09 April, 2007

Drawing the typefounder’s mould

The typefounder’s mould was first described during the 16th century, and the principles of its construction do not appear to have changed greatly thereafter. Since the mould consists of two L-shaped pieces that slide together, these principles are essentially simple, but it is not an object that is at all easy to represent graphically.
Philip Gaskell’s New introduction to bibliography, first published in 1972 and reissued with some radical corrections in 1974 (a few more were made later), is still the most reliable and comprehensive historical guide to the processes of printing and the related trades. Its account of the setting of type and printing at the hand press was based not only on his reading of technical works but on personal experience at the Water Lane Press, the printing office that he set up in a cellar at King’s College, Cambridge, and where some undergraduates joined him. I was one of them.
When I first saw Gaskell’s book in 1972 I thought that his illustration of the mould on page 11 of the New introduction (shown above) was one of the clearer and better images that I had seen. I still think so, but the more I looked at it the more puzzled I became. Typefounding was not one of the trades that he knew at first hand, and this probably accounts for his omission of any discussion of the illustration, or indeed any indication of its source. I found that this was a more complicated matter than I had thought.

The origin of the image is evidently plate 7 in the first volume of the Manuel typographique of Fournier le jeune (Paris, 1764), above, but it is reversed laterally. It does not show the mould that was in common use in France and Britain, which is on his plates 5 and 6, but an alternative design, ‘the mould used in Germany, Holland, and elsewhere’, and although this operates on the same principles it does have some different features. But Gaskell’s image, although it shows Fournier’s German mould, does so at a distance, via other interpretations.
His figure includes not only a mould but several other objects, a punch, a matrix, a piece of type, and also a ‘schematic diagram of the casting mould’ to show its principles. These had all been published some years before in an article by Otto Fuhrmann, ‘A note on Gutenberg’s typemetal’, in the Gutenberg-Jahrbuch, 1950.

As his caption acknowledges, Fuhrmann’s images had in turn been assembled from several separate unattributed illustrations in an article on typefounding by Friedrich Bauer, ‘Schrift und Schriftguß’, in Das moderne Buch, the third volume, issued in 1910, of a series published in Stuttgart under the general title Die graphischen Künste der Gegenwart. The drawing of the mould is shown above.

The mould in Bauer’s article may possibly have been suggested by an engraved plate, above, that was published a century earlier by C. G. Täubel in the third volume of his Allgemeines theoretisch-praktisches Wörterbuch der Buchdruckerkunst (Vienna, 1809). There is no doubt that Täubel’s image of the German mould was taken from Fournier’s plate, since it also shows the different parts of the mould in an exploded view on the same plate. In the process of copying the image was ‘flipped’ horizontally. However, the angle at which the mould is represented was altered, and so was the relative position of the two parts, so that at first glance the reversal is not apparent, and the hooks which are used to extract the types that stick in the mould after casting had their position switched so that they seem to match those of Fournier’s original plate.
Why the slight deception? Images are often reversed when they are copied, but in this case, I suspect, one reason for the reversal may be that Fournier’s ‘German’ mould has its ‘body’ on the right when it is in the casting position, as in French moulds. German moulds seem generally to have had the body on the left, in the manner shown in the plate in the handbook Die so nöthig als nützliche Buchdruckerkunst und Schriftgießerey, published by C. F. Gessner in Leipzig in 1740, and some surviving later German moulds seem to confirm that this was the German practice. By flipping Fournier’s image, Täubel was able to show a mould that looked more convincingly German.
The mould shown by Bauer reproduces Fournier’s image more faithfully, although also in reverse, and since it keeps the hooks in their correct relative position, it may have been derived directly from the Manuel typographique. But there are some changes too. A wooden sheath is added to the end of the spring that holds the matrix in place, perhaps as insulation (since all metal parts of the mould get hot). This is a feature that is often seen in German moulds but not in French or English ones. And only one turn is given to the coil of the spring where it is fixed to the insulating wood.
In adapting Fournier’s image, Täubel had perhaps inadvertently preserved one very French feature: the mould that is shown on his plate, and in Bauer’s illustration and Gaskell’s, would cast type with the nick above the letter, in the French manner, and not below it, as was the practice in Germany and England. But the separate example of a cast printing type shown by both Bauer and Gaskell has the nick below and could not have been cast in the mould that is shown (for which it is in any case too big), something that must puzzle the observant reader of either text.

Acting no doubt with the best of intentions, Gaskell himself introduced one last element of confusion. For his extensively revised impression of 1974, he had the image of the mould redrawn with a ‘pecked’ line, presumably to show its internal structure more clearly. He also added ‘a matrix, held in place by the spring’, in order to show how it fits in the mould. The matrix that is shown is a simple slab of copper with a round indentation at the back. It is not a matrix for casting with a hand-mould. In his notes to the reprint of Moxon’s Mechanick exercises (2nd edition, London, 1962, at page 161) Harry Carter observes that ‘matrices meant for casting-machines have a shallow hole at the back instead of a notch. The hole is no good in a hand-mould: the bow [or spring] tends to pull the matrix off its seating on one side.’
Gaskell died in 2001. Since his text is still in current use, it seems useful to continue to make corrections to it where they are needed. The image of the typefounder’s mould in his book is clear and has some value, but Gaskell, as befits a bibliographer, was a scrupulous writer, identifying even the individual copies of the books he photographed for their type and crediting the many individuals who offered corrections to his text. He clearly had no idea that this image had been altered from its original model to the extent that is set out here.