When the Calais-Dover and other short cables were first worked, it was found that the ordinary needle instrument in use on land lines was not sufficiently sensitive to be affected trustworthily by the ordinary current it was possible to send through a cable. Either the current must be increased in strength or the instruments used must be more sensitive. The latter alternative was chosen, and the mirror galvanometer was the result.

The principle on which this instrument works may be briefly described thus: the transmitted current of electricity causes the deflection of a small magnet, to which is attached a mirror about three-eighths of an inch in diameter, a beam of light is reflected from a properly arranged lamp, by the mirror, on to a paper scale. The dots and dashes of the Morse code are indicated by the motions of the spot of light to the right and left respectively of the center of the scale.

The mirror galvanometer is now almost entirely superseded by the siphon recorder. This is a somewhat complicated apparatus, with the details of which we need not trouble our readers. Suffice it for us to explain that a suspended coil is made to communicate its motions, by means of fine silk fibers, to a very fine glass siphon, one end of which dips into an insulated metallic vessel containing ink, while the other extremity rests, when no current is passing, just over the center of a paper ribbon. When the instrument is in use the ink is driven out of the siphon in small drops by means of an electrical arrangement, and the ribbon underneath is at the same time caused to pass underneath its point by means of clockwork.

If a current be now sent through the line, the siphon will move above or below the central line, thus giving a permanent record of the message, which the mirror instrument does not. The waves written by the siphon above the central line corresponding to the dots of the Morse code, and the waves underneath corresponding to the dashes.

The cost of the transmission of a cablegram varies from one shilling per word, the rate to New York and east of the Mississippi, to ten shillings and seven pence per word, the rate to New Zealand. In order to minimize that cost as much as possible, the use of codes, whereby one word is made to do duty for a lengthy phrase, is much resorted to. Of course those code messages form a series of words having no apparent relation to each other, but occasionally queer sentences result from the chance grouping of the code words. Thus a certain tea firm was once astonished to receive from its agent abroad the startling code message - "Unboiled babies detested"!

Suppose we now follow the adventures of a few cablegrams in their travels over the world.

A message to India from London by the cable route requires to be transmitted eight times at the following places: Porthcurno (Cornwall), Lisbon, Gibraltar, Malta, Alexandria, Suez, Aden, Bombay.

A message to Australia has thirteen stoppages; the route taken beyond Bombay being via Madras, Penang, Singapore, Banjoewangie and Port Darwin (North Australia); or from Banjoewangie to Roebuck Bay (Western Australia).

To India by the Indo-European land lines, messages go through Emden, Warsaw, Odessa, Kertch, Tiflis, Teheran, Bushire (Persian Gulf), Jask and Kurrachee, but only stop twice between London and Teheran - namely, at Emden and Odessa. Messages from London to New York are transmitted only twice - at the Irish or Cornwall stations, and at the stations in Canada. Owing to the great competition for the American traffic, the service between London, Liverpool, and Glasgow and New York is said to be much superior to that between any two towns in Britain. The cables are extensively used by stock brokers, and it is a common occurrence for one to send a message and receive a reply within five minutes.

During breakages in cables messages have sometimes to take very circuitous routes. For instance, during the two days, three years ago, that a tremendous storm committed such havoc among the telegraph wires around London, cutting off all communication with the lines connected with the Channel cables at Dover, Lowestoft, etc., it was of common occurrence for London merchants to communicate with Paris through New York. The cablegram leaving London going north to Holyhead and Ireland, across the Atlantic to New York and back via St. Pierre to Brest and thence on to Paris, a total distance of about seven thousand miles.

Three years ago, when the great blizzard cut off all communication between New York and Boston, messages were accepted in New York, sent to this country, and thence back to Boston.

Some time ago the cables between Madeira and St. Vincent were out of order, cutting off communication by the direct route to Brazil, and a message to reach Rio Janeiro had to pass through Ireland, Canada, United States, to Galveston, thence to Vera Cruz, Guatemala, Nicaragua, Panama, Ecuador, Peru, Chili; from Valparaiso across the Andes, through the Argentine Republic to Buenos Ayres, and thence by East Coast cables to Rio Janeiro, the message having traversed a distance of about twelve thousand miles and having passed through twenty-four cables and some very long land lines, instead of passing, had it been possible to have sent it by the direct route, over one short land line and six cables, in all under six thousand miles.

Perhaps some of our readers may remember having read in the newspapers of the result of last year's Derby having been sent from Epsom to New York in fifteen seconds, and may be interested to know how it was done. A wire was laid from near the winning post on the race course to the cable company's office in London, and an operator was at the instrument ready to signal the two or three letters previously arranged upon for each horse immediately the winner had passed the post. When the race began, the cable company suspended work on all the lines from London to New York and kept operators at the Irish and Nova Scotian stations ready to transmit the letters representing the winning horse immediately, and without having the message written out in the usual way. When the race was finished, the operator at Epsom at once sent the letters representing the winner, and before he had finished the third letter, the operator in London had started the first one to Ireland. The clerk in Ireland immediately on bearing the first signal from London passed it on to Nova Scotia, from whence it was again passed on to New York. The result being that the name of the winner was actually known in New York before the horses had pulled up after passing the judge.

It seems almost incredible that such information could be transmitted such a great distance in fifteen seconds, but when we get behind the scenes and see exactly how it is accomplished, and see how the labor and time of signaling can be economized, we can easily realize the fact.

The humors of telegraphic mistakes have often been described; we will conclude by giving only one example. A St. Louis merchant had gone to New York on business, and while there received a telegram from the family doctor, which ran: "Your wife has had a child, if we can keep her from having another to-night, all will be well." As the little stranger had not been expected, further inquiry was made and elicited the fact that his wife had simply had a "chill"! This important difference having been caused simply by the omission of a single dot.

 -.-. .... .. .-.. .-..

c h i l l = chill

-.-. .... .. .-.. -..

c h i l d = child 

- Hardwicke's Science-Gossip.

[1]

Cables not fully described in the text, Map B. Eight cables at the Anglo-American Company: 7, Heart's Content to Placentia, two cables; 8, Placentia to St. Pierre; 9, St. Pierre to North Sydney; 10, Placentia to North Sydney, two cables; 11, St. Pierre to Duxbury; 18, Charlotte's Town to Nova Scotia; 19, Government Cable, North Sydney to Bird Rock, Madeline Isles, and Anticosti; 21, Halifax and Bermuda Cable Company's proposed cable to Bermuda.