Dates form such an important feature in colophons that this essay cannot be brought to a close without some attempt to explain the difficulties which arise in connection with them. As regards the method of expressing the year there is very little to say. Theodoric Rood (see page 61) preferred to speak of 1485 as the 297th Olympiad from the birth of Christ, being under the impression that Olympiads consisted of five years instead of four. Other printers showed great ingenuity in finding elaborate synonyms for what we are now content to express in the two words “Anno Domini,” and among other phrases employed “Olympiades Dominicae” (see page 79), but without any attempt to introduce the intervals between the Olympic Games into the Christian reckoning.

As an additional method of dating we occasionally find a reference to the year of the indiction, a method of dating by cycles of fifteen years, instituted by the Emperor Constantine in 312. To find the indictional year, 312 must be subtracted from the year of the Lord (the same results will be obtained by adding 3), and then after dividing by 15, the remainder will give the number of the year in the indiction. Thus (1488-312)/15 or (1488+3)/15 in each case leaves a remainder of six, and A.D. 1488 was thus the sixth indiction.

According to different methods of reckoning, indictions began in September or October, at Christmas or on January 1st. In colophons, I believe, they are always used in conjunction with years of the Lord reckoned from January 1st, and they have only the effect of a chronological flourish.

A much more important supplementary method of dating is that by the names of ruling popes, emperors, sovereigns, or princes, or still better by their regnal years. I have long cherished an ambition to compile a kind of “Bibliographer’s Vade-mecum,” one section of which would be devoted to exhaustive lists of the smaller as well as the greater sovereigns of Europe during the period when their names in old books are of chronological value. Here, however, it must suffice to offer lists of popes, kings of England and France, and doges of Venice, for the periods which concern us, and to use these as illustrations of the way in which such information can be brought to bear on the dating of early books.

POPES

KINGS OF ENGLAND

KINGS OF FRANCE

DOGES OF VENICE

As our first example of how these tables may be used we will take a colophon where no year of the Lord is given, and sovereigns are mentioned without their regnal years. We shall find that even the mere names may help us to a close approximate date. Our instance shall be Wendelin of Speier’s edition of the “Supplementum” of Nicolaus de Auximo, the colophon to which ends:

Sixtus IV became pope early in August, 1471, NicolÒ Tron was elected doge on November 23d of the same year, and died in July, 1473. We can thus date the book as “about 1472” with absolute confidence.

Writers of poetical colophons are naturally more inclined to use regnal dates than the year of the Lord, which it is seldom easy to get into a verse. In the “Moral Prouerbes of Cristyne” Caxton gives us the month, the day, and the regnal year, together making a precise date. This colophon runs:

The seventeenth year of Edward IV ran from 4th March, 1477, to 3d March, 1478, so that the “Moral Proverbs” were finished on February 20th of the latter year.

When a change of sovereigns occurred in the year in which a book was printed, the mere name of the earlier or the later of the two shows in which part of the year the colophon was written, and regnal dates supply the same information for years in which no change of sovereigns took place. Thus the colophon to Wynkyn de Worde’s edition of the “Vitas Patrum” ends: “Enprynted in the sayd towne of Westmynstre by me Wynken de Worde, the yere of our lorde M.cccc.lxxxxv. and the tenthe yere of our souerayne lorde Kyng Henry the seuenth.” As Henry VII’s reign began 22d August, 1485, its tenth year would cover the twelvemonth, August, 1494, to August, 1495, and we are thus told not only that the book was issued in 1495, but that it was printed before August 21st of that year.

A subsidiary date, of course, as a rule loses its usefulness when the printer explicitly mentions also the month and day on which the book was completed. It may, however, have a special value as furnishing a means of fixing the day from which the printer reckoned his year. In the fifteenth century the year could be reckoned as beginning on Christmas day, on January 1st, on March 1st, on March 25th, or at Easter. In arranging the books issued from any press in chronological order, it is of vital importance to know which reckoning the printer followed, and we may now give some examples to show how regnal years can be used to settle this.

The fourth year of Sixtus IV began on 9th August, 1474, and ended 8th August, 1475; therefore January, 1475, in his fourth year must be January, 1475, according to our modern reckoning, not January, 1476, as it would be had the year been calculated from March 25th or Easter day—two similar examples will be found in Hain’s “Repertorium Bibliographicum” under the numbers 255* and 2050*.

On the other hand, Caxton’s colophon to the “Mirrour of the World” ends:

As the twenty-first year of Edward IV ran from 4th March, 1481, to 3d March, 1482, Caxton’s 8th March, 1480, must clearly be 1480, old style, or 1481 of our reckoning, and Caxton is thus shown to have begun his year on March 25th.

So again the long colophon or epilogue to the “Cordyale” tells us that the book “was deliuered to me William Caxton by my saide noble lorde Ryuiers on the day of purificacion of our blissid lady, fallyng the tewsday the secund day of the moneth of feuerer. In the yere of our lord M. cccc. lxxviij for to be enprinted.… Whiche werke present I begann the morn after the saide Purification of our blissid Lady, whiche was the daye of Seint Blase Bisshop and Martir, And finisshed on the euen of thannunciacion of our said blissid Lady, fallyng on the Wednesday the xxiiij. daye of Marche in the xix yere of Kyng Edwarde the fourthe.”

Earlier bibliographers got very confused over this book and made absurd mistakes as to the time which Caxton took to print it. But Mr. Blades had no difficulty in showing that the different dates follow closely on each other. Caxton received the book on February 2d, began printing it on February 3d, and finished it on March 24th, all in the same year 1479. We have a double method of proving this, by the two week-days mentioned and by the regnal year, which covered the period March 4, 1479, to March 3, 1480. The only March 24th in this twelvemonth was that in 1479, and in 1479 March 24th, as Caxton says, fell on a Wednesday. In 1479, moreover, February 2d fell on a Tuesday, in 1478 on a Sunday. It is thus clear that the Tuesday, February 2, 1478, of the colophon must be an old-style date, answering to 1479 of our reckoning.

The occasional mention of the day both of the week and the month in German colophons offers us, in the absence of regnal years, almost the only proof we can obtain that German printers began their year either at Christmas or on January 1st,—I am not prepared to say which. Thus the colophon of an edition of the “De remediis utriusque fortunae” of Adrianus Carthusiensis reads:

In 1471 February 8th fell on a Friday, in 1472 on a Saturday. Therefore it is clear that in therhoernen’s reckoning January and February were the first months of the year, as they are with us.

Before inquiring as to what printers reckoned the year as beginning at Easter, we must give the following table:

EASTER DAY, 1470-1521

That Pierre Gerard and Jean DuprÉ at Abbeville reckoned the year from Easter to Easter, we get a broad hint in the colophon to their magnificent edition of Augustine’s “De Ciuitate Dei” in French. This is in two volumes, the colophon to the first of which is dated “le xxiiii jour de Nouembre l’an mil quatre cens quatre vingt et six,” while the second runs:

That the second volume of so large a work must have been printed after the first is so nearly certain that this alone might have caused us to look out for a means of making April 12, 1486, later than November 24th of the same year. The words “auant pasques” put the matter beyond doubt, for Easter in 1486 of our reckoning fell on March 26th, but in 1487 on April 15th. Clearly, therefore, the book was finished on Holy Thursday, 1487, and Easter was the date from which DuprÉ and Gerard reckoned their year.

We can obtain an equally neat proof of the French year beginning at Easter from a copy of Pierre Gringore’s “Chasteau de Labour,” in which, underneath the name of Philippe Pigouchet, appears the colophon:

This edition consists of sixty leaves and does actually contain a long interpolation not found in the first edition of 22d October, 1499, or the second, which is dated 31st December, 1499, or in yet another edition dated 31st May, 1500, all three of which have only fifty leaves instead of sixty. Thus it would appear at first sight that Pigouchet and Vostre printed Gringore’s additions in March, 1500, and omitted them again two months afterwards in May. Inasmuch, however, as the French year 1500 ran from Easter Sunday, 19th April, 1500, to Easter Sunday, 11th April, 1501, it is obvious that the only 31st March in it fell in 1501 according to our reckoning, and that the edition of 31st March, 1500, was really produced in March, 1501, and is ten months later than that of May, 1500. We thus get an orderly sequence of three unaugmented editions of fifty leaves, followed by an augmented one of sixty, and all difficulties vanish.

In Italy the year appears generally to have begun on January 1st, but in Florence on Lady day, March 25th. At Venice the legal year is known to have begun on March 1st, and most writers on Aldus have asserted positively that this was the date to which he conformed. That other Venetian printers observed January 1st as the first day of the year can be proved from the mention of Pietro Mocenigo as doge in the colophon to an edition of the “Istoria Fiorentina” of Leonardo Aretino. This ends:

As our table of Venetian doges shows, Mocenigo died on February 23, 1476, eleven days after this colophon was printed; and it is thus clear that February, 1476, meant the same to “Maestro Iacomo de Rossi” as it does to us.

That the antiquarian Aldus troubled his head about the beginning of the Venetian legal year seems a strange inconsistency. But the late Mr. R. C. Christie, who proved conclusively, in an article in “Bibliographica,” that in his later books Aldus began his year on January 1st, was yet obliged to admit that the Lascaris, which is dated “M.cccc.lxxxxiiii ultimo Februarii,” was probably finished only a few days before the Supplement, which bears date March 8, 1495, and that the Theodore Gaza of January, and the Theocritus of February, 1495, both really belong to 1496. I would suggest that in adopting March 1st as his New Year’s day in these three volumes, Aldus pleased himself with the idea that he was reckoning not “more Veneto” but “more antiquo Romano,” since (as the names of our last four months still testify) the Roman year originally began in March, and it was only the fact that after B.C. 153 the consuls entered office in January that caused our present reckoning to come into use, the sacerdotal year continuing to begin on March 1st. If Aldus, after adopting the Venetian legal year because it agreed with the earliest Roman reckoning, was convinced that he was being a little more Roman than the Romans themselves, it is easy to understand his change of practice.

It would appear, then, that the only books for which we must reckon the year as beginning later than January 1st are a few early books of Aldus (March 1st), all English books of the fifteenth and sixteenth centuries, and books printed at Florence (March 25th) and in France (Easter). I strongly suspect, moreover, that in Florentine and French editions of learned works written in Latin there would be a tendency toward January 1st, but I cannot offer any proof of this at present, though it is a question which I hope some day to work out.

As the examples quoted in our text will have abundantly shown, the days of the month are expressed either according to our present use or by the Roman notation, reckoning from the Calends, Nones, and Ides. The Calends were always the first day of the month, the Nones fell on the 5th, and the Ides on the 13th, except in March, May, July, and October, when they were each two days later. Days were counted backwards from the Nones, Ides, and Calends, both the day from which and the day to which the reckoning was made being included in the calculation. Thus March 2d was called the sixth day before the Nones (ante diem sextum Nonas Martis), and March 25th the seventh before the Calends of April (ante diem septimum Kalendas Aprilis, or a. d. vii. Kal. Apr.). July and August are sometimes called by their old names Quintilis and Sextilis.

In Germany, more especially at Strassburg, and in Strassburg more especially by an unidentified craftsman known as the “Printer of the 1483 Jordanus de Quedlinburg,” we often find books dated on such and such a day of the week before or after a festival of the church or a particular Sunday, the Sunday being indicated by quoting the first word of the introit used at high mass. Thus in 1485 the anonymous printer of the Jordanus finished a “De Proprietatibus Rerum” on S. Valentine’s day (in die Valentini, February 14th), the “Historia Scholastica of Petrus Comestor” after the feast of S. Matthias (post festum Matthie, February 24th), the “Postilla” of Guillermus on Thursday (March 9th) before the feast of S. Gregory (quarta feria ante festum Gregorii), a “Casus breues decretalium” on the day of SS. Vitus and Modestus (in die Viti et Modesti, June 15th), and Cardinal Turrecremata’s “Gloss on the Psalter” on S. Michael’s eve (in profesto Michaelis, September 28th). To another edition of the “Postilla” of Guillermus he adds the imprint:

“Reminiscere” is the beginning of the introit for the second Sunday in Lent, and as Easter in 1493 (see our table) fell on April 7th, this was March 3d, and the “Postilla” were finished on Monday, March 4th.

The colophon to a Strassburg edition of the sermons known by the title “Dormi secure” tells us that it was issued “secunda feria post Laetare” in the same year 1493. “Laetare” being the first word of the introit for the fourth Sunday in Lent, it thus appeared on Monday, March 18th, exactly a fortnight after the “Postilla” of Guillermus. So again we find Hans Schauer of Augsburg dating an edition of a “Beichtbuchlein,” or manual of confession, “am Samstag vor Invocavit in dem XCij. iar,”—on Saturday before Invocavit, 1492,—which gives the date (Invocavit marking the first Sunday in Lent and Easter in 1492 falling on April 22d), Saturday, March 10th. It is generally only the introits of the first four Sundays in Lent (Invocavit, Reminiscere, Oculi, and Laetare) and that of the first Sunday after Easter (Quasimodo) that are used in colophons in this way.

We may bring this chapter to an end by noting one or two fruitful causes of error in dating books which arise from misunderstanding the reference or meaning of the dates in their colophons. In Chapter VI it has already been noted that where an author or editor has given the date on which he finished writing, such a date has often been confused with the date of imprint. More dangerous but much rarer than such a pitfall as this is the case of the reprinted colophon (see Chapter VII), which can be detected only by experts in typography. The majority of mistakes, however, arise from very simple misreadings. In many fifteenth-century fonts of type the symbols x and v are very imperfectly distinguished, so that the five has often been mistaken for a ten. Modern eyes, again, being used to the symbols iv, ix, xl, are very apt to read the fifteenth-century iiii as iii, the viiii as viii, and the xxxx as xxx. On the other hand, as they neared the end of the century the printers not only expressed ninety-nine by ic, but also used the forms vc, iiiic, iiic, iic, to express the years ’95 to ’98; and, as has been done here for the sake of brevity, occasionally omitted the precedent Mcccc., as in the “in dem XCij. iar” of the colophon of the “Beichtbuchlein,” quoted a page or two back. They also, it may be noted, frequently expressed eighty by the reasonable symbol for fourscore, or quatre vingt—namely, iiiixx. These latter methods of writing dates, however, though they may puzzle for a moment, can hardly mislead; but in the case of books issued in the years 1470, 1480, 1490, and 1500 (more especially the last) there is one error so easily made that it has left its mark on every old catalogue of incunabula. Thus when Hermann Lichtenstein dated an edition of the “Opuscula” of S. Thomas Aquinas “anno salutis M.cccc.xc. vii Idus septembris” he encouraged any ignorant or careless cataloguer to misread the date as 1497 on the “Ides of September,” instead of 1490 on the seventh day before the “Ides of September.” The mistake may be made just as easily when words are used instead of numerals, for “anno nostre salutis millesimo quadringentesimo octogesimo quinto kalendas Iunij” is very easily read as 1485. It is, of course, equally easy to make the opposite mistake and transfer to the record of the month a number which relates to the year. As a rule, the printers, by interposing “die” or “vero” or both, or by a change of type, put their meaning beyond dispute; but sometimes they got confused themselves, and by leaving out either the last numeral of the year, or that of the day of the month, produced a puzzle which can be solved only by independent knowledge of the years during which a printer worked.