The Varityper (also known as the Vari-Typer or VariTyper) was a highly ingenious "word processor" of the pre-digital age. This machine could use over 300 different type styles and write in 55 languages; it could adjust the space between characters, and even produce right-justified copy. Even though the Varityper enjoyed a successful career of about 60 years, you may never have seen one, for the machine was not generally adopted as a standard typewriter. Instead, it found a niche as a "cold typesetting" or "office composing" machine: it was generally used to produce neat, camera-ready copy for offset printing, at a cost much lower than that of conventional printer's methods.
The Varityper is based on one of the greatest early typewriters, the Hammond. James B. Hammond died in 1913 and willed his patents to the Metropolitan Museum of Art (an act of self-congratulation which, in my opinion, is justified!). The history of the Hammond after that point is a little obscure, but the Hammond Multiplex was produced until the mid-twenties, when the company was bought by the Frederick Hepburn Co. and renamed Varityper. This company failed in the Depression, and it was sold in 1933 to a partnership headed by Ralph C. Coxhead, a business machine salesman. Throughout these convulsions, the company remained headquartered in New York City.
Like the Hammond, the Varityper uses an interchangeable type shuttle, and uses a spring-driven hammer behind the paper to hit the paper against the ribbon and shuttle. The keyboard is a three-bank, double-shift QWERTY arrangement, as on the Universal models of the Hammond. As in the Hammond, the paper is held between two rollers and feeds down into a basket.
The earliest Varitypers look much like Multiplex Hammonds, but they soon went through some important transformations. The biggest step in the modernization of the Hammond design was electrification (introduced 1927), which operated in a very simple way: after every 13 keystrokes, an electric motor in the rear of the typewriter would rewind the spring which hit the hammer against the paper and drove the carriage forward. All other operations were manual. (If you ever type on one of these Varitypers, you'll see that the operation is very quiet until you reach the 13th stroke, when there's a startling noise as the spring is rewound.)
Under Coxhead's leadership the company gradually introduced many more improvements (the changes were said to number 3000). The traditional Hammond vulcanized-rubber shuttles were replaced by metal shuttles. Right-margin justification was achieved by a mechanical system in which each line would have to be typed twice. It sounds tedious, but it was a clever solution to a challenging mechanical problem -- although the results were not comparable to professional typesetting.
By 1937, the standard Varityper looked like the behemoth pictured at the top of this page, with an 18-inch carriage supported by a heavy frame, and lots of new growths -- scales, dials, even an electric lamp -- grafted onto the original Hammond design. This specimen is finished in dark grey crackle paint, but I've also seen them in glossy white paint. The big disk to the left of the type shuttle holds a single-use ribbon, which is essentially a very long strip of carbon paper. This feature was introduced in 1934. Work produced with this ribbon is clearer than work produced with a reusable cloth ribbon, although not as sharp as work produced with modern plastic ribbons. It is also possible to use a conventional cloth ribbon.
This smaller version of the late-1930s Varityper has a short carriage and is not capable of justifying the right margin. A nice feature, however, is a drawer that opens from the front base of the machine, where you can store a dozen type shuttles.
The Varityper's most glorious moment came in 1945, when one of them typed the Instrument of Surrender which was signed by the Japanese on the battleship Missouri. After the war, the machine received further improvements. The design was updated, and the machine was refined with devices such as an automatic forms ruling device and a more precise line spacing device. In 1968, the VariTyper 1010 (above right) offered a powered keyboard--a feature which had always been available on electric typebar typewriters. Varitypers always retained the three-bank keyboard.
An interesting feature of single-use ribbons is that you can read what someone has typed by looking at the used part of the ribbon. The last thing typed on my big Varityper was the address of a funeral home. That was appropriate, I'm afraid, because with the advent of the personal computer, around 1980, the Varityper machine was mortally wounded. However, its life was a good one -- in fact, one of the greatest successes in typewriter history. The Varityper nobly carried forward James Hammond's brilliant ideas, and ensured that these ideas served humankind for nearly a century.
As for the Varityper company itself, it still exists, as a division of AM (Addressograph-Multigraph) International, and it produces sophisticated printers which are very much a part of the digital age.
"I can recall Sales Department pep talks where they would show us enormous blow-ups of IBM's direct impression copy and the carbon spattering that a 20x enlargement would show. Then they'd show us the same enlargement of OUR copy with it's perfect, well-defined characters. Then we'd go out on a service call and repair a tape reader that had just chewed ten feet of paper tape (and a hour or so of work) into confetti. A great example of good technology derailed by a bad business decision.
"The 1010 was the last of the classics, as far as I know (I left in 1970). It had a powered keyboard but if I remember correctly it still rewound the hammer spring every 13 keystrokes.
"One of VariTyper's favorite selling points was their interchangeable type. You could have two fonts in the anvil at a time and could rapidly change a font if you chose. IBM's golfball kinda took the steam out of the interchangeability feature. Also, the VariTyper fonts were a half-moon shape and after hammering on the characters toward the center of the font for a couple of years, the fonts had a nasty tendency to lose some of their curvature. This would cause the characters toward the ends to get a less than sharp look to them as the (now sloppy) ends of the font moved back when the hammer struck. Also, this distortion of the font altered the character spacing. I remember us service guys scratching our heads about the choice of 1010 as a model designation. When the sales types would demo a machine with an older font and type the model number, it would come out '10 10' with the 1 and zero nearly on top of each other and exaggerated space between the zero and 1."
Art Protin writes: "After reading your page on VariTyper, I decided to share some more of its history. I worked there from 1981 to 1985, and followed some of its trials after that.
"When I went to work there, they no longer made a mechanical printing device. Rather the device was a computerized 'composer' which controlled a photographic process of 'flashing' the images of the characters onto photographic paper. During all the time I knew it the output was this paper which would be taken up in a cartridge and then the paper would be cut off, the cartridge move to a processor which would develop the 'film' (only it was paper not film). I saw three different generation of this composer and two different generations of the output engine. The first output engine used an optical disc with all the images of the characters for that font. The disc spun and the mechanism would time the 'flash' through the disc to put the character on the paper. The mechanism would also control optical components that sized and positioned the image. The second generation replaced the optical disc with a computerized CRT on which the character would be drawn and the optical components no longer expanded the character onto the paper but now reduced it. The composer that was being built when I arrived was based on a CPU built in house from LSI and SSI ICs that was upward compatable with the 8080 but faster and with a couple of extended instructions. The second generation replaced this home-made CPU with the Z-80. A third generation went to a multiprocessor arrangement with multiple composer stations driving one output engine. The output engine was unsurpassed for many years, being able to produce 2 point type, and with a resolution of 2400 dots to the inch. (With typographic output, there are two standards: flash & escape, or escape & flash. I do not remember which VT used but it is the opposite of the CAT typesetting engine that was used at Bell Labs for the output of troff. The whole process of producing a document of n characters involves making n images, 'flashing' and positioning the paper/optics, 'escaping', either n-1, n or n+1 times, depending on whether any escapement was needed before the first character, after the last character, or only between characters. Given this relationship, it should come as no surprise that the two operations would be tied together, nor that there might be two orders chosen.)
"The death of VT/Tegra was the shift of work from the back room, where professional 'typesetters' did final preparations for printing, to the front office, where authors took control with troff, LaTeX, WordPerfect, etc., and output came from 600 dpi laser printers. I saw this coming, but could not explain/show the issue to management at VT. It was as described in 'The Innovators Dilemma', no one at VT could get interested in that new market because none of their existing customers had any interest in it. It was not the changes in the technology per se that killed VT, it was the changes in workflow enabled by the technology that killed them."
I heard from a former actuary for the Varityper Plant Pension Plan:
"The plan was, of course, frozen and consisted of about 325
participants about half of whom had retired and half who were waiting
to collect benefits when they reached retirement age. The plan sponsor
finally divested itself of responsiblity for the plan, when it was
taken over by the Pension Benefit Guaranty Corporation (PBGC) on March
13, 2008. The people I dealt with attempted to discharge their
responsibilities in a thoroughly honest and responsible manner. As best
I can tell, Varityper became a corporate 'orphan' and the folks I
worked with attempted -- again, in the most responsible and above-board
manner -- to discharge their responsibilities until such time as the
PBGC would agree to accept the plan under their trusteeship."
Update (January 2012): Jim Cunningham writes: "I worked for
VT. for about 26 years, 1967 - 1993. I do have good memories of
VT., particularly the 'old Varityper,' before Tegra. It was much
like a big family in those days.