had 2 to 56. which yields about 26 for the whole tribe with !>T9. A
list

comparison of the ephebe-lisl with the
probably yield
Antigonis.
It

of the prytanies will

a

safer result.

Part of I'aiania was assigned to
part, 88 which,
in
a

was undoubtedly the smaller
but one
'.

prytany of
-.
I:
.

fifty, gol

member
L903, p
p.

in

I

(i

11

871 and 865, while

Koeb n

:

Gott. g(

.!«:•

vii s;

Corru " Studies

VIM.

L2.

Vol.1]

Ferguson.

The Priests of Asklepios.
Pandionis, had regularly

165

the other part, which remained in

12.

We

do

not
in

know how many Waiaviels
305
4,

from

Pandionis

were
This

ephebes

bu1

Erom Antigonis there eame three.

suggests thai the ephebe-lisl was
list.

much

larger than the prytany
it

So. too,
a1

Phegus had one ephebe
in
I

senator

all

(i

Il~>

871b.
in
I

in 305 1. whereas had qo There are four names and one
II 5

fragmenl of

a

name

extant

G

251b

Erg. k

from an un-

known deme.
rhinus;
for

Since there were so many, the deme can have heen
It

only Kydathenaion, Oe, or Myrrhinus.
three87
of the

was certainly Myr-

names are Eound among the 'SXvppii'ovcnoL in Kirchner's Prosopographia, and none among Hence there were at those from cither of the other two denies.
Eour
least
live

ephebes From Myrrhinus

in

305

4.

This deme was rep-

resented by six

members
in

in

the senate in the fourth century.

All this evidence goes to
in

show
in

that there were as

each tribal

list

305/4 as
five

334/3, and that the total
six

many names number
in

must have been between

and

hundred.

At the time of the

census of Demetrius of Phaleron there were 21,000 citizens

Athens.

This, on the ratio of the Belgian census, 88 calls for 1,176
in their

young men
tains
that
a list

eighteenth and nineteenth years, or aboul 588
It

for either of these ages.

is

thus likely that
age,
S11

1
il

<i
is

II 5 251b conto he ohserveel
in
1

of

young men of only one

and

this

document takes cognizance of the registration
eirl

the
IT'

denies (oi evypacfievTes

Kopoifiov ap^opro<;)

,

just as

(i

563b does, whereas
S7

in
</'.

283 2 the younu
I'.A.
sic;',;
(Z.
f,, r

men
a.

are referred to as
I'.A.

For

Kallistlicncs

Atlien.i.lnrns

276;

;

[Sundwall Aristokrates I'.A. L921. does not belong to Myrrhinus.]
ss

c.

below

89)

shows thai

tl

Fraxcotte: L 'Industrii 'dans
philosophic
il*

la Grdc<

culte d(
89

I'wniversite di

ancienm (BibliothSqiu Liige, Fasc, VII, L900, p.

ii<
1

lafa-

i

e]h of Demetrias appear side by side the ephebes Among 'Ep7o]\\e'oi's. 'Ep~\yoK\(ovs ami The lasl name may be restored in many Among the Kacvvveis ways, e.g., #iXo]k\&ws, Oe/jaaTo]K\eovs, 'IepoJ^Xeoi's, etc Here too there are ^]dfiax oi Kij0«roK\[A>vs] and [Etf]j8oi/Xos Kr)<pi<ro[K\tov;]
thr
. .

|

.

one name, .g., Iv7;0t<ro[o6rou] through the assumption that we have to do with two pairs of brothers thai the restorations in the Corpus became current. Brothers are, of course, common in the same ephebe-lists when the service was for one year only and there was no compulsioi

is possible a number of different restorations of K770t(ro[0tDvTos] Iv7j0t(ro[5u)poi/] kt\. It is simply

its
.")

i'.-iII.

will

of

the

•; II or registration. [Some of the details of this treatmenl of have to be altered because of Sundwall's clever rearra s istotelis documenl (l>< instituiis reipvbliccH Athei
I I

aetatem commutatis. Acta Societatis fenniccH
era! conclusion is substantiated.]

XXXIII

(1907), bul

I

166

Univi rsity of California Publications.
£<$>!) l3evaavTti<;

[Class. Phil.

toi>9

eiri

Mert
1

o
in

/cXt'01/9

ap%ovro$.
it

If

the service

were voluntary
wli\

in

305

and

283 2

is

impossible to explain
the earlier year and
in
its

1,100 (a1 Leasl 5 600)
in

came forward

in

only 33

the latter; for the city was equally popular
in
;i

gov-

ernmenl and equally involved
years.
I

serious foreign

war

in

the two

conclude therefore thai the national ephehe system was
in
Mil.")

still

in

existence
in

4.

and

this result

finds substantial confirmaff.,

tion

thai
in

the sophronistai,

who

are Lacking in 283/2
in

are

found
(I
It

this

year

still.

The sophronistai appear
in

303 2 also

G

II 5

565b), so that the change had not occurred

a1 thai date.

therefore took place

the following twenty years.
in

Had

a

national ephebe system been

existence

when

Alliens regained

her independence in 289,

it

the democrats in the war-time which followed.

would never have been abolished by Nor is conceivit

able that

it

was abolished in 289
in that

itself.

On

the other hand,

if

done away with prior to 295/4, the democrats on recovering the
government
year would have been unable to restore
it

be-

cause of their relations to Demetrius Poliorcetes.

A

Macedonian

garrison in Museion and a restoration of universal conscription

do not harmonize.

The only occasion

suitable for this

momentous
its

change came
lished in
called
it

in

301 B.C.

In this year a government was estab-

Athens on

a moderately aristocratic basis

enemies

an oligarchy
all

— which had as

its

foreign policy the aban-

donment of

imperialistic

notions, and, without sacrifice of

independence, the maintenance of friendly, neutral relations with
all

the powers/'"

It

system voluntary.

was this government which made the ephebe The number of ephebes instantly fell to a

mere handful. Ten sophronistai for about three times as many The sophronistai were therefore discharges seemed absurd.
pensed with.

But the old habit of registering the ephebes under
It
it

deme-captions persisted.
of government in 276/5,

existed in 283

2,

but

upon the change
in
27.")

was also discarded, and

4

\Y.

the tribe-captions alone are used.

Had Kimon
employed

II

belonged

in

292

1.

the old system should have been
I

in

EG

II

330.
19 or

10.

(i

115 37ic
Gesch.

w

ii]

have to be dated
IV.;

in either
ibid.

250
L80

alt.

V, pp. L55

Eduard Meyer:

pp.

iv.

Vol.

i
|

Ferguson.

Th I'm sis

of Asklepios.

L67

249

8;

for the secretary's
Eilpeo-tSr}?
is

deme began with

'Et,

which cau be

re-

stored onlj as
Eiresidai,

or EtreaZo?.

For Akamantis, the tribe of
Eitea,

there

do place between 256 5 and 243 2

which
a

a1

this time

belonged

to

both Antiochis and Antigonis, has

place

in

250

sion of ihc

The decree was passed a1 the conclu 4!) or 249 8. war between Athens and Argos, friends of Antigonos

Gonatas, on the one side, and Alexander, his rebellious nephew.

on the other.

It

commends Aristomachos
in

of Argos for insisting

on including Athens
cessful rebel.
in

the pence he had purchased from the sucin l'4:!: "
:

Alexander was dead

he bad nol rebelled

256,

i.e.,

Museion.
iil.mi

when Antigonos withdrew Hence the dating above given.
was reached
in

his
It

garrison
is

from the

obvious thai the

of Alexander's success
i.e.,

a

short time before the

passing of the decree,

either 250 or 249.

Diogeiton has been assigned to 252/1 because 'AfcpoTifios Aicryjiov 'I/capiew,
in

who moved
earlier
is

the passing of I

G

II

Add. Nov. 352b
)

this

archonship,

was rapta?

(™v

o-TpaTuoTi/cwv!

in

255

4.

Twelve years

out of the question, Tor thai takes us back

of the Macedonian regime.
sible,
It

Twelve years

later in

240 39

is

pos-

but
is

much

less

probable.
1

evident that the archon-list
in

<i

II

859 was begun, as

Zhebelev and Kirchner claimed/'officers lor the first

the year 230 29 with the

year of Athenian independence.
in

Three boys who were 18
7rpa)T?;9
r)XiKia<;

Phaidrias' archonship were

tT;?

in the

year of Anthesterios. 93
a

Anthesterios was
SjXtKt'a*;

archon
in

in
ii

160 59 or 158/7-156/5; for

boyrfj's Bevrepas
all

161

won

the boxing-match open to boys of

ages

in

Anthe-

sterios' archonship.

159 8
in

is

excluded, because comic exhibitions,

which were not given

and

in

the year which preceded thai of Anthesterios. are
in

two successive years, were given in 161/0 The possi154 3 and

bilities

Phaidrias

in

Anthesterios
in

in

L60 59, or
L5 has

Phaidrias
little
a

153 2 and Anthesterios
in
;i

158

7.
1!>.

A

boy of

chance

boxing match with others of

A boy
is

of IT

is

much more

likely winner.

Hence the

lasl

possibility

tobepre-

'" Corinth was taken by Aratos iii 243 from Antigonos, no1 from Alexander. [Kirchner (Berl. Phil. Woch. L906, p. '.".mi gives ground fi ing [ G II 5 371c to 250 49.]
i

.

gel.

Am.

1900, p. 448.
f.

For the references sec Cornell Studies X. pp. 67

168
ferred.

University of California Publications.

[Class.Phil.

It

is

assumed with Rangabe {Ant. Hell, [1,678
?}XiKia^
;iii<1

IT.

thai

TralSes Trfi irpioTiys
i)\iKi'a<i
1!)

were L2 and
t/3j't?/?

13,

TraiSes

T/79

ScvTepas
IS.

14

and

15,

Trai&es ttjs

i)\ikicls Hi,

17,

and

years old.
Tlic result of

Wilhelm's combination" of
is

F

c

;

II 5 :>*:„.

496c,

and

II

Add.

4:»:ili

thai

tint

name
in

of the priesl
]

for

Timarehos'

year (138/7 B.C.) was

X</co[>
fact,

<£\vew.
is

No

Attic

names begin with
as

Ni/eo[t>,

ami,

what
other
the

read by Koehler

T

can be equally well K.

Then the restoration [ZWX09] Nt-

kok[pcitov] <\>\vev$
<P\veti
in

may

be

made

No

name among

the

thf

Prosopogrophia

fulfills

conditions.

Zoilos'

father "SiKOKpdrryt
II

1224).
in

Serapis
In
I
< ;

H5

<\?\vevi was an ephebe in 172 1(1 <t Zon\o? ZatXov <P\vevi was priest of and of ayvrjs 'A^/ooSiV?/? in 105 4:'. 117 6 P.A. 625] One 373c (230 29?) ZwtXo? ZmtXov 3>\vet/9 is found.

ZmXov

His

cousin

(

)

item

in

the catalogue of dedications to Asklepios given in
<>)

I

G

II

403 (Thrasyphon 221
virep tov iraiBiov.

is

as follows:

tvttov ov avedrjicev

ZW'Xo?
priest

The two

arc probably the

same

The

of Asklepios in 138 7 will be the great-grand-son of the

donorof

221

i)

B.C.

I

Lave to thank Johannes Sundwall of the Royal Alexander
in

University

Helsingfors for an admirable monograph, entitled
Beitragi

Epigraphischi
im
Zeitalter
1

iwr sozial-politischen GeschichU
(Leipzig:
in

Minus
1906
.

des

Demosthenes

Kreysing,
the press.

which

received while

my
I

study was
pp. 47
f.
I

Sumlwall

has also

made
to
in

the discovery
distribute the
section
9

thai the official

employed
tribes,
priests.

priestship of

Asklepios
tabulates

^hIt was among the
the

and

(pp.
a

75

ff.)

he

extant

The matter had only
his failure to his

subsidiary interest for him, how-

ever,

and

examine
list

Mill
the

836 with sufficient thorpari

oughness has made
cardinal error was
Berl. phil.

for

most

incorrect.

His

in

not distinguishing

between the priests and

AJso published

Woch., L902, pp. L908 f. in Beitr. mi. Gesch. as Beihefi

IV.

vol..

1

Ferguson.

The Priests of Asklepios.

L69

ex-priests of Asklepios found in this document.

And

ye1

they
offer-

are marked off with

all

reasonable precision.

The annual

ings to the temple are invariably catalogued under the headings

koI rd8e e0' iepew Ilpo/cXeovs
(1.

lleipaiea)*;

(1.22), ^tXeov

Wneaiov

36), etc., the priests

in office

foreach year being thus clearly deslike

ignated.
e.g.

The ex-priests simply make dedications
iepev$
Av<riic\rj$

other people,
etc.
is

atcdfaov,

IvrraXiJTTios

(1.

-2)

and

when

the officiating priest donates anything, this, loo.
it

recorded

by entering

regularly

;is

an item
is

in

the section to which bis

mime
which

gives the date.
is a

There

qo1 the least difficulty in deciding

priest

and which an

ex-priest,

and yet their confusion
Snndwall's
table.

vitiates the entire disposition of the priests in

Sundwall
II

(p. 70. n. 1) sui^ests that the

archon-name K[v- of
This
is

835,

1.

8,

he restored Buxenippos (305/5).

practically

impossible.
'

The

secretary
(I

tor

305/4 was

[
:

Jo?

Avkov

Wcoire/c^delp]
II is

G
,

II

Add. 252b; II- 252c)

that for the year
i

of
It

835

KXe(7[t;/'//9]
[

no other restoration of
]o>

line

being possible.

Avkov 'AXco-rreKfjOev is found only in inscriptions from which the archon-name is lost. l>nt \Lv%eviTnrov tills the lacuna in these exactly, and there is absolutely no place, except 305/4, in the entire neighborhood in which a secretary
true that

from Alopeke can he
miist
In
n. 3)
:

placed.

There can he no doubt that E[uMe]\t(T€V?) Sundwall says,
sicker.
(p. 78

be restored Eu[bulos].

regard to

AuoWa
ist

[<?

Die Ergdnzung
It
is

gam
1.

Von

M

ist

noch

<

int

Spur
a

iibrig.

true

thai a

faint scratch like the

lower Limb of

M

appears

in the

lacuna of
letters,

MM.

Bui the space certainly
96

calls for

more than three
Avaavia[<;

and on other -rounds
is

also the restoration

\lpo/3a~]\i(o-io<;)

much

preferable.

The juxtaposition of Nikomachos (1. 33) and Nikomachos tempted me to make the same resUaiaview of II 839, though is probably deceii fnl. toration as Sundwall has made p. 78, Q. 2
it
I

I,

It

would require Nikomachos

to

have been

priest
(1.

prior to 276

5.

The
makes

restoration

Ti/xokX^ E[tTeatos]

16)

Sundwall

also

So, too, be assigns Telesias of Phlya to 336 5

and Euni-

kides of Halai to 341/0.

To

Teisias

338/7

be likewise gives the
is

M See above,

pp.

L49

f.

[where the restoration of Sundwall

accepted].

L70

University of California Publications.

[Class. Phil.

demotikon Ke$«\f/tfer and
reasons given above
i

to
i

Pataikos bhal of 'EXewr/wos.
thai

The

p.

1

t5

show

Lysitheos of Trikorynthos
«

was

not priesl in 334/3.

Number V
:///

:;

of Klio
I

the

new and eonvenienl
after

title

of B< ftragre

alten GeschichU
printer.
I

also reached

my

study had gone
in

to the

am

pleased to find thai
(p.
a

Beloch

his article

Griechischt Aufgeboti
conclusion, though by

352) arrives

a1

approximately the same

slightly different method, as to the

ber of ephebes listed in I

G

II 5 251b.

numThe comparison which
I

have instituted between this ephebe-lisl and the prytany-lists can

now

be carried further by the aid of Sundwall's tables [op.
ft'.).

cit.,

pp. 86
(p. 89)

It is

perhaps worth noting that this same scholar

has proved the correctness of Bates' conjecture (Cornell
p.

Studies VIII,

12) that the pari of Paiania transferred to An-

tigonis (see above p. 164)

was the smaller one of the two.
(the distinguished historian, C. F. Lehits

C. F.

Lehmann-Haupl
and

inanii; the Beitrage
al

founder being,

it

seems, rebaptized

the

same time)

in his well considered article
isctu n

Politik vor <h

m

Chri monid(

Kriege, which this

Zur attischen number of

Klio also contains (pp. 375
in

ff.),

has doubtless done a good service

showing that Athens
301
ff.

in

274/0 had the same foreign policy as
all

in

—the

establishment of friendly relations with
Its

the

greal powers of the time.
4,

embassy
its

to

Pyrrhus (Justin,

XXV.

4) probably sought respect for

neutrality.

And

in

lad the

city

had the friendship of Ptolemy and
at
it

his allies, the Spartans

and others; of Antigonos,
seemingly of Pyrrhus, Tor

this time

Ptolemy's friend; and

was not molested by him.
I

Lehmann-Haupt
and Antigonos don

s

explanation of the estrangemenl of

'tolemy

the designs of Arsinoe

upon the throne of Maceit

is

also plausible.

And

I

do oo1 think that

is

invalidated
in

by the fad that the Chremonidean War did
Philokrates cannol
ture of Athens

not heuin

268.

now
in

be ejected

fir

268/7, and since the cap-

came
it'

the

fall

of 262. five military seasons had
in

then elapsed,

the

war began

the
the

summer

of 266 (Peithia

demos).

Pausanias' remark thai
(inrl fMa/cporaTov)

Athenians resisted Eor
no more than
this.

very long time
is.

calls for

It

of course, none the

less possible, as

Lehmann-Haupl maintains.

voL.i|

Ferguson.

Tht

Priests of Asklepios.

1<1

thai the treaty

deferred result

made between Athens and Ptolemy in 266 was the while of an understanding aimed at in 274
alive.
to join the series of

Arsinoe was
|
|

s1 ill

have endeavored
ilogy in 288

Asklepios priests and the series
to
a

of prytany-secretaries

for

the earlj

third
0.
:

century before Christ
It

fixed

ehr

7.

262/1, and 221
is

may
thai

be granted cheerfully thai

no one of these joints
for the end of the

absolutely fasl

thai 263 2

Chremonidean War, and
to

are alike "i"' and 262 wffectio and reelection of
1

:i

magistrates are alike adequate
the priestly series;
thai

explain the reduplication of Ajatiochis

in

and 284/3 are both possible for [saios, and thai there is no necessary parallelism between the series of secretaries constructed by working backward from 221/0 and the series of priests during I have not sought to make a mathematical proof: the period 262/1-229/8.
288
7

believe that In this, however, an historical demonstration is sufficient. have succeeded, otherwise, moreover, we bave to do with a mosl astounding It must be an accident that upon the restoration of the series ef accidents.
I
I

official

tribe,

Erechtheis.

order of the priests' tribes in 307/6, the rotation began with the firsl It must be an accident that both priests' and secretaries"

It must be an accidenl thai the official order tribes locate [saios in 288 7. of the secretaries' tribes demands Antigonis, again the coryphaeus of the

sequence, in 261/0; that the Macedonian coins
in

make

their

first

appearance

year; that the end of 263 2 was chosen by the cataloguers of the Asklepios' dedications as the point at which to enter a lol of semi official offerings, made by priests during the preceding thirteen years; and

Athens

in thai

f that the legislative activity of Athens, and state dedications to the shrii the year in which, according to Eusebius, Asklepios began anew in 256/5


I

the Athenians regained their autonomy.
division between I

It

must be an accident

thai
oi

the

G

II 835

and

G

II 836 occurs in the archonship

and that no priests are mentioned in the latter half of this joint catalogue who cannot be located preferably after 276 5 (there is no place for one more), while a change of government suited to explain both the division and
the absence of earlier priests took place in the archonship of

Eubulos,
ff.)

in

276/5.
lieve in

Professor Kolbe (Deutsche Literaturseit. L907, pp. 932 I do not.] the possibility of such accidents.

may

be-

University of California Publications.

In

Appendix
IV century B.C.

I

List of Priests.

WfjKTTapxos

Ko$a>Ki8i)<;
[etc

II

1466, L468.
II

'ApxU:Bov
•E[X]9r6^s

Koi\] V

<i

147!).

II 1446.
II

\\vtfv8r)fxo<; 'E\evo-ivio<;

1651.
II

Kti)<tik\?)[s

'

A]jvovaio<i
II

1481;

III

144.

M.e\dva>iTO<; \oXapjev<i

1472.
II

M ew'cTT/jaTO?
NiKoSrj/jbO'i

'AyyeXrjOev
II

1447, 144S.

:;."il)

4!»

.'

1440.

Tt>a)v

II

147:;.
II

<l>i\ofc\r)>i "E-vireraicov

1475.

[V or

III

century B.C.
II 1491.

'()Xu/x7r/^09 Kv8a07]vaiev<i

Ill

centurj

B.C.
II 14! 16.

Atcr^paWSf?;?]
Ayfxayevi]^
II

1350 296 5?
II 1496.
8. II

Eu#v8?//io[<?] 'At'Ti/cXeou? e£ [Oiou]

Nucavfap
1ifi]vXo<;
<I>tA.iOs

<S>Xvevs

II

14!):. [iic

301 0, 289

NiKoarpdrov
II

«K] oi\i)<i

1500.

$a\77/>eu?

L505.

End

of century.
II

<\>opfj.[{o)]i>

'HBuXou

['EX~\evcrivio<i

1504.

End

of century.

II

century B.C.
'Aijixov

'A6r)vayopov MeXiTet/?
II

II

1204.
Pleistainos.

Xewinbi^ <£\vevy

840.

Archon

I

century B.C.
5 N [k] oarparo'i Acf)i8vaio [sr] Io^okX^ <\>iXo)tov Houinevs, yovw
'

i

1

1

1

1

1

8e AiovvaoBiopov Aeipac.

Suotov

Atk. Mitt. XXI,
i
I

p.

297,

100 B.C.

\\vppiv[ov(Ti(>s)

<i

II

Add. 477

c

Archon

K;il-.

Vol.1]

Ferguson.— Tht
II

Priests of Asklepios.

17:5

I

and

centuries A.I).
<\>\vev<
III

WyaOdirois

693.

Aivhi.n Peiso
III

c.

175 A.D.
C.
<»1

'Acr&J7r[o'S&)/)o?]
A/o'(/)a;'//9

KXeofievovs <]>\u(ev>)
III
III

102a.

A.I).

\\7ro\\u>viov 'A^i/wew

228, 228a, -2'.K 229a.

H

t

<>'[<£t\o?]

EuSogov 'KXeucra'/os
III

132 D.

<I>\a(ovio?)

729.
III

KjoXAirrew
c.

181

li.

Archoii Q. Trebellius Rufus

100 A.D.

UNIVERSITY OF CALIFORNIA PUBLICATIONS

CLASSICAL PHILOLOGY
Vol.
1,

No.

6. pp.

175-204

March 23, 1907

HORACE'S ALCAIC STKOPHE
BY

LEON JOSIAH RICHARDSON.

The

Alcaic' strophe as

employed by Horace involves the

fol-

lowing quantities

A
J3

'


---~ _w
el
I

A--~
die

Q
I

tescende caelo

age

tibia

regina longum

!alliope melos,

sen voce nunc mavis acuta, sen ndibus citharave Phoebi.
III. 4,
I

I.

./.

THE ELEVEN-SYLLABLE

AL< Ale.

Wha1 word-arrangements are possible in a line of eleven syllables and how many of them did Horace actually employ ? It
1

is

evident thai there ace two ways of arranging words in

a

line

of

two syllables
>.

i

namely, either monosyllabh
in a

monosyllabU

or
in

dissyllabh
a

four ways

line of three syllables, eighl

ways

line of four syllables,

and

so on.
in
a

In short,

we are

able to

make

out the total possible

ways

given line by means of the forin

mula -'

'

'a

beingthe Dumber of syllables
in a

the line

.

Thus

it

appears thai

line of eleven syllables 1,024 differenl

arrange-

ments are
1

possible.

Vet

among

his

634 examples of

.1

Horace

This meter is found in thirty-seven of Eorace's Odes, __ _ strophes or 1,268 lines. Ten of these Odes, containing 60 strophes, are in Book I; twelve, containing 86 strophes, are in Book II; eleven, containing lis strophes, are in Book 111: and four, containing 53 strophes, are in Book
IV.

17.;

University of California Publications.

[Clabs. Phu..

employed onlj
the 19 thai
1.

117 arrangements, confining himself generally to
:

follow

2.
3.

doctus Bagittas tendere Sericas quicumque terrae munere vescimur

(47 cases)

(46 cases)
i

4.

laetum theatria ter crepuil sonum audire magnos Lara videor duces
mutarel umbras
atqui sciebal
odi
el

29

a

(25 cases) (25 cases) (23 cases) (18 cases)
(18 cases)

5.
6.
7. ».

iuga demeret
sibi

quae

barbarus

profanum volgus el arceo me oec Chimaerae spiritus igneae
hie

9.

Lnnocentis pocula Lesbii

(17 cases)

10.
11.

donee virenti canities abest

(17 cases)

o matre
\ixi

pulchra

filia

pulchrior

(16 cases)
(16 cases)

12.
L3.

robustus acri militia puer
puellis

Quper Ldoneus
el

(15 cases)
-

M.
15.

delevil

urbem Dacus
[uerellis

A.ethiops

(12 cases)
(11 cases) (11 cases) (10 crises)

perire quaerens nee muliebriter

16.

17.
L8.

exanimas tuis temptare multa caede bidentium
cur
i

quamquam

choreis aptior et iocis

(10 cases) (10 cases)

19.

favete Unguis: earmina non prius

2

Long syllable almost always in the
fifth space.

Eorace departed from his Greek models by putting a first space- and always in the
See Table VII. 3
oote.

On

the reading of III.

5,

17. see

K

iessling

's

form .1 was without a fixed caesura or diaeAugustan age the "derivation theory" of meters held sway (see Gleditseh, Metrik, pp. 70 and 73) and Horace, apparently under its influence, resolved the Eleven-Syllable Al(3
In its (.reek
in the
resis.

Bu1

caic into

did by making
I.

two phrases of sound, each constant in length. This he See Table a word end regularly in the fifth space.
five

Only
II.

exceptions occur:
l'1
:

I.

37, 11

;

IV. 14. 17:
last

I.

16, 21; I.

37, 5;

17.

and
is

in

three of these (the

three as cited)
ii

the regular division

not

wholly absent, falling as

does be-

lie admitted syllaba tween the members of a compound word, anceps in the Una space of A, and hiatus occasionally be1 ween .1
I

and

.1

or between

.1

and

B.

See Table V.
is

I

Latin quantitative versification
is

based on

a

number

of

principles, one of which

importanl

for

our presenl purpose:
a

Lable,
:

-This term is used to designate any pari of a verse occupied by whether Long or short, there being eleven Buch spaces in I.

By]

The

tables are to be

£

I

al

the close of this paper.

vol. l]

Richardson.

Horace's Alcaic Strophe.

177

namely,

in

the initial portion

(generally two or more feet) of a
no1
is

verse rhetorical elements should
s]

often coincide with correto say, coincidence,

ling metrical elements.
Is

Thai

when

i1

does occur,

generally preceded or followed by non-coincidence.

And
imt

so

it

happens,
each
a

among

other things, thai
;

seldom

till

single fool
ictuses;

successive

successive words word accents usually do
the average
in

coincide with

caesuras on
al

outnumber
IX.
vel
alii
I.

diaereses.

These facts are hinted

by Quintilian
[i.e.,

90:

plerique enim ex commissuris eorum
sione limit pedes: ex quo
Bant.
tit

verborum
alii

divi-

nt

isdem verbis

atque

versus
first
:

The principle under consideration

is

obeyed

in

the

and second of the following
virginibus Tyriis mos
sparsis hastis longis
esl

verses, bu1 disobeyed in the third
(Verg. Aen.
(lb. [,309.)
[,

gestare pharetram.

quaerere constituil sociisque exacta referre.

campus

splendel

el

horret.

(Ennius, Varia L4.)

Thus we have an importanl

cine to the metrical structure of

any given poem. By way of brief illustration, let as suppose thai we are trying to discover the meter of the Aeneid. The itiiti
;l

|

portion of the verses

is

composed

in n

greal variety of ways, bu1

seldom or never with any of the following word-arrangements:
denique
<

laesare.

primae terrae. denique terrae.
primo
I

aesare.

This
in

is

all

the

more

significant because such groups
is.

i

ur often

Latin prose.

The fad

these word-arrangements are not

allowed to begin the verse

in

question because the rhetorical

ele-

ments would each exactly coincide with corresponding metrical elements throughoul more than one foot. Tl onclusion is therefore to be

drawn

that the feet at the outsel of Vergil's verse are
I,

either dactyls or spondees or

oil,

combined.

We may
Within the
at

reach this same result by another method of analysis.
initial

portion of the verses word-breaks tend to occur

certain points with

marked frequency.
must
he

Thes,. points, accordfeet.

ing to the principle above described, must be within
wise

Otherthe

expressed,

they

caesuras.

Knowing where

caesuras are located, we are able to differentiate them from diaereses

and

so to identify the feet.

17s

University of California Publications.

[Class. Phil.

5

Verse

.1

is

nowadays often divided
I

>;-~|-:
1

II

— -Hfilia

into feel

as follows:

>

1

1

1

we

find in

1

torace
t

many

verses. Like
pulchrior,

>

matre pulchra

where there would be an overwhelming correspondence of words

and

feet.
a

Again, according to the theory represented

in

this

scheme,

trochee would end with the third space, and yel aboul
a

half the verses have
the unity
this

break there;

if

this

were really

a diaeresis,

and flowing character of the verse would vanish. By theory, the fundamental foot would be trisemic. despite the
thai

fad
at

mosl of the

feel

as represented have syllables thai are

variance with such

a

norm.
is

Furthermore the

tine

would begin

with anacrusis, which
dence.
direct

here unsupported by any genuine evi-

In short, this interpretation of Horace's verse rests on no

ancient authority,

it

disregards well established laws of

quantitative verse structure,
those

and alto-ether

is

a

false

guide for

who would read

the Alcaic strophe in the

manner intended

by the
(6)

Roman
Let us

poet.

verse

.1

and

analy.se

now regard what we have called the first phrase of according to the method outlined in section
it

4 above.
First Space.
syllabli

In 119 verses this space

is

occupied by

a

mono-

Second Space.
space.

O -)

In 291 verses
is

a

break occurs after this

Among
o

these cases the break

preceded by:

dissyllable

-"
monosyllablt
v"

times times

p

monosylldbh
|

Third Spun.
space.

i

In

308 verses
is

a

break occurs after this
:

Among
a

these rases the break

preceded by
1!
'

trisyllable

!l

times times times

~

p y
5

monosyllabh
dissyllabh
three

dissyllabh

(i

monosyllable

•'•"•

monosyllables
to

9 times

The

fact
is

that

Horace allows words

end here with greal

fre-

quency
space
is

significant.

H implies

that

the break after the third

a

caesura.

This and the sequence of quantities involved

v,,, -

]
l

Richardson.

Horaa
a1

's

Alcaic Strophe.

179

poinl

to

iambic movemenl

the outsel

of

.1.

Significanl

also

are the differenl degrees of favor represented

in

the

numbers 67
than by

and
menl

33,
is

which

resull

in

pari

from the fad

thai

an iambic move(3

thrown into
y.

less

bold relief 4 by cases under

those under

Fourth Space.
this space.
a

(----)

In 53

verses a
is

break occurs after

Among
trisyllabh

these cases the break

preceded by:
1

quadrisyllable

times

monosyllabh
trisyllabh

22 times
LO
:;

monosyllable
5
e

times

dissyllabh

dissyllabh

times
times times

<
t?

monosyllable dissyllable monosyllable monosyllable monosyllable dissyllable
dissyllable

'.<

5

monosyllable monosyllable

times
limes

four monosyllables
'I'lic

cases

under
,-ti

a are [1, 17,

6

;

III, 5,

10 21, 10 IV,
;
;

4, 69.

Two

of them,

Least,

may

be neglected:

in

III.

:>.

Id. the

quadrisyl-

lable exists only so
in

far as results from ;m elided

pentasyllable;

IV

.

1.

ill),

the verse begins Carthagini iam, where the noun and

particle are closely joined

and the
.1

effect
a

is

much
single

the

same

;is

though the
word.
taken

first

phrase of

embraced
;it

pentasyllable
of verse
is

The
in

rarity of quadrisyllables

the outset

.1.

connection with the succession of quantities,

an indi-

cation of iambic

movement.
in

Noteworthy
the
less

also are the differenl
l'l*

degrees of favor repr< sented

numbers

and

Hi.

an iambic
f3

movemenl being thrown
.

into

hold relief by cases under

than by those under y The unwelcome character of the cases under 8 is made evident not only by their rarity hut also by the

'An iamb is thrown into relief when it is occupied by two monosyllables; u diiamb when it is occupied by:
u b
<
(I

a

dissyllable, or by

quadrisyllable.

dissyllabh dissyllable. dissyllable monosyllabh

f

g
h

monosyllable. monosyllable monosyllabh dissyllable, lonosyllabh dissyllabh monosyllabh monosyllabh trisyllable. trisyllable monosyllable. four monosyllabh s.

As

m

rule,

it
;i

is

less

auslaul oi

foot.

tendency
quently,
c

to

render
<

and

objectionable to throw into relief the anlaut than the niter the penultimate Byllable of the fool has objectionable a break after the auslaut. less objectionable than d; and are g less objectionable

A break
less

;i

I

than

/'.

180

University of California Publications.
occur.
In

1-i.ass.i-hm..

way they are disguised when they do
ii,. s
i

III.

29, 5, the
trisyl-

f the pair exists only so far as results In
I.

from an elided

lable.

L6, 21,

the pair arises from two elided trisyllables.
In
I.

The

verse has no break after the fifth space.

37, 5, the

firsl

f the pair arises

from

a

trisyllable affected

by

synizesis.
In
is

This

also omits the usual break after the fifth space.

shun.
found.
in

oo real case of two dissyllabic words beginning
'Phis
is

a

verse

stroii- evidence of an iambic
is

movement.
under
e

Pointing

the

same direction under £ and
ry.

the fad

thai

cases

outnumber those

Fifth Space.
this space.
a
j8
-

(--

)

In 629 verses
is

a

break occurs after

Among

these cases the break

preceded by:
1

pentasyllable

"
times

quadrisyllabh

monosyllable

4

monosyllable quadrisyllable
trisyllabh
dissyllable
trisyllable

42 times

5
e

dissyllable
trisyllable

176 times
1" :
'

''"""-

j
7)

monosyllable

monosyllable

20 times

monosyllable trisyllabh

monosyllabh

10 times
66 times
....

6
t

monosyllable monosyllable trisyllable dissyllable dissyllable monosyllable
dissyllablt

1

time

k

monosyllabh
monosyllable
dissyllabh

dissyllabh
dissyllable

33 times 58 times

\

monosyllable dissyllabh
lissyllable

monosyllable
monosyllable

mot imea

nosyllable
v

monosyllable
nosyllable

mo9
1i,, "' s

I

monosyllable
nosyllable

monosyllabh

dissyllable

mo5
dis-

times

o

monosyllable monosyllable monosyllable
syllable

9 times ° times

k

five

monosyllables

Thai verse

.!

begins with iambic meter

is

evidenced by the char-

Table IV acter of the monosyUables falling in the fifth space. shows 49 such cases. The resulting break after the fourth space
is

usages:
in

generally bridged over and softened by some of the following a In twelve cases elision lakes place, being Located as
the following

example:
el

dulce

deconwn

est

pre patria

men.
III.
•_'.

13.)

Vol.

i
|

Richardson.

Uoraa

's

Alcai<

Strophe.

L81

(6)

In twenty-nine eases the break

in

question

is

concealed by

another break after the third space.

Tims
of

the metrical
is

phrase

doses with two monosyllables.
a

One

them

no1

infrequently
the promi-

proclitic or an enclitic,

which also serves

to
in

lessen

nence of the break after the fourth space, as
,

in. -cutis

ad

Si

cuncta pe
i

IV.

9,

38.)

(c)

Only

a

few cases remain, and

in

some of them the monosylknit together, as in

lable

and the preceding word are closely
iamdudum apud
m<
i

st.

eripe te morae.
III.

Again, significant of iambic meter
y so greatly
£

is

the fad thai cases under
thai 6 thai
v

outnumber those under /?; that A outnumbers k; outnumbers
,,-.
;

outnumbers
and
«

£; thai
<

outnumber

;iml

;

:

that

S

and

e

are strongly in favor.
.

If the first phrase of

1

(five syllables)
in

is

compared with whal
Nor;
*^

precedes the main caesura

the iambic trimeter of

Epodes
will be

i

normally

live syllables),

the words occurring in one case

found

to accord with those in the other as regards their

form, length, and arrangement.
III.
t

This
il

is

well illustrated by

Epode

where the word-a rra
:

1

1

Lionel

s

in

verses Tree from SUbstitu-

ions are typically

-I I-

1-1II-

l-l
(7)
tli

I-to he
in

The following points are

noted for the

lighl
.

they
.-

row on the nature of the rhythm
(a)

the second phrase of

I

Breaks within the phrase occur freely after the sixth,

seventh, eighth, and ninth spaces, most freely, however, after the
eighth.
/,

(Table

I.)

The

favorite combinations of

words within the phrase

are, in

order of preference, as follows

182

University of California Publications.
trisyllabh
trisyllable,

'

lass.Phil.

monosyllabh

trisyllabh

dissyllabh

.

yllabh

trisyllable,
.

quadrisyllabh dissyllabh
disyllabic quadrisyllabh

(Table

III.

Monosyllables are

abundanl

in

the
a
is

sixth

space

alone.
in

Only seven linns does
monosyllable
occiii's al
!t.

a

verse end with
is

monosyllable and
to say, in
II.

six of these cases the effeel
id,,
is

veiled.

Thai

11, 13,

preceded by another monosyllable, as someBu1

times
In
in
I.

the close of the dactylic hexameter or pentameter.
5; III. 26, 9; 29, 9;

13;
9,
1.

II. 15,

and

4!)

there

is

elision.
a

IV.

the monosyllable stands out boldly after

pentasyl-

lable, mii effeel that is

probably intended to reinforce the striking
the seventh or eleventh space. the eighth or eleventh space.
in the

character of the thoughl
d)
Dissyllables end freely
in in

Trisyllables end freely

Quadrisyllables end freely
I

ninth or eleventh space.
al

'entasyllables

and bexasyllables occur occasionally

the

close of

t

he verse.
to be

The conclusion
same manner

drawn from
him
first.

this evidence

is

as follows:
in

The poet's Peeling has
the

not Led

to treal the

second phrase

as he did the

Be

has not here studiously
foot, since

avoided the coincident termination of word and
occur freely
tion
a1 all

breaks

points, excepl after the tenth space, an excep-

due

to the fact that

monosyllables are oo1 welcome

in

final

position.
8

We

are

now

in

a

position to
lirst

make
ii

oul the meter of the

whole verse.

As regards the

phrase,

has been shown that
to the
fol-

words are frequently chosen ami arranged according
lowing divisions:

--I --" -l-l

We

rarely find

:

vol..

1

Richardson.

Horace's Alcaic Strophe.
Bu1

183

The meter,
particular

therefore,

is

iambic
?

in

character.
firsi

whal

ia

the

form of the

feet

Do the
IT

four syllables constitute

two iambs or our diiamb?
istically as

these syllables appeared character-

there would be ground for recognizing two
are normally is
-

iambs, bu1 asa matter of fad the}
nineteen verses begin
t

(onlj

)

and the conclusion
lias

inevitable

h,it

.1

begins with
its

a

diiamb.
unity, which

A verse by

very nature

implies thai
t

it

embraces homogeneous elements. Therefore, since he firsi phrase of .1 contains a diiamb plus one syllable it is probable thai this
syllable

introduces

a

second

metrical

division,

no1

necessarily
in

identical with the

first,

but similar in kind

and commensurate
is

duration.
in effect

Keeping

in

mind

thai a
a

diiamb
foot

quadrisyllabic and

hexasemic, we find that

having these two propfifth,

erties

is

made up by

the syllables in the
it

sixth
a

seventh, and
foot

eighth spaces.

Moreover,

assumes the form of

to

which

ancient writers on metric frequently refer, namely, a major ionic

(-in

-

-).

Three syllables remain, long short long, respectively, and they
turn answer the conditions of
a

quadrisyllabic hexasemic foot.

one.

however, that has been

dilied

by catalexis
this foot

in

the final

cadence of the verse.
ently be a ditrochee, as

In acatalectic

form

would apparIvt

may

be gal hered from the Tw<
I

-SyllabU

Midic

cited by

Bephaestion

Emit..

XIV,

4. ('.

16tt\ok\

ayvd

/xeAAt^d/i-etSc "SaTrcpol.
fr.

With

this verse

(=

=Aleaeus
aXXd
/i.e

34) compare:
uiOws.
.

6e\w

tl f€LTrr]i\

Kwkva

AlcaeUS

fr.

lib
fr.

KOiAan/u'x<iJi' i-rnron'

TrpvTavis [HormOui']

Stesirhonis

21.

Verse

.1.

then,

may

be classed as an

Epionic Trimeter Catalectic

and

is

to be represented

thus

:

reader being always

al

liberty

to treat

the last foot as

This conclusion
'Hephaestkm [Eneh.
8t airb
6
'

is

not

only supported by ancient authority, 5
'<

C. deseribea .1 in its Greek form as fo AAxou kov KaraX-qurinbv iari, rb Ka\ovp.(vov i^aar^pav f) irpuTyv ovfvyiap e\ei. ia.fx.JiKr)i\ tfroi rijv Of devrepav iojviktjv airb p.d'{ovos i) bevripav iraiuivixriv, ttjv 5t {TTTaff-q/jLOv, A,-arcU\ei5a (k rpo\aiov xai 7-77S &5ia.<pbpov.
V,TTLioviKbi>

WW

ixu'^ovos

Tpip-erpov

ivbtKaevWoLtiov,

tt)v

p.ev

184

University of California Publications.

[Class.Phil.

bu1 corroborated by

numerous
motive
8

parallels in allied verse forms; for
is
:

example, the
Y,

init ial

found
VI.
1:

in
1

Pindar,
;

mea,

I.

str. 2

and 4; V.
124c
,

str. 5

Isth.,

I.

str. 5;

sir.

IV.

29,

1; fr.

122, 1:

!'i.

1: Bacchylides,

VIII, ep.

XI.

1

and

8;

XIV,

ep. 1; et passim.

The
found

final motiv<

-- is also abun<i*J7.

dant; see for example Christ, Metrik, section
in acatalectic

This motive
fr.

form

is

in

Sappho,

fr.

50; Pindar,

75, 4;

Kurip.. M< din. 151

3;

and elsewhere.

/;.

THE NINE-SYLLABLE ALCAIC.
ways
<>r

(1) There are 256 possible

arranging words
I

in

;i

verse

of nine syllables.
48, confining
1.

I

n lliis Alcaic,
;is
;i

however,
1<»

[orace

employed only

himself

rule to the

following:
(60 cases)
(26
c;isrs
i

cantemus Augusti tropaea
rugis
cui
el

2. 3.

Lnstanti

senectae

laurus aetemoa honores

(26 cases) (21 cases) (20 cases)
(17 cases)

4.
5.
(I.

redegit in veins timores
oblitus aeternaeque Vestae

quantis fatigarel

ruinis

7.
8.
i).

excepit ictus pro pudieis

(15 eases) (13 cases
(12 cases)
(

qod Seres infidive Persae
sortitur insignis
el

imos

10.

sumptu Lubentes

el

deorum
this verse
is

8

cases)
in

(2)

The metrical character of

revealed

what

follows:
First Space.

In 84 verses the initial
his

word

is

a

monosyllable.
a

Horace departed from
lable almost

Greek models by putting
See Table VII.

long syl-

always

in this space.

Second Space.
space.

(- -)

In 83 verses a break occurs after this
is

Among
a
(8

these cases the break

preceded by:
75 times

dissyllablt

monosyllabh
(-

monosyllabh

8

limes

Third Space.
this space.
a
j3

-

)

In

259 verses
is

a

break occurs after

Among
trisyllablt

these cases the break

preceded by:
L55 times

monosyllabh
dissyllablt

dissyllable

65 times 35 times
I

7
5

monosyllabh

three

monosyllables
is

times

term motivt

use.

I

to

designate any dominanl metrical design or

Bequem

e.

Vol.1]

Richardson.

Horace's Alcaic Strophe.

185

Evidences of iambic movemenl are seen in the sequence of space, quantities, the greal frequency of breaks after the third
the frequency of trisyllables as

shown

in

a,

and the fad

thai

ft

outnumbers
this space.
a
i
t

y.

Fourth Space.

(~

--)

In 51

verses a break occurs after
is

Among
trisyllabh

these cases the break

preceded by:
"

quadrisyllable

times

monosyllabh
trisyllable

34
'-'

times

y
5

monosyllable
dissyllabh

,m "' s
times

dissyllable
13
(l

monosyllable dissyllable monosyllable monosyllable monosyllable dissyllabh dissyllable monosyllable monosyllabh four monosyllables
in
-

times times
times

f
Tj

2

e

times

Evidences of iambic movemenl are seen

the sequence of

space, the quantities, the infrequency of breaks after the fourth way e outfacts under a and 8, the way p outnumbers y, and the

numbers

£.

Fifth Spaci
this space.
a
,3
-

.

(~ - - - -)

In 52 verses a break occurs after
is

Among

these cases the break

preceded by:
:;

pentasyllabh

times times

monosyllable quadrisyllable quadrisyllabh monosyllable
trisyllabh

•">

" times

5 e

dissyllabh
trisyllabh

23 times
11

dissyllable

times

f
t,

6
,

monosyllabh monosyllable trisyllabh trisyllable monosyllable monosyllable monosyllable trisyllable monosyllabh
monosyllabh dissyllable dissyllable dissyllable monosyllabh dissyllabh dissyllable dissyllabh monosyllable
dissyllable

time
times

"
1

time
times

5
1

«

time
tin

\

n
,.

monosyllable
dissyllabh

monosyllable monosyllable
dissyllabh

mo"

nosyllable

times

monosyllabh
nosyllable

mo,im,,>

£

monosyllable
nosyllable

monosyllable

mo
u "
!

o

monosyllable monosyllabh

monosyllable

dis]

sylabh
7T

''"'"

ti\

«•

monosyllabh

s

° times

and

Evidences of iambic movemenl are seen in a comparisoi The relatively small number k, and A. y. of 8 and t. and of
t,

L86

University of California Publications.

[Class.Phu*

of breaks after the fifth space

makes
has
a

it

clear thai this verse
/»'.

is

no1

divided into

se1

phrases of sound.
.!.

Tims
a

though beginning

with the same quantities as
Kiessling pointed ou1 thai
the effecl of the break
is

differenl

opening cadence.
in

when

word ends

the fifth space,
a

generally subdued by the presence of

monosyllable

in

the sixth space.

This feature affords

a

good exI

ample of the way Borace's arl underwenl change. In Book four verses have a word ending in the fifth space withoul a following monosyllable
there are
11;
lit.

16, 3; 26, 7:

29,
3,

11

;

35, 11

I.

In
;

Book

II

sewn such
in

verses (1, 11;
III
in

3; 13, 27: 14. 11

19, 7: 19,

19).

Books

and IV they disappear altogether.
the fourth or fifth space,
in
a

Since words seldom end
labic
is

monosyl-

not likely to occur often
is

the fifth space.
27.

Only one
neutralized
in
''>.

example
by

found, namely

et

in

II. 3,

and

this

is

elision.

Owing
and only

to the general

avoidance of words ending
a

the fifth space, only
3;
!!•.

two verses end with
eighl

quadrisyllable
|

i

II.

lit)

end with two dissyllables

I.

16, 3; 26,

7; 29, 11; II. 1. 11: 13. 27: 14, 11: 19. 7: 19, 11).

It is

an inter-

esting fact, as
first

.Mr.

Page points out. that
is

in six

of these cases the

dissyllable of the pair

repeated at the outset of the suc-

ceeding verse.

For example:
AJcaee, pleetro </""/ nai
is,

dura fugae mala, dura

belli.

ft!,

t:;.

27 28.)

Horace departed from
;i

his

Greek models by putting invariably

Ion- syllable in the fifth space.

Sixth Space.
after this space, of such
a a

(-

-~

)

In 251

verses

a

break occurs

verse

mark of iambic movement, for toward the close the usages of diaeresis and caesura undergo a
words
at

change, breaks after the even syllables becoming numerous.

The
ill

favorite combinations of

the close of

a

verse are.

order of preference
trisyllable trisyllabh
.

monosyllablt dissyllabh

.

monosyllable trisyllabh

(Table 111.)

vol.1]

Richardson.

Horace's Alcaic Strophe.

Lffi

Words

of more than three syllables occurring

in this

verse are

interesting as regards both their rarity and their position.

Only
space.

one hexasyllable occurs and

that ends in
in

the seventh

Among
and
5

11
in

pentasyllables, 3 end

the fifth space, 3 in the sixth,
in

the seventh.

Among
in

69 quadrisyllables, 5 end
the seventh,
in

the

fifth space.

26

in

the sixth, 36

and

2 in the ninth.

Tims

these polysyllables lend to occur

the middle of the verse.

(3)
p.

By
IV..
it

a

process of reasoning similar to thai
Pell

followed on
firsl

177

appears that Borace

the

rhythm of the

four

syllables as a foot in the shape of

a

diiamb.
a

Especially signifi-

cant

is

the fact thai not
it

a

single verse has

word ending

in

the
aexl

fourth space unless

be

a

monosyllable or trisyllable.
to a

The

four syllables also conform

diiamb.
in

This

foot,

i1

should be remembered, occurs

Alcaeus and
of

Sappho both
the latter

as
in

and
Horace's alcaic
in

form

The extreme raritj strophe may be due to
.

the

abundance of Long syllables
by
fool
itself
is

Latin, to the fad thai this form

metrically ambiguous, being either a quadrisyllable

or two dissyllabic feet, and to the fact that the gravitas Roinvested his
(

mano with which Horace
the foot
it

hies

is

better served by the
In

form of the diiamb containing three long

syllables.

reading

does not stand to reason that the ancients
initial syllable.

consciously shortened the
in

Any

positive reduction
sense.

Length

at

this point

would often confuse the

For

ex.

ample, shortening the

initial syllable of

canes ('thou arl hoary'

which mighl conceivably be the word concerned, would resull in canes 'dogs' I. The same applies to scores of words subjed to a
I

similar change of meaning, should the

The

fact that

firsl

syllable be shortened.
fool
is

-

-

is

in effect a

hexasemic

rather to

be explained on other grounds.

To be

sure, this diiamb.
a

when
mora;
and

exactly measured, seems to be overlong to the extenl of
bu1 since the overlength
is

in
is

the

firsl

syllable of the foot,
is

since the

compass

n\'

the fool

Large, the excess

aeither enough
in
tl

nor

in

; i

position to unbalance the rhythm.
is

Compare

nection the ditrochee, which

also hexasemic in effed
(

and

fre-

quently
in

lias tliree

long syllables

).

thus being overlong

the

hist

syllable of the foot.

188

/

niversity of California Publications.

[Class. Phil.

The

final

syllable of this verse remains to be accounted

for.

some scholars (Masqueray: Metrique, section 276, and C of the Greek Alcaic strophe GleditsCh: section L50, 3),
According
to
/»'

.,,,.

held to be

in

effed one Long verse.
Eel1

Bowever

thai
is

may

be,

Horace certainly

B and C
between

as separate verses, as
a1

shown by
to

the Pad thai he admitted syllaba anceps
as interverse hiatus

the close of B, as well

B and C

(Table V).

According

0. Schroeder {Berl. Philol. Wochenschr., 1904, Nr. 51),

B

is

an

iambic pentapody
iamb.

i

Fiinfkeber), the

final syllable

representing an

The conclusions reached
by Kiessling and

in this

paper supporl the view held
/>'

many

others, namely, that

is

hypercatalectic.

The
as
is

transition

from the ascending rhythm of

this verse to the
syllable, just

descending rhythm of
the ease
iii

C

is

facilitated

by the extra

the following examples:
Bacehyl.

----,-

Vn,

b. 14-15.

Cf. XI, 2-3.

_^__L_^

«

XII, str. 2-4.

Cf. ep. 3-4.

XVIII,

str.

1-2.

XIII, ep. 6-7.

\'~_

«

VIII. str. 8-9.

<'f.

ep. 3-4.

Especially significant are the following:

Bacehyl. XIV,
Cf. Pindar,

str.
fr.

3-5.
,

I24 c

L-2;

fr.

126, 1-2.

•-,-

____'__^t (

Bacehyl. XIV, ep. 1-3.
Cf.

Pindar,

fr.

122, 1-3.

In the last
is
It

example
having
a

line

1

nearly equals

.1.

line 2 equals
is

/>'.

line 3

like

C

in

descending rhythm part of which
line 2
I

trochaic.

seems fair to say thai

which equals

II

i

is

an Iambic

Dimeter Hypercatalectic

oo1 a

pentapody

i,

since the nezl to the

Vol.i]

Richardson.

Horace's Alcaic Strophe.

189

Last

example clearly shows
iiiri

thai a

dimeter may Legitimately occur

in this

rical context.

Vrrsi'
talectic

ll.

then,
is

may

be classed as an [ambic Dimeter Kyperca-

and

to he represented thus:

C.

THE TEN SYLLABLE ALCAIC.
in a

(1)

Then- are 512 possible ways of arranging words
In this Alcaic the

verse of ten syllables.

pud employed hut

ol,

confining himself as
1.

a

rule to the 10 following:
(49 rases)
(36

egit

equos volucremque currum

2.

divitiis potietur heres

3.

vis rapuit rapietque gentis

(33

easi

4.
.">.

purpureo varius colore
levia

-

(27 cases) (18 cases)
(16
..-,-

personuere saxa

6.
7.

Pegasus expediet Chimaera fronde nova puerum palumbes
Delius '! Patareus Apollo pomifero grave tempus anno in domini caput inmerentis

(13 cases) (13 cases) (13 cases)
(11 cases)
is

v

.

9.

1".

2

The metrical character of
:

this ver.se

revealed

in

the

following analysis
First Space.
syllablt

In 69 verses this space

is

occupied by

a

mono-

Second Space.
this space.
a
j3

(

)

In

101

verses
is

a

break occurs after

Among
dissyllable

these cases the break

preceded by:
99 times

monosyllabh

monosyllable

2 times

Third Space.
this space.
a
/S

(- - -)

In

79 verses a break occurs after
is

Among
trisyllable

these cases the break

preceded by

:

60 times
9

monosyllable dissyllable
dissyllable

7
5

monosyllable

10 times
n times

three

monosyllables
is

Dactylic

meter

here suggested,

for an

initial

trisyllable

occurs

less

frequently than an
/;
/»'
.

initial dissyllable

99 cases

com-

pare pare

.1

and

and

o

and

-.

enjoy virtually equal favor

.1

and

.

190

University of California Publications.

[Class-Pho-.

Fourth

Space.
It

(is

-

-)

In
:

252 versea

I

break occurs

after this space.
a

preceded by

quadrisyllable
dissyllable dissyllablt

s"

times

s

l

times

7
5
e

monosyllable
trisyllablt

trisyllable

55 times 20 times
9 times

monosyllabh

other combinations

The unequal favor enjoyed by
Fifth Space.
this space.
It
(

— —
-

y

and

8

points to dactylic meter.
a

)

In 14 verses

break occurs after
four limes.

is

preceded by trisyllablt
fifth

dissyllablt

A break after the
is

space

falls

between two short syllables and
a

so situated in the verse as to produce

weak

effect,

which seems

to

account for
Sixth

its

infrequency.
(Tt

Space.

—is

~)

In 52 verses a break occurs

after this space.
trisyllablt

never preceded by Tiexasyllable, and by
4.

trisyllablt

bu1 once, namely, in IV.

72,

where the

second trisyllable exists only so far as arises from an elided quadrisyllable.

This

is

strong evidence of dactylic meter.

Seventh Space. (- — — -->
this space.
It
is

)

In 112 verses a break occurs after
trisyllablt

never preceded by either trisyllablt

monosyllabh or hexasyllablt
tylic meter.

monosyllable, which points to dacto be

The general conclusion
is

drawn from the
syllables of
in

fore-

going points
dactylic feet.

that Horace felt the

first six

C

as

two

Since words are seldom allowed to end
a rule,

the fifth

or sixth space, polysyllables are barred, as
in

from beginning

the sixth or seventh space and. therefore, tend to gravitate to

the initial or middle parts of the verse.

The usual combinations of words
in

at

the close of the verse are.

order of preference

:

quadrisyllabh
trisyllabli

dissyllabh
.

.

trisyllabh

dissyllable dissyllabh

.

quadrisyllable trisyllable.

(Table

III.')

3)

We

are

now

in a position to identify the meter through-

out the whole verse.
ff.

Hephaestion (quoted by Gleditsch,

p.

L73

applies the term logaoedic to dactylic or anapaestic verses in
initial or final parts (or

whose

both) the arses consist, uo1 of pairs

of short syllables, Iml of single short ones.

Be

cites

C

as an ex-

Vol.1]

Richardson.

Horace's Alcaic Strophe.

191

ample (Ench., VII. s. C.). more rapid tempo than
;i

By reading the dactylic dipody with
is

given the

final

ditrochee the time

relations of the verse are as a
Allg. Metrik d. Gr., III.
1.

whole kepi true (see Westphal:
:

p.

366; Masqueray

p.

328; Gleditsch:

section

142).

Compare

in

this connection

the substituted ana-

paests and dactyls in Horace's Epodes; these feet, which are ordi-

Further lighl thrown on the metrical structure of C by the following verses, some of which are logaoedie and some rochaic
is
t

narily tetrasemic, are there given trisemic values.

a

I


.

\
fi

__

Eybrias {Antk.
275).

Lyr.

p.

_ ~
-•

~-^
^^-w.
s

Baechvl.
<-f.

fr.

20,

str.

2;

XV.

str. 4.
fr.

7
5
e

SimoTiidcs.

57.
l">.

-._

^_-^_. — -^„____ w _
_

«_--

Aristoteles, fr. 5,

_„____

Simonides,

fr.

30.
;

f v

--


,

--

-

—— —

Praxilla, fr. 5

ef
str. 5.

Bacehyl. XV,

-, - - -- -,

AJcman,
I:

fr.

5, str.

'J-

it.

Here
This
verse
('.
8-£.

— ~and-~seem
especially evident
its

to
in

Ik-

made equivalent
Examples

in

time value.

is

The alternative forms of the same
»-y end somewhat like having the exad form of C
fr.

included within braces).
quite in

manner.

A

line

concludes three of the strophes in Alcman,
[bycus,
frr. 1. 9; 8c,
1
;

5;

it

occurs also in

8e,

1

;

L3, 4: 15, 2;

Bacehyl., [V,str.6; ami
l>\

eighteen passages of Greek dramatic poetry cited

W. Christ:
270f.

Grundfragen der melischen Metrik
Akad. dm- Wissensch.,
It is

>/>

r

Griechen, Abhandl. der

Philos.-philol.

CI..

Munchen, 1902,

found
it

to follow
is

iambic and oilier kinds of verses; qo1 infrea

quently
a

used to conclude

strophe.

Since strophes having

distinct kind of yerse as clausula are

abundant, nothing stands
die verse in the shape of
;i

in

the

way

of our taking

C

as a Logj

192

University of California Publications.
ditrochee.

[Class. Phil.

dactylic dipody

followed by

a

H may be termed

a

7 Dactylotrochaic Dimeter, being represented thus:

THE ST Km pi IK AS
The pud's

A

WHOLE.
is

feeling for the strophe as a whole

reflected

in

the following points:
l

Elision occurs 69 times in the
in

firsl

verse of the strophe,

59 times

the second, 38 times in the third,

and
3,

31 limes in the

final vers.',

[nterverse elision occurs twice (II,

27-28; III, 29,

35-36
(2)
il,,w

.

Since interverse hiatus works against the unity and even

of the strophe,

Less

often as Horace's art develops.
in

we should expect to find it occurring Less and Such turns ou1 to be the fact,
firsl

as

appears
(3)

Table V.
verse of the
so in the second, infrequenl

Sense-pauses are numerous within the
still

strophe,

more

in

the third,

and

The majority are not coincident with the main rhythmical pauses, the sense being made to run on from verse to misc. and strophe to strophe. (4) Long words tend to occur in the latter part of A, but in
rare in the fourth.

the middle of

B

and of

G.

Furthermore, as regards word-lengl as
thai
i

and combinations of words, Tables II and III show verse has different habits of beginning and ending
of B, however,

I

1

I

each

the extremes

do not differ greatly)
/»'

;

(2)

A

has characteristic
(3)

ways of beginning,
same may be said of
other verses.
(5)

has others,

C

still

others;

much

the

their closing, the final effects of the clausula,

however, being especially well differentiated from those of the

We may

here consider the question whether the Alcaic
differs

strophe of

Book [V

materially

from thai of

Borace's

earlier work.

A comparison shows

results

somewhai

as follows:

Type

1.

as recorded on p. 176, occurs in

Book IV ten

times, type

2 eleven limes, type 3 once,

type 4 not
it

at all,

type 5 twice, type 6

once, type 7
7

mx

1

inies.

In short,

turns out that certain forms
is

Justification for Gleditsch: section 65,

bringing two dactyls within one meter
1, fin.

found

in

vol.11

Richardson.

Horace's Alcaic Strophe.

193

of verse abundantly

represented

in

Books I-III arc relatively

Book IV. and via versa, the general resull is being thai in the pod's later work the range of lyric effects somewhal narrower. The bold and exceptional features of the strophe, cited passim in the foregoing pages, poinl to the same

much

less

frequenl

in

conclusion, since they are in
[-III.

large measure confined to
<

Books

Lighl

is

sometimes thrown on the date of ;m
lines.
is

>de's

compo-

sition

by

tests

along these

(6)

The

location of the ictus

a

matter ool so easily deter
to be

mined

as the

form of the

feet.

However, we seem

war-

ranted in holding that an ictus belonged to each foot, and thai if it belonged to the first half of a given foot, it belonged to the same
half of all the feet alike.

An

ictus

hardly belonged

to the final

two syllables of the major (____). This leads one to infer that in each foot the ictus longed to the first half. The interpretation of the Seikilos
I

ionic (--

~~)

or

he dactylic dipody
hein-

scription

and Anonymus Bellermannius, section 85, given by F. Blass {Hermes, 35 [1900], 342; Neue Jahrb. Mass. Altertum, 3 [1899]. 42) points to the first half of a diiamb as the place of the
ictus.

(7)

The strophe

as a

whole may he represented, from the

standpoint of reading, thus:

194

University of California Publications.

[Class. Phil.

TABLE
The number
the strophe
is

1.

of times thai
in

a

word cuds

a1

any given poinl

in

shown

the following table.
.1.
1!»!)

For example, among
;i

the 634 verses included under

begin with

monosyllable,

291 are so

composed that

;i

word ends with

the second space, 308

with the third space, and so on.

Vol.

i

Richardson.-

Horace's Alcaic Stroph

l!tr.

TABLE
Sin
•d-leneths

II.

he beginning of

;ill

four verses.

L96

University of California Publications.

[Class.Phil.

TABLE

III.

Summary
four verses.

of word-lengths in the concluding portions of

all

monosyllable monosyllable.

monosj

liable dissyllable. liable trisyllable.
*

'

monosj

1

monosyllable quadrisyllable.

'monosyllable pentasyllable.
'

dissyllable monosyllable. dissyllable dissyllable.

'

'

dissyllable trisyllable.
dissyllable quadrisyllable.

'

'

dissyllable pentasyllable.

'dissyllable hexasyllable.
'

trisyllable monosyllable.'
trisyllable dissyllable.

'

trisyllable trisyllable.' trisyllable quadrisyllable.
*

'

'

trisyllable pentasyllable.
t

'

ri syllable hexasyllable.

'quadrisyllable monosyllable.

'quadrisyllable dissyllable."
' •

quadrisyllable trisyllable. quadrisyllable quadrisyllable. quadrisyllable pentasyllable. quadrisyllable hexasyllable.
pentasj liable monosj Liable.

'

'

pentasyllable dissyllable.'

'

pentasyllable trisyllable.

'

' •

pentasyllable hexasyllable.

'

hexasyllable monosyllable. hexasyllable dissyllable.
hexasyllable trisyllable.'
'

'

Vol..

1
|

Richardson.

Eoraa

's

Alcaic strophe.

197

TABLE
This table takes aceounl of
all

IV.

words found

in

Borace's Alcaic

strophe, showing their Length in tonus of syllables, their relative

frequency, and the places of the verse
table
first
is

in

which they end.
in
.1

The
the
;

to be read as

follows:

199 monosyllables stand
in

space of

A;

211 dissyllables end
in the

the second space of
so on.

60

trisyllables

end

third space of

C; and

L98

/

niversity of CaUfornia Publications.

[Class.Phil.

TABLE
[nterverse hiatus occurs

V.

fifty times, as

shown

in

the follow ing
.1

enumeration.
closed

Cases falling between verse
parentheses,

C and

verse

are en-

within
a

being

less

objectionable

than

those

within

single strophe; cases involving an exclamative monosyl-

lable, likewise little objectionable, are
I.
!'.

marked with an

asterisk.

7

;

I

I.

I. I. I.
I. I.
I. I.

16,

(16); 27.
6;
L3;
I

17.
i!<).

L6)

;

25.

None.
Noi
e.
1 i

Total:
eases
\\

27.

[thin

si

raphes.

29. 31,

X
i

eases between strophes.

5;

II.

34.
35,
37,

None.
9;
11.

I.
I.

(12

i

:

(32)

;

38.

II. II.

1.

(12).

3,
5, 9.

II.

11.7.
II. II.
II.
. .

None.
3;
(li').

9,

11. L3,

None.
S
r; s " s
'

Total: (8);
11:

(4); 7j

21;

26;

(28).
9 cases

within stroP hes

-

between strophes.

II.
II. II,
I

15.
17.
1!>,

None.
(4*);
31.

(20).

I,

l'u.

None. None.
17:

III.

1.

Ml.
III.

2,

(24).
i

3,
A,
5, 6.

(8); (4);
1":

10).

III.
III. III.

(16);
11
:

(28);

('

!);

(7(5).

(12

None.
None. None.
i

III.
III. III.

17.

21.
i'::.

L6

,

111. III.

26.
2!>.

None.

None.
I

IV. IV, IV. IV.

I.

9.
I

None.
l.

None.
In.

I.",.

vol.

i|

Richardson.

Eoraa

's

Alcaic Strophe.

199

Many
turity
in

instances of interverse hiatus

in a

poem indicate immaII.

the poet's art, intractability of material, or conditions

of composition some

way unfavorable.
a

Ode
form

L3,

for

example,

shows

not only seven cases of interverse hiatus bu1
:

the following
L

unusual points
in

Verse 22 {A) has
.1
i

no1

fo

elsewhere
II.
1 1

Borace; thai of verse 33
i

occurs again only
in
I.

in

7.

L3;

thai of verse 14

.1

occurs again only
is

-'A.

10,

and

1.

4.

17

;

thai of verse 27

(/>')

unique; thai of verse 19 (B) occurs again
is

only

in III. 6,

11; that of verse 8 (C)
I, 9,

unique; thai of verse L2

occurs again only in

24.

200

University of California Publications.

[Class. Phil.

TABLE

VI.

Features of rare occurrence may by their very rarit}
lighl

ilimw

mi the oature of the verse.

One may thus
as

see whal the poel

generally avoids and, by contrast, whal he seeks.
verses

The following
of
of

of

Borace are each unique

regards arrangemenl
a

caesuras and diaereses.
Alcaic ant ibarbarus.
1.
(

Rightly interpreted they form

sorl

)

Thaliarche
sit

merum

diota.
et.

i

I

quid

futurum eras fuge quaerere
in

I

fervor
el

el

eeleres iambos.
est.

I

musa eordi

hie tibi eopia.

I

hunc Lesbio saerare pleetro.
teque tuasque deed sorores.

B
(C)
I

mercede. quae
N(
lie.

te

cumque domat Venus.
ter et quater.

dis earus Lpsis,

quippe

(.1)

me

ciehorea levesque malvae.
e1

(C)

quo Styx
Spes

Lnvisi

horrida

Taenari.
et.

(A)
(/>')
I

regumque matres barbarorum
te
el

albo rara

Fides
in.

relit.
i

Lneude diffingas retusum

n

Ills

avitis

dum

Capitolio.

B (A)
I

mentemque Iymphatam Mareotico.
I

[aemoniae daret

nt

catenis.
et

I

principum amicitias
res ordinaris

anna.

(C)

grande munus

nun decoloravere eaedes.

(B) (B)
(C)
(/>')

quae caret ora eruore aostro.

ab

insolenti
et
et

temperatam.

hue vina

unguenta
aetas
et

et

nimium
iiitiin.i.

brevis.

dum
nil

res

sororum.
et
;
1

(B)
i

interest

an pauper
et

I
I

sni's exitnr.-i

lies

iii

et

emum.

(/>)

None.

depone
a1

snl>
e

lauru

mea

nee.

/.'
i

Arineiiiis in oris.

|

dob ter aevo functus amabilem.
I

(A)
(C)
I

quaerere nee
\

repides

in

usum.
vel line.

ernis neque ano luna rubens oitet.
lien

cur

sub

:ilt;i

vel

platan
<

I

hospitis;
en:i\

Qle renena
six e

lolcha.

I

(

[ganda

reges.
\

B
et
. I

e\ incet

lllllies.

tlllll

iehllill

nee dis
nt

amicum
uos!

est

nee mihi te prius.
Lncredibili

(A)
|

rumque

rum

modo.

l
I

Vol.

I
|

Richardson.

Horace's Alcaic Strophe.

201

nodo eoercea viperino. In. que did us nun sal idoneus.
I.

I

None.
d(
\

scendal

in

campum
o el
el

petitor.

B
I '

itamque Bub dn
populo

trepidis agal

('inn

duce fraudulento.
situm.

aurum inrepertum e1 sic melius coniuge me l"\ i- el sorore.
qui
s
i

I

t

aquae subeunl
(lis

el

aurae.

'

me fabulosae Vblture
inni
\

in

Appulo.
dato.
t.

1

sin.'

animi -us Lnfans.
el

(C)

os lene consilium

da1
I

is

e1

Volcanus hinc ma1 rona
sententiarum
miss,
s

u
i

(B)
'

do1 as el

Lntegri

.

ad

<

>rcum oec peredit.
el

B
I

anciliorum
iiir.iluiiii

nominis
'!

e1

tii»;n'.
'

[ove
is

ml
i'e
\

i

Ron
,i\

III.
III. III.
II.

5,

14 21

dissentienl
di
.-ill
i

condicionibus.
\

I

5, 5,
5,

epta
s,.

\

i'li.

i'li
,-t

(l

ium.

43

r< -mi. .\

Us,,

Lrilem.

B
'

56
18

;mt
I

Lacedaemonium Tarentum.
inquinavere
el

II. II.
II. II.

6,

rin urn

genus

e1

domos.

I

17 21, 21,

\
in
Jl

i

e
fce

sermonibus
te

oeglegel horridus.
laeta aderil

I

Liber

el

si

Venus.

III.
III. III. III.
ill.

23
26,
29,

None.
9
3
.->

29, 29, 29,

7
!i

quae beatam diva tenes Cyprum et. cum flore Maecenas n sarum et. Lamdudum apud me est. eripe te morae. declive contempleris an um .).
()
t'.-isti.li.

1

B
I

III.

sam desere eopiam
et
di

et.

rrr.

29,
29,

12

fumum

opes strepitumque Romae.
cte premil

IN.
III.
III.

30 32

caliginosa

deus.

I

29, 29, 29,
1.
'.»

Pas trepidat.

quod adesl memento.
et.
1

40
49

cum
venti

i

era diluA Les quietos.

III.

Portuna saevo laeta negotio
|

IV.

avei

i.

in
i

ii

i

x

in

ovilia.

IV.

I.

22

st

omnia sed
esl

diu.

IV.
IV.
IV. IV.

4.
I.

52 56
72
I

fallere

<-t

effugere

triumphus.

pertulit Ausonias ad urbis.
inniii
is

!.

Hasdrubale Lnterempto.

!).

ne forte credas Lntei itura quae.
imiii
i

IV.
IV. IV. IV. IV. IV.

9,
1

26
1.

s.'.l

omnes inlacrimabiles.
>

I

5
17

aeten

ei

(

qua
in

bo]

babitabilis.

1

1.

Bpectandus
mittere
te

certamine Martio.
medii
a

11.
1

24

.'.11111111

per
«'i
I

I.

33

copias te consilium

I

17.

202

University of California Publications.

[Class.Phil.

Ii

miisi qo1

I"'

snpposed

thai because a verse

is

rare

in

form
as,

il

is

aecessarily crude.
resull

Some

verses are rare

i

1

by design,

tor

instance, mighl

from onomatopoeia; some (2) by chance,
;

the unusual features being purely accidental

some
arise

3

i

by defect.

Like

tlif

cases of interverse hiatus, these
lli«'

lasl

from imma-

turity in

poet's art, intractability of material, or conditions

of composition

some way unfavorable,
In

as

may

be inferred

from

the circumstance thai where such verses
ities

abound other

irregular-

are likely to be found.

I. -".7.

for example, along with the

unduly similar word-arrange

ots of verses 21, 22,

and

23, the

objectionable fifth-space division of verse 23, the absence of the

regular division
outsel of verse
5,

in

verses 5 and 14. the two dissyllables
oilier features

a1

the

and

shown
In

in

the three examples
11

above quoted, we find interverse hiatus after verse
initial

and

a

short

syllable in verses 15

and

22.

111.

4.

along with the
in

monosyllable closing verse 59 and other features shown
seven examples above, we Mud
41
a

the

prosodic irregularity

in

verse

and

five

cases of interverse hiatus.
a

In III, 29. along' with
in

lour verses closing with

monosyllable 'three
at

the

first

three

strophes

.

the two dissyllables
in

the outset of \rrsi'

5,

and the

other features shown

the nine examples above,
elision.
is

we and verses
verse 52, hut

35-36 connected by interverse
in this

It

should be noted also
in

poem

thai the

form of verse 36

repealed

nowhere
only
in

else in
III. 4.

Horace, and the form of verse - occurs elsewhere

<i-~>.

vol..

1

Richardson.

Horace's Alcaic Strophe.

203

TABLE
Horace allowed
st
;i

VII.

short syllable to begin
:

a

verse of the Ai

rophe

in

t

he following eases
9,

[,

1

mi|

University of California Publications.

I'

1

''

hil1

'

Summary

of eases

in

B:
3

Book
Book

I
II

....

Book III Book IV
Total

2 cases

cases

10 cases

UNIVERSITY OF CALIFORNIA PUBLICATIONS
IN

CLASSICAL PHILOLOGY
Vol.
1,

No.

7, pp.

205 262.

June 17, 1907

SOME PHASES OF THE RELATION OF THOUGHT TO VERSE IN PLAUTUS
BY
1

KXRY

W.

PRESCOTT.

In his study of the Saturnian verse Leo has recently stated
his conception of the relation of thought to verse in early Latin

poetry: "

in

early Latin verse," Leo says, with reference espec-

ially to the

Saturnian, "verse and sentence are identical; art-

poetry in

its

beginnings" (and he refers

to

Plautus as

illustra-

tive of the principle),"

when

sentence-structure was developing,
itself to

resisted this inherent
bj

requiremenl and Limited

the

norm
justi-

which words

in

the sentence intimately connected in thoughl

were not separated by the verse unless the separation was
fied

by special considerations: externally, by reason of length, or

by the colligation of words through alliteration or other means
of connection: internally, by reason of emphasis or
effect of

some

stylistic

the

word thus separated."
fallen urspriinglich

1

'" Vers und Satz
dieser

zusammen;

.

.

.

Die Kunstpoesie

hat in ihren Anfangen, wie sick die Satzbildung machtig entwickelte, mit

der Poesie
beschrankt,

umewohnenden
dasa

Norm
(lurch (lurch

den Vers get

Forderung gekampft und Bie auf die im Satze eng zusaniinengehbrige Worter nicht remit werden diirfen, wenn sich nicht die Trennung

einen besonderen

Omstand
ler

Lange, durch allitterirend
des gesonderten Worts.

ala berechtigt erweist; aiisserlicli durcb andere einander suchende und anziehende

Wortverbindungen, Lnnerlich durch Nachdruck oder Bonsl atiliatische AJbaichl So erscheiiit der Gebrauch bei Plautus ausgebildet. ' Abhandl. Gotting. Gesell. (19 Der saturnisehe Vers 14 In 1SS1 Buecheler reminded School! thai only pronominal adj(

=

were aeparated from their uouna by the verse-end, thai almosl ao other (Truculentus, ed. of Plautus adjectives were so treated, in the texl Buecheler repeated this admonition in Schoell, praefatio XLV, n. 1).

206

University of California Publications.

[Class Phil.

Deo has
I

lefl

to

others the task of testing the validity of his law.
to

have attempted

gather and study the evidence offered by one
in

group of examples
w«»rds.

Plautus,

the

cases

in

which adjectival

whether ordinary attributives, pronominal adjectives, or numerals, are separated from their substantives by the verse.
In

many

respects the study

must be

descriptive:

the

Lack

of

similar studies in Greek poetry, and the fragmentary remains of
earlier Latin poetry, usually of uncertain metrical constitution,

retard
ical

a

convincing account of Plautus's position

in the

histor-

development of verse-technique.

Nor

will

it

be just to con

firm or refute Leo's theory until other phases of the problem in

Plautus. and the corresponding
investigated.
of

phenomena in Greek poetry are For the present, the study may suggest points view and methods of approach, which will doubtless need readis

justment as the problem

studied

in

its

larger aspects.

I.

Among
tion
is

the features that Leo enumerates as justifying separa-

length: this element

may

be a matter of syllables, or in

addition to syllables
is,

may

include an extension of thought.

That

a given

word may be
Ion?.

long, or a thought-unit involving several
it

words may he

In either case,
If.

is

not at once clear that

length occasions the separation.
KM..

however, as appears to be
his

.Mus.

41

(1887)
Plauti

312.

In

1893 Appuhn published
inter

dissertation:

Quaestiones
intercedant

Plautinae.

Quae
in

rationes

versus

singulos

sententiasque
Interpretative

exemplo
this

comprobatui
attempt
to

(Marburg).
cover a
large

analysis was impossible

field

within the

compass of a doctor's dissertation. Norden summarizes the usage of Vergil
For
references
to

in

Aeneis Buch AT. 390-391.
£

studies
.if

of

the

general

questi

the

collocation
cf.

of

words, as well as

the special question under consideration,
J

the
n.

same
1.

work 3S2
In

n. 1,

and the same anther's Die antike Kunstprosa
1

68

the present paper the song-measures are excluded;

have not

knew

in^ly
poses,
not

included

examples from such passages except
is

fur

comparative purthe

and then their provenance
with
but

stated.

I

may

!» open to criticism in
verses conthe
tech-

dividing the material
the
results

reference to the metre of no

cerned;
nique
of

shew

important

differences

between

iambic and trochaic verses, or of the shorter and longer verses, except such as may more conveniently be described parenthetically, and a metrical classification interferes with clearness of presentation.
the

Vol.

i
i

Prescott— Thought and
more

Vers*

in

Plautus.

207

the case, 2 words of five or

syllables thai are metrically suit-

able regularly tend to the end of the verse, or less frequently to
the beginning,
it

follows that,

it'

such

a

word

is

a

substantive or

adjective, the difficulties in

combining the two members of the
consisting of

pair in one verse are

much

greater than they otherwise would be.
a

And
will

similarly,

;i

thought-unil

substantive and

several adjectives, wherever they

may

be disposed in the verse,
into

by reason of the number of

syllables, easily overflow

the next Verse.
In a

thorough treatmenl of Leo's theory predicative expresshould
be
included.

sions

The consciousness of verse-unity
in

could not be better illustrated than

these two couplets:

isque hie eompressit virginem adulescentulua
(vi), vimilcntus,

multa aocte,

in

via.
in

(Cist.

L5S

quoin hasce herbas huius modi

suom alvom congerunt
(Ps.

formidulosaa dietu, non essu raodo.

823)

Bui such cases of predicative expressions, involving Long words,
are apart
f<

from our immediate purpose.

There

are,

however, a

w

cases of adjectives following their substantives
is

either adjecpre-

tive or substantive

of greal

Length)
it

and

not

so clearly

dicative.

Their position makes
in

possible thai they amplify the

meaning,
justify

which case
separation.

this

amplifying force as well as Length
almost
cases the adjectives
it

the

Most of these adjectives are derived
in
all
'<\'

from proper nouns; and since
stand
at

the beginning
in
:

the second

verse

is

significanl

to

note that
their

Oscan and I'mhrian proper adjectives usually follow
8

nouns

Philopolemum vivom, salvom
villi

el

sospitem

in

publics

A
-

l.uiii

una

el

ibidemque ilium adulescentulum tuom Stalagmum servom (Capt. S73)
eeloce,

Iti

tin'

Mostellaria,

for

example, out of 90 cases of words of
a1

live

<>r

syllables,

two-thirds

stand

the

end

of

,-i

verse;

of
<

the

remain-

>n the other nil but tun are metrically impossible at the end. hand words of four syllables are freely disposed in the interior of the Five syllables is, therefore, assumed to be the minimum oi verse.

ing third

which
10

may

1"-

regarded as offering

difficulty.

*Nilsson,

de

collocatione

pron.

adj.

apud

Plautum

et

Terentium

= Lunds

Qniversitets Ars-skrifl 37

(1901).

208
ihin

University of California Publications.
ego te ad ilium duco dental um Lrum it Macedoniensem, qui te nunc flnitcm
\ t'.-n

[Class Phil.

:

(I's.

lolu)

quern

propter,

o

mea

\

Lta

.'

propter

militem

Babyloniensem, qui quasi uxorem sibi (True. 391)
Bed
illi

patruo huius qui

\i\ii

aenex

Carthaginiensi duae Euere

filiae:

(Poen. 83)

These examples are of somewhal differenl
case, tlic length of

value.

in

the first

adulescentulum and

its

consequent position
at the

(of fifteen occurrences of the

word thirteen are

end of the

verse) are the controlling factors:

than in

vs.

169 of the
(intus)
i

Aleum is no more amplifying same play (nam cecum hie capiivom
\

adulescentem

Aleum,

summis
its

ditiis

where the adjective

prognatum genert summo ct is kept in the same verse with
alike in

shorter noun.

The next two examples are
In the last example,

having the

separated adjective followed by the caesural pause and an ex-

planatory gm'-clause. 4
caesura]

too.

we have the
to

pause.

Plant ine usage of these adjectives points

Length as the influential factor.

Carthaginiensis occurs only at

the beginning of the verse (Poen. 59, 81, 963, 997, 1377) with

one exception

(1121).

Babyloniensis

is

less

constant: at the
(here,
;

beginning

in

True. 81, penultimate

word

in True. 203

however, iambic septenarius; in the other cases, senarii)
three eases the

in all

same phrase occurs.
|

So we get militem

|

Baby
are

loniensem (391),
miles
|

Babyloniensem militem (81). Babyloniensis

(203).

It is clear that length
ilmiii usis

and metrical conditions and
in
all

potent.
its

Muci

does not occur again: Macedonius takes
the

place

(Ps. 51, 316, 616, 1090, 1152, 1162),
it

cases except one (316)

stands

at

the end. differenl
position;
in

metrical
all

constitution

making

it

convenient in thai
is

the

cases of Macedonius. however, separation

avoided except in the

following couplet:
1

Cf.

'I'

rue.

83:
sil>i

quern antehac odiosum

esse

memorabat mala,

Babyloniense
venturua prregre:
here the adjective
scription.
is

ilitem:

is

mine dicitur

not

separated, and

a

demonstrative

resumes the de-

For relative clauses denning separated adjectives cf. Seymour, Earv. Stud. Ill (1892) 98 If., and for explanatory clauses after a separated
:

r.it

ive

in

Plaut US Cf.

Im'Iiiu

.

p.

-')-.

Vol.i]

Prescott.

— Thought
(Ps.
1

and

Verst

in

Plautus.

209

Pseudolus tuos allegavii hunc, quasi
milite esset.

a

Macedonio

L62)

In this case ihc adjective precedes,

and the unity of thoughl
noun,
is
it

is

seriously affected.

Such

a

case strengthens our feeling thai
its
it

in

the examples

in

which the adjective follows
is

not so
is

much

the amplifying force, which

difficull to

prove, as

the

length thai conduces to separation.
In a

few cases of ordinary attributives, however, the thought,

quite as

much

as the Length, justifies the separation:
milia

quom sexaginta
TOlaticorum

bominum
oecidi

lino

die

manibua

meis.

(Poen.

472)
1

The swaggering antithesis of G0,000 and a single day" occupies the first verse, and crowds out volaticorum; but this adjective is in itself of a length that makes it most adaptable to the extremes
of the verse

—so in the conversation
volaticorum

that
dico

follows our passage:

—an,
Plautus
is

opsecro,

hominum? usquam

ita

sunt

quidem. homines volatici?

no slave to such external conditions, however, for the
its

adjective by

separation and prominence produces the climax

of surprising absurdity after the antithesis of the preceding Nor is it far-fetched to suggest that the juxtaposition of verse.

" hands," volaticorum and manibus, " wings " and
dental.

is

not acci-

In both of the following cases the

rest

of the second

verse
tive,
a

is

which stands
:

an explanation of the separated adjective or substanat the beginning of the second verse before

Strong pause

ut in oeellis hilaritudo est, heia, corpus cuius modi,

subvolturium
novi,

illud
ita

quidem,
solct,

Bubaquilum

volui
7

dicere.

(Kud.

121)

Neptunus
si

aedilis est;
6 6

quamvis t'asthliusus quae improbae sunt merces, Lactal omnis.

(Hud. 372)

Cf Aul.
.

70, Aul. frag. 3.

Leo. Analecta

Plautina:
volui.

de figuris sermonis
F<>r p.

II

31, refers

t<>

the wordof a

play in

subvoiturium
cf.
is

a

Blightly

different

explanation

separated adjective
hi

below,

l'24.
.

More
.

like

our presenl

example, but
the

with a play mi verbs,

Fri\ "la ria

frag. s

tl

nly other occurrence of the adjective, fastidiosua

\*

in

same

position

(M. G. 1233).

210

University

oj

California Publications.

[Class Phil.

There may be

a

difference of opinion in the interpretation of the

second example: perhaps the second verse explains fastidiosus
rather than aedilis.
prise and, as in the
ot

But
first

in

any case

aedilis

comes

in

;is

;i

Bur-

example, the separation and

tin-

position

the unexpected idea enhance the effect.
It
is,

of course, true that the separation seems more violent in
first

the second case than in the

because the adjective precedes

Similarly in these examples:
quo niodo
fidicina?
rue Indus ti-eist
i

do

ilia

conduct
recte

icia

— factum
LUunc,

hercle vero, et

factum

iudico.

(Ep. 706)

volo deludi

dum cum

hac usuraria

uxore nunc mini morigero.

(Ainph. 980)

In both of these the long' prepositional phrase, quite

apart from

the long adjective, makes separation almost inevitable. 8
a preposition the accusative case fidicinam

—conducticiam

Without
is

ac-

commodated
entire verse

in

a

single

verse in

Ep. 313: whereas the same

phrase with uxoraria escapes separation only by occupying an

cum Alcumena uxore

usuraria.

(Ainph.

498)

The
cia

significant fact

is

that in all the few occurrences of conductiat the
is

and usuraria the adjectives stand

end of the verse

(Cure. 382, True. 72).

The same

position

the regular habitat

of praesentarius, so that the following separation

may

in Large

measure be referred

to the length of the adjective:

vendidit tuos natus aedis.

argenti minis numeratis.

— — praesentariia —quot? — quadragiuta.
perii.

oeeidi.

(Trill.

1081)

(For other cases of
Poen. 705, 793.)

this adjective at

the end.
in

Most. 361, 913,
the
first

The explosive

alliteration

verse

may, from Leo's standpoint, partially reestablish the unity of
that verse; indeed, from an English point of view the idea "cash

down"
b

is
si

a

separable

idea.''

but

we may

not safely attribute

it

prm The
The

a In rins.

fact

that argenti minis constitutes an almost inseparable
Ep.
p.
7« >7
\.

alliteration in

is

also to be note.

I.

f.

mutuos, below,

23

Vol.

i|

Prescott.

—Thought
to

and

Verst

in

Plautus.

-11

unit

(usually

al

the end or beginning of

a

verse)

adds

to

the

difficulty.

This brings us

examples of long thought-units
a

Such thought-units may be
by
rascally, devil of

of two sorts:

substantive attended
e.

a succession of adjectives of equal value,
,-i

g.

"a

Long, lean.

fellow "
<>l'

;

<»f

a

substantive
value,
e.

a

impanied by
twin
richer

attributive
sister.

modifiers

unequal

g.

"my own
a

"

Our author
first
it

is

fond of billingsgate, and offers

store of the

variety of

compounds than we may

quote.

In
is

general

may

he said that such a succession of adjectives

usually so disposed as to accentuate the unity of the verses: the

substantive usually precedes or
attributives; the thought
is

is

embraced between groups of

measure complete, and the virtues or vices or indifferent qualities either run over into several
in a

verses or occasionally are

bound within

a single verse, in either

case without serious disturbance of verse-unity.
will illustrate these characteristics:

A

few examples

nisi inihi supplicium virgeum (MSS. virgarum) de te datur (M. <;. 502) longum, diutinumque, a mane ad resperum.
si.-it

propter virum

fortem atquc fortunatum

et

forma

regia.

(M.

<;.

'.'.

ef.

56

ceijUOm

recalvom ad Silanum senem, statutum, ventriosum, tortia superciliis, contracta fronte, fraudulentum, deorum odium atque hominum. malum, mali viti probrique plenum, (Kud. 316) qui duceret muliereulas duas secum satis venustasl

For other examples, Bacch. 280 Cas. 767, Men. 402, 487, M. G.
True. 287.
be noticed: in the
in
first,

(if

Leo's strigosum

is

accepted
12:..

.

88, Ps. 724. 974,

Rud.

313,
will

In the examples quoted other obvious features
intensification of one idea
in

one verse;
a
is

the second, initial rhyme.

There are
in

a

few case. n(

succesqo1
SO

sion of

two or three adjectives

which the unity

obvious
ut aliquem hominem strenuom Ps benevolentem adducerem ad t>-.
i

post

altrinsecust Becuricula
ibi

ancipes, itidem aurea
in

litterata:

matria aomen
11-

Beeuriculast. 1*

(Rud.

r

I

i.

Bud. 478, L156

212

University of California Publications.
nunc stafuam
dare auream auro Philippo, (Cure. 439)
!

[Clasb Phil.

ibi

voili

Bolidam" faciundam

<'.\

Tn

all

of these the
s<»

noun and one adjective
is

or two) stand

m

the

firsl

verse

thai the though!

practically complete; benevolen-

tem, and aurea

(as

we

shall see presently), are metrically con-

venienl in the places which they occupy; the separated adjectives
all

stand

;i1

the beginnings of their respective verses
it

and are
is

not without emphasis;

is

also to be noticed

thai

litterata

explained

in

the rest of the verse.

Of

the second variety of thought-units, two occur with sufficient

frequency to be of significance.

These are the expressions for

" own twin sister, brother, son," often accompanied by a pleonastic numeral when the expression is in the plural; and the
phrase Tor a sum of money
in

which

nummi
'Phis

aurei Philippi

in

various arrangements, with an accompanying numeral or further
attribute,

makes an elaborate complex.

latter

phrase
12
:

is

usually from eight to thirteen syllables in extent, and on five
occasions the longer varieties run over into a second verse
sunt
trecenti
tilii
i

lit

us aurei

nummi

Philippi?

— descent

i

quoque.

(Poen.

165)

qui milii mille

numimim

crederet

PhUippum,

(Trim 954)
illius
in

atque etiam Philippum, numeratum
mille

mensa manu,

nummum.
nummi

(Trin. 965)
hie sunt numerati aurei

trecenti

qui

vocantur Philippei.

(Poen. 713)

nam

ducentis aureia

Pbilippis redemi vitam ex flagitio tuam.

(Bacch.

L010)

On

the contrary,

in

a large majority of cases similar varieties
not always with aureus, are included in a

of the

same phrase,

single ve.se: As. 153, Bacch. 230, 590, 882, 934, 1026, Poen. 670,
13 732, T.in. 152, 959, 1158.

"
'I'lif

pmximity of faciundam gives solidam predicative force
cf. Cicero,

in

our

passage:

de div.

I,

24, 48.
I

In I'crs. CIS r ,,,!,, i, iiiik rati arc proliaMy amplifying, as out in his punctuation: <-f. I'crs. 526.

...»

brings

h
Kritik

ia

separation
u.

thai in any .T these phrases there was any violent. for the usage of the various forms Langen, Beitriige zur Erklarung des Plautus 85 11'., Brix on Trin. S44). At least in
nut
likely

(cf.

Vol. l]

PrescotL— Thought and
a

Vers*

in

Plautus.

213

There are
than
six

do/en instances of the

firsl

phrase, including more
into
;i

syllables,

and of these only two escape

second

verse; these two are of eleven and ten syllables:

geminam germanarc meam
hie

sororem esse indaudivi: earn veni quaesitum.
est

(M.
duos

(!.

441)

spea milii

vos inventurum fratres germs
el

b

geminos, una matre uatos

patre ono uno die.
first
14
)

(Men. 1102)
is

The second of these (and possibly the
separation: geminos
sizes the idea
is

only apparenl

followed by

a

sense-pause which empha-

as amplifying,

and the elaboration of the same
gives
in
a

idea in the rest of the
verse.

same verse
is

distinct

unity to thai
a

Indeed, geminus

elsewhere

the

same play
is

sub-

stantive: .Men. 26, 4o. 68, 69,

and

if

the prologue

of dubious
vs.

authorship

in

parts, at least once in the

play

itself,

1120.

In nine cases long- forms of this complex are confined
verse:

to a single

Amph.

480.

cf.

1070, Men. 18, 232, 1082,
sure,

112."..

M. G. 238,

383, 391, 717.

To be

our impression that
is

this situation

points to a sensitiveness to verse-unity

momentarily disturbed

when we
over

find a

much

shorter Eorm of the same phrase running
sicul

soror
earn.

eius hue

gemina

venil

Ephesum
for

el

mater accersuntque

(M. G. 974

Only momentarily,
Palaestrio

again

gemina
the

may
soldier

be

substantival;
tactfully,

may

be

working upon

very

the separation of

nummus

Philippus, the use of

Philippus alone, and the

examples above (Trin. 954, 965, with qui vocantur Philippei in Poen. 711 i, ther. suggest that the wonts are separable, either one amplifying tl When aureus (convenient at the verse-end, cf. above and As. 153, Bacch. 230, 590, 934, Trin. 1139) is a part of the phrase, the separation seems more violent; if, however, Bentley's emendation of Bacch. 230 is right, there would be some evidence of a substantival aureus, similar to the usage
as a and one should compare the usage of in the fragments of Greek comedy: Jacobi. comicae dictionis The separation of aureus is do more than that of a material genitive as in Hipponaz, 22, 1:

of later times;

-

i

i

without

btclttip

ku\
-

can
-

ia'/iGKd KaonepioKQ
-

But Plautus does not separate the genitive auri in this phrase. 14 The resumptive earn in the same verse with sororem may help
-then the unity of the verse.

to

•jij

University of California Publications.

[Class Phil.

slowly unloading his ammunition,

";i

sister,

her twin."

(So,

perhaps, also

in

vss.

47:<

474.)

And

Leo mighl add thai the
first

alliteration in sicut soror reasserts the unity of the

verse.

15

II.

In so

far as he

overcomes the obstacle offered by length

in

a

large majority of cases, Plautus
the integrity of the verse.

may
is
I.

be said to show respect for
to separation

But the poet's aversion
est

or his indifference to verse-unity

which there are no obstacles

in

by conditions in the length of words or thoughts.
tested

Some general
amples.
In

considerations will help us to appreciate the ex-

the later Republican prose the substantive
its

is

often sepa-

rated from

by intervening words, and much more frequently in poetry, so far as I know, no effort has been made to discover whether such separation is regulated by any laws or not" whether, for example, certain attributives are more separable than others, whether the intervening words are of some
attributive

special character, etc.

such separation in early Latin prose

Xorden 17 has already pointed out that is, as regards the number
to limitations.

and the nature of the intervening words, subject

Altenburg 18 has collected the material: usually only one word From intervenes, or if more, they constitute a unit of thought. our present point of view we should like to know whether the
attributives themselves
a

show degrees of separability: whether.

the genii

units should come Ep. 559, in which and t lie adjectives constitute an inseparable compound and perhaps account for the escape of mulierem:

Under the head of long thought
i\.-

accipe,

aerumnosam

et

miseriarum

eompotem

mulierem

retines.

The same would apply to Nonius's reading aerumnarum, 18 Even the interpretation of the material under discussion
would
lie

in

this

paper

study of the collocation of adjective and substantive within the verse, quite apart from the question of separation by
facilitated

by

a

the verse.
Die antike Kunstprosa
18
I

179-180, and 180

n. 2.
II

De sermone

pedestri Etalorum vetusl issimo

=J

B.

Supp ll»d.

24 (1898)

523

ff.

Vol.

i|

Prescott— Thought and
of

Vers<

in

Plautus.

215

for

example, the separation
not

certain
thai

pronominal
ordinary

adjectives

does

appear earlier than
is

of

attributives.

Perhaps the material
alization;
tiif

too scanty to lead to convincing generis

fact

that in Oscan the relative adjective

very

regularly separated

from

its

noun and stands

a1

the opposite

extreme of the clause lends significance
Plautus. 18

to a similar situation in

Such observations as Kaibel makes in his study of 20 would affect our interpretaAristotle's Athenian Constitution tion of many examples if early Latin prose showed similar characteristics:

he

notes

that

certain

pronominal

adjectives

are

separated from their substantives with greater frequency and

by more intervening words than ordinary attributives;
tions in the order of such
relative, too-ovtos,
00-0?,

lie

men-

frequency
uvtos,

o5tos,
.•

7m?,

0A05,
last

uAAoi, the

ov&is,

tU

but the

seven are
re-

naturally represented only by one or two examples; he also
fers
to

numerals,

but

without

separation in such cases.
of the
21

mentioning the frequency of Altenburg's examples show that some
in

corresponding words

Latin
that

are separated
in

in

early

prose.

When we add

thereto

Plautus,

quite

aparl

from the question
show
a

of separation within the verse, the cases of

separation by the verse and, often, by intervening words as well,
relatively

large

number

of pronominal adjectives and

numerals, we

may

suspect that some influence
less violent
it

made

the disturb-

ance of verse-unity either
it

or more imperative than
in the case of

appears

to us

and than

perhaps was

ordinary

attributives: in

Plautus 20 per cent, of the cases of separation

sive adjectives.

by the verse-end are pronominal adjectives. 25 per cent, possesThat is, more than half 5 per cent, numerals.
1

arc pronominal words and numerals.
lena A step towards the explanation of some of these phei has been taken by Wackernagel, 22 though without reference to

the mattei- of verse-unity.

His investigations

in

[ndogermanic
Ugh1 survivals

Languages, especially Greek and Latin, bring
"Altenburg,

:i
1.

to

c.

."">:ii>

;

Norden,
\

1.

c.
.

[

181

n. 1.

Stil u.

Text

<1<t

Ho/

dea Axistoteles 99ff.

For example,

ceteri,
I.

mums, numerals

including

nullus,

alter,

tantus,

qui (rel.), quis (indef.

B Indog. Forsch.

I

t06ff. Cf.

Delbriick, Syntakt.

Forsch. Ill

17.

216

University

oj

California Publications.

[Class Phil.

of
a

a

law by which shorl enclitic words tend

to the

beginning of

sentence, usually to the second place.
enclitics,

Pronominal words are
adjectives
are
directly

ofteu

and

some

pronominal

affected by this law.
of
will

Others are indirectly affected; for the law

pronominal

attraction,

combined with

Wackernagel 's law,

sometimes bring pronominal words thai

may

or

may

not be

enclitics to at least the third place in the sentence.

Such laws
to its

have precedence of the natural attraction of the adjective
substantive.

A few other laws
tally that
flicts.

affect the collocation of

verse-unity must
of the

waive

its

claims,

words so fundamenwhenever it conto oik
1

Words

same category are attracted
in

another.
e.

Certain formulas exist for the expression of certain ideas,
of oaths.

g.,

Groups of words

Plautus have been studied and

peculiarities of collocation discovered.

refled
ficial

the usage of ordinary speech.

Most of these conditions But there are other artiin

combinations

not
of

we may not always say

—whether due the influence of rhetoric or —resulting often the interlocking
to

words and the consequent separation of words that are synAll such factors must he appreciated. tactically connected.
Apparent violation of verse-unity may he only conservation of
these natural or artificial collocations. 23

Some

of these general considerations account for the separate

treatment of ordinary attributives, possessive adjectives, other

pronominal adjectives, and numerals.

All of them will

make

more

intelligible the discussion of individual passages.

In this discussion I do not wish to he understood as representing the attendant features to be the cause of separation or atone-

ment for separation;
tion.

that

would he begging an important questo

In viewing the problem of verse-unity with reference
it is

Leo's theory,

apparent that the cases of separation are often
to be justifications

attended by such features as Leo regards
-3

for

On the various matters here briefly referred to cf. Langen, Rh. Mus. Stude12 (1857) 426 ff.; Kellerhof, de collocatione verborum Plautina mund-Stud. II 49 ff.; Kampf, de prononiinum personalium usu et collo-

=

catione ap. poet, scaen.

Rom. 16

ff.

= Berliner
133;

Studien III
u.

(1886); Leo,

Bemerkungen

iiber

plautinische

Wortstellungen
132

Wortgruppen

= Nach-

richt. Gotting. Gesell.

(1895) 416,

Norden, A.eneis Buch VI. 386.

Vol.

i|

Prescott.

— Thought

and

Versi

in

Plautus.

217

separation: a descriptive paper notes
features.

tin-

a

ppearance of such
treatment
is

Quite apart from
question

this

descriptive

the
is

important

which

Leo's

theory

involves,

namely:

Plautus, under the influence of earlier Latin
Of verse-unity in the sense that
justified
all

poetry, conscious
lie

cases of separation must

by special considerations?
is

Granting that

these

fea-

tures attend separation, there
or
all

the further question:

may any
in

of these be proved to be necessarily involved in the relation

of thought to verse?

For example,
not
its

alliteration

is

inherent
to

Plautus's style:
verse-unity?- 4

may

appearance have nothing

do with

Furthermore, granting that Plautus

is

conscious
denied,

of the individuality of each verse, which

may hardly be
ways:
a

such consciousness

may

arise in one of several

port

may

be under the influence of a primitive form of verse in which 's theory; or verse and sentence are identical so Plautus in

1

he

may

be far removed from any such influence and yet pre-

serve the unity of the
of entirely even in

verse— which is not necessarily Los1 advanced stages of verse-development

sighl
eithei

for the purpose of bringing into relief units of thought, or a- a

On a priori concession to an artificial tendency of bis time.-' grounds Plautus's attitude towards verse-unity may well be suspected of being affected by the Saturnian verse; he is, however, adapting Greek comedies, and the verse-technique of his Greek
sources

had reached

Latin verse.

a much higher point than contemporary This counter-influence must be reckoned with in

any

a priori reasoning.

Leo would be the

first

to

recognize the

validity of this contention.

None of these important questions is begged in the following Some of them may be considered by way descriptive treatment.
of conclusion, but

many

of

a few phases of verse-unity.

them cannot l.e settled in a study of The division of adjectives is but a
is

small part of word-division, ami word-division

hut

a

part of

"Of course the fact thai alliterative groups are usually limited to a single verse in itself shows a consciousness of verse-unity. The question the purpose of at issue is whether a noun or adjective is separated for bringing it int.. an alliterative group. 3uch an artificial preservation of unity appeal- in Bion: cf. Wilamowit/..

Adonis 38

39.

218

University

oj

California Publications.

[Class Phil.

a

Larger topic which includes the division of the larger units of

thought, phrases and clauses.
III.

When
thai

;in

attributive follows
is

its

substantive

it

is

often possible

the adjective

amplifying; each case must be interpreted

with reference to the context, hut the mere possibility justifies
us
in

distinguishing between (a) adjectives that follow, and (6)
that

those

precede

their

substantives.

Further

classification

mighl he desirable, for example, with reference to whether or
hut this not words intervene between the adjective and nonn would confuse the discussion. I have persuaded myself from
:

an inspection of the Mostellaria that the
of the

number and

the nature

words that intervene between adjective and noun within the verse are the same in the corresponding situation when a In some cases it may well be argued verse-end also intervenes. that verse-unity was sacrificed to the normal collocation of The equally important question whether within the words.
verse the collocation of adjective

and noun and intervening
is

words

is

ever abnormal for the sake of preserving verse-unity

not within the limits of this paper.

(a)
It is

not easy to

draw

a line

between purely predicative and
as
a

amplifying adjectives.

The former,
set
oft'

we saw

in

examples of

long adjectives, are often
participial
is

in

separate verse;

many

are

ex Be lmnc reliquit qui hie nunc habitat filium

pariter

moratum

ut

pater avosque huius fuit.
litteris

(Aul. 21)

cm- Lnclementer dicis lepidia

lepidis tabellis lepida eonscriptis
\

manuf

(Ps. 27)

ilicus

is

cum corona, candide
(Cas.

vestitus, lautus, exornatusque ambulat.

767)

Somewhal

differenl

in

effect,

bu1

equally separable are these

participial adjectives
miles lenoni
Ballion: epistulam

conscriptam mittil

Polymachaeroplagides, (Ps. 998)

Vol.

i
i

Prescott— Thought and

Verst

in

Plautus.

219

hominem cum ornamentia on nibus
exornatum adducite ad me
et
tarn

ad trapezitam Aeschinum.

(Ps.

756)

ornatam adduce lepide

in

gnatam tuam peregrinum modum.
tu

(Pera 157)

effect of

"Writes and scuds." "dress ap and bring" may suggesl the such separation. Such examples, in which the verbal
is
it

clement
T

prominent, are hardly within the scope of tins paper. 28
that the following group of eases will nol he regarded
:

take

as illustrating real separation; predicativ please,

amplifying
distress

as
is

you
an

the suggestion of

physical

or emotional

afterthought, which separation by the verse-end and intervening

words, and position in close connection with caesura or diaeresis accentuate
item parasiti rebus prolatis latent
in

occulto miseri, victitant suco suo.

(Capt. 82)

eeastor lege dura vivont mulieres

multoque iniquiore miserae quam
itaque
ih>s

viri.

(Merc. 817)

ventisque fluctibusque

Lactatae exemplis plurumis miserae21 perpetuam noctem;
ill;i

(Bud.

autem virgo atque

altera itidem ancillula
in

de navi timidae desuluerunt
ibi

scapbam.

(Bud. 74)

me

uescio quis arripit

timidam atque pavidam, aec vivam aec mortuam.

(Cure. 648)

A

similar pathetic effeel

is

evident in

video sedentis
-•

in

mulierculas scapba sulas duas.
in

(Bud. L62)

Nor

present participles as

nam
te
-T

istaee
-lie

quae

tilii

renuntiantur, (ilium
I'-.

\

amantem argento circumducere,
a

s.i,

preceding

pronoun,

in

a

lyrical

i

text:
in

Bed muliebri
mill i

animo sum tamen: miserae (quom venit)

mentem

mortis, metu

membra

occupat.

(Bud. 685)
couplet

Note the alliteration carried through the Another example, of misera following a pr
pol

with

pathetic

effect.

un:

me quidem
i

miseram odio enicavit.

As.

920

I

220

University of California Publications.

[Class Phil.

\'<>r

will
:

there be any doubl

thai

these adjectives are inde-

pendent

nunc equoa iunctos Lubes
capere

me
in

Lndomitos, ferocis, (Men. 862)

Conviva commodus
in

M. G. 642 does

riot

prevent the same

adjective from becoming an amplifying expression with the
in miii

same

eonvivas volo
reperire nobis

commodos, qui una

sient.

(Poen. 615)

tence after

Here the noun and adjective appear at the extremes of the senand before pauses. 28 In the following case the conis

text shows that frigidam
thesis in

predicative; calefieri finds

its

anti

adponi frigidam:
calefieri iussi reliquias

ius est

peruana quidem adponi frigidam postridie. (Pers.

10.1)

"Served up cold" is clearly the idea. 29 Nor may I admit as indubitable cases of real separation such substantival adjectives as virgo and posticum:
eiua eupio filiam

virginem mihi despouderi.
est

(Aul. 172)

etiam hk* ostium
(St.

aliud posticum nostrarum harunc aedium:

449)

Filiola virgo (Rud. 39)

and virginem gnatam suam (Trin. 113)
first adjective,
filia

may

support the adjectival force of the
in

but in

any case the separation

our passage defines

and

contrasts
of

the daughter of Euclio with the middle-aged

woman
is

Megaclearly
con-

dorus's previous remarks (162). 30
28

As

for posticum,

it is

The adjective molestum
ci

in

the

following verses

more

closely

nected with the infinitive:

Lmpudicum et impudentem hominem addecel molestum ultro advenire ad alienam domum, (Rud. 115)
will not take luculcntum

Ami one
tive
I

(luculentc P)

as anything but

predica-

Ep. 158) after comparing vs. 341 of the same play.
t.

memini: ut muraena et conger ae calefierent: nam nimio melius oppectuntur frigida. (Pers. 110)
30

So, but

with clearly expressed contrast
civis
is
.lis

Ln

the second

verse,

the com-

pound virgo

ided in

an paulum hoc esse til>i videtur, virginem (Ter. Eun. 857) conservam esse credidi. vitiare civemf

Vol.

i|

Prescott.

— Thought
if

and

Versi

in

Plautus.

221

a substantive in

Most. 931, and so
not

its

diminutive
it

in

Trin. 194,

1085;

in

the Stichus,

an appositive,

defines ostium.

The separation
In

of aliud docs not here concern us.
in

connection with substantival adjectives another passage
is

thr Aulularia

to be considered
[ui

namque
dotatae,

I

dicat: quo illae nubeni divites
ius

si

istud

pauperibus

poniturf

(Aul.

189)

The contrasl between

divites
it

and pauperes suggests

thai the for-

mer
is

is

substantival: but

does not at once follow thai dotatai
vss.

purely adjectival.

For

534-5 of the same play show how

easily the participial adjective becomes substantival

nam quae

indotata
et

est,

ea in potestate esl
et

viii

;

dotatae mactant

malo

damiio

viros.

Similarly Ter. Phor. 938, 940.

II'.

however,

it

is

adjectival in
as facti

our passage,

it

adds

to

and explains

divites very

much

osum

in
in

venit hoc mini, Megadore,

factiosum,

me autem

esse

mentem ted esse hominem divitem hominem pauperum pauperrimum.

(Aul.

226)

In

both passages we have the contrasl between rich and poor,
in

and

factiosum as

in
a

dotatai

the

happy

isolation a1 the begin-

ning of the verses of

more

specific attribute of the rich class:

in each ease the emphasis is accentuated by the scum '-pause which follows the separated adjective. From a differenl point

of view

hominem divitem factiosum should benevolentem (Ps. 697, hominem strenuom
Mos1
of

be

compared with
p.

above,

211).
into

such amplifying

ideas
;it

are
the

similarly

broughl
s<

prominence by their position

beginning of the

nd

verse; often they are followed by a decided sense-pause; some-

times this separation brings them into the vicinity of contrasted
31

The verse Lmmediatly following
partem, but
this

in
in

the Stichus

(450a)
division
is

contains pos-

ticam
4.">i

verse
.-t'.

is

aot

A. and
it'

the

in

B
to

is

suspicious:
it

l.m mi

/..,.

vs.

150a

of 450a and genuine, as Lindsay

seems
32

regard

in

his

Oxford

text,

some support.

Cf. Pauli Pestus, 220 276 de Ponor. In a similar context Menander (585 K.) has a similar separation:

M=

a

purely

adjectival

foro

_

.

,

-

222
83

University of California Publications.

[Class Phil.

ideas.

All

of these

features,

with attendanl
te,

alliteration,

are

illustrated in
ego
Ealse,

Philoerates

faciam

ul

verus hodie reperiare Tyndaras.
a

(Capt. 609)
is
i

The separation of an adjective from
ranged,
p. luit in

vocative

similarly arto

here

in

a

succession
Quid

of epithets
]nmin

referred

on

211

i,

:iis,

levior

quam pluma, pessun

t

aequissume,

flagitium hominis, subdole ac minimi pretif

(Men.

1^7

i

The surprise of the opprobrious epithet is made more effective by separation and prominenl position. The element of surprise, which false and levior, like subvolturium and volaticorum

among

the long adjectives, illustrate, recurs in another example
a

of the vocative; the parasite greets his patron as

veritable god

on earth
terrestris,

te

o mi Iuppiter eoepulonus eompellat tuos.

(Pers. 99)

Without the element of surprise and without
pause, but,
I

so distinct a sense-

think, with emphasis paterni

is

separated

in

nonne arbitraris eum aduleseentem anuli paterni signum novisse. (Trin. 789)

So

in

Poem 10S0
position,
is

the

same adjective stands with emphasis
in
ea
.108)

in the

same

though not separated.
heightened by alliteration" 4

Contrast

quodque concubinam erilem insimulare ausus probri pudicam meque summi flagiti, (M. G.

and here prominenl position
the adjective, thai
For contrasted
of an adjective
cf.

is

given to the crime rather than

the two crimes
into
:

may occupy
the same
verse

the extremes of
by the separation

Ideas brought

Caecilius 22]

R

egon vitam tneam

Atticam contendam cum
unless
it

istac

rusticana

(tua), Syra?

is

an octonarius, as

<

'.

P.

W. Muller supposea
. .

wrbanatvm brings oul the font rust more plainly: ef. rusticatim ponius 7 l; (Leo, Analects Plautina: de figuris sermonis II 32).
.

Bergk's asticam in Pompartitudo

Cf.

propt

.

.

prdbrum, propinqua partitudo (Aul. 75), probrum palam Aul. 276).
.

.

.

.

I

Vo]

i
|

Prescott.

Thought and

Versi

in

Flautus.

the verse

meque.
separati

Contrasl
E

and the two abused innocents be juxtaposed in pudicam and comprehensiveness are obtained in tins

dexU ram:
age push stende manum laevam ostende. quin equidem ambas profero.
I
i

dexteram.— em.
649)

— nunc

\ni.

Somewhal

differenl

is

the collocation
iii

in

nixus laevo

femine babel
203)

laevam manum,
ferit

dextera

digitis
|

rationem
•;.

computat,

femur

dexterum.

M.

Here the contrasted

parts occupy

different

verses;

dexterum

echoes dextera of the preceding verse, 88 and the actor's gestures

doubtless contributed to the effect; the alliterative features arc
plain,

whether or

no1

pari of the poet's intention in separating

the adjective.

An
ative;

adjective expressive of size
in this

is

naturally Liable to separation
is

and prominence; 86

example maxumi

practically prediceffect:

number and

size are

postponed with dramatic
in in

postquam devolanl anguea iubati deorsum

curias condit usl

Lmpluvium duo
capita.

maxumi: continue

extollunl

ambo

(Amph. H07)

Essentially attributive, bu1

in

effective juxtaposition, the
to verse-unity in

same

adjective

is

postponed with more injury

sumne probus, sum lepidus civis, qui Atticam hodie civitatem maxumam maiorem feci atque auxi civi feminal (Pers. 17)
i

The postponement

of the verb

makes the thought

less

complete,

but the alliterative juxtaposition 87 of the superlative ami comparative more than compensates for the separation.

When
is

the

verb comes

in

the

firsl

verse, the adjective escapes into the second
in

verse with less violence to unity, ami
Cf. usqiu
.

this

example

brought
\-.

.

.

>is<in<

...

faciebatis
. . .

.

.

.

fugiebatis ...
. .

210-

213);
(Aul.

iuMtfi

(As.

i-i

126);

deam

</-"/,i

.

(As. 781
1-1
i

TSi

Hi

L15);
1
1 ;

itaqiu

(Cist.

513-515);
.

/•"•<
<

(Merc.
~<7\

125);
perqin

egotnet

(Merc. 852 85
il-

ferreas, ferream, ferreas
. .

\'<r<.

~<7'.'.
;

120), pater

.

pater

.

.

<

Poen. L260 1261).
390.
<;.

"Cf. Norden, Aeneia Bueh VI.
'
l

f.

Can.

1

;.

Amph.

:»\. Capt. L034, M.

1218, Bud. 71. St. 739.

224

University of California Publications.

[Class Phil.

into associations of thoughl
a

and sound

thai give the second verse

unity of

its

own

:

nulla Lgitur dicat: equidem dotem ad te adtuli (Aul. 498) erat pecunia. maiorem multo quam
t
i I
>

I

So with elaborated emphasis on
verum
nunc
si

size:

qua

mi

obtigerit

hereditas

magna atque

luculenta,"
is

(True. 344)
to the

A
in

necessary specification
opinor,

added

noun

in

quam

ex

me

ul

unam faciam

litteram

Ion (gam.

me)um

laqueo collum quando obstrinxero.

(Aul.

77

Alliterative possibilities

may have

helped attract the adjective
litteram

into the neighborhood of laqueo; the alliteration in

longam

is

merely an unavoidable accident.

This prominent position, combined with a sense-pause, some40 expressed in th^ times introduces an elaboration of the idea

separated adjective: so
si

in

the elaboration of a joke:
hodie digito tetigerit
istia

hercle

illic

illas

invitas,

ni

istunc

invitassitis

(Rud.

810)

or with further explanation of the idea as in the

examples quoted
372.
is

above

(p.

211) in Rud. 1158, and (p. 209)
in

-421.

In two examples
at the

which the long adjective inhonestus

se1

beginning 41 of the verse the amplifying idea occupies the
:

entire second verse with predicative effed

nunc hi'- occepit quaestum nunc fili gratia Lnhonestum e1 maxime alienum ingenio buo.

(Capt. 98)
luculenta

•Note the balance between
atqut

magna

atqui

(345)

and dula

amarum

(346).

According to the reading of the MSS. Bacch. 279 belongs here:
ego lembum conspieor longum Btrigorem maleficum exornarier.
I'.ut

strigorem

is

dubious.

STorden, Aeneis

Buch VI, 391.
Eun.
357.

stands in the same position in Ter. For the occupation of the entire second verse cf. Trim 750:

"The same

adjective

S ed

ut

ego nunc adulescenti thensaurum indicem
ai
'is

indomito, pleno

ac

lasch iae

.'

vol.]]

Prescott— Thought and

Versi

in

Plautus.

verum quom multos multa admisse acceperim
Lnhonesta

propter

amorem atque
in

aliena

a

bonis:

i

M.

<:.

1287)°
off a1

A few

cases remain

which the added

ideas,

se1

or

aear the beginning of the second verse, are rather conspicuously linked by alliteration to aeighboring words in the same verse;

some such cases have been already mentioned, bul ing the alliteration is even more conspicuous:
turn

in

the follow-

quae hie sunt scriptae litterae, boc [uo insunl armati atmir animati probe." (Bacch. 941)
i

militea

quid istic' verba faeimus.
«•<•!.

huic homini opusl quadraginta
et

minis

Titer calidis, danistae quas resolvat,

eito.

i

Ep. 141)

sacres sineeri

quibus hie pretiis porci veneunl (Men. 289
.'
I

Diaeresis or caesura contribute to the emphasis

and independent

unity of the amplifying ideas;

in

the second

example the entire
is

second verse has
a

a

unity of

its

own. of which the alliteration
the following example,

superficial

indication. 44

In

referred

to anion-' the cases of successive epithets, the alliteration in both verses brings into relief the distind unity of each, and the separated adjective, being only the last in an accumulation of epi-

thets, escapes into the

second verse without violence:

iam hercle ego Lstos fietoa eompositoa crispoa concinnos tuoa unguentatos usque es cerebro exvellam. (True. 287
|

In M. G. 508 we noted a certain artificiality

in

probri pudicam

mequt summi
ends of
a

flagiti
set

(above,
in

p.

222).

The employment of the
balanced
id, .as

verse to

relief a pair of

appears

in

malacum

et

mi vir. lanam, unde tibi pallium calidum confieiatur, tunicaeque hibernae bonae," (M. G. 687)
cm,',

"

The adjectives here are

less

evidently amplifying, though conis

ceivably separable: the striking feature

the position of each

"Omitted in A. "Cf. Accius 308 K 3
ut nunc,

:

cum animatus

iero,

satis

armatua sum.
I

"For
and

alliterative
in

groups including calidus

cf.

as.

255, 309,

Ep.

256j

especially,
reperi,

connection with cur passage:
ee<lo

coinminiscere,

calidum consilium

cito,

(M.

G.

226

University of California Publications.

[Class Phil.

pair of adjectives

a1

the opposite extreme of the verse, the
et.

firsl

pair varied by the connecting particle

The two substantives

are divided between the verses; the verb

common

to

both Btands
is

before the diaeresis of the second verse; the alliteration
paratively unimportant.
Cf.

com-

Nbrden, Aeneis Buch VI, 383 on

similar

phenomena

in

Vergil.

The regularity with which adjectives, following their substantives and separated, st mid at the beginning of the second verse.
In

qo1 appreciably disturbed by a
ideas.

few examples of different

dis-

positions of the separated

So the adjective sacerrumus,
which constitutes
the oppro-

which regularly appears
a unity by itself

a1

the end of a verse in Plautus (Bud.

158, Most. 983), is effectively placed in a verse

and with

alliteration that hisses ou1

brious epithet

45
:

praesenti argento hoinini,

si

leno est homo,

quantum hominum

terra

sustinel

sacerrumo.

(Poen. 89)

Similarly Plautus sets

off

the accomplishments of the parasite's

sun-dial; again superlatives, and to be sure in one case metrically

convenient

i

cf.

Merc. 206)
this

;

and again

in a

verse that

is

an

in-

dependent unit; both
tially predicative:

and the former example are

essen-

nam(unum) me puero venter erat solarium. tnulto omnium istorum optimum el yerissumum.

(Boeotia,

1,

4)

The separated adjective stands after a diaeresis, with reiteration of the same idea at the end of the same verse and in the next
verse, in
quia enim
filio

lms oportel opitulari unico.
nihilo

— al

magia

ille

unicust
Hit

mini

quamquam unicust, films quam ego illi

pater:

(Caa.

262)

(Cf. Capt. 150: tibi

unicust,
less
si

mi etiam unico magis unicus.)

A somewhal similar but
conferre

explicable separation occurs in
tesseram

itast,

si

vis

hospitalem,
is

eccam

attuli.

(Poen.

L047)
(cf.

Here

the

adjective
81

not

demonstrably amplifying

953,

« Cf. Ter. Hec.

minime equidem me obleetavi, quae cum milite Corinthum nine sum profecta Lnhumairissiimo

Vol,

i|

Prescott.

— Thought

and

Versi

in

Plautus.

227

1052, where

it

precedes the noun), though

it

may

be

fell

as an

afterthought; the association of thought

in

eccam attuli maj

have drawn
end, of

it

from

its

nou\i; bu1 the interruption, by the verse-

tin- artificial

interlocking of tesseram

confem

si

vis hos-

pitalem—a thought-unil embraced between noun and adjective Poen. 615, Pers. 105, p. The examples above is striking. 220) are similar, but the adjectives in those cases are more
I

clearly amplifying or predicative.

We
stood

bave reviewed the cases

in

which the separated adjectives

follow their substantives: 46 such adjectives have very regularly
at

the beginning of the second verse and usually with

a

caesura or sense-pause immediately following; with
tions they have been

few excep-

added

ideas, the separation of
to

which was
of
of

accomplished without
indeed,

violence
if

verse-unity;
predicative;

many
mosl

them,

were

almost

uot quite

them

gained by separation, through acquiring emphasis, or producing There is perhaps only one doubtful antithesis or sound-effects.
case
:

bonam abeat aba
It

quin potius per gratiam M. G. te. 125)
I

1

may hardly
is

be said that

bonam adds

to the thought,

for per

gratiam

sufficient in itself

M. 6. 979, 1200, and
pi.

St. 71 accordetc.

ing to Leo, Bemerkungen fiber

Wortstellungen

418 and

Lindsay. Class. Rev. 8 [1894] 159).
Plautine
the
i

Bacch. 1022. Rud. 516).
vs. 979,
a

Bona gratia is. of course, The same idea, expressed in

same play,

vin tu illam actutum amovere,

te

ul

abeat

per gratiam!

makes us suspect

thai

in

1125 the poel availed himself of the

pleonastic adjective and of separation for the sake of the reitera" Most. 501 should be added:
hospes
defodil

me

hie aeeavit, isque

me

Lnsepultum elam (ibidem) in bisce aedibus, scelestus, auri causa, mine tu bine emigra:
scelestae hae sunt

aedes,
is

Lmpiasi
in

habitatio.

The afterthought
at:
ef.

scelestus
in

echoed

scelestae.

defodit

terrain dimidiatos in Ca1

3

lnsepultum w • .XX.W'II

."..

228

University

oj

California Publications.

[Class Phil.

tion

of a-

and b-sounds,

just

as a

consideration
,?

for a-

and

t-

sounds affected the structure of

vs. !»7!).

(&)
It is

obvious thai the cases of separation
in

tive

appears

the

firsl

verse,

in which the adjecand the substantive in the second,

necessarily

involve

the

incompleleness of

the

first

verse.

In

enumerated in the previous paragraphs the adjectives ranged from purely predicative to loosely amplifying; the thought was in a measure complete in the firsl verse,
mos1
of the cases

especially

it'

the verb

came
real.

in

thai

verse; the separation

was

apparent rather than

The examples about
in

to he discussed
it

may
is it

seem, per

se,

to

impair the validity of Leo's theory;

is

important, therefore, to note that they are few

aumber.

Nor

impossible that in spite of the separation the noun or adjec-

tive

may

be so related to the context as

to

reinforce to some

extent the unity of the verses.
It

may

be well to quote at once

a

striking example of the reali-

zation of this possibility.

In one passage already quoted we
687, above, p. 225).

have seen some evidence of a rather studied disposition of adjectives

and substantives (M.

<i.

The case

before us shows evidence of even more care in the collocation of

words
aequo mendicus atque
ille

opulentissimus
(Trin. 493)

censetnr censu ad Acheruntem mortuos.
I,

is

perhaps annoying
sufficiently

which are

enumerate the features of this couplet, plain to any sympathetic reader or hearer
to

In the first place, the

thought

is

incomplete until the caesura of

the second verse

is

reached.

Yet the separation of aequo from
in

censu
first

is

attended by an effective juxtaposition of ideas

the
sep-

verse,

which gives
1.

to that verse a partial unity.

,s

The

distinguishes sharply between dissyllabic ami and maintains thai the former may qoI be separated. There does not seem to me to lie any evidence to warrant such a 'lis tinction, and it lacks inherent probability. His contention thai bonam is unemphatic and absorbed in the first foot, may ease the separation, but
c.

"Appuhn,

67-68,

trisyllabic adjectives,

Woes not explain

it.

«Cf.

<'ist.

532:
postrei

[uando aequa lege pauperi cum divite

nun

licet,

Vol.

i
i

Prescott.

—Thought
m

and

Verst

in

Plautus.

229

aration of censu results in

figura etymologica and consequent

unity of sound- and sense-effect.
ries

And mortuos
firsl

al

the
a

end caras to

us back to the nouns of the
l0

yerse in such

way

establish
ideas.

the unity of the couplet

by the dose interlocking of

A

phase of
decel

faro

kowov

is

illustrated in the following case:

innocentem qui sit atque innoxium servom superbum esse, apud erum potissumum.

(Pa 460)

The thought
servom
as

is

again incomplete until we reach the caesura of
is

the second verse; yet there
to the side of

a

fitness

in

the transference of
it

superbum, with Which
attract

belongs as
to

much
allit-

with the adjectives of the preceding \rrsv, and
it.

which

erative opportunities (cf. As. 470) of this
of the

The

significance

example is clearer on comparing it with the recurrence same thought without separation of the adjective in
innocentem servom atque innoxium
esse,

decel

confidentem

suom apud erum potissumum.

(Capt. 665)

In both passages the verse preceding the couplel

contains the

adverb confidenter, and this adverb prompts the commonplace

m
in

each case: in the Capt. the poet repeats the idea

<>f

the adverb
a

the corresponding adjective; in the Ps. he chooses

synonym.
the latter

It is not,

of course, possible to discover whether

in

case his choice

was determined by

a

desire to avoid the recur*<

rence of the same stem or whether the alliterative unit

mmi

superbum came to his mind independently of any consciousness of monotony in the repetition confidenter Bu1 confidentem.

in

any
is

case the comparative artificiality of the couplet from the

Ps.

evident

clear.'""'

the development in freedom of technique is Without discounting the value of other factors may we
:

'"Nor
lent in

is

tl

mphasis on aequo

to

!»•

overlooked;

cf.

the <;rc<k equiva

Menander 538 K
Kotvoi

'
tina.

\

A

'I'll.-

tragic

seriousness of
style,

speaker

in

the
t"

Trinummus perhaps
the

explains
Plaut.

the
50

artificial

which

Ms
is

dignity

expression

I

Leo,

Forsch.

L22 and

oote 5).
in

The hiatus

Capt. 665

perhaps

a

pan

of the crudity of composi-

tion.

230

University of California Publications.

[Class Phil.

qo1

say thai

when once
|

the alliterative unit occurred to him the

unity of sound prove,
for
its

superior

to the affinity of the attributive

noun, and that this conservation of unity of sound was
thai there

made

easier or perhaps suggested by the fact

was

a

strong unity of thoughl as well which linked servom to super-

bum?

By

this question

we do not imply any conscious
to

intent

on the poet's part; we mean simply

suggesl

thai

the two

to us to prove thai the poet's technique on occabeyond the point of preserving the more natural and obvious unity of thought, and shows here as elsewhere a sensitiveness to unity of sound and to the more artificial phases

examples seem
sion

had

go1

of unity of thought.
In
this connection,
in

properly, we should note the isolation of
verse by the transposition of
its

an adjective
t
>

the

first

noun

a

relative clause that occupies the second verse: 51
nisi qui meliorem adferet quae mi atque amicis placeat eondicio magis, (Capt. 179)

It will be

granted that this
the noun, to

is

analogous to our previous examis

ple: again

which two attributive ideas belong,
in

•x

pressed with the second.

Somewhat similar, too, are these cases mon to two adjectives is separated from
stands
;it

which a noun comfirst

the

adjective,

and

the beginning of the second verse before a sense-pause;

the second adjective stands in the same verse with the noun:
multis et multigeneribus opus est
militibus:
tilii

primumdum opus

est

Pistorensibus;
ita

(Capt.

159)

quam ego postquam

aspexi, nun
ut

amo

ut

sani solenl

homines, sed eodem pacto

Lnsani so-lent.

(Mere. 262)

The

sound-effects, especially

in

the tetrasyllable

rhyme

in

the

second case, are obvious.
''

Tin-

figure

of
in

speech

involved,

without

separation

by

the

verse,

is

easily paralleled
tract]
. .

Plautus and other poets:
ap. scriptores

for examples cf. Bach, de atL6;

.

inversa

Latinos

Vahlen, Eermes
I

17

(1882)
however,

598 599;
separation
cf.

Leo,

A.nalecta

Plautina:
the
1

de figuris sermonis
adjective
p.
is
|.

20.

If.

by

the

verse occurs,
1

usually a

demonstrative:

Rud. L065, Poen.

19

(quoted

elow,

254

Vol.

i
i

Prescott.

Thought and

Versi

in

Plautus.

231

Equally studied

is

the juxtaposition of differenl case-forms of

the same word; the separation that results
the attraction of words of the same stem

may
for

indicate that

each
its

other 52
i

is

stronger than the attraction of the attributive to

or

than any sensitiveness to verse-unit y

:

nam

ex uno puteo similior

nunquam
est,

potis
ista

aqua aquai8' sumi quam haec

atque

bospita.

(M.

<>.

55]

|

Again the thought reaches

a

partial completion at the caesura;
in

the four objects in two pairs are grouped

the second

verse;

and the sound-effect
to the audience.

aqua aquai was doubtless oo1 ungrateful This example, too, gains in significai from
in
in

the occurrence of the same thoughl

another form:

nam

ego hominem hominis similiorem aunquam vidi alterum. neque aqua aquae nee lacte esl lactis, crede mi, usquam similius,
hie tui est, tuque huius
is

quam

autem; (Men. 1088)
is

Bere

it

worth aoting that the second example, which
all

with-

out separation, shows

the simplicity and explicit

fulness of

an early and undeveloped style; the identity of sentence and
verse
first

is

almost as exact as
the

in

the early Saturnian verse.

The
the

example, on
is

contrary,

shows
;

a

i'vrcr

technique:
is

thoughl

more condensed,

less explicit
a

verse-unity

less scrua

pulously preserved.

We

have

suggestion before us of

dif-

ference, if not of a development, in verse-technique in the course

of the poet's activity.
Artificiality in the disposition of
in
lien inciiiiiiisti me auream ad te afferre oatali die lunulam atque anellum aureolum in digitumf (Ep. 639)

words

is

dearly discernible

The
with

chiastic

arrangement of the pairs of substantives and adjecin

tives,

the consistent diminutives in the second verse in contrast

annum

the

first

verse,

and the

artificial

interlocking of
is

the words are the noticeable features.

So far as any unity
to the

discoverable,

it

I

sists

only

in

such unity as appeals
Rh.

ear
Kel-

'For other examples
lerhof,
I.

ef.

Kiessling,

Mus.

23

L869

Ultl'..

c.

58 60.
in

Tin- traces of aequt

A and B (both, however, corrected

t"

aquae)

oeed net detain us:

cf.

Men. L089 quoted above.

University of California Publications.

[Class Phil.

t'n. hi

the differenl sound-effects of
in

<-;i

<-h

verse—a-sounds predomin

inating

the

firsl

verse,

1-, 1<»

m-,

u-,

and a-sounds

the second;

certainly there does seem

be something conscious in the change

we form and of soundconeffects in the second verse could have arisen only from At the sciousness Unit the second verse was a distinct entity.
from auream of the
firsl

verse to aureolum of the second;
the unity of

may properly maintain

thai

;i

same time the fad
in
;i

thai

the consciousness expresses itself only
that,

superficial

oi-

external [(reservation of verse-unity, and
is

unity of thought

interrupted, suggests that "art-poetry*' in

Plautus's hands was on occasion further advanced
chronological
suspect.
In contrast
a

than
us

the
to

proximity of Saturnian verse would Lead

with merely superficial observance of unity stand
in

few cases of separation

which the thought serves
hosticum
line

to reasserl

the unity of the verse
niihi

domieilium

est.

Athenis domus

est

Atticis;
sitis

ego istam

domum
(M. G.450)

neque moror aeque vos qui homines

novi noque scio.

Alliteration, to be sure,

may have

attracted hosticum to hoc, but

the

dominant factors are emphasis and contrast. Hosticum is firsl in the sentence because emphasis brings it to that position. Domieilium is first in the verse' because emphasis again de1

it must stand in the same domus to bring out the contrast between ••house" and "home." The effect may be suggested in English by In, use you point me to; my hom< 's in "Stop! a stranger's who you gentle Athens; for your home I care not, nor know men may be."

mands
verse

for

it

a prominent position:

with

\

1

|

Another passage

in

which

at first sighl

unity seems to be dis-

regarded, when studied in the light of the context, shows considerable consciousness of the intimate association of verse-unil

and thought-unil
babui

:

aumerum
in
:ie<iis

eedulo:
id

angiportum,

hoe est sextum a porta proxumum angiportum me devorti iusserat;
i<l

quotumas

dixerit,
in

eye

admodum

Lncerto Bcio.

(Ps.

960)

"But

est

domieilium

CD.

Note also hostium (ost—) BCD.

Vol.

i|

erescott.— Thought and

Versi

in

Plautus.

233

Ilcrc again,
tive

it

may

be said

th;il

porta has attracted the alliteraa

proxumum, but
to the
its

the verse-division represents

correspondverse Leads

ing division of thought.

The beginning
a1

of the

firsl

up
ond
in

number and

precise location; angiportum, separated

from

two adjectives, stands oul

the beginning of
is

lie'

secre-

verse, again with emphasis,

and

repeated55 with the
contrasl

sumptive pronoun
right: the sixth,
the alley-way
thai
I

all

of

which

heightens the
is:

with

dis of the third verse.
(in

The

effect

"I've go1 the number
alley-way, lhat"s
of the house,

going from the gate),
hut the

was

told to take;

number

t've clean forgotten."
in

Perhaps the existence of any unity
will be less readily

the following example

granted:
maiorem
filius

eoepi observare eequi

mini honorem habere! quam eius habuissel

pater. in

(Aul.

L6)

There seem

to be

two prominenl factors
is

the separation: the
dif~
:

comparative degree
Yet

attracted to the ablative of degree of

ference; 58 alliteration
iit

brings together honorem
in

and haben

t.

too fanciful to say thai

spite of the separation the

position of filius
a

and pater

at

8 the ends of their verses-" suggests

unity of thought quite apart

from and above the syntactical

and

alliterative unity of each verse.'

The two

verses are com-

parable to the two pans of the scale, the son balancing the father,

and maiorem alongside of. filius marking which the expectant Lar anticipates. 50
B Examples of such repetition
demonstrat.
50

the turn of the balance

= Studemund-Stud.

may

be

found

in

Bach, de

usu

pron.

II

353 354.

31

See the examples in Fraesdorff, de comparative gradua usu Plautino Other factors, external <>r internal, may have precedence over tinthe comparative, but the generalization above is not thereby endangered. honos homini Trin. 697, meqw honorem illi habert True. 59] honores sum domi habuit maxumos Pers. 512, habuit, rru hdberi honorem
ff.

natural juxtaposition of the ablative of degree and
;

i

i'.

ks. 81.
:s

To be
is

sun-, they
ras.

owe
that
I

their position
-•

in

sum.-

metrical

cue
(but

venience:

cf.

i_. 21,

30 of the
the
i

prologue.
involves separation

'It

not

likely
in

following example
:

m Chium
nl
>i

urc 7^

tu

Leucadio, Lesbio, Thasio, Ohio,
vino

vetustate

edentulo

aetatem

inriges.

i

234

University of California Publications.

[Class Phil.

Ndr can

I

be sure thai

my

understanding of the nexl case
is

will

prove convincing.

The adjective mutuos

occasionally sepa-

rated in expressions of the ideas of borrowing and Lending; in

two of
tive

tlif

cases the adjective Hollows the noun, in one the adjec-

precedes.

For purposes

of comparison
in

I

include them

all

here, although the

former belong
nummos

the previous section:

tecumque orari
dares

ut

sescentos mihi

utendos

mutuos.

(Pers.

117)

sed quinque inventia "pus es1 argenti minis mutuis, quas hodie reddam: (Ps. 732) scl potes nunc

mutuam
til>i?

drachumam dare imam

mihi,

quam

eras

reddam

(Ps. So)

The frequenl
in

collocation of this adjective with dare

and rogan

commercial phrases
to the

may have

given

it

a

substantival force

corresponding

English "loan":

so,

for example, exoran

mutuom
that

in

Pers. 43 (with argent'um far distant in 39) suggests

the

adjectival

force

is

approximately substantival, 60 and
in

eventually this substantival usage becomes established; even

Plautus we have tuU
819).
if

si

pudoris egeas, sumas

mutuom (Amph.

If this

is

granted, the separation becomes innocuous, even

the adjective precedes; the alliteration in the last example

perhaps adds to the unity of the verse, but no such additional
feature
is

necessary

it'

mutuam

is

in

effect appositional.

The
t.

cases hitherto discussed have shown, in varying decrees,
it

consciousness of verse-unity and conservation of
nt in

to

spite of the separation of the attributives.
to consider

some exThe examto a sensi-

ples

we have now
to

do not so plainly point
verse-

tiveness

the

identity

of

and sense-unit.
in
a

There are

often

extenuating circumstances, but

most cases we must
interruption of a
the features thai

admit that the separation involves
tlioimlil-unit
in

distinct

with

less effectual

employment of

other examples reinforced the unity of the verse.
»Cf.
Ps.

Prominent

294:

quUus

est

tibi

mutuom
A-.

:i

rgent

am

7

-

quin nomen quoque Lam

Lnteriil

quem roges " mutuom."

248 and

stantival.

Trim 1051 also show mutuos in a sens.' approximately BubThe various forms of facen mutuom are hardly parallel.

Vol.

i
i

Prescott— Thought and
these

Versi

in

Plautus.

235

among

is

;i

group of superlatives of
in

cretic

measuremenl

which may owe their separation
to separation, bu1

pari to metrical convenience;

occasionally there resull sound-effects thai
in

may have condu
is

1

general the violation of unity

unmistaksuperficial.

able,
It
is.

and the palliating or counteracting features are
however, always
to be

remembered
a1

thai

the cases of sepa-

ration are extremely few in proportion to the

number
verse.

of occur-

rences of a given adjective

the end of
is

a

The mosl

important member of

this

group

maxumus, which we have
its

already found separated, bul
a1

following

uoun and standing
This adjecat

the beginning of the second verse with emphasis.

tive

appears 86 times

in

Plautus:

:'»!)

times

the end of the
It

verse, 38 times in the interim-, nine times a1
is

the beginning.

not likely that, under uormal conditions, the position at the
is

verse-end

prompted by

a

desire to emphasize; 63 generally unposition.

emphatic words occupy

this

examples of the phrase open
ration, will
illustrate the

A collection of all the maxumo, with and withoul sepa-

62 feature of metrical convenience:

nam
nunc

rex Seleueua
te

me

opere oravif

maxumo (M.
(Most.

G. 75)

hoc

orare

Lussit

opere

maxumo
(Ps.
(St.

752)

pater Calidori opere edixit

maxumo

897)

rogare

iussit

ted ut opere

maxumo

248)

iussit

maxumo
120)

opere orare,

ut

patrem aliquo absterreres modo, (Most.
null

hercle vero taceo.

nam

tu

maxumo

me

opsecravisti opere,

Casinam

ut

poscerem oxorem mihi (Cas. 992)

Cf. Terence,
Thais
te
oratiat

maxumo
532)
once by
the

opere,
for

ut

eras

redires.

(Eun.
al least

Such a position emphasis is occupied words with which we are now concerned:

very

ego miserrumis periclis sum per maris
vectus, capital] periclo per praedones

maxuma plummos

me

servavi.

(Trin.

L087)
isomewhat more apparmaion opere, nimio opere, tanto opere

The
ent

significance
that

of the cases of separation

when we note

magno

opere,

are never separated in Plautus by the verse-end.

_':;<',

/

niversity of California

Publications.

[Class Phil.

It

is

evident thai open

is

attracted to oran
is

and opsecrare, but
oothing
in

so far as the

thoughl
in

is

concerned, there

to

diminish
TJi>.

the violence

the divisioE of

maxumo open
in

Must.

or

the division of the
532.

Larger word-groups

(';is.

992 and

Eun.

And

in

the

firsl

of the two following e;ises of

maxumus
in

there are qo sound-effects to relieve the separation;

the sec-

ond, separation brings together m- and a-sounds; these are, however,

from

lyrical

passages:
ibi

ubi quisque Lnstiterat, eoncidii crepitu.

aescio quia

maxuma

voce exclamat:

(Amph. L063)
quid machinerf quid comminiscar?
(Capt. 531)

quam malum!

maxumas

augas Lneptua incipiaso. 68 haereo.
(

'!'.

Terence,
Geta, bominem
preti

maxumi
(Ad. 891)

M

te

esse

hoclie

iudicavi

animo meo;

Consideration for sound and

the artificial

arrangement of words

may have

played some part in the structure of these verses:

Alexandrum magnum atque Agathoclem aiunt maxumas
duo res gessisse: quid
inihi fiet
tertio,

qui solus facio facinora

mmortalia?
in

(Most.

775)

The a-sounds are prominent

the first verse:

magnum and
same verse;
contrasl with

maxumas

are perhaps not unintentionally put in the

duo, interlocked between
tertio a1 the other

maxumas and

res,

is

in

extreme of the same verse. 68
at

Another superlative optumus occurs
in one third of the total

the end of the verse
in

number

of

its

occurrences;

only one

case does
I,,,

its

position result in separation:
is

1,1ms incipissi

the reading of the

Mss.

"Contrast with this the stereotyped position ;it the end of the verse, without separation, of minimi preti, parvi preti, magni preti, quantivis preti in Plautus (cf. Rassow, de Plauti Bubstantivis s. v. pretium GS.= JHB.
Supplbd. 12 (1881) 710).
or. altera
.
. .

.

.

.

altera, Aul.
ii;
It

195; superi
. .
.

.

.

.

inferi,

Aid. 368;

n
.

dignius, Baceh.

malefactorem

beneficum, Baceh. 395;

meam

.

.

hmm.

Capt. 632.

is

interesting to uote in this connection a couplel

in

bacchiac verse:
sed

vero duae,

sal

scio,

maxumo

uni

populo cuilubel

plus satis dare

potia

-nut.

(Poen.

226

Vol.

i
i

Prescott.

—Thought

and
sed,

Versi

in

Plautus.

237

it.',

optuma
(Ep.
202)

70a

video

opportunitate

ambo

advenire.

Wiih

this

should he compared
adveniistis.

optuma opportunitate ambo

(Mere. 964)
is

Next

iii

significance to the rarity of the separation

the fact,
-

attested by the verse from the
v

——
v

Mere,

thai the initial

sounds op
is
IS

<>i>

are the external manifestation of unity which

cer-

tainly interrupted by the end of the verse.

Such

;i

case

far

from disturbing Leo's theory.

Such interlocked complexes of
within the

thought and sound, which are characteristic of the language,

must hurst the bonds
verse; that they do
it

that

confine units of thoughl
is

so rarely

significant.
at

A

third superlative that, like optumus, stands

the end of
is

the verse in one third of the total
(illiniums.

number

of

its

occurrences

The singular and
so,

the plural of this
is

word are perhaps
to the

on

a

different footing: the plural

conceivably analogous
in

separation of omnes;™
in
ii

for example,

this ease of

plurumi

the interior of a verse, the separation seems less yiolenl than
eases of the singular
67

plurumi ad ilium
periere
pueri
liberi

modum
988)

Carthagine.

(Poen.

Whether
is

this

is

true

in

the ease of the following feminine plural

not at once patent
()

to

an English auditor:
hominum plurumae
(Bud. 1235)

Gripe, Gripe, in aetate

fiunt

transennae, ubi decipiuntur dolis.

Tn

any

ease, the singular serins at first to he rather

ruddy

sepa-

rated in
miles Lyeoni
in

Epidauro hospiti
(Cure. 4:M>)

suo Therapontigonua Platagidorua plurumam
salutcni dicit.

naturally and result
Cf.
in
I

Here the conventional phrases of epistolary address run along in two separations, with the first of which
elow,
p.

258.

Eph.

.".!M

pluruma (plurwnum .Mss.)

is

predicative.

238

University of California Publications.

[Class Phil.

wr are

qo1

qow concerned,

bu1

verse-unity

is

suggested

in

the
is

alliterative colligation of Platagidorus
if

plurumam;
dicit

the effecl

as

plurumam were an adverb and salutem
it

oo more than

sain

nihil, as the following

example suggests:
Philto Lubet (Trin.
135)

erum atque Bervom plurumum
salvere,

Lesbonicum

el

Stasimum.

in

which, again, we have similar alliteration
Pilto.

plurumum
is

Philto,

pronounced

So, too, our explanation

confirmed by

multam me
salutem
in
Lussit

tibi

Therapontigonus

dicere

(Cure.

t20)

which, as

in the

other cases, multam mt are attracted to each

other, while salutem iussit like salutem dicit
68 at the beginning of the second verse.

and salven stands

The adjective parvolus occurs
the

thirteen times: nine times

a1

end of a verse, three times with separation. Of these three. one belongs in our examples of adjectives following their nouns, and is a mere afterthought:
nam mihi item gnatae cum
nutrice una sunt
cluae

surruptae parvolae.

(Poen. 1104)

The other two. both from the same play and of the same situation, are cases of violent

and absolute separation: 69

nam ego illanc olim quae hine flens abiit parvolam (Cist, L23) puellam proiectam ex angiportu sustuli.
mihi ah hippodromo memini adferri parvolam (Cist. 552) puellam eamque me mihi supponere.

nam

A

comparison with three cases
in

in

Terence

justifies us in

attrib-

uting the separation

Large measure to metrical convenience:

ilii turn matri parvolam puellam dono quidam mercator dedil

I

Eun.

L08
in

68

On

the other hand, withoul
in

separation, but again

alliterative

col-

ligation

multam
09

meis

verbis

Veneri dicito salutem.

(Poen.
in

406)

The
is

alliteration,

interrupted by the verse-end,
it

has
idea

no significance, for

is

accidental:

the
it

puellam range of expressions for the

parvolam

too limited to admil our regarding

as genuine alliteration.

Vc "-

"I

Prescott

Thought and
periit
ut

Verst

in

Plautus.

239

nisi

>i

iihi

forte quae olim

parvola
esl

Boror, hanc se intendit

esse,

audacia.

(Eun. 524)

ah, stuli

it i.-ist

istaec,
!
I

aon
27
I

pudor. tam ob parvolam
|

rem paene
In the second

e patria

A.d.

parvola
effects

—"died

example sense
in

;is

well

;is

sound may eom
the
lasl

t

periil

infancy," and

in

there are sound-

th.-it

reassert
this

the unity of the verses. 7 "

So much for
participial

group of

cretic adjectives

T]

the

following

adjectives
lie

may
said

he

more

easily separable
force
in

because of

it

could

'"Something mighl hardly apply

for a
last

substantival

parvola, though
a

to

the

example from Terence: Buch

force

is

possible in Ter. Eun. 155:

par v.. la
nine
the
est

abrepta;
is

substantival

force

evidenl

in

Terence's "
to
it

parvolo

(Ami.
is

35,

A.i.

48)=a
1346, but

puero.
it

The nearest approach
not

in
is

Plautus
Ps.
7s.".

in

Poen. 896,
I

is

certain

in

either place;
p. 59.

nor

a

clear case.

f.

Lorenz, Pseudolus, Einleitung

Before
to

leaving
in

these

examples
I

in

which
call
is

metrical

convenience
to
;i

seems
re-

be a
the

factor

the

separation.
it

may
me.

attention
not
in

closely

lated

phenomenon which,
stereotyped
positi

seems
t'

to

always
the

recognized.

Is

not

certain
|

words

verse

often
of

nothing
in

""" v tban the working of the t's mind along "grooves." as a psychologist might express it.'
cases above
i'

the

path

well-worn
the

For example,

which salutem iussii or dicit, or salvert mint, appear, the positi salutem ami solvere (rather regularly at the beginning of the rerse, though not uniformly) can hardly lie attributed to metrical convein

nience alone:
is

it is to S0 ] stent a matter of hat.it. A better example furnished by these examples from Euripides's [phigeneia in Tauris:

-n/in/rniv
(';;

mi

j-EOTOV EH VCLOV >u

hlV

<i/iia

~aaaq
/.'.

Trfinoipqim-rr u/,\,n,n.

(Ill)

I".

.'...

//i.'f

devpo, (hn-FTt r'/ai: iv
-'

I'lui'/u'

'Atiqvuv
,-•

e)

Kadtdpvoat x^ovi.

('.»77

i

n

i

to c

voiaiv ot^erot,
fff/v

oepvm

8
(

aya2.fi'

Ixovaa- ddXta
.

Kaddpfiara.

1315

u-ii'/iiii.

in,,,

-r

(1384)

Those of us who are reluctant to admit metrical convenience as a may find some comfort in emphasizing the part that mental hal.it plays in the regular appearani f certain words in the same part ..f th( 'AyaXfia in the verses above seems to me to owe its position to this rather than anything else.

240

University of California Publications.

[Class Phil.

the peculiar nature of the adjective,
of
/"//' r:

and the balanced
salve, insperate nobis

isolation

pater, te compleeti nos Bine.—cupite atque exspectate
pater, salve.
I

Poen.

1259

I

The greetings are from two
ranged
verse, 7
-'

sisters

with artificial

variation of

the conventional terms: the imperatives and
in chiastic

vocatives are ar-

order; pater stands

a1

the beginning of each

leaving the adjectives at the end in
is

each

case.

The

collocation

the same as in
conspieoi

o salve, insperate multis annis post quern
frater,

(Men. 1132)

to MS. B, but the other members of the Palatine fam(and A apparently agrees) introduce a change of speakers Even it' we agree with the editors in following before frater.

according
ily

A and

participial vocative,

the majority of the Palatine family, the isolation of the and the relative clause that modifies it, may

tives insperate, cupite,

point to a certain degree of separability in the participial adjec78 and exspectate in our passage.

seems to be
the verse:

There remains a small group of cases in which verse-unity lost sight of, and which are alike in that the adjecof tives are of four syllables metrically convenient at the end
pol istic

me haud eentesumam
ut

partem laudal quam ipse meritusl

laudetur laudibus.

(Capt. 421)

partem

di.xi

atque, otiuro
si

rei

si

sit,

haud eentesumam possum expromere.
liberal]

(M. G. 763)

quisquam hanc causa manu assereret, (Cure. 490)
ne epistula quidem ulla nee errata adeo tabula
pictura
sit.
sit
;

in
si

et
:
I

aedibus qua inutilis
763
I

earn rendal

A.s.

Centesumus occurs only
causa occurs
in

in

these two places in Plautus; liberali

the interior of the verse in Poen. 906, 964, 1102,

-c{. above, p. 223, n. 35. "Ferger, de voeativi usu Plautino Terentianoque insperatt f b in Men. L132 on the ground thai ing tus without an accompanying noun.

32,
is

not

defends the readfound in Plau-

Vol.

ij

Prescott.

— Thought

and

Verst

in

Plautus.

241

and
Ps.

so liberali many, in Cure. 668, 709; inutilis occurs again in

794 and
is

a1

the end of the verse.
a

Bu1 the separation

in

these

cases

not

entirely

matter of Length and metrical conveniin
is

ence: the collocation of the other words
fixed

the sentence
not

is

so

by almosl

inviolable

Laws thai

if

surprising thai

the adjective should escape into the second verse.

For

to

anystudj

body
of Hie

familiar
position

with
in

Plautus
the

and
of

with

Wackernagel's

sentence
/><>!

enclitic
>n<.
si

words

it

will

lie

clear that the collocations
*

istic

quisquam hanc, and
the usage of the
is

i

si

qua are

to

;i

considerable extent

fixed in

Language; the increased difficulty of conserving verse-unity
obvious. 74

The very

fact that in

some 15,000 verses

so few c;ises of sepa-

ration occur

— and

this in spite of the fondness of the

Roman

for interlocked

complexes which would seem to make the preser-

vation of verse-unity difficult

— clearly attests
thai
in so

the sanity of Leo's

contention.

The further fad

many
Large

of the few cases

of separation the unity of the verse reasserts itself through a>sociation

of thought or sound confirms in

measure his

re-

quirement of special justification when separation does occur. The existence of a few cases in which unity is not apparent need
not affect the validity of the principle; the essential unity of the
verse so far as attributive adjectives are concerned
is

clear at

once from comparison with a tragedy of Euripides or of Seneca

— clearer
The
of

than any

statistics

could

make

it.

IV.
large proportion of possessive adjectives anion- the cases
leserves

separati

an

explanation.
if

They

represenl

one-

fourth of the total: indeed

we eliminate cases of merely apparent separation the proportion would lie even Larger.
\'o

small part of the explanation

is

found, of course,
in

in

the

relative
74

frequency of the possessive adjectives
As.

the conversaposition

In

763

ff.

there

is

perhaps some
pictura
earn
at

effect

in

the

of

the

nouns epistula, cerata
cessive
last

tabula,

or also

near

the

beginning
unity
of the

verses.

The resumptive

may

reinforce the

verse.

242

University of California Publications.

[Class Phil.

tional

Latin of the plays.

Thai among 3000
60 should

75

cases of posses-

sive adjectives only aboul

be separated
in

from their

substantives by
tribute to

the verse-end

may seem
;

itself

some
it.

slighl

verse-unity

rather than

ntravention of

Yel

the obvious violence to the unity of thought, al Leasl from an English standpoint, in dividing "thy son'" between two verses

makes even
must
us.
is

a

small

percentage seem

inexplicably

Large.
to
in

We

not,

however, allow our English standpoinl
separation of

influence

The separation of "thy son" by the verse-end
not alto-ether
in
is

English
tuos.

analogous
of an

to the

filius

from

For
son'"

the

Latin sentence the phrase corresponding to "thy
less

much

independent unit of thought than

in

the English sentence: in the Latin sentence, largely because the possessives metis, tuos, stios are generally unemphatic and often

without accent

in

the verse-end does not separate

the phrase- or sentence-unit, the division by "thy" from "son.*' hut rather
It
is

divides
tuos

a

larger unit of thought.
aedis
filius

clear, for

example, that
a

emit

(Most.

G70)
til

constitutes
ins
\

unit

of

thought;

and

so.

too,

does

aedis

tuos

emit

(Most.

637, cf. 997).
is

rather that of aedis
tuos.
sttos

The separation in this latter case, if any is felt, filius from tuos emit than merely of filius
it

from
hms,
sible

Furthermore, since the possessive adjectives mens,
are generally unemphatic in our examples,
likely that in this
is

pos-

and

example tuos was absorbed

in

the

rhythmical
violence
in

unit

tuos

<

mil without

much

consciousness of any
fre-

separating tuos from

filius

by the verse-end; the

quency and
sessives

ease with which words intervene between these postheir substantives (quite apart

and

from separation by

the verse-end)
sessive

may

support this contention.

Even

if

the pos-

had some
Cist.

slight stress
in

upon

it,

as in the beginning of
{filiam

trochaic verses

and rarely
600),

an iambic verse

suam

despondit,

certainly

such stress was subordinate: have been

suam, despite some quantitative prominence, must

merged
:

in

the surrounding words.?
I.

8

Of course

it

may

he ob-

NiNsuii,

c.

L2.
i^

Some
faila

such

idea

expressed by Appuhn,
verses

1.

<.

63,
I

but

in
it

a
is

to

account

for trochaic

ami

<'ist.

tine.

hope

way that dear that

Vol.

i]

Prescott.

-Thought and

Versi

in

Plautus.

243

jected thai the thoughl would lend us to
ingj

merge

i1

in

the

pn

d

rather than

in

the following word, in the
is

example quoted,
For our presabsorbed
is

and
eni

thai the possessive

enclitic,

7'

do1

proclitic.
is

purpose

it

is

enough

thai

the

possessive

in

a

larger unit, and thai

the separation by the verse-end

by no

means
of the

the

same

as thai

involved
its

in

the division between verses

English possessive and
place
it

substantive. 78

In the second

is

to

be noted thai the possessives are

subject

to at

least

one influence from which ordinary attribuothers before him, observed
to

tives are free:

Kampf, 79 and
a

the

attraction of pronominal
tion

words

one another.

Such

attrac-

appears
cam meae

in

relatively small
(

number

80 of our examples:

uxori

Men.

I

s"

1

illam quae nicaiii
|

gnatam
I

(Cist. 547),

tu milii tua

oratione

A.s.

L12),
(

ad illam quae tuom
forea eonservaa

... nlium

Bacch.

106

l,

tneaa a te (As. 386),
is

the

paragraph above

aot

intended

to

offer

any complete explanation
aeparation,

of the separation, but
is

only to suggest

that

the

such as

it

is.
I

The poinl thai probably by no means ao harsh as it appeara to us. wish to make is that the unemphatic possessive has very little independent
force and
is

aot

merely " swallowed

up " (Appuhn)
cf.

metrically,

bul

ab-

sorbed

in

larger thought-units even of ordinary speech.
K.

"Lindsay, Latin Language 167; but
Poreningen:
Trans.
articles

Wallstedt,

Fran Pilologiska
ff;

Spr&kliga
Phil.

dppsatser
36
in

MI

(Lund L906) 189

also

Eiadford,

A.ncr.

Assoc.
to

was accessible
fact

me

last two (1905) 190 ff. time to use them for the discussion above.

Neither of these

'•The
separation

that

the genitive case

is

use,

l

in

appositional relation to the

possessives

(e.g.
ia

mea unius opera) might lead to the suggestion that the This not more serious than that of a possessive genitive.
auggestion
if

would be
regularly

a

helpful

the

possessive

genitive

in

Plautus

were

noun by the verse; cases Bac.-h. 901, Bud. 1079, Cist. 544), but rarely: and the do occur (e. g. possessive genitive with pater, uxor, filius, mater, which are the nouns most
or even

frequently separated

from

its

frequently
is

appearing
e.

in

our cases

of

the
its

Beparated

possessive

adjective,

in

Plautus almost

inseparable from
(V.

noun even by intervening words.

Kampf,
1

1.

16

though too few to be significant, of a verse-end internu vening between pronominal words thus combined are worth noting: tua meat ia-. 279 280, meam meam Ep. 180 181, mea Cist. ,s 99, nu tua Pa 112 113. suamqut Trin. 1"'.' 11". tibi M. (.. 738 739, at

A

few

cases,

m

'.

jii

University of California Publications.
men

[Class Phil.

Blio
si,,,,

te esse

amicum

el

ilium intellexi tibi

(Capt.

11"'.

dispendio

tuo tuam libertam
(

I

Poen. L63

I,

servos ...

buos mihi

Most. L087

i.

[f alliteration

appears

in

such cases,
il

il

is.

of course, incidental

and

results from the attraction;
I

is

not n

primary

factor.

Wackernagel
tuos,
it,,'

[ndog. Forsch.
his

I.

i06ff.) does qo1 include mens,
enclitic

suos

among

examples of

words

thai

drifl

to

There are cases of separation thai mighl have been affected by liis law, bu1 they are too few i,: suggesl the dired influence of his Law; these few show the
beginning of the sentence.
enclitic

immediately following the introductory possessives word; they seem more significant when other words intervene Aul. 733, between the possessive and the noun: e. g. True.
:'>.">.">.

Since Wackernagel's law affects particularly certain monosyllabic and dissyllabic pronouns, it follows that in comSI.
41fi.

bination with the law of pronominal attraction there results in many cases the necessity of placing the possessive in the third

or fourth place; take, for example, these two cases, one of separation,

one without separation
tu tua

:

eonteris

me

oratione, limlicr. quisquis es.

(<'ist.

609)

profecto nemo est quem lam dehine metuam mihi ne quid aocere pes^it. cum tu mihi tun (As. ill oratione omnem animum ostendisti tuom.
S ay

i

T,,

makes
to a

tu second in the sentence, in

aothing of other features, the rule of collocation that combination with the attrac-

tion that joins tu tua mi

and

tu

mihi tua, undoubtedly regulates
it

considerable degree the disposition of the words; and
consideration

is

clear that the existence of such laws of collocation

musl appear
of
verse-

seriously
unity,
at

to
leasl

interfere
in

with
cases.

the

poet's

many

Such laws affed the spoken language;
servant of them than of verse-unity,
expeel of
a
it

if

Plautus

is

more

ob-

is

no more than we should

dramatic poel who
.lay.

is

reproducing the conversational
to

Latin of Ins

The same general truth applies
a

ordinary

attributives, bu1 they are qo1 a-

class

subjed

to thess particu-

v.m..

1
1

Prescott— Thought and
In addition to the
in
is

Vers<

in

Plautus.

245

lar regulations.

observance of laws controlis

ling the

arrangemenl of words
It

speech the pool

governed by

the conditions of his verso.
of metrical convenience.
factors.
It

easy to overestimate the force

is

seldom more than one of many
the iambic or pyra1

Bu1

it

may hardly

be denied thai

rhic possessives found a comfortable habital

the

end81 and

a1

the

beginning of certain

iambic and

trochaic

verses,

[ndeed,

quite aparl from the metrical convenience of the possessives thai

do not involve separation, the cases of separated possessives of
iambic or pyrrhic measuremenl lead
1) in all cases of separation
a

to

two conclusions:
tuos, or suos fol-

in

which mens,

lows

substantive, whether with or without intervening words,
at

the possessive stands
2) in
all

the beginning of the second verse;

eases of separation in which metis, tuos, or suos pre-

cedes a substantive, whether with or without intervening words,
the possessive stands
at

the end of the

firsl

verse.
test

83

The exceptions
It is

to

these principles 84 only

their

validity.
firsl

of course evident thai in the eases covered by the
is

rule

there
i>t

no reason

why

the possessive should qo1 stand
is

at

the end

the second verse: such a position

unusual, probably because
is

the
greal

separation
;

by

intervening

words
is

thereby

abnormally

an example from Terence
qui tuna illam amabant, forte

ita

ui

fit.

filium

perduxere

illuc,

secum

ul

una

esset,

meum.

(And.

5

Similarly under the second rule there

is

do reason

why

the pos-

sessive should not stand at the beginning of the
here,

firsl

verse; bu1

again, such position

is

unusual

probably because of the

extent of the intervening words: an isolated
B

exampL
L41,

For

statistics ef. Nilsson,

1.

c.

37.
A.ul.

"Amph.
998,
less.

134,

L35,

As.

387,
130,

134,

289,

Bacch. 880, Capt.

B73,

cist. 586, 601, Cure. 347,

Ep. 391, M)l, 182, 583, M. G. 543,
Ps.
183,

M
L101,

Poen.

L64,

L92,

1375,

650, 850,

Bud. 743, Trin.
L84, 547,
77

lilt. True. 293.

As.

16,

112,

785, Anl.
7 1".

733,

Bacch. 406, 777. Cist.
7'.'!'.

279, Men.

120,

180, 518,

M. G. 563, 635,
first

Bud. L392, St. U6. Trin.

1147. True. 355.

•"The hiatus, therefore, after the
suspicious.

cured by changing suam hue to hue suam

word of Ps. 650 is not to be (Bothe), and Trin. ltl becomes

246

University of California Publications.

[Class Phil.

meamne
per vim

hie
ul

Mnesilochus,
retineal

Nicobuli

filius,

mulieremf
is

(Bacch.

842)
in

In both cases the rare position

attended by other features:

the

first,

the postponemenl of mi urn perhaps suggests the pathos

of the situation; in the second, emphasis, alliteration,

and

collo-

cation with hie are contributory

factors.
in

Finally, such an ex-

ception to these rules as appears
to the peculiar

the following

example

is

due

nature of the formula and the greater conveniat the

ence of obst cro

end

:

adsum,
genua,
u1

<

lallieles:

per

fcua

obsecro

tu istue insipienter

factum sapienter feras

(True. 826)

Gf.

Cure. 630, where per tua genua

U

obsecro concludes the

and Poen. [1387], where, again 85 we find per <</<> tua U genua obsecro.
verse,

at the

end of the verse.

We
st
I.

tively large

have thus noted several features that make the comparanumber of separated possessives more easily under-

As

in

the case of ordinary attributives, there are occa-

sionally special

conditions which emphasize the unity of the
of the

verse in
arising

spite

separation.

The accidental

alliteration

from pronominal attraction we have already noticed
a
'Tit

there are
ubi

few cases of genuine alliteration:
empta,
ut
(

aliquo ex urbe amoveas;

nisi

quid est tua

sccus sententia.

Ep. -7i»)
suos

nam hominem servom
domitos habere oportel oculos
et

manua

(M. G. 563)

oculoa vein

meos deleetare munditiis

meretrieiis.

(Poen.

1!»1)

There are
in

;i

tew cases, allied

i<»

those of pronominal attraction,
not

which pronominal words are
in

immediately juxtaposed hut

are grouped together

the

same verse:
all.

salus

in. 'a.

servavisti

me.

|

Bacch.

879

|

vel ego, qui
i

dudum (ili causa coeperam excruciare animi, quasi quid lilins
Langen,
Beitrage
zur

i

r.

Kritik

a.

Erklarung

.1.

PI.

335;

Kamp?,

.-.

21.

Vol.

i
i

Prescott.

-Thought and
Ep. 389

Versi

in

Plautus.

24*3

mens
mea,

deliquissei

me erga

(

o

filia

quom hanc video, mearum meabsens

tniseriarum

commonea; (Rud. 742)

Tn the following example meat

belongs

to

both aouns:

inscitiae

meae

el

stultitiae

ignoscas.

|

M.

G.

542

The possessive adjectives of the plural pronouns of the firsl and second persons occur naturally with much less frequency
than meus, tuos, suos, and cases of separation are proportionately fewer.

They are

subjecl to fewer special

regulations and

conditions: they are not enclitics; metrical convenience does not
affect their position so significantly; they are 1" be sure subjecl

to the principle of

pronominal attraction: 87

salute

te,

vicine Apollo, qui aedibus
te,

propinquos nostris accolis, venerorque
tonstricem
DOvisti imstraiu?

(Bacch. 172)

Suram

(True. 405)

eredidisti

qua re filiam nostram? (Ep. 597)
volel

meritissumo eiua quae

faeiemus, qui hosce amores
(As. 737)

nostros dispulsos eompulit.

nam meus
ibi

formidat animus, aostrum tarn diu
(Bacch. 237)

desidere neque redire (ilium.

In these eases there
verses.

is

little

to suggesi

the entity of individual

and its noun in every example bul one bracket other words, and the word-group thus formed shows no Such word-groups appear in very simrespect for verse-unity.
possessive
ple form in Altenberg's

The

examples from early prose;

in

Plautus's

verse

— we may

not here enquire into the causes
i

they are often

N'etc

also it/o,

on mill, mi us

at

or

near the beginning of sua

verses.

This does not Haut. 711
:

"7

happen

to

appear

in

our examples, but

i

ut

quom

narret senes

voster nostro esse istam

amicam

gnati, non credat tamen.

-l<s

University

oj

California Publications.

[Class Phil.

elaborate, as the

cant

illustrates. 88 The signifiemployment of such interlocked phrases the poel so seldom allows them to escape into the second is verse. II true that when the ordinary attributive escapes, verse-unity seems more often to reassert itself than when a posInst

example above

fad

is

thai

in

spite of the

sessive

is

separated, bul such difference as there

is.

is

accounted

for by the relative frequency of the possessives, the

tmemphatic
which

nature of most

of

them, and
to

their

metrical

character,

draws some of them
as noster,
ences,
hick of

the extremities of the verse.

Inasmuch

voster are subject only to the second of these influ-

emphasis may properly be regarded as the mosl
in

important factor
88

the separation. 89

from

In the cases of mens, tuos, suos, usually the possessive is separated its noun only by a verb (Aid. 733-734, Ps. 849 850). There are a few eases of more elaborate interlocking:
ad ilium quae tuom /ilium unice unicum.

perdidit,

pessum dedit

tibi

(Bacch. 406)

Special effects are usually
ease
is

produced by such arrangements; an interesting

sicut

tuom

vis

unicum gnatum buae
et

superesse vitae sospitem

superstitem, (As. 16)

Eere the couplet
the

is

securely

linked

tenet her

by the connection

between

and the adjectives of the second; but as the connection is predicative, the unity of the second verse, reinforced by the sound- and sense effed is paramount; tuae is separated from vitae, and the separation also divides the group tuae superesse vitae, but if our conelusions above are correct, the weak force of tua< made the separation
first

noun of the

verse

.

inoffensive to the

Roman.
quid
ais.'

Another interesting case

is

ecquam scis (ilium tibicinam meum amare? Ps. 182
I
I

The
weak

criss-cross

ecquam

.

.

.

filium tibicinam

meum

brings together the con-

trasted
to

objects and
interrupt

suggests the

father's

indignation, while
verses except
so

meum
far as

is

too
n
is

seriously the unity of the

already interrupted.
evidence dees not
tional
factor.
suffice

to

include Greek

influence as an addi-

The ways of expressing the possessive idea in Creek are more varied, and th uditions inherent in the words are different from those of their Latin equivalents. The fragments of the New Comedy offer
almost
no parallels to the separation
in

Plautus.
-i

In
, t

Mienander's (307 K.)
it
li
t

-

n

he art icle w

he pos

Vol.

i
I

Prescott.

— Thought

and

Vers<

in

Plautus.

249

V.

These special conditions also affed
adjectives, so thai
it

many

other
for

pronominal

is

not
in

surprising that,

example, the
in

demonstrative pronouns

their adjectival
to

usage are second,
adjectives.

frequency of separation,

the

possessive

Again
to

however, the cases of separation, viewed with referei
total

the

few.

number of occurrences of such adjectives, The fact that these words are pronominal

are extremely
as well

as ad-

jectival

may

in

many

cases have mitigated the separation; and

the effect of Wackernagel's law and of the law of pronominal
attraction,

working either separately or

nounced
Bach,

in

many

of our

examples.
taken
in

in common, is very proThe studies of Langen,

Kampf,

Kellerhof,

connection

with

Wacker-

nagel's different and broader point of view, explain the position

not only of the demonstratives, but
of the indefinite quis and
its

<>f

the determinative,
If

and
find

derivatives.

these

words

their natural habitat immediately after the introductory
of the sentence,

word
is

and

if

the closeness of the adjectival relation

something much
nagel's

less
is

binding than the operation of
-it
is

Wacker-

law— as

quite evident

remarkable thai cases of

separation are so infrequent.

The examples
nunc

that

follow will

show the pronominal word

in

close connection with the introductory
is

word

of the sentence; so

immediately followed by hoc:
nunc hoe deferam

argentuin ad hanc,
sessive genitive

quam mage amo quam matrem meam.
suggest

(True. 661

l

may

an amplifying idea.
In

I

have not

found any
ur,

cases of ipdg, coq thus separated.

but they are

less

Euripides, however, parallels frequent than in Blautus; e. g. yi vf Med. s77. possibly with emphasis
;
|

'»>,>

itfv

<

[ph.

T.

L362),

|

(Iph. T. o$5), iraoav

-

etc aip>
,,,„

<™> 9eolt ayalfun
.

[ph. T. 1480).

8

-

925
.

[on 725, Med. 746), rap
1

rac cag (Ion 1271),

riwa

.

.

(Med. 792),

-

(Med. L132).
likely to

On

the whole, inherenl

have been the dominant

features of the Latin words are more influences, although the agreement points

to

an inherited separability.

250

University of California Publications.
nunc hoc vos00 rogal

[Class run..

Qomen Trinummo
nt

fecit,

liceal

possidere banc

aomen fabulam.

(Trin.

In close association with qui or with si:

oam servom
hi

misi qui ilium™ Bectari solel
is

meum gnatum:
qui ilium

ipse hanc destinavit

fidicinam.

(Ep. 486)

dudum

coneiliaverunl
|

niihi
l

peregrinum Spartacum,

Poen.

769

oimis ecastor facinus minim

meo

viro sic

illi conlibitum insimulare false facinus tarn malum.

est.

qui

siel

(Amph. 858)

qui ad

ilium deferal
<i.

meum
ut
si

erum, qui Athenis fuerat, qui hanc amaverat, (M.
illic

131)

eoneriminatus

sit

advorsum militem

meus conserves,

earn vidisse hie

cum

aliene escularier, (M.

<i.

242

nam
edepol ne
illic

si ille

argent urn prius
xelusi
si

hospes hue affert, continue nos ami

sumus.

(As. 360)

pulchram praedam agat,
(Aul. 610)" 3

quia illam inveneril

aulani onustam auri;
>li

tilii

propitii sunt,

nam

hercle

si

istam semel amiseris
restitues locum.

libertatem, haud facile in

eundem rusum

(M. G. 701)

The regularity with which the separated noun
ni

in these

and many

her eases stands at the beginning of the second verse, with
it

many

words intervening between
mystery followed by
sonal
its

and the pronominal adjective

solution

pronoun and an my son." etc. Such interpretation may he purely subjective, 84 but in any case
. . .

— conveys — "him appositive
iiber

the effect of a per-

1,0

But A reads
'in
Tlii—

vos hoc.
ct'.

verse

Leo,

Bemerkungen

plautinische Wbrtstellung

u.

Wortgruppen 430.
'•'-

ilium qui P.

n Features reinforcing the unity of the verse are apparent in the previous example {hospes hue), ami here particularly where aulam onustam
auri an- undoubtedly
Linked

together by

a

unity of sound-effect:

ft'.

Aul.

763, 617, 709, 809, 821.
"<'f.
relative

Appuhn,
clause
in

I.

c

59.

In a case like th.-

following, the noun with
i

its

the second verse seems te
first

intensity the sulist ant val ef-

fect of the demonstrative in the

verse:

• quam
mulieri

facile

el

quam fortunate
v<>lt

evenit

illi.

obsecro,

quam
is

liberare

amator."

(Ep. 243)

Occasionally this effect

brought out explicitly:
te
(

em

istic

homo

articulatim concidit, senex,
188
I

tuns servos.

Ep.

Vol.

i
i

Prescott.

— Thought

and

Versi

in

Plautus.

251

the rather constanl attraction of these pronominal

words

to the

second place

in

the sentence, withoul

regard

to

any association
the sepa-

with the noun, was certainly the usage of the spoken language;
ii

is.

therefore, unlikely thai there

was any violence
the division
in

in

ration

by the verse comparable
. .

to

English of

"that

.

son of mine."
force,

Many pronominal
a
:

adjectives seem to

have an independent

closer affinity

with other words

than with their substantives

in

any consideration of verse-unity

they are almost non-existent.
In isolated cases the separated demonstrative appears in com-

pany with nam and quid;
than with the noun
eonicit
in

the indefinite quis

and

its

derivatives

are similarly connected with the
:

introductory
liliain

particle

rather

11:1111

is

Qlius

uavem miles elam matrem suam, (M.
quid
hi'-'
ii.ui

G.

ill)

poterat de suo

scuex obsonari

filiai

nuptiis?
:,u
'

(Aul. 294)
'

sed speculator ae quis

llill(

;u "

;i1

'

l:

"'

va

:iut

dextera

uostre consilio senator adsit

rum

auritia plagis.

(M.

<;.

607)

nam

cogitato,

si

quis hoc gnato tuo

tims servos faxit. qualem
nescio quid
istuc negoti
alius.
(

haberes gratiam?
nisi
si

(''apt.
est

711

dicam,
|

quispiam*"

Amphitruo

Ajnph. 825

ibo in Piraeum, visam ecquae advenerit
in

portum e\ Epheso aavis mercatoria.

(Bacch.

23"))

ecquem
recalvom ad Silanum senem, statutum, ventriosum, (Bud. 316)

Some examples have already
pronominal
particles

illustrated the juxtaposition of
a

words:

in

the

following case

lyrical

passage
that

and pronouns are grouped together
the separation of istam
quid? hie
etc.,

in a

way
if

readis

ers of Plautus will admit to be almost inevitable;

there
it

any

violence

in

—which
but

I

doubl
I

is

easily

Usually punctuated
case the stress
is

unnecessarily,

think; in any

on quid, and hie is uot the first word of the sentence-unit, as the metre slums. '"There is. however. Dothing regular in the collocation

see

ii

samples

in

Prehn, Quaestiones Plautinae de pronominibus indefinitis

7-8.

252
Porgiven

University of California Publications.

[Cum

Phil.

for the sake of scelestam, scelus,

linguam and the

di-

vision only brings into relief thai

phrase:

quid est! quo modof iam quidem nerele ego til.i istam (Amph. 556) scelestam, scelus, linguam abscidam.

There are other examples of the demonstrative which have none f the attendant features illustrated above, but which for
other reasons are hardly to be considered as disturbing the unity Among these is a small group of cases in which of the verse.
the noun
is is

in the first verse,

and the demonstrative
in

in

the sec-

ond verse

defined in a relative clause-, thus the second verse
the
firsl

simply amplifies the meaning of the ih.hu
'

verse:

immo
quem

apud trapezitam situm
dixi

est

ilium

Lyconem,' (Cure. 345)

continuo arbitretur uxor tuo gnato atque ut fidieinam illam quam is volt liberare, quae ilium corrumpil tibi,
ulciscare atque
it

a

curetur, usque ad

mortem

ut serviat.

(Ep. 267)

oboluit

marsuppium

huic istuc quod habes.

(Men. 384)

So, too, with idi

m
(luxit

uxoTem

hie sibi (Cist. 177)

eandem quam olim virginem

hie compresserat,

There

is,

of course, oo
sed

more

separation in these cases than
ipse advenit

in'"

optume cecum
ille.

hospes

qui has tabellas attulit.

(Pers. 543)

According to the earlier punctuation with a comma after singularias, the following verses would no1 concern us:
eis indito
ist

catenas singularias

as.

maiores

quibus

sunt

Luncti

demito.

(Capt.

L12)

But

Bach (Studemund-Stud.

11

322)

offers

valid

reasons

for

is diffireferring istas to the previous verse; such a separation examples are wide of the mark. cult to parallel, and Bach's

There
"

is.

to

be sure,

a

contrast suggested by the juxtaposition

or

in

quid aisl tu nunc
ilium

si

forte

eumpse Charmidem conspexens

quem

til.i

istas

dedisse

commemoras

epistulas,

(Trin.

950)

Vol.

L]

Prescott

Thought and

Verst

in

Plautus.

253

ot
ii
:

istas ;iinl
is

maiores, which

may accounl

for the separation, bu1
if
it

certainly very vaguely suggested; the demonstrative,
is

follows the noun and

in

the second verse,

is

usually attended

by features thai more evidently justify separation:
quia Lstue quaesof an
ille

ille

quasi ego?

is

ipse quasi tu.

quasi ego

"

si

vis," inquil
ei

" quattuor sane dato "
rei

(tum)senex (St. 552)"

dies

haec praestituta

est,

proxuma Dionysia?
(

cms
tu

ea

quidem

sunt.
et

Ps. 58

abduc hosce intro

una autricem simul
te.
(

iube hane abire bine ad

Poen. L147)

qua pro

re argentum promisil hie tibif- si vidulum hune redegisseni in potestatem eius, iuratusl dare in Iii talentum magnum argenti. (Eud. L378)
i

Such analogies as there are to istas according to Bach's punctuation must be found in these examples: the contrasl in ilh cms:"-' and the resumptive force of hane and its ego, haec
.
. . . . .

proximity
verses;
eius
is
it

to hinc -all these features reinforce the unity of the

may
is

be doubted whether in the

lasl

example hunc

.

.

.

a
it

feature that has any bearing upon the separation of

hunc:
that

an unusual example

(cf.

Trin.

ll'JM-4 according to
to

Lindsay's Oxford text), and the nearest parallel
I

Bach's

istas

have found.

A
tics

few examples do not admit of grouping under characteris-

common
postremo,

to
si

any Large number of cases:
ut

dictis aequis perduci

vera

haec credas
L98)

mea

dicta, ex

factis uosce rem.

(Most.

haec sunl atque aliae multae

in magnis dotibus (Aul. 532) incommoditates sumptusque intolerabiles.

an

te ibi

\

is

inter istas vorsarier

prosedas, pistorum arnicas, (Poen. 265)

" The whole context should be read
quasi in.

to

gel

the

play on quasi ego and

"Contrasl with this verse a later reference

in

the same play:

nam
praestitutast,

olim

quom

abiit,

argento haec dies
[ue

quoad referret aobis,
be included,
if

dum

rettulit.

(Ps. 623)

Most.
correct.

618 should

Leo's supplementary

reading

"j:>4

University of California Publications.
mulier profecto aataat ex ipsa Mora

[Class Phil.

;

nam quaevia

quae morast aeque, mora minor ea videtur quam quae propter mulieremst.
alia
]>n>
ili

(M.

('<.

1292)

Lmmortalea, similiorem mulierem
ut

magisque eandem,
In

pote quae nun

sit

eadem,

|

M.

(;.

528)

none of these

is

the separation

violent: effective antithesis,

Long words grouped in one verse, alliteration, the combination
of associated ideas
.
.

ea

.

.

.

qua<

propter mulieremst, 1 ™ eandem
all

.

eadem

—-are

compensating features,

of which testify to

the individuality of the verse.

noun

The freedom with which the relative is separated from in Oscan and Qmbrian (Norden, Kunstprosa I 181

its

n.

Altenburg,

De sermone

pedestri

Etalorum

vetustissimo

530)

suggests that the relative adjective has an inherent separability;

and

in several of the cases there

is

some evidence of unity

de-

spite the separation:
nimis paene manest.

aegotium agere,
ut
in

id

mane quod tu occeperia totum proeedit diem. (Pers. 114)
(M.

tabellis quos eonsignavi hie heri

latrones, iluis

denumerem atipendium.

Gr.

73)

cui servitutem di danunt lenoniam

puero, atque eidem
Lta

si

addunt turpitudinem,

(I's.

7(>7)

nt

occepi dicere, ilium quern
hie eius

dudum

(c

fano foras)
(Rud. 1065)

lenonem extrusisti,
di

vidulum eccillum (tenet).

ilium infelicent omnea, qui pest hunc

diem
(Poen.
149)

leno ullam

Veneri

unquam immolarit
peiurio

hoatiam,

qui hie litem apiaci postulanl
mali, res falsas qui

impetrant apud iudicem,

(Bud. 17)

quin tu tuam rem eura potiua

quam

Seleuci, quae tibi

condicio aova et Iuculenta fertur per
ni

me

interpretem.

|

M. G. 951

I

lierele

diffregeritis talus postliac queinque in tegulis
1

videritis alienimi.

"(

M.

<i.

\~<<>

)

qui

omnea

ae

amare

credit,

quaeque aspexerit
viri

mulier: 103
"" Tliis
si\.'

eum oderunl qua
exhaust
at
tintl

qua mulieres.
note

(M. G. 1391)
at

dues net

ilVc-ts:

mora

the ends of sueees-

veraea;

and mora

end of the aecond verse

may

lie

in

close

re

lation

quaevia alia ut' its nun verse aa well as with ""Similarly, l>ut without separation by the verse in
with
a

the

next

verse.

quemque

milite line videritis
in
I'..

hominem
( I

in

nostris tegulis.

(M.

(I.

L60)

V ah, n

mulit

n

s

i

ii

in

).

Vol.

i|

Prescott-

Tim,,,, hi

and

Versi

in

Plautus.

The aniformity with which the separated substantive stands
the beginning of the sec
I

;ii

verse

is

rather striking: the mystery
its

.suggested by the anticipatory relatives makes
of
;i

solution worth]
in

prominenl position; the resumptive pronoun
;it

many

cases

makes the noun
negotium
. .

home
s

in
.
.

its
.

verse
ibus,
is

in

spite of separation
. . .

.

.

.

id,

latrom

puero

eidem, lenonem

.

tins: other evidence of unity

visible in the

fad

thai

mali

(Rud. IS)

belongs as

much with

the qui of
in

its

own

verse as

with the qui of the preceding verse, 104 and
.

th

sho

mulier

.

.

mulieres

I

M..

G. 1392). 105
is

Occasionally the interrogative adjective

similarly separated:

quem amplexa sum

hominem?

(M. G. L345

I

euia ad aures

vox mi advolavit?

(Rud. 332) m

The

indefinite adjectives, too,
;i

now and then appear
and
cases

in
its

verses

by themselves; such

separation of nescio quis from
1

noun
quis-

hardly impairs verse-unify:

"7

of aliquis and

quam, 106 by the very nature of
nam
sibi

the words, are inoffensive:

lamlavisse hasce ait architectonem
exaedificatas insanum bene.
(

nescio

quem

Most. 760)

atque ego

illi

aspicio osculantem Philocomasium

cum

altero

nescio quo adulescente.

(M.

<;.

288)

aliquem arripiamus, prandium qui percoqual

(Men-. 579)

ego si allegavissem aliquem ad hoc negotium minus hominem doctum minusque ad hanc rem callidum,
1

i

Ep.

r.

Leo, Analects Plautina: de figuris sermonis

I

20.
its

The position of
verse,

mali
I

(18)
''"iii

and bonos (21), each
r

at

the

beginning of

brings oul

he

rast.
<*\'.

For the repetition of mulu

ecce ad

me

advenil
pulchrior:

mulier,
lM

qua
lint

mulier
the
I

alia

oullasl

(Merc.
usually

Cui

Mss.

same

or

similar

phrases

occur

without

separation: Triu.
107

15,

'arc 229, Mere. 864.

Cf. T,r. A.d. 657-658.
Cf.

Ter. A.I.

716 717.

256

University of California Publications.
peiorem ego hominem magiaque vursute malum

[Class Phil.

aumquam

edepol

quemquam
ueqi

ridi
tael

quam

hie

esl

Simia;

(Ps. 1017)

go

riorem beluam
fce

vidisse

me umquam quemquam quam

censeo.

(Most. 607)

There are some noteworthy features: the balanced alliteration ii: Merc. 579-580; in Ep. 427 aliquem is really substantival,
in ''somebody else," and the next verse a separable elemenl the two cases of quisquam, the regular juxtaposition of words
;

ending

in
110

-quam

is

illustrated.
is

109

Alter,

when separated,

in effect an

added idea:

at ego nunc, Amphitruo, dico: Sosiam servom

tuom
(Amph.

praeter

me

alterum, inquam, adveniena faciam ut offemlas domi,

612)

eho tu,

quam

vos igitur filiam
(Ciat.

nunc quaeritatis alteram?

602)

The separation (omitted in A)
verse in
the
(

of alterum from Ionium in the following couplel
is

more

violent;

cf.

the same phrase within the

Bacch. 1184, an anapestic passage, and in frag. 4 of

laecus:

immo etiam

si

alterum
(Ep. 7518)

tantum perdundumst, perdam potius quam sinam

So

in this ease of

tantulum

unquam

intermittit

immo, Chrysale, em non tantulum tempus quia eum nominet.

(Bacch. 209)

It would be difficult to prove that any emphasis is attained by the position of these cretic words at the end of the verse and

by their separation from the substantives. bu1 the context

in

111 cadi case suggests considerable emphasis upon the adjectives.

I'liis

hardly needs further evidence, but to quote only our
100,
I

play,

cf.

Men. L92,

17.

518, 613, 780, 959.
in in

""c\\ the separation of alius

St.

449-450; Tor. And. 77s 779
Inst
is

{alia

aliam), Hec. 365 366, Ad. 52 53,
ul

the

two cases preceding the noun.
confirmed
215, and for the demonstratives

The

inherent

separability of these pronominal adjectives
:

by the same phenomena in Greek Menander 567; Philemon 7; 58;
tls; 580; 71s; for
-»;

cf.

above,

p.

Diphilus 30; 3; for abr6q
for barii
3.

Menander 140;

for

Menander 325, 8; aXXog Menander 535,

Menander 117Menander 393; for rooavra

VmI

'I

Prescott.

— Thought

and

Verst

in

Plautui

VI.

The numerals,

also,

have an independent existence which may

accounl for the cases of separation by the verse-end:
scelestiorem ego
.-1111111111

argento faenori
hie

Qumquam

ullum

vi<li

quam

mihi annus optigit.

(Most. 532)

nullum"-

fecit.

(

Bacch. 982

1

feral

epistulas

duas, eas DOS

1

sigiii'inus.

quasi siut a pat re:

(Trim

77 1)
ei

filiae

duae craut. quasi aunc meae sunt; eae eranl duobus nuptae fratribus, quasi nunc meae sunt vobis. (St. 539)

Alexandrum magnum atque Agathoclem
linn

aiunl

maxumas

res jjessissi':

<

1

1

1

i

1

mihi

fief

tertin,

qui solus facio facinora inmortalia.
hie dico, up

(Most. 775)

fanum Veneris
|

qui mulierculas

duas secure adduxit,

Rud.

128

sepere aliae mulieres

duae

]><>st

me

sic-

fabulari inter sese

I

Ep. 236)

mulieres

duae innoeentes intus

hie

sunt, tui

Lndigehtes auxili,
in

(Rud. 641)

quia
rediisse turn

viis

patriam
el

domum
(St.

rideo bene gesta

re

ambos,

te

fratrem tuom.

506)

eaptivorum quid ducunt secum!

pueros, virgines,

binos, ternos, alius quinque;

(Ep. 210)

ubi saepe eausam dixeris pendens advorsus oeto
artutos, audacis vdros, valentis rirgatores. ubi saepe ad

(As. 564)

Ianguorem tua duritia dederis octo
\-

validos lictores, ulmeis adfectos lentis rirgis.

(atque)
solere

auditavi saepe hoc volgo dicier,

elephantum gravidam perpetuos decern St. esse annos lion quinquaginta modo, quadringentos filios lial <-t atque equidem lectos sine probro:
;

"-

\'i rliiini

emendation), Ter. Ter. .M. 85.

milium without separation by the verse Ullus, with nequt precedi Eun. 88.

in

Bacch. 785
rated

(by
in

University of California Publications.

[Class Phil.

The

lasl

passage

is

from

a

eanticum, and

is

ascribed by Leo to

;m amplificator.
are worth noting.
i

In the other examples some attendant features
Respect for unity
tertio
is

shown
in
in

in

duae

— duobus
of

st.

539

.

duo

,,; and (Most. 77.n.
1

the

isolation

adjectives and nouns in the see

vrerse

the two examples

from the Asinaria.
the noun, or
if
it

In most of the cases the numeral follows

precedes the separation brings into prominence
(As.
:"

bnportanl

elements
in

564, 574, St.
4

168).

A

few eases of

omnes are

place here
(

hariolos,

haruspices

mitte ciniics;

'Ainpli.

1

L32

|

certum

quin edepol serves, ancillaa domo (<as. 521) est omnia mittere ad te.

deartuasti dilaceravisti atque opes
confecisti omnes, res ac rationea meaa:
ita res divina mihi fuit:
res serins

(Capt. H72 ap.

Nonium)

oninis extollo ex hoc die in alium diem.

(Poen. 499)

Rhodum venimus, ubi quas merces vexeram (minis ut volui vendidi ex sententia: (Merc. 93)
servoa pollieituat dare
.sues

mihi omnis quaestioni.
ubi ego

(Most. 1087)

omnibus
(Ps. 771)

parvis magnisque miaeriia praefulcior:

atque

minoris facio prae ill", qui o mn ium Ep. 522 leguin atque iurum finer, eonditor duet

me

;

I

fateor

me omnium
minimi
preti.

hominum

ease A.thenia A.ttieia

(Ep. 501)

The

first

six

examples,

in

which

the adjective

follows

in

the

second verse, involve no violation of verse-unity; the last three. however, are certainly, from an English standpoint, more destructive of
582.)
It
is

unity.
likely

(Cf.
that

also

the separation
is

of

tot

in

Poen.

the adjective
in

more separable than
for
this
is

the corresponding

word

English: the evidence

found
1,3

in

general, the apparenl separability of numerals in
.

and

Cf. Poen. 898.

For omnes
Ter. Ad. 229
aliquod, Ter.
(ef.

Ln

Te,-.

ef.
in

An. I. 77. 667, Bun. 1032.
Plautus, above,
p.

Similarly complures,
,

.,

plurwni

237); paud, Ter. Eec. 58;

Phor. 312.

Cf. Norden, Aeneis

Buch VI. 390.

Vol. l]

Prescott.

Thought and

Verst

in

Plautus.

259

in the

usage of the corresponding words

in

Greek verse.
of
of

1

Ceris

tainly
likely
.such

the explanation of the separation
to

numerals
the

more

be
in

found

in

inherenl

qualities

numerals as

than

such attendanl features as the metrical convenience

of the cretic

omnium

al

the end of a verse.

VII.

Proper and improper numerals, pronominal adjectives, and
in

particular possessive adjectives were separated withoul essen-

tial

disturbance
to be
l»nt

of

verse-unity.

This

inherenl

separability
in

seems

proved not only by the treatmenl of these words
by the evidence furnished by early
Latin

Plautus,

prose,

and by Greek prose and verse: the nature of the evidence suggests
ih.it

this separability

was an inherited

trait.

The opera-

tion of

Wackernagel's Law and of the law of pronominal attraca

tion

is

further manifestation of the looseness of the bond that
t<>

binds pronominal adjectives
possessive adjectives

their nouns.

The separation of

was probably promoted by the unemphatic
;i

nature of the words, which suffered

loss of their individuality.

These conclusions do no1

differ essentially fr

those of A.ppuhn
I

In the treatmenl of attributive adjectives, however,
that

hope

something

lias
set

been

gained

by

an

attempl

to

interpret,

within the limits
aration.

by the paper, the passages illustrating sepattributives

We

found thai

following the

noun ami

separated were regularly expressions of ideas panging from predicative to amplifying, and the separation was usually attended by features that reinforced the unity of the vrerse. We found,
too. that

when
fr

the separated attributives preceded their nouns.

although

an English standpoint the unity of the verse was
in

For the ordinary aumerals

Ter. cf.
.

Bun. 332, Phor. 638, A. I. 46.

For Greek examples
128, 3;

ef.

Philemon

4.

Id; 28,
_

548;
104,

Philemon L2;
7:

B9, 7;
17.

r,Diphilua
532,
l:
iro/Lte,
1.

;

Menandei 535, '': 282; 382; 397; 179; other aumerals, Menander 7. 1; 357; 5 index 67, 2; Menander 13, 2; 17 Philemon 91, 7: Menander 292, t: 363, 7;
,

Menander
>-.

593.

And

Eor

aumerals

in

early

Latin

prose, ef. Aitenburg,

524

ft.

260

University of California Publications.

(
\

'i-^ Phil.

impaired,

there

were almosl

always associations of sound or

sense thai reasserted the unity of the verse;

more often the unity

was apparenl

in

the organization

<>f

the thoughl

Hum

in

the

superficial colligation

resulting from sound-effects.
effects repis

We may
resenl
cially
true.
t<»
;i

no1

always be confidenl thai the resultanl

efficienl

causes: in the matter of alliteration this

espe-

The confinement,

in

most

cases,

of

alliterative

groups
eration

siimlc verse attests the entity of the verse, but allitin

is

seldom more than an incidental factor
evident

separation:

usually other and stronger factors appear along with alliteration.

Metrical convenience

is

in

the position of some words,
cretic

especially those of considerable length,
possessive adjectives of
position convenient
ration.
for

words, and
:

the the

pyrrhic and iambic measurement

such words

may have conduced
is

to sepa-

Again, however, other factors are usually discernible.
effect

Indeed, the total

of

a

verse or couplet
is

a

product of

many
a

factors:

it

is

not easy to say that one
it

more important
that
in

than another.

But
is

seems

to

me noteworthy

so large

number

of separated attributives, the unity of the verse, if
correct,
is

my
like

interpretation

effected by internal organization

rather than by superficial colligation.

So much
\

so that in cases

maxumo
the

\

me
\

opsecravisti opere, opturna
conferri
a
si

vos video opporT

tunitate, tesseram nize

vis

hospitalem

prefer to recog-

beginnings of

\'n^T

technique

rather than

admit
factors

metrical convenience and alliteration as really dominant
in

the separation.

Such

c.-iscs

are rare; nor

may anybody deny
The
effect
is

the essential unity
in

of verse, the practical identity of verse

and thought,
often

the ex-

amples under discussion.
hut in

crudely simple,

many

cases the poet
little

man; he shows no
means
same
cate
;i

is far from being a clumsy craftscompetency in making verse-unity a

of bringing into effective relief associated

thoughts and

sounds; and occasionally he uses the beginning and theend of the
verse, the

beginnings of successive verses,

in

ways

thai

indi-

consciousness of the opportunities, not merely of the limi-

tations,
It

presented by verse-unity.
also significanl that

is

we can

find so little positive proof of

Vol.

l]

Prescott.— Thought and

Verst

in

Plautus.

261

the influence of his Greek sources: 118

lie

Beems rather

to be

workfash-

ing out his own problems

in the spiril of Ins

own Language,

ioning his verse with nice adjustmenl of sound-effects peculiar
to Latin,

often producing

;i

oeal balan<
;i

r

antithesis which has

yet to be proved to result from

study of Greek rhetoric, and
set

happily conserving, even within the limits
the

by verse-unity,

simpler

forms of interlocked

word-groups,

which

are

as

characteristic of the organizing

power of the Roman mind

us

any phase of

their political administration.

These same wordthe barriers, and

groups, however, must sometimes break

down

maxumo
t<il<
,

\

m( opsecravisti opere,
<-<ni(<iTt
si vis

<>/>hini<i

vos video opportuni
poinl the

tesseram

hospitah

m perhaps

way

which Leads

to greater

freedom.
is
it

Only after further investigation

safe to take the historical
is

point of view and ask ourselves what
tion
in

Plautus's precise posiIn

the development of verse-technique.

the answer to

that question

we must

not be too hasty in placing him aear the
in

beginning of art-poetry
sion
is

Latin: the comic verse under discus-

the most capacious of the

commoner forms
the
effects

of metre; and
to

this verse

conveyed the conversational Latin of the day

an

audience that
thought.
conditions.

must catch

at

once

of

sound

and

Epic verse and tragedy were created under different

Some

of the simple directness of Plautus's verse

i>

perhaps

to

be attributed to these conditions rather than to the

chronological

proximity of the Saturnian

verse.
in

Bui

in

the

present paper we have been interested only
1115

suggesting some

Without further investigation of Creek technique the statement must remain in this vague form. It would be easy to find parallels from Euripides, and some cases from the New omedy, of Plautus's postponement
<

of adjectives and nouns to the beginning of the second verse, and of post-

ponement for antithetical

effects, hut

the running over of the though! to the
try,
i

caesura of the second verse, familiar to readers of Greek tragic
the exception rather than the rule
to
in

is

Plautus; uor are the features

common

Ureek and Plautine verse too hastily
of
the

to be regarded as merely imitative

in Latin verse, especially in the case of antithetical effecta

[nvestigation,

particularly

technique

of

Aristophanes,

Euripides,
pr.

and
le

the

Not
more

Comedy, based upon sympathetic interpretation, must
precise statement of Plautus's relation to his models
in

any

these respecta

262

University of California Publications.

[Class Phil.

ways
i(ui

of interpreting

a

small pari of the evide]

thai bears

upon

the question which Leo has answered, forestalling the investiga
t

of the subjed

in

his

admirable statemenl of the historical

position of Plautus in this phase of verse-technique.

i

x

i

>i

:x

Accius .".us ir, 225 aote. Adespota, LO. Ataxpw Upol-ivov, given
citizenship, L51. Alcaeus, 2, 9. Alcaic st rophe Eorace Alcinan. 2, 9.
(

Athenian

Prov. Cons. 20, 47, 68 Qote. Sulla 13, 38, (17. Sulla 30, 83, 68 aote.
Tull.
v,
I.".,

36, 68 aote.
I,

M.

ii,

27, 70, 59,
6,

i',s

note.

),

17."..

Amphictyons, Board Anacreon, !».

of,

139.

Antigonis, established, Ahtigonis and Demetrias,
I

141.

abroga
of,

in

1

1

of,

Ml'.

Verr. Verr. Verr. Verr. Verr. Verr. Verr.

ii, ii. ii,
ii, ii.
ii

1.

17.1.

77..

2, 2,
3,
::. :;.

ii.
I-;.

4.

88 aote. 40, 99, 68 aote. 13, 32, 68 note. 40, 91, 68 aote. 72, 169, 68 in.tr. 7,1, I. 68 aote.
L5,
1 1 I

Antipatros,
153.

arehon,

date

L40,

Archons,
list.

Athenian,
L32.

chronological

Ariphron, :'.4. Arrheneides, arehon, date of,
L53.

Hiatus in Greek I'... Clapp, Melic Poetry, 34. lomparative grammar, Concessive sentences, Intensive, 40, simple, 37, 13, 4 1; with 43, 4 4
(

'.»•">.

;

140,

imlicat ive apodosis, 66.

Concessive
35
in.

si

clauses,
in
in:'.,

in

Plautus,

Asklepios, catalogue of dedications, 169; first dedication by the to, 147; priests of, chronological
list,

Conscription menl of,
i

Athens,

abandon

131

;

quence, Athens, before

priests of, official se144.
<

lorinna,
I

7..

In.

temel

i

ias,

established, Ml.

Jhremonidean War.
1
;

Demetrias ami
tin ii
I

Antigonis,
12.

170; change of governmenl in, 14!); eleven ribes in, 14i! fall
of,
17.4
;

i.l'.

I

Macedonian money
139; of war with
1

in,

I

Hagoras of Melos, ligamma, 2.

1".

147;
tiaics.
L58,

reflections
L59.

tnagis

I

dgammated
in
::.

words,
hors,

in

1

Aratos,

other

.-iiit

I"

:

in

[omer, s Pindar,
:

Attalis. creation of,

12.

Diogeiton, arehon,
I

.late of,

l»:7.
17.."..

Attic coins, old
1

st

\

le

of abandoned,

>i,

.inc. h.n,

aivlu.il.

date

of,
"t'.
|

18.

t

hiatus in. 1. Bendis, duplical inn of cult. L57. Bopp, I''., influence on Language study, 95. 3 Gaecilius 223 Et , 222 aote. 'luuikles, arehon, date of, 161. Charinos, arehon, date of, L50.

Bacchj

lides,

!»;

Ekphantos, arehon, date Alcaic Eleven syllable
177..'

159.
>.

Bon
st

Elision, in
1!»U.

II..

race

's

Alcaic
L70.

rophe,

Ephebes in 305 i. 164, Ephebe lists. 162. Bphebe system, made
ir..;.

voluntary,

Ohremoniuean
17(i.

War,
ii.

.late

of.

139,

Erinna,
Att.
17.
13,

1".
17.
I.
1

Cicero,

Ad
7,

18,

:'..

7.7.

Etiam
Etsi,

si.

16.

Arch.
Caecil,

68.

13

i

Caecil, L3,

68 aote.
<;*>

Eubulos, arehon, .late of,
Ei>pud5r)s
'AXcueife,
1

l!'.

169.

L9, 62,

note.
lis

pries!

of

As-

—Lael.
liii|>.

-

Div.

i.

24,

i--.

212 aote.
7.ii.

klepios,
n..tc

iii.

Pomp.
i'7.

17.

L04, 68.
81, 68
1

-

Mil. 27, 72, 68 aote.
:•::;.

Ferguson, W. S., The Pri A aem Method of Asklepios Athenian Archons), Dating
131-173.

—Pis.

aote.
35, 68 note.
1'ul.l.

Prov. Cons.

l.

Glaukippos, arehon,
Phil.,
v.. I.
1.

.late of,

156.

Univ. Calif.

CI

[263]

huh
Grimm,
J.,

X.

influence

on

language

1,

11.

Btudy, 95. Hiatus, after a diphthong ot long 32; after a shorl rowel, apparent, 3 vowel, 32 34 Greek in Bacchylides, L; in Poetry, L-34; in Bomer, \l,.|,,(inter1 Pindar, 1. 2; in verse), L92; relative frequency in Greek authors, 2.
1
i

9,

20,

lis; 6, 6, 11-; L19; 11. 2, L19;
II.
1:1.

'•'.

1.

L19;
I

L3,

I.

119;
1.;,

ll

1

.';

16,

31,

L19;
17.

is.

L19;
L3,

L6,

54,
2:;.

119;

I

I

;

66,

L19;
i,

Sat.

l.

Hi;

1.

ill;

1.

;

25, 111; 1, 28, 112; 1. 50, L12; CI. 112; 1. 68, 112; 1. 7". 112; I. 98, 112; 1. 117. 112; 1.
1

1-1. 11:;;
2, 56,

2,

s,
2,

Historical
8;

grammar, 96. Homer, 27; digammated words

11-;
2,

104,

in,

li:;
:;

:

I::-.
3,

U3; Hi;
114;

ll-; 2, 32, li:;; L13; 2. 119; 1. 113; 3,
1

hiatus in, I. Horace, Alcaic strophe in, 175. Horace, ('arm. i, L, 2, 119; L, 27, 119; 2, 9, Hi'; 3, 22, 119; 3, 24 119; 3, 30, 119; 4, 1, 120; 4,:.. 120; 7. 7. 120; 7. 15, L20;
7

26, L13;
3,

38 52, 113;
3,

3,

40,
3,

114;
66,

45,
3,

56,
2..

Hi;

HI;
3,

98, 114;
3,

98 112,

114;
3,

99,

100,

HI:

111;
;;,
1

101, 111; 3. in:;, 114; 3, L05, mn. Ill; 3, ins. Ill; 3, Mil,
1

16

L20;
11,
2,

8,

L0,

120;
11,
5,

8,

1

I,

114;
li:,;
4.

3,
1 1

111,
">
;

HI:
1

:;

-

120;

120;

120;

12,
1

14; 4,
:,,

7(5,

4, 87,

L5;

12, 3, 120; 13, L2, 7. L20; 11 L20; It!. 8, 120; 22, 17, 120; 22 21, 120; 24, 2, 120; 26 (whole ode), 124; 26, 2, 120; 26 6, L20; 28 (whole ode), 124; 28, 2. L21; 28, 5, 121; 28 7 121; 28, 16, 121; 28, 18, 121; 28, 19, 121; 31, 8, 121; 34 ( w hole ode), 124; 34, 2, 121; 34,5, 121; 23, 12, 121; Carm ,i. 1. 17. 121; 1, 30, 121; 122; ... 12. 122; 6, 14, 3 12221, 122; 7, 18, 122; 8, 10 122; 9, 3, 122; 10, 9, 122; 10! 18, J 22; 11, L3, 122; 13, 122; 14, 11, 122; 14, 1."., 13 122; 14, L8, 122; 14, 21, 122;
.

in:,,

106,
5,

115;
102,

5, 7:;.

11.-,;

101,
6,

ll-";

ll-;
10,

6

4,

115;

18,

115;

8,

115; 8, 46, 115; 9, 24, 115; 9, 34, 115; 10, 49, 117,; 1. Sat. ii. 1. 17. 116; 1. 27,. 116; 52, 116; 1, 77, IKi: 2. 17. 116;
.,

28,

IKi;

2,

83,

116;

2,

88,

1

2, 104, IKi; 2, 105, 116; 95, 3, 49, 116; 3, 91, 117; 3, U6; 3, 141. 117; 3, 17,:;. 117; 117; 3, 199, 3, 191, 117; 3,193.

116;

117;
3,

3, 200, 117; 3, 269, 117; 283, 117; 4, 90, 117; 4, 94,

16 (whole oclr), 122; Hi, '.». 122; Hi, 17. 122; 16
22^

124;
Hi,
1!»,

16,

2.

13,

122;
16,

12:';

123;
iii,

is.

1.

123;
1.

20,

21.

123;

0, 6, 1, 117; 6, 59, 117; 117; 0. 101, 117, 126; 7, Si, 118; 28, 117; 7. 49, 117; 7. 7,1. 7. on, lis; 7. in.",. 118; 8, IIS; 8, 77,, US; 121: -Bpist. i. 1. is. 124; 1. 42. -2. 52, 124; 1. 07,. 124; 1. 124; 2. .",1. 12 1; 2. 40, 124; 2.

117;
til.

1

Carm.
2

1.

10,
3,

12:'.;

41, 123;
11, in,
1

"it

12::-.

49,

123;

!'•».

47 124 125- 3
I.

;

2.

54,

1

1

1.

127,;

2,

56,
(i,

K»,

127,;

1.

16,

127,;

123; 17, 12. 123; 27. 28 (whole ode), L24;
12:;-.

S3; 4,

28,

Carm.
d
1;:;,

iv.

•"

(whole ode), 129;
is.

•'•.

128;
I.

3,

124.
1.

128;
24,
;

3,

22,
I.

128;
7,
1

13,
>.

128;
29,

128;
9,
I

125; 6, 3. 125; 6, 4, 12V 0. II. 125; 0. 24, 12-; 6, 27. 125; s. 7 125; 7, -J I, 127,; 7. 7,. 125; 8, 12, 125; 127,; 7. si. in 7, 127,; 11. in. 125; 11, 21. 11. 27. 126; 11, 29, 126; 126;
12
2

128;
l,

12s
9,
1

7,

12V
18;
9,

1-0;
12.

12.

1.,,

126;
14. 13,

12,
8,

17,.

125,

L28;

5,
u

126;

16,

120;

126;
1

25
11,

(1,

128; 1" 128;
1

"I"'

'•

" x;
'

14
14'

12,

126;
14.
1

H,
22,

126;

1.

11.

11.
1

12V

L3,

20,

— Carmen
23,

128; 1. 28, L28; 1. 6, Saeculare, 127. lis; 2, Hi. L18; 2, Epode 2. 7. Us; 2. II. Ms; 2, 16, 11-:
128;

120; 14, 26, 1. 35, L26; is. 9, 124, L26; 126; is, 71. 120; 1-, ins. 126; 19, 21. 126; K.i. ll. 126;

120;

[264]

Indi

Episl
L3,

ii,

I.

8,
I,

L27;
L01,

I.

LI,

1-7;
2,

1.

Book
21,

iii,

2,

Us. L28;
36,

1-7;
L35,

1-7;

32,

L28;
69,

1-7; 2, 125, 1-7; 2, 58, L27; 2, L38, 127; 2, L27; 2, L59, 1-7: 2, L75, L27; 2, 207, 127; 2, 213, L27; Ars Poetica, 19, 129; 61, L29; 7i), L29; 111, L29; 17:;. 129; 359, L29; 393, L29; 167, L29; L27;
L51,
2,

II.
1

I. 17. 126; L28; 62, 122; 65, L15; 71, L2]
I

!•_';

371,
l-l.

1

L3;

L24,

l

L2;

1

19,

L15;

Ill:

123;
i

i;

1

1!);

111',

[bycua of Rhegium, L0. Ictus, in Eorace's Alcaic strophe,
L93.
1

L13;
L26,

ndefinite 2nd singular, 82.

[ndo-European languages, 95. oscripl Lones Graecae ii, 310, 151 I": A.I.I. Nov. 373 b, L38, ii.
1

981, 1025, 1027, 1034,
128,

1

785, L19; 871, 117; 894, ] 904, L26; 915, L3, !6; L27; 939, 955, 123; 959, 1-7; 969, 129; 971, 1-; 992, 122; I"..:;. 17, 125, 128; 1026, lis. L028, L21; I":;:.. L22; L16;
I
I 1

569, 830,

1-7; L29;

581, -17.

941,
L27;
I

1

I

13;

1

L16;

1

129;

I":'.-.

129;
.

115; 1037, L045,

ii.

A.I.I.

Nov. 567
L45;
L5;
ii.
ii,
1

b,

L40;
15;

ii.

766,

L38,
b,
1

A.I.I.
1

Nov.
ii.

766
835,
ii.

767,

ii

8
,

859, 25]

ii\
e,

L48; ii. 836, L38; L67; ii'. 17s l>. 138; 5 164; ii 371 c, L66; 385 e, 196 373 c, 158; Li
Hi.

b,

,

.

and

ii

A.I.I.

453

b,

168.

archon, date of, 141. Jason, archon, date of, 154.
Isai.is,

Junggrammal

iker,

103.

Kallimedes, archon, date of, L55. Kiiiion I, archon, date of, 159, 166. date of, L59, 166. Kimon Ii. arc! Kleanthes, headship of, L54. 1!'. KXei7[^r7s], secretary in 276 5, Kleomachos, archon, date of, L55.
,

1

L20; In. -J, L049, 127; 1059, 126; L058, L24; 125; 1063, 117. 11-, 125; L068, 117; L067, 126; L082; L24; L26; L071, 1085, L22; L090, L28; Book iv, 304 (329), L27; L29; 378, L20; 450, L16; 115; 160, 122; 509, 127; L20; 571, 120; 576, 119; 124, 128; 712, 129; 789, mil', L13; 867, 117; 1005, L034, L18; 127; 1023, 116; L109, 1080, 113; 1155 L169, 113; 1129, HI; L162, III: 1156,
1

117;
1

I

"'.".

L066,
L23,

L16;
376,
158,

548,
-

1

L5; 15;

I

L071, 120; 113; 1177.

Lucretius, Book i, 8, L22, 128; II. 128; ID, 126; 36, L12; 72, 126; 84 LOO, 117; L27, L26; 74, 1-1 L38, 129; 200, 119; 211, L12; 228, 117; 257, L22; 259, 118; 305, 120; 275, 122; 276, 1-1 311, 116; 318, 128; 326, 120; 11 1. 116; 422, 114; 474. 119; 639, 1-7; 641, 127; 656, 121; 680, lis; 923, 11!); 926, 126; 927, 121 L22; 928, 117, 120; 947, 127; 929, L15; 936, 111 1111. L25; 1031, 12] Book ii, 1, 125; 7, 123; 8, 122; 10, L16; 11, 123; 17. 116; 27, 123; 29, 11-: 30, L22, 126; 34, 123, 1-4; 37, 122; 39, L27; L9, 123; 50, 11-: 264, 126; 3] I. 17. 128; 413, 122; 430, 11-:
;

Booh v,l, L25;

;

:

;

L198, II!); L282, L15; 1". 121; 37, 127; 115; 83, 125; 203, L19; 82, 204, 120; 206, 126; 220, 122; 259, 251, 11-: 256, 119, 1-1 115; 272, 119; 326, III, 120, L28; 381, 121; 386, 128; 109, 128; 727, 120; 737, L! 746, 120; 7-:'. L29; 821, 114; 839, 126; 864, 119; 936, 118; 939, lit: 951, 125; 962, lit: 963, lit: 979, L26; L007, 112; L26; L044, L16; L034, 125; 1085, ill: 123; L084, HI". 1-7: 1113, III; L108,
I-:
:

123,

1-1;

1114,

1

122; 122; 1204, 125;
1

582, 117; 636, 120; 643, 119; 646, L15; 7":;. 116; 831, 117: .'i'7. Hi'; 1031, 121; 918, 1-1 1167, 123; 1040, 125; 1104, L29; 1170, 125; 1172, 117;
( :

1119, 121; lit; 1144, 125; 1213, 127; 1230, 11): 1429,

HI-. 116; 1127, 121, lit: 1151, 117: 1218, 1276, 115; 1454, L13;

25

27,

,. 17. 111. 125; 25, 125; 116; 50, 115; 58, 115; 111, 120; 129, 115; 152, 11! 1:'-: 247, 1-1 259, 121
:

[265]

Index.
note B3 Asinaria, v. 16 17. 245 note 88); 81, 233 284
I
I

121; 421, 122; 429, 122; 608, 11'.'; Hi:'.. 120; 1114, 126; 117; 1222, L187, L29; L149,
L18;

I,

52); 111-13, 244; 112, 243, 84 120-21, (note 245 83);
I

Awowa[s
AvfflOeos

M<

\\t\ revs],
1

priest
priest

of of

note

i"i

;

153*,

ll

I

note
164,

I

I

;

Asklepios,
1

19,

L69.

TptKOpi'irtos.

1:5. Asklepios, ,\ sanias, archon, da1 e of, L40. 10. Lysiades, archon, date of, Lysias, archon, date of, 159. Melanippides of Melos, 10. Merrill, W. A.. Od the Influence of -129. Lucretius on Horace, Mesogeia, importance of in 3rd
I I

55; 210 13, 223 (note L93, 93; II (note 4); 24] 12, 237, 35) 83; 248, 234, 318, 10, 66; 360 61, 250; 386, 243; 387, 245 note 82 403, 36; 405, 40, ss in; i| ||:;. I;,, 17. 71. 72,
212,
13)
; ;
i

213

(note

I

;

:

|

88;
134,

124-26,

223
i

(noti
:

1

1

1

Ji-

i

note 82

170, 22'.':

century, B.C., 160.

Mesogeia,
161,

Koivbv
162.

tCjv

~SU<roytiwv,

Mirari

|

mirum

),

in apodosis, 86 s

,

in Language, 106. Nine-syllable Alcaic (Horace), 184. Nutting, II. <'.. Studies in the SiClause, 35 94.

Mixture

i

(

12. break in, Ubios, archon, date of, 161.
(fficial

order,

1

'Qvf)Tu>p

MeXiretfs,
1

priesl

of

Ask-

lepios,

16.

HdrcuKos,
14.",.

priest

of

Asklepios,

Peiraieus, decline of, 157. Phaidrias, arc! date, 167.

Phaidros of Sphettos, politics
L50.

of,

528-29, 77; 536 38, 92; 564 65, ^.s; 7,74 75, 257, 258; 1/7,7. 589 90, 52; 603, 38; 699, 59, 60; 720, HI; 72,7 38, 247; 744. si; note 13) 763, 240, 241 noti (note) 783 82, 223 7s:,. 245 (acte 83) ^7s 55; 920-21, "Jilt 894, II Mmte m te 27 933, 38, 66; Aulularia, v. 12, 233 (note 58) Hi 17. 233; 21-22, 218, 22,:; (note 58); 30, 233 (noti 7". 209; note 34); 75, 222 77 78, 224; 98, 40; 100, 41, 47; 114-17,. 223 (note 35); 172-73, 220; 195, 162, 220; 236 (a,,tr 65); 209, 80; 226 127. 223 228, 50 (ante 2); 247.
i

:

;

I

:

I

|

;

I

I

;

:

|

;

<t>(\oxdpr7S

Philippos, archon, date of, 150. 'Oadev, priest of Asklepios. 16. Philoxenus of lythera, 10.
l

<

I

In

Hi

tii

s,

98.

Pindar, digammated words in. 4. Pindar, diphthongs and long vowels
in

hiatus, 15.

Pindar, hiatus in, 1. Plautus, Amphitruo, v. 21, 44; 13 l. 245 (note 82) 135, 245 (note 82); 313, 94 (note 52); 336, 58; 391 92, 93; 450, 42, 71, 72 27 198, 88, 180, 213; 210; 556 57, 27,2; 612 13, 256; oote 33); 621, 78 675, 50 note 2) note, 94); 701, 92 te 703, 51, 83; 704, 223 37 705, -1 (note 40) 819, 825-26, 93 234; 251; 849, note 50); 858 59, 250; 880 81, 78, 80; 891 92, 57; 908, 37; 947 18, 61, 62; 977, II; 980 81, 210; 1048, 10, 88; 1051, II L063 64, 236; 1070, 213; 1107!», 223; 1132 33, 258;
;
:
I

(

;

i

i

;

;

;

note 42) 254, 38; 276, 222 (ante 34); 289, 24E 41 294-95, 311, 251; 82); til. 60; 320, 5 te i. (note 380 368, 236 (note 65) 2) s|. 83 (note); 421, 43; 489 90, 226; 498 99, 224; 523 24, 51; 532-33, 27,2,; :,\\AW^. 221 I, 555, 40, 69; 557, 55; 610 250; <;17. 250 (ante 93 oil' 50, 222,; 21. 7s (note 33) 2:0 (note 93) 7:;:; 34, 70!). note 211. 217, note 83), 248 7i;::. 250 note 93 768, 88) 77ti. te 50) II 93 809, 27,o (note 93) 821, 250 (note Frag. 3, 209 note ' 93 note), '.'I Bacchides, v. .".7,. 92 7,1 i; 11. 236 m te (also note L28, l". 17, 66, 70; 17265) 7:;. 247; note 23) 179, 38, 68 209-210, 256; 230, 212. 213 (a,,te 13); 235 36, 251; 27.7 38, 2 17; 279 80, 211, 22! (note 39); 364, 39; 365, 38; 395, 2::ti (note 65) 406 7. 243, 245
85
;
l

I

;

:

:

1

;

;

1

I

;

1

;

;

:

:

1

;

1

<

;

1

l

;

l

;

;

[266]

Indi
(note

(note 88); 83), 248 oote 39) 140-41, 82; 179 80, 56; 529, 85; 80; 563, 64; 590, 212, 213 be L3 635, 51; note 13 697, tO, 17, 69; 698, 86; 73] 32, 9] 777. 245 (note 83) 785, 257 (note) 818-19, 61, 62; Ml' 43, 246; 879-80, 246; 880, 245 (note - !) 882j 212; 885, 10; ss 7
426, 447-48,

M

I

<

:

istellaria, \. ::. 'ill. '17. Us. 7".

in.

52;
;

27,
.
|
:

:i7. ,;;.

71

:;.;,

38; 98 99
24,
-::s
:

j

i;,-,

;;s.

:.;

;

I

:

207;
18
l.

177 7s,
76,
"
i

252;

L83 84,

76;

I

1

;

;

;

;

53; 321, 90, 91; 513 note 35) 523 :;:;. 228
( ;

I

18)

;

,te

78)

:

547,
(inn.

243,

245
586,
i

(note

-

;

.

38; 901, 243 (note 7m 913, 84; 934, 212, 213 (note L3); 941-42, 225; 973 74, 257, 258; 982 83, 257; L004, 42, 17. 71. 1HU 11. 212; L013, 38; 88; L022, 227; L026, 212; L045, 40, 66; L102, 40; 11-1. 79;
; 1

238; 242
(

;

aote note

H».
(

ills., note 76 601, 2 15 602 03, 256; 609 82) 244; "17. 52, 79;
|

;

;

-

1

'

:

683, 5]
15
i

note
772,
'J

I

I

;

734,
i

i

te

;

l.".

I,,

te

83 Curculio,
I

;

\.

::.

12;
I'll.
(

L60,

I.".:

I

L72, 60

ao1
I

I

!);

— Boeotia,
(
i

L84 85, 93, L193, 38,
I

256;
te
1.

L91,
;

13;

59 90;
7-i;
:

;

142, 85
L86, 60;

•"

2)
I

229, 255

note 42) ; ] 15, 40; 226, r,-, note 106 246
I

:

Frag.
I.

\.

5,

226
(

17.

54;
Hi

259,
i

II;

265,

-

-

laecus, Frag.

256.

-Captivi, \. L2, 37; 27-28, 52, 77. 80 note) 82-83, 219; 98 99, 224; lnuiil, 76; 112-13, 252; L6, 84; 140, 244; 111. 245
; 1

Hi fce 150, 226; 82) 159-60, 230; 169, 208; L79 80, 230; 202, 82; 206, 58; 221, 82; 223, 38; 259-60, 59; 321, II; 417L8, 7:: noti 121-22, 240; oote 53] 32, 236; 9, 41 543, 13; 599, 91, !•_' (note), 609-10, 222; 612, 90; 613, 94; His. 80; 632, 236 i.92
;
l

;

i

l

i

:

:

65
38,

l

665 66, 229 (also 672 7::. 258; 683 - 1.
:

-Us 69, 53; 59; 303, 89, 90; 321, 86; Hi. 252; ::i7. 245 note 351, 89; 382, 210; 238; 429-31, 237i 130, 10, 212; 12, 7:'. 73; 190 91, 240 630, 246; 648 19, 219; 241; 701, 7!»; 709, 24] Epitlicus, v. Ill \-:. 225; "I'll note 58 202 03, M. 257; 236-37, 'i. 250 (note 94 225 279 80, JIT 313, -In; 341, 22
ii-tr
:
|

345-

82)

;

-it,
I

-

l.-.'.i

.

19,

I

I

:

668,
158,

;

i

I

:

237;

.

246;

88;
ii;
;

711-12,
IS;

52,

-."I

;

742,

389

38; 744,
854,

842,

13;

850, 59;

82)

906,
v.

873 75, 207, 245 59; 996, 86;
I

237 91, n< te 67 (note M.. 247; Ml. 245 D7 28, 255, 256; 451 52, 82)
:

L034,

86

-

13);
;

180-81

223
Casina,
:

unit,. 37

;

93, 10; 255, 225 (note A4) 262-64, 226; 269, 93, 94 279 80, 243 (note 80 :il l. 38; 309, 225 (note ID
:
.

is-. 24£ (note 80) 250; 188 89, 2£
.

;

522-23, 258; •l note
;

543,
.

90;
599,
71.

I

-

38, 66; 324, HI 345, 94
;

II
|

note note 52)
i

l

;
I

335,

-.

247;

-

58,

89,

90;

ID
77.;

II.

ill.

52] 22, 258; "-in. 75; 542, 86; 721, 82;
lis.

528 29,

357 62; 57, 60;
;

88;
ii7l.

;:::'

10,

610, 23]

D.
;

72,

7::.

mi

l-j.

257;

591, 80; 668,

B2; 706 "7. -Frivolaria, IV

I'll.

742 13, 86; 7»17 218; 795, 16; 806, 15 -7s Q( te 33 i, 89; 811,
957, 13, 16 992-93, 235,

Menaechn
I".
I

i.

\.
"is.

213;
:

Is. 213; 26. 213; 213;

I

1".-:.

82;

1

12,
15,

80;
I

1

12),

17;

236;

13;

238,
-

ll

1006, 223

225

:

-

Indt

(note
79;

L09) note
-

;

102,
i

120,
I

245

d< te

211; 117 83);
I.

18,
I

I

'.

te);

1

127; 26,

227;

17.

256

L09),

180,

187, -I
!

243, 245 222; 518,
te

158, 78, 80; L200, 227; L207us. 7s; Hi: L218, 223 L209,
I

dint.' 37)
12(12,

:

I,

256
i

L09)
i

HI;

;

L233, 209 (note 7,: L263, 7,H;
I

note L2 566, 80; 576, 85 670, 38; 613, 256 (note L09) note 83) 740, 245 746, 35, 38; 751, 40, 17: 760, 59; 780, note I": i; 844, 94 (also 256 oote ."-I I; 862 63, 220; 959, 256 oote L09) L048 19, 79; L060, L2, 89; L082, 213; 1088 (note 53); L102-03, 90, 233 L120, 213; L125, 213; 213;
;
; |

;

121»2 in. L345 16, !54; 225; 255; L356 7,7, 62; L391 92, 254, nut,- 52) 27,:,; 117. HI 129,
I
|

;

1

1

l

:

note L-3) 86 Mostellaria, \. 12. :;7; r,r, 7,7. 60; |ii 17, ins nit. 253; 229, 12. 89; 211 42, 12. 7:i: 351, 41, 12, 71(1; 361, 210; 393, 89 21, 235, 236; 1112. 56; 50] 04,
i

;

1

:

L132 33, 240 (note 73) Mercator, v. 93 94, 258; LOO 01, 255 (note L65) 124-25, 223 L55-56, 80; 206, 35) 226; 262-63, 230; 298-99, 85; im. 91, 92 (note) 130, 57; 445, 85; 197, 59; 517, 55; 564, 7,7s, 94 note 52) QOte 579 80, 255, 256; 595, 42, 72: 622, 78; 636, 38; 650-51, (in; 692-93, 7.;); 694, 42. 7:: 7 11. 85 (note 42); 817-18, 219; 819, 38; 838, 40; 841, 66, ii7, 68, 71 852 54, 222, (note (note L06) 35); 864, 27,7, 890, 92; '.hi; us. 93; 941, 7s
; ; : ;
I

;

m

,

.

(lu.tr) 532 33, 27,7: 580, 94; 607-08, 256; 609, 13 din. 13 note) note) 618, 253 (note LOO); 637, 212; 2 17, L3 (note); 666, 638, (note); 670, 2-12: 752, 235; 7(1imH. 255; 777,-77. 236, 2717. 27,s 827 38, 78; 854, 13; 912, 2Ki; H14, 38; 931, -1(1; Hi:;, Hits. Its:;. 22(1; 997, 212; 221 247, (note 82); L087 88, 211
;

227
93,
1

;

-

:

:

;

;

(note 82), 271s. 245; L0! 11(17. id; (note 52) Persa, v. 39, 241; 40, 42. (id; 43, 99-100, note 33) 234; II, 7s
;
,

:

222; 220

(note 33) 964, 227; 52Miles Gloriosus, v. 9 1", 21 7:;. 64, 66 56-57, 211 te) 73-74, 254; 75, 235; 88, 21! 111 L2, 251 L31 32, 65 52, te), 250; L56 7,;. 254 160, 254 (note L02) L88, 10 05, 222: 226, 227; (note 44); 238, 213; 242-43, 250; 288 89, 27,;,; 293, 61 298, 38; 306, 38; 383, 213; 391, 213; 107, 43; II 150-52, [2, 213 (note ID: 232; 17:; 71. 211; 475-76, HI. (12: 502 03, 211 508, 222. 227,; 2.71: 532, 43; 9, 217, oote 82), 217: 551-52, 231 563 64, 245 (note 83), 2 Id: 7,71. 60, "H. 66 (note) 607 08, 25] 631, 37, 38; 635, note 83); 642, 220; 673, 82; 685 86, 58; 687 22s 701-702, 250; 717. 213; 738 39, 243 te 80); 711. II: 717. 38; 763 64, 7::. 240; 7!Ui. 217, (note 83) 803 04, I". 72; 1117. 84; 951 52, 27, 974 75, 213; !»7D. 227. 228; 1111
; I

105-06, 220, 227: 110-11, 114-15, 254; note 29) 117 is. 234; 157-58, 219; 272.
i

:

;

;

;

;

;

;

;

:

1

:

;

:

:

-

I; 438, 212 43; 282, 42; 2(12. 17 (note 12): 449-50, 82; note 7,:, 75, 22:;; 512, 22:; 543-4 1. 526, 212 (note 12) 27,2; 571-73, 22:; (note 35) 594 95. (12. 63, 64; 601, 12: (111. so; 612-13, 92; (177,. 13; 724, mi; 825, 80; 771; 51, 38; 59, Poenulus, v. 12. 208; 83 8 l. 208; 89 90, 22d Id:;. 244; KM. 2 17, ( n »te 82); 165-66, 212: 191 92, 245 82), 246; 207 08, 80; 212-13, 83; 226-27, 236 265 dd. 27,:; 330, 89; 2 12. 44; 38 351, 56; 274. 38; 406 "7. lis 20, 22:: te 68 te), "71: 27s 472-771, 499-500, 209;
I

1

(

;

;

:

:

:

!

1

:

;

:

••

;

1

:

516-17, 54, 74; 550, 58, 63; 615 Id. 220, 227. 582, 258; (127 36, 82, 83; 670, 212: 699 7(i(i. 22.:; 7(17,. 210; ih.tr 59) 7H7. 89, l»n; 712 I. 212. 212 (ih.tr 72] 2l\ 92; 728, L3) mi; 729, 60; 722. 212; 769 70,
( ; 1
:

[268]

Indi

250; 793, 210; 812-] 864, 56; 896, 23! 906 te L13) 898, 258 921, 59; 958, 226; 963, 208; 988-89, 237; 997, 964, 240; 208; L047 18, 80, 226; 227; L063-64, 79 (note a< bi L080, 222; L08 l. L3, 16 his:,. L102, !40; 60; L2) L04 05, 238; L124, 208; 11 17 L92, 84; 48, 254; L162 63, 89; -lit. 89, 92 note) L201, 44; 1259 61, 223 53 L25] 52, L346, 240; (note 35), (note L375, 245 (note 70); L387, 246; 1377, 208; 82);
I

680, 86
'7

note
:

I

I

t

;

685
94;
17;

31
i

;

721,
II.
;

;

744,

7,1

(note
89;
66,
I"

i

I

;

B10
ti

1014,
104
I.

!0,
|

7,7;

13;

230
i

note), 254;
I
l

L075, 38;

I

;

1

1

I

I

;

L3fi 39 92 note) not( L56 7,7. 21 211, 224; 1235 36, 237; 1-71. 52 |; 1311,
;
I I
l

;

I

;

13;

1378 80, 253;
I

0; 1353, 35, 38, 39; L361, L392, 245 (note
l

100, 37,
\.

45;
16;
1

L39] 92, 77; Pseudolus, v. 27 28, 218; 51, 208; --7. 12; 58 60, 253; 85 86, 234 (note 80); 24 I. L2 13, 243 285, 62; 286, 92 44; 264, 42;
: 1

Stichus,
38, 66;

27, 71,

II.

I

I

;

t:

;

.

L2, 5 227; note 33 2); L51-52, 78 tilt. 257, 258; 171 72, 59 i* 248, 235; 287, l".
I

I

:

167

l

I

te), 94; 290, 35, 38; 291, note 346, 66; 294-95, 234 208; 430 31, 219 160 61, note 16 133, 86, -7 ivi' v'.. II; 245 171. 229; (note 38); (note 82), 248 in; 514, '.'I te t99, 62 616, 208; 623 24, 253; 626, 15; 650, 245 (notes 82 and 84); 697 98, -1 I. 221; 724, 2] 749, 732 33, 234; 740, 59, 9] sii; 7:,ii 7,7. 219; 7C7 68, 254; (note 771-72, 258; 783, 239 70 792, 42, 73; 794, 241; note 823 24, 207; 849 50, 245 note 88) 897, 235; 82), 248 960-62, 232; 974, 211; 976, 998-99, 218; 86; 997, 86; note 1017-18, 256; 1033, 50
|
I

i

L6,

-I

I,

24!

;

lit 50, 220, note 3] 150 a, 221
I

|

I

;

L10) 451, 22]

;

i

i

;

257, 258; 552
i

31); 506 ''7. 257; 510 512 L3, 64, 65; 539 II. 7,:;. 253; II. 7n 739, 223
II.
7,1
;
"

;

Trinummus,
i,,i;
(.-,

I

;

:

:

20-21, 250; note L06 1". 24 109 L10-20, 78; 220; L13, 80); III. 34); 148, 75; 347 18, 83 152, 212; 194, 22]
v.

255

-

i

I

;

98, 77,;

1

:

(

;

349, 8
-

:

383,

13;

II

I

I

:

L5,

-

.

135 36,

238;
38;
13;

165, 1-7. 6S

38;

474,
17.

4

193-94,

I

228; 507, 38; 516
531-32,
;

80;

-"l'7.

L040 II. 2) L090, 208;
;
I

I

208; L070, 7,1 L3, t3; L148, 7,

76,

77:
566,

;

:

208; L162 63, 208, 209; L241-42, 62; -Kudens, v. L7-18, 254, 255 (note mi 21, 255 (note l"i I; 39, 220; 71. 223 note 37); 74 77,. m ti -lit; L15 I':. 220 -11 158, 1-7,. L28 29, 257; L62-63, 219; 226; L59, 28, 72; 313,211; 316 20,211,25] 329, 332 33, 255; 79 (note 35 369 70, 219; 372-73, 209, 224; 421-22, 209 te 379, 65 224; 172, 92 6 (not- 52 I; 178, 21 94 552, 86 535, '.">: 516, 227; 566, 55, 7,.;. 63; (note
1

L52,

i:; 557 58, 59; 62; 593, t3; 600, 13; 697, 233 679, 77,ii 51, 22 7 is, 81; -I 771 75, 7.;::. II
I

-I

I

;

I

|

:

;

257;

i

;

7-1'
15 )

90
;

I

S

:

;

51,

252
-I":
;

5,

212,

:

;

213 212; 962,
13

965 66, 212, 213
L051,
I

83,

i

;

L059 60,
in-.-,.

92;
;

L081 82,
; :
-

210;
1101,

1

22]

1087-89

245
1139,

253;

M

213

"II.

I

:

I

Indi

III.. 245 (note 245 82); L184, L158, 212; (note 83) L87, 86; L85, 40, 67, 70; 36; Truculentus, v. 6, 90; 66, 38; 72, L40, 210; 83 85, 208 (note) 287 88, 62; 203, 208.; 234, 85 not* 211, 225; 293, 245 45, 224 315, 42, 69, 7n; 3 note 38) note 38 346, 224 39] note 83) 355, 244, 245 16] 62, 92, -us; 405-06, 247; 82; 527, 40, 69, 70; 591, 233 613, 38; 615, 38, (note 57) 39; (iiil 62, 249; 692 93, 77 (note) 766, 91, 92 (note), Ml 826 27, 815, 13; note 7,1 246; 830, 52; 833, 38; 854, 38, 68 note 23) 877, 38, '17; 80 v. 56-57, 78, Vidularia,
; I I

Subjunctive
tive

protasis,
in
Hi.

;i|inil(isis

with indicaPlautus, 50.

Tamen

<

t

<

i

,

;

Tamensi, Tametsi,
Tetir/as

46.
I

I

15.

;

Kc<pa.\7j9ei>,

priest priest

"f

As

i

I

;

1

I

klepios, 14ii. TeXtaias <p\v[e6s),
1

of

As-

l

;

I

;

I

;

Mi. klepios, Telokles, archon, date of, 152. Eorace), 189. Ten-syllable Alcaic note T< rence, Adelphi, v. 46, 259
l I

I

;

Is.
I

239

i

;

note 70)
85,
i

oote

M"/;

;

i

I

;

i

;

-

note); 68, 77; L06, 44. Polysl ratos, archon, date of,TL57.
I

IloXi'^ews

^.owievs,
1

priest

of

As
-

klepios,

1(1.
iii

Posse

apodosis, 53. l'ratiims uf Athens. L0. Prescott, H. W., Sum,' Phases of to the Relation of Thoughl Verse in Plautus, 205-262. versus future Presenl subjunctive,
(

po1 is),

- Heauton

258 note note L07 239; 657 58, 255 n. .IL08 716 17. 255 761, I- (nut.' ID; S91-91', L'.-.li; 77. Andria, \. 35, 239 note f0) 258 (note 111): 80 81, 245; note HI,; 667, 258 639, 7,1 77s 7:1. 256 note 110); Eunuchus, \. 88, 257 (note); '< 108-09, 238; L55-56, 239 332, 259 (note); 357, 224 70) (note 41 I; 524 25, 239; 532 note :;:!, 235, 236; 857, 220 30); 1032, 258 (note 11
-l".i.
;
I
I

52 53, 256 mite) 257 274 75, L14)
:
I

;

:

I

I

;

I

;

;

(

I

;

I

I

1

;

Timorumenos,

v.

711-12,

indicative, 87.

Ptolemais, creation <>t', 158. Ptolemais, given secretaryship

ir-

regularly, 143. Pure conditional sentences, 50-66.

247 (note 87). fce Terence, Eecyra, v. 58, 258 365114); 85-86, 226 (note) 66, 256 (note 110); 71.",, ."1
;

(note 3)

Quid

si.

89.
.1..
7.")

Richardson, L. Strophe,
1

Eoraee's Alcaic

_*o4.

Phormio, v. 312, 258 (nut,- 114) 44 7, 51 (note 3); 638, 259 (note); 736-37, 51; 93S. 22]
;
;

S.-inskiit.

95.
in. 2.

940, 221.

Sappho, 2, 9. Sappho, digamma
Schlegel,
Friedr.,
st inly.

Terpander,
influence on
95.
lan-

10.

Thersilochos, archon, date of, 155. OovydvTjs, priest of Asklepios, 149.
L69.

guage

Science of language, 95. Athenian, demes of, Secretaries, chronological list, 131. Secretaries to the Treasury-board n\- Athena, 15.
I

Timocreon,
Ti^okXtJs

.".1.

Elreahi, priest klepios. 149, 169.

of

As-

Timotheus, 9. Treasurers of
of,

Athena, secretaries

Serapis, priests of, 139. Short quantities, the outset of Eorace 's eleven syllable and nine syllable Alcaic verses, 203. Si (concessive), 37 42. si clauses, loosely attached, 85 86.
.-it

L31.

Urios, archon, date uf. 152.

Wl

,,\'

I.. The Whence ami ler, Whither of the Modern Science Language, 95
1'..
l

11 '.'.

Word

analysis,
of.

1"".

si

objed

clauses,
<

7.".

Zeno, death

L54.
cult, 156.

Simonides of leos, 9. Sophronistai, abolished, 166. Stesichorus of Eimera, I".

/ens Soter, duplication of

[ZanXos] NiKO/c]pdTOi']</)Xi,6i''s, of Asklepios, L68.

priest

[270]

Sign up to vote on this title
UsefulNot useful