On The Nature of the Universe, by Lucretius

ON THE NATURE OF THE UNIVERSE

APPENDIX D

The Text

Like all classical texts, our editions of Lucretius rely mainly on manuscripts (MSS) that go back to the Middle Ages. By a detailed comparison of the different MSS - especially their mistakes - it is possible to argue a case for their relative antiquity and reliability - although 'older' does not always mean 'better', as the intelligent scribe emended as he copied, either from his own head or by consulting other MSS.

The actual text L. wrote is lost. So are all the copies made of it for the first few hundred years after its composition, as the earliest MSS date from the ninth century - a gap of eight hundred years between composition and earliest text, the sort of gap between (say) Chaucer and ourselves.

The archetype or source of the MSS we have, then, has not survived. It is, however, known in greater detail than that of almost any other Latin text. In his edition of 1850 the great scholar Lachmann deduced that the archetype was written in capitals in the fourth or perhaps the fifth century. The symmetry of the passages omitted in one major manuscript and stuck in at the end - each of them fifty-two lines long - shows that the pages were of a standard twenty-six-line length, the fifty-two-line gaps being the result of a whole leaf falling out of place. (Cf. for example 4.299-322, misplaced after 323-47 because a leaf of the archetype had fallen out and been replaced the wrong way round.)

The earliest sources for the text are the two Leiden manuscripts O and Q: three sets of loose pages (schedae) are also extant from the same century, known as G, V and U. After OQGVU there is a further gap of about 400 years until the next MS.

In 1417 the scholar Poggio Bracciolini discovered a text of L. (known as π) and sent it off to Niccolo Niccoli to be copied: Niccoli's copy survives in Florence and is known as L, and a host of other MSS exist copied from Poggio's text and known collectively as 'Itali'. The consensus of the best of them is known as P (for Poggio) and is the best reconstruction possible of that long lost manuscript π found in 1417. The comparison of P against OQGVU suggests that Poggio's text derives from O, although there are occasions where P agrees (rightly and wrongly) with Q against O, and although some scholars argue that the Itali derive from an MS tradition independent of OQ. The 'family-tree' of MSS (the stemma) as suggested by recent research looks like this:

Below are listed the places where the text translated differs from the Oxford Classical Text of Bailey. Where the text is certainly corrupt but no emendation has been devised that is acceptable, then the corrupt words are printed with the so-called obelus (†), before and after them.

BOOK ONE

14	insert comma after ferae.
44-9	omitted by Bailey, as it is repeated at 2.646-51, but translated here.
199-600	Bailey notes a lacuna, filled in the translation by Munro's conjecture: corporibus, quod iam nobis minimum esse videtur debet item ratione pari minimum esse cacumen On this see Furley, Two Studies in the Greek Atomists 31-3, and for a different view, Long and Sedley, The Hellenistic Philosophers vol. 2 p .36.
657	Bailey prints †Musa† .The translation reads Merrill's quae sint (cf. 4.510).
721	Bailey's ltaliae is the easier reading, altered by him in his own 1947 edition to my own reading Aeoliae, referring to South Italy, close to the Aeolian islands and near Rhegium, which was said to have been founded by Aeolus' son Iocastus.
744	Bailey reads ignem with the manuscripts, but altered it in his later edition to imbrem, which nicely completes the four Empedoclean elements.
860-61	Lambinus noted the lacuna and supplied the supplement, which I translate: et nervos alienigenis ex partibus esse
873-4	Munro noted the lacuna and Diels transposed 873-4 to produce sense: the translation fills in with the suggested supplement of Bailey in his 1947 edition.
998-1001	Bailey puts these lines in between 983 and 984: the translation restores them to their original place.
1013-14	lacuna: translation follows Bailey's suggested supplement.
1068-75	defective in the manuscripts as a page has obviously been torn, removing the ends of these lines. The translation attempts to reconstruct the sense from the words remaining.
1093-1102	Manuscript O has a lacuna of eight lines. The translation supplies the minimum required to join the two ends together, but the lacuna may be considerably longer than eight lines.

BOOK TWO

42	†epicuri† OCT. I read et equum vi (Munro).
43	†itastuas† OCT. Read Bernays' pariter, omitted by haplography. No need for the line composed by Bailey as 43a.
159	read unum (OQ), not una (OCT).
252	read motu (OQ), not semper (OCT).
356	read quaerit (Bailey) for OCT †non quit†.
462	read Grasberger's sensibu' dentatum for OCT's †sensibu' sedatum†.
515	read hiemum usque (Munro) for OCT †hiemisque†.
529	read ostendens (Munro) for OCT †ostendam†.
600-601	No lacuna as marked in OCT.
630	read Diels': quos memorant, Phrygias inter si forte catervas instead of OCT reading: quos memorant Phrygios, inter se forte <quod armis>
681-2	lacuna. Supply Munro's: quis accensa solent fumore altaria divom
748-9	lacuna filled by Bailey's: corpora prima omni semper privata colore
749-50	lacuna. 749 should read et omnis, then the lacuna should be filled by Bailey's: res sese mutat, mutat quaecumque colorem
805	OCT reads curalium (red coral) as suggested by Wakefield. OQ reading caeruleum better.
903-4	lacuna filled by Munro: ipsi sensilibus, mortalia semina reddunt
1082	geminam (OQ) preferable to Marullus' genitam.
1174	OCT †scopulum† is correct, not to be emended.

BOOK THREE

84	OCT †suadet† I translate Lambinus' reading fundo.
97-8	The lacuna is filled by Bailey's: at quidem contra haec falsa ratione putarunt
240	OCT †quaedamque mente volutat† I translate the reading of T. J. Saunders (Mnemosyne 28 (1975) 296-8): et quaecumque ipsa.
377	I place a comma before dumtaxat, not after as in OCT.
444	MSS have †incohibescit†. Read Woltjer's emendation incohibens sit.
492	OCT reads vis. I read vi (Brieger),
493	OCT has †animam spumans†. I read anima spumas (Tohte).
658	The MSS reading utrumque, obelized in OCT, is at least arguably correct and has been translated 'both ends of ..."
823	The lacuna is filled by Bailey's: hoc fieri totum contra manifesta docet res
962	I read Marullus' iam aliis for OCT's †magnis† 1012-13. There is no need to postulate a lacuna as marked in OCT.

BOOK FOUR

26-13	OCT attempts to make sense of the lines by transposing groups of lines. I translate the lines in the MSS order and mark where the duplication of thought indicates that the second proem should be discarded.
79	OCT obelizes the MSS reading †patrum matrumque deorum†. I translate Konrad Miiller's personarumque decorem.
127-8	Munro argued that a whole leaf of the text is missing here - a loss of 52 lines of verse. The translation supplies the minimum required to complete the sense.
133-42	No need for Bailey's transposition of 141-2 as in OCT: but Lambinus' transposition of 135 after 132 is correct.
144-5	Lacuna to be filled with a line such as 2.66: expediam: tu te dictis praebere memento.
166	Voss' omnis makes better sense than oris (Q) printed in OCT.
289-90	Lacuna filled by Bailey's:' hoc illis fieri, quae transpiciuntur, idemque.
437	OCT prints Lachmann's undae: better sense is provided by the reading of F undis.
546-7	One of the major cruces in the text. The OCT reads: et reboat raucum †retro cita† barbara bombum, et †validis necti tortis† ex Heliconis Many attempts to emend the lines have been made: I have finally gone for Buchner's buxus cita in 546 and Richter's gelida volucres nocte hortis e fruticosis, both argued for in my edition of Book 4 ad loc.
632	OCT prints Lachmann's umidulum (MSS read umidum). Better by far is Orth's emendation validum. Sturdy stomachs, not digestive juices, are called for here.
638	OCT prints the MSS reading †est itaque ut†. The smallest change required is Howard's et for ut, and I have translated this.
858-76	is printed within square brackets in OCT, but there is no strong reason to doubt the authenticity of the passage.
990	OCT obelizes the MSS reading †saepe quiete†, obviously copied mistakenly from 991. I read Richter's fundere sese.
1026	OCT's puri is weak, much improved by parvi (Clarke).
1130	OCT reads atque Alidensia (OQ), although the only place corresponding to the word (Alinda in Caria) would not give that adjective and did not produce the goods referred to either. A much better reading is Lambinus' ac Melitensia. Similarly Chiaque (OQ) ought surely to be emended to Coaque -Coan garments were de rigueur for the elegant ladies of fashion. (Griffin J.R.S. (1976) 92 -n.1).
1271	The OCT reading pectore is nonsense, the best emendation being corpore (Clausen).

BOOK FIVE

29-30	The OCT reading is correct to posit a lacuna here and to transpose 29 and 30. The translation supplies the minimum sense to fill the intervening gap.
312	OCT reads †quaerere proporro sibi cumque senescere credas†: better with cumque emended by Munro to sene (omitted by the scribe by haplography sene sene-).
396	OCT reads superat et lambens: better to read the MSS reading superavit and emend lambens to ambiens.
412	Purmann's vitas is wrong: the MSS multas needs simply to be altered to multos to make sense.
571	I read the MSS fulgent instead of Lachmann's mulcent adopted in the OCT.
704-5	Lacuna noted by Munro, best filled with Bailey's suggestion: pluribus e causis fieri haec qui posse putarunt (cf. 752).
948	OCT vagi (Lachmann) should be restored to the MSS reading vagis.
1012-13	OCT marks lacuna noted by Marullus, filled by Munro's: hospitium ac lecti socialia iura duobus.
1094	OCT prints Marullus' incita; better to restore the MSS insita.
1442	OCT prints the MSS reading †propter odores†: better to emend to navibus altum with Merrill.
1451	Emend OCT's polire (131) to polita (Bergk).

BOOK SIX

47-8	a difficult crux: the translation merely plugs the lacuna.
56-7	is 90-91, which is 1.153-4: OCT omits the lines here but prints them at 90-91. These lines are identical to 56-7 and 1.153-4 and should be omitted.
131	Voss' magnum is preferable to the MSS reading parvum.
242	The OCT obelizes ciere: but it is a possible reading either in the simple sense of 'displace' or perhaps to suggest the 'summoning forth' of the dead from their tombs.
453	OCT reads Lachmann's moris: better to read modis (OQ).
490	OCT †montis†: I prefer Richter's nebulis.
550	OCT obelizes †es dupuis†: the whole line should read: nec minus exsultant, fissura ubicumque viai. Fissura was first proposed by Rusch.
608-38	Bracketed by Lachmann and OCT because of the abrupt transition in thought. L. might have placed the lines elsewhere in the revision that the poem evidently lacks.
697-8	Munro first noted the lacuna and filled it with: fluctibus admixtum vim venti: intrareque ab isto.
712	Removing OCT's comma after terris makes better geographical sense, as the Nile was not the only river to flood in summer.
762	OCT †poteis†. Emend to primo his (Richter).
778	The MSS reading tactu (obelized in OCT) is still the best reading available.
804	OCT prints MSS corrupt †fervida servis†. Much better is Lambinus' emendation fervida febris.
839-40	Lachmann argued that a whole page of the archetype had fallen out between these lines: but the lacuna has not been universally accepted.
870	Read miscente (OQ) for OCT/Wakefield's unnecessary gliscente.
899	OCT prints Bernays' latentis for the unexciting MSS reading tenentes. I prefer to read Romanes' natantes, agreeing with taedae.
927	The MSS reading auras is preferable to Lambinus' emendation auris printed by OCT.
954-5	The lacuna was first noted by Brieger. I translate Bailey's suggestion to fill the gap: corpora nimborum penetrant et semina nubis
971	OCT reads ambrosia: I prefer Avancius' ambrosiam.
972	OCT prints Lachmann's frondeat esca: I prefer the recent suggestion of M. L. Clarke, fronde virescat.
1135	OCT prints †corumptum†, easily emended to coruptum with the Itali.
1195	OCT prints: in ore truci rictum: a better reading is Richter's in archiatri tactum.
1281	OCT prints pro re <compostum >. I prefer Housman's propere pro tempore.

Go to Next Page