Publication: Journal of the Society of Telegraph Engineers
INDIAN AND AMERICAN TELEGRAPHS.
In March, 1873, Mr. W. E. Ayrton read a paper, entitled " On some Points in connection with the Indian Telegraphs," before the Society of Telegraph Engineers. This paper is published in No. 5, vol ii. of the Journal of the Society.
Mr. Ayrton first refers to the application of Mr. Schwendler's formula for reducing the apparent or observed resistance of either insulation or conductivity to its true value. Soon after the publication of Mr. Schwendler's formula I applied it in the measurement of insulation and conductivity of overland lines, but never found it to agree, or even approximate, to the result of a careful measurement, taken under circumstances where the insulation resistance was so large as not to affect the result, as measured in the ordinary manner; that is, to take the mean of the observed resistance with alternating poles of the battery.
Mr. Schwendler's formula is based upon the supposition that the insulation conductivities are uniformly distributed throughout the line, a condition of affairs that never exists in this country except in clear or cold weather, and when the escape of current by the insulators is so small as not to enter into the result.
In summer, long lines are often affected by local showers: it may be clear on some portions, and raining at others. During general rains, "north-east storms," humidity that affects insulators is very unevenly distributed — raining in some portions and only cloudy in others. The conductivity of insulation is never stationary under these circumstances, but as varying as the amount of rainfall at different intervals. I refer to the ordinary glass insulators of this country.
On very old wires, whose resistance is great through bad joints, conductivity is immensely improved by rain or dampness. So that, in my experience, it is idle to attempt the measurement of wire resistances or line conductivities, except by short sections or under favourable conditions of weather.
Mr. Ayrton says, "We have learned two very important facts," one of which is, "that the insulation of sections varies enormously under the same climatic influences." If this be true, how do they arrive at the real conductivity of the wire ? It also speaks badly for the insulators, provided we are to understand by the term "same climatic influence" that there is sensible leakage on any portion in clear weather.
A porous porcelain insulator conducts in clear weather; a glass insulator under same circumstances does not . By this is meant, when both are exposed to the sun's rays and not affected by moisture on the surface.
Mr. Ayrton enters into a calculation by which he is led to believe the consumption of battery material is increased, in his particular case, 60 per cent. by the introduction of the 1 per cent. of bad insulators. This calculation leads to erroneous results, inasmuch as it is based upon the supposition that there is no consumption when the circuit is open or the exterior resistance is infinite. Consumption of material is going on when the circuit is open, and this is particularly the case with the kind of battery he refers to as used in India, the Menotti.
In our country this battery without the intervening layer of sand or sawdust is most used. It is best adapted to our mode of working the closed circuit system. The size of the elements is sufficient to bring the resistance down to five and sometimes two Siemens' units per cell. If the battery is much worked this fact tends to keep the solutions separate. The more wires worked from the same battery, or the less the exterior resistance, the less the local action, and in this sense the more it is worked the greater the economy.
It answers admirably for working many lines from the same battery.
This battery is called the Callaud, and is essentially the same as used by the French at their central station in Paris.
Mr. Ayrton describes the manner of testing insulators at the factory and upon their arrival in Bombay, showing the great care bestowed upon the insulation.
Are the benefits derived commensurate with the care bestowed?
This question was put in another form by Mr. Preece, and the answer of Mr. Ayrton is, "In dry weather the insulation varied from five millions a mile to over three hundred, in wet weather it was as low as half a million;" and that "the average was two to three millions in wet weather."
I infer that better results are obtained in this country with our poorest insulation. In measuring say 100 miles we seldom find it as low as one million per mile in rain. Leaving out the city portions, it will run as high as eight and seldom as low as four millions.
If we measure the insulation from this city to New York, the insulation of the same wire will measure twice as high when taken from the outskirts, and say six miles of city lines excluded as compared to the entire line, including city portion, where the measure is taken from the central station.
The insulation in cities is greatly reduced by smoke and gases of combustion which "short-circuits" the insulation, making the remit appear below the average. Again, measuring from the country into a city, where the insulators are thus affected, the results appear above the average. Comparing the insulation of the lines in India, I should infer they could be improved by using the common insulator of this country.
As regards progress in insulation, are they to be compared with many lines in this country that average at least one hundred millions per mile in rain? We have thousands of miles of such lines in which I have a personal interest, and it would be out of place for me in this article to refer to them in detail.
The second " important fact" developed by the mode of testing in India was that the earth-plates had "a high resistance and were intensely polarised." "This was remedied by using large copper earth-plates with leading wires insulated from the ground."
Many people think copper better for ground-plates than iron in an electrical sense, but copper on account of its better conducting qualities possesses no advantage over iron, and zinc-coated iron is more economical and equally durable.
What advantage copper possesses over iron for ground-plates I am unable to see. A ground-plate is simply a joint between two conductors, the wire and the earth, the latter specifically being a very poor conductor. If the sections of an iron and copper wire were joined, the iron being six times as large as the copper to ensure equal conductivity, the section of iron in contact with the copper would not conduct sufficiently to make a joint equal in conductivity with a section of either the iron or the copper wire. But if we enlarge the end of the copper so as to cover the section of the iron then the joint has a capacity for conduction equal to a section of either. We can, after understanding the nature of this joint, draw upon the imagination to form an idea how immensely an iron wire must be enlarged, and the extent of surface exposed, to fulfil the conditions of a joint with the earth. Supposing the earth in contact with the plate to be one-fourth saturated or contain one-fourth moisture, the plate would require at least ten thousand square feet of surface in contact with the earth to ensure a joint electrically perfect, or equal to a section of No. 8 iron wire.
With perfect joints the immense section of the earth, considered as a conductor, adds nothing to the resistance; or, in other words, if we make two perfect joints with tho earth, there is nothing added to the resistance of the circuit.
The defects of ground-plates in connection with lightning-rods have been much discussed by the meteorological section of the Franklin Institute of this city latterly. It has been the custom to terminate lightning-rods in dry earth, often without any plate attached. It is the belief of those who have given the subject much attention that rods thus constructed are the cause of more harm than protection.
Every summer many valuable barns with their contents are destroyed, and in a majority of cases these barns are supplied with rods improperly connected with the earth. These rods draw the discharge from the clouds, and, there being insufficient means of getting from the rod to the earth, the charge flies to neighbouring conductors, often log-chains, ploughs, cart-tyres, or other metallic substances, igniting the combustible material with which these buildings are stored.
As an experiment, two plates, containing nine square feet of surface each, were buried fifty feet from each other, using the two plates and the earth as a portion of the circuit. They gave a resistance of 150 Siemens' units. Allowing a stream from a hydrant to flow directly over each plate, with holes in the ground above, so as to completely saturate, the resistance of these plates or rather the earth in contact was reduced to 32 units, or 10 units each, compared to an entire metallic circuit of No. 10 wire, 100 feet in length. By substituting 50 feet of wire in place of the earth, total resistance was reduced 32 units.
Instead of lightning-rods being connected to the earth in an electrical sense, they are, in truth, insulated from the earth.
People have often found it difficult to work short lines of telegraph in this country where the earth was used as a portion of the circuit, and connection was made or attempted by an ordinary ground-plate, exposing insufficient surface.
The writer was called to explain a difficulty of this kind about two years ago. A line, one mile and a half in length, connecting a factory with the proprietor's residence, was constructed with type-printing instruments. They failed to work, from insufficient current. The battery had been increased to 100 cells, the ground-plates overhauled and replaced with pure copper of increased surface, but all to no purpose, the apparent current was feeble.
The resistance of this line including the earth-connections was found to be 1600 units. Owing to the dryness of the soil the earth-plates made poor connection. Within a hundred yards of each terminal there happened to be a railway track; connection was made to this track, and the strength of current increased thereby to the extent that ten cells were ample to work the line.
A railway track, as the rails are joined in this country, makes a perfect ground even when the earth is dry, owing to the immense surface in contact.
A ground-plate, as usually constructed in this country, seldom interposes sufficient resistance to affect ordinary lines. The longer the wire or the greater the resistance, the less the effect in the proportion this resistance bears to the total resistance of the circuit.
In this sense also the polarization of the ground-plates is small, in proportion as the resistance of the line-wire and instruments is large. In the case of the India lines or circuits of 500 miles, I am unable to see how its effects are even appreciable. Insulating the wire leading to the plate to prevent it would do so, so far as this action is local, or confined within the limits of the plate and its connecting wire.
How this effect could now be shown on the instruments of the line, were the leading wires uninsulated, I am unable to comprehend.
The portion of Mr. Ayrton's paper that elicits most remark in this country is as follows: "To facilitate this reading by sound, the receiving signaller gives an acknowledgment by sending a dot at the end of every word, and the sending signaller continues repeating the word until he gets this acknowledgment."
By this method, the receiving " signaller " would be obliged to drop his pen to acknowledge every word as transmitted, or at least work under great disadvantages. We are unable to see how two good operators could by this method get more than twenty messages per hour over a line, these messages to average twenty words each, while by our method of closed circuit forty messages per hour is a moderate rate of speed. The sounder by the American plan responds to every touch of the sending operator. Any accident or change in the condition of the line involves a change of adjustment of his relay. If such change takes place either by increase or decrease of current the sending operator pauses for a reply from the receiving operator. On the circuits where much business is performed the operator sends for hours without a reply, yet he has assurance that every word is being correctly received. Especially is this the case when despatches are going in the other direction from the same station. The sending operator is assured of the correct transmission, otherwise he would be advised by the other wires. So it is not the custom in this country to stop sending merely for the purpose of ascertaining whether the despatches are correctly received. Whenever the sending operator stops and closes the circuit, then the receiving operator replies or acknowledges.
By the duplex system, lately introduced, this method is increased. We may say the average is from 60 to 70 messages per hour on a single wire.
I would like to hear from Mr. Ayrton how this speed of transmission compares with that of the Indian lines.
Perhaps a more pertinent question would be, Do those interested in the Indian lines get as great a return for their expenditure in erecting and operating a No. 1 or No. 3 wire by their method as obtained in tins country with lines of equal length, but of less than one-third the size, and the American method?
If I am able to judge fairly I should say lines in this country, costing half as much as the Indian lines, do twice or more than twice the business.
The methods of operating and maintaining the Indian lines are referred to in the light of progress or advancement in telegraphy — we are considering them in that sense from an American standpoint.
In the discussion following Mr. Ayrton's paper, Mr. Preece states "that sounders had been introduced almost entirely in America."
As early as 1860 nine-tenths of the instruments in the larger and more important stations were sounders, and now there is scarcely a Morse recorder to be found in any of these stations.
The recording instruments are used only to a very limited extent, in the way-stations of railway lines where the least business is performed. They are used by operators who are learners until they are able to read by sound.
In the year 1850 "Jemmy" Leonard, of Louisville, got the reputation of being the most expert "receiver" in the country, and entirely by ear. Since that time, or soon after, it became tho ambition of operators to emulate Leonard, and receiving by sound from that date has been popular.
In 1852 I designed the first sounders used upon the lines of which I had charge. They are now made with from 4 to 8 units resistance, and usually worked by two cells Callaud battery.
In the years 1867-8 Mr. Varley made a report on the lines of the Western Union Company. In speaking of relays he says, "Keep down the resistance of everything is a golden rule in telegraphy." He found the relays in the New York office "to vary from 69 to 1182 ohms," and that "the 69 ohm relay is, where it should be, on a long line No. 2 Chicago;" and further, "The resistance should not exceed 130 ohms per relay, and it will pay you to have them all re-wound." Acting upon this advice the resistance of the relays was reduced.
For circuits of one hundred to two hundred miles they averaged about 50 units, and from 100 to 200 cells of Grove battery were required to work the lines.
Mr. Ayrton says, in speaking of relays, "We have not, however, used any relays having more than about 3,600 or 4,000 ohms resistance." I merely refer to these opinions to show that "doctors disagree," and that there is no question upon which we have such a diversity of opinions as upon the proper resistance of the relay. I would like much to see a relay made to fulfil the requirements as demonstrated by Eisenlohe, Dumoncel, and others.
The resistance of the relay, I will venture an opinion, has much less to do with the good working of lines, practically, than other conditions.
The Philadelphia and Reading Railroad work, a circuit of 140 miles with 40 relays, average resistance 100 units each. This line is used in connection with its traffic and the running of trains. It works well in all weathers, and would were the resistance of the relays twice as large or twice as small.
The Pennsylvania Central Railroad has three similar circuits, with relays averaging 150 units each, at intervals of about five miles; this resistance is larger than necessary, but the excess, whatever it may do, occasions no inconvenience.
With ordinary insulation, or say an insulation resistance of two million units per mile, these lines could not be worked by our American system of closed circuits, with a battery at each terminal and the currents from both flowing when the circuit is closed.
When the sender opens the line the receiver's battery still sends a current through his relay over the line and insulators, which is greater or less as the insulation is good or bad; the affective or working margin is the difference between this "escape" and the whole line current. When the line current is small, owing to the resistance of wire and relays, this margin is small, and the "escape " must be kept at a minimum.
The wire in connection with all stations on the main trunk of the Philadelphia and Reading Railroad before referred to has a resistance, including relays, of over 8,000 units; with such a resistance the current passing over the line is unavoidably small, and to be able to work such circuit the current escaping by the insulators must be small in comparison to that passing through the line.
The Philadelphia and Reading Railroad Company makes a more extensive use of the telegraph than any other railway in this country; they employ ten wires on the main line, with branches in all directions, as soon as they enter the coal regions, 130 miles distant from Philadelphia.
I refer to the railways in this country because they are the only parties operating the telegraph upon anything approaching what is usually termed scientific principles ; and further, all that Mr. Ayrton says of the railway telegraphs in India is of a negative character.
The foregoing are suggestions upon reading Mr. Ayrton's paper, and the remarks in connection therewith of the other Members of the Society.
The Chairman said that the Indian Telegraphs at one period were the "roughest thing possible." "He believed at the present time the Indian Telegraphs were amongst the most scientifically worked telegraphs in the world."
In a free and open discussion I should say it was more scientific than practical.
Real advancement in telegraphy is best shown by comparing the results obtained with the means employed.
In this connection it is proper to refer to the Automatic Telegraph Company, that operates a single wire extending from New York to Washington, a distance of about 300 miles, and doing a local business with stations at New York, Newark, Trenton, Philadelphia, Baltimore, and Washington.
In this circuit there are fifteen relays. The line is worked in one circuit; or, in other words, while transmission is going on between any two stations, the balance of the line is unemployed so far as transmission of the business of the other stations is concerned.
This Company does an average daily business of upwards of 700 paid despatches. They have secured this business simply by promptness and efficiency, in competition with the old Company, having its wires extended to every considerable town in the country, with hundreds of branch offices in the cities mentioned. Ninety per cent. of this business is done during the six active