miswriting secured a place in the text at an early time and has been perpetuated through a series of transcriptions; the other, a much more likely explanation but often ignored by editors, that the mediaeval scribe accidentally fell into the same mistake of copying as the ancient scribe. The text of Horace has been examined in this light by Vollmer, whose edition will soon be published. He makes the bold assertion that our MSS. all came from one ancient copy, a volume belonging to Mavortius (consul A.D. 527), and containing Porphyrion's commentary. From two separate transcriptions of this ancient volume arise the two families (three, according to Keller) of our MSS. The text of Horace rests, therefore, on a slender and insecure basis, and offers suitable material for conjectural emendation. In C. I. xii. 31 our MSS. exhibit the unmetrical version found in their common original: "Et minax, quia sic voluere, ponto Unda recumbit." From the remarks of various ancient grammarians it is clear that the ancient and true reading in C. I. viii. 2 was: "Lydia, dic, per omnis Hoc deos vere." Our MSS. offer (1) te deos oro, (2) hoc deos oro, variants which Vollmer refers to a suprascript explanation— He complains with justice that the evidence of ancient grammarians has not been properly appreciated by editors of Horace, and that in the huge mass of details in Keller and Holder's critical apparatus the wood sometimes cannot be seen for the trees. They have, he says, fallen more than once into the error of assigning to lemmas (headlines) the same traditional value as to the accompanying scholia 1 F. Vollmer, "Die Ueberlieferungsgeschichte des Horaz" (in Philologus, Suppl. X. ii.), 1905. 2 C. IV. xiv. 28, diluviem meditatur agris, seems to me an example of this. The remarks of grammarians show that the ancient texts had meditatur, and that this unusual meaning of the word was commonly glossed (i.e. explained) as minatur. Our MSS. exhibit minitatur, which looks like an attempt to suit to metre the suprascript gloss minatur which had found its way into the text. (notes), and of forgetting that these lemmas were usually either left blank by a scribe for the rubricator to supply, or indicated merely by the initial letters of the words, so that the rubricator (or the next transcriber) had to guess at their true form. Thus they quote Caesius Bassus as evidence for the reading te deos oro, because these words appear, through the pardonable error of a rubricator, in the headline of the paragraph; whereas, from Bassus's remarks, "Horace makes the second foot a spondee, hoc deos vere; but, if he had followed Alcaeus's type, he would have used some metrical arrangement like hoc dea vere," it is clear that Bassus's text had hoc deos vere. As example of an ancient variant he cites Seneca's quotation of Sat. I. ii. 27, with Buccillus for Rufillus. Horace himself confirms Rufillus (Sat. I. iv. 92, "Ego, si risi, quod ineptus Pastillos Rufillus olet, Gorgonius hircum ""). The origin of Seneca's variant, whether a "first thought” of Horace, or the real name as opposed to the pseudonym, or a joke at the expense of some contemporary, is not clear. Vollmer's bold theory has roused a storm of opposition. Its weakest point is that he is forced to assign to some mediaeval scholar the interpolation' at the beginning of Sat. I. x. : "Lucili, quam sis mendosus, teste Catone, Defensore tuo, pervincam, qui male factos Emendare parat versus," etc. And yet the theory might be saved by the supposition that these lines were taken from some MS. of, let us say, Suetonius which was extant in Charlemagne's time. Vollmer makes the famous Blandinius codex a mere member of the second family of MSS., and suggests that its "bright jewel," the reading campum lusumque trigonem (Sat. I. vi. 126), may have been shared by the better MSS. of the first family also, since only inferior representatives of this family (with rabiosi tempora signi, possibly a marginal variant in the archetype) are available for this portion of the text. Our traditional text of Plautus can be traced back to two It relates to Valerius Cato, the teacher of Catullus and of the "Celtic" school of the "novi poetae " in Cicero's time. Cato published an edition of Lucilius (see Marx's Prolegomena, p. li). ancient editions. Leo declared that the common source of these was an edition made in the second century A.D. from the material collected by Valerius Probus, the famous" sospitator” of the Republican authors; that the numerous errors which they share-e.g. Acropolistidem for Telestidem, Epid. 568; in ius vos volo (the regular verb in this phrase is voco), Poen. 1225-indicate that Probus could not find more than a few copies of the different plays, and these with a very corrupt text; further, that their divergences are often mere capricious alterations by third or fourth-century editors. Thus at Stich. 223 he thinks that Probus's copy (or copies) had some corruption of hercle aestimavi, which he believed to be what Plautus wrote; one editor conjectured Hercules te amabit, the other hercule iste amavi, or something which has assumed this form in the extant palimpsest. In a recent pamphlet1 I have taken a more optimistic view, that the variety of the text really goes back to the Republican period, the one text being the "Genuine" (or " Author's ") text, the other the "Revival" (or "Theatrical ") text. The latter embodied the alterations introduced by stage-managers at the revival of Plautus's plays (from Terence's time onwards). What appear to be errors common to the two editions are really "inevitable" errors of transcription, made by transcribers at different times and independently. In the Epidicus the "fidicina" Acropolistis has passed for Periphanes' daughter under the name Telestis. Periphanes, of course, speaks of her as Telestis, while in the Sceneheading and in the "nomina personarum" throughout the dialogue she naturally figures as Acropolistis. The passage in question stands as follow: PERIPHANES. Iube Telestidem huc prodire filiam ante aedis meam, Ut suam videat matrem. PHILIPPA. Remigrat animus nunc demum mihi. (Act. IV. Sc. ii.) ACROPOLISTIS PERIPHANES PHILIPPA FIDICINA ACROP. Quid est, pater, quod me excivisti ante aedis? 1 The Ancient Editions of Plautus. Oxford (Parker), 1904. (St. Andrews University Publications, No. III.) Would not every transcriber, who was not familiar with the plot, be likely to regard Telestidem here as an error for Acropolistidem, and to correct his original accordingly? The presence of this error in all our MSS. does not imply that it stood in the Probus-edition. On the contrary, we can believe neither an editor nor an intelligent reader to have tolerated it. Can we imagine an edition of Shakespeare to remain century after century unaltered, if it contained in Twelfth Night a line in which the Duke was made to address Viola as "Viola" instead of "Cesario"? Where no "inevitable" error is there, the consensus of our two families of MSS. can be followed with great confidence. My edition of the Comedies observes these conservative principles very strictly. Thus in Poen. 1225 I refuse to alter in ius vos volo. Where hiatus appears in a line in both editions, I regard the hiatus as Plautine. Perhaps I have clung too closely, like a limpet, to the rock of tradition; but the course of Plautine textual criticism since Ritschl's time has shown that a reading shared by our two families of MSS. generally turns out in the end to be correct. How many ancient editions do our MSS. of Juvenal represent? Probably only one; for all the variants can perhaps be explained on the hypothesis that a single ancient copy survived to modern times-a copy whose text has been most faithfully transcribed in the codex Pithoeanus. It seems to have contained a number of marginal (or interlinear) variants, as well as a body of scholia which probably supplied transcribers with additional variants. This text was "doctored" by some mediaeval scholar-e.g. viii. 148, sufflamine multo (i.e. MVLIO) consul was re-arranged as multo sufflamine consul, to save the metre; ix. 106, taceant (i.e. FAC EANT) omnes was thoughtlessly altered to clament omnes, to save the sense. From this "doctored" text come the "codices deteriores." Another hypothesis makes the good text (of the Pithoeanus, etc.) and the bad recognovit W. M. Lindsay. Oxford 1 T. Macci Plauti Comoediae: . (Clarendon Press), 1904-5. text (of the "deteriores") come from two ancient editions, one good and one bad. If they do, Housman, in the preface to his edition,1 is partly justified in his somewhat vigorous indictment of some previous editors for unduly ignoring the readings of the inferior text; although he certainly falls into the opposite error of unduly favouring some poor readings. At xi. 148 the Pithoeanus-reading— Non a mangone petitus Quisquam erit in magno cum posces, posce Latine, is confirmed by an ancient Greek-Latin conversation-manual, which shows us that in magno poscere, in magno miscere, etc., were current phrases of the wine-table. The "doctored" text had changed in to et; a fifteenth-century copy has omitted the second syllable of quisquam (a common error in MSS.), and from this "corruption of corruption," quis erit et magno, Housman elicits (and prints) qui steterit magno, "who has cost a good round sum." But he is right in repeating Jahn's warning, that the text rests on insecure basis, and may contain defects where none are suspected. Winstedt's fragment is apparently the content (29 lines) of a page of the ancient copy, which had been omitted in some early mediaeval archetype, while the omission was cleverly concealed in a subsequent transcript. To quote Housman: "Twenty-nine verses were omitted after vi. 365, and left this headless sequel: consilia et veteres quaecumque monetis, amici, hac mercede silent? crimen commune tacetur. In our MSS. these lines have been battered into the following shape: audio quid veteres olim moneatis amici, 'pone seram, cohibe.' sed quis custodiet ipsos 1 D. Iunii Iuvenalis Saturae : (Grant Richards), 1905. edidit A. E. Housman. London |