The Machine Was Done Again

A sentence beginning with T could become a whole row of Ts. The first typebar had stuck at the printing point, and the other letters were striking its back. Charles Edward Weller remembered having to prise tangled bars apart before work could continue.

Weller was a shorthand reporter trying an experimental writing machine in St Louis in 1868. He had known its maker, Christopher Latham Sholes, in Milwaukee, Wisconsin, where he had worked for Western Union. The machine offered a way to turn rapidly taken notes into a readable transcript without writing the whole thing out again by hand.

There were jobs on which appearance mattered less than others. Weller recalled using it to prepare a transcript of an impeachment trial for the printer. The printer needed copy to set in type; a court expecting a finished record would have demanded neater work. Even in its rough state, the machine did a considerable amount of that job.

In his account, published in book form in 1918, Weller was careful about the failures. They were intermittent. When everything ran properly, the speed impressed him. When it did not, someone with a transcript to deliver had to stop being a writer and attend to the mechanism.

A workshop in Milwaukee

Sholes already knew the trade of putting words into print.

He had been a printer and newspaper editor before becoming Milwaukee's collector of customs. In the machine shop run by Charles F. Kleinsteuber, he worked among men with other mechanical projects of their own. Carlos Glidden and Samuel W. Soulé joined him in developing the writing machine. Matthias Schwalbach, a mechanic, helped turn successive arrangements into working parts.

By the summer of 1867 Weller had seen a device with a single letter. A key sent a piece of type against carbon paper and made an impression. Extending the principle meant giving the alphabet separate moving parts, all capable of reaching one printing point, then moving the paper exactly far enough for the next letter.

The patent granted to Sholes, Glidden and Soulé in June 1868 shows an arrangement very different from the decorated production machine. The paper lay in a flat frame above a circular assembly of typebars. The keyboard resembled a small piano. Beneath it, levers and connections translated the movement of a finger into a blow against the paper.

Moving the paper required its own restraint. A hanging weight supplied the pull; an escapement, a mechanism releasing movement in small steps, metered it out. The paper had to remain still during the impression and advance a letter-space as the key and type returned. If that control failed, Weller recalled, the carriage could jump an inch or two, even half a line.

His early machine sometimes needed more pull than its weight supplied. A jackknife, a paperweight or a pair of shears could be added to the load. Too much could break the cord. Weller's account included a practical warning about keeping toes out of the way.

The ink supply involved another kind of preparation. When a ribbon wore out, he bought silk or satin ribbon from a dry-goods shop, immersed it in black ink in a washbowl, then strung it over chairs and other furniture to dry overnight.

An entirely new machine

Sholes kept writing to Weller about improvements. In April 1870 he began a letter with Nil desperandum, followed by his own translation: don't despair. Although he had already had the machine done, he was still making valuable changes.

“You had better have an entirely new machine.”

He had reduced two spacing wheels to one and connected the weight directly to the printing shaft. The revised machine, he said, would run thirty lines without winding. Another key made paragraphs. There was little point in rebuilding Weller's old machine, with its limited set of characters, when a new one could give him more.

In July Sholes announced completion again. In September he described another major change. The slotted disc holding the typebars had gone. Each bar now swung on an independent steel bearing around an open ring. A brush could reach in to clean the type, and a hand could get at anything needing attention. The large type he had used for the experiment was simply a set he already had available.

Weller received a replacement in the autumn of 1870. The paper was now held around a rubber roller, and the machine looked so different that he would scarcely have recognised it as a relation of the old one. He used it for ordinary correspondence. For paid reporting, breakdowns could still consume too much time; when deadlines pressed, he returned to writing out his shorthand notes by hand.

James Densmore had joined the enterprise as a financial backer and pressed its makers to keep developing it. With the promoter George W. N. Yost, he took the machine to the Remington works at Ilion, New York. By April 1873 Sholes could report a contract to manufacture a thousand machines. He also described a model developed to print from a continuous roll of paper for the Automatic Telegraph Company.

Printing underneath

Remington manufactured guns and sewing machines.

Its mechanics had to make the writing machine reproducible in a factory. Jefferson M. Clough and William K. Jenne worked on that task; Jenne came from the sewing-machine department. Manufacture began in 1873, and the Sholes & Glidden was offered commercially in 1874.

Their improvements extended to small parts under the keys. In a patent filed in 1875, Clough and Jenne described light wooden key-levers strengthened with thin strips of paper glued to their sides. Pine kept the moving part light, but could split. The paper reinforced it without appreciably increasing the weight. Glass keytops covered printed characters held in metal cups.

The same specification proposed spring-steel typebars that could bend sideways in a collision and return to position when freed. Colliding bars were still something for a maker to accommodate.

The finished machine printed capitals only. Its typebars swung upwards against an inked ribbon beneath the paper and roller. The impression was made out of sight. To inspect a fresh line, the operator lifted the carriage, bringing the underside of the roller and its writing into view. Early machines on stands also had a foot treadle for returning the carriage; the fingers still worked the letter keys.

Weller remembered a different arrangement on the early flat-frame machine: the thin paper was struck against a ribbon above it, so the inked impression appeared on its upper surface. Putting the ribbon directly next to the type, with paper wrapped around a roller, allowed thicker paper to be used. That improvement came with the hidden printing point.

A closer look · 1:23 · English · automatic English captions

Lifting the carriage to read

Play film

Keeping the commercial machine moving also meant daily attention. Instructions recorded on a surviving Sholes & Glidden at The Henry Ford tell its owner to keep dust out and wipe the top rods, especially the shifting rod, with a greasy cloth every day.

A familiar row of letters

The keyboard was changing alongside the mechanism. The early piano-like keys gave way to rows of round buttons. On the commercial machine the upper letter row begins QWERTY, but the whole arrangement is not quite today's: C and X exchange places, M sits beside L, and the letter keys also supply the missing numerals 1 and 0.

The machine's troubles do not establish that the letters were deliberately arranged to make a typist slow. Researchers Koichi and Motoko Yasuoka compared surviving machines, keyboard drawings and early typing methods, challenging that familiar explanation. Their reconstruction emphasises the requirements of telegraph operators as well as changes made by inventors and manufacturers. Some proposed intermediate key positions remain inferences. What the successive arrangements do establish is that the familiar keyboard emerged through revisions, rather than appearing complete on the first patent drawing.

A beginner in Hartford

In December 1874 the American writer Samuel Clemens, better known as Mark Twain, tried his new machine in a letter to his brother. Its uneven beginnings survive on the page: stray characters above the address, repeated letters in the first sentence, and words running together. He was trying to get the hang of it. This was his first attempt, he said, and he expected to acquire facility soon.

He listed advantages. He could lean back in his chair; the machine put a great many words on a page; it did not scatter ink blots. He hoped it would print faster than he could write. Susie had pressed the keys once or twice, he added, and had no doubt contributed letters where they did not belong.

By March 1875 the new instrument had produced a different inconvenience. Twain asked that his name not be used to advertise it. He had stopped using it because a typed letter brought a request for information by return post. Answering the questions took more trouble than handwriting the original letter. He called it a “curiosity-breeding little joker”.