The first six books of Homer's Iliad online

. (page 5 of 48)
Online LibraryHomerThe first six books of Homer's Iliad → online text (page 5 of 48)
Font size
QR-code for this ebook


21 a. I10MEK1C SYNTAX xlix

order of the words in the translation or by the tone of voice in
reading. To translate pa </.s- f/v/.s- natural (or even you, see or you
knoiv) or ye at least, often throws upon the particle very dispropor-
tionate emphasis. The student can most easily and clearly appre-
ciate the force of a particle by comparing a number of examples
which have become familiar to him ; he will then see the impor-
tance of these particles to the character and tone of a speech or of
the narrative.

/?. T is used far more freely than in Attic prose. A single re is
often used to connect single notions, as KWCO-O-IV | olwourl TC A 4 f.

y. 6(f>pa is the usual particle to introduce a final clause.

b. Interrogative Particles, a. The general interrogative particle
in Homer is rj, but in a double question (where the Attic Greek
uses -rrorepov . . . >/) -rj or rjt stands in the first member, ^ or ?c in the
second ; cf. A 190 ff.

/?. When 5 introduces a single question, it is rarely used as in
Attic, as a mere interrogation point. It regularly implies emotion
of some kind, as A 133, 203.

21. a. Parataxis. The Homeric language is far less distinct
than the Latin or the English in the expression of logical relations,
and gives less prominence to the logical forms of syntax ; but it is
seldom difficult to appreciate the ancient idiom if an attempt is
made to find the Homeric point of view.

The Homeric poems contain many survivals of the simplest form
of sentences. In the earliest stage of the Greek language, clauses
were not combined with each other as secondary and principal ;
they were simply added one to the other. To use the technical terms,
mil i'<l unit inn <>r parataxis (TrapoYa^is) was the rule, not subordina-
tion or hypotaxis (VTTOTO&S). Originally the relatives were demon-
stratives, and relative sentences have been called ' parenthetic
demonstrative sentences.' Thus 8c was used in the apodosis of
relative and conditional sentences. This was especially frequent
when the relative or conditional clause preceded, as ct 8c KC /x^
&MKrtr, e 'yw 3 e ' KCl / at>ro5 \o>/xat A 137 but if they shall not give it,
(but) then I myself shall take, etc., ctos 6 ravff cop/uuve . . . rj\6f 8'
A 193 f. while he was pondering thii . . . (but) then Athena

iwl in



came, OL-TJ Trep <f>vXXwv yeveiy, rovrj Se /cat di/8p<ov Z 146 as is the race of
leaves, (but) even such is also the race of men. So avrdp and dAAa are
used with stronger emphasis than 8e, as ei 8e < Kaprepos eo-o-t, 0ea 8e
(re yuvaro /Arjrrjo, | dAA' 6'8e cfreprepos ecrriv CTTCI TrAeoVeo-o-tv avacrcrei A 280 1.
#w if thou art mighty and a goddess is thy mother, (but) yet, etc.,
where the apodosis is really contrasted with the protasis ; cf. A 81,
quoted in the next paragraph.

b. Compare with the foregoing the use of /cat in the conclusion
of relative sentences, to mark the connection of the clauses. Thus
also re was freely used in subordinate clauses, as os *e 0eots eVwreitfiyrai
fjidXa r" 1/cAvov avrov A 218 whoever obeys the gods, (and) himself the
gods readily hear; and re ... re is found in both protasis and
apodosis, marking their correlation, as et irep yap re ^oAov . . . Ko.ro.-
vr&f/r), | dAAa re Kat />ieTO7ri<r0ev e^et KOTOV A 81 f . for even if he should
restrain his wrath, (but) yet even hereafter, etc.

c. The first part of a paratactic sentence may introduce the
cause or reason for what follows, as in Andromache's words to
Hector, "Exrop arap crv /xoi eoro-t Trarrjp KCU Trorvio. fjn^rrjp, | lySe KO.CTI-
yvryros, crv Se JJLOL OaXepos 7rapaKOtrrj<s' | dAA' aye vvv eAeai/ae Kat avrov
pifjiv 7rt 7rvpyu> Z 429 ff. but thou, Hector, art my father, etc.,
which implies " Hector, since thou art my all."

d. Correlative Constructions. The Greek language was always
fond of a parallel or antithetic construction, a contrast, a balance,
where the English subordinates one thought to the other ; but the
adversative relation, where the English idiom would use a subordi-
nate clause introduced by for, although, when, while, or since, is
more frequent in these poems than in later Greek, as dAAa 7ri'0eo-0'-

a/A<a> Se vetorepw ecrrov e/xeto A 259, <vAAa ra /ACV r' oVe/uos ^a/aaSt? X^ t >
aAAa 8e G* vXrj rrjXeOowcra c^vet, eapo? 8' eTrtyt'yverat CO/DT; Z 147 f. . . .
when the season of spring comes on, ^/xev 8>J TTOT' e/xev TTCI/OOS
/x,voto ... -^8' Tt /cat vw />tot ro8' fTriKprjrjVov eeA8w/3 A 453 ff.
didst hear my former prayer so now also fulfill this my desire.

e. avrap also is used where a causal particle would be used in

English, as oSuV/^crt TreTrap/xei/os, avrap GIOTTO? | to/xw Ivl cm(3ap<S rjXi]\.aro
E 399 f . thrilled with pains since the arrow was fixed in his stout



f. In these contrasted clauses, av, avre, avrdp, dra/o, aXXa, as well as
Sc, may be used in correlation with /u.eV. And Kat TOTC and apa (pa,
ap) are used as well as Se to mark the apodosis.

g. A copulative conjunction is sometimes used where the English
uses a disjunctive or, as Tpi-n-Xfj rerpaTrXrj TC A 128 threefold or (and)
fourfold, in which prominence is given to the second member. Cf.
cva Kat ovo B 346, xOi^Oi TC Kat 7rp<oia B 303, Tpt^a re Kat rerpa^Od
T 363, t <!-'/ tt i- ijutitfi'que beati Verg. Aen. i. 94.

h. The Homeric poet sometimes puts into an independent clause
the incidental thought which in later Greek would be expressed

regularly by a participle, as Xaoi 8' rjprja-avro 0eoTs t xetpas dvca-^ov
T 318 the people prayed to the gods with uplifted hands (lit. and
I iff >'<1 tltt'ir lunnlii) ; for ^etpas ava<r\6vT<;, cf. /xcyaA.' ev^ero, ^apa? di/a-
A 450; (Zeus /xep/x^/oi^e ws 'A^tA^a) TLfj.y<Tr], 6\(rrj 8e TroAeas CTTI
i/ B 4 was pondering how he might honor Achilles by
(/t'sf /<>// i ny (lit. timl (/rsfroy), etc.; cf. ri\6e <epa>v, l^wv A 12 f. ; aAA'
aKeovcra K<i6r)<TO, /xa> 8' 7ri7rei'$eo /xv^a> A 565 for c/u-w Tret^o/xei/r; /jivOw.

i. Conversely, the participle, as in later Greek, often contains

the principal idea, as KaTc'i/cvo^v | "IXiov eKTreptrai/r' cvrc^eoy aTroveecr^at
B 113 promised that I should wk Hins, and return, but in the very
next verse is the English idiom, vvv 8 KOLK^V a-jrarr^v /3ov\cvo-aro, Kat
/xe KeAeva | Svo-KXe'a * Apyo? tKeVflai B 1 14 f . planned an evil deceit and
bids me go, etc., for aTraTrjv /?ovAevcras.

22. a


22. a. The dialect of the Homeric poems is in one sense arti-
ficial : it was spoken at no place and at no time. But it is not a
mosaic composed of words and forms chosen capriciously from the
different Greek dialects ; it is a product of natural growth. The
poets retained many old words and forms which had disappeared
from the spoken language (cf. hath, loved, I ween, etc., in English
verse), and unconsciously excluded all that was not adapted to
dactylic verse ; but they did no violence to their language ; they
did not wantonly change metrical quantities, nor did they intro-
duce new grammatical terminations, nor violate syntactic usage.

duce ne'


b. The student must always remember that the Homeric dialect
was not a modification of the Attic dialect, and was not derived
from it, but that it represents an older stage of the language.
Many Attic forms can be derived from the Homeric forms. Thus
Homer uses the longer forms of the dative plural, as /coiXflo-iv [/cot-
Xais] A 26, otooj/ouri [oiwi/ot?, 35 d ] A 5 ; of the genitive singular
of the second declension, as n/ata/xoto [ Ilpia^ov, 35 a] A 19; and
of the infinitive endings, as xoXoxre/Acv [xoXwo-eiv, 44 /] A 78,
So/xevai [Sowai] A 98 ; and uncontracted forms generally, as oXyea
[aXy?7, 24] A 2, ereXetero [ereXetro] A 5, 'ArpeiS??? A 7. Even
where a shorter form is used, as lo-rav [ecmyo-av] A 535, erapcov
[cratpcov] A 349, this is not to be regarded as shortened from the
Attic form, but as nearer the original.

c. The Homeric dialect is essentially Ionic and seems to have
been developed among the lonians of Asia Minor, influenced pos-
sibly by the speech and certainly far more by the old poems of their
Aeolian neighbors. The oldest form of Greek epic songs seems
to have been Aeolic, but the lonians brought epic poetry to per-
fection. Even the Pythian priestess delivered the oracles of
Apollo in epic verse and Ionic dialect, and the Dorian Spartans
sang about their camp fires the Ionian songs of Tyrtaeus.

d. Some forms seem to be borrowed from other dialects ; but the
student must remember that when the poems were Composed, the
difference between the dialects was less than at the earliest period
when we have monumental evidence concerning these.

e. The conservation of old forms together with the introduction
of new forms was very convenient for the verse ; e.g. for the infini-
tive of the verb to be, Homer could use l/x/xei/at as dactyl, _ w w ;
l/xevat as anapaest, w w _ ; l/x/xev as trochee, _ w ; l^v as pyrrhic,

w \j ctvat as spondee, Naturally, the choice being offered,

metrical convenience determined which of these forms should be
used. No difference in meaning exists between KpoviW, son of
Cronus, and Foovi&js. Metrical convenience often or generally
decides between the use of 'AX<HOI or 'A/DyeToi. If prominence is
to be given to the name of the Greeks, at the beginning of the
verse, 'Apyctoi must be used. On the other hand, the verse can


close with 'A^atoi, but not with Aaraoi or 'Apyeiot, with 'Apyei'oiorv,
but not with 'A^atoio-tv, with *A^atoiv, but not with Aarawv or
'ApyeiW. Vergil also uses Argi, Achivi, Danai, Dorici, and
Pelasgi as synonymous.

f. Synonyms and stock epithets or phrases, also, are used
according to the poet's convenience. ai/a avSpw 'Aya/xe/xi/<ov is used
after the feminine caesura ( 58 /) of the third foot, but tvpv /cpeiW
'Aya/xe/xv<oj/, 'Aya/xe/xvovo? 'ArpeiSao, Ol* 'Aya/xe/xvora Troi/xera Aaaiv after

the masculine caesura of the same foot. n^A^iaSew 'A^A^os is used
after the masculine caesura of the third foot ( 58 e), TroSas WKVS
'AxiAAev's after the masculine caesura of the fourth foot ( 58 ^),

but TroSapKT/s 8tos 'A^tXAevs, TroSwKeos AiaKi&ao or d/xv/xovos Aia/a'Sao,
on, 7ro8(o/cca II^Aeuura, d/xv/xora II^AeiWa, Or 'A^tAAiya
after the feminine caesura of the third foot, with Stos
a tag when the verse is filled up to the bucolic diaeresis

( 58 h). Cf. the epithets of Apollo, e/caroto A 385 w w _ w ,

Kr)(36\ov A 14 w w w , eKacpyos A 479 w w w , Ka.Tr)(36\ov

A 370 w w w w , fKaT-qflfXtTao A 75 w w w w w . See 12 &.

g. Some anomalies of form (as of verse) are as yet unexplained,
but the assumption is justified that all which remain either (1)
were supported by the usage of the people and might be explained
by more complete knowledge of the history of the language, or
(2) followed the analogy of what was in use, or (3) are errors
which have found their way into the text during the course of
transmission to the present time. As the poems were handed
down among the Greeks at first orally, and afterwards still uncrit-
ically for centuries, errors unavoidably crept in, and when the older
forms were unprotected by the meter, the obsolete forms were
gradually assimilated to (or replaced by) what was later and more


is regularly used for d, as ayoprj, o/xooy, VT/VS, except in
0eo, yor/f/tts-,s, Aaos, prople, some proper names (as AiVcias), and where
a consonant has been lost, as /Jas, /a>x<k. Occasionally, as B 370,


fjidv is found instead of the less frequent prjv (the strong form of
/xeV). dA.ro A 532 (from dAAo/xcu) is another instance of d, unless it
is to be written dAro. a remains when it is the product of con-
traction or ' compensative lengthening/ as 6pa, Trao-a?. (H. 30 D.)

b. The final d of the stem is retained in the genitive endings -do
and -dn/ of the first declension, as 'ATpei'Sao A 203.

c. do and -rjo often change to ecu, with transfer of quantity :
'Arpei'Sao, 'Arpei'Seco. Cf. /JaxnA^os with Attic /foo-iAetos, iepvja and
tepea. But the frequent Ado? never has the Attic form AedW.

d. Compensative lengthening is sometimes found where it is not
in Attic, as e/os (^eV/ros), eiW/ca (Lesbian |j/i/e/ca), Kovp-rj (/cop^ra),
ynovi/os [/xoVos], ovpos (op/:os\ Soupos [Sopv], icros (^ricr/ros).

e. Diphthongs occasionally preserve t where it is lost in Attic
before a vowel : aiei', CUCTOS, ereAet'ero ( 47 y), veiKeieovce, oAoir;, Trvoiy.

f . But i is lost before a vowel in oWa (wKeTa) *Ipts B 786, in -oo for
-oto as genitive ending of the second declension ( 35 &), and in ep:eo
for e/xeto, e^. ; cf. xpuo-euns A 246 with XP VO " V A 15. As in Attic,
the penult is sometimes short in DIG? (as A 489, A 473). In these
cases i has turned into y. Likewise v is sometimes dropped between
two vowels, becoming w. See 59 k 8.

24. Contraction. Concurrent vowels generally remain uncon-
tracted : de'/cw, oAyea, Wt? (in nominative and vocative singular),
ots (o/ris = ovisj ewe). Attic cv is regularly ev before two conso-
nants, and the adjective is always evs or rjvs. Patronymics from
nouns in -evs form -efSqs, -etwv, as 'ArpeiSrys A 7, IIi;\eici)va A 197.
These uncontracted vowels were originally separated by a conso-
nant. (H. 37 D ; G. 846.)

25. Synizesis. a. Vowels which do not form a true diphthong
may be blended in pronunciation into one long sound, for example,
'ArpetSeco w w , OeoeiSea. F 27, 877 OVTCOS A 131, Sr) avre A 340,
TToAto? B 811, 'loruuav B 537, o-xerXir) T 414, in. which t must have
had very nearly the pronunciation of its cognate semivowel y.
The genitives of the first declension in -ecu, -ew are always pro-
nounced with synizesis. (H. 42 D ; G. 47.)

b. Synizesis often served the purpose of the later contraction.
yfjictav did not differ in metrical quantity from

28b. V<)Wi:i> AND VoWKL CHANGES lv

26. C ft f tn's is not frequent. Xote roweica (TOV /eKa) A 291,

wvVds E 396, XW& B 238 ( Kal 4/**)> Ta>AAa A 465 ( Ta " AAa )'
(H. 76; G. 42 if.)

27. ///'//"x (H. 75 D; G. 34) is allowed:

a. After the vowels t and v, as lyx^' o^voort E 50.

b. When the two vowels between which it occurs are separated
by a caesura (KU^O-TO eViyva/xi/'cura A 569) or by a diaeresis ( 58 h) :
seldom after the first foot (avVap 6 cyj/w A 333), more frequently
after the fourth foot (yx ea ovoWa E 568). Hiatus between the
short syllables of the third foot is allowed nearly as frequently as
in all other places together, more than two hundred times. This
freedom of hiatus emphasizes the prominence of this caesura
( 58 d).

c. When the final vowel of the first word is long and stands in
the accented part of the foot ( 57 a), as ro> o-e KUKYJ alo-y A 418.
See 59 1: e.

d. When a long vowel or diphthong loses part of its quantity
before the following vowel ( 59 k), as TT)I/ S' eyw ov Xvo-w A 29,
/XT; v\> rot ov x/Wo-/^ A 28. The final and initial vowels may be
said to be blended in the first example, while in the second the
final letter may have been pronounced as y. This is called weak
or improper hiatus ; it is essentially the same as the following.

e. When the last vowel of the first word is already elided, as
fjivpC 'A^diol? oAye' ZOrjKev A 2.

N. I'. Hiatus before words which formerly began with a con-
sonant ( 32) is only apparent.

The poet did not avoid two or more concurrent vowels in the
sain.- word ( 24).

28. Wtxinn. (H. 79 ; G. 48.) a. a (in inflectional endings and
in apa and pa), c, t, o may be elided, at is sometimes elided in the
v-rb endings. 01 is elided seven times in /xot, three times in rot,
once in <rot A 170 (unless oi'Sc o-ot ota> or ov <rot oi'u> should be read
there for ovSe </ <KO>).

b. TO, Trpd, dm', Trcpi, rt', and the conjunction on do not suffer
elision. oY is for ore (cithrr the temporal conjunction or the rela-
tive o with TC affixed ; 42 q), T' for rt or rot.


c. i is seldom elided in the dative singular, where it may origi-
nally have been long.

d. Oxytone prepositions and conjunctions lose their accent in
elision; other oxy tones throw the acute accent upon the preceding
syllable, as TO. KOLK [Ka/ca] A 107.

N.B. Elision is not left to the reader, as in Latin poetry.

29. Apocope. (H. 84 D ; G-. 53.) a. Before a consonant the
short final vowel of apa and of the prepositions avd, Kara, irapd may
be cut off (cbroKOTn?, aTroKOTTToo). The accent is then thrown back
upon the preceding syllable (although it might be more rational to
consider it lost, as it is in elision).

b. After apocope, the v of dvd and r of Kara follow the usual
rules for consonant changes : d/ATreTraXwv T 355, a/x TreSiov E 87,
Ka/3/?aXev E 343 (/care/JaXcv), KaS 8e (Kara 8e) frequently, KOLKTOLVC
Z 164 (Kara/crave), KaTTTreo-e-nyv E 560, Kappe'ovcra E 424, KaXXiTre Z 223

c. avepva-av A 459 is explained as derived by apocope, assimila-
tion, and vocalization of /:, from avd and ptpvu. Cf. 32 h.

d. Apocope was no mere metrical license ; it was common in the
conversational idiom of some dialects. More striking examples of
apocope and assimilation than any in Homer are found in prose


30. a. Where collateral forms appear, one with single and the
other with doubled consonants, the form with two consonants
is generally the older, or justified etymologically, as TTOO-O-I, TTOO-I
(from 7ToS-(rt) ; vci/ceo-cre, vei'KCfre (vet/cos, vetKe<r-), OTTTTW? (OK/TOOS, cf. Latin
quis, etc.) } OTTI, KT\.

b. Single initial consonants, especially X, /A, v, p, <r, are often
doubled (as p is in Attic) when by inflection or composition a short
vowel is brought before them (see 59 A), as eXXtWero Z 45, !XXa/2e

c. But sometimes p is not doubled where it would be in Attic, as

E 598, Ko,T/oeev A 361.


d. Palatal and lingual mutes often remain unchanged before /x,

e. Lingual mutes are commonly assimilated to a following o-, as
(7ro8-<ri). <r is sometimes assimilated to /x or v : cti/xevai
for eoyxevcu, dpyevvds, white, for dpyetr-vos, as dpyevvdw F 198,
for /reorv/xt ( 32 a), fptfitwrj E 659 dar&, c/. "Epe/3os.

f . o- is frequently retained before cr, as Wo/x<u from the stem
er-, cTcA.co-0-e (r/. reXos from the stem reAeo - ), /2e\co-o-tv, from the
stem /SeAeo-.

g. Between /x and X or p, /? is sometimes developed, as a/x/Sporos
from stem /xpo or /xop (murder, Latin mors, morior), while in /?pords,
mortal, the /x of the stem is lost ; /xc/x/JXwKe A 11 from /xXo or ttoA
(c/ 1 . etioAoi/) ; i^/x^poTes, aorist of d/xaprayu). (T/*. the 8 of dvSpos and
the # in English chamber (camera).

h. KdfjifiaXe E 343 is found occasionally in the Mss. as a variant
reading, a softer pronunciation for Ka/3/?aA.e ( 29 b).

L A parasitic T appears in TTTOA.IS, TrroAe/xos for TroAis, TroAe-
/u.09. (7/ 1 . 8ix0d, TP^OL with Attic St^a, rpt'xa. The proper names
Neoptolemus (NcoTrrdXctto?) and Ptolemy (IlToAc/xatos) preserved this
T to a late period.

j. The rough breathing (h) has no power to prevent elision or
weaken hiatus. The smooth breathing is found with several words
which have the rough breathing in Attic, as <J)x/xe
[^tiepa], ttAro (from dAAotuu), ^eXtos [ijAios], 'AtSTys

k. The v movable was written by some ancient critics after the
ending - of the pluperfect, as /3e/3A.7J/cav E 661, ^i/diyai/ Z 170 ; cf.
iJo-Kciv T 388 (^O-KCCV), >dpeiv (impf. of <j>opo>) A 137. It is freely
used before consonants to make a syllable long by position ( 59/).

1. The final o- of adverbs is omitted more often than in prose.
Not merely and CK, OVTOJ? and ovr<o, but also TTWS and TTW, TroXXaKi?
and TroAAaKi, d/x^t'? and d/x<^t (adverbial), are found as collateral forms.

31. Jllefitf/it'ftiti of a and p is frequent (H. 64; G. 64): KapSoj
I'. l.^L', KpaOLr) a 353 ; Kaprio-TOt A 266, Kparo? A 509, KpaTraOos B 676,
and KapTra^os. Cf. rpaTTcib/xcv F 441 from repTrw, TcpTTiKcpawo? from

For the shifting of quantity from -do and -rjo to -eo>, see 23 c.


32, The Digamma. (H. 72 D.) a. The following words seem
to have been pronounced by the Homeric poet more or less con-
sistently with initial diganiina (consonantal u, van, p, pronounced
as English w) :

dyj/u/u, break, aXts, enough, dAwvat, am captured, dva, king, dvSdveo,
please, d/ocuos, thin, d/>vos, lamb, do-, city, e, ov, ot, /WTW, ec., with a
possessive pronoun o?, 77, 6V (eo? KrA.), cap, spring, 28ra, wedding gifts,
&vo<s, tribe, UKOCTL, twenty, CIKCD, yield, cipco, say (future epe'a>), e/cds, /ar,
Kao"ros, eac/?., eKvpos, father-in-law, e/ov, willing, lASo/uu, desire, eAtWco,
wind, eATTO/xat, hope, ei/vv/xi (/recr-vr/xt), clothe, I<r0r)<s, fl^ara, clothes, CTTOS,
word, tpyov, cpSa), work, fpvu, draw, eWe/oas (vesper), evening, %, six,
ITOS, year, lrr;s, companion, fj8v<s, sweet (dvSdj/w, please), rj&os, haunt,
rjpa, favor, 16.^, cry aloud, te/xat, desire, strive, tSetv, see (and otSa, et8o?),
iKeAo?, /i'/ce, lotKa, am Zi^e, ts, strength, sinew, l<f>i, mightily, to-os, equal
(cf. cfto-os), irvs, felly, and irer; (withe), willow, OLKOS, house, oti/o?,

b. Probably *IAtos, *Ipts, and several other words also were pro-
nounced with initial /r.

c. dvScivw, c, cKvpos, *,, rjOea, and others seem to have begun
originally with two consonants, cr/r.

d. In more than two thousand cases 'apparent hiatus 7 ( 27 /)
is caused by the omission of initial p. Less frequently a /r must
be supplied in order to make an apparently short syllable long by
< position 7 ( 59 j).

e. The verse alone affords no sufficient test for the former exist-
ence of p in any word ; it only indicates the loss of some conso-
nant. This is not conclusive evidence for p, since tr and y were
also lost. Which consonant originally was present has to be
learned in each case from inscriptions of other Greek dialects,
from a few notes of ancient grammarians, and from other cognate
languages; cf. Zpyov with work, otvos with wine, ouco? with vicus and
wich (in Norwich), ZTTOS and cty with vox.

f. The sound of p evidently was going out of use in the Homeric
period. It is not infrequently neglected in our texts, and sometimes
this neglect seems to be due to the poet himself, but p can be
restored in many passages by minor changes. For vtov c

33 c. DECLENSION lix

A 21 it is possible to read via ptKrjfioXov, for Travreo-cri 8' dvdcrcretv
A 288 it is easy to read iraa-w Sc pa.va.a-o-f.iv, and x 6 / 30 ^ f^mfftoXav for
xepvw fKrjfioXov A 14. Perhaps fcretrat fj.v p dA.ai/e Z 167 may have
been /crewu /xe'v pf. KT\.

g. That the sound of p was still alive in the Homeric age is
shown by the accuracy of the poet in its use where comparative
philology shows that it once existed. But it had disappeared from
some words, and was often neglected in others.

h. p sometimes leaves a trace of its existence in its cognate
VOWel v : avepvo-av A 459 for appepvo-av ( 29 c), raXavpivov E 289 for
TaXd-ppLvov. So doubtless aTrov/aas A 356 for a.7ro-ppd<s.

i. Some irregularities of quantity may be explained by this
vocalization of p. Thus dTroeiTrwv T 35 may have been
pronounced nearly as aTroveurw. avia^oc finds its analogy in
ta^T/ A 456 (ycVeTOvia^).

j. A neighboring vowel sometimes seems lengthened to compen-
sate for the loss of p ( 59 c).

k. An c sometimes was prefixed to a digammated word and
remained after the p was lost, as ec'ASwp, ceucoo-t, e'epyei, .

1. Sometimes the rough breathing represents the last remnant
of a lost consonant (especially in the words which once began with
<rp, as dvScu/w KT\. ; cf. c, above), as exoiv, co-Trepos. Often the same root
varies in breathing, as av&dw and ^8v?, but 5805, ei/w/u, but eo-fl^s.

m. For the augment and reduplication of digammated verbs, see

n. For Sfci'So), S/TT/V, see 59 h.


33. Special Cast' KwUny*. (H. 217; G. 292 if.) a. The suffix
-</u(V), a remnant of an old instrumental case, added to the stem,
forms a genitive and dative in both singular and plural : dye/V^i,
in the herd, l<f>i, wit It miyht, rarity, from the ships.

b. The (old locatival) suffix -Oi is added to the stem to denote
place where: oOi [ou], where, rrjXoQi [TT/AOV], far awn if.

c. The (old ablatival) suffix -Ow is added to the stem to denote
place whence: 50tv, whence, *lorj0tv, from Ida, ovpavoOtv, from

Online LibraryHomerThe first six books of Homer's Iliad → online text (page 5 of 48)