First, let me acknowledge up front that this is a VERY long post consisting of numerous [related] questions I've had for some time about how dollar bills were designed in previous eras. If you can answer even one of these, I'd be hugely appreciative. Also, I'm posting this here because Google has absolutely failed me so far (which is unusual) and Typophile seems to be a good source of people who would know about this rarely discussed topic. If anyone can refer me to some other more appropriate resource for discussing the details of pre-computer typography and engraving, that'd be great too. Thanks!
My basic goal here is to understand—with as much detail as possible—precisely how U.S. paper currency was designed in the pre-computer era. The typography and detail is utterly exquisite, and while I have some decent freehand skills myself, I still can't quite wrap my head around how this was done. For the sake of this conversation, take a look at this:
Not only is this an extremely resolute example, but this gorgeous specimen is from 1891—entire generations before anything we'd call "computer design" was available. So it makes for a good reference of what's possible in the pre-digital world.
1) My first question is regarding the Guilloché patterns, which are the fine geometric, radial designs seen within various elements (like a modern-day Spirograph). Thankfully, I at least understand Guilloché itself: various machines (rose engines, linear engines, etc.) are used to etch these highly accurate patterns into what I'd imagine are separate plates that can stamp each Guilloché element into wherever the overall bill design is being drawn. That much I think I understand. But honestly, a lot of the overall shapes of these elements are so complex that I wonder, is it all being done with a traditional rose engine? I didn't realize such machines could follow such dramatic, tightly-cornered contours. I guess I don't have a specific question here, I'm just looking for more precise information on how a rose engine or similar device is integrated into the overall bill design and layout process, since the two seem to be so tightly and ornately coupled.
2) The biggest mystery to me, of course, is all the curved typography. I'd imagine that a lot of the linear type (or even type set around a circle) is done by placing metal sorts by hand (or something along those lines). So while I'd imagine the average dollar bill artist has amazing hand-lettering skills, I'd still assume they draw from an existing alphabet of sorts whenever possible, for the sake of consistency if nothing else. However, look at the enormous, curved "United States" type set in blackletter with a crosshatch fill inside, that seems less likely.
Not only is the type set around a very smooth, uniform curve, but the hatch pattern inside follows the curve as well. I suppose I can imagine using a compass in very tiny increments to build it up layer by layer, following guide outlines that define the direction and volume of the drop shadow, but I keep thinking there must be some additional tool or technique beyond that. The accuracy, symmetry and general consistency of it all just seems so remarkable. Not quite impossible, but worth asking about.
3) Furthermore, the curved type just above "United States" that reads "THERE HAS BEEN DEPOSITED IN THE TREASURY" is a bit mysterious to me as well. This has a bit more of a hand-drawn quality to it, but I still have some logistical questions about how exactly they pull this off so cleanly. Again, the crosshatch drop shadow is a bit too clean for me to automatically assume it was all done by hand. Also, while I can understand the hand-lettering process itself, and even how it might be done along a curve (a compass should do the trick, really), what about planning out the spacing itself? It all seems pretty well kerned, especially when you examine specific pairs like S-I and I-T in the word "DEPOSITED". Again, I might be answering my own question, but is it simply a matter of measuring the width of each letter and its spacing within the curve, then drawing them one by one? Would this measuring just be done by eyeballing it, or did they refer to some kind of character width reference even in the case of hand lettering? It just seems so easy to make some minor miscalculation that you don't notice until you've sunk an hour or two into the process.
Another reason I ask about 2) and 3) is because if you look closely at the filigree patterns around, say, the "ONE SILVER DOLLAR" element in the center of the obverse, you can see that the illustration quality doesn't have nearly the same level of accuracy as seen in the lettering and cross-hatch details. It's nicely done in its own right, but it very much looks "hand made". The lettering doesn't, at least not to the same extent.
4) How was the layout produced overall? Was this all done by one guru, or did a team break it into individual tasks? Were the elements prepared separately and then somehow merged, and if so, how was that done? Was this remarkably detailed piece all created at actual size, or was it created on a larger canvas and then somehow shrunk down for transfer to the plate? And how would that have been done in the 19th century? Some kind of optical process? Lastly, about how long would a piece like this take from start to finish, and how were mistakes handled/corrected?
In general, I'm clearly ignorant to the details of how such intricate design and typography was done in this era. I think I can intelligently guess my way through a lot of it, but I can't shake the feeling that there are some extra techniques and tricks that would make it all seem a lot more feasible. Sorta like the rose engine—it demystifies so much of this process once you understand it even exists, where before the patterns seemed superhuman for a freehand illustrator.
Lastly, any overall resources I can look at, be they books, message boards, whatever, that discuss this kind of design production in more detail would be excellent. I've really been interested in learning how entirely non-digital design can be taken to these heights, and starting with some solid reference material would be a big help.
Thanks to anyone who managed to trudge through all this! I apologize again for the length of the post, but I think this is a worthwhile conversation that would be of interest to a lot of people if some answers start to materialize.