Signatures are the sign or mark which Printers place beneath certain pages for the convenience of the Binder, and to distinguish the sequence of the sections (sometimes styled quires or gatherings) which they print.

The doctrine of development and the survival of the fittest has thrown floods of light upon many dark places in the natural history of plants and animals, and I believe the same doctrine may be made equally useful in the study of bibliography. The halfpenny newspaper of to-day, with its rotten material and blurred impression, seems, at first, to have nothing in common with the beautiful vellum manuscripts of the middle ages; and yet the one is the true descendant of the other, and it was only by slow degrees that the printer’s progeny parted with their family likeness to the aristocratic products of the professional scribe.

The survival of the fittest is plainly shown in the development of signatures. The simple consecutive number which is used by modern printers to indicate the sequence of the sheets, is the true survivor of various ways of signing books from the 9th to the 19th century. It has been customary among bibliographers, especially the older writers, to consider the idea of signing sheets at all as the invention of printers. M. de Marolles says that signatures were first used by the printer Jean de Cologne in 1474; and that among all the ancient MSS. which had passed through his hands, he had not found one with signatures. He declares that many other bibliographers, even more conversant with MSS. than he was, were of the same belief.⁠[8]⁠

De la Serna corrects Marolles, but only as to date, which he takes back to 1472, in which year he says that J. Koelhoff, of Cologne, first used signatures.

Meerman runs wild in his Origines Typographicæ; believes in Corsellis, the first Oxford printer, and that he was earliest in using signatures, referring to the well-known Expositio of 1468-(1478) as a proof.

Fischer, in his Beschreibung typographischen Seltenheiten, devotes forty-two pages to the question, coming to the same conclusion as De la Serna, and, like him, was quite unacquainted with the use of signatures before the invention of printing.

In England, Conyers Middleton, T. Hartwell Horne, and others discuss at length their origin and first appearance. Middleton, indeed, describes a copy of Baldi lectura super Codic., in the University Library of Cambridge (of which he was chief), in which there are no signatures to the first half of the book, but regular signatures through the second half. Although the worthy librarian was sadly out in considering this volume as showing the earliest use of signatures, the fact that we here find the first instance of printers removing signatures to a place close up to the text is a curious instance of the transition from one custom to another, and adds great bibliographical interest to the book. Its teachings will be still further considered later on.

The chief use of signatures was and is for the binder. Binding is certainly as old as books. Signatures are certainly as old as binders. It is conceivable that the early monastic scribe, who made his own parchment, concocted his own writing-ink, copied leisurely, with his own hand, the Bible or Psalter, and, lastly, bound them propriâ manu, might complete his work without wanting any signatures to help him; or, at any rate, might be satisfied with placing a catchword at the end of each section as a guide to their sequence. But when the manufacture of books passed from the monk’s scriptorium into the hands of trade guilds, and the increased demand for books caused a great subdivision of labour; and when, instead of one, a manuscript would pass through a dozen workmen’s hands before completion,—then signatures became a necessity, as much for the scribe as for the binder, as necessary for the collation of the early MS. as for the steam-printed novel of to-day.

Let us then begin with the professional scribe, and consider the use of signatures to him. In commencing a book he had first of all to calculate how many pieces of parchment or vellum he would require, which he would then get from the “parchmenier,” who made it his business to cure, dress, sort according to quality, and cut up skins to size ready for his customers. The next process was to rule the down margin lines and the cross lines, between which the text was written, allowing for two pages on each side of the vellum sheet, and leaving space in the middle, between the pages, for the folding. He would then determine how many pieces should go to a section, and counting out his vellum sheets in fours or fives, he would sign each piece at the extreme bottom edge of the right-hand corner. If in quaternions, his signatures were a j for the first, a ij for the second, a iij for the third, and a iiij for the fourth sheet, and so on with every four sheets, through the alphabet. Everything being now ready, the scribe would take the piece marked a j, and having written that page and its verso, would lay it aside, and do the same with a ij; and not until he had passed the middle of the section would he return to and complete the earlier written pieces, the signatures on which would guide him as to sequence. Thus proceeding from signature to signature, he would finish his manuscript, and hand it over to the binder, whose first duty would be to carefully fold each piece in the centre, and then, having got all his sections in order, he would scrupulously check the sequence by the signatures before beginning to sew them on the bands.

The intention of writing the signatures at the extreme edge of the paper was that, being unimportant to the bound book, and impertinent to the text, they might disappear under the knife of the binder. All the workmanship of that period being honest and thorough, the binding was expected to last as long as the book itself, so that the possibility of a book being rebound and requiring the signatures a second time was never thought of. This position of the signatures is why so few manuscripts show them plainly; although they are still to be found, half cut away, in many books, if the student knows how to look for them.

It is by no means uncommon to find in early books, both manuscript and printed, which have all the rough edges, and have certainly never been under the binder’s plough, that the signatures have altogether or partially disappeared. This, at first, is very puzzling, but, in fact, both parchment and paper varied somewhat in size, and often when the sheets were sewn on the bands by the binder, the irregularity of the edges would be so obtrusively ugly, that the shears were used freely, and the redundancy, often including a part or the whole of a signature, disappeared; yet, to the eye, the volume appears uncut. Uncut by the binder’s plough it certainly was, for although the exact date when the plough first came into use is unknown, there is evidence that it was not used in the 15th century. In Jost Amman’s Book of Trades (Frankfurt, 1534) we have the earliest representation of a binder at work. He has a book securely fastened between two strong pieces of wood, by means of screws, and holding it between his knees, he is “ploughing” with a sharp knife through the edges. This, of course, would make the leaves perfectly even, a characteristic never, I believe, found in any “fifteener” which retains its original binding.

In a splendid copy of an early 15th-century Missal at the Mazarin Library, Paris, the scribe has adopted the unusual plan of placing all the signatures in the very back of the sheets, where they could not offend the eye; but, as the binding is loose, they can be collated with little trouble throughout the volume, each signature consisting of four sheets or eight leaves, the first four having bold signatures.

Dr. Ginsberg tells me that signatures are common in early Hebrew MSS. and printed books, although, of course, they run the reverse way.

When printing was invented, no new method of signatures was at first adopted. The Mazarin Bible, for instance, which is a large folio, was printed page by page, and signed by the pen at the foot of the first four rectos of each signature, just as if it had been a manuscript. In the Perkins copy these signatures are visible throughout. When the first printers wanted a smaller size than folio they treated their paper at first just as the scribe treated his skins, cutting it up in half sheets and printing their quarto pages one at a time. We must remember that the first printing-presses were very small in the “platen,” which is the flat surface lowered by the screw to squeeze the paper upon the type; so small, indeed, that although a whole sheet was put in the press, only half could be printed by the first, and half by a second “pull.” Moxon, who in 1693 wrote the first book on the Mechanics of Printing, gives the size of the platen in his “improved” presses as 14 inches by 9 inches, which is much smaller than a sheet of foolscap. This double pull was the source of much bad printing, and made it easier to print single pages than two pages by a double pull.

Returning to the manuscript signatures on printed books, we note that Caxton’s early books show the same treatment. The first book from his Bruges Press, The Recuyell of the Histories of Troye, a fine copy of which is in Her Majesty’s library at Windsor, is signed throughout at the very foot of the pages. So is another book from Caxton’s press, the Quatre Derennieres Choses, and a fine copy of Gulielmi de Saona Rhethorica in the Upsala library. Before me is a Cologne-printed book, Eusebius de morte sancti Jeronomi, from the press of Ulric Zell, circa 1470, and here the signatures are in MS. throughout, although many of them are half cut away in binding. Perhaps the most interesting instance is the work at Cambridge, already mentioned as quoted by Middleton, who, while noticing the printed signatures, which are not used till the work was partly printed, seems entirely to have overlooked the manuscript signatures, which are all at the foot and run through the work, the MS. signatures being correct, when those that are printed are occasionally wrong. Perhaps nothing will be found to show the transition more forcibly than this; for the printer blundered over the plan of printing the signatures, which was new to him, and had to fall back on the old system. We must here note that the printers could not, without difficulty, copy the custom of the scribes, and print their signatures at foot, because two or three types at a distance from the body of the page would certainly be broken off by the pressure; so finding the MS. signatures troublesome and often hard to read, they tried the plan of stamping them in with types by hand at the extreme edge, nearly always at foot, though sometimes at the fore-edge. This development was scarcely an improvement, and is only found in a few books from the Italian press of the years 1475-76. Then the printers, instead of hand-stamping, tried printing them at the very foot, and by the same pull of the press. This plan had no life in it, and it was then that the bright and bold idea struck a Cologne printer to ignore the ugliness and place his type signatures close up to the solid page. The custom soon spread and became general, and curious it is to notice how this slight development has given rise to numerous mistaken arguments on the so-called “invention of signatures.”

We may now safely conclude that the idea of books without signatures is a bibliographical delusion.

---

The following is a list of books in several of our public libraries, by which the reader may, if he so please, verify for himself the foregoing statements. They show the various steps in signature development.

CLASS I.

MSS. WITH WRITTEN SIGNATURES AT FOOT OF THE PAGE.

1.—Latin Theological Treatise. XIII cent. (Bod. MSS. No. 1840.)

In quaternions. The sections are all numbered at foot of last page, but have nearly all been cut away. The leaves were all numbered i to viij in each section; see especially for a plain instance section 8. This plan is also seen in Bod. MS. 1841 and others.

2.—Bible in Latin, with French illuminations. Early XIV cent. (Bod. MSS. No. 1848.)

A small thick volume, made up in quinternions. The illuminations are all deficient in the ornamental flourishes, which are cut off at foot. The sections were all numbered on the first recto, but some are gone. The first rectos of each section are signed at foot a, b, c, d, e.

3.—Wicliff’s Commentary on St. Luke. English MS., XIV cent. (Bod. MSS. No. 1913.)

The sections are all signed on the first folio, at top, in red numerals, and signatures, letters with numerals, appear throughout, at foot.

4.—Prayers, etc. Small vol., XIV cent. (Bod. MSS. No. 1851.)

Cut in binding, but some signatures left, which appear thus, the top numerals showing the sequence of the sections, and the bottom, the first four rectos of the section, the former changing, and the latter the same in every signature:—

an example of stacked signatures as described in the text

5.—Poor Caitiff. English, XIV. cent. (Bod. No. 1843.)

All the quaternions signed on first four pages at extreme bottom corner.

6.—Hannapes, N. de. Exempla Virtutum. Small volume, written about 1400. (Bod. No. 1860.)

Signatures throughout at extreme bottom corner.

7.—The Old Testament and Apocrypha. Englished by John Wicliff. Written about 1420. (Sion Coll., London.)

Signatures nearly all cut away, but plainly visible at signature g and onwards, when the following notation is adopted:—

example of signature g with varying number of horizontal lines above

8.—Psalterium. Large folio, dated 1327. (Sion Coll., London.)

This splendid XIV cent. MS. is now nearly deficient in signatures, although they are still visible at s and t.

9.—Biblia. 2 vols. large folio, dated 1360. (Guildhall Library, London.)

A beautiful MS. with a few of the signatures remaining in the second volume. All those in the first volume have been cut away in the rebinding.

10.—Albertus Magnus. Large vol., XV cent. (Bod. No. 1897.)

Signatures throughout at extreme bottom corner.

11.—The Chronicles of England. Folio, XV cent., on vellum. (Lambeth Palace.)

Sections in quaternions, signed in MS. at bottom right-hand corner.

12.—Speculum Vite Cristi. Folio, early XV cent. (Brit. Mus., Arundel, 112.)

Sections in quaternions, signed in MS. with Arabic figures, at extreme edge of bottom right-hand corner. On the first recto of each section the number of the quaternion is also given thus: ij q, iij q, iiij q, and so on, q meaning “quire” (?).

13.—Boethius de Consolatione Philosophiæ. Translated into English. Vellum, XV cent. (Brit. Mus., Harl., 2421.)

Sections in quaternions, signed in MS. on the first four rectos, at bottom right-hand corner.

14.—Oriental Manuscript. “Markemath Pettrī,” “Making wise the simple.” Arabic, vellum, XI cent. (Brit. Mus., No. 2568.)

Each quaternion is signed with letters and Arabic figures on the first four leaves.

15.— Opus Bedæ, and other pieces. Vellum MS., IX cent. (Brit. Mus., 15 B xix.)

The sections, but not the leaves, signed with the Greek alphabet.

16.—Joh. Scolastici Scala cœli. MS. on Vellum, dated 1473. (Sold at Sotheby’s, Feb. 25, 1889.)

Signed throughout, at the extreme foot of the page, and often thus, 1 d, 2 d, 3 d, 4 d, instead of d j, d ij, d iij, d iiij.

CLASS II.

MANUSCRIPT SIGNATURES IN EARLY PRINTED BOOKS.

17.—Biblia (Fust and Schœffer). Large folio. Mentz, 1462. (Bod. Show Case.)

Signed throughout in plain MS. at the extreme corner of the first five rectos in every section.

18.—Plinius Secundus. Historia Naturalis. Jenson, Venice. Large folio, 1472. (Bod. Auct., N. 1, 2.)

Signed in MS. at extreme edge all through. A copy at the University Library, Cambridge, is the same.

19.—Platea, F. de. De restitutione. Large 4to. Venice, 1472. (Bod. Douce, 147.)

MS. signatures throughout at extreme corner, although many have been cut away.

20.—Plinius Secundus. Historia Naturalis. Large folio. Rome, 1473. (Bod. Auct. Q. I 1.)

Much ploughed, but remains of MS. signatures plainly visible.

21.—Gerson, J. Super Magnificat. (Types of Fyner, Esslingen.) Folio, 1473. (Bod. Auct. VI Q. III 43.)

A few MS. signatures left.

22.—Sermones notabilis Magistri Alberti. Folio. Coloniæ, Arnoldus Therhoernen, 1474. (Cam. Univ.)

The peculiarity of the signatures here is that they are not quite at the foot, but written about half an inch from the bottom.

23.—Repertoriū Milis de Verona. Folio. Impressū ꝑ Nicolaū Gotz de Sletzstat, 1475. (Cam. Univ.)

Signed at extreme corner in MS.

24.—Panormitanus. Louvain. Folio, 1475. (Bod. Auct. III Q. V 10.)

MS. signatures at extreme foot.

25.—Gerson, Joh. Opuscula. Folio (Strasbourg?) 1475. (Bod. Auct.    .)

MS. signatures throughout at extreme bottom corner.

26.—The Recuyell of the Histories of Troye. Folio. Printed at Bruges by Caxton, c. 1475. (H.M. Lib., Windsor.)

In quinternions, with MS. signatures at foot of every first five rectos.

27.—Les Quatre Derennieres Choses. Folio. Printed at Bruges by Caxton, c. 1476.

Signed throughout at foot in MS. (Sold at Messrs. Sotheby’s a few years ago, where I examined it.)

28.—Epistola Sācti Jeronomi Presbiteri ... de Libris Salomonis. Folio. No place or date, but with the types of C. de Homberch of Cologne, c. 1475. (Cam. Univ.)

Volume II only in the library, which begins at Proverbs. In quinternions, the book beginning with signature JJ, written at the extreme bottom corner. A peculiarity worth noting, although by no means uncommon, is the use of a + upon every fifth recto, to show the middle of the section. Another peculiarity, not noticed by me in any other book, is the originality displayed by the scribe who added the signatures. They run through the alphabet of double capitals, then follow the double small letters, as usual, up to zz, ⁊ and ꝰ; but instead of then commencing the alphabet afresh, with another series, the scribe beat his brains as to what he should do next. ZZ (which was really a duplicate) did for one, and then a sign like the Greek letter ξ, thus: | ξj | | ξij | etc.,⁠[9] followed by the words “est” and “per.” His imagination now exhausted, he bethought him of the Lord’s Prayer, the use of which I should never have imagined, till the sequence of the words forced itself upon me. It is used thus, with contractions: p’ter j—p’ter ij—p’ter iij—p’ter iiij—four leaves unsigned; qui j—qui ij—qui iij—qui iiij—four leaves unsigned; and so on through—es—Jn—celis—sāficet’—nomē—tuū—adūerat—r’gnū—tuū—fiat—volūtas—tuā—sicut—and here the book ends.

29.—Aristotelis Ethica. Folio. Louvain, 1476. (W. B., from the Klemm Sale.)

Plain MS. signatures through every quaternion.

30.—Platea, F. de. De restitutionibus. 4to. Venice, 147-. (Bod. Auct., Q. inf. I 15.)

In Roman type. Signed throughout in MS. at extreme bottom corner. Many cut away.

31.—Quatuor Novissima. Folio. Cologne. P. de Olpe, n.d. (Cam. Univ.)

Signed at extreme corner in plain MS.

32.—Legenda Sanctorum. Large folio. Printed at Toulouse about 1480. (Sion Coll., London.)

Signed throughout in bold MS., beginning with a, and ending at mm.

33.—Historia Scholastica. Folio. Sine ullâ notâ. Types of Homberch, Cologne (?), no date. (Cam. Univ.)

In double columns, signed in MS. throughout at extreme foot. The leaves are also numbered in MS. in a peculiar way. The first 100 folios are numbered 1 to 100; the second lot a1 to a100; the third b1 to b101, which ends the volume. Another copy has MS. signatures also, but well up on the page.

34.—The earlier Prophets. A printed Hebrew book. Folio. Soncino, 1485. (Brit. Mus.)

Signed throughout in quaternions in Hebrew MS.

35.—Balbi de Janua Catholicon. Folio. Gutenberg. Moguntiæ, 1460. (Sold at Sotheby’s, Feb. 25, 1889.)

About half an inch having been ploughed off the foot, only portions of the signatures are here and there to be traced. In the original state a bold MS. signature, at the bottom right-hand corner, went through every section.

36.—Calderini, J. Filii Consilia. Folio. A. Rot, Rome, 1472. (Sotheby’s, Feb. 25, 1889.)

MS. signatures throughout at extreme foot.

37.—Joh. de Imola. Folio. Venice. J. de Rubeis, 1475.

MS. signatures throughout at extreme foot.

CLASS III.

MS. AND PRINTED SIGNATURES IN THE SAME BOOK.

38.—Aquinas, Thos. Summa Angelica. Per Albert Standael. Folio, 1473. (Bod. Auct. I Q. IV 30.)

Ploughed, but MS. signatures at b 1 and b 5. On c signatures stamped in by hand begin.

39.—Horatius Flaccus. Folio. Milan, 1474. (Brit. Mus. C 3, b 3.)

Up to d the signatures are all inserted by the pen, after which printed signatures are used, both being at the extreme edge. There are manuscript catchwords.

40.—Panormitanus. Folio. Venice, 1475. (Cam. Univ.) Printed by Johannes de Colonia.

Both MS. and printed signatures run on together through the volume, and their capricious character must have troubled the binder. The printed signatures are close up to the text, and the MS. signatures quite at the foot of the paper. The printed signatures have Arabic figures, thus a 2, and the MS. signatures Roman numerals, thus + ij. The two sorts run together thus:—

Note the curious custom often seen in manuscripts, where, instead of sig. z being followed by a series of capitals or duplicated small letters, certain contractions and signs are interpolated, and then the new alphabet is begun. The most common are ꝰ the contraction for “us,” ∴ an algebraical mark, and “et” or “est,” both very common words in Latin books. The abnormal use of the letter w in both, as a manuscript and printed signature, is certainly against typographical custom, which has always followed the Latin usage.

41.—Baldi, Ubert. Lectura super 123 Cod. Folio. Venice, J. de Colonia et J. Manthen de Gherez, 1474. (Sold at Sotheby’s, Feb. 25, 1889.)

Manuscript signatures A to J; K to N have normal printed signatures; O to U have signatures in MS.; while X, Y, Z, ⁊, ꝰ, aa, and so on to nn, which is the last, have printed, with occasional MS. signatures.

42.—Augustinus. Cita d’ Dio. Without place, printer, or date, but c. 1478. (Bodl. AA. c 7.)

Normal signatures throughout, but supplemented by MS. signatures, which are placed at extreme foot.

CLASS IV.

SIGNATURES STAMPED IN WITH SEPARATE TYPES.

43.—Platea, F. de. De restitutionibus Usurarum et Excōicatorum. Folio. Venice, 1473. (Bod. Auct. II Q. inf. I 40.)

The signatures to De restitutionibus are stamped in by hand with single types. They are at the extreme edge of the sheet and always close on the “point holes,” the position of which at that time was always at the extreme top and bottom of every folio, making four holes for each sheet. At signature g the plan is altered, and signatures written in by hand, in the same place, supersede the hand-stamp. Many signatures have been cut away.

44.—Horatius Flaccus. Folio. Milan, 1474 (?) (Brit. Mus. C 3, b 3.)

Here there are MS. signatures up to d, where the stamping begins at the extreme edge of the paper, many having disappeared under the binder’s knife.

45.—Seneca. Tragœdiæ. (Ferrariæ), Per Andream Gallicum. Large Folio. 1474 (?) (Cam. Univ.)

Stamped signatures at the extreme edge of the bottom corner, rugged, uneven, sometimes upside down, and plainly stamped in by hand. Signature a is all there; b is all cut away; c is omitted; d, which follows b, is all there; e is signed 2 d on the first recto, and dd on the third recto; e all cut away; f all there; g all there, and so on up to n. Upon n 3 for the first time the signature is raised up to the line beneath the text, and so continues to u, which is the last. Brunet notices the copy in the National Library, Paris, thus: “Nous y avons remarqué, à l’extrémité inférieure des feuilles des 12 premiers cahiers, et d’une partie du 13^me cahier, des signatures qui y ont été apposées après l’impression, avec des caractères d’imprimerie.”

46.—Propertius. Folio. Milan, 1476 (?) (Brit. Mus. C 19, a 9.)

Signatures plainly stamped in.

47.—Horatius Flaccus. Folio. Venice (impensis P. de Lavagnia civis Mediolanensis), 1476. (Bod. Auct. O. 2. 2.)

The signatures begin with cap. A, stamped in, and this plan is pursued through all the quaternions. The signatures are at various distances from the foot of the text, and all at extreme edge, many being cut off.

48.—Aquinas, T. de. Summa Angelica. Per Albert Standael. Folio. No place. 1473. (Bod. I Q. IV 30.)

The volume has been ploughed, but MS. signatures remain at signatures b and b 5. On signature c the signatures are stamped in by hand, many, however, being cut away.

49.—Conradus de Allemania. Concordantiæ bibliorum. Folio. Strasbourg. Mentelin. No date. (W. B.)

This volume is very peculiar in its method of signature, and shows in an interesting manner the transitional period between the manuscript and printed methods of signing. The book is a large folio, and the sections are quinternions. On the margin, at the foot of the first recto of each section, the number of the section is written boldly in black ink, thus: 1^us, 2^us, 3^us, etc. = primus, secundus, tertius, etc. Lower down, in Arabic numerals, is the number of the sheet, which runs from 1 to 5 in each section, there being five sheets to each section. Thus, the first sheet of each section has its specific place plainly marked upon it, while the succeeding four sheets have a figure only, and nothing to identify them with any particular section. There are type signatures also, but these do not assist in this respect, for they are all stamped in by hand at the extreme edge of the paper, and consist of four varieties only. Thus all the first quarter of the Concordance, occupying 12 sections, are stamped a 1, a 2, a 3, a 4, a 5, leaving five leaves unsigned; all the second quarter, from 13 to 24, are b 1, b 2, b 3, b 4, b 5; the third quarter, from 25 to 32, are all c; and the fourth, 33 to 42, all d. These stamped-in letters are not true signatures, for, with the exception of the first sheet of each section, which bears its rotation number, there is nothing to discriminate them. The type-letters are stamped very carelessly, often quite omitted, often so close to the fore-edge that there was no room for them on the sheet, varying in position with every leaf, and sometimes half an inch and sometimes two inches from the bottom line. This irregularity may account for the sheets having MS. numbers, identical with what are, or should be, the type numbers. Occasionally, at the extreme foot, is also a third repetition of the numbers in bright red ink, evidently added by the rubricator. It is difficult to perceive what particular use either the MS. or type numbers could serve in this instance. There is certainly no difficulty in concluding that the printer’s plan was useless, in case the sheets became mixed, to himself as well as the binder.

CLASS V.

SIGNATURES NOT STAMPED IN BUT PRINTED AT THE SAME TIME AS THE TEXT, A LONG DISTANCE BELOW IT, LIKE THE MS. SIGNATURES.

Every one practically acquainted with Typography will see the difficulty of printing one or two letters standing alone at a distance from the page of type; they were certain to be battered from want of support and the wonder is that any printer ever tried the plan.

51.—Catullus. Opera. 4to. Venice, 1475. (Brit. Mus. C 19, d 9. Cracherode copy.)

The third book in this volume is Propertius, and here, although they might easily escape the eye, are bold signatures, printed so near the corner that only a few have escaped the binder’s knife. They are B 1, B 2, B 3, C 1, C 3, D 2, D 3, D 4, E 1, E 2, E 3, and F 1. In the first two and last books there are no signatures remaining.

53.—Nider, J. de. De contractibus. 4to. Homborch, Cologne. No date. (Bod. 2 Q. VI 25.)

Signature a j is blank; a ij is without signature; a iij is signed in type, but at a distance of two full lines below the text. Type signatures also appear at signatures b j, and b iij. The volume has been ploughed.

THE NORMAL SIGNATURE, PRINTED IN THE LINE JUST BENEATH THE TEXT.

55.—Mamotrectus. Folio. Printed at Ergow (Switzerland), 1470. (Bod. Auct. V Q. V 41.)

A book quoted as the earliest known instance of printed signatures; but this is a mistake, for the numeration of the various columns of text by a series of letters under each column has nothing in common with signatures. The MS. signatures, if any, have been cut away.

56.—Nider, Johann. Expositio Decalogi. Folio. Lubeck. Printed by John Koelhoff, 1472. (Brit. Mus. C 14, b 2.) Also in the Bodleian.

This is a puzzling book, for it is at least two years earlier than any other book so signed. In this city, too, many works were issued with MS. signatures with a later date than this. It is dangerous to assert that a book is wrongly dated because you cannot make it fit into a bibliographical theory; but I feel inclined, from the general aspect of the book, to date it as 1482, rather than 1472.

57.—Platea, F. de. Folio. Cologne, 1474. (Brit. Mus. 1275, d 5.)

Here the signatures are printed close up to the text in the usual way.

58.—Lucan’s Epigrams. Folio. Venice, 1475. (Brit. Mus. C 16, i 9.)

Here the signatures are printed close up to the text in the normal way.

---

From these examples, it appears that (1) the scribes, who made books before the rise of printers, used signatures; that (2) the printers began by signing like the scribes, at the very foot of the pages; that (3) for some years they used both MS. and stamped signatures; which (4) gave way to stamping the signatures in by hand; then (5) by printing them in at extreme corner; and, lastly, in 1474, by placing them up close to the page, where they have remained ever since.

We have seen how every book had its signatures, and that the law was to sign the leaves of the first half of every section. We will now trace the development of signatures caused by printing more than one page at a time, and its effect upon the size notation of books.

Paper, although invented many centuries before the discovery of printing, was little used for library books, being looked down upon as inferior in every way to vellum. It was used for school-books such as Donatuses and Cato’s distichs, but not for good books. With the advent of the press this was soon changed. Even if the expense had not been too great, all the skins of Europe could not have supplied the rapacious jaws of the new giant book-maker. So it is that the use of paper for standard books is synchronous with the invention of printing. But with sheets of paper a new development arose: the scribe writing on, page by page, section by section, required his vellum already cut to size, and in this the early printers followed suit—from necessity where the pages were large, and from old custom where small; but paper, being made in moulds, differed from parchment in being always regular to certain fixed sizes; and being easily folded, a new nomenclature for the various foldings became necessary.

Until the invention of printing there appear to have been no particular names for the sizes of books. There are numerous catalogues extant of large manuscript libraries, especially those of Burgundy and France in the 14th and 15th centuries; but although the catalogue writers are minute as to certain particulars, such as the character of writing and the ornamental binding, there is never any mention of size beyond the vague “ung grand liure” for a big History of the World, or “ung petit liure” for a diminutive Horæ. With the use of paper, however, subdivision came in the natural course of events, and sheets folded in half were called folio, in half again, quarto, and in half again, octavo,—a nomenclature as precise as it was novel. Another development was printing two pages at a time, followed soon by four pages. This was done by an arrangement which allowed one half of the sheet to be printed by a first pull, and the other half by a second pull, without lifting the sheet from the press. It also required a special arrangement of the pages of type, which had to be placed head to head, in order to make them read aright when folded.

We have seen how, with folio books, the various sheets were placed inside one another like a quire of paper, and now with the printing in quarto a system of signatures arose which has puzzled bibliographers, viz.: where the first recto of a section is signed a j, the second recto blank, and the third signed a ij, with all the rest blank. Let the reader take two sheets of common note paper, and folding each, separately, in half, make a representation of quarto. Now mark on the first recto of the first sheet the signature a j, leaving three leaves unmarked, and on the first recto of the other sheet a ij, and then place the second sheet inside the first, and at once you have a quaternion, with the following series: sig. a j—none—sig. a ij—none—none—none—none—none. This was really an excellent plan: it answered all the purpose of the binder in collating, and was a natural development It did not, however, take deep root, and the old plan of signing the first four leaves prevailed long after its usefulness had vanished. The octavo size was treated in a similar manner. It is common to find 16th and 17th century books printed in the whole sheets octavo, and signed like the folios and quartos on each of the four first rectos. We may here just notice another peculiarity of signing, and that is where, although the sections have eight leaves only, they are signed upon the fifth as well as the first four rectos, leaving only three rectos unsigned. This had a slight use in telling the binder which was the central sheet of each section, and that he need look no further. It was also a distinct imitation of the scribes, who used a + mark for the same purpose, as already noticed. The Aldus family did not print smaller than 8vo, but the Elzevirs and Plantins used 16mo and 32mo abundantly, the old custom of signing still surviving, and the small alike with the large sections consisting of eight leaves, of which four were signed and four not signed.

The tendency to print small books developed another practice which ought to be understood. Say that it is decided to print a volume in 24mo; this, of course, is 24 pages on one side of the sheet and 24 pages on the other. Take now a sheet of any size, a sheet of note paper will show the working of it as well as any: folded once it is folio, making 4 pages; now fold carefully across into three, that makes 6to, with 12 pages; fold into half again, and that is 12mo, with 24 pages; and again in half is 24mo, with 48 pages. Here the back is too thick and clumsy for binding, so cut the paper into three even parts where you made the second fold, and then treating each third part as if it were a distinct sheet, and signing each on the first four rectos, you have 16 pages to each third. Thus, three complete signatures were printed on one sheet, and then cut up into thirds when bound, with the result that the 32mo book, if judged only by the visible signatures, ought to be classed as 8vo, all the signatures having four signed and four unsigned leaves.

We now see that all sizes being signed alike, the signatures cannot with early printed books be any guide as to size.

So much for signatures. There are many other peculiarities which might be noticed, but when the above are understood, other variations explain themselves.