Harvard Studies in Classical Philology, Volume 93

Google

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project

to make the world's books discoverable online.

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover.

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the

publisher to a library and finally to you.

Usage guidelines

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to
prevent abuse by commercial parties, including placing technical restrictions on automated querying.
We also ask that you:

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for
personal, non-commercial purposes.

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the
use of public domain materials for these purposes and may be able to help.

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find
additional materials through Google Book Search. Please do not remove it.

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner
anywhere in the world. Copyright infringement liabili^ can be quite severe.

About Google Book Search

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web

at |http: //books .google .com/I

IRVARD STUDIES

CLASSICAL PHILOLOGY

SDrrtii jtr a cotnatTEH or nrx classical

mSTBUCTOKS Of HARVARO VttlVRRSITr

\'oLt;Mri IV

BOSTON, U.S.A.
PlJBLtSHKn BY GINN & COMPANY

WnoS; OISN * COUPAMV
IT A M U>WMTK tliu.

'893

upKic; Otto HARicASSowrrz

ywt",

it^

Q

HARVARD STUDIES

IN

CLASSICAL PHILOLOGY

EDITED BY A COMMITTEE OF THE CLASSICAL
INSTRUCTORS OF HARVARD UNIVERSITV

- • » »

• ••«■•

Volume IV

BOSTON, U.S.A.
PUBLISHED BY GINN & COMPANY

LONDON: GINN & COMPANY LEIPSIC: OTTO HARRASSOWITZ

57 & 59 LUDGATB HitX QUKX STXASSB S4

1893

EDITORIAL COMMITTEE.

•

•

• • •

• •

•

t

,.••

••:.

?//!•!

• • • •

• •

•

• • •

•••••

• • •

m

• • •

•-,

• • •

• •• 4

t •

•

• • • <

V

• .

:•;!:

• •

••:

• • • •

%

•

• •

IV/V

• ■ •

•••:•

• •

\:V.

• • • a

• •••

■ • •

• •

• •

• • •

• • •

••••

• • •

• -

;:•

;.'.;•!

•

• •

• •

•

• • •

••:..

••

• •

•

• • ■ •

• • ••

•

••..*

?;//!

. •••

• •

••••

JAMES a GREENOUGH.
FREDERIC D. ALLEN.
MORRIS H. MORGAN.

1005 29

•..••.

Typography by J. S. Gushing & Co., Boston, U.S.A.

PRBSSWORK BY GiNN & Co., BoSTON, U.S.A.

CONTENTS.

PAGB

The Avkoi or Tibia i

By Albert A. Howard.

The Tragedy Rhesus 6i

By John C. Rolfe.

The Use of HERCLEy EDEPOLj ECASTOR, by Plautus

AND Terence 99

By Frank W. Nicolson.

Accentual Rhythm in Latin 105

By J. B. Greenough.

On the Omission of the Subject-Accusative of the Infin-
itive IN Ovid 117

By Richard C. Manning.

Latin Etymologies 143

By J. B. Greenotjgh.

On ireipap cXco^cu (2 501) AND THE Manus Consertio of the

Romans 151

By Frederic D. Allen.

Herondaea 169

By John C. Wrigln.

NOTES 201

INDEXES 211

ui

2 Albert A. Howard.

All of these instruments may be classed, according to their acous-
tic properties, as ' open ' or as ' stopped ' pipes. An open pipe gives
the complete series of harmonics, the octave, the twelfth, the double
octave, and the third above the double octave of the fundamental
tone of the pipe. A stopped pipe produces a fundamental tone one
octave lower than the corresponding tone of an open pipe, but gives
a series of harmonics including only the twelfth and the third above
the double octave of the fundamental tone.

FINGER-HOLES, THEIR POSmON AND NUMBER.

The notes of the musical scale lying between the fundamental tone
of the pipe and the first harmonic can be produced only by opening,
in the side of the pipe, holes of proper size at appropriate distances
from the end of the pipe and from each other. To produce the
diatonic scale through continuous octaves, beginning with the lowest
register of the instrument, six such holes are required for an open
pipe, eleven for a stopped pipe. To produce the chromatic scale
even more holes are necessary. Theoretically, the position of each of
these holes for a given scale can be mathematically determined, but
in practice the position of the holes is largely determined by experi-
ment, and the holes are often slightly out of place, the result of which
is a defect in the pitch of every note dependent on a hole improperly
placed. But every performer on a wind-instrument has it in his
power, by means of well-known devices, to correct minor defects
in the pitch of his instrument caused by improper location of the
finger-holes, and, no matter how perfectly his instrument is con-
structed, he is under the necessity of making these corrections in
pitch. That this was equally true of ancient artists and instruments
is apparent from a passage in Aristoxenus (ed. Marquard, p. 60,

chap. 42-43)-

The extent to which the pitch of a note can be altered by these
various devices is, however, limited, and the proper position of the
finger-holes, especially before the adoption of mechanical means for
closing such holes as were not in use, must have been a very difficult
problem for the ancients to solve. Doubtless many of the treatises
mentioned in Athenaeus IV. 80 and in other writers dealt, at length,

III sqq., and wherever in this article Egyptian instruments are mentioned the
facts are drawn from that source, unless statement to the contrary is made.

The Kv\A% or Tibia, 3

with this subject ; one treatise mentioned by Athenaeus XIV. 36, the
irepi avAcuv rpTfTtta^ of Aristoxenus, was perhaps devoted exclusively
to this subject. This work comprised not less than two volumes, for
the reference is Iv irpwr^ ir€pl avktav rpi/o-ccus, implying that, at least,
a second volume existed. The loss of this treatise is all the more to
be regretted from the fact that it was the work of such an eminent
authority on ancient music.

No treatise on the boring of pipes has been preserved to us even
in a fragmentary form, and I have found but two direct statements,
in ancient writers, as to the proper position of the finger-holes. The
first is from Aristotle, Prob. XIX. 23, Aia ri SivXafria ij n/n; 1% xnraTrfs;
^ vpSnTOv fikv on in tov i^fticrcos 1} X^pS^ ^oAAoficn; koI [17] 0X1; avfiffxi}-
vc/vat Sea Tncurcuv ; ofUMcos Sk i\€i koI ivi rcuv avpCyytav * 1} yap 8ta rot)
fuaov T^s trvpiyyoi ^ TprjfuiTOi <f>uivrf ry &' oXiys rrj^ avpiyyoi avfJLffiuwtl
Sea Traawv. in iv Tci^ avXx>i^ rep SurXaaiff Sioonz/Luxri Xafifidverai to Sea
waclav, Ktu oi avkorpvirax ovrcu XafJifidvovo'iVf opjUiOi Sc kcu ro Sea irhrrt
rw i7fUoAt(|». Irt ol ras oijptyyas appjorrofuyoi cig fikv rrp^ xnraTrjv oKpav
TOV Ktfpov ifjivXaTTovo'i, Trp^ d!k vrjfnpf /tic^pi tw "^fuatoi dvawkrfpolwnv,
opjoitoi 8i icoi r^v Sea vhrrt t^ "^fuokiff koI ttjv SicL TtTTopiov rep iiriTpiTf^
Sioon/fuiri Xafipdvovciv. Irt at ^v rotg rptycdvoig ij/aXTrjpiois ttjs To^
^irircurecDC yivofitirrj^ <rvfiff>ii)v6wn Sea Trcurcdv, 1} fiiv SeTrXoo'ca ovoti, 1} Sk
ij/iurcea r<^ priKV.?

From this passage we learn that the finger-hole by which the octave
was produced was bored in the middle of the tube ; the hole to pro-
duce the fifth was so placed that the length of the entire instrument
and the length of the portion above the hole should be in the ratio
3:2, and the hole to produce the fourth was so placed that the entire
instrument and the portion above the hole should be in the ratio

^ The discussion of the syrinx here mentioned I have reserved for a later place
in this article (page 19).

^ The text of this problem in the various editions is in such condition that cer-
tain changes in words and in punctuation seem necessary in order to make the
passage intelligible. The changes in punctuation are so obvious that they need
no justification. I have inserted t^ before oXiy and have restored avfi^vwoikri which
is found in one Ms. in place of ovfA^vwov^a. For adXocf the Mss. read AXXott and
for Irt of the following sentence 0rt; both of these changes were suggested by
Aug. Wagener, as I learned after I had myself decided upon them. Finally ol
must be read for oZ in the Mss. in the last sentence, as is evident from the ^ fUp
and the ^ 94 which follow.

4 Albert A, Howard,

4 : 3. Assuming that the finger-holes have a diameter equal to the
internal diameter of the tube itself, these are the positions which they
should theoretically occupy in an instrument of cylindrical bore.

The second statement is from a late Roman source, Favonius
Eulogius in the commentary to the * Somnium Scipionis ' (Cic. ed.
Bait, et Orell. Vol. V. pt. I. p. 412) : si tibiam cuiuslibet longitudinis
sumas, et octaua eius portione deducta cauemam imprimas, tonus
auditur ; si sextam decimam ex reliqua metiaris hemitonium conse-
quitur. This however is a mere ' rule of thumb,' and it is doubtless
of late origin, since it implies the existence of a tempered scale in
which all of the tone-intervals are equal, and in which the interval of
the semitone is half as great as that of the tone. Even as a ' rule of
thumb ' this is too inexact, for instead of * octaua * we should have
' nona eius portione deducta.'

From Macrobius, in Somn. Scip. II. 4, 5, we get the further infor-
mation that not all finger-holes had the same diameter : nee secus
probamus in tibiis de quarum foraminibus uicinis inflantis ori sonus
acutus emittitur, de longinquis autem et termino proximis grauior,
item acuta per patentiora foramina, grauior per angusta. This infor-
mation is of value since it proves that the ancient instrument-makers
were familiar with an important principle utilized .to-day in the manu-
facture of wind-instruments, namely, that a hole of small diameter
nearer the mouth-piece may be substituted for one of greater diam-
eter in the position which the hole theoretically should occupy.

The earliest form of avXo9, to whichever of the types of instrument
mentioned above it belonged, must have been, from a musical point
of view, a very imperfect instrument. Pollux, IV. 80, says that,
until the time of Diodorus of Thebes, the avX<k had only four finger-
holes, and Aero,* in the scholion to Hor. A. P. 202, quotes Varro as
having twice made the statement that the tibia anciently had only
four holes, while others maintained that it had only three. No Greek
or Roman pipes with only four finger-holes have, as yet, been found,
although they are not infrequently represented in works of art ; but
unquestionably, until some mechanical device was employed for

^ Varro ait in tertta disciplinanim et ad Marcellum de lingua Latina, quattuor
foraminum fuisse tibias apud antiques, et se ipsum ait in templo Marsyae uidisse
tibias quattuor foraminum, quare quatema tantum foramina antiquae tibiae habu-
erunt; alii dicunt, non plus quam tria.

The AiiKo^ or Tibia. 5

closing the holes not in use, the pipe, when used in pairs, cannot
have had more than four holes, unless the thumb was used to close
one hole.

Of thirty-four Egyptian pipes described by Loret * twelve have but
three holes each and thirteen only four holes each, while the primi-
tive pipes of the various ethnological collections show, with few ex-
ceptions, four holes each, — indirect evidence which helps to confirm
the truth of the statements of Pollux and Varro.

On an instrument with but four finger-holes it is, of course, impos-
sible to produce a continuous scale beginning with the lowest register
of the instrument Either, then, this lowest register of the avXos was
not used in the earliest times, or the musicians made use of a scale
in which several tones of the lowest octave were wanting. In the
absence of any direct testimony showing which of these methods was
followed, any attempt to solve the problem is simply a matter of
conjecture. Perhaps the simplest solution of the problem would be
to assume that the notes of the pipe corresponded with the notes of
the tetrachord, although in this case but three holes would be neces-
sary, since on a pipe with four holes five notes can be produced.
By neglecting the lowest register of the instrument and playing solely
in the harmonics a complete diatonic scale can be produced on an
open pipe with three holes, or on a stopped pipe with four holes,
with the exception that on the stopped pipe the octave is a semitone
too high.

It is, however, chiefly from the historical point of view that the
determination of the scale of this simple instrum/snt is of interest to
us, for the avAos did not long retain its primitive character, but was
constantly improved and perfected ; in fact, the names of some of
the artists by whom improvements were made have been preserved
to our time. Diodorus of Thebes is said to have improved the avAos
by increasing the number of finger-holes ; Pollux IV. 80, iroKvrrprjrw
IvoCrjat . . . irAaytas dvoiifdg T<p irvcv/Luxri ras oSous.

Whether or not this improvement was actually the work of Diodo-
rus, the ancients certainly did possess avAoi with more than four
finger-holes. This is clear, not only from works of art, but fi-om
the actual remains of musical instruments to be seen in various
museums of Europe.

^ Joarnal Asiatique, 1889, pp. 197 sqq.

6 Albert A, Howard,

Probably the increase in the number of holes was not a laige one
at first, for in the existing instruments the number of holes varies
from five to fifteen, and the oxiKin certainly did not reach its final
development until a later period than that of Diodorus, as is shown
by the statements of the ancients themselves. Pronomus, the cele-
brated musician and the teacher of Alcibiades, improved the avXoc
by making it possible to play on a single pair of instruments the three
musical scales in use at that time, whereas before his time a separate
pair of instruments was required for each scale. Pausanias IX. 12.
5, 'AvS/Has T€ l(m Upovofwv dvSpos auXiyoxivTos ciraycuyoTara c? tovs
iroXkovs* rccDS fi€y yc iScas av\Siv rpcis CKrcuvro ol avXTfTcu, koI rois fuv
avXrjfJLa rpikow ro Aatpiov, Bidft>opoi Sk avrois i^ apfiovCav r^v ^pvyiov
irrerroirjvTo 6L avXoi, to Sc KoXovfUvov AvStov iv avXots rjvXtiTO aXXoiOi^.
UpovofWi ^^v oi 7r/Da>ro$ ivevotfaev avA.ov9 ii airav apyuov[a% clSo9 cxovras
^TTiTi/Scuaf, irpioTO^ 3^ Bidff>opa cs roaovro fiikrf xnr* avXoi9 rpj\rja€ roit

avroi9. This same story is told of Pronomus by Athenaeus XIV. 31,
and although there are other ways in which this improvement could
be effected, the simplest method would be to increase the number
of finger-holes, closing temporarily such holes as were not required
for the particular scale in which the musician wished to perform.
It is also possible that some of the mechanical attachments to the
instrument, which will be described later, were at this time brought
into use.

In works of art of the later period the instruments are sometimes
represented as provided with a large number of finger-holes, as for
example in Plate 1. 1, at the beginning of this article, which represents
the pipe in the hand of a muse from a sarcophagus in the Louvre.
This pipe has at least nine lateral openings. Of the Egyptian instru-
ments described by Loret, two have five holes each, five have six
holes each, one has eight, and one eleven holes. Of four Greek pipes
in the British Museum three have six holes each, one five, and of the
four pipes found at Pompeii in 1867 two have ten, one has twelve,
and one fifteen holes. These instruments from Pompeii can very fit-
tingly be called ttoA vrpiyroi * or multiforatiUs^ adjectives which are not
infrequently used by ancient writers to describe the avAos or tibia.

1 iroXtJrpirrof, Pollux IV. 80., Antfa. Pal. 9. 266, 505. 5.

2 Multiforatilis, Apul. Flor. 3, p. 341, Met. 10. 32; multifora, Ov. Met. XII. 158,
Sen. Agam. 358; multiforabilis, Apul. Met 10, p. 254, Sidon. ep. 8. 9.

The Ai\<fe or Tibia, 7

BANDS.

When the number of holes in the pipes was increased to such an
extent that the fingers could no longer close them all at the same
time, naturally some device for closing the holes not in use had to be
adopted. In the case of the pipes found at Pompeii this device con-
sisted of bands of silver encircling the tube of the instrument, one for
each hole, and pierced by a hole which corresponded exactly with the
hole in the tube itself. These bands could be turned round on the
tube and made to cover the holes in it, as is indicated by the position
of the bands on the instruments as they were found. In some cases
the band is so placed that the hole is open, and in other cases the
hole is partially or entirely covered by the band. The bands are
fitted neatly together, so that originally the whole body of the instru-
ment was covered by them, and it is probable that Horace had in
mind such an instrument when he says, A. P. 202, ^^ tibia non ut
nunc orichaico uincta^ Two of th^ pipes in the British Museum are
also provided with metal bands, and the same is true of a fragmentary
instrument in the same Museum ; cf. C. T. Newton, Halicamassus,
Vol. II. pt. I, p. 339.

Similar bands are represented in at least two reliefs on sarcophagi
in the Louvre, one No. 240 of Frohner's catalogue, representing a girl
blowing on two pipes, each of which has a number of lines cut on
it to represent the places where the bands join each other, although in
this case no holes are represented in the bands ; the second, Frohner,
No. 378, a sarcophagus representing the Muses with their attributes,
Euterpe holding in each hand a pipe, the one in the right hand broken.
The instrument in the left hand is shown in Plate 1. 1, and the bands
and the finger-holes in them can be easily distinguished. Perhaps
the lines which are occasionally seen on the pipes in vase paintings
are intended to represent these bands (cf. Baumeister's Denkmaler,
Fig. 592, and von Jan's description). Not infrequently there is repre-
sented in works of art an additional device to enable the performer
to turn the bands.* This device consisted of a hook-shaped projec-
tion from the surface of the band and seems to have been called

^ Relief in Naples Mus. No. 6684; relief in the Vatican, cf. Baumeister, fig.
598; sarcophagus in the UfBzi Gallery, Florence, representing contest between
Muses and Sirens.

8 Albert A, Howard.

by the Greeks Kcpos or PofAfivi (cf. Arcadius de accentibus, p. i88, ed.

Barker) aXX* CKoorcp Trvcv/utan ovk dre^o)? ovS* (l/Aov(ra>9 rot <rrffi€ui
irrfffero ' KaOarrtp 6L rots avXot^ to, rfn^fiara cvpdfKvoi, ^iri^paiTCiv avrct
Kol xnravoiytiv ottotc /SovXotvro, K€p<uri ruriv rj ^ofuPv^iv v<f>opKiot9
(v^oAKtoif) €7rer€;(vaoravro, avo> koI Kara), ical ^v8ov re kcu c^o> OTpiffHuv"
T€5. Tavra ovTOHrt, KcJUcctVoif wirtp Kcpara ra o^/Licta CTroci/oiaro T<p
irvtvfiaTi, h^ Tt (ryrnjuoi kKaripta (rrffM.r}vdfA€voi. tovto 3^ to cv wnrtp avXt^
ioiKoq, oirtp €v8ov KOi lfo> orpc^cov CTri^parretv re kcu viravotyciv ro

The K€para themselves are not in existence on any of the instru-
ments which have been found, but the pipes from Pompeii have,
attached to many of the bands, Uttle sockets into which such arrange-
ments could be inserted, and these sockets were undoubtedly intended
to serve this purpose. Probably such a socket was originally attached
to e^ch band of these Pompeian instruments ; the oxidization of the
metal of the bands has, however, caused all trace of the socket to
disappear in some cases, although on thirteen of the bands the
sockets, or traces of them, are still to be seen.

The Kfpara are frequently shown in works of art in which bands
are not represented on the instrument.^ It seems reasonable to con-
clude in all such cases that the omission to represent the bands was
due solely to careless work on the part of the artist, for without
the bands the Kipara would be useless. Furthermore, the frequent
representation of the Kipara is conclusive evidence that the bands
were an almost universal addition to the instrument, although unfor-
tunately the exact name by which they were called has not been
preserved to us, at least in any recognizable form.

SIDE-TUBES.

Still another device is frequently represented in works of art, but
has not been found on any of the instruments hitherto discovered.
It consists of a short tube, often represented with a flaring end,
directly over a hole in the pipe. This arrangement is often rudely
represented in works of art, and has given rise to the theory that
holes not in use were often stopped with a plug (cf. von Jan in Bau-
meister's Denkmaler, p. 557). That this arrangement is to be re-

^ This is notably the case in wall-paintings; e.g. Helbig Wandgem^de, Nos.
56, 69, 730, 765, etc.

The KiiKji^ or Tibia. 9

garded as a tube and not as a plug is, however, shown conclusively
in bas-reliefs ; ^ for, wherever the sculptor has paid any attention to
detail in his work, the hole in the tube is clearly represented, and
inferences drawn from works of art in which the details are but
rudely executed are, in the absence of literary evidence, of very little
value.

On the whole, the representations of avW in works of art are as
accurate as could be expected, for naturally the artists did not strive
for absolute perfection in all the minor details of a musical instrument
which played a very subordinate part in the entire work. Such care
in detail as the artist has shown in representing the pipe in the hand
of the muse in Plate 1. 1, is exceptional, although of the greatest value
to the student.* The fact, therefore, that the artists have so frequently
represented these attachments as tubes is the best of evidence that
they were regularly tubes, and not plugs to stop the holes. These
tubes were sometimes, if not always, mounted on bands such as have

^ The following list includes only a few of the reliefs in which the holes in the
tubes are. represented : Paris, Louvre, Frohner's Cat. No. 85 (photograph in H.
d*£scamps, Marbres du Musee Campana, pi. 25) ; Frdhner, No. 378 (heliotype
from photograph pi. I. I, of this article) ; Munich, Glyptothek, No. 188 (reproduc-
tion from a photograph, Baumeister, 6g. 11 86); Rome, LAteran Museum, No. 751
(inexact engraving in Mon. dell' Inst. VI.-VII. 80, 2) ; Vatican Museum, sar-
cophagus representing Cupids as Muses (inaccurate engraving in Museo Pio
Qem. IV. 15); Florence, sarcophagus representing contest between Sirens and
Muses (cf. Baumeister, fig. 1704, from an engraving); London, British Mus.,
sarcophagus from Townley Coll. (inaccurate engraving in Anc. Marb. of the B. M.

n. 35)-

^ Engravings of ancient works of art are, with few exceptions, very inaccurate
in matters of detail, although they give, with sufficient accuracy, the main features
of the originaL Not one of the various engravings of the instrument shown in
Plate I. I, represents either bands or holes above and below the side-tubes of the
pipe. In the Museo Pio Qementino, Vol. IV. pL 15, the pipes held by the youth
who represents Euterpe have each six side-tubes; in the original there arc only
four represented. In Millin, Galerie Mythologique, Plate XX. Fig. 64, Euterpe
holds two pipes which might be mistaken for the rounds of a chair, while in the
original one of the pipes has five side-tubes, the other seven, and the holes in the
tubes are clearly shown. Even in such a work as the Museo Borbonico, Vol. XI.
tav. 59, the female figure is represented holding pipes with five side-tubes, while
in the original only four are shown. Wherever in this article reference is made
to a work of art, I have seen either the object itself or a photograph of it, and
have not depended on the engravings.

lo Albert A, Howard.

already been described, and tnbes thus mounted with a separate band
for each tube are shown in the plate to which reference has been
made. It is, of course, not impossible that such tubes were origi-
nally attached to the instruments found in Pompeii, and that the
oxidization of the metal has caused aH trace of them to disappear^
although this is yery improbable, especially if any considerable num-
ber of such tubes were in use on these instruments. The number of
tubes represented on a single instrument varies from one to seven,
the latter number occurring on the pipe of a muse in the British
Museum and inaccurately shown in MiOin, Galerie Mjrthologique,
Plate XX. Fig. 64.

The words of Pollux quoted above, p. 5, probably refer to this
very device : vAayiac dbotfas rw wo^^utn ras o&Nrs» for surely the
words irAoycoft 63oi might weU be used to describe these tubes.

There is no direct evidence to show that the bands on which diese
tubes were mounted could be turned round on the pipe so as to
cover the holes in it, but it is almost impossible to conceive that diey
could not be thus turned, for not only are the tubes occasionally
mounted on separate bands as in our plate, but the ic^Msra, or arrange-
ments for turning the bands, which have abready been described, are
often shown on works of art in connection with die tubes, and occa-
sionally, as in the relief from Zoega, Baumeister's Denkmaler, Fig. 594,
there is such an arrangement corresponding to each of the tubes.^
The efifect produced by these tubes when in use would be to lower
the pitch of the pipe, for the tone would be produced at die outer
end of the tube and not, as ordinarily, at the sur&ce of the pipe.

A somewhat similar arrangement is to be found on some of the
Boehm flutes manu£u:tured in England, and is described by E. Sw Ro(^-
stro in his book on the Flute, p. 183. The C* hole of the ordi-
nary flute serves both as a note-hole in producing C' and as a
vent-hole in producing several other notes. It is ordinarily placed
near its proper position as a vent-hole, whkrh is somewhat nearer
the embouchure than the proper position of the C^ hole, and the
diameter of the hole is diminished to allow the productioQ of C ^, as
a result of which the intensity of this note is weakened. To remedy
defect, in the flute described by Rockstro, the C ^ hole is bored

^ Tha refief I hsve not seen* bat in die foOowing waU-patBtiagt which I bftve
sees there is a wkpn kx each tnbe : Hefibi^ Nq&. 7^0^ 765, 1 t^a

The AvXo9 or Tibia, ii

with nearly the same diameter as the other holes, in its proper position
as a vent-hole, and into this hole a short tube is inserted. The note
C ^ is produced by the combined length of the flute to this point and
the tube, but when the hole is used as a vent-hole in producing the
harmonics, the connection with the outer air is of course at the
bottom of the tube, and the length of the tube does not influence
the sound.

The pipe held by the muse, Plate I. i, shows five of these tubes
and in addition several bands with holes but without tubes, from
which it must be inferred either that the tubes could be attached to
the instrument and removed from it at the will of the performer, or
that only certain holes of the instrument were provided with this
arrangement There is no direct evidence by which this point can
be settled, but in case the tubes were permanently attached to the
bands there would be no satisfactory reason for their existence, unless
the band was provided with a second hole, which could be brought
into use by turning the band until this second hole, and not the
tube, covered the hole in the instrument. There is nothing in any
of the reliefs to indicate the existence of a second hole, but in Proclus
(comm. in Alcibiad. p. 197, ed. Creuzer) the statement is made
CKooToy yap rpvmffui rwv avkaw rpui ^tfoyyous, <ik ^ouriv, rovXAxurrov
dxf^itfotv, u Sk Ktu ra irapaTpvinjfi4iTa dyoixBtirjf irXciOvs* A second
hole in the band, such as has been described, might well be called
a vapaTpvtnjfm and would double the number of notes that could be
produced on the instrument. With such a double arrangement and
with tubes of proper length, a difference of half a tone in pitch
could be produced by employing alternately the tube and the hole.
Unfortunately, the word waparpvmifui is found in no other place in
Greek literature, so that any attempt to explain its meaning is purely
conjectural, and while the explanation given above satisfies all the
requirements of the passage in Proclus, I offer it simply as a suggestion.

KOIAIAI.

As regards the KOiXxatf which are mentioned by Aristoxenus (Harm,
p. 60, ed. Marquard) as a part of the avXo9, probably the main bore
of the instrument is meant There is nothing against this view
except the use of the plural in this passage of Aristoxenus, while the
following passages from ancient writers favor such an explanation.

12 Albert A, Howard.

Porphyrius ad Ptol. p. 217, ed. Wallis : iraXiv & cav Xa^27^ ^vo avXovs,
rois \ikv fiiJK€cnv icrovs * rots Sk cvpvnpn rSiv kocXiuiv Sca^cpovras * KoBdr
rrtp Ixova-iv ol ^pvyioi Trpos rov? "EAAtviicoi^ * cvpiycrct? TnipawXrfaitaq to
evpvKOtXMW oivrepov wpoiifuvov iftOoyyov tov anvoKOtXiov * Otwpovficy yc
rot rovs ^pvytov9 orci^ovs rats kocAxcus ovra^ eirt iro\X<jp /Sopvrcpovs i^X^'^
7r/:>o^aXXoKTas rwv 'EAXi^vixuiv. Here the kocXou can hardly be any-
thing else than the bore of the instruments, and the same is true of
NicomachuSy p. 9, ed. Meibom : dvaTraXiv yi, rStv ip,irv€wrTUiv ai /mct-
{ovcs KoiXi<o<r€i^f Kol rh. fui^oya prfKift vinBpov xoi IkXvtw. Further-
more, the word cauemae is used in Latin with the same meaning,
Servius ad Aen. IX. 615 : tibiae aut Serranae dicuntur, quae sunt
pares et aequales habent cauemas; aut Phrygiae, quae et impares
sunt et inaequales habent cauernas,

KINDS OF pn>fes.

A long Hst of names of avXoi has been preserved in Pollux IV.
74 sqq. and in Athenaeus IV. 76 sqq., and from these lists it is
apparent that the number of different varieties of this instrument was
very great. In some cases, however, two or more names for the
same variety of instrument have been preserved, as can be clearly
shown, and this is probably true in many cases where proof is impos-
sible. In the following pages an attempt has been made to distin-
guish, so far as possible, between the different kinds of instruments
with reference to the manner in which the tone was produced on
them, and then to classify the instruments themselves with reference
to the pitch of their tones.

SINGLE PIPES.

The performer either made use of a single pipe, or he played upon
two pipes at the same time. Several names of single pipes have been
preserved, although perhaps not one of them is a name which was
intended to distinguish the entire class of single pipes from the class
of double pipes. The word /AovavXos seems to have as its first element
the adjective /lovos, and unquestionably Martial had this view when
he wrote XIV. 64 : —

Ebria nos madidis rumpit tibidna buccis :
Saepe duas pariter saepe monaulon habet.^^

The A^Xo9 or Tibia, 13

It is apparent, however, from a passage in Pliny N. H. VII. 204,
that not all single pipes belonged in the class of /AomvXoc His words
are : musicam Amphion (inuenit) fistulam et monaulum Pan Mercuri,
obliquam dbiam Midas in Phrygia, geminas tibias Marsyas in eadem
gente. The Hbia obliqua was undoubtedly a single instrument, but it
is not classed by Pliny among the /AomvXoc. This same distinction
between the fuWvXos and the irXayuivXos is made by Juba as quoted
by Athenaeus IV. 78, and it is possibly from Juba that Pliny's state-
ment is drawn. Pollux, IV. 75, distinctly states that the ^utw-vKvi is
of Egyptian origin : fiovavXos evfnffw. yuiv ianv AlyvmCtav^ and since
among the Egyptians the word ma-it^ was the general name for avXo9,
which was, however, used in a restricted sense to indicate an instru-
ment held like the clarinet or oboe, as distinguished from one held
as the flute is held, von Jan, Baumeister's Denkmaler, p. 562, is of
the opinion that the /xok- of /xovavXos is derived from the stem mam
or mom of this Egyptian word, and that it has no connection with
the Greek word /xoyof .

There is no etymological reason whatever for assuming that this
word is not composed of /tovos and avXds, exactly as it seems to be ;
and, furthermore, the fuWvXos was, as is implied in this derivation,
always a single pipe.

From the fact that the fioKivXo$ is thus contrasted with the irXayi-
avXo$, it would appear that the /AomvXof was a single pipe blown at
the end.

The single pipe is very rarely represented in Greek and Roman
works of art ; in nearly every statue in which it is represented the
instrument has been arbitrarily restored, and in very many instances
it is absolutely impossible to prove that the hand originally held a
pipe of any sort.' In reliefs and paintings where the instrument is
not a restoration two classes of aixW can be distinguished, one held

1 Cf. Athenaeus IV. 78. « Cf. Loret, p. 126.

' The single pipe blown at the end like a clarinet is represented on a sarcoph-
agus in the Louvre, FrOhner, No. 240, and in the following wall-paintings, H el-
big, Wandgem&lde Nos. 227, 767, and possibly it is not a restoration in a statue
of Euterpe in the Naples Museum (no No.). All of these representations I hare
seen, as also the following in which the instrument has been restored; Louvre,
Frohner, Nos. 287,382,395; Rome, Cap. Mus., Helbig's Fiihrer, Nos. 430(12),
521(10); Vatican, Helbig, No. 209(406); Munich, Glyptothek, Nos. 105, 106;
Berlin, Antiken Skulpturen, 258, 259, 602.

14 Albert A. Howard.

horizontally like the modem flute, ami the other held in front of the
perfonner like a clarinet or oboe. If the name fMmvXos has been
correctly interpreted in the preceding paragraph, it would natoraUy
apply only to the second class of instruments named above ; but as
instruments of this class seem to have the same general character-
istics as the double pipe, they will be treated in connection with the
double pipes.

Instruments of the first class were probably designated as a rule
by the name vAaTftiBivXos, Latin tibia obSqua^ although there were
undoubtedly separate names for the difierent varieties of instruments
of the class.

The TAayuivXiK is frequently mentioned by Greek and Roman
writers. According to PoUux ^ IV. 74, it was invented by the Liby-
ans, and was made of lotus wood ; according to Hiny, VII. 204^ it
was invented by Midas in Phrygia. A passage in Apuleius, Met. XL
9, shows that the instrument was held horizontally^ exacdy as the
modem flute is held : calamus odHquus ad amrem porrectus dexitram^
a £act which might have been inferred from the adjectives vAsyios
and abliquus which are used to describe it

Instruments of the vAayuivAos-t3rpe are extremely rare in works
of art, and, wherever they are represented, the entire instrument is,
with very few exceptions, a restoration, and therefore of no value in
determining the actual appearance of the ancient instrument.^ Woris
of art in which there has been no restoration show at least two, and
probably three, varieties of the instrument A wall-painting from
Pompeii' and a little gold ornament^ from a tomb in Russia repre-

1 Cf. Athenaeus IV. 78.

3 The v-XaryiauAof is represented on two reliefs in the Lonvre, Fr5hner» Nos.
88^ 575» and in the hands of a little tem>cotta figore. cf. Hemey, Les figurines,
etc., pL 136; also in a wall>painting, Helbig, Wandgemilde, No. t^cx It has
been partially restored on a sarcophagus in the Vatican, cC Mus. Pio Clem. V. 13^
and on a Herma of the British Mos., cf. Anc. Marb. II. 56, but in both of these
cases enough of the instrument is preserved to make the restoration certain. All
of these repre se ntations I have seen except the one in the B. M., and of that I
have seen a casL I give also a few of the restorations: Louvre, Frohner» Nos.
262, 265; Berlin, AnL SIculpturen, Nos. 260, 261; Rome, Vatican, Helbig's Ftlhrer,
No. 19 (56 a).

3 Helbig» Wandgcmaldeb No. 7€o» Zahn III. 31, Mus. Bor. XV. 18..

« Compte Rendu, 1867. p. 45t ^^^^ P^ ^-

The hv\6<i or Tibia, 15

sent musicians performing on an instrument held horizontally in
front of the mouth of the performer exactly as the modem flute is
held. The instrument in the painting is a straight tube with flaring
ends, that in the little ornament is a simple bar of gold, bent, at
present, at the end farthest from the mouth, although, as we are
informed in the account of it, the bar was probably originally straight.
Neither of these instruments shows holes either for the fingers or for
the mouth, but the fingers, in both cases, rest on the tube as though
they were intended to cover holes in it. In modem instraments of
this character the end nearest the mouth-piece is stopped with a
cork, and this was probably true of the ancient instrament, in
which case it would correspond exactly with the modem flute ; but
as no instrument of this character dating from ancient times has yet
been discovered, it is, of course, impossible to decide this point
except by inference.

A second form of this instrument is represented in a relief on a
puteaP in the Louvre, and by a terra-cotta figure,' also in the Louvre.
In both of these representations the performer is blowing, not across
a hole in the side of his instrument, but across the open end of the
tube. The tube is held horizontally, or nearly so, and the hands
occupy the same position as on the modem fiute, although in neither
of these instruments is there any trace of finger-holes. One might
be tempted to think that the artist, from carelessness or ignorance,
had represented inaccurately the manner in which the instrument
was held, were it not for the fact that on Egyptian monuments the
transverse pipe is always ' represented as being held in this way. The
vXaytavXos, as we are informed by Pollux. and Athenaeus (cf. page
14), was invented by the Libyans, and if this statement is true,
there is sufficient reason for believing that these two works of art
represent accurately the way in which the pipe was sometimes held.
The Arabs of to-day perform in the manner described* on a simple

^ Frohner, No. 88; cf. Garac, pi ijo, 139, n. 141.

* Heuzey, Les figurines, etc., p. 136.

' Loret, p. 209. La fliite oblique ^tait on simple tube ouvert aux deux bouts,
dans lequel on soufflait en biais. Jamais, en effet, on n'a trouve de fldte munie
d'une ouverture lat^rale semblable k ceUe par laquelle on souffle dans nos fldtes
modemes; jamais non plus, dans les bas-reliefs, la fldte oblique n'est representee
depassant leg^rement la bouche de Tinstnunentiste.

^ Loret, p. 219.

i6 Alb^n A. Howard.

pip^ of r^^ provided with fiD^r>holes and open at both ends^ and
by careful experimeot the reader cao convince himself that it is pos-
sible in thi& way to produce not only the tundamentai tones of a pipe,
but also the hariBonicsw^ Instruments belonging to either of these
varieties would have the acoustic properties oi the open pipe^ and
with six tii^^-holei^ would give the complete diatonic scale thxongh
at least two octaves^ : they might properly be called tlutes«

What has^ been described as probably a third form oi the vAoTt-
owW difters from the other two vaheties in that it has a month-piece
inserted, not into the end« but into a hole in the side of the instm-
meat oear the upper end, which is^ dosed. Instntments of this type
are represented on a sarcophagus^ in the Vatican and by a Hexma^
in the British Museum, and although in the second of these represen-
tations the instrument, with the exception of the mouth^piece^ is a
restoratKH^ tiiere can be ik> doubt that the instrumeftt has beea
pco|H:rW restored. In both of these cases^ the mouth^piece consists
apparently of a short tube, and was perhaps a reed, like the reed of
an oboe or of a clarinet, for in the Herma» winch I have sees ooicf. in
a castp the mouth-piece seems^ to have been held betweesr the lips^
There 5eemj> to be authority for such an iDd^rence in Axistotfce^de

6i Aft^iur^y where y^Mxnftu. Tr^yuik couki very m»]l reler to a. mootb^
piece extending obliquely fxom. the instnuneni i& the manner de-
scribed above.

There are, farther, in the Bcitish Miseuok^ two instruments wihich
were pos&ibiy intended to. be used as a pair> akhough in thek present
tona it would be impossible to play on them both at once. One o£
these pipes is closed at tiie upper end«. and. although the corsesptuidr
iag end of the second pipe is now bioken» it was undoub^HUy origi?^
nally closed as in the liist. uuuxumenk. Near this closed, end there is
on eaidi pipe the raised bust oi. a Maenad, above the forehead of

3Utt|bi>w«c«. i^itfirei/ nn«n'r.f^']ai. It was^ ooljf AiijOL xoAay I 'ai l igcfc ihfti. I su£CQ«ii«ii
ia pwtioiaog cieui toam^ Tlic iiytfrnownr ua<!ki wa&.a jaun oi tmatiiQo with oa

- MtKw Pio Oeiu* ^'. 13. * i^o»>tctl by PofpbjFcuft^aii Ptoi., p. .24^^

The Av\J9 or Tibia. 17

which a hole is bored obliquely into the main tube, the slant of the
hole being toward the lower open end of the instrument. One of
these pipes has six finger-holes, the other five, and there are lines
around the pipes between the finger-holes which indicate the points
of meeting of bands such as were found on the Pompeian instruments.
The hole above the head of the Maenad is evidently intended for
the reception of the mouth-piece ; the fact that it is bored obliquely
clearly shows that it was not itself the mouth-piece, but that the
latter was inserted into it. The arrangement of bands and finger-
holes differs slightly in these two instruments, so that the scale of
both cannot have been exactly the same ; but the resemblance between
them in size and in appearance seems to indicate that, in spite of the
slight difference in scale, they were used as a pair. The presence of
bands on instruments with so few finger-holes is a further indication
that they formed a pair, for on a single pipe the fingers of both hands
could be used to close the holes, and on such an instrument, with
but six holes, the bands would be an encumbrance rather than a help,
unless, as is hardly conceivable, they were immovable and were
intended merely as a kind of ornament.

Without some additional form of mouth-piece the two instruments
cannot have formed a pair, since it would be impossible to perform
on both of them at once in their present state. It would, of course,
be possible to insert into each of these pipes such an S-shaped tube
as is used on the bassoon of the modem orchestra, and thus to per-
form on both of the instruments at the same time, but there is nothing
in the literature or in the works of art to indicate that mouth-pieces
of this description were ever used with pairs of pipes. On the other
hand, the two works of art which have been mentioned do show a
form of TrAaycavXos with a mouth-piece inserted into the side of the
tube; and, in the absence of any evidence showing that a similar
form of mouth-piece was ever used on the double pipe, I am inclined
to think that these instruments from the British Museum were used
as single pipes, and belong to the type of TrXaycavXos shown in these
two works of art.

By fiar the greater number of representations of the av\o$ show the
performer blowing directly into the end of the pipe or pipes, which
are held in front of him, pointing downward like the clarinet or oboe ;
and, with very few exceptions, the works of art show pairs of pipes
and not the single instrument.

1 8 Albert A. Howard,

The single pipe is, however, represented in works of art and is
mentioned in the literature, so that it would be impossible to deny
the existence of such an instrument. In works of art it is repre-
sented as a straight tube, and in the Pompeian wall-painting, Helbig,
767, it is provided with a bell like the bell of the modem clarinet or
oboe. The mouth-piece, of which a fuller description will be given
later, has occasionally the appearance of an oboe-reed.^ The single
pipe was undoubtedly in most cases a reed-instrument, and was then
constructed exactly like one of the pipes of a pair, but there is occa-
sional mention of a single pipe which may possibly not have been a
reed-instrument.

The reasons for believing that such an instrument existed are as
follows : Athenaeus, IV. 82, in speaking of the inventors of the vari-
ous musical instruments, says : Mi/rpoScapos 8* 6 Xtos h Tp<oiKoU <rv-
ptyya, fi€y fftrfciv cvpciv Maponxiv fcat avXov iv KcAxuvaiS, rwv irportpov

kvi KoXdflff (TVplioVTiOV' Evi^OpiW S* 6 CTTOTTOIOS ^V r<p TTCpi fJLtXoTTOUWV

Trfv pkv fjLOvoKaXapjov avpiyya *Ep/A^v cvpciv, riva^ S* iaroptiv litvOrjv koI
"PwvoKiyv Tov^ Mi;Sovs, rrp^ Sk iroXvKaXapLOV SciAi^vor, Map<rvav Bk r^v
KrjpoSerov* A distinction is here made between the <Tvpiy$ povoKoXa-
/xos and the avpiyi 7roXvKaAxi/Ao$, and the fact that the name <rvpiy$ is
used for both clearly points to some resemblance between them.
The <Tvpiy$ 7roXv#caXa/A05 is the ordinary pan*s-pipe, in using which
the performer simply blew across the open ends of the tubes, which
were held vertically in front of the mouth. The tubes of the pan*s-
pipe were ordinarily simple joints of reed, and it is clear from a
passage in Porphyrius (ed. Wallis, p. 237) that they were not always
closed at the bottom,' and that when they were not closed, two dif-
ferent tones could be produced on each tube, one when the tube was
open at both ends, and the second when the lower end was stopped
by the finger of the performer.. It is but a step from such an instru-
ment as this to a single tube open at both ends and provided with
finger-holes, and the new instrument could appropriately be called

^ Mus. Borb. XVI. 3, cf. Baomeister, fig. 595.

2 Perhaps Plin. N. H. VII. 204 was copied from this same source, fistulam et
monaulum Pan Mercuri (inaenit), and we should read fistulam monaulum = ffvpiy^
fjMvoxdXafiot.

' In the Museum fUr V51kerkunde in Berlin, I remember having seen, in the
collection from Bolivia, a panVpipe in which the tubes were open at both ends.

Tlu A^Xo9 or Tibia. 19

a crvpcyj fuvoKaXofuoq, The <rvpiy$, mentioned near the begmning of
Aristotle's problem XIX. 23 (quoted on p. 3), was probably a avpiy^
fiovoKoXofUK, How many finger-holes this instrument had is not
stated ; it certainly had one in the middle of the tube, and the note
produced when this hole was open was an octave above the note pro-
duced when it was closed.^ That such an instrument belongs among
the avXoi has been seen in the case of the irXaytavXo^, and furthermore
Pollux, IV. 69, twice calls the tubes of the pan's-pipe avXoc. It is
possible on such an instilment, provided with six finger-holes, to
produce a continuous scale through two octaves; the instrument
having the acoustic properties of the open pipe. The scholion to
Pindar, Pyth. XII. i, favors this explanation, and the wonderful ac-
compHshment of Midas of Agrigentum can be easily explained ; for,
after the reeds of the pipes were broken, he blew on the pipes alone
T/Kwnp avpiyyoi, simply holding them vertically before his mouth and
blowing across the open ends of them. The difficulty would consist
simply in the necessity of blowing on both of them at once.*

^ This instrument cannot be the same as the <rvpiy^ mentioned later in this
same passage of Aristotle, and in which the tubes were either stopped at the end
or partially filled with wax, for if such an instrument had a hole in the middle of
one of its tubes, the note when the hole was open would be not one but two oc-
taves above the note produced with the hole closed. With the hole closed the
instrument is a stopped pipe, but with the hole open it is an open pipe, one half
as long as the original stopped pipe, and would therefore give as its note the
double octave.

The use of the same word ffOpiy^ with two different meanings within so short
a space can be easily explained. In illustrating the ratio of the note to its octave
(2: i), the writer began with the string of a musical instrument which, if plucked
so as to resound through half its length, gives the octave of the note produced by
the whole string. Applying this same principle to the other instruments he shows
that on the avpiy^ fioyoKdXafiot the octave is produced by boring a hole in the
middle of the tube, and in the same way also in the ai\6i; but in the aOpiy^ iroXv-
irdXajuos, in which the tubes are stopped pipes without finger-holes, the octave is
produced by filling the tube half full of wax.

' Gevaert, Hist, de la Musique, II. p. 275, explains the ffvpiy^ fWPOKd\afwi as
an instrument with a flageolet mouth-piece like that of the ordinary whistle, and
says that these instruments occur no less frequently in works of art than does the
pan's-pipe. I cannot, however, admit that this proves the existence of a flageolet
in ancient times; for after a careful examination of all such instruments which I
could find in Naples, Rome, and Paris, I am thoroughly convinced that in every
case the instrument is a restoration and of no value in determining this question.

20 AIb€rt A, Ha^ward.

The #«^ pimmim%mp%\ was aBK> called *lry^ (c£ ^m. JTd^. xj^.
^wji, Bekker^ Antedate GraecOj L 265 )^ and if k l e ce i fed this name
fan die fxX diat its note resemUed die ooCc of die bod, it most
have been a pipe of ingh pildi.

Egy ptiaa refie^ and paintingv even die oldest of them, most
freqaentiy show perfiannanrrs oq the doable pipe. Assrriaii ^ rcfie^
with no excepdoo, show the doable pipe, azid with tcit few cxcep-
tioos the same is true of Gredc and Roman woris of art. These
Greek and Roman representatioos arc tcij nomeroos. EspedaQj is
this true of irase-paincings, of which serexai handred coold easily be
casalogoed; and Tct in nearij crcij instanrr the doable pipe is
shown. In the liteiatxzre the mention of the doable pipe is eqaally
predominant ; the words obXok and tibiae are regnlarhr used in the
plxxial. The natoral inference from this evidence is that the musi-
cian osaaDj peribrmed on a pair of pipes.

Gevaert, Histoire de la mosiqae, VoL II., p. 2S5, asserts that the
great artists, Olympos, Sacadas, and Prooomos, played onlr on the
single pipe and that this was the instrument used by the rirtoosi at
the Delphic contests. The evidence tending directly to disprove
both of these statements is, however, of the most amvindng charac-
ter. In the scholia to Pindar, Pyth. XII. i, we are told that Midas
of Agrigentam, who won the prize at two Delphic contests, once had
the misfortane, in the midst of a contest, to break the moath-pieces
of his instrnments, bat that he continued his performance ^unrois rots
naK&fuM% r^tf crvyxyyof and won the contesL It is to be nodced
that this contest took place at Delphi, and that the pipes are men-
tioned in the plural In the 48th Olympic* contest Sacadas the
Argive won i-wi rA avXA, where again the plural is used. Pausanias

This it distinctly the case in Paris, where the restorations are consdentioiisly
stated tn the catalogues, and in Naples and Rome every such instrument is either
restored or bf oken. It is interesting to note that in Wilkinson's Jfammrrs and
Customs of the Egyptians^ Vol. II. pL 183, there is represented from a relief in
Thebes, a musician performing this very feat attributed to Midasw I cannot, how-
ever, vouch for the trustworthiness of this representation, not having seen the
original.

1 Rawlinson, Ancient Monarchies, VoL L, p. 533.

' Pausanias, X. 7. 4.

The Aii><j6<: or Tibia. 21

also describes a statue of Sacadas^ holding pipes (again the plural)
which were fully as long as the body of the musician. As for Pro-
nomusy the passage from Pausanias which has already been quoted
(p. 6) shows conclusively that he, too, used the double pipe, and
the fact is further attested by Athenaeus, XIV. 31, 8co7rcp tftrav i&oi
Koff kKoxmiv apfJMviav avXoe koX ^iccurrotf avXtfrStv \nnjp)(ov avAol iKwrrg
apfiovui, rrpocifkapoi cv roi$ dywri. UpovofjuK ^ 6 ©i;)3ai6$ irfHoro^ rfiXrf
orcv &iro rStv avrwv [av\S)v] iraxroi apfiovui^. The following passage
from Lucian imphes that the musicians Timotheus ^d Ismenias also
made use of the double pipe (adv. Ind. 5) cT ns avAciv fi^ IvurrdiLt-
vos KTqaajLTo roi)^ Ti/juoOiov avXovs y rov^ *Icrfirjviov, oS^ irrrk raXdvrtav
6 *l<rfjLi^tai iv KopCvOt^ irrpCaro, dp* &v &a rovTO kcu avAciv BwaiTo;
Even the vase-paintings ' in which the musician is represented wear-
ing a wreath, evidently as a token of victory, show him performing
on a pair of pipes and not on the single pipe. There is no lack of
further evidence on this point, but enough has already been said to
show conclusively that the double pipe was used at the various musi-
cal contests and by the chief musicians of antiquity.

REEDS.

In the overwhelming majority of cases, the avAos must have been
a reed-instrument. Aeschines' says distinctly that if the reeds
(yAwTToi) of the pipes are taken away, what remains is useless.
Aristotle* describes the eggs of the skate and of the dog-fish by
saying that they resemble the yAtorrai of pipes, mentioning, of
course, something perfectly familiar to his readers. Theophrastus,*
in describing the icoAa/ios avXr/TiKoq, assumes that the yXSyrrcu are a
regular feature of the pipe. The word ^fiyvAi/ficKo? • in its colloquial
meaning, "worthless," "played out," acquires this meaning from the
fact that the worn-out reeds of pipes were thus designated.

Two types of reed-instruments are known to modem times, — the
clarinet-type with a tube of cyUndrical bore, and the oboe-type
with a tube of conical bore. The mouth-piece of the clarinet has
a rectangular opening in the side, over which a thin wedge of reed is

1 Pausanias, IX. 30. 2. ' Aesch. contra Ctes. 229, Bekker.

* Mon. deU' Inst. V. 10. * Aristot. Hist. an. p. 565 a 22.

* Theophrastus, Hist. Plant. IV. 11.

* Aristophanes, Achar. 681, schol; cf. Etym. Mag., s.v. i^vXri/Uvos,

22 Albert A. Howard.

bound in such a manner as to leave the thinner end £ree^ so that
when caused to vibrate by the breath of the performer it alternately
closes and opens the aperture in the mouth-piece.

The mouth-piece of the oboe is composed of two thin strips of
reed which are bound firmly around a small conical tube, so that the
ends of the strips project beyond the smaller end of the tube and
flatten against each other, thus forming the apex of the cone. A
reed of the oboe-type can be easily made of straw by splitting the
straw for a sho]:t distance, and pressing the two halves together at
the end.

There is a marked difference between instruments of these two
t)rpes, those of the oboe-type having the acoustic properties of the
open pipe, while those of the clarinet-type have the acoustic proper-
ties of the stopped pipe. This difference is caused by the bore of
the instrument, and not by the form of mouth-piece employed ; for
the instrument invented by Adolph Sax, and known as the Saxo-
phone, which is of conical bore and on which the tone is produced
by means of a clarinet mouth-piece, has the acoustic properties of
the open pipe, and the experiments with pipes of cylindrical bore
described later in this article show that with either form of mouth-
piece the cylindrical tube has the acoustic properties of the stopped
pipe.^ All Greek and Roman instruments, and all fragments of such
instrutnents which have as yet been found, are of cylindrical bore,
and have, therefore, assuming that they are reed-instruments, the
acoustic properties of stopped pipes. Furthermore the statement of
Aristotle as to the position of the finger-holes of the avXd« (cf.
p. 3 sq.) applies only to pipes of cylindrical bore.

In works of art the pipes are occasionally so represented that
it seems absolutely necessary to assume that the bore must have
been conical, and this is especially apparent in a relief ot which
a photographic reproduction is shown in the Bulletino Communale,
1880, PI. VII.-VIII. Until, however, some instrument or fragment
of an instrument is discovered which settles this question in the
affirmative, it is only possible to say that the ancients may have had
instruments of conical bore. Even the pipes found in Egypt must
be regarded as cylindrical tubes, for they have only the natural taper

^ Friedrich Zamminer, Die Mosik und die musikalischen Instrumente, Giessen,
1855, p. 505, confirms this statement.

The AvXJ? or Tibia, 23

of the material (reed) out of which they are constructed, and this is
very insignificant.

It does not, however, follow that the mouth-piece was alwa3rs a
clarinet mouth-piece, as has been assumed in all of the experiments
hitherto conducted with pipes made in imitation of those found at
Pompeii, and as is implied by von Jan ^ in his description of the avAo« ;
in fact, there is fully as much evidence in favor of the oboe mouth-
piece as there is in favor of the clarinet mouth-piece.

The split straw is the simplest possible form of oboe mouth-piece,
and was certainly used as a reed by the ancient Egyptians. Loret
(p. 207) says that in the case in which were found all of the Egyptian
pipes now in the museum at Leyden was found also a number of
short pieces of straw, and on p. 209 that two pipes, one in the Brit-
ish Museum and one in the museum at Turin, still have such a piece
of straw inserted into the end as a mouth-piece.* The straw is in
each case inserted into the smaller end of the pipe, although in
neither case is the difference in diameter of the two ends more than
one millimeter.

Vergil * and Tibullus * several times speak of the performer on the
auena^ or oat-straw, and, in view of this evidence from Egypt, it
must not be rashly assumed that- they are speaking figuratively. A
reed of such simple construction may very well have been used by
the shepherds on their pipes.

Theophrastus, Hist. Plant. IV. 11, describes at length the plant
KoXafioi, from which the pipe and its mouth-piece were made, but
unfortunately he does not tell exactly how the mouth-piece was
made. After having been properly seasoned, the KoXafioi was cut
into pieces corresponding to the joints of the plant, and on each
piece the joint nearest the top was left. These pieces were not less
than two palms in length, and the pieces from the middle of the
stalk were considered the best for the mouth-pieces, {cvyiy. The
mouth-pieces for both of the pipes of a pair were made from the
same joint, for if not so made the reeds, we are told, would not
sound in harmony. These joints were next cut in two, and the por-

^ Baumeister, Denkm., p. 553; cf. Gevaert, II. p. 282.

^ Cf. Wilkinson, Ancient Egyptians, Vol. II., p. 308.

' Vergil, Eel. I. 2, X. 51, and the spurious lines at the beginning of the Aeneid.

* Tibullus, II. I. 53, III. 4. 71.

24 Albert A, Howard,

tioQ toward the root was used as the moath-piece for die left-hasd
pipe, the other for the right-hazKi pipe; the oroyia (doubdess the
end taken into the mouth) was in each moath>piece at the point
where the joint had been cut apart, and here, again, if this course
was not foOowcd, the reeds woold not sound in harmony. Farther
than this Tbeophrastos does not go in describing hov die mouth-piece
was made, and whatever more is to be kamed about these mouth-
pieces most be learned by experiment. One of the pieces, as left
br the descripticm. is a simple tube of cane, open at both ends and
about a pahn in length : the other piece b of the same length, but
stopped at one end by the joint of the cane. This stopped end is,
h o we v e r^ to be inserted into the instrum«it, and cofxsequenthr a hole
most be bored through die joint to alk>v die air to pass through it.
It is then possible either to s^t the tube and, after scraping the
eais to be, held in the mouth until they are very thin, to bind the
two [aeccs together so as to make a reed like that of the modem
bassoon, which b an instrument of the oboe-cype ; or by stopping
the end to be takei into the mouth, making a rectangular opening
in the s&de of the tube near thb end, and binding over it a reed, as
in the chrinet mouth-piece, to produce an instrument of die chrxnel>
type. The first of these methods ts the simpler, and it b exacdy in
this way that the reed of the bassoon is made. Sbch a mouth-piece
woulid, furthermore, be the most natural dereiopoKnt from the pcimi-
tcve mouth-piece of straw. Too much stress must not be bid upon
works of art. but in some of these the mouth-piece is so represented
that one ts almost cocnpelled to believe that the artbt had in mmd
this form of reed. This is especiaEy tn» of a vase painting refwo-
doced in the Moo. dell* Inst. XI. 27, aztd of the single pipe in d^e
Fbmpeian waH-painting^ Zahn, III. 45, to whkh r efe rence h^s abrcady
been m^ide. There b^ moreover, nothing in die passage brom Tbeo-
phrastos incoQsstent with thb view, and ooe or two expressoos used
by dtb author in stating the best time for harvesting the miXtfmK are
best explained by assuming that the reeds were of djoe oboe-type.
These expressions are found in the folfowing pi^ssage^ which for the
sake of cleamess b given in full : Hist. Ffant. IV. 1 1. 4. r^ SI f'oi^niir

fumr^xk -jipp^fTVitMm mm r]MMBTavA}f(r«ws iw^^^ vtiAA^ ^v/K/mur & r»

The AuXo? or Tibia, 25

OTOfUL rtav yAcarrwv, o irpos r^v itaropCav c7vcu y^prfciiuov, iiro. 8c cis r^v
irAaoiv fiertPtfoav koL 17 rofi^ fiertKin^Orf ' rifJLVOwn yhp Svf vvv rw
Sicippo^pui>vo9 Kflu 'Eicarofi)3ai(tfvo9 cucnrcp Trpo rpowSiv fUKpov ^ xnro
rpoiw, yiv€a$(U 8c (^ocri rpCtvav rt xprjaifiov kcI Karavkrjaeto^ Ppa^tios
hwBoJL Ktu KarafnrdxrfULTa ras yAwrras tcr^civ * rovro Sk dvayKaiov reus
fura irXda-fULTOi avAovcri. Pliny, N. H. XVI. 171, who attempts to
translate this passage, translates <rvfjLfivav 8^ to crrofm, etc., compri-
mentibus se iingulis, and Karaanrda-fULra ras yXiarrai ifrxtiv, apertio-
ribus earum lingulis ad sonos flectendos. It is difficult to see what
these words can mean if applied to the mouth-piece of a clarinet, but
if applied to that of the oboe they might mean that when the tube
was split to make the reed, if the material had been cut at the proper
season of the year, the two halves of the reed, when flattened against
each other by pressure, retained this shape and did not have a ten-
dency to curl up into their original form, — a tendency which is ex-
pressed in Pliny by the words apertioribus earum lingulis ad sonos
fiectendoSf since if the reeds curled they would separate in the middle.
Even the name {cvyo« as applied to the mouth-piece seems to imply
reeds of the oboe-type. This word naturally means a pair of like
things and would here mean, when used in the singular, a pair of sim-
ilar reeds united into a mouth-piece. The plural fcvyiy would then
mean the two mouth-pieces of a pair of pipes. It is not easy to see
what {cvyo$ could mean if applied to a clarinet mouth-piece. It is
further to be noted that the clarinet has but one reed, while here
Theophrastus uses the plural, to crropA twv yAcuTToiv.

Since the discovery of four ancient instruments at Pompeii in
1867, experiments with tubes made in imitation of them have been
conducted by Kraus in Florence and by Gevaert in Brussels, with a
view to determining the scales of these instruments. These experi-
ments seem, however, to have been made in every instance on the
assumption that with instruments of cylindrical bore it was necessary
to make use of the clarinet mouth-piece.* The exact form of mouth-
piece used in the experiments at Brussels and Florence on the fac-
similes of Pompeian pipes, is described at length by von Jan in

^ Banmdster, Denkm., p. 553; Gevaert, Hift de la musique, II. 280. The
clarinet type of mouth-piece is used to-day by the Arabs (cf. Baumeister, p. 554;
Loret, p. 234 sq.), who perform on an instrument corresponding in many respects
to the double-pipe of antiquity, and it is possibly a survival of a very ancient usage.

26 Albert A. Howard.

Baumeister's Denkmaler, p. 554. This form of mouth-piece certainly
does give a fuller and clearer tone than the reed of the oboe-t3rpe ;
but, as has already been stated, the form of mouth-piece has no effect
on the harmonics of the pipe, and with the Pompeian instruments
the scale is exactly the same, no matter which form of reed is used.
The works of art give very insufficient material from which to decide
this question, and it is, of course, possible that both forms of mouth-
piece were used.

The reeds were taken directly into the mouth, and were not
enclosed in an air-chamber, as is shown by the following passage
from Aristotle ^ (de audib. p. 802 b, 18) : Set y^ ical roiv avAcuv cZmu
TQH yXorrraf irvKvhs fcou Actas kcu OfiaXd^, oma^ &v icoi to irv^fui ScantH
ptmjTOJL &* avrSiv Xidov fcoi ofjuaXov fcoi yai &c<nra<rficvoK. &o icoi rh.
fiefiptyfjiiva rwv {cvywv, kcu ra ircircDicara ro crtaXxw, cv^vorcpa yiyvt"
rat * ra Sc iniphi KaKOtfxova ' 6 yap arjp &a vypov koX Aciov <f>€p€Tiu fiaXa-
Kos fcat 6/AaAo9. The following passages show further that the entire
mouth-piece was not taken into the mouth so that the reed swung
perfectly free in the mouth of the performer, but that the reed or
reeds were held between the teeth or lips of the performer, so that
he could use them to govern the tones produced on his instrument.
Aristotle de audib. p. 804 a (cf. Porphyrins, p. 252) icat yap &v
irUay ns ra {cvyiy, /ioAAov 6ivT€pa ij if^wvrf yiyverai icoi \enroT€pa ;
Porphyrins, p. 249, iv Sk rats (rvyfcporc/xus yAo>(raat$, 1) ^cdv^ yiyvcroi
CKXr/poripa fcal Xa[JLTrpoT€pa, &v vUayi ri9 avra« fioAAov rocs x^^^^ l ^^^
Festus (ed. Mtiller, p. 116) s.v. lingula — alias insertae, id est intra
denies coercitae ut in tibiis.

This control of the reeds exercised by the lips is one of the princi-
pal means for correcting the pitch of the pipe to which reference
has already been made, and these passages show clearly that this
device was well known to the ancient musicians.

The entire mouth-piece, including the reeds, seems, from the vari-
ous passages quoted, to have been called the {cvyo?; but as the
important part of it was the reeds, the name yAcorra was often used to
designate the entire mouth-piece, as is shown by the use of this word
and by the fact that the mouth-piece, which fitted into a socket in
the end of the tube and which could be detached from the instru-
ment, when not in use was kept in a little case called the yAoirroico-

* Quoted also by Porphyrins, p. 250.

The kvKo^ or Tibia. 2J

yudo¥} Such a case for mouth-pieces is shown in a number of
vase-paintingSy' and appears to have been simply a small bag.

Where two pipes were employed at the same time, each had a sep-
arate mouth-piece, as is distinctly shown in every* work of art which
has been preserved to our time. This statement is confirmed by the
Egyptian and Assyrian works of art, for in them the instruments are
never represented as uniting in a single mouth-piece.

In addition to this, experiments which I have made with pairs of
pipes of various sizes united by a single mouth-piece, into which
could be fitted reeds of both the types described above, made it very
apparent that the pipes were not intended for use in this manner.
With either form of reed the pipes cannot be made to sound inde-
pendently, and they admit of the production of but one tone at a time.
If the open end of either pipe is closed, a note an octave lower than
before is produced, which would seem to indicate that such an instru-
ment was acoustically an open pipe. The harmonics of the instru-
ment, however, when both pipes are open, are those of the stopped
pipe. That the pipes do influence each other is shown by the fact
that, although with pipes of equal length the resultant note is the note
which either pipe alone would give, with unequal pipes the resultant
is not the fundamental note of either pipe but the average of the notes
of both. I further observed that the angle which the two pipes made
with each other had absolutely no influence on the pitch of the tone.

These experiments were made both with pipes of the same length
and with pipes of different lengths, and the result was always the
same. With pipes of unequal length the result is by no means sur-
prising; for, theoretically, it seems impossible that a single reed
should be able so to vibrate as to produce different tones on two
pipes at the same time. In the light of these experiments, there-

1 Pollux, II. io8, VII. 153, X. 153-4.

3 Baomeister, Denkm., p. 554, Fig. 591.

* In Bartholinus, de tibiis vetenim, pp. 5i> 52, there are two engrayings which
represent pairs of pipes, each pair uniting in a single mouth-piece. The first is
a reproduction from J. J. Boissard, Antiquitates Romanae, in which a sacrifice to
Priapus is represented; the second is from J. P. Bellori, Admiranda Rom. antiq.
▼estigia, pi. 47, in which is shown a female figure playing on a pair of curved
pipes. I have sought diligently for the original reliefs from which these engrav-
ings were made, but have never been able to find them, nor have I ever seen in
an ancient work of art two curved pipes in the hands of a musician.

28 Albert A, Howard.

fore, it is safe to say that the two pipes were never made to sound
with the same reed, and that, if they were ever united by a single
mouth-piece, the mouth-piece must be regarded simply as an air-
chamber, within which each pipe had its own reed.

OAMOS, Y«OAlfION.

There are two portions of the instrument, near the mouth-piece,
for which names have been preserved, — the oX/1109^ and the v^oX/uok.'
Two explanations of these parts of the instrument, neither of them
entirely satisfactory, have been offered, — one by Gevaert, Vol. II.,
p. 285, in which the 6A/A09 is described as the mouth-piece, the v^oA-
IJLiov as the upper part of the instrument, into which the oAfu>9 was
inserted ; the other by von Jan, in the Allgemeine musikalische Zeitung^
1 88 1, No. 31, in which the oX/ju>9 is described as the beak or mouth-
piece of the clarinet-type, the v^dX/xtov as the reed itself, which
rested on the under lip of the performer, and was therefore under
the oAfu>s. But there is absolutely no evidence that the oAfu>9 was
ever taken into the mouth of the performer, as would be required by
either of these explanations, and a much simpler explanation than
either of them is possible. The word oAfu>9 signifies a mortar, or a
peculiar shaped cup, and the word v^dX/xtov signifies originally a
support or stand for the oKyjoi ; cf. Aristophanes, in Pollux, X. 114;
Hesychius, s.v. The mortar and mortar-stand, as shown in a vase-
painting,^ have a very striking resemblance to the upper end of the
instruments from Pompeii (PI. II.), and it is not improbable that
these parts of the pipes received their names on account of this
resemblance. According to this view, the oXfu>s is the pear-jshaped
bulb, and the v^dA/xtov the flaring tube inserted into it.^ The bulb,
or ^Afu>9, is very frequently shown in works of art,* and it is interest-

1 Pollux, IV. 70.

^ Pollux, IV. 70; Hesycb. s.v. yJkpo% ti tou affhov Tp6t rf a-rSftaTi^ rj al yXwrrlSci.
Kal inrbBtiiA ri. For ¥1 Aug. Wagener reads {. Ptol. Harm. I. 3, p. 7, ed. Wallis.

' Jahn, Bericht. der Sachs. Acad., 1867, taf. i. 4; cf. Schreiber, Bilderatlas, taf.
LXVII., Fig. 3.

* Von Jan, Baumeister's Denk., p. 556, suggests the possibility of some such
explanation as is here offered.

^ Both SKiMt and ^^Xfiu^p are shown in a Pompeian wall-painting (Helbig,
Wandgem&lde, No. 114), and in a mosaic in the Museo Kircheriano ; for the
dXftot, cf. Mon. dell' Inst VI. 18; H. d'Escamps Marbres du Musee Campana,
pL 25; Baumeister, fig. 598; etc.

The AvXo9 or Tibia. 29

ing to note that on the modern clarinet this part of the instrument
has, in all countries, received its name from its resemblance to some
familiar object, being called in America and in England the ' pear/
in Germany * die Birne/ in France * le baril/ etc.

The part which has been called the v<^oAfuov is not always shown
as a flaring tube, but is often represented as a cylindrical tube. It
is evident from the statement of Hesychius, fi^Dos n rov avAov irpos
Tip crrofmn, that the v<^oX/uov was near the mouth-piece of the instru-
ment ; but it is equally clear from the catalogue of the various parts
of the avkoq as given by Pollux, IV. 70, that it cannot be identical
with the reed, for both of these parts are mentioned, and in such a
way as to imply that they were not identical.

According to this view, Gevaert is correct in assuming that the
vit>6\fjuav was the flaring tube into which the mouth-piece was in-
serted, but is incorrect in assuming that the o\/ju>9 was the mouth-
piece.

♦OPBEIA.

In performing on the double-pipe a great deal of pressure was
exerted by the breath on the cheeks and lips, and to relieve this
pressure the performer made use of a bandage, passing over the
mouth and cheeks and provided with holes through which the
mouth-pieces of the instruments could be passed. Such a bandage
is frequently shown in works of art, and was called by the Romans
capisfrum, by the Greeks ^op/9cta, wtpurrofjuov? It is only with reed-
instruments that the capistrum can have been used for the purposes
here stated ; with any other type of instrument the pressure of the
breath would be very slight and the bandage would be an inconven-
ience rather than a help.

It has occurred to me that possibly this bandage was intended to
serve still another purpose ; that of holding the instruments to the
mouth of the performer so that the hands might be left free to move
up and down on the instruments, as would be necessary in turning the
bands, and in opening and closing the finger-holes. Inasmuch as but
one hand could be employed for each pipe, some such arrangement
was highly desirable if not absolutely necessary and especially if the
number of finger-holes was very great

^ SchoL At. Vetp. 580; Hesych., s.v.; Suidas, s.v.

30 Albert A, Howard.

A capistrum could easily be constructed in such a way that the
mouth-pieces would fit tightly into it and support the instraments,
leaving the hands absolutely free to move.

The instruments themselves, when not in use, were kept in a case
called avfii^} Such cases are frequently shown in works of art,
and regularly have a separate compartment for each tube. No
Greek or Roman cases of this sort have ever been found, but two
instrument-cases have been found in Egypt and are preserved, one
in the Louvre at Paris, the other in Leyden (Loret, pp. 199 and 201).

HARMONICS.

There are very convincing proofs that the acoustic properties of
the avX.6q were well known to the Greeks and Romans, and that they
made use of the harmonic tones of their instruments in their music.
In all vase-paintings and in many of their reliefs, the instruments are
represented as tubes of small diameter in comparison with their
length. The pipes represented in Egyptian works of art show this
same peculiarity, and of the existing instruments described by Loret
(p. 197 fg.) only one has a diameter so great as 1.8 cm. while with
few exceptions the diameter is less than 0.6 cm. The four pipes
found at Pompeii have an internal diameter of 0.95 cm. and the
diameter of the Greek instruments in the British Museum is corre-
spondingly small.

It seems clear, therefore, that these instruments were ordinarily of
small diameter as compared with their length. Tubes of such dimen-
sions facilitate the production of the harmonic series of tones (cf.
Zamminer, die Musik und die mus. Instr., p. 218), and the ancients
very probably had this object in view in constructing their instru-
ments with these dimensions.

Aristoxenus * gives the range of a single pipe as two octaves and
a fifth, TO 8ia ircvrc icai ro ^s &a murSty, and this range for the diatonic
scale alone would, without the use of harmonics, require eighteen
finger-holes. Not only is this a larger number of finger-holes than is
found on any instrument preserved to our time, but the difficulties
of performing on such an instrument would be very great, on account
of the fact that the holes nearest the mouth-piece would be very close

1 Pollux, VII. 153, representations Mon. dell* Inst XI. 27 et al.
^ Aristoxenus, Harm., ed. Marquard, p. 28.

The AvXo9 or Tibia, 31

together. Furthermore, Prodos, in his commentary to Plato's Alci-
biades, chap, d^^^ says that from each hole of the pipe at least three
tones could be produced, and that if the vaftarpvm^fiara are opened,
even more tones are possible. A possible explanation of the irapa-
rpvinifULTa has already been given (p. 11) ; the important statement
is that three tones can be produced from each hole, which can hardly
be anything else than a reference to the harmonic tones of the pipe.

In view of all this evidence it is somewhat surprising to observe
that Gevaert^ and others are of the opinion that the ancients could
not produce the harmonics on their instruments, and that even with
the pipes found at Pompeii it would be impossible to produce more
than one tone for each of the lateral holes. The inaccuracy of this
statement can be shown by actual experiment ; for with tubes made
in exact imitation of the pipes from Pompeii, with the aid of either
form of mouth-piece described above, I have produced three tones
from each hole, as stated by Proclus.

The modem clarinet has a small hole near the mouth-piece, called
the " speaker," which, when open, facilitates the production of the
harmonics, but it is not absolutely essential to their production, as
can be demonstrated by actual experiment. It is reasonably certain
that such an arrangement also existed on the ancient instruments, as
will appear later, but its absence would not be proof that the har-
monics were not used by the ancients.

In three different passages' in Greek writers the statement is made
that if the performer presses* the {cvyiy or the yAwrroi of the pipes,
a sharper tone is produced; and, although we have two different

^ Gevaert, Hist de la Musique, II. 285; von Jan, Baumeister, Denk., 556.

^ Aristotle, de audib., p. 804 a; Porphyrins, ed. Wallis, p. 249; id. ib., p. 252.

* K. von Jan, in Philologus, XXXVIII., p. 382, interprets the words Ar irUvjn rcr
rd ^e^i? as ' wahrscheinlich wenn man die Klappen dflhet.' In the other pas-
sages quoted it is, however, the 7\wrTai which are operated upon, and in Por-
phyrins, p. 249, the words are Ar Tc^<r|7 ret cArhA /ifiXXor roi% x«^<r<, which cannot
possibly refer to opening the keys of the instrument, but must refer to pressure
with the lips upon the reeds. It is therefore to be assumed that where the ^e^i?
are mentioned, the same operation is described as in the other two instances, and
that the word ^eir/17 is simply a more comprehensive term, meaning mouth-pieces
including the reeds. Moreover, it does not seem possible that either of these
expressions can mean openmg the keys of the instrument, for the keys, if we call
the bands by that name, were not opened by pressure.

3^ Alb€rt A. If^zvmni

''*^****"'^ §3r dxc port oixhc uistiiunextt oa whicii tiifC piuwuc is to be
ciet t et iy the actml opentiaa in all of these cases was andoabcedlj
the sazne. Br pcessmg with the Ups or teeth oa the reed of a dan-
net ^ or oboe aear die b^e of the reed, it is pomibtr to pto dace tibe
higher hiUxnoiiBcs of the tostnnneot, and this is oi all probobi&iT
what is meant bfdfee passages qmted above.

Forthennocev witjh seedless izfitntments^ socb as me ^iptfi jhwob-
J^Mc asfed the rXacjuBiAa^ it seesfts impossible that the penianner
shooLi not hove «£BCQvered the existence of ck harnaoQac tones; fiar
oa lay iostrxnKnt of thcs character a irenr si^t^^ aiiiimnmL e^bit to
increise tf^e power of dae tocae ruevitabhr leads to die proAactaon of
haRBooflc tooes^ It is;, theretbce,. nicaoosbntb^ that this
ootobserred b^ the lUBcfeocsw

THE snaDcx.

.^ Oils besi sabi above. e£iQe iBxkcii daniet hasL luar die
piece, a sauH hcle caibi the ''spesiker/ whkhL when opei^ rmi?ti>rs
the pecsjcmsr s> pcoiftxce wiiSBCot edibct thie haottmnr ttsoes of tiie
sDstrxmesi&. >\Ifcfti.7q^ this ^^mce &s^ nx absofixiidiy aecessaiT, it is
of ^oe |:ctsio:sc asscsCBzce tt> ^le p e r fo ciae t^ eapecaffiy n p t ^Trftiyfmg
the icsc ^iKacods cc chie B^ins^ iiaii£&^

Tbege jce a sew passuiggs itt vo^ (iffit m^ifagis wttnc^ aagijt'tne €aa>taoe
^saihuaacganggnieacqi chae Gceek jai R>3tnatt uihflLiniwag^ ilm-
«3SKSiis ^ed. Mv^tnuxL pi^ ^> :^eais cf jn .iixn?ilntm<L id> tte m£km
whii:hL w3t> called de ^rvt^^ ami w^bch^ when in Txae^ hmi t^ «fect
«fcHsax^as; 7m:h:^ :^ aBscnnneBC ; hts wxnc^ ace: taL

sbk;. ^tf w^en :Sz$ aeac hrng g ft was in ise the^ oao^ >^ chif pi^es w-^

aai£ .Qcr&is a?" ± :iif stmtf ;sS^<^n:^ ^i^oe v^^hl^ in&crxniissifi^ ii»

iwt »i i/Bryn- -^^bA^^am -i^f^tr^D* f- ■H i—* ). y^iM^mt •mil Im^yiiiiifik. «^

^u;S.2sBimsMft,4it3te^% j^mtUk I^ntt^Tk. >Bft

»j • 'i'

The AuXo9 or Tibia, 33

The effect produced by pressing with the lips on the reed, as has
already been said, is to produce the harmonics of the pipe, and it is
clear from this passage that the syrinx also served this purpose.
Again, in Plutarch (non posse suav., p. 1096 a) the syrinx is men-
tioned as a part of the avAds : &a rl r^v lorcuv avXwv 6 <rr€V(i>r€po9
PapvTtpofv ^cyycToi • icat 8ia rC T79 (nJipiyyos dFacnroificnTf, irasnv oivye-
rot roi9 <fi06yyoi%, xXivofJiivrfs Sk voXav Papwei koL awa)(Oui xpos rov
€T€po¥, itaxOus Sk d^cpov rjx^i. From this passage is derived the
additional information that the syrinx affected all of the tones of the
instrument, and although words of precisely opposite meanings, avatr-
rrav and Karaxnray^ are used in these several passages to denote the
method of employing the syrinx, the effect in all three cases was the
same.

A little hole in the instrument, like the 'speaker' of the clarinet,
would raise the tone of each hole to the harmonic * overtones,* and
such a hole could easily have been covered by a sliding band, which
in some instruments was pushed up to open the hole, and in other
cases pulled down for the same purpose, — a fact which would account
for the difference in the use of the two words, dmcnrav and Karaxnrav}

In the Anecdota Graeca Oxoniensia, Vol. II., p. 409, several defi-
nitions of the word <Tvpiy( are given, and one of them, o^/xaiW t^v
6jnp^ Twv fujvatKwv avXiav, shows that some hole of the av\6q was
actually called a syrinx.'

If the syrinx was such a device as has been described, light is
thrown on a passage in Plutarch (de mus. 21), Avruca Ti/Ac^avi/s 6
Mcyopixof ouTioi i7roX,€fjLij(r€ Tots crvpiy^iv, cuotc tov^ avAoTroiovs ovS*
iviOuvai TriairoT€ cZaorcv M ravi avAous, dAAa icat rev UvSikov <iya>V09

^ The words dwaairaw and Karaffwaw in these passages undoubtedly both mean
' to put the S3nrinx in operation/ although the manner of doing it is expressed by
words of exactly opposite meaning. It seems impossible, therefore, that von Jan
(FhiL XXXVIIL, p. 382) should be right in thinking that a joint at the lower
end of the instrument, which could be detached, was called the syrinx, and his
explanation, * Der Theil also, auf welchem man nach Abnahme der Syrinx weiter
blasen kann, heisst selbst Syrinx und das Blasen darauf a-vplrrtiP,* certainly does
violence to the passage of Aristoxenus which he quotes. It is far more probable
that the syrinx was itself a part of the instrument which could be put into opera-
tion by one or the other of the actions implied in the verbs dtnurrciw or Kara*
ffwaw, and that avptrrtip means simply to perform on the instrument when the
syrinx is in operation. ' Cf. Etym. Mag. s.v. vvfuy^.

34 Albert A. HawanL

*

fioXurra 8ca roSr' drcim;. It would seem, then, that this musician
considered the addition of the ' speaker ' to his pipes as detracting
from the dignity of his art, by making it possible for a poorer per-
former to produce, with ease, the harmonic tones^ — an art which
before this innovation required great skill on the part of the per-
former. The objections of this ancient artist to innovations can be
paralleled by the objections of modem artists to the Boehm mechan-
ism on the Ante.

There is, further, a passage in Quintilian (I. ii. 6) which may
possibly refer to this same syrinx : ' ne illas quidem circa s litteram
delicias hie magister feret, nee uerba in faucibus patietur audiri nee
oris inanitate resonare nee, quod minime sermoni puro conueniat,
simplicem uocis naturam phniore quodam sono circumliniri, quod
Graeci KararcjrXaa'fifvw dicunt ; sic appeUatur cantus tibiarum^ qucu
praeclusiSy quibus clarescunt, foraminibus^ redo modo exitu grauiorem
spiritum redduni^ If the foramina quibus clarescunt can be inter-
preted to mean the * speakers,* or holes by means of which the pitch
is raised {i,e, the harmonics produced), this passage gives additional
reason for believing that the ancients were acquainted with this de-
vice. The effect of closing the holes would be to lower the pitch of
the instrument by causing the node in the instrument to disappear,
so that the fundamental tones would be produced {recto modo exitu
grauiorem spiritum reddunt).

The instruments found at Pompeii lend support to the theory that
the 'speaker' was known to the ancients. Two of these instru-
ments (Nos. 76891 and 76892, see PI. II. figs, i and 2) have narrow
bands of silver encircling the tube near the mouth-piece, and on one
of the instruments the band is supplied with the mechanism to en-
able the performer to turn it round on the tube.*

The metal band of No. 76891 is badly oxidized, but even in the

^ In view of the fact that von Jan, in Baumeister's DenkmSler, p. 261, says that
not the slightest indication of such a 'speaker' is to be seen on any of the
remains of ancient instruments, it seems proper to state that only in the case of
the instrument No. 76892 is there any indication of a hole in the tube. The band,
however, exists on both of the instruments mentioned, and can be seen in the orig-
inal photographs of these instruments in my possession. Further than this, I have
personally examined the instruments themselves and have measured them. I also
inserted a pin into the hole in this instrument and convinced myself that the hole
actually extended into the main tube of the instrument

The A^X,o9 or Tibia, 35

photographs of the instrument it is distinctly visible. The band of
No. 76892 is very well preserved, and through it and into the tube
of the instrument is a small hole, not larger than an ordinary sized
pin. This hole is also visible in the photographs of the instrument ;
and even assuming that it is the result of accident rather than design,
the fact that bands are found in this particular place on the instru-
ment leads to the conclusion that there was a hole in the tube to be
covered, and if so, what is more probable than that it was intended
to be used as a ' speaker ' ?

KINDS OF DOUBLE PIPES.

In two passages^ from Greek writers, which have already been
quoted, the statement is made that, until the time of Pronomus,
three different kinds of pipes were used for the three different scales,
the Dorian, the Phrygian, and the Lydian, and that Pronomus im-
proved the pipes so that all three of the scales could be played on
the same pair of instruments.

This statement, however, must certainly not be interpreted as
meaning that, from this time on, all instruments were of exactly the
same length, for even after this improvement pipes of many different
sizes were manufactured and used, as is shown by the statements of
Greek and Roman writers, by the works of art, and by the instru-
ments which have been preserved.

The Romans recognized two main classes of double pipes, which
are thus described by Servius, ad Aen. IX. 615, ' tibiae aut Serranae
dicuntur, quae simt pares et aequales habent cauemas aut Phrygiae,
quae et impares sunt et inaequales habent cauemas.' What follows
in this scholium is directly attributed to Varro, and very possibly
these words, too, are taken from the same source.

In works of art the double pipe is represented either as two straight
pipes of equal length, or as two pipes of unequal length, in which
case the longer pipe is always curved at the end farthest from the
mouth, and terminates usually in a bell like the bell of a clarinet.

The curved pipe * is frequently mentioned in Roman writers, and

^ Plosanias, IX. 12. 5; Athenaeus, XIV. 31.

«Tib. II. I. 85, Phrygio tibia cunia sono; Vcrg. Aen. XI. 737, curua chores
indixit tibia Bacchi; Ov. Met. III. 533, adunco tibia coma (cf. Met. IV. 392);
Or. ex P. I. I. 39, ante deum Matrem coma tibicen adanco; Ov. F. IV. 181,
infiexo Berecynthia tibia coma.

36 Albert A, HowanL

VBsaaSif m such a maimer as to impij that it was of Plirjgian or^m,
and akiioQgfa such an instrument is not focmd in anj distinctij Greek
work of art, the use of the curved pipe in Greece can be proved from
the literatnre. Hesjchins, &t. cyB^nXifs, mentkxis diis instrument,
and GaQs it the Phr3rgian^ pipe : o rmn ^^p u y CoK mAmt. c^^a 711^ o
^urt^ yygTO>OT<r s^as, as does also Athenaens, lY. 84, cr & t^
iorripm ^Qanta o cvric*lMr ^^v t "TB ^fa op Symt P^pmr csAor r pt^ fa m
pn9fti^ oorw AcyMT n^ ^p m y v m. p ^ws fop orros* ««p o on to c^ps?
ovTM TpotroDrrownr oiaAoTOvr tw tmt qgAy i yyir niAmwL, and PoQnx, IV.
74, AvAmt & oSiy . . . IkufitH ripr fthr ^Xjpr rvftvo?, to ^ cufn^ux ^jpirywr.
d^HS 2r OBBrr^p^ ' TMT osAmt «wu >€wcir vpui/ciiiar, avAc? & rii ^prfVf, A^

^ Tbete is a pusage in the scholia to VergtU in whkh a different cxptaaatioa
of die curved pipe is given : Servins ad Aen. XL 737, kanc ti6utm Gnuci xXc^C-
sdXov t^*-»r^wf, Laiimi mascam HHam. U^ however, die xXa7tairAaf was a cur b ed
pcpCr the ezplanatiaa of it which has alreadr beea given (p. 14 sq.) is incacrect;
for, w^ die esuxpdon of the Ettie gold ornament there desctibed* all representa*
tk>ns of the tzansvexse Sate show perfectly straight tobesy and we are given to
imdeiatand that die carve in this ornament was the resolt of accident. The very
name of the instnzment seems to indicate the manner in which it was held« and
cspeci a fly if we compare It with the Latin Dame tiHa abiiqma^ which can scarce^
mean anything ebe dian transverse fiote. Farthermore, both these naznes^ xX«-
7(fnrXaf and tibia abliqwi^ are regolarly osed in the singolar, as in Pbllax, IV. 74,
and in Fliny, N. H. VTL 204, obliqaam txbtam Midas in Phrygta, (inaenit) geminas
tibias >larsyas in eadem gente, and in soch a manner as bo indicate that the
™»'«'^a« osed only one of them at a time; while die carved pipe» althongh the
name is osed in the singolar, is regularly represented in works of art as one of a
pair» die other pipe being straighL It therefore seems &ir to *«8^»im^ that the
schofiast is misrakfn in his statement, and that the curved pipe was teaify the
Berecynddan horn which formed one of the pipes of the Phrygtan pur, and
particniariy as di^ xkaiimm is foami only in the faUer commentary to Vergil first
edited by Daniels^ and is possbly of late origin. In only one other passage is
dtere any mention of the tibia uaaca (Solinns^ 5. I93» and firam diis passage no
fordier inf or ma tion can be gained.

* This is the most probable conjecture for die htarifttn tmv o^w* of die 3Cs8^
althoo^ even this b inconststent with the passages quoted from Hesychius and
Adienaens^ in that it implies that bodi pipes of the pair were curved. I know of
no ancient worit of art in which both pipes are curved* although two old engrav^
xng% whkh claim to represent such works of art» do show both pipes curved.
Both of diese e ng r a v i ngs are in P. S. Bartols* admtranda Rom. antiq. uestig.« PL
17 and 47» and the latter is reprodnced* with the figure turned from right to left,
in BartDlxnus^ de tibia ueterum, p. 52. In the original relief however^ from
which the first engjcavtng was made» <me of the i astminea ts is straight (c£ Heibig^

The AvXo9 or Tibia. 37

The natural inference from this evidence is that the impares were
always Phrygian pipes, exactly as was stated by Servius, and, further-
more, that one of these pipes was always curved at the end.

Pollux, IV. 80, describes the ya/ni/Xiov avXi^/ia as two pipes, of
which one was larger than the other ; and if this statement is cor-
rect, these pipes, according to the theory here advanced, were
Phrygian pipes. It is to be feared, however, that, as has been sur-
mised by von Jan,^ Pollux has here explained as a variety of instru-
ment what was merely a form of composition for the instrument, for
avXvffia is the name, not of an instrument, but of a composition. A
possible explanation, however, is that this particular form of com-
position was always played on the Phrygian pipes.

The regular name in Greek for the Phrygian pipes seems to have
been avAoc iXv/ioi, as is shown by Athenaeus, IV. 79, rov^ yap ikvfwv^
avXovs . . . ovK SXXovi Ttvas dvat ixovoyitv rj rciv^ ^pvyCovs, and by the

passage already quoted from Pollux.

In the didascalies to the plays of Terence there is no mention of
the Phrygian pipes, but according to this view, the impares which
were used to accompany the Heautontimoroumenos and the Phormio
were, nevertheless, Phrygian pipes.

In performing on the Phrygian pipes, according to Hesychius (s.v.
cyiccpaiiXi;^), the curved pipe was held in the left hand ; and although
the works of art give no conclusive evidence on this point, the curved
pipe being represented now in the right hand and now in the left
of the performer, there are reasons for assuming that Hesychius is
correct in his statement. The Egyptian monuments, although they
do not represent the Phrygian curved pipe, do show pairs of straight
pipes in which one of the pipes is longer than the other. In all these
cases the longer pipe is held in the left hand of the performer (Loret,
p. 139). Again, if, as is implied by Servius in the passage quoted,
the impares were always Phrygian pipes, the didascalies to the plays
of Terence support the statement of Hesychius ; for, although two
right-hand pipes are mentioned in these didascalies, a pair of left-
hand pipes is never mentioned. It would seem, therefore, that the

Ffihrer durch die Sffentlichen Sammlangen Rom's, II., No. S05, note). The other
eDgraving I have been unable to identify with any existing work of art in Rome.
Possibly the author wrote hkp^ rwr adXwr.
^ Banmeister, Denkm., p. 563.

38 Albert A. Howard.

musician never changed the instrament wiiich he held in his right
handy but that, when the character of the music changed, he changed
the pipe held in his left hand. Furthermore, Varro, as quoted by
Servius, Aen. IX. 615, tibia Phrygia dextra unum foramen habet,
sinistra duo, quorum imum acutum sonum habet, alUrutn graueniy
implies that the left Phrygian pipe produced the lower tone, and, as
the curved pipe is always longer than the straight one in the represen-
tations, it would of course produce the lower tone ; and therefore, if
Varro's statement is correct, the curved pipe was held in the left hand.

Theophrastus, H. p., IV. 11. 7, says that of the two mouth-pieces
which are to be made from the same joint of the reed, the one
towards the root was used for the left pipe, the one towards the top
of the reed for the right pipe. If any inference can be drawn from
this statement, it is that the pipe held in the left hand was intended
to produce the deeper and more voluminous tone, for the internal
diameter of the joint would be slightly greater in the part toward
the root than in the other part.

As compared with the other forms of pipes, the Phrygian pipes are
said by Porphyrius ^ to have had a bore of smaller diameter, and to
have emitted much deeper tones. The latter of these two state-
ments agrees exactly with what is known of the tone of the Phrygian
pipe from other sources, in which stress is always laid on the deep
tones * of the instrument.

SIZES OF PIPES.

The other main class of double pipes consisted, according to Ser-
vius, of a pair of pipes, equal in length and with the same internal
diameter, spares sunt et aequales habent cauemaSy but there is noth-
ing in this statement which precludes the possibility of a variety of
instruments composing this class. The works of art and the testi-
mony of ancient writers show clearly that such variety existed, and
even in ancient times there existed a general classification of the
instruments, which has been used by Gevaert ' as the basis of his
description of the avXot. This classification is found in Athenaeus,

^ Porphyrius, p. 217, ed. Wallis; cf. Plut non posse suav. 13.

^ Sophocles, frag. 468, ^piicurra fip6fiop, explained by Hesychius as referring
to the Phrygian pipe; Athenaeus, IV. 84; cf. Eurip. Hel. ly^g, fiap^fipofutw ci^X6r;
Anstoph. Nub. 313, /3aptJ/3po/40j adX6t; etc.

' Hist, de la musique, II. 273.

The AvXo9 or Tibia, 39

XIV. 36, and is taken at second hand from Aristoxenus, ircpt avXcuv
rpva-€o>9. Five classes of instruments are there enumerated, — the
mpOhfUAy ina&ucoi, Ki&apum^pioi, rc\eioc, and vTrcprcXcioi, but in Athe-
naeus, IV. 79, the last two of these classes are called dlvSpcioc.

This classification would seem to have been made with reference
to the pitch of the instruments ; for Aristoxenus (p. 28, ed. Mar-
quard), after giving the range of a single pipe as two octaves and a
fifth, says that the highest note of the irapOtvioi compared with the
lowest note of the vircprcXcioi gives an interval of more than three
octaves, and evidently he is here comparing the two classes of instru-
ments which show the greatest variation in pitch. In further proof
of this may be quoted Aristotle, H. A. 7. i. 7, where it is stated that
the vapOeytoi were higher in pitch than the Trat&Kot, and Athenaeus,
IV. 79, where it is stated that the i^fuWoc, which were the same as
the miSucoi, were shorter than the rcXctoi, and would therefore pro-
duce higher tones. As the virc/:>rc\cuK were used to accompany men's
voices (cf. Pollux, IV. 81), the pitch was evidently lower than that
of any of the other classes named.

Not all of the varieties of instrument mentioned in ancient writers
can be classified with certainty, but so far as possible an attempt will
be made to arrange them according to this scheme of Aristoxenus.

As pipes of very high pitch, which doubtless belong in the class of
the vap^cFioi, are mentioned the yCyypoi^ (Latin gingrinae), which
were not more than a span in length, and gave very shrill tones ; the
miluina, described by Festus (p. 123) as genus tibiae acutissimi soni;
the o-icvTaXta,' which were very short pipes ; and probably also the
'rapdrprfToi, of which Pollux (IV. 81) says : Opi^vois rjpfunrQv^ 6(v jcou
vtaSl^ nrcorres, and the viyXapoi, Pollux, IV. 82.

The ^/Atairoc' were identical with the inu&Kot, and were shorter
than the rcXciot, as appears from Athenaeus, IV. 79. They were
used at banquets. Probably the irapocvtoi also belonged to this class,
for of them Pollux, IV. 80, says, a-futcpdi pkv, taoi 8* cE^^a>, rrp^ yap
hrvnfra avp.voo'Cif irpcrctv. Gevaert adds to this class the avXoc ip.fia-
n/pcoi and the SoktvXucoi mentioned by Pollux, IV. 82, the former
of which was used in processions and the latter to accompany the

^ Athenaeus, IV. 76; Pollux, IV. 76; Festus, p. 95; Solinus, 5. 19.
« Athenaeus, IV. 79; Pollux, IV. 82.
* Ct Heiychius, s.v. Iifdowoi.

40 Albert A. Howard.

hyporchemes; but it is doubtful whether such separate varieties
actually existed, and Pollux says that by some they were said to be,
not instruments, but forms of melodies, — which shows that he, at
least, knew nothing about them at first hand.

The Ki6(Lpi<mjptoi were used in accord with the lyre, as their name
indicates (Pollux, IV. Si), and in this class ought probably to be
placed the fua-oKoiroi, which, as we are informed by Photius and
Hesychius, were somewhat shorter than the rcXccot : vTroSecorcpoc riav
rcXciW, Kal ovrcs ficcroc. Here belong also the Phrygian pipes, at
least in Roman times, for Horace twice mentions them in connection
with the lyre (Od. IV. i. 22, and Epod. 9. 5), and, furthermore, the
/Aaya&9 avXos (Athenaeus, XIV. 35-36). The testimony in the pas-
sage cited makes it clear that such an instrument existed and be-
longed to this class, although the word fiaya&iq when used alone
probably meant a kind of harp or lyre.

The TTvOiKoi, the instruments used in the Pythian contest, belonged
to the class of rcXctoi (Pollux, IV. 81). According to Diomedes
(p. 492, ed. Keil), these instruments were used to accompany the
cantica of the Roman comedy.

The avXoi v7rcprc\ctoi were used to accompany men's voices, and
were the lowest in pitch of all the instruments. In this class belong
probably the pipes used in the temples ; cf. Mar. Victor., p. 44, ed.
Keil, ' spondaeus dictus a tractu cantus eius qui per longas tibias in
templis supplicantibus editur, unde et spondaulae appellantur qui
huius modi tibias inflare adsuerunt.'

There remain a number of varieties of the pipe which, in the
absence of any testimony concerning their nature, cannot be classed
with any degree of certainty; these are the wkvoi (Poll. IV. 76),
the StWoi (Athen. IV. 79, Poll. IV. 76), the iBovOoi (Poll. IV. 77),
the p6iJ.pvKt% (Hesych., s.v.), and the vmyrprjroi. (Athen. IV. 79),
while the words vTroTrrcpoc (Poll. IV. 77) and inroOtarpoi (Poll. IV.
82) are probably corrupt readings for xnrorprjroi.

The fl>Srny$ is said by Athenaeus, IV. 78, to liave been a form of
irAayiavXo9 ; its home was in Egypt.

The entire range of these five classes was, as stated by Aristoxe-
nus (cf. p. 39), more than three octaves ; and the classes themselves
must have corresponded pretty closely with the ranges of the human
voice, — soprano, alto, tenor, baritone, and bass, — as is clear from

The kv\o% or Tibia, 41

the names of some of the classes and from the direct statements
about the others. Although the combined range of the five classes
is nowhere stated, we can perhaps ascertain it approximately by the
following method of reasoning.

The longest straight pipe which could be used, without keys to
close the holes beyond the reach of the fingers, and produce a
continuous scale, would be about %% cm. in length ; the hole nearest
the lower end of the instrument would then be 78.1 cm. from the
mouth of the performer. Such an instrument, if it belonged to the
class of stopped pipes would give as its lowest tone G of our scale.

The shortest instrument mentioned in ancient writers is the yiyypo?
one span or 32.18 cm. in length. A stopped pipe of this length would
give as its lowest tone g' of oiu: scale ; an open pipe would give g".
Possibly Aristoxenus did not include these short pipes in his classifi-
cation, for he seems to be speaking of instruments in ordinary use
and the ytyypoc do not seem to have been instruments of this charac-
ter. The normal range of soprano voices is given in RiemantCs
Musik'Lexicon, as from c' to 3!\ and it is not improbable that c' of
our scale was the lowest note of the irap^moc as a class. The rcXeioc
were probably about an octave below the iroptfcyioi, for this is the nor-
mal interval between soprano and tenor voices. Their lowest tone
would then be c of our scale ; their length, if stopped pipes, would
be about 62 cm.

The Toi&icoc were also known as ^fuoiroc, a name which suggests
that they may have been half as long as the instruments of some other
class, perhaps the vwtfyrikaou On this supposition their lowest tone
was g ; their length, if stopped pipes, about 44 cm. As the lowest
note of the KifiapurnjpuK we might assume e of our scale, for one vari-
ety of this class, the fua-oKmroi, seems to have had a range midway
between that of the re\cioi and the inu&xoc.

It is further probable that within these different classes there were
slight differences in pitch, and that, for example, not all vircprc\eioi
gave a tone so low as G of our scale.

Of the instruments in use among the Romans, those which have
for us the deepest interest are the ones which are mentioned in the
didascalies and in the commentaries to the plays of Terence, and it
is desirable, so fisir as possible, to ascertain the exact character of
these instruments.

42 Albert A. Howard.

In the didascalies the following names of instruments are men-
tioned : tibiae pares^ tibiae serranae^ duae dextrae^ and tibiae im-
pares; and this list practically coincides with the statements of
Donatus at the beginning of the several plays. If it be assumed
that these are all different instruments, there are at most four va-
rieties; but as a matter of fact this number must be somewhat
modified.

Placing side by side the statements of the didascalies and the
statements of Donatus, which probably do not in all cases refer to
the same performances of the pla3rs, the following scheme is pre-
sented, in which I. represents the didascalies as they are found in
the Bembine Ms. ; II. the same, as they are found in the other Mss. ;
and III. the form in which they are preserved by Donatus.

I. II. III.

Andria • paribus (dextris ut

sinistris)
Eunuchus duabus dextris duabus dextris dextra et sinistra

Heauton primum imparibus primum imparibos

deinde duabus dextris deinde duabus dextris
Hecyra tibiis paribus tibiis paribus tibiis paribus

Phormio tibiis imparibus tibiis serranis

Adelphoe tibiis serranis tibiis serranis tibiis dextris id est

Lydiis

Only in the case of the Hecyra is there absolute agreement in all
three statements, — a fact which may be due to a change in the
instrumentation of the other plays for later performances.

From Servius it has already been seen that the Serranae were
pares, and the duae dextrae, as the very name indicates, must also
have been pares. The imparesy according to Servius, were always
Phrygiacj and therefore the dextra and sinistra y if they differed in
length, were Phrygian pipes impares, or, if they were of the same
length, were pares. This reduces the actual number of varieties to
three, which accords perfectly with the statement in Diomedes
(p. 492, ed. Keil), where we are told that the comedies were
accompanied by pares, impares, or serranae. Furthermore, in the
treatise de comoedia, it is said of the comedies : agebantur autem
tibiis paribus aut imparibus et dextris aut sinistris. Dextrae autem
tibiae sua grauitate seriam comoediae dictionem pronuntiabant.

The AvXo9 or Tibia. 4J

Sinistrae et serranae acuminis leuitate iocum in comoedia ostende-
bant, ubi autem dextra et sinistra acta fabula inscribebatur, mistim
ioco et grauitates denimtiabantur. Two left-hand pipes are nowhere
mentioned in the didascalies and, if the sinistra is the curved Phry-
gian pipe, cannot have been used as a pair, so that even Donatus
seems to mean only three different varieties.

Of these three varieties the duae dextrae are said by Donatus, in
the introduction to the Adelphoe, to have been the same as the
Lydian pipes : modulata est autem tibiis dextris, id est, Lydiis, ob
seriam grauitatem qua fere in omnibus comoediis utitur hie poeta.
If this quality of grauitas is prominent in nearly all the plays of
Terence, and if it can be expressed only by the Lydian pipes,^ it
seems necessary to assume that the pares mentioned by Donatus, in
the introductions to the Andria and the Hecyra, were also Lydian
pipes. In the impares the right pipe was perhaps the same as in the
duae dextrae, the left pipe longer and curved at the end. The ser-
ranae, although pares, were probably both shorter than the dime
dextrae.

METHOD OF PERFORMING ON THE DOUBLE PIPE.

In deciding how the musician performed on his two pipes at once
there are three possibilities to be considered. He may have pro-
duced the same tone simultaneously on both pipes in unison; he
may have produced tones separated by an octave ; or he may have
played the melody on one pipe and an accompaniment in accord,
with smaller intervals than the octave, on the second.

The third of these methods seems the most probable, for the fol-
lowing reasons. The auAoc cXvfioc {tibiae Phrygiae^ were, as has been
shown, impares, or unequal in length, and consequently on one of
them could be produced tones lower than could be produced on the
other, but this was an absolutely useless featiu'e in case the pipes were
played in unison. Again, there would be very great difficulty in
keeping the instruments absolutely in tune, as would be necessary in

* Aero (ad Hor. Carm. IV. 15, 30) assigns to the Lydian pipes a different
character : Lydiis tibiis laeta canebantur, Phrygiis tristia. But in the commen-
tary to Carm. III. 19. 1 7, both Aero and Porphyrio say that banqueters regularly
danced to the music of the Phrygian pipes; and, as these instruments were also
prominent in the orgiastic worship of Cybele and of Bacchus, it seems improbable
that their music should have been of a very doleful character.

44 Albert A. Howard,

|^]ring them in unison. Forthennore, a passage in Yano (rer. rnst
X. 2. L5) shows OMichisiTeljr for Roman times that the pipes were
not played in miison : ' certe, inqmt Fandaninsy alint pasdo et aliot
agri coltmay sed adfinis et at dextra tibia aha qnam sinistra, ita nt
tamen sit qoodam nxxio conimicta, quod est ahexa exosdem car-
minis modorom incentima aha smccenttua.

These words of Varro, it is tme, might be apphed to the second
mediod of performance mentioned above, but assaming that the pipes
were reed instruments c^ cyhndncal bene and that each was provided
with finger-holes enough to enable the performer to {ffodnce a con>
tinoous scale by the use of the harmonics^ if the pipes were equal in
length the range, when playing in octaves, would neverthdess be
confined to exactly four tones and their semitcHies for each series
of harmonics of the instrument; for, admitting die possibflity of
performing on two pipes at once, before the invention of the syrinx^
it was impossibie to produce simultaneously fundamental tones on
one pipe and harmonics on the other, an operation whkh would
necessitate blowing gently into one pipe and violently into the other
at the same time. Either, then, the practical range of the pair must
have been Hmited to four consecutive tCHies and their semitones, or
the pipes were played in accord, with smaller intervals dian the octave.
It is inconceivable, from similar reasoning, diat, with instruments of
equal length belonging to the class of open pipes, the performer
should ever have played in octaves.

Furthermcnre in the worics of art the two hands of the performer
are always near together, and this would never be the case when
playing in octaves on stopped pipes. This method of representation
was perhaps merely a conventional one ; but it is difficult to see how
the convention could arise if^ in actual performances^ the bands were
alw^rs at considerable distances from each other.

The existence of polyphonic music in bodi Greek and Roman
times can no longer be doubted,' and to judge frt^m Phitarch (de

^ With the * spesker/ or syrixn* this difficulty i» entirely obvisted; and if the
eiatence of this mechanism is admittsd» there is no ph;rstcal impoasibtlity in
phi]rxng the entire scale in octaves.

* Ct Karl von Jan, in Fteckeisen's JahrbOcher, 1S79, p. 585; R. Westj^ial,
Die Mnsik des Griechischen AIterthams» p. t68 ^. The following pasnges I
have neves seen qnoted in support of this theory : Boeotios* Inst mns. L 3, Sn^

The kxiXM or Tibia, 45

mus., ch. 39), it was on the avXos that music in two parts was first
performed. Apuleius, Flor. III., assigns to Hyagnis the honor of
having first performed on his instruments music of this sort : ' primus
Hyagnis in canendo manus discapedinauit, primus duas tibias uno
spiritu animauit, primus laeuis et dexteris foraminibus acuto tinnitu
et graui bombo concentum miscuit.'

It is only necessary to add that there is no unsurmountable difficulty
in performing on such instruments as were found at Pompeii a simple
melody and its accompaniment ; the fingering of the instruments is
the only difHculty^ and that could be acquired by practice.

The accompaniment in Greek music, as is now universally agreed,
was higher than the melody.* This is conclusively shown for the
lyre by Aristotle, probl. XIX. 12, and it is safe to conclude that it
holds true for the pipes as well. The same relative position of
melody and accompaniment is still preserved in the music of the
Greek church' and in much of the music of the early Christian
church. The natural inference from this fact is that the melody was
performed on the pipe held in the left hand ; for, as we have seen
above, in the case of the Phrygian pipes the longer instrument, and
the one which would consequently produce the lower tones, was held
in the left hand ; and if on the Phrygian pipes the melody was per-
formed with the left pipe, probably the same was also true of the
other varieties of the instrument.

It is hardly conceivable that the Romans did not hold their pipes
exactly as the Greeks did, more particularly as the art of playing on
this instrument was believed by the Romans themselves to have been
borrowed from the Greeks. The following passage from Varro (rer.

' Est enim consonantia dissimiliam inter se uocum concordia '; id. ib. I. 28, ' Quo-
tiens enim duo nerui uno grauiore intenduntur simulque pnlsi reddunt permixtum
quodammodo et siiauem sonum, diiaeque uoces in unum quasi coniunctae coa-
lescunt; tunc 6t ea, quae dicitur consonantia, Cum uero simul pulsis sibi quisque
ire cupit nee permiscent ad aurem suauem atque unum ex duobns compositum
ionum, tunc est, quae dicitur dissonantia.' In the experiments of Pythagoras, as
described by Censorinns (de die nat, X. 10), it would seem that two pipes of
different lengths were blown upon at the same time to produce the accord of the
fourth, the fifth, and the octave.

^ Westphal, Harmonik, p. 1x3; Gevaert, Hist, de la Mus., I. 364; K. von
Jan, Jahrbficher, 1879, p. 583.

* Christ, Anth. Carm. Christ., p. 1 13.

46 Albert A. Howard.

rust. I. 3. 15 sq.) would seem, however, to show that the accompani-
ment at Rome was in his time played on the pipe held in the left
hand : ' Certe, inquit Fundanius, aliut pastio et aliut agri cultura, sed
adfinis et ut dextra tibia alia quam sinistra, ita ut tamen sit quodam
modo coniuncta, quod est altera eiusdem carminis modorum incentiua
altera succentiua. Et quidem licet adicias, inquam, pastorum uitam
esse incentiuam agricolarum succentiuam auctore doctissimo homine
Dicaearcho, qui Graeciae uita qualis fuerit ab initio nobis ita osten-
dit, ut superioribus temporibus fuisse doceat, cum homines pastoriciam
uitam agerent neque scirent etiam arare terram aut serere arbores
aut putare ; ab iis inferiore gradu aetatis susceptam agri culturam.
Quocirca ei succinit pastorali [L], quod est inferior, ut tibia sinistra
a dextrae foraminibus.' The only difficulty lies in the last sentence^
for no inference can be drawn from the first sentence, since the
natural tendency would be to mention the right hand before the left,
and further to mention the incentiua before the succentiua, as is indi-
cated by the repetition of these words in the same order in the very
next line. In the last sentence the word pastorali must be omitted,
or ei must be changed to ea, as is shown by the context, pastorum
uitam esse incentiuam; but the words ut tibia sinistra a dextrae fora-
minibus are incomprehensible. The use of the preposition ctb with
inferior is very unusual, if not unparalleled, and it is not impossible
that the whole phrase beginning with ut is an interpolation.

The passage from Apuleius, Flor. III. (quoted on page 45), seems
further to imply that the higher tones were produced on the instru-
ment held in the left hand, the deeper tones on the one held in the
right hand. In the words, 'primus laeuis et dexteris foraminibus
acuto tinnitu et graui bombo concentum musicum miscuit/ laeua^
dextera, foramina are clearly equivalent to laeua, dextra, tibia, and,
unless we have in this passage a singular instance of chiasmus, the
inference is unavoidable that the acutus tinnitus was produced on
the left pipe, the grauis bombus on the right pipe. If, therefore, the
accompaniment was higher than the melody, it must have been per-
formed on the instrument held in the left hand.

It is very improbable that the Romans made any change either in
the manner of holding the pipes, or in the relative pitch of melody
and accompaniment, and therefore the conclusion seems inevitable
that both Greeks and Romans regularly played the accompaniment
on the pipe held in the left hand.

The kvlsM or Tibia, 47

But the Phrygian left pipe was longer than the right, and therefore
the accompaniment, if played on it, would at times be lower than the
melody, if any use whatever was made of the lower tones. That the
lower tones were used is indicated by Servius in the passage quoted
on page 38, where he attributes to Varro the statement that the right
Phrygian pipe had one hole and the left two, one of which gave a
deep and the other a high tone. Although it is very improbable that
the number^ of finger-holes on either pipe was so small as is here
stated, it is significant that both high and low tones are attributed to
the pipe held in the left hand. The Phrygian pipes were a prominent
feature in the orgiastic worship .of Cybele and of Bacchus, and possi-
bly in the music which accompanied this worship the accompaniment
was performed on the pipe held in the left hand, and alternately rose
above the melody and droned below it.

DESCRIPnON OF EXISTING INSTRUMENTS.

The instruments found at Pompeii in 1867, ^"^^^ ^ number, are
straight tubes of ivory, which were originally covered with close-

^ The use of the present tense ' babet ' in this passage implies that this descrip-
tion of the Phrygian pipes applies to instruments in use in Varro's time; but it
must be remembered that Varro, in stating that the tibia in ancient times had only
four holes (cf. page 4, note), adds the statement that he had actually seen such
instruments in the temple of Mars3ras, leaving us to infer that all instruments in
use in his time had a greater number of 6nger-holes. Not only is it incredible
that any attempt was ever made to play a melody and its accompaniment on a
pair of pipes such as is here described, one of which would have a range of two
notes and the other of three, but the works of art which I have seen, with possibly
a single exception, do not confirm this statement as to the number of finger-holes.
The single exception is found in a relief on a sarcophagus in the Louvre, Frohner,
No. 500, where a bacchant is represented playing on Phrygian pipes. The curved
pipe, which is held in the ri^ hand, has two side-tubes near the lower end,
but the straight pipe has neither tubes nor holes.

Although it is very improbable that the number of finger-holes was so limited
as is indicated by this statement of Varro, it is not unlikely that the left pipe had
a greater number of holes and a wider range of notes than did the right pipe.
Festus, p. 109, says : impares tibiae numero foraminum discretae, and although
I have seen but two works of art in which the pipes of a Phrygian pair differed in
respect to the number of holes, in both cases the curved pipe had the greater
number. These two works of art are a sarcophagus. No. 751, in the Lateran
Museum at Rome, and the sarcophagus, No. 300, in the Loavre.

48 Albert A, Howard,

fitting bands of metal, probably silver. This metal is in some places
badly oxidized, and in a few instances has entirely disappeared ; but,
on the whole, the state of preservation is remarkable. Each tube
has, at the end most remote from the finger-holes, a pear-shaped
bulb, also of ivory, 3.8 cm. in length, into which is inserted a flaring
tube of ivory, 3.5 cm. in length, provided with a socket into which
the mouth-piece was inserted when the instrument was in use.

Through the kindness of the director of the Naples Museum I was
permitted to examine these instruments and to make careful meas-
urements of them, and from these measurements it is possible to
reconstruct the ancient instruments, and to ascertain by experiment
their scales.

The bore of all four of the instruments is cylindrical and 0.95 cm.
in diameter, the external diameter of the tube being in each case
1.42 cm.

The following measurements were all made with a standard rule
measuring to the sixty-fourth of an inch, and were all repeated to
insure accuracy. The measurements as given in the text have been
reduced to the metric system from the original measurements in
inches, which are added in the foot-notes. The distances were all
measured from the lower end of the tube to the nearest side of the
hole or band. In some instances the holes in the tube are still
covered by the bands, and in such cases the measurement indicates
the distance to the hole in the band and the diameter of the hole.

I. Catalogue No. 76891.

Total length, 49.68 cm. A, 4.28, diam. 0.63 ; B, 7.46, diam. 0.63 ;
C, 11.74, diam. 0.63; D, 13.97, diam. 0.63; E, 15.87, diam. 0.63;
F, 18.09, diam. 0.79 ; G, 19.52 to 21.9, a band in which no hole was
visible ; H, 23.49, diam. 0.79 ; I, 26.51, diam. 0.71 ; K, 28.74, diam.
0.79 ; L, on the under side of the tube, 31.75, diam. 0.79.*

Close to the bulb in this instrument is a narrow band, somewhat
corroded, but which can be seen distinctly in the representation
(Plate 11. i).

1 Totol length, 19^. A, iH. <iiam. J; B, 2^. diam. \\ Q 4f, diam. \\ D, sfc
diam. \\ £, 6}, diam. \\ F, 7^, diam. ^; G, 7}^ to 8|, a band in which no hole
was visible; H, 9f, diam. ^^\ I, iO]V> diam. A; K, ii|<^, diam. ^^\ L, on the
nnder side of the tube, 12}, diam. ]\.

I

J

The Av>u6<: or Tibia. 49

The bands covering holes F, H, and I have traces of the mechan-
ism described above to aid in turning the bands. This was probably
the case with all of the bands originally, but the silver is so badly
oxidized that the traces of the mechanism have disappeared. The
holes Qy Dy and E are covered by the bands ; the measurements are
of the holes in the bands. In the other three instruments there is
a hole corresponding to the band G in this instrument, yet I failed to
discover any hole in this band.

II. Catalogue No. 76892.

Total length, 52.7 cm. A, a band 1.9 cm. wide, in which there
is no hole ; B, a band from 1.9 to 4.44, with turning mechanism, but
no hole ; C, 6.03, diam. 0.63 ; D, 9.20, diam. 0.87 ; £, i i.i i to 13.33,
a band in which there is no hole; F, 13.65, diam. 0.79; G, 16.03,
diam. 0.79; H, 18.09, diam. 0.79; I, 20.63, diam. 0.95; K, 24.76,
diam. 0.95; L, 27.94, diam. 0.63; M, 29.21, diam. 0.63 (this hole
is on the under side of the tube) ; N, 32.22, diam. 0.79.^

Close to the bulb on this instrument there is a band of silver
1.58 cm. in width and provided with the mechanism for turning.
The silver is slightly oxidized, but close by the bulb is a hole in
the tube itself not larger than a medium-sized pin. The band and
the hole can be seen in Plate II. 2. The bands of B, C, D, F, G, H,
have the mechanism for turning, although in the case of G there is
only a trace of it still left. In each of the other instruments there
is a hole corresponding to the band E of this instrument, which has
no hole.

III. CAtalogxje No. 76893.

Total length, 49.21 cm. A, 1.27, diam. 0.63 ; B, 4.12, diam. 0.79 ;
C, a band without hole or mechanism for turning ; D, 9.68, diam.
0.79; E, 12.06, diam. 0.55 ; F, 13.81, diam. 0.63; G. 15.87, diam.
0.63; H. 18.09, diam. 0.63; I, 20.16, diam. 0.79; K, 23.33,

' Total length, 20}. A, a band } in. wide, in which there is no hole; B, a band
from f to l}, with turning mechanism, but no hole; Q 2}, diam. \\ D, 3f , diam.
^; £, 4| to 5}, a band in which there is no hole; F, 5}, diam. ^; G, 6^,
diam. -f^i H, 7|, diam. ^; I, 8|, diam. f ; K, 9}, diam. f ; L, ii, diam. \\ M,
1 1 1, diam. \ (this hole is on the under side of the tube) ; N, 12ft, diam. A.

50 Albert A. Howard.

diam. 0.63; L, 25.71, diam. 0.79; M, 27.95, ^^"^^ ^*^\» ^^ ^c
under side of the tube ; N, 30.95, diam. 0.63.^

The bands of B, D, F, G, H, L have the mechanism for turning,
which has, however, almost disappeared in the case of D, F, and H.
The holes £ and H are still covered by the bands. This instrument
is shown in Plate II. 3.

IV. Catalogue No. 76894.

Total length, 53.65 cm. A, 1.4, diam. 0.63 ; B, a band without a
hole from 2.54 to 5.08; C, 6.03, diam^ 0.71 ; D, 8.89, diam. 0.79;
£, 12.22, diam. 0.63; F, 14.28, diam. 0.79; G, 16.51, diam. 0.63 ;
H, 18.41, diam. 0.63; I, 2047, diam. 0.63; K, 22.70, diam. 0.63 ;
L, 24.76, diam. 0.63 ; M, 27.30, diam. 0.64 ; N, 29.36, diam. 0.64 ;
O, 33.17, diam. 0.64 ; P, 35.87, diam. 0.64 ; Q, 37.95, diam. 0.56.'

The hole Q, as can be seen from Plate II. 4, is now slightly elon-
gated, but this is clearly the result of an accident ; all of the holes in
this and in the other instruments are circular, or very nearly so, and
with very few exceptions the edges and inner surfaces of the holes
are well preserved in all of these instruments.

There is no hole on the under side of this instrument, and conse-
quently the engraving of it shown by Gevaert (Vol. II., p. 280) is
incorrect in that particular, as also in the representation of hole A,
which in the original is very near the lower end of the tube. Only
the band of hole H has now the mechanism for turning the band,
although originally most of the bands were doubdess thus equipped.

The holes A, C, E, G, H, I, K in this instrument are still covered
by the bands. The metal casing near the bulb is so badly oxidized
that it is impossible to say whether there was a separate band at this
point, as in Nos. I. and II.

Each instrument has a separate band for each hole, or did have

^ Total length, 19}. A, |, diam. \\ B, i}, diam. ^\ Q a band without hole
or mechanism for turning; D, 3^, diam. -j^; £, 4}» diam. ^; F, 5^, diam. \\
G, 6i, diam. J; H, 7J, diam. J; I, 7J{, diam. ^\ K, 9A» diam- \\ L, loj, diam.
^; M, II, diam. ^,.on the under side of the tube; N, 12^, diam. \.

* Total length, 21}. A, -j^, diam. \\ B, a band without a hole from i to 3;
C, 2|, diam. /j; D. 3 J, diam. ^\ E, 4}|, diam. \\ F, 5}, diam. ^^\ G, 6^,
diam. \\ H, 7^, diam. \\ I, 8|^, diam. \\ K, 8({» diam. \\ L, 9f, diam. \\ M,
to}, diam. J; N, ii|^, diam. J; O, x^i^^ diam. J; P, 14^, diam. \\ Q, 14^4,
diam. j^.

The AiXd? or Tibia. 51

originally ; but a careful search failed to reveal any band in which
there were two holes as described by Gevaert (Vol. II., p. 646), al-
though the fac-simile of No. II. in the Naples Museum does have
one band in which there are two holes so arranged that both cannot
be open at the same time.

It is to be observed that no two of these instruments are of exactly
the same length, but that I. and III. differ in this respect by only
0.42 cm., and that II. and IV, differ by 0.95 cm., while between
I. and IV., the two extremes, there is a difference in length of
3.44 cm., or enough to make a difference in pitch with pipes of these
dimensions of practically ah entire tone. It would appear, therefore,
that these instruments belong in two different classes. Experiments
were first made with a pipe corresponding to No. IV. Into this a
clarinet mouth-piece made from a slender rod of bamboo was in-
serted, in form like the one described in von Jan's article in Baumeis-
ter's Denkmaler. Repeated experiments showed that the proper
length for the mouth-piece was not to exceed 5 cm., for with a
mouth-piece of this length the complete chromatic scale can be pro-
duced through one octave and without the aid of the three holes
nearest the mouth-piece. If these three holes are then opened suc-
cessively, the second, third, and fourth of the octave can be pro-
duced ; and then, with all the holes closed, the first harmonic of the
closed pipe,^ i.e, the fifth above the octave, is produced, which is the

1 Both Gevaert (Vol. II., p. 285) and von Jan (Baumeister^s DenkmSler, p. 556)
say that this is impossible without the aid of the ' speaker/ and the latter says :
" Ein cylindriches Rohr von so einfacher Konstruction wie die aus Athen und
Pompeji erhaltenen Auloi ermSglicht nach Aussprache der SachverstSndigen fUr
jeden Griff gar nnr einen einzigen Ton," and farther on : " Ob nun jene Auloi
eine andere Konstruction hatten als die von uns in Athen und Pompeji gefun-
denen Reste, oder ob die Kunstfertigkeit der Bl&ser auf denselben Resultate er-
zielte, die wir jetzt nicht nur praktisch sondem auch theoretisch nach den Gesetzen
der Akustik ffir unm5glich halten, das l&sst sich vorl&ufig nicht entscheiden."

It is, however, a mistake to suppose that the laws of Acoustics apply only to
elaborately constructed pipes. These lavi^ apply to all pipes, however rudely
constructed; nothing could be simpler than the brass tubes with which these
experiments were conducted, and yet I have repeatedly produced the harmonics
on them, without the aid of a ' speaker.'

Even on the clarinet it is possible to produce the harmonics without the aid of
the ' speaker/ although it is by no means easy, as I have been informed by a prac-
tical manufacturer of clarinets. The tube of the clarinet, however, has a much
wider bore than that of these pipes from Pompeii.

52 AAa^ A. &mmn£

OBBt Qoie WL the ifiiifniiii. anif* ^ Lff\uiieit same skifi. 2m1
tD pimhw^fc tixis partxcaiic irnifr, bnt witit case ic can. be pnadaced*
Tlbe HiUimmic.?* of the amis promi c ed aim tbe oiiier oaks t e quk e
ttr lies jitdL joii the iw.timf Fianmrmc, dr ssdi above the tspeffih^
cm. be pmniccti tromt. jclv vit coe miics — "f*wr tfiffioiit^ bv *S|^»>f»

T!ie !mie j i mLeat ±e suiixh-pece can be maast m
' wea k e r ' bf nrrmng ±e bamr « due ±e baie s but aighihr
seiL sni ± 5 3iit rmptrsBoie ±at ±e mcrtmr : im?tfl . :aipv befpeii
^e m aeL ves in ans ^ror mm. ±e " jgeakar ' mas jiycuc tL Tbe au^
itSQ^naciai bv A Jt :he itsazxmxiiiL of ±is assanmesic s ^s near me
^«iir at iie :nne nwr x '32K» ^ncrv jttie <afe:r oit die

Tbe ^caie jf :ms mba ' Amtai i;. ca imKB al irzth die -AuiirMnf gitLTt; . :^
i* ^eZ i 1 1^^ 4; X ^ i. i^ b»^ c. c ^w £. ^. f*^ ^^ joiL cuiminniig
It dxe :hst bacmxixn:^ x. b2l b. c\ c'^w £'. £'^. s^V f^. f*^. ^\
^'^. i\ b*. c"^. -?*'. Jtdie :sec-jiiit Ittnnumc die :iiilcn«zii^ iznes
,2at be 7TO%mi:«£ ^eat sse. f ^. ^*. ^j-'^w -t' x' ^ b \ c"^ r**^. tf*.
:1111s ^H'HK "ins Jisanmeiir 1 3tQ^ >tt iiil :xuee ^zof^s^ ami anlir
jr m nir iitn>r -nt* j cmr4 «g t ^ JtfssQmKSDis xs :u :ne ssoi^xt die 7ig«.
-votie xis^f ^ jonmni i^ :ne ^amg ne at <tt ?^»:*iis tnac duee :Dins vjiiiit
be in^nhic^i. TOut ^aidii^ une.

TUiut; bi^ sai»EStimxii|^ xdi aQiK^^t«>£ izr »r
xDStfxsanrtm^ ,snt x b«s»Anr xr ni^ime-^nRi
stsassLL rhs 3S«?i -^ttca ^ xt "^am rescue ike die ^we^aseL but
^jiBBteR&nr at ^ g. iitiutn:^. ^mc^xcaily Tie >isiie ^sate^ as Bit :Qe
rlgmegKt»>i :ne sttne N»1mflltl^&s:^S(;altt^;aabe TnMiKiftbmtduramr
aunwoic ^saxes^ ^ imeis^ — x :JKt -witc:! »»» be xuuittsftt as iuther
:;»7Utsic:t :XMc: :ne UDtnmics: <)i die iMie xce mit^ w an itgM: jf :fieiiiziK
n :ww :saifn(rv^i. ^'\xt :ne <;:)iniie«: TMutib^nect. bm^e^ie!; die :iiitt
^ -Tuaer aou iukrr aou Tie nti:*! x% Tie '««n^e^ imes 5^ nus^ -snfc-
<»m Tir wUt Tie 3atesw^j«*^«ea^ die ^•ssc ^zm^Mxtm it Tie Higaaitt it
Tie i$i^ ip«2ft Tie :e<Q^ ^^soie^ x ;\{Sf«^$t^»iim^ <«in«Daii: at Tie aiic^
A Tie ai2«e.

sr^nsdr :9!a|i«tnne^ i«r Tle»>e u^Kmoiew^s iir «ic:«»ic: Ti«e^ ^ititt t^ck^
ai|« X :a$K: TI1c^Klh«)t^ {t ibyikK- 4\ He 4Syna«^r^:)t>^ «i»i;:r "«!«»£

The AvXJ? or Tibia, 53

The scale of the instrument No. 76892, with a clarinet mouth-
piece long enough to bring it in tune with the instrument just
described (/>. 0.95 cm. longer), is as follows : fundamental tones, d, e,
f, g, g*, a, bb, c', c'», d', e'; first harmonic, a', b', c", d", d"«, e",
f", g", g"*, a", b"; second harmonic, f"«, g" , a", b", c'", c'"», d'".
It is to be observed that all of these tones correspond to tones which
can be produced on the instrument first described — a fact which can
be further illustrated by blowing on both pipes at the same time. The
scale of the second instrument is, however, less complete than that
of the first ; the tones f and g' of the fundamental series are Jacking,
as are also several of the semitones. There are two bands on this
instrument in which I could discover no hole, and one of them cor-
responds exactly with the band of the f ^ hole in the instrument first
described. The other band is provided with the mechanism for
turning, and if provided with a hole would give the note ei2. If
both these bands were provided with holes, the second instrument
would lack only the tones b, f, and g' of the first instrument.

The scales of these instruments agree so exactly in respect to the
other tones that it is at least conceivable that we have here preserved
the two pipes of a pair, of which one for some reason had a scale less
complete than the other. Gevaert (Vol. II., p. 280) expresses his
belief that these instruments from Pompeii were intended to be used
singly, and he is evidently influenced in this decision by the fact that
these pipes have so large a number of holes. In fact, however, this
is insufficient ground on which to base such a belief. As has been
shown above, the double pipe was far more frequently used than was
the single pipe, and to judge from the price paid by Ismenias,^
these instruments were very expensive, which makes it improb-
able that four single instruments should have belonged to one
man; and yet these pipes were all found together, and evidently
belonged to the same musician. The pipe in the hand of the muse,
PL I. I, was one of a pair, and the number of finger-holes actually
represented on this instrument is at least nine. Of the pipes shown
on a sarcophagus in the British Museum one has seven, the other
five, lateral tubes ; and as these tubes are near the middle of very
long instruments, and there is considerable space between the lowest

^ Lucian, adv. ind. 5.

$4 Albert A. Howard.

one of them and the end of the instnunent, it is natural to suppose
that there were several finger-holes in this space, which the sculptor
did not attempt to represent There are many other relief and
works of art, representing the doable ppe, in which a similar space
is shown, and, assoming that these representations are fidriy ccnrrect,
it is clear that the ancient musician often performed on two pipes
each of which had a considerable number of holes.

It is by no means certain that the two pipes of a pair of pares had
each the same number of finger-holes, similarly placed, and, as a
matter of fact, no two ancient instruments have ever been found
which were exactly alike ; but in the case of these two instruments
the scales are such that with shght limitations they could be used as
a pair, and it therefore seems not improbable that they were so used.

The instrument Na 76891, with a mouth-piece 4 cm. in length,
gives the following scale : fundamental, e, f^ f^, g^, a, a<, b. d', e'^
r, g'; first harmonic, b', c", c^'l, d"l, e", T, f'^ a", b", c'", d"' :
second hanmrnic, g"«, a", a"^ c'", c'"^ d'", d'"^.

There is a band on this instrument which covers the proper posi-
tion for a hole to produce the note c', but no hole is visible in this
band.

With a similar mouth-piece the instrument Na 76895 gives the
foQowing scale : fundamental, e» xe, i, g, g^, a, a^, b, c', c'^, d'^, e',
r« ; first harmonic, b', xb', c", d\ d"^ e", f, f ^ g", g"<, a"t, b'',
c'"« ; second harmonic, g"^ xg"J, a", b", c'", c'"^ d"', d'"^ c^".

There is a band on this instnmient in which no hole is visible, but
which covers the proper position for a hole to produce the note f ^.

The hole 1.27 cm. firom the k>wer end of this instrument has
only a very slight influence on the pitch of the pipe — less than half
a tone. The two instruments last described are of almost exacdy
the same length, difiering in this respect by less than 0.5 cm. The
scales of the two instruments agree with respect to a majority of
the tones which can be produced, and where there is lack of agree-
ment, the scale of one instrument supplements that of the other, so
that by the aid of both a complete chromatic scale con be produced.
There are misang, however, the holes necessary to produce the tones
from g' of the fundamental scale to b^ the lowest tone of the first
harmonic series.

It is perhaps less probable that diese two instruments formed a

The AvAxfc or Tibia. 55

pair than in the case of the two instruments first described, but even
here it is not impossible that we have preserved a pair of pipes of a
different variety from the first pair.

The scales of the instruments as here given are in terms of stand-
aid pitch, and have not been changed to conform with the theory
that the pitch of ancient times was a minor third below that of mod-
em times.*

There are two instruments in the British Museum, of which a
partial description has been given on page 16. These instruments,
which were acquired by the Museum from the Castellani collection,
consisted originally of an inner tube or lining of wood covered with
a bronze casing composed of bands similar to those found on the
instruments from Pompeii, but without the mechanism to enable the
performer to turn them. The wooden lining has almost entirely dis-
appeared, but enough of it is still preserved in one of the instruments
to give an idea of the original form and of the diameter of the bore.
The instruments are shown in Plate I., Figs. 2 and 3, and in the hole
near the middle of Fig. 2 a portion of the wooden core is still visible.

^ Unfortunately, these scales do not agree with the scales given by Gevaert
(Vol. II., p. 295) for these same instruments. The scales there given are raised a
minor third, and still the lowest tone of each instrument is said to be d, which
assumes an actual b (=:h of the German notation) of the modem pitch (cf. p. 286).

I am inclined to think that the fac-similes of the Pompeian instruments on
which the scales of Gevaert were determined are inexact, and for the following
reasons. On page 280 it is said of the Pompeian instruments: "tous ont la
wStoit forme, la m6me dimension et ne different entre eux que par le nombre de
trons dont ils sont perc6s"; and on page 286: *'toutes quatre ont la mime
longueur.** As a matter of fact, no two of these instruments are of exactly the
same length; the longest is 1} inches, or nearly 4.5 cm., longer than the shortest.

To produce the note b in a stopped pipe the vibrating column of air should be
about 66 cm. in length, and as the longest of these instruments is only 53.6 cm.
long, it would require theoretically a mouth-piece nearly 13 cm. in length in
order to produce this note, while, as Gevaert himself says (p. 292), the mouth-
pieces were very smalL The so-called fac-simile of No. 76892 in the Naples
Museum does not have the narrow band and hole near the mouth-piece which is
distinctly shown in Plate II. 2, and it has eleven holes, two in one band, while in
the original, which I examined with especial reference to this point, I failed to
discover any band with two holes.

It is for these reasons that I have ventured to disagree with Gevaert in his
as to the scales of these instruments.

56 Albert A. Howard,

The lines which mark the divisions between the bands are not very
distinct in the plate, a fiact which is due to the unsatisfiactory nature
of the photc^raph from which the reproductions were made, but on
close inspection the lines can be distinguished. One of the instru-
ments is now broken just above the hole for the mouth-piece, and
shows at the point where it is broken a portion of the wooden lining
which originally extended from end to end of the tube. The other
instrument, which is unbroken, is closed at the end above the mouth-
piece, and undoubtedly the broken instrument was originally closed
at this end. Mr. Arthur H. Smith, to whom I am indebted for the
measurements of these instruments, thinks that both of them are
broken at the end farthest from the mouth-piece ; but inasmuch as
the distance from this end to the mouth-piece varies in the two instru-
ments by only one millimeter, and as this can hardly be the result of
accident, it is probable that both instruments are practicaDy complete
at the lower end, and that the imperfect edge of the metal gave rise
to the belief that they were broken.

In the measurements which follow the dimensions are all given in
centimeters. The distance is given from the lower end of the instru-
ment to the nearest side of each hole, and the diameter of the hole
in two dimensions, the first lengthwise of the instrument, and the
second measurement at right an^es with the first.

V. Catalogts No. S4, 4-9» 5.

Instrument broken at the upper end. Length, 27.6 ; diameter at
the upper end, Ext. 1.7 x 1.62, Int. 0.8 x 0.8, this being the diameter
of the wooden lining idiich is preserved ; lower end, Ext. 1.5 x 1.5,
Int. 1.25 X 1.30 (the wooden lining has entirely disappeared at this
end). Holes : A, 1.15, diam. 0.75 x a? ; B, 4.3, diam. 0.8 x 0.78
(edge (tf the hole imperfect) ; C, 6.85, diam. 0.8 x a73 (edge of
the hole imperfect) ; D, 12.12, diam. 0.75 x 0.78 (below, part of the
wooden lining with hole, diam. 0.7 x 0.4) ; £, 19.1, diam. 0.8 xo.72
(edge of the hole imperfect) ; F, bust of Maenad raised above the
sorfiure (tf the instrument, and above the forehead a hole sbniing
toward the fewer end of the tube 24.4, diam. 0.6 x a6.

Tbere is a separate band for each of the finger-holes, and in id<ii-
tkHi a band with no hde between D and £. .\Q of the finger-holes
in this instramem are on the same side of the tube.

The k\)\o% or Tibia. 57

VI. GlTALOGUE No. 84, 4-9, 6.

Instrument closed at the upper end. Length, 29.8 ; diameter at
the upper end, Ext. 1.52 x 1.6, the internal diameter cannot be
ascertained as this end of the instrument is closed ; lower end, Ext.
1.4 X 1.5, Int. 1.3 X 1.36. The wooden Iming of this instrument
has entirely disappeared. Holes : A, i.o, diam. 0.55 x 0.55 ; B, 3.2,
diam. 0.75 x 0.8 ; C, 6.8, diam. 0.6 x 0.55 ; D, 13.75, diam. 0.78 x
0.79; E, 16.0, diam. 0.83 xo.76; F, 19.3, diam. 0.75 x 0.73; G,
24.5, hole for the reception of the mouth-piece exactly as in the
instrument last described, diam. 0.63 x 0.62.

The hole E is on the under side of the tube, and was doubtless
closed with the thumb; the other holes are all on the upper side
of the tube. Each hole has a separate band, and there are two
additional bands between C and D, but neither of them has any trace
of a finger-hole.

In my experiments to determine, if possible, the scales of these
instruments, I assumed, for the reasons stated on page 1 7, that they
were reed-instruments of the irAayuivXos-type, and that the main bore
of the tubes was originally cylindrical and 0.8 cm. in diameter, this
being the internal diameter of the wooden lining where it is preserved.
I used a very small clarinet mouth-piece, and connected it with the
instrument by a rubber tube. With such an arrangement about 10
cm. in length, the note produced when all of the holes are open is
the octave of the note produced when all are closed.

Although this mouth-piece is fully one-third as long as the instru-
ments themselves, and is double the length of the one used with the
Pompeian instruments, the results obtained are such as to justify the
assumption that mouth-pieces of this length were used on these in-
struments. The scale of fundamental tones of instrument No. VI.
is, in terms of standard pitch, a, xa*, xb, c', e', f'*, a', and each of
the last three notes can be lowered a half tone by partially closing
the hole through which it is produced. The notes a^ and b are a
very little flat The harmonics can be produced, and are those of
the stopped-pipe.

With a similar mouth-piece the instrument No. V. gives the follow-
i|ig scale of fundamental notes : a, a^, b, c', d ^, a', and by partially
dosing the proper holes d' and g ^ can be produced.

58 Albert A. Howard.

In the description of these instruments attention was called to the
presence on each of one or more bands not provided with finger-
holes. The single band on No. V. covers the proper position of a
hole to produce the sixth, and the two contiguous bands of No. VI.
cover the positions of holes to produce the fourth and the semitone
between the fourth and the fifth. Similar bands without holes are
found on the Pompeian instruments I., II., and IV., and on the in-
strument shown in Plate I. i ; but in every instance the band without
a hole is placed between two other bands, each of which is provided
with a hole.

Under these circumstances it is natural to suppose that the instru-
ment-maker, to lessen the friction and the consequent wear of the
instruments, made the movable bands with finger-holes narrow, and
that, to prevent them from slipping up and down on the tube, he
inserted between them fixed bands without holes.

This explanation does not, however, meet the case of the instru-
ment No. VI., in which two contiguous bands are left without holes,
since there is no apparent reason for inserting two fixed bands where
one would be sufficient ; and inasmuch as one of these bands covers
the position of a hole to produce the fourth, the only missing tone in
the diatonic scale, it has occurred to me that possibly this instrument
is unfinished, and that when finished one or both of these bands
would have been provided with a finger-hole. With an additional
hole for the fourth, the complete chromatic scale could be produced
on this instrument by the method of fingering suggested above.

The arrangement of the finger-holes on this instrument is such as
to imply that it is a single pipe, and not one of a pair. There are
two distinct groups of three holes each, one near the mouth-piece
and the other at the lower open end of the instrument. The space
between these two groups is so great that when either set of holes is
closed by the fingers of one hand, only with great difficulty can a
hole of the other group be covered with a finger of the same hand.

There are two other instruments in the British Museum which
were found in a tomb near Athens, and were brought to London with
the Elgin collection. These instruments are simple tubes of wood
(sycamore), open at both ends, and each tube has six finger-holes,
one of which is in each case on the under side of the tube, and was
evidently closed by the thumb of the performer. No bands or other

The kvKo^ or Tibia, 59

mechanical devices, and no traces of such arrangements, are found
on either of these instruments. The main bore of the tubes, although
not exactly cylindrical, is nearly so, and the instruments were cer-
tainly not conical pipes like the modem oboe, for in one of them it
is the end nearest the mouth-piece which has the greater internal
diameter.

The measurements of these instruments were made for me by
Mr. Smith, and were made in exactly the same manner as were those
of the two instruments last described.

VII. Elgin Instrument A.

This instrument is not perfectly straight, but has somewhat the
appearance of an elongated S. The length measured along the cur-
vature is 35.6 cm., in a straight line 35 cm. The upper end is at
present very much injured, and the sides of the pipe are pressed
together ; the internal diameter is 0.8 x 0.4. The diameter at the
lower end is, Ext. 1.7 x 1.5, Int. i.o x 0.85. The distances to the
holes are, as before, measured from the lower end of the instrument
to the nearest side. A, 4.1, diam. 0.85 x 0.75 ; B, 9.3, diam. 0.95
Xo»7S; C, 12.8, diam. 0.9 x 0.85 ; D, 16.3; diam. 0.94x0.78;
E, 19. 4 (on the under side of the tube), diam. 0.9 x 0.73 ; F. 22.8.
diam. 0.9 x 0.8.

VIII. Elgin Instrument B.

This instrument is straight, and is 31.2 cm. in length. Diameter at
the upper end, Ext. 1.45 x 1.20, Int. 1.25 x 1.03 ; at the lower end,
Ext. 1.65 X 1.45, Int. 0.9 X 0.83. Holes : A, 6.4, diam. 0.9 X 0.73 ;
B, 10.5, diam. 0.93 X0.83; C, 13.7, diam. 0.93 xo.80; D, 16.8,
diam. 0.88 X 0.77 ; E, 20.4 (on the imder side of the tube), diam.
0.88 X 0.73 ; F, 23.0, diam. 0.85 x 0.75.

These pipes are apparently reed-instruments of the same general
t3rpe as those found at Pompeii, but are of much less elaborate con-
struction. Without some arrangement for closing the holes not in
use, they cannot have been used as a pair, since it would be impossible
to close all of the holes of either instrument with the fingers of one
hand. Furthermore, the pipes are of unequal length, and although
the longer pipe is slightly curved, it has not the shape of the Phrygian

6o Albert A. Howard,

pipe as shown in works of art ; so that if it has been correctly sup-
posed that the impares were always Phrygian pipes, these instruments
cannot have formed a pair.

To determine the scales of these instruments I used a clarinet
mouth-piece,^ which, for the instrument No. VII., was made 7.4 cm.
long, because with a mouth-piece of that length the note produced,
when all of the finger-holes are open, is the octave of the note pro-
duced when all are closed.

The finger-holes of both instruments are oval and of large size, so
that from each two tones can be readily produced, one when the hole
is partially uncovered, the other when it is entirely open.^

The following scale of fundamental tones was produced, the letters
in parenthesis indicating tones produced when the holes were only
partially uncovered : g, (g«) a, (a*) b, c', (c'*) d', (d'*) e', (f) g*.

With the instrument No. VIII. I used a mouth-piece 8.2 cm. in
length, and produced the following scale of fundamental tones : a,
(b) c', (c'») d, (d'«) e', (f) f'«, (g'«) a', (a'«) b'. The harmonics
can be easily produced on both of these instruments, and are those
of the stopped pipe.

In order that the note produced with all of the holes open should
be the octave of the note when all are closed, the instrument No.
VIII. would require a mouth-piece over 12 cm. in length, and as
such a mouth-piece would make this instrument longer than No. VII.,
it seemed necessary to adopt the shorter form here described.

^ The mouth-pieces with which the scales of all eight of the instruments here
described were finally determined, were made exactly like the mouth-piece of an
ordinary clarinet, but very much smaller. The opening over which the reed was
bound was in each instance 2.3 cm. in length.

* Cf. Gevaert, Vol. II., p. 646.

THE TRAGEDY RHESUS.^
By John C. Rolfe.

ALTHOUGH the Rhesus has been handed down to us as one of
the tragedies of Euripides, its genuineness was suspected even
in antiquity, for in the first of the two arguments which we possess,
after a brief outhne of the plot, we read these words : tovto to Spofm
ivioi voOov xnrevovfcravy EvpiiriSov 8^ firf eTvcu* tov yap So^okAciov fiaXXjov
viroifMuvuv )(apaKT^pa, iv fihrroi roTs &8a(rKaXcats w yvrjaiov avayi-
ypavToiy KoX i) ir€pl ret furdpaui 8^ iv avrcp iro\\nrpayfxoavvri tov Evpi-
vi&qv o/AoXoycT.

In modem times the discussion was first revived over two centuries
ago by Joseph Scaliger, whose conclusion is : * * auctor Rhesi vetustissi-
mus, qui sine dubio non est Euripides.'

Since his time the discussion has been vigorously carried on, and
almost every one who has given his attention to the study of Euripides,
or of the Greek drama in general, has had something to say on the
question. In 1863 Frederic Hagenbach, who took the authorship of
the Rhesus as the subject of his inaugural dissertation,^ did a service
to future investigators by giving * a full list of those who had discussed
the question up to his time, together with a brief summary of their
views.

To this dissertation the reader may be referred for fuller partic-
ulars. It is enough to say here that while the majority of critics are
of the opinion that the play is not the work of Euripides, hardly any
two agree as to the author or the time in which he lived. It has

^ This paper, written in Latin, was accepted in 1885 by the classical faculty of
the ComeH University as a dissertation for the degree of Doctor of Philosophy.
While the general plan of the dissertation remains unchanged, it has been wholly
rewritten, and a number of additions and corrections have been made.

* ProUg, ad Afanilium, pp. vi. fol.
» De Rheto Tragoedia, Basle, 1863,

* pp. 6, 7, and 51-62.

61

62 John C. Rolfe,

been ascribed to Sophocles,* to the younger Euripides,' to one of the
Alexandrine Pleiad,' and to an obscure and tasteless imitator,^ about
whose period there is a wide divergence of opinion. One distin-
guished critic ^ thought that the many absurdities which he saw in the
Rhesus could best be accounted for by supposing that it was a com-
bination of a tragedy and a comedy, intended for the fourth place in
a tetralogy, in lieu of a satyric drama. This last view has found only
a single supporter.*

The opposite view, that our Rhesus was written by Euripides, was
ably sustained by Vater' and Hartung,* and is still held by some
competent critics.

Hagenbach's dissertation was reviewed favorably by Rauchenstein,'
and unfavorably by Schenkel.*® The question has since been made
the subject of special discussion by Menzer,*^ who supported Her-
mann's view ; by Albert," who thought the Rhesus a youthful work of
Euripides \ by Noldecke,'* who merely decides that the drama was
not written by any one of the three great tragedians, nor in their
time ; and by Eysert," who does not attempt to determine the author
and date of the play, but only to show that it is not abnormal in its
language, and that it is not the work of an imitator.

Views on the date and authorship of the Rhesus have also been
pronounced incidentally in histories of Greek literature and works
of various kinds dealing with the Greek drama. Such views are
more likely to be unbiased, and the present state of the controversy
may be shown, and incidentally that the question is still an open one,
by quoting three opinions of this kind.

* Gruppe, Ariadnty pp. vii-x.

* M. Anton. Oelrio, Proleg. in Semcae Tragoedias^ p. xxi.
' Hermann, Opuscuia^ III. pp. 262-310.

* Hagenbach and others.

* Dindorf, EuripideSy Ed. Oxon. pp. 560 fol.

* Spengler, De Rheso Tragoedia^ Program d. gym. z. Dfiren, 1857.

' Vindiciae. * Euripides ResHtutus,

* JahfCs Jahrb.f. Phil. 89, pp. 569-571.
M Philologus, XX. p. 484-

^ De Rheso Tragoedia, Berlin, 1867.

^ De Rhiso Tragoedia, Halle, 1876.

^* De Rkesi fabulae aetate et forma ^ Schwerin, 1877.

^* Rhesus im Lichte des Eur. Sprachgebrauches^ B5hm. Leipa, 1 89 1.

The Tragedy Rhesus. 63

Bergk in his Griechische Literaturgeschichte^ discusses the matter
at some length ; his view may be summarized as follows : —

The Rhesus is the work of an imitator of Aeschylus, who lived
after the close of the Peloponnesian war, but before the time of
Alexander the Great. He has followed his model closely in the lan-
guage and in the external details, but has missed the Aeschylean
spirit. The play does not deserve the excessively severe criticism
which has been passed on it. The choruses, especially the beautiful
one beginning with v. 527, are deserving of the highest praise.* It
must be admitted, however, that the author lacked dramatic power,
and that he has not made the most of his subject, which is well
adapted to a drama.' The play shows not the slightest sign of the
style of Euripides} It cannot be Alexandrine. The author has
been indirectly influenced by Euripides and his school. Bergk*s
view with regard to the alleged S(H^oKAciov xfipaKrrjpa of the Rhesus
had best be quoted in full : * Fon dem Geiste des Sophokles ist hier
nichts wahrzunehmeny und wenn uns auch keine von den friihesten
Tragodien des Sophokles erhalten ist, so konnen wir doch zuver-
sichtlich voraussetzen, dass sie des grossen Namens nicht imwlirdig
waren. Indes enthalt jene Bemerkung, richtig verstanden, einen
beachtenswerthen Fingerzeig. Sophokles hat in der ersten Periode
seiner dicterischen Thatigkeit sich vorzugsweise an Aeschylus an-
geschlossen und namentlich den Stil jene Meisters sich angeeignet,
jedoch in der massvoUen Weise, die jedes Werk des Sophokles kenn-
zeichnet. An diese alteren Tragodien des Sophokles mochte der
Rhesus hinsichtlich der Behandlung der Sprache erinnem ; denn nur
diesen Punkt hattenjene Kritiker im Auge,^

The other side of the question is supported by Christ in his Grie-
chische Litteraturgeschichte? He says : * Der Rhesus ist nichts anderes
als ein Jliadis carmen diductutn in actus. Die Echtheit der Tragodie
ward nach den Didaskalien schon in dem Altertum angezweifelt, in-

* Vol. III. pp. 612-619.

' Bergk thinks that this chorus may be derived from an old ' Volkilied!

' The opposite view is held by many critics. See especially Beck, Diatribes^
p. 266.

^ P. 615. 'Im Uebrigen hat der Rhesus nicht die entfemteste Aehnlichkeit
mit der Weise des Euripides . . . wovon sich nicht die geringste Spur zeigt.*

^ In Iwan MfUlers Handbuch, VoL VII. pp. 203, 204.

64 John C. Rolfe.

dem die alexandrinischcD Kunstrichter in ihr mekr den sophoklei-
schen Character finden wollten. Das kann sick nun kaum auf etwas
anderes als den Mangel an euripideischem Pathos beziehen ; denn von
der eigentUcken Kunst des Sophokles idsst sick noch weniger etwas
in der Tragodie finden, Aber dieselbe weicht so sehr von der Art der
Medea, der Troades, und aller erhaltenen Tragodien des Euripides
ab, dass sie entweder aus ein ganz anderen Kunstperiode nnseres
Dichters stammt oder tiberhaupt falschlich demselben zugeschrieben
wurde. Fiir die Unechtheit sprachen sich Valckenaer und Hermann ;
aber dass Chorlieder ^ von so kunstvollem und reichem Versbau wie
die des Rhesos sind, in der Zeit der alexandrinischen Pleias, an
welchem Hermann dachte, noch gedichtet worden seien, hat durchaus
keine Wahrscheiniichkeit. Glaubwtirdiger ist daher die Ansicht der
alten Gramatiker Krates, Dionysodorus, und Parmeniskos,' denen sich
in unserer Zeit Vater und Hartung angeschlossen haben, dass der
Rhesus ein Jugendsttick des Euripides sei. In der That hatte Eurip-
ides nach den Didaskalien, wie in der Hypothesis bezeugt ist, einen
Rhesus geschrieben, und konnte demnach hochstens nur davon die
Rede sein, dass der euripideische Rhesos durch das gleichnamige
Sttick eines anderen Tragikers verdrangt worden sel Auf die Jugend-
zeit des Euripides fUrht aber auch der politische Hintergrund der
erhaltenen Tragodie, der mit der Griindung von Amphipolis £Cm
Strymon (um 453) zusammenhangt. Der Rhesos ist also das alteste
Stuck des Euripides, und aus dem Vergleich desselben mit der Medea
kann man ermessen, welche ausserordentliche Fortschritte der
Dichter in der Darstellungen der Leidenschaft und der Erregung
tragischer EfTeckte gemacht hat.'

If beside these views we set those of v. Wilamowitz-Moellendorff,
we shall see how litde the best scholars agree about the Rhesus. In
his brilliant monograph entitled De Rhesi scholtis disputatiuncula he
sa3rs : ^ ' Constat banc tragoediam circa Demosthenis aetatem ex imi-
tatione cum Sophoclis turn Euripidis ortam esse Athenis.' In his
edition of the Herakles ^ of Euripides he adds : ' Die nachahmung
des Sophokles ist in den motiven und der stiliserung der personen
nicht minder greifbar als in der diction und namentlich der metrik.'

^ Cf. Bergk's opinion of the Choruses, p. 63 above.

^ As V. Wilamowitz says (p. 71 below) it is not certain that Parmeniscns held
this view. ' p. 12. ^ p. 41.

The Tragedy Rhesus, 65

Foreseeing that this view will be assailed, he says : ^ ' Quaesiverit quis-
piam ex caecis praecipue quorum ferax haec aetas est SophocUs
admiratoribuSf quomodo is qui tot et tam gravia vitia Rheso expro-
baverit, de Sophoclea cogitare potuerit imitatione.' His answer is
that the ^plumbeus imitator^ did not do justice to his model. He has
but a slight opinion of the play, which he characterizes as ' mediocris
ingeniifetus,^

That two centuries of discussion have led to so little result, and
that there is still an excuse for writing on the question, is due in a
great measure to the methods of the disputants. Almost without
exception they have begun with a preconceived theory of the author-
ship of the play, and have supported their theory without regard to
any other possibility. This is especially true of Valckenaer, Beck,
Hermann, Gruppe, Vater, and Hartung. These earlier disputants,
too, have argued largely on what are called aesthetic grounds ; that is
to say, they have attempted to show that the Rhesus is or is not
worthy of Euripides. How subjective and how thoroughly unsatis-
factory this kind of criticism is, especially when used to support a
preconceived view, may be judged from the results. Valckenaer,
Hermann, and others of that faction saw absolutely no merit in the
drama ; it seemed to them a pitiful piece of patchwork, made of bits
taken from Homer and the three great tragedians, put together with-
out taste or skill. To Vater and Hartung, on the contrary, it seemed
a very meritorious piece of work, and Gruppe reached the climax by
regarding the Rhesus as an early work of Sophocles, part of a trilogy
with which he won his first dramatic victory ! In supporting these
views their authors have heaped on the play the most extravagant
praise and the most unreasonable condemnation, and have thus offered
opportunities to their adversaries, without strengthening their own
cause.

The problem has also been attacked from the point of view of the
language and style ' and of the metre,' but no more satisfactory con-
clusions have been reached, mainly because no one apparently has
begun the investigation with an unbiased mind.

In this paper the non-aesthetic arguments, from the didascaliae and

^ De Rhesi Sckoiiis, p. 12. ^ Hermann, Hagenbacb, Menzer, Albert

' Spengler, Menzer, and others.

66 John C, Rolfe,

scholia, the language, the metre, and the syntax and style will be
mainly relied on, but before taking them up, it will be well to give a
brief outline of the play, especially as recent criticistn has thrown
some light on many of the disputed points.

The action of the drama goes on at night in the plain of Troy
before Hector's tent. The Chorus, composed of Trojan guards, comes
to arouse Hector. They tell him that the watchfires of the Greeks
are still burning, and that the generals are hastening to Agamemnon's
tent to take council. As a fair specimen of a good deal of the crit-
icism which is used to show that the Rhesus is unworthy of Euripides,
it may be said that the poet is censured for representing the whole
guard as going to the general's tent, instead of sending one of their
number ! Another critic says : * In Rheso Hectorem excitent vigiles
nunciantes frequentissimam congressionem ducum ad tentorium
Agamemnonis, quod inventum est ineptissime, nam si tanta sunt acie
oculorum vel tam prope a Graecorum castris stationem habent, ut
ista tam distincte cemant, vix opus est exploratorem mitti.'

Surely this is to inquire too curiously and to hold the poet too
closely to details.

When the guards make their report. Hector at once concludes that
the Greeks are intending flight, and proposes an immediate attack
on them. The Chorus endeavors to dissuade him, as does Aeneas,
who has been aroused by the confusion. Aeneas urges that it is
better to allow the army needed rest and to defer the attack until
dawn, but in the meantime to send a spy to learn the reason of the
disturbance in the Grecian camp. Hector is finally persuaded and
calls for a volunteer for this dangerous mission. Dolon offers himself,
but demands a substantial reward. Hector offers him successively
the privilege of becoming one of Priam's sons-in-law, a large amount
of gold, and one of the famous Greek generals as his slave. Dolon
refuses them all and demands as his reward the horses of Achilles, and
these Hector finally promises him.

Then Dolon explains his plan for disguising himself. He will
cover his head and his whole body with a wolfs skin, and running
on all fours will enter the Grecian camp without exciting suspicion.
When secure from observation he will walk erect. This stratagem,
which the Chorus highly commends, has been ridiculed by all who
have denied that Euripides wrote our play, and even by most of the

The Tragedy Rhesus, 67

opposite faction. And yet, as v. Wilamowitz-Moellendorf points out/
our author follows a tradition which, though different from the
Homeric account, was current in the fifth and fourth centuries
before Christ. He says : * Atqui Dolonem re vera indutum pelle
lupina rcrpairow fUfniaraarOai \vkov KfXtvOov (uti similes ihsidiae
passim de barbaris Americae incolis narrantur) vulgari per quintum
quartumque saeculum ferebatur fabula, cuius in litteris quidem nulla
praeter banc tragoediam pervenit ad nos mentio, sed vascula Attica
et antiquissima et recentia pictam luculentissime banc scaenam
exhibent.' Some illustrations which fully confirm this statement are
given by Schreiber in the Annali deW Institutiox 1875.' ^^ ™^^"
tions thirteen vases which represent the Doloneia, although Vogel*
seems to know of but one, and that too one which follows the
Homeric tradition. There can be no question, however, that in
the scene depicted by Schreiber in Tav. R. i, exactly the disguise
described by our poet is represented.

After the departure of Dolon a shepherd comes from Mt. Ida,
to announce the arrival of Rhesus, the Thracian king. Hector at
first refuses to listen to him, and insists on supposing that he has
come at this inopportune time to bring news of his fiock. When at
last the shepherd is allowed to tell his story, he gives a vivid descrip-
tion of the mighty army and wonderful horses of Rhesus. This
scene also has been severely criticised. Von Wilamowitz sees in it
an indication that the Rhesus is an imitation of Sophocles's lost
drama Pastores, the points of resemblance being that in the play of
Sophocles the scene is laid in the Trojan camp, there is a double
plot in the deaths of Protesilaus and Cycnus, and finally ' chorum
audimus de pecoribus suis multa narrantem.'

After the shepherd has told his story, the Chorus in a beautiful pas-
sage celebrates Rhesus, the son of the river Strymon and one of the
Muses. Then Rhesus appears and is met by Hector with reproaches
for his long delay. Rhesus urges as his excuse a protracted war
against the Scythians, saying that as soon as he had finished it he
had hastened to the relief of his allies. In a boastful speech he
declares that on the morrow he will finish the war which has raged

1 De Bhesi Scholiis, p. 1 1. > See etpecially Tav. R. i.

' Scenen aus Euripides in Vasengemaldent p. 21.

68 John C Rolfe,

for ten years with his own army alone. Then he will go with Hector
to conquer Greece. Hector, before so boastful, becomes modest in
comparison with this new miks glariosuSf as Vakkenaer calls him, and
informs Rhesus that to conquer Greece is not so easy a matter.
Rhesus finally consents to let the Trojans take part in the coming
battle, but insists that he be placed with his force opposite the bravest
of the Greeks.

While Hector accompanies the Thracian to the resting-place as-
signed him for the night, the guards are reheved. The approach of
dawn is described in a charming ode. Of this Patin' says: 'On
y voit les astres qui s'effacent k Thorizon, les pl^iades qui com-
mencent k paraftre, la constellation de Taigle planant au milieu du
ciel, la lune illuminant tout de sa lumi^re : on y entend les plaintes
du rossignol aux bords du SimoTs, la fiiite des pasteurs qui d^j^ m^-
nent leur troupeaux dans les paturages de Tlda. Ces souvenirs de la
nature et la vie champetre, ainsi jet^s, k la fa^on d'Hom^re, parmi
les scenes de guerre, ont beaucoup de charme.'* As the Chorus
departs it expresses anxiety about Dolon.

Meanwhile Odysseus and Diomedes have made their way into the
Trojan camp. When they appear on the scene, we learn fh)m their
conversation that they have killed Dolon and have learned from him
the pass-word and the position of Hector's tent. They propose to
slay Hector, but find him gone. As they are on the point of depart-
ing without accomplishing anything more than Dolon's death, Athena
appears, to tell them of the arrival of Rhesus, and how they may
slay him and carry off his steeds. This they proceed to do, while
Athena, in the guise of Aphrodite, reassures Paris, who is hastening
to his brother's tent. Paris, Aeneas, and Athena are really not neces-
sary to the action of the drama, and their introduction has been
severely criticised. The multiplicity of characters, however, which
Bergk^ ascribes to the influence of Euripides and his school, give
variety and liveliness to a drama somewhat lacking in action, while
the intervention of Athena, as Patin* points out, lends a fatalistic turn
to the events which follow.

1 Tragiques Grecs^ Vol. III. p. lOo.

* Valckenacr, Hennann, and critics of their school saw no such beauties in this
passage. It is fair to say, however, that it has been generally admired by all who
have not had reasons for refusing to see any merit in the Rhesus.

8 I.e. * l.c. p. 162.

The Tragedy Rhesus. 69

Meanwhile the guards are aroused, and returning surround the two
Grecian heroes, who escape by the coohiess of Odysseus and his
knowledge of the pass-word. Then the charioteer of Rhesus, severely
wounded, enters and informs the guards in a simple and dramatic
narrative that the Thracian king has been killed with many of his
followers. When Hector appears, the charioteer accuses him of
treachery and of the murder of Rhesus, saying that the Greeks could
have had no idea of the arrival of the Thracians, and of their position,
without supernatural intervention.^ Hector listens with a patience
which seems to most critics excessive, and orders that the charioteer
be taken to his house and his wounds attended to.

Then the Muse, the mother of Rhesus, appears with her son's
body in her arms,' bewailing his death in a speech which many critics
believe adds to the drama the pathos which it otherwise lacks, though
others find it only laughable.

After this Hector deplores the death of his ally in simple and noble
language, the dawn appears, and an attack on the Greeks is ordered.

One of the things which has been especially censured in the play
is the boastfiibess of Hector and Rhesus. Others have attempted
to justify it, by saying that the author wished to represent the bar-
barians as despisers of the gods, and as finally visited by their
vengeance. This view is examined at length and disposed of by
Noldecke.' Von Wilamowitz sees in this proof of imitation of Soph-
ocles ; he says : ^ ' Nihil est quod in Rheso legendo plus taedii moveat
quam tumidissima Hectoris et Rhesi gloriatio : at Sophocles consimili
artificio barbaros mores pinxerat, nee dubitamus magno poetae bene
cedere potuisse, quod in plumbeo imitatore bilem movet.'

Before passing to the non-aesthetic arguments, it may be well to
give one or two opinions to offset the flood of condemnation which
has been poured on our unlucky author, opinions which are not given
to support a theory. Patin^ finds the Rhesus beautifiil in details,
carefiiUy worked out, but lacking in dramatic power, especially in

^ Patin considera this an ingenious justification of the appearance of Athena
on the scene.

* Recalling Aurora and Memnon in the Psychostasia of Aeschylus.

* De Rhesi fabulae aetate et forma, pp. 5 foL

* De Rhesi Scholiis,^. 12.
*Lc Vol. III. pp. 178 foL

70 John C Rolfe,

pathos.^ MahafTy* compares it to Schiller's WaHenstein as a picture
of camp life. Klein ^ considers that in many of its details it is not
unworthy of Sophocles ; that the scene in which the guards gather
round Hector's tent, to intercept the fleeing Odysseus, is sui generis^
reminding one of a camp scene of Salvator Rosa.

Passing now to more tangible evidence, let us first consider what
light is thrown on the question by the arguments and the scholia. In
the first argument we are informed that the didascaliae recorded a
Rhesus of Euripides. The origin and trustworthiness of the didas-
caliae are treated by Boeckh.^ Their testimony regarding the Rhesus
has never been questioned, and it is generally admitted that Euripides
wrote a Rhesus. If our play was not the work of Euripides, it was
substituted for his Rhesus, and to that fact undoubtedly owes its
preservation. That the substitution was accidental seems clear, for
a play deliberately designed to be fraudulently assigned to Euripides
would surely not show so many departures from his usual manner.
The substitution would seem to have been made before the play
came to Alexandria. Certainly the work of one of the Alexandrine
Pleiad could not have deceived the grammarians of the time.

The authorship of the first argument is disputed. KirchofT^ main-
tains that it is all derived from a ^iKcudpxov wroOtm^, such as is
prefixed to the Medea and the Alcestis^ but Hagenbach's' view seems
more probable. He thinks that the fact that the writer of the argu-
ments cites Dicaearchus especially as authority for the statement that
there were two prologues is a proof that the other parts of the argu-
ment were not drawn from him. He says : ' Nihil aliud sumere
possumus nisi compilatorem appellasse auctores, ubi maximi ponderis
erant, non nominasse ubi vel inferioris erant notae, vel omnino eorum
nomina evanuerunt.'

The view that the w. Ivwl of the argument were less known gram-
marians, though of fine critical sense, is supported by v. Wilamowitz.'^
From an examination of the scholia he reaches the following conclu-
sion : 'Tenemus igitur extitisse editionem Rhesi iccxuMr/ACKi/v, extitisse
(rvyypofifui explicandis signis scriptum ad demonstrandum spuriam

^ It may be compared in the latter regard with the Pernans of Aeschylus.
» Hist Greek Lit, Vol. I. p. 376. * Philologus, VII. p. 536.

• Geschichte d. Dramas^ Vol. I. p. 302. • De Rheso Tragoedia^ p. 9.

* C. I. G. Vol. I. p. 350. T j)g Rkesi ScMiis, p. u.

The Tragedy Rhestis, 71

eius originem. Nihil est cur huius editionis auctorem post Chr. nat.
fuisse credamuSy cum ex ipsius Aristarchi discipulis complures com-
memoret, minorem natu neminem. Deinde alius extitit grammaticus
qui in priore commentario novum extruxit refiitatis quantum potuit
dubitationibuSy servato Euripidis nomine . . . talem editionem nul-
lam fuisse contendimus nisi unius Rhesi ; scilicet neque est Euripidis
haec tragoedia neque antiquitus eadem est credulitas.'

He sees in the argument traces of these same critics, ' quorum
alter dubitationem movet, a qua alieni fiierunt Aristophanes, Crates
(de Parmenisco dubitari potest), alter satis earn credit refeUi e studio
astrorum, quod quantas turbas excitasset e scholiis noverat et ex
didascaliarum testimonio.' The view of Hagenbach that these cvum
must have lived ' aliquot saecula post Alexandrinorum aevum ' seems
to lack evidence.

The question of the prologue is an important one, for all the playr
of Euripides, except the doubtful Iphigenia at Aulis^ have prologues.
Unfortunately the evidence is conflicting. As it is expressly stated
in the argument of Aristophanes that the guards speak the prologue,
and as the Rhesus is the shortest of all Greek tragedies, it has been
maintained that the play has come down to us in a fragmentary
state, and that the prologue has been lost. This view has been
examined by Menzer,* who shows very ingeniously that none of the
characters of the Rhesus could have spoken such a prologue. His
conclusion is : 'Nullo modo verisimile est intercidisse Rhesi prolo-
gum, Euripidis more compositum.* Perhaps, as Hagenbach sug-
gests, the line quoted from Dicaearchus may have been preserved
from the lost play of Euripides. The origin of the second prologue
seems to be correctly stated in the argument. If this view be
accepted, the Rhesus began with an anapaestic system like the Per-
sians and the Suppliants of Aeschylus. While this must be admitted
to be doubtful, the deductions of v. Wilamowitz from the scholia are
certainly deserving of consideration. In the light of his arguments
the -Kt^ rk yutrdfxrta woXvwpayfiwrvvri would seem to be a desperate
attempt to defend the view that our play is the Rhesus of Euri-
pides, like that of Crates : * Kpdrrj^ dyvoctv ^lyo-i tov EvpiwCSvjv t^
vcpc ra fUT€u>pa OttapCav, 8ia to vtov in c&ou ort rov 'Vrjaw ^SciScuTKC.

^ De R. T, pp. 45 foL * Scholiam on v. 528.

72 Jokn C. Rolf€.

This passage is the only su gges ti oo of 4 v^ t«

No light seems to be duovn od the anthotship of the Rhesus bf
the fragments of the Xp^it^eru* of Accias^ or by the mectiacral
Ckwisims Aiuns^ whose aatbor takes fiftr fines from the Rhesos.

In taking the hngaage of the Rhesos as the basis of aigoment,
critics hare made the same enor as in the aesthttiL ciitici&m of the
plaT. The indices of Hermann, Hagenbach, and Albert are evi-
dendr made to soppoct a precooceired theorr. Those wlach fbflov
hxvY been made on a didferent pbn. An attempt has been made to
cbssth' the woids and expcessiQas vsed in the Rhesos w ith o ut form-
log a&T pieriotEs optmon. and to get from die ittiiccs thas made
vtote^ner fight is pcissiMe. No great lefiance most be placed oo
tab aigametit alooe* ^y in the &rst plsce ve do oot possess all the
«oiks of the three gneat tza^ednas : aai secoodiT, tifec occnrrcnce
eneaot an onssaal vutd onbrin the Rhesss and ia Aeschyios, for
iflfestmce^ does not necessaruT prove thit the vrtter of die Rhesos
boROved thtt vocd Koa Aeschvisk TS^ sacertiaiCT of
of Gcteicssm is pot&ced cct Nr Xaack : ^ ^ Wire d&e
E^eictra icas^ cubc^ eriadKak so v^nie oe Form fcw<fcyiw ab ein
5fei5ci^ ^!e«x^E£^» .\::^:'r!ient & «& Uvcfuafect des Raesos gekend
gemcht vecvksx. tsmi icoer Vesrsoic^ dieset^e sx, bc agisg ie n ab miTi
dKS crscgidiicg*

I: lUT S^ «ai Sr «3T of pr^iisc^ Q^ic Hi^^ss^iiic^''^ im&es are not
v^scibr oxt^t : ix iia t^ Ihsc oc «vxri> vtio:^ seem ^ koa to show
vmcicca of zsx'^vu;^ «e lioil sscc x Ssv «^a:a xze »xiad afeo in
Ae$cd::5^3Sw Jmi Sie sumie^ 3$ ruie of ^ E^rpsiejix Ibc ^ikveovcz. he
g:n» jkb:^(fc^i:? xc loac^ «^k^ ^ nss bm£ cc acscanacr. Hb
xnioces Jce criDC2«>£ H .0<rt* jmi Fii^sec^.* uhi I mr?
seme cc ^ot ^jc«it *

cfiH^ ^^^»e «vc3> w^^iiT^ jce 3»x i.'tsiii ^x :» essmc ««k^ cr ft

Hi^«:^»r>.. J^csC ^ Hj^j^sn^iftc^^ jfe^^itc x^ H«rnima> I&^

.acs: XT HJtt!«!X)^«rJ:^ I iur(^ l^mii. ;i; lacocsssrv v .u£ :sev«E3L «m£

The Tragedy Rhesus. 73

dflif/iau ^ and a number of airof rpayuSov/Acm. A separate list of char-
acteristic expressions ' has been made. The other indices are new.'

I. aira^ elprffiipa.

1. ai^otf. 417.* AlbertcomparesJi/fuijAesch.Eum. 906; Soph. Aj. 674.

2. d^pXAMJ/. y^y. Cf. dfipXunroiy Aesch. Eum. 956.

3. ivOpumo&ufifav. 971. In this sense.

4. SCPcLfioi. 215.

5. i&tvyji^. 304- Albert compares xpv<7avyij«, Soph. Oed. Col. 685.

6. $oivan^pu)v. 515. Cf. ^otmTiJp, Aesch. Ag. 1503.

7. KaxoyofipfHK. 260.

8. icoAAiyc^poc. 349. Albert compares KaAAiir/o^po9y Aesch. Ag. 245.

9. Ko/Mncm^. 817. Cf. KOfiayum^py Aesch. Eum. 186.

10. KopfmnroLo^. 964.

11. iccpd8cro9. 33. Albert compares xP^aoScToq, Soph. El. 837, and

*;(aXxo3ero9, Aesch. Sept. 146.

12. fn/vas. 534.

13. wicrCPpofJUK, 552.

14. oivfneXavtjfroi. 363.

15. o^ctXcrif. 965. Cf. 6i^i\mi^j Soph. Aj. 590.

16. iroKi/icpcvcii. 361.

17. xovXwtviTS. 716.

18. vpcnr a rrji* 361.

19. v/MxmvXciog. 273.

20. vporaivi. 523.

21. irpov^tpcwrfTi^. 296. Cf. irpov^pewao), Eur. Phoe. 92.

22. paxiivTOi. 712. pajcoSvnji occurs in Chrysostom.

23. raxvjSaras. 134. A. compares dfipopdrtf^y Aesch. Pers. 1073.

24. rtrpdpjoipoq. 4.

25. rcvxo^opos. 3.

26. ^AXooT/xkyros. 8. Cf <^vAAo(rrpa)Sy [Theocr.] Epigr. III.

27. xpvo-dj9«aXo9* 921. A. compares KoXXiPwXo^, Eur. Or. 1382.

28. xpvarorevx^' 340. A. compares XP^^^^^^4^^Tf^^ Aesch. Ag. 300.

29. ilfa^KLpoxpotK' 716.

^ Eysert adds eight dra^ tlfnifUwa to Hagenbach's list. It is fair to say that all
these appeared in my original list of 1885. I am somewhat indebted to E3rsert's
Talnable monograph in the revision of lists I. and II., bat as I was unable to
consult it until my paper was in type, I have been unable to give it the con-
sideration it deserves. It is a very important contribution to the subject.

» III. • IV.-VIII.

* The following editions have been used: Wecklein*s Aeschyitis^ Berlin. 1885;
Campbell's Sophocles^ Oxford, 1871 ; Nauck's Euripides^ ed. 3, Leipzig, 1870.

74 John C. Rolfe.

II. JTra^ rpayipSovfieva.^

1 . d^ovXox. 761. The adj. occurs frequently in Sophocles and Euripides.

2. drfiovk' 550.

3. aJ$09. 990.

4. dic(K3wo)9. 588. The adj. occurs Eur. Med. 248.

5. &Kfjudi(ov. Used of a person. 795. dicfia{(tf occurs in Aesch. Sept. 95,

and Choe. 722.

6. dvaveipfo, 514.

7. airXi/icros. 814.

8. diro(y(Ufuu. 177, 466.

9. dpurroroKO?* 909.

0. dcnrourros. 348.

1. dvrqppi{os. 288.

2. d<^tnrvii{[<o. 25.

3. yairw€w, 75. yaTrovo?) Eur. Supp. 420.

4. ytpovaia, 401, 936.

5. yccopycctf. 176.

6. Bix/Sai, 525.

7. SiJ3o\o5. 374.

8. SioTrras. 234. In this sense.

9. So^a. 780. With the meaning < vision.'

20. Soprj ( = Sopara) 274.

21. Svo/iiot. 529.

22. eycpri. 524. Occurs with a different meaning Soph. Antig. 413.

23. cSpaibs. 783. In the sense of * furnishing a seat.'

24. ciO'Spofii;. 604.

25. ^KKcavres. 97. Kcarres, Aesch. Ag. 840.

26. cKTvircw. 308.

27. iKTp(mrj. 881. In this sense.

28. cAcv^cpi09. 358.

29. cvToura-o). 492.^

30. i$afru)$€(o> 8 1 1.

31. ^fciKm;s. 322.

32. €7ri3c^i09> 364.

33. iwiOpuxTKio. 100.

^ tfypi/TMf, which Eysert gives as dva^ rpaytfidodfuwow, is found in Aesch. Pro.
374. Eysert rejects d/3ot;\wf because Euripides uses the adjective, but retains
dKip96pbft where the conditions are the same. He also diminishes the list by
rejecting all *//om. J^eminiscenun,* such as fUpfupos, S^x^cu, etc.

> So the MSS. and Liddell and Scott. Nauck adopts Musgrave's conjecture,
drrapai.

Tke Tragedy Rhesus, 75

34. Iirifiofi^. 327. Occurs in Aesch. Choe. 817, and elsewhere in a

different sense.

35. iwixpdofioi. 942.

36. cvSofco). 496.

37. tvawXayxyici. 192. A. compares OpamtairXdxyniif Aesch. Pro. 756.

38. cvcrra^co). 317.

39. KoparofUK* 606.^

40. KaTOKOwn. 553.

41. Kardyrrii, 318.

42. KtifiT^Xujv, 654. Used of a person.^

43. icXonrucos. 205, 512.

44. KpvaTaXXoTTrfKTOi* 441. KpwmXXcnnK, Aesch. Pers. 504.

45. KvPtwo. 446.

46. XtiOffMfXK. 881.

47. fLcAoirocas. 550.

48. fuX/^Sia. 923.

49. fUfiPXuMciaq, 629.

50. fUpfupoq, 509.

5 1 . fjufioi. 256. Used by Aesch. in a different sense.

52. fLvyBuryuoi' 789.

53. yav#cXi7piai'. 233.

54. yawrroBiiOv. 136, 244, 448, 582, 59 1 , 602, 673.

55. wKTTjyopla.* 19. Cf. wKTriyop€2v, Rh. 89, and wKTriyop€ur0aij Aesch.

Sept. 29.

56. vapoKoipfK. 830.

57. w€Xnurrq9, 311.

58. TrpotcaBfffioJL 6.

59. trAoony^. 303. In the sense of 'yoke.* Used in a different sense

by Aesch. Choe. 289.

60. woXiapxK' 382.

61. vpTinTS. 797.

62. «p<^8cro¥. 307.

63. vpocrAof. 210.

64. vrmxudi' 503.

65. irv/KTcC (heterogeneous plu.). 97.

66. ^fiff* 64.

1 See Eysert, Lc. p. 18.

* Sa Liddell and Scott {Greek- English Lexicon t.v.) and others. This inter-
pretation does not leem certain.

* Schirlitz {De Euripide navi sermonis ^onditore^ Halis, Sax. 1864) considers
this Aral 9i^fa4wo9 in this sense.

76 John C. Rolfe.

67. o^/M. 688. In the sense of ' watchword/

68. oi^fAtpw, 683.

69. mnma. 245.^

70. <nr€paf. 701.

71. awaOpoiOa. 613.

72. awtfattfurpnifAA, 489.

73. roXvrcvt*. 744. Cf. iKToXmrnm, Aesch. Ag. 1017.

74. rpoficpof* 36. In this sense.

75. ^^Spoct&l}«. 353.

76. U^lu^ 730«

77. ^oi-oTos. 355.

7$. ^orroXtf. 158. ^ovoXi^, Aesch. Sept. i6a
79. x^Mrnk. 414. Cf. tqAvx«mttos« Aesch. Choe- 35a

Whjlt some of the words in these two lists, especially in the latter,
ought not to be considered of great weight as evidence, the number
of rare words which occxir in the Rhesus has ahravs been deemed
significant* and a weighty argument against the view that the play
was written by Euripides. Those who support that view have at-
tempted to diminish the number or to show strong resembbnces in
dorm and compo^tion to words whkh are found in the other tragic
writers^ Albert * showed that very similar words occur in Aeschylus,
Sophocks» and Euripides^ but without greathr strengtbenii^ bis case.
Vater mjide^ an examination of the B^icckae of Euripides^ with the
view of showing that it did not di£&r from the Rhesos in dot regard,
but couki ficfed in a phy of 1370 lines (as against 996 in the Rhesos)
only it^^f Jvaf t^fnffutMm^ socne of which were at least doubtfoL The
Ixr^ really scientiSc examination of the questiOQ has been made
recently ^ by Eysert» wboise results are most interesting; and in viev
of the weight which his ahrays been bid on the nxxmber of rare words
in the Rhesus^ almost starthzxg. As the result of an extesscve lexico-
graphical examinatioa of aH the tragedies of Eunpiiies*. he in;^ that
while dlie Rhesus,, accocding to his tabie. has 1^^ «ra^ lyyinm, it
contains only i/mr-^g^UiKn^ of the 4^ Xjts^ mifnunmm to be fbtrad in
ail the plays of Eun^Hides. The Rhesus bte tiois to be euceeded in

The Tragedy Rhesus. yy

that respect by the Phoenissae (40), the/?« (36), the Iph, Tau. (35),
the /ph. AuL (32), and the Bacchae (31).* In the percentage of
airo^ fifirj/UinL to the number of verses in the plays the Rhesus stands
higher, though it is surpassed by the Cyclops and approached by some
of the tragedies. His table is as follows : Cyclops, 3.10 % ; Rhesus,
2.81 ; Iph. T., 2.34; Phoen., 2.27; Bacch., 2.23; Ion, 2.22; Here.
F., 1.97 ; Iph. A., 1.96 ; Hiket., 1.86 ; Elect., 1.84 ; Hel., 1.59 j Troad.,
1.35 ; Hipp., 1.29 ; Orest, 1.28 ; Hec, 1.23 ; Andr., 0.79 ; Heraclid.,
0.76 ; Med., 0.63 ; Ale, 0.60.

In view of these results the argument from the number of rare
words in the Rhesus is certainly deprived of much of its weight. On
the other hand an examination of the above table of percentages
makes it difficult to suppose that the Rhesus was the earliest work of
Euripides, and most of the upholders of the Euripidean authorship of
the play have admitted their position to be untenable without that
hypothesis. Eysert does not take sides on the question of the author-
ship of the play, although he seems to lean toward the view that it
was written by Euripides. His aim is to show that in the number of
its rare words the Rhesus does not differ greatly from the other plays,
and in this he has certainly succeeded. Even the number of airaj
rpayi^fioviuya in the Rhesus cannot be said to be very significant
against the 103 in the Bacchae.

Eysert's list of * Voces Euripideae Kar iioxriv * is also interesting, and
justifies his claim that the author of the Rhesus was no servile imita-
tor. Whether the presence of eight such words in the Rhesus, out of
a list of seventy-four, indicates, as he believes, the Euripidean char-
acter of the play may well be doubted. It would seem to indicate
that the author of the Rhesus was familiar with the writings of Eurip-
ides, but it is not of sufficient weight to offset the arguments against
the Euripidean authorship of the play.'

The next list contains some expressions which seem to Hagenbach
and Menzer to be pecuUar to the Rhesus. Those in division a, it
seems to me, can be paralleled in the other tragic writers, and I have
subjoined some expressions which appear similar. Those in division

1 It will be remembered that Vater fonnd only 8 dr. tip. in the Bacchae; Eysert
finds 31, and 103 dsra| rpayffdodfitpa !

* See es{>ecially Wecklein's review of Eysert's monograph quoted below, p. 82.

78 Jokn C. Rolfe.

b I have been unable to paraDd. To the latter list I have added a
few cases. Some of the expressions given by Hagenbach have been
transferred to other lists.

IIL EXPRESSK>NS SAID TO BE PECULIAR TO THE RhESUS.^

a.

54. a2p€a$at ^uyi/K. Cd cupomu ^nrfipty Aesch. Pers. 4S4.
173* frrparriKaL-nfi wtSht. Cf. arpaTifkamip vmr, Aesch. £um. 640.
318. €fiwu Karamfi (vfi^opa wpoi r&yaBd. Cf. wpo^ to¥ c^oi^ 6 ^Awos

€fiwu. Soph. Aj. 157.
388. deo$ icaravm <rc Cf. $€6&€w KarmrwtCa waBm fioXatiyy Aesch. Ag. 107.
512. i^€i¥ KXmwtKoi cSjpos. Cf. rums iro6^ cSjpas ridurfie fUH Aoa{crc; Soph.

O. T. 2.
566. ^[fw^ ardiu &* «rrMr.* Cf. errata ^ h ff inrvy vpo xopSuis Topog,

Aesch. Ag. 189.
875. di o-c rciFcnu. Cf. cts ri^rSc rcu^ ^0¥Wy £or. Hec 263.

30. a^yiW f ^opoc.

3 1 . yvfjivrjrnify iiavofiXpL.
III. vuicroc ^ Karcurraxru,

184. ipCnrri y 6yTtpi% unr<oy c/uum.

194. (ipurrcvcii with the genitive.

210. P6jmv X^fifTl vpocrOiav KiAxpfuxrac.

217. "EpfJt^ ifni\rjTSiy aya(.

254. ircSocm/?^ o'^ycus.

395. ScirAovs dn/p.

487. viXrqv ip€umi.

538. rif iKrjpv)(Orf with the accusative.

560. ctcnrato) Xdxov*

568. xAcC^ci (Ti^pov.

612. v6$€y r croicrai Pappdpov orpartvfiaTfK :

787. dfiwcDv ^pas i(€y€ipofuu wtaXounv.

929. iriTyauu KOpax.

932. dAxcu (^lAapparM.

While some of the expressions in 3 must not be pressed, it seems
fair to conclude from the lists already presented that the author of

1 By ' peculiar to the Rhesus ' is meant an expression which is a dwa^ rpaytpBo^-
fieyo¥ (see p. 72).

^ Hagenbach's comment on this expression is ' inaudite dictum.*

The Tragedy Rhesus,

79

the Rhesus aimed at originality in his diction, and that he was not a
servile imitator of the three great tragedians, or of any one of them.
So far as he was influenced by others, his model appears to have been
Aeschylus.

IV. Words found only in the Rhesus and in Aeschylus.*

1. dydg. 29.

2. ety^nnryos. 3, 825.

3. d&ifiavro?. 697.

4. dvrcpaoi. 1 84.

5. SucraXios. 247.

6. SiNTOi^ai. 724, 805.

7. lrcio$. 435.

8. la. 553.

9. cmnjAan^s. 117.

10. icaipuus. 339.

11. jcoro^. 828.

12. jcftfSctfv. 308. Used in a differ-

ent sense and in the fem. in
Soph. Aj. 17.

13. fidfiayva. 817.

14. fwvapxp^. 31.

15. WKTrfYop€ia* 89.

16. ^Ipyavoy,ofa musical instrument.

922.
17* iroXtaxoS' 82I.

18. irpewovTioi. 202.

19. irpoiriW. 405.

20. ^fyiCQ).* 785.

21. <rvpirf¥. 58.

V. Words found only in Aeschylus, Sophocles, and the Rhesus.'

1. al$w. 122. As adj.

2. dfMaxpii. 456.

3. avXfiiv. 112.

4. hOvfpoi. 289.

5. ii^opoi. 30.

6. Opaxrata, 863.

7. wavSuciOi, 720.

8

9
o

I

2

3

4

wp6(njfjua, 390.
^woivcitf. 172.
^tXi/tits. 217.
tf>pvKTwp{a. 55, 128.
<^vTaXfiio9, 920.
XOLfMJfvvrf. 8, 852.
Xvoi;- 118.

VI. Words found only in Aeschylus, Euripides, and the Rhesus.'

1. dyicoXi;. 948. 7. A^aata. 792.

2. afii;ir. 480. 8. 6mip^, 143.
3< dfLinfffUi¥€w. 647. 9. diroiHroi. 467.

4. AmufiOKToq. 222. lo. dpCi<^ro$. 124.

5. divuno?- 828. II. dpjcovrrtt?. 499.

6. dvoicrftiip. 516. 12. aiSimji. 873.

1 It has been thought best to make lists IV.-VIII. complete, rather than to
select those words which seem especially significant. The meaning of course is
that the words are found in no tragedy except the Rhesus, and the works and
fragments of Aeschylus.

* Menzer says (p. 33) that ^jcw used of animals is a Ar. r^

> See note to list IV.

8o

John C. Rolfe,

13
14

15
16

iS
19

20

31

23

H

25
36

27
3$

3<5

3«
32

53
34

35
36

37
3S
39

avravc^tof. 944. 40.

dvrc<i». 668. 41.

ax\iia, 452. 42.

P^y, 719. 43.

Puuvi* 901. 44.

jSoXoc. 730. 45.
yopY«Mroc. 7. yo|fry<tf>irt99 Soph. 46.

Aj. 450- 47-

Scny. 928. 48.

Sunro9* 741* 49.

fiox^uuo^. 372. 50.

SvcrcypcTOC. 212. 51.

fii«OTvx^* 961. S2-

iKwyi$^. 472. 53.

hn^$ti¥Oi. 334. 54.

^i||iiO«. 278. 55.

cvKXcik. 758. 56.

€V^^X^, 51a 57.

C\7l)4<yKK. 303- 58-

$€a^ 676, 687, 784. 59.

^l^KTO^, 669. 60.

<^t^. 57. 61.

MifiOi^ 492* 62.

^tJ^^^ 23s* 63

KtSoSit^ 643. 6$.

Kftf>ft#cf«c« 60$. 66.

KATMn^m^ 3^8. 67.

icaT07m;s. 1 34, 1 50, 155, 558,632.
iccAaSscD. 385.
jcvicvos. 618.
XoKTiiia. 411.
Xcmif. 287, 921.

Xvirpo9, 759> ^3-
/jueXdyxifUK* 962.
IMOvovfUvo^. 871.
fivpia^. 913.
voi&)A.crfti{p. 549.
voytyrXog. 22.
wunmXcvw. 180.
vc&xm^7$. 254, 763.
TcXa^. 557.*
xeXarog. 430.
xi/yaitK. 929.
vXcitM. 834.
Tpcv/HicnfC. 646.
vpoffoioafa*. 696.
Tv/DyM*. 122.
OTpan|Xara0. 276.
armlpofMoi. 495.

438.
656.
rpoam^' 40.
r^MiiyML. 751.

i/^aXaor. 27.

VIL \V<>RDS fXX'ND C»CLY IK SOTHOCLKS AXD IK THE RhESUS.'

I. ^y^j^M^. 503,715. to,

3. ^y^rc^«(i'. 436. 11.

3. XnW{«. 925- 95* • »2.

4. &4vn^Arv«». 7S1. &^yM^n^

Acsch. Eutn. 156. 13.

5. €t/rfmi^' ^60. 14.

6. ivmtyim, 715. I5.

7. <VjS<>vXy)^. 105. c{^/3«^X«^ t6.

Ae?»c>». Oi*>e. 693. 17.

S. ^^i$ti^. 77^- ^^'

447-
pea^rccoXM^ccy^BU 518^

wrrvmiC^ 61 1« 614. In Eur.

in a ^Sfierent sense.
Kponjpo. 499*

oX<»^iiyMi|pMifc. S96.
fww. 784^

^ ^ tk« MSS.; NmcIc tvudft i-)UM».

• See wAe to Van IV.

The Tragedy Rhesus,

8i

VIII. Words found only in Euripides and the Rhesus.^

I.

2.

3-
4.

5-

6.

7-
8.

9-

10.
II.

12.

14.

16.

17.
18.

19.
20.

21.
22.

23-
24.

25-
26.

27.
28.

29.
30-

31-

ayofuu. 244. 32

oiviy/ioc* 754* cuviyfuiinAesch. 33

and Soph. 34

dicXcttf. 752, 761. 35

OfLVOTtf . 419, 438. 36

av^iycvi/s* 895. avroycvi/Sy 37

Aesch. Supp. 8. 38

PaXm. 356. 39

fiovfipofUw. 333, 412. 40

yv/injs. 313. 41

yi^. 515. yvircasy Aesch. Supp. 42

804. 43

ScfUtf. 232. 44
SiairpcmTS. 617.

Suwen^, 43. 45

Si^pcvo). 356. 46

SvflrAnTO'lcci). 79I . 47

i(taofULi, 872. 48

imiapifo. 441 . 49

hnKOvpim. 937, 956. 50

^nrairopos. 529. Compounds of 51

this kind are common in Aesch. 52

e.g. IvTdirvXoi, Sept. 150. 53

cvacAfio$. 97. 54

€vSo(ta. 760. Cf. cvSofof, 55

Aesch. Choe. 302. 56

cvv£p09. 927.

Idxpvaoi. 370,439. Such com- 57

pounds are common in Aesch. 58

e.g. Idrrvpo^y Pro. 1 1 18. 59

KoBapfAO^w. 767. 60

KoOapCu. 35.

KapaSoKJu). 144. 61

Koparofiifa, 586. 62

Marotricoirog. 125, 505, 592. 63

KXtmlf. 777. 64

KOtytayCa, 904. icocvo>vo$, Aesch. 65

Ag. 1 02 1, and elsewhere. 66

jcopvovQ). 933. 67
KWfffmfi. 325.

XaCvfuu. 877.

/uXySos. 35>>393-

vavaiiropos. 48.

w^cva). 520.

ii^njptf:. 713.

6py6s. 282.

opurpa. 437.

TituSoirouk* 980.

irofiirav. 855.

ire&upo). 372.

irc\Ti7 (=ireXTa<rT<u). 410.

irXrfppjeXfj^. 858.

iro\v^ovo9* 62y 465. Cf. TToXv-

ff>^opo9y Aesch. Pro. 660.
iroXvxop&x. 548.

TTOpTFtipa. 442.

irpiaPevpa. 936.
/kuVco. 73.

traBpoi (Xoyos). 639.
aicoXoircs. 116.
(TKvXevpa. 593.
OTiX/3<i). 618.
avpirvpwa. 960.
(TvyicaToo'icaTrTa). 39I .
avvOripa. 572, 684.

T€Tpa7rOUS. 255. Cf. TtTpaXTKt'

kq^y Aesch. Pro. 411.
ro(riprfi. 226.
rpiPiov. 675.
xmdpyvpoi. 970.
viTcunrwmy?. v^i. vmunrumfpy

Aesch. Supp. 188.
^aKou. 943.
^opcrpa. 979.
if^povpoq. 506.
<l>wrrfpa. 440.
Xa<r/ia (ft^pck), 209.

X<VMW- 771-

Xpv(roicoXA7ro$. 305. Cf.x/wxr©-

KoXAof, Soph. Fr. 68.

> See note to list IV.

82 John C. Rolfe.

It may not be amiss to collect the above results in tabular fonn.

I.

11.

IIL

IV,

V.

VI.

VII.

vni.

Total.

1K1

■m

Wonb found in Acschjlns

«

•4

67

Wordi found in Sophoclei

M

la

3>

Words found in Euripides

e?

67

'34

Words DO[ found in AHchylus. . .

»9

79

la

67

'93

Wordi not found in Sophocle*. . .

"9

79

21

67

67

=63

Words not found in Euripides, . . .

=9

79

41

M

iS

161

Words found only in Aeschylus. .

ai

ai

Words found only in Sophocles. .

18

IS

Wordi found only in Euripides, .

67

t?

In considering this table it should be bome in mind that we
possess more plays of Euripides than of Aeschylus and Sophocles
tc^ether. Remembering this, we are justified in saying that the lan-
guage of the Rhesus is Aeschylean rather than Euripidean, while the
resemblance to the language of Sophocles is slight. There is certainly
00 servile imitation of any one of the three.

But as Wecldein says,' ' Hier muss weniger gezahlt als gewogen
werden.' The following words, forms, and expressions seem to him
to be strong evidence that our Rhesus is not the work of Eurip-
ides; —

Ai^ltmnAufuat'. 97I-
tcimitAu (If rt. 875.
vponun. 523.
S^. 274-
8^(ftu. 525.

vtm tv 'ApytuMi paKti^. 150, 155, 321, 589.
^(Tw vavi iw' 'Apytimi ro&i. 203 .

artixm: 86, 138, 201, 2gl, 296, 299, 582, S94t 628, 992, 993.
iroWal fiir hru^, TroXAa TrtXraarwy WXi^ voUw 8^ ttrpaxruv rofonu,
ToXvt S ix^^ yv/iviit. 311 fol.

1 £trl. JFHU. IVoch., 19 Dec. 1891, b >

» of Eyserl'* monograph.

TIte Tragedy Rhesus. 83

Of these the use of the word ireXToorai is, perhaps, the most sig-
nificant, carrying us down as it apparently does to the reorganization
of the Athenian army by Iphicrates in 391 B.C. The word, which first
occurs in Thucydides II. 29, does not seem to have been formed
before the time of the Peloponnesian war, when the Athenians had
Thracian allies. Its use in the Rhesus seems to show that the play
was not written before the time of the Peloponnesian war, but not
necessarily, as Menzer claims,^ * multo post.* Hence it would seem
that the Rhesus was not written by Euripides in his youth, and there-
fore, that it was not written by Euripides at all.

Before leaving the consideration of the language of the Rhesus
it will be well to scrutinize carefully the lists of expressions which
seemed to Hagenbach to show imitation of Sophocles and of Eurip-
ides, and to see whether they will bear investigation. Finally, as
the result of the examination so far seems to suggest imitation of
Aeschylus, a list of expressions which seem to have been derived
from Aeschylus will be given.

IX. Expressions which seem to Hagenbach to suggest iMrrATiON

OF Sophocles.'

55. cmiVet /t* Ivwyyi <lipvKT<Dpia. Cf. 7rcuSo9 /u cmivci f^oyyos, Soph. Ant.

1 214, but also axuvofim S* vtt* i\mSoi, Aesch. Choe. 193. Note

the third example in list X.
82. iv rpoirj ^po^. Cf. Iv rpourg ^p6^, Soph. Aj. 1275, but also iv

y^Xn^ Tpoir^, Aesch. Ag. 1236.
145. irpoafju((u with the dative. Same construction Soph. Ph. 106, but

also Eur. Fr. 903.
158. hnmrufJLO^* pkv Kopra. Cf. opdla^ S* 'O^aacvs €lfi circowfux Kojcoisy

Soph. Fr. 877, but also Kopra S* &v iirtawfw^, Aesch. Eum^. 90, and

iirunrvfuf 8c Kopra IIoAvveticci Xcyo), Sept. 645.
183. ^X^ irpoPaXXovr Iv Kvfioun Saupovoi. Cf. dci yap cS mirrovcriv ot

Atof Kvpoi, Soph. Fr. 809, but also Hpyov iv kv^oi^ "^PV^ Kpivti,

Aesch. Sept. 401 ^^ and Rh. 446, ptirrci? tcvPfwav rov irpo^ 'Apyaous

"Xprp^.

> De R. T,, p. 53.

' In aU cases the parallel first given is Hagenbach*s, though in some cases I have
quoted more at length than he. Those which follow are my own, though it is more
than likely that some, if not aU, of the cases have been noticed by others. When
no parallel hut Hagenbach's is given, I have been unable to find a similar expression.

' Hagenbach considers iirdinfftot a Sophoclean word !

* This may well have been a proverbial expression.

84 John C. Rolfe.

329. ApKovfiey o2 aioiaims "IXujv ttoXm, Cf. dpxiata ^vrfirKawr iyio, Soph.

Ant 547, but also <Jipicovfiev i^/xets oi irpoOvi^K€vr€qy Eur. Ale. 383.^
389. iraAoi^ 'it^^P^ ^^' ^roAoi^ o^^'^pq^ Soph. Aj. 624.^
476. ^ KofiTa. Cf. 9 Kofyroy Soph. £1. 312. A frequent expression in

Aeschylus. See Ag. 597, 1251 ; Choe. 928 ; and elsewhere.
690. Poqv iytpriov, Cf. Bp^voy iyuptrtj Soph. O. C. 1778. For a meta-
phorical use of iyuptOy cf. rfyupty aXXip cicSox^ iro/iirov Trvposy

Aesch. Ag. 311.
732. (TVfJLf^opa PcLpeloL. Cf. Papuaa^ (vfiffnifidy, Soph. Tr. 746^ but also

Pa4>€ui y a&€ avfiifH>pdf Aesch. Pers. 1045.
819. Tov*^fcropa to fJLtjStv c?ku vofu{crc. Cf. v/ia? ro firjSkvwTa^y Soph.

Aj. 1275.
883. Tpouiy dydyu vdXxv ct$ vtvOoi ^aifiwv, Cf. tas ^fitpa icAtVa re jcdyaya

voXiv airayra rdy^ponrcuxy Soph. Aj. 131.
892. iraiiSa tokS^ opSta oucrpStq &ay6yff \nr i)($piav* Cf. Soph. £1. 102, ao/vy

mrtpy ovTVi (uxok oucrpSt^ re OaycvrxK.
965-6. ^^iXcri? 8c fwi Tois *OpifU<aq ripJutau ^cuVco^ou ^iXovs. Cf. ov

KoTour^t iyfa ^oif ck ov8cv ^kciv ci/a* o^iXctits crt ; Soph. Aj. 589.

In those cases in which an expression is paralleled in both Aeschy-
lus and Sophocles, there is certainly no reason for considering that it
shows imitation of Sophocles. It is certainly fairer either to give the
earlier poet the credit, or to consider the expression a literary com-
monplace, which any writer might use without making himself liable
to the charge of imitation. The remaining cases, even if we accept
such far-fetched comparisons as the one on v. 883, are surely not
sufficient to show systematic imitation of Sophocles, much less to
justify Hagenbach's charge ' nostrum pariter ex omnibus fabulis hau-
sisse quae opusculo suo intexeret.* The conclusion may &irly be
drawn from this list and from the word lists, that the Rhesus shows
little resemblance in language and style to the extant works of Soph-
ocles.

^ This expression from the AUesHsy with Rh. 329 as parallel, is quoted by Hagen-
bach among the expressions which seem to him to show imitation of Euripides,

^ In this case the parallel is only apparent* for though the words are identical,
they are used in very difi^erent senses.

The Tragedy Rhesus. 85

X. List of expressions which seem to Hagenbach to suggest

IMITATION OF EURIPIDES.^

6. ofiBov KC^aAi/v. Cf. SpOwaw Kopoj Eur. Herac. 635, SpOovrt Kopa,

Eur. Hipp. 19$, opOov vpoa-amov, Eur. Ale. 388.

7. kwroy p\€<f>dpwv yopynnrov c3pav. Cf. (TTvyvrfv dtfipvv Xixrocra, Eur.

Hipp. 290.
55. atuvu fi Ivwxof ifipvtcTfopCaL,^ Cf. ov yap fu atuvti OtafffaTo, Eur.

Ion, 685.
59. ^acwoi -SjXiov \afi7rTrjp€9' Cf. <f>a€wa2q ijXtov v€pi'trrvxo.hy Eur.

Ion. 1 5 17.
80. irdvT Slv 4^Prfiu^ ladi, 8ct/uuuVci)v roSc. Cf. Eur. Hipp. 519, iravr &v

ffioPrfitlu l(t6i • ScifuuVei? 8^ rl; ' but also iravra hiipaivtiv f^iXci,

Aesch. Pers. 603.

84. QTrXovs iir ixOpoi^ fivBo^ airAt^av X^P^* ^^* dirXoOs 6 fivdoi lij^

6Xrj$€uiq i<f)Vy Eur. Phoe. 469, but also airXoOs 6 fiv$09 ' n^vSc fjuty
<Tru\tiv iaiOy Aesch. Choe. 552.

85. fuiXa OTTOvSg noSoi. Cf. koI firp^ 'OSvacrcvs lp\trai (nrovSjg ttoSoSs

Eur. Hec. 216.

90. irvKoiov T€vx€(Tiv BifjuoL^ fTiOcv* Cf. Kovfjuff irvjco^ov r<p3e, Eur. Herac.

725, and 9rvica{c jcpar' ifxov viKrj<f>6poVf Eur. Tro. 353.

91 . TL 8* loTi ; fuav Tis iroAcfuW dlyycAAcrcu \6)(oq ; Cf. ti 8* lort row

irapovTOii iK7rX^a-(Toy Aoyov ; Eur. I. T. 240, and ti 8* loriv, 'I<^yc-

veca, icoivov ^v SofUKs, Eur. I. T. 11 60.
105. tiff fyrff &yrjp tufiovXtK, m Spaccu x^pt* Cf. dff ^<r&a Swaro^ Spav

oaw irpoOvfUK <?> Eur. Herac. 731.
122. vtirvpyuyran OpaxTti* Cf. o$ ireirvpyoxrcu BpaxTti^ Eur. Orest. 1568, but

also x5 pkv T^S lirvpyc/vTO oroA.^, Aesch. Pers. 195.
144. <TaXiriyy<K auSiJi'. Cf. <raA.7riyyo9 ^ixn^i 'Emv. Tro. 1267, but also Pmp^

aaXmyyoi, Aesch. Sept. 381.
154. TTpo ytuaq rovSe kiv^vvov fti^^. Cf. iciVSvvov i^ Snjufixavtav plwrovrt^,

Eur. Herac. 148, and kivSvvok /uyav piirrovrt^, Eur. Fr. 406. The

metaphor is the same as in ipyw iv kv/3oc$ 'Apvi^ Kpivtl, Aesch.

Sept. 401 . See note to this expression in list IX.
168. yofuiv iK.^ Same expression in Eur. Andr. 1279. ^^^ ^' ^^ third

example in list XI.

> See note to list IX

3 Note that Hagenbach gives this expression with a Sophodean parallel in list
IX. It cannot be evidence of imitation of both Sophocles and Euripides.

' Lacbmann suggests that this is a proverbial expression; it seems highly
probable.

* There is no significance in such a parallel as this.

86 John C Rolfe,

178. coTi •)^pvGQfi Iv &>/iOi9.^ Cf. Crj S IfAOiyt firjrt xpvcros cv 8o/iOis> Eur.

Med. 542.
186. Oovptov yovov. Cf. Oovpio^ "A.prj^, Eur. Phoe. 240, but also Bovpw^

Bipiry:, Aesch. Pers. 720 and 756.

201. iX$u>v S cc 66fjLov^ i<f>€<moi, Cf. 8i(ai ^ X^P? '^ Sofjuoii ^^coriov,

Eur. Med. 713, but also Sofjunw i<l>€<moi Ifmvj Aesch. Eum. 580.

202. (TKCVQ irpeirovTio^ aiafi ifiov koBojiI/o/jixu. Cf. iv t^ KaBajtfnuT &fJL<tn

waiSi awfiaroi ; Eur. Ion. 1006.
204. cTir' €1 Tiv aXXrp^ dyrl rija^ c^as 0T0X17V. Cf. <rroX^v 8c riva <lnj^
AfiifH xf^* ^f^^ PaXtlv; Eur. Bacch. 830. crroAi; in the same
sense in Aesch. Pers. 195 quoted above on 122.

208. \vK€UJ¥ &fi<t>i vCiTov a^ofuu hopay ic.rX Cf. (ttoXtv re Brfpoi <Sfi^€-

jSoAAc <T^ Kdp% Xeovros, ^ircp avros i^iowXiierOf Eur. H. F. 465.

209. xd.<Tpa Offpo^. Cf. 8eiv<p X'^'^f^'^^ Oripo^y Eur. H. F. 363.

211. rcrpairovv pifiTjtropM, \vkov iccXcv^oy. Cf. rcrpduro8oc Paxnv Orfpo^

Tt$€fi€v<Ky Eur. Hec. 1058.
216. dAA* €v a o Maui9 muq ^kcutc jcoi iraXtv iripifrnxv "Epfc^. Cf. dAAa cr*

6 Maui? irofiiratbs aKi^ ircAocrcu Sdfioi?, Eur. Med. 759, but also

(vXXafioi 8* ^^VcQK ntu? 6 Mcua? cin^p<irraro9« Aesch. Choe. 808,

and *£p/ui^ 8* cvXoyo)? orvv^yoiycv, Aesch. Sept. 495.
274. fuixfi^ wpo x<V^^ '^ ^^V'^ PauToCofuv. Cf. 8cXror re ypa^9 n/^S*

ijv irpo X^^*' ^^* /SaoTofcis, Eur. I. A. 35.
278. Tovoi^ warp^i ytj^ iptfixtixra^ iri&ov ; Cf. icci pof roST itcXimlvfr ipij-

fioMTCts trcSon Eur. Andr. 314, but also to¥^ iprffUMnur oxwy Aesch.

Ag. 1054.

281. koyov 8c 8i9 rocrov fi ifcow^um^' Cf. Xoyov 8c oc paxptA ^brmraixrw,

Eur. Hik. 638.

282. «'f>o9*l87;9 c^pYa8a9. Cf. irpo9 ^pydL8a9. Eur. Bacch. 445.

287. 'I8ar<n' Xciras. Cf. *l8aror Xciras, Eur. Andr. 295 and Fr. 415, but also

irpo? Ki^aip^i\>9 Xciras* Aesch. Ag. 310.
287. ^Pw frap^rx€^ Cf. ^fiw irapctrxc, Eur. Hec. 1 113, but also rww

iRifxurxw, Aesch. Pers. 330, and w€vO(k wapanrxw, Pers. 325.
296. (rruxwr 8* aiufcro9 Trpouftpcvnfnotf 6S01). Cf. tk ii:v vpov£cpcvviy(r»

iTTLpovy^ Eur. Phoe. 92.
323. ficya? TTCw. This reading of Naudc^ spoils H."^ parallel with Andr.

189, and Bacch. 64a which he dtcs.
329. ApKoiypxv oi <rw(oirrf^. Cf. ApKO^^pufv ^p^U oc TpoAn^cncorrcs a«9cK,

Eur. .-Mc. 383, but Hagenbach himself in his S<^hocleaQ list

compares iSp«r€(r» Onjaianxr fy<w, Soph. Ant. 547.

^ There is no significmnce in such a parallel as this.

' There is no sigT^tBcance in this parallel except in the use of rpov^^pcvmrr&t
and r^oti^^Mvv^'w* which h« alretdy been noted in list I.

The Tragedy Rhesus, 87

331 . (TcAas 0eov. Cf. 0cov €r€)iji^y Eur. Hik. 469, but also Atoc o-cXas, Soph.

Oed. Col. 95, and <^(u8pov oXtov crcXa?, Aesch. Eum. 927.

332. voXX* AyaxTTpi^^v. Stoii, Cf. 6 yc^ Sto^ mvr dycurrpc^ci iraXiV) Eur.

Hik. 331.
357. i Trorpcs <S ^pvyuL. Cf. (3 koXXiotof cS Kv^XoiTrtov, Eur. Cycl. 266,^

but also <3 iroXis i mrpta, Soph. Phil. 12 13.
370. daixpwrov ircXrav. Cf. {^axpixrov ircXn;$ avo^y Eur. Ale. 498.
397. Toifwi o-e. The same expression in Eur. I. A. 1557, but not appar-
ently in the same sense.
399. ov yap n Xc^ets «k SjcXriroi &v ^lXoc?. Cf. ov p.riv ip€h y< fi C09 ixifjuar

l^utVy Eur. Ale. 658.
403. TTOLiuiv 8e hdiptav Kwrpav ovk iiripAlmfuv ; Cf. ttcmv 8^ yaia^ cpKOf ovk

d<l>iyfu6a; Eur. Herae. 441.
409. Kara aropa. Same expression oeeurs Eur. Herae. 801 , but also

Aeseh. Choe. 571, and Fr. 434.
415. vixTTVi ov apiKpa iroXci. Cf. opKovsy mariv ov cfUKpav, Eur. Hipp.

1037. Aeeording to Stobaeus we should read irioris in this sense

in Aeseh. Sept. 54, instead of Tnxms.
419. afiwmv Sc^tov/icvoc. Cf. afAvariv cXKixra?, Eur. Cyel. 417.
421. Xcyitf KOT ofipa aov. Cf. jcar* ofipi i\Oiav paXV* ^ur. Andr. 1064, and

)(£ fuv KOT ofipa aras trpoacvx^Toi Bt^^ Eur. Andr. 1 1 17.
423. ^xBtlav XoyaiK Ttpvinv KtXxvOov. Cf. ovpavov ripviov o&6v, Eur. Phoe.

I, and also rCv &p.^ avras cri iropov rifivu) ydfiov Xtrr^pa;

Aeseh. Supp. 814.
427. Focrror rbv irpos *IXu)y. Cf. irpos *IXtov kootos, Eur. I. A. 966, and

vooTOi 'IXiou irvpyov9 lire, Eur. I. A. 1 261. Aeeording to the

seholiast vooroc is used in the same sense in im <l>opPrj^ vwrrov,

Soph. Ph. 43.
430. aLpanipoi ireXavos. Cf. alparrfpov ircXavov, Eur. Ale. 851. Cf.

seventh example in list XI.
438. ck (TV KOfurcis. Same expression Eur. Orest. 57*. Cf. also ro(r6v&€

ixofuru fivOov, Soph. Aj. 770.
446. prrrcis icv/3ciW tov irpos 'Apyctous 'Aprpr. Cf. It' avrw oXXa pXrjpar

iv icv/Softs jSoXciv iremiffy Eur. Hik. 330. See on 154 above.
453. TOW5 ficy' av^ovFra?. Cf. ci <ru /liy' avxct?? Eur. Herae. 353**
467. poKpas darovcrais* Cf. Sik pxucpaq dirovcrai?, Eur. I. A. 1 172.'
498. loTi ST alpvXutrarov Kporrfp^ 'OStKrcrcvs. Cf. ov rip *08uotrcug i<mv

oifuxXos povony Eur. Fr. 709.

1 H. might have found a closer parallel in Med. 643, i warplt, i iihftara,

* The parallel is not exact in either case.

' There is no significance in such parallels as these.

88 John C Rolfe.

530. Iirrairapoi IIAciaSef aWipuu. Cf. iyyvq r^ iwrcanpov IIActaSo^y

Eur. I. A. 7, and lei S i^* iwra TlXudBw lx<uv Sp<Viory Eur. Fr.

779. In the same way Aesch. (Fr. 304) uses iwrdfiooi of the Nile.
580. rt Sfjra Spa>fiev. The same expression Eur. I. T. 11 88. ri S^ra is

frequent in Aesch.
596. Kop&ov &€SfrfYfi€Vou Cf. icoi Spoiv yc Xvwrj KopBiay Svp(Oi^aofwi, Eur. Ale

1 1 00. A conmion metaphor. Cfl <rv/A<^opa Sajcyo, Aesch. Pers. 848.
608. Setnroiv* 'Aftiyo, K^ryfAaroi yap "QfrBofirp^ rov <rov awi^&rf yrjpw» Cf.

icXvci? yap av8^ Kouvtp ov iropcuy ^eas, Eur. I. T. 1447^ and xXway

fjLsv avSn^y ofipa ^ ovx opSiv to oof, Eur. Hipp. 86, but also ical

raXAa voXX* ^vcuccunu hUausv ^, ci pai iropoirc ^^oyyos ^ 6

<ni/iayw¥y Aesch. Supp. 250.
618. oHTTC Torofuov Kvfcvov iTTcpdv. Cf. &* oWipoi xiayaxpt'^ kvkvou irrcp^

Eur. Hel. 216.
625. Tpifiw¥ Kol 00^09. Cf. ot>^o9 «ciu TpififaVf Eur. Med. 686.
639. auBpdis Xcyounv d^i^ofuu. Cf. rois Xoyov$ cTku aoBpaviy Eur. Hec

1 190. A common metaphor. Cf. Pindar N. 8, 59, aoBpoi icvSos.
647. ovff AfiYifpaySt Ttfof^, Cf. KOKuw yop TiSiy TOT ovK afun^poiWy Eur.

I. T. 361.
693. Tif ^b'Sip«»v 6 j8a9 ; Cf. use of 6 in *£ptt9 6 kot* ofApanav Ids vo^or.

Eur. Hipp. 525.
701 . vyjfniOTrp awopa&x Kiicn/pui jSior ; This reading of Naudc's spoils H.^s

parallel with Eur. Herac. 84.
72 1 . hn yoF ^pvywy to^ '^XY^ jSoXcif. Cf. kot ^Apyos tXi^os ^i/k T08099

Eur. 1. T. 752.*
730. io-ois yap CIS p6Xo¥ ns Ip^crai. Cf. dvifp cts /3a\or jaiA<rniTaiy Eur.

Bacch. 848.*
732. frvfK^opa pjLp€'ta. Cf. ficLptuf, crvfi^op^ TcrXify/Aoos» Eur. Ale 856.

See note on this expression in Est IX.
751. Tus Slv 6\oifjaj¥ ; Same expression in Eur. Ale. 864, and Med. 97.*
756. fcoKtok ircrpoirnu. Cf. same expression in Eur. Med. 364, but also

KOLK^ Sk TpOfjas, Aesch. Pers. 2i6» and c^ inwpaKnuy Aesch. Ag.

556.*

772. <l^^dry x^- ^^* same expression Eur. Med. 612.^

1 These expressions are not paraUeL Here, as elsewhere, the anthor of the
Rhesus shows his originality in coining new phrases.

* This seems likely to have been a proverbial expression. If not, the parallel
b strikingy especially as Schwartz (Z>r metapkaris e mart et re navoH pttiiis qmaeS'
Homes Euripideae, Keil, 1 878, p. 20) says that this metaphor is not fotmd In
Aesch. or Soph.

* There is no significance in this.

^ There can be no significance in snch a paraUel, unless one beeves that our
anthor literally made his plav a piece of patchwork.

The Tragedy Rhesus, 89

796. PaBtuiv SXoKa rpavfjuvroi, Cf. Sopo^ Tax€iav otXoKo, Eur. H. F. 164,

but also owxK &X0KI vtoTOfjua, Aesch. Choe. 25. PaOtiav aXoKa is

used metaphorically in Aesch. Sept. 580.
796. ^acr/ayov TrkrjyrjS' Cf. t^curyavoiv irXifyas, Eur. Andr. 1074, but also

irkriyg Sopo^y Aesch. Pers. 307.
803. tbcdmu irapcoTi. Cf. same expression, Eur. Hel. 421,^ but also ovk

^X^* ^v dKoffoL, Aesch. Choe. 516.
834. w\iKta¥ Xoyovf. Cf. irouzs fu/xt^vots irXcKovotv; Eur. Andr. 66, but

also SoXov irXcxcis, Aesch. Choe. 219.
855. TO TTofkirav. Cf. same expression, Eur. Fr. 196.
870. aXif rviv TtOvtfKOTwv. Cf. same expression, Eur. Hec. 278, but also

aXi« XtXriyfUvwv, Aesch. Eum. 678, and wrffioyfji S* aXt9, Ae%eh.

Ag. 1656.
877. XdCwrff ' aycvr€s tk 8o/iov9 ifwts . * . iropovkcrc. Cf. XdCvaOe

n^vSc KUi 8d/Aovs KOful^€T€y Eur. Phoe. 1660.^
879. vfiSx S* lorra? rourtv iv rtixti. )(p€<t}v Ilpuxfup re kcu ycpovcri (n;/A^vai

vcicpovs ftiirrctv kcXcvciv. Cf. (rrjfiaivtiv ov ravpov dXXa impOnrov

cf^diai, Eur. Herac. 489, but also *Ayafic/tvovo9 ywcujci €rrffiav(a

ropa>9) cw^ ^iravreiXaaav . . . ^irop^co^eiv, Aesch. Ag. 26.
904. ycyov9 Kocvflovtay. Cf. inuScDv icocvo>vuiv, Eur. Phoe. 16.^
949. <ro4>urTriv S* aXXov ovk iirdiofmi, Cf. fidvnv ovx crcpov a$ofiai, Eur.

H. F. 9I2.»
974. ircy^ ouro). Cf. oioro) Sk irtvOoi, Eur. Ale 336, but also mjiijavhs

<f>€fyav, Aesch. Pers. 296.
974. pfo¥ oTcTtt). Cf. pfoy aiati^j Eur. Hipp. 205,' but also <f>ip€iv ioq

p^aroy Aesch. Pro. 104.
980. HI flxuSoirocoi avfi<f>opaC irovoi pportavy K.r,\, Cf. {^17X01 8* dlyafu>v9

drcicvov9 re Pparwv, Eur. Ale. 882, 8eiv6v ro rtVcrciv icai ^epa

^tXrpov fiiyoL, Eur. I. A. 917, Scivoi^ ywoi^v (u Si* wSiviov yuvai,

Eur. Phoe. 355.*

In general the same criticism of Hagenbach's method may be
made as in the case of the expressions in list IX. There remain,
however, in this list of Euripidean expressions more cases in which
no other parallel than his can be given. Is this evidence of deliberate
imitation of Euripides? To my mind it is not. In the first place
many of the parallels are of no significance, as has already been

^ There can be nothing significant in such a parallel as this.
* See note *, p. 88.

' It seems to me that these expressions are not parallel. The idea in the
ptSMge from the Rhesus is quite a different one.

90 John C. Rolfe.

pointed out, others are far-fetched, and the number of those which
show such imitation or adaptation as is shown in the sixth example
in list XI. are very few. Even were they more numerous, they would
not prove our author to be a servile imitator of Euripides. Euripides
himself is justly charged by Aristophanes^ with diligently reading
and imitating Aeschylus and other poets, and Schirlitz^ gives a list of
writers who ' partim consulto partim fortuito ' have drawn on Euripides
for words and expressions, without laying themselves open to the
charge of being servile imitators. The judgment of Schirlitz on the
Rhesus is interesting : ' ' Praeterea examinavi Rhesi fabnlam : quam
qui composuit, tum ^ Aeschylea noimulla, Sophodea, Euripidea recepit
nusquam praeterea ob\*ia, tum ipse nove dictomm partimque andacter
fuit artifex.*

This judgment is supported by oar word lists. It seems fidr to
coQchide that there is ix> evidence in the language and style of the
Rhesus str(»)g e]K>ugh to prove that Euripides wrote the play, or that
its author was a servile imitator. On the other hand, this last list,
added to the other evidence, makes it probable that the author of
the Rhesus was £umliar with all the works of Enr^Mdcs, and hence
liv«d after his time.

In consideriiig the following list of Aeschviean eipr e ssio ns, the
leiader shouM remember that many expiessioDs whkh seem to have
been deri\>evl from Aeschvhis have alreadr been

XL IJST \\r KXn^f^l^i^VC:^ ^t^UKtl aXXXST DOTATIOCC Of AESCHTUS.

^2. ikWrfr vhtmAm^ mirt\««T«€. CI waTr| iwT m ^y gw fp , Aesch. Ag.

The Tragedy Rhesus. 91

308. iroXAoiOt avv xwiwnv iicrvwa t^fiw. Cf. xaXxiiXaToi icAa^ovcn

jcttSoiKCf f^Pavj Aesch. Sept. 373.
430. o^n;^ vcXayos. Cf. viXavoi axfUkrwr^yrfij Aesch. Pars. 818.
514. d/Avccpaf ^X^^' ^^' ^"^ P^X^^ iray€vT€9y Aesch. £um. 190.
934* SffTv iccXinu. Cf. iccXinu yauavy Aesch. Supp. 15.

Most of these expressions are not of great weight. They certainly
do not indicate servile imitation of Aeschylus. The parallel in 308,
as has been said, is striking, and all the evidence thus far supports
the view that if our author took any one of the tragedians as his
model, that one was Aeschylus.

The metrical structure of the Rhesus has been so thoroughly
worked over that nothing more can be done than to give a summary
of the conclusions which have been reached. It is generally agreed
the metre is such, with regard to the resolution of long syllables, as
was usual before the 89th Olympiad. Although agreeing on this
point, Vater and Hermann draw dififetent conclusions from it. Vater
contends that the Rhesus was therefore written before that date, a
view which Spengler * supports. Hermann however says : * ' Quoniam
numeri sunt quales ante Olympiadem LXXXIX, cetera autem qualia
nee ante illam Olympiadem nee proximis post eam temporibus scripta
esse credibile videtur, mu/Zlt? recentiorem essefabulam contendo numeris
ad optima exempla conformatis.^ Lachmann' sees no signs of imita-
tion of Sophocles or Euripides, but of Aeschylus. The metre, how-
ever, he thinks corresponds to that of Sophocles and Euripides at
their best.

The case is well summed up by v. Wilamowitz : ^ ' In anapaestis
Tw So^icXctov vro^cuva xopaxT^pa scriptor Rhesi, in senariis vero ne
antilabas quidem admittit (quas in trochaicis tetrametris habet 686
sqq.) neduro laxiori indulgeat disciplinae. Docemur igitur hac
quoque in re superstitibus etiamtum Sophocle Euripideque sensim
relaxatum esse veteris tragoediae severitatem, duce ac principe in
numeris Sophocle, in renim tractatione Euripide, a duum virorum
magnorum exempb cum aequales pependisse turn eos qui proxime
sequerentur (Agathonem, Critiam, Chaeremonem, Carcinum dico ; de
ceteris parum constat) at fuisse tempus quo Athenienses diffluentif

^ Df Rhtso Tragoedia, p. 12. ^ De Choruis sysUmatit irag. groic^ p. 1 1 6.

* Lc p. 28a * AnaUtta Euripidea, pu 198.

92 John C Ro/fe.

levitatis taederet; rediisse igitur quantum possent ad antiquiora
exempla ; atque suspicare licet hac ex causa Theodectae gloriam non
minimam promanasse. Cuius severioris Musae unum poetam inlus-
tra\nt Meinekius Moschionem, cum doceret eum ab omni pede tri-
syllabo in senariis cavisse; mediocris ingenii fetus at consimili
subtilitatis amore commendati extat Rhesus, saeculo quarto exeunte
baud dubie Athenis scripta.'

It will be seen that Hermann and ▼. Wilamowitz substantially
agree, except that while the latter assigns the Rhesus to the end of
the fourth century, the former attributes it to an Alexandrine writer.
The principal argument against the latter view is the number of the
choral passages. We know that the Alexandrine writers so subor-
dinated the chorus that Lycophron in his Cassandra dispensed with
it altogether. In our play the chorus forms an essential part of the
play, and the choral passages are marked by a simplicity and beauty
which does not suggest Alexandria.

A line of argument which has been wholly neglected in all special
disseitations on the Rhesus, but has been touched on incidentally by
writers on the grammar of Euripides, consists of deductions from
syntactical peculiarities. Evidence of this kind seems especially
\*ahiable; unfortunately, however, it is difficult to collect. Some
of these writers ha\^ lelt the Rhesus oat of consideration altogether
as noQ-Euripidean, and in the other cases it is difficult to o^ect
and classifv the scattered references. Soch material as I have been
able to gather is gix-en below.

l\\Sho Momntsen,^ in exuninii^ the use of ew and of furm with
the g<»uii>>^ in Euripivies, finds the btter coostrKtioii more frequent
in Euri)\ide$ than in the vxher tn^>ediaiis. Of the Rbesos he says : *
* IVt Rhesus hAt \\« allem am wenigneii jmtb c, gen. and venith
skli a^x^h ^Uiloich ab xmec^t.* He cdntinties : * Deoa die Zeit dcr
ei«ea .VWxjmdriner werwArf dkse CccBSTsctkwi wieder fist voU-
^^xaik)^. Hienuoh ru >mhe)k«i konnte der Vf. des Rhesos mit
I yk\>pduvMV A)v>4k>«iiu$ RlKiidmSs KallinuKiixB&, LecoidjKS k\xi Tarent
ii<evK*e«^ geKiM*Ni $<^n/

Harms«^n* $ji>^: • Muhc^ vej\> r&n^ qxaoa Jiixctxrom mraiitnr

The Tragedy Rhesus,

93

genitivus post praepositionem coUocatus. . . . Rhesus hoc in re
longe recedit ab usu dicendi Euripideo in quo quidem inveniuntur
octo huiusmodi exempla.* ^ Further on, speaking of anastrophe, he
says : ' ' Inter Euripidis singiilas fabulas hac in re nullum fere dis-
crimen est; tantum id commemoratione dignum est, Rhesum
fabulam multo minorem exemplorum numerum praebere, quam
genuinas fabulas Euripidis. Sunt enim in Rheso tres loci (72, 397,
930) » genuinae fabulae vero singulae octo minimum exempla conti-
nent ut Ale, vel novem ut HeracL, vel tredecim ut Cyclops, quam-
quam multo brevior haec fabula est ceteris.'

Tachan,' speaking of the use of the infinitive with final force, says :
' Quam clarissime elucebit, id quod supra iam monui, exemplorum
copiam eo maiorem fieri quo recentiore tempore scriptae sunt
fabulae.' After stating the principle on which his infinitives are
selected,^ excluding such cases as the infinitive with 8tiS<i>/Lu and <^-
IScvfuUf Seivos Xcyciv, Koipo^ SjcowTaif etc., he gives a table, from which
the following selections are interesting : —

All cases.

fait, Swut,

Participle.

Infinitive.

Alcestis, 1 162.

«3

8

2

3

Medea, 1419.

21

16

3

2

Hippolytos, 1466.

24

19

2

3

Elcctra, 1359.

44

26

10

8

Here. 142S.

41

29

6

6

Phoc. 1766.

43

1

25

13

5

Orest 1693.

4t

34

3

4

Rhesus, 996.

25

8

9

8

1 150, 155, 203, 221, 471, 502, 598, 660. « p. 25.

» De enuntiatumMm ftnalium apud Eurip, ratione atque usu, p. 72.

* ' Eos tantum commemorabo infinitivos qui re vera pro enuntiatione final! extant.'

94 John C. Rolfe.

Tieuel ^ shows no important variation in the Rhesus from the pla3rs
of Euripides. He accepts the Rhesus as written by Euripides, and
assigns it to the years 445-442 B.C.

Professor Goodwin/ speaking of the independent clause with /uii|
or ^ 01^ expressing desire to avert an object of fear, notes that after
Homer we find no examples of the independent clause either with
^ or with iM^ ov until Euripides, who has three cases of the former
and one of the latter.' An example with ftif ov occurs in the Rhesus.^
Cases of this kind are so rare that the coincidence in usage must be
admitted to be striking. It is possible that the author of the Rhesus
followed Homer directly, whom he has followed in other respects.'
It is perhaps more reasonable, however, to admit diat he ft^wed
Euripides.

Speaking of the use of vpur with the indicative in poetry. Professor
Goodwin* notes in Aeschylus' one example, after a negative; in
Sophocles one,' after an affirmative; in £uri[»des ^[<r^ all after
affirmatives; while die Rhesus has twa**

Hus bst coincidence is very striking, for unless our author is here
influenced by Pindar, who has three cases, he ts certainly following
Euripides. ImiutioQ of Pindar seems unhkdy, since neither of die
cases » in a hric passage.

FVom diese iUustrations it may be seen that dits fidd is a promising
one^ Hie examples are not sheeted, but aie all that I have been
able to collect. It will be seen that three of the six give evidence
aiiamst the view that Euripides wrote the pity, two give eqoaDy strong

• Ak. i{l5» l>c*l. r:^ H. r, i^^kw Ttw «&i. <^ m^aA libe )ut ii «i& ^ mk

• 115.

fi)l^ «^ yw^}^>^«^ ><«ft fis-vn tW )wipuM!e «*< die either utOc it wc«^ «oeHi tint
^ i> nit>i(r ^ <vNiix«t«i<^>e d>M ^(v*:fe«» ^ Vv4:s. .fn die Isifacsss is d^ ««k of

• At>A, u<v ^ A ,^^ >«^A. n^;^ H^^. v^ .Ut, iSa. l^ die secaai s&d

The Tragedy Rhesus. 95

evidence on the other side, while the other throws no light on the
subject at all. Taken alone the last two examples are strong argu-
ments for the view that Euripides was the author of our play. Taken
in connection with the other evidence, they seem merely to show
Euripidean influence.

After this presentation of the evidence, it may be well to examine
all the hypotheses and possibilities, and to see what conclusion can be
reached. In the first place, was the play written by Aeschylus or in
his time? Is there any sign of the influence of Aeschylus? To the
latter question the answer must be in the affirmative. The number
of unusual words, many of which seem to be suggested by Aeschylean
words, the anapaestic beginning and the absence of a prologue, the
close connexion of the chorus with the action of the play, the num-
ber of words and expressions which may be traced to Aeschylus,
make any other view unreasonable.

That the play was actually written by Aeschylus has been main-
tained by no one. That it was written in his time might be inferred
from the language of Scaliger^ and Lachmann.' It will be shown
below that so early a date seems impossible.'

Next we may ask the same questions about Sophocles. That the
Rhesus was written by Sophocles was maintained, as has been said,^
by Gruppe ; that it shows imitation of Sophocles, by v. Wilamowitz.*
The answer to the former is the same as to the corresponding
question about Aeschylus. The latter is a more difficult question.
The comparison which v. Wilamowitz makes with the Pastores^
does not seem convincing, and the argument which he derives from
the metre' cannot be held to be conclusive evidence of imitation
of Sophocles. On the other hand, the examination of the words
and expressions which Hagenbach regards as Sophoclean,' taken
in connection with the large number of words which do not occur
in Sophocles at all,^ seems to show that Christ's statement is hardly

^ Sec p. 61. * De choricis systematis trag, graec.^ p. 1 16.

* Cf. also the remarks on TcXro^rcU and on the metre pp. Z^ and 91 above.

* See p. 62. • See p. 64. • See p. 67.
' See p. 91. ' pp. 83 and 84.

* See p. 82, and Hagenb., De. R. T., p. 30. It should be noted that a few of
the words cited by Hagenbach are found in Sophocles; x^t*^^9 for instance,
appears in his own list of Sophoclean words.

96 John C. Rolfe.

too strong. There are certainly no signs of direct imitation of
Sophocles.

Was Euripides the author of our play? Are there signs of imita-
tion of Euripides? To the latter question the answer must be in
the affirmative.

The answer to the former question is made easier by the fact that
all those who now support that view admit that if Euripides wrote our
play, it was his earliest work. Even they ^ acknowledge that there is
a great difference between the Rhesus and the later plays of Eurip-
ides. On this point Hermann's remark,' though not conclusive, is
significant : ' Mutat profecto aliquid aetas, facitque saepe ut quis
alius \idetur \\x foetus quam adulescens fiiit ; at non mutantur omnia,
sed est etiam quod sibi constet, manentque eiusdem ingeni vestigia.'

But there are more tangible proofs that the play could not have
been written so early. The list of words and expressions given by
Wecklein,' the fact that the action of the play demands four actors,^
and the evidence that the author of the Rhesus was acquainted with
the later pla\*s of Euripides,^ seem very strong arguments. That the
play was not the work of Euripides, though perhaps not necessarily
that it was later than his time, is shown by other evidence. First, by
die apparent lack of a prologue.* Hartong made so much of diis as
to say ^ : * Aut Euripidem Rhesi auctorem esse negandom aut pro-
oemio banc fabulum instructam esse credendom.* Then he betravs
himself into die hands of his adversaries by isSaD% utterly to prove
the fonner existence of a proiogoe by an examination of die frag-
ments of the Sy^tt^rsi4^ of Accius. Next die arguments from the
fifst diree cases cited in die examination of the smtax* seem to be
of some weight. And finally we can haidhr believe that Aristophanes
woqM have missed the c^f^poctunitt of lidkuhog a jonthiul efiusion
of Euripides.

The idea of DindoH, that the Rhesus was written for the foorth
plaoe in a tetrilo^^ which Bergk characterixies as ^gsrnx mm^td^h^
nuy readily be answered by aesthetic ax$;nixieiits. for it is evident that
the Rhesus has not the ^d^htest rKiemUance to the extant woiks of

^ See Om«'^ reauo^ jk ao^. * 1 c p. 574. * p. Sa.

The Tragedy Rhesus. 97

that kind. More conclusive is the argument from the metre, and
perhaps Bergk's suggestion, that the conclusion of the Rhesus sug-
gests that it was followed by another play, deserves consideration.

Was the play the work of a servile imitator ? Reasons for answer-
ing this question in the negative have already been given.^ The
Rhesus with its many departures in language and style from the
normal Greek tragedy, and with its anapaestic beginning, may be
the work of an imitator of Aeschylus, but hardly of a 'plumbeus imi-
tator ' of Euripides and Sophocles.

Was it then written by an Alexandrine, one of the famous Pleiad ?
Aesthetic arguments against this view might be multiplied. Crutt-
well says : ' ' The drama could find no place at Alexandria. Only
recondite legend and frigid declamation, almost unintelligible from
the rare and obsolete words with which they were crowded, were sent
forth under the name of plays.' This certainly does not apply to the
Rhesus.* Moreover, all the evidence indicates that if our Rhesus was
substituted for a play of the same name written by Euripides, the
substitution must have taken place before the play reached Alexan-
dria.^ The argument from the position of the chorus ^ also deserves
attention.

After this consideration of the hypotheses, the following con-
clusions may be reached. Our Rhesus is not the work of Eurip-
ides. It was written by an Athenian who lived between the end of
the Peloponnesian war and the time of Demosthenes, who made a
well-meant but not wholly successful attempt to write a play of the
old school, strict in its metrical structure, and avoiding the peculiar-
ities of the school of Euripides. He naturally took Aeschylus as his
model. Being familiar with the plays of Sophocles and Euripides,
he consciously or unconsciously followed them somewhat, especially
the latter, which probably set the fashion in his time. He had more
poetic than dramatic ability.

^ pp. 82, 84, 89, 90, 91. ^ Hist, Rom, Lit., p. 220.

* Especially since Eysert has pointed out that the Rhesus is not greatly ex-
ceeded in its number of rare words by some of the plays of Euripides.

* See p. 61. * p. 92.

THE USE OF HER CLE (MEHERCLE), EDEPOL (POL),
ECASTOR (MECASTOR) BY PLAUTUS AND TERENCE.

By Frank W. Nicolson.
I. Comparison of the Use of the Words in general,

Hercte, MehercU, Edepol, Pol, Ecastor, Mecastor,

Plautus 639 3 356 250 101 17 (=1366

Terence 96 2 22 55 5 2 (= 182

735 +5 378 + 305 106 -h 19 (=1548

Grand total of all cases, 1548.

Plautus uses these words 1366 times in 20 plays (20,888 lines).
Terence uses these words 182 times in 6 plays ( 6,074 lines).
i,e. Plautus averages 66.3 to a play, or i in 15^ lines.
Terence averages 30.3 to a play, or i in 33 lines.
Ratio of Plautus to Terence more than 2:1.

II. Comparison of the Use of the Words in particular,

HercU (Mehercle). Edepol (Pol). Ecastor (Mecastor).

Plautus 642 606 118

Terence 98 77 7

HercU, Edepol (Pol), Ecastor.

i,e, Plautus 5.432 5 '135 ' i

Terence 14 11 : i

III. Comparison of the Use of Forms of the Same Word.

HercU, MehercU, Edepol, Pol, Ecastor, Mecastor,

Plautus 639 3 356 250 loi 17

[i,e, 213 : I 1425 : I 6:1]

Terence 96 2 22 55 52

\i,e, 48 : I I : 2.5 2.5 i]

99

100 Frank W. Nicolsan.

i>. Plautms prefers form Edepolxo Pdl in ratio 1.5 : i.
Terence prefers form Pol to Edepol in ratio 2.5 : i.

Note. — In both poets, women prefer the form Pol to Edepol
(P/ttaAfs, 81 : 27, i>. 3:1; Terenee^ 46 : 10, i>. 4.6 : i) ; while men
prefer the form Edepol to /^/ {PtamHu, 329 : 169, i>. a : i ; Terence y
la : 9, i>. 1.3 : i). See § VIII., Table 3, below.

IV. The Use or T^ese Words by Men and Women Distinguehed.

A. (Me>hercl£ is used altogether by men in PLiutus and Terence,

Note* — There are^»«r possible exceptions to this rale; cases
where, according to the Mss^ htrck is ised by jssnmm^ These have
an be^:i emended by various editors;, to conform to the general rule,
as bid down by Gellius^ 3W/. Alt XI. 6.

These cases are the folfowiog : —

I. Flaut. Oi^VM9$2. Ritschl and Ussicig both give the speech
^ Hand mentire hercle ; nun pa. « . .'* to a woman, die former to
Ckvstmsita^ the latter to Afrrr^mM^ The passage b rciy corrupt;
the kitnr^ of the odker M5&. b indecipberab^ in the Amhrosian,
thoQ^ the space wooM admit it. The state ci the test also makes
it doQbdoI as to the pefstD(b> to whocn this aztd the two preceding
speeches s2kkiU[ be re&^red. Several ot the Mss^ rocItMliii^ the
Ambcosaan^ <dk^ inoc sepiixate this speech &v»oi the E^tme just pre-
cedcQg. The IVUpihia Edctioa j^tves the wocvi^ "^ haoid oKntire herde,
etc^* t<;> CKyff^^. Sk> r>tE$sakie.

i. F^ttt. Cis^X. 50W Cr ^«MMjKMt«K. "^ Ei^isKJem hercie addam op«ram
ijeduikX**' EnsKswied by Seydfest ^Stiad. p-^ 15 k '*'kitmr iK^fffm> ctc.^

5t. PlsKtt* Mf^(.\ 719k £hrrippui^ •'^Orpw httdie sence : sed ta me
t!en(^t» :§ciieB&^ Bl>thfe ^:tve$ the ^uist poxt ^ IjissmtA-xms^ aad the
Ibst tK> JPk/f^pui (^ceoiini^ a)^ i^^ Imt stn^}. Ussai^ bcae&ecs ^&e Hne
as Tznstiiii^ tv> tbie cvwtfiext. We&se teciioz^ the Sme^ ccosniectBig this
qse o£ ^i^ciir a itcz^ v^^' thie poet^ ajad as teiadBn^g: amikn^ aTCfioa- d&iB^
tio :$dDO«ir tbtiifi tbae p^y wad i»x wtittea by Planfitss.

4^ PltoL, J>au.-a6c. ito^ MiE^fm»nL. ^Hx* lii! (h)erdie cjoapiL'*
T^ is dh^ ceod^tog: oi mpst oi ^ \l»^ Tb# AntbccssBZ^ hiswcver.

The Use of HercUy Edepol^ and Ecus tor. loi

(M)eca8tor.

B. (M)ECASTOR is used altogether by women in Plautus and
Terence.

Note. — There are two possible exceptions to this rule, as follows :

1. Plant. Asin. 930, where the Mss. give the speech " Ecastor qui,
etc." to Argyrippus, Given to Philaenium by Pareus.

2. Plant. Asin, 898, where the Mss. give the speech " Ecastor dig-
nus est " to the Parasitus. Given to Artemona by Acidalius.

(Sde)pol.

C. (Ede)pol is used by both men and women, in Plautus and
Terence. The ratios are as follows ; —

Used by men. Used by women. Ratio,

Plautus 498 108 4^ : I

Terence 21 56 1:2^

i,e, Edepol ( Pol) is rather a man^s oath in Plautus and a woman^s
in Terence. It has been shown above that men preferred the form
Edepol to Poly and women the form Pol to EdepoL

V. Comparison of the Use of {Me)hercle and (^Ede)pol by Men

IN Plautus and Terence.

In Plautus y men swear (i) by Hercules 638 times; (2) by Pollux
498.

In Terence y men swear (i ) by Hercules 98 times ; (2)* by Pollux 21.
/>. Ratio {Me)hercle to {Ede)por\n Plautus, 1.3 : i.
Ratio {Me)hercle to i^Ede^polxn Terence ^ 5:1.

VI. Comparison of the Use of {Ede)pol and {M) ecastor
BY Women, in Plautus and Terence.

In Plautus, women swear (i) by Pollux 108 times; (2) by Castor
118.
In Terence, women swear (i) by Pollux 56 times ; (2) by Castor 7.
/>. Ratio {Ede)polio {M) ecastor in Plautus i : i + .
Ratio {Ede)polio {M) ecastor in Terence 8:1.

VII. General Statement of §§ V. and VI.

In Plautus, men swear about equally {1.3: i) by Hercules and
by PoUuXy and women about equally (1:14-) by Pollux and by
Castor,

Frank W. Nicolson.

In Tekence, men prefer Hereuki to Pollux in the ratio 5:1; and
women prefer PoUux to Caslor in the ratio 8 ; i.

VIII. Tables showing the Use of these Words by Various
Classes of Characters.

(N.B. Tb* miuBbcn iiftn tba lign + indiciB Ibe occuRuca oT tlM for
I. ECASTOR (MECASTOR).

^

J

i

s

3
1

i
i

1

1

1

1

s

,...

AHmria

S

9

u*

9

7+1
£+)

6+.

S

*+'
17+6

■

"

.„

B«ch-d«

•

Cptiri

!)+»

Cutellaiu....

Curculio

Epidicu.

'

s

'

.'.

4+t

L

'l

.„

MiluCIa

■

'"

Fteudolu.....

■

6+.

3

Trinumroiu...

..+,

'S

8

J7+6

■■t+3

«+s

3

4+1

.

4+<

.«.„

T

3

„.

.

*.

S-

SocilV-, A.,>bore.

I I,. WOMEN.

...

1

s

i

1

1

1

1
j

.43

II

S3

I»+i4

15+19

4+6
9* 4

1*+ 9

is+i

■■+6*

+J

+a

+*

US

..'*!,

.+.

+1

<

*;■

H-l
■*4

+»

+1

■+3

■+1

!+«

4+11

/•x-

.=^„

u

'*•!

1+4

,+.

+.

»j+a.

3j6+»jo

r«

I+U

+.

.t«-46

«+„

Crux

T«

-Ui

lull

UM

dT.

Irf+W.

ACCENTUAL RHYTHM IN LATIN.
By J. B. Greenough.

IN the great controversy as to whether the Satumian verse was
quantitative or accentual, each side maintains that the facts which
support the view of the other are an accidental result of the laws of
accent in Latin. It is said, on the one side, that a given scheme
of quantity necessarily produces in some degree a certain succession
of word accents, and, on the other, that a given succession of word
accents must cause a certain scheme of quantity.

Now it so happens that we have one notable example of the
connection between the two schemes in the history of the Sapphic
verse. That this verse was originally purely a quantitative one is,
of course, obvious. Yet throughout Christendom the commonly
received scheme of the verse, except with professed scholars, is an
accentual one, and has been for 1500 years. One familiar example
is the fine German air by Fleming, to which Integer vitae is sung by
men's voices, and, curiously enough, sometimes at funerals. Another
is a poem by Southey, The Widow^ better known perhaps through
the parody by Canning, —

Needy knifegrinder^ whither art thou going f

Here we have an example of a quantitative verse changed by the
effect of the word accent as it depends on the succession of long
and short syllables into an entirely different one, a sapphic turned
into a dactylic metre. So the rhythm

rrif rirr rirrirr

In- te- ger vi- tae see- le- ris- que pu- rus
became as soon as it was read by the word-accent,

r r r I r r 1 r r r r 1 r r '

In- tc- ger vi- tae see- le- ris- que pu- rus.

105

io6 % B. G re enough.

The causes which made this possible are two, one the permanent
spondee universal in Latin in the second foot, and the other the
caesura after the fifth syllable, which is almost always observed in
Latin, but not in Greek, though it may be that Alcaeus practised it.
If either of these phenomena is absent, the dactylic effect of the
word accent is almost sure to be lost, though occasionally it happens
to be preserved by some accidental succession of word accents.

In this dactylic scheme there are naturally four principal ictuses
on the I St, 4th, 6th, and loth syllables, with two possible secondary
ictuses on the 2d and 8th, thus : —

In' te' ger vi' tae || see' le ris' que pu' rus.

f r r I r r 1 r r r r I r r '

As the fourth syllable is long in the Latin form, it must, if the caesura
is observed, be a long penult and therefore accented ; as, —

lam satis ter'ris;
Grandinis mi'sit;
Aut in umbro'sis;
Nee venena'tis;
Unde voca'lem.

The only exceptions would be (i) the case in which one monosyl-
lable alone (not two together) should stand before the caesura, — a
case which, however, does not occur, whether designedly avoided or
not, — and (2) the case where an elision throws the word accent out
of its proper place. Of the latter case, there are four examples only
in Horace, and none at all in Catullus, and it would seem as if it
were purposely avoided. The cases in Horace are : —

Thessalo victo'r« et (H. 4. 10) ;
Oderit cura'r« et (IL 16. 26) ;
Imbrium divi'n* avis (III. 27. 10) ;
Pegasus terre'n'i" equitem (IV. 11. 27).

This is obviously too small a number to be of importance.

If the next word after the caesura is a word of one, two, or three
syllables, the first syllable must be accented, inasmuch as the next
syllable is short ; thus : —

. . . a'vidum domando;

. . . et' avitus apto;

. . . pri'us an quietem;

. . . si'mul alba nautis.

Accentual Rhythm in Latin. 107

If the word has four or six, there must always be a secondary

accent on the first. This results from the length of the third syllable,

thus : —

. . . su"pera're pugnis;

. . . me"diocrita'tem.

A five-syllable word does not occur in this position.
The last syllable but one of the verse must, unless the last word is
a monosyllable, be a long penult, and so receive the accent, thus : —

. . . atque di'rae;
... at rube'nte;
. . . fabulo'sus;
. . . mediocrita'tem.

Even if the verse ends in a monosyllable, the conformity of accent
may be retained by another monosyllable preceding ; as, una mo^rs
est: ff SoL

There are only four cases in Horace and three in Catullus where
the conformity is destroyed by a monosyllabic close. These are,
in Horace : di'em qui, ne^fas heu, mi^nus iam, pur^pura-ve (C. s. 9 ;
IV. 6. 17; I. 25. 6; II. 16. 7). In the last case the accent is
practically restored by a natural secondary accent on the a.

Here again the cases are so few as to be insignificant, and we have
three out of the four dactylic ictuses agreeing with the word accent.

In the other perceptible word accents, the primary one of the
first measure and the secondary (or half) one of the third, there is
more chance for variation. In the first measure, as the second
syllable is short, the word accents must always conform at least to the
primary ictus of the dactylic measure, as : —

In'teger; — A'cer et; — la'm satis;
Si've neglectum; — Ne' sit ancillae.

If a word accent comes on the second syllable (as in lam satis) ^

this naturally falls in as the secondary ictus of the measure, K T T •

Even such a rare case as nuptiarum expers is not felt as an irregularity
on account of the necessary accent in ixpers, taken in connection
with the secondary accent on the first syllable {nu^^ptia^rum). The
second word accent after the caesura, —

lam satis terris nivis at'que dirae,

io8 y, B, Greenough,

is likely to vary from the dactylic scheme, as there is a greater possi-
ble variety of combinations. But in the great majority of cases this
accent conforms, because the proceleusmatic measure (the third of
the dactylic scheme) may always have a secondary ictus or not, as
the word accents suggest. If the form is like sceUris^que purusy the
accent and ictus change places, the secondary accent of sce^Ueris^que
becomes the primary ictus, and the primary accent falls on the
secondary ictus. If the form is nHvis afqtu^ both accents conform,
one to the primary and the other to the secondary ictus. If the
form is gra^vida sagiftisy the secondary ictus disappears, as also in
iaUuHs ne'que ar'cu. In lo^ca fab^^ulo'sus the secondary accent falls
on the secondary ictus. Examples are : —

Per meos finis et' aprica rura ;

Si tener plane ca'dit haedus anno ;

Larga nee desunt Ve'neris sodali ;

Lenis incedas ab'^eas'que parvis ;

Nee dabunt quamvis re'deant'' in au'nim ;

Auream quisquis me'"dio"crita'tem.

The few cases of monosyllabic endings may somewhat disturb this
part of the verse also, but the cases, as seen above, are not numerous
enough to count.

Thus if the caesura is preserved, the word accents of the verse
must practically correspond to the dactylic scheme into which they
have since developed, so that the moment a sapphic is read as prose,
it at once becomes dactylic.

If the caesura is neglected, the verse does not usually conform at
all, and were it not that Horace had set the pattern in his rigid
form, we should probably never have had this dactylic measure.
Catullus' Sapphics are not numerous enough to be of any account,
but out of thirty verses he has eleven in which the masculine caesura
is disregarded. One more has a compound divided at the caesura.
One of those with feminine caesura happens to conform to the
dactylic scheme on account of an elision : —

Gallicum Rhenum horribile aequor ulti-
( Mosque Britannos).

In Horace, out of 612 verses, 47 have a feminine instead of the
masculine caesura, as in —

Phoebe silvarumque || potens Diana.

Accentual Rhythm in Latin. 109

These 47 are curiously distributed, being for the most part confined
to a few (only nine) odes. Thus there are nineteen out of 57 in
Carmen Saeculare, twelve out of 45 in IV. 2, six out of 33 in IV. 6 ;
the remaining ten are scattered, but in only six odes, I. 10, 12, 25,
30, II. 6, and IV. 1 1 . This curious distribution could hardly have
been accidental, but it does not seem to be chronological nor caused
by the subject.

Seneca, the tragedian, out of a very large number of sapphic
verses, has not a single feminine caesura, and no verse that does
not conform to the dactylic scheme.

Ausonius, out of 48 verses, has only one with feminine caesura : —

Lesbiae depelle modum quietis.

In all cases the Adonic must conform to the dactylic scheme,
unless there is a monosyllabic close, which seems not to occur.

To show the difference in treatment between the Latin and the
Greek, in the first ode of Sappho only eight verses out of 21 are
constructed after Horace's model. None of the rest can be read
with the Latin word accent so as to be rhythmical.

The definite and intentional following of the dactylic scheme seems
to appear first in Christian hymns. One attributed to Ambrose (t397
A.D.) is the earliest I have found (wrong quantities in italics) : —

Christe cunctorum dominator alme
Patris aeterni geniti^ ab ore
Supplicum MoXa parit/r ac hymnum
Cerne benignus. — Daniel, XCVL

So others by Magnus Felix Ennodius (tS2i a.d. ; Daniel, CXXI.)
and Gregory the Great (6th cent.; Daniel, CXLVL). The famous
hymn to St. John by Paulus Diaconus (8th cent.), which gave the
names to the notes of our scale, is written in this form : —

Ut queant laxis resonare fibris
Mira gestorum £Eunuli tuorum
Solve polluti labii reatum
Sancte Johannes.

This metre became one of the most common forms of Latin hymns
throughout the Middle Ages, and a very large number are to be
found in Daniel's Thesaurus. They are variously treated according

no y, B, G re enough.

to the learning of the writers. Some of them conform entirely to
the Latin rules of prosody, though obviously intended to be read
and sung by accent ; others seem to have little regard for quantity
except so far as the scheme of accent produces the converse effect
to that discussed above, securing a general conformity to prosody.

The prose rhythm of the Alcaic is not so simple, but there is a
strong tendency in this also to an accentual rhythm. The scheme
of this accentual rhythm is —

r r r I r r II r r r I r f r

and the syllables by quantity for comparison are —

As the diaeresis after the fifth syllable is almost imiversal, there are
only 59 verses out of 664 in Horace's Alcaics which do not have an
accent on the fourth syllable, as must be the case from the penulti-
mate law of accent. The exceptions are cases where the diaeresis
follows a monosyllable, as matu^rior vis, or where the word accent
is disturbed by elision, as dixi sacral ment'^y insania audi^r*.

Of these 59, 19 have two monosyllables together, as, securus Oquae;
— pugnare Thracum est, so that the accent is practically restored.
This leaves 40 which would be defective, but two of these, through
having no diaeresis, happen accidentally to conform. Thus we have
only 38 out of 664, a little over 5 per cent, that fail to have this
principal and decisive accent.

The first word accent may be either on the first or second syllable,
as in vides ut aita, and Soracte nee iam^ but in all accentual verse
the first accent is constantly misplaced, so that this practically makes
no difference. In case the word accent falls on the second syllable,
the first is almost inevitably treated as an anacrusis, and the measure
becomes —

ill r 1 1 I

So we have practically two accents fixed in all but the small propor-
tion given above.

As to the last half of the verse, the rhythm is not so marked as
in the first There are 236 out of the 664 that at first seem not to

Accentual Rhythm in Latin, in

conform to the dactylic scheme in one or both of the accents. But
19 of these have one accent right and a monosyllable in the place of
the other, as cla^vidus ef me^ro; — auf ma* re Cas^pium, and may be
treated as conforming. The rest are cases where the first word after
the diaeresis is a monosyllable, a dissyllable, or a word of four sylla-
bles, as eras' fo' His ne'mus; — li^na su'perfoUo; — sustC neant 0' nus ; —
or where the accent is disturbed by elision, as tristftiam et meatus; —
parti' cul^ u'niiique. But of the first kind, 1 7 have monosyllables in
both of the accented places, and it will be noticed that a monosyllable
may always receive the accent, and in that case the following accent
(in the present sequence of longs and shorts) falls in with the sec-

/ //
ondary ictus of the dactyle, as nam* ti}bi; — dis' pi'etas ^ P j^. So

the verse substantially conforms, as in eui licet in diem ; — et pecus et
domos.

Of the remaining 200 there are 85 which contain one monosyllable
in the accented position followed by a trisyllable. This arrangement
has one of the two required accents, and the want of the other is
little felt because of the tendency of the rhythm to generati^ a sec-
ondary accent in the proper place, as non' adytis'* quatit This ten-
dency is helped by the falling of the last word accent on the secondary
ictus of the dactyle. Thus, te' pro^ugus'* Scythes becomes

r r r r r r

Thirty-three more have three dissyllables in this part of the verse, as
le'ne me' rum ca'do ; — ifle po'tens su'i. These limp a little, but they
only have one accent out of four misplaced. The first word accent
fails properly, and one word accent besides fails in with a secondary
ictus. Thus we still have

r r r r r r

Deducting ail these as tolerably rhythmic, we have 82 left, of which
three or four are disturbed by elision, as particul'^ undique, consi-
li*" et tuos. These have only one misplaced accent, but it comes
in such a place that it breaks the rhythm, at least to my ear, more
than any of the cases thus far treated. The rest are formed of a

112 y* B, Gre enough,

four-syllable word followed by a dissyllable, so that neither accent
conforms.

This then represents the whole number of verses that do not
conform to the scheme sufficiently for an ordinary ear, only about
12 per cent of the whole.

But the Alcaic never seems to have been chosen as a Christian
hymn-form, probably on account of the want of uniformity in the
verses of the strophe.

Of the Asclepiadean, on the contrary, there are many examples
among mediaeval hymns, though that rhythm seems much less suited to
adaptation than the others to an accentual scheme. It has a general
dactylic flow, but cannot be made to conform without violence, as : —

Maece'nas a'tavis e'dite re'gibus,

O et' praesi'di"* et duPce de'cus me'um,

Sunt' quos curri'culo pul'ver* Olym'picum.

The mediaeval examples proceed in this form ; —

Exul'ta nim'ium tur'ba fide'lium
Sollem'ne ho'die mar'tyris in'clyti.
Est fes'tum mo'dula car'mine for' titer
In lau'dem do'mini at'que poten'tiam.

— CyxilUiy Daniel, CLVII.

Fes'tum nunc ce'lebre mag'naque gau'dia
Compel'lunt a'nimos car'mina pro'mere
Cum Chris'tus so'lium scan'dit ad ar'duum
Caelo'rum pi'us ar'biter' (?).

— St, Rabanus Maurus, Daniel, CLXXXVII.

It is probable that such were not written until the use of the more
flexible metres had established the style of treatment

The question arises whether the classical writers of these metres
could have been conscious of the accentual rhythm of their verse.
In regard to Catullus, it can be assumed with confldence that he was
not, or at least did not care for it in these complicated metres, of
which by the way he made no particular study. But of Horace
it is not so certain. The rigid spondee, to be sure, may have been
forced in a manner by his material But in the case of the
caesura, that seems impossible. The distribution of the verses
without caesura is so strange that it could hardly be accidental.

Accentual Rhythm in Latin, 113

Three- quarters of the cases are in three odes. One-third are in the
Carmen Saeculare, which we know to have been written for music,
and actually sung, no doubt to a Greek melody, in which the prosaic
rhythm would, of course, disappear. This phenomenon has been
variously explained. The view of Kiessling is that the rigid caesura
is due to a return to the original constitution of the Sapphic. But
the chronological distribution of the cases of neglected caesura seems
to be inconsistent with this view. Most of the cases are in Horace's
later efforts, and we can hardly suppose that he adopted a theory
of this sort as more profound, and then abandoned it in his later
compositions. The only explanation would seem to be that he used
the freedom of Sappho in the odes which were to be sung, or in
which he had the idea of music in his head, and restricted himself
for the most part in those which were to be read. May it not be
that, wishing to reach a wider audience than the little circle whose
ears were tuned to Greek music, he purposely adopted the form which
was readable as prose ? There seems to be no other reason for his
adopting the restriction of the caesura which, as we have seen, is the
most decisive factor in determining the prose rhythm.

Horace, as a man of the people, was quite capable of so doing,
and he represents himself as charged by his enemies with wishing ta
be read by all classes. One might almost think he had this in mind
when claiming to have introduced the metres to Latium. Then, again,
the Sapphic and Alcaic, in which alone the prose rhythm is apparent,
were his favorite metres. There was evidently no violent change
in the transformation of the classical metres, at least as they were
felt by the lay public, into the Christian hymns. Whatever we may
think of the Satumian verse in particular, it is clear that the Latin
ear was tuned to a word-accent verse. Most of the early Latin
poetry is so rhythmic in its prose accent that it could hardly have
been written by any one whose ears were not accustomed to verse
constructed on that principle. Ennius'

O magna templa caelitum commixta stellis splendidis
reads like a Christian hymn, and

Extemplo acceptum me necato et filiam
has a very modem tone.

114 y^ B, Greenough.

In a number of iambic verses taken consecutively from the remains
of Ennius and Naevius, as they are given in Merry's collection, out
of 1500 ictuses, only about 22 per cent fail to conform to the word
accent, and this counting all cases of verbs compounded with prepo-
sitions, though it may well be that the preposition was at that time
accented, and all cases of a dissyllable at the end of a verse, though
the last verse ictus must have been very weak.

The hymns attributed to St Hilary (t368 a.d.) show 14 per cent
of such non> conforming ictuses. And yet these hymns are the earhest
beginnings of our system of modem poetry.

It can only be that an early accentual feeling of rh3rthm was
partially superseded among the learned by the purely quantitative
Greek rhjrthm ; but in the decline of scholarship or the levelling up
of the lower classes, the old sense of accentual rh3rthm began to
assert itself more and more. This is the view of Keller and many
others ; and see also Klotz, Altramische Meirik,

It seems quite possible that Horace was aware of this sense, and
appealed to it in his two £Eivorite rh3rthms. And no doubt the reason
why the Sapphic afterwards became the only fiavorite one out of all
the proper lyric forms was that it was metrical in prose as well as in
poetry. It is not without significance that Seneca, the tragedian,
has, in a thousand or so verses, not one that cannot be read in the
Christian fiishion. The &ct that a verse preserves the quantities is
no sign that it was not to be read as prose, for many of tiie Chris-
tian hynms are perfect in this respect It is evident from the above
discussion that the Hoiatian treatment of the Sapphic and Alcaic
roust have had a powerful influence on Christian hymnology, and
was a potent fsictor in the introduction of accentual rhythms.

Note. — One thing appears in this investigation which, though not
strictly belonging to the subject, is yet worthy of notice. We have
an explanation of the peculiar Hcense of modem so-caDed iambic
verse, by which we can write : —

From afl' that dweO' bdow" the skies',
Lef the Crea'tor^ praise arise.

Accentual Rhythm in Latin. 115

In Latin iambic verse written according to accent of course no dis-
syllable could properly stand in the first place. But in Christian
poetry there were many such words almost necessary^ such as Christe,
Deus, Pater, Lucis. And these often would naturally demand the
first place. Hence the license came in (perhaps encouraged by the
practice in earlier popular poetry) of allowing such words in the first
place, giving a trochaic accent to the first foot instead of an iambic
one. The frequency of this practice has caused it to be followed in
all Christian poetry, so that in modem times it seems to us as regular
as any other form of the verse.

ON THE OMISSION OF THE SUBJECT-ACCUSATIVE
OF THE INFINITIVE IN OVID.

By Richard C. Manning.

THE aim of this paper is to discover the extent to which Ovid
allows the omission of the subject of the infinitive, and to find
the conditions under which he allows this omission, including the
person and number of the omitted pronoun, the tense and voice of
the infinitive, the position and construction of the word from which
the subject-accusative is to be supplied.

The quotations from Ovid are in accordance with the text of the
following editions : —

P. Ovidius Naso. Ex Rudolphi Merkelii recognitione edidit R.
Ehwald. Tom. I. Leipzig: Teubner, 1888 (Amores, Medicamina
Faciei Femineae, Ars Amatoria, Remedia Amoris).

P. Ovidi Nasonis Heroides. Edidit H. S. Sedlmayer. Leipzig:
Freytag, 1886.

P. Ovidi Nasonis Metamorphoseon Libri XV. Edidit A. Zingerle.
Leipzig: Freytag, 1884.

P. Ovidi Nasonis, Fasti, Tristium Libri, Ibis, Epistulae ex Ponto,
Halieutica, Fragmenta. Edidit O. Glithling. Leipzig: Freytag,
1884.

The arrangement of my material is to some extent the same as
Anton Funck's in his article on ' Die Auslassung des Subjekt-pro-
nomens im accusativus cum infinitivo bei den lateinischen Komikem,'
in Fleckeisen's Jahrblicher, 1880 (121. 725).

In the prose passage in which Seneca (Controversiae, 2. 2. 9)
professes to quote from Ovid (he says, 'haec illo dicente excepta
memini*), there is an instance of omitted subject-accusative after
doUo,

This paper deals only with passages in Ovid where the subject of
the infinitive is omitted ; but the construction is only half understood
when these passages are not minutely compared with those other
passages where the pronoun that is the subject-accusative of the

"7

llS Richard Q Mamnng.

imilEitive might have been omitted, but \s. mt. — oimpazed verb biy
verb, cense bf tense, person br person.

Some of the cases in which the in&nitxve \s osed v i idMmL an ex-
pressed snbiect retpnre litde more than pafftmg oodce.

With impexsonai verbs the Tntlninve, if it meeir invoires an in-
•ie^nite sot^etx and refers to no speriai person or persons, is rcsn-
larty used without an expressed iobject ; there are. moreovm; ^Tith
impexsonai verbs^ many cases whoe it is imposabie to teil whedier
the soi^ect dL the infininve is conceived as detnnte «3r jxdenmte.
Therefiue I sfaaiL in genernL take into considexntion only snrh Jt
these verbs as are the equivalents of verbs of knowing: and the like
(soch as liquet. cinstat\. It is, h o w ever, to be abscrved rhat. ^-rcsi
when the snt^ect of an innnitive which \s. osed in. ainnettin n wrtk an
impersonal verb is not expressed, x prediratc nomi or jifiect i ve
stands in the accasative (or occaaonaily the •iative : jce beiow).
This predicate accnsatxve ]& used by Ovid in the cbilowxng pa^wages^ : —

With est, in phrases like uxtis. ct, at tiigimi, and the Hke;. —

>fet. 5. jaS : '[lumiim est

£5S£ lovis /rairtsm.

Her. 14. 4,1 est mihi soppficxi <3EasaL/iass£ puam.

Ako 3i(fet. ir 283: rx gt3: rj. 319: Fas. 6. 27: Tr. t. 2. 56; 4* c 7a:
4. 6- 3D ; 5. t4- ^ : ^o^- 2. 7. 65 : 4, 6. 32.

With v^catar, — yUex^ i§, 256:

aasoaae vo canir

tMOpene asig afrmi, cpxam (|aad nnt antp. manqQe^

With frdUStat^'-BeT. 14- 7'-

£iS€ ream. gaesasOL. qpam. sc pfaoirwe parmti.

With mout^^ytex. i. 66r :

sed aocetgay >rtB—
AlscSCet. 9. 479.

Widi prjdest.'-^MS. i. 4;^; Met. 11.333.

— Tr. 4- 3' 56-

tempos nbt escc^xo te — vas. aoii vis 3£ardeErx —

Omission of the Subject-Accusative, 119

Other instances are : —

With iuvo, — Met. 13. 964; Fas. 6. 589.

With (Ucet, — Her. 19. 144; Tr. i. i. 3; i. i. 10; Pon. i. 6. 20.

With paenitet, — Pon. 2. 3. 14 :

gratis paenitet esse probam.
Also Her. 7. 134.

With ptget, — Her. 14. 14; Met. 11. 778 (the manuscripts vary).

With pudet, — Met. 7. 617; Tr. 4. 3. 52; Pon. 4. 8. 14.

The accusative of the pronoun, in several of the examples cited
above (e.g. Tr. 4. 3. 56), is the object; a second accusative as sub-
ject of the infinitive is, of course, quite unnecessary in most cases ; it
is, however, necessarily used in the following passage : —

Tr. 2. 70 :

fima lovi superest : tamen hunc sua facta referri

et se materiam carminis esse iuvat.

Instances in Ovid in which a predicate noun or adjective is
attracted to the dative are few in number, but are found with various
verbs. As they are so few, and as the fluctuations of the manuscripts
in some passages between the dative and the accusative are interest-
ing, I quote them all : —

Am. I. 6. 23:

redde vicem mentis, grato licet esse, quod optas.

It is remarkable that here the attraction is caused by an unex-
pressed dative.

Her. 14. 64 : quo mihi commisso non licet esse piae,

Tr. 3. 1 1 . 21 : in causa facill cuivis licet esse disertum.

The manuscripts vary between disertum and disertOy and to some
extent even between cuivis and quemvis.

Tr. 5. 2. 6 : an magb infirmo non vacat esse mihi?

Am. 3. 2. 8 : ergo ill! curae contigit esse tuae?

Met. II. 219:

siquidem lovis esse nepoti

contigit hand uni, coniunx dea contigit uni.
The reading nepotem is also found.

.i*.i6x

IXUB

■ j-

:55

rise:: mr.r'ft iffiaiiy.iiirm jt ~xcsei im:

-5* C

laiun ;■? !

.fcr- r -2^-

:*^- :rw ^fc

V

TT^f'.M 'xsr

TT

-ac

^-. z. >

^n^j IT

unnm

■«m: tie

li» J-

IT "SJOL

Tie :hBC ji^nns^ ji jsef wnnttgn

Omission of the Subject- Accusative, 121

Fas. 4. 804 :

num tamen est vero propius, cum condita Roma est,

transferri iussos in nova tecta Lares
mutantesque domum tectis agrestibus ignem

et cessaturae supposuisse casae,
per flammas saluisse pecus, saluisse colonos ?

Saiuisse pecus is a secondary idea. The last line is equivalent to
per flammas saluisse cum pecore colonos, •

But the passages that form the special subject of consideration are
those in which the subject-accusative is entirely omitted, and of
primary importance among these are those in which the infinitives
depend on verbs (or equivalent expressions) of knowing, of saying,
or other verbs of similar construction, with which the connection
between the infinitive and the accusative is a close one. These are
the examples, therefore, that require particular attention. They
divide themselves into two classes: Class I., comprising examples
in which the (unexpressed) subject of the infinitive and the subject
of the verb on which the infinitive depends refer to the same person ;
Class II., containing the examples in which there is a change of
subject

Class I.

The instances where the omitted subject of the infinitive and the
subject of the verb on which the infinitive depends are the same can
be conveniently classified according to the person and number of the
omitted pronoun.

a. In the first person singular occur : —

Pen. 4. 1.5:

sive trahis vultus, equidem peccasse fatebor .

Met. II. 453:

sed tibi iuro

per patrios ignes, si me mode fata remittent,

ante reversurumy quam luna bis inpleat orbem.

A. A. 2. 171 :

nee puto nee sensi tunicam laniasse, sed ipsa
dixerat.

122 Richard C, Manning.

Met. 14. 844:

quae si modo posse videre

£ata semel dederint, caelum accepisse videbor.

Here the usual reading is videbor, but there is manuscript authority
ioxfaUbor, which Merkel adopts.

Tr. 5. 10. 48 :

quod patriae vultu vestroque caremus, amid,
atque hie in Scythids gentibus esse queror.

The text is uncertain.

Also, with /^i/^r, — Met. 6. 357; Fas. 4. 321 ; Tr. 5. 5. 63.
With confiteor^ — Met. 7. 164.
With profiteoTy — Am. i. 7. 33.

In the first person plural no instances of the omission of the pro-
noun are to be found.

b. In the second person singular occur : —

Am. 3. 14. 15 :

quae £acis, haec fadto ; t2XLtam fecisse negate.

Met. 6. 41 :

neve monendo

profecisse puiesy eadem est sententia nobis.

Her. 4. 176:

addimus his predbus lacrimas quoque : verba precantis
perlegis : et lacrimas finge videre meas.

Rem. 540:

iam quoque, cum credes posse carere, mane.

A. A. 2. 349:

cum tibi maior erit fiducioj posse requiri,

cum procul absenti cura fiiturus eris,
da requiem.

Here Hbi erit fiducia is practically equivalent to some such form
as credes.

Also, with neff>y — Am. i. 4. 70; Met. 2. 693; Fas. 6. 557 (the text is
not certain).

With pu/Oy— Met, 10. 354.

Omission of the Subject- Acaisative, 123

With/tf/^^, — Am. 3. 14. 37; A. A. 2. 591 ; Pon. 4. 3. 19.
With queror, — Fas. 2. 855 ; 6. 473.

There are no instances to be quoted for the second person plural.

c. In the third person singular are found : —

Met. 6. 601 :

ut sensit tetigisse domum Philomela nefetndam,
horruit.

A. A. 3. 733 :

ille, feram vidisse raius, iuvenaliter artus
corripit.

Met. 2. 599 :

dominoque iacentem

cum iuvene Haemonio vidisse Coronida narrat.

Met. 7. 832 :

saepe tamen dubitat speratqxit miserrima/^/^'.

Met. 1 1 . 74 :

utque suum laqueis, quos callidus abdidit auceps,
cms ubi commisit volucris, sensiic^t teneri.

It is also possible to refer teneri, not to voiucris, but to crusy in
which case the omitted pronoun would be id.

Met. 10. 132:

et ut saevo morientem vulnere vidit,

veUe mori statuit,
Velle could also be interpreted as a complementary infinitive.
Also, with negOy — Am. 3. 14. 5 ; Met 14. 151 ; Fas. 6. 336 ; Pon. i. 7. 54,

though nego can in this last place be taken in the sense of ' refuse/
as well as in that of ^ deny.'

With/a/^^, — Met. II. 134; 12.407; Tr. 2. 449; Pon. 1. 1. 51 ; 4.5.31.

With putOy — Her. 18. 32 ; Met. 9. 459.

With sentiOj — Tr. 4. i . 46.

With queroTj — Her. 9. 43.

With proftieoTj — Fas. 5 . 671 .

With iuro, — A. A. i. 425.

With riferoj — Tr. 5. 4. 22.

124 Richard C. Manning,

With credo^ — Am. i. 8. 71.
With indignor, — Met. 1 1 . j^j.

(i. For the third person plural occur the following examples : —

Met. 2. 389:

omnesque dei non posse fatentur .

Am. 1.2. 18:

acrius invitos multoque ferodus urget,
quam qui servitium /errd /aientur^ Amor.

Met. 3. 573 :

Bacchum vidisse negantnt.

A. A. 2. 558:

peccent, peccantes verba dedisse putent.

Her. 16. 261 :

quae mihi non aliud, opa^m. formidare^ locutae,
orautis medias deseruere preces.

There are in all, then, 53 examples of this construction in Ovid,
all but 5 of them in the singular. It occurs with 18 different verbs ;
the ones with which it occurs as many as four times are fateor (14),
nego {g),puto (6), queror (4), and stntio (4). More than half the
examples (27) have the perfect active infinitive (the perfect passive,
on the other hand, is not found at all in this construction). There
are 18 occurrences of the present active and 2 of the future active
(A. A. I. 425 and Met. 11. 453), both of them with the verb iuro,
5 with the present passive infinitive (Am. i. 8. 71 ; Met. 7. 832 ; 11.
74; II. 787; Fas. 6. 473). The wtihs faieor and tugo^ which, as
has just been seen, have a special tendency toward the simple infini-
tive, i.e* the infinitive and the main verb represent actions of the
same person, show also a marked preference for the perfect tense
of the infinitive ; of the 14 instances of the construction ynXh fafeor
10 have this tense, and all 9 of the instances with nego. Closely
parallel to this mode of expression are the common English idioms
* I admit having done so * and ' I deny having done so.' It is
noteworthy that the verbs video and scio^ which are frequently found
in Ovid with the accusative omitted when the infinitive denotes the
action of another subject, are not found in this construction when
the subject of the infinitive is the same as that of the leading verb.

Omission of the Subject-Accusative, 125

This is perhaps because other constructions are often used instead
of the accusative and infinitive with video and scio^ when the accusa-
tive would be the same as the subject of these verbs. The accusative
and infinitive, however, occur with scio in the following passages :
A. A. 2. 247; Met. 10. 427; 15. 142. Examples of the same sort
with video have thus far eluded search.

Class II.

The extent and variety of the examples included in Class II., that
is, those in which the subjects of the infinitive and the verb on which
it depends are not the same, will be most clearly seen by a classifica-
tion according to the construction and position of the word from
which the subject of the infinitive is to be supplied.

a. The omitted subject is often the antecedent of a relative clause,
which more usually precedes the infinitive, but sometimes follows it.
In the latter case the infinitive is usually accompanied by a predicate
word in the accusative (though not at Her. 21. 25 and Met. 2. 393).
It will be seen that the instances in the neuter are particularly
numerous.

( 1 ) Neuter singular : —

Her. 4. 133:

luppiter esse pium statuit quodcumque iuvaret.

Her. 17. 130:

nam mea vox quare, quod cupit, esse neget ?

Pon. 2. 4. 15 :

quod tu laudabas, populo //rz^///.rx^ putabam.

Also, Am. 2. 2. 14; A. A. 3. 720; Met. 3. 417 (bracketed by some
editors); 15. 263; Tr. 3. 5. 28; 4. 7. 11 ; 5. 13. 19.

(2) Neuter plural : —

Pon. 4. 3. 58 :

tu quoque fac timeas, et quae tibi laeta videntur,
dum loqueris, fieri tristia posse puta.

Ibis n 27:

certe ego, quae voveo, superos motura putabo.

126 Richard C, Manning,

Her. 20. 242 :

effigie pomi testatur Acondus huius,
qitae fuerint in eo smptaL, fttisse rata.

The comma is sometimes placed after eo.
Also, Am. 3. 14. 45 ; Met. 4. 476.

(3) Masculine singular: —

Met. 2. 393 :

turn sciet, ignipedum vires expertus equonim,
non meruisse necem, qui non bene rexerit illos.

Fas. 3. 804 :

viscera qui tauri flammis addenda dedisset,
sors erat, altemos vincere posse deos

Tr. 5. 8. II:

vidi ego naufragium qui mersit in aequora mergi.

The manuscripts vary.

Pon. 4. 13. 27 :

nam patris August! docui mortale liusse
corpus, in aetherias numen abisse domos :

esse parem virtute patri, qui firena coactus
saepe recusati ceperit imperii.

(4) Masculine plural : —

Met. 8. 220 :

quique aethera carpere possent

credidit esse deos.

Her. 21. 25:

iamque venire videt, quos non admittere durum est.

(5 ) Feminine singular : —

Met. 9. 724 :

quamo^^ virum putat esse, virumy2;r^ credit lanthe.

Met. I. 586:

quam non invenit usquam

esse putat nusquaniL

A. A. 3. 715:

iam iam vetUuram^ quaecumque erat Aura, putabas.

Omission of the Subject- Accusative, 127

Tr. 4. 3. 17:

non mentitura tu tibi voce refer :
esse tui memoreniy de qua tibi maxima cura est.

The passage just quoted is the only one of all these (i out of 26)
in which the relative is neither nominative nor accusative.
(6) Feminine plural : —

Am. 3. 12. 38:

Protea quid referam Thebanaque semina, dentes ;

qui vomerent ilammas ore fuisse boves,
flere genis electra tuas, Auriga, sorores,

quaec\\XQ rates fuerint, nunc maris esse deas.

d. Sometimes the antecedent, instead of being made the subject
of the infinitive, is incorporated in the relative clause : —

Met. 15. 363:

nonne vides, quaecumque mora fluidove calore
corpora tabuerint, in parva animalia verti ?

Her. 1 . 74 :

quaecumque aequor habet, qtuucumque pericula tellus,
tam longae causas suspicor esse morae.

Met. 15. 133:

inponiqwQ suae videt inter comua fronti
quas coluit fruges,

Pon. 2. 2. 68 :

tempus adest aptum precibus ; valet ipse videtque,
quas fecit vires, Roma, valere tuas,

c. The omitted subject-accusative is sometimes to be supplied
from an accusative used in another part of the sentence as the
subject of another infinitive : —

Am. I. 2. 12:

vidi ego iactatas mota face cxtsotxt flammas
et vidi nullo concutiente mori,

Tr. I. 2. 40:

nesdt in immenso iactari corpora ponto,
nesdt agi vends, nesdt adesse necem.

12S Richard C. Manning.

Pon. 1 . 7. 70 :

Fas. 5. 167:

et mala Ndsamemj quoniam meroisse videtiuv
si non ferre doles^ at menasse dole.

ora micant Tauri septem radianffa flammxs,
navita qaas Hyadas Grahis ab imbre vocaL

pars Bacchum nutrisse patat, pars credidit
Tethyos has neptes Oceanique senis.

In this example it is to be noted that the subject is expressed only
with the second of the two mfinitives, though for die first it can be
easily supplied from die preceding sentence.

Tr. I. I. 19:

vivere wu dices, salvom tamen esse negabis :
id quoque^ quod vivam^ maims habere dei.

Here we have three infinitives^ of which the first, with a subject,
depends on dices : the second, without a subject, on negabis ; while
the third, also without a subject, depends cm die positive idea ex-
presed in dices and implied in negabis,

Tr. 3. 7. g:

vivere tne dices* sed sic, ut vivere nolim,

nee mala tarn longa nostra levata mora,

et tamen ad musas, qoamvis nocuere, reverti,

aptaque in altemas cogere verba pedes.

The pronoun serves as subject-accusative of three infinitives, vivere^
revertiy and cogere^ but the last two are separated from the first by an
interposed infinitival clause with a subject of its own.

Also* Met. g. 259; Pon. i. 5. 77^ 4- 6. 50: 4. g. 99.

d. More firequendy it happens that the omitted subject-accusative
has already been expressed as the direct object of a verb : —

Met. lo. 256 :

oscula dat reddio^ putat.

Pon. 2. 8. g:

est aliqoid spectare deos et adesse patare.

Met. 3. 245 :

et velnt absentem certatim Actaeona clamant —
ad nomen caput ille revert — et abesse quenmtur,
nee capere oblatae segnem spectacula praedae.

Omission of the Subject-Accusative. 129

Met. 15. 851 :

natique videns bene/ii^/a fiatetur

esse suis maiora.

Her. 21. 223:

si me nunc videas, visam prius esse negabis.

Tr. I. I. 14:

neve liturarum pudeat ; qui viderit illaSj
de \zicrimv& facias sentiat esse meis.

Met. II. 438:

quo magis hos novi, — nam novi et saepe patema
parva domo vidi — magis hoc reor esse iimendos.

Hoc is the accepted reading, though there is manuscript authority
for hos.

Met. 15. 278:

et Mysum capitisque sui ripaeque prions
paenituisse ferunt, alia nunc ire^ Caicum.

This is the usual punctuation. If the comma after ire were re-
movedy the construction of Caicum would become undetermined.

A. A. 2. 265 :

adferat in calatho rustica dona puer :
rure suburbano poteris tibi dicere missa.

Pon. 4. 9. 45 :

haec modo te populo reddentem iura videbit

et se decretis finget adesse tuis :
nunc longi reditus hastae supponere lustri

credet et exacta cuncta lacare fide :
TLyxnz facer e in medio facundum verba senatu

publica quacrentem quid petat utilitas :
nunc pro Caesaribus superis decernere grates,

albave opimorum coVLaiferire bourn.

Cemet is a variant for credet

Also Am. 3. 7. 75; Her. 15. 92; Rem. 226; Met. 4. 376; 8. 191;
9. 446 ; 13. 28 ; 14. 162 ; 15. 847 ; Fas. 4. 900 ; 4. 930 ; 6. 238 ; Tr. i. 7. 5 ;
I. 7. 34; 2. 72; Ibis 598; Pon. 2. 2. 125; 2. 6. 24 (bracketed by some
editors).

e. In the following instance the subject-accusative is to be supplied
from an accusative which comes under neither of the preceding cate-

tyy Richard 61 Mannimg.

gories ; that is, it is neither a direct object nor the sabject of an
mfinitiire.

Met. 4. 609:

Acrisius saperesty qui moenibos arceat orbis
Argolicae, contraqne iUmm ferat anna, geniisqae
noa potet esse lovb.

/. The instanrcs in which the sabject of the infinitive is to be
snpplied from a genitive, a dative, or an ablative, are very few : —

A. A. I. 372:

tnin de te oarret^ tom persuadentia verba
addat et insane inret amore wwri.

Fas. 3. 874 :

flebat, at amissa gemini cansorte peridi,
caenileo iundam nesdns esse dec.

Met. 14. 230 :

ille refert [line 223]

invidia sodos praedaeqoe coptdine victos,
esse ratos aamm, dempsisse ligamina ventis.

Fas. 2.419:

constitit et canda teneris blanditnr alummis

et fingit lingua corpora bina sua.
Marte solas sdres : timer afiiit.

Met. 5. 471 :

illo forte loco dHapsam in gorgite sacro
Persefikones zonam stimmis ostendit in ondis
quam sinitil agnovit, tamqnam tunc denique raptam
sdsset, inomatos laniavit diva capHlos.

g. There are many cases, on the other hand, ^ere the sabject of
the infinitive is to Jbe supplied firom a preceding nominative : —

Tr. I. 2. 20:

me miserum, quanti mantes volvontur aquarum!

iam iam tacturos sidera samma putes.

qaantae diducto subsidunt aeqnore vaUis I

iam iam tacturos Tartara nigra putes.

Tr. 3. 7- 40:

sunt tibi apes modicae, cum sis dignissima magnis :
finge sed innumeris censibus esse pares.

A-3i

Omission of the Sudject-Accusative. 131

Met. 6. 442 :

vel sorar hue veniat ! redituram tempore parvo
promittes socero.

Met. 10. 27 :

vidt Amor, supera deus hie bene notus in ora est,
an sit et hie, dubito. sed et hie tamen auguror esse.

Also Am. 3. 6. 36; Her. 11.87; 20.154; A. A. 3. 153; 3.288; Met. 3.
453; 4- 272; 6.583; 7.378; 9-5^? 10.255; 11.546; II. 718; 14.488;
Fas. 2. 405 ; 3. 658; 4. 316; 5. 634 (bracketed by some editors); 6. 434
(also suspected by some editors) ; 6. 704 ; Tr. i . 7. 24 ; 3. 4. 66 ; 5. 6. 25 ;
Pon. I. 3. 88; i. 6. 4; i. 9. 3; 3. 3. 74; 3- 5- 10; 4. 13. 6.

h. It is noteworthy that, in two instances, the omitted subject-
accusative is represented by a nominative in the following clause : —

Tr. 5. 10. 5 :

stare putes, adeo-procedunt tempara tarde.

Met. 6. 269 :

lacrimaeque suonun

tarn sobitae matrem certam fecere ruinae,

mirantem pduisse^ irascentemque, quod ausi

hoc essent superi.

/. The sabject-accosative is sometimes to be supplied from a reli^
tive whose nominative antecedent is unexpressed : —

Am. 3- 9. 46:

avertit vahnsy Eryds quae possidet aroes :
stmt quoqtiey qui lacrimas cotUinmsu ocgant.

A. A. 2. 522 :

dkta erit itse ibras, quam to fortasse videbis :
isse is3M et te £dtt videre puta!

Here the peculiar lepctitkm is rttpooable fof the cooftructioiL

/ Or the xKmunative, tbougjb escpresced at some diMmce^ may be
only implied in the feotence in which the ioftnitive stands : —

lleL 9. 38 :

iUe cauis hauslo sparfh me pulveie palmis
tsque vKffsn lixli-ae lactu fiavesdt hareittr.

c:

::.-• :a: ia»: Jit- ^oi^i J.

-• . *. A -

^ Jr

-,41 -,.1 1%

jhc z :a

ict3e Tiiifr

am r

J- 15: '!«. "'fAi 7. rrt

-/

■wKBt TmacxL icrrcs iSt is imecetusiu 3 oi
IS somemne^ looicaecL '.if on. aitiectxre or -jt tae pamcnne •.cn^

rt r^

J?C»U-

J15I

: Ptm. :- i. 1:

"iiVtm tiieae cases sanud jc :oiii|nieti ^nofic ^vmse an atiiei:tn^ or

Omission of the Subject- Accusative, 133

Pon. I. 9. 8:

ante meos oculos tamquam praesentis imago
haeret, et exstinctum vivere fingit amor.

Also Met. 7. 577 ; 8. 743.

m. The omitted accusative can sometimes be supplied from a
word of slightly different grammatical form : —

Met. I. 162:

sed et ilia propago

contemptrix superum saevaeque avidissima caedis

et violenta fuit : scires e sanguine natos.

Fas. 6. 505 ;

quaerit ab his I no, quae gens foret. Arcadas esse
audit.

A. A. 3. 167:

femina procedit densissima crinibus emptis

proque suis alios efficit aere suos.
nee pudor est emisse : palam venire videmus.

n. In

Am. I. 13. 47 : - scires audisse : rubebat,

and

Am. I. 12. 2 : infelix hodie littera posse negat,

the prominent position of the heroine of the elegy in the poet's
thoughts is more effectively brought out by the omission of any word
expressly referring to her. Somewhat similar instances are : —

Met. I. 242:

ocddit una domus : sed non domus una perire
digna fuit : qua terra patet, fera regnat Erinys.
in facinus wrasse putes. dent odus omnes
quas meruere pati, sic stat sententia, poenas.

Also Tr. 3. 5. 53:

spes igitur superest, £acturum, ut moUiat ipse
mutad poenam condicione loci.

Here, however, the reading facturum is thought not to be free
from doubt

134 Richard C. Manning.

This completes the list of instances in which the omitted suhject
of the infinitive is not the same as the subject of the verb on which
the infinitive depends. Their number is 146. In about .seven cases
out of eight the subject of the infinitive is to be supplied either firorn^
a nominative (which is itself sometimes unexpressed), or from an
accusative which is the subject of an infinitive or the direct object of
a verb.

When these passages are classified according to the person and
number of the omitted pronoim, the result is as follows :

In nine instances the pronoun of the first person singular is omitted, —

Am. I. II. 13; 2. 7. 8; Her. 21. 223; Tr. i. i. 19; i. i. 20; 3. 7. 9;
5* 9- 35 0^^ uncertain) ; Pen. 4. 6. 50 ; 4. 9. 99.

In not one case is the pronoun of the first person plural omitted.
In the second person singular four instances, —

Her. 3. 113; A. A. I. 372; Met. 14. 842; Pen. 4. 9. 45.

For the second person plural only one instance, —

Tr. 3. 4. 66:

V05 quoque pectoribus nostris haeretisy amid,

dicere quos cupio nomine quemque sue.
sed timer officium cautus compescit, et ipsos

in nostro poni carmini nolle puto.
ante voUbcUiSy gratique erat instar honoris

versibus in nostris nomina vistra legi.

Even here it is possible to supply a verb of the third person from
suoy but because of the preceding verb hcuretis and the following
voUbatis such a construction seems less natural.

The remaining 132 instances are all of the third person, 27 in the
masculine singular, 36 in the feminine singular (a notably larger
number than in the masculine), 19 in the neuter singular, 19 in the
masculme plural, 11 in the feminine plural, and 20 in the neuter
pluraL It was found impracticable to distinguish between persons
and things, or between things animate and things inanimate, as, in
the Metamorphoses especially, the dividing line between these cate-
gories is often far from clear.

There are also a number of passages where an indefinite id^ refer-
ring not to a word but to a clause, is to be supplied as the subject of

•

Omission of the Subject- Accusative, 135

an infinitive, and other passages where the infinitive is impersonal,
and no subject is to be supplied : —

A. A. I. 316:

adspice, ut ante ipsum teneris exultet in herbis !
nee dubito quin se stulta decert putet.

Tr. 4. 9. 4 :

fac modoy te pateat paenituisse tui.

Met. 2. 788:

successurum(^t Minervae

indoluit.

Snccessurum is the accepted reading, though the manuscripts differ.
Tr. 5. 7. 26:

carmina quod pleno saltan nostra theatro
versibus et piaudi scribis, amice, meis.

Met. 15. 388:

et genus omne avium mediis e partibus ovi,
ni ^oxtXfieriy quis nasci posse putaret?

Fas. 3. 168:

si licet occultos monitus audire deorum
vatibus, ut certe fama licere putat.

Also Am. 3. 7. 46; Her. 5. 40; 17. 34; A. A. i. 426; i. 637; 2. 592;
Met. 3. 639; II. 788; 14.725; 15.278; Pon. I. I. 58; 1.9. 24; 2. 5. 16;
4. 5- 13-

For the sake of convenience a summary is made of the number of
times the various pronouns are omitted : —

Pronoun OMnrED. Class I. Class II. Total.

1st sing 10 9 19

1st plur o ' o o

2d sing 14 4 18

2d plur o I I

3d sing 24 82 106

3d plur ^ JO J5

53 146 199

The classification according to the voice and tense of the infinitive
is as follows — the classification being made according to the meaning

136 Richard C. Manning.

of the form, so that deponent verbs are regarded as actives, and
meminisse is grouped with the present infinitives : —

Mood and Tense. Class I. Class II. Total.

pies, act 18 81 99

perf. act 28 26 54

fat. act 2 II 13

pres. pass 5 10 15

perf. pass o 16 16

pass, periphrastic . . o 2 2

53 146 199

Esse is not expressed in any of the 13 instances in which the
subject of the future active infinitive is omitted ; in 11 of the exam-
ples we find the simple future participle ; and fore in the other two
(Met. 9. 724 ; A. A. i. 425).

In the perfect passive infinitive the unaccompanied participle is
found in 1 1 of the 16 cases, esse is expressed four times, and/uisse once.
In the five passages in which the full form is used, the infinitival
element always occupies the same position in the verse, which is, in
one instance, the hexameter, in the other four, the pentameter : —

Her. 21. 223:

si me nunc videas, visam prius esse negabis.

Fas. 3. 874 :

caeruleo iunctam nescius esse dea

Fas. 4. 930 :

adstrictum longa sentiat esse mora.

Tr. I. I. 14:

de lacrimis factas sentiat esse meis.

Tr. I. 7. 24:

pluribus exemplis scnptA/uisse reor.

The same rule holds true in one of the two instances where esse is
combined with the gerundive : —

Met. 1 1 . 438 :

parva domo vidi — magis hoc reor esse timendos.

The other example has esse in the fourth foot : —

Tr. I. 7. 34:

si praeponendos esse putabis, habe.

Omission of the Subject-Accusative, 137

We have seen that in the compound forms of the infinitive, esse
and the subject-accusative are often omitted, and the participle only
is expressed ; in a similar way we, in several cases, find a predicate-
adjective or predicate-noun, in the accusative, standing alone, without
infinitive or subject-accusative. The construction with two accusa-
tives and no infinitive is plainly the original one after verbs of know-
ing and the like ; but it is impossible to tell how far the Romans of
the classical period realized this, and to what extent they thought
that such expressions involved the ellipsis of an infinitive. The
passages concerned are as follows : —

Fas. 6. 335 :

dubium, nymphamne putant

an scierii Vestam,

Am. 3. 7. 83 :

neve suae possent intaciam scire ministrae.

Her. 16. 309:

nee, si bona msLgnzpufaret,

quae tenet, externo crederet ilia viro.

A. A. 2. 597 :

ista viri captent, si iam captanda putabatU,

Also, with/^/^, — Am. 2. 8. 14; Met. i. 502; 7. 69.

With credo^ — A. A. 2. 287; Pen. 2. 7. 11.

With reoTj — Her. 5. 145 ; Met. 4. 674.

With experior, — Ibis 322 :

quosque putas fides, ut Larissaeus Aleuas
vulnere non fidos experiare tuo.

With dicoj — A. A. 2. 646:

omnibus Andromache visast spatiosior aequo :
unus, qui modicam diceret, Hector erat.

With faieor, — Met. 4. 736 :

gaudent generumque salutant
auxih'umqne domus servaforemqae fatentur
Cassiope Cepheusque pater.

Also Met. 12. 596.

With profiteor^ — A. A. i. 181 :

primisque ducem profitetur in annis
hellaque non puero tractat agenda puer.

138

Richard C. Manning,

With queror^ — Her. 7. 30 :

non tamen Aenean, quamvis male cental, odi ;
sed queror infidum questaque peius amo.

With mentiory — Met. 9. 707 :

iussit ali mater fuerum mentita.

With fingpy — Rem. 504 :

qui potent sanum fingere^ sanus erit.

The following is an alphabetical list of the verbs and equivalent
expressions with which Ovid, as shown in quotations and citations
above, omits the subject of the infinitive. The number of instances
that occur with each verb is also given : —

argue . .

•

I

(Pon. 4. 6. 50)

audio . .

•

I

(Fas. 6. 505)

auguror . .

I

(Met. 10. 27)

cemo . .

I

(Met. 4. 376)

comperio

I

(Pon. 3. 2. 82)

confiteor

I

(Met. 7. 164)*

crcdibile est

I

(Tr. 2. 72)

credo . .

. 10

dico . . .

. 6

doceo . .

I

(Pon. 4. 13. 27)

doleo

I

(Pon. I. 7. 70)

£&teor . .

. 18

fcro . . .

2

(Met 15. 278 ; Fas. 6. 434)

fiducia est

I

(A. A. 2. 349)

fingo . .

6

gaudeo .

2

(Met. II. 546; Fas. 4. 900)

indignor

I

(Met. II. 787)

infitior .

I

(Pon. I. 3. 88)

iuro . .

3

latet . .

I

(Pon. 4. 13. 6)

liquet

• 5

loquor .

I

(Her. 16. 261)

meminl .

I

(Fas. 6. 238)

mentior .

2

(Met. I. 614: 8. 251)

miror

• 3

minim .

2

(Met. 6. 583; II. 731)

oarro

I

(Met. 2. 599)

ncgo. .

. 17

nescio

I

(Tr. I. 2. 40)

Omission of the Subject-Accusative.

139

nesans .

•

2

(Met. 7. 3»o; i-as. 3. 874;

probo . .
profiteor
promitto
puto . . .

I
2
I

45

(Tr. 3. 5. 28)

(Fas. 5. 671; Am. i. 7. 33)

(Met. 6. 442)

22 % of the whole number.

queror .
refero

8
• 3

reor . .

6

scio .

8

sentio

• 9

sors est

spero

spes . .

statuo

suspicor

testor

video

I
2

2
2
I
I

• 13

(Fas. 3. 804)

(A. A. I. 470; Met. 7. 832)

(Fas. 6. 393 ; Tr. 3. 5. 53)

(Her. 4. 133 ; Met. 10. 132)

(Her. I. 74)

(Her. 20. 242)

200

Deduct ... I

for passage inserted twice (A. A. 2. 171).

Total

•

•

•

. 199

.

All the passages that have, as yet, been taken into consideration
are such as are used with verbs of "knowing," of " saying," of "per-
ceiving," or other similar verbs, with all of which the connection
between the infinitive and the accusative is a close one. With many
verbs, however, the accusative has no such close union with the in-
finitive, and differs only to a slight extent from an ordinary object-
accusative. Verbs of this latter class are readily and often used with
the infinitive and no accusative. Here, however, no attempt must
be made to treat this topic ; we must rather turn to the consideration
of passages in which Ovid combines the infinitive with a predicate-
nominative, where in prose a^ predicate-accusative would have been
the natural construction.

With verbs indeed that. are followed sometimes by the accusative
and infinitive and sometimes by the complementary infinitive, it
causes no surprise when a predicate-nominative appears, — as in the
following examples with gaudeo and despero : —

A. A. I. 345 :

quae dant quaeque negant gaucUnt tamen esse rogatoi.

140 Ric/tard C. Manning,

A. A. I. 295:

Pasiphae fieri gaudebat adulter a tauri.

Fas. 5. 241 :

cur ego desperem fieri sine coniuge mater
et parere intacto, dummodo casta, viro.

Ovid also combines prccor and posco with the predicate nominative
and infinitive, somewhat after the analogy of verbs of wishing : —

Her. 5. 158 :

sed tua sum tecumque fiii puerilibus annis
et tua^ quod superest temporis, esse precor.

Pon. I. 7. 6:

ecquis in extremo positus iacet orbe tuorum,
me tamen excepto, qui precor esse tuus.

Met. 8. 697 :

esse sacerdotes delubraque vestra tueri
poscimus.

In connection with these should be cited : —

Fas. 4. 381 :

dux mihi Caesar erat, Sub quo meruisse tribunus

glorior;'^

for the principle of construction with glorior is practically the same
here as that with gaudeo, A more remarkable example is : —

Pon. I. 5. 66^:

hoc, ubi vivendum est, satis est, si consequar^ arvo
inter inhumanos esse poeta Getas.

Sometimes the nominative and infinitive is used with verbal phrases
containing a noun, such phrases being in meaning equivalent to verbs
that govern the simple infinitive : —

Am. 2. 19. 14:

A! quotiens finxit culpam, quantumque licebat.
insonti, specietn praebuit esse nocens.

Speciem praebuit = visa est or simulavit.

Am. 2. 4. 14:

spemqMe, dot in molli mobilis esse toro.

1 Cf. Hor. Epod. 11. 23. 2 qi Qc. de. or. I. 150.

Omission of the Subject-Accusative, 141

DifTerent from these and much more difficult is : —

Tr. 2. 10:

deme mihi studium, vitae quoque crimina demes,
acceptum refero versibus, esse nocens.

Here the metre is largely responsible for the nominative, as, indeed^
it is to greater or less extent in all the examples.

A very strange construction, and unjustifiable except on metrical
grounds, is found at

Tr. 4. 3. 51^:

me miserum, si turpe putas mihi nupta videri.

Specially interesting are the instances where verbs that usually
govern the accusative and infinitive are found with an infinitive and
predicate-nominative, evidently in imitation of the Greek. This
nominative is sometimes a noun or adjective, sometimes the participle
contained in the infinitive. This is not the place to discuss the
origin and character of the construction, as Ovid yields one example
only for each of these classes. They are, with predicate noun in
the nominative : ' —

Met. 13. 142:

sed enim, quia rettulit Aiax
esse lovis pronepos, nostri quoque sanguinis auctor
luppiter est.

With the perfect participle in the nominative (and esse unex-
pressed)': —

Met. 9. 546 :

pugnavique diu violenta Cupidinis arma
effugere infelix, et plus, quam ferre puellam
posse putes, ego dura tuli. superata fiateri
cogor, opemque tuam timidis exposcere vatis.

^ But compare Lucan. 9. 1038 :

utque fidem vidit sceleris tutumque putavit

iam bonus esse socer, lacrimas non sponte cadentis

eflFiidit

« Cf. Cat 4. 2; Verg. Catal. 8. 2; Hor. Ep. 1. 7. 22; Verg. Catal. II. 24;
Prop. 3. 6. 39.

• Cf. Verg. Aen. 2. 377.

LATIN ETYMOLOGIES.
By J. B. Greenough.

I. Auctor,

THE ordinarily received course of development of auctor from
augeo has always seemed to me contrary to what we know of the
methods by which words are formed. In Lewis and Short's {^Harper's)
Dictionary, we read under auctor, * he that brings about the existence
of any object or promotes the increase or prosperity of it, whether
he first originates it, or by his efforts gives greater permanence or con-
tinuance to it ; to be variously translated,* etc. This is a very good
description of the meanings of the word, but it is evidently intended
to suggest a course of etymology from augeo which cannot be right.
In the great revision of Forcellini by De Vit, we have, in like manner :
proprie significat qui auget, h,e, gignit ac producit. But we can hardly
suppose the ideas expressed by the word can all come directly from
the verbal idea of increase. Not that the two sets of ideas are in-
consistent or far removed from each other, abstractly considered, but
they do not correspond in any tangible direct sense, such as is neces-
sary for the development of a word so old as auctor. •

ITie word-making process moves in lines of thought often frivolous,
superficial, ' unessential,* but always obvious and immediately appre-
hensible. When you find an abstruse and profound connection of
thought between two words, one which appeals to the sense of the
true inwardness of things, you may be sure it is not the right one.
Many etymological notions resting on associations of ideas that fit
very well with the essence of things are erroneous. As a fact, slang
is the only example we have of the real growth of words.

The names of Peelers and Bobbies applied to policemen, both
derived from Sir Robert Peel's constabulary reforms in Ireland, and
even the regular name oi police, — really the name for the sanitary

143

144 ^ 3- Greemrngk.

cleaning: of cities and anmes^ — indicate the kind of associations hf
which new words are formed and deveiopecL These aasociatzons aze
constantiy distorted in popular language, and twist words into coeon-
ings which are very far from the original etrmon.

The methods by which a philosopher, ^is stuA^ — for everr philos-*
opher is a mere man tar more than he is a piiilosopiier. — wotdd
make v¥ords» are precisciy the methods by winch words are not made.
.\nd it is just the popular, tranaent, ami superricial associations rhnr
must be followed in the tracing ot the etymologies oi words, wiien
the origin of the words is so remote tiiat the actual associations have
been forgotten.

The word Imcker in Eogiish, for instance, setfins at rirst sigiit to be
iounediately connected widi the idea oi some one l^tktnd as a. sup-
porter, and probably in many cases is used widi that idea ; but in tact
It comes through one or two much less abstract a^ociations of ideas.
meaning in some uses originaiiy the endorser or man who * backs '
( writes iiis name on the back oi\ a ooie, and in some 'oses one who
puts ills money on a card. These two streams have 'inited in die
word. Is there, then^ any way in which auctar^ with its muitx&noos
meanings, can come ax>m aui^ro /

Now. in the case oi this word* we have a complete set of ihe
^formations, aucto^ auctus^ .luitar. In what i>alpable and natnrai
sense could auctor origmally tiave meant an imrta^icr^ so as to »igree
wnh the verb of the set? The suggestion is found in the kindred
word auctw. If auctw, from being an 'enhancement' lot the price),
comes to mean an * auction sale,' auctor may well be the agent ox
the same acdon. In such a proceeding who would be the t-mkamcr ^
VVho enhances the price at an auction ? We think at '>nce of the
buyer, but there are many indications that the opposite meuiod. often
in vogue nowadays, was the original one. * Shall I have twenty-iive? •
the auctioneer says. If any one nods or makes any intelligible s^n,
the auctioneer <:ontinues* * Twenty- live ; who'll give thirty? *

In ancient times the bidding ^vas regularly made by a sign. See
digtia iicert^ as in Verr. 2. 3^ 27, and tiH^rt iiigtium^ Ibhi. 2. r. 141.
See also the story m Stiet. CaL 39, where CaiiguU piayeil a joke on
Apontus Satuminus by causing, a worthless lot oi property to be
knocked on to him, because, being asleep it in auction, he con*
tinned to nod vraxi^ che auctioneer raised the pnce.

Latin Etymologies. 145

So, again, in the locus classicus, Stichus, 193 :

Haec verba subigunt me uti mores barbaros
Discam atque ut faciam praeconis compendium.
Itaque auctionem praedicem ipse ut uenditem.

(221) Logos ridiculos uendo. Age licemini,

Quis cena poscit ? Ecqui poscit prandio ?

That is, the seller asks for a price, and if that is offered, he immedi-
ately raises it. Of course if the price is not offered, he comes down,
as in this case ; but the usual custom doubtless was to raise the
price.

If this view be correct, the auctor would be originally the selUr at
auction^ which was the common method of selling at Rome. This
auction was apparently managed by the seller himself (cf. mores bar-
baros, above) ; so that the two words auctor {seller) and auctio {sale)
would be a natural pair. And in fact auctor is regularly used in
juristic language for vendor. See, P. Caesennius^ auctor fundi,
Caec. 10, and also Quaero an pila quae ab auctore domui coniuncta
erat ad emptorem quoque iure emptionis pertineat (Dig. 19. i. 52,
quoted from Scaevola) .

The derivative auctoro, with the distinct meaning of sell, and its
derivative auctoramentum, price, strongly confirm vendor as the
original meaning of the word. Cf. est enim illis {mercenariis) ipsa
merces auctoramentum servitutis, Cic. de Off. i. 150.

The first step in the wider development would be the use of such
expressions as, turn illi inteHexerunt se id quod a malo auctore emis-
sent diutius obtinere non posse, Verr. II. 5. 56. Here an immunity
has been purchased by the Mamertines from Verres, and the pur-
chase is not secured to them. So the idea of guarantor in reference
to title becomes attached to the word ; cf. bonus auctor ^ gravis auctor,
used in a figurative sense, a reliable guarantor, a good authority, Cf.
non si mihi Juppiter auctor spondeat (Aen. V. 17), where, as so often
in other cases in Virgil, the original meaning shines through. Closely
akin are certissimus auctor (Geo. I. 432), non futtilis auctor (Aen.
XI. 339), Italiam petiit fatis auctoribus (Aen. X. 67). In fact, from
this use comes the very general use of the word in all connections
as an authority, the commonest meaning of the word, from which the

146 y, B, Grecnoiigh,

abstract auctoriias gets its meaning. Closely akin, though perhaps
assisted by the use in the next stage, is the meaning adviser, where
really there is no clear idea of a responsible party.

The meaning of originator and the like naturally comes through
auctor legis, which means properly th6 proposer, that is, the authority
of a bill; in other words, the voucher for it before the people.
With this official use belongs auctores fieri, used of the senate,
originally an action preceding the offer of a measure for the popu-
lar vote. In this manner of development all the various uses attach
themselves to the original idea without any forcing. It is curious
enough that the ordinary grouping of meanings under the word
exactly reverses this natural development ; so that if it had occurred
to anybody to read the article in Harper's Dictionary, for instance,
backwards, he would have had the whole history of the word admira-
bly set forth.

II. Opto,

Every one must have noticed the tendency that Latin writers have
to use opto in a religious or semi-religious sense. Thus, we find con-
stantly such expressions as a dis immortalibus optabo, Cic. Cat. 2. 7.
15 ; equidem tibi bona optatn omnia. Plant. Rud. 639 (this follows a
jocose adjuration). So also :

Di tibi omnes semper omnia optata offerant, Ter. Ad. 978 ;
Di tibi dent quaequomque optes, Plaut. Mil. 1038 ;
Cupimus optamusque, Cic. Phil. 14. i. 2;
Dari votis optat aprum, Ae. IV. 158.

Very commonly it gives the sense of the English hope and pray.

This tendency suggests that the word may have once had a quite
different meaning. It seems not unlikely, at any rate, that the
primitive ^opio (see praedopiont, praeoptant, in Festus) is akin to
ops, opus, and opera, and meant serve, with a special religious sense,
either very early developed, — in which case we may compare San-
skrit apas and apas, both meaning religious ceremony, — or else later
acquired, as is the case with facio, pkl^vi, and others. We may com-
pare the technical use of operor as worship. So operari est deos
religiose et cum summa veneratione sacrificiis litare, Non. 523. 9.
Cf. Virg. Geo. I. 339.

Latin Etymologies, 147

The transfer from serve to pray is natural enough, as we see by
the acquired force of veneror^ as in si veneror stultus nihil horum^
Hor. Sat. II. 6. 8, and often elsewhere in the same sense.

The primitive \opio must have very early become specialized in
the meaning of choose^ as we see by opHo, This development might
naturally come from some such association as is found in the three
wishes of fairy lore, which are of course really alternative prayers
granted by some superior power. Cf. Theseo cum tris optationes
Neptunus dedisset, Cic. Off. 3. 25, 94.

The whole group, though obviously thus connected, does not give
any natural analysis in reference to the sequence and relationship of
its various aspects. The idea of service ^ however, certainly runs
through all the members. There are many phrases with opus where
it is parallel with the English word use, i.e. service, as Nil opus est,
its of no use ; atque haud sciam an ne opus sit quidem nihil unquam
omnino deesse amicis, Lael. 14. 51 (entirely equivalent to expediat).
So also opera, whatever its exact relation to opus, often has this same
sense. Thus operant dare is almost render service ; non operae est
often means /'/ isn't worth while, its of no use. The use of operae,
laborers, is of the same character, properly services, like the English
help.

The derivative operor runs in the same groove. The simplest
representative of the common root, ops, is the most puzzling in its
various meanings. The idea of seroice is apparent enough, but it is
certainly highly colored with various other religious ideas, as is seen
from the personification of the notion as a divinity, taken in connec-
tion with the obscure rites by which the divinity was worshipped and
the various identifications which she underwent. She may have been
a corn-demon, or a personification of worship or of some other prim-
itive idea or act. But without more knowledge of the religious
usages connected with the word, it is impossible to analyze the asso-
ciations that belong to it in its common use. Perhaps in time, in
the increased attention to early religious ideas and rites, some solu-
tion may be found.

III. Exerceo, Exercitus,

The development of the ordinary meaning of exercitus has gener-
ally, it seems to me, been wrongly traced, or perhaps better, assumed.

t4& "jF- B. GreeTwugk.

We happen to have in English words which suggest a more natnml
association of ideas for the wonL

^ A tr can-band captain eke was he
Of fiunous London town,*^

means that John Gilpin was a captain of militia. So the 'May
training' of New England was the meeting of the militia for mili-
tary exercise. In my childhood I only knew soldieia as * trainers.^
Following this analogy, exercitus as the abstract of exerceo might well
mean ' /^ training,' i.e. the army of the city assembled, fibr military
exercise, as in English ' a nmneroos following/ for instance, gives a
concrete sense to an abstract word. This view in a somewhat dififtf-
ent form has also been held by others ; see KeHer, Lateinisclu Ety-
mologi^n, p. 52.

The underlying notion of exerceo is however less obvious. The
group of words to which it belongs has a very clear fundamental
conception, though the notion is not easily expressed in English.
The idea of * setting a barrier/ either from outside in, as it appears
in coerceOy or from inside out, as in arceo and arx^ or perhaps both
ways, as in area, is plain. But what can exerceo ^ meaning * exercise,
train,' keep in or off? A suggestion is given by the peculiar use of
the Greek iKirovtiy and itcwovutrBau

An example from the Education of Cyrus (Cyr. i. 2, 16) shows the
notion that the Greeks attached to these words :

Ka2 Kw 8i In ififiiyu fAoprvpCoL koI t^ furpuK hairrp avrStv #au rov
ii(ww€ur$ai rrpf iuarav, Aurxp^ H^ 7^ In koI vvv ion TLtpaai^ #oqu
T^ irrvtiv Kid r6 &wofivrr€frOai tcai to <t>wnp fitarois ^cuycoAu • ajurxpay
a i<m Kol TO lovra troi <^v€pay ytvttrOai, ^ rov avp/qtrak creica rj mu oXAov
nvo^ rouAno\). Tavra 8^ ovic &v cSwavro inubf d fitf kcu. Suurrf /urpiq.
iXp*ovro Kol rb vyphv iKWOvwyr€^ &inqXurKOv wotc oAAj; irg dxo;(0)-

p€iV.

The same association of dryness with physical strength and endur-
ance existed also among the Romans. The words siccus and siccitas
are regularly used of men ' in good condition ' or * in training,' as
the modem athletes say. Thus, in describing the robust old age of
Masinls«a, Cicero says : summam esse in eo corporis siccitatem (LaeK
34). So Nonius explains sic cum by exercitum^ and cites Varro as

Latin Etymologies, 149

saying, evidently in allusion to the Greek passage quoted above,
Persae propter exercitationes pueriles modicas earn sunt consecuH
corporis siccitatem ut neque spuerent tuque emungerentur (Non. 4.
26). So also Lucilius in the same passage is represented as say-
ing, cum studio in gymnasio duplici corpus siccassem pi/a, Cf.
Corpora graciliora siccioraque, Plin. 34. 19. 15. In fact, the idea
is a very old medical one, of which our word humor preserves a
souvenir. The general currency of this idea suggests that exercere
was originally exercere umorem^ i.e. siccare corpus; in other words,
exercise. The thing that was kept off was the ' peccant humors of the
blood.' The use of the word in the later sense becomes more natural
when we consider the ease with which the object of a verb used in a
special phrase is dropped. One might hear a New England girl say,
"No, I don't play (the piano) in public, but I take (lessons) of
Professor A, and I practise (playing?) a great deal." Nor need we
go farther than the Latin, in which we have putare (rationes), facere
(rem divinam), and numerous other verbs used absolutely ^ as we say.
At first sight it seems an objection that exercere is not merely used
absolutely but takes a personal object {exercere se, exercere tironem).
But nothing is more common than such a shifting of the relation
between a verb and its object. The word defendere has undergone
a change precisely similar. From meaning ward off it has come to
be used very naturally as defend^ a change of the same kind and in
the same direction as the one supposed in exercere. In fact the
change from defendere iniuriam to defendere vitam is on the whole
greater than that from exercere umorem to exercere se.

This explanation seems a more natural one than that given by
Keller in the discussion cited above. The borrowing of a Greek
word like dcncccu seems very unlikely when we consider the rest of
the compounds of arceo and its kindred nouns.

ON mlpap kUaOai (S 501) AND THE MANUS
CONSERTIO OF THE ROMANS.

By Frederic D. Allen.

Ilcioap *rope* and vtlpap 'end* — these words (or meanings, if
you will, of the same word) the expounder of Homer has much ado
to apportion justly. But of several crucial passages involved, the
most difficult is certainly that in the description of the Shield of
Achilles, 2 497 ff. : —

Aoot 8* €iv &yoprji iauv aSpoot ' ivOa 3^ v€lk<k
iaptop€i ' Svo 8* aySp€9 ivuKtov ctvcxa ttocv^s
&vhpo^ AiroKTafievov ' 6 ftcv €V)(€to iravr airoSovvaif
500 8i;/LUiH TTi^vcricaiv, 6 8* avoLVtro firfSkv IKiixOai '
dfi<f>(o 8* liaOrp/ iirl iCTOpi iriipap iXiaSai '
Aaoi 8* afi4^or€poi<nv iTrqirvov, ofufn^ dpcuyot.

Almost every one resorts to the meaning * end.* 'Ilcpas Xafitiv, * finire
litem,* * take an issue,* * accept a decision,* ' obtain consummation,*
*Ziel d. i. Entscheidung gewinnen* are translations given. The
ancients usually took laropi as pAprvpi (' by the aid of a witness *),
explaining that to rcA.09 r^ &ki/s oi SiKcurrot iriOevro coorc rov irapi-
Xovra fmpTvpiav vlkSlv (Schol. A). Modem scholars oftener under-
stand it as ' umpire,* ' referee * ; it certainly has this sense in 4^ 486.

But whichever meaning of lUrropi i^e adopt, the interpretation is
forced. Supposing for the moment that v€ipap can mean * decision,*
still i\€<rO€u is not an appropriate word.^ It implies, especially as
reinforced by UaBrfVy much too active a taking.^ Ilcipap Ikco^ would
be a more suitable phrase. It would perhaps be captious to ask how

^ Zenodotus wrote dp4a$ai here (or was it in 500?), but this affords no relief.

^ This is not disproved by the foregoing verse, where, as in 367, ^ 297,
ta 334, i\fff$ai is idiomatically used (like toucher in French) of the receipt (tak-
ing possession) of moneys due. The meaning is probably * he refused to lay
hands an a single sheep.' I incline to accept Leafs exposition of this passage,
6,9alprro being best taken as = ' recusabat.'

I5«

152 Frederic D. Allen.

a longing to ' accept a decision ' on the part of two wranglers coold
find expression in sculpture. It is, however, a pertinent inquiry
whether just this would be the ' eager longing ' of a man engaged in
a public quarrel, backed by a crowd of shouting adherents.

These difficulties must have occurred to others, and I do not
doubt that many have felt what only Doederlein (Homer. Glossar.,
vol. II, p. 138) has spoken, — that these words naturally mean
* seize the rope in presence of an umpire.' But what rope ? A meta-
phorical rope, Doederlein answered. Rope-pulling, he thought, was
used as a figure for a judicial contest, just as it is elsewhere in Homer
for a battle ; * seize the rope,' meant * begin the trial.* A harsh and
abrupt metaphor, surely. I should prefer it to the usual explanation,
but I cannot think it satisfactory.

But need the phrase be figurative? I propose, with diffidence, to
take it literally. In support of this I can offer only an hypothesis,
based on rather remote analogies. It makes no claim to be ccHisid-
ered as a proof. I would translate verse 501 * and both were hasten-
ing to grasp the rope before the umpire.' The rope-grasping I
conceive as a symboUcal act, tjrpifying an actual bodily contest, and
preliminary to a trial before judges, — analogous, in short, to the
ceremony called manum conserere among the Romans.

Pulling-matches must have been familiar to the Homeric Greeks.
The conception of a battle as a rope-puUing between Trojans and
Achaeans, with gods at the extremities of the line, meets us more than
once in the Hiad.^ Of course some actuality fiimished the basis for
this metaphor. It may have been a mere game or gymnastic exer-
cise. In later times there were such games. In that called SicAjcv-
crriV&i, mentioned by Pollux (IX, 112), Svo \ua^iax muSwy curtv cXxov*
otu rovs cr^K>v9 oi ci-cpou A rope is not mentioned here, but in
another variety of the game described by Pollux (IX, 116), Photius,
Hesychius and Eustathius,^ under the name of frtmx^&ajf c\jcetv, a
rope is passed through a hole in a post, and two men, back to back,
pull on the ends. A simple ** tug of war " between two youths, with
a short rope provided with handles, is depicted on a gem in Flor-
ence (Krause, Gymnastik und Agonistik, plate VI, i*).

But this game may once have been more than a game. Organized

1 H 102, and elsewhere. Sec below, p. 165, * On P 3S9 (p. nil, 24.).

On irelpap kXeaOai and Manns Consertio, 153

pulling-contests for the decision of controversies are not unknown in
primitive society. It will be remembered that Zeus proposes such
an one at 19, to settle the question of superiority between himself
and the other gods. The trial suggested in this famous passage is
only a " tug of war " with the advantage of gravity on one side. I
give such other instances as I have been able to come by; it is very
likely that others exist and can be pointed out by those versed in
folklore. In the Pali version of the story of Solomon's judgment the
" Future Buddha " (Bodhisatha) finds two women quarrelling about
the possession of a child. He proffers his services as umpire, and
they agree to abide by his decision. " Then he had a line drawn on
the ground, and told the Yakshini to take hold of the child's arms
and the mother to take hold of its legs, and said ' the child shall be
hers who drags him over the line.* " ^ Mr. James Deans * relates
that in Hoiduk Land, Queen Charlotte's Islands, Indians of different
tribes used often to unite in killing a whale. Possession was then
decided in the following manner. Cedar-bark ropes were fastened
to head and tail. " When all was ready, at a given signal, every man
pulled with might and main, the representatives of each tribe by
themselves pulling in a different direction to the others. The tribe
who pulled the whale furthest were the victors. . . . The winning
party took the spoil home to their village . . . ; the losing party
good-naturedly started for their homes, well knowing it might be
their lot to be victorious in the next pull they had." I suspect from
the context that " pulling " here means rowing, but the principle is
the same. A curious pulling-match is reported from among the
Eskimo. The community divides itself into two parties, the " ptar-
migans," those who were bom in the winter, and the " ducks," or
children of summer. "A large rope of sealskin is stretched out. One
party takes one end of it and tries with all its might to drag the
opposite party over to its side. The others hold fast to the rope and
try as hard to make ground for themselves. If the ptarmigans give

1 T. W. Rhys Davids' " Buddhist Birth Stories," vol. I, pp. xiv-xvi. I am
indebted to Professor Lanman for this instance, and to Mr. W. W. Newell for the
three following. Dr. Hayley remembers an account, in the narrative of an
Afirican explorer, of an organized pulling-match for the possession of a woman, in
which the woman sustained such injuries that she died; but he is unable to
famish the reference at present.

' American Antiquarian, vol. X (1888), p. 42.

154 Frederic D. Allen,

way, the summer has won the game, and fine weather may be ex-
pected to prevail through the winter." ^ This is perhaps little more
than a game, though it appears that a religious significance is
attached to it.* More, to the point is the method by which the
Passamaquoddy Indians of Maine decided between rival claimants of
the chieftainship. "If there were two candidates, the matter was
decided by the candidates* joining hands over a mark drawn between
them, their adherents forming two lines by each clasping his arms
round the waist of the one in firont of him. The party which suc-
ceeded in pulling the opposition candidate across the mark had the
right to elect the chief.'*'

We have here pulling with and without a rope ; and we have — it
will be further observed — both the pulling of the object in dispute,
and a pulling pure and simple to decide an abstract question. These
puUings are no more absurd than the old European trial by combat ;
in fact they are simply mild and bloodless forms of such a trial. We
might suppose our Homeric rope-seizing to be an actual puUing-
match for the settlement of the dispute, did the description end with
the passage I have quoted above. But there follows a fiirther de-
scription, apparently representing a new relief on the shield, which
shows us a later stage of the same proceedings. Here a court of
justice, consisting of -^ipovrviy is trying the cause. The description is
not very clear, and the exact meaning of several phrases is in doubt.
But it is certain that no such rudimentary justice as rope-pulling is
there dispensed. I have therefore imagined the rope-seizing to be
symbolic, — a purely formal contest, to furnish the needed ground
for a magistrate's interference. For this we have the significant
analogy of Roman judicial procedure.

The theory of the Roman legis actio was, as Maine has convincingly
shown, that a magistrate must see a quarrel going on before he could
interfere. When litigants appeared before him, therefore, the first
step was the enacting of a sort of mimic contest, called vindication

1 F. Boas, in the Sixth Annual Report of the Bureau of Ethnology, p. 605.

2 Mannhardt, Zeitschrift fur deutsche Mythologies vol. 4 (1859), p. 301 ff.,
interprets our children's pulling-games (" London Bridge " and the like) as
originating in religious notions of a contest between the powers of light and
darkness for the possession of souls.

^ Mrs. W. Wallace Brown, in the Journal of American Folklore, V. 57, Jan.-
Mar., 1S92.

On irelpap eXiadai and Manus Consertio. 155

Gaius, IV, 16, gives in full the formalities employed in case the
object in dispute was a movable chattel, like a slave. Both parties
laid hold of the slave, each in turn touched him with a stick (festuca^
vindictd)^ and claimed him in a form of words. The magistrate then
said mittite ambo homin^m, and both loosed their hold. Then fur-
ther parley, which does not concern us. We are not told that they
pulled the man, but the joint seizure certainly seems to typify pulling,
and we are told that the stick represented a weapon. So much for
movable property ; now how about immovables, which could not be
brought into court? Gaius goes on to make provision for these
(IV, 17) in a passage which is incomplete, as a leaf of the book is
gone. He says that a part of the object was brought before the
magistrate, — a tile from a house, a clod from a farm, etc., — and
that proceedings went on as over the whole (Jn earn partem perinde
atque in totam rem praesentem fiebat xnndicatio). But this cannot be
the whole of the story, at least as regards land. For we happen to
hear from two other sources, Gellius XX, 10 (a vexed and vexing
passage), and Cicero pro Murena 12, 26, about another act, per-
formed at the land itself. Respecting this act the practice was
different at different epochs. The old and simple way was for the
praetor to go to the spot himself to be present at the ceremony.
Then when distances became too great and magistrates too busy,
the litigants went there with witnesses to perform the ceremony,
and came back into court.

Now what was done at the land ? The ceremony was called, we
know, manum conserere. Just in what it consisted we shall inquire
presently, but I wish first to enforce the point — not generally recog-
nized — that the manus consertio * applied only to land, or to land
and other immovables. For observe first that it was something done
out of court. The phrase ex iure invariably accompanies it, wher-
ever it is mentioned.* The locution in iure manum conserere was
evidently unknown to Gellius, except as an antiquity in the Twelve
Tables. Now a thing done out of court could not apply to movables.

* We may allow ourselves this phrase, as a convenience, although it occurs in
no ancient writer.

* Thus in the well-known passage of Ennius quoted Cell. XX, 10, 4 (and by
Gcero in two places); in Varro L. L. VI, 64; Cic. de Orat. I, 10, 41; pro
Murena 12, 26; Cell. XX, 10, x and 9; Probus in Keil's Gram. Lat. IV, p. 273;
these (with Lactant. Inst. I, x, 12) being all the occurrences.

156

Frederic D, AlUn,

for movables, Gaius tells us, were always brought into court Sec-
ondly, both in Gellius and in Cicero's Murena — the only places
which show an3rthing about the application of the phrase — the talk
is solely of land-litigation. When, as in Decemviral times, the act
was performed in iure^ it still had reference to land, for Gellius evi-
dently means that the praetor went to the spot. And even in a third
passage, Cicero de Orat. I, lo, 41, where the phrase is employed in a
bantering metaphor (te ex iure manum consertum vocarent quod in
aHenas possessiones tarn feme re inruisses)^ the figure is that of a
trespass on land. Finally, any remaining doubt should disappear on
a comparison of the Murena passage with Gaius's description of the
vindication of movables. I print the two dialogues side by side.
Cicero, it must be understood, is quoting in a touch-and-go fashion,
simply to illustrate useless legal verbiage.

Gaius.

A. {holding the slave'\. Hunc
ego hominem ex iure Quiritium me-
um esse aio secundum suam cau-
sam. Sicut dixi, ecce tibi, vindictam
inposui. {Touches him J]

B. {also holding the slave^. Hunc
ego hominem, etc., etc. {Touches
him!\

Praetor. Mittite ambo homi-
nem.

A. Postulo annc dicis qua ex
causa vindicaveris.

B. lus peregi sicut vindictam
inposui.

Cicero.

A. Fundus qui est in agro qui
Sabinus vocatiu* . . . eum ego ex
iure Quiritium meum esse aio.

[B. repeats the same formula, '\

A. Inde ibi te ex iure manu
consertum voco.

B. Unde tu me ex iure manu
consertum vocasti, inde ibi te re-
voco.

Praetor. Suis utrisque super-
stitibus* praesentibus istam viam
dico. Ite viam. {They go ^ Red-
ite viam. {They return. "]

A. Quando te in iure conspicio,
. . . anne tu dicis qua ex causa
vindicaveris ?

[B . Answer not given .]

^ Compare Festus, p. 305 M. : super stites testes praesentes significat, cuius ret
testimonium est quod super stitibus praesentibus i, infer quos controversia est, vin-
dicias sumere iubentur. For the meaning of vindicias sumere, see p. 161 below.

On welpap ekitrOcu and Manus Consertio, 157

No one can help seeing from this correspondence that manus con-
sertio was to land what joint seizure and touching with the wand was
to movable property, and that it was done out of court because it
had to be done at the land itself.

The point established that " consertion of the hand " pertained to
immovable property only, let us ask next in what the act consisted.
We ought to know this from the chapter of Gellius (XX, 10) which
has already been several times referred to, for Gellius wrote this
chapter on purpose to tell us, and there is in it a passage evidently
meant to give an explanation of the phrase. In its present shape,
however, it passes the human understanding; one may read and
re-read it, and still ask himself what manum conserere really was.
We shall have to attack this perplexing passage directly, but we will
put it off as long as we can. Meanwhile let us consider possible
answers. The litigants evidently could not pull the land, as they
could a man. They might, nevertheless, lay hands on it simultane-
ously. If manus consertio is this, it is identical with the act for
movable property, an act to which, as we have seen, the name was
never applied. But a different name implies rather a different act,
and this brings us to another possibility, that the litigants joined
hands over the land. Now ' link the hand * is the natural meaning
of manum conserere. It is very hard to conceive it as meaning any-
thing else. And for etymological reasons alone I cannot help having a
strong conviction that this was the act performed. Herein I agree
with Poste,* who says " the object grasped seems to be the hand of
the adversary " ; and I hope to show that Gellius's language confirms
this view, and does not, as has been supposed by some, favor the
other alternative.

Now we must address ourselves to the Gellius passage. It is as
follows. After saying that he is going to tell us what he has found
out about the meaning of the phrase ex iure manum consertum^ he
goes on, according to our manuscripts : —

7 Manum conserere. Nam de qua re disceptatur in iure [in re added
by editors'] praesenti, sive ager sive qmd aliud est, cum adversario simul
manu prendere et in ea re omnibus [sollemnibus editors'] verbis vindi-

8 care, id est vindicia. Correptio manus in re atque in loco praesenti

^ Gaius, 2d edit^ p. 499-

IjS Frwdgric D.

apod praecorem ex dnodedzn scnlis Trhar, in qnfbcs vsl ac r iptum est :
9 si qui in ixxre Tiannin canseronr. 5ed pastir.azzL praetores, prapogatis
Italoe inibos dans iarfsdirnnninns 'ja£X£ tsto zoords vurumily emsmJdd]
negotils occrpar. padsd. ^pra&zsci ii£ijn\ vrnffiriannn cficrndanun
cansa Tad addid by eJitjril laa^mqnas res gravabannrr. instltutum est
contra duodedm Tahnlas Tacfm sensa ^^^nsensa ddd4jrs\^ at Ihigantes
noQ in iiire apod praerorern iianuni consercirent. sed ex hire mannm
consertmn vocarent. id est alter alterazn ex rore ad coosereodam mannm
in rem. de qua. ag<Kvcar. vocaret. atqize profecti simol in agrnm, de quo
litigabatur. terrae aliqnfd ex eis. od anani gtebam. in ius in orbem ad prae-
torem deferrent. et in ea. j^ieba. tamqoam in toto agnx vindicarent.

The p>assage begins with two loose words, oat of all connexion with
their context. No probable way of connecting them with the fore-
going or the following words has ever been suggested, and Hertz is
undoubtedly right in assuming a lacuna after them. Now the care-
ful reader will perceive that the promised definidon of manum con-
serere is altogether lacking. The sentence banning with nam is
not this definition, although it has sometimes been taken to be, and
writers have consequently inferred that manus consertio was joint
seizure, and even that the joint seizure of mo\'ables described by
Gains was called by this name. This sentence does not even pretend
to be a definition of manum conserere^ but it does pretend to be a
definition of inndicia, Vindicia cannot, however, be meant as a s>Ta-
onym of manus consertio^ since it applies to land and movables (sive
ager swe quid aliud est) , whereas the manus consertio relates to land
only. More than this, the vindicia here described cannot even in-
clude manus consertio. Possibly the term vindicia would, in its
wider use, include the manus consertio, though I know of no clear
case of thi». Vindiciarum dicendarum causa in § 9 below is not one,
for vindicias dicere is merely a technical phrase for 'pronounce
Judgment.* At any rate it seems quite certain that vindicia here is
employed in a narrower sense, and designates a ceremony distinct
from manum conserere. For in the first place, cum adversario simul
tNiinu prendere is irreconcilable with correptio manus in the next
Bcntcuce (that is, the two phrases cannot designate the same act),
but UsS we read on it becomes evident that correptio manus relates to
the manus consertio. Secondly, this vindicia takes place in court
{in iure)^ whereas manus consertio took place — except in the earli-
CMt timc9 — out of court. I should not lay much stress on this />

On Tret pap eXiadac and Manus Consertio, 159

iure, because the text is defective just here/ if it were not upheld by
what follows in § 9. At the end of that section the custom, known
also from Gaius, of the vindication of land by means of the gleba is
mentioned. Let no one suppose that Gellius means this to be the
manus consertio. For this takes place apud praetorem^ whereas he
has said just above that the manus consertio did not (ut non in iure
apud prae tore m manum consererent). There was, then, a vindication
of land quite apart from the manus consertio — a vindication which
took place in iure and was performed over a clod. Surely of this
and of no other Gellius is thinking when, in § 7, he speaks of claim-
ing land by joint seizure, and lumps it with the claiming of mova-
bles. In iure is therefore in all probability sound.

It is quite impossible, we see, that this »a/n-sentence in § 7 should
be a definition of the manus consertio or have anything to do with
it. The following words, correptio manus etc., certainly refer to it,
but they simply give additional facts, and presuppose a definition.
The definition must absolutely have preceded, and it is plain that the
loose words manum conserere at the beginning are a remnant of the
missing passage. In the light of the foregoing discussion we may
imagine that the first two sections once read, in substance, somewhat
as follows : —

Manum conserere est manum adversarii corripere in agro de quo liti"
gatur ; ita enim maiores nostri pro vindicia faciebant quae posteriore
tempore in iure fiebat et volgo sic vocaiur. Nam de qua re disceptatur in
iure in re praesenti, sive ager sive quid aliud est, cum adversario simul
manu prendere et in ea re x^?//emnibus verbis vindicare, id est vindicia.
Correptio manus in re atque in loco praesenti etc.

There is still one incoherency in Gellius's account. The litigants, he
tells us, when the praetor became too busy to go with them to the
land, ' no longer did the manus consertio in court, but one summoned
the other out of court to perform the ceremony. They then went
together to the land, and * — what next ? Joined hands, we expect to
hear. No, but ' took a clod, brought it back into court, and vindicated
over it there.' Gellius appears to mean that, although there was a sum-
mons ad conserendam manum, the consertion itself was entirely omitted

^ It would obvioasly be open to us to write disceptatur in re praesenti mstead
of disceptatur in iure {in re) praesenti.

l6o Frederic D, Allen.

and a different ceremony sabsdtnted. They employed, — so we are
given to onderstand — the clod-form of vindication for knd, while
retaining the verbiage of another, obsolete, form. But this statement,
even if true, cannot be historically complete. GeDias evidently omits
at least one intermediate stage. There most sorely have been a
time when the challenge to leave court ad conserendam manmrn was
followed by an actual performance of the diing specified ; die substi-
taticm of the clod-ceremony most have been later. Bat GeUias's
language is, perhaps, capable of another interpretation; he may
mean that both acts were performed, — the carreptio manms and the
bringing of the clod. The ceremony with the clod would then be an
addition, not a substitution. Now a double process of vindication
would seem to us needless; nevertheless such a thing might have
been thought necessary at the period when the praetor first ceased
visiting the land and witnessing the ' hand-grappk ' in person. In
itself this alternative is more probable than the first, whether it be
Gellius's meaning or not

I cannot forbear mentioning a third possibility. It may be that
GelHus has confiised two things ; that vindication with the dod was
really a matter quite apart from manus cansertio^ the one being
employed in some cases, the other in others. It must be understood
that the whole afiair — clod as well as hand-grapple — was to Gellius
a thing of the past. This is clear firom the language of Gains, a con-
temporary of Gellius, who uses the past tense in describing all these
acts of vindication. Gellius, himself, at the beginning of the chapter,
says of the words ex iure manum consertum that they are verba ex
antiquis acttofubus^ q^<^f cum lege agitur et vindiciae contenduntur^
diet nunc quoque apud praetorem soleni^ and goes on to describe the
difficulty he had to ascertain their meaning. Though stiU in use, the
words were a mere form, corresp>onding to no real act. Even at
Cicero's time the same was probably true. From the dialogue in the
Murena we may suspect that the whole visit to the land had become
a fiction, the parties simply going a short distance and returning.
The jurist L. Cincius, in a passage to be quoted unmediately, speaks
of the clod-bringing as if it were an obsolete custom, and Cincius
lived at about Cicero's time. It would, therefore, not be an incred-
ible supposition that Gellius, reading his lawyers' books, had got two
processes confounded.

On irelpap iX^cOai and Mantis Consertio. i6i

With this possibility in mind, we naturally turn to the scene in the
Murena oration, to see whether it gives any hint of a clod. It cer-
tainly does not. With the return of the parties, the ceremony of
vindication seems to be complete. However, it may be that Cicero
simply skips this part of the process. One rather striking coinci-
dence could be adduced in support of this opinion. In Festus,
p. 305 M. (quoted in footnote on p. 156), we are told that litigants
were bidden vindicias sumere in the presence of witnesses [supersti-
tibtis prcusentikus). We should hardly know what vindicias sumere
meant, were it not for the statement of Cincius (Festus, p. 376 M.),
vindicicu olim dicebantur illae quae ex /undo sumptae in ius adlatae
erant From this it appears that the clod was called vindiciae^ and
that vindicias sumere is to take the clod from the farm. Now the
command to do this superstitibus praesentibus reminds us strongly of
the command of the praetor in the Murena scene (see p. 156), and
suggests that in that command the fetching of a clod may be implied.
On the other hand, it may be urged that the simple manus consertio^
without the clod process, would also have required witnesses.

I can see no way of deciding positively among these three under-
standings of the matter, but confess to a leaning towards the second,
— the assumption that in post-Decemviral times, when the magis-
trates gave up visiting the land, the clod ceremony was superadded
to the other. This clod process was called vindicia^ and was prob-
ably performed by joint seizure, as on a movable.

This discussion of the manus consertio has necessarily been a little
complicated, and a word of recapitulation may not be out of place.
The points made are these. Manum conserere was a form of vindica-
tion that related to land (and presumably other immovables), and
corresponded to the joint seizure of movables. It was performed
out of coiut, before witnesses, except in very early times, when the
magistrate used to go to the land. The words naturally mean * link
the hand,' and Gellius calls the act correptio manus. We infer that
the act was a joining of hands, and no statement of the ancient
writers controverts this view. We may distinguish at least three
periods of the usage. First the magistrate went with the litigants
to the spot to see the ' hand-grapple ' performed. Secondly the liti-
gants went there with witnesses and performed it, and (if we have
righdy chosen among puzzling alternatives) brought back a clod for

i62 Frederic D, Allen.

a second process in court. Thirdly the actual performance of these
ceremonies was omitted, an empty form of words being still kept up.

The usual term for what we have called the ' joint seizure ' of
movables, as opposed to the manus consertio^ appears to have been
vindicia. • But we may shrewdly suspect that it was originally called
adseriiOy adserere^ — words which in the literary period were con-
fined to a single kind of process, the vindicatio in Hberiatem or in
servitutem}

We must not forget to speak of the variant manu consertum^ which
manuscripts give us in several places.* This may of course be noth-
ing but archaic spelling. But it may also have been an actual form,
used along with the other.' Its meaning would be just the same.
Either consere tnecutn manu or consere mecum manum might be said.
In like manner adserere alicui manum^ and adserere aMquem manu
were both known.

The interpretation of this hand-joining as a pulling-match is not
absolutely necessary, and I have wondered at times whether it might
not be taken as symbolizing a scuffle or a wrestling-match. This
would be the form which a quarrel about land would naturally
assume ; each claimant would treat the other as a trespasser, and try
to put him off the premises. But the fact that one hand was used
seems fatal to this. The plural manus is never used in this locution.
A pulling seems indicated after all, and we must suppose that it arose
as a substitute for a free scuffle, suggested by the analogy of the pull-
ing of movables. This hand-pulling, it is interesting to observe,
became a common figure for a battle, as did the rope-puUing of the
Greeks. The oldest occurrence of the metaphor, contra conserta
manuy Plant. M. G. 3, shows the singular, in agreement with the legal
formula. Varro L. L. VI, 64, manu conserere cum hoste, also uses

' In the Oscan Tabula Bantina, the words manim aserum (= manum adse-
rere) seem to be employed in reference to seizure of the person in execution for
debt (jnanus iniecHo) .

^ The Mss. give manum in Cell. XX, 10 in six places, also in Cic Fam. VII,
13, 2. Manu in Varr. L. L. VI, 64; Cic. Mur. 12, 26 (twice); GelL XX, xo, 9
(once). Both forms Cic. de Orat. I, 10, 41; Mur. 14, 50; Cell XX, xo^ 4; Pro-
bus in Keil's Gram. Lat. IV, p. 273.

' Hardly, however, the only form. Note ad conserendam manum in Gellius.

* Festua s.v. sertorem (p. 340), Paulus s.v, asserere (p. 25).

On ireipap ekiaOai and Manus Cansertio. 163

the singular, and so alwa)rs Caesar, Nepos, and probably also Cicero.^
Expressions like manus consereret first appear in Sallust and Livy,
who, however, also use the singular.' It is hard to say whether this
is a simple perversion of the original term, or is due to an admixture
of a wrestling-metaphor. Manus conserere (' lock the arms ') might
have been said of the i^-q of wrestlers, but there is no evidence that
it was. Another faded-out metaphor based on the pulling-contest is
contendere^ contention

This inquiry has taken us far afield, but the patient reader will now
understand the nature of the hypothesis — I repeat that it is only an
h3rpothesis — by which I should like to account for the enigmatical
Homeric expression from which we started. *Faustrecht' is the
oldest law, and the earliest administration of justice consisted simply
in seeing fair play between combatants. The free fight gave way
(with great saving to life and limb) to a regulated pulling ; the dis-
puted object belonged to him who could pull it away from the other
man. The pulling was then extended to contests about immovables
and abstract questions, the victory being his who could pull his oppo-
nent over a scratch. Then came interference of magistrates, and
questions of Suci; and ius. But it was still thought needful that a form
of pulling should be gone through, before the magistrate began his
inqmry. To this point the Romans had got when we first know
them, and this point, as I imagine, the Greeks had reached in the
Homeric age, the difference being that the Greeks used a rope when
a movable object was not in dispute, while the Romans did not. I
am Ceut from asserting that every civilized community has gone
through all the above stages. But such, I conceive, was the course
of things in ancient Italy and Greece.

There are various other difficulties in the trial scene in S, which
we shall not attempt to deal with. The relation, for instance, of the
MiTQip to the ytpovrt^, and the significance of the two talents of gold.
Skilful discussions of these, by Hofmeister and Leaf respectively, may
be read in the 2^tschrift fur vergleichende Rechtswissenschaft^ vol. II,
p. 443 ff., and in the Journal of Hellenic Studies ^ vol. VIII, p. 122 ff.

^ In the tpnrioiis oration Antequam iret (8, 20) occurs manus conseratis,
* Noteworthy is conserunlqtu dextrtu, Stat Silv. x, 6, 50.

164 Frederic D. Allen,

T wish I could strengthen this surmise of mine by citing some sur-
vivals of the terminology of the supposed custom in Greek. There
are indeed references to pulling-matches. One of these is the much-
vexed passage Pindar Pyth. II, 90, araBfUK 3c rcvo? cXKo/ttcKoc irtfMvd^y
where the meaning is * dragging at some rope beyond their strength.'
But I have not found anything that looks like judicial pulling.
Nevertheless I will cite one group of words. The game called
cKawip^v €\k€iv has been mentioned above (p. 152). Hesychius

adds that wav t6 Svaxepik^ (TKaircp&L Xeycrou, #cal 6 ttoo^cdv CKatrip^rf^,
He has also a gloss aKawapSoi * 6 rapa;(o>8i7s fcol &^aytayo^. That is,
cKairtpSa or aKawdpSa meant ' tug-rope/ a desperate job was called a
*tug/ a man in straits a * tug- puller,' a headstrong horse a *tugger.'
Now in a well-known fragment of Hipponax (Frag, i Bgk.), the poet
calls on Hermes with the words Stvpo fioi o-icairapScvcrac. Above this
last word, in the codex which contains this fragment, is written the
gloss orv/Afuxx$otu, which is right enough, for o-KairopScvoxu certainly
means 'pull on the rope for me.' Now there are three other Hesy-
chian glosses that contain similar forms :

(TfcaTTCpScvonat * XoiSo/DiJaai.
oncapiraScvfToi * Kpivai.
KairapBakrai ' poLvrtwraxTOax,

In the second, orKapTruScixrou is surely miswritten for (nca^rapScixTitu.
Here one might imagine that there lurked in Kplvai an allusion to
judicial rope-pulling, if it were not far more probable that all these
glosses relate to the Hipponax passage, and are wild guesses about
the last word in it. It seems hardly chance that all end in -ocu.

As a corollary to this discussion it may be pointed out that the
proposed interpretation of 2 501 makes it possible to separate for-
mally the word meaning * rope ' from that meaning * end.' Outside
of Homer the word for * end ' is an j-stem, and with S 501 out of the
way there is no reason why it should not be so in Homer. The form
TTtlpap can be restricted to the meaning * rope.' In short, we may
distinguish two words :

1. wtlpapf plural ircipara, * rope.*

fntipa^ (Pindar), plural wupara \

2. j Aeol. ir€ppa%f plural rrippara > * end.'
I. Att. wipai, plural rripara J

On irtlpap ekiaOai, and Manus Consertio. 165

The Homeric occurrences of these words are :

' Rope:

2 501. Already discussed.

N 35^* "^^ ^ Ipt&K Kpartprj^ icou o/ioiiov ttoXc/aoio

irccpap iiraX\aiavT€^ itr* ifxifxyripoto'i rawca'aVf
appffKTov r SXxirov tc, to iroAAwv yowar' cXixrav.

A difficult passage. I incline on the whole to understand a rope-
pulling between the two armies helped by the gods, — substantially
Heyne*s interpretation. 'Etta then goes with Taia;o-<rav. This con-
ception is strongly suggested by the expression Ipiho^ #cai iroKkpm
TTtipy), especially when the similar phrases in A 336, O 413, M 436,
Y 101/ H 389, n 662, are taken into account. In all these places
there is the underlying idea of a pulling-match. The difficulty is that
aXvToy in the next line is irreconcilable, and if we adopt this view,
verse 360 must be bracketed as a later addition.*

€ 289. iKff>vy€€iv pr/a wtipap Sif^voi, rj /uv Ixava.

' Noose of tribulation.*

H 102. aVTCLp VTr€p$€V

viKfj^ irtCpar l\ovrax iv d.Oavdroia'i Otoiaxv.
The gods hold the ropes which decide victory. A * tug of war ' is
certainly intended here.

H 402, M 79, X 33» 41- 6\«6pov vtipar iff>ri7rrai (^^^irro).
The ' noose of death ' is fastened to them. The metaphor from
snaring birds or animals.

Z 143, Y 429. iamv Iff ii Kty Oaffxrw 6\€$pov wtCpaff uci^.
' That thou mayst be caught in death's noose.'

^ Here, however, r Aot does not mean ' rope-end,' but wokifiov riXos is used
as in n 630.

* The only possible alternative is to revert to the conception of older editors,
and nnderstand a ' noose of destruction.' I should then understand that Zeus
and Potidon ' throw reciprocally (that is, each to the other side, ^aXXd() the
noose of war, and tighten it round Greeks and Trojans.' Two nooses would, of
course, be intended. I do not feel sure that this is not right. Quite impossible,
to my mind, is the complicated explanation which has found favor with most
recent editors, by which the two gods pull at opposite ends of a rope in which
the combatants are somehow implicated. This is meaningless and corresponds to
nothing in actual life.

i66 Frederic D. Allen.

fi $if 162, 179. iK S avTciv vtipar dan^^Ot^ {iyijwTw).

' Let the ropes be made fast to the mast/

y 433' ^^' ^^ x^P^"^ ix"^ X*^^^> vkipara rixyrii.

The smith's tools are called the * ropes of his art' — a figure
borrowed from seamen's parlance. So our figuratiye expression
' know the ropes.'

^ 35 0« ^'r*^ ^ TToxjSi iKacrrov vtlpar iairtv,

Nestor ' showed the ropes * of each part of the race to Antilochus.
The metaphor as in the last.

'End:

B 200, 301, 3 563. vtlpara 701179.

c 284. vp^% iirl irctpaorc yalrj^, •

X 13. "fj S l^ irtipaff uiavt fi<i$vpp6ov cuxcavoiO.

478. ov8* ct Kc ra vetara vtipaB" iici^cu.

yoii/s Kou irovroio.

1^ 248. ov yap TTCtf irarroiv ^irc ircipar* ac^Xcuv

ijX^ofiev.

As the plurals of these words agree in form, so do the verbs
derived from them: vctpcuvw 'fasten,' 'bind' (x 175, 192; cp.
Hymn. Horn. Ill, 48), and w€ipalvia (= irtpaCvta) * complete ' (ft 37 ;
borrowed Soph. Trach. 581).

It will be perceived that in the above distribution, we have made
more of the ' rope '-meaning than is ordinarily done, and recognized
it in more passages. Modem editors, it seems to me, show a certain
disinclination for this meaning, and will admit it only under compul-
sion. They follow in this the ancient interpreters of Homer. I may
be wrong about one or two places. Only I must enter strong protest
against one thing, — a straddling interpretation, which assumes both
meanings (say ' rope ' and ' issue ') at once. Either the word means
' rope ' in a given place or it does not.

With v€ipap * rope ' thus formally distinguished from ircpas, its ety-
mology becomes a separate question. It may be an equivalent for-
mation from the same root, with a specialized meaning. It is just
possible to get from ' end ' to * rope-end,' if we suppose the word to

On irelpap iX^cBai and Manns Consertio. 167

be a technical sailors' term. This would differ but slightly from the
common view, which throws the two words entirely together. Merry
and Riddell on fi 51 make a vigorous plea for this. They cite endje
as a German sailors' term. * Rope-end ' will certainly fit (instead of
' rope ') in all the Homeric places listed above, except N 359, where
both gods (if we accept the tug of war interpretation) pull on one
icufap. But TTcifop may also be wholly unconnected with ircpas.
Doederlein suggested that it might have to do with ircipci/9 ' basket.'
The notion of ' plaiting ' would be common to the two.

However this may be, both the meaning * rope ' and the form
Tccpap disappeared from Greek at an early time, ncipop indeed is
not found after Homer ; ircipara is used in the Homeric Hymn to
Apollo 129, of the swaddling-bands of the infant god (Xvovro 8i
trtifxiTa frdvTa), Also, Still with the ' rope '-meaning, in a verse of an
unknown poet, Stob. Eel. I, 2, 9, Zm 6 koI (co^s koI Oavdrov wupara
viOfjLWf} Pindar's wtipar AWkinv Scockucv, Pyth. IV, 220, looks back to
^ 350 as well as to ^ 248. It is hard to tell which meaning the poet
had in mind. Not quite so doubtful is voWiav ircipara awravwrtu^ iv
Ppaxu, Pyth. I, 81. Fennell recognizes that this somehow pertains
to the " handling of ropes " ; Gildersleeve suggests that the twisting
of strands by a ropemaker is meant. But the context of this last
passage warns me to desist.

1 Pat by Nauck among the Tragic Adespota (472). But is it not a dactylic
hexameter, ZciW (7^) 6 ical Bapdrov ical ^io^t wilfiara puffiuw?

HERONDAEA.
By John Henry Wright.

I.

PUNCTUATION IN THE PAPYRUS.

I. The spaces, — It is an important peculiarity of the papyrus
manuscript of the Mimiambi of Herondas recently discovered (Papy-
rus No. CXXXV., British Museum) , that, while as a rule the letters
of the several verses are written continuously, without break or pause
between the different words, now and then — in about twelve per
cent of the verses — slight breaks or blank spaces do occur, never
amounting however to more than the space ordinarily taken up by
from one to two of the letters of average breadth. The significance
of these breaks for the punctuation of the text was first emphasized
by Blass, and has been recognized by several critics of the poet;
but thus far only sporadically. In this article I propose to present
all the examples, and to discuss the dpubtful ones, not neglecting at
the same time the examination of a few related topics, important in
their bearing on the text-criticism of our author.*

It should be observed, in the first place, that these breaks are
never intended to mark words as words, nor to suggest the proper
combinations of letters into words in ambiguous instances : this work
is performed, but without system, incompletely, and only very rarely
direcdy, by the marks of the rough breathing,* the accents,* the coro-

^ For convenience ordinary t3rpe will be regularly used in the notes for the
readings of the manuscript It is to be regretted that in the text the font of
inscriptional type so imperfectly represents the cursive majuscules of the papyrus.

' Only the rough breathing is written, and always in an angular form (*~),
except in the late wh'kw^ VI. 3. The cases are : II. 70 (wra7iya) ; V. 20 (irw-
ffir); VI. 25 (^ Biror<Hr); VI. 68 (<iM*[XX]i,) ; VII. 46 (6i).

' The accents, acute, circumflex, and grave, exhibit interesting peculiarities.
The circumflex and acute are used with many proper names, but not with all ; also
to distinguish between words spelled alike but differently accented, and to indicate
the correct grouping of letters into words, etc. : e^.^ I. 29 (tf/cu, not ^ea^); I. 85

169

I/O John Henry Wright.

nis/ and perhaps once or twice by the use of the dot or point.* The
last, however, has a more extended use — in peiiiaps a dozen exam-
ples, collected below — as a sign of punctuation (ortyfuy), having here
a value not wholly unlike that of the spaces, but not so strong as that

of the jrapdypa^oi.

Punctuation within the verses is indicated mainly by these spaces :
indeed, these spaces have no value except as signs of strong inter-
punctuation, and they alwa]^ have this value, when not accidentally
made.

(juif not ftd) ; iHd, (ocvwi = ot tf'ov, not iwov) ; IV. 42 (avrif = ^vn^ not
m&Tii) ; II. I {irri = i9r4, not i^TM). At V. 41 (o*^) perhaps the accent may
indicate odif. At II. 9 (lif^iMv) the accent seems to suggest the synalcepha of the
final syllable (Crosiiis) : and in V. 49 the acute on the ultima (cunycovcdt) may
be intended to indicate a rising tone of voice, necessary in a question (Diels).

The grare accent regularly appears to be used to provide against mtsapprehcn-
sions. The examples are I. 60 (rdroXi . . . : to show that this is not rmra; it
also shows that we have a longer word here than rmra) ; I. 76 (IIv#iM 9i) ; L 70
(tf nryiff : 6 + ipayit^ not Jr rrX.) ; II. I (i^ri : *rrf, not lore) ; IL 24 (*/*' :
ifU, not ifid) ; HI. 74 (Wpmf : Le. rcpvdf, not r4p»as) ; IV. 91 (riXmmw : i.e.
rt\a,96p); Wl. 46 (biz perhaps taken for article — wrongly; hardly "^xxiae
indicium,^ Crusius).

^ The coronis (*)» usually written at the top of the line (at I. 15 and 11. S^ at
the bottom), always appears to indicate elision at the end of words^ It seems to
have been put in by the first hand« except at VI. 3 (ovf ^). The other cases are :
L 15, /oftjoaow = fu/tm 8009; IL 24, i/i* otr = I/ao^ = ifti 9^; Uh 49^ mXiptfcv* =
KiXifdlpa; IV. 5y t^twrtp = cJrrcp = col wvrcp; IV. l6» oXcxrop nfr/m = iXa^
T6pa tifrpa; IV. 41, cv^tXX'iotftf'a = Kv^^XXa loS^a. The only doubtful case is II.
83. Kavrorraa-^aumw : this must be ad aSrbt rd vit atkov, hardly rd reatrrov, cer>
tainly not, in the light of all our examples* ris avrov (Doric short a) . The cases
show that the mark is not necessarily used, as in the Codex Alexandrinns, to sepa>
rate words as words, nor like the duwroX^^ (vvoSumtoXi^) : iraw 8ui/mSXau tml
BiAx^V^^ 6^e£XM;^^ Tiwa \d^ i^-^-* l^rur, d^un, not Irrt N^^ck : Bekker,
Anted. Graeea^ IL p. 675 j ; some of the examples, however, m^t be explained
as iiwtanrm of the diastole: ^^., II. 24 (^pi'ov, not amv), HL 49 (coXaf^v^, not
KoXiftfTm), etc CI Gardthausen, Gritdk. PnUSograpkit^ pp. 273 C

^ In Proem. II (Cr.) ra«vXX-«i3ijr, the point appears to be intended only to
mark off the words rk cvXXi and iM^wi it can have no force for punctuation
here; cf. FV. 50 (M'tfvr-if/M^), but see p. 182, note i. (In L 3 [rtr] the dot is
merdy a part of the sigma : likewise at IL 6 [cX^awoc], the mark above tiie X is
part of an unfinished a, begun too near the X; c£ 1. 51; at VIL 48 [&CMf], die
mark on o is part of a ^ in the preceding line.) For some remarks on the use of
the signs ( " , ") see below, pp. 177, 178 notes.

Herondaea, 171

An examination of the photographic facsimile of the manuscript
discovers about one hundred and ten cases of such intentional spac-
ings. Of these not more than from two to six are in any way ambig-
uous, and a fair consideration, it seems to me, would deny ambiguity
to all.^ The remainder, over one iiundred and four, are nothing but
indications of punctuation, for which we have in our modem editions
our various signs. A classification of these examples according to
the punctuation adopted in the latest text-edition of Herondas {^Bib-
Uotheca Teubneriana : ed. Otto Crusius, 1892), which errs by no
means on the side of excessive punctuation, yields the following
groupings : —

a, — The breaks accompany a change of speaker^ in the dialogue at
I. 7* (coTiv; TuXXt?), 20, 82; II. 48; III. 58, 78*, 81, 82, 87, 93;
V. 31, 55 ; VI. 15, 17, 19*, 22*, 23*, 25 ; and are thus represented
by our period, colon, dash, or question-mark (the last indicated by
the*).

b, — They stand at the end of a question in the examples starred
above, and in the following additional cases where no change of
speaker occurs after them : I. 9, 48 [?] ; III. 43, 60 ; IV. 5 7 (per-
haps an exclamation, ola l/>ya;) ; V. 10, 18, 41, 75 ; VI. 10, 44, 45,
75, 76 ; VIII. 4, 5. In these cases they are represented by a
question-mark,

c, — Many have the value indicated by Crusius by a period: most
of the unstarred cases under a^ and the following additional examples :
I. 8, 79 ; II. 68 ; III. 59 ; IV. 33 ; V. 20, 56, 66, 67, 74 ; VII. 4, 117.

d, — They have the value of a colon at I. 15, 66, 82 ; III. 11, 26 ;
IV. 21, 55, 92, 93 ; V. 6; VI. 5, 31, 61; VII. 65, 128; VIII. 11 ;
and of something like it at IV. 58.

^ The apparently exceptional cases are discussed below; see pp. I73f. Occa-
sionally, bat extremely rarely, when the large bulk of the writing is considered, we
find other slight breaks. In most of these instances the letters of a verse have
been written more sprawlingly than usual, and thus give the appearance of spac-
mg where no pause is intended. I have observed only these examples : yiip^ijuhtw

(L 46); VO0Jl0W (I. 60); fieT€\$€l9' fi9fivfiJI9 (11. 50); KUWOvfik (IV. 30) *,

oC^rot (V. 43); roO^row (V. 58); IjfUtt^v (VL 82). (In the apparent ^XcUr^ov,
I. 5, the letter iota has disappeared, leaving only a slight trace.) Such is not the
explanation of the pause in I. 55, discussed on pp. 186 ff.

' Change of speakers is usually indicated, but with many omissions, by the
rapdyfitifot; see pp. 178 if.

172 John Henry Wright,

e. — Cnisius represents them by a comma at I. 13 (^«), 67, 89;

11. 22, 49, 77; III. 49, 81 (Travcrai^ucaviu) ; IV. 43, 46, 90; V. 9,
25> 34. 42 (tovSc^kou <rv), 53, 69 (ran \<iXAa) , 70 ; VI. 3 (aunyv^crv),

12, 18, 49, 77, [96, after rt] ; VII. 57 {bis), 58 (before KawafiurKa),
60 (after dKpo<r<^v/>ia), 61 (after c<^)3(x), 98, no.

/. — In the following verses, where the spacings are indicated by
the sign of caret, Crusius inserts no mark of punctuation ; but no one
can deny that at least a strong phrasing, if not punctuation, was dis-
tinctly intended : II. 2 (ovk iarl \ ^/acW Kpiral B-JKovOfVf^pvSk r^s
So^c) ; III. 10 (t6v fiicrBov airtif^K^v to, Nawoucov fcAaixro)) ; IV. 83
(cv/ACv^s u-qs I Kcu\ci)9 iir ipois ToMrBcf^Kd tikcs rcavSc | uur ^irvii/rou) ;
IV. 42 (ov <ro4 Xeyoiy^auTi; Tg . . . xaaKOwrg ;), and 55 (avny crv^
fjudvov) ; III. 25 (rpiOriikipq. Maptava ypoLfifiTrliovTO^ | roG irarpof avrcpy^
rov MopcDva ivotrfctv \ ovro^ 'SiCfimva) ; IV. 24 {ov\ 6p§i icciva | Iv rg
Paxr€ifja ypafifmra ; i.e. * Don't you see those [things] on the pedes-
tal, the letters?') ; perhaps also IV. 59 (t6v muSa Brj^{rov) yv/xvov).

At III. 80 there is a pause which taken in connexion with the cor-
rections at this point is extremely significant. As first copied, uncor-
rected, vv. 79, 80 read (in part): (79) ITICOIZUIHN (80) <t>eP OC-
ACANHKAKHCeeNHIBYPCAI. The corrector, evidently die
first hand, having previously designated 80 as corrupt (by an oblique
line in the margin opposite ^GP ; see p. 181) sets himself to correct
it. He puts a mark of erasure over the N of ZUIH N (superior dot) ;
writes in, in the upper part of the space after ♦€ P, the letters GIN;
and draws his reed through the two Vs (at C0GNHI mistakenly; at
BY PC A I apparendy correctly ; probably these I's were earlier can-
celled by the copyist, as he wrote) . These facts show that the original
manuscript from which the papyrus was transcribed read something as
follows, of course metrically an improbable reading :

Metrotime (/o the master), et rl o-oc C<^i
<^€p', — wTo.% &v r\ KOKYj (rOtiTQ fivpo'a :

i.e. *If I am anything to you (cf. V. 70), come, — [give him] all the
blows his vile hide may bear.' The corrector, however, so radically
modifies the text that it seems obvious, either (i) that the manuscript,
at the time it was copied (for the hands in GIN and ♦GP are the
same), was corrected by comparison with another manuscript contain-
ing different readings, or (2) that our copy was made from dictation,

Herondaea, 173

the scribe not distinctly hearing the words. The second alternative
cannot be adopted ; see p. 183, note 2. The first alternative is sup-
ported by other inserted readings. Now the reading <^^ty could not
have been in the original: otherwise the space after ^ would not
have been made. We infer accordingly that in the manuscript used
for correction the text read :

Metrotime (to her son) . ct n o-oc {anj,

' While your life holds out, you'll have to get all the blows your vile
hide can stand (or that the cowhide is good for).'

There are now left six cases, which upon first examination appear
to be exceptions to the law that spacing always indicates punctua-
tion.* These are I. i and 64 ; V. 68 ; VII. no and 118 ; and VIII.
3. (i.) In the difficult and corrupt passage I. 64 we are not shut up
to one reading ; the space after irpiT^os favors a reading like that first
proposed by Crusius : & wpi^s, i)8c* iarl kt\., or Blass's Soca irpiT^eis,
1780K7V kt\., rather than Bttcheler's or Cnisius's in the text-edition,
although it is not wholly impossible with the latter.* (2.) At VII. 1 18,
the papyrus reads: YUJPH (ii8) APHPeNOTTAH BOYCOAAKTI-

CACYMAC. Bttcheler's ^lapfj | ifirjpev ojrXrj, fims 6 KoLKTiaxis vfJLai —

'scabra congruit ungula, bos pressit vos calce' — gives excellent
sense, besides preserving the punctuation, and therefore may be pre-
ferred to Crusius's ^a>^ | Sifitfpcy SwXjj )3ovs 6 XoKTicnK vfiSs — ' Der
Ochs der euch versohlt hat, fUhrt eine raudige Klaue.* The cases
VII. 1 10, V. 68, and VIII. 3 belong together. (3.) In VII. 1 10 (Ixew
yapf^cvxi yXaatray, rfSoyr}^ 8* rjOfjioy), there is a strong rhetorical pause
before ov^i whereby ovxji ykaarauv becomes parenthetical. (4.) At V.
68 (^tcarrffyn^trBio ovria Kork fivos wnr€pf^ ^duov rifirf), the interesting
pause seems to be a rhetorical one, due to the verb that must be
supplied, of which rifirf is subject. (5.) In VIII. 3 {fj wpoafievti^ irv,
fuxpi axVf^rjXuK A&X^ci | [^rw kJiktov io-Sus), the space is not large and
may be accidental, but a rhetorical pause is quite probable here also,
especially if we read fUxpis c2. It is not unlikely that the passage is

^ Among the excq>tioiis I should not include I. 55 (dBucrot is Kv&JiplriP^a^prf'
fit) ; see below, pp. 187 fi The text at VIII. 28 ([d]wa\ic0ai^ini : J^a^. 2. 7), and
af Proem. 9 (JUvrtpri^yw) is too fragmentary to be taken into consideration.

^ Diels proposes r/H^^it, ^d^cM i^ rtp^lan.

174 John Henry Wright,

slightly corrupt, and that the pause may be not originaL^ (6.) There
remains I. i {ofMOKra, rip ^pf/v^ns' ovk o^ | cT #ctX.).' If we are to
adopt this punctuation as one originally intended in thought, we may
suppose that the scribe, by whom the original was made of which our
manuscript is a transcript, was misled by the form of ns : he took
the pronoun with the ov as an interrogative beginning the sentence,
and therefore spaced it off from the preceding word (cf. IV. 21 ;
VI. 18). Our scribe merely copies what he has before him.

Of the six doubtful cases, then, one is probably due to a copjrist's
mental confusion; two cease to be exceptions upon the adoption of
otherwise approved readings; and three, if not purely accidental,
likewise cease to be exceptional if we admit the possibility of the

^ The synizesis of cv + 17, across a pause in the sense, is not an objection : cf.
III. 81 (rautf'Oi, UaFcU), and IV. 50 {futprvpo/uu, ^yifd ' Itf'tf'er' ^iWpiy nlwij). Bat
the place is otherwise open to criticism, and from several points of view appears
to be corrupt. As it stands it would probably be better to take it as fUxjM ev
jXiof MX^ci (for €v compare VII. 125, — where read r^r . . . fialrtf^ \ BdXrova-aw
€v M * w9cw . . . ffol ^rreir — and for the position of cv compare also Dem. Cor. 144,
cv xpayfUL <rvrr€$4w, or Plat /^gp. I. 329 C, ev ovw ftoi ical r&rt Ido^r ^jcccpof c/ircir) ;
or, possibly, lUxpi 9 cv MX^cc But the synizesis cv + iy is perhaps too harsh to
be allowed even to Herondas. The i>apyrus has only one other instance, ro ftcv
oT/ia (V. 7), but a similar synizesis in Uirr^wa (III. 71) was avoided by the cor-
rector by erasing the v. And at II. 43, where ftdxpit w — a not dissimilar diph-
thong, though elsewhere freely suffering synizesis — is used, hiatus is permitted
(ftdXP^t oO cfri;). Perhaps even t6 ftw aX/uLf in V. 7, is an analogical form, and
should be written t6 fuo aX/aa : compare r4o in VIII. i (rev, II. 98) and a4w
vfiij^it (=aio if rpij^it, Cr.), VII. 96, if reference may be made to so problemat-
ical a passage. If, now, we reject the present reading because of its extraordinary
synizesis, the words will be seen to be an easy palaeographical corruption of
M e X P I COY H A I OC (cf. ndxpit o5 cfri,, II. 43, and dxpif ^Xiof d(hB, II. 88), or,
since that combination is objectionable because it made hiatus at II. 43, and must
not here, M G X P I CO H A I OC (cf r[ov if\]lov «vrroi, II. 13). In the latter case
the CO might have been taken for £0 (Cobet, Nov, LecL pp. 178 f.), and this
easily written into the more familiar £ Y* This process was, of course, helped by
the M6 X P ICG Y» a few lines below (m6cP< •'«'» l>ttt perhaps yjkxp^'i ev). In this
line (VIII. 3) we can hardly make the letters = fUxpit ev, as an hyperionism for
fidxpif o^t under the influence of the foregoing fUxP*- '''^^ (y- i)*

^ Can we take dp4^0'ct impersonally, and read the verse : QpUffffa, dpd0'ff€i r^v
06fnip, rlt ; oiiK 6^c» rrX. ? (Cf. K&hner, Ausf. Gramm, II. p. 30.) It is per-
haps better, however, to explain it as above, if after all the pause be not^ an
accidental one, like those in ov^rot (V. 43), and toO^top (V. 58) mentioned at
p. 171, note I.

Herondaea, 175

use of the space to suggest merely a slight rhetorical pause, such as
was regularly indicated by the oriyfi^ /icot; in the writing of the
Roman period.*

It may, therefore, be reaffirmed with emphasis, that in the Heron-
das papyrus the blank spaces between certain words in the verses
always have the value of strong interpunctiiation, and must be care-
fully heeded by all who would seek to construct the text or to inter-
pret the poet.

But the punctuation by spacing goes only a little way. If it had
been applied consistendy and completely, we should have had more
nearly a thousand than a hundred cases to register.

II. Sriyfuu. — Punctuation is also indicated in the papyrus by the
use of the dot or point in the line. This method of punctuation has
not the significance of the former for purposes of text-criticism, since
it may be in large part the arbitrary work of later correctors or
readers, whereas the spacing must have been made by the original
scribe, and can have been only a reproduction of what he had before
him. Punctuation according to spacings may go back to Herondas ;
but that by points hardly.

Some of the points or dots in the text may be mere blots, — for we
find others like them in the middle of words, or hanging on the tips
of thickly-written letters, — or even parts of letters detached from
the body of their letters on fibres of papynis slightly shredded off.
Again, very frequently the intentional dot does duty in Herondas to
indicate omission or erasure, and is then regularly placed above the
letter or letters to be rejected : occasionally it is also placed, both
above and below, and once in a while at the right side or on both
sides ' of the objectionable word or letters. In the latter position it
may lead to a confusion with the use of the point for punctuation.
Actual cancellation is effected by drawing a line obliquely, or some-
times horizontally, across the undesired letters, syllables, or signs : it
is sometimes combined with omission as indicated by a superior dot

1 On the use of the vAvn see Blass, Griechische PaloiographU, in I. Mllller*s
Handbuch, I«. pp. 311, 312, 323.

^ For example at I. 50 (naraiJC(ov*rvXXov*), where the marginal rpvX(Xof)
is to replace the word in the text. Possibly the point in II. 98 (after ^tfiri) has
a like value; bat its mate is not visible at the beginning of the word, nor has any
sabstitate or gloss been written on the margin.

1 76 John Henry Wright.

(see IV. 67). It is not unlikely, though not certain, that this cancel-
lation was done by the first hand, in the progress of his writing : see
IV. 83, where in eMTT the M is cancelled, and the correct TT placed
just after, though it is possible that the scribe here wrote out eMTTP
at first. The dots indicating omission were added on the revision by
the corrector, who was apparendy the first hand (see on III. 80,
above, pp. 172 ff., also p. 184).

The points as distincdy used for punctuation ^ may be grouped as
follows : cases where they are by Crusius represented by periods, by
interrogation-points, by colons, and by commas. Where the orty/AiJ
falls at the close of the verse it is designated in my list by an asterisk.

a,— Periods: 1. 3 (erUIAG. ; point at middle) ; I. 4 (ACCON'*) ;
I. 8 (AOYAH.; middle); I. 82 (176101.; middle); II. 98
(<t>OIBH.; perhaps middle); VI. 5 (MeTPeUI* with H- written
above 6 ; unless the point here merely indicates erasure of H on
second thoughts [so Crusius, and cf. III. 62], it means that we are
to read MeTPH- with fuU pause, and not MeTPHUJ) ; VII. 76
(TTPHSr: if this be a oriy/xi;; if a line, there are no similar uses in
the papyrus) ; VII. 113 (OUIMeN*).

b, — Questidn-marks : I. 3 (0YPH N. ; interrupted question) ; 1. 3
(CY); 1.4 (TTPOCeAGIN.).

c— Colons: IV. 21 (ArAAMATUIN') ; VII. 114 (TTA5. ; mid-
dle).

d— Commas: I. 8 (Tl-; middle) ; IV. 37 (BATAAHN.).

A glance at this list shows at once the futility of attempting to
identify these points with any ancient system of ariyimi (rcXcta,
{nrooTiyfuj [and /ico^]). They are inserted with little discrimination.
Thus all three are used to indicate a strong pause ; the "rcXcta" at
I. 4 has very strong force ; less at IV. 21. The " wrooriy/xiy " is weak
at IV. 37, less weak at I. 3, and rather strong at VII. 113. Prob-
ably the papyrus is not carefully enough written to justify us in very
nice distinctions between the " fUcnrj '* and either of the others, but a
difference is certainly to be observed between the top and bottom of
the line as places to receive the points. And we must also bear in
mind that some of these cases may well be those of accidental

^ The following cases appear to be accidental: in V. 21 the point uiider fi of
UPoa; and the points on both sides of X in VIII. 42 {o-X'rji), where the sense
demands ovXtu.

Herandaea, 177

#

blots. Ptobably some early owner of the papyrus began with the
good intention of putting the points in {haxrrliax rov 'Hpa)v8av), but
soon gave up the task. It will be observed that eight out of the
thirteen or fourteen trrvy^wl are found in the first mime, and most of
these near the beginning.

11.

The 7rapaypa<^09 AND o)9eA.09.^

J. IIapaypa(/>09. — A short horizontal line, drawn distinctly, firmly,
and usually with full reed, is frequently met with in the papyrus, and
has various values.* Within the verses and between the lines, where
it occurs rarely, it is placed close above certain letters, regularly
vowels : in this position so miscellaneous seem to be its functions
that we cannot speak more definitely of it than to say that it calls
attention to something noteworthy in the letters or words marked.*

^ On this name see p. 180, note 4.

' The sign (*) is used five times in Herondas, and, as — with perhaps one
exception — it is always over short syllables, it may be identified with the sign
invented by the Alexandrine metricians to indicate a short syllable (/3pax«<a,
rpdros XP^"^)* 1^^ ^^S^ cannot be taken as a rhythmical sign, since while
ordinarily in the ipffit, at VII. io8 it stands on one of the resolved feet in the
$4ctt. The cases are I. 50 (6 Moreric[^]jn7f) ; 1. 56 (Ufffiis) ; IV. 30 (rdi^ yiporrd ;
rp^f Moip4(aif) ; VII. 108 {[Sv]w<ur6 fi iXaffai) ; and the puzzling IV. 62, which
has given rise to a spirited controversy (see Cnisius, Philol, 50 (1S91), p. 446;
Ludwich, Berl, Phil. IVoch. S., 1892, pp. 642, 1 349, and L. MUUer, idid. p. 995).
Here the original draft had TTYPACTON. or TTYPA^TON : over T a P is
written and upon T and A stand the marks','. Meister*s wvpaarpoi' is now
adopted by both Kenyon and Cnisius. The first syllable of the word, contrary to
usage, is here metrically long: hence it is marked; it also has the acute accent.
The second sign perhaps refers to the original or natural quantity of the syllable
in w^paypoPf which — on this theory — the scribe must have thought he had before
him, in his original : otherwise the sign is unintelligible to me. For wvpaypoi',
cf. wvpdypTf, Anth, Pal. VI. 117.

'The examples of this sign (') in the papyrus are the following: over iota^

III. 74 (itf' = els); III. 79 (i = e/, followed by enclitic); V. 5 (wpo^offiff =
rpo^dfftii) ; V. 18 {<p€piff=<p4p€is, Cr. : probably ^4p* eff ) ; Proem. 1 1 (rrtov<rt= ?) ;

IV. 43 {iMrlVf a short vowel : perhaps a mark of cancellation ?) ; VI. 25 (Bcrarotf-,
a short vowel); and perhaps in the obscurely written I. 82 {fiti^v—Y)^ unless
here it be meant for the superior dot indicating erasure, the scribe mistakenly
thinking of hi^. The only other cases are: over alpha^ III. 79 (rara), and

178 John Henry Wright.

But the chief use of the horizontal line in the papyrns is to indi-
cate a change of speaker in the dialogue, and in this function it may
be identified with the very ancient sign known as the sopoypa^og.
In cases of this sort it is always placed just under the b^:inning of a
line, slightly projecting into the margin, and shows that within the line

IV. 56 (ffaf«c9), both at the beginning of the line; over T, IV. 62, T^iporrpor
(rvpcrypor?) cited in the previoos note.

It will be noticed that, in all the cases where the sign is used with iota having
the value of et, there exists a second form, with which confusion might arise :
thns at III. 74, c7t and e^t; III. 79, cl and e^; V. 5, irpo^do'cu, not rp6^asu.
Since iptU is at IV. 28 given by €pta', it is probable that a different expression
was meant in the ^epitf* of V. 18, Le. ^p* dt. The form at Pnxm, 1 1 (eriotwi)
is enigmatical. Cmsiiis takes it for irtov^tj but Dieb and Bncheler render it bj
ir4ov€L. Perhaps the mark over the iota merely calls attention to the anomalous
quantity of the voweL

Since the sign C)» ^ ^'^d i^ the papjnras, may with probability be identified
with the Alexandrine sign for the short syllable, one is tempted to connect this
sign with the Alexandrine (*) used to designate a long syllable (Auucpd, XP^"^
dlaiiftct). Bat the data will not support such a conclouon. The sign has not
metrical valne, since it stands over long and short (IV. 43, VI. 25) syllables with-
oat distinction. Nor has it rhythmical valne, since, while on syllables onder the
$4ffa at [I. 82}, ni. 74, 79, IV. 62, V. 18, and Proem. 1 1, in an equal number
of cases its syllables are in the Apea (III. 79; IV. 43, 56; V. 5; VI. 25). Its
peculiar use in connexion with duplicate values of iota which could not be or
had not been differentiated by the addition of an accent, its possible use with
9€ifyp (I. 82) taken in an unusual sense, and its erratic application to other
syllables lead one to believe that, as inserted by the scribe of the papyrus, it was
nothing more than an intermarginal "obelus,^ intended to call attention to
dubious or peculiar forms ' and uses. Unlike the " obeli " discussed below,
these cases were probably a tradition firom the original manuscript (see pp. 180 ff.).

Several apparent "obeli" of this sort require attention. In L 7 the mark
after xaXi is hardly a ** paragraphus '^ (Crusius), at least in the sense of a sign
indicating punctuation; it is rather part of the upper bar of the following r;
the papyrus fibres (vertical) have shredded loose at this point, and sagged down,
as a comparison of the writing above and below will demonstrate. — The mark
over the first a of Mara- in L 50 does seem to me an intended': it is rather a
thickened fibre of the papyrus. — At V. 17 (fu&pa), the mark is probably an acute
accent, the scribe taking the word as fu&paWf not fuipaw, — The peculiar line
over the first w in VII. 77 (r^r riftov) is nothing more than a part of the follow-
tng r. — In IL 73 the line over the much blotted r (?) in ^tXA[]T[ ]t must be the
remnant of a letter suggested for the place, perhaps a sprawling r. — In I. 54 the
line over r in T[d ffaX6r] appears to be the horizontal stroke of a r begun too
high.

Herandaea. 179

above, or at the end of the line, there is a transition to a second
speaker. In this place it never has any other meaning. Not taking
into account the ornamental forms of the sign found under and
adjoining the closing lines of each mime,^ there are sixty-three cases
of the use of the -rapdypa^ in Herondas to indicate change of
speaker. These do not, however, comprise the total number of
necessary changes of this sort; hardly more than from sixty to
sixty-five per cent.

The -rapdypa^ indicates a change of speaker at the end of the
line in the following verses : I. 66 ; III. 70, 76, 83, 85, 86, 88 ; IV.
18 [?], 3^1 S^f 53. 71, 78 ; V. 3, 7, 9, 18, 19, 25, 28, 34, 36, 38, 39,

62, 68, 79, 80; VI. II, 21, 26, 36, 56, 73, 78, 79, 84, SS, 92 ; VIL

63, 76, 78, 82, 90, 92. It indicates a change of speaker in the middle
of the verse, there being none at the end, in I. 7 ; III. 58, 81, 87 ;
IV. SS ; V. 73 ; VI. 19, 20, 22, 23, 25, 47, 97 ; VIL 3. At II. 78,
it shows, like our marks of quotation, that the speaker has finished
his own remarks, and is now about to introduce a citation from the
laws of Chaerondas.

Only at one place, out of the sixty-three cases, is the -rapaypai^
certainly wrongly applied : at V. 55 it comes a line too soon. (At
L 65 it also comes a line too soon, but it is there cancelled, and
given correctly below under line 66.) In the distribution among the
speakers proposed by BQcheler and Crusius, though not in that pro-
posed by Rutherford, it would seem that also after I. 81 and IV. 34
the trapaypa^ had been wrongly used. But in view of the number
of correct examples and of the nature of the blunders made, we ought
to be slow to admit exceptions here. The verse I. 82 may well be
put into the mouth of Threissa,* and the words in FV. 35-38 could
have been said by one of the maids. At all events it can be urged
that such was the distribution of parts in the manuscript from which
the pap3rrus was copied, though this may not have been the original
intention of the poeL

1 At the ckite of a book the ornamental finial sigin waf called the icapi^plt i
Isdore, Ortg^, I. 21; ct Bfaut, CrUch. PaUuographu, p. 31 1. In thli papyrus
it often resembles the ItrXS^ Ar^pUrucr^t, with additional Bourishes,

* This has been proposed, on other grounds, by O. Ribbeck, Khsin. Afustum
47 (1892). p. 629.

l8o John Henry Wright.

There appears to be little doabt that the mark (~) abo^e letters
and the vofldypoi^ were inserted by the first hand.^

II. 'OyScXoc — With the wupaypa^os mtzst not be confused other
short lines — not marks of accent or of quantity — found both in the
text and on the margin of the papyrus : they are usually drawn from
right to left obliquely downward.^ Wlien placed in the body of the
text, a line of this description — sometimes here taking a horizontal
position, and ordinarily roughly drawn — actually cancels an objec-
tionable letter or group of letters ; I think these marks were, a:i a
rule, made by the scribe in the progress of his writing, whereas eras-
ures suggested on the revision are designated by the superior dot.
In one place this mark appears to cancel a ^ultily placed mpck-
yptuf>ai (L65).

But the chief function of this obliquely drawn line is to call atten-
tion to verses ' requiring examination for one reason or another : and,
since in this function — though hardly in its form — it resembles the
o/Sc\o9 of the Alexandrines, it may provisionally receive this name.^
In these cases it is placed on the left margin directly opposite, or
near, the first letter of the line in question. While it signalizes many
verses it by no means calls attention to all corrupt readings or
obscure passages. In many instances, if not in all, it appears to be
the work of the first hand or of an immediate contemporary, since
it not seldom calls attention to omission of letters, or to incorrect
letters, where the correction is made by the first hand. But not all
of the corrections that it points out as necessary are actually made,
nor when made are they invariably in the first hand. The cases of
the use of this obelus, which is extremely important for the text-
critidsm of our poet, may be grouped as follows :

a, — It designates verses where Hr/ttrs have been omitted^ or

^ The nature of the blunders made in inserting the irap(i7pa^ appeaxs to
prove that the signs were copied by die scribe after he had mitt e n a considerable
part of the text, and were not doe to his own conjecture.

3 In rV. 51 the line has the opposite slanL

'In 11. 36 the mark is placed opposite a word in die v ene (oucuur). This is
the only clear case where it is found not in the margin, in this function.

^ It is probably forcing language a litde to name this sign an ^/ScX^r. The
obelus of Homeric and Platonic text-criticism was used distinctirely to indicate
athetesis; combined with other signs, h o we ve r , it had many other vakus. Thns

Herondaea, l8i

wrongly given, in the first draft, but are supplied or corrected either
by the first or by a later hand : II. 3 (NYN becomes viyw, with H
written above by first hand) ; III. 45 (HMG0A — 17/Aai^a, Al above,
late hand?) ; III. 46 (KAAIOYCAGKACTOY— A before 6 erased
in line) ; III. 80 (^GP OCAC — ^^cpawxra?, av written above, first
hand ; see pp. 172 f.) ; IV. 10 (lAGLU-A above the A, first hand?) ;
IV. 67 (avJACIMOC-CIAAOC, first hand?; earlier in the line also
erasures by cancellation and superior points) ; IV. 76 (after GPFA,
TA inserted, late hand [Crusius] ?).

b. — In the following, marks of accent are added: II. 83 (KAY-
TOCTACAYTOY0AH ; acute on first 0, circumflex on H; also
coronis at bottom of line after TAC : all probably by first hand) ;
III. 6 (XAAKINAA, acute on I) ; and VIII. 14 (ANNA, circum-
flex on ultima; first hand). See also IV. 2, under d, below.

r. — At III. 49, KAAH0IN UJCXe, after N above the line, in the
first hand, a coronis is inserted, probably in first draft ; also at II.
83 (see under b above) .

d. — K short vowel \^ designated as such in VII. 108 (6 A AC A I ;
a " over A, in first hand?), andat IV. 2 (TTYPArTON ; a short " is
put over A, but at the same time a P is written above the T ; the Y
also bears a" : see p. 177, note 2, above) .^

e, — At the following places a corrupt text is indicated but no
attempt is made to correct it, either by the first hand or by later

among the rk rapariBifUPa rots 'OfiripiKoit vrlxoit 'Apiffrdpx'ui fftifuia we read of
the obelus, figured as a short horizontal line : 6 Si'6^\bt xp6t rd dBerovfiepa Irl
TttQ woiifToOf iyovp 9€vo$ovfiiva ^ inrofiefiXrifUva (Osann, Anted, Homanum, p. 3;
qL Naack, Lex. Vindob. pp. 271 f., also 274 f., 277 f.). In Plato texts: 6/3eX5f
Wfi^ T^p d^/ri^tr * 6/3eX5s wtpucTiytUpot xpbt robt tUalovt d^cr^eis (Diog.
Laert III. 66). Our ** obelus '* in lierondas may sometimes be used with this
Talne, bat probably not : its various uses, as we have surveyed them, better fit
the Aristarchean dcrX^ direp/tf-rurrof (irp^t rh. ivdwria xal /lax^f^i'fik, Kal frtpa
ex'^futra rdfircWa xal irfrijfULTa), It also differs from the obelus of the classical
manuscripts, in that on our theory it is merely a conventional sign, originally
adopted by our scribe and used by him as a memorandum, whereas the ordinary
obeli represent a tradition of literary criticism going back usually to the Alex-
andrine age, and were copied from manuscript to manuscript ; cf. Weil, Melanges
Craux, pp. 1301, on obeli in the MSS. Z and B of Demosthenes.

1 Except at this place, which was probably obelized for other reasons, no (')
is found at all in obelized verses. This suggests that these (*) marks were in the
tax before the obeli were written on the margin.

i82 J^m Hemy Wright.

hands: V. 59, Rndierfdid supplies <9)c ; VL 65, 0IK6IN conec ted
\yf Cnisixis to oucofr [ie. ouciufrj, by Rntherfdnl and ocfaeis to oucofr \

VII. 35 a fragmentary Tase : obdns of pecailiar finrm, inseitcd at
fiist diafk? ; VII. 46, if not a grave accent, the obdns caOs attentkm
to ambigiiaas groaping of letters ; VIL 8Sy 96, cormpc lines : restoiafc-
tioo uncertain ; VIL no, end of fine nnfntrffigible to scribe : prob-
ably ^j/^pim ; VXI. 126, conecikni is attempted bat left incooqilete :

VIII. 21, fragmentary line.

/. — In three i^aces there seems to be nothing the matter with
tfie text ; all of the fines, however, appear to hare somediing inter-
esting to the scribe : IV. 32, its ambigaoas coostmctioQ ; IV. 50^
periiaps, its droQ Homeric reminiscence, and VIL 71, the extraordi-
nary form of oath.^ Except for the coosstent and exclusive use of
the obliqae fine elsewhere to indicate corruption of text, we might
infer that it was hoe used, fike the haf)ai hn^irnxn^ ascribed to
Aristophanes of Byxandnm, to point out passages remarkable %2x
some reason or other. Otherwise it may hare been only an accidental
memorandom sgn arbitrarily adopted by the scribe when engaged

in.

THE EARLIEST CORRECTIOXS.

The sabject of the earfiest corrections in die popyrosy — dieir rda-
tion to the original, and to later corrections, and their chronological
sequence — is important in determining the character of the original
of oar copy and that of other manuscripts possibly nsed for coQation.
It is at the same time an extremdy difficnit subject, especially when
investigated through the medium of a photographic facsimile of the
popynxsy in n^ch many pecofiarities of the original £iil to reioodnce
themselves. My remarks on thb subject are offered tentatively :
diey are doubtless open to correction in detail, though I trust not to
serious modificaticHi.

^ PosBibiy, however* we onght to nile out tkcse abo. For in IV. 50 there
to be a mrii»di tfter im^rr^ which, howerer, wxf hare been put in, not as
a piim tiiaHo tt mark* but Hke die dvo^MvrvXif — to be sore, not dsewhere occur-
nxtg — to mark off the r from ^lu^m. (c£ Pnttm, 11, fiir the only odier certaiii
nBtance of thia sort in the papyma; bat see above^ p. 170^ w)te 2). And in VTL
ji, atteiitHin xoxf ha:ve been merdy caOeii to tti^ as cfifferent from die ^o, with,
drcmnflex accent, eisewhere fiiimd in the manuscript {t^. L $5* IV. acv etc.)-

Herondaea. 183

An independeiit examination of the internal evidence available, in
the collection of which the notes of Kenyon and of Crusius have
been most helpful, makes it possible for us to reconstruct the early
history of our papyrus somewhat as follows :

The scribe had before him, as the original to be copied, a manu-
script in which the verses were written line by line, with occasional
spadngs to indicate punctuation; it was also provided, at least to
some extent, with diacritical marks — -rapdypai^, the signs',**, but
not completely at least with signs for accent. It was written on the
idiole legibly, and in a style of writing not differing essentially from
that of the papyrus, and exhibited peculiarities of orthography such
as prevailed only in late Alexandrine times, and afterward. Not to
take into consideration the perishableness ^ of papyrus manuscripts
whta much used, this original could not have been prepared much
before the first century b.c, if even as early as that

This original manuscript the scribe now copies,' with reasonable
fidelity, cutting himself a new reed once or twice. In copying he
makes mistakes of various sorts: occasionally he unconsciously
changes the Ionic forms of the original into the more familiar Attic
forms,' and sometimes slightly blunders in his grammar and S3mtax ; ^
here and there he appears to be carrying the thought, and not the
exact words of the original in his mind, and thus when he writes he
unconsciously substitutes a new word for the word first read;' of
course he makes mistakes in reading the letters, and occasionally
gives us nonsense, and also writes verses metrically impossible.

^ Cf. Plin. M //, XIII. 83 : he speaks of papyri two hundred years old as /mm
gimqma mcmmm^nia^ rarely met with.

*That our copy was not written from dictation is clear from the nature of
lererd bhmders, where the forms of the letters, not their sonnds, are misappre-
heiided: e^. L 2 (ATTOIKI HC for ArPOIKIHC); I. 76 (AI for H. in A«-
«»); V. 65 (GAGIN for GAGIN : A#«iir); ill. 19 (AAI«^op«#r«^ : AI for
AI : Le. M X^mp^tptu) ; IIL 34 (AT for AF in aypev) ; IV. 94 (Aim for Aaac).

» For example: I. 39 (x^fttpa^ for mifttpoff); II. 7 {[r6]Ki%M^ for [v6]\u)c);
XL 36 (•UEuiF for •uMifr} ; III. 59 (rov for icov) ; V. 63 (av^ir for avrur), etc.

* He makes \i0ot masculine in IV. 21; writes aorist snbjoncthre for future
indicative m VUL 3 (MXf 9 after fUxP^(i)\ and present su^nnctive for aorist
optative in IIL $2 (fioKKt changed to /9aXoi).

* At II. 64 be writes ftoipaw, but at once changes it to /air$ow by drawing his
pen through the middle letteis and writing t^Bo over oi^cu At III. 82 he wrote

l84 John Henry Wright.

Some of his enors he detects just after they have been made, and
these he corrects on the spot, either, when possible, by changing the
actual forms of the letters, or by drawing his reed across the wrong
letters and writing the correct ones just above. In the actual prog-
ress of writing the first draft he probably does not copy the n^-
Y^44M, possibly not all the diacritical marks, and certainly not all the
accents.

His draft now completed^ he takes it in hand for revision. That
the original scribe revises the manuscript, and not another hand, is
clear from the handwriting of many of the corrections. At first he
carefully collates his copy with the original, and corrects innumerable
blunders. It is at this time^ that he puts in the rapaypfs^akj and
some of the diacritical marks : letters and words to be omitted he
now neatly indicates by putting points over them ; letters or words
to be substituted he now writes in between the linesy just above those
that be had mistakenly written. Some of the errors or obscurities in
his own written copy he cannot correct fix>m his original : in these
instances he dashes an ^ obelus ** in the margin to mark the vexse as
one requiring subsequent attention.^

This collation now finished — a hurried collation, siiHre he leaves
a number of corrupt passages^ not only uncorrected, but also un-
noticed — he examines the ^ obelized " lines in detail, and here for
the first time appears to have called in the aid of a second mann-

iratt<tf Jmu of TuiC^i probably tbiokiog of what be had w iilieu at 65; for the
form, cf. Anin. Fai. XI L 2t 1, Amztre^ni. ^» S*) : the correct word was wpi^m.
At III. 05, wh<!re he tir»t wrote 'm^ikm^tjt^ probably fuUowing his copy, he at once
changes the wotii to ««i$>u>, apparendy a sudden coQjecenxaL emendation sag-
g«:sted by the comext; -w^iunKi^ is more ;>r»>bable: cf. CraaitB >ui mmt.

^ The tact that the v^M^^Mi^ot are twice put in a line too soon saggests^ that
the scribe's, eye ran >iown the coiumn a» he inserted then, and thia woald nut
have been the case if he had written them in each time after writing the line
V^cf. l. 05, V. 55 \

^ Chides wiiere the obeiiied Unes contain conrectiottft^ certainly writtes by the
tkst hand ore II. 36^ TIL :k>, IV. ti and 67: perhaps- also IV. 70. Thcfe is
unctfrtainty about some of the other ItaeSk

It mig^ht be utged that the obeias was inserted by a late hand tn caU atten^
tion to muA^h-corrected verses^ 3tu it may be replied, tirst, that the obeii hare
the charaLCtri^tic^ oi the tint hand. anU, secondly* that many -)ther rerscs soow«
ioi^ much greater c*>crectioo ate not obeiieed. The expiaaatioa gtven above
acc*JuiKs :oc ail the poeoomena; the other one does noL

Herondaea, 185

script : i.e., he uses a second manuscript only to correct otherwise
obscure passages, not for the purpose of preparing a critical edition.*
In this second manuscript the accents in particular were more fully
given than in his original, and the reading of the text was different
in a few places ; for the obelized Unes in question he adopts the read-
ings and corrections suggested by the manuscript, though occasionally
he appears to reject them on second thoughts.

From the spasmodic way in which the o-rty/juu are put in, we might
infer either that the scribe began to copy these marks while first
writing, but soon wearied of the effort and gave it up, only now and
theii later in the progress of this writing copying a oriyftiy, or, what
is more probable, that he or another later hand at a subsequent time
began, but did not complete, the task of punctuating with the points.^

The following examples, taken with those mentioned above and in
the notes, will at once bear out and elucidate some of the positions
here taken.

IV. 83. xKAAOICieMTTPOIC- At first examination and compari-
son with his original the line looked £aulty, and was obelized, but on
doser comparison he found that by inserting t after TT it became intel-
ligible. He thereupon cancels the obelus. (On the M, see p. 176, top.)

IV. 10. Here he had written IA6UJ,which could not be right Appeal
to the original failed to solve the doubt. An obelus is dashed in : on
comparison with another manuscript, or perhaps as a result of his own
conjecture, he now writes IA6UJ.

1 1 1 . 36. 1 K t A N . After comparing his original and correcting A to H ,

^ Except in obelized lines, there are no first-hand corrections in the manu-
icrtpt that most be accounted for on the theory of an appeal to another mana*
tmglL (For in VI. 38 koMv for 99^9 is in a later hand, and in I. 15 — fwt,ocop
— the coronis was inserted merely to indicate an elision of a (f.^., not vl6s), of
course not to differentiate fwV 6cow from the other reading fws 6cop, preserved in
Ttrions prorerbial forms, here given in the margin in a late hand.)

» If the insertion of the myfuU had been undertaken by the scribe, it prob-
ably would have been carried out to the end, as were the other parts of his colla-
tion. He conU hardly have inserted these marks, at least at the earliest stage,
except as he copied them; but it is hardly conceivable that the original manu-
scripC could have been as erratically punctuated as the earlier my/tal indicate.
The points were certainly put in after the verses were written, since no space is
allowed for them. In view of all these facts it seems more likely that the rrtyfuU
were, in the main, the work of later owners of the manuscript

1 86 John Henry Wright

the word is still puzzling: he obelizes it; later, on comparison with
another manuscript, he inserts the acute accent, which shows that this is
o2mV> 1^0^ oiicciV (cf* VI. 63, and p. 182).

V. 19. AO is corrected to AG (Sovfuu, i.e. Sco^uu, to Scv^aoa) : hence
and 6 in the original manuscript must ha\'e resembled each other. See

on VUI. 3, abo\^, p. 174* note i.

YIU.6, KAIACTHCON. This reading, suggested by the &m,ft at
the beginning of the line« is on revision seen to be ^Ise ; the scribe points
C H and C, and changes T to Y, restoring the correct reading 2^.

HI. 45. The scribe wrote HMe0A (for ^^adh:)y probably through
association with ist pi. mid.> and not because he proaoonced 6 and A I
alike ; the latter is not to be expected in a manuscript of this date, and
there are no other cases of this confiisioQ in the pap3nrus : of coarse eariy
8 1 is often given by 1^ and not sekiom even 6 1 as written is oonected* by
a superior dot> to L The conectioik at III. 45 was made by a later hand.

The manuscript^ thus prepared for use» passes into other hands.
In its later history it soffeis more or kss modification. Ezzors
previously undetected are now corrected (IV. 61, So^ etc.) ; coi^ec-
tuzal emendation is attempted* somedmes mihappihr. Readings,
interhnear or maiginal» are appaiendy imported firom other mann-
scripts* firom Hecondaean qootadons in other authors* or* in the case
of some proverbial expressions* fixxn variant forms in literatur e or
hie. The glossator appears widi his bits of scholia* very foir in mnn-
ber* and in abbreviated form.

In making this attempt to ascertain the oldest accessible ""a^ing^
on record or reasonably to be deduced fixrm the record* we by no
means would assume that text-criticism shouki cease upon the com-
pletion of tiiis task. Indeed the large work wiE yet remain of tracing
the text bock to the pen oi the author* and in dns more interest-
ing work conjectural emendation must play a large part. Bot the
conjectural reconstruction of the text can never sa&ly begin until
the utmost possible has been made of the record*

Herondaea, 187

IV.

l,4>priyi^ IN HeROND. I. 55.

The facsimile of the papyrus at I. 55 reads :

KINeUiNAGIK-^^-^^^HPIMN C4>P^C

The gap at the middle, between r and t), in which there is room for
from seven to nine letters, has been filled by Bticheler and others so
as to read ^ucr[o9 vol KvO]rfpirp^ ; ^ by Crusius and others, a^ticr[o9
is Kv$]f}pirp^. The latter is palaeographically more probable. The
close of the line is universally understood to be <r4>prpfk ; but the
traces of the ink quite as well agree with (r<^piy$is, or even possibly
with a^piy^ The very distinct break in the continuity of the
writing before the letters <r^p shows that there is a pause in the
sense at this point, i.e. that the last word cannot be taken closely
with the foregoing. It is mainly in the light of this consideration
that the interpretation here offered is new.^

Now a^piyo), with its short penult in classical usage, is impossible,
and is hardly to be justified by Oppian, Cyn. III. 368, where
o^ptya^ might be read for MS. <r4>piy^, or by Draco Stratonicensis
(p. 119. 7 Hermann), who gives o-^plyui in a list of words with long
penult, — a list teeming with demonstrably false quantities.*

^ The final letter is probably ff, bat it may be a blotted t.

' There are traces of the a* of dducTot^ and Kv$ is fairly certain. The space
between this a* and Kv$riplrip appears to me much too small for rai, at least as pal
is written a few lines below, and elsewhere (I. 66, 86; VII. 71, etc.).

* Ratherford has proposed A$un'09 iiip KvHipris * J^f, ff^fnfylt, but it cannot be
wholly right : it offends against the metre besides being too much of a departure
from the clear traces of the letters on the papyrus. All other editors have com-
bined €^pfnylt closely with the foregoing words : either with Mucroif or with 1^
lL»$npliit (B&cheler's first proposition).

* If €^p€yi^i were possible, it would refer to the manly vigor and strength of
the athlete Gryllos, lover of Metriche. In an epigram of Leontius we read
of an aged athlete vanquishing his vigorous younger rivals : rpi^^vt 5ri ff^fnyhtap^
rat ip trwoddfuf rXiop dX«$ | puc^at, A nth. Pal, XVI. 359; cf. also ti^xi ^^fH'
"yQpra ituropt&oPTtUf said by Achaeus xwpl riff edc((at tQp dffXrrrQp 9irfyo6fiMPOS,
Atbcn. X. 414 c, D. (Nauck, p. 747). — Z^piyr^it, if admissible, could be taken
either as a parenthetical interrogative (like 7cX$t in II. 74), addressed to Me-
triche, — * Don't yoo glow with desire? * (at this description) ; or as a parenthetical
ffonark — 'Ah I yon glow with desire, I see.'

t88 John Henry Wright.

Rriecttng ir^yiiif or o^yijiy and accepting the reading cr^pfyts,
we ha\*e yet to find a wholly satisfactory interpretation of the word
in this context It is possible in classical Greek to nnderstand
ir4»>f)«t ^<r^fM>it)» *seal«* in the literal sense, as either the metal
^e^l or the stone ^with inscribed device (mt legend, kx nninscribed)
v^r as the impression made by whatever kind of a seal, often also
exptrssevi by ^^^^y^i^pm} Horace's ^mAx x^Z&i fmdico {Efist I.
^lX ^>« citevi by Blicheler« b hardly apposite, at least in the meaning
attav-lievt to it by Hv^ace. Here the n^terexxre cs to seals impressed
u^vHi the l\Anl^t \Uvrs v>t the apartments of the chaste one. who
x^'C^-.^^t^ tn the prxectioa assured by rfjezi. Horace probibly had

t» tn;:vd $;x^« |\Jk<sSA^^ ^ as Ar;StOf ^Wi:ae>'5 raw ynrn AM B ximia \ v^^^arfir
&t^ ^Tx^VVdv^fvw r*^ woi^ •^\X««V x^ t m n z^ yiK? \Tkesm, 414—6) ;

<^<% ^^(MMur<Ni V ri <r. 5;«* E^. tCl^^. o X£3C^ '• : cc Licoftiizvs's rk €
«\Xa lW«iTv>j<««<vK «L'^«yr^vi( lioAMM r«ij«r;«« awteM 1 Aicx. 50S. wiieze
?>^e j%K*» t>>c S»^V4»ft<^. >*t Vjcv^y t2>f Tiftsssxre in Her.3Bidt^ wiiich

Xt^^rrs^V s>wiy. x"^j&«^ K** >e *m*it.'^ is^ dy^^s ^vncgnrian jmu azsnd.

TV iTsc *x #-<(6wr^>c ir. ?V ^sfTssr *t ar nnnc soonf — "^a fiesa far
Xr^hrosl^r^rS so^-vv ' ,1K.. Fl)is> .^^i»:i»:.7 H-n.-^- TSfidiete > firs: nrcnv
^>?*:v*n' --?s TvtssiMic hcrsc. iuR hayd^v csraur*. iii rirw of ccber

■ V,>v v »Kc <:Miifn*^Vfs v*vv 1/ tSt ^t^yrtT?*: nvrix a: satn. hic ^isr: art &
^^"^ t- K-'N: »Kc wfKH*: ^i! ^*v is m^r^n; -.>* itvr^TchAti;;tpm;tfasissi& a:" the twr ssnsss
*>."v*ii ■•'*•. ' i'n''*"'e*^«k>r «?< ^vr ir VoK>r»h. fi'l^m, • ^ ^ ati^ ATI. l 30:
;•> tK »\»«^« ■««k««t*«^.r«% \t tK H\*'9': f«>m>h »»t as?*- a: hk RnrirssuSL d

Herondaea. 189

more probable possibilities. If the lexica and word-lists are to be
trusted, this sense of <r4^payii is mainly petrographical and technical,
and not popular. There remains to be considered the interpretation
which takes the expression — dBucroi h Kv$rfpir)v a<f>priyh — in a
figurative sense, ' a seal unbroken in love/ or ' a seal of inviolate
virginity.' In support of this view of the passage Crusius cites
Nonnus, Xwrofuvif S* aalfavarov c^ o-^paylSa Kopurii {J?i'onys, II. 305),
and compares Paul the Silentiary/ ^(pva'toq d^voroto Sicr/Aaycv oft/xa
Kop€UMS I Zcu9, SiaSv9 AaFoas ;(aAxeAarovf ^oAo/iovs {^Anth, PaL V. 217 ;
also Suid. s.w, Kojonw opoi, a/A/Aara) . These examples appear to be
very apposite, and almost silence objection, especially if we group
with them the o^ucrros . . . <r4>payU of Lycophron. But they
obtain compelling force only on three rather violent assumptions,
viz. ( I ) that the expression * inviolate seal of virginity ' in the words
SBucToq (o^avoTos) cr^payis with some word for love or maidenhood,
had become a stereotyped phrase in early Hellenistic poetry ; (2) that
as such it was here used by Herondas, and (3), that as such it was,
centuries later, reproduced by Nonnus and Paul The truth of these
assumptions it will be impossible to demonstrate, at least from these
examples or from others like them. No one would dream of turn-
ing to Lycophron as a mirror of current usage, and both Nonnus
and Paul, Christians of the fourth century a.d., are quite too far
removed frt)m the Hellenistic age to require us to explain the
phenomena of their art only on the theory of an imitation of Hellen-
istic models. The collocation SBikt<k cr^payis is not in itself so
extraordinary as to require us, finding it in Lycophron, to view it as
already a stereotyped one, or to prevent our taking the words sepa-
rately under some circumstances. The words oOikto^ (o^votck)
o^pi7yt« vapBtntfiy Kopuji^, ^^ ^^^ ^^» ^^ ^^^ occur in the Anthology,

^ It is not impoisible that the receiyed text of this mach-qaoted epigram may
be incorrect, and that we should read xpvcreof dSpavffroio 8iir/iay€if Afifia xoptlas
for d^avrroio. This is the reading of Cod. Leidensis of Soidas, s, Kdciow, though
elsewhere we have d^avrroM. Prot>ably the situation is conceived by Paul in
this epigram, about Danae imprisoned in a tower, much in the way that a cor-
re^xmding situation is represented by his contemporary Agathias in AtUk. Pal.
V. 294. 19, l^oXdro^ ^^iff vvfTfiitiUL Kopelns, and a classical adjective for
vipyt^/ui and a word used in the sense of Tvpyufta is ABpav^rot, rather than
A^oMTOf : £nr. //re, 17, Tvpyoi ddpoiwroi.

r I- 1 » I I I I ;.

190 Jnim H€mwy Wrigki,

wEucjic if t3BC cx|vcsBK3ii nMi ocoomc
St vrasM cetxi^ihr lurre liceii receded, so
tius v^:^ ««I1 ctltl &v k: iu&ecd, t^ iio q um i ci of the
^i^ktt ssa&iir MM v^^bpMs KudcB m
cc p&nmes vs^ v^di^KyeK. Is secsK to
£i^fl*cssBCi& «itapirf^*c r <hp« jwi ^Tss iuiggiused to XonBas, if ws/i bjr
LvvccdairaaL br csannesfi itsace ai ias o«« 3aBe;3 ii 'v^ack die rad
^iftd ^lOKd. ixrpeOr i^ipcnu;^ <I>rygiKa lufiwmiy , isnnr ncv

<3ic» t^ THBi^nikx^ Tnftnmrgif 3£ juct ^ lin!«jc:i anx^ck ^doi the

<c Xoaoms^ joi^ ^ AaciruiKS' cc Si3ciil nudxQc iiiiriinr^ siese md <fdoa
IfiCif m ^' .c « s . imi nee necessssrDr m^iiss of si^ VimiiHiVu ig j^ ; be is

e;)ii?csLif<xc t:" ^^ ^fc t^ r iit ^ iL^rfh|»n* f*«MirMsK i£ fiaondBs Iv

nt^oss' Tc.^xss :^uc tbf Isosr snsc ^ TMrrr. in ^le scbk of

:s^it ^mriiwucva: Ix i [ < n ^* i» v^f^vi^ «vqw; a ^ ' ik &ii^ ItaL A

^ •

Wi ^N *V tVxsSs H >^

Herondaea, 191

795 ; yuKUKOf 0iyciv, Eur. EL 255 : and in the gloss S$iKT(}i * -fj irap-
$€vo9 in Bekk. Anecd, 828, where the word is quoted from Araros,
a poet of the New Comedy, the reference is, of course, to a maiden.
Tliese and other examples justify us in taking o^cktos cs KvBrjpirfv,
like aBiKToq Kvirpt8o9, as ' [hitherto] untouched of love, heart-free.'

It may be that in the appended <r<f>p7iyU we have only an emphatic
appositive, — * untouched by love, — a very seal,*^ — but I am disposed
to believe that there is here an added thought, coordinate with the
leading expressions : viz. the thought of secrecy which often attaches
to <r4>payh and its derivatives, rather than that of inviolateness or
purity. This sense — not sufficiently noted in L. and S. — maybe
illustrated by the following examples:* o-^/Miyt^c t6v Xoyov <^*727>
Solon afi. Stob. Serm, III. 79, p. 87 Mein. ; ^p-qriav iirttov yXdnray
<r<f>pfrfyU iviKutrOta, Lucian of. Anih, PaL X. 42 ; a[\Aa S^ Oavfuvra
iroAAA <roifty <r<f>piiytairaTo criy^, Nonn. loA. xxi. 139 ; x^iXco-i 3* a
^ayyounv lir^fr^fnjyio'axiro frvyrfv, Nonn. Dionys, XLVII. 2l8 ; dXXa c
tk^yrq ypXKMfi ^WSiyo-cv vn-o tr^ptqyt&a. ciowr^, Christod. Ecphr, 31,
Le. Anth. PaL II. v. 31. Probably it was in large part the idea
of secrecy associated with the seal that lent special force to a<f>payi^
and its derivatives in reference to the Greek mysteries : e.g., iwi-
<r4>pay£Cw$ouL means 'to initiate,' 'to make one of the fiwrrai (fivw,
*to be closed').' Of course the term has chiefly the connotations
of authority and completeness, and these meanings develop especially
in the numerous applications of the words to Christian usages. (Cf.
Steph.-Dind. Thes., s.w,) ^

This interpretation — whereby <r^pi;yis is understood to suggest the
idea of secrecy — is quite in the spirit of Herondas. It furnishes
an additional example of a motive elsewhere found in the mimes,
that of caution and silence in matters of love and intrigue (I. 47,

^ To PAuI the SilentUiy the expression might mean ' untouched of love, yet
bearing love's own image or seal * : cf. r^w Tplw iww^p-i^ifftw 'Epwt[6ptur^]iU6pa,
Antk. PaL V. 274. — Rutherford*s {f, ff^pijylf, * look, his seal/ is rather abrupt
and har^ but it has the advantage of preserving the punctuation*

* In Aeschylus the same thought is expressed by kXtSs : dXX* Ifm xd/Ml kkijt
iwl yXiiv^if 0^Xa( (^Frag, 316 Nauck), with which compare Soph. O, C, 1052,
^psroSrir Jr koX x^utf'^a icX|^f ^2 yX^aio ^4^m TpoffT4\<aw EdftoXridfir, and
^V' 849. 2 Nauck. Ct Lobeck, Aglacph, I. p. 36, note. Ancient rings made
of key and teal combined have been sometimes found : cf. Daremberg et Saglio,
Did, Ami. I. p. 295, >^. 349.

V^ y^, j: ft
I 4 fif^M^tar. toe

M4ifMMe* — avc

IffM/Xffj, ii* jinmyiHim Hi ir smiiasr msE^ a: m. «, r- h

^/ 4(#e yisqnPTvt woatt imn alcnr is. -n: 390: ^nnsr

pHf^y^ t. IM^^HtlVe furs tUSMt ^WDIUL ie ni nXMCCXinC 1X> rimiM»iMm|;

Mlt0-$¥^ iOiC '0<itpirva&, ' imiiraet ssal * lac tosat Ji^iji g a r id iie
^j^M^i>v^)MipfiiiaIiD on: vc -ne gwrsTinr.

M-C'^> suff:?* jjd

't$$0, V'iwaraB as Tbwx. /x ML 125 is Cod. Ab2k. 222 ii), js

fM ^Micatirxi of the Aisbrona Scbosi. Mcsieke Ind alreadT pro-
1^,^ t/> em<mi tbc rTiIgztc to ItOw if S^^m., 'A^tf«r ^r^l^r^
( \M fe^^ding, apparecthr coc ili en ed by that of Ambr. i; viiere, hov^
^r^f, 'Apsroc irrc^ttm^ stands (not 'Afimrwr mmpmn^)^ has been
H4itidtdf as defimtdj rslahifshfd, br Zaegier. UiUer. Maass, and
ttfhen. It has been suggested by HiBer' with much ^rfansibility

I The gioM in Diogentauni (VL 67) on Ae pf oiqbul f iffni i w . . . ^i^
§i0f0t MiPtU, h hrl r&p ii^^vp. Siudas has iwl rm irrx**-

' Thif reading, at lout 3f/6X«nr f 2;i#Mr9, is giren also in Par. L (Reg. 2S31).
* On this theory of Hiller, I tbotild be disposed to explain Zitaup as originally

Herondaea, 193

that a Simon might have been mentioned by Aratus in one of his
lesser poems * as a rival in love, and thus may have been regarded
by the Scholiast as identical with Molon (17 Si/ion^). Meineke's
suggestion that Mo\<»y in the text of Theocritus is a corruption of
^^MK9 is hardly probable in view of the impossible quantity of the
penult of the latter word.

The viilgate reading goes back to the manuscripts used by Cal-
lierges in his editio princeps of the Scholia (Rome, 1506) ; these
were several in number {Ik &a<^ipiay ^m-iypaffHuv) , and at least one
of them appears to have belonged to the same family as Ambr. k*
If we bear in mind the easy confusion of the ancient abbreviation
for icoi with majuscule 1; it is not difficult for us to believe that even
Ambr. >^'s MiXtw ^ Idfiwv may be a mistake for an earlier MoXoik koI
2cfM0r. On palaeographical grounds then we might accept as the
original reading something like this: MoXoiv iccu Stfunv* *Aparo9
db^cpoon^ ('Molon and Simon: Aratus was their rival in love'))
which involves the least possible departure from the manuscript
tradition ; or the vulgate reading MoXoiv kcu Scfuov * ^Apdrov drrtfiaarai
(' Molon and Simon : Aratus's rivals in love ').

It is well known that in the Scholia Vetustiora of Theocritus lurk
several pieces of extremely explicit information upon matters in Cos,
which may safely be ascribed to an early commentator on the poet^
himself a resident or native of the island, apparently recording and
reporting stories and traditions locally current. This was Nicanor
the Coan : he is certainly the authority for several items in the long
Scholium on Theoc. Id. VII. 6, where he is cited by name (Niicavoip
6 K^ vrofimffftariitw), probably also for much in Scho/, Idd, I. 57^

m margiiial explanatory giofs in a test in which /ioXc&f (participle) was read or
nadentood : see below, p. 197, note 2. The Scholiast of Ambr. i, endeavoring to
stand on two stools and to reconcile the older and better text-tradition of M6Xwr
(proper name) with the suggested Zlfutw, connects the two names in his remark.
OB the verse. Bnt I do not believe we are forced to such a conclusion.

^ On Arattts*s Arycuu, iriypd/jifiArOf and rai7Fta, see now Maass, Aratea^
pp. 250 ft (WOamowitz-Kiessling, PkiL Uni. XII., 1892). In the epigrams
FhikNrles was celebrated: Antk, Pai. XII. 129.

' For some remarks on the very complex soarces of Callierges's Scholia, see
Ahrens, Bucolic<>rum Craecorum . . . Reliquiae, vol. II. pp. hd, Ixii. — I regret
that it is impossible for me to identify the manuscript sources at the place under

194 John Henry Wright

V. 123, VII. I, 5, 10, 21, 45, XVII. 68, 69, Syr. 12 ; and doubtless
to him also we owe some of our information as to Theocritus's family
connexions at Cos.

Now it seems to me highly probable that among the minor chro-
niques scandaUuses of the prominent men of the little island was a
piquant story to the effect that the great Aratus/ and two other per-
sons known as Molon and Simon were rivals in certain love-affairs in
which one Philinus figured; and that this story, gaining doubtless
greater currency from the fact that the liaison may have been cele-
brated in part byAratus in one of his minor poems, was recorded by
Nicanor in his commentary, and lies at the bottom of the Scholium
on Id, VII. 125. It is a matter of indifference to the argument
whether the names Molon, Simon, and Philinus were the actual'
names of the persons concerned or were partially fictitious, though
the former seems to me more probable. At all events it was under
the names of Molon and Simon that the story was current, and was
reported by Nicanor. Molon, from the fact of his mention in such
good company ' as that of Id, VII., which appears to have included.

^ Maass, AraUa, c. viii (de Coo poetarum ^bdalico), discusses the question of
Aratus*s sojourn in Cos, and his friendships in the island, where he passed several
years in his youth. The Phaenomena were there composed, and were read and
recited to the literary coterie, mainly pupils of Philetas, among whom Aratus was
a leading figure. — Were Herondas, and, after an interal, Artemidonis, the editor
of Theocritus, later members of the same fraternity^ X

^ From the fact that so many of the persons mentioned by Theocritus in Id,
VII. appear under fictitious names (see the next note), and commonly in forms
shorter than those of their actual names, Maass suggests that Molon is a pseudonym
for an otherwise unknown Anchimolus (M^Xwr A7xoiro : 125). He and Knaack
associate Philinus with Philocles, ibid. pp. 230 f., 322 f. But the identification
of Philinus and Philocles is by no means certain : Philinus may well have been
the actual name of a real person; and certainly Aratus's own name appears in
this idyl in an undisgiiised form, as does also that of Philetas. The presence of
the name Molon in Coan legend is an argument for the name Molon rather
than Anchimolus : Dibbelt, Quaestiorus Coae mythologae, Greifswald, 1891, cited
by Maass.

' Philetas (v. 40); Aratus (v. 98, 122); Theocritus (2i/Mx'9af, w. 21, 50, 96;
cf. Syrinx 12) ; Dosiades (Avic/dat, w. 12, 27, 55, 91 ; unless Lycidas be O. Rib-
beck's Astacides; he cannot have been Gercke's Callimachus) ; Alexander
Aetolus (Tlrvpos, i.e. Sdrvpot, the name of Alexander's father, 72) ; Asclepiades
(2tireX/dat, 40). With 'Apiffrit (v. 99) Maass (/.r. p. 320) would identify Aris-
totherus the astronomer; Bergk makes of Aristis the astronomer Aristarchus of

Herondaea. 195

besides Theocritus, Philetas, and Aratus, the names of Dosiades,
Alexander Aetolus, Asclepiades, and possibly Hegesianax, Alexus,
and Aristothems, was doubtless a person of some distinction. And
the same might have been true of Simon. Unless he was a Coan
citizen, perhaps we have in this name a vague reminiscence of
another hitherto unsuspected member of the Coan fraternity of
poets, viz. Smias* of Rhodes, the author of the Akte^ Ovum

Samot. Haberiin {Carmina figurata Graeca, pp. 53, 54) finds Horxnesianax
referred to in *A7ffdNi| (w. 52, 61); Alexus (Athen. xiv. 620 E; this naihe may
be the doable for Alexander Aetolus; cf. Crusius, JahrbKf. Philol. 143, p. 3S7)
in *A4dwra% or 'A/a^ftixos (w. 2, 132) ; and a possible Pericles, brother of Theoc-
ritus, in Evirpirof (w. I, 131).

Probably H&berlin is not right in identifying <f>tX?M>t (w. 105, 121) with the
nmner of th^ same name, friend of Daphnis, in Theoc. Id, II. 115. The latter,
as Wilamowitz has suggested, is certainly the famous Coan sprinter who won
the prize in the ^lavKot at Olympia in at least two successive Olympiads (B.C. 264,
260: Enseb. Chron, I., SchSne, vol. I. pp. 208, 209; cf. also Paus. VI. 17. 2, who
makes him winner at five Olympic contests — boys' race, B.C. 268? H. Forster,
Dii Suger in den Olympischtn Spielen, nos. 440-445). If there is at w. 98 ff.
a reference to an actual love-affair of Aratus's youth, — and this seems highly
probable, since with all its anachronisms Id, VII. gains its main charm from its
reminiscent character, — this Philinus, in the prime of his youthful powers in
260 B.C, could hardly have been old enough, if actually then bom, to have been
the object of Aratus's affections as early as circa B.C. 292-288, when Aratus appears
to have sojourned in Cos as a young man. Perhaps, however, unless the name
be wholly fictitious or a substitute for that of Philocles or of some other person,
— it is the type of the youthful lover in Eupolis (^PoL Fr, 206, p. 314 Kock; so
Crusius), — Aratus's Philinus may have been, as H&berlin suggests, the one named
by Strato {C.A, III. p. 362 Kock), or the glossographer of Athen. xvi. 681, 682
(pupil of Philetas?). But the extreme frequency of the name ^f^tXcwt, espe-
cially in Coan inscriptions, should make us pause before insisting upon an iden-
tification. The name, referring to different persons, occurs in the following
inscriptions, not later than the third century B.C. : Paton-Hicks, Inscriptions of
Cos, nos. 10 ^ 48; 10 r 36, 70, 75, 83, and 45 a 9.

It is an interesting coincidence that on the same set of stones, to be dated
not far from B.c 260, we find the names of Nannacus, Aratus (of course not the
poet, who had long since left Cos), Philinus, and Simus (see the next note),
referring each to more than one person. One of the older inscriptions (Paton*
Hicks, no. 149) is that of a family Simonidae (Ai^t *\kwIov Zifuapid&w).

^ Of the date and literary affiliations of Simias we know little. He preceded
the tragic poet Philicus (Hephaest EncA, p. 58, Gaisf.: in Athen. v. 198 B.C.
his name appears as Philiscus); wrote in his carmina figurata a kind of
poem, on which Dosiades and Theocritus tried their hands, and like Asclepiades

196 John Henry Wright

and Securis, companion-pieces of Dosiades*s Ara and Theocritus's
Syrinx,

Have we not in Herond. III. 25, 26 another covert reference, if
not to this particular story, at least to the two citizens or residents of
Cos named in it ? The Coan affinities and connexions of Herondas
are everywhere evident in the mimes.^ And in this same third mime
we have at least two passages where we may safely see local allusions.'
At III. 10, in rjv Nawaicou icAaucra), there is probably a hit at a Coan
worthy, if at the same time a personal application of a proverbial
expression. The extremely rare proper name Nannacus is found on
a Coan inscription of the same period as Herondas. And in ras
iphofjuas T ofiavov wdj^a^ r oJSc | ra>v afrTpoSi^itav (III. 53, 54),
with its novel dorpo&^cv?, it is extremely likely that there is an
allusion to the Coan school of astronomers, established by Aris-
totherus, if not earlier, and represented at the time of Herondas
apparendy by Dositheus.* In the light of these parallels it does not
seem to me too violent to assume that in the Molon and Simon of
III. 25, 26 — which I suggest for the Mdptav and IXfuav of the papy-
rus — we have a third local touch, which would be highly appreciated
by Herondas's Coan readers. At the same time we must not forget
that the word iifiuw might carry with it, at this place, several second-
ary suggestions, since it is not only the name of many very respec-
table people in antiquity, but also has some other connotations at
once ludicrous and otherwise objectionable.^ Names from the circle

gave his name to a metre. His date and birthplace, his poetic tastes and bis
activity as Homeric glossographer make it probable that he was, like Theocritus,
a pupil of Philetas at Cos, circa 300-290 B.C. Cf. Susemihl, GcscA, d, Griech,
Literatur in der Alexandrinerteit, I. pp. 179-182; H. p. 660.

The name St/xXat might well be disguised in XlfAww, or the two could easily
interchange: compare Uavvapldt, Havfflatf Havataw referring to the same person;
Zi/u>f = Zl/iww, Strabo xiv. 648. Cf. Crusius, Jahrbb, 143, pp. 385 if.

^ Cf. Crusius, Untersuchungen %u den Mimiamhen des Herondas^ pp. 186 fl,
^ 34t 5^ ^ ii3> 125, and the index to the same scholar's text-edition, where
words found both in Herondas and in the inscriptions and other Coan records are
designated by an asterisk.

^ The fact that the ifiUdfiti and tUdt are spoken of as holidays both in this
mime (53; cf. V. 80) and in Coan inscriptions (Paton-Hicks, idid. nos. 369. 3,
402*. 6, etc.) cannot be pressed, since these days were also elsewhere holidays.
Cf. Crusius, Uniersuchungen^ pp. 68, 113.

' Maass, Araiea^ p. 321, note 56.

^ Crusius, Un£ersuchungen, p. 60.

Herondaea, \<yj

of the doctus poeta Aratus, itself the school of the poet-y/oafi/uaTiKos
Philetas, might very well be chosen by the fond father in his attempt
to examine his son on the rudiments of letters, the first step in litera-
ture {ypafjifULrZoirroi rov irarpos avr<p). Possibly also in the ^iXaivwv
of Herond. I. 5, daughter of the go-between Gyllis/ we may see
the double of the frail youth who had stirred the emotions of Aratus
and his friends.

If, now, Molon (or Maron) and Simon belong together in the
Coan story, it is clear that if the MoXoiv ^ of Theocritus is correct,
die MofMuv of Herondas must be wrong; or, vu^ versa, that the
Mo\a>v of Theocritus must be a corruption of MapoDv. In my opinion
MoXofK is too strongly fortified to be dislodged from Theocritus and
his commentator. In its favor are the tradition of the best manu-
scripts, and, apparently, the text at the bottom of the Scholia Vetus-
tiora. It is perhaps also sustained by Eustathius, who is full of
Theocritean reminiscences, in the words MdXoives o& irapa r<^ KtofUK^,
o T€ yptoi [read ipStv] koI 6 aKunrrofitvoi (p. 882. 24). Now a hero
Molon is nowhere mentioned in Greek literature, so far as I know,
unless he lies behind the word Molon which is found in Coan
mythology. I suggest that rjpioq is here a corruption for ipdv (' the
lover'), and that in appending this epithet Eustathius had in mind,
though vaguely, the Molon of Theoc. Id, VII. 125. The MoXa>v
6 aKunrrofuvoi is the one mentioned in Aristoph. jRan, 55. Eusta-
thius might very well have here connected both the Molons with
the poet of comedy, through a slightly confused recollection of a
sentence in the Didymean commentary on Aristophanes, of which we

^ The original form of the name here is ^iXalptow. The marginal variant
^iKauwUht probably suggested itself to a late corrector of the papyrus because of
the notorious betaera of this name (AnlA. Pal, V. 202: cf. Crusius, Unteriuch-
^^fV^i PP* 43« 129). Perhaps, however, there is in this daughter of the athlete
Gryllus's friend, a covert reference to the great athlete and runner Philinus named
above, whose career resembles that of Gryllus.

* The reading fioXcir, participle, adopted by Ahrens and others from inferior
manuscripts, and from a varia lectio of the Scholiast, is hardly probable. As the
lecHc facilior it probably arose from a misunderstanding of the proper name
M6X«#r, wen attested by Ambr. k — text and Scholia, — by the first hand of Medic.
/, and by the Juntine, which is based in part upon a manuscript of the same
family as Ambr. i, as good as i, if not better. This confusion was not a little
helped by the /loXoii^'a | r^pmi^rw vorl riiP Tiftayil^oto vakaiarpaw of Id, II. 96, 97.

198 John Henry Wright,

have traces in the Scholiast on Aristophanes and in Suidas.^ In this
commentary Didymus had said that there were two Molons in an-
tiquity, respectively actor and thief, and that Aristophanes here {^Ran.
55) means the thief, since he was small of stature. Now in the pas-
sage cited above from Eustathius we are also told that there were two
Molons, and that both were celebrated by the comic poet ; whereas
in fact only one Molon is mentioned by the poet^ while it is the com-
mentator that discourses of two Molons. This duality of Molons in
Greek comedy according to Eustathius, arises from a misrecoUection,
on his part, of the Didymean commentary, since elsewhere he refers
apparently to only one Molon as mentioned by a comic poet.' All
these facts with others show, first, that Eustathius read his Aristoph-
anes, his Theocritus, and his Didymus, and, secondly, that at least in
two cases — where by a fialse association of ideas he gives to Aris-
tophanes what Didymus had said, and where he turns a thief into a
lover (or hero) — his recollection of his reading was of such a nature
as to make it quite probable that the Theocritean Molon came into
his mind and was duly noted as he endeavored to recall and record
a bit of dimly remembered Didymean lore.

Retaining, then, the Molon of Theocritus, the question arises
whether the Mapaw of the Herondas papyrus can be traced to an
original MdXwv as written by the mimographer. There is no uncer-
tainty about the reading of the pap3rrus: MAPUJN is unmistak-
able in both places where the word occurs. If an error was made
by this or an earlier scribe, it must have come about in one of two
ways, either through a misreading of the letters of the original text,
or from some probably unconscious mental confusion, on the part of
the copyist. The manuscript from which the papyrus was copied,
though in the main quite legible, was at places obscurely written,
and abounded in orthographical errors, among which misread letters
figure largely, all of which may be seen from the corrections made

^ Schol, Aristoph. Jian. 55 : AlivjiSt 4^riinw tfri 9io MhXufpH tUriPf h inronpir^t
Kal h \tinroli&rn% * koX /AaXXoF rhw XuroS&rriP X/7ei, 5t iari fxucpbf r6 ffQfta. Said. s.
M6\tap : M6Xayi«t dvo, {nroKpiral koI XanroSvrat.

^ Eustath. p. 1852. 1 1 : rapd t6 fioXctr H & MovXios 'IwfucJ ixty$4ffti rod v • icai'd
Kal 6 ToO KwfwcoD MXwp Kal ol fjLo\low€i, — Eustathius*s remark that Molons were
large persons is probably to be traced to some other source, if not one of his own
etymologies (MoXaycs ol voXv^uyiBeis dirb roto&rov MoXctfrot, p. 1 834. 32).

Herondaea, 199

by the first hand in his revised copy ; this has been pointed out on
pp. 182 ff. Now the letters OA in the writing of circ, b.c 100-
iUD. 100, or even earlier, might well have been dashed off by a scribe
so as to be taken by a copyist for AP : interesting examples of these
letters blindly written occur in our papyrus itself at IV. 29 (MHAON),
and II. 78 (eAPCeUJN).

But we are not reduced to the necessity of explaining the probable
corruption on palaeographical grounds alone. As we have already
seen, the scribe of this manuscript did not slavishly copy his original,
letter by letter, but appears often to have carried the words in his
mind, dictating them as it were to himself, and writing sometimes
not the word he saw, but the word he thought he heard. Now in
such a process it is quite possible that, in the case of an unusual
proper name, the cognate sounds of the liquids X and p might have
become interchanged,^ — as in the classical example of Alcibiades's
pronunciation of Ocwpo? and Kopai as OccuXos and Kokai — and that
while our scribe saw MoXcuv he wrote Map<i>v. The mistake may
have been made the easier by an association of ideas with Virgil.
The writer of the papyrus manuscript, "who may be provisionally
assigned to the second or third century a.d." (Kenyon), when Virgil
had already become a text-book in the schools and was well known
in the ancient world, might well have associated the supposed Maro
of the original mime, whose name is there spelled out to a lazy school-
boy, with the famous Roman.' It should finally be remarked that the
Map<av of the Coan inscriptions, to which reference has been made
in illustration of the name in Herondas, cannot be taken into consid-
eration in this connexion. Unlike the Nannacus, Simus, Philinus,
and Aratus mentioned as foimd on stones of the third century
B.C., this word occurs only in a late Christian inscription ; ' perhaps

^ For Alcibiades's mispronunciation see Aristoph. Vesp. 44, 45; Plut. AU. i.
Cf. *kitaprf6t . . . Xfyeroi Kal 'A^X7or, Stephan. Byz. s,v. In one of the
modem Cretan dialects dXXo is arro»

'^To a scribe writing in Egypt after B.C. 50, the name of the Alexandrian
Marion, the Olympic ropado^rfnyr of B.C. 52, who won the prize for the pancra-
tium and the wrestling match on the same day, and thus became the fifth Hera-
clean double-victor, would also have its associations. Forster, Die Sieger^ nos.

579. 580.

' "Mdpwwf. h'^Qv) K, Small stele, with aedicula in the centre of which is

a cross within a circle " : Paton-Hicks, Inscriptions of Cos^ no. 339, p. 219.

200 John Hemy Wright

the young man on whose gravestone it stands received his name,
which is not a fieqoent ooe among die GredLS, in honor of the
author of the Aeneid.

In view, then, of all these considerations, I do not hesitate to pro-
pose as, at least, a probable, if not a certain, reading at Herond. II.

34-26 : —

rpiB^|Up^ MoAwra ypof^iart^ornx

rov varpoc avry ror MoXwra jiou f acy

oSroc ^^fHi^^r^. 6 XP'V^^'

NOTES.

Eur. Alcest. 229, 230.

KaX ttXcov y Pp^'K^ hiprqv ovpavltf irtKojiTata. The word ovpavii^ is

suspicious, as the expression "sky-hung halter" is too extravagant
for Euripides, though Aeschylus might perhaps have used it. Bacch.

1064, iXanis ovpaytov cucpov KkaSov, El. 1 1 58, ovpdvta Tti\€a, Tro. 1 08 7,

rcixca ovpaa^uLy are not parallel cases. To call a />r^ or raaU ** high as

heaven " is a common poetic hyperbole in every age. An instance

more to the point is El. 860, ovpdviw m^/uui, but this is surely a far

more natural expression than oifpdvtoi Ppox^ would be. Wecklein

has suggested the reading <iyxoKi<p instead of wpavitf ; but the two

words have little resemblance to each other, and it is hard to see

how the change could have arisen. Possibly we should read ovXx>fUi^

" fatal," " deadly," instead of ovpavuf. The Epic form ovkofuyoq is

found in three passages of Euripides — Iph. Taur. 1109, irvpycov

ovXofUvuw (so the Cod. Florentinus; the Palatine has 6ko/jL€vuw,

which is metrically impossible here), Iph. Aul. 793, rrarptSog ovXo-

ficyas (so the Mss.), and Phoen. 1526, wXofuv alKCa/mra, About

the two first of these there is some dispute, and many editors read

6XXvfUyiav and oXXv/iiya^ with Erfurdt, believing that ovXofiei^ cannot

mean "lost," "ruined"; but cf. Aesch. Prom. 397, ovXoficms Ty^as.

About the third, in which ovkofum has the desired meaning of "fatal,"

"baneful," there is no question. Sophocles has one v^ry doubtful

case, Antig. 840, ov#c ovkofuifav vj^/M^^as. Here the Mss. have oAo-

fUmvy but the metre requires a long initial syllable. If ovXofUmv

is right, it must mean "dead," as it is opposed to ItiiIkivtov ; but

Martin's oixpfuyay is not a violent change, and brings out the antithesis

better. To sum up, Euripides uses the word once in the required

sense, and has two other cases where the form probably occurs,

though with a different meaning. Aeschylus has the form once,

201

202 Notes.

Sophocles probably not at alL When the inflnmrr of Homeric
passages like Od. lo^ 394, ^dpfuuaw owXofuwoi^, is taken into account,
it seems not mireasonaMe to suppose that Euripides may have
written Pp6x;u ou X o fio^

PeTRONIUS, C. 65, BUECUELER.

Ego mates tate conterritus praetorem puiabam venisse; ibid, prae-
iorio loco s€ posuit From the use of the word prtutorem in the
former of these two passages Mommsen (Hermes, XIII. p. 109)
inferred that the scene of the Cena Trimakhionis was laid in
Cumae, because it is certain from inscriptions (C.I.L. X. 3685,
3693) that in that city the chief magistrates were styled prae tores.
Friedlaender in his admirable edition of the Cena (p. 6; cf.
" Wochenschrift f. klass. Philologie," Nov. 25, 1891) adopts the same
view. If it were certain that Petronius in these two passages used
the words praetor and praetorim in their strict sense, the argument
would be conclusive. But there are several facts which make this
extremely doubtful. The word praetor was one in the use of which
there was great latitude. Originally denoting any leader or chief^
it was long appUed throughout a large part of Italy to the chief
magistrates of towns. How widely this usage prevailed may be seen
ftx)m the instances collected by Marquardt (" Staatsverwaltung," L*
pp. 149, 150). It continued among the common people even after
their towns became colonies or municipia. There is evidence that in
at least three Campanian cities— Cales (C.I.L. 4651, 4657, 3923),
Capua (Cic. de Leg. Agr. II. 34, 92) and Cumae — the chief magis-
trates were called praetors ; and what proof have we that this was
not the case in others as well, e,g. Puteoli? The inscriptions, it is
true, do not show this; but they do not in the case of Capua,
although they are very numerous.

Again, the title " praetor " is one which a stranger like Encolpius
might very naturally apply to a city official of whose precise rank he
was ignorant, but who was attended by a lictor and a large retinue.
We have as litde reason to suppose that the chief magistrates of
Triraalchio's city were really called praetors as that the same title
belonged to those of Saguntum (Liv. XXI. 12, 7) or of Fundi (Hor.
Sat. I. 5, 34). In the words of Teuffel (Rom. Lit ed. Schwabe,

Notes, 203

p. 745), "auch der praetor c. 65 kann nichts beweisen." In the
second passage from Petronius quoted above, Friedlaender has tried
to use the words praetorio loco in support of his position (see
" Wochenschrift," /.r.). But praetorius locus is simply another name
for the locus consularis — the place of honor at the table — and the
expression has undoubtedly survived from the time when the chief
magistrates at Rome as well as elsewhere were called '' praetores."
Cf. Sen. Cont. IX. 25, 2, meretrix uxoris loco accuduit, immo prae-
torts.

On the other hand, the difficulties of the view that Cumae is the
scene of the Cena are very great. Trimalchio says (c. 48), nam
Sibyllam quidem Cutnis ego oculis meis vidi in ampulla pendere.
Strange language, surely, if the speaker is in Cumae ! Mommsen
thought that Petronius wished to hold up Trimalchio to greater
ridicule by making him relate events supposed to have occurred in
his own city as if they had been seen on a distant journey. This
view has found few adherents. On the other hand, Friedlaender
frankly admits that if Cumis in c. 48 is right, the scene of the Cena
cannot be laid in Cumae. Following Studnitzka, he holds Cumis
to be a gloss which has crept into the text (see his edition, note
ad loCy and " Wochenschrift," /.^.). It is, of course, possible that
this is so, but to assume that it is true without further proof is to
suit the fiacts to a preconceived theory. Moreover, even if Cumis
is a gloss, there is another passage that occasions difficulty, for in
c. 53 the praedium Cumanum, quod est Trimalchionis is mentioned
along with horti Pompeiani; a fact which Friedlaender vainly strives
to explain away in his note ad loc. Either one of these passages
might perhaps not be decisive, but to disregard both is surely to go
too far, and is contrary to the principles of sound criticism.

Soph. Track. 56, 57.

/AoXurra S* ^vrcp dnxn^ *YAXov, ci rrarp^
v^bUM rtv* cSpaF rtni KoXiai vpdnra'uv Sok€iv ;

So the Laurentian. Vat. and Harl. have ve/ici, which some editors
prefer, following Matthiae. The principal difficulty is with toO koXus
irpduro-av Sokuv. Those editors who retain these words unchanged

204 Notes.

regard them as an epexegedcal addition, the whole being equivalent
to ci ytyjOL rt¥ tapay rdv ror var^pa KoXm^ wpd/nruw Stmuv. To this
there is the fatal objection that Soiccur is unsuitable. Hyllos would
be concerned about the rta/ welfare of his &ther, not his apparent
prosperity. The lines have been emended in many ways. The
easiest of these changes is perhaps that suggested by Nauck, ci
varpo9 I vifLor vty «jpaF rov KoXm^ wpa/nruw Soku^. KoXms wpdatmv
usually means ^ to prosper/' " be fortunate/' but may also mean
"to act rightly/' or "fittingly." Cf. Soph. O. C. 1764, nu Tuvm
fL 1^ vpaatroyra xaXJuq x^f"^ cfov aJufv oXinrory where Professor Jebb
acutely observes : " koXS^ with wpaxnromu (not with Ifcir), ' in a
seemly manner/ 'duly' (Lat. nVSf). The fact that wfwnrorra mX^
usually meant ' hiing well ' is no objection. The ancient Greek
instinct for words was remarkably free from bondage to phrases."
C£ also Plat Gorg. 507, C ; Charm. 172, A. If now we read

d wdpo^
V€fui Tiv wpay rov icaAak wpaaxruv Sokuv,

and take icaXm wpaavuv in the sense of " to act rightly/' the diffi-
culty seems to be in great part removed. The idiomatic use of
wapoi (like the Ger. sons/) with a present is well known. The most
familiar case is doubdess the Homeric wdpoq ye pkr am Ai/u^as,
II. 18, 386. The sense wiU then be "if on other occasions he has
(habitually) shown care for his reputation (for being thought to act
righdy)." The change is certainly a very slight one.

H. W. Hayley.

On Horace, Sat. I. 4. 39.

Primum ego me illorum, dederim quibus esse poetas,
Excerpam numero.

Bendey (and before him N. Heinsius), against the almost unani-
mous testimony of the manuscripts, changed to poetis, citing 5. I.
I. 19 atqui licet esse beatis; S. I. 2. 51 munifico esse licet; Ep, II.
3. 372 mediocribus esse poetis non homines, non di, non concessere
columnae; and the passage which he found already quoted as a

p*^*^^*^ b^* AtTffx. 3L L. 6.. 24 <{EiD ccCi. Itutk, .yiiiwic ijsposbcm cihnnm

fteiCT C jgi*i^ D g uil e i aas beea feflb««<i bece br O^dliL Kjnpc

YakioL ScaiScir Kirsrims jad ocbos. DuIieBfiiicser. Kiti^ier. KcOier.

Widduiny jsd Heves ^in die bst e&im o€ CMSl) netmi ^tw^isf^

bat ao oat a£ t&en poimcs osc die &nr aa Bcnt&nr s aEjpnacBC. oc

gcrcs JUT reasoa fer tta pufcHHir onepc cfic v«^s&fi ct onDttKztpt

ainhmui .

I B2ve B3 d3i=bc mac Beacfer was ligbs ia bofc&ag tiixt Hence

aivaj^ osed a pcedkace diche, and nerer a predkate accasafive^

after Mcef esse ami cqarrakac exfrtssims ff prmmissiirm^ In iS^^ L

16. 61 da XBO^ £&Ifere, da iaurip itfi«L'M]ae Tideru ^le amnsschpts are

prettj ctrekf dnrided b cfccn imstt s*Mm:$Kpft and ia^stMim Mtmki^tm-

qpc ; but the femier b nghtlr prtcued in afl die cditioiiL Tlie octhr

odier f uny l f ^ diat occur in die poems are diose cilcd br Bender.

in which tibe text mar be leg^arded as ceitauL Bender*^ aigiiuneiil

£ulsy becanse the case here is not paiald to those vhkh he cites^

bat imoli e s a difi e ient principle. ZX^ in our passage does not

opress p c i i HtniM iOy as e^uced^ does in E^, II. 3. 371* whkh b

Bender's nearest panDd; it c a pi esscs aJmussUm of an assertion.

the giamiug of a daim ; and the construction it introduces s diat

of indirect discuunc . It is not the d^ of E^, I. 16. 6i. or of «!& IL

3. 191 di tibi dent da s Km rednccre, bat of E^. IL 1. 125 si das

hoc, panris qooqoe rebos nu^na iovarL 'Esse poetas* stands fo

the ' saoras poetae ' of the writeis who claimed dat distinction for

themsehrcs. This use of i^ s common enough in aigumentadre

discooTK; c£ Cic Tmsc. L 25 M, Qoid hoc? dasne aut manere

animos post mortem ant morte ipsa interire? A. Ek) vera M. Quid

si maneant? A. Beatos esse coocedo. For the dative yw^«^« d

Cic Jftv. I. 53 Socrates . . . nihil ipse adfene ad persuadendum

volebat, sed ex eo, qaod sibi iDe dederai qaicam disputabat, ahquid

cooficere malebat qnod iDe ex eo qaod iam coocessisset necessario

adprobare deberet.

C. L. S>irrH«

206 Notes.

In Liddell & Scott's lexkon, under the word -rijyrviMA, we find : —
amp^ w., to fix^ pitch a tent^ Andoc 33, 9, Plat. Legg. 817 c (so
in Med. crjnjr&s siifio^ai, to pitch their tents ^ Hdt. 6, 12).

In the passage from the pseudo-Andoddes as well as in that from
Herodotus, the expression is a purely military one ; not so in the
Laws. Here we have to do with the answer to be given to tragic
poets who may request to be aUowed to produce plays, and part of
this answer is : —

fii) 07 oo^Tp't "^fia^ pfykmi ye ovrm vfiai vorc vap tf/ui^ iatruw aicrpm^
T€ in^(avT€ui tear ayopay «u icaAAi^wrovs vroxpiTas utnyayofierovi ktK,

I have never seen this passage referred to in the discussion of
theatrical antiquities, yet it is obviously of interest, and it may be
of importance, in the debate between the Old-Stagers and the No-
Stagers. It is true that cnci/mc w^$ai here may simply mean that
actors on coming to a town ' camped out ' in the ayopa and lived
there during their stay in the town. Yet at the time when the Laws
was written, inns were not so rare that travellers were reduced to
this necessity; and further the theatrical word dcmyoyofimns so
closely following seems to belong or to point to aicrpms. But it is
possible to look at the phrase in two other ways. It may be a sur-
vival from the time when as yet there were no permanent stage-
buildings, and when axriyrj meant merely the hut or booth used by
the actors as their dressing-room. (I use the word 'survival' here
because the ideal city of the Laws was already provided with Bwrpa
(p. 779 d), and these, at the time when the Laws was written, must
have included stage-buildings, whether the stage was raised or not.
In a note on aKtpfhs irrjiajL in my article on aicfp^wAy a-KiivWf and
aKtp^ in the American Jotimal of Philology, XIII, p. 79, 1 did not
observe that trtcrfvaa taken in this sense must be a survival.) From
this hut would be made the entrances of the actors, into it their
exits, and on its front would be hung the scenery. The hut itself
would be set on the edge of the circular orchestra, which might
naturally be in or near the lyopL The old ' market-orchestra ' of
Athens is an instance in point. A third view may be to take o-ioTvas
in the sense of ^wagons,' — the wagons in which the travelling

Notes, 207

troupes of actors carried round their scenery, costumes, property, and
the like. These would be covered, to protect the goods from storms ;
they would be atctpnl TpoxqXaroi, a phrase used by Aeschylus, /Vrj.
1000 ; cf. also Ar. Ach. 69, ifncrpmnLhtoL^ said of the envo)rs travelling
in the covered carriages of the Persians ; and tncrprfi as used of the
tilt of such a wagon in Xen. Cyr, 6, 4, 11.

Whatever be the meaning of the phrase, the whole passage seems
to be our earliest mention of travelling troupes of actors.

M. H. Morgan.

MeXo? *song.'

Curtius in his Greek Etymology connects this word with fietXca and
fieiXtx^o9 ; it would then mean originally something like ' softness.'
Similarly Vanicek. This is a conspicuous example of how things are
not named. There cannot be any doubt that fic\os ' song ' is one
and the same word with fic\o9 ' limb.' The meaning ' song ' is post-
Homeric, appearing first in Archilochus, Alcman, and one of the
smaller * Hpmeric ' Hymns. The transition from ' limb ' to ' song '
is illustrated by the Sanskrit word pada. Pada, properly * foot,*
means a quarter of a slaughtered animal. Then it means a line of a
four- verse stanza. Thence it comes to mean * verse ' outright, even
of some different stanza.

So in Greek the rhythmical divisions, or phrases, of a song were
once called its fic\iy, or ' limbs,' precisely as long afterwards, by the
same figure, they were called its iccoAxu The strophe of four phrases
always predominated in the simpler sorts of lyric poetry. Such a
strophe, for instance, was the elegiac distich, at the time when ele-
giacs were sung. It needs no great effort of the imagination to
conceive it as a quadruped.

ucypif Tcv fcaroiccior^c : ) . ...
'• ifx "t £L ' Morehmbs,

K^T OAKlfJLOV C^CrC UVfJLW )

. . , > hmd limbs.

And I need only mention the stanzas of Alcaeus and Sappho, the
form of the Attic scolion, and such early lyric scraps as Archil, frag.

2o8 Notes.

Z'^ and 94, to make clear the prevalence of the tetracolic strophe.
The termmology in question grew up, I think, in the singing-school ;
boys were taught their songs phrase by phrase ; they were made to
sing and play them limb-meal {nxkifiav) ; they called their singing
lesson their yJXri, as they did their Homer lesson Imi. The antith-
esis of Imi and lilKri in such places as Plat Rep. II, 379*, X, 607*,
rests on ancient tradition.

How then did /tcXo? come to signify a whoU song? At first by a
sort of " synecdoche," just as we use * stave * or * strain ' for a whole
song. Indeed, these very words will fit perfecdy as translations of
iUXjk in nearly all the earlier occurrences. As when Archilochus
says, fcoAov iidp(ai fi€XK o28a, SiBvpofifiov (frag. 77), or Theognis
(761), 'Let lyre and pipe soimd a 'ufioy fitXK, or Alcman (frag, i)
calls on the muse to begin a fiiXxK vtoxfioy. So a score of other places.
Often the plural, * strains of music,' is used : Horn. Hymn. XIX, 16
(ovK aK rovyt irapaZpafioi iv fuXita-anv), Aesch. Suppl. 809, Pind. O.
II, 47, etc. ; sometimes when a single composition is evidendy meant
(Pind. O. X, 84; I. V, 2). Nevertheless there are places in the
earlier poets where the later sense of ficXos begins to appear : Alcman
frag. 25, Imi roSc koI fic\o9, ' these verses and this tune ' ; Echembrotus
in Pausan. X, 7, 6, ficXca koI cAeyou?. And the sense ' lyric compo-
sition' is clear in Herod. II, 135 ; V, 95. The idea of music is
always present, whether melody as opposed to words be intended
(Alcman frag. 25, quoted), or instrumental music only (Theogn. 761,
quoted ; Pind. P. XII, 19 ; Alcman frag. 82 ; Sophocl. frag. 226 D ;
Simonides frag. 46), or, finally, song as opposed to instrumental
accompaniment (Pind. N. IV, 15 ; O. X, 84).

Significant is the adjective i/ifitXT^, established in the metaphor-
ical sense 'suitable,' 'harmonious' in Simonides's time (^fificXcci)^,
frag. 5), and recurring in Aristophanes and Plato. It is based on a
phrase iv fieXa, which we find in Plato, Soph. 227** (iv fjJXti iftOeyio-
fuOa). This phrase also is of the singing-school. To Plato it doubt-
less meant * in tune.' But originally, we may well believe, to sing ^v
fiiXu was to sing * in time,' following the rhythmical divisions. There
are indications that ^fi/xcXiTs once referred to rhythm. First Sappho's
ifjLfieXiioq iroScaratv <op\(vvTo (frag. 54). Then the dance cfifieXcia.
This name was not confined to the tragic dance, but was used by
Aeschylus (see Hesychius s.v.) of the a-uawis of the satyric drama.

Notes. 209

In Herod. VI, 1 29 ififjiAaa means simply * dance-tune/ Lastly, on
Tiva Xdy<i>v i/jLiJLfXfULv, Aristoph. Ran. 896, the Scholiast remarks,
KaraxfyijoTiKSts vvv rrp^ evpvOfuav ', perhaps wrongly, for ^ftfteXcca may
here signify only ' harmonious combination.'

The opposite of ifjLfifXrj^ is vXrjfifieXrjs (first, I think, Eurip. Med*
306), which points to a phrase ir\rjv fie\ov9, * out of time * (and tune).
For this we have nap fie\o9. Find. N. VII, 69, and irap^ /liXoq several
times in Plato.

Of ficAi^o), 'sing rhythmically,' we have spoken above. It first
occurs Aesch. Ag. 11 76 (and Pind. N. XI, 18?). I suspect that Hor-
ace's carmina divides (Od. I, 15, 15) is a translation of this. Marini
and Henzen interpret carmen descindentes in the Acta Fratrum Ar-
Valium (ann. 218 ; see p. 33 of Henzen's edition) in a similar way.

The parallelism of the German glied and lied^ which has sometimes
been brought forward, is striking but illusory. Lid, * limb,' and liody
* song,' are separate words in old German, and without etymological
connexion. Nor has fie\o9 anything to do with fieA.Trctf.

7rXai^oSta9, Homeric Hymn, III. 75.

irAavoSiac 8* i^Aawc Sect ^nafia^ci>8ca ypipov

is an " acephalous " verse, which I think has passed unchallenged
hitherto, though its refutation lies in plain sight in Hesychius : ttXi/vo-
hiax . . . T^i 7r€wXavrffi€vrfi rrjs 6p$rfs oSov, rovrcoTiv dSucoM. This gloss

relates to another occurrence of the word, in a figurative sense. In
our passage irkrp^o&uK should be read, as an adjective referring to
Povq understood. We have spoken in the preceding note of irXtffifu-
Xi/s, and the phrase vkTfv fitXov^ which it presupposes. In like way
TrXi/vd&os comes from a irkrp^ o&w, ' out of the road.' The two words
together show that TrX^v in former times had a wider range of mean-
ing than ' except'

Frederic D. Allen.

GENERAL INDEX.

A, A» confused, 183.

Accent, acute, indicating a question,
170.
example of change to, from quan-
tity, 105.

Accents in Herondas, 169 f., 181.

Accentual rhythm in Latin, 105 ff.

Accompaniment, higher than melody,

45-
by left hand, 46.

Acoustics of pipes, 22.

Acute accent, in Herondas, 169 f.

Adonic, conforms to dactylic scheme,

109.

tuber ere^ adsertio^ 162.

Aeschylus, Euripides and the Rhesus,

words peculiar to, 79.

imitations of, in the Rhesus, 90.

and the Rhesus, words peculiar to,

79-
Sophocles and the Rhesus, words
peculiar to, 79.
All Gf not confused in Herondas, 186.
Alcaic, not fayorite for Christian hynms,
112.
prose rhythm of^ no.
scheme of accentual rhythm, 1 10.
Alcibiades, his mispronunciation of p,

199.
Alexander Aetolus, 194 f.
Alexis, 194 f.

Allen, Frederic D., on reiptip A^-
vBvx (2 501) and the manui
conserHo of the Romans, 151 £
on ^Aot, song, 207 ff.
on Hom. Hymn (III, 75), 209.
dEfMM wapOtwlyis, 190.

Amorgtts and Amolgus, 199.

dwMTToi', 33.

Anchimolus, for Molon, 194.

dro^ €lfnifUwa in the Rhesus, 72, 73,

76, 77-
dw«^ Tpay(pdo^fUPa in the Rhesus, 72,

74.
d^auoTOf, 189.

dpdffff€i, impersonal?, 174.

Aratus, youth, 191 E

at Cos, 194.

friendships, 193.

lesser poems, 193.

PhaenomenOf 194.
Aristarchus of Samos, 194 f.
Aristis, 194 f.

Aristotle, on finger-holes, 3.
Artemidorus, of Coan schools of poets,

194.
Asclepiadean, in hymns, 112.

general dactylic, flow of, 112.
Asdepiades, 194 f.
dtfTpodc^if, 196.
Astacides, 194 f.
darriBi, darriffow, I $6.
dSucTot, 189 ff.
SiSpavoTOff 189.

Attraction of predicate to dative in
Ovid, 119 f.

in Horace, 205.
AY, Ar, confused, 183.
auctiOf meaning of, 144.
auctor^ etymology of, 143 f.
a^Xifruc^f, KdXa/iof, 21.
A^X^, The, or TUna^ i E
adX6f, bands on, 7 ff.

curved, 35.

211

212

General Index,

aiikbf, in Didascalies of Terence, 37.
the double, 20 ff., 35 ff., 43.
experiments with, 27 ff.
existing instruments, described,

47 ff.
finger-holes, 2 ff., 19.
Greek and Roman, cylindrical bore,

22.
harmonies of, 30 ff.
kinds of, 12 ff., 39 f.
icoiMai, II f., 22.
Lydian, 43.

method of playing the, 43.
5X/Aor, 28 f.
0op/3e(a, 29 ff.
Phrygian, 42.
pitch of, 41.
Pompeian, 30.
range of, 30, 44.
reeds of, 21 ff.
side tubes in, 8 ff.
single, 12.

sizes of the, 38 ff., 41.
speaker in the, 32.
stopped, 2.
syrinx, 32.
avena, 23.

Bands on aUXAf, 7 fL

Bassoon reed, 24.

Berecyntian horn, 36.

Blots, in Herondas papyrus, 175.

Boehm flutes, 10.

/5pttx«ia, sign, 177.

Breathings, rough, in Herondas, 169 f.

Caesura of Sapphic, 106, 108 f.
Callierges, scholia on Theocritus, 193.
Callimachus, 194 f.
€apistrumf 29, 30.
carmina figurata^ in Greek, 195.
Change of speaker indicated in Heron-
das by spacing, 171.
by Tapd7pa0oi, 178 ff^
Circumflex accent, in Herondas, 169 f.

Clarinet, bulb of, 29.

mouthpiece of, 21.

speaker in, 31.
Classical writers, were they conscious

of prose metre ? 112.
Coronis, 170, 179, 181.
Cos, traces of, in Herondas, 195.

Nicanor of, 193.

school of poets at, 193 ff.
Curved pipe, 35 ff.

haKTvKiKol (adXoQ, 39.

A» A> confused, 183.

Dactylic metre, Sapphic changed to,

105.
Datiye for accusative in predicate after

infinitive in Ovid, 1 19 f.
in Horace, 205.
buLffri^i., 177.
iiavTokiif 170.
Diatonic scale in a^X6f , 2.
Didymus Chalcenterus, 197 f.
dieXjn;<rr(rda, 1 52.
dcrX^, 181 f.
Dosiades, 194 f.
Dositheus, at Cos» 196U
Double pipcs» J^ 35, 43.

G» Of confused, 186.

£, All not confused, 186.

ipd6fi,ii, holiday at Cos, 196.

ecastar, use of, by Plautus and Terence,

99 ff.
ecUpol, use of, by Plautus and Terence,

99 ff.

Egyptian pipes, i, 23, 30.

EI, corrected to I in Herondas, 186.

e/icdr, holiday at Cos, 196.

fkvftoi (adXo/), 43.

ififiaT^jpioi (atlKoi), 39.

Emendations :

Aristotle, ProM, xix (23), 3, 19.

Eur. Ale (229), 201.

Eustathius (p. 882, 24), 197.

Gellius (xx, 10, 7), 159.

General Index.

213

Emendations :

Hcrondas,il/i«i. I (i), 174; (55),
187 ff.. 192; (64), 173; (82), 179;

III (24-26), I92ff.; (80). I72ff:;

IV (35-38). »79r (62), 177; V
(7). 174; (18). 178; VII (1 18),
173; f«23). 174; VIII (3),

173 <.
Horn. Hymn (III, 75) , 209.

Schol. Thcocr. VII (125 Ambr.)i

193-
Soph. Trach. (56), 203.

^fdoTot (ai\oi), 39.

ifjt4u:\iis, meaning of, 208.

iriff^^yli^eaOai, to initiate, 191.

esse, omitted in compound fonns of

infin. in Ovid, 137.

Etymologies, Latin, 143 ff.

e9, position of, 1 74.

€v 4- 11, synezesis of, 174.

Eucritus, 194.

Euripides, Aeschylus and the Rhesus,

words peculiar to, 79.

Ale. 229 emended, 201.

final sentences in, 93.

imitation of, in the Rhesus, 85 ff.

and the Rhesus, words peculiar
to, 81.
Eustathius, reader of Aristophanes,
Theocritus, and Didymus, 197 f.
exerceo, etymology of, 147.
exercitus^ etymology of, 147.
^(lyvX/i^pot, 21.

figurata carmina, of Simias, Dosiades,

and Theocritus, 195.
Final sentences in Euripides, 93.
Finger-holes in adX6ff, 2, 5, 6 fil, 30.
in o'vpi7(, 19.

7eXf r, parenthetical, 187.
Gellius, XX, 10, 7, emended; 157.
gingrinae (Hbiat), ylyyp^i, 39.
Glosses, in Herondas, 186.
irX^ra, in a^X6ff, 21, 26, 31.

yXurroKOfutop, 26.
r P» TT» confused, 183.
Grave accent, in Herondas, 170.
Greenough, J. B., Accentual Rhythm
in Latin, 105 ff.
Latin Etymologies, 143 ff.

Harmonics of the ai)X6t, 30 ff.
Hayley, H. W., on Eur. Ale. (229),
201.

on Petron. (65), 202.
Hercle, use of, by Plautus and Terence,

99 ff.
Hermesianax, 194.
Herondaea, 169 ff.
Herondas papyrus, punctuation, 169 ff.

rapdypa^t and d/3eXot, 177 ff*

earliest corrections, 182 ff.

not written from dictation, 183.

traces of G>s in, 196.

ffippTfyls in, 187 ff

Molon, Simon, and Aratus in,
192 ff.
Homeric Hymn, III, 75, emended, 209.
Horace, his possible feeling for accent-
ual rhythm, 114.

Sat. I, 4, 39, reading, 204.

treatment of Sapphic verse, 107 ff.
Howard, A. A., The Ai)X6ff or Tibia,

Iff
inrtpriXttoi (adXo/), 39.
^^Xfu/ow, 28.
inrocriyfiij, 176.

Iambic verse in Ennius and Naevius,
how it conforms to word accent,
114.

impares tibiae, 37.

Infinitive of purpose, 93.

inUger vitae, 105.

firy^, 2a

cdXafiot, 23.

Mirs^vAr, 33.

Kipara, in bands of pipes, 8.

214

General Index.

KiBaptffr'^iptoi (a^XoOi 39> 40.
cX^f = ff^fMyli, 191.
KoOdoif main bore of pipe, 11.
Kopwwls, 179.

A, A» confused, 183.

X and p interchanged, 199.

Lycidas, in Theocr. (Id. vii), 194 f.

Lycophron, 189 f.

Lydian pipes, 43.

Macrobius, on finger-holes, 4.

fidyadis a(f\6Sf 40.

fjMKpd, the sign ?, 177 f.

Manning, R. C, Omission of the Sub-
ject-accusative of the Infinitive
in Ovid, 117 ff.

manum consererty 155 ff.

manu{m) consertum^ 162.

manus conserere, 163.

Manuscripts, papyrus of Herondas,
169 ff.
Z of Demosthenes, obeli in, 191.

Marion, Olympic victor, 199.

Maro (Vergil), 199.

Maron, 197 ff.

/i^> independent clauses with, 94.

IjAi 0^, 94.

Mecastor, use of, in Hautus and Ter-
ence, 99 ff.

Mehercle^ use of, in Plautus and Ter-
ence, 99 ff.

fuXl^ta, meaning of, 208.

Melody, by right hand, 45 f.

/U\ot, song, 207 ff.

fiiari oriyfi^, 175 ff.

IkurbKoiroi (a^X«/), 40.

Metrical structure of the Rhesus, 91.

Midas of Agrigentum, feat of, 19 f.

Mistakes, various, in Herondas papy-
rus, 183.

luo\(M)Vt in Theocritus MSS. 197.

Molon, rival and friend of Aratus, 192 f.

Molonis, at Cos, 194.

M^ravXof, 12 f.

liOvoKd'Ka.iAOi, 19.

Morgan, M. H., on Plat Legg. (81 7 C),

206.
Mouthpiece of pipe, 18, 24, 25, 26, 27.

Nicanor, of Cos, conmientator on The-
ocritus, 193.

NICOLSON, F. W., The Use of HercU
{MehercU), Edepol {Pol), Ecas-
tar {Mecastor) by Plautus and
Terence, 99 ff.

Nominative for accusative in predicate
after infin., in Ovid, 139 fil

Nonnus, 190.

Notes of the scale, origin of, 109.

Of Gi confused, 186.

d/3eX6}, in Herondas, 178, 180 ff.

in Homeric texts, 180 f.

in Plato, 180 f.
Oboe, mouthpiece of, 22.
SKiioi, 28.
Omission of subject-accusative of infin.

in Ovid, 117 ff.
Omissions, how indicated in Herondas,

Open pipes, 2.
ops, meaning of, 147.
opto, etymology of, 146.
orichalco vincta, refers to bands, 7.
Ovid, omission of subject-accusative oi
infin. in, 1 1 7 ff.

ratdtjcof (ai^XoQ, 39.

PanVpipe, i, 18.

vapdypaiftos, in Herondas, 177 ff.

irapdrpriTOi, 39.

waparp^rilM, II, 31.

jrapOiviOi (aUkoV), 39.

Paul the Silentiary, imitator of Non-
nus, Antipater of Sidon, and of
Latin poets, 190.

Tttpap iXiirBat (2 501) and Jtfanus
Corner Ho, 151 ff.

vttpap and vipat, 164 ff.

ireXra<rrcU, appearance of the word, 83.

General Index.

215

Pericles, brother of Theocritus ( ?), 194.

v€piffT6/uop, 29.

Philaenion, 197.

Philetas, 194.

Philicus (Philiscus), tragic poet, 195.

Philinus, of Cos, 194 if.

Philocles, favorite of Aratus, 194.

0op/3e£a, 29.

Phrygian pipes, 37 f., 42.

Pipes, see ai\6t.

irXa7(avXof, 13 &

wXawodlas, vXi^voSlat, 209.

Plato, Legg. (817 C), explained, 206.

Plautus, use of Herde, Edepol, and

Ecastor in, 99 ff.
vXri/jifuX'fis, meaning of, 209.
irX^r, older meaning of, 209.
Po/, use of, by Plautus and Terence,

99 ff.
Polyphonic pipe, 44.
praetor^ use of the word, 202.
Prologue of the Rhesus, 71.
pronomtiSy 20 f.
Pulling-matches, 152 ff.
Punctuation in Herondas, 169.
vvpdypri, 1 77.
vjipaypop, 177.
vCpnarpoPf 177.

Quantitative changed to accentual

rhythm, 106.
Quantity, how connected with accent

in Latin, 105.

Range of pipes, 50, 40.

Reed instruments, types of, 21.

Reeds, 21, 26.

Rhesus, dwa^ ttprifUpa in the, 72, 73,

76.

dva^ rpaytpdod/mwa in the, 72, 74.
authorship of the, 95 E
expressions peculiar to the, 78.
metrical structure of the, 91.
prologue of the, 71.
S3mtax in the, 92 ff.

Rhesus, words pectdiar to, and Aeschy-
lus, 79.

and Euripides, 81.

and Sophocles, 80.

and Aeschylus and Euripides, 79.

and Aeschylus and Sophocles, 79.
Rhythm, accentual, 105.
RoLFE, J.C., The Tragedy Rhesni»

61 ff.
Rope-pulling, 152 E

<rdXT«7$, I.
Sapphic verse, 105.

changed to accentual, 105.
in Christian hymns, 109.
Sappho, treatment of Sapphics by,

109.
Satumian verse, 105.
Saxophone, 22.
Simias and Simon, 199.
Simias of Rhodes, 195.
Simon, rival and friend of Aratus,

192 ff.
Simonidae at Cos, 195.
Single pipes, 12, 18, 20.
aKairapSid^aif 164.
ffKairipdaw IXirtcr, 152, 164.
o-jciyrdf t^^, 2o6.
amraXLaf 39.
Smith, C. L., on Hor. Sat (1, 4, 39},

204.
COf G0» confused, 174.
Sophocles, imitation of, in the Rhesus,

83 f.
Trach. (56) emended, 202.
words peculiar to, and the Rhesus,

80.
words peculiar to, and Aeschylus
and the Rhesus, 79.
Speaker, in pipe, 31 ff.
Spondee, permanent in Latin Sapphics,

106.
a^paylt, 188.

rrty/Aol, in Heronda% 1 75 ff., 185.
Stopped pipe, 2.

2l6 General Index.

SdbfC Ct-iiJMiit 'e of mfim. wiittcd a P^ra^ 42.

«ij^t I, iS^ I9» 5:^ 33. plvilaa

Syilnrpfci, mdittted byccet. 17a | Ai !■■•!, 35,42.

(«At0.39^

Cmmb, pqKS Bot pfa^ed a, 44.
oC Hcf^ Edepol, ud v«Ki«liiip, dorfie pipes m, 21.
pqxs ■aaed a tte itntgiff i lirt of, . ^^^j^^^^j^ ,

Tcatnu tiiiM of HcnadH^ Io6l ;_ _t - .£.

itCos^ 194 C
sdKifiAOB, 191 C
{see al» aAfo), i.
., 42.

INDEX' OF IMPORTANT CITATIONS.

Aero, Hor, A, P, (202), 4.
Anth. PaL (v. 217), 189.
Apuleius, FJcr, (iii), 45, 46.

Met. (xi), 14.
Arcadios, de accent, (p. 1S8), S.
Aristophanes, Achar, (681), 21.
Aristotle, de audit, (p. 801 b), 16 ; (p.

802 b), 26 ; (p. 804 a), 26, 32.
/r<^/.(xix, I2),45; (xix, 23), 3, 19.
Aristoxenus, Harm, (p. 28), 30; (p.

60), 2, II.
Athenaeus (iv, 76}, 12; (iv, 12), 13;

(>▼. 79). 37; (>▼» 80), 2; (iv,
82), 18; (iv, 34), 36; (xiv, 31),
6,21; (xiv, 36), 3.
Bo€thius, Jnst, mm. (i, 3), 44.

Cicero, Mur, (12, 26), 155 f.

Euripides, Ale, (229), 201.
Eustathius (p. 882, 24), 187.

Favonius Eologius, 4.

Gaius (iv, 16), 155; (17), 155.
Gellius (xx, 10), 157 it

Herondas, ilftmt,! (i), 174; (3), 170;

(7). 178; ("5). 185; (47)»i9i;

(50). 178; (54). 178; (55). »73.
187 E, 192; (64), 173; (65),

179 f.; (82), 177 f., 179.

II (2), 172; (6), 170; (64), 183;

(66, 67), 192; (73), 178; (74),

187; (83), 170.

III (10), 172, 196; (25), 172;
(26), 192 ff., 196; (36), 185 f.;
(45), 186; (53, 54), 196; (63),
184; (71). »74; (80), 172 ft;
(82), 183 f.

IV (10), 185; (24), 172; (32),
182; (35, 36), 179; (42), 172;
(43)» 177 U (50), 170, 182;
(55). »72; (59). 172; (62),
177; (83). 172. 176, 185-

V (7), 174; (9). 186; (18), 178;
(21), 176; (49), 170; (55), 179;

(68), 173; (70), 192.

VI (5). 176; (38). 185.

VII (35), 182; (48), 170; (71),

182; (76), 176; (77), 178;
(no, 118), 173; (123), 174.

VIII (3), 173 f.; (6), 186; (28),

173; (42), 176.
Proem (9), 173; (li), 170.
Hipponax {frag, i), 164.
Homer (Z, 143), 165; (H, 102), 165;
(402), 165; (M. 79). 165; (N,
358ff.).i65; (2,497 ft), 151 ft;
(T, 429), 165; (*, 350), 166;
(7. 433). "66; (€, 289), 165;
(m, 51, 162. 179), 166; (x, 33.
40. 165.
Hymn. Del, Apol, (129), 167;
Hermes (75), 209.
Horace, Carm, (iii, 16), 190.
Sat, (i, 4, 39), 204.
EpUt, (i, 20, 3), 188.

A, P, (202), 7.

217

2l8

Index of Impartamt Citations.

loacnpCkxis: of Cot (FUon-Hida),
195 fi^ '99-

LadsB* adv. Imd. (5), 21.

Timam (13), 18&
Lycophron, AUx. iSoS)^ 18&

Macxobras^ SMt*. Sap. (ii» 4, 5)9 4.
Muioi VktofiDiB (p. 44 K.}, 40.

NkomachiB (p. 9), 12.
NoimiB^ Diamys. (ii, 305), 189; (ihrii,
218), 191.
/M. (xxi, 139), 191.

Cjrn. (55, 368), 187.
Ond, FasL vf (381). 140; (804). 121 ;
T (167), 128; Ti (662), I.

Mti. ix (546), I4« ; »fi (M2), I4i-
i^ii/. i (5, 66), 141OL
Trist. i (1, 19), 128; n (10), 141;
in (4. 66), 134; (7, 9). 128; hr

(i 50. HI-

Bnil die Sikiitiaij, Amtk. FiaL ▼ (217),

189; (274), 191.

ix (12). 6; (30^ a), m;

« (7. 4),
(65).
Pindsr, PfA. i C8i), 167; ii (90). 164;

hr (220). 167; zii (i), 19, aa
Fliaj, iV. If. Tii (204)* 14, 36; zvi

(17O. 25.

Pbtob /«r- (817 C.). 206.

FfauitiB» Asm. (898). lOl; (93D), lOI.

Cms. (982), loa

Cu^ (50)9 100.

Merc. (719)9 loa

Trair. (210), loa
VisaiMtf^Xomfossesmaviiir{p. 1096 A)^

33-
de wnts, (r), i; (xxi), 33.

PtJhix, IT (69), 19; (70). 29; (74), 1.

12, 14, 30; (75), 13; (80), 4.

5.K^37; x(iL4),28w
Fbqphjno (p. 217), 12; (249), 26.
PlrodiB, AU. (p. 197), II; (cbap. 68),

31-

Qointiliiii, i (11, 6), 34.

Sdiofia, HieocntiB* Tii (6), 193; (125,

Ambr. JT), 192 L
Serrias, md Aem. ix (615), 12^ 35, 38;

n (737). 36^
Sophocks, Track. (56), 203.

Stobsen, EcL i (2, 9), 167.

Fin-. hxxT (27), 18&

TbeociitBS^ n (96^ 97). 197; to, 194;

(125), 192 ff.
TheophzastiB, Hist. PkmL rr. (11), 21,

TlnKfdides, n (29), 83.

Varni^ R. R. L (2, 15), 44, 461

3 blQS QQ7 SAB bSB

1

^

1

o

,

CO S

■

o» O 1.

00 H

•

H ?

■

d .

ja ■

• p#

O ! 5

■

-H ■ /.

o> «

'

in «

■

•

Q H -! 1

o o

;

H ^

■

^

•

•H

•

a

; .^ ?

•SI3 '5

■S'-^

U «

1 « ■ ■

9

' V ^>^

it

5|

■ 0: •■

1

u a:

I

oepl9 4!*. I