BC5UGHT WITH THE INCOME FROM THE SAGE ENDOWME'NT FUND DHE GIFT OP Itenrg US. Sage 1891 /<; JuIMT^A. %..3/..f..%. THE PALATINE TEXT i OF PLAUTUS I BY W. M. LINDSAY, M. A. FELLOW OF JESUS COLLEGE, OXFORD. Ham^B f arte anb ®0. 27 BROAD-STREET, OXFORD ; AND 6 SOUTHAMPTON-STREET, STRAND, LONDON. i8q6. THE PALATINE TEXT OF PLAUTUS BY W. M. LINDSAY, M. A. FELLOW OF JESUS COLLEGE, OXFORD. Ifanwa f arte anit €0. 27 BROAD-STREET, OXFORD ; AND 6 SOUTHAMPTON-STREET, STRAND, LONDON. 1896. PRINTED BY JAMES PARKER AND CO., CROWN YARD, OXFORD. C^e 'llalathte' ^t^t at "^hntuB, r^UR MSS. of Plautus are, as is well known, divided ^ into two families. The sole representative of the one is the famous fourth century Palimpsest in the Ambrosian Library at Milan, a MS. written in Capitals. The other, comprising all the remaining MSS., which are, without exception, in minuscules and date from the tenth century onwards, is called the ' Palatine ' family, because its two most important representatives belonged to the old Palatine Library. Some of the divergences of the ' Palatine ' text from the text of the Ambrosian Palimpsest are divergences which are known to have existed between rival editions or ' recensions ' of Plautus in the period of the early Empire (e.g. in Paulus' epitome of Festus mention is made o{ gravastellus, the ' Palatine ' reading in Epid. 620, as well as of ravistellus, the reading of the Palimpsest). Our two families of MSS. are therefore generally sup- posed * to embody two early rival ' recensions,' the ' Palatine ' and the ' Ambrosian recensions,' as they have been called. The ' Ambrosian recension ' we know, as far as it can now be known, from Stude- mund's Apograph of the Ambrosian Palimpsest (Berlin, 1889). The texts of the MSS. which exhibit the ' Pala- tine recension ' have been made available to us in the very careful and accurate ' apparatus criticus ' of the new Ritschl edition (completed in 1894)''. But a good ' See on this point Leo Plautinische Forschungen, recently published. ^ This is probably as accurate an apparatus criticus as has ever been compiled for any classical author. Still it is impossible but that some errors and omissions must have crept into a work of this extent ; and there B 2 4 The ' Palatine' Text of Plautus. deal has yet to be done in order to get a knowledge of the ' Palatine ' text as it existed in the Archetype of these MSS., an Archetype referred by general consensus to the 8th or 9th century, and, if such a thing be pos- sible, to get a glimpse at a still earlier form of the text, the form namely in which it may have existed in a proto-Archetype contemporary with the Ambrosian Palimpsest. In this paper I propose to trace the his- tory of the transmission of the ' Palatine ' text with the is no doubt that a facsimile by photography of the chief minuscule MS., the Codex Vetus (B) of the Vatican, would be of immense service to students of Plautus. Here is a batch of ' corrigenda ' which I noted in a fairly rapid examination of the earlier plays, and a more leisurely inspec- tion of the Captivi: — Amph. H qui demidiam D ; 600 und B (not unam. It is not uncommon in B for the Adverb to be distinguished by an accent, e.g. una Capt. 479, ilU Amph. 197, 203, Capt. 359, Cure. 340) ; 619 tibi W ; 669 comodum W (not commodum) ; 770 thessala J (the t has the 'daseia ' over it to indicate th) ; Aul. 664 sereliam 'S^.feret iam B^ ; Capt. 102 ut imperet J ; 138 Ergo sile D (as well as B) ; 208 Apage om. J ; 291 estopus1\ efiT occeperif]; /i,'j(> condempmnt Vl \ i,<)a, His 'Q O \ ^16 nimio in ras. B'' (so B' probably had nemo like f-Ey) ; 521 JVe B E, JVec J ; 562 ak meus (eu ex corr. ) B (so that there is no trace in 5 of an alcumms) ; 900 Cura quam (ex Cura aquam) B, potes B, petes E J ; 940 uti B ; 994 gnatus meust B, gnatus meus est J (what of E ?) ; Cas. 260 aut B" in ras. (so that B' probably had et like VEjf) ; 374 uotat is also the reading of B ; 634 the horizontal stroke drawn by B° through the upright stroke of the d of B^ seemed to me not to indicate a correction of d to t, but to be a reproduction of the Archetype's cadamabo, a corruption of coda amabo^ i.e. cadam amabo ; 759 ne meaefx. meae meae B ; 846 does the marginal sign by B' indicate a marginal variant oliocabo in the Atchetype? ; Cure, 99 the correction of nausea to nautea seemed to me to be by the hand of Camerarius ; Epid. 52 Totus (? Totis) B (not Tot. There is a contraction- sign after the second t) ; Baech. 539 antedhac ex andedhac B ; 695 aee B', haec B^ ; 743 for "post e " read "ante e " ; 940 " geruntur B^ " seemed to me doubtful. The supposed contraction-sign for -ur appeared to be a part of the punctuation-mark; Men. 432 for "sciscitari" read " scisscitari" ; the marginal susurra is, I think, in a 1 5th cent, hand ; 479 parti B', parte B° ; the suprascript words I read as palla eri. Modernized spellings like -uus for -uos, cum for quom, &c., are not always recorded, e.g. Capt. 149 tuumBD^, 303 cum BDJ, 323 minime BDJ, 468 esuriales BDOJ, 714 seruus B E, (but seruos J). Nor are the junctions of small words like Prepositions with neighbouring words, e.g. Amph. Arg. ii. 2 informam D, 3 Propatria D, and so oftai with D throughout the play, e.g, 3 inrebus D, 10 inrem D, 12 abdis D, 20 aduos D. I use B^ in these 'corrigenda' in the same sense as it is used by the Ritschl editors. The ' Palatine' Text of Plautus. 5 help of some clues which were suggested to me by a recent inspection of the chief minuscule MSS. The first thing about the minuscule MSS. of Plautus that calls for explanation is the fact that while MSS. of the first eight plays (in this order: Amph., Asin., AuL, Capt., Cure, Cas., Cist.,Epid.) are fairly numerous, MSS. of the last twelve " (in this order : Bacch., Most., Men., Mil., Merc, Pseud., Poen., Pers., Rud., Stick., Trin., True.) are very few. The last twelve plays of Plautus were practically unknown until the Renais- sance, when the discovery of the codex Ursinianus {D), now in the Vatican collection, was hailed as the dis- covery of a lost classic. To explain this curious fact in the transmission of the ' Palatine ' text we must, I think, suppose that the immediate Archetype of our minuscule MSS., an Archetype which I will call Pp, had been divided, probably for convenience of copying, at the end of the Epidicus. The first volume of the copy, containing at the beginning the title of the work and the name of the author, kept its place in monastery Libraries, and was duly entered in their catalogues, while the second, being without such indication, fell into neglect. Now an inspection of our three earliest minuscule MSS. reveals ^ clear traces of an Archetype which had been divided in this way. These three MSS. are : — B, the Codex Vetus, in the Palatine collection at the Vatican ; saec. x. C, the Codex Decurtatus, in the Palatine collection at Heidelberg ; saec. xi. in. D, the Codex Ursinianus, in the Vatican collection at the Vatican ; saec. xi. " Of the Vidularia only the title remained at the end of the Archetype of our ' Palatine ' MSS. ^ See Ritschl Prolegg. ; also Chatelain's account of these MSS. with photograph specimens in his Pallographie des Classiqices Latins. 6 The ' Palatine ' Text of Plautus. B is generally described as a MS. containing all the twenty plays with the Querolus, a late imitation of Plautus, prefixed. But when we look at the MS. itself we see that the scribe, or rather scribes, when beginning, the task of copying, had an original which contained ^ only the first eight plays. They began their work on what is now fol. 9, and on the reverse side of that leaf wrote a title : in hoc volumine continentur comediae plauti numero via, with a list of the first eight plays (Amph. to Epid.). When the last twelve plays were added and the Querolus prefixed, that title and list were erased, and on the first page of the quaternion which was prefixed with the Querolus a new title was written : In hoc volumine continentur comediae plauti numero xxihr with a full list of the plays, beginning with the Querolu^ ;^„ (hence ' twenty-one,' not ' twenty '), and with a slight break in the list between the Epidicus and the Bacchides. This addition of the last twelve plays and the Querolus to the first eight must not be regarded as a case of binding in one volume two entirely different MSS. For the Querolus overflows into the first page of the Amphitruo-Q^z.\.&xn\on, and the Epidicus ends on the second leaf (fol. 74 v.) of a quaternion of which the Bacchides occupies the remaining leaves, the transition from the last play of the first eight to the first play of the last twelve being apparently made without a break "- Everything points to the staff of scribes, em- " That a new scribe began, as Chatelain appears to say, on the first page (fol. 73 r.) of this quaternion (Epid. 532 sqq.) seemed to me a matter of doubt. The vellum of which this quaternion is composed is somewhat soft and ' furry,' differing from the hard, glistening vellum of the preceding quaternion ; and this change of the material may account for the slight change in the appearance of the writing. It certainly is not the case that the writing of the last twelve plays is of a later style than the writing of the first eight. Whether the scribe of Men. 381— fin.. Mil., &c., was the scribe of the first part of the MS. {Amph Capt. 190) I could not decide. The Bacchides begins a new page, as is a common practice with new plays, but is written in the same hand as the preceding pages of the Epidicus. The 'Palatine' Text of Plautus. 7 ployed on the copy, having gone on continuously from the first part of Plautus to the second part. Our next oldest MS., C, contains, it is true, only the last twelve plays. But a single glance at the quaternion-marks, the numerals added at the bottom of the last page of each quaternion, shews us that the staff of scribes (three in number), who copied out the MS., numbered their quaternions with reference to a full collection of all the twenty plays. The original mark of the first quaternion of our MS. was xvii, of the second, xviii, and so on till the last, which would be No. xlvi, though in the MS., as it stands, these quaternion-marks of the scribes have been erased, and the marks (equally old marks, it seems) i, ii, &c., remain. The third MS., D, contains, in addition to the last twelve plays, the first three and a part of the fourth {Amph., Asm., AuL, Capt. i — S^S); the quaternions of the first part being marked with numbers as the quaternions of C, those of the second part with capital letters. A, B, C, &c. This MS. presents every ap- pearance of being in reality two separate MSS., which have been bound together. It seemed to me that the writing of the last twelve plays presented a later ap- pearance, while the writing and general style of the first three and a half had a considerable affinity with C. The text of C and of D is so similar that they are admittedly derived from one and the same original, and that a copy of the original of B for the last twelve plays. The home of C was the Abbey of Freising, near Munich, as we know from the entry on the fly- leaf at the beginning : liber iste est sancte Marie et sancti Corbiniani Frisingensis. D, as is well known, was found at the Renaissance in Germany. The home of B is not ascertained ; but its style of writing clearly proclaims it to be a German MS. Our supposed Archetype (/••?), which was divided at the end of the 8 The ' Palatine' Text of Plautus. Epidicus^ for convenience of copying, may thus with some likelihood be located in Germany. An examination of the text of B and of D for the earlier part of Plautus shews us that the two MSS. were copied from the same original, a MS. which had, e.g., arte Nik incabis Amph. 440, hci 700, qua me Asin. g6, ititidem 220, ipsos Amph. 754, instans Asin. 54. The band of connexion between the two MSS. is drawn closer by the analogous pagination in B and D (see Ritschl Pro- legg. xxxii.) ; the place where the writing of this part of D abruptly stops » {Capt. 503), apparently the exact half of the first eight plays in the original, ends a leaf in B (fol. 43 v.). This is an important fact to bear in mind in studying the text of the first three and a half plays, that the testimony of D stands on an absolute level with the testimony of B in respect that both are copies of the selfsame original ; and I do not know that D is to be called a worse copy ^ than B. So that the reading of D must be regarded as of equivalent value to the reading of the scribe of B. I say " the scribe of B" for in the Codex Vetus the text of these first eight plays has received a large number of corrections (cf. Seyffert B.P.W., 1895, p. 682) ; and to the nature and origin of these corrections, a matter of immense importance in the history of the text of Plautus, we must now turn our attention. It is much to be wished that some expert in Latin ' This seems an unequal division ; though some of the inequality is removed if the Querolus stood at the beginning. e At the end of a leaf (fol. 69 V.), the last two leaves of the quaternion (the ninth), foil. 70, 71, being ruled for writing but left blank. This is very near the place where the first scribe of B stops, viz. Capt. igo. •■ The chief fault of D in this part is a. tendency to omit small words, e.g. Amph. 6n sum, 612 nunc. D is however superior to B in indicating the lacuna at Amph. 1034, and in Asin. 96 reproduces the qua me of the original, where B incorporates the gloss with the text and gives qua me arte. The ' Palatine ' Text of Plautus. g Paleography, on whose judgment scholars could rely, would make a thorough study of these corrections and settle once for all the difficult question of assigning each correction to the particular corrector to whom it belongs. I could have wished for a longer time for the examination of them than was at my disposal ; but still I convinced myself that it was possible to distinguish two outstanding" classes of corrections: (i) those which appear to be the work of the scribe himself^ either at the moment of writing, or after the completion of the page; these I would label B^ ; (2) those which come from the hand of the official 'corrector,' who used for the purpose another MS., not the original of B ; these I would label B^- That the B^-writer was the official 'corrector' of the MS. I infer from this fact among others, that the scribe (B^) has here and there left a blank for an unintelligible word and has written d[eest] in the margin. These blanks (e.g. Cas. prol. 64, 347, 361) have been filled up by B^, who also supplies a good deal of the punctuation, scene-headings and 'personam m notae.' In Aul 455-8 for example, the 'personarum notae,' EVCL(io),COC(us), EVCL(io), are in the same handwriting and the same colour of ink as v. 454. Temperi . . . caput, a line not found in any other MS. and added in the margin oi B. Similarly in the Asinaria the 'personarum notae' at vv. 652-3, LEON(ida), ADOL(escens), are in the same handwriting and the same colour of ink as the ' There are a very few others, scattered up and down the plays, in different shades of ink (on which see Goetz's Preface to the Epidicus, p. xii, &c.). But they are so few that we can ignore them in this discussion. ■■ It is possible that some of the B' corrections may be the work of the •rubricator,' supposing him to be different from the scribe. Corrections that can be with certainty referred to him are a few in red letters or with addition of red dots ; e.g. in Capt. 195 the id of id voluerunt is separated from the following syllable by a red line ; in Epid. 661 the t of tres is expunged by red dots. B 3 lo The ^ Palatine'' Text of Plautus. suprascript si in v. 651, a word omitted by all the other MSS. The eye soon becomes accustomed to the upright, fine writing of this corrector (B^) and the grey tint of his ink'; and I think that anyone who gives a little time to the examination of the MS. will agree with me that all such corrections as are evidently derived from a different original than the original of B^ D come from one and the same corrector'^. Wherever a reading in the text or a ' personae nota ' in the margin is found in B but not in the other MSS., that reading or ' personae nota ' will be found to be in the handwriting of this corrector whom I call B' : so that this corrector must have used a MS. which was not available to the writers of the other MSS. At the same time a large number' of these i?^-corrections are merely restorations of the text of B to the form in which the text appears in D, and in which it without doubt appeared in the original of BD (e.g. Aul. 48, 257, Capt. S'^. Are we then to suppose that B^ used two originals to check the errors of B, viz. (i) the original of BD, which we may call P-*-°, (2) the original from which the unique readings come, an original which we may call P^^ ? To me it seems that the more likely explanation is that only one was used, viz. P^^, and that all the corrections in B hy B'^, whether corrections of mere deviations of the scribe from P^'^ or additions of entirely ' Towards the end of the Captivi it has a darker hue ; but the writing is clearly the same. " About the corrections in the Amphitruo, e.g. 156 quasi, 175 onus, 178 potivii, 199 turn, 218 se, 294 detexere, 303 pridem, 323 gestiunt, 375 cuius, 406 hie, 647 clueat, 755 ego . . . credis, I speak with more reserve, for the ink has here a yellower tinge, and the letters have not the pronounced B^-character. Still I am inclined to refer them also to the same corrector. Even if they came from another hand, that would only mean that the oiiScial corrector of the Amphitruo was a different person from the oiiScial corrector of the other plays, though the original used by him was no doubt the same. B'^xs, not the rubricator, for, e.g., to the ru- bricated headings of Epid. III. iv B^ has added in ordinary ink the word MVLIER. The ' Palatine ' Text of Plautus. i i new readings, come from this single original. This original was clearly a MS. of the 'Palatine' family, having, for example, the same omissions as the other ' Palatine ' MSS. at Cas. 86$ sqq., Epid. 593, 624 "- Nay more, I believe it to have been the actual original of P^^. For now and then we get glimpses of a close relation between it and the original of the other MSS. Thus P^^ seems to have had in Asin. 266 obscaeuauit, a reading from which the variants of the other MSS. are easily explained ; similarly in Aul. '537 it had edi (in P^^ di) with the suprascript gloss audiui (cf. di audiui B', di D) ; in Cas. 634 it had necadamabo (a cor- ruption of ne cada amabo (see above), and so on (cf Cas. 347). These are specimens of the indications which a study of the B^-readings affords that P^^ was nothing but the Archetype of all the other MSS. The same corrector continues his work in the second half of the MS. (we see his handwriting clearly, for example, in the addition desubito, Bacch. 79) ; though the extent of the part he has played in this section of the MS. is not of much moment to determine, for the corrections here seem to be taken from the actual original copied by the scribe and to bring no new codex within our range °. Still the fact that he shews traces ° In this last passage the Ambrosian Palimpsest supplies us with the whole line : Estne consimilis quasi cum signum pictum pulchre aspexeris. The Archetype of the ' Palatine ' MSS. had estne cons signum ficlum, &c., the gap before signum being due to a hole in the leaf of this or a previous Archetype. The MS. used by B^ had evidently the same gap; and his addition of idei-a, so as to make the word considera, is a mere conjecture, like his change of latis (for blatis) of B^ in Epid. 334 into latras, or his gloss upon oppo (which he mistakes for oppido) of the Archetype in Cas. 766 (where the Ambrosian Palimpsest has oportuit), id abundanter. » Occasionally it would seem as if a correction came from an original of the C D-type, e.g. Most. 316, 413, Mil. 363. I ascribe to mere conjecture (like Most. 264 effigiem) the few good corrections in Most, isi, 187, 355, &<:• 12 The ' Palatine ' Text of PlautuS. of himself in the second part is of importance, for it i5Uggests a likely account of the composition of the whole MS., namely this : — The scribes employed in the transcription of B began with an original (P^^) which contained the first eight plays only, the same original as was used by the scribes of the first three and a half plays of D. The 'corrector^ of the scriptorium corrected this copy not from the original {P^^) itself, but from another MS. (P^^), which contained the whole twenfy plays and the Querolus prefixed, and ■which probably was itself the original of P^^. It was from this MS. P^^, brought into requisition by the correc- tor, that the remaining plays {with the Querolus') were copied, and the old title with list of eight plays replaced by a new title with a full list. For this second part of the MS. no fresh original was available, and so the various correctors of this part used the actual original from which the copy was made. The corrections then in the second part of B have no more value i? than the corrections which we label B^ in the first part, and may conveniently be all included under the sign B'^, even though they come from various hands. They persist till the middle of the Miles and are very numerous indeed in the Mostellaria. It is un- fortunate that the corrector did not continue his work till the end of the Mercator\ for the scribe of this part was a very careless one. There is probably no part of Plautus where the testimony of B is of less repute than the latter half of the Miles and the whole of the Merca- tor (see, e.g., Merc. 981-3). The new scribe who be- gins at the end of the Mercator (at v. 1013) and the numerous scribes who follow him are much more care- T> No authority, for example, should be attached to the reading culinae in Most. 5. Utcumque (not quaecumque) was the reading of the Archetype in Bacch. 662, and so on. 1 The order of the plays in the second part of B (as of C and D) is Bacch., Most., Men., Mil., Merc, Pseud., &c. The ' Palatine ' Text of Plautus. 1 3 ful ; and the reason why the corrections are not continued through the remaining plays may well have been that the plan of dividing the last seven plays between several scribes, to each of whom was assigned only a small portion to copy, ensured such accuracy of copying that the services of the corrector were not necessary. Our examination of the three oldest MSS., B, C, and D, has thus taken us past and behind them to four (or three) lost MSS. :— (i) P^^, which I identify with (2) P^, the original of B for the last twelve plays, and suppose to have contained all twenty plays with the Querolus prefixed. (3) P^^, containing the first eight only. I make this MS. a copy of P^^ (4) P^°, the original of C and of D for the last twelve plays. It contained the last twelve plays only, and was a copy of P^- If I am right in supposing P^^ to be the first part of P^, this whole MS. (P^^+P^) is nothing else than the Archetype (P^), which we are in search of, an Arche- type divided at the end of the Epidicus for convenience of copying. P^^ and P^^ are the two-volumed copy. The remaining minuscule MSS. of Plautus, which, as we have seen, contain only the first eight plays, do not point to another Archetype, but are all derived from a common original, which we may call P^, and which was itself a copy of P^°- The chief of these MSS. are: — E (Milan, saec. xii ex.). V (Leyden, saec. xii in.). / (British Museum, saec. xii in.) f. ' y has a curious epigram at the end : Exemplar mendum tandem me compulit ipsum Cunctantem nimium Plautum exemplarier istum, Ne graspicus [leg. graphicus) mendis proprias idiota repertis Adderet, et liber hie falso patre falsior esset. 14 The ' Palatine ' Text of Plautus. To these we have to add two leaves of a lost MS (C), which are bound up with a MS. in the Ottobonian collection at the Vatican. They contain Capt. 400 — 555 in minuscules of the eleventh century. seems to have belonged to the V E J group, and may have been more closely connected with J (e.g. Capt, 512 tandem om. O J). The O V E J group, being derived from an original which was itself a copy of the original ol B and D, is of little value where we have the evidence both of B and oiD (in Amph., Asin.,Aul., Capt. 1—503), and need only be appealed to when B and D are at variance '. For the remainder however of the first eight plays {Capt. 504 — fin.. Cure, Cas., Cist., Epid.) it has more impor- tance, for in the absence of Z* as a check upon B, we cannot always be sure that a peculiar reading of B is not a mere freak of the scribe, destitute of all authority'. Of course where B^ intervenes with a correction, its authority is paramount ; and I am in- clined to accept its prode in Cure. 487, where B^ corrects T tu pro e uirgo oi B^ to Ei tu prode uirgo (on prod{e) of prod-ire, &c., see my Latin Language ch. ix § 44) ", and its eopso in Cure. 538, a line left by B^ for B^ to fill That this cannot be, as it is often supposed to be, the composition of the scribe of J, but has been copied by him from his original, we see from the fact that the word graphicus has been wrongly copied graspicus. I would refer this corruption to a method of writing ph, of which there are many traces in the minuscule MSS. of Plautus, I mean the superposition of the ' daseia ' (Isid. Etym. I. xviii. 10 Ot.) over the consonant. An ill-made • daseia ' would be easily mistaken for the early minuscule form of s. ' So in Asin. 765 abs te was the reading of the Archetype, not ex te ; similarly in Aul. 406 Opiati uiues (cf. Seyffert in Burs. Jahresber. Ixxx. 233). ' Thus in Cure. 507 the his saltern of B is a, corruption, due to the fact that in P"" (as in P^) the words hi saltern were written his altem Cf. Cure. 198, Capt. 971. " The reading of /"" seems to have been proae, the scribe having mistaken the d of the Archetype for the tall-backed form of a. A similar mistake explains the confusions of aio with dico in the Captivi (e.g. 72, The ' Palatine ' Text of Plautus. I S in (on eopso, eumpsum, &c., see my Latin Language, ch. vii. § 20) "- One other MS. remains to be considered, the lost 'codex Turnebi' {T), and its relation to our hypo- thetical Archetype. All that we know of this lost MS. is the account of certain of its readings, given by Turnebus in support of his emendations of the text of Plautus ; and knowledge of this sort is always more or less uncertain. From these readings (they have been collected by Goetz in the Preface to the Poenulus) we see that in the first eight plays T shews close relation with the other minuscule MSS., sharing their corruption Libyco for Liberi in Cas. 640, while in Cure. 485 it departs, apparently through wilful conjecture on the part of the scribe, from the true reading Oppiam, which they offer, and substitutes operia^n^, and in Cas. 729 it intervenes to justify the reading of ^ against V E J. In the last twelve plays it shews a similarly close relation with our minuscule MSS., agreeing with B against C D in cases of variance. Its peritent Poen. 30, where B has pertant and C D have pereant, is a mere conjectural emendation of pertant of the Archetype (a corruption of the Plautine pereant, due to confusion 694). Is it possible that this is the explanation of the famous corruption in Capt. 201, mulla oculis multa miraditis (rniraclitis) ? Should we read : Eiulatione baud opus est : Multa oculis muti mira aitis ? The corruption of muti to multi is common enough in MSS., and a further grammatical alteration of multi to multa causes no difficulty. But the form aitis is open to question. ' Similarly in Capt. 420 I am inclined to accept the B'-correction laudibus, and read : quantis seruum laudibus, Suom erum seruus coUaudauit. Cf. also Amph. 207 reddere ; Asin. 38 qui, &c. y Maeoni aper of Cas. 523 seems to me another conjecture of the scribe for menuiaper, as he misread merui aper of the Archetype. That merula per of Festus is the true reading I have tried to shew in the Classical Review, 1892, p. 124. i6 The ' Palatine ' Text of Plautus. of E and T in capital script). But its usefulness appears in Poen. 977, where it retains Punicast guggast homo, words which were omitted hy B C D as unintel- ligible '■, and which may have had some discrediting mark accompanying them in the Archetype, and (to omit lesser instances like Poen. 1033 tnicdilix T, micdilia BCD; Rud. 122 exicas T, exigas B, exagitas C D ; Rud. 641 praeuortere T, reuortere BCD) in Rtid. 613, 724, 671, where it supplies (editors say, correctly supplies) words or letters to gaps of the original of BCD. In Rud. 613 5 shews fanom, with a blank space following, C D shew fano, but T completes the line with fano meae uiciniae ; in Rud. 724, where B leaves a lacuna at the beginning of the line, and C D begin with Est lex, T reads Non licet est lex ; in Rud. 671, where BCD leave a lacuna before scelestus, T reads quin scelestus (cf Pers. 606, 622, 843). Do these readings which Turnebus ascribes to T compel us to suppose that T came from a different Archetype than the Archetype which we have posited for the MSS. preserved to us ? To me it seems that there is no great necessity for this supposition. It is quite conceivable that they came from that very Archetype and no other, possibly having been taken from it at an earlier period, that the copy B was made when the gaps in it were not so pronounced, or when obscure passages in it were more legible. I will there- fore continue to speak of the composite MS., P^^+P^, as identical with Pf, the Archetype of all minuscule MSS. of the ' Palatine ' family. It remains to discover what may be inferred about the form of this Archetype from a study of the several minuscule MSS. An inference of this kind must always be made with caution. For a peculiarity shared by all ' It is possible that B left them for the corrector to supply, as in Poen. 770. The' Palatine' Text of Plautus. 17 the MSS. may be a peculiarity, not properly of the immediate parent MS. {P^), but of a still earlier ancestor. Thus their omission in certain plays of letters, syllables, or whole words at the beginning of a line, and some twenty verses later at the end of a line can be ascribed with certainty to a hole in the page, and can be used as a criterion of the number of lines contained in that page ; and it is from evidence of this kind that Schoell, Seyffert and others {Bursian. Jahresb. Ixxx. 236) have estimated that the Archetype con- tained from nineteen to twenty-one lines to a page. But who can say whether the hole was in the page of P», or of the MS. from which P^ itself was copied, or of the original even of that MS. ? Such evidence only enables us to say that some Archetype or other had 20 lines to a page, not that this or that definite Archetype had. I can produce a surer criterion of the number of lines in a definite Archetype (the original of B, an original which I identify with P^) for one play at least, the Poenulus. The excessive conscientiousness of the German monk who copied a part of this play led him to copy the very headings of the page of his original at the place where they stood in the original ! So that in the margin at w. 1222-3 we fiadplauti, at vv. 1255-6 paenulus, at vv. 1288-9 plauti, and again at vv. 1354-5 plauti, at v. 1385 penulus, sure indications that the Poenulus in P^ (i.e. in Pf) was written on pages of some 33 Hnes. (After v. 217 B omits 2, not 3, pages.) The same MS., B, gives us another clue, and a very important one, to the place held by this Archetype in the transmission of the ' Palatine ' text. At the beginning of the Bacchides, that is to say, where the copying of P^ {P^) into B began, we find a curious spelling, usually corrected, of words like huic, hoc, hoc, with k instead of h. Thus in v. 107 huic was first written kuic ; in v. 108 the spelling kac remains un- 1 8 The ' Palatine ' Text of Plautus. corrected ; in v. 105 kinc has been erased, and hinc remains (the original had both words apparently); in V. 1 10 ^fc was first written koc, and so on (cf. w. 169, (177, 444, &c.). The substitution of k for h can hardly be traced to any other source than an Archetype in capitals, like the capitals of the Ambrosian Palimpsest where the letter H is written exactly like the letter K. Now it is at the beginning of a copyist's task that we generally find most fidelity in the reproduction of the original ; and the strange spelling preserved in B seems to indicate that in P^ (Pf) a new scribe began his work at the Bacchides, and with curious fidelity reproduced the K-forms of his original, which must accordingly have been a MS. in capital script, and therefore of an early date. (This Archetype we may call P".) The word ' must ' may seem too strong to be used of a conjecture of this kind ; but there is surely very great improbability that misspellings of so familiar words, kuic for huic, kac for hac, koc for hoc, &c., would be transmitted through more than two generations of MSS. If the conjecture be right, our task of recon- structing the Archetype F^ is enormously simplified. Every Latin MS., and the MSS. of Plautus like the rest, contains corruptions which can only have arisen from confusion of minuscule letters, and side by side with them corruptions which can only have arisen from confusion of majuscule letters" ; but it is usually impossible to decide how many minuscule Archetypes may have shared between them the production of the first class of corruptions. If, however, only one MS., Pp, stands between B (for the last twelve plays), and the Archetype in capitals, P", this difficulty disappears. • A good example in the ' Palatine ' MSS. is the curious corruption idum for Jmem in Mil. 1414, which is at onqe explained if we write Jovem as it would be written in capital script, lOV. The O was confused with D, and the rest of the misspelling followed as a matter of course. The ' Palatine' Text of Plautus. 19 The size of page inferred from the numerous lacunae in our existing' ' Palatine ' MSS., viz. 19 to 21 lines, I would refer, not to Pf, but to the earlier stage. For although it is possible that the pa^es of other plays in Pv contained a quite different number of lines from the pages of the Poenulus, it is not probable. Nineteen is the number of lines ^ to a page of the Ambrosian Palimpsest [A); and there are several indications that P had a form very like A. For example, the word omitted by all the 'Palatine' MSS. in Epid. 710, inpudentiast, occupies a separate short line at the bottom of the page in A (cf. Mil. 169) ; the lacuna (cf. A) at the beginning of v- 87 and at the end of v. 100 of the same play in P postulates a page (with a hole in it) of the same arrangement of the lines (nineteen in number) as A ; similarly in Cas. 742-54 and 881-3'= the arrangement of lines in P and A must have been the same (cf. Epid. 19) ; and so on. To sum up then, our investigation of the minuscule MSS. of Plautus has led to these conclusions regarding the history of the transmission of the 'Palatine' text : — (i.) There was in some German monastery library in the tenth century a MS. of Plautus [P^^), from which the unique readings added to the first eight plays in B by the corrector (B^) are derived. (2.) This MS. may have been the original from which was copied P^^ (the original of B [Amph. — Epid.'\ and oi D[Amph. — Capt. 503] ), and therefore the Archetype ^ That the original from which A was copied had a page of twenty-four lines is suggested by the fact that vv. 232-3 of the Stichus are wrongly written after v. 208, as well as at their proper place, an error which was likely due to the scribe's copying, when he turned over the page, the top- lines, not of the left-hand page, but of the right-hand page. "= When one looks at passages like this in the Apograph of A where lacunae occurred in P, and tries to picture to oneself the appearance of the gappy page of P, the suggestion is often forced upon one that P may have been a papyrus MS. 20 The ' Palatine ' Text of Plautus. also of P^ (the original of O VEJ), which was a copy (3.) It may have been the first part of a MS. of which P^, the original of B {Bacck. — True.) was the second part. The composite P^^+P^, divided for convenience of copying at the end of the Epidicus, I identify with the immediate Archetype (P') of all the ' Palatine ' MSS., including T. (4.) This P^ (Pfi) was immediately copied from a MS. in capitals (/'"), and (5) was the Archetype of /"^^, the original of C and of D (Bacck. — True). If we put these results in the form of a ' stemma codicum,' we have : — {Amph. — Epid) pB% = pBD /"" (written in capitals) pp T '■ (Bacck. — True.) P'i B P^° B D P Feb. 1896. OJ V E W. M. LINDSAY. Crintet) bs James Market anb (to., Crown i^att, ®tfot5. f r" riQ 1 "'''*— NOV PA 6585!l74 " """"'"'"' """^ Palatine text of Plautus, 3 1924 026 478 051 Ml Oftim;. mpi-;^^