ATF's Machine & Hand Engraved Working Patterns

1. Problems Ahead

The identification of this method of making working patterns involves several concerns. Here is my current understanding of it (in my own words):

Between 1899 and 1905, for use in direct matrix engraving only, ATF used a method in which a drawing was traced pantographically and reduced 25:1 to a working pattern. In the first version of this method, the working pattern consisted of a lead plate of greater thickness than initially necessary. The pattern itself was "incised" into this lead plate as an outline (whether by drag or rotary engraving I do not know); the area "between the lines" was not cut away and remained at the same height as lead plate blank. This pattern was followed in matrix engraving first by using a pointed stylus to follow the incised outline and cut a corresponding outline in the matrix blank, and then by using a flat-bottomed stylus kept by operator skill between the incised lines on the working pattern. This operation wore the pattern sufficiently that after a single use its surface was shaved away and the process of making and using a working pattern repeated until the working pattern blank was too thin to use again. In a later version of the process a zinc plate capable of repeated use was used instead of a lead plate.

But this seems to be a needlessly difficult and inefficient method, and its description raises more questions than it answers.

2. Other Identifications of Benton's "First" Method

The earliest reference of which I am aware which can be considered authoritative concerning ATF's early working pattern making techniques is that of Linn Boyd Benton himself in his chapter "The Making of Type" in Frederick H. Hitchcock's The Building of a Book. (NY: The Grafton Press, 1906) In it, Benton describes a wax-plate (over typemetal substrate) and electroforming method.

In Typographical Printing Surfaces (1916) , Legros and Grant make two references to "Benton-Waldo" working patterns ("formers"). They mean, I believe, actual Benton-Waldo machines made by either Benton, Waldo & Co. or ATF (Legros & Grant manufactured a competitor to this machine, and were not likely to employ "Benton-Waldo" as a generic term for any vertical pantograph, as has sometimes been done.)

First, they say "The first formers [that is, working patterns] for the Benton-Waldo machine known to the authors, and indeed, the formers still generally used in that machine, are produced by electrotyping in the following manner." (p. 208; they go on to describe wax plates on a typemetal substrate from which electroformed working patterns are made, just as in Benton's 1906 description).

Slightly later, when referring to the methods of several early matrix making companies (Monotype, (Rogers) Typograph, Monoline, Victorline) they say that these companies are using "the old Benton-Waldo electrotyping methods". (p. 209)

It may be significant that they use the word "shaving" in conjunction with the making of these wax-plate patterns. First they note: "Type-metal plates of equal and uniform thickness are coated with a wax composition which is shaved off on a machine to the thickness required for the raised portion of the letter." (p. 209; presumably these are relief patterns, which would be correct for use in cutting punches with a Benton vertical pantograph). Later they say that "The burr on the wax [raised by drag engraving the wax] is dressed off on the shaving machine..." (p. 209)

3. Rehak's Account

But in Practical Typecasting, p. 108, Theo Rehak describes a very different process. I've attempted to summarize my understanding of it earlier, but here is what he actually says:

"The patterns originally used by the Benton engraver for cutting punches or type characters on blank types bodies [1] had to be raised (i.e., in relief). Later, the raised offset patterns were discarded for right-reading incised ones, first cut by tracing the design on a flat square of lead and firmed up by hand engraving. The Benton Delineator was used for this purpose (no trace of this device has survived); it used a traced working darwing with a 25:1 ratio. After each use the thickness of the lead plates was reduced by shaving away the old surface and a new one was prepared. When the plate was too thin to use further, it was discarded." (p. 108)

This account (along with the next paragraph in Rehak, not quoted in full here) raises several questions:

First, this is the only instance I have been able to find of any mixed machine-engraving / hand-engraving method. It would be good to have an independent confirmation of such an unusual method.

Second, it dates this method to the introduction of "incised" right-reading patterns. These would have been appropriate for using the third form of Benton's vertical pantograph as a matrix engraving machine. Benton was using the earlier form(s) of his vertical pantograph for patrix and punch engraving from at least 1884 (probably 1883), but did not begin using the later form for direct matrix engraving until 1899. What method was he using for the first 15 or more years when raised patterns were required?

The duration of use of this method of incised lead plates (and later zinc plates; see below) seems very short: no earlier than 1899, but no later than 1905 (Rehak's date; certainly no later than Benton's 1906 chapter).

Third, Rehak's account continues with a description of the basic problem with this method: The working pattern had only "incised" lines on it, which were followed easily enough with a "pointed stylus" (tracer) for cutting the matrix. But when it came to cutting away the bulk of the matrix betwen the incised lines, this was done with a "flat surfaced" follower on the pattern. Only the skill of the operator kept this operation "between the lines." Rehak notes that this problem disappeared with the introduction of the wax-plate method "by 1905." (In wax-plate methods, the wax "between the lines" is removed and an electroformed working pattern made which presents an intaglio pattern of the entire letterform.)

This makes sense so far as it goes, but it raises a question: Why didn't Benton simply rout out the lead working pattern between the lines, as (say) Goudy did a few decades later? Keeping a cutter between the lines by operator skill alone is hard (= expensive in industrial terms). Doing this hard process once for a working pattern that would be re-used for many matrices makes sense. Doing it for every matrix does not. It seems to me that Benton was too smart for this.

Third, the description of the "lead" plates as single-use doesn't make sense. Rehak continues, after describing the shaving away of the lead plates after one use: "Later, zinc plates were incised and used, replacing the lead ones. Zinc could be stored more easily than the lead plates." (p. 108)

Yet the working patterns of Goudy and Rimmer were cut in typemetal, and while they might have been as durable as electroformed plates, they certainly weren't single-use.

The shaving away of the lead plates also doesn't make sense. Casting a lead plate is a trivial operation for a typefoundry. Shaving a plate away is more expensive, and also implies that in its original size the plate is thicker than necessary.

Fourth, the 25:1 ratio seems too large. The two original ATF working patterns in my possession (wax-plate method patterns) are nearly five inches high. That would imply a finished drawing 125 inches (10 1/2 feet) high. Clearly that was not done. Yet if one assumes, say, a finished drawing about 12 inches high (not uncommon in other practice), the working pattern would be a tiny 1/2 inch high.

Fifth - and I'm not sure if this is a question or an observation - the assertion that the switch from raised to incised patterns required a shift in working pattern technology is puzzling. One can easily imagine a relief pattern in which the full letterform is raised up (and you trace around it); similarly one can imagine an intaglio pattern in which the full letterform is cut away (and you trace within it). The first fits well with Benton punch-engraving, where both tracer and cutter "come to form" from the outside and there is no tracing "between the lines." The second is what, say, Goudy did. But the existing wax-plate method ATF patterns that I have differ from either of these: they consist of a single raised outline, and are right-reading. They may be used to cut punches by tracing on the outside of the raised line with a pantograph which reverses (such as the early punchcutting version of Benton's). They may be used to cut matrices by tracing the entire area on the inside of the raised line with a non-reversing pantograph (such as the later matrix engraving version of Benton's). If the raised patterns used in the first fifteen years of Benton's work consisted of a raised outline (like the later ATF wax-plate patterns) then no change would have been necessary. This implies that they must have been solid raised letterforms. [2]

Finally, there is some question as to the identification of the pantograph used. It may or may not have been Benton's "1899/1905 Optomechanical Pantograph". But while the patent for that machine envisioned its use for cutting, the version illustrated in 1909 in Kaup's article "Delineating the Characters" is a drawing, not engraving, machine. This machine is basically an enlarging pantograph; how it might be used for reducing a working pattern from a larger finished drawing is not at all clear. Kaup also shows a fairly conventional pantograph "Delineating on Wax Plate" (that is, drag engraving wax-plate working patterns). Whether the "Benton Delineator" is one of these machines or a third as-yet unidentified machine isn't certain.

Of course, Rehak may be absolutely correct. After all, he apprenticed at ATF and I did not. But I simply don't understand what is being said sufficiently.

4. Notes

1. That is, patrices in the terminology used in the present book.

2. From a historical point of view, this supports Rehak's account of this method. If it was in fact an intermediate method used for only five years between the introduction of direct matrix engraving and the development of the wax-plate method, then it could easily have been just as awkward as described.