6 3. Materials
Dd is made up of paper folios with parchment inner and outer bifolia. The paper is worn at the edges and damp or water has marked the top margins, damaging the ink of some of the rubrics in the header (e.g. fol. 142). The parchment leaves are also in poor condition. They seem to be cheap left-overs: e.g. fol. 109 has a hole at the bottom, and the scribe wrote the last three lines of the text around it. On fols. 180/181, glue seems to be visible at the bottom of the bifolium, as if the single bifolium was taken from another manuscript or another block of sheets of bigger size and here folded into two. The paper and the parchment leaves were carefully cleaned in 2004, and a detailed discussion and report of the conservation work can be found in 'The conservation of Dd.4.24' by Kristine Rose.
The size of the folios is variable depending on the material. The paper folios are regular 200 mm in width x 290 mm in length; the parchment folios vary from 144 mm in width x 275 mm in length (fol. 13) to 190 mm in width x 290 mm in length (fol. 120). The parchment bifolia used externally are more regular in size and bigger than those used in the inner part of the quires.
1 3.1. Paper-stocks
The paper is a thin laid watermarked paper in quarto format. The sheet was first folded along a line parallel to the shorter side, and then along a line parallel to the original longer side; consequently the chain-lines are horizontal and the watermarks appear in the gutters. Most probably all sheets were cut at the top of the quire (see section 3.2.).
From an approximate calculation each sheet before folding was ca. 580 mm x 400 mm after it was trimmed, which indicates the paper-size called Royal. The size called Royal or a variety of it provides the nearest match, whose size is 250 mm x 200 mm with variation 225 mm x 180 mm and 260 cm x 200 mm (Ornato et al. 2001). This size is also confirmed by the opening of fol. 182r and its conjugate after disbinding.
The manuscript was possibly written on two paper-stocks, as two main types of watermark can be traced across Dd. It is difficult to ascertain whether both paper-stocks are complete, but it is likely that they are. Bibliographically, a paper-stock should be identified considering both the number of watermarks present in the quires and its twins (Needham 1994, pp. 24-64). This is important because in a papermill the vatman uses two moulds to make paper, the marks attached on these moulds are similar, although not identical, and known as twin marks. These twin marks identify those sheets produced with those two moulds (Stevenson 1951-52, pp. 24-64 and 1968, 1, p. 16).
Two dragons appear in quires 1-8 and a dog in quire 9. The watermarks of quires 1-8 are alike, but not the same and they could be twins. As the eleven congener folios of quire 9 are lost, it is uncertain whether the paper in this quire constitutes another complete paper-stock. I have argued elsewhere that the fragments of the dog watermarks on the remaining folios do not reveal any twin (Da Rold 2003, p. 112). However, after the partial disbinding of the manuscript it has been possible to examine in more detail the indentation of fols. 198-199. Fol. 199 contains the only example of the tail of the dog, and it seems to appear on the right side of the mould, rather than the left and therefore could be the twin of the other head whose tail is lost.
Prior to the disbinding of Dd, I adopted a system to indicate the two halves of the watermark visible on the conjugate leaves, which I retain here despite the availability of reproductions with transmitted light, which show the two halves as a whole. It is noticeable that the watermarks in the gutters are damaged and therefore it can still be difficult to obtain a more precise measure. It was not possible to obtain a reproduction of the dog, therefore each half-mark is reproduced and measured separately .The two halves of the watermark are designated A and B, and A1 and B1 when a twin is found, and as such they are represented in section 3.2.
The size of the watermarks is expressed using Tanselle's discussion of Stevenson's suggestions (Tanselle 1971, pp. 46-48 and Stevenson 1962, p. 200). Using this system it is possible to indicate both the maximum height and width of a mark, including 'the distance between the chainlines and the position of the watermark in relation to the chain line. … whichever dimension of watermark crosses the chainlines is recorded in brackets, with the distance to the nearest chainline on either side entered on each side of the brackets' (Tanselle 1971, p. 46). As the chainline cuts through the marks, the bracketed measurement indicates the distance between the chainline and the mark. The mark cannot be measured at one time, therefore "at least" is added to indicate approximation in the size which is expressed in mm. In the second paper-stock, I preferred to indicate the size of each half of the mark, rather than adding them up, as they probably belong to different paper-moulds.
The first paper-stock includes both the dragon mark A/B, measuring at least 61 x 7[22/20]9, and also the dragon mark A1/B1, measuring at least 68 x 4[25/21]9. A/B is smaller than A1/B1. In addition, B has a slightly higher number of visible sewing-dots than B1 and the tail of the dragon is 13 mm from the tranche-file as compared to B1’s distance of 1 mm. Further proof that these two watermarks constitute a paper stock is evinced by the position of the watermark on the mould. Sheets of paper from a medieval mould have a mould side and a felt side. The mould side is the part of the sheet which touches the wires of the form and retains the indentation of the wires, whereas the felt side is the opposite side which was usually placed on a felt so that the sheet could dry before being pressed. An analysis of the running of the mould and the felt side of the paper in Dd has revealed that the twin marks are ‘reverted’, that is the twins are attached one on the left and the other on the right of the mould which made the original sheet. Not an unusual feature as Coleman emphasises when analysing a paper manuscript of Boccaccio’s Il Teseida. He comments: ‘The moulds [i.e. the marks] were attached to different ends of the two paper forms when the paper was manufactured. As a result the twins pairs are often mirror images of each other on the paper sheets or are upside-down from each other’ (Coleman 1997, p. 23).
The second paper-stock has the dog mark A/B, measuring at least 28 + 36 (fol. 199 was repaired) x 4[25/24]5.
dog mark A, measuring at least 28 x 4[25/24]5
dog mark B, measuring at least 36 (fol. 199 was repaired) x 4[25/24]5
3 3.2. Fingerprinting the dragon paper-stock
The reconstruction of the two moulds of the dragon paper-stock from the folded paper confirms that the laid wires of each paper folio are no more than 1 mm apart. Chain-lines are 40 mm apart with the exception of the one used to chain the watermark. This chain-line or tranche-file is 30 mm distant from the preceding and following chain-line. From the end of the fourteenth century it is customary in Italian paper to insert an additional chain-line tranche-file to which the watermarks were attached. This may be an initial indication that the paper is of Italian origin, although the practice was not exclusively Italian (Briquet 1923, 1, p. 14). The mould includes eight chain-lines on the watermarked folios and seven on the unmarked ones, for a total of 15 as a whole.
4 3.4. Dating the paper-stocks
Scholars in the past have tried to identify and date the watermarks in Dd. It is not plausible, I believe, to offer a precise date for the production and use of the paper-stock in Dd because we do not have a comprehensive dataset of medieval paper used in Britain in the late medieval period (Da Rold 2007b). However, for comparative purposes it is possible to rely on publications based on paper found in archives in the rest of Europe. Here I offer an overview, hoping that more precise data may become available in the future.
Manly and Rickert (1940, 1, p. 100) identify the watermark in quires 1-8 as a Basilic similar to Briquet 2630, dated 1384-92 (Briquet 1923, vols. 1-2). Although this match is the closest of the marks in Briquet, it does not show sufficient similarities to be the right one. Mosser (Mosser 2011) suggests that the mark is unidentified in Briquet, but it is near the twins Drache 266/319, which are dated 1401 in Piccard (1980). Mosser (2011) also points out that there is a variant state from the same mould as a tracing in the Briquet Archive in Geneva: Papiers Briquet, Basilic 9024, Udine, dated 1402.
These two propositions are plausible, however, the dragons in these reproductions have too big a lower jaw to be the same as that in the paper of Dd. Moreover, the position of B and B1 in respect of the tranche-file differs from the watermarks considered by them. Perhaps a better analogue in Piccard (1980) could be 322 or 324 and its twin 318, for the position of the watermark to the tranche-file and also for the tracing of the jaw. The date proposed by Piccard for this watermark is 1392, 1393 for the first mark and 1394 for the second one. It seems that the dragon watermark belongs to a paper-stock datable to the end of the fourteenth century.
The paper of the last quire is later than that used in quires 1-8. Manly and Rickert (1940, 1, p. 100) suggest an unidentified Basilic. Mosser (2011) proposes that the watermark is close to a Chien, Briquet 3597 (Palermo, 1413-16), and a closer match can be found in the Briquet Archive in Geneva: Papiers Briquet, Chien 6652, Archiv, Palerme dated 1416: fols. 194-203 (Mosser 2001). Mosser (2011) also notes the Zonghi mark No. 989, dated 1400 (Gasparinetti 1953). The tracing of the Chien, Briquet 3597, dated 1413-1416, is close to the one in Dd, but as Mosser (2011) remarks ‘if both were made from the same mold [sic], it would appear that the Dd stock was made earlier since the Dd watermark preserves considerably better detail’. Zonghi 989, dated 1400, offers a good match, but the resemblance to the tracing in the Briquet Archive, dated 1416, is better still. The date on the Briquet Archive paper, 1416, however, does not reflect the date of the production of the paper, rather the date of the document in which the watermark is found. Further comparison on these possible variants must be carried out to see how many similarities they share and whether it is desirable to exclude one or the other in an attempt to date the paper in Dd.
I super-imposed the mark of the Briquet Archive, Chien 6652, over that of Zonghi 989. Many details are similar even if the two pictures as a whole are not identical. Their position is also alike in respect to chain-lines and tranche-file. This exercise confirmed that the earlier watermark, Zonghi 989, dated 1400, cannot be rejected.
Briquet (1923, 1, p, 231) had already compared his tracing 3597, dated 1413-16, with Zonghi 989, dated 1400. This comparison was further considered by Mosin and Traljic. These two scholars published a collection of watermarks in 1957, taking into consideration and reproducing marks from other albums such as Briquet and Zonghi. Their ‘chien entier’ 2547 and 2548 appears in the catalogue as ‘2547-48. 1400, Fabriano (Z 988, 989); vs 1413-16, Palermo (B 3597) (Mosin 1957, p. 97).
Briquet 3597, 1413-16, is very similar to the tracing in the Briquet Archive, Chien 6652, dated 1416; it is probably a later variant from the same mould. This implies that the two tracings are related and also similar to the tracings in Zonghi. The fact that Briquet 3597 existed in earlier variants is also documented in Stevenson’s edition of the Briquet album in 1955. In the Addenda and Corrigenda, Stevenson (1955, p. 65) writes that a variety of Briquet 3597 is found in a document in Spain in the Diocesan Archives ‘Curia Funda’, Manual Codina, dated 1402/9.
To summarise the argument, there appears to be a connection between Briquet’s tracings, Zonghi’s example and the dog watermark in Dd. I suspect we are dealing with the production of paper from twin moulds. The watermark that was found in the Spanish archive also suggests that the time lag between production and use might be shorter, and also proves that variants appeared in Europe at a fairly early time. It is accepted that a mould could have life of two years, but more research is necessary on how quickly paper-stocks could be in use after it was made, especially in Britain, including time lag from shipping to the point of sale. If all the dog paper-stock came from the same papermill, it would mean that it was produced around 1400 or earlier and was still available in 1416.