of the siphon dips into a pot of ink, and the other end hangs close above a moving strip of paper. The signals are so feeble that if the end of the glass siphon rested on the paper it would not move at all, but by causing the siphon to vibrate continuously against the paper the free motion of the siphon is not interfered with, and the ink is spluttered upon the paper so that the siphon traces a line of very fine dots and thus records the signals transmitted through the cable. This instrument, though crude at first, has gradually been perfected. It is now the most important part of modern cable appa

ratus.

The Duplex System.-The next improvement, undoubtedly the greatest ever made for increasing speed, was the invention of a successful system of "duplexing" cables by Dr. Alexander Muirhead and Herbert Taylor in 1875. This invention rendered it possible simultaneously to send messages both ways through a long ocean cable. In 1878 the Direct United States cable across the Atlantic was successfully duplexed, and a speed of 16 words a minute obtained each way at the same time. Duplexing cables has now become such a fine art, chiefly through the labors of Dr. Muirhead, that the capacity of cables, and therefore their commercial value, has been practically doubled. Since 1875 about 80,000 miles of ocean cables have been duplexed almost entirely on the Muirhead system.

The increasing traffic across the Atlantic and the pressure of competition led next to an increase in the size of the copper "core" which conducts the electric signals. The resistance of a wire delays the electric current and therefore the speed. By doubling the size of the copper core the resistance is halved and the speed greatly increased. The copper wires used for telegraphy on land weigh about 200 pounds per mile. In 1894 two cables were laid across the Atlantic, one for the Commercial Cable Company and the other for the Anglo-American Company. The copper core of the former weighed 500 pounds per mile, while the latter weighed no less than 650 pounds per mile, or as much as three ordinary land wires. The result was that the speed obtained with these two cables was as high as 40 to 50 words a minute, or, working duplex, from 80 to 90 words a minute. On previous Atlantic cables 25 to 28 words a minute was the maximum each way. Owing to the reduction of rates the benefit of this tenfold increase of speed since the early days has gone almost entirely to the general public.

Automatic Transmission.- The increase in speed brought up another difficulty. No human operator can send so fast. The key used for signaling through cables by hand is practically the same as the ordinary Morse key used for land telegraphy, except that two keys are used side by side, one to send positive signals and the other negative signals, the letters of the alphabet being indicated by various arrangements of the two kinds of signals. First-class cable operators can send as many as 30 words a minute for a few minutes, but a sustained speed of 20 words a minute, when working by the hour, is regarded as very good. To take full advantage of the speed of a modern Atlantic cable therefore, it is necessary to have some automatic method of transmitting. The advantages of automatic transmission are higher speed, greater

uniformity of signals, more legibility, and fewer mistakes.

-

The method adopted is simple and beautiful, a modification of the Wheatstone system. The message is first punched as a series of holes in a paper tape. This perforated tape is then run through an automatic transmitter, and by means of a system of small levers the required signals are transmitted at any desired speed. The operator has a wooden stick in each hand with which he strikes one or other of the three keys of the small perforator directly in front of him. One key punches a right-hand hole, another key a left-hand hole, and the middle key makes a space. In this way the cablegram before him is transmitted at the rate of about 20 words a minute into a perforated tape. From the perforator the tape runs into an automatic transmitter, or "auto." There is a row of small central holes in the tape, and on each side is a row of larger holes. The latter represent the message. A small star wheel in the "auto" engages with the central line of holes and feeds the tape along at a uniform rate. A couple of small steel rods about the size of a knittingneedle, one for each of the two rows of message holes, continually vibrate against the paper. When either of them enters a perforation in the paper, a lever connected with it moves and makes an electric contact, sending a short, sharp signal into the cable.

Cable Relays Recently several still more wonderful inventions have been perfected. There is good reason to believe that it is now possible to work a typewriter in New York by playing on a typewriter keyboard in London, and vice versa.

of the series has three keys. These have to be The little tape perforator in the first machine struck on the average four times for each letter, and much practice is required to become skilful in using it. Several tape-perforators with ordinary typewriter keyboards have been invented. The success of a machine of this kind will mean that cable messages will be transmitted by simply playing on a typewriter keyboard, the striking of the keys perforating the transmitting tape, which then runs through the "auto," which sends signals through the cable to the other end, where they are written in ink by the siphon recorder. It is at this latter point that has lain the great difficulty that has baffled cable inventors for years. By the time that an electric signal has passed through a long section, say 1,000 miles, of ocean cable, it has become so feeble that it can be recorded only by the extremely delicate mechanism of the siphon recorder. It has not been possible, until recently, to retransmit automatically into another section. On land, relays are used. For instance, messages from New York to Chicago are automatically repeated at Buffalo or Meadville, and by automatic repeating every 600 or 800 miles it is an every-day occurrence to telegraph direct between New York and San Francisco. A relay capable of performing similar work for cables has been a dream of cable engineers and inventors for years, and in default of such an instrument "human relays" have been employed; that is, at the end of one section of a cable an operator takes the paper record of a cablegram as it comes from the siphon recorder and retransmits it.

a

But the cable relay is now an accomplished fact. The only hope of constructing such an instrument was to utilize the siphon recorder. One difficulty has been that the movements of the siphon, as shown by the paper records, have till recently been most irregular. There has been what photographers would describe as "lack of definition" about the signals, rendering it hopeless to attempt to relay them automatically by machinery. The first thing to do was, therefore, to straighten and sharpen up the signals a bit, and a very able group of cable engineers, including H. A. C. Saunders, electricianin-chief of the Eastern and its associate cable companies, his assistant, Walter Judd, with Dr. Muirhead, inventor of the cable duplex, and Messrs. Brown and Dearlove, succeeded in sharpening them. They secured very regular signals, usually described as "square signals." This result was obtained by means too technical to be described here, but the chief device used is known as an "inductive shunt." Having squared the signals, it was now possible, though by no means easy, to construct cable relay. Two have recently been perfected. One is known as the Brown & Dearlove relay, the principal inventor of it being S. G. Brown. The other has been invented by Dr. Muirhead. In both a fine wire terminating in a platinum contact-point takes the place of the ink in the siphon of a recorder. The contact-point, instead of resting on the paper tape, rests on a rapidly moving metallic surface divided into two parts. In the Brown & Dearlove relay this contact-surface consists of a constantly revolving metallic drum or wheel. The siphon, with its wire and contact-point, "skates," as the inventor describes it, with the utmost freedom on the periphery of this wheel. The drum looks like a phonograph cylinder. As the siphon skates upon the right or left half of this drum it makes a positive or a negative electric contact and automatically transmits a corresponding signal with renewed energy into the next section of cable. In the Muirhead relay the moving metallic surface consists of a small plate vibrating rapidly. The result is the same. Able in this way to make definite electrical contacts through a long ocean cable, an operator can easily work, by means of these contacts, local apparatus moved by more powerful currents. In this way both Mr. Brown and Dr. Muirhead have devised perforators which reproduce at the receiving station perforated tape identical with that used for transmitting the message at the sending station. This tape is available for retransmission through an "auto," this plan having the advantage that the signals are retransmitted in as perfect form as the original signals; and, theoretically at any rate, the process may be repeated indefinitely, so that it would be possible to send a cable message automatically through a dozen stations from England to Australia. This will no doubt be done in time, but it is a very slow process getting such complicated and delicate inventions into commercial use. It is a question of time and growth. The Brown & Dearlove relay has been adopted by the Eastern Company, and has been in commercial use for some months at Mediterranean stations. Dr. Muirhead's relay has also proved very successful in several longdistance tests.

From this description of cable relays it will be seen that an operator, by playing on a typewriter keyboard in London, can now produce a perforated tape in New York. A machine invented by the writer of this article is so arranged that by simply turning a handle it works a typewriter automatically under the control of a perforated paper tape, something after the fashion of a mechanical piano, at a speed of 90 words a minute. In order that this machine may print messages from a perforated tape produced by the cable relays it is necessary that all the letters shall be of equal length, and the writer has devised a new cable alphabet that not only fulfills this condition, but is also about 12 per cent shorter than the cable alphabet at present in use.

Hence, it is now possible, at any rate theoretically, automatically to typewrite a cable message across the Atlantic in page form at a speed 12 per cent faster than the cables can at present be operated. More than this, by the same mechanism it is feasible to operate a linotype or typesetting machine automatically, so that the fantastic possibility presents itself of playing on a typewriter keyboard in London and setting type automatically in New York. DONALD MURRAY.

Salem, Mass., 3 Dec. 1751; d. Boston, 18 April Cab'ot, George, American statesman: b. 1823. He was educated at Harvard College. In 1791 he became United States senator for Massachusetts, and proved a steadfast friend of the Washington administration. He yielded essential aid to Hamilton in perfecting his financial system. In 1814 he was chosen a delegate to the memorable Hartford Convention, and was elected president of that assembly. See Lodge, Life and Letters of George Cabot (1877).

Cabot, James Elliott, American biographer: b. Boston, Mass., 18 June 1821; d. 16 Jan. 1903. He was the friend and literary executor of Emerson and in 1887 published A Memoir of Ralph Waldo Emerson,' a work undertaken at the request of the Emerson family.

Cabot, John, or Giovanni Cabota (in the Venetian dialect, ZUAN CABOTA), an Italian navigator in English employ; the discoverer of the continent of North America. On 5 March 1496, he was given by Henry VII. of England letters patent authorizing him to take possession of any countries he might discover. Under this charter, in May 1497, he embarked in a single vessel, accompanied by his son Sebastian, and sailed west, as he said, 700 leagues, when, on 24 June 1497, he came upon land which he reported to have been a part of a continent, and which he assumed to be in the dominions of the Grand Cham. A letter of that year represents him as having sailed along the coast for 300 leagues; he landed, but saw no person, though he believed the country not uninhabited. He