Munro Notes on Lucretius - General
H. A. J. Munro (1819–1885) was the great English editor of Lucretius. His edition (first published 1864) gave the Latin text, a translation, and two sets of notes. “Notes 1” are the critical notes: they record how each reading of the Latin text was established. “Notes 2” are the explanatory commentary.
This page gives the essay with which Munro opens Notes 1, “On the Formation of the Text.” It tells how a single lost manuscript was copied and rediscovered, how the Italian scholars of the fifteenth century and the editors who followed them treated the poem, and why Munro follows Lachmann’s text and spelling. The last pages explain the abbreviations used in the notes (A and B for the two Leyden manuscripts, Lach. for Lachmann, and so on). The line-by-line notes themselves are on the companion page, Munro Notes on Lucretius - By Section.
Source: T. Lucreti Cari De Rerum Natura Libri Sex, with notes and a translation by H. A. J. Munro, fourth edition finally revised, Volume I: Text (London: George Bell and Sons, 1900), pages 1–38. The text was prepared from a new scan reading of that volume, checked against the volume’s older text layer, and corrected by hand against the page images where the two disagreed. Munro’s book references (I to VI) are printed in his small capitals as I, II, III and so on; “notes 1” and “notes 2” are his own names for his two sets of notes. Each printed page can be linked, for example #p12 for page 12.
If Lucretius had come down to us with a text as uninjured as that of Virgil and a few other ancient writers, he could scarcely have been reckoned among the most difficult Latin poets. Certainly he would have been more easy to explain than Virgil for instance or Horace; for he tells what he has to tell simply and directly, and among his poetical merits is not included that of leaving his reader to guess which of many possible meanings was the one he intended to convey. Fortune however has not dealt so kindly with him. Not that the great mass of his poem is not in a sound and satisfactory state: in this respect he is better off than many others; but owing to the way in which it has been handed down, his text has suffered in some portions irreparable loss. It is now universally admitted that every existing copy of the poem has come from one original, which has itself long disappeared.
Of existing manuscripts a fuller account will presently be given: let it suffice for the moment to say that the two which Lachmann has mainly followed and which every future editor must follow, are now in the library of Leyden. One is a folio written in the ninth century, the other a quarto certainly not later than the tenth. Large fragments of one, if not of two others, of the same age as the quarto and very closely resembling it are also still preserved, partly in Copenhagen, partly in Vienna. These manuscripts and at least one more must have lain for centuries in the monasteries of France or Germany, where they found at different periods several correctors, more or less competent. It is to be presumed then that they had some readers, though few if any traces of them are to be met with in the voluminous literature of the middle ages. In my previous editions I said that my friend Professor Mayor had given me a reference to Honorius of Autun in the bibliotheca maxima patrum xx p. 1001, who is there made to quote II 888 in this way, Ex insensilibus me credas sensile gigni, the context proving that he meant to say ne, not me; and asked whether this writer who flourished in the first half of the twelfth century had taken the line from the poem itself; or had borrowed it from Priscian inst. IV 27 who cites it with nasci instead of gigni, the editor of the bibliotheca having thought fit tacitly to substitute gigni from Lucretius. The latter is proved to be the fact by Mr Julius Jessen in the philologus, vol. 30 pp. 236—238: he quotes what follows from Barthius’ very learned note on Stat. silv. II 7 76: ‘nec vero cadentibus aut collapsis iam rebus Romanis auctoritatem suam amisit Lucretius noster, ut videre potes apud Magnentium Rhabanum pracfatione Laudum Daedalarum Crucis, Gulielmum Hirsaugiensem in Institutionibus Philosophicis et Astrologicis, Honorium Augustodunensem in Historia Mundi, Ven. Bedam Libro de Metris. — Referring to the only printed edition of this work of William of Hirschau, who lived from 1026 to 1091, he shews that Honorius copied from him the passage in question, and that William cites it thus: ‘Ex insensili credas sensile nasci’, getting it clearly then from Priscian. Hrabanus Maurus and Beda seem just as little to have known Lucretius at first hand. [Hermes vol. 8 p. 332, it is said that in Brit. Mus. ms. 377 (13th century) of Daniel de Merlai’s Philosophia, in p. 90 a, Lucretius is quoted.]
In Italy he was even more completely unknown. A catalogue which Muratori antiq. III p. 820 assigns to the tenth century, proves that the famous library of Bobbio contained at that time librum Lucretii I; but before the fifteenth no Italian poet or writer shews any knowledge of him whatever. In the year 1414 the celebrated Poggio Bracciolini went as apostolic secretary to the council of Constance and remained on this side the Alps in different countries, Switzerland Germany France and England, until 1420, with one short interval passed in Milan and Mantua, During these years he procured from various monasteries many most important Latin works hitherto totally unknown in Italy: see Mehus’ preface to his life of Ambrosius Traversarius p. xxxiii foll. Among these was a manuscript of Lucretius, obtained apparently from some German monastery either by him or his companion Bartholomew of Montepulciano, about 1417 as his letters seem to indicate, and transmitted the same year to his intimate friend the Florentine Nicoló Niecoli, a most zealous scholar and patron of the revived classical studies. ‘This manuscript, which Poggio wrongly supposed to be only a part of the poem, has itself disappeared, but was the parent of every copy written during the 15th century, that is to say of every one now extant with the exception of those specified above: it must have very closely resembled the Leyden folio. ‘Et te, Lucreti, longo post tempore tandem Civibus et patriae reddit habere suae’ says Landinus in his poem in praise of Poggio. Niccoli having such a treasure in his hands was in no hurry to part with it. We find Poggio writing to him in December 1429 to remind him that he had kept his Lucretius twelve years. A few days later it is ‘you have had Lucretius now for fourteen years; I want to read him, but cannot get him; do you wish to keep him another ten years? He had already tried in vain what coaxing could do: he promises him at one time to send the book back in fifteen days, at another time in one month, if he will only let him have it for so long; but feels sure his book will come to him at the Greek calends. Much as Niccoli loved Poggio, he loved still more to have the sole possession of a newly discovered Latin poet, and I doubt whether Poggio saw his Lucretius at all events before his return from Rome to Florenco in 1434. — Niccoli died in 1437 and left behind him a manuscript written by his own hand and now in the Laurentian library, the truest representative of Poggio’s lost original, as is abundantly proved by the critical notes of the present edition. Between this date and that of the earliest printed editions a knowledge of the poem was diffused through Italy by many incorrect copies. Eight of these, including Niccoli’s, are preserved in the Laurentian library, all of which I have examined, two with care, as being of no small importance for the text; six are in the Vatican, all seen by me as long ago as the autumn of 1849. Of the copies in England I have had in my hands at least seven; one of these belonging to our Cambridge library has been open before me the whole time I was writing my critical notes. Those manuscripts which have been of any importance in forming the text will be more fully spoken of, after the printed editions have been discussed.
The editio princeps, of which only three copies are known, was printed about 1473 by Ferandus of Brescia. It is the only one of the early editions which I do not possess; I have had to trust therefore to the very unskilful collation of Gerard at the end of the Glasgow edition of Wakefield. As it was printed from a manuscript a good deal corrected, but yet inferior to such amended copies as the Cambridge ms. for instance or that which I call Flor. 31, it is of little importance in the history of the text: of far less than the two next editions, since they by accident came to be the foundation of the vulgate. The former of these was published by Paulus Fridenperger at Verona in 1486 ‘die vigesimo octavo septembris calen. octobris’. It was printed from a ms. closely resembling the one written by Niccoli, as may be seen by the most cursory inspection of my critical notes. It is therefore very rude and inaccurate, but being less interpolated than the editio princeps or the majority of existing mss. it represents the archetype more faithfully than these do, though there is hardly a line without some monstrous blunder. The next edition was published in Venice ‘per theodorum de ragazonibus de asula dictum bresanum’ 4 september 1495. From some elegiac verses at the end one C. Lycinius would appear to be its editor, if editor he can be called; for it exactly reproduces for the most part the Verona edition even in the minutest points of its perverse punctuation. There are however throughout the poem not a few differences in the two editions, some of little, others of greater importance; for example IV 125—190 are wanting in the Verona, but not in the Venice. The reason why I dwell on this fact will appear presently.
In December 1500 Aldus published his first edition of our poem, the first systematic endeavour to make it intelligible throughout. The editor was Hieronymus Avancius of Verona, who dates his dedication ‘Kalendis Martii. M.ID’, old style I presume, and really therefore 1500: an interval of twenty-two months between the two dates would not be easy to understand. Avancius is known by other works also, especially the Aldine edition of Catullus. A slight inspection will shew that he took either the Verona or the Venice edition, upon which to form his text; a more careful examination will prove that it must have been the latter. My critical notes will furnish many other instances; let me here only mention that in III 994 he and Ven. have torpedine for cuppedine, while Ver. reads turpidine; 1011 he takes from Ven. its remarkable reading egenus, which Lachmann adopts and wrongly assigns to Marullus: Ver. follows the Leyden and all other known mss. in reading egestas; 1015 he and Ven. have the absurd reading numela for luella, where Ver. has the equally unmeaning biela. Ven. therefore is the ‘ante impressus ‘ spoken of by Aldus. Avancius’ preface shews that for his day he was a good and well-read Latin scholar, and had studied Priscian Nonius and Macrobius for the illustration of his author. Aldus in his prefatory letter to Albertus Pius confirms this, and says that he knew Lucretius by heart, ‘ut digitos unguesque suos’. Avancius in his preface asserts much the same; and the few critical remarks he there inserts shew that this was true at least to a certain extent. At the same time he admits with a seeming candour that owing to the immense difficulty of the work he has left much for others to do. Much indeed he has left undone; and it would have been a herculean task for one man fully to correct the desperately corrupt Venice edition, especially in those days when there were but few extraneous aids and the art of systematic criticism was still in its infancy, two generations having yet to elapse, before it reached its full growth in the hands of the illustrious school of French critics. What he has done however is very great and entitles him to high praise, if it is indeed his own. But this shall be considered presently. The next edition is that of the well-known scholar Ioannes Baptista Pius, published 1511 ‘kal. Maii’ in his native Bologna. Lucretius’ text is embedded in an enormous commentary which displays amid much cumbrous learning no slight acquaintance with the Latin poets, several of which he edited before and after his Lucretius. He thus describes what he has done: ‘contulimus non sine aerumnis vigiliisque diutinis eodicem veneti Hermolai: et Pomponi romani: codicemque non omnino malum: qui servatur Mantuae in bibliotheca quadam suburbana: qui fuit viri non indocti gentis clarissimae Strotiorum. non defuit Philippi Beroaldi praeceptoris quondam mei: nunc collegae: impressus quidem: sed tamen perpense examinatus. Codri quoque grammatici Bononiensis: cuius copia mihi per Bartholomeum Blanchinum virum eloquii excultissimi facta est: Marullique poetae industria mira castigatum non defuit exemplar Severo Monaco Placentino graece latineque perdocto musarum athleta non gravatim offerente’. He makes no mention at all of the man to whom he was most indebted, Avancius: for his text is a reprint of the first Aldine, with however not a few changes of words or phrases, often for the better, often for the worse, either inserted in the text or proposed in the notes, and derived it may be presumed in many cases from one or other of the sources just mentioned. But strange to say when he makes a change in the text, the lemma of his note nearly always contains not this reading, but that of Avancius as if he had meant it to stand: thus I 9 he rightly reads diffuso lumine; but his lemma has diffuso numine with Avancius, which the latter however corrects at the end of his Catullus: and he adds ‘sunt qui legunt lumine’. 15 for capta he wrongly inserts in the text quodque; but his lemma has capta, and his note rightly explains the construction and makes no mention of quodque. 34 his text properly has Zeiicit, his lemma Kefficit. after Avancius; 3D his text wrongly gives suspirans, the lemma suspiciens; and so throughout the poem. This very singular circumstance I explain in this way: he was living a& Rome when his edition was printed and seems to have sent the text and commentary separately; for the bookseller prints at the end a long page of errors with this notice prefixed, ‘Hieronymus Platonicus Bononiensis bibliopola ad lectorem. contuli Pii exemplar cum edito Lucretio: labeculasque pauculas notavi cet.’ Pius edition was reprinted by Ascensius in 1514 with not a few changes in the text, some of them taken from the notes.
The next edition must be ever memorable in the history of Lucretius, that published by Philip Giunta ‘anno salutis. M.D.XII. mense martio. Whether this means 1513 new style I cannot tell; but I know that he dates a Gellius and a Romualdi vita as published in January 1513, ‘Leone pont. max. christianam R.P. moderante’ and ‘ Leonis X anno primo’. Now Leo X only became pope in March of that year; so that here he must be speaking of 1514; and in Florence At all events this mode of dating seems to have been in common use. The editor was Petrus Candidus who, great and important as the corrections are which he has introduced, has yet used a copy of the first Aldine upon which to make them, though he has never mentioned the name of Avancius. It seems to have been the practice of those times to take at least whatever was printed without acknowledgment: thus Giunta regularly made booty of Aldus, Aldus of Giunta in turn. What is said in the present case is grounded on a close inspection of the two volumes. Candidus, where he does not designedly leave him, follows Avancius in the minutest points of spelling and punctuation. The latter for instance says in his preface that he writes ‘veteres imitatus repertumst, itemst, necessest’ and the like: Candidus in his preface that ‘in tam culto, tam nitido, tam undecunque castigato poeta’ he will not admit archaisms like volgum, volnera; or nullast, haudquaquamst and the like. And so in his text while rejecting Avancius’ patefactast, volnere, etc., he keeps his frugiferenteis, rapaceis and a thousand such forms which have no authority in their favour, while those which he discards have much. Lachmann always so hard upon Avancius says ‘huius ineptissimam scribendi rationem Eichstadius studiose imitatus est’, but has not a word of blame for Candidus.
But whence has the latter got his many and brilliant corrections? for few or none appear to come from himself. He says in his address to Thomas Sotherinus that what he did was to collate all the vetusta, exemplaria that were in Florence and to expunge what was condemned by the obeli of Pontanus and Marullus, ‘praestantissimorum aetate nostra vatum?. He refers of course to John Iovianus Pontanus and his friend and pupil Michael Marullus, after Politian two of the first scholars and Latin poets of the most flourishing period of Florentine learning, the latter half of the fifteenth century: ‘Marullo ed il Pontan’ have the honour to be mentioned together by Ariosto in the Orlando xxxvii 8. But Candidus goes on afterwards to speak only of Marullus ‘cuius in hoc opere censuram potissimum secuti sumus’; and in a note at the end he says that in changing the order of verses here and in most other places he has followed the arrangement of Marullus. To Marullus therefore everything which is peculiar to the Juntine has usually been assigned, whether in the way of praise, or of blame as by Victorius and by Joseph Scaliger who inherited among many other of his father’s antipathies his dislike to Marullus. But Lachmann has gone much farther than this, and has given to him not only by oversight, as will be seen in notes 1, much that belongs to older authorities; but everything that first appeared in Avancius’ edition as well, calling the latter ‘fur improbus’ and other opprobrious names. That he got much assistance from the labours of Marullus is certain; but by ascribing to the latter everything that is in the Juntine, in some respects more, in others less credit is given to him than he deserves. As I can throw some light on this interesting question, I will examine it at some length here and in various parts of notes 1.
The scholar, poet and soldier, Michael Tarchaniota Marullus Constantinopolitanus, as he calls himself in the editions of his poems printed during his life, appears from this title and his epitaph in San Domenico of Ancona, where he and so many of his ancestors are buried, as well as from the epithet Bizantius given to him by his friend Petrus Crinitus, to have been born in Constantinople. As he can hardly have passed middle life when he perished in the river Cecina near Volterra April the 10th 1500, he must have been a mere child when on the capture of his native city he was brought to Italy, probably to Ancona. He received his training however in Florence, and he found a Maecenas in Lorenzo de’ Medici. Though he never printed anything on Lucretius, his manuscript emendations appear to have been well known during his life, and a copy of the poet to have been found on him at his death: ‘ex miseranda illa in mediis Cecinae undis Latinarum musarum iactura cladeque insigni unus est Lucretius receptus’ says Candidus in his preface; and his friend Petrus Crinitus in his de honesta disciplina xv 4, published in 1504, but mostly written it would seem before Marullus’ death, after well refuting an alteration of his which shall presently be referred to, adds ‘quae ab eius quoque sectatoribus recepta sunt pro verissimis’. This intense love of Lucretius he seems only to have conceived in the latter years of his life. Candidus, whose preface full of feeling shews that he greatly loved Marullus and deeply deplored his untimely end, strives to make the most of what he did: he says ‘Lucretianae adeo veneris per omnem aetatem studiosus fuit, ut cet) But this must be an exaggeration: the first edition of his poems, published without a date, but not later than 1490, containing only two books of epigrams, shews so far as I can see no trace of any acquaintance with Lucretius. Catullus is chiefly imitated even in the elegiacs, and next to him Tibullus and Horace. Six pages from the beginning there is a poor poem of eight lines ‘de poetis Latnis’ [sic], in which he says that Tibullus Maro Terence Horace Catullus, each in his kind, are the only good Latin poets: Hos si quis inter caeteros pomet vates, Oneret quam honoret verius. The Roman editions of 1490 and 1493 I have not access to; but in December 1497, two years and a few months before his death, he published at Florence a much enlarged edition. A third and fourth book of epigrams are added: in these too I find no trace of Lucretius. Then follow four books of hymni naturales. In these, especially such as are written in heroics, the strain is ‘of a higher mood’, and we meet with frequent imitations of Lucretius, even in the lyrics, as Opibusque late pollens tuis which recalls Ipsa suis pollens opibus. But in these heroics it is to be noticed that the rhythm is Virgilian, not in any respect Lucretian even where he closely follows the latter’s language, as in the hymn to earth: Anfe repentino caeli quam territus haustu Vagiat aetheriam in lucem novus editus infans. Cum protectus humi nudus iacet, indigus, exsors Auwili, infirmusque pedum infirmusque palati. Then imitating at once and contradicting Lucretius’ ut aecumst, Cui tantum in vita restet cet. he goes on Atque uno non tantum 4nfeliv, quod sua damna Non capit et quantum superat perferre laborum. This the last poem published in his lifetime is full from beginning to end of Lucretian phraseology. In this edition too he inserts two new verses in the poem ‘de poetis Latinis’ spoken of above, NVatura magni versibus Lueretii Lepore musaeo illitis, the best in the poem and recalling musaeo contingens cuncta lepore. Crinitus l.l xxiii 7 quotes this poem and mentions a conversation he had with Marullus in which ‘factum est iudicium nuper a nostro Marullo de poetis Latinis egregie perfectum et prudenter’, and Ovid and other poets are blamed; and then it is added ‘itaque legendi quidem sunt omues inquit [ Marullus]; sed hi maxime probandi pro suo quisque genere Tibullus Horatius Catullus et in comoedia Terentius. Vergilium vero et Lucretium ediscendos asserebat?. — Let what has just been said be at once applied to a striking interpolation. After I 15 the Juntine first inserted the v. Illecebrisque tuis omnis natura animantum, which long kept its place in the common editions. Lachmann of course attributes it to Marullus, as do most editors. Lambinus says of it ‘neque eum Naugerius neque Pontanus habuerunt. Marullus unus vir doctus ex auctoritate veteris cuiusdam codicis, quemadmodum mihi religiose asseveravit Donatus Tanottus, nobis eum restituit. amicus quidam meus ingenio et doctrina praestantissimus putat esse ab ipso Marullo factum cet. What his authority is for that which he says of Pontanus I do not know, but Naugerius editor of the Aldine of 1515 properly omits the line, though he in general minutely copies the Juntine. Now this line is written by the hand of Angelo Politian in the margin of a manuscript which belonged to him and forms xxxv 29 of the Laurentian library. Politian died in September 1494, when Marullus could hardly yet have done much for Lucretius; and besides this as he had been long the deadly enemy of Politian, it is not likely the latter would have inserted in his manuscript one of his verses. I infer therefore that it is Politian’s own; and as Candidus says in his preface that he collated all the ‘vetusta exemplaria’ in Florence, he could not have neglected this manuscript which was then in the famous conventual library of San Marco. I conclude therefore that Candidus’ taking it from the margin of Politian’s ms. is the right explanation of lanottus’ assertion that Marullus got it from an ancient codex. It is quite possible indeed that Marullus copied it himself from this ms. which passed to San Marco immediately after Politian’s decease, and thus robbed him of his verse after death, as he is said to have robbed him of his bride during life. Naugerius has in his first page another variation from the Juntine, but that a perverse one: in I 7 he reads Adventuque tuo and joins it with what follows. This corruption I believe to proceed from Marullus; for his hymn to the sun contains a passage evidently imitated from Lucretius: Cum primum tepidi sub tempora verna favoni Aura suum terris genitalem exuscitat auctum: Adventuque dei gemmantia prata colorat: At pecudum genus omne viget, genus omne virorum Perculsi teneras anni dulcedine mentes. I can shew in other cases that Marullus corrupted Lucretius, where he has not been followed by Avancius or Candidus: VI 650—652 are quite correctly given by Avancius, and in his learned preface he says with reference to 652 Nec tota pars cet. ‘totus prima brevi, quia quott redditivus est’. Crinitus Ll. xv 4 quotes 650 651 rightly, and adds ‘qua in re grammaticorum nobis authoritas patrocinatur, quando et centesimus et millesimus probe dicitur: partem multesimam inquit Nonius nove positum est à Lucretio pro minima, ne quis forte paulo incautius atque audacius a veteribus decedat. quae a me vel ob eam rationem sunt adnotata, quoniam Marullus Dizantius aetate nostra, vir alioqui diligens, paulo improbius delere haec et alia pro ingenio subdere tentavit; quae ab eius quoque sectatoribus recepta sunt pro verissimis’. —Candidus gives these two verses rightly and says in note at end of Junt. ‘citatur Nonio locus’: he has got this clearly from Crinitus, who in the same chapter correctly quotes and illustrates I 640 Quamde gravis cet. which the Italian mss. and editions had corrupted: this too Candidus took from him: for Marullus appears to have read Quam gravior Graios inter as does Pius in his notes, and Gryphius of Lyons. Again VI 332 Avancius rightly gives per rara viarum, Candidus perversely after Marullus per operta - see his note. But fifty instances like the last might be quoted. Candidus has also missed some of the best of Marullus’ conjectures: see for instance notes 1 to I 1013 where I have got from the margin of one of the Florentine mss. perhaps the most brilliant example of his critical acumen. Then again unless I greatly err I have shewn in my notes that Gifanius in preparing his edition had before him a copy of the Venice ed. of 1495, lent to him by the zealous scholar Sambucus, as he testifies both in his preface to Sambucus and in his address to the reader. In the former he says ‘exemplum Lucretii ad nos dedisti, non illum quidem calamo exaratum, sed ita vetustum et idoneum, ut vicem optimi manuscripti fuerit, siquidem in eo vidi omnium paene mendorum origines, quae magnam partem a Michaele Marullo, cuius immutationes in eo adscriptae erant omnes, primum parta, mox admiserunt Florentini cet.: in the address he speaks of the ‘Sambuci liber quem ipsius Marulli manu adnotatum magno pretio vir ille praestantissimus paravit’. Why then Lachmann p. 6 should write ‘neque enim facile Gifanio credere possum Marulli ipsius manu annotatum fuisse illud exemplar impressum quod se ab Iohanne Sambuco utendum accepisse scribit’ I cannot comprehend. Gifanius was a dishonest plagiary, but at the same time a most astute man. Why should he tell a gratuitous falsehood which Sambucus would at once detect? He was writing only two generations after Marullus’ death; and even if Sambucus gave his money for what was not the handwriting of Marullus, it was at least a genuine copy of his notes. But notes 1 furnish abundant proof of what I say: see for instance those to I 806 II 16 V 44 and especially III 994. It appears then that Avancius got from Marullus much which the Juntine does not record, and on the other hand, that Candidus took from Avancius without acknowledgment much that Lachmann and others assign to Marullus. Candidus, as I have said above, formed his text on a copy of the first Aldine: in doing this he must have had before him another edition with the ms. notes of Marullus, perhaps the very one which he tells us was found on him at his death. If now all that is common to the first Aldine and the Juntine comes from Marullus, as Lachmann maintains, surely Candidus must have been struck with this coincidence and would have recorded it against Avancius the editor of the great rival publisher. Yet Avancius did borrow largely, very largely from Marullus, especially in the case of interpolated verses made by the latter. How is this to be explained ? Evidently even before his death Marullus’ labours on Lucretius were known; and probably there were more copies than one of these, the one not always agreeing with the other. On this point compare notes 1 to I 551—627, where Candidus makes some perverse transpositions of verses, on the authority of Marullus he says in his note at the end; but the learned annotator of one of the Laurentian mss. states that some put 551—564 after 576, and adds ‘verum Marullo parum referre videtur quomodo legatur’. This annotator and Avancius Pius Candidus Gifanius can hardly all have had the same copy: perhaps all were different. Avancius then may have had his notes in the very copy of Ven. on which he formed his text; and may have looked on them as public property which he might make use of without acknowledgment according to the practice of the time; for neither Pius nor Candidus acknowledges in his turn what he got from Avancius; nor does Naugerius the editor of Ald. 2 say a syllable of Candidus whose edition he copied with few variations.
But Lachmann to III 98 cites in proof of his charge that Avancius was a dishonest plagiary three interpolated verses which doubtless were composed by Marullus and are corruptly given in Ald. 1. In notes 1 to III 98 I have attempted to shew from Gifanius that Marullus perhaps wrote pufarit, and that Avancius intended to read the same: Avancius was probably as good a Latin scholar as Marullus, if less versed in Lucretius. In the line inserted after IV 102 multae for multas may be an error of the printer or an oversight of Avancius. In that inserted after IV 532 there can be little doubt that he purposely wrote suis, imagining that oris was a plural. ‘The correcting of texts was then in its infancy, and Avancius had so grievous a task before him in making sense out of the monstrously corrupted Venice edition, that much must in fairness be excused: we cannot tell what were the exact relations between him and Aldus and his printers. At the end of his Catullus published two years later he has taken occasion to give four pages of Lucretian criticism, in which he has proposed many excellent alterations of his former text, though I do not find that any editor before me has noticed these which are very important for his reputation: see notes 1 to II 422 and other passages. The inference then I draw from all this is that both Avancius and Marullus did much for Lucretius, Marullus doubtless more than Avancius; that much which is peculiar to the Juntine is not from Marullus, and much of what Marullus did is not in the Juntine. Between them they vastly improved a grievously corrupt text; and though they introduced many perversities, we ought in simple justice to take into consideration only what is good. In my notes for obvious reasons, when Ald. 1 and Junt. agree in a reading, I mention both; when a reading is peculiar to Ald. 1, I assign it to Avancius by name; when it first appears in Junt., I still say Junt., though it is always to be inferred that the best readings are most likely due to Marullus. By assigning to him all alike one would often be doing him less, sometimes more than justice.
[In my second edition however I was able to throw fresh light on the history and criticism of Lucretius’ text by the undoubted corrections of Pontanus and Marullus, still existing among the books and manuscripts of Peter Victorius which have formed for centuries so valuable a portion of the Munich library. They were examined by me and copied out in the summer of 1865, my attention having been directed to them by a Goettingen program of Prof. Sauppe. The learned writer informs us that he had examined the Munich ms. of Lucretius and found it corrected throughout by some Italian scholar, Where Candidus the editor of the Juntine mentions in his note a reading of Marullus, this reading invariably appeared among these corrections. From this and other indications he concluded, and the conclusion seemed most reasonable, that these were the very corrections of Marullus which Candidus had used for his edition. He makes the probable suggestion that the long connexion of Victorius with the Giuntas would readily explain his possession of a manuscript which had belonged to that firm.
At Munich through the courtesy of the librarian I had the full use of the following important documents: 1. the manuscript just mentioned: 2. a copy of the Venice edition of 1495 with corrections by Pontanus in the handwriting of Victorius who describes them in the first page as ‘emendationes ex Pontani codice testantis ipsum ingenio eas exprompsisse’: 3. another copy of the same edition likewise corrected throughout by the hand of Victorius who says at the end ‘contuli cum duobus codicibus, altero Ioviani Pontani, altero vero Marulli poetae Bizantii, impressis quidem, sed ab ipsis non incuriose, ut patet, emendatis, quos commodum accepi ab Andrea Cambano patritio Florentino M.D.XX. Idibus Martiis. Petrus Victorius’. What the printed edition was from which he copied these emendations of Marullus I do not know: very likely it was this Venice edition itself which must have had a large circulation and was the very edition containing Marullus’ notes which Gifanius made use of, as has been already told. Victorius says of grando in V 1192 ‘glando in Pont. libro’: now since not only the Venice edition, but the Verona and first Aldine have grando, and also the Brescian as I learn from Earl Spencer’s librarian, Pontanus must have used some printed edition now unknown. The pains which so eminent a scholar as Victorius has taken in copying out twice the emendations of Pontanus and once those of Marullus would prove the high estimation in which those two learned men must have been held when he was a young man of twenty. As he has also filled a copy of the Juntine with long parallel passages from the Greek, he must himself at one time have contemplated an elaborate edition of the poet and has to be added to the long list of scholars with whom this remained an unaccomplished design.
We have then an undoubted copy of what Pontanus himself asserted to be his own original emendations; and as they are accurately repeated by Victorius in his second copy, if we subtract these we have in what remains the undoubted corrections of Marullus. Now the latter with only a few variations, easily to be accounted for as being earlier thoughts or in other ways, all reappear among the alterations of the Munich ms. which are however much more numerous. When we consider all this, and remember that wherever Candidus in his notes mentions the name of Marullus, the reading which he assigns to him is found here; that he tells us in his preface his text is grounded mainly on the revisions of Pontanus and Marullus, the latter more especially; that, as the present edition will demonstrate, the numerous readings which first appear in the Juntine, good bad and indifferent, where not taken from what we now know to be those of Pontanus, nearly always agree with the corrections of this manuscript; and finally that Candidus not unfrequently gives a new reading peculiar to this of all manuscripts known to me, as in his note to V 826 where he mentions pariendo as a variation, we may fairly conclude that Candidus in preparing his text had the use of this corrected manuscript, and that the corrector was Marullus. It would be natural too to conclude that this is his own copy amended by his own hand; and for the most part I do not doubt that this is so. However they cannot all have been written at the same time, as the ink differs in different places; and as so many of the emendations agree with those of Pontanus, it seems not improbable that the ms. was in his possession before it came into the hands of Marullus. As the Italian hand writings of that age resemble each other so much, at least to our eyes, the writing of the pupil may not have differed much from that of the master. However that may be, we must conclude that the corrections common to both belong to Pontanus, as he was the elder and we saw above that he claims them for his own, and the scholar would naturally borrow from the master. The emendations too of Pontanus, valuable as many of them are, have the appearance of being earlier and more rudimentary than those of the other: he not unfrequently too sees that something is wanting and says ‘fragmentum’, where the latter supplies a whole verse with more or less success. The scholar therefore completed what the master commenced; and the emendation of Lucretius links their names together not less honourably than does the verse we quoted from Ariosto. Upon the whole this fresh information has greatly raised my estimate of both, especially of Marullus. His industry is at least as conspicuous as his sagacity: he has evidently carefully collated manuscripts and editions and gathered materials from all accessible sources. Throughout the poem the many verses omitted in the Munich manuscript are supplied with unfailing diligence. He evidently was acquainted with several of the existing Florentine manuscripts; among others that of Niccoli I believe, as well as that of his enemy Politian, and Flor. 31 whose readings Lachmann so strangely assigns to the notary Antonius Marii. Upon the whole he must be placed as an amender of Lucretius immediately after Lambinus and Lachmann, if not indeed in the same front rank, when we consider the circumstances of his age and the imperfection of his materials; and Pontanus perhaps may rank after him. Lambinus, as well as Avancius and Naugerius editor of the second Aldine, must have had access to some copy of Pontanus’ corrections.
What I said of Marullus in my first edition with much more imperfect materials from which to draw conclusions, I find now confirmed in essential points. There were in circulation different copies of Marullus’ emendations; Gifanius had access to one of these: see notes 1 to I 274 Iv 1005, as well as II 16 125 465 III 994 V 201 1151 VI 25: what I inferred from his own poems is borne out by II 719 and 749 and some other passages: the assertion of Crinitus quoted above that Marullus had corrupted VI 651 652 is fully confirmed here: Marullus for the correct multesima most unskilfully proposes multa extima; and for tota reads sit: the latter Candidus adopts. In other cases his more mature judgment as seen in the Munich manuscript doubtless differed from his earlier notions. A man who studied Lucretius so long and earnestly cannot fail to have often changed his mind on further reflexion and with new sources of information. Candidus does not by any means follow Pontanus or Marullus in his orthography: that is formed as I proved in my first edition mainly on Avancius. In many cases they might have taught him better; to avoid for instance such embellishments as his amneis, virenteis. He rightly however avoids such barbarisms as the Aymbres and sylva of Marullus. From succus littus arctus and the like, which the latter carefully introduces, we may infer that he and Pontanus had some share in bringing such corruptions into common use. The careful collation which I have given in this edition of the emendations of Pontanus and Marullus will prove their importance, and shew how rash and unfounded Lachmann’s procedure is in assigning everything that is new in the Juntine to Marullus: even in the many instances where he and I are in agreement, it must be remembered that he speaks without authority, while I possess the testimony of Marullus himself.]
The Juntine closes the first great epoch of improvement in the text of Lucretius: the second Aldine edited by the well-known scholar Andrew Naugerius and dated ‘mense ianuario M.D.XV.’ is for the most part a mere reprint of it without however one word of acknowledgment according to the usage of the time. Yet the changes are not few, mostly for the better, not always: two instances are given above from the first page, the one a gross corruption, the other a right rejection of an interpolation. For the next fifty years Ald. 2 appears to have been the model edition. Gryphius of Lyons published several texts, three of which I have before me: they generally follow Naugerius, but not always, often recurring to Avancius. Those of 1534 and 1510 have many marginal readings, most of them taken from Avancius or the notes of Pius, a few from sources not known to me: see notes ] to I 977 officiatque. Yet even these two editions do not always agree with each other.
Little advance however was made on the Juntine before Dionysius Lambinus. He dates his address to Charles IX 1 November 1563 and afterwards speaks of his first edition as published in that year; though the title-page of my copy has 1564. Lambinus was among the most illustrious of the great Latin scholars who studied and taught at Paris in the sixteenth century. His knowledge of Cicero and the older Latin writers as well as the Augustan poets has never been surpassed and rarely equalled. — Whoever doubts that the nicest critical and grammatical questions can be expressed in Ciceronian Latin without effort or affectation, let him study the commentaries of Lambinus. Scaliger says of him ‘Latine et Romane loquebatur optimeque scribebat’: his ease and readiness are astonishing. He made use he tells us of five mss.: four of these appear to have been Italian mss. of the fifteenth century: the fifth, of which he used a collation by Turnebus, and which he calls the Bertinian, was the same as the Leyden quarto. In his preface and throughout the work he acknowledges his obligations to Turnebus and Auratus. His Lucretius is perhaps the greatest of his works: there was more to be done here, and therefore he has done more. He had moreover a peculiar admiration for this author, of whom in the preface to his third edition he says ‘omnium poetarum Latinorum qui hodie exstant et qui ad nostram aetatem pervenerunt elegantissimus et purissimus, idemque gravissimus atque ornatissimus Lucretius est’. If his boast that he has restored the text in 800 places goes beyond the truth, though I am not sure that it does, yet the superiority of his over all preceding texts can scarcely be exaggerated; for the quickness of his intellect united with his exquisite knowledge of the language gave him great power in the field of conjecture, and for nearly three centuries his remained the standard text. Lachmann says he did much less than Marullus. But so far as there is truth in this, it is merely saying that the one lived before the other: nine tenths of what Marullus effected, Lamnbinus could have done currente calamo; but I doubt whether Marullus could have accomplished one tenth of what Lambinus succeeded in doing. Lachmann accuses him of strange levity and rashness. But it must be remembered that in a short life he got through an amazing amount of work in conformity with the wants of his age. He only gave two years and a half to his brilliant edition of the whole of Cicero: and probably did not spend many more months on his Lucretius than Lachmann spent years. Nor was it possible in that age even for a Lambinus to apprehend the true relation of the mss. of Lucretius to one another. His copious explanatory and illustrative commentary however calls for unqualified eulogy, and has remained down to the present day the great original storehouse, from which all have borrowed who have done anything of value for the elucidation of their author. Scaliger says ‘Lambinus avoit fort peu de livres’: if so he made good use of them, as his reading is as vast as it is accurate, and its results are given in a style of unsurpassed clearness and beauty. His notes observe the mean between too much and too little: he himself calls them brief, while his thankless countrymen, thinking however more perhaps of his Horace than his Lucretius, have made /«mbin and lambiner classical terms to express what is diffuse and tedious. A second and much smaller edition with only a few pages of notes, but with many variations from the first, was published in 1565.
Scarcely could this first edition have issued from the press, when the well-known scholar and jurisconsult Obertus Gifanius of Buren began with systematical and unprincipled cunning to pillage it and convert it to his own purposes. His Lucretius was printed by Plantin of Antwerp in 1566 as stated at the end, though of two copies before me the title-page of one has on it 1566, the other 1565, which is the date of his own address to Sambucus, and of the two privilegia at the end; for the March 1564 of the first must be old style, as it is later than the February 1565 of the second. He brought nothing new to his task, except the ms. notes of Marullus in the old Venice ed. fully spoken of above; for the emendations and readings of Antonius Goldingamus homo Anglus, which he speaks of in his preface, and the veteres libri and the like which occur throughout his book are mere blinds to conceal his thefts from Lambinus. The way in which he contrives at once to bestow empty praise on this scholar and yet to extenuate his merits and put him as a commentator of Lucretius on the same level with other learned men, Turnebus for instance, is a marvel of astuteness. In the preface to his third ed. Lambinus states the truth with great terseness: ‘omnia fere quae in eo Lucretio recta sunt, mea sunt; quae tamen iste aut silentio praetermittit aut maligne laudat aut sibi impudenter arrogat’. Yet so great was the skill with which all this was done that he deceived many and was thought to be a rival worthy of Lambinus. Contrary to what many believe, the age loved brief notes; and his were brief, the other’s copious. Even the great critic of that generation Joseph Scaliger, who well knew the character of the man and accused him of gross deceit towards himself, says ‘Gifanius estoit docte, son Lucréce est trésbon. Lambinus however knew the truth, and his wrath was as signal as the provocation. In 1570 he brought out a third edition greatly improved and enlarged; much of the additional matter however consists in invectives against the aggressor. In a long preface of great power and beauty of style he states his wrongs. There and throughout his commentary the whole Latin language, rich in that department, is ransacked for terms of scorn and contumely. The same charges are repeated in a hundred different shapes with curious copiousness and variety of expression. Gifanius with consistent cunning attempted no public reply to all this. Many years afterwards, when Lambinus had long been dead, a new edition of the other’s book was brought out at Leyden in 1595, in which many additions are made to the brief notes, but not a word is said of the charges brought against him by Lambinus. He was rewarded for his reticence, and for a century or more opinion was divided as to whether he or Lambinus did more for Lucretius. In private he corresponded with the cankered and unhappy Muretus: the two exchanged futile charges of dishonesty against the dead critic, who was far too genuine a scholar to be capable of being a plagiary.. Lachmann so stern with Avancius has nothing to say of this much more flagitious case: ‘qui quo iure’ he observes ‘aut Lambinum aut alios compilasse dictus esset non quaesivi’. Gifanius had no business whatever to edit a poet: he was without poetical taste and grossly ignorant of metre.
For a century after Lambinus nothing was done for Lucretius: the common editions followed either Lambinus or Gifanius. In 1658 the singular labours of Gassendi were given to the world. Deeply versed in the works of the fathers and the philosophy of all ages down to the latest discoveries of Descartes he devoted himself with the zeal of a disciple to the dogmas of Epicurus. The two first of his huge folios are given to this philosophy, and a large portion of them to the exposition of Lucretius. Much that is curious may be gathered from them, and I have perused them with attention; but to say the truth I have not found much to my purpose in them. The author was utterly devoid of the critical faculty, and all that is of value in him on this head is borrowed from Lambinus; as well as the most useful of his illustrations: his corrections of the text are almost without exception worthless. In the 17th century several distinguished scholars, Salmasius J. F. Gronovius Nic. Heinsius Isaac Vossius, turned their attention to Lucretius; but their labours were only desultory. Of the ms. notes by the two last which are in my possession I will speak afterwards. In 1662 Tanaquillus Faber or Tanneguy Lefebvre published at Saumur a text of Lucretius followed by emendationes and notulae. He was a clever but vain man, who seemed to think such work rather beneath him; he takes care however to inform his readers that he spent but little time or pains on it, and had only Lambinus and Gifanius before him, though he owed nothing to either. The truth is that without Lambinus he could not have advanced a step: clever man that he is, he affords a good proof how much Latin scholarship had deteriorated in France during the century between him and Tambinus. Of Pareus, Nardius, Fayus nothing need be said.
Had Bentley in 1689 or 1690 succeeded in his efforts to obtain for the Bodleian Isaac Vossius’ famous library, he might have anticipated what Lachmann did by a century and a half. As he was at that very time working hard at Lucretius, if he had once got into his hands the two mss. now at Leyden, he would at a glance have seen their importance and would scarcely have failed to complete the edition which he was then meditating. The great knowledge of Epicurus’ system which he displayed two years later in his Boyle lectures and his zeal for the recently published Principia of Newton would have aided him in expounding the tenets of the poet. This however was not to be; but his marginal notes published in the Glasgow edition of Wakefield prove what he could have done, if he had gone on with his design. I cannot doubt that Lucretius would have suited him better than Horace, and offered a fairer field for the exercise of his critical divination.
In 1695 there came from the Oxford press a Lucretius edited not by Bentley, but by Thomas Creech Fellow of All Souls, a man of sound sense and good taste, but to judge from his book of somewhat arrogant and supercilious temper. The text is nearly always a reproduction of one or other of the editions of Lambinus: such criticism indeed he seems in his preface to look upon as unworthy of him. His notes are in most cases mere abridgements of those of Lambinus or copied from Faber, and his illustrations are usually borrowed from the former. All this he does as if it were a matter of course, not thinking it necessary either to avow or conceal his obligations. — His interpretatio is his own: how far it is of assistance to a student must depend upon what he seeks for in it. His Lucretius however owing to the clearness and brevity of the notes has continued to be the popular one from that time to the present.
The worthy London bookseller Jacob Tonson published in 1712 a finely printed text with various readings at the end collected from many quarters with a great deal of trouble, some of value, most quite worthless. This I chiefly mention on account of what follows. In 1725 Sigebert Havercamp Professor at Leyden gave to the world his variorum edition in two large volumes. Though his reputation has never been great, my readers will hardly perhaps be prepared for what I am going to say. As Professor in Leyden he had the full use of the two Vossian mss. there, the main foundation of a genuine text: how did he use this advantage, which in profession he makes so much of? The chief feature of his edition is a vast and cumbrous apparatus of various readings, derived from about thirty-one sources professedly distinct. Of these thirty-one twenty-two are simply the various readings of the London edition just mentioned which Havercamp has taken and tumbled into his own without changing the notation. Most of these are of the most futile nature, taken from worthless editions which reprint or ignorantly depart from those of Giunta, Aldus, Lambinus or Gifanius, such as that of Pareus, Gryphius, Fayus, Nardius and the French translator the Baron de Coutures: the more worthless the authority, the more fully it seems to be given. There are also some collations of the mss. of Vossius and that in the Bodleian which it did the London bookseller credit to get together. The nine remaining authorities are these: a certain Basil edition of 1531, its marginal readings, a collation of the Verona edition of 1486, also jottings in its margin from three unknown mss., a second collation of the Bodleian, and lastly the two all important Leyden mss. The two last are the only authorities he has collated himself. How has he performed this task ? he has not noted one reading in three; the most important variations he usually omits; and the readings he gives are as often wrong as right. That which he has borrowed from others and thrown in a lump into his edition is for the most part as worthless as the scribblings of a schoolboy. So incredibly careless is he, that the Vossian collations which he borrows from Tonson are or should be those of his Leyden mss.: see note 1 to V 471 for a glaring instance of a false reading which he slavishly copies from Tonson and ascribes to his Leyden quarto. Nay more the Bm of the London edition and his own X are one and the same Bodleian ms. so that we have this ludicrous result, that the same ms. is cited twice over as two independent authorities. His various readings are therefore not only cumbrously inane, but are a snare and delusion, and have led astray those who like Wakefield have trusted to them. Thus in his hands the two unrivalled Leyden mss. have been worse than useless. What he does himself is always worse done than what he borrows from others, poor as that generally is: he has collated none of the old editions except the Verona, and that was done for him, and better done than he would have done it for himself. Nor are his explanatory notes much better: he has heaped together in a crude mass those of the chief editors; but except in the case of Virgil and Horace and one or two others of the best known poets, indexes to which are in everybody’s hands, he has not even supplied the references to Lambinus’ learned notes who from the circumstances of his age could not himself furnish them; nay in one case he has given Lambinus’ own words as those of Cicero. In his two bulky volumes there is not one week’s genuine work beyond what scissors and paste could do: seldom has performance fallen so far short of profession and opportunity.
There is nothing to detain us between Havercamp and Wakefield who in 1796 and 1797 gave his three volumes to the world, rivalling the other’s in magnitude. Yet the work, such as it is, is his own, and is not a mere slothful compilation from others. Gilbert Wakefield possessed one quality which a critic can ill dispense with, that of despising any amount of authority which did not rest on some real foundation, and refusing to admit that, because a reading had appeared in edition after edition for centuries, it might by that alone claim recognition. He therefore set about a new revision which was to be based on manuscript authority alone; but neither his knowledge nor his industry nor his ability nor his taste sufficed for such a work. He professed to collate five English mss., among them our Cambridge one, and most of the old editions. This task he executed with incredible carelessness. As he had the full use of the Cambridge ms., one might have expected that his collation of it would be done with some care; but it is quite untrustworthy. From this as well as the evidence of his own notes and the nature of the case I infer that his other collations are not more to be depended upon. Had this labour been faithfully performed, it would still have been of little use, as he had no notion of the true relation of these late mss. to one another. He looked on each as an independent authority and thought he could not do wrong, if the words he put in his text were found in one or other of them. Then he had to take from Havercamp the readings of the Leyden mss., and therefore could gain no true insight into their character. As he had no knowledge of the language or philosophy of his author, he undertook to explain whatever words he put into his text in long turgid notes of unmeaning verbiage. His work was got through with a strange precipitancy: when engaged on the first part, he had never read the other parts of the poem; when he came to them, he had forgotten what went before. Morbidly vain and utterly unconscious of the immeasureable distance between Lambinus and himself, he assails the most brilliant and certain emendations of the unrivalled scholar in a hideous jargon and with a vehemence of abuse that would be too great even for his own errors. Thus by some fatality or other, by its falling into the hands of a Gifanius Havercamp Wakefield instead of those of a Salmasius Gronovius Heinsius Bentley, the criticism of Lucretius remained for centuries where it had been left by Lambinus, nay even retrograded. And yet Wakefield did display occasional flashes of native genius, and our notes will shew that not a few certain corrections are due to him; but from the first to the last of his 1200 quarto pages there is not a single explanation of the words or philosophy of his author for which a schoolboy would thank him: so incurably inaccurate and illogical was his mind. Yet owing to the boldness with which he asserted his pretensions he was thought even by scholars to have done something great for his author: he received complimentary letters from Heyne and Jacobs, ‘hominibus modestis et ab omni iudicii subtilitate abhorrentibus’; and more than thirty years afterwards Forbiger in preparing his compilation for the use of the general public took him for his supreme authority. Even later than that so great a scholar as Ph. Wagner often appeals to him in his notes to Virgil. But though long in coming the avenger was to be.
Already in 1832 Madvig in a short academical program, afterwards republished among his opuscula, exposed the futility of Wakefield’s criticism and gave some intimations of the right course to pursue. Stimulated by his example more than one scholar followed up the attack. The most important contribution of this kind was made by Jacob Bernays in an article printed in the Rhenish Museum of 1847. This able paper would have produced a greater effect than it did, if it had not been so soon superseded by Lachmann’s more complete and systematical work. This illustrious scholar great in so many departments of philology, sacred, classical and Teutonic, seems to have looked upon Latin poetry as his peculiar province. Lucretius his greatest work was the main occupation of the last five years of his life, from the autumn of 1845 to November 1850. Fortunately he had the full use for many months of the two Leyden mss. His native sagacity, guided and sharpened by long and varied experience, saw at a glance their relations to each other and to the original from which they were derived, and made clear the arbitrary way in which the common texts had been constructed. His zeal warming as he advanced, one truth after another revealed itself to him, so that at length he obtained by successive steps a clear insight into the condition in which the poem left the hands of its author in the most essential points. Like many other great scholars he seems to have kept few or no common-place books, Resolved to master his subject he perused the grammarians and poets and nearly the whole of the older writers in order to illustrate Lucretius through them and them by Lucretius, and the Latin language by all. He had an almost unequalled: power of grasping a subject in its widest extent and filling up the minutest details. One mark of a great original critic, which eminently belongs to Lachmann, is this: even when wrong, he puts into your hands the best weapons for refuting himself, and by going astray makes the right path easier for others to find. — Another test is this, when his influence extends far beyond his immediate author. Now hardly any work of merit has appeared in Germany since Lachmann’s Lucretius in any branch of Latin literature without bearing on every page the impress of his example. When he is better known in England, the same result will follow here. Though his Latin style is eminently clear lively and appropriate, yet from his aim never to throw away words, as well as from a mental peculiarity of his, that he only cared to be understood by those whom he thought worthy to understand him, he is often obscure and oracular on a first reading. Had his commentary been twice the length it is, it would have been easier to master. But when once fully apprehended his words are not soon forgotten. His love for merit of all kinds incites in him a zeal to do justice to all the old scholars who have done anything for his author; while his scorn and hatred of boastful ignorance and ignoble sloth compel him to denounce those whom he convicts of these offences. In one instance, that of Forbiger, this sternness passes into ferocity: most of his errors that scholar could hardly avoid in the circumstances in which he was placed.
Hermann warns us, when we disagree with Lachmann, to think twice lest we, not he, be in fault. His defects however must not be passed over. “While the most essential part of his work, the collation of the two Leyden mss, has been performed with admirable skill and industry, he has not been so happy in the use of secondary evidence, that of the Italian mss. and the older editions. Much he has taken on trust on insufficient evidence, and much that he had before him he has not always accurately used. Some proofs of this have been given, more will be seen below. But a still more serious defect must be told: he meant his book to be a critical revision of the text, and left to others the task of explaining and illustrating the meaning. So far good: but as the text of an author in the condition of that of Lucretius cannot be always rightly constructed without a sufficient knowledge of his system and its literature, he has not unfrequently strangely blundered and grossly corrupted the poet’s words: for examples of this see I 599—634 II 522—529 1010 foll. V 513—516. His consummate knowledge of the Latin language as well as of the manner of Lucretius in particular enables him often to amend his author with great success. As he wishes too to produce, where it is possible, an intelligible text, many of his corrections he must himself have looked upon as only provisional ^ Yet his greatest admirers must concede that he has not Madvig’s ‘curiosa felicitas’ in emendation. He has however achieved a work which will be a landmark for scholars as long as the Latin language continues to be studied, & work, perfidiae quod post nulla arguet aetas.
Jacob Bernays in 1852 edited a text of Lucretius for the Teubner series. There can be little doubt that carried away by the strength of his admiration for Lachmann he has followed him too faithfully; yet he not unfrequently differs from him. Where he recalls the old reading he is generally right; where he deserts him for a conjecture of his own, he is often very successful Had he prepared a more elaborate edition, as he appears to have once had thoughts of doing, there is no doubt that Lucretius would have owed him much. The impulse given by Lachmann to the study of our poet has called forth numerous papers either inserted in the German philological reviews or published by themselves. Some are of more, some of less importance: my notes will shew where I have been indebted to them. One English publication of eminent merit, as it criticises not the text of the poem, but its matter and poetical beauties, shall be mentioned elsewhere.
To return now to the manuscripts whose history was sketched above. Though I examined the two at Leyden for some days so long ago as the autumn of 1849, what will now be said of them is borrowed from Lachmann who had them in his hands for six months and during that time applied the whole force of his practised and penetrating intellect to unravelling all their difficulties and obscurities. Both, as already mentioned, belonged to the magnificent collection of Isaac Vossius. The older and better of the two is of the ninth century written in a clear and beautiful hand: I call it A. It has been corrected by two scribes at the time that the ms. was written, as Lachmann tells us. One of these is of great importance: in most essential points he agrees minutely with the ms, of Niccoli, the oldest of the Italian mss.; and doubtless therefore gives the reading of the archetype. It will be seen in notes 1 how often I make the united testimony of A and Niccoli to outweigh all the rest. The other Leyden ms. which I call B is of nearly equal importance: it is of quarto size closely written in double columns, apparently in the tenth century. It is probable that it and the ms. next to be mentioned were copied from some copy of the archetype, not like A from the archetype itself. Four portions of the poem are omitted in their place, but come together at the end in this order, II 757—806 V 928—979 I 734—785 II 253—304. Lachmann has demonstrated that these sections formed each an entire leaf of the lost archetype: 16 29 39 115 are the numbers of these leaves. It is manifest then that after A was copied, these leaves of the archetype had fallen out of their places and been put together without order at the end, before B, or the original of B, was copied from it. More will be said on this point presently. B has had several correctors, but all of the 15th century; one a very brilliant critic for his age, to whom are due many of the finest emendations in the poem, as will be seen in notes 1. This ms. was once in the great monastery of St Bertin near St Omer. “Turnebus collated it in Paris and his collation as we saw was used with much effect by Lambinus: it afterwards came into the possession of Gerard John Vossius, Isaac’s father. A large fragment of another ms. closely resembling B in everything double columns and all, except that it is said to be a small folio, not quarto in shape, is now at Copenhagen: it contains book I and II down to 456, omitting however the same sections as B, viz. I 734—785 and II 253—304, and doubtless for the same reasons, because copied from the same ms. from which B was taken. It usually goes by the name of the Gottorpian fragment from the place where it once was. I have three collations of it, one published by Henrichsen in 1846, another in the handwriting of Nic. Heinsius, another in that of Isaac Vossius. Formerly it had a very high reputation: in truth it much resembles, but is more carelessly written than B, and is seldom of much use, except once or twice to confirm A against B. Strangely enough there are in the Vienna library fragments of a precisely similar ms. containing large portions of the later books, viz. II 642 to III 621 inclusive, omitting however in the proper place II 757—806 exactly as B does; then VI 743 to the end; then follow, precisely as in D, the four omitted portions given above, proving this to be copied from the ms. from which B was taken. Naturally enough these Vienna fragments were assumed to belong to one and the same ms. as the Gottorpian; but Dr Ed. Goebel, from whom I have borrowed this description of them, seems to prove in the Rhenish Mus. n. s. xii p. 449 foll. that the two portions now bound together are of different sizes and belonged to different mss. However that may be, the former part seems to be the same ms. as the Gottorpian, and the other if not the same, is precisely the same in internal character; and in either case is of the same, that is of very little value. Probably therefore a more accurate collation would hardly repay the labour.
All other mss. known to exist were, as has been already said, copied mediately or immediately from Poggio’s lost ms. which must have resembled A almost as closely as the Gottorpian resembles B. The most important are among the eight preserved at Florence in the Laurentian library, numbered 25 26 27 28 29 30 31 32 of desk xxxv. 30 was written by Nicoló Niccoli himself, who had Poggio’s ms. so long in his possession: this we are told in the learned Mehus’ preface to his life of Traversarius p.L. As he studied so many of Niccoli’s manuscript letters, he must have known his writing better than anybody else. There are many corrections in a much later hand, but Niccoli himself seems on the whole to have copied Poggio’s ms. faithfully, and not to have made many changes. His ms. therefore, as will be seen in notes 1, is of great value in deciding between A and B. It is unfortunate that Lachmann could make no use of it: I collated it with some care in the summer of 1851: the old Verona and Venice editions have a text closely resembling Niccoli’s. 31 is next in importance to 30, but of a widely different character, having a text much more corrected than Niccoli’s or even Ver. and Ven. It is clearly written and in excellent preservation, and much resembles in general character the manuscript in our public library which I had open before me all the time I was composing notes l, and which is as well preserved and as distinctly written as the other. It excels the Cambridge on the whole, though the latter has many good corrections not in the other. These two therefore I have used as good examples of corrected codices. From whom come the many excellent emendations contained in these mss. is quite unknown. Lachmann used a not very complete collation of Flor. 31, and to it he attributes the corrections which it has for the most part in common with the Cambridge and doubtless some other mss. Having been told too by H. Keil from whom. he got the collation that it was written by Antonius Mari filius, he fills his commentary from one end to the other with the name of this worthy Florentine notary. I can only say that I compared it with ten or more voluminous mss. written in magnificent style and signed by this man between 1420 and 1451 all closely resembling each other; and neither in general appearance nor in the form of particular letters nor in their abbreviations have they any resemblance to the ms. of Lucretius. This scribe’s name therefore I have excluded from my notes. Of the other Laurentian mss. 29 is to be noticed for the marginal annotations of Angelo Politian spoken of above and often referred to in notes 1: it twice over has this note ‘liber conventus Sancti Marci de Florentia ordinis Praedieatorum habitus a publicis sectoribus pro libris quos sibi ab eodem conventu commodatos Angelus Politianus amisit seu qui in morte Angeli Politiani amissi sunt’. 32 has some learned marginal remarks on the first book from which I have derived some facts about Marullus. The six mss. of the Vatican I collated as long ago as the autumn of 1849, but not with much care or skill; yet it will be seen from notes 1 that they have been of considerable service to me: their marks are as follows, 3275 and 3276 Vatic. 640 Urbin. 1136 and 1954 Othobon. and 1706 Regin., at the bottom of the first page of which are the words ‘Nicolai Heimsi. As further helps I have had Gifanius’ ed. of 1595 with ms. notes by Nic. Heinsius which I bought from H. G. Bohn many years ago: it will be seen that I have derived from it some valuable emendations not in Heinsius’ adversaria nor elsewhere so far as I know. It has also a complete collation of A all through, of B in the first four books, and of the Gottorpian fragment. It contains too a complete collation of the codex Modii, which Heinsius denotes by s: he says of it ‘variantes lectiones excerptae sunt ex libello edito Paris. an. 1565 quem Fr. Modius cum ms, suo contulit, ut ipse testatur fine lib. 1 inquiens: Collatus cum ms. meo 26 Jun 1579 Coloniae’: it was lent to Heinsius by Liraeus; Liraeus had it from Gruter, Gruter from Nansius, Nansius from Modius himself. Heinsius says ‘codex Modii non est idem cum B Vossiano, nam pag. 8 [I 227] ubi ex Modiano notatum ad lumina, Vossianus in’. — Heinsius speaks I presume of the small 2nd ed. of Lambinus, as the one which Modius used: it has like others in lumina: if then Modius’ codex is B, either he or Heinsius has made a gross mistake. I have noticed several other instances, where s is made to differ from B; but in these cases Lambinus’ 2nd ed. has the reading which Heinsius gives to s, so that Heinsius may have here been misled by Modius’ negligence. It would seem certainly that s and B are the same: if they are two, then their agreement is very extraordinary, much closer than that between B and the Gottorpian fragment. I also possess a copy of Faber’s Lucretius with a poor collation of A. and B and the Gottorp. as well as many other notes and illustrations in the writing of Isaac Vossius. Havercamp had a copy of the same notes, but has employed them with his usual carelessness. Notes 1 will shew what important use I have made of them: they have enabled me to strip him of several of the most showy feathers with which he had decked himself either from negligence or worse. Spengel, Christ, Goebel and some others have in various journals and publications made much ado about a codex Victorianus as they call it, once belonging to P. Victorius, now in the Munich library, as if it were a rival, or nearly so, of A and B. From the readings cited I see clearly that it is a common Italian fifteenth century ms. neither better nor worse than twenty others, much resembling the Verona and Venice editions and of no importance whatever. [As the reader has been already told, I examined this manuscript myself at Munich during the summer of 1865 and can confidently affirm that what I said of it is the truth. It is much interpolated: its corrections are not so valuable as those of Flor. 31 or I think of our Cambridge manuscript. I have recorded some of its readings in my critical notes under the term ‘Mon.’, and have occasionally spoken of it as the codex Victori. Strange that learned men should have taken so much trouble about its own readings and said not a word of the much more important emendations of Marullus which it contains. This is not the case with Prof. Sauppe in the program spoken of above; but I cannot help citing from its first page a few lines which I read with no slight surprise: ‘unum addo, quod ab aliis nondum quod sciam animadversum coniecturam de victoriani codicis origine propositam valde confirmet. Post l. 3 enim v. 360 versus novem scripti sunt, quos delendos esse homo quidam doctissimus in margine monuit, qui versus 403—411 per errorem hic illatos esse vidisset. ratione autem subducta inter v. 360 et 403 versus bis vicenos senos interpositos esse invenimus, ut facile intelligamus in singulis archetypi paginis versus vicenos senos scriptos fuisse eiusque inter scribendum cum unum vellet scriptorem victoriani duo folia vertisse. in archetypo vero oblongi C. Lachmannus ostendit p. 3. 49. 233. aliis locis eandem versuum rationem fuisse’. Probably before this time the learned writer will have discovered that his arithmetic is at fault and that he has counted forty-two as fifty-two; and that his theory is thus entirely upset. But it is not for such a trifle as that, that I have quoted his words. Does he really mean to assert that this Munich is not like every other fifteenth century ms. a descendant of the one brought into Italy by Poggio? that the long lost archetype was preserved by some mysterious intervention for the special use of the copyist of this codex? Again I would ask whether he looks upon the laborious and sagacious calculations, by which Lachmann demonstrated the number of pages in the lost archetype and the number of verses in each page, as a mere plaything thus to be trifled with; and not rather as a key to unlock many secrets of criticism and not to be understood even without some slight effort of mind. I beg to tell him that the number of lines in the archetype between III 360 and 403 was not either 52 or 42, but 44; that is to say 42 verses of the poem + two headings: and that III 360 did not commence, nor III 402 terminate a leaf; but that III 360 was the fifth line of page 108 of the archetype, and III 402 was the fourth line from the bottom of page 109. Further study too will perhaps make him regret that he has put the unfortunate paragraphs between I 503 and 634 toa fresh torture, and permit him to see that in no part of the poem is the argument or text in a sounder state.—But even while revising the 2nd ed. for the press I found in a recent program by Th. Bergk, with the name of Ed. Heine on the title-page, a fresh attempt to magnify this much vexed Victorian codex: in p. xiv it is said that it ‘solus iusto ordine exhibet libro iv locum antiquitus archetypi schedae paginis inversis perturbatum: nam post v. 298 Atque ea continuo sequuntur v. 323 Servet. …347 Ac resilire, tum v. 299 Splendida…322 Quae sita sunt, denique v. 348 Quod contra. hoc igitur insigne est virtutis documentum, atque possit aliquis inde colligere librum Poggianum ex archetypo descriptum esse, antequam illae paginae inversae sunt’etc. Is it not strange that so definite an assertion should be printed, when in truth this manuscript has the verses in just the same inverted order, in which they are given by the Leyden and all other known mss.? Nay more the corrector (Marullus without doubt) arranges the disordered lines, whether after Politian or not, exactly as Candidus does in the Juntine, who beyond any question adopted his arrangement from this manuscript. And yet the main purpose of the program spoken of is to prove Marullus not to be the corrector of this Victorian codex, and to prove it from this very passage!] It will be seen that by the materials which I have collected and just described I have in many important cases got nearer than has been done before to the readings of Poggio’s ms. which was a worthy rival of the Leyden two.
But Lachmann’s long experience and disciplined acuteness have enabled him to go beyond existing mss. and to tell us much of the lost archetype, as I call it after him, of all existing mss. Notes 1 will shew that many difficulties are cleared up by this knowledge. This archetype then, though it is not certain that even A was immediately taken from it, was written in thin capitals, like the medicean of Virgil; the words were not separated, but in the middle of verses points were put at the end of clauses. Ancient mss. as a rule keep with singular care to the same number of lines in a page: ours had 26 lines in a page, excepting only those which concluded a book. But remember there was a heading or title at the beginning of each section; and each of these headings occupied a line. Lachmann brings many proofs of this being the number. When this ms. was copied, it was clearly much torn and mutilated. It was stated above that four portions, omitted in their place by B, come together at the end, and that these each formed a leaf of the archetype which had fallen out of its proper place. Each of these alone or with its headings consists of 52 lines. Then turn to note 1 on IV 299—347 (323—347 299—322) where this inversion is explained in the same way, by the accident that is of a loose leaf being turned the wrong way: see also note 1 to I 1068—1075 and 1094—1101, where the mutilation is accounted for in the same manner. Thus we obtain six certain landmarks in different parts of the poem. The archetype therefore consisted of 300 pages, or admitting, as seems to be an undoubted fact, that a whole leaf is lost between VI 839 and 840, of 302; of which the first was not written upon, as well as one for some reason or other somewhere between I 785, which ends one of the loose leaves at the end of B, and 1068 which, as shewn in note 1, begins a fresh leaf. Page 190 which followed the end of IV was left blank. I may also note that pp. 137 and 191 contained an index of the headings of IV and V respectively, while the headings of VI are crowded into the lower part of p. 249, the upper part of which contained the last 13 lines of the text of V: see Lach. p. 398: although the different titles come in their places in these books too, as well as in the first three which have no such index prefixed. Having made for myself a list of these pages after the rules stated in various places by Lachmann, I have found it of great use; as the ends of lines throughout the poem towards the bottom of the several right-hand pages had been specially exposed to mutilation in the damaged archetype. Verses also omitted in their proper places were apt in this as in other mss. to be put afterwards at the bottom of pages. Besides the injuries which it had received from accident or ill usage, our archetype must have been carelessly enough written, though A and B prove that it retained many valuable vestiges of great antiquity, especially in the spelling of words, and though there may have been few stages between it and the age of the author. There is one point, the nature of the hiatus after IV 126, as to which it is not easy to accept Lachmann’s theory. That there is a hiatus there, is indisputable and the special questions connected with it are fully discussed in note 1 to that passage. As the accidental loss of a whole leaf would not suit his system of pages, he boldly declares that twenty-five verses and one heading have perished, that is one single page of our archetype. Now it is easy enough, as we have seen, to explain the accidental loss of a leaf, by which every subsequent copy must necessarily want the contents of that leaf: it is easy enough to conceive any one ms., A or B or Poggio’s, passing over by mistake one whole page. But it is in the highest degree unlikely that different copies, A B and Poggio’s, neither of which as Lachmann admits was copied from the other, should all pass over a single page of their original; or that this single page should be wholly illegible, while that which preceded and the reverse page of the same leaf should be entirely uninjured. It seems to me therefore much more natural to assume that our archetype or one of its predecessors accidentally omitted an uncertain number of verses; or rather that a whole leaf of the archetype had been lost, as after VI 839. Lachmann’s system of pagination would then be set right in this way: only books IV and V have an index capitum prefixed filling one page; for that of vi as I have said is crammed into the lower part of the last page of V: before this index in V Lachmann has shewn that the archetype had one blank page. Assume now that one page was similarly left blank before the index of IV and all will be right: the pages of the archetype would then be raised to 304. The index capitum prefixed to VI I accidentally omitted to notice in former editions, as it had no bearing on the question of pages. The assumption here made which is commended by Mr Polle, but blamed by Mr Susemihl in philolog. xxix p. 427 foll., I still think probable. The latter asks what conceivable reason there could have been for the two blank pages before V, except to begin the new book with a new leaf, as was done with all the rest. Why, I which had no index prefixed began on the second page of a leaf; so did III, which had no such index; so did VI whose index is crowded together in the manner spoken of. Other mss. such as the medicean of Virgil, seem to have no preference for beginning a book on a new leaf. Why the two pages were left for the index between IV and V I do not know, any more than why a page was left vacant somewhere between I 785 and 1068: it had something to do perhaps with calculations about the parchment required. I was going to say more, but forbear to enlarge on so fruitless a topic.
But we are able to advance even beyond the archetype: in many parts of the poem there are manifest undoubted interpolations, which must have been inserted by some reader who wished at one time to confirm what is said, at another to convict it of inconsistency and the like. Generally, not always, these passages are repetitions of genuine passages; sometimes they consist of several, sometimes of a single verse: I 44—49 and III 806—818 are good and incontrovertible examples. But enough is said of these throughout our notes. Lachmann however still unsatisfied has not paused even here, but has gone up to the very times of the poet. No careful reader will refuse to admit that he has proved not a few passages, some of them among the finest in the poem, to have been subsequent additions made by the author, which he did not live to embody properly with the rest of his work. Lachmann has gone too far; and unless I err, I have shewn that not a few sections thus marked by him are properly connected with what precedes and follows. Yet it is certain that his theory applies to II 165—183, and more than one long paragraph of IV V and VI. It has been shewn sufficiently in the notes to these passages that the most important of them have a close connexion in matter and manner with each other. Like Lachmann, I have marked them off by [ ] All through the poem many single verses and passages of some length are designedly repeated by the poet, some of them again and again. It is probable that he would have removed many of them, if he had lived to revise his work: the exordium of IV for instance could hardly have been left.
Some readers may be surprised at the number of verses which have been transposed in the poem; but they should remember that every ancient writing which depends finally on one ms. is in a similar plight. When a scribe omitted by accident a verse, in order not to spoil the look of his book, he wrote it at once after the next verse, if he immediately discovered his error; if not, he omitted it altogether, or added it in some other place, often at the bottom of a page; he would then affix an a, b to mark the right order; the next scribe would not notice or would purposely omit these and so on: see Bentl. to Hor. ars 46. Every one of these errors has been committed again and again by the copyists of our poem. — Most of these transpositions are certain and were made long ago by Lambinus Marullus Avancius and others; many were first made by Lachmann. Some of these I have not followed: not a few I have first ventured on myself. But connected with this question I must draw attention to one point which seems of importance. You would expect as a rule single verses to be thus transposed; and this is the case in Lucretius’ mss. as in those of other writers: sometimes too one or more verses are repeated after the misplaced verse, which ought to follow it in its proper place, as if to shew the reader whither it ought to be transferred: comp. IV 991 i.e. 999 of the mss. followed in them by 1000—1003, which are only the vss. which follow it in its right place repeated after it in its wrong place: see also V 570 (573) and what comes after. But besides such usual instances of transposition there are throughout the poem many small groups of verses, forming generally sentences complete in themselves, which have got quite out of their right place: comp. I 984—987 (998—1001), II 652—657 (655—659 680) and iv 1227 1228 (1225 1226), three passages first transposed by me; also II 1139—1142, 1168—1170, III 686—690, IV 50—52, V 174 175, 1127 1128. Now that a seribe should so often transpose several consecutive verses always forming an entire and independent sentence by mere casual carelessness, is to me in the highest degree improbable. Again most of these passages read to me like possible additions not necessary to the context, though they improve it. I believe them then to be marginal additions by the poet, inserted on the same principle as the longer sections discussed above: these too the first editor, faithfully preserving everything in his copy, but not caring always to find the right place for what the author left ambiguous, has inserted out of their order. Add to these V 437—442 which the context could dispense with: these vss. are found out of place in Macrobius as in our mss. This increases the probability that they were out of their proper order from the first, two apparently independent authorities Macrobius and our archetype quoting them in the same way. Perhaps these single vss. might be added to the list, IV 202, 205, VI 957, 1225, 1237, 1270 as they might all be dispensed with. Look too at IV 129—142, so strangely disordered in the mss.: 133—135 may be all marginal additions by the author afterwards wrongly placed by the editor. The ms. arrangement of IV 299—348 has been already accounted for. If all these passages are subtracted, there will then be left a not very unusual number of single verses transposed by the ordinary negligence of copyists. The numbers occasionally given on the left hand of the page denote of course the order of the lines in mss. which Lachmann follows in his edition: where spurious vss. of the mss. are omitted from the text, he still allows them to count. For obvious reasons I have followed him in this, as he will be the future standard of comparison, and there is great advantage in a uniform numbering of the verses.
Since many special questions of orthography are noticed as they occur in the notes, I should have thought it unnecessary to say more in this place than that in essential points I follow Lachmann, if it were not for the apparent unwillingness of scholars in this country to accept even the smallest change in what they look upon as the usual or conventional rules of spelling. The notion of any uniform conventional spelling is quite a chimera: I never find two English editors following any uniform system; nay the same editor will often differ in different parts of the same book. But whence comes this ‘conventional’ system, so far as it does exist? from the meritorious and considering their position most successful endeavours of the Italian scholars in the fifteenth century to get rid of the frightful mass of barbarisms which the four or five preceding centuries had accumulated. They sought indeed to introduce rigorous uniformity in cases where variety was the rule of the ancients; and though these cases embraced only a few general heads, they yet comprised a great multiplicity of particular instances, because involving the terminations of cases, the assimilation of prepositions in compound verbs and the like. But where there was only one right course, they generally chose it; yet from the utter confusion into which the use of the aspirate had fallen, their own language having entirely lost it in sound, but at this time retained it in spelling; from the almost complete identity both in sound and writing of c and t before i, and the like, they never could tell whether humor or umor, humerus or umerus, spatium or spacium, species or speties was correct; and consequently as a rule chose the wrong. Their general principles however were not accepted by the most thoughtful scholars in any age, so far at least as concerned the text of ancient authors, unless it be during a part of the present century; neither by an Avancius in the 15th nor by a Lambinus or Scaliger in the 16th nor by a Gronovius in the 17th nor by a Bentley in the 18th. Yet this system gradually established itself, because it came to be used by scholars in their own writings, some of the barbarisms being gradually eliminated; new ones however being introduced, such as coelum coena moereo sylva caetera for caelum cena maereo silva cetera in order to derive them preposterously from Greek words: Marullus, as we said above, writes sylva and hymbres.
Many attempts were made in various directions to change this state of things: the best and most systematic was that of Ph. Wagner in his orthographia Vergiliana published in 1841. With admirable industry he amassed all the evidence afforded by the medicean and, so far as it was accessible to him, of the other ancient mss. of Virgil. As these, like other old mss. are as a rule very tenacious of the true spelling in those cases where there is only one right method, he performed this part of his work with eminent success, and still remains one of the best authorities on the subject. In those other cases however referred to above, in which variety is the rule of the ancients and which include a great multitude of particular instances, he has chosen to abandon the safe ground of evidence and experience and has made Virgil write what he decided on a priori principles he must have written. This seems to me the reason why his system was not more generally followed. Still less satisfactory was Madvig’s spelling in his de finibus published in 1839: it was utterly unlike that of the mss. and yet in many points it was not what Cicero used; in still more you could not be sure whether it was what he used or not. Here too Lachmann bringing into play his extraordinary ‘power of asking the right question,’ and joining with it a minute knowledge of the whole evidence upon the subject, saw at once what could be obtained and what could not, and shaped his course accordingly. The Leyden mss. of Lucretius, imperfect in many respects, are on the whole admirable in their orthography, at least equal to any of the mss. of Virgil, confirming them in what is true and confirmed by them in turn: in some nice points, such as the frequent retention of the enclitic st, they far surpass them. With their aid he was able to confirm those improvements in spelling which Wagner had so well established in opposition to the system in common use. But in regard to the other class of words in which the usage of the ancients varied in different ages or even in the same age, he did not dogmatically determine what his author wrote and thus close the door to all future change; but knowing that certainty was not here attainable, he carefully sifted the evidence offered by his mss. and made the best approximation he could to what his author might have written, always taking the most ancient form for which his authorities supplied any testimony direct or indirect. Thus the question was not foreclosed; nor were we left to vague generalities, but a firm historical groundwork was gained upon which future improvements might be built, if better evidence hereafter offered itself. Lachmann then in this, as in so many other departments of philology, seems at once to have produced conviction in the minds of the majority of the most thoughtful scholars, in Germany I mean; for in our own country most seem to scout the question as unworthy of serious attention: a great mistake; for Latin orthography is a most interesting and valuable study to those who care to examine it, and touches in a thousand points the history grammar and pronunciation of the language. Let me give two examples of the effect at once produced by Lachmann. Otto Jahn in 1843 published his elaborate edition of Persius in which he adopted throughout the spelling then in common use, though he had so many excellent mss. to guide him to a better course: in 1851, the year after Lachmann’s work came out, he published the text of his Juvenal and followed in it most minutely the principles of Lachmann; and fortunately he had a most excellent authority in the codex Pithoeanus; so that the spelling is probably not very far removed from the author’s own. In the years just preceding Lachmann Halm published several orations of Cicero with elaborate critical Latin notes: and yet, though his spelling was somewhat better than that of Jahn’s Persius, it is still essentially ‘conventional’ and arbitrary: in the years following Lachmann he published a series of school editions of Cicero’s orations with brief German notes, and now in these the spelling was wholly modelled on the system pursued by Lachmann. The same system too he has carried out in those volumes of the elaborate edition of Cicero edited by him and Baiter, which came out after Lachmann’s Lucretius. Stimulated by the examples of Madvig Ritschl and Lachmann the rising generation of German scholars has pursued the critical study of Latin with eminent success; and nearly all of them follow in orthography the guidance of Lachmann. This system then may fairly I think be now regarded as the true ‘conventional’ system; for surely the Lachmanns and Ritschls of the nineteenth century have a better right to dictate to us in the present day what shall be accepted as ‘conventional’ than the Poggios and Vallas of the fifteenth. Since my first edition came out, Madvig has published the last books of his text of Livy, of which there exists but a single and very ancient ms.: these he has edited in a form differing from that of the other books, and has now given his very weighty authority in favour of adhering to the spelling of the oldest mss., with some reservations which I do not understand. And now too in his new edition of the de finibus he has entirely abandoned the orthography of the other.
In following Lachmann then I am sure that I have authority on my side; I believe that I have reason as well. In those cases indeed to which I have already referred, where the universal testimony of inscriptions and of mss. beyond a certain age proves that there is only one right way and about which the best scholars are all now agreed, there cannot be any doubt what course should be taken: we must write querella loquella luella sollers sollemnis sollicito Iuppiter littera quattuor stuppa lammina; on the other hand milia conecto conexus coniti conixus coniveo conubium belua baca sucus litus and the like; condicio solacium, setius artus (adj.) autumnus suboles: in many of them an important principle is involved: obeying the almost unanimous testimony of our own and other good mss. we cannot but give umerus umor and the like: also hiemps. I have heard it asked what then is the genitive of hiemps: to which the best reply perhaps would be what is the perfect of sumo or supine of emo. The Latins wrote hiemps, as they wrote emptum sumpsi sumptum and a hundred such forms, because they disliked m and s or t to come together without the intervention of a p sound; and our mss. all attest this: tempto likewise is the only true form, which the Italians in the 15th century rejected for tento. Then mss. and inscriptions prove that d took an n before it, tandem quendam eundem and the like, with the exception of circumdo in which the mss. both of Lucr. and Virgil always retain the m: and generally, though not invariably, m on the other hand remained before q: quemquam tamquam and so on; though the new corpus inscrip. Lat. has I find nunquam; and so has Augustus in his res gestae, but quotiescumque. Then always quicque quicquam quicquid (indef.), but generally quidquid (relative), though quicquid is found in the lex Rubria and once in our AB: always peremo interemo neglego intellego etc. Above all we must scout such barbarisms as coelum moestus sylva caetera nequicquam. In these points Wagner is as good a guide as Lachmann; but in regard to the cases in which ancient usage varied shall we follow the former who deserts the mss. for preconceived general rules, or Lachmann who here also is content to obey the best evidence he can get? I have unhesitatingly come over to the views of the latter: ‘hypotheses non fingo’ should be the rule in this as in other matters. As said above, all these uncertain spellings fall under a very few general heads. One of these is the assimilation or non-assimilation of prepositions: inpero represents the etymology, impero the pronunciation of the word. From the most ancient period of which we have any record, centuries before Cicero or Lucretius, a compromise was made between these opposing interests: words in common use soon began to change the consonant, those in less common use retained it longer. In the first volume of the corpus inscrip. Lat., the most recent of which are as old as the age of Lucretius, most of them much older, imperator occurs 26 times, and is always spelt with m, proving that in a word, which must daily have been in everybody’s mouth, etymology in remote times yielded as was natural to sound: imperium again occurs three, inperium six times, being doubtless in somewhat less common use. Now in Lucretius imperium impero or imperito occurs six times, and the mss. always spell with m, and so Lucretius spelt I have no doubt: indeed many of these common words the silver age I believe more frequently wrote with n, than did that of Cicero. Then Virgil uses imperium 40 times; and Ribbeck’s capital mss. have m in every instance, except M which twice has inp., though one of these two cases is at least doubtful: for Aen. VIII 381 Fogginius prints imperiis. Yet in defiance of all this evidence Wagner gives us inperium, surely without reason on any view of the case; for the foundation on which we must build is thus withdrawn from under our feet. To take another common instance, commuto occurs 9 times in the corpus inscrip. and always with m; 12 times in Lucretius and always with m. Other words are more uncertain: we find in the mss. impius and inpius, immortalis and inmortalis, conligere and colligere, compleo and conpleo; and so with other prepositions, ab, ob, sub, ad: all tending to prove that usage was in most words uncertain. Again we have exsto, and exto, exsolvo, exulto expiro expecto cet., s being generally omitted; and this agrees with Quintilian I 7 4 who implies that it was a learned affectation of some to write exspecto in order to distinguish ex and specto from ex and pecto; it agrees too with all other good evidence: the mss. of Virgil furnish precisely the same testimony as those of Lucretius; yet Wagner in all such cases writes exs: surely we should keep ex where the mss. keep it, exs where they have exs: and so with supter or subter, suptilis or subtilis, ab- or ap-, ob- or op-, sub- or sup-, succ- or susc-, and the like: appareo occurs ten times in Lucr. and is always spelt thus by our mss.: so apparo, appello (both 1 and 3 conj.); but adpetitur and always adpono, adporto or atporto, in which words the separate force of the preposition continued to be felt: in exact conformity with this the first volume of the corp. inscr. has twenty times appareo and also apparitor, proving that in the earliest times the prepos. had been assimilated in this common technical word: thus too in the 21 instances of appareo in Virgil all Ribbeck’s mss. always have app. except M once, Aen. XI 605, misled by the usage of its age: comp. the suggestive remark of Servius to Aen. I 616 ‘applicat: secundum praesentem usum per d prima syllaba scribitur: secundum antiquam orthographiam …per p’: yet in defiance of all this Wagner makes Virgil always write adpareo and the like. We find haud and haut, and sometimes aliut aliquit quicquit and the like, sound and etymology carrying on an undecided battle in the mss. of Lucretius, as in inscriptions and elsewhere: adque is sometimes but rarely found, sound having here as might be expected gained the victory: Wagner cannot be right in always forcing adque on Virgil. Lucretius seems to have recognised only sed: he once has elabsa, and once praescribta: see notes 2 to VI 92: in such forms sound must have at an early period prevailed; and b d g gave way to p t c before s and t: lapsus for labsus is the same principle as rex (recs), rexi (recsi) written sometimes recxi, rectum from rego: to judge from the best mss. labsus and the like became again much more common in the silver age. Mommsen has recently published an admirable copy and exposition of the res gestae of Augustus from the Ancyra monument. Augustus was somewhat of a purist in spelling and cashiered an officer for using in a dispatch the vulgarism isse (not ixe) instead of ipse. His system quite bears out what has just been said: he always writes imperium and imperator; he has immortalis, but inmissus; impensa, while the heading of the work, not written by him, but perhaps by Tiberius, has inpensa: generally conlega and conlegium, but once collegium and collaticius; exilium, but exstinguere; on the other hand sexsiens as well as sexiens, proving that x and xs were identical: he writes appellaverunt; but adque the only time he uses the word.
Another question involving a multitude of details is the use of -is or -es in the accus. plur. of participles and adjectives and substantives whose gen. plur. ends in ium, as well as of some other classes, doloris or dolores, maioris or maiores: here too Wagner involves himself in inextricable perplexities by his eclectic system, when his mss. were admirable guides, had he chosen to follow them. The mss. of Lucretius are no less admirable and probably represent very fairly the author’s own usage: they offer -is five times out of six; and -es is somewhat more common in substantives in very general use, as ignes vires aures. Inscriptions quite bear out our mss.; and the sole relic of Latin yet disinterred from Herculaneum contains this v. Utraque sollemnis iterum revocaverat orbes. Pertz recently printed in the Berlin transactions the few remaining leaves of a ms. of Virgil, which he assigns to the age of Augustus and which may really be of the second or third century: we there find the acc. plur. of adjectives and participles ending 18 times in -is, 3 times in -es pares felices amantes; of substantives we find sonoris, but 4 times vires, and artes messes crates classes aves, quite confirming the testimony of our A and B. Varro de ling. Lat. VIII 66 writes item quod in patrico casu hoc genus dispariliter dicatur civitatum parentum et civitatium parentium: in accusandi hos montes fontes et hos montis fontis; and in Lucr. II 587 we find potestates, V 1239 potestatis: then ib. 67 he says quid potest similius esse quam gens mens dens? quom horum casus patricus et accusativus in multitudine sint disparilis; nam a primo fit gentium et gentis, utrobique ut sit i; ab secundo mentium et mentes, ut in priore solo sit i; ab tertio dentum et dentes, ut in neutro sit i; well our mss. six times have the acc. gentis, never gentes; dentes four times, never dentis; mentes five times, once only, II 620, mentis. As for the nomin. plur. of such words, Varro l. l. 66 says sine reprehensione vulgo alii dicunt in singulari hac ovi et avi, alii hac ove et ave, in multitudinis hae puppis restis et hae puppes restes: the fragment of Virgil just cited has the nomin. plur. putris and messis, though we saw it had messes in the accus.: in accordance then with these high authorities the mss. of Lucr. not unfrequently retain this nomin. in -is, which it would be monstrous to extirpate: I have always therefore kept it. In precise conformity with Lucr. Augustus has in the accus. agentis and labentes, finis and fines, consules and once consulis: once too the nomin. pluris. On the other hand he always uses the accus. gentes, departing in this word from the rule of Varro and Lucr. We see from the corpus inscr. that -eis -is -es were all in use: it is probable that Lucr. occasionally employed the termination -eis, intermediate in sound between -es and -is; but, if so, his manuscripts have left few or no traces, and it would be most perverse to follow Avancius Wakefield and others in thrusting it into his verses in season and out of season: V 1280 B has mortaleis, perhaps from Lucr.: Augustus more than once has this -eis in the abl. plur., quadrigeis, emeriteis; and the inscription in his honour still existing on the arch of Rimini erected in 727, midway therefore in time between his res gestae and Lucr., has celeberrimeis, vieis, redditeis. A and B have however not a few indications expressed or implied of the ending -ei: see note 2 to III 97 oculei: these have of course been carefully preserved.
On another question, comprehending a multitude of particular instances, I have followed Lachmann and our mss. which here too are on the whole excellent guides: I speak of the vowel or consonant u followed by another u. The old Latins appear to have been unable to pronounce uu; and therefore the ancient o long kept its place after u; or for qu c or q was used: quom qum or cum, never quum; linquont linqunt or lincunt; sequontur, sequntur or secuntur; equos (nom.) equs or ecus; volgus divos divom aevom and so on. They appear to have begun soonest to tolerate uu in terminations, when both were vowels, suus tuus and the like. Now the mss. of Lucretius have retained in very many instances divom volnus volgo vivont cet.; equos (nom.) and ecus, ecum, aecum; relinquont relinqunt or relincunt oftener than relinquunt, so sequontur secuntur secutus locuntur locutus; but with Lachmann I retain the uu, when the mss. offer it, in order not to get lost on a sea of conjectural uncertainty like Wagner and some others, who not only desert mss. but in many cases intrude a spelling older than the age of their author: thus Augustus has rivus rivum annuum (not once -uo or -vo); why not then Virgil, or at least Varius and Tucca? The mss. of Lucretius are also very pertinacious in retaining the genuine old forms reicit eicit or eiĕcit cet. and never offering reiicit eiicit and the like: Grai Grais, not Graii Graiis. But further details on the most interesting points of the ancient orthography will be found in various parts of my notes. Again in those many cases where the sound was intermediate between u and i and the spelling therefore uncertain, such as the penult of superlatives and certain other adjectives, and words like lubet or libet, dissipat or dissupat, quadrupes or quadripes and many others, I have of course submitted to the guidance of our mss. which only once for instance have u in the superl. and once manufesta, and twice offer arbita, not arbuta. Augustus, having learnt it probably in boyhood from the all-accomplished dictator, for whose apprehension nothing was either too little or too great and who, Gellius tells us, first introduced the i for u in superlatives, invariably writes frequentissimus septimus vicensimus, as well as finitimus manibiae: comp. what Suetonius says of his use of sĭmus for sumus, probably from a wish to be consistent. He would scarcely have thanked Varius and Tucca for bringing him the Aeneid embellished with Wagner’s maxumus, septumus and the like, introduced so often in spite of his mss. [Both forms are found in the same sentence lex colon. Genetivae c. 66 (Ephem. Epigraph. vol. III p. 93) optima lege optumo iure.] I have likewise followed AB in the adoption of e or o in vertere or vortere and the like: e is naturally the more common, yet vorti vorsum divorsi vortitur convortere vortex are all found: also in reading reddunda gignundis dicundum cernundi faciundum agundis cet. or the more usual agendum quaerendum cet. Do I then claim in all these doubtful cases to reproduce the spelling of Lucretius or his first editor? Certainly not; but still in most of them Lucretius and his contemporaries undoubtedly allowed themselves much latitude; and I have not intentionally permitted anything to remain which might not have been found in one or other ms. before the death of Virgil. By adhering tenaciously to the mss. where not demonstrably wrong one gains a firm resting place from which to make further advances, if better evidence offer itself. However that may be, I cannot bring myself to accept the arbitrary and eclectic system of a Wagner, much less the hideous barbarisms of a Wakefield; nor on the other hand, after feasting on the generous cereals of a Lachmann and a Ritschl, can I stomach the ‘conventional’ husks and acorns of the Italians of the 15th century. At the same time it will be seen that my spelling differs less from this system, than does that of Wagner in his standard text of 1841, or even his subsequent modification of that text for common use which Prof. Conington has adopted in his Virgil.
Most of the abbreviations and marks used in the notes are sufficiently explained above: A and B denote the two Leyden mss., Gott. the Gottorpian fragment, Nicc. the Florentine ms. written by Nicolò Niccoli, Flor. 29, 31, 32 the mss. of the Laurentian library forming nos. 29, 31, 32 of desk xxxv; Camb. our Cambridge ms.; Vat. or Vatic. the Vatican mss.; and Urbin. Othob. or Reg. with the number attached identify more nearly the mss. contained in those several departments of the library: one Vat. 2 Vat. 3 Vat. mean one, two or three of the Vatican mss. where it was not worth while specifying them. In this new edition Mon. denotes the codex of Victorius in the Munich library. Brix. Ver. Ven. Ald. 1 Junt. Ald. 2 are the editions fully described above, where it is explained when and why the names Avancius, Candidus, Marullus, Naugerius are or are not used instead of that of one or other of these editions. The ms. notes of Heinsius and Vossius, which are often cited, indicate the notes by those scholars which are in my private possession and have been described above. Lamb. Wak. Lach. Bern. Bentl. need no explanation after what has been said; and in this edition Pont. and Mar. designate Pontanus and Marullus, whose readings I have got from the sources mentioned. The dots … imply that one verse, * that more than one or an uncertain number are lost; such interpolations as it has been deemed advisable to retain in the text, are printed in small capitals; the letters syllables and words which are omitted in the mss. but can be restored with more or less certainty, are given in Italics. In quoting Ennius the last edition, that of Vahlen, has been used; for the fragments of the Roman scenic writers, except Ennius, that of Ribbeck: in citing Cicero the smaller sections are referred to as far the most convenient for reference: for Terence the several recent editions; for Plautus Ritschl and Fleckeisen in the plays they have published; in the others the old variorum ed. has been employed: in Pliny the sections of recent editions are cited, as the older divisions are intolerably awkward. Notes 1 have been made as short as is consistent with perspicuity: unless the contrary is expressly stated or implied, the word or words which appear first in the note are those of our text; thus ‘genitabilis. genitalis etc.’ signifies that genitabilis is the right reading and is found in A and B and the other chief authorities, but genitalis is mentioned for the reasons given. Again ‘281 quam Lach. for quem. quod Junt.’ means that Lachmann first gave the correct reading quam instead of quem which is the reading of A and B and other mss. as well as editions before the Juntine of 1512 which prints quod, the reading generally followed by the old editors. Of course if any one before Lachmann had read quam, he, not Lachmann, would have been cited for it. ‘Ed.’ means the present editor. Let it always be remembered that the corrupt reading, cited in a note, is that which appears in A and B, unless the contrary is expressly stated.
The passages which were first added to the second edition have been enclosed within [ ] in cases where ambiguity or awkwardness might be occasioned, if no distinction were made between the old and the new matter; but not otherwise.