I figured I'd post something on this since, there hasn't been a response yet.
Based upon the fact that the patent number is in English, I took the chance that it was probaby originally from an English speaking country. The British patent for number 4615 is dated 1821 and references a method to destroying smoke in a chimney. The US patent is dated 1826 and is labeled as a machine for making hubs for wheels for carriages. I would speculate that it is the latter as opposed to the former, assuming that either one has any relation to the item shown. The timeframes for both patents would be consitent with the manufacturing appearance and apparent age of the device.
You can see in the closeup that the bar at the top is ruled. There is a strip with teeth below so the device can be positioned in fixed increments. Clearly a device for marking letters either with ink, or for embossing.
From this and previous comments and finding the patent number under the US Patents:
the pinch rollers near the bottom of the picture are for holding a wide strip of leather in place.
the lever on the left of the pinch rollers is to permit the leather to be pushed through the pinch rollers a specific distance (like the height of a line of text).
the indexed bar near the top of the photo is to allow the user of the tool to correctly space the lettering from left to right on the leather.
just below this indexed bar is what looks like a jagged saw blade, like you might see on your box of cellophane wrap in your kitchen. It's likely to cut the leather piece off after advancing it far enough to have all the desired lettering in place.
the the leather piece is later mounted on the wheel to indicate who built the wheel, when, and maybe whose wagon it is for. Maybe even information about how the wheel is constructed so someone else can repair it.
@Duane: I found an image of a very early typewriter. It has a dial for selecting letters instead of the plate-type mechanism in your photo. It had a serrated bar, as does yours.
Some of the early typewriters were ... interesting to say the least. As I wrote in one of my books (so it must be true):
Some of these weird and wonderful contraptions were as difficult to use as a Church organ, while others printed by keeping the paper stationary and hurling the rest of the machine against it – the mind boggles.
If this is that old, is it possibly a manual punch for something like loom control cards? They were functionally similar to the Hollerith punch card. That might be the link to 'modern' computers that used punch cards.
In the enlarged image, you can see characters in the grid so it appears that the stylus is used to select a letter, perhaps moving around a die for embossing. This is similar to the earliest typewriter prototypes and not that different from Dymo embossed label makers. Given the small size of the array, which would result in an even smaller size of the die being moved around by the stylus, it would appear to be for making really small print. Perhaps it's for writing up contracts? :-) Or maybe for the fine print in currency or coins
"It's clearly not QWERTY. But it has 81 characters, both upper and lowercase letters and the digits 1-9 (where is 0?- uppercase O?). It's a Norwegian/Danish keyboard as it includes the letters æ, ø and å (ÆØÅ). You clearly see that the system is that the less used uppercase letters are placed to the left and right, leaving the more frequently used lowercase letters in the middle. It would interesting to study the position of the lowercase letters relative to their overall frequency in the Norwegian language, but I'll leave that to others :-)"
I would have guessed some sort of code generator, but an index typewriter must be close. And for me, Danish would be a secret code. But part of it was missing in the initial picture, which did not help at all. But6 it was fairly obvious in the closeup view that it had something to do with character printing, rather encrypted or not.
Probably the most interesting part is that such a function was fabricated using blacksmith tools.
Blog Doing Math in FPGAs Tom Burke 23 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...