Book Excerpt



Foreword by Howard M. Nixon.
New Castle and London: Oak Knoll Press & The British Library, 1996 (but 2008). 5.75 x 9 inches. hardcover, dust jacket. 386 pages followed by 14 plates. Foreword by Howard M. Nixon. Fourth edition. This is a classic reference work about decorative and commercial English bookbinding techniques written by one of the foremost experts on bookbinding. Each chapter covers various aspects of bookbinding techniques as well as historical information. Each of the chapters describes the... READ MORE

Price: $65.00 save 50% $32.50  other currencies  Order nr. 44862



Worked Headbands

Evidence of Anglo-Saxon methods of headbanding is scanty due to the dearth of bindings of this early period, and to the fact that the comparatively light construction of the Coptic-style bindings then in use has resulted in deterioration of these vulnerable points. However, we have the excellently preserved seventh- or eighth-century Stonyhurst Gospel, the headbands of which may well have been fairly typical. In this case the leather was turned in at head and tail of the spine, as usual. The turned-over leather was stitched through with a white thread which was tied down in the middle of each section; a blue thread, which was worked with it, passed through the leather only. This manner of headbanding provided narrow bands of decoration both where modern headbands are seen and also outside on the spine, and they were the only direct connections between the leather covering and the sections.1

After the introduction of raised-thong sewing the headbands usually consisted of double thongs similar to those along the spine. The thread was usually white, but it was not uncommon for another colour to be worked in alternately. At least as early as the twelfth century it was usual for the headbands to be sewn independently of the main sewing, and they were always tied down in the middle of each section until the end of the fifteenth century when the greatly increased output of books resulting from the invention of printing made it necessary to speed up working methods. They were then tied down much less frequently. Because the boards of early bindings were flush with the leaves, it was a common practice to cut away the corners of the leaves at each end of the spine to accommodate the headbands and prevent them from sticking out too far and becoming damaged.

In the twelfth century and early part of the thirteenth the headbands were combined with a leather tab, which extended beyond the spine. This is illustrated in fig. 54 and briefly described in the chapter on covering.

From the middle of the thirteenth century until the end of the fifteenth the leather cover at the ends of the spine was often cut so that it just covered the headband. This was then sewn through from front to back, or back to front, underneath and along the length of the hidden headband. The result is a series of stitches on the spine, and another series where one would normally see the beading (fig. 55). Uncoloured linen thread was normally used for this and for the headband, which is usually single and sewn independently of the main sewing of the book.

Very interesting is the plaited headband made with strips of leather, usually tawed and stained pink. This is rare in English bindings, but is common in German examples of the late fifteenth and early sixteenth centuries. These headbands look rather crude when drawn (fig. 56), but in fact they are always very neatly executed. They are curiously satisfying in appearance, and they are the strongest headbands yet devised. In making them, two thongs were plaited round a former of rolled vellum, or the like, which was already sewn to the book with thread as in earlier types; and in addition to going round this, the thongs passed through holes made in the leather at the top of the spine where it was cut off instead of being turned in. The closely formed and set appearance of these headbands makes me suspect that the thongs were worked wet. It will be seen from what has been said that in this type, as in the previous one, the leather cover is sewn (by means of the plaited thongs) to the primary headband which had previously been sewn to the book, thus constituting a first-class connection at a vital point.

The conventional headband, sewn with coloured silk or thread, with the beading showing at the bottom, and with the turned-in leather knocked over to form a cap, came in very early in the sixteenth century and quickly became popular. I have noticed more blue and white headbands than any others, but pink and blue, pink and brown, and other combinations were not unusual. It was not, generally speaking, the practice in England in the sixteenth century to employ more than two colours, and they were not arranged in different proportions to make a more interesting pattern. Usually, they were twisted two-and-two with a simple cross-over bead in front. I have noticed that in some cases where each colour has two twists round the cord only one twist was in fact made by the headbander because double strands of thread were used, but this seems to have been a minority practice.

Early stuck-on headbands are dealt with separately, pp. 107–8.

Double headbands are rarely found in English sixteenth-century bindings, but are more common in continental work. In the seventeenth century they became more popular in England, the colour combinations became more interesting, and finer material was used. Until the beginning of the present century it was usual for the two parts of the headband to consist of two rolls, one smaller than the other, with the smaller one being placed above the larger. About sixty years ago it became more usual to place the smaller roll in front and below the larger one. About the same time it became the general practice to use a flat strip of material instead of the larger roll.

Headbands were invariably laced into the boards until pasteboard superseded wood during the early part of the sixteenth century.

In nineteenth-century ‘extra’ work, finer silks and more elaborate patterns were employed than is now the case. The headbanders were very fast at their work, and it has been said that a girl working simple headbands on average-size books could keep up with a man ‘lining up’ with hollow backs, but I think the man’s work must have included ‘banding’. Part of the beauty of worked headbands depends on even tension of the silks, and this was usually very well maintained.

From at least as early as the first ten years of the seventeenth century until well into the nineteenth nearly all headbands, both single and double, were formed on rolled paper, which replaced leather. Cane was also used as a foundation material, but comparatively rarely. If the paper for the roll was wholly pasted, which it very often was not, this was done in its flat state. The paper was then rolled up with the fingers and finally consolidated and given a good shape by being rolled under a flat board. In the case of unpasted paper one edge must have been tipped with paste or glue to prevent it becoming unrolled. According to Cowie’s Bookbinders’ Manual (1828), p. 44, these could be bought in London ready prepared, from 6d. to 2s. the hundred. According to the same authority the paper was rolled round a piece of thread or twine, but, although I have seen a great many broken headbands of this kind, I have not often found a core. The reason why rolled paper was used in preference to cord must be that it could very easily be made in many more different sizes than the limited range of cords available. Rolled paper was replaced by cord or catgut for double headbands, and for single headbands it was replaced in about 18301 by flat strips of vellum, board and other materials. This change has proved to be a wise one, for the rolled paper is inflexible and tends to break in half if the spine ‘gives’ when the book is read.

Since the fifteenth century very few binders have troubled to ‘tie down’ headbands at every section, or tried to go through the middle of the section where the tie-down is being made. The usual practice has been to put the needle down between the leaves at intervals of about 1/4 in., but 1/2-in. intervals are not uncommon, even in very expensive work produced in the nineteenth century. It seems ludicrous that so much highly skilled labour should be put into very elaborate double headbands with several colours, which were tied down with very fine thread at wide intervals, but this was a common practice about a hundred years ago. It is usually regarded as a good thing to tie down so that the thread goes under the kettle-stitch which, theoretically, will help to anchor it. Generally speaking, this has been done all through the centuries, but it is probable that in many cases, particularly in cheap retail work, it has been fortuitous rather than the result of conscientious endeavour. It would seem that no weakness has resulted when this has not been done.

Stuck-on Headbands

It is quite widely but erroneously believed that stuck-on headbands were introduced early in the nineteenth century, but in fact they were in use at least as early as the last years of the sixteenth century in German bindings—but not, I think, in English bindings until the early decades of the seventeenth.

One assumes that they were used for economic reasons, but when the method of making them is considered, the actual saving in time is questionable. Two threads of different colours, or of the same colour, were sewn in the usual manner with a cross-over beading on to a strip of vellum (the upper edge of which was sometimes bent over to give greater bulk) which was probably held in a press of some kind. This was often arranged so that there was about three-quarters of an inch on each side without sewing (fig. 57), and this was stuck to the board, usually on the outside. On other occasions the bulk of the headband was cut to the width of the book and just the part under the sewing extended across the joints (fig. 58) or was laced through the joints of some vellum bindings, but this, I think, was solely a continental practice. Sometimes the extended part consisted of a piece of rolled vellum, which was rested at the top of the strip while the headband was sewn. Mr Roger Powell has noted that some headbands were sewn double the required width for head and tail and were then cut in half to make two headbands.

The main advantage of these headbands from the binder’s point of view would seem to be that they could be made at odd moments or perhaps at slack times when there were no books on the premises. They were not much used in England, certainly very little after the seventeenth century, but they were quite popular among German binders until the nineteenth century, and were sometimes used in bindings of fair quality.

Hand-made stuck-on headbands of a different type did become popular in England during the early decades of the nineteenth centurn when the greatly increased output of books created production difficulties. No doubt, other factors were also involved. The new type of headband consisted of a strip of pasted linen or calico (sometimes striped, but more often monochrome) folded round a piece of string. This is still available.

Machine-made stuck-on headbands were in use at least as early as the 1850s1 and have been widely employed ever since in all but the best leather work.




Tight Backs

A book is said to have a tight back when the covering material is stuck to the backs of the sections, either directly or when there are layers of other material sandwiched between cover and spine. Generally speaking, the use of tight backs was usual until the beginning of the nineteenth century, when hollow backs began to come into use in England, but this statement needs to be qualified. It seems that in the early days of the English codex there was no attempt to stick the leather cover to the book, or, indeed, to stick the backs of the sections to each other. Examples of this are to be found in the Stonyhurst Gospel1 and in some bindings of the twelfth and thirteenth centuries. The tying-up marks on either side of the raised thongs on some bindings of this later period suggest that attempts at adhesion were being made for the first time about then. It is possible that some books had no covering on the spine at all.

Apart from the early bindings mentioned above, and limp vellum and some stiff-board vellum-covered stationery bindings, the tight back was always used in England until about 1800.2

In France in the seventeenth and eighteenth centuries (and to a lesser extent much earlier) it was a common practice to line the spine with strips of parchment manuscript which overlapped the joints and were pasted down under the endpapers. These linings, employed in conjunction with sewing on cords rather than on thongs, and with covering leather which has had less tendency than ours to become powdery, have helped French bindings to survive in much better condition, and with much less tendency for the boards to become detached, than English ones, which comparatively seldom have been reinforced with parchment. These strengtheners are found in some English bindings of the late fifteenth century, and in some early sixteenth-century ones, but in these they are often attached to the end panels, or middle and end panels only, and are not stuck down on to the board. I have occasionally seen the vellum reinforcements in English bindings of the seventeenth century,1 and first half of the eighteenth.2

The tight back was comparatively rarely used after about 1820, except by fine binders who often employed it in conjunction with false raised bands, a style known as ‘mock flexible’.

All ‘boarded’ books had tight backs and laced-in boards from the 1770s, when this style was introduced, until the 1820s, when casing came in. Many more had tight backs until the late 1830s when ‘boards’ more or less went out.

Interest in tight backs for ‘extra’ work was revived, to a limited extent, by Cobden-Sanderson and Douglas Cockerell, both of whom deplored the recessed cords and hollow-back combination which, at the time in question (late in the nineteenth century) was employed on many expensively bound books which justified more sound construction. Their strictures influenced an army of amateur binders for several decades, but had little effect on trade practice. Today, few books which are finely bound by the trade have tight backs, though lectern Bibles and other books which are going to have hard use and are intended to have a long life are usually sewn on raised cords, and therefore have tight backs.

I should explain for the benefit of the uninitiated that the objection to the tight back is that if the leaves are thicker in relation to their area than they should be, so that they will not lie down without persuasion, the restricted movement of the spine causes the book to open like a fan. The hollow back, on the other hand, allows the spine to throw up in the middle, thus allowing thick leaves to lie more or less flat. This is good for the appearance of the book because it prevents the leather on the spine creasing and breaking up the gold, but has the disadvantage that the throw-up is sometimes angular, thus putting a strain on the sewing. Many thin-paper books are thoughtlessly given hollow backs even when there is no movement in the spine when the book is opened, but this should be avoided because the adhesion of the leather to the back of the book relieves strain on the cords which have been laced into the boards.

Hollow Backs

In France books were bound with hollow backs from about 1770,1 but in England this device is seldom found in books bound before about 1800.2 The fact that Hüttner wrote3 in 1802 that the English do not use hollow backs is a clear indication that they were far from common, allowing for the possibility that he was not conversant with everything that was going on.

The hollow back4 was, I think, still comparatively rare in finely bound books before about 1820, when the heavily-shouldered semielliptical spine (fig. 41) began to give way to the structurally and aesthetically less satisfactory flat spine (fig. 42), which had virtually no shoulders and consequently provided little support. The combination of the weak spine and the use of the hollow back on books which at that time did not normally need them has resulted in many of the bindings becoming exceedingly dilapidated, while those bound with tight backs thirty or forty years earlier are still sound, apart from broken joints.

An early description of the technique is to be found in Parry’s The Art of Bookbinding (1818), pp. 23–4:

open backs and extra bands

When the book is intended to have an open back and extra broad bands, the back must always be lined with a piece of thin sheep leather, then cut a piece of strong paper the size of the back, and mark thereon where the bands are to be placed, with the compasses. Paste a piece of rough calf or any thick leather on thin pasteboard, and press them in the standing-press; when dry, the leather may be cut for the bands and pasted on the before-marked paper, putting it between two boards till dry, and before covering the book, to be exactly placed on the back, and tied round the bands with thin cord.

A typical binding of the 1820s with broad bands is shown in plate 4.

My copy of Dibdin’s Bibliographical Decameron (3 vols., 1817), in a paste-grain morocco binding which may post-date the book by a few years, has hollows made in the way described above, but the spines are lined with coarse canvas and the hollows are edged down at each side.

Although, in my experience, the hollow back seems to have been comparatively rarely used before about 1820, its use spread rapidly soon after this date. Confirmation of its popularity is to be found in the following firm statement in Cowie’s Bookbinder’s Manual (1828), p. 48: ‘This method has come into general use for almost all books of a tolerably good thickness… .’ The technique described in this work (p. 49) was to cut a piece of paper twice the width of the spine and the exact length. This was folded in half longitudinally, one part being glued to the spine and the other taking the raised bands.

Arnett, writing in 1835,1 advocated gluing the paper on, leaving at one side enough to cover the spine, and at the other a long unmeasured piece. When the short piece had been folded back over the spine the long part was taken back over it and trimmed level with the joint. The two loose pieces were then to be glued together.

Later, Zaehnsdorf’s1 method was to glue the first lining on, and trim off. He then glued the second lining on, leaving uncovered a narrow strip of glue along one joint. This held the first ‘off’ while it was glued, the paper was folded back and rubbed down to take the second ‘off’, thus making ‘two on and two off’,2 as shown in my diagram (fig. 59).

The anonymous author of two excellent articles on ‘Linings’ in his series on ‘Forwarding’ in The British Bookmaker,3 regarded the foregoing as a ‘slop’ method. He advocated three or four preliminary ‘on’ linings, each of which was to be progressively stepped back from the joints to equalize the support given to the spine, bearing in mind that the greatest strain is in the middle. The credit for the introduction of this system of graduated linings, he says (p. 58), should be given to Francis Bedford, or rather to his workmen. The author emphasizes the importance of using as little glue as possible, and of sand papering each lining separately. He then describes the sliding-off method—one of the earliest descriptions of this particular technique—which is in fairly general use today in fine binderies. Briefly, this involves gluing the spine, laying a strip of paper on it, level with one joint and overhanging the other. The overhanging part is folded back on itself level with the joint, slid off the spine, and then folded over to make a tube. This is replaced on the spine and the superfluous paper is trimmed off. Depending on which way the hollow is glued on this gives ‘two on and one off’ or ‘one on and two off’, but extra linings can be added as necessary.

In high-quality work it was not unusual until recent years to give the spine a preliminary lining of thin leather, the purpose being to gain a more flexible opening than would be possible with paper linings giving comparable support. This is still done in exceptional cases.

In the best nineteenth-century binderies brown paper was normally used for hollows, but in most cases the quality of the paper was very poor and it is clear that many binders regarded this operation as an opportunity to use up any bits and pieces which happened to be lying around. This applies also to the stiffeners in the backs of publishers’ cases, for which it was common practice to use waste printed matter—advertisements, and the like.

Spring Backs

The most significant date in the history of stationery binding is probably 1799, when John and Joseph Williams were granted a patent for a method of binding the chief point of which was the spring back. The specification shows that the new style of binding was, basically, very similar to the present-day ledger binding. The following is the whole of the published abridgment:1

‘New-invented improved method of binding all sorts of books.’ A back of semi-circular, semi-oval, ‘or any other semi or curved form, turned a little at the edges,’ and made of any metal or other material ‘capable of retaining a firm situation,’2 is put on the book before it is bound, so as ‘just to cover over (but not to press) the edges of the paper.’ This back will, when the book is opened, ‘prevent it spreading on either side, and cause it to rise in any part to nearly a level surface.’ If necessary, ‘the vacuity from the firm back to the angle of the paper may be hid by a cap at each end, made of the aforesaid or any other material, with a spring catch to fasten within the back, for which the back must be sufficiently large to admit the cap, so as to prevent the paper being fretted against, and tin or some other matter placed over the catches.’ The method of binding is as follows:- ‘The paper forwarded in the usual manner, sewed on vellum strips, glewed, cut, clothed and boarded (or half boarded), and the firm back put on by being fastened at the sides through holes by vellum, forrell wrappers, or other matters pasted down upon or drawn through the boards; and the other half the board pasted over, the cover put on, and book finished.’

It is clear that the new principle was an immediate success, for Hüttner, writing1 about two years later, says that Williams made a lot of money out of it.

Before many years had passed it was found that spring backs could be made of materials other than iron. In Parry’s Art of Bookbinding (1818), p. 43, the reader is recommended to make the ‘elastic back’ with strong pasteboard. This was to be soaked in water and dried in shape on a piece of wood. When dry, it was to be glued and covered with strong paper, and again dried on wood.

Cowie2 refers to the ‘hollow back’ and says that millboard should be dipped in water and then glued on each side. After about ten minutes it is to be shaped by placing it on a sheet of paper and around a roller. The roller and millboard are to be rolled up in the paper and worked backwards and forwards so as to mould the board,3 and is then to be placed near the fire to dry.

Arnett, in 1835,4 employs the term ‘spring back’, which is still used. He says that many methods of making the backs are in use, and describes three of the best of them. His first method is to take a piece of pasteboard or thin millboard a little more than twice the width of the spine, this to be folded on each side so that the middle portion is about a quarter of an inch wider than the spine. Several more strips are cut, the first a little less than the width of the spine, the next a little narrower, and so on. The flaps of the very wide piece are then pared; the centre section is glued, then the next largest strip is laid on it, and so on until all the pieces are glued together. Finally, the whole thing is glued and the flaps are brought over and rubbed down (fig. 60), when it is shaped on a roller and allowed to dry.

Arnett’s second method is to cut a piece of firm millboard to the size required, pare the edges, and then soften it before the fire, after which it can be moulded as previously described.

His third method of making a spring back is that patented by Williams in 1799, which involves making it of metal. Although this type had become obsolete it has been used in recent times for extra-large ledgers.

The methods now employed are similar to those described, one to about five thicknesses of board being used, according to the size of the book and shaped with the aid of a roller and paper.


1 Described by Roger Powell in ‘The Stonyhurst Gospel: the Binding’ (p. 370) in The Relics of Saint Cuthbert, Durham Cathedral (1956).

1 Flat ones dating about ten years earlier than this are not infrequently found.

1 A patent, part of which covered the mechanical manufacture of headbands, was taken out by one Mark Bingley in 1846 (No. 11, 495), but I cannot recall actually seeing any machine-made material on books of about this date.

1 ‘The Stonyhurst Gospel: the Binding’, by Roger Powell, p. 370, in The Relics of Saint Cuthbert, Durham Cathedral (1956).

2 Until the last thirty years or so of the eighteenth century, it was not usual to line the spine under the cover. Then it became a common practice to use three or more thicknesses of paper for this purpose in the case of the bandless style which was then coming into fashion on fine bindings, although this was still not done with very cheap ones. Canvas was sometimes used in addition to paper.

1 For example, in a Samuel Mearne binding made for Charles II, the property of Mr H. D. Lyon.

2 Whereas in France in the seventeenth and eighteenth centuries calf and sheep bindings were often reinforced, in England the linings were mostly confined to morocco bindings.

1 The technique is described by Dudin in L’Art du relieur doreur de livres (1772), pp. 81–2, and may well date back several years.

2 Graham Pollard, ‘Changes in the Style of Bookbinding, 1550–1830’, The Library, June, 1956, p. 87.

3 Hüttner, a German resident in London, wrote an article on English Bookbinding, which was published in Englische Miscellen, Band 6 (Tübingen, 1802), a journal devoted to English subjects for German readers.

4 Not to be confused with the loose backs of the limp vellum cases which covered some books when they were issued, mainly during the sixteenth and seventeenth centuries. These were not formed with a tube of paper, and there was no deliberate intention to separate cover from spine for a specific purpose as there was when the true hollow was introduced. This purpose, which may have been two-fold, was to preserve the solidity of the gold-tooling on the spine when the book was opened, and, less likely, to enable the book to open more easily. I think little credence should be given to the second explanation because most books of the period could open very well with tight backs, though later this was not the case.

1 Bibliopegia (1835), p. 58.

1 The Art of Bookbinding (1880), pp. 83–4.

2 This means that two thicknesses of paper are glued directly to the spine, and that two are not. It could be ‘two on and three off’  or any other combination, depending on circumstances.

3 Vol. VII (1894), pp. 58 and 85.

1 Abridgments of Specifications relating to Books, Portfolios, Card-Cases, &c. A.D. 1768–1866 (1870), no. 2355, Nov. 4, 1799.

2 Metal backs of this kind, one without its binding, are in the Public Record Office.

1 In Englische Miscellen, Band 6 (Tübingen, 1802). See below, p. 255.

2 Bookbinder’s Manual (1828), p. 67.

3 It is suggested that the board should go right round the roller so that it can be cut in half to make two backs, but this method is never employed now because it would be difficult to make the back the right size for a particular book, and because the work involved in cutting through the board would be greater than the trouble of making two separate backs.

4 Bibliopegia, pp. 151–2.