Moulds and Paper Sizes E. J. Labarre, in his excellent Encyclopaedia, lists over three hundred named sizes of paper but concludes these are all variations of only fifteen names and only half a dozen actual sizes. Some of these variations are quite bizarre, like "Double double small hand" (40" x 30", which is the same as Quad Crown) and "Pinched Post" (18 1/2" x 14 1/2"). Yet we have plenty of moulds in stock whose sizes do not correspond to anything in Labarre's list. What practical reason can there be for this? There is none, because the driving factor behind paper sizes must surely be market forces with which, somehow, the papermaker had to comply. Practical factors do, however, affect the sizes of paper moulds and this may account for the common sizes Labarre refers to. In his analysis, he points out that most paper sizes have a length to width (or more often a width to half the length) ratio of about 1.6, which is similar to the Golden Section. This can be partly explained by market requirements, in that books and letters would tend to be sought in familiar and aesthetically pleasing proportions. However the shapes and sizes of handmade paper that can be produced depend on what a vatman can handle, which, in turn, is influenced by the span of his arms and, to go by many old photographs, the size of his beer belly! Anyone who has made paper on square or nearly square moulds will know that they do not feel quite right. This is partly because a square mould does not seem to balance as well as a more extended shape. In addition, any mould width of more than about 25 inches is difficult to hold away from the body, which is necessary while forming sheets. One of the largest English sizes made by hand (but not at Hayle Mill) was Grand Eagle (42" x 28"). Double Elephant (40" x 27") was more common. Small moulds could be a problem because to make small sheets one at a time was grossly uneconomic. The answer was to use two, three, or four sheet moulds with divided deckles or to subdivide the sheets with tearing wires (which formed heavy watermarks) to facilitate the tearing down of the sheets when dry. Dard Hunter states that two-sheet (or double) moulds were introduced in Europe by the Whatman Mill in 1826 but this is refuted by Schulte who ascribes the invention to Holland, although he is not specific about the date. Because of their considerable width and weight, two-sheet moulds were strongly resisted by the workers at first and indeed were never accepted in some countries. Our collection of 258 sets of moulds included 82 pairs of two-sheet moulds of which only two had sheets abutting on their long sides, like this-- [first diagram] --making nearly square moulds, with the deckle divider running away from the vatman. Of these, one was quite small and clearly experimental. The other measured 34 1/4" x 27" overall, producing two sheets of 25" x 15 5/8". These moulds were made in 1986 and not only did we find them very awkward to handle, but also the 25" long deckle divider did not cut well. A third two-sheet mould was divided from hand to hand, which almost certainly gave problems as the wave crossed it. [second diagram] Most of the other two sheet moulds were very long, the largest measuring 49 3/8" x 20 1/2", producing two sheets of 23 1/4" x 18 1/4" ("Medium"). In later years we found these moulds very difficult to use and inevitably, without decades of experience, it was hard not to make wedgy paper (thicker at one end than the other). [third diagram] We had only two pairs of three-sheet moulds and three four-sheet moulds, the sheets meeting at their long edges in all cases. Most paper produced at Hayle Mill until the late 19th century was made to customers' specifications and a lot of it was for security papers. These often required sizes that might seem unusual until you realize that they were to be folded and might include a counterfoil. However, there is an indication that we used mostly standard sizes. In the archives we have a handwritten list of prices from Edwin Amies and Son, mouldmakers, dating from 1878, which reads as follows: Size Without With Inches watermarks watermarks ONE-SHEET MOULDS Medium 23 x 18 L5-5-0 L5-10-6 Royal 25 x 20 L5-5-0 L5-10-6 Double Foolscap 27 x 17 L5-5-0 L5-17-6 Super Royal 27 x 19 L5-7-6 L6-2-6 Imperial 30 x 22 L6-6-0 L7-0-0 TWO-SHEET MOULD Foolscap 17 x 13 1/2 L6-0-0 L7-5-0 Small Post 18 x 14 1/2 L6-17-6 L8-3-0 Large Post 21 x 16 1/2 L7-15-0 L9-0-0 Demy 22.5 x 17.5 L7-5-0 L8-10-0 Copy 20 x 16 L7-10-0 L8-0-0 Medium 23 x 18 L8-8-0 L10-0-0 Other sizes were quoted by area; e.g., 150-200 square inches for L2-15-0 and 450-500 square inches for L5-10-0, both without watermarks. It is interesting to note not only the fairly small difference between large and small moulds but also how cheap watermarks were and that there did not seem to be any limitation on their complexity. Also they did not bother to quote for one-sheet moulds for the small sizes. Mould Construction and Quality Other authors have dealt with details of mould construction (Coleman and Moore), and Loeber not only covered this in great detail but also set out important design criteria. At one time I would have thought that the features they describe were all absolutely essential to making good paper. As I was brought up with beautifully crafted moulds, I always assumed that nothing else would do. At Hayle Mill this was certainly the case and we became very unhappy at any deficiencies in moulds, old or new. Now I would have my doubts, at least about some details in some circumstances. Having visited many papermills, workshops, and small factories in the last few years, I could not fail to notice the variation in mould design and quality. Most British moulds aim for the same standard, although there are many interesting variations in details. Moulds in other European mills differ significantly from British moulds. They are often much lighter, with quite slender deckles, softer wood and sometimes of fairly loose fit between deckle and mould. The differences are certainly not obvious in the papers, however. In many parts of India, European-style, rigid, wire-covered, wove moulds are used, albeit with a different dipping technique. Generally the construction is simpler and deckles usually have a plain rectangular cross section and are held on to the mould rather than fitting around it. As deckle edges are trimmed off in India and quality targets are not as high as in Europe, the moulds seem more than adequate for the papermakers' needs. Closely linked to functionality is precision. There are various aspects of this. The deckle needs to fit well, not too tightly or loosely and without allowing pulp to seep under the edge and make unduly feathery edges. It needs to start off with exactly right angled corners and stay that way for perhaps hundreds of thousands of sheets. Printers in particular are rightly fussy that sheets be precisely rectangular. Lack of squareness can be especially obvious in folio books which are not trimmed. Paper can be out of square for many reasons, but to have any chance of success in making paper with right angles the deckle must be as close to perfect as possible. Watermarks must also be precise in their location, shape, and weight. Laid moulds can present a problem as the watermark sometimes slides along the laid lines to which it is sewn. Andrew Hoyem was very unhappy when this happened with the whale watermark in the paper we made for his Arion Press edition of Moby Dick, as the whale's mouth is gaping open in an unduly large number of sheets. Hoyem was not mollified when I pointed to medieval examples, so I did not add that paper historians and descriptive bibliographers are always thrilled with these little peculiarities. Moulds have a hard life and, considering this, it is remarkable how long they last. It is always interesting to look at well-used moulds and see where they are suffering. The bottom left hand corner of the deckle is worn, being the first point of contact with the mould. Cracks appear in the corners and the joints work loose (leading to loss of squareness). Brass and copper reinforcements crack and wear and are sometimes replaced. Laid lines are fragile, especially near to the parallel deckle edges. For this reason there is additional sewing when new and it is not uncommon to have to replace parts of the twist wires and several laid wires. Brass wires become brittle so that sewing comes undone. For this reason, phosphor-bronze is preferable. Ribs are vulnerable to damage. It is common for them to be broken where the pin goes into the mould frame. This is sometimes rectified by binding with wire. The coucher's finger nails and the ass can damage ribs in the middle of the long edges and the moulds have reinforcing wire panels fitted near the bottom of a few ribs. Mould Materials It must have been tempting for mouldmakers to use harder wood to resist damage, but the timber must also be dimensionally stable and light. Mahogany has been the timber of choice for the last century or more. Its abrasion resistance is not very high and boxwood inserts on the bottom of the frame help compensate for this. We experimented with nylon to replace the boxwood, but found that these moulds slid too much on the bridge and stay. We also varied the metal parts, using stainless steel instead of brass for reinforcing the deckles and stainless steel, Monel metal, and aluminium mesh for wove covers. We returned to brass, phosphor-bronze, and copper, finding little advantage in the alternatives. There would seem to be plenty of opportunity to experiment with other new materials for wearing surfaces, cover fabric, and the like. To date, synthetic textile covers seem to either sag or slacken with time but it may well be the choice of exactly the right material that counts. Of course such materials, alloy flyscreen, and many other types of mesh work very well when production quantities are small or with smaller moulds. The Filipino paper industry uses such materials almost exclusively and, although I have given much advice there on mould construction, in many cases they are more than adequate. If you are selling abaca paper for 25 or 50 cents a sheet, you cannot readily justify expensive moulds. Conclusion One of the curious aspects of mouldmaking is that its history is so undocumented, especially prior to 1800. Papermakers and mouldmakers were not only secretive but seemed disinclined to write anything down about their tools. Often the constructional details would have seemed obvious to them and many of the early craftsmen were illiterate anyway. Furthermore, very few papermill records have been kept over the centuries. There are limits to what can be inferred by studying paper, although that has not hampered some self-styled filigranologists. This brief article may shed some light from a twentieth century papermaker's perspective on European moulds that have changed little in a hundred years. Perhaps a reader with access to eighteenth or seventeenth century or even earlier moulds and records can add to our knowledge in the future.   . Labarre, E. J., Dictionary and Encyclopaedia of Paper and Papermaking, Swets & Zeitlinger, Amsterdam, 1952. . Hunter, Dard, Papermaking Through Eighteen Centuries, New York, 1930, p. 236; repeated in Papermaking: The History and Technique of an Ancient Craft, Alfred A. Knopf, Inc., New York, 1947, (republished by Dover Publications, 1978), p. 229. . Schulte, A., Gutenberg Jahrbuch, 1934, p. 20. . Sizes were not given by Amies in inches. By way of comparison, I have taken the most common sizes quoted by Labarre. Demy is particularly variable from 18" x 14.5" to 24" x 18". . Prices are in pounds, shillings (1/20th of a pound), and pence (1/12th of a shilling). In 1878, L1 was US$4.80. . Coleman, S. A., "The Craft of Making Moulds for Hand-Made Papers," Papermaking by Hand in 1967, J. Barcham Green, Maidstone, Kent. . Moore, Tim, "Western Mold Making," Hand Papermaking, Volume 2, Number 1, 1987. . Loeber, E. G., Paper Mould and Mouldmaker, The Paper Publications Society, Amsterdam, 1983. . We have moulds from at least twelve British mouldmakers, although all of the moulds shown here were made by Amies, whose moulds make up about two-thirds of our collection. The other mouldmakers represented in our collection are: Bacon & Wayman, Brewer & Co., W. Green, W. Green Son & Waite, Knowles Trotman, R. Larner, J. Marshall, T. J. Marshall, Sinclair, G. Tovey, and W. & R. Balston Ltd. We also have a mould from Tumba, in Sweden, and Tadao Endo, of Shiroishi, Miyagi, Japan. . This is an alloy containing 67% nickel, 30% copper, and small amounts of iron, manganese, carbon, silicon, and sulphur. It is stronger and more expensive than phosphor-bronze and may have been used because of wartime shortages of bronze.