159
36
PUBLIC RECORD OFFICE
Reference -
TLC.O. 885
6
PUBLIC RECORD OFFICE, LONDON
ALEY WITHOUT PERMISSION OF THE BE REPRODUCED PHOTOGRAPHIC- COPYRIGHT PHOTOGRAPH-NOT TO
1 December 1896.]
Mr. CARSON.
Sir Donald Smith-continued. 2551. But under the control of the Western Union Telegraph Company? Possibly the Great North Western Telegraph Company may be under the control of the Western Union, I do not know; they may or they may not.
2552. Their lines are leased to the Western Union?-So.
2553. And also the lines of the Dominion Telegraph Company? And the Dominion; I know the Dominion are.
2554. They are both leased to the Western Union; that is the case? Yes. There is this, however, to be said, that that message does not leave British territory.
2555. No, no; but under the control of the Western Union Company from the point at which the landing is made in Cape Breton, Nova Scotia, to Montreal?-To Montreal, that is so.
2556. Under the control of the Western Union ? Yes.
Chairman.
Can
2557. Quite another point, Mr. Carson. you tell me how many letters you reckon there are in a telegraphic word, on the average Fight.
?__
2558. And what proportion of paying words do you reckon there are in the words that you have to send?-It bears upon how many words there are in a message; that is the first thing.
Mr. Jones.] Say 10.
Chairman,
2559. What is your average length of a message?--Nine words at the outside.
2560. How many additional words have you to send?-Three.
2561. Therefore, on the whole of your business your paying words are three-fourths ?--Nine- twelfths-three-fourths.
com-
2562. Those words, of course, may be either press-words or Government words, or inercial words; the value of them may vary, but nine words out of every twelve are words for which you are paid ?—Yes. The reason of that is that we have a system by which we do not waste the capacity of our cables by repetition. You take a message, say, of ten words, and it contains three or four groups of figures in it. Well, in the ordinary run of telegraphy that message would be acknowledged by the receiving station, the number of the message and all those figures would be repeated back again. We do not do that; we repeat nothing; we send the message a continuous string of messages- during the whole of the working day.
2563. What is the working-day ?-Our work- ing-day is from nine o'clock in the morning till four o'clock-that is about 19 hours-18 to 19 hours is our working day, owing to the difference in time between the two countries; but, as I was saying, we do not repeat back, we have a system of checking. We send it on the instrument, another instrument records what is sent, and a second man checks what is sent. In 99 cases out of a hundred, what is sent on the cable will reach the other end. At the other end we have a man who checks the receiver, that is on those occasions when we write down the messages; but
Chairman-continued.
[Continued,
at some of our stations we never write down the messages, we send them from the slips upon The which they are received at the station. clerk can read those telegraph signals as easily as he can manuscript, in fact, more easily. They are sent on, and that eliminates all chance of
error.
Mr. Gillies.
2564. Then do you consider that that increases the carrying capacity of the cable ?-Decidedly, because the cable never stops, it goes on in one continuous stream,
2565. In what proportion would you consider that that increases the carrying capacity of the cable over any other system?-It all depends upon the amount of figures that there may be in your traffic; still, a very sensible increase in the carrying capacity of a cable may be made by omitting those repetitions; very sensible, I should think. I cannot say just at present what it would be; of course it would be the a fluctuating quantity dependent upon amount of figures which you have to repeat. We repeat nothing; not even a Government if it were composed entirely of figures message; we should not repeat a word of it.
Chairman.
2566. What should you think of a cable, that only half of the words sent were paying words?
-It is a matter of system.
2567. Do you call that a good system?—I call my system the best system.
Mr. Gillies,
2568. I suppose there is something to be taken into consideration about the quickness and the general ability of the person who manipulates the sending of the message through the cable; the expertness of the officer ?—Yes, if you were working the lines by an officer. We do not work our lines by officers; we work them by apparatus--automatic apparatus. With reference to the capacity of a cable, there is the capacity of the cable and the speed at which a clerk can work. The clerk may be anything you like; or he be a he may be a very clever man
We have very heavy-handed man and so on. cables that will carry 150 letters a minute; for speed we generally take five letters to the word, 150 letters would be 30 words per minute. have seen clerks working that cable at 60 letters per minute. I visited all our stations years ago in America and England. I found that upon one of our best cables the clerk was working at 60 letters a minute. I knew very well that that was a great waste, and eventually we got rid of the sending by hand and introduced machinery which would maintain the speed regularly; we keep the machine running at a speed nearer to the theoretical capacity of the Now we cable than any clerk could work. are running that particular cable at 130 to 135 letters a minute all day long.
1
2569. Are the mistakes numerous;
I mean that the receiver cannot make out, cannot trans- late, cannot make out the words and write them down ?—You mean failures of the apparatus.
1 December 1896.]
Mr. CARSON.
Mr. Gillies-continued.
2570. Yes?-No; because, as I have said, we have a checking clerk, who checks the signals as they are sent in, and then as regards errors, and what we call "rubs out" and waste of that kind we have rules that no clerk is to put a slip on to the cable at all unless it is free from errors to a certain extent. Say it is a message of 10 or 15 words, he shall punch it without making any errors at all. If he makes errors he must destroy that slip; but whatever he does, his blundering must not go upon the cables. That is our system, and we do not find many errors.
2571. Is that practice extensively adopted? -I do not know, it is our system.
Chairman.
2572. Is it a system that could be adopted any- where Anywhere, wherever you like.
2573. Across the Pacific? — Oh, decidedly. Certainly, anywhere.
Mr. Jones.
2574. Your company, the Anglo-American, is what is called a British company, is it not?—It is a British company, it is the oldest- the l'imecr Atlantic Cable Company. It laid its first cable
in 1858.
2575. But you have an arrangement with the Western Union, which is an American organisa- tion, is it not?—Yes.
2576. Your cable landing at Sydney, then you hand over your business to the Western Union? -Yes.
2577. And from that point to Montreal it is so to say in the hands of a foreign company?—-
Yes.
2578. From Cape Breton, how do you get to Canso; you have a cable across to Canso, have you not? We have a cable there in case of interruption of our ordinary route, but we have no cables at Canso. There are Western Union cables coming from Penzance, they land at Canso and the Commercial Company's cables land at Canso. The Direct Company's cable lands at Halifax.
2579. So that there seems to be a feeling entertained that it would be desirable, if pos- sible, to have the whole route under English control, so to say?—What do you call the whole route?
2580. Well, from here to Vancouver, or if a l'acific cable is built, the Canadian Pacific could, of course, extend their lines to meet you? Yes, I believe the Canadian Pacific (whose railway terminates at St. John, New Brunswick) have a wire to North Sydney, Cape Breton, on the Government Railway-messages for the Pacific could be transferred to the Canadian Pacific at North Sydney with the permission of the Western Union Company.
2581. One question more. You said with reference to the non-paying words-how many letters would you consider in a message were non-paying? The non-paying matter in a message is the number of the message, that will be three letters; the station from, that will be in a code of three letters; then there is the
[Continued.
Mr. Jones--continued. number of words, which is two letters; that is eight. Eight letters is equal to one word.
In
2582. One word for the message ?—Yes. 2583, Which would be sent free? - Yes. the previous estimate I counted the number of the message, the station from, and the number of words, as words; and then the message itself, we take as nine. That would be nine-twelfths; but what we send free only amounts to a single word, so that we really send nine-tenths of our work as paying matter.
Chairman.
2584. Do you send the time of day?—No. A cable is a very expensive thing and its carrying capacity very limited, therefore as little as possible should be wasted in unpaying matter. For several years we did not give even the place of origin free, and it was only through competi tion was granted.
Mr. Jones.
2585. In the event of a Pacific cable being laid from Vancouver to Fanning Island, could the same system be applied over 3,300 miles ?— Yes absolutely the same system.
2586. Are you an electrician?-1 have had 44 years' experience.
Chairman.
2587. Let us get this quite clearly, Mr. Carson; you consider nine-tenths of your words paying words, as a matter of fact ?-If you take what we call the preamble as one word of eight letters we get nine-tenths.
2588. In fact you give us nine-twelfthe ?--By counting the preamble as three words.
2589. Whereas by number of letters it is only one-By number of letters it is only one.
But you may take it in this short way that we repeat nothing. Of course under the Convention rules you are obliged to repeat figures and acknow- ledge the receipt of messages, but those are not
our rules.
Mr. Jones.
2590. How many letters per minute do you think would be sent over the Pacific cable sup- posing it to be laid?-What is the length?
2591. Three thousand, three hundred miles.— And what is the size?
2592. Say about 550 lbs. of copper and 350 lbs. of gutta-percha.-I suppose you would get about ten words a minute through.
2593. Ten paying words? Well, if you adopted my system.
2594. You would get ten paying words a minute with the duplex system, ten words each way?-Yes, there is no reason why you should not get ten each way.
2595. That seems considerably in excess of some other estimates; I think it only fair to mention that some have estimated much less.— It is a question of system as to how much waste you get upon your cable; we have gone very carefully into this matter of waste, and we believe that we have run it down to the lowest,
U 4
160