https://web.archive.org/web/20110409171021/http://www-03.ibm...
It has both the original typewritten scan and a searchable-text version right underneath.
https://web.archive.org/web/20110409171021/http://www-03.ibm...
It has both the original typewritten scan and a searchable-text version right underneath.
Does anybody have any other insights?
Yes, and in fact one of the most popular was from IBM itself [1], released in 1944.
> The letterhead and body text aren’t aligned, so if it did go through a press it took two passes.
It was pretty standard practice to have pre-printed letterhead, hence the cachet of something being issued on "company letterhead". Take a sheet of company letterhead, pop it in the ol' Executive, and type-type-type.
> The signature is also in ink, so that’s either a third pass for color, or an actual signature, and the letter doesn’t have the notation to indicate that it was signed by the secretary, so that leads me to think that it wasn’t widely distributed.
I'm not really sure what potential significance you see in this. It was likely typed by the secretary and signed by the CEO. It's the original copy. Any copies required for the personal reference of the supervisory personnel affected would be made in the standard 1950s ways - a few carbon copies for the top executives, mimeographs further downstream if necessary.
They also charge enough for their services, as senior partners of big firms for it to make no sense for them to do their own typing.
This was indeed standard until colour laser printers became cheap (and physically printing letters became less common), well into the 2000s.
I should probably use it as scrap paper, there's no way it will ever be used for sending letters at the current rate.
https://old.reddit.com/r/typewriters/comments/178vjlf/sample...
There were variable-width font typewriters starting in 1930:
https://old.reddit.com/r/typewriters/comments/3ltlgn/have_va...