STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. STUDIES IN THE MSS OF THE THIRD DECADE OF LIVY By P. W. Shipley In Certain Sources of Corruption in Latin Manuscripts, pub- lished in 1904, I sought to illustrate by means of two MSS of the third decade of Livy (the Puteanus,and its ninth-century copy, the Vaticanus Reginensis 762) the errors which crept into the texts of the Latin authors in the copies made by ninth-century scribes of the uncial MSS of the fourth, fifth, and sixth centuries. The scope of the work precluded the consideration of a number of questions arising out of my study of the two MSS. These questions will be discussed in the present series of papers under the following headings: I, The Relationships of the MSS in the Puteanus Group; II, Addenda and Corrigenda in Mommsen's Analecta Liviana for the Readings of Vaticanus Reginensis 762 ; III, The Codex Mediceus; IV, The Date of the Corrections by Erasure in the Puteanus; V, Addenda and Corrigenda in the Readings of the Puteanus as given in the Critical Apparatus of August Luchs. I. THE RELATIONSHIPS OF THE MSS IN THE PUTEANUS GROUP 1 Since 1869, when Heerwagen published his Commentatio critica de T. Livii xxvi. 41. 18-44. 1, the attention of students of the text of the latter half of the third decade of Livy has, naturally enough, been concentrated upon the problem of restor- ing, in so far as it is possible to do so, the readings of the Spi- rensian tradition, and the relationships of the MSS in this group have been worked out in great detail. The MSS of the Puteanus group have received but little attention, since they are descended from an existing original and are, for this reason, of no value for the text of Livy for the greater portion of the decade. But, at 1 The MSS of this group which are earlier than the end of the thirteenth century are given below, together with their designations : P = Puteanus sive Parisinus 5730, saec. V. R = Vaticanus Reginensis 762, saec. IX, a copy of P. This manuscript is not used in Classical Philology IV, October, 1909] 405 406 F. W. Shipley the beginning and end of the decade, which are missing in the Puteamis, some of the earlier MSS of this group possess an inde- pendent value, and the readings of C, M, B, and \ are for these portions given in the critical apparatus of August Luchs. The determination of their relationships to each other and to P is thus of some consequence. A stemma is given by Luchs on p. vii of his Prolegomena, as follows: Vat a 762 . Vit' o I This stemma, as I shall proceed to show, is not entirely cor- rect, although Luchs can hardly be held responsible for the error. His classification is no doubt based upon the readings given in Mommsen's Analecta Liviana. This work, while it laid the foundations for a sounder criticism of the text of the second half of the decade, is full of inaccuracies of detail, due partly to the haste with which some of the collations were made, and in part to the almost unavoidable blunders in arranging and digesting by means of numerical designations the readings of eighty-four manuscripts. It will be seen from the readings of the Vaticanus Reginensis given below (p. 413), what a considerable margin one must allow for possible error in the readings of most of the other MSS as given in the Analecta. The Vat. Reg., in the critical apparatus of the editions, and has therefore no letter of designation. In the following pages it is referred to as Vat. Reg. and its readings are given for convenience under the designation R. Its number in Mommsen's Analecta is 66. M= Mediceus Laurentianus LXIII. 20, saec. XI, a copy of Vat. Reg. according to Luchs. No. 36 in the Analecta. O = Parisinus 5731 Colbertinus, saec. XII. No. 1 in Analecta. B = Bambergensis M. IV. 9. saec. XI. No. 9 in Analecta. X = Laurentianus LXIII. 21, saec XIII. No. 37 s in Analecta. Londiniensis Harleianus 2493, saec. XIII. No. 27" in Analecta. Vaticanus 1847, saec. XIII. No. 55 in Analecta. Oantabrigiensis, Trinity College, R. 4. 4, saec. XIII incip. No. 26 in Analecta. Parisinus 5732 Colbertinus, saec. XIII. No. 2 in Analecta. Parisinus 5736, saec. XIII vel XIV. No. 6 in Analecta. The numerous MSS of the fourteenth and fifteenth centuries are all referred to under the numbers given by Mornmsen in the Analecta, and for these the reader is referred to the Analecta, pp. 34-37. MSS of the Thibd Decade of Livy 407 the portion covered by the readings given in the Analecta, fol- lows its original with great fidelity and has comparatively few departures from the readings of P. Of these few variations a considerable proportion are either wrongly or inadequately re- ported in the Analecta, and, in consequence, the real position of this MS in the Puteanus group was overlooked. This oldest descendant of P is of no value for the text of Livy owing to the fact that it is more mutilated at the beginning and end of the decade than the Puteanus itself. Its readings do not appear in the critical apparatus, and the interest in it has been of a purely paleographical nature. For this reason the errors in the report of it in the Analecta have passed unnoticed. In the stemma of Luchs it is recognized as a copy of P, and the parent of M, but it can readily be shown that, of the four manuscripts of the P group of which Luchs makes use in his critical edition, all but C are descended from P through Vat. Keg. It can further be shown, even from the Analecta, that the majority of the great number of manuscripts of the P group trace their descent through this ninth-century copy. In his stemma Luchs recognizes that B and X are descended from P through a common ancestor, but fails to recognize that the common ancestor is Vat. Reg. The descent of \ from Vat. Reg. is easily shown, and to avoid the errors of the Analecta I shall make use of the readings of \ as they are given in the criti- cal apparatus of Luchs's edition, and my own readings of Vat. Reg. I shall give those readings only in which Vat. Reg. shows a departure from P. For convenience the readings of Vat. Reg. are designated by the letter R. non reuehi P,* reuehit omitting non R, X. reum P, rerum R, X. Scipionem bellis P, scipion | bellis R, scipio non bellis X. in acie P,* acie R, et ex aciem X. umbrium P 2 , ubrium R, X. insiciliamessenam (=Messanam) P, in siciliam esse nam R, X. doleraeuitae P, dolere \ uitae R, dolore uitae X. accepere P,* acceperet R 1 , acciperet R 2 , X. Reading of P inferred from Luchs's silence. 28. 24. 7 25. 7 8 26. 14 27 5 28. 6 29. 2 33. 14 2. 2 2. 13 12. 8 17. 9 22. 5 IB 408 P. W. Shipley 33. 15 nequ.es P l , neque eques P 2 , neque acies R, X. 16 ca | pessendae steterat P, capessendaest et erat R, capessenda est et erat X. 34. 12 pecuniam ilergetes P,* pecunia | milergetes R, pecuniam inilergetes X. 35. 12 continentem] contionemtem P, contionem tern R 1 , contio- nantem R 2 , contio natem X. 29. 2. 2 peruenere et] perueneret P, R 1 , perueniret R 2 , X. neu resutui acie P 1 (neu restitui acie P 4 ) neu res ut ui acie R, neu res ut in acie X. epirotae] epiroe P epikoe R, X. respecietis P, recipietis R, X. consulem P* (probably written cos), quos R, X. fauorem] fatio | rem P 1 , supra o prima manu u uidetur scripta fuisse, facilio | rem P 2 , facillionis R, X. From the above examples it will be seen that the readings of \ either agree with those of Vat. Reg., where they depart from those of P, or have their starting-point in errors which originated with it. The readings rerum 28. 27. 7; neque acies 28. 33. 15; epikoe 29. 12. 8; and facillionis 29. 22. 15 are especially strong evidence of the descent of \ from Vat. Reg. That X is not a direct copy but is descended from Vat. Reg. through an intermediary of as late a date as the eleventh century is shown by the progress of error in 28. 34. 12 pecunia milergetes R., pecuniam inilergetes X, and in 29. 2. 13 ut ui acie R., ut in acie X. In both instances the error points to forms of m, n, u, i, which are at least as late as the eleventh century. That B and X were descended from a common ancestor had already been recognized by Luchs, and may readily be shown without resorting to the Analecta by a comparison of the readings of C, B, X, given in the critical apparatus of Luchs's edition at the end of the decade, which is missing in P. A few omissions and lacunae will be sufficient to show the almost constant agreement of B, X against C 30. 30. 20 ferrum utrimque C, om. B X. 25 ea B C, om. X. 32. 4 spes B X, om. C. 'Only those passages are cited in which Luchs has given the readings of all three MSS. 35. 4 36. 8 8 9 37. 5 11 39. 6 40. 8 9 41. 1 MSS of the Third Decade of Livy 409 proelio et C, et om. B A. mille vel oo C, • X • B, om. X. mille et quingenti] (probably oo et D in P) C, • X • et B, decern X. quo magis] quam magis B X, magis C. quinquaginta C, om. B X. arguerentur] argueretur P C, spatium relictum B, om. X. C. Sulpicius Galba] c. C, om. B X, sed in B spatium relictum. prouincia C, om. B X sed in B 15 litt. spat, relictum. anno C. om. B X. prorogatum imperium] pro X, inter prorogatum et imperium B 7 litt. spat. rel. 4 acidinum esset uti consules C, ace (spatio relicto) sule B, om. X. 4 his uideretur ut plebem rogarent cui iuberent C, his uidere- tur ut pleb (spatio relicto) berent B, om. X. 5 ex duobus exercitibus C, om. B X, sed in B spatium relictum. These lacunae and omissions clearly point to a common ancestor for B and X later than P, and, as the descent of X through Vat. Reg. has been proven, it is clear that B also is a descendant of that manuscript. B like X is not a direct copy of Vat. Reg., but is removed from it by at least one intermediary of the tenth or eleventh century, as is shown by such errors as 30. 6. 9 where B has niagis for magis.' The stemma of the older manuscripts of the P group as given by Luchs should accordingly be corrected as follows: V.t Beg. 762 . Vat. Beg.' and C remains the only one of these manuscripts whose descent from P is not through Vat. Reg. 762. These relationships have all been established without resorting to the Analeda, on the basis of the readings of P, C, B, X, as given 1 In a paper in the Transactions of the American Philological Association for 1902, on "Numeral Corruptions in a Ninth-Century Manuscript," I had already pointed out that the freakish errors of the scribes of Vat. Reg. were duplicated by similar errors in the Bambergensis ; but as the readings of the Bambergensis were from a portion of the text that was missing in the Vat. Reg,, and I had then no other reason for calling in question the stemma of Luchs, I was obliged to regard it simply as a coincidence. 410 F. W. Shipley in the critical apparatus of Luchs, who is usually most careful, and my own collation of Vat. Reg. The fact that Vat. Reg. is the ancestor of B and X. as well as of M having been established, it is possible to show that it is also the ancestor of a much larger number of manuscripts. But the read- ings of these manuscripts are either not given at all in the critical editions, or are only given sporadically, when they happen to show traces of the Spirensian tradition, and one is forced to fall back upon the Analecta as the only available evidence. In order to make use of the Analecta for this purpose, it is necessary to elimi- nate from the readings there given for Vat. Reg. the inaccuracies in the collation submitted to Mommsen either by Schoell or Jeep, who collated for him the manuscripts in the Roman libraries, and the clerical blunders in digesting its readings under the numerical system which Mommsen adopted. The corrected readings are given in paper II of this series on p. 413. With the readings of Vat. Reg. thus corrected, one can trace with greater accuracy the relationships of the later MSS of the P group, bearing in mind, of course, the possibility of similar errors in the Analecta, in handling the reading of these MSS. It will here be necessary to make use of the numerical designations adopted by Mommsen. Among the passages given in the Analecta the key to the MSS descended through Vat. Reg. is to be found in 29. 22. 8 memoria- que absentis Scipiom's fauorem ad uulgum conciliabat. The reading of P is given by Luchs as follows: fatio | rem P 1 (supra o prima manu u videtur scripta fuisse) facilio \ rem P 2 . The scribe of Vat. Reg. wrote facillionis, a blander resulting apparently, by a parablepsy, from facilio | of the reading of P 2 , and the nis of Scipionis. This senseless reading, which thus originated in R, and which is not of the sort likely to be perpetuated except in direct copies, is repeated either as it stands, or with modifications, in a large number of manuscripts, as follows: facillionis M, B, X, 2, 12 a , 26, 27 a , 62 a , 81; facilionis 38, 67 b , 72 a , facilioris 44; facilliomnes 29 a , facili omnes 74; facile omnes 6 a , 29 b , 79; facilior dis 67 a . Equally significant is 27, 33, 10 in necopinatam fraudem MSS of the Third Decade of Livy 411 lapsos esse, but its significance is lost in the Analecta, owing to the fact that the reports for both P and Vat. Reg. are incorrect. Luchs gives for P innecopianas | tarn P 1 , sine copia nas tarn P 2 ; R 1 had sinnecopia nata, which was altered by R 2 to sine copia nata. This reading is, according to the Analecta, found in the following MSS: M, B, X, 12 a , 14, 23, 26, 27 a ( ?), 29 a ( ?), 31, 35, 38, 44, 46, 49 a , 51, 52, 59, 62, 67, 70, 71, 74, 75, 79, 80, 81. In 27. 34. 5, P has opsoleta, Vat. Reg. had also opsoleta, but this was changed by the first hand to obsolita (the report in the Analecta is here misleading). This reading obsolita occurs, with modifications, in the following MSS: obsolita M, B, X, 3, 4, 12 a , 38, 44, 46, 56, 71, 74, 77, 80; obsoluta 18, 49 a , 70; absoluta 11, 14, 75; absoleta 82; absolita&l, 81; abseleta 34; absoleta 53 b ; obselata 58. These passages alone give as probable descendants of Vat. Reg. the following MSS: M, B, X, 2, 3, 4, 6, 7, 11, 12. 14, 23, 26, 27 a , 29 a , 30, 33, 35 (?), 38 a , 39, 44, 46, 49 a , 51, 52, 55, 56, 59, 60, 62, 67, 70, 71, 72 a , 74, 75, 77, 79, 80, 81. There are in all ten MSS of the P group of as early a date as the thirteenth century, namely, Vat. Reg., C, M, B, X, 27 a ( = H of Luchs), 55 (Vat. 1487), 26 (Cantab., Trin. Coll., R. 4. 4), 2 (Paris. Colbert. 5732), 6 (Paris. Colbert. 5736). Of these, C alone does not trace its descent through Vat. Reg. M had already been recognized as a copy of the latter MS by Luchs. Luchs had already shown that 27 a and 55 were copies of X, though he failed to recognize their relationship to Vat. Reg. The other thirteenth-century MSS, 26, 2, 6, form a group which is closely related to the Bambergensis. It is highly probable that the Bam- bergensis was itself the archetype of these three MSS. The fol- lowing readings in the Analecta, which are peculiar to these three MSS and to B, point pretty clearly to this relationship, but, owing to the possibility of further errors in the Analecta, absolute proof is not to be had without a new collation of this portion of the MSS in question. 27. 33. 10 praepropere P, R; praepraepropere B, 26, 2. 34. 2 et conlega P, K; set conlega B 1 , sed conlega B 2 , 26, 2. 8 uiro iussa R; uiri iussa B(?), 26, 2, 6a. 412 F. W. Shipley The remaining MSS of the fourteenth and fifteenth centuries, which I have shown to trace their descent through Vat. Keg., are so late and show so many signs of interpolation, emendation, and crossing that it is hard to trace their further relationship definitely. Practically all, however, are more closely related to the Lauren- tianus than to the Bambergensis or the Mediceus. They fall into groups as follows: (a) 7,12,74,31; (b) 67,4,23; (c) 33,3,72; (d) 11, 14, 71, 75; (e) 19, 30, 44, 53, 56, 59, 60, 61, 69, 77; (/) 10, 18, 29, 39, 47, 49, 51, 70. The stemma given below indicates in detail the genealogy of the MSS of earlier date than the end of the thirteenth century. P - P.» (s. ») V»t. Reg. 762 . Vat. Beg.' (a. ii) M (e. xi) o) 7,12,74,31 b) 67,4, 23 c) 8, 33, 72