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 byJSTOR. 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.istor.org/participate-istor/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. ON THE EARLY PRINTED EDITIONS OP PLINY'S CORRESPONDENCE WITH TRAJAN By Elmbr Tecbsdell Mbkrill In a paper entitled "Zur friihen Ueberlieferungsgeschiclite des Brief wechsels zwischen Plinius und Trajan," which was read before the fiftieth Versammlung deutscher Philologen und Schulmanner (Graz, 1909), and was printed in Wiener Studien XXXI (1909), 250- 258, 1 discussed the traces of the descent of the text concerned during the centuries of its existence in MS form, and set down without argu- ment a few conclusions regarding the history and comparative value of the early printed editions. Of this later period the present paper is designed to treat more fully, even at the risk of some vain repetition. After some centuries of an independent but precarious existence, the book of PUny's correspondence with Trajan was appended as a tenth book to the nine books of the miscellaneous letters. From this stage in the history of Pliny's text no complete MS has come down to us, though cod. Ashburnham R. 98 (B) preserves a trace of the condition of its archetype in its title, C. PLINI SECVNDI EPISTVLARVM LIBRI NVMERO DECEM. The sister MS of B, cod. S. Marci 284 (F), is similarly abbreviated, and is without title. But from the same ultimate ten-book source as the immediate parent of B and F was descended a complete MS, which was found by Fra Giocondo, of Verona, apparently at or near Paris in the first years of the fifteenth century.* Apparently very soon after its discovery a copy was made of the correspondence with Trajan by (or for) a certain Petrus Leander,^ and sent or brought by him to Hieronymus Avantius, of Verona, who, as is well known, published in May, 1502, " C. Plinii lunioris ad Traianum Epistolae. 46. nuper repertae cum eiusdem responsis.'" 'On the relation between B and P, and that of both to this codex Parisinus, see two articles by a pupil of mine, Mr. Prank Egleston Bobbins, published in this num- ber of Classical Philology. 2 On Petrus Leander cf . Echard Scriptores Ordinis Praedicatorum, t. 2, p. 137 (Paris, 1721). 8 The first letter in the edition of Avantius is headed "9De lacu Nicomedensium Epistola . xiyii ." and is the one numbered 41 in all editions since the one first issued [Classical Philology V, October, 1910] 451 452 Elmer Tbuesdell Merrill Of the reason why only these letters were available for publica- tion in his edition, Avantius vouchsafes us not a word of explanation. This silence, added to the frankness with which he numbers his first letter as xxvii, and continues the numbering in due order there- from, suggests that he hoped to repair later the loss of the missing first part of the collection. To these indications may be added yet another that looks in the same direction. The first letter given by Avantius begins with its paragraph mark, lemma, and number at the top of the page (fol. Aiii'') without the intervention of any book- title, extra spacing, unusually large initial letter, or other sign of the opening of a work. Should it prove possible to print later the epistles i-xxvi, the sheets might readily be prefixed to those now issued, and form therewith a consistent whole. It is perhaps idle to guess what had happened to the copy of letters i-xxvi; but that Leander's copy was made from the complete MS that Giocondo had discovered seems indubitable; therefore it is reasonable to suppose that, being able to copy part of the MS, he was able to copy all, and did copy all, and that the loss of the earlier section was due to some accident in the journey of the sheets from Paris to Italy, or to some untoward event in the study of Avantius, or in the printing-office of Johannes de Tridino,' but the find was too important to await the possibility of securing another copy from Paris, and Avantius therefore determined to issue the work in its fragmentary form. The book exhibits a rather unusual number of by Heinrich Keil. From this the letters proceed in due order, as in Kell's edition, through the final one, which is headed " q De Diplomatibus .Ixxiii." (121 in Keil). There would thus appear to be 47 letters, according to the numbering of Avantius, instead of 46, as he avers in his title. The apparent discrepancy is due to the omission altogether of the number .xxxvi. at the place where it is due (58 in Keil), the serial numbering continuing with . xxxvii . in its proper place. The total of numbered let- ters, is, therefore, actually 46, and not 47. But another explanation is of course possi- ble, that Avantius simply erred in his computation, subtracting 27 from 73. The numbering in Avantius may be brought into accord with that of Keil by noting that Avantius assigns a single number to a letter of Pliny and to its answer, taken together; that he assigns no number to 58 (Keil), though he does assign numbers to each of the four enclosures that accompany it ; and that he also assigns no numbers to 86A, 86B, and 87 (Keil). ' It is of course possible that the words of Avantius in his preface, " Petri Leandri industria ex Gallia Plinii iunioris ad Traianum epistolas licet mancas deprauatasque habuimus," mean that the copy was incomplete when it came into the hands of Avan- tivis ; but it seems hardly justifiable to push the meaning so far. Eaely Editions of Pliny's Cobbespondence 453 simple errors, which appear to be due partly to the inaccuracy of the copy (perhaps arising from furtiveness, but certainly also from difficulty in reading; cf. the preface of the edition of Aldus ia 1508), and partly to carelessness in proofreading. The haste of the publi- cation may well be responsible for this deficiency in editorial work. Only eight months later* than the edition of the correspondence with Trajan published by Avantius appeared another by Beroaldus, who, in 1498, had published the eight books of Pliny's miscellaneous letters with such letters of the ninth book (properly the eighth) as were then known, and, in 1501, had issued them in a second edition, joining thereto the Panegyric and the Liber illustrium uirorum. Beroaldus is discreetly silent concerning the source of his text of the correspondence with Trajan. "Hae sunt Epistolae aliquot," he says in his preface, " quae nuperrime in lucem prodierunt .... has pro^ime recognoui, emaculatasque una cum Panegyrico impri- mendas dedi." It will be observed that Beroaldus at any rate does not profess to draw upon any MS source for his text. It would have been more ingenuous of him to state distinctly that his only source was the printed book of Avantius. He prints precisely the same letters as Avantius, and an inspection of the readings wherein his text differs from that of his predecessor shows so clearly that I need not here rehearse the details that his was merely the work of a corrector, which he performed in a very creditable fashion.^ It would have been a curious coincidence indeed, if two independent copies of the Paris MS had suffered loss in precisely the same amount. It is hardly credible that any copy of only these last 46 letters from the Paris MS existed which could serve as the archetype of both the copy made by (or for) Peter Leander, and a copy used by Beroaldus. It would have been more wonderful yet if these two equally truncated copies had been made, one from the Paris MS, and the other from an independent source. And it would have been most wonderful of all if the publication of Avantius, with its evident marks of haste, if not iThe edition by Arantius was published in May, 1502, that by Beroaldus in January of the same year ; but March was then the first month of the year, and the edition of Avantius therefore preceded that of Beroaldus by eight months. 2 The readings of Beroaldus and Oatanaeus are very imperfectly cited in the apparatus criticus of Keil's edition of Pliny's Letters (1870). I hope to present them with accuracy and completeness in the apparatus criticus of my own critical edition, taking them from copies of the works in my possession. 454 Elmer Teuesdell Meeeill of surreptitiousness (note also that Avantius says merely that the MS was ex Gallia), could have had as its final source quite a different codex from the Parisinus that we know to have been discovered about this time. Precisely as, in 1498, Beroaldus had simply appro- priated the work of Pomponius Laetus, so now he appropriated that of Avantius. He furthermore gave an air of finality to his edition by omitting the serial numbers of the letters (by which Avantius had so frankly called attention to the loss of such a considerable amount from the beginning of the book) , and he inserted a brief address before each letter, while Avantius gave none, but only a lemma before each letter or pair (in one case a larger group) of letters. In this respect Avantius doubtless followed Leander's copy, who, in turn, had in this particular accurately transcribed the Paris codex ; for there is the same lack of addresses in the transcript made for GuHlaume Bude of the first 26 letters from the Paris codex. Of this transcript I must now speak. Soon after the publication of the correspondence with Trajan by Avantius, a copy of his edition and one of the 1498 edition by Beroaldus of the nine books of Pliny's miscellaneous letters were taken by Guillaume Bud6, and the parts needed to complete the entire body of correspondence were interpolated in manuscript from the Paris codex that Giocondo had discovered. The whole was bound together in a single volume, which Budaeus apparently thence- forth used as his handexemplar of Pliny's Letters. He inserted in the printed portions of the volume very many text-corrections, taken, as were the MS supplements, from the Paris codex. This source of the corrective readings, and of the MS supplements, he himself makes clear in an appended note. To the notes of readings from Parisinus he also added from time to time a few emendations of his own, and, in later years, a few readings drawn from other sources, especially from the 1506 edition of Catanaeus. His book is now preserved in the Bodleian Library at Oxford.' Meanwhile a copy of the entire codex Parisinus had been made by Giocondo at Paris and brought by him to his friend Aldus, at Venice, together with six other copies of Pliny's Letters, partly • Of. B. T. Merrill "On a Bodleian Copy of Pliny's Letters," in Classical Philology II (1907), 130-56. Eaely Editions of Pliny's Cobbespondence 455 in manuscript, partly in print corrected from manuscripts.' But it is evident that none of these except the copy of Parisinus contained any trace of the correspondence with Trajan. Two years later Aloisio Mocenigo, who had been serving as Venetian ambassador at Paris, returned home, and brought Aldus the codex Parisinus itself. On the basis chiefly of this manuscript, about the antiquity and correctness of which Aldus speaks in rapturous terms, he issued in 1508, at Venice, the first text of Pliny's Letters complete in ten books. The after fate of the codex Parisinus is unknown. In 1518 Aldus issued a second edition of his work, but the only changes con- sisted in the correction of a few typographical errors, and the improve- ment by conjecture of a few readings. No fresh MS aid was invoked, and many passages susceptible of easy improvement were left as first printed. But we must turn back to follow a step farther the fortunes of the truncated text of the correspondence with Trajan. In 1506 lohannes Maria Catanaeus published at Milan a text of the same letters of Pliny that had been issued by Beroaldus in 1498, appending to them as a book labeled Epistolae ad Traianum .C. Plinii Caecilii Secundi the correspondence with Trajan issued by Avantius and Beroaldus in 1502 — forty-six letters, no more, no less. On the source of his text Catanaeus touches only in the single remark that he had at first thought of refraining altogether from publish- ing these letters to and from Trajan, "et quia uno tantum exemplari praeter impressa nee illo admodum uetusto adiuti fuimus, et ne iccirco ansam maledicis obiurgandi daremus." The exemplaria impressa were of course the editions by Avantius aild Beroaldus. But Catanaeus distinctly claims to have had also a MS at hand, though it was "not very old." Keil (ed. 1870, praef. p. xxxvi) did not entirely reject this statement, conceding that Catanaeus might have had a copy, with some added conjectures, of the same sheets that Avantius had used, though insisting that Catanaeus made no changes that could not have been reached by conjecture. But Keil proffered no proof of his judgment, and since in later days Wilde,^ and even 'Cf. the preface to the edition of Pliny's Letters by Aldus (Yenice, 1508). 2 G. I. Wilde De C. Plinii Caecilii Secundi et Imperatoris Traiani Epistulis Mutuis Disputatio (Leyden, 1889) 123. 456 Elmer Truesdell Merrill so distinguished a student of such matters as Remigio Sabbadini,' have asserted their belief in the truthfulness of Catanaeus regarding his MSS, at least as set forth in his second edition, the question may perhaps deserve some slight examination. For the purposes of this examination the correspondence with Trajan may be considered in two divisions: one comprising x. 41-121, the letters that first appeared in the editions of Avantius and Bero- aldus, and were issued by Catanaeus in his edition of 1506, and the other comprising x. 1-40, the letters that first appeared in the edition of Aldus in 1508, and were first included by Catanaeus in his edition of 1518 (the same year that saw also the second Aldine edition.) The method of examination concerning the first division of the text (x. 41-121), in the lack of other evidence, must rest upon two things; first, as Keil saw, upon a comparison of the readings of Catanaeus in the letters in question with those of Avantius, whose publication preceded his; and second, upon a comparison of the readings in the given letters adopted by Catanaeus in his second edition (1518) that differ from those of his first edition (1506). The reason for this second point I will specify when I come to it. In taking up the first point, it must be borne in mind that Cata- naeus in 1506 distinctly speaks of using the exemplaria impressa in the preparation of his text of the correspondence with Trajan, and many of his variations from the text of Avantius are found to agree with those previously adopted by Beroaldus. As Beroaldus neither had nor professed to have any manuscript for his authority, it is very evident that in the examination of the readings of the text of Catanaeus for the purpose of helping to determine whether his profession of having a manuscript at command is true, all his variations from Avantius that agree with those of Beroaldus must be left out of consideration. The total number of variants in Catanaeus from the readings in Avantius, exclusive of probable misprints, differences in mere orthography, and alterations in mere grammatical forms (like -arunt for -auerunt), is about 215. Of these, 73 are already found in Bero- aldus. The remaining 142 are too many to be recorded in the pages ' B. Sabbadini " Storia e Critica di Alcuni Test! Latinl," Museo Italiano di Anti- chitii Classica III (1890), 357. Eaely Editions of Pliny's Ooeeespondenoe 457 of this journal. They must be examined in detail in the promised apparatus criticus to the Letters. Comparatively few of them would find a place on any theory of their origin in the text of the Letters. The large majority are not merely wrong readings, but they are not such as suggest from their nature that they must have had a manu- script behind them. They smack rather of an eagerness for arbitrary emendation. Even though, for example, Gatanaeus did well (as against Avantius and Beroaldus) in perceiving that 86A must be separated from 86, and 86B from 86A, yet he did very ill in striking out the words qiiam ea quae speret at the mutilated opening of 86B,. and inserting what was clearly a mere guess of his own, Fdbium Valentem (cf. Wilde loc. cit. 119 ff.). His zeal for similar emendation in the immediate neighborhood of this success led him still farther astray in the desertion of Beroaldus, and the building upon Avantius in the division of 87 into two letters. Avantius had begun Nymphi- dium in 87. 3 with a large, ornamented, capital letter, such as he uses at the beginning of a new epistle, and had concluded the pre- ceding sentence with a period. Beroaldus rightly corrected the error; but Gatanaeus reverted to Avantius, and even outdid him, dividing the epistle and prefixing an address. The total impression created by the study of the entire mass of variants may be summed up by saying that if this part of the exami- nation were all that could be made, and there were no other reason to suspect the truthfulness of Gatanaeus, it might be believed, as he claimed, that he had at command a manuscript "not very old" — probably a sister of that used by Avantius — and that some of his aberrations were due to a wilful preference of his own guesses to the leading of his guide. At the same time it would seem curious and a bit disturbing that out of some 140 variants from Avantius none were of such a character as clearly to indicate their dependence upon a manuscript. There the first part of the examination must rest. To turn now to the second part of the examination, that of the readings in x. 41-121 in the edition of 1506 that differ from those in the edition of 1518; this appears to be of some value because the first edition of Aldus (1508) had been published in the interval between the two editions of Gatanaeus. 458 Elmer Tbuesdbll Merrill Whatever changes Catanaeus made in the text of x. 41-121 in his second edition as compared with his first must have come from con- jecture or from the copying of Aldus; for, as I shall point out later, Catanaeus does not claim to have any new MS authority for the text of these particular letters. Now, the number of these changes is very large. It is difficult to give the precise sum of them, because not infrequently changes in consecutive, or nearly consecutive, words might be counted as one variation, or as more than one. But according to a conservative enumeration the number of changes of reading in x. 41-121 made by Catanaeus in his second edition as compared with his first is 229; and 196 of these readings agree with those of Aldus, or substantially so. Of the remaining number, which of course must be due to conjecture, none are of any consider- able importance for our discussion. In this count are charitably omitted changes of addresses in Cat.^ from Cat.*, the addition of lemmata to the letters in Cat.^ (taken, of course, from Aldus), all merely orthographical alterations, and finally, misprints clearly such. The degree of the copying from Aldus is very striking. Yet Catanaeus ridicules unsparingly both Aldus' estimate of the age (and consequent value in his eyes) of his MS from Paris, and also Aldus' editorial work. By this action Catanaeus seriously impeaches his own integrity. If he had had a MS on which to found his first edition, even such a "not very old" one as he professed, it is with difficulty conceivable that he would forsake it in such a very large number of instances in behalf of the work of his decried competitor. The only conclusion appears to be that he had no MS support for his first edition of x. 41-121, but later, finding that Aldus had (and a wonderfully good MS too) he was shrewd enough to appropriate his rival's achievement, while hoping to distract attention by abusing the man from whom he committed the theft. In 1506 Catanaeus appropriated the editions of x. 41-121 by Avantius and Beroaldus, and built his own upon them, professing to have a MS in order to enhance his dignity, disguise his obligations, and gain credit for his emendations; but conceding that his MS was "not very old," in order to forestall blame for not curing more difficulties. In 1518, nine years after the appearance of the edition of Aldus, he did not hesitate to discard Avantius in favor of Aldus, Eaely Editions of Flint's Cobeespondenoe 459 even though thereby he discredited his own former pretensions. He perhaps hoped that after the lapse of so many years the deadly parallel would not be appealed to. If Gatanaeus had professed to have for these letters x. 41-121 any further MS authority in his second edition than in his first, it would be necessary only to point out in how many of these instances the readings of Aldus with which Gatanaeus agrees are clearly only conjectural expansions by Aldus, after his confirmed manner, of the briefer and sincerer readings of Parisinus, which Avantius more truly presents. They could not have stood in any MS that Gatanaeus had. They were conjectures of Aldus, and Gatanaeus copied them from him. But under the actual circumstances this detailed exami- nation is certainly not necessary. Let us now turn to the discussion of the other division of the cor- respondence, namely x. 1-40. In his second edition of the letters of Pliny, published in 1518, Gatanaeus prefixed in due order to the letters to and from Trajan that he had printed in his first edition the 26 (x. 1-40, as we now commonly reckon them) that had first seen the light by the care of Aldus in 1508. Of the source of these additional letters he says in his preface,' Emiseram iampridem commentarios in eas Plinii Caecilii epistolas, quae forte tunc impressae circunferebantur. Caeterum cum postea Eomam uenissem, ac cum Dominico Mamiliano uiro in libris antiquae lectionis perquirendis diligentissimo beneuolentiam contraxissem, deseriptas mihi de uetustissimo codice germanico plures ad Traianum, et insuper quasdam eiusdem Plinii ad amicos epistolas legendas obtulit, quae de libro octauo exciderant, etc. A little farther on, Gatanaeus abuses Aldus roundly (though with- out mentioning his name) for his ignorance as an editor of Pliny's letters, and especially for his stupidity in supposing codex Parisinus to be so ancient. Incidentally he refers to the codex Germanicus from which Mamilianus had made the copy as pemetiistum. Here again Gatanaeus does not profess to have more than a recent transcript of a very old MS "from Germany," or perhaps we must 1 Not having constant access to a copy of the 1518 issue of the second edition by Gatanaeus, I have quoted from my copy of the reprint of that edition issued by Ascen- sius at Paris in 1533, which appears to represent accurately the second recension. 460 Elmek Teuesdell Merrill understand "in Germany"; and this transcript embraced only those letters that the Roman edition of 1474, and all following editions up to that of Aldus in 1508, had omitted from the eighth book (viii. 8. 3 — 18. 11), and also the letters to and from Trajan that Avantius, Beroaldus, and Catanaeus himself in his first edition had not printed, but which had later appeared in the 1508 edition of Aldus. Only thus can we understand the words quoted from the preface of Catanaeus. That a transcript of only these letters should have been made to supplement and complete the printed editions is, to be sure, by no means impossible. Budaeus had precisely such a partial copy from the codex Parisinus made for his own use, to supplement the editions of Beroaldus (1498) and Avantius (1502) ,* and although there is no reason to suppose that any copy was ever made from his copy, it is of course conceivable that the copyist who transcribed the letters for Budaeus might also have had opportunity to make another copy which finally reached Catanaeus. But Catanaeus declares his transcript to be from a German codex, and not from a Gallican one. That is, we are asked to believe that out of Germany appeared not, indeed, a MS of the correspondence with Trajan, but a transcript of a part of that MS, and from this time on nothing more was ever heard of the remarkable original! Catanaeus ridicules Aldus for claiming such great age for his Parisinus; but at least Aldus was perfectly explicit about the provenience and journeyings of the MS. "Why was Catanaeus so vague — not to say reticent — about the circumstances surrounding his Germanicus? His assault on Aldus challenged explicitness from himself. This he avoids. In place of openness he snarls and growls. The circumstances arouse suspicion. Moreover, he stands under a cloud regarding the claims made by him about the MS authority for the correspond- ence with Trajan in his first edition. But the ultimate appeal must be made to the text itself. And here we may disregard x. 41-121, for which Catanaeus does not profess to have any new authority, and consider merely the letters X. 1-40, which he prints for the first time in his second edition. In this matter the examination is established on a basis that was not 1 See Classical Philology II (1907), 132, 135. Early Editions of Pliny's Oobkespondence 461 accessible to Keil; for the handexemplar of Budaeus preserved at Oxford contains a transcript of x. 4-40 from the Paris codex. By this we may check the readings of Aldus, knowing that Aldus had no other authority than Parisinus for the letters in question. And by this also we may check the readings of Catanaeus. The result of the comparison of the text of Gatanaeus (Cat.) for X. 1-40 with the texts of Aldus (a) and of the Budaean transcript (I; or Budaeus as corrector, i) is somewhat interesting. There are only some 60 variations of reading between Catanaeus and Aldus (disregarding, as before, misprints, or their corrections, and orthographical differences). These lend themselves to easy arrange- ment in three categories, as follows (references are by page and line of Keil's edition of 1870): A. Under the first heading may be grouped these cases, inc'uding a trifle over one-third of the total number of variants: status la statuarum Cat. at multas et omnes publicas la et priuatas multas et omnes publicas, Cat. et la etiam Cat. VOfUOV IJ,€IJ,VTOV a. VOIXOV /It/K^tTtKOV Cut. VO/XOV /H£I/<^tTOV tfl lOC, i Abascanto la Abascantio Cat. quoque indulsisti la quoque saepe indulsisti Cat. Accium a Attium I Actium Cat. quietissimum la quietissimam Cat. morum la maiorum Cat. mi la mihi Cat. minima la nimie Cat. rauerentissime et oflSciosissime la reuarantissimae at officiosis- simae Cat. scripturum la scriptum Cat. intribuantur la intribuant Cat. intribuatur a^ et . . . . cumulat la ut . . . . cumulet Cat. ita la, om. Cat. oportebit la oportabat Cat. inquisitio la inquisitor Cat. exegit la exig^t Cat. meminerimus la memineris i Cat. f uerint . hetariae * [* om. I] quae breues fiant la fuerint heta- riae quamuis breues fient Cat. 283. 10 tempus la opus Cat. 273. 9 16 26 28 274. 10 15 16 18 24 275. 12 28 277. 13 19 278. 14 25 279. 14 280. 4 6 11 281. 3 282. 3 462 Elmee Tbdesdell Merrill Many of the above instances, taken by themselves apart from the context, look as if they might be mere palaeographical variants of a common enough sort. But studied in their respective contexts they smack not of palaeographical error, but of attempted emendation of the text of la. It might be remarked that in all these cases the emendation is a bad one. This, to be sure, is not the real point at issue; but one would not expect to find in Catanaeus' professed "codex Germanicus uetustissimus " such an array of simple but bad emendations. The expectation would be rather to find more readings that are clearly not due to any attempts at emendation. But in hope of this we must evidently pass to other categories. B. About as many variants as in class A are very plainly to be judged otherwise as to the quality of the reading. Not all of them are clearly better readings than those of Aldus, but none of them are impossible. Some are very striking improvements upon the read- ings of Aldus. In some cases I, which may be taken as certainly representing Parisinus with fidelity, agrees with Catanaeus as against Aldus; in others Aldus himself, in his second edition (1518), con- temporaneous with that of Catanaeus, reached the same end by conjecture. The list follows: 270 [III B] Cat. distinguishes from preceding letter (in note does not profess MS authority!) 271. 19 sollicitudine (soli- a) la soUicitudini a^ Cat. 23 matronam habet la patronam habuit Cat. 273. 4 locationem la locatio cum Cat.'^ i^ 19 quod la quo a^ Cat. 274. 27 uacat la uacant Cat. (probably right but can't be sure) 31 Imp. Maxima a imp. maximae I Cat. 275. 8 itinere la itineri a' Cat. 13 in om. la, add. Cat. 276. 33 regere a regerere I Cat. 277. 3 sit la est Cat. 278. 3 habent est sordidum la (perhaps rightly) habent et sordidum Cat. 279. 5 meam la mea a^ Cat!^ 280. 6 dati lecti sunt la dati lecti si sunt Cat. 10 dies pro quo la dies pro quo I uel i dies quo Cat. 17 similia in his la similiaque his 1 Cat. 26 relinqui a reliqui I Cat. 31 quid a quod I qui a^ Cat. Early Editions of Pliny's Correspondence 463 281. 5 quo qui la quod qui 1 Cat. 23 ut la et a^ Cat? 30 complurimum la complurium Cat. 283. 11 inchoauerint quid itaque compereris per aquaeductus et relique- rint far la inchoauerint aquaeductus et reliquerint . quid ita- que compereris perfer Cat. 284. 7 confenmt a conferent I Cat. 24 illi la illis i Cat. No one is competent to deny that such of the above readings of Catanaeus as are surely in agreement with what Pliny himself wrote might have survived in a MS of the Letters till the fifteenth century. But all the variants in Class B, as well as those in Glass A, are of such a character as to be easily reached by conjecture (good, bad, or indifferent) based merely upon the text of Aldus. Had Catanaeus a MS at command of different source from that of Aldus, one would certainly expect to find in Classes A and B some few readings that would appeal to the critic as not so possible to attain by mere con- jecture. The text of Aldus is not so impeccable as to make this expectation unreasonable. And one would further expect to find in the notes of Catanaeus some definite statements that his manu- script read thus and so in contradistinction to that of Aldus. But I believe he makes no such statement anywhere, though in iii. B he caustically remarks: Qui hanc epistolam cum praecedenti coniunxerunt, ut esset una et eadem cum superiore, ex professo arguuntur non habuisse exemplar tem- poribus Plinii scriptum, ut asserunt, neque sensum uUo modo intellexisse. Yet not even here does he definitely cite his " codex Germanicus uetustissimus," or its transcript. It is a cause of legitimate wonder that Aldus did not, even in his second edition, see that iii. B, what- ever the lack of separation in Parisinus, was in reality the first in the series of Trajan's replies.* But Catanaeus might easily have made this correction without the aid of a MS. As he himself sug- gests in neque sensum ullo modo intellexisse, he needed only to under- stand the sense in order to assign the few lines to their proper source. From this examination of a full two-thirds of the variations of Catanaeus from Aldus is it hard to reconcile the patent facts with ' Probably in the original MS of book x there were no addresses given, but only a lemma for each letter, or for each pair of letters, where the letter was accompanied by an answer ; cf. also x. 57 in A. 464 Elmeb Truesdell Mebrill the editor's profession of having in possession a transcript of x. 1-40 from a German MS. This difficulty becomes an impossibility on consulting the final class of examples. C. In this final class of variants the Budaean transcript pretty surely represents Parisinus correctly, but Aldus has emended Paris- inus — almost always wrongly — sometimes, indeed, preposterously — or else has blundered otherwise; yet Gatanaeus copies Aldus, or builds upon him ! I append the list: 271 [lemma of V] Ciuitatem Komanam latraliptae suo petit I Ciuitatem Romanam Harpocrati latraliptae suo petit a Cat. 271. 7 ndnerte 1 adit te a. SMget Cat. apud te a' (rightly). 272. 5 at si eum I eum scilicet a Cat. esse eum i (rightly). 34 Kal. septembribus I Kal. Septembris a Calend. Septembris Cat. 275. 2 cum I ut a Cat. 275 [lemma of XV] in Bithyniam profligate I in Bithyniam a Cat. 276. 7 agantur I aguntur a Cat. 8 cum maxime I cum Maximo a Cat. 13 et I ut a Cat. 21 sufiicientes I suflicienter a Cat. 277. 10 milites om. I, add. a et Cat. 26 te poscat an homines in se ut latius uelint (uelit I) la res poscat an homines imperare latius uelint Cat. 278. 4 et uetus. itaque tamen aestimans (aestimans corr. ex aestima- mus) nouum fieri quod uideris mihi desiderio eorum indulgere posse I et uetus. Id itaque indulgentia tua restituere desyde- rant, ego tamen aestimans fieri debere, uideris mihi desyderio eorum indulgere posse a Cat. 28 dabitque I dabit ipse a Cat. 279. 29 dixerant sacramentum ita nondum I dixerant sacramento mili- tar 1 nondum a dixerant sacramento militari nondum Cat. dixerant Sacramento ita nondum i (rightly) 281. 11 solent et ad I solent enim eius modi ad a Cat. 282 [lemma of XXXVII] De aquaeductu I De aquaeductu Nicome- densium a Cat. 282. 20 h 8 • XXX • CCCXVIII I • H-S-xxx • ccc ■ xxviiii • a • HS • xxx • ccc • xxix • Cat. 21 imperfectus adhuc emissus destructus etiam est I imperfectus adhuc relictus ac etiam destructus est a Cat. 284. 6 addit I addunt a Cat. [additi Casaubon, rightly] In view of these readings it is impossible to believe that Gatanaeus had either any transcript from a MS from Germany, or even any Early Editions of Pliny's Cobrbspondenoe 465 copy from Parisinus. Either one of these would have left more evident trace of itseM upon the text of Catanaeus; and especially, he would not have abandoned its readings for the readings of his decried rival, Aldus, in so many instances where it must have read differently from Aldus. In 1506, and again in 1518, Catanaeus put forward a vague claim to support by MS authority in order to con- ceal the undignified fact that he was depending for his text on Avan- tius and Aldus respectively, as corrected by his own conjectures. He had no manuscript of any part of the correspondence with Trajan. At the very best the professed transcript, received from Dominicus Mamilianus, of viii. 8. 3—18. 11 and x. 1-40, if it existed at all, could have been only a transcript of these previously unknown letters from the 1508 edition of Aldus, and not from the Galilean codex itself — still less from a German codex'. Sabbadini's faith in the truthfulness of Catanaeus was due in part to the charity that "beheveth all things" (" bisognerebbe sup- porre che fosse un grande imbroglione questo signor Cattaneo, se uogliamo negargli fede che il suo codice venisse dalla Germania e non dalla Francia ") ; yet his belief was doubtless made more easy by his conviction that a tradition of the correspondence of Pliny and Trajan was extant at Ferrara before 1450;^ but this conviction he has later, and wisely, retracted,^ and his charity is left unsupported to bear the moral shock to which charity is too often exposed in this wicked world of ours. Wilde's belief in the truthfulness of Catanaeus concerning his use of MSS was based oji the existence of three readings of Greek words or phrases found in Catanaeus, which Wilde believed must have come from a MS, because Hardy* said they occur among the manuscript corrections in the volume at Oxford which has before been mentioned. The conclusion was that these readings must, therefore have stood in Parisinus, although they do not stand in the 'For the text of -riii. 8. 3-18. 11 Catanaeus of course depended in 1518 on the 1508 edition of Aldus and on the codex Mediceus (M), which he mentions in his preface, and was the first to use. He needed no transcript of these letters from any other source on which ^o found his readings, ^Museo Hal. Antieh. Class, loo. dt. SR. Sabbadini La Scuola e gli Studi di Gttarino Guarini Veronese 111 (1896).