1 Dibdin, Thomas Frognall, An Introduction to the Knowledge of Rare and Valuable Editions of the Greek, and Latin Classics (London, 1827), ii, 141–143.Google Scholar

2 For a recent appraisal of Dibdin as a bibliographer, see Munby, A. N. L., The Formation of the Phillipps Library up to the Year 1840 (Cambridge, 1954), p. 113.Google Scholar

3 Juvenal the Satirist (Oxford, 1954), p. 206; one finds here the entirely correct statement that “Juvenal's satires were first printed by Ulrich Han, probably in Rome between 1467 and 1469”. In the note attached to this quotation, Professor Highet has confused the two Han printings.

4 The Andrew Fletcher copy of Hain 9660; compare Adams, Frederick B. Jr., Fourth Annual Report to the Fellows of the Pierpont Morgan Library (New York, 1953), pp. 24–26.Google Scholar

5 Dibdin (loc. cit.) cites only two copies, that in the collection of Count D'Elci of Florence and the one “in the Magliabecchi library in the same city”. The first is now in the Biblioteca Laurenziana (D'Elci 981), but there is no copy in the Biblioteca Nazionale Centrale in Florence, which includes the Magliabecchi library. Apparently Dibdin was misled by the notice in Ferdinando Fossi's catalogue (Catalogus codicum saeculo XV impressorum qui in publica bibliotheca magliabechiana Florentiae adservantur [Florence, 1793—95], iii, 91). This does not, however, refer to a Magliabecchi book but to a volume belonging to the Duke of Cassano. See note 11 below.

6 The Catalogue of Books Printed in the XVth Century now in the British Museum (London, 1908-49), v, 582, maintains that neither Johannes nor Vindelinus de Spira printed this edition, but that it is the work of an anonymous press.

7 An Index to the Early Printed Books in the British Museum: from the Invention of Printing to the Year MD. with Notes of those in the Bodleian Library (London, 1898-1903).

8 Hain, Ludwig, Repertorium bibliographicum (Stuttgart, 1826—38).Google Scholar The numerical order, of course, implies that Hain considered his no. 9660 as the earlier; he assigned his no. 9661 either to Riessinger at Naples or to Han at Rome. Hain 9661 was printed with Han's second type and Hain 9660 with the same printer's third type. Proctor (p. 226, no. 3344) incorrectly notes that Hain 9661 was printed with Han's type 3.

9 Katalog der Inkunabeln der kgl. Universitäts-Bibliothek zu Uppsala, (Uppsala, 1907). On p. XI, Collijn explains: “Ferner sind überall Verweise auf Proctors Index gegeben worden, sofern die betreffende Inkunabel dort verzeichnet ist, und seine Bestimmungen sind benutzt worden, wenn nicht meine eigenen Untersuchungen zu abweichenden Ergebnissen geführt haben.”

10 “Il Fossi (III, 92) aveva assegnato ad Ulrico Han quest’ edizione … Gli studii recenti hanno mostrato che il Fossi s’ era bene apposto (Pr. 3344)” (II, 189).

11 The Rylands copy of Hain 9661 was the copy formerly belonging to me Duke of Cassano (see note 5 above). Compare the note in The John Rylands Library Manchester: A Brief Historical Description of the Library and its Contents, with Catalogue of the Selection of Early Printed Greek, and Latin Classics Exhibited … MCMVI (Manchester, 1906), pp. 9-10: “The three books in the [Cassano] collection that had special attractions in Lord Spencer's eyes were an unique edition of Horace, printed by Arnoldus de Bruxella at Naples in 1474, an undated Juvenal, printed by Ulrich Han at Rome before 1470, and an Aldine Petrarch of 1501, on vellum, with the manuscript notes of Cardinal Bembo. Could he have obtained these three volumes, there is reason to believe he would have been willing to forgo the rest of the Cassano library, fine as it was, but the fates decreed otherwise.”

12 “Of the fifty principal Greek and Latin authors, we are fortunate in being able to show the first printed edition.” The John Rylands Library Manchester: Catalogue of an Exhibition of the Earliest Printed Editions of the Principal Greek and Latin Classics (Manchester, 1926), p. v. The prefatory note is signed by Henry Guppy.

13 1906 Catalogue, p. 56, and 1926 Catalogue, p. 32. Ironically enough, the edition displayed was not that represented by the Cassano copy, one of the three prized items of this collection, but Lord Spencer's copy of Hain 9660 (cf. Dibdin, , Bibliotheca Spenceriana [London, 1814-15], II, 117–119 Google Scholar).

14 “The Marston Juvenals”, Yale University Library Gazette, xii (1938), 76.

15 Guaraaschelli, Teresa and Valenziani, Enrichetta, Indice generale degli incunaboli delle biblioteche d'Italia (Rome 1943-54), iii, 203-04.Google Scholar

16 This and the copies at Florence, Manchester, and Uppsala seem to be the only survivors of this edition. The copy at Naples bears the press-mark: S. Q. IX B 19.

17 Mead, Herman R., Incunabula in the Huntington Library (San Marino, 1937), p. 169, no. 3946.Google Scholar

18 A copy of Hain 9676 (Indice generale 5568) is also found, and has been consulted, in the Pierpont Morgan Library; cf. Thurston, Ada and Bühler, Curt F., Check List of Fifteenth Century Printing in the Pierpont Morgan Library (New York, 1939), p. 70 Google Scholar, no. 806. Yale now also has a copy of the Fivizzano edition.

18 An Essay on Colophons (Chicago, 1905), p. 89.

19 In PML 44865, the colophon reads:

21 Thus BMC, iv, 18: “In use in 1470 and 1471”.

22 The famous phrase “Know thyself” (Juvenal, Satire XI, 1. 27) is printed in Hain 9660 as: vωθη αvτov. The Naples copy of Hain 9661 has the following text supplied in manuscript: γvωσ i σε αvτωv. The longer Greek passage in Satire IX, 1. 37, reads in Hain 9660: ηθovζ Γαρ Γλvxεoσ αvδρα xιvαιδov; in the blank space of the Naples copy (of Hain 9661), there is written: ηθovζ γλvxεισ αvδρα xηρα ιδov. This should read: αvτòζ γἁρ ἐϕἐλxεται ᾀvδρα xívαιδoς.

23 According to BMC, iv, 18. Robert Proctor (The Printing of Greek, in the Fifteenth Century [Bibliographical Society, 1900], p. 29) was of the opinion that the Greek type “seems to occur for the first time in the Tortellius of 1471”.

24 No example of a book printed in type three during the life of the partnership is known to BMC, nor was it used by Han after the partnership had lapsed.

25 Burger, Konrad, The Printers and Publishers of the XV. Century with Lists of their Works (London, 1902).Google Scholar

26 The correct order is found in the 1470 Venetian and Fivizzano editions, as well as in that by the printer of Duns. In the Naples copy, “neroceronte” has been corrected in manuscript to read “rinocerote” (rhinocerote—Duns), while Hain 9660 prints “kinoceronte”. The first Venetian edition has “rinocerunthe”, and the two Vindelinus de Spira editions, as well as the Fivizzano printing, give “rinoceronte”.

27 Professor Edward Robertson, librarian of the John Rylands Library, kindly informs me that their copy has been washed and that it is impossible to tell whether or not such marks were once there. For the use made of these letters by early printers, compare my “Stop-press and Manuscript Corrections in the Aldine Edition of Benedetti's Diaria de Bello Carolino”, PBSA, XLIII (1949), 365-373, and my “Corrections in Caxton's Cordiale”, PBSA, XLvni (1954), 194-196.

28 Fivizzano prints “cresi” as Han, while the “Duns” and 1470 Venetian editions have “croesi” and Vindelinus offers “cresci”. In die Naples copy “sufficiens” (also found in the second Han printing) has been corrected by hand to read “sufficient”, thus agreeing with all the other editions. “Narcisci” is also corrected to “narcissi” in the Naples volume.

29 The line appears thus in all the other early editions. The Naples copy of Hain 9661 supplies the following text in manuscript: “Induxit cesar cui claudias omnia: cuius”.

30 In line 226, Hain 9660 also reads “stabunt”. This has been corrected in the margin to “stabant”, the reading of the other texts.

31 The other Juvenals here discussed also have the correct order.

32 See, for example, the “rhinoceros” of note 26 above. Then, too, the edition of Hain 9661 lacks the printed Greek words found in Hain 9660. One would hardly expect a printer to revert to a more primitive practice once he possessed the technical equipment for a more perfect product.

33 Without exception, the other editions have “imperiis”.

34 In Juvenal VII, 233, the Morgan copy reads: “Dum petit aut termas phebi balnea dicat”. Hain 9661 has the second “aut” before “phebi” as required. Again, in line XI, 103, the second edition prints: “Vt phaleris gauderet equs. celataque cassis”. Here the “equs” is doubdess a misprint for “equus” (as die odier editions). It seems most unlikely that the “equs” is a graphic variant for the “ecus” preferred by Housman, (Saturae, Cambridge, 1931, p. 105).Google Scholar

35 For Ulrich Han's career, compare also Haebler, Konrad, Die deutschen Buchdrucker des XV. Jahrhunderts im Auslande (München, 1924), p. 18 ff.Google Scholar Haebler stresses the fact (p. 20) that “klassische Texte sind seit dieser Zeit [partnership with Chardella] aus der Werkstatt des Ulrich Han fast gar nicht mehr hervorgegangen”. This is still another reason for believing that the Juvenal (Hain 9660) appeared before the partnership began. In that case, too, one must assume that the first appearance of the Greek type was not, as Proctor had stated, in the Tortellius of 1471 (a product of the partnership) since it is used in the second Han Juvenal.

36 For convenience in citing examples, sigla have been assigned to the several editions. The printer of V is discussed in note 6.

37 Concerning the curious repetitions of lines in V¹ and V², compare the BMC's annotations.

38 Since the publication of Miss Stillwell, Margaret B.'s Incunabula in American Libraries (New York, 1940)Google Scholar, the Huth-Ascherson copy has been acquired by the Yale University Library. It is a pleasant duty to acknowledge the kind help cheerfully rendered by the staff of that library, particularly the interest taken in this study by Mr. Thomas E. Marston.

39 Compare Haebler, , op. cit., p. 27 Google Scholar, and BMC, v, ix.

40 For furdier particulars, see BMC, v, 212 and VII, 1-li.

41 Edition V has its own share of errors, which do not reappear in the other editions. Thus in the lines comparable to VII, 142-145 of a modern text, V omits line 144 (sardonyche, atque ideo pluris quam Gallus agebat); this line is present, however, in the later printings. The Yale copy had the line supplied in manuscript, but this volume has been thoroughly washed and cleaned so that the readings are now partly obliterated.

42 In place of the proper name Iunco (at XV, 27), all Han and de Spira editions have “vinco”; V prints “iunio”, a variant admitted by Housman; F has “iunco” and D offers “iuno”.

43 The de Spira editions do not invariably follow the Han readings. Thus, at VII, 59, the word “enim” is omitted (from: “fontibus Aonidum. neque enim cantare sub antro”) by both Han editions and by V and D; the de Spira and Fivizzano editions have the “enim”.

44 In order to save space, lines have been quoted whether they provide complete sense or not. If not, the reader is referred to the Loeb edition (New York, 1930), which has provided the modern quotations.

45 Occasionally, all the Venetian editions have the same reading. Thus in XIII, 62 (“prodigiosa fides et Tuscis digna libellis”), the last two words are “digna tabellis” in V V¹ V² D, while the Han and Fivizzano editions provide the standard text.

46 V and D have the same error (“Has pariter“) in XIII, 208 (“has patitur poenas peccandi sola voluntas”); the other editions give the correct word.

47 In VII, 138 (“sed finem inpensae non servat prodiga Roma”), the Han and Fivizzano editions have “obseruat non” for “non servat”. In the Naples copy this has been corrected in ink to “non seruat” and the text thus corresponds to that found in all the Venetian luvenals.

48 The copies of F at the Huntington and Yale libraries have the misprint “longfa eras”. In VIII, 79, only F and D (and the Han editions) have “esto” in the line “esto bonus miles, tutor bonus, arbiter idem”; the reading is “est” in V, V¹ and V².

49 The problems that have here been investigated, although based upon a classical text, belong wholly to the Renaissance. It is through the study of printings such as these that we may come to a realization of the numerous difficulties and obstacles which beset the Renaissance printer, as well as the means that he found to solve and eliminate them. By investigations of this sort, we may arrive at a full knowledge of the nature of book-production and the means of distribution of printed books in Italy in the fifteenth century.

50 Although we now have reason to believe that the use of the Greek type by Ulrich Han anticipated the date (August 1471) set by Proctor, it seems doubtful that the fount came into Han's possession as early as 1470. It is therefore not likely (though not, of course, impossible) that the second Han printing preceded the first Venetian one. In any event, this second edition by Han is the first to include the text of Persius, and this is the true editio princeps of that work ( Morgan, Morris H., A Bibliography of Persius [Cambridge (Mass.), 1909], no. 1Google Scholar). Professor Morgan, however, added his small contribution to the confusion surrounding the earliest Juvenals which was just then growing. In describing the first Persius, he stated (p. 1): “Printed in Rome by Udalricus Gallus (Han) in 1469 or 1470, in the same type as his Juvenal, and the two are sometimes found bound together, whence it has been wrongly thought that they were so published.” Morgan referred, of course, to the second Han Juvenal, not the first (no copy of which contains a Persius). The date for the first edition of Persius, therefore, is not “Rome, 1469 or 1470” (as Morgan asserted), but [Rome, summer 1471]. Incidentally, Morgan's second edition of Persius is that issued with the second Juvenal by Vindelinus de Spira (Hain 9672); this should now be dated in the summer of 1473. Morgan no. 3 appeared with the first de Spira Juvenal (Hain 9673); this is now the second Persius, printed “after 1 August 1471”.

51 The Greek words in VI, 195 (ζωή xαì ψvχή) are printed in V² while a blank space is left in V¹. This is still another bit of evidence pointing to the conclusion that V¹ preceded V². The Greek words are printed in the second Han edition and are supplied in manuscript in the Naples copy of the editio princeps. For the significance of these details, compare note 32 above.