صور الصفحة
النشر الإلكتروني

begins without an Introduction or Dedication of a formal kind, and the one which of all the books we should most expect to be so adorned, seeing that it is on a more elaborate scale and concerned with higher matters than the rest.

To recapitulate. Concentrating upon one region of the text, already suspected by the prince of Latinists because of the simple meaning of two Latin words, we have seen that II. x. xi. refer to our Books I. II. III. as already concluded; that xiii. refers to Book II. as already concluded (13) and to Book III. as well under way (25), but at the same time that much later in the period of our Book II. nothing but the Cynthia in one volume has yet appeared, and that we have not passed the seventh month since its appearance.

Without calling more evidence I submit that it is proved that our text has suffered from dislocation of a complicated kind.


Lucian Mueller' claims to have sought, but in vain, for the pagination of a Propertian archetype. Scaliger repeated the story of Alessandro Alessandri to the effect that the archetype of Propertius was discovered in the fifteenth century in a battered condition; he made it an excuse for the inauguration of a school of jig-saw which is not yet dead. (But the key-picture may be a necessity for this art). I have said elsewhere (Journ. of Philol. XXXI. pp. 175, 195) that some of the MSS. of the family I call C are themselves dislocated, and that the story may refer to a really old and really battered copy, which was not however the archetype.

Lachmann's reconstructed archetype of Lucretius was the immediate parent of two very old but extant MSS. The isolated fragments copied out at the end of the later MS. were still in their place in the earlier. Certain leaves fell out of the archetype in the period between the two, and were inserted together at the end. The number of lines upon the leaves of the archetype could thus be easily inferred; in old MSS. this is usually constant throughout.

The archetype of Propertian MSS. is very far behind our extant copies, and the order of its text was theirs. If dislocation has taken place, it was at a point in the tradition still further back than our archetype, which was, of course, the latest from which all extant MSS. are descended. The pagination of the archetype is likely to be immaterial; that of its damaged ancestor must be the object of our quest. Fulgentius in the sixth century quotes as from Propertius the verse diuidias mentis conficit omnis amor, which is not found in our MSS. Between him and the scribe of our Neapolitanus at Wolfenbüttel over 600 years intervene, plenty of time for the decay of a binding or for the neglect of so unascetic a poet, even to the tearing out of his leaves to serve a holier purpose. No instance can be quoted of a classical writer's works being deliberately mutilated and disarranged. The damage in our case was presumably due to accident, ignorance, or mere age. The critical faculty is more likely to come into play when the damage has to be made good as far as possible-that is,

1 Preface to his Teubner text.

in a copy of a dislocated exemplar. Signs of editing then in our archetype would confirm the theory of dislocation at an earlier stage. There are many such signs.

Signs of Dislocation and of Editing in our IV. 1.

I have briefly pointed out in C.Q. for April, 1917, XI. p. 103, some of the reasons which have led me to the assurance that in our II. xxxiv. 61-94 we have two distinct references to Virgil, of which vv. 61-66 belongs to the year 23-22 B.C., while vv. 67-94 are a complete elegy removed from the end of the Cynthia and of date 26 B.C. The junction of such passages from kinship of subject (and that a favourite subject) would be a natural idea for an editing corrector, whose exemplar had fallen to pieces. I also refer there to the fate of the passage on the poet's birthplace, our IV. i. 61-66. In connection with this latter I propose now to discuss briefly the whole region of the text, IV. i. 38-70.

(i.) Lachmann postulated a lacuna after 38 and after 54. The intervening passage, 39-54, is certainly inconsistent with what precedes and what follows. But let us take certain points in detail :

(a) In vv. 1-38 the poet has sung only of the small beginnings of Rome, the nakedness of the land of Romulus. In v. 48 Italy is already a felix terra; the omens of Troy have been fair (huc melius, 39; iam bene, 41). There is no hint of lacrimae in 1-38, or 39-54, or in 55-70; yet in 73 Propertius is rebuked for the sadness of his note.

(b) But vv. 87, 88 sound like an anticipation of such a passage as contains our vv. 47, 48, and 53, 54. Yet they are themselves inconsistent with the lacrimae of 73; for the soothsayer rebukes Propertius for his sadness, and then proceeds to describe Aeneas's pilgrimage as longa sepulchra.

(c) vv. 71-76 and 87, 88 prepare one for a recitation by the soothsayer of the fata Aeneae: Propertius himself, the speaker of 1-38, 55-70, seems rather interested in the moenia of Romulus (57). When the soothsayer proceeds to his recitation, vv. 109-118 take us only to the scattering of the Greek fleet, and stop short exactly at the point where the fața become of interest to Rome. Our vv. 39-48 pick up the tale at this very point.

(d) The style of vv. 39-54 is more like the soothsayer's than like that of 1-38; but style alone is a dangerous criterion.

Many scholars have wished to connect 38 and 55, 56: sanguinis altricem non pudet esse lupam, optima nutricum nostris lupa Martia rebus. . . . If we could show cause for the rejection from this passage of vv. 39-54 the juncture would be made.

(ii.) But vv. 55 to 70 constitute a still more vexed passage.

(a) Baehrens (at 57) speaks of a corrector, qui fragmina dissita collegit. Luetjohann proposes to strike out as spurious vv. 65, 66. Housman would insert them in the last poem of the Cynthia, much as Postgate would

join the most recalcitrant couplet of II. x. to the poem now under discussion. But can they be severed from 61-64? The patria of 60 can only mean Italy, whose centre is Rome, in view of moenia (56, 57). The poet claims to be a good Augustan by now and no more a provincial Umbrian. With this 61-66 are quite at variance. In 65 I accept Professor H. E. Butler's correction of my conjecture; for the evidence on which scandentis qui Asis ... is based see my article, Journ. of Philol. XXXI. p. 170. Verses 63 to 66 become thus altogether Umbrian: the patria of 64 cannot be the patria of 60. Again, 57 and 60 invoke the poet's powers to attempt a quasi-heroic strain. It is therefore strange that in 61, 62 he should deprecate 'shaggy' epic and call upon Bacchus for a smooth wreath of his ivy, the more so that in 73 Apollo, the founder and the seer of fate, but not Bacchus of the lighter lay, is in question. Of all occasions on which Bacchus might be invoked surely the least likely is when a song of foundations is to be assayed pio uersu; not even Ennius' addiction to Bacchus (61) will make this tolerable in a city where Bacchic rites were still forbidden in public. Here then is another passage, 61-66, which we must mark off from 55-60 as clearly as we mark off that from 39-54.

(b) Verses 67-70 are inconsistent not only with 61-66 (Roma, tibi surgit opus: superbiat Vmbria libris), but with both 55-60 and 71-74, which agree together that the poet's intention has been to sing of foundations and of fate (71); now he inopportunely asserts that his plan is rather to sing a Hesiodic strain, of things holy, of days, of ancient place-names. The soothsayer was hard of hearing if he thought him rushing at all rashly to sing of fate.

I should not give space to all these difficulties, if I had no solution of them. The first suggestion which I desire to emphasize is that vv. 61-66 do not appear here by chance. The reference to the patria in each passage has led an editing corrector to insert them, where a superficial likeness tempted him. But their removal from another place was surely conditioned by a dislocation there, which obscured meanings and caused the critical faculty to come into play. This combination of autobiographical passages is to be likened to that of the passages concerning Virgil (p. 64, above); nor is that the only parallel in our text. In order to understand how vv. 61-66 came to be removed, we must consider, and if possible visualize, the conditions under which the copyists of a damaged exemplar would work.

The Copying of a Damaged Exemplar.

It is most improbable, for instance, that a scribe copying three consecutive and complete quires would extract from an intelligible context a passage within these clear bounds and insert it at a distance of eight whole quires. If on the other hands, after three complete quires, he came upon a number of loose leaves comprising part of the next quire and parts of quires further on which had at some time fallen out, he might well hesitate to copy them in the haphazard order

in which he found them, and search for contexts or for kindred subject-matter. He would be particularly puzzled if, after a gap in the sense, he came upon a series of consecutive pages the first of which was headed by the last lines of a lost elegy. What to do with this headless and bodiless tail, which his honesty forbids him to omit? He will first look for a body it might fit; but failing that he must either spoil his consecutive series by a fragmentary opening, or hold the short passage in reserve till a somewhat similar context arrives, or till he reaches the end of the series, when he will perhaps insert it at the bottom of a leaf, as he would do with an accidental omission later detected.

Lachmann and Schrader offer us a good example of what I mean at II. iv. In the MSS. his saltem ut tenear iam finibus follows without sign upon v. 44 of elegy iii. Scaliger perceived that this passage cannot belong to iii. At the same time the MSS. begin a new elegy at v. 11, multa prius; and ten verses are too few for a complete elegy from 1 to 10. Lachmann, Schrader, and others therefore begin the elegy at his saltem and ignore the MSS. at v. 11, though the conspicuous illuminated letter now given to Multa was the last thing likely to have shifted on this page. Imagine our scribe, perhaps less of a Latinist than Schrader or Lachmann, faced with such a problem as this. He has ended a page of the exemplar at Hesperios (II. iii. 44); the next leaf is a loose one followed by others also loose. He has no certainty that the next was the next always. He dare not pretend that the M of Multa is not illuminated; shall he begin his next page with the beginning of iv. from that point and keep the ten first verses of the exemplar's page in reserve? He decides that the fines indicated in the first lines are the East and West of iii. 44, and that these ten verses are but the end of iii.

Let us imagine the case slightly different. He comes to a full stop at the bottom of a leaf of the exemplar at iii. 44: he has no idea which of the jumbled loose leaves following ought to come next. So he sorts them through, and finds one at some distance beginning his saltem ut tenear iam finibus. He snatches at a straw, and infers that the fines are our East and West, and brings the whole leaf up to follow iii. 44. If he does not alter the disordered units within themselves and respects always the illuminated letters, he will on the whole be doing less harm than Lachmann and Schrader. But he may not be consistent any more than later editors. He may, in such a disordered region as we conceive, begin to copy a leaf next in order and after a line or two perceive that it is a false join. He may then set this leaf aside for a more favourable moment, and when that comes insert the rest of the page, with a reference back to its first verses already copied. Or he may treat certain of the loose and isolated leaves as units in themselves, to be doctored into some sort of literary shape. He may indicate the beginning of a new elegy, when really it is only the beginning of an isolated page, to show that at least the following verses do not belong to the leaf last copied. This supposition would explain, for instance, the appearance of an otherwise inexplicable illuminated letter in cod. Neapolitanus at II. xxvi. 29: Seu mare per longum mea cogitet ire puella. If in this neighbourhood a

new leaf began which was isolated from the context once preceding, a scribe might indicate this isolation by giving the sign for a new elegy to the first extant verse which could by any manner of means stand at the head of the passage. But for suppositions of this sort to become concrete, it is absolutely necessary first to recover the number of lines which stood upon each page and leaf of such an exemplar. I shall omit the experimental stages, the false trails, and the misjudged evidence among which I have from time to time wandered, and present briefly some of the facts of our case.

The Pagination of the Damaged Exemplar.

We return first to the region whence we set out, II. x. xi. xiii. Is there any hint hereabouts of a common factor to the various sums of verses which on literary grounds awoke suspicion? If x. and xi. seem to belong between our Books III. and IV., perhaps they stood upon a loose leaf which fell out from that region. The sum of verses is 26+6=32; this would mean a page of as few as 16 verses, that is an uncial page. But we notice at once that xiii., which seems to belong to a third book (v. 25), breaks off in sense (and in most modern editions) after 16 verses. Turning on to xiv. we find a poem apparently of 32 verses, but on examination we perceive (with Fischer and Postgate) that the last four of these ought to be the conclusion of another elegy, and that xiv. ends naturally at v. 28 with the inscription. Is this, or is it not, a parallel case to x., xi. ? There too we seem to detect the end of an elegy in the six lone verses of xi., and that of an elegy which, judging by x. 7, 8, should precede the one it now follows. At the end of xviii. is a passage treating of false hair and paint, which Kuinoel and most succeeding scholars divide from the rest of the context. It is 16 lines long.

After v. 10 of xxii. Fonteine and Baehrens indicate a lacuna; after v. 42 all critical editors do the same. From point to point are 32 lines. From xxii. 43 to xxiv. 16, where Scaliger first marked a lacuna, are 32+ 16 verses. There are three more isolated passages of 16 lines before elegy xxxii.: i.e. xxvii., xxviii. 47-62, xxxi. So far as our Book II. is concerned 16 seems a likely unit enough.

We turn on to III. xviii. and following poems. After xviii. 8 most scholars suppose a number of lines to have been lost, for Marcellus is nowhere named in the extant poem nor is his illness, to which our vv. 9, 10 refer. On our view the most probable explanation of some such loss is that a page has not been copied or a leaf has fallen out of the ancestor MS. Now after xx. 10 is an equally clear case of a lacuna: the wooing and the winning are two incidents, not one, as Scaliger pointed out. A page or leaf, on our view, has been omitted here. From xviii. 9 to xx. 10 are 26+28+10=64 verses or 4 pages of 16 (I am not entering into all details here or anywhere, but giving the broad outlines).

Let us count on from this point to the end of the book. 42+24+38 lines = 158, or two short of 160, 10 pages of 16.

We have 20+ 34+
But after v. 36 of

xxii. at least one couplet has fallen out of the catalogue of monsters, for arboreasque cruces Sinis (37) has no construction. (The suggestion is as old as Livineius.)

« السابقةمتابعة »