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. HIEREMIAS DE MONTAGNONE AND HIS CITATIONS PROM CATULLUS By B. L. Ullman Hieremias de Montagnone, a Paduan judge, put together, about the year 1300, a book of quotations from ancient and mediaeval writers which he called Compendium moralium notabilium. Per- haps the most interesting fact about this work for the classical student is that Hieremias quotes seven passages from Catullus, and thus furnishes us one of the first testimonia to that author after the resurrection of his work early in the fourteenth century. The quotations are all the more important because none of the extant MSS of Catullus was written before 1350. 1 The definite contributions which I have to make on the subject under discussion are, first, the report of an examination for the Catullus citations of a number of hitherto unused MSS of Hieremias' work; second, a solution, which I believe to be both simple and con- clusive, of the problem which these citations present to us and with which all students of the MS tradition of Catullus are familiar; and third, some interesting sidelights on the " lost Verona" MS of Catullus. The net result of this investigation will in my opinion be of great importance to Catullus students, for it will end in the exaltation of O to a point higher than any reached by it heretofore as a witness of the text of the Veronensis, and thus of Catullus himself. First of all, it is necessary to date the work of Hieremias as closely as possible. The authority of Scardeonius,' 2 who says that Hiere- mias died about 1300, was followed by Ellis in his large edition of Catullus (1878) 3 and by Wheeler, the most recent writer on the subject. 4 It has, however, been definitely settled by Rajna, 5 by means !This is the earliest date that has been suggested for the Oxoniensis (O), Hale Classical Review, April, 1906. But he has since given reasons that make against so early a date, Class. Phil. Ill (1908), p. 243. 2 De Antiquitate Urbis Patavii, ed. Basil. (1560), p. 235. *Proleg., p. ix. <" Hieremias de Montagnone and Catullus," AJP. XXIX (1908), p. 186. * Studi di Filologia Romanza V (1891), p. 193 ft. Ellis Catullus in the XlVth Cen- tury (1905), p. 8, mentions Rajna and adopts his conclusions, but curiously enough fails to give the reference. I am indebted for it to the late Professor Traube. [Classical Philology V, January, 1910] 66 HlEEEMIAS DE MONTAGNONE 67 of official documents of the city of Padua, that Hieremias became judge in 1280 and died in 1320-21. Between these two dates the Compendium must have been produced, for Hieremias was already judge when he published the work, as we know from the title itself. Rajna adopts 1290-1300 as a safe compromise between the two ex- tremes, though admitting that he has no real argument for his choice. Granting that this date is approximately correct, must we therefore assume that Hieremias read Catullus before 1300? By no means. There is direct evidence that the Catullus citations and perhaps others were added later. One of the MSS of the Compendium, Bodleian Canon. Lat. 212, of the fifteenth century, 1 omits every one of the seven Catullus passages. Citations from other authors also are lacking, which appear in the other MSS. 2 That these omissions were accidental is out of the question. That they were intentional on the part of the copyist is very unlikely. It would seem, then, that this MS is descended from an early edition of the Compendium, made before the discovery of Catullus, or at least before Hieremias had read that author. How many editions in various states of completion there may have been it is not possible for me to say. There is, however, evidence of one intermediate edition. Casana- tensis 312 (C. iv. 11), a MS dated 1398, omits the two Catullus passages that Hieremias cites last. It is conceivable, of course, that these omissions are accidental, but it is more probable that this MS is descended from one which was copied from the original before Hieremias had inserted all the quotations in his work. 3 Further light should come from a thorough investigation of the MSS of the Compendium. In my opinion the Catullus passages were added after 1300, perhaps as late as 1310. The Verona MS of Catullus does not seem to have been discovered before that time; at least, 1 This MS was evidently known to Ellis, for in his book, Catullus in the XlVth Century, he speaks of two Bodleian MSS — the other being Oanon. Miscell. 186, used for his large edition of Catullus. He, however, says nothing of the omission of the Catullus passages. 2 1 noted omissions of part of the Sallust quotations and the one from Vergil's Georgics in iii. 4. 8, of those from Ovid De Rem. Am. and Paul Ad Eph. in iv. 4. 8, and of the one from Ecclesiastes in iv. 6. 3. 3 The omission of one passage (51. 15, 16) in Paris, N. a. 1. 1779 (dated 1475), and of another (39. 16) in Paris, lat. 6469 (of about the same date), I consider accidental. 68 B. L. Ullman the earliest datable reference to it is by Bencius Alexandrinus, who examined it shortly before 1315. 1 The puzzle about the Catullus passages in the Compendium has been the manner of citing, not according to poems as we cite from Catullus at present, but according to chapters, capitula, which do not correspond to our division of the poems. There cannot be the least doubt that Hieremias used the term capitulum, not liber, which is found only sporadically in the MSS, as may be seen from the reports given below. That the abbreviations c. and ca. are to be taken as capitulo and not capite, is perhaps not entirely proved by the abbreviations ca° and c° which are found in some MSS, nor by the occurrence of the full word in one passage in the Venice edition. These facts may merely show that the original abbrevia- tion (probably c.) was taken as capitulo by the scribes. Hiere- mias' use of the same form in quoting from other works, prose and verse, and the interpretation of it as capitulo by the scribes and the Venetian printer, in these passages also, serve to confirm this interpretation. Moreover, a common use of capitulum as applied to poetry not only suggests that this is the word that Hieremias meant, but also shows how it is to be understood. The use referred to is in designation of one of a number of poems. For example, in codex R of Catullus, R 2 (fourteenth century) has written cap" 1 in the margin of poem 67 to indicate that a new poem begins there. Codex F of Propertius (fourteenth century) uses capitulum as a term for a new poem. 2 The word is regularly used in this sense in a MS of Petrarch's Trionfi (Bodl. Canon, ital. 70) . 3 The forms carmen and charta (carta) , which may be thought of as possible expansions of the abbreviation c, may be summarily dismissed. The latter will not solve the riddle of the Catullus citations. The former seems not to have been used by Hieremias at all, even in quoting from Horace's Odes, if the Venice edition is to be trusted, for in this we find such forms as Od. li. 2. c. 10 followed by Od. li. 2. cap. 16 — the passages being from odes 10 and 16 respectively of the second book. 1 Sabbadini, " Bencius Alexandrinus und der cod. veronensis des Ausonius," Rhein. Mws. LXIII (1908), p. 225. 2 Lib. Hi. c. 4. So also Baehrens' critical apparatus in his edition of Propertius (1880). 8 The term, found also in other MSS of the Trionfi, was no doubt used by Petrarch himself. See Appel Die Triumphe Francesco Petrarcas (1901), p. 2, note. HlEBEMIAS DE MONTAGNONE 69 We see, then, that the natural way to take the word capittdum is as a synonym for carmen, poem. Since, however, this explana- tion does not correspond with the poem division of Catullus as shown by the MSS, various other interpretations have been suggested. By water 1 and Baerhens 2 thought that the quotations were taken from a florilegium divided into chapters, but Ellis 3 rightly objected to this explanantion on the grounds that we have no evidence of the existence of such a florilegium, and especially because the capitula are numerically arranged to agree with the order of the poems in our Catullus MSS, i. e., the farther on in the Catullus collection a given poem occurs, the larger is the number of Hieremias' capitulum quoting from that poem. Ellis himself thought that the citations were taken from a complete MS of Catullus, though he did not explain the numbering of the capitula. In his later work* he suggested that the "lost Verona" MS, from which he believed Hieremias drew his citations, "was divided into short books or sections, which fell out from the later transcripts, giving way to the division into separate poems, with their titles, which also formed part of the same codex." He also suggested that these sections were the original libelli of which the collection was formed. Wheeler 5 put forth the theory that Hieremias' MS was entirely different from the MSS we possess, that it was, in fact, divided into ten sections on the basis of the meter employed, and that these were the capitula of Hieremias. This theory is based on the false premise that Hieremias speaks of ten capitula, whereas he speaks of twelve. 6 A great deal of twisting and squeezing is also necessary to make the scheme fit the facts. Besides, it is inherently improbable that Hieremias' MS was not a descendant of the Veronensis. Peiper 7 expressed the opinion that Hieremias himself divided the work into convenient sections, though making use of divisions already existing in his MS, and perhaps merely numbering sections marked off by titles. It was only inherent probability that led Peiper to this conclusion, for he found no support 1 Ellis' edition (1878), Proleg., p. xi. 2 Catulli Veronensis Liber (1876), Proleg., p. Iviii. s Ibid. 4 Catullus in the XlVth Century, pp. 7, 8. &AJP. XXIX, pp. 186 f. 6 See the reports from the MSS at the end of this article. 7 Q. Valerius Catullus, Beitr&ge zur Kritik seiner Gedichte (1875), p. 22. 70 B. L. Ullman for it in the MSS of Catullus, as he himself confesses: "Sein codex musste dann allerdings sehr stark in dieser beziehung von den uns bekannten abgewichen sein." He goes on to say that the Veronensis probably had fewer titles than the existing MSS. If Peiper had examined Hieremias' usage for other authors more thoroughly at the time when he wrote his book, he would have found strong support for his suggestion. He did, it is true, note that the Tobias of Matthaeus Vindocinensis had been divided by Hieremias into chapters for the purpose of quotation, but he did not work out the details of this division. It seems strange that no student of Catullus, and especially Peiper, has called attention to the fact that Seneca's Tragedies and Terence are quoted by capitula exactly as Catullus is. The Seneca citations were later published by Peiper himself, 1 who noted that Hieremias numbered the separate scenes and cantica in his MS as "chapters." The same is true for Terence. A few illustrations from the Andria and the Phormio will suffice. And. 67 is cited a number of times (ii. 2. 1 ; iii. 3. 5, etc.) as being in chap. iii. In the MSS, the periocha forms "chapter" i; the prologue "chapter" ii. and Act I, sc. 1 (where vs. 67 is found) "chapter" iii. Vs. 191 is said (iv. 5. 5) to be in chap, iv, and it is in the fourth division of the MSS; 266 is in the sixth " chapter," accord- ing to Hieremias (iii. 1. 8) and the MSS; 305 and 307 are in the seventh (iv. 4. 10; iv. 5. 11) ; 426 in the tenth (i. 3. 4) ; 555 in the fourteenth (iv. 5. 5). Phorm. 41 is quoted from chap. iii. (ii. 5. 3; iv. 3. 1). In the MSS the periocha is the first section (the didascalic notice, written in capitals, being part of the title), the prologue is the second, and Act I, sc. 1 (in which 41 occurs) is the third. Thereafter 77 is in the fourth chapter according to Hieremias (ii. 6. 12) and the MSS; 203 in the sixth (v. 1. 1); 241 in the seventh (v. 3. 1); 454 in the tenth (i. 3. 7) ; 562 in the thirteenth (ii. 3. 9) ; 696 in the seven- teenth (iii. 6. 5). The agreement between Hieremias' statements and one group of Terence MSS continues throughout. I shall have more to say about the citations from Terence as well as some from other authors at another time. We may state Hieremias' principle of numbering in the case of those poetical works in which no scheme of numbering according to 1 De Senecae tragoediarum uulgari lectione (A) constituenda, Breslau (1893), p. 21. HlEBEMIAS DE MONTAGNONE 71 books or otherwise existed, as follows: he gave a consecutive number- ing to the sections formed by the interposition of red-letter titles or headings. For the sake of convenience I group here the passages which Hieremias quotes from Gatullus and the "chapters" to which he assigns them. The text is given farther on. 22. 18-20 = cap . V 39. 16 = " V 51. 15-16= " V 64.143-48= " viii 66. 15-16= " ix 68.137 = " ix 76. 13 = " xi, et penult. Hieremias' MS, therefore, was divided, presumably by red titles, into twelve sections. Did the "lost Verona" MS present any such appearance? We know from the agreement of MSS O and G, 1 that many, though by no means all, of the poems were separated from each other by a blank space of one verse. This divided the MS into at least 27 sections — obviously not the ones Hieremias numbers. Other poems were distinguished merely by some mark, as in O. In these places neither O nor G left a space. The omission by of all titles, even at the beginning, makes it more difficult to say what titles the Verona MS had. Nearly all the titles in G and all those in R are by second hands. Before inserting these red-ink titles, R 2 (in a few cases R 1 ) put catch-titles of black ink in the margins as a guide. R 1 supplied the catch-titles for 4, 5, and 6. 2 R 2 altered the form of these somewhat in copying them. In G at these points it was G 1 , not G 2 , who wrote the titles, and what is more, in the form in which R 1 gives the catch-titles. The bookheading at the beginning (Catulli Veronensis liber Incipit) was also written by G 1 , but no other titles were made by him. Evidently these titles were found by GR in their archetype, and probably they were also 1 It is universally conceded that O and G are very close to the " lost Verona " MS. That R is the ancestor of most of the other MSS, including M and D, and is on a par with G, is a conclusion which is bound to be accepted sooner or later by all scholars. Professor Hale believes that O is a direct copy of the Veronensis, and that G and R are copies of a lost copy of the Veronensis (Class. Phil. Ill, pp. 233 f.). His position is adopted in this paper. 2 Hale Class. Phil. Ill, pp. 246 f. 72 B. L. Ullman in the Veronensis. Probability becomes certainty when we examine O. This MS leaves room for a colored initial, by putting the first letter in the margin as a catch-letter and by indenting the first two lines in poems 1, 2, 4, 5, 8 (three lines are indented in the first poem). In poems 7 and 9-60 there is no space left for an initial, and there is no catch-letter, the first letter being part of the line. 1 The only thing that distinguishes the poems is the blank space of one line preceding them and the use of very slight catch-paragraph marks (//), which, however, are missing in 7. The colored initials were supplied for 1 and 2 (at a later time?) but not for the others. In as faithful a copy as certainly is, 2 we have a right to believe that the difference in treatment of poems 1, 2, 4, 5, 8 is based on a difference in its archetype, the Veronensis. We see from GR that the Verona MS probably had titles for 1," 4, 5, and 6. Thus OGR point to an unusual appearance in the Veronensis for 1, 4, and 5. This unusual appearance, to my mind, was the presence of colored titles and initials. According to O, the same was true for 8, judging from its indention there, but according to GR, it was true, not of 8, but of 6, judging from their first-hand titles to 6. O does not even leave a space before 6, but merely has the catch-paragraph marks. In other places where this occurs in O, G 1 has left no indication of a separation, G 2 putting in a marginal title and a paragraph mark. In this case, however, G' left a space, inserted a title, and indented for the initial. The title which R'G 1 have is ad se ipsum, changed by R 2 G 2 to ad flauium. The ad se ipsum title obviously does not belong to 6, but to 8, which is just a page farther on, and there is where it was placed by R 2 G 2 . O is therefore right: the Verona MS had no break at 6, but did have a break with title ad se ipsum and initial at 8. The archetype of GR wrongly placed this title one page farther back, before 6. Again the faithful points to title and initial in the Verona MS for poem 2, though no title is given by R'G 1 . No doubt the Veronensis had one, and the archetype of GR failed to copy it. These titles and 1 Except that in 13, 14, and 15 there is a catch-letter, but without indention. 2 See below, p. 77. 8 Bather a book-heading, not title, above 1. That the Veronensis always had a book-heading is not certain. See below, p. 78. The argument here made is not affected. HlEEEMIAS DE MONTAGNONE 73 initials thus divided the first 60 poems of the Veronensis into five sections, as follows: Poem 1 = Section 1 " 2-3 = " 2 « 4 = " 3 " 5-7 = " 4 " 8-60= " 5 It will be seen that Hieremias cites 22. 18-20; 39. 16; 51. 15, 16 from capitulum 5, and that all three fall into our fifth section. Thus the greatest difficulty about the citations has been explained, the fact that four capitida preceded poem 22, while one capitulum included at least 22 to 51. Let us now see how the rest of the Veronensis must have appeared. In O poem 60 ends on fol. 14v, the rest of the page (five lines) is left blank, and 61 is started on the next page after another space of one line. Nowhere else does O leave a space at the bottom of a page merely in order to begin a new poem at the top of the next page. For example, poem 50 begins two lines from the bottom of fol. 12» after a space of one line. It seems to me that we are justified in assuming that there was more than a mere space of one line in the Veronensis before 61, that very probably there was a title. This would be the beginning of section 6. It may well be that the Veronen- sis also began 61 on a new folium, after leaving several lines blank on the previous page, as in O. In this state of affairs we may perhaps see a trace of the putting together of the Catullian liber. Perhaps one of the ancestors of the Veronensis, or even the Veronensis itself, was put together from two (or more) MSS, the first one (or ones) containing poems 1 to 60, the second one (or group of ones) con- taining poems 61 to 116. Possible confirmation of this suggestion is found in the Explicit epithalamium which follows 61 in O. This gives a distinctive air to the poem, as if it had once stood alone, for nothing similar to it is found with any other poem. 1 At any rate, the phrase (perhaps in red ink in the Veronensis) emphasizes the division between poems 61 and 62. Thus our section 7 begins with 62. G begins 63 as a new poem, with space and initial, but O merely has the catch-paragraph marks without break. We may > Another hint of the same sort is found below, p. 76. 74 B. L. Ullman follow the more trustworthy and consider 63 as part of section 7. In the case of 64, O simply leaves a space of one line. This would not be sufficient to permit us to assume a new section beginning with 64. But at the beginning of this poem there are glosses and variants in by the same hand which wrote glosses and variants at the beginning of the book. 1 These notes at the beginning are surely derived from the archetype, the Veronensis. (Cf. in 3. 14 the superscribed i. fulcra in both O and G [by G 1 ] and the striking agreement of the marginal notes to 2 with the remarks of Guilelmus de Pastrengo, who, therefore, must have used the Veronensis or O itself). 2 It is thus extremely likely that the glosses in 64 also were to be found in the Verona MS. These alone would attract attention to the poem and would serve to distinguish it from the preceding poem, even if there was no title. The marginal note to the first line, especially (narrat hie ystoriam aurei velleris), attracts attention to the break. Thus the eighth section of the Veronensis began with 64. Hieremias quotes 64. 143-8, assigning the passage to capitulum 8. Beginning with 65 a different method of indicating new poems is adopted in O. In addition to the one line space, there is an illumin- ated initial (different in style from those in poems 1 and 2) in the margin; the second letter is a capital in line with the first letters of the other lines; then the rest of the word is written close to it without the usual space between. I give an illustration from 72: klcebas L esbia D ilexi Other poems are distinguished in the same way, with two exceptions: there is no one-line space preceding them (nor is there any in G) and instead of an illuminated initial there is only a catch-letter. It is possible that the striking difference between the treatment of new poems in the earlier and later parts of the book is due merely to the caprice of the scribe of O. Two reasons lead me to believe that this is not true; first, because O in other respects is so faithful in reproducing just what he found; and second, because the reading Vitas instead of Multas, in 101. 1, which both G and R (and therefore 1 See facsimile of the page in Chatelain PaUographie des classiques Latins XV. A, or Merrill's Catullus (1893). 2 See Schulze in Hermes XIII, p. 57. D! HlEREMIAS DE MONTAGNONE 75 their archetype) originally had, makes it certain that the Veronensis had the word in a form very similar to that of 0: m Vitas (catch- letter m) . A second possibility is that the scribe of the Veronensis introduced the new system into the later poems, or that these were written by another scribe. Lastly, it may be that the Veronensis or one of its ancestors was put together from separate libelli, as has been suggested above (p. 11), and that a new libellus began with poem 65. However this may be, O gives us little help from 65 on, in determining the sections of Hieremias. We may assume that 65 with its new system began a new section — section 9. Hieremias quotes 66. 15, 16 from capitulum 9 (there is no break at 66). 68. 137, also, is quoted from the ninth "chapter." There is no break at 67 in OG, but there is one at 68 which we must ignore. There are breaks in OG at 69 and 72, which began sections 10 and 11 of Hiere- mias' division, for 76. 13 is quoted from "chapter" 11, which at the same time is called the penultimate "chapter." After 76 there are breaks at 77, 80, and 89 in OG. The twelfth and last chapter must begin with one of these. Our choice is not difficult: in R there is found in the margin of 77 a catch-title ad ruffum by the first hand. 1 As this title no doubt comes from the Veronensis, we are safe in letting section 12 begin with this poem. G's failure to preserve the title suggests that he failed to preserve other titles. If the archetype of GR was equally negligent occasionally, we can see why no titles have come down to us for some of the other sections. The scheme of division, then, that Hieremias, following the indica- tions of his MS, used was as follows: oer a 1 = Capitulum i u 2-3 = ii u 4 =r iii u 5-7 = iv a 8-60 = V a 61 = vi a 62-63 = " vii a 64 = " viii u 65-68 — ix u 69-71 — " X a 72-76 = xi a 77-116 xii Whether Hieremias imposed this numbering on the Veronensis itself or on a copy it is impossible to determine. 1 Hale op. cit., p. 247. 76 B. L. Ullman Let us stop to summarize briefly. 0, by its indentions, shows us that the first sixty poems of the Veronensis were divided into five sections. In four cases out of the five, confirmation is added by three first-hand titles in G and R which must go back to the Verona MS. After poem 60 there are the following indications of new sections : the unusual space of five verses in after poem 60, with the conse- quent beginning of 61 on a new folium; the " Explicit epithalamium " at the end of 61 in 0; the glosses in the margin of in 64; the new method of beginning poems in O from 65 on; the first-hand catch- title in R at 77. Thus we have indications of various kinds for the beginnings of ten of the twelve sections. We have had no serious conflicting evidence to be explained away, nor do we have any for the remaining two sections. We are confronted here merely by a lack of evidence. We know from the MSS that there may have been new sections at 68, 69, 72, 80, and 89, and that there could have been no others. Now that we have established the agreement of Hieremias' sections with those indicated by OGR for the Veronen- sis, we may be allowed, on Hieremias' authority, to select 69 and 72 as the beginnings of sections, and to reject 68, 80, and 89. It is to be noted that the argument is based on the numbers which Hieremias indubitably assigned to the capitula, according to the evidence of the MSS, and that in no case has emendation been resorted to. Two possible objections to the explanation given must be antici- pated. In a Verona florilegium of 1329 we find the phrase Catullus ad Varum followed by a quotation of 22. 19-21. This quotation was no doubt taken from the " lost Verona " MS. Since in G also we find the title Ad Varum (in the margin by G 2 ) it has been generally assumed that this title existed in the Veronensis. But in R the second hand first wrote Ad suffenum in the margin; later he deleted the second word and replaced it by Varum. This is the form which he used for his colored title. Professor Hale has shown that G 2 corrected G on M, while the latter is a copy of R as corrected by R 2 . Therefore if the Veronensis had any title here, it must have been that which R 2 originally wrote, Ad suffenum, not Ad Varum. I have no hesitation in saying that it had neither. The author of the Flores had no difficulty in making up his title from the first line, possibly following the analogy of the title which he found HlEEEMIAS DE MONTAGNONE 77 for 8, ad se ipsum. In the same way R 2 arrived at his title — after being misled into making it .Ad suffenum,by the first word of the poem, which is Suffenus. The other possible objection is similar. Bencius Alexandrinus 1 introduces a quotation of 35. 1—4 with the words "Catullus poeta Veronensis ad amicum Aurelium scribens." As poem 35 is not addressed to Aurelius, Sabbadini was at a loss to explain the words. But if we look at 0, we see that there is no break at 35 and that the first preceding break is at 21, which is addressed to Aurelius, as can readily be seen from the first word, the vocative Aureli. R 2 and G 2 give titles Ad Aurelium, but I do not believe that these go back to the Veronensis. Bencius, like the author of the Flores, invented the title. 2 We may make here another comment on Bencius' quotation. In quoting 35. 2 he gives the reading occilio. This is the reading of O alone. Bencius could not have used O which was written at least a generation after his time. The other MSS give cecilio, which is the right reading. O's reading, then, is not an error on his part but is derived from the Veronensis. The reading cecilio must be an emendation, 3 perhaps from vs. 18, where the name occurs again, though corrupted in the MSS to cecilia. I mention this here to illus- trate O's striking trustworthiness, thus supporting my case as just presented. If we had not the evidence of Bencius, no one would hesitate to say that the Veronensis had cecilio, not occilio. Further evidence of O's being in many ways the most faithful descendant of the Veronensis that we have is presented by an exami- nation of the Hieremias quotations as I restore them from the MSS. The form nee in 22. 18 which Hieremias surely wrote is the reading of O as against the neque of GR and the editors. I believe that O has the reading of the Veronensis. In 64. 145 Hieremias seems to have had ftgestit, as in O. The abbreviation probably was meant for pregestit, but in O should be taken as postgestit, which GR originally had. In the same line O has adtpt'sd P a(Upiscl , which the Veronensis probably had in the same form, since G originally had 1 Sabbadini loc. cit. Bencius certainly used the Veronensis. 2 See also, in the present number of this Journal, Hale's fuller discussion of Bencius' citation, in the article " Benzo of Alexandria and Catullus," which is based upon an argument of the same nature. 3 1 do not believe that the Veronensis had a double reading. 78 B. L. Ullmak adipisci, while R has apisci. Hieremias seems to have had apisci a1 ' adi Pi s ci. Only a reading very similar to O's will account for the readings of GR and Hieremias. In the same way in 60. 5 the conteptam of R and the contentam of G (corrected by G 2 to contep- tam) are best explained by assuming a reading like O's (contenf- tam) for the Veronensis and the archetype of GR (the caret perhaps being omitted in the latter) . The form Catulus which Hieremias used is probably not a slip on his part. Perhaps the Veronensis did not have a book-heading at the time when Hieremias examined it or got his copy from it. O has none by the first hand. It is to be noticed that Hieremias is the only one of his period, except the compiler of the Verona Flores, to mention Catullus without adding Veronensis or poeta Veronensis. 1 This designation seems to have come into the Verona MS as a book- heading with the epigram of Benevenuto de Campesanis (cf. the title of this poem in G: — Catulli poete Veronensis). If Hieremias examined the Verona MS before it had a heading he would have had to look into the poems themselves for the name of the author. O in the majority of cases where the name occurs in the poems has the form Catullus, but in the first three cases (6. 1; 7. 10; 8. 1) has the single I. The last case is particularly striking. It is in the first line of the poem which, as has been shown above, was introduced by the title ad se ipsum in the Veronensis. If in our faith in we can trust it to have reproduced here the spelling of the Veronensis, we can easily understand how Hieremias came to use the form Catulus. In turning over the first few pages of the MS his attention would immediately be attracted to the title ad se ipsum and the vocative Catule in the first line of 8. The surprising faithfulness of O as shown in one case by the evi- dence of Bencius and in a number of cases by that of Hieremias is of considerable importance for any attempt to reconstruct the " lost Verona" MS of Catullus. This task is made much more difficult 1 So Bencius, Petrarch, Pastrengicus. This, of course, applies only to those who speak of Catullus in such a way that it is evident that they read some of his poems. It is significant that in the two places in which Pastrengicus' mention of Catullus is certainly the result of an examination of a MS of that poet he uses the expression Veronensis poeta, while in the two quotations of Catullus which are drawn from Pliny and Isidore simply the name of the author is given. HlEREMIAS DE MONTAGNONE 79 in that we can not be sure that an apparently correct reading of GR (such as cecilio), differing from O, existed in the Veronensis. In other words, the value of O is not equivalent to the combined value of G and R but is greater than it. There may be dozens of cases like occilio and nee. The following are the MSS of the Compendium that I have exam- ined: B,= Bodleian Canon. Miscell. 186, s. XV B 2 = Bodleian Canon. Lat. 212, s. XV (omits the Catullus passages) C = Casanatense (Rome) 312 (C. IV. 11) dated 1398 (omits the last two citations) Cm = Cambridge Univ. Library E e II. 29, s. XV E = Escorial II h 11, dated 1402 M = Madrid H. h. 21, s. XIV-XV 1 Mn = Munich 14317, s. XV N = Naples, Naz. VII. E 2, s. XV New = New College, Oxford (Bodleian), 100, s. XV 2 P,= Paris lat. 6469, ca. 1475 (omits 39. 16) P 2 = Paris N. a. 1. 1779, dated 1475 (omits 51. 15, 16) R = Reginensis (Vatican) 1526, s. XIV (?) V 1= Vatican lat. 4278, s. XIV (?) V 2 = Vatican lat. 1168, s. XI V ( ?) (fragmentary; ends IV. 3. 3, omitting the last four Catullus quotations) Ven — Venice edition, 1505. I take from Ellis 3 readings from: Br = British Museum 22, 801, s. XV. Prom Peiper 4 1 take readings from: Vra = BreslauIF 129 Vrb= " IF 246 Vrc = " IV P 50. 1 My notes on this MS are incomplete ; they fail to report on the second and seventh passages. 2 Ellis (Catullus in the XlVth Century, p. 9) says that this MS is dated 1400. That date is found at the end but seems rather to refer to the composition of the sermon to which it is appended. 3 Catullus (1878), Proleg., p. x. Neither Ellis nor Peiper is absolutely trust- worthy. Their silence, especially, cannot be trusted. Ignorance of this fact led Wheeler astray. 4 Loc. cit. 80 B. L. Ullman Rajna' mentions the following MSS not examined by me: Florence, Laur. Gadd. Rel. 46 Florence, Riccard. 250 Florence, Riccard. 816 Florence, Magliab. Palch. IV, cod. 128, fol. 121 (containing only the Italian proverbs) Venice, Marc. Lat. CI. VI, 100 2 Modena, Est. XII, K, 12 Valladolid (Carini Gli Arch, e le Bibl. di Spagna, Palermo (1884), I, 269) Trotti collection (now dispersed; see Novati in GHorn. Star. d. Lett. It. IX, 147) Padua, Antoniana (Tomasini Bibl. Patav. MS, p. 56; the MS has now disappeared) Milan, Ambros. P. 29 sup. (the work is there attributed to a certain Giovanni de Giapanis [Grapanis], but it is taken bodily from Hieremias). Other MSS known to me are: Leipzig, Karl W. Hiersemann (Manuscripte des Mittelalters u. sp&- teren Zeit, 1906), No. 61, dated 1376 (?) Darmstadt, dated 1410 (Osann Vitalis Blesensis Amphitryon et Aulularia, p. vii, cited by Peiper, op. cit.). The text of the quotations from Catullus is now given. No attempt at a complete critical apparatus is made, since no critical examination and comparison of the MSS was undertaken. Perhaps there are better MSS extant than any of those used, perhaps even Hieremias' own copy still exists. It seems, however, that the text of the Catullus quotations can safely be restored from the MSS listed. The oldest of my MSS seem to be the best: the Casanatensis, the two Vaticani, the Matritensis, the Escorialensis. I report only such readings as are of interest for my purpose. It is to be remembered that the text has strong MS support where no readings are given. Several facts can conveniently be grouped together at the outset. Catulus is found in 63 cases in my MSS " " " " 15 " " Peiper's MSS " " " " 4 " " Br, according to Ellis Catullus " " " 6 " " my MSS (4 in P„ 2 in E) " « " " 6 " " Ven " " " " 2 " " Br, according to Ellis. 1 Loc. cit. 2 Valentinelli Bibliotheca Manuscripta ad S. Marci Venetiarum IV (1871), p. 188, gays that this MS was written by Hieremias himself, but Rajna (p. 198, n. 3) emphati- cally denies this. c. isf ounc i in 51 case; c. a « « 5 « c. a u u lg « c. a ii " 6 " c°. a it " 8 " ca. a a « 2 " ca°. a a " 13 " capitulo a it " 1 case HlEBEMIAS DE MoNTAGNONE 81 Ven Peiper's MSS Br, according to Ellis my MSS (5 in C, 3 in V,) B, my MSS (6 in Mn, 7 in V,) Ven. 1. or li. is found in 6 cases in my MSS (1 in New, 1 in Cm, 1 in Mn, 2 in R, 1 in M; 4 of these are in the third quotation and probably point to a common archetype for some of the MSS in which they occur) 1' is found in 1 case in Ven (first quotation) li. is found in 1 case in Br (first quotation), according to Ellis. I. 3. 8 Catulus c. V. Omnes fallimur nee est quisquam quern (Cat. 22. 18-20) non in aliqua re videre suffenum possis suus cuique at- tributus est error. V CV 1 V i EN New Cm B, P 2 Mn Br 5 MR Pi Ven 7 Vra Vrb Vrc 1 nee CV 1 V 2 MEN New CmB^i Mn Vra Vrb Vrc neque R P 2 Ven. II. 1. 5 Catulus c. IX. Ne minium simus stultorum more mo- (Cat. 68. 137) lesti. IXCV^E New Cm P 2 Villi V x N B x Mn 9 i? Pi Ven Br Vra nono Vrb Vrc. III. 4. 8 Catulus c. V. Ocium et reges prius et beatas perdidit (Cat. 51. 15, 16) urbes. V C Fi F 2 E N New Cm B, Mn 5 MR Px Ven Br. IV. 4. 8 Catulus c. V. Bisu inepto res ineptior nulla est. (Cat. 39. 16) V CEN New Cm B t P 2 Mn 5 ViMR Ven Br 7 Vra Vrb Vrc 2 . IV. 5. 11 Catulus c. VIII. Nulla viro iuranti femina credat (Cat. 64. 143-48) Nulla viri speret sermones esse fideles Quis dum aliquid cupiens animus j?gestit apisci * 1- * dipl,cl Nil metuunt iurare nil promittere parcunt Set simul ac cupide mentis saciata libido est Dicta nichil metuere nichil periuria curant. 1 Peiper probably misread his MSS. 2 Peiper probably again misread the MSS. 82 B. L. Ullman VIII CEN New Cm Pi P 2 Mn 8 Vi M R Pi Ven Br Vra Vrb Vrc pgescit E pgessit New B\ Mn Ven pregessit Br Vra Vrb pgrescit R pgressit M N pgrossit Vrc pregressit P 2 pigresoit C Vi Cm pegre scit Pi aspici. al' adipisti (in text) M apisci Vi Pi P a Br apisti N aspici R Vra Vrb Vrc adipisci C Ven ipsi E New Cm Mn 1 et ipsi Pi. (Cat. 76. 13) Idem c. XI et pe. Difficile est longum subito depo- nere amorem. XI E New Cm Bi P 2 Mn 11 Vi M R Pi Ven (Ven has Arabic 11, not Roman II, as Ellis gives) Vrc - (sic) N 12 Br Vra Vrb et pe. N New Cm B t Pi Mn Ven et pit' M P 2 Vrc et ple'P et pult' Vra Vrb et penl't F, et peult' E om.Br(f), IV. 6. 3 Catulus poeta c. IX. Est ne novis nuptis odio venus (Cat. 66. 15, 16) atque parentum frustrantur falsis gaudia lacrimulis. IXEBiPiMn Villi N 9 ViRPt Ven Br Vra Vrb Vrc i New Cm. University op Pittsburgh 1 In these four MSS, and in these only, animus (abbreviated) follows the reading iotpregestit instead of preceding it.