.r'^' f^\. '^■•%"A»:-,' ^■^l^i' .k^*:.«v m^w- '^. ^jm * •( 9f^ •IP, fytmll llmv«itg pilr«g THE GIFT OF yf- U)' Hjvt/u^ A , \'^'\o%z -^ ?/;/ II o( Cornell University Library Z251.G7 P96 + Printing of Greek n the fifteenth centu 3 1924 029 500 414 olin Overs Early Oxford Bindings. By Strickland Gibson. (Printed for the Bibliographical Society at the Oxford University Press.) — Ta most students of English bookbindings, save to those frequenters of the National Art Library v?ho are familiar with its gresit collec- tion of rubbings made by Mr. Weale, the illustrations of early Oxford bindings vehieb Mr. Gibson has here brought together, in their variety and beauty, will come as a sur- prise. No such series of illustrations has pre- viously been published, and the photographic department of the Clarendon Press deserves great credit for the admirable quality alike of the collotypes and the photolithographs. j Fine, however, as the illustrations are, they I owe their interest very largely to Oxford con- servatism. Owing to the lamentable dis- appearance of Duke Humphrey's library and to the havoc wrought by rebinding, the earliest example which Mr. Gibson is able to show cannot be dated earlier than 1460. This i» the more disappointing as, by a strange chance, the deed executed about H80, which offers the earliest evidence of a "studium generale" at Oxford, contains a mention of a certain " Laurencius ligator" as the tenant of property in Cat Street. In his ' Chronological List of Oxford Binders,' Mr. Gibson is able to quote references to as many as ten different binders of the thirteenth century and seven of the fourteenth, most of them dwelling in the same street. Of the work of these men, and of their successors of the first half of the fourteenth century, no trace now remains at Oxford, though in the British Museum there is a manuscript of the Pan- dects of Justinian, once the property of Read- ing Abbey, which bears the tantalizing in- scription, in a fourteenth - century hand, " istum librum Oxonie fecit Bicardus de Eedyng ligari." But at Oxford, if anywhere, old traditions linger, and two of the bind- ings here reproduced (plates 13-16), both in their stamps and the arrangement of them, have all the appearance of the twelfth-century work which gave English binders for a time the pre-eminence in Europe. The conserva- tive spirit which preserved this tradition for three centuries was equally potent in resist- ing innovations. The large panel stamps which were so generally used by the London stationers at the beginningof the sixteenth cen- tury apparently gained no footing at Oxford, and gold tooling is not found on Oxford books until Stuart days, when, too, we hear of a binder named Nicholas Smith leaving half his "boxe of gilding toles" to his brother John. No doubt the gilding tools were at first regarded as dangerous innovations, but the Oxford binders soon learnt to make good use of them, and we hope that Mr. Gibson will speedily follow this monograph on the early Oxford bindings with another on the gayer specimens of later date. Apart from the excellence of the illustrations which he has here brought together, his own personal work in the present volume is of considerable interest and value. The stamps and rolls which make np every binding are all carefully described and numbered, and repro- duced separately.by means of photolithographs N°3960, Sept. 19, 1903 from drawings, as well as in combination. Be- sides the chronological list of Oxford binders already mentioned, Mr. Gibson has also made extracts from the Bodleian Day Books and Account Books which supply the prices of bindings in the seventeenth century, and a further appendix will enable any one to see at the Bodleian specimens of the work of seven- teen different binders. His book is thus a good specimen of what a monograph should be, a small subject taken up with enthusiasm, and »' treated with a fulness and accuracy which ' leaves it in no danger of ever being super- seded. OUR LIBRARY TABLE. VMk. John Murkat publishes what to (nany WJ^l be an interesting, and what all wili find an^dmirably written volume in The Life of MiMat Pasha, by his son, All Midh^t Bey. No modern Turk has ever made so |;reat a nameXfor himself in Western Europe as, between the murder or suicide of one Sultan, the deposition of another, the accession of the present Sultan in 1876, and his own murder ^1883, Midhat won. The impression which h& produced here when he came to London i^ the height of his fspae was not, generally,'lfavourable, but none doubted his ability, anc^ the book before us.ls an essential part of the, history of the Eastern Question in our time. V :'. Those who «xpect that Lope and Lovers of tlie Past will ^rrespond to its alluring title will be disappointed. The portrait prefixed to the book, which 'is that of/Charlotte Corday, will prepare reacfiprs for thte fact that the tales are a varied coUe^ion of .odds and ends from the French archives ; ihey mostly concern either the reign of ^oui/ XVI. or the Terror. M. Paul Gaulot wsoliS them for a French review ; they are fai)^ translated by Mr. F. O. Laroche, and are pub^hed by Messrs. Chatto ■ & Windus. ' ^ Canon Welldon's s^mons addressed to ! Harrow schoolboys,' entittted Yoiith and Duty (The Eeligious Tract ^ciety), are telling ' addresses, the majority o^ them being con- 1 structed on the same simple plan. The first : necessity of a scbool sermon We take to be a i clear scheme, capable of bekig remembered. ' Canon Welldon always leaves with his young ) hearer the impression of a\proportionate ' whole. He has a power of cl^r exposition, i and knowledge of the sort of illulktration which ) boys understand. His exampleat are mainly ] drawn from the kind of interest ttiat in nine I cases out' of ten appeals to boys-\the life of action. Add to this that his styl« is simple and direct, never redundant, but amays self- restraJjned and compressed, vigorous, and abovfe all earnest, and it is obvious thTat these twenty addresses provide an excellenB| model for, school sermons. They deal witlt very various subjects, but are most of them\sym- p^thetically adapted to the duties, tempta- tions, and responsibilities of public-school life. TJIE PRINTING OF GREEK IN THE FIFTEENTH CENTURY By ROBERT PROCTOR Illustrated Monographs issued by the Biblio- graphical Society. No. VIII. The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924029500414 ILLUSTRATED MONOGRAPHS No. VIII THE PRINTING OF GREEK IN THE FIFTEENTH CENTURY By ROBERT PROCTOR PRINTED FOR THE BIBLIOGRAPHICAL SOCIETY AT THE OXFORD UNIVERSITY PRESS December, 1900 K THOMAE GVILELMO DVNN PRAECEPTORI INCLITO GRAECARVM LITERARVM EXIMIO FAVTORI PIETATIS DEBITAE PRISTINAE DISCIPLINAE HAVD IMMEMOR ALVMNVS PREFACE I CANNOT let this book go forth without an apology to my readers, in that I have ventured to attempt a work dealing to a very large extent with technical details, without that knowledge of printing which can only be acquired by practical experience. In this respect I owe much to the kindness of Mr Horace Hart, the Controller of the Oxford University Press, who made many obscure points clear to me by ocular demonstration. The greatest care has been taken to make the illustrations facsimiles as exact as possible ; but it will often be found that minute indications to which I have drawn attention in the text are not visible in the reproductions. I must ask those who wish to study the matter closely to turn to the originals in these instances, and to remember that in the first printed books Greek type generally printed with an exceptional lack of sharpness, and that since even the best process blocks magnify and harden any defect due to imperfection of inking or presswork, while they almost always thicken the lines to some extent, the appearance of the type is sometimes materially altered. Mr E. Gordon Duff gave me valuable help with the unique books in the John Rylands Library, especially in connexion with the photograph from the Batrachomuomachia shown in plate VH. To him, and to the other friends by whose knowledge and advice I have profited, my best thanks are due. R. P. CONTENTS PAGE I. Sketch of Hellenism in Italy ... . . . . r II. General Remarks on Early Greek Printing .10 III. Books in which Greek Types are found up to 1476 . 24 IV. Books of the Earlier Greek Class, up to Aldus . . 48 V. Books of the Graeco-Latin Class -83 VI. Books of the Later Greek Class, from Aldus 93 VII. Books in which Greek Types are found, 1476-1500; Greek Printing outside Italy . 126 List of Illustrations ....... ... 149 Plates and Analyses i55 Appendix . . . ■ 207 Index . 209 THE PRINTING OF GREEK IN THE FIFTEENTH CENTURY I. In the present essay, after a short sketch of the history of Hellenic Division of culture in Italy up to the time when the printing press began its work, _, I propose to give a general survey of the early Greek press, touching on the principal groups of types, their relation to the writing hands on which they are based, the special difficulties which the printers had to overcome, and their methods and technique, where they differ from the usual practice of the time. I shall then attempt to enumerate and classify the Greek types used by the printers of Latin books up to the year 1476, when the first book printed wholly in Greek made its appearance. This will be followed by a chapter devoted to the Greek printed books from 1476 to the end of the century, preceded by a short list of these books in the order in which they are described. Lastly will come notes on a few of the founts used in later editions of Latin books, especially those produced in the smaller towns, and some mention, in the shortest form, of the first Greek printing in other countries — Germany, the Netherlands, France, Spain, England. The illustrations, like the books, fall into three divisions, though The plates. the arrangement is somewhat different. First come representations of the types used by the Italian printers in Latin books up to 1476 ; these I have endeavoured to make as complete as possible, and I hope that few early founts of importance have escaped me ; but those who know the difficulty of searching a whole literature for isolated words will excuse omissions. The second and principal series, which is, with one exception, placed together at the end of the book, contains facsimiles of every known type used in a Greek book, that I B is, a book in which the text is Greek, whether it be accompanied by a Latin translation or not, up to the year 1500. The third series is more composite. It consists of (a) selected typical specimens of the Greek type found in Latin books by Italian printers from 1477 to 1500; {d) the first types of the same class in other countries; (c) a few examples of notable sixteenth-century founts which are described in the text. Analyses. To the plates of the second series I have added what I believe to be a new feature, in the shape of an analysis, appended to each, of all the different letters or sorts found in the text therein reproduced. These analyses are in certain cases only part of a larger plan, where I have attempted, always very imperfectly, no doubt, to draw up a list of all the sorts included in the type in question, with the object of indicating the nature of the fount, or the relative size and complexity of the case with which the compositor had to deal I have done this chiefly where a fount is both important in itself or representative of a class, and at the same time the books printed in it are of sufficiently small volume to be analysed without the excessive toil involved in the minute examination of a large mass of letters. Thus, in order to give some idea of a type of the later or Aldine class, I have chosen the type of Bissolus and Mangius in its first state, rather than one of the Aldine founts or that of Kallierges, because it is found only in two small books. In other instances, when a type is used first in a small book, and afterwards in a larger one without alteration, as the lower-case letters of Lorenzo di Alopa with which the ApoUonios of Rhodes and the Lucian of 1496 are printed, I have enumerated only the sorts found in the smaller of the two books. Again, if one fount is based on another, even if it be of comparatively small extent, such as the second Milano type on the first, I have not given more than the letters which occur on the page photographed; nor have I given a list of the variations in a recast type, such as that of the Homer of 1488 or the Vicenza type of 1490, when the original founts (Milano 1476, or Vicenza 1489) have been set out in full. These analyses must be accepted with very large allowances on lihe score of accuracy or completeness, but, so far as concerns the plates them- selves, the reader will have the remedy in his own hands ; in any case, the plan seemed to me likely to be of use both to students of the books or of Greek printing generally, and to those engaged in the good cause of trying to raise modern Greek founts from the mire of dull ugliness in which they are for the most part sunk. Thus much being premised, I will proceed to my subject-matter. In an essay which deals with the printing of Greek almost entirely Greek from its technical side, there is little need to describe in any detail J"a|y^^ '" the growth of that order of things which made it possible to produce books printed in that language with reasonable hope of profit. The facts are well known and easily accessible ; they form part of the history of the Renascence movement, and as such may be studied in the works dealing with that period. But a short summary of some kind is desirable, not only to enable the reader to obtain a clearer view of the matter in hand in its bearings on literature generally, but also because many of the men chiefly concerned in the Hellenic revival were the authors or editors of the books chosen by the printers, or were the teachers of the younger generation of scholars who set the printers to work, or revised and corrected the texts as they passed through the press. All the authorities are agreed that the first Italian of eminence Leontius to study Greek was Boccaccio, at whose invitation his teacher, the^'^*"^' Calabrian Leontius Pilatus, came to Florence, and delivered lectures on the Greek language during the years 1361 to 1364. Though his stay was a short one, Leontius opened the eyes of the cultured Florentines, who began to collect Greek manuscripts; these were im- ported in large numbers, and in this way, there can be little doubt, many treasures have been preserved to us which would otherwise have been lost altogether. This measure was due in great part to the wealth and enterprise of Palla Strozzi, who also in 1396 induced the Signoria to write a letter to Manouel Chrusoloras, offering him Manouei the Greek chair for ten years at an annual salary of one hundred C'^™5°'°''*^- florins. Chrusoloras, who is connected with the early press by his Erotemata, the book most frequently printed of all Greek books in the fifteenth century, was at that time a teacher of Greek at Con- stantinople, -where Guarinus of Verona, afterwards his successor at Florence, was one of his pupils. A few years before this time Chrusoloras had visited Italy as an envoy from the Emperor, and though his mission was unsuccessful, his fame as a teacher induced many to resort to Venice to hear him. One of these, Jacobus Angelus, returned with Chrusoloras to Constantinople, and it was he who was 3 B 2 mainly instrumental in persuading his master to accept the invitation of the Florentine Signoria, After three years, however, Chrusoldras left Florence, and joined the Emperor at Milano. He continued in his service for some ten years, during which he visited several countries, including England, it is believed in 1405 or 1406. In 1414 he accompanied the Pope (John XXIII) to the council of Konstanz, and died there in the following year, being buried in the Dominican monastery, now the Insel-H6tel. His pupils. Great as was the influence of Chrusoloras himself on Hellenic learning, it was greater still from the celebrity of his pupils. It was the same with all the Hellenic professors of the time; themselves at best tolerable grammarians or copyists, with little or none of the higher feeling towards literature, they succeeded in creating through the eager enthusiasm of their hearers a distinct school of humanists of wider culture and larger outlook than themselves. Even if their learning and polish was but skin-deep, or chiefly manifested in a strong tendency to virulent and scurrilous abuse of one another, they were men who succeeded in saving and in a measure making Aurispa. known what was left of the ancient Hellenic literature. Besides Guarinus, Chrusoloras numbered among his pupils Giovanni Aurispa, the translator of Hierokles, Francesco Filelfo, Niccolo Niccoli, cele- brated for his library (in the purchase of which he ruined himself), Lionardo Bruni of Arezzo, Omnibonus Leonicenus, the elder Vergerius, Gregorius Tifernas, Giannozzo Manetti, and Ambrogio Traversari, who played an important part in the council of Florence. Aurispa and Filelfo, not content with what they could obtain in Italy, went to Constantinople to pursue their studies. Aurispa returned in 1423, bringing 238 manuscripts with him. Eugenius IV, who patronized Hellenic learning, appointed Aurispa Apostolic Secretary ; and at the council held at Florence in 1438 to seek a reunion between the Eastern and Western Churches, he acted as interpreter between the Greeks and the Latins. The council was attended by a very large number of Greek statesmen and scholars, and gave the greatest impetus to Hellenic studies in Italy that they had yet received, popularising them for the first time among literati of the second rank. It was the presence of these Greeks, and especially of the venerable Georgios Gemistos, called Plethon, that led to the foundation of the Platonic Academy, and to the special education of Marsilius Ficinus 4 in connexion with it, Filelfo went out to Constantinople in i4i9Fiieifo. at the age of 21, as secretary to the Venetian Consulate, worked at Greek under loannes Chrusoloras there, and married his daughter. He afterwards became professor of Greek at Florence, where he became an opponent of the policy of Cosimo de' Medici. Cosimo, who was after Palla Strozzi the great patron of Hellenism at Florence, placed politics before letters, and in 1433 tried to assassinate Filelfo, besides banishing Palla Strozzi in the next year. Strozzi retired to Padova, and there continued his support of Greek scholars, notably loannes Arguropoulos and Andronikos Kallistos. Of Arguropoulos Arguro- nothing is known between 1441, when he was with Strozzi, and''°"°^" 1456, when Cosimo appointed him Greek professor at Florence. He held this post for fifteen years, then went to Rome (where his lectures were attended by Reuchlin), and died there at the age of seventy. Kallistos was held to be second in learning to Theodoros Gaza only. Kallistos. The life was an unhappy one. After the death of Palla Strozzi in 1462, he taught Greek at the University of Bologna; thence he went to Rome in 1469 ; driven from Rome by poverty, he moved on to Florence, where Poliziano, who was his most eminent pupil, endeavoured to obtain a fixed salary for him from Lorenzo de' Medici. The application was apparently unsuccessful; in 1475 Kallistos was compelled to sell his manuscripts at Milano to obtain sufficient money to journey to Paris. Fate was still against him, and the next year he died, poor and epH^oq 9tAcov, in London. Theodoros Gaza, just mentioned, whose grammar was one of the Gaza, first books printed by Aldus Manutius, was a native of Thessalonike. He> came to Italy between 1430 and 1440, and after studying Latin at Mantova under Vittorino da Feltre, taught Greek at Pavia, and afterwards at Ferrara. In 1450 or 145 1 he entered the Pope's service, and became a close friend of Bessarion until the death of Nicholas V in 1455. From 1455 to 1458 he lived at Naples, but lived in retirement from the death of king Alfonso till 1464, when Paul II summoned him back to Rome. After Bessarion's death in 1472 he left Rome finally, and died in 1475. In the days of Chrusoloras, and for some twenty or thirty years Second later, the chief object of those who had mastered the Greek language P^"°"^- was to secure from destruction the treasures of Hellenic literature. But by the middle of the fifteenth century the movement had passed 5 into other channels, and scholars were devoting their energies to the dissemination of that which the earlier generation had rescued from the Turk. This was done partly by making copies of the manuscripts, a task to which most of the Greek teachers of this period, Kallistos, Gaza, Demetrios Laskaris, Chalkondulas, actively devoted themselves; partly by means of translations into Latin, which multiplied exceedingly Nicholas V. at this period, chiefly owing to the enthusiastic patronage of Nicholas V, who collected at Rome a large number of scholars of both nations, including among the Greeks Gaza, and Georgios of Trebizond, who came to Italy in 1420, and had taught Greek at Florence, and among the Italians Giannozzo Manetti, who had been made Secretary to the Pope, Guarinus of Verona, Lorenzo Valla, Poggio, Perottus, Tortellius, Petrus Candidus and Gregorius Tifernas. All were busily engaged in translating; and to their efforts, as well as to those of Lionardo Bruni.his pupils Rinuccini and Acciaiuoli, to Carlo Marsuppini, Filelfo and others, we owe the surprising list of Latin versions from the Greek which the Italian printing presses issued during their first years. Didot, on page xliii of his Aide Manuce, enumerates forty-three works by twenty-one different authors as printed up to 1492, and his list is very incomplete. Third period. With the death of Nicholas V the second period of the Hellenic revival, the age of the pupils of Chrusoloras, may be said to end. The third period, which concerns us most directly, was not a time of literary productiveness. The rise of the press turned the attention of scholars to the emendation of the Latin and Greek classics, and the few names of note which emerge are those of able and industrious editors rather than of writers. Among the Hellenic editors four are ChSkon-^ most prominent — Demetrios Chalkondulas, Konstantinos Laskaris, dulas. loannes or Janus Laskaris, and Markos Mousouros, three of whom were closely connected with the early presses. Demetrios Chalkon- dulas or Chalkokondulas, of a noble Athenian family, born in 1424, came to Italy in 1447, and after a visit to Rome taught Greek at Perugia, among his pupils there being Campanus, afterwards bishop of Teramo, who has left us a panegyric of his master. Thirteen years later we find him Greek professor at Padova, at an annual salary of 400 florins, and in 1471 he succeeded Arguropoulos at Florence. Here he remained for twenty years, the poet Tarchaniota MaruUus being among his friends, and Poliziano the most brilliant of his 6 disciples. While at Florence he edited the great Homer of 1488, and when he moved to Milano in 1492 he seems to have induced Heinrich Scinzenzeler to establish a Greek press, from which issued the next year the orations of Isokrates under his supervision, and his own Erotemata, together with two other grammatical treatises. Similarly in 1498-99, he patronised Bissolus and Mangius when they fled from Venice, and edited for them the great Souidas, the largest Greek book printed in the fifteenth century. He remained at Milano till his death in 151 1, and a monument was erected there to his memory by his pupil Trissino. Both as author and editor, Chalkondulas was closely connected with the early press, and in the second capacity he stands easily ahead of his contemporaries. It was quite otherwise with the man Konstan- who was his only serious rival as a teacher, Konstantinos Laskaris, Laskaris. whose grammar received the honour of being the first entirely Greek book printed. Laskaris was a Byzantine, and was nineteen when the Turks took Constantinople and made him prisoner. After being ransomed he lived for some time at Rhodes, but in 1460 he was at Milano, teaching the language and writing Greek manuscripts. Five years after this Ferdinand I summoned him to Naples, but he was not successful there, and sailed for home. When by an accident the ship touched at Messina, Laskaris found an opening for himself, remained there, and was appointed to one of the professorships which had been established in 1462 for the instruction of the Basilian monks in Sicily. He spent the" rest of his life at Messina, and died of the plague in 1501, leaving a large collection of manuscripts, seventy-six of which are now in the National Library at Madrid. Laskaris seems, except as author, to have had no con- nexion with the press ; as a diligent copyist he may have, scorned the printed book, though he possessed a copy of the Milano Souidas, which he left in his will to a Sicilian monastery. Probably however the remoteness of Messina from the places where Greek printing was carried on, which were all in North Italy, is sufficient to account for the fact. The interesting epilogue of Laskaris to his TTepl ovojwaToq kqi pHiuaroc is printed as an appendix to this book. His namesake loannes loannes Laskaris, who called himself Janus when writing in Latin, was not only the moving spirit in the second Florentine Greek press, that of Lorenzo di Alopa, but himself designed the majuscule fount which 7 distinguishes the books issued from that press from any others. Born in 1445, he began his career in Italy as a prot6g6 of Bessarion, who sent him to study under Chalkondulas at Padova. Left without resources, like so many of his fellow-countrymen, by the death of his patron in 1472, he followed Chalkondulas to Florence; gained there a great reputation by his lectures, and the favour of Lorenzo the Magnificent, who appointed him his librarian, and sent him on two journeys in the East to buy manuscripts. A list of the places he visited and the persons from whom he made purchases is still extant, and has been printed in the first volume of the Centralblatt fur Bibliothekswesen. While he was absent on his second voyage Lorenzo died, and on his return to Florence Laskaris undertook the editing of the Anthology and other Greek classics for Lorenzo di Alopa. But this task was soon interrupted by the arrival of the French; he attached himself to Charles VIII, and returned with him to France. He died in 1535, at the age of ninety. Markos Markos Mousouros was the most prominent man connected with Mousouros. ^jjg press during the later years of our period. A native of Rhe- thumnos in Krete, and therefore a fellow-townsman of the printer Kallierges, he seems to have been born about 1470. He studied under loannes Laskaris at Florence. We also know from his own statement that he copied Greek manuscripts in his youth ; the copy of Galen's works written by him was bought by Blastos and Kallierges from Nicolaus Leonicenus for the text of their edition. But nothing more is known of him till he appears as chief editor for the Aldine press in 1497. The Latin version of Mousaios, which was printed about that time, and interleaved with the unsold copies of the Greek text printed some two years earlier, is attributed to him ; he edited the Dictionary of 1497, the Aristophanes of 1498, and the Letters of 1499 for Aldus. He was very active at this period, for at the same time he was editing the books printed by Kallierges and Blastos in 1499 and 1500, and undertook journeys on their behalf to purchase manuscripts ; and in the same year, 1499, he was appointed Greek tutor in the household of the Prince of Carpi. From 1503 to 1516 he filled the office of censor of Greek printed books at Venice ; he was appointed assistant professor at Padova, also in 1503, and succeeded to the chair itself in 1505, without however ceasing his labours as editor. On the outbreak of war in 1509 he withdrew to 8 Venice and occupied himself in preparing the works of Plato for the press. When peace was restored the Venetian Senate offered him the professorship there, and he held it till after the death of Aldus in 1515. He then obtained leave for a temporary absence, and accepted an invitation to Rome, to co-operate with his old master loannes Laskaris : but he never returned to Venice. Having taken orders, he was created by Leo X bishop of Hierapetra and soon afterwards archbishop of Monembasia, but died at the end of 1517 while still a man of middle age. His mastery of Latin was praised Ijy Erasmus, who said that Gaza and loannes Laskaris were the only other Greeks who succeeded in learning the language. He had a remarkable turn for the writing of Greek elegiac verses, of which the chief monument is his long poem in praise of Plato. But he was very careless in his use of priceless manuscripts : the unique manuscript of Hesuchios shows still how they were ruthlessly scrawled over by him and then sent into the compositor's room to serve as copy. The only Italian scholar at all prominently connected with the Joannes Greek press was Joannes Crastonus, a Carmelite monk of Piacenza, C""***"""^- who enjoyed a great reputation for learning in his day, though there is little information concerning him. Tritheim, writing in 1492, mentions his letters as being very numerous and elegant, and believed that he was then still alive ; but, except occasional mentions by his contemporaries, no more personal knowledge seems to be preserved. He was evidently a close friend of Bonus Accursius, and all his published work was brought out by him ; he was the author of the Greek and Latin dictionary, which though overlaid by the accretions of successive editors, held the field till the time of Henri Estienne, and he compiled a shorter vocabulary in Latin and Greek ; both of these were printed three times before 1 500. He also trans- lated the grammar of Laskaris for the edition of 1480, and revised the Latin version of the Psalter for the edition of 1481. Thus all the evidences of his literary activity that we possess are included in a period of four years, 1 478-1481 ; but of course the collection of materials for his lexicon must have been the work of his lifetime. We have thus rapidly passed in review the principal agents in Greeks and the revival of Greek literary studies in Italy; the patrons, Palla Strozzi, Cosimo de' Medici, Bessarion, and others; the teachers and professors — Greeks, generally restless, always exacting, continually in 9 c money difficulties ; their influence was immense, but was communicated chiefly in oral discourses which died with them; their writings are scanty, and of no permanent value as literature. The Italians, eagerly striving for a mastery of the language of Homer and Plato, and hanging on the lips of their golden-mouthed teachers, carried the knowledge thus acquired into wider fields, and turned it to practical use, seeking culture rather than learning, and adding elegance and polish in composition to the grammatical niceties and dialectical hair-splitting which often satisfied Greek tastes. The contrast between the master Chalkondulas and the disciple Poliziano is a case in pointy an extreme one, no doubt, but the more essentially typical of the tendency just referred to. Practice of the early printers of Greek. II. So much may suffice as to the men who prepared the way for and carried on the Hellenic revival ; an account of the printers themselves is best given in connexion with the books they produced, and we must now pass to the next division of our subject, and inquire how the first printers of Greek worked, what were their methods and their instruments ; why they employed these methods and instruments ; what kind of results they aimed at obtaining, and how far they succeeded or failed. Descriptions and discussions of the peculiarities of individual types or books will come later ; at present we are con- cerned only with a general survey of the field, and the inferences to be drawn thejefrom : and some attempt must be made first of all to trace the origin and models of the various Greek types which we find in early printed books, and to determine their relation to the manuscripts from which they are derived. The types of the Greek presses in Italy up to the year 1500 may tion of types, ^^g divided by their form into three well-defined series. The first class, which may be called the Older or Early Greek class, includes all books printed under definitely Hellenic influences down to the establishment of the Aldine press in 1494, and excludes all other books. It comprises in reality only two distinct varieties of type. The one is the Milanese type of the 1476 Laskaris (pi. I), designed by or under the eye of Demetrios Damilas ; the second Milanese type 10 Classifica- (pi. II) bears marks of the same hand, and may be considered an adaptation from the first. The other type is that which the Kretan printers, Laonikos and Alexandres, used at Venice in i486 (pi. V). As is well known, the Italian scribes had in the Renascence period Their gone back to earlier models of handwriting, and revived for classical wrking. ° texts the book-hand of the twelfth century. The earliest printers in Italy, reproducing in the form of type the writing of contemporary manuscripts, exercised a wise discretion in rejecting certain character- istics of handwriting, and thus created the roman types, which soon freed themselves from manuscript tradition, and in the hands of a long succession of craftsmen rapidly developed, and degenerated almost as rapidly. In Greek writing, a movement somewhat analogous to that of the Italians, but not strictly parallel, took place at a rather earlier date. From the middle of the fourteenth century onwards two influences seem to have been at work, the one conservative, the other progressive. The writers of vellum manuscripts kept the twelfth- century hand comparatively unchanged, advancing in the direction of greater freedom by very slow degrees ; the writers of service-books and other manuscripts for church use were most rigidly opposed to innovation, and were helped by the continued use of vellum for service- books, while its decreasing use for secular books gradually brought the more moderate representatives of both schools into closer touch. The progressive school, consisting at first only of those who wrote on paper, adopted with the new material a freer hand, which tended to replace simple ligatures by complex abbreviations, and to reduce whole words to a labyrinthine tangle of flourishes. While the two hands continued to exist side by side, it was open to Demetrios Damilas, as designer of the first type of genuinely Hellenic character, to choose which he would. It was probably the greater simplicity of the older style rather than any more aesthetic consideratiops that determined his choice, because the experiment of Aldus, eighteen years later, in fashioning a fount based on the current hand, met with almost universal approval from his contemporaries. As the first book printed with the new type was the work of a writer then living, we may surmise with a tolerable amount of assurance that the type was not imitated directly from the writing of the actual manuscript which served as copy for the press, as is usually held, perhaps on somewhat insufficient evidence, to have been the case with the printers of Latin books. II c 2 Imitation of In his first fount Demetrios made a mistake which the greater "'uihedtoo e'^P^^'i^nce of his Italian fellow-craftsmen enabled them to avoid; fa"! ^ **"' but he was followed in it by most of the early designers of Greek founts, though he seems himself to have tried later to escape from it. The blunder lay in the attempt to reproduce not merely the forms of handwriting, but also the effect of continuity naturally produced by the motion of a pen over paper, and therefore right and proper in manuscript, but unsuitable for impressions made on paper by separate stamps laid side by side. Though this practice was largely abandoned in the second Milanese type (pi. II), and many of the changes in the first type, when it was recast for the Homer of 1488 (pi. Ill), were made with a view to its modification, it reappears in a different shape in the Venetian fount of i486 (pi. V), and it was without doubt one of the principal inducements in the adoption by Aldus Manutius of the later style, in which the close imitation of writing is essential to its success. Aldus repeated the mistake some years later in his introduction of italic type, which in his hands, like his Greek founts, revolutionised the whole history of printing with disastrous consequences. The Venice The Milanese types represent more or less a simplified book- 1486 and the ^and of the moderate older school ; the Venice fount is very distinctly MSS. derived from a church-hand of archaic appearance, and resembles the writing of the Gospels dated 1305, reproduced in plate 205 of the first series of the publications of the Palaeographical Society. The ecclesiastical character of the type is so strong that it may be con- jectured that it was cut for a projected series of service-books, which for some reason never went beyond the Psalter, and that the first book for which it is used is of the nature of a type specimen. The book in question, the Batrachomuomachia, is printed in alternate lines of red scholia and black text, so that, while its small size and popular character made it suitable for its purpose, it was also an experiment in the art of printing in two colours, which was indispensable for a liturgical series. The Milanese types, by their greater divergence from the influence of handwriting, show that their designer was a man of considerable originality and ingenuity; the Venetian fount, while agreeing with them in its genuinely Hellenic character, differs widely both in design and execution. In the Venetian type the appearance of continuity is sought by an elaborate system of ligatures, two, three 12 and four letters being commonly cast in one piece, and in an Immense variety of forms and combinations, so that the number of sorts found in the two books exceeds twelve hundred, and even this is probably far from representing the fount in its completed state as projected. The effect was unsatisfactory ; because the ' case ' was complicated to an extraordinary extent by the enormous quantity of boxes required, and the use of so many ligatures resulted (in practice, though not of necessity) in splitting up the longer words into disconnected syllables, a result which makes the books very difficult to read even after considerable experience of them. The first Milanese type of 1476 Construction (pi. I) is constructed on an ingenious plan, which enabled the resem- °ypg ''*^ blance to writing to be kept without the immense labour involved in cutting, casting and composing a vast number of unwieldy sorts. After the fashion adopted later for certain italic or script types, each letter was, with few exceptions, cut and cast separately, and the white space between the letters was reduced to the smallest possible amount by producing the end or connecting stroke of certain letters to the very edge of the type body, or, where the letter following began with a concavity or hollow, such as ^ or x. by kerning, or bringing the first letter rather over the edge of the second by means of a pro- jecting shoulder which carried the connecting stroke, and fitted in to a corresponding depression in the hollow of the second letter. In consequence of the adoption of this plan, the number of sorts found in the Laskaris of 1476 is not much more than one-sixth of those in the two Venetian books. The second group or main division of Greek types consists ofGraeco- those used by Italian printers who habitually printed Latin or vernacular books, and were not under direct Hellenic influence, but who used Greek letters to print the passages in that language which occur in such books as Gellius or Lactantius, and sometimes printed a Greek text in parallel columns with a Latin version. These types, which may conveniently be called Graeco-Latin, are easily distinguish- able from the types of the first or Hellenic group. They are as a rule very haphazard in their use of accents and breathings ; the forms of certain letters are often clumsy, and their employment rather wild ; each letter is cut separately, after the fashion of a Roman or Gothic fount, without any attempt at continuity. They were made by work- men accustomed to Latin types, who faced the problems connected 13 with Greek letters from the outside, from a different point of view from that of the Hellenic type-founder, and applied their technical knowledge and practice to the production of Greek forms in a Latin spirit. The Graeco-Latin group as a whole may be subdivided into three sections differing from each other in their origin. The types of the first class, which form the great majority, must be held to be copied from the writing used in the manuscripts for the Greek quota- tions in the text of classical authors, and rest wholly on Western tradition ; of these founts there are two essentially distinct varieties : the Roman-Greek, wide and spreading, without any accents or breath- ings, and the Venetian-Greek, a more compact and regular kind of fount, in which the accents are generally arranged on the 'cutting-out' system, which I shall explain shortly. Those of the second class, repre- sented by the type of 1489 used by Leonardus Achates at Vicenza (pi. XII), are copies of Hellenic types of the older style; the Vicenza type, for instance, is derived from the second Milanese fount. The third class is represented by only one type in the fifteenth century, that used at Reggio and Modena from 1497 to 1499 by Dionysius Bertochus. This class consists of copies from Hellenic types of the new style, and the type of Bertochus (pi. XVIII) is a rude imitation of the first Aldine fount. Of these three classes only the first has any claim to historical continuity with the past, and with manuscript tradition. Their rela- A priori it would Seem probable that an Italian printer, in adapting MSS° ^^^ ^^^ ^^^ purpose of his Roman type the Latin writing of the manuscripts, would also adopt for his Greek letters the writing used for the incidental Greek passages in his texts ; and that the two clearly defined varieties of our Graeco-Latin types, the Roman and Venetian, would each represent in a modified form the writing found in one distinct class of manuscripts. But as far as can be ascertained this seems not to be entirely the case. The fifteenth-century manuscripts of such writers as Aulus Gellius, Macrobius or Lactantius can be for the most part divided into two classes as regards the treatment of the Greek passages in them. In the one these passages were left blank by the writer of the manuscript, and were supplied by another hand, either, as in the case of the finer and more carefully executed books, at the same time and by a skilful writer of Greek, or, in ordinary cases, later, by some one, probably an owner or reader who knew and could write the language more or less. In the other class of H manuscripts the Italian scribe knew the elements of Greek writing, and did his best with it, so that it differs mainly in a want of firmness and clearness from the hand of a practised writer of Greek. It is of course impossible to say definitely that there are no manuscripts of the period which exhibit forms of writing similar to those of the Graeco-Latin types; it is even possible that the difference between the Greek type of Jenson (fig. 8) and the Greek in an Italian manu- script of Gellius in which the Italian scribe wrote in the Greek himself, is no greater than that between the Roman type of the same printer and the Latin writing of the same manuscript. At present, however, there seems to be no evidence that the distinction between Roman and Venetian founts rests on manuscript tradition, or that the curious parallelism between the arrangement and style of writing of the Graeco-Latin manuscripts of the tenth century, and the arrangement and topography of a Graeco-Latin printed book of the fifteenth, is anything but a coincidence ; though it is tempting to assume the existence of some succession of links connecting such a book as the tenth-century bilingual Psalter in parallel columns from the library of Nicolaus de Cusa, which is figured on plate 128 of the first series of the Palaeographical Society's publications, with the similar Greek and Latin texts produced by Italian printers in the fifteenth century, or the Gospel of John (in the next plate), in which the two languages are written in alternate lines, with a book such as the similarly arranged Batrachomuomachia of circa 1475 (pl- VII) in the John Rylands Library. With Aldus Manutius a new era began. He had, like Damilas, Aldine and to make his choice between the ancient and the modern style. The types. '^^ sound commercial instincts, which were always so prominent in him, led him to depart from precedent, and to bid for popularity by choosing as his model the current modern hand, with all its luxuriance of contortion and extravagance of meaningless abbreviations. With a lesser man the choice would have signified less ; with Aldus it was disastrous. The enormous vogue of his publications and the great number of them exercised an overwhelming influence, affected the whole future history of Greek printing, and inflicted on its aesthetic side a blow from which it has never recovered. With the traditional conservatism of the Church, in this one instance justified by its results, those who printed and read the Orthodox service-books 15 vigorously resisted the innovation, with the exception of a few printers, such as Kallierges, who issued liturgical books only incidentally ; the older and purer type which the majority continued to use, by gradual contamination with the descendants of the Aldine founts, in the Course of time developed into that which is commonly employed at the present time in Hellas itself, and is the standard type of the majority of continental printers. This class of types, which owes its final form largely to Didot, differs in many respects from the so-called Porsonian types generally used in England. These last, which owe their origin to Baskerville in the last century, and were only slightly modified by Porson, represent a revolt against the degenerate de- scendants of the Aldine class of type : superior as they are to the continental founts in legibility and evenness, their dull monotony, ungraceful forms, and general lack of firmness and dignity in spite of all their precision, make them far removed from any standard which can now be accepted as satisfactory from any point of view. The well-designed though somewhat thin-faced and spiky types made by Messrs Decker of Berlin (which have been used in the present work, and are found in a few other books printed in England, such as Wharton's Sappho) and that of Mr Selwyn Image are a welcome sign of a return to better things ; though the modifications to which the last-named has been subjected have deprived it of much of its charm, and the excessive smallness of the scale on which it has been carried out has obscured some of its finest features, and made it difficult to print from. Character- At the time, however, the Aldine types carried all before them : Aldinetypes. *^^ ^^'^^^ forms cease abruptly, as it became a point of honour (and of business) with every printer of Greek books who wished to be in the running, to follow the fashion by basing his type on the common writing hand of the day, the chief characteristics of which, whether written or adapted to the needs of the printer, are an absence of dignity, and a restlessness expressed in the want of restraint in the voluminous curves, the endless variety in the size and form of the letters, and an incredible complexity of abbreviation which makes the deciphering of a Greek text no small difficulty to the inexperienced. These faults are of course hardened and emphasized by their trans- lation from the freedom of handwriting into the fixed mould of type. The loss of dignity is not compensated by the unrestrained freedom ; i6 the vigorous beauty of form so striking in the older types is replaced by letters which at their best are ungraceful, and in all but the most careful hands degenerate into wiry thinness and nerveless imbecility. The earliest founts of the Aldine class vary much in quality ; thus the first of all, the largest of the Aldine types (pi. XV), is one of the worst possible ; the lower-case Florentine type of Lorenzo di Alopa (pi. VI) is comparatively simple, restrained, and solid ; that of Kallierges, though most elaborate and rather thin-faced, is so well designed and cut with such firmness and evenness that the feeling of pleasure in its technical excellence predominates over the dislike produced by the defects which it shares with the rest of its kind. Two types which stand apart from the foregoing may be classed Abnormal together as abnormal. One of these is that of the undated edition ^°""'^' of the Erotemata of Chrusoloras (pi. VIII), which is one of the very first of all printed Greek books. The Greek type of this book is in most respects an anticipation of the later or Aldine class, though framed on Graeco-Latin lines ; in the form of the letters it resembles the later types, and the method adopted for the accents is the same in kind as that of Aldus ; but in other respects the fount is made up of separate letters after the fashion of the Graeco-Latin types of the older class. The second anomalous type is far better known ; it is that (pi. VI) which was specially designed by loannes Laskaris, nominally on the basis of ancient inscriptions, for the Greek press established at Florence in 1494 by Lorenzo di Alopa. As originally planned and at first used, it consists only of large and small capitals, without lower-case, and has an ingenious contrivance for inserting the accents which will be described shortly. Two years later, when the editor and printer desired to add the scholia to the texts of Kallimachos and Apollonios of Rhodes, a lower-case fount was designed, of which some mention has already been made. The types of Greek books having thus been in some sort classified according to their form, we must now proceed to examine the con- struction of the founts, with special reference to the difficulties which their designers met with, the problems which they had to solve, and the various ways in which they overcame or solved them. The great stumblingblock in the way of the designer or founder Treatment of a Greek type is of course the presence of the accents and breathings, byX^elriy and of the iota subscript, which last, however, was usually omitted printers. 17 D in the early types, and need not be considered here. If the different combinations of accents and breathings with the letters be cut on single punches and cast solid, the number of punches to be made is very large, and the size of the case is increased in consequence ; on the other hand, any device for combining them during composition increases the difficulty of composing, and of obtaining an even impression, and also adds greatly to the wear of the type. Modern printers have decided, especially in view of the steady decline of the average compositor into a machine and the altered conditions of labour and supervision in modern printing-offices, that the expense of the additional punches required is preferable to the trouble and loss of time caused by any more intricate method; so that Greek types now, with a few exceptions in the way of kerned letters for the more complex sorts, especially those with iota subscript, are always cast solid. It was quite otherwise with the first printers, who were continually planning new ways of economizing in this direction, though they did not hesitate to multiply different forms of the same letter in order to produce greater variety. Omission of The existing fifteenth-century founts, when this test is applied to accents. them, fall into four groups.' The first plan tried is the simplest of all, and consists in merely omitting the accents altogether. Though this method has distinguished modern support, it must be considered unscientific, and was at any rate not likely to satisfy a native Greek printer. It is in fact a characteristic feature of the Roman class of Graeco-Latin founts, but is not used in any book printed wholly or mainly in Greek, except in the first impressions of Wittenberg (151 1, 1 5 13). But the text of the New Testament in the Complutensian Polyglott is not far removed, as the type, there found in its original state, has no breathings at all, and no accents except an acute. Accents cast Next come the types in which the accents are cast on the letters ; letters ^-^re two ways of lessening their number were tried. One plan, sanctioned by Damilas himself and used in the Laskaris type of 1476, was to omit certain of the less common combinations, and where they occurred to substitute others for them. The details of this practice, which cannot be commended, will be found in the 'Cutting-out' description of the Laskaris itself given in a later chapter. The other ™^' ° ' method is more interesting. It was the prevailing practice in Graeco- Latin types of the Venetian or Jensonian class, and is not unknown 18 to some extent elsewhere. To explain it, a concrete instance will be best. Take the letter a. For a complete set of this letter and its accents and breathings (without the iota subscript) twelve sorts are required: namely, adddaciaaaciaa. A fully developed Jensonian type would have only the last six of these sorts made, and would form the rest by cutting off with a knife from the face of the type what was not required. Thus a can be formed from any of the six by cutting out everything except the letter ; d is made from a or d by cutting out the accent ; d from d and a from d by cutting out the breathing, and so on with the rest. The process is easily seen, because not only is the breathing too far back in an d, and too far forward in an d formed from d, but the cutting was hardly ever done thoroughly, a blur being left, and it was often not done at all, a word like HvLoypq being not at all uncommon in certain founts. The consonants were similarly treated; a large proportion of them were cast with an apostrophe attached, which was intended to be cut out when not wanted. Many examples of this practice may be seen in the plates on which Jensonian founts are reproduced. Modifi- cations of it exist even in the Hellenic class of types, for instance in the Venetian fount of i486, where unaccented sorts are frequently made from the accented form by erasing the accent ; some instances of this are given later in the detailed description of the books in question. A third group consists of the types in which some form of separate Separate combination of letter and accent is adopted. In these the accents are ^""^^'"S- cast separately, and are combined with the letter by the compositor. There are many different ways of doing this. In the most elementary device (pi. XXIII), found in most of the first Greek books printed at Paris (1507), the accents were cast on a body of the same size as that of the letters, and were nlade to occupy only the lower half of its face. The page was then made up of alternate lines of letters and accents ; the latter were arranged over the letters in the line below to which they belonged, and the line was filled up with spaces or quadrats. Thus each line of text occupied twice the depth of the body of the type, and is arranged thus : the upper half of line i is white, the accents occupy the second half, and the letters take the whole of line 2. The effect thus produced is that of a heavily leaded page, such as were printed at Leipzig and elsewhere for school use, to be interlined with writing. But this method of printing Greek, 19 D 2 besides being clumsy, wasted a great deal of space, and was soon abandoned by Gourmont in favour of a more advanced system. The next plan is, instead of having a second line for accents, to work by means of a split body, somewhat in the manner now adopted for printing Hebrew with vowel points. Two slightly different varieties of this system exist. One of these is adopted in the early Vicenza Chrusoloras (pi. VIII), the other in the majuscule fount designed by loannes Laskaris (pi. VI). In the latter, which is the simpler, and there- fore the more typical, the accents occupy a space equivalent to the difference in height between the large and the small capitals. Thus, while the large capitals occupy the whole depth of the body, the small capitals have a body only some two-thirds of their depth, and the accents are cast on a minute body about one-third of the whole ; they are thus arranged in a kind of trough, and are placed above their letters and blocked up in the same way as that described above in the case The Aidine of the Paris books. The first two Aldine types and the three founts method. i,ased on them are arranged on a new system, which includes not only accents and breathing, but also abbreviations. From the evidence of slightly later types, which still exist, such as the French Greek founts originally commissioned by Francis I, there is no doubt that the Aldine method was to fit both accents and abbreviations over the letters to which they belonged by means of kerning, very much as the modern printer does in extreme cases, as for instance when he has to make an ^ ; he has
&fco)J TToNx) u&i^opct Mi^Oop rois* S^&xaKoi0' a^iicoi^rb
xoXop Kai GvMop &r&if cop. id efl; icorruptibilis ec condicor fternns
m acre babicins . bonis bonii^fercns.&(hs mulco niaiore mercedem.ni''
itMsaut Si malis imm ecBirorem ex:dmn9.1lui*ru5aliolocoenumeKins.
^bnsmaxtefdcinoribusihcirecdaisb|cninilic.(J3&vrfc ^& Narf&iatf
apowTovc Geco^copri Nocrftvc noixtiatf-rfc^^urNoctfce. i^cti
afr&pos* ocKf i'rop &vpHp i^i»pre.p&apiT may be from
u> or (S, both of which are used. In the facsimile the u of tcu in
line 3 comes from v (see line 6), which itself may be formed, like i,
from a. v. In line 5 the tt of ^mooto, and in line 6 the v of diev and
the second jw in eiujwev, have had an apostrophe partially cut away.
Other instances of this are the i of aiev in line 6, and the second
letter of the last line. The pointed v is alone found, not the square
Jensonian form. The use of a Latin k with the upper part cut
away, as in line 5 of the facsimile, is frequent.
Before leaving this type of Zarotus one more point may be
noticed, which his Cicero shares in part with other books, though it
is very prominent here. It will be seen from the facsimile that the
Greek sentences do not always fit the spaces in which they stand;
in other cases they are left out altogether. The explanation of this
must be that a compositor who knew no Greek left this to be inserted
by a second who did know a little, and made a rough estimate of the
space likely to be required ; in some instances the second man missed
the places where he was wanted, and hence the blanks ; in others the
space wanted was miscalculated, whence comes the appearance shown
in the facsimile. Two examples will serve to show the extent of his
knowledge, or of his inability to read his copy. One is the word
K'eXMoTOiG, which is well supplied with accents, breathings, and
apostrophes; the other is the quotation from Hesiod: —
TH<; h' dpexfiq ibpajra Geol npoirdpoiGev eGHKav
dedvarof iwaKpoq he ml cpGioq otiwoq kc. quthv
Kal TpHXlJg TC TTpCOTOV eiTHV b' ek dKpOV IKHTOl
pHibtH 6h eireiTa weAei, xaAenn nep touoa.
This^ appears in the following form: tho bdp er s ibpcoxa 9eoi npo
apoi e Kav d dvarol jwoKpoc; be kI 6 105 oljwoo endvrco Kal t hx'is to rr cotov
40
its. Ioquu£'juLEJUtetifctor "crov Jsjsi ye Had etKAsfocr ATtoXoijULivdai me
minus iam mouent ut uides Itaque ab bomerimagtu eloquentia co
£ero me ad ueca pcaecepta tov gvpiTTi^ov juliVS o-otp^gi ir
quem uetfiim Ccntx. Prealiiis laudat e^f egjie 8c
ait pode eunde 8C askx. ■mpocrto Kcti oiriVo-o^
uidere 5C tamen oibitominus take up almost the whole of the face. The
accents and breathings are cast separately on a minion body, which
represents the exact difference between the height of the E and that
of the moderately tall letters, such as nr or 6. In the Florence type
the capitals are in the position of the ^, E, and v here, as the accents
fill the exact difference between their height and that of the small
capitals ; but the absence of low letters made the problem simpler,
84
as it was not necessary to cast the other letters in an unusual position
on the face of the type. The Paris type of 1507, again, is much less
ingenious, as there the accents have an independent line to themselves
between each line of text. In our present type the low letters are
very prominent, and as may be seen in lines 15, 16 of the facsimile,
are clear, but only just clear of the accents in the line below. In
the case of the accented \|/ (line 6), it is necessary to assume the
dovetailing of the accent into the letter by means of a file, or some
similar instrument. We must suppose that the non-accented letters
were filled up with minion quadrats ; but the process of adjusting
these and the accents in their proper positions must have been a
wearisome one. The number of variations in the individual letters
is small. The second e and are very rarely found ; the former
only once (leaf 2*, line 3) ; the is introduced from the Latin fount.
The secondary forms of i and x (lines 6, 1 1) are intended to fit closely to
a preceding or following letter respectively. Of tied letters there are
twelve, for the most part insignificant ; among the accents the " is the
most worthy of attention.
Where and when was the book printed ? It is clear from the
character of tire Latin type, that it is of a comparatively early date ;
not later, that is, than 1480 or thereabouts. But both questions
might be difficult to answer, if by a fortunate accident two pieces
of evidence were not forthcoming to determine them within narrow
limits. The roman type, though I have not met it elsewhere, agrees
generally with a class of types used in certain towns of North Italy,
especially Bologna, Vicenza and Treviso, though it is more irregular
than other types of its class. In the Chrusoloras it is heavily leaded,
being a type of english body. Of the Greek type we find the h
(line 4 of plate VIII) and on (line 2) appearing as strays in the Greek
type used at Vicenza by Dionysius Bertochus in 1483 (plate X),
and the ei (line i) occurs in the same type when used by the same
printer or his partner at Venice in the following year (plate XI).
Bertochus, a native of Bologna, was throughout his long and
chequered career emphatically a Philhellene, as he established two
distinct Greek presses, the first at Vicenza in 1483, which he removed
to Venice in 1484, the second at Reggio d'Emilia in 1497, which was
moved to Modena in 1499, and back again to Reggio in 1500 or 1501.
He was continually on the move ; we find him at Bologna from 1474
85
to 1476; during the years 1477 to 1480 he disappears for a time; he
reappears at Vicenza in 1481, moves to Treviso in 1482, is back in
Vicenza in 1483, and settles in Venice in 1484. We may then assume
that there is prima-facie evidence in favour of Vicenza as the place
of origin of the Chrusoloras, as it is there that the stray letters turn
up in 1483. Fortunately, however, more is known of the type than
this. In 1 48 1 Bertochus was in partnership with the printer Giovanni
da Reno, who in 1475 had moved into Vicenza from Santorso, a
village some ten miles to the north. The earliest dated book of his,
printed at Vicenza, that I have seen, is the commentary of Omnibonus
Leonicenus on Cicero, De Oratore, which has no printer's name, but
a colophon giving the place of imprint and the date 11 kal. Ian.
(22 Dec.) 1476. This book actually contains a considerable number
of Greek words printed in the type of the Chrusoloras. The roman
type of the book is english, with a pica face, so that, with the exception
of the three tall letters, our Greek fount matches it well, with very
little trimming, when used, as it is here, without its accents and breath-
ings. The result of the short letters being cast at the top of the body
is clearly seen, the level of the Greek words being decidedly higher
than that of the Latin text. Of the three tall letters, £ does not
appear to be used : the other two have been filed to make them fit
in, without however curtailing the face of the letter ; but the y, even
when it comes under a short letter, such as an a, pushes it up out
of its place in the line above. The ^ is managed better, and has
produced no observable dislocation.
But from the occurrence of the Chrusoloras type in the Omnibonus
of 1476, it seems certain that the fount, in which the accent-system
is an integral part of its design, was not made for that book, to which
some of the letters have had to be forcibly adjusted ; it is necessary
therefore to conclude that in all probability our Chrusoloras is earlier
than December, 1476. But if so, it has a good claim to contest
with the Batrachomuomachia of Ferrandus the honour of being the first
Greek book ever printed. The Laskaris, of course, which is dated
30 Jan. 1476, whether this means 1477 or not, will retain its pride
of place as the first book wholly in Greek, and stands on a different
level as being genuinely Hellenic in its design and in all of its actual
production except the mere press-work ; but the Chrusoloras is never-
theless a Greek text, and is printed with remarkable care and accuracy
86
in view of the difficulties involved ; it is of great interest, as being
almost certainly the work of Italian craftsmen, whether Giovanni da
Reno or another were the printer, and on account of the ingenuity
with which the new problems involved were faced and solved. Though
the solution may not be wholly satisfactory, the designer succeeded
in providing in a fount of less than sixty lower-case sorts for every
possible combination. The book is interesting, moreover, for the
character of its type, which, rough, straggling and over-sloped as
it is, is unique as a deliberate attempt, forestalling Aldus Manutius
by nearly twenty years, to reproduce for the purposes of the press
not the bold and graceful forms of the calligrapher, but the ordinary
familiar penmanship of the time; and for the curiously-linked chain
which leads from the Chrusoloras, which is without any mark of origin,
through the Omnibonus, in which place and date are given, but the
name of the printer must be inferred from the type in which it is
printed, to the partnership of the same printer with Bertochus in 1481,
and the appearance of stray letters of the type in books printed by
Bertochus in 1483 and 1484.
The next Italian printer who produced a Greek book was one The Parma
at Parma, who has not yet been identified. He printed several ™^° °'^^^'
Latin books from the beginning of 1480 to the end of 1481, among
them being editions of Nonius Marcellus, Festus, and Varro, in
which Greek type is used. The Greek book, which is another edition
of the Erotemata of Chrusoloras, printed like the earlier one in parallel
columns of Greek and Latin text, is without any indication of its
origin or date; but both roman and Greek types are identical with
those of the Nonius. But as a considerable number of blank spaces
in the Nonius and its companion volumes point to a deficiency of
Greek letters at that time, the Chrusoloras may be assigned to the
next year, 1481. The Latin type is easily recognisable by certain
peculiarities, some of which it shares with other early Parmese founts,
while some are found in this type alone. Such are the contractions
for us and rum, the e with a cedilla of unusual form, used for ae,
though ae is also common; the 1 with a stroke through it, and the
wide h. The Greek (plate IX) must be derived from that of Wendelin
of Speier or Jenson, as is shown by the double form of n and other
distinctive marks, and rather from the first than the second; but
though boldly designed, and not without merit, it is somewhat rough
87
Dionysius
Bertochus
at Vicenza
and Venice.
and uneven, and does not keep very close to its original. It is
a two-line brevier fount of the cutting-out class (the 1 of lines i , 6, for
instance, is made out of the I of line 8 or ii), and of simple con-
struction, without any kerns or similar devices except the occasional
undercutting of the high t. There are no capitals or iota subscript ;
the accented vowels are very incomplete and rather carelessly used ;
for instance h is almost always printed h or h, the accent not having
been erased, and so with the other rough breathings. The stops
(period, colon, hyphen and question) are supplied from the roman
type. There are only four tied letters, tttt, c9, gg, or, and very few
instances of more than one variety of the same letter ; except it and t,
which are always found, the only one seems to be co, one variety
of which, wide and standing rather on one side, is a feature by which
this type can be easily identified. Both kinds of n and o> are seen
in the word nepicsncoiuevcov in line 2 of plate IX ; the last (o is the
one to which I refer, but it is not usually tipt up so much as in this
instance. The c final goes below the line, while the or rises above
it, just the reverse of what is the case in Jenson's type; the
absence of any v except the pointed form, which is as usual rather
clumsy, and more like a roman v, may also be noticed ; other features
of the type are enumerated in the analysis of the type opposite
the facsimile.
The first Italian printer who printed Greek books with types of
the Graeco-Latin class, and set his name to them, was Dionysius
Bertochus, of whom something has already been said (p. 85). He
is first met with as the associate of Ugo Rugerius in 1474. He
led a wandering life among the towns of Northern Italy, but seems
not to have been his own master till 1483, when his Greek books
printed at Vicenza appeared. In the next year he renewed a former
partnership with a fellow-townsman, Peregrino Pasquale, and moved
his press to Venice. Here too the first book produced by the firm
was Greek, and though it has the name of Pasquale only, it is in
the type of the Vicenza books of the previous year. After this no
more Greek printing was done by Bertochus till 1497; this later
press of his will be spoken of in its place.
At Vicenza in 1483 his work consisted of reprints of two books
first issued some four or five years before by Bonus Accursius, and
the Venice book is a third edition of the Erotemata of Chrusoloras,
88
so that he did not (either then or later) break new ground. The
first of the two Vicenza books has neither imprint nor date ; it is
a reprint of the Latin-Greek vocabulary by Crastonus, which was
one of the first books in which Bonus Accursius used his second type.
In the second place Bertochus reproduced the Lexicon of Crastonus,
the first book with which Bonus Accursius was connected. It has
a full colophon, and is dated November lo, 1483. In this edition
the preface of Crastonus was omitted, but that of Bonus Accursius
was allowed to remain. It was probably fifteen months later, the
date being February 5, 1484, that the Chrusoloras appeared at Venice.
The same type is used for both languages as in the Vicenza books,
but the composition and press-work are inferior. Though Pasquale
was no novice, we might have supposed that the omission of the
name of Bertochus from the colophon represented an actual absence,
did we not know from his books of 1497 how badly, in spite of all
his experience, he could print when he chose.
The type used for the Greek text of these books is a mixture of His type,
at least two founts with a differently-sized face. Of the larger letters,
many if not all appear to be identical with those of the Tortellius
printed at Treviso by Hermann Lichtenstein in 1477, in many copies
of which the name of Michele Manzolo or Manzolino is substituted
for that of Lichtenstein ; the smaller letters are in part at least those
of the type used by the same Manzolo after his removal to Venice
in 1480. In the Vicenza books a large number of the letters are
found of both sizes ; some are of the smaller, and others of the
larger only. Many of the duplicate forms are used in the Lexicon
only, and do not appear in the Vocabulary, Besides this mixture,
which pervades the whole fount, several letters from the type of the
early Chrusoloras appear in these books, though they are entirely
out of keeping with the rest. These letters are h, ei, AA, and cstt.
Of these four only on occurs in the Vocabulary, while the Venice
book is the only one of the three in which M and et are used. The
ordinary ligatures of the type, which are very few, are common to
all the books alike, except ou and oC, which appear only at Venice.
This very composite fount is of english body, with certain of its
accents kerned, as may be seen in lines 18, 19, and 29, 30 of the
page reproduced from the Lexicon (plate X), and lines 23, 24 of
the page from the Venice book (plate XI). Its approximate extent
89 N
may be seen in the analysis, although many of the lesser variations
are difficult to distinguish. It has a set of capitals, K, Y, and Y being
absent ; K is replaced by the roman K, Y by an A reversed, with the
cross-stroke cut out, or by a roman Y, and Y by its corresponding
lower-case letter. Returning to the mixture of types, some instances
of the presence or absence of certain letters may be given. Of the
consonants, the short, rectangular r occurs in the Lexicon only; in
this also a third form of b is once found, as well as the smaller ^
and the larger k; both forms of the last are shown on plate X.
Certain vowels also, including i, are used in that book only. The
variations of £ are instructive. In the Vocabulary the only E is a tall
letter occupying the whole depth of the face ; in the Dictionary, after
being used at the beginning of the book, it was found inconveniently
large, and was replaced by a shorter letter, which in the rest of the
Lexicon and in the Chrusoloras is used to the exclusion of the larger
form. Of the two sizes of letters, a good many, for instance /u, n, u, o,
may be easily distinguished in plate X. Some letters, such as 6,
occur in the smaller size only, though the great mass of the type
belongs to the larger-faced fount. The tall t, rarely used, is not
undercut. In places where a kerned letter abuts on a letter with
a long tail in the line above, especially the x. the file seems to have
been used. At Venice a roman v and c are sometimes found for
V and c, ; this, which does not happen at Vicenza, is a sign of inferior
workmanship. There is also an i from a gothic type, in which the
dot is replaced by a stroke ; this is used both at Vicenza and Venice
as [, and also (only at Venice) for i in the Latin text. Similarly
a gothic i (i. e. m) is found in place of T. Lastly, certain letters, such
as H and d, are of a different form at Venice from those used at
Vicenza.
Leonardus Besides Bertochus, the only printer of Latin books in this period
Vicenza. who ventured on printing Greek, was one Leonhard, of Basel, who
latinised his surname as Achates. He began at Padova in 1473, but
moved to Vicenza in the next year, and renjained there till 1497,
though there are long intervals (1482 to 1489, 1491 to 1497) during
which nothing is known of him. His four Greek books just fill up
the time between these two gaps, though they do not represent
half his total output during those years. The first book was issued
in June 1489, and is a reprint of the Milanese Greek and Latin
90
Laskaris of 1480, the Latin version being by Crastonus. The type His type:
in which it is printed has the same origin, and is an obvious imitation ^"* ^'^'^"
of the second Milanese fount, in which the earlier bilingual Laskaris
was set up. It is two-line brevier in body, a fairly careful copy and
certainly the best type of its class. The analysis shows the extent of
the type; there are few ligatures, but a tolerably complete set of
accented sorts, and a large number of variations in individual letters.
The capitals are not complete ; the 0, X and Y are wanting, and the N
(in 1489; it was supplied in 1490) is that of the roraan type; the A
is simply an A with the cross-stroke cut out. The type is not much
kerned, but the existence of something of the sort is shown in certain
places, where an exceptionally tall letter, like b, comes below a low
letter or a capital ; there is an instance with p and 6 in lines 10 and
II of Plate Xn. On the same page there are many instances of
a letter with a stroke over it. That these strokes do not form part
of the letter as cast can be seen from the fact that they differ every
time the same letter recurs (as p in lines 8, 13, 19, 24, or t and 0,
each twice in line 15), and also because, when attached to a high letter
like T, V, in line 14, or r in 24, they are clearly outside the body of
the letter, and encroach on the preceding line. The difficulty of
attaching small pieces of lead to the letters as required will account
both for the diversity of form, and also for the way in which some
of the strokes, such as those above the a in line 21, or the p in 24,
are bent. Following its model in this also, the Vicenza type has
the iota subscript. The high r and t are very common ; the former
is undercut (rK, line i ; fi", line 1 3), but not the t ; in one place an
i of the roman type is found under the r, a sufficient proof, if any
were needed, that these undercut letters and the short ones which
are joined to them are always separate sorts,
Plate XII is not from the Erotemata of Laskaris, but from another
work by the same writer, entitled TTepl ovo/uaroQ Kai pAjwaTog pipAiov
TpfTov, there printed for the first time. It is often found bound up
with the Erotemata, and may be supposed to have been printed
shortly after the larger work, towards the end of 1489. In the
epilogue, dated from Messina in 1466, Laskaris surveys the field of
Greek grammar and lexicography, and gives some account of his
own writings. As this is almost the only Greek document of any
interest in these books which has not been reprinted by M. Legrand
91 N 2
in his Bibliotheque Helldnique, I have given a transcript of it in an
appendix.
Second state After the appearance of the Erotemata, Achates discarded in his
of the type. Qj-eek books the two-line brevier roman used for that work, and
replaced it by a smaller fount of a body not much larger than pica,
which had been in his possession since 1482. It is not certain whether
this was the cause or the effect of the alterations he proceeded
to make in his Greek type ; but probably it was the cause, since
the body of the Greek was made the same as that of the previously
existing roman type. The result of this recasting was, as might be
expected, lamentable. The old punches were used for the most
part, but various changes had to be made in order to compress the
face of the type to fit the smaller body. An elaborate system of
kerning disposed of most of the high and low letters, but unfortunately
there were now so many kerns that it was not always possible for
the compositor to prevent them from coming against each other in
succeeding lines. The x of line i and the tl, of line 2 on Plate XIII
show what the natural result of this was, the two kerns being left to
fight it out between themselves. A new 9 (line 12) was adopted, and
the smaller ^, which even in the second book by Laskaris had largely
replaced the taller one (only the short one occurs in Plate XII), is
now the only one used ; the or is usually found with its tail chopped
off; new forms of (o take the place of the old ; the iota subscript
is dropt; the upper part of A is cut off and used for A; a roman v
is found for v besides the original form of the letter, and the tall f is
almost entirely discarded.
The effect of these changes is so astoundingiy bad that one can
only wonder how any printer, even if, for the sake of cheapness, he
wished to saye space, could have consented so to ruin a really fine
type. Two books were printed with the recast fount, both editions
of the Erotemata of Chrusoloras ; one of them is dated Sept. i, 1490,
the other Dec. 23, 1491. Both are badly printed; the surface of
the forme was uneven, and did not take the ink or meet the paper
properly ; and for the same reason a single small impression caused
so much injury to the face of the letters, that in the second edition
the type looks like one which had been printed from for years.
The facsimile given in Plate XIII from the edition of 1490 shows
some of the marvellous shifts to which the compositor was reduced.
92
Besides the instance already noted, the second r in rerpa(^a, line 4,
had to be curtailed on account of the b below. The different ways
in which M is treated are a curious study ; in line 8 the second A is
docked, in line 20 the first; in 27 the first is pushed up, and in 30
the two seem to be run together. In lines 8 and 9 the ei has pushed
the i below it out of place ; similarly in line 23 the 6 has displaced
ju in the line above. In line 12 the lower part of the T is entirely
broken off, while in 18 it is bent.
VI.
With Aldus Manutius a fresh period in the history of printing Aldus
opens. Concerning this celebrated man so much has been written ^^""*'"''
that it is unnecessary to -do more here than to refer those who wish
for an account of his life and work to the volumes of Renouard
and Didot. For the business enterprise and eager scholarship of
Aldus no praise could be too high ; the ingenuity and resource dis-
played by him as a printer and the general excellence of his press-work
are beyond question; but the new founts of his invention, whether
Greek, roman or italic, are in each case lamentably devoid of beauty
of form other than that conferred on them by good cutting, and his
overwhelming influence among his contemporaries and successors
secured the ultimate disappearance of the older and purer models.
The list of the Greek books printed by Aldus up to 1500 will be
found on pages 50, 51 ; I add here for comparison an abstract, taken
from the facsimile published in 1892 by M. Henri Omont, of the
price list of such as were then published, which Aldus issued in
October, 1498 ; a document of great interest, only known from a
single copy at Paris. It will then be necessary to discuss certain
questions relating to the order in which the earliest of the Aldine
classics were issued, before describing the founts used for those books,
their peculiarities, and their difference from those which preceded
them.
' Libri graeci impressi. Haec sunt graecorum uoluminum nomina. Abstract of
quae in Thermis Aldi Romani Venetiis impressa sunt ad hunc usque J^su^d bV^'
diem, scilicet primum octobris, m.iid. (A) In grammatica. (i) Ero- Aldus in
temata Constantini Lascaris . . . Venduntur marcellis quattuor. (2) ''*' "
Grammatica Vrbani . . . Venduntur non minoris marcellis quattuor.
93
(3) Canonismata quae thesaurus et cornucopiae appellantur . . . Ven-
duntur minimo, nummo aureo et semis. (4) Grammatica . . . Theodori
Gazae . . . Veneunt aureo nummo, nee minoris. (5) Dictionarium
graecum . . . Minimum pretium est aureus nummus. (B) In poetica.
(6) Theocriti eclogae triginta . . . Venduntur non minus marcellis
octo. (7) Aristophanis . . . comoediae nouem . . . Minimum pretium
Venetiis, aurei nummi duo et semis. (8) Musaei . . . de Herone et
Leandro amantibus, cum interpretatione latina. Venditur marcello.
(C) In logica. (9) Logica Aristotelis . . . Venduntur aureo et semis.
(D) In philosophia. (10) Primum uolumen. Vita Aristotelis . . . Aris-
totelis physicorum libri octo . . . Venduntur ad minimum nummis
aureis duobus. (11) Secundum uolumen. De historia animalium libri
octo . . . Minimum pretium Venetiis nummi aurei duo et semis. (12)
Tertium uolumen. Theophrasti de historia plantarum libri decem . . .
Minimum pretium nummi aurei tres. (13) Quartum uolumen. Aris-
totelis magnorum moralium ad Nicomachum libri duo . . . Minimum
pretium nummi aurei duo. (E) In sacra scriptura. (14) Psalterium
graecum. Venditur marcellis quattuor. (15) Officium in honorem
beatissimae uirginis . . . Venditur marcellis duobus.'
Order of the There is some difficulty in determining the order in which the
first Aidines. ^^.^^ \^qq]^^ ^f Aldus were issued. Mr R. C. Christie, in an admirable
paper contributed to the first volume of Bibliographica, has proved
beyond question that, as regards the dated books, the solution is to
be found in the hypothesis that Aldus at first used the Venetian
method of dating from March i, but soon abandoned it for the
modern style in which the year begins on the first of January.
Mr Christie showed that the time of his change in this practice dates
from the beginning of 1497, at the time of the issue of the second
and third volumes of the Aristotelian series, and that the Venetian
method of dating is used in all books before this time, and in no
books after it, with the doubtful exception of the Grammar of
Urbanus Bolzanius, dated January 1497. Thus it is possible to be
reasonably certain as to the order in which the great majority of
early Aidines appeared.
But the position of the three undated books, the Mousaios,
Galeomuomachia, and Psalter, still remains undetermined. These
were once considered to be the first productions of Aldus, and to
have preceded all the dated books ; this position has been usually
94
abandoned as regards the Psalter, but is universally held of the
Mousaios and Galeomuomachia. The early date attributed to the The Psalter.
Psalter rests on the phrase used by loustinos Dekaduos in his
preface : eboEe /uoi thv GeonveuoTov ptpAov r&v Geicov nptoTOv evruncooai
YaAiW">v . . . uxsnep Tivd wpo&poiwov Kal KHpuKa feioTTpuoiov tcov juer ou rroAu
TuncoGHGOjwevcriv HiwTv Getcov npoeKnejwyai fpacppiv. It is clear that this only
refers to the priority of the Psalter to the rest of the Bible ; Dekaduos
has just been speaking of a projected edition in three languages,
and it is of this that the Psalter is described as a precursor. All
that is definitely known of the date of the Psalter is, that it is earlier
than October 1498, as it appears in the first price list.
The date assigned to the Galeomuomachia rests on similar evidence. Galeomuo-
Didot (Aide Manuce, p. 57) quotes from the preface of Aristoboulos '"^'^^'^'
Apostolios, as follows : ' il crut devoir le publier comme un hdraut,
KHpuKo, prdcurseur des oeuvres de la Grece qui vont 6tre imprimdes.'
This of course if true would be the strongest evidence in favour of
its priority ; but unfortunately Apostolios says nothing of the kind.
These are his words : oiov Tiva KHpuKa TTpoeKrTejw\|/ai rfiq [oii] /lAef ou noAu
TunooGHOOjuevHq 'loivfao The Ionia was a collection of apophthegms
compiled by the writer's father, Michael Apostolios ; and there is
no ground whatever for Didot's paraphrase in general terms. The
Galeomuomachia does not appear in the catalogue of 1498, and was
probably intended for private circulation.
Then we come to the Mousaios, which has a preface by Aldus Mousaios.
himself The statement of Aldus is perfectly clear and precise :
MouoaTov tov naAaioTOTOv holhthv fiGeAHoa npootjwid^eiv Ttj) t6 'ApioToreAei Kai
Tcov (309U)v Totc erepoic auT^Ka hi ejwoO evTuncoGOjwevoiQ. The Mousaios is
thus earlier, but not much earlier, than the first volume of the Greek
philosophic collections which were published between 1495 and 1498.
This first volume appeared on November i, 1495 ; and all that
Aldus' own statement permits us to say is that the Mousaios must
be before that date, while his words infer that the Aristotle, at the
time the preface to the Mousaios was written, was within a measurable
distance of completion. There is no internal evidence whatever
for assuming any one of these three books to be earlier than the
Laskaris of 1494/5.
But any one who has handled the Mousaios must have noticed
the curious make-up of the book. The Greek text is accompanied
95
by a Latin translation on alternate leaves. The first leaf has a title
on the recto, and on the verso notes of omissions in the Greek
text. The second leaf, signed a, contains ^he preface, and two
epigrams by Mousouros. The third leaf is signed b ; it contains on
the recto a translation of the epigram facing it, and on the verso
the beginning of the Latin text, corresponding with the beginning
of the original which is on the recto of leaf 4, signed a 11. Leaf 5
(Latin) is signed c, leaf 6 (Greek) a ui ; leaf 7, b iiii, leaf 8 a lui ;
leaf 9, V, leaf 10, a luu ; leaves 11 and 12 from the middle of the
book, and are both Latin, the inner pages being occupied by two
woodcuts, and an epigram in Greek and Latin; leaf 11 is signed
b vi. Thus the book, if taken to pieces, falls into two sections
entirely independent of each other ; the Greek text, a quire of ten
leaves signed a ; the Latin version, twelve leaves, signed b. How
do these two parts stand towards each other typographically ? The
Greek text is printed all in one type, which is identical with that
used for the Galeomuomachia, the text of the Psalter, the Gaza,
Theokritos and the Aristotle, But three things in the types used
for the Latin portion are noteworthy. In the first place, the roman
shows decided signs of wear, and is the same that first appears (in
a dated book), in a perfectly new condition, as a few lines on the
titlepage of the Theokritos of February 1495/6. Secondly, the only
Greek type used in the Latin part is the smaller fount which is
found first in the Thesauros of August 1496. This also is by no
means new. Thirdly, a paragraph mark is used which I have
found nowhere else earlier than the Grammar of Bolzanius dated
January 1497. That Aldus did actually not possess the smaller
Greek fount at an earlier date than August 1496 can I think be
proved. It is used in the last four volumes of the Aristotle series,
beginning with January 1497, but not in the first. The Latin preface
to the Gaza (Christmas, 1495) is printed in small roman type. In
this preface two Greek words occur, nd9H and jwesa. The first word
in most copies is printed in ordinary Venetian Graeco-Latin letters,
while the second is actually left blank to be filled in by hand. In
other copies there are blanks for both words. The preface to the
Dictionary, printed just two years later, also has Greek phrases,
but here they are set up in the smaller Aldine type.
For these reasons I am convinced that the Latin part of the
96
Mousaios cannot be earlier than 1497, and is probably not before
1498 ; in the Psalter, though it has no roman type, large use is made of
the smaller Greek fount, and it may therefore also be assigned to 1497
at earliest. Like the Psalter, the Mousaios figures in the list of
October 1498 in its completed form ; its position at the end of the
section in which it stands may or may not indicate its recent com-
pletion, as the books are not placed in strict chronological order.
So too the position of the Psalter before the Greek Hours of
December 1497 cannot be relied on as evidence of its priority.
There remain then the Galeomuomachia, and the Greek text
of Mousaios, which are very similar in many respects. There is
a slight difference in the way the signatures are printed; in the
Galeomuomachia the iota adscript of the capitals is used for the
numbers, in the Mousaios the lower-case iota. The paper is similar,
though the watermark differs. The page is much longer in the
Galeomuomachia, having twenty-three or twenty-four lines to the
twenty of the Mousaios. Both however consist of a quire of ten
leaves, and begin on the recto of the first leaf, without a titlepage,
as indeed was natural with such small pieces ; the absence of a title-
page cannot then be taken as an argument in favour of an early
date. Typographically there is little or nothing to choose between
them; the type is in much the same condition in both books. We
may therefore assume without much risk that both were produced
at nearly the same time. The Mousaios is, as we have seen,
earlier than the Organon of November 1495 '> what is its relation
to the Laskaris of February and March in the same year ?
With regard to the Laskaris, two points have to be noticed. Laskaris,
The first is that the Greek type of the Laskaris, while identical in j^^j^ ^'"^ '
design and some other respects with that used in all the other books,
has peculiarities of its own. The body is shorter, being little
larger than great primer, while that of the other books is nearly as
large as double pica; it is also wider, so that there is a greater
amount of white between the letters ; it has a number of letters found
nowhere else, and does not contain a number of ligatures and
abbreviations used by Aldus in all his other books. But the
Mousaios and Galeomuomachia do not agree in this respect with
the Laskaris, but with the Organon and the rest. Secondly, the
preface of Aldus seems conclusive. ' Constantini Lascaris (he writes)
97 o
uiri doctissimi institutiones grammaticas introducendis in literas
graecas adulescentulis quam utilissimas, quoddam quasi praeludium
esse summis nostris laboribus et impendiis, tantoque apparatui ad
imprimenda graeca uolumina omnis generis, fecit cum multitude eorum
qui graecis erudiri literis concupiscunt (nullae enim exstabant impressae
uenales, et petebantur a nobis frequenter), turn status et condicio
horum temporum, et bella ingentia quae nunc totam Italiam infestant.'
Again at the end of the preface he says : ' rudibus et ignaris peritus
literarum graecarum Lascaris institutiones imprimendas curauimus ;
mox eruditis et doctis optimi quique graecorum libri imprimentur.'
It would require strong evidence to upset the claims of the Laskaris,
backed by the evidence of Aldus himself, and of the type, to be
the first of the Aldine series.
There is however a difficulty, which is not affected by the presence
or absence of two small pieces like the Mousaios and Galeomuomachia.
How can it have been possible to recast the type, and to print
a book of the size of the Organon, between the March and November
of the year 1495 ? We may suppose if we like that the letters
used for the Laskaris had been already tried and rejected for the
larger work as unsuitable, and that Aldus used them here to avoid
the entire waste of the fount before putting the metal back into
the melting-pot ; but there is no grain of evidence for this. As
things stand, the time required seems quite incredibly insufficient
for all that had to be done.
First type of Something must now be said of this Greek type of Aldus, so
Aldus. praised both by himself and his contemporaries, and even by modern
writers who were still in the thraldom of the Bodoni and Didot period.
In the preface to the Aldine Psalter, Dekaduos speaks of it thus :
"Aa6oc ToiintKAHv MavouTioq . . . apexfiq ^HAcp koi th irpoq to Hjuerepa KHfeejuov(a
re Kai. OTOprfi thv t&v rpaiuiudrajv lovjcov euapiuoortav Kal ouvGeoiv t!^ toC oketou
vooq e9eOpev oEuthti' e& fdp Aefeiv tov xapoKTHpa, ounep ouk aunq r&v enl to
KaMirpacpelv xeipio69a3v evexdpaSev tbpaiorepov. This passage gives the
clue to the success of the Aldine Greek type. Aldus broke away
from the usage of his predecessors, and produced a type based not
upon the noble and beautiful older book-hand, but on the ordinary
correspondence or business handwriting of his day, involved and
contracted to an extreme degree, but, as writing, not without merit
for its freedom and flowing lines ; and for that very reason eminently
98
unsuited for fixing in the rigid uniformity of type. To avoid this as
far as possible, variants without end of the same letter or contraction
were made, and new combinations, each more extravagant and con-
torted than the last, were incessantly added. The Gaza of December
1495 is an example of the extreme point to which the use of con-
tractions was carried ; in that book long words like eveoTcbc, naparaTiKo^,
wapoKeijuevoQ, jueAAoiv, dopioTO? are represented by a single intricate and
unmeaning convolution. So KecpdAaiov elsewhere, even in the smaller
types. The developments of these exaggerations may be well studied
in the alphabets of the French Royal types, of which some account
will be given in the last chapter. The result of this tendency was
a partial remedy of the first trouble at the expense of the compositor,
whose cases threatened to assume a bulk and complexity likely to make
his work physically impossible except with immense labour. Thus it
was necessary to endeavour to reduce the number of sorts, both for
the compositor's sake, and also doubtless on account of the expense
of cutting so many punches, without diminishing the number of possible
combinations, on which the success of the fount depended. For this
end certain modifications of the usual methods of kerning were
invented, and it is probably to this that the expression in his applica-
tion of 25 February 1495 to the Signoria for privilege partly refers.
The date of application corresponds exactly with that of the publica-
tion of the Laskaris, and is shown by the colophons to the Aristotle
of 1495, the Gaza and the Theokritos, which mention the privilege,
not to be reckoned, as Baschet and Didot assumed, more Veneto,
i.e. 25 February 1496, according to our reckoning. In this document,
Aldus, who in the preface to his volume called Thesauros, issued
in August 1497, states that he had been engaged for more than
six years (annus enim agitur iam septimus) in perfecting a system
of printing in Greek, applies for a copyright in his Greek characters
for ten years on the ground that 'havendo facto intagliar lettere
greche in summa belleza de ogni sorte in questa terra, ne le qual
habbia consumato gran parte della sua faculta cum speranza de doverne
qualche volta conseguir utilita, et za molti anni chel ha consumadi nel
intaglio de le dicte lettere, habia trovato, per la dio gratia, doi novi
modi, cum i qual stampira si ben, e molto meglio in grecho de quello
che se scrive a penna.' Here, besides the comparison with writing His two new
again insisted on, mention is made of ' two new methods ' invented ""^
99 02
by Aldus in connexion with his experiments in type-founding. One
we may feel fairly confident is the adoption of the new style of face
in place of that based on older models ; the other is probably the
contrivance by which the types were cast in such a way as to enable
the compositor to unite a letter with a breathing, accent, mark of
abbreviation or contraction, into a whole which should have the
appearance, as printed, of a single letter. Existing founts of a some-
what later date show what the method of Aldus was. The practices,
already known to printers, of kerning one line of type into another,
of interlocking letters in the same line, and of working the accents
separately by placing them in a trough above their letters, led up
to the invention of Aldus, which was only a combination or develop-
ment of these as regards the lower-case. But the process was not
completed at the time his first book was issued. On sig. 18^ of
the Laskaris of February 1495, in the last line but one, a space has
worked up, as is common in all printed books, and stands level with
the face of the type, so that it has been printed. A space must of
course represent the size of the type-body, if it stands straight and
is unbroken, as is this space. In an ordinary fount it reaches from
a point level with the head of a high letter to one level with the
tail of a low one, e.g. from dot to tail of a j, or roughly speaking,
halfway between a line of type and the lines immediately above
and below it. The space in the Laskaris, however, reaches from
the foot of the short letters in one line to the same place in the
line above ; it is therefore clear that this first attempt of Aldus
differs from all other types in the position of the letter on the body
of the type. In view of the preponderance in Greek of the high
strokes, and to give as much room as possible for the insertion of
the kerned sorts, Aldus had the short letters cast at the foot of the
face, in the lowest possible position ; he shortened the low strokes
as much as possible, and kerned them on to the line below, while
developing the high strokes greatly ; and he also provided in this
way a very long shoulder to support the projecting parts of the accents
and the numerous contractions. But the plan was a failure ; it was
at best a makeshift, an intermediate step in the full development of
the new method, and was probably condemned not less by the
discovery of a way to overcome the difficulties of combining the
separate working of the accents with a type cast on the ordinary plan,
100
than on its own demerits. The fact that the privilege was applied
for at a time when the Laskaris was practically completed, and after
Aldus must have determined to abandon that fount, and to recast
it on a slightly different system, seems to prove that his second
invention, if indeed it is to this point that he refers, must be the
plan (of separate accents attached by kerns) taken as a whole, and
not only that stage of growth marked by the Laskaris type.
The extent of this first type of Aldus in its earliest form is shown The first
to some degree by Aldus himself. In the ' Alphabetum Graecum cum Laskaris!**
multis Uteris ' printed at the end of the Laskaris, Aldus has given
examples of all the varieties of simple letters which his type at that
time contained. Of the capitals there are two forms of E, TT, and Q,
but of no others. In the lower-case, of v there are seven varieties,
of a, (p, CO, five ; of p, t, four ; of f, e, h, 9, A, E, u, three, and of
h, {, I, K, n, 0, TT, p, 8, q, x- W> two. Thus the twenty-five letters
{counting o and c) are increased to seventy-five, and this irrespective
of all accents, breathings, &c. A little further on Aldus gives what
is even more interesting to us, a list, on two pages, of the principal
contractions used, both alone and in position, upon a word of which
they form part. The second of these pages is reproduced on plate
XIV, and I need not further refer to it; the first contains contractions
for av (two), dv, aq, ac, 'aq [i. e. aq with an acute accent on the preceding
syllable], aiq, a"i<;, eq, ev, ev, fiq, hq, hv (two), hv, iv (two), eXq, eiq (two), 6v
(two), 'ov, &v, twenty-five in all. The abbreviations, which were of
course cast with a thin shank, were probably supported on spaces
when printed by themselves, as on this page. The facsimile, together
with the page from the Mousaios which is shown on the next plate,
will give some notion of the complexity of the first Aldine fount and
its wealth of ligatures and contractions.
The fount as finally completed is shown in the page from the and in its
Mousaios reproduced on plate XV. Kerning between the lines
is almost wholly abandoned, though still used to a very small extent,
as in the oGai of line ii, and a space which is found on sig. A i^
of the Psalter proves that the type was now cast in the usual position.
In preference to a longer discussion of the peculiarities of the Aldine
types in general, I have thought it best to point out in the case
of each fount separately the problems it presents, and how far these
support or conflict with the conclusions here arrived at as to the
lOI
methods followed by Aldus and his fellow-craftsmen. In this way
it happens that less detailed notice is taken of the first type than
of the latter ones, because its larger size makes the amount of
it which it is possible to reproduce here very small; and the
greatest attention is bestowed on the 1498 type (which is not Aldine
at all, but a careful imitation), because this has been chosen for
extended analysis on account of its smaller volume. In plate XV
the accented capitals first call for notice. I have already spoken
in the fourth chapter, while treating of the press of Lorenzo di
Alopa, of the way in which the accents were fastened to the capitals
in later types on the Aldine model, and therefore presumably in
the Aldine types themselves. In the lower-case, while the majority
of the accents and other marks are clearly cast separately from
their letters, there are a few which appear to be solid. Such are
the TO in line 3 and line 10, while to in line 7 has its accent separate.
The contraction for tcov in line 4 is almost certainly cast in one
piece ; and the Tcji in the same line, and in line 5, has the circumflex
placed suspiciously low down. Again, it seems that on no other
hypothesis can the presence be explained of sorts with a horizontal
line above them, which seems quite out of place, and cannot have
been intentionally added. See pi, line 4, i in Ouib({p, line 7, and e
in line 10. There is probably a considerable number of different
accents used, but they are difficult to distinguish, as such sorts must
have been specially liable to get bent (hence difference of slope) of
broken, with consequent difference in length. The variant forms
of the same letter can be best discovered by the help of the analysis
facing plate XV, and I need not dwell on them here. The number
of separate abbreviations is fairly large for the size of the page, but
the only one that is at all elaborate is the etvai in line 5, where the
ai, with preceding circumflex, stands above two letters, ei and v, and
has made it necessary to place the breathing in front of the word,
instead of over the ei. The use of the iota subscript is common,
though not universal (Ouibfci) in line 7) ; it seems impossible to
determine whether, as in the French founts of 1544, the iota was
attached by kerns, or was cast in one piece with the vowel ; but the
latter alternative seems the more probable.
Character of But what is to be said of this much-vaunted type of Aldus ? I fear
t etype. ^j^^^. j^^ resemblance to the writing to which they were accustomed,
102
which endeared it to his contemporaries, does not appeal with equal
force to us to-day, nor can we any longer see with the eyes of
a Bodonist, to whom everything beautiful was 'barbarous' and only
the misshapen and ugly were admirable. In truth, in spite of all
his estimable qualities, Aldus seems to have been a man of phenome-
nally bad taste for his time^ and unfortunately the blunders which in
a lesser man would have been unnoticed, the enormous influence of
the books which he produced perpetuated and sanctioned. It was
in vain for Doukas and Ximenez to produce at Alcala, as a striking
antithesis to the prevailing tendency, the most splendid Greek type
ever designed, at a time when the work of Aldus had reached its
fullest development ; or for the Venetian printers of the Greek service-
books to persevere in keeping up the older and better traditions ;
the stream was too strong, the great professional printers, such as
Froben, Estienne, and their contemporaries, caught up the prevailing
fashion and the cause of Greek printing was lost, as that of Latin
was soon to be. To us, whether from the point of view of beauty
or usefulness, the first type of Aldus has no redeeming feature. It
is not even a good specimen of its own class, as may be soon proved
by comparing it with the lower-case Florentine type of the Apollonios
of Rhodes, or that of Kallierges, in which grace and regularity help
to atone for their deficiencies in other respects; that of Aldus, on
the contrary, is not only illegible, but is slipshod and ragged to the
last degree.
This double pica Greek, recast from that used in the Laskaris, Second type
was the only Greek fount possessed by Aldus till 1496, and it con-
tinued in use to some extent till 1498, the Aristophanes of that year
being the last book in which it is found. The second type first
appears in a few words of Gre^k in the Latin preface of Aldus to
the Thesduros of August 1496. It is between two-line brevier and
great primer in body, but the face is disproportionately small. Prac-
tically a reduced copy of the larger type, it shows a firmer and more
practised hand, and avoids many of the extravagances of the earlier
type, while still full of minute variations and elaborate contractions.
Its character can be seen from the page reproduced on plate XVI, from
the ' EnisToAai hacpopaiv of 1499. I have chosen a page which partly
corresponds with that taken from the Phalaris of Bissolus and Mangius
issued the previous year (plate XIX), for purposes of comparison;
103
it is evident that the other type is a copy of that of Aldus. This
Phalaris type in its first state, i. e. as used at Venice, I have chosen
as an example for analysis, to discover if possible the approximate
number of sorts contained in a type of this later class, because
of the small bulk of the only two books in which it is found. To
read through all the books printed by Aldus in any one of his types
would have been an impossible task. But the close relation between
the two founts makes it unnecessary to dwell on this second Aldine
type further than to call attention to its greater simplicity, especially
in the smaller number of detached contractions, as compared either
with its predecessor or with the Phalaris type.
There can be no doubt that in this type, as in the larger one,
the accents are added to the letters by means of kerns ; we con-
tinually find letters clearly the same, differing only in the accent
over them. Instances of this on the page reproduced in plate XVI
are numerous ; in jwh, lines 4 and 5, the slope of the accent over the
H differs; in bet, lines 2 and 10, the accent is rounded in the one,
curly in the other, but the letter is identical in both ; compare also
dAA with oAA in lines 19, 20. About the jwev in line 3, and those in
18, 20, it is hard to be sure whether the accents are different or
identical. The e in line 4 is a different letter from that in line 5,
but that in line 17 seems to be the same as that of 5 with a lower
breathing. The in lines 3 and 4 is the same letter with differently
sloped accents ; on the other hand, in a>, lines i and 6, both letters
and accents differ. In the e of avapaAeoGai, line 4, and of Trepijwevcojwev,
line 5, the letter is different, but the accent, which is eccentric both in
shape and position, appears to be the same. The iiq in roiiq, line 6,
seems odd ; it is possible that the kern has not been fitted on to its
letter properly. Attention may be called to the four kinds of 6 noticed
on this page. The 6 of jwovov, line i, and of avfepoq, line 10, seems to
be the same letter, and the same is the case with TTp6(paaiQ, line 9, and
KoojMOv, line 1 5 ; but the accents are different, though those of /uovov
and TTp69a(3i^ may be identical. The oi-forms are, as usual, arranged
so that the accents stand above both letters, as in ol, line 2, 01, line 5,
o"i, line 8.
Third type The third and last of the fifteenth-century Greek types of Aldus
is used for the Scholia to the Alexipharmaka of Nikandros, annexed
to the Dioskorides of July 1499, which is the last of the Greek books
104
of Aldus.
of Aldus printed in the fifteenth century. This forms an independent
section of the work, and seems to have been added later, as an
afterthought; it is not mentioned on the titlepage, and is wanting
in a large number of copies. But as the type used for it is that in
which the Philostratos of March 1501 was set up, it is certainly to
be reckoned among the fifteenth-century types. It is a pica fount of
very flowing character, with a large number of ligatures, but fewer
contractions than in the larger founts, because the smallness of the
type made the kerned sorts very difficult to handle. As a specimen
of the art of type-founding it is a marvel of skill and ingenuity, and
considering its small size, very legible, owing to the fineness and
uniformity of the lines, and the care with which it is printed. The
capitals are however very unsatisfactory, being both too small and
out of character ; they appear to be largely identical with those used
in the early books (Vergil, Martial) printed in the italic type of Aldus.
In the portion of this type reproduced on plate XVII, two spaces
which have worked up can be seen ; the first, in line 7, shows the full
body of the type in the original, but the facsimile fails to indicate
more than a small part of it; the other, in line 21, is shorter, and
as it is evidently standing properly on its feet, it may be a space
of which the shoulder has been accidentally broken off. But Dr P.
Schwenke has lately observed spaces of similar form in the forty-
two-line Bible ; and he thinks, no doubt rightly, that these are the
spaces belonging to that part of the fount which was modified by
cutting away the shoulder, in order that the letters might stand under
the high f. It is possible, then, that we have in this space an example
of one adapted for holding up a kern, like those in plate XIV, referred
to above ; though, if this were so, one would expect it to be somewhat
shorter than it actually is.
The analysis of this type clearly shows a marked difference from
the larger founts in the number of varieties of single letters ; it is
however accidental that there are exceptionally few contractions in the
piece chosen for reproduction. In smaller variations the minuteness of
the face makes it difficult to distinguish them without a photographic
enlargement of the type, and there are probably far more than those
noted in the analysis. A number of those which are recorded there
differ from each other only in the shape or position of the accents ;
it seems incredible that any separate method of working these could
105 P
have been adopted here, from the difficulty of manipulation, and the
later Greek founts of similar sizes were certainly made without kerns ;
still, the phenomena agree with those observed in the larger types of
this class. Compare for instance the i in oibel (line 1 5) with those in
Totq, lines 25, 26. The first two seem to have the same i, and the same
accent, differently placed ; in the second and third the accent is in the
same position, but the form of the letter differs. In line 29 the h in
EiKH and in ei kh is the same letter, but the accent is different. So
also is the of toO in line 10, and of outoC in line 11. The oi-forms
show the usual peculiarities; for instance the i in oibeT, line 15, and in
dxTic,, line 25, is identical, except that in the second the is partly under
the accent, while in oibei the fee is not. In all other cases where i comes
after a ligature the other T is used, as rotg in lines 25, 26, 27; and
this is a point against separate accents. In the same word oifeet the
breathing is between the and the i, while i, as in line i, differs. It
is possible, if we can place the cutting of this type after October 1 500,
at which time the press of Kallierges and Blastos ceased working, and
their privilege lapsed, that the present fount was made on their system
of separate punches and combined matrices ; this would help to explain
the facts, but involves assumptions which have no direct evidence to
support them.
Imitations of The Aldine press having thus been dealt with in detail, we have
ine types. ^^^^ ^^ consider the imitations of its types which appeared during
Bertochusat our period. Of the two printing firms which come under this head,
Mo^^ena^" the first is our old friend Dionysius Bertochus, who abandoned his
ancient ways to follow the new fashion. After remaining at Venice,
when he returned thither from Bologna in 1489, till 1494, he moved
once more to Reggio d'Emilia, where in company, partly at least,
with a fellow-townsman, Marcantonio Bazalieri, he established in
1496 a new press for printing both in Greek and Latin. While
at Venice, he had used (for instance in the Perottus, Cornu copiae,
of 1494) an ordinary Venetian fount of the Graeco-Latin class; at
Reggio he set himself to reprint line for line the books issued by
the first Milanese press (as he had already done before, in 1483), with
a fount copied from the first Aldine Greek type. Only two Greek
books, so far as is known, were issued from his press at this time ;
one is a second reprint of the Latin-Greek vocabulary of Crastonus,
which was one of the first books printed for Bonus Accursius in the
106
later type ; the other is the third part of the same editor's Aesop,
containing select fables with a word-for-word Latin translation. Both
books are reprinted from the earlier editions without any attempt at
revision, and are moreover very badly and carelessly done. They
are both dated 1497, without any month or day being given, but
from the state of the type it seems fairly certain that the Crastonus,
in which the name of Bazalieri is joined with that of Bertochus in
the colophon, is the earlier of the two.
Of the first quire in the Crastonus there are two different editions, The two
probably due to an accident with the formes. Both are equally voc^uiary^
incorrect, some mistakes being common to both, others appearing
in one only. Some of the variations are interestmg as proving
beyond dispute that at least some of the accents were inserted above
the letters during the actual composition. The edition I call A has
si^atures A ii, A iii to the third and fourth leaves, containing the
beginning of the text (the first leaf is blank, and the second, which
has no signature, is filled with the Latin preface) ; edition B has
the signature Aii on the second leaf, and no signature on leaves
3 and 4. A few examples out of many of errors common to both
A and B are cmroTTTTaTToc (3^), although the second n differs in the two
editions; dnTOGjWHSiQ (4*), TrposirAipoo) (5*), jwoikoq, irapaKAuxoc (6*), ra£o-
(puAaKiov (6^) ; on 'j^ fecoprfa is printed reoopv(a in both, though with
a different v in the two editions. The word loobuvajweco is loo&Hvajueo)
in A, oobHvajweo) in B. Errors in A, correct in B, are : anonejwno for
otTOneiWTra), 3*; TipoGKiAtco for npooKuAio), and emppHaaxiKoq, eTTippijwaTiKooq,
dTtKeijuat for avrfKeiiuai, 6*; e9aiwvAAoc for eqjajuiAAog, 6^ ; cum with m
upside down, 8*. Mistakes in B, correct in A, are equally numerous;
such are cKpaipeicTiKoq for dcpaipeTiKoq, 3*; nposepKOjuai, 6*; AeuKOOTHg (no
accent) for Acukoth^, 8*. In B the word buoxuto on 7^ has a roman i in
place of the second iota. Of other differences not involving spelling,
those in which the position of the accents differs (and these are very
numerous) need alone detain us. On the first page of the text pafruAoq
has its accent higher up and further back in B; dTTepxojwai has the
accent in front of the letter and tipped back in A ; eAdnvoq the same,
more pronounced ; dneAasThK; in A has the accent beyond the letter, in
B over the first limb of the eta ; dnopdAAco in B is like eAdjivoc in A.
In oxytone words ending with a consonant the accent is frequently
placed over, or even beyond the consonant in one edition, over the
107 P 2
vowel in the other; thus, leaf 8\ nrepcoTOQ, kuPcutikoq, 7^, aiwvoq, 6^,
oiKobojUHTHC, have the accent wrong in A, and afoiv on 7^ in B. But
these last are of course not so decisive as medial accents ; on 6\
Kaioj has the accent on the i in A, between a and i in B, the diphthong
being in both editions a single sort with ligature ; on the same page,
jwoixeta is exactly similar; on 5^, the accent of eioiroieoj is behind the
e in A, before it in B. It would be easy, but is needless, to multiply
instances of this sort. It is quite clear that Bertochus copied the
Aldine scheme for accents, but did not know how to cast the kerns
properly, or how to compose them when cast. The spaces which
stand too high and have been printed are numerous (there are two
on leaf 5^ of the Crastonus, edition B) ; they stand rather low, so
that an accent comes some distance up a space standing in the line
above. Thus the second space on the page referred to stands over
an d ; the accent of the d rises in front of the space the greater part
of a millimetre above its foot. The accents rise above the tailed
letters, such as p, in exactly the same way. Some of the accents,
however, are not independent of the body of the type in this way,
for instance the ii in line 4, the 6 in line 6, the e in line 7, and the
H, H in line 8 of the facsimile, plate XVIII; compare these with the
H, 6 of line 3, the o) of line 5, or the 0, d of line 15 ; or compare e of
line 13 with e in line i ; and it will be clear that the accents are of
two sorts, separable and inseparable. Examples of the setting for-
ward of the separable accent are found in the page of the Aesop
reproduced in plate XVIII, in nepi, line i, eTieibH and Kudv, line 7,
and aiiTOQ, line 1 1 .
Type of The type, a rather large two-line brevier, in which the Crastonus
and Aesop are printed is, as already mentioned, an imitation of the
first Aldine type ; it is however exceedingly rude and unskilful, and
being a bad copy of a bad fount, is of unspeakable baseness. Most of
the larger and all the most elaborate ligatures are not reproduced, and
there seem to be none of the contractions which could be placed over
or fixed to the preceding letter, except ov, and perhaps one or two
more. On the other hand, kerning after the older fashion, both vertical
and horizontal, is more used than in any of Aldus' books after the
Laskaris. It is easy, in fact, to discern that the type-founder, if he
were not Bertochus himself, was, like him, accustomed to the older
Graeco- Roman founts ; and, while he thought himself compelled to
108
Bertochus.
follow the fashion in the form his type took, he was unable to carry
out the new methods in their entirety, and has, so to speak, grafted
the modern shape of the letters and the new way of accenting on
habits and practices formed or learnt in an older school of printing.
In 1498 Bertochus moved once more, and took his press and Bertochus
types with him from Reggio to Modena. Established here in theModena;
basement of a house, he printed an edition of the Lexicon of Crastonus
' impressum in aedibus Dionysii Bertochi bononiensis subterraneis,'
which was finished in October 1499. This was not his first book
since his departure from Reggio, as an edition of the poems of
Tibaldeo had been issued in May of the same year. The lexicon is
a far more ambitious work than either of the Reggio books. It is
much more accurately printed, and the press-work shows great
improvement ; it seems strange, indeed, that an experienced man
like Bertochus, who had been a pirinter for a quarter of a century
and had been associated as an expert with some of the best printers
of his time, should have produced work so unworthy of him as the
two books printed at Reggio. Their inferiority may have been due
to difficulties experienced with the new and no doubt extremely
perplexing way of printing the Greek, or to accidental and temporary
circumstances of which we know nothing. The Lexicon, a folio of
considerable size and bulk, was, as originally planned, a mere copy
of the Milanese edition, or of Bertochus' own earlier reprint of
1483, and it is printed in the same types, both for Latin and Greek,
as had been used at Reggio. In this form it was finished on
20 October 1499. Afterwards, however, there was added at the
end a Latin index, adapted by one Ambrosius of Reggio from the
similar index appended to the Aldine edition of i49T- I" this index,
besides the two earlier types, a new small roman fount is used for
the text and preface. The preface is dated, 'Regii Lepidi tertio
nonas lulias. M.D. ;' thus it was printed at least nine months after
the Lexicon itself. Now Bertochus was still at Modena in May 1500,
the date of his Martianus Capella ; that book does not contain any of
the smaller roman type ; but this is found in an undated book printed
by Bertochus, the poem of Demetrios Moschos entitled To Ka8' 'EAevHv
KOI 'MeEavbpov, dedicated by Ponticus Virunius to Louis XII of
France. The colophon is, ' Rhegii Lingobardiae presbyter Dionysius
impressit.' In this little book three types are used; the small roman
109
in question ; a larger roman, different from that of the three Greek
books just considered, and a small Greek type. What then is the
date of this book, printed by Bertochus at Reggio ? First we may
notice that when the index to the Crastonus was printed, Bertochus
possessed no small Greek fount, and was much hampered in con-
sequence. Secondly, the use of a new larger roman type seems to
indicate a later date than the Lexicon. Thus we may conclude that
it is later than July 1500. Thirdly, the small Greek of the Moschos
is actually the same as that used for the Souidas, printed at Milano in
November 1499, under the superintendence of Chalkondulas by Bissolus
and Mangius of Carpi. When we ask how this Milano type came
into the hands of Bertochus, the answer is given by a book of which
I copy the description from Panzer (viii. 243. 2) : ' Erotemata Guarini
cum Libanii opusculo de modo epistolarum, Graece. In fine : Impensis
nobilis Simonis Bombasii et sociorum Pontici Virunii et Presbyteri
Dionysii Bertochi, Benedictus Manzius impressit Regii Lingobardiae
MDI. die X. lulii.' From this we may be reasonably certain that
the Moschos is not earlier than the beginning of 1501 ; and that
and back at Some time between May 1500 and July 1501 Bertochus returned
Reggio° from Modena to Reggio; but at which place the index to the Crastonus
was printed there is nothing definite to show, though the circumstantial
evidence from the preface and the use of the small roman type points
to Reggio rather than Modena, and the shortness of the interval
between May 15 and July 5 is, in view of the small distance between
the two towns, not an argument of much weight on the other side.
Bissolus and The Greek fount just mentioned as being used for the Milano
Ve^nice^i^fqS. Souidas of 1 499 made its first appearance a year earlier at Venice,
in connexion with a press which seems to have been intended to
become a rival to Aldus, if we may judge from the deliberate way
in which he is ignored in the prefaces to the two books which alone
appeared as the result of the efforts of the promoters. These books
are both thin quartos, one containing the Life and Fables of Aesop,
the other the letters of Phalaris, Apollonios, Brutus and Krates.
The names of those composing the firm appear in both books, but
more fully in the Aesop ; they were Bartholomaeus Pelusius of
Capodistria, Gabriel Braccius, or Braccio, of Brisighella, loannes
Bissolus and Benedictus Mangius of Carpi. Of these the two last
wfere the printers ; the first two were editors. The Phalaris,
no
dated 18 June 1498, was the first of the two books to be published,
and is dedicated to Pietro Contareno by Braccio in a Latin preface,
which is amusing for the studied insult to Aldus contained in it.
' Cum^ omnium (he says) atque adeo cotidianis querelis rei literariae
calamitas deploretur, quae librariorum impressorumque incuria indies
diffunditur latius, incredibile dictu, nee minus foedum, nuUos tarn
diu bonarum artium cultores exstitisse, qui sacratissimarum literarum
numen uelut a profanis assererent, mysteriumque hoc, ut ita dicam,
imprimendorum librorum si non studiorum antistites, at initiati uel
cum sordium suspitione susciperent ; hoc uero tempore non desunt,
qui hoc uere publicum negotium priuato otio libentissime praeferant,
hoc maxime freti, quod inuidorum impetus, quos non defuturos iam
nunc satis perspectum est, te patrono facile sustinere posse confidant.'
He goes on to say that they intend to print a Latin version of the
Letters, so arranged as to interleave with the Greek text (after
the fashion of the Aldine Mousaios as finally completed), and to
correspond line for line and page for page. The preface to the
Aesop, which has no date beyond that of the year, is also by Braccio.
He refers to the intention of himself and his companions to print
both Greek and Latin authors, and to begin with the Greek, as the
foundation of Latin literature : speaks of the Phalaris as printed,
and goes on thus : ' Vitam Aesopi, fabulas, et epistulas Phalaridis
noster Bartholomaeus lustinopolitanus uertit in latinam ita ut uerbum
de uerbo expresserit seorsum, alioque uolumine, id quod decentius
et commodius uisum est, haberi uoluimus graeca a latinis, perpetuoque
ordine et paginarum et uersuum sibi singula respondere.' It is note-
worthy that in these two prefaces Braccio adopted the system of
accenting Latin w'hich is usually attributed to the initiative of Aldus
two or three years later : his opinion may have been current before
he ventured to carry it out in his printed books.
These two quartos, with an edition of Ficinus, de triplici uita. They leave
in Latin only, dated 1498, but without printer's name, represent the suddenly
whole output of this ambitious undertaking; and there are ^^"yiJnano*"
signs that some disaster overtook the firm. To begin with, the Latin
versions of Phalaris and Aesop spoken of in the prefaces were never
published ; and the letters of Phalaris, though protected by a ten
years' privilege, were reprinted with impunity by Aldus the very next
year in his collection of the Greek letter-writers. Nor can it have
III
been long after June 1498 that Bissolus and Mangius, the two
printers, left Venice and betook themselves to Milano, where they
were employed by Chalkondulas on the great Souidas which was
finished in November of the next year. Thus there was a dissolution
of partnership, and a sudden flight of the printers, which involved,
we must suppose, the withdrawal of the privilege granted to them ;
and the large differences between their type as used at the two
places suggests, though this may be illusory, that they were forced
to abandon their stock and could not carry away even the whole
of their punches. What is certain is, that a large proportion of the
letters are new in 1499 ; that a new, larger type appears first in
the Souidas, and though used only on a single page, and so far as
I know never found again, it was clearly a complete fount of similar
size to the smaller one. The cause of the catastrophe was most
probably some action by Aldus, intended to protect his copyright
in the method of printing Greek invented by him, a copyright
certainly infringed by Bissolus and Mangius ; and the false assertion
under which the privilege quoted below was obtained is likely to
The Souidas have made matters worse for them. The Souidas is an enormously
'■^gg- voluminous book. It has 516 leaves, of which four have only so
much Greek between them as would fill one ordinary leaf, so 513
may be taken as the number, that is 1,626 pages. Each page has
45 lines, and each line has about 45 to 55 letters, or single sorts ;
allowing for the blank spaces at the end of paragraphs, of which there
are few, 45 may be taken as an average. Thus we have (513x2 = )
1,026 pages with an average of (45 x 45 = ) 2,025 letters, or 2,077,650
letters in the book. To do all the punch-cutting and casting required,
and to edit and print off a volume of this size in a time which cannot
in any case exceed fifteen to sixteen months, was a marvellous feat, and
justified those concerned in the production of the book in the laudatory
dialogue reproduced on plate XX and the epigrams which they addressed
to each other and to their readers. The persons in question were
the printers, Bissolus and Mangius, whose device, representing two
flowering branches on a black ground, with the motto ' Sudauit et
alsit,' and the initials IB BM, is at the end of the book, and replaces
the mark (apparently intended for a pine-cone) which had been used
at Venice. Both devices are reproduced by Kristeller (Italienische
Buchdruckerzeichen, 67, 68). The editor Chalkondulas, who had
112
been, as we have seen, professor of Greek at Milano since 1492, is
associated with the printers in the colophon, and wrote a Greek
preface which gives some interesting information about the printing
of the book. Giovanni Maria Cataneo wrote the Latin preface, and
Antonio Motta some epigrams inserted at the beginning; but as
the writer of the poem addressed to Chalkondulas, printed after the
colophon, names them as participators in the book, they probably
had some more intimate connexion with it. In this poem there is
a typographical curiosity which I have not seen elsewhere. Two
words in different lines, que and iam, having been accidentally
omitted, they were stamped in with types by hand on the margin,
and the place for their insertion was indicated with a pen. The
Latin preface of Cataneo mentions the changes made in the type,
though no previous work of the printers is spoken of After
describing the hitherto unsatisfied desire of scholars for an edition
of Souidas (which Aldus had previously intended to produce, as is
shown by the incomplete document reproduced by Baschet, Aldo
Manuzio, Lettres et documents, p. 3, apparently of the early part
of 1499), he continues : ' tandem ad hanc prouinciam reseruatus uir
atticae facundiae princeps Demetrius Chalcondyles praeceptor noster
non, ut ceteri, graecorum studiosis tantam felicitatem inuidit, sed
ducem se constituens egregios huius artis et industrios artifices
loannem Bisolum et Benedictum Mangium Carpenses accersiuit ;
per quos, typis in melius reformatis, additis etiam plerisque et magnae
et admirandae gratiae, quippe qui in eo genere praestantissimi sint ;
et praeter conditionem et aetatem suam, plurimis multoties coUatis
exemplaribus emendandum, immo excolendum et renouandum Suidam
aggreditur, tanto studio et diligentia usus, ut . . . in illo expoliendo
auctorem ipsum superauerit.' The Greek preface of Chalkondulas
begins thus : To napov pipAfov Zoutba TeriincoTai jwev uno BevebiKTou Ma^ou
KOI ' loidvvou BisoAou tcov Kapnafcov &v 6 juev eucpuHq a)v koi netpav ouk oAffHv
eoxHKcbc ev th tcov eAAHviK&v fpajUA^oiTcov euapjw6oT(p ouvBesei, Gnoubfi xe kql
npoGujuia xRHOoiuevou, oufeev irapHKev eq buvajwiv tAv eiq opGHv guvtoHlv
KQi cjujujweTpfav Ttov TTpog aAAHAa GTOixeicov koi ouAAapcov ouvreivovTcov ei jmh
TTOLi Ti ev TOGOUTti) ouvTafjwaTi napecoparai. ' IcodvvHq he apioToq ojv rpajwjwaTO-
rAu90c, Kai TO KoAAiGxa tcov fpajUjwoiTcov kqG" ogov oiov Te hv eiq aKpov
eKiuijwHGdjwevoq toioCtov xapoKTHpa rpaMM«TCov anojehesac, eyei, oiov ecTiv opav
ev T b* dvrnrdAouc. veOoe Zeiiq- ot fdp 09 ipfig
eAAdbog lAAdvcov naiol npenouoi Tijnoi.
'Appearing out of the unknown, the soaring eagle on a sudden
turns to flight a host of lesser birds; mounted on his car, the sun
dims his sister's beams, and effaces the light of the stars. So
before these characters shrink back the former letters, creatures of
file and reeds. I marvel how by the cuts of the fashioning graver
one shaped thus the row of intertwined types, and how he fixed
the minute accents between the straightest of lines, hanging them all
on the vowels. But why wonder I at Kretan wit ? for aforetime
by the best of her sire Athena learned them many crafts. A Kretan
fashioned the letters, and a Kretan joined together the pieces of
brass; a Kretan pricked them into one, and a Kretan cast them
in lead. A Kretan pays for all, who bears a name of victory ; he
121 R
who sings now is a Kretan. To Kretans the Kretan aigis-bearer
is kindly. Wherefore let us pray with one accord, that the sire
of our patron may have given a name of true prophetic meaning
to his child, and may he vanquish his rivals. Zeus nods yeasay:
for to the sons of Hellenes the types from sacred Hellas excel'
The lesser difificulties in this need not detain us long. The
last clause seems to mean that Hellenes should prefer books printed
by a Greek firm to those of Italians like Aldus. The expression
'creatures of file and reeds,' as applied to the older types, is
puzzling ; the reed must apply to writing, and the phrase will then
mean ' the first printed books and the manuscripts.' But this is
hardly satisfactory, because a reference to writing, especially as
rpa^A^ara, seems out of place. It is at least interesting to have
contemporary evidence of the large use made of the file by the
compositors of the early press. The name of victory in line 15 is
of course NikoAooc (BAasToq). We now come to the middle of the
poem, which fcontains the real crux, and it seems best to give the
interpretation of Didot (Aide Manuce, p. 549 sqq.), who had actual
experience of printing, together with his notes, placed in brackets,
and to comment on that, rather than to attempt any explanation
of my own.
' Ainsi ont disparu les caracteres ant^rieurs, ces produits de la
lime et du roseau [il indique par \k les essais plus ou moins informes
des types grecs que Ton rencontre quelquefois dans les Editions
princeps des auteurs latins imprimis par Jean Schoefer a Mayence,
et par Vindelin de Spire et Nicolas Janson a Venise], et j'admire
comment a I'aide du burin fut sculpt^e et ciselde cette rangde de
types si compliquds, et comment on est parvenu a fixer les accents,
presque insaisissables, suspendus et si bien d'aplomb sur les voyelles
entre ces rangdes de lignes. [Prdc^demment on fondait s^par^ment
les accents, et on ajustait ces petites pieces dans les entrelignes,
en les pla9ant sur les lettres plus ou moins exactement. II fallait
done, en composant une ligne de ces accents, les disposer de maniere
que chacun d'eux se trouvit placd juste au-dessus de la lettre qu'il
devait completer. Mais ce procdd^ imparfait, employe k Paris par
Gourmont pour ses impressions, et k Anvers par Martin d'Alost, et
par d'autres, fut bient6t abandonnd. Le moyen si ing6nieusement
et si exactement ddcrit par Musurus, qui ddija avait et^ adoptd par
122
Aide, fut un immense progres pour rimpression du grec] . . . C'est
un Crdtois qui a cisel6 ces poin9ons [les accents graves s^par^ment
et aussi sur acier devaient 6tre disposes de mani^re h. pouvoir
s'adapter sur le poinfon au moyen dune encoche], c'est un Crdtois
qui a adapt6 les petites pieces d'airain [ces accents, ajout^s successive-
ment, KaBf Iv, sur chaque poin9on, qui ^tait encochd, ne formaient
plus qu'un seul et m^me poin9on de ces deux pieces lides ensemble
par un fil solide], c'est un Crdtois qui les a rdunies, c'est un Crdtois
qui les a accoupl^es [c'est de la reunion des accents en les liant sur
les poin9ons qu'il est ici question], c'est un Crdtois qui les a enfonc^es
[dans une matrice a cuivre], et c'est un Cretois qui a obtenu la
fonte des lettres en plomb [au moyen de ces matrices].'
Mr Didot, though probably right in his general conclusions,
seems to me unfortunate in many of his details. His translation
of the two critical passages suffers in the first by a looseness of
paraphrase which fails to render the original, and in the second by
his making six processes out of the four described in the text. His
explanation of the pfvH kqI hbvaKec, seems insufficient ; what do the
words mean ? And Mousouros could not refer to the Graeco-Latin
types only, and deliberately ignore all the preceding Greek types,
when he speaks of rd npocGev fpdjWjwaTa ; nor would there be any
point in a comparison of a book wholly Greek with a Latin text
containing Greek sentences. Then the process described in Didot's
second note is, as we have seen, one of the most uncommon ways
of inserting accents, and found in only two types of the fifteenth
century, both of Didot's examples (as to the Antwerp type he is
quite wrong) being of the sixteenth century, and therefore not
applicable in the present case. Again, whatever view we may
take of the present poem, there can be no question that the Aldine
process was entirely different. The three questions that have to be
answered are ; first, what is the meaning of the passage eHeOjuai . . .
eniKpe/udcag ? second, what process is described in the second passage
(KpHq rap . . . jMoAupboxtTHQ) ? third, how can the two be reconciled ?
In the first passage, the first two lines are clear enough, despite
a doubt as to the exact sense of kottIq (the knife-edge, or the cuts
produced by it) ; the word ijeptnAeKToc describes admirably the general
effect of the type on the reader, and probably also on the unfortunate
compositor. It does not much matter whether aonrroQ is to be
123 R 2
rendered ' untouchable,' i. e. because of their smallness, or ' invincible,'
i. e. unrivalled, which is the more usual meaning. It is the action
described as fixing, or making firm the accents between the lines
of type, and hanging them above the vowels, which seems inapplicable
to a type of this character, and irreconcilable with any sense that
can be extracted from the second passage. In this four things are
described ; the first, which is the cutting of the punches, and the last,
the casting of the type, are not to be mistaken ; the other two
processes, described as ouvefpeiv xd yaKKia, to string together the
pieces of brass (which cannot therefore be steel punches, as Didot
says, but may conceivably be copper, which seems to have been
the usual metal at that period) ; and ko^ ev orf^eiv, to prick, or inlet
them so as to make one piece of them, must be intermediate. The
second probably describes the sinking of the punch into the bar
of softer metal by striking, Katf ev, so that the two pieces of the
punch make one matrix; and Didot's explanation will be the right
one, that the punch-cutter did something of this kind. He cut
a punch for a letter, say a, of some two-thirds the height of the
body which it was intended the type should have, and made a small
hole in one side of it. He also cut an accent of, say, half the height
of the letter, with a pin on one side corresponding to the hole in
the letter-punch. The accents could thus be used with any letter
by simply fixing them on to the punch, and the matrix could be
struck from the combined punch, while the unaccented letters could
be provided for by a second punch, of the full height, or possibly
by a simple adjustment of the type-mould. In this way (though in
fact a good many accents of each kind were cut for the sake of
variety) the necessary accents and breathings would be less than
a dozen, and the work of the punch-cutter much lightened, though
of course the size of the case for the compositor would be much
larger than in the Aldine types, and his work of adjustment simpler.
But I confess that I cannot see how the words xd ya\Kla ouveJpeiv
can refer to this process Under any interpretation ; nor how the
two passages can be reconciled with one another; because a type
cast in this way would have a solid body, and how then could the
accents be said to be fixed between the straight lines of type and
hang over the vowels ? Having attempted to state the problems
and difficulties raised by this poem, I am compelled to leave them
124
unsolved: it must not be forgotten that the exigencies of metre
may have hampered Mousouros greatly in an effort to be clear and
precise, and that he had to find Greek words sufficiently dignified
for verse, to describe technical processes which it would have been
difficult to make plain in prose — processes, too, which are unknown
to the modern type-founder, and can only be guessed at by us.
If we turn to the books themselves, we find confirmation of Description
Didot's explanation of the process in the uniformity of the accented °^*^^*yP^-
letters ; the same accent is always found on the same letter, though
there is a considerable number of different accents. For instance,
in the facsimile on plate XXI, the circumflex accents of wac (line 2),
aC and kh (line 3), too and C (line 4), are all different ; so also with
the acute accents of 6, v, hv in line i, and with the grave accents of
a, fdp, TO, also in line i.
The number of variant letters (except £) is few, however, in
comparison with other late founts, and the iota subscript is not
used. A comparison of the type, and of the few spaces which have
worked up shows that the letters are set low on the face, but not
to such a degree as in the Aldine Laskaris. The interlinear space
being four millimetres, three of these belong to the lower, and only
one to the upper line ; a few tails, such as those of x and p, fall
below this limit, and a large number of letters, especially abbreviations,
rise above it ; these are probably kerned in both instances. The
kerns are sometimes extremely complicated ; thus in lines 16 and
17, where Kai is immediately above enl, the tail of the i in kqI projects
so as almost to touch the breathing on e of erri, while the accent '
of etii, which comes just beyond the end of koi, runs much higher
up. The dAX and hm at the end of lines 26, 27 is a similar example.
There is a large number of long and intricate abbreviations, some
of them most unusual, as jwaAAov, jwdTtov, fA^pS>, uirap, y'vpiejai, Travxa,
ndvTcov, AofO, nepl, rpdtperai, exei, SHjwaivei, dvri toO (as in the Souidas)
and oiov, of which there are two different ones; besides commoner
forms like eorl. The number of three and two-letter ligatures must
be very large, probably two hundred or more.
With this remarkable type of a noteworthy printer we take
leave of the Greek books of the fifteenth century. It only remains
to say a few words about the Graeco-Latin types found in books
of Italian origin from 1476 to 1500; to mention briefly the first
125
books printed in Greek in other countries ; and to give a very
short sketch in conclusion of the later history of Greek types.
VII.
Latin books The mass of Latin books which contain Greek words and
words^'^^^'' sentences, printed from 1476 to the end of the century, is so
1476-1500. enormous that to make a complete examination of them would
require half a lifetime, and the harvest would be very small, on
account of the great sameness in the Greek types which run
through them, and their almost uniform want of originality. I propose
therefore, in dealing with the Italian books, to confine myself to
a few specimen founts, and to treat them as briefly as possible,
taking the Venetian books first, and afterwards those of the other
Italian towns.
The larger types in the later Venetian books are all modelled
on Jensen's or Wendelin's founts, but as time goes on depart more
and more from their exemplars. The letters used by Jacobus
of Fivizzano in 1477, in the Paradoxa of Cicero, seem to be based
on Wendelin, but are not very well cut. Those of Andreas de
Venetian
larger types.
bropauca fumgfi:8iadfcrpfi.2i,iaTi T
m|/b> eu|icoT.> . o0*|? KQtt t-TrotyM 'e-rromiiox.* • 'vcldkL
Vtlojc •TjxtAouojc on/MOXJ YjoLcjsxuiiMoxj . oBiy KO*
OfUHpoc M.v0o|i €-7ran»Hff04fTic o^oriJoc 9^oio ."'ots
^Ofr^CerUC fS^pOL^^OLirObpciANYcij ToTe O fUW^WJI
'aroTcMfl'lo h t^oto^^iT^co Toq, HTiq TpETtslou c(<; O IXiKpiy l^r too EpEpyH'
7lKw iraprtK* "c^c '^pHfoU.g^cJ otxTiKcp. oioif q[;pg(pa3 ESp«>^
5>K Eq^pOfJUJUldU.TpE^a) T^^paJUlloa Tpe'zSCJO TETpOCUXJlaU. (TE/
owjucjooTou To (SpEXco GEC)pX)Xa fts'SpEHioa ^^ ^6 & TE'ro^'otl
Ik Tov tetuttTou rpiTou TpoTr§ lov 7r7 c| 5 $. tjS'S irpo tow
T EvpE^-H a l^r 7co TpiToa.TpE'ireTou rh t «[<; o^.Koi t^ a ottro'
fta AAeTcxi. h otTtofia AAojlie'jjou tov t y juetou 7o S . otoji
yEyEAccq;ouyiyEAoc L.haf>j.utIi(>ato
KA.Y. H.KT.610 ji Agyw' TtA ucl .KT. ut df CO . ipUco
SKca. TfJ8>{«.TiKTO . Kou ett'i currcpartttfio Biinia
TDv lieAAovToc;. to .f ; e V^ turo,f . habet ut dicam
610V Ae^co . 7tA£^«. Kou ETTi implicabo:K i prateri /
Tov TTotJJ ocKEiJLiEVotr TO • X • topcr fcflo .jf. uc dixi .
oiop AeAfl^oc . TTt-ff A5X« • ftnplicui .
yplTtt 2^ot -ioij . 21.H 9 h.t; Terria p.2l.ucl.0.ucl t
piovotS^ca ttAhOco ocpiTttKol uc cano abundo feftio *
I'm toOheAAovtoct TO . [s, 19] ta)[io,
ir, 12, 12].
Stops, etc.
Grave accent [i, 7, etc.]
colon [12] period [17].
comma [5]
194
•ar Avuov (Tinvfxfitti
varcsf«<^4.w«ft^eMH ^
Al/JtflQ if WJ'fe.W AflTArH
DcPaftorc'K
Lupo«
p Aftornouellnm
Lupicatuluin
inueniens K
accipknscum
canibus nutrieb^.poftqua
autem au Aus e fl;*fiquando
lupus pecudem
rapait*cam
canibus 8CJp(e
infedlabatar.ac canibus
nonuakntibus
aliquando copraehendere
lupuffl*8Cpropterea
teuertentibus
illefequebatur*
donee uti^hunc
coprachendens tanqaam
lupus partiapet
uenatiDn(s.po(tea
teuertebamr.ii autem non
lupus extra rapuiftet
pecudeni'ipfe
dam interficiens Jimul
XVIII. REGGIO, D. BERTOCHUS, I497.
195
C C 2
PLATE XIX.
Phalaris and Aesop, Venice, 1498.
[The whole fount as it occurs in these two books. Sorts, not combinations, are shown;
but the combinations occurring on the page here reproduced are added in parenthesis
after the sorts of each letter.]
Capitals.
24 letters.
Lower-case.
a[i. 1, ii. i] a ai[i. 2, ii. 6] aiq, aXA
av[2] au[i3] — (a[i. 2, ii. 8] d[i.
II, ii. 15] a[i. 4, ii. is] a[i7]
at [6] aii[3] div[i] av [25] dv[i]
dv[s] civ [16] au[i3] a0[i4]).
p [i. I, ii. 3, iii].
r[i. 25, ii] ra[8] rap rap[i- i, "• 13]
rr re [31] rei[i9] rev r»[6]
ri[25] fv rotp] rp r^^bg] r"
—0-6 [4] re [26] rea?] rfi[6]
ri[25]).
b[i. I, ii] ?)a[7] feai bag be [11] be
bei bH[i4] bi[3] bo [3] bp hv
bo>— (be[3] be [2] btW bl[2]
bt [11]).
€ [i. I, ii. 2] €1 [i. 8, 11. 10, iii, 16, iv. 17]
ev em [2] ep ecn emiU] eu[27]
— (€[i. I, ii. 14] e' [26] e [i. 7, ii- 13]
I [7] I [28] €[29] el[i. 16, ii. 17]
ei[i. I, ii. 18] eUi. 6, ii. 8, iii. 9]
et[2i] enl[2] ee[27] €t)[28]).
^[i. 10, ii. 16].
H[i. 2, ii. 4] Hv[7] — (H[i. I, ii. 6]
iH[io] H[i. 3, "-4] fi[5] H[9]
fi[9] '^[14] fiv[7]).
0[2].
i[i. 3, ii. 4]— ('[i- 2. "• 6, 1". 8] U22]
i[i. 10, ii. 14] I [6]).
K[i. 2, ii. 3] Kal[i. 3, ii. 7] Kara.
A[i. I, ii] M[i. 16, ii].
jw[i] iwev[i. I, ii. 18, iii] — (Mev[2]
|wev[i]).
v[i. I, ii. 3, iii, iv] vCv.
£[i. 9, ii. 17].
o[i. i,ii. i] CD [4] — (6[i] 6 [9] o[io]
6[i2] ot[4] otj23] o?[i6] oi[9]
OU[20] O0[4] 0t[22] ou[6] 0C[24].
n[i. 2, ii. 6] na[2] we [5] nei[i9]
TTH TTi TrA[2] TO [i. 3, ii] mro np[i]
npo[i3] Ttr nto[i. 9, ii. 27] — (tt [i]
nd([4] TTci[3] net [19] nel[2o]
TTeT[9] •jt6[7] np6[23] np6[5]
TTU>[27]).
p[i. I, ii. 2] pi [12].
cs[i. 26, ii. 10] c[i. i, ii. 3, "i- 5, iv]
(3a[i] 6av[i6] oe[4] oei ch[22]
00 [6] (39ai ci[23] 00 [21] csn cc
OT OTa oral c5Te[8] orei 6th[25]
STL GT0[3] OTU GTOi OU [7] OX [3]
oxp 00) [6] — (o[[i8] 06 [5] cnj[75.
T[i. 4, ii] Ta[i. lo, ii. 13] Tai[3]
Te[l2] Tei TH THV[2] THQ THQ [4]
Ti[6] To[i. 4, ii. 8] T0[g] TOO[8]
Tp[22] TT T(0[ll] T(S[i. 15, ii. 25]
TCOV — (Ta[l3] TO [15] TO [3]
197
fd[29] Te[8] THV[2] tl[9] Tl[6]
TO [8]).
u [i. I, ii. 6, lii. 12, iv] vi uv [2] — (u [8]
u[i6] u[i. 4, ii, 19] II [i. 4, ii. 20]
C[26] Cv[ii]).
(p[i, 4, ii. 13].
X xa[2i] Xai Xe[2o] X^iLvl X"[i]
X9 XI [12] XV X0[8] Xp[i- i^ii]
Xuv x«>[i]— (Xi["] xUu] xoM
Xa)[i]).
o)[i. 2, ii, 14, iii, iv]— ((b[24] a)[3]
a)[i, 19, ii. 12] &[i. i, ii. 5]).
Stops, etc.
Comma [i] colon or period [2]
mark, sloping hyphen [6]
hyphen. »
question
straight
Contractions, etc.
(a) Letters superposed : a = a, ra or axa.
at (two forms) = ai, ov = ov. =
or OQ. T over ou = ToO, over v = via,
over a = ara or to, etc., to = (o,
(6) Signs : ai aq aq (with preceding acute)
ag eiv eTv elq ev ev eg (two forms)
Hv Hv Hq fig oiq olg 6v ouq oug'ouq
(with preceding acute) Tai cov (over
jw = A^ou) &v cog (Im; (two forms).
Accents, etc.
Number of simple sorts given above : —
Capitals [24] lower-case [185] contrac-
tions [31] accents, etc, [17], Total 257.
Number of simple sorts and combinations
of all kinds (lower-case) noted in Phalaris
and Aesop: — a [80] ^[3] r[4l]
h[55] e[85] ^[2] h [39] 0(5]
I [25] k[22] a [19] iu[29] V[29]
£[2] 0[39] tt[58] p[29] o[97]
t[85] u[38] 9 [7] XUS] V[i]
CO [36]. Total 880.
Specimen letters (e ju t), giving all combin-
ations found : — e [i, ii] e [i, ii] e
e[i, ii] e[i, ii] e l[i-iii] l[i, ii]
e 'I 'e ei[i-iii] ei[i-iii] ei[i, ii]
et[i-iii] eUi, ii] el[i-vi] eiL[i, ii]
el[i-iii] ei[i, ii] ei[i, ii] el eiaq
ev ev ev ev ev ev ev ev eni eni
em ep ep ep ep[i, ii] ecsn eon
eort eoTi eTai[i, ii] eu eu eu eu
eu eO [i, ii] eu eu eu eu eu eucov
ecov ecoQ.
ju jwa (or jwara) jwai [i, ii] jucic iweiC
jwev[i-iii] iuev[i, ii] jwev[i-iii]
jwevai luevaq juevov [i, ii] ^evoc, [i, ii]
JWeVOUC iUHV JWHV JWHq juov juog jwou
jwoug jwcov.
T[i, ii] T Ta[i-iii] Ta[i, ii] Tc<[i,
ii] Ta[i, ii] Ta[i-iii] to Ta[i, ii]
TO Tai[i, ii] Ta( Ta'i[i, ii] xaq re
Te re f e xei rei rei [i, ii] xeq
TH TH TH TH [i-iii] T H THV [i, ii]
THV THV[i, ii] THC THg [i, ii] tl t[
tI Tl[i, ii] Tiov TO [i-iii] T6[i, ii]
TO TO TO [i, ii] TO TO TOTC TOV TOV
Toq Tou [i, ii] touq Tp tt too too toj
T(o [i-iii] toov Toiv [i, ii] tcoc tuk-
198
ofbocj)
jUHvoq AeKejwppiou evvdxi;!. Further we learn from the preface of Chal-
kondulas that he edited the text for this edition. His share in the
work and that of the Nerli brothers are clear ; what remains doubtful
is the position of Damilas in regard to it. For this we gain no help
from the two later books, which only name Giunta, the publisher;
' impensis et cura Phylippi de zunta Florentini,' in 1497 ; ' impensa
Philippi luntae bibliopolae' in 1500.
There can be no doubt that the Greek type is the same in all
66
three books. Are we to suppose then that Damilas was the printer
of all these ; that he remained in Florence as a printer for thirteen
years, without producing anything for nine years consecutively ; or
was he the printer of the Homer, not of the Giunta books ? It has
always been taken for granted that he was the actual printer of the
Homer, though the fact that not he, but Paravisinus, printed the
Laskaris of 1476, might have suggested a doubt. We may seek
a solution of the problem in an examination of the Roman types
used in the books under discussion. In the Homer two such types
are used; an english roman for the dedication to Piero de' Medici,
and a larger roman for the signatures. In the Zenobios the Latin
portion is in gothic ; the Orpheus has a colophon in the english roman
of the Homer, issued twelve years before.
Looking through Florentine books in search of these types used
elsewhere we are not long in finding a considerable number, though
it is more difficult to come across any in which either date or the
name of the printer is given. Such are however to be found. To
begin at the wrong end — the book entitled Scrutinium Consiliorum,
by Agostino da Novi, an Augustinian canon at Padova, dated
April 25, 1500, is printed in the two roman types of the Homer
(the smaller, as already mentioned, being used in the Orpheus of
September, 1 500) ; and the printer's name is there given as
' Bartholomaeus pres. Florentinus.' Three years earlier the Logic
of Savonarola, printed ' per Bartholomaeum de Libris,' is set up in
the same smaller roman in combination with the gothic found in
the Zenobios of the same year. The Florentine Histories of Bruni
and Poggio are in the larger roman, ' impresso per Bartholomeo p.
fiorentino,' and dated 1492. Lastly, in 1487 editions of the Corbaccio
and the Epistola a Pino de' Rossi, both by Boccaccio, were printed
by ' B. di Francesco Fiorentino ' in the same type as the books of
1492, that is in the type used for the signatures in the Homer
of 1488 ; and in the same year the smaller roman is used in
a ' Lamento di Costantinopoli,' which has no printer's name.
Now since there can be no reasonable doubt that these variations
of name always refer to a single man, this result has been attained ;
that all three types we are in search of were used by a single printer
at intervals ranging from 1487 to 1500, the exact period covered by
the Greek books. The wide intervals between the signed books, in
6^ K 2
the case of those in Latin and Italian, is apparent only, not real;
for every book with name or date there are at least twenty without
one or the other, and a continuous series of at least two hundred
books may be drawn up, reaching from 1482 to 1500, and including
very various branches of literature. Bartolommeo di Libri was in
fact, if the evidence may be trusted, one of the most prolific and
most reticent of all early printers. As the intermittency of the
Share of Greek books is a strong argument in favour of a single printer whose
the work.'" Staple productions were in other languages, and the fact that the
larger type in the Homer is used only for the signatures points
conclusively to its not being a loan (for the smaller fount would
have answered the purpose equally well), it is necessary to discover
the exact meaning of the words tt6v({> Kai beEioTHTi as applied to the
share of Demetrios Damilas in the Homer. To me it seems clear
that they refer primarily if not solely to his possession of the matrices
or punches, and to his labour and skill (novog kqI beEioTH?) in recasting
the type and making the new sorts used in the Homer and its
successors. That the type is recast, though on the same body, may
be proved by tracing a word in one book and placing it on the same
word in the other; for instance, the word klikAco^ occupies just one
millimetre more in the Homer than it does in the Laskaris. If the
same process be repeated with the Aesop, the identity of the type
in the Laskaris and the Aesop is seen at once. But Damilas may
well have taken a more active share in the actual composition of
the work. The correctness of the text in the Homer implies a pro-
portionate amount of experience and skill in the setting-up, and we
know that, like Paravisinus, Libri, though an excellent printer, can
have had no previous experience in Greek printing, unless he learnt
his business under Bonus Accursius at Milan. This is quite possible,
as far as dates go, because Libri's first dated book appeared eleven
months after the Psalter of Dec. 1481, but the character of his types
points to a connexion of Libri with Naples, and he was more probably
the companion of his fellow-citizen, Francesco di Dino, who after
being for some years a printer at Naples, returned to his native
place in 1481, or at the end of 1480* So that it is quite open to those
who consider the words of the Homer colophon respecting Damilas
to be insufficiently accounted for by his connexion with the type,
joined with Libri's habitual reticence, to contend that he must also
68
be taken to have exercised a direct supervision over the printing of
the Homer. But it is as certain as anything inferred from indirect
evidence can be that the Homer was produced by the press of
Bartolommeo di Libri, and there is nothing to show that any one else
except Giunta had any part in the printing of the other three books.
The points in which the type, as recast for the Homer, differs The type i
from the original fount of 1476 may be briefly summarised. The J^^g *"*^
iota subscript reappears, and the second sort of e, found as e in the
Laskaris, and without the breathing in the Aesop, is used, also without
a breathing, in 1488. The Homer also agrees with the Crastonus
and Aesop in the rejection of the angular a sorts used in the Laskaris.
A considerable number of variant forms of the same letter, especially
those adapted to fit closely on to the succeeding letter, have vanished
along with that practice ; on the other hand, accented sorts lacking
in 1476, such as e, i, have been added, together with double letters,
as av, Hv, Cv; the at may have existed in the first fount, as av (in
the form aC) is used in the Aesop. The number of consonants with
an apostrophe is increased by a second h', tt', and others ; and several
accented capitals have been added to the limited number existing in
1476. These and the other additions fall, taken roughly, into two
groups, the one consisting of those which harmonise with the rest
of the type, and the other of those which are more or less out of
keeping with it in point of size or character. Of the former, the
sorts which may be called new, and are not (like hv, vl, Cv) merely
supplementary varieties of previously existing combinations, are a
certain number of tied letters, such as dAA and aAA (plate HI, line 8),
^ev (line 2), rp (line 2 ; this is undercut on the left side for the insertion
of short letters), and a series of ou-forms. Of the second class the
most striking are a new semi-capital k (line 3), much larger than the
older one of the same kind, er (line 8), ct, er (line 30), and eii (line 15),
eij (line 28). A few other points may be noted before passing on;
a second p (line 7) has been added, and is used side by side with
the older one; the open 9 seems to be absent from the Homer, but
is found again in the later books. A new £ (line 6) has almost ousted
the original letter, which nevertheless occurs occasionally ; it is quite
common in the Zenobios and Orpheus. The 6 (line 4) is the same
as that in the Laskaris, though the Crastonus and Aesop contain one
of the right size. The i first appears in the Homer ; the concurrent
69
use of h' as one sort (line i) and as two (line 9) is also noticeable.
In the analysis of the sorts used in plate III I have marked with
an asterisk those occurring on that page which appear in the Homer
for the first time.
Other Of the three remaining books of the first Florentine press the
bookT "^ undated Chrusoloras, an octavo, may be taken first. It contains
the Greek text only. It is more in agreement with the Giunta group
than with the Homer, and may perhaps be placed about 1496. Com-
pared with the Homer, the principal differences are the introduction
of a third, rather squat £, and the revival of some of the older letters
largely or altogether disused in 1488. Thus the open 9 is common,
and the oldest E is here more often used than the two later sorts.
The large new eu-forms have almost vanished in favour of the originals,
eii alone being found, and that only towards the end of the book.
The Zenobios and Orpheus agree substantially with the Chrusoloras ;
both are quartos, with Latin colophons ; the first has also a Latin
preface addressed to Giorgio Dati by Benedetto Ricardini. The
Orpheus is distinguished by a woodcut headpiece and initial printed
in red, doubtless a tribute to the influence of Kallierges.
Revival of Meanwhile at Milano the printing of Greek had been once more
at Milanof^ taken up. Demetrios Chalkondulas had, as has been mentioned,
returned thither from Florence in 1492, and he seems to have given
an immediate impulse to both printers and scholars. His arrival
happened at a fortunate time, for the workshop of the brothers Honate
had been broken up in 1490, when they finally moved to Pa via, and
the punches of the second fount used by Bonus Accursius no doubt
came to light at the same time. They seem to have been bought
by Ulrich Scinzenzeler, then the leading, though not the senior
printer in the city. He was naturally not disinclined to make use
of his acquisition, and fell in readily with the suggestions of the
newly appointed professor. A new type was cast from the punches,
Isokrates, reproducing the older fount with few variations. Three citizens,
secretaries to the Duke, were found willing to share the cost, and
in January, 1493, which may be 1494 by modern reckoning, the
orations of Isokrates, a small folio, were issued with the following
colophon : * EreAeiajGH ouv Gecji to wapov pipxfov ' looKparouc ev MebioAavcp,
?)iop9(o9ev juev Cno AHjUHTpfou toC XaAKOvbuAou, TUTra)9ev be koI ouvTe9ev uno
'EppfKOU ToC fep/uavoO Kal ZepasTiavoC toO Ik TTovTpejwouAou" to h' avdAcoiwa
70
1493
nenoiHKOCiv of toC Xa/wnpoTOTOu Hrejwovoq MebioAdvou rpajWjwaTeTq BapeoAoiwatoq
iKuaooq, BiKevTiOQ 'MfnpavToq, BapGoAojuatoQ 'Po^covoq, erei Ttp dno thq XpisroC
revvHoeojq x^^^OGTtp reTpaKOOiooTcp evevHKOortj) Tptrcp jWHvoq 'lavouap(ou eiKOOTiJ
TerdpTi;!. The device of Ulrich Scinzenzeler follows : but the printers
named in the colophon are Henry the German and Sebastiano of
Pontremolo. Henry the German, or Heinrich Scinzenzeler, was prob-
ably a brother of Ulrich, and worked in his office. Four Latin books
containing his name as printer were issued at different dates from
1488 to 1496, and the types used are (probably in every case) those
of Ulrich. Of Sebastiano, nothing is known ; it is possible that he
had experience in Greek printing, gained from the older Milanese
press, or from one of those afterwards established ; some one accus-
tomed to such work must have been employed in the composition
of the Isokrates, the skilful hand being apparent throughout; and
his share in the work may have entitled him to an association with
the master-printer's brother in the colophon. The venture seems to
have been unsuccessful, if it be not that an unusually large number
of copies were printed ; for M. Legrand, in his Bibliographie Hell^nique
(I. p. 17), has drawn attention to the fact that the Bibliotheque
Nationale contains a copy which was reissued in 1535 with a new
title-page and colophon in Greek and Latin ; the last leaf of the
first quire and the first leaf of the last quire being also reprinted.
The Latin title runs thus : ' Isocratis Orationes XXL ahas a Demetrio
Calcondylo primum Mediolani correctae, et editae : nunc autem iterum
accurate recognitae et impressae emittuntur. Venetiis M.D.XXXV.'
To make it worth while for a publisher to do this, the 'remainder'
bought by him must have been of considerable size. But in face of
the completer Aldine edition of the preceding year, even a lying title-
page did not succeed in selling the book; and the copy at Paris
seems to be the only one in this state known to exist.
One volume besides the Isokrates was printed with the same The three
types. It consists of three independent parts, usually bound together, fj^^^j^g^'*^^^
but shown by the separate errata to each, and the signatures, which
are neither continuous nor supplementary, not to be connected other-
wise than by identity of format and subject, and approximately
identical date of production. They are three grammatical treatises;
the first, containing signatures a to 9, is by Chalkopdulas himself;
it is entitled epcoTHjwaTO cuvonriKd twv okto) toO Aorou juepcov /uexdi tivo)v
71
these books.
XpHcfjuajv Kovovcov. The title of the second (sig. a to i) is as follows :
ToO C09toTaT0u kqI AoriojTOTOu KupoO MavouHA toO MosxonoijAou biopGcoGevroiv
epajTHjwaTcov. The third, irepl biaAcKTCOv j&v irapd Kop£v9ou TrapcKpAHGeiodiv,
is anonymous, and has signatures a to r- None of the three has any
date or imprint, and it is a question whether they precede or follow
the Isokrates. The absence of capitals tells in favour of the priority
of the latter, and though Chalkondulas might have been expected to
begin by printing his own work, which was doubtless used by him
in his lectures as a text-book, the type certainly seems more worn
in the grammar volume. The errata appended to the grammars
also point in the same direction. The difference in the signatures,
which are printed in roman capitals in the Isokrates, in Greek
lower-case in the grammars, does not help to solve the problem.
Type of The type of these books is that of 1480-81, which is larger than
great primer, recast on a two-line brevier body, and consequently
modified in several ways. These may be classified thus : (a) adjust-
ment without alteration; (d) substitution of new forms for old;
{c) disuse of older sorts; (cT) addition of new sorts. The first of
these processes was carried out partly by trimming the punches,
partly by kerning the type in the mould. The lower-case letters
which have been trimmed in such a way that part has been cut
off are r, h, 0, p (usually), high t, "] /"« ['• _i.
ii, iii] jud jwa i^ai[i, ii] A'a([i, ii]
MOi juat ^lav[i, ii] mQV jwdv
)wap[i, ii] /ud(p[i5] |uac jwac
iue[i. i6, ii. 17, iii] i^e[i, ii] /we
Me jueraq i<\ei[i, ii] fxel ^td ^ei /ueA
/ueA /wev[i. 12, ii-vi] jwev[i-iv]
iuev juep Aiep[i, ii] jmh [i-iii] /km
/UH/MH[i, ii] MHv[2o] jMHv jwi[i, ii]
/uf[i, ii] jwl jui juv juo[i. 3, ii, 8]
jwo jwo [i, ii] ju6 /u" [i, ii] /wu /uO
A/xfe^DO(;»ou(ftf;f?t5if; V'itrs?eiTpiH'nu<; voTfc'JW^i 55eiKoi« <£ Aojo^et(?ioi(;
52t K9M^H0'/*-fi>l«.«rf tp-i A • oun 01 o9-flc toe t^w ^K6iT •wtw'jitjyt itfl^uMlt.-
|SiV OVO •ZJO/W,* Act AM V -OTM,* d OZpct M€ laU &I • 0?(A* TTO Q S ou • « a'h Kl 531 U*
loo oBUi