the binding


The manuscript was probably first bound soon after it was written. (For a detailed account of medieval binding structures and techniques see Szirmai 1999) The quires were sewn to thirteen supports, with kettle (or change-over) stations at head and tail, about 65 mm apart. The stations, one for each support, through which the thread passed were made with a knife, with cuts extending inwards about 8 mm from the spine-fold of the outermost bifolium of each quire. The supports were probably made from tawed skin straps (none of these have survived).

Thread stains in the spine-fold of the innermost bifolium of each quire show that the sewing pattern was all-along, and that the binding had end-bands at head and tail with their thread passing through the kettle stations. (How the back corners of the quires were treated is unknown as the manuscript was retrimmed when rebound in the nineteenth century.)

The present wooden boards of the binding are almost certainly from the first binding and therefore of thirteenth century date. (No new boards are mentioned in an invoice for the rebinding of the manuscript in 1819, and this confirms the antiquity of the present boards.) They are about 12-13 mm thick and slightly rounded at the exterior edges. Both boards would have been originally square, but they are now slightly curved (with the curve running vertically and the inner part of the curve of both boards facing the manuscript) having moved a little over time. (This curve suggests that the boards were carefully cut, with the grain of the wood running vertically, and that the boards were each made from single planks.)

A brief description of the binding made in 1811 (Hammarsköld, pp. 157, 163, footnote i) suggests that the upper board was then broken, and there is some evidence on the upper paste-down (f. 1) that this board had split down the centre. If this was so, it was presumably repaired in 1819. 

The sewing supports passed into tunnels in the thickness of the boards in the fashion common in Romanesque bindings. (The tunnels are clearly visible in the upper board, where the board has moved a little away from the manuscript.) However, it has been impossible to determine how the boards were cut to accommodate the remainder of the supports (or slips) as the present paste-downs and present cover hide all the evidence at the interior and exterior surfaces of both boards.

At some stage this binding had metal furniture applied, certainly to the lower or back cover and probably to the upper or front cover, but the date of this furniture is unknown. There are rust marks on the last leaf of the manuscript (f. 312v) from the nails used to attach the furniture, and this shows that there were probably four bosses or corner pieces and a central piece. This leaf also has other rust marks in small groups at the three outer edges of the leaf, and these must have been from straps or clasps to keep the book closed when not in use, one each at head and tail and two at the fore-edge. These are perhaps most likely to be additions rather than to belong to the original binding campaign.
The manuscript was rebound in 1819 by Samuel Sandman, a binder active in Stockholm between 1813 and his death in 1838, who worked for the royal library for many years. The binder numbered all of the quires in arabic numerals in red crayon at the foot of the first rectos close to the spine. However, the first quire (ff. 2-7), formerly one of eight leaves with the first and last leaves lost, has the first three surviving leaves numbered 2, 3 and 4. This suggests that the absence of the two outermost leaves had been noticed, but also that the surviving leaves were detached as they are now. The manuscript was resewn on thirteen supports again, made of tawed skin straps, with the thread passing through circular stations, two for each support. The new stations are close to the older ones, and some of the older stations were reused. The sewing appears to have been, in part at least, packed, and, when completed, all three edges trimmed, and the spine rounded, lined and glued.

The original boards were reused and covered with thick tawed skin that was tooled with an elaborate design in blind front and back and on the spine. The covers have a conventional design of a frame within which are diagonal lines with small tools stamped in the spaces between the lines. Prominent among the tools is a crown, but this and the other tools have not been identified elsewhere.

Both boards have thick pierced metal furniture, each with four corner pieces and one centre piece, with a green-blue paper between the furniture and covering skin. All the pieces incorporate raised circular elements acting as bosses to lift the cover from the surface on which the book was laid. The corner pieces incorporate pairs of griffins, facing each other, but the centre pieces contain no animals or figures. In addition, the lower or back cover has two small additional pieces of circular furniture above and below the centre piece. These have a central hole through which passed an iron raised element also with a pierced hole (the one closest to the foot of the book has been lost) that could have been used with chains or something similar to attach the book to a piece of furniture in a library.

The innermost corner pieces on both boards and the lower outermost corner piece on the upper board  are more carefully finished than the other three. This suggests that other three pieces are copies, made to supply lost or missing corner pieces. The presence of some metal furniture is made clear in a brief description of the binding made in 1811 (Hammarsköld), although the description is not entirely clear as to whether all of the furniture was present or not. However, the invoice of the binder Samuel Sandman for his work on the binding states that there were three new corner pieces, and these must be the two outermost ones on the lower board and the top outermost one on the upper board. This written evidence confirms that the rest of furniture must date from before 1819 and that it was reused. The furniture that survived until 1819 is certainly not thirteenth century, and is perhaps most likely, to judge from its appearance, that it dates from some time before the arrival of the manuscript in Sweden in 1649. However, what is peculiar is that the centre pieces look more like the three replacement pieces than the five original ones, and their status is uncertain. 

The invoice of Samuel Sandman survives as a copy in the Royal Library accounts. He was paid 30 riksdaler for repairing and cleaning the leaves, covering the binding with skin and tooling it front and back and on the spine. For the necessary parchment and the covering skin he received 15 riksdaler (the skins were bought from a tanner named Holmberg), and for three new corner pieces he received 10 riksdaler (these were bought from a bellfounder named Horner). He was paid one riksdaler for new nails and for fixing the metal furniture, and, finally, he was paid 22 riksdaler for a leather case bought from a sadler named Rydholm. This leather ‘case’ survived until at least 1906, for a photograph of the manuscript made in that year shows what appears to be leather cover visible at one of the board edges and at the spine. This appears to have passed from the fore-edge of one board and around the spine to the fore-edge of the other board. There must have been holes cut in it to allow the metal furniture to project through the cover, and these holes and the metal furniture must have helped to keep this cover in place, although it would have been easy to remove and replace as it appears not to have been fixed in any permanent way to the present cover. (This, of course, is why it was easy for it to have been detached and mislaid or lost at some date after 1906.) In total Sandman received 78 risdaler for his work on the Codex Gigas. (This is difficult to relate to modern currency, but in 1820 a cow could cost 45 riksdaler and this gives some idea of the value of Sandman’s work.)




































Tooled Crown







Metal furniture







The invoice of Sandeman

The invoice of Sandeman 


stereoscopic picture from 1906

Stereoscopic picture from 1906

Leaf Marks

One of the few pieces of evidence that shows that the Codex Gigas was used, or at least intended to be used in the late middle ages, are leaf-marks of tawed skin at the fore-edges of leaves containing the beginning of a new text. However, the Isidore text has two (one for book 1 and one for book 10), and the Old and New Testaments have the beginnings of most books marked. Titles were written on the leaf-marks in an expertly written formal textura quadrata script, and this looks to date from in or about the fifteenth century.