TRANSACTIONS OF THE VOL. VI. Part I. THE APPARATUS CRITICUS OF THE CULEX. By a, E. HOUSMAN. CAMBRIDGE UNIVERSITY PRESS WAREHOUSE, C. F. CLAY, Manager. LONDON : FETTER LANE. GLASGOW: 50, WELLINGTON STREET. 1908 (^All rights reserved.) Price One Shilling and Six Pence {net). THE APPARATUS CRITICUS OF THE CULEX. BY A. E. HOUSMAN. VOL. VI. PART I. I THE APPARATUS CRITICUS OF THE CULEX. In the Classical Review for 1902, p. 340 a, I selected, out of the MSS of the Gulex then known to scholars, the three following as the three main pillars on which the text should stand: Vaticanus 3252 saec. ix, surnamed Bembinus from its discoverer and owner Petrus Bembus, collated by Zange- meister for Ribbeck's edition of 1868 and there designated B; Leidensis Vossianus Lat. Oct. 81 saec. xv, first used by Schrader in his liber emendationum of 1776^ collated by Ribbeck and more accurately by Baehrens poet. Lat min. uol. II an. 1880, both of whom call it V; Corsinianus 43 F 5 saec. xv, brought to light by Prof Ellis in the Journal of Philology for 1887, pp. 153-6, cited as F by Ribbeck in his edition of 1895, and not yet collated. The main pillars are still three in number, but one dis- appears and another takes its place : the Vossianus must now make room for its grandfather, Vaticanus 2759 saec. xiii. The first mention of this MS, so far as I know, appears in uol. II fasc. 1 (1905) of the^oe^i Latini minori of Mr G. Curcio, who has used it for his recension of the Moretum and the Copa. Mr Ellis's edition of the appendix Vergiliana, where its lections are often cited in the notes to the Cidex, was published in July 1907 ; and I saw at once that Vat. 2759 must be an ancestor of Voss. 81. There were, to be sure, some things in the adnotatio critica not compatible with this relation, but I guessed that these would prove to be mistakes I — 2 UNlVthSIIY Of lUlNOlS LIBRARY AT URBAHA-CHAItfAiGN 4 CAMBRIDGE PHILOLOGICAL TRANSACTIONS. of Mr Ellis's. I wrote immediatel}^ to Rome for photographs of the MS, which I received when the Vatican awoke from its summer slumbers in October ; and it is as I have said : the Vossianus is a descendant of the Vaticanus. To Mr Ellis the Vaticanus is merely one of his sixteen MSS. In his apparatus criticus he leaves many of its chief readings unrecorded ; in the table prefixed he excludes it, as he also excludes the Corsinianus, from his list of the five best codices (four of which are negligible) ; and in a preface of ten pages he finds no occasion even to mention it. So far is he from perceiving its relationship to the Vossianus that in places where the two MSS give the same lection he often quotes them as separate authorities and often quotes Voss. without quoting Vat.; while at u. 214 he prints ex in his text because Voss. has et, though Vat. like all other MSS has e. Yet the relationship is so manifest that in proceeding to demonstrate it one is even embarrassed by the wealth of proof. Out of the eleven readings which I cited in 1902 to show the special merit of Voss., ten (60 spretis, 93 liget, 237 tue... ire, 249 uecordem, 304 talis... belli, 330 ipse, 340 ne quisquam, 343 argoa petens, 378 mali nec, 390 propter) are now found in Vat., as Mr Ellis notes in six cases and omits to note in four. Instead of the eleventh there is found something better still. At 332, where most MSS give ' metuenda Charybdis', Voss. has ranolea : ' hinc statim suspicabamur, uel potius intelligebamus ' said Schrader lib. emend, p. 26, 'poetam scripsisse Zanclaea Charybdis '. Vat. has zanclea^, with the c so written that it is hardly distinguishable from o. Two lines below, at 334, where most MSS give '(Tanta- leae) gener amplis (Atrides)', Voss. has g ener amen pr obis, the 1 Mr Ellis, following a practice common among editors, has used this opportunity to obliterate, so far as in him lies, the memory of Schrader's emendation: h\H note is, ' zancleaN : rawoiea Vossianus: metuenda'BM. G&ni.: et uerida Cors., h.e. {m)etuenda^ : nothing more. (The reader, let me warn him, must not infer that the rest of Mr Ellis's mss,— FEPTS b Vatic. 1586 Mediol., — agree with his text in reading Zanclaea: not a single one of them does so ; Mr Ellis's silence only means that he has omitted to ascertain what they read or forgotten to record it.) THE APPARATUS CRITIC US OF THE CULEX. 5 truth all but one letter. Vat. has the truth itself, gener amen prolis^. Three of the verses omitted by Voss., 27 and 319 and 403, are likewise omitted by Vat., and both Mss place 99 before 98. But the verses 217 and 412 — 414, also omitted by Voss., are preserved in Vat., and it does not, like Voss., repeat 392 sq. after 396. At 168, for the aurae or aure of most MSS, Voss. gives r arte. Vat. has ante : the corrector, wishing to restore aure, wrote r over the t, but did not trouble to write u over the n, because it was so much like u already. In Voss. this r has been substituted for the wrong letter. At 189, instead of spiritus, Voss. has spem, ' ex spc' as Baehrens explained. Vat. has that very abbreviation of spiritus. At 361 *hic est et Horatia uirtus ' the first hand of Vat. wrote oratia : o was then made into g by adding a tail to it, 1 Mr Leo states on p. 20 of his edition of the Culex, and repeats the statement on p. 98, that Eibbeck ' egregie ' and * praeclara emendatione ' restored generamen prolis (i.e. generam^lis) for the gener amplis of the mss. Kibbeck neither did nor pretended to do anything of the sort : what he did was simply to borrow generamen from the Vossianus and prolis from two other MSS, Voss. 96 and Koelerianus ; and, like an honest man, he said so. Mr Ellis, while consigning to oblivion the fact that Schrader conjectured Zanclaea at 332, preserves in memory the fable that Ribbeck conjectured generamen prolis at 334 : ' generamen prolis Ribbeck ' ; and Ribbeck, in virtue of this legendary exploit, is mentioned on p. xi of the preface as the only person whose services to the text of the Culex take rank with Mr Ellis's own. Fiction is again preferred to history at 294, where Baehrens conjectured ' dignus amor uenia, si Tartara nossent ' for the gratum of the MSS, and Mr Leo, instead of ascribing this emendation to its author, re- presented it as a suggestion of his own, ' debuit sane ueniam.' Mr Ellis found the truth declared in two places (Baehrens's own edition and Ribbeck's), the falsehood insinuated in one ; yet what he transmits to his readers is * " debuit sane ueniam " Leo '. There seems indeed to be an unwritten law that the corrections of Baehrens should be assigned to anyone rather than their rightful owner : the conjecture animi simplicitate tui for nimia s. tua in el. Maec. i 22 was proposed by Baehrens in 1879 and by Chatelain in 1880, and Mr Ellis allots it to Chatelain ; the conjecture somnia for omnia ib. 50 was proposed by Baehrens in 1872 and by Birt in 1876, and both Mr Ellis and Mr Riese say ^somnia Birt.' 6 CAMBRIDGE PHILOLOGICAL TRANSACTIONS. and c was inserted between a and t, producing gractia. Voss. has gracchia. Many other details point the samS way. At 3 and 8 Voss. has lusim and posteri for lusimus and posterius: in Vat. the termination -us is written compendiously. At 9 Voss. has dicta for nostra : in Vat. the capital N of Nrd is so formed that it might possibly be mistaken for D and cause the error Bca. At 20 Voss. has parens for pales: in Vat. the a and I are joined by a high arch which makes the word look something like pares. At 78 Voss. has mariantia for ma- nantia: in Vat. the first n much resembles ri. At 226 Voss. has uita for the iure of most MSS : Vat. has a word which probably is meant for iura but is indistinguishable from uira. At 302 Voss. has manihus for a nauihus : Vat. has anauihus written as one word, and the first half of the first a is in- conspicuous. At 326 Voss. has cecidit for concidit : Vat. has cocidit. At 340 Voss. has dims for diues : in Vat. the e is very like o. At 358 Voss. has parides for pariles: in Vat. the I is thickened at the base and might well be mistaken for d. Much else of this sort could be added ; but I will now set out on the other hand all the facts I have been able to find which might seem to indicate that Voss. is not derived from Vat. 93 iocundo Vat., iucundo Voss. ; 120 tue leto Vat., tuo letae Voss. ; 275 obtemptu Vat., ohtenta Voss. ; 307 sanxti w Vat., xanti Voss.; 315 curae {ce) Vat., contra Voss. In all these cases Voss. has the right reading ; but that reading occurs also in some or in all of the other MSS and may there- fore have been borrowed by Voss. from a foreign source ; or it may even have been recovered by conjecture. At 75 Vat., perhaps rightly, has suhtus, while Voss. like the other MSS has suhter ; but suht^, as it is written in Vat., might easily be mistaken for suht'', the commoner word. The only phe- nomenon which puzzles me is the following : both MSB omit u. 319, but Voss., according to Baehrens (for Ribbeck is silent on the point and Mr Ellis does not even note the omission), leaves a gap of one line, while in Vat. the sequence is un- broken. If a scribe observed the presence of this verse in THE APPARATUS CRITICUS OF THE CULEX. 7 another MS, you would expect him to insert it, not merely to mark its absence ; but perhaps he might be deterred by its evident corruption and its repetition of sigeaque praeter from 307. I said above that Vat. was the grandfather, not the father, of Voss. This appears from the following examples, which require the assumption of at least one intervening copy : 42 Ignibus (quite plainly) Vat., Lenihus Voss. (through Icnihus or Inibus)', 144 corimbos Vat., conbos Voss. (through coribos); . 7 ^ 162 inctos Ysit., inceptos Voss. (through incertos) ; 389 tn... piignans Vat., temen. . .pugnas Voss. (through tamen. . .pugnds). Few direct additions are made by Vat. to our knowledge of the Gulex, for most of its best readings have been preserved in Voss. Apart from 332 zanclea and 334 generamen prolis already mentioned, it seems to contribute nothing new and true except transcendat in 84, where the other MSS have transcendit, and sociate (i.e. ' sociatae gloria sortis ') in 301, where Voss. has societate and most MSS sociat de ; and here the truth had long ago been divined by the early editors. It is of much more service in enabling us to make subtrac- tions from the testimony of Voss. and to sift apart the true and the false, which has hitherto been difficult or impossible. A MS of the 15th century, even when copied straight from an ancient original, will often contain the embellishments of the educated scribe ; and many such can now be detected in Voss. First, its orthography. It presents quom at 174, 187, 204, 223, 306, 308, 379; quoi at 171, 188, 221, 408; quoius at 206 ; aduorsum and aduorso at 85 and 243 ; subcincta at 331, conpellit at 29, inminet at 57, and similar forms in many other words compounded with prepositions : most of these are received into the text by Baehrens and some of them by Ribbeck. But wherever Voss. has quom, Vat. has cu or c; i wherever Voss. has quoi, Vat. has cui or c ; Vat. has cuius at 206, the usual contractions of aduersum and aduerso at 85 and 243, succincta at 331, gpellit at 29, iminet at 57, and so throughout. Baehrens again, whose addiction to Voss. was excessive and pernicious, embraced, or used as substance for 8 CAMBRIDGE PHILOLOGICAL TRANSACTIONS. conjecture, several other readings of this MS which we now ascertain to be late corruptions. Because Voss. at 9 had dicta he conjectured docta\ but Vat. has nostra like the other MSS. Because Voss. at 214 had cetris he conjectured certis; but Vat. like the rest has tetris. In the same verse he suggested ec for the et of Voss. (as Mr Ellis even now suggests ex) ; but Vat. like the rest has e. At 226, where Voss. has uita for the iure of most MSS, he conjectured uicta ; but Vat. has inra. At 248 he adopted the turbas of Voss., and at 361 its Gracchia ; but Vat. has turmas in the one place and originally had oratia in the other. My own con- jecture uires for the unusual sibi uires of the MSS at 388 can no longer derive any support from the uires sibi of Voss., for Vat. has sibi uires like the rest. In Mr Ellis's report of the readings of Vat., though his omissions are numerous, capricious, and deceptive, distinct misstatements are comparatively few, and I will here set right the most important. At 10 Vat. has not spoliantur, as Mr Ellis says, but poliantur ; at 55 not qua but que ; at 168 not ' uisus an nisus ' but uisus quite plainly ; at 264 not cd (causa) but the usual abbreviation of cura (as at 394) ; at 370 not scipiatos but scipiadas ; at 409 not armis but artus. Six or seven of less moment will be tacitly corrected in the following pages. All things considered, I thought it advisable to obtain photographs of the Corsinianus also. Mr Ellis printed what he termed his collation of this MS in the Classical Review for 1892, pp. 203-5. With what text he had collated it he did not say, and it was therefore impossible to use his collation as collations are used : in u. 14 for instance, where he noted nothing, one could not tell whether Cors. had alma or acta or Arna, since editions are divided among the three. Still, Ribbeck and I took Mr Ellis at his word; and in u. 2, where all accessible editions, all texts of the last four hundred years or so, have orsum, we ventured to infer from the silence of the alleged collation that Cors. had orsum too. We were deceived ; it has ursum. The fact is that Mr Ellis does not collate MSS : he transcribes from them, without reference to TEE APPARATUS CRITICUS OF THE CULEX. 9 any particular text, such things as attract his attention. That Cors. has numina for tartara at 294, and that it omits the two whole verses 7 and 125, are things which did not attract his attention ; but fortunately no new lection of any value is missing from his excerpts. Of actual misstatements I have noticed only eight, one of which, 243 saxum (really saoco), he now corrects in his edition, and only three of the remainder have any importance : 88 herbas (really herhis)^ 283 hiiuges (really hiiugos), 304 tali (really talis). My photographs of Vat. 2759 and Cors. 43 F 5 shall be put at the disposal of anyone who wishes to collate them ; but what I now have in view is a task of more public utility. Both MSS, especially Cors., contain a great number of mere blunders and falsifications which are of no use for the con- stitution of the text; and instead of adding to the vast stores of unprofitable matter collected at the foot of their pages by Ribbeck and Ellis I design to furnish something new : an adequate apparatus criticus to the Gidex. Of the three cardinal MSS two were unknown to Ribbeck in 1868, to Baehrens, and to Mr Leo; one was unknown to Ribbeck in 1895 ; and Mr Ellis's adnotatio ci'itica of 1907, indispensable for the moment because nowhere else can lections from all three MSS be found, is so often misleading or unintelligible that it is both dangerous and difficult to use. I propose then to take as a standard the text of Mr Leo (1891), which is best suited for the purpose as adhering more steadily to a single MS and admitting fewer conjectures than any other, and to note firstly, like Mr Leo himself, all variants of B which are not quite insignificant, and secondly such variants of Vat. and Cors. as either preserve the truth or point to it, or are specially characteristic of the MSS, or claim notice by any other quality, depravity included. Where Vat. and Cors. agree in a false reading, even if it deflects but slightly from B's, I shall generally record it, and where all three MSS differ I shall generally note the readings of all three ; but where either Vat. or Cors. coincides with B in a true reading I shall ignore the false reading of the solitary recusant, unless its divergency is somewhat conspicuous. 1—5 lO CAMBRIDGE PHILOLOGICAL TRANSACTIONS. Where the true reading is a matter of uncertainty I shall be more minute: for example I mention tliat Cors. has memora- hili for memorabilis at 37 but not that it has idtem...actam for uitae. . .acta at 276. Such trifles as the difference between ae and e, i and y,ph and /, ti and ci, the omission and insertion of h, or the usual misspelling of proper names, I shall seldom notice, nor shall I as a rule record the corrections of a second hand when the first hand is legible. That the reader may see for himself my practice in selection I subjoin a full collation of Vat. and Cors. for the first 41 lines, omitting nothing but details of orthography \ The Bembinus was collated by Zangemeister for Ribbeck's first edition and has again been examined by Mr Ellis : the two reports sometimes differ, and I will note the differences. When one scholar, having before him a collation made by another scholar, consults the MS anew and contradicts the earlier statement, it is generally assumed that the second report is correct and the first erroneous ; but here this assumption must not be made. I have taken the precaution to test, in the Giris, the collation of Arund. 133 which Mr Sievers made for Baehrens's edition of 1880, and the 1 1 Liisimus] usimus Vat. Cors. 2 orsuni] usum corr. in orsum Vat., ursum Cors. 3 culices] culicis Vat. Cors. docta] dicta Vat. 6 erit] eris Cors. 7 om. Cors. 9 cum] com corr. in cu Vat. 10 digna tuo'] dignato Vat. Cors. 11 proles] prolis Cors. 12 auctor] autor Cors., actor Vat. 13 reciriente] retinente Vat. 14 Arna Chimaereo] Alma (supra scripto terra) chimereo Vat., Almachi mereo Cors. 15 Asteriae'] asterie Cors., astrigerum Vat. 17 sonans liquido] sonas liquida Cors. 18 laticis] nemoris Vat. 19 Naides] Naiades Vat. Cors. 21 secura] fetura Vat. Cors. tenentis] tenentes Vat. 23 uagusl uagos Cors. antra] astra Vat. Cors. 27 om. Vat. ponitque] ponitque canit non pagina helium Cors. 28 Phlegra] Flegraque Cors. quae] quo Vat. 29 nec] Hec Cors. enses] ensem Vat. 30 urit Erichthonias] Vritur hie thomas Cors. arces] axes corr. in arces Vat. 31 perfossus] con- fossus Cors. 32 iacta] Leta Vat. Cors. quaerent iam] querentia Cors.- vxylumine] uolumina Vat. Cors. 34 cum timuit] coiitinuit corr. in cum timuit Vat., cum timuit {timuit erasa priore scriptura ut uidetur) Cors. e uenientis] u£nienti8 Vat. Cors. 35 currere caimina] carmina currere Vat. 37 hoc] Hec Vat. memorabilis] memorabili Cors. 38 mansura] uel e mensura Core. 40 felicis]felicis Ya.t. ,felices Cors. jnemoretur] remoretur Cors. 41 lucens supra lin. add. Vat. THE APPARATUS CRIflCUS OF THE CULEX. it collation published by Mr Ellis in his edition of 1907, by comparing both with the MS itself; and I find that although the later is, as it ought to be, more exact than the earlier, there are nevertheless ten places where Mr Ellis's contradic- tiQns of Mr Sievers are contradictions of fact. 93 trihuistis man. 1, tribuisti man. 2 Sievers rightly: trihuisti man. 1, ^W6m5^is man. 2 Ellis. cortimus ^: cortinius 240 uno S : una E. 275 uersa repetes corr. in uersare potes S : uersa re petis (suprascr. potes) E. 295 ei corr. in est S : est E. 301 minois S : minoris E. 326 flumina S : flumina E (what the MS has is flumid, that is flumla with the stroke lying too far to the right, as often). 356 inempte corr. in inepto S : inempte corr. in inempto E. 533 namque corr. in iamque S : iamque (suprascr. namque) E. 534 sidere S : munere E\ When therefore Mr Ellis con- tradicts Zangemeister's report of the Bembinus, although we may assume that he is oftener right than wrong, we cannot in any given instance assume that he is right. Compared with B and Vat. and Cors. all other MSS of the Gulex are insignificant, and to cite their readings throughout would be a sinful waste of time and paper. Still, there are in the poem some 30 passages where a true reading not found in any of the chief three is found in one or more of the others. Probably some of these readings are conjectural, and possibly all of them, but not one of them necessarily; for it is certain that the majority of the MSS are in no way derived from B or Vat. or Cors. Many of them are much older than Cors. and even than Vat. ; and though none of them is so old as B, they yet prove their independence of B by preserving u. 133, which B omits. One would therefore expect them to 1 Most of these errors now reappear in the edition (1908) of Mr Curcio, who says on p. x 'la collazione riportata dal Eibbeck e dal Bahreus differisce in piu di un luogo da quella fatta dall' Ellis ; io mi sono attenuto a quella di quest' ultimo.' I may here add that Messrs Ellis and Curcio both employ a new and not quite worthless ms of the Giris, Vrbinas 353, and that their reports of it disagree in 25 places, not reckoning matters of orthography. This codex therefore, which was barely worth collating once, now wants collating a third time. 12 CAMBRIDGE PHILOLOGICAL TRANSACTIONS. retain here and there some fragment of the truth which the three chief authorities have lost : the wonder is that they retain so little. The most conspicuous among them is one of the latest, Helmstadiensis 332 saec. XV, which alone and un- supported contributes some half-dozen of these true readings, and a second half-dozen in company with one or more of the rest. Mr Ellis has set himself, with unusual care and ingenuity, to withhold from his readers a knowledge of these facts, and indeed of all facts which might tend to raise the credit of Helmst. 332. He receives into his text two readings (334 Atrides, 402 surgens decus) for which Helmst. is the sole MS authority, and he mentions four more (42 peneti-abat, 96 poetae, 221 horrent, 356 peremptae) in his notes ; but in every case he either conceals their true origin or assigns them a false one. In the Ciris he pursues the same course, which is there rendered peculiarly difficult by the fact that Helmst. (except where the Brussels fragment survives) is the best MS of the Ciris. Mr Ellis however, by the use of such bold expedients as declaring anubus at 375 to be a conjecture of Ribbeck's, contrives to mention it no more than nine times in his whole apparatus criticus, employing in its stead a grossly interpolated offshoot of the same stock, Vrbinas 353. The remaining MSS of the Culeoj, though some of them belong to the 10th and 11th centuries, are of still less use. I will here give the names and dates of all that I shall even once have occasion to cite : Par. 7927 saec. x, Par. 8069 saec. X uel XI, Par. 8093 saec. x, Cantabrigiensis Kk. v. 34 saec. X, Petauianus (lost), Mellicensis saec. xi, frag. Stabulense (Par. 17177) saec. xi, frag. Par. 8207 saec. xiii, Vatic. 1586 saec. XIV uel XV, Vatic. 3255 (date not stated), Voss. 81 saec. XV, Mediol. (Ambr. O 74 sup.) saec. XV, Koelerianus bibl. Goetting. saec. xv, excerpt. Paris. (Par. 7647 saec. xii uel XIII, 17903 saec. xiii), excerpt. Escor. saec. xiii uel xiv. The apparatus criticus which follows takes no count of the emendations of editors old or new ; but it records every MS reading which a student of the Cideoc needs to know. Nay it records much more : succinct though it appears in C(Hnparison with the notes of Ribbeck or Baehrens or Ellis, THE APPARATUS CRITICUS OF THE CULEX. 13 it errs deliberately on the side of fulness, and contains a great deal which, when it has once been recorded, need never be mentioned again. A future editor, preparing to equip the Gulex with what Cobet calls an 'apparatus uere criticus', should lighten it by throwing overboard many variants of Vat. and Cors. which are certainly false, several variants of B which are altogether trivial, and lections such as that of Cant, at 15 and frag. Stab, at 357, which, though instructive as tracing the progress of corruption, are useless for the constitution of the text. One word is necessary on the notation employed. The Bembinus we call B; but what am I to call the Vaticanus and the Corsinianus ? The reader wishes, and so do I, that they also should each be designated by a single letter ; but that is forbidden by the character and conduct of our fellow men. I cannot, as Ribbeck did, call the Corsinianus V, because Mr Ellis has transferred this letter to the cod. Harl. 2534' which Ribbeck called O, so that henceforth no editor of the Gulex can apply it to any MS whatsoever. I cannot, as Mr Ellis does, call the Vaticanus V, because this letter has been used for the last forty years to signify the Vossianus, and I will have no hand in the work of promoting confusion. I cannot take two unappropriated letters from the alphabet and assign them to these MSS, because, no matter what letters I choose, the amour propre of my next successor will forbid him to acquiesce in my choice and compel him to choose two other letters : this is a field in which everyone can display originality, and in which therefore originality is unfailingly displayed. There is nothing for it but to use signs which carry their meaning on their face, Vat. and Cors. : seven letters where two would serve, * arbitrio si natura finxisset meo I genus mortale.' 2 orsum corr. ex ursuni B, ex usum Vat., ursum Cors. 3 culices] culicis B Vat. Cors. docta B Cors., dicta Vat. 6 erit B Vat., ms Cors. 7 om. Cors. 10 digna tuo] dignato B Vat. Cors. poliantar Vat. Cors., spoliantur B. 14 Arna\ Alma B Vat. Cors. 15 asterie Cors., astrieri Cant., 14 CAMBRIDGE PBILOLOOICAL TRANSACTIONS. astrigeri B, astrigerum Vat. 18 laticis B Cors., nemoris Vat. 21 secura B, fetura Vat. Cors. tenentis Cors., teiientes B Vat. 22 uirentes Vat. Cors., uirentis B. 23 a?i^m] as^m B Vat. Cors. 27 om. Vat. canit non pagina helium add. B Cors. 28 quae B Cors., quo Vat. 30 FriY erictonias Vat., Fra^ erechtonias (teste Zang., erechth. teste Ell.) B, Vritur hie tliomas Cors. 31 perfossus B Vat., co?i/bs5ws Cors. 32 /ac^a Par. 7927 et 8093 (teste Silligio, negante Ell.), Mediol. (teste Ell.), Laeta B, Leta Vat. Cors. uolumine] uolumina B Vat. Cors. 35 currere carmina B Cors., carmina currere Vat. 37 Hoc B Cors., Hec Vat. memorahilis B Vat., memorabili Cors. 40 me- moretur B Vat., remoretur Cors. sicut coni. Baehrens. 42 Igneus B Cors., Ignitus Vat. penetrabat Helmst., penetrarat Vat. Cors., penetrat B. 44 fugarat B Vat., fugahat Cors. 45 patula] pabula B Vat. Cors. 47 co/^es B Cors., ^ore5 Vat. 48 abdunt B Vat., addunt Cors. 49 iamque omni B Vat., mm iamque Cors. e B Cors., Vat. 51 desertas herebant...rupes B Cors., desertis erra- bant...rupis Vat. 55 Fe^ Par. 8093 (teste Sill., negante Ell.), frag. Par. 8207, Mellic, fTaec B Vat. Cors. quae B Vat., ^i^a Cors. 57 in riui praestantis B Vat., intuici (ut uidetur, pictis potius quam scriptis ultimis qnattuor litteris) perstantis Cors. 60 pretiis B, praetus Cors., spretis Vat. 64 domus Helmst., Par. 8069 (testibus Sill, et Ell., negante Ribbeckio), frag. Par. 8207, excerpt. Par., domos B Vat. Cors. anget] angit B Vat. Cors. 67 boeti Vat. Cors., coeti B. 68 at B (teste Zang.) et ex corr. Vat., a B (teste Ell.) Vat. Cors. 71 notat B Cors., nouo Vat. dulci Helmst., excerpt. Par., dulcis B Cors., dubiis Vat. 72 recinente Cc»rs., recanente B, retinenente Vat. 73 inuidia B Cors., inuidie Vat. Helmst., Petau., frag. Par. 8207, excerpt. Par., om. B Vat. Cors. 75 subter B Cors., subtus Vat. uelat B, i/o/a^ ut uidetur Vat., uelut Cors. 77 intus B Vat. Cors., imis Helmst. 81 agnouit B Cors., agnoscit Vat. 83 excerpt. Par. et Esc, frag. Par. 8207 (teste Sill., tacite negante Ell.), Non B Vat. Cors. 84 uel] nec B Vat. Cors. transcendat Vat. sicut coni. Bembus, trans- THE APPARATUS CRITIC US OF THE CULEX. 15 cendit B Cors. 87 panchaia B, panchasia Vat., pancheia Cors. 88 herbe Vat., herhis B Cors. uariantibus B Cors. et praecedente unius litterae (fortasse c) rasura Vat. adsunt B Vat., addunt Cors. 90 hue B, haio Vat. Cors. 91 Derigit Cors., Dirigit B Vat. 92 requie Vat., requiem B Cors. itic^w Vat. Cors., uictus B. 93 Zi^e^ Vat., ^icei B Cors. 95 c^^^^it Vat. Cors., culte B. 96 pastor sihi B Cors., pastori Vat. poe^e Helmst., jooe^a B Vat. Cors. 97 traducit Vat. Cors., traducis B. 98, 99 inuerso ordine Vat. 100 solitum Par. 7927 (teste Sill., negante Ell.), Par. 8093 (teste Sill., tacente Ell.), solidwn B Vat. Cors. 103 occeanum Vat., oceanus B, occeanus Cors. 105 repete- bant B Cors. et corr. in repedabant Vat. 112 infandas Vat., infanda B Cors. e] et B Vat. Cors. 114 nati de] nati se Vat., nati se-^ B (teste Zang.), natise B (teste Ell.), ae Cors. datura'] futuram B, futurum Cors. et (supra /it scripto ite^ a. d.) Vat. 115 /ier6a Cors. et corr. ex herbas Vat., herbal B. 116 ^ite choros Helmst., Cant., alii ex Silligii Ribbeckiique silentio, que chorus B Cors., chorosque Vat. 117 horpheus Vat. et man. rec. B, horridus Cors. et in marg. B, erasus B. 118 ripis B Cors., riuis Vat. 119 pernix remorantem Vat., permigre morantem B, per nigre morantem Cors. 120 tuo lete Cors., tue leto Vat., fi^ae laetae B. 124 platani Vat., platane B, platanus Cors. 125 om. Cors. 128 Ambustus Helmst., Mediol. (teste Ell.), Ambustos B Vat., Ambusto Cors. matauerat Vat. Cors., motauerat B. 129 inter lineas additus in Vat. i??2- j^feae] amplexae B Vat. Cors. 132 perfide Vat., perfida B Cors. 133 om. B. Perfide Vat. cum plerisque, Perfidie Cors. Cors. cum plerisque, i Vat. (iice?ic?e] defende Vat. Cors. 137 etii^a B, addita Vat. Cors. sicut coni. Heinsius. 138 proceras B Vat., proceratus Cors. 139 motibus] montibus B Vat. Cors. 140 r?ec] B Vat. Cors. 141 ligantes Vat. Cors., ligantis B. 148 5it6em^] superat B Vat. Cors. 149 ac^a Vat., orta B Cors. Zi'gwo- rum Cant., liquorem B Vat. Cors. 150 qiiaqua] quamquam B Vat. Cors. 151 querulae B Vat. Cors., queridas excerpt. Par. et Esc. 152 aeris B, etheris Vat., alit.. echo om. l6 CAMBRIDGE PHILOLOGICAL TRANSACTIONS. Cors. 155 suhter] supra excerpt. Par. et Esc, que super B Vat., que supra Cors. quos excerpt. Par. et Esc, quae B Vat. Cors. leniter Cors., leuiter B Vat. 161 quietem B Cors., soporem Vat. 163 isde7n] idem Vatic 3255 (teste Ell.), Koeler., ide Cors., idae B, yde Vat. 165 subsideret] siibsideris Vat. et (teste Zang.) B, sub sideris Cors. et (teste Ell.) B. \.6Q grauis... lingua om. Cors. 167 motibus Cors., montibus B Vat. 168 irae] aurae B, aure Cors., ante supra t scrip to r Vat. 170 pectus B Cors., corpus Vat. ecce] et se B Vat. Cors. 174 sese B Cors., late Vat. 176 toruos^ toruo B Vat., torua Cors., tornus Helmst. 178 comparat B Vat., computat Cors. 179 intonat B Vat., insonat Cors. sicut cooi. Schrader. 180 euersis torquetur... orbis B Vat., euexis torquentur. . .orbes Cors. 185 gemmis B Vat., gemmas Cors. 186 senioris an semoris incertum Vat., 56 moris B Cors. naturae B Vat., natura Cors. 187 icto corr. ex I acta Vat., I acta B Cors. 189 sensus B Cors. et corr. in sensim Vat. 191 exanimus B, exanimis uel exammis Vat., exanimi Cors. refugit B Cors., resurgit Vat. 192 truncum detraxit ab orno Vatic 1586 (teste Ell.), truncum detraxit ab ore B Vat., detraxit ab arbore truncum Cors. 193 c?^i] B Vat. Cors. 50cmriY] sociaret B Vat. Cors. numenue B Cors., numen ne Vat. 194 ualuit] uoluit B Vat. Cors. ^a^is] ^a^es B Cors., ifaZ^i (radendo factum ex ^a^es ut uidetur) Vat. 197 cristae Par. 8093, cristam B Vat. Cors. 198 mVte] tardus B Vat. Cors. 199 ?iec ^^rius] Nescius B Vat. Cors. obcaecauerat B Cors., excecauerat Vat. 200 implicuit B Cors. et pro uar. lect. Vat., impleuit Vat. 202 ereboeis B Vat., erebois Cors. 203 aurata B Cors., aurato Vat. 207 requierunt B Vat., requiein dare Cors. 210 Qitis inquit] Quid inquit Cors., Inquit quis Helmst., Inquit quid B, Pnquid quid Vat. acZ (/we Vat., acZ ^item Cors., 6i^(/i an ancestor of Voss. 81 . 3 — 8 Vossianus 81 a descendant of Vat. 2759 . 3 — 8 ,, ,, interpolated . . . . 7, 8 Zanclaea Charybdis 4 I I MEETINGS FOR 1908-9. Lent Term, 1908. 1. Thursday, February 6, at 4.15. 2. Thursday, February 20, at 4.15. 3. Thursday, March 5, at 4.15. Easter Term, 1908. 1. Thursday, May 7, at 4.45. Michaelmas Term, 1908. 1. Thursday, October 22, at 4.15. 2. Thursday, November 5, at 4.15. 3. Thursday, November 19, at 4.15. Lent Term, 1909. 1. (Annual Meeting) Thursday, January 21, at 4.15. In the Lent Term the Society will meet in Mr Giles's rooms, imanuel; in the Easter Term in Mr Harrison's rooms. Trinity; in Michaelmas Term in Dr Jackson's rooms. Trinity ; for the Annual ting in Mr Giles's rooms, Emmanuel. '^embers wishing to communicate papers or points for discussion jijtiested to inform either of the Secretaries, ae Society's Library is in the Museum of Classical Archaeology, a St Mary's Lane. COUNCiL FOR 1908. p. GILES. M.A., EVMANCEL. J. E. NIXON, M.A., King's. PROF. H. JACKSON, Litt.D., Teinity PROF. A. A. BE VAN, M.A., Tsinitt. ©rOmarg i«embm of Council. PROF. F. C. BURKITT, M.A., Trinity. MISS J. E. HARRISON, Newnham. REV. PROF. W. W. SKEAT, Litt.D., Chkist's. A. W. VERRALL, Litt.D., Trinity. PROF. W. RIDGEWAY, M.A., Gonville and Caius. PROF. E. J. RAPSON, M.A., Sx John's. REV. R. G. RURY, M. A., Trinity. J. P. POSTGATE, Litt.D., Trinity W. G. HEADLAM, Litt.D., King's. R. D, HICKS, M.A., Trinity. E. HARRISON, M.A., Trinity. R. K. GAYE, M.A., Trinity. E. C. QUIGGIN, M.A., Gonvjlle and Caius. C. F. ANGUS, M.A., Trinity Hall. J^^. SLEEMAN, M.A., Sidney Scsse-v. S. G. CAMPBELL, M.A:, Christ's. l^anfetrg : MESSRS Barclay & CO. (Mortlock's bank), Cambridge. The object of the Society is the promotion and publication of researches into the Indo-European languages and literatures, and the promotion of Philological studies in general. All Students of Philology are eligible as members. The subscription is £1. Is. annually, or £10. 10s. life composition. Subscribers receive all the publications of the Society, including the Proceedings and Transactions : and also the Journal of Philology, which appears twice a year. Members of the Society who wish to complete their sets of the Journal of Philology can now obtai" back volumes at half price (excepting Vols. II, V, VIII to XI and XIV which are nearly out of print) by applying directly to the publishers, Messrs Macmillan & Co. Ltd., St Martin Street, London. Those who wish to join are asked to communicate with one of the Secretaries, who will give further information. publications of tfte CambnUge philological Society since 1880. s. Transactions, Vol. I (1872—1880) ' 15 Transactions, Vol. II (1881, 1882), with Index to Volumes I and II (and Proceedings for 1882) ... 12 Transactions, Vol. Ill (1886—1893) Part I 3s. ; Part II 2s.; Part III 25. M.; Part IV Is. 6c/.; Part V with Index ... ... ... ... ... ... ... 1 Transactions, Vol. IV (1894—1899) Part I 3s. U.; Part II 6 Transactions, Vol. V Parts I and II, each ... ... ... 3 Transactions, Vol. V Part III {net) 2s.; Part IV {net) ... 1 Proceedings, 1882 to 1888 (except Parts I, Hand IV— VI, which are out of print), each {net) Proceedings, 1889 to 1902 (with Laws and List of Members), each {net) Spelling Reform and English Literature by H. Sweet (out of print). PkONUNCiATioN OF Latin IN THE AuGUSTAN Period (Third Thousand) 0 3 ALL PUBLiaHED AT THE Cambridge University Press Warehouse, Fetter Lane, London. m I UNIVERSITY OF ILLINOIS-URBANA flTivQn YH C001 THE APPARATUS CRITICUS OF THE CULEX. CAM