In the 1830s, the introduction of the direct cutting of wood type by pantographically controlled routers provided the first easy solution to the problem of large type. (Large metal type made a comeback later in the 19th century with the technologies of stereotyping and electrotyping and with the direct pantographic engraving of matrices. Wood type remained viable, however, until the end of commercial letterpress printing.)
Prior to inexpensive wood type, the production of large types was difficult. If cut by hand in wood, the process was labor-intensive. The various methods of producing metal types in large sizes all had issues. Rob Roy Kelly, in his American Wood Type: 1828-1900 (NY: Van Nostrand Reinhold, 1969; reissued 2010 by Liber Apertus Press) provides an excellent survey of early large types (mostly wood, some metal) with some notes on the methods used to make them. Since then more research has been done and it is now possible to list several different methods by which large cast metal types were made:
To this list might be added various hand and (simple) machine methods of using cast-lead matrices made from existing types in "makeshift" typecasting as practiced from at least the 18th century through the 20th, although these methods were not specifically for large matrices.
From at least the 16th century through the end of the 18th, large metal types were made using traditional sand-casting techniques (the sand here performs the role of the matrix). Typically, hand-carved wooden patterns were used for molding the sand. In American Wood Type: 1828-1900, p. 29, Rob Roy Kelly describes two instances of this.
One of these methods, attested by "records preserved in the files of the Plantin-Moretus Museum" was employed by the "sixteenth-century Flemish punch-cutter" van der Keere. He carved patterns in wood. Alignment was done in the wood pattern stage of the process. These patterns were used in conventional sand-casting of presumably relatively thin printing faces; these were in turn "fused to metal bodies or fastened to wood blocks" and made type-high.
In the 18th century, these methods were used by various typefounders including Caslon and, especially, Cottrell. Reed says 'It is well known that until comparatively recently the large "proscription letters" of our foundries, from three-line pica and upwards, were cast in sand. This practice died out at the close of the last [that is, 18th] century." (10) See Talbot Baines Reed. A History of the Old English Letter Foundries (London: Elliot Stock, 1887): pp. 10 ("proscription" letters), 248, 250 (Caslon), 291, 292 (Cottrell), 305 (Fry).
Ray Millington, in Stephenson Blake: Last of the Old English Typefounders . (New Castle, DE / London: Oak Knoll Press / The British Library, 2002.) says that Cottrell used brass patterns, but I'm not sure that this isn't a misinterpretation of ambiguous source material. Millington does not cite his source for this.
(So far, the only reference I have been able to find to brass with regard to Cottrell is a citation in Kelly that must be traced back several levels. Kelly is at this point discussing the "Sanspareil" matrices (not sand-cast matrices) and notes that there is some suggestion that Cottrell may have introduced them. He cites James Mosley (in "English Vernacular," Motif. [No. 12 ?] (Winter 1963-1964): 18) who in turn cites the 1764 volume Universal Director: Or, the Nobleman and Gentleman's True Guide to the Masters and Professors of the Liberal and Polite Arts and Sciences (Thomas Mortimer, 1763), who says that Cottrell invented "a new method of cutting letters in brass for monumental and other inscriptions." This says neither that Cottrell used brass patterns for sand casting nor that he used matrices of the "sanspareil" form.)
[Note to self: I need to track this down further. I have not yet acquired the issue of Motif in which Mosley's essay appeared. The 1763 Universal Director has been scanned by Google four times, but they are suppressing these scans (presumably because Gale offers a print-on-demand version from microfilm).]
Reed indicates that sand-cast types were supplanted relatively quickly around 1810 by the so-called "sanspareil" laminated matrices. These, in turn, fell out of use after 1834 and the introduction of direct pantographic engraving of wood type.
This process for making matrices for casting large types has been reconstructed by James Mosley. It involved first carving a wrong-reading relief letterform (in some material not recorded). This was then used as a pattern for sand-casting a brass wrong-reading relief letterform (a rather flat affair, like the tip of a punch without the body of the punch). This cast-brass "punch" was then driven into a lead planchet to form what I am here calling a lead "intermediate strike" (Mosley does not use this term). This intermediate strike was then used as a pattern in conventional sand casting to cast a brass matrix. This brass matrix, finally, was used to cast type. Mosley dates types made by this method to the middle of the 16th century. Miraculously, three components of the matrix making process for this type (cast-brass "punches," lead "intermediate strikes," and brass matrices) survive in the Joh. Enschedé en Zonen museum in Haarlem (Netherlands). See: James Mosley's "Typefoundry" blog posting "Big Brass Matrices Again: the Enschedé 'Chalcographia' Type" (2008-03-21): http://typefoundry.blogspot.com/2008/03/big-brass-matrices-again-ensched.html and also his earlier "Big Brass Matrices: a Mystery Resolved?" (2007-03-19) for context: http://typefoundry.blogspot.com/2007/03/brass-matrices-mystery-resolved.html
This process for making matrices for casting large types has been reconstructed by James Mosley and Stan Nelson. Insofar as both use lead "intermediate strikes," this method is a variation on the method of Brass matrices from lead intermediate strikes (cast-brass punches) known from at least the 16th century.
The problem that both "lead intermediate strike" methods address is that driving a very large punch into a copper planchet is not practical with hand striking techniques. This method, as reconstructed by Mosley, involved first cutting a very large steel punch conventionally. Rather than using it to drive a matrix in copper, however, it was used to produce a strike in (much softer) lead. The lead "intermediate strike" (my term) involved in this method was not (to judge from surviving examples) of sufficient mass to have been used itself for casting as was sometimes done with heavier lead matrices. Instead, it was used as a pattern for sand casting a brass matrix (using conventional sand casting techniques). The advantage of the steel punch method over the cast-brass "punch" method was that the steel punch could then be used to "clean up" or refine the cast brass matrix by striking it into the casting cavity of the cast brass matrix. This brass matrix, finally, was used to cast type.
Mosley finds evidence for this method as early as the well-known "Fell" types imported from the Netherlands into England in the early 1670s. It may well have been practiced earlier. See: James Mosley's "Typefoundry" blog posting "Big Brass Matrices: a Mystery Resolved?" (2007-03-19): http://typefoundry.blogspot.com/2007/03/brass-matrices-mystery-resolved.html , and also its followup, "Big Brass Matrices Again: the Enschedé 'Chalcographia' Type" (2008-03-21): http://typefoundry.blogspot.com/2008/03/big-brass-matrices-again-ensched.html
The use of steel punches to "clean up" the cast brass matrices has both been confirmed in the literature (by Ambroise Firmin-Didot, writing in 1851; Mosley translates what he says of this) and in actual practice by Stan Nelson. Here is a snapshot, taken at the 2010 Conference of the American Typecasting Fellowship held that year in Piqua, Ohio, of a large punch (my recollection is 72 point, but I could be in error), lead intermediate strike, and brass matrix made by Stan Nelson. This photograph does not do Stan's work justice - it is breathtakingly beautiful.
At present I know of these only from a brief reference in a blog posting by the noted English type historian James Mosley. "Big Brass Matrices: a Mystery Resolved?" in his blog "Typefoundry" on 2007-03-19, at: http://typefoundry.blogspot.com/2007/03/brass-matrices-mystery-resolved.html Mosley notes that this practice was known in Germany, "and there are surviving examples of such leaden matrices in the Norstedt Collection in Sweden." He does not give a date, but the context of his remarks suggests that this was prior to Fournier (1764).
The "sanspareil" matrices were composite matrices for large types, cut directly by hand. The letterform was first cut out of a plate of brass with a saw; it was cut right reading (as are all matrices). The parts of this pierced plate other than the letterform itself were then riveted to a backing plate to form the matrix.
The term "sanspareil" would appear first to have referred to a specific group of types offered in characteristically large sizes (the concept of a typeface did not exist at that time). In the auction description of the Fann Street Foundry when sold by Thorne and purchased by Thorowgood (1820) the matrices were categorized as "Roman and Italic ... Black ... Shaded ... Flowers ... Ornamented ... Egyptian ... Script ... Engrossing ... German ... Two-line Letters ... and Sanspareil ..." (Reed, p. 295). It seems later to have been applied generally to any matrices made in this manner, regardless of the style of the type they cast. Thus Millington (see below) illustrates a 24-line Pica Antique sanspareil matrix. In this case "Antique" is the classification of the type (like "Egyptian" or "German") while "sanspareil" designates the matrix technology ("24-line Pica" designates the size; this works out to a nominal 288 points).
Thomas Curson Hansard, in Typographia (London: Baldwin, Cradock, and Joy, 1825) attributes the invention of this kind of matrix to William Caslon IV (p. 354-355). He describes it as "the greatest improvement in the art of type-founding that has taken place in modern times." Talbot Baines Reed, in A History of the Old English Letter Foundries (London: Elliot Stock, 1887), p. 327, agrees (though he finds Hansard's description extravagant). Reed dates the Caslon foundry "Circular" announcing these types to Jan 1, 1810. He describes them as "pierced, or rather built-up matrices, in place of the old sand moulds hitherto in use".
However, Ray Millington, in Stephenson Blake: Last of the Old English Typefounders . (New Castle, DE / London: Oak Knoll Press / The British Library, 2002) argues that this attribution to Caslon IV is in error, and that this form of matrix was invented by William Henry Garnett when working for one of the partnerships which later became the well-known foundry of Stephenson, Blake. I confess that I cannot follow this argument, but it does seem a matter worth attention. As noted earlier, Millington illustrates a Blake & Garnett 24-line Pica Antique sanspareil matrix and a type cast from it.
As noted in the section on sand-cast types, above, there is reference in a 1763 text to Cottrell having invented "a new method of cutting letters in brass for monumental and other inscriptions." This involves the concepts of cutting, and of brass, and of letters, but it by no means indicates that he developed the "sanspareil" matrices, as Kelly suggests.
It seems quite likely that from an early period printers employed various methods to make matrices from existing types with which to cast sorts in short supply. Very little is recorded about these "makeshift" matrix making and casting techniques. In the 18th century, Benjamin Franklin says that he did this using matrices made of clay and the printers Peter Miller and Isaiah Thomas (as reported by Thomas) used matrices cast in lead. See the CircuitousRoot Notebook on "Makeshift" typecasting.
All portions of this document not noted otherwise are Copyright © 2013 by David M. MacMillan and Rollande Krandall.
Circuitous Root is a Registered Trademark of David M. MacMillan and Rollande Krandall.
This work is licensed under the Creative Commons "Attribution - ShareAlike" license. See http://creativecommons.org/licenses/by-sa/3.0/ for its terms.
Presented originally by Circuitous Root®
Select Resolution: 0 [other resolutions temporarily disabled due to lack of disk space]