MA S TER NEGA TIVE NO. 92-80497-20 MICROFILMED 1992 COLUMBIA UNIVERSITY LIBRARIES/NEW YORK as part of the "Foundations of Western Civilization Preservation Project" Funded by the NATIONAL ENDOWMENT FOR THE HUMANITIES Reproductions may not be made without permission from Columbia University Library COP^IRIGHT STATEMENT The copyright law of the United States - Title 17, United States Code - concerns the making of photocopies or other reproductions of copyrighted material... Columbia University Library reserves the right to refuse to accept a copy order if, in its judgement, fulfillment of the order would involve violation of the copyright law. AUTHOR: COULTER, CORNELIA C. TITLE: RETRACTIO IN THE AMB SIAN AND PALATINE... PLACE: BRYN MAWR, PA. DA TE : 1911 COLUMBIA UNIVERSITY LIBRARIES PRESERVATION DEPARTMENT Master Negative U BIBLIOGRAPHIC MICROFORM TARGET Restrictions on Use: Original Material as Filmed - Existing Bibliographic Record I- J. N j.u I ^^ r r< \3 .»:: m- *t o / ... i.) - o I wrrt..!.^ • \mJ ••• -^• i.~7 ID5MYCG92-B99A CC:;9665 BLT.Tim CP s pau PC SB GIG G40 050 :IGG 1 L : ©ng PDx 191 .1./ DCF- INT RTYPsa CSC." GPC: REP: • » Wb. •«»• •«■ S"r : p MODS D 1 : CPI sO OR.'! POL: DM: RRs .1.21091 MrxK>- I cNNC Qll=^lb.B9 vol. 10 Coulter H Cornelia Catlin , ~ I d 1885 FRNs BNRs F 1' C s G FSI JiG COL s MS EL s u ATC: CON : b ILC: EMLs AD: 02- 11 -92 UD: 02- 11 -92 1 I s N3 BSE r ij 245 IG Retract^it :i. o in the Ariitaro:-;ian and Palatine recensions of Plau tus- I hCmi c o f o r m ] ; -'• ' b a s 1 1 1 d y o f the P e r s a , P o e? n i-i 1 ii b m P b e ti d o 1 ti s , S t i c: h u s and T r i n u m m 260 3GQ B .^ =■•■ I c b y C D I" n e 1 i a C a t D.. i n (2 o u 1 1 e r . Bryn Mawr ^ Pa - w = I bBryn Mawr CcO.lege, -101911 P I 118 p^:::^ I c23 cm "J- 90 Bryn Hawr college? monographs. Moriograpl-i ser ies., == I vvol It ft ^GA B i bl i o g r* a p 1 1 y s p » 1 1 5 1 1 8 TECHNICAL MICROFORM DATA REDUCTION RATIO: ^^^ IMAGE PLACEMENT: lA (HA; IB IIB DATE FILMED: P3_l^_iQ_Q2.__ INITI ALS__^A>XA;^a^ FILMED BY: RESEARCH PUBLICATIONS. INC WOO DBRinriR. C.T 1 r Association for information and Image Management 1100 Wayne Avenue, Suite 1100 Silver Spring, Maryland 20910 301/587-8202 Centimeter 1234 56 789 10 n nil iiiiliiiiiiiiiliiiiiii::^^ I 1 Tf 1 I _ ^^^^-^y^-^™- I I I I 1 .^T I I I I I I II I Inches I 1.0 1.25 ■ 10 |63 tiS. I.I i 2.8 1 3.2 3.6 4.0 lUIAU 1.4 II I 1 1 I I 4 2.5 12 13 14 15 mm ii|iili|iili[iililiil 2.2 2.0 1.8 1.6 TTT llilllllllllilllllllll I I I I MflNUFflCTURED TO fillM STRNDfiRDS BY fiPPLIED IMRGE, INC. Bryn Mavvr Collhge Monographs Monograph Series, Vol. X ietractatio IN THE Anibrosian and Palatine Recensions of Plautus A STUDY OF THE PERSA, POENULUS, PSEUDOLUS, STICHUS AND TRINUMMUS BY CORNELIA CATLII^ COULTER Bryn Mawr, Pennsylvania, U. S. A., November, 1911 BRYN MAWR COLLEGE Bryn Mawr College Monographs Monograph Series, Vol. X ANNOUNCEMENT The Bivn Mawr College Monographs are issued in two series : the first, the Monograph Series, contaiuing articles that appear here for the first time ; the second, the Reprint Series, contain- ing reprints of articles that have appeared in other journals. The monographs are edited by a committee of the Faculty of Bryn Mawr College, consisting at present of President M. Carey Thomas, ex officio; Professor Elmer P. Kohler, chairman; Professor Arthur Leslie Wheeler, and Professor Carleton Brown. Retractatio IN THE Ambrosian and Palatine Recensions of Plautus A STUDY OF THE PERSA, POENULUS, PSEUDOLUS, STICHUS AND TRINUMMUS BY CORNELIA CATLIN COULTER Bryn Mawr, Pennsylvania, U. S. A., November, 1911 BRYN MAWR COLLEGE CONTENTS M _ PAGE Introduction. 1) The Sources of the Inconsistencies and Contradictions in the Text of Plautus. a) Freedom in the Handling of the Greek Original . . 1 b) " Plautine Carelessness " ...... 6 c) Retractatio g d) Interpolatio 20 2 ) The Manuscripts of Plautus 23 I. Persa 27 II. POENULUS 42 III. PSEUDOLUS 67 IV. Stichus g3 V. Trinummus 97 Conclusion 209 Bibliography 215 INTEODUCTIOK I The twenty-one plays of Plautus, in the form in which they have come down to us, present contradictions and inconsistencies which must impress even an uncritical reader, and which demand explanation at the hands of the Plautine scholar. For some defects we must hold Plautus himself directly responsible. We know that he wrote in an age of rude technique, and that he allowed himself great liberty in adapting Greek comedies for the Koman stage. Terence (Adel Prol. 6-10) says that Plautus, in presenting his Commorient es, a version of the I.vpa7ro0uj^. In addition to these secondary versions, the manuscripts sometimes present apparent attempts to shorten scenes by omission.23 People who had to stretch their legs before a Plautina longa fahula came on the boards (Pseud. Prol.) would be likelv to welcome some cutting. The method of indicating such omissions, according to Lindsay, was to adscribe in the margin, opposite the beginning of the passage in question, the line or lines which were immedi- ately to follow.24 This seems to be the purpose of the repetition in the first scene of the BaccJiides.^^ Line 73 appears after 1. 64, as well as in its proper place, and the verses that intervene (65-72) contain a series of puns and quibbles which might easily be spared. Most. 553, 557-559 are repeated after 1. 549, as if to indicate the possibility of shortening this rather wordy passage.^^ 23Gootz in Acta soc. phil. Lips. 6 (1876), pp. 268 f. W. Lindsay, Ancient Editions of Pluutus, Oxford, 1904, p. 1, note a (end). ^Lindsay in Amer. Journ. Phil. 21 (1900), p. 27. ''Goetz, Ed. Mai. (1886), ad loc. ^H. Kellermann in Comm. phil. Jen. 7 (1903), p. 134. Cf. Lindsay, Ed. (1905), ad loc. INTRODUCTION 13 In the Trinummus, some stage-manager seems to have cut out 11. 362-368 — pious reflections which are somewhat tedious after the protracted moralizing of the first part of the scene — in order to come to the point in 1. 369.^"^ This line is inserted in A after 1. 361 ; and in P, not only 1. 369, but 1. 368 as well, has slipped into this place. To this stage revision of Plautus has been given, in recent years, the technical name retractatio. It corres- ponds to the Greek hao-Kevrj (which was Eitschl's term for it), and though, like the word contaminatio, it has no support in classical Latin, it is so convenient that it may well be retained. The first work on retractatio was done by Osann,-^ in the last three chapters of his Analecta Critica. He summed up the evidence for per- formances of Plautus after his death, with especial em- phasis on the Casina-prologue, and argued that many variations in our texts must go back to these later pro- ductions. The actors' versions were, in his opinion, responsible for the citation by grammarians of verses not in our text, or of different forms of verses which we have, as well as for parallels in the manuscripts of Plautus. His first two points are open to question, and so is his assumption of post-Plautine subject-matter in such passages, e. g., as Bacch. 1072-1075, which speaks of a triumph as a common thing, or Cas. 699, which men- tions a vilicus. But his principle that where parallel versions exist both should be preserved, is correct; and his opinion agrees in several cases (e. g., Bacch. 511-520; Pers. 442-443, 433-436) with the results of more recent scholarship, while in others (Capt. 1022, Most. 548 ff.. "Lindsay in Amer. Journ. Phil. 21 (1900), pp. 27 f. ^F. Osann, Analecta Critica, Berlin, 1816, pp. 141-204. 14 EETEACTATIO IN PLAUTUS Poen. 1042 ff.)? ^^^^ though his solution may not be correct, he has recognized the difficulty. The weak points in Osann's theories were patent enough, and it was these, rather than his real contributions to the study of Plautus, that scholars of his generation chose to consider. Ladewig -^ pointed out the danger of infer- ring from discrepancies in grammarians' citations that different versions of a play existed, when the same result might have been brought about by many other causes — lacunae in our text, carelessness in quoting, or error in the text tradition of the grammarians themselves. EitschFs Parerga ^^ contain a number of valuable obser- vations on the period of the Plautine Kevival ^^ and its effect on the manuscript tradition — especially changes of title,^- new versions of the prologues,^^ and the second ending of the Poenulus,^^ But in reality Eitschl was rather conservative in applying his theories to single points in the text. He notes various kinds of addition and repetition in the Trinummus, but, even though he observes that 1. 312 " perbene, si numeros spectas, f actus est versiculus," he takes this merely as proof of " vetustas interpolationis et exercitatio interpolatoris." ^^ The sec- tion De Plauti Bacchidihus ^^ contains a spirited reply to Osann on this one play. Eitschl argues that the difficulties of the play are caused mainly by the loss of a large portion at the beginning, and that the confusion in the text which leads Osann to suspect parallel versions, 29 Ladewig in Zeitschr. f. Alter tumswiss. 1841, coll. 1079 ff.; Rhein. Mus. 3 (1845), pp. 179-205; 520-540. ^ F. Ritschl, Parerga Plautina et Terentiana, Leipzig, 1845. «Pp. 89 ff. ^Pp. 157 ff.; 233 ff. »3Pp. 180 ff. ^Pp. 601 ff. ^Pp. 509 ff. ^Pp. 389 ff. Cf. Ritschl in Rhein. Mus. 4 (1846), pp. 354 ff.; 567 ff. (=Opusc. II., pp. 293-374). INTRODUCTION 15 is due to the usual carelessness of scribes. It is possible, however, to trace a gradual shifting of position in Eitschl's editions of the plays. When he edited the Trinummus, he committed himself no further than to say: ^'^ ^' Ceterum diversarum recensionum tenues quasdam tanquam reliqui- as non infitior hodie quoque superesse, sed earum ex anti- quioribus ut puto saeculis repetendarum, partim autem sua sponte natarum, partim critiea opera grammatiooruin. par- atarum "; and the note on Bacch. 377, 378 (Ed. 1849) is equally cautious : ^^ " ;N'on possunt ab eodem posita esse qui versus 380, 381 scriberet: quamquam iam a Ils'onio lecti, qui priorem affert." Eitschl recognized traces of SLacr/cevrj in the Sticlius,^^ especially in the first scene, and in the names of the sisters, which appear in different forms in the two families of manuscripts ; he thought that Pers. IV., 9 must have suffered shortening ; "^^ and finally, in his preface to the Mercator/^ he not only distinguished between Bcao-fcevi] and the dittography of the scribes, but assumed that the prologue, either as a whole or in part, and three certain cases of " parallels," dated from the Plautine Eevival at the beginning of the seventh century A. U. C. Bergk, in his reviews of EitschPs edition of Plautus,^- had noted that certain cases of dittography were due, not to the changes of a grammarian or copyist, but to altera- tion for the stage; and he observed with some pleasure Eitschl's gradual conversion to his own point of view. "Ed. (1848), Praef., p. livii. {=Opusc. v., 325). ^Ed. (1849), ad loc. ^Bd. (1850), Praef., pp. x ff . 40 Ed. (1853), Praef., p. ix. « Ed. (1854), Praef., pp. vii f. *^Th. Bergk in Zeitschr. f. Altertumswiss. 6 (1848), coll. 1124- 1149 {=Opusc. I., pp. 3-29) on Trinummus ; 8 (1850), coll. 325-348 {—Opusc. I., pp. 29-53), on Miles, Bacchides, Stichus. 16 EETEACTATIO IN PLAUTUS Even Ladewig so far changed his opinion about this " grundverkehrt " theory of Osann's/^ that he published in 1861 ^^ a series of conjectures of his own, made on the basis of Ritschl's text, in which he recognized a number of parallel passages and later additions. Thus at the end of half a century, retractatio had won a place as a recognized phenomenon in Plautus. The work since 1870 has consisted chiefly of a more detailed study of single plays, and of theorizing as to when and how the later versions entered our text tradition. Oskar Seyffert '^ in 1874 suggested a theory which he has since elaborated: namely, that the variant lines of our text of Plautus were, at one period in the history of the manuscripts, written in the margin; and that their present position, sometimes before, sometimes after, the place where they belong, is due to careless copying from the margin into the text. Seyif ert's statements were made on the basis of a few suspicious passages. Two years later Goetz ^« made a study of a large number of ditto- graphies, dividing them into four general classes: (1) where one version immediately follows the other; (2) where one is inserted in the middle of the other; (3) where the two are separated by an interval; (4) where only the later version is preserved. As characteristics of the rewritten passages he notes an effort for rhythmical correspondence, humor of a very poor quality, and especi- ally the attempt to shorten excessively long scenes. He observes that the dittographies generally appear in both families of manuscripts, and he therefore concludes that *3Lade\vig in Rhein. Mus. 3 (1845), p. 523. « Ladewig in Philol. 17 (1861), pp. 248-269; 452-480. «0. Seyifert, Studia Plautina, Berlin, 1874, pp. 10 ff. *«G. Goetz in Acta soc. phil. Lips. 6 (1876), pp. 235-326. INTRODUCTION 17 they probably stood in the common archetype, into which the smaller ones at least were introduced as marginal adscripts. Another general treatment of the plays was undertaken by Langen,"*^ who, in 1886, discussed the troublesome points of each play under three headings: (1) repetition of thought; (2) discrepancies in subject- matter; (3) spurious and suspected passages. The great value of his work lies in its completeness. Where two versions unquestionably exist, Langen's comprehensive study enables him to decide which is Plautine; and on the other hand he can argue that certain faults which are characteristic of Plautus everywhere are not to be charged to retractatio. In the ten years between Goetz's article and Langen's book students of Plautus produced numerous articles and studies on the individual plays. Many of them made some valuable contributions to our knowledge of retrac- tatio, but nearly all went too far in their search for traces of dittography.*® Within the last ten years there has been a revival of interest in the subject, and a second (and perhaps a more moderate) set of dissertations has appeared. Recent study of retractatio has been more or less closely connected with critical estimates of the manuscripts. Since the time of Ritschl,^^ it had been a generally accepted idea that the two families of manuscripts which we have to-day — A, the Ambrosian Palimpsest, written in rustic capitals and dating probably from the fourth cen- 4T P. Langen, Plautinische Studien, Berlin, 1886. "See, for example: W. Brachmann, De Bdcchidum Plautinae retractatione scaenica, in Leipz. Stud. 3 (1880), pp. 59-187; A. An- spach, De Bacchidum Plautinae retractatione scaenica, Bonn, 1882. «Ritschl, Ed. Trin. (1848), pp. xxxviii flf. 18 EETRACTATIO IN PLAUTUS tury, and P, the Palatine family, consisting of a half dozen minuscule manuscripts— went back to a common archetype ; and numerous studies were made to determine which of these families was the more trustworthy.^^^ In 1885, in the preface to his edition of Plautus,"^ Leo stated a theory (more fully developed later ^2) which has given the impetus to much of the recent work on the manuscript question. The theory is : that our two fam- ilies of manuscripts, A and P, represent reading copies of an edition of the twenty-one plays made by Probus (a grammarian of the Flavian period) or his school. Both this edition and the first published edition of Plau- tus, in the age of Lucilius and Accius, were made on Alexandrian principles— i. e., everything in the sources was preserved in the text, and critical symbols indicated spuriousness or referred to notes in the commentary. The variant readings and parallel versions in our manu- scripts accordingly owe their preservation to the gram- matical work of these two periods. The last point was disputed by Seyffert. He had evidently been collecting material to support his theory of marginal variants during the twenty years that had elapsed since the publication of his Studia Plauiina.^^ He now took as a starting point Leo's remark ^^ that the repetition of Men. 1037- 1043, in slightly different form, between 1. 1028 and 1. 1029, was due to a grammarian of the second or third ^Such as M. Niemeyer, Be Plauti fahularum recensione duplici, Berlin, 1877; B. Baier, De Plauti fabularum recensionibus Am- brosiana et Palatina commentatio critica, Breslau, 1885. "Leo, Ed. (1885), Praef. *^Leo, Plant. Forsch., pp. 1-53; Plmit. Cant., pp. 5 ff . ^E. A. Sonnenschein in Trans. Amer. Phil. Ass. 24 (1893), p. 7, quotes Seyffert " in a private communication " on this question. ^Leo, Plant. Forsch., pp. 15, 16. INTRODUCTION 19 century, who copied this version into the margin of his manuscript. Seyffert ^^ questioned Leo's statement, and, with a wealth of suggestion and illustration, advanced the contrary theory: that the variant in question appeared in the common ancestor of A and P, and that at least a considerable number of the differences between our two recensions can be traced back, not to the activity of grammarians, but to marginal or interlinear variants in the archetype. Leo's theory was again attacked by Lindsay, who de- voted his " Ancient Editions of Plautus " ^® to a state- ment of his own views. According to Lindsay, the text- tradition of Plautus followed after his death ^^ two main divergent channels," " the one adhering to the genuine ^ ipsa verba ' of the poet, the other exhibiting all the alterations, curtailments, or amplifications introduced by the stage-managers of the Revival time in order to make the performance pleasing to the audience of the day.'' There was a certain amount of " mixture " of these two versions, and of addition from grammarians and com- mentators, but in general the Ambrosian Palimpsest represents the first of these traditions, the genuine " ipsa verba " of the plays, and the Palatine text shows the " Revival " adaptations. This view, " conservative " and " optimistic " as Lind- say thought it, has aroused much opposition. The Italian reviewer ^'* who criticized the book found himself " piena- mente d'accordo " with the views expressed there, but English and German critics have treated it less kindly. 56 Seyffert in Berl. Phil. Woch. 16 (1896), coll. 252-255; 283-288. "Lindsay, The Ancient Editions of Plautus, Oxford, 1904, espec- ially pp. 35-37; 142-150. " Aurelio-Giuseppe Amatucci in Riv. di Fil. 34 (1906), pp. 605-608. 20 EETEACTATIO IN PLATJTUS 21 Leo ^^ thinks that Lindsay has too little regard for the views of other scholars on the passages which he dis- cusses, and is too anxious to find "stage alterations" wherever P differs from A, even in single words. " Und wenn man Lindsay recht geben will," he concludes, " dass meistens A die urspriingliche Lesart bietet, so folgt damit doch nichts fur seine Hypothese, dass A in ungebrochener Linie auf Plautus, P in eben so ungebrochener Linie auf die gleich nach Plautus eingetretene Uberarbeitung zu- riickgehe." And Sonnenschein ^^ puts his opinion con- cisely: "Mr. Lindsay's conception of the independence of the two recensions from so early a date will not, I think, be found to hold water." «^ One class of difficulties still remains, a class which is found to a greater or less extent in all classical authors, and which may therefore be briefly dismissed. This kind of alteration, which goes by the name of interpolation originates later than any of the others, and is due chiefly to the work of grammarians and commentators. It often arises through the addition in the margin of a parallel passage from some other play, which in the course of time is taken into the text. The earlier stage of this process appears in the manuscript B. Cure, u., 1 is con- cerned with the physical condition of the leno Cappadox. In the margin opposite 11. 222, 223 (though apparently intended as an adscript to the phrase oculis herheis in 1. 231) are the words: «Leo in Gott. Gel. Anz. 166 (1904), pp. 358-374. ^» Sonnenschein in Class. Rev. 19 (1905), pp. 311-316. •"One great objection to Lindsay's theory, the existence of a large number of common errors in A and P, was emphasized by Leo {Gott. Gel. Anz. 1904, pp. 364 flf.), and has since been investigated in detail by Eugen Sicker {Philol Suppl.-Band. xi. (1908), pp. 179-252). \ r i [ INTEODUCTION Solent tibi oculi duri fieri censesne locustam esse, and opposite 11. 242, 243 : album atrum vinum potas quid tibi quaesito opus est. Both couplets are confused versions of lines from Men. v., 5 (923-924; 915 ff.), the scene in which the physician is examining Menaechmus I. for symptoms of insanity ; and Eitsehl ^^ thinks it probable that they were set down as parallels for the similar scene in the Curculio. So Stick. 722 Quid igitur? quamquam gravatus fuisti, non nocuit tamen seems to have been added as a parallel to 1. 763 gravate, and then to have been copied into the text after 1. 766.^^ In the same way, a gloss on a single word or the ex- planation of a difficult phrase may be added between the lines or in the margin, and so creep into the text. This accounts for some extra lines and some divergence of tradition in our manuscripts. Poen. 1020 Ut hortum fodiat atque ut frumentum metat is apparently an explanation of palas and mergas two lines above; and Pers. 321 rogasti P (for orasti A) f^ 408 periure A (for iniure P and JKTonius) ; Poen. 342 occuUo A (for abstruse P) all seem to be glosses. We are indebted to Eitsehl for clearing away many of these intrusions into our text, and in particular for point- "Ritschl in Philol. 1 (1846), pp. 300 ff. (=0pM5C. ii., pp. 274 ff.) "Ritschl in Philol 1 (1846), p. 305 (=Opusc. n., p. 281). ""Cf. Lindsay, Anc. Edd., p. 73. 22 EETRACTATIO IN PLAUTUS ing out how often the citation of a parallel passage might cause confusion.^^ Goetz, in his discussion of Ditto- graphien im Plautustexte,^' and Kellermann in the article entitled De Plauto Sui Imitatore,''^ made similar studies, but in each of these the work on interpolatio was simply a preliminary to the main investigation. The century of philological work on Plautus, the course of which has just been outlined, makes it possible to sum- marize the causes of difficulty and inconsistency in the plays as follows: (1) free treatment by Plautus of his Greek originals, resulting in omissions or in the combina- tion of two originals into one Latin play; (2) Plautine carelessness in detail ;«^ (3) changes in the text made during the Kevival, a generation after the death of Plautus; and (4) parallel adscripts, explanations, and glosses, added by later scribes. The mass of difficulties grouped under the third head of this summary (retradatio) forms the subject of the present investigation. An effort has been made first of all to determine whether the two great families of manu- «*Ritschl in Philol 1 (1846), pp. 300-314 {=Opusc. ii, pp. 274- 291). «= Goetz in Acta soc. phil. Lips. 6 (1876), pp. 236 flf. '^Kellerman in Comm. phil. Jen. 7 (1903), pp. 131 ff. •'Inconsistencies due to the first two causes are traceable to Plautus himself. Further back than Plautus it is almost im- possible to go. But as the charm of ISIenander shows through even the mutilated Cistellaria, so it is possible that a few of the defects in Plautus may be referred to his Greek models. Wilam- owitz {Index schol. Gott. 1893-1894, pp. 13 flf.) has made it probable that the poor technique and crude character-drawing of the Persa go back to an original in Middle Comedy, and it may be that Acts I.-III. of the Poenulus, which are distinctly poorer than the other half of the play, merely reflect the weakness and verbosity of the Greek original. (See the analysis below). INTRODUCTION 23 scripts differ in the amount of retractatio that they indi- cate, and secondarily to throw new light on the general problem of retractatio. Since the two groups of manu- scripts can be compared only where it is possible to know the contents of the Ambrosian Palimpsest, the work is limited to the five plays best preserved in that manu- script (Persa, Poenulus, Pseudolus, Stichus, Trinum- mus), and primarily to those portions of the plays the text of which is contained in both A and P. Except in cases where the source of the confusion is doubtful, diffi- culties due to other causes than retractatio are excluded. Of the passages suspected of retractatio, only those in which its presence seems fairly probable are discussed; others, w4iich can lead only to questionable conclusions, are listed in footnotes. ^^ Citations are made from the Goetz-Schoell text of Plautus (Editio Minor), ^^ and Studemund's Apographon "^^ is taken as the basis of the work on A. Mention has already been made of the two families of manuscripts on which our text of Plautus is based. A fuller description is a necessary preliminary to an in- vestigation w^hich must constantly refer to the manu- scripts. Until the early part of the nineteenth century, only one of these families, the Palatine, was known. This group, which received its name from the fact that •* Variations of a word or phrase are not considered. Of course Bome of these may be due to retractatio {e. g., Pers. 597 me inpulsore atque inlice A, sti^isu atque inpulsu meo P; Poen. 343 palpas et lallas A, caput et corpus copulas P) ; but the majority of cases are probably to be assigned to scribal error or interpolation. ~G. Goetz and F. Schoell, Plauti Comoediae, Leipzig, 1892-1896 (revised 1904-1909). '" W. Studemund, T. Macci Plauti Fahularum Reliquiae Amhros- ianae, Berlin, 1889. 24 EETKACTATIO IN PLAUTUS its most important representatives, B and C, were at one time in the library of the Elector Palatine, comprises six or seven manuscripts, ranging in date from the early tenth to the late twelfth century. Though the manu- scripts vary greatly in authority, their common origin is a recognized factJ^ Therefore, for the purposes of this paper, variations between the individual manuscripts are disregarded, and the whole family is designated by the symbol P. Of the other family of manuscripts there is only one representative, a palimpsest in the Ambrosian Library at Milan, written about the fourth century, and covered in the seventh or eighth century with parts of the Book of Kings. It was discovered by Cardinal Mai, under whose direction an imperfect collation was published in 1815 with the title M. Accii Plauti Fragmenta Inedita. Even at that time the manuscript was in a bad condition. Some leaves were lost altogether, and those that remained had been injured by the cleaning process, by the ink of the second writing, and by the chemicals that Mai used to restore the original text. With careless handling and with the passage of time, some of the pages were torn, and others decayed so as to leave only a border around the edge, while the ink gradually faded. Ritschl, writing in 1837,'^^ lamented the loss of some passages (e. g.j in the Cistellaria), which could have been read when the palimpsest was first discovered, and censured Mai for not having published a complete collation at " See the stemma codicum in Ritschl's edition of the Trinummus (1848), and the discussions in more recent critical editions — e. g., Lindsay's edition of the Captivi, London, 1900. "Ritschl in Zeitschrift f. Altertumswiss . 1837, coll. 737-758 {=Opnsc. II., pp. 166-197). INTEODUCTION 25 that time ; and Geppert found in 1846 "^^ that much that had been visible even the year before had faded away. Studemund, "that scholar-hero," as Sonnenschein calls him,*^^ spent the greater part of twenty-five years in deciphering the palimpsest,"^^ only to die before his results were given to the world. They were brought out by his friend Oskar SeyfFert, and are now accessible to the stu- dent of Plautus, in a form which is of infinitely more value to him than the manuscript itseK would be. Fortunately, even the smallest fragments are of value for the study of retractatio, A few letters at the begin- ning or end of the lines, even where all the rest of the passage is gone, show whether or not a certain passage appeared in A, and what was the order of the lines. Even the contents of a missing sheet may sometimes be estimated accurately. Difficulty arises when the missing section contained a canticum, in which case there can be no certainty as to how the lines were divided, or unusually long verses, like trochaic septenarii or iambic odonarii, which are run over in varying proportions, sometimes only one line out of thirty-five, sometimes three lines out of four. Passages containing scene-head- ings also cause difficulty. In general, the scene-division "^^ of A corresponds to that of our printed texts; but some- times, (e. g., Pseud, IV., 5, 6) two scenes are run together under the same rubric, and sometimes (e. ^., Pers, IV., 7), where a single character leaves the stage, there is no 73 K. Geppert, Vher den Codex Ambrosia/nus u. seinen Einfluss auf die plauti/nische Kritik, Leipzig, 1847, p. 28. "Sonnenschein in Trans. Amer. Phil. Ass. 24 (1893), p. 10. "W. Studemund, T. Macci Plauti Fabularum Reliquiae Amhros- ianae, Berlin, 1889 (Edited by 0. Seyffert). Prooem., p. xxii., Seyffert's note. "There is no division into acts in any of the manuscripts. 26 EETRACTATIO IN PLAUTUS new scene-heading. On the other hand, A leaves a space of one line before the speech of the Caterva (CapL 1029), and makes a similar break at the entrance of Satiirio (Pers, 726). The amoimt of space left for the scene-headings is not always the same. In about 75 per cent of the cases extant the scene-headings occupy two lines, but the exceptions occur so irregularly that one can never be sure how much space was occupied by the missing headings. In the present paper, calculation of the amount of text in lost sheets of A is for the most part confined to continuous passages of trimeter without scene-divisions. Even here, of course, the results are not absolutely certain ; in other places, though the calculation has occasionally been made, it has even slighter claims to accuracy. CHAPTER I. PERSA The hero of the Persa is a slave, Toxilus, who holds a position of trust in his master's household, and has been left in charge during the master's absence. Toxilus is in love with Lemniselenis, a girl in the service of the leno Dordalus, and is anxious to obtain her freedom. But as he himself, being a slave, can neither purchase another slave nor be patronus of a freedwoman, he ar- ranges with Dordalus that on a certain day he shall pay the required sum of money, and that Dordalus shall then go through the form necessary to set her free.^ The day approaches, and Toxilus has not succeeded in getting the money. He therefore persuades his friend Sagaristio, who is likewise a slave, to lend him six hundred nummi, promising to repay him in a few days. Sagaristio pro- vides the sum in question by appropriating funds given him by his master for the purchase of cattle; the money is paid, and Lemniselenis is set free. In the meantime, Toxilus has arranged another scheme to make the leno himself pay back the sum that has been borrowed from Sagaristio. The daughter of the parasite Saturio is dressed up as if she came from the Far East; Sagaristio puts on Persian garb, and is introduced to Dordalus as a messenger from Toxilus' master, who has an Arabian girl for sale. Dordalus sees the girl, is so charmed with her that he is willing to make the purchase suo periculo, ^ Cf. U. V. Wilamowitz-Moellendorff in Index schol. Gutt. 1893- 1894, p. 18. 27 28 EETRACTATIO IN" PLAUTUS PEESA 29 and pays the exorbitant sum of sixty minae ^ on the spot. Of course the natural sequel follows. As soon as the pseudo-Persian is out of the way, the father of the girl appears and threatens a law-suit; and the play ends with an uproarious banquet of Toxilus, Sagaristio, and Lemniselenis. The Persa stands alone among the plays of Plautus in showing indications of an original in Middle Comedy. Wilamowitz ^ was the first scholar to suggest this date for the original of the play, and his theory, which has been accepted by Seyffert,^ Hueffner,^ and Leo,^ seems to be well established, in spite of a recent attempt to combat it.*^ Wilamowitz's strongest argument is that the play represents the Persian Empire as still intact. The story of a messenger who comes ex Persia (1. 498), with news of the capture of Chrysopolis in Arabia by the Persians, points to a date before the conquests of Alex- ander. For, while it is true that the account of the Persian campaign makes no pretensions to truthfulness,^ still the Athenian public must have demanded a certain amount of verisimilitude, and a story in which Persians figured as the chief actors would have had no point after the downfall of the Persian Empire. Therefore we may ^The regular price was twenty or thirty minae. Cf. Ritschl, Opmc. II., p. 308, note. ^Wilamowitz in Index schol. Gott. 1893-1894, pp. 13-26. *Seyffert in Bursian's Jahresber. 1895, pp. 39 fiF. »F. Hueflfner, De Plauti Comoediarum Exemplis Atticis, Gottin- gen, 1894, pp. 70 f.; 74-76. «T^eo, Plant. Forsch., p. 110; of. Leo in Hermes 41 (1906), pp. 441 ff. 'M. Meyer, De Plauti Persa, in Comm. phil Jen. 8 (1907), pp. 145-191. * Cf. Meyer, De Plauti Persa, pp. 181 ff. assume that the original of the Persa was a Greek play of the age of Demosthenes. To the unfinished technique of the Greek original may be due certain defects in the plot and the character- drawing of the Persa. The exceptional subject-matter of the play attracted the attention of Camerarius, who commented on it : " Argumentum f abulae est exile, ama- tionis servilis." Although servilis amatio is touched upon in some of the other plays (Amph, 658; Mil 1007-1008; Stick, 431-434), there is no other play in which the love of a slave forms the main theme. ^ But this argu- mentum exile may be merely another indication that the play originated in Middle Comedy. Slaves are promi- nent in several of the earlier plays of Aristophanes (we remember Dionysus and Xanthias in the Frogs, the Paphlagonian and his rival in the Knights); and in the Plutus, which stands on the border-line between Old and Middle Comedy, the slave Carion is one of the most important figures. It would have been only natural, therefore, if this tradition had continued and slaves had played an active part in the plots of Middle Comedy. ^^ The composition of the Persa is poor, although the theory of contaminatio, suggested by Ladewig ^^ and dis- cussed at length by van Ijsendijk,^^ j^^g ^^^ ^^^^ proved. Van Ijsendijk thought that the Persa was made up of two plays: (A) the Persa (our present Acts IV.-Y.), in which a free-born girl is put through a form of sale, "Meyer, De Plauti Persa, p. 152. '**W. Stiss in R}iein. Mus. 65 (1910), p. 456; cf. Wilamowitz in In^ex schol. Gott. 1893-1894, pp. 18 flf. "Ladewig, Vber den Kanon des Volcatius Sedigitus, Neustrelitz, 1842, pp. 38 ff. "A. van Ijsendijk, De T. Macci Plauti Persa, Utrecht, 1884. 30 KETKACTATIO IN PLAUTUS PEESA 31 and with the proceeds of this sale a lover gets his arnica; (B) the Boaria or Sagaristio, in which a slave steals the money given him by his master to buy cattle, and so helps the erilis filius to get his arnica free. But, as van Ijsendijk's critics have observed,^^ the two lines of action start together, and are connected throughout the play; and the whole object of the pretended sale is to get the leno to pay for Lemniselenis with his own money, and so repay the loan of Sagaristio. Anyone who is not convinced of the unity of the Persa need only compare it with one of the certainly contaminated plays {Miles or Poenulus). The length of the Persa is only 857 lines, while each of the others has 1400 lines or more ; and the two threads of the story are closely inter- woven here, while in the others the two plots are quite distinct, and the line of division clearly marked. The banquet scene at the close is only loosely con- nected with the rest of the play, but there is no reason for suspecting a separate origin. The persons involved are the same as in the earlier scenes, and their conduct is quite consistent; Paegnium, for instance, shows him- self just the same merry wag in Act V. as he was in Act II. To one who remembers the festivities with which Aristophanes ends his plays — the banquet spread for Dicaeopolis, the weddings of Peisthetairus and Try- gaeus — the last act of the Persa will sound a familiar note ; and the discomfiture of the leno in the final scene will seem closely akin to the horseplay of the Old Comedy.^* ''Seyffert in Bursian's JaJiresher. 1886, pp. lllf.; Meyer, De Plauti Persa, pp. 159 ff. " Wilamowitz in Index schol. Gott. 1893-1894, pp. 22 f.; Suss in Rhein. Mus, 65 (1910), p. 456. The character-drawing shows the broad lines of the Old Comedy rather than the finer touches of the ^N'ew. The leno Dordalus, stupid and gullible to the last degree, is not far removed from some of the exaggerated figures of Aristophanes. The daughter of Saturio, too, with her elevated diction and her rhetorical sentences, may reflect the technique of the transitional period, when a free-born woman was as yet an unfamiliar figure on the comic stage, and the playwright had to borrow some touches from the heroines of tragedy. ^^ 440 445 TO. Cape h6c sis. DO. Quin das? TO. Nfimmi sexcenti Mc erunt 437 ProW, numerati: fdc sit mulier libera, Atque lific eontinuo addfice. DO. lam faxo hTc erit. Non li^rcle quoi nunc hoc dem spectandfim scio. TO. Fortasse metuis in manum concr^dere? tMirtim quin citius iam a foro argentdrii Abefint quam in cursu r6tula circumu6rtitur. Abi istSc trauorsis Sngiportis fid forum: Eadem fsta[ha]ec facito mClier ad me trfinseat Per h6rtum. DO. Jam hie faxo aderit. TO. At ne pr6palam. DO. Sapi6nter sane. TO. Sdpplicatum crSs eat DO. Ita h^rcle uero. TO. Dflm stas, reditum opdrtuit. JfJfO ff. The difiiculties of this passage have puzzled scholars for the last hundred years, and the number of explanations is almost equal to the number of writers on the subject. The dialogue runs quite smoothly down to 1. 439 ; then Dordalus says (1. 440), evidently referring to the money that Toxilus has just paid : Non hercle quoi nunc hoc dem spectandum scio — i. e,, spectandum, utrum prohum sit necne,^^ Toxilus' reply (1. 441): Fortasse " Wilamowitz in Index schol. Gott. 1893-1894, pp. 25 f. "Langen, Plaut. Stud., p. 334. 32 EETKACTATIO IN PLAUTUS metuis in manum concredere, is disjointed and hardly intelligible, since it lacks both direct object and dative of the person with in manum concredere. The couplet which follows (11. 442-443) bears a striking resemblance to 11. 433-436 : DO. Mirtim quin tibi ego cr^derem | ut id6m milii 433 Facer^s quod partim fdciunt argentfirii: Ubi quid credideris, citius extempld foro Fugiflnt quam ex porta Ifldis quom emissQst lepus, but it has no grammatical construction and no logical connection with the rest of the dialogue. Then (11. 444- 448) come certain directions about the freeing of Lemni- selenis. Out of the tangle we can separate 11. 442-443, which are clearly a fragmentary parallel version (probably with a line lost between quin and citius) of 11. 433-436. The first version, a comparison of the money-changers to a hare at the games, may be Plautine, since we know of the use of animals in the games before the death of Plautus. (Liv. 39, 22 tells of a venatio data leonum et pantherum in the year 186 B. C.) This reference is paralleled by one to the ostrich in an earlier scene (1. 199 marinus passer per circum), and the two together make it probable that there was a keen interest in venationes at the time that the play was produced. The second simile (11. 442-443), which is much more general, may have been added after this interest had declined. "With this intrusive couplet removed, it is possible to trace a certain connection in the dialogue. Toxilus pays the money, saying, " Set the woman free, and then bring her to me immediately." Dordalus murmurs, half to himself, as he examines the money, " How am I to know PERSA 33 whether this is good or not ? " and Toxilus adds, seeing his hesitation, "Are you afraid to hand her over to me?''^^ Then he goes on rapidly: ''Take this side street to the forum ; [have your money tested and set her free] ; then bring her back secretly by the same path. [I want her company to-day myself] ; to-morrow she may go to sacrifice for her liberation." But even so there is much that is not clear. We do not quite get the force of metuis . . . concredere; no explanation is given for the injunction to secrecy (11. 444-446) ; and the reason why the sacrifice is to be postponed until the morrow must be inferred. It seems probable that the end of this scene, like a later section (11. 738 if.), has suffered considerable cut- ting, which has left only the bare essentials of the dia- logue, without the necessary connection. If this is so, we may put down 11. 442-443 as a parallel version of 11. 433-436, and say that 11. 440-441, 444-448 are a shortened version of a scene now lost. P has the verses in the order of the Gtoetz-Schoell text. A agrees with P through I. 443, then deest 444-478 (35 lines of senarius). The passage contains three scene-headings, which, if given in the regular form, would have occupied 6 lines more, making a total of 41 lines. The one sheet missing in A would have given space for only 38 lines in all. But it is possible that some of the headings were missing in A (Cf. iv., 9, where A makes no division), or that the scene-headings occupied only one line each (Cf. iir.. 2; "The object of the verb has commonly been taken as arr/entum, but there is no reason why mulierem should not be supplied from 1. 438. It is the woman who is Toxilus' great concern at this point (Cf. 11. 438, 445, 447.) The use of concredere with a person as the direct object is supported by Capt. 348 (filium) ; Cist. 245 (ami- cam); and manus in the technical sense of the power of a man over the women of his household is too well known to need comment. a4 EETRACTATIO IN . PLAUTUS m., 3 in A.^' In any case, A has 11. 442-443, which are the most suspicious verses. TO. Sagarlstio, heus, exi atque educe ulrginem Et istds tabellas quSs consignauf tibi, Quas tu attulisti mi Sb ero meo usque e Persia. 460 Jf60-Jf61, This couplet seems to be a later insertion, introduced to prepare the audience for the sudden appear- ance of the letter in 1. 497. No mention has been made of this letter in the previous plans of Toxilus,^^ and when the trick is finally played on Dordalus, it is not Sagaristio, but Toxilus himself, who hands over the tablet (1. 497). The two lines are an awkward addition after 1. 459, and the dependence of tabellas upon educere is unparalleled.^*^ A deest. TO. Age illfic apscede pr6cul e conspectG: [SA.] taee Vbi ctim lenone m6 uidebis c6nloqui. Id erit adeundi tempus. nunc ag4rite uos. 467 J,67-J^68. These verses are identical with 727-728, and must therefore be regarded with suspicion in one of the two places. Some connecting link is necessary be- tween IV., 2 and IV., 3, and we must either retain 467- " The latter explanation is suggested by Studemund. See his note on fol. 57 5v. " A discussion of all the details of the plan was of course not necessary. The trick is perfectly clear without a previous mention of the letter, as is the similar trick in Bacch. TV., 9, where 1. 941 is the only preparation for the presentation of the letter in 11. 997 flF. *"Langen, Plant. Stud., p. 178. PERSA 35 468 or assume that these verses have crowded out other directions to Sagaristio. A deest. DO. Itibedum ea hoc acc^dat ad me. SA. I sfine ac morem illf gere 605 P^rcontare, exqufre quiduis. TO. Age, age nunc tu: in pro6lium Vfde ut ingrediare atispicato. VI. Liquidumst auspicitim: tace. Ctirabo, ut praedSti pulcre ad cfistra conuortdmini. TO. Concede istuc: 6go illam adducam. DO. Age, Hi rem ease in nostrum putas. TO. 6hodum hue, uirgo. ulde sis quid agas. VI. TSceas: curabo Ut uoles. 610 605-610. The reader who tries to imagine the stage "business" of this part of the play finds himself in difficulty. Dordalus has expressed a wish to question the girl, and the pseudo-Persian Sagaristio has given his permission (11. 605-606). Then Toxilus speaks 11. 606 f. (Age, age nunc tu: in proelium vide ut ingrediare auspi- cate) to the girl, and she replies in a low tone. But in the very next line Toxilus is saying to Dordalus, Concede istuc; ego illam adducam. His next speech, an aside to the girl, and her reply (609-610) repeat the sense of 11. 606-608, and have in several places identical phrasing (607 vide, cf. 610 vide; 607 tace, cf. 610 taceas; 608 curabo, cf. 610 curabo). Clearly we have here two alternative versions (605-608; 609-610), of which the first is the cleverer and therefore probably the genuine. ^^ • A has 605-610 in the order in which they appear in the Goetz- Schoell text. P. omits 608, 610 {i. e., one line of each version). ^Cf. J. H. Gray in Class. Rev. 14 (1900), p. 24. 36 EETRACTATIO IN PLAUTUS DO. T6xile, quid ag6? TO. Bi deaeque te Sgitant iratf. [et] scelus, 666 Qui hac non properes d^stinare. DO. tHabeto. TO. Eu. prae- datu's probe: abi, argentum ecfer hue. N6n edepol minis trecentis cirast: fecisti lucri. SA. Hetis tu, etiam pro u6stimentis hflc decern acoed^nt minao. DO. Abscedent enim, non accedent. TO. Tfice sis: non tu illiim uides 670 Qua^rere ansam, inf6ctum ut faciat? abisne atque argenttim petis? ♦ ***♦*** atque ut dignflst perit. 67 P DO. Heus tu, serua istfim. TO. Quin tu intro is? DO. Abeo atque argentum adfero. 666 ff. A shortening of the scene seems to be indicated by the half -line abi, argentum ecfer hue, which is appended to 1. 667 in P. Apparently this phrase was substituted for eu, praedatus probe (1. 667)/^ with the result that the Palatine manuscripts have retained both endings, while in A the line ending probe appears just before 1. 669. There are further signs of change at the very end of the scene. The fragment of a line atque ut dignust perit (671^), which is preserved only in A, seems to belong to a bit of comment after the Ze/io's departure, and is hardly suitable before 1. 672.-^ To the first half of this fragmentary line may belong the adverb interibi, cited by the Glossarium Plautinum from some point in the Persa between 1. 588 and 1. 677, but not to be found in our present text.^* It is quite possible, therefore, that we have here traces of some such short- ening as this: "Cf. Leo, Ed. (1896), ad loc. =«Cf. Leo, Ed. (1896). ad loc. 24 The strictness with which the author of the glossary follows the order of lines in the text of Plautus makes it improbable that interibi in this position refers to 1. 165 of the play. Cf. Ritschl. Opusc. 11., pp. 266 ff.; Schoell. Ed. (1892), Praef.. p. xx. PEKSA 37 667. DO. Habeto. TO. Abi, argentum ecfer hue. 67P (Schoell's supplement) DO. Interibi opperire. SA. Aegre avidus abit atque ut dignust perit. A is very illegible at this point. The end of 665 is preserved, and then follows a space of three lines, which may have contained 666 and 667, the latter divided so as to take up two lines. After this it has a line ending probe, and then the ends of 669-671, 67P, 672. So A apparently omitted 668, and added 67 1^ which does not occur in P. P has 667 in the form - - - habeto, eu praedatus probe, abi argentum ecfer hue, and omits 671^. Vanlloquidorus Virginisuendonides t Nugiepiloquides Argentumextenebr6nides Tedigniloquides NQmmosexpalponides Quodsemelarripides Ntimquameripides : €m tibi. 702 705 70Jf. The absurd patronymics of 1. 704, which mean either : " Talk-to-you-as-you-deserve-son, N^onsense-son, Flatter-son," or " Talk-to-you-as-you-deserve-son, Flatter- money-out-of-you-son," according as we follow the reading of A or P, repeat the sense of the preceding line : " Talk- nonsense-to-you-son, Bore-your-money-out-of-you-son." ^^ The line may have been substituted to expand the joke. The line appears in both A and P, though A reads nugidespalpo- nides, and P nundesexpalponides. DO. Immo equidem gratiam 719 Tibi, T6xile, habeo: n^m te sensi s^dulo Mihi dare bonam operam. TO. Tlbine ego? immo t sedulo. DO. Attfit, oblitus sum fntus dudum edicere Quae u6lui edicta. ads^rua banc. TO. Saluast ha6c quidem. VI. Pat^r nunc cessat. TO. Quid si admoneam? VI. T6mpus est. =5 Cf. Leo, Ed. (1896), ad loc. 38 BETEACTATIO IN PLAUTUS TO. Heus, Saturio, exi. nfinc est ilia occ&sio 725 Inimlcum ulcisci. SA. Ecce me: numquld moror? TO. Age illGc abscede pr6cul e conspectti: tace. Vbi cflm lenone m6 uidebis c6nloqui, Turn tflrbam facito. SA. Dictum sapienti sat est. TO. Tunc, quando abiero— Quln taces? scio quid uelis. 730 DORDAIiVS. TOXIIiVS. VIRGO. LENO SERVOS DO. Transcldi loris 6miiis adueni^ns domi: Ita mlhi supellex squSlet atque aed^s meae. TO. Redls tu tandem? DO. R^deo. TO. Ne ego hodi6 tibi Bona mdlta feci. DO. Ffiteor: habeo grdtiam. 722 if. Much has been written about the defects of the play at this point. ^^ It is strange that Dordalus should go off the stage immediately after he has purchased the supposed Persian girl, and should leave her in the care of Toxilus, instead of taking her inside the house with him; even more strange that he should reappear six lines later, announcing (1. 731), Transcidi loris omnis adveniens domi. In the meantime, Saturio has emerged from his hiding-place and has been given some brief and quite inapposite directions about the line of action to follow when the leno returns. Two verses of this dialogue (727-728) repeat a couplet earlier in the play (467-468). Of the remaining lines, several show the stock phrases of comedy (726 numquid moror, cf. 1. 462; 729 dictum sapienti sat est, cf. Ter. Phorm, 541; 734 bona multa feci, cf. 1. 263). Immediately after the return of Dordalus (1. 734) the conversation goes back to the subject under discussion before he left (1. 721) ; and the dozen lines intervening seem to have no purpose ^Qi. especially Goetz in Acta soc. phil. Lips. 6 (1876), pp. 300 f.; Wilamowitz in Index schol. Gott. 1893-1894, p. 21; Meyer, De Plauti Persa, pp. 172 ff. PEKSA 39 but to prepare for the entrance of Saturio at 1. 738.^"^ But the plan had been that Saturio should appear sud- denly (cf. 11. 162-164), and his opening lines suit a first appearance. Therefore 11. 722-734 are probably a later addition. A has the whole passage. P. omits 1. 730. SA. Nisi ^go illun hominem pdrdo, perii. atque optume 738 Eecum fpsum ante aedes. VI. Sfilue multum, ml pater. SA. Salu6, mea gnata. DO. Ei, P€rsa me pesstim dedit. 740 VI. Pater hlc mens est. DO. Hem, quid? pater? perii 6ppido. Quid ego igitur cesso inf^lix lament^rier MinSs sexaginta? SA. 6go pol te faciilm, scelus, Te quoque etiara ipsum ut Idmenteris. DO. 6ccidi. SA. Age ambula in ius, 16no. DO. Quid me in itis uocas? 745 SA. Illi[cJ Spud praetorem dicam: sed ego in i(is uoco. DO. Xonne dntestaris? SA. Tudn ego causa, cfiruufex, Quoiqudm mortali llbero auris dtteram, Qui hie commercaris cfuis homines llberos? DO. Sine dIcam. SA. Nolo. DO. Atidi. SA. Surdus sum: fimbula. 750 Sequere hSc, sceleste, f6les uirginSria. Sequere hdc, mea gnata, me tisque ad praetorem. VI. Sequor. 7 88 ff. Ritschl ^^ observed that the movement of the play after the close of IV., 8 was very hurried, and other editors have agreed with him. In particular, IV., 9 seems to have suffered from shortening. The discovery of the pseudo-Persian girl's identity is made far too quickly (11. 739-740), and the question of Dordalus: Hem^ quid? pater? (1. 741) is absurd after 11. 739-740. Satu- rio, his daughter, and Dordalus, are suddenly dismissed from the stage, and we hear nothing of what takes place between them in the forum. Then, too, we are puzzled " Meyer, De Plauti Persa, p. 174. =« Ritschl, Ed. (1853), Praef., p. ix. 40 KETKACTATIO IN PLAUTUS PEESA 41 by the fact that the parasite, whose sole object in under- taking the deceit of the leno has been to provide himself with a dinner (11. 140-147; 329 ff.). does not appear at the banquet in Act V., while Dordalus, who left to defend himself before the praetor (1. 752), re-enters at 1. 778. The difficulties are explained if we suppose that IV., 9 is the shortened form of a scene in which the leno plead for mercy and finally, by the offer of a sum of money, induced Saturio to drop his legal proceedings, take the money and his daughter, and go home.^^ A and P have 738-741 in the same form, end of the play. Then A deest to the The Persa offers comparatively slight evidence for the changes of the Plautine Kevival. There are a few in- stances of parallel versions— 11. 433-486, 442-443; 605- 608, 609-610 ; 703, 704. An attempt to bring the scene to a close seems to be indicated by the confusion in the manuscripts after 1. 666. The play as a whole is notice- ablv short, and the dialogue in two of the scenes (11. 440-448; 738-752) is so hurried and disconnected that we are justified in assuming a shortening in which the original version was lost. On the other hand, there are a few places (11. 722-734, and possibly 460-461), where an insertion seems to have been made in order to prepare for a later scene.^® 2» Meyer, De Plauti Persa, pp. 177 flf. ^In addition to the passages discussed in detail, the following lines have been suspected: 60, 240, 280-295, 453-454, 562, 673-682, 694-699, 833-851. Of these lines, A is missing for 60, 240. 453-454, 833-851. Both A and P have 280-295 (except that P transposes 293, 294), 562, 673-682, 694-699. The following lines show minor variations: 399, 485, 498, 500, 515-516, 574, 597. N The indications of change are not confined to the Pala- tine manuscripts. A has the beginnings of the two short- ened scenes (11. 440 ff. ; 738 ff.) in the same form as P, and probably contained the whole. On the other hand, there are three passages in which A and P alike show traces of retractatio, but A has preserved a larger number of the suspicious lines. Far from showing the purer text, therefore, A gives all the later versions that P gives, and has some of them in fuller form. POENULUS 43 CHAPTEE 11. POENULUS The Poenulus tells the story of two sisters in the service of the leno Ljcus, who gain their freedom through Agorastocles, the lover of the elder girl, and his slave Mil- phio. In the first scene, Milphio evolves a scheme which he promises will give Agorastocles not only his sweetheart, but the whole household of the leno as well. The vilicus of Agorastocles, who happens to be in the city, and who is a stranger to Lycus, is to be dressed up as a foreign soldier and sent to the leno with a request for an evening's pleasure. Then Agorastocles is to appear and demand his slave, and, upon the leno's denial of all knowledge of the slave, Agorastocles is to drag him off to court. This plan is carried out with the help of advocati from the forum, who introduce the newcomer to Lycus and witness the whole proceeding. But Milphio, the author of the scheme, disappears from the action, and the pro- posed law-suit never takes place. Instead, Milphio enters at the beginning of Act IV., raging against the leno and threatening his destruction, as if Acts I.-III. had no existence. He learns from the slave of Lycus that the sweetheart of Agorastocles and her sister are really free- born Carthaginians, and, knowing that his young master is also Carthaginian by birth, he confidently plans their release. Just at this moment the Carthaginian Hanno appears, on a search for his two daughters and his nephew, all of whom were stolen away as children. The nephew is discovered to be Agorastocles, the adopted son of Han- 42 f, no's old guest-friend. Milphio then proposes that Hanno aid in outwitting the leno and obtaining the freedom of the two girls by pretending that they are his daughters. The fiction proves to be only too true, and the happy father promises the hand of his elder daughter to her lover. When the leno returns, desperate over the ruin that has already come upon him, he finds three-fold vengeance waiting at the hands of Hanno, Agorastocles, and the soldier to whom he had promised the younger girl. Even this hasty sketch is sufficient to show the break between Act III. and Act IV., and the repetition in the two plots against the leno, A more detailed study of the play brings out other difficulties. The first act pre- sents the two girls as meretrices of the ordinary sort, who are already familiar with their calling (cf. especially 11. 233-23G; 265-270); in the last act, they are repre- sented as entering upon their profession on the very day that the play opens (11. 1139-1140), and their conversation is full of sentiments befitting their noble birth (cf. 11. 1185-1186; 1201-1204). In the first half of the play, too, they have no prospect of freedom aside from the lover of the elder sister (11. 360-363) ; in the second, the prophecy of good fortune immediately makes them hope for assistance from their parents (1. 1208). We note, too, that, though the prologue gives the scene of the play as Calydon (11. 72, 94; cf. 1057, 1181), Milphio pro- mises Adelphasium that she shall become civis Attica. (1. 372). It is almost certain that the Poenulus combines two different plots, one covering roughly the first three acts of the play, and the other the last two. The scene of the first was laid in Athens; that of the second (the •*«•» 44 KETRACTATIO IN PLAUTUS POENULUS 45 Kapxv^ovio^, of which the plot is given in the prologue) in Calydon. The two originals seems to have had as common elements two sisters in the service of a leno, and a festival of Venus, at which the sisters offered acceptable sacrifices, but the offerings of the leno were rejected. The two plots are necessarily bound together to some extent, especially at the beginning and the end of the play, but their general outlines can still be recog- nized. The first presents two Athenian meretrices, with the elder of whom a youth is in love. To obtain her freedom, he and his slave play a trick which results in a law-suit and the leno's ruin. In the Kapx^3oVto9, on the other hand, the girls are Carthaginians of noble birth, who are to enter upon their calling on the day that the play begins. Their release comes about through their father, who arrives in Calydon at the proper mo- ment, finds his daughters, and discovers his nephew in the lover of the older girl. The line of division between the two plots in the Poenulus is so clear that it attracted attention compara- tively early. ^ Teuffel,^ noticing the distinctness of the two plots against the leno, and the contradiction about the scene of the play, decided: '' Beim Poenulus lage die Annahme einer Contamination ziemlich nahe, wenn dadurch etwas gewonnen ware." But Teuffel's hint was not followed up for fully twenty years. Then Keinhardt ^ and "Francken ^ studied the composition of the play with ^ G. Langrehr, De Plduti Poenulo, Friedland, 1883, p. 14, says that Rapp was the first to suspect contaminatio in the Poenulus. = W. Teuflfel in Rhein. Mus. 8 (1853), pp. 35 ff. { = 8tud. u. Char"., pp. 337 ff.) *L. Reinhardt in Studemund's Studien auf d. Gehiete d. archai- schen Lateins, Vol. i. (Berlin, 1873), pp. 109 ff. *C. M. Francken in Mnem. 4 (1876), pp. 146-175. great care. The analyses which they made have been modified and corrected in detail by Langen,^ Leo,^ and Karsten,"^ but the main lines of their division still remain unchanged. Earum hlc adulescens Alteram efflictim perit Suam sibi cognatam inprtidens, neque sit qua6 sit, Neque earn timquam tetigit: ita eum leno mdcerat: (Neque qufcquam cum ea f^cit etiamnfim stupri, Neque dtixit umquam : n^que ille uoluit mlttere : ) Quia amdre cernit, tdngere hominem u6lt bolo. 96 100 99-100. This couplet, which gives the substance of 1. 98 in a little fuller form, is probably to be set down to retractatio. A deest 1-281. (Ehem, pa^ne oblitus sfim relicuom dlcere. Ille qui adoptauit htinc pro filio sibi Is lUi Poeno, huitis patri |, hosp6s fuit.) 118 120 118-120, These lines may have been added to explain a little more fully the situation indicated in 1. 75 emit Jiospitalem is filium inprudens senex. The passage is especially disturbing because it breaks the connection between the subject of the next sentence (1. 121 is, or 1. 124 hie — cf. below) and its antecedent, unquestionably the Carthaginian who has been under discussion in 11. 104-115. A deest. ^Langen, Plant. Stud., pp. 181 ff. "Leo. Plant. Forsch.. pp. 153 ff. 'H. J. Karsten in Mnem. 29 (1901), pp. 363-387. 46 KETRACTATIO IN PLAUTUS POENULUS 47 Is h6die hue ueniet r^p[p]erietque hie fllias 121 Et htinc sui fratris fllium, utquidem didici ego. Ego Ibo, ornabor: u69 £iequo animo noscite. (Hie qui h6die ueniet. r§p[p]eriet suas fllias Et hane sui fratris fflium. dehinc c6terum 125 Quod r6stat, restant aii qui faeiSnt palam. Val^te: adeste. ibo: alius nune fieri uolo.) Val6te atque adiuuSte ut uos seru6t Salus. 121-128. The close of the prologue undoubtedly con- tains two versions. There is almost exact verbal repe- tition in 11. 121-122 and 124-125; two announcements are made of the speaker's proposed change of dress (123; 127) ; and the farewell to the audience is given in two different forms (127; 128). Seyffert « is probably right in arranging the two versions: (1) 11. 121-123, 128 ; (2) 11. 124-127. Of the two, the second is probably the genuine. Alius nunc fieri volo (1. 127) is more vivid than orrmhor (1. 123), and the collocation restat, restant (1. 126) is quite in the style of Plautus. P has the verses in the order of the Goetz-Schoell text, except that 1. 126 appears after 1. 127. A deest. A deest. AD. Xegoti sibi qui uol^t uim parSre, Nauem et mulierem haec duo comparato. Nam ntillae magls res dua6 plus negOti Hab^nt, forte si 6eceperls exornfire, Neque umquam satis hae duae res ornantur, Xeque els uUa orn^ndi satis satietfis est. 210 215 2H. The thought of 11. 214, 215 is exactly the same. The phrase duae res in 1. 214 is an awkward repetition from 1. 212, and the line is unmetrical. In line 215, on the contrary, the quibble satis satietas sounds Plautine. The first line is therefore probably due to a later hand. •0. Seyffert, Studia Plautina, Berlin, 1874, p. 11. Atque ha^ ut loqu6r, nunc dom6 docta dico. Nam n6s usque ab afirora ad h6c quod di^ist (Postquam aurora inluxit numquam ooncessauimus ) Ex Industria fimbae numquSm concessdmus Laufiri aut fricfiri aut terg^ri aut orndri. 216 220 218, Another variant, also unskillful metrically, seems to be presented by 1. 218. This line repeats 11. 217, 219, and was probably intended to take their place. A deest. AD. fnuidia in me nflmquam innatast n^ue malitia, m4a soror : 300 Bono me esse ing^nio ornatam quam afiro multo mduolo. (Atirum id fortuna Inuenitur, ndtura ingenitim bonum: B6nam ego quam bedtam me esse nimio dici mduolo.) M^retricem pudorem gerere mdgis decet quam ptirpuram. (Mdgisque meretric6m pudorem quam atirum gerere c6ndecet.) 305 Pfilcrum ornatum ttirpes mores p^ius caeno c6nlinunt: L^pidi mores ttirpem ornatum fdcile factis c6mprobant. 300 ff. Fond as Plautus was of sententiae, he would hardly have made Adelphasium utter the whole of this speech as it stands in our text. N'early every line in it has been suspected by some one of the editors. We can, however, be sure of retractatio only in 1. 304, which repeats 1. 305 almost word for word, omitting the neces- sary conjunction -que^ and substituting decet for the Plau- tine verb condecet. The second version seems to have been composed to introduce a new detail {purpura) into the list of the courtesan's ornaments.^ The whole passage occurs in both A and P, but A has the order: 303, 305, 304. •Langen, Plant. Stud., pp. 338 f. 48 KETRACTATIO IN PLAUTrS POENULUS 49 Sfc enim dicer^s, sceleste: hulus uoluptas, te 6psecro, 387 Hulus mel, huius cor, hulus labellum, hulus lingua, huius sauium, Hulus delicia,, huifis salus amo^na, huius festluitas, Hufus colustra, huifls dulciculus c^seus, mastlgia: 390» (Htiius cor, huitis studium, huius sauium, mastlgia.) 390^ ^90^ The terms of endearment in 1. 390,* except for studium, merely repeat those of 1. 388, and the epithet mastigia addressed to Milphio is taken from 1. 390. It is possible that the word studium was a new bit of slang in the Kevival Age, and that the alternative line was composed for the purpose of introducing this novelty. P has the order: 389, 390*, 390^ A omits 390* (a genuine verse), but writes dulciculus caseus above savium mastigia of 390^ showing that some form of the line must have stood in the archetype of A. AG. fta me di ament, tSrdo amico nll[i] est quicquam ina€- 504 505 510 quius, Pra6sertim homini amSnti qui quicquid agit properat omnia. Slcut ego hos duco aduocatos, h6mines spissigradlssumos, Tardiores quSm corbitae stint in tranquill6 mari. Atque equidem hercle d^dita opera amlcos fugitaiil senes: Sci[e]bam aetati tfirdiores. m^tui meo amorl moram. N^uiquam hos procos mihi elegi loripedis, tardissumos. Quin si ituri hodie 6stis, ite aut Ite hinc in malSm crucem. Slcine oportet ire amicos h6mini amanti operam datum? Nam fstequidem gradtis succretust crlbro pollinfirio: Nisi cum pedicis c6ndidicistis Isoc grassarl gradu. ADV. Hefls tu, quamquam nos uidemur tibi plebeii et pati- peres, 515 SI nee recte dlcis nobis dlues de summo loco, Dluitem audact^r solemus m^ctare inforttinio. N^c tibi nos obnlxi[i] sumus istCic, quid tu ames aut oderis. Quom argentum pro c§pite dedimus, nostrum dedimus, non tuom. Llberos nos 4sse oportet: nos te nili p^ndimus: 520 N€ tuo nos am6ri seruos [tuos] #sse addictos c^nseas. Llberos homines per urbem m6dico magis par 6st gradu fre: seruile ^sse duco f^tinantem ctirrere. k Pra^sertira in re populi placida atque Interfectis hostibus N6n decet tumultuari. 86d si properab^s magis, 525 Prldie nos te dduocatos hGc duxisse op6rtuit. N6 tu opinere, hatid quisquam hodie nostrum curret p6r uias N6que nos populus pro cerritis Insectabit iSpidibus. AG. At si ad prandifim me in aedem uos dixissem ducere, Vfnceretis c^ruom cursu u^l grallator^m gradu. 530 Ndnc uos quia mihi fiduocatos dixi et testis dCicere, P6dagrosi estis kc uicistis cocleam tarditCidine. An uero non idsta causast qu6 curratur c6leriter, [ADV.] Vbi bibas, edfis de alieno quantum uelis usque adfatim, Qu6d tu inuitus nGmquam reddas domino, de qu<(o>io Meris? 535 S6d tamen cum e6 cum quiqui quamquam sumus paup^rculi, fist domi quod edimus: ne nos tdm contemptim conteras. Quicquid est pauxlllulum illuc nostrum t id omne intus est: N^ue nos quemquam fldgitamus n§que nos quisquam fldgitat. Tud causa nemo nostrorumst suos rupturus ramites. 540 AG. Nlmis iracundi estis: equidem haec uobis dixi per iocum. ADV^. P6r iocum itidem dicta [m] habeto, qua^ nos tibi respon- dimus. AG. 6bsecro hercle operdm celocem banc mihi, ne corbitdm date. Attrepidate sAltem: nam uos adproperare baud postulo. ADV. Slquid tu placide 6tioseque Sgere uis, operam damns: 545 SI properas, cursores meliust te dduocatos dticere. AG. Scltis, rem narrdui uobis, quod uostra opera, mi 6pus siet, D6 lenone hoc qui me amantem Ifidificatur tam diu: El paratae ut sint insidiae de afiro et de seruo meo. ADV. 6mnia istaec scimus iam nos, si hi epectatores sciant. 550 Horunc hi[n]c nunc caflsa haec agitur sp^ctatorum fibula: H6s te satius 6st docere ut, qudndo agas, quid aga[n]s sciant. N69 tu ne curfissis: scimus rem omnem. quippe omn^s simul Dldicimus tecum fma, ut respond^re possimfis tibi. AG. fta profectost. ^d agite igitur, flt sciam uos scire, rem 555 Expedite et mihi quae uobis dtldum dixi dlcite. ADV. ftane temptas dn sciamus? non meminisse nos [tjratu's, Quo modo trec^ntos Philippos Collabisco ullico DMeris, quos def^rret hue ad lenonem inimicuni tuom, fsque se ut adsimuldret peregrinum [esse] aliunde ex alio oppido. 560 "Vbi is detulerit, tu eo quaesitum s^ruom aduent6s tuom Q(im pecunid. AG. Meministis m^moriter: seruSstis me. 50 KETRACTATIO IN PLAUTUS POENULIJS 51 ADV. file negabit: Mllphionem qua^ri censeblt tuom fd duplicabit 6mne furtum: l6no addicettir tibi. Ad earn rem nos 6sse estis uls tibi. Ten^tis rem. 565 ADV. VLx quidem hercle | — Ita pauxillast — digitulis primoribus. AG. (Hoe cito et curslmst agendum. pr6pera iam quantflm potest. ADV. B4ne uale igitur. te aduocatos m6lius celeris dficere: Tdrdi sumus nos. AG. 6ptume itis, p^ssume hercle dicitis. Quln etiam declderint nobis f^mina | in talCs uelim. 570 ADV. At edepol nos tibi ( in lumbos llnguam atque oculos In solum. AG. H§ia, hau uostrumst Iracundos 6sse quod dixl ioco. ADV. N^e tuom quid^mst amicis p6r iocum iniust^ loqui. AG. Mtttite isUec. quid uelim uos, scltis. ADV. Callemfis probe : L#nonem ut peritirum perdas, Id studes. AG. Ten^tis rem.) 575 5Jf.O if. The long tedious scene between Agorastocles and the advocati was apparently shortened for later pro- ductions. The beginning of the scene (11. 504-542) and 11. 567-573 show exactly the same development of thought. In both, Agorastocles rebukes the old gentlemen for their slowness, and they resent the reproof ; then he apologizes, saying that his words were meant only in fun. The plan against the leno which is reviewed at length in 11. 547-566 is summarized in 11. 574-575, though, as Goetz ^« observed, the second version would be incomprehensible if we had not the first as well. Moreover, a third version of the beginning of the scene is probably preserved in 11. 543- 546.^^ Here again we have the remonstrance of Agor- *• Goetz in Acta soc. phil. Lips. 6 (1876), p. 269. "Goetz {Acta soc. phil. Lips. 6, p. 254), noting the inappropriate- ness of these verses in their present position, made them precede 11. 541-542. Leo [Plant. Forsch., p. 161, note) thought them a part of the same shortened version that we find in 11. 567-575. Kellermann {Comm. phil. Jen. 7, p. 134) agreed with (5oetz that 11. 543-546 and 567-575 could not belong to the same recension. astocles and the reply of the advocati (cf. 11. 504-540; 567-571). The parallelism is particularly close between 507 (corhitae) and 543 (corhitam) ; between 521-523, 546, and 568; between 541-542 and 572-573; between 565 (tenetis rem) and 575 (tenetis rem). There seem therefore to have been three versions of the scene: (1) the Plautine version, 11. 504-542; 547-566; (2) 11. 567- 575; and (3) 11. 543-546, probably followed by 11. 547- 566.21 P has the passage in the order of the Goetz-Schoell text. A deest 501-571 (=71 11.) Two sheets of A are missing (=76 11.). If we allow two lines for the scene-heading of III., 1, and assume that a few of the long verses were run over, we find that the whole passage could very well have been contained in A. But in any case, the fact that A preserves 572 ff. in the same form as P, would argue that it had the rest of the passage as well. ADV. Aet6li ciues t6 salutamfis, Lyce: Quamquam hSnc salutem f^rimus inuitf tibi. [Et quamquam bene uolumus leniter lenonibus.] LY. Fortdnati omnes sitis: quod cert6 scio Nee f6re nee Fortunam Id situram fieri. ADV. Istic 6st thensaurus stflltis in linguS situs, Vt qua^stui habeant mdle loqui meli6ribus. ViSm qui nescit qua deueniat fid mare, Eum op6rtet amnem qua^rere comit^m sibi. Ego mfile loquendi u6bis nesciul uiam: Nunc u6s mihi amnes 6stis: uos certtimst sequi. Si b^ne dicetis, u6stra ripa u6s sequar: Si mfile dioetis, u6stro gradiar Ifmite. 621 622t» 625 630 " Langrehr, De Plauti Poenulo, p. 19, suspected 11. 523, 567, 728, 733 (to which he should probably have added 1. 730) because the advocati, for whom Plautus regularly uses the plural, speak or are addressed in the singular in these lines. Of the suspected verses, 1. 567 is probably not by Plautus; the rest occur in passages which are otherwise free from suspicion. 52 EETRACTATIO IN PLAUTUS ADV. MalO bene facere tantumdemst perlculum Quantlim bono male facere. LY. Qui[d] uero? ADV. Scies. Malo slquid bene facifis, id beneficium Interit: 635 Bon6 siqiiid male facias, aetatem 6xpetit. Fac^te dictum. s4d quid istuc ad me Sttinet? Quia nos honoris tuf causa ad te u^nimus, Quamquam bene uolumus Igniter lenonibus. 622''. This verse is the same as 1. 639, except that 1. 622'' begins et quamquam, and 1. 639 quamquam alone. Lindsay ^^ is probably right in thinking that the repe- tition indicates shortening. The actors of the Revival could easily pass from 1. 622 to 1. 639, and thence to the end of the scene. A deest 604-634. LY. Sed Mec latrocinSntur quae ego dixi omnia. 704 CO. Quid ita? LY. Quia aurum p6scunt praesentfirium. 705 CX). Quin hercle accipere tti non mauis quam €go dare. Quid, si 6uocemus hflc foras Agordstoclem, Vt Ipsus testis sit sibi certlssumus? Heus ta, qui furem cSptas. egredere 6cius, Vt tlite inspectes aQrum lenonf dare. 710 AGORASTOCLES. ADVLESCENS ADVOCATL LYCVS LENO COLLABISCVS. VILICVS AG. Quid ^st? quid uoltis, testes? ADV. Specta ad d^xteram: Tuos s^ruos aurum | Ipsi lenonf dabit. CO. Age Recipe hoc sis: helc sunt numerati atirei Trec^nti nummi qui uocantur Phflippi. Hinc m6 procura: pr6pere hosce apsuml uolo. 715 LY. Edep6l fecisti prodigum promtim tibi. Age eamus intro. CO. T6 sequor. LY. Age age ^mbula: Ibl quae relicua ilia fabuliibimur. POENULUS 53 CO. Ead^m narrabo tfbi res Spartiaticas. LY. Quin s6quere me intro. CO. D(ic me intro: addictdm tenea. 720 AG. Quid ntinc mihi auctores 6stis? ADV. Vt frugf sies. AG. Quid, si Animus esse non sinit? ADV. Esto tit sinit. AG. Vidfstis, leno quom atirum amus. AG. Rem aduorsus populi t saepe leges? ADV. Sciuimus. 725 AG. Em ista^c uolo ergo uos commeminisse omnia. Mox quom Sd praetorem | tisus ueniet. ADV. M6minimus. AG. Quid, sf recenti re a^dis pultem? ADV. C^nseo. AG. Si pflltem, non recMdet? ADV. Panem frangito. AG. Si exferit leno, c6nsen[t] hominem int^rrogem, 730 Meus s^ruos [si] ad eum u^neritne? ADV. Qufppini? 706 ff. The repetitions in the manuscripts at the close of III., 3 seem to indicate a shortening of the prolix scene that follows. After Collabiscus had agreed to pay Lycus the sum demanded for an evening's entertainment, there was no reason for delay ; the money could be handed over at once, and then Agorastocles could appear and demand his slave, in accordance with the prearranged plan. It was apparently with the intention of cutting out the intervening dialogue that some manager adscribed 11. 720, 730 at the close of III., 3. But if, as Leo ^* and Lindsay ^^ think, 1. 720 followed directly upon 1. 706, and 1. 730 in turn upon 1. 720, Agorastocles must have appeared upon the stage without any summons or any notice of his coming. Some announcement seems neces- sary, and the fact that 1. 720 appears in A one line later than in P perhaps indicates that it was intended to be spoken after the summons of x\gorastocles. (IL 707-710). Id Lindsay, Ed. (1905), ad loc. " Leo, Plant. Forsch., pp. 7 f . Lindsay, Anc. Edd., pp. 43 f. p IS 54 EETRACTATIO IN PLAUTUS POENULUS 55 P has the order: 706, 720 (in the unmetrical form Quin sequere me introf Due me ergo intro. Addictum tenes.) , 707 ff., repeating 720 in metrical form in the proper place. A has the order: 706, 707, 720 (in the same form as P, except that it omits addictum tenes), 730, (with the readmg gmd tumf hominem interrogemf) , 708. Then the sheet which must have contained 709-745 is missing in A. The 38 lines on the sheet would not give room for these 37 verses and a two-line scene-heading; but it is possible that the scene-heading for III., 4 took up only one line in A,- as the heading for III., 5 actually does. At any rate A, in inserting 730 before 708, shows more evidence of change than P. CO. Absc6dam hinc intro. AG. Factum a nobis cdmiter. 805 Bonam dedistis, aduocati, operSm mihi. Cras mane, quaeso, in c6mitio estote 6buiam. Tu s6quere me intro. 1163 ualete. ADV. Et tti uale. 805-808, The command of Agorastocles to CoUabiscus in 1. 808 {tu sequere me intro) is impossible after Col- labiscus bas announced in 1. 805 ahscedam hinc intro, Langen^'^ is probably right in thinking that a later version substituted 1. 805 for 11. 806-808, in order to shorten the scene. He observes that the vocative advocati comes in much more naturally if 1. 805 is omitted, and that the salutation valete . . . vale (1. 808) forms a suitable close for the dialogue. A deest 783-849. MI. Illic hinc abiit. dei Immortales m€um erum seruatGm uolunt 917 fit hunc disperdittim lenonem: tSntum eum instat 6xiti. Satine, prius quam untimst iniectum t6lum, iam instat dlterum? fbo intro- haec ut m6o ero memorem: nam htic si ante aedes guocem, 920 ^•This is evidently the assumption of Seyffert. Woch. 16 (1896), col. 253, note 3. Langen, Plaut. Stud., p. 342. Cf. Berl. Phil. Quae afidiuistis modo, nunc si eadem hic iterum iterem, inscitiast. firo uni potius fntus ero odio quam hfc sim nobis 6mnibus. Dei Immortales, qudnta turba, quanta aduenit calamitas H6die ad hunc len6nem. sed ego nunc est quom me commoror. fta negotium fnstitutumst : non datur cessatio: 925 Nam et hoc docte consulendum, quod modo concr4ditumst, fit illud autem ins6ruiendumst c6nsilium uernaculum. R6mor si sit, qui malam rem mlhi det, merito f^cerit. Isunc intro ibo: dum erus adueniat a foro, opperiar domi. 917-929. The close of the scene undoubtedly shows two parallel versions. There is very close correspondence between 917-918 and 923-924 (note especially the repe- tition of dei immortales) ; between 920 iho intro and 929 nunc intro iho. But 1. 920 implies that Agorastocles is in the house, while 1. 929 states that he is to return a foro. We must therefore recognize two versions: (1) 11. 917-922; and (2) 11. 923-929. Evidence for the genuineness of the former version is found in the jingle ero . . .ero in the last line ; ^^ and this version agrees with 1. 808 (which, as we have just seen, is probably genuine) in representing Agorastocles as inside the house. Both A and P give the entire passage. Ythalonim ualon nth sicorathisyma comsyth Chym lachunythmumys thral mycthi baruimy sehi Liphocanethythby nuthi ad edynbynui Bymarob syllohomaloni murbymy syrthoho Byth lym mothyn noctothu ulechanti clamas chon Yssid dobrim thyfel yth chil ys chon chem liful Yth binim ysdybur thinnochot nu agorastocles Ythemaneth ihychir saelichot sith naso Bynny idchil liichilygubulim lasibit thim Bodialytherayn nynnurs lymmon choth lusim 930 935 « "Lindsay, Anc. Edd., p. 55. 56 EETRACTATIO IN PLAUTUS 945 Ythalonimualoniuthsicorathiisthymhimihymaooinsyth Combaepumamitalmetlotiambeat lulecantheconaalonimbaluinbar ff dechor Bats* * * *hunesobinesubicsillimbalim Esseantidamossonalemuedubertefet Donobun*huneccilthumucommucroluful Altaniinauosduberithemhuarcharistolem Sittesedanecnasotersahelicot Alemusdubertimurmucopsuistiti Aoccaaneclictorbodesiussilimlimmimcolus. 930-9Jf9. A has only one version of the Carthaginian passage; P gives two. Lindsay ^^ regards 11. 930-939 as retradatio; but it is hard to imagine any stage-mana- ger, no matter how industrious, remodeling a speech in a foreign tongue, or calling in a professor of Semitics to correct the grammar. It is much more likely, there- fore, that the second version is the work of a later grammarian. A omits 930-939. 1045 HA. O mi popularis, saliie. AG. Et tu edepol, quisquis es: 1039 Et sfquid opus est, qua^so, die atque Impera 1040 Popuiaritatis caCsa. HA. Habeo grStiam: Verum 6go hie hospitium | hdbeo: Antidamiie fllium Quaero — commostra, si nouisti — Agor^stoclem. (Sed eoquem aduleseentem tu hie nouisti Agordstoclem ? AG. Siquidem Antidama[t]i qua^ris adoptatleium, Ego sum Ipsus, quem tu qua^ris. HA. Hem, quid ego aQdio? AG. Antldamae gnatum me ^se. HA. Si itast, t6sseram Conf^rre si uis h6spitalem, eccam attuli. AG. Agedum, htic ostende. est pfir probe, uam habe6 dorai.) HA. mi h6spe9, mlue mtiltum. nam mihi tGos pater 1050 Patrltus ergo | h6spes AntidamS.s fuit. Haec mlhi hospitalis tessera cum ill6 fuit. Ergo hie apud me hospitium tibi praeb^bitur. u Lindsay, Anc. Edd., p. 44. POENULUa 57 101,2 ff. This passage has long been considered one of the most certain examples of retradatio in Plautus, but scholars are by no means agreed as to the extent of the dittography. There are two inquiries about Agor- astocles (1042-1043; 1044), each introduced by an ad- versative particle {verum, sed), and two presentations of the hospitalis tessera (1047-1049; 1052). The saluta- tion in 11. 1050-1051 is concerned with the identity ol Agorastocles as adopted son of Antidama, Hanno's ances- tral guest-friend (cf. 11. 1042-1043), and the tessera (1. 1052) is brought in almost as an afterthought. Seyifert's^^ solution of the difficulty is on the whole the most satisfactory. He finds two parallel versions: (1) 11. 1042-1043, followed by a reply of Agorastocles, similar to that in 1. 1046, and then by 11. 1050-1053 ; (2) 11. 1044-1049, followed by 1. 1053. Each vqrsion begins with an inquiry about Agorastocles and a dis- closure of the young man's identity; but in the first the recognition rests upon a simple statement, m the second there is a careful examination of the tessera. If Seyffert ^' and Langrehr ^^ are right in thinking Anti- damas (1. 1051) an un-Plautine form,^ the second ver- sion is the genuine. P has the verses in the order of the Goetz-Sehoell text. A has 1041, 1044-1048, 1042-1043, 1053, 1049-1053 (i. e., A trans- poses the couplet 1042-1043, and repeats 1053). 90 Seyffert, Stud. Plant., pp. H f. "SeyflFert, Stud. Plaut., p. 12. « Langrehr, De Plauti Poen., p. 21. » The same form oecurs in 1. 955. The final -s is m both cases necessitated by the meter. 58 RETEACTATIO IN PLAUTUS 1155 1160 1166 HA. Tu abdlic hosc intro et Una nutric^m simul 1147 lube banc abire bine ad te. AG. Fac quod Imperat. MI. Sed quis illas tibi monstrabit? AG. Ego doctlssume. MI. Abeo Igitur. AG. Facias m6do quam memores mauelim. 1150 Patruo fiduenienti c4na curettir uolo. MI. Lachanam uos, quos ^go iam detrudam fid molas, Inde p6rro ad puteum atque fid robustum c6dicem. Ego ffixo hospitium hoc Igniter lauddbitis. AG. Audin tu, patrue? dleo, ne dicttim neges. Tuam mlhi maiorem flliam desp6ndea-s. HA. Pactum rem habeto. AG. Sp6ndesne igitur ? HA. Spondeo. Mi patrue, salue: nfim nunc es plan6 meus. Nunc d#mum ego cum ilia fabulabor llbere. Nunc, -patrue, si uis tufis uidere fllias. Me s^uere. HA. Iam dudum ^quidem cupio et U sequor. (Quid, si edmus illis 6buiam? AG. At ne int^r uias Praet^rbitamus m^tuo. HA. Magne Itippiter, Restltue certas mlhi ex incertis ntinc opes. AG. tEgo quidem meos amores m^cum confid6 fore.) Sed eccas uideo ipsas. HA. Ha^cine meae sunt flliae? Quantae 6 quantillis iam sunt factae. MI. Scin, quid est? Thraeca^ sunt * * * ♦ onem sustolll solent. (OpIno[r] hercle hodie qu6d ego dixi p6r iocum, Id 6uenturum esse #t seuerum et s^rium, 1170 Vt haec Inueniantur h6die esse huius flliae. Pol istfic quidem iam c6rtumst. tu istos, Mllphio, Abdtice intro: nos hasce hie praestolabimur. ) 1162 ff. There are traces of alteration at the close of this scene. In 11. 1160-1161 Agorastocles bids Hanno follow him if he wishes to see his daughters. In 11. 1162-1163 Hanno proposes, as if the plan were quite new, that they go to meet the girls, and Agorastocles objects that they may miss them on the way. At 1. 1166 the two girls are seen approaching, and at 1. 1173 Agoras- tocles and Hanno announce that they will wait for them. Milphio has been commanded to enter the house at 11. 1147-1148, and has apparently done so (cf. 1. 1150 abeoj 1. 1154) ; to our surprise we find him on the stage at P0ENULU3 59 11. 1167 ff., making a belated comment on the identity of the two girls, and receiving the same directions that had been given to him before. There are certainly two versions of the command to Milphio and of the proposal to look for the girls; and it is possible that the whole passage (1162-1173) represents an alternative ending for the scene, intended to follow 1. 1146. A and P give the scene in the same form. Quaeso, qui lubet tarn did tenere collum? 1266 Omltte saltem tu altera: nolo 6go istue — Enicas me. Prius quam te mihi desponderit. AD. [0] Mitt6? sperate, salue. Ck>ndamus alter alterum ergo in n§ruom bracchialem. 1261-1268. There is much confusion in the latter part of the play about the betrothal of Agorastocles and Adel- phasium. It takes place in full form at 11. 1155-1157 (cf. Aul. 255-256; Trm. 571-573), and the promise made at that time is recalled to Hanno's mind at 11. 1278-1279. But at 1. 1268 and again at 1. 1357 Agorastocles speaks as if the betrothal were still to take place. This last line, as we shall see, probably belongs to the latest version of the play, and it is possible that 11. 1267-1268, which are awkwardly introduced at best, also represent a later addition. If the close of V., 3 was changed by the substitution of 11. 1162-1173 for 1147-1161 (cf. above), the betrothal was omitted in the later performance of the play, and the changes in the other scenes may have been made to agree with this. A has the passage in the order of the Goetz-Schoell text. P has the order 1266, 1268, 1267. POENULUS 61 60 RETEACTATIO IN PLAUTUS 1315 1320 1325 AG. Num tlbi, adulescens, mfilae aut denies prtiriunt, Qui huic ^ molestus, fin malam rem qua^ritas? ANTA. t Qur non adhibuisti, dtim istaec loquere, tympanum? Nam t^ cinaedum ease drbitror magis qu^m uirum. AG. Scin qufim cinaedus sum? Ite istinc, serul, foras, Ecf^rte fustis. ANTA. Heus tu, siquid p6r iocum Dixl, nolito in s6rium conuortere. ANTE. Qui[d] tlbi lubidos, opsecro, Antamo^nides, Loqui Inclementer nostro cognato 6t patri? Nam hie n6ster pater est: hfc nos cognoult modo Et htinc sui fratris fllium. ANTA. Ita me Ilippiter Bene am6t, bene factum, gafideo et uolup 6st mihi, Siquid lenoni | optigit magnl mall, Quomque ^ uirtute nobis fortuna 6ptigit. ANTE. Credlbile ecastor dieit: crede huic, ml pater. HA. Credo. AG. 6t ego credo. sM eccum lenonem 6ptume: (Credo. AG. At ego credo |. 6depol hie uenit c6mmodus.) 1331 Bonum uirum eccum uideo: se recipit domum. HA. (Quis hie 4st? AG. Vtrumuis #st, el leno \ijkos. In s^ruitute hie fllias habult tuas, 1334 Et mi [hie] atiri fur est. HA. B^Uum hominem, quem n6ueris.) AG. Rapiamus in ius. HA. Mlnume. AG. Quapropt^r? HA. Quia Inifiriarum mtilto induci sfitius est. LYCVS. AGORASTOCLES. HANNO. LENO ADVLESCENS ANTAMOENIDES MILES POENVS LY. Decipitur nemo m^a quidem sent^ntia, Qui suls amicis nfirrat recte r^s suas: Nam omnibus amicis mels idem unum c6nuemt, 1340 Vt m§ suspendam, ne fiddicar AgorSstocli. AG. Leno, efimus in ius. LY. 6psecro te, Agorfistocles, Susp^ndere ut me llceat. HA. In ius t6 uoco. LY. Quid tlbi mecum autem? HA. Quia [ hasce aio llberas Ing^nuasque esse fllias ambSs meas. 1345 Eae sflnt surruptae cdm nutrice pfiruolae. LY. lam prldem equidem istuc scfui et miratfls fui Nemlnem venire qui Istas adserer^t manu. Meaequid^m profecto n6n sunt. ANTA. Leno, in itis eas. LY. De prSndio tu dicis: debettir, dabo. 1350 AG. Duplum pro furto mi 6pus est. LY. Sume hinc quldlubet. HA. Et mlhi suppliciis multis. LY. Sfime hinc quldlubet. (ANTA. Et mlhi quidem min[im]a[m] arg^nti. LY. Sume hike quldlubet. Coll6 rem soluam iam 6mnibua quasi bfiiolus. AG. Numquld recusas contra me? LY. Hand uerbtim quidem. AG. Ite Igitur intro, mQlieFes. sed, pfitrue mi, 1356 Tuam, (it dixisti, mlhi desponde flliam. HA. Haud aliter ausim. ANTA. B6ne uale. AG. Et tu b^ne uale. ) ANTA. Leno, firrabonem hoc pro mina mecfim fero. LY. Perii h^rcle. AG. Immo haud mult6 post, si u^neris. LY. Quin 4gomet tibi me addloo: quid praetore opust? Verum 6bsecro te ut llceat simplum s6luere, Trec^ntos Philippos: crMo, conradl potest. Cras atictionem f^ciam. AG. Tantisper quidem Vt sis apud me llgnea in cust6dia. LY. Fiat. AG. Sequere intro, pStrue mi, ut hunc festum diem Habefimus hilare[m] huiUs malo et nostr6 bono. MultGm ualete. mtilta uerba f^cimus: Maldm postremo omne Sd lenonem r^cidit. Nunc, qu6d postremumst condiment um f fibulae. Si plficuit, plausum p6stulat como6dia. 1371 m IUS 1360 1365 HANNO. POENVS AGORASTOCLES. LYCVS. ADVLESCENS LENO ANTAMOENIDES MILES ANTERASTYLIS. ADELPHASIVM. MERETRICES II AG. Quam r^m agis, miles? qui lubet patruo meo Loqui Inclementer? n^ mirere mCilieres Quod efim sequntur: modo cognouit fllias Suas ^sse hasce ambas. LY. H6m, quod uerbum aurls meas Tetiglt? nunc perii. ANTA. Vnde haec periertint domo? AG. Carthfiginienses stint. LY. At ego sum p^rditus. 1375 i. 62 EETRACTATIO IN PLAUTUS POENULUS 63 1390 1395 IlMc ego metuei semper, ne cogn6sceret Eas aiiquis: quod nunc factumst. uae miser6 mihi. Peri#re, opinor, du6deuigintl minae 1380 Qui hasce 6mi. AG. Et tute Ipse peristl, Lyce. HA. Quis hie 6st? AG. Vtrumuis 6st. [noluel leno u6l XiJ/cos. In s^ruitute hie fllias habuit tuas Et mi atiri fur est. HA. B^llum hominem, quern n6ueris. AG. Len6, rapacem te 6sse semper cr6didi: 1385 Verum €tiam furacem qui nortint magis. LY. Acc^dam. per ego t te tua te genua Obsecro Et hiinc, cognatum qu6m tuom esse int^llego: Quando boni estis, tit bonos faeere addecet, Facite tit uostro sCbueniatis stipplici. lam pridem equidem istas sclui | esse llberas Et ^xpeetabam, slqui eas asserer^t manu[m] : t Nam meae prosum n6n sunt, tum autem aurtim tuom Reddam quod apud mest €t ius iurandum dabo Me malitiose nil fecisse, Agordstocles. AG. Quid mihi par faeere sit, t cum egomet c6nsulam. Omltte genua. LY. MItto, si ita sent6ntiast. ANTA. Hetis tu, leno. LY. Quid lenonem uls inter negCtium? ANTA. Vt inam mihi arg^nti[m] reddas, prius quam in neruom abdticere. LY. DI meliora fdxint. ANTA. Sic est: h6die cenabis foris: 1400 Atirum, argentum, e6llum, leno, [s]trl3 re nunc deb^s semul. HA. Quid me hae re fdcere deeeat, Egomet mecum c6gito. SI nolo hunc ulclsei, litis s6quar in alieno 6ppido, Quantum audiui ing^nium et mores 6ius quo[d] paet6 sient. AD. MI pater, nequld tibi cum istoc rel t siet ac massum 6b3ecro. 1405 ANTE. Atisculta sor6ri. abi, diiunge Inimicitias cum Inprobo. HA. H6c age sis, len6: quamquam ego te m6ruisse ut pere^s scio, N6n experiar t^cum. AG. Neque ego, ei atirum mihi reddfis meum. Leno, quando ex n6ruo emissu's, c6npingare in cSrcerem. LY. iam autem, ut sol4s? Ego, Poene, tIbi me purgattim uolo. 1410 Slquid dixi irfitus aduorsum animi tui sent^ntiam: fd uti ignoseas, qua^so, et quom istas Inuenisti fllias, fta me di ament, mihi uolup[tatis] est. HA. fgnosco et cred6 tibi. ANTA. tLeno, tu autem amicam mihi des facito aut auri mihi reddas minam. LY. Vin tibieinam meam habere? ANTA. Nil moror tiblcinam: N6scias utrum el maiores bticcaene a mammal sient. 1416 LY. Dabo quae placeat. ANTA. Ctira. LY. Aurum cras[3] a te referam tuom. AG. Facto in memoria habeas. LY. Miles, s6quere me. ANTA. Ego uero sequor. AG. Quid ais, patrue? quando hinc ire c6gitas Carthaginem? Nam tecum mihi una Ire certumst. HA. Vbi primum potero, Ilico. 1420 AG. Dum atietionem facio, hie opus est aliquot ut manias dies. HA. Faciam ita ut uis. AG. Age sis, eamus: nos curemus. platidite. ) Double Ending (1315-H22), The manuscripts of Plautus give two versions of the close of the Poenulus, each ending with a request for applause (11. 1315-1371 ; 1372-1422). Early editors tried to fix upon one of these endings as genuine, and to discard the other ; ^^ but the prevalent view to-day is that both contain a mixture of the genuine text and the text of the Plautine Kevival, with some still later additions and interpolations. The older of the two versions is represented by 11. 1322-1355 ; 1398-1422 ; but even this is probably not preserved in the form in which Plautus wrote it. Substitutions were made for both parts of this ending, though not necessarily by the same author or at the same date. The first half was displaced by the confused dialogue of 11. 1372-1397, and the last by 11. 1356-1371, the only final scene in Plautus which is written in iambic senarii, and the only one in which the abstract comoedia asks for applause.^^ Both A and P give both endings. A makes no division between the two; P leaves a space as if for a new scene before 1. 1372. '^See Goetz-Loewe, Ed. (1884), on V., 7. ^'On the whole question, see Langen, Plant. Stud., pp. 343 ff.; Leo, Plant. Forsch., p. 158, note 3. 64 EETEACTATIO IN PLAUTUS POENULUS 65 Except for some slight differences which will be discussed below, the form of the two endings is the same in both families of manuscripts. 1315 ff. The general discussion of the Double Ending- has left untouched 11. 1315-1321. The starting-point of both endings is the insolent speech of Antamoenides to Hanno (11. 1309-1314). As a matter of fact, this speech is rebuked three times: (1) by Agorastocles (11. 1315 fF.) ; (2) by Adelphasium (11. 1322 ff.) ; (3) by Agorastocles (11. 1372 if.). The second reproof follows the apology of the miles (11. 1320-1321), and is clearly out of place; the third, as we have already seen, belongs to the later version of the ending. Moreover, there are three an- nouncements of the approach of the leno in three successive lines (1330-1332). The second of these so closely res- embles the first that it must be regarded as an alternative version, and the repetition of eccum in the third throws suspicion on that line also. The passage which follows L 1331 in P (11. 1333-1335) is identical with 11. 1382- 1384. It seems more appropriate in the first position, for we should expect Hanno to inquire about the ap- proaching stranger f^ in the second, since Lycus is already taking part in the dialogue, there is much less reason for the question. It is probable that in the first two versions Antamoenides was rebuked for his rudeness, apologized, and received forgiveness; then as the leno was seen ap- proaching, Hanno inquired about him. In the third ver- sion, Lycus entered abruptly after 1. 1314, just in time to hear the reproof of Antamoenides and to learn that Hanno was the father of his two slaves. The three versions therefore ran as follows: (1) 11. 1315-1321, ^'Langen, Plant. Stud., p. 351. followed by 11. 1329-1330, 1333-1335; (2) the genuine version, 11. 1322-1328, 1329-1330,1333-1335; (3) 11. 1372- 1381. P omits 1331 (which is certainly due to retractatio) , but aside from that omission gives 1315-1337, 1372-1384 in full form (1333- 1335 = 1382-1384). A omits 1333-1335, but writes quis hicst at the close of 1332, showing that the lines must have appeared in full form somewhere in the A-family. If we assume that 1322-1328 and 1372-1381 were alternative versions, the repetition of 1333-1335 after 1381 is not surprising. 1378 if. The repetition of 1. 1377 after 1. 1381 in P may perhaps indicate a shortening of the scene by the omission of 11. 1378-1381.^'^ P repeats 1377 after 1381. A deest after 1381, but apparently did not repeat 1377, for the 38 lines of the missing sheet would have given just enough space for 1382-1419. The Poenulus shows more extensive alteration than any other play of Plautus. It contains the longest single case of dittography — the Double Ending — and both this ending and the body of the play have received an unusual amount of revision. In two places (11. 504 ff. ; 1315 ff.) there are clear traces of three parallel versions. Scat- tered through the play, too, in passages which would otherwise arouse no comment, are a number of suspicious forms (the singular verbs in 11. 523, 728, 730, 733 ; the nominative in -s in 1. 955). Then there are, of course, examples of the more common kinds of retradatio: var- iant lines (214, 218, 304, 390'') and longer alternative versions (11. 121-128; 917-929; 1042-1052), and possible ^'Leo, Ed. (1896), ad loc. Lindsay. Ay\c. Edd., p. 45, note c. 66 EETRACTATIO IN PLAUTUS attempts to shorten scenes (11. 622", 706 ff., 805, 1377 ff.)-^' For a number of these passages, the evidence of the Ambrosian Palimpsest is lacking. In the long Double Ending, the text of A and P is practically the same, and it is the same also in 11. 917-929, 1147-1173. A and P give the same text, though their order differs, in 11. 300-305, 1267-1269; and in two places, though A does not preserve the whole of the passage, it gives enough to show that the passage was present in some form in its archetype (11. 389 ff . ; 1333 if.). In two instances where retractatio undoubtedly exists, A shows slightly more evidence of it than P (11. 706 ff., 1042 ff.) ; and the alternative line 1331 is preserved only in A. The one change which A apparently fails to indicate is the short- ening at 11. 1378 fF. On the whole, then, A shows slightly more evidence of retractatio than P. ^The following lines have also been suspected: 5-10, 14, 27, 37-40, 46-58, 79-82, 159-163, 176, 225-232, 242, 244 ff., 263-274, 275-282, 289-296, 313-316, 327-328, 330-408, 352-354, 360-364, 370-380, 385, 408, 419, 456^S 518, 521, 550, 551, 576, 631-632, 669-678, 717- 718, 721-745, 831, 865, 877-878, 950-960, 967-970, 982-984, 988- 989, 990-991, 1020, 1036, 1075, 1079-1085, 1108, 1116, 1125-1126, 1159, 1192-1204, 1212-1236, 1277-1279, 1328-1330, 1336-1337, 1349, 1350, 1353-1358, 1393 f., 1401, 1403-1404, 1408, 1417. The evidence of A is lacking for 5-10, 14, 27, 37-40, 46-58, 79-82, 159-163, 176, 225-232, 242, 244 ff., 263-274, 275-282, 518, 521, 550, 551, 631-632, 717-718, 721-745, 831, 1075, 1079-1085, 1393, 1401, 1403-1404, 1408, 1417. Both A and P have 289-296, 313-316, 327-328, 330-408, 352-354, 360-364, 370-380, (P has the order 376, 375), 385, 408, 419, 576, 669-678. 865, 887-888, 950-960, 967-970, (P has the order 969, 968), 982-984, 988-989, 990-991, 1020 (in different order), 1036, 1108, 1116, 1125-1126, 1159, 1192-1204, 1212-1236, 1277-1279, 1328-1330, 1336-1337, 1349, 1350, 1353-1358 (A omits 1353). A omits 456b^ The following lines show minor variations: 331, 342, 343, 587, 690, 875, 975, 1332. CHAPTER III. PSEUDOLUS The Pseudolus was a favorite with Plautus himself, if we may believe Cicero (De 8enect, 14, 50), and its lively action and exuberant humor have made the play popular ever since. The situation is the old one of the youth in love. This time his name is Calidorus, the girl is called Phoenicium, her master Ballio, and the title- role is taken by Pseudolus, the slave of the young lover. In the first scene, Calidorus shows Pseudolus a letter from Phoenicium, announcing that she has been sold to a Mace- donian soldier for twenty minae; three-quarters of the sum have already been paid, and she is to be handed over, on the payment of the last quarter, to the soldier's mes- senger, who is to bring an impression of his master's seal- ring as identification; the limit set for the payment of the balance is the next feast of Dionysus, which falls on the morrow. Ballio is next introduced — a brutal char- acter at best, and especially brutal to-day, because it is his birthday, and he is planning to give a grand dinner. Calidorus appeals to him for six days of grace on the payment of twenty minae for Phoenicium, is told, to his surprise and delight, that she is not for sale, and then hears that she has already been sold to a Macedonian soldier — with all the details just as they were given in the first scene, except that the time set for the payment of the other five minae is " to-day." Pseudolus under- takes to help his young master, frankly confesses his purpose to the father, Simo, and warns the old gentleman 67 68 EETEACTATIO IN PLAUTIJS that he intends to cheat him. He adds that he is plan- ning a second campaign, against the leno Ballio. If he succeeds in both plots, Simo is to pay for the girl. Simo's friend Callipho promises to remain at home that day and lend his aid to the schemer. Lnck throws into the hands of Psendolus a sealed letter from the miles; so a rogue who rivals Pseudolus himself in cleverness is dressed up like the soldier's messenger, and sent to Ballio, with the letter and five minae, Ballio recognizes the soldier's seal, delivers the girl to the messenger, and is then so sure of safety from attack that he offers to give Simo twenty minae if Pseudolus succeeds. But when the real mes- senger of the miles appears, it becomes evident that the trick is already played, and that Ballio is the victim. Simo acknowledges that Pseudolus has won his twenty minae, and the money is handed over,— to a very drunk and very impudent slave. The plot of the Pseudolus shows obvious inconsistencies. The explanation given in the letter (11. 51-59) makes the later account of the sale of Phoenicium superfluous, and the attitude of Calidorus — his tender solicitude for the girl (1. 231), his unboimded joy when he hears that she is not for sale (11. 323 ff.), his despair when the truth is finally told (11. 342 ff.) — is quite incomprehensible in one already familiar with the contents of the letter. There is a discrepancy, too, about the time when the final payment of the miles is due — hodie in the one case (11. 373 f., 623 f.), eras in the other (11. 60, 82). Calli- pho, whom we expect to take a prominent part in the action after his promise of aid to Pseudolus (11. 547-560), disappears from the stage at the end of the scene and is never mentioned again. PSEUDOLUS 69 Ladewig ^ noticed some of the discrepancies, and sug- gested that the Pseudolus really contained two plots : (1) the duping of the old man to get the money; and (2) the duping of the leno to get the girl. But other scholars paid little attention to this suggestion. Langen ^ noted a number of difficulties in the plot, without attempting to explain them by coniaminatio, and Leo ^ confined his hint of two originals in the Pseudolus to a foot-note. In 1897, Bierma* made a detailed analysis of the play, and this analysis, modified by Seyffert,^ Leo,^ and lately by Schmitt,*^ is the basis of work today. It is evident that the Pseudolus does not, like the Miles and the Poenulus, combine two full plots, but that certain scenes from one plot have been worked into the other. The two plots agree in general features : in both, the difficulties of the young man in love are removed by the clever slave ; in both, the father has received some warning of the slave's plans, and has to pay over a sum of money at the end. The two plots may be sketched as follows : (A) The girl with whom the young man is in love has been sold to a Macedonian soldier ; part payment has been made, and the balance is due on the day that the play opens. The slave outwdts the leno by getting possession of the soldier's letter, and sending a messenger ^ Th. Ladewig, Vher d. Kanon des Volcatius Sedigitus, Neu-stre- litz, 1842, pp. 32 f. * Langen, Plant. Stud., pp. 198 flf. ' Leo, Plant. Forsch., p. 153, n. 2. * J. W. Bierma, Quaestiones de Plautina Pseudolo, Groningen, 1897. 'Seyffert in Berl. Phil. Woch. 18 (1898), colL 1511-1515. •Leo in Nachr. Gbtt. Ges. 1903, pp. 347-354. ' A. Schmitt, De Pseudoli Plautinae exemplo Attico, Strassburg, 1909. (Accessible to me only in a review by M. Niemeyer in Berl, Phil. Woch. 30 (1910), coll. 870-873). 70 EETBACTATIO IN PLAUTUS with it to fetch the girl. The actual cost of the girl is defrayed by the young man's father, who has previously agreed to pay this sum if the slave's ruse succeeds ; but he is reimbursed by the leno, who has in the meantime promised him twenty minae on the same conditions. So the leno is ruined, and his fall is all the greater because the scene is laid on his birthday.^ (B) The young man needs twenty minae to pay for the girl, the morrow being the last opportunity. The slave gets this sum by cheat- ing the father, after he has announced his intentions and warned the old man to be on his guard. In the trick (the details of which we can not fill out) he has the help of the father's friend Callipho. To the first plot (A) belong I., 2, 3 ; II., 2-IV., 8 ; to the second (B) : I., 1 (ex- cept 11. 51-59), 4, 5 (except 11. 522-546) ; V., 2. Plautus himself made some changes; he added 11. 51-59, and 522- 546 ^ (elements from plot B) to plot A; he combined the two plots in II., 1 ; and he added V., 1 ^^ and certain details (e. g. 1. 1308) in V., 2. •It is possible that this motif had even more prominence in the original.- It is emphasized in 11. 165, 167, 179, 234, 243, 1237, and forms the sole connecting link between Act III. and the rest of the play. Cf. Bierma, De Plant. Pseud., pp. 40 ff., 87 f.; Leo in 'NacJir. Oott. Ges. 1903, p. 352. •The insertion is unskillfully made, and has resulted in some confusion. As the dialogue stands, Simo agrees that if Pseudolus succeeds in both plots (i. e., the duping of Simo himself and the duping of Ballio) Simo will pay him twenty minae. The original bargain in plot B must have been that Simo would pay the money if Pseudolus succeeded in cheating Ballio. Cf. Leo in Nachr. Gott. Ges. 1903, pp. 349 ff. *•* Hence the double motivation of the entrance of Pseudolus (11. 1282, 1283), and of Simo's return to the stage (11. 1239 ff.. 1285). Cf. Leo, Plaut. Cant., p. 41, note, in Ahhandl. Gott. Ges., 1896-1897. PSEUDOLUS 71 Exp<5rgi meliust Ifimbos atque exstirgier: Plautlna longa fibula in scaenSm uenit. Prologue (II 1-2). Both the Ambrosian Palimpsest and the Palatine manuscripts give these two lines, which bear the mark of the Plautine Kevival in the phrase Plautina longa f alula. (Cf. Cas. Prol. 12). A pro- logue is really unnecessary for the exposition of the Pseu- dolus,'^^ and Lindsay ^^ therefore concludes that the ex- tant couplet formed the whole of the later stage-man- ager's introduction. But though the play may be per- fectly clear without the prologue, the prologue in its present form is far from clear. The comparative melius (1. 1) demands as least an implied positive, and there is no expressed subject for the infinitives exporgi and exsur- gier. The extant prologue must therefore have been preceded by several lines, and it was probably followed by others giving the name of the Plautina fahula and making the usual request for quiet.^^ P prefixes the prologue to Argument I. (omitting Argument II.). A places the prologue immediately before Scene I. (omitting Argument I. and giving Argument II. in a later hand). PS. 'Nunc n6stri amores, m6res, consuetfldines, locus Itidus, sermo, suaui[s]sauiatio, Compr6Svsiones drtae amantum comparum, Tenerls labellis molles morsiflnculae, Nostrorum orgiorum ♦ g * * * itinculae, Papillarum horridulJlrum oppress! tinculae'. Hartinc uoluptatum mlhi omnium atque itid^m tibi Distractio, discldiura, uastiti^s uenit. 64 65 67* 70 "Leo, Plaut. Forsch., p. 196. "Lindsay, Anc. Edd., p. 1, note a. "Leo, Plaut. Forsch., p. 197. 72 RETRACTATIO IN PLAUTUS 67". The verse, as Lowe ^* noted when he tried to re- construct its reading from the Palimpsest, is not appro- priate here, for it inserts a phrase of general nature in a list of specific details. He likewise objected to the word orgia, which does not appear elsewhere in Koman literature before Catullus. Leo ^^ surmised that the line might be a later addition to 1. 64, and this hypothesis is possible, although in the broken state of the text it is hard to reach definite conclusions. P omits 67*'. Nempe ita animati | ^stis uos: uincltis duritia hoc fitque me 151 Hoc 8ls uide ut alias r6s agunt. hoc agite, hoc animum aduortite. Hue adhibete auris quae ^go loquar, plagigera genera | homi- num. Numquam €depol uostrum dfirius tergum 6rit quam terginum h6c meum. Quid ntinc? doletne? em sic datur, siquis erum seruos sp6rnit* 155 Adslstite omnes contra me et qua€ I6quar aduortite finimum. 151-156, Even the general wordiness of Ballio's speech does not excuse the three commands to pay at- tention (11. 152, 153, 156) in so brief a space. Vincitis duritia hoc \_terginum'] atque me (11. 151), and numquam edepol nostrum durius tergum erit quam terginum hoc meum (1. 154), have the same idea at base, though the turn given to it is different ; and the two lines are intoler- able together. Hoc (1. 151) must mean terginum (cf. 1. 154), and though the action on the stage probably helped to make the sentence clear, still the postponement of the noun is at least noticeable. There are two cases of verbal "G. Lowe, Analecta Plauiina, Leipzig, 1877, pp. 153 ff. >*Leo, Ed. (1896), ad loc. PSEUDOLUS 73 repetition: 152 animum advortite, cf. 156 advortite animum; 153 quae ego loquar, cf. 156 quae loquar, Therefore it seems possible that we have two versions : (1) 151-153 ; (2) 154-156. Of these, the second is probably Plautine, for the phrase contra me (1. 156) to which Lorenz ^® objected, is paralleled in Pers, 13, and the clearness of 1. 156, as opposed to 1. 151, together with the play on tergum . . . terginum, counts in its favor. P has the verses in the order of the Goetz-Schoell text. A has the whole passage, but puts 153 after 154. Tfbi hoc praecipio ut niteant aedes: habes quod facias: propera, abi intro. 161 TO esto lectist^rniator. tfl argentum eluito, Idem exstruito. Ha^, quom ego a foro reuortor, facite ut offendSm parata, versa sparsa, t^rsa strata, latitaque unctaque 6mnia uti sint. Nam mi hddie nataUs dies est: decet 6um omnis uos conc^le- brare. 165 Pernam callum glandlum sumen facito In aqua iaceant. satin audis? Magnlfice uolo me ulros summos acclpere, ut mihi rem ess6 reantur. Intro abite atque haec cite celebrate, ne m6ra quae sit, cocus qu6m veniat [mihi]. 166. This verse breaks the close connection between 1. 165 and 1. 167, and is quite out of place here under any circumstances. Ballio has finished his individual com- missions in 1. 162, and 11. 163-168, with the sole exception of 1. 166, are general directions in the plural. ^"^ The line is given in both A and P. "A. Lorenz in Philol. 35 (1876), p. 159. " Cf. H. Usener in Index schol. Oryphiswald., 1866, pp. 13 ff. 74 EETRACTATIO IN PLAUTUS t Xytilis, face ut Snimum aduortas, quofus amator6s oliui 210 Dynamin domi habent maxiimam: SI mihi non iara hue cfilleis 61euin deportatum erit, Te fpsam cuUeo 6go eras faeiam ut d^portere in p4rgulam. Ibi tibi adeo 16ctus dabitur, tibi tu hau somnum cdpias, sed ubi 215 ■^sque ad languor^m — : tenes Qu6 se haec tendant qua6 loquor. Aln, exeetra tu, qua6 tibi amicos t6t babes tarn probe 6leo onustos ? Num quolpiamst hodi^ tua tuorum 6pera conseru6ruin NltidiuscuMm caput? aut num ipse ^go pulmento ut6r magis 220 "C^nctiuscule ? sed scio, tu | 6leum hau magni p6ndis: uino T6 deu[i]ngis. sine modo: Rgprehendam ego cuncta h^rcle una opera, nisi quidem tu hodie 6mnia Fficis eflFecta haec tit loquor. 21 0-22 Jf. As the text stands, this whole passage is addressed to Xytilis. The length of the tirade is quite out of proportion to the speeches addressed to the other meretrices (11. 188-193 ; 196-201 ; 225-229). The sense of 11. 210 f. is repeated in 1. 218, and the threats of 11. 222-224 are an anticlimax after 11. 212-217. It is pos- sible that 11. 218-224 are a second version, intended to avoid the brutality of 11. 210-217.^8 The passage has the same form in A and P. PS. H6c ego oppidum Sdmoenire ut h6die capiatdr nolo. 384 Ad earn rem usust h6minem astutum d6ctum, cautum et cfiUi- dum, 385 Qui Imperata ecf^cta reddat, non qui uigilans d6rmiat. OA. C^o mihi, quid 63 facturus? PS. T^mperi ego fax6 scies. "E. Norden in Rhein. Mtis. 49 (1894), pp. 197 ff. PSEUDOLUS 75 N6lo bis iterdri: sat sic 16ngae fiunt fabulae. CA. 6ptumum atque aequlssumum oras. PS. Pr6pera, adduc homin6m cito. Patici ex multis sfint amici, h6mini qui certi sient. 390 PS. figo scio istuc: 6rgo utrumque tibi nunc dilecttim para Atque ex multis 6xquire illis tinum qui cert(is siet. lam hie faxo aderit. PS. Potin ut [h]abeas? tibi moram dictis creas. 385-386. The end of this scene is almost hopelessly confused. The request of 11. 385-386 is repeated in 11. 724-728, where it is much more in place ; for Pseudolus has by that time gained possession of the letter and has formed a plan in which he needs the help of a clever rogue. The reflection of Calidorus (1. 390) and the re- sponse of Pseudolus (11. 391 f.) sound as if Pseudolus had really asked, not for a rogue, but for a trusty friend ; and 11. 697f., spoken when Calidorus brings in his friend Charinus, agree with this interpretation: Pseudolus mihi ita imperavit, ut aliquem hominem strenuom Benevolentem adducerem ad se. LI. 385-386, then, are inappropriate in their present position. It seems possible that they were made up on the model of 11. 724 ff., and got into this position by mistake, crowding out a bit of the original dialogue, in which Pseudolus asked for the aid of a friend. The passage has the same form in A and P. PS. Atque 6go me iam pridem hulc daturum dixeram 406 Et u6lui inicere trfigulam in nostrtim senem: Verum Is nescioquo pScto praesensit prius. Jf.06-Ji.08. The vague allusion of these verses is not 76 EETEACTATIO IN PLAUTUS PSEXJDOLUS 77 explained by anything in the earlier scenes of the Pseudolus, and the difficulty is not materially lessened by the assumption of contaminatio}^ It seems probable that the passage is a later insertion, suggested by 11. 421- 422 : SI. Si d^ damnoseis aut si de aiiiat6ribus 415 Dictator fiat ntinc Athenis Atticis, Nemo anteueniat filio cred6 meo. Ita nCmc per urbem s6lus sermoni omnibuat, Eum u^lle amicam llberare et qua^rere Argentum ad earn rem: hoc dlii mihi rentintiant Atque Id lam pridem s^nsi et subolebat mihi, 421 Sed dlssimulabam. PS. lam illi fe * * fHius: Oeclsast haec res, ha^ret hoc negotium. One notices the repetitions : 406 iam pridem, cf. 421 iam pridem ; 408 praesensit, cf. 421 sensi ; and the fact that huic (1. 406) has no antecedent. The passage occurs in both A and P. PS. Vin €tiam dicam quod uos magis mir^mini? CA. Studeo h#rcle audire: nam ted auscult6 lubens. SI. Agedtim: nam satis lub^nter te auscult6 loqui. 522 623»' 523^'', These verses are clearly doublets. Abraham ^° has shown conclusively that the second is un-Plautine. Auscultare in Plautus never takes a complementary in- finitive, and is never joined with an adverb, but always with an adjective ; agedum is never used alone, but always with another imperative; and satis lubenter does not occur elsewhere in Plautus. The intention of the author "Bierma's explanation. Cf. De Plant. Pseud., pp. 38 f., 56 f. ^•W. Abraham, Studia Plautina, Leipzig, 1884, pp. 182 flf. of the line may have been either to supply a variant for 1. 523', or to add another verse, and so give both the old men speaking parts. P has the verses in the order of the Goetz-Schoell text. A deest 2 sheets =76 11. A deest 477-551 = 77 11. (including 523*' and 543^. Therefore A probably omitted 523^. ConcMere aliquantlsper hinc mihi intro lubet, Dum concenturio in corde sycophantias. ***** non ero uobls morae. Tiblcen uos int^ribi hie delectauerit. 571 573* 573'' 57S\ The beginning of 1. 573' is lost in the Palimp- sest, but enough is left to show that it contained much the same announcement as 11. 571-572. Interihi (1. 573") must, as Baier ^i points out, refer not to the line imme- diately preceding, but to dum concenturio (1. 572). Therefore 1. 573' is probably due to retractatio, its object being to announce clearly the return of Pseudolus in the next scene. P omits 573'. SM iam satis est phllosophatum : nimis diu et longum loquor 687 Di Immortales, atirichalco c6ntra non carlim fuit MeGm mendacium, hlc modo quod sfibito commentus fui, Quia lenonis m6 esse dixi. ntinc ego hac epfstula 690 Trls deludam, erum 6t lenonem et qui banc dedit mihi epistulam. Etige: par pari aliud autem qu6d cupiebam c6ntigit: V6nit eccum Calid6rus: ducit nfecioquem sectim simul. 688 ff. The exultant di immortales, with which 1. 688 begins, echoes 1. 667, and the tone of the next few ^^B. Baier, De Plauti Fah. Recenss., p. 128. 78 EETRACTATIO IN PLAUTUS lines is what we should expect at the beginning of a speech. This is particularly strange, because Pseudolus has just announced (1. 687) : sed iam satis est philosophatum, Leo 22 therefore advances the plausible theory that the speech of Pseudolus originally consisted of 11. 667-687 sed iam satis est philosophatum, followed by 690 nunc ego hac epistula -693; then 11. 688-690 quia lenonis me esse dixi were substituted for the early part of this scene; and the editor who combined the two versions, finding a vacant half line at the end of 1. 687, supplied nimis diu et longum loquor. The piassage has the same form in A and P. CA. Dtilcia atque amSra apud te sum 4locutus 6mnia: 694 Scls amorem, scis laborem, scis egestat6m meam. CH. C6mmemini omnia: Id tu modo me quid uis facere fdc sciam. 696* [CA. Quora haec tibi alia sum elocutus uis scires si scis de symbolo. CH. Omnia inquam tu modo quid me facere uis fac ut sciam.] 696* 696''\ These two lines, which are in part a verbal repetition of 11. 694-696', seem to offer an alternative version: 694 apud te, cf. 696*^ tihi; 694 sum elocutus, cf. 696' sum elocutus; 696' omnia, cf. 696" omnia; 696' tu modo me quid vis facere fac sciam, cf. 696" tu modo quid me facere vis fac ut sciam. The symholum which is introduced so abruptly here (1. 696**) could have had no special significance for the story of Calidorus, though it was of importance in the trick that Pseudolus planned =2 Leo, Ed. (1896), ad loc. PSEUDOLUS 79 to play. The purpose of the alteration may have been to emphasize this detail. A omits 696'"' SI. Viss6 quid rerum m6us Vlixes 6gerit, 1063 lamne hdbeat signum ex firce Balli6nia. BA. f6rtunate, cMo fortunatam manum. 1065 SI. Quid €stt BA. Iam — SI. Quid iam? BA. Nil est quod metufis. SI. Quid est? Venltne homo -ad te? BA. N6n. SI. Quid est igittir boni? BA. Mina6 uiginti s^nae et saluae sdnt tibi, Hodi6 quas aps te 6st stipulatus Psetidolus. 1069 SI. Sed c6nuenistin h6minem? BA. Immo amb6 simul. SI. Quid alt? quid narrat? qua^so, quid dicit tibi? BA. NugSs theatri: u4rba quae in comoMiis Solent lenoni dici, quae puerl sciunt: Malum 4t seelestum et p6iurum, aibat 6sse me. SI. Pol ha(i mentitust. BA. 6rgo baud irattjs fui. Nam quSnti refert el nee recte dlcere^ Qui nll faciat quique infitias n6n eat? 1079 1080 1085 1079-1086, These lines were suspected of being due to retractatio by Kiessling,^^ in 1868, and have been bracketed by most editors since then. The question of 1. 1079 is a surprise after 1. 1067, and the answer an absolute contradiction. The sense of the two questions is the same; for Simo is evidently referring, not to the meeting with Pseudolus in L, 3, which Ballio describes, but to a meeting in which Pseudolus might have tried to trick Ballio. So 1. 1079 repeats 1. 1067, and the whole passage is probably a later addition. A and P give the passage in the same form through 1084 — then A deest. 23 A. Kiessling in Rhein. Mus. 23 (1868), pp. 425 f. 80 BETEACTATIO IN PLAUTUS Hefis, ubi estis uos? Hioquidem ad me r4cta habet rectam uiam. Ai^t> /HA.> Hefi«, ubi estis uos? Hefis, adulescens, quid istic ^ ^ debetfir tibi? B6ne ego ab hoc praedatus ibo: noui, bona scaeuSst mihi /HA> Acquis hoc aperit? Heus, chlamydate, quid istic ^ ^ ^ debettir tibi? A6dium dominflm lenonem Ballionem qua^rito. lUO llSl, L. 113 Y, which combines 1. 1136 and 1. 1139, was probably intended as a substitute for 11. 1136, 1138- 1139. If 1. 1137 were genuine, we should expect it to be followed, not by another aside by Ballio (1. 1138), but by the answer of Harpax (1. 1140).'^^ Both A and P give 1. 1137. B\ fidepol hominem uerberonem Psetidolum, ut doct6 dolum 120o C6mmentust: tanttindem argenti qu&ntum miles d^buit D4dit huic atque hominem 6xornauit, mtilierem qui abdu- ceret. 1207. 1208 1205-1201. An attempt to cut out 11. 1162-1204 may be indicated by the repetition of 11. 1205-1207 after 1. 1161. BA. Quid ogimus? manuf^sto teneo hunc h6minem qui argen- tum attulit. 1160 SI. Quidum? BA. An nescis qua€ sit haec res? SI. Tflxta cum ignarlssumis. BA. Peefidolus tuos dllegauit htinc, quasi a Mac^donio Mllite esset. A has the passage in the order of the Goetz-Schoell text. P has 1205-1207 both after 1161 and after 1204 2* The first scholar to suspect the line was Fleckeisen. Cf. RitschVs Ed. of Pseud, (1850), Praef., pp. xiiif. PSEUDOLUS 81 BA fidepol ne istuc mSgis magisque m^tuo, quom uerba «a(i- dio. 1214 [BA.] Mlhi quoque edepol ifim dudum ille Sfirus cor perfrige- facit, Sflmbolum qui ab hoc accepit. mfra sunt, ni Pseiidolust. 121Jf-1216. The first of these verses is the same in sense as the last two, and the word edepol is repeated. It is worth noting, also, that the best manuscripts of the Palatine family prefix BA. to 1. 1215. Perhaps 1. 1214 was substituted by the same reviser who cut out 11. 1162-1204, who accordingly wished to avoid the name Sums. (Cf. 1. 1203).2« The passage has the same form in A and P. The results of work on the Pseudolus are somewhat unsatisfactory. After the imdoubted evidences of retrac- tatio in the prologue, we expect to find further traces of change in the text of the play. But only 11. 523,** 688 ff., 1137 can be put down with any degree of certainty as later versions. In many other places (e. g., 11. 166, 385 ff.), though there is unquestionably something wrong with the text, it is hard to locate the trouble exactly. Certain lines preserved in A alone (11. 67^ 573*) are probably due to retractatio, but are so fragmentary that any statement about them must be qualified. Two pas- sages (11. 151-156, 210-224) seem to present parallel versions; 11. 406-408 are apparently a later addition; and 11. 1205-1207 were perhaps inserted after 1. 1161 to shorten the scene. ^^ 25 Ribbeck suspected the verse. Cf. Ritschl's Ed. of Men. (1851), Praef., p. xv. ^In addition to the passages discussed in detail, the following lines of the Pseudolus have also been suspected: 65, 82, 91 ff., 97, 82 RETKACTATIO IN PLAUTUS In these cases of retractatio, possible or probable, the two families of manuscripts are almost evenlj balanced. Most passages have the same form in both. A is the only one to preserve 11. 67', 573". P alone has 11. 696'' and repeats 11. 1205-1207, and A probably omitted 1. 523' also. But at least the Pseudolus shows that A is not a purer text than P. 116, 142, 176, 177, 205-208, 238, 259-263^ 269, 284, 292-295, 307, 336!., 384, 398, 403, 422, 467, 485, 497-499, 502 f., 527, 530, 543S 544, 550, 565 f., 576 f., 585^ 586, 599, 600, 737-750, 759-766, 767- 904, 768, 781 f., 842 f., 866-889, 936-939^ 944, 1002-1008, 1025-1031, 1043, 1073, 1093, 1097, 1098, 1196, 1204, 1245, 1259-1261, 1277 f., 1314. These suspected passages are preserved as follows: A deest: 238, 259-263^ 485, 497-499, 502 f., 527, 530, 543^ 544, 550 737-750, 759- 766, 767-904, 768, 781 f., 1025-1031, 1043, 1093, 1097, 1098, 1259- 1261, 1277 f. Both A and P have 65, 82, 91 ff., 97, 116, 142, 176, 177, 205-208, 269, 284, 307, 336 f., 384, 398, 403, 467, 565 f., 576 f., 585^ 586, 599, 600, 842 f., 866-889, 936-939^ 944, 1002-1008 (A transposes 1002, 1003), 1073, 1196, 1204, 1245, 1314. A omits 293- 295, and places 292 after 296. P omits 422. The text of the Pseudolus shows an unusually large number of slight differences between the two families of manuscripts. Varia- tions of a word or phrase occur in the following lines: 85, 152, 208, 223, 298, 308, 315, 321, 340, 372, 375, 385, 389, 391, 392, 397, 418, 432, 433, 451, 621, 627, 631, 659, 669, 700, 723. 841, 843, 856, 864, 873, 889, 901, 954, 955, 975, 978, 992, 997, 1175, 1204, 1220, 1294, 1295, 1299. ^2 Hi CHAPTER IV. X STICHUS The opening scene of the Stichus introduces two sisters, whose husbands have left home three years before to repair their damaged fortunes, and have not been heard from since. Their father, Antipho, wishes them to marry again, but they firmly refuse. The elder sister, Panegyris, sends for the parasite Gelasimus, explaining that she wishes to dispatch him to the harbor for special tidings. But before he reaches her door, the slave who is regularly on the watch comes from the port with the news that Epignomus and Pamphilippus have arrived. Epignomus and his slave Stichus appear on the stage; Stichus asks for a holiday, and receives permission to go to a banquet with Sagarinus, the slave of Pamphilip- pus, and Stephanium, who is arnica amhohus. Then follows a series of scenes in which the parasite makes desperate efforts to get an invitation to dinner, but is repulsed, and the two brothers, who have in the mean- time become reconciled with their father-in-law and with their wives, plan a banquet together. After this Stichus comes on again, ready for the feast, and through the last six scenes of the play he and his two friends drink and dance and sing. " Ein ratselhaftes Stiick," said Teuffel ^ of the Stichus, and the play has remained " a puzzle '^ to scholars ever *W. Teuffel in Ehein. Mus. 8 (1853), p. 38 {=8tud. u. Char.* p. 340). 83 84 EETEACTATIO IN PLAUTUS since. All are agreed ^ that the play which we possess is very (liferent from the " Adelphoe Menandru " which the didascalia names as its source. Eitschl ^ char- acterized the Stichus as " ansserst fluchtig skizzirt," and Leo ^ has said of it in recent years, " Die Teile sind sehr hiibsch, das Ganze nnmoglich." The play opens as if it were to hinge upon the faith- fulness of two wives to their husbands. The second, third, and fourth acts, in which the husbands return and become reconciled with their father-in-law, are slightly connected with the theme of the first act, though they give much less prominence to the women than we should expect. But in the fifth act all these characters disap- pear from the stage, and the rest of the play is taken up with the banquet of their slaves. The slight connection of Act Y. with the rest of the play led Goetz ^ to suspect retradatio, and Winter ^ even assimied that a later author combined two plays of Plautus to make the Stichus. But such composition is not impossible for Plautus himself."^ We remember that he was sometimes unsuccessful in combining two Greek comedies, and that he sometimes left out important scenes at the close of a play. (Cf. Cas. 641; CisL 782 ff.) So the lack of unity in the Stichus is not a sufficient reason for denying the play in its present form to him. ^'W. Siiss in Rhein. Mus. 65 (1910), pp. 452 fF., tries to prove that the Stichus accurately represents the Greek original. 'Ritschl, Parerga, p. 280. * Leo in Nachr. Gott. Ges. 1902, p. 376. ^ Goetz in Acta soc. phil. Lips. 6 (1876), pp. 302 ff. «F. Winter, Plauti Fahularum Deperditarum Fragmenta, Bonn^ 1885, pp. 82fiF. ^Cf. Leo, Plant. Forsch., pp. 150 ff.; Leo in yachr. Gott. Ges, 1902, p. 377. STICHUS 85 Such a lack of unity would, however, have been im- possible in the Greek original. Even Aristophanes makes his plays center around one or two principal characters, who take part in the riotous scenes at the end as well as in the earlier action; and the Persa of Plautus, which is probably based on an original of the Middle Comedy, preserves the unity of characters through the banquet-scene.® From all that we know of 'New Comedy, and especially of Menander, we can infer that unity of character was still more essential there. Before the discovery of the Cairo papyrus, Wilamowitz declared that Menander could never have joined humano capiti cervicam equinam; ^ and we to-day can make the state- ment even more positively. The fifth act of the Stichus, then, must contain some alteration by Plautus, and the passage (11. 419-453) which prepares for Act V., was probably original with him.^^ We notice that, as the text stands, Stichus remains on the stage after he has been dismissed, and Epignomus waits awkwardly through the entire monologue of Stichus (11. 436-453). If 11. 419-453 were cut out, the transition would be perfectly easy, and Epignomus would be on the stage for the begin- ning of the next scene. In other words, 11. 419-453 are a necessary preliminary to Act V. as it stands, but would be quite superfluous in a play which did not end with a merry-making among slaves. But while we recognize the faulty construction of the play, we are not justified in assuming that Act V. was original with Plautus, ^^ or even that its presence here 'Leo in Naehr. Gott. Ges. 1902, pp. 376 f. •Wilamowitz in Neue Jahrh. 3 (1899), p. 516. "Leo, Plant. Forsch., p. 152; Nach/r. Gott. Ges. 1902, p. 383. "Siiss in Rhein. Mus. 65 (1910), p. 453, notes the large number of Greek details in this act. 86 RETEACTATIO IN PLAUTUS STICHUS 87 is due to contaminatio. Leo has dissected the Siichus ^^ and found in it material from three Greek plays: (A) a play on the theme of the faithful wives; (B) a play with a parasite as its central figure; (C) a play furnish- ing material for the banquet-scene. It is improbable that Plautus used so many different sources as this. The more likely theory is that a single original, the Adelphoe of Menander, is the basis of the Siichus, but that the plot has been disturbed by omissions, alterations, and additions. ^^ In particular, Plautus seems to have changed the last act, perhaps, as Teuffel suggested,^^ sub- stituting a slaves' banquet ^^ for the masters' banquet in the original play. The Stichus is unique in showing evidences of re- tradatio in the names of the characters. The elder of the two sisters appears in the text of both A and P (11. 247, 331) as Panegyris, and the same name occurs in the scene-heading of II., 2, in A and in P, and in the heading of L, 1 in P. A, on the contrary, gives her name as Philumena in the scene-heading of I., 1. The name of the other sister does not occur in the text, but is given in the scene-heading of I., 1 as PampJiila in A, and as Pinacium in P. A recent study of the scene-headings in the manuscripts of Plautus ^^ has made it evident that "Leo in l^achr. Gott Ges. 1902, pp. 381 ff. "Schanz, Rom. Literaturgesch. I., l^ p- 91. "Teuffel in Rhein. Mus. 8 (1853), pp. 39 f. (= Stud, u. Char.,^ p. 342). ^The slaves' banquet followed Greek, not Roman custom. Cf. 11. 446-448. "H. W. Prescott in Harvard Studies 9 (1898), pp. 102-108. Cf. Lindsay, A7ig. Edd., pp. 102 f. r the names of the characters as they appear in the scene- headings, do not belong to the direct tradition of the Palatine manuscripts, but were at some period filled in from the text. So the form Pinacium (which is appar- ently due to a misunderstanding of 1. 284) has no authority, and the only evidence to be considered is that of P and A in the text, and of A in the scene-headings. From this evidence we may infer that Panegyris was the original name of the elder sister, and that the change to Philumena was made for a later production. As to PampJiila^ there may be some doubt. Since this sister is nowhere named in the text, there is no direct evidence against the name, but it is open to suspicion, because it appears in company with Philumena in A. PAN. Credo ego miseram fuisse Penelopani, Soror, suo ex animo, quae tam diu uidua Viro suo caruit: nam nos eius animum 2" De n5stris factis noscimus, quartim uiri hinc apsunt, tQuorfimque nos negotiis aps^ntum, ita ut aequomst, SolHcitae noct^s et dies, soror, sumus semper. 5 PA. Nostrum oflficium nos faeere aequomst: Neque Id magis facimus quam n6s mo net pietas. Sed hie, m6a soror, adsidedum: multa nolo tecum 7** Loqul de re tuiri Salu^ne, amabo? PA. Spero quidem et uol6. sed hoc, soror, crticior: Patr^m tuom metimque adeo, unice qui Onus 10. 11 Ciufbus ex omnibtis probus perhib^tur, E(im nunc inprobl uiri officio titi, 13. 14 Virls qui tantas aps^ntibus nostris 15 Facit fniurias inm^rito Nosque ab eis abduc^re uolt. Haec r6s uitae me, soror, saturant, Haec mlhi diuidiae et s^nio sunt. PA Ne lacruma, soror, neu tfio id animo 20 Fac qu6d tibi [tuos] pater facers minatur. Spes ^st eum melius f^cturum. 88 BETRACTATIO IN PLAUTUS STICHUS 89 25 30 Noui 6go ilium: ioculo ista^ dicit: Neque ill6 sibi mereat P^rsarum Montis qui esse aurei p6rhibentur, Vt i«t(ic faciat quod td metuis. Tamen si faciat minume Irasci Decet: n6que id immerito eu^niet. Nam uiri nostri domo ut iibierunt, Hie t6rtius annus — Ita fit memora^. Quom ipsi Interea uiufint, ualeant, Vbi sint, quid agant, ecquld agant, Neque participant nos n^ue redeunt. An id d6les, soror, quia illi suom 6fficium Non e6lunt, quom tu tuom facis? PAN. Ita pol. 35. 36 PA. Tace sis: caue sis audlara ego istuc Posthac ex te. PAN. Nam quid iam? PA. Quia pol meo animo omnis silpientis Suom offleium aequomst col ere 6t facere. 40 Quam ob rem 6go te hoc, soror, tam etsfs maior, Moneo (it tuom memineris officium: Et si Illi improbi sint dtque aliter tNos fficiant quam aequomst, tam pol tNequid magis sit omnibus obnixe opibus 45 Nostrum 6fficium meminlsse decet. PAN. Placet: tficeo. PA. At memineris fScito, ( Nolo 6go, soror, me cr^di esse inmemor^m uiri: Neque ille e6s honores mlhi quos habuit p6rdidit. Nam p6l mihi grata acc6ptaque huiust benlgnitaa: 50 Et m6 quidem haec condlcio nunc non pa^nitet Nequ^st quor [non] studeam has nQptias mutarier. VerCm postremo in patris potestat#st situm: Faciendum id nobis qu6d parentes Imperant. Scio atque in cogitando maerore atigeor: 55 Nam pr6pe modum iam ost^ndit suam sent^ntiam. Igitlir quaeramus n6bi9 quid facto (isus sit.) Jf8-57, This passage was one of the first in Plautus to be suspected of dittography. It gives, in briefer form and in dialogue-verse, the substance of the preceding canticum. It must therefore be considered a variant for the lyrical passage, probably introduced in order to dispense with the musical accompaniment. The author seems to have contented himself with presenting the gen- eral situation of 11. 1-47, without attempting to explain it in detail (there is no direct statement, e. g., of the father's plan to give his daughters in marriage again). On the other hand, he has borrowed the idea of 11. 53, 57 from the following scene (11. 68 ff.). A omits 48-57. Prlncipium ego quo pficto cum illis ficcipiam, id rati6- cinor : 75 Vtrum ego perplexim lacessam ordtione ad htinc modum, Quasi numquam quicquam adeo adsimulem, an qudsi quid in- daudluerim Efis in se merulsse culpam: an potius temptem Igniter An minacit^r. scio litis f6re: ego meas noui 6ptume. SI manere hie s6se malint p6tius quam alio ntibere, 80 N6n faciam: quid mi 6pust decurso aetatis spatio cfim ei8 G^rere bellum, qu6m nil quam ob rem id fdciam me ruisse firbitror ? ^yilnume: nolo tflrbas. sed hoc mihi 6ptumum factu drbitror, Sic faciam: adsimuUbo quasi quam clilpiam in sese admlserint: P4rplexabilit6r earum hodie p^rpauefaoiam p6ctora. 85 P6stid tagam igitur delude ut animus m^us erit faciam palam. Mfllta scio faciflnda uerba: ibo Intro, sed apertast foris. 75-83, The order of these verses is confused in P, and the whole passage is full of difficulties. Langen ^"^ notes the peculiar use of perplexim (1. 76), which must refer only to quasi quid indaudiverim Eas in se meruisse culpam (11. 77 f.), and not to quad numquam quicquam adsimulem (1. 77) ; the unusual construction eas in se meruisse culpam (1. 78) ; the position of potius in the first instead of the second alternative clause (11. 78 if.) ; "Langen, Beitrage zur Kritik u. Erkldrung des Plautus, Leipzig, 1880, pp. 147 ff. 90 KETRACTATIO IN PLAUTUS STICHUS (>1 and the adverb minaciter (1. 79), not found elsewhere in Plautus. Langen rejected altogether about one-third of 11. 75-79, and rearranged the rest; Leo ^^ thought that 1. 79 was an interpolation, and leniter (1. 78) a corruption for saeviter. The large number of repetitions from the verses imme- diately preceding and following is noticeable also; 76 perplexim (a very rare form), cf. 85 perplexahiUter ; 77 adsimulem, cf. 84 adsimulaho; 77 quasi quid indau- diverim Eas in se meruisse culpam, cf. 84 quasi quam culpam in sese admiserint; 79 ego meas novi optume, cf. 73 novi ego nostros; 81 non faciam, cf. 84 sic faciam. The content of the verses does not in itself suggest re- tradatio, but the confusion in the manuscripts, combined with the many irregularities in construction, makes one suspect that 11. 75-83 (or at least 75-79) are by a later hand. A has the verses in the order of the Goetz-Schoell text. P has the order: 80-83; 75-79; 70-74. GE. Famem ego fuisse stispicor matr^m mihi: Nam postquam natus sfim, satur numquam fui. Neque qiifsquam melius r^feret matri grStiam, (Quam ego matri mea6 refero inuitlssumus. ) Neque r^ttulit quam ego refero meae matri Fami. 155 158«' 157 ff. There are here, as Seyffert recognized,^^ two versions : (1) 157 Neque quisquam melius referet matri gratiam 158" Quam ego matri meae refero invitissimus; i^Leo, yachr. Gott. Ges. 1895, p. 420, n. 3; Ed. (189C). ad loc; 'S/(whr. Gott. Ges. 1902, p. 377. ^"Seyifert, Studia Plautina, Berlin, 1874, p. 11, n. 10. ¥ (2) 157 Neque quisquam melius referet matri gratiam 158'' Neque rettulit quam ego refero meae matri Fami. Of the two forms of the second line, 158'' has most often been taken as the genuine. It is less awkward than 158*, and the citation by Charisius of fami in the dative, from the Stichus of Plautus, proves that 158*^ occurred in his sources, and that Charisius himself, writing in the fourth century A. D., regarded it as genuine. A has 157, 158" — i. e., preserves the first line of the couplet, and the substitute verse for the second. P omits 157, but has 158", 158** — i. e., omits the first line of the couplet, but preserves both versions of the second line. Nam ilia m^ in aluo menses gestauft decem: At 6go illam in aluo g^sto plus ann6s decem. Atque flla puerum m^ gestauit pSruolum, Quo minus laboris c^pisse illam existumo: Ego n6n pausillulam In utero gest6 famem, Verum h6rcle multo mdximiam et graulssumam. Vterl dolores tmihi oboriuntur cotldie: Sed mStrem parere n^queo nee quid agSm scio. 159 160 165 160 ff. The position of 11. 165-166, between 11. 160, 161 in A, may perhaps indicate a shortening by the omission of 11. 161-164. ^^ A places 165, 166 after 160. P has the verses in proper order. tAuditaui sa^pe hoc uolgo dicier, Sol ere elephantum gntuidam perpetuus decem Esse ^nnos : eius ex s^mine haec certost fames : Nam idm comp lures iinnos utero haer^t meo. 167 170 167-170, The parasite's account of himself and his 20 Lindsay, Anc. Edd., p. 55. 92 RETEACTATIO IN PLAUTUS mother comes to a suitable close in 1. 166. Then 11. 167- 170 add a new and hardly consistent thought, and 1. 170 is particularly objectionable because complures annos repeats, in weaker form, the plus annos decern of 1. 160. It is possible, therefore, that these verses are a later in- sertion,2i made in order to expand the coarse wit of the passage. The passage appears in both A and P. Geiasimo nomen mlhi indidit paru6 pater, 174 Quia inde iam a pausillo pliero ridicultis fui. 175 Propter pauperiem hoc adeo nomen rgpperi, Eo quia paupertas f6cit ridicultis forem: Nam ilia artis omnis p^rdocet, ubi quem &ttigit. 17Jfif. Gelasimus gives two explanations for his name: (1) that his father gave it to him because he was a droll child; (2) that he received the name because poverty taught him to be witty. A connection between these two thoughts,-^ though possible, is rather strained, and it seems more likely that we have here two parallel versions: (1) 11. 174-175; (2) 11. 176-178. A and P both contain the whole passage, in the order: 174, 176, 175, 177, 178. Dicam atietionis catasam, ut damno gafideant — Nam cfiriosus n6most quin sit mftleuolus — : [Ipse ^gomet quam ob rem | atictionem pra6dicem:] Damna 6uenerunt mSxuma miser6 mihi. 207 208" 208'', The verse is impossible after dicam auctionis =^Langen, Plant. Stud., p. 372; Leo, Nsm

erd cenam meat, Adu6rsitores p6l cum uerberibtis decet Dari, (iti eum uerberfibundum abducent domum. Parfita res faciiam tit sit. egomet m6 moror. Atque Id ne uos roir^mini, hominis s^ruolos Potare, amare atque Sd cenam condlcere: Licet ha6c Athenis n6bis. sed quom c6gito, Potitis quam inuidiam inu6niam, est etiam hie 6stium Alitid posticum n6strarum harunc aMium: [Postfcam partem mfigis utuntur a^dium.] Ea ibo 6bsonatum atque eddem referam ops6nium: Per h6rtum utroque c6mmeatus cCntinet. Ite hac secundum u6smet: ego hunc lacer6 diem. Jf-Jfl-Jf-Jf-B. The end of the scene is unduly protracted, and Stichus announces his departure three times (11. 440, 450"^ 451 23 Lindsay, Anc. Edd., p. 55; Ed. (1905), ad loc. 94 EETRACTATIO IN PLAUTTJS 445 453) before he finally leaves the stage. His impa- tience with the delay of Sagarinus, too, is quite out of place here. Baier ^^ and Leo ^^ are probably right m thinking that this motif was taken over from 11. 641- 648 where it is much more appropriate, and that 11. 441-445 were inserted at the time of the later production. P omits 441-445. 483 485 GE. Quando quidem tu ad in6 non uis promlttere, (Sed qiioniam nil process! sat ego hac, lero Ap^rtiore magis uia: ita plan6 loquar.) i8S-Jf85, The first of these three lines (483), begin- ning quando quidem, and the last two (484-485), be- ginning sed quoniam, are undoubtedly parallel versions. Of the two versions, the second (11. 484-485) is the more subtle, and therefore probably the genuine. P omits 484-485. ST. (Proin tu lauare prSpera. SA. Lautus sum. ST. 6p- tume : 668 Sequere ^rgo | hac me | Intro. SA. Ego uer6 sequor.) Volo ^luamus h6die: peregrina 6mnia ' Rellnque: Athenas ntinc colamus: s^uere me. Sequor 6t domum redetindi principidm placet: Bona sca6ua strenaque Cbuiam occessit mihi. 668 ff. The close of this scene seems to have been shortened by the substitution of 11. 668-669 for 11. 670-673. P has the verses in the order of the Goetz-Schoell text. A deest 1 sheet = 38 11. =* Baier, De Plauti Fah. Recems., pp. 123 f. =* Leo in ^Sachr. Gott. Ges. 1902, p. 379. STICHUS 95 A deest 648-681 = 34 11. -f 2 scene-headings (=4 11.). Total, 38 11. Therefore A probably had 668, 669. Tlbi propino. d^ciimum a fonte tibi tute inde, si sa- pis. 708 B^ne uos: bene nos: b^ne te: bene me: b^ne nostram etiam St6- phanium. Blbe[s], si bibis. ST. Non mora erit apud me. SA. 6depol conuiui sat est: 710 M6do nostra hue amfca accedat: fd abest, aliud nil abest. L^pide hoc actumst. tfbi propino canthanim. Vi- ndm tu habes: Nlmis uellem laliquid ptilpamenti. ST. Si h6rum quae adsunt pa^nitet, Nil est. tene aquam. SA. Melius dicis: nil moror cuppMia. Bibe, tibicen: ^ge siquid agis: bfbendum hercle hoc est: n6 nega. 715 i '10-711. These verses disturb the connection equally here and in the position to which Ritschl transposed them (after 1. 735). Langen^^ noted also that the use of mora (1. 710) in the sense of " delay " was un- Plautine. The couplet seems to be introduced for the purpose of shortening the scene. ^"^ P has the verses in the order of the Goetz-Schoell tezt. A deest 709 to end of play. The Stichus shows no extensive changes due to re- tractatio, but there are traces of slight alterations all through the play, from the name of the elder sister at the beginning to a proposed shortening of V., 4. The changes seem to have affected especially the third scene of the first act. There are a number of passages showing parallel versions, the second version in one case (11. 48- 57) evidently being intended to dispense with musical ^Langen, Beitrdge, pp. 171 ff. "Leo, Ed. (1896), ad. loc; Nachr. Gott. Ges., 1902, p. 378. 96 EETEACTATIO IN PLATJTUS accompaniment. A shortened version of the close of \ ., 2 is given, and a couplet (U. 710-711) is inserted to shorten V., 4. On the other hand, the revisers seem to have made some trivial additions to the original thought (11. 167-170; 441-445). 2» The evidence of the Stichm is particularly valuable because we can consult both families of manuscripts for nearly all the questionable passages. Only in the case of 11. 710, 711 is it absolutely impossible to tell v^hat A contained; for calculation makes it probable that 11. 668, 669 appeared on a lost page of A. This passage, then, would belong in the same class with 11. 174- 178 and with a less certain case of retradaho (11. 167-170), where A and P have exactly the same amount of text. Both P and A have 11. 75-83, though the lines appear in different order; and the mixture m 11. 157 ff. indicates that both the Plautine and the substitute version were at one period represented in both families of manu- scripts. There is one place (U. 48-57) in which P gives a second version not preserved in A, and several cases (11 441-445 ; 483-485 ; 160 ff. ; 232 f.) in which A shows more evidence of retradatio than P. The evidence of the SUchus, therefore, is decidedly against the theory that A is the purer text. 28 - In addition to the pa^ges discussed in detail, the following lines have been suspected: 84, 118-120, 12M25, 135, 179-180, 22o, 294, 321, 330, 387, 425-435, 427-429, 450^ 473-482, 535, 555, 590- 59l! 681, 684, 746-747. _ ^,a-±-. Of these lines, the evidence of A is lacking for doo, 681. '46-^4^, A omits 450^ P omits 387, 427-429, 535, 590-591; A and P both have 84 118-120, 121-125, 135, 179-180, 225, 294, 321, 330, 42o- 435 (except that P omits 427-429), 473-482, 684. The following lines show variations of a word or phrase: ^6 f., 90, 163, 166, 189, 202, 237, 253 f., 255, 262 f., 282, 342, 350, 373, 374-376,' 390 f., 586, 594, 632 f., 640, 688. CHAPTER V. TRINUMMUS The Trinumnius of Plautus is translated, so the pro- logue tells lis, from the ^rjcravpo^ of Philemon. It is a comedy without female parts (except for the two abstractions who speak the prologue), and without the erotic element which is so prominent in other plays of Plautus. When the play opens, the old man Charmides has gone off on a business voyage, leaving his daughter and his dissolute son in the care of his friend Callicles, with special instructions that a treasure buried in his house be kept intact for the daughter's dowry. In the meantime, the son, Lesbonicus, goes from bad to worse, and finally advertises his father's house for sale. In order to preserve the treasure without betraying the secret, Callicles buys the house. Lysiteles, a young man of exemplary character and good family, and a devoted friend of Lesbonicus, now sues for the daughter's hand, and Callicles feels in duty bound to produce the treasure. So a rogue is hired to play the part of a messenger from Charmides, to bring forged letters for Lesbonicus and Callicles, and a sum of money to serve as the daughter's dowry. Unfortunately for the success of the scheme, Charmides arrives unexpectedly, and meets the supposed messenger in front of his own house. But matters are explained, Lysiteles receives both bride and dower, and Lesbonicus is pardoned, on condition that he take the daughter of Callicles as his wife. 97 98 KETKACTATIO IN PLAUTUS CA. Quid uenis? ^7 ME. Malls te ut uerbis mliltis multum obitirigem. CA. Men? I^IE. Namquis est hie filius praeter me iitque te? CA. Xem6st. ME. Quid tu igitur r6gita« tene obitirigem? 70 Nisi t(i me miMmet censes dicturfim male. Nam si In te aegrotant dries antiqua6 tuae 72* [Sin Immutare uls ingenium m6ribus] 72 (Aut si demutant mores ingenilim tuom Neque eos antiquos s^ruas, ast captds nouos,) Omnibus amicis morbum tu incuti^s grauem, 75 Vt te uidere audlreque aegroti sient. 12 if. Most editors have taken 1. 72' as an explanation of 1. 73, or an adscript parallel to it. But 1. 73 is per- fectly clear without explanation, and 1. 72*" is too closely related to it in thought and phrasing to be merely an accidental parallel. We are therefore led to suspect the hand of the retraciator. The manuscript-reading sin can not stand, since the strongly adversative idea which sin demands is lacking ; and KitschFs ^ emendation sive is therefore probably to be accepted. But Kitschl him- self observed that the Plautine conjunction was not sive, but aut d (the form which we actually find in 1. 74), and this fact supports the other evidence against the genuineness of the line. The next two lines (73-74) have been regarded as due to dittography. But they can not be simply an alternative version of 1. 72*, for the first words (aut si) are impossible at the beginning of a sentence. Kitschl noted the irregular use of the word mores — in the sense of "temperament " instead of " the (proverbially corrupt) morals of the day,'' as elsewhere in Plautus (cf. 11. 28 fF., 1037, 1045, etc.). But this criticism applies only to 1. 74. Without this addition, ^Ritschl, De Inter polatione Trinummi Plautinae, Bonn, 1844 {= Parerga, pp. 513ff.)« TRINUMMUS 99 •1. 73 is unobjectionable; the conjunction aut si is Plau- tine, and mores has its customary meaning. Therefore it seems probable that 11. 72," 74 were inserted as a more emphatic substitute for 1. 73. P has the whole passage. A omits 72^. Nil 6st profecto sttiltius neque st6lidius Neque m^ndaciloquiGs neque arguttim magis Neque c6nfidentil6quius neque peiifirius Quam urb^ni adsidui clues quos scurrds uocant. 200 200, One would not object so much to the repetition in this passage, were it not that mendaci-loquius antici- pates the compound confidenti-loquius in the next verse, and that the circumlocution with magis interrupts the series of simple comparatives. ^ It seems possible that the verse was intended as a substitute for 1. 201. The line occurs in both A and P, but in A has the reading adeo argutum, in P argutum magis. Qui homo cum animo inde Sb ineunte aetate depugnfit sue, 305 Vtrum itane esse mfiuelit ut eum finimus aequom c^nseat, An ita potius fit parentis 6um esse et oognatl uelint: Si finimus hominem p^pulit, actumst, animo seruit, n6n sibi: Si Ipse animum pepullt, dum uiuit, ulctor uictorflm cluet. Tfi si animum uiclsti potius quam finimus te, est quod gati- deas. 310 Nlmio satiust tit opust te ita 6sse quam ut anim6 lubet. ( Qui Snimum uincunt qufim quos animus semper probior^s cluent. ) SO 5-8 12, Philto is delivering a sermon on the text, ^J. Brix, Ed. (1879), ad loc. 100 EETEACTATIO IN PLATJTTTS "Better is he that ruleth his spirit." He generalizes broadly (11. 305-309) and then makes a personal appli- cation to the case of Lysiteles (1. 310). After this he crives a weaker turn to the last statement (1. 311), and then returns to generalization (1. 312), repeating mnch of the phrasing of 1. SIO.^' Bergk * was probably right in thinking that the last two lines were an alternative version for 11. 305-310. The passage has the same form in A and P. [PH.] Is probust, quern pa^nitet quam pr6bus sit et frugt bonae : oZ\f Qui ipsus sibi satis pla^^et, nee pr6bus est neo frugt bonae: Qui Ipsus se cont^mnit, in eost Indoles industriae: Benefacta benefSctis aliis p^rtegito, ne p^rpluant. S22, Without 1. 322, the speech of Philto makes a very neat antithesis (11. 320-321), followed by a line of practical application (1. 323). The line which inter- venes (322) merely repeats 1. 320, and is especially disturbing because it returns to the first half of the con- trast after the second is finished.^ The line occurs in both A and P. PH. Qu6i[us] egestat^m tolerare uls? loquere audact^r patri. 358 L^sbonico huic adulescenti, Charmidi filio, Oui lllic habitat. PH. Quin comedit quod fuit, quod n6n ^ fuit? 360 /LY.> Ne opprobra, pater: mfilta eueniunt hdmini quae uolt, ^ qua6 neuolt. /PHS M^ntire edepol, gnSte, atque id nunc fdcis baud consue- ^ '^ ttidine. ^Ritschl, Parerg., pp. 522 AT. ^ Bergk in Zeitschr. f. Alt. 1848, coll. 1137 f. ( ^Langen, Plant. Stud., pp. 374 f. TEINUMMUS 101 Nam sapiens quid^m pol ipsus fingit fortunam sibi: E6 non multa qua6 neuolt eueniunt, nisi fictor malust. Mtilta illi opera optist ficturae qui se fictor^m probum 365 Vltae agundae esse ^xpetit: sed hie ^dmodum adulesc^ntulust. Non aetate, u6rum ingenio aplseitur sapi^ntia. S^pienti aetas condimentum, tsSpiens aetati cibust. Agedum eloquere, quid dare illi ntinc uis? LY. Nil quicquflm, pater: TO modo ne me prohibeas acclpere, siquid d6t mihi. 370 361 if. The insertion of 1. 369 after 1. 361 probably indicates the omission of the moralizing in 11. 362-368.^ A inserts 369 after 361. P inserts 369, 368 after 361. = Opiisc. I., p. 17) PH. I hac, L^sbonice, m4cum, ut coram nflptiis 580 Dies c6nstituiatur : eadem haec confirm^bimus. tTu istuc cura quod iu^i: ego iam hie ero. Die C^llicli me ut c6nueniat. ST. Quin tu 1 modo. LE. De dote ut uideat quid pus sit facto. ST. I modo. Nam c^rtumst sine dote ha<;G>d dare. ST. Quin tu I modo. 585 LE. Neque enim Illi damno umquam 6sse patiar — ST. Abi modo. Meam n^glegentiam. ST. f modo to pater LE. Aequ6m uidetur quin quod peccarim — ST. f modo. LE. Potlssumum mihi id 6psit. ST. I modo. LE. 6 pater, Entimquam aspiciam te? ST. I modo, 1 modo, I modo. 582. After Lesbonicus has finally uttered the long- delayed spondeo which betroths his sister to Lysiteles, and Philto has left the stage, Lesbonicus remains and resumes the discussion of the dowry — to the great disgust of Stasimus. It is possible that the audience may have grown impatient too, and that consequently 1. 582 was •Lindsay in Amer. Journ. Phil. 21 (1900), p. 27; cf. Anc. Edd., p. 47; Ed. (1905), ad loc. m 102 EETKACTATIO IN PLAUTUS substituted for the original ending of the scene (11. 583-601)/ A deest 568-636. 668 67a Itast amor balUsta ut iacitur: nil sic celerest n^que uolat: Atque is mores h6minum moros 6t morosos ^flficit. Minus placet magis qu6d suadetur : qu6d dissuadettir placet. Quom Inopiast, cupifls: quando eius copiast, turn non uelis. (Ille qui aspellit, Is compellit: Ille qui consuad6t, uetat.) 672, The suggestion of Bergk,^ that 1. 672 is a second version of 1. 670, has been followed by most editors. Except for the fact that 1. 672 has a personal subject, the thought of the two verses is the same, and ille (1. 672) must twice be scanned either tile or lll\^ A deest 672-735. ME. Homo c6nducatur aliquis iam quanttim potest. 765 [Quasi sit peregrinus. CA. Quid is scit facere p6stea?] Is homo 6xornetur grSphice in peregrintim modum: tlgn6ta facies qua6 non uisitata sit (Menddcilocum aliquem. CA. Quid is [i]scit facere p6stea?) tFalsIdicum, confid6ntem. CA. Quid tum p6stea? 770 765 if, Brix ^^ was probably right in bracketing 1. 766 as an interpolation. Quasi sit peregrinus seems to be merely an explanation of 1. 767 in peregrinum modum, and quid is scit facere postea? is apparently borrowed from 1. 769 to fill out the line. But 1. 769 presents a reasonably certain case of reiractatio. The objections *Leo, Ed. (1896), ad loc. "Bergk in Zeitschr. f. Alt. 1848, col. 1141 ( =Opusc. i., pp. 20 f.). »Cf. Langen, Plant. 8tud.. p. 376; Niemeyer, Ed. (1907), ad loc. "Brix, Ed. (1879), ad loc. (Brix numbers 770). TRINUMMUS 103 which Brix makes to quid is scit facere postea? in 1. 766 hold equally for it here: the question could properly be asked only when the man had already been found and his ability to carry out the scheme was under discussion. (See Pseud. 745 for an instance of scit properly used in a similar situation.) Plautus uses mendaci-locus only here and in Trin. 200 ^^ (where, as we have seen, it is also suspicious). The meaning of the word is exactly the same as that of falsi-dicus, in the next line, and it is possible that the new compound was introduced in both places (11. 200, 769) for the sake of novelty. P omits 769. [CA.] Sed epistulas quando opsignatas adferet, [Sed quom obsignatas attulerit epistulas] Nonne firbitraris ttim adulescentem finuli Pat^rni signum n6sse? 788* 788*' )ab 788''°, The two lines are unquestionably variants, the second giving the idea " sealed " a little more emphati- cally. The scansion attulerlt may perhaps be explained as the lengthening of a short syllable before the final metrum of the line,^^ but it is at all events sufficiently rare to throw suspicion on the line.-^^ A deest 774-834. ME. In hulus modi negotio Di^m sermone[m] t6r[r]ere segniti^s merast: Quamuls sermones p6ssunt longi t^xier. Abi il then[a]saurum idm confestim clSnculum. 795 " The statement is based on a collection of the adjectives in Plau- tus made by the Latin Seminary of Bryn Mawr College, 1907-1908. "Cf. Lindsay, Ed. Capt. (London, 1900), Introd., p. 42. "Lindsay, Anc. Edd., p. 47; Ed. (1905), ad loc. 104 EETRACTATIO IN PLAUTUS TRINUMMUS 105 796, 797, Here again we have two lines which are very similar in meaning. Megaronides might be allowed to repeat himself if he were not at that very moment doing his best to put a stop to the conversation ; but as it is, it seems probable that 1. 796 is a later version, which borrows much of its phrasing from 11. 806 f. A deest. the later version must have crowded out the original form of the line. CH. Faciam ita ut uis: Sgedum, nomen tu6m primum memory mihi. 883 Magnum f acinus incipissis p6tere. CH. Quid ita? SY. Quia, p(at<(e>r[em], Si ante lucem tire occipias k meo primo nomine, 885 Coneubinm sit noctis prius quam ad p6stremum peru^neris. CH. 6pug tfactost et uidtico ad tuoni nomen, ut tu pra^dieas. 6st minusculum alterum quasi tiuxillum uinSrium. Qufd est tibi nom6n, adulescens? SY. Pfix, id est nom^n mihi: Hoc cotidianust. CH. Edepol n6men nugat6rium: 890 Quasi dicas, siqufd erediderim tfbi, ' pax ' periisse Ilico. 889-891. These three verses appear in P after 1. 937. Meier ^^ transposed them to their present position, and altered the reading of 1. 889 to quid illud est nomen? The order of P is manifestly impossible; the inquiry about the name must follow directly after 1. 882, before Charmides goes on to ask about the sycophant's facta et itinera (11. 893 ff.). But if 11. 889-891 are preceded by 11. 883-888, the question in 1. 889 should be, not quid est tihi nomen f but quid est alterum nomen? It is possible that the stage-manager of the Eevival used only 11. 889- 891, substituting quid est tibi nomen? for a question about the minusculum alterum}^ If this was the case, "M. H. E. Meier, Commentatio de Plauti Trinummo, Halle, 1845, pp. 7 f. ^'^Cf. Leo, Ed. (1896), ad loc. P has 889-891 after 937. A deest 864-1044. •SY. Hanc me iussit L§sbonico suo gnato dare eplstulam 898 6t item banc alteram suo amico CSllicli iussit dare. CH. Mihi quoque edepol, quom hlc nug[ur]atur, c6ntra nugarl lubet. 900 ^bi ipse erat? SY. Bene r6m gerebat. CH. :6rgo ubi? SY. In Seleucia. CH. Ab ipson istas Sccepisti? SY. E mSnibus dedit mi ipse in manus. 901, The question and answer of 1. 901 anticipate the long dialogue (11. 928-947) in which Charmides asks his own whereabouts. The shorter answer is really the correct one (cf. 11. 112, 771), but is for that very reason the less likely in the mouth of the sycophant. It is im- probable that Charmides would ask the question at 1. 901, and then devote so much time to it later, or that he would fail to comment on the inconsistency of the sycophant's two answers. It is much more likely that 1. 901 repre- sents another part of the same shortened version which we find in 11. 889-891. A deest. CA. Quid hoc hlc clamoris afldio ante aedis meas? CH. O Collides, o Collides, o Cdllicles, Quallne amico m6a commendaul bona? CA. Probo ^t fideli et fido et cum magnS fide: Et sSlue et saluom te Sduenisse gjatideo. CH. Credo 6mnia istaec, si | itast ut pra^dicae. Sed qufs istest tuos ornfitus? CA. Ego dicfim tibi: Tliensafirum effodiebam fntus dotem filiae 1093 1095 1100 106 KETRACTATIO ITT PLAUTUS Tuae qua6 daretur. s$d intus narrab6 tibi Et h6c etalia: s^quere. OH. Stasime. ST. Hem. CH. Str^nue Curre In Piraeu[u]m atque tinum curriculum fa«e. Vid6bis iam illic n^uem qua aduecti sumus. lub^to Sa[n]gari6nem quae imperauerim HOo Curare ut efferlintur, et tu it6 simul. Sol(itust portit6ri iam port6rium. Nil €st mora. cit ambula: a<;tuttim redi. ST. Illlc sum atque hie sum. CA. S^quere tu baxj me intro. CH. Sequor. 109S ff. The rapidity with which this scene draws to a close is equal to that of certain scenes in the Persa. For the audience, to be sure, no explanation of the motives of Callicles was necessary, but it seems incredible that Charmides should rest satisfied with the simple assurance of 11. 1096-1097. Leo ^« is probably right in thinking that P gives only a shortened version of the scene, although we cannot assume, as Ritschl ^"^ did, that the passage had any fuller form in A.^^ A deest 1079 to end of play. ST. Hie m6o ero amicu 86lus firmus r^stitit 1110 Neque d^mutauit Snimum de firmfi fide, Quamqufim labores mtiltos * ♦ ♦ * ♦ Sed hie tinus ut ego sfispicor, seruSt fidem. tOb rem laborem eum 6go cepisse c^nseo. 1110 ff. There seem to be traces of two versions here: (1) 1110-1112; (2) 1113-1114. As far as we can tell from the mutilated text, the sense of the two passages was about the same, and some of the phrasing is repeated. "Leo, Ed. (1896), ad loc. "Ritschl, Ed. (1848), Praef., pp. xxv f . *" Studemund's Apog., note on fol. 464v. TKINUMMUS 107 A deest. B marks a lacuna after multos (1112). The abundance of sententiae and moral reflections in the Trinummus makes the play a difficult one for the student of retractatio, A sententious line would natur- ally invite every later poet to try to turn the phrases a little more neatly. On the other hand, even the origi- nal author might be open to the same temptation, and might add to a pithy sentence another in slightly differ- ent form. Indeed, wordiness and repetition are so char- acteristic of the moralizing style that it is frequently impossible to say whether a given line is an extension by Plautus himself or by a later author. The suspicious passages are extremely limited in ex- tent, in no case covering more than half a dozen lines, and generally not more than one or two. The majority are concentrated in certain scenes, especially II., 2 and III., 3. Of the possible variant lines, most are of the moralizing type: 200, 311-312, 322, 672. Dittography also seems to be present in 11. 796 f., 1110 ff. ; and almost certainly exists in 11. 72 if., 769 f., 788"'. The transpo- sition of 1. 369 seems to indicate the omission of a passage, and 11. 582, 889-891, 901 are probably intended to furnish substitutes for lengthy scenes. It is probable that at 11. 1093 if. only the shortened version is pre- served.^^ "The following lines have also been suspected: 6-7, 18-21, 60, 64, 92, 93, 126, 206-209, 223 ff., 231-232, 248-249, 263, 321, 368, 414-415, 420-424, 427^ 470, 471, 527-528, 562-568, 587-589, 660 ff., 702, 707- 708, 756-762, 764, 792, 808-814, 816, 831, 852, 857-860, 872, 929 ff., 980, 982, 1005, 1033, 1043-1045, 1053-1054, 1130-1131, 1164-1166. The evidence of A is lacking for 126, 587-589, 702, 707-708, 792, 808-814, 816, 831, 872, 929 ff., 980, 982, 1005, 1033, 1043-1045, 1130- IQ^ KETKACTxYTIO IN PLATJTUS Unfortunately the Trinummns affords slight opportu- nity for a comparison of the manuscript-tradition. In several of the most certain cases of retractaUo, sheet, are missing from the Palimpsest, and it is impossible to calculat: their contents. In most of the o hers, the reading of the two families of manuscripts is the same. At 11. 361 fi., where both A and P indicate an omission, P has -transposed one more line than A. A omits 1. 7. , though it gives the remainder of the suspicious passage, and P omits 1. 769, which is almost certainly a later addition. But on the whole, where we can compare A and P, the testimony of the Trinummns strongly supports the theory that the two families of manuscripts had a common origin. 1131 1164-1166. The rest of the passages are preserved in both i and P 6-7 18-21, 60, 64, 92, 93, 206-209, 223 ff., 231-232 248- 249 '>63 321 (preserved in B, but omitted in the other Palatine Ll'scripts). 368 (in different order), 414-415, 420-4f , 427Mm different order), 470, 471, 527-528, 562-568 (except that Ade^t 568ff.), 660ff., 756-762, 764 (in different order), 852, 857-860, 1053-1054. Q^ „,. Tlie following lines show minor variations: 52, 61, 70, l«b, Zi4, 238, 256, 328, 339, 351, 537, 660, 842, 1064, 1069, 1078. coisrcLusioN The five plays which have been discussed in detail show a marked difference both in the amount and in the kind of retractatio that they present. By far the most extensive changes, as well as the greatest number of cer- tain examples of retractatio, occur in the Poenulus. In this play we find a secondary ending of about forty lines ; two passages, each of which shows three parallel ver- sions ; and numerous others which present two parallels. The Persa seems to have suffered shortening, and to have preserved only the shortened version of certain scenes. The Stichus has several alternative versions, only one of which is of any length, and a few small additions. The changes in the Pseudolus and the Trinummus are for the most part confined to single lines. It is evident, therefore, that theories about the general problem of retractatio should be based, not on a few selected cases, or even on all the cases in a single play, but on the whole body of text. However, even the study of five plays has led to some general conclusions. It appears that, on the whole, the retractatores made no very important contributions to our text. Even in the second ending of the Poenulus they used to a large extent material that Plautus had supplied, and in other cases their debt to him was even greater. Pers. 722-734 is a patchwork of Plautine phrases, and Pseud. 406-408 is borrowed directly from the next scene. The later poets often seem to have con- tented themselves with making slight changes in phras- ing, either to improve upon the original form of a line 109 no KETEACTATIO IN PLAUTUS {Pers. 704; Trin, 788^ Sticlu 158^) or to give a more modern turn to an old phrase {Poen, 390'; Pers, 442 f.). The alternative versions sometimes have the effect of shortening the scene, and once, at least (>S^^ic/i. 48-57), there is a change in order to dispense with mu- sical accompaniment. Occasionally a transposition indi- cates that a wordy passage was to be omitted {Poen. 622^ Stich, 160 ff.; Trin. 361 ff.)- Such shortenings by means of simple omission generally occur in the body of the scene; alternative versions which are intended to shorten a scene most often occur near the end {Pers. 666 ff.; Poen, 805; Sticli, 668 ff. ; Trin, 582). The chief object of this investigation, however, has been to determine as far as possible the relation of the Ambrosian and Palatine recensions to the phenomenon of retractatio, New light has been thrown upon this question by a study of the five plays in which the Ambro- sian Palimpsest is best represented— for any solution must rest primarily upon the evidence of these plays. The discussion has taken up altogether 66 cases of retractatio.'' In 17 of these 66 cases, the evidence of the Palimpsest is absolutely lacking : Pers. 460 f . ; Poen. 98-100; 118-120; 121-128; 214-215; 217-219; 622^ 805-808; Stick. 710 f.; Trin. 582; 672; 788'^ 796- 797 ; 889-891 ; 901 ; 1093 ff. ; 1110 ff. We are therefore reduced to 49 passages on which to base our conclusions. ^This summary includes only certain or fairly probable cases of retractatio. Pers. 467-468; Poen. 930-939; Stick. 208^ are excluded, since the difficulty in these passages is probably not to be charged to retractatio, and the passages listed in foot-notes under each play are omitted as well. In the summary, the Double Ending of the Poenulm (11. 1315-1422) counts as a single ease, but two additional oases are listed from the same portion of the play (11. 1315 flf.; 1331). CONCLCJSION 111 Of these 49, we find 15 preserved (aside from slight verbal differences) in exactly the same form in A and P: Pers. 704 ; Poen. 917-929 ; 1162 ff. ; Pseud. 166 ; 210-224 ; 3851; 406-408; 688 ff. ; 1137; 1214-1216; Stick. 167- 170 ; 174 ff. ; Trin. 200 ; 305-312 ; 322. In 6 other cases it is probable that A, if preserved, would give the passage in the same form as P. These are cases in which the Palimpsest breaks off after giving part of a suspicious passage, or else, even though the text is entirely missing from A, calculation makes it probable that A had the same form as P: Pers. 440 ff. ; 738 ff. ; Poen. 504-575; 1315-1422; Pseud. 1079-1086; Stick. 668 ff. Also, at Poen. 389 and 1333 ff. (the latter passage discussed under 1315 ff.), though A has at first omitted one or more lines, part of the verse or the passage in question is added between the lines or in the margin, showing that somewhere in the A-family the passage was given in full form. Altogether, then, we find 23 cases in which the text-tradition of the A-family and the P-family is virtu- ally the same. Of the remaining 26 cases, some show differences in the amount of text preserved, others in order only, and a few differ both in amount and in order. There are, in all, 8 instances in which A shows evidences of retrac- tatio not found in P : Poen. 706 ff. (the insertion of 730 after 706, 707, 720, indicating a further omission) ; 1331 (an alternative for 1330) ; Stick. 441-445 (an addition modeled on 11. 641-648) ; Trin. 769 (a variant for 770) ; and, less certain cases: Pseud. 67" (an addition to 1. 64) ; 573" (likewise an addition) ; Stick. 160 ff. and 208 ff. (transpositions to indicate omission). On the other hand, P presents 5 cases of retractatio of which there is no 112 EETRACTATIO IN PLAUTUS trace in X: Pseud. 523^ ' and Stick. 48-57, both undoubted examples of later versions; and also Poen. 1378-1381 (a proposed shortening);^ Pseud. 696^ (an addition); and 1205-1207 (a shortening). We also find a number of instances in which both families of manuscripts give a parallel version or a short- ened scene-ending, but one or the other omits some of the lines. In such cases it must, of course, be assumed that the whole passage originally stood in that family, but that in some way (perhaps through errors due to homoeo- ieleuton or homoeokatarUon, or through other forms ol carelessness in transcribing) part of it was lost. Thus in Pers. 605-610, P omits 608 (a genuine line) and 610 (part of the later version) ; 666 ff., P omits 67f ; 722-734, P omits 730; Poen. 1042-1053, P gives 1053 only once, whereas A gives it twice; Stich. 483-485, P has only the spurious version, A has both. Against these 5 passages in which A gives the fuller form can be ranged 2 m which P is more complete: Stich. 157 ff., in which A gives only one (probably the substitute) version for the second line of the couplet, whereas P has both (but omits the first verse of the couplet) ; and Trin, 72 ff., in which A omits 72'. In 3 of the cases just mentioned (Pers. 666 ff. ; Poen. 706 ff.; 1042-1053) there is a difference in the order of the lines as well as in their number. There are also 6 passages which show the same text in A and P, arranged in different order: Poen. 300-305; the prologue of the Pseudolus (which appears in a different position in the two families) ; Trin. 361 ff. (in which P transposes 368 as well as 369) ; and, in addition to these fairly certain 2 The contents of A are calculated. CONCLUSION 113 cases, others which are less sure: Poen. 1267-1268; Pseud. 151-156; Stich. 75-83. On the whole, the difference between the two families of manuscripts in the amount of retractatio preserved is slight. In 23 cases out of 49, A and P seem to show the same text-tradition. In 6 additional cases, the only difference is in the order of the lines. A has 8 cases of retractatio which P does not give, and P has 5 which A does not give. In 5 cases, though both families of manu- scripts show traces of retractatio^, A gives more text; in 2, P has the fuller form. The few differences that exist indicate, not that A presents the " ipsa verba " of Plautus, and P the " Eevival text," but that A, as the older manuscript, has kept more of the " Revival " altera- tions than P. We are therefore forced to the conclusion that the source of A and P was the same; that the two families had originally about the same amount of re- tractatio, but that, in the course of centuries, some lines and passages have dropped out ; the Palatine manuscripts, being the later, have naturally lost more than the Am- brosian Palimpsest. When we try to account for the omission of a passage in one family of manuscripts and its transmission in another, for the confused order of half a dozen lines, or the mutilation of a substitute passage, we find the most satisfactory explanation in Oskar Seyffert's theory that at one time in the history of the common archetype the passages due to the Plautine Revival were adscribed in the margin, ^ot only is the complete loss of certain passages easier to understand on this hypothesis ; but the disappearance of single lines like Pers. 610, Trin. 72*, is intelligible, if we assume that the whole passage once stood in the margin and was introduced from there into 8 114 EETEACTATIO IN PLAUTUS the text. It sometimes happens, too, that the spurious passage is preserved in full, but that, in being taken into the text, it has crowded out a genuine line {Stich. 157 in P, and probably Pers. 668 in A). Confusion in the order of lines may have arisen in the same way. Twice a substitute passage has been inserted in the wrong place in both A and P : Pers. 442 f . ; Pseud. 385 f. ; and twice (in portions of the text for which A is missing) P has put an alternative version in the wrong position: Trin. 889-891; 901. The differences of order in A and P {Poen. 300-305 ; Pseud. 151-156 ; Stich. 75- 83) also point to variants which were written in the margin and were taken into the text at different points. This investigation of retradatio in five plays of Plau- tus therefore supports the view that the Ambrosian Palimpsest and the Palatine manuscripts were descended from a common archetype; that substitute versions were written in the margin of this archetype; and that the introduction of these marginal adscripts into the text was responsible for the omission of whole passages and of single lines, for differences in order, and for confusion in the genuine text of Plautus. BIBLIOGKAPHY EDITIONS OF PLAUTU8 GOETZ, G., F. ScHOELL, and G. Loewe — Ed. Mai., Leipzig, 1871-1894. GoETZ, G., and F. Schoell — ^Ed. Min., Leipzig, 1892-1896 (revised 1904-1909). Leo, F. — Berlin, 1885 (incomplete). Leo, F.— Berlin, 1895-1896. Lindsay, W. M.— Oxford, 1904-1905. EiTSCHL, F. — Begun in 1848 (incomplete). Ubsinq — Copenhagen, 1875-1892. GENERAL LITERATURE Baieb, B. — De Plauti Fahula/rum Recenmonihiis Amhrosiana et Palo- tina Oommentatio Critica, Breslau, 1885. Bebok, Th. — Reviews of Ritschl's editions, in Opusc. I., pp. 3-53. Bbachmann, W. — De Bacchidum Plautinae Retractatione 8caenica, in Leipz. Stud. 3 (1880), pp. 59-187. DziATZKO, K. — De Prologis Plautinis et Terentianis Qudestiones Selectae, Bonn, 1863. Cher die plautinischen Prologe, Luzem, 1866-1867. GoETZ, G. — Dittographien im Plautustexte, in Acta soc. phil. Lips. 6 (1876), pp. 235-326. Gbauebt, W. H. — Historische und philologische Analekten, Munster, 1833. Kellebman, H. — De Plauto Sui Imitator e, in Comm. phil. Jen. 7 (1903), pp. 129-197. Ladewig, Th. — Zum Epidicus des Plautus, in Zeitschrift fiir Alter- tumsioissenschaft, 1841, coll. 1079-1099. Oher den Kanon des Volcatiu^ Sedigitu^, Neustrelitz, 1842. ' Einleitungen und Anmerkungen zu plautinischen Lustspielen, in Rhein. Mus. 3 (1845), pp. 179-205; 520-540. *A selected list of the books and articles which have been most helpful in the preparation of this dissertation. Authorities for minor points are given in the foot-notes. 115 ^ 116 BIBLIOGRAPHY A and P^, Ladewig, TK,—Plautinische Studien, in Philol. 17 (1861), pp. 248 269; 452-480. Langen, p. — Plautinische Studien, Berlin, 1886. Leo, Y.— Plautinische Forschungen, Berlin, 1895. Lindsay, W. M. — Th£ Tioo Recensions of Plautus, in Amer. Journ. Phil. 21 (1900), pp. 23-37. The Ancient Editions of Plautus, Oxford, 1904. OSANN, F.—Analecta Critica, Berlin, 1816. Reinhabdt, L.— De Retractatis Fahulis Plautinis, in Studemunds Studien auf dem Gehiete des archaischen Lateins i., Berlin, 1873, pp. 79-111. RiTSCHL, F.—Parerga Plautina et Terentiana, Leipzig, 1845. Parallelstellen im Plautus als Ursache von Glossemen, in Opusc. II., pp. 274-291. Seyffert, O. — Studia Plautina, Berlin, 1874. Zur Oherlieferungsgeschichte der Komodien des Plautus, in Berl Phil Woch. 16 (1896), coll. 252-255; 283-288. Sicker, E.— Novae Quaestiones Plautinae, in Philol. Suppl-Bd. 11 (1908), pp. 179-252. SoNNENSCHEiN, E.—Thc Scientific Emendation of Classical Texts, in Trans. Amer. Phil. Ass. 24 (1893), pp. 5-16. Spengel, A.—T. Maccius Plautus, Gottingen, 1865. Studemund, W.—Zur Kritik des Plautus, in Festgruss der philo- logischen Gesellschaft zu Wurzhurg, Wiirzburg, 1868, pp. 38-76. y. Macci Plauti Fahularum Reliquiae Amhrosianae, Berlin, 1889. Teuffel, W.— studien zu den romischen Komikern, in Studien und Charakteristiken,^ Leipzig, 1889, pp. 315-352. PERSA VAN IJSENDIJK, A.— De T. Macci Plauti Persa, Utrecht, 1884. Meyeb, M.— De Plauti Persa, in Comm. phil. Jen. 8 (1907), pp. 145-191. V. WiLAMOWiTZ-MoELLENDOBF, U.— De Trihus Carminihus Latinis, in Index schol. Gott. 1893-1894, pp. 13-26. P0ENULU8 Bbachmann, W. — De Bacchidum Plautinae Retractatione Scaenica, in Leipz. Stud. 3 (1880), pp. 73-79. BIBLIOGEAPHY 117 Fbancken, C. M. — De Poenuli Plautinae Compositions, in Mnem. 4 (1876), pp. 146-175. GOETZ, G. — De Compositione Poenuli Plautinae Commentariolum, Jena, 1883. Haspeb, Th. — De Poenuli Plautinae Duplici Exitu, Leipzig, 1868. Kabsten, H. T. — De Compositione Poenuli, in Mnem. 29 (1901), pp. 363-387. Lanqbehb, G. — De Plauti Poenulo, Friedland, 1883. Legband, Ph.-E. — Pour Vhistoire de la Com^die Nouvelle, 4. U original du Poenulus de Plaute, in Revue des Etudes Grecques 16 (1903), pp. 358-374. Reinhabdt, L. — De Retractatis Fahulis Plautinis in Studemund's Studien auf dem Gehiete des archaischen Lateins i., Berlin, 1873, pp. 109-111. ScHUETH, K. — De Poenulo Plautina Quaestiones Criticae, Bonn, 1883. PSEUD0LU8 BiEBMA, J. W. — Quaestiones De Plautina Pseudolo, Groningen, 1897. Kabsten, H. T. — De Plauti Pseudolo, in Mnem. 31 (1903), pp. 130-156. KiESSLiNQ, A. — Plautinische Miscellenen, in Symhola Philologorum Bonnensiwn, pp. 835-839, Leipzig, 1863-1867. Zur Kritik und Erkldrung des plautinischen Pseudolus, in Rhein. Mus. 23 (1868), pp. 411-426. Leo, F. — Vber den Pseudolus des Plautus, in Nachr. Gott. Ges. 1903, pp. 347-354. LoBENZ, A. 0. F. — Zum Pseudolus des Plautus, in Philol. 35 (1876), pp. 153-180. Ed., Berlin, 1876. NoBDEN, E. — Sprachliche Beohachtungen zu Plautus, in Rhein. Mus. 49 (1894), pp. 197-203. Sauppe, H. — Quaestiones Plautinae, in Index schol. Gott. 1858-1859. Schmidt, F. — Bemerkungen zum Pseudolus des PlautUrS, in Misc. phil. Gott. 1876, pp. 20-31. Schmitt, a. — De Pseudoli Plautinae exemplo Attico, Strassburg, 1909. UsENEB, H. — Pseudoli Plautinae Scaena Secunda, in Index Schol. Gryphiswald., 1866. STICHUS Leo, F. — Vher den Stichus des Plautus, in Nachr. Gott. Ges. 1902, pp. 375-391. 118 BIBLIOGBAPHT TRINUMMUS Brix, J.— Ed., Leipzig, 1879. ——Edition revised by M. Niemeyer, Leipzig, 1907. RiBBBCK, O.—Zu Plautua' Trmummus, in JRhein. Mus. 27 (1872), pp. 177-180. RiTSCHL, F.— De Jnterjiolatione Trmummi Plautvnae, m Parerga, pp. 509-579. Teuffel, W.—Zu Plautus' Trinummus, in Rhein. Mus. 30 (1875), pp. 472-475; 632-633. Vahlen, J.—yoria, in Hermes 16 (1880), pp. 257-259.