There appears to be conclusive
evidence for including also the Culex, though I was long
prevented from examining the language of this poem by
erroneous impressions that I had at first formed respecting
the treatment of the caesura in this work.
evidence for including also the Culex, though I was long
prevented from examining the language of this poem by
erroneous impressions that I had at first formed respecting
the treatment of the caesura in this work.
Ovid - 1869 - Juvenile Works and Spondaic Period
org/access_use#pd
? Vol. li] The Juvenile Works of Ovid 157
Index, examined the language minutely and pronounced it
wholly Ovidian. It should be added that Kleemann used
only a part of the available material from the Index and the
Ovidian corpus perhaps not more than one-half but
even this, in my judgment, is far more than is strictly neces-
sary for purposes of valid proof. Kleemann's study was much
praised by the reviewers, but otherwise it has received little
attention. 14 The reasons for this neglect are not far to seek.
Many scholars had accepted without question and without
independent study the distorted views of Voss which came
to them in a slightly diluted form through the voluminous
commentary of Dissen ; others lay entrenched in fancied
security behind the barrier of the metre which Hertzberg
had so conveniently and so confidently provided. Strangely
enough, the Index to Ovid remained practically a closed book.
As soon as the Burman Index is used, the whole Messalla
Appendix is seen to be unmistakably the work of the youthful
Ovid, aetate eighteen to twenty-four. 15 There is complete
identity of vocabulary, and all the most characteristic Ovidian-
isms are in the Appendix, except only those which were de-
veloped later for the sake of the virtuosity and which are
conveniently enumerated by Eschenburg. The evidence
which is drawn merely from repeated tags and half-verses,
without distinctive peculiarity of usage, is of course incon-
clusive for an author like Ovid who borrows so freely from
contemporary poets, but the proof that rests upon the plain-
est and simplest idioms often recurring is one that scarcely
admits of doubt or question. The Lygdamus poems, com-
11 An honorable exception is Professor K. P. Harrington's scholarly edition
of the Roman elegiac poets (Selections, 1914). Professor Harrington com-
ments on the possible identity of Ovid and Lygdamus with complete candor
and with an open mind, although he does not commit himself definitely (Introd.
36).
15 In justice to myself, it is only fair to state that I reached my conclusions
as to the identity of Lygdamus and Ovid at a time when I was acquainted
neither with Gruppe's results nor with those of Kleemann and Ehrengruber,
but was obliged to rely upon the simple text of Ovid and Eichert's lexicon to
the Metamorphoses (Hanover, 1886).
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? 1-8 Robert S. Radford [1920
posed at eighteen or nineteen, contain about ninety specific
and striking Ovidianisms ; Kleemann gives about half of
these. The Panegyric Upon Messalla contains probably more
than a hundred Ovidianisms ; Ehrengruber has already called
attention to all of these, and in order to explain their occur-
rence, has propounded the ingenious theory that the Pane-
gyric was a school exercise composed in a later age by some
pupil of the rhetoricians who had access to all the works of
Ovid and pilfered most freely from them all. 16 Next, as the
metre shows, come the six little letters of Sulpicia, the kins-
woman of Messalla, in which this lady (with Ovid's sympa-
thetic assistance) undertakes the part of wooing the shy
youth, Cerinthus; they contain more than twelve Ovidian-
isms. About 21 B. C. , aetate twenty-two, Ovid composed the
five charming elegies giving in fuller form the story of the
same pair of happy lovers, Sulpicia and Cerinthus ; they
show more than forty Ovidianisms and 47. 4% of dactyls,
thus approaching closely to the proportion of the first Amores
(about 48. 5%). The exquisite imitation of Tibullus, Nulla
tuum nobis (iv, 13), that closes the collection, has ten Ovidian-
isms. Since Ovid edits the unfinished second book, the two
spondaic elegies, with Ovidian language and thought, are in
all probability his work, and can no longer be ascribed to
Tibullus either in whole or in part. The poem in honor of
Messalinus (n, 5) has, in fact, long occasioned difficulty, and
has been known for nearly a century as the ' suspected elegy. '
The other Ovidian elegy, n, 2, is evidently a continuation of
the Sulpicia group, and celebrates the birthday of the shy but
sorely smitten lover Cerinthus, who is now happily married
to Sulpicia and is therefore given his true name of Cornutus.
This poem is rightly assigned to the ' fourth book ' by Gruppe.
The conclusions which we have reached upon grounds of
language and metre are supported also by strong external
16 Op. cil. x, 71 ; n, 28; in, So, etc. This treatise of more than 700 pages
in all is a magnificent collection of material and, in spite of its technically
erroneous conclusion, is truly a masterly piece of work.
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? Vol. li] The Juvenile Works of Ovid 159
evidence. Thus in the letters of the exile Ovid refers in
unmistakable terms to the issuance of the Appendix, when
he writes that it was Messalla who first induced him to ven-
ture upon the publication of his works. 17 He mentions
expressly the epicedion which he had composed upon the
death of his patron and which was sung in the forum (Pont.
i, 7, 29 fL), but he also refers again and again to poems which
he had composed in Messalla's honor in his early youth.
Thus he writes that not even the eldest son, Messalinus, can
remember the time when he " first began to venerate Mes-
salla " (Pont, n, 3, 79 ff. ). This last statement scarcely
applies, I think, to our Panegyric, but can be more fitly
referred to an earlier eulogy which we fortunately still possess
-I mean Catalepton ix (xi), contained in the Appendix
Vergiliana, a poem which celebrates Messalla's triumph over
Aquitania, and which was therefore written in the year
27 B. C. Ribbeck, Appendix Vergil. Proleg. 12, and Gesch.
d. rom. Die/it, u, 200, has already, on stylistic and metrical
grounds, identified the author of this elegy with ' Lygdamus,'
and both Marx (Pauly-Wissowa, i, 1326, s. v. Albius) and
Schanz, op. cit. 282, p. 233, speak approvingly of this view.
I find on examination that these judgments are strongly
supported by the language of the poem, and this elegy may
therefore be confidently regarded as the earliest extant work
of Ovid, written in his seventeenth year ; in fact, in a brief
monograph and admirable commentary which was published
some years ago, but has just come into my hands at the
moment of writing (De Ovidio elegiac in Messallam auctore,
Budapest, 1909), Nemethy, I find, has already clearly per-
ceived and, in large measure, convincingly demonstrated the
Ovidian authorship of the Catalepton. Messalinus, the eldest
son, quindecemvir 19 B. C. , consul 3 B. C. . was probably born
about 38-36 B. C. , and can therefore scarcely have been more
than ten years of age when this eulogy was composed.
I have not yet spoken of the date of the second eulogy
17 Pont, u, 3, 75 ff. (addressed to Cotta, the younger son).
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? 160 Robert S. Radford [1920
which we possess, namely, the elaborate Panegyric of the
Messalla Collection. This brilliant and highly rhetorical
work is metrically more advanced than the Lygdamus elegies
and was certainly composed at a later date than these poems.
Undoubtedly, by an ingenious literary or artistic fiction, the
Panegyric itself purports to be written in the year of Messalla's
consulship, 31 B. C. , and the events of the years 30-27 B. C. ,
namely, the expeditions to Aquitania and the Orient, which
had already been expressly celebrated by Tibullus (i, 7) and
by the youthful Ovid (Catalepton, ix) , are nowhere mentioned
as having actually occurred. Special students of the Pane-
gyric, however, have long seen that the Gallic and Egyptian
campaigns are well known to the clever writer, and are most
skilfully introduced into the poem by way of prophecy (vati-
cinium ex eventu}. 18 The Panegyric was therefore composed
after the Lygdamus elegies, and the first draft of the poem
must have been drawn up about the year 23 B. C. ; it treats
chiefly the earlier career of Messalla, which had not previously
been made the subject of poetic encomium. We are not at
present fully in a position to state how Ovid was occupied
in the interval between the composition of the Lygdamus
poems and the Panegyric. It is true that H. de la Ville de
Mirmont confidently assumes (Jeunesse d'Ovide, 209) that
it was shortly after the two early marriages and about his
twentieth year that our poet visited Asia Minor and Sicily
in the suite of the poet Macer, and at first glance the Pane-
gyric also, in its present revised and perfected form, appears
to contain probable or possible references to Sicily (vss. 197,
200). We must carefully refrain, however, from drawing
hasty conclusions at present and must frankly admit that
we cannot at once determine the exact date of the year which
Ovid spent with Macer in Asia and Sicily (Pont, n, 10, 21 ft". ).
A brief word must be said, however, upon Ovid's relation
to the Ciris (the legend of Scylla and Nisus) in the Appendix
18 See Hartung, op. ell. 38 ff. ; Ehrengruber, op. cit. i, 7 ; x, 71 ; also Belling,
Albius Tibullus: Untersuchung, 205 ff.
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? Vol. li] The Juvenile Works of Ovid 161
Vergiliana. As is well known, this poem offers special peculi-
arities and difficulties. Besides being partly a translation
from Greek sources, it everywhere closely follows the manner
of Catullus and, owing to the poet's prodigious memory, in
many passages it presents almost the appearance of a cento
compiled from Catullus and Vergil. It was composed after
the publication of the complete Aeneid (19 B. C. ), the whole
of which it imitates. The agreement therefore of the Ciris
with the usual Ovidian vocabulary is not quite so close as
we find in the other juvenile works, yet it is sufficient, I
believe, easily and conclusively to establish Ovidian author-
ship, especially when we consider that, by a species of /eeW<m
and in a purely temporary stage of his art, the poet has di-
vested himself of a part at least of his usual and natural
manner. Certain it is that the situation described in the
poem suits Ovid and Ovid alone. The work is addressed
(vss. 36 and 54) to the young " Messalla," by whom Mes-
salinus is evidently meant. The writer definitely renounces
(vss. 1-2) the public career to which he had formerly devoted
himself and of which he has now grown weary, but there is
not a word in the poem to warrant the usual assumption that
the author was a man of advanced years who had reached the
age of forty-five or fifty. 19 On the contrary, the situation is
precisely that described in Trist. iv, 10, 33-40, i. e. , the author
has held certain minor offices in the cursus honorum and now
refuses to go further in the pursuit of public honors. Hence
no one who has followed the career of Ovid with genuine
interest can read the opening lines of the Ciris without some
thrill of emotion. 20 Refusing to become a candidate for the
quaestorship and the senatorial rank at the age of twenty-
19 Teuffel-Schwabe, Hist, of Rom. Lit. , Eng. trans. 230, 2, i; Ribbeck,
Gesch. d. rb'm. Diclit. n, 355; Schanz, op. cit. 241.
20 Propertius also (in, 21, 25 ff. ) finely pictures himself as a student at
Athens in the school of Plato or the garden of Epicurus, but the scene is prob-
ably an imaginary one. At a more youthful age and with less experience of
the world, Horace too visited Athens and " sought for truth amid the groves
of the Academy " (Episl. n, 2, 45).
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? 162 Robert S. Radford [1920
six, the poet whose influence upon subsequent European
literature was to be so vast, hastens to Athens, 21 the " fair
garden of Cecrops," ! 2 and the true home of the intellectual
life, in order to drink, at the fountain source, of the ever-
living waters, and to worship at the shrine of the world's
four great Teachers. 23 With the over-sanguine temperament
of youth, he even dreams of composing at some future time
a great epic upon nature and the creation that shall rival
the sublime and majestic work of Lucretius. 24 The Ciris is
to be placed then about the middle of Ovid's spondaic period ;
it precedes the first Amores and also three or four other works
belonging to the carmina iuvenalia.
40
II. Transition from Sulpicia Elegies to Amores. Spondaic
Character of First Amores
The percentage of dactyls and of dactylic beginnings which
the juvenile poems of Ovid exhibit may be seen in a summary
form from the table below. In the case of elegiac verse the
percentage is here given for the whole distich, that is, it has
been obtained by combining the first four feet of the hex-
ameter and the first two feet of the pentameter ; in the case
of epic verse, the percentage is for the first four feet of the
hexameter. In the Lygdamus elegies, since the style of the
youthful poet is still imperfectly formed and he vacillates
between two proportions, I give the six Lygdamus elegies
first as a whole and secondly as forming two groups. It will
be found that the second group (4 and 5) gives results almost
! 1 Trist. i, 2, 77 : nee peto, quas quondam petii studiosus, Athenas.
22 Etsi me vario iactatum laudis amore
Irritaque expertum fallacis praemia vulgi
Cecropius suavis exspirans hortulus auras
Florentis viridi sophiae complectitur umbra. Ciris, 1-4.
23 Plato, Aristotle, Zeno, and Epicurus; cf. Ciris, 15: sapientia . . . quat-
tuor antiquis heredibus edita censors.
24 Ciris, 12-41. The plan was later fulfilled probably in the Aetna (the
language of which I have not yet examined in all its details) and in the pro-
oemium and fifteenth book of the Metamorphoses, also perhaps in the lost
Phaenomena and in parts of the Fasti.
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? Vol. li]
The Juvenile Works of Ovid
identical with the Sulpicia letters. The figures for the Pane-
gyric and the Ciris are those of Ehrengruber (op. cit. x, 5
and 21). The Aetna, which, according to every probability,
is also the work of Ovid, is omitted from the comparison
only because the exact figures are not accessible to me at the
present writing ; its proportions are. however, not very differ-
ent from those of the Ciris.
There appears to be conclusive
evidence for including also the Culex, though I was long
prevented from examining the language of this poem by
erroneous impressions that I had at first formed respecting
the treatment of the caesura in this work. Finally, Pliny,
N. H. xxxii, 152, was wholly mistaken in his conjecture
(" fortassis ") that the Halieutica relates to the " fish of
the Black Sea " and was consequently written at the close
of the poet's life ; the schemata show conclusively that the
poem, wholly dependent as it is upon Greek books, belongs
to the Lygdamus and Sulpicia period.
THE JUVENILE WORKS OF OVID
Catalept.
TV
Halieut.
Lygdamus,
all six
Lygdamus,
four
Lygdamus,
two
Sulpicia
Letters,
IX.
62 vss.
130 hex.
elegies. 25
elegies,
i, 2, 3, 6.
elegies,
4i 5-
IV, 7-12.
290 vss.
1 60 vss.
130 vss.
40 vss.
SS. Dact. 2
53 44-8
222 42. 8
359 41-3
216 45. 0
144 36. 9
45 37-5
SS. Spond.
75 55--'
298 57. 2
5" 53. 7
264 55-o
246 63. 1
75 62. 5
I. 27 Dact.
46 71. 0
62 47. 7
1 66 57. 2
ioi 63. 1
65 50. 0
20 50. 0
I. Spond.
1 8 28. 1
68 52-3
124 42. 8
? ? 59 36-9
Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? 164 Robert S. Radford [1920
A few words of explanation may be added to the tabular
statement. Ovid began in Catalepton, ix with a proportion
of dactyls which was wholly normal in the year 27 B. C. Thus
he has 44. 8% in the whole distich, which is as good as the
second and third books of Propertius, 28 as good also as the
eighth and third elegies of Tibullus' first book. 29 Yet he has
no limitations whatever in Catalepton, ix upon polysyllabic
endings in the pentameter; on the contrary, like Catullus
and like Propertius in his earlier work, he fairly revels in
their use (50%). In the Lygdamus elegies, however, and
in the Sulpicia letters, the ambitious and aspiring youth
seeks suddenly to pass from the longer endings to the more
elegant dissyllables of Tibullus, and is evidently preoccupied
with this problem and its difficulties. Under these conditions
throughout the Lygdamus poems he wavers greatly in his
composition, and, in one hundred and seventy verses (Lygd.
4 and 5, [Tib. ] rv, 7-12), he even sinks to the proportion of
Catullus, namely, about 37% in the distich, and to only
50% of dactylic beginnings. 30 This is of course excellent
Latin elegy, but it is not the kind that was most in vogue in
24 B. C. 31 It is only in the qualified or limited sense just ex-
28 Propertius has 44. 8% and 44. 7% in the second and third books respec-
tively. My figures are taken from Hultgren, op. cit. 23, who follows the five-
book division of Propertius.
29 Tibullus has 44. 8% in the eighth elegy (Pholoe) for the whole distich
and 45% in the third elegy. It is scarcely fanciful to see in the decline of
dactyls in the third elegy an expression of Tibullus' sadness and depression of
feeling during his illness at Corcyra ; we have the same phenomenon in Lyg-
damus' fifth elegy. My figures are taken from Cartault, Le distiqitc elegiaque
chez Tibitlle, Sulpicia, Lygdamus (Paris, 191 1), 7.
30 Figures for Catullus are given by Hultgren, op. cit. 15 ff. Catullus has
usually 58% to 6o r i of dactylic beginnings, but in Carm. 69-119 (319 vss. )
he has 36. 6^ of dactyls in the distich and only 47^ of dactylic beginnings.
31 Catullus wrote some of the best Latin elegy, and the naturalness and
directness of his style is due in part to his not exceeding the proportion named.
As I have shown in my "Licensed Feet in Latin Verse," op. cit. 251-27? ,
even the best Latin poets, such as Catullus and Horace, experience some diffi-
culty in always providing one required dactyl, and therefore they occasionally
admit without metrical ambiguity in such a foot exceptional or vulgar short-
enings and even short vovels (without m) in hiatus, as Lucilius, ix, 243 Bahr. ,
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? Vol. li] The Juvenile Works of Ovid 165
plained that Ovid can be said to have ever been a disciple
of Catullus in the matter of dactylic proportions, but un-
doubtedly in his youth he paid this brief tribute to the freer
and more natural style of his great predecessor. A secondary
cause for the low proportion may perhaps be found in the sad
and almost despairing character of the two elegies.
It will be noted that Ovid advanced in the Sulpicia and
Messalinus poems which are composed in the elegiac metre
to about 47% of dactyls in the who! e distich ; the hexameter
lines alone of these poems show 46. 1% and 45. 9% respectively.
On the other hand, the hexameters of purely epic poems like
the Ciris fall back to 43. 5%, which is normal for epic verse
o 5 re cSrupto (necessary dactyl of the hexameter); Lucr. vi, 1133, natu 5 ra
cSruptum; Cat. 10, 26, istos c6mmoda; nam volo d Sar&pim (necessary
dactyl of the Phalaecean) ; Hor. Cann. in, 14, n, iam virum expertae, malS
6minatis (short vowel in hiatus, necessary dactyl of the Sapphic) ; Pers. 3, 66,
di'scitfi, 6 miseri (license of the first foot, with greatly preferred dactyl) ; Lux-
orius, 302, 4, magnum depre 5 nderg usum. If difficulty is experienced in supply-
ing one required dactyl for the Sapphic or the hexameter, it is clear that two
necessary dactyls, as in the pentameter, constitute a very exacting demand
upon the Roman language, and if the virtuosity is also insisted upon, a very
elegant but very artificial form of verse is the result. For example, for metrical
reasons, Ovid (like Tibullus) constantly uses ab arte and similar phrases drawn
from the vulgar language, with its analytical tendency, instead of the simple
ablative, and he also uses, by poetic license, to an unparalleled extent, the
simple ablative for the ablative of the agent with ab, as Her. 4, 64, capti
pare 5 ntg sor6r, ' my sister was captivated by your parent,' where neither
parenti nor a parente could enter the verse ; upon the whole subject see in part
Guttmann, Sogenanntes instrumentales ab bei Ovid, Dortmund, 1890. I was
mistaken therefore in my former discussion (p. 271) in thinking that Ovid had
perhaps actually used the spelling parente for the dative parenti, as he so often
uses mare and caeleste for mari and caelesti, and as Statius, Sih. iv, 2, 28, uses
glaucae certantia Do b ride saxa, and Propertius, v (rv), 8, 10, writes cum temere
anguino creditur o^rl manus (Xeue-Wagener, Formenlehre, i 3 , 301). The
honest 37% of Catullus and Lygdamus does not compel a resort to such un-
usual constructions, and is therefore by no means to be despised. Dissen and
Postgate (Selections from Tibullus, p. XLIII), on purely subjective grounds and
without consulting indices, lexicons, or Latin authors, have discovered that
Lygdamus is an author of " poor Latinity. " On the contrary, his Latinity is
more natural and in some respects better than that of the mature Ovid. Yet
even the latter like Euripides, an unrivalled master of the graceful and
pleasing forensic style could give most of us lessons in correct Latinity to
our great and lasting profit.
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? i66 Robert S. Radford [1920
and almost precisely the proportion of Vergil. In fact that
strange and anomalous thing, the actual preponderance of
dactyls in a Latin epic poem, could never by any possibility
arise in the more vigorous epic hexameter (versus fortis or
grams) ; it must appear first in the soft and tender strains of
the elegy (versus mollis or levis) , and may be transferred thence
to epic verse, as Ovid transferred it in the Metamorphoses
and Valerius Flaccus later employed it in imitation of Ovid.
Even Ovid in his mature period has recognized this difference
to some extent; for the hexameter of elegy in the Heroides
(first series) reaches 56. 4%, while the epic hexameter of the
Metamorphoses does not exceed 54. 8%.
It is evident then that in the first edition of the A mores
which was published in 14 B. C. , 32 only a few years after the
Sulpicia elegies, the Ciris, and the Aetna, Ovid had had little
opportunity to develop a marked dactylic virtuosity and to
become a highly artistic elegiac poet. We may be sure then
that the first Amores of 14 B. C. showed only a very moderate
advance upon the 47. 4% of the Sulpicia elegies, and they
may even have contained, like the latter, a few polysyllabic
endings in the pentameter. Up to 14 B. C. Ovid was in fact
fully as much an epic as an elegiac poet (see Am. i, i, 1-18
on his early ambition to excel in epic themes), and he was
still far from having attained in the Sulpicia poems the high
virtuosity which appears in three of the four genuine elegies
of Tibullus' second book, namely, 55% or more. 33 Two of
the short Sulpicia elegies and the equally short Cornutus
poem show, it is true, an equal or a superior number of dac-
tyls, 34 but four Sulpicia poems and the long Messalinus elegy
32 Schanz, op. cit. 270 (293), suggests that these elegies probably came
first separately before the public, but later were collected in the edition of
14 B. C. This seems quite probable, especially as the spondaic Amores show
in general the predominance of the same schemata or figures of the hexameter
as the Sulpicia and Messalinus poems.
33 The first elegy of the second book has 55. 2% in the distich, the third has
50%, the fourth 55%, and the sixth 56. 8%; see the figures of Cartault, op.
cit. 7, and Ehrengruber, op. cit. x, 5.
34 iv, 6 (20 vss. ) has 50% of dactyls ; iv, 5 (20 vss. ) 58. 3% ; n, 2 (20 vss. ) 50%.
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? Vol. li] The Juvenile Works of Ovid 167
(n, 5) have a decided preponderance of spondees. A marked
change in Ovid's whole attitude took place, however, after
14 B. C. For unlike the Ciris and the Aetna, which did not
greatly increase their author's reputation, the Amores, im-
mediately upon their first publication, achieved a prodigious
success; they at once became popular favorites, and, like
the Eclogues of Vergil, were frequently sung in the theatre
with accompanying dance (Trist. n, 519; v, 7, 25). The
favorable reception thus accorded to the elegies naturally
determined the direction of the poet'3 genius and led him to
devote himself uninterruptedly (with the exception of the
tragedy of Medea) for sixteen years (14 B. C. to 2 A. D. ) to the
perfection of an elegiac style which surpassed even the later
work of Tibullus.
Having developed a remarkable and distinctive technique
in the Heroides and the Ars, Ovid resolved to make the first
Amores conform fully to his later rules of art. The revision
was not due to the fact that this popular and successful work
was poetically immature, nor even chiefly to the desire to
add a series of new elegies ; as he himself tells us in the pre-
fixed epigram, the question was primarily one of more careful
finish and elaboration : " hoc illi praetulit auctor opus "
(cf. Gruppe, op. cit. 377). In brief, his principal purpose
was to eliminate all polysyllabic endings, to introduce every-
where the full dactylic virtuosity, to multiply the dactylic
beginnings, and to remove or greatly reduce those schemata,
such as DSSS, SDSS, 5555, 55, SD, and the like, which
were no longer fully acceptable. The view adopted by
Heuwes in his discussion of the matter (op. cit. 31) is that
the revision " consisted in here and there substituting a word
that was more suitable for one that was less suitable to the
metre and sense, or in changing the collocation of words o 1
verses, or in doing all these things at the same time. " These
conclusions are partly correct, but there can be no doubt that
both the additions and the alterations were far more extensive
and thoroughgoing than Heuwes supposes. The first edition
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? 1 68 Robert S. Radford [1920
was in fact subjected to a drastic revision and was almost
entirely rewritten in conformity to the new rules of art. 35
Nevertheless abundant traces of the character of the first
edition still remain ; for in not a few cases the poet has
triumphed over the versifier and the artist, and a number
of the original poems of the first edition have been retained
with very moderate changes. This fact may be shown very
clearly in two ways, if the writer correctly understands the
true nature of Ovidian versification. For the full and har-
moniously developed virtuosity in the Ovidian sense means
first that there shall be a considerable preponderance of
dactyls in the distich taken as a whole, and secondly that
there shall be a similar preponderance of dactyls in each
member taken singly, that is, in both the hexameter and the
pentameter lines. It is evident therefore that virtuosity is
lacking in the Amores in two ways: (i) when the spondees
are either equal or predominant in the whole distich ;
(2) when they are equal or predominant in the hexameter
line (versus fortis or diirus) alone. With this understanding
of the nature of virtuosity, I may summarize the facts briefly
as follows.
? Vol. li] The Juvenile Works of Ovid 157
Index, examined the language minutely and pronounced it
wholly Ovidian. It should be added that Kleemann used
only a part of the available material from the Index and the
Ovidian corpus perhaps not more than one-half but
even this, in my judgment, is far more than is strictly neces-
sary for purposes of valid proof. Kleemann's study was much
praised by the reviewers, but otherwise it has received little
attention. 14 The reasons for this neglect are not far to seek.
Many scholars had accepted without question and without
independent study the distorted views of Voss which came
to them in a slightly diluted form through the voluminous
commentary of Dissen ; others lay entrenched in fancied
security behind the barrier of the metre which Hertzberg
had so conveniently and so confidently provided. Strangely
enough, the Index to Ovid remained practically a closed book.
As soon as the Burman Index is used, the whole Messalla
Appendix is seen to be unmistakably the work of the youthful
Ovid, aetate eighteen to twenty-four. 15 There is complete
identity of vocabulary, and all the most characteristic Ovidian-
isms are in the Appendix, except only those which were de-
veloped later for the sake of the virtuosity and which are
conveniently enumerated by Eschenburg. The evidence
which is drawn merely from repeated tags and half-verses,
without distinctive peculiarity of usage, is of course incon-
clusive for an author like Ovid who borrows so freely from
contemporary poets, but the proof that rests upon the plain-
est and simplest idioms often recurring is one that scarcely
admits of doubt or question. The Lygdamus poems, com-
11 An honorable exception is Professor K. P. Harrington's scholarly edition
of the Roman elegiac poets (Selections, 1914). Professor Harrington com-
ments on the possible identity of Ovid and Lygdamus with complete candor
and with an open mind, although he does not commit himself definitely (Introd.
36).
15 In justice to myself, it is only fair to state that I reached my conclusions
as to the identity of Lygdamus and Ovid at a time when I was acquainted
neither with Gruppe's results nor with those of Kleemann and Ehrengruber,
but was obliged to rely upon the simple text of Ovid and Eichert's lexicon to
the Metamorphoses (Hanover, 1886).
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? 1-8 Robert S. Radford [1920
posed at eighteen or nineteen, contain about ninety specific
and striking Ovidianisms ; Kleemann gives about half of
these. The Panegyric Upon Messalla contains probably more
than a hundred Ovidianisms ; Ehrengruber has already called
attention to all of these, and in order to explain their occur-
rence, has propounded the ingenious theory that the Pane-
gyric was a school exercise composed in a later age by some
pupil of the rhetoricians who had access to all the works of
Ovid and pilfered most freely from them all. 16 Next, as the
metre shows, come the six little letters of Sulpicia, the kins-
woman of Messalla, in which this lady (with Ovid's sympa-
thetic assistance) undertakes the part of wooing the shy
youth, Cerinthus; they contain more than twelve Ovidian-
isms. About 21 B. C. , aetate twenty-two, Ovid composed the
five charming elegies giving in fuller form the story of the
same pair of happy lovers, Sulpicia and Cerinthus ; they
show more than forty Ovidianisms and 47. 4% of dactyls,
thus approaching closely to the proportion of the first Amores
(about 48. 5%). The exquisite imitation of Tibullus, Nulla
tuum nobis (iv, 13), that closes the collection, has ten Ovidian-
isms. Since Ovid edits the unfinished second book, the two
spondaic elegies, with Ovidian language and thought, are in
all probability his work, and can no longer be ascribed to
Tibullus either in whole or in part. The poem in honor of
Messalinus (n, 5) has, in fact, long occasioned difficulty, and
has been known for nearly a century as the ' suspected elegy. '
The other Ovidian elegy, n, 2, is evidently a continuation of
the Sulpicia group, and celebrates the birthday of the shy but
sorely smitten lover Cerinthus, who is now happily married
to Sulpicia and is therefore given his true name of Cornutus.
This poem is rightly assigned to the ' fourth book ' by Gruppe.
The conclusions which we have reached upon grounds of
language and metre are supported also by strong external
16 Op. cil. x, 71 ; n, 28; in, So, etc. This treatise of more than 700 pages
in all is a magnificent collection of material and, in spite of its technically
erroneous conclusion, is truly a masterly piece of work.
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? Vol. li] The Juvenile Works of Ovid 159
evidence. Thus in the letters of the exile Ovid refers in
unmistakable terms to the issuance of the Appendix, when
he writes that it was Messalla who first induced him to ven-
ture upon the publication of his works. 17 He mentions
expressly the epicedion which he had composed upon the
death of his patron and which was sung in the forum (Pont.
i, 7, 29 fL), but he also refers again and again to poems which
he had composed in Messalla's honor in his early youth.
Thus he writes that not even the eldest son, Messalinus, can
remember the time when he " first began to venerate Mes-
salla " (Pont, n, 3, 79 ff. ). This last statement scarcely
applies, I think, to our Panegyric, but can be more fitly
referred to an earlier eulogy which we fortunately still possess
-I mean Catalepton ix (xi), contained in the Appendix
Vergiliana, a poem which celebrates Messalla's triumph over
Aquitania, and which was therefore written in the year
27 B. C. Ribbeck, Appendix Vergil. Proleg. 12, and Gesch.
d. rom. Die/it, u, 200, has already, on stylistic and metrical
grounds, identified the author of this elegy with ' Lygdamus,'
and both Marx (Pauly-Wissowa, i, 1326, s. v. Albius) and
Schanz, op. cit. 282, p. 233, speak approvingly of this view.
I find on examination that these judgments are strongly
supported by the language of the poem, and this elegy may
therefore be confidently regarded as the earliest extant work
of Ovid, written in his seventeenth year ; in fact, in a brief
monograph and admirable commentary which was published
some years ago, but has just come into my hands at the
moment of writing (De Ovidio elegiac in Messallam auctore,
Budapest, 1909), Nemethy, I find, has already clearly per-
ceived and, in large measure, convincingly demonstrated the
Ovidian authorship of the Catalepton. Messalinus, the eldest
son, quindecemvir 19 B. C. , consul 3 B. C. . was probably born
about 38-36 B. C. , and can therefore scarcely have been more
than ten years of age when this eulogy was composed.
I have not yet spoken of the date of the second eulogy
17 Pont, u, 3, 75 ff. (addressed to Cotta, the younger son).
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? 160 Robert S. Radford [1920
which we possess, namely, the elaborate Panegyric of the
Messalla Collection. This brilliant and highly rhetorical
work is metrically more advanced than the Lygdamus elegies
and was certainly composed at a later date than these poems.
Undoubtedly, by an ingenious literary or artistic fiction, the
Panegyric itself purports to be written in the year of Messalla's
consulship, 31 B. C. , and the events of the years 30-27 B. C. ,
namely, the expeditions to Aquitania and the Orient, which
had already been expressly celebrated by Tibullus (i, 7) and
by the youthful Ovid (Catalepton, ix) , are nowhere mentioned
as having actually occurred. Special students of the Pane-
gyric, however, have long seen that the Gallic and Egyptian
campaigns are well known to the clever writer, and are most
skilfully introduced into the poem by way of prophecy (vati-
cinium ex eventu}. 18 The Panegyric was therefore composed
after the Lygdamus elegies, and the first draft of the poem
must have been drawn up about the year 23 B. C. ; it treats
chiefly the earlier career of Messalla, which had not previously
been made the subject of poetic encomium. We are not at
present fully in a position to state how Ovid was occupied
in the interval between the composition of the Lygdamus
poems and the Panegyric. It is true that H. de la Ville de
Mirmont confidently assumes (Jeunesse d'Ovide, 209) that
it was shortly after the two early marriages and about his
twentieth year that our poet visited Asia Minor and Sicily
in the suite of the poet Macer, and at first glance the Pane-
gyric also, in its present revised and perfected form, appears
to contain probable or possible references to Sicily (vss. 197,
200). We must carefully refrain, however, from drawing
hasty conclusions at present and must frankly admit that
we cannot at once determine the exact date of the year which
Ovid spent with Macer in Asia and Sicily (Pont, n, 10, 21 ft". ).
A brief word must be said, however, upon Ovid's relation
to the Ciris (the legend of Scylla and Nisus) in the Appendix
18 See Hartung, op. ell. 38 ff. ; Ehrengruber, op. cit. i, 7 ; x, 71 ; also Belling,
Albius Tibullus: Untersuchung, 205 ff.
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? Vol. li] The Juvenile Works of Ovid 161
Vergiliana. As is well known, this poem offers special peculi-
arities and difficulties. Besides being partly a translation
from Greek sources, it everywhere closely follows the manner
of Catullus and, owing to the poet's prodigious memory, in
many passages it presents almost the appearance of a cento
compiled from Catullus and Vergil. It was composed after
the publication of the complete Aeneid (19 B. C. ), the whole
of which it imitates. The agreement therefore of the Ciris
with the usual Ovidian vocabulary is not quite so close as
we find in the other juvenile works, yet it is sufficient, I
believe, easily and conclusively to establish Ovidian author-
ship, especially when we consider that, by a species of /eeW<m
and in a purely temporary stage of his art, the poet has di-
vested himself of a part at least of his usual and natural
manner. Certain it is that the situation described in the
poem suits Ovid and Ovid alone. The work is addressed
(vss. 36 and 54) to the young " Messalla," by whom Mes-
salinus is evidently meant. The writer definitely renounces
(vss. 1-2) the public career to which he had formerly devoted
himself and of which he has now grown weary, but there is
not a word in the poem to warrant the usual assumption that
the author was a man of advanced years who had reached the
age of forty-five or fifty. 19 On the contrary, the situation is
precisely that described in Trist. iv, 10, 33-40, i. e. , the author
has held certain minor offices in the cursus honorum and now
refuses to go further in the pursuit of public honors. Hence
no one who has followed the career of Ovid with genuine
interest can read the opening lines of the Ciris without some
thrill of emotion. 20 Refusing to become a candidate for the
quaestorship and the senatorial rank at the age of twenty-
19 Teuffel-Schwabe, Hist, of Rom. Lit. , Eng. trans. 230, 2, i; Ribbeck,
Gesch. d. rb'm. Diclit. n, 355; Schanz, op. cit. 241.
20 Propertius also (in, 21, 25 ff. ) finely pictures himself as a student at
Athens in the school of Plato or the garden of Epicurus, but the scene is prob-
ably an imaginary one. At a more youthful age and with less experience of
the world, Horace too visited Athens and " sought for truth amid the groves
of the Academy " (Episl. n, 2, 45).
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? 162 Robert S. Radford [1920
six, the poet whose influence upon subsequent European
literature was to be so vast, hastens to Athens, 21 the " fair
garden of Cecrops," ! 2 and the true home of the intellectual
life, in order to drink, at the fountain source, of the ever-
living waters, and to worship at the shrine of the world's
four great Teachers. 23 With the over-sanguine temperament
of youth, he even dreams of composing at some future time
a great epic upon nature and the creation that shall rival
the sublime and majestic work of Lucretius. 24 The Ciris is
to be placed then about the middle of Ovid's spondaic period ;
it precedes the first Amores and also three or four other works
belonging to the carmina iuvenalia.
40
II. Transition from Sulpicia Elegies to Amores. Spondaic
Character of First Amores
The percentage of dactyls and of dactylic beginnings which
the juvenile poems of Ovid exhibit may be seen in a summary
form from the table below. In the case of elegiac verse the
percentage is here given for the whole distich, that is, it has
been obtained by combining the first four feet of the hex-
ameter and the first two feet of the pentameter ; in the case
of epic verse, the percentage is for the first four feet of the
hexameter. In the Lygdamus elegies, since the style of the
youthful poet is still imperfectly formed and he vacillates
between two proportions, I give the six Lygdamus elegies
first as a whole and secondly as forming two groups. It will
be found that the second group (4 and 5) gives results almost
! 1 Trist. i, 2, 77 : nee peto, quas quondam petii studiosus, Athenas.
22 Etsi me vario iactatum laudis amore
Irritaque expertum fallacis praemia vulgi
Cecropius suavis exspirans hortulus auras
Florentis viridi sophiae complectitur umbra. Ciris, 1-4.
23 Plato, Aristotle, Zeno, and Epicurus; cf. Ciris, 15: sapientia . . . quat-
tuor antiquis heredibus edita censors.
24 Ciris, 12-41. The plan was later fulfilled probably in the Aetna (the
language of which I have not yet examined in all its details) and in the pro-
oemium and fifteenth book of the Metamorphoses, also perhaps in the lost
Phaenomena and in parts of the Fasti.
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? Vol. li]
The Juvenile Works of Ovid
identical with the Sulpicia letters. The figures for the Pane-
gyric and the Ciris are those of Ehrengruber (op. cit. x, 5
and 21). The Aetna, which, according to every probability,
is also the work of Ovid, is omitted from the comparison
only because the exact figures are not accessible to me at the
present writing ; its proportions are. however, not very differ-
ent from those of the Ciris.
There appears to be conclusive
evidence for including also the Culex, though I was long
prevented from examining the language of this poem by
erroneous impressions that I had at first formed respecting
the treatment of the caesura in this work. Finally, Pliny,
N. H. xxxii, 152, was wholly mistaken in his conjecture
(" fortassis ") that the Halieutica relates to the " fish of
the Black Sea " and was consequently written at the close
of the poet's life ; the schemata show conclusively that the
poem, wholly dependent as it is upon Greek books, belongs
to the Lygdamus and Sulpicia period.
THE JUVENILE WORKS OF OVID
Catalept.
TV
Halieut.
Lygdamus,
all six
Lygdamus,
four
Lygdamus,
two
Sulpicia
Letters,
IX.
62 vss.
130 hex.
elegies. 25
elegies,
i, 2, 3, 6.
elegies,
4i 5-
IV, 7-12.
290 vss.
1 60 vss.
130 vss.
40 vss.
SS. Dact. 2
53 44-8
222 42. 8
359 41-3
216 45. 0
144 36. 9
45 37-5
SS. Spond.
75 55--'
298 57. 2
5" 53. 7
264 55-o
246 63. 1
75 62. 5
I. 27 Dact.
46 71. 0
62 47. 7
1 66 57. 2
ioi 63. 1
65 50. 0
20 50. 0
I. Spond.
1 8 28. 1
68 52-3
124 42. 8
? ? 59 36-9
Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? 164 Robert S. Radford [1920
A few words of explanation may be added to the tabular
statement. Ovid began in Catalepton, ix with a proportion
of dactyls which was wholly normal in the year 27 B. C. Thus
he has 44. 8% in the whole distich, which is as good as the
second and third books of Propertius, 28 as good also as the
eighth and third elegies of Tibullus' first book. 29 Yet he has
no limitations whatever in Catalepton, ix upon polysyllabic
endings in the pentameter; on the contrary, like Catullus
and like Propertius in his earlier work, he fairly revels in
their use (50%). In the Lygdamus elegies, however, and
in the Sulpicia letters, the ambitious and aspiring youth
seeks suddenly to pass from the longer endings to the more
elegant dissyllables of Tibullus, and is evidently preoccupied
with this problem and its difficulties. Under these conditions
throughout the Lygdamus poems he wavers greatly in his
composition, and, in one hundred and seventy verses (Lygd.
4 and 5, [Tib. ] rv, 7-12), he even sinks to the proportion of
Catullus, namely, about 37% in the distich, and to only
50% of dactylic beginnings. 30 This is of course excellent
Latin elegy, but it is not the kind that was most in vogue in
24 B. C. 31 It is only in the qualified or limited sense just ex-
28 Propertius has 44. 8% and 44. 7% in the second and third books respec-
tively. My figures are taken from Hultgren, op. cit. 23, who follows the five-
book division of Propertius.
29 Tibullus has 44. 8% in the eighth elegy (Pholoe) for the whole distich
and 45% in the third elegy. It is scarcely fanciful to see in the decline of
dactyls in the third elegy an expression of Tibullus' sadness and depression of
feeling during his illness at Corcyra ; we have the same phenomenon in Lyg-
damus' fifth elegy. My figures are taken from Cartault, Le distiqitc elegiaque
chez Tibitlle, Sulpicia, Lygdamus (Paris, 191 1), 7.
30 Figures for Catullus are given by Hultgren, op. cit. 15 ff. Catullus has
usually 58% to 6o r i of dactylic beginnings, but in Carm. 69-119 (319 vss. )
he has 36. 6^ of dactyls in the distich and only 47^ of dactylic beginnings.
31 Catullus wrote some of the best Latin elegy, and the naturalness and
directness of his style is due in part to his not exceeding the proportion named.
As I have shown in my "Licensed Feet in Latin Verse," op. cit. 251-27? ,
even the best Latin poets, such as Catullus and Horace, experience some diffi-
culty in always providing one required dactyl, and therefore they occasionally
admit without metrical ambiguity in such a foot exceptional or vulgar short-
enings and even short vovels (without m) in hiatus, as Lucilius, ix, 243 Bahr. ,
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? Vol. li] The Juvenile Works of Ovid 165
plained that Ovid can be said to have ever been a disciple
of Catullus in the matter of dactylic proportions, but un-
doubtedly in his youth he paid this brief tribute to the freer
and more natural style of his great predecessor. A secondary
cause for the low proportion may perhaps be found in the sad
and almost despairing character of the two elegies.
It will be noted that Ovid advanced in the Sulpicia and
Messalinus poems which are composed in the elegiac metre
to about 47% of dactyls in the who! e distich ; the hexameter
lines alone of these poems show 46. 1% and 45. 9% respectively.
On the other hand, the hexameters of purely epic poems like
the Ciris fall back to 43. 5%, which is normal for epic verse
o 5 re cSrupto (necessary dactyl of the hexameter); Lucr. vi, 1133, natu 5 ra
cSruptum; Cat. 10, 26, istos c6mmoda; nam volo d Sar&pim (necessary
dactyl of the Phalaecean) ; Hor. Cann. in, 14, n, iam virum expertae, malS
6minatis (short vowel in hiatus, necessary dactyl of the Sapphic) ; Pers. 3, 66,
di'scitfi, 6 miseri (license of the first foot, with greatly preferred dactyl) ; Lux-
orius, 302, 4, magnum depre 5 nderg usum. If difficulty is experienced in supply-
ing one required dactyl for the Sapphic or the hexameter, it is clear that two
necessary dactyls, as in the pentameter, constitute a very exacting demand
upon the Roman language, and if the virtuosity is also insisted upon, a very
elegant but very artificial form of verse is the result. For example, for metrical
reasons, Ovid (like Tibullus) constantly uses ab arte and similar phrases drawn
from the vulgar language, with its analytical tendency, instead of the simple
ablative, and he also uses, by poetic license, to an unparalleled extent, the
simple ablative for the ablative of the agent with ab, as Her. 4, 64, capti
pare 5 ntg sor6r, ' my sister was captivated by your parent,' where neither
parenti nor a parente could enter the verse ; upon the whole subject see in part
Guttmann, Sogenanntes instrumentales ab bei Ovid, Dortmund, 1890. I was
mistaken therefore in my former discussion (p. 271) in thinking that Ovid had
perhaps actually used the spelling parente for the dative parenti, as he so often
uses mare and caeleste for mari and caelesti, and as Statius, Sih. iv, 2, 28, uses
glaucae certantia Do b ride saxa, and Propertius, v (rv), 8, 10, writes cum temere
anguino creditur o^rl manus (Xeue-Wagener, Formenlehre, i 3 , 301). The
honest 37% of Catullus and Lygdamus does not compel a resort to such un-
usual constructions, and is therefore by no means to be despised. Dissen and
Postgate (Selections from Tibullus, p. XLIII), on purely subjective grounds and
without consulting indices, lexicons, or Latin authors, have discovered that
Lygdamus is an author of " poor Latinity. " On the contrary, his Latinity is
more natural and in some respects better than that of the mature Ovid. Yet
even the latter like Euripides, an unrivalled master of the graceful and
pleasing forensic style could give most of us lessons in correct Latinity to
our great and lasting profit.
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? i66 Robert S. Radford [1920
and almost precisely the proportion of Vergil. In fact that
strange and anomalous thing, the actual preponderance of
dactyls in a Latin epic poem, could never by any possibility
arise in the more vigorous epic hexameter (versus fortis or
grams) ; it must appear first in the soft and tender strains of
the elegy (versus mollis or levis) , and may be transferred thence
to epic verse, as Ovid transferred it in the Metamorphoses
and Valerius Flaccus later employed it in imitation of Ovid.
Even Ovid in his mature period has recognized this difference
to some extent; for the hexameter of elegy in the Heroides
(first series) reaches 56. 4%, while the epic hexameter of the
Metamorphoses does not exceed 54. 8%.
It is evident then that in the first edition of the A mores
which was published in 14 B. C. , 32 only a few years after the
Sulpicia elegies, the Ciris, and the Aetna, Ovid had had little
opportunity to develop a marked dactylic virtuosity and to
become a highly artistic elegiac poet. We may be sure then
that the first Amores of 14 B. C. showed only a very moderate
advance upon the 47. 4% of the Sulpicia elegies, and they
may even have contained, like the latter, a few polysyllabic
endings in the pentameter. Up to 14 B. C. Ovid was in fact
fully as much an epic as an elegiac poet (see Am. i, i, 1-18
on his early ambition to excel in epic themes), and he was
still far from having attained in the Sulpicia poems the high
virtuosity which appears in three of the four genuine elegies
of Tibullus' second book, namely, 55% or more. 33 Two of
the short Sulpicia elegies and the equally short Cornutus
poem show, it is true, an equal or a superior number of dac-
tyls, 34 but four Sulpicia poems and the long Messalinus elegy
32 Schanz, op. cit. 270 (293), suggests that these elegies probably came
first separately before the public, but later were collected in the edition of
14 B. C. This seems quite probable, especially as the spondaic Amores show
in general the predominance of the same schemata or figures of the hexameter
as the Sulpicia and Messalinus poems.
33 The first elegy of the second book has 55. 2% in the distich, the third has
50%, the fourth 55%, and the sixth 56. 8%; see the figures of Cartault, op.
cit. 7, and Ehrengruber, op. cit. x, 5.
34 iv, 6 (20 vss. ) has 50% of dactyls ; iv, 5 (20 vss. ) 58. 3% ; n, 2 (20 vss. ) 50%.
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? Vol. li] The Juvenile Works of Ovid 167
(n, 5) have a decided preponderance of spondees. A marked
change in Ovid's whole attitude took place, however, after
14 B. C. For unlike the Ciris and the Aetna, which did not
greatly increase their author's reputation, the Amores, im-
mediately upon their first publication, achieved a prodigious
success; they at once became popular favorites, and, like
the Eclogues of Vergil, were frequently sung in the theatre
with accompanying dance (Trist. n, 519; v, 7, 25). The
favorable reception thus accorded to the elegies naturally
determined the direction of the poet'3 genius and led him to
devote himself uninterruptedly (with the exception of the
tragedy of Medea) for sixteen years (14 B. C. to 2 A. D. ) to the
perfection of an elegiac style which surpassed even the later
work of Tibullus.
Having developed a remarkable and distinctive technique
in the Heroides and the Ars, Ovid resolved to make the first
Amores conform fully to his later rules of art. The revision
was not due to the fact that this popular and successful work
was poetically immature, nor even chiefly to the desire to
add a series of new elegies ; as he himself tells us in the pre-
fixed epigram, the question was primarily one of more careful
finish and elaboration : " hoc illi praetulit auctor opus "
(cf. Gruppe, op. cit. 377). In brief, his principal purpose
was to eliminate all polysyllabic endings, to introduce every-
where the full dactylic virtuosity, to multiply the dactylic
beginnings, and to remove or greatly reduce those schemata,
such as DSSS, SDSS, 5555, 55, SD, and the like, which
were no longer fully acceptable. The view adopted by
Heuwes in his discussion of the matter (op. cit. 31) is that
the revision " consisted in here and there substituting a word
that was more suitable for one that was less suitable to the
metre and sense, or in changing the collocation of words o 1
verses, or in doing all these things at the same time. " These
conclusions are partly correct, but there can be no doubt that
both the additions and the alterations were far more extensive
and thoroughgoing than Heuwes supposes. The first edition
? ? Generated for (University of Chicago) on 2014-12-26 11:20 GMT / http://hdl. handle. net/2027/uc2. ark:/13960/t45q4vr7s Public Domain / http://www. hathitrust. org/access_use#pd
? 1 68 Robert S. Radford [1920
was in fact subjected to a drastic revision and was almost
entirely rewritten in conformity to the new rules of art. 35
Nevertheless abundant traces of the character of the first
edition still remain ; for in not a few cases the poet has
triumphed over the versifier and the artist, and a number
of the original poems of the first edition have been retained
with very moderate changes. This fact may be shown very
clearly in two ways, if the writer correctly understands the
true nature of Ovidian versification. For the full and har-
moniously developed virtuosity in the Ovidian sense means
first that there shall be a considerable preponderance of
dactyls in the distich taken as a whole, and secondly that
there shall be a similar preponderance of dactyls in each
member taken singly, that is, in both the hexameter and the
pentameter lines. It is evident therefore that virtuosity is
lacking in the Amores in two ways: (i) when the spondees
are either equal or predominant in the whole distich ;
(2) when they are equal or predominant in the hexameter
line (versus fortis or diirus) alone. With this understanding
of the nature of virtuosity, I may summarize the facts briefly
as follows.
