curred to him that he could devise a machine to perform this work much more neatly and quickly. He discussed this project with a friend, Samuel W. Soule, like himself a printer, and a man of decided ingenuity. They began work at once in a small room on an upper floor of a mill owned by Henry Smith, an old friend. This two-and-a-half story building, in simple ashlar, stood on a narrow strip of land between the Milwaukee River and the Rock River Canal. Here, day by day, Sholes drew his plans with Soule's aid, and here their model gradually took form, proving to be a thorough success in a final test. On the same floor of the mill was the workshop of another tenant, Carlos Glidden, the well-to-do son of a retired ironmonger. Glidden was an inventor, too, and he was developing a spader which he believed would outdo the work of any plow on the market. Naturally, there arose many a colloquy betwixt the three inventors regarding their plans, with much debate of the weak points disclosed as their experiments followed one another.

Sholes and Soule duly patented their numbering machine on November 13, 1866. Shortly afterward they showed it to Glidden, as it turned out capital work at a pace far outstripping that of manual labor at its best, and with infallible correctness. Glidden exclaimed: "Sholes, why cannot you build a machine to print letters and words as perfectly as these figures are struck off here?" This query had doubtless often been put to other inventors, but now it was asked of the man who was to give it a triumphant response. But not at once, although the idea took firm root in Sholes' mind, and kept him on the lookout for any information that would serve his turn. He who seeks, shall find. In July, 1867, Sholes came upon a description, in the Scientific American, of a writing machine for which a great deal was claimed. It had been exhibited in London by its inventor, John Pratt, of Centre, Alabama. Its de

scription was accompanied by an editorial prophecy since fulfilled in all but its closing words: "A machine by which it is assumed that a man may print his thoughts twice as fast as he can write them, and with the advantage of the legibility, compactness, and neatness of print, has lately been exhibited before the London Society of Arts, by the inventor, Mr. Pratt, of Alabama. The subject of typewritting is one of the interesting aspects of the near future. Its manifest feasibility and advantage indicate that the laborious and unsatisfactory performance of the pen must, sooner or later, become obsolete for general purposes. Legal copying, and the writing and delivering of sermons and lectures, not to speak of letters and editorials, will undergo a revolution as remarkable as that effected in books by the invention of printing, and the weary process of learning penmanship in schools will be reduced to the acquirement of the art of writing one's own signature, and playing on the literary piano above described, or, rather, on its improved successors."

Pratt's machine struck Sholes as complicated and liable to get out of order. He believed that he could devise mechanism more simple, and at least as efficient. Soule had been a helpful partner in the numbering machine, a success from the start; would Soule embark with him in this second project? Yes. Glidden, who had given Sholes his first push from the shore, was received as a third partner: he was to contribute the necessary funds. A conference was held as to plans, which were sketched in a preliminary way. First of all a writing machine must write, but how was its paper to be imprinted? Soule suggested the scheme, never excelled, of placing convergent typebars on the rim of a circle, so that each might strike the center. Whether this design was original with him, or borrowed, is not to be ascertained at this distant day. It first appeared in the writing machine of Xavier Progin, in 1833; it presented itself

again in the embossing machine of Alfred E. Beach, in 1856. Other inventors had gone astray in sliding their typebars through a horizontal circle, rotated on a vertical axis, as Charles Thurber did, in 1845. When an operator wished to print "A" he turned the ring until "A" stood over the printing point. He then depressed the "A" typerod so as to leave "A" printed on the paper beneath. This mechanism, much too slow for business, survives in toy machines.

And yet the Thurber design, faulty in the disposal of its typerods, displayed a feature of cardinal value; its paper was borne on a cylindrical carriage, or platen, and this Sholes adopted in his second model. It remains to this hour an indispensable part of every standard machine. Sholes devised the letters, all capitals, a spacer, and other details equally important. But no one of the three partners undertook any systematic inquiry as to what their predecessors had done, so they troubled themselves to devise novelties which worked badly, when they might have laid hands on old contrivances that worked well. In their first model Sholes built a keyboard resembling that of a piano, with two rows of keys:

3579 N O P Q R S T U V W X Y Z

2 4 6 8. A B C D E F G H I J K L M

He did not know that Dr. William Francis, of New York, in his remarkable machine of 1857, had introduced keys of the peg form now universal, and arranged them in four rows so as greatly to shorten the journeys taken by an operator's fingers. Sholes at length abandoned his piano keyboard at the instance of his model-maker, Matthias Schwalbach, a builder of tower-clocks in Milwaukee. As we have just seen, Sholes in his first keyboard gave his characters a strictly alphabetical and numerical order. He soon changed this for the present order of disposal which,

like the compartments of a printer's case, places the characters oftenest used nearest to the working center. As patented on July 14, 1868, the claims of Sholes, Glidden, and Soule were: (1) A circular annular disc, with radial grooves and slots to receive and guide the typebars so that they struck the center. (2) Radial typebars to correspond with this disc. (3) A ratchet to move the paper-carriage by the breadth of a tooth when a key was struck. (4) A hinged clamp to hold the paper firmly on its carriage.

Frederick Heath, of Milwaukee, as a lad was employed as a messenger by Mr. Sholes as he began to devise his typewriter. On the wall of Mr. Heath's office he has framed a rough, uncouth model of the first machine invented by Mr. Sholes. "His original idea," says Mr. Heath, "was to have his keyboard fashioned after that of a piano, and there you have it. The first row is of ivory, duly lettered; the second row is of ebony; and then, as you see, a third row, made up of letters and characters that are little used, is in the form of pegs. The framework is of wood, with the leverage below, and the basket form of typebars above closely resembles those of some machines in use to-day. The original model was very clumsy and weighty. The writing was on a tape of tissue paper, and the platen was fastened to the body of the boxlike affair. The writing could not be seen till it was completed, and when the document was once removed from the machine there was no way by which it could be replaced with any degree of certainty that the lines would correspond with those previously written.

"Mr. Sholes was collector of customs of the port of Milwaukee during most of the time that he was engaged in devising his typewriter, and later he was Comptroller of the city of Milwaukee. While acting in this latter capacity, it fell to his lot to enter into a contract, on behalf of the city, for the paving of certain streets. He had the contract

FIRST PATENT, SHOLES, GLIDDEN, AND SOULE, JUNE 23, 1868

Key-levers, L, vibrating on the fulcrum, M, with the inner fingers, u, reaching under the typebars, so that the keys act directly on the types.

The spacer or ratchet, I, combined with the bifurcated lever, H. connected with the bar, T, pivoted at s and resting across the arms of the keys, L, so that striking the key-faces will work the teeth of lever-forks up and down and into the notches of the spaces, so as duly to move the paper-carriage.

The pins, e, fastened to the table A', combined with the pawl, h, and the spring, ', to give the paper-carriage a certain and regular cross-line movement at a right angle to the space movement from line to line.

The spring-clasps, b, attached to the bars, C and C', on a line through the middle of the platen, G, combined with the springs, a, attached to the bar, E, hold the paper on its carriage smoothly and tightly.

The spools, m, combined with the gudgeon, s', the shaft, 7, the pulleys, and R, the band, v', the cord, v, the weight, W, the ratchet-wheel, V, the pawl, t, and the bar, P. pivoted to the back of the case, A2, feed a fresh part of inking ribbon to each type successively.