The Codex Gigas contains 310 parchment leaves and two paper leaves as follows: i (paste-down) + 309 + ii (paper end-leaf and paste-down), foliated 1 (paste-down) + 2-7, 8-272 and 274-312 (between 272 and 274 is a slip sewn to 274 foliated 273). The leaves are now 890 mm x 490 mm but, because a little has been lost when the manuscript was re-trimmed, were originally about 900 mm x 500 mm.
The lower or back paper end-leaf and paste-down were formed from several pieces of paper. Underneath the front (or upper) paste-down (f. 1) are at least one layer of paper and another of parchment, the edges of which are visible protruding from the edges of the present paste-down. However, the extent of these leaves is unknown and they have not been included in the leaf-count above. The first quire (ff. 2-7) is now detached from the manuscript.
The manuscript was collated almost entirely in regular eights: a (f. 1, the paste-down) + i8 with 1 and 8 lost (ff. 2-7), ii8 (ff. 9-16), iii8 (ff. 17-24), iv8 (ff. 25-32), v8 (ff. 33-40), vi8 (ff. 41-48), vii8 (ff. 49-56), viii8 (ff. 57-64), ix8 (ff. 65-72), x8 (ff. 73-80), xi8 (ff. 81-88), xii8 (ff. 89-96), xiii8 (ff. 97-104), xiv8 (ff. 105-112), xv8 (ff. 113-120), xvi8 (ff. 121-128), xvii8 (ff. 129-136), xviii8 (ff. 137-144), xix8 (ff. 145-152), xx8 (ff. 153-160), xxi8 (ff. 161-168), xxii8 (ff. 169-176), xxiii8 (ff. 177-184), xxiv8 (ff. 185-192), xxv8 (ff. 193-200), xxvi8 (ff. 201-208), xxvii8 (ff. 209-216), xxviii8 (ff. 217-224), xxix8 (ff. 225-232), xxx8 with 8 cut out (ff. 233-239), xxxi8 (ff. 240-241, 142, 243-247), xxxii7 2, 4 and 5 are singletons and 6 and 7 have been cut out (ff. 248-252), xxxiii8 (ff. 253-260), xxxiv8 (ff. 261-268), xxxv8 + 1 5 is a slip sewn to 6 (ff. 269-277), xxxvi8 (ff. 278-285), xxxvii8 (ff. 286-293), xxxviii8 (ff. 294-301), xxxix6 4, 5 and 6 cut out (ff. 302-304) and xl8 (ff. 305-312).
New items usually begin on the first leaf of a quire. There are three exceptions. One new work (Josephus, Antiquites) begins on f. 118r (q. xv leaf 6), another (Josephus, Jewish Wars) begins on f. 178v (q. xxiii leaf 2), and one lost work (Rule of St Benedict) began and finished on one or more of the three leaves lost following f. 304 (q. xxxix leaves 4, 5 and 6). A fourth, and special exception, concerns the first item in the manuscript, the book of Genesis. The prefatory material to Genesis was written on f. 1v (an end-leaf) and Genesis itself began on the first recto of quire i on a leaf now lost, no doubt so that the effect of a grand opening initial was not limited by script above it.
The verso of the upper or front paste-down (f. 1) contains text (the recto is blank) and this may once have been part of a preliminary quire (perhaps of four leaves) of which this is the only leaf to have survived. The first quire is now detached and it has lost its two outer leaves (the leaves before f. 2 and following f. 7). The leaves cut out (q. xxx leaf 8 and q. xxxii leaves 6 and 7) were probably blank and came at the ends of items, and these may have been cut out early. One item has been lost (Rule of St Benedict) and this was on all or some of three leaves cut out following f. 304 (q. xxxix leaves 4, 5 and 6).
The only irregular quire (q. xxxii) is the last of several containing one item and it appears to have been assembled in an ad hoc fashion from at least some leaves (singletons) otherwise regarded as unsuitable. The only slip in the manuscript (f. 273 in q. xxxv and sewn to f. 274), approximately the size of a quarter page, appears to reveal a miscalculation on the part of the scribe concerning the amount of text that he could fit into a predetermined space.
The present foliation is in arabic numerals in dark ink at the centre of the rectos of each leaf, and it may date from the seventeenth century. There is an earlier foliation in a combination of arabic and roman numerals in pale ink at the lower inner corner of every tenth leaf.
The first leaf with this early foliation is f. 30v (= 30), and this means that this foliation included the two missing leaves in quire i. (No foliation is visible on ff. 10v and 20v.) The foliation has arabic numerals to 90, but the next figure is C for 100, followed by C10, C20 and so on until CC for 200, and then this pattern is continued until CCC (f. 304v). The foliation is accurate as far as f. 130v (= 130), but the next numbers are on f. 141v (= 140) and f. 152v (= 150) meaning that eleven leaves were counted as ten. The next two numbers are in their proper places, ff. 162v and 172v (= 160 and 170), but then there are another eleven leaves counted as ten for the next number is on f. 183v (= 180). The sequence then continues to f. 263v (= 260) when another eleven leaves are counted as ten for the next number is on f. 274v (= 270), but the rest of the sequence is accurate, ending on f. 304v (= 300). This shows that the leaves cut out after ff. 231 and 252 were already lost, and it also appears to mean that the foliation did not include the slip (f. 273).
The scribe who foliated the leaves wrote the total number of leaves on f. 312v (308), and this shows that the three missing leaves, some or all of which had text, following f. 304, were lost when this foliation was made. It is difficult to date arabic numerals, but the the likeliest date for this foliation is either the late fifteenth century or the early sixteenth and certainly before 1561.
Above the total number of leaves written by the scribe who foliated the manuscript in 10s is a note dated 1561 stating that the manuscript contained 3011 leaves. However, this is clearly a mistake for the 11 was written over erasure, and presumably what was erased was a single numeral. A later hand has corrected this total by writing 308 above the note, and this is accurate if the slip (f. 273) is not included in the total.
The parchment might be calfskin and it is of moderate quality. One skin made a bifolium or two leaves, and originally a bifolium was about 890 by 1000 mm. This is very large, but it is smaller than the thirteenth-century Mappa Mundi at Hereford Cathedral in England which is also thought to be calfskin (Clarkson 2006, p. 96). The parchment varies a little in thickness and colour, and the lower outer corners are usually a little thicker than the rest of the leaf. The hairsides tend to have a pale yellowish cast and the fleshsides are near-white. The arrangement of the parchment in the quires is regular, with hairsides facing hairsides and with hairsides on the outside of the quires and therefore follow Gregory’s Rule.
The outer margins of most leaves have been brushed with a liquid that left a slightly yellowish stain. This might be parchment size and it may have been applied when the manuscript was rebound in the early nineteenth century. Some of the outer edges of some leaves were repaired with parchment, probably at the same time, and these repairs are clearly visible as their parchment is much whiter than the original parchment.
In a few places what appears to be a reagent was applied in the nineteenth century in the hope of making particular passages of text more legible (ff. 1v, 305r and 308v). The reagent, as is usual, quickly turned dark brown and whatever text is underneath now cannot always be seen or recovered.
Small holes were pricked in the four margins of each leaf to guide the ruling of lines to guide the scribe in writing the texts. The ruling was done one leaf at a time with a fine point that left a furrow on one side of a leaf and a ridge on the other. The arrangement of the lines on the leaves, or the ruling pattern, is complicated and appears to be consistent throughout the manuscript. The overall pattern has two columns, usually of 106 horizontal lines, with six vertical lines between the two columns and six at the outer edges of each column. In addition, the first and last six horizontal lines in each column extend across the entire width of the leaves. All the horizontals usually extend the full width of the space between the two columns.
There are several curious features about the ruling. First, the use of a point, rather than plummet or crayon, in the early thirteenth century is very conservative. Secondly, the ruling was usually done on the rectos, and this means that ridges appear on versos and furrows on rectos, and this is contrary to the usual procedure in Romanesque manuscripts in which ruling by point was always done on hair sides so that facing pages were always the same, with either ridges or furrows.
There is one exception to this standard ruling pattern. The quire with the Calendar + Necrology (quire xl ff. 305-312) has a different ruling pattern because of the need to arrange the text in columns. Holes were pricked in the top and bottom margins so that five columns could be ruled, a wide central column flanked on either side by two narrower ones. Each column was defined by one vertical on either side but for the outer edges of the area occupied by the columns where there are six verticals ruled in the same manner as the rest of the manuscript. Another distinctive feature to the ruling of the Calendar is that after the leaves had been ruled in point every third horizontal line, the lines on which the Calendar entries were written, was ruled in yellow, and on the leaves 305v, 310v and 311r one vertical yellow line was ruled either side of the area ruled for writing.