Transactions of the American Philological Association

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 marginalia 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 this resource, we have 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 individual
personal, non-commercial purposes.

and we request that you use these files for

+ Refrain from 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 attribution The Google “watermark” you see on each file is essential for informing 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 liability 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
ai[http: //books . google. com/|

da
fe
wand
te
te
=]
1S)
=|
FF

Grujamin

ad

TRANSACTIONS

OF THE

AMERICAN

PHILOLOGICAL ASSOCIATION,

1x -X
1878.

PUBLISHED BY THE ASSOCIATION.
PRINTED BY THE CASE, LOCKWOOD & BRAINARD CO.,
HARTFORD.

1879.

TRANSACTIONS

OF THE

AMERICAN

PHILOLOGICAL ASSOCIATION.

1878.

. PUBLISHED BY THE ASSOCIATION.
PRINTED BY THE CASE, LOCKWOOD & BRAINARD CO.,
; HARTFORD.
1879.

(kL 2-13

VI.

CONTENTS.

. Contributions to the History of the Articular Infinitive.

By Professor B. L. GitDERSLEEVE.

. The Yoruban Language.

By Professor C. H. Toy.

. Influence of Accent in Latin Dactyhc Hexameters.

By Professor M. W. Humpureys. .

. Observations on Plato’s Cratylus.

By Juuius Sacus, Px.D.

. On the Composition of the Cynegeticus of Xenophon.

By Professor T. D. Szymovk.

Ehlsion, especially in Greek.
By Professor M. W. Humpureys. .

PRocEEDINGS :—Tenth Annual Session, Saratoga, 1878.

M 9824

qn

84

TRANSACTIONS

OF THE

American Partorogrcat AssocraTtion,
1878.

I.— Contributions to the History of the Articular Infinitive.

By BASIL L. GILDERSLEEVE,

PROFESSOR OF GREEK IN THE JOHNS HOPKINS UNIVERSITY.

In examining, with some advanced students, the current
statements that Thucydides was a pupil of Antiphon (see Blass,
Geschichte der Attischen Beredsamkeit, 1. 85), and that Demos-
thenes was an admirer, and to some extent, an imitator of
Thucydides, I marked out some grammatical and stylistic
categories which I thought worthy of special observation ;
and among these the use of the articular infinitive, partly
because the history of the combination had interested me for
several years, partly because I thought I had noticed that
there was a certain coincidence, both in special handling and
in proportionate employment. From the examination of these
authors I proceeded to look into the usage of the other
orators, and thence to the closer study of the general question
involved. My treatment has not been exhaustive, and hence
I dare not formulate with confidence. My examination has
been limited to Pindar and the dramatic poets, Herodotus,
Thucydides, and the Attic orators, except Hypereides, and I
only undertake to register progress for the sake of those who

2

6 -B. L. Gildersleeve,

may be richer in leisure, or more accurate in method; and
the modest title of contributions must be my cover against
the charge of rash generalization.

I have no new theory:of the infinitive to advance, nor shall
I venture on the comparative side of the study, although I
know that in the present drift of research it will be hard to
gain consideration for any view which does not take in a
large group of languages. Yet I am convinced that in this
investigation the only safe course is to follow the special:
development of Greek, and in support of this conviction, I
would offer a few preliminary remarks. |

After considerable debate the form of the Greek infinitive
seems to be regarded by most of those who are qualified to
discuss the question as a dative abstract, although Curtius
has still something to say in behalf of a locative element,
as Westphal had done before him. - Now it matters very little,
so far as this investigation is concerned, whether the case-
form was dative or locative, or a blending of both, or whether
the infinitive forms in the kindred languages are perfectly
- parallel with the Greek infinitive or not. The use of the
article with the infinitive completes the deorganization of the
infinitive—deorganized before, it is true, yet, so to speak, not
confessedly deorganized. By assuming the article, the Greek
infinitive, though comparatively late, sunders its inflexional
connection with the substantive by a formal act, and bases its
claim to the character of a substantive on a foreign element.
That this divorce between the infinitive and its form did not
take place without a certain struggle, that there was a dim
half-consciousness, is shown, I think, at more than one point,
and a striking analogy to the uneasy conscience of the Greek
appears in our English handling of the verbal in -ing. The
very attempt to attach the article gave a little shock to the
sense of language, and it is not until we reach the Attic time
that there is any freedom, any license in the use. It is true,
as I have said, that the infinitive was deorganized before;
that it had become what some scholars have rather unhappily
called an adverb before. But the article is confession, and
that is a long stride, and one in which the other Indo-European

On the Articular Infinitive. 7

languages have not kept pace with the Greek. Add the
article, and the next step is the use of the preposition with
the article; and for a time this must have been to the Greek
a strange thing. He had taken liberties with the infinitive
before—he had construed it directly with piv, which, theo-
retically, would require the genitive or #, and there are traces
of an early attempt to combine avri directly with the infinitive ;
but with this additional innovation the sense of form revived,
as it were, and the prepositions seem to have worked their
way into use by degrees. Compare with this phenomenon
the limited use of prepositions with gerund and gerundive
in Latin. Thucydides, it is true, lets them in as a flood; but
Thucydides is abnormal here as elsewhere. With the use of
article and infinitive, the Greek language paused. There was
no further mechanical handling of the infinitive. The article
might take a demonstrative besides, as in ARISTOPH. Vesp., 89:
ipg re rovrov rov duaZev, but in classic Greek there is no parallel
for certain Latin constructions, such as are usually set down as
Hellenisms. See my remarks on Persius,1, 9. In such
Greek as that of Ignatius, we are not surprised to find (Ep. ad
Eph. 3): 7d adtaxptroyv hyper Civ; 11: ré ddAnSuvdv Ziv; ad Magn. 1:
Tov dcaravroc ipo iv; 5: 16 Ziv abvrod, all vulgarisms or Latin-
isms. It would, then, be perfectly justifiable to begin this
study on purely Hellenic soil, with the articular infinitive as
essentially an Hellenic product, and yet it may be worth while
to go a little further back and see how far the infinitive was
deorganized before it was thus stamped as a fossil.

A language retains its habits long after it has lost its con-
science. So in phrases and formule the Greek infinitive may
be regarded as having retained its dative, or, as some would
say, its locative sense throughout the whole history of the
language. So the complementary infinitive, the Saipa idéofac
of Homer, the Saipa dxotoa of Pindar, the aoc Saupdca of

.. Thucydides, and all the so-called ‘ loose”’ infinitives belong’ to

this earlier category. So the occasional use of the infinitive
after verbs involving motion is a remnant of the older time
and perfectly consistent with the function of the pure dative
in cognate languages, nay, in Greek itself. With a verb of

8 B. L. Gildersileeve,

motion, the dative represents the personal object for which
rather than the object to which, and as the strictly personal
character of the dative was effaced in the subsequent contumina-
tio, it is not surprising that this construction becomes less and
less common. Still, such verbs as zéprev, arooré\Accr, deddvat,
are found at all stages with the infinitive. But not to
consider classes in detail, the great mass of verbs that take the
infinitive as an object, may be summarily comprehended under
the title of Verbs of Creation, by which I mean verbs whose
office it is to bring about a result. And here it is well worth
notice how the original dative (for which) and the accusative
of the inner object meet—how the object for which, and the
object to be effected coincide. The chief of these verbs of
creation are verbs of will and endeavor—call them verbs of
asking, persuading, teaching, exhorting, or what not. They
all convey the notion of effort to an end, of will, of purpose.
Of will, of purpose, I repeat, not tendency, because the primi-
tive conception knows nothing of tendency in the modern
impersonal sense.

To these combinations the dative notion may not have been
foreign. So in English, when by dint of frequent use the to
had become a mere “sign,” there was added, in order to bring
out the final sense, a “for,” which was dropped when the
conscience had become seared. But while I have just shown
how dative and accusative might meet as to sense, the question
recurs: What was the infinitive to the Greek himself? If
anything definite, an accusative, it would seem. If Homer
says: Gobdkro vicny H 321, M 347, was that other to him than
éBovrero vexav. In AR. Eccl. 807 foll.: aGAX’ feev xaarog év
doxdly gépwy | reiv Guar’ dprov ad| ov cal dvo Kpoppuw | cal rpetc ay
éddag, the infinitive ‘drink ”’ is parallel with a loaf of bread,
two onions, and three olives—and so we can hardly recognize
a shifting of cases in Philem., frag. 167: airé &’ tyieay xparor,
elr’ ebmpatiay, | rpiroy b€ yaipecy, ctr’ dpeidery pyderi.

The use of the infinitive as an object, and as an accusative
object, led in time to its use as a subject. It became to the
Greek an accusative neuter. Now the neuter has no nomina-
tive, because the nominative implies a sentient agent, or one

On the Articular Infinitive. 9

so conceived, but the Greek language was not at the pains of
developing a special form for this occasional use of the neuter,
and in its capacity as an accusative neuter the infinitive was
treated as a nominative, despite its dative form. This use of the
infinitive as a nominative may be considered the final efface-
ment of the infinitive as a dative from the consciousness.

Having now followed the infinitive to the perishing of its
case-form, we must next examine the shifting of its temporal
relations, and this carries us to the consideration of the
use of the infinitive in oratio obliqua, a difficult subject, but
one which cannot be avoided in treating of the articular
infinitive.

As an abstract noun we should expect the infinitive to have
but three tenses—present, aorist, and perfect, say xocir,
rojoa, xexonxéva. The future infinitive, although formed
from the beginning of our record, seems to have been as much
due to the necessities of oratio obliqua as the future optative,
which is post-Homeric.

The three forms of oratio obliqua develop in the following
order: First, the infinitive form, secondly, the optative form—
which is chiefly post-Homeric, the optative for the indicative in
Homer being restricted to a narrow class—the interrogative—
and thirdly, o¢ with the participle.

The verb which controls the oratio obliqua clauses is a verb
of saying or thinking—which in the first two forms is almost
always expressed; the third form corresponding largely to
“‘ partial obliquity’? in Latin.

This oratio obligua construction, in which the infinitive no
longer represents the stage of the action, the kind of time, but
the relation of the action to the present, the sphere of time,
seems to have arisen gradually from the other class—the
verbs of creation—the verbs of will and endeavor. The
connecting link remains, and consists of the verbs of swearing
and witnessing, hoping and promising, verbs in which the will
is the deed. Two indications of this survive in the normal
language. The negative of the infinitive after these verbs is
with reasonable regularity pj, and the tenses follow largely
the older scheme—so that the aorist is used for the future—

A

10 B. L. Gildersleeve,

especially with verbs of promising in which ambiguity is
impossible. .

The deflection of the verbs of saying and thinking from the
verbs of asseveration and the like scems to be due to the
image of oratio recta before the mind of the speaker. Hence
the negative is oi, and the future infinitive represents the future
indicative. And here it is important to notice for the differ-
ence between Greek and Latin the closeness with which oratzo
obliqgua follows oratio recta in the one, the looseness in the
other. In Latin there is often no oratio recta present to the
mind, but the Greek is far more plastic, far less tolerant of
oratio obliqua than the Roman, and it is interesting to watch
how he feels his way to new combinations. So in Homer II.
9, 684 &» with the infinitive, which you can find anywhere in
prose Greek, is timidly used with direct reference to an
existing ay with the optative (v. 417).

As to the much debated accusative and infinitive, I will
simply say that by frequent use it formed a kind of abstract
compound, such as we find in the Latin gerundive, and to some
extent in the Greek participle, and so was employed as a
totality in various combinations and even as a subject. Cur-
tius’s explanation by a kind of prolepsis and confusion of
two constructions is to my mind utterly unsatisfactory. But
not to dwell on these points, let us hasten to the real matter
at issue. If the oratio obliqua infinitive is older than the
articular infinitive—as we all know—how can the articular
infinitive be limited to the category of the pure abstract
noun? Of course this is its ordinary function, and scores
of passages may be cited to show that the infinitive and the
abstract were considered parallels by the Greeks themselves,
and I shall have occasion to revert to this fact myself. If
Thucydides says (2, 87, 3) ro rdre rvyeir, that does not prove
that rvye7y has a real aoristic past time, for Demosthenes (37,
43) says just as readily rd» SdpuPov réy rére. Besides, the negative
vy which marks the difference between the infinitive proper and
the oratio obliqua infinitive shows that the articular infinitive
remains essentially an abstract. But inasmuch as (1) the
article may be prefixed to those forms of the infinitive which

On the Articular Infinitive. 1

are due to oratio obliqua alone, notably the future infinitive
and the infinitive with av, inasmuch as (2) the tense of the
infinitive is often suggested by the indicative context, inas-
much as (8) the general sense of a verb of saying or thinking
often seems inseparable from the complex, it is not going
too far to say that the articular infinitive may be used now
as a pure abstract, now as an abstract form of oratio obliqua.
That it is often used as substantivizing the imperative use of
the infinitive is also worth notice.

But it may seem hardly worth while to linger on so minute
a point as this difference between an abstract noun and the
abstract expression of an oratio obliqua relation, and I now
proceed to consider some chapters from the history of the
articular infinitive.

Homer: There is no articular infinitive in Homer; but
the nisus is there asin Od. 20, 52: avin cai r6 puAdocey ravvvyoy
éypnooovra: and assuredly in an Attic writer we should have
no difficulty in recognizing the articular infinitive here, espe-
cially as there is a marked tendency to use the articular
infinitive of disagreeable things. But we must interpret with
Nagelsbach: “It is another nuisance, this thing, keeping
guard all night awake.” Compare the familiar use of the
preparatory rouro.

In PinpDaR, in whom the articular infinitive occurs for the
first time, to any extent, the use is restricted. According to
Erdmann De Pindari usu syntactico, p. T5, there are but ten
examples; all of these, except one, in the nominative, and
that one in the accusative. Noteworthy is the position of the
articular infinitive, which in all the passages cited, except Ol.
.8, 60,—and that to be explained by chiasm,—is put at the
beginning as an object of thought, a real accusative, after all.
The aorist preponderates largely; seven times out of ten.
The rest are presents. There is but one instance of articular
accusative and infinitive.

Tae Dramatic Poets: The next group that I have inspected
is that of the dramatic poets. Aeschylus, Sophocles, Eurip-
ides, and Aristophanes. The statistics for the first two have
been furnished by Dindorf and Ellendt, respectively. Mr.

12 B. L. Gildersleeve,

J. H. Wheeler, late Fellow of the Johns Hopkins University,
has collected the instances of the articular infinitive in Eurip-
ides, and I have read Aristophanes myself to that end, my
_results being compared with those of Dr. E. G. Sihler, who
read the same author, independently, for the purpose. While
there may have been some oversight in detail, the general
result can hardly be wrong.

AESCHYLUS uses the articular infinitive within modest_
limits; chiefly in the nominative and the accusative—the
latter largely in the stereotyped combination ro pi (ro pi) od)
after verbs and phrases of negative result. There is, I believe,
but one example of the articular accusative and infinitive as
the subject of a sentence, whereas the accusative occurs
repeatedly with ré pf, which I consider a hint as to the way
in which the construction spread. The preposition is used
sparingly ; in fact, only three times, and in one of these the
reading is doubtful. The tenses of the articular infinitive in
Aeschylus are the present and aorist. There seems to be no
trace of substantivized oratio obliqua.

Aeschylus then is conservative, but less conservative than
Pindar. |

In SopHocues the vast mass consists of nominatives and
accusatives; there are very few genitives and datives, not
- more in proportion than in Aeschylus. There is a considerable
increase in the percentage of use—say one occurrence in one
hundred dnd twenty verses, whereas we find in Aeschylus
one occurrence in one hundred and fifty-nine, but the handling
is essentially the same. So prepositions are used sparingly
(xpd, €v, etc). The tenses are all present or aorist, except such
perfects as cidévac (Antig. 263), which does not count, and-
mepevyévat (Vv. 487), which hardly counts. Remarkable is the
substantivized oratio obliqua, Antig. 285. 6.

rijc EAwidog yap Epyopat dedpaypevoc
TO py Tadety Gv GAO TAY TO popotpor.

Here in the mouth of the watchman the article may be
considered deictic, and the twist in the expression may be
excused. The articular accusative and infinitive subject
oceurs Phil. 968.

On the Articular Infinitive. 18

In Evripipes there would seem to be a marked falling off.
According to Mr. Wheeler’s count there are not so many
articular infinitives in all Euripides as in Sophocles; and the
bulk of Euripides ig two and a-half times as great. The
occurrences number but one in three hundred and twenty
verses. It would be rash to account for this by the closer
approach of Euripides to every-day speech, as I did two years
since in the matter of é4» with the subjunctive. Still it is
worth noticing. Over forty per cent. of the whole number are
nominatives, but the genitives bulk much more largely than in
the others. Prepositions (cic, dea) and quasi-prepositions (xapos,
évexa, iw) are sparingly used. The tenses are present and
aorist, counting ¢i9eSa as a practical present. The articular
accusative and infinitive, as a subject, is rare, but not so rare
as in Aeschylus and Sophocles; and, on the whole, there is
somewhat greater freedom in the handling of the construction,
but it would seem as if it had not become pliant enough for
the poet’s purpose—who is iypdéc, if anything. The largest
number of articular infinitives occur in the Iphigenia at Aulis,
—by some considered his latest piece,—but this is not to be
urged.

Theory would require that ARISTOPHANES should not use
the articular infinitive so much as the tragic poets; and the
theorist who should begin his search with the Acharnians
would be gratified to find none in that play. But an examina-
tion of the other plays will not bear out this theory. Aris-
tophanes does use the articular infinitive less frequently than
Aeschylus and Sophocles,—once in two hundred and fifty-eight
verses,—but still much oftener than Euripides. The bulk
consists of nominatives and accusatives. Prepositions are
not very common (dz, é:a, év, xepi, bxd), nor are the quasi-prep-
ositions (évexa, xAnv). "The tenses are all present and aorist,
ciwSévac being a practical present. A large proportion of the
articular infinitives in Aristophanes are purely deictic, or
anaphoric; yet another class is exclamatory, both of these
belonging to what may be called the popular side of the con-
struction; and a considerable number are parodic. So the
cluster in Ran. 1477. 8: ro Jay pév éore xarSaveiv, | ro wveiy Ee

3

14 B. L. Gildersleeve,

Secnveiv ro d€ caQevdev xwdeov: and other gnomic passages bear
the same imprint. The variations in the different plays may
be of some interest. The Acharnians, as I said, contains
none; the Peace but one, the Lysistrata but two. In the
Clouds they congregate in the latter part, as might be sup-
posed. In the Plutus there are more in proportion than in
any other comedy ; a fact which may or may not be significant.

I now turn to prose. And first to the Historians; and first
of the ‘historians, Herodotus. HERropotos uses the articular
infinitive very rarely in comparison with Thucydides, who was
the first writer to appreciate its possibilities. According to
the count of Mr. Allinson, Fellow of the Johns Hopkins Uni-
versity, Herodotus uses it only thirty-two times; eight times
(probably) in the nominative, six times in the genitive
(three of these without manuscript authority, according to
Abicht), eighteen times in the accusative (largely negative). |
The tenses used are present and aorist; the perfect being used
of resulting condition, 4, 6: 76 éoriy3ac. The prepositions are
avri, pera. and éc. In three passages, 1, 210; 6, 32; 7, 70, the
manuscripts construe a»ri directly with the infinitive; a phe-
nomenon which [I have found again in later Greek. The
greater part of the examples-occur in the latter part of the
work.

A remarkable contrast is presented by THUCYDIDES. For
his usage I have depended on Forssmann: De infinitevt tempo-
rum usu Thucydideo in Curtius’s Studien, vi, 1. The articular
infinitive rises to an important element of the peculiar style
of Thucydides. While his bulk is only six to Herodotus’s
seven, he uses the articular infinitive more than eight times
as often and with great freedom. The genitive and dative
are liberally employed. Instead of a sparing use of preposi-
tions, he indulges in the construction without stint (fifteen
different prepositions), and absolutely riots in the use of &a
ré, which occurs seventy times. Of course present and aorist
tenses preponderate, but the perfect is also used, and besides the
articular infinitive with a», he uses the articular future infini-
tive, which is a bold step—every time, be it noted, with a quasi
oratio obliqgua dependence on such words as “hope,” “trust,”

On the Articular Infinitive. — 15

“ proof.’’? Of the use of the articular infinitive in Thucydides
as a kind of substantivized oratio obliqua in the other tenses, I
have not time to treat, and I regret exceedingly that I have not
been able yet to analyze the usage of Xenophon in respect to
the articular infinitive, especially as 1 am very much inclined to
think that he was influenced by Thucydides. The rest of this
paper must be devoted to the use of the articular infinitive in
the orators, who are of especial value to the student bent on
ascertaining the normal range of the language. I have studied
all of them except Hypereides to this end, but my statistics
are not so full as I could desire in regard to Lysias and
Tsocrates, although so large a proportion of each orator has
been read that there can hardly be any very great error. For
Antiphon, in addition to my own reading, I have had the
advantage of lists made by Mr. J. H. Wheeler, late Fellow of
the Johns Hopkins University. Mr. F. G. Allinson has done
me the like service for Isaeus and Aeschines, Mr. W. H. Page
for Lycurgus, Dr. E. G. Sihler for Dinarchus. Of Demos-
thenes I read about three-fourths myself, and the whole was
read by Messrs. Page, Wheeler, and Savage, Mr. Page under-
taking orations 1-18, Mr. Savage orations 19-34, Mr. Wheeler
the rest to 59 inclusive. The standard of measurement is the
page in the Teubner edition, which is fairly uniform for all.
contained in that series, except Antiphon and Andocides, who,
together with Dinarchus, had to be estimated. Of course,
I have: excluded from the count of the pages, documents,
introductions, and the like. It is not claimed that the result
is absolutely accurate, but sufficiently so to show the bearing
of the investigation. |

Quantitively the comparison of the orators shows the follow-
ing order of occurrences : ’

Lysias, . ; . . ; 12
Andocides, . . . . 20
Isaeus, . . , . , .25
Aeschines, - . . - . .30
Antiphon, . . . ; ; .00
Lycurgus, . ; ; . . .60
Isocrates, . . . . . .60

Dinarchus, . ; . . .80

16 B.D. Gildersleeve, ;

The variation in Demosthenes is remarkable, and I will
recur to it presently.

In the public orations the occurrence is . 1.25,
In the private orations .80,

‘which certainly serves to bring out very foreibly ‘the well-
known preference of Demosthenes for this construction. The
nearest approach to him is made by Dinarchus—the homespun
Demosthenes, the rustic Demosthenes, the xpiSivog AnpooSévnc
of the ancients. Bookish Lycurgus, umbratic Isocrates come
next. Then Antiphon, who uses it rather more freely than
Thucydides. Low down stand Aeschines, Isaeus, Andocides,
Lysias—Aeschines, the man of mere native cleverness, Isaeus,
the man of practical business talent, Andocides, by no means
a littérateur, and Lysias, in whom oc reigns and in whom
the narrative is the great thing. To come back to the
‘variations in Demosthenes. They are indeed great, and
would have furnished an illustration for the text from which
his admirer, Dionysius of Halicarnassus, is never weary of
preaching. Of course his case demanded especial care, and
I separated the public orations from the private, inasmuch
as it was to be expected that the difference of theme would
show a difference in the number of occurrences. I then
excluded from the count those of the public orations which
are open to suspicion, and found the average of the remainder
to be 1.25. The lowest average of the undoubted public
speeches is presented by the Second Philippic (vi), in which
the average is .87, the highest by the First Olynthiac, in
which the average is 2.75. Both of these are short orations,
and it may be unfair to judge by them, but it cannot be a mere
fancy that the large number of articular infinitives in the First
Olynthiac gives a peculiar tone to the oration. The long
speeches vary as follows:

XXIV. Contra Timocratem, . ; 1.06
XXI. Contra Midiam, ; , 1.10
XIX. De Falsa Legatione, . ; 1.18

XVIII. De Corona, . . , 1.85

XX. Contra Leptinem, ; , 1.54

XXIII. Contra Aristocratem, . . 1.62

On the Articular Infinitive. 17

The De Corona is almost exactly the mean between the
highest and lowest. If the private orations be taken without
criticism as they stand, the average will be about .80, but as
_ the genuineness of so many of these is assailed, statistics wil]
be of little satisfaction to those who share the popular tendency
towards a9érnorc. In the earlier speeches the average is lower
than in the later. So the first speech against Aphobus goes
as low as .26. In the speech against Spudias, which, to be
sure, 1s questioned, there is no occurrence, nor any in the
speech against Callicles, in which Demosthenes approaches
nearer to Lysianic Soc than in any other. The two lnghest
are the speech against Conon, a masterpiece in which the
decvornc Of Demosthenes and the simplicity of the supposed
speaker are curiously blended (1.07), and the speech against
Pantaenetus (xxxvii), in which the occurrences are: 1.06, and
which is a specimen of the grand manner by which Demos- .
‘thenes sometimes betrays himself even in his private orations.
The proémium is a massive period, better suited to a stately
public oration. | °

If I had time, I might treat of the variations in the other
orators, as for instance in Lysias, who ordinarily has no
fondness for the construction, and yet crowds an extraor-
dinary proportion into the speech against Philon (xxxi),
which, according to Blass 1, 477, marks an epoch in the
history of Lysianic art. So I might call attention to the
apparent coquetry of Isocrates with the construction in the
Panegyricus, but it is high time to bring this paper to a
close, and I must suppress what I had to say about the effect
of massing the infinitive, and about the rare construction
of the articular infinitive with a, and the articular future
infinitive, for which Demosthenes, of the orators, is our chief
warrant.

In conclusion, then, suffer me to say a few words as to the
stylistic significance of this construction. Is it a mere acci-
dent that one author employs the articular infinitive much
more frequently than another? Is the use determined as
much by the department as by the individuality ? It would
be rash, as I have said, to formulate, but the following con-

18 B. L. Gildersleeve.

siderations may be of some weight for the further investiga-
tion of the subject, if it should seem worth the while.

The infinitive has sundry advantages over the abstract
noun.

1. Language is capricious as to the development of its other
abstracts, while the infinitive is always ready, and not only
positively, but, which is a great thing, negatively. We find a
goodly number of negative abstracts formed by the help of
ay- (a privative) ; but they do not supply the needs of expres-
sion. We have an ddvvapia, an ddvvacia, an aduvaria, but how
often do they occur in comparison with pa) dbvac$ac? What have
we for p27 BovAcoSat, py EMEAELv, pa) pédAAecy ?

. Then the abstract noun often wanders off to a transferred
signification, while the infinitive has the original meaning of
the verb ; yrepn and yryrwoxey are not necessarily equivalents.

3. Besides, the abstract noun does pot always sharply indi- |
cate the stage of the action, as the infinitive does. [aSo0c can
be analyzed into racyey and raSeiv, A\dyoc into A€yerr and Ada,
xpdéaodoc Into rpocépxeoIac and wpoceASeiv. pagar is sharper than
mpakc. Tupavvevoar gives an element that ruparvic does not.
Biog and Savaroc are not so clear as Zv and Saveiv.

There. are similar considerations as to the voices. TeAa-
oSijva is Clearer than yéAwc—dAdorpiwSijrac than addXorplwece.

4. Finally, the infinitive takes up, with greater and greater
ease, into this abstract relation its subject and its modifiers,
and a whole complex is thus made the object or subject of
the verb, whereas the regular abstract with its dependent
genitive is less compact.

The only drawback then to the infinitive was the absence
of the article; and as soon as the article was added the
infinitive went on its new life.

The change was, as we have seen, prefigured in Homer, and
the deictic or demonstrative use seems to have been the popu-
lar use. So in the exclamatory infinitive, and the numerous —
turns in which the demonstrative is contemptuous just for
the same reason that otroc is contemptuous, and pointing 1s
contemptuous and object and objectionable are used in a bad
sense. Outside of this rudimentary. popular use, the spread

On the Yoruban Language. " 19

of the articular infinitive seems to be due to conscious ratioci-
nation, to the increasing tendency towards the employment
of abstract nouns in varied relations; and the articular infini-
tive is consequently a gnomon of the reflective element, and
cannot be left out of consideration in estimating the character
of style.

Il.—The Yoruban Language.

By CRAWFORD H. TOY,

PROFESSOR OF HEBREW IN THE SOUTHERN BAPTIST THEOLOGICAL
SEMINARY.

The main body of African languages, omitting the Shemitic
and the Hamitic, (the Berber, the Ethiopic, and the Egyptian, )
fall into three groups: the Hottentot in the south, the Bantu
occupying the whole center about as far north as the equator,
and the Negro lying in Senegambia and Soudan, the last of
which has as yet received little attention, while the structure
of the others has been carefully studied and satisfactorily
exhibited. On the Guinea coast, however, is found a group
of dialects almost wholly different in vocabulary and structure
from all of these, and offering interesting linguistic features.
This group includes the Basa and Grebo of Liberia, but its
most important member is the Yoruban, which is spoken by
a partially civilized population of about two million people
inhabiting the territory included between Dahomey, Borgoo,
the Niger, and the bight of Benjn. Its literature, which is
wholly the work of Christian missionaries, consists of collec-
tions of proverbs, Bible-translations, and a few other religious
books. Grammars have been written by Crowther, a native
Yoruban, now Anglican Bishop (London, 1852), and the
American missionary Bowen (Smithsonian Institution, 1858).

I. PHONOLOGY.

The phonetic system consists of letters and tones.
The letters are exhibited in the following table:

20 * | C.H. Toy,

VowsELs. | Ha .rFr-ConsonaNntTs. CONSONANTS.

g\¢\4/z E
giB/ 8] & r=
a2iaAa|;mila 7)
Guttural a | ai | h
, a | au
é | éi
Palatal, . : el | y
i
ft | | | Sonant. | Sard.
Linguo-dental Oo | oi. 4 . 8 njd t
rit ' 2 jz (in dz) 8 (sh)
Oe —- —— |---| — > -- |e flO. J
Labial, .

; oi | Ww f | mj| b /p(in kp)
| u | |

Besides the three primary vowels a, t, u, there are the sec-
ondary (diphthongal) e, 0, and the closed modifications of these,
a (in bat), e Cin let),7 Cin bit), o Cin not), u Cin full), and
perhaps, also, w (in but). The sounds of ai and au seem to
be real diphthongs ; in the others the second vowel is very .
slight, and wa, we, ui, wo are mere combinations of unmodified
vowel-sounds ; au is found in a few words only, mostly adverbs,
and probably compounds. There is a partially prevalent law
of vocalic harmony whereby vowels of personal pronouns and
prefixes are made to accord with those of verbs and roots ;
the form of the objective pronoun is dependent on that of the
preceding verb, thus: em: sha a “I wounded it,” 0 se e “ thou
shuttest it,’’ on ti ¢‘‘ he struck it,” enyin ro o “‘ ye provoked
it,” awon ru u “they stirred it,” and so on through all*the
vowel-sounds. The subject pronoun mo “1I,’* is used only
before verbs containing the vowels e,2, 0, u, while mo “I,” is
used only before a, e, 0. This law prevails to some extent in
prefixes to nouns, as, aba, ebe, bt, obo (there are no- words
beginning with « except the object-pronoun, uv * him’’), but
is not strictly adhered to, since it would make it impossible
to derive more than one noun from each verb. In this obedi-
ence to a law of vowel-harmony the Yoruban stands almost
alone among the African languages, and so far represents a

On the Yoruban Language. | 21

stage of linguistic development in which the sense of euphony
prevailed over the sense of signification, a fluid condition of
speech corresponding to the facility with which some bar-
barous tribes change their vocabulary.

The consonants, including the breath or aspiration A and
the semi-consonants, are eighteen in number, and represent,
as will be seen from the table, all the varieties except the
guttural (not counting the simple aspiration as a guttural
proper). There is only one spirant, the. labial f, but the
scheme of liquids, sibilants, nasals, and full breaks or conso-
nants is nearly complete ; the sonants occur somewhat more
frequently than the surds. The surd labial p, however, is
found only in the compound kp, which seems to come from
word-composition ; a similar origin is probably to be supposed
for three other consonant-combinations of frequent occurrence,
dz (7), yb, mb. 7 |

In its law of tone, whereby words spelled alike are distin-
guished in meaning, the Yoruban stands, to some extent, on
the same plane with the Chinese, though the system is less
. elaborate than in that language. There are three main tones:
the middle or ordinary, the acute or rising, and the grave or
falling. Thus: ba is “to lie in ambush ;” bd “to meet, over-
take;’’ ba@ “to bespeak.’”’ In general, each word in the Jan-
guage has its own tone, which it retains unchanged, and
which is‘a part of its form, whence the Yoruban has a distinctly
musical sound. But in connection with the personal pronouns
there is a law of tone-harmony by which these pronouns are
brought into a certain relation with the verb. The personal
pronouns all have, normally, the middle or ordinary tone, and
this they retain when they express the subject or nominative
" case, and in the shorter objective forms when they are gov-
erned by verbs having the acute tone; but after verbs with
middle or grave tone the object-pronouns take the acute tone,
as: okonri: bd mi “the man met me;”’ but ode ta 6 “a wasp
stung him ;” babba la 6 “father beat him.” In compound
words the rule of tone becomes more complicated ; in general
there is an attempt to contrast and distinguish by difference
of tone.

4

“4

23 C. H. Toy.

Polysyllabic words usually have the accent or stress of voice
on the penult, with a secondary accent on the second preceding
syllable ; but in derivatives a verb having the rising inflection
commonly takes the accent.

Changes of letters. Vowel-elision is common, and is gov-
erned by rules, in general a short vowel yielding to a long
one (diphthong or grave) ; but there are many complications
which have to be learned by practice. In addition to the law
of vowel-harmony above mentioned, the o-vowel is subject to
a modification, becoming o before e and 9, and there are
other interchanges not governed by recognizable law. Among
the consonants the only regular euphonic change is that of 2
in nt, which becomes iz before vowels; and the interchanges
of & and g, s and sh are found. With these slight exceptions
the language shows fixedness and precision of form.

II. MORPHOLOGY.

1. Roots and Words. The roots are probably all mono-
syllabic, most polysyllabic forms easily resolving themselves
into simpler elements, as, olubodzuwo “ inspector,’’ made up |
of olu ** chief” (from lu * to beat’’), be * to go,” odzu * eye,”
wo “visit” =‘ chief man who goes to visit with the eye.”
The roots consist mainly of consonant and vowel with or
without nasal appendage; a few pronouns and adverbs con-
sist each of a single vowel, but these may not be original.
The words that begin with two consonants seem to be com-
pounds (but in some cases these double consonants seem to
be phonetic derivatives from single ones). No root is made
by vowel plus consonant.

Inasmuch as the roots are monosyllabic and begin with con-
sonants, there is a partial form-distinction between root and
word in the case of many nouns formed by vowel-prefixes ;
and this derivation serves also to distinguish so far between
verbs and nouns, since no verb begins with a vowel. Mono-
syllables beginning with consonants are either verbs, pronouns,
or particles (which last are originally Verbs or pronouns or
interjections emotional or imitative); apparent exceptions
come from contraction. Reduplication also distinguishes the

On the Yoruban Language. 23

noun-word from the root. The Yoruban, in this respect,
occupies a midway position between the inflecting languages
proper, and those in which no difference is made between root
and word. .

2. Word-composition. Here the language is very rich.
Composite noun-forms are made by derivation, by reduplica-
tion, and by composition proper; in fact, we are warranted
in saying that all nouns are thus formed, the simple root now
appearing only as verb or pronoun. We find, therefore, as is
to be expected, a great variety of prefixes, monosyllabic and
dissyllabic, which are combined with verbs simple or com-
pound. ‘The commonest prefixes are a concrete and abstract,
with the dissyllables ati abstract and ad: concrete; as abo
‘‘shelter,”’ ¢fo “ the act of washing,” atibo ‘the act of coming,”
abila ‘“‘ that which is striped.’"” The vowels e and o are also
used to form concrete nouns, and there are other dissyllabic
prefixes, as ada, which denotes result, abu, mostly concrete,
and afi=“the maker” (from fi “to make’). The mono-
syllabic prefixes are probably demonstrative pronouns ; abo =
“that which shelters ;’’ we find similar uses of the pronoun
in the present syntactical construction of the language. The
law of vowel-harmony, by which the vowel of the prefix
would be assimilated to that of the verb, has been modified
by the necessity for variety of signification; usage has, in
some cases, fixed one meaning to the noun formed by a, and
another to that formed by e, though in other cases these
prefixes interchange without change of meaning. The dissyl-
labic prefixes are themselves nouns formed from verbs, but
no longer used separate; adi is from the verb di “ to beget,
be,” ati probably from a root tz “to finish”? (compare titi
‘wholly, continually,” and the auxiliary t expressing com-
pleted action), abu from bu “ to give,” or 57 “to be.” Some-
times the verb without the prefix appears as the first element
of a compound ; but in this case a prefix seems to have been
dropped, as buba “a hiding-place,” from 752 ‘‘a place,” and
tba “hiding.” By she combination of @ with the verb i
(= nz) ‘to have,” is formed the prefix al (el, ol, etc.), signi-
fying ‘ possessor,’ as alake ‘‘the lord of Ake” (title of the

24 C. H. Toy,

king of Egba), alatye ‘‘ owner of the world,” from atye “‘ the
world” (from ye “to live’’). The prefix a is also used as a
negative: zda ‘‘created ’’ (from da “‘ to create’), aida “ un-
created’ or “ uncreatedness,”’ to which may be prefixed al,
alaida ‘‘ possessor of uncreatedness,”’ “ uncreated.”’

Reduplication. The original and simplest form of redupli-
cation in the formation of nouns is the doubling of the verb-
root, a8, kpejakpeja ‘“‘a fisherman,” from kpeja “ to fish,”
sisesise “a laborer,’ from sise “ to labor.’”’ In order to obtain
variety of signification the vowel of the root is sometimes
changed: from ga “ to be high,’ comes gaga “ closely,’’ and
giga “hight,” also gigagiga “ great hight ” and “ loftily,” from
le “to be hard,” lile (and lilelile) “‘ hardness.” Where the
verb is a compound of verb and noun, only the. former element
is doubled, as from dara “‘ to be good” (= da + ara), didara
‘‘ goodness.’’ In.this formation of nouns by simple repetition
of the idea of the root, the language retained consciousness
of the distinctness of the elements so far as to insert various
descriptive syllables between the two components; thus, by
inserting the indefinite pronoun 42, an indefinite sense (easily
passing into one of contempt) was obtained, as eiyeketye ‘* any
bird,”’ from ezye ‘a bird,” entakenia“ a contemptible ‘person,”’
literally: ‘ person—any—person;”’? the demonstrative iyi
_ gives emphasis, as ekuruiyekuru “the very dust’? (= dust—
this—dust) ; the adverb 72 “ always’ (perhaps from 77 “ to
see, appear, be ’’), expresses perpetuity, as in Zyeraye “ always
living,’ (from aye “ alive’’).

Noun-composition proper. Besides the simple juxtaposition
of two nouns, the defining following the defined, as in Hebrew
(as, omo ehin “ child of back,” “ follower, disciple,” tle tubu
‘house of prison,’ ‘jail,”) long agglutinations are made,
the various clements of the complex idea following in general
the English order, as, zbaiyedze “a turning the world upside
down, spoiling the world,” literally, “‘ meet-world-consume,”’
afibikpore “an ungrateful person,” literally, ‘* one-make-evil-
call-good,”” that is, one who returns evil for good ; adogunsille
‘‘a revolutionist or disturber,” literally, a-da-ayun-si-ille ** one-
make-war-against-land” ; afemojumo “ morning,” is “that ,

On the Yoruban Language. — 25
which desires the light of the eye of light (dawn).” Com-

pounds are made also by combining nouns and adverbs, and

verbs and adverbs.

The following list of nouns from the root tu ‘ease, recon-
cile,”’ will illustrate the facility of the Yoruban in noun-form-
ation :

itu ‘‘ ease.” letutu ‘‘ reconciled.”

aitu ‘‘ uneasy.” eletututu ‘‘a reconciler.”

laitu ‘‘ uneasy " (prefix ¢ ‘‘to have”). letututu ‘state of reconciliation.”
alaitu ‘one who is unreconciled.”. atletutu ‘‘ unreconciled.“

etutu. ‘‘ reconciliation.” laitetutu ‘“unreconciled.”

alatletutu “one who is unreconciled.”’

Compound verbs. Inthe formation of composite verb-forms

also, the language proceeds in an agglutinative way, the dis-
tinct. signification of the different components not being lost
sight of. Sometimes the root is simply doubled, as in be, bebe
‘‘ beg,” belebele “to be thin,” toto.‘ to be whole,” but this is
rare, most. such doublings being nouns (or adverbs). More
commonly two different verbs are combined to express a single
(tho complex) idea (as occurs frequently in Hebrew, less
frequently in English). But in such combinations the signifi-
cation and force of the verbs remain always separate and
distinct, so that nouns and prepositions may intervene be-
tween them in the sentence; only the second verb sometimes
occupies the primitive neutral position, neither noun nor verb,
but the bare conception, which is the basis of both. #2 lu
‘* beat’’ is, literally,*‘ make-beat,” fidzo ‘ make-burn,” “burn” ;
the idea ‘“‘show ”’ is expressed by fi .. han ‘“‘ make .. appear,”
o fi won han mi “ he-make-them-appear-me’’ = ‘‘ he showed
them to me”; nwon mu u wa si Damascus “ they-make-him-
come-to-Damascus”’; mo ba iwe dze** I-meet-book-consume’’ =
“I destroy the book’’; or, the object of the action being the
subject of the verb: iwe ba dze “ book-come-consume,”’ *‘ the
book is spoiled”; ha ti re loh “ come-of-thee-go”” — “ go thy
way”; ta ni yio si duro nigbati o ba fi ara han “ who-it is-
shall-and-stand-when-he-come-make-himself-appear,” “ and
who shall stand when he appears’? o be ole lori (lit ort “as
to head ’’), ‘* he-cut-thief-asto-head "’ = “he beheaded a thief.”

-

26 OF H. Toy,

The insertion of a preposition sometimes gives a transitive
sense to a phrase: eru ba mi“ fear comes me” — “I am
afraid’; o ba mi leru (= li eru) “ he comes me in fear’? =
‘he frightens me”; ino mz badze “ mind-my-is troubled’ =
“Tam grieved’; 0 ba mi nino (= ni ino) dze “‘ he-meets-me-
in mind-spoil”” = “he grieves me.”’

Foreign words. There are a few words taken from Arabic,
as tuba ‘ repentance,” kefert (kaphir) “‘ unbelievers, gentiles,
heathen ;’’ from English, as dann (of marriage), and the
Bible-words baptisz ‘‘ baptize,” temple “ temple,” furlong, and
from the Haussa language, sinkafa “rice,” takarda “ a book.”’

3. Inflection. As may be inferred from the forms already
given, the Yoruban does not belong to the class of inflecting
languages proper. It has some prefixes that have almost or
quite lost their independent character, but the duty they perform
is simply to convert verb-roots into nouns. As we shall see,
there is a slight attempt to mark relation by different forms
of pronouns, and there are certain temporal and modal words
that have almost droppeq their original meaning and become
signs of relation, though they almost all retain their form
unchanged, and are independent words. The language is,
therefore, not agglutinative in the sense in which that term is
used of the Turkish, for instance; it rather exhibits the first
simple attempts of a primitive root-language to employ certain
of its words to mark the distinction between the two main
‘classes of substantive words, nouns and verbs, and to indicate,
in a general way, the temporal and modal modifications of
the latter, to which must be added that it has distinguished
the particles (prepositions, adverbs, conjunctions, ) with clear-
ness. |

A. Nouns. Nouns are without inflectional signs of gender,
number, and case. Gender is marked, as in English, by
different words, or by sex-words prefixed, but only where
there is real distinction of sex; English, by a long process of
unburdening, has reached the primitive position beyond which
Yoruban has never passed. But while English retains the
idea of grammatical gender as a heritage of its long historical
development, this idea does not properly exist in Yoruban at

On the Yoruban Language. 27

all; it has only sex-distinction, and this is,-of course, con-
fined to substantive nouns (there is no mark of gender in
adjectives or pronouns). Difference of sex is sometimes
marked by different words (as sometimes in Indo-European
languages), as baba “ father,’’ cya “‘ mother,” Okonri “ man,”
obtrt ““ woman,” where there is no indication of gender, but
the objects are regarded as distinct and independent, and
have names given them in accordance with some prominent
characteristic ; obirz is from 62 “ to bear,’ okonrz (from oko)
perhaps from xo “to rule, guide.’ However, the words ako
and abo naturally came to mean “ male animal ”’ and “ female
animal” respectively, and were then used to mark sex where
there was only one word for the species of animal, akomali
“bull,” abomalu ‘‘ cow,” ako esin “ horse,” abo.esin “ mare”’ ;
ommo “ child ”’ is defined by an added sex-word, as ommo
konri “ boy,” ommo binri “ girl,’ and so several other words.
In the case of compound words, the sex will, of course, he
marked by the sex of the principal component, as bale (=
oba ile) ‘lord of the house,” iyale (iya ile) “ mistress of the
house.”” As the language marks only real sex-distinctions,
the conception of tke “neuter” gender does not exist.

For number also there is no inflectional sign ; but plurality
is marked with sufficient distinctness, usually (where the con-
nection does not. make it clear) by the personal pronoun
awon “they”? put before the noun, as Saulu nmi ilo ati pina
st awon ommo-ehin Oluwa‘* Saul-was-breathing-accusation-and- |
slaughter-against-them-disciple-of the Lord,”’ “ against the dis-
ciples of the Lord” ; this awon must be repeated before each
separate word; or, if the noun refer to a person addressed,
the second personal pronoun enyzn “ ye”’ is used, as ki enu ki
omase ya nyin, enyin arakonri,*‘ do not wonder, ye brother,’ =
‘‘brethren’’; the more emphatic demonstrative pronouns
wonyi * these”’ and wonni “ those ’’? may also be employed in
the third person. Plurality is expressed, also, by simple
repetition of the noun, and this repetition is necessary when
the idea of reciprocity enters.

The relation of case is, in general, determined by the’ posi-
tion of the noun; the subject standing before the verb, and the

28 C. H. Toy,
object, usually, after it. For the sake of emphasis the objéct
sometimes stands first, and its syntactical character is made
clear, either by the connection of thought, or by the collocation
of words, as on li a wi fu * him-it is-we-spoke-to’’ = “‘ we
told him.’’ So, the defining relation in which one noun
stands to another is frequently expressed by position merely,
the defining word following the defined (as in Hebrew):
okko obba * ship-king’’ == “ king’s ship.” This relation is
sometimes expressed more exactly by the relative pronoun &
(quite as in Aramaic), as ona ti ilu * road of town,”’ literally,
‘‘ road-which-town.”’ All sorts of - relation-are thus indicuted,
but the insertion of the ¢i is necessary when possession is to
be predicated, or where without the ¢i the second noun might
be supposed to be in apposition with the first: Atiba oba
“Atiba the king,” but Atiba ti oba “ Atiba (the servant) of
the king.” In this case the relative pronoun seems to have |
its own proper. force, = ‘‘ which” or ‘‘ that which”; so in|
such a sentence as: nwon se ti orisa, literally, ‘* they-do-what-
idol,” that is, “ they do what pertains to idols, they worship
idols.”” The ¢ is much employed to express this general,
indefinite sort of relation, which is left to be understood from
the connection. The person addressed is indicated by the
simple noun, with or without personal pronouns, or by some
interjection, as o (put after the noun), or some demonstrative
pronoun indicating greater or less nearness (y?, when the per-
son addressed is quite near the speaker, na, when he isa short
" distance off). a

Adjectives. The Yoruban seems not to have differentiated
the adjective proper; its descriptive words are predicative,
that is, they are treated as if they included the copula, and
therefore fall, technically, into the class of neuter verbs. To
indicate that a quality pertains to a substantive, the name of
the quality itself (abstract noun) is appended by way of defi-
nition, as ohun didara “ thing of goodness’? = “ good thing”’;
or the possessor of the quality (concrete noun) is preposed in
apposition, as alagbara enia “‘ strong one-person’ = “ strong
man’’; or a descriptive relative phrase is added, in which the
neuter verb is used, as zda tz o mu “ sword-which-it-is sharp ”’ =

On the Yoruban Language. 29

‘a sharp sword”’; or, when the qualified substantive is the
subject of a verb, the quality is predicated of its subject in a
simple sentence, and the affirmation of the verb is added in a
relative sentence, as enia re lt o se e ‘“‘ person-is good-it is-that-
did-it ’” = “‘a good person did it.” It may be said, then, that
the function of our adjectives is wholly performed in Yoruban
by substantives and verbs; the abundant use of neuter verbs
is a sign of failure to differentiate either the adjective or the
copula.

Comparison. Hightening of intensity is expressed by dew
and dzu loh = “ exceedingly”’ and “‘ more”’; in connection
with numbers le is used. The highest degree of intensity
may be expressed by dzu gbogbo (= “ surpassing all”’) or tan
(= “completed ’’), put after the adjective.

Numerals. The numeral system is very well developed, beth
in extent of numbers and in the expression of the various
relations of the numbers. The ten units are as follows: enz
(and okan), edzt, eta, erin, arun, vfa, edze, edzO, esan, Ewa,
twenty is ogun, thirty ogbon, two hundred tgba; of these the
last-named =“ heap, ’’so called because cowries (the shell-coin
of the country) are counted in heaps of two hundred each ; the
origin of the others is obscure. From 11 to 14 the numerals
are formed by adding la (= “ great’’) to the units (okanla,
edzila, etc.); from 15-19 by subtracting the proper unit from
20; from 21-24 by adding units to 20; from 25-29 by sub-
tracting units from 30, and so on; multiples of 20 are used
up to 180 (40 is ogodzi = twice twenty, and so on), the inter-
mediate tens are made by subtraction of ten from the next
higher (fifty, adota is sixty, ogota less ten ewa); in the same
way multiples of 200 are used. The first unit enz is used
only in counting ; Okan is employed independently, =“ one
person,” san with anoun. A singular usage in connection
with the other units is the prefixing of m to them when they
are attached to nouns expressed or understood, as enia mewa
‘‘ten men’”’; in Bowen’s Grammar it is suggested that this
m is from the verb mu “to catch”’ in the sense of “ amounting
to,’ on the ground that when an African [Yoruban ¢] speaks
in English he generally says ‘he catch ten” for “ there were

=

v

80 C. H. Toy,

ten.”’ Ordinals from 1 to 19 are made by prefixing ek (or ek,
according to the law of vowel-harmony) to the cardinals: as,
ekini “first,” ekedzi “second,” etc. (or, kint, kedzi); they
follow the noun. Examples of distributive numerals are:
metameta “ three by three’’ (doubling of meta, which is from
eta by prefixing m); okokan “‘one by one” (reduplication of
first syllable of okan, where the m is not prefixed); dkdkan
“one cowry each”’ (from dkan “ one cowry’’).

B. Pronouns. The substantive relational material of the
language (which, for convenience, is treated here before the
verb) is comparatively full, yet simple; the varieties of form
seem to come from the working of euphonic rules rather than
from effort after delicate distinctions of thot. 1. Personal
Pronouns. These are entirely without inflection, and lacking
not only in the rich generic development of the Hottentot,
but in all distinction of gender. It will be more convenient
to take the persons one after another.

First person. The singular is emt, the plural awa, between
which we should not expect to find any obvious relation (since
‘“‘we’’ is not the plural of “I,” but =‘ and others””). These
may be used everywhere, under all circumstances. But, in
accordance with the law of vowel-harmony already explained,
there are two modified forms, the open mo used optionally .
when the first vowel of the following verb is e or 0, and the
' close mo, used when that vowel is e, 2, 0, or vw. If, however,
the verb is future, the euphonic forms are not allowed ; another
form n (ng) is then employed, and this is found also before
the negative ko (ko) “not.” This ng seems to be the nasal-
ization of the m of emz, apparently a euphonic change induced
by k and the o of the future: ng o ri “J shall see,” ng ko ri
“I do not see.”” The abbreviated form mi sometimes stands °
absolutely at the beginning of a citation, the verb of saying
being omitted: mz nibo “I (asked) where ’?’’ (which comes
possibly from contraction with the verb wi“ to say”). When
the pronoun stands as object, frequently the full forms are
employed, but sometimes the abbreviated forms mi, wa (“me,”’
‘‘us””), where the abbreviation is euphonic and not inflectional,
(and possibly these shorter forms are the original). These

On the Yoruban Language. sil

latter are used in general when the pronoun follows the verb
or the noun in a simple and unemphatic way ; but if there be
any emphasis, as, if the pronoun stand at the beginning of
the sentence, or if the relative ¢ precede or follow it, or if two
pronouns be connected by a conjunction, or if the reflexive
na (‘self’) follow, or often for euphony, the full forms are
employed.

Second person. Full forms: zwo “thou,” enym ‘“ye;”
euphonic, open 0, close 0; there is no special form before
future verbs. The short forms after verbs (object) are 0 and
nyin ; but after nouns, while the plural form is the same, the
singular is not 0 but re, the origin of which is not clear; it
seems to be a demonstrative pronoun or a noun (it is found
in the third person also).

Third person. Full forms, used for subject, on (on), awon
(and nwon); the plural has also the short form a; the
euphonic open is 0, and close o; the short ¢ used in citations
(as mz above), possibly out of o we “he said.”” Before future
verbs is sometimes found the demonstrative pronoun y? in the
sense of ‘“‘ he,’’ for which in some cases a is used. The usual
plural subject is nwon; awon is used before the relative t (and
frequently as plural sign before nouns, as above explained).
As object after verbs the plural third person is won; the
singular shows a great variety of forms, 0, 0, u, a, e, e, ¢,
conforming itself, according to the Yoruban law of vowel-
harmony to the vowel of the preceding verb. After a noun
the objective or defining form is re, the same as in the second
person.

Emphatic, Reflexive, and Reciprocal. The simplest emphatie
addition toa personal pronoun is the demonstrative na “this,”
as, iwe ti emt na, “book of me this one,” “ my own book.”
Greater emphasis is given by the substantive kpdkpa: emi
kpakpa “I myself, my very self.” More common is ara
- (=“body,” as the Rabbinic esem “ bone’’), which is treated
as a noun and followed by the shorter, defining form of the
‘personal pronoun: 0 fe ara re, “he loves body of him” =
‘*he loves himself’’; it is also reciprocal: nwon fe ara won
‘they love one another” (literally “body of them’’). Out

82 C. H. Toy,

of ati “and” eki *‘ only” and ara is formed tikara, which is
used as subject emphatic pronoun: as, on tekara re di ara re,
‘he himself binds himself.”’

Demonstratives. The simplest forms are y2, na, ri, ‘ this,
that,” with the plurals wony: “ these,” wonni “ those ’’ (com-
pounded of the won found in the third personal pronoun) ;
they follow the nouns they qualify. From yi are formed eyz,
eyiy2, eyint, alayi, eleyt “‘ this,’ with plurals awonyt, wwony?,
nwonyi, and from na comes onna “that,” plurals awoni,
awonna; these are used as independent substantives. Na
and ni, especially the latter, have the force of the definite
article. | |

The Relative ti is without variation of gender, number, or
case, and (as in Hebrew) a personal pronoun is often intro-
duced for the purpose of defining the subject of the relative
clause, as emi tt mo mo “ I-who-I-know,” “I who know” ;
or, the indefinite o (or 0) is employed, as 0 st damu awon Ju
ti o wa ni Damaskus “ he-and-confounded-them-Jew-who-he-
lived-in-Damascus,” “the Jews that lived in Damdscus” ;
this o (properly third person singular) is used for all persons
and numbers. In like manner if the relative stand in a
defining relation to a following noun, this relation is expressed
by a personal pronoun, as osonri ti omo re de “ man-who-
son-of him-came’”’ = “the man whose son came.” As ante-
cedents to the relative are employed enz * one,” and eyt “‘ this,”’
or sometimes nouns.

Interrogative and indefinite. First the demonstrative ta
(probably connected with the relative ¢) is frequent as inter-
rogative pronoun, usually having the demonstrative nz attached
to it: wo ta ni Oluwa “ thou-who-Lord,” “ who art thou,
Lord?” After a transitive verb it introduces the dependent
or indirect interrogative clause, as emi mo tani o lu “ I-know-
what-he-struck.”” Tani may follow a noun as a defining term,
tle tant “house of whom.” A further illustration of the
demonstrative origin of the interrogatives is found in the fact
that t2 is used in questions as = *‘ what?” alone or preceded
by the interrogative 62: 62 emi ti nse “ whether-l-what-am
doing ?’’ = “‘what am Idoing?”’ Here the 67 precedes as a

On the Yoruban Language. . 83

general interrogative sign (as Latin an), then follows the per-
sonal pronoun as subject, then the question-word proper, and .
then the verb. Other interrogatives are ki and wo variously
combined (in origin demonstrative), the former inserted be-
tween the parts of a reduplicated noun giving an indefinite
sense. . .

C. The Verb. The Yoruban verb has no form-distinctions :
for all persons, genders, numbers it remains the same, but
this lack is readily supplied (as in English) by the use of the
personal pronouns. In respect to modifications of the idea of
the stem, the language has pursued an entirely different path
from that taken by the southern families of dialects, the Hot-
tentot and Bantu, having developed no system of derived
stems (Causals, etc.,) such as is found in them. Yet, while
it maintains its isolating character, its words standing sharply
apart in almost Chinese separateness, it has means of express-
ing the ordinary temporal modifications of the verb-idea with
sufficient distinctness by the insertion (prefixing) of words
that have almost lost their independent signification and may
thus be called half-inflections; it even makes an approach to
modal expression, and has a number of agglutinations corre-
sponding to the English ‘“‘ may, can, ought,” etc., expressing
permission, ability, obligation, desire, etc., these auxiliary
words retaining their full force, yet idiomatically sinking into
appurtenances of the main verbs to which they belong. The
following are the principal forms employed in this auxiliary
way. First, the conditions of completedness and incompleted-
ness are distinguished, the former being marked by-tz, the latter
by the sharp nasal » (ng); ti is naturally the sign of past
time (in which we naturally think of actions as complete),
but also of finished action in present time (our perfect); » is
used of present or past, and may be prefixed tot. As to the
origin of these forms n is probably connected with the sub-
stantive verb nz to be mentioned below, and ¢é with a root ti
or ta meaning ‘* Gnish’’ (see the verb ta7 = ‘to be finished,”’
with which also may be brot into connection to “to be suffi-
cient, attain to”’). Zi may be used with any time-combina-
tions to indicate that an action is past with reference to any

84 | C. H. Toy,

other. In some idiomatic uses it seems to stand as an inde-
- pendent verb (not a time-auxiliary): mkpa ise owo ti wah
‘“ by-labor-money-finish-come ’’ == ‘‘ by labor money comes” ;
on ko le ti so eso * it-not-able-finish-bear-fruit,’’ == “it cannot
hear fruit."—Future time is expressed by 0, the origin of
which is doubtful; it seems probable that it is connected with
the third singular personal pronoun 0, in which case emo ni
‘‘T shall have ”’ is literally, ‘“ I-he-have,” or, ‘I am he who is
to have.’ Often it is preceded by the demonstrative y: (com-
pare the similar combination tz 0 above mentioned): ow yio
ton ona mi se * he-will-again-way-my-make’’ = “he will pre-
pare my way’’; in on yio se the yi and o act as demonstrative
and relative: *‘ he-that one-who-make.” In certain cases a
is used instead of 0. There is another word ma which seems
to mean intensity and repeatedness, and is used to express
habitual or continued action: on yio tt ma bo * he will have
been coming ”’ ; it also expresses desire in the first and third
persons, and permission in the second, and this use, perhaps,
points to a different root (the word occurs also as a negative).
With all these forms may be employed the substantive verb
ni, which gives fullness to the expression : emi ni ri “‘ l-am-
see’ = ‘*] am occupied with seeing.”

Of modal forms there are no very clear examples. Certain
words are used in combinations out of which modal ideas
naturally arise ; but it does not appear that these are connected
with these words except in a very general way. Such a word
is ba (= “ reach, attain’), which occurs in conditional sen-
tences, as:°6t zwo ba rit kpa a “ if-thou-reach-see-it,-kill it”
= “if thou see it, kill it’; here the sense of uncertainty is
involved in the whole sentence, and does not seem to connect
itself particularly with ba, which also is, in other cases, used
in @ pure indicative sense. A similar remark may be made
of the dependent sentences introduced by the particle Ai _
‘‘ that’: the modal sense comes out from the general struc-
ture of the sentence, and is independent of the ki. A pecu-
liarity of this construction is that when the nominative begins
with a consonant or consists of two or more syllables, the ki
is repeated (for the sake of clearness or emphasis), and may

On the Yoruban Language. 35

be when the pronoun is on: ki on ki ole imu u “ that he should
seize them.” Je ‘can, be able,” gbodo“ dare,’ and some
others are used as independent verbs, without modal force.

Gerund. The above construction of dependent sentences
with ki (introducing substantive, telic, and other clauses) is
employed when the subject of the dependent clause is different
from that of the principal. When the two subjects are the
same, a gerund is used, made by prefixing 2 to the verb, as
wo “seeing (videndum) from wo “to see”: mo wa iwo nyin
‘« T-come-seeing-you””’ = “‘I come to see you.” Sometimes
this form occurs in independent sentences, and then appears
to be an emphatic assertion of the act instead of the ordinary
verb-root: kt (= ko) ise awodi “ not-the being-a hawk” =
“it is not a hawk.’ Along with this may be mentioned the
forms aba and iba (from ba “‘ to meet ’’), expressing obligation :
emi aba (or, iba) se e “‘ I-the being bound-do-it ’’ = ‘* I ought to
do it.” 16a is also used for ‘if,’ and tba... iba = “ whether
(either) ... or’: tba ise okonri, iba ise obiri “‘ whether men
or women, ’ literally, ‘‘ coming on (supposing )-the being-men-
supposing-the being-women.” Besides this abstract noun
with prefix 2, others formed by prefixes a and atz are similarly
employed. °

Passive. The passive is not made from the reflexive, nor by
the addition of a modifying root, but by a simple use of the
ordinary verb or noun. Most commonly the active with the
indefinite subject a “they” is employed: @ riz “he is seen”’
(literally, “they see him”). Or, a gerundal construction is
used (the abstract noun of action): ile se imz, “the earth-is-as
to shaking ’’ = “the earth is shaken.”’ In the same way may
be employed (with nz) the reduplicated nouns made by doubling
the first syllable of a transitive verb, as rivt from ri “to see”’:
rirt li (=nt) emi; “as to seeing-am-I”’—‘“I am seen.”
Finally, the compound transitive verbs may be used as pas-
sives: thus, from ba... dze (=“ meet... spoil’’) ‘‘ consume,”’
we have: iwe ba-dze “book consume’’ = “the book is con-
sumed.”
_ Participle. Our participles may be rendered by the continu-
ous form made by prefixing x to the root, or by independent
clauses.

86 C. H. Toy,

Substantive Verb. The Yqruban shows a profusion of sub-
stantive verbs, such as is natural in a primitive, unliterary
language ; the different forms are however distinguished by
usage with some clearness. The most common form, and the
one that approaches nearest the simple copula is nt (before
vowels lz), the origin of which is not clear; the only word that
offers a probable explanation of it is nz ‘“‘ to have,” from which
the substantive force may have come somewhat as in French
iya. Nis frequently used as =“ it is,” and appears often
where its only effect seems to be greater fullness of expression,
tho it has of course fixed itself in various idiomatic phrases ;
there is sometimes a heaping up of auxiliary words, such as
we find in Aramaic or in some French phrases. Examples of
its use are: emi ni ri, “ I-am-see”’== “1 see;” awa lio ae e
“‘ we-are-that-did-it”’ = “‘ we did it’’; swe nz yt o ri, * thou-art-
he-shall-see ’’ = ‘‘ thou shalt see.”’ Of verbs expressing exist.
ence proper, there are mbe, gbe, wa. The first of these (from
bi “to beget’’) is used for absolute existence, =“ exists,” the
second (also from 62) merely takes the place of mbe in the
imperative and in certain dependent sentences, the third,
(meaning “to dwell’’) is used of existence ina place. Modal
existence is expressed by ri, which seems to be connected with
the verb ri “to see”’: it is used with such modal words as
behe “thus,” 6: “as.” Se (=“to do”) denotes the occu-
pying a position, which calls for exertion; thus, oba li (= ni)
on means “he is (is described, known as) a king,”’ but on se
oba, “he fills the station of king’; so also nearly dze. To
these may be added 82, expressing existence in a place, and
used chiefly in negative sentences, and dz = ‘‘ become.”

SYNOPSIS OF THE VERB 1% ‘‘ TO 8EE.”’

Iseeorsaw . ; , emt ri or emi ne rt.

I saw or have seen . ; emt ti ri or emi lio tt ri.

I am or was seeing . ; emt nri.

I have or had been seeing emi tt nri or emi nui nri.

I shall or will see . ; emi o (or @) ri or emi nio (or ni yt 0) r2.
I shall or will have seen emio ti rt or emi ni yi o ti re.

I may or would see . emt ma Ti.

I might or would have seen emi ma t@ rt.

To these might be added potential forms with le, conditional

On the Yoruban Language. 37

forms with ba (introduced by 07), and others that do not prop-
erly belong to the synopsis. - The above shows sufficiently the
isolating method of the language, and the fullness of its verbal
forms. .

D. Partigles. Adverbs are derived from verbs and nouns.
They follow the words they qualify except those expressing
negation, those formed from fi“ with” and nouns and some
others. The language abounds with adverbs of time and
place, and such as express some particular quality, and a
degree of some particular quality, but has failed to abstract
the conception of degree, and to supply words for it; it has,
therefore, as is the case with many half-developed languages,
a host of locally descriptive words, and a paucity of general
terms.

Prepositions also come from verbs and nouns. The same
minute local descriptiveness is found among them as in the
adverbs. Local prepositions proper are susceptible of three
forms; one used when the sentence expresses rest, the second
when there is motion from the object to the subject, the third
when the motion is in the opposite direction.

Conjunctions. The language has several words for the sim-
ple copulative, and a good store of adversative, illative, con-
cessive, causal, telic, and temporal connectives.

II. SYNTAX.

-In conclusion, some more general observations on the syn-
tactical structure of the language may be added.

In the simple sentence the usual order is: subject, copula,
predicate; the attributive adjective (or pronoun) usually
follows its substantive; the substantive verb is usually inserted,
whether the simple ni (lz), or one of the forms that express
existence in a more definite way. The language is, indeed,
fond of the repetition of these words, and of pronouns, per-
sonal and demonstrative; in the absence of inflections it
resorts to this repetition to secure distinctness or emphasis,
and so constructs sentences that seem ungainly to us. It is
unfortunate to call this use of pronouns and other words
‘‘ nleonastic,”’ a term that, so far as it is accepted, shuts out

6

88 C. H. Toy.

the investigation of the peculiarities in the modes of expres-
sion of the language. In general, when a noun (whether
subject or object) is separated from its verb, clearness of
reference is gained by the insertion of a personal pronoun:
awon alagbase baba mi melomelo li o li ondze “ they-hireling-
father-my-how many-it is-he-have-food’’ = “how many hire-
lings of my father have food,” where the vividness gained by
the “it is”? and the general subject “he”’ is obvious. So the
insertion of the personal pronoun after the relative: “give
me the inheritance ”’ t 0 tort mi “ which-it-belong-me ”’ ; and,
after verbs of saying: 0 tenumo o kpe on ko ae e “ he-said-it-
namely-he-not-do-it’ = ‘he said that he did not do it.”” The
substantive verb is often employed to represent a subject or
clause, as in the first example above given, summing it up and
holding it separate before the mind, somewhat as in the Eng-
lish expression ‘ there is.’

The various parts of the composite sentence are commonly
regarded merely as standing to one another in the relation of
sequence (as in Hebrew). Dependent clauses, however, are
frequently introduced by appropriate conjunctions, expressing
relations of time, manner, purpose, cause, and the like. Con-
ditional clauses are introduced by 6:7 “ if” or iba (= “‘ obliga-
tion’), the verb da then standing regularly with the main
verb, and the shade of the idea (as to certainty or uncertainty)
is left to be inferred from the general connection. Our par-
ticipial constructions are expressed by a simple verb in an
independent clause, or by a separate clause introduced by a
conjunction, or by an abstract noun of action. The latter
plays an important part in the language, like that of the
infinitive and gerund in English. It occurs as simple subject
or object, and often expresses purpose; in many cases it
stands instead of the finite verb. It may then be introduced
by the preposition /: “‘in, in regard to” or not; if the prepo-
sition be absent, the noun is to be taken absolutely, as defin-
ing the verb simply by the expression of the idea.

IlI.—Influence of Accent in Latin Dactylic Hexameters.*

By M. W. HUMPHREYS,

PROFESSOR OF GREEK IN VANDERBILT UNIVERSITY.

In my paper on trimeters (Transactions Am. Phil. Assoc.,
1876), the investigation had much to do with the word-feet
were SS LS SSS Ye YS, — KH; the object
being to ascertain on what syllables the ictus could fall. In
hexameters we have nothing to do with this question, for in
the first three the ictus cannot fall on the accent, and in the
last two it cannot fall anywhere else. In words, on the other
hand, that are composed of long syllables, we are not to
expect so much influence of accent as was found in iambics,
for in the latter a so-called spondee partakes of the nature of
an iambus, having its thesis (i. e., &pe¢) shorter than its arsis
(Séorc), whilst in dactylics the two are equal, and the ictus is
more nearly uniform from beginning to end of the verse.
Further, when Ennius introduced hexameters, he imitated
Greek models and instituted a more artificial sort of verse, in
which quantity could not be so much neglected, and so did
not compose so much after the norm of popular usage.

Having examined the extant hexameters of Ennius, I find,
accordingly, that he entirely disregarded accent. This is evi-
dent from the fact that in the fifth and sixth feet, where the
usage of later poets shows it to be easy to place the ictus on
the accent, he has discord between ictus and accent as fre-
quently as it occurs in Homer or Hesiod, read with Latin
accent; and in the latter case the coincidence of ictus and
(Latin) accent is due entirely to the system of accentuation
and the structure of the verse. And, further, Ennius pre-
ferred the masculine caesura (with conflict between ictus and
accent) to feminine, where there is no such conflict.

* This paper contains the substance of a dissertation published in Latin
at Leipzig in 1874. The edition was so small that its reproduction in Eng-
lish seems justifiable. Asa review of the dissertation in Bursian’s Jahres
bericht for 1877 misrepresents it, great care has been taken to make no
changes of any importance, except such as were necessary in order to
reduce its size. '

40 M. W. Humphreys,

But in the fifth and sixth feet, the form of the verse gen-
erally causes ictus to fall on accent without any effort on the
part of the composer. Any one, by a little reflection, can see
why this is so, and that it is so,is shown by Greek verses
read with Latin accent. The caesurae, on the other hand,
cause conflict up to the fourth foot, where their relation is
variable. Hence we see that, whether the composer wills it
or not, there will generally be conflict in the earlier feet of
the verse, and coincidence in the last two,—a sort of strife
followed by a reconciliation. Consequently, in the course of
time, when the ear became accustomed to this, it appeared to
be a property of the verse, so that verses in which it did not
happen, seemed strange and harsh. Hence poets began to
seek conflict followed by coincidence, and the more they did
this, the more objectionable became verses in which it was
neglected. Accordingly, the poets of the Augustan age have
conflict in the first few feet more frequently than Ennius, and
in the last two feet more rarely ; nor can any one read hexa-
meters much without coming to feel that this peculiarity
renders the verse pleasing, and peculiarly so, when words
employed in one line are repeated in the next with the rela-
tion of ictus to accent varied. A beautiful example of this
is found in Catullus (LXII, 20-22) :

Hespere, qui coelo fertur crudelior ignis?

qui natdm possis compléru avellere mdtris,

cémplexé matris retinentem avellere ndtam.
(See also Virg. Bucol. VIII, 47-50 and in the poets generally.)
The frequency of this shows that it was purposely done by
the poets.

§ 2. So far the discussion has been general; but now I
proceed to examine the different authors, and shall begin with
the origin and trace out the development of the artificial
relation of ictus to accent, and shall briefly consider the argu-
ments of those who deny to the accent any influence.

In order to havea rule by which to measure the phenomena,
I shall examine Greek verses read with Latin accent; for by -
this means we come at what would have been the state of
affairs, had all been left to chance. The general discussion

' On Accent in Latin Dactylic Hexameters. 41

will be confined to the last two feet; and in ascertaining the
relation of ictus to (Latin) accent in Greek, and comparing
with the same in Latin, I confine myself to the fifth foot. In
Hom. Od. I, containing 444 verses, the ictus conflicts with
Latin accent in the fifth foot

in 15 verses where | ~ — | follows.

“32 CO “| yY~ — — follows.

“16 * ‘< something else follows.

But the word-feet ~ ~ and ~ ~ — — are more numerous in
Greek than Latin. Toshow this, since many verses of Ennius
are not entire, I employ 444 verses of Virg. Aen. I, and com-
pare with Od. I, containing 444 verses.

In Od. Il wehave~~Y- _.. ; . 176
we S . 101
In 444 of Aen. I, ~ — ; ; . 134
~rn-wS C=. . 21

IT omit words combined with -que into the form ~ ~ — ~ ,
as I should otherwise have to recognize Greek words of the
form — — — followed by monosyllabic enclitics. Now to
find what would be the relation of ictus to Latin accent in
Greek, if the forms ~~ and ~ ~ — ~ were not more
numerous, we reduce thus:

for ~ ~ , 176: 184::15: 114+

for ~ ~ — ~ , 101: 21:: 82: 7—.
Then 11 + 7 + 16 = 34, which would be the total number of
discords in the fifth foot. Now in the 541 verses of Ennius
there are 36 conflicts, which in 444 would be 31, which is
practically the same as in Homer (34). But I have also
examined Iliad III in the same way, and, omitting proper
names, I find the conflicts a little rarer than in Ennius. We
can affirm, therefore, that in the aggregate the verses of
Ennius do not differ in this respect from those of Homer,
from which it appears that Ennius paid no attention to accent.
In examining others, therefore, I shall compare all with En-
_ nius’s usage as ‘being acvidental. Conflict in the sixth foot
being occasioned by a final monosyllable, and in the fifth foot
by caesuraé in that foot, it will be important to note all
instances of monosyllables at the end and of caesura in the

42 M.;W. Humphreys,

fifth foot, even when elision or a preceding monosyllable pre-
vents conflict. I shall also record other peculiarities. When
a dissyllable becomes a monosyllable by elision, the fact will
be noted. Of course no special notice will be taken of verses
which exhibit no peculiarity. The annexed table* shows the
result of the examination of Ennius’s Fragments containing
541 verses, Lucretius III, 1,092 vv., Hor. Sat. I, 1,025 vwv.,
- Hor. Ep. I and part of II, 1,000 vv., and all the works of
Virgil, 12,869 verses with complete endings. Of course these
numbers must be taken into account in comparing the result
for the different authors. These are selected for the table as
representative poets, but others will be included in the dis-
cussion that follows. It is scarcely possible that all the fig-
ures should be exactly correct, but they are nearly enough so
for the present purpose.
1. For Ennius we collect the following result:
Caesurae after fifth arsis (1) with conflict 28
(2) without conflict (a) monosyl. 32
(b) elision (a) diasyl. 5

(B) polysyl. 5 42
In sixth foot, caesurae (1) with conflict 40
(2) without conflict (monosyl.) 5

In fifth foot conflicts without caesura (1) -que 4
| ' (2) otherwise 8 12
Spondees in fifth place (orovdeaZovrec) 13
Verses with both feet contained in one word 20
From this we see that in the sixth foot the conflict takes
place (in proportion to the number of caesurae) much more
_ frequently than in the fifth, and is more frequent, even, than
in Homer. MHere we are not to infer that he strove after con-
flict, but that he frequently imitated certain Homeric endings
(with caesurae in sixth foot) which especially pleased him ;
and this imitation he sometimes carried to an extreme, as
when, induced by vicg éucy dd, yadxoBaréc da, etc., he wrote
endo suam do, altisonum cael, laetificum gau, éetc.; and, in imi-
tation of Tmesis, “ saxo cere- comminuit -brum.’? Those who

*Sce page 43.

On Accent in Latin Dactylic Hexameters.

Lucr. Hor.8. Hor. BE. Virg.

Form. Ea.
aspen al SE LA 6 | 1 380
sso ftloTCv AS UN 1 3
sr tlolo zs 0 0 5
sr tloTU Als 2.06(COOUCOO
cso eleoTUlzls 0 0 5
se IN ts 4 0 0
lIf!loCIl4zs 138 24 78
I! lov es 13 17 8
I! JoU4i[s 1 7 14
Ito! 4le 0 0 0
It4[olo 4s 0 4 2
if|_e4Is 1 oO 0
IéloOLTIJ42Is 1 4 7
Ff fe 3 2 0
lIzloUl eats 0 o 1
I-~JOUlOC4Zs 0 0 3
Iz lOUlO[zAs 0 0 1
lv’ to Ll 4s 0 0 5
lp ts 3 18 8
level oo Leds 2 4 3
IV [ToLIlezls 0 0 1
so lo CU ts 4 10 1
—~/locvtzAls 1 1 = 1
—~A’|lllLI4zs 0 2 0
JoulLItLls 25 26 29
JoULtLI|s 7 2 14
[If 4ls 1 0 0
JovltAls 3 16 38
“olor ls 0 Oo O
JoloIl4zls 0 oO 7
AolJVIAls 1 2 0
Soot es 20 46 15
4 ts 5 D 0
fol szs 0 4 5
tA2oOlL=s 4 7 90
AoOf'|Ls 0 oO 1
JoOf’| 4 ls 0 oO O
L2’°lloLes 0 oO 1
'2’l|loltis 0 0 0
Aolofr_’ 0 oO 2
<n ee 0 oO O
ar tof Oo ww ee

14
1
1
1

S

ed

oo
CGCorcoonwoooaonrrrF QQ OF DMO OHNMReHF Co OCOore Ceo MoN Oo Mwy

14
29

~mwoe

43

Examples.
ignis mare ferrum
rubens hyacinthus
amatorem quod amici
stolidi soliti sunt
medicum roget ut te
purpureo narcisso
te quoque dignum
an Meliboei
rem facias rem
si qua tibi vis
aut quod ineptus
et magnis dis
iam data sit frux
sic compellat
aut etiam ipse haec
non ego avarum
stans pede in uno
saepe ego longos
mentem animumque
atque oculi sunt
quanti olus ac far
antestari ego vero
solidoque elephanto

‘scripsere alii rem

texere et in illam
adfixit habes qui
exiguus mus

sublatae sunt

obstitit et nox

iugera centum an
ridere decorum et
audivit at in se

isque pium ex se
Alcimedontis
incrementum
indicium illud
promissaque barba
Ephyreique aera
omniaque in se
servareque amicos
arvaque et urbis
caléremgvue | Inter, ete.
tétasque | Advolvere, etc.

| — | <= make up all the rest, that

is, Ennius 405, Lucretius 894, Hor. Sat. 738, Hor. Ep. 833, Virgil 12,372,
the per cent. of peculiar endings to the entire number being, Ennius 25,
Lucr. 18, Hor. Sat. 28, Hor. Ep. 17, Virgil 4?

44 M. W. Humphreys,

think that such a treatment of words was common at Rome,
~ are called upon to accept

‘* Massili- portabant iuvenes ad litora -tanas.”’

In the fifth foot the ratio of caesurae with conflict to those
without it ig about what we should expect (28 : 42) if it were
left to chance. In Hesiod I have found the ratio a little
larger, but this is due to causes already explained.

2. Now we come to Lucilius; but we need not wonder if
no great progress was made by a man who prided himself on
‘‘ standing on one foot and composing two hundred verses in
an hour.” Yet it will not be uninteresting to see how he
differs from Ennius, especially as he did not follow Greek
models so closely. The 925 verse-endings (in which the text
is often doubtful) show this result:

Caesurae after fifth arsis (1) with conflict 10

(2) without conflict (a) monosyl. 45
: (b) elision (a) dissyl. 8
(8) polysyl. 11 64
Caesurae in sixth foot (1) with conflict 35
(2) without conflict (a) monosyl. 22
(b) elision (a) dissyl. 1
. (B) polysyl. 4 27
Conflicts in fifth foot without caesurae, |

QyJ|trvr+ |= 10

(2)-—~-~-’|-T 5 15
Fifth spondees, | ——— ~,1,—-~’ | — +~,1 2
Both feet in one word 31

Comparing this with Ennius we see (1) that caesura in the
fifth foot is somewhat rarer, and that conflict is much rarer,
but that accent on short syllables is disregarded; (2) that
caesura in the sixth foot is not quite so frequent, and that a
considerably less proportion have conflict, though the conflict
still predominates; (8) that fifth spondees are much fewer ;
and (4) that both feet are something less frequently contained
in one word (for 925: 541 is greater than 81:25). Of
course no great importance is attached to slight differences,
but in the two main points (caesurae with conflict in the fifth
and in the sixth foot) the difference is considerable, the ratio

On Accent in Latin Dactylic Hexameters. 45

being, in the fifth 1 : 5, and in the sixth 1:2. But Lucilius
did not hesitate to place an ictus on a grave syllable, provided
there was another on the accent. Instances are rarer than
in Ennius merely because he did not indulge so much in high-
sounding compounds, such as “ altisonantes,” “ sapientipo-
tentes.”” But there are two pointa to be specially noted:
first, while conflicts grow rarer, the caesurae without conflict
also grow rare, but not in the same proportion; and secondly,
already in Lucilius, if the verse ends in an ionic word, | ~ ~
— ~, 4 polysyllable before it is carefully avoided (—— + |
~~ +); but if the ending is | ~ ~ | + —, the polysyl-
lable before it is not so rare as in later poets.
8. In Lucretius (II, containing 1,092 verses) we find :
Caesurae after fifth arsis (1) with conflict 2
(2) without conflict (a) monosyl. 58
(b) elision (a) dissyl. 17 |
() polysyl. 13 88

Caesurae in sixth foot (1) with conflict 40
(2) without conflict (a) monosyllables 22

(b) elision 0 22

Conflicts i in fifth without caesura bee |-~~-—|—2 |

2) —~~~|=S (aque) 7

(3) 2 —-~—’|— —» 34 18

Spondees in fifth place T

Both feet in one word 51

From which we see (1) that conflict in the fifth foot is
more carefully avoided than in Lucilius. These conflicts,
however, are more rare in Book III than in the rest, which
have four or five apiece. (2) In the sizth foot, however,
accent is disregarded as much as in Lucilius. (3) Words
containing both feet are not avoided, and accent on a short
syllable is disregarded, so that the form — “ WY WY suffers
elision four times in the fifth place. These words, however,
are all infinitives of verbs compounded with prepositions, and
are placed so that the ictus falls on the emphasized preposi-
tion, as “défluere hilum.’’ (This happens, also, in iambics.
See Transactions, 1876.) Similarly an unusual accent is
neglected ; for such forms as “ mutareque”’ receive the fifth

7

46 M. W. Humphreys,

ictus on the antepenult seven times, although the (artificial)
accent is on the penult. Not a few, however, deny that -que
creates this accent. (This subject is also discussed in Zrans-
actions, 1876, and comes up again in this paper.) Elision is
carefully avoided before a monosyllable at the end of a verse,
although it would prevent conflict. It was, no doubt, avoided
on account of its roughness. It is not rare in Ennius, and
Lucilius has some abominable instances of it, as ‘“‘ consciu’
sum mz; at”’—. In the fifth foot, however, Lucretius admits
elision in order to prevent conflict ; for before caesura in that
foot we find the form — = — | only twice, and —— —’ thir-
teen times. Even this elision was too harsh for Virgil and
his contemporaries.

4. Corssen, to establish his theory that accent was entirely
ignored, counted the conflicts in Lucr. IJ, where he found
sixteen (in the fifth foot), of which twelve were conflicts with
the unusual accent on a short penult caused by -que, as in
‘“arbastaque lénta”’; and from this he jumped to the con-
clusion that accent was entirely disregarded. If he had merely
asserted that the phenomena, whatever they might prove to
be, were due to other causes, there would have been no need
of making the count. Hence, as he made the count, he cer-
tainly meant to conclude from it whether accent was regarded
or not. What, then, did he demand? That all ictuses in those
. feet should fall on accents? By his reasoning we can prove
that Virgil did not avoid hiatus (cf. Aen. XII, 31, 535, 648,
etc.), and that he regarded final short syllables as common
(XII, 18, 68, 263, 550, 667, 772, 883, etc.). But it is use-
less to reply to such arguments. The very fact that conflict
is so much rarer in Lucretius than Ennius shows that this
rareness is not due entirely to accident.

In Horace Sat. I, Corssen finds eleven (11) conflicts in the
fifth foot and fifty-five (55) in the sixth and compares them
with those in Virgil to show that the strictness of the latter
was not due to his following popular usage, that is, due to his
observing accent. Of course not; but what was it due to?
_ Legibus aestheticis,” says Lucian Miller. Very good: but
what do they relate to? The truth is, Horace wrote his

On Accent in Latin Dactylic Hexameters. 47

Satires carelessly —much more so than his Epistles, and con-

sequently we find not only numerous conflicts of the sort, but

also frequent neglect of main caesura, and other licenses.
From Hor. Sat. I, containing 1,025 verses, we collect the

following :
Caesurae after fifth arsis (1) with conflict 43
(2) without conflict (a) monosyl. 114
(b) elision (a) dissyl. 14
(B) polysl. 3 131
Caesurae in sixth foot (1) with conflict 61
(2) without conflict (a) monosyl. 58
(b) elision (a) dissyl. 4
(B) polysyl. 0 62
Conflicts in fifth foot without caesura 21
‘“ “sixth “ by synaphea 2
Spondees in fifth place, absolutely 0

Whence it appears that in the Satires, Horace, in compari-
son with Ennius, guarded somewhat against conflict in the
fifth foot, and much more than Ennius in the sixth. The
number of caesurae without conflict in this foot is the same
as that with conflict, but if it were left to chance the number
with conflict would be as much more numerous than those
without, as there are more polysyllabic than monosyllabic
words. But if he avoided them at all, why did he make any
conflicts ? Simply because this is not an absolute law, and
there was no necessity to observe it strictly, and it would have
cost more labor than it was deemed worth.

Horace’s Epistles, being written more as monuments of
literary art, were more carefully composed than the Satires.
In Ep. I, and part of II, making 1,000 verses, we find:

Caesurae after fifth arsis (1) with conflict 19
(2) without conflict (a) monosyl. 70
(b) elision (a) dissyl. 0

(B) polysyl. 2 T2

Caesurae in sixth foot (1) with conflict 36
| (2) without conflict (a) monosyl. 49
(b) elision (a) dissyl. ‘1

(B) polysyl. 1 51

«48 M. W. Humphreys,

Conflict in fifth foot without caesura (1) -que,-ne 8
(2) otherwise 11 19
Both feet in one word 6
Spondees in fifth place 0
Here we see that in both feet the caesurae are rarer, and
the conflicts much rarer than in the Satires. The very fact
that the same author, under different circumstances, should
compose verses so different in this respect, shows that it is
not a result of accident or necessity. Horace does not em-
ploy elision to prevent conflict in the fifth foot. This becomes
evident from a comparison with Lucretius.
Hor. Sat. Epist. Lucr.
Conflict (-— + |-~-~—~—) 83 19 2
Elision (ST +’ | ~~ 4+ —) 3 2 13,
' which gives a ratio of 94: 1 against elision in Horace as com-
pared with Lucretius.
5. In all the works of Virgil (12,869 complete verses), I
find :
Caesurae after fifth arsis (1) with conflict 57
(2) without conflict (a) monosyl. 153
(b) elision (a) dissyl. 138
(8) polysyl. 17 1838
Caesurae in sixth foot (1) with conflict 47
(2) without conflict (a) monosyl. 48
(b) elision (a) dissyl. 2(?)
(3) polysyl. 0 50
Conflicts in fifth foot without caesura(1)-que,-ve 131
(2) otherwise 9 140
Spondees in fifth place 6-80
Both feet in one word, 30 proper names + 7 37
The verses of Virgil containing conflicts either end in
proper names or Greek words, or are composed in imitation
of Homer, Theocritus, or Ennius. The ending “ et magnis
dis,” ‘which occurs twice in Virgil, is found also in Eunius, and
not a single other instance even of the ending | — | — + | —
occurs in all the poets examined. The ending | ~~ ~ ~
eis frequently represented by “ hymenaeus,”’ before which a
short syllable is sometimes lengthened and hiatus occasionally
admitted. The same things occur with this word in Catullus.

On Accent in Latin Dactylic Hexameters. 49

The caesurae themselves, without conflict, are rare in Virgil.
Langen says they would not be rare in the fifth foot if they
were allowed after polysyllabic words of each metrical form
as frequently as they are after monosyllables. But then
polysyllabic words of each form are not so numerous any-
where as monosyllables ; and, moreover, earlier poets have the
caesura more frequently even after monosyllables. Where-
fore we must concede that Virgil avoided the caesurae them-
selves to some extent, but the conflicts still more. Of this
presently. |

If enclitics cause accent to fall on a short ultima, as “ pro-
miss&que (a question discussed in Transactions, 1876), it is
clear that Virgil disregarded such accents as being unusual,
or rather, artificial; for in such cases the conflict is more fre-
quent than in any other. It is true, ictus cannot fall ona
short syllable, and such words compelled conflict. But the
word-foot — — — ~— suffers elision seventeen (17) times in
the fifth place, causing conflict, and of these instances siz-
teen have -que, as “‘omniaque in se,” the only other being
“‘¢ntremere 6mnes,’’ which is like the examples in Lucretius
(prepositions in composition receiving ictus): That such
words have to suffer elision proves nothing, for the same is
true of words of this form without an enclitic. Nor is it
necessary to assume that the original accent remained, as
“émniaque”’ (like rfara ye); but the true explanation seems
to be this. First, the roughness of elision at that place
(as we shall see hereafter) was avoided; but -que, -ve, and -ne
suffer total elision (while other words do not), and so cause no
roughness. Secondly,in these forms the ictus does not, fall on
a syllable adjacent to the accent, while in other cases it does
(cf. omniaque = + — (~) and colligere — —“ ~(~)).
And might.it not be, after all, that a sort of secondary accent
did remain on the original tone syllable ?

6. The difference between Virgil and Horace in respect
to conflict ia due to the fact that their works are of different
natures. MHorace’s Satires were written to effect sométhing
at the time,—were practical and oljective ; whilst the works.
of Virgil were designed to be permanent literary monuments,

50 M. W. Humphreys,

or works of art. No oneexpects a dépét-building to be like a
memorial hall. The one is a means or an instrument to
accomplish an end; the other is its own end. The one is
useful, the other ornamental. But the success of an instru-
ment may make it as great an object of admiration as a
monument, and the useful may also be ornamental. The
fact, however, of a work being in verse at all, makes it toa
certain extent a work of art. Lucretius is didactic, but he
could have taught better in prose. Hence his writing verse
at all required that he should make his verse at least endura-
ble, and if possible, attractive. Horace’s Satires were attacks
upon the follies of men, but had to be made readable. His
Epistles had a less definite immediate object, were more
nearly a pure work of art, and so had to be more readable.
The Georgics of Virgil are somewhat didactic, but more
monumental. His Bucolics (which have a lyric tone) and
his Aeneid are purely monumental.

The Elegy also is artificial or monumental. Hence in the
Elegies of Catullus conflicts are rare, in Tibullus and Proper-
tius still rarer, and in Ovid they almost vanish. As far as
Elegies seem to be practical (e. g., the love-poems of Tibullus
and Propertius), their effectiveness depended in great measure
on their perfection as works of art. Besides, they were writ-
ten to be published, and in the case of Propertius the real
name (Hostia) cannot be substituted for the fictitious (Cyn-
thia) without creating frequent hiatus and false quantity.
Hence he probably wrote only for publication.

(a) In the 3823 hexameters of Catullus’s Elegies there
are:

Caesurae after fifth arsis (1) with conflict, 5
(2) without conflict, 18

Caesurae in sixth foot (1) with conflict, 6
(2) without conflict, «= 7

One conflict (cxv, 5) is caused by synaphea.

In his Heroic poem (Epithal. Pel.) Catullus, though perhaps
the best of Roman poets, allowed himself to imitate Alexan-
drian models, and admitted many Greek peculiarities, such
as fifth spondees (See Cic. Att. vu, 2). In that poem there

On Accent in Latin Dactylic Hexameters. ‘51

are eight conflicts in the fifth foot, always caused by ‘“‘ hyme-
naeus,” or some other Greek word, or a proper name. In
the sixth foot there is one discord with caesura, and one with
synaphea. But there are only three monosyllables before
fifth caesura, and none before sixth.

(b) In the 855 hexameters of Tibullus I find :

Caesurae after fifth arsis (1) with conflict, 6
(2) without conflict, 17

Caesurae i in sixth foot (1) with conflict, 0
(2) without conflict, 3

In Books 11 and IV, which were composed with more art than
inspiration by some late versifier, I find no discord at all. In
the 211 verses of ‘“‘ Messala,” a silly Heroic poem by an
unknown stupid poetaster, there are six conflicts in the fifth
foot, and only three caesurae with a monosyllable.

(c) In the 1,572 hexameters of Propertius, there are:

Caesurae after fifth arsis (1) with conflict, 4
(2) without conflict, 18

Caesurae in sixth foot (1) with conflict, 1
(2) without conflict, 27

Conflicts in fifth foot caused by -que, 2
Spondees in fifth place, i)

(d) In 500 hexameters of Ovid’s Heroides, and the same

number from the Metamorphoses, I find:
Elegiac. Heroic.
Caesurae after fifth arsis (1) with conflict, 0

(2) without conflict, 5
Caesurae in sixth foot (1) with conflict, 0
(2) without conflict, 1

Conflicts caused by -que, 5

Spondees in fifth place, 1
From all this it is evident that in this artificial sort of poetry
great care was used. Propertius does not appear to have
avoided caesura itself, without conflict, as much as the other
Elegists. .

7. In order to give a better comparative view of the poets
examined, I have reduced the more essential points to a
uniform scale of 1,000 verses: |

CO fF N © Se

52 M. W. Humphreys,

. | ms i
ala|& a/R Fis].
Blalel@)/8) oiela - Sia
e318 8) es) Pile) 42/3).
jR lA Al Mi/ Mie; olo |e E ©
Caesurae after fifth arsis.
(1) With conflict, .. . |62110) 4) 42;19) 4) 2/15!) 7} 8] 0
(2) Without conflict, . . « | 27) 661 80 1128) 72);14; 2) 54) 21; 12/10
Caesurae in sixth foot.
(1) With conflict, . . . . | 74} 86/36) 58) 36; 4; 0/18; 0| O| 0
(2) Without conflict, . . .| 9/25/20); 59] 51 un 4/21} 4/18; 2

The number (4) for conflict in fifth foot in Lucretius is not
taken from Book 111, but from the reading of several books.
The result for Ovid is from too small a number of verses
to show anything more than that the conflicts and caesurae
are extremely rare. The average for all his works would no
doubt be different. Thus, in Virg. Aen. 11 there are no
conflicts in the fifth foot, whereas the average for all. his
works is four in 1,000 verses. This shows also that where
things are very rare, mere accident may affect them consid-
erably. I should say, therefore, that there is no appreciable
difference between Ovid’s Met. and his Elegies in respect to
conflict, but that both of them differ widely from Horace’s
Satires.

Virgil admitted the fifth and sixth caesurae, whether with
or without conflict, much more rarely than his predecessors ;
and he carefully avoided an ionic word-foot at the end pre-
ceded by a monosyllable, | » |~- +=. This ending
occurs only five (5) times, and in each instance the last word
is a proper name, while -~ / | ~~ 4. = with conflict is
much more frequent. Hermann (Elem. p. 344, Epit. § 322)
says that the cause lay, not in the last word, but in the
preceding one, it being unpleasant to have ictus on the unac-
cented ultima. But then, why the still greater aversion to
a monosyllable in that position? Hermann thinks that the
effort to secure coincidence in the last two verses was because
the lungs were exhaustsd, and so a smooth ending desirable.
One might say thatewhen the lungs are exhausted, the ictus
must be weaker, and so the conflict-would be less objection-
able, and in support of this, the iambic trimeter might be

On Accent in Latin Dactylic Hexameters. 58

cited. The true explanation of the above-mentioned phe-
nomenon seems to be this. In the first place, for reasons
already stated, conflict in the fifth foot was to be avoided,
and this could be done, if there was caesura, by placing a
monosyllable before it; and then this monosyllable forbade
a long word being placed after it. Virgil preferred even
conflict with two polysyllables together, to harmony with a
monosyllable and polysyllable combined. This is perfectly
evident from the following exhibit:

a | ees 41,| +] ~-+ a 5 (proper names).
ST t|orTvl[+co 4,]/+)]-~~-|+— 180

That is, the tendency to use polysyllables with long words
and monosyllables with short ones, as compared with the
converse is as seventy-six to one, and that, too, in spite of the
tendency to avoid conflict. Similarly in

Horace Sat. I, we find|“.|J~-~.— 8,|~|-~VY|+< 78

andin Ep. I,j7/60- 425 2,[+4|/06-|/4£— 49
Why some poets found |“ |~— + = more unpleasant
than other poets did, it would be idle to inquire.

§ 8. I shall now make a few observations on special points.

1. Spondaic verses generally end in a proper name of the
form __. “+. The ending _ — | + —, although the
coincidence of ictus with accent is perfect, was not employed.
This is because there was rarely occasion to use a double
name, as ‘Gaius Gracchus.” If, again, the name consists of
three long syllables, it creates conflict in the fifth foot, and if
it is two long syllables it can be put in the sixth place.
Hence only / __ “= is left. . Of course exceptions, such
as |“ | _— <, occur.

2. A word'of special importance is frequently reserved for
the end of the verse. If such a word is ‘monosyllabic it
naturally creates conflict; that is, in my opinion the poets
placed such words at the end, not. always because they were
monosyllables, but frequently although they were monosyl-
lables; as

parturiunt montes, nascetur ridiculus mus.
This verse, 1 am aware, is often cited to illustrate surprise

8

o4 M. W. Humphreys,

caused by a final monosyllable. That the monosyllable is
sometimes so employed I do not deny, but in this case the
sense of “ridiculus” prevents surprise in “mus.” If the
verse were |

parturiunt montes, nascetur magnijficus mus,

there might be surprise, though in this instance irony would
be suspected as soon as the adjective was read. In order to
create surprise we should have to read the verse

parturiunt montes, nascetur—ridiculus mus.

And similarly in “ procubuit viridique in litore conspicitur—
sus,” there is no more surprise than there is in

‘quantum lenta solent inter viburna—cupresst.”

But if a surprise is to be caused, the monosyllable is well
adapted to this place; for every one feels that, being an
emphatic word and having accent, it is not to be read like an
ordinary thesis, and consequently it gives the verse a novel
ending ; as

dat latus, insequitur cumulo praeruptus aquaé méns.

In such cases I suspect that the Romans unconsciously made
an entire foot of the sixth arsis, and placed a seventh ictus
on the monosyllable, thus: aquae mons = _ —. ,. This
would be a heptameter; but the ancients made a similar
blunder in regard to the so-called pentameter (which is a
hexameter). .

3. When two verses have too close a connection to admit a
pause between them, a monosyllable is frequently placed at
the end to prevent the voice from falling and destroying the
sense. Hermann (Elem. p. 842) teaches that when there is
a pause near the end, this monosyllable causes a sufficient
prolongation of what follows the pause, to make it “ compara-
ble’? to what precedes, as

at Boreae de parte trucis cum fulminat, é cém, etc.
But who will prolong ‘“‘et cum” so that ‘“‘apte ad praecedentia

comparari possit”’? It seems to me that the accent on “cum”
(or if it has none, then its proclitic character) prevents a

On Accent in Latin Dactylic Hexameters. 55

cadence of the voice which would mar the sense; and this
close connection with the next verse is most likely to exist
just when there is a pause near the end, but when the close.
connection does exist without such pause, the monosyllable is
still employed, as,
his me consolor victurum suavius 4c sf
quaestor avos pater atque meus patruusque fuisset.

For such instances Hermann’s explanation is unavailing, while
the explanation just given accounts for all alike. Horace in
his Satires very often closely connects two verses in this way,
whether there is a pause near the end of the first, or not; as,
Book I, Sat. I, 17, 46, 50, 56, 69, 81, 82, 96, 101, etc. This
is one thing which contributes materially to the large number
of sixth caesurae in Horace. Lucian Miiller (De Met. Hor.
p. 61), speaking of monosyllabic prepositions and conyunctions
at the ends of verses, says: “mitigatur haec inelegantia
addita, quod saepius fit, elisione.” In my original disserta-
tion I criticized his statement as referring to all monosyllables,
and so far did him unintentional injustice. His remark applied
also to main caesura after such monosyllables, and in that it
is strictly correct ; but there is so great an aversion to elision
in the sixth foot that such elisions as those mentioned by
Miller, as,

naturae fines viventi, iugera centum an, etc.,

are not at all frequent. I have counted such cases as this
and “ porro et,’’ and find that no greater proportion of such
monosyllables, when final, have elision, than when found else-
where in the verse; and as to final monosyllables generally,
the comparison is so striking that I give it:

In all Virgil, without elision, 48, with elision, 2 (atqu’).

“* Hor. Sat. I, “ “c 658, SS “U4

6¢ Tt Kpist. I, 6¢ cs 49, ce 6é . ]

§ 4. I now proceed to reply briefly to the arguments of
those who deny that accent has any influence.

1. As conflict grew rarer in the sixth foot, the caesura even
without conflict also grew rarer, but not to the same degree.
‘“‘ Why, though, did it grow rarer at all, if the offense lay in the

56 M. W. Humphreys,

conflict ?”’? Weanswer: Because there was a sort of conflict ;
for the monosyllable at the end has an accent which interferes
_ with the cadence, which is objectionable unless the two verses
are so closely connected as to make it desirable (See also end
of 3 below).

2. ‘“‘ Conflict with caesura in the sixth foot might be prevented
by elision, as in ‘‘decérum et,’ and yet this elision is very
rare, occurring only twice in Virgil.” To this we reply that
elision in this place gave the verse so rough a termination that
the offense was greater than that of conflict. This is shown
by the fact that of the fifty (50) endings of the form | ~ | =
in Virgil, only two have elision (‘‘atque”’ each time, where the
elision is total and hence not unpleasant), and in these fifty
cases there can be no question of conflict at all. But if the
elision was objectionable between two monosyllables, how much
more so between a long word and a monosyllable.

8. Lucian Miller attributes the rareness of fifth caesura
after polysyllables to “ esthetic laws,’ whatever they may be.
But be they what they may, if they are,laws, they must refer
to something, and this ‘‘something’’ I take to be the relation
of ictus to accent. I am willing to admit that the objection
was to ictus on a weak ultima, but it is weak because of its
relation to accent. Moreover, the unpleasantness was merely
a result of contrary usage, and that usage excluded fifth ictus
from the ultima just as much as it included coincidence of
ictus with accent.

“ But when the ictus falls on a monosyllable there is no conflict,
and still this is rare in the most careful writers.” This objec-
tion is not exactly true. There zs a species of conflict when
the monosyllable is followed by another, or by “ WY, thus,

| _|+~Y~| + <, as the ictus is immediately followed by
an accent (the case when it is followed by a long word,
| 2. | ~~ + &, has been already explained). The ending
|. | ~~ | S creates conflict also in the sixth foot. And
further, in avoiding conflict the poets no doubt would uncon-
sciously avoid that which causes conflict—namely, caesura.

4. “ But the monosyllable under the fifth ictus sometimes
is @ proclitic,and so has no accent.”’ This is true. Out of

On Accent in Latin Dactylic Hexameters. 57

the hundred and thirty (130) verses in Virgil which end in
|t~-|042| +2, six (6) have prepositions in the fifth arsis,
as ‘“‘ ab Jove summo,”’ to say nothing of the seven (7) which
have “non,” and the eight (8) which have the relative
pronoun, and the many others which have monosyllables
regarded as proclitics by the grammarians. But as we have
‘better means of discussing the prepositions, I shall confine
myself to them, and the conclusion will apply to other words.
I cannot agree with those who, in their eagerness to maintain
the influence of accent, deny that prepositions were proclitic.
Quintillian says CI, 5, 25) that in ‘circum litora” and in “ab
oris”’ there is but one accent. He cannot mean ictus, for there
are two on “circum litora” (and, by the way, his statement
proves that the ancients observed both ictus and accent in
reading —if it needed any proof). This being our best
authority, I need not cite others. Among other evidences,
however, I may mention that in inscriptions of all periods of
antiquity, prepositions are found joined into one word with
their objects (Corss. 1, p. 863, etc.). So “ antidhac,’’ “ pos-
tidhac”’ are results of the proclitic nature of prepositions
before they lost final -d. The analogy of Greek prepositions
Cf that is of any value) supports this view. Some of them,
év, éx, etc., are confessedly proclitic, and inscriptions show close
union with their objects by euphonic modifications in the case
of other prepositions, and when they lose their vowel, no
accent is written; so that the written accent (as in rapa
rovrovc) must have been very slight, though zn other parts of
speech Dion. Hal. (De Comp. Verb. C. x1) implies that it was
a “rao déeia.” In view of all this I must assume that prepo-
sitions had noaccent. Ihave also made a careful computation,
and find that monosyllabic prepositions are placed under the
fifth arsis about as often, in proportion to their entire number,
as words are which have an accent. But after all, this is not
surprising. The monosyllable, though unaccented, prevents
the ictus from falling on a weak ultima; and though it has
not the musical elevation belonging to accent, it has the stress
belonging to the ictus, as it is an independent word ; so that
the difference between “ab Jove”’ and “‘armaque’”’ is not great,

58 M. W. Humphreys.

the latter having the root-stress on “ar-’’. The same explana-
tion applies when a trochaic proclitic receives the fifth ictus, as
‘“‘propter eundem,” “unde Latinum.”’ But this union of a
proclitic with the next word is not entirely so close as that
between two syllables of the same word, as is shown by such
endings as “inter eundem,” which are numerous, while by
the more elegant poets a single word like “ ingemuerunt”’ is
avoided in this position. Moreover, the caesura may fall
between a preposition and its object, as: |

et inde tot per ] impotentia freta;

unless we try to believe that in this iambic trimeter alone
Catullus neglected caesura. Other examples occur in other
poets. |

Here I close. It would be impossible to sum up all the
conclusions in a brief space. I wish merely to repeat and
emphasize the statement that the influence of accent in dactylic
hexameters was a result of usage, and not of an original aver-
sion to conflict between it and ictus; and that this very
conflict, which got to be unpleasant in the last two feet, was
quite agreeable in the first few feet of the verse.

IV.— Observations on Plato’s Cratylus.

By JULIUS SACHS, Pux.D.,

NEW YORK CITY.

The student of the science of language who wishes to take
a comprehensive view of the theories advanced regarding it,
cannot fail to take cognizance of Plato’s writings as the ear-
liest detailed embodiment of speculation and observation on
this subject. Not but that among the predecessors of Plato
and Socrates valuable suggestions on the nature of language
were offered, but they were isolated flashes across the field of
intellectual vision rather than systematic discussions; neither
Herakleitos nor Parmenides formulated. their inquiries in a
manner calculated to emphasize distinctly the difference be-
tween thought and speech. Strange though it may seem,
the Greek philosophers busied themselves considerably with
hypotheses on the origin of the reasoning faculties, before
they convinced themselves that the final results of such inves-
tigations must, of necessity, be futile, unless they attacked
the problem of the origin of language, since language was
the vehicle of reasoning, and thus the most essential charac-
teristic of human kind. Plato’s age was fully alive to this
inquiry, and in the Cratylus we have by no means a tentative
effort in this field of speculation, but a résumé of prevalent
theories which a master in the art of dialectics sifts, indorses,
modifies, or rejects. The very art of the writer, however, his
consummate use of the various devices of oratory, satire,
modest doubt, etc., have rendered a correct appreciation of his
position all the more difficult, as we lack almost completely
the evidence for the real opinions held by those philosophers
whose views he introduces as foils for his argument.

Hence various modern writers on comparative philology have
been able to interpret the position of Plato in consistency with
their favorite theories, and the Cratylus has been represented
as the precursor of those linguistic treatises that proclaim
the study of language a physical science as well as of those
that make it a historical science. One point we may lay down

60 J. Sacha,

even at this stage: Plato’s Cratylus, whatever its object or
tendency, cannot be disregarded in any discussion on the
science of language ; it forms the landmark around which the
speculations of the ancients on the subject may be grouped.
From Herder on through Schleiermacher, Ast, Steinhart,
Benfey, Miller, Whitney, Steinthal, Geiger, down to the most
recent expositor of these issues, Ludwig Noiré (Ursprung der
Sprache), all seek to establish their relation to the Platonic
dialogue ; nay, the last-named philosopher, whose estimation
of his own results is significantly presented in the sentence:
“Thus language must have arisen; it cannot have arisen
_ otherwise,” finds in Plato’s exposition the germs of most
advanced modern thought, as of Schopenhauer, and a series
of linguistic and philosophical discoveries that thenceforward
-became an heir-loom to all later speculative research. Now,
notwithstanding the discrepancy of opinion as to the ulterior
significance of the dialogue, it is a fair question, Are there not
a number of points, generally adopted by all commentators,
from which a consistent interpretation ought to be possible ?
A review of the various discussions on the Cratylus, casually
undertaken by me, has convinced me that opinions are still
almost hopelessly divergent on the problem proposed in the
dialogue, and that yet there have appeared two discussions
that merit a more thorough consideration than they have
received for their bearing upon the main issue; I| refer to
Benfey’s ‘“‘ Ueber die Aufgabe des Platonischen Dialogs Cra-
tylos,”’ and Dr. Herm. Schmidt’s “‘ Plato’s Cratylus,im Zusam-
menhange dargestellt.”” The reasons for this neglect seem to
me to constitute a special plea in their favor ;. neither of them
seeks to establish a relationship between the Cratylus and the
general system of Platonic philosophy. I urge this as a point
in their favor, for the much-vexed question of the Platonic
philosophy, with its numerous subsidiary issues, is too apt to
bias the judgment on the import of the single dialogue; and
it seems to me incompatible with the nature and purposes of
these dialogues, that they should all represent one and the
same line of thought, uninfluenced by the exigencies of a con-
versational exposition. Two circumstances that have, respec-

On Plato's Cratylus. 61

tively, been prejudicial to these essays in the eyes of the
German philological world will not influence our estimate of
them. Dr. Schmidt’s essay does not present a connected
theory of the meaning of the Cratylus, but analytically takes
up the various passages, and, disregarding the final result,
discusses fairly and acutely the interpretation which is pre-
sumably the best. Whilst Schmidt then has no special theory
to advance, Benfey, who dves look to the claims of the work
as a philosophic whole, too modestly pleads ignorance as a
metaphysician, and as an exponent of Platonic phraseology.
Here, then, has been found the vulnerable point by the spe-
cialist-critics ; and though it must be admitted that now and
then there occurs an impossible rendering of some minor
passage in the Greek, his sound qualities as a linguist more
than compensate for this deficiency. "

To those parts of Schmidt’s work that do not tend to eluci-
date the questions which Benfey has also treated, nothing
more than a passing notice can be given ; let it suffice that
many a passage, involving knotty, grammatical construction,
has been capitally set forth by Schmidt: On the main issues
of the dialogue, Plato’s opinion of the origin and formation
of language, the contributions of the two writers seem to me
specially valuable.

In this direction Benfey has developed in succinct argument
a point that is particularly timely just now, when other Ger-
man critics, like Schaarschmidt and Krohn, apply the crucial
test to every one of the dialogues, and attempt to deny the
Platonic origin of the majority. If Plato is not the author,
he argues, it would remain for Schaarschmidt to prove that
the dialogue is of much later origin, the product of a time,
when the study of language was more thoroughly developed,
say, the Aristotelian; and as this can never be done, the
inherent excellence of the treatise as the oldest comprehensive
work on the subject of linguistics remains unimpaired ; the
question of Plato’s authorship is, under all circumstances,
secondary to the internal consistency of the views expressed.
Let it not be supposed that the treatment of this question of
authenticity is a purely speculative one.- Schaarschmidt’s

Q

62 J. Sache,

criticisms on so-called inconsistencies in the Cratylus must
stand or fall, in several instances, with the accuracy of trans-
lation in a given passage. Thus, when he ascribes to the
author of the Cratylus the assertion that’ in a sentence each
word embodies a judgment upon an object, and that, if a
statement is false, every single word contained in it must also
be false, a careful study of the previous passage would have
led to a more rational conclusion. With Schaarschmidt,
many others err in trying to ascertain what they call ‘“ den
verhillten Sinn”; this license once granted, the way is open
to various mystifying interpretations, and the natural course
of reasoning may as well be abandoned. No more striking
instance of this warping of the logical faculties could be
found than Steinthal’s exposition of the object of this dialogue
in his “Geschichte der Sprachwiss. bei den Griechen und Roe-
mern.” ‘The first part of the dialogue, where Plato proves
that a name is the sound-complement of the fundamental
idea of the name (die Ausfiihrung der Idee des Namens im
Laute), and supports the view with the greatest sincerity
(mit seinem Herzblute),’’ all this serious exposition we are,
according to Steinthal, to regard as not serious, and in -the
famous second or etymological part whatever is sportive,
conceals under it the reverse of sportive observation, is, in
fact, exceedingly sober. Now, whither will such methods of
interpretation lead, if, without any clue in the writings before
us, such renderings are possible? But why are such tours de
force ascribed to Plato? Because, though anxious to establish
a science of etymology, he has so little confidence in the cor-
rectness of his derivations that he finds it safest to ridicule
them all, good, bad, and indifferent. Stranger still, however,
is it that these philosophical critics have generally failed -to
observe carefully the exact meaning of the technical terms
used ; and it is peculiarly meritorious that Benfey has estab-
lished these conceptions beyond a doubt.

The question whether Plato considered language to have
originated and developed dice or déoe, for which latter word
Evyixn is frequently used in the Cratylus, could not be
answered satisfactorily, so long as it was not definitely under-

On Plato’s Cratylus. 63

stood that £u-6j«n has varying technical and popular significa-
tions. Benfey has carefully discriminated its three respective
significations, as (1) “an arbitrary agreement, unlimited in
every respect, perfectly optional,’ (2) “the agreement or
accord of a number of persons, bound by natural ties,’’ and
(3) “such agreement as has become conventional,”’ and we
recognize the vast difference between the v»0i«n or accord of
society, by means of which the originally manifest meaning of
a word is retained, notwithstanding the changes and modifica-
tions in etymological value, and that arbitrary fuv@ij«n which
e.g. decides upon certain sound-combinations as proper desig-
nations of various numerals. Jowett recognizes the difficulty,
and in his latest edition renders it often by “ convention and
agreement.” Plato’s time is preéminently the period of tran-
sition to a special philosophical terminology, and works in
which this process of evolution is being perfected, require a
more faithful interpretation than others with a fixed technical
vocabulary. In deciding these questions, the aid of kindred
sciences is often very desirable, and that were an unworthy
sense of exclusiveness that would forego the information
likely to be attained from such a source. Not unconsciously,
however, is this evolution of terms brought about. Plato’s
tendency toward nice distinctions appears, for instance, from
a survey of the verbs he employs in the sense of “‘ to mean ”’ ;
and one cannot fail to notice with what consideration for the
requisite shade of meaning he employs voeiv, iyyetoOar, déyecv,
dvopalev, kadeiobat, elvat, BovrrAErIar, Sndovy, pnvuecy, onpaiveryr, arEk-
alev, pupeiobar, daivesBar axeixacpa, toxev. A similar definite
conception of Plato’s leading terms seems to me an absolute
necessity, where he himself has not made matters as plain as
in the instance just quoted ; ovoua and pyjyua are the veriest
by-words of the dialogue, and yet the translations given by
Schleiermacher, Steinhart-Miller, and others are ambiguous,
since they are confused by the later application of the word
by grammarians, with whom oévoua = noun, pijpa = verb.
That dvoua here means “ word’”’ in its wider sense, and not
the noun-forms merely, is of no slight importance in the con-
sideration of the main question, for, if we admit that the verbs

64 J. Sachs,

‘are also cvéuara (and this has, I believe, been unhesitatingly
conceded to Benfey), we are forced to admit that sfpara can
no longer be rendered, as all translators have done, by “ verb,”
that the phrase cvéuara cai pauara would be tautological, and
that pixa must indicate an intermediate stage between the
“word” and the ‘‘ sentence’’ in a logical ‘and grammatical
sense; the logical sense being differentiated from the gram-
mnatical in this fashion, that the same word may in turn serve
AS AN orope OF Piya, according as it is accepted as an appellation,
or conceived of as a condensation of a logical phrase. So Povd4
is the ovopa to fod} (shot) as paya and if /30A) can be analyzed
still farther, it becomes the ovoza to another pia. Benfey
contends, and not unfairly, that the later meaning of pijpa (=
verb) comes more naturally from this original application,
that the pia contains that part of the sentence which is
independently intelligible. . Not only is Plato’s usage of philo-
sophical terminology often the cause of mistaken conclusions,
but the instances are not infrequent where a modern investi-
gator will be oblivious of the development and growth of cer-
tain ideas since Plato's time. How else could a distinguished
scholar like Steinthal sneer at John Stuart Mill’s statement
that ‘‘ words are important for the comprehension of things,”
and identify this with Cratylus’s statement that “‘ a knowledge
of the names of things involves a knowledge of the things
themselves,” seeing that Cratylus refers to the original phys-
ical nature of words in which he presumes to find a genuine
reflection of the objects they refer to, whilst Mill has in mind
the logical meaning that has gradually developed out of a
word. Benfey and Schmidt, whilst cognizant of such princei-
ples as have here been stated, have proceeded to the solution
of other difficult questions by throwing upon the words
involved the light of comparative grammar.—A link in the
argument, so urges Schaarschmidt, is wanting in the cele-
brated passage (388 B.) where, after speaking of the fune-
tions of various instruments, the shuttle, the awl, etc., Socrates
recurs to the name as an instrument, and draws analogous
conclusions. Let us examine for a moment the text and
Jowett’s translation, which is no stronger here than any of

On Plato’s Cratylus. 65

the other versions. Socrates asks: xepxiZovrec b& ri dpwper; ov
THY KpOKNY Kal Tove oTipovac cuyKEexupevouc duaxpivouev; ‘*What do
we do, when we weave? Do we not separate or disengage
the warp from the woof?” and shortly afterward, dpyavy évre
T® Ovdpare Gvopacorrec ri wovodper; Hermogenes: ovx txw dAéyery.
And Socrates : "Ap ouy OwaaKxoper Te adAHAGUE; 6 Do we not teach
one another something ?’’ Now with such a translation there
ig an unwarranted transition from dtaxpivoper to deddoxopev. AN
analysis of the verb édacw shows, however, that in its primi-
tive root-form 6a we have the true signification of separation
which underlies even the forms éaiw “to burn” and daivupe
‘¢to entertain as guest,’ and it is in accord with the etymo-
logical character of- the whole dialogue that Socrates should
thus delicately make the logical transition. On the other hand,
I do not believe that it will be easy to find one word which
in the translation would carry the same suggestiveness with it,
and yet not transcend the scope of meaning, usually ascribed
to deddoner. Of the salient points in the dialogue which,
stripped of the dialeetic form, betoken a substantial knowledge
of certain principles, current now among students of compar-
ative grammar, Benfey has made an interesting list, and
without giving way to the enthusiasm usually connected with
such observations, has also dropped various claims that had
been previously made for Plato’s linguistic insight. Among
these prominent points I single out the following: “ that
word would be most correct which would contain completely
its etymological elements ;’’ again, ‘“ words are overlaid by
the addition or stripping off or twisting of letters for the sake
of euphony’’; ‘“‘onomatopoietic origin of words is to be disre-
garded almost completely.” With the acknowledgment of
Plato’s grammatical insight must be coupled, however, the
warning that whether in sport or ignorance, or from other
motives, the illustrations of these principles are in many cases
untrustworthy.

Have Benfey and Schmidt, you will probably ask, taken any
new position on the central question, that of the purpose of
the Cratylus? I may as well state that 1 look upon Benfey’s
judgment in this question as the most valuable recent contri-

66 J. Sachs,

bution to its solution. All preceding commentators, from Pro-
clus to the moderns, have assumed as Plato’s purpose the
treatment of the question, ‘““Has language, as it exists, come
into being guce or béoe ¢’’ and have, with an expenditure of con-
siderable ingenuity, maintained the one or other issue. What
curious methods of procedure were necessary to make Plato a
doctrinarian on either side of this question! That Socrates
is represented as finding fault with the views of both Cratylus
and Hermogenes, the typical expositors of the two opinions,
was undeniable. Now in the one of these critical analyses,
Socrates, so say Steinthal aud others, does not mean what he
says; he criticizes, and yet at heart supports a certain view.
Whence this knowledge of the attitude of Socrates? The
solution is simple; not from the work itself can such incon-
sistency be gathered, but from the desire of the modern
theorist to confirm his experiences from this ancient product,
of literature. Others, less metaphysical, find Plato’s indi-
vidual opinion in the golden mean between the opposing
views. But for this intervening opinion no statement can
be found in the Cratylus. On the contrary, the very sup-
porters of this theory confess, as Schleiermacher does, that
Plato’s language indicates that he cannot give satisfactory
account of his opinion; and thus, also, honest doubts as
to the cogency of his own opinions seem to have presented
themselves to Deuschle in his work “ Die Platonische Sprach-
philosophie”’ who confesses that to himself it is not clear,
how in the concrete application gic and Géor¢ can correspond
respectively to %@o0¢ (custom) and éuvOjcen (agreement). I
cannot understand why a point of primary significance has not
been urged as the final answer to these speculative fancies ;
that the language of Socrates, naturally interpreted, proves
him to be opposed to the views of both Cratylus and Hermo-
genes is indisputable. Again, if Socrates would wish us to
accept the reverse of what he says, the language with its facile
particles would afford unmistakable proofs of such intentions;
why, then, this vacillation instead of a frank confession of the
situation ?

Neither gice nor Gécee can language, as it exists, be proved

On Plato’s Cratylus. 67

correct; in other words, language, as actually used, neither
conforms in its origin and growth to the natural meaning of
words, nor to the agreement of mankind regarding them.
An ideal language only might be constructed conformably to
these principles; in it the veritable opOérnc dvoparwy would
have to be sought; whatever correctness of appellation actual
language shows forth, is purely accidental, is, as it were, a
reflection from the world of ideas; and yet, it is desirable to
extract from language, as it exists, whatever traces of sys-
tematic development can be definitely established; hence
Plato enters as far as possible into an analysis of existing
language, and scrutinizes its laws. . In seeking for analogies
to this method of treatment, Benfey has, strange to say, over-
looked that Platonic work which is most strikingly similar
in conception and execution, more so than the Politeia and
Politikos that he mentions. I have in mind the Noo: a treat-
ise far more comprehensive, it is true, than the Cratylus, but
equally impelled by the desire to extract an ideal code of
laws from the existing and opposite systems, prevailing in
Greece. Not for a moment can Plato have assumed that such
a code would take effect without extensive modifications and
adaptation to the limiting circumstances of time and people,
nor, I take it, was that at all his purpose, but rather to evolve
from imperfect and contradictory methods something higher
and consistent in itself. And such is the case with language. ,

Under this assumption, however, it must be evident to every
student of Plato, that the relation of the second part of the
dialogue, the so-called etymological part, must be established
with respect to Benfey’s theory. Views have diverged widely
respecting its importance from Dionysius of Halicarnassus
who considers it the cardinal point, as the additional super-
scription he gives to the dialogue: repi érupodoyiac proves, to
Schleiermacher, who looks upon it as ‘‘ Nebensache,’ and with
whom many others fail to find any purpose in this exposition.
That Steinthal alone had endeavored to fathom this curious
mixture of gravity and irony has already been referred to,
but his reasoning has been shown to be exceedingly faulty.
According to Benfey it is not only no minor part that has

68 J. Sache, .

e

assumed in consequence of Socrates’ tendency to ridicule the
etymological fashions of the day undue proportions, but it is
a legitimate outgrowth and further exposition of the first por-
tion of the work. ’Op@érnc cvoparwy he has there defined as
existing, when name and object mutually suggest and cover
each other. To the practical illustration of this mutual kin-
ship he devotes himself in the second part, but language, as
it actually exists, bristles with imperfections, and hence the
application of his principles does not result in a consistent
series of etymological analyses. Many absurd conceptions
obtrude themselves, but it is to be remembered that the sense
of the ludicrous is not what he panders to; it is rather the
weakness of language, unphilosophical as it needs must be,
that Socrates demonstrates in this extensive series of etymol-
ogies. The sense of proportion that Plato elsewhere displays
so uniformly, could never have permitted him to ignore the
limits within which ridicule proves effective; so prominent a
part as this second must have served some higher purpose ;
and if Benfey’s efforts had succeeded in establishing this
point merely, his treatise on the Cratylus would seem to me
a noteworthy performance, worthy of general recognition and
study.

V.— On the Composition of the Cynegeticus of Xenophon.

By T. D. SEYMOUR,

PROFESSOR OF GREEK IN WESTERN RESERVE COLLEGE.

Xenophon at Scillus, as Diogenes Laertius reports, spent
his time in huntiny, entertaining his friends and writing his
histories—eteréAee Kurnyer@y Kat rouc gidoug éoriwy Kat ruc toropiac
avyypagwy. Kven without this express statement we might
safely infer his devotion to the chase from the frequent and
loving references to hunting in his larger works. From the
Anabasis to the Oeconomicus no one of his writings is with-
out some allusion to this pastime, or some illustration drawn
from it.

In the first book of the Anabasis we have a digression upon
the chase of the wild ass and the ostrich, and a comparison
of the flesh of the ass with that of the partridge, a bird which,
as we know from the ancient monuments, was often hunted
and shot on the wing in Persia. Cyrus the younger is praised
AS Ppiodnpdraroc Kai mpoc Ta Sypia geAoKtvdurdraroc, aNd an anecdote
is told of his prowess in conflict with a bear. In one of the
villages of Armenia the Greeks captured the Komarch’s daugh-
ter, but her bridegroom was “ off hunting hares’’— Aaya
yxéto Snpdowr. In the fifth book of the Anabasis, Xenophon
says in praise of his home at Scillus that there are Sijpac
rartwy Grooa EaTiv aypevoueva Inpia. ’

In the Memorabilia, Socrates is represented as often compar-
ing and contrasting the acquisition of friends to the pursuit
of game. Friends are not to be taken xara rédac like hares,
nor ardry as birds. He tells Theodota that she needs some
one to act the part of a hound (éyri xuydg) for her—to scent
out the rich who are susceptible to the charms of beauty and
drive them into her nets. In another place he says that men
of the best natural endowments need the most careful train-
ing, as the best dogs, if neglected, become the worst.

In the Hellenica Xenophon mentally smacks his lips as he
tells us (av, 1, 15) of the palace of Pharnabazus, where

10

70 T. D. Seymour,

Agesilaus found such good hunting in the parks and forests.
An observation like this we could hardly find in Thucydides.

The writer of the Spartan State remarks the care bestowed
on the hunting dogs, and the importance attached to hunting
in the education of the Spartan youth.

But it is in the Cyropaedia, where the writer’s fancy had free
sway, that his love of the chase is most conspicuous. Cyrus
as a child fawned on his grandfather like a puppy on his.
master. On his first great hunt he cried out like a blooded
puppy on approaching the game. His first battle was on
occasion of a hunt of the Assyrian prince. In the celebrated
sixth chapter of the first book, Cambyses directs his son how
to take advantage of the enemy by recalling the arts which
he had used against the hares and larger game, describing the
pursuit undoubtedly much as it was carried on in Greece, just
as elsewhere in this romance many Spartan regulations are
ascribed to the ideal Persians.

The Armenians were more willing to yield to Cyrus because
they had hunted with him years before. Chrysantas urges
the other Persians to enroll themselves for the cavalry, that
they may be better able to pursue a man or a wild beast.
The son of Gobryas lost his life hecause by his success in the
chase he excited the jealousy of the Assyrian crown prince.
Finally, as soon as Cyrus was established at Babylon, he
appointed masters of the hounds and took his court out to
hunt.

Such evidence of devotion to venery prepares us to accept
the further statement of Diogenes that Xenophon wrote a
treatise on hunting. <A tract under that title is found in
MSs. of Xenophon’s works, and is referred to as his by
authors and lexicographers since the early part of the second
century of our era. It covers about thirty-three pages of
Teubner’s text, and is divided into thirteen chapters. The
first is introductory; the next describes the nets; the next
six chapters describe the dogs and methods of taking the
hare; the ninth is devoted to the chase of the deer; the
tenth to the wild boar; the eleventh, only a few sections, to
lions, leopards, lynxes, panthers, and bears; the twelfth and

On the Cynegeticus of Xenophon. 71

thirteenth are a defence of the chase and an attack upon
sophists. ro

The external evidence for the Xenophonticity of thts work
is strong. We know that Xenophon was devoted to hunting;
that he was an eminently practical man; as he wrote treatises
on kindred subjects, as horsemanship, he might be expected
to write on this subject, and Diogenes Laertius tells us that
he did write such a treatise. Arrian of Nicomedeia, who
flourished at the beginning of the second century of our era,
was not content with writing another Anabasis (of Alexander)
and a second Memorabilia (of Epictetus), but reasserted his
right to the name which he bore of Eevopwy oO "AInraiog by
writing a short Cynegeticus as a continuation of the work of
the son of Gryllus, on the ground that the elder had not
known the Celtic dogs and the Libyan and Scythian horses.

This work of Arrian is in itself most insignificant, but its
authenticity has not, to my knowledge, been questioned. It
begins with an evident allusion to Xenophon’s first chapter—
Zevopwrre ry Vpuddov AfAexrar..... ot mawWevdevree TO Neipwre ripy
maidevery Taurny Orwe Seopidreic re Haar wat Evrqsoe Kare thy "EAdCa.
Thro the whole work also, Arrian refers to the views of his
master, occasionally confirming and occasionally correcting
them. K.g., he says (1v, 5) that he has no objection to yapora
dupara, which Xenophon (11, 28) considers bad. Again,
Arrian does not consider a uniform color a fault, while Xeno-
phon calls it Snypeddec.

Aelian, living at the same time as Arrian, says (de nat. an.
XII. 24) Zevopay dé ixéip cuvor d€yer vai radra — quoting from
Cyn. tv, 9, and elsewhere refers to this work.

Hermogenes, in the latter half of the same second century,
quotes Xenophon’s description of the hounds smiling and
scowling and doubting. One expression is not a verbal quota-
tion, but the rhetorician was probably quoting from memory.

Athenaeus, Libanius, Pollux, Harpocration, and Suidas also
refer to the work, and have words and phrases from it.

The tract presents many peculiarities, so many indecd that
there is but a poor basis for conjectural emendation. But
Valckenacr is said to have been the first to suspect the author-
t

12 | T. D. Seymour,

ship of the work. In his notes to Euripides’s Hippolytus
(published in 1768) he says: “ Xenophon aut quicumque
scripsit Cynegeticon.”’ Afterwards he seems to have confined
his. suspicions to the proémium, in which Schneider agrees
with him. lL. Dindorf also in the preface to the last critical
edition (Oxford, 1866) says that Valckenaer was right in
limiting his suspicions to the proémium and the epilogus,
which is no better, “nam quod in hoc libello et-imperativorum
formae sunt Macedonicae potius quam Xenophonteae et aliae
multae non Atticae, non sufficit ad eripiendum illum Xeno-
phonti, nzsz alia accessertnt argumenta.

Here apparently the case stands to-day. No one claims
the authenticity of the introduction, as Bernhardy says in his
“ Wissenschaftliche Syntax,” who does not have a mean
Opinion of Xenophon’s understanding; and most agree with
Haupt (Opp. 1, 195) in saying that the original work must
have begun with the last section of the first chapter; but so
far as 1 have seen, critics have, with the exception of intro-
duction and epilogue, affirmed or denied the Xenophonticity
of the treatise as a whole, and mainly on general grounds.

There seems indeed much uncertainty in discussing this
question in detail. Xenophon spent much of his life out of
Attica. If we adopt the view which seems to me most. proba-
ble, that he was not much more than thirty years of age*
when he went to join Proxenus and Cyrus, he spent most
of his life in campaigns in Asia Minor and in Peloponnesus.
It is not strange then that Sauppe finds in his writings three
hundred and sixteen poetic words, ninety-nine ionic, and
sixty-three doric. A large number of these unattic words are
in the Cynegeticus, but from this alone no inference can be
drawn, especially as some allowance may be made for the influ-
ence of the subject in introducing unusual words. So those
who have rejected other opuscula of Xenophon have based their
judgment on the matter or the style, not on the un Xenophontic
use of words. Thus Boeckh rejected the Athenian State
because it must have been written during the Peloponnesian
war. Later authorities are still more definite. Kirchhoff

*See the argument by Professor Morris, Transactions for 1874.

On the Cynegeticus of Xenophon. 3

assigns it to 424 B. c., while Moritz Schmidt and Faltin set it
430-429. Their arguments are based on the allusions to the
taxes, to the comedy, and to the naval supremacy of Athens,
and they are convincing.

In a work on hunting, however, we do not expect such
references to public affairs, and in fact we find in our tract
no hint of the kind. Nets are described as used as they
were in the Middle Ages (as is shown by the allusions in old
German literature), and as we find them pictured on the
monuments at Koyunjik. Dogs are described as showing
their proximity to the game in the same way as at the present
day. Horses and bows are not used, but that seems a pecul-
iarity of place, not of time.

The Xenophontic authorship of the Agesilaus has been
disputed because of the florid style of the rhetorical encomium,
and because, tho Xenophon died at an advanced age only a
year or two later than the Spartan king, the work bears few
marks of the old age of the writer. Some have assumed the
existence of a grandson of Xenophon, of the same name, as
the opponent of Deinarchus (this can hardly have been our
Xenophon, for Deinarehus made his first public speech 836
B. C.), and as the author of the Agesilaus, the epilogue of the
Cyropaedia, the treatise on the Revenues of Athens, and the
editor of the Hellenica and the Spartan State. But there is
no reason for assigning the Cynegeticus to a younger Xeno-
phon. In fact, the only prominent stilistic peculiarity of the
Agesilaus and some of the other opuscula is (as Blass says) the
immoderate use of piv and ye pny (see de re equest., §§ 4-16:
piv thirteen times on two pages) ; but this particle is not once
used in our treatise.

Moreover it is impossible to decide upon the authorship of
this work from the statement of Diogenes Laertius that Xeno-
phon wrote /3BAa zpd¢ r& rerrapavorra, for he immediately adds,
tAAwy adAwe Statpobrrwy. There is no help to a decision from
the position of the treatise in the MSS. In the Florentine
MS. (53, 21), the only one which contains the Cynegeticus
with other opuscula, this tract is placed first, and such was
its position in the earliest edition (in Latin at Florence, 1504).

T4 T. D. Seymour,

The Aldine edition (1525) was the first to place it at the end
of the works, where it has since remained.

All these things make it difficult to refute the Xenophontic
origin of the work, or any part of it. But on the other hand
the authority of Arrian and the rest in support of its authen-
ticity proves too much. Arrian and Libanius referred to and
quoted the proémium, the genuineness of which no one would
now claim. We only infer that the work existed in its pres-
ent form, and was accepted as Xenophon’s, at the beginning
of the second century after Christ.

Further, the discussion of this question on the ground of
the internal evidence of style and constructions is made easier
and surer by our having more than eleven hundred pages of
Xenophon’s writings, the authenticity of which has never been
questioned. We are thus able to observe the minute details
of his style as well as the general features which are set forth
by Hermogenes (Spengel, Rhet. Gr. 11, 418): “Eare roisvy obrog
ageric per Gre peadtora...... KaSupoc d€ Kal edxperic, etmep Tee ErEpo.
é6 Zevopoy. His constructions are simple. He avoids all
involved sentences, as he does all abstruse thought.

But we have not merely voluminous writings of Xenophon,
but works of every period of his life, and on various subjects,
from the Memorabilia and Anabasis, written soon after his
return from service with Agesilaus in Asia Minor, to the
Cyropaedia, Hellenica, and treatise on the Revenues of Athens,
which occupied his later years. We are almost admitted to
his study. We see how ready he is to use a second time, in
almost the same language, a good thought. We see how
several experiences of his own and sayings of Socrates are
combined to form incidents and speeches in his romance, the
Cyropaedia. ;

Cyrus the Great, before Babylon, is made to extricate his
forces from a difficult position by the device which Agesilaus
used before Mantinea (Hell. v1,4,18). Cyrus the elder gains
the affections of his subordinates by the same attentions as
Cyrus the younger. The sanre thoughts on the Delphic motto,
Prae cavrdy are found in the Memorabilia, in the dialogue
between Socrates and the beautiful Enthydemus, and in the

On the Cynegeticus of Xenophon. T5

Cyropaedia in the conversation between Croesus and Cyrus.
These books contain the same warnings to young officers that
a knowledge of tactics is a small part of military science ; the
same remarks on the gradual change of the seasons; the same
views of prayer and the gods’ unwearied care for men, of
ingratitude, of ob doxeiy dd’ eivac; the same thoughts on catch-
ing hares. The list might be indefinitely extended, and all
these examples are in language so similar as to show the
identity at a glance. It would be easy to show a similar
connection between Xenophon’s other works.

If then Xenophor writes a treatise on a subject to which,
as was shown at the beginning of the paper, he has referred
so often, and especially in the Cyropaedia, like this, one of his
later works, we should expect to find many of the same
thoughts, in the same style, and not infrequently in the same
words. The objection that tlie subject excuses unusual words
and style in this point of view has less weight. I find but
two passages (VI, 26 and vil, 11) which could be considered
in any sense parallel to anything that we find in the other
works of our author. These are directions to the master to
feed the dogs himself whenever it is possible, and to rub down
the dogs before leaving the hunting-ground. Similar advice
i8 given in regard to horses in the treatise Hepi immeuie.

As regards words, I may say that we are surprised to find
here so few hunting words which had been used in the other
works. Siparpor, ciySnpoc, curSnpevric, tAéypara, moddypat, cpre-
dovat, 6 it Tote apKuat = O apKuwpos and others, which are found in
the Memorabilia and Cyropaedia, are not met with in this tract.

But I will proceed to notice certain peculiarities of the
Cynegeticus. | |

Remarkable is the frequent occurrence of asyndeta. Xeno-
phon on occasion uses the asyndeton effectively. Addressing
the soldiers after the treachery of Tissaphernes, he speaks of
those who trusted the Persians a8 watpevot, cevrovperot, b/3pico-
pevo In the Anabasis v, 2, 14, the soldiers ran together xai
ra BEAN Gpovd Epepero, AoyXat, rocevpara, opevddvae KTA. In other cases
there is somewhat less animation, as in Anab. vI, 6, 1, where
the Greeks at KaAmng Aquiy plundered mupove Kai xpcdac, civor,

76 T. D. Seymour,

doxpta, pedivag, cvka- But in all of Xenophon’s larger works
there are not so many instances of asyndeta as in these thirty-
three pages; and no example like Cyn. v. 30, which section is
remarkable in many ways, but does not stand alone in this
little work. Cf. v, 18, rove AiSouc, ra dpn, ra Péd\AEa, Ta dacéa.
Cf. also vi, 1, Kuvay dé xdopoc dépaca, ipavrec, reAapwriar’ torw O€
ra peév dépaca padaka, xharéa crA. Some passages may easily be
emended, as VI, 8, paxpa [cai] i~nAd. Others are in themselves
unobjectionable, as perhaps 1x, 1 and x1, 1, but taken together
they are extraordinarily frequent, and the first mentioned, v, 30,
is desperate. There is no rhetorical animation to excuse it, nor
a long list of qualities of one object, but the sentence is made
obscure by the frequent juxtaposition, without conjunction, of
two or three nouns or adjectives.

I notice next the use of prepositions. Professor Tyler says
(Transactions fur 1873) that thirty-six per cent. of Xenophon’s
verbs are compounded with prepositions. Beginning with 1, 18,
the part of the work most Xenophontic in character, we find
that thirty-seven per cent. of the verbs in the first nine sections
are compounded with prepositions ; while in chapter v. we find
that fifty-seven of the first hundred verbs are so compounded.
This can hardly be mere chance, especially as many of these
compound verbs do not differ sensibly in meaning from the
simple. Thus eddy and xcadebdy, evel and broxeet are used in
parallel passages ; ércyrwpilw like yrupifw; ixayw like ayw.

This of course points clearly to a later origin for the pas-
sages in which the unusual number of compounds is found.

Further. On the twenty-seven pages which are devoted to
the treatise proper, excluding the proémium and epilogue,
there are twenty-one verbs which are compounded with two
or more prepositions, thirteen of the twenty-one being on the
ten pages which begin with chapter im. The last twenty-
seven pages of the seventh book of the Cyropaedia, which I
took up at random, have but one verb so compounded. Other
passages have more, but that the large number here is not due
to chance or the nature of the subject, is obvious from a glance
at some of the verbs; éyxaramdécw being equal to épurdékcw,
éyxarappanrw to evparrw. Compare zpodietéASwor, V, 4. It is

On the Cynegeticus of Xenophon. 717

evidently the result of the growing tendency, noticeable e. g.
in New Testament Greek, to make the verb more definite by
prefixing a new preposition.

Moreover, there are on these twenty-seven pages forty-three
cases (thirty-one different verbs) ‘of the repetition of the
preposition with which the verb is compounded, before the
nouUN, a8 avo rwv Kuynytoiwy aradXarrovet, Smeppopei brép Ta ToWUTW,
and others. In the Memorabilia, one hundred and forty-two
pages, I have noticed but thirteen examples of this repetition ;
and of these thirteen, two are in passages suspected by
Valckenaer and Dindorf. In the three hundred and thirteen
pages of the Cyropaedia I noted but fifty-five examples. At
this rate the twenty-seven pages of which we are treating
should have not more than five, instead of forty-three. This
of course may indicate hasty preparation as well as interpola-
tion, but we are hardly prepared to find it in Xenophon.

A few instances of irregular constructions with prepositions
and verbs compounded with prepositions, deserve our notice.
Chapter v, § 18 we find droywpmor rove AiS9ouve. The first exam-
ple I find of an accusative after this verb'is in the scholia to
Euripides’s Phoenissae, 105. Two lines farther on aroxwpizover
is found, and aroywpover might easily be emended, but Dindorf
has remarked on the transitive use of xwpeiv in this sense in late
Greek. Perhaps this accusative (v, 15) is better explained
as the limit of motion, but one would be puzzled to parallel
that from Xenophon.

For agierayrat rév fdov, 111, 8, Dindorf, following Schaefer
compares Anab. 11, 5, 7, a well-known sentence: ror yap Sear
wédEpov ox olda vir’ ard Tolov av TaxoUC PEevywy Tic aTOpUYoL,...
arodpain,...anosrain, But surely the Greek usage did not
demand the repetition of a noun in another case because the
third verb in such a series did not govern the accusative. So
our construction, roy cov, is unusual in Xenophon.

The use of azo in such expressions as IV, 4, yrwpiZovea: aro
Tov Supod, awd rijc KEpadijc, awd TwY Ouparwy Krd, is not Xenophontic,
Compare Xx, 12, riv Kivnow amo Tig Kepadiic, instead of the simple
genitive.

Peculiar also is the use of cea with the genitive. III, 5, éea-

11

78 T. D. Seymour, |

Tpéxovet Ota rov txvouc, tho vil, 6 we have dtarpéxery ra tyvn, which
is obviously the normal construction. Compare with this rv, 3,
Tpotrwoay Ova Tov tyvouc, and VI, 22, Carr wor Oa Tov tx voug, and
strangest of all, x, 16; agpicur’ ay dia rij¢ paBdov. We have an
example of this in St. Matt. vir1, 28—wapedSeiv dua rife dd00, but
it is not classic usage, and by no means parallel to da ray
dpéwy, dua rov rakewy and the like.

Mera is used eight times, iy but once, except in composition.
One use of pera is unquestionably not Xenophontic. x1, 3, the
the wild beasts descending to the plain by. night are caught
pera ixxwy cai Ordwy.

The preposition is sometimes irregularly omitted, as Iv, 9,
cic ra Bon modAAKte, Ta O€ Epya Hrrov. Of. v, 15, rove Aepwvag, rac
vanac.

In connection with these may be noticed V, 27, aya rotvrac =
‘‘besides this reason.” VIII, 1, tw wodvy ypévoy seems clearly
corrupt, as éw cannot be naturally joined with éj\a. Another
particle to be noticed is oré in éré péy, and dre dé. Never used
in the larger works of Xenophon, it is found in this treatise
four times, v, 8 and 20; 1x, 8 and 20.

In rv, 1, ra peyéSn peraty paxpov cai Bpdyéwr, we notice that
Xenophon regularly uses the singular of péyeSoc, and that
peraéy can hardly be found in Xenophon used to denote what
is between two qualities, as here, “long, short, between these.”
Compare algo V, 8, doer rodt, puxpdv, peragy rovrwy, “ far away,
near, between these.”

Another peculiarity is the omission of the reflexive pronoun,
especially with pexreiy and its compounds. vV, 4, yaipovrec yap
Te peyyet éravapoinrouvTeEc pakpa dcatpovary avrervaizovrec, Where we
expect atrouc with both dcacpotory and éxarapperrobvrec. Cf. Vv, 8,
éré d€ kai év ry Sadarry ciappimrey. Also VI, 22, émipperrovea, and
1X, 20, pexrotor. Where Theognis speaks of poverty he says
(175):

iv O€ xpi) pevyorra Kai Ec peyaxyrea KOvrov
pimrety cal werpwy, Kipve, car’ HAdiarwr.

Two similar examples of the use of pra are found in Eurip-

ides, and one in Menander, but I have met with none in
Xenophon’s unquestioned works, nor in other classic prose.

On the Cynegeticus of Xenophon. 79

We find in this treatise, moreover, an unusual number of
periphrastic expressions, specially with gyer. pndev dv h yi
avinow (VI, 25) is not unlike wayra dca dpa pvovm (Anab. I, 4,
10), but x, 23, dv av dav dupw as equal to “ both the parents”
of the wild beast, is not so natural. Many periphrases with
éxey are found in all of Xenophon’s works. They are not
uncommon also in Isocrates, as in his Panegyricus, § 67, we
find gor: yap dpyixwrara kat peyiorac duvacreiac Exovra. The rhyth-
‘mical argument for the construction is quite lacking, however,
in sentences like Iv, 1, of our tract, mpwrov per ovv xpi eivac
peyadag elra éxovoag rac xepadag éAagpac xrA. Cf. 1, 8, aobvraxra
Exovoat ra owpara. Stranger still is 1v, 8, ai pev ody wuppai zxov-
oat torwoay AeEvKiyy rpixa «rd, and VI, 1, of dé inavrec [éorwoar]
éxovreg ayxiAag xrX. Most awkward of all, however, is the
beginning of VI, 5, rv d€ orodny 6 dpxuwpog ELirw Exwv txt Sypay pa)
éxoveay Bapoc, where the zxoveay so near Yijpay, and far from
oroAny, is & Clumsiness which we can scarcely impute to Xeno-
phon, especially as the same short sentence has another case
of that participle. I can give parallel examples only from
later Greek, as Pausanias v. 18, avip rn perv deca xidcca rip de
Exwy éoriv Sppoy.

Another peculiarity of this opusculum in its present form
is the use of the infinitive. Perhaps eiva in 1x, 1, émi dé rove
veBpouc Kai rac éXagove Kuvag elva Iyduac, Will serve as an exam-
ple. Compare xexrijoSa, X,1. It is evident that these are
not exactly like [epi Irmecfic 111, T, reipay AapPavev, where the
infinitive is in apposition with the AauBave» contained in
Anrréov of the preceding clause. At the beginning of chapters
1X and x the subject is changed, and after the break it is not
a mere matter of course to carry on the force of the de in
vil, 8. In v,15 also we have this infinitive A\apBdaveyv without
any word on which to depend. No yp# or dei has been used
in the whole chapter; and that this is not the imperative use
of the infinitive is shown by the accusative of the participle
urayovra, Which must agree with the subject of AapPavev.
This example in v, 15 is the first in the work. Before this
the imperative and the infinitive with yp/ are used. Thus in
chapter Iv the imperative is used nine times; xpi twice;

80 T. D. Seymour,

duecvov tore ONCE, aNd ayadoy gore Once. After v, 15 the next
18 aréyeodat, V, 84, which, if it were alone, might be taken as
used for the imperative. I, 8, aye» may be taken to depend
on the xpf in § 2. Soalsoin§4. But v1, 11, we have ro» é
xuyyyerny ékevac after twelve imperatives. This infinitive is
constant thenceforward to the end of chapter x. Chapter x!
is brief, and xi and xi do not need it or have it. This
infinitive must depend on the idea of advice stated, 11, 2, dca
dé kai ola det rupeoxevacpeévov ehdetv ex’ abro ppacw Kat abra cal ray
émoripny Exdorov iva mpoedwc éyyespy re Eoyy. But it is more
than seven pages after this that the first infinitive is found
which depends on this introductory sentence.

If some of these peculiarities seem slight, and the argument
to be founded on them weak, I would call attention to the
cumulative force when several of these unusual constructions
are found in one passage.

For myself, then, I am convinced that Xenophon did not
write this treatise in the form in which we have it. A com-
parison, however, of the passages in which the most marked
peculiarities to which I have referred occur, shows that most
of the solecisms and difficulties are contained in certain
sections and chapters which may be omitted without inter-
fering with the symmetry of the work; and further, such
omission will remove certain difficulties in what remains.

The results of my investigation are as follows:

Xenophon began with 1,18. The long list of heroes who
excelled in the chase, found in the proémium, is not so much
in the style of Xenophon as of the later rhetoricians; and,
as Mure remarks, it is absurd to preface with so much pomp
a tract mainly devoted in its present form to the pursuit of
hares, which were not the game of Hercules and Theseus ;
avayopeveijvac for dvappySivac (1, 14) is not Attic; and the
style in general of these first seventeen sections is not that of
the Memorabilia, nor of the Cyropaedia. To begin with éys
pev ovv (1, 18) seems at first abrupt, but is not unlike the
beginning of the Revenues of Athens, éya péy rovro erd.3 nor
the introduction to the Hipparchicus, Uparer per Soorra yph xra.

On the Cynegeticus of Xenophon. 81

Moreover, according to Schneider the Breslau manuscript
omits the ovv of the éyw pev ody, and until within the last
century the editions began the second chapter with what is
now § 18 of the first chapter.

To the close of 11, 8, Xenophon writes of nets and their
props. Then follows a long interpolation, out of connection
here and containing many deviations from Xenophon’s style,
to v1, 7, where the dropped thread is again taken up and
directions given for fixing the props for the nets. In these
interpolated chapters are found the most unusual cases of
asyndeta, the most remarkable periphrases (as vI, 5), and
the greatest license in the use of prepositions. |

From vi, 7, our author tells how the hunt is to be begun.
The infinitive in that section can now, assuming this long
interpolation, be easily made to depend on dea é¢ xai ota dei...
gpacw of 11, 2. After vi, 16, there is an interpolation of six
sections to the beginning of § 23 which resumes the narrative,
and the close of chapter vi brings the hunter to his home
after the chase of the hare.

Chapter vii is devoted to the care of dogs and their breeding.
Ac OtcaXredrey 2... Epa, in § 2, 8§ 5 and 8, and ra edn in § tT;
I consider interpolations.

Chapter vill treats of tracking hares on the snow. Part of
§ 1, cid évéora. .. agavizea, may be from a later hand.

Chapter 1x, on hunting deer and fawns, has interpolated
§§ 8-10, 18-16, 19-20. In §§ 8 and 20 we find the use of
ére 6€é, which is unknown to the other works of Xenophon, and
the reflexive use of purreiv.

In chapter x, on the wild boar, I hold §§ 4-18, and xpnoréor
. ++. maoxot in § 22, to be interpolated.

Chapter x1 treats of hunting panthers, lynxes, etc., which
were not at all in the line of Xenophon, who always writes
from his own experiences. Moreover, the second paragraph
speaks of poisoning water and food for the wild beasts, a
procedure far from the sportsmanlike spirit of our author. In
the third section, also, is the late use of sera, of which I have
previously spoken ; and it is worth mentioning, perhaps, that
pera iS found in the whole treatise eight times, but seven of

82 T. D. Seymour,

- the eight times in passages rejected by me on other grounds.
ovv is used but once (VI, 16, ov» raic oipaic) out of composition,
and that in a passage which I hold to belong to the original
work. It is known that Xenophon used pera less, and cw»
more than his contemporaries and later writers.

Chapter x11 may well be genuine as far as § 17, where the
original ends, ending as it began with the praise of Madeéa.
The eighteenth section contains a direct reference to the
proémium than which, as Dindorf says, the epilogus is no
better. .

This scheme attributes to Xenophon less than half of the
treatise before us; but it removes or explains a much larger
proportion of the difficulties and’ solecisms, while what is left
is in a condition to receive emendations which the wretched
_ state of the MSS. renders necessary.

This theory relieves Xenophon of the responsibility for the
following statements: that hares do not see well because they
rush past everything with such rapidity that their eyes have
no practice in examining objects carefully, and because their
vision is injured by so much sleep (v, 26, 27); that their
tails are too short to be of much use to them as rudders in
running, but they make use of their ears, dropping them to
the ground and bracing themselves upon them when they turn
quickly to avoid the hounds (v, 32); that the breath of the
wild boar is so hot as to scorch the hair of the dogs which
approach him; and that a hair laid upon his tusk immediately
after his death will shrivel up (x, 17). These statements,
which savor of Aelian and the later writers, are all in rejected
passages. My theory also relieves Xenophon of the responsi-
bility for a few stupid puns and unnatural rhetorical clauses.
Compare V, 17, Séovet yap padtora peév ra dvavrn ij ra Gpadd, ra be
avopoa avopotws Cuneven places unevenly, i. e. less than up hill,
more than down hill), ra 6€ karavrn ijeora. Here avopoine iS ,
used, obviously introduced solely for the sake of the Parono-
masia. Compare VI, 20, where the hunter robvopa peradddovra
(literally changing the name, where he means calling the name
of each in succession) éxdarne rij¢ kuvdc, is to shout, making the
sound of his voice 620, apt, puxpor, péeya. Of what advantage

On the Cynegeticus of Xenophon. 83

it would be to give a little call, when the hounds are supposed
to be at a distance, we are not informed. But it would take
us too long to consider every example of slovenliness or stu-
pidity of thought and construction in the work as we have
it. The same portions of the work, then, which contain
statements and thoughts which we are not ready to ascribe
to Xenophon, also contain the unusual constructions to which
I have called attention. The theory propounded in this paper
claims attention on the ground that it so largely removes
what is unlike to or unworthy of Xenophon, and still leaves —
a framework far more symmetrical than the traditional form,
with a beginning, a well-arranged middle, and an end.

From the external evidence in its favor, as well as from
certain internal marks of style, I am inclined to believe that
the Cynegeticus is from the hand of Xenophon. If that be
still disputed, I claim that the evidence here brought forward
for an earlier and a later hand in its composition is still
unshaken.

The interpolator generally contented himself with inserting
chapters and paragraphs. Only occasionally are we obliged to
cut out from a sentence which seems Xenophontic a word or
‘ two which, as is evident from other passages, proceeded from
the second hand. Only once is it necessary to the construction
of the sentence to supply anything from an interpolated section.
In VI, 7, 6 dpxuwpey, the subject of éx:Baddérw, must have been
dropped by the diasceuast when he wrote §§ 5 and 6.

Who this interpolator was, it is perhaps idle now to inquire.
We only know that he must have lived not later than the
beginning of our era; for Arrian early in the second century
after Christ seems to have accepted this tract in its present
form as the work of Xenophon.

VI.—Elision, especially in Greek.

By M. W. HUMPHREYS,

PROFESSOR OF GREEK IN VANDERBILT UNIVERSITY.

I propose, in this paper, to discuss the nature of elision in
Greek ; and, in so doing, I shall first examine the views of
others, and then present my own.

Corssen, Westphal, Heinrich Ahrens, and many others,
hold that elided vowels were not entirely suppressed, but
merely diminished, and one argument they employ is that the
Greek name itself, cvvadoih or cvyxpeocc, does not signify total
expulsion, but implies that the vowels were united in a rapid
pronunciation, and did not suffer what was called &SAuwke, or
expulsion. To this I reply: 1. That the ancient writers are
not always to be interpreted literally ; for, as one vowel, or
rather one syllable, appeared to result from a combination of
two, and the elided vowels sometimes were slightly sounded (a
point to be explained hereafter), there was no reason why
they should not, in a loose way, designate the process by the
word ouvadowpy, OY avyKpiate, which does not necessarily mean
anything more than “conjunction”; and besides, already at a
tolerably early day they employed the term ixSAuhe to denote
elision. Even modern writers are not exempt from much
more inaccurate applications of terms than even ovvadcg7 in
the sense of elision, to say nothing of the other words. As,
for instance, the German grammarians call final m and ” in
French a “ Nachklang,”’ or “ after-sound,” as if they were
pronounced after the accompanying vowel, while every one
knows that they merely give the entire vowel a nasal tone.
Many illustrations of: this could be cited, but this one must
suffice.

2. Moreover, when the ancient grammarians speak of
the suppression of hiatus, they frequently fail to distinguish
between the various processes, or else between the words that
denote them; and in the very passage cited by Corssen,
clision is confounded with crasis. The passage is: “Eore de

On Elision, especially in Greek. 83

auvadotp) dvo dwrnévrwy dinpnpévwv eic piay svdrAaBjv Evrworc, oloy
ro Gvopa. robvopa. If the metrician had not added his
example, the inference would have been that cuvado) was
always the combination of two vowels into one, whilst the
example he gave shows that he had only crasis in mind when
he cast his definition; and yet Corssen wishes to apply the
definition to elision.

But as it is the custom of many now-a-days to dismiss the
question of elision with the statement that Ahrens has shown
it to have been only a partial expulsion, and as Ahrens has
given about all the arguments for that view, I proceed to take
up his arguments and examine them one by one. In the first .
place, Ahrens says that if elision is total, the letter immediately
preceding the elided vowel closes the word as thus modified,
and he calls attention to the fact that we then find not a few
unpronounceable combinations, as éo3)’, céu»’, etc., and others
which the Greeks would not tolerate, as vv«r'’, reid’, etc. But
if, as he asserts, elision does combine two vowels into one,
then the two words become one; and why then may we not
be allowed to combine the words after expelling one of the
vowels? And this is exactly what happens, except in some
instances about which I shall presently speak. Secondly,
Ahrens says that adye é3nxev, avr’ eueto, and similar combina-
tions, would still have a “‘ hiatus offensionem, quam non inesse
constat.”’ How does he know? The Greeks did not suppress
two syllables, by elision, because this would have maimed the
word too severely, so there was nothing left them but to
tolerate the new hiatus, as custom required them in poetry
to remove the original hiatus. And besides, does Alrcns’s
diminution-theory remove his own difficulty? It seems to
me to increase it, for who will pronounce dAye* éSnxev for us
(pronouncing the final a of adyea, yet making it of inap-
preciable length)? And in 4yri’ éueto there is surely a less
offensive hiatus than in avri* ¢ueto. Moreover, it is a well-
known fact that languages, in removing one hiatus, sometimes
create another no less offensive; as, in Sanskrit vané 4sit is
resolved into vanai Asit, and the i being elided, we have vana_
dst, where the hiatus appears even worse than at first (Bopp,

’ 12

86 M. W. Humphreys,

Crit. Skt. Gram. § 38). But after all, I am willing to admit
that there is nothing offensive in the remaining condition of
things when elision has been made; for the two words are
pronounced continuously, and the vocal muscles do not have
to arrest themselves and then renew the exertion as they do
in case of real hiatus.

Thirdly, Ahrens draws his conclusion from the scholia on
Kurip. Or. 279:

"Ex Kupatay yap avrsic av yard’ opus.

The scholiasts on this passage (and also on AristopuH. Frogs,
804—not cited by Ahrens) say that Hegelochus, getting out
of breath, passed rapidly. over the elision, and the spectators
thought he said yadjy dps, which circumstance gave Aris-
tophanes (Frogs, 304), Strattis (Anthroporrhaestes), Sannyrion
(Danaé’), and others an opportunity to amuse their audience
at the expense of Euripides and his great actor. But, as I
_ shall show hereafter, when occasion demanded, the Greeks did
sometimes slightly sound elided vowels, and one of the most
natural places to do this is where ambiguity might result
from total elision; and the statements of the scholiasts show
that what Hegelochus did was nothing unusual under ordinary
circumstances ; and as to his breath failing him, that seems
- to be one of the many inventions of the very fertile minds of
the scholiasts. If he did not have enough breath to utter a
‘¢ diminished’’ vowel, how could he add ope so as to be heard
by thirty thousand people? And if he stopped to take breath,
then he did not pass rapidly over the synaloephe. The fact
may be that, getting out of breath he lowered his voice, thus
making yadjr out of yady’, and then took a breath and added
ép#, Which would complete the transformation, since the elision
should not be complete where any pause is made. This,
however, is a mere conjecture, and is not necessary to the
explanation of the matter. The ‘mistake on the part of the
spectators was anyhow quite natural, because dpayv yaAnva was
a forced and unnatural expression. |

Fourthly, Ahrens observes that evra) does not denote
expulsion, and that clision takes place before a pause, and at

On Elision, especially in Greek. 87

the end of a veise, and even between two speakers. I have
already spoken of the meaning of ovradu~4; but here I shall
discuss the subject more at length. That the word é&Auhte
was employed to designate elision is well known, and the only
question is how early ft was so used. I shall not attempt,
however, to settle this question, for it is clear that it was so
used sufficiently early to show that whatever it denoted had
an existence in classic times. Draco enumerates seven kinds
of synaloephe, among which he places é«3Acuc, which he defines
thus: kai Ane pév gore Evdog gwrige vroc amwwAera, and
illustrates by iz’ guov for iwéd éuov, although he defines syna-
loephe itself thus: Suvari} cé i rod mpoepnpévov Kal évredovc
gupmrucic re kai Evworc: a definition which shows how
careless the ancient grammarians could be in their state-
ments; and, in my opinion, they had crasis also in mind, or
even exclusively in mind, when they appear to apply the word
cvvadoupy to elision, except that when employing it as a generic
term, they sometimes apply it specifically to elision, just as —
one may call a temporal sentence a relative sentence. Here
is another statement of the subject: Zurudowy éore db0 cvAAa far
Kava pwrvyevra Evwarc KarajoAy Tévwy. yiyverac O€ KaTa TpdTOUG EMTa,
drhovg péy tpeig Kara ExSAcWer, ew Eué avti rod éwi éué’ xara
Kpaouy, Taya avril rou Ta Ea’ KaTa ouvaipency, vnpHydec avTi Tov ynpnidec.
auvstrove O€ régaapac, KaTa EKSALWeY Kal ouvaipecty, E“ovUTodUrEt
arrirov éuot broduver’ Kara Kpaay Kal cuvaipemy, waddo¢g avTi Tov
G@ aimddog KkaTa ExSrAtWev Kal Kpdotr, Kayw arti Tov Kat Eyw'
Kara éxddunpv kai Kpaow Kai aoaipentr, Ev tgSworig avrt rou ey Ty
AiSworig. Here it is evident that the grammarian by &xSAuhe
means total expulsion, for in éx’ éué he says we have éxSrAne,
and in cay& both &SAche and kpiow; that is, the « in cai being
elided, we have xa’ éy® which then suffers xpaowe; and no one
will deny that this « was totally suppressed. Hephaestion
therefore rightly distinguishes between ovvexpeornacc, by which
whéwy (Il. A. 183) is reduced to one syllable, and synaloephe
(generic, including elision), by which a vowel is rejected, as
wy’ Exardyyxepor (Il. A. 402), Str’ ep’ GAde (Il. A. 850). But
these two processes would have been the same, if elision had
only been a diminution. And the scholiast on this passage
does not err when he says: Acagéper 6€ cuvarvwpy suvexpurijcewc,

88 M. W. Humphreys,

ityouv ovviZioewe. fh pev yap cvvadoion ac ypagetactotrw ai
Exdwretrac’ i 6& avvignme oby we ypaperat Expwreirat, GAN’ év Tw
Baivecy rag do avAXdafsac Gpov rep Tic TOU pérpou SenaTElac GuvEKgwrET
dco Kal ouvilnore A€yerac KTE.

§ 2. So far my arguments have been negative. I shall
now present my own views, and support them with a brief
discussion of the evidence in their favor.

In prose, as is well known, the Greeks tolerated hiatus,
except that some rhetoricians tried to banish it in artificial
compositions, an illustration of which we have in the orations
of Isocrates. But then, if they chose, they could elide.
Hence we draw the important conclusion that the Greeks
could elide or not elide, as suited their convenience. One
might assume this as a matter of course, but Cicero, while
testifying to this peculiarity of Greek, denies that the same
privilege exists in Latin. He says (Or. 44, 152): ‘Sed
Graeci viderint: nobis ne si cupiamus quidem distrahere
voces conceditur,’’ etc. But in poetry the Greeks avoided
‘ hiatus for the most part, and in tragic trimeters banished it
entirely, except (apparently) after ri. as AESCH. Sup. 306, ri
ovv; SOPH. Philoct. 917, vi eimac; and rarely after «d in close
combinations. But frequently it was difficult to prevent a
word which ended with a vowel from preceding one beginning
with a vowel, even when there was a pause between them.
In that case they did not totally expel the vowel, nor even
necessarily reduce it to inappreciable quantity. Whenever
this happened the clision was indicated as if total, while in
recitation the elided vowel was either pronounced in full or
merely diminished, just as the sense required or permitted.
Another instance of partial, or apparent clision is where an
emphatic monosyllable apparently loses its vowel, as Eurip.
Tro. 945: ob 0’, GAN Enauriy roi ree? eonoopa. So Alcest. 984.
Also where the sense would be obscured, as Herc. Fur. 972:
aAdoc GdXoa’, ec wETAOUE O pEY pnTpos kre. CF. SopH. Elect. 1499,
Eurip. Ion 3, etc., etc. And thus it came about that if
for any reason they desired it, they felt themselves at liberty
merely to diminish a voiel, even when there was nothing but
metrical considerations to prevent its total expulsion. This fact
is of special importance in determining certain effects of elision

On Elision, especially in Greek. 89

in the construction of verses—a subject on which I propose
to present a paper at some future time.

But that vowels could be, and actually were, entirely
expelled by elision, is shown by the following considerations :

1. When the second word begins with an aspirated vowel,
then the aspirate affects the final consonant of the first word,
if it can be aspirated without changing its character, as vix¥
dANY, EP Hiv, Swpay’ drwc, which seems to me impossible if the
elided vowel was pronounced ever so little, for then it would
have separated the consonant from the aspirate. This, it is
true, does not happen in Herodotus ; but then in H. it does
not happen in compound words, like arinj, where all admit
total elision.

2. If the ultima has the accent, it goes back to the next
syllable when -elision takes place, and enclitics retain their
accent when elision takes place before them. This would
hardly have been the case if the elided syllable had only been
diminished, for the Greek accent was merely an elevation of
the voice, and not stress. (This recession of the accent from
an elided ultima is found in some of the examples used by
Ahrens to prove that the vowel was not entirely suppressed,
as iwSd', céuv’.) We have in the Greek language itself an
instance of the accent remaining on a merely diminished
vowel (or at least not seeking another syllable), and that is
in aphaeresis, a8 éxeivw dwxev or exeivy “Cyxev. Thiersch ridi-
cules such accents, calling them “ accentus depofarovrrac,’ but
if he had put on his phrontistic spectacles he would have
detected a xpexdSpa on which they ride; in other words, the
omission of the vowel in this case only indicated its diminution
to inappreciable quantity, while the accent still remained on it,
just as in Sanskrit we find an accent (the svarita) partly on v
and Jj, although these not only fail to make syllables them-
selves, but even do not lengthen a short syllable, as in svar,
kva, nadjas (Bopp, Crit. Gram. § 30, 1); and similarly even
in Greek where an accented vowel suffers synizesis, as in
Aivéag (Rhes. 85), dproréwr (Alcest. 921), revyéwy CAndrom.
167), OOTEWY (Tro. 1177). So Oirkwe, "Aywrdéwe, and in Hom.
Od. Aiyerriove, with hundreds of instances everywhere. In

90 M. W. Humphreys,

all these the accented vowels become virtual consonants.
This view is further sustained by the fact that when the first
syllable of a yord is entirely lost, the accent on it is removed
to the next syllable, as in the Homeric Arc for E“adde, where
no vowel precedes. Corssen, indeed, denies that such forms
have lost the augment, but G. Curtius more successfully
maintains that they have. But as it is important to establish
the position that in aphaeresis the vowel was thus diminished
and yet.retained the accent, I must not leave unnoticed the
fact that Thiersch, Buttmann, and others deny the existence
of aphacresis, and assert that all the apparent instances of it
really belong to crasis. This view, though, cannot stand in
the face of the following facts:

First, such combinations as divaya *yw, which are fre-
quently found in Mss., and are not wanting in inscriptions,
would have to be written dvvapays, with omission of « and
contraction of a with « Secondly, when the word suffering
aphaeresis begins with an aspiratéd vowel, the consonant
beginning the syllable preceding would become exposed to
the aspirate as in Soipdrir for ro iparior, Sypépa for rH jpeg,
_ Sarépq for rH érépg Conce drépa), whilst in reality we find such
examples as airy ’réoa (ARISTOPH. Lys. 736) for airn érépa,
which, by crasis, would ‘become abSijrépa (ai3arépa?). I am
willing, indeed, to concede that some instances of aphaeresis,
as found in the texts, are to be written otherwise, as x2 ’yyovsa
(Lys. 48), which ought to be written xiyyovea, for the article
loves crasis, and I suspect that the usual way of writing these
words is due to the fact that the double crasis seemed rather
bold, and obscured the words. Felton’s a ’Aaje (Clouds
1268) with long « is certainly wrong. Thirdly, aphaeresis
sometimes takes place after a long pause where crasis is
impossible, as Clouds 1354: éyw gpisw* weedy xré.; Philoct. 591.
Aeyw' "rt rovroy «ré.; Rhesus 157: Féw" “wt rovrae xré. 3 Iph. in
Aul. T19: péddXw ‘rt ravry «re, etc., etc. It is sometimes
regarded as taking place at the beginning of a verse, but a
careful examination of all the Greek dramatic poetry convinces
me that this may have been a mere omission of the augment
of verbs (which frequently occurs in psec dyyedcwaé), although

On Elision, especially in Greek. 91

in the great majority of cases the preceding verse ends with
a vowel. This vowel is sometimes short, as in Oed. Colon.
1605-6, and sometimes we find a consonant, as Oed. Rex
1245-49. I am not so sure, however, that aphaeresis may
not take place after a short vowel; and I shall presently have
occasion to cite a case of similar aphaeresis in Latin. But to
return:

3. Diphthongs are frequently elided, and especially in the
verbal ending -a. Now can a whole diphthong be reduced to
inappreciable quantity ? It is difficult to reduce a diphthong
even to a short syllable; nor is there any reason why the first
vowel should be diminished unless the second is entirely
removed, so that those who assert that elision is mere diminu-
tion are compelled to affirm that « is dropped entirely and
a diminished; but if: in a diphthong can be dropped entirely,
why cannot any elidable vowel be thus dropped, as a in aAyea
éSncev, yadkmra ops? One might reply that the « becomes a
sort of consonant or semi-vowel, like y; and I believe that
this is what actually happens when a diphthong is shortened,
&S IN vik toe oF rpvywr, Where «= y, and in ailerov év vepédnor, -
and isev é¢ueio, where v = w, since v is never elided, the well-
known exception in a quoted oracle in Herodotus being only
apparent. But if this is what becomes of the second vowel
in case of elision of a diphthong, there is no reason at all for
the shortening of the first vowel, as there is no longer hiatus.
In such instances, therefore. as cowaad’ éy rode (for xoupaoSac),
KoAdo ékeore (for xoAdoac), dovv’ éveore for (dovvar), yay’ éxijpe (for
yijpa), déou éyw (for déouac), necessarily the second vowel,
and in fact the first, too, was elided, unless for some special
“reason it was desirable to make the first audible.

4. Epicharmus, as quoted by Athenaeus (vil, p. 338, d;
see Ahrens, de Crasi ct Aphaer. p. 2) plays upon y éparo¢g and
yépavos, from which it appears that the « in yé was suppressed.
Aristophanes (Clouds 1273) appears also to play upon az’
évov and azo vou. Further, Dion. Hau. (De Comp. Verb. c. 11)
calls xrumeir’ (for crumeire) “two syllables.”” Jam not disposed
to make much of this, as an inappreciable vowel might be
omitted in counting syllables metrically.

92 M. W. Humphreys,

dD. The words érar, érérar, yap, your, etc., for dre av, dwore ay,
ye ap, ye ovr. etc., show that the vowel was entirely suppressed ;
and after they had been a long time in use, the combinations
began to be regarded as single words. This might happen, it
is true, merely from long juxtaposition, as we have in Latin,
(where elision does not appear to have been total) tantdpere,
magnopere, for tantd Gpere, magno dpere. But this is much
rarer than in Greek, and we have in Latin two vowels united
into a diphthong, as in neuter, neutiquam, deinde, etc.

6. Finally, if elision had been only a diminution of the
vowel, as in Latin, it would not have been subject to so strict
limitations, but would have been as universal as it was in
Latin. But, as is well known, elision in Greek was strictly
forbidden under certain circumstances. For instance, a,
and o in monosyllables were not elided (except a in od);
and v was never elided at all. The seeming exception in
Herodotus (vit, 220) dor’ éoxvdéc should most probably be
doru 'pxvdéc (a sort of aphaeresis after a short vowel), or perhaps
the oracular poet or priest was at his wit’s end for a verse,
and admitted diminution where expulsion was not tolerated.
When v closes a diphthong, it does, indeed, seem to be elided;
but in that case, as I have already said, it was probably
pronounced somewhat like ww, just as 6 in Sanskrit before a
vowel becomes av, where v was, no doubt, pronounced like w.
Further, érx and zept do not suffer elision, possibly because
they would then sound like ére (with its « elided) and zep,
which would not be the case if their vowels were sounded
ever so little. Some words, however, with long vowels, did
suffer a partial elision; but this is one of those exceptions
that prove a rule; for if all elisions were only partial, then
joy ob (as one syllable) should be written yp’ od, and would be
an ordinary case of elision. Nor is it crasis, for then it would
be pe (ef. por for pu) ody); and moreover the combination may
occur when a slight pause intervenes, as Oed. Tyr. 944:
réSyyxev' ci CE pry, abrog aba Yarveivy, and also where crasis
would utterly obscure the sense, as Trach. 85: ji’ i) ceaws-
preca | keivou j3tov owaarroc, i 0X0 peas apa. The contraction
of jo) ob into pe would itself be rather obscure; but if so in

\
On Elision, especially in Greek. 98

writing, then certainly also in speaking; and that contractions
of the sort, when made in speaking, were also indicated in
writing, is shown by yovv for cai 6 év, Korie. Heracl. 178;
SdéxX' amcévac for ra Orda ameéva, Birds 449; and even «ai for
cai ai, Lysist. 1105. We have, however, an instance of crasis
not indicated in writing, in Eurip. Orest. 599: ei yi) 6 wedeboag |
pvoerai pe pi) Saveivy, unless with Witzschel and others we omit
6, or admit synizesis of a long vowel with a short one, result-
ing in a long syllable.

Further instances of non-elidable vowels are found in the
genitive ending -oo,-ao, and to a great extent in the dative
ending ~; and third singular endings in -e are not elided
before av, unless we admit cix’ ay pérpov (Ion 354); for cvréoy’
ay (Alcest. 901) is an impossible conjecture, and éAavSav’ ay
(Sopa. Elect. 914), though desirable as to the sense, is not
the mss. reading. Other instances of forbidden elision might
be cited. All this proves conclusively that elision was recog-
nized as having power to remove a vowel entirely; for,
otherwise, there was no reason why elision might not have
been as general as it was in Latin, where the restrictions, as
far as they exist, merely have reference to too great a mutila-
tion of the word, and were a refinement of artificial writers.
The vowels which could not be elided entirely in Greek, were,
for the most part, not even allowed to suffer diminution to
inappreciable quantity (Latin elision), because this was a
modified form of ordinary elision, and was written as elision,
i.e. the vowel was omitted in writing, except in those few
cases where the eclision was never total, as uo’. An inves-
tigation of these latter cases would lead to a discussion of
synizesis and synaeresis, which is foreign to the object of this
paper. "

§ 3. These arguments seem to prove that vowels could be
and frequently were entirely expelled by elision. It now
remains to be shown that not unfrequently they were, for
special reasons, only partially elided; and indeed they some-
times had nearly or quite their full time, although they counted
for. nothing in the structure of the verse.

In the first place, elision takes place before a strong punc-

13

94 M. W. Humphreys,

tuation, as (Birds 990) ovx ei Supad’; é¢ xépaxac xré. TKurip.
Androm. 459:
Kreivec pe. = axoKktev” wo aQwrevrov ye oe.
Sopu. Elect. 662:
rao toriv, w bey. avroc nxacag cadwe.
Cf. also 671, 1041, 1112, 1470, etc.
I am not disposed to attach much importance to elision at

the end of a verse, as in Birds 1716:

wpe, kador Jéapa’ Jupeaparwr 0

avpat dtaWaipover KrE.,

and Oed. Col. 1164:

gui ~aciv avTov é¢ Adyouc EASEiv podrovr’:

aireiy amwe\Seiy 7 acpadwe ric devp’ Ooov:
for I doubt whether this ever happens unless the sense requires
the verses to be closely connected together; and that being
the case, the two verses can be read continuously as one long
verse, and the vowel can be dropped.

Again, elision takes place between two speakers, as Birds
846, 1015:

EYEAT. otpwe wap ip’. THEO. 19, dyad’, of réurw a Eye.
MISO. pa rov A’ ob df’. MET. adda rac;
So Sopu. Elect. 1431:

0) eloupGre mov
rov avop; EA. 颒 Hpi ovrog éx xpoactiou—.
Ibid. 1502:
OP. add py’. AIT. vonyov. OP. coi Badeoreor xapoc.

Elisions of the sort just now mentioned—those at a full stop
not between two speakers—are comparatively rare ; for there
was something harsh about them; and although we may use the
interrogation point or the period, still the pause is really short
in most cases. That they were in some measure unpleasant
is shown by the fact that Isocrates, who did not tolerate hiatus
in his orations, also banishes this sort of elision—a thing
which he could do more effectually than the poets, who were
somewhat trammeled in the arrangement of their words by

On Elision, especially in Greek. 95

metrical considerations. But elision between two speakers
does not appear to have been avoided at all; for, in fact, it
was not a real elision. The second actor began to speak just
as the first one struck his last syllable. To this it may be
objected that the same thing could have happened just as
well, if the first speaker closed with a consonant; and this is
certainly true. But there would have been nothing to indi-
cate that it was to be so recited, and, as I have already said,
verses were so composed that, when written, they looked
perfect, which could not be done if the first speaker’s final
syllable had been disregarded when it was closed by a conso-
nant. We find something analogous to this effort to make
the verse appear perfect in the classic French drama, where,

without affecting its pronunciation, the mere spelling of a
word is sometimes altered, so as to make it look like the
word with which it rhymes, as Le Cid, v. 771, where voi (for
vois) rhymes with toi, and 851, where voi rhymes with mot.
Somewhat analagous is also the method of indicating a pause
at the end of a piece of music when the last measure is
incomplete.

In view of all this it is safe to assert that elision between
two speakers was relatively more frequent than at a full stop
in a speech of one person. (I say relatively, because this sort
of elision only has a chance to occur when a verse is divided
between two persons.) In fact it was not avoided at all, but
sometimes appears even to have been sought, as it gave one
actor an opportunity to fall in before the other had entirely
finished his last word—a thing to be desired when the dialogue
is animated, or for any reason rapid. If any one doubts this
let him examine such passages as Kurip. Orest. 1598-1612,
where in fifteen lines this elision occurs seven times.

_ These arguments prove conclusively that elision was some-
times only partial, and sometimes ayen only apparent, the
vowel omitted in writing being pronounced in full, but counting
for nothing in the structure of the verse.

§ 4. Although it was more especially designed to investi-
gate Greek elision in this paper, it will not be irrelevant to
append a few remarks on elision im Latin. It is conceded by

96 M. W. Humphreys,

nearly all that elision in this language was only partial.
Hence it was subject to less strict laws than in Greek, and
hiatus is more rarely admitted, it being easier to avoid it.
Indeed Cicero (Or. 44, 152), as quoted before, says that the
Romans were not allowed to neglect elision even if they
desired to; and then finds fault with poets for doing it. But
this very fact that poets did sometimes allow it, shows that,
the law Cicero announces was not without exception, and of
course it was no physical necessity, but merely convenience
and usage. Yet Cicero’s remark shows that in prose tt was
practically universal. Such words as neuter, deinde, etc., show
that elision was not total, at least, in some cases where there
was nothing to prevent its being total if it ever was; and
those cases where the vowel was entirely lost (tantopere,
magnopere, animadvertere, etc.) are mere results of long
usage, the vowel having been slightly pronounced at first, just
as we say “extrordinary”’ instead of “ extra-ordinary.”’ In
tantopere and magnopere this process was hastened by the
identity of the two vowels brought in contact, “‘ tanto-opere,”
as elision and contraction are more necessary under these
circumstances. This is illustrated by the Greek second
declension in the genitive plural, which contracted before the
existence of the law that a long ultima should prevent the
accent from falling on the antepenult; while the same con-
traction did not happen in the first declension until after this
period, the vowels not being so similar; thus Acyowr became
Adywy, while povcawy remained. Afterwards the long ultima
removed the accent, and then they said Adywr, povoawy (an
extant form); and finally poveadwy contracted into povedr.
Similarly nihil became nil, and mihi, mi; but we must not
carry the illustrations too far; for phenomena from within
words, simple or genuinely compound, will not always hold
for separate words; and tantopere and magnopere are not to
be regarded as genuine compounds, such as cogere, degere,
in which crasis seems to have been employed. (Corssen, by
the way, Ausspr. Voc. Beton. 11, 889, writes ‘ tantépere,”’ or
possibly his printer did it for him.) And so in mht, nihil,
the process was different from that in tanto opere, but they

On Elision, especially in Greek. 97

illustrate the aversion of the vocal organs to a consecutive
repetition of a vowel. But of genuine crasis between two
words, not combined into a genuine compound, I know of no
example in Latin. In fact, elision being only partial, and so
being allowable under almost all circumstances, there was no
need of crasis; and the nearest approach we have to it is what
we find in degere, cogere (just mentioned),.further examples
of which are déesse (two syllables), deerrare (three syllables).
in which Velius Longus (p. 2227) says the e or ee was long
(by nature), which of course we should expect, as the prepo-
sition sometimes formed a syllable to itself. Contractions.
such as amatast, integratiost, are not crasis, but a species of
uphaeresis, as is shown by Tibullus (1, 9, 03, and 77):
at te qui puerum donis corrimpéré’s ausus—
blanditiasne meas aliis tu véndéré’s ausus.

_ The later Roman grammarians speak of elision as if it were
‘a total expulsion of the vowel; but their authority is not of
any importance. The name “elision,” it is true, strictly
interpreted, would imply total removal; but the Roman gram-
marians employed terminology that was adapted to Greek,
and sometimes even mistranslated Greek terms. So we now
speak, and I have just been speaking of ‘‘elision”’ in Latin ;
and while doing so, I have been trying to show that it is not
elision, but diminution.

But there are good reasons for believing that the particles
-que, -ve, -ne lost their vowels entirely through elision; and
-ne is sometimes written without its vowel even before a conso-
nant; just as face, duce, dice lost their e, and even cave (being
much used) lost its e sometimes, as shown by Cicero’s well-
known remark implying similarity of sound between caunéas
and cave ne eas. The elision of these particles will come up
in my next paper.

Briefly, then, to sum up the whole matter:

1. In Greek, elision was the total suppression of a vowel;
but it could be only the partial suppression, and sometimes was
required to be only partial, or even merely apparent.

2. In Latin, elision was the partial suppression of a vowel ;

but in a few special instances it was total.
14

American PaitoLocicaL Assoctation,

1877-8.

MEMBERS IN ATTENDANCE AT THE TENTH ANNUAL SESSION.

N. L. Andrews, Madison University, Hamilton, N. Y.

H. C. G. Brandt, Johns Hopkins University, Baltimore, Md.
Charles J. Buckingham, Poughkeepsie, N. Y.

Henry F. Burton, University of Rochester, Rochester, N. Y.
William C. Cattell, Lafayette College, Easton, Pa.

B. L. Gildersleeve, Johns Hopkins University, Baltimore, Md.
W. W. Goodwin, Harvard College, Cambridge, Mass.

8. 8S. Haldeman (University of Penna.), Chickies, Pa.

Albert Harkness, Brown University, Providence, R. I.

Charles R. Hemphill, Theological Seminary, Columbia, 8. C.
M. W. Humphreys, Vanderbilt University, Nashville, Tenn.
A. C. Kendrick, University of Rochester, Rochester, N. Y.

F. A. March, Lafayette College, Easton, Pa.

A. C. Merriam, Columbia College, New York City.

Edward North, Hamilton College, Clinton, N. Y.

C. K. Nelson, St. John’s College, Annapolis, Md.

J. Sachs, Classical School, 649 Madison Ave., New York City.
A. Duncan Savage, Johns Hopkins University, Baltimore, Md.
William C. Sawyer, Lawrence University, Appleton, Wis.

T. D. Seymour, Western Reserve College, Hudson, Ohio.

J. B. Sewall, Thayer Academy, Braintree, Mass.

L. A. Sherman, Hopkins Grammar School, New Haven, Conn.
W. A. Stevens, Rochester Theological Seminary, Rochester, N. Y.
C. H. Toy, Southern Baptist Theological Seminary, Louisville, Ky.
J. Hammond Trumbull, Hartford, Conn.

AMERICAN PHILOLOGICAL ASSOCIATION.

0 @ o___—_

Saratoaa, N. Y., Tuesday, July 9, 1878.

The Tenth Annual Session was called to order at 7.30 o'clock
P. M., in‘ the audience room of the Opera House of the Grand
Union Hotel, by the President, Professor B. L. Gildersleeve, of
Johns Hopkins University, Baltimore, Md.

The Secretary being absent, Professor W. A. Stevens, of Roch-
ester Theological Seminary, Rochester, N. Y., was appointed Sec-
retary of the meeting.

The Treasurer, Charles J. Buckingham, Esq., presented his
report, showing the receipts and the expenditures of the past
year. [See p. 31.] .

On motion, President William C. Cattell and Professor M. W.
Humphreys, were appointed Auditors of the Treasurer’s report.

The Secretary presented a report from the Executive Committee,
announcing the election to membership of,

Rev. T. T. Eaton, D.D., Petersburg, Va., and Mr. 8S. E. W. Becker,
Wilmington, Del.

Professor C. H. Toy, of the Southern Baptist Theological Semi-
nary, Jouisville, Ky., read a paper on “The Yoruban Language.”

The Yoruban language, spoken by a partially civilized people living near
the western coast of Africa, east of Dahomey, belongs to an isolated
linguistic group, which shows little or no resemblance to the great Hot-
tentot and Bantu families in the south, or the Negro and Berber dialects
on the north, but is nearly akin to the Grebo and other dialects of Liberia.
The literature consists of collections of proverbs and Bible-translations
made by Christian missionaries. Grammars have been written by Crow-
ther (London, 1852) and Bowen (Smithsonian Institution, 1858).

I. Phonology. The phonetic system consists of letters and tones. The
consonants are eighteen in number, namely: the aspiration h, of palatals
the surd &, the sonant g, the nasal ng, the semi-vowel y, of linguo-dentals
surd ¢, sonant d, nasal n, semi-vowels 7, /, and the compound linguo-sibilant
j (dz), of labials surd p (only in the combination Xp), sonant 0, spirant /,
nasal m, semi-vowel w, of sibilants s and s (sh), to which might be added
z (zh) in dz=j. There are no gutturals. The combinations dz (J), xp,
gb and md occur frequently; of these the first may be a weakening of g or
d, mb (only in mbe and mbi) probably comes from word-composition, and

4 Proceedings of the

the origin of the others is doubtful. The vowels are the three primitive
a, t, u, the secondary ¢, e (as in /et), 0, o (as in not), and the diphthongs ai,
é7, ot, 6¢, and apparently aw (in a few words, mostly adverbs and probably
compounds). There is a law of vocalic harmony (partially prevalent) by
which vowels of personal pronouns and prefixes are made to accord with
those of verbs and roots. In its law of tone, whereby words spelled alike
are distinguished in meaning, the Yoruban stands to some extent on the
same plane with the Chinese.

II. Morphology. 1. Roots and Words. The roots are probably all
monosyllabic, most polysyllabic forms easily resolving themselves into
simpler elements. The word is not differenced in form from the root, and
there is no essential difference in form between Noun and Verb, tho a
partial difference is made by the system of nominal prefixes; thus, there
is not a verb of less than three syllables (that is, not compound) beginning
with ¢ in the language, ¢ being a noun-prefix. Roots consist of a single
vowel, or consonant and vowel (with or without nasal appendage); a few
words beginning with two consonants are probably compounds. 2. Word-
composition. Composite noun-forms are made by derivation, by redupli-
cation, and by composition proper. The derivation is by prefixes only
(a, €, €, t, 0, O, abt, att, etc.), and the language is rich in these forms, as
oku ‘‘corpse” from ku ‘‘to die,” atilo ‘‘a going” from lo ‘‘to go,” ese
‘*sin,” lese ‘‘to have sin,” elese ‘‘sinner,” dese ‘‘the state of having sin,”
ailese ‘‘sinless,” and sevcral others from se ‘‘to sin.’”’ These prefixes
were probably originally independent words, and the derivation is true
composition. There are compound verbs, made up of verb and verb, or
verb and preposition. 3. Inflection. The Yoruban may be called semi-
inflecting, there being a number of agglutinations that have more or
less lost their independent character. (1) Nouns are without inflectional
marks of gender, number, or case. Gender is sometimes denoted by
prefixed sex-words (as in English), as ako, ‘‘ male,” abo ‘‘female” (from bi
‘‘to beget”); thus: akomalu ‘‘bull,” abomalu ‘‘cow.” Case-distinction
is marked by position (as in Hebrew), or by the Relative Pronoun &, equiva-
lent to ‘‘ of” (as in Aramaic). The comparative and superlative degrees
of Adjectives are made by the affixes ju and julo respectively (= ‘‘ beyond,
more”), (2) Verbs show no.distinction of gender, number, or person, nor
in themselves of time, completeness or modal conception, and there:are
no derived stems (Causals, etc.), as in Hottentot, Bantu, Woloff. But
temporal and modal distinctions are expressed by prefixed verbal or pronom-
inal words, continuous action by 7 (probably the substantive Verb v2) and
sometimes by ma (perhaps = *‘do”—so in Basa and Grebo), past time and
completeness by %@, future time by 0 or yio (the origin of the yz is doubtful;
it may be the pronoun yi = ‘‘this,” or a verb = ‘‘turn, revolve,” as Basa
dyi, Grebo di ‘‘come,” mi ‘‘ go,” yt ‘‘purpose’’). In the expression of
modal conceptions the Yoruban stands about on the same plane with
modern English — certain agglutinations have acquired modal significa-
tions: ba is employed in the protasis of conditional sentences involving
uncertainty, and it has such agglutinations expressing obligation, desire,
permission, ability, and the like (as English shall, must, will, may, can).

American Philological Association. 5

Participles are made by prefixing 7 to the root, or by abstract substantives
(as English ‘‘I go a-fishing”); a Passive is formed by the indefinite
construction with Pronoun @ or non ‘‘ they,” as afo o ‘‘ they broke it” =
‘‘it was broken,” or by the use of the substantive verb nz with a redupli-
cated noun, as 7irt (from ré ‘‘to see") ni baba ‘father is seen.” The
language shows a primitive exuberance of substantive verbs, which, how-
ever, are distinguished in use. The simplest form is 2é (before vowels @2),
perhaps connected with nz ‘‘to have,” and frequently employed in a merely
emphatic way; mbe (from bi “ beget”) and wa make prominent the idea of
existence; ri, si, ya express modal being; there are others less definite.
(3) The pronouns are without the rich generic development of the Hot-
tentot, and are in other respects flexionless. Personal: Singular, emi (mo,
mo, ng), ro (0, 0), on (0, 0), Plural awa, enyin, awon (nwon), with the
objective forms mi, 0, a (¢, €, t, 0, O, #4), wa, rnyin, won. Besides the law
of vowel-harmony above-mentioned, there is this difference in the forms
of the first person singular, that ng is used only with future verbs.
Emphatic forms are made by adding na, and reflexives by prefixing ka and
ara. Demonstrative: na, ni (wonnt), yi (eyi, eyiyt). TInterrogative: ta, t,
ni, ki,wo. Relative: & The preponderance of the simple dental, palatal,
and labial forms (found in most languages) is obvious, and the original
identity of all the pronominal forms may be regarded as probable, but
nothing further as to their origin cau be said. Their particular uses must
here be passed over.

III. Syntax. The syntax is very simple. The usual erder of werds in
the simple sentence is: subject, copula, predicate; the attributive adjective
(or pronoun) usually follows its substantive; the substantive verb as mere
copula is generally omitted. The order in relative clauses is the same.
Relational particles are few; various parts of a coemposite sentence are
commonly regarded merely in the relation of temporal sequence, as in
Hebrew, whereby a naive and vivid coloring is given to narrative and
proverb. Purpose is expressed by an abstract noun alone, when the sub-
jects of the principal and dependent verbs are the same; when they are
different, the dependent clause is introduced by ké. Conditional protasis
is introduced by 6! = (‘‘if”), or by a (= “obligation ’) followed hy jepe
Or sepe, as iba, jepe (or, sepe) emi ni, ‘‘ had it been I,” literally: ‘ obligation
that it is that Iam.” Substantive clauses, as subject or object, are intro-
duced by «(probably the Relative Pronoun).

Remarks were made upon this paper by Professor S. S. Halde-
man, of the University of Pennsylvania, Philadelphia.

Professor F. A. March, of Lafayette College, Easton, Pa.,
Chairman of the Committee, appointed in 1875, ‘tu take into
consideration the whole matter of the Reform of English Spelling,”
and continued after its Reports in 1876 and 1877 [see Proceedings,
1875, pp. 8, 13, 23; 1876, pp. 35, 36; 1877, pp. 30, 31], presented
the Report of the Committee, as follows:

In accordance with the plan of preparing a list of words for which an
amended spelling may be adopted concurrent with that now in use, as

6 Proceedings of the

suggested by President J. Hammond Trumbull, at the session of 1875,
and favorably reported upon by the committee of that session, the com-
mittee now present the following words as the beginning of such list, and
recommend them for immediate use:

Ar. Giv. Tho.
Catalog. Hav. Thru.
Definit. Infinit. Wisht.
Gard. . Liv.

On motion, it was

Resolced, That the Report of the Committee on the Reform of English
Spelling be accepted, and the Committee be continued for one year.

Professor 8S. S. Haldeman, of the University of Pennsylvania,
then read a paper “On Virgil’s Hexameters.”

This paper presents a discussion of the principles of the Hexameter,
illustrated by an English version of a hundred lines from the opening
of the Aneid, in which an attempt is made to present every syllable
and every natural or prose accent of the original, presenting a specimen
sufficiently long to familiarize the listener to the nature of the versification,
if he is not so much accustomed to the powerful accent of German or
English as to prevent him from appreciating the lighter effect of a rhythm
of quantity, which seems proper only in languages where the accents are
light. Even if the classic accent was, as some believe, a change of pitch
rather than of force, the stress which the derived languages exhibit at
the points of accent, indicate at least a concomitant stress originally.

The pronunciation of words is essentially the same in prose and verse,
and there is sufficient evidence (Quintilian, Priskian, Donatus) that Latin
differed from Greek in being without final accent, so that CAN’o could
not be pronounced CANO! in the first line of Virgil. But while the quan-
titative nature of classic rhythm is admitted, the average speaker with an
English vernacular knows so little about quantity, that he will assert that
in the pairs fate fat, deep dip, note not, three are long and three short,
where all require the sume time, and knowing no rhythm but that of
stress, the character of his own emphasized verse is intentionally or
unconsciously forced upon classic examples, and there are Latin-English
grammars in which the beginning of every foot is marked with an accentual,
instead of confining it to the last two, which enable the listener to deter-
mine the metre.

The most prominent feature of the hexameter line is the two closing
feet of a dactyl and spondee differing from the preceding feet in having
their natural accent at the beginning. In the first half of the line, the
accents may occur at any point in the foot, or a single foot may have
two natural accents, and to connect these with the two final feet (the
adonic close), Virgil endeavors to interpose a fourth neutral foot which
shall be without natural accent, as in bk. 1, 1. 8—

Mii’s& mi’|h! cavs’|as mém’djra qvo | na’miné | 1x’so |
where the fourth foot is neutral.

American Philological Association. T

In efforts to bring Latin and English prosody into correspondence, and
to get rid of the detinite statements of the ancient grammarians, Professor
Key says they ‘‘ were dealing with a language which was already dead;”’
and Richard Roe asserts of the ancients that ‘‘there is reason to believe
that their perceptions of quantity were confused and imperfect” !

A few lines would enable a Greek or Roman listener to detect hexameter
verse, a test which fails, not only with the spurious English caricatures,
but with the ordinary English heroic measure, where, in many cases, the

_Supposed five-foot line is equally of four or of six feet, and when rhyme
is rejected, the close of the lines must be indicated by non-metric methods,
such as punctuation and syntax. Take, for example, the opening of
‘* Paradise Lost,” when it will be found that the first line may be broken
at several points without injury to the rhythm, which is rather that of
rhythmic prose than of a given metre—

Of Mans
First disobedience, and the fruit of that

Forbidden tree, whose mortal tast brought death
Into the world, and all our woe, with loss....

And the fruit of that forbidden tree, whose
Mortal tast brought death into the world, and
All our woe, with loss of Eden, till one... .

Whose mortal tast brought death into the world
And all our woe, with loss of Eden till
One greater man restore us, and regain... .

Mans first disobedience, and the fruit of
That forbidden tree, whose mortal tast brought
Death into the world, and all our woe, with....

Mans first disobedience and the
Fruit of that forbidden tree, whose
Mortal tast brought death into the....

—which resembles a line of ‘‘ Evangeline” (2:249)—

Louis|burg is | not for|gotten, |
nor Beau | Séjour, | nor Port | Royal. |

Ovid's line (Metam. bk. 6, 1. 451)—
ecce venit magno dives philomela paratu;
may be thus imitated in English—
she is coming, cloth’d | with pru|dence, Philo|mela the | careworn | —
but this will be likely to strike the English ear as seven accentual feet—
she is | coming, | cloth’d with | prudence, | Philojmela the | careworn. |
The following lines (87-91) are selected from the paper—

Then follow men’s loud cries, an’ echoes are heard from the cordage.
Clouds quickly bédim’ the expanse from all eyes of the Teucri,

and Nox fuliginose incibates—broods on the high sea.

Poles of the orb have thinderd, and air carries numerous lightnings;
all things thréaten {nstant and widespread death unto mankind.

8 Proceedings of the

The author gave a version of 33 lines (1-33) in the ‘‘ Literary World,”
New York, Nov. 6, 1852; of four lines in his ‘‘Analytic Orthography,”
1860, § 646; and of fourteen lines in the article HEXAMETER of ‘‘ Johnson’s
Cyclopedia” New York, 1876.

The Association thereupon adjourned to 9 o’clock Wednesday
morning.
Wepnespay, July 10—Mornina Session.

The Association resumed its session at 9 o’clock a. m., the Presi-
dent in the chair.

The minutes of the previous session were read and approved.

The Secretary presented a report from the Executive Committee,
announcing the election to membership of,

Professor Henry F. Burton, University of Rochester, Rochester, N. Y.
On motion, it was

Resolved, That the Association approve of the list of words reported by
the Committee on the Reform of English Spelling, as judiciously selected _
for the purpose mentioned in the Report.

On motion, it was

Resolved, That a committee be appointed by the Chair to recommend a
suitable time and place for the next meeting.

The President appointed as such committee, President William
C. Cattell, Professor W. W. Goodwin, and Professor C. H. Toy.
On motion, it was

Resolved, That a committee be appointed by the Chair to nominate
officers for the next year.

The President appointed as such committee, Professor A. Hark-
ness, Professor S. S. Haldeman, and Dr. J. H. Trumbull.

Professor H. C. G. Brandt, of Johns Hopkins University,
Baltimore, Md., then read a paper on ‘“‘The Roman Alphabet in
German.” *

Ther is a strong movement in Germany for the establishing of the
roman alfabet. The Berlin conference, on orthografy, of 1876, did not
do its duty towards it. The question coms home to us in this country
with our german-english schools, on account of the extensiv study of

* Printed according to the five following rules of the Spelling Reform Association. The
capitals of proper adjective ar dropt:

1.—Omit @ from the digraf ea when pronounst as e-short, as in hed, helth, etc. 2.—
Omit silent ¢ after a short vowel, as in hav, giv, etc. 8—Write / for ph in such words as
alfabet, fantom, etc. 4.—When a word ends with a donbl letter, omit the last, as in shal,
clif, eg, etc. 5.—Change ed final to ¢ where it has the sound of ¢, as in lasht, imprest, etc.

American Philological Association. 9

german, and our intimate relation with german scholars and german
learning. The movement gains not smal momentum from the objection
of foreigners to the old german type. It goes hand in hand with ger-
man spelling reform. While we in this country cannot move faster in
regard to the latter than the fatherland and certainly not so far as radical
fonetic spellers, we ought to do al in our power to help on the former.

We must insist upon it that the use of the roman alfabet is a re-in-
troduction of it, not an innovation. The socald gothic character is
the misshapen roman character. It became angular in the twelfth and
thirteenth centuris, partly through the influence of gothic architectur
with its straight lines and pointed arches. Unfortunately printing was
invented when this changed form was the prevalent one and the type was
shaped after it. The first books in Germany, Italy, Holland, Spain, and
England wer printed in it. In Italy, first of al, it was dropt and the
pure roman character resorted to, owing largely to the round writing of
the classical manuscripts.

In the sixteenth century, the latin classics wer printed in latin type
in every country that could boast scholars. Popular books like the
bible wer printed in gothic; in Italy not even these, in France to a smal

extent. The Dutch and English dropt it entirely in the seventeenth .

century. Germany, Denmark, Sweden, and Bohemia stil retain the gothic
type or german, as it is deservedly and fondly cald.

Originally it was no more german than it was dutch or english or
italian. It is german now, because Germany is the only nation of the
first rank that stil clings patriotically, as she thinks, to an old abus.

For international reasons, Germany should establish the roman alfabet.
It is the only international alfabet in the world. German scholars, of
whatever branch of learning, print their publications — at least ninety out
of a hundred — in latin type. And why? Becaus the results of german
scholarship ar red the world over and latin type makes them more
accessibl and palatab] to foreigners.

Ther ar certain drawbacks which hamper the rapid strides of this
movement. The german books in latin type ar not printed alike. Ther
is great irregularity in the use of the capitals, and ther ar five different
signs for german sz ({).

As to capitals, som leav them, as in german type, to every noun.
Som reject them entirely except in the word beginning a paragraf, and
som reject them in al nouns except proper names.

To bring about uniformity, the following ‘‘ rules” are proposed, which
also include a few changes in spelling, generally approved of:

1. Drop the capitals of nouns, excepting proper names.*

The french system is preferabl to the english. Hence, do not
use capitals for adjectivs derived from names of persons, places, and

countries, for the names of the months, the days of the week, the
points of the compass.

e.g. preussischer grenadier, hollindische heringe, ostfriesische butter,
bairisch bier.

“* To giv capitals to the personal and possessiy prononns in address and letters is a
matter of etiquette.

10

Proceedings of the

‘« Der erste tag im monat mai

Ist mir der gliicklichste von allen.”
Hagedorn.

‘‘Am sonntag bet’ und sing, am werktag schaff’ dein sach.”’

2.

Write s for { and 8; ss for ff, {8, £.

e.g. ‘‘Mit musse kommt man auch fern.”
‘‘ Er muss das mus essen.”’
‘Grosse kinder, grosse sorgen.”
‘* Mass ist zu allen dingen gut.”
‘‘ Massig wird alt, zuviel stirbt bald.”

Ther being now no consistency in the use of {8, A fs, g, g, nothing
would be gaind by retaining the awkward sz. e signs {s g, g.
used to som extent in german books in roman type, should be
discarded, as other nations do not use them. They ar only so many
more signs for the same sound—voiceless s—for which ther ar two
alredy, ss and s final. If it wer the office of the consonant to
indicate the quantity of the preceding vowel, one of the abov signs
might be used after a long vowel and ss after a short one, as the
Berlin conference proposed. The diacritical mark [-] over the
vowel would indicate that much better.

. Drop d in the adjectiv tebdt and all its derivativs. Hence, tot, toten,

der tote, totschlagen. But der tod, todkrank, todfeind, todfehde,
todsiinde.
Embdte and gefebeidt occur only rarely now for ernte and gescheid.

. Drop h after t,—e. g. Thurm, turm; Yeirath, heirat; Bithfel. ratsel; suffix

thum, tum, as kénigtum, reichtum.
Retain h, however, in borrowed words, e. g. kathedrale, katheder,
athlet.

e.g. ‘‘Doch bin ich auch nicht der, der alles, was
Er tat, als wohlgetan verteid’gen mochte.”

. Write a sing] consonant in the aftixes —nif} or nif8, -inn, oig— or mifé-,

e. g. Regriibnig, begribnis; Wifg- or Wifsbranch, misbrauch; Bonigina,
konigin. An added vowel restores the double consonant, e. g.
gefangnisse, freundinnen, missetat.

The Berlin conference excepts Wiijé-, because it is a ‘‘ stammsilbe.”
But it is difficult to see, why mis— is not as truly an ableitungssilbe
as -nis. Besides —-nif8 does not lose one s because it is a derivativ
syllabl, but becaus s is superfluous.

. In deff or def8, feB and their derivativs, in borrowed words ending in §,

or j8 write only one s:—e. g. des, wes, indes, deshalb, desfalls,
weshalb, compas, atlas, firnis, kiirbis, as, kiras. An added vowel
restores the doubl consonant as in 5:—e. g. des compasses, atlasse,
kiirasse, dessen, wessen.

. When in compound words the same letter would occur three times,

drop one,—e. g. Betttnch, bettuch, Schiffjabrt, schiffahrt ; Stammanatiter,
stammutter.

. Drop one] in @allfijey, Wallnaj, Wallrog, and write after the analogy

of himbeere, damhirsch, etc., walfish, walnuss, walross.

. Write the foreign infinitiv ending —ieren uniformly —iren, e. g. studiren,

turniren, einquartiren, regiren, spaziren, barbiren.
The Berlin conference wants to write still regieren and spazieren,
, because it is “‘ablich,” and barbieren, turnieren, because ther ar
nouns, barbier, turnier, But why not write all alike —iren?

American Philological Association. 11

10. Words are divided into syllabls, in general according to pronunciation,
somtimes according to etymology, and if compound according to
composition,—e. g. be-sainf-ti-zgung, ii-ber-ein-stim-mung, wand-te,
reis-ten, du ris-sest, un-gern, da-rin, ras-ten (raged), wes-pe, has-pel,
hat scheln, wach-sen, haus-tiir, ret-ter, was-ser.

The consonant combinations ch, ck, ph, sch, st, tz, ar inseparabl—
e. g. ma-chen, we-cken, so-pha, wa-schen, ra-sten (rest), ka-tze,
schwatz-te, wach-te.

As to rule 1, which concerns capitals only, it may be remarked that
they wer at first used only for the beginnings of paragrafs and pages. In
the earliest printed books space was left for the capitals, which wer
added by the hand of the illuminator. When they came to be printed
with the rest, they became very common, especially in Germany.

At first proper names receivd them, then appellativs, then neuter and
abstract nouns. In the seventeenth century every noun, as is now the
case. It is proposed to limit them to the word beginning a sentence and
to proper names, because all nations, except the english, using the roman
alfabet, do so for the very good reason, probably, that a more extended
use of them is of no advantage.

A few changes in spelling are added, because in the change of type
they will find a more ready acceptance than in the old alfabet.

Remarks were made upon this paper by Dr. Julius Sachs,
Professor S. S. Haldeman, Professor M. W. Humphreys, and
Professor W. C. Sawyer.

Professor W. C. Sawyer, of Lawrence “University, Appleton,
Wisconsin, next read a paper on ‘‘Some Contributions of the
Phonograph to Phonetic Science.”

The Phonograph, though but an indistinct talker as yet, converts
audible into visible form with such marvellous exactness that the latter
can be reconverted into the former. This affords the basis of a new
demonstration of the compound character of the @ of fate. Long a
distinctly and forcibly pronounced into the mouth-piece of the phonograph
yields, besides the fine and confused indentations upon the tin-foil at the
beginning and end of the utterance, two distinct series of uniform groups
of indentations. The first is the series representing the principal element
of the letter. It is composed of miniature human tracks, the hollow of
the foot being quite plainly marked in every other one, and altogether
wanting in those which are intermediate. These are followed by impres-
sions which change gradually into uniform groups of three indentations
each, of which the first is long and the second and third are round. The
space between the long mark and the round ones is about twice as great
as that between the round ones themselves. The length of the long
impression is equal to about two-thirds of the space occupied by the two
round dots together with the interval between them. In the tin-foil
examined, the first series is about twice as long as the second, the groups

12 Proceedings of the

in each averaging about seven to the inch. It is evident that the phono-
graph cannot give the same result if the needle passes backward over the
impressions described as when it passes forward. If we get the sound
ei (Italian) when the needle passes one way, we ought, on @ priori
principles, to get ze when it passes in the opposite direction. So long as
like causes produce like effects, a simple and uniform vowel sound cannot
produce two such distinct series of impressions as I have described, nor
can the phonograph produce a single uniform sound from two series of
indentations that are so different one from the other.

With this invention we enter upon a new era in phonetic and orthoépic
science. A dull ear, even more than an unconquerable conservatism,
holds the spelling reform in check. An exact symbol implies a definite
conception of the thing symbolized, and, until scholars can agree con-
cerning the elements of our specch, all systems of notation must remain
crude and ill adapted to the purposes of culture. It is vain to shout, “A
sign for a sound,” till we discover for what precise sounds signs are
required. The phonograph, however, comes to the aid of this reform at
the point of its sorest need, and, as illustrated above, brings the eye to
the help of the ear in so effective a manner as to promise the settlement
of all, or nearly all, our phonetic disputes.

I venture here a query upon which I dare not yet express any opinion
of my own: ‘‘ May not the essential forms produced by the phonograph
under the impulses of the voice in articulate utterances be advantageously
substituted for our barbarous alphabet? ”

Numerous unforeseen difficulties may arise, but those which now appear
are not insuperable. We are told, for instance, that we cannot even read
the writing of the phonograph, on account of the slightness and constancy
of its variations. Its impressions and traces must resemble each other
precisely as much as its utterances,—an invaluable correspondence—and
when we enlarge the traces of the stylus upon a surface easy for the eye,
we shall see that they differ as much In form as our vowels and consonants
differ in sound. Variations which correspond to the peculiar overtones
of individual voices may be neglected, since they are as unessential to
distinctness of writing as the latter are to the distinctness of speech. It
may be thought that the writing of the phonograph is too complicated
and difficult of construction to be adaptable to general use for script and
print. Against this objection two considerations tend to reassure us:

1. The profile of the depressions and elevations upon the foil is a
continuous curve, easily traced, and probably contains everything essen
tial to the writing.

2. <A brief section of the curve corresponding to a single sound in the
phonograph—for instance, the curve covering the two dots and dash
composing one of the groups described above—would sufficiently define
that sound.

I will allude to one more possible benefit that the phonograph may
confer upon linguistic science. The pronunciation of foreign languages
is wretchedly taught in the great majority of our schools, even of the
highest grades. When the phonograph is brought to perfection, the

American Philological Association. 13

voices of the best orators and orthoépists of all living languages may be
heard in our class-rooms, and, if Mr. Edison himself has not over-
estimated the possibilities of his invention, the children in our homes
may, in their most susceptible years, huve their very toys so selected that
they shall acquire considerable familiarity with colloquial French or
German, or both, in the time devoted to play.

The phonograph is still too immature, and my own study of its results
quite too slight, to indicate the real value to philology of this invention;
but I seem to see a clear promise that some of its best fruits will fall
within our province.

Remarks were made upon this paper by Professor S. 8.
Haldeman.

A paper entitled “Observations on Plato’s Cratylus” was then
read by Dr. Julius Sachs, of New York City.

The Cratylus gives a resumé of the theories, prevalent in Plato’s time,
on the relations between thought and speech; to appreciate Plato’s own
views is rendered difficult both by his style and the meagre knowledge we
possess of contemporaneous philosophical speculation, hence the divergent
interpretation put upon the Cratylus by modern critics. Some salient
points, however, susceptible of common acceptance, the works of two
recent critics seem to contain, Benfey’s ‘‘ Uber die Aufgabe des Platonisch-
en Dialogs Kratylos,” and Dr. Herman Schmidt's ‘‘ Plato’s Cratylus im
Zusammenhange dargestellt.” The vindication of Plato’s authorship for
the dialogue seems fully carried out by Benfey; discarding the traditional
speculative fancies on the ‘‘underlying meaning” and connection of the
various parts of the dialogue, as Steinthal, Steinhart, etc., have elaborated
them, Benfey has evolved, with less brilliancy perhaps, but with more
trustworthiness, the exact meaning of the technical terms used. A care-
ful arrangement of the various grades of meaning that the word fuvfyxn
shows throughout the dialogue, and in like manner a concise discrimina-
tion between Plato’s use of the various verbs, indicating ‘‘ thought,”
between the terms évoua and pjua are among the merits of Benfey's work;
the instincts of the student of Comparative Grammar have confirmed by
many valuable suggestions the continuity of reasoning in the dialogue,
notably so in 388, B., where the perception of the original root da in
d:ddoxecv affords the justification for the transition from dcaxpivev to
diddoxecv. Language, as it exists, cannot be correct either gtoe or Géoes;
an ideal language only might be constructed conformably to these prin-
ciples, and whatever correctness of appellation actual language shows
forth, is purely accidental.

Dr. J. Hammond Trumbull, of Hartford, Conn., presented
papers on ‘The Name Oregon,” and on “ The Inflections of the
Micmac Verb.”

The last paper of the morning session was read by Professor

14 Proceedings of the

M. W. Humphreys, of Vanderbilt University, Nashville, Tenn.,
on “The Influence of Accent in Latin Dactylic Hexameters.”

The observance of accent in dactylic verse is not due to the same cause
that it is in iambic verse. In the latter, the shortening and weakening of
the thesis (¢. ¢. dpoc) rendered certain relations of ictus to accent
unpleasant on their own account; whilst in dactylic hexameters, where
the thesis retains the full time of the arsis, there was originally no
influence of accent at all, as is shown by comparing Ennius with Homer
read with Latin accent. But the mere form of the verse caused accent and
ictus generally to fall together in the last two feet, and to come in conflict
in the earlier feet. In the course of time this conflict or strife, followed
by the agreement or reconciliation, was regarded as a peculiarity of the
verse, and any verse not presenting it, seemed unusual and harsh. As we
advance from Ennius to Ovid, we find each kind of poem becoming
more and more carefully composed in this respect, but in such a way that
one sort of composition may be rougher at a certain period than another
sort at an earlier period. Thus, the Satires of Horace are not so carefully
composed as the De Rerum Natura of Lucretius; and the Elegies of
Catullus are more carefully written in this respect than even the Epistles
of Horace; but they are not so carefully written as later Elegies.

In the following table are found some of the principal results of an
examination of the Roman poets. The table is the average for every
1,000 hexameter verses, but does not in every case profess to be absolutely
exact, as it is derived from only a partial reading (except for Ennius,
Lucilius, Virgil, Catullus, Tibullus, and Propertius).

i
Z i
E16
Caesurae after fifth arsis.
(1) With conflict, . 3; 0
(2) Without conflict, 12 | 10
Caesurae in sixth foot.
(1) With conflict, . i o
8

(2) Without conflict,

It will be observed that the more perfect the art of composing became,
the more the caesurae themselves were avoided in the last two feet, even
when there was no conflict. This was, first, because in avoiding conflict,
they unconsciously avoided that which was likely to lead to conflict; and,
secondly, because the ending | “ | ~~ -. ~ was harsh on account of
a monosyllable being followed immediately by so long a word, and the
ending | _ | ~ ~ | - = really caused a sort of conflict between the fifth
ictus and the accent in the fifth thesis.

With regard to monosyllabies at the end of a verse, it is to be observed
that they generally do not cause surprise. Even in ‘‘ridiculus mus,” the
meaning of ‘‘ ridiculus ” leads us to expect something like ‘‘ mus.”

American Philological Association. 15

This monosyllable, when immediately preceded by another, is much
employed to force the reader to connect two verses closely, when the sense
requires. The accent and stress of the monosyllable prevents the voice
from falling and pausing. (See Hor. Sat. passim).

The varied relation of ictus to accent in the two hulves of the verse is
made use of frequently to produce a pleasant play upon ictus and accent
by repeating the same words with their relation changed, as Catul. LXIT,
20-22:

Hespere, qui coelo fertur crudelior ignis?
qui natém possis complécu avellere mdtris,
cémplerié matria retinentem avellere ndlam.

(See also Virg. Buc. VIII, 47-50, and in the poets generally.) Many
special points, not contained in this abstract, were discussed.

The Secretary presented a communication from the Secretary of
the American Anthropological Society, inviting the correspondence
and codperation of this Association.

On motion, it was

Resolved, That the Secretary be instructed to acknowledge and reply to
this communication.

A recess was then taken till 4.o'clock Pp. M.

AFTERNOON SEsSSION.—4 P. M.

The Association met after the recess.
Mr. A. C. Merriam, of Columbia College, New York City, read
& paper on “The Homeric @éAo¢.”

This was an argument against the critical canon that g/4o¢ in Homer is
used as a synonym for the possessive pronoun. The first three books of
Anthon’s Homer being most widely used in this country in preparation for
college were brought under review. Everywhere in the translations there
given giAoc is rendered by the possessive pronoun, except four times in I.
This destroys in a great measure the tone of the passage in the episode of
Chryseis, A 20, 98, 441, 447; in that of Hephaistos, A 572, 578, 585, 587;
of Patroclos, A 345, where the keynote of the later books is struck; and
of the patriotic sentiment B 158, 162, 174, 178, T 244. With like strictness
the rule is followed by Derby, Pope, Sotheby; less constantly by Herschel
and Merivale: while Bryant, Blackic, Newman, Chapman, and the Latin
version of the Ernesti-Clark edition, respond almost uniformly to the
feeling of the poet. The scholia B. L. on A 20, have éAceea dé 4 xpoobixy
tov giAyvy, Autenrieth, accepting a derivation from ode, makes the original
meaning possessive; but the derivation of ¢i20¢ is too doubtful to deter-
mine its meaning.

¢iAoc occurs in Il. and Od. between 500 and 600 times. As adjective not
predicate or vocative with proper name, its usage may be divided into two
classes. In a count only fairly exhaustive, 287 examples were found for

16 Proceedings of the

first class, 112 for second. Of the former number zaryp claims 30, pirnp
11, radrra 1, réxvov 9, réxoc 16, vidc 69, maic 14, xaciyvytor 8, Exupd¢ 1, pATpUs
1, G@xatic 2, axoirng 2, xovpidtog 1, GAoxor 12, réore 8, praia 5, trpoddc 6, vixge 1,
Eraipoc 84, xegdAn 2, Seivoc 4, avip B, yépwr 2, Gedc 1, wAdea 1, yaia, aia, warpic
45. None of these words give offence in the English when connected with
dear, and this translation is forced upon us when a limiting genitive or a
possessive pronoun qualifies the same noun. This with vidc very common;
’Odvaozo¢ giAog vidc at least 14 times. See also A 354, B 260, x 222. Simi-
larly N 427, 0 689, M 355, B 564, 713, H 44, ¥ 289, K 50, » 259, 7 455, = 502,
B 51, 1455. The possessive pronoun E 314, 318, T 182, Z 474, IT 447, T 4,
@ 330, 378, o 214, » 505, Q 416, £ 177, v 418, 7 350.

The remark of Liddell and Scott on I 555, was met by showing that
giAy meant once dear, normally dear. Compare épateivd 218, oryaddevra
{ 26, pidov Aesch. Choe. 616, 4 327, waxdpioc Eur. Ores. 4, lébero Hor. C. 3,
5, 22. Homer feels that the ties of affection between parent and child can
never be destroyed. The evidence of the strength of the domestic affec-
tions in all the relations of life is cumulative throughout the poems. See
Glad. Juv. Mund., pp. 896-8. Are we not then actually wronging Homer
and the spirit of the Heroic Age by nullifying epithets, which, rightly
understood, open to fair fields in that distant past?

The second class comprised 6vudé¢ found 12 times, aidv 1, Frop 48, xjp 12,
orjGoc 8, Aatude 1, xeip 9, yotvara 7, yvia 8, BAédapa 2; also eivara 1, dpa 3,
yépag 1, déua 1, oixia 1, déuvia 1, véorog 1. With these words the possessive
translation of giAoc is almost universal; but, if we find the possessive
pronoun or possessive genitive at the same time, this must show that the
sense of dear was in such cases necessary, and consequently that the same
sense was not incompatible when the pronoun was not used. With 4ydéc
these not found, but a guast possessive dative. With 7rop a genitive is
found e 297, 406, 7 147, @ 114, w 344, 6 425, d 708, y 205, 7 68, 0 166, 182,
the last a possessive. The pronominal dative iscommon. «7p has a geni-
tive d 270, « 485; a possessive « 413, 7 274, and datives. or#ecor has
genitive v 9, and datives; yobvara genitive »v 231, and datives; yvia genitive
N 85, and datives. This gives fair ground for the conviction that 9iAos
was to Homer no empty epithet, nor even a possessive pronoun, but really
meant with the words of the second class what it did with those of the
first, a plump dear. The explanation of this phenomenon was conceived
to be this: In many ways it is the youth of the world which the Homeric
poems set before us in their childlike directness of expression and awkward
fondness for calling a spade a spade, quite foreign to ourselves but often
heard from the mouths of children before a conventional schooling has
repressed the undisguised utterances of the feelings. The child will say
**my poor dear hand” because it knows only to call that dear which is
dear. The man avoids it as savoring of egotism. The Homeric man calls
his hand dear, just as he calls it stout » 174, or his thigh thick A 231, or his
house beautiful @ 41, or himself culiant A 393; cf. ¢ 19, 20, 7 188, H 75, © 22,
etc. Were we to speak straight out from the heart, we should acknowledge
that the bodily organs are dear to us; see Xen. Mem. 1, 2, 54, Cic. Lael. 81.
Epigram of Maec. in Suet. Wesay, ‘‘run for dear life,” ‘‘dear me.” In

American Philological Association. 17

Hm. ¢iAvv buudv regularly of loss of life; cf. K 495, P 17, 4 109, v 323, IT 82,
X 58, ¢152. This idea of loss is to be considered B 261; cf. Z 272.

Another point to be taken into account in the explanation is the great
fondness of the poet for personification. This figure is applied to the feet
N 75, the hands « 434; and the members of the body are sometimes treated
as separate and distinct individualities apart from the person himself as
v 237, A 314, e 355, v 18-22. As soon as this was done, there is as good
reason for applying the word dear to one’s own hand, for instance, as to
that of a second person.

Some kindred uses of our word dear were cited from Shakespeare, where
it is used above 400 times; but the German exhibits the exactest parallel
in das liebe brod, lieben tag, liebe gott (cf. w 514, Theogn. 373) and in the
application of Web to gelt, vieh, gut, rock (cf. B 261), sonne, heraz, etc.

In later Greek the application of giAo¢ to words of the first class is com-
paratively common, but with those of the second it is rare. The following
were cited: Hes. W. & D. 3860, Theog. 163, 568, 283, W. & D. 608;
Hym. Ap. 113, 524, Epigr. 4 15, Tyrt. 10 25, 12 23, Theognis 531, 877,
983, Pseud. Phocyl. 98, Sim. Ceos 37 4, Iby. 4, Pind. Ol. 1 6, 24, Pyth.
3 109; Aesch. Choe. 276, 410, Agam. 983; Eur. Hek. 1026, Elec. 146;
Theoc., 17 65, 7 104, 21 20; Mosch. 4 1, 15, 32, 51; Apoll. Rhod., 2 712,
1 281, 3 492.

These examples appear to show that a sense of the quaintness of this
usage began to prevail long before the Attic period, but of the feeling that
it was equivalent to a possessive pronoun, no evidence was seen. Neither
does the canon appear to be laid down in the oldest scholia on the II. the
Ven. A. Dind.; but this and the scholia of the Od. treat ¢iAo¢ as if to them
it meant dear, cf. scholia on Od. 1 238: nor in the Lex. of Apollon., nor
Hesych., nor Etymolog. Magnum, nor Suidas. But it appears in scholia
B. L. V. on Il. 1 569 (see on 1 555), and in Eusth., B 261, and most lexicons
since.

In conclusion an earnest protest was entered against that kind of criticism
which, in translating from the classic poets, would root out all the poetry,
would carefully prune away all the peculiarities of the original, would in
fact strip Homer of his ¢gapo¢ and y:7éy and array him in dress coat, vest,
pants, immaculate tie and shirt-front, before permitting him to appear in
modern society. There is wide difference of opinion as to methods ina
poetic translation, but in the class-room the aim should be to reproduce
the original with all possible fidelity. If Homer says that the wave shouts,
let us not translate it roar; if he calls wine honey-hearted, let us follow
Tennyson; if he calls the heart dear or shaggy, let us so translate. Like
true archaeologists, let us dig for the genuine treasures of that distant day,
and carefully preserve all we find. Shall Schliemann, shall Cesnola, put
their treasures into the hands of the artist of to-day, to remold and refash-
ion till all the pottery becomes Wedgwood or Sevres, and all the gold
might have come from Paris or London?

Remarks were made upon this paper by Professor B. L. Gilder-
sleeve, Professor W. W. Goodwin, and Professor E. North.

3

18 Proceedings of the

Dr. L. A. Sherman, of the Hopkins Grammar School, New
Haven, Conn., then read a paper on ‘The Greek Article as a
Pronoun.”

This communication grew out of the conviction that the identity of
the Greek and English articles, perhaps commonly assumed in teaching
Greck, and certainly often implied in Greek grammars, was incorrect
and misleading. The article of Greek is demonstrably stronger and nearer
its pronoun-original than that of English. It is clear that the 6 9 76 of
Homer had commenced the same carcer of progressive weakening which
‘ is common to the history of the article generally; but it had not in the time
of Plato and Xenophon descended through all the stages and touched
bottom in the shape of a genuine article, as has the English from its
equivalent original se seo that of Anglo-Saxon. It still remained a demon-
strative in ol uév, of dé and some other expressions in Attic prose, and in
varying instances in Attic poetry, as is admitted by all scholars. There
was, therefore, a lingering consciousness in the Greek mind of a pronom-
inal potentiality in the article. In the light of certain examples it was
maintained that the article of Greek was very nearly like that of modern
German, which retains so much of its old pronominal strength as to
admit of standing as the representative of a noun alone. It was then
urged that the article in prepositional phrases like Mévwy xai ol civ avre,
oi aug? TiZarwrva, was more likely pronominal than article to an omitted
avdpec; ag also the second article in of &’ immo: ararreg of eta Ki:pov, and in
like examples.

It was argued further that, if these conclusions were correct, the inter-
pretation of the article with the participle would need to be amended. It
is unnatural to suppose that the participle was always substantived over
the article, when the latter being so nearly a demonstrative had already
so much of the substantive in its nature; but rather in an unknown propor-
tion of instances it stood in predicative agreement with its so-called article.
This view scems really suggested by the very statement of the best grammars
(see Curtius’ Schgr. § 581, anm.; Kiihner, Ausf. Gr., § 461, 4, 5). With
Kihner’s statement that eiaiv of Aéyovor sometimes gives way to eisiv ot
Aéyorvrec is to be associated the constant rendition of the article with par-
ticiple in German and English grammars by a relative clause. If the
participle be really, in the thought of the Greeks, a nomen agentis, it
should be easier to deny the article a substantive value than we find it. In
regard to the facts of the language, it seems to be clear, first, that the parti-
ciple is not infrequently found in undoubted predicative agreement with
an article known to be demonstrative, as the following familiar sentences
illustrate: "Evraia diéoyov aaAhawy Baatrcic¢ te Kat ol “EAAnveg o¢ Tpedxoyra
orddia, of piv dtaKovrec tov¢ Ka attorc, Og TavTag vikavTes: ol Hé dpmacovres
wc Hon mavreg vixovTec (Anab. I. x. § 4). Secondly, there are occasional
instances where the participle must be regarded as in predicative agree-
ment with article-forms not admitted to be demonstrative: . . . Kai ri
duxaida, ... THY Gogyy Kai Kadi Peyonévyy civat, Aaujavee (Anab. T. x. § 2).
Thirdly, there are passages constantly met with in which not only greater

American Philological Association. 19

difficulties ‘are encountered, but the thought seems distorted and shorn of
its naturalness and force if the participle be taken as attributive. A
single example will suffice: of dé woAéusor dpavreg pév Todo augi Xetpioodov
evmeTac Td Wdwp TEpavTas, dpavrec dé TOVE audi Zevoddvra Blovrec etc robuTradad,
deiaavreg pu aroKdAeodeinaav, debyovo ava kpatoc.. (Anab. IV. rr. § 21).
There seems little doubt that the lack of nomina agentis in Greek is in
some measure supplied by the participle, but it would also seem a matter
of judgment in each case whether the participle be so used or not.

Remarks were made upon this paper by Professor B. L. Gilder-
sleeve, Professor W. W. Goodwin, and Professor A. Harkness.

Professor J. B. Sewall, of Thayer Academy, South Braintree,
Mass., read a paper on “The Greek Indicative, Subjunctive, and
Optative Moods: what is the distinction between them?”

I. It is obvious that the indicative, in general, predicates fact as actual.
It is that form of the verb specially employed for the assertion of what is,
was, or will be. E. g. Dem. 4:1, éesd# xrA.—The different uses of the
indicative mood in the different kinds of sentences shows the same. Dem.
4:5,—4:30 (relative); 1:15,—4:47 (result, cf. Goodwin, Greek Moods and
Tenses, § 65. 1, N. 5, and § 65. 3.); 4:86 (causal). After temporal particles
signifying until and before that cf. also Goodwin G. M. T., § 66.

When a clause denotes a result not attained in past time, or an unattain-
able purpose in past time (after iva, etc.), or a wish for the contrary of what
is taking or what has taken place, or a condition or conclusion contrary to
fact, the same is apparent because the result not attained, the unattain-
able purpose, the contrary wish, and the contrary condition or conclusion,
are all brought before the mind as the opposite actual facts.

We may call the indicative therefore the mood of actual fact.

II. The Subjunctive. Can we detect that essential character in those
relations (condition, purpose, temporal limit, deliberation, etc.), which made
the verb-form we call the subjunctive a mood by itself, not the indicative,
not the optative, not the imperative? Let us sce.

In Dem. 4:3, iv’ eidre xrA, what is the element, essential and common,
in edge, a subjunctive in a clause of purpose, and in oAcywpyre (same sen-
tence), a subjunctive in a conditional clause? Not simply futurity. They
do not predicate actions which are actually to, or will certainly, take place,
nor actions which are mere possibilities, potential—actions cxisting only
in thought or conception as fof2owbe (same sentence). The facts they
predicated were rather before the speaker’s mind as hoped or expected, in
the one case, and in the other, as feared or deprecated—facts something
more than mere conceptions, and much less than actual—rather, lying in
the region, so to speak, between actuality and mere conception; the region
of doubt, uncertainty, dependence; the region of facts dependent in some
way upon will or other power and determination than the speaker’s; a
kind of fact to which the term contingent may very well be applied.

The same is clearly to be seen in Dem. 4:17, 22, 41,—Plato 230, E., and
all similar examples.

20 Proceedings of the

This being the nature of the subjunctive, it is easy to see how it came
to be used in the form called deliberative, and in exhortations and prohibi-
tions, and with the value of an emphatic future after ov u#, and also why
the indicative sometimes replaces it in final clauses (after Swe, etc.), and
after verbs of striving and fearing by which the form of statement was
made more vivid, viz.: for the reason that the indicative was the mood of
actual, while the subjunctive was the mood of contingent fact. And to
say with reference to #4 with the perfect indicative expressing a fear that
something bas already happened, that ‘‘the difference between this and
the perfect subjunctive is often very slight, the latter expressing rather a
fear that something may hereafter prove to hace happened” (Goodwin,
G. M. T., § 46, note 5, b), is only to say, that, in this case, the indicative is
used for actual, and the subjunctive for contingent, fact.

III. The Optative. The position of Rost, Ktthner, Donaldson, and
others, that this mood is nothing but a peculiar form of the subjunctive,
and that they differ in tense only, is very properly discarded by our best
and latest Greek scholars. This can only be however on the ground that
there is an essential difference—a modal.rather than a temporal difference.

The optative received its name from the early grammarians from its use
in wishes, but this evidently was not its original, as it was not its principal,
use. Its principal use lay in the protasis and apodosis of conditional sen-
tences, in final clauses to denote past purpose, in oratio obliqua and after
Ewe, wéxpe ov, ete., after historical tenses. In these different positions it is
easy to see that the optative was the form of the verb employed when the
act or state to be predicated was merely conceptional, not brought before
the mind as actual either in the present, past, or future, nor as contingent,
but as merely conceived. E. g. Dem. 4:25, e yap éporrd reg ** * * cizrocr’ dv,
‘for if any one should ask **** you would say,’ the fact predicated in
both condition and conclusion is not predicated as actual, nor in any way
contingent, but as merely conceived. It is pure supposition, mere thought
as Madvig says. So elsewhere. For an example of final clause (taken
at random) see Xen. Anab. II. 6, 21, and oratio obliqua Thucyd. 2:13.
In the latter example our English idiom has no other form for the clause
dre ’Apzidauog pév oi Sévog ein than the blunt factual indicative, that Archida-
mus zas his friend; and if Thucydides had made the statement on his
own authority, he would have said ’Apyidapog pév ol févoc Fv. But he
attributes it to Pericles, and that carries it out of the region of actual fact
as far as he himself is concerned, and he employs the mood which his
mother tongue provides him with to express it separate from all actuality,
fact as it exists merely in the conception.

We may call the optative mood, therefore, the mood of conceived fact.

My conclusion accordingly is, that the distinction between the Greek
indicative, subjunctive, and optative moods is an essential one, one of
kind and not of degree merely, the indicative being in general the mood
of actual fact, the subjunctive that of contingent fact, and the optative that
of conceived fact.

And a corollary from this would be that the distinction between the
subjunctive and optative moods in conditional sentences is not one merely

American Philological Association. 21

of degree, but of kind—not one merely of greater or less vividness, but of
essential nature, which supports a position assumed in a former paper
(Trans. Phil. Asso., 1874).

‘The Auditors of the Treasurer’s Report reported that they
found it correct; and it was, on motion, accepted.
A recess was then taken till 8 o’clock p. m.

EVENING Session.—8 P. M.

The Association met after the recess.
The annual address was delivered by the President, Professor
B. L. Gildersleeve, of Johns Hopkins University, Baltimore, Md.

At the meeting of the Philological Association in New York two years
ago, the President, Professor Harkness, gave a comprehensive survey of
the progress and results of philological study,during the last century, in
which he laid especial stress on the origin and growth of comparative
philology and linguistic science, and impressed upon the members of
the Association the duty of carrying forward the good work that had
been begun by others. The concluding exhortation of that address has
suggested the theme of this—the Special Province of the American
Philologian.

Many fields of philology are as open to Americans as to any devotees of
linguistic science, and some philological work is peculiarly our own. So
it would be sheer laches in us to resign the department of American
languages. And although the history of this Association shows that there
is no danger of this, the very fact that so much has been done, pledges
us to still greater activity inthis department, which is always challenging
exploration. As the aboriginal languages of America demand our special
care by reason of our local relations, so our historical connection makes
English a matter of prime interest to us, and American scholars have
done admirable work here, and in some of its forms the historical and
scientific study of English has more votaries in America than can be
found anywhere else. But outside of these departments, which are ours
by local and historical ties, the power of individual effort and individual
example has been strikingly manifested in the Sanskrit studies that have
made a name all over the world for the distinguished scholar, whose
absence we deplore, and have established the science of comparative
etymology in this country on a sound basis. In all the leading branches
of philological work there are gratifying signs of life, and our associates
who are pushing forward the study of the Germanic and the Romance
languages, and the Orientalists of this Association and of our sister society
show no lack of activity. But it must be admitted that the prospect is
not so good for the classical philologian as for those who are at work in
less crowded fields, and as they in an especial manner need a word of
encouragement, this discourse is addressed especially to them. There are
indeed some grounds for the despondency of the classical philologian,

22, | Proceedings of the

who is aiming at higher work, but there is no reason for despair. The
preparation furnished by our schools and academies is very defective, but
it was a great deal worse a few decades since. The isolation from other
workers is chilling, and yet it is not so entire as it once was, and this
Association has done much to bring philologians nearer to each other,
although the intervals of meeting are so great. The want of a channel of
intercommunication other than the annual voluine of transactions is one
cause why philological productivity is so slight in this country, for men
cease to produce, if there is no outlet for production; but as soon as the
want is properly presented, it will certainly be supplied, and the establish-
ment of a philological journal will not and cannot long be deferred. The
most scrious drawback that we have to encounter is the want of apparatus;
but perhaps even that is exaggerated. In the matter of occasional mono-
graphs, the increasing facilities of the book-trade brings certainly as
many within our reach as can be procured in small German towns, in
which excellent work is done for all that; and besides it is unwise to
attach too much importance to these dissertations, a large proportion of
which are written by very young men and have no great scientific value.
But even if all the literature were accessible, every edition of every
author, every treatise on every subject, it would not be desirable to dull
the freshness of appreciation which can only be gained by direct employ-
ment with the text—with the theme. The field of antique literature is
vast, but it is a narrow range as compared with the continent of com-
“mentary and dissertation, and any competent man can survey with his
own eyes large stretches of the original sources of all our knowledge and
so gain new points of view as well as new illustrations for the work he
may bave in hand. Let any man try what can be done by close study of
a text, and a wide range of reading in cognate directions, before he says
that Americans have nothing to do except to repeat the references in
German books, or at most to run over the indexes of German editions.

Of course, if in our authorship we persist in treading the eternal round
of school-books, there will be less room for individual effort, but even in
the most thoroughly beaten track of classic literature, there is something
yet to be settled; and if we look at our work from its historical and
acsthetic sides, all of it requires to be done over every few years. With
the progress of social science, with the advancing knowledge of historical
evolution, the problems of antique culture, of antique legislation, appear
in new lights. Not to speak of the positive gain to be derived from the
newly-discovered inscriptions and monuments, which are adding more and
more definiteness to our conceptions of the antique world, and are helping
us to a better understanding of the dialectic life of the classic languages,
and the cantonal and provincial life of the classic peoples, ancient history
has to be interpreted into terms of American experience; and it is not
saying too much to say that some phases of American life enable us to
understand the ancients better than some contemporaries on the other
side can do. But apart from the special aptitude of Americans for the
appreciation of the political and social relations of antiquity, due partly
to our peculiar endowment, partly to our peculiar position, the aesthetic

American Philological Association. 23

problems involved in the study of classical philology shift from time to
time; the great masters ever need new interpreters. Even the best work
done forty or fifty years ago leaves us thoroughly dissatisfied. Not only
is there that sense of shortcoming which we feel in all translations, but
there is often a repulsive, often a ludicrous incongruity, which shows the
change of aesthetic basis. Now Americans have proved and are proving
every day that they do not lack acuteness, subtlety, delicate appreciation,
and just comprehension in their literary criticisms, but, so far as I know,
there has been little independent treatment of the antique authors in this
regard. Nor is it unworthy of consideration whether the exact study of
function—to use a wider word than syntax—may not be destined to give
us a firmer foundation and a clearer outline for the whole structure of
style than would have been thought possible some years ago. Indeed this
study of syntax or of function—comparative syntax, historical syntax—
is large enough to occupy all the force that classical philology can spare
for generations to come. No index will serve the turn of the true
investigator, because no index-maker can possibly anticipate all the points
of view which the thoughtful student will assume, so that it is simply
indispensable that the student shall have immediate vision, immediate
intercourse with the authors themselves, and if a second-hand acquaint-
ance is of little use in this field of study, it is of no possible avail in yet
another direction—the exploration of the linguistic consciousness of the
great classic authors—a direction in which something yet remains to be
discovered. The conclusion of the whole matter is that the classical
philologians of America are in nowise debarred from high scientific
work, and especially in the province of grammar, this Ypryxo¢ padnyatwv
as Boeckh has called it, may the American philologian find abundant
room for the native sagacity, the unresting energy, the quick inventiveness
that have distinguished our people in other departments of science.

At the conclusion of the President’s address, the Association
adjourned to 9 o’clock Thursday morning.

Tuurspay, July 11.

The Association resumed its session at 9 o'clock a. m., the
President in the chair.

The minutes of the sessions of the previous day were read and
approved.

The Secretary presented a report from the Executive Committee,
announcing the election to membership of,

Professor C. R. Hemphill, Theological Seminary, Columbia, S. C.

The committee on the time and place of the next meeting recom-
mended that the next meeting be held at Newport, R. I, during
the second week of July, 1879.

24 Proceedings of the

On motion, it was

Resolved, That the report of the committee on the time and place of the
next meeting be accepted, and that the Secretary be instructed to send to
the members a printed circular, mentioning the place appointed for the
next meeting, and requesting each member to send word, whether he
would prefer the date appointed, or a later week in July. [See page 36.]

On motion, it was

Resolred, That the Secretary be instructed to send to the members a
printed circular embodying the substance of the report presented on Tues-
day evening by the Committee on the Reform of English Spelling.

The committee to nominate officers for the next year presented
nominations as follows:

For President—Professor J. B. Sewall, Thayer Academy, Braintree,
Mass.

For Vice-Presidents—Professor C. H. Toy, Southern Baptist Theological
Seminary, Louisville, Ky.; President William C. Cattell, Lafayette Col-
lege, Easton, Pa.

For Secretary and Curator—Professor Thomas C. Murray, Johns Hop-
kins University, Baltimore, Md.

For Treasurer—Charles J. Buckingham, Esq., Poughkeepsie, N. Y.

For additional members of the Erecutive Committee—

Professor W. W. Goodwin, Harvard University, Cambridge, Mass.

Professor M. W. Humphreys, Vanderbilt University, Louisville, Ky.

Professor F. A. March, Lafayette College, Easton, Pa.

Dr. J. Hammond Trumbull, Hartford, Conn.

Professor W. D. Whitney, Yale College, New Haven, Conn.

The report was accepted, and the persons therein named were
declared elected to the offices to which they were respectively
nominated.

Professor B. L. Gildersleeve, of Johns Hopkins University, Balti-
more, Md., then read a paper entitled ‘Contributions to the History
of the Articular Infinitive in Greek.”’

The use of the articular infinitive completes the deorganization of the
infinitive in Greek. Deorganized before, the infinitive had virtually become
an accusative to the Greck consciousness, yet in many of its combinations
retains traces of its dative origin. Indeed the dative of the object for
which and the accusative of result mect in most of the familiar construc-
tions of the infinitive. Not only in case, however, but also in tense there
must have been a change in the relation of the infinitive to meet the neces-
sities of oratio obliqua. So the future infinitive, old as it is, seems to be
younger than the other tenses of the infinitive, as the future optative is
demonstrably younger than the other tenses of the optative. The connect-

American Philological Association. 25

ing link between the verbs of creation (verbs of will and endeavor), which
take the abstract infinitive (neg. 44), and the verbs of saying and thinking,
which take the oratio obliqua infinitive (neg. ov) is formed by verbs of
swearing, witnessing, and hoping, in which the use of the negatives and of
the tenses seems to show the transition. Now as the articular infinitive is
younger than the orate obligua inf., it would seem to follow that the
oralio obligua inf. would be susceptible of the article, a point which must
be admitted to some extent for the articular inf. and dy and the articular
fut. inf.—both comparatively rare combinations. The truth seems to be
that the articular infinitive may represent the contents of an oratio obliqua
sentence without losing, however, its abstract character, which abstract
character is sufficiently indicated by the negative 9.

As to the history of the articular infinitive, the construction does not
occur in Homer, the only apparent example (Od. 20, 52) being in apposition
to a demonstrative article.

In Pindar it occurs ten times—all except one example being apparently
in the nom. but all in such a position as to vindicate a virtual accusative
use.

In Aeschylus the occurrences are one in 159 verses. The cases are
chiefly nominative and accusative. Many of the examples are due to the
stereotyped grouping of rd “4, Td wu) ov. Prepositions are very sparingly
used. The tenses are present and avrist.

In Sophocles the occurrences are one in 120 verses. The vast mass
consists of nominatives and accusatives; there are very few genitives and
datives. Prepositions are used sparingly. The tenses with no exceptions
worth considering are present and aorist. A remarkable instance of what
may be called substantivized oratio obliqgua occurs in Antig. 235. 6.

In Euripides there is a marked falling off; but one occurrence in 320
verses. Forty per cent. of the whole number are nominatives, but the
genitive bulks much more largely than it does in the others. Prepositions
and quasi-prepositions are not much used. The tenses are present and
aorist, ei ¥ioda: being a practical present. There is somewhat more freedom
in the handling. The largest number occurs in the Iphigenia at Aulis.

In Aristophanes we note an increase as compared with Euripides, one
occurrence in 258 verses. The bulk consists of nominatives and accusa-
tives. Prepositions are not very common. The tenses are present and
aorist (eiwévac being a practical present). A large proportion of the artic-
ular infinitives in Aristophanes are purely deictic or anaphoric, some
exclamatory, others parodic. The largest number occurs in the latest
comedy, the Plutus.

Of the historians Herodotus uses the articular inf. very rarely in com-
parison with Thucydides. Few prepositions are employed. Remarkable is
the use of avri directly with the infinitive. While the bulk of Thucydides
is only about six to Herodotus’s seven, Thucydides uses the articular
infinitive with great freedom and more than eight times as often. -All the
cases and fifteen different prepositions are freely used. Also all the tenses.
Especially noteworthy are the articular inf. with dy and the articular
fut. inf.

4

26 Proceedings of the

In the orators the usage varies greatly, the occurrence to the Teubner
page being for
Lysias, about 12
Andocides (estimated), 20
Isaeus, 25
Aeschines, . . . . . .80
Antiphon (estimated), . . . . . .50
Lycurgus, . 60
Isocrates, 60
Deinarchus (estimated), 80
Demosthenes (private orations), . . . .80
(public orations), . . 1.25
The lowest average of the undoubted public speeches i is presented by the
Second Philippic .87; the highest by the First Olynthiac 2.75. The long
public speeches vary from 1.06 to 1.62. In the private orations there is
considerable variation, the highest being contra Cononem 1.07, and tn
Pantaenetum 1.06, the lowest contra Calliclem in which there is no occur-
rence.

Professor M. W. Humphreys, of Vanderbilt University, Nash-
ville, Tenn., next read a paper on “Elision, especially in Greek.”

In this paper the position was taken that elision in Greek was usually
total.

1. The words employed by the Greeks to denote elision were discussed.

2. .The ancients used their terms carelessly, and sometimes used the
same word to denote both crasis and elision. ‘The passage cited by
Corssen to prove partial elision really referred to crasis, as the illustration
shows: "Eore dé cuvaroign dbo guvntvtev digpnuévuv ei¢ piav ovdAdAaBgv evoot,
otov TO dvopa, Tohvopa,

8. Discussion of Ahrens’ arguments.

(a). Ahrens says that some words would be unpronounceable tf elision were
total, as to92’, céuv’, aicxp’, etc., etc. But as he asserts that elision combines
_two words, why may they not be combined by omitting the vowel and
pronouncing a consonant at the beginning of the next word: ofp-vian?

(0). Sometimes a hiatus would remain ‘causing an unpleasantness that
evidently ts not tn tt.” How does he know there is nothing unpleasant
in it? Compare Sanskrit rané dsit, vana ast. But granted there is no
offense in it; the two words being uttered continuously would prevent
offense except such as exists where two consecutive vowels occur in one
word. And there would be great offense in dAye’ E0nxev, if we pronounce
the elided a a little accordiifg to Ahrens’ theory.

(c). The scholia on éx svpdrwv yap atSic ad yaAf' dp& were discussed.

(d). Ahrens’ interpretation of cvvadorgy discussed, and ancient definitions
cited.

4. The Greeks allowed hiatus in prose. They could always neglect
elision (in prose); but they could also elide under certain conditions.
Hence they had the choice between the two. Not so in Latin, where
elision was almost universal (Cic. Or. 44, 152). In poetry Greeks avoided

American Philological Association. 27

hiatus. This restriction with the requirements of metre sometimes caused
elision even before a long pause. 7'hen the elision was only partial, and
sometimes even only apparent; i. e. the verse merely appeared perfect to
the eye. Hence they felt at liberty either to elide totally or partially as
circumstances demanded (a subject to be discussed in another paper).
But that vowels could be and often were totally suppressed, is shown by
these considerations:

(2). The aspirate of an initial vowel passes over to the (remaining) final
consonant, as 颒 jyiv.

(b). The accent, when final, is thrown back, as 097’, céuv' (words cited
by Ahrens). Aphaeresis of accented vowels (aphaeresis is partial) and the
Sanskrit svarita show the possibility of accenting a reduced vowel; hence
the elided vowel was not merely reduced. (Arguments proving aphaeresis
parkal, omitted.) Analogous to svariia is Synizesis, as Atyuntiove (8 syl.)

(c). Diphthongs are frequently elided. If the second vowel merely
combines with the next vowel, then the first retains its full quantity.

(d). Such plays upon expressions as 7 épavog = yéparog (see Ahrens de
Crasi et Aphaerest, p. 2) show total elision.

(ce). ‘Orérav, érav, ydp, your, etc., show that a vowel was totally expelled.
This happens in Latin much more rarely, as in tantopere, whereas we find
neuter, deinde, etc., resulting from partial elision (for in Latin elision was
partial except in -que, -ve, -ne).

(f). If elision had been only a diminution, it would not have been sub-
ject to such rigid restrictions, but would have been more nearly universal,
as in Latin.

5. But vowels were also only partially elided under certain circum-
stances,

(2). We find elision at the end of a sentence. That we find it also at
the end of a verse proves nothing, for when this happens the sense always
requires the verses to be read continuously.

(0). Elision takes place between two speakers. Here, in fact, the elision
is entirely for the eye of the reader. As the first speaker uttered his final
vowel, the second speaker began his first syllable, and so a rapid exchange
of words is secured.

6. That elision was only partial in Latin is shown by the fact that it
was so universal (Cic. Or. 44, 152). In tantopere, magnopere, the long-
continued juxtaposition led to the suppression, especially as the vowels were
similar. The absence of crasis also proves it. The nearest approach to it
is in cogere, deerrare, etc. In amatast, the quantity results from position, as
is shown by ‘‘ corrimpéré’s ausus,” ‘‘ véndéré's ausus ” (Tibullus, I, 9; 53,
57). Some statements of grammarians imply partial elision. Some late
grammarians speak of it as if it were total; but*their authority is worthless.
That -que, -ve, -ne suffered total] elision will appear in another paper.

Mr. A. D. Savage, of Johns Hopkins University, Baltimore,
Md., next read a paper on “"Padapav9voc épxoc, or Did the notion
of irreverence in swearing exist among the Greeks?”

This paper is a discussion of Greek views of the moral side of swearing.
*‘Padaydydvoc opxog is ‘the name found in scholia, lexica, etc., of oaths by

28 Proceedings of the

animals and plants. The reason given by the scholiasts is, a wish not to
use the names of gods in swearing. Of the passages in scholiasts, etc.,
which mention this oath, two may be regarded as the sources of the
others, Schol. Aristoph. Av. 521, and Schol. Plat. Apol., 21 E. The
earliest appearance of the above-mentioned reason is in a fragment of
Sosicrates given by Schol. Aristoph. Av., 521. Sosicrates flourished not
later than 150 B. c. It would be expected that such a notion as the
irreverence of swearing among pagans would be made the most of by the
Greek fathers of the church. But even the fathers do not condemn
swearing for that reason. Again, when the moral side of swearing is
touched upon in the literature of Greece and Rome, the grounds for
condemnation are those of the fathers; either, that it is better to be on
the safe side of perjury, or that a man of honor should on common
occasions expect his word to be sufficient. This might lead us to suspect
that the notion of irreverence (provided we thus interpret the words of
the scholiasts) came to them from the grammarians of Alexandria, for
this notion existed then in that part of the world, among Israelites and
Egyptians. An examination, however, of the passages in classical Greek
authors, in whom the oath of Khadamanthus is found, leads to the view,
first, that the names of dogs, geese, and plants were substituted reveren-
tially for the names of gods by some persons whose piety unhappily was
tainted with weakness; and in the second place that such oaths gained
only the sneers of the more enlightened. Socrates swears by the dog and
the plane-tree, but he also swears by the gods. Hence the inference that
with Socrates it was not in earnest. In Aristoph. Av., 521, we are told
that the prophet Lampon swears an oath by the goose, when he has a bit of
swindling todo. And in Aristoph. Vesp., 88, the speaker swears by the
dog, and calls Nicostratus a dirty beast (xarazfyor). Hence in the eyes of
Aristophanes this oath was silly. A fragment from Cratinus preserved by
the Schol. Apol., 21 E. says oi¢ qv péytoroc bpxocg arravre Ady xiwr, erecta xh
Seovo F eciyav. This makes it plain that there were people who swore by
dogs and geese instead of gods. We may take the words of the Scholiasts
to mean reverence, and use them here. The writer of the paper would
incline to the view that Cratinus’s mention of the oath is satirical.

Remarks were made upon this paper by Dr. Julius Sachs and
Dr. J. H. Trumbull.

Professor T. D. Seymour, of Western Reserve College, Hudson,
Ohio, then read the last paper of the session, ‘‘ On the Composition
of the Cynegeticus of Xenophon.”

It was the aim of the paper to show that the assumption of certain
interpolations removed or explained most of the difficulties which abound
in the treatise, and restored the work nearly to the form which its author
gave it.

Since Valckenaer, few have maintained the authenticity of the prooe-
mium or the epilogus; but L. Dindorf, in the last critical edition (Oxford
’66) thinks that sufficient evidence has not been brought forward to decide
against the genuineness of the treatise itself, and it is to be observed that

American Philological Association. 29

all the critics seem to have assailed or defended the Xenophonticity of
the work as a whole and on general grounds.

So much of Xenophon’s life was spent on campaigns in Asia Minor
and in Peloponnesus that we are not surprised to find a large number of
poetic and dialectic words in his writings. Hence it is impossible to
attach much importance to the unusual words in the Cynegeticus. But
the characteristics of his style are well known. We have from his pen
voluminous works on varied subjects and written at every period of his
life.

In this opusculum one peculiarity of style to be noticed is the frequent
occurrence of Asyndeta. Xenophon is not averse to the rhetorical employ-
ment of Asyndeton as Hist. Graec. rv. 3, 19—ewdoivro, éudyovro, aréxretvov,
ané3vynoxoy and even where there is less animation, as Anab. vi. 1. 6, the
Greeks at KdéArne Acufyv are said to plunder rupoi¢ xa? xpr9dc¢, olvov, dorpua,
pedivas, ctxa. But in all of Xenophon’s larger works there are not so many
examples as in this one treatise and none like those in Cyn. v. 80. cf. also
Iv. 1 and v1. 1.

Another peculiarity is the use of the infinitive, e. g. 1x. 1, elvac and x. 1
xexrjoS8a, Inv. 15 we have AauBavey tho no ypf or dei has been used in
the chapter. In most cases a direction is clearly implied, and the infinitive
must depend on the general idea of advice which pervades the work, and
which is stated at the beginning of 11. 2: 50a d2 xa? ola dei rapeckevacpévor
tAdeiv éx’ aird dpdow nai avta Kai tiv Excorhuny éxdorov. But it is more than
seven pages after this that the first example of the unusual infinitive occurs.

The paper next noticed the peculiarities in the use of prepositions in
this treatise. Beginning with 1. 18, the part of the work most Xenophontic
in character, forty-six of the first one hundred verbs are compounded with
prepositions—-nearly the proportion in Xenophon’s larger works—while in
Chap. v., fifty-eight of the first one hundred verbs are so compounded,
often with no sensible difference of meaning from the simple verbs.
Further, an unusual number of verbs are compounded with two or more
prepositions; and the same preposition is repeated with the noun about
ten times as frequently as on the same number of pages in the Memorabilia
or Cyropaedia. This indicates, of course, a later authorship for the
passages where such peculiarities are found. The preposition is sometimes
irregularly omitted, as rv. 9; et¢ ra dpy woAAduec, rad2Epya Frrov.

Acé with the genitive is noticeable in m1. 5: dtatpéxovor did rov Lyvove.
Cf. rv. 8: mpotrwcay did Tov lyrouc, also VI. 22 and x. 16. The use of ueragt
is not Xenophontic in v. 8 aroVev roAt, pixpdv, perakd tobruy, cf. Iv. 1.

Another particle to be noticed is 67é. Never used by Xenophon, it is
found here four times, v. 8 and 20, rx. 8 and 20.

The use of the plural of abstract nouns was noted, e. g. 11. 7: aobuperpoc
Ta TAY TPd¢ Ta wHKH. ITI, 3: oxAnpal ra eldy. IV. 1: Ta peyDn peraksd paxpav
cal Bpazéwv. The omission of the reflexive pronoun, specially with perra
and its compounds, is unusual. The use of periphrastic expressions, most
frequently with évecv, as Iv. 8: éxovoa: farwoar xrA, was noticed.

These peculiarities will be recognized at once as common in a later age,
and where many of them are found in dny passage they afford a presump-
tion of later authorship than Xenophon.

30 Proceedings of the

The result of the investigation is as follows: Xenophon began with I. 18.
After II. 8 there is a long interpolation reaching to VI. 7. After VI. 16
six §§ are inserted by the reviser. Chap. VII is genuine with the excep-
tion of two or three short clauses and most of §$ 5 and8. Chap. VII is
doubtful. Chap. IX has interpolated §§ 8-10, 13-16, 19-20. In Chap. X
SS 4-18 and part of § 22 are late. Chap. XI presents few peculiarities of
style, but the unusual use of yerd § 8, would suggest that it is not Xeno-
phon’s. Moreover our author is accustomed to write from his personal
experiences, of which there are few traces here. Chap. XII may well be
genuine as far as § 16 where the original work ends.

This scheme assigns to Xenophon less than half of the work before us,
but it removes or explains nine-tenths of the difficulties and leaves us a
more systematic treatise with a beginning, an end, and a well-ordered
middle. It deprives us of Xenophon’s authority for some stories which
savor more of Aelian, and removes many sentences full of unnatural
rhetoric, but does not remove anything in which the style of Xenophon is
marked.

Who the interpolator was it is perhaps useless to inquire. He evidently
lived before the time of Arrian, as the prooemium is referred to by the
latter author. He seems to have changed the original text in but a few
places, generally contenting himself with inserting whole sections or
longer passages.

On motion, it was

Resolved, That the thanks of the Association be tendered to the proprie-
tors of the Grand Union Hotel, for the use of this Opera Hall, and for
their courteous attentions to the members of the Association.

On motion, the Association, at 1.30 o’clock, Pp. m., adjourned.

81

Association.

gical

alolo

American Ph

‘ooRUULOD bunspny | ‘SATUHdMWOA ‘A W (peuig)

O8[B DAVY OAA

TTHLLVO OM

‘suodnod predun pus anp-JaA0 day YA ‘PBOIIEYY 1.19489 4A WNOIQDQMUOYH 9q} JO puog 9y} poulurexa AT;euOBIEd

"4091100 0q 07 OUIBE OY} AJI7100 94 ‘BIOYONOA 94} YA 31 poredui0d pus ‘yuNo0dd8 9AOQY oq} poulMExs JulALT]

"a[QidaT[00 yUaseld 48 jou ‘YORI OG*LTY Jo ‘aurles oq} Jo suodnod onp-19A0 aaG
QUA ‘sIB[[OG peipuny VAL 10J ‘pvorpey 1938944 JNIIaUUOD 24} Jo puog oUO ‘JaINSBaL], 9Y} JO SPULY 94} Ul OS[¥ SI DIT],

‘saunsve4l, “KVHONIMOOG ‘£f SATHVHO a a
eS SULTS
L8'19F " ‘ezgt ‘g Aine ‘Ainsvoiy, uy souvfeg
000° ° . . 1Bj9.109§ pred sasuedxa 10430

pus ‘arly 419]0 10y ‘HIOWH [eG 8 ‘poyoa yuNoury

00° ‘sMONDaT[0) pus sozmieod, ‘JWINSBILL, ;, »
OL'9L , SUOTOVSUBI LL» UNNQLSIC ,, »
SFT ‘sodusvo1dxg pus saseisod ‘AIBIBINIE ,, »
00‘S1 . ‘Q1OWII}[8G UI UOIsses Jo sasuadxy
L809¢3 SUOTIVSUBL], ,, PUB ,, SPuIpedooig ,, Zu

a/)

28 Sorts

"SLEI ‘8 ftne-LLgl ‘or fine

‘NOILVIOOSSY TVOINOIONHG NVOIYINY 27 YIM juNODp us ‘saunsvo4y

‘£1819100§ aq} Wor} poatadar souRlYY
“48919]U]

‘suonBorjqnd Jo se[sg

‘poataoar QOUIS S}UIUISsaSsB PUB S2a,T
2181 ‘OL Ainge ‘Aamsvory, ul oouueg

‘aq

‘AVHONINONG ‘( SHTHVHO

OFFICERS OF THE ASSOCIATION.
1878-79.

PRESIDENT.
JOTHAM B. SEWALL.

VICE-PRESIDENTS.

CRAWFORD H. TOY,
WILLIAM C. CATTELL.

SECRETARY AND CURATOR.
THOMAS C. MURRAY.

TREASURER.
CHARLES J. BUCKINGHAM.

EXECUTIVE COMMITTEE.

The officers above named, and—
WILLIAM W. GOODWIN,
MILTON W. HUMPHREYS,
FRANCIS A. MARCH,

J. HAMMOND TRUMBULL,
WILLIAM D. WHITNEY.

MEMBERS OF THE AMERICAN PHILOLOGICAL ASSOCIATION.

Frederic D. Allen, Mt. Auburn, Cincinnati, O.

Joseph H. Allen, Cambridge, Mass.

William F. Allen, University of Wisconsin, Madison, Wis.
Joseph Anderson, Waterbury, Conn.

N. L. Andrews, Madison University, Hamilton, N. Y.
Albert N. Arnold, Baptist Theological Seminary, Chicago, II.
Cecil F. P. Bancroft, Phillips Academy, Andover, Mass.
Stephen G. Barnes, Iowa College, Grinnell, Iowa.

Charles C. Bates, Plymouth, Mass.

H. Louis Baugher, Pennsylvania College, Gettysburg, Pa.

S. E. W. Becker, Wilmington, Del.

J. B. Bittinger, Sewickly, Pa.

Porter C. Bliss, office of ‘‘ Library Table,” New York City.
James R. Boise, University of Chicago, Chicago, I].
Charles E. Brandt, Farmington, Conn.

H. C. G. Brandt, Johns Hopkins University, Baltimore, Md.
Fisk P. Brewer, Iowa College, Grinnell, Iowa.

John A. Broadus, Southern Baptist Theol. Seminary, Louisville, Ky.
Charles J. Buckingham, Poughkeepsie, N. Y.

L. H. Buckingham, English High School, Boston, Mass.
Elihu Burritt, New Britain, Conn.

Henry F. Burton, University of Rochester, Rochester, N. Y.
Richard E. Call, Mohawk, N. Y.

W. B. Carr, Petersburg, Va.

Franklin Carter, Yale College, New Haven, Conn.

William C. Cattell, Lafayette College, Easton, Pa.

Talbot W. Chambers, 70 West Thirty-sixth street, New York City.
Henry L. Chapman, Bowdoin College, Brunswick, Me.

Elie Charlier (Life Member), Central Park, New York City.
Francis J. Child, Harvard University, Cambridge, Mass.
Lyman Coleman, Lafayette College, Easton, Pa.

George F. Comfort, Syracuse University, Syracuse, N. Y.
Jacob Cooper, New Brunswick, N. J.

A. Crittenden, Packer Collegiate Institute, Brooklyn, N. Y.
Howard Crosby, University of New York, New York City.
Edward P. Crowell, Amherst College, Amherst, Mass.

Mrs. M. W. DeMunn, Providence, R. I.

Schele de Vere, University of Virginia, Va.

Mary C. Dickinson, 178 Madison street, Toledo, O.

5

34 American Philological Association.

Martin L. D’Ooge, Michigan University, Ann Arbor, Micb.

T. T. Eaton, Petersburg, Va.

B. H. Engbers, Mt. St. Mary’s Seminary, Cincinnati, O.

George R. Entler, Franklin, N. Y.

Carl W. Ernst, Providence, R. I.

QO. M. Fernald, Williams College, Williamstown, Mass.

Gustavus Fischer, Rutgers College, New Brunswick, N. J.

M. M. Fisher, University of the State of Missouri, Columbia, Mo.
Alexander Fleischmann, Cazenovia Seminary, Cazenovia, N. Y.
H. H. Furness, Ph.D., 222 West Washington square, Philadelphia, Pa.
James M. Garnett, St. John’s College, Annapolis, Md.

B. L. Gildersleeve, Johns Hopkins University, Baltimore, Md.
D. C. Gilman, Johns Hopkins University, Baltimore, Md.
William W. Goodwin, Harvard University, Cambridge, Mass.
Richard T. Greener, Howard University, Washington, D. C.
James B. Greenough, Harvard University, Cambridge, Mass.
Ephraim W. Gurney, Harvard University, Cambridge, Mass.

8. 8. Haldeman (University of Pennsylvania), Chickies, Pa.
Albert Harkness, Brown University, Providence, R. I.

William R. Harper, Denison University, Granville, O.

Samuel Hart, Trinity College, Hartford, Conn.

Charles R. Hemphill, Theological Seminary, Columbia, 8. C.
Theophilus Heness, School of Modern Languages, Cambridge, Maas.
Thomas Wentworth Higginson, Newport, R. I.

George O. Holbrooke, Trinity College, Hartford, Conn.

E. S. Holden, U. S. Naval Observatory, Washington, D. C.
Selah Howell, Christian Biblical Institute, Stanfordville, N. Y.
Milton W. Humphreys, Vanderbilt University, Nashville, Tenn.
John T. Huntington, Trinity College, Hartford, Conn.

Edwin E. Johnson, Trinity College, Hartford, Conn.

Asahel C. Kendrick, University of Rochester, Rochester, N. Y.
Robert P. Keep, Williston Seminary, Easthampton, Mass.

John B. Kieffer, Mercersburg College, Mercersburg, Pa.

D. B. King, Lafayette College, Easton, Pa.

Louis Kistler, Northwestern University, Evanston, III.

Charkes R. Lanman, Johns Hopkins University, Baltimore, Md.
Albert G. Lawrence, Newport, R. I.

R. F. Leighton, Brooklyn, N. Y.

John M. Leonard, University of Missouri, Columbia, Mo.

John R. Leslie, Newport, R. I.

William 8S. Liscomb, Providence, R. 1.

Thomas R. Lounsbury, Yale College, New Haven, Conn.
Rebecca 8. Lowrey, 162 West Forty-seventh street, New York City.
Merrick Lyon, University Grammar School, Providence, R. I.
W. G. McCabe, Petersburg, Va.

Joseph H. McDaniels, Hobart College, Geneva, N. Y.

Francis A. March, Lafayette College, Easton, Pa.

D. 8. Martin, Rutgers Female College, New York City.

American Philological Association. 85

R. H. Mather, Amherst College, Amherst, Mass.

Charles M. Mead, Andover Theological Seminary, Andover, Mass.
John Meigs, Lafayette College, Easton, Pa.

Augustus C. Merriam, Columbia College, New York City.

Charles D. Morris, Johns Hopkins University, Baltimore, Md.

F. A. Muhlenberg, Univ. of Penn. (4807 Walnut street), Philadelphia, Pa.
Thomas C. Murray, Johns Hopkins University, Baltimore, Md.
Joseph H. Myers, Milton-on-Hudson, N. Y.

C. K. Nelson, St. John’s College, Annapolis, Md.

Edward North, Hamilton College, Clinton, N. Y.

Howard Osgood, Rochester Theological Seminary, Rochester, N. Y.
Charles P. Otis, Massachusetts Institute of Technology, Boston, Mass.
W. B. Owen, Lafayette College, Easton, Pa.

Lewis R. Packard, Yale College, New Haven, Conn.

William A. Packard, College of New Jersey, Princeton, N. J.

E. G. Parsons, Sumner Academy, South Byfield, Mass.

Tracy Peck, Cornell University, Ithaca, N. Y.

William T. Peck, High School, Providence, R. I.

William C. Poland, Brown University, Providence, R. I.

Noah Porter, Yale College, New Haven, Conn.

Samuel Porter, National Deaf-Mute College, Washington, D. C.
John W. Powell, Washington, D. C.

De Witt T. Reiley, Rutgers College, New Brunswick, N. J.
William A. Reynolds, Wilmington, Del.

Leonard W. Richardson, Trinity College, Hartford, Conn.

W.G. Richardson, Austin College, Sherman, Texas.

Ezekiel G. Robinson, Brown University, Providence, R. I.
Lawrence Rust, Kenyon College, Gambier, O.

Julius Sachs, Classical School, 649 Madison Ave., New York City.
George W. Samson, Rutgers Female College, New York City.

A. Duncan Savage, Johns Hopkins University, Baltimore, Md.
Wesley C. Sawyer, Lawrence University, Appleton, Wis.

Philip Schaff, Union Theological Seminary, New York City.

Henry Schliemann, Paris, France.

Jotham B. Sewall, Thayer Academy, Braintree, Mass.

Thomas D. Seymour, Western Reserve College, Hudson, O.

Joseph Alden Shaw, Highland Military Academy, Worcester, Mass.
L. A. Sherman, Hopkins Grammar School, New Haven, Conn.
Charles Short, Columbia College, New York City.

E. G. Sihler, Johns Hopkins University, Baltimore, Md.

Ephraim G. Squier, 185 East Thirty-ninth street, New York City.
A. B. Stark, Logan Female College, Russellville, Ky.

Benjamin F. Stem, Classical Institute, Easton, Pa.

Frederick Stengel, School of Mines, Columbia College, New York City.
William A. Stevens, Rochester Theological Seminary, Rochester, N. Y.
Edward F. Stewart, Easton, Pa.

Austin Stickney, Florence, Italy.

J. Henry Thayer, Andover Theological Seminary, Andover, Mass.

36 American Philological Association.

William E. Thompson, Genesee Wesleyan Seminary, Lima, N. Y.
Crawford H. Toy, Southern Baptist Theological Seminary, Louisville, Ky.
J. Hammond Trumbull, Hartford, Conn.

Joseph A. Turner, Hollins Institute, Botetourt Springs, Va.

William 8. Tyler, Amherst College, Amherst, Mass.

Milton Valentine, Pennsylvania College, Gettysburg, Pa.

James C. Van Benschoten, Wesleyan University, Middletown, Conn.
Addison Van Name, Yale College, New Haven, Conn.

C. Osborne’ Ward, 486 Adelphi street, Brooklyn, N. Y.

_ W. B. Webster, Military Institute, Norfolk, Va. °

R. F. Weidner, Philadelphia, Pa.

James C. Welling, Columbian University, Washington, D. C.

J. B. Weston, Antioch College, Yellow Springs, O.

Mrs. A. E. Weston, Antioch College, Yellow Springs, O.

Albert 8. Wheeler, Sheffield Scientific School, New Haven, Conn.
John Williams White, Harvard College, Cambridge, Mass.

W. H. Whitsitt, Southern Baptist Theological Seminary, Lexington, Ky.
William D. Whitney, Yale College, New Haven, Conn.

J. Colver Wightman (Life Member), Taunton, Mass.

Alonzo Williams, Brown University, Providence, R. I.

Edwin H. Wilson, Yale College, New Haven, Conn.

The Executive Committee herewith announce, in accordance
with the votes of the members communicating with the Secretary
[see p. 24], that the Eleventh Annual Session of the Association
will be held at Newport, R. I, beginning Tuesday, July 15, 1879,
at 3 o'clock Pp. mM.

Members intending to read papers at the next session of the
Association are requested to notify the Secretary at as early a date
8 possible.

TRANSACTIONS

OF THE

AMERICAN |

PHILOLOGICAL ASSOCIATION,

1879.

PUBLISHED BY THE ASSOCIATION.
PRINTED BY THE CASE, LOCKWOOD & BRAINARD CO.,
HARTFORD.

1880.

CONTENTS.

I. Modal Development of the Shemitic Verb.

By Professor C. H. Toy. ; . 6
II. On the Nature of Caesura.
By Professor M. W. Houmpureys . 25
ITT. On Certain Effects of Elision.
By Professor M. W. Humpureys. . . . 82

IV. Studies in the Heliand.
By Professor A. S. Coox. . ; ; . 60

V. On the Development of the Latin Subjunctive in Prin-
cipal Clauses.

By Professor A. HARKNEsS. . . 76

VI. The Original Recension of the De Corona

By Professor M: L. D’Oogr.. - | 92
VII. The Authorship of the Dialogus de Oratoribus. __
By Professor T. Peck. . ; ; . 105

VIII. On the Date of the Prometheus of Aeschylus.
By Professor T. D. Seymour. . . . 111

ProcEEDINGS:—Eleventh Annual Session, Newport, 1879.

TRANSACTIONS

OF THE

AMERICAN PHILOLOGICAL ASSOCIATION.
1879.

I.—Modal Development of the Shemitic Verb.

By CRAWFORD H. TOY,
NEW YORK CITY.

The verb-structure in the Shemitic family of languages is
one of the simplest in the world—simpler than that of many
uninflecting tongues. This is true, whatever may have been
the history of the family before it became what we now call
Shemitic. Through what changes it has past we do not
know, but in its present form it is markt by poverty of
verbal expression; or, we may say, it has, by a process of
sifting, reduced its verbal apparatus almost to a minimum.
We may consider it as reasonably certain that the verb in this
class of languages was originally a noun or (what amounts
to the same thing) that it began its development at a time
when there was no distinction of form between noun and
verb.* This noun or noun-verb advanced in two directions :
first, by the addition of syllables (prefixes, probably originally
independent words) that attacht substantiv-ideas to the
signification of the root, producing derived stems, which are
in a sort new verbs; secondly, by the purely inflectional
modification of the noun-stem by additions at the beginning

*See my article on ‘‘the nominal basis of the Hebrew verb” in vol.
vin. of the Transactions (1877).

2

6 C. H. Toy,

and end, and the differentiation of these different forms into
various modal expressions of the action.

By modal modification in the broadest sense is meant any
modification that does not add a substantiv idea to that of the
verb-root, but only expresses an accident of the action. It
may include all expressions of the time and completeness of
actions and the conception of them as real or unreal. The
material that the primitiv Shemitic selected for this expression
consists of two noun-stems, one simp! triliteral and another
provided with a prefix. Whatever the original force of these two
forms may have been, they were in the earliest time of which
we have information devoted to the representation of the ideas
of completeness and incompleteness respectively. The attempt
to establish a separate form for the expression of temporal
accidents in actions was, for some unknown reason, abandoned,
and such temporal expression came in only as a secondary
application of what may be called the completional force of
the two forms.* The second of these, the incomplete, or
better, the inchoativ, ingressiv, received farther inflectional
modification by terminations which in the noun exprest case-
relations, or a strengthend, intensified condition of the object.
The subject-relation or Nominativ was represented by the
ending uw (to take the form in which we actually find it), the
object-relation or Accusativ by a, the possessiv-relation or
Genitiv by 7, while in each of these the state of independ-
ence and intensity was indicated by the addition of an m or n,
and the state of dependence not only by the omission of this
letter, but sometimes also by the dropping of the vowel-
endings. We should thus have seven possibl forms of the
Inchoativ, but in fact the language has chosen to use
commonly only four, namely, the two in wand 2, the vowelless
form and the strengthend form in a (am or an and anna)f.
The modal material thus comprises five forms, the complete,

*The Assyrian offers no real exception, for it is generally agreed that if
it had a true present tense it has borrowed it from the Accadian, or made
it under Accadian influence, so that the tense cannot be regarded as
properly Shemitic.

+ Um, umma, tmma also in Assyrian (Oppert, Assyrian Grammar, p. 50) ;
and there are traces of an t-form in Arabic.

On the Shemitic Verb. 7

katala, and the Inchoativ, yaktulu, yaktula, yaktul, yaktulan
(and its variation yaktulanna). As has been said, we do not
know how the completional distinction came to reside in these
two forms; but, once there, it would easily lend itself to the
expression of temporal distinctions (as it did in some cases,
particularly in later stages of the language, the idea of past
time attaching itself to the Complete or Perfect, and that of
future time to the Inchoativ) ; and so also it came to set forth
those distinctions that are usually specially characterized by
the name “ modal,” those that deal with the subjectivly real
and unreal. We may designate the five forms as follows:
1. the Complete or Perfect (describing an action simply as
finisht) ; 2. the Inchoativ in w, or the w-form (describing an
independent inchoativ or ingressiv action); 3. the Inchoativ
in a, or the a-form (in which the ingressiv act is conceivd
of as dependent); 4. the vowelless or Jezma-form (repre-
senting the action as sharply detacht); 5. the Inchoa-
tiv in an (or anna), or the an-form (the strengthend,
intensiv presentation). But in the four last the essential
inchoate ingressiv signification remains always prominent,
and the shades of difference between them may sometimes
become almost invisibl. Let us look at the actual uses of the
forms in the various Shemitic dialects, taking first the
expression of the several modal conceptions and then the
functions of the several forms. The Imperativ may be
reckond (as it is allied in signification to the Perfect on
the one hand and to the Jezma-form and the an-form on the
other) as a halfway-form between the Perfect and the Inchoativ.

We begin with the expression of modal ideas, and first
command. 1. This is renderd in all the dialects by the
Imperativ in both its forms, the simpl, and the strengthend
or emphatic in an. But the Imperativ is employd in positiv
command only, not in prohibition—a fact the discussion of
which we shal come to below. It is also confined to the
second person. 2. The Jezma-form is employd extensivly
to express command. Thus in Arabic it occurs frequently
with the prefix lz, which indicates a nominal conception of
the verb; dt-yaktul is “‘ for his kiling,” ‘let him kil.” The

8 C. H. Toy,

conception of command, as involving something as yet non-
existent, usually attaches itself to the Inchoativ, and, as
carrying with it decision, naturally also to the shortest,
sharpest form. This Jezma-form is generally used in the
third person, but sometimes in the others; the greater
brevity of the Imperativ suits the greater sharpness of
the direct address to a person. The construction with i
occurs apparently also in Jewish Aramaic in a jussiv or
optativ sense, and Ethiopic has a similar form with da.
Without the prefixt particle this form (which, however,
everywhere except in Arabic represents not merely the original
vowelless Inchoativ, but the others as wel) is used for positiv
command in all the dialects, usually in the third person,
and, with a negativ particle, for prohibition. The Hebrew
distinguishes between negativ command and negativ exhorta-
tion—the former is introduced by Zo, the latter by al. 3. The
long Inchoativ in an and anna (principally in Arabic). With
the negativ Ja it expresses prohibition, differing from the
short form only in being emphatic.

The similarity in meaning between these three forms is
obvious, but the dissimilarity in form is no less obvious. The
short Inchoativ has dropt the original final vowel, while the
long form is the strengthend Accusativ, and the Imperativ
has the simple stem without the preformativ ya characteristic
of the Inchoativ. It has been suggested by some grammarians
thatthe Imperativ is derived from the Jezma-form by dropping
the preformativ, but this explanation is clumsy, and therefore
improbabl. It is simpler to go back to the original form and
meaning of the stem of the noun-verb. The Imperativ is the
simp! noun inflected in gender and number, and, because
used on account of its vigorous brevity in direct address,
confined to the second person ; it is the bare noun or name of
a thing spoken sharply and decisivly. Perhaps the activity
involved in the expression excluded the passiv sense; or the
absence of Passiv Imperativs in Shemitic may be nothing more
than a formal accident, the result of the choice of another
construction to avoid ambiguity of form. Some such fact
also may serve to account for the avoidance of the negativ

On the Shemitic Verb. 9

Imperativ, possibly coupled with the feeling that in such cases
there was less sharpness in the expression. However this
may be, there is no need to suppose the Imperativ a derivativ
from the short Inchoativ. Stil less reason is there for
regarding this latter as derived from the Imperativ. It is
based on the form out of which the latter was developed ;
but the two took their several ways, reaching the same point by
different lines, but always retaining their distinctiv character-
istics. The Jezma-form remains an Inchoativ, and out of the
inchoativ signification develops its jussiv force (we shal see
that this is not its only force) in the way above indicated.
The an-form with a like inchoativ signification comes naturally
to the same use, only with its added emphasis (as also in the
Imperativ there is a similar emphatic long form in an) ;
whatever peculiarity of meaning may exist in the a-form
which is its base, does not appear in this use. We are thus
forced to go back to the original significance of these forms,
and to allow them large latitude within the bounds of the
essential meaning.

Let us next take the expressions of wish. 1. Here we find
the Perfect widely used. In Arabic and Ethiopic it is
employed in the largest sense, of any wish, and quite answers
to the use of the Greek Optativ. In this construction of the
Perfect there is no distinct formal expression of wish; it
merely represents the act as a completed thing (in the
intensity of desire that it shal be) and leaves the precise
meaning to be suggested by the tone or context. In Hebrew,
however, the Perfect expresses only a wish that is known to be
unfulfild—the action stated to be complete is contrasted by
the context with the present reality, and thus recognizd as
non-existent. In another Canaanite dialect, the Phenician, the
usage agrees with that of the Arabic. 2. Arabic employs
the long Inchoativ in an, the emphatic force of which is
appropriate to the expression of wish, in which there is usually
more or less of intensity. It is obvious also that the desired
thing, as in the nature of the case not yet existent, is properly
represented by the inchoativ form of the verb, which exhibits
the action as one just entered on and incomplete; the

10 C. H. Toy,

incomplete stands close to the non-existent. 3. In Ethiopic,
Hebrew, and Aramaic we hav the Inchoativ, which in these
' dialects is at present a vowelless form (except with suffixes).
but is a representativ of all the forms of the Inchoativ,
except the an-form, and therefore does not specially correspond
to the Arabic Jezma-form. Hebrew, however, pursuing its
course of abbreviation, has an apocopate form still shorter
than this last, and employs it in the expression of wish. In
both the Hebrew forms the inchoate sense is the prominent
one; and the brevity of the shorter form is appropriate to
_the energy and excitement of the state of mind involvd.
4. The Assyrian has a special form made by prefixing lu or lt
to the Inchoativ. Whether this is imitated after the Accadian
(which makes a Precativ by prefix ga) is uncertain; such a
precativ form is found in Jewish Aramaic, and is not unlike the
Arabic Jezma-form with prefix 22 and the Ethiopic with prefix
la. In any case, however, the modal force is in the verb-
stem, and the same remarks apply here as above.

These differences of construction exhibit considerabl flexi-
bility in the Shemitic conception of wish, yet are easily
explaind by reference to the original force of the verb-forms.
We can also understand how the Arabic, desiring to bring out
the element of energy and intensity, does not employ the
u-form of the Inchoativ, nor the Jezma-form (having besides
appropriated these to other uses), but confines itself to Perfect
and long Inchoativ in an.

The voluntativ form, that used to express determination of
wil, resolution, is of course the an-form of the Inchoativ.
This use is fully developt only in Arabic; it is, however, not
infrequent in Hebrew (in the first person), and is found in
Assyrian. The relation between form and meaning is
obvious; the Inchoativ suits the non-existent character of
the verb-act, and the energetic an is appropriate to the
natural energy of the thought. Here again we fail to see
any trace of the dependent force of the a-form on which this
longer form seems to be built; there is, however, a discernibl
relation between the voluntativ and the objectiv sense of the
a-form, the latter expressing the object towards which the

On the Shemitic Verb. 11

determination is directed. In Hebrew the voluntativ ends in
a, and is probably the representativ of the old an-form,
the Hebrew throwing away the nunation or mimation, as
it does habitually in singular nouns. It would indeed be
possible to regard this Hebrew voluntativ as the original
a-form, which had not been confined to the dependent sense
that we meet in its use in Arabic; but the analogies obtained
by a comparison between the Hebrew and the other Shemitic
dialects rather point to the former explanation—the Hebrew
verb everywhere shows signs of phonetic decadence.
Purpose is naturally exprest by the Inchoativ, since it
looks to an unaccomplisht object. As telic constructions
are necessarily syntactically dependent, the Arabic employs
for these its a-form, the Subjunctive, which especially indicates
the object aimd at. It has, however, somewhat petrified the
construction, always introducing the verb by /z or some other
particl meaning “‘in order that.’””’ The Arabian grammarians
also insist that there is always an elision of an “ that” after
li; but this is a mere grammatical fancy, the real power of the
modal expression being in the verb, or rather in the combina-
tion of the real preposition /2 and the verb: the expression
“he went li yaktula that he might kil’? means “he went to
kiling.”” In Ethiopic also the shorter form of the verb (com-
monly called Subjunctiv) is employd in this construction,
either alone or after the particles kama and za “in order
that.”” This Ethiopic verb-form, tho now without final vowel,
represents formally all the original unemphatic Inchoativs, but
performs a part similar to that ofthe Arabic a-form. Hebrew,
Aramaic and Assyrian use their own Inchoativ, which also
represents the original three; and there is no means of
determining whether the peculiar Arabic force of the a-form
ever existed in these languages; whether, that is, it was a
part of the original Shemitic material, or is a special develop-
ment of Arabic. In Hebrew, tho the introductory conjunction
is common in this construction, it is sometimes omitted, the
telic form being suggested by the juxtaposition of the words,
and the same omission is found in Arabic with the w-form;
the objectiv nature of the dependent verb, which is elsewhere

12 C. H. Toy,

represented in Arabic by the termination a, is here given by
the inchoativ sense and by the position. The ordinary Arabic
use of dz and the Inchoativ is parallel with the common
Shemitic telic construction of preposition and Infinitiv, and
the comparison between the two brings out the nominal
character of the Inchoativ.

The expression of general result or limit is nearly the same
as that of purpose. Arabic employs the a-form and Ethiopic
its shorter and longer forms, with the appropriate conjunc-
tions. The relation of dependence is as obvious here as in
the telic constructions, but the Ethiopic is freer in the use
of the verb-forms, and shows that the process of petrifaction
had not advanced very far. This long form in Ethiopic
(which has a vowel under the first radical of the stem), tho a
true Shemitic Inchoativ in its function, is of doubtful origin ;
whether its inserted vowel is of nativ production, or is
an imitation after a non-Shemitic language, is not clear.
But in any case its syntactical force is beyond doubt, and
there is no ground in the usage of the Ethiopic for supposing
that this long form carried with it any non-Shemitic idea,
or playd any other part than that of the Inchoativ. The
Arabic further employed the w-form in the ecbatic constrac-
tion, when the conjunction was omitted, whence we must
infer, not that it confounded purpose and result, but that a
certain liberty in the use of the verb-forms existed. The
form set apart as the expression of the object (the a-form)
was employd after the preposition, but in the absence of the
preposition the relations of incompleteness and dependence,
given respectivly by the form and the position, were considered
to be sufficiently expresst in the w-form. Such uses point to
a time when the present stif differentiation of verb-use in
Arabic grammar did not exist.* When, however, it is desired
to represent the result as an accomplisht fact, the Arabic uses
the Perfect. In Hebrew, result is usually expresst by the
construction of sequence, and the verb follows the ordinary
Hebrew laws of sequence.

*A similar phenomenon exists in the Latin use of guum with the Sub-
junctiv.

On the Shemitic Verb. . 13

In the construction of object-sentences (in which one clause
is the object of the action containd in the preceding) the
form of the verb in the dependent clause is determind strictly
by the nature of the thought. 1. Where the action in the
dependent clause is conceivd as really existent in past or
present, the Perfect is used in all the dialects, as after verbs
of saying, seeing, thinking and the like. If the action lies in
the present and is to be represented as continuous, Arabic
permits the use of the w-form. 2. In the case of future
action, after verbs of thinking, supposing, etc., the w-form
and the a-form are found in Arabic; these set forth the action
as non-existent, with the difference that the a-form expresses a
close dependence. The Ethiopic employs its long form (called
in the grammars the Imperfect) in some similar cases, as
after verbs of beginning and ceasing. The Hebrew prefers
the Infinitiv-construction, which is also found in the other
dialects. 38. Where the act of the dependent clause is in the
highest degree unreal, as after verbs of wishing, expecting,
etc., we find the a-form in Arabic (introduced by the conjunc-
tion an “ that’’), and in such cases the Ethiopic has its shorter
form (Subjunctiv) with or without a conjunction.

Conditional and other correlativ sentences show a great
variety of constructions, yet always under the control of the
proper force of the various verb-forms. I. The simplest case
is where the condition or the act of the antecedent clause is
represented as really existent, and the apodosis or consequent
act also real; the rule in this case is that the Perfect shal be
used in both clauses. Where there are seeming exceptions,
they are the result of some peculiar conception of the action
in the mind of the writer. Instead of the Perfect the parti-
cipl is sometimes employd, especially in Aramaic, when it is
desired to express a present or continuous act. II. The usage
is the same when the condition is determind as unreal. The
act is represented as complete, and the context indicates its
true character. III. When the condition or antecedent action
is put merely as a supposed existing fact, or as in general
undetermind and ideal, the form of the verbs in the two
classes depends on the special coloring that it is intended to

3

14 C. H. Toy,

giv the action. 1. It is not uncommon to find the Perfect
in both clauses in Arabic and Ethiopic, by which the condition
and the result, tho from the context obviously future, are put
as finisht or as actually present. This construction is not
found except where the condition is patent and near at hand,
or where for the sake of energy and vivacity the speaker or
writer desires so to represent it. 2. In those dialects that
hav reduced their Inchoativs to a single form, Hebrew,
Aramaic, Ethiopic, the use of this form is the prevailing one.
Hebrew uses its Imperfect in protasis and apodosis; but, in
accordance with its laws of sequence, often expresses the
apodosis by Waw withthe Perfect. Inthis case the Perfect does
not abandon its proper signification; tho Hebrew in its law
of sequence has petrified its constructions, it is always possibl to
recognize the original meanings of the verb-forms, and in
this case the Perfect acts as the same form in Arabic describd
above. In Aramaic the participl often takes the place of
the Imperfect in the apodosis, with a force not very different
from that of the Perfect. The Ethiopic moves more freely,
and varies its verb-forms according to the demands of the
thought, especially in the apodosis. In the protasis the verb
is usually Perfect—the language has chosen as a rule to look
on a condition as something already settled, as a mere
assumd preliminary to the result, and then the time of the
result fixes the form of the verb of the apodosis: if the time
is future, the verb is Imperfect; if past, the verb is Perfect.
But, if the time of the protasis be present, the verb is
commonly Imperfect, in order to express the incomplete
character of the action. 3. The above examples of the use
of Perfect and Inchoativ respectivly to set forth conditions
and results conceivd of as real and unreal or ideal are easily
intelligibl from the nature of the verb-forms. In Arabic we
find further a differentiation in the use of the Inchoativ not
possibl to the other dialects, and especially a peculiar use of
the Jezma-form (the Jussiv of the grammars). This form is
subject to various special rules of use, being, like the a-form
(Subjunctiv) brought into stif connection with certain par-
ticls; but its employment in conditional sentences obliges us

On the Shemitie Verb. 15

to recur to its essential inchoativ signification, and to lay aside
that special jussiv force that has given it its ordinary name.
In sentences in which condition and result are represented as
merely supposed facts we find this Jezma-form sometimes in
both clauses, sometimes in only one, the Perfect commonly
standing in the other. Further the wform and the an-form
are found in place of the Jezma-form. If this makes it
necessary to regard this last as in these cases performing the
part of a simpl Inchoativ (an expression of incomplete,
ingressiv action), it does not prevent us from recognizing
something special in its character and force. Its distinction
from the Perfect, with which it is often brought into con-
tact in these conditional constructions, is clear enuf: the
Perfect represents the act as really complete and present, the
Jezma-form puts it as something just entered on or to be
entered on. We can also understand how it differs from the
an-form, which is always emphatic, and always so emphasizes
the incomplete nature of the action as to locate it in the
future distinctly. The a-form lias its special function of
dependence in Arabic, tho it sometimes leaves this in the
background and brings forward its original inchoativ force.
But how does the Jezma-form differ from the long Inchoativ
in #, with which it sometimes alternates in these constructions ?
Certainly not by any element of command supposed to reside in
it, for if this explanation would serve in constructions where
an Imperativ stands in the antecedent clause, or for the
apodosis in general, it would be wholly insufficient for the
protasis, in which a command would be out of place. Nor
can it be said of the Jezma-form that it interchanges in sense
with the Perfect. It is tru that after the negativ particles
lam and lamma it has what seems tous a present or
proper perfect signification, but the explanation no doubt is
that the language came to regard the action after these
particls as an incipient one, a thing that from the nature of
the thought could not be existent, and yet was thought of as
about to be, on the point of beginning. It is this feeling that
controls the verb-use, and has made the Jezma-form the rule.
This projection of the feeling of a speaker, or of that mas of

16° | C. H. Toy,

speaking that constitutes a language, into the circumstances
of an action is not uncommon, and the particular feeling
may often appear strange to one accustomd to the modes of
thought of a different language. We are so far removed
from Shemitic methods of conception that it may wel be
hard for us to comprehend and explain their grammatical
constructions, especially when they take a petrified shape,
that is, a shape that is doubtles the product of a natural
feeling, but the isolation of which and the absence of free °
movement in the language disguises its force and conceals its
origin. It is so to a great extent, for exampl, with the
Hebrew usage of sequence, in which the verb-forms seem to us
to shift in an arbitrary and surprising way, and in which,
tho we may be abl to discern its general signification, there
remains after our best efforts a certain unknown something.
It would not be strange, then, if we should find it not easy to
explain all the uses of the Arabic Jezma-form, which the
language has evidently dealt with in a somewhat peculiar way.
We cannot explain historically how its construction with lam
arose any more than how the Hebrew use of verbs after Waw
arose, and we cannot determine the precise feeling of the
conditional use of the Jezma-form. But we know enuf of its
application to enabl us to giv a general statement of its sig-
nification. When we observ its use as a jussiv, its employment
after certain negativs in what seems to us to be very nearly a cate-
gorical sense, and its function in some conditional sentences,
we are led to the conclusion that it is the extremest expression
in Arabic of the purely inchoativ sense—it is the most
delicate presentation of that peculiarly Shemitic conception
of an action as being just on the point of beginning, so that
to us it seems to hover over the dividing line between the
existent and the non-existent. Its curtailed form may be
connected with this peculiar significance, either by virtue
simply of the resulting brevity, or by the comparativ isolation
that the absence of the vowel suggests. Whatever may be
the relation between form and meaning, this view of its
signification offers something like an explanation of its uses.
The explanation of the construction with the negativs lam and

On the Shemitic Verb. 17

lamma is suggested above. Its jussiv force may easily come
from its exhibition of an act as being on the point of happening,
as in English the future tense is sometimes used where a
command is involvd. The conditional use follows in the
same way: in the protasis the Jezma-form givs the act as
incipient, and this suggests its immediate occurrence and also
its present non-existence—it thus represents the condition as
a supposd fact, lying near to the speaker and calling forth an
immediate interest. Its range in actual use is wide—it occurs
in constructions that in Greek would include Indicativ, Sub-
junctiv, and Optativ; but it always maintains its own force,
and must be interpreted not according to our usages, but
according to the modes of conception of the Shemitic peo-
ple. The longer Inchoativ form in w does not emphasize
the.idea of incipiency so sharply and delicately as the Jezma-
form, and is rarely used in conditionals. It occurs in the
apodosis when that is separated from the protasis by the
connectiv participl fa “then.” In this case the separation
effected by the particl confers a certain independence on
the second clause, and it adopts the more general expression
of the inchoativ sense. Stil more rarely does the a-form
occur, only where a second parallel verb follows the Jezma-
form in protasis or apodosis, and that under certain condi-
tions in the use of particls. It appears therefore that the
Arabic treats the constructions with the negativ particles lam,
lamma, and the conditional particles im and others, alike,
regarding the action in all of them as a thing imminent, not
existent, but on the point of beginning. It is not, indeed,
confind to this view—the Perfect, as we hav seen, is some-
times employd to vary the conception by representing the
action by anticipation as really existent, and other shades
of meaning are given by the employment of the forms in
wanda. This last construction is rare ; practically the ideal
conditional forms in Arabic are the Perfect and the Jezma-
form, the forms in wu and a being devoted to other uses, and
the selection is based on the significations of the verb-forms
and the conception of the conditional above describd.

We may sum up this rapid view of the modal constructions

18 C. H. Toy,

in Shemitic by a statement of the modal functions of the
several verb-forms. 1. The Perfect is primarily the expres-
sion of an existent complete act, in present, past, or future
time, and.thus covers the ground occupied in the Indo-European
languages by the Indicativ Perfect, Aorist, Pluperfect, and
Future-perfect. But it also performs the part of an Optativ,
the object wisht for being represented by anticipation as
actually in complete existence. In Hebrew it is used in those
optativ sentences only in which the thing desired is located
in the past, and known to be impossibl. Further it is
generally employd in Shemitic in conditional sentences in
which condition and result are known and declared to be
either real or unreal, and also frequently where the condition
or the result or both are put simply as ideal or supposed cases.
It is a favorit conditional form in Arabic and Ethiopic. 2.
The Inchoativ in ~ is commonly employd in what we call the
Indicativ sense, and stands contrasted with the Perfect by
representing the action as ingressiv or incipient in present
past, and future, answering to our Present, Imperfect, and
Future. But as the Shemitic and Indo-European conceptions
of the verb are very different, the former distinguishing only
the completional and not the temporal element of the act,
these two Shemitic forms are in fact used each over the whole
ground of the Indo-European verb, the Perfect often standing
where we should use Present or simple Future, and the
Imperfect or Inchoativ in the place of our Aorist or Future-
perfect. Tho commonly occurring in this Indicativ sense
in Arabic (and it is not found as a grammatically distinct
form in the other dialects), it is used also, as we have seen,
in telic and conditional sentences to express relations of
dependence and subjectiv unreality. Asto the name Indicativ,
it belongs not only to this form, but sometimes to the Perfect
and to the Jezma-form and an-form also. 8. The Inchoativ
in a, modally distinguisht only in Arabic, is devoted to the
expression of relations of dependence, such as ideal result,
purpose, limit, and sequenée. This use flows from its inchoa-
tiv sense, and from the objectiv force proper to it as the
objectiv case of the noun-verb. It looks forward to a point as

On the Shemitic Verb. 19

yet non-existent; when the object of the action is to be
represented as already attaind, one of the properly Indicativ
forms is used, either the Perfect or the w-form, according to
the speaker’s conception of the complete or continuous char-
acter of the action. There is no reason in the form itself
why it should be so rarely employd in conditional con-
structions (as is stated above); usage alone has determind
its restriction to its particular class of constructions; only
an occasional deviation enabls us to recognize the broader
signification that underlies its present special use. 4. The
vowelless or Jezma-form is appropriated to the expression of
command (chiefly in the first and third persons) and to condi-
tional and certain negativ sentences, its form and meaning
permitting, indeed, a wider use, but suiting very wel the com-
parativly restricted range that usage assigns it. Its jussiv
sense passes naturally in some cases (particularly in negativ
sentences) into an optativ. 5. The longest Inchoativ form in
an or anna follows the Jezma-form so closely in signification
that we should naturally think of it as a derivativ from the
latter, but for its vowel a, which rather connects it with the
a-form. It is, however, a true Indicativ in the first instance,
and often acts as an emphatic extension of the form in w, tho
always as a Future. Its uses in prohibition, wish, and in con-
ditional sentences are to be explaind, as above, by its inchoa-
tiv sense, to which is always added the emphasis proper to its
form. It is an emphatic Imperativ and Optativ. In Hebrew
it occurs in a fragmentary way as a Voluntativ ( the so-called
Imperfect with paragogic a). 6. The proper Imperativ of
the second person has already been mentiond, the nearness
of its relation to the form in an and the vowelless form
pointed out, and reasons given why it should be regarded,
not as aderivativ from the latter, but as an independent
formation, which has advanced in its own way to a point
nearly identical with theirs—nearly, but not quite, for there is
a perceptibl difference in the coloring of the command as
given by the different forms; the Imperativ simply states the
act (or, more probably, the actor) as an object of thought, and
leaves it to be inferd from the tone that it is to be done, and

20 C. H. Toy,

is thus more peremptory than the others which represent the
act as something that is about to be done. There is a sim-
ilar difference, as is remarkt above, between two imperativ
constructions in English, and so also there is a difference in
the coloring of two Shemitic expressions for the present, one
of which uses the Perfect, and the other the Inchoativ.

Most of the Shemitic dialects now possess only one form
for the three unemphatic Inchoativ forms found in Arabic,
and this is without final vowel, not answering to the Jezma-
form, but representing a merging of the three into one.
This form must execute the functions of the original three.
But under the stres of this poverty, various languages have
created new forms for special purposes, or have devoted to
special uses the forms that arose from phonetic usages or
from imitation of other languages. Hebrew has made a short
Inchoativ by dropping a final consonant, or reducing a vowel,
and employs it as a jussiv or optativ, and without the prefix
wa. Ethiopic, on the other hand, has a lengthend form
gotten by inserting a vowel a@ under the first radical, which
expresses the incomplete in present, past, and future (answer-
ing in general to the Arabic wu-form), while the older,
shortend form is used in the expression of command and wish
(somewhat as the Arabic forms in @ and without final vowel)
in dependent and independent sentences; but no very sharp
difference between these two is maintaind as in Arabic, the
Ethiopic preserving a considerabl freedom in the employment
of its forms. The Assyrian Precativ (made by prefixt Zu or
li) has a distinct function, and its relation to the original
Shemitic scheme is obvious. In Aramaic, particularly in its
modern dialects, the old modal expressions hav been largely
expunged by the use of the participl. The Amharic shows
nearly the same modal development as the Ethiopic, using the
‘old shortened form for command and wish, and the lengthened
form in telic sentences, while the Tigrifia exhibits a more
extensiv employment of the lengthend form in conditional
sentences than the Ethiopic, which shows a preference for the
Perfect.

It may reasonably be inferd from the examination of the

On the Shemitic Verb. 21

existing Shemitic languages that the modal material of the
primitiv Shemitic was about what we find in classic Arabic.
It is certain that the five forms above discust were in the
mother-tongue, for they can be traced in all the members of
the family. It may be that there were others not now found
in Arabic, such as the forms in wmnma and imma, of which
there are traces in Assyrian, and the simple Inchoativ in 2,
remains of which are found in Arabic. We should in fact
not be surprised to find that the seven or eight forms of the |
Singular noun, together with the Perfect, constituted the origi-
nal modal material; but if this were so, the language early
dispenst with all but the five, by means of which it was abl
to expres its modal ideas with sufficient distinctness. It is
certain that the mother-language exprest these distinctions,
since the identity of modal development in the various
dialects could not be otherwise accounted for.

We hav almost no data for tracing the historical genesis
of the modal expressions. We are warranted in holding that
the verb-forms began as nouns and noun-verbs, and that the
modal development proper began at a time when the cases of
the nouns were already in existence. The mode-expression,
however, started from the completional difference of the two
main forms (one with and the other without preformativ), the
only mode-difference that has held its place in the Shemitic
languages. The Perfect naturally connected itself with the
idea of the real; the Inchoativ, with that of the unreal.
Beyond this point it is not certain how far the mother-
language advanced. If we could suppose that the original
state of development has been preservd in Arabic, we should
hav to say that the primitiv Shemitic had so differentiated
the forms that the a-form was devoted to the expression of the
relation of dependence, the vowelless form to command, wish,
and the most delicate shade of the inchoativ conception, and
the long form in na or ma to emphatic assertion, command,
or wish, while the more general expression of the unreal was
assigned to the w-form, and the real, with connected optativ
and conditional uses, to the Perfect. On general grounds
this may be considerd probabl, but the absence of the modal

4

22 C. H. Toy,

a-form in the other dialects leads us to leave the question
undecided—it is possibl (as is suggested above) that this form
is a special creation of Arabic.

We have treated the Perfect as a proper modal form, tho it
is usually said in the grammars that the modal development
attaches itself only to the Inchoativ. This is perhaps
nothing but an affair of phraseology, but the Perfect has a
function as truly modal as the other. The Shemitic did not
originate special agglutinations for its modes. It. took its
derived noun-verb (made by a preformativ ya from the simple
stem) and used its cases for the expression of modal ideas;
these cases in all probability at an earlier stage played the
part of mere nouns, and, as they advanced to the verb-state,
gradually and naturally transformd their case-relations into
mode-relations. Similarly, the Perfect, which was also a
noun (with pronouns attacht) without diversity of case-
relations, transferd its nominal conception of completedness
and transformd it into the corresponding mode-relations.
It seems to work as real a confusion of ideas to confine the
name Mode to the Inchoativ, while the Perfect is called an
Optativ and a Conditional, as it was, according to the old
nomenclature, to call the two main forms Preterit and
Future, explaining that the first was also a Present and
Future and the second a Preterit and Present. The Shemitic
mode-development went out from Shemitic conceptions, and
our terminology must be made to conform to these concep-
tions, not only for the sake of grammatical exactness, but also
that we may learn to comprehend the true shades of meaning
exprest in the literature of the language.

From what has already been said it is clear that the genera]
tendency in the Shemitic languages has been to drop formal
mode-distinctions, and indeed to a compression of all the
senses flowing from the Inchoativ into the shortest or Jezma-
form (leaving, however, the Imperativ unaffected). In the
other direction a compensating process of modal formation
has also been going on, but to a less extent.

Ancient Arabic retaind or developt the fullest modal
material, whether precisely the complete primitiv material,

On the Shemitic Verb. 28

cannot be certainly said. In the earliest remains of Hebrew,
reaching back perhaps 1200 or 1300 years B. C., only the
Jezma-form is in ful use, the u and a-forms being preservd
only in connection with suffixes, and the am (or an) form in
a petrified state as a Voluntativ (similarly the noun-forms
in w, 2, and @ exist in classical Hebrew only in a petrified
state, and with suffixes, the Jezma-form being the common
one). The further shortening of this form into the so-called
Jussiv is another illustration of the tendency to abridgment,
and helps us to understand the prehistoric decadence of the
Inchoativ. Of course Hebrew, tho it dropt the forms, retained
the ideas, economically reducing the material of expression to
what it considerd the minimum, tho afterwards obliged to
create a new form suited to its peculiar needs. There is a
further step in post-biblical Hebrew, where the main forms
have largely sunk the original completional in the derived
temporal sense, and the flexibility of the modal expression
has suffered corresponding diminution, the Perfect being
appropriated to all real and the Inchoativ to all unreal con-
ceptions.

The Assyrian in its earliest known stage shows apparently
less formal degradation than the Hebrew in the Inchoativ,
inasmuch as it retains the forms in « and a as wel as the
Jezma-form. But it seems to hav quite lost the sense-
distinction of these forms. It has also maintained the
ma-form (and indeed more fully than the Arabic in the three
cases umma, amma, imma), but its ordinary Inchoativ is the
Jezma-form as in Hebrew. Its new formation of a Precativ
has already been mentioned. The curious question, whether
it had a Perfect in historical times, must be considered as yet
undecided. It certainly brought this form from the mother.
tongue, and if its remains do not show it, we may conclude
that the Perfect was dropt either from unknown syntactical
considerations peculiar to the Assyrian, or through the influ-
ence of another language. As its optativ expression is
assigned to a peculiar form (the Precativ), it has practically,
as far as is now known, comprest its modal material into
the Jezma-form.

24 C. H. Toy.

Pure Aramaic does not appear as a literary language til
after the begining of our era, and then shows the same
general state of form-degradation as Hebrew and Assyrian,
without having developed, like Hebrew, a shorter Inchoativ, or,
like Assyrian, a Precativ. There is, indeed, in biblical
Aramaic a precativ and future form made by prefix /, but
there is no trace of it in classic Aramaic and it is probably a
peculiar Jewish form, either a dialectic modification of the
Aramaic Inchoativ (with preformativ n) or a combination of
the preposition 7 with the verb. Aramaic has, however,
compensated for the loss of the original verb-forms by the use
of periphrastic (participial and other) expressions, and in
general by advancing towards an analytical structure.

The process of abridgment has been carried by the Ethiopic
(whose earliest written remains belong to the fourth century
of our era) even farther than by the Hebrew and the Assyrian,
but its new lengthend form describd above has brought it
back nearer to the original Shemitic modal development. In
the other members of the African branch there is no new
modal material, except that the Ambaric shows, like the
Aramaic, a disposition to adopt compound periphrastic forms,
and an analytical structure.

Modern Arabic shows about the same stage of formal deg-
radation as ancient Hebrew, and its modal expression has
been modified accordingly, and the modern Aramaic dialects
exhibit an exaggeration of the tendencies of the classic
language.

It appears, therefore, that the loss of primitiv forms and
the origination of new forms took place in many of the
Shemitic dialects before the historical period; when they
appear as written speech, they hav already traversed a long
course of growth, decline, and new growth. The primitiv
tongue of the family developt a respectabl set of mode-
forms out of very simpl material, and the dialects hav
curtailed these til there has been left the smallest possibl
subjectiv element in the formal verbal expression. A minute
examination of the modal expression in any one of the dialects,
as Hebrew (which does not belong to the design of this

On the Nature of Caesura. 25

paper), would nevertheless show a considerabl power in the
expression of delicate subjectiv shades of thought, not by
distinct forms, but by the suggestions arising from the main
completional element of the verb. This element itself may
be considerd a peculiar Shemitic modal conception, or at
least developt in Shemitic speech to an extraordinary
degree, and permitting very delicate distinctions of thought.
By it the language is enabld to characterize an action as
finisht, or as just entering on existence and in all the stages
of incompleteness. It has seizd on and formally fixt the
period of ‘ becoming,” the stage of advance from non-existence
to existence, and has thus given a peculiar dramatic coloring
to its ordinary style, while it has groupt around this idea
the various conceptions of the ideal that constitute the
material of modal thought in our family of languages. These
last it has in common with other tongues; but the funda-
mental conception of completional distinction may be regarded
as the Shemitic contribution to the modal material of speech—
a conception that it has workt out more fully than any other
linguistic family.

TI.—On the Nature of Caesura.

. By M. W. HUMPHREYS,
PROFESSOR OF GREEK IN VANDERBILT UNIVERSITY.

While this paper presents an independent discussion of the
nature of caesura, it is so shaped that it also serves as an in-
troduction to the following paper On the Effects of Elision. —

1. Caesura in general serves two purposes. (a) One of
these is to allow the reciter in long verses to catch his breath,
in such a way, however, that he shall not be permitted to
pause too long for the purpose; and accordingly, in such
verses, we usually find a pause at the proper place, or at least
the liberty of making a pause without impairing the sense.
In the latter case occurs a slight xpévoc xevde or tempus inane,
which may fall even between words closely connected. This

26 M. W. Humphreys,

use of caesura is not necessary in iambic trimeters, and con-
sequently the caesura may fall where a pause is not to be
thought of, and very frequently falls where the sense, though
permitting a pause, does not require it. (6) The other gen-
eral use of caesura is, not to separate, but to link together the
two halves, or rather principal portions, of the verse. If the
verse is divided exactly in the middle, it at once falls apart
into two shorter verses; and if every word-foot constitutes a
verse-foot, the whole verse falls to pieces, very much as a
brick wall would do, if the bricks were laid the one exactly
upon the other, without any over-lapping. A thread-bare
illustration is afforded by
sparsis hastis longis campus splendet et horret.
An example in trimeters is Agam. 948: °
miQuv" Kparoc pévroe wapec vy’ Exwy époi.

Hence, somewhere near the middle of the verse, a word must
end in a foot, so that the foot, which the reciter feels to be a
unit of the verse, may connect the two portions together. This
. prevents the reciter from pausing too long; for if he did so,
he would destroy the rhythm of the foot in which the caesura
occurs. And so with the other feet in the verse: the more
numerous the caesurae, the more vigorous will be the recita-
tion, as they prevent too great a pause and so insure care and
attention on the part of the reader or reciter. A better way,
therefore, of indicating caesura, would be to use a vinculum of
some sort, thus: ~ —~— ~~ —~— ~ —, rather
than ~ —~—~]|--~—-~—w~—. For the sake of
convenience, however, I retain the usual method. So it is not
to be wondered at that we find many verses which require no
pause from beginning to end. To say that such verses have
no principal caesura is, in the first place, to beg the question ;
and, in the second, to overlook the fact that these verses regu-
larly have a caesura at the place for the principal one. Of
course I do not mean thereby to say that, for instance, Aristo-
phanes did not write verses without any main caesura, for he
certainly did, and frequently allows diaeresis after the first
dipody to pass for the chief caesura—a thing not unknown
in the Tragedians, especially in Aeschylus, where it is quite

On the Nature of Caesura. 27

common. But this is not so frequent as is usually supposed,
for it is erroneously assumed that the principal caesura must
be at the longest pause. Still another use of caesura, with
which the present discussion is not concerned, is to conceal
the cause of the pleasant rhythmical effect of verse.

Let us now proceed to illustrate the whole subject. The
examples cited are not exhaustive, but merely such as I picked
up when reading for other purposes. We find caesura between
the subject and the verb, as Agscu. Theb. 15:

Bwpoter, repag || py ’EarecpSivai xore.
Between the verb and its object, ibid. 270:

Sapaoc pidorc, Avovea || roAcpiwy poor.
Between an adjective and its substantive, ibid. 19:

éSpeWar’ oixnrijpag || aorcdnddpouc.
So Alcest. 513, 856:

Sarre rev’ év 7790 || hépg péddw vexpov.

kairep Bapeig || Euppopg wexrAnypévoc.
Even between the article and its substantive in various rela-
tive positions, Choeph. 658, Philoct. 964, Hel. 708:

&yyedAe rotor | cupiow dwparwr.

#6n ’ari cai roic rovde || rpooywpeiv Noyorc.

ovy de poySwy || rev ev "Tig BpaBevc.
After a preposition, Oed. Rex 615, Troad. 946, 1211, Iph.
Taur. 1174:

kaxov O€ xav év || duepg yroing pug.

rl 83) ¢povovo’ Ex || Swydarwy ay’ éoxcpny.

rys@my, ov é¢ || rAnopovac Inpwpevor.

“AroAXor, ovd’ Ev || GapBapors rd’ HAmo' av.
Even after ov, Iph. Taur. 684:

koux Eo Sxwe ob || xpy Cuvexrvevoai pe cor.
Before postpositive words (év, yap, etc.), Orest. 360, Eur.
Elect. 35:

"Ayapepvovog ev | yap ruyag yriordpny.

Sdpapra, rarépwy || pev Mucnvaiwy aro.
So Eumen. 478, Hec. 549, 736, Heracl. 39, 729, 743, Herc.
Fur. 69, 1126, 1396, Iph. Taur. 96, 955, 1161, 1379, Iph. Aul.
425, etc., etc.

28 M. W. Humphreys,

Caesura, even between an enclitic and the preceding word,
is better than no caesura: thus. Ion 574, Eur. Suppl. 727, Iph.
Taur. 696, Choeph. 181, 738, Antig. 1256:

éyw 9 Groiag || pat yuvacog éképuc.
O¢ Ev re roic. decvotaiy || gorcey GAcepoc.
KTnoapeEroc, iv EdwxKa || cor Odpapr’ Exe.
ovx Hoaor evdaxpura || poe Néyerc rade.
Aven 8 GpaIo¢ || Eori coe Evvéuropoc.
kal rijc G&yav yap || ori rov aryiic Bapoc.
Before postpositive wc, Theb. 53, Antig. 256:
Emvet, Neovrwy || dc “Apn decopxdrwr.
Aexrn 0 &yo¢ pEvyorroc | Oc EXHY Kore.
It is not necessary to multiply examples of caesura between
words closely connected. Suffice it to observe, that if we
reject these caesurae, we shall have a vast number of verses
without any main caesura, almost all of which have this sort of
caesura. This cannot be attributed to chance. And to fur-
ther strengthen my views, I shall adduce some illustrations
from other verses. Noone will deny that the trochaic tetra-
meter catalectic of the Tragedians requires diaeresis after
the second dipody. There is one apparent exception to
this,—Agscu. Pers. 165; but this will be explained in my
next paper. Although this diaeresis is required, and corres-
ponds in a certain way to the main caesura of the iambic tri-
meter, still it takes place between words closely connected,
and that, too, in spite of the considerable length of the verse.
I give here a few of the numerous instances of this: Troad.
451, 454, Ion 530, 1252, Iph. Aul. 871, 877:
® arégy rov grradrov por || Sea», ayddApar’ ena.
db Joatc aipaic pépeaSai||aoe rad’, & parted’ avak.
kai ré pot Ac~ecc; warnp adc || cipe cai ov waic eudc.
topev, & TdAdatva, rag aag || Evpdopac, iv’ et rvyne.
WO Exe Kai gol per evvore | cipi, op O fooov rae.
aprippwv, rAqnv é¢ oé Kai any || raida’ rovro & ov gpovel.
See also Iph. Aul. 860, 868, 1342, 13867, and passim. In
some of the above examples the division takes place between
an enclitic and the preceding word. This is not so strange as
might appear at first sight; I can produce examples of a

On the Nature of Caesura. 29

grammatical pause immediately before an enclitic, as Androm.
T47, Sopa. Elect. 647-8:

Hyryou, réxvov, pot devp’ ix’ wyxadate oradeic.

kai ph, pe AOUTOV Tov Rapdvrog Et TLvEC

ddAowe Bovrevovary éxHadeiv, pac.
Some punctuate the last example differently ; but unquestion-
able are Hec. 432, Hel. 1166, Heracl. 78, Antig. 544, etc., un-
less the vocative is read without a pause.

An enchtic can even stand at the beginning of a verse when
the preceding verse is closely connected, as Heracl. 280-81 :

Aaprpoc 0 axoveag ony UBpv davijoerai
oo xai roXiratc Yi TE THOE Kai Huroic.
Dindorf, however, writes gavijeerat | coi xré.

Similarly it may be shown that a proelitze admits a grammati-
cal pause after it, from which fact it is evident that it was not
a necessity that it should be read as a part of the word following
it, and hence could admit caesura after it. Examples of pause
after proclitics are Sopn. Elect. 348-9, Phoen. 1280-81:

firec N€yetc prev Gpriwe we, ei Aaac
aJévoc, TO TouTwY pcos ExdetLecac av. —
émevy” emecye, Suyarep’ we, iy per Iaow
ratoacg TPO AdyxNS, ovpoc Ev Paee ioc.
A proclitic may also stand at the end of a verse, as Plut. 878.

To illustrate further the fact that caesura may take place
between words intimately connected, I shall now cite some
examples from Latin poets. In Horace, who certainly did not
neglect the main caesura, we find, Epod. V, 88, XVII, 6, 13, 36:

sub haec puer iam ]j non, ut ante, mollibus.
Canidia, parce | vocibus tandem sacris.
postquam relictis | moenibus rex procidit.

> quae finis aut quod | me manet stipendium;

to say nothing of the well-known verses, Epod. XVI, 8, I, 19,
XI, 15:
parentibusque avominatus Hannihal.
ut assidens znplumibus pullis avis.
. quod si meis ¢7aestuet praecordiis.
In Catullus, who never neglects caesura anywhere else, we
find, 1V, 18:
et inde tot per | inpotentia freta.

5)

30 M. W. Humphreys, —

Of course this sort of caesura is not to be expected so much
in the dactylic hexameter, on account of the length of the
verse, and the consequent desirableness of having a pause in
it. Still many instances do occur, as Hor. Sat. I, 4, 2-5;
Kpist. I, 11, 21; 10, 14:

atque alii, quorum | comoedia prisca virorumst,

siquis erat dignus | describi, quod malus ac fur,

quod moechus foret aut | sicarius aut alioqui

famosus, multa j cum libertate notabant.

Romae laudetur | Samos et Chios et Rhodos absens.
novistine locum | potiorem rure beato.

In the Homeric Poems are found many such verses as Il. N,
49, 71:

GAAn pev yap Eywy’ || ob deidta xeipac aaxroue.

txvia yap permease || rodwr Oe kynpdwr.
And even in the so-called pentameter, whose incision (with two
or three peculiar exceptions) is invariable and fixed, it may take
place where there is no grammatical pause, as CaTuL. 84, 12:

iam non Ionios ] esse, sed Hionios.

ARCHIL. 16, 2 (Bergk):

kiovac, & peyaAn || yal’, drévepSev Exec.

2. There is a commonly received error that caesura only
takes place where a polysyllabic word-foot extends across the
space between two verse-feet, and ends in the latter of them.
This is utterly false. Caesura is where a word-foot, be it
monosyllabic, dissyllabic, or polysyllabic, terminates in a
verse-foot. There is a tendency, it is true, to avoid diaeresis,
and especially strong is this tendency in that part of the verse
where the main caesura is necessary. Consequently diaeresis
in the dactylic hexameter is not very frequent after the second
foot, but when it does occur there and a monosyllable follows
it, the verse has both diaeresis and caesura. To say, then, that
caesura excludes diaeresis, or rather includes the absence of
diaeresis, is to confuse the whole matter sadly. Besides, not
a few instances (although, to be sure, not very many) occur
even in hexameters, as Hor. Epist. 1,6,40; 7,16; 7,52, etc.:

Ne fueris hic tu. | Chlamydes Lucullus, ut aiunt.
Tam satis est. At tu, | quantum vis, tolle. Benigne—.
Demetri, (puer hic | non laeve iussa Philippi—).

On the Nature of Caesura. ° 31

And they are especially common in the Satires of Horace, as

I, 1, 8, 18, 28, 32, 40, etc., etc. This is most usual in Latin

where the monosyllable is preceded by elision, as Epist. I, 2, 8:
Stultorum regum et] populorum continet aestus ;

in which case the monosyllable is often closely connected with

the word after it.

But what is of especial importance for the present discus-
sion, in iambic trimeters the caesura is frequently effected by
means of a monosyllable, as Orest. 662:

Wuxny & épny dd¢ || re radaurapy xarpi. ~

In the 801 iambic trimeters of the Alcestis this occurs 117
times ; and even if we omit instances where the monosyllable
is enclitic, or a monosyllable preceding it is proclitic, there
still remain 71 instances; that is, the matter was left to take
care of itself. So in the last Epode of Horace, containing 81
verses, there are 7 instances, verse 30 being an appropriate
specimen : ;

Quid amplius vis ? | O mare et terra, ardeo.

3. It is a happy circumstance that G. Hermann denied that
caesura in the fourth foot of iambic trimeters was ever to be
regarded as the main caesura. This one view of Itts justifies
us in disagreeing with him in anything we please, if we can
support our views with arguments. But when J. H. H.
Schmidt denies that the caesura in the third foot is of any
importance, it is time to begin to get out of patience. In his
Leitfaden he says: “Unter 100 Versen pflegen etwa 50
Theilung (i. e. hephthemimeres), etwas mehr als 25 Kin-
schnitt (i. e. diaeresis in the middle of the verse), etwas
weniger als 25 Bruch zu haben. * * * Die sonst angenom-
menen Gliederungsarten, welche dem Rythm widerstreiten,
haben keine wesentliche Bedeutung.”’ To reply to this would
be like arguing with a man who insists that twice two is five.
Schmidt has thrown great light upon the reading of Lyric
Poetry, but he should not have tried to make everything lyric.

Iil.—On Certain Effects of Elision.
By M. W. HUMPHREYS,

PROFESSOR OF GREEK IN VANDERBILT UNIVERSITY.

Although some of the facts which are discussed in this
paper have been mentioned by others, still it may not be out
of place to state that all of them were observed by myself
before I knew that attention had been called to them, and
that I have arrived at all my conclusions by independent in-
vestigation.

§ 1. QUASI-CAESURA IN GREEK.

Having once had occasion to find examples of iambic tri-
meters without any main caesura, I observed that nearly all
such had an elision so placed that if the elided vowel were
pronounced there would be caesura. This, as I have since
learned, was observed by Porson, who calls it quasi-caesura
(a name which I adopt), but offers no explanation of the
phenomenon. This elision may take place either at the end
of a polysyllable, as Ajax 435:

ra mpwra kadXore? || avtsrevoac orparoi,
or in the postpositive or enclitic monosyllables re, ye, pe, dé,
etc., as Theb. 538:

ob pujyy axopraorog Y Egiorarae woAate.
In such instances I have no doubt that the Greeks (who ordi-
narily made their elisions ¢otal) slightly pronounced the elided
vowel, so that the effect of caesura was in some measure pro-
duced. But to this view there seems at first sight to be an
objection. When the reader had come to the place for the
penthemimeral caesura and found elision instead, how was he
to know whether to make his caesura there or not, as the main
caesura might be hephthemimeral? That is, in such a verse as,

Qo 9

kayw padove’ Edel’, 6 0 éoabdn povos,
how was the reader (and especially the reciter, for in reciting
you cannot think ahead) to know whether he was to make
caesura by slightly sounding the -a of paSvtea, or was to wait

On Certain Effects of Elision. - 33

for the hephthemimeral caesura? This difficulty would evi-
dently debar writers from employing quasi-caesura at the end
of the first dipody under ordinary circumstances, and hence
we actually find that it is admitted as the equivalent only of
the hephthemimeral caesura. Certain exceptions which were
to be expected will presently be explained.

It should be observed that verses with quasi-caesura seem
to have diaeresis in the middle, which would be a grave fault.
Now G. Hermann, who rejects the caesura in the fourth foot,
must of course reject quasi-caesura also, and this he does by
attributing it to chance. Accordingly he asks how it came >
that Aeschylus and Sophocles neglected the elision (when
they used diaeresis in the middle) more than Euripides, who,
he says, was so much more careless than they. We ask in
reply how it came that there was any great difference between
the usage of Euripides and that of the others, if it was all due
to chance. And besides, a verse containing what Hermann
regards as an unsuitable substitute for caesura, might well be
made by one whom he considers an inferior metrician. But
I deny that Euripides was more careless than the others. In
fact he is in some respects the most polished and versatile
metrician of the three. His frequent resolutions which give
variety and life to the verse, being subject to strict limitations,
are no evidence of carelessness or of deficiency. Moreover,
he does not admit the quasi-caesura more frequently ; whereas
the diaeresis in the middle, without elision and without any
main caesura elsewhere, the rest sometimes admit, but
Euripides virtually never.

Hermann’s position demanded of him to show that there
were many: verses entirely without main caesura, and in at-
tempting to do this (Elementa, p. 111) he produced the follow-
ing supposed instances from Oed. Rex: 326, 449, 598, 599,
615, 788, 744, 785, 809, 1290, 1476. But of these eleven
verses, four have the quasi-caesura, and one more (449) has the
ordinary hephthemimeral caesura:

AEyw Cé cor, roy avdpa || rovror ov mddat,
and in another (508), airoto: rayra is to be read for airoic
ararra: |

TO yap TuyxEtv ubroter || ravr’ évravd’ En.

34 -, MW. Humphreys,

(Some write zay for rav7’.) And so the number of verses
wanting caesura is reduced from eleven to five; and one of
these five (615) has a break between a preposition and a noun,
which is not the same as no caesura at all; for if in this
instance we regard the preposition and its object as one
metrical word, the third and fourth feet would then be made
up out of one word:

kaxov b€ kay | év Hueog | yvoine pea,
which Hermann himself in another place correctly regards as
the worst sort of verse. Still another of the five has a break
between an enclitic and a word preceding it, which break is
better than total absence of caesura for the same reason that
applies to the verse just mentioned, for if the two words form

one metrical word, we have
| capa duxdoig | cévrpooi pov | xaSixero.
So that there are really three verses without main caesura, and
four with quasi-caesura ; or even allowing him to count the
two I have just mentioned, the ratio is still only five to four,
and that too in a play that seems to have been especially
selected for making the ratio seem great. Now if the four
out of nine instances suffer elision by chance, and the same
ratio is sustained in other plays, we must conclude that regu-
larly four words in nine suffer elision, which is not true. But
in fact, the other plays of the Tragedians not only sustain the
ratio in favor of elision, but show that the instances of elision
vastly exceed those of its absence in such verses.
Hermann, further opposing quasi-caesura, compares the

verses

GX’ oy wédeg oTVyEl, ov TYyhoec VEKPOY ;

drav yap ev dpovic, rd9’ iyyijoet ov vy,
with these,

KevrTetre, py petoegd” eyw"rexov apex.

yuvati rapdévore 5’ awd/3AenTog pera,
affirming that they are of the same sort. In reference to the
two former he says: ‘ Who will believe that the actor re-
cited them in any other way than as was suited to the sense?
for if rhythm and not sense is to be observed in reciting,
there is no reason why breaks should not be made even in

On Certain Effects of Elision. 35

the middle of words” (a thing, by the way, which he himéelf
had just done on the same page in discussing another sub-
ject). Accordingly he regards the main caesura as being
replaced in all these verses by diaeresis in the middle. With-
out denying that this ever happens, I think that, although
in the above two verses which have no elision we place a—
comma after the third foot, still the actor, in order to em-
phasize od and ré3" and so express their relation to réAy and
éray, made his caesurae respectively after of and 769’; for if we
read these verses according to the sense we are almost compelled
to make that sort of rhetorical pause and change of tone which
best suits caesura. And then we may ask in turn who will
believe that the actor recited . . . geidecS’: éyw .. . without
pronouncing the elided vowel, especially as Hermann himself
requires us to make a pause at such places.

In order to prove beyond all doubt that this elision at the
middle of the verse cannot be attributed to accident, I shall
first collect from one play of each Tragedian the examples of
verses that have no main caesura, but have diaeresis in the
middle, and see how many of them have elision at the diaere-
sis; and then I shall give the results of a similar examination
of all the extant tragedies. And in so doing I shall first take
account of those verses where the sense seems to require the
chief pause to be at the diaeresis, whether there be caesura or
not, and then I shall drop out the verses which contain a
caesura of any sort in the third or the fourth foot, so that
there can be no misunderstanding as to what I mean by
caesura. I shall, therefore, temporarily place caesurae at
places where I do not believe they belong. I use Dindorf’s
text.

First, then, 1 find in Arscg. Theb. the following:

1. With elision after the third foot :

252: ob &¢ PIdpov ovyoo || avaoynoee rade.

385: cele, Kpavouc yairwp’, || ix’ | dowidoc dé rg—
410: repdvra cai arvyovrd' || ixépppovac Adyouc.
426: mipyots 8 awedei | deiv’, | a | pom xpaivor rin.
544: we wAetor’ éx’ avdpi | rgd’ || idwrec Sat BEAD.
562: Sedv Jedovrwv | av || dAnSedoap’ eyo.

36 M. W. Humphreys,

° 635: ddAwompov maar’ i ereiaxyaoac.

637: % Zarr’ aresacrip || Gxwe avdpnrarny—

799:” cadwe Exec Ta | wAetor’ | év | & wvrdAdpaorw—
1005: doxovvra cai ddfav7’ || awayyédAEey pe ypip—
1007: ’Ereoxhéa pev | rove’ | éx’ | edvoig xSovdg—
1012: obrw per apgi | rove’ || éxéoradrat Aéyerv.
1053: add’ abropfovadrog | iad’, | arevvérw 0 éyw.

2. Without elision :

457: xat py rov évrevdev || NXaxdvra mpoc wvAacc.
632: détw, rov abrov | cov || caciyrynrov, roAE—
695: @idrov yap éxSpa | por || rarpoc raAar’ dpa—
702: YJeoig per Hon | wwe || rapnperdHpeda.

1046: GAN’ ov woreg orvyei, | ov | Tipo TAagy.

Here we have thirteen instances with elision against five
without it. Now let us drop out all the verses that have any
break in the third or fourth foot: in the first group, 385, 426,
544, 562, 799, 1007, 1012, 1053 being dropped, five remain ;
in the second group all but the first’ being dropped, only one
remains ; and it will be observed that in 682 cov might well be
written, and I have already shown that caesura belongs after
ot in 1046. In 695 gidov goes with wrarpéc and in 702 xwe
modifies the word after it, so that these two verses, at any
rate, must be read almost continuously; and hence the rejec-
tion of all four of these verses is proper enough, while most
of the verses rejected from the other group would have to be
strained in order to place the caesura elsewhere; and seeing
that elision at the diaeresis certainly answered for caesura, we
should be justified in retaining the whole list as instances of
quasi-caesura.

Secondly, I find in Sopa. Antig. the following:

1. With elision after the third foot, vv. 44, 57, 74, 77, 80,
307, 399, 407, 408, 473, 515, 658, 732, 7383, 764, 1012 2.—
sixteen in all. The peculiar verse 544,

Piro, Kaovyvirn, pe aripayc TO pi} ov—,
which really belongs here, I omit entirely, as it might be dis-
puted.

2. Without elision, vv. 827, 518, 555, 723, 899, 997, 1021,
1073,—eight in all. In 55 dvo is to be read with pia, and in

On Certain Effects of Elision. 37

71 oo should be emphatic, and is read with doxei, and 718
should begin aA’ eixe Suug; so that I omit these three verses.
Rejecting, as before, in the first group 307, 407, 473, 515,
658, 733, 764, 1012, we have eight left; and in the second
group we reject all but 1021, thus leaving the ratio 8:1.
And in 327 (continuous), 555 (emphasis on Zjr), and 899
(sui emphatic), the removal of the caesura from the middle
is not so violent as it is in any of the first group; conse-
quently the above ratio should really be twice as large.

Finally, from Eur. Elect. I gather the following:

1. With elision after third foot, vv. 4, 14, 31, 64, 78, 96,
284, 305, 382, 504, 510, 555, 570, 642, T70, 782, 837, 980,
1008, 1012, 1036, 1065, 1087, 1262,—twenty-four in all.

2. Without elision: 438, 248, 1042, 1094,—four in all.

In the first group we reject 770, 837, 980, 1003, 1036, 1262,
and in the second group all, leaving eighteen against none.

And yet these elisions at the middle of the verse, when
main caesura is otherwise wanting, have been attributed to
accident! These three plays arc quite enough to show that,
whether the author was conscious of it or not, he allowed this
sort of diaeresis to pass for caesura. But:I have gone fur-
ther and examined all the plays of the three great Tragedians,
-omitting the Cyclops, but including Rhesus; and the follow-
ing tables show the result:

1. Including all possible cases :

With elision: Withoat elision: Percentages :
Aeschylus, 89 instances; 39 instances, 69+ 31—
Sophocles, 150 i 53 “ 74 26
Euripides, 315 “ 101 “ 76 24

2. Excluding all doubtful cases:

With clision: Without elision : Percentages :
Aeschylus, 42 instances ; 19 instances, 69— 31+
Sophocles, 44 “ 9 “ 83 17
Euripides, 123 “ 1(%) ‘S 9 01 (7?)

I have omitted Aristophanes because of the uncertainty of
the caesurae in Comedy.
In excluding doubtful cases, 1 took no note of verses which
have diaeresis after the first dipody in lieu of cacsura,and it
6

38 M. W. Humphreys,

is the considerable number of these that apparently increases
the instances without elision in Aeschylus in both tables.

In Euripides the only examples of verses with diaeresis in
the middle without elision and without any break at one of
the places for the principal caesura, are Hel. 86, and Bacch.
1125. But the former verse is corrupt, having in the MSS.
an anapaest in the fourth place; and, by the way, it has a good
caesura in the fourth foot; but the critics, in removing the
anapaest, destroyed the caesura. Such an ‘“emendation” is
utterly unworthy of consideration. The other verse is:

AaPovea 3d wAEvace aparepar yxEpa.
This being the only instance, one is tempted to remove it by
writing wAévato’.

The statistics show that j in Euripides a limit was attained,
or nearly attained, towards which we see a tendency in pass-
ing from Aeschylus to Sophocles.

In collecting the examples | observed a few facts to which
| call attention. 1. Verses which have diaeresis with a
pause in the middle, especially when there is no elision, very
frequently have one or more of the following peculiarities :

(a) There is an antithesis between the two parts. This
may be expressed by wer ..... dé... as Antig. 555:

ou perv yap eidov ij, Eyw d€ ward
(Cf. Sopw. Oed. Rex 785, Elect. 1036, Philoct. 503, 1009,
1021; Kur. Rhes. 161, Hippol. 313, Phoen. 521, Ion 742,
Hel. 575, Iph. Aul. 827, etc.; and with elision, AEscH. Prom.
500; Sopu. Oed. Rex 1163, Elect. 696, Philoct. 359, 676;
Kur. Alcest. 625, Med. 1141, Hec. 497, Hel. 49;) or it may
be expressed by dre... rére...., or in some other way, as
ArscH. Theb. 1046, Pers. 251, Suppl. 401, Agam. 1353, 1396;
Sopo. Ajax 1377, Antig. 518, Oed. Rex 968, Oed. Col. 1038,
Elect. 1038, Philoct. 907; Eur. Alcest. 789, Androm. 656,
Hec. 232, 253, Suppl. 268, 379, Heracl. 424, Hel. 987, Bacch.
507, 682, 975, Iph. Taur. 674, Iph. Aul. T47, etc. In these
cases the real caesura is generally found in its proper place,
and the emphasis of antithesis causes the caesural pause.

(6) There is a long pause near the beginning. When
this pause occurs, the rest of the verse is naturally read con-

aVvVEenv—

On Certain Effects of Elision. 39

tinuously, so that the mere break at the place for the prin-
cipal caesura is sufficient, as Antig. 997:

rid taney; we éyw | roll cov dpicow ordpa.

So Philoct. 736, Phoen. 1005, etc., etc.

(c) The arsis (Soc) of ‘the third foot is a dissyllabic
word, that is, it is resolved, and there is caesura in the
- foot, as Androm. 47:

a SY ow me ; e , ,
UG C eaTt Talg pot poros, | UTEKTEUT YW AaSpa.

The relative frequency of this in Euripides, where resolutions
are frequent, leads to the suspicion that a computation might
show it to be due to accident; but it is certainly striking,
if we examine Orest. 1585, Phoen. 449, 846, Suppl. 1060,
Here. Fur. 321, 1181, lon 742, 828, 1030, Hel. 267, 290,
1027, 1028, 1241, 1899, 1449, Elect. 43, 1084, Bacch. 297,
353, 841, 975, Iph. Tau. 371, 484, Iph. Aul. 747, etc.

2. (a) In Lyric passages when an occasional ‘iambic tri-
meter occurs, as in other respects, so in regard to caesura, it
is not subject to the laws of the ordinary verse ; consequently
I have omitted them in the count. As examples see Troad.
1305 and its corresponding verse 1320.

(6) A few verses present neither main caesura nor diaere-
sis, as AESCH. Suppl. 244, Pers. 501 (both ~ - ~ —| ),
SopH. Ajax 969 (~ - ~ -—| with elision), Oed. Col. 373
(~-- ~.. |), Eur. Suppl. 803 (which has a break in the
middle, but pause after the second dipody).

As the collection of all these statistics was a mere parergon
while I was reading the Tragedians for another purpose, I do
not pretend that the figures I have given are absolutely cor-
rect. In fact, in some cases, I know that they are slightly
erroneous; but they approximate the truth sufficiently to leave
no doubt as to the correctness of the general result. I hold
that I have fully established the fact that quasi-caesura must
be recognized ; and the explanation of it which I have given, it
seems to me, is not only satisfactory, but is the only possible
one. But we are not to imagine that the vowel exposed to
elision received its full sound, and that a caesural pause was
made in addition; for this would be like an anapaest with

40 M. W. Humphreys,

caesura after its first syllable. The elided syllable was pro-
nounced enough to render the first part of the verse somewhat
similar to that of a verse having the main caesura in the
fourth foot. The voice then passed rapidly on to the next
word, unless the sense demanded a pause ; and even when this
was the case, the elided vowel, receiving a fuller pronunciation
~ than was customary (as it was usual to suppress elided vowels
entirely—elisions before long pauses being generally avoided),
in a certain measure supplied the place of a pause. If the
objection be made that this would make the verse like one
having its caesura after the first syllable of an anapaest, I
reply that this must be the case whether we recognize quasi-
caesura or not, for it is universally admitted that before a
strong punctuation an elision cannot be total This sort
of caesura, then, is rather of the sort which serves merely as
a link or bond to hold the two parts of the verse together,
than of the sort which gives the reciter a space to catch his
breath in. So that the portion of the verse after the caesura
is like that of a verse having no caesura, except that most
probably the vowel (which is always short) following the
elision suffered a partial aphaeresis. But in all cases the
elided vowel and the one after it were so pronounced as not
to interfere with the proper time of the foot.

An apparent difficulty is presented by those verses which
have elision with a long pause at the ordinary penthemimeres ;
for if the slight pronunciation of an elided vowel at diaeresis
creates caesura, why does not the same thing at caesura de-
stroy it by creating diaeresis? The reason is found in the
fact that the diaeresis in the middle of the verse is always
followed by a single mora whose place is, in a manner, partly
filled by the elided vowel, there being no ictus, whilst the
penthemimeres is followed by a double mora with ictus, whose
place cannot be even approximately supplied by the elided
vowel. For instance in Alcest. 381:

xpovocg padabe oa” ovdév tod’ 6 KarSavwr,
unless we slightly pronounce the elided vowel, we have
padaéece-—which is hardly admissible ; but the barely audible
e could not be mistaken for the arsis (Sé¢) of the foot.

On Certain Effects of Elision. 41

Still it somewhat impairs the flow of the verse, so that some
of the Romans, who in almost all cases had only partial elision,
appear to have avoided this elision. Hence, in the only ode
of Horace composed entirely of trimeters (the last Epode),
containing eighty-one verses, there is not a single instance of
this elision, and it is very rare in the trimeters in his other
odes, which odes were composed, as I shall presently show,
under laws less strict in other respects than the last Epode.
The rarity of elision at the caesura in Horace can hardly be
attributed to accident ; for in the twenty-ninth Ode of Catullus,
containing twenty-four verses, there are seven such elisions,
three of which precede a polysyllable, which is more objection-
able than before an unimportant monosyllable. In Horace, on
the other hand, there are in all the trimeters only two instances,
one of which (Epod. V, 97, where vicatim loses its ultima),
takes place to allow the ictus to fall on the first syllable
(Transactions Am. Phil. Assoc. 1876, p. 121), and the other
(VI, 11) is
cave, cave : namque in malos asperrimus,

where, even if we place the caesura after namque, we must read
continuously, and when this is done, -gue really suffers total
elision, as will appear hereafter. It must indeed be admitted
that Catullus allows more elisions in general than Horace
does, but not so many more as to account for this disparity.

That elision at the main caesura is so frequent in Latin
dactylic hexameters is no matter of surprise, because the
feminine caesura is also admissible. These cases are not to
be confounded with those where elision takes place at the end
of the second foot before a monosyllable, so that this mono-
syllable is by the elision closely connected in sound with the
preceding word, and so admits caesura after it, as Hor. Epist.
II, 1, 46:

Paullatim vello et | demo unum, demo et item unum.

(Or is this a kind of aphaeresis of the vowel of the monosyllable,
which is usually et?) In other verses of sufficient length to
make a breathing-place desirable, an elided vowel at this place
was sounded a little even by the Greeks. Since this elision,
frequently occurring in the versug politicus, as Nub. 1362:

s A ‘ wf ’ t ‘ ,
Kai Tov Lapwridnv Epack | Elvat KaKoY TonrTHy,

42 M. W. Humphreys,

caused the first half to sound somewhat as if it had feminine
caesura, this latter came actually to be admitted, as Nub. 1411:

ov Kapé got Cikatdy ear | EvYOETY Opotuc ;
and finally even with a long syllable, as Nub. 1366:

éyw yap Aisyvroy vopizw || rperor év momrate.
But this one concession being made to the influence of elision,
no further elision was tolerated at the feminine caesura.
Again: even in the trochaic tet. cat., Aeschylus seems to
have allowed elision, in one instance, to substitute apparent
caesura for the otherwise universal diaeresis: Pers. 165:

raura pot deTAH peppy’ || appacrde éorey Ev dpecir.

From all this it is evident that in iambic trimeters, quasi-

caesura is to be expected after the second foot (as well as the
third), provided some circumstance compels the reciter to
sound the elided vowel; and this actually occurs, (1) when
there is a long pause at the end of the first dipody, and (2)
when at that point the verse is divided between two speakers.
The former kind is of rare occurrence, especially in the Trage-
dians, since the latter portion of the verse was rather long,
and, besides, the syllable following the elision is not neces-
sarily a short one (as in the middle of the verse). Still I
suspect that Hel. 818:

Epet O€ ric p's ov yrwoeral y Oc Eig’ Eyu,
and a few other verses, such as Oed. Col. 1475, Trach. 449,
1186, 1208, Philoct. 1035, Ajax 969, etc., are of this kind.

Verses of the other sort are scarce, indeed, in the Trage-

dians (see Trach. 418), since they do not very often divide
verses between actors, but in Aristophanes I find these in-
stances: Lys. 911, Eccl. 1094, Ach. 832, Equ. 726, Aves
$46, Nub. 726, 729, Pax 283, 367, Plut. 3874. In my paper
on Elision (1878) I called attention to the fact that, when
elision seems to occur between two speakers, there is in fact
no elision at all, but when the first speaker is uttering the
vowel marked as elided, the second speaker utters his first
syllable, there being not only no pause between them, but an
actual overlapping. Hence, when there seems to be an elision
of this kind after the first dipody, we really have the ordinary

On Certain Effects of Elision. 43

penthemimeres, the overlapping being employed to secure

quick retort, or sudden reply of some sort, thus (Ach. 832,
Nub. 726):

Kai yaipe wodAd.

"AX

"ANN wyay’ arodwnr’ apriwc.

‘.3 SY ° , ,
AX Quty ovK Extywoury.
Xe

® ~ ,
GQTONEL KAKUOTA.

Here the caesura tuok care of itself without any effort on the
part of the first speaker, and hence this kind of caesura is as
common at the end of the second as at the end of the third
foot,—in fact more common in proportion as caesura in the
third foot is more common than in the fourth.

_ § 2. Quasr-caeEsura IN LatIn.

As the Romans were accustomed to hear vowels that were
exposed to elision slightly pronounced, their ears were less
delicate as to these vowels than were the ears of the Greeks,
who were accustomed to the total suppression of elided vowels.
Consequently, in Roman authors the quasi-caesura is not. to
be expected so frequently. But, as they always made clided
vowels audible, if quasi-caesura occurs we may expect. it
occasionally after the second foot, no effort being required to
produce it; but the fact that the latter portion of the verse is
so long, and that the elision is usually followed by a long
syllable (which cannot be ‘represented by a diminished sytla-
ble), renders it rarer than at the middle of the verse.
Accordingly there are a few examples of it in both places.
But when the (apparent) elision takes place between two
speakers, of course the same principle applies as in Greek,
and we find not a few instances. (See Casina 352, 397, 509,
etc.) In the other plays of Plautus it is not rare between two
speakers. In Terence also we:find a number of examples, as
Andria I, 1,7; J, 1, 92; II, 5,5; IV, 4,46; V, 3, 23; ete.

Sometimes the caesural pause seems to have prolonged a
vowel exposed to elision into a whole mora, so that there was
no need of this mora in the next word; whence arose hiatus
at caesura, —a question too vexed for discussion here.

44 M. W. Humphreys,

§ 3. PRINCIPLE UNDERLYING QUASI-CAESURA.

I have said that, in case of quasi-caesura in Greek, the
reader, having passed to the middle of the verse without
encountering caesura, unhesitatingly makes it at this point
by slightly uttering the elided vowel, because he knows that
it is the last chance for a main caesura. I now propose to
illustrate this by giving some further applications of the same
principle.

1. An iambic word (as Lachmann pointed out) rarely suffers
elision in Latin, especially in dactylic hexameters, the cause no
doubt being that the word would thereby be too much modified.
(The elision sometimes occurs before et and rarely before
other monosyllables ; in which case I suspect aphaeresis rather
than elision.) And when a reader has become accustomed
to finding iambic words always unelided, he acquires the
habit of boldly pronouncing them in full without reference to
the following word. This gave the poets the opportunity of
admitting hiatus after such words, as it removed one of the
ohjections to hiatus—the danger of leading to a false reading.
Consequently there is a considerable number of iambic words
with hiatus, as “‘ndvd auctus hymenaeo,” “Simd; hic fllius
arma.” For many examples see LACHMANN (ad Lucret. III)
and Corssen (Ausspr. Voc. u. Bet. I], p. 785). Unless one
is acquainted with this fact, he will hardly read correctly at
the first attempt Vere. Geo. IV, 463:

atque Getae atque Hebrus et Actias Orithyia.
Also, when an iambic word is immediately preceded by the
arsis, the final syllable may, for a like reason, be shortened
before a vowel, as CatuL. CXIV,6: dum ddmd ipse egeat ;
Ovip; Metam. III, 501: . . valé, valé inquit. Here the short
syllables dd- and vd@- force a short to follow.

2. The lengthening of a short syllable under ictus in hexa-
meters is due to the same principle. For when you have
read the thesis (apac), you know already that the long arsis

Séoc) must follow; so that you make the syllable long even
though it be a short one; whilst, on the other hand, if you
finish the arsis and find a short syllable after it, of course you
pronounce it short expecting a dactyl, and this prevents the

i

On Certain Effects of Elision. 45

composer from substituting a short for a long vowel in
thesis—a thing which he can do in arsis without any danger
of being misread. It is for this reason that common syllables
(éixpovor) also are made long much more frequently in arsis
than thesis. The statement that the ictus lengthens a short
syllable, if literally meant, is absurd. But for the difficulty
mentioned, the lengthening would be much less of a license
in thesis, for in arsis it-is not only lengthened, but also
receives special stress, which distorts the word all the more.
But in iambic verse, a short syllable cannot be lengthened
under ictus (a good proof that ictus does not lengthen
syllables); for the arsis may be resolved, and when you
strike a short syllable you at once anticipate a resolution,
and would go wrong were a short syllable put for a long one.
For instance, if in the verse, |
. seis; feci ex servo ut esses liberttis mihi,
you put ego for mihi, the reader would put a dactyl in the
fifth place, and the verse would come out defective :

scis: feci ex servo ut essem libértiis égo.

_§ 4. ELISION aT THE END oF A VERSE.

1. Of elision at the end of a verse no example is known to
me in the case of Latin iambic trimeters. In Aeschylus I
find no instance of it. In Nophocles it occurs, as far as I
know, ten times (Antig. 1031, Oed. Rex 29, 332, 785, 791,
1184, 1224, Oed. Col. 17, 1164, Elect. 1017), six examples
being in one play. It is confined to verses which are so con-
nected grammatically with the next, as to forbid a pause ;
and, as was to be expected, the syllaba anceps is not admitted.
Several of these verses end with «’, which some _ editors
strangely transfer to the beginning of the next line, as if you
would not have to read continuously in either case.

In Euripides there appear to be no examples. Orest. 1489
and Elect. 1184, in both of which it takes place before a
pause, are not to be counted, as they are mere accidental
trimeters in Lyric passages. This is the second feature in
which we have found Lyric trimeters to differ from ordinary
ones.

T

46 M. W. Humphreys,

In Aristophanes I recall two examples (Aves 1716, Eccl.
351), in both of which & stands at the end, the preceding
word being closely connected with the next line.

2. In Greek hexameters I do not know of any instances of
elision at the end. Of course I take no notice of Nauck’s
attempt to banish -ec and -nc (Dat. pl. endings without final :)
from Homer, which attempt leads sometimes to -oo’ and -y0
at the verse-end. In Iliad © 206, = 265,0 331, Aristarchus
wrote Ziv, and not Ziq’.

In Latin hexameters we find this elision occasionally under
one of two conditions: (a) When the sense and grammatical
structure allow no pause, then a vowel with -m, or a simple
short vowel (especially in -que) may be elided. It is my
opinion that, to insure a correct reading, the poets sometimes
sought this elision, as Vera. Aen. VII, 160, X, 781:

Iamque iter emensi turris ac ¢ecfa Latinorum
ardua cernebant iuvenes, muroque subibant.
Sternitur infelix alieno volnere, coelumque
adspiet(, et dulcis moriens reminiscitur Argos.

Other examples are Aen. I, 448, V, 422, VI, 602, VIII, 228,
XI, 609. Also, Geo. I, 295, which is the only one | where the
elision is not needed.

(6) When -que stands at the end and occurs again between
the main caesura and the end, then the last -que may suffer
elision, even before a pause, as VerG. Geo. II, 443, Aen. II, 745:

navigiis pinos, domibus cedrumque cupressosque ;
hinc radios trivere rotis, hinc tympana plaustris.
Quem non incusavi amens hominumque deorumque,
aut quid in eversa vidi crudelius urbe ?

Other examples are Geo. II, 344, III, 242, 377, Aen. I, 332,
IV, 508, 629, V, 753, VII, 470, 1X, 650, X, 895. So Ovmn,
Metam. IV, 11, 779, VI, 507, etc. The examples referred to
in Vergil are exhaustive; so that it cannot be regarded as
an accident that another -que is always found near the elided
one. The cause is that the too frequent repetition of -que
was unpleasantly monotonous, on which account the Romans
often subjected one of them to this elision even at the verse-
end, when it happened to be convenient. The aversion of
the Roman ear to -que shows itself in several ways. Even

On Certain Effects of Elision. 4T

when not repeated it is exposed very often to elision in the
body of the verse. This happens frequently at the main
caesura, and especially before et (cf. Hor. Epist. II, 8, 145,
162, 165, 196. 199, etc.). Again, after Lucretius, it became
rare after short e (Horace, in Sat. I, 1, 89, has servareque
amicos, with elision) ; and finally went out of use after words
ending thus. Still another striking evidence of their aversion
to the monotony above mentioned is found in the fact that,
when -que is found two or more times in the same part of the
verse, it is frequently leugthened in one of its positions. We
often see it stated in the elementary Prosodies that -que
is sometimes lengthened; but I have never seen it stated
when this occurs. Some speak, indeed, of its being lengthened
before two consonants ; and in fact it does usually occupy this
position when lengthened, but the consonants are generally a
mute and a liquid, which very rarely cause position when
initial. In fact even the strong position is exceptional in
Latin when it acts on a final vowel of a preceding word. In
Greek Comedy the weak position never lengthens a syllable,
and even in Tragedy, a final vowel to secure quantity always
takes movable v (if it can) before the weak double consonants ;
as Hel. 135, 656: GAgoev cd€oc, HAxicevy Boorsy, and in hundreds
of other places in all Greek poetry. And, what is more to
the point, Vergil and other Latin poets do not under these
circumstances lengthen final syllables of other words. Rare
exceptions occur, as CaTUL. (in iambics) IV, 9: Propontidd
trucemve, etc. One example of -qué occurs before a simple
sin Vergil (Aen. XII, 363):
Chloreaqué Sybarimque Daretaque Thersilochumque,

where Wagner attributes to s the force of a double consonant!
He cites, by way of proof, Aen. III, 464:—gravid sectoque
elephanto. But if s had this weight, why do we meet no
more instances of it? Besides, every one knows that occa-
sionally any final vowel (and especially neuter plural -a) is
lengthened before a single consonant. And of the letters,
s is least likely to have such an effect. In Ennius and
Lucilius final -s with an initial consonant in the next word
hardly ever makes position in thesis, and with its vowel is

48 | M. W. Humphreys,

elided very often. Cicero elided it in verse, and Catullus
(CXVI, 8) drops it before.another 8. Priscian, on “ distinctt
smaragdo,” says: “S enim in metro consonantis vim saepe
amittit.”” Even in the middle of a word before a conso-
nant it fails to make position not unfrequently in Plautus and
Terence. It is the weakest of all the consonants, except final
m before a vowel. Moreover, we have an example before 7 in
Vergil (Aen. III, 91): liminaqué Jaurusque, etc.; and even
before p in Ovid (Metam. VII, 225): Othrysqué Pindusque,
etc. But, as I said, it usually precedes two consonants, as
Vere. Kel. IV, 51, Geo. I, 158:
lerrasqué tractusque maris, coelumque profundum.
. lappaequé tribulique, interque nitentia culta.

The remaining examples in Vergil are Geo. I, 164, 352, 371,
II, 885, IV, 222, 386, Aen. IV, 146, VII, 186, VIII, 425, IX,
767, XII, 89, 181, 863. In Ovid there are a good many
instances, as Metam. III, 530, VIII, 526, ete.

The enclitic -ve, when repeated, also may suffer elision at
the end, as Hor. Sat. I, 6, 102,—riisv% ptrégréve | Exirem,
etc. When we consider all that has been shown about -que
and -ve, and remember further that neve becomes neu, etc.,
and that -ne drops its -e sometimes even before a consonant,
and that such forms as “omniaque ”’ suffer elision in fifth foot
of hexameters incomparably more than other words, we can
hardly doubt that when these enclitics were expused to clision,
their vowel was tvtally suppressed.

When elision takes place at the end of a dactylic hexameter,
the catalectic pause is destroyed by the continuity of the two
verses, and so we can have no syllaba anceps, but the last foot
must be an actual spondee. I am not disposed to insist on
this, but apparent exceptions (such as Vere. Geo. II, 69, III,
449, Aen. XI, 333) have been removed not without Ms.
authority.

That the last syllable must be long after the elision has
taken place is not disproved by the fact that, at the end of a
Sapphic verse it may, under similar circumstances, be either
long or short (cf. CaruL. XI, 19, 22: “nullum amans vere,
sed identem OmnYum | ilia rumpens,’’ and “qui illius culpa

On Certain Effects of Elisiun. 49

cecidit velit pratum | ultimi flos,” etc.) ; for the last foot in
this verse may be either a trochee or a spondee, as is shown
by comparing these two verses:
~ Gallicum Rhenum horribile aequor ##ti—mosque Britannos.
Labitur ripa Jove non prodante tr—orius amnis;
or in Greek :
wixva Guvedvreg mrép car’ wotvw ats i—pog due pécow.
iLavet cat wAaaioy adv pwr éi—oag braxovet.
Nor is anything proved by the well-known distichs:
. *H péy’ ASnvawiet dowe yéved’ Hix’ ’Apiori-
yeirwy “Inmapyoy xreive kat “Appdcroc.
Otrog 64 oot 6 KAeevdg av’ "EXAdOa racar ’ AmoAAo-
dwpng’ ytyywoxeg ToUvopa ToUTO KAUwY.
Oijxe 6 Opto vovowr re Kaxov Cwaypia Nixé-
Hhone, Kal yepwor detypa wadaryevéwy.
For it is too plain that the poet, in these verses, was driven
to his wit’s end to get the proper name with one short syllable
between two long ones into an elegiac distich at all. The
reason that no such division of word is found at the end of a
hexameter ending in a long‘ syllable, is that such a word
could be incorporated elsewhere into the verse.

§5. THe Porsonic PAvseE.

The substance of the well-known law promulgated by Por-
son in his Prolegomena ad Hecubam, which is usually ex-
pressed in a rather clumsy way, is this: caesura in the fifth
foot must not be preceded by the long ultima of a polysyllable ;
or, to make it applicable also to the trochaic tetrameter:
caesura cutting off three half-feet from the verse-end must not
be preceded by a long ultima. (By polysyllable I mean a
word of more than one syllable.) If the break is followed by
an enclitic, the caesura is not so decided; so that some ex-
ceptions occur in this case, but not the tenth of what would
naturally occur. Hence we infer that the enclitic was not
regarded as preventing the ¢aesura—a fact which sustains my
views before expressed with regard to main caesura before an
enclitic. The law applies to some extent, as was observed by
Elmsley (Review of Porson’s Hecuba), when the caesura is

50 M. W. Humphreys,

preceded by a long monosyllable which is more closely con-
nected with what goes before than with what immediately
follows. If it isa postpositive word, the offense is still greater,
and if it is an enclitic, the offense is next to that of a full
violation. Similarly, if a postpositive word follows the
caesura, the offense is greater if it is not enclitic, but not so
great as if it were not postpositive, and the offense of a pro-
clitic approximates that of a trisyllabic word at the end.
Hermann denies that the postpositive character of a word
after the caesura excuses a violation of the law, except where
av is separated by the caesura from a verbal form suffering
elision, as eiroup’ | av rére. My theory, however, that elision
in Greek was ordinarily a total suppression of a final vowel,
leading to a close connection of the two words (as ror avé-
p’ dpa, dx-d’ agampeiv, etc.) led me to expect to find occasional
violations of the Porsonic law excused by elision, even though
the word following should not be postpositive ; for whatever be
the reason for the rule, it has reference to a pause or break
between words. My search for examples was at first almost
fruitless ; but finally I went to.the old editions and Ms. read-
ings, and found my theory fully sustained. Since my investi-
gation I have found that Munk had already stated that the
rule does not hold “wenn ein apostrophirtes Wort zur Ver- -
kniipfung mit dem folgendem zwingt.”” Of course, under the
partial elision theory, this announcement went unheeded. To
show to what extent elision had influence, it will be necessary
to collate the principal exceptions to the rule, and briefly
discuss the commonly received views, as first set forth by
Hermann.

In the first place, though convenient, it is not accurate to
speak of the forbidden spondee, for it has not even the form of
a spondee, but of an anapaest, when the arsis of the fourth foot
is resolved. Thius, if in Bacch. 495,

éxecra Jupsov rovee wiipdcdc ix yepwr,
we substitute répddouc for rapadoc, the Porsonic law is violated.
One violation of this sort occurs in Mss., Ion 22:

gpoupw rapuZevéuca pudAdcac owparog,

where Porson very properly wrote pvAake.

On Certain Effects of Elision. 51

Again, all the apparent violations where fpir and ipir pre-
cede the pause are to be removed by writing jpiv and ipir
or er and ir. Dindorf calls attention to the possibility
of this change, and yet he does not make it in his text.
Aeschylus has one instance, Prom. 821,—jpir | at xan; but
there is another instance of the short ultima in Eumen. 347,—
éy’ div éxpaySn, where the verse demands this quantity. In
Sophocles these forms are quite common before the Porsonic
pause, as Oed. Rex 1482, Oed. Col. 25, $4, 81, 1038, 1167,
1408, Elect. 1328, 1882, Philoct. 581; but the short ultima is
met at every turn in other parts of the verse, and it looks
strange to see in Dindorf’s text (Oed. Rex 1482) the ending
ipiv | od’ dpdy, and two lines lower down (1484), the beginning,
dc tiv, w réxv’, KTE.

In Euripides there seems to be no shortening of this ultimate,
and consequently these forms are not admitted before thie
Porsonic pause. One exception appears to occur in his Frag-
ments (Dind. 711) ; but the verse is quoted by Aristophanes
who, it is well known, frequently fails to reproduce the exact
words.

Hermann, looking upon hephthemimeres as almost a fault,
thought it required another pause in the latter part of the
verse, so as to make this part more nearly equal ‘to the first
part. Hence he divides thus when this (34) caesura occurs:

xeivn yap GAEaéy vey, | etc Tpot | avr’ aye.
This theory he applies in explaining certain violations of the
rule in question. In fact it looks as if he got up the theory
for this purpose; and he would have you believe that the
neglect of the rule was intentional, in order to increase the
weight, so to speak, of the latter part of the verse. Under
certain circumstances, the presence of the hephthemimeres
does seem to excuse a violation of the rule, but the violation
could not have been sought, for then we certainly should
have had more instances. The explanation of the simultaneous
presence of hephthemimeres and disregard of the rule is
mainly due to the fact that a trisyllable preceding the pause
causes the objectionable diaeresis in the middle, and quadri-
syllables of the suitable form (=~ -- .-) are rare, and

52 M. W. Humphreys,

hence only the dissyllable ( — — ) is left, and this creates
hephthemimeres. With a monosyllable there, the law ordin-
arily does not apply. There are, however, two ways in
which the hephthemimeres might have failed to exist, the one
when there is quasi-caesura, the other when a monosyllable
follows the penthemimeres ; and instances of both these actu-
ally occur in the few examples cited by Hermann (Jon 633,
quasi-caesura, and Iph. Aul. 1212, monosyl.) :

ao’ év9ad’ elyov ayad’, || dxovedy pov, mareo.

reigey éxgdova’, || do 3’ ouapreiy poe nérpac.
And the hephthemimeres, accordingly, is very common, when
a polysyllable ending in a short syllable precedes caesura in
the fifth foot, although the Porsonic law is then intact. But
Hermann calls attention to the fact that often the caesura has
along pause. In the first place this is not so frequently the
case as he assumes; for he puts a strong pause where he
would not otherwise have placed it, as after cipx’ in Rhes. 715:

Biov 8 axaray, eipx ayuprne ric Narre.
(This particular example, however, is lyric, and proves little
of itself; but it serves to illustrate.) And so, frequently,
when the Porsonic pause is followed by an enclitic. But
when it is followed by yap, pév, ody, and other postpositive non-
enclitic words, there generally 7 a pause at the hephthe-
mimeres. But this seems to me to be chiefly due to the fact
that these particles generally have a pause preceding them by
a word or two. Hence we generally find the pause whether
the law is violated or not, as, with violation, Trach. 932:

idwy 8 6 waic wpwlev’ tyvw yap rakac—
and without violation, Antig. 771:

ov THY YE pn) Styotcar’ eb yu | ouy A€yete.

(Cf. Antig. 96, 255, 270, 407, 448, 478, 567, 989, 1023, 1048,
1108, 1165, 1255, 1302.) Besides, Aristophanes, who cer-
tainly disregarded the law, has hephth. like the Tragedians,
when he has Porsonic pause. Still it may be that the pause,
allowing the reciter to catch breath, justified the non-observ-
ance of the rule. For, although I cannot see why comic

On Certain Effects of Elision. 58

actors should be presumed to have better lungs than tragic
actors, still we know that harder tasks were imposed upon
them, and so, for want of a better, we accept Hermann’s (and
others’) explanation of the law, viz., that it was to prevent
too heavy a drag when the lungs of the speaker were nearly
exhausted. If they had time tocatch breath near the middle,
the observance of the law, then, became less necessary.

I now proceed to examine all the violations, as far as I know
them ; and it will become quite evident that elision is one of
the principal causes of disregarding the law. In collecting
examples I have been greatly aided by ELMsLEY’s Review of
Porson’s Hecuba. (To collect examples would be the simplest
thing in the world if the Ms. readings had not been tampered
with.) Ishall first briefly allude to instances which have no
apparent excuse, or at least, one that was not much applied.
The emendations to most of these seem to have sufficient
ground. In Pers. 321, proper names are concerned. In
Rhes. 731, the same thing seems to occur; but the correct
division into verses removes it. In ArscH. Suppl. 198, and
one or two other verses, the origin of the false reading is
evident. Oed. Col. 664, which a preposition seems to excuse,
was changed by Porson. Oddeic and ovdé», changed by Porson
into oid’ ic and odd’ év, occur in Oed. Col. 1022, Alcest. 671,
Phoen. 747, and Herc. Fur. 1338 (spurious?). Trach. 1136,
had quantity before prwpévn—changed by Heath to pupern.
Philoct. 533 had rpocxicarrec (referring to two)—changed to
dual; 731 was changed before law was known; Rhes. 928,
Androm. 346, Hec. 729, Heracl. 640, Herc. Fur. 983, Ion
22, Iph. Aul. 530, 1456, and a few others in Euripides have
received obvious emendations. For Ion 1, see Dindorf. Iph.
Aul. 665 is manifestly corrupt, the verse being defective. In -
Androm. 346 and Hec. 729, where there is quantity before
w (in Weddouar), Dindorf accepts the emendations, but in Iph.
Aul. 580, he retains «gra | ~evdouar, but rejects the whole pas-
sage. Munk retains all three, regarding the quantity by posi-
tion as being less effective. In the Fragments of Euripides,
364, 2, Porson changes xamocwoac to xarocwoar’; 364, 28, is
doubtful ; Witzschel’s reading of 499, 4, is unworthy of notice ;

8

54 M. W. Humphreys,

in 594, 8 (eur yap HAS pnrpi Keovily ™poc Aéxoc) Conington writes
kedvov ele Aéxoc,—received by Dindorf; 699 was no doubt
modified by Aristophanes, by whom it is quoted; the verse
(—kxaprev; | carSaveiv) in Plutarch, referred by Valckenaer (ad
Phoen. 1331) to Eur. Palamedes, need not be considered ;
T07,—ri xpav | cizare is quoted by Aristophanes, where 7
throws doubt on the passage; 773 has roro (for rovrov) in
Mss.; in 1019 ~oAAa (for roAd}v) is found in one Ms., and is
to be retained as being the more obscure reading ; in 1068,
3,—perapédacay | AauBaver, the emendations (Heimsoeth adgave.,
Meineke perapéAeca) appear to be quite arbitrary. The fact,
however, that relatively so large a number occur in quoted
fragments casts doubt upon their accuracy.

The great body of instances of disregard of the law, nearly
all of which are genuine, but have been much tampered with,
I take up in this order: 1, Where there is elision; 2, an en-
clitic ; 3, yap, pév, ovr, etc., after pause; 4, enclitics and post-
positive words before pause ; and I shall include the trochaic
tet. cat.

1. First, then, where there is elision :

AESCH. Suppl. 752: Kade ay iptv Evppépor ravr, w Téxva,

where Elmsley writes rad’; in fact he always writes raé’ and

760° when he finds rair’ and voor’, although he is compelled to

leave rovd’ (as Oed. Rex. 219) and révd’ (as Iph. Aul. 895),

etc., unchanged.

AESCH. Pers. 762: é obre ripy Zeve Gvak ravd’ dracey.
Sopu. Ajax 1101: éeor’ avacoeey, Sv 66 fryeir’ otxoSev; Pors. fyer,
others iver’.

“ — Oed. Rex 219: yw E€vac bev rou Ndyou Tove’ Ekepw.

“ — Oed. Col. 505: rotkeiSev adroove, © “evn, rovd™ jy d€ rov—:

changed by Elmsley.

“© Elect. 413: ct poe Aéyoee roy Oey, Etxoup’ My Tore.

Trach. 718: rac obk cAei cai révde; Edn y' obv éun—(your 7).

‘¢ Philoct. 22: & pot rpoceA$av siya ohpay’ elt’ Exe. All

efforts to change this, as far as I know, have failed.
Eur. Alcest. 1080: éyvwxa xabriécgt add’ Epwe ric p ebayer: Elms.
ric ELaYyEet.
Hippol. 294: YUPULKEC aide ouyxadioraur ay vooov; MSS.
ovyxadvioravrag, etc.

Eur.

66

66

Gé

ée

66

66

66

66

6s

66

oé

6

6¢

66

6é

w~
w~

On Certain Effects of Elision. 55

Androm. 875: xpodovg édeet Swudrwy ravd’ éxreceiv.
“ 935: Bréxove’ ay abyac rap’ éxaproir’ ay Aéyn.
“1184 : obrog pév odv ex révd' éripdr’ av, yépor.

Troad. 464: obk ayrAipeoS’, } peSnoeoS’, & caxai; where
Musgrave and others have peSieer’.

Orest. 91: ofrwe Eyer rad, dor’ ameipnx év xaxoic: Pors.

awelpnxer.
“615: paddor & éxeivn cot Saveiv éor’ atia, and

Bacch. 246: rai?’ ovyxi decvigg ayyévnc Eor’ aia, in which
verses Elms. proposed ézaiia, éxaia, and Dindorf
accepts.

Phoen. 522: Zevyrvede 8 irrove, redia riptAacd dpparwr:

Pors. xipriad’.
Phoen. 1619: add’ ert vedlwy abrog eiporm’ Gv Bior;
1626: éyw dé vate o” odx tdoap’ av ySéva. |
Heracl. 456: padtora 8 EtipuoSeucg pe Bovdoer’ av AaBav—
6-29: cat orepparovre Kai karapyeoy ei Soxet, Which
Porson does not mention, and [Dindorf is unable to
change.
Ion 1016: cic ty 8€ xpaySév zavrov ixap’ eiageperc, well
changed into eic &v dé kpadévr’ abrov, i} xwpic, popeic.
“1426: grey re xpoc rgd, i] pory red’ ebruyeic ;

Hel. 1628: oiwep h dixn xedever p’ adr’ adioracY Exrodév:
Pors. apicrao’.

Bacch. 1272: «Adore ay obv rt Kamoxpival ay cogwe.

Iph. Aul. 380: we adeAgér dvr” avip yap aicxpdc aideioy ob

gedct, where Markland puts xpnordc aideioSat
gAci. Some change was needed.

rT 523: dy py ov dpacec, rwe trodaBouw’ ay Adyor ;
‘< 635 : éymw 6€ BovAopat ra oa orépy’, & rarep: Dind.
rejects.

“858: SotAoc. ovx i Spvvopat Two” i rvyn yap HL’ ovK
ég: Elms. yap oix.
6 895: MevéArewe adeired’ Hude, O¢ kaxwv reve’ arwoc:
Pors. revo’ o¢ alriog Kaxwr.
Frag. 1045, 5: od’ ay yévoro ypappa raovr’ ev ypagn,
changed by Nauck.

I have omitted such cases as Trach. 592,—wsc oid’ | ei doxeic ;
for it does not matter how closely a word bears upon a mono-

56 M. W. Humphreys,

syllable before the pause, the rule does not apply, if that
monosyllable bears upon what follows. Hence, there are
many such endings, as Sop. Elect. 574—oid ei¢ | “Idcor,
D96—we ry | pnrépa, 1411—ddN’ oi | &x ofServ. Cf. Oed. Rex
388, 515, Oed. Col. 1443, 1646, Philoct. 385, Rhes. 418, 765,
Alcest. 320, Hippol. 79, Androm. 378, Hec. 592, Heracl. 181,
2050, 270, etc., etc. (All these, by the way, sustain my theory
that a break between a proclitic and the next word is not to
be disregarded in discussing caesura; for here it is clear that
the two are not rhythmically one word. )

In the examples collected above, it will be seen that there
are many instances of a» with elision. Let it now be observed
that in all the Tragedians, to the best of my knowledge, there
ig no instance of ay without elision in such positions; and it
was quite easy for the poets to construct such endings, and
they would certainly have done it by accident, had they not
avoided it. The frequency of elision where there is no 4a»
cannot be the result of accident either, and it shows clearly
that we are not, without some other sufficient ground, to
attempt to make emendations. It would be a strange thing
if so large a per cent. of corrupt exceptions to the rule hap-
pened to have elision.

2. I shall merely refer to other violations of the Porsonic
law, without quoting them in full. The instances where the
pause is limited by an enclitic, as in

ri wapSevever Capov etor | coe yapou,

are AESCH. Prom, 648, Agam. 1052, Choeph. 903; Sopa.
Ajax 995 (?), Oed. Col. 982, Elect. 482, Philoct. 593, 788,
801; Eur. Rhes. 715 (lyric), 868, Alcest. 1085, Hec. 507,
Orest. 111, Heracl. 516, Ion 633, Hel. 471, Elect. 1119, Iph.
Taur. 942, Iph. Aul. 1207, 1212, Frag. 126. (Frag. 794, 4,
is corrupt.) In Agam. 1052, Elmsley reads mei3€ | »e» Aoyw for
reidw xré; Alcest. 1085, Valckenaer }@acxe: caxov for i, '¢ | ove
caxov; Iph. Taur. 942, évSév | yo ééa variously emended ; Iph.
Aul. 1207 changed by Porson :—only four attempts at emen-
dation in twenty-three examples. If the commentators had
spared those which have elision as much as they did these,
the condition of the texts would be much better.

On Certain Effects of Elision. 57

3. The following verses have pé», yap, etc. immediately after
the Porsonic pause, with objectionable preceding spondee, as in
ov 0 ipiv h pucovea juoeic | pev AOyy.
idwy 0 0 Tate @ puter éyvw | yap radac—,

namely: AgscH. Prom. 107; Sopa. Oed. Rex 142, Oed. Col.

265, Elect. 357, Trach. 808, 932, Philoct. 422, 466, 546 ;

Eur. Heracl. 803, Ion 954, Hel. 1552, Iph. Taur. 678, Iph.

Aul. 391, 1146 (which has both avaxadiyw | yap Adyouc and

dvaxadinpouer Adyovc). Tothe other fourteen examples I have

encountered no emendations.
4. When an enclitic or postpositive monosyllable precedes
the Porsonic pause, as in
w pijrep, nvdac, } roAvY out | Boorpyywy—,
Cnrei wapedSetv’ roy caxwy yap | pnrépwr—,

the rule has, among others, the following exceptions: Sopu.

Elect. 376, Oed. Rex 435, Oed. Col. 115; Eur. Androm. 230,

Troad. 1182, Phoen. 403, Elect. 275, Frag. 162, 2, Frag. 716

(Dind. coi, Witzschel aor). Phoen. 403, ij» reg | dvorvyi, varies

in Mss. This list, being collected by a rapid perusal, cannot

be exhaustive. The examples show two things: first, that
in such cases the law did apply; out, secondly, that the offense
was not so great as in polysyllables.

From all this we deduce the following conclusions:

1. All departures from the Porsonic Law, as I at first ex-
plained it, are to be regarded as exceptional.

2. These exceptions may take place under the following
conditions :

(a) When the break is followed by an enclitic, in which case
there is frequently a weak hephthemimeral pause or
quasi-caesura.

(6) When the break is followed by a postpositive particle
(pév, yap, etc.), in which case there is generally, from
the nature of the case, a strong hephthemimeral pause.

(c) When there is elision at the break, whether it be followed
by a postpositive word (av), or not. This is the only
excuse for a real polysyllable with long ultima fol-
lowed by a real or virtual amphimacer. There is, in
this case, no restriction as to caesura, because the elision

. 58 M. W. Humphreys,

renders a word foot of any possible form admissible
before the break. The pause is always short enough
to allow ¢otal elision. ;

(ad) Note: When, instead of a polysyllable, a postpositive or
enclitic monosyllable precedes the break, the law is not
so rigorous.

§6. Tat Porsonic Pause In Latin.

The Roman Dramatists did not observe the Porsonic law,
the structure of their verse with respect to quantity being
looser even than that of the Greek Comedy, where the rule
does not hold. The statement which I have seen that Catul-
lus carefully observed the rule, is somewhat ridiculous, as his
senarii are all pure, rendering a violation of the law impossible.
Horace, in Epode XVII, observed the law; but in v. 10 isa
violation, an anapaestic word coming before the break. It is
true, there is elision in this case, but the effect. of this elision
in Latin is not the same as in Greek. In Epode XVI the
senarii being pure, and other Epodes containing no violation
being very short, we conclude that Horace, with the above
exception, neglected the law. Of course, when the law is
violated, you will generally find the hephthemimeres, for the
reason already stated, the exceptions being when there is a
word of the form -—- ~ - -— before the break, or a monosyl-
lable of proclitic nature (so as to prevent diaeresis in the
middle) followed by the word foot ~ .- --., as Epode V, 17,
XI, 27: )

iubet sepulcris | caprificos erutas.

Sed alius ardor | aut puellae candidae.
This makes it, if possible, still more evident that this (33)
caesura came unsought, for Horace did not feel himself under
the necessity of mitigating what he did not regard as an
offense; for when I speak of violations of the law I mean
merely what would be violations if any attempt were made to
observe it. The long pause at the (33) caesura is not so fre-
quent as in the Greek Tragedies ; but the cause of the pause in
Greek (yap, per, etc.) did not exist in Latin. Still, as I said
before, there was probably some mitigation in the pause,

On Certain Effects of Elision. 59

although it thus originated; that is, the pause renders possible
a violation of the law. |

In Latin, elision would be no mitigation, as the elided word
was not entirely suppressed.

In Seneca, the phenomena attending caesura in the fifth
foot stand in no relation whatever to those in Greck. He
seems to have desired a cumbersome ending. Elision, indeed,
is common, but, as I said, is no mitigation, for that is exactly
what he did not wish, as the facts will show. In Herc. Fur. a
word ending in a trochee never precedes the break, whether
there is elision or not, nor does a short monosyllable nor a
dactyl without elision ever precede the break; that is, we
never find the forms ~|--~-,—- ~|-~ ~—, -~
(| vy HY | = 5 but only. | ~~,

= -[- vo, + (| ov = -MI-~ >,

and, in the play mentioned, in such manner that we find

Total caesurae without elision, . , . 20
c “c with ‘ Ly . 129
Violations of Pors. law without elision. . 11
“ “é “with “ . 59
Conformity to Pors. law as nearly as possible, 0c!)

From this it is evident that he was fond of elision at the break,
or of anything else that. would make the ending drag.

$7. ReLations oF Existion TO ACCENT.

In composing verse, two things must be observed: first.
each verse must preserve its proper feet, caebura, rhythmical
accents (ictus),—in short, everything that distinguishes it
from other verses; that is, regard must be had to the form of
the verse. Secondly, the words which form a verse, must not,
in order to become adapted to the verse, be distorted too much
in their pronunciation; that is, regard must be had ta the
words and the sense. If an awkwardly composed verse be so
read that its metrical form shall be preserved, the sense is
lost and the words sound ridiculous. If, on the other hand,
such a verse is read according to the form of the words and
the sense, no one will suspect that it is a verse at all. Now
this paper has only dealt with the form of the verze. But the

60 A. S. Cook.

form of the words, and the sense, are also concerned with
elision. In Latin, for instance, it enables the ictus or stress
of voice to fall on the root-syllable, as in

ita mé vetuistas 4mplexu, 4nnorum énecat,
since &4mplexé, (especially in the second and fourth places)
would distort the form of the word too much. So in dactylic
hexameters, the ending =~ — | ~ ~ — ~ is to be avoided,
while it is more admissible (though not much, for another
reason) with elision: == — (~)|~~ + dv.

These are mere illustrations. The whole subject of accent
and ictus in trimeters is discussed in Zransactions Am. Phil.
Assoc. for 1876; and for hexameters, the subject is discussed
in Transactions for 1878.

IV .—Studies in the Heliand.

By ALBERT 8. COOK,

ASSOCIATE-PROFESSOR FOR ENGLISH IN THE JOHNS HOPKINS UNIVERSITY.

Since the Heliand was first made accessible to scholars in
general by the publication of Schmeller’s edition in 1830, a
number of editions have appeared, and several critical and
exegetical essays of high worth have contributed to its
elucidation, especially in what relates to its age, origin, and
place in literature. A general survey of these productions
has been given in the last and most complete edition of the
Heliand, that of Eduard Sievers, which, anxiously awaited
by his fellow craftsmen, at length issued from the press early
in 1878.

Sievers, by printing the carefully collated text of both
MSS. upon opposite pages, and accompanying it with the
prose passages on which the poetical version is founded, has
deserved well of all Germanists; but he has gone much far-
ther: for, however the conception of the alliterative formula
or of the poetical formula in general may be modified by
future investigators, it is undeniable that he has, with much

Studies in the Heliand. — 61

labor and tact, made the first collection of the standing
epithets and phrases employed by the old Saxon singer, and,
in so far as they furnish parallels, by the old Norse and
Anglo-Saxon poets.

His earlier studies on this subject bore fruit in a monograph
entitled Der Heliand und die Angelsdchaische Genesis (Halle,
1875). In this he seeks to dismember the Genesis formerly
ascribed to Caedmon, and to demonstrate that vv. 234-8382
rest upon an old Saxon original,—upon a lost poem by the
author of the Heliand. By a consent which is nearly unani-
mous among scholars he has made good his theory, and it
was the acumen thus displayed that marked his eminent fit-
ness to be the future editor of the Heliand.

Lastly, his volume ‘contains’ a body of annotations at once
learned and suggestive.

The Heliand has been unaccountably overlooked in England
and this country ; though it appeals alike to lovers of poetry,
antiquity, and religion, yet no English translation of it has
ever been made for the reading public; even its relation to
Paradise Lost, through the poem of Caedmon, has been but
incidentally remarked, if at all.

Since, however, increased attention has of late years been
bestowed upon the Teutonic languages and literature, and the
researches of Grein in Germany, March in our own country, ©
and Sweet in England, have revived, or in some sense
created an interest in the beginnings of English speech, it
may not be unadvised to prophesy that the Heliand, as the
most important literary monument bequeathed to us from the
original seat of the Saxon race, will be as deeply and fruitfully
studied among the English-speaking peoples as by the Ger-
mans. . /

Vilmar, Deutsche Alterthiimer als Kinkleidung der Evan-
gelischen Geschichte (Marburg, 1862), and also Windisch,
Der Heliand und seine Quellen (Leipzig, 1868), have recog-
nized the epic structure of the Heliand. The former has
pointed out some of its more obvious relations to Beowulf and
the older heroic poetry, while Windisch, on the contrary, has
dwelt most on the poet’s art in selection and arrangement,

9

62 A. S&. Cook,

his strivings after an organic unity suggested neither by
Tatian’s Gospel Harmony nor the triad of commentators, and
his fusion of heterogeneous elements to a compact and well-
ordered whole.

No less, however, than by his tact in choosing out of this
somewhat chaotic evangelical history the most striking and
mutually consentaneous passages, does our unknown poet
betray the hand of a master in his original additions and
. the treatment of individual scenes. By original additions we
must not be understood as meaning those amplifications of a
thought that consist in the piling up of synonymous expres-
sions around a central core—a well-known feature of Saxon
poetry, both continental and insular—nor do we refer to the
national coloring dyed, as it wére, in the grain, to that trans-
forming light which, emanating from the ancient ethnic words
employed, suffuses the whole composition with tinges and.
tones caught from the dawn of history. Rather do we desig-
nate those lines or longer sections for which no manuscript
authority stands responsible, but which have flowed spontan-
eously from the mind of the author as informed by the tradi-
tion of his people and a vital faith in the conquering and
already pervasive Christianity. So far as known, no attempt
has yet been made to separate and systematize these passages,
so significant from the culture-historical point of view.

Not less deserving of attention is his manner of enlivening
the Gospel narrative by dramatizing every scene which admits
of interlocutors, substituting dialogue for narration, and _pic-
tures for history. Here he has full command of his resources,
giving rein to his imagination, but never permitting it to lead
him into extravagance or a disregard of the limitations im-
posed by the sacred and veracious character of his theme.
Not alone in the introduction of strictly dramatic form, but
generally in the disposition of accessories, in a motivation of
some hitherto isolated occurrence, in the omission of a cir-
cumstance deemed inconsistent with the heroic tone of the
composition or with the conception of certain personages, or,
finally, in the insertion of some natural reflection excluded
by the severe and self-restrained Evangelists, do we perceive —
the touches of a master hand.

Studies in the Heliand. — 638

Reversing the order of presentation of these two topics, this
paper will treat: (A) of the chief modifications undertaken
with the design of adding vividness or life-likeness, but in no
true sense extraneous to the sacred text and the comments
possessed by our author; and (B) of the accretions which
bear the stamp of newness, and are evidently of his own in-
vention, though they vary in intrinsic worth, and in their rela-
tion to the fabric of the poem; while a third section (C) will
be devoted to a few syntactical observations upon characteristic
idioms and constructions.

The quotations refer to Sievers’ edition, and generally to
the Munich manuscript.
A.
Verses 106-8. So he tho thana uuiroe drog
ald after them alaha, endi umbi thana atari geng
mid tx rocfatun rikiun thionon.
The ceremonies of the Roman Catholic Church, rather than
of the Jewish Temple, seem to have been present to the
author’s consciousness.

144-58. A monody on the ravages of Old Age. The plaint
is true to Jewish habits of thought, yet derives much of its
elegiac sentiment from German pensiveness. Worthy of note
the specification of age at marriage—twenty winters—and of
the time which had since elapsed—seventy winters. With
gibenkeon endi gibeddeon, cf. our English law-phrase bed and
board.

159-63. The angel’s grief and surprise at Zacharias’
unbelief.

185. Butan that he mid is suidron hand ; ‘ With his right
hand ’ is added for picturesque effect.

208 ff. The impersonality of numbers merges into the per-
sonality of one, and he an old, wise man. So 22] ff. A
kinsman, but now, to mark difference, an overweening and
presumptuous man, insists upon disregarding the mother’s
wish, and (225 ff.) is answered by the first speaker, who—
and not Zacharias—proposes the use of a writing-table.

231-6. He, the old man, goes nearer, lays the book in

64 A. S. Cook,

Zacharias’ lap, and entreats him to write wisely with word-
marks what the name of the holy child shall be. Here is
doubtless a reminiscence of the Runic scratchings and gravings.

287-8. Mary is noways double-minded. Mis mi hugi tuifli.
Rather is she clear in her perceptions, trustful and unshaken.
Cf. Vilmar, pp. 82-3.

293. ‘Said to whom she would.’ Makes no secret of her
condition, being strong in the consciousness of her purity.
Characteristic of German womanhood and confirmatory of
Tacitus, Germania, 8 and 19.

380-1. Poetry and popular speech touch and mingle.
‘With her two hands.’ Cf. English, J saw him with my tro
eyes, and Hel. 980, 1177, 1194, 2042.

388. KHhuscalcos.: Grooms or horse-herds, instead of
shepherds. Cf. 2400, Hrosso hofslaga. There is no allusion
to horses in the Gospels; this trait is peculiarly German.
Hengist and Horsa are the names of the traditionary Saxon
chiefs who first settled in Britain.

481. ‘Now I am so well stricken in years.’ May perhaps
be inferred from the Bible narrative, but is nowhere explicitly
stated. Of. 493. )

548. The whole interview between the Magi and Herod is
dramatically conceived, but properly falls under B.

587-92. Worthy of remark is the precision: ‘Self-same
day,—out of the East.’

601. ‘Each morning’ protracts the time, and consequently
the distance, to the imagination.

641. Uuestar. But Bethlehem is nearly south of Jerusa-
lem. Of. also 597, both times because the wise men have
been described as coming from the Kast.

656. ‘White stars.” Similarly 590, 668, 2605, 4313.

732 ff. The whole account of the Massacre of the Innocents
is highly wrought. Similarities have been detected between
this narrative and that in Otfrid’s Krist, as elsewhere between
the two works. A reference list of correspondences may be
found convenient, and is here given.

Those for which a common basis has been made out are as
follows :

Studies in the Heliand. 65

Heliand. Otfrid. Common Author.

48 ff. II.9.11 ff. Alcuin in Joh.
420 1.12.24 Vulgata.
464 1.15.11 Beda in Luc.
545 1.17.15 Hrab. in Matt.
1024 II.4.1 ff. Hrab. in Matt.
1046 ff. 11.5.5 ff. Hrab. in Matt.
1805 II.16.7 Hrab. in Matt.
2028 ff. II.8.28 Alcuin in Job.
3053 ff. ITT.12.28 Hrab. in Matt.
4956 ff. IV. 12.34 Alcuin in Job.

The following still await explanation :

Heliand. Otfrid. . Heliand. Otfrid.
734 ff. 1.20 passim | «4446 ff. V.20.113
803 ff. 7.22.28 ff. 4572 ff. IV.12.5

1597-9 IT.21.28 4833 IV.16.52

1604 ff. 11.21.31 ff. 5478-9 IV.24.27

2925 TIT.8.24 5535 ff. IV.27.8

3843 III.17.18 5566 _IV.80.8

4027-8 TII.24.11 5571 IV.30.238

4040-1 III.24.21 5607 ff. IV.32.1-2

4065 ff. II. 24.47 ff. 5638 IV.33.18

4880 ff. V.20.5 5642 ff. IV.33.20

4385 V.20.19 ff. 5723 ff. IV.35.7 ff.

4396 V.20.71 5762 IV.36.19

1193 | V.20.78

758-60. Accurate notions of geography evinced.

964-7. Cf. Sievers’ note to 251, where a long list of simi-
lar constructions is given. |

968. John is blithe of heart when he sees the approach of
Christ. Cf. 1163.

983. ‘Fair from the flood ;’ a happy alliteration.

985. Himiles doru. Elene 1230; Salomon 37; Ps. 77:25.
Probably biblical. :

1121. The wilderness of temptation is represented as an
illimitable wood.

1178-9. It is natural to suppose that the nets should have
been broken the night before, but the statement must be looked
on as embellishment.

1197-8. Matthew leaves the ‘gold and silver and many
gifts, precious treasures,’ and chooses Christ as his liege. In
these and ‘similar words is contained the germ of many a
later tale of loyalty and devotion, of Charlemagne and his

66 A. 8. Cook,

paladins, of Arthur and his knights. Christianity is begin-
ning to leaven the Middle Age. By it the national virtues of
the simple-hearted but warlike Germans are confirmed, their
rudeness and ferocity mollified.

1610-2. The poet cannot conceive of God as the tempter,
and therefore prays for His aid against the machinations of
the evil spirits. —

1840-8. The power against unclean spirits, to cast them
out, is not delegated to the apostles, but is reserved to him-
self by their Lord,—the policy of an earthly prince strenuous
in maintaining his authority. |

1854-5. The apostles are forbidden to take with them
gold and silver, for they will gain nothing thereby.

2026-7. To be exhorted before the whole assemblage is
conceived of as prejudicial to Christ’s dignity.

2258-9. The wind and the sea are in a manner personified.
‘They fulfilled his command, Wielder’s word.’

2279-92. No allusion is made to the swine into which the
devils entered.

2572. ‘Bitter fire,’ i. e., biting, devouring fire.

2707-9. Philip is considered as having died before the
marriage of his wife with Herod.

2720-4. It is Herodias who casts John into prison.

2750-2. It is at the request of Herod that the daughter of
Herodias dances for the diversion of the company, and he
makes the promise before she begins. Upon this her mind is
inclined toward him, and the dancecommences. Vv. 2763-4
prove that this sort of dancing was strange to the Germans,
for it is these words that are always attached to every descrip-
tion of foreign and unfamiliar customs.

2813-4. ‘Was their curiosity great concerning wise words’
explains why they had congregated in such numbers.

2852. ‘The folk bode still.” Not found in the sources.

2856-7. Christ orders the meat to be borne away by his
disciples and dealt out. Matthew simply says that they distri-
buted to the multitude.

2906-9. ‘Then let they on the strengthful stream
High-horned ship the clear waves
Part the sheer water. Sank light of day,
Sun neared its setting.’

Studies in the Heliand. 67

For the same sense of sheer in English, cf. Rich. II. v. 3, 61.

3100-1. The reproof is softened in tone, to be more in
keeping with Peter’s dignity as primate.

3135-6. The transfiguration extends likewise to the moun-
tain where Christ and the disciples are standing.

3157-9. An antidote against fear is provided in the assur-
ance that nothing of what they had seen should harm them.

3200. Christ’s discussion with Peter as to the propriety of
his paying tribute was omitted, lest he should seem to be com-
promised by doing at another’s behest wliat he was under no
moral obligation to perform.

3261-2. Thoh he mildean hugi bari an is breostun. May
be gathered from the context.

3324. Obar that. Cf. English, Over and above that.

3306. The letha uuihti who sink the rich man’s soul into
the swart hell are introduced as the counterpart to Godes
engilos (8850), who bore Lazarus to Abraham’s bosom.

3418. The command is given but once, while in Matt.
20:3 it 1s repeated. ‘

3564-6. An argument is drawn from Christ’s universal
beneficence.

3076-8. This expression is singularly beautiful. They
ask to behold man’s busy doings, the light of the sun, and the
splendor of the earth.

3671 ff. The particulars concerning his mode of entry and
the animal that bore him are omitted.

3676. Mid berhtun blomun. A figment of the author’s.

8691. Uue uuard thi. Cf..English, Woe worth the day.

3709-10. <A loud voice, the most powerful of songs, is
raised when they reach the Holy City.

3822-5. The coin is brought sensibly before our eyes by
the realistic handling.

3828. The powers of Caesar are dwelt upon, and the
extent of his dominion.

8865. The writing with the finger is passed over as of
small moment, or difficult of explanation.

8980. John 11:5 omitted. Indeed, the whole story is
much condensed, to its manifest improvement for the poet’s

purpose.

68 A. S. Cook,

4234-7. Olivet is pictured as one of the mountains of Ger-
many. In truth, it is neither broad nor high, neither green
nor beautiful.

4284-5. The transition from the fate of the temple to that
of the world is made with much skill, and affords the necessary
bridge to-the disciples’ next question.

4305. Here, and again 4809, the coming is spoken of, not
as Christ’s, but as the Father’s.

4339. The’fig-tree, being unknown to the Saxons, i is not
distinguished by name.

4486-9. The rulers promise and actually give to Judas
whatever he demands as the price of his treachery. By this
we are made aware of their eagerness to destroy the Saviour.

4499 ff. Account of Jesus’ washing his disciples’ feet much
abridged. Only the essential features are retained.

4501. Skred uuester dag, sunne te sedle. Poetical addition.
With wuester compare the English westering, ‘toward the west.’

4577-80. Not only the treachery of Judas, but his delib-
erate betrayal of the Master for money is insisted on.

4750-2. Tears fall from him, his sweat drops to the
ground, as gore comes boiling from wounds.

4983-5. Thar uut an themu bomgardon
herron thinumu hendt bundun,
Jastnodun 13 folmos.
Not in the original. |
5086-8. The high priest adjures Christ to declare whether
he is the Son of God, and adds these words:
‘Who this light created,
Christ eternal King. We can perceive naught thereof,
Neither in thy words nor in thy works.’

5416. To designate the character of Barabbas it is said
that he perpetrated crime often by dusky night.

5449-52. Pilate’s wife is terrified by the vision.

5035-8. It is true that Hrabanus Maurus is authority for
the details of the crucifixion, but none the less is the poet’s
vigorous language worthy of admiration:

‘They drove cold iron
New nails hatefully (7?) sharp
Hard with hammers __ through his bands and through his feet,
Bitter bands.’

Studies in the Heliand. ° 69

5578. They revile the Saviour, and infer from his apparent
powerlessness that the subjects of such a ruler would be
miserable indeed: ‘ Woe worth the wortd, quoth he, if thou
shouldst have control of it.’

5607-10. In John 19:25, the names of all the women are
given, and that of Mary is but codrdinated with the rest.
Here special prominence is given to the latter, the others
being mentioned only in general terms, and not till after the
description of Mary’s grief and her view of Christ’s sufferings:

Blec under them bome: —giaah tro barn tholon,

uuinnan uuunderquala,
The verses quoted are not without a touch of that pathos
which finds its embodiment in the Latin Stabat Mater.

5798-9. The earthquake is represented as the effect of the
angel’s appearance.

B.

The portions independent alike of Scripture text and com-
mentary will be found in this section. It is convenient to
arrange them under six heads. The grounds upon which
certain doubtful passages have been assigned to particular
classes will recommend themselves, it is hoped, to all who
have scrutinized the language and pondered the thought of
the Heliand, though different critics will naturally vary in
their estimates of the import and character of any extract
whose place is not at once decided by the most cogent
internal evidence.

I. In the first rank are to be placed such verses as contain
the artistic motive or explanation of a subsequent part, and
which therefore belong to the organic structure of the poem.
In this regard a suggestive and regulative influence is to be
attributed to the commentators, since the author must have
reflected on their mode of dealing with disjointed, but weighty
statements of the Evangelists, especially when the truth con-
veyed is unusually important or startling.

But making all allowance for hints thus derived, we can
scarcely help seeing in the poet of the Heliand an artist with
exquisite perceptions of sequence and relation, one who aims

10

T0 A. S. Cook,

first at perfection of design before attempting the lighter task
of adding grace and roundness to the representation in detail.

Thus he accounts (43-5) for the universal sway of the
Romans by assuming a decree of Divine Providence, (239-42)
for Zacharias’ blindness, (478-80) for Simeon’s reverential
_ eagerness, (865-72) for the career of John the Baptist by
supposing a communication direct from heaven, prefaces
(1146-50) the calling of the disciples by a general statement,
accounts (1163-5) for the readiness of Andrew and Peter to
forsake their previous vocation, (3118-22) for the Transfigur-
ation, (3954-6) for Christ’s departure across the Jordan,
gives (4807-8 ff.) distinctness to our view of Judas with the
approaching band, by causing the apostles to wake from sleep
and look upon the troop, explains (4964-6) Peter’s faint-
heartedness, comments (5111-3) upon the malicious cruelty
of the Jews, and (5003-5) upon Christ’s willingness to endure
it, and assigns the reason (5794-6) of the women’s presence
at the sepulchre.

Allusions to the Judgment occur (2609-20, 5096-7) apart
from the general description of Doomsday. Cf. also the
formulae under der Jiingste Tag in Sievers’ catalogue.

II. Intimately connected with the foregoing are such veri-
fications of prophecy and seguels of incipient action as are
omitted by the Evangelists, but, being probable in themselves,
satisfy the natural demand for poetic justice and completeness.
Here also we find express assertions of that which is contained
only inferentially in Scripture.

Zacharias (170-4) is stricken with dumbness, according to
the word of the angel, the birth of Christ (371-5) fulfills
prophecy, (1984-93) general conclusion of Christ’s connected
discourses, (2066-74) effect of the miracle at Cana, (3029-
33) joy of the Syrophenician woman, (3275-7) synopsis of
Christ’s reply to the rich young man, and (5460-4) mention
of the messengers sent to Pilate by his wife.

With 5424-6, which contains an allusion to the retribution
that overtook Pilate, compare an article by Wilhelm Creize-
nach, in Paul and Braune’s Beitrdge I, entitled Legenden
und Sagen von Pilatus, p. 94 ff.

Studies in the Heliand. — 71

III. Not adduced in the way of motive or effect, but sim-
ply to mark a transition, are the following: 1486-7, 1613-5,
2462-4, 2491-3, 2513-6, 5245-6.

IV. More noteworthy than any, except those under I., are
the portions which relate to manners and customs, or to modes
of thinking and theological opinions among the German people.

(a) The high estimate placed upon womanly purity is indi-
cated by the emphasis put upon the penalties attached to its
loss (805-12, 3843-5).

(6) Light is thrown upon the universally Germanic, but
specially upon the English unwritten constitution, the binding
force of precedent, and the development of common law in
England by the fact that landuwise and gibode, eo and aldsidu, .
custom and law, are coupled as synonymous terms. Cf. 454,
796, 2768, 4549-53, 5258, 5404, 5739.

(ec) The Germanic notions of fate and predestination are
illustrated (4617-20) by Christ’s words to Judas:

Frumi so thu thenkis, quad he,
do that thu duan scalt: thu nt maht bidernien leng
uuilleon thinan. Thiu uurd ts at handun,
thea tidé sind nu ginahid.
Likewise by 2187-90, 4778-80, 4784-5, 4823-8, 4978-80.

(d) The wedding feast at Cana (2001-12) and Herod’s
birthday banquet are portrayed with great minuteness and
zest. One side of the Saxon nature is here displayed. Of.
Beowulf 612-652, 1981-4, 2015-25.

(e) Generosity and condescension are praised in the chief
(628-9, 1199-1202), while loyalty, gratitude, fidelity, and
tenacity of purpose are regarded as indispensable virtues in
his retainers. Cf. 675-7, 1169-72, 1187-9, 2154-8, 3215-28,
4002-4, 4521-5, 4556-9, 4773-5. In 5000 ff. the repentance
and sorrow of Peter in view of his unfaithfulness is depicted
with moving pathos.

Riches in abundance are the material reward of obedience
and courage, 1845-7, 1649-52. The relation between lord
and vassals is transferred to Christ and his disciples, so that
when he is seated and speaking with authority, they surround
him, and are in one sense his supporters, his eazlgesteallan,

72 A. 8. Cook,

the executants of his will. So 1272-8, 1281-90, 1381-38,
1580-6, 2167-75.

The wounding of Malchus by Peter (4877-82) is amplified
_in the relation, in accordance with the tastes of the age and
nation.

(Cf) The foreknowledge and omnipotence of God are em-
phasized, 644-8, 3239-41.

The primacy of Peter and the literal interpretation of Matt.
16:18, 19 (8066-82) are accepted as beyond peradventure.
Again, Peter is represented as answering in the name of all
the apostles (3054-6) and is exhorted to mildness, since to
him is consigned the charge of Christ’s flock upon the earth
(8253-6). His denial of Jesus is excused because of God’s
foreordination (4978-80) and explained to be the means of
teaching him man’s weakness and the duty of forbearance
(5028 ff.). |

As to the cultus of Mary, there is much less ground for
forming au opinion, though tradition and the awakening
sentiment of chivalry would appear to have grafted more
than one strange slip upon the simple Bible stem. That she
is a lovable and virtuous maid (252) need surprise no one;
the two Marys at the sepulchre are also lovable. She is called
Christ’s mother (2018, 5607), and our Lord’s mother (264) ;
but in the latter case the qualifying phrase, mid mannun, is
added. ‘Fairest of women’ (270, 379, 2017, 2032) and ‘ Fair-
est woman’ (438) are standard epithets in Anglo-Saxon
(Gen. 626, 700, 821, Men. 148, 168, El. 1170), but are applied
to no one else in the Heliand.

The more weight may be attached to this fact, since our
author is always sparing of his superlatives, never weakening
their force by indiscriminate application.

Mary, even after the birth of Jesus, is called thiorna (436,
665, 802, 1998, etc.), a word which elsewhere in the Heliand
must be translated virgin.

Was the perpetual virginity of Mary accepted at that time
among the Saxons as an article of belief? A negative infer-
ence might be deduced from the use of magad,—once for
Mary (1997), once for the sisters of Lazarus (3967), and

Studies in the Heliand. 78

five times for the daughter of Herodias (2760, 2766, 2770,
2777, 2784),—but also for the woman taken in adultery
(3861). It cannot be concluded, from the evidence furnished
by the Heliand, that Mariolatry was already established in
Northern Germany during the first half of the ninth century.

V. This class contains didactic or moral generalizations,
often couched in the form ‘so each man does,’ ‘ will do,’ or
‘shall do.’ They are such as naturally arise in the consid-
eration of topics bearing most directly upon the duties of life.
Here are also included predications of good or evil as attend-
ant upon contrary actions and dispositions. Instances are:
1072-5, 1458-60, 1769-70, 1824-6, 2226-31, 3659-60, 4114-7,
4375-T. This moralizing strain is in imitation of the com-
mentators (cf. 5046-9).

VI. Under this head are collected poetical expansions and
additions, often original and of great beauty, which do not fall
under any of the preceding divisions.

197-8. Flight of the winter-year :

Skred the uuintar ford, geng thes geres gital.

199-20. The comeliness of John the Baptist :

Lik uuas im sconi,
uuasim fel fagar, fahs endi naglos,
uuangun uuarun tm uulttige.

Gentility was indicated among the Germans as well by
naglos as by fel and fahs.

Further: 292-5, 827-9, 350-6, 383-6, 438-40, 447-9, 526-8,
548-62, 7382-44, 790-2, 1020-4, 1049-52, 1482-3, 2077-87,
2097-9, 2119-24, 2136-8, 2161-7, 2206-12, 2238-41, 2264-8,
2284-90, 2524-8, 2543-4, 2639-46, 2696-8, 2796-9, 2805-10,
2952-60, 8198-5, 8207-15, 3345-7, 3405-11, 3428-31,
3749-50, 3755-7, 4103-14, 4203-5, 4256-69, 4331-3, 4440-3,
4663-6, 4757-60, 4946-8, 5117-21, 5134-6, 5142-4, 5172-4,
5286-91, 5298-5303, 5365-7, 5376-9, 5894-6, 5418-20,
5562-4, 5681-8, 5827-31.

C.

A complete syntax of the Heliand is still wanting, but
Behaghel’s Mod: im Heliand will be of much service to the

74 " ASS. Cook,

future grammarian who shall undertake to supply this desid-
eratum. Only a few scattered observations are presented below.

The abundance of reflexives must instantly strike the
scholar. Following Matzner, Englische Grammatik, they may
be divided into reflexives with intransitive verbs:

(a) of Rest; (6) of Motion; and (c) of Mental Action.

(a) Verbs of Rest:

Wesan: 79, 87, 253, 506, 654, 782, 962, 987, 1027, 1052,
1121, 1175, 1198, 1227, 1233, 1234, 2112, 2187, 2219, 2401,
2465, 2495, 8294, 3329, 3953, 3969, 4239, 4632, 5695, 5716,
5865, 5964, 5983. Sittian: 988, 1176, 1286, 1291, 3332,
4273, 5946, 5976. Witan: 653, 4184, 4558. Standan: 1811,
2378, 3758. Werdan: 1198, 2401, 2408. Lnbéian: 81, 4034,
4113. Gisittian: 5805. Wonén: 989. Btan: 2706. Lig-
gian: 3336. Bidan: 5721.

(6) Verbs of Motion:

Giwitan: 458, 531, 650, 677, 712, 780, 806, 832, 873, 960,
1024, 1118, 11384, 1189, 1248, 1994, 2088, 2167, 2236, 2282,
2290, 2305, 2698, 2799, 2802, 2973, 2982, 3053, 3110, 3163,
3170, 3182, 8458, 3585, 3663, 3706, 3906, 4010, 4185, 4198,
4212, 4237, 4554, 4628, 4715, 4718, 4769, 4786, 4796, 5159,
§312, 5440, 5729, 5748, 5762, 5870, 5910, 5974.

Gangan: 102, 477, 1127, 1150, 2000, 2334, 2381, 3878,
3893, 3913, 4270, 4478, 4526, 4798, 4804, 4838, 5001, 5061,
5150, 5176, 5232, 5584, 5693, 5703, 5715, 5722, 5906. Faran:
683, 718, 796, 1136, 1228, 2292, 2488, 2676, 2698, 2894,
3482, 3541, 51638, 5776, 5956. Gehnigan: 981, 8122, 4744.
Wendian: 699, 3298. Kuman: 1235, 3184. Fiscén: 1156.
Kinun: 2409. Talén: 2471. Farféhan: 2503. Stigan:
2681. Thurugangan: 3488. A’rtsan: 4714. A’wahsan: 859.

(ec) Verb of Mental Action:

Andrddan: 116, 1908, 1907, 2252, 3157, 4882, 5818.

Besides, a reflexive is found with diginnan: 312, 1148,
2389, 2395, 2402, 2499, 2710, 2942, 3233, 3325, 3478, 3485,
8495, 5062, 5072, 5959.

In 859, 4100, the dative should probably be coupled with
was and not with the participle. The frequent use of a re-
flexive with wesan might easily be the occasion of this idiom.

Studies in the Heliand. — 75

Appositives with the definite article after a vocative are
common (March, Anglo-Saxon Gram. §289,a). So Herro the
godo (1588), Drohtin the godo (1607), Fro min the godo
(2099). Further, 2105, 2428, 2550, 2824, 2935, 3258, 4082,
4080, 4292, 4403, 4509, 4517, 4685. —

Genitives with the definite article or demonstrative are also
of frequent occurrence. The place of the latter varies.

(a) Before the genitive: Thana godes sunu (1384), themu
is liohte (1548), them is godun uuercun (1687), thin godes
lera (2479), thene godes sunu (2671), them is godun tungarun
(3176), them is lerun (4196), them is uuordun (4205), them
18 wungarun (4635). .

An adjective precedes the genitive: T'hie guodo godes suno
(4011, 5089).

An adjective follows the genitive: ANd thiu 1s godum gum-
scepi; cf. the attributive genitive in Greek (Hadley, § 531).

(6) After the genitive: Is thana fader (228), is thane
endi (1856), 78 thero gesteo (2045).

.The repetition of the same thought in two consecutive lines
often brings two synonymous words or phrases to stand on
opposite sides of a part of speech relatively superior, which
accordingly may be regarded as goverying or supporting either
one of the two, while the other is appositional. Abbott deals
-with a somewhat similar phenomenon in his Shakespearian
Grammar, § 513. An example of the peculiarity in question
is found at 1306:

Thie motun thie marion erde
ofsitiven that selbe rik,
Other instances occur at 1988, 2018, 2672, 2711, 4000,
4114, 4204, 4216, 4337, 4879, 4612, 4742.

T6 A. Harkness,

V.—On the Development of the Latin Subjunctive in Principal
Clauses.

By ALBERT HARKNESS,

PROFESSOR OF THE GREEK “LANGUAGE AND LITERATURE IN BROWN
UNIVERSITY.

It is the object of this paper to give a brief outline of the
probable development of the Latin Subjunctive in Principal
Clauses. It aims to trace the steps by which the various
meanings of this mood, as seen in Latin authors, have been
gradually developed out of the simple etymological force of
its original forms.

The Latin Subjunctive contains the forms of two moods,
originally distinct, though closely related,—the subjunctive
proper with the sign a, and the optative with the sign ¢.
These forms, however, are used without any difference of
meaning, and are made to supplement each other. Thus
the subjunctive forms are found in the present tense of the
second, of the third, and of the fourth conjugation, as moneas,
moneatis, regam, regamys, audiat, audiant, while the optative
forms are used in all the other tenses of these conjugations,
ahd in all the tenses of the first, as monerem, rexerim, audit.
vissem, amem, amarem, etc.

Moreover the Latin Subjunctive contains the meanings, as
well as the forms, of two distinct moods—the subjunctive and
optative, of the cognate tongues, and what is especially note-
worthy in this connection is that while in Sanskrit and Greek
the subjunctive and optative meanings are denoted by separate
forms, in Latin they are both expressed by the same form.

Again in Latin the subjunctive and optative forms are not
by any means confined to the significations ordinarily ascribed
to those moods in the cognate tongues, but have a much wider
range of application. Thus:

1. They supply the place of the future indicative in all
verbs of the third and of the fourth conjugation. Thus the
subjunctive form regam, ‘let me rule,’ also supplies the future

On the Latin Subjunctive in Principal Clauses. TT

‘TI shall or will rule;’ audiam, ‘let me hear,’ is also future,
‘I shall or will hear.’ But in the other persons, the optative
forms supply the future. Thus reges, reget, etc., though opta-
tives in origin and form, are never used in an optative or a
subjunctive sense, but only with the force of the future indica-
tive, not ‘may you rule,’ ‘let him rule,’ etc., but, ‘ you shall
or will rule,’ etc.

2. The forms of the Latin Subjunctive not only thus supply
the future indicative, but they are also used with their own
proper force in various subordinate clauses which originally
must have taken the indicative. They thus occur in causal
and temporal clauses, in dependent questions, and in the
subordinate clauses of the oratio obliqua.

Such are some of the peculiarities of the form and use of the
Latin Subjunctive, peculiarities whose complete explanation
can be found only in the general development of the mood
itself. Special discussions of separate points are of course
indispensable, but they often rest upon too narrow a basis to
give the best results. Such discussions of special usages in
the Latin Subjunctive have often been for this very reason but
partially successful. Constructions, closely connected both
with each other and with the general subject, have been taken
out of their proper connections, as parts of one complete
whole. But any trustworthy discussion of the Latin Subjunc-
tive, or of any vital points connected with it, must rest upon
a much broader basis. It must gather light, not only from
the whole range of the Latin language, but even from the
cognate tongues. Such a discussion will accordingly involve:

1. The general development of the subjunctive and optative
moods, as shown by constructions common to the Sanskrit,
the Greek, and the Latin.

2. The later special development of the mood on Latin
soil.

What then, we inquire, was the origin of the two classes of
forms appropriated to the subjunctive and optative moods, and
what was their original force ?

It is well known that the Indo-European language developed
various methods of forming the present stem from the root..

11

78 A. Harkness,

The stem thus formed differed in meaning from the root simply
in the fact that it denoted continued action, while the root
expressed only the general idea of the action itself, without
any reference to its continuance. Of the various forms of
the present stem originally used, three became especially
important in the subsequent development of verbal inflections.
These were:

1. The reduplicated present stem, from which was devel-
oped the Indo-European perfect. For a discussion of this
subject, it is only necessary to refer to an able paper read
before this Association four years ago by Professor Alonzo
Williams.

2. The present stem in a, from which was developed the
Indo-European subjunctive.

3. The present stem in ja =z, ‘to go,’ as seen in eipe, eo.
From this were developed the optative mood and the future
indicative.

Thus both the subjunctive and the optative mood are in
their origin only special developments of certain forms of the
present tense, a view now generally accepted, I think, by
the leading scholars of the new school. Let us now briefly
illustrate this point in its relation to each of these moods.
We shall thus find, [ trust, that these etymological forms
throw light upon the subsequent development of the Latin
Subjunctive.

It will be remembered that in the mother-tongue of the
Indo-European family, from which the Latin, the Greek, and
the Sanskrit alike derived their various inflections, the original
type of a verb consisted simply of the union of a verbal root
with a pronominal root or stem. Thus from the root da was
formed da-ta, Latin dat, ‘he gives;’ from root bhar, bhar-ta,
Latin fert, ‘he bears.’ Subsequently these roots were devel-
oped into stems in various ways, especially by the addition of
the determinative a. These stems were then inflected by the
addition of pronominal roots or stems. Thus from the root
bhar was formed the stem dbhara, from which comes the verb
bhura-ta. The language then contained two sets of forms,
one denoting simply the action of the verb, the other that

On the Latin Subjunctive in Principal Clauses. 79

action in its continuance, progress, as bhar-ta, ‘he bears,’
bhara-ta, ‘ he is bearing,’ ‘is trying to bear,’ or ‘is one who
bears.’ This latter form, denoting as it. does continued or’
prolonged action, is especially fitted to emphasize the idea of
effort. Indeed the conative use of the present indicative is
distinctly recognized in all ages of the Latin literature. But
earnest effort readily suggests deszre, as one strives only for
that which one desires to attain. Hence bhara-ma, ‘1 am
bearing,’ or ‘ trying to bear,’ comes also to mean ‘I desire,
intend, purpose, to bear.’ But vigorous effort suggests not only
desire, but also possibility or probability, as one will very likely
accomplish that which one is already attempting. Here then
we have, as I conceive, the original form and meaning of the
Indo-European subjunctive, a special form of the present
indicative used in a somewhat special sense to denote, first,
an attempted action, and secondly, a desired, possible, or
probable action. The development of the mood was doubtless
exceedingly slow and gradual. At first the difference in
meaning between this particular form and the other forms of
the present tense was scarcely perceptible, but it gradually
became more and more marked until at length a new mood
was recognized. Then for the first time in the history of the
Indo-European language did the subjunctive mood have a
recognized existence, as distinct from the indicative.

But the Latin Subjunctive, as we have already noticed, also
contains the Indo-Kuropean optative, a mood developed from
that form of the present stem whose formative element is the
verb ja, or 2, ‘to go.’ The stem in ja, which, like all forms
of present stems, denoted originally duration, continued action,
became the basis of several important verbal inflections, of
which we notice the following :

1. The present indicative, as ayyé\Aw = ayyéAjw, ‘I go asa
messenger, am a messenger, [ announce.’

2. The future indicative, as ertt —es-it for es-i-ti, ‘he is
going to be;’ dwow = ewo-jw, ‘1 am going to give.’

3. The optative, as sit = est = es-ie-tz, identical with the
original form of the future, and like that meaning originally
‘ he is going to be;’ Soing = b0-in-or, * you are going to give.’

80 A. Harkness,

Now these etymological facts are instructive in various
ways. They show

1. That the Indo-European optative was identical in origin
both with the future indicative and with one form of the
present.

2. That it was used originally to denote a contemplated,
future, or probable action, as doiny, ‘I am going to give,’ almost
synonymous with desu, ‘I shall give.’ Here we recognize at
the very outset in the development of the mood one of the
familiar uses of the Greek optative, viz., its potential use
with ad», often best rendered into English by the future
indicative with which it was identical in origin.

But this etymological meaning of the optative naturally
suggests desire, wish, as doiny, I am going to give, intend to
give, readily suggests the kindred thought, J would like to
give, may I give.

We have thus reached for the Indo-European optative two
distinct but closely related meanings, which may be regarded
as primitive and etymological,—meanings, moreover, which
accord very exactly with those previously found for the
subjunctive.

From the facts now presented it seems clear that originally
the subjunctive and the optative were closely related, both in
form and meaning, not only with certain parts of the indicative,
but also with each other.

But we have already noticed the fact that the Latin Subjunc-
tive contains the general meanings of the subjunctive and
the optative of the cognate tongues, and that these meanings
are denoted sometimes by subjunctive forms, and sometimes
by optative forms, 1. e., indiscriminately by either. We have
also observed that these same forms in a large class of verbs
regularly supply the place of the future indicative.

What now is the explanation of this remarkable confusion
in the use of forms, a confusion so great that different forms,
subjunctive and optative, occur with precisely the same mean-
ing, and the same forms with different meanings? Did the
Latin lose important distinctions which it had inherited from
the Indo-European, or were these kindred forms which we

On the Latin Subjunctive in Principal Clauses. 81

have been considering so closely related in meaning when that
language first began to have a separate existence, that they
might be used with comparatively little difference of meaning ?
That the Latin may have lost some distinctions that it once
had, is not at all improbable, but that it ever possessed the
nice and delicate distinctions seen in the Greek subjunctive
and optative is in the highest degree improbable. Indeed
these distinctions are, I think, now regarded by those whose
opinions are entitled to the greatest weight as the special
product of the Greek mind. Accordingly, in my judgment,
the indiscriminate use in Latin of subjunctive and optative
forms in connection with the regular use of those same forms
to supply the place of the future indicative in many verbs,
contains an important historical fact in the development of
these moods. It shows that, when the Latin first became a
separate language, the forms of the subjunctive, of the opta-
tive, and of the future indicative were used with little or no
difference of meaning, a view fully confirmed, as we have
just seen, by the etymology of the forms themselves.

Having thus examined the etymological meaning of the
Latin Subjunctive forms, we proceed, in the second place, to
inquire what relation this meaning sustains to that which is
actually found in the works of Latin authors. In conducting
this inquiry we shall find it necessary to study the Latin
Subjunctive, first in simple sentences and principal clauses,
and secondly in subordinate clauses.

What then appears to be the primitive or fundamental
meaning of the Latin Subjunctive as seen in Principal Clauses ?
That the primitive features of the mood were first developed
in principal rather than in subordinate clauses scarcely admits
of adoubt. The old theory which made all subjunctive clauses
subordinate by supplying supposed ellipses, is so completely
exploded as scarcely to require a passing remark, yet traces
of its influence may be discovered, even in valuable works, of
recent date. The most trustworthy scholars, however, are all
agreed in the opinion that the subjunctive and optative moods,
in all their general and characteristic features, as seen in the
Sanskrit, the Greek, and the Latin, were developed in principal
clauses. |

| 82 A. Harkness,

But what has already been said in regard to the etymological
meaning of the subjunctive forms renders it clear that we
must not expect to find, as the result of this inquiry, any broad
and well defined line of distinction between the provinces
originally occupied respectively by the indicative and the
subjunctive. Accordingly, when we turn to the Vedas, which
furnish us our earliest specimens of Sanskrit, and to the poems
of Homer, which furnish us our earliest examples of Greek
syntax, we often meet the subjunctive in uses much more
closely related to that of the future indicative than in later
works. Indeed, in such examples as the Homeric

ov yap zw Tolove idov avépac, ove idwpat, Lliad 1, 262,

‘for I have never seen, and I do not expect to see such heroes,’
the substitution of the future indicative for the subjunctive
would scarcely make a perceptible change in the thought. It
is also well known that in the Homeric poems the aorist sub-
junctive and the future indicative are often identical in form,
and that the difference in meaning between them is sometimes
so very slight that it is not possible even to distinguish the one
from the other. We have already noticed the familiar fact that
the optative with &» is often used in a future sense. But the
Latin, Greek, and Sanskrit, in the use of moods, conform to the
same general analogy. Indeed the use of subjunctive and opta-
tive forms in Latin to supply the place of the future indicative
does not differ at all in kind from the Homeric use of similar
forms in a future sense ; it is in fact simply the result of carry-
ing out that analogy on a larger scale. In view of these facts
the conclusion seems inevitable

1. That when the Sanskrit, the Greek, and the Latin first
became separate languages, the distinction between the indica-
tive, on the one hand, and the subjunctive and optative on the
other, so clearly recognized in the classical period, was only
partially developed.

2. That among the various meanings denoted by subjunc-
tive and optative forms in the earliest writers, those which are
most closely related to the general meaning of the indicative
may safely be regarded as the earliest. We recognize, there-
fore, in the signification just noticed, as belonging to the early

On the Latin Subjunctive in Principal Clauses. 83

Sanskrit, Greek, and Latin, one of the most primitive uses
of these moods, the potential use denoting expectation, likeli-
hood, a contemplated action.

But the Latin Subjunctive, like the Sanskrit and Greek
subjunctive and optative, also denotes desire, wish. It is thus
used:

1. In prayers and wishes: Di bene vertant, ‘ may the gods
cause it to turn out well,’ Plaut.

2. In exhortations and entreaties: Consultamus bonis, ° let
us consult for the good,’ Cic.

3. In commands, admonitions, warnings, and especially
in negative commands, prohibitions: Scribere ne pigrere, ‘do
not neglect to write,’ Cic.

4. In admissions and concessions: Fuerint pertinaces,
‘grant or admit that they were obstinate,’ Cic.

This use of the subjunctive, like the potential, is readily
developed out of the etymological signification of the subjunc-
tive form, as we have already seen, since earnest effort natu-
rally implies desire. Indeed,in the Vedas and in the Homeric
poems we sometimes find the subjunctive of desire apparently
in the very first stage of its development, scarcely distinguish-
able, on the one hand, from the future indicative, and on the other
from the potential subjunctive, but perhaps differing from the
latter very much as the two ordinary signs of the English future
— shall and will—differ from each other. In the Homeric
expression already quoted, ov ydp ww ivr, ovde iowpat, the sub-
junctive involves no idea of desire; it is entirely potential in
its nature, but is as yet so imperfectly developed as to be
scarcely distinguishable from the future. It may even be
rendered by the future shall, but not by will. In some pas-
sayes, however, as

devre, dvw prot Exeobov, idwp’, driv’ Epya réruxrat, Tliad 22, 450,

‘ here, you two follow me, I wish to see what has happened,’ the
subjunctive denotes desire, yet it is so closely related to the
future that it may be rendered by our auxiliary wilt, J will see.
Indeed, in some instances in Homer, it is very difficult to
determine even from the context whether a given subjunctive

84 A. Harkness,

should be interpreted as potential, or as expressive of desire,
or even to say why the subjunctive is used at all, as
aA’ ay’, éywr abrocg Tephoopac Hoe tdtwuar, Odys. 6, 126,
‘but come, I will myself try and see, or let me try and see ;’ or
again, |
add’ aye viv éxipecvor, apiia revyea Sdw, Iliad 6, 340,

‘but come, now wart, let me put on my martial armor,’ or ‘I will
put on, etc. The subjunctive in such instances appears to be

still in embryo.
We have now seen that these two uses of the Latin Sub-

junctive, and of the corresponding moods in the cognate
tongues, the subjunctive and optative, appear in different
stages of development in the very earliest literary records
that have come down to us from any branch of the Indo-
European family, and that they are readily and naturally
derived from the etymological meaning of the forms themselves.
Moreover it is well known that these two meanings, which for
convenience we may call potential and optative, run through
the whole range of Sanskrit, Greek, and Latin literature, and
that in their several forms they embrace all the meanings
known to the Latin Subjunctive in Principal Clauses.

If the doctrine of this paper is correct, it is quite clear that
the view once so generally received that the potential subjunc- -
tive was first developed in conditional sentences is no longer
tenable. We have already seen that the potential use of the
subjunctive springs directly and naturally from the etymolog-
ical force of the forms, and that it is also found fully devel-
oped in simple sentences in the earliest literary records extant.
I may also add that there are strong reasons for believing that
it was long used before conditional sentences were known.

But before we close this part of our discussion, I must
anticipate an objection from those who adopt without qualif-
cation the views of Professor Delbriick, as set forth in his
masterly treatise on the use of the subjunctive and optative
in Sanskrit and Greek, a work, let me add, greatly superior,
in my judgment, to all others upon the subject which it treats.

Delbriick, in his attempt to find the primitive meaning of
these moods, assumes without authority, as it seems to me,

On the Latin Subjunctive in Principal Clauses. 85

that it is to be sought only in the first person singular, thus
excluding from his investigation all plural forms, and all forms
in the second and third persons. He does so simply because,
as he himself says, it is generally admitted that the primitive
meaning of the optative is that of wishing, and that this
meaning is expressed in its greatest purity in the first person
singular. His own words are: ‘“‘ Es wird vermuthlich jetzt
allgemein angenommen, dass die dlteste Bedeutung des Opta-
tivs der Wunsch sei.’’” But if the common opinion that the
idea of wishing lies at the basis of the optative is correct,
what occasion is there for a special investigation on the part
of Professor Delbriick to ascertain the primitive use of that
mood? Qn the other hand, if the correctness of such an
opinion is questioned, what right has the Professor to limit
his inquiry to the particular person and number in which that
meaning appears, and to reject the other persons in which a
different idea is prominent? How does he know in advance
that the latter may not be as truly primitive as the former ?
This assumption, it seems to me, has led Professor Delbriick
to overlook one important element in the primitive use of the
subjunctive and optative. Assuming that the original force
of each is seen in the first person singular, he reaches the
conclusion that the subjunctive originally denoted will, deter-
mination ; the optative wish, desire.

Now if we examine subjunctive and optative forms in the
second and third persons in the identical classes of sentences
examined by Delbriick, we shall find the potential idea, in
the form of likelihood, probability, possibility, just as distinctly
and clearly expressed as is that of desire in the first person.
I claim, therefore, that we have the strongest possible reasons
for believing that two distinct meanings, apparently equally
primitive, were developed at a very early period, that of desire
in the first person, and that of possibility, likelihood, in the
second and third, and that subsequently these two meanings,
though especially conspicuous in the particular forms in which
they were respectively developed, ‘were extended to all the
persons.

We have thus far considered the development of the Latin

12

86 A. Harkness,

Subjunctive only in those clauses which remained independent
throughout the classical period, but the subjunctive in many
subordinate clauses also belongs to our theme, as it was
developed while those clauses were yet independent. Modern —
linguistic research has clearly established the fact that orig-
inally the syntax of sentences in the Indo-European family
of languages was exceedingly simple. Even in connected
discourse, thoughts were presented separately in simple sen-
tences, or followed each other in codrdinate clauses, and yet
the thoughts thus expressed were by no means equally impor-
tant. In fact, one of them might be quite subordinate to
another, and yet be presented as entirely coordinate with it.
Homer and Plautus furnish abundant illustrations of this.

Subordinate clauses, originally unknown, were subsequently
developed, by a slow and almost imperceptible change, out of
simple sentences and principal clauses. Every subordinate
clause therefore represents an independent sentence or clause,
and accordingly in examining a subjunctive in such a clause,
our first inquiry must relate to the history of the mood in this
special instance. We must ascertain whether it owes its
origin to the nature of the thought itself, or has been devel-
oped by the subordinate character of the clause. Upon
investigation I think we shall find that in conditional sentences
and in concessive, final, and consecutive clauses in Latin, the
subjunctive is entirely independent of the character of the
clause in which it stands, and was in fact developed before the
clause became subordinate, but that in causal and temporal
clauses, in dependent questions, and in the subordinate clauses
of the oratio obliqua, the mood in many instances has been
developed simultaneously with the subordinate character of
- the clause, and is entirely dependent upon it.

We proceed to examine the cases in which the subjunctive
appears to be original, that is, to have belunged to the clause
in its original and independent form. Some of these in the
use of moods conform, to a considerable extent, to the corres-
ponding constructions in Sanskrit and Greek. For this par-
ticular part of our discussion, therefore, we shall find abundant
inaterial and important aid in the work of Delbriick to which
I have already referred.

On the Latin Subjunctive in Principal Clauses. 87

We begin with conditional sentences. That the mood in
the condition does not depend upon the conditional particle is
obvious from the following facts:

1. The conditional particles ci, st, nist, are used with differ-
ent mgods. In fact, in Latin # is used, or may be used, in
every conceivable form of condition.

2. <A conditional particle is not even an essential part of a
conditional sentence. The force of such a sentence may be
expressed in Greek, Latin, or English without any particle
whatever. Thus: NMegat quis, nego, ‘does any one deny, I
deny’ = if any one denies, deny. Roges me,nthil respondeam,
“ask me, I will make no reply.” Lacesse; gam videbis furen-
tem, ‘ provoke him, i. e., if you provoke him, you will see him
Frantic.’ In these examples, it will be observed, we have the
full force of conditional sentences expressed without any
conditional particles, in clauses entirely independent, yet
containing all possible varieties of finite moods, the indicative,
subjunctive, and imperative. Moreover, in such examples as
these, we find, I conceive, the original type out of which
conditional sentences were developed. Originally the two
clauses, the condition and the conclusion, were entirely inde-
pendent, as in the examples just noticed, and the mood in each
was determined by the ordinary principles which regulate the
use of moods in principal clauses. The indicative was used in
treating of facts, and the subjunctive or imperative in all other
cases. 2, probably the locative case of an indefinite pronoun,
meaning at any ime or in any manner, has nothing whatever
to do with the mood, but merely serves to show that the action |
in the conclusion is connected in time or manner with that in
the condition. Thus, without the particle, negat, nego, ‘he
denies, I deny ;’ with the particle, s¢ negat, nego, literally ‘ he
denies (at some time, then or at that time), J deny.’ Here
we have the indicative, whether with or without the conditional
particle. Let us now analyze a case with the subjunctive: Dies
deficiat, si velim numerare, etc., ‘ let me at,any time wish to
recount, etc.. then the day would fail me.’ Here we have
distinctly before us the subjunctive of desire in the condition,
and the potential subjunctive in the conclusion, the identical

88° A. Harkness,

forms which we have met in principal clauses. The subjunc-
tive in conditions is always a subjunctive of desire, that in the
conclusion is generally potential, though that also sometimes
denotes desire, as, peream, si poterunt, etc., ‘they will some-
time be able, etc., then let me perish.’

Concessions are so closely related to conditions that they
scarcely require separate notice. The subjunctive is the same
in both. Thus the subjunctive of desire is easily recognized
in the following examples: Quamvis sit magna exspectatio,
tamen eam vinces, ‘let the expectation be as great as you
please, even then, or even thus you will surpassit.’ Absolvite
Verrem qui se fateatur pecunias cepisse, ‘acquit Verres, let him
admit that he accepted money.’ We must not forget that the
clause introduced by gut was originally independent. ncet,
irrideat, plus tamen ratio valebit, ‘it is permitted, let him
deride, with all this derision reason will avail more.’ In this
short sentence we have three clauses, each one of which retains
the identical mood which it adopted when it was an independent
sentence.

Again final clauses need not detain us long, as purpose nec-
essarily involves desire, and is readily developed out of it.
Such clauses therefore always contain a proposed or desired
action, and accordingly in Latin require the subjunctive, as:
Punit ne peccetur, ‘he punishes, let not crime be committed.’
Here the second clause contains the desire, command, purpose
involved in punit. The independent sentence out of which it
was developed already contained the subjunctive of desire, and
in fact remains unchanged in form in the final clause before
us. Servis imperat ut filiam defendant, ‘he commands his
servants, so let them defend his daughter.’ By thus restoring
the final clause to its original independent form, we can
appreciate, in a measure, the true force of the subjunctive of
desire. Ze rogo ut eum juves, ‘I ask you,so aid him.’ Legum
edeirco servi sumus, ut liberi ease possimus, ‘we are servants of
the law for this reason, accordingly let us be able to be free.’
Vereor ne laborem augeam,‘I fear, let me not increase the
labor.’ Periculum est ne ille te verbis obruat, ‘there is danger,
let him not overwhelm you with words.’

On the Latin Subjunctive in Principal Clauses. 89

Moreover, in the Homeric poems, where final clauses are
less fully developed than in later Greek or in Latin, we some-
times meet with passages in which it is difficult to say whether
we, have before us an independent clause containing the
the subjunctive of desire, or a dependent clause expressing
purpose. Such passages are especially instructive on the
question of the origin of final clauses. Thus

pw’ olov édoare,...
. ixéof? exi vijac ’Ayawy,
Aioowp’ avipa rovroy, Iliad 22, 416,
‘let me go forth alone to the ships of the Achaeans that I may
supplicate this man.’ But perhaps the original conception is
here preserved, let me supplicate this man. Again, take a
negative sentence,
pndé trv’ darvoc
aipeirw, py xappa yevapueda dvoperécoory, Iliad 10, 192,
as purpose, let not sleep take possession of any one, lest we
become a joy to our enemies, but perhaps better as independent
clauses, let not sleep take possession of any one, let us not
become a source of joy to our enemies.

Consecutive clauses are in origin, form, and history closely
related to final clauses. Both refer to the result of the action,
the former contemplating it as probable, as something to be
expected, the latter as something desired, intended, proposed.
While, therefore, the subjunctive in final clauses is found to be
a subjunctive of desire, that in consecutive clauses, denoting
simply expectation, is potential: Mon is sum, qui his utar, ‘I
am not the one who is likely to use these things.’ Aristides
tta virit ut Atheniensibus esset carissimus, ‘Aristides so lived
that he was likely to be very dear to the Athenians.’ Observe
that the subjunctive, in examples like this, did not originally
express an actual result, but only a contemplated one, that
which would follow as a matter of course. Then, in relation
to a past event, the transition became very simple and natural
from a contemplated result which was to be expected in a given
case, and which would follow as a matter of course, to that
which did follow as a matter of fact. Hence the Latin, which
meant, originally, he 80 lived that he would of course be very

90 A. Harkness,

dear to the Athenians, may be not improperly rendered, he so
lived that he was very dear to the Athenians.

An additional argument in support of the view that the
subjunctive in consecutive clauses is potential in origin and
force, is found in the fact that it takes the negative non, the
regular negative of the potential subjunctive, while final
clauses take ne, the regular negative for the subjunctive of
desire, as: lta vixi, ut non frustra me natum existimem, ‘I
have so lived that I do not think, or, more literally, that I am
not likely to think, that I was born to no purpose.’

In the cases which we have thus far examined, the Latin
Subjunctive appears to be original, that is, to have been
required by the very nature of the thought contained in these
clauses in their original and independent form. We find here
no new use of moods, but simply the two distinct and well
recognized uses of the Latin Subjunctive which run through
all departments of Latin literature in all ages of its history,
uses moreover not at all peculiar to the Latin, but belonging
equally to the Sanskrit and the Greek.

In conclusion, it only remains for.us to notice the fact that
these two primitive uses of the subjunctive sometimes occur in
gtill other classes of subordinate clauses, having been devel-
oped in them while they were still independent. As they
appear, however, in forms much less disguised than those
already examined, they will in general be readily recognized.
A few brief examples will be a sufficient illustration of this
point, as: Htzam tum vivit, guom esse credas mortuam, said of
infamy, ‘even then it lives, when you would suppose it to be
dead,’ the potential subjunctive in a temporal clause. Dif-
ferant, dum defervescat ira, ‘let them defer it till their anger
cools, i. e., that it may cool, or to come still nearer the
original conception, let them defer it, in the mean time let
their anger cool.’ In most other cases, as in dependent
questions, and in the dependent clauses of the oratio obliqua,
an original subjunctive is so easily recognized that it does not
even require illustration.

We have thus endeavored to trace in outline the development
of the Latin Subjunctive in Principal Clauses from its origin

On the Latin Subjunctive in Principal Clauses. 91

in the Indo-European tongue to its latest uses in the works of
Roman authors. We have scen:

1. That its forms, etymologically examined, denote effort,
attempted action, and are closely related in origin and force
to the conative present.

2. That attempted action suggests :

a. Desired, proposed action.

6. Probable, possible action.

3. That the uses of the subjunctive actually found in
principal clauses in the works of Latin authors naturally
arrange themselves in two classes, corresponding to the two
meanings already derived from the etymology of the forms,
and authorize us to distinguish :

a. An optative subjunctive, or a subjunctive of desire.

6. <A potential subjunctive.

4. That to one or the other of these two classes belong:

a. The subjunctive in all principal clauses.

6. The subjunctive in conditional, concessive, final, and
consecutive clauses. |

e. All original subjunctives in other subordinate clauses.

But our discusssion now brings us to a use of the Latin
Subjunctive in subordinate clauses, which is peculiar and
special, which finds few analogies in Sanskrit or Greek, which
is not strictly potential or optative, and which seems to have
few claims to be regarded as original. It occurs in causal
and temporal clauses, in dependent questions, and in the
subordinate clauses of the oratio obliqua. Unlike the uses
already considered, it seems never to have been developed in
principal clauses. It does not therefore fall within the scope
of this paper, and the discussion of it must be reserved for a
future occasion.

92 ML. D’ Ooge,

VI.— The Original Recension of the De Corona.

‘By M. L. D’OOGE,

PROFESSOR OF GREEK IN THE UNIVERSITY OF MICHIGAN.

Just what relation the speech of an Athenian orator deliv-
ered in court, in answer to a counter-plea, holds to the same
speech as subsequently revised and publislred, is a question that
can never be settled as to details. But as regards the more
general features, such as consistency of argument, evenness
of finish in all parts, unity of plan, direct reference to points
made by the first speaker,—in all these matters it would seem
that the internal évidence gained from a careful comparison
of two rival speeches would be sufficient to determine, with
some degree of accuracy, not only the relation of two such
speeches to each other, but also of each speech in its revised
form to the same speech as originally planned and delivered.

An examination, for example, of the speech of Demosthenes
vs. Timocrates shows verbal repetitions, contradictions, and
less careful finish in the second half. These facts led Benseler
to suppose that this oration is a patchwork made up of the
speech of Demosthenes vs. Androtion, of that of Euctemon
vs. Timocrates, and of a speech of Demosthenes vs. Timocrates.
Schaefer regards the speech as a combination of two drafts
or sketches by Demosthenes, an earlier and a later; the
earlier was directed, he thinks, against both Androtion and
Timocrates ; but, since Androtion and his associates paid the
prize-money while the action was pending, the orator prepared
the second draft to meet the changed situation. This is
substantially the opinion, also, of Blass. Thatin the prooemium
the payment of the money should be denied, in the statement
of the case (cf. § 11-16) be granted, then again denied in
the second part of the speech (cf. 121, 131), not to mention
the cases of hiatus and violations of the, Demosthenic law of
eipuSpia to be found in the second half of the oration, is suffi-
cient evidence that this speech cannot originally have been
cast in a single mold.

On the Original Recension of the De ‘Corona. 93

This oration against Timocrates affords at once both an
illustration of Kirchhoff’s theory of the original recension of
the de Corona, and, by way of contrast, as we shall attempt
to show, a reason for doubting the soundness of that theory.

Kirchhoff’s. theory of the origin of the de Corona in its
present form, is contained in the Abhandlungen der Kénig-
lichen Akademie der Wissenschaften zu Berlin, 1875, and is
substantially as follows: It is well known that the trial of
Ctesiphon was, for some reason, delayed for several years.
Demosthenes, according to Kirchhoff’s theory, wrote out a
plea soon after the bringing of the indictment, in 336 or 335
B.C. This plea could have been directed only against the
formal writ of complaint, for beyond this nothing could be
definitely known of the attack of Aeschines. The de Corona
in its present form plainly divides itself into four parts: first,
§ 1-8; second, § 9-58; third, § 53-121; fourth, § 122-324.
The third part is that which contains the reply to the specific
charges of the indictment.

It is this part (538-121), plus 3, 4, and 8 of the prooemium,
which constitutes the original plea.

But when the trial came off, some six years later, the
situation was changed. Ctesiphon and Demosthenes then
agreed to divide the material of the defence, Ctesiphon taking
the legal points and Demosthenes the political issues. Accord-
ingly, Demosthenes, after careful preparation, but in extem-
porized language, replies to his antagonist’s attack upon his
political career, and makes no use of the original plea drawn
up six years before. After his case was won, Demosthenes
concluded to publish his speech. |

In preparing his oration for publication, it was the purpose
of Demosthenes to incorporate with the speech actually deliv-
ered by him at. the trial, the original plea on the legal points,
so as to present the entire defence in a complete form. But
in remodeling the old plea he must, of course, notice the
statements of his opponent, and, by interpolated passages,
meet the points he had not anticipated. In proof of this
recasting and interpolating, Kirchhoff cites § 73-79 and
§ 95-101. In § 75—Toiro peév roivuy...... Aéye—Kirchhoff

13

94 M. L. D’ Ooge,

discovers an interpolation out of all harmony with the context:
A. Mommsen, however, takes this, it seems with good reason,
as a passage inserted by the author of the spurious documents
immediately following.

But Demosthenes soon became convinced that this attempt
to recast the old plea, so as to make it seem an organic part
of the speech spoken at the trial, could not succeed, and the
project was abandoned. Thereupon he contented himself
with reducing to writing the speech actually delivered by him,
reproducing, as closely as possible, the language and the
arguments employed. Thus arose the second or younger
speech, which is preserved in § 1, 2, 5, 6, 7, 10-52, 122-324.
It was the intention of Demosthenes to publish only this
younger speech and to keep the older in retirement. But cir-
cumstances unknown to us prevented this intention from
being realized. The unknown editor of the de Corona found
among the literary remains of the orator both the older plea,
bearing the marks of an uncompleted recasting, and the
- completed transcript of the speech delivered at the trial;
and, in the belief that the orator intended that both these
pleas should be united, the editor took up the task of inter-
jecting the older into the body of the younger speech, so as
to form one united and complete defence. The work, though
mechanical, was done with discretion. In conclusion, Kirch-
hoff expresses the hope, that, while many may not be inclined
to adopt his theory without more proof, he has at all events
succeeded in shaking the universal admiration of the de Corona
as a faultless masterpiece of rhetorical art. __

Kirchhoff starts out with the assumption that Demosthenes
could not have arranged the subject matter of the de Corona
in its present form, without possessing a minute acquaintance
with the contents of the speech of his opponent. With this
assumption we cannot agree, except so far as it may apply to
certain direct references to the speech of Aeschines which will
be noticed below. Furthermore, Kirchhoff thinks that Demos-
thenes had no expectation that Aeschines would charge him
with being the promoter and author of the peace of Philocrates,
and regards, therefore, the entire second division, § 9-58, as a

On the Original Recension of the De Corona. 95

part of the later and actually delivered speech inserted in the
earlier plea. .

But this charge of Aeschines is only a degree more bold
than that which he had made some thirteen years before, in
his defence concerning the embassy. From a comparison of
Aeschines, de Falsa Legat., § 58-61 and 79, with contra Ctesi-
phontem, § 60-71, it appears that the difference in the attitude
of Aeschines is simply this: in the latter speech he is almost
silent about his own relation to the peace, and in the former
he charges upon Demosthenes, in company with Philocrates,
not the authorship of the peace, but bribery in its negotiation.
When we take into the account the course of events after 344,
it is not at all surprising that Aeschines, in his desire to charge
the policy of Demosthenes with as many disastrous results as
possible, and with his talent for misrepresentation, should, in
330, have modified the more general charge of copartnership
with Philocrates in bribery, to the more specific charge of being
the joint cause of the peace. It does not seem at all likely
that Demosthenes was ‘‘ completely surprised,” as Kirchhoff .
supposes, at the charges of his opponent in relation to the
peace; and, bating a few direct allusions to the very words of
Aeschines, no good reason can be shown why Demosthenes
should not have treated this topic from the very first substan-
tially as we find it. Nor can the supposed surprise of Demos-
thenes be inferred from the language in § 225, as Kirchhoff
would have it. These words refer, if to anything definite, to
the Theban alliance, but are best understood as a rhetorical
remark upon the desperate means to which Aeschines resorts.
Somewhat similar are ric otk ay devnoe x.r-é., § 126, and the
language in § 209 where the allusion is decidedly obscure.
The whole passage (§ 9-53) on the peace may be regarded
as antedating the trial in origin, with the exception of the fol-
lowing direct allusions to the words of Aeschines: § 27-28,
beginning with raira ra ywpia, § 41, § 51, 52, and possibly
one or two more that are not so clearly distinguishable.

Kirchhoff passes now to examine the formal reply to the
bill, which is contained in § 53-121. From the nature of
the case, Demosthenes could more easily anticipate the course

96 M. L. D’ Ooge,

of the attack upon the legal questions involved than upon the
political issues to be discussed. So completely had Demos-
thenes anticipated in his preparation this part of the case,
that he needs to stop but seven times in this part of his
defence to notice the very words of his opponent. This is
just what we should expect. But in discussing the relation of
this part of the de Corona with the speech of Aeschines,
Kirchhoff leaves it an open question whether the final recen-
sion of the speech of Aeschines was prior or subsequent to
that of the speech of Demosthenes. Now a comparison of
these two speeches justifies the belief that the reviser of the
Aeschines had before him the present revision of the Demos-
thenes. Though not proved, it is properly inferred from a
comparison of Aeschines, § 225, 226, with Demosthenes, § 243,
and of Aeschines, § 189, with Demosthenes, § 319. So again
from a comparison of the arguments of the two rival orators
on the question of the place of proclamation, it seems rea-
sonable to infer that Demosthenes could not have had the
argument of his opponent in its present form before him when
he made his own; for, while he might not have made a satis-
factory defence upon this point, it does not appear probable
that he would have made so little show of defence against the
elaborate argument of Aeschines.

But the main support of Kirchhoff’s theory les in the
relation which the passage § 53-110, in which Demosthenes
treats of his public career in answer to the first count of the
indictment, is made to bear to the rest of the oration. It is
worthy of notice, to be sure, that Demosthenes should not, in
this part of his speech, review his administration beyond the
reform in the trierarchal law (Olymp. 110,1). But is it rea-
sonable to suppose that the orator, when he composed this part
of his oration, had no intention of making any allusion to the
important réle he played in the events just prior to Chaeronea ?
The part Demosthenes had in withdrawing Athens from the
war about Amphissa, in bringing about the alliance with
Thebes, in fortifying the city after the capture of Elatea, the
confidence the people continued to repose in him after the
great catastrophe—all this Demosthenes had no idea of men-

On the Original Recension of the De: Corona. 97

tioning in vindication of his policy, if we must believe the
theory of Kirchhoff. And why? Because Demosthenes says
in § 110:

kairot ra peywrra ye Tay weroXtTEvpevwy Kal WENpaypEevwY spauT@

mapaXeirw, YrokapBavwr xpwrov pey epetiic rove wepi abrov rov wapa-
vouou Adyoug axvduuvai pe deiv, cira Kav pndey eirw wept rey ora
moXrevparwy, dpoiw, rap iar Exdory TO cvvedoc Urapyety pot.
But this language is nothing more than a rhetorical artifice
common enough in transitions and introduction of new topics.
Compare the expression ovdevog TUUTWY pépynpat, § 69, and similar
turns in the seventeenth and twentieth prooemia of Demos-
thenes. We understand the orator to mean that he passes by
these most important of his public deeds for the present
(rapadkeizw = I am [now] leaving out of view), without imply-
ing that he does not intend to speak of them later. To account
for this silence about events subsequent to Olymp. 110, 1,
Kirchhoff supposes that Demosthenes, at the time of the
composition of his original plea, judged it to be politically
inopportune to discuss the affairs connected with Chaeronea
when they were still so fresh in memory. But supposing the
trial had occurred when the action was first brought, is there
not a moral certainty that the connection of Demosthenes with
these very events would have formed a prominent part of the
attack, and would have required a correspondingly vigorous
defence ?

In confirmation of his-theory, Kirchhoff cites those passages
in this division of the oration (§ 58-121) which are direct
answers to the points of Aeschines, and claims that they are
recognizable as later insertions and additions, since they can
be detached from the context without breaking the connection,
and In some instances seem to interrupt the course of the
narrative or argument. Kirchhoff mentions § 70, 73-79 as
far a8 Kai rovrote jvarrivpny, 82, 85, and 95-101 as clearly of
later origin than the body of this part of the oration.

It is not with Kirchhoff’s view of these passages that we find
fault, but with his argument; for it involves the position that in
what he holds to be the real speech of Demosthenes—§ 10-52
and 121-324—the direct allusions to the words of Aeschines

98 M. L. D’ Ooge,

are not thus easily separable and do not break the connection,
on the ground that they are not later insertions in the body of
a speech composed before the trial. It is here that Kirchhoff’s
theory seems to break down. If passages of a like character
with those above mentioned can be found in the supposed
extemporized oration, then is not only Kirchhoff’s theory
disproved, but the opposite claim that the entire oration has
the same genesis and was composed and written altogether,
either before or after the trial, is established.

To this question let us now turn. There are in the de
Corona twenty-nine distinct references to the language of
Aeschines. Of this number seven are found in the formal
reply to the indictment, three in the prooemiui, and the rest,
nineteen, are divided between divisions two and four, i. e. five
in § 9-53, and fourteen in § 122-324. This seems like a
distribution of points controlled rather by a pre-arranged
division of the entire subject matter, than like one resulting
at the moment of the trial from the order pursued by the
first speaker. The speaker who extemporizes in reply to an
attack is likely to follow, to some extent at least, the order
of his antagonist.

It is quite remarkable how widely Demosthenes departs,
alike throughout his entire speech, from the order which his
rival chose in attacking him. Numbering the allusions made
by Demosthenes to the words of Aeschines in the order of
succession in which they occur in the speech of Aeschines,
we find the corresponding sequence in the de Corona to be as
follows: 22, 28,5, 9,7, 15, 8, 8,12, 26, 27, 10, 2, 4,18, 29, 13,
14,6,17,11, 19, 21,16, 24, 1,28, 25,20. Itis just this disagree-
ment in order, as compared with the speech-of the plaintiff,
that we should expect to find in the speech of the defendant
if the structure of his speech was fixed in advance of the
trial.

The question now is, are the allusions in what Kirchhoff
calls the real speech of Demosthenes any less separable from
the context, and any more consistent with the connection than
those in the supposed original plea. While Kirchhoff does
not claim this, it is yet the necessary inference from his

On the Original Recension of the De Corona. 99

argument. What Kirchhoff claims to have shown in respect
to the allusions in § 54-121 to the language of Aeschines,
may equally well be shown, we think, in respect to similar
allusions in other parts of the de Corona. Without pretending
to he able to point out the stitches which fasten these passages
to the body of the speech, we may find traces of the seams.
Take, e. g., § 126-128. A glance at this passage shows that
its original structure has been disturbed for the sake of ridi-
culing the high-flown peroration of Aeschines. The anacolu-
thon occasioned by 3éé in the second line (a reading of the
best authority ),1s possibly an incidental mark of the departure
from the original form.

The addition in § 161-2, beginning with ob« amo rij¢ euavrov
yrwpne, the object of which is to show the false conduct of
Aeschines toward Aristophon and Euboulus (cf. Aeschines,
§ 25, 189), weakens the sentence whose natural close is with
dteréXour.

Examine next the passage from § 218 to § 247, inclusive.
The consummation of the alliance with Thebes has just been
discussed. There were two points connected with that trans-
action which required defence: first, the favorable terms
granted to Philip; second, the results of the alliance. The
passage has these peculiarities: § 227-231 is evidently a
later insertion. Demosthenes turns the point of the illustra-
tion, but with reference to an entirely different matter.

The reference in § 232, rapadciypara m\drrwy x.r. é. Seems to
have suggested the entire passage from § 232 to § 237, inclusive.
It is worthy of notice that in § 241 Demosthenes repeats the
enumeration of the results of his policy already named in § 230,
adding only one new item, sc. cat rij¢ otrowopwiac. § 244-246
is apparently called forth by the taunt of Aeschines upon the
bravery of Demosthenes. The connection of § 247 with § 239
is so close that they seem to belong together. The original
part of this passage (i.e. the part composed in advance of the
trial) may possibly be § 218-224, 238, 239, 247.

Demosthenes had still to point to the survival, after Chaero-
nea, of the confidence reposed in him, as evidenced by his
election to the office of grain commissioner and of public

100 + ML. D’ Ooge,

eulogist of those who had fallen. We cannot suppose that
he prepared beforehand any plan of his speech without
including these two points. They are treated in § 248-290.
But this passage contains also three direct allusions to the
speech of Aeschines, and a compgrison of the fortunes of the
two rivals. The reference to the case of Cephalus (§ 251) is
rather loosely joined to the context by Nai, gnoiv. If we take
§ 252-275 as a part of the original composition, ravraydSer
of § 252 seems to fit in well with the closing sentence of § 250
§ 276-284 seems to be one of the clearest instances of later
insertion in the whole oration. The opening sentence, cai xpoc
roic aAAacc, helps make the transition, but does not hide the
seam. The break between § 284 and 285 is the most abrupt
in the oration, and has been noticed as a rhetorical defect by
many critics.

The allusions to the language of Aeschines found in the
remainder of the speech are so brief and so unimportant to
this discusssion as not to require any notice. We need not
examine Kirchhoff’s analysis of the Prooemium. He has con-
fused there, as elsewhere, the notion of a subsequent revision,
in which the orator incorporates extempore passages spoken
at the trial, with the idea of two distinct speeches welded
together by some later hand.
~ The theory of Kirchhoff involves other difficulties :

(1) Demosthenes is supposed to have prepared at the
outset a plea on the legal points of the indictment, but when
the trial occurred to have assigned this part of the defence to
Ctesiphon. But for this change of plan no reason can be
found, and nowhere is there any allusion by Demosthenes to
a plea made by Ctesiphon.

(2) In § 238 Demosthenes repeats the allusion to the
Byzantians by Aeschines already referred to in § 95, which
favors rather the supposition of Schaefer, that Aeschines
omitted from his revised speech what he had said on this
point, out of regard for the Rhodians, than that of Kirchhoff,
who believes (apparently overlooking the second allusion in
§ 238) that Demosthenes misquoted Aeschines.

But on Kirchhoff’s supposition that Aeschines had not

On the Original Recension of the De Corona. 101

mentioned the Byzantians, the allusion in § 238 as well as in
§ 95 would be an instance either of erroneous anticipation or
of later insertion, and would go to prove that the parts of
the de Corona in which they occur are of contemporaneous
origin.

(8) The theory of Kirchhoff forces us into this dilemma:
either the unknown arranger and combiner of the two speeches
succeeded remarkably well in-accomplishing a task which the
orator himself abandoned as impracticable, or Demosthenes
did not aucceed, when he put his speech in writing, in giving
it that unity of structure and closeness of connection in all its
parts that should characterize a production carefully composed
out of material fully known and at hand. The only explana-
- tion of these instances of loose connection and abrupt insertion,
is to suppose that the body of the oration was composed before
the trial, and that the orator in a subsequent revision inserted
these originally extemporized passages as best he could.

It may be objected that too much stress has been laid in
this discussion on the supposed blemishes of structure in the
last division of the oration, and that the hope expressed by
Kirchhoff, that he has succeeded, at least, in shaking the
universal admiration with which critics have regarded the de
Corona, seems to be confirmed by our own demonstration.
The answer to this charge is three-fold: (1) In combating
the theory of Kirchhoff we have employed his own weapons,
and made the most of every point. The evidence brought to
sustain the theory might properly be assailed as an illustration
of a kind of criticism which ‘strains at a gnat and swallows
a camel.” (2) On our theory it is to be expected that the
interpolated passages should be easily separable from the
context, and should sometimes show the seam by which they
are, so to say, stitched into the whole. The skill with which
this is done in most instances is itself worthy of admiration.
(3) A comparison of the de Corona with other speeches of
Demosthenes argues the essential unity of origin of this
oration.

Take, for example, the oration of Demosthenes on the False
Embassy.

14

102 M. L. D’ Ooge,

In this oration critics have discovered grave faults of
arrangement and composition, and have proposed bold changes
and excisions. One of the latest and boldest of these critics
is O. Gilbert (die Rede des Demosthenes TEDL THC naparpeo,*iac,
Berlin, 1873). The latest defender of the present plan of
the speech is Blass, who seconds the views of Schaefer except
as regards the relation the written speech holds to the spoken.
From the discrepancies between the speech as we have it and
the allusions to it found in the reply of Aeschines, Blass
argues that the written speech was prepared some time before
the trial, and was not afterward revised so as to meet the
change of situation brought about by the condemnation and
flight of Philocrates; while Schaefer, on the other hand,
explains these discrepancies by supposing that Demosthenes,
in revising his speech, suppressed some passages and added
others.

However that may be, the point that concerns us more
particularly is that the most radical view of the origin and
character of the de Fals. Legat. goes no farther than to point
out displacements and later insertions in the text, and throws
no doubt upon the original oneness of origin of the oration.

It would be interesting in this connection to make a minute
comparison between the oration of Aeschines de Fals. Legat.
and the de Corona, particularly with reference to the direct
allusions which each orator makes to the words spoken by the
other. From such a comparison it would appear that Aeschines
pursued, in the subsequent revision of his speech, the same
course that we suppose in the revision of the de Corona. We
have space to notice only a few of the twenty direct allusions
made by Aeschines to the words of Demosthenes.

In § 56 Aeschines refers to the attack of Demosthenes upon
his speech before the assembly (rije dé imo... .. wept Tijc eiphync),
and proceeds to show that Demosthenes falsified, referring
particularly to the statement of Demosthenes, de Fals. Legat.,
§ 13-16. This statement covers two points: (1) That Aes-
chines spoke against the peace of Philocrates the first day, and
in favor of it the second day of the Assembly; (2) that he
thus changed his attitude in the presence of the envoys

On the Original Recension of the De Corona. 103

sent by the Greek states to act in concert with Athens. The
answer of Aeschines to (2) is contained in § 57-62; to (1),
in § 63-69. The sentence in § 69, éredy dé cai ry dnunyopiay
x.r.é, is plainly intended to form the connecting link between
the inserted passage and the body of the speech.

An examination of Aeschines’ de Fals. Legat., § 144-171,
warrants the belief that this entire passage is a later insertion.
The beginning of this passage marks what is, perhaps, the
most abrupt transition in the oration. Aeschines has been ex.
plaining how Phocis came to be ruined, “ first through fortune
which is the arbiter of all, then through the length of time
and the ten years’ war.”’ He calls for testimony from Boeo-
tians and Phocians to show how he had benefited their states,
and then asks wc ody oix x.7.é, § 143. Then without any
introduction, without any sentence to make the transition,
excepting the words éroApnoe S cizeiv we ey roig..... KEPLTINTW,
he passes to consider, in § 144-171, matters of a personal
nature in response to the attacks of his opponent. The close
of this passage is but little less abrupt than its beginning.
In § 170 Aeschines calls up testimony to his honorable record
as a soldier; § 171 forms the connecting link between these
personal considerations and a review of the history of Athens.

An analysis of the passage itself, including § 144-171,
shows the sutures of the different parts. In § 144-145
Aeschines tries to turn the point of the quotation of Demos-
thenes about gyn In § 146-152 he meets the charge of
betraying his country into the hands of Philip. In this
passage occurs the allusion to the attack of Demosthenes
on Philon, no reference to whom is found in the speech of
Demosthenes as handed down. § 153-159 is occupied, as
far as the sentence aAN’ ota «.7.é., with a reply to the con-
trast Demosthenes had drawn between the conduct of Aes-
chines and of Satyrus towards the Olynthian woman and the
prisoners of war. What follows in § 159-161 seems intended
to lead the hearer (reader) back to the main issue ; but pres-
ently this is forgotten, and in § 162-163 the charge is noticed
that he joined with Philip in singing paeans after the destruc.
tion of Phocis. The abruptness of the opening sentence in

104 M. L. D’ Ooge.

§ 162, its want of connection with what precedes, is apparent at
a glance. In § 164-166 Aeschines answers the charge that he
so suddenly changed his attitude towards the peace. Then
follows, in § 167-170, the defence of his military record,
already mentioned above.

The last part of this oration, from § 172 on, seems originally
to have been directly connected with the passage immediately
preceding § 144-171. In § 180-138 Aeschines aims to show
that through the agency of Demosthenes and his associates
Athens was prevented from joining with Philip in an honora-
ble attempt to put an end to the Phocian troubles. In
§ 172-178 he proceeds to show that it was in the line of
ancestral precedent to make peace as well as to carry on
war, and that peace had generally been productive pf more
good than war.

The sentence od rove AnuooSévove ipdc obk éwy mpoyovouc pypetaSac
x.t.é, in § 171, is plainly intended to recall the expression in
§ 138, rove mpoydvoug ixwAveu rov Sypoyv pupeioSar, and seems to
have been written for the very purpose of connecting these .
disjointed parts.

Enough has been shown of the structure of the de Fals.
Legat. of Aeschines to make it appear that in its present
form it is to some extent—just how far may not be exactly
determined—a reconstruction and recension of the one original
speech delivered at the trial, and a fit parallel of the de Corona
in its genesis and composition.

While we may never be able to show just how far this process
of recasting and revising an oration for publication extended,
we trust that this discussion has, at least, made apparent the
wide distance that separates an oration like the de Corona,
and the de Fals. Leyat. of Aeschines, from a production like
the Zimocratea, which, as was shown at the outset, is con-
fessedly not simply a revision of what was originally one
speech, but a combination of two or three different pleas,
characterized, in spite of the skill shown in the work of
combination, by contradictions, by literal repetitions, and by
unevenness of finish.

On the Authorship of the Dialogus de-Oratoribus. 105

VII.— The Authorship of the Dialogus de Oratoribus.

By TRACY PECK,

PROFESSOR OF LATIN IN CORNELL UNIVERSITY.

As far as our knowledge extends, it was Beatus Rhenanus
who, in his editions of Tacitus early in the sixteenth century,
started the discussion in regard to the authorship of the
Dialogus de Oratoribus. About fifty years later, Lipsius main-
tained with more force that Tacitus could not have written the
work, and his judgment—based mainly on the wide difference
between the style of the Dialogus and that of the admitted
writings of Tacitus—greatly influenced his contemporaries,
and is still not without adherents. But the broad question
whether the work is the composition of Tacitus, or of Quin-
tilian, or of the younger Pliny, or even of Suetonius or Maternus,
by thé consent of modern scholars has been narrowed to this
question,— whether Tacitus was or was not the author. The
arguments once advanced in favor of all others except Tacitus
are so easily overthrown by considerations of chronology,
literary style, judgments of men and things, and mental
‘characteristics, that, if I mistake not, these quasi-claimants
are now without champions. The current of criticism, so far
as it has been publicly expressed, is to-day setting strongly in
favor of Tacitus. It is hoped by this paper not so much to
add materially to this discussion—now nearly three hundred
and fifty years old—as to excite among us greater interest in a
work which singularly merits and rewards study as the outcome:
of and protest against the literary condition of Rome in the first
century of the Empire, and because of its fertile suggestiveness
of thought, its exquisite Latinty, and the freshness of its style.

The Dialogus is nowhere distinctly mentioned in extant
classical Latin, but in a letter of the younger Pliny (1x. 10) to
Tacitus it is in all probability quoted from. Pliny’s words are
these: Poémata quiescunt, quae tu inter nemora et lucos com-
modissime perfici putas. Now, the exact idea and the essential
words appear in chapters 9 and 12 of the Dialogus, nor are

106 T. Peck,

the words in such a use elsewhere found in Tacitus. This
contemporary evidence should seem conclusive: but it has been
urged that the quotation is from some work or part of a work
of Tacitus now lost. It is hard to see how the thought could
have found pertinent expression in any of his historical com-
positions. Of course the passage might have been in thé
vanished correspondence of Tacitus, but this explanation can
hardly appear probable to one who examines the question with
a judicial rather than a partisan frame of mind.

It seems like trifling to mention the other ground on which
the authority of this letter has been impugned. Heinrich
Gutmann maintains that the letter was written by Tacitus
himself in reply to Pliny’s (1.6) communication. But in the
earlier letter there is nothing to which this could serve as a
fitanswer. The style of the letter is through and through that
of Pliny himself. There is no evidence that Tacitus wrote
poetry, and had he thus written, Pliny would in all probability
have added his name to the list of versifiers in Ep. v. $.. And
by what strange chance should this single letter of Tacitus
have alone been preserved, while all other epistles from Pliny’s
many correspondents (save only those from the Emperor
Trajan) were excluded from the collection? If Pliny had
thought to honor Tacitus by the insertion of one of his letters,
would he have chosen this exceedingly light specimen ?

The testimony of the manuscripts is in favor of Tacitus
alone. All existing Mss. of the Dialogus have the important
lacuna between chapters 35 and 36, and other corrupt passages,
and are doubtless derived from the copy that was brought to
Rome in or very near the year 1457. In all Mss. the title runs,
Cornelit Taciti Dialogus de Oratoribus. Additions to this are
of late and known origin. Thus Gronovius in his edition
increases the codex-superscription with the words stv de causis
corruptae eloquentiae. This was the subject of an early work
of Quintilian (Inst. Or., v1. prooem. 3), but Spalding C1. c.)
has clearly proved that the Dialogus cannot be this treatise of
Quintilian. Furthermore, the Dialogus is not found with the
mss. of the Histories or Annals, so that it cannot thus by acci-
dent or blunder have been identified with the works of Tacitus.

On the Authorship of the Dialogus de Oratoribus. 107

Diplomatic tradition thus speaks only in favor of Tacitus, and
if we ignore this, the door is opened wide for scepticism in
regard to the authorship of nearly all ancient literature.

Of no little weight is the testimony of Pomponio Leto of
Calabria, who died not later than 1498. His words are:
C. Tacitus scripta Maecenatis appellat calamistros. The epi-
thet is found in the Dialogus, cap. 26, nor is it elsewhere in
extant Latin applied to this effeminate littérateur. Lipsius
endeavors to rebut this evidence for Tacitus with a contemptv-
- ous remark about the ignobleness of the witness. Of illegiti-
mate birth he certainly was: so, too, is he freighted in history
with many names; but nothing is known of him that justifies
us in supposing him guilty of falsehood or of easy credulity
in regard to our present question.

In the chronology of the Dialogus there is nothing incon-
sistent with the theory that Tacitus was its author. The work
professes (cap. 1) to be an accurate reproduction of a discus-
sion held in the sixth year (cap. 17) of Vespasian’s reign.
The reporter was at that time juvenis admodum,—a tantaliz-
ingly elastic expression. Tacitus applies (Agr. 7) the words
to Domitian when he was eighteen years old, and to Vespasian
(H. 11. 78) and Helvidius Priscus (H. 1v. 5) with much vague-
ness. The birth-year of Tacitus is not known, but it was
probably 54 a. D., so that in 74 or 7d a. D. he could properly
have been spoken of as juvenis admodum. The time of
composition of the Dialogus can only be conjectured. The
objective way in which the writer speaks of himself in the
first two chapters, implies, to my mind, a considerable interval
between the discussion and its report. Nor does it seem
natural that Justus Fabius, to whom the work is addressed,
should have ‘frequently’ (saepe ex me requiris, cap. 1)
requested a guvenis admodum to undertake the discussion of
so weighty a theme. But the fact of a youthful writer seems
unmistakable from the entire warp and woof of the piece
itself, and if Quintilian (Inst. Or., x. 8, 22), as is altogether
probable, alludes to the work, it must have been given to the
public some time before the death of Domitian in 96. The
Dialogus may thus have been published late in Vespasian’s

108 T. Peck,

reign, or under Titus. Nor do I see anything fatal to the
view that it came out in the first years of Domitian’s rule.
Tacitus in the Agricola (3) certainly says that the fifteen
years of that monster’s reign were passed in silence; but he
is apparently characterizing the entire reign from its last terri-
ble years, and is referring to the capital danger of faithfully
writing history or the lives of outspoken, freedom-loving
individuals, rather than to the composition of such inoffensive
pieces as the Dialogus. Under Domitian Quintilian certainly
wrote, as did Statius, Silius Italicus, Martial, and many others
whose names are better known than their writings. I do not
think that the style or temperament of the piece will allow us
to place its composition after Domitian’s death. Too late,
also, appears the mysterious absence of Tacitus from Rome
for four years (Agr. 45) just before Agricola’s death in 93.
So late in Domitian’s jealous and vindictive period Tacitus
would hardly have ventured to speak in such praise of Ves-
pasian (capp. 8,17), nor could he then have spoken with such
buoyant cheerfulness of the imperial régime (41). Even the
work of the calm and impartial Quintilian (I. O., x. 1, 91, 92)
is marred by the servile flattery of the times, but we need not
suspect that the grave Tacitus—and that, too, gratuitously—
thus prostituted his powers.

Leaving these external considerations, we are now prepared
to question the personal motive, the sentiments, and the literary
style of the Dialogus. Reading between the lines, I think we
may find that Tacitus intended the work as a kind of pro
domo sua, @ vindication of his withdrawal from the career of
a forensic orator and devotion to literature. As Cicero in his
de Oratore speaks his own views through Crassus, so here
Tacitus probably puts forth Maternus as his representative.
Maternus is certainly the leading character: the discussion
takes place at his house; he is introduced as a well-known
person ; his tragedies give rise to the debate; he directs and
closes the argument, and twice recalls the disputants from
digressions. In early life Maternus had gained prominence
as a lawyer, but was now absorbed in the writing of plays;
Tacitus, too, was educated as a lawyer and entered upon the

On the Authorship of the Dialogus de Oratoribus. 109

practice of the law, but abandoned this career for that of a
historian. That the writer of the Dialogus had had a legal
training like that of Tacitus is beyond a doubt: the discussion
is introduced as an actual trial, in which we have plaintiff,
defendant, advocate, and judge; the closing speech of Mater-
nus bears the character of a judicial verdict ; in the body of
the work are many legal terms, as cut bono? formula, inter-
dictum, and frequent reference is made to speeches that deal
largely with legal technicalities and processes.

The Dialogus and the acknowledged writings of Tacitus
show a quite remarkable consistency of views of men and
events, and a like conception of life. In all we see a regret
for the grander life and opportunities of the republic, but at
the same time a philosophic acquiescence in the empire as the
inevitable, and thus a strong tendency to fatalism. Here as
there we find a cordial recognition of woman’s influence in
the prudent training of her children; a rich vein of satire—
the genuine Roman satire—worked against the present, but
with an earnest desire for its improvement; a disposition to
concede greatness in men of the times, provided they are not
measured by the superior standards of the past; a keen insight
into human nature and motives, and a portrayal of character
and action with real dramatic effect; pregnant, epigrammatic
utterances—not accumulated and paraded, as in Seneca, as
sheer rhetorical tinsel, but naturally suggested by the course
of thought. Here as there we see that Cicero—particularly
his rhetorical writings—and Vergil had been sympathetically
studied, and have left their stamp upon thought and expression.

In the youth of Tacitus there were two antagonistic schools
of style, whose chiefs were Quintilian and Seneca. There
seems no doubt that Tacitus was first drawn to the former—
a modified type of Ciceronianism. The younger Pliny (Ep.
vit. 20) clearly implies this, and the Dialogus is the best extant
specimen of this literary renaissance. The prevailing fulness
- of expression, the rhythmical periodic structure, the thoughts
brocaded with an almost gaudy rhetoric, and at the same time
frequent instances of tautology and an almost cloying richness
of language, show both the author’s success and failure in this

15

110 T. Peck.

tour de foree. But Tacitus was not one to wear long the garb
of any school or model, and from the Dialogus on through the
Agricola, and the Germany, and the Histories, and the Annals,
we can trace the steady development of that marvelous method
of expression which finally puts Tacitus apart from all who
have ever written. The difference between the styles of the
extremes in this series is certainly very great. But style
naturally varies with the subject, changes with the times,
and is modified by maturer years and wider experience. The
historical studies of Tacitus, and his observation of the whim-
sical and brutal use of power, and of the apparently inevitable
decline of his country, had profoundly affected his whole soul,
and his style could not have remained stationary. Tacitus
had not always been the austere, almost gloomy man that
might be inferred from his latest writings: from the letters of
Pliny he appears to have been a very genial and companionable
friend ; there is even a tradition that he once wrote a book of
witticisms (liber facetiarum),and certainly the Dialogus shows
in its author a cheery and hopeful disposition.

The Latinity of the Dialogus, though based upon a close
study of Cicero, reveals in all its texture the traits of the
“ Silver Age,”’ and has many of the peculiarities of the later
writings of Tacitus. Among these peculiarities may be men-
tioned a fondness for using pairs of words of like meaning—
especially nouns and adjectives; a decided poetic coloring,
as in the free use of metaphor and of personification; the
reversal of the traditional order of the nomen gentile and
cognomen ; and the reckoning of the reign of Augustus from
his first consulship, instead of from the battle of Actium or
from the time of receiving the title of Augustus.

On the Date of the Prometheus of Aeschylus. — iil

VITI.—On the Date of the Prometheus of Aeschylus.

_ By T. D. SEYMOUR,

PROFESSOR OF GREEK IN WESTERN RESERVE COLLEGE.

All efforts to determine the date of the representation of
the Prometheus Bound have failed to discover convincing
arguments. Concerning the date of no other Greek drama
do scholars differ so widely. While Schoemann and others
hold this to be the earliest of the extant plays, Moriz Schmidt
classes it with the Rhesus as the rear-guard, the stragglers—
der Nachtrab—of Greek tragedy. To Westphal and Dindorf
the character of the metres seems to point to a late epoch in
tragic art; to Wecklein the careful formation of the trimeters
seems to indicate that the play was written before the Per-
sians. Gottfried Hermann believes that only two actors were
employed, and assigns the play to an early date; Karl Otfried
Miller thinks that three actors were necessary, and that the
play is one of the poet’s latest works, produced after Supho-
cles had introduced the third actor. Donaldson selects for
the date of representation ‘the year 464 B.c., when the news
would reach Athens that ‘Themistocles had entered the service
of the Persian king. The warrior of Marathon and Salamis,
and the friend of Aristides, would at such a time with peculiar
force utter that abomination of treason, which the poet puts
into the mouth of his chorus.””. Who, unacquainted with the
play, would suppose that this “abomination of treason”
referred to lines 1067 fg. :

pera rovo G re xpi) wacyery Esedw*
Toug mpodurac yap pugeiv Euadoy,
KOUK EaTt YOoog

Tijd ijvrev’ anéxrvoa paddov !

It is indeed childish, as Haupt says, to seek or accept in the
plays of Aeschylus or Sophocles allusions to the politics of their -
time. This notion of Donaldson is shown in its full absurdity
by the equally improbable hypothesis of a later English writer,

112 T. D. Seymour,

that this play was intended, on the other hand, as a glorifica-
tion of Themistocles. In that event it would be best dated
before the Persians.

All the above arguments are based upon the metre, the
language, and the construction of the play, or upon some
uncertain political allusion. From the very nature of the
arguments and the known facts of the poet’s life they are
inconclusive. According to the Marmor Parium, Aeschylus
was born Ol. Lx 4, 525 B.c., and contended with Pratinas
Ol. Lxx, 500 B.c. With the latter date agrees the statement
at the beginning of the Biog AisyiAov—véoe dé piaro ray rpaypduwy.
The Persians, the first tragedy of which the date is fixed, was
represented in the archonship of Meno, 472 B.c.; the Orestean
Trilogy, the last presented by Aeschylus at Athens, was put
upon the stage in the archonship of Philocles, 458B.c. These
fourteen years were by no means the Lehrjahre of Aeschylus.
At the beginning of this period he had been writing tragedies
for twenty-eight years, and was probably then fifty-three years
old. While he was ready to accept suggestions from his
younger rival, Sophocles, and modify his stage arrangements,
we have no reason to suppose that he materially changed his
style of composition or metres. The history of the Athenian
stage of that period is by no means clear. We do not know
even whether the third actor was first introduced by Aeschylus
or Sophocles. Much less are we informed as to the date of
other innovations. We have too few of our author’s works
to justify us in the assertion, from internal evidence, that one
play belongs to his fifty-fifth year and another to his sixty-
fifth year. We do not know what is accidental in the play
and what belongs to the period of development and work.
The simplicity of language and construction of the Prome-
theus may be explained as natural to the earliest age of
tragedy, or as the result of Sophoclean influence.

A comparison with the Suppliants has been suggested as
affording an indication of the date. In both plays the
fortunes of Io are referred to in similar Janguage, and the
return of the Danaids to Greece is alluded to in the Prome-
theus. We may compare Suppliants 29: roy 3ndvyevij orédor,

On the Date of the Prometheus of Aeschylus. 113

38: Aékrpwy dv Séucc eipye, B12: cat Zeve y' épaxrwp yerpi gerdec
yovov, 46: Zyvag ipafir’ exwrupia 8 éwexpaivero pdporog aiwy |
evrAdywe, Exagoy 6° tyévvacev, 814: “Exagoc dAnSe¢ puciwy éexwvupoc,
179: aive gurdada rap’ Exn Sedroupévac, 228 : Eopog we wederaewy |
ileoSe xipkwy Trwv dporrépwy poy | éxSpwr opainwy cal peacrovrwy
yévoc, 208: yévoc TeXaryev rivde xzprovrar yGora, 804: wavdrrny
oioBouxdArov, . . puwra KivATHpLOY, . oloTpor,. . paxpp Spdopy, 311:
KavwBov xaxi Mépugey ixero, 320: xevrnxovrarac, 467: wpparwoa,
540 fg.: "Iw | otorpy Epeoaopéva | peiryer Gpuprivooc, | rohda PBporwv
dcaperBopeva|gvra «rr. with Prometheus 846:

torey wOdALC Kavwio¢ Eoyarn x Sovec,
évravda on oe Zeve ridnow Eudpova
Ld ~ 9 ~ X ‘\ N ¢
[ erapey arapBei XEtpt Kat Styor povor]}.
éxwvupov O& Tar Avog yerynp ddov
régecc KeAct vor” Exagor, o¢ Kapwwoerat
donv xXaruppouc NeiAog apdever yIdva"
néuntn 0 ax’ atrov yévva nevrnxovrarac
Ld ‘ w 9 € ~ & /
madwv tpog Apyog ovy Exovo’ éEhevoerat
Sndvoropoc, pevyovea ovyyerh yapov
aveyor® of 0 éxronpévor dpévac,
Kipxoe weAEL@y ov paxpav AEdEtppErot,
itouve Snpevovrec ob Inpacipove | yapoug «rd.

Also 789: jj éyypagou ov pripoow Sédracc pperav, O69: pupmwror
eicopwoa Povravy, 61D: piwme ypicSeio Eupavet oxipripar, O80:
otorpnArary de deipart, 591: rove iweppijxerc Spduouc, 499: eLwuparwoa,
709 fg. and 792 fg., a list of the many tribes through which
Io must pass before she finds rest.

Aeschylus shrank no more than Homer from the repetition
of a thought in similar words, and might be as willing to give
the substance of his Suppliants in a later tragedy as to amplify
into a new tragedy the sketch given in the Prometheus. It
is not easy, then, to draw from this comparison any definite
conclusion as to priority of composition. Moreover the exact
date of the Suppliants is unknown.

A surer criterion we may perhaps find in a comparison with
the works of Pindar, of which Aeschylus was a careful student,
as he was a student of the other lyric poets and of Homer,
and did not hesitate to borrow from Phrynichus. He trans-

114 T. D. Seymour,

ferred é¢ ro wadov éx rov xadov. His originality was so great
and undisputed that he could afford to borrow. We think no
less of Milton’s genius when other nations claim the original
plan of the Paradise Lost, and when we find in Spenser the
germs of some famous passages. The great Teutonic poet is
in our time not accused of lack of originality, though Greene
claimed that Shakespeare was decking himself in borrowed
plumes. So Aeschylus’ genius is manifested in all the changes
which he made in the plan of the Phoenissac. He gave the
prologue of the Persians (if it may be called a prologue, in
the parodos) to the chorus of old peers, and not to the eunuch
dusting the throne. Wisdom was shown also in the constitu-
tion of the chorus from Persian peers, and not from Tyrian
women brought to Susa without sufficient motive. It is an
evidence of our different estimate of Euripides that we are
not quite sure how much of the Medea’s beauty is due to
Neophron.

As then the Persians, perhaps the earliest tragedy which
has come down to us, bears witness to the willingness of
Aeschylus to study and profit by the works of other poets,
and as he himself said that his dramas were but scraps from
the sumptuous banquets of Homer, we need not be prejudiced
against the thought that he not only studied but borrowed
from his great contemporary.

The resemblance between the most tragic of lyric poets and
the most lyric of tragic poets has been often noted. Born in
the same lustrum, educated in the schools of the same city,
though widely separated by their different relations to the
wars for the freedom of Greece, and differing in some minor
points, as in their treatment of the strife of the Gods, they
were alike in their ethical views, in their sublime, rugged,
and sometimes grandiose style, and, as we should expect,
were alike in their use of words and constructions. Thus
a comparison with Pindar has not only restored axAdrov for
the axAyjorov of the mss. in Prometheus 371, but has explained
many other passages.

Both poets use the present tense in predictions. So Prom.
H5LY: dvacg re cappreic See deopa quyydarw, SAB: Evravda dy oe Zeve

On the Date of the Prometheus of Aéschylus. 115

riSnov Exdpova Agam. 126: xpdry pev aypet Mprapou modu ade
xthevdoc, Pind. Ol. vil, 42: Mépyapoc api reaic, fipwe, yepoc
épyaciag Grioxerea, Pyth. 1V, 48: rore yap peyddag | eLaviorarrac
Aaxedaipovoc.

The emphatic position of the proper name at the end of the
sentence is not indeed peculiar to Aeschylus among the tragic
poets, but withe. g. Prom. 612: rupoc Aporoig Sorijp’ dude MpopnSia,
we may compare Ol. vn, 13: ray zorriay | tpvéwy raid ’Adpodirac,
"AeAloid re vuppar, ‘Pdcor, Pyth. VI, 80: érapiuBporoy | avapeivacc
orpurapyoyv Aiddxwy | Mézvora, Isth. 1V, 53: OnBav aro Katyeiav
Hoppav Bpayuc, Puyay & axapxroc, tpooxradaiowy HAY’ avi | . . vide
"AAKUHVac. | °

Common to both poets is the use of a noun as an adjective.
Agam. 403: domioropag «Advovc, Isth. 1, 28: érXiratc dpopae.
Both have the figurative use of woujy. Supp. 167: vady xor-
pévec, Ol. xX, 88: exei rrovroc 6 Aaywv worpéva | Exaxrov adrdérptor |
Svdexorre arvyepwraroc, Nem. VIII, 6: womévec .- | Kuxpiac dépur.
So with zpirang, Prom. 169: paxapwr xpiranc. Pyth. vi, 24:
Bapvdray areporay Kkepavyoy re xpvrany. Both poets use aica
frequently and in various constructions. Ol. 1X, 42: Ace aieg,
Pyth. tv, 107: «ar aloay, Pyth. vitt, 18: zap’ aioay, Frag. 1, 2:
ouv Seay aiog, Supp. 545: éy aiog, T9: wap’ aiaay, Choeph. 927 :
marpoc aiga. In Pyth. x, 66: giréwy giréorr’, Gywr ayorra, the
opotoxarapxra remind of Prom. 19: dxovra o° &kwv, 29: Sede Sear,
192: orevdwy oxevdorrt, 218: éExdvS’ Exdyvrt, 276: apoc &AXor’ Gddor,
671: dxovoay dcwr.

Both fancy the ‘appositional use of xo») and similar words.
Supp. 626: ebyac ayadac, ayadar mavac. Isth. m1, 7: ebxAéwy
& Epywy drava xp) per ipvijoa rov éodov xrh.

Constructions which were later less frequent were not
uncommon in these two pocts. Such is repi with a dative of
cause. Persians 695: o¢Bopat & avria rébar | céIev apyaiy repi
répBe, Pyth. v, 58: Agovrec wepi ceipare puyor.

We may further compare Pers. 1056: yeveiov répSe Aevejpy
rpixa and Pyth. 1x, 80: EtpuoSioc éwel xepadray | Expade gaoyarov
acpg. Agam. 113: oiwrév Bacrede and Ol. xm, 21: oiwvey
Baoréa. Supp. 998: répecr’ crwpa (of the maidens) & evpidracrog
ovdapec is illustrated by Isth. 1, 4: “Agpodirac | ebSpdvou prdo-
reipay ddioray oxnwpar, Prom. 150: deapoic advrare dypiwe medaoag,

116 TD. Seymour,

by Pyth. Iv, 227: rove ayaydv CevyAq wédaooey. Examples of
this similarity might be indefinitely multiplied.

But not infrequently we find coincidence of thought or
language, which seems to be the result of conscious imitation,
and with perhaps only a single exception, we find from the
known dates of ode and tragedy that it is Aeschylus who
borrows. We may perhaps regard as accidental the similar
allusions to the punishment of Aesculapius, Ag. 1022, Pyth.
11, 55; and to the crime of Ixion, Eum. 738 (apwrocrérow
mpostporaic ‘Itiovec), Pyth. 1, 382 (éu@bdcoy atua xpwresrog ob« arep
réxvac tréuite Svaroic) ; and the same figure of a cock fighting
on his own duhghill, Agam. 1671, Eum. 861 (but in a sus-
pected passage), Ol. x11, 14. In other cases, however, the
connection is more evident.

Pindar, Pyth. vii, 95: éwdpepor ri dé rics ri 8 ob reg3 oma
évap | avSpwroc was in the mind of Aeschylus, we may well
believe, when he wrote Agam. 839: dpAdtac caroxrpoy, etdwdor
oxac, unless we accept the possibility that both are later than
Aesch. fr. 390: ré yap Bpdreov oxépp’ ep’ tptpav ppovel, | kai xeorov
obdey paddoy i karvov ond. The resemblance to Pindar is per-
haps as close, though not so obvious, in Prometheus 545 fg.:

gép Orwe dyaptc xapic, @ pidroc’ ele wou rig GAKa 5

tic épapepluy dpniic; obd eépy snc

O\Lyoopaviay aKiKkuy

iadvetpou, ¢ TO purer

adadyv (déderac ins. Hermann) yévoc éurerodiopévor;

and 448: dveparwy | adiyxo poppaio. That chorus of the
Prometheus, 545 fg., it may be remarked in passing, has
been noticed by Heinrich Schmidt for its Pindaric form; and
in the first strophe we find the dactylo-epitritic measure which
prevails in Pindar, but which is found in Aeschylus only
there and in the same play 887 fg., and perhaps once in the
Suppliants.

But among the more specific points of resemblance between
odes of Pindar and the Prometheus of Aeschylus, the most
remarkable is the description of the hundred-headed monster,
Typhon, Pyth. 1, 15 fg.:

On the Date of the Prometheus of Aeschylus. 117

dcr’ év aivg Taprdpy xetrat, Sewy wodémog,

Tupwe éxarovraxapavocg’ rév wore

Kidixtov Speer roAvwvupov ayrpor* viv ye pay ..

Lexeria 7 abrov mele ortpva Aayvaevra® Kiwy 5’ otpavia ouvéyet,
vupdeco’ Airva .. Tag épevyorrat pey axdarov rupoc ayvérarat,

Ex pvyoy Payai «rd.

The Prosodion to Aetnaean Zeus, of which we have fragments
92, 93, seems to have been written about the time of the first
Pythian ode, when Hiero honored in every way the city which
he founded on the site of Catana: Keivw peév Airva decpoc imep.
piadog | apoixerrat.

. « GAN’ olog GxAarov Kepailec Seay

Tupi’ Exarovraxdparoy avayxg, Zev warep,

év ’Apipotc roré.

The monster is mentioned alse in Pyth. vi, 15: Bia dé cai
peyadavyoy Eogader év xpdvy.| Tupac Kites Exardyxpavoc ov vey Gdviev.
Here we see that Typhon was not yet Aetnaean, and Strabo
intimates that Pindar was the first to transfer him to Sicily.*

With these passages we have to compare Prometheus
301 fg.:

toy yqysvi, te Kelixlay olxttopa

dytgwy iddy wxtecga, ddior tégas

ixatoyxkgavor meds Bla» yeipotusvoy

Tugava Fodpor, maar O¢ dvéoty Fsotg xth.
364: xeitas OTsywnov niyoloy Padacalou

inotuavos gltasow Altyalass dno,

. . Svtay dxgayijoovtal note

motauol mupds ddntorvtes dyolass yradors

Tis xallixdgnou Sexellas devgods ybas . -
371: Pequois anidrov Bélsos nugnvdou Cadys.

That the resemblance here is beyond the workings of chance
is evident; but which is the original? Perhaps the answer
to this question might be left to the scholarly instinct of each

*In Hesiod, Theogony 860, for a:dvyc, the reading of the mss., Schoe-
mann conjectured 'Airvy7s. This has been received into the text by Flach.
But, besides the uncertainty of the emendation, the passage in the Theogony
is quite distinct in style from the rest of that work, and may be of later date
than Pindar and Aeschylus.

16

118 T. D. Seymour,

reader. Schoemann says that Aeschylus’ description of the
eruption might have been written if the poet had never seen
Sicily. It would hardly have had its present form if he had
never seen or heard the first Pythian ode of Pindar. We may
be supported in our decision by an examination of the ode.
Hiero of Syracuse had gained the victory in the chariot-race
at Delphi. Six years or possibly only two years before, he had
driven out the inhabitants of Catana and had founded there
a Doric city, named from the mountain on whose slope it lay.
He was therefore to be celebrated as Aetnaean, and the glory
of the city Aetna is the real subject of the ode. The poet
begins with an address to the lyre: Xguoéa qégucys, ’Andddwro;
xa ionhoxéuwv obvdixoy Moody xtéavov. ‘©The minstrels obey
thy biddings. Thou dost quench the thunderbolt of ever-
living fire. The eagle, dgyic otwvav, sleeps on the sceptre of
Zeus. Thy notes soothe the hearts of the divinities. But all
the creatures which Zeus does not love are frightened at the
voice of the Pierides.’’ tocu 68 ui, aegllnxe Zeds, atitortas Bode |
Iegiiwy diovta, yav te xal ndvtoy xat’ dpuoiudxetor, | 0s 1 dy alg
Tagtépw xeitas xtd. Krom this creature who 4irvas éy uelaugilios;
dédstas xogvpais, the transition is natural and easy to the moun-
tain and the city named from it, which, with the founder’s
victory, is to be celebrated in the ode.

It is obvious that Typhon is an essential link in Pindar’s
chain. He is not introduced as a mere ornament or illustra-
tion, or as Philoctetes, 52 fg., to exalt the Syracusan king.
Let us now look at the parallel passage in the tragedy. After
speaking of Atlas, Prometheus continues: 1» yyyevq te Kedsxlow
olxijtoga xti. At the conclusion of this description, Prometheus
abruptly changes the subject by addressing Oceanus:

ad 5 ovx dnegos od euov diduaxdhov | ronters.

I am so far from agreeing with Wecklein that the mention of
Atlas is only a transition to the description of Typhon and
the eruption of Aetna, that the mention of Atlas seems to be
complete in itself, while Typhon is only introduced to give a
local allusion which would be appreciated best at the court of
Syracuse. For an Athenian audience Typhon did not stand
on the same footing as the Titan Atlas, and the Athenians

On the Date of the Prometheus of Aeschylus. 119

felt no particular interest in the eruption of Aetna. It is to
be noted, moreover, that in the epode of the following chorus,
Typhon is ignored and Atlas alone referred to in udvor dy
noda tev cdhov dv ndvots | .. soWduay xd. This could hardly be,
if it were true that the Titan was mentioned only as a transi-
tion to the monster. Considering then, the necessity of the
myth to this ode of Pindar’s, and the loose and episodical way
in which it is introduced in the tragedy, I think we may find
some confirmation of our opinion that the Prometheus was
written after the ode, which latter is dated by Boeckh 474 B.c.,
but by Bergk assigned to 470 B. c.

In the second Pythian ode, which was composed for Hiero,
probably 477 or 476 B.C., we find:

93. pégesr 0 thagews énauyévtoy duBdvta Cuydy

dgiyes: ott xévtgor 0é toe

haxtiadausy tehéder

dleatnods oluos.
We find the same thought in Prometheus 322: odxouy tuorye
yowusvos didaoxddy| ngds xévtga xGlov exteveis, and Agamemnon
1624: gig xévtga we Adxrte. Euripides uses the expression
twice; once in Bacchae 795: agig xévtga daxritouus Frvytds dy Few,
and again Frag. 607 : 9d xévtga fi; Aéxtete toig xgutotel gov. By
the time of Euripides it was becoming proverbial, as it was a
mere maxim in the time of St. Paul. In this connection it
is to be remembered that in Acts 1x, 5 the words oxAdngd» oor
amg0g xévtga Aaxtiteey are not found in the Greek ss., but are
inserted in the teztus receptus from Acts xxv1, 14. In the
latter passage St. Paul is giving to Agrippa an account in the
Greek language of the vision and the voice whith addressed
him “in the Hebrew tongue.”’ The Apostle, then, undoubtedly
gave a free translation, in a current Greek proverb, of the
words which he had heard. Commentators have been, perhaps,
too hasty in assuming that this was already such a common
expression in the time of Pindar and Aeschylus. In this
they have neglected the gégerr & ghaqgas xri. It would not be
in the manner of Vindar to develop the figure as he has,
if it were not his own fresh expression.

In the same second Pythian ode, & # douate necosydleva

120 T. D. Seymour,

_ xatalevyrin | oFévos Innov, verse 11, reminds of Prometheus
465: dq doua t Fyayor pedyvlovs | &nnous.

In the same ode, verse 34, in speaking of Ixion’s passion
for Hera, the poet says: zi; 8 xut’ adrdy alel navtds dgay jétgoy.
The same thought is expressed in Prometheus 890: dg 1
xndevoas xa sautdy aproteder Uaxea.

With a hyporchema written for Hiero at the same time as
this second Pythian ode, Frag. 105: Wouddsoor yag gv Sxb Fasc
Glaus Sredtav, | 8s duaSopdgntoy ofxoy ov zixatas, may be compared
Prometheus 709-710:

Lxidas J dgl&es vouadas ot nhextas atéyas
nedugoros valovo én’ edxixdhoss dyoss.

This last: seems like a dramatic development of duaSogég;tor
oixov.

It is instructive, also, to compare the prophecy of Themis
concerning Thetis, as given in the two poets. Isth. vim, 34 fg.:

eins 3 stBovloc (cf. Prom. 18: dgtofotlov Géusdoc) ev usooror Oducs,
Elvexer nenguusvor hy, pégtegdy xe ydvov &vaxta matgds tExEiv
novtiay Sadv, O¢ xEegauvod te xgéacor Kido Bélos

dungee yegt tgeddovtds t duasuaxétov, Ji dauatlouévav

4 dds nag’ ddcl~soiory.

This evidently is the prophecy to which reference is made in
the Prometheus 908:
tovov sagristae
f2uov yausiv, S¢ adtdy éx tvgavvidoc
Sodvwv t Giotoy &xBadsi xth.
920: Ttoioy nalasotiy viv nagacxevuterae
én’ attds adtg, dvopayutatoy tépas-
Oc 0} xegauvod xgelocor sigices pidya,
Boortis F sneghddovta xagregdy xréinoy-
Salacolay te yig tevdxterpay ydcor
Tolasvay, alyuny th» ToceWivoc, oxedg.

The Titaness Themis is referred to in the tragedy once and
again as the source of her son’s knowledge of futurity, and
we notice that Aeschylus adds two lines which did not affect
the danger of Zeus. The poet had in mind Pindar’s # Atdr
wap dodedgeciocy.

On the Date of the Prometheus of Aeschylus. 121

It is noteworthy that all but the last of these Pindaric
passages to which parallels have been found in the Prometheus
are from the odes and the hyporchema which were composed
for Hiero of Syracuse. For this there is no easier explanation
than that the tragedy also was prepared for the court of
Syracuse. No other reason appears why the Prometheus
should be fuller of Pindaric expressions and allusions than
the Agamemnon. This is. especially true of the Typhon
episode. This monster is mentioned again in the Seven
against Thebes, but the name is not found in any other Attic
writer, except once in the Phaedrus of Plato, Aristophanes’
Clouds 386, and, according to Hesychius, in a lost play of
Sophocles. Neither in these passages just mentioned, nor in
Herodotus, is he brought into connection with Sicily. As
the description of the Aetnean eruption serves a dramatic
purpose, to show the prophetic power of Prometheus, the
prediction concerning the Amazons, verse 723 fg., has been
compared with it. This prediction is very brief, however, and
the more natural as the Amazons were far more interesting
than Typhon to Aeschylus as an Athenian.

That Pindar should have seized upon the story of Typhon
is natural. Sicily was colonized after the beginning of the
historical period. The island is not rich in myths, and
strangely enough, in all his Sicilian odes Pindar never men-
tions the Cyclops. Moreover the Syracusan tyrant was not
descended from the ancient heroes. When an Aeginetan
received the victor’s crown the poet had an embarras de
richesse in the justice and honor of Aeacus, in the exploits
of Telamon and marriage of Peleus, and the brave deeds of
Ajax and Hector. When Diagoras of Rhodes was victor we
have the story of the first appearance of the island, and the
betrothal of the island’s nymph to the sun. For Thebes cf.
Frag. 29:

‘Iounvor } yovoaldxatoy Mella»,

} Kadpoy, } onagtay tegdy yévocg davdgay,
tay xvavaunvxa OBay,

}} 10 navtohpor otébvoc ‘Heaxiéoc,

4 tay duvicou noluyatéia touay,

} yduor Asuxwhdvou ‘Aquovias Surijoous» ;

122 T. D. Seymour,

This abundance of mythical subjects was not to be found in
Sicily, and thus Aeschylus’ conscious imitation is more natu-
ral and obvious. In the prediction concerning the eruption
of Aetna, in the mention of the “smooth fields of fertile
Sicily,” and of the monster, where, as Wecklein says, the
poet speaks rather than Prometheus; in the warning not’to
kick against the pricks; in the exhortation to marry in one’s
own rank; in the Scythian wheeled houses—in all this, Hiero
heard allusions to his victories, and to the Epinikia of Pindar
in his honor—allusions which were well understood by his court
and the Greeks generally. We cannot suppose that these allu-
sions were introduced for an audience at Gela or at Athens.
The abundance of nautical metaphors in the Suppliants

indicates to Teuffel the possible Sicilian origin of that play.
There is no lack of such figures in the Prometheus ; e. g. 149:
olaxovéuot, 182: dédia yuo duql caic téyats | 1a note tHrde advew | yor
a6 téigua xéhoavt éoWsiv, 190: tiv 0 atégauroy otogécas dgyi», 515:
olaxoorpéqos, 1001: dyhets udtyy us xim’ Snore nagyyopav, 1015: olds
oe yemdy xal xaxdy torxuula | Execo’ dpuxtos. But in neither,
perhaps, do we find more sea words and phrases than in the
Seven against Thebes, which begins:

Kadpov xolitas, ygi Aéyesy td xaigee

Oates puldooes neayos éy ngduwvn mbdEws

olaxa ywumy,
and ends,

ueta yop udxagac xal Jid¢ laydy

ids Kaduetow gvie addev

udvatoanivas yd dddodanwy

ximate gutay xataxhvo diy.

From this then we draw no conclusion. Aeschylus has merely
proved himself what Dionysus was not, a true Zadapimoc. I
would lay no stress either upon the Sicilianism of dppoi of
verse 615 of the Prometheus.

I would rest the argument upon the comparison with Pindar.
This alone makes it probable that the tragedy was composed
in Sicily, and after the first Pythian ode and the hyporchema
which Pindar wrote for Hiero. This ode is in honor of the
chariot victory which was gained at the 29th Pythiad. Accord-

On the Date of the Prometheus of Aeschylus. 128

ing to Bergk and the earlier authorities, who reckon from
the first cregavirnc ayoy in Ol. XLix 3, this would be in the
early autumn of 470 B.c. Boeckh, however, reckons from the
establishment of the ayw#» yxpnparirnc, and his date for the ode,
474 B. c., is adopted by Dissen, Schmidt, and others. This
fixes the latter part of 474 or 470 B.c. as the earliest date
possible for the tragedy. But in the Spring of 472 Aeschylus
produced the Persians at Athens. It has been suggested that
it was the fame of that play which gained for Aeschylus an
invitation to the court of Syracuse. The Persians would
naturally arouse the tyrant’s interest. He loved to hear the
conflict with the Carthaginians, in which he had a share,
compared with the battle of Salamis. As the conflicts were
near in time, they were alike in results. It does not surprise
us, then, to find Himera and Salamis united in that first
Pythian ode. Aeschylus reproduced (gaciv avadiddbac rouc
Tlépoag tv Xexedig) his Persians on the stage of Syracuse. We
may well believe that it was during this visit that he com-
posed the Aetnean Women and the Prometheus Bound.
The satyric drama of the Persian trilogy was, as is well
known, the Prometheus Mup¢dpoc or Tlupxacic. That there
was a trilogy devoted to Prometheus has been assumed
since Welcker, and is perhaps probable. In that case it
is natural to suppose that the satyric drama which accompa-
nied the Persians was written before the Promethean trilogy,
and in the IIvpxacig the poet saw how the subject could be
treated in a trilogy. It is not easy to believe that a Pro-
methean trilogy was produced in 473 B.C., with a satyric
drama on another theme, while the Ivpxaeve was put upon the
stage a year later. Moreover there is no positive evidence
that Aeschylus was in Sicily between 475 and 472 B.c.

The duration of the visit of Aeschylus at Syracuse after the
spring of 472, is uncertain. According to Plutarch he was
again at Athens in 468, when Sophocles gained his first
dramatic victory. The Seven against Thebes was presented
in 467. In this latter year Hiero died and his court was in
confusion.

If, then, we accept the Pythiad reckoning of Bergk and the

124 TD. Seymour.

older authorities, there is reason to believe that the Prome-
theus Bound was written at Syracuse between the autumn of
470 and the early spring of 468 8.c. If we follow the Pythiad
reckoning of Boeckh and Dissen, we may believe that it was
written between 472 and 468 B. c.

AMERICAN PHILOLoGIcAL ASSOCIATION.

1878-9.

MEMBERS IN ATTENDANCE AT THE ELEVENTH ANNUAL SESSION.
(From the autograph register. )

Stephen Pearl Andrews, New York, N. Y.

William Hyde Appleton, Swarthmore College, Swarthmore, Pa.
Albert N. Arnold, Pawtuxet, R. I.

Charles J. Buckingham, Poughkeepsie, N. Y.

L. H. Buckingham, English High School, Boston, Mass.

W. C. Cattell, Lafayette College, Easton, Pa.

Albert 8. Cook, Johns Hopkins University, Baltimore, Md.
Edward P. Crowell, Amherst College, Amherst, Mass.

Martin L. D’Ooge, University of Michigan, Ann Arbor, Mich.
T. T. Eaton, Petersburg, Va.

James M. Garnett, St. John’s College, Annapolis, Md.

B. L. Gildersleeve, Johns Hopkins University, Baltimore, Md.
W. W. Goodwin, Harvard University, Cambridge, Mass.

A. Harkness, Brown University, Providence, R. I.

Samuel Hart, Trinity College, Hartford, Conn.

M. W. Humphreys, Vanderbilt University, Nashville, Tenn.
Mary H. Ladd, Chauncy-Hall School, Boston, Mass.

Charles R. Lanman, Johns Hopkins University, Baltimore, Md.
Robert F. Leighton, Brooklyn, N. Y.

John R. Leslie, Newport, R. I.

F. A. March, Lafayette College, Easton, Pa.

A. C. Merriam, Columbia College, New York, N. Y.

Wilfred H. Munro, Bristol, R. I.

C. K. Nelson, St. John’s College, Annapolis, Md.

Charles P. Otis, Institute of Technology, Boston, Mass.

L. R. Packard, Yale College, New Haven, Conn.

Tracy Peck, Cornell University, Ithaca, N. Y.

William T. Peck, High School, Providence, R. I.

J. B. Sewall, Thayer Academy, South Braintree, Mass.

T. D. Seymour, Western Reserve College, Hudson, Ohio.
William E. Thompson, Genesee Wesleyan Seminary, Lima, N. Y.
Crawford H. Toy, Norfolk, Va.

Julia E. Ward, Mt. Holyoke Seminary, South Hadley, Mass.
Benjamin I. Wheeler, High School, Providence, R. I.

J. Colver Wightman, Taunton, Mass.

Alonzo Williams, Brown University, Providence, R. I.

AMERICAN PHILOLOGICAL ASSOCIATION.

Newrort, R. I, Tuesday, July 15, 1879.

The Eleventh Annual Session was called to order at 3 o’clock
P.M., in the hall of the Rogers High School, by the President, Mr.
Jotham B. Sewall, of Thayer Academy, Braintree, Mass.

An address of welcome was made by His Excellency G Governor
Van Zandt, to which the President replied.

Mr. Sewall announced the death of the Secretary, Professor
Thomas C. Murray, of the Johns Hopkins University, Baltimore,
Md., and the appointment of Professor Charles R. Lanman, of the
same institution, to serve in his stead from March until the next
election of officers.

The Treasurer, Charles ‘J. Buckingham, Esq., presented his
report, showing the receipts and expenditures of the past year.
[See p. 38.]

The Chair then appointed Professor A. C. Merriam and Professor
J. M. Garnett a committee to audit the Treasurer’s report.

The Secretary pro tempore, Professor Lanman, presented a report
from the Executive Committee, announcing the prompt publication
of the Transactions for 1878, and the election to membership of

Professor Albert 8. Cook, Johns Hopkins University, Baltimore, Md. ;
Professor Walter Q. Scott, Wooster University, Wooster, Ohio; Rev.
Ambrose J. Faust, Ph.D., Washington, D.C. ; and Mr. Joseph R. Ander-
son, Jr., Richmond, Va.

On motion, Professor B. L. Gildersleeve, Professor C. H. Toy,
and Professor F. A. March were appointed a committee to draw
up resolutions in commemoration of the late Professor Murray.

On motion, Professor F. A. March, Mr. Charles J. Buckingham,
and Colonel John R. Leslie were appointed a committee to arrange
the hours for the sessions.

The first paper was by Dr. EK. G. Sihler, of New York, on “The
Critical and Rhetorical Labors of Dionysius of Halicarnassus and

4 Proceedings of the

the Ars Rhetorica.” In the absence of the author, it was read
by the Secretary.

\

I. The extant rhetorical and critical works of Dionysius were all writ-
ten at Rome, with one probable exception. They are not a continuous
and systematic exposition of rhetorical theory and practice, but detached
treatises illustrating the mature convictions and tenets of the practical
teacher. Mr. Sihler endeavored to explain the principles and practices
therein embodied. Of the latter, the most important is the Atticism of
Dionysius, the rigorous and absolute exclusion of any models but the
' standard-bearers of classic Attic prose, of the century from Andocides to
Demosthenes. The practical instruction of Dionysius was therefore neces-
sarily merged into literary criticism. His criticism, again, to be appreciated,
must be judged from the one-sided, practical (non-historical) stand-point
of its author. The degree of suitableness of the several classic Attic
authors for practical imitation and the purposes of rhetorical culture
seems to have served as the basis of Dionysius’ canon. This explains
his strictures on the orations in Thucydides, on the fullness and poetical
flights of Plato, and on the padding and the monotonous rise and fall of the
Isocratean periods.

What we may call the system of Dionysius was then set forth: viz., 1.
His ofvSecorg or construction, rising from the analysis of sounds and the
metrical consideration of prosody in prose to that of clauses and periods;
2. His diction and vocabulary (@véuara), -

II. Very many critics have denied the Dionysian authorship of the
po-called Ars Rhetorica. It is an aggregate of detached pieces referring
exclusively to the epideictic kind of oratory. It holds in high esteem the
models of Plato and Isocrates. The former of these is often imitated in
minute terms and phrases, and even Demosthenes is made an imitator of
Plato. ;

Mr. Sihler suggested that the Ars may have been a performance of the
younger years of Dionysius, written in his Greek home (this piece alone is
dedicated to a man bearing a Greek name); and that the Lysianic and
Demosthenean standard influencing the other and later writings of Dion-
ysius may to some extent be due to his association and close coéperation
with the Atticist and purist, Caecilius of Calacte.

Mr. Stephen Pearl Andrews, of New York City, read a paper on
“ Ideological Etymology as a distinct Method in Philology.”

Mr. Andrews’ paper concerned the classification of words on the basis
of the Jdeus which they express. He stated as a fact that from Jacob
Grimm to August Fick, the opposite method of etymology, that of classi-
fying words upon the basis of their phonetic structure and affiliations,
had exclusively prevailed, as if there were no other possible method.
Still, it is perfectly obvious, on reflection, that every etymological process
has two factors, or concerns itself with two things: first, the words in
question, as phonetic structures or bundles of sounds; and secondly, the
deas involved, the meantngs-of the words; and that either of these two

American Philological Association. 5

may be given the first rank, and the other be subordinated to it. Conse-
quently, there are two elementary methods of etymologizing possible,
besides an ulterior compound method resulting from the coaction of the
former two.

For example, break, breech, and brag are three words which have a
certain double relation to each other: first, phonetically, which is pointed
out when we say that k, ch, and g are consonant sounds capable of being
interchanged by certain phonetic laws (and so of the vowels, ea, ce, and
@); so that one or another of these words is, presumably, an earlier form,
and the other two later, and ‘derived from it. This is the prevalent, and
what we may call the Historical, Physical, or German Method of etymo-
logical investigation.

But we may consider the whole matter from the opposite point of view,
thus: A breach is something brok-en, or which has undergone the process
of break-ing. Breach is, therefore, a derivative idea from break, and the
phonetic similarity of the two words may be tncidentally alluded to, as a
consequence of this natural alliance of the two ideas. So, to brag is to
break out, to throw one’s self into notice (Lat. jacto), and the similarity of
the word brag to breach and break may be again incidentally brought in,
to illustrate and‘confirm the natural or inherent alliance of the three ideas.
This latter method, which deals with the nature and affiliations of ideas or
meanings, as embodied in words, as of the first importance, and of the
phonetic phenomena of the words as secondary, is the New Method now
proposed, which Mr. Andrews calls Ideological or Psychical, or again,
the American Method. In this latter case, there is a distinct place for a
Science of Ideology, as underlying and controlling the study of the higher
aspects of Etymology; but an Ideology, still, studied and perfected through
the reflex potency of the direct study of words, and hence not merely
metaphysical, but distinctly philological.

Mr. Andrews does not wish it to be understood that the term American
Method has reference to himself (and his own nationality) as the first to
propound it, as a distinct method; it is a recognition of the merits, in
this regard, of Noah Webster, who somewhat unconsciously inaugurated
it when he reduced the meanings of the verbs of the English and allied
languages to thirty-four in number (see Introduction to Webster’s Dic-
tionary). Mr. Andrews, taking his departure from this spontaneous first
effort, of Webster, in the right direction, analyzes and further generalizes
his thirty-four classes of ideas, reducing them all to no more than three
grand major classes: 1. The idea of division or apartness (of, off, from),
Spencer’s Differentiation; 2. The idea of wnity or togetherness (at, to, with),
Spencer's Integration; and, 3. The idea of transition or vacillation between
these two (over, through, across, between).

The gist or core of Mr. Andrews’ paper, which was long and elaborate,
dealing extensively with the Indo-European Roots, is contained in the two
following propositions:

I. The Prepositions (as the words of Relation) contain in themselves
or represent all the primitive ideas, from which are derived the meanings of
all the other parts of speech (or words in language).

6 Proceedings of the

II. The Prepositions themselves, though very few in all languages,
are nevertheless susceptible of being reduced, in respect to their meanings,
to a very much smaller group, ending, indeed, in the absolutely funda-
mental difference between of and at (or from and %&), and their hinge-wise
connection with each other (between, etc.).

A recess was then taken from 5 until 8 o’clock.

Newport, Tuesday, July 15, 1879.

EVENING SESSION.

The Association met at 8.15.
The Annual Address was delivered by the President, Mr.
Jotham B. Sewall.

After some congratulatory words by way of introduction, ‘‘Our Duty
to our Mother Tongue,” was announced as subject.

First. As to its purity. Ought we to labor for it? We wish to speak
a pure language and to maintain its purity. But what does “‘ purity of
the English language” mean? If we mean by it freedom from admixture
with other languages, it is to be remembered that there is no such thing
as an English language by itself, simple and pure. From the time when
it could first properly be called ‘‘the English language,” it has been a
composite. Trench’s estimate is sixty parts Anglo-Saxon, thirty Latin,
five Greek, and five parts from other languages. The Anglo-Saxon is the
chief part. The other parts have been added and assimilated, a process
not of a moment and then done, but still going on and always to go on as
long as English-speaking people are active-minded and inventive, and
come to need terms which their own tongue does not supply. The purity
of the English language, therefore, does not require an obstinate resistance
to the admission of words from a foreign source, nor to the formation of
new ones. Indeed, it would be useless, because a resistance of the law of
growth. A needed word, or an apt one, must and will come. All that
can be meant by ‘‘ purity of the English language,” besides this, is freedom
from pedantry, vulgarisms, and barbarisms, and its correct use, and dis-
tinct and full enunciation as a written and spoken language. And for
this, obviously, it is important to labor.

Second. As to its restoration. By this is meant the bringing forward
of the Anglo-Saxon element to such predominance as to displace the other
elements as fur as possible—indeed, completely. The advocates of restora-
tion wish to bring back and re-instate the language of England in the time
of Alfred the Great as the English language, and banish Latin, Greek, and
all other elements from use. There arise two questions: Can it be done?
and, Is it desirable? As to the first, a work so completely revolutionary
would seem impossible. Not that the laws of the growth and change of
language would have to be set aside, but that the manipulation required
isa greater one than could possibly be brought into exercise. Something,
much indeed, is possible if desirable. The law of stimulus avails here as

American Philological Association. T

well as elsewhere, and influence can be exerted upon language for its
growth and change as upon other things. An eminent illustration of this
is the act of Edward III. (whereby English was substituted for Norman-
French in law use), and its result. If an increased interest in the study of
Early English should arise and the English language should take a more
important place in the studies of our higher schools and colleges, the
knowledge thus more commonly acquired would put the earlier language
back within more common reach, and familiarity with its store of words
would tend to bring back into use all such as were apt and needed; and if
it proved as available a source for new terms as the Greek or Latin, it
would naturally be so employed. We resort to Greek and Latin for new
words, not so much because they are a superior store-house of material,
but because their superior place in our methods of education has made
them more familiar and put them within easier reach. Is it then desir-
able? The revolutionary work aimed at by some is not desirable. There
is neither reason nor sense in shutting ourselves out from other stores and
denying ourselves the liberty of borrowing and assimilating from other
languages, if we wish. Why deny the discoverer and the scientist their
resource in the Greek, or refuse to accept so apt a word as cavion from the
Spanish, or even taboo from the savages of Polynesia? We shall not all
allow that the contributions to the English from other languages have not
been an enrichment rather than otherwise. Nevertheless, has not some-
thing been lost which it would be well to regain, and is there not a labor
in this direction worth performing to avail ourselves of the riches of
strength and beauty which once dwelt in the tongue? From the nature
of things, the language of the tenth century could not be the language of
the nineteenth, and the labor to make it so would be an attempt to do the
impossible. There was, however, an undue influence exerted upon the
language from exterior sources: first, through the Norman conquest; and
secondly, through the advent of the ‘‘ new learning,” to the detriment of
the language as the language of the English. There was a loss on the
Anglo-Saxon side, and while we take the position that purity does not
require of us unyielding resistance to the income of foreign words, we
may also take the position that we may well labor to recover what has
been lost. No one needs to be convinced that simplicity, strength, and
beauty will be gained by the recovery of old English words.

Third. Ae to its erthegraphy. ‘here ie need of reform. Our ertheg-
raphy ie truly a kakography. What ar tht objectione to referm? Nun,
save prejudic and inertia. But thtee ar great ebstaclee. ‘Dhey will yield
only to rationally directed and unremitted effort. Sumthing will be gaind
perhaps by tht introduction of new charactere fer sounde repreeented by |
several lettere. Sumfthing too mey be dun by returning to tht original
spelling ef wurde which hav been corrupted.

Fourth. Asto itsstudy. Our mother tongue has not the place it ought
to have in the curricula of our schools and colleges. It ought to have a
place of equal importance with any. Considering it asa language, and
its stores of history, poetry, oratory, etc., why should not a man of equal
abilities be able to do as much with it for himself as Demosthenes did for

8 | | Proceedings of the

himself with his mother tongue? In these days of March, Morris, Skeat,
and others, there is no want of means and instrumentalities, Aside from
the direct results which it is easy to see would flow from the study of
English with the same thoroughness and exactness which are now given
to Greek and Latin as to discipline and culture, obvious arguments in its
favor are to be found in the aid thus furnished to the work in behalf of
purity, restoration, and reform of orthography. Indeed, it is easy to see,
how, beginning with a historical study of the language, a logically con-
nected and well-defined course might be laid out, embracing literature,
rhetoric, and logic, and ending in these.

Professor A. C. Merriam, of Columbia College, New York City,
then read a paper ‘“‘On some Passages of the Odyssey.”

It has appeared to some weak and pointless to tell Nausicaa, ¢ 35, either
that she is a Phaeacian, or that she is noble, and this consideration has led
Bekker to omit the line. On the contrary, it was argued in this paper that
the poet had composed the line for the specific purpose of hinting that the
princess was wooed by native suttors and those only, and thus to intimate
at this early stage how distasteful they are to her, and the effect which the
coming of Odysseus may produce. For, the distinction in Greece between
the aristocracy and the royal family was so broad a one in the Homeric
period that the latter were accustomed to contract a marriage only with
royalty beyond their own borders, as is the custom among the reigning
families of Europe at the present day. Since Phaeacian manners and
customs are essentially Greek, simply a trifle more god-favored and effemi-
nate, the same custom may be supposed to obtain in Scheria. Indeed,
there is presumptive evidence of this in the history of the royal family.
Nausithous, a grandson of the king of the Giants, becomes king of the
Phaeacians perhaps by marriage, as Menelaus obtains the throne in Sparta.
His son Rhexenor may likewise have married from a distance before their
removal to Scheria, but he dies young, leaving an only child, Arete.
Soon after, the migration probably took place to their isolated home in
Scheria, and there Alcinous, loath to wed among the aristocracy, in time,
partly driven by the custom and their isolation, matches with royalty by
espousing his niece Arete. Compare the case of the sons of Aeolus,
x 5-7. Two sons of Alcinous have taken wives in Scheria, but there was .
no other resource if they were to be married at all. Yet there remains a
lingering hope in the family that for the only daughter, the darling of
them all, some Nausithous may be thrown in their way by Providence,
and she may thus be rescued from the nobles she despises, and sustain the
dignity of her station. It is such a state of affairs that the line in ques-
tion seems to disclose by reminding us that she is wooed by native suitors,
and if so, it is far from weak and pointless, since it explains the uttered
wish of the princess, ¢ 244-5, that Odysseus might remain and become her
husband—lines which were rejected by Aristarchus as too bold to suit the
maidenly character of Nausicaa; but it is to be observed that they are
spoken to her attendants, and the position of the group is such that they
are quite out of earshot of Odysseus. The thought of her approaching

American Philological Association. 9

marriage is continually before her, and is naturally a subject of frequent
conversation between herself and her maids. The change wrought in the
squalid sea-waif has been so wondrous that he must needs have been sent
by Heaven to their land, a thought which awakens at once the hope that |
he may be the one looked for to bring her deliverance from an alliance
beneath her dignity. He has already hinted his former importance in
the world, and it is not unusual in the poems to find the outer comeliness
taken to argue mental and moral qualities of worth, which in this case
have been proved as well by his speech as by his bearing throughout. Al]
this conspires to present the thought to her in the light of a possibility,
and the frankness of her nature reveals it to her companions, but the poet
has been careful that it shall not reach the ears of Odysseus.

Again, closely connected with these two passages are the lines ¢ 276-88,
which were likewise rejected by some of the ancient commentators for the
same reason as 244-5. But Goethe, with his true poetic instinct and clear
insight into the workings of the human heart, has divined the real char-
acter of the maiden, and vindicated both the passage and the purpose of
the elder poet. Besides a frankness and a naiveté so open that the thoughts
which spring in the heart fall naturally from the lips, Homer here develops
_ more fully, though covertly and by the dramatic method, that predilection
for Odysseus which has already been seen to be springing up in the maiden’s
breast. Itis this first feeling of love which makes her so sensitive to
the thought of the gossiping tongue that would couple her name with
Odysseus, and the poet with consummate art has veiled it carefully by
causing the maiden to put into the mouth of another what she censures in
her words, though she wishes in her heart. An engaging forwardness is
thus rescued from the verge of boldness by an expedient which Pope
declares to be ‘‘an instance of the great art of Homer in saying every-
thing properly.” These lines, too, contain as a whole the strongest con-
firmatory evidence of the theory advanced on 35. Nitzsch would retain
the remainder of the passage, while disposed to reject 280-1; but these are
exactly to the point. Despsiring of a release from the threatened indig-
nity, the maiden’s prayer turns even to the gods. Nor need such a prayer
be considered as presumptuous on her part. Poseidon himself is her
great-grandfather, and the gods are wont to come familiarly to fraternize
and feast with the Phaeacians who are of their kin. The poet who rep-
resents the: sea-nymph Thetis as married to Peleus and living for many
years in his palace, would surely feel no difficulty in the idea of sucha
union among the semi-divine Phaeacians. In fact we have a similar
prayer from the Greek Antiope of the Odyssey, 4 261.

Lastly, 7 311-16 was discussed. These six lines fell under the suspicions
of Aristarchus, and doubtless because of Alcinous’ startling offer of his
daughter’s hand to a total stranger; but the king is plainly quite capti-
vated by the commanding presence and bearing of his guest, and by the
eloquence and delicacy of sentiment he has displayed. His high station
in life has been directly asserted, and the immortality which Calypso de-
signed for him argued a lofty lineage befitting such a fate and such a love.
His sentiments have already met the approval of the courtiers, and

2

10 Proceedings of the

though Alcinous dissents from the thought that he might be angered at
Odysseus’ return with Nausicaa, yet he respects and approves the delicacy
that prompted the action of the hero, to the degree that he feels assured
that Odysseus’ feelings are at one with his bearing and eloquence. To
infer kingly station in a stranger from his appearance and comportment
only is not confined to the impetuous Alcinous in the Odyssey; cf. 6 68, »
223, » 253. A fair parallel to the case of Alcinous was cited from the
Gaelic poem of ‘‘ Evir-allen,” claimed to have been discovered by Baron
de Harold in Ireland and translated by him. That Alcinous believes
Odysseus to be no impostor, but the hero his words and appearance pro-
claim, he tells us himself 4 863-7. When to all this there is added the
strong predilection to a foreign marriage with royalty, the offer from so
impetuous a person loses its startling features, and becomes eminent!y
characteristic of the man. That the offer is not pressed is doubtless due
to a strong gesture of dissent from Odysseus at 7315. No allusion is made
to it later, and on the following day it is assumed by king and court that
Odysseus has a wife and family, The princess, too, has settled back to
the same conviction, and her few simple words of parting are tinged with
the melancholy of a shattered hope, though her dream had been all too
short to leave a sting behind. Homer has not handled this episode as
many a follower of his has done. Here we have no Medéa, no Dido, no—
but their name is legion. It was quite within his purpose to enthrall us
with his beautiful creation of the Phacacians, but it must in no wise
thwart the grander scheme of his greater epic.

In relation to the theory advanced this much was claimed: For three
long-suspected passages it supplies a thread that runs through them all,
knitting them together into unity and coherence, and connecting them
back to a fourth to which it gives a weight and significance that can
scarcely be overestimated, standing where it does and striking the key-
note of so many strains throughout the whole episode.

The committee on the hours of meeting reported. The report
was accepted with slight amendment, so that the hours were ar-
ranged as follows: from 9 o'clock to 1; from 4 to 6; and from 8 to 10.

The Association adjourned to 9 o'clock, Wednesday morning.

Newport, Wednesday, July 16, 1879.
MornNING SEssION.

The Association resumed its session at 9.20 a. m., the President,
Mr. Sewall, in the chair.

The appointed committee presented the following resolution
commemorative of the late Professor Murray, which was unani-
mously adopted:

The American Philological Association desire to give expression to
their sense of the eminent services of their late Secretary, Professor

American Philological Association. 11

Thomas Chalmers Murray, and to their deep sorrow at his untimely death.*
From the beginning of his connection with this Association. Professor
Murray showed on every occasion the utmost readiness to further the
interests of our body, and as Secretary distinguished himself by his
promptness, his accuracy, and his uniform kindness. A scholar of rare
attainments, an investigator of great acuteness and excellent balance, an
admirable expositor of the results of his studies, a teacher of unusual
power and suggestiveness, Professor Murray seemed destined to eminence
in his chosen department; but by the members of this Association his
memory will be cherished especially for his faithful discharge of duty
and his winning and self-sacrificing courtesy. Therefore,

Resolved, That this testimony to the worth of Professor Murray be
entered upon the proceedings of the Association and a copy be sent to the
surviving members of his family.

The Secretary announced, in the name of the Executive Com-
mittee, the election of six new members:

Mr. John Tetlow, Girls’ Latin School, Boston, Mass.; Mr. Wilfred H.
Munro, Bristol, R. I.; Mr. William C. Collar, Boston, Mass.; Dr. George
W. Ingraham, Cornell University, Ithaca, N. Y.; Mr. Benjamin I.
Wheeler, Providence High School, Providence, R. I.; Mr. Edwin De
Merritte, Chauncy-Hall School, Boston, Mass.

On motion, Professor W. W. Goodwin, Professor M. L. D’Ooge,
and Professor E. P. Crowell were appointed a committee to nom-
inate officers for the ensuing year.

On motion, Professor L. R. Packard, Professor T. D. Seymour,
and Professor Samuel Hart were appointed a committee to recom-
mend a suitable time and place for the next meeting.

Professor M. W. Humphreys, of Vanderbilt University, Nash-
ville, Tenn., read a paper on “The Nature of Caesura.”

1. Caesura serves two general purposes: (@) in long verses it gives the
reciter an opportunity to inhale, but falling in the midst of a foot prevents
him from taking too much time; and (6) it serves asa cinculum to hold
the two portions of the verse together, and docs not separate them. In
short verses, the latter is the chief office of caesura, and hence it can fall
between words closely connected: between the subject and the verb, e. g.,
Theb. 15; between the verb and its object, ibid. 270; between an adjective
.and its substantive, ibid. 18; between the article and its ‘substantive,
Philoct. 964; after a preposition, Oed. Rex 615; after ov, Iph. in Taur.
684; before pév, yap, etc., Orest. 3860; even before an enclitic, a break
is better than no caesura, e.g., Ion 574; and in other similar positions it

“Mr. Murray was Professor of Shemitic Languages at the Johns Hopkins University.
After an illness of a week, he died at Baltimore on the 2th of March, 1879, aged 20 years.
He Jeft in manuscript a couree of lectures on The Poctical Books of the Old Testament,
soon to be published: papera on The Home of the Shemitic Peoples, and on The Original
Case-Form in Shemitic; and a work on Hebrew Synonyms, well advanced towards
completion.

12 Proceedings of the — °

must be recognized. If it be claimed that such verses lack chief caesura,
we refer to verses which are admitted always to have a fixed caesura or
incision, as the trochaic tetram. cat.; cf. Troades 454, 458, etc., etc. In
454, it falls before an enclitic: this is not so surprising when we consider
that a grammatical pause may immediately precede an enclitic; as, An-
drom. 747; Hec. 432, etc. Many other arguments to sustain such caesurae
are here omitted, and many illustrations from Latin, and from the so-
called dactylic pentameter. The examples are very numerous, those
cited above being mere illustrations.

2. It is not necessary that a break, to constitute a genuine caesura,
must follow a polysyllabic word; but in some verses an incision or diae-
resis immediately before a chief caesura is objectionable. This is proved
by many examples taken from authors who otherwise never neglect
caesura: as, e. £.,

Quid amplius vis? | O mare et terra, ardeo.

8. The last section of the paper contained strictures on Hermann’s
ridiculing hephthemimeral caesura in trimeters and J. H. H. Schmidt's
rejecting the penthemimeral.

Dr. Robert F. Leighton, of Brooklyn, N. Y., then read “An
Account of a New Manuscript of Cicero’s Letters ad Familvares.”

Scholars have felt renewed interest in Cicero's letters ad Familiares since
the discovery of the manuscript of these letters at Vercelli in 1474. But
little is known of the history of this original MS., except that it is the old-
est MS. of these letters now known. The question has long been agitated
whether this MS.—the coder Mediceus xlix, No. ix, now in the Laurentian
library at Florence—is the only standard in determining the text of these
letters, or if other MSS. exist whose text is independent of this, and can
therefore serve as a means of comparison and correction. Orelli, after a
critical examination of all the MSS. in Italy and Germany known to him,
maintained that, with the exception of one page of a Turin palimpsest,
all MSS. of the letters ad Fam., were directly or indirectly copies of
the Mediceus. Baiter agrees with Orelli; Hofmann, however, eonsiders
the coder Parisinus Notre Dame 178, to be of independent authority for
the books which it contains: viz., i to viii, 8, 6 as far as the words tmped-
tendi moram. In 1827, the Erfurt MS. was discovered, and Orelli pro-
nounced it a copy of the Mediceus; but Meyncke and Buecheler after a crit-
ical examination proved that it was independent of the Mediceus. The
passage chiefly relied on by M. and B. was ad Fam. xv, 2,5, where, instead
of the one word cohortatus, the five words et tamen adoleacentem essem.
cohortatus are found. Orelli had noticed this; but he regarded the words
as a gloss of cohortatus.

In 1874, M. Thurot had the Tour MS. brought to Paris and examined.
This MS. had been known since 1829, having been mentioned by Haenel
(Cic. quaest. acad. crit. Ep. ad Fam., Sec. xii, m. 4—p. 482); but Orelli
contested the date assigned to this MS.; the other editors of Cicero's letters
have not even mentioned it. M. Thurot proved that it is a copy dating
from the twelfth century, and that it is independent of the Mediceus.

American Philological Association. 13

In 1839, Oehler called the attention of Orelli to the codices Harlejans in the

. British Museum; but these MSS. remained unnoticed until 1874, when

Franz Ruhl! gave a brief account of them in a note to Ritschl, published
in Rhein. Mus., vol. 30, pp. 26 et sq.

In 1876, the writer had an opportunity to examine these two codices, and
carefully collated them. They were purchased by Lord Harley in 1750,
and were brought to England from the monastery of Cusa, in Holland.
One is from the eleventh century (coder Hariejanus 2683), and the other
(codex Harlejanus 2773), from the twelfth century. They are beyond a
doubt independent of the Mediceus. In xv, 2, 5 are, besides the word
cohortatus, the same four words chiefly relied on by Meyncke and Buecheler
to prove the independence of the Erfurt MS. A number of passages from
these letters where the text is doubtful were quoted, and the value of
these new MSS. proved by the aid they furnish in revising the text. The
reading of several passages which have hitherto baffled the ingenuity of
critics is satisfactorily settled by these new MSS. The conclusion is that
all known MSS. of these letters are either copies of the Mediceus, or like the
codd. Par., Erfurt., Harl. primus et secundus and Turonenats, independent
of the Mediceus, but copies of the same archetype, and therefore servicea-
ble in correcting the Mediceus and thus settling the original text of these
letters.

Professor Albert Harkness, of Brown University, Providence,
R. I., read @ paper on ‘‘The Development of the Latin Subjunctive
in Principal Clauses.”

The paper aimed to show the steps by which the various meanings of
this mood as seen in Latin authors were developed out of the simple ety-
mological force of its original form. The Latin subjunctive contains both
the forms and the meanings of two moods, originally distinct, the subjunc-
tive proper with the sign @ and the optative with the sign7. These forms,
however, are used without any difference of meaning, and are made to
supplement each other. While, therefore, in Sanskrit and Greek the sub-
junctive and optative meanings are denoted by separate forms, in Latin they
are both expressed by the same form. Moreover the subjunctive and opta-
tive forms in Latin supply the place of the future indicative in the regular
verbs of the third and fourth conjugations. What now is the explanation
of this remarkable confusion in the use of forms, a confusion so great that
different forms, subjunctive and optative, occur with the same meaning and
the same form with different meanings? In this singular anomaly we rec-
ognize an important historical fact in the development of moods. It shows
that when the Latin first became a separate language, the forms of the sub-
junctive and optative, and of the future indicative were used with little or
no difference of meaning, a view fully confirmed by the etymology of the
forms themselves, The subjunctive and the optative moods are in their
origin only special developments of certain forms of the present indicative,
and originally denoted continued or prolonged action, from which was de-
veloped the idea of effort—attempted action. But earnest effort readily
suggests on the one hand desire, as we strive only for that which we desire

14 Proceedings of the

to attain, and on the other possibility or probability, as we may very likely
accomplish that which we are already attempting.

But secondly, what relation does this etymological meaning of these
forms sustain to the meaning actually found in the works of Sanskrit,
Greek, and Latin authors? If we have interpreted the etymology aright,
we must not expect to find any broad and well-defined line of distinction
between the provinces originally occupied respectively by the indicative
and the subjunctive. Accordingly, in the Vedas and the Homeric poems,
our earliest specimens of Sanskrit and Greek, we often meet with the sub-
junctive in senses much more closely related to that of the future indica-
tive than in later authors. Indeed, in the works of Homer, the aorist sub-
junctive and the future indicative are often identical in form and so closely
related in meaning that it is not always possible to distinguish one from the
other. But the Latin, Greek, and Sanskrit all conform tothe same analogy
in the use of moods. Indeed, the use of subjunctive and optative forms
to supply the place of the future indicative does not differ at all in kind
from the Homeric use of similar forms in a future sense. It is in fact
simply the result of carrying out the analogy on a large scale. In all these
facts we discern the germ from which was developed the potential sub-
junctive.

But the Latin subjunctive, like the Sanskrit and Greek subjunctive and
optative, also denotes desire, wish, a meaning which like the potential may
be readily developed out of the etymological signification of the forms.
Indeed, in the Vedas and the Homerjc poems, as well as in the early Latin,
we sometimes find the subjunctive of desire apparently in the very first
stage of its development, scarcely distinguishable on the one hand from
the future indicative and on the other from the potential subjunctive, dif-
fering perhaps from the latter very much as the two ordinary signs of the
English future, shall and :rill, differ from each other.

The two meanings now noticed for the Latin subjunctive, and for the
corresponding moods in the cognate tongues, the subjunctive and optative,
appear in different stages of development in the very earliest literary rec-
ords that have come down to us from any branch of the Indo-European
family, and are readily derived from the etymological meaning of the
forms themselves.

Moreover it is well known that these two meanings, which for conven-
ience we may call potential and optative, run through the whole range of
Sanskrit, Greek, and Latin literature, and that in these several forms they
embrace all the meanings known to the Latin subjunctive in principal
clauses.

But the subjunctive, in many subordinate clauses, also belongs to our
theme, as it was developed while those clauses were yet independent.
Thus in conditional, concessive, final, and consecutive clauses in Latin the sub-
junctive is entirely independent of the character of the clause in which it
stands, and was, in fact, developed before the clause became subordinate.
The paper closed with an illustration of this point.

Remarks were made upon this paper by Professor Alonzo Wil-
liams, by Mr. S. P. Andrews, and by Professor W. W. Goodwin.

American Philological Association. — 416

A paper by William A. Goodwin, C.E., of Portland, Maine, on
‘‘Chaucer’s Cecilia,” was read for the author by the President, Mr.
Sewall.

Dhis paper discust tht several interpreteation2 ef the name Cecilia (ae
recited in the Seconde Nonnes Tale of the Canterbury Telez), of which
the commentatore hav taken no notig.

Accerding to the nun’e first interpretation of tht neme,

‘It is to seye in englissh | heuenes lilie:”
i. e. caeli lilium, by reaeon of her chastity; er ‘lily,’ becauee of her whit-
ness of henor, vitality of consciens, and swvet sevor of good fame.

In thé next stanza sht interprets the name az ‘u wey to thet blind’
(caects via ?), becauee by good teaching sht wae a guide. ‘Dhis derivation
ie far-fetcht, and tht vowel2 alone of tht wurdz thus connected agrte
together; but what more closely analogous wurde ar ther ef tht requieit
meaning?

_ ‘Dhe rest ef tht stanza give a derivation from caelum and ‘‘lia.” Ia

“ta” Greek? Aiav mtane ‘‘exctedingly, very much;” and Aeia (which,
when pronoungst with iotaciem, might bt fairly repreeented by ia),
‘smooth, flat,’ may be concvivd of with tht same transition ef mtan-
ing a2 in the Latin plane, German glatt weg, dialectic platterdings, and our
flatly, and so positioly, complétely. Dht first part, then, saye the nun, ‘‘is
set for thoght of hoolynesse;” and the secund, ‘‘for hire lastynge bisy-
nesse” or activity. ht whole would then mtan ‘complétely intent on
holiness.’

Fer tht nun’2 next hermeneutic venttre,

‘*Cecile | may eek be seyd | in this manere .
Wantynge of blyndnesse,” * * * *
we mey adopt the desperate suppozition that caecifatis and sum ferm of
Aelrw (‘to lack’) floated vaguely before tht poet’s mind.
. Dht nun esseys, for the fifth and last time, to ‘‘expowne” the neme
ae followe:
‘Or elles loo | this maydens name bright
Of heuene and leos comth.”
She adde that men might rightly call her tht heven (caelum) ef people
(Aeoc), and illustrates in detail (stanza 16) the aptness ef the name in this
signification. Spenser’e understanding of the name may bt gatherd from
the couplet in Faerie Queene, i. x. 4:
‘*Dame Coelia men did her call, as thought
From heaven to come or thether to arise.” .

Remarks were made upon this paper by Professor F. A. March:

Het fthoght ther coud be no doubt that the ‘“‘Lia” ef the third
interpretation ie Leah, Latin Zza, Laban’e daughter, who wae a familiar
repreeentativ of laborious activity in contrast with Rechel, the reprezenta-
tiv ef the centemplativ life; ae may bt sven in the Latin Hymne; fer
example, in Bernard of Clugny’a rhythm, well known in Dr. Nevale'e
translation under tht name of ‘‘Dhe Celestial Country” (March’e Lat.

16 Proceedings of the

Hymne, p. 280); and, ae Dr. Lanman suggests, in Dunte (Purgatorio, xxvii,
101, and the commentetore).

For tht uther interpretation2, in which it ie taken for granted by the
poet that he will be understood, familiar wurde ar to bt preferd. TDhe
_ French late, Old French lée with which Chaucer would associate hia aley
(alley) ie better than via fer tht secund interpretation; and in tht fourth
lees i. e. -less ie better in itself than Aeizxw, and nattrally leade to the leos
ef tht fifth.

Professor M. W. Humphreys, of Vanderbilt University, Nash-
ville, Tenn., presented in brief abstract a paper on ‘Certain
Effects of Elision in Versification.” The character of the paper
involved too much quotation to admit of reading in full.

1. Diaeresis with elision in the middle of iambic trimeters serves as a
sort of hephthemimeral caesura, the elided vowel in this case not being
entirely suppressed. The author gave at length the reason why elision at
the end of the first dipody can not create penthemimeral caesura (except
when the elision is between two speakers), and also a reply to Hermann’s
objections. Elision at the middle creates caesura or quasi-caesura ; and
this may be shown by statistics.

(a) If we take the strict but false view of caesura, as explained above—
page 11, we find, in verses apparently without chief caesura,

with quasi-caesura: without quasi-caesura: Percentages:
In Aeschylus, 89 instances; 39 instances. 69 to 31
In Sophocles, 150 “ 53 “s 74 “* 26
In Euripides, 815 “ 101 “ 76 ‘* 24

(0) If we take the liberal and correct view of caesura, the results, tab-
ulated as before, are:

with quasi-caesura: without quasi-caesura: Percentages:
In Aeschylus, 42 instances; 19 instances. 69 to 31
In Sophocles, 44 “ 9 “ 83 ‘* 17
In Euripides, 123 “e 1(?) 99 ** O1

These tables were followed by a list of instances of quasi-caesura after
the first dipody, in which elision seems to take place between two
speakers, but where there is in reality no elision at all.

2. Elision among the Romans never being total, the partially pro.
nounced vowel produced less effect than in Greek where the elision was
usually total; hence quasi-caesura among the Romans was unusual,
except between two speakers, and then it took place after the first dipody
as well asin the middle of the verse. Many examples might be given from
Plautus and Terence ; thus, Casina 352, 509, etc. ; Andria i. 1. 7; i. 1.
92, etc.

3. The quasi-caesura was explained and illustrated by many other
metrical phenomena.

4, Elision at theend of the verse. Thisis not employed in Latin iambic
trimeters, nor in Greek heroic hexameters ; but in Greek trimeters and

American Philological Association. 17

Latin hexameters, especially those of Sophocles and Vergil. The condi-
tions under which it is admitted or required were explained, and a note
was given discussing the lengthening as well as the elision of -que.

5. This section contains a full discyssion of the Porsonic Law, and
shows that the Greek Tragedians, before they were ‘‘emended,” allowed
elision to excuse the neglect of this law, the elision being total, and forcing
or allowing the two words to be pronounced nearly as one word.

6. Discussion of the Porsonic Law among the Romans, with a special
note on the metres of Seneca.

7. Relations of ictus to accent as modified by elision. See Transac-
tions of the Am. Phil. Assoc. for 1876.

This paper is a sequel to the one on ‘‘ Elision, especially in Greek,”
published in the Transactions for 1878, and to the paper on ‘‘ The Nature
of Caesura,” of which an abstract is given above, page 11.

Remarks were made upon this paper by Professor D’Ooge, Mr.
Sewall, and Professor Merriam.

A paper by Professor S. S. Haldeman (University of Pennsyl-
vania), Chickies, Pa., “On Spurious Words,” was read by Dr. C.
K. Nelson.

Among the phases of words to which I hav givn names, is one termed
parop’sis (implying a false view),* due to bad writing and consequent type
errors, as in printing ‘Hebrides’ for ‘Hebudes,’ where, as in other cases,
the false form has been legitimised.

In the useful, condenst, and generalli accurat English Dictionary of
Hyde Clarke, D.C.L., we find, in its alphabetic place— ‘ag’non’ (for
‘ag'rion,’ a kind of dragon-fly) and ‘inli’dan’ (for ‘tu’lidan,’ a milliped),
which Webster and Worcester giv wrongli as ‘ta’lidan,’ with ‘1,’ which
Webster also puts in ‘fa’lus’ (for ‘talus’ or ‘julus’), a concurrence of long
sounds opposed by a law of English speech, and seldom ad missible.t

In some dictionaris, we find ‘mis’gum’ or ‘mis’gurn,’ defined as a fish
like an eel in size and form. The nativ countri is net mentioned, which
leavs the language and etimologi in doubt. The spellings hav a Celtic
(‘c’ as k) appearance; ‘misgurn’ has a Welsh look, but this language seems
to giv noclu. In Gaelic we find ‘easgann’ (an eel; ‘easg,’ a ditch), which,
when badli writtn, accounts for both forms of the supposed word.

‘FYorin’ (a kind of pasture-grass), is thus givn, with Italian ‘fiére’
(flower) as the probabl etimologi, which the assignd accent contradicts.
The plant is a nativ of Ireland, and we may associate its name with Irish
‘féur’ (grass, fodder), ‘fiuran’ (a weed eaten by cattl), ‘feoran’ (a green, a
grassi field); Welsh ‘pori’ (to graze), ‘ pawr’ (pasture, grass), pl. ‘ porion.s
Compare ‘ g-wyr-dd’ (green), Latin viridis.

‘Frorite’ (a kind of mineral) follows the preceding word. Being derived

*Onutlines of Etymology. Philad., J B. Lippincott & Co., 1878.

t Clarke gives wrongli the genus of the spider-monkeys (Ateles) as a plural, defined as
‘monkeys.’ The name means imperfect, and is in allusion to the bands, of which the
thum is incomplete, or wanting.

3

18 Proceedings of the

from ‘ Fiora’ (an Italian localiti), the proper word seems to be ‘fid’rite,’ in
two syllabs.

Altho ‘pregnable’ is from French ‘prenable,’ it seems to hav borrowd
the gay of ‘expugnable.’

When ‘ daalder’ (= dalder, the d@ educed from 2, a Dutch form of
‘dollar,’ is given as da-al’-der by the lexicographers, they invent a spurious
word and demonstrate the absurditi of pronouncing foreign words, and
even certain English speech-words, according to pretended ‘‘ analogies”
of the English alphabet.* |

A ‘ctipel’ (= cipl) is a kind of cup used by refiners and by them calld
cup'l and cop’l, but the dictionaris indicate the sound as ci-pel. Pryce
(Mineralogia Cornubiensis, 1778), describes the ‘‘cuppel” and ‘‘ cuppella-
tion ”—thus pronounst in the lectures of the late Professor Hare, Univer-
sity of Pennsylvania.

Lecturers on anatomi hav a term ‘ poplite’al,’ which lexicographers (who
do not-hear anatomic terms) misred as ‘poplit’eal.’ So the speech-word
‘either’ is sometimes misred ‘ eye-ther.’

Formerli Worcester had ‘ X&ng’ti’ which he markt as ‘zang’te,’ with
English z and obscure e. This is an important word in theologic contro-
versi, upon which volumes hav been writtn. The ‘x’ (= 8h) is Portuguese,
and the word has such spellings as ‘Shang-Te’ and ‘ Shangti.’

If the word ‘sacciform’ (= sakkiform, from sac or sack), is spoken as
‘saxiform,’ it should mean rock-shaped. As Latin discus has given ‘disc’
and ‘dish’ to English, a supposed word ‘dissiform’ would be spurious,
because, to follow the law of English speech, the first syllab of disc-i-form
shud be pronounst ‘disk’ or ‘dish,’ but not ‘ diss,’

In Old English, ‘e’ had the Europe’an power, and ‘ meet’ was pronounst
‘mate’ as in ‘ helpmeet,’ a helper, or in modern spelling—‘ helpmate.’

In some encyclopedias the accent of words is given, except where it is
not known; but the orthoepists pretend to supply the correct sound and
accent for the entire vocabulari, including such spurious forms as never
had a pronunciation.

A paper by Dr. Anton Sander, of Lawrenceville, N. J., “On
Greek Negatives,” was read by Professor Packard, of Yale College.

This paper maintained three propositions : :

1. That the sequence in Greek of compound negatives after a simple
one may be more intelligible to us if we observe that these compound
negatives are not disjunctives (like ‘‘neither—nor’’), but correspond to
‘‘and not” or ‘‘ but not,” “‘and never,” etc. They simply repeat the neg-
ative with each repetition of the conjunction, though in actual use this is
Sometimes overlooked and another conjunction used also.

2 The much discussed combination ov 7 is not, as is commonly thought,
a stronger, but a weaker negative than ov. It is not used where an
emphatic negation would be expected, e. g., in oaths, in answer to a ques-

* See further upon this word and its affinities in my address, Spelling Reform Bulletin,
No. 1, April, 1877.

American Philological Association. 19

tion, with the imperative, or with the present or past indicative. The 44,
as a weaker negative than ov, can not reverse it entirely, but does so par-
tially, and the phrase od uf is therefore a weaker nevation than ov alone,
and may be translated hardly, scarcely, not always, etc.

8. In expressions of fear (dédocxa 2?) [nu ob] EAD, témeo ne [ut or ne non]
ventat) the negative “4 or ne reverses the negative idea of the verb to fear,
thus leaving an affirmative result, ‘‘ I fear he will come ;” but the second
negative, ov or non, restores the negation, ‘‘I fear he will not come.” The
same negative idea is expressed in Latin by wé alone, for then the negation
in the verb témeo is not destroyed by a negative after it.

Remarks were made upon this paper by Professor Gildersleeve.
The Association then took a recess until 4 o’clock.

Newport, Wednesday, July 16, 1879.
AFTERNOON SESSION.

The Association resumed its session at 4 P. M.
Professor T. D. Seymour, of Western Reserve College, Hudson,
Ohio, read a paper ‘On the Date of the Prometheus of Aeschylus.”

All efforts to establish the date of the representation of the Prometheus
have failed to discover convincing arguments. While Schoemann and
others hold it to be the earliest of the extant plays, Moriz Schmidt classes
it with the Rhesus, as the stragglers—der Nachtrab—of Greek Tragedy.
To some the simplicity of language and construction seems to savor of the
earliest times; to others it seems to be the result of Sophoclean influence.
To Westphal the character of the metres seems to point to a late epoch in
tragic art: to Wecklein the careful formation of the trimeters seems to
indicate that the play was written before the Persians. Some think that
only two actors were employed, and assign the play to an earlier period;
others think that three actors were necessary, and that the play must have
been produced after Sophocles had introduced the third actor. A much-
quoted English authority selects for the date of representation 464 B. c.,
‘‘the year when the news would reach Athens that Themistocles had
entered the service of the Persian king,” and refers to this act ‘‘the abom-
ination of treason which the poet puts into the mouth of his chorus;” but
another Englishman supposes that the play was intended as a glorification
of Themistocles, in which case it must have been written before the
Persians.

These arguments are all based upon the metre, the language, and the
construction of the play, or upon some uncertain political allusion. From
the very nature of the case they are inconclusive. Between the Persians
and the Orestean trilogy only fourteen years passed. These were by no
means the Lehrjahre of Aeschylus. He had been writing tragedies for
twenty-eight years, and was probably fifty-three years old. While he was
ready to accept suggestions from his younger rival and modify his stage
alrangements, we have no reason to suppose that he materially changed

20 Proceedings of the

his style of composition or his metres. We have too few of his works to
justify us in the assertion, from internal evidence, that one play belongs
to his sixtieth year and another to his sixty-fifth year. We do not know
what is accidental in the play, and what belongs to his period of work.

A surer criterion we may perhaps find in a comparison with the works
of Pindar, of which Aeschylus was a careful student, as he was a student
of Homer, and did not hesitate to borrow even from Phrynichus.

The resemblance between the most lyric of tragic poets and the most
tragic of lyric poets has been often noted. Born in the same lustrum,
educated in the schools of the same city, though widely separated by their
different relations to the wars for the freedom of Greece, they were alike
in their ethical views, in their magnificent and sublime style, and, as we
should expect, were alike in their use of words and expressions. Thusa
comparison with Pindar has emended at least one line of Aeschylus, and
has explained many another.

But not infrequently we find coincidence of thought or language, which
seems to be the result of conscious imitation, rather than of the like time
and character of the poets. We may compare Pindar, Pyth. viii, 95:
émdpepoe’ ti dé tic; ti & ob tig; oxtac dvap | dv¥pwroc, with Aeschylus, Aga-
memnon 839: eldwAov oxcac. The resemblance is perhaps as close, although
not so obvious, in Prometheus 545 fg.: gép’ duc adyapic yapic, & gidog,
eiré mov tig Gand; | Tic épapuepiny dpnttc; ord’ édépyxdne | dAtyodpaviay aacxvy |
iadévetpov, d Td gwrav | adady (déderat) yévog éurrerrodiopévoy KrA.; and in Prom.
448: GAA’ ovecpdruv | GAiyKioe wopdaicr: KrA.

But among the more specific points of resemblance between odes of
Pindar and the Prometheus of Aeschylus, the most remarkable is the
description of the hundred-headed monster Typhon. Pindar, Pyth. i, 15
fg.: ¢ 7 tv aivg Taprdépy xeirac Sedv rodéutoc | Tupac Exatovtaxdpavoc: Tév Tore |
Kidixiov Splpev rodvovvpoy dvtgov' viv ye pudv |... ZeneAia 7’ avrov reklee orépva
Aayvdevra: xiwy J ovpavia cuviyet | vipdeco’ Alrva... Tag épetryovra: pév arAdrov
mupo¢ ayvérara: | éx puyov wayai xrA. See also Pyth. viii, 16 (where we see
that Typhon was not yet Aetnaean), and Fragg. 92, 98, Bergk. With
these we have to compare Prometheus 351-354 fg., 364 fg.: rév yyyevq re

Kidixiwv oixhtopa | dvtpuv .. . ddtov répac | Exatoyxdpavov . . . | Tugdva Sodpov,
aot 0S avréory Seoic |... xeitat orevwrov TAnciov Sadaaociov | imotpevog pilaery
Airvaiay iro |... Evdev ixpayhoovral more | worapoi rupic ... amAdrou BéAect

wuprrvéov CdéAnc.

That the resemblance here is beyond the workings of chance is evident;
but which is the original? From the structure of the first Pythian ode,
and from the cardinal importance of the mention of Typhon there, viewed
in connection with the fact that it is no essential link in the chain of the
tragedy, the writer of the paper considers it probable that the tragedy, or
at least that passage, was written after the first Pythian ode, which is
dated by Boeckh 474 B. c., but by Bergk 470 B.c. The only limitation of
date for Fragg. 92, 98 is that the hymn from which they were taken must
have been written before the death of Hieron in 467, or at all events before
the overthrow of his dynasty in the next year.

In the second Pythian ode, in honor of a victory gained by the same

American Philological Association. 21

Hieron, we find, 93 fg.: g¢épecv & thadpic éravyéviov AaBdvra Cuydv | apfyec’
wort xévrpov dé rat | Aaxtiodéuev redédec | dAca9npd¢ oluoc. We find the same
in substance in Agamemnon 1624: wpdc xévrpa pa) Adxtile, and in Prom.
822: obxovy Euorye ypOuevog didacndAw mpdc xévtpa KOdAov Exreveic. This soon
became a proverb, but that it was not such already in the day of Pindar
is shown by his treatment of it. He would hardly have developed the
figure if it were not his own fresh expression.

In this same second Pythian ode, verse 11, we find: év 9 dppara reiorxd-
diva xaralevyvig | odévoc ixmuv, by which we are reminded of Prom. 465:
tg’ apa 7’ Fyayov gLAnvioue in Trove.

In the same ode, verse 34, in speaking of Ixion’s passion for Hera, the
poet says: yp7 dé nar’ avrov aici mavroc dpav perpor, | evvai dé rapdrpora é¢
xaxérat’ adpdéav | &8aAov, In Prom. 890 the same thought is expressed: o¢
TO kndevoat xad’ éavTov aploreber paxpy.

The paper calls attention further to the prophecy of Themis concerning
Thetis, as given in Isth. viii, 84 fg., and Prom. 908-909 and 920 fg.; also
to the hyporchema written for Hieron, Frag. 105 Bergk, according to the
scholion on Pyth. ii, of the same date as that ode, as compared with
Prom. 709-710, which seems like a dramatic development of Pindar’s
expression.

How are we to account for the fact that Aeschylus borrows expressions
and allusions from Pindar more frequently in the Prometheus than in any
other play, and mainly from the odes or hymns in honor of Sicilians?

Not long after 473-2 s. c. Aeschylus visited Syracuse, and at the request
of Hieron presented his Persians. How long he remained in Sicily at that
time is uncertain. We know he was again in Athens 467 B. c., when he
brought upon the stage his great Theban trilogy; moreover in that very
year Hieron died and his court was in confusion. During this visit to
Syracuse, between 472 and 467 B. c., besides giving a new representation
of the Persians, it is intimated that Aeschylus wrote his Aetnaean
Women, and may well have written the Prometheusalso. That the Prome-
theus was written after the Persians, i. e. after 472 B. c., is made probable
not merely by the fact that the first Pythian ode may not have been
composed until 470 B. c., but also by the probability that the satyric drama
of the Persian trilogy, Prometheus the IIvp¢époc, was written before the
Promethean trilogy.

No reason appears why the Agamemnon should not be as full of Pindaric
expressions as the Prometheus, except that it was not written for the court
of Hieron. In the prediction concerning the eruption of Aetna; in the
mention of the smooth fields of fertile Sicily and of the monster, where,
as has been remarked, it is the poet rather than Prometheus who speaks;
in the warning not to ‘‘kick against the pricks;” in the exhortation to
wed in one’s own rank; in the Scythian wheeled houses—in all this, Hieron
heard allusions to his victories at Delphi and the Epinikia of Pindar in his
honor—allusions which were well understood by his court and the Greeks
generally. From this, then, it seems probable that the Prometheus was
written in Sicily between 471 and 468 B. c.

22 Proceedings of the

Professor C. H. Toy, of Norfolk, Va, read a paper “On
Shemitic Derived Stems.”

Professor Toy said that one peculiarity of the Shemitic family of lan-
guages is its symmetrical system of derived verbal stems, Reflexives,
Causals, and Intensives. Such forms exist in other languages, but not
with the regularity of the Shemitic. The latter approaches in this respect
the agglutinative tongues (as Turkish, and especially Hottentot), which
put out their strength on the expression of objective verbal conceptions
to the comparative neglect of the subjective.

This paper offered some remarks on the general table of Shemitic
derived verb-stems. First, the close similarity of those forms in all the
dialects. Omitting a few rare and doubtful stems, we may infer that the
system was the same in all, and that it therefore existed in primitive
Shemitic. This makes it the harder to reach the original forms and mean-
ings, since there are no different forms with which to compare these.
Next, the regularity of the formation, and the simplicity of the material.
The little phonetic degradation discoverable simplifies the statement, but
increases the difficulty of analysis. The material falls into two divisions
with two corresponding classes of significations: (1) Internal modifications
of the stem by doubling a letter or a syllable, by broadening a vowel, or
by inserting a weak consonant (w, y, n), or by combinations of these.
They express intensity, or the affecting an object by the action, and are
called intensives and affectives. (2) External modifications of the stem
by various prefixes, na and ta making reflexives, and sa and ha (or a)
making causals.

As to the origin of these modifications not much that is definite can be
said. The reduplication is apparently symbolical, and some insertions of
ew and y may be referred to reduplication, while others seem to involve
the addition of separate syllables. The vowel-broadening also may be
symbolical. The prefixes offer great difficulties, so that we have either to
assign them very general meanings (demonstrative, for example), or else
to suppose that they are remnants of earlier longer forms now lost. All
that seems probable is that the Derived Stems were originally nouns: we
may compare them with other nouns in Shemitic, but no satisfactory re-
sults are obtained.

In the local distribution of the stems some diversities are observable.
The na is most used in Assyrian, the fa in Assyrian and Aramaic, the sa in
Assyrian and Ethiopic, the ha in Aramaic, Hebrew, Arabic, and Ethiopic,
the reduplication and vowel-broadening in Arabic and Amharic. The
north-eastern group (Assyrian-Babylonian) has the greatest development
of reflexives, which, however, it often uses as passives; the northern
(Aramaic) regularly makes passives of its reflexives; the central (Hebrew
or Canaanitish) has made the greatest reduction in the number of the
stems, but kept two passives made by internal vowel-modification, while
the southern (Arabic and Ethiopic) exhibits the greatest general richness
of form and signification, the Arabic alone also having a regularly-formed
passive. The tendency has been to drop stems, not only in the ante-
historical period, but also later.

American Philological Association. 23

Professor F, A. March, of Lafayette College, Easton, Pa., read
a paper “On the English Dictionary of the Philological Society.”

An ‘‘Apptal” hae just been issiied to tht Epglish-sptaking and Epglish-
reading public to read books and meke extracts fer Dht Philological
Society’e new Kpglish Dictionary. ‘Dht dictionary hae now been more
than twenty ytare in preparation. It began with a reaoltition to prepare
a supplement to existing dictionariee, adopted en the suggestion of a
peper by Dean Trench, red in November, 1857. Dhe wurk upen this
opend up so many deficiencie2z, and suggested a dictionary so different
from the old that tht plan wae soon enlargd, and in January, 1859,
a ‘‘ Propozal fer a new Epglish Dictionary” wae issied. It wae to be
etymolegical and histerical. Dht original propoeal made much ef the
etymological side. Every bedy wae askt to send in etymologiee. But
tht historical sjde provee to be tht more impertant. It ie propoeed to
rvad substantially all the books in tht language, and make quotatione
for all the wurde which occur in them. ‘Dhtze quotatione ar to be made
en a tiniform plan, tach on a slip ef paper ef tht sjze ef a half shtet of
note paper. TDhoee for tach wurd will bt breght together frem all the
books, classified accerding to their mtaninge, and arrangd in histerical
order, so ae to giv tht history of the wurd.

Dhe wurk of preparing this mattrial went on with zeal fer ytare. Dhe
original editor, Mr. Herbert Coleridge, djed, but hie plage wae taken by
Mr. F. J. Furnivall, who net only urgd the wurk ferward, but started
the Early Epglish Text Sogiety to print and reprint rare old texts and put
them into tht hande of rvadere. Mattrial accimileted so fast that it
stemd impossible to fjnd any publisher to print the Dictionary, and for a
number ef yeare interest in it had almost died out.

Lately, however, ernest efforts to arrange fer publication hav been
meade, and, finally, the Delegatee of the Clarendon Press, in tht Univer-
sity ef Oxford, hav assiimed tht entjre financial responsibility of the
undertaking. Dr. Murray, the Prezident of tht Sogjety, will edit it, with
a number of sub-editore. ‘Dhe letter A, four hundred pegee, i2 to be out
in 1882, and tht rest ie to fellow in tht course of ten ytare, if pessible.

Dr. Murrey finde sume two tune ef slips en hand. In tht erliest
ptriod, when tht books ar few, the wurk ie fairly dun or premist; but
in the later centiiriee, many books remain untucht. C new ‘‘Apptal” ie
now isstied for help. (1 thoueand volunttvere ar askt fer to cempltte tht
reading within the next two yeare.

American books hav hardly been tucht. Dhey ar all left for American
rtadere, ae ar also books of British authore of tht 18th Gentiry. Four er
five hundred American rtvader2 ar nteded in order to compltte so soon
this liberal allotment. Dr. Murray says that any wun can help, especially
with medern books; hie pipile hav supplied him with fjve thouezand good
quotatione ina munth. But ef course persone who hav access to original
editione ef authore ef tht 18th Centiry, and who hav sum scholastic
preparation for the wurk, must do tht most important part ef it. Mem-
bere of this Assogiation ar lookt to with hope. Voeluntvere ar requested

to giv tht tjtlee of four or fjve books which they hav at hand and ar

24 Proceedings of the

inclined to undertake, so that a selection may bt meade at wunce. If
original editione ef eighteenth centtiry books ar to bt had, they ar to be
preferd; if not, American authore of uther date ar to be taken.

In American: books tht first theght ie to seciire quotatione for all the
wurde tized to name the fieical featiree, productione, and uther objects,
and tht peciiliar acts, habits, and relatione to bt found htre. Erly books
of travel, law, er recorde ar to be seght, in which such namee would bt
likely to make their first apptvarang; so also books and pamfiets on natf-
ral history, surveye, and exploratione.

Besjdez thtee neveltiee, it ie to be rememberd that quotatione ar
wanted for all the commun werde in their cemmun mbtaninge fer tach
generation. We ought to send in a large bedy of thtee frem our great
American authore, our statesmen, lawyere, and fheologiane, and our men
ef scjen¢, a2 well ae our poets, novelists, and historiane. Every happy
expression of thoee theghts which Americana most valte and act upen,
which can bt found in print in suitable cumpass for quotetion, mey well
be put upon its slip and sent to Dr. Murray, so that American theght
mey bt fairly and fully repreeented in the Dictionary.

Local Dialect ie net to be included in this Dictionary. (Qfter it ie dun,
it ie propoeed to begin on a Dialect Dictionary tniform with it. Rtadera
in the United States ar notifjed by Dr. Murrey that ‘‘ they will save time
by first commiiniceating with Prof. F. A. March, ef Lafayette College,
Easton, Pennsylvenia, whom,” ht saye, ‘‘wt hav askt to erganjze and
gujde the wurk ef our American frende.”” Dr. Murrey mentione that ht
will furnish rtadere with printed slips, bearing tht date, author, and tijtle
ef their books, so az to save mechanical labor ae much ae pessible. Dhe
Reference list of Booke at tht end of tht Dictionary will record the
nameez ef their Rvadere.

Any wun eble and willing to act ae Sub-editor in arranging, classifying,
and completing tht preparetion ef the mattriale for tht Editor’e revizion,
i2 requested to commiinicete with Dr. Murray. Hie address ie Mill Hill,
Middlesex, N. W., Epgland.

The Association took a recess from 6 to 8 P.M.

Newport, Wednesday, July 16, 1879.
EVENING SESSION.

The Association came together again at 8.20 P. m.

Professor M. L. D’Ooge, of the University of Michigan, Ann
Arbor, Mich., read a paper examining critically the views of Pro-
fessor Kirchhoff, of Berlin, ‘On the Final Recension of the De
Corona.”

In the Abhandlungen der Kéniglichen Akademie der Wissenschaften eu
Berlin, 1875, Kirchhoff presents the following theory of the origin of the

de Corona of Demosthenes in its present form:
Demosthenes wrote out his plea soon after the indictment was first

American Philological Association: 25

brought, 336-5, B. C. This plea could have been directed only against the
formal complaint; beyond this nothing was definitely known of the
attack of Aeschines. The original draft of the oration, accordingly, is
found in §§ 53-121 plus §§ 3, 4, 8 of the Prooemium. But when the
trial occurred, some six years later, Ctesiphon and Demosthenes agreed to
divide the matter of the defense in such a way that Ctesiphon should
make the reply to the legal points and Demosthenes should pay exclusive
attention to the political issues of the case. Accordingly, Demosthenes
replies to his opponent without making use of the old draft drawn up in
defense of Ctesiphon, and in extemporized language. In writing out his
oration subsequently, Demosthenes, in order to give completeness to the
defense, recasts the old speech and seeks to engraft it upon the body of
the speech actually delivered by him. Kirchhoff regards § 75 and §§
95-101 as interjected into the body of the older speech. Demosthenes,
finding it impossible to make one consistent whole out of the two speeches,
contented himself with reproducing, as accurately as he could, the speech
on the political issues. This later speech is contained in &§ 1, 2, 5-7, 10-
52, 122-824 of the de Corona as we have it.

There were found, therefore, among the literary remains of Demos-
thenes these two distinct speeches with traces of an attempt at combina-
tion. The unknown editor of the oration in its present form completed
what Demosthenes set out todo. One of the connecting links is § 9.

Kirchhoff believes that the plan of the oration and the frequent allusions
to the words of the rival speech imply an acquaintance with the contents
of the speech of Aeschines that argues a recasting of the de Corona sub-
sequent to the trial. He further thinks that the entire passage on the
peace and the embassy, §§ 9-53, was composed after the trial, and con-
nected by Demosthenes, or his editor, with the earlier speech.

This paper aims to show that Demosthenes had reason to expect a
renewal of the charges with reference to the peace of Philocrates previ-
ously made by Aeschines in the de Falsa Legatione, and that, excepting a
few direct allusions to the words of Aeschines, there is no good reason
for supposing that this part of the de Corona was not composed in advance
of the trial.

But the main support of Kirchhoff’s theory lies in the supposed rela-
tion between that part in which Demosthenes treats of his public career,
8§ 53-110, in answer to the first count of the indictment, and the rest of
the oration. Kirchhoff infers from the language in § 110 that when the
orator composed this part of his speech he had no expectation of entering
upon the vindication of his policy subsequent to the trierarchal law,
Olympiad 110, 1.

The unreasonableness of this inference is shown (a) from the nature of
the case, and (6) from the natural interpretation of the language in § 110 asa
rhetorical artifice, when compared with similar expressions found else-
where. From Kirchhoff's theory it must follow that the direct allusions
to the words of Aeschines found in the supposed older part of the ora-
tion, sc. $$ 53-120, are more easily recognized as later insertions, and do
more violence to the connection of thought than is the case with similar

4

26 Proceedings of the

allusions found in the supposed younger part of the speech, sc. §§ 10-52,
121-324. This point is fully considered. It is shown that out of the
twenty-nine direct references in the de Corona to the language of Aes-
chines, eight are found in the supposed older part, and nineteen in the
supposed younger part of the oration, not counting the prooemium; that
the distribution of these allusions seems controlled by a predetermined
arrangement of the subject-matter; and, above all, that these allusions
are not any more consistent and more closely connected with the context
in the part of the oration supposed to have been written after the trial,
than in the supposed older part. The comparatively loose connection of
SS 126-128, 225-226, with the context is noticed. §§ 227-231 are discussed
as a later insertion. The passage §§ 232-237 seems to owe its existence
to an allusion in Aeschines. § 247 seems displaced, naturally following
close upon § 239. § 241 seems to be a repetition of § 230; it may be that
§ 230 is the later passage inserted in response to the illustration of Aes-
chines with reference to accounts. In §§ 276-284 are pointed out clear
instances of later insertion. It seems natural to connect §§ 285-290 di-
rectly with § 250, and to suppose that in the original draft the honors
conferred upon Demosthenes immediately after Chaeronea were named in
close connection.

The theory of Kirchhoff involves also the following difficulties: (1)
There is not the slightest intimation by Demosthenes that the legal points
are to be treated by Ctesiphon, or have been treated by him. (2) The
dilemma is presented that either Demosthenes was not successful in giv-
ing his oration unity and finish of composition, or the unknown editor
and arranger succeeded to a wonderful degree in accomplishing a task
which the orator himself gave up as impracticable.

The oration is further compared with the de F. LZ. of Demosthenes,
and it appears that that oration is less compact and consistent in its struc-
ture than the de Corona. ,

From a comparison with the de F. ZL. of Aeschines, which holds the
same relation to the speech of Demosthenes as a rejoinder which the de
Corona holds to the speech of Aeschines against Ctesiphon, it appears
that Aeschines pursued the same course as we believe Demosthenes to
have done with regard to the subsequent insertions into the body of his
speech, in direct reply to the language of his opponent.

Several passages from the de F. LZ. of Aeschines are examined, to show
that in its present form this speech is a later recension of the speech
delivered at the trial, bearing all the marks of original unity and harmony
of composition, in spite of later insertions, and presents therefore a fit
parallel to the de Corona in its genesis and structure. A comparison with
the Timocratea shows the wide difference between a revised speech like
the de Corona, cast originally in a single mould, and a production like the
Timocratea, which is plainly not a simple recension of what was originally
one speech, but a combination of two or three distinct drafts, character-
ized by contradictions, by repetitions, and by unevenness of finish.

American Philological Association. 27

Professor L. R. Packard, of Yale College, New Haven, Conn.,
read a paper “On Geddes’ ‘Problem of the Homeric Poems.’”

The main difference between this theery and Grote’s, with which it
coincides in its division.of the Dliad, is that Mr. Geddes regards the non-
Achillean portions of the Iliad as written by the same poet who composed
the Odyssey. This poet he regards as properly entitled to the name
Homer and the traditions therewith connected, and as an Ionic Greek of
Asia Minor. The poet of the Achilleid, on the other hand, he regards as
earlier in time and belonging to the Dorian stock of Thessaly.,; The
theory seems open to several objections. 1. It assumes the connection in
time and authorship of the various Ulyssean portions of the Iliad with
one another and with the whole Odyssey. 2. If it is answered that this
is not an assumption but a conclusion based on proofs, it is to be observed
that in many cases the proofs are found in a single book or in two or three
books only, whereas the peculiarity shown by them is ascribed to the
whole Achillean or Ulyssean portion of the poem. 8. This theory seems
to ignore too much the influence of the poet’s subject upon his choice of
words and his representations of life and character. A number of the
differences insisted on are probably due to the change from the compara-
tive barbarism of war to the civilization of peace. 4. This theory, like
that of Grote, fails to take due account of the inconsistencies that run
through the whole texture of the poem. It emphasizes some contradic-
tions, but ignores others of no less importance.

Professor C. H. Toy read the last paper of the evening. It was
upon “ Expressions of Modal Ideas in Shemitic.”

I. Command: 1. Imperativ; 2. Jezma-form; 8. Imperfectinan. (The
Imperativ probably collateral with Imperfect, and not derived from it.)
II. Wish: 1. Perfect (in Arab., Eth., Phen., any wish; in Heb., a fulfilled
wish); 2. Imperfect in an, 3. In Heb. and Aram., Imperf.; 4. In Assyr.,
a special form, the Precativ. III. Determination of will: Imperf. in an
(Voluntativ). IV. Purpose: Imperfect (in Arab., the a-form; in Eth., the
shortened form). VY. General result or limit: Arab., Imperf. in @ and
Perf.; Eth., both forms of Imperf. VI. Object sentences: 1. Perf. in
Arab. and Heb. (real act); 2. Imperf., Arab. a-form after verbs of wishing,
etc., u or a-form after verbs of supposing, etc., Eth. short form. VII.
Conditional: 1. Ideal: Arab. Jezma-form in protasis and apodosis, or Im-
perf. in an (emphatic); Heb., Imperf. in both clauses; Eth., usually Perf.
in protasis, Perf. or Imperf. in apodosis. 2. Present existing fact: Arab.,
Perf. in protasis, Jezma-form in apodosis (or Imperf. in u); Eth., as in the
Ideal; Heb., Imperf. (or, sequence-construction); Aram., Imperf. or Parti-
ciple. 3. Real: usually Perf. or Partcp. 4. Unreal: the same.

Thus the functions of the several forms are: 1. Perfect: pure Opt., and
ideal, real, and unreal condition; 2. Imperf. in uw. Opt., and exhibition
of an act as merely an object of thought, whether condition or result; 3.
Imperf. in a. dependence—ideal result, purpose, limit or sequence; 4.
Jezma-form: Opt., command, condition; 5. Imperf. inan - emphatic wish,
command, condition; 6. Imper.: affirmativ command.

28 Proceedings of the

Examination of these uses shows: (1) that the modal senses belong to
both Perf. and Imperf., and (2) that they flow from the completional (Perf.)
or inchoativ (Imperf.) force of the twoforms. The particular employment is
determined by usage, but the orfginal sense is never lost sight of, and must
always be considered in interpretation, each dialect being studied for itself
under the guidance of the facts obtained by general comparison.

As to the historical development: It may be inferred from the usages of
the various dialects that the mode-distinctions existed in primitiv Shemitic.
The tendency has been to drop these distinctions, indeed to compress all
senses flowing from the inchoativ into the Jezma or shortest form (leaving,
however, the Imperativ unaffected), as is evident from an examination
of the different Shemitic languages, ending with modern Arabic, which
stands in this respect about on the same plane with Hebrew and Aramaic.

The Association then adjourned to Thursday morning.

Newrort, Thursday, Ju.y 17, 1879.
MorRNING SEsSION.

The Association resumed its session at 9.15 a. .

The minutes of the proceedings of Tuesday and Wednesday were
read and accepted.

In the absence of Professor Crowell, Professor March was ap-
pointed to fill his place in the committee on nominations.

The Secretary announced the election of four new members:

Mr. Winfred R. Martin, Jersey City High School, Jersey City, N. J.;
Miss Mary H. Ladd, Chauncy-Hall School, Boston, Mass.; Dr. B. Perrin,
Hartford High School, Hartford, Conn.; Mr. Eben Alexander, East Tén-
nessee University, Knoxville, Tenn.

The committee on the time and place of meeting recommended
that the next session be held at Philadelphia, Pa., July 13, 1880;
but left the matter open and subject to modification by the Execu-
tive Committee.

The Society then listened toa paper by Professor Albert S. Cook,
of the Johns Hopkins University, Baltimore, Md., entitled “ Studies
in the Heliand.”

Since the investigations undertaken by Grein, Windisch, and Sievers to
determine the sources of the Heliand, and the proportions in which they
were utilized in its composition, the poem is understood to rest upon the
basis of the Pseudo-Tatianic Gospel Harmony, supplemented by the com-
mentaries of Hraban, Alcuin, and Bede.

This paper has for its object:

I. To separate those expansions or embellishments that have originated
with the poet, from the body of the composition and particularly from the
merely periphrastic portions to which they are adjoined or with which
they are incorporated.

American Philological Association. 29

II. To collect and classify those passages in which the author displays
his full originality, ¢. ¢., is entirely independent of his sources.

These passages fall under six heads:

1. Those that motive a statement or event which otherwise would
want an explanation.

2. Those that present an issue or sequence, not expressly warranted by
the sources, but contained in them by implication. |

3. Those that mark transition, having of themselves no distinctive char-
acter.

4, Passages descriptive of Old Saxon manners, morals, or religion, or
indicative of current theological views.

5. Didactic and moral generalizations.

6. Poetic ornaments, often assuming the shape of poetic formulae.

III. To call attention to a few of the more remarkable syntactical pe-
culiarities of the poem.

The committee to nominate officers for the year 1879-80 pre-
sented nominations as follows:

For President—Professor Crawford H. Toy, Norfolk, Va.

For Vice-Presidents—President William C. Cattell, Lafayette College,
Easton, Pa.; Professor Lewis R. Packard, Yale College, New Haven,
Conn.

For Secretary and Curator—Professor Charles R. Lanman, Johns Hop-
kins University, Baltimore, Md.

For 7reasurer—Charles J. Buckingham, Esq., Poughkeepsie, N. Y.

For additional members of the Hrecutive Committee—

Professor Basil L. Gildersleeve, Johns Hopkins University, Balti-
more, Md.

Professor William W. Goodwin, Harvard University, Cambridge, Mass.

Professor Milton W. Humphreys, Vanderbilt University, Nashville,
Tenn.

Dr. J. Hammond Trumbull, Hartford, Conn.

Professor William D. Whitney, Yale College, New Haven, Conn.

The report was accepted, and the persons therein named were
declared elected to the offices to which they were respectively nom-
inated.

Professor F. A. March, chairman of the Committee on the
Reform of English Spelling, appointed in 1875, and continued in
1876, 1877, and 1878, reported:

Dht Cemmittee haz net been calld on diring tht last year for any
official action. Dht Philolegical Society in England, which it waz theght
might appeint a Committee ef Cenfereng on tht subject, hae net dun
so. We can however report pregress in tht reform. ‘Dhe Memorial to
Cepgress in favor ef the appeintment ef a Commission to examin and
report on tht referm, and on tht expediency ef moving the Guvernment
ef Great Britain to finjte in a jeint Commission, which wae prepared by

80 | Proceedings of the

membere ef this Committee, hae been sijgnd by reprezentative ef more
than fifty Cellegee and Universitiee. It hae also led to similar memoriale
from Dht American Instittite ef Instruction, and many uther ttachere’
associations. TDhte Department of Public Instruction of tht city of Chicago
finanimously rezelvd to cerrespend with uther Boarde on tht subject.
Great reading ef papere and discussion ef them went on all over tht coun-
try at tht winter mtetinge of tht associatione. Action in favor of reform
hae been taken by the State Ttachere’ Assogiatione of Pennsylvania,
New Jereey, New York, Massachtisetts, Ohjo, Illinois, Towa, Michigan,
Wiscensin, Missouri, Virginia. Action hae also been taken by tht State
Legislatiiree of Cennecticut, Pennsylvania, Wisconsin, lowa. Dhet report
ef the Cemmittee ef the Legislattire of Wisconsin in favor of putting a
fonetic dictionary in tht public schoole hae been printed, and ie an eble
argiment fer the reform. Many uther reform papere hav been printed
in tht journale, magazinee, transactions ef lerned socjetiee, and, within a
few ytare, in books ljke thoeze ef Hadley, Whitney, Miller, Ellis. Vol-
timeez devoted to fonetics and reform hav apptard from Mr. Swéet, sum-
time Prezident ef .the Philolegical Soviety ef London, and Mr. J. H.
Gladstone; pamflets also from Prof. J. L. Johnson of the University of
Mississippi, Prof. L. H. Carpenter ef tht University ef Wiscensin, Prof.
Edward Nerth ef Hamilton Cellege. Dhe Spelling Referm Association
ie to hold its anniial méeting this year at Philadelphia under tht most
favorable auspicee ae a Department of the National Ediicational Asso-
giation. Max Miller, Dr. Murray, Prezident ef the Philolegical Sogijety
ef Epgiand, and ex-Preeidents Mr. Sweet and Dr. Morris, with Rev.
W. W. Sketat, Prof. ef Anglo-Saxon in tht University ef Cambridge, and
Rev. A. H. Sayce, Prof. of Philelogy in the University of Oxford, hav
censented to act ae Vice-Prezidents.

In Epgiand it will be rememberd that in 1876 tht National Union ‘of
Elementary Ttachere, repreeenting sum 10,000 teachere in England and
Welee, past almost iinanimously a reeolition in favor ef a reyal Com-
mission en Spelling Referm. More than a hundred and thirty School
Boarde, incliding thoee ef London, Liverpool, and Birmingham inijted
in the rezolition, and the movement led to a cenferen¢g in London, at
which Prof. Saycse ef Oxford preejded, and Dr. Murrey, Mr. Sweet, Dr.
Merris, Mr. Ellis, Mr. J. H. Gladstone, Sir Charlee Reed, and mauy uther
dignitariee ef Church and Stete took part in the discussione. Cd cem-
mittee from the Cenfereng waited on the Lord Preeident ef the Council
in the Department of Edication with a strive of reeoliitione. An organi-
zation ef theee reformerez for permanent prosecition ef tht reform hae
been fermd, with a Secretary and plenty ef uther effigere, and they will
publish an organ.

Dht National Assogietion ef Great Britain fer the Promotion of Social
Scjeng, before whom tht subject wae broght by Prof. Newman at their last
Copgress, after long deliberation by a Committee, haz a report before it in
favor of an alternativ spelling for scientific purposee, and fer teaching, and
to guide tht progress of reform. Dr. Murrey, Prezident ef the Philoleg-
ical Sogjety, who hae undertaken tht editorship of tht Socjety’e Historical

American Philological Association. 31

Dictionary, wishee to hav tht pronunciation ef that wurk ina key alfabet,
which mey bt an agrved fonetic alfabet a2 ie propozed by tht Social Scj-
eng Association. It ie a time of rapid progress, aud it seeme not unlikely
that sum occasion mey arjze diring tht next year when a Committee of
the Association may be needed. Perhaps it may bt wiee to continie tht
Committee anuther year.

The report was accepted, and the committee continued for another
year.

The Auditing Committee reported that the accounts of the
Treasurer had been compared with the vouchers and found correct.
The report was accepted.

The Society then returned to the regular order of the day, and
listened to a paper by Professor B. L. Gildersleeve, of the Johns
Hopkins University, Baltimore, Md., ‘On the Encroachments of
wf upon od in Later Greek.”

Every one who has read much Greek of the post-classic period must
have noticed that the negative 4 is used in various relations in which it
would not be employed so readily, if at all, in model prose. It is easy
enough to charge all such variations to the account of a gradual breaking
down of the language, and indeed the whole matter may be learnedly des-
patched by calling these misuses specimens of the so-called soloecismus Ala-
bandicus. But such corruptions do not come in without cause. If the
appreciation of the negatives was indeed so much enfeebled, we should
expect the two to be interchanged pell-mell, whereas it is “4 that has
encroached on ov and ov has troubled “4 comparatively little.

Asa slight contribution to the history of these encroachments, I have
examined anew the usage of one of the best of later Greek writers, with a
view to employing the results thus gained as categories for further investi-
gation. This author is Lucian, who was a careful student of Attic Greek,
and in his Soloecista notices not only such gross blunders as d¢eAov duvgcy
but such pardonable lapses as ovv#owv av, so that it could hardly have been
absolute heedlessness of the earlier usage; and, indeed, we find him every
now and then reverting to the classic norm. The explanation is to be sought
in the popular speech of the time. Lucian, man of the world as he was,
avoided all affectation and followed the drift of the spoken language, so far
as it was not rude or solecistic.

The classic differences between ov and yf are here assumed as well °
known, even if not sufficiently well formulated. Between these two nega-
tives there is a certain border land which in the classic period was occa-
sionally invaded by uf. It is just this border land on which 4% has
squatted so resolutely in the post-classic time. In less figurative language,
the later use of #47 is not so much an innovation as an extension ; and the
following seem to be the lines of intrusion :

First, #7 with the infinitive of oratio obligua. The natural negative of
the infinitive as such is uf; and it was not until the infinitive had begun to
represent the indicative that the negative ot could have been tolerated. But

82 Proceedings of the

this toleration was established before our record, and we can now only
guess at the primal state before the incoming of the future infinitive,
which marks unmistakably a new function of the infinitive, just as the
incoming of the future optative marks a new function of the optative.
Still there is a group of verbs of saying and thinking, which retain the old
negative. Such are verbs of asseveration and belief, such verbs as ouvina:,
paprupeiv, moretery, mercioSac and the like. Cf. Il. 9, 182-3; Od. 5, 179;
Hdt. 1, 165; 2,179; Ar. Vesp. 1047, 1281; Andoc. 1, 90; Lycurg. 76;
Dem. 21, 119, etc. So paprvpd ph, Dem. 45, 15, cf. 40, 47; meoretw uy,
Andoc. 1, 2; Dem. 21, 221; xéro:Sa wh, Pind. Ol. 1, 104; xérecopas ph,
Plat. Apol. 37 A. Occasionally ¢dévaz and Aéyecv, occasionally olecda: and
voutCey join the ranks of these verbs, which involve the will, where the
utterance strives to make the statement good and the thought is at once a
wish ; although it must be observed that grammarians have not always
been careful to distinguish the legitimate use of “4 with the infinitive in
apposition, from this extended use of #4 with the infinitive.

Now it is evident that this form of expression carries with it the emphasis
of the witness on oath, so to speak, the emphasis of desire, and hence the
tendency to use it in the later time, which always leans to the impressive.
Mf with the infinitive is equivalent to ‘‘I swear,” ‘‘I vow,” ‘‘I bet,”
instead of quieter forms. How common this oratio obliqua u7 is in Lucian
is known to every reader of the Pantagruelist of Samosata, as George
Saintsbury has happily called him.

Again in clauses with 47, “4 is sometimes found in classic times, a
phenomenon due to the influence of the leading verb. So when the lead-
ing verb is a verb of swearing, as in the well-known passage : ovd’ oudoat
Xp?) Tous’ Ore uh Tote Tpaypa Té6c° Eotat, Theogn. 659 (cf. Il. 10, 329), or an
imper., as in the famous instance Antiphon 5, 21. But such deviations are
so rare that we must not insist on them as possible misleaders. We must
rather connect the 6re u4 in declarative sentences with the use of #7
with the infinitive in oratzo obliqua. It is clear that in a period when p74
could be used freely after a verb of saying this form 67: “7 would suggest
a convenient equivalent for an oratio obligua expression, especially after a
principal tense from which the oratio obliqua optative is excluded. It
were indeed worth inquiry whether this form 6r¢ 47 with ind. did not help
to thrust out érc ov with the optative. At all events we find the opt. form
of oratio obligua becoming rarer and rarer.

Another important extension is to be noticed in the relative sentence.
Even in classic times the negative of a relative clause is “4 when the rela-
tive gives the notion of characteristic, and as the characteristic sometimes
gives a ground, the clause with “#7 seems to be causal outright. Here the
subjective element represented by u would appear in standard Latin as
the subjunctive. Causal relatives then began to take “4, and with causal
relatives adversative relatives, which are thus fused with concessive rela-
tives, and this is extended to the integral parts of the relative sentence.
And not only so, but we must further take in the equivalent of the rela-
tive, the participle. So, often in later Greek, where we should expect the
negative ov with a participle, we find the negative “4, which is a phenom-

American Philological Association. 38

enon analogous to the familiar Latin combination in which qué with
subjunctive is used as 4 parallel for a characteristic adjective.

The widest divergency from classic usage has been touched on already,
the use of causal particles such as é7eé (4r:) and the like with the negative
#4. This variation is due sometimes to the oratio obliqua element just
recognized, and sometimes to the relative characteristic, which elements run
into each other. So that even here we have a development of doctrine
rather than a bold and bad heresy.

Remarks were made upon this paper by Professor W. W. Good-
win, Professor M. W. Humphreys, and Mr. S, P. Andrews.

Professor Tracy Peck, of Cornell University, Ithaca, N. Y..,
then read a paper “On the Authorship of the Dialogus de
Oratoribus.”

After briefly sketching the history of this discussion since its rise early
in the sixteenth century, the speaker examined the external and internal
bearings of the subject and reached the conclusion that Tacitus must have
written the work.

A passage in a letter of the younger Pliny to Tacitus (IX. 10), was cited
and shown to be in all probability a quotation from the Dialogus, and thus
conclusive in regard to its authorship.

. It was seen that the manuscripts of the Dialogus testify only in favor of
Tacitus, Changes in, or additions to the MS. title of the work were shown
to be of late origin, and with the evident purpose of supporting theories of
a non-Tacitean authorship. Pomponio Leto’s quotation from the work, as
a work of Tacitus, was given as significant of the belief in the fifteenth
century.

From a discussion of the probable time of composition of the Dialogus
it appeared that there are nu chronological objections to Tacitus as the
writer.

A strong personal motive that may be detected in the Dialogus—the
justification of the withdrawal of the leading interlocutor (Maternus) from
forensic pursuits, and devotion to literature—was claimed to be in harmony
with a natural impulse of Tacitus to vindicate his own change in literary
work.

Attention was Called to a like conception of life and to a consistency of
judgment of men and events in the Dialogus and in the admitted writings
of Tacitus.

The literary style of the Dialogus was admitted to be widely different
from that of the latest compositions of Tacitus. But it was shown that
. Tacitus was, in his earlier years, an advocate of the revived Ciceronianism
of the Dialogus, and it was maintained that the difference in subject and
motive as well as in the age and experiences of the writer would sufficiently
account for the great dissimilarity or development of style.

Professor W. W. Goodwin, of Harvard University, Cambridge,
9)

84 Proceedings of the

Mass., read a paper “On Greek Verbs which add Epsilon to the
Stem in certain Tenses not belonging to the Present System.”

The only question raised here is a practical one of classification, and
there is no thought of discussing the origin of the phenomenon. George
Curtius, in his Grammar and in his treatise on the Greek Verb, includes
under one class (the E class) those verbs which add e to the stem in the
present, like dox-£-w, and those which have no e in the present but add ¢ in
some or all of the other tenses, like BotAoua:, BovAfooua: (BovAe-), parOave
(ua8-), pabjoopar (uabe-). It is obvious that, although these two phenomena
may depend on the same principle, the latter cannot be made the basis of
a classification which proposes to show the relation of the present to the
simple stem of the verb; for va#e- has nothing to do with the formation of
pavéavo- from yuad-, and the stems in ¢ have no claim to the title of ‘‘ simple
stems.” Hadley saw clearly this weak point in the E class of Curtius, and
rightly excluded from this class all verbs except those which have e in
the present (like doxé-w), and which alone can be said to form the present by
the addition of e to the simple stem. But he does not introduce under a
single head the far more numerous class of verbs which have e in other
tenses than the present. He gives under the ‘‘ First Class” a list of such
of these verbs as belong to this class; but, as he does not do the same in
the other classes, the impression is given that the phenomenon in question
is in some way specially connected with class I., although this false
impression is guarded against by a remark under 331. Now the verbs of
other classes which take « are nearly as numerous as those of the first
class. Further, examples of this formation are found in every one of the
eight classes of Curtius (except of course the seventh), and it cannot help
confusing any classification to introduce into it a peculiarity which is
common to all the classes already marked off on other grounds, Curtius
includes in the second division of his seventh class only those verbs which
would otherwise belong to the first, and those which form the stem ine
from the present stem. As, now, no verb can belong to this division
which does not belong also to some one of the other classes of Curtius, it
is plain that it can add nothing to this classification to bring the division
into it at all.

I have therefore thought it better to class the addition of ¢ to the stem
in other tenses than the present among the many modifications of the
stem which are peculiar to certain tenses of verbs, like the lengthening of
the stem-vowel in ridw, rivfow, the insertion of o in reréAe-o-nas, and the
change of ¢ to o in orépyw, éoropya, and to a in oréAAu, oraAxa.

Remarks were made upon this paper by the President, Mr.
Sewall.

The next paper, on the ‘Nomenclature of Early California,” by
Mr. E. L. Williams of Santa Cruz, California, was read by Prof.
F. A. March. :

Namee of placee in California ar generally henorary, scriptiral, or
descriptiv Spanish namee, or original Indian namee. Honorary namee

American Philological Association. 35

ar few. Only three wer mentiond: Mendocino, a cape, named in henor
ef the Viceroy of New Spain, under whom it wae discuverd, in 1542:
Monterey, also in honor ef a Viceroy, in 1602; Branetforte, also fer the
Viceroy, 1797.

Scriptiral namee ar very commun. Erly missionariee introdtict the
custom ef ising them, and it hae been fellowd. QM large number wer
mentiond. Descriptiv name2 wer mentiond in still larger numbere, and
many of them expleind. Sum of tht descriptive, which appear little
grafic now, wer verified a2 having been sumtime fitting. TDhoze explaind
ar Spanish.

Into tht Indian namee tht peper did net go at lepgth.

The last paper was by Mr. Albert S. Gatschet, Linguist of Pro-
fessor J. W. Powell’s United States Bureau of Ethnology, Wash-
ington, D.C. It was “On Syllabic Reduplication as observed in
Indian Languages, and in the Klamath Language of South-western
Oregon in particular.”

The author of this paper had peculiar facilities for studying the redupli-
cative feature of Western Indian Languages on his trip made in 1877 to
Oregon. In no other linguistic family of the West studied by him does
this mode of grammatical synthesis hold a more prominent place than in
the Klamath language, which is also spoken by the Modoc tribe. But redu-
plication is common to all languages of the world, and a more profound
knowledge of all the facts relating to it must prompt scholars to distin-
guish between two different kinds of reduplication: (a) the sterative, used
mainly in the formation and derivation of words, and (0) the distributive,
used for inflectional purposes. Instances were given from several of the
Indian tongues, in which either one of the two kinds or both have been
observed. Both kinds are very prominent in the Malay-Polynesian dia-
lects, and Steinthal has given a very lucid exposition of duplications
observed in Dayak. Fr. Miller has done the same for the Eastern and
Western Malay-Polynesian dialects. In America we find distributive redu-
plication in the Flathead-Selish of Montana, in Klamath, and in the Santa
Barbara dialects of Southern California. Some instances can be traced in
Kalapuya_and Algonkin,dialects, and it seems to be or to have been gen-
eral in the dialects of the Nahuatl stock (Aztec, Tarahumara, O’ pata, Pima,
etc.). Jterative reduplication exists in all the above languages, in the
Maya, Yuma, and Shoshoni linguistic families, and in the Wayiletpu,
Pomo, and Mutsun (Olamentke dialect). Many Sahaptin dialects, like
the Warm Spring and Nez-Percé, use it to form diminutive nouns and
_ adjectives describing the surface-quality of objects of nature. As stated
above, the Klamath language is prominent in the use of both kinds of
syllabic reduplication, as will appear from the following particulars:

In Klamath the iterative reduplication repeats the entire radical syllable
without vocalic or consonantal changes, a few instances excepted. Redu-
plication of the first two syllables of the word in an iterative sense is
observed only in words (nouns and verbs) commencing with the sounds

36 | Proceedings of the

k, i, n, u (or ou, w), and some of them show a diphthong of an adulterine
character, as te-wkté-wkeh, ‘long-tailed chicken-hawk.’ Terms formed by
iterative reduplication also assume the distributive form. We find it in
onomatopoetic terms: yatyaua, ‘to be noisy,’ wekwékash, ‘magpie;’ in
frequentative and usitative terms: nidshonidshua, ‘to make grimaces,’
tughtishla, ‘to shiver,’ cf. Latin ttwbare; in adjectives of colors: mets-
métslt, ‘sky-blue, purple;’ in adjectives describing surface-quality: l4k-
lakls, ‘polished, smooth;’ in adjectives describing external shape or form:
Kéixolst, ‘round, spherical, annular, cylindrical;’ in adjectives marking
intensity: Mtchltichli, ‘strong, powerful.’

The distributive reduplication in Klamath doubles the first syllable or the
first two syllables, but does not extend beyond the vowel which is in-
cluded in the doubling process. A prefix may be reduplicated as well as
the radical syllable; and if its vowel is short, and the reduplication mono-
syllabic, the vowel of the second syllable will be a. This grammatical
synthesis prevails throughout the whole language; for not only verbs and
nouns, but even most particles are subject to it. The idea of severalty or
distribution is expressed by it: thus, ktékna, ‘to cut a hole into one object,’ or
‘to cut holes into many articles by one single cut;’ distributive, kektdkna,
‘to cut holes into different, separate objects by cuts repeated at different
times, for every object separately.’ Vép means ‘hand, hands, the hand, the
hands, a hand;’ nénap, ‘each of the two hands, the hands of each person
considered as an individual distinct from any other individual.’ This will
suffice to show that the distributive form in Klamath has nothing to do
with what we call plural. No plural exists in that language as a regular
form.

Different modes of distributive reduplication exist, all of which are
dependent upon the phonetic laws of the language. Thus in monosyllabic
reduplication we find seven different modes. (1) Regular doubling, with
-a- in the second syllable: thus, apltplt, ‘dusky;’ distr., ttaptipls, ttpa,
‘to lay down;’ distr., ¢-dtpa. (2) Duplication with syncope of a. télak,
‘waistcoat;’ distr., tédak, instead of tétalak. (8) Duplication without vo-
calic change: tmz, ‘grouse;’ distr., tnétmu, lékanka, ‘to go astray;’ distr.,
lolo’kanka. (4) Duplication of diphthongal radicals: teéni, ‘ recent;’ distr.,
telint; tuexa, ‘to perforate;’ distr., tuétoya; ydki, ‘seed-basket;’ distr.,
y4-ikt. (5) Duplication with vowel inverted: tchuaish, ‘buzzard;’ distr.,
tchdtchutsh, instead of tchi-tcha-ish; puélya, ‘to throw down;’ distr.,
pepuélya, instead of pupd-éiza. (6) Duplication with elision of consonant:
tléxo, ‘brain;’ distr., Wtlyo; tmékil, ‘ green lizard;’ distr., ié¢mkiw. (7) Du-
plication with apocope of verbal suffix: kshéna, ‘to carry on the arms;’

distr., Xshéksha, instead of kshékshana.

' Dissyllatic distributive reduplication is subject to the same phonetic
laws as monosyllabic duplication. The following examples illustrate this
fact. Thus udélgatko, ‘checkered,’ has for its distributive, wde-uddlgatko;
utchin, ‘to fish with a net,’ utcht-utchdn,; uddma, ‘to cover a vase,’ udd-
udma,; udumichna, ‘to swim on the surface,’ udiddmichna; uldyue, ‘to
scatter,’ ula-ultwe; kawakdga, ‘to rip up with the teeth,’ kawakaukdga,;
ibéna, ‘to dig,’ ipépa, tbépa.

American Philological Association. 87

Both forms, the absolute as well as the distributive, go through all the
declensional and conjugational inflections of the noun and verb.

On motion, it was

Resolved, That the thanks of the Association are due and are hereby
tendered to Colonel John R. Leslie for his unwearied efforts for the comfort
of the members, and to the Master and Trustees of the Rogers High
School for the use of their building.

On motion, the Association then adjourned.

ogical Association.

American Philol

38

sensunop Susppny | ulENAVO Wf (peu3ig)

WVIWHdH 1 “V

‘BLOT ‘OT Aine “TY ‘twodMaN

‘suodnoo prsdun pus onp-10A0 UOA0S TIT ‘PROITIH YY 1018944 IND1O9UUOD 943 JO puog 9q) pauTmMExea A||vuOsi0d osTe
GAB DAA “3001109 9q 03 OMe 043 AJT]I00 OM ‘SIOYONOA 04} YIM Woy} polvdu0d pue ‘syUNOD08 GAOgS oq} PoulMIVxe Zuyavy

aro ‘9[Q109T[09 yUaseId 48 JOU ‘WOVE OG'ZTS JO ‘oures oy} JO suodNod onp-10A0 U2AOB OIA ‘SIBTIOG
paipuny oearg toy ‘Aueduiog pvolprey uisqs9 A, JRONIOMUODH 9q} JO puog oOUO ‘JOINSBOI], 94} JO SPpUBY 94} Ul O8[¥ SI O10T J,

‘soinsvedT ‘NVHONIXONG ‘f SATHVHO a a
C9 TL68 C9 TL6$
core |° ; , ' ‘AInSvel], Ul suse
00°8T . . “ ‘SUOTOBSUBL], " SuyNQWsIp ,, »
0¢'T ae ‘Suyupd ,,
CSS ° ; . . . ‘ZeINSBOL], jo sosuodx7y
8198 ‘sadeisod QS OF ’ ‘\sa19)U]
pus ‘Axons Suyuud Joy A1¥}0100g ,, - 00 'FE ’ * ‘guonsorjqnd jo saleg
STST ° “eZoyereg 48 UOIsses Jo sosuadxiT 00 SEP ‘peaqeoal QOUIS BJUSUIsSsessB PUB 800.7
1g90e8 | ° * BUOMIVSUBL], ,, PUB , SS3UTpesoolg », SUYULIG LS 1973 ‘g/9T ‘6 Aine ‘Ainsvaly, Ul aousyeq
LA) a@®

6L8T ‘YT Ampe-8.si ‘6 fine
‘NOILVIOOSSY ‘IVOINOIONIHG NVOIMENY 947 Y71M yUNODID ur ‘sounsDaL]

‘AVHONINONG ‘f SaIUVHO

OFFICERS OF THE ASSOCIATION.
1879-80.

PRESIDENT.
CRAWFORD H. TOY.

VICE-PRESIDENTS.

WILLIAM C. CATTELL,
LEWIS R. PACKARD.

SECRETARY AND CURATOR.
CHARLES R. LANMAN.

TREASURER.

CHARLES J. BUCKINGHAM.

EXECUTIVE COMMITTEE,

The officers above named, and—
BASIL L. GILDERSLEEVE,
WILLIAM W. GOODWIN,
MILTON W. HUMPHREYS,
J. HAMMOND TRUMBULL,
WILLIAM D. WHITNEY.

MEMBERS OF THE AMERICAN PHILOLOGICAL ASSOCIATION.

Eben Alexander, East Tennessee University, Knoxville, Tenn.
Frederic D. Allen, Yale College, New Haven, Conn.

William F. Allen, University of Wisconsin, Madison, Wis.
Joseph Anderson, Waterbury, Conn.

Joseph R. Anderson, Jr., Richmond, Va.

N. L. Andrews, Madison University, Hamilton, N. Y.
Stephen P. Andrews, New York, N. Y. .
William H. Appleton, Swarthmore College, Swarthmore, Pa.
Albert N. Arnold, Baptist Theological Seminary, Chicago, Ill.
Cecil F. P. Bancroft, Phillips Academy, Andover, Mass.
Stephen G. Barnes, Iowa College, Grinnell, Iowa.

Charles C. Bates, Plymouth, Mass.

H. Louis Baugher, Pennsylvania College, Gettysburg, Pa. —
8. E. W. Becker, San José, Cal.

J. B. Bittinger, Sewickly, Pa.

Porter C. Bliss, office of ‘‘ Library Table,” New York, N. Y.
James R. Boise, University of Chicago, Chicago, IIl.

Charles E. Brandt, Farmington, Conn.

H. C. G. Brandt, Johns Hopkins University, Baltimore, Md.
Fisk P. Brewer, Iowa College, Grinnell, Iowa.

John A. Broadus, Southern Baptist Theol. Seminary, Louisville, Ky.
Charles J. Buckingham, Poughkeepsie, N. Y.

L. H. Buckingham, English High School, Boston, Mass.
Henry F. Burton, University of Rochester, Rochester, N. Y.
Richard E. Call, Mohawk, N. Y.

W. B. Carr, Leesburgh, Loudoun Co., Va.

Franklin Carter, Yale College, New Haven, Conn.

William C, Cattell, Lafayette College, Easton, Pa.

Talbot W. Chambers, 70 West Thirty-sixth street, New York, N. Y.
Henry L. Chapman, Bowdoin College, Brunswick, Me.

Elie Charlier (Life Member), Central Park, New York, N. Y.
Francis J. Child, Harvard University, Cambridge, Mass.
Lyman Coleman, Lafayette College, Easton, Pa.

George F. Comfort, Syracuse University, Syracuse, N. Y.
Albert 8. Cook, Johns Hopkins University, Baltimore, Md.
Jacob Cooper, New Brunswick, N. J.

A. Crittenden, Packer Collegiate Institute, Brooklyn, N. Y.
Howard Crosby, University of New York, New York, N. Y.
Edward P. Crowell, Amherst College, Amherst, Mass.

Edwin De Merritte, Chauncy-Hall School, Boston, Mass.

American Philological Association. 41

Mrs. M. W. De Munn, Providence, R. I.

Schele de Vere, University of Virginia, Va.

Mary C, Dickinson, 173 Madison street, Toledo, O.

Martin L. D’Ooge, Michigan University, Ann Arbor, Mich.

T. T. Eaton, Petersburg, Va.

B. H. Engbers, Mt. 8t. Mary’s Seminary, Cincinnati, O.
George R. Entler, Franklin, N. Y.

Carl W. Ernst, Providence, R. I.

Ambrose J. Faust, Washington, D. C.

O. M. Fernald, Williams College, Williamstown, Mass.
Gustavus Fischer, Rutgers College, New Brunswick, N. J.

M. M. Fisher, University of the State of Missouri, Columbia, Mo.
Alexander Fleischmann, Cazenovia Seminary, Cazenovia, N. Y.
H. H. Furness, 222 West Washington Square, Philadelphia, Pa.
James M. Garnett, St. John’s College, Annapolis, Md.

B. L. Gildersleeve, Johns Hopkins University, Baltimore, Md.
D. C. Gilman, Johns Hopkins University, Baltimore, Md.
William W. Goodwin, Harvard University, Cambridge, Mass.
Richard T. Greener, Howard University, Washington, D. C.
James B. Greenough, Harvard University, Cambridge, Mass.
Ephraim W. Gurney, Harvard University, Cambridge, Mass.

8. S. Haldeman (University of Pennsylvania), Chickies, Pa.
Albert Harkness, Brown University, Providence, R. I.

William R. Harper, Baptist Theological Seminary, Chicago, II.
Samuel Hart, Trinity College, Hartford, Conn.

B. J. Hawthorne, State Agricultural College, Corvallis, Oregon.
Charles R. Hemphill, Theological Seininary, Columbia, S. C.
Theophilus Heness, New Haven, Conn.

Thomas Wentworth Higginson, Cambridge, Mass.

George O. Holbrooke, Trinity College, Hartford, Conn.

E. 8. Holden, U. 8. Naval Observatory, Washington, D. C.
Selah Howell, Christian Biblical Institute, Stanfordville, N. Y.
Milton W. Humphreys, Vanderbilt University, Nashville, Tenn.
John T. Huntington, Trinity College, Hartford, Conn.

George W. Ingraham, Sing Sing, N. Y.

Edwin E. Johnson, Trinity College, Hartford, Conn.

Asahel C. Kendrick, University of Rochester, Rochester, N. Y.
Robert P. Keep, Williston Seminary, Easthampton, Mass.

John B. Kieffer, Mercersburg College, Mercersburg, Pa.

D. B. King, Lafayette College, Easton, Pa.

Louis Kistler, Northwestern University, Evanston, II.

Miss Mary H. Ladd, Chauncy-Hall School, Boston, Mass.
Charles R. Lanman, Johns Hopkins University, Baltimore, Md.
Albert G. Lawrence, Newport, R. I.

R. F. Leighton, 109 Lefferts Place, Brooklyn, N. Y.

John M. Leonard, University of Missouri, Columbia. Mo.

John R. Leslie, Newport, R. I.

William S. Liscomb. Providence, R. I.

6

MEMBERS OF THE AMERICAN PHILOLOGICAL ASSOCIATION.

Eben Alexander, East Tennessee University, Knoxville, Tenn.
Frederic D. Allen, Yale College, New Haven, Conn.

William F. Allen, University of Wisconsin, Madison, Wis.

Joseph Anderson, Waterbury, Conn.

Joseph R. Anderson, Jr., Richmond, Va.

N. L. Andrews, Madison University, Hamilton, N. Y.

Stephen P. Andrews, New York, N. Y. |

William H. Appleton, Swarthmore College, Swarthmore, Pa.

Albert N. Arnold, Baptist Theological Seminary, Chicago, Ill.

Cecil F. P. Bancroft, Phillips Academy, Andover, Mass.

Stephen G. Barnes, Iowa College, Grinnell, Iowa.

Charles C. Bates, Plymouth, Mass.

H. Louis Baugher, Pennsylvania College, Gettysburg, Pa.

8. E. W. Becker, San José, Cal.

J. B. Bittinger, Sewickly, Pa.

Porter C. Bliss, office of ‘‘ Library Table,” New York, N. Y.

James R. Boise, University of Chicago, Chicago, III.

Charles E. Brandt, Farmington, Conn.

H. C. G. Brandt, Johns Hopkins University, Baltimore, Md.

Fisk P. Brewer, Iowa College, Grinnell, Iowa.

John A. Broadus, Southern Baptist Theol. Seminary, Louisville, Ky.
Charles J. Buckingham, Poughkeepsie, N. Y.

L. H. Buckingham, English High School, Boston, Mass.

_ Henry F. Burton, University of Rochester, Rochester, N. Y.
Richard E. Call, Mohawk, N. Y.

W. B. Carr, Leesburgh, Loudoun Co., Va.

Franklin Carter, Yale College, New Haven, Conn.

William C. Cattell, Lafayette College, Easton, Pa.

Talbot W. Chambers, 70 West Thirty-sixth street, New York, N. Y.
Henry L. Chapman, Bowdoin College, Brunswick, Me.

Elie Charlier (Life Member), Central Park, New York, N. Y.
Francis J. Child, Harvard University, Cambridge, Mass.
Lyman Coleman, Lafayette College, Easton, Pa.

George F. Comfort, Syracuse University, Syracuse, N. Y.
Albert 8. Cook, Johns Hopkins University, Baltimore, Md.
Jacob Cooper, New Brunswick, N. J.

A. Crittenden, Packer Collegiate Institute, Brooklyn, N. Y.
Howard Crosby, University of New York, New York, N. a
Edward P. Crowell, Amherst College, Amherst, Mass. -
Edwin De Merritte, Chauncy-Hall School, Boston, Mas

42 American Philological Association.

Thomas R. Lounsbury, Yale College, New Haven, Conn.

Rebecca S. Lowrey, 162 West Forty-seventh street, New York, N. Y.
Merrick Lyon, University Grammar School, Providence, R. I.

W. G. McCabe, Petersburg, Va.

Joseph H. McDaniels, Hobart College, Geneva, N. Y.

Francis A. March, Lafayette College, Easton, Pa.

D. 8. Martin, Rutgers Female College, New York, N. Y.

Winfred R. Martin, 20 East Twenty-eighth street, New York, N. Y.
R. H. Mather, Amherst College, Amherst, Mass.

Charles M. Mead, Andover Theological Seminary, Andover, Mass.
John Meigs, Lafayette College, Easton, Pa.

Augustus C. Merriam, Columbia College, New York, N. Y.

Charles D. Morris, Johns Hopkins University, Baltimore, Md.

F. A. Muhlenberg, Univ. of Penn. (4307 Walnut street), Philadelphia, Pa.
Wilfred H. Munro,’ Bristol, R. I.

Joseph H. Myers, Milton-on-Hudson, N. Y.

C. K. Nelson, St. John’s College, Annapolis, Md.

Edward North, Hamilton College, Clinton, N. Y.

Howard Osgood, Rochester Theological Seminary, Rochester, N. Y.
Charles P. Otis, Massachusetts Institute of Technology, Boston, Mass.
W. B. Owen, Lafayette College, Easton, Pa.

Lewis R. Packard, Yale College, New Haven, Conn.

William A. Packard, College of New Jersey, Princeton, N. J.

E. G. Parsons, Sumner Academy, South Bytield, Mass.

Tracy Peck, Cornell University, Ithaca, N. Y.

William T. Peck, High School, -Providence, R. I.

Bernadotte Perrin, High School, Hartford, Conn.

William C. Poland, Brown University, Providence, R. I.

Noah Porter, Yale College, New Haven, Conn.

Samuel Porter, National Deaf-Mute College, Washington, D. C.
John W. Powell, Washington, D. C.

DeWitt T. Reiley, Rutgers College, New Brunswick, N. J.

William A. Reynolds, Wilmington, Del.

Leonard W. Richardson, Trinity College, Hartford, Conn.

W. G. Richardson, Austin College, Sherman, Texas.

Ezekiel G. Robinson, Brown University, Providence, I. I.
Lawrence Rust, Kenyon College, Gambier, O.

Julius Sachs, Classical School, 649 Madison avenue, New York, N. Y.
George W. Samson, Rutgers Female College, New York, N. Y.

A. Duncan Savage, Metropolitan Museum of Art, New York, N. Y.
Wesley C. Sawyer, Lawrence University, Appleton, Wis.

Philip Schaff, Union Theological Seminary, New York, N. Y.
Henry Schliemann, Athens, Greece,

Walter Q. Scott, Wooster University, Wooster, O.

Jotham B. Sewall, Thayer Academy, Braintree, Mass.

Thomas D. Seymour, Western Reserve College, Hudson, O.

Joseph Alden Shaw, Highland Military Academy, Worcester, Mass.
L. A. Sherman, Hopkins Grammar School, New Haven, Conn.

American Philological Association. 43

Charles Short, Columbia College, New York, N. Y.

E. G. Sihler, Classical School, 649 Madison avenue, New York, N. Y.
E. Snyder, Illinois Industrial University, Champaign, Il.

Ephraim G. Squier, 185 East Thirty-ninth street, New York, N. Y.
A. B. Stark, Logan Female College, Russellville, Ky.

Benjamin F. Stem, Classical Institute, Easton, Pa.

Frederick Stengel, School of Mines, Columbia College, New York, N. Y.
William A. Stevens, Rochester Theological Seminary, Rochester, N. Y.
Edward F. Stewart, Easton, Pa.

Austin Stickney, Florence, Italy.

John Tetlow, Girls’ Latin School, Boston, Mass.

J. Henry Thayer, Andover Theological Seminary, Andover, Mass.
William E. Thompson, Genesee Wesleyan Seminary, Lima, N. Y.
Crawford H. Toy, 43 West Twenty-fourth street, New York, N. Y.
J. Hammond Trumbull, Hartford, Conn.

Joseph A. Turner, Hollins Institute, Botetourt Springs, Va.

Milton Valentine, Pennsylvanig College, Gettysburg, Pa.

James C. Van Benschoten, Wesleyan University, Middletown, Conn.
Addison Van Name, Yale College, New Haven, Conn.

C. Osborne Ward, 486 Adelphi street, Brooklyn, N. Y.

Minton Warren, Johns Hopkins University, Baltimore, Md.

W. B. Webster, Military Institute, Norfolk, Va.

R. F. Weidner, 1330 Franklin street, Philadelphia, Pa.

James C. Welling, Columbian University, Washington, D. C.

J. B. Weston, Antioch College, Yellow Springs, O.

Mrs. A. E. Weston, Antioch College, Yellow Springs, O.

Albert 8. Wheeler, Sheffield Scientific School, New Haven, Conn.
Benjamin I, Wheeler, Brown University, Providence, R. I.

John Williams White, Harvard College, Cambridge, Mass.

W. H. Whitsitt, Southern Baptist Theological Seminary, Louisville, Ky.
William D. Whitney, Yale College, New Haven, Conn.

J. Colver Wightman (Life Member), Taunton, Mass.

Alonzo Williams, Brown University, Providence, R. I.

Edwin H. Wilson, Yale College, New Haven, Conn.

oo —_

The Executive Committee herewith announce that the Twelfth
Annual Session of the Association will be held at Philadelphia,
Pa., beginning Tuesday, July 13, 1880, at 3 o'clock p. m.

Members intending to read papers at the next session of the
Association are requested to notify the Secretary at as early a date
as possible.

amo RE OS ARS
a

14 DAY USE

RETURN TO DESK FROM WHICH BORROWEDE

LOAN DEPT.

RENEWALS ONLY—TEL. NO. 642-3405
This book is due on the last date stamped below, or
on the date to which renewed.
Renewed books are subject to immediate recall

DA
INTERLIBRARY LOAN

MAY 8 1976

' General Library
LD21A-60m-6,'69 University of California
(J9096s10)476—-A-32 Berkeley

GENERAL LIBRARY - U.C. BERKELEY

p00056952%