that what it had seen devoured and dissipated by the flames could reunite, even in another world. There is an uncomfortable feeling in the description that Cynthia's ghost was pieced together; compare v. 12 pollicibus fragiles increpuere manus. Once more it gives occasion to subtle etymological uses of words. Thus in III. 5. 8 plebei paruae funeris exsequiae (the procession of the funeral), v. 9. 24 lucus ab umbroso fecerat orbe nemus 'a circle of sacred shade trees (lucus) formed a glade' (nemus). nemus is an open space in a lucus'. There is a term, already used by others? with reference to Propertius, which seems to me to be an appropriate description of the foregoing phenomena; and I have consequently used it throughout my commentary with reference to them. I should, however, be sorry to see it limited to them, as it gives very happily the general disparateness of the Propertian genius. The word I mean is disjunctiveness.'

3

False echoes.

I

This may be justly called perverting language and forcing it into channels along which it was never intended to flow. But here, as elsewhere, there is the inevitable retribution; and Propertius on his part is often snared in his words. have already pointed out instances where his choice of expressions has been irrationally influenced by what he or others have written: and examples of these 'false echoes,' as Professor Campbell in his Sophocles appropriately calls them, might be multiplied still further. Thus we have in 11. 1. 57, 58 omnes humanos sanat medicina dolores: solus Amor morbi non amat artificem, an obvious echo of 1. 2. 7, 8 crede mihi non ulla tua

1 Compare p. xci.

2 I do not know who used it first, or in what connexion. Nor can I now find out where I saw it quoted.

3 By this I mean that, though the meaning is changed, the expression is the same.

P. P.

f

'Fixed phrases.

est medicina figurae: nudus Amor formae non amat
artificem (where see note); in III. 3. 20 uapulat
umbra mea is a reminiscence of Plautus (quoted
in the note). II. 6. 26 et quaecumque uiri femina
limen amat is from Horace Od. 1. 25. 4 amatque
ianua limen. I have pointed out the same tendency
in IV. 24. 7 (where see note) and iv. 7. 581,
which shews the influence of metre.
With all his love of variety and inclination to change,
Propertius has a large number of phrases into which
his thought seems naturally to fall. Of these I have
already given some examples and drawn some infer-
ences from them (p. xl.). Their number need not sur-
prise us. It is a mark of a certain phase of disorga-
nized intellect to set round anything which has even a
comparative fixity. Still less, as the foregoing examina-
tion might show us, need we wonder that their meaning
often fluctuates and sometimes disappears. They are
the buoys in Propertius' troubled sea of thought and
they share in its heavings. There was another cause
for these repetitions which we must not neglect.
Propertius was deeply imbued with the spirit of
Greek literature, and he felt with a Greek influence.
true poet's instinct that his own lan-

guage was in itself incapable of producing the qua
lities which attracted him. Consequently he called in
the assistance of art, and, following in the steps of his
Alexandrine masters, he has succeeded better perhaps
than any other Roman poet in giving to his verse
something of the harmony and symmetry of the
Greek. I cannot spare space to demonstrate this at
length: for one illustration will suffice and Hertzberg
(pp. 107 sqq.) has given many. It is very apparent in

1 I am aware that these and the following remarks may seem fanciful, especially to persons who have not considered the subject.

II. 3. 43, 44 siue illam Hesperiis siue illam ostendet

Eois, uret et Eoos' uret et Hesperios. Compare 1. 8. 25 and indeed the whole of the poem, I. Love of symme16. 30, 20. 26, and the notes. The attrac- try and correspondence. tion of these correspondences in metre and syntax extends yet further; and under their influence Propertius often assimilates the form of one sentence to another, even when there is little similarity in substance between them. Examples are 1. 12. 20 Cynthia prima fuit, Cynthia finis erit (Cynthia et ultima erit would have made the correspondence real); II. 5. 28 Cynthia forma potens, Cynthia uerba leuis (the discord between the real construction and that which the words seem to suggest is very marked). In III. 13 (11). 18 the result is a very forced expression ambos una fides auferet, una dies. While discussing the influence of mere expressions in Propertius, I will add a tolerably complete list of the instances

Literal and me

fused.

where in consequence of the use of a word taphorical conin more than one sense we find either a confusion, or more strictly a non-differentiation of ideas, or a more or less conscious play upon words. III. 17 (14). 23, 24 (a passage generally misunderstood) libertas quoniam nulli iam restat amanti, nullus liber erit si quis amare uolet 'since we now see (iam as in III. 32 (26). 24 omnes iam norunt quam sit amare bonum) that no lover is free (i.e. his own master), he will be no free man (i. e. free-spirited) who chooses to love,' ib. 19. 24 (16. 40) ferre ego formosam nullum

1 Observe the double quantity of Eous, and compare what Martial says 9. 12. 13 sqq. on the unsuitability of Earinus for verse, dicunt Earinon tamen poetae, sed Graeci quibus est nihil negatum et quos Apes "Apes decet sonare; nobis non licet esse tam disertis qui Musas colimus seueriores.

2 The repetition of the same words in similar positions is the chief agent in producing these effects.

onus esse puto 'to bear with the beauteous is no burden,' so in IV. Î. 6, IV. 6 (7). 33, iv. 8 (9). 23, 24, iv. 23. 7 and notes, v. 5. 54 uersibus auditis quid nisi uerba feres? (a play on uerba dare to cheat)'. In some of these cases the literal and metaphorical are blended. So also in Iv. 3 (4). 4 Tigris et Euphrates sub tua iura fluent, where the idea of the actual motion of the rivers is united with the metaphorical idea of their passing under Augustus' jurisdiction. We have, however, nothing as bad as Ovid's experiments in this line, of which I will quote two, in order to shew how hopelessly false and hollow the thing may become when not guided by genuine poetical feeling. Trist. 3. 5. 45, 46

non mihi quaerenti pessumdare cuncta petitum

Caesareum caput est quod caput orbis erat (two bodies and one head)

Pont. 2. 5. 38

sed sunt tua pectora lacte

Explanation of the phenomena.

et non calcata candidiora niue (a good metaphor spoilt, material and metaphorical 'fairness' being confused). I am sure that my readers will pardon this somewhat long, though imperfect, discussion if I have shewn them that in Propertius we are dealing with no ordinary phenomena. These contrasts, these extravagancies, these fluctuations and incoherencies, these half-formed or misshapen thoughts, what do they signify? What is the secret of this chaos? It is that here we are looking on a stage in the realization of thought which is not usually presented to our view. In other writers we only see the full-formed crystals, sometimes flawed and dim, sometimes bright and clear. In Propertius thought is crystallizing still. It is still comparatively amorphous

1 These must be distinguished from cases where the confusion is in the idea itself; where, for example, a God and his statue (v. 1. 7) or a river and river-God are not discriminated (Iv. 3. 45 n., v. 2. 7 n.).

and still turbid with development. At such a stage there is hardly any limit to possibilities; and almost anything is possible with Propertius. At such a stage the susceptibility to impressions is extreme; and we have seen how potent even the smallest influences are in attracting and deflecting Propertius. But I will leave my readers to pursue the metaphor farther for themselves'.

self-criticism.

It is hardly necessary to point out that with thought, thus unconsciously developing itself, a conscious self-criticism was impos- Incapacity for sible. And I cannot find the slightest proof or hint in the poems that they were subsequently submitted to its test. It is true that we find traces of double readings occasionally (see Appendix A); but these are to be differently explained. Poets are not usually good critics, especially of their own compositions; and Propertius certainly was no exception to the rule. There is the greatest difference in his work. By the side of poems which shew the highest flights of his imagination, we find, as in III. 18 (15), the flattest level of conversational prose. Even if he had wished thus to improve his poems, it is very doubtful whether he could. For his best work is that which is done at once under the strong and controlling influence of immediate inspiration3.

1 This, though an early stage in the development of poetical thought, is not the earliest. The curious in such matters may find the earliest in some of the poems of Blake.

I trust these remarks will be intelligible. They are as clear as the nature of the case allows. It is impossible here at any rate to distinguish between the genesis of thought and its embodiment in language.

2 There is nothing that shews so much spirit as v. 8, nor so much grace as I 9 (10), a birthday poem to Cynthia; and they are as finished as anything in Propertius. Yet they were written immediately after the events that they commemorate.