CO885-(6-7) — Page 325

91

330

PUBLIC RECORD OFFICE

Reference :-

C.O. 885

6

PUBLIC RECORD OFFICE, LONDON

ALLY WITHOUT PERMISSION OF THE BE REPRODUCED PHOTOGRAPHIC- COPYRIGHT PHOTOGRAPH-NOT TO

24 November 1896.]

Chairman-continued.

90

Mr. PREECE.

ship you can regulate the slack within, say, 5 per cent. and practice has made it now, as I say, about 10 per cent.

1346. Would you say that an excess of slack was an evil? It would be waste.

has been a

1347. The waste I can understand; but you have heard the evidence which has just been given by Captain Goodsall as a repairer of cables. He evidently regarded slack with a somewhat unfriendly feeling? He did, for the very simple reason that at the present day, with these cutting grapnels, the art of grappling a cable and cutting it and bringing it to the sur- face has become a very accomplished art, and it causes them no trouble at all; so that, as a matter of fact, slack now is not laid so much with the idea of facilitating repair but as an absolute necessity, as I have tried to explain.

1348. Now, I should like to take you to the There question of the speed of working.

good deal of misconception on the matter of speed; would you explain how this misconception has arisen, and what are the real facts of the case? Well, hitherto the practice has been very varied indeed. It has been the custom always to speak of speed of working as Bo many words per minute. Now, the practice has become general to drop the words, and simply speak of letters per minute. It arises from the simple fact that if you take a leading article of the Times" and count up the number of words or letters, you will find the average is a little under five letters per word, but now tele- graphic words have grown in dimensions; code words are introduced, as you heard yesterday; the little words "and"and" if," and conjunctions and all those things, are abandoned, and a mes- sage itself is a very different thing to what it was in the early days of telegraphy. Well, now, I wanted to get up absolute facts, and bring absolute We have facts before you on this word business. the control of the whole of the American traffic passing over the Anglo Company's lines. We supply them with wires; we have relaying stations in liaverfordwest, in South Wales, at Llanfair, in Wales, and at a place called Nevin, also in Wales, and I had at these stations for two hours for six days every message counted, and it came out that the average num- ber of letters per word in the American traffic was 7-3. That is my own observation. I have had that confirmed since from observations from the Eastern Telegrapli Company, which, doubt less, will be given you authoritatively by Mr. Heese, and there for the same words, that is, the mere words in a message, not taking into con- eideration the official codes, they make it out to be 7-6 little longer. In addition to that, every letter requires a certain number of electrical impulses to form it. The elements of the letters themselves are separated from them- selves by impulses, and the words are also sepa rated from each other by another number of impulses. The result of it is that for calculating purposes the number of letters forming a word must be taken as eight. Then, from my own ob- servations on the Anglo traffic, it is 7.3 letters per word, plus the spaces, making the number for cal-

Chairman-continued.

[Continued.

culating the number of letters to a word eight; so that in my calculations I am going to give you I take eight letters per word,

1349. I think an interesting experiment was made in August last by the Commercial Cable Company? Yes, it was mentioned in evidence, and it was so important that I asked the Com- tercial Company to give me the facts of the case, and they have done so.

1350. Would you describe exactly what that experiment consisted of?-I will give it you exactly in their own words :-" In August last looped two of our cables at Canso, making

11

line of 4,511-802 knots, and obtained up to 27 letters per minute; 50 volts and so "microfarade, the extreme limit we allow for

ordinary purposes, were used to send with. "The receiving was done through 140 micro- "farads, and we think it possible to work up to 35 letters per minute under more favourable "conditions."

1351. On that I should like to ask two ques- tions. That would be, on your calculation of eight letters per word, about three and a-half words per minute --That is so, as they got 27 letters, and they say under hetter conditions (that is, of apparatus) they might have got 35. I take it that on that length of 4,512 nautical miles we might have had 32 letters per minute. That means four words.

1352. Or, on the duplex system, eight words? -Eight words in all. I refer to single working. I shall refer to duplex working afterwards.

1353. What was the core of that cable ?- That cable is of the Atlantic pattern. It is what Mr. Siemens in his table calls the Mackay-Bennett. He gave us a table; here it is. It is about- you may roughly take it to be about 400 and 400.

1354. Then the core of this is 400 and 400 ?— I am not sure about the exact figures, but I think you might take it as 400 and 400. could get it accurately and let you have it.

Sir Donald Smith.

I

fact

1355. The 1884 cable ?-These are the two 1884 cables. Now why I am anxious to put this in is this: that taking it at 32 letters, or four words, it exactly confirms the standard that I have used in all my calculations, which is 20 words through 2,000 miles of the Atlantic core. That is a standard of speed that I have always used in my calculations. Now, this experiment. gives us a very good standard to work from. It is an absolute experiment-it is a that we have got under our control, and it is not mere calculation from laboratory ex- periments or from mathematical deductions. From that I have prepared a table, and in this table, taking 2,000 miles of the 1884 American core at this four words per minute, I have taken the Anglo-American of 1894 as the 400 and 650, I have taken the Commercial cable (that is, Siemens) as 500 and 320, the Silvertown pro- proposed 533 and 365, and the Post Office posed 940 and 940; and on the length of the

24 November 1896.]

Mr. PREECE.

Sir Donald Smith-continued. Fanning Island cable I have only taken 10 per cent. slack for this calculation, so as to make it com- parable with the others. I make it 3,628. The Anglo-American type of 1884 will give us 6:08 words per minute. I will put in this table.

Chairman,

1356. You will hand it to me presently ?- Yes, I will hand it to you. The 1894 American

core

will give 785 words per minute; the Commercial Siemens core will give 63 words per minute; the Silvertown proposed core will give 67 words per minute; and the Post Office will give you 14·3 words per minute.

1357. What is the fallacy underlying the evi- dence which you have heard from Mr. Gray and Mr. Siemens as to the rate of speed ob- tainable with their cable?I think it is solely due to the fact that they have taken their words at five letters per minute. If you take their figures and multiply them by five, and divide by eight, you will find that their speeds will agree almost exactly with those that I have taken; in fact, I may say at once, on the question of speed there is absolutely no possibility of error. cable speed is one of the most certain things we have to deal with.

Mr. Jones.

Chairman-continued.

Continued.

was the figure specified which the cable had to do; that was the figure which the Silvertown people and the Siemens people tendered to; well, I say that those 12 words per minute on the 8-letters standard mean 7.5 words per minute. You simply multiply 12 by 5, divide by 8, and you get 75. So on the Silvertown cable we cannot count the 8 letters per word on a greater speed than 7.5 words per minute, and with these causes of delay that I named to you, that is brought down to 4 words minute.

per

1362. Can you do the same for each one of those, and write it by the side ?—Yes; I will do it afterwards for you. I would not undertake to do it at this moment.

1363. Will you complete the table in that way?—Yes.

is

1364. Would it be easy to maintain this reduced speed of four words per minute?—No We find by experience in working telegraph that it is almost impossible to maintain high pressure. It depends a good deal on the clerks, and they tire. It depends also a great deal on the business hours, and the variation in the business transacted. The

Even taking the case of the Atlantic cables, which are certainly worked at the highest pressure the greatest amount of trate exists hetween this country and America-and there well-marked maximum and minimum of work dependent entirely and solely on the hours of business in London and the hours of business in the United States. I have a letter on that point: I have no gures to hand by which "a curve of our daily traffic could be made, but approximately speaking, the traffic from this "side to America attains a maximum between four and seven in the afternoon, and a minimum *between 8,30 and 11 in the morning. That from America is at its highest between 9 p.m. and 4.30 a.m., culminating about midnight. "The lowest being from 4.30 a.m. to 2 p.m.,

falling to zero about 8 a.m., English time.'

135. That is what Dr. Muirhead said yester- day?—Yes, it is one of the most certain things. If it is an error, it is an error due to our starting with different standards, and the whole and sole difference between my figures and Siemens', and Gray's, and the rest, is merely due to our starting from different standards.

Chairman,

1359. Now these speeds; are they actual working speeds obtainable, or are they the highest possible speeds? They are the highest possible speeds.

1360. What allowances must you make to obtain from the highest possible speed the prob- able working speed P-Well, there must be some allowance made for errors in transmission, errors in reception; words have to be repeated; the messages themselves are prefaced by the official prefixes to indicate what they are, where they are to go to; they are coded for time and for other purposes, and there are also, in the transmission of business, always a great number of service messages, of course, necessarily sent free. The result is that all these sources of additional work mean additional delay, and my calculation is that with a maximum speed of 75 words per minute, you could not possibly count on getting through more than four words per minute when working at high pressure.

1361. Could you reduce the figures in each one of these cases to what you consider the working actuality 7-What I did was this, which I ought to have said, ignoring that for the moment. If we start with the assumption that our cable, the Pacific cable, is to do 12 words a minute, that

$4

LL

*

1365. Now, what, in point of practice, can you inform the Committee, is the amount of working hours that can be got out of those cables ?--Well, Dr. Alexander Muirhead yesterday took it at 18 hours. I think he was quite wrong there. I do not think he can have sufficient experience in the traffic working of the line to have given authoritative figures, and I am sure you will be able to get absolute figures from the Eastern Telegraph Company. My experience is this, that if the total work during the 24 hours were con- centrated at high pressure for a certain number of hours you could not take it at more than 10 hours a day. What I mean is, that between London and Australia, although the business would be transmitted during the whole day, owing to these maxima and minima periods, it would be the same as though the whole work were concentrated for a certain number of hours at high pressure. Now I take that, and I make this calculation, of an ordinary pressure a mean pressure-you that owing to time, and to the always existence

cannot take the work between London and Australia at a greater rate than three words per minute, nor for a longer period than 10 hours, and

M 2

92