ill 111 1 1 w
i ll! 1 (11 ill! I
ii liiiiiiiirif ■
i ! Ill I i 1 J i
i (i i i if i
i II ill
BOUGHT WITH THE INCOME
FROM THE
SAGE ENDOWMENT FUND
THE GIFT OF
Benrg W. Sage
1891
fiM9^S /^^f^
3 1924 031 255 007
olin.anx
( 3 )
Electricity, in our Homes, Walker,
6s.
Static, Atkinson, 6s. 6d. [6d.
Supply, Gay and Yeaman, los.
Municipal, Gibbings, ly.
and Magnetism, Ash worth, 2s.6d.
Bottone, 2s. 6d.
Gerard, loj. 6d.
Hi)uston, y. 6d.
Maycock, 2s. 6d.
Electro-motors, Bottone, y.
Houston, $s.
Electro-platers' H. B., Bonney, y.
Electrolytic Analysis, Neumann, 6s.
Elliott's Gas and Petroleum Engines,
2J. 6a?.
Industrial Electricity, 2s. 6d.
Emmet's Alt. Current Wiring, t,s.
Engineer Draughtsmen's Work, is. 6d.
Fitting, Horner, ^.
Locomotive, McDonnell, is.
English and American Lathes,
Homer, 2ls. net.
Explosives, Guttmann, 2 vols., 2l. 2s.
Farman's Auto Cars, p.
Farrow, Specifications for Building
Works, 3j. 6d.
Stresses and Strains, $s.
Field Work and Instruments,
Walmisley, "js. 6d.
Findlay's English Railway, 7^. 6d.
'Fitting, Homer, ^s.
Electric, Maycock, 6s.
Electric Light, AUsop, S^.
Fitzgerald's Nav. Tactics, is.
Fletcher's Architectural Hygiene, 5^.
Carpentry and Joinery, ^s.
Steam-jacket, "Js. 6d.
Foden's Mechanical Tables, is. 6d.
Forestry, Webster, y. 6d.
Forth Bridge, Is.
Foster's Central Station Bookkeeping,
los. 6d.
Galvanic Batteries, Bottone.
Garratt's Modern Safety Bicycle, 3J.
Gas and Petroleum Engines, 2s. 6d.
Gaseous Fuel, is. 6d.
Gatehouse's Dynamo, is.
Gay and Yeaman's C. S. Electricity
Supply, los. 6d.
Gearing, Helical, "js. 6d.
Geology, Jukes-Browne, 2s. 6d.
Gerard's Electricity and Ms^netism,
los. 6d.
German Technological Dictionary,
4J. net. .
Gibbings' Dynamo Attendants, is.
Electricity Supply, it,s. net.
Gore's Electro-chemistry, 2,s.
Metals, los. 6d.
Gray's Electrical Influence Machines,
. 4f. 6d.
Griffiths' Manures, "Js. 6d.
Guttmann's Explosives, 2 vols., 2/. 2s.
Hatch's Mineralogy, 2s. 6d.
Hatton's Mathematics, 2s. 6d. and 3^.
Haulbaum's Ventilation, is.
Hawkins' and Wallis's Dynamo,
los. 6d.
Heat Engines, Anderson, 6s.
Helical Gears, "Js. 6d:
Herbert's Telegraphy, y. 6d.
Telephone System of the Post
Office, y. 6d.
Hering, Electric Railways, Ss.
Hertzian Waves, Bottone, 3^.
Hewitt's Organic Chemical Manipula-
tion, 4J 6d.
Hobbs' Electrical Arithmetic, is.
Hoblyn's Medical Dictionary, loj. 6d.
HoltzapfFel's Turning, 5 vols., S^- 9^-
Hopkinson's Dynamo Machinery, $s.
Horner's Helical Gears, "Js. 6d.
■ Iron-founding, 4J.
Metal Turning, 4J.
Pattern Making, y. 6d.
Principles of Fitting, y.
English and American Lathes,
2ls. net.
Hospitaller's Polyphased Alternating
Currents, 3^. 6a!'.
Houston's Electrical Terms, 3 1 J. 6d.
Electrical Measurements, $s.
Electricity and Magnetism, 3^.
6d.
Transmission of Intelligence, 5^.
Hurler's Alkali Makers' Handbook,
10^. 6d.
Hutton's Mathematical Tables, 12s.
Hydraulic Motors, Bodmer.
Incandescent Lighting, Houston, 5^.
Indicator Handbook, Pickworth, y.
net.
Induction Coils, Bonney, y.
Industrial Electricity, 2s. 6d. [net.
Inventions, How to Patent, 2s. 6d.
Iron Analysis, Arnold, las. 6d.
Analysis, Blair, i%s. net.
THE
TELEPHONE SYSTEM
OF THE
BRITISH POST OFFICE.
Manual of Telephony, by Preece and Stubbs, 15s.
Practical Telephone Handbook, by J. Poole, 5s.
Telegraphy, Electricity in its Application ;jo, by
T. E. Herbert, 3s. 6d.
Telephone System of the British Post Office, by
T. E. Herbert, 3s. 6d.
ELECTJiic Telephone, by E. J. Houston & A. E. Kennelly,
Electric Telegraphy, by E. J. Houston & A. E. Kennelly,
Ss.
Electrical Instrument Making, by S. R. Bottone, 3s. 6d.
Wireless Telegraphy, by S. R. Bottone, 3s.
Electricity and Magnetism, by W. Perren Maycock, 2s. 6d.
Practical Magnetism and Electricity, by J. R. Ashworth,
2s. 6d.
WHITTAKER & CO., LONDON, E.O.
^
THE
TELEPHONE SYSTEM
OF THE
BUITISH POST OFFICE.
il ^xAc(iC(x2 35anb6ooa.
BY
T. E. HERBERT, A.M.I.E.E.,
ENGINEER, POSTAL TELEGRAPHS ;
LECTURER, MUNICIPAL TECHNICAL SCHOOL, MANCHESTER ;
AUTHOR OF "Electricity in its Application to Telegraphy."
Second Edition, Revised,
with additions.
If/TH ONE HUNDRED AND FORTY-SIX ILLUSTRATIONS.
WHITTAKER AND CO.,
2, White Hart Street, Paternoster Square, London ;
and 66, Fifth Avenue, New York,
1901
T
Bntbrbd A.T Statioksrs' Hali,
PREFACE.
The acquisition of the telephone trunk lines by the State led to
the design of a system of working which differs in many particulars
from any previous system. Since the transfer many developments
have taken place, and as no adequate description of the system
existed, a series of articles dealing with it were undertaken by the
writer of the present volume. These articles in their crude form
proved to have filled a distinct want amongst the employees of the
Post Office — so many of whom, in one form or another, were for the
first time brought into touch with telephony.
The primary object of the articles was to describe in detail the
trunk system ; but it was considered desirable to give some account
of first principles, and accordingly the transmission of speech,
transmitters, receivers, and, in fact, telephone stations generally,
were briefly described. The writer is aware that much of this has
been described times without number, but his experience as a
teacher leads him to believe that the inclusion of this matter will
render the book useful to a larger circle of readers.
The first eight chapters may be said to be introductory to the
subject of telephony generally. The permanent current systems
nextly recfeive attention, and their various classes of switch
sections. After this follow special switches, such as transfer
boards, record table switch sections, etc.
A chapter upon inductive disturbances has here been interpo-
lated with the object of rendering the subsequent chapter, dealing
with the conditions under which superimposing is practicable, more
intelligible. The writer begs to express the opinion that too much
attention cannot be paid to the questions arising from a considera-
tion of these chapters, and especially so to any whose lot it is to
have anything to do with the testing and fault localisation of
telephone circuits.
vi. Preface.
Test boards, lightning protectors, etc., now receive considera-
tion; and at this point it may be said that the original articles
ended.
In order to reduce the large number of batteries incidental to
the permanent current system, arrangements will shortly be made
to work exchanges from a single set of cells. The complications
thereby introduced form the subject of Chapter XXV.
The writer has also considered it desirable that a short descrip-
tion of the Company's local systems should be added to facilitate
the comprehension of all the operations involved, and with this
object Chapters XXVI. to XXIX. were written.
The Newcastle system is next briefly described. This chapter
may at first sight appear somewhat meagre, but it should be recol-
lected that the system is only in force in some half dozen exchanges,
of which Newcastle is by far the largest. Apart from that, the sys-
tem is interesting inasmuch as it presents some very novel features.
The most recent arrangements are dealt with.
In Chapter XXXI. is given a description of several special
arrangements for which it was somewhat difficult to find places
anywhere else, entailing, as some of them did, a more detailed
knowledge of the Company's local systems ; and the volume closes
with a chapter on switch section faults.
A very brief mention of the KR law is included in the ap-
pendix. A list of the numbers of the tags and their uses is supplied
for purposes of reference, as also is a table of the resistance and
capacity of aerial and underground conductors, for which latter
the writer begs to acknowledge his indebtedness to Mr. A. Eden,
one of the Technical Officers to the Post Office.
It will be observed that the testing of trunk lines has not been
dealt with. This differs little from the testing of telegraph lines,
and with the aid of the chapter upon inductive disturbances no diffi-
culty should arise. The subject of telegraph testing has been dealt
with in so many other works that the writer deemed it advisable to
exclude it as beyond the scope of the present volume, the original
object of which was merely to describe the trunk line system.
As this little volume was originally designed for the especial
use of officers of the Department, the official nomenclature has
been rigidly adhered to. The familiar "jack" or "spring-jack"
gives place to "switch-spring," and similarly, "plugs," "trans-
lators," "Usteningkeys," are replaced by "pegs," "transformers,"
Prefaa. vii.
and "speaking keys." The latter two designations appear to be a
distinct improvement.
In its present and somewhat extended form it is hoped that the
book will also prove useful to students for the City and Guilds
Examinations.
The writer begs to express his thanks to Messrs. Ericsson and
Co., Stockholm, the General Electric Co., and the Consolidated
Telephone Co., for the illustrations of apparatus manufactured by
them ; also to the Journals of the Institution of Electrical Engineers
for much information. ,
In conclusion, the writer begs to here place upon record his
deep sense of obligation to Mr. G. W. Bannister, Engineers Depart-
ment, Manchester, who took entire charge of the illustration of both
the original articles and of the present volume, and who has also
rendered other most valuable aid.
PREFACE TO SECOND EDITION.
In the present edition a chapter entitled " Recent Advances in
Switchboard Design" has been added. Its primary object is to
briefly state the principle of the central battery system which the
Department propose to use for the competitive telephone scheme in
London. Since it will probably be some considerable time before
the details of this system, adapted to the varying requirements of the
trunk and local services, have finally settled down, it has been con-
sidered undesirable to include arrangements which can only be
regarded as tentative. In the same chapter a brief account of the
working of one or two non-multiple or divided board systems has
been included in order to indicate the lines upon which opponents
of the multiple board are working.
An index has been added, but with these exceptions few changes
have been made.
Engineering Department, T. E. H.
Postal Telegraphs,
Manche.ster.
CONTENTS.
PAGE
CHAPTER I. — Introductory i
Principles involved*! the transmission of speech — Pitch, volume
and timbre,
CHAPTER II.— The Bell Telephone 4
Lines of force — Effects of an electric current — Production of
current — Bell telephone, description and theory — Trans-
formations of energy involved — Relative dimensions of parts.
CHAPTER III.— Early Transmitters 10
Edison transmitter — Necessity for induction coil — Principle of
induction coil — Resistance of various parts of circuit — Hughes
microphones, descriptions and theory — Microphonic joints —
Necessity for divided core in induction coil.
CHAPTER IV.— Modern Receivers 17
Swiss — Phelp's pony crown — Gouloubitzky — Ader — Du Moncel's
investigations on design of telephones — Gower — Double pole
Bell— Watch.
CHAPTER v.— Modern Transmitters 24
Classification of transmitters — Gowrer — Ader — Mix and Genest —
Pendulum — Blake — Hunning's transmitters — Packing diffi-
culties — Decker: — Solid-back — Ericsson transmitters.
CHAPTER VI. — Battery Telephone Stations 34
Trembler bells — Switches, etc., and their purposes for telephone
stations — Connections in skeleton — Switch-arms.l
CHAPTER VII.— The Leclanche Cell 40
Chemical action of cell — Agglommerate type — Six block type —
Shallow circular zincs.
CHAPTER VIII.— Magneto Telephone Stations .. .. 44
Magneto bell — Magneto generator — Theory of instrument — Arma-
ture cut-outs — Complete station— Ericsson table telephone.
CHAPTER IX. — Post Office Telephones 53
Gower-Bell — Post Office telephone.
CHAPTER X.— The Local Exchange System 57
History — Requirements for exchange working — Secret systems —
Switch-springs — Flat pegs — Polarised indicator No, z — Tele-
phone relay — Subscriber's instrument — Switch board tele-
phone — Telephone tablet — Polarised indicator relay — Working
pf system — Circular pegs — Local switch for tri)nk exchanges,
y. Contents.
PACK
CHAPTER XI. — Principle of Permanent Current System . . 70
Normal arrangement — Automatic calling — Cfoss-connection strips
— Five and eight-point switch-springs — Actual connections of
trunk line.
CHAPTER XII.— AoxiLiARY Apparatus 73
Telephone Exchange galvanometer and self-restoring indicator,
CHAPTER XIII. — Operating Connections 79
Ringing Ttey — Skeleton of cord connections — Complete connections
of speaking and ringing keys and cords — Mechanical design of
speaking key — Switch telephone connector circuit — Operator's
telephones — Breast-plate transmitter — Double pegs — Head-
gear receiver.
CHAPTER XIV.— The " A " Switch Section 88
Arrangement of apparatus — Post Office subscribers — Service wire
— Junction circuit — Local contacts.
CHAPTER XV.— The " B " Switch Section 93
Arrangement of apparatus — Junction clearing — Generator and
reed ringers — Transfer circuits — Visual indicator — Principle
and connections of transfer circuits.
CHAPTER XVI.— Up and Down Call Wire loi
Service circuits — Record table — Up and down call wires— Call
key — Record table tablet.
CHAPTER XVII.— The " C " Switch Section loj
Disposition of apparatus' — Transfer circuits — Local switch junctions
and call wire.
CHAPTER XVIII.— The Transfer Board 107
A and B circuits — Principle and design of board — Combination key
— A circuit diagram — B circuit diagram — Cross-connection
strips — Fault tracing,
CHAPTER XIX.— The Record Table Switch Section.. .. 115
Object and design — Method of working — Connections — Night calls
CHAPTER XX.— Direct Junction Circuits 118
Arrangements for terminating upon trunk switch sections — Record
table transfer section — ^Junction transfer sections.
CHAPTER XXI.— Call Office Circuits 123
Silence cabinets — Counter communication switch for one cabinet;
— for five cabinets — Stock Exchange circuits. ,
CHAPTER XXII.— Inductive Disturbances 130
Static induction — Effects upon circuits — Dynamic induction —
Combined effects of static and dynamic induction — Crossing
and symmetrical twist systems — Effects of faults.
CHAPTER XXIII.— Superimposed Circuits 137
Transformers — Principle of superimposing — Conditions necessary
—Signalling— Non-polarised indicator relay— Complete cpij-
nections — Testing.
ConUnis. Xi.
PAGE
CHAPTER XXIV.— Test Room Appliances 144
Test cases — Test boards — Lightning protectors — Line tablet —
Battery tablet — 40-circuit tablets — Switch-spring test boards —
Small protectors — Cross connections — Manchester arrange-
ments.
CHAPTER XXV.— The Universal Battery System .. .. 152
Object and theory — Secondary cells — Speaking fuse tablet — Charg-
ing switches — Overhearing — Permanent current working —
Connections and theory — Signalling fuse tablet — Primary
battery universal working — Test board arrangements.
CHAPTER XXVI.— Small Local Exchanges 160
Design of switch board — Connections — Junction arrangements.
CHAPTER XXVII.— The Series Multiple 165
Junction and multiple boards — Appearance of multiple board —
Engaged test — Connections — Working.
CHAPTER XXVIII.— The Self-Restoring Board .. .. 171
Principle of parallel multiple — Engaged test — Operating connec-
tions — Flat boards.
CHAPTER XXIX.— The Call Wire System 175
Principle of system — Single cord working — Engaged test — Operat-
ing connections — Night control switch — Service peg — Cross-
connection fields — Advantages and disadvantages.
CHAPTER XXX.— The Newcastle System 179
Distinctive features — Subscribers' station — Description of multiple
— Operating connections — Method of working — Short circuit
wire — Engaged test-«Call and detector switch — Trunk working.
CHAPTER XXXI. — Miscellaneous Special Arrangements . . 185
Trunk engaged tests at Manchester — Common return — Night
arrangements — Intermediate stations on trunk lines — Single
wire trunks— Supervisor's switch section — Special keyboard—
Cord testing — Auxiliary receiver.
CHAPTER XXXII.— Switch Section Faults 195
Method to be adopted in tracing — Faults in telephone sets — Dis-
connections in long lines — ^Trunk circuit apparatus faults —
Cords — Keyboards — Earths on junctions — Operator's tele-
phone — Local circuits — Record table switch sections— Transfer
circuits.
CHAPTER XXXin.— Recent Advances in Switchboard Design 200
Central battery system— Kellog system— Chicago express system—
Sabin's divided board system— Engaged subscribers— Automatic
ringing — Equal distribution of work.
APPENDICES—
A.— The KR Law 206
B.— Resistance and Capacity of Wires 209
C— Switch Section Cross-Connbction Strip Numbers . . 210
INDEX 214
THE TELEPHONE SYSTEM
OF THE
BRITISH POST OFFICE.
CHAPTER I.
Introductory.
The invention of the electric telephone dates back to 1854, when
M. Charles Bourseul published a paper in which the following remarkable
passages occur : — " Suppose that a man speaks near a movable disc sufficiently
pliable to lose none of the vibrations of the voice, that this disc alternately makes
and breaks the currents pom a battery ; you may have at a distance another disc
which will simultaneously execute the same vibrations." •'.... It is certain that
in a more or less distant future speech will be transmitted by electricity. I have
made experiments in this direction; they are delicate, and demand time and patience,
but the approximations obtained promise a favourable result." This undoubtedly
contained the germ of the electric telephone, but to Philip Reis, in
1861, belongs the credit of actually carrying the idea into effect. His
telephone took advantage of the fact that an iron bar when magnetised
emits an audible " click." If the rate and intensity of these clicks
can be controlled, musical notes and other more complex sounds may be
reproduced. For instance, a musical note is often produced by a circular
saw when cutting through wood. This is caused by the excessive
frequency with which the teeth strike the wood. Reis's transmitting
instrument consisted of a membrane with a contact arranged in its centre
by means of which the current was made and broken by every movement
of this diaphragm.
The first practical telephone was due to Graham Bell, in 1876, but
before describing this remarkable instrument, which is to-day very much
as originally invented, a little information upon the subject of sound will
prove advantageous, as it is impossible to have a clear idea of the
transmission of sound without a knowledge of the essential principles
involved. A sound is any disturbance such that, when communicated to the ear
B
2 Saund Waves.
and' transmitted by the nervous system to the brain, the sensation which we call
sound is experienced. A sound is produced by any action which throws the
surrounding air into a state of sufficiently vigorous vibration. When a tuning
fork is struck the prongs are thrown into a state of rapid vibration. Let
us for the moment confine our attention to one of these prongs. This
prong moves backwards and forwards at a great speed, and in so
doing alternately condenses and rarefies the air in contact with it. This
effect is passed on to the next layer of air, and this on to the'next, and so
on, the disturbance thus spreading outwards in every direction.
Wave motion consists in any regular periodic disturbance. Sound
waves are not up and down vibrations, but are propagated in straight lines.
The immediate cause of the sound is the vibration of the sounding body,
and these vibrations are communicated to the surrounding air, which is
carved into a series of waves of alternate condensation and rarefaction.
These waves are maintained so long as the sounding body vibrates.
The sound emitted by a tuning fork is greatest just when it is struck, and
if it is watched it will be seen that the amplitude of vibration, or the
distance through which the prong is moving, gradually grows less and less,
consequently the surrounding air will not receive such vigorous pushes,
and the sound will not be so loud. In fact, the subsidence of the sound
may be seen from the amplitude of vibration of the tuning fork as well as
heard. Now a tuning fork is an instrument so designed that when once
thrown into a state of vibration it will continue to vibrate at the same rate,
but with gradually decreasing amplitude. It can only vibrate at one rate,
and to this fact it owes its utility for tuning instruments by.
If the surrounding air were thrown into vibration by means of some
instrument which vibrates at the same rate as the tuning fork, the fork
would be thrown into vibration, emitting its own distinctive note. The
annoyance caused by a loose gas-globe when singing is an excellent
example of this. It is frequently only one note which will cause this
globe to vibrate, and trouble is sometimes experienced in finding the
offender. Now nothing of the nature of a tuning fork can be used for
receiving the vibrations caused by articulate speech, which vary in loud-
ness, in pitch, and in quality. The tuning fork is what may be termed a
persistent vibrator, and as has been already pointed out the amplitude of
vibration determines the loudness of the sound. The rate at which the
fork vibrates backwards and forwards determines the pitch of the note.
The reader will doubtless have noticed the great amount of vibration
which is actually felt in addition to the sound produced by the very low
notes of an organ. These vibrations are much lower even than the deep
bass of a man's voice, which latter may cause sixty or seventy sets of
waves of alternate rarefaction and condensation to succeed each other
in one second. In the case of a woman's voice the vibrations are much
nuicker. reaching, perhaps, one thousand per second.
Principles Involved in the Transmission of Speech. 3
There is yet another way in which sounds differ from one another, and
this is the difference which exists between a piano and a violin, both of
which sound the same note, and that is the characteristic of the sound.
This characteristic is called the timbre of the sound. Timbre is a French
word meaning tone, but conveying far more than its English equivalent.
The timbre of the voice is due, not to the amplitude of the wave or its
frequency, but to the shape of the wave. The vibration is seldom, if
ever, a simple regular backward and forward vibration, but may make
several small excursions before moving bodily over to the other side to
execute a similar vibration. Articulate speech causes waves of the most
complicated nature. The difference between a spoken word and a note
of the same pitch and volume is due to the difference in the shape of the
waves.
If these waves are allowed to impinge upon some mobile plate suffi-
ciently flexible to lose none of the vibrations, the plate would be thrown
into vibration, executing a series of movements which would be syn-
chronous with the movements of the sounding body. If these movements
could subsequently be reproduced in the plate we should have the
surrounding air carved into waves which would exactly resemble those
produced in the first instance by the sounding body, and these waves
would produce the same effect upon our auditory senses. The principle of
the transmission of speech to a distance depends upon our ability to cause
a plate or disc to vibrate in exactly the same manner as a distant disc
which is thrown into a state of vibration by means of the waves produced
by the human voice. A certain toy, which the majority of my readers
have seen at some time or other, may serve to illustrate the above pro-
position. The toy consists of a couple of tins with the bottoms knocked
out and replaced by paper diaphragms. The centres of the two drums
are connected by means of a tightly stretched string. On speaking into
one tin the paper diaphragm is alternately slackened (moving inwards)
and tightened (moving outwards). The movement alternately slackens
and tightens the string, and the distant diaphragm is slackened
and stretched— !.«., if the pull on the string is slackened the
diaphragm moves inwards, and if tightened outwards. Thus whatever
movement is made by one diaphragm is reproduced at the other. The
sound waves impinge on the first diaphragm, causing it to execute vibra-
tions in unison with those waves. The distant diaphragm executes
vibrations in synchronism with the first one, therefore causing waves
precisely similar to the original ones. Its effects upon the auditory senses
are also similar.
The electric telephone merely consists of an electro-magnetic mechanism
by means of which two diaphragms are made to control each other so
that a movement of one causes an exactly similar movement of
the other.
B 2
The Telephone System of the British Post Office.
CHAPTER II.
The Bell Telephone.
In order that the modus opiranii of the Bell telephone may be clearly
apprehended, it is essential that we should feel perfectly familiar with the
theory of the electrical phenomena upon which it depends. As the
majority of my readers are familiar with the fundamental principles of
magnetism and electricity, it is proposed to pass quickly over them.
In the region around a magnet there is a field of force, since a compass
needle will be deflected. To produce this deflection force must be
required, since the needle is held in its north and south position by the
earth's magnetic fleld. It is not possible for any thinking person to
believe that action at a distance, without any connection between, is
possible. The magnet creates some peculiar disturbance in the ether*
which permeates all matter, and this disturbance in the ether may be
made to re-act on magnetic bodies, and thus produce an effect. To the
practical man these theories are of very small moment, but if, from a con-
sideration of these points, we can deduce a working hypothesis, we shall
not have wasted our time. Faraday explained nearly all the magnetic
phenomena by means of " fe'««s of force." The amount of magnetic dis-
turbance at any point will depend upon the strength of the magnet originat-
ing the disturbance, and upon its proximity. Clerk Maxwell gave to
these lines of force definite quantative values, that is to say, the strength
of a magnetic field was expressed in lines of force present. A certain
magnetic field was chosen, and this was termed -a. unit magnetic field.
Through this unit field we suppose one line of force passes in every unit of
area. A field twice as strong would have twice as many lines per unit
area, and so on.
Three effects may be produced by an electric current, viz., the magnetic,
the thermal, and the chemical effects. The magnetic and thermal effects
are always present when there is an electric current, but chemical effects
are not produced unless the current is made to pass through a conducting
chemical compound. The thermal effect is not noticeable unless special
means are taken to render it prominent. A wire conveying an electric
current is surrounded by a magnetic field. This magnetic field we repre-
sent by lines of force, as shown in Figure i. If we wish to increase this
* Lodge's ''Modern Views of Electricity" contains a non-mathematical and thoroughly
readable account of these ideas.
Production of Electric Currents. 5
magnetic field, we may either increase the strength of the current, or we
may wrap the wire round and round into the form of a coil or solenoid,
thereby bringing a greater length of current-conveying wire together.
This results in an increased number of lines of force, as every added inch
of wire adds its lines to those already there. The necessity for bringing
the wire together is merely to concentrate the effect.
There are many ways in which an electric current can be produced.
Firstly, there is the chemical method, i.e., by means of a voltaic cell. A
voltaic cell is an arrangement which in return for chemical energy vrill
furnish electrical energy. Such cells usually consist of two dissimilar
metals, immersed in some conducting liquid, which will attack one of the
metals, or elements as they are usually called in this connection.
Secondly, there is the electro-magnetic method, and this is the method to
which it is desired that careful and particular attention be paid. When-
ever a conductor is cut by a line of force, or whenever a line of force cuts
a conductor, an electromotive force is generated in that conductor, and if
the ends of the conductor are joined by a wire (while the lines are cutting
the conductor) a current will flow through the circuit thus formed during
the time that the lines are cutting the conductor. Generally, we may say
that whenever any alteration in the number of lines of force embraced by
a conducting system is made an E.M.F. is set up. The value of the
E.M.F. generated depends upon the strength of the magnetic field, and
upon the rate at which the conductor is cut. For instance, the absolute
unit of E.M.F. is the E.M.F. produced between the ends of a conductor,
6 Theory of the Bell Telephone.
which cuts a unit magnetic field at unit rate — i.e., that cuts one line of force
per second. In a unit magnetic field there is one line of force per unit
area. The volt, or practical uilit of E.M.F,, is made very much larger
than the absolute or C.G.S. unit. The direction of the current depends
upon the direction in which the lines cut the conductor.
The Bell telephone consists of a mobile iron plate placed in a
magnetic field with a system of conductors, e.g., a coil of wire. This iron
plate, or diaphragm, usually of ferrotype, is made very Ihin, so as to be
flexible. Sound waves impinging upon this diaphragm result in movement
or vibration, and this movement causes an alteration in the distribution of
the lines of force in the magnetic field. The coil of wire is so arranged
that this disturbance in distribution results in some of the lines cutting
through the conductors, and as we have seen that whenever a line of force
cuts a circuit, a current is the result.
,t»w«^
Every movement of the diaphragm causes an electric current. The
magnetic field will be disturbed to the greatest extent at or near the
poles of the magnet : consequently, the coil of wire is made to surround
one pole of the magnet (as shown in Figure 2) . The lines we may imagine
as moving laterally through a small distance. In order that by these
movements the lineg may cut as many turns of wire as possible, we must
make our coil of wire very small ; that is, bring the greatest possible number
of conductors into the region where the maximum disturbance takes place.
We see then, from these considerations, that very fine wire must be em-
ployed, and that if the space taken up by the wire is large,' there will be a
large number of turns of wire which will remain in a practically constant
field, and will therefore merely be useless resistance. The coil must
have the largest possible number of turns which can be got into the space.
Beyond this space additional turns of wire will be worse than useless.
Theory of the Bell Telephone. 7
It is now necessary to show clearly how movements of the diaphragm
produce currents proportional to those movements, also how these currents
again reproduce upon a second diaphragm the movements of the first. The
Y
E
• Figure 3.
magnetic circuit, it will be seen, includes the diaphragm. The lines of
force starting from the magnet pass through the air gap between the
magnet and the diaphragm into the diaphragm, then out again, and back
to the other pole of the magnet. The diaphragm is consequently attracted
and is thus strained inwards to the magnet. The distance to which the
centre of the plate is pulled in depends upon its thickness or stiffness, and
upon the strength of the magnet. There is a very delicate balance between
the two, the diaphragm remaining in such a position that the pull of the
magnet inwards is exactly equal to the "fly-back" tendency of the
diaphragm, which is tightly clamped round its outer edge. When sound
waves impinge upon this delicately-balanced diaphragm, vibrations which
are in synchronism with the sound waves are executed by it. Let us
suppose that the effect of the first wave is to move the diaphragm
further inwards. The nearer approach of the diaphragm causes several
lines of force, which formerly entered the coil, to pass straight from the
magnet to the diaphragm. It will also cause fresh lines at the sides to
pass through the coil. These lines in taking up their nejv positions cut
through the conductors of coil, and, as we have seen, an electric current is
the result. The strength of this current will depend upon the amount of
movement which the diaphragm executes. In this instance the lines cut
through the coil from the sides to the centre. A rarefaction wave will
cause lines which formerly passed through directly from the magnet to
the diaphragm to move towards the sides, thus cutting the coil of wire in
the opposite direction, and so producing a current in the opposite direc-
tion. From a consideration of these points, it is obvious that currents of
varying strength and direction, in direct proportion to the strength and
direction of the sound waves, are sent out, and thus complex sounds re-
sult in complex currents.
AS we have seen (Figure 3) the transmitting instrument is precisely the
same as the receiver, or, in other words, each instrument will, in turn, act
as either transmitter or receiver. The undulatory currents which pass
through the coil of the receiving instrument alter the distribution of the
8 Mercadier's Researches.
lines of force in precisely the same manner in which they were altered by the
movement of the transmitting diaphragm ; hence the receiving diaphragm
executes precisely the same movements as the transmitting diaphragm. For
instance, a current in one direction tends to reduce the strength of the
magnet, causing the diaphragm to move outwards : whereas a current
in the opposite direction would act with the magnet, and thus bring about
the near approach of the diaphragm. The effects produced are directly
proportional to the strength of the currents ; therefore the movements
executed are in exact synchronism with the movements of the transmitter,
and thus the exact sounds will be reproduced — i.e., the movements
of the receiving diaphragm will carve the air into waves, which are pre-
cisely similar to the original waves. It is quite Immaterial in which way
the two telephones are joined up, since it will only mean that the diaphragm
will move inwards to the same extent that it might, if carefully joined up,
move outwards. This will not alter the sound produced.
The energy of the sound waves is, in the transmitter, converted into
electrical energy in the form of current. A certain proportion of this energy
is wasted in overcoming the resistance of the line. The greater proportion
of the energy is utilised in moving the diaphragm at the distant end. Here
then there are several transformations of energy ; hence there is bound
to be a proportion lost ; that is to say, the reproduced sounds cannot be as
loud as the original sounds. They will be precisely similar in all save
this one respect. If the line is very long, then the sounds will be almost
inaudible. The energy of the voice is used to produce 'the distant effect
in the Bell telephone, and therefore the arrangement is not as efScient as
those arrangements in which the received sounds are produced by some
secondary set of arrangements in which the energy of the voice is merely
used to direct.
The question as to the relative dimensions of the various parts of a
telephone is i.n exceedingly important one. Unfortunately, the ex-
planation of' the Bell telephone is by no means satisfactory, and it is
not possible to treat the subject with the same certainty that we may
treat an action which is thoroughly understood. It is not, at present,
possible to use our mathematical knowledge to determine the best possible
dimensions. However, the following general rules have been deduced
by Mercadier : —
(a) The stronger the magnetic field the thicker should the diaphragm
be.
(b) Having ascertained the correct thickness of the diaphragm there
is a certain diameter which will give a maximum effect. The stronger
the magnetic field, the larger should the diaphragm be,
(c) The relative positions of the coils, magnets and diaphragms must
then be determined by considering what arrangement will, for the smallest
movement of the diaphragm, produce a maximum disturbance of the
Design and Theory. g
magnetic field. The coils must then be so arranged as to be cut by as large
a number of lines as possible for a given alteration in the magnetic field.
It would at first sight appear that we might indefinitely increase the
sensitiveness of our receiver by increasing the strength of the magnet,
but this is not so, as the diaphragm loses much of its elasticity under the
stronger attraction between it and the magnet. These general remarks
as to designing telephones should be borne in mind when considering the
various forms of receivers. Mefcadier's researches led him to design
telephones with exceedingly small diaphragms, whereas the telephones first
introduced had diaphragms over three inches in diameter. Present day
practice, as we shall see, seems to indicate a medium between these two
extremes.
It was previously remarked that the theory of the Bell telephone was
by no means satisfactory. It is not conceivable that the exceedingly
small currents generated by the transmitting instrument will be sufficiently
powerful to cause the distant diaphragm to move bodily backwards and
forwards, thus generating the vigorous sound waves which we recognise.
It would seem that the slight alterations in the magnetic field at the
receiving end were out of all proportion to the magnitude of the effects
produced. It is suggested by Du Moncel,* who has made a special study
of the theory of the telephone, that there are many causes which act
together to reproduce speech. There is firstly the bodily movement of
the diaphragm such as has already been described. Secondly, it has been
suggested that there are molecular movements' taking place in the
diaphragm. Thirdly, the molecular vibrations of the magnet consequent
upon its alternate magnetisation and demagnetisation by the undulatory
currents. In regard to this last effect it should be mentioned that when
an iron rod is magnetised or demagnetised it alters very slightly in shape,
at the same time emitting a slight " tick." Telephones have been con-
structecl without diaphragms upon this principle. It will be obvious that
a series of ticks will form a musical note if they succeed each other with
sufficient rapidity. The magnitude of the tick produced depends upon
the strength of the current producing it. It will thus be seen that the
action of the telephone is somewhat obscure.
♦ Du Moncel, "Le Telephone," 4me edition.
10 The Telephone System of the British Post Office,
CHAPTER III.
Early Transmitteks.
In the Bell telephone the energy of the voice is used to produce the
distant effect, therefore the effects produced cannot be as strong as
the original ones. In order to increase the effects, Edison in 1877
invented his carbon transmitter. This was the first of the carbon trans-
mitters, and opened quite a new vista in telephonic research. The
instrument is illustrated in Figure 4. It consists essentially of a piece
Figure 4.
of carbon so arranged that sonorous vibrations will cause varying
degrees of pressure upon it. The electrical resistance of carbon decreases
under pressure, and upon this fact it was thought that the action of the
instrument depended. A ferrotype diaphragm DDi is used to receive the
sonorous vibrations. These vibrations are then communicated to a
platinum plate G by means of the ivory button C. Between the platinum
plate and the table BE is placed a circular disc of lamp black F.
The table B is capable of adjustment by means of a screw thread
which is cut upon its longer extremity. In this way the pressure upon
Edison Transmitter. ii
the carbon disc can be varied at will. The circular ring H, which is
screwed on the table B, serves to keep the carbon button in position.
The case of the instrument is made of iron, and forms one terminal and G
the other. When the diaphragm DDx is thrown into vibration, these
vibrations are communicated to the platinum plate G, which in turn alters
the resistance of the instrument. It is now considered that it is merely
the resistance of the contact between the platinum plate and the carbon
which is altered, and not that the resistance of the carbon itself is reduced
by being alternately compressed and relieved of compression. However,
this is a point to which it will subsequently be necessary to return.
The Edison transmitter, then, so far as we have now considered it,
consists in an apparatus the resistance of which will vary in exact
synchronism with the sonorous vibrations which we cause to impinge
upon its diaphragm. If now a battery and a Bell telephone be joined in
circuit with this instrument, and we speak on to it, we shall have varying
currents passing through our receiver. These currents are in every way
proportional to the effects producing them, consequently we shall have
speech perfectly reproduced. It should be noted that in this case we do
not employ the energy of the voice to produce the distant effect. It
merely directs the movement of the diaphragm, and the battery which
we employ furnishes the energy utilised at the distant end, so that it is
theoretically possible to have the effects produced at the distant end
greater than the original effects. Now let us consider what the proportions
of this circuit should be as regards resistance, In order that we may obtain
maximum variations of current from changes in the resistance' of the
transmitter. Ohm's law states that the current in any circuit vari s
directly as the E.M.F., and inversely as the resistance. With a constant
E.M.F., then, the current through a circuit varies inversely as the
resistance of that circuit. Now, in order that the variation of resistance of
one fart of that circuit may produce a large variation in the value of the
current, we should desire that the remainder of the circuit should have as
little resistance as possible. For instance, let us suppose that the resist-
ance of our transmitter in its normal condition, battery, and receiver
amounts to 20". of which the transmitter is responsible for lo"- If now
the note C is sounded, let us suppose that the resistance of the transmitter
varies between g" and ii<» at the rate of 240 variations per second.
Consider now one of these variations — for instance, the decrease in the
resistance of the transmitter to qm- In this case the resistance of the
circuit is 19M1 as against 20" when normal. This represents a variation
of 5 per cent, of the original current, Had the resistance of the remainder
of the circuit been nil, the variation would have been 10 per cent. On
the other hand, if the line were of considerable length the variation would
have been far less. Had the resistance of the circuit been iooo<°, the
variation would only have been i-ioth per cent., or the percentage
12
Induction of Electric Currents.
variation of current would only have been one hundredth of the percentage
variation of resistance of the transmitter. Thus vfe see that the longer the
line the weaker is the speaking, since the speaking depends not upon the
current flowing through the telephone, but upon the extent of the variations
produced in the value of the currents passing through the instrument. If the
line is of any length, then such an arrangement as has been depicted would be
utterly useless.
Figure 5.
The solution of this problem was found in the induction coil. The
induction coil consists of an electro-magnet around which a second set of
windings are placed. The first set of windings is termed the primary coil,
and the second set the secondary coil, and the currents sent through the
primary induce currents in the secondary. In order that we may appreciate
the reason of this we must refer back a little.
It was previously stated that whenever a line of force cut a conductor
an E.M.F. was generated in that conductor. Figure 5 illustrates a coil
of wire joined to a galvanometer. A magnet N S is placed above, and a
Theory and Utility of Induction Coil. 13
few lines have been drawn to represent lines of force. If the magnet is
moved down into the coil of wire the magnet's lines will cut through the
various turns of wire ; thus an E.M.F. will be generated in each. These
E.M.Fs. add together and produce a kick upon the galvanometer, but the
effect is only a transient one — it lasts only while the magnet is moving.
Immediately the magnet ceases to move the lines cease to cut through the
conductors and the E.M.F. ceases. When this magnet is withdrawn the
lines of force again cut through the several conductors, but in the opposite
direction, viz., from bottom to top, and a current in the opposite di-
rection, lasting only so long as the magnet is moving, is the result. It
will at once be remarked that in order to prolong the current the magnet
should be moved very slffwly. That is so, but the E.M.F. induced varies
directly as the rate at which the conductors are cut, consequently if we
do this the current produced will be correspondingly small, though lasting
longer. If now we wished to produce a series of currents in opposite
directions we might devise an arrangement by means of which N S is
moved rapidly up and down. The essential point is that the lines of force
shall cut and re-cut the coil of wire.
If the magnet N S be replaced by an electro-magnet, we may magnetise
or demagnetise it at will by making or breaking the circuit of a battery
connected to it. In this case we should prefer to place the magnet actually
inside the coil in order that a maximum number of lines shall cut through
the windings of the coil on making and breaking the circuit. On making
the circuit it is obvious that these lines of force have to spring into exist-
ence, and that in doing so they will cut the windings of the coil, thus
generating an E.M.F. When the circuit is broken the lines have to col-
lapse, as it were, back into the iron again, cutting through the coil, but in
the opposite direction, thus generating an E.M.F. in the opposite sense.
If now the current is increased in strength more lines of force are added to
the magnet, and these lines in springing into existence cut through the coil,
thus generating an E.M.F. When these lines fall back into the iron a
current in the opposite direction results. If an Edison transmitter and
battery be included in the circuit with the electro-magnet (or primary
circuit), currents will be generated in the secondary coil which, by their
direction, indicate increase or decrease in the resistance of the transmitter
and by their strength indicate the amount of variation which takes place.
The advantage of this system is that the transmitter, battery and primary
coil only are in circuit. The line circuit is quite. separate.
We may make the resistance of our battery low, and we may also make
the resistance of our primary coil low, since we can provide a heavy cur-
rent. One quarter of an ampere through 40 turns produces the same mag-
netic effect as 10 milliamperes through 1,000 turns. At the same time we
can employ thicker wire for our ten turns than we could possibly do if we
had to provide a thousand. The E.M.Fs. generated in the secondary coil
14
Hiighes Microphone.
axe directly proportional to the current variations in the primary. The
resistanceuDf the secondary circuit, which includes the lines and receiver,
is constant ; therefore the current produced in it will be directly propor-
tional to the E.M.Fs. induced in it. We have thus accomplished our object,
for we have made the primary circuit of low resistance, the transmitter
itself being the principal item, and thus variations in the resistance of the
transmitter are not masked by a large constant resistance. To put the
matter plainly and simply, the difference is something like the difference
between throwing a cup of water into a small basin of water and throwing
it into a pond. In the latter case we do not observe any very large altera-
FlGURE 6.
lion in the quantity of water in the pond, whereas in the former case we
may cause the small bowl to overflow.
Professor Hughes, in May, 1878, read a paper before the Royal Society
which greatly added to our knowledge of the subject of telephony.
Professor Hughes disco\fered that any system of loose contacts acted as k
telephone transmitter. He discovered that three iron nails laid upon one
another, as indicated in Figure 6. would transmit speech perfectly, and
gave to all such instruments the name of " microphone." This was done
as it was thought that these transmitters actually amplified the original
sounds, but this is an exceedingly doubtful point. An explanation of the
Principle of Action.
15
action lies in the fact that these loose contacts vary in resistance when
shaken by sound waves which are directed towards them — i.e., the
resistance of the arrangement varies in exact synchronism with the
sound waves. Now, whether the explanation of the phenomenon lies in
the fact that when the joints are shaken the pressure on the nails is varied,
and that, therefore, the resistance is varied, is somewhat doubtful, The
nails are not very efficient as a transmitter, but such sounds as the ticking
of a clock and the pitch of a note are reproduced, but the timbre is not.
Professor Hughes experimented with various substances, but finally found
that carbon acted far better than any other substance. The most common
form of carbon instrument is shown in Figure 7. It consists of two
carbon blocks, between which is fitted a third carbon rod with tapering
points. The carbon pencil is fitted loosely, touching both blocks. Sound
Figure 7.
is perfectly reproduced by this instrument, notwithstanding its simple and
primitive look. The question now arises — Why should carbon behave so
much better than iron ? Shelford Bidwell investigated this subject, and
came to the conclusion that the effect was not solely due to the improve-
ment and deterioration of the electrical contact between the two
conductors. The electrical resistance of carbon, unlike that of all metals,
decreases when the temperature is increased. The passage of a current
through carbon decreases its resistance by increasing its temperature.
The increase of temperature is obviously greatest at the points of contact
where the resistance is greatest. Let us consider what would take place
when the contacts were slightly shaken apart by a sonorous vibration.
Firstly, the contact is rendered worse, and the current is thereby reduced,
therefore the temperature at the points of contact is reduced. The effect
i6 Design of Induction Coil.
of the variation of contact is thereby greatly enhanced. When the current
is increased the temperature is increased, and the resistance of the contact
thereby reduced. It is also thought that some of these effects are due to
the formation of minute arcs at the points of contact. These theories are
supported by the fact that a microphone grows sensibly warm Under
continued use, and, further, that if the battery power is increased beyond
a certain point hissing sounds result.
We thus see that a microphone consists in any apparatus which will,
when subjected to sonorous vibrations, vary in resistance. To such
contacts Professor Hughes gave the name of " microphonic joint." They
may be formed in hundreds of ways, both by intention and by accident. A
telephone switch arm with a defective spring, or a loose joint in a gutta-
percha covered conductor, will occasionally form such a joint. Carbon is
infusible, inoxidible, a bad conductor, and its resistance decreases with
increment of temperature ; and to these properties it owes its special utility.
It has been pointed out that the core of an induction consists of a bundle
of thin iron wires. The object of this form of core is to meet two diffi-
culties, the first being that due to residual magnetism, and the second
that due to a solid core acting as a secondary circuit. The current in
the primary is uni-directional, and thus residual magnetism is very much
in evidence. Residual magnetism tends to mask and reduce the changes
in the magnetisation of the core and thus to reduce the values of the
E.M.Fs. generated. If the magnetic circuit were completed as in a
transformer (see Chapter XXIII.), the values of the residual magnetism
would be far higher.
When a current is induced in the secondary of an induction coil, a
magnetic field is created due to that current in such a direction as to
weaken the inducing field (viz., that due to the primary). If a hi avy
copper cylinder is placed around the primary of an induction coil
little or no current will be induced in the secondary, as our copper
cylinders will take up the majority of the available energy. Now the core
of the primary circuit acts like a closed secondary circuit if composed of
solid iron, The iron wires, owing to a thin coating of oxide, are not in good
electrical contact, and therefore their effect in taking up energy which we
desire only in the secondary circuit is not considerable. In a perfect
induction coil the energy given to the primary circuit will all be repro-
duced in the secondary circuit or circuits, and if one of these circuits is
taking up energy, the smaller is the amount of energy at our disposal in
the other circuit.
The Telephone System of the British Post Office. 17
CHAPTER IV.
Modern Receivers.
The Bell receiver is very much to-day as it was in 1876, when invented.
Of course, the instrument has been vastly improved, but the improvements
consist rather in matterstif detail than in the essential parts. One of the
earlier modifications was the receiver of the Swiss administration. In
this instrument the simple bar magnet was replaced by a compound
magnet formed by placing eight thin bar magnets side by side. Into the
end of the magnet was fixed a soft iron core, and over this a small bobbin,
containing about 2,500 convolutions of No. 38 wire, was placed. The re-
sistance of the instrument was 100 ohms.
Now let us consider in what respect this receiver differs from the original
Bell receiver and what is the efifect of such differences. Firstly, a com-
pound magnet can always be magnetised to a higher degree than a bar
magnet of equal size. This improvement then results in a stronger mag-
netic field. Next the iron core upon which the wire is wound is only
3-i6ths of an inch in diameter, and the coil of wire round it has a diameter
of 5-8ths of an inch. The lines of force are concentrated in the iron core,
as nearly the whole of the lines of force from the compound magnet
pass through it. Thus we have a very intense magnetic field round the
coil of wire, and the slightest motion of the diaphragm modifies the distri-
bution of the lines. The coil is very small, and therefore a large number
of lines cut through it at every movement of the diaphragm. The receiver
is an excellent one, though not used so much now as formerly, it having
given place to the double pole receivers.
The idea that increases in the strength of the magnetic field always
resulted in increased efEiciency in the receiver led to the construction of
Phelp's " Pony crown " receiver, in which six magnets were used. These
magnets were bent into circular form, the similar poles being attached to
the iron core round which the wire was wound, and the other poles
touching the edge of the diaphragm. The instrument was, however, not
a great improvement upon the original Bell, and is certainly inferior
to many simpler modifications in use to-day. Besides, the shape of the
instrument is objectionable.
A telephone was constructed by Gouloubitzky with four coils and two
circular magnets. The two ring magnets terminated in four coils, the
opposite cores being of opposite polarity ; in short, the instrument con-
C
1 8 Ader Recetvev.
sisted of two telephones with four coils acting on one diaphragm. The
reasoning which led to its construction was based upon the fact that when
several receivers are joined up together to receive music, etc., the sound is
quite as loud on any one of the telephones as when only one is joined up.
It was thought that the combining of two receivers in one would result in
double the volume of sound. This was not the case, however, and this
and many kindred forms of receiver have now disappeared from the region
of practical telephony.
The Ader receiver is one of the most sensitive receivers in use to-day,
and though not very extensively used in England certainly merits de-
scription. It consists of a ringed-shaped circular magnet (E, Figure 8)
Figure 8.
the poles of which terminate in soft iron cores round which two coils
C and Ci are wound. Above the diaphragm D is placed a circular ring
of iron B B^, the object of which is to concentrate the magnetic field.
This ring is termed the "surexcHateur" (over exciter). The term is a
somewhat awkward one to translate, but its general meaning will be
gathered from the foregoing. This ring reduces the magnetic resistance
of the magnetic circuit and consequently increases the number of lines of
force passing through it. Du Moncel found that the nearer the mass of
an armature approached that of its magnet the greater was their
mutual attraction— i.e., the greater is the magnetic field between them ;
consequently if having designed our ring and magnet upon these lines we
place our diaphragm between them we shall have it placed in the strongest
Adev Hecelver.
t9
magnetic field which it is possible to obtain with this particular magnet.
If the diaphragm itself is made heavier it loses much of its flexibility.
Figure g.
The ring answers precisely the same purpose without affecting the flexibility
of the diaphragm, amd to this may be traced its greater efficiency. The
magnet is usually nickel-plated, and thus presents a very good appearance.
Figure io.
Figure 9 shows the external appearance of the Ader receiver made by
the Consolidated Telephone Company.
20
Gower Receiver,
Another most successful modification of the Bell telephone is the Gower
receiver. It consists of a large semi-circular magnet, NS (Figure lo), the
poles of which terminate in two soft iron cores turned downwards and
fixed into the magnet at right angles to it. The diaphragm is placed above
the cores, but instead of having the instrument separate and distinct from
the transmitter it is usually fixed, as shown in Figure ii, and the sounds
are conveyed to the ears by means of flexible tubes. In the construction
of the instrument we shall notice several features in which the instrument
differs from other modifications of the Bell telephone. The diaphragm is
very large, being three and a half inches in diameter, and is correspond-
ingly thicker. The magnet is very massive and compact, being half an
inch in thickness, three-quarters of an inch broad, and nine inches in length.
The receiver is a most efficient one, the speaking being very loud, but
the flexible tubes are somewhat expensive to maintain. This instrument
was formerly used by the Post Office to the exclusion of all others, but
now the Gower receiver is replaced by two double pole Bell receivers.
The complete Gower Bell Telephone Station will be described subsequently.
£,L.£\/ATION
Figure ii.
The double pole receiver was invented by Bell, but there is a vast
difference in detail, though not in principle, between his instrument and
those in use to-day. The Bell instrument was large and heavy, and was
fixed upon a stand. Siemens modified this further, rendering it lighter
and handier, and from this instrument the present form of double pole
receiver has been evolved. This instrument is used by the Post Office,
for whom it is frequently manufactured by the General Electric Company.
A section of the instrument is shown in Figure 12.
The magnet A A is of the horseshoe shape with parallel 4imbs, and
terminates in two soft iron cores P P, which are secured to it by means
of two screws. Upon these cores the coils B B are wound. They are
wound with No. 38 silk-covered wire, and have a resistance of i2o»> . The
diaphragm D D fits on to the top of the metal case, and is held firmly in its
position when the mouthpiece E is screwed down, as shown in the above
diagram. The position of the magnet with respect to the diaphragm
Double Pole Bell Receiver.
21
D D Is readily adjustable by means of the screw S. The instrument
illustrated is the metal-cased receiver, the shaded portion of the case
being metal and the black part ebonite, but the receiver specially made
for the Post Office does not in any essential feature differ from this.
It is, however, encased completely in ebonite, as shown in Figure 13,
and is much lighter. Prior to the adoption of this latter form the metal-
Figure 12.
cased receiver was employed. The double pole receiver is certainly one
of the most successful and efficient modifications of the Bell instrument,
and it is light and of a convenient shape and size.
The Collier-Marr receiver, which has two diaphragms, is more sensitive
but its somewhat awkward shape and heavy weight has prevented its
22
CoUier-Marr Receivev.
genera,! adoption for commercial purposes. It consists of a large horseshoe
magnet terminating in two soft iron pole-pieces. A diaphragm is placed
close to each pole-piece, and between these diaphragms is the somewhat
large coil of wire. This coil of wire is wound upon a bundle of iron wires,
the centre portion of which is replaced by a round rod of ebonite. The
powerful action of the instrument may be due to a resonant action between
the two diaphragms. It is wound with a very large number of turns of
wire having a resistance of 350" . The efficiency of this instrument may
be gathered from the fact that upon one occasion it was found possible to
speak to the Head Post Office from an underground box 800 yards distant
thro.ugb a pipe containing some thirty wires, many of which were actually
Figure 13.
working Wheatstone, and this with an earthed return wirel This
experiment was tried almost as a joke, but since then the instrument has
been much used in Manchester for this purpose. It is, however, true
that no double pole receivers have been wound to this resistance with the
other parts on a large scale, so that it is scarcely fair to term it the best
receiver.
There are many receivers of the " watch " pattern in use to-day. They
are exceedingly small and compact, being, as their name implies, about
the same size as a watch. The receiver adopted by the Post Office
for the use of linemen is of the watch pattern. The instrument is used by
the linemen to speak from a pole to the Telephone Exchange, so as to
Watch Receiver.
23
locate faults with certainty where there is no intermediate test-box.
A watch receiver is shown in Figure 14. The magnet consists of a ring of
steel, from opposite sides of which come pole-pieces around which are
wound two coils of wire. The circular magnet may be regarded as two
bent bar magnets placed with similar poles together.
Figure 14.
The instrument is very efl5cient, and is frequently provided with an oval
ring fixed at right angles to the back of the case, thus making it some-
what resemble an Ader receiver. In some forms of this instrument a
siir excitateur is also added.
24 The Telephone System of the British Post Office.
CHAPTER V
Modern Transmitters.
There are many different forms of transmitters now in daily use, but
they may be readily divided into three classes : —
I. Modifications of the Hughes Microphone (carbon only).
II. Modifications of the Edison Transmitter (carbon and metal).
HI. Granular Transmitters.
The first two classes are very nearly identical — in fact, every transmitter
might with propriety be regarded as a modification of the Hughes
instrument. The third class is very different in form, to the first two,
and to it belong many of the most efficient instruments now in use.
Class I. — Modifications of the Hughes Microphone.
The first transmitter of this class was probably the Crossley, in which
four carbon pencils were used. It was for a time used by the Post Office,
C*
3=
Figure 15.
but rapidly gave way to the more powerful Gower. The Cower consists
of a central block of circular carbon C (Figure 15), from which radiate
eight carbon pencils, each terminating at a carbon block. The four right
hand blocks are connected by a strip of copper, and form one terminal of
Gower, A der, and Mix and Genest's Tfansmitters. 25
the instnimeat, and the left hand four blocks form the other. This
system of pencils and blocks is fixed upon a pine board nine inches wide
by five inches deep and one-eighth of an inch in thickness. This pine
board acts as a diaphragm, and is fixed inside the box immediately below
the porcelain mouthpiece. The action of the transmitter is precisely
similar to that of the original Hughes microphone, save that as more
contacts are employed its action is rendered more uniform and more
reliable. Movement of the diaphragm causes disturbance of the contacts,
and thus increases or reduces the resistance of the instrument. For
instance, when the diaphragm moves downwards the centre blocks and
the outside blocks move downwards, but owing to inertia the carbon
pencils do not immediatelytollow, and this causes a looser contact, which
results in increased resistance. When the diaphragm moves upwards the
pencils tend to remain in the position they occupy ; thus pressure is caused
between the blocks and the pencils, and thus the contact is improved —
i.e., its resistance is decreased.
There are a very large number of carbon pencil transmitters in use, of
which the Gower may be taken as a prototype. The Ader transmitter
consists of three parallel blocks of carbon, between which are placed
twelve parallel carbon pencils, six on each side, the two outer blocks form-
ing the terminals. It, however, possesses no advantage over the Gower.
A pencil transmitter much used in Germany is that made by Mix and
Genest, in which two parallel carbon blocks are bridged across by three
carbon pencils. In this case the transmitter is placed in a vertical
position. The pencils are held up by means of a sort of felt brake, which
prevents the pencils from remaining on the bottom contacts of the two
carbon blocks. This is a great advantage, as it entirely eliminates the
jarring sounds present in all other pencil transmitters when placed in
the vertical position. The jarring is due to the pencils rolling
along the contacts. The brake effectually prevents revolution of the
pencils, yet leaves them sufficiently free. Of the carbon transmitters
this latter instrument is the only one which presents any very novel
feature. Many modifications have been introduced merely for the sake of
avoiding patents, etc. If the pre cess of cutting bread with a knife were
patented someone would immed ately discover that the process of cutting
was more neatly and effectively carried out by means of a saw, which he
would be happy to supply at a cheap rate. Similar remarks apply to
many transmitters. If we thoroughly understand our prototypes we shall
have no difficulty in realising the modus opermidi of any instrument which
we may chance to come across.
There is yet another class of transmitter used somewhat upon the
Continent, known as pendulum transmitters. The Berliner microphone
may be taken as the prototype of these instruments. It consists of a
ferrotype diaphragm, to which is secured a cylindrical carboji button. A
26
Blake Transmitter,
second button rounded off at the end-and carried by an articulated lever is
held against the first button by its own weight. The diaphragm is con-
tained in a circular indiarubber ring, and a spring coated with indiarubber
is also employed to prevent persistent vibration.
Class II. — Carbon and Metal Transmitters.
The Edison was the first of this class. Usually a ferrotype diaphragm is
employed to collect and mechanically transmit the sonorous vibrations to the
electrical system of the transmitter. The Blake transmitter, which belongs to
Class II., is even to-day in very extensive use in this country. It consists of
a carbon button (Figure i6) carried by a spring and holder B, and held
opposite the centre of a ferrotype diaphragm D, the vibrations of which
influence the contact between the carbon button and a platinum point
carried by a second spring S. The two springs are mounted on an adjust-
able frame F, which is secured to the two pillars W and W It is secured
to W by means of a spring M fixed to both. The screw V serves to regulate
the amount of pressure between the platinum point and the carbon button.
The diaphragm is fitted into the indiarubber ring RR', and is held in
position by means of two springs E and E' (Figure 17) fixed to the circular
Blah Transmitter.
27
ring which carries the two pillars to which M and V are secured. It
should be noted that the various parts in Figures r6 and 17 are labelled
to correspond. The action of the instrument is explainable upon the
ground that the contact between the carbon button and the platinum
point is improved and deteriorated by movement of the diaphragm. The
indlarubber ring is most useful in that it prevents a vibration from con-
FlGURE I
tinning after the originating sound has ceased. If a long loose chain is
shaken from one end the movement travels slowly to the end. The chain
is, in fact, moving after Qne has ceased to shake it. If our diaphragm
28
Hunnings Transmitter.
vibrates after we have ceassd to send sound waves on to it, we shall have
jarring and blurring of sounds, and to get rid of this the indiarubber ring
is employed. It effectually damps out these after-vibrations. Our dia-
phragm must be the exact converse of a tuning fork, which latter is a
persistent vibrator. The whole transmitter is contained in a square box,
the front of which is hinged, and a bole is cut and turned to serve as a
mouth-piece. The external appearance of the Blake transmitter, as
manufactured by the Consolidated Telephone Company, is shown in
Figure i8.
For short distances the Blake is very efficient, but is by no means
satisfactory for long distances. It was used by the National Telephone
Company until recently to the exclusion of all others.
Class III. — Granular Transmitters.
The granular transmitters are instruments in which carbon in granular
form is used. The great trouble with transmitters of this class is that
they become practically useless when the granules become damp and
clogged. The Hunnings transmitter was the first of this class, but it
is not at all satisfactory as a practical instrument. It consists of two
platinum plates D and B (Figure 19), the space between which is fillled
Figure 19.
with little carbon granules. The front plate D, held in position by means
of the ring RRi, acts as the diaphragm. The cover carrying the mouth-
piece M serves to clamp the diaphragm D in. position. A small fine wire
grid is put across the mouth-piece to arrest moisture. The principle of
action is that when the diaphragm moves inward the granules are brought
into closer contact and thus the resistance between the two plates is reduced,
and when the movement is in the outward direction the converse takes
place. Undoubtedly the Hunnings transmitter is more powerful than any
of the transmitters which have been describe4. Now let us examine its
^'Packing*' Difficulty. 29
defects and see what remedial measures can be adopted. The transmitter
must of necessity be vertical, otherwise the granules would not always be
in contact with the diaphragm, since the effect of working the instrument
is to cause the granules to settle down into the smallest possible compass.
When the instrument has been in use some time the granules at the bottom
of the diapiragm become closely packed, and correspondingly loose at the
top. The part of the diaphragm which moves most under sonorous vibra-
tions is the centre. At the outside all round the diaphragm the movement
is inappreciable, consequently we have the granules all round remaining
in practically the same condition as regards resistance of contact. Further,
at the bottom the granules are very tightly packed, and form a sort of
short-circuit across the cerftre part. The contact resistance of the gran-
ules between the front and back plates at the centre of the diaphragm
is varied to a very large extent, but the granules round the edges, particu-
larly at the bottom, do not vary, and therefore we have our variation very
much reduced. For instance, let us isolate, say a half-inch circle of
granules at the centre of the plate, and assume the resistance ofiered in the
normal position to be, say, lo" . Let us now start an organ pipe, which
causes the resistance to vary from 5" to I5»> ; we have here a variation of
100 per cent., but now if the resistance of this outer set of granules be 5">
we shall merely have the variation between the joint resistance of 5" and
the lower, and 5" and the higher resistance — 1.«., between 2J" and 3J" , or
334 per cent. Thus it will be realised how serious is the "packing " trouble,
as it isxalled. In the first case the transmitter is three times as e6Rcient as in
the second.
As has already been stated, the Runnings transmitter originally con-
sisted of a platinum diaphragm, between which and the backplate of
carbon is placed a layer of granulated carbon. This granulated carbon
consists of oven-made engine coke powdered and sifted free from dust.
Many improvements were introduced by Charles Mosley, of Manchester.
The platinum plate was replaced by a very fine thin board, in the centre
of which was fixed a small carbon disc, forming one electrode, the back of
the instrument, as before, forming the other electrode. The effect of this
improvement was to confine the current to the central granules— that is to
say, the granules in contact with the small carbon disc attached to the
diaphragm. Further, the space was made wedge-shaped, the point of the
wedge being at the bottom— i.e., the diaphragm and the back of the trans-
mitter were inclined together. These improvements resulted in very
greatly enhanced efl5ciency. owing largely to the greater uniformity of
contact amongst the loose granules. The packing trouble was not even
here entirely eliminated. The granules became set and wedged together,
thus little variation in the degree of contact took place. The packing
trouble does not manifest itself until the transmitter has been in use some
little time. The transmitter must always be in the vertical position,
30
t>eckert Transmttier.
otherwise the granules will settle down, and the diaphragm will not always
touch them when at rest — a very necessary condition.
The " Hunnings-cone " invented by Deckert is, perhaps, the most
reliable of the many granular transmitters. This is the instrument
FiGnRE 20
adopted by the Post Office, for whom it is manufactured by the General
Electric Company. The object of all the improvements which have been
Figure 21.
made from time to time upon the original Hunnings transmitter has
been to eliminate the packing difficulty. The Deckert transmitter
shown in section in Figure 20 consists of a carbon diaphragm M
Deckert Transmitter. 31
and a. backplate of carbon K, which is cut into the shape of little
pyramids, as shown in elevation in Figure 21. Further, the space
between them is filled with granular carbon. It will be noticed that
the ridges between one pyramid and the next fall opposite the centre
of the pyramid above and below (Figure 21). This prevents the
granules from running from the top to the bottom along the ridges.
Figure 22.
The cones keep the granules well distributed, and thus uniformity of
contact is secured. In order to prevent the trouble which arises from
the granules becoming packed at the bottom of the cones, the diaphragm
is fitted with a woollen ring F F, so that only the granules in the centre
of the transmitter are active. A plan of the diaphragm is shown in
Figure 22. The central set of pyramids opposite m in Figure 22 are
Figure 23.
fitted with little tufts of silk at their points, and tend to damp out any
vibrations which are not continuously sustained, and, as we have seen, an
32
Solid-Back Transmiiley.
efficient transmitter must on no account be a persistent vibrator. One
terminal of the instrument is formed by the metal ring R R', upon
which the polished carbon diaphragm is mounted, the backplate of
carbon forming the other terminal. The external appearance of the Post
Office form of this transmitter is indicated in Figure 23.
A new Deckert transmitter works admirably and leaves but little to be
desired, but after it has been in continuous use for some time, say three
or four months, its efficiency becomes much impaired, due to bad
connection between the ring R R' and the diaphragm, also to the granules
becoming wet and packed ; but this can easily be remedied by a skilled
mechanic. The Deckert, or, as it is more commonly termed, the
Figure 24.
Hunnings cone transmitter, is certainly one of the best long distance
transmitters in the market. The National Telephone Company are
fitting these transmitters in connection with their new systems, and it
is hoped that this course will result in increased efficiency when speaking
over long distances.
Another form of transmitter, very much used in America, is the solid'
back. It consists of two very small discs of carbon A and B placed in a
small chamber or cell filled with carbon granules. The movement of the
diaphragm D causes the nearer approach of the discs and thus more
intima e contact between the granules. The instrument is shown in
Solid-back Tmnsmitter. 33
section in Figure 24. The diaphragm, which is of ferrotype, carries the
carbon disc A, bolted to it by means of a small screw N, and is clamped
around its periphery. The second electrode and containing cell is carried
by the mica washer, and is adjustable by means of the screw S. The
interior of the cell, which is lined with insulating material, is filled with
carbon granules. It will be noticed that the electrodes are somewhat
smaller than the chamber, and the object of this is to cut the granules out
of circuit at the bottom of the chamber. This gets rid of the packing
difficulty in this case. The current passes through the granules directly
between the electrodes. This, of course, tends to render the transmitter
very efficient. A small spring R coated with indiarubber serves the
purpose of damping out an^ vibrations which tend to persist after the
originating cause has ceased to act. The ferrotype diaphragm forms the
one and the back electrode the other terminal of the instrument.
This instrument is undoubtedly one of the most efficient transmitters
in use to-day, but unfortunately it is somewhat delicate, and will not
stand rough usage. From the above description it will be gathered how
very small the working parts are, and hence how liable to damage. It is,
however, used somewhat extensively by the American Bell Telephone
Company.
The Breastplate transmitter used by the Post Office is of the Ericsson
form, and a full and detailed description will be found in Chapter XIII.,
when the various switch-board accessories are dealt with.
34 The Telephone System of the British Post Office.
CHAPTER VI.
Battery Telephone Stations.
Having examined some of the leading types of transmitters and
receivers, it now becomes necessary to consider what arrangements will
have to be made in order that the instruments may be pat to practical
uses. First of all some means of calling attention to the instrument must
be arranged, as a telephone is not usually required sufficiently often to
justify continuous attendance, and a bell and a key must therefore be
added at each end for calling purposes.
Figure 25.
The trembler bell electrically consists of an electro-magnet, the circuit
of which is completed through its armature. The electro-magnet MM'
(Figure 25), consisting of two coils, is fixed to a brass frame F, which
is in turn secured to the wooden base. To the top of the brass frame a
steel spring S, carrying the armature, is secured by means of screws
Trembler Bell, 35
It will be seen that the spring is prolonged beyond its junction with the
armature and rests upon the platinum pointed screw which is carried by
the binding pillar C. This pillar is insulated from the brass frame of the
instrument by means of an ebonite washer. The bell-hammer is screwed
into the end of the armature.
The path of the current is from the terminal A, through the electro-
magnet, on to the contact screw C, and along the steel spring S, out on to
the brass frame of the instrument, thence to the second terminal B.
Thus the circuit is only complete so long as the spring touches
the contact screw. The connection of a battery between the terminals A
and B will cause the hammer of the bell to vibrate backwards and
forwards. The current, in ps&sing through the coils of the electro-magnet,
causes it to become a magnet, and the armature is attracted towards it,
causing the hammer to strike the bell. Immediately the armature moves
towards the coils the spring S is pulled from the contact C and the circuit
broken. This caiises the soft iron cores to become demagnetised, and
the steel spring then restores the armature to its normal position, when
the contact is again made. This alternate make and break goes on, and
the armature is thus kept in a state of vibration.
It will be equally obvious that we might arrange our bell so that when
the armature was attracted the coils of the electro-magnet would be short-
circuited. For some purposes, where it is undesirable that the circuit
should be broken, this arrangement is preferable ; but ordinarily this is
not important, and the disadvantage of running the battery the whole
time and the risk of bad contact at the short-circuiting points far out-
weigh any advantage. Unless the contact is perfect the arrangement is
most inefficient, far more so than would be the case in the ordinary form.
The spring contact screw is tipped with platinum at the point of con-
tact in order that the sparking, due to self-induction when the circuit is
broken, may not impair the contact. Platinum is selected, as it is inoxidi-
ble and infusible at all but enormously high temperatures. It should be
noted that the distance between the armature when down and the electro-
magnet is never less than i-32nd of an inch. This distance can be adjusted
by means of the pole pieces which are screwed on to the cores. The mag-
netic circuit is completed through the armature and through the piece of
iron which joins the lower extremities of the cores. The mechanism is
usually protected by means of a wooden cover. The resistance of the
bell varies with different makes from about I5«> or 20" to loo" , which
latter figure represents the resistance of the Post Office form. In this
pattern of bell the ends of the coils are brought to four brass connection
plates, so that the coils may, if desired, be connected in parallel, thus re-
ducing the resistance from ioo» (in series) to 25<>' for short or local circuits.
An arrangement must also be introduced by means of which the bell,
ordinarily connected to the line wires, is cut out of circuit and the
D 2
36
Battery Station.
speaking apparatus joined up. The key for calling the distant station
must cut out both the telephone and the bell and put a battery direct on
to the lines. Our key must then have two positions. Firstly, at rest one
line wire must be joined through to the second switch, the purpose of
which is to insert either the bell or the telephone as required. Secondly,
on depression a battery must be joined up between the lines. Lastly, a
third switch must be provided, by means of which the microphone battery
is cut off when the instrument is not in use in order to prevent unr^cessary
waste. The two lines necessary for satisfactory telephone working are
termed the " A " and " B " wires respectively for purposes of distinction.
A double wire circuit is termed a "metallic circuit" in contradistinction
to an earthed or single wire circuit (see Chapter XXII.)
The A wire (Figure 26) is led to the end of a brass spring A, which
normally rests upon the top contact. The depression of the brass spring
establishes a contact with the bottom stop, to which is joined the positive
pole of the battery B', the negative being jcSned to the B wire. The top
contact is joined to the centre of the switch-arm D, which is very similar
A ;
;:iju=^
A LINE
Figure 26.
to an ordinary single current Morse key. The end of the switch-arm is
shaped so that the receiver may be hung upon it. The weight of the
receiver holds the arm down in opposition to the spring S, thus establish-
ing contact between the lever and the front stop. Raising the receiver
from the arm allows the spring to raise the lever, and thus establishes
contact with the back stop.
The instrument is only in use for speaking purposes when the receivers
are lifted off the rests, consequently the back stop is joined to the
secondary of the induction coil and the receivers (which are joined in
series), the other end being connected to the B line. The front stop is
connected to the bell, the other terminal of which is joined to the B wire.
The right hand switch-arm is arranged to join up the speaking battery
through the microphone and primary circuit of the induction coil by
completing the circuit between the back and central contact of the lever.
Complete Telephone Circuit. 37
In the case which we have been considering two switch-arms were
employed, and this necessitates the use of two receivers. The two re-
ceivers are always joined in parallel, but in series with the induction
coil, as shown in Figure 26. When both receivers are raised from
the rests the microphone and receivers are joined up. Speaking on
to the diaphragm causes variations of the current strength in the primary
coil which leads to the production of E.M.Fs. in the secondary. The
currents sent out flow through both receivers on the B line, thence through
the distant station's receivers and secondary of induction coil on to the
A line, thence to the ringing key A, through the top contact to the left
switch-arm, and through its back contact to the secondary coil where the
E.M.F. originated. This is tfee speaking circuit. The receiving speaking
circuit is from the A line through the ringing key, left switch-arm, receivers
and secondary of the induction coil to the B line, as shown by the dotted
lines. When the diaphragm of the transmitter is quiescent no E.M.Fs.
are produced in the secondary, consequently the received speaking is not
interfered with. The speaking currents sent out pass through your own
receivers, and this fact may be demonstrated by tapping the microphone
or by blowing upon it, when a distinct sound will be observed. By de-
pressing the ringing key or push A, the battery B' is connected between
the A and B lines, no matter what position the switch -arms are in. A
current received when the switch-arms are down passes from the A line
through the top contact of the ringing key to D through its front stop,
thence through the bell on to the B. A current of from 20 to 30 milliam-
peres will be required to ring the bell.
B ILIhl E
Figure 27.
A diagram (Figure 27) shows the connections of the speaking circuits
when two stations have raised their receivers from the rests. The line
circuit consists of two sets of receivers, R and R', two secondary coils, and
the A and B lines themselves. The primary circuits are quite separate
and distinct from the line circuit, and it is thought that this diagram
renders the matter sufficiently clear. ,
It is preferable to use two receivers so far as the speaking is concerned,
as the addition of the second receiver does not sensibly reduce the volume
of sound produced in the first, and, further, both ears are closed by the
receivers, and are thus less influenced by external sounds. One receiver
38
Switch- Arms.
is, however, quite sufficient, and it can be held to the ear whilst writing
down a message. When only one receiver is used a switch-arm of a
somewhat less simple character is required. In addition to joining up the
receivers and secondary or the bell, according to its position, it must also
complete the microphone circuit. In the Crossley telephone an ordinary
switch-arm, with the addition of an insulated metal plate and two springs,
is used. The switch-arm (Figure 28) has two positions. When down.
TO B£LL I .
TO •S£CO/^DAKr
4 «£C£-/l/£«
^%J>
Wzz^lS&zzz^
TO riiCfi1of**iOr^e CtRCiJIT
Figure 28.
the lever makes contact with the top stop, thus joining up the bell.
The two springs, which are the extremities of the microphone circuit, rest
upon the ebonite plate, and are disconnected. When the receiver is lifted
off, the steel spring causes the lever to rise, making contact with the bottom
stop, and thus joining up the receiver and secondary coil. The two springs
now rest on the insulated metal plate, and are thus connected, completing
the microphone circuit. It was soon discovered that this insulated plate
was quite unnecessary, as the switch-arm itself might be used for the
purpose of completing the microphone circuit, and in many modern single
receiver stations this is accordingly done.
TO MICROPHONE
TO SBZONDAR*
8t R£CCtVEPS
TO l^ienOPHONB
Figure 29.
In the Post Office form of the instrument the two circuits are kept
distinct, two springs being provided. The switch-arm (Figure 29) consists
Switch-Anns. 39
of a brass lever, the end of which is in the form of a circular rod. The
two right hand springs correspond to the top and bottom contacts of the
Crossley form. A small ivory stud projects from the side of the arm, and
when the receiver is taken off pushes the two microphone springs together,
thus joining up the speaking battery. In the double receiver form two
similar switch levers with two separate pairs of springs are employed.
Nearly every telephone has a different form of switch-arm, but only
three forms have been described. A very superficial examination of actual
apparatus wi'l at once render the particular device adopted and the object
of each part perfectly obvious. The differences are purely mechanical in
the majority of cases. The three types described are three forms used at
various times in the Post Oftce.
Instruments of this class with trembler bells present some disadvantages.
In addition to the two cells required for speaking, a ringing battery has
to be provided, and this battery has to be maintained. In the case of a
private wire from residence to manufactory this is not a very serious
consideration to the owner of the wire, but where perhaps a couple of
thousand subscribers have to be joined to a central exchange the matter
becomes all important. The batteries require periodical examination,
whereas a magneto generator does not. This at once resolves itself into
a commercial question of pounds, shillings, and pence.
40 The Telephone System of the British Post Office.
CHAPTER VII.
The Leclanche Cell.
For battery telephone stations the Leclanche cell is found to be most
suitable on account of the absence of local action and the small amount
of attention required. The cell consists of a square glass containing
vessel in which is placed the zinc rod. The porous pot carries the carbon
plate packed round with manganese dioxide. The exciting fluid is
salammoniac, which has no effect upon the zinc until the circuit is closed,
when the following reactions take place : —
Outer cell : —
+ 2NH4CI + 2H2O
together two molecules of and two molecules of
with Salammoniac Water
+ 2NH4HO + H2
together two molecules of Ammonium and two atoms
with Hydrate of Hydrogen
Zn
One molecule of
Zinc
= ZnCla
form one molecule of
Zinc Chloride
Inner cell : —
Hs +
Two atoms of and
Hydrogen
aMnOa = MuaOg + HjO
two molecules o^ form one molecule oi and one molecule
Manganese Manganese of Water.
Dioxide Sesquioxide
The manganese dioxide is frequently made into a solid block and placed
next the carbon, thus doing away with the porous pot.
A battery used for telephonic purposes must be free from local action, as
the time during which the instrument is in use is very small as compared
with the time it is idle. The battery is usually employed only for a very short
time, and is then given plenty of time in which to recover before it is again
used. The Leclanche cell possesses no local action, but polarises some-
what rapidly if the current is kept on for any length of time, because the
hydrogen arrives at the carbon plate faster than the depolariser — i.e., the
manganese dioxide — can get rid of it. This is particularly the case with
the ordinary porous pot form in which the manganese is packed round the
carbon plate in the porous pot. An illustration of this type of cell as
manufactured by the Consolidated Telephone Company is shown in
Figure 30. The agglomerate form represents a distinct advance in that
the internal resistance of the cell is greatly reduced by the absence of the
porous pot. The manganese dioxide is mixed with carbon and is pressed
into the form of a solid block. The blocks are held in position on the
LeciancM Cell.
41
carbon plate by means of indiarubber rings. Sometimes the zinc is held
in a small porous pot, as shown in Figure 30. This precludes all possibility
of the zinc touching the carbon and thereby causing a short-circuit.
Figure 30.
In the form of cell used by the Post Office the zinc rod is provided
with indiarubber rings at top and bottom, and thus the same object is
accomplished without the resistance introduced by the porous pot.
Figure 31.
There is far less polarisation in this form of cell than in the ordinary
type, and it is therefore more efficient. A special form of agglomerate
cell is used for speaking purposes.
It consists of a thick carbon rod
42
Agglomerate Cells.
grooved to receive the edges of six agglomerate blocks composed chiefly
of carbon and manganese dioxide (Figure 32). The blocks are placed
in position, and coarse cloth of a canvas nature is wrapped round
the arrangement, and the whole is then held in position by means
of two indiarubber rings. A circular zinc plate surrounding the
negative element reduces the internal resistance of the cell by dis-
tributing the current through a very much larger quantity of the liquid
than is the case with a zinc rod. These large zinc plates were found
to be very wasteful , as the current is not evenly distributed throughout
the area of the zinc, the result being that the zinc is worn away very
Figure 32.
unevenly. A hole occurred about an inch and a half from the bottom
and also at the point where the liquid ceases at the top. It was found by
experiment that the size of the zinc might be greatly reduced without
proportionately increasing the resistance of the cell by using the form of
shallow circular zinc shown in Figure 33. This zinc, instead of resting
on the bottom of the cell, is suspended from the top by the lug, which is
bent backwards. The whole of the zinc is immersed below the liquid.
The complete cell is termed the six-block agglomerate Leclanche, and is
contained in a circular earthenware jar. Two cells are used with the
Deckert transmitter and give most excellent results, but, in the Post
Power Necessary for Speaking and Ringing,
43
Office exchanges, it is found that, with a new and carefully adjusted
breast-plate transmitter, one cell gives better results, as the speaking
with two cells is exceedingly heavy and muffled, a man's voice coming
through in the form of an indistinct roar. For ringing purposes a cell
Figure 33.
with a higher Internal resistance answers the purpose equally well. The
two six-block cells are usually supplemented by sufficient cells of the
ordinary form to provide the ringing current. The six-block cell does not
polarise to so great an extent as the ordinary cells.
44 The Telephone System of the British Post Office,
CHAPTER VIII.
Magneto Telephone Stations.
The magneto generator and bell is used by the National Telephone
Company upon all their systems, and in order that the working of both
the trunk and local systems may be apprehended it is necessary to discuss
this apparatus. The object gained is to reduce the battery necessary at
the subscribers' offices. The generator provides alternating currents in
retu'n for the mechanical energy expended in turning the crank handle.
The bell is designed to ring with these currents and will be described first.
It consists of an electro-magnet A B, with a pivoted armature J rendered
^~T ^"^ \^ _^ ^'
Figure 34.
magnetic by induction from a permanent magnet N S (Figure 34). A thin
steel rod carrying a brass ball is screwed into this armature at right
angles to it, and plays between the two bell domes L and M. The diagram
sufficiently illustrates the mechanical construction.
The south pole of the permanent magnet is immediately above the
centre of the armature, thus rendering the centre north and both ends
south. The electro-magnet is joined up in the ordinary way ; thus a
current in one direction flowing through it will cause A to become north
and B south. The armature is repelled at B and attracted at A, thus
holding the hammer upon the right dome. Reversal of the current makes
A south and B north ; the armature is therefore attracted at B and repelled
Magneto Bells. 45
at A, the result of which is that the hammer passes over to the left hand
dome. It will be seen that a succession of currents in opposite directions
will cause the hammer to vibrate between the two domes, thus giving a
ring. Adjustment of the play of the armature is made by moving the
domes bodily, screws for that purpose being provided. The distance
between the armature and the electro-magnet is adjustable within small
limits by means of the nuts on the screwed pillars of the frame work.
Figure 35.
The external appearance of a magneto bell is shown in Figure 35. Its
resistance is now 1,000 ", though formerly many of lower resistances, 100 ■*
to 500 ", were used. The high resistance bells require far less current,
and in a large exchange this is an important matter, as the pulling up
of the generator is very serious. During the busy hours of the day the
generator would almost stop were ioom bells universal, and it is obvious
that the lower the speed of the generator the less is its E.M.F., and thus
many of the bells would not be rung at all. In order that the principle of
the magneto generator may be clearly comprehended, it will be well to refer
back to the theory of the induction coil. It has previously been stated
that whenever a conductor is cut by magnetic lines of force E.M.Fs.
are generated in that conductor. It would, however, be more accurate to
say that whenever the number of lines of force passing through a
system of conductors is altered, an E.M.F. is generated. The magneto
generator takes advantage of this fact. A coil of wire is twisted round
in a strong magnetic field in such a way that the number of lines of
force passing through the coil is continuously changed. The magnetic
field is produced by the strong permanent magnet (Figure 36).
The system of conductors, together with its mechanical support, is
termed the armature. The form of armature now universally adopted is
that used in the dynamo invented by Dr. Werner Siemens in 1858, and
consists of a bar of iron wrought into the shape of the letter " H." The
46
Magneto Generator,
two sides of the H are curved, as shown in Figure 36, in order that they
may the more closely fit the pole pieces. In fact, it consists of an iron
cylinder into which are cut two wide deep slots so as to leave only a
relatively small central bar, around which a large number of turns of wire
are wound. This is the shape of the armature in section. It is usually
about an inch in diameter and from three to four inches long, and is
wound with silk-covered copper wire of about 40 S.W.G. — i.e., 0048 inch
in diameter. When the curved ends of the armature are at the top and
bottom the lines of force nearly all pass from the north pole of the magnet
through the curved ends of the armature to the south pole, scarcely any
Figure 36.
lines of force passing through the central or cross portion of the H — i.e.,
there is the minimum number of lines through the conductors. Imagine
now that the armature is revolved until the curved ends of the armature
are precisely opposite the two pole pieces, as shown in Figure 36. The
lines of force all crowd through the small air-gap into the curved end and
through the central portion out on to the other curved end, thence through
the small air-gap to the south pole. Here, then, nearly all the lines pass
through the conductors. In moving from the first position, where very
few lines passed through the conductors to this one, where a very large
number pass through the conductors, E.M.Fs. have been induced as the
number of lines through the conductors have been gradually increased
Magneto Generator.
47
from none to the maximum. This represents a quarter of a revolution,
and at this point the E.M.F. falls to zero. On moving forward till the
first position is reached there are again no lines through the conductors.
The lines, on threading out of the circuit, produce an E.M.F. in the
opposite direction to that produced when threading in. On moving for-
ward to the second position, lines have again been threaded through, but
in the opposite direction, as the faces of the armature have been reversed ;
this produces an E.M.F. in the same direction as the decaying Unes. At
this point the E.M.F. again falls to zero, and on turning to the first
position is again at its maximum value, but in the opposite direction, fall-
ing to zero as the second position is reached. Hence it will be seen that in
Figure 37.
one revolution of the armature fresh currents in opposite directions are
induced.
A very small movement of the armature from the first position causes
a much larger alteration in the number of lines passing through the arma-
ture than an equal movement from the second position. A consideration
of these facts will show us that the largest E.M.Fs. are produced when
the movement is from the first position, and smallest when from the
second. The number of lines through the armature is continually chang-
ing as it revolves; consequently E.M.Fs. are continuously induced.
48
Short Circuiting Arrangements.
These E.M.Fs. rise to a maximum at the first position, fall to zero at the
second, then rise to a maximum in the opposite direction, and again fall to
zero as the second position is reached ; hencs we see that the current is
reversed in direction twice per revolution. The external appearance
of a combined magneto generator, bell, and switch lever is indicated in
• Figure 37.
There is yet another point which will have to be considered in the
design of our magneto generator, and that is as to how it should be joined
up with the bell. It is not desirable that we should have the bell and the
generator in series, since we should have to waste a goodly proportion of
the energy of our distant generator in overcoming its resistance. A switch
is therefore placed on the armature of the generator, by means of which
the generator is short-circuited when not in use, but immediately the
handle is turned the short-circuit is removed. There are many forms of
switch for this purpose, but the one most generally used is the one illus-
trated beljw in Figure 38. It consists in so arranging the spindle, upon
Figure 38.
which the wheel driving the armature is fixed, that when rotated it moves
over slightly to the right, thus breaking a contact. The spindle has a pin
put through it, and this pin rests in two triangular grooves. It is held to
the narrow end by means of a spiral steel spring. When the handle Is
rotated the pin rides to the top of the groove, where it catches the top
edge, and the armature then begins to rotate. When the pin rides to the
top the spindle inside the tube upon which the wheel runs moves for-
ward, thus breaking contact between this spindle and the spring which
rests upon it at the end.
The ends of the armature coil are connected to the spindle and spring,
which are in turn connected to one side of the magneto bell and the
switch-arm. It will be seen that a current from the distant station passes
from the spring to the spindle, thence on to the magneto bell without
Connections.
49
traversing the resistance of the armature. On rotating the armature the
short-circuit is broken, and alternating currents pass out through the bell
to the distant station, there actuating the distant bell. The bell at the
calling station rings, and this gives an indication that the circuit is com-
>=S^
1 ' ^
'♦
i-S
CwtT
/vo ;
FionRE S3.
circuit. It should be noted that normally the exchange indicator is
deflected to the right.
The acquisition of the trunk lines rendered a new form of local switch
absolutely necessary ; but before going into its details it will perhaps be
advisable to first describe in outline the modifications which have been
made, and their object. First, the drop indicators have been replaced by
the indicator relays. Secondly, provision is made for junction wires to
the trunk switch sections, in order that renters may be put through on to
trunk lines. Thirdly, each renter's line is also brought to a plug switch-
spring placed next to his switch-spring for the purpose of joining up the
operator's set. This is accomplished by pulling out the plug a distance of
New Form of Local Switch. 67
about half an inch. This puts the operator's set across the lines. The
pegs and cords are of the circular pattern and have no apparatus connected
to them. The operator can listen to any conversation by withdrawing the
speaking plug.
A circular peg is illustrated in Figure 56. The metal parts of the pegs
are made in three parts. The part A. is hollow, and a sleeve of ebonite C is
placed inside in order to insulate the rod B. Upon the end of this round
rod is fixed a small circular ball D. Thus the screw marked E is in
connection with the tip, and F with the body of the peg A. The cords
consist of two twisted tinsel strands insulated with cotton, wrapped round
outside with wire ; over this is a. braided cotton cover. The wire gives
mechanical strength to the cord, thus preventing its wearing out as
quickly as it would do were the wire omitted. The conductors are
l^IGURE 56.
soldered to two small copper tags, which are secured under £ and F. The
brass peg is covered by a red or black celluloid sleeve, which is held in
position by a small screw. The tip of one peg is connected to the tip of
the other, and shoulder to shoulder. The pegs and cords have no other
connections, and the shoulders of the pegs rest upon brass brackets.
The cords are held taut by means of weighted pulleys, in order to prevent
tangling. The weights are arranged so as to run up and down guided by
thin steel rods. The brackets are removable from the back of the switch ;
thus in the event of a fault it is not necessary to go to the front of the
switch, thereby impeding the operator.
The renter's lines pass through the indicator relays, thence to the switch
springs. Each switch-spring is also "teed" on to the plug switches,
which when withdrawn also " tee " the line on to the operator's set.
In Figure 57 the connection of two renters are indicated. Line i passes
along to switch-spring No. i with the indicator " teed " across. The
straight line is the A and the curved hne the B line. The two springs are
" teed " on to the two inner springs, and the round ebonite plug holds the
two outer springs apart and prevents their making contact with the two inner
springs. The two outer springs are connected to the operator's apparatus
68
Switching Arrangements.
The withdrawal of the plug allows the two outer springs to make contact
with the inner ones, thus joining the subscriber's line to the operator's set, as
shown at No. 2. The subscriber's switch-spring consists of two springs, the
top spring being slightly longer than the lower one. The top spring is slightly
curved, and when a peg is inserted the curved end rests upon the tip of
the peg. The lower and shorter spring makes contact upon the shoulder
of the peg. Thus the two lines of the subscribers are connected together
by means of a pair of pegs and cords. The tips of the pegs connect the
two "A" lines together, and the two shoulders connect the "B" lines
N? I > IWftM«lJU
Figure 57.
together. The permanent currents behave exactly as they do in the case
of the old switch. In the normal condition the renter's permanent current
from the subscriber deflects the indicator to the right. The removal of
the receivers stops the permanent current, thus causing the indicator relay
needle to hang vertical. This calls the attention of the operator to the
circuit, who withdraws that subscriber's plug, thus connecting the
speaking set, and obtains the number of the subscriber wanted. The
speaking plug of that subscriber is then withdrawn and the operator
obtains his attention, and then completes the connection by means of a
pair of cords and hears the renters speak. The speaking plug is then
pushed back, thus cuttmg the operator out of circuit. When the renters
put back their receivers both renters' needles are again deflected to the
right, thus serving as an instruction to the operator to remove the plugs,
which is accordingly done. It should be noted that the relay tongue is
biassed by means of the spring and held on to the right contact ; thus the
stoppage of the permanent current (which holds it off the left hand
contact point) allows the spring to close the local circuit, thus effecting a
ring.
The operator's connections are shown in Figure 58. The outer springs
from the plug-switch pass through the central springs of the springs
operating Connections.
60
labelled "direct." through the inner springs, through the "reversed"
springs, through the generator springs to the secondary coil and receiver
of the operator's telephone. The receiver is joined to an ordinary peg and
is inserted as shown, and the transmitter is similarly connected. The
right-hand inner spring is connected through the secondary of the
induction coil S to the left spring, through the tip along the shoulder back
AEvCReeO qff^ERATOII
Figure 58.
to the inner left spring. The path of the primary circuit is from the
right hand battery to the right hand long transmitter spring, through the
transmitter back to the other pole. When the transmitter and receiver
are reversed the other batteiy is inserted. The pegs must be inserted
diagonally. The depression of the first peg puts a battery to line in one
direction, and the second key puts on the battery in the reverse direction.
The generator is put on to the circuit by depression of the third key.
70 The Telephone System of the British Post Office.
CHAPTER XI.
Principle of Permanent Current System.
In dealing with trunk, i.e., inter-town wires, we confront quite a new
aspect of the question. The National Telephone Company have estab-
lished exchanges in nearly every town of any size, and subscribers' lines
terminate at these centres. The trunk lines connect together the different
centres or districts, and terminate at the nearest Post Offices. In order
that a subscriber of the Company may be put through upon a trunk line
he must be connected to the Post Office, who will put him through to the
distant Post Office. The distant Post Office in turn puts him through to
the distant exchange of which his correspondent is a subscriber.
The lines between the Post Office and the Company's exchange are
termed "junctions." A junction is provided for every trunk terminated
4t the Post Office, so that should calls for all the trunks be received delay
may not occur through there not being a junction at liberty. It has been
found advisable to set apart a special wire or wires in addition for passing
calls from the National Company to the Post Office and tice oersd, thus
rendering indicators upon the junctions unnecessary. The apparatus
required at the Post Office must be such as to enable the operator to see
whether a trunk is calling, engaged, or disengaged ; also to connect junctions
and trunks together, and to be able to speak on the trrnk and on the service
wires. Further arrangements must be made for a ring-off signal.
>? L/At£
/? c. I I A e.
' Lec/ii.
fxe. r-
«o*^
Y
»»-
Figure 59.
First, it will be well to cor-sider the principle of the automatic signalling
upon trunk lines. A polarised indicator relay (described upon page 64)
and two sets of batteries are employed, which normally (trunk disengaged)
are joined up to the trunk, as shov/n in Figure 59. The larger battery
Automatic Calling and Clearing.- 71
usually consists of six Leclanche cells, and is joined up in series with
the indicator and the trunk line. The battery and indicator are similarly
joined up at the other end of the line ; thus the two batteries oppose each
other, and no effect is produced upon the two indicators. The circuit
is thus in a state of equihbrium. The smaller battery consists of three
cells, and is joined across the right hand coil at either end, and sends a
current through this coil in such a direction as to deflect the needle to the
right — «■ e., in the opposite direction to the main permanent current. The
local permanent current battery has a resistance coil of 350" joined in
series with it. It will be seen that the two main permanent current
batteries oppose one another, and that the local permanent current batteries
produce a deflection in no way interfered witti by the other batteries. In
the normal condition, then, both indicators are deflected to the right.
When a peg is inserted into the trunk switch-spring the apparatus, at
the end where the peg is inserted, is entirely removed, and the lines are
Figure 60,
connected through the ring-off indicator {which will be described later)
connected across the cords. This condition is illustrated in Figure 60.
The main permanent current battery is made sufficiently large to over-
come the local battery and send a current through the indicator at the
undisturbed end in such a direction as to cause a deflection to the left.
The local permanent current flows through one coil only, but the main
permanent current flows through both, thus doubling the magnetic effect
of the relay upon the armature and needle. The 35o"> coil is inserted
in preference to using a smaller battery in order that the current due to
the main battery may not pass through the low resistance of the battery
in so large a proportion as to rob the right hand coil of current. It will
thus be seen that the insertion of a peg at the distant end allows the main
permanent current to overcome the local, thus reversing the deflection of
the needle from the right (disengaged) to the left, which position indicates
a call
yi Connections of Trunk Switch-Spring and Relay.
Next it will be most convenient to consider the form of switch-spring.
And here it may be mentioned that the word "switch-spring" has been
substituted for the word "jack " by the Post Office. In the normal condition
the switch-spring must maintain the connections as shown in Figure Sg.
The insertion of a peg must cut the circuit of the local battery and
remove the indicator and main permanent current battery. It must also
connect the trunk line itself to the peg. The switch-spring consists of
seven insulated brass springs, in two sets of four and three springs,
mounted upon a brass foundation. The lower four springs consist of two
outer springs of different length and two inner springs making contact
with them. The short spring comes into contact with the tip, and the
long spring with the body, as shown in Figure 63. Thus the shoulder is
connected to the B line and the tip to the A line. The central spring of
tlj.e three springs at the top is moved upwards when the peg is inserted,
Figure 61.
thus breaking confact with the lower spring. This cuts the local per-
manent current circuit. The insertion of the peg also causes the two
springs connected to the trunk to break contact with the inner springs,
and this cuts the indicator and main permanent current battery out of
the circuit.
The whole of the connections of the switch section terminate upon
"connecting tags," which merely consist of thin brass stampings fixed
edgewise into grooves in an ebonite block and secured by an ebonite cap.
The ends of these brass stampings project beyond the ebonite, and have
holes of sufficient size to take a No. 16 wire at either side of the -strip.
Each strip, which is 6 in. x i in. over all, holds twenty-four tags in four
groups of six. These strips are screwed upon the inner parts of the side
of the section. The left side (viewed from the front) is devoted exclusively
to batteries, signalling wires, etc., whilst the other side contains the twisted
speaking circuits Each tag is numbered, and thus joining up is an ex-
Switck-Springi.
n
ceedingly simple matter. The manner in which the tags are appropriated
in A, B, and C sections will be found in Appendix C.
The switch-springs used by the Post Office are of two forms, the one
having four and the other seven springs, the former being illustrated in
Figure 62. The body or foundation of the switch-spring is of brass,
Figure 62.
one-iighth of an inch in thickness, the four springs fitting into straight
jicks made in two thick ebonite washers, which are held together by
screws. The shape of the body of the switch-spring is shown in Figure
62, and the circular brass tube fits Into the ebonite foundations of the
switch. When a peg is inserted, it passes through this tube and pushes
the two outer springs open, and when pushed home is connected, as shown
in Figure 63. It will be seen that the ball of the peg rests upon the
Figure 63.
curved end of the shorter spring, and the long spring rests upon the back
part of the brass peg. This raises the two outer springs off the inner
springs, thus leaving the inner springs disconnected. In Figure 62 it will
be seen that the long spring is in contact with the nearer inner spring, and
the short spring is in contact with the other inner spring. These switch-
springs are termed "five-point switch-springs," the fifth point being the
body of the switch-spring, which is occasionally employed and serves to
connect the body of the switch-spring to the long spring. .
74
Switch-Springs,
The other form of switch-spring is the " eight'foint switch-spring," which
consists of seven springs. The three additional springs B, C, D are
placed above the four springs, but at right angles to them (Figure 64).
Figure 64.
The central spring carries an ebonite projection with a steel ball — resting
in the body of the switch-spring. The insertion of a peg raises the ball,
thus lifting the central spring from the lower to the upper spring.
The Telephone System of the British Post Office. 75
CHAPTER XII.
Auxiliary Apparatus.
It has already been stated that two indicators— viz., the self-restoring
indicator and the telephone* exchange galvanometer — are bridged across
the cords in order to register the ring-off signals. The telephone exchange
galvanometer (Figure 63) consists of two coils of wire mounted upon brass
frames placed side by side, in the centre of which hangs a soft iron needle
rendered magnetic by induction from the permanent magnet. The
axle of the indicator passes through the permanent magnet at either end,
.
s
L>^
B
1
Figure 65.
terminating in pivots, one of which is secured in the removable pivot box
fixed to the end of the magnet and the other at the back of the coils. The
axle is partly brass and partly steel. The portion passing from the left
hand pole of the permanent magnet to the small forked soft iron needle is
of steel. The remaining portion between the two poles is of brass. The
coils of the instrument are wound upon two rectangular brass bobbins,
which are removable from the brass base of the instrument. It will thus
be seen that we have here a galvanometer with an induced needle. A
light aluminium needle is attached to the other end of the pivot, serving
y6 Telephone Exchange Gatvanomeiey.
as a pointer. It will be noticed that we have both the pointer and the
needle depending from the axle, and as this weight would render the
instrument unsensitive, small balancing weights are provided. The first
consists of a thin brass rod passing through the axle of the instrument,
with round brass washers on either side. Should the needle not hang
vertically, compensating adjustments can be made by screwing the rod
through the axle and clamping by the brass nuts. The second balancing
system consists of a long thin brass screw, which passes through the axle
in a straight line with the needles. By raising or lowering the head of
this needle, the weight of the aluminium and iron needles is to some
extent balanced. The most sensitive condition of the instrument is when
the arrangement is dead beat ; this is to say, if, when momentarily
deflected, the needle slowly returns to the central position without
swinging from side to side. If the top screw is too high the weight of the
screw will hold the needles over, even when there is no current flowing
through the coils. The coils terminate in four insulated collars, through
which four brass screws are passed, thus serving to secure the instrument
to the ebonite strip and also to connect it to the terminals at the back of
the strip with which the screws are in contact. The resistance of each
coil of the instrument is 500o>, as the coils have a very large number of
tarns in order that the very small current may produce a sensible magnetic
effect. The coils of the indicators are always joined in parallel, thus
offering a resistance of 250M. In order that the black needle may be
clearly visible, a circular green celluloid disc is fixed between it and
the permanent magnet by means of two small brass pillars carrying screws.
In order to render the indications of the telephone exchange galvano-
meters more conspicuous, and at the same time increase the sensitiveness
and reliability, it is in contemplation to provide a black disc in place of
the present green one and to paint the needle white. A strip of blackened
brass of the same width as the needle will normally cover it. Stop-pins,
limiting the deflection of the needle, will be provided ; thus the small
deflection will produce a conspicuous signal, inasmuch as a white needle
will appear upon a black background. Signals, when they do not indicate
a call, should be as unobtrusive as possible, and this object appears to
be attained.
It has been previously stated that a self-restoring indicator is joined in
series with the exchange galvanometer. This indicator (Figure 66) consists
of two separate electro-magnets, each of which is iron-clad. The back
electro-magnet is nearly three times as long as the front one, and before it
is pivoted a heavy soft -iron armature held away from it by the weight of a
long brass lever. This lever is so shaped as to catch the armature placed
before the front electro-magnet. The armature is of soft iron, pivoted at
its lower extremity, and tilted slightly' outwards, so that when the lever
is raised it falls forward. A very light aluminium disc is suspended
Self-restoring Indicator.
77
in front of this armature, and when it (the armature) falls forward the
shutter is thrown forward by it, thus indicating a call. When the long
coil attracts its armature the lever is raised, and the front armature
being released throws the shutter forward. In order to restore the normal
state of a£fairs a current is passed through the short coil, and this attracts
the front armature, bringing it beneath the lever, where it is caught.
The long coil is placed across the lines ; thus any current passing along the
lines leaKs through this coil and drops the shutter. The shutter is re-
stored by completing the circuit of a battery through the short coil by
means of the speaking key when in the speaking position. It will be
obvious that the indicator is not required to be down when speaking on the
circuit to which it is attached.
The object of surrounding the electro-magnets with an iron sleeve is to
concentrate the lines of force, thus preventing them from passing through
contiguous indicators connected to other circuits. Were this precaution
FlGDRE 66.
not taken the lines of force would pass through the contiguous indicators,
thus inducing currents which would reproduce the conversation upon
the first circuit. A very little consideration shews that two con-
tiguous unclad indicators are equivalent to an induction coil. Besides
preventing overhearing, the iron sheathing also makes the magnetic
circuit almost wholly of iron, thus rendering the indicator more sensitive.
It also completely separates the two electro-magnets, so that there
is little or no magnetic leakage from the line to restoring coil. The
path of the lines is along the core into the iron disc at the junction of the
coils along the sleeve through the air into the armature, and back through
the air to the core. The line coil has a resistance of looo" and the restormg
coll of 45o»>. The former, when carefully adjusted, requires 5 milliamperes
to actuate it. and the latter from 20 to 25 milliamperes. The local contacts
consist of a very small light spring connected to the body of the-instru-
78 Self-restoring Indicator.
meat and placed beneath the bottom of the restoring armature. When this
latter is thrown forward the spring is brought into contact with a small
screw passing through an insulated brass projection. The body of the in-
strument and this projection thus form the local contact. The two electro-
magnets are fixed upon either side of an iron strip, which is secured at
either end to the woodwork of the switch section.
These indicators, which are bridged across the cords, do not sensibly
diminish the speaking currents, as their resistance is practically infinite
to currents of such high frequency. The high self-induction is, of course,
responsible for this state of affairs. (Vide Chapter XXIV.)
Tht Telephone System of the British Post Office. 79
It
CHAPTER Xni.
Operating Connections.
is necessary to provide a means of ringing National subscribers.
Tliis is effected by means of a ringing key, which cuts off the blaclc peg
and joins a generator to the red one. The current passes over the junction
to the subscriber, thus ringing his bell. Similarly, a battery is provided
for the purpose of ringing Post Office subscribers, and also in order to
assist the distant main permanent current in the event of its not being
Figure 67.
sufficiently strong to both reverse the deflection and close the local circuit.
The ringing keys (Figure 67) consist of a brass block carrying four
insulated springs and two insulated contact studs. The two inner springs
normally make contact with the studs, and project beyond the brass block.
These springs are curved inwards in such a manner that when the key is
depressed the triangular-shaped piece of ebonite which is attached to the
rod of the key pushes them outwards, where they come into contact with
the outer or generator springs. The long inner springs are connedted to
the tip and shoulder of the red peg respectively. The studs are connected
to the corresponding studs of the black ringing key, whose inner springs
are connected to the tip and shoulder of the black peg. The outer springs
of the black peg are connected to the ringing battery.
8o
Connection of Ringing Keys.
The connections of a pair of cords and ringing keys are shown in
Figure 68. Depression of the black key connects the battery springs to
the peg springs ; thus the negative pole of the battery is connected to tha
tip of the peg which comes into contact with the A line. The positive is
connected to the B line ; thus it will be seen that the ringing battery is
joined in series with the distant main permanent current. Similarly,
1)6,1
■U'
m
ijf '^ ■if sj/
'o *>UT£i> Sfftines ^ ouTKn s/'/tmae
o* m-H^n at./teK o* o-ntet) nee
Figure 68.
depression of the red peg connects the generator to the tip and shoulder.
In the normal condition the tip of the black peg is connected to the tip
of the red peg, and shoulder to shoulder. The connection between the
inner studs is also joined to the ring-off indicators and to the speaking key.
The battery is teed along to all the outer springs of the black keys, and
the generator is similarly connected to the outer springs of all the red
A key is attached to each pair of cords which, when thrown forward,
puts the operator's speaking set across the cords at the point indicated
Compute Operating Connections. 8i
in Figure 68. The general form of the key consists of six springs and
four studs ; in fact, is very similar to two ringing keys placed side by side,
but operated by one lever. The triangular piece of ebonite used to push
apart the springs is twice as long as that used in a ringing key. On the
left there are two springs, which rest upon insulated studs at the front
Figure 6g.
and back of the key respectively. Similarly upon the right, two springs
rest upon contact studs, but in this case two springs are placed outside
these, so that when the ebonite block descends the studs are left discon-
nected, and the in^er spnngs make contact with the outer. This is the state
of affairs when the key is in the normal or upright position (see Figure 69).
G
82
speaking Keys.
These outer springs are disconnected, not at present being in use. The
two inner springs upon the right are Connected to the operator's secondary
receiver, and may therefore be termed the " receiver springs." The
corresponding studs are joined to the connection between the ringing Iceys —
i.e., practically across the cords. Upon the left the front spring and stud
are used to complete the primary circuit of the operator's speaking set.
This prevents the battery being worked when the operator is not actually
speaking. The back spring and contact are used to restore the indicator
attached to the particular cord. When the indicator drops, the operator
goes in circuit by throwing the speaking forward, thus at once bringing in
the speaking apparatus and restoring the indicator to its normal position.
Figure 70.
We are now prepared to consider the whole of the operating connections,
and these are indicated in Figure 70. It will be seen that each pair of
cords has a self-restoring indicator and a galvanometer bridged perma-
nently across it, and that the two ringing keys merely cut off the other
side and join the generator or the battery to the cords as the case
may be. The generator is connected to the outer or generator springs of
all the ringing keys, and the battery is similarly connected to all the battery
springs. The connection between the ringing keys is joined to the back
and front right hand studs of the corresponding speaking key. The front
receiver springs are all connected together to one side of the secondary
and receiver of the operator's speaking set, the back receiver springs
being connected to the other side. The transmitter springs are all con-
Switch Telephone Connector, 8^
nected together and go to one side of the transmitter, battery, and primary
of the induction coil. The transmitter studs are all similarly connected
to the other side of the primary circuit. The left hand back or restoring
springs are all connected together to the negative pole of the lo-cell
Leclanch^ battery used to actuate the front coil of the self-restoring
indicator. The right hand side of all these coils are connected together
to the positive pole of the battery. The restoring studs are each connected
to the left hand side of the corresponding indicator, and it will thus be
seen that throwing forward the speaking key restores the corresponding
indicator. The speakirg key is shown conventionally in Figure 70. The
receiver springs and studs are in the centre, with the two disconnected
springs outside them. Tlie receiver springs are shown with an insulated
projection, which, when in the speaking position, allow the transmitter
and restoring springs to make their respective circuits. When in the
normal position all the contacts are severed.
The only point in regard to the operating connections still remaining to
be considered is the connections of the operator's telephone itself. Pro-
vision has been made so that in the event of the failure of one of the
speaking batteries a second one can immediately be substituted. In fact,
the majority of the apparatus is in duplicate. The transmitter and
receiver are connected to ordinary circular pegs and fit into switch-springs.
There are four switch-springs — two for the transmitter and two for the
receiver. Each set of two switch-springs is joined up with a battery, and
thus changing from the one set to the other introduces another battery,
Formerly two induction coils were also provided, but this has been found to
be unnecessary, and all sections have now only one. Spare transmitters and
receivers are also provided, so that it is possible to change practically the
whole of the speaking apparatus instantly. This prevents the existence of
a fault In the speaking apparatus from stopping the section. The switch
telephone connector, as this set of switch-springs is termed, is one-and-a-
half inches square, and is placed beneath the desk of the section. The
front appearance of the connector is indicated at A in Figure 71. The
right hand side and the bottom is of brass and forms a protection to the
springs, the other sides being closed by the woodwork of the section. The
pegs must always be inserted diagonally — i.e., the receiver in Rj and the
transmitter in Ti. or R2 and Tj. The switch-springs each have four
springs, but only the outer ones are actually used. When the peg is in
the position shown the second battery is in use. The path of the current
is from the positive pole of No. 2 battery (joined to tags 62 and 63) through
the primary of the induction coil to the outer transmitter springs of the
speaking keys, thence, if a key be down, to the inner spring through short
spring of Tj, through the ball of the peg, through the transmitter, back
to the shoulder of the peg to the negative pole of the battery. The
secondary circuit Is from S, through the short receiver spring to the tip
a a
Connections of Switch Telephone Connector,
a--
speaking Battery Circuit.
85
of the peg, through the receiver to the shoulder of the peg, to the long
receiver spring, thence to the front receiver stud oi the 'speaking key,
through the front receiver spring to the tip of the pair of pegs, through
the line connected to the peg to its shoulder, thence back vid the back
receiver spring and stud to S. This circuit, also, is only complete when a
speaking key is down. In brief, the speaking key attached to each pair
of pegs and cords completes the primary circuit and tees the secondary
Figure 72.
Figure 73.
circuit across the pair of cords to which it is attached. It will be seen
that putting the pegs in Ri and Tj brings in the first battery. In the case
of small offices having only one section, the switch-board telephone pre-
viously illustrated is frequently used, but with circular pegs in place of
the four-section peg. In this case the two pegs are put into one cover, as
shown in Figure 72. This prevents the possibility of the pegs being
put into Rj and T,. or R» and Ti. Another form of hand telephone
86
Head-Gear Receiver.
used is that made by the Western Electric Company, described under
the heading of " solid back " transmitter. The receiver is of the
ordinary watch pattern, but the button placed upon the handle when
depressed completes the primary circuit. This enables the operator to
listen to a conversation without introducing the noise from the switch
room upon the wire.
At the larger exchanges the transmitter and receiver are designed for
attachment to the person of the operator in order that both hands may be
left free for operating. The receiver is of the ordinary watch pattern, but
with a steel band attached. The ether end of the band has a soft leather
acorriMQ fm^e^
Y/i/inisnee OJtuze
■rMfiite.ce .
CA»B0I1 BLOCK IHTeKSTICES
viTH FLyinnE. I.
Figure 74.
pad attached, which rests above the operator's right ear. The connection
is effected by means of a two-conductor flexible cord which terminates in
a red-covered peg. The general appearance of the apparatus is shown in
Figure 73, but it should be noticed that the Post Office form of receiver has
only one steel band. It has been found that many of the operators trans-
ferred from the Company prefer the receivers which are fixed to the head by
means of elastic bands, and arrangements have been made to supply these
when asked for, but it must be confessed they are scarcely so neat in appear-
ance as the other form- The transmitter is of the Ericsson pattern, and is
Ericsson Breast-Plate Transmitter. 87
shown in Figure 74. It is of the granular type. The ferrotype diaphragm
forms one terminal, whilst the circular block of carbon forms the other.
Two triangular grooves are cut in this block of carbon, which is covered
with flannel, the edge of which lightly touches the diaphragm. A round
hole is also cut into the centre of the carbon block, and a light spiral
spring in compression, also covered by flannel, is placed in this recess,
thus putting a very slight tension upon the diaphragm. The space
between the carbon and the diaphragm is filled with finely granulated
carbon. A small circular brass disc is attached to the inside of the
diaphragm, and this it is which is in contact with the granules. A
varnished silk diaphragm is placed in front of the diaphragm proper
in order to prevent the access of moisture to the granules. The whole
arrangement is placed in an aluminium case, and may be rotated
upon an axis at right angles to the diaphragm. The case forms one
terminal, and an insulated spring, which rests upon the continuation of the
insulated screw securing the carbon electrode, forms the other. A key
upon the transmitter permits of its disconnection, for the purpose already
mentioned in the case of the switch board telephone. In some types of
this instrument the switch is replaced by a spring and insulated contact
upon the periphery of the case, so arranged that the circuit may be
broken by rotating the transmitter by means of the mouthpiece.
A transmitter of this pattern having an oiled silk diaphragm placed at
the end of the mouthpiece in a small circular box at right angles to the
tube is now being tried, with excellent results. The diaphragm keeps the
inside of the transmitter dry, and to clean it it is only necessary to remove
the mouthpiece.
88 The Telephone System of the British Post Office.
CHAPTER XIV.
The " a " Switch Section.
Having now examined every part of the simpler switch section, it will
be well to consider the section as a whole. At the very small offices, where
there are less than three trunk lines, a section of the A type (Figure 75)
a^t 'juvamtftm
Figure 75.
is used. Upon an A switch section three pairs of pegs, with their
corresponding three speaking keys, six ringing keys, and ring-oflf indicators,
are fitted. The ring-off indicators are placed at the top of the section,
the self-restoring indicators being placed uppermost. Beneath the exchange
Sialvanometers come the polarised indicator relays corresponding to the
Post Office Subscribers Connected to Trunk Switch. 89
trunks, the corresponding switch-springs for which are at the bottom of
the board. Next to the trunk relays comes a space for ten polarised drop
indicators of the form previously described. Below these are ten five-
point switch-springs, which are joined up as shown in' Figure 76. The
lines from the subscribers are connected to the outer springs, whilst the
inner springs are joined to the indicator. The shutter is held up by the
co/v/vecrz/va
Figure 76.
permanent current, which also deflects the needle to the right. When the
permanent current is stopped by the removal of the receivers, the shutter
drops, and calls the attention of the operator.
There are three self-restoring indicators attached to the three circuits
between the switch and the National Company's Exchange. These are
joined up as sho vvn in Figure 77. It will be seen that the two lines from the
^osr Office
t4^r^of*At. Cos
' SjtcM^yvas
IHOtcJt-rott
OFi FtM Tons
FlGUBE 77.
National Company come to the two outer springs of an eight-point switch-
spring, In the normal condition these two lines pass through the back or
■line coil of the self- restoring indicator by means of the inner springs
(Figure 77). A ring from the Nationtvl causes the shutter to fajl forward
90
Junction Arrangements,
thus indicating a call. The insertion of a peg at the Post Office breaks tho
contacts of the inner springs, thus cutting out the line coil of the indicator
and at the same time causes the two top springs to come into contact
thus closing the circuit of a local battery and restoring the indicator to its
normal condition.
The mode of proceedure is as follows : — Subscriber io5i rings up,
dropping the indicator in connection with his line at the National Ex-
change. The National operator inserts a peg into his switch-spring, and
ascertains the nature of the service required by throwing forward the
speaking key. If connection with a local subscriber is desired, the corre-
sponding peg of the pair is inserted in the required subscriber's switch-
spring, and he is rung up in the ordinary way, and the connection is
^osr-
fieo fi£
^VJver-fo^ J"
fie liu^f,
_ !C SubaoriDer,
Figure 78.
complete. If a trunk connection is required, the operator takes par-
ticulars, removes the peg, and places it into the service wire switch-spring,
and rings the Post Office, when the particulars are entered upon a numbered
white ticket. When this subscriber's turn arrives for the use of the trunk
line required, the Post Office operator rings up the Company on the service
wire, and asks that the subscriber may be connected to one of the junction
lines, and when this is done the connections are as indicated in Figure 78.
The Post Office operator then rings up the subscriber with the red peg,
and the connection is completed by the insertion of the black peg in the
switch-spring of the required trunk.
The trunk ringing battery consists of ten No. i Leclanch^ cells— 1.«.,
about I? yoltg. The Company's subscribers have magneto bells so that
Generator Ringing.
91
when their attention is required for trunlts alternating currents are necessary to
ring ihem up. In the smaller exchanges, where there is only one switch sec-
tion, a magneto generator is fixed on the right hand side of the switch section.
ojt* aet.L
MiaH-r aeuL
T^vrt/l iMOIC^T^fSj
Figure 79,
This generator has three magnets, and has no cut-out switch upon the armature.
In order to ring upon a junction, the handle is revolved, and the red button
corresponding with the red peg in the junction switch-spring is depressed,
92 Local Contacts.
For periods of the day when constant attention at the switch is not neces-
sary, a bell must be provided, in order to give an audible call. The complete
external connections of the local contacts upon the self-restoring indicators,
polarised indicators, and trunk indicator relays, together with the numbers of
the connecting tags, are indicated in Fig. 79.
These connections appear somewhat complicated, but a very little con-
sideration will show that, bearing in mind the general principle, they are
remarkably simple. First let us suppose there is only one bell. The cell
is joined in series with the bell and the two points of the local contacts.
When the tongue touches the base of the self-restoring indicator the cir-
cuit is completed and the bell rings. In many cases a second bell has to
be fixed in an upstairs room, in order that when no one is in the room
attention may at once be called to the switch when a distant office calls.
One side of each bell is joined together ; the two other sides of the bells
are connected to either side of a two-way switch. The centre of the two-
way switch and the left hand side of the bells corresponds to the two
terminals of one bell, but the position of the switch determines which bell
shall be in circuit.
The body of all the self-restoring indicators are connected together to
tag 44, and the insulated screws are also teed together to tag 43.
The trunk indicators are fixed upon a copper strip, which is connected to
tag 54. The insulated screws of the five indicators are joined to tags 49,
50, 51, 52, and 53 respectively. If the drop indicators used for sub-
scribers were joined up in precisely this way the bell would ring con-
tinuously while two subscribers were through. In order to avoid this, a
short-circuit piece is inserted in series with the two local contacts of the
drop indicators. This consists of two pieces of brass, so shaped that
a brass plug may be inserted between them. When the plug is inserted,
the drop indicator local contacts are joined up precisely as the self-
restoring indicator locals. Thus when any local contact is made, the
circuit of the battery is made through either the day or the night bell, as
■he case may be.
The Telephone System of the Briiish Post Office. 93
CHAPTER XV.
The "B" Switch Section.
Where there are from three to five trunks, one B switch section is used.
The B section is of exactly the same size and shape as the A section
previously described ; but a strip of five five-point switch-springs is added
to accomodate the additional junctions. In addition to this six pairs of
cords, together with the necessary ringing and speaking keys, are furnished.
In certain cases two visual indicators are added ; but this is a matter
which will be dealt with later.
A e- jft# j«4/«««v -
Jan* r/an# •
y^ftwvj* ^tttrep J»minma ,
^
yii.m^ma>r9 £••/»<•**« a^i^w^rrOM,
Figure 8o.
The junctions all terminate upon five-point switch springs, and no
indicators whatever are fitted upon them. The local contacts are precisely
as described in the case of the A section. The arrangement of the
apparatus upon the switch is shown in Figure 80, and requires no further
94
function Clearing Arrangement.
remark, save that the service wire to the Company terminates in one of
the self-restoring indicators. In the event of anything going wrong a.
junction may be used by leaving a peg in at either end, and thus leaving
the ring-off indicators in circuit.
Upon a busy trunk switch section much time would be wasted in
calling off each of the connections to the National. In order to avoid this
the two inner contacts of the junction switch springs (Figure 8i) are
connected to an earthed battery, which, when the peg is withdrawn, sends
a current along both lines to the National, where it passes through an
indicator, and thus gives a disconnecting signal.
At the National end each junction terminates in a peg, the A and B
lines being connected to the tip and body respectively. Across the lines
yf. T Co
r-m
AO.
^£E:
Figure 8i.
two bridging coils are joined in series. These coils are wound to looo »
each, and have a very high self-inductance. The coils are wound upon
Hn iron core, and are surrounded by iron closed at both ends, thus forming
a magnetic circuit wholly of iron. The self-inductance possessed by this
arrangement is so high as to prevent any appreciable portion of the speaking
current being drained away.
The junction peg rests upon a spring and contact, which, when the
peg is at rest — i.e., when the junction is not in use— breaks the circuit of
the junction-clearing battery. When, however, the peg is in a subscriber's
switch-spring the withdrawal of the peg at the Post OflBce joins up the
junction-clearing battery, and the visual indicator (see page 97) shows
white until the peg is removed and placed in its normal position at the
Company's exchange.
Reed RinaerS.
95
When an operator has a really busy section, containing five trunks, the
ringing of subscribers by means of a hand generator would cause much
delay. In order to avoid the necessity for this a special ringing apparatus,
known as a reed ringer, is frequently employed. This apparatus provides
an alternating current from primary batteries. Two sets are used, joined
up in opposite directions, and are alternately connected to the line.
The apparatus is shown in the diagram (Figure 82). It consists of an
electro-magnet with long flat steel armature, with a contact breaker
similar to that of an electric bell. The steel reed is connected to
one line, and the centre of the two sets of the batteries is connected
to the other side of the ringing keys. Thus it will be seen that the
left hand battery is joined*up when the bar makes connection with the
bottom stud, and the right hand battery wrhen touching the top
•tud. These contacts consist of fine springs placed upon the reed in
/?<..
ysieni ut t M.British Post Office. 103
CHAPTER XVTI.
The
Switch Section.
At all the larger offices switch sections of the C type are fitted.
Where it is thought that<»the number of switch sections is likely to grow
to four, i.e., sixteen or more trunks, it is not desirable to fix B sections, as
to change a B into a C section is somewhat difficult and expensive.
To take a case in point, a certain office at the time of the transfer had
only some twelve or thirteen trunks ; but here C sections were fitted, and
the wisdom of this policy is shewn by the fact that an additional section
has recently been fixed. In another case the number of trunks was about
the same, but their growth was so improbable that B sections were fitted.
Thus it will be seen that fixed and immutable rules are not wise, and that
each case should be taken upon its merits.
T»«» TO RKMAtttina fow aPAiNaa
auTTOM a
Figure go.
The chief point of difference between a B section and a C section lies in
the transfer circuit provision ; indicators and switch-springs, for ten transfer
circuits being fitted instead of two, as in the B sections. These are fitted in
four strips. The top strip consists of five visuals, underneath which is
placed the strip of five switch-springs corresponding to them ; then comes
the second five visuals and switch-springs.
Two transfer circuits are provided from Section i to Section 3 and two
from I to 4. Section 2 can reach all save Section 4, between which two
circuits are provided. In the case of five sections two circuits between the
104
Transfer Circuits.
following sections must be provided : i and 3, 1 and 4, i and 5, 2 and 4,
2 and 5, 3 and 5. For six sections : i and 3, i and 4, i and 5, i and 6,
2 and 4, 2 and 5, 2 and 6, 3 and 5, 3 and 6, 4 and 6.
Where the number of sections exceeds this a special transfer switch is
usually employed ; but this will be described in Chapter XVIII.
The connections of the visuals and switch-springs to the tags is in-
dicated in Figure 90. It will be seen that they are precisely as shown in
the case of the B switch section, and that when joining up a transfer
<^=^
8'^o%pt Srriia^ Sprtpf^
S y/atai fpJicator-S
S draipf Sr„'fc^S/>rxpg»
S S-fotpf -
Figure 91.
circuit between sections i and 3, tag 21 goes to tag 21, and tag 23 to 23 of
the other section. Similarly the ends of tags 47 and 48 are connected.
The reason of this is obvious. Tags 47 and 48 are respectively connected
to the positive and negative poles of the battery. Tag 47 is connected to
the top springs of all the switch-springs teed together, and tag 48 is con-
nected to the teed bottom springs. The visual signalling wires are con-
nected to the respective circuits to which they are intended to work. The
corresponding loops are 83 and 84 to loi and 102.
Junctions to Post OJtce Local Switch.
105
Where there are more than three sections the tags 47 and 48 should be
taken to the test board and there joined to the battery by means of U
links. If this is not done very great difficulty will be experienced in
tracing faults, owin|[ to the multiplicity of teed connections. It should
be noted that only one battery must be used for the whole exchange. If
a single set of cells proves inadequate, other single sets joined in parallel
must be added.
The disposition of the apparatus is indicated in Figure 91. It will be
seen that the space for Post Office subscribers is reduced from 10 to 5, as
operators cannot attend to Post Office renters in addition to trunks ; and,
indeed, at all the larger offices a separate switch is provided for them.
The transfer circuits have'^lready received suitable notice.
In the case of C sections, record tables are always provided, and hence
we notice the disappearance of the three service self-restoring indicators.
Lon^L St^irc-
?ir,UFE 92
A strip of five-point switch-springs is provided for junctions to the Post
Office local switch. Where the renters are all accommodated upon slack
sections, these junctions are, of course, not required ; the switch-springs
are, however, always fitted.
Two junctions are provided from each section to the local switch, where
they terminate in switch-springs with two springs only. These latter are
fitted in strips containing twenty switch-springs. No. i switch section's
junctions are numbert d i and 2 ; No. 3's, 5 and 6 ; No. 20's, 39 and 40,
and so on. A call wire is provided to the local switch and this passes
round all the sections, being connected to each call wire key. The local
operator listens continuously, and receives all instructions upon it. Thus,
if Section 3 wants subscriber No. 20, she depresses the Post Office call
wire key and says " 20 on junction 5," or, if junction 5 is in use, on 6.
The call wire arrangements are indicated in Figure 92.
io6 Local Contacts and Vibrating Sounder.
From what has already been said, the arrangement of the down call
wires will be obvious. The Company's operators are allotted twenty-live
junctions each, i.e., the junctions of five switch sections, and the down call
wire is therefore passed to each of the five sections where call wire keys
are fitted as already explained.
The local contacts are quite the same as in the case of the A and B
sections, and do not require any especial mention. The vibrating sounder
is substituted for bells, as the sound produced is sucti as not to cause
interruption of service. It has also been found that one sounder for four or
five sections is quite adequate, but provision is made so that one, two, three,
four, or five sections may be put on to it. The vibrating sounder consists of
an electro-magnet with a trembler bell make and break contact. The
armature is attached to a stiff steel spring, and thus the sounder merely
makes a buzz. In order to prevent sparking at the contact, and for other
reasons, a condenser is joined across it. This condenser is placed in the
base of the sounder.
The Telephone Syslein, oj ihe British Post Ojice. lo^
CHAPTER XVIIt.
The Transfer Board.
It will be quite obvious that where a large number of trunk switch
sections have to be provided with transfer circuits, direct circuits
from each section to everj^other section within the exchange would be
utterly unfeasible. Where it is possible to adopt this system it is
obviously desirable to do so, as it saves switching operations, which,
however smartly performed, occupy time. It is somewhat difficult to say
how many sections may be dealt with in this way, as conditions vary so
much. For example, an office may have forty trunks, twelve of which go
to one town. Again, by carefully considering the arrangement of the
trunks upon the sections, the transfer work may be reduced to a minimum.
Generally we may say that where the number of sections exceeds six or
eight a transfer board will be essential.
The transfer board is the connecting link between the various sections.
There are circuits from each section in the room to the transfer board, and
means for connecting them together in any prescribed manner are
provided. In order to render the working as rapid as possible the circuits
between the sections and the transfer board are divided into two groups,
known as A and B circuits. The A circuits are those upon which the
section operator passes forward a call to the transfer board operator.
The B circuits are those upon which the latter again passes forward the
call to the required switch section. The A circuits then are used outwards
from and the B circuits inwards to the sections. To each section there
are allotted three A circuits and two B circuits — i.e., the section operator
can pass forward three calls and receive two simultaneously if necessary
The A circuits are those upon which calls come inwards to the transfer
board, and these are placed upon the slanting desk (Figure 93). The
visuals in connection with these circuits are at the top of the slanting desk,
and immediately below them come the combination keys, the object of which
is to enable the transfer operator to speak and signal upon any circuit.
The loop of each circuit terminates in a peg, and upon this circuit there
is absolutely no apparatus. These pegs are to be seen above the desk.
Each A circuit is provided with a peg, a visual indicator, and a com-
bination key. It will be noticed that the board shown is only half fitted ;
but, it is, of course, a very simple matter to add more apparatus when
required.
loS
Trunk Line Transfer Board.
Figure 93.
Combination Key.
109
The insertion of a peg at the section drops the visuals at the section
and transfer board. The operator depresses the combination key, thus
connecting up her telephone. Having ascertained the nature of the
service, the corresponding peg is taken up and inserted into a B
circuit of the required section. The combination key is then raised, thus
putting the visuals at normal. The B circuits are accommodated upon
the upright part of the transfer board.
The combination key consists of thirteen springs arranged in two sets.
These two sets are placed opposite each other, with a round piece of
ebonite, controlled by a lever, playing between them. In the normal
condition the springs are as shown in Figure 96. The upper long springs
are connected to the A circuit, and the outer ones to the operator's
secondary ; thus when the ebonite moves upwards and pushes apart the
long springs, the operator's telephone is connected to the circuit. At
the same time the central pair of springs make contact, thus com-
pleting the operator's primary circuit. This is shown in the upper part
of Figure 94. The inner spring on the left is not used at all. The
bottom springs are arranged exactly like a ringing key, and when the
ebonite descends, forcing the long springs apart, they leave the inner
springs and make contact with the outer ones. Of these six springs only
the right hand three are utilised. Tbe effect of depressing the key is
no
Principle of "A" Circuits
indicated in the lower part of Figure 94. The mechanical design of the
key is illustrat :d in Figure 95.
Figure 95.
The connections of an A circuit are shown in Figure 96. It will be
seen that the speaking circuit passes from the line switch-springs to the
peg at the transfer board, whence if is teed off to the long springs of the
combination key. This provides a' method of connecting the operator's
telephone to the circuit. Only when the key is depressed is the operator's
telephone in connection with the circuit. These are the whole of the
speaking connections.' -
-mA
,-j]
Switch Sktic
. TuAIiSFSR BoANtf
Figure 96.
It has previously been explained how visual signalling from section to
section is accomplished (page 104). It is for this purpose that the three
lower right hand springs are used. The long spring corresponds to the
central spring of the eight-point switch-spring, the inner spring to the
Icwer, and the outer to the top. This long spring passes through the
Method of Working.
Ill
visual indicator, along the signalling wire, through the section visual,
to the middle spring. The top and outer springs and the lower and
inner springs are connected together to the positive and negative poles of
the ^ignalling battery respectively.
When a peg is inserted at the section, the central spring and the upper
spring make contact, thus throwing both visuals. The transfer operator,
seeing the signal thrown, pulls forward the lever of the corresponding
f of ot^r circuit
dis
Figure 97.
combination key, thus putting her speaking apparatus across the circuit.
The section operator states the town required, and the transfer operator then
puts the peg in connection with the circuit into a B circuit in connection
with the section upon which the required trunk line is to be found. Having
done this the combination key is pushed upwards, thus connecting the
lower long and outside springs. This causes both visuals to go black
again. Having finished, the section operator removes the peg, and both
visuals are again thrown. The transfer operator removes the connection
112 "B" Circuits.
ind returns the key to the central or normal position. Everything is
again normal.
In order to pass forward the calls received at the transfer board, B
circuits are provided as already explained. The B circuits are each
equipped with an eight-point switch-spring and a visual indicator. These
circuits are joined up in precisely the same way as the pair of circuits
shown in skeleton in Figure 84. There is no apparatus upon the loop, and
the whole of the signalling is done upon the separate wire. These B
circuits with their corresponding visuals appear above the A circuits upon
the vertical portion of the board. The lowest row consists of two strips
of five eight-point switch springs, which correspond to the second visual
of the first ten sections. Next come the B circuits of Sections XI. to XX.,
and above this, XXI. to XXX. By putting two strips of ten visual and
switch springs (as in Figure 93) in place of strips of five, it will be seen
that we may go on to sixty sections, and this course was pursued at
Manchester.
The actual connections are shown in Figure 97. The principle of the
connections needs no comment, since it is precisely the same as that indi-
cated upon page 98. The insertion of a peg automatically calls the wanted
section, and the transfer operator sees when the call is answered by the
visual going black.
As in the case of trunk switch sections, the cross-connection strips upon
the left accommodate the signalling and those upon the right the speaking
circuits. By left and right are meant left and right looking at the section
from the front.
Every transfer board is fitted with fifteen cross-connection strips, each
containing four groups of six tags, placed upon either side, thus pro-
viding accommodation for sixty sections. As the boards were originally
designed for a different method of signalling, half the tags are not now
required, and are left disconnected. In future boards this will, of course,
not happen. Each set of six tags accommodates three A or two B circuits,
and are marked with the number of the section to which they correspond.
An accurate A circuit diagram is given in Figure g8.
The first two rows of B circuits upon the transfer board are those
corresponding to Sections I. to X., and the tags accommodating these
circuits appear opposite a label marked " Row I." From XI. to XX. is
marked " How II.," and so on. Each set of tags is marked by the side in
white, with the number of the section to which it corresponds.
The ten A circuits appear beneath a label marked " Vis. ccts. A," and
occupy two-and-a-half cross-connection strips. The remaining twelve tags
are in use for the one negative and the ten positive leads coming from the
ten A circuits. These ten tags are, of course, all teed together.
The speaking batteries are to be found upon the last set of tags upon
the central bottom cross-connection strip.
Actual Connections.
113
The tags which are left disconnected will readily be seen from a very
brief consideration of Figure 96. The speaking circuits are similarly
arranged and marked. The call wire occupies the same relative position
as the first two A circuit positive lead tags. Thus it will be seen how easy
it is to trace the tags corresponding to any particular circuit.
A single battery must be used to work the whole exchange, but the bat-
tery leads to each section should pass through U links upon the test board,
so that a short-circuit upon one of the switch-springs may be readily
localised. Where secondary cells are in use, the leads pass through fuses,
■^ StfW^aiy d^Tfr
Figure 98.
which are blown in the event of a fault. This fuse indicates upon what
section or transfer board the fault lies. In the case of transfer boards the
A and B circuits are frequently divided into groups, so that the position of
the fault may the more readily be ascertained. The top spring is con-
nected to the positive and the bottom spring to the negative pole of the
battery, so that if the top spring touches the central spring in the normal
condition the battery is short-circuited. Upon a switch section this fault
blows the section's fuse. The fuse is then replaced by a glow lamp
and a peg inserted into each switch-spring in turn until the fault disappears,
t
114 Faults and their Localisation.
due to the contact being broken with the bottom spring. A short-circuit,
which occurs immediately the peg is inserted, due to a faulty bottom spring,
may similarly be p.'.oved. As a rule, this is instantly discovered by the
operator.
Now, at the transfer board a precisely similar proceedure may be
followed. It should be remarked that owing to the formation of a derived
circuit all the circuits which are through — i.e., indicator black— have their
indicators thrown. The faulty circuit does not drop its indicator. The A
circuits may be dealt with by manipulating the combination keys.
The transfer boards now in use are, as a rule, of far larger dimensions
than is necessary, and it is probable that in future designs the size will be
considerably reduced. Space for two operators was provided, but only
one is required. The combination key might be replaced by a speaking
key of the plug pattern and the signalling accomplished by means of a
socket contact, exactly as described in the case of the R.C.J, circuits
(page 120). Of course, the present signalling wire would replace the
bridging coils.
It may be mentioned that a key for connecting two transfer boards
together is added. This key merely connects the two speaking circuits
together so that an operator may answer calls upon the second board
without removing her telephone.
The Telephone Sjslem of the British Post Office. 115
CHAPTER XIX.
Record Table Switch Sections.
When the trunk lines were in the hands of the Company, subscribers
who desired trunk connections were at the larger exchanges put through
to a special table where tickets were made out. At the smaller Post Office
trunk exchanges the Company's operators repeat demands to the Post Office
record table operators, but as the Company's operators also have to attend
to local requirements, there is some chance of error. Accordingly, at the
larger offices, the subscribers who ask for a trunk are put straight through
to the record table, where they repeat their demand. Having taken their
demand, the Post Office operator gives instructions for their disconnection
by pulling forward a plug, which lights a little lamp placed by the side of
the ticket wire which has been utilised. This is the Company's signal to
disconnect.
Each record table switch section accomodates ten or twenty ticket
wires (officially designated "record table circuits"), according to
requirements. One such circuit is shown in Figure 99. The speaking
lK)=t^
Figure gg.
circuit or loop passes from the short spring of a five-point switch-
spring to the short spring of the Company's switch-spring, and from long
spring to long spring. It will thus be seen that any subscriber connected
at the Company's end will be through to the Post Office. The Post Office
record table operator is provided with a pair of pegs. One of these pegs
is placed in a switch-spring in connection with the ordinary record table
ii6
Connections.
tablet, to which the operator joins her telephone. Thus it will be seen
that the record table operator is able to speak to any subscriber connected
to this circuit by means of the other peg of the pair mentioned. It
will at once be apparent that if a subscriber who calls up has merely
to say "Trunk," and is put through to the record table, where he is
immediately attended to, the system will involve but little delay.
The fact of the insertion of the peg, connecting the subscriber at the
Company's end, drops the visual indicator in connection with the circtiit.
This is accomplished by using the B line of the loop as signalling wire
also. Normally, a battery of about sixteen volts is connected through a
visual indicator to the B line at the Post Office. The insertion of a peg
at the Company's end earths the B line through a relay, which is, however,
biassed against this current. The visual at the Post Ofl&ce shows white,
and the record operator pegs on to the loop. A white ticket is then made
m mm @ in [n h hi m ei
c=3 CI3 c=s cn c=3-c=3 era c:3 i=3 c=i
OI"»'
M
rf '
oooooooooo
HHIlllllIlElllSlIllEl
/OSloilcl,S/m:pO O O O O O
o o o o
oooooooooo
T£f4 CtHCUtTS
Figure ioo.
out, and the operator withdraws her speaking peg, at the same time
pulling forward a plug-key, which increases the voltage of the battery to
about thirty, the current due to which overcomes the bias upon the
Company's relay, thus lighting the small lamp attached to its local circuit.
The visual at the Post OflBce also shows white. The lighting of this lamp
is the Company's clearing or disconnecting instruction. The subscriber's
peg is withdrawn, thus removing the earth from the B line, the lamp goes
out, and the visual at the Post Office goes black, since there is no path for
the battery. The Post Office operator sees this, and pushes the plug-key
back, thus putting on the smaller battery, and everything is in status
quo ants.
The switch-spring at the Post Office end consists of four springs, as
shown. The plug-key consists of a similar switch-spring, but with the
outer springs of the same length. A brass plug is fixed to the front of the
Method of Working. 117
section so, that when pushed home, as shown in the figure, the outer
springs are joined together through it. When pulled outwards the outer
springs make contact with the inner. At the company's end a. switch-
spring with five springs is employed. The insertion of a pe^ there causes
the top sp ing to touch the contact, which is joined to earth ; thus a
current fljws from the smaller battery to the bottom plug-key spring,
through the brass plug to the outer spring, through the visual to top inner
spring of the loop switch-spring, and thence along the B line to the Com-
pany. It there passes through the retard coil and relay through the top
spring to earth. This causes the Post Office visual to show white, but the
Company's relay is not affected. The insertion of a peg at the Post Office
end cuts out all the signa^ng apparatus by disconnecting the top inner
spring and visual. The visual now shews black. Having taken the demand
the plug-key is pulled forward and the speaking peg withdrawn ; thus a
current from the whole battery leaves via the top inner and top outer
springs of the plug-key through the visual and line to the Company's
relay, which is to earth. The relay is now actuated and the lamp lights.
The Company disconnect and the Post Office indicator goes black. The
plug-key is pushed back, and everything resumes its normal condition.
The object of the retard coil is to minimise overhearing between the
several circuits when at normal.
It will be seen that the battery connections are teed along, and this is
done by means of bare copper wire in the actual apparatus.
The disposition of the apparatus is shown in Figure 100. ■
To receive calls at night a relay is placed in the positive lead of the
calling battery, and the local contacts are connected to a vibrating
sounder and battery. In the daytime this relay is short-circuited by
means of a two-way switch. The arrangement is similar to that described
in Chapter XX.
1 3.8 The Telephone Systttn of the British Post Office.
CHAPTER XX.
Direct Junction Circuits.
In the suburbs and outlying districts, instead of bringing the sub-
scribers' lines into the central exchange, the Company sometimes establish
sub-exchanges. If a subscriber at one of these exchanges desires to be
connected to a trunk line, he would ordinarily have to go through the
central exchange to the Post Office. Where there is a considerable
amount of trunk traffic to any particular sub-exchange, direct junctions
from the sub-exchange to the Post Office are provided exclusively for this
traffic.
These junctions are dealt with in three ways at the Post Office. Where
there is no record table, or where there are only two or three sections, the
circuits are placed upon the least busy section, or upon the central section
in the latter case. Where, however, there are several sections, a special
switch, in size comparable with a record table switch section, is used.
This is placed upon the record table, and is attended to by one of the
record table operators. Where the number of junctions to be dealt with
exceeds ten, a junction transfer section has to be employed.
All these devices are worked upon the same principle, and having
described the simplest form, it will be but a step to the more complicated
arrangements.
Since only two wires are provided for each circuit, the signalling has to
be done over the circuit itself. It is not practicable to do this by sending
currents round the loop, and therefore the earth has to be requisitioned
as a return wire. In Chapter XXII. it is explained to what trouble an earth
upon one side of a loop gives rise. In order to avoid this difficulty
bridging coils are used. These coils are merely electro-magnets with their
magnetic circuit permanently closed. A coil of wire is wound around an
iron core. An outer iron sheath is next passed over the whole, and both
ends are closed by discs of iron touching the core. Two soldering tags
project through holes in one end of this disc. The resistance of these
coils is usually either 600" or looo". If now we join two of these coils
in series and tee them across a telephone loop, we may put an earth upon
the central point of the two coils, as we shall thereby treat both lines alike,
and thus shall not disturb the balance of our circuit. The object of using
these magnetically short-circuited electro-magnets in preference to resist-
ance coils Ijes in the fact that the former possess a large amovint oj
R.C.J. Circuits Terminated upon SwitchSections. 119
self-induction, and thus their resistance to a speaking current is enormously
enhanced, and the speaking volume of sound, therefore, is not sensibly
diminished by their application.
If now bridging coils are placed at either end of our circuit, we may
perform whatever signalling we choose over the centre of the loop without
interfering with the speaking round the loop. It is, of course, assumed
that the circuit is not faulty.
One R.C.J. (Record Call Junction) circuit terminated upon a switch section
is illustrated in Figure loi. It will be seen that the loop passes to the long
and short springs of an eight- point switch-spring, the three top springs of
which are used for signalling, as in the case of the ordinary transfer circuits.
Across the loop is placed, the bridging coils, the centre of which passes
through a visual indicator to the middle spring. At the Company's .end
the circuit terminates upon a five-point switch-spring, the inner springs of
Ro
NT.C?
Figure 101.
which are connected together to a visual indicator, having an earthed
battery joined to it. In the normal state of affairs, the negative pole of
the battery A at the Post OiBce passes through the relay, bottom and
central springs through the visual and along both lines in parallel. The
battery B is exactly equal to the A, and as both have their negative poles
to line, no current whatever flows through the indicators. At the
Company's end a special pair of cords having bridging coils with an
earthed visual indicator across them is used for connecting subscribers to
the R.C.J, circuit. Immediately a peg is inserted at the Company's end
the inner springs no longer make contact with the line springs, and in
place of this the lines are connected to the ring-off visual across the cords,
and thus the circuit of the battery A is completed. This drops the Post
Office visual also, and the operator's attention is thereby called. A peg
is inserted at the Post Office in order to answer the call. This disconnects
1 2 - Method of Working
ilm battery A and substitutes an earth in its place ; thus both indicators
go blaclc, showing the Company that the subscriber is receiving attention.
The withdrawal of the peg at the Post Oifice again drops both visuals.
The Company clear [the line by withdrawing the peg, and the opposing
battery B is again inserted. The signals are now normal.
The insertion of a peg at the Post Office removes the battery A, and the
battery B drops both the calling and the Post Office visual. When the
Company insert a peg to answer, the battery B and the calling indicators
are cut out, and the Post Office visual goes black, thus showing that the
Company have attended. The withdrawal of the Post Office peg gives the
clearing signal.
It will thus be seen that the whole arrangement is eotirely automatic,
the insertion or removal of the pegs giving all the requisite signals. It
if
a i
&/ ,o=(E3=n
*
UootfVf^rf^
RlH fi-OPrlHO g Vta>^ \m^
t
Figure 102.
may be well, however, to trace the ordinary procedure. A subscriber calls
up and asks for a trunk connection. He is then connected to the R.C.T.
circuit by means of the special pair of cords. This act calls the Post
Office, who answers him and takes particulars of his request. The peg is
withdrawn, and the Company receive a clearing signal upon the ring-off
indicator, which is instantly obeyed. When this subscriber's turn comes
for the use of the trunk the Post Office operator inserts a peg in the R.C.J.
circuit, thus caUing the Company. This subscriber is asked for, and is
then connected to the R.C.J, circuit. He is rung up by the Post Office,
and after having been put through, is cleared by the withdrawal of the
peg at the Post Office.
The use of the relay at the Post Office end is the only point which
has not been dealt with. Its object is to provide an audible signal at
tiroes of the day or night when continuous attention is not given. It wjlj
R.C.J. Switch Section.
121
be seen that whenever the Post Office visual is actuated a current flows
through this relay, closing the local circuit, which is joined to the vibrating
sounder of the section upon which the circuit is placed.
Upon a B section five circuits may be placed by the addition of a strip
of five visuals and one of five eight-point switch-springs. Ten circuits
may similarly be accommodated by using strips of ten indicators and switch-
springs. It will be obvious that this can only be done where the section
is not a very busy one. If these circuits are placed upon the central sec-
tion of three sections, the central operator will make out all the tickets. The
•
oaoobooooooooooooooo
ecoooooooooooooooooo
OOOOOOOOOOOODOOOOOOO
j&iu
..
i I
VISUAL IND^^C ■<:'-■'' CeT>)
. Sj^^JV/TRArOFCBCc-ia;
Sw.Sfrs. apT^ ,
Sw , SPttS .8 CT "^ .
Vlft. iNDhS *^
6w. SPRtt.8 PT. %p
Sw St>R.^PT COPCRATOtn SB«^
Peas Cm. With Vi'^img\M>ar*>i
Figure 103.
outer operators will easily be able to reach the circuits when they want
sub-exchange subscribers.
Where there are a large number of sections such a system would be
quite impossible, owing to the delay in getting through the transfer boards.
Let us suppose that there are three of these circuits to A, B and C sub-
exchanges to be dealt with at an office having twenty sections. In order
to avoid the necessity of the section operator speaking to the R.C.J, oper-
ator, one of the visual indicators upon each section is marked A, another
B, and a third C. At the R.C.J, switch the ten circuits marked A at the
122 Junction Transfer Section.
sections are placed beneath a label marked A. These circuits are equipped
with visuals and eight-point switch-springs, and are joined up and worked
exactly as explained in the case of ordinary transfer circuits. The ten
circuits marked B are collected under a heading B, and the circuits are
similarly dealt with. Suppose now an operator wants B sub-exchange.
A peg is inserted in the transfer switch-spring marked B, thus dropping
the indicator in connection with this circuit at the R.C.J, switch. As
this indicator appears below the heading B, the operator at once knows
that B is wanted here. If the circuit is disengaged the required connection
is made by the insertion of a peg, and both indicators go black. When
the section operator has done with B, she removes her peg, and thus the
visuals both show white uiltil the R.C.J, operator disconnects (which this
signal instructs her to do) .
It now only remains to show what modifications it is necessary to make
in the apparatus for this purpose. Firstly, then, there are these circuits to
the sections. These are accommodated upon strips of twenty visuals and
twenty . switch-springs, there being space for three such sets of strips.
Above these come the five visuals in connection with the circuits terminated
upon the sections. At the base (Figure 103) are five pegs in place of the
switch-springs used when the circuits were terminated upon the sections
These pegs are furnished with a socket-contact, which takes the place of
the three top springs in the switch-springs. The peg (Figure 102) rests
upon a lever, which, normally, it holds on to the lower or battery contact.
Raising the peg allows the lever to rise to the earth stop, thus calling the
Company automatically. The remainder of the signalling is precisely
as explained before, the replacement of the peg giving the clearing signal
upon the ring-off indicator.
Upon seeing a signal under C drop, the operator takes up the C peg and
puts it into the switch-spring of the calling circuit. The act of raising
the peg calls the Company, and the operator at the section sees she is con-
nected by her visual going black. A peg connected to the operator's
speaking set is provided for answering calls from the sub-exchanges, and
also for informing the operators that the lines are engaged.
The sections are placed upon the record tables, and accommodate five
circuits. Occasionally, however, ten circuits are put upon one switch by
the addition of five more pegs and visuals.
Where the number of sub-exchanges is large a section of a larger type
is used. This stands separately, and accommodates thirty circuits, but in
principle it is precisely the same as the record table transfer section just
described.
In London, where there are several exchanges, this system is adopted,
as obviously it would be out of the question to give to every section five
junctions to each of I^opdon's ten or twelve ^xch^nge;
The Telephone System 0/ the British Post Office. 123
CHAPTER XXI.
Call Office Circuits.
At all Post Office trunk exchanges a silence cabinet is provided, in order
that the public may be able to use the trunk lines. These cabinets
contain a telephone, and a.fk made sound proof, in order to ensure privacy
of conversation. At the counter a switch is provided, so that the clerk
Figure 104.
may speak to the exchange or to the cabinet, also that the cabinet may be
put through to the exchange. Further, an automatic arrangement is made
use of, which rings an elegtric be}l iiqmediateljr the speaker rises.
124 Counter Communication Switch for One Cabinet,
It is the duty of the counter clerk to ascertain from each applic int for
the use of a silence cabinet the details of the service required, ind to
communicate these details by telephone to the trunk exchange operator
in charge of the counter circuit.
A four-cell battery is joined to the cabinet telephone, and this sends a
permanent current to the exchange, where it holds up a shutter, as
previously explained in the case of local subscribers. This battery also
serves for the counter telephone speaking to the exchange.
The arrangement is depicted in Figure 104. A bell, relay, switch, and
telephone are mounted upon a wallboard at the counter. This is known
as the " counter communication switch." The current flows from the
positive pole of the silence cabinet four-cell battery to the fifth terminal
of the counter switch, where it passes through a loo" resistance coil
through the telephone to the seven-terminal switch. In the normal
position the battery goes on through the switch to the A line, thence back
along the B line to the negative pole of the battery. The two last
terminals of the counter communications switch are used for the lines, and
it will be seen that they are also connected to the non-polarised indicator
relay. The indicator relay is thus bridged permanently across the ex-
change lines and will be deflected to the right when the permanent current
is flowing. When the counter clerk depresses the button upon his
telephone or removes the receivers, the permanent current is stopped and
the exchange is thus called. The counter is calle4 by ringing from the
exchange by means of a battery which acts in the same direction as the
permanent current battery, and it is therefore necessary to have resistance
in circuit with it, so that it (the permanent current battery) may not short-
circuit the relay and thus prevent the receipt of the ring. The local
circuit of the relay is from the positive pole of the two-cell battery attached
to the switch to terminal 6 of the telephone, thence through the bell to
the right hand contact of the indicator to the tongue, and thence back to
the negative pole of the battery.
When the clerk speaks from the counter switch the two-cell battery is
again used. It passes to terminal 6 through the primary circuit to
terminal 5, and back through top right and left terminal of the 7-ter-
minal switch back to the negative pole of the battery. The secondary
circuit' is from terminal 4 through the lower two terminals of the switch
to the A line, back along the B line to the top centre terminal of the
switch, back to terminal 5 of the telephone. This provides for all the
operations from the counter.
In order to put the cabinet through to the exchange, the switch is tu:ned
to position z, and the connections are altered so that the cabinet telephone
is joined to the exchange Unes. Terminal 5 of the cabinet telephone is
permanently connected to the B line, and terminal 4 passes through the
two lower terminals upon the right of the switch to the A line, Thus the
Counter Communication Switch for Five Cabinets. I25
secondary circuit is completed, and when the conversation is finished the
battery sends a current along these lines, thus giving the clearing signal.
The primary circuit is completed through the instrument in the usual way
by means of the agglomerate cells in the cabinet. The caller is able to
speak to any place to which he is extended by the switch room. It will
be noticed that when the switch is at position 2 the permanent current
passes through the cabinet relay, and thus ringing currents from the
exchange will ring the bell in the silence cabinet when necessary.
When the caller rises, the contact which is placed upon the cabinet seat
is made, and the circuit of the two-cell battery is completed from the
positive pole of the battery, through the bell, etc., through the seat con-
tact, back through the „7-terminal switch to the negative pole of the
battery. The bell rings until the switch is put back to the normal
position. This is the counter clerk's clearing signal, and by turning the
switch the local circuit is broken. The cabinet battery gives the clearing
signal to the exchange as explained.
Where it is necessary to provide more than one silence cabinet a special
switch is used which will accommodate five cabinets. The lines to
the exchange terminate upon polarised drop indicators upon the switch
sections or upon indicators upon the local switch. At the counter these
lines terminate upon the long and short springs of a five-point switch-
spring. The inner springs of this switch spring are connected to batteries
to supply a permanent current, and in series with the battery, resistance
coils of 200" are inserted in order that ringing currents sent from the ex-
change may not be short-circuited by this small battery. Non-polarised
indicator relays are, of course, connected across the loop, and it is upon
these indicators that calls are received. Normally, then, the indicator is
deflected to the right by the permanent current sent from the counter.
When the line is connected the permanent current battery joined to the
inner springs is cut out and the needle hangs vertical. Ringing currents
from the exchange deflect the needle to the left and at the same time close
the local circuit attached to the armature, thus ringing a bell. A diagram
of these arrangements is given in Figure 105.
The lines from the cabinets to the counter switch terminate upon eight-
point switch-springs (Figure 106). It will thus be seen that any cabinet
line may be joined to the exchange lines by means of a pair of plain cords —
•.«., a pair of pegs connected together by two-conductor cord without any
apparatus connected across: in fact, cords such as are used upon the record
table switch sections previously described. At the cabinet a speaking
battery only is employed, and no permanent currents are sent. In order,
that the counter clerk may speak to the exchange or to any cabinet, a
telephone connected to the outer springs of a five-point switch-spring is
provided, and this may be connected to any line by means of a pair of
pegs. In order to actuate the cabinet bell, a ringing battery is also
125 Counter Communication Switch for Five Cabinets.
added to the counter telephone. In the case of the exchange the insertion
of the peg cuts off the permanent current, and thus automatically register;
a call at the exchange by allowing the shutter to drop.
The only point which now remains to be dealt with is in reference to
the automatic ring-off given when the caller rises from the seat. This is
accomplished by the speaking battery and a non-polarised drop indi-
cator at the counter. The B line of the pair of wires between the
counter and the cabinet is connected to the negative pole of the speaking
battery at the cabinet as terminal 5 serves for one pole of the speaking
battery and also for the B line. The positive pole of this speaking battery
Figure 105.
passes through the contacts of the cabinet seat to the counter, and thence
through the non-polarised indicator in connection with the cabinet on to
the top spring of the eight-point switch-spring. The central spring of
this switch-spring is connected to the B line of the circuit to the cabinet,
and thus when a peg is inserted in this switch-spring the speaking battery
from the cabinet sends a current along the seat wire and through the
indicator, thus attracting its armature and dropping the shutter. This
closes the circuit of the local battery and rings the bell, thus calling the
attention of the counter clerk. In the normal state of affairs (as shown in
Figure 106) the circuit is broken at the switch-spring, but when a caller is
speaking the circuit is broken at the cabinet seat. When, however, he
rises the circuit is completed, and the indicator falls. For five cabinets,
we should require five lines to the exchange, each connected with an indi-
cator relay and a five-point switch-spring, and five lines to the cabinets,
Connections. 127
each connected to an eight-point switch-spring. The seat wires (one from
each cabinet) would each have to be provided with a non-polarised indi-
cator. One bell and one local battery would serve for the whole board,
but separate permanent current battery resistance coils would have to be
provided for each exchange line. One operating telephone and half a
dozen pairs of pegs and cords would also be required.
At the base of the board containing this switch there are two rows of
twenty terminals, which are appropriated as follows : —
Row I. — Five loops (ten wires) to the exchange, and five loops (ten
wires) to the cabinets.
CABINET
*'ci.enont
B a
o o
m^^ L rP
Figure 106.
Row 2. — Ten wires to accommodate the five permanent current
batteries ; three wires to take the speaking and ringing batteries for the
operating telephone ; two wires for the bell local, and five seat wires.
The arrangement of the apparatus upon the switch is now the only
point which requires consideration. At the top are placed the five indicator
relays in connection with the exchange. Below this are placed the five
non-polarised indicators, and then come the strip of five-point switch-
springs, and below this the eight-point switch-springs. Beneath all this
the operating telephone, 200'» resistance coils and terminals are placed.
The bell surmounts the whole arrangement.
At the Stock Exchange a very rapid system of working is in vogue. It
is of first importance to members of the Stock Exchange to be able to get
through upon trunk lines without having to return to their offices. Now
the amount of traffic justifies the use of an operator at the Stock Exchange,
and a most rapid system has been designed.
The member gives to the operator his number (t.e., his number upon
128 Stock Exchange Circuits,
the Company's local system) together with the town and number of his
correspondent. This is repeated to the record operator at the Post OiEce
upon an up call wire, and a ticket is there made out. Since the call is
boolced to the subscriber's number no money transaction has to be entered
into by the operator.
The whole of the cabinets at the Stock Exchange are multipled on to
tvery switch section in the central exchange. A call wire is provided
running round all the sections to the Stock Exchange operator. This
operator specifies the cabinet to which any call shall be put through, and,
therefore, the fact that every cabinet appears upon every section does not
lead to confusion or the design of engaged tests.
When the call matures, the section operator enters the down call wire
to the Stock Exchange by depressing her, call-key, and asks where to
connect the subscriber. The Stock operator replies by mentioning the
subscriber's number and that of a disengaged cabinet. The section
operator forthwith makes the necessary connection. An attendant stands
by the Stock Exchange operator and hears the information given to
the section operator, which he repeats by shouting. This is the method
in which the subscriber is called. An incoming call is dealt with in
precisely the same manner. In reply to the section operator's request for
a cabinet for the required number, the Stock operator gives the number
and that of the cabinet, which is loudly repeated by the attendant, and
thus the wanted subscriber is advised if he is in the exchange.
Now, in order that the Stock Exchange operator may know which
cabinets are engaged, a current is normally sent from the Post Office over
the loop through bridging coils and an earthed galvanometer. When a
peg is inserted at any switch section, this current is stopped, thus
indicating cabinet engaged. At the switch sections, the cabinets are
connected to eight-point switch-springs. The cabinet lines are, of course,
teed on to the respective line springs of all the switch sections. The
indicating or clearing battery for each circuit passes through the lower
and central springs of each switch-spring on to the inner springs of the
last switch-spring; thus the insertion of a peg at any switch section stops
the current, and thus serves as a reminder to the Stock Exchange
operator that that particular cabinet has been engaged. In the busiest
part of the day, the operator can always see which cabinets are vacant
by glancing at the galvanometers.
The silence cabinets are joined up similarly to those shown on page 123,
save that the seat contact is omitted and no counter switch provided.
The permanent current is restored when the speaker puts back the
receivers. This actuates the section operator's exchange galvanometer,
and the peg is withdrawn, thus restoring the indicating current, and
thereby showing the Stock Exchange operator that the cabinet may now
be used again.
stock Exchange Cif cutis. lag
It may be mentioned that the cabinet lines each pass through two five-
point switch-springs, in order to provide a means of cross-connecting
lines and cabinets when necessary. Such a necessity would arise if two
circuits were faulty, one in the cabinet and one in the line or Post OfiSce
switch-springs. One good circuit could be made up from the two faulty
ones.
130 The Telephone System of the British Post Office.
CHAPTER XXII.
Inductive Disturbances.
Unless special precautions are taken it is impossible to obtain a silent
telephone circuit. Neighbouring wires act inductively upon the circuit,
thereby introducing currents which cause disturbance. The telephone
receiver is so extremely sensitive that these currents produce effects
which very often render speech impossible. In order that we may clearly
apprehend the way in which these induced currents are produced, it may,
perhaps, be desirable to consider one or two very simple cases of
induction.
©
Figure 107.
Consider three insulated brass balls, A, B, and C (Figure 107). A and
B are close together, and C is at some distance from both, but is joined to
B by a copper wire. If now the positive pole of an earthed battery be
joined to A, it (A) will become positively charged. This will attract a
negative charge in B, which it will hold bound, the positive being repelled.
This positive charge will flow along the wire to the ball C, which will
receive a positive charge. In other words, charging A will cause a
current to flow from B to C. It will be noticed that the effect of A upon
C has been neglected; its effect is merely to reduce the quantity of
electricity flowing from B to C ; in fact, we may say that A exercises
precisely the same influence upon B and C, but that owing to the greater
Electro -static Induction. 131
proximity of B the charge induced in B is correspondingly greater. Since,
then, the two spheres are at different potentials, a flow of electricity takes
place.
Let us now consider what happens when A is discharged. The bound
negative charge upon B is set free, and is neutralised by the free positive
formerly existing upon B, and by the positive upon C. In order that the
positive upon C may get to B, it has to flow along the wire, so that a
current in the opposite direction to the first current flows along this wire.
Thus we may say that any change in the electrical condition of A pro-
duces currents in the wire between B and C,a positive increment in charge
producing a current from B to C, and a decrement of charge producing a
current from C to B. If these balls were sufficiently large, a conversation
could be held between A and B by placing a telephone in the wire
between B and C, and one between A and earth. This is a typical
example of electro-static induction.
Now to apply this knowledge to the case of a circuit, imagine a metallic
circuit B C, beween X and Y, with telephones at either end (Figure 108).
+ + +A+ -f +
~" B "~
+ + + +
+ c +
>
Figure 108.
The wire A is the disturbing wire, and is much nearer to B than to C,
In the normal condition there are equal quantities of positive and negative
upon B and upon C. Now, immediately A is positively charged, the
negative electricity in B is attracted and the positive repelled to C, since
the repelled positive charge upon C is so much smaller than that upon B.
Obviously this can only take place by currents flowing through X and Y.
If the charge upon A be still further increased, then a further current
flows from B to C. If now the charge upon A be reduced, some of the
bound negative upon B is released and flows to C through X and Y. A
negative charge passing from B to C is obviously the same thing as a
positive current flowing from C to B ; therefore we see that every varia-
tion in the charge upon A causes corresponding currents through the
telephones at X and Y.
The connection of the positive pole of an earthed battery to A at either
end charges it positively. If the wire is disconnected at the distant end
K 2
132 Dynamic Induction.
the potential is the same at both ends, but if the distant end be earthed
the potential at the battery end will be that of the battery, but will be
zero at the other ; hence the charge upon an earthed wire is half that upon
the same wire disconnected, or, what comes to the same thing, we may
say the capacity of the line is half that in the former case. Any change
in the value of the current flowing along A will induce currents in X and
y, and conversations taking place along A will be heard at X and Y ; and,
similarly, conversations taking place between X and Y will be overheard in
A. This is static induction. If the circuits did not possess capacity,
obviously charges could not be induced upon them, and thus overhearing
due to this cause could not take place.
There Is yet another kind of induction, viz., electro-magnetic or dynamic,
and this is due to the magnetic lines of force due to a current cutting
conductors. Let us consider the circuits A and B C (Figure 109). Let us
■<
Figure 109.
suppose that a current is suddenly started in A in the direction shown.
The lines of force from A cut through B and C, thus inducing currents in
them flowing in the opposite direction to the originating current in A.
Since B is nearer to A than is C, the magnetic field is greater at B,
consequently a higher E.M.F. is induced in B than in C, and this E.M.F.
overpowers that in C, thus causing a current to flow round the loop as
indicated by the arrow heads. Thus, cmy changes in the value of the
current flowing along A will produce corresponding currents in X and Y.
This is a typical case of dynamic induction.
Both static and dynamic induction take place at the same time in the
cases shown where one wire of a loop is nearer to any disturbing source
than is the other. Referring to and comparing Figures 108 and 109 it
will be seen that a current as shown in Figure 109 will charge the wire A
positively ; hence we shall have currents Induced in B C due to static
induction as shown in Figure 108, and at the same time currents due to
dynamic induction, as shown in Figure 109. At X these currents add
together, since they are flowing in the same direction, but at Y they are
flowing in opposite directions and oppose each other. Here we see an
Symmetrical Twist System. 133
explanation of tfae curious phenomenon that whilst one end of a circuit
may be quite silent the other may be very noisy indeed. At the silent
end the static and dynamic induction is equal and opposite, whilst at the
other they add together. This state of affairs is somewhat rare, but there
are certain faults which will bring about this result.
If now the disturbing wire were as near to B as to C, the inductive
effects upon the two wires would be the same, and no effect would be
produced (Figure no.) Static induction would rot affect X and Y, since
4- + +^>-f- f 4-
Bi ca
Figure iio.
the repelled positive charges upon B and C would be equal. Dynamic
induction could not produce any effect upon X and Y, since the currents
induced in each wire would be equal, and none could flow from one wire
to the other.
In order to overcome the trouble due to induction from neighbouring
circuits, two courses are open to us : The currents induced upon each wire
must be made equal, and then, since they would tend to flow round the
loop in opposite directions, no effect would be produced. Prof. Hughes
solved this problem by his suggestion of symmetrically twisting each
4. I .3 s^ljZ a, I .-I '. I ^^
1-PoLE assooLP 3a=PBLE 4™ Pole
Figure hi.
pair of wires. The other alternative is to cross the A and B lines ; but
these crosses become exceedingly complex where a large number of wires
have to be dealt with, and though the system is much used in America it
is unquestionably inferior to the symmetrical twist as carried out by the
British Post Office.
The wires are twisted in fours, each wire forming the meeting-point of
two sides of a twelve-inch square. The diagonal wires each form a loop.
The four wires perform a complete revolution every four poles. The
positions taken by two pairs of wires are shown in Figure iii. Nos. i
134 Symmetrical Twist System.
and 3 are the A and B lines of the first circuit, and Nos. 2 and 4 those of
the second circuit. By following the changes of No. i it will be seen that
in turn it occupies each one of the insulator positions, but with No. 3 always
diagonally opposite to it. Nos. 2 and 4 similarly revolve between Nos. i
and 3. At the fifth pole the wires assume the position shown for the first
pole. The distance from No. ij to No. 4 is twelve inches, as also is the
distance from No. ± to No. 3. Similarly the distances from 4 to i and
to 3 are equal, thus the currents induced will be equal. Now let us
consider the effect of bad regulation of the wires. Suppose at the first
pole No. 2 dips far more than do the others, No. 2 will then be closer
to No. 3 than No. i, hence induction will take place. Again, the canting
of the arms will bring about a similar result. In fact, anything which
tends to destroy the square will give rise to inductive disturbances. Lack
of uniformity in the spans may sometimes cause trouble owing to local
circumstances, such as the presence of trees. This increases the capacity
of the nearer wires more than the more distant ones by bringing the earth
Figure 112.
up nearer to the wires. All these difficulties are to a very large extent
overcome by symmetrical twisting and careful maintenance.
When through any fault more current passes along one wire of a
telephone circuit than the other, manifestly this is equivalent to a single
telephone circuit. If now we indicate a telegraph circuit upon the same
poles as telephone loops, we shall see that it will have no effect upon the
perfect and symmetrically twisted telephone circuits (Figure 112).
The dotted wires form one loop, and the full lines the other. It will be
seen that A, B, C, and D balance a, b, c, and d. The effects produced
are equal and opposite the two wires of each loop.
Let us now see what happens in the case of telephone lines which are
not perfectly insulated. Where the lines are equally and uniformly
insulated the insulation resistance of the lines is equivalent to a fault at
the centre of each line equal to the insulation resistance of the lines.
Where the insulation resistance of the two lines of a loop is equal, no
effect will be produced upon the circuit provided it is uniformly twisted.
Effect of Faults.
135
It will be seen (Figure 113) that each earth fault clears half of each line's
charge. The neutral point is at each telephone, and therefore no d scharge
passes through the telephones. Where the insulation of one line is lower
than that of the other, the lower resistance fault clears some portions of
the other line, and, in order that this may occur, current must pass
€
'<--
V
f f f V f
^S_r -
>
/
Figure 113.
through the telephone. Where the capacity or resistance of the lines of
the loop are unequal, the neutral point of the discharges is altered, and
thus more or less current flows through the telephone and we have noise
from neighbouring circuits. Where a definite fault upon one line exists,
the neutral point would be moved, as shown in Figure 114. Here we see
that currents pass through the telephone, but it should be noted that all
these troubles are due to static induction. In only one case in twenty can
inductive disturbances be traced, even in part, to dynamic induction,
and then the majority of the trouble is due to static inducfion. Again,
it is somewhat difficult to separate dynamic and static induction and
direct leakage from circuit to circuit.
It should be noted that under no circumstances should circuits be
twisted with the pole between them, as shown in Figure 112. Forty-eight-
inch arms are used and four wires upon either side of the pole fill two
arms. To maintain the insulation as perfect as possible, Cordeaux's
double shed white insulators are employed. This is necessary with
even the shortest trunk lines, as they are liable to be connected to
long lines. From a consideration of the pnnciples enunciate J pre-
viously it will be seen that a small fault upon a short circuit may not
cause any disturbance upon that circuit, yet when connected to a long
trunk in perfect condition the disturbances caused will be sufficiently
pronounced to prevent speech. A full earth upon a junction line or
136
Symmetrically Twisted Cables.
.'.ubscribers' line may not be observed till it is connected to a trunk
line having a large capacity. In fact, the conditions of Figure 114 will be
reproduced. Thus we see the cause of the phenomenon that two circuits
Figure 114.
apart may each be silent, yet when connected together are excessively
noisy.
It may be mentioned that the wires in cables are always symmetrically
twisted either in pairs or in fours. In the latter case the diagonal wires
'hould always be chosen for forming each loop.
The Telephone System of the British Post Office. 137
CHAPTER XXIII.
Superimposed Circuits,
It has been found possible to obtain three circuits where only two pairs
of wires existed between two points. The third circuit is termed a super-
imposed circuit, since the principle of its application consists in using
the first trunk as an A line and the second as the B line, bunching
together the two lines of each trunk for this purpose.
In order to accomplish this, four transformers are used. These trans-
formers are in reality specially designed induction coils. The primary
coil is wound upon a bundle of thin iron wires, the ends of which project
far beyond the windings. Over this is wound the secondary coil, con-
sisting of twice the number of windings, but with the centre of the coil
specially connected to a separate wire, known as the tap or centre of the
secondary. The iron wires are now bent over the coils from either end,
carefully covering each other, so as to form a closed magnetic circuit.
The instrument is indicated in Figure 115 as made by Messrs. Ericsson.
Figure 115.
The two terminals marked S are the ends of the secondary, the single ter-
minal the centre of the secondary, and the central terminals the ends of the
primary. The iron wires from the left hand side extend to the outer
margin of the right hand brass band, and those from the right to the
outer margin of the left hand side.
When a current is stopped, started, or altered in strength, the strength
of the magnetic field flowing through the iron wires is altered in value,
138 Principle of Superimposing.
and E.M.Fs. are thus generated in the secondary in exact accordance with
the variations in the primary. Thus it will be seen that if telephones be
connected to the primary and secondary terminals of the transformer, it
will be possible to speak from the one to the other just as though a
metallic connection existed between the two circuits. Similarly generator
currents will be reproduced by the secondary, and will drop an indicator
or ring a bell as the case may be.
Two trunks with a •• + i " circuit superimposed upon them are
shown in Fig. 116. The thick windings terminating upon switch-springs,
marked Nos. i and 2, at A and B are the primary circuits of four trans-
formers. The trunk lines are connected to the secondary windings. If
now two subscribers are connected to No. [ switch-springs at A and B,
they will be able to converse with ease. The A subscriber's currents will
pass through the primary of transformer i, thus inducing currents in its
Figure 116.
secondary which pass along the trunk line through the secondary of
transformer 3, thence back along the trunk line to the secondary of trans-
former i. The currents passing through the secondary of transformer 3
alter the value of the magnetic field circulating round the iron wires and
thus currents exactly corresponding to the original ones are induced in the
primary of transformer 3 at B. These currents circulate through the
apparatus of the subscriber connected to No. i switch-spring at B. Similarly,
B subscriber speaks to A subscriber through the two transformers. No. 2
circuit is arranged in a precisely similar manner and subscribers connected
to this circuit may similarly converse. If now two subscribers are con-
nected to the superimposed circuits they will be able to converse without
producing any effect upon the circuits No. i and 2, and without those
circuits producing any effect upon them. The path of the current is from
the long spring to the centre or tap of the secondary of transformer i,
where the current splits equally through the two windings, passing in the
Conditions Necessary for Superimposing. 13$
same direction along the A and B lines of the trunk to the ends of the
secondary at B, at the centre of which the currents re-unite and flow
through the subscriber's apparatus connected to B to the centre of trans-
former 4, where it again splits, equally flowing along both lines of the
trunk to the centre of transformer 2, where the two portions unite and
flow back to the subscriber at A. Thus, it will be seen that the subscribers
connected to the + i circuit are able to converse. No effect will be
produced upon circuits i and 2, since equal currents flow through each
half of the secondaries, but in opposite directions. In the case of No. i
circuit it will be seen that the currents flowing from that circuit pass
through the primary, thus inducing currents flowing , from end to end of
the secondary in the sime direction throughout, whereas the currents
from No, i circuit would, in the one case, be in the same direction, and, in
the other case, in precisely the opposite direction to this current ; and, since
currents circulating in the two halves of the secondary are equal, no effect
is produced, since the magnetic fields would be in opposite directions.
This has the further beneficial effect of annulling the self-induction of the
coil for the -f- i circuit, the winding being, so far as this circuit is con-
cerned, exactly like that of a resistance coil, where the wire is doubled
back upon itself. Since the + i circuit is connected to the centres of No. i
and No 2 circuits, there is no current flowing through the -|- i circuit from
Nos. I and 2. Thus it will be seen that all three circuits may be speaking
at once without interference the one with the other.
Two perfect circuits have been postulated in this case, but as these are
things which exist only in imagination it will be well to oonsider what
happens with ordinary circuits. Upon each circuit there will be a certain
amount of leakage, owing to the insulation resistance of the circuits not
being perfect. In the case of circuits in good order, we may imagine this
insulation leakage as due to single faults at the centre of each line. In
the case where the insulation resistances are all equal no trouble is caused,
but in the event of unequal insulation derived circuits from one circuit
to another are formed. If the two lines of circuit No. i differ in insulation
resistance, conductor resistance, or capacity, the current splits unequally
from the superimposed circuit, thus causing overhearing.
The conditions for obtaining a satisfactory circuit are : —
The circuit shall not exceed 50 mites in length.
The two circuits must form the diagonals upon two arms through-
out, and must be carefully twisted and regulated.
The circuits must not differ in conductor resistance, insulation
resistance, or capacity.
These conditions will be rendered more apparent by a re-perusal of the
last chapter.
It will be apparent from a consideration of Figure 116 that it will not be
possible to employ the permanent current system upon circuits No. i and
140
Non-polarised Indicator Relay.
No. 2, since the primary circuit is not in direct connection with the trunk
line. Thus it is necessary to ring through by means of a generator.
Only the two line springs with their inner springs are used to join up the
indicators. The inner springs are connected to the indicators. Upon the
superimposed circuit it is possible to work with permanent current, since
circuit at either end is connected directly to each other, via the trunk
lines. This is accordingly done.
The form of indicator used upon circuits upon which a circuit is super-
imposed is known as the non-polarised indicator relay (Figure 117). It
differs but little from the polarised form used upon trunk lines for
real size.
Figure 117.
permanent current working. The chief point of difference consists in the
absence of the permanent magnet which magnetises the armature in the
case of the polarised form. The armature of the non-polarised relay is of
soft iron, and passes over the projecting core of the right hand core
and under the left hand one, therefore a current in either direction
attracts the armature and thus closes the local circuit. An alternating
current will therefore permanently attract the armature. The needle is,
however, rendered magnetic by the permanent magnet upon which the
soft-iron needle is pivoted. An alternating current, such as is furnished
by a generator, therefore causes the needle to oscillate, and at the same
time closes the local circuit.
Actual Connections.
141
The complete connections for a superimposed circuit at the test board
and section are shown in Figure 118. It will be seen that five sets of four
test holes are taken up at the line test box (see Chapter XXIV.). The
two circuits upon which a + i circuit is superimposed are marked circuits
I and 2. Circuit No. i passes through the top pair of U links, through
the secondary of the first transformer ; the primary passes through the
second set of U links on to the line springs of the trunk switch-spring. It
Section
CcT.ZCO/mf -)
Figure 118.
will at once be obvious that ringing will have to be done by means of the
generator, since there is no direct connection between the switch-spring
and the circuit. A non-polarised indicator relay is substituted for the usual
polarised relay, and the batteries which would be necessary for permanent
current working are cut out. The connections at the sections are not in
any way altered. At the battery tablet the main permanent current leads
142 Ringing Arrangements upon the Transformer Circuit.
to the section are connected together by the insertion of a U link
horizontally. This puts the relay directly on to the inner sptiigs without
any other connection. The local permanent current battery is not re-
quired, and therefore the U links are removed altogether. It will thus be
seen that the tpp three springs are not used at all upon circuits i and 2,
Circuit 2 is precisely the same as circuit i .
The centres of the secondaries of the two transformers go to the lower
pair of the fifth set of test holes, and thence through the U links to the
line springs of the trunk switch-spring. Since this circuit is in metallic
connection with lines at both ends, the ordinary permanent current system
is employed. The top and bottom holes upon the battery group are
respectively the positive and negative poles of the battery, and it will thus
be seen that the main permanent current battery is joined up in series
with the polarised indicator relay, and the whole arrangement connected
to the inner trunk springs. The local permanent current is bridged across
the right hand coil of the relay, with 350™ in its circuit. These circuits
are, of course, broken when a peg is inserted.
The needles of circuits i and 2 hang vertically, whilst that of -{- 1 is
deflected to the right in the normal condition. It has previously been
pointed out that the generator must be employed for ringing upon circuits
I and 2, and in order to emphasise this distinction the lower half of the
indicator relays are painted black. As the generator is here required both
to ring the trunk and to ring upon the junction, the two pairs of cords
opposite circuits i and 2 are both connected to the generator. The black
peg is fitted with a red cord, and the ringing key in connection therewith
has a black ring round it. Half of the black peg is also coloured red.
The circuits upon which a + i circuit is formed are termed the " trans-
former circuits " on account of the transformers included in their circuit.
Such circuits should not be used for long through calls.
Occasionally, owing to climatic conditions or to faults upon circuits i
and 2, it is not possible to work a superimposed circuit, and it then be
comes desirable to .cut it off. At the same time the transformers which
are in the line circuit are not necessary. By removing all the U links
and inserting four links in the dotted positions the two circuits are con-
nected to the trunk switch-springs direct. Where there is a slight fault
upon say No. i we shall be able to use it, and of course No. 2, but if the
-I- I circuit were also connected up we should only be able to use one
circuit, as faults cause what is said upon any circuit to be overheard upon
the remainder. In actual practice it is not found possible to obtain a
satisfactory -|- 1 circuit upon circuits exceeding fifty miles in length.
Before a superimposed circuit is pronounced satisfactory it should be
subjected to the following tests : Ring with the generator upon circuits
1, z, and -I- I in turn, listening upon the remaining circuits. If the
generator can be heard the circuit should be abandoned. The cause of
Testing Superimposed Circuits. 143
tfais should be traced, and measurements taken of the conductor and
insulation resistance of each wire. Dry joints are fatal to superimposing.
Next, the circuits should be carefully listened upon for overhearing from
other circuits upon the same set of poles. If this is observed, then crosses
should be made upon the line till the overhearing vanishes, unless, of
course, it is due to bad regulation of the wires. The behaviour of the
circuits should then be watched for a week or two, noting how often it is
necessary to suspend the superimposed circuit.
144. The Telephone System of the British Post Office.
CHAPTER XXIV.
Test Room Appliances.
In order that faults upon trunk lines, etc., may be rapidly localised and
remedied, the provision of a suitable test board at once becomes essential.
At the smaller offices where a single section only is necessary a very
simple form of test case is employed. This accommodates the trunks,
junctions, and other circuits, but no provision is made for batteries, the
leads being taken direct from the battery rack to the switch section.
Each circuit passes to two test holes which are connected to the two
similar test holes in connection with the switch section by means of U
hnks. Each circuit has then four test holes allotted to it. In order that
the apparatus upon the switch sections may not be damaged by lightning,
protectors are fixed, but upon the instrument or switch section test holes.
The object gained is that by no crosses is it possible to leave an unpro-
tected circuit connected to the switch section. If a through wire is ter-
minated upon the section the protector is upon the section, and thus no
damage can be done.
At the larger offices test boards are employed, and here provision for
batteries is included. First, however, it will be well to consider the
design of the test holes. These test holes are made of brass chemically
tinned for the sake of appearance. They consist of a brass tube a secured
to an ebonite base b by means of a nut and washer d. A hole e is drilled
through the tube, and through this the connecting wire is passed. It is
bent over the tube and securely soldered, thus making an excellent joint.
Into the end of the test hole is screwed a brass plate, carrying a carbon
disc /, which is one of the pairs of plates of the lightning protector. The
mica disc has three small circular holes cut into it, through which dis-
charges may take place from the plate / to h, which latter plate is con-
nected to earth by means of the clamping spring g, which is used to
mechanically hold the plates together. Thus it will be seen that the two
plates are separated only by a small air gap. The distance between the
centres of these test holes is one inch, and therefore it is necessary to
make the test pillars of different lengths in order to accommodate the two
protectors required for the A and B lines. This will be amply apparent
from the lower part of figure 119,
Lightning Protectors.
145
The action of all lightning protectors is very much the same. When
dealing with lightning we confront quite a new aspect of our subject.
The currents are of an exceedingly sudden character, hence protection
which will suffice for slow heavy currents will be useless in the case of
lightning. If an explosive such as gun-cotton is laid upon a stone in the
open air and is ignited with a match nothing beyond a "fizz "will take
place, but if fired with enormous rapidity by means of a detonator, the
air above it, possessing inertia, will not move instantaneously with sufficient
rapidity, i.e., it will offer an enormous amount of inertia resistance, with
Figure iig.
ihe result that the stone will be blown to pieces. Lightning behaves
in an analogous manner to the gun-cotton fired by a detonator. The
electro-magnetic inertia resistance offered by a coil of wire to a lightning
discharge is enormous, and, if given a chance to rebound and avoid
traversing the coil by cracking through a small air gap, it will naturally
do so as a path of lesser resistance. This means that a current cannot
reach a finite value in an infinitely short time, and the reason of this lies
in the self-induction of the apparatus. Wherever a current exists there is
a magnetic field in existence. Now this magnetic field is the result of the
current, and, in coming into existence, cuts through the current-convey-
ing conductors to which it owes its origin; and we have seen that
L
146
Tweniy-Five-CircuU Tablet Test.
wherever a magnetic field cuts a conductor an E.M.F. is generated. This
E.M.F. opposes the current to which it owes its origin. The more
suddenly the current is applied the greater is the opposing E.M.F., or, in
other words, the greater is the resistance offered. This resistance,
which differs from the ordinary ohmic resistance, is frequently termed
impedance.
The test board for two sections consists of two tablets of roughly the
same size and shape, having respectively twenty-five and thirty-two sets
of four test holes. The upper tablet is arranged in five rows of five sets of
z
3
K
10
z
o
F
o
r
ie in
z
h
,
^0
z
3
f! ?
Z
u r
3
a.
LINE -tablet-
Figure 120.
four holes to accommodate the lines. The upper pairs of each set of four
holes carry protectors, and are utilised for the lines from the test board to
the switch section. The lower holes are the trunk or junction lines
coming into the exchange from the roof or underground cellar, as the case
may be. The A line is on the left and the B line on the right.
To each switch section are brought five trunks and five junctions,
together with service circuits and, perhaps, a Post Office subscriber or
two, but the number of these is distinctly limited, and at the majority of
the smaller exchanges there are none. The five rows are appropriated as
indicated in Figure 120, the centre row being appropriated to the up and
Battery Test Tableii.
H7
down service circuits, the silence cabinet circuit, telegram circuit, and
perhaps a through trunk. Where the number of through circuits is large
a larger tablet is usually provided.
The battery tablet (Figure 121) is appropriated to ten main and ten local
permanent current batteries, together with the two batteries used to work
the visual and self-restoring indicators upon the two switch sections. In
addition there are other batteries which are required, as the bell local, reed
ringing and driving, junction clearing batteries, etc.
The top and bottom holes are respectively the positive and negative of
the batteries, and it will be observed that it is impossible to short-circuit
the batteries by the insertion of a U link in any position. The links are
inserted top to right and bottom to left.
\^
00000000000000 1
0000000000000000 1
0000000000000000 1
0000000000000000 1
° 1
BATTERY TABLET.
Figure 121.
These test holes, both line and battery tablet, are connected to cross-
connection tags at the base of the test board, to which the trunk lines,
batteries, etc., and the lines from the switch sections are joined. These
cross-connection strips each consist of five groups of four tags. The tags
are merely thin strips of brass fitted into ebonite blocks and clamped
thereinto by means of a thinner strip of ebonite secured by screws. Into
the end of each tag a hole is bored at right angles to its length. For the
line tablet ten strips are employed. The wires from .the test-holes pass to
one of the tags. This first set of tags is connected to its fellow by means
of tinned copper wire (Figure 122). The object of this arrangement will
be dealt with subsequently.
La
148
Cross-Connection Slips.
The battery tablet is only accommodated with one set of four tags for
each set of test holes. Eight strips are provided, and thus the four lower
tags are not required.
-o
A
O INSTRUMENT.
B
P LINE..
A
6
A
FiGDRE 122.
The general arrangement of the cross-connection strips is shown in
Figure 123.
Here it may be mentioned that by the use of protectors in separate
cases, or by the use of a smaller form of protector directly affixed to the
cross-connection strips, it is possible to place eight sets of five groups of
four test holes— i.e., forty circuits — upon a line test tablet of the present
dimensions. By this means, the size of the test board can be greatly
reduced.
The protectors are of similar design to those already described, but
differ in dimensions. Two small oblong carbon blocks (i" x J") are
separated by perforated paper, and held together by a spring which forms
the line terminal. The other block rests upon a strip of brass, which is
earth-connected, and upon the other side of which two similar protector
blocks are fitted.
Later Forms of Test Board, 149
A test board of a totally different character is now being tried. This
consists of nothing more nor less than rows of ten live-point switch-
springs. The trunk line is connected to the left hand switch-spring,
and the switch section side of that trunk is connected to the right
hand switch-spring. The inner springs are connected across. Thus
in the normal state of affairs the trunk line passes from the line springs
to the inner springs, thence to the section inner springs, and to the long
and short springs, which are in turn connected to the section. Thus
every two switch springs correspond to a, set of four holes upon an
ordinary test board. Now it is possible to get twelve of these strips into
the space of a line tablet — i.e., 60 lines instead of 25 or 40. In order to
perform crosses, etc., ciftular pegs and cords are used. To loop wires, a
II
MI .. 4| Line
CROSS - CONNECTION.
TAGS
il
ill] ill
Battery
TAGS
Figure 123.
brass peg is provided. All the testing apparatus is so arranged that
tests may be made with facility by the insertion of suitably connected
pegs. It will of course be quite obvious that, by plugging into the left
switch-spring of a pair, the trunk line side is connected, and by plugging
the right the section is obtained. Through wires take up two holes, one
for each side.
In a large exchange considerable trouble would be caused if the trunk
lines appeared upon the test board in the order in which they appear upon
the sections. For instance, three trunks to the same town might follow
three different routes. In tracing, say, a contact, it would be necessary to
traverse the board from end to end in order to pick out the wires
following the same route as the faulty wire. In all large offices the lines
are arranged in the order of the routes which they follow. It would not
be desirable to place the lines upon the sections in this order, and even
if it were so the removal of a trunk from one section to another would
be a very awkward matter.
150
Line Tablet Cross-Connections.
It is, then, necessary to have the wires both in geographical and in the
section order. In order to accomplish this object the double row of line
tags is utilised. The right hand set of tags contains the whole of the
trunks in the exchange, and the left hand set of tags contain the whole of
the positions upon the switch sections. As jive trunks are placed upon
each section we see that each strip corresponds to a section. It has
already been pointed out that the central set of five boles upon the test
board are used for through wires, counter-cabinet, and such-like circuits.
The Junctions and service wires take up two rows, and thus there are only
ten positions left for the trunks (i.e., the equipment of two sections). The
extreme left hand and extreme right hand strips are therefore devoted to
trunk lines. Upon the left side of each of these strips are the trunks and
sections in geographical order, and upon the right side in the section
order. The right strip then accommodates the wires from section I. The
Figure 124.
right strip at the right hand side accommodates the wires from sec-
tion II. If there are forty sections to be accommodated, then the extreme
right and left strips of each test tablet will be marked consecutively,
section I. to XL.
To take a case in point, let us suppose that we have to deal with two
trunks, Todmorden and Bacup, coming into our ofBce by different routes,
the former being No. i on section I, and the latter No. i on section XII.
The Todmorden lines come in upon the right hand tags and are joined
through to the left ones by means of bare wire ; thence they pass to their
position upon the test board. Here they are linked through to the switch
section— i.e., the top set of holes must be joined to the first position on
section XII. This is accomplished by taking a pair of wires from the
right hand side of the left tags to the right hand side of the first position
on section XII. Similarly, Bacup section side i§ crosg-cpnnected to its
proper positioi).
special Arrangements at Manchester. 151
The ten wires accommodating the iive trunks upon section I. are all
taken to one strip ; in fact, the right of the left sides of the trunk strips
are the positions upon which all the sections terminate. This enables us
to use a six-wire and a four-wire cable to take the whole of the trunks
upon a section. Had the trunks upon a section to go to different places,
five separate pairs of wires would be necessary. The same remarks apply
with equal force to the incoming trunks. Since the wires following the
same routes are put together, cabling is again possible.
In order to remove trunks to different places upon the sections, it is
only necessary to alter the pairs of wires used for cross-oonnecting pur-
poses. Suppose that it is necessary to place Todmorden upon the third
position of section XXXVI., that the trunk occupying that position is to
go to position i, section I., and that Bacup is to go to position 5 on
section V. (unoccupied at present). A pair of wires is soldered in the
place of the present pair upon the instrument tags of the route side of
Todmorden. These wires are then taken to the left side of the row of tags
accommodating section XXXVI., and are soldered to the third position.
The wire taken off in order to perform this last operation is taken to the
left side of section I., position i tags, and the wire removed from there is
taken to the fifth pair of tags on the strip appropriated to section V.
All this appears complicated, but in practice it is the simplest imagin-
able matter. The crossing of a few trunks or their removal to other
positions is readily dealt with without incurring either joints in the
conductors or a large expenditure of wire.
Where there are only a few trunks these cross-connection strips are not
always utilised, but in a large exchange they are essential.
At Manchester, owing to want of space, trunk lines only are placed upon
the line tablets. The junctions are accommodated upon a special set of
tablets placed above the ordinary line tablets. The record table circuits,
etc., are placed upon the battery tablets, and the batteries are temporarily
taken direct to the sections. Accumulators will subsequently be used for
the permanent currents, and these will be accommodated upon the fuse
tablet.
X52 The Telephone System of the British Post Office.
CHAPTER XXV.
The Universal Battery System.
The universal battery system consists in employing a single set of cells
to provide power for several circuits. Its object is to provide the necessary
electrical energy as cheaply as is possible. The maintenance of a large
number of primary batteries, both as regards stores and attention, is most
costly. At the larger exchanges where a great number of batteries would
be required, accumulators are employed, thus saving not only money
but also space.
In order to illustrate the principle of the system let us imagine that we
have a battery whose E.M.F. is 20 volts, and whose internal resistance
is absolutely nothing. Through a resistance of iooo» this battery would
send a current of 20 milliamperes. If, now, a second circuit of 1000 "
be added, a current of 40 milliamperes will be sent cut, owing to the total
resistance of the circuit now being half its former value — i.e., there will
be 20 milliamperes flowing through each of the iooo» circuits. Thus the
application of the second circuit does not alter the value of the current
Howing through the first. Similarly it may be shown that the applica-
tion of circuits of any resistance whatever will not alter the 20 milliamperes
flowing through the first circuit. If, however, our battery has a large
amount of internal resistance, the application of other circuits will largely
alter the value of the current flowing through the first circuit. This will
mean that the values of the current through the various circuits will
largely depend upon how many circuits are closed at the time, or, in other
words, the current will greatly fluctuate. Now this must not occur if the
working is to be reliable, and thus it will be seen that batteries whose
internal resistance is very low can only be employed.
The batteries tmployed must also be capable of supplying large currents
for a considerable time. Now, the accumulator or secondary cell fulfils
all these conditions. The smallest size of accumulator ig more than
sufficient for the drain of even the largest exchanges. As regards internal
resistance, this may for all practical purposes be considered as negligible.
The type employed are the E.P.S. K7 cells, having a capacity of about
100 ampere hours.
Since the internal resistance of the cells is negligible, the calculation of
voltages necessary for any particular purpose is remarkably simple, it
being merely necessary to multiply the current required by the resistwce
of the circuit.
speaking Fuse Tablets.
153
For working the operator's telephones a single cell, U., two volts,
sufiSces. About fifteen transmitters are worked from each cell. The
positive and negative leads from the cell pass through a system of switches
to the speaking fuse tablet, which consists of four sets of brass bars
carrying fuse terminals. These fuse terminals (Figure 125) consist of a
tinned brass screw with a long stem having a washer held firmly against
its inner surface by means of the spring (in compression). The general
arrangement of a single set of bars is shown in Figure 125. One lead
from the cell passes to the top bar and the other lead to the bottom bar.
Between these two bars are placed a set of fuse pillars corresponding
exactly with those upon the bars. The inner pillars are connected to tags
JUATOR
THROUGH SWITCHES
^}-^~^
'Ho
Iff
2li
Fuses
I e 3
u
Fuses
i
To Section I
Figure 125.
61 and 62 of the respective sections, and thus by placing pieces of fuse
wire, as indicated in the ab^ve figure, the cell is joined up to the sections
wuich it supplies with current. Tags 61 and 63 are connected together,
The use of secondary cells renders fuses essential, since the maximum
current which they will supply is very large — in fact, sufficiently large to
burn up the conducting wires, and thus run a great risk of fire as well as
injury to the cells. Normally, of course, the currents supplied are very
small, owing to the resistance of the circuits, but in the event of a fault
which causes a short-circuit a very large current may be supplied. In
order to avoid this, thin platinoid wires are placed in the circuit, and
when the strength of the current reaches one ampere the thin wire is
fused and the circuit broken, thus avoiding all risk of fire. It will be
()uite obvious that H is not possible to insert a fuse between the cells and
154
Prevention of Overhearing.
the point of distribution, owing to the necessity for keeping down the
resistance. The speaking fuse tablet is placed inside the case containing
the charging switches in order that the length of wire between the cells
and the point of distribution may be as short as possible.
Usually two sets of four cells are provided for speaking purposes. This
enables sixty operators' speaking circuits to be dealt with without
diflSculty. The speaking tablet, arranged in four rows, also accommodates
sixty circuits.
It is absolutely necessary that the resistance of the cells and leads,
together with the necessary contacts up to the bars, shall be very low — ^in
3' 'A
Figure 126.
fact, under one-tenth of an ohm. Let us suppose that there are fifteen
transmitters joined up to these bars. If now the resistance before the
point of distribution is large, then speaking upon any one circuit will be
heard upon all the remainder. The working of a telephone depends upon
the transmitter varying the current flowiag through its primary circuit.
If now these variations in current alter the value of the current flowing
through other circuits then clearly overhearing takes place. It is only
where the resistance of the cells, etc., to the point of distribution is ex-
ceedingly low that overhearing does not take place. There is always a
slight amount Qf overhea.ring present, but un4er noriQP'l cooditions it i$
Universal Permanent Currents.
155
not discernible unless the room is perfectly silent and the speaking is
carried on in a different room to the listening. The leads between, the
cells and the bars are usually of seven-sixteens stranded copper. The
points where the resistance comes in is, of course, in the various contacts
and connections, which are, on account perhaps of the acid-laden air, more
or less liable to rapid deterioration.
In order to work the main and local permanent currents from the sig-
nalling accumulators, several modifications have to be made. Let us
suppose that all the permanent current leads were grouped, just as they
are, on to a set of accumulators. We should now find that all the circuits
were noisy until a peg was inserted which cut off the signalling circuits.
This noise is due to laqk of symmetry, since the A lines of all the circuits
not in use are directly joined together, whereas the B lines have to pass
through the trunk relay to become connected. This ill-balanced circuit
would cause trouble by overhearing. An operator at say Bury, listening
upon a disengaged Manchester-Bury circuit, might hear an operator at
Carlisle calling out upon a disengaged Manchester-Carlisle circuit, due
to the fact that the Manchester-Bury and Manchester-Carlisle trunks
were put in contact through the lead of the main permanent current
battery
All difficulty may be avoided by placing the main permanent current
battery in the centre of the coils of the relay, instead of at one side. This
necessitates the changes in the connections of the apparatus indicated in
figure.
In the first place the coils of the relay have to be divided so as to place
the main permanent current battery in the middle of the coils. The point
formerly used for the local contact is appropriated for the end of the left
coil. The local contact is joined to the end of the left coil, which is
connected to the battery. Now in order that the local permanent current
may not cause trouble, a resistance coil of 1200" is used in place of the
350o> coil and a larger voltage is therefore required. In order to preserve
the balance as regards the main permanent current the right hand coil
is permanently shunted by 1200" — i.e., by the same resistance used for the
local permanent current. Eight volts is required for the main and also
for the local — i.e., 16 volts for both. Now, in order to avoid adding
fresh terminals to the relays, the bell local contact is connected to the
left hand coil, which is in turn joined to the positive pole of the main
permanent current battery — i.e., to the i6-volt lead. The 12-volt lead is
connected to one side of the day bell, and thus when the local circuit is
completed the i2-volt and i6-volt leads are connected to either side of the
bell, thus giving a ring due to four volts. Here, then, we have o, 8, 12,
and le-volt leads carried to each section. Now 12 volts is used for trunk
ringing, and 16 volts for the self-restoring indicators. One other lead is
^l$o required — viz., for the tr^insfer circuits and for the junction cleiurinp —
156 Signalling Fuse Tablet.
for which 32 volts are used. The generator has also to be provided for.
This occupies another two wires. Thus if two 4-wire cables are run to
each section we can provide all the signalling power necessary.
In the test room, near the test boards, a fuse table is placed. This fuse
tablet consists of rows of brass bars and fuse terminals arranged in a
fashion very similar to that described in the case of the speaking fuse
tablet. Each row, with the exception of the last two, has forty-five fuse
terminals, and there are six rows of pairs of bars. They are usually
appropriated thus: —
Row \ . — Transfer circuits. Sections i to 45.
Row 2. — Generator. Sections i to 45.
Row 3. — \- Main P.C., + S.R. indicators, + Local. Sections i to 45,
+ Trunk ringing. Sections i to 45.
Row 4. Main P.C., + Local P.C. Sections i to 45.
—Trunk ringing— S.R. indicators-Local P.C. Sections 1 1045.
Row 5. — Transfer circuits, transfer board, record table, transfer sections,
and odd voltages (30 circuits).
Row 6. — Junction clearing, sections i to 25. Odd voltages, 10 fuses.
Junction clearing, sections 26 to 50. Odd voltages, 10 fuses.
Thus there is a fuse in every signalling lead to every section. Now on
row 5 are placed the transfer board leads, the record tables (calling
16 volts, clearing 32 volts), the motor generator (30 volts provided as a
stand-by in case of necessary shutting down of electric light pla^it), the
electric clocks, etc., and, in fact, any odd voltages required for special
purposes. There is further space at the end of the sixth row for ten
circuits.
In order to prevent the fuses being blown every time a faulty circuit
is connected to a switch section, a 200" coil is placed at each section in
the generator leads and a. 30" coil in the trunk ringing lead for a like
pui'pose.
At the end of the fuse tablet, rows 5 and 6, is placed a tablet containing
switches for crossing the dynamotors. The sections are divided into
groups of fifteen, and four switches are provided, thus permitting the
number to grow to sixty without any alteration being necessary. The
switches are remarkably simple, merely consisting of six test holes
arranged in circular form with a test hole placed in the centre. This
central test hole carries a U link which is capable of connecting the
central hole to any one of the six outside holes. Now there are eight of
these arrangements in four rows. The top central hole is connected to the
top fuse tablet bar for the first fifteen sections, and the bottom central hole
to the lower first fifteen section bars. Similarly, the other three sets
AnangeikeHi of Cells. 157
of two switches are connected to sections 16 to 30, 31 to 45, and 46 to 60
respectively.
One pole of the first dynamotor is joined to No. i test hole upon each
of the four upper sets, and the other pole to No. i test hole upon each of
the four lower sets. No. 2 dynamotor is connected to test holes No. 2, top
and bottom, and the motor generator to No. 3. Thus we can put any
particular fifteen sections on to any particular generator by means of this
switch. The three spare holes are provided in case they should at any
future time be required.
The only point which remains to be dealt with is the accumulator
switches. These are placed in a special case in the same room as the cells.
In this case are the speaking and signalling switches, together with the
speaking fuse tablet already described. It will, perhaps, be most con-
venient to deal with the speaking battery switches first. Two sets of four
cells are employed. One switch, practically consisting of a series of
specially designed two-way switches, is used for connecting up either set
No. I or set No. 2 to the speaking fuse tablet bars. A second switch in
one position connects the cells not in use to the charging leads usually
taken from the ofiice lighting plant. In the second position the set of cells
not in use is insulated. An ammeter for observing the charging current
(10 to 15 amperes) and a switch for inserting resistance to vary it are also
interpolated in the charging circuit. An automatic switch is also inserted,
so that should any fault occur the cells may not discharge through it.
The signalling battery usually consists of two sets of sixteen cells, tapped
off as follows : — o, 8, 12, 16, 20, 24, 28 and 32 volts. The local permanent
currents put very much more work upon the cells than the main
permanent currents, and therefore it is necessary to carefully group the
cells so as to avoid having half the cells completely discharged whilst the
other half are practically fresh. In order to do this the sections are
divided into three groups : —
I'rom o to 8 takes Local P.Cs. , Group i.
„ 8 „ 16 „ Main P.Cs., Group i ; also Local P.Cs., Group 2.
„ 16 ,, 24 ,, Main P.Cs., Group 2 ; also Local P.Cs., Group 3.
,, 24 ,, 32 ,, Main P.Cs., Group 3.
It will be obvious that the sixteen volts taken to each section is taken
in three leads — e.g., Group II., 8, 16, 20 and 24 volts leads are utilised
8-16 local P.C., 8-20 trunk ringing, 16-24 main P.C, 20-24 hell local,
o and 32 are taken to each section for transfer circuits. The junction
clearing is taken from the telegraph accumulators at offices where these
exist, or from primary batteries where they do not. This is due to the
objection to having any earth connection upon the signaUing accumu-
lators.
158
Charging Switches.
The object of the switches provided is precisely the same as in the case
of the speaking circuits. In one position No. i set, together with all its
taps, is connected to the signalling fuse tablet in the test room, and in the
second position No. 2 set is similarly connected. The charging switch
insulates or charges the set of cells not in use. It is, of course, needless to
say that all leads to the points of distribution, viz., to the fuse tablets, are
of heavy wire, usually seven-sixteens stranded copper insulated with
vulcanised rubber. Fuses are inserted in all the leads from the cells
before they pass to the charging and controlling switches. In the case of
the signalling leads the slight resistance inserted is not of such vital con-
sequence. A fault at a switch section blows the small fuse on the fuse
tablet, but a fault between the fuse tablet and the cells blows the fuse in
the leads of the cells affected.
It has been found possible to considerably reduce the number of primary
batteries in use at the various exchanges by suitably grouping the various
set!
r'l'ij'M'itilj
Figure 127.
circuits on to a single set of batteries. The whole of the permanent
currents are arranged precisely as described in the case of accumulator
working, but as no fuse tablet is necessary the batteries are grouped upon
the battery tablet as shewn in Figure 127. Two sets of batteries are used,
one being a stand-by in case of failure. The left central holes upon the
first row of test holes are used for the leads from the sections. The top
holes are the leads from set i, and the bottom ones from set 2, so that it
will be seen that, by reversing the U links from the top holes to the
bottom ones, set No. 2 will be joined up in place of set No. i (as shewn).
The local and main permanent current tags are teed at the sections as
already indicated. Each section occupies two sets of test holes, so that
one battery tablet will accommodate the permanent current leads for
twelve sections.
The trunk ringing, self restoring, and bell local may also be worked
from this battery precisely as indicated in the case of accumulators. This
pyimary Batteries Used for Universal Working. 159
leaves four sets of holes for junction clearing and any other powers, such
as reed ringing or generator, as may be required.
The contacts upon the speaking keys for closing the self-restoring indi-
cators have been modified, so that the current only flows momentarily upon
raising or depressing the speaking key. This has been accomplished by
rounding off the side of the plunger which works the self-restoring contact.
The contact is only made when the key is half way up or down. This
improvement has greatly reduced the work upon the self-restoring battery,
and it has now become possible to group it upon the common primary
battery.
At all the larger exchanges one or more sets of bichromate batteries in
parallel will be used.
It has been pointed out that one battery tablet will now serve for twelve
sections. At an oi£ce of this size a six-panel test board having six line
and six battery tablets would at present be required. Now under this
arrangement five battery tablet spaces would be left idle, and by filling
these up with line tablets the size of the board could be reduced to four
panels.
A form of lightning protector has been designed which will fit on to the
cross-connection strips of the test boards. These protectors consist of
small rectangular carbon blocks insulated by perforated paper. Now by
dispensing with protectors upon the test holes it is possible to get forty
sets of four holes into an ordinary tablet — ».«., one of twenty-five holes.
By replacing the present tablets an exchange of twelve sections could be
accommodated upon a test board of three panels — i.e., five line tablets and
one battery tablet. Thus, the size of the test board may be reduced to
about half its present dimensions in this case. With a switch-spring
test board a still further reduction is possible. This is a most important
development in view of the value of space in Post Offices.
It will, of course, be obvious that it is not possible to work more than
one operator's telephone from a. primary battery owing to the latter's
internal resistance.
Post Office telephones cannot be worked from the secondary cells owing
to the primary and secondary circuits being in nnsymmetrical contact,
and hence telegram telephones are always worked by primary batteries
Similarly the primary and secondary circuits of switch-board telephones
have to be separated.
i6o The Telephone System of the British Post Office.
CHAPTER XXVI.
Small Local Exchanges.
The system adopted by the Company for dealing with their subscribers
varies very considerably with the requirements. Where the number
of lines is under 200, separate switch-boards, each accommodating 50
subscribers, are employed.
aSD
¥iHffi3HH^
FiGUiSE 128.
The general appearance of the switch-board and a sectional drawing
are shewn in Figujre 128. Each subscriber's line terminates upon a switch-
spring and an indicator. The switch-springs are acommodated at the top
Indicators and Switch-Springs.
i6i
of the board, and it will be observed that there are six rows of ten.
These ten extra switch-springs are provided for various junction circuits,
should they be necessary. Similarly the five indicators below the fifty
subscribers' indicators are provided for similar purposes. The pairs of
cords are placed upon a, small shelf above the indicators and below the
switch-springs, the object being to avoid the cords crossing in front of
and thus obscuring the view of the indicators. The last row of ten
indicators are for the ring-off signals, and upon the shelf below these
indicators are placed the ringing and speaking keys.
The subscribers' hnes terminate upon five-point switch-springs, the inner
springs of which are connected to low resistance (loo ») indicators, the
form of which is indicaftd in Figure 129. It consists of an electro-magnet
Earth
Figure 129.
A B with a small iron armature C carrying a lever D pivoted in front of
it. The lever catches the pivoted shutter E, which falls forward when
the armature C is attracted. Upon a brass disc, normally covered by the
shutter, the subscriber's number is painted in black figures. It may be
mentioned that these indicators occupy about one square inch each.
Thus, the subscriber turning the crank, discloses his number at the
exchange.
The ring-off indicators, which are in circuit whilst subscribers are con-
versing, are of the tubular form, very similar to the back coil of the self-
restoring indicator, but much smaller and more compact. Of these there
are ten, one for each pair of cords. The indicators are covered with an
outer iron sheathing to prevent overhearing from one circuit to another
ma the contiguous indicators. The low resistance line indicators are not
in circuit during conversations, and therefore do not need this provision.
The pegs are of precisely the same construction as those used by the
Post Office and need no comment,
M
1 62
operating Connections.
Only one pair of ringing keys is provided for the whole ten pairs of pegs
and speaking keys, and thus it becomes necessary to put these ringing
keys in with the operator's telephone— i.e., the speaking keys switch the
particular pair of cords through the ringing keys to the operator's
telephone. In order to accomplish this, a speaking or, as it is usually
called, a listening key of more complicated structure is required. The
ringing keys are of the same type as those used by the Post Ofl&ce, but the
mechanical design is slightly different.
:^ Ti*«(SMiTT6a.
Figure 130.
The complete connections of a pair of cords, speaking key, and the
operating connections are shewn in Figure 130. The five springs of the
speaking key are furnished with four contact stops, upon which they rest
when in the speaking position shewn.
In order that we may ring upon either pair of cords, both lines have to
pass through the ringing keys ; hence each speaking key, when in the
speaking position, passes the A and B lines of each peg to the four bars.
The four lines A B A^ B^ of the two pegs then pass to the long springs of
the ringing keys, whose inner springs are connected together. Teed across
Mixed Metallic and Single Wire Circuits. 163
Ihis connection comes the induction coil and operator's receiver. The B line
of the pegs is dotted throughout the diagram in order to facilitate tracing.
Depression of the left key rings on to A and B, and depression of the right
key rings A* B'. The depression of either key cuts off the other one as in
the case of the Post Office trunk system. It will be obvious that the four
bars marked " operating circuit tags " are common to the ten speaking keys.
The ring-off indicator is permanently teed to the A lines, but is only
joined up to the B lines when the key is in the normal position, in which
case the two left springs are in contact, thus connecting the A lines of the
two pegs, and disconnecting them from the operating circuit bars. The B
lines are similarly connected by the corresponding springs on the right
hand side, but in addiflbn to this the outer spring also makes contact, thus
joining up the ring-off indicator in bridge across the cords.
It will be seen that the induction coil has six terminals and that the
secondary coil is thus divided into two half-colls, the operator's receiver
being placed between them instead of at one side as is the case in the Post
Office system. The object of this will be seen later when the multiple
board is considered, but suffice it to say that in this case such arrangement
is quite unnecessary and has only been utilised for the sake of uniformity
of connections.
The subscribers' lines are all brought to soldering tags and lightning pro-
tectors placed in the space above the switch-springs. The protectors
merely consist of a series of serrated brass plates placed very close to
an earth-connected bar.
Where there are three or four of these switches, junctions between the
outer switches are provided to enable subscribers upon those boards to be
connected. These junctions may be worked by signalling or by speaking
direct to the operator concerned since they are so close together.
It has been assumed that these subscribers all had metallic circuits, but
where this is not the case the same board may easily be utilised. Let us
suppose that the subscribers all possess single lines. It is then only neces-
sary to earth the B line soldering tags and the side of the indicator
connected to the B line inner spring.
Where metallic and single subscribers exist together, earthing the B
line soldering tag would mean that when a single line was connected
to a metallic the two A lines would be connected together, and the B line
of the metallic circuit earthed. If the lines were short, this would matter
little, but were the metallic line a long one — as, for instance, a trunk line —
then the noise created due to static induction would probably preclude
conversation (page 136).
In order to avoid this, transformers have to be utilised. The primary
is connected to the single line with its other side earthed, and the
secondary is connected to the metallic circuit. This metallic circuit is
thus not interfered with, there being no earth connected to it, and thus
M 2
164 Translator Keys.
the only disturbance upon it is that from the single line. Where only a
few metallic circuits exist upon each board they are marked with a white
ring, and a special pair of cords joined up with a transformer is used when
it is necessary to connect them to a single line subscriber. When two
single subscribers are connected together, the B line of the pegs is not
required, but when two metallic subscribers are connected it is necessary.
To connect a subscriber to a Post Office junction, a transformer is
inserted or not inserted at the Company's exchange as may be necessary,
but the junctions are all metallic, and when a subscriber is connected
earth must not be used upon it.
At exchanges where all the lines are single ones, transformers are
introduced upon the junctions.
Where the subscribers are half metallic and half single, a set of
translator keys, in appearance very similar to a speaking key, is inserted
by the side of the speaking keys. In the normal position, this key
connects the A and B lines of the two pegs across directly. When
depressed, a transformer is inserted between the cords for connecting a
single to a metallic subscriber or junction.
The Tehphne System of the British Post Office. 165
CHAPTER XXVII.
The Series Multiple.
Where a very large number of subscribers have to be dealt with in one
exchange, a different method of working has to be adopted. If 3,000
subscribers were connected to sixty boards similar to those described in
the last chapter, only about if per cent, of the connections demanded
would be for subscribers upon the same board, and thus a certain number
of junctions would be required to each of the other fifty-nine boards. This
c-
B-
A-
9 9
9 ?
n . ' _ - 1 '
' F— '
u 1
A - - 1 '
r 1
•-T i" !"
iba
.
I' 't
( ^0
I
.iii
Figure 131.
could obviously be simplified by giving junctions to the central of each
three operators, but even then scarcely 5 per cent, of the calls could be
dealt with by one operator, and nineteen sets of junctions would be
required to the twenty sets of three boards. This would render the
working of the boards very slow, and the number of junction lines required
would be very large. The junctions between the switches might terminate
1 66
Principle of Multiple Switch.
upon switch-spfings only, and then nineteen call keys would be required
upon each section in order to instruct the various operators to put certain
numbers on to certain junctions. This class of system may be termed
the divided board system in contradistinction to the multiple system.
The difference between the two systems is illustrated in Figure 131,
where three subscribers each terminating upon three different boards are
shewn. In the divided board or junction system, A, B, and C's lines
terminate upon three different boards and are connected together by means
of junctions. In the multiple system every subscriber's line is brought
Figure 132.
to each board, and thus A may be connected to B or C at the first board
without the services of another operator being necessary. The multiple
system then consists in placing the whole of the subscribers within the
reach of every operator. Each section of the multiple board is six feet
six inches in length and contains the switch- springs and indicators of the
200 subscribers — i.e., the operators working this section answer the calls of
200 subscribers. Besides this the switch-springs of every subscriber upon
the exchange are repeated upon the board. The switch-springs belonging
Necessity for Engaged Test. 167
to the subscribers attended to upon any section are termed the home
switch-springs, and these are placed below the ordinary multiple. It will
be obvious that it is not essential to the making of connections that the
subscribers' numbers attended to upon a section should appear in the
multiple upon that section, but in practice this omission is never made, as
it would tend to confusion in operating.
The general appearance of a multiple switch is illustrated in Figure 132,
and it will be seen how similar in section it is to the board described in thr
last chapter. Its length is practically six times as great. The horn
indicators, pegs, and speaking and ringing keys occupy the same position!
as in the small board. The home switch-springs appear at the bottom of
the multiple, the duplteated switch-springs appearing above them.
It has now been shown how a subscriber ringing up may be connected
to any other subscriber he may ask for without the services of a second
operator. It therefore remains to explain how an operator may know
whether the subscriber asked for is or is not already connected and
speaking. For instance, say No. 17 asks for No. 1236. Now No. 123(1
may be connected at his own home switch-spring, or at any of th&
multiples. How, therefore, is the operator in charge of No. 17's line to
know whether he is engaged or not 7
This is accomplished in a very simple manner. The barrels of the
switch-springs upon each subscriber's line are connected together — i.e., the
barrel of No. 17's switch-springs on each multiple and upon bis home switch-
spring are all put into contact. The pegs used for making the connections
have three instead of two points. The tip is the A line, the neck the B
line, and the shoulder, which makes contact virith the barrel of the switch-
spring into which it is inserted, is connected to a small earthed battery.
Thus the insertion of a peg anywhere, whether in the multiple or the home
switch-spring of any particular subscriber puts this earthed battery upon
the barrel of this subscriber's switch-spring throughout the exchange. The
operator's telephone is earth-connected, and thus by touching the barrel of
this subscriber a click is received in her telephone, thus indicating that the
subscriber is engaged. The absence of the click shows that the subscriber
is not engaged, and that the required connection may be made.
Five-point switch-springs are used in the series multiple. The subscri-
bers' lines pass through the whole of the multiple switch-springs in series, as
indicated in Figure 133. The A and B lines pass from the line springs to the
inner springs, and thence to the next line and inner springs, and so on,
the final inner springs (those of the home switch-spring) being joined to the
home indicator. The B line has been dotted throughout to facilitate
tracing. Thus it will be seen that by inserting a peg in any one of the
switch-springs the A and B lines are picked up by the tip and neck of the
peg. A bad connection between any one of the line and inner springs
renders speaking difiicult, but such faults are extremely easy to localise.
i68
Connections of Engaged Test.
The test wires, as they are termed, are the wires connecting together the
barrels of all the switch-springs belonging to a particular subscriber.
Figure 133.
A connection between No. 17 and No. 1236 is shewn in skeleton in
Figure 134. The connection is made at No. 17's request, Since a peg is
%
'M r-if'-
Figure 134.
inserted in his home switch-spring and one in No. 1236 in the multiple.
Since No. 17 has his home switch-spring below the first multiple the
Connection is made upon the first multiple. It will be seen that the two
lines are connected together by the tip and neck of the two pegs. The
bodies of the pairs of cords are connected together to an earthed battery ,
and thus the test wires of No. 17 and No. 1236 are both joined to it. By
touching the barrel of either of these subscribers' switch-springs upon any
of the multiples with the tip of the peg a click is received in the telephone,
the centre of which is earthed for that purpose.
The complete operating connections are precisely the same as in the
case of the small 50-line board (Figure 130). We now see the object of
dividing the induction coil and of the earth upon the centre of the
receiver. In order to obtain a click in the receiver when the tip of
Method of operating. 169
the peg touches the barrel of an engaged subscriber, it is necessary
to have an earth upon the receiver. Now this earth can only be placed
in the centre of the circuit, as, were it placed elsewhere, disturbances
would be caused whenever an operator entered a circuit. The earth
is placed in the middle of the two coils of the receiver, and as there
is half the secondary of the induction coil upon either side no disturbance
is occasioned. When the speaking key is down the tip of the peg is con-
nected through one coil of the receiver to earth, and it is in this way that
the engaged click is received.
It has previously been pointed out that each multiple section contains
200 home switch-springs and indicators. These are dealt with by
four operators. A m'ultiple of the whole of the subscribers connected
to an exchange is given to every four operators. Thus in an exchange of
2,000 subscribers, ten multiples and forty operators would be needed.
This is, of course, exclusive of junctions, etc.
Each operator attends to the demands of fifty subscribers, and has
twelve pairs of cords with speaking keys, etc., allotted to her. Frequently,
however, the 200 lines are attended to by three operators, and the cords
then have to be slightly re-arranged.
The procedure in case of a call is : Say No. 17 rings up, his indicator
falls upon the home section ; his operator pegs in on the home switch-
spring and gets the number of the wanted subscriber. Suppose this is
No. 1236. She takes the corresponding peg of the pair and touches the
barrel of No. ia36's switch-spring on the multiple. If she gets a Click (the
speaking key is meanwhiledown)No.i7isinformed that No. 1236 is engaged.
If disengaged, the peg is pushed into the switch-spring, and the corres-
ponding ringing key depressed. Upon hearing the subscribers speak the
operator raises the speaking key and does not come in circuit till the
ring-off indicator is dropped by the subscribers turning their generators
at the conclusion of their conversation.
A good point in connection with this system is that a triple connection
is impossible. If two subscribers are connected and a peg is inserted into
one of their switch-springs upon the multiple, one of the subscribers may
be cut off, but it is not possible for the third man to listen to the conver-
sation of the other two.
An objection to the series multiple is that in an exchange of 5,000 sub-
scribers, the A and B lines each pass through twenty-six moving contacts.
It may be mentioned that with the object of avoiding the use of three
section pegs the seven-point switch-spring illustrated in Figure 100 (upon
record table switch sections) are sometimes used. The barrel of the .switch-
spring is electrically separate from the remainder of the switch-spring.
The long moving spring is connected to the test-wire of the subscriber to
which the switch-spring refers. The other point with which the moving
spring makes contact upon the insertion of a peg is connected to the
1 70 Mixed Systems,
engaged test battery. Thus upon the insertion of a peg in any subscriber's
switch -spring, the barrels of his switch-springs are connected to the
engaged test battery. The barrel where the peg is inserted is, of course,
disconnected from the moving spring by the insertion of that peg.
In mixed systems — i.e., partly metallic and partly single — translator keys
are introduced, and the metallic circuits specially marked by a white ring.
In single wire systems transformers are introduced upon the junctions.
Sometimes two switch-springs are allotted to each Post Office junction,
the one for single and the other for metallic subscribers, a transformer in
the former case being interpolated.
The Telephone System of the British Post Office, 171
CHAPTER XXVIII.
The Self-Restoring Board.
The self-restoring indicator makes it possible to arrange our switch-
springs upon the branching or parallel system. The A and B lines pass
right along the line of B6ards and are teed off at each switch-spring. As
the calling indicator upon the home switch-spring is also permanently
teed across the loop, it is essential that when a peg is inserted in the
multiple the corresponding home indicator should be locked, as otherwise
calling currents would drop it, and the operator would never know whether
the subscriber required attention or whether he was being rung up by
another operator upon another multiple. Also it is clearly necessary to
provide an engaged test.
The switch-springs employed are of a somewhat different form to any
we have previously considered. There are three springs, and the barrel is
divided into two parts. Three switch-springs and their connections are
indicated in Figure 135. The front part of the barrel is used for the test
Figure 135.
wire, and the back part for the B line connection. Two springs of the
switch-springs are placed opposite each other, and upon the insertion of a
peg are connected together by the neck of the peg. The third spring of
the switch-spring is the A line spring. The connections of the pegs are
slightly different in this form of board. The two tips and the two bodies
of the pegs are respectively connected together, the neck being left dis-
connected in each case. The insertion of a peg connects the A line to the
tip and the B line to the body. Thus is the connection between the two
lines of the metallic circuit effected.
172 Connections of Switch-Springs and Indicators. — -Engaged Test.
It now remains to trace the engaged test and method of locking the
self-restoring indicator. In Figure 135 a peg is inserted in the second
multiple switch-spring. The battery which is used to lock the self-
restoring indicator is also used for giving the engaged test. The circuit
of the battery is along the battery lead through the two springs con-
nected together by the neck on to the test wire, and thence through the
restoring coil to earth back to the battery. This current locks the indi-
cator, and thus ringing upon the circuit will not drop the home indicator.
At the same time, the front barrel is connected to the earth battery.
Figure 136.
The operating connections are very similar to those previously de-
scribed, and so far as the engaged testing is concerned are precisely the
same. The middle of the receiver is earthed exactly as in the connections
previously described ; thus by touching the front barrel with the tip of the
peg a click in the receiver is heard if a peg is anyiVhere inserted in that
particular subscriber's switch-spring. In the normal condition of affairs,
the subscriber by ringing drops the self-restoring indicator. To answer the
call, the operator inserts a peg and restores the indicator, and at the same
time connects the battery to the test wire, thus engaging the subscriber
throughout the room. The insertion of the corresponding peg into the
operating Connections.
173
wanted subscriber's switch-spring, after testing, locks his home indicator,
and also engages him throughout the room.
The front barrel of the switch- spring is made larger than the inner or B
line barrel, and thus the test wire is never connected to the body of the peg.
Were this done, an earth would be introduced upon the B line of the loop,
and this would, of course, cause disturbances.
In order to make the connection with the B line barrel perfectly secure,
Section of Table.
Flat Board with Self-Restoring Indicators.
Figure 137.
an " umbrella spring " is employed. This consists of a brass spring let
into the body of the peg.
The operator's connections beyond the four connection bars (Figure 136,
are precisely the same as previously described, and the connections
between the pegs and bars are not widely different. The ring-off indicator,
which is of the self-restoring pattern, is teed across the cords permanently.
The restoring coil is operated by a separate contact upon the speaking
174 General Arrangement, — Flat Boards.
key. From the Figure 136 it will be seen that when the speaking key is
depressed the indicator is restored.
In this system it is quite unnecessary to touch the indicators, since they
are all worked automatically, and therelore they may be placed above the
multiple, thus placing the switch-springs within easy reach of the
operators. The relief afforded to operators by not having to restore
the shutters by hand is not inconsiderable. The other details of the
boards are precisely similar to the boards previously described.
In order to reduce the number of multiples necessary, the flat board has
been introduced. Imagine the multiple switch-springs placed upon the
flat. We may now put operators upon either side of the multiple, and
thus roughly speaking only half the number of multiples will be necessary.
The keyboards and one cord of the pair (Figure 137) are placed upon
a flat shelf below the level of the flat board. At right angles to the flat
board are placed the home switch-springs. In the canopy suspended
from the ceiling above the flat board are placed the self-restoring
indicators corresponding to the subscribers' lines. Here also appear the
corresponding cords of the pairs and the ring-off indicators. The
electrical principles of the system are precisely the same as the system
which has just been described. Two hundred subscribers, each dealt
with by four operators, are placed upon either side of the multiple, i.e.,
one multiple is required for every four hundred subscribers. As before,
twelve pairs of cords are allotted to each operator. These multiples are
placed side by side, and thus the end operator reaches to the next multiple
for subscribers which would be beyond reach were one multiple only
provided. Thus, at the point where the multiple ends, a part of the
multiple, termed a. dummy section, has to be repeated for the use of the
outside operators.
The flat board has not met with much favour save in this country,
the objections urged against it being that it is impossible to effect repairs
upon the multiple during the daytime, and the accumulations of dust
which, being upon the flat, naturally result. Again, such things as pins,
bits of black lead [i.e., pencil points), etc., cause trouble. With the upright
board it is a. very easy matter to remove any strip of switch-springs
without disturbing the operators. A pump connected to a funnel is used
to draw the dust from the switch-springs by suction, and to a great extent
this gets over the diflBculty. However, the question as to the desirability
of flat boards is one which time alone can settle, but it may be mentioned
that the National Telephone Company's Engineer-in-Chief, Mr. Dane
The Telephone System of the British Post Office. 175
CHAPTER XXIX
The Call Wire System.
The call wire system consists in giving to every subscriber, besides his
ordinary circuit to the exchange, access to a second circuit. This second
circuit is common to about fifty subscribers, any one of which, by depres-
sion of a key, may pla£e himself upon it. At the exchange the circuit
terminates upon the speaking apparatus of the operator attending to the
fifty subscribers' lines to which it is common. It is upon this circuit,
termed the call wire, that the subscriber makes his demands for
connections.
i5et,'s
"* *"TESTv/infi
Test..
eJtSl
ivnn-xipuEa
Figure 138.
Since all instructions as to connections and disconnections are given
over the call wire, it is clearly unnecessary to provide indicators or
speaking keys at the exchange. The subscriber rings up his correspondent
himself and thus speaking keys are not required. In the single cord
system each subscriber's line terminates upon a peg, and is also connected
to the multiple switch-springs, which merely consist of two springs and a
barrel. The arrangement is depicted in Figure 138. No. 26x's lines are
176 Engaged Test Arrangements.
connected to a peg and also to the multiple switch-springs shewn. Thus,
to connect No. 1321 to No. 261, No. I3zi's peg is inserted into No. 261's
multiple .switch-spring, and the connection is complete.
In front of each operator are placed the cords of the subscribers who
have access to the call wire in that operator's charge. These cords are
placed in the canopy of switch, as the flat board system is again used here.
The question as to the engaged test has now to be considered. The
pegs are made in three sections as before. The third point is connected
permanently to an earthed battery, thus the barrel of any switch-spring
into which it is inserted is similarly connected. The barrels of the switch-
springs belonging to a subscriber are all connected together, and thus the
insertion of a peg joins the barrels of all his switch-springs to the earthed
battery, thereby engaging him upon all the multiples. Thus the called
subscriber is engaged. When the peg of the calling subscriber is pulled down-
wards to make a connection, the socket-spring S makes contact with the
point C. Now C is joined to the earthed test battery and the spring S is
joined to the subscriber's test wire : thus the pulling down of the peg
engages the calling siibscriber throughout the room.
It has previously been pointed out that the subscriber asks for connections
upon his call wire. At the conclusion of the conversation the calling sub-
scriber should again go upon his call wire and ask that the connection may
be severed. Let us suppose that he omits to do this or that the operator
fails to sever the connection. As this request was made by No. 1321 (Figure
138), No. 261 is powerless to get the connection severed, sinceit is very im-
probable that he is on the same call wire as No. 1321. Another point is that
No. 261 may require a connection shortly after he has spoken to No. 1321.
If now the connection were made by the operator, three subscribers' lines
would be joined together and that without No. 261 's knowledge. Therefore, it
is necessary not only to test the line of the called subscriber, but also that of
the calling subscriber. Accordingly every subscriber's test wire is also taken
to a small brass stud by the side of his peg. If now No. 261 asks for a con-
nection the operator touches the stud and, finding him engaged, tells him
to ring up his correspondent and ask him to call off.
The operating connections are shewn in Figure 139. The call wire passes
to the operator's telephone, and also through a ringing key to the service
peg. It is also teed to a night control board and to a switch-spring upon
the chief operator's desk. The operator's telephone is permanently bridged
across the call wire, so that she is always listening and ready to receive
demands for connections. The centre of the receiver is earthed and one
side is connected to a flexible cord and thimble, which is usually fitted upon
the second finger of the right hand. This ebonite thimble carries a small
brass stud, with which the testing is done.
The service peg is used for speaking to subscribers, or for other special
purposes when necessary. In the normal state of afifairs it is not required.
Night Anatigenients.
177
The operator is unable to disconnect from the call wire save by the en-
tire removal of her apparatus, and thus private conversations with sub-
scribers are practically impossible, since they may be overheard by anyone
coming upon the call wire or at the chief operator's desk.
The call wires are also teed to a night control switch so that at night
calls may be quickly attended to. Normally the call wire is disconnected, and
it is only when a subscriber places his instrument across it that it is con-
nected. At this board a battery and indicator (as described in the case of
Post OfiSce up call wires) is connected to each call wire when the divided
lever is moved over to the left (Figure 139). Immediately a subscriber
depresses his call key this indicator falls. The night operator then moves
the lever downwards (Over to the right in the figure), thus connecting his
telephone to the call wire. Having ascertained the numbers of the calling
and called subscribers, the connection is made by a special pair of
t*-t
Figure 139.
cords upon the nearest multiple. Of course, when the day operators leave
they remove their telephones, which would otherwise bridge across the call
wire and actuate the night indicators.
Having considered the electrical details of the system, we may now
pass to the construction of the system. To economically work a. town
with the call wire system, areas must be most carefully considered. Lead-
covered paper cables containing either 153 or 204 metallic circuits are laid
down to the various areas. These cables pass to standards where they are
distributed to the various subscribers. The smaller the area the shorter
is the length of wire required for the call wires. Since the call wires have
to be arranged from geographical considerations — i.e., the distribution of
the subscriber's lines — the subscribers upon any call wire are not arranged
numerically; for instance, No. 3004 may be next door to No. 13. These
two subscribers would, of course, be upon the same call wire.
N
178 Advantages and Disadvantages. — Cross-Connection Fields.
At the exchange the cables are cross-connected, so that the subscribers'
lines appear upon the test board in numerical order ; thence they pass to
the tags which are connected to the multiples round the room. These
must clearly be in numerical order. Now the position of the pegs is that
of the call wires, since the subscribers' pegs must all be within the reach
of the operator dealing with the corresponding call wires. The wires
have then again to be cross-connected to the cord positions. This board
is termed the intermediate cross-connection field.
The advantages of this system may readily be summarised : We have
the advantages and disadvantages of the flat board. The fact that in a
breakdown of a call wire fifty subscribers are stopped, and are quite
powerless until the call wire is repaired. There is also the possibility of
the call wire being stopped by two subscribers disputing. Again, office
boys can cause endless trouble by singing, whistling, or shouting upon
the call wire, and since it may emanate from any of the fifty subscribers
it is difficult to locate. Another point is that by listening upon one's call
wire one might ascertain with whom every subscriber upon that call wire
does his business. If two subscribers in precisely the same business are
upon the same call wire much mischief may accrue from this source. The
fact that the subscribers ring their own correspondents is a notable
advantage.
In regard to the switching, the main objection is the fact that the called
subscriber cannot get disconnected without calling up the calling sub-
scriber. This, however, is a trouble which may readily be overcome by
the use of a device which would drop a ring-off indicator upon either
subscriber depressing a key.
The Telephone System of the Bntisk Post Office. 179
CHAPTER XXX.
The Post Office Multiple System.
At Newcastle it was early necessary to provide a multiple switch, and to
still retain the permanent current system. The advantages of this latter
system are that an automatic ring-off signal is obtained from the sub-
scriber, and, further, ti)at the operator can always tell by glancing at any
indicator exactly what is happening upon the circuit to which that indicator
is attached.
The Newcastle system presents some notable advantages over any other
system which has yet been propounded, inasmuch as it combines the
advantages of the parallel multiple without its disadvantages. There is an
arrangement by which, when two subscribers are conversing, the accidental
connection of a third entirely short-circuits all three, thus stopping, and
at the same time advising, the operator. This is the most notable
advantage. The system is also a secret one.
At the subscriber's office a telephone having two receivers and switch-
arms is used. This is of the universal type previously described, but has
two extra contacts upon the right hand switch-arm. In the normal state
of affairs a permanent current flows through the telephone relay to the
exchange, there deflecting the needle of the non-polarised indicator relay
permanently connected to the circuit over to the right, thus signifying
that the circuit is disengaged. In order to call the exchange the left hand
receiver is raised from the switch-arm. This stops the permanent current,
and causes the needle at the exchange to fall to the vertical, which, iii
this system, indicates a call. At the same time the receiver and secondary
of the induction coil are joined to the line, and thus when the operator
speaks the subscriber hears. Should the operator not answer immediately,
the subscriber can vibrate the needle at the exchange by alternately
raising and lowering the left switch-arm. Upon receipt of a reply from
the operator the right hand receiver is removed and the demand is then
given to the operator. The raising of this lever not only joins up the
primary of the subscriber's telephone, but also reverses the direction of
the permanent current which deflects the needle at the exchange to the
left, thus indicating engaged. We have then three distinct indications
for normal or disengaged, engaged, and calling. When a subscriber is
through to another subscriber both their indicators are deflected to the
left, showing engaged. When they have concluded their conversation
they replace their receivers upon the switch-arms and thus reverse their
M 2
iSo
Arrangement of Multiple.
permanent currents, which again flow in the normal direction, deflecting
the indicators to the right. This is the operator's instruction to disconnect,
which is done without remark.
The subscriber's lines pass along the run of the multiple boards and are
teed off to the respective switch-springs in each multiple as also to the
nome indicator (Figure 140). En passant, it may be remarked that it has
been found unnecessary to duplicate the home switch-springs upon the
multiple. The switch-springs consist of four springs, of which the two
inner ones are teed to the line and come into contact with the two sides
of the peg, which is exactly similar to the peg described upon page 59
but for the fact that the end near the handle is rounded so as to fit into
the circular opening of the switch-spring. The two outer or short-circuit
springs project beyond the line springs and in the normal condition are
connected together by means of the metal contact pin which passes
through holes cut into the line springs but without touching them.
SuBScniBKA^y^.
Figure 140.
In the case of trunk line operating in the Post Office multiple system
the pairs of pegs and cords and apparatus by means of which the con-
nections are made are of a very complicated nature. It is here necessary
to distinguish between the two pegs of each pair. The answering peg is,
as previously stated, of very similar construction to that described upon
page 59. The other or calling peg is illustrated in Figure 140, and is of
the same shape and size as the answering peg, but the centre of this
peg contains two pieces of brass insulated by ebonite. The ends of
this peg Li and L3 come into contact with the inner or line springs, thusi
picking up the A and B lines of the subscriber. The two segments Si and
S2 pick up either side of the short-circuit wire by means of the outer
springs. The object of this arrangement will be appreciated later. The
outer or short-circuit springs are normally connected together by means
of the brass contact pin. The short-circuit wire runs from the A short-
circuit spring to the B short-circuit spring of the following switch-spring
throughout the multiple, as indicated in -Figure 140.
operating Connections. i8i
Each pair of cords has attached to it a speaking key, which, in the
normal condition " through " connects Lj and Si, also I^ and Sg of the
detector peg and thus renders it precisely similar to the answering peg,
which does not possess the separate segments. The top and bottom sides
of the pegs are respectively connected across, and thus the arrangement
is now identical with the simple pair of pegs and cords described on
page 59.
Upon the receipt of a call (needle vertical or vibrating) the operator
first moves ovar the speaking key of the pair of cords which it is proposed
to use for the connection and then inserts the answering peg. In this way
the operator's telephone is connected to the calling subscriber's line.
Having ascertained the number of the required subscriber, the detector or
calling peg is inserted into the wanted subscriber's switch-spring. Now
Li and Lg of the detector peg are connected to a detector which consists
of nothing more or less than a galvanometer of high resistance. If the
subscriber is disengaged, his permanent current will flow through the
detector in such a direction as to deflect it to the right. If engaged it
will be flowing in the opposite direction, and thus n. deflection to the left
will be observed. If the subscriber is calling, the needle will either hang
vertically or will vibrate ; thus it will be seen that by the insertion of this
detector peg the operator can tell at a glance precisely what is happening
upon any circuit. Now this is the engaged test. Should the required
subscriber be engaged, the distant operator is informed and the answering
peg is withdrawn. The detector peg is, however, allowed to remain, so
that the operator may be made aware of the conclusion of the conversation
by the reversal of the permanent currents.
If, however, the wanted subscriber be not engaged — i.e., if the detector
is deflected to the right — the operator momentarily depresses a key attached
to the breast-plate transmitter. This energises the electro-magnets of the
" combined call and detector switch " and the armature is attracted, thus
making a series of fresh connections. Firstly, the detector is cut off, and
a calling battery of sixteen cells is connected up to the wanted subscriber's
line, thus ringing his bell. This calling battery does not cut off the
operator's speaking set, and consequently upon the subscriber's response
it only remains for the operator to go out of circuit and thus put the
subscribers through. This is accomplished by the simple act of turning
back the speaking key to the normal. By the deflection of the calling
subscriber's indicators to the left the operator is automatically informed
that the subscribers are through. The two subscribers' batteries are now
both sending reversed permanent currents, and thus both their indicators
are deflected to the left, and the insertion of a detector peg at any other
multiple in either of the subscribers' lines will give a deflection to the left
(engaged) upon the detector. Upon the conclusion of the conversation
the subscribers replace their repeivers, and this r«vers^ the permanent
i82 Combined Call and Detector Switch. — Triple Switching.
current, deflecting the indicators attached to those subscribers to the
right. The operator who made the connection observes this, and removes
the pegs.
It only now remains to be explained how triple switching is prevented.
The insertion of an answering peg connects the long springs to the line
springs, as indicated by the dotted lines in Figure 141. At the same time
the contact between the two springs is broken by their being pushed apart
by the insertion of the peg. If now two pegs are inserted into the switch-
springs of the same subscriber upon two switch sections, the conditions of
Figure 141 obtain — i.e., the line is short-circuited. The indicator imme-
diately falls to the vertical, and thus the operator is made aware of the
mistake. Now this is the reason for the different construction of the
detector peg. It is necessary to bridge the detector across the lines with-
out disturbing a speaking subscriber. If now a detector were attached to
an answering peg the lines would be short-circuited by this device. In
order to avoid this the detector peg is made in four sections. When
detecting Sj and Sg are disconnected, but immediately the detector is cut
A
• >
1
^
/^
— s-
A
]
c
ri^ —
.V
7\
<
1
>
Figure 141.
out of circuit by depression of the call key, Si and Sj of the detector peg
are joined together, so that the insertion of a peg in the through position
will at once short-circuit the subscribers, and the operator will at once
observe the error. It is scarcely necessary to add that Si and Sj are
connected in the through position. It will now be seen that, by causing
the short-circ«iit springs to project beyond the line springs, operators are
prevented from tapping circuits, as thereby the speaking subscribers and
the operator would be short-circuited.
The combined call and detector switch is so arranged that after the first
depression of the transmitter call key it is independent of it, the separate
calling battery current flowing through its coils, and thus keeping them
energised ; in fact, the whole arrangement is bridged across the lines.
This prevents the operators calling by means of a series of short rings.
In order that the introduction of the operator's telephone may not cause
false signals upon the circuit of a oalling subscriber, a battery is inserted
jn series with it, thus Gauging it to exactly resemble an engaged subscriber.
Trunk Line Working. 183
This battery also serves the double purpose of working the call switch
upon the depression of the transmitter call key.
The system of trunk line working is very similar. The switch-springs at
either end of a trunk line are arranged so that in the normal condition
opposing permanent currents are sent to line through a relay. The inser-
tion of a peg reverses the direction of the current. An indicator relay is
permanently bridged across the lines, but its circuit passes through the
local contacts of the first relay (Figure 142). In the case shewn the trunk
is disengaged. The indicators are deflected to the right by reason of the
derived circuit across the lines. If now a peg be inserted at A the battery
there is reversed. The two batteries now combine together, and the local
circuit of the polarised lalay at B is broken, and thus the indicator needle
hangs vertically, indicating a call. At B the current in the relay is in the
wrong direction to break the local circuit, and this needle goes over to the
left. Upon the insertion of a peg at the distant end in order to reply that
battery is reversed also, and the tongue of the relay goes back to its stop,
thus again joining up the indicator. It should have been mentioned that
the relay is biassed against the small current which normally flows through
A p
'*iOtC»TOf
L\
^^^^ NoHMftL
t3^ Wi B
T-]
r
T
Figure 142.
It, and it is only upon its augmentation by the distant battery that the
bias is overcome. Both the indicators are now deflected to the left,
indicating engaged.
It is scarcely necessary to point out that the battery upon the operator's
telephone is so arranged as to connect the same poles to the lines as the
trunk automatic calling batteries, in order to avoid derived circuits through
it. The subscriber's permanent current batteries are also similarly con-
nected with respect to the trunk batteries.
Upon the conclusion of the conversation the subscribers replace their
receivers, thus reversing their batteries, which causes them to join in series
with the trunk battery at their own end. This current overcomes the bias
of the relay, and the indicator falls to the vertical, thus advising the com-
pletion of the conversation. Precisely the same thing occurs at the
distant end. In any case the removal of the peg at the controlling office
brings about the same result.
In working with the Company's local system, batteries are inserted
across the junction lines to make the system resemble the above as closely
184 Trunk Line Wovhing.
as may be. A generator ringing key is added to each junction, and the
connections are obtained by means of a down call wire. If a subscriber
tA the Company rings off, the local circuit of the junction indicator is
broken, and the needle goes over to the right. This is accomplished by
means of a weighted bias.
Connections from trunk to trunk are made by means of special switch to
switch junctions. A special circuit to each switch section is fixed upon
every section. To transfer a connection to section five say, the operator
inserts a peg into number five section call wire. This deflects the indicator
attached to that circuit to the left, and the operator answers and names
a junction wire for the connection to be extended upon. Upon this junction
there is no apparatus in bridge.
In the case of the local system the operating is slightly different firom the
method described in the case of trunk lines, but since exactly the same
method may be used it was considered well to give it prominence. Instead
of the pairs of cords being joined up with the operator's telephone and com-
bined call and detector switch, they consist merely of pairs of plain two-section
pegs and cords. The operating peg is quite separate, and the modus operandi
is as follows : — Subscriber calls up ; operator answers by inserting her peg ;
ascertains requirement, and withdraws ; then inserts it into the called sub-
Eciiber's switch-spring, and tests wilh the detector. If disengaged, she presses
the telephone button, thus actuating the call-switch, and so rings up the called
subscriber. Upon receipt of his reply her peg is withdrawn, and the connec-
tion made by means of a pair of plain cords previously described. It will thus
be seen that the functions of the plain cords and the special operating cord
have been combined in the trunk-switch.
The Telephone System of tlte British Post Office. 185
CHAPTER XXXI.
Miscellaneous Special Arrangements.
Manchester Junction Arrangements. — Where the call wire
system is in use at the local exchange some modifications in the method
of working are desirably One of the consequences of the call -wire
system is that the called subscriber is powerless to clear the connection
himself. Hence when a trunk call arrives for a subscriber who is engaged
locally as the called subscriber, some delay arises before the calling
subscriber clears the connection. Now trunk lines are not kept waiting for
this, and therefore an excess of junctions over the trunk lines is required.
Six junctions are allotted to each section in place of the usual five.
In the ordinary system when a subscriber who is engaged locally is
asked for by the Post Office, the Company's operator connects him to the
junction line, but also comes into circuit herself when the subscriber is
asked if he will take the trunk call. If he agrees, a twist of his generator
will sever the connection, and the Company's operator by testing will at
once observe when this has been done. Here it does not matter whether
the subscriber is calling or called subscriber.
In the case of the call wire system each junction terminates in a peg at
the Company's end, and upon a request being made for a certain sub-
scriber to be placed upon a particular junction, that junction peg is pulled
down and inserted into the multiple switch-spring bearing the number of
the wanted subscriber, without testing to see whether he is engaged
or not. Now not only is the subscriber's loop extended to the Post
Office, but also his test-wire. There it passes through a telephone
exchange galvanometer, one of which is allotted to each junction. If a
subscriber is engaged locally the test battery sends a current along the
test-wire through the galvanometer, producing a deflection to the right,
thus advising the Post Office operator. The tip and neck of the junction
peg when inserted into a subscriber's switch-spring extends the loop, and
the body of the peg connects the body of the switch-spring to the third
wire. The junction clearing arrangements are precisely the same as in
the case of the ordinary system. The socket contact upon the junction peg
joins up the junction clearing relay, so that when the peg is removed at
the Post Office the lamp may be lighted.
This system throws upon the Post Office operatgr tbe responsibiUtjr q(
getting connections cleare4.
i85
Trunk Engaged Test at the Post Office.
It will be observed that the third wire passes from the top contacts of the
eight-point switch-spring, through the galvanometer and through the three-
point key (Figure 143) to earth. Where there are a large number of junctions
to be dealt with, the whole of the switching cannot be done upon a single
FiGDRE 143.
multiple board. Hence it is possible to get a subscriber connected to
more than one junction. For instance, a subscriber who is speaking, say,
to London from Manchester would not appreciate having half or one
minute of his three minutes wasted by an explanation with a second
Third wire
22i
-o-
TMIRb "WIRS
L-1 :
Figure 144.
operator, who wishes to tender a threepenny call from say Rochdale. It
was therefore necessary to design an arrangement which would enable an
operator at the Post Office to ascertain whether a wanted subscriber is
engaged upon another junction. This is the object of the three-point key
Vibratory Trunk Engaged Test. 187
shewn In Figure 143. Depression of this key causes the exchange galvano-
meter to he deflected to the left if the subscriber is also connected to
another junction.
The principle of the arrangement is shewn in Figure 144, where one
subscriber is shewn connected to two junctions, i.e., so far as signalling is
concerned. The test-wire of a subscriber who is not engaged locally is
disconnected, and therefore the depression of the trunk engaged test key
does not affect the galvanometer. Now in the case illustrated it will be
seen that the current passes along the first third wire, along the sub-
scriber's test wire, which is connected to earth through the second third
wire to which that subscriber is connected. Thus, when a subscriber is
connected to two or mot^ junctions the test wire is earthed, and, there-
fore, the depression of the trunk engaged test key informs the Post Office
operator, who waits till the conversation is finished.
Now the Company's engaged test battery is of the same voltage as the
trunk engaged test battery, therefore a subscriber engaged locally only
will merely cause the needle to fall from a deflection to the right to the
vertical position. If engaged on a junction also the needle deflects to the
left.
If a subscriber is asked for (locally) several times, and tests engaged,
it is usual for the operator to enter his circuit by means of the service
peg, and inform him that someone wishes to speak to him. It would be
most objectionable to have conversations upon trunk lines interrupted in
this way.
The test wire of any subscriber connected upon a junction is put
to earth through the telephone exchange galvanometer upon the junction
to which he is connected. The Company upon testing get a click, due to
this earth. Now, in order to render the test distinctive, the primary of
an induction coil is included in the main earth at the Post Office. The
secondary is joined up with a vibrating sounder and battery, so that an
alternating current of sufficient frequency to produce a musical " hum "
is always included in the Post Office earth ; thus any of the Company's
operators testing a subscriber who is connected to a junction line hear a
musical note, which is their instruction not to enter the circuit as the
subscriber is speaking upon a trunk line. If the subscriber is engaged
locally then a click merely is heard. The arrangement is indicated in
Figure 144.
Glasgow Junction Arrangements. — At Glasgow the Company's system
is a mixed one, and in addition to this it is operated upon the call wire
system. Here some very awkward arrangements ■ have to be made. Every
junction is furnished with three pegs. The first takes the junction loop direct
for metallic subscribers (distinguished by a white circle round the switch-spring
upon the multiple), the second includes a transformer for single wire sub-
1 88 Glasgow Jimction Arrangements.
scribers, and the third peg is connected to the third wire of the junction and
terminates at the Post Office in an exchange galvanometer.
The procedure is as follows : — The Post Office ask for a number to be
placed upon a particular junction. First, the special peg, which is short,
connects the subscriber's test wire through the Post Office exchange galvano-
meter, in circuit with which is placed a small battery. If engaged locally the
subscriber's test wire is to earth, and a deflection is observed upon the gal-
vanometer. If no deflection is observed, the Post Office operator asks for the
subscriber's line to be joined through to the junction. If a deflection is
observed, the short peg is left in the switch-spring until the Company's
operator is told upon the call wire that the subscriber is clear.
This cumbrous arrangement has been found to be imperative as the
simpler system in use at Manchester cannot be employed on account of the
form of a switch-spring used by the Company. This system may be regarded
as a tentative one, pending alterations which would render the simpler system
applicable.
Differential Self-Restoring Indicators. — In the case of A and B
sections, in order to guard against the possibility of failure of the service cir-
cuit, one or more of the junction lines are equipped with self-restoring indi-
cators ; indeed, this is why the three self-restoring indicators are provided.
It has, however, been pointed out that the connection of the indicator between
the two inner springs of the switch-springs prevents the use of automatic
clearing, and it b to avoid this difficulty that the differential indicator is pro-
vided. The centre of the line coil is connected to the junction clearing
battery, and the inner springs of the junction switch-spring are connected to
the ends of the line coil in the ordinary way. Since the indicator is differential,
the clearing current splitting equally through the two half coils produces no
effect upon it.
The Indicator is wound to a much lower resistance than is usual, each half
coll having a resistance of sow only.
Three-Contact Generators. — In order to avoid the complications
shewn in Figure 8z, a three-contact generator has been designed. Instead
of the power or hand generator being connected up to all the sections,
and exchangeable by means of a single two-way switch, the hand generator
itself makes the requisite changes. In the normal position of the three-contact
generator the power generator is joined to the particular section to which it is
attached. Upon revolving the hand generator the power generator is cut off
and the un-short-circuited hand generator substituted. This avoids all switches
and makes «ach section complete in itself.
Night ^Arrangements. 1 89
Night Arrangements. — The majority of the smaller Post Offices close
at 8 p.m., and it becomes necessary to provide a night service, and at the
same time retain the control of the circuits. At 8 p.m. all the smaller Post
Offices which are not open all night are put through upon one or two main
trunk lines to the Company's exchange. The trunk line is then extended to
the larger Post Office, which is open all night. For instance, Barrow-in-
Furness puts the Blackburn-Barrow trunk through to the Company by means
of a junction line, which at the Company's end is fitted with an indicator.
At Blackburn, this trunk is extended to Manchester by means of a Blackburn-
Manchester trunk. A peg is inserted at Manchester so as to remove the
permanent currents, thus working the circuit as an open one, and ringing by
generator. Similarly a ^uple of Manchester-Blackburn trunks would be
extended to the Company's exchange at Blackburn. Now one end of these
lines is under the control of a Post Office open all night. If a Blackburn,
subscriber wished to speak to a Barrow subscriber, the Blackburn National
would ring up Manchester Post Office, who would ring Barrow National.
This seems somewhat roundabout, but it achieves the required object —
viz., dispenses with night attendance at the small Post Offices, and yet
leaves the complete control of the night traffic in the hands of the Post
Office.
It will be seen that at night there are a very large number of trunk lines
which are not in use. For instance, three or four of the twenty-seven circuits
between Manchester and Liverpool would amply suffice to carry the traffic
even at the busiest periods. Now in order to reduce the space occupied by
the working trunks, a switch is arranged which throws all the working trunk
lines on to, say, the first ten sections. A large removable label is provided for
indicating the alterations which have been made.
At present the switch used for concentrating the trunks at night merely
consists of a trunk test tablet, the test holes and U links being used as a
switch. In course of time it may be found worth while to concentrate the
trunks upon a. special board accommodating from thirty to fifty lines, and in
this event it is probable that a simpler switching arrangement could readily be
designed.
Intermediate Stations upon Trunk Lines. — It is occasionally
advantageous to place an intermediate station upon a trunk circuit. This
can only be satisfactorily done where the traffic of the intermediate office
is small.
The principle of the method consists in dividing the trunk circuit into
two halves at the intermediate office and connecting them together by
means of a pair of cords. A condenser is included in the ring-off circuit
at the intermediate office, so that alternating currents only affect the
apparatus. Thus the primary battery signalling between the terminal
offices does not affect the intermediate one. To call this office, the terminal
tgd tntermediate Stations upon Tfunk Lines.
oflSces rings with the red button. The disengaged side is cut off by removing
the peg, thus connecting up a polarised indicator relay and permanent
currents. The circuit is then entered by means of the speaking key. This
virtually divides the circuit into two. The disengaged side is joined up to the
permanent current relay, and thus the terminal office is not disturbed. Any
call would be received upon the relay and would then be answered by the
intermediate office pegging into the second switch-spring. Now when the
intermediate office connects a subscriber the condenser is short-circuited by
inserting a solid brass plug in the special switch-spring. Thus when the
Figure 145.
terminal office removes the peg the exchange galvanometer is deflected. In
order to receive calls a magneto bell is joined in parallel with the self-restoring
indicator and exchange galvanometer.
If the intermediate office wishes to enter the circuit to speak to one of the
terminal offices it is necessary to first ascertain whether the circuit is disengaged.
By inserting the brass peg the exchange galvanometer is directly joined up.
If the circuit is disengaged the permanent currents deflect the galvanometer.
If not, no deflection is observed, and in the latter case the peg is left in, and
the circuit is entered upon the conclusion of the conversation, when the
permanent currents are restored.
The whole arrangement is indicated in Figure 14.1;, where the circuit is
Single Wire Trunks. 191
shewn in the normal condition. Obviously, if the exchange galvanometer
were left directly in the circuit, both the terminal circuit indicators would
be actuated ; hence the condenser, which, whilst it stops direct, permits
alternating currents to pass.
Single Wire Trunk Lines. — At the transfer of the trunk lines to the
Post Office a few single wire trunk lines existed, and it was therefore necessary
to slightly modify the present system of working in order to meet this case.
As it is the intention to double all these circuits as early as possible, a very
brief mention of the arrangement adopted will suffice.
In the normal state of affairs the circuit is joined up precisely the same as
an ordinary trunk circuit, ^ve that the B line is to earth and that the main
permanent current battery is increased to nine cells, so that the distant local
circuit of the relay is closed upon the insertion of a peg at the home end.
In this respect it is independent of the trunk ringing battery. Upon the
insertion of a peg a transformer is joined up, through which the speaking
takes place.
Supervisor's Switch Section. — Experience has shewn the desirability
of providing a means of communication with the supervisor of the exchange,
and also a means by which the working of any section might be checked,
and therefore a switch for the purpose has been provided at all the larger
exchanges. It consists of a transfer circuit from each switch section to the
control switch, a series of switch-springs each teed on to the operator's
telephone, and also a set of switches for bunching and throwing the up call
wires or record table circuits, as the case may be, on to special indicators
placed upon the local switch. The transfer circuits are placed upon one of
the vacant switch-springs and visuals upon the sections, and are marked
"supervisor." At the control switch the switch-springs and visuals (which
are in strips of twenty) are marked with the number of the switch section to
which they correspond.
The receiver springs of each switch telephone connector are taken to a
two-point switch-spring in every way similar to those used for the local switch
junctions. Thus every operator's speaking circuit is permanently connected
to the control switch, so that by inserting a peg connected to the speaking
set it is possible to listen to and check the working of any section in the
room. The instructions or other remarks addressed to the operator, as well
as the operator's own conversation, may be listened to at this control board,
thus rendering it possible for an inexperienced operator to receive assistance
from the superintendent, which is a most valuable aid.
At the top of the supervisor's switch-section is placed a row of ten plug
keys. These are of exactly the same pattern as the plug keys shewn on page
68. The up call wires are each teed on to the long springs of the plug keys,
the top and bottom inner springs of which are respectively teed together to a
192 Supervisor's Switch Section.
single circuit going to the local switch. In the normal condition, when all the
plug keys are in, the teed connections from the call wires to the supervisor's
switch-section are disconnected. When all the plug keys are pulled out the
whole of the up call wires are bunched on to the special circuit going to the
local switch. This circuit passes on to a plug key having six springs, and
being, as regards contacts, exactly similar to a ringing key. In the normal
position the bunched call wires are connected to a battery and indicator
arranged as described in the case of the Company's system (page 177) — i.e.,
immediately a call wire is connected, by an operator going upon it, the
indicator drops. When the key is pulled forward the call wires are thrown on
to the local switch operator's telephone.
At the supervisor's switch section the plug keys are provided with a brass
bar and screw which will clamp the plug keys either in or out, thus preventing
the disconnection of a call wire by an accidental touch.
Special Keyboard. — Experience has shewn that the provision of separate
ringing keys for each pair of cords is not absolutely necessary, and it is in con-
templation to reduce the number. As it is necessary to ring with the generator
upon transformer circuits (the outsides of superimposed circuits), three ringing
keys are necessary — viz. , two keys (a black and a red) for the black pegs and
a red one for the red pegs. This set of three keys would replace the present
twelve. The connections would have to be re-arranged and additional con-
tacts added to the speaking keys — in fact, the connections would be very
similar to those employed by the Company (page 162). The speaking key
would have to pass on both the lines from both pegs on to the common ringing
keys, whence they would pass to the operator's telephone.
At the same time it has been proposed to put in two common keys through
which the cords would pass after leaving the depressed speaking key. The
object of this is to enable an operator to cut off either cord, whilst speaking
upon the other, without removing the pegs. As the system is an automatic
one the removal of pegs gives clearing signals, and this causes delay when it
becomes necessary to ask the originating subscriber if, say, he will have
his conversation extended to six minutes. The arrangement is being tried
experimentally, as experience only can shew whether the improvement is of
practical value.
Cord Testing. — In any telephone exchange the weakest point is un-
questionably the cords. These are usually of one of two types, the one
having the two conductors lightly covered and wrapped spirally vidth wire
covered by a spiral braiding ; the other cord does not contain any wire, being
packed with cotton.
It is in contemplation to replace the present form of brass pegs by
aluminium, and at the s-ime time to use a cord of about half the diameter
of the present form. This cord consists of two tinsel conductors suitably
Cord Testing. 193
insulated with cotton and encased in a woven covering. The openings at the
ends of the pegs are considerably narrowed, and the cord can readily be
knotted inside. At the same time the weight which holds the cords taut is
very considerably reduced. However, some time will have to elapse before
it can be said definitely whether this is in reality the improvement it appears.
The testing of cords and pegs is a most important matter. A special piece
of apparatus, designed to detect cord faults before they become sufficiently
pronounced to cause disturbance to the working, has recently been designed.
It consists of four switch-springs, a press key, and two 20 resistance coil to the third point of the peg, and thence through the
body of the switch-spring S,, through the cut-off relay C R to earth, and so back
to the negative pole of the battery. This causes the lamp L^ to glow and also
cuts off all the subscriber's calling apparatus by actuation of the cut-off relay C R.
The clearing lamp L, glows until the receiver is raised from the switch-hook,
MMIii e i U DILI.
EXCHANGE ENW *
OPCRATINO CONNECTIONS.
Figure 146.
when another circuit is formed from the positive pole of the battery through
the coil of the transfomer T„ and cut-off relay R, along the B line through
the subscriber's transmitter and primary of his induction coil, along the A line
and back to the battery through T^. This provides the speaking current ; and
also by actuating R^, places a shunt of 600) across the lamp L^, so darkening
it. Precisely the same remarks apply to R^ and the other side of the trans-
former. The speaking currents induced in the secondaries of the induction
coils pass through the transformer and induce similar currents upon the other
side.
For the sake of simplicity the operator's speakmg apparatus has been
omitted from the operating connections, and the calling and clearing batteries
Kellog System. 203
have been shewn separately for the same reason. In practice only one battery,
consisting of secondary cells of very large size, is employed for the whole
exchange. The speaking key is merely a tap across the lines at the point
marked by crosses, and the ringing keys are interpolated in the usual manner.
It will now be apparent that the calling apparatus is removed from a sub-
scriber's circuit immediately a peg is inserted either in the home or one of his
multiple switch-springs ; also that the lamps L, and L, shew exactly what is
happening upon either side of the cords. It is not until both clearing lamps
glow that the operator disconnects.
Junctions to exchanges where the same system is being operated are worked
in a very similar manner. Each junction terminates in a peg, with a four-coil
transformer interpolate(^ Two lamps are provided, one of which glows when
the distant calling subscriber hangs up his receiver, and the other when the
called subscriber does so. At the first exchange the two lamps in the pair of
cords light in the ordinary way, one for the calling subscriber and one for the
called subscriber.
Junctions to exchanges where a different system is employed are worked by
the adoption of an altogether different plan. Condensers are inserted in either
side of the junction-loop, and the signalling upon the junction side is effected
by means of a self-restoring indicator placed upon the sub-exchange side. To
work with the trunk line system of this country .some such arrangements
may be necessary, but these details will probably not be finally settled until
the system has been in operation some time.
The Kellog System. — This system has for its object the increase of the
capacity of a telephone exchange. When the number of subscribers to an
exchange reaches 10,000, the limit of the multiple is very nearly obtained.
Again, were it possible to construct a multiple of say forty or fifty thousand,
the cost of providing for new subscribers would be a largely increasing quantity
as the exchange grew.
Four multiple boards are provided : the first, or A board, containing a multiple
of subscribers, l to lo,ooo ; the second, or B board, 10,001 to 20,000 ; the C
board, 20,001 to 30,000 ; and the D board, 30,001 to 40,000. Each sub-
scriber has four indicators and four home-jacks, one upon each of the multiple
boards. A, B, C, and D. He is provided with a generator of special con-
struction, and four keys, A, B, C, and D. Depression of the A key sends a
positive current along the A line upon turning the generator. Depression of
the B key sends a negative current along the same line. Similarly C and D
respectively send positive or negative currents along the B line. At the ex-
change each line passes through two polarised indicators, and it will thus be
seen that by depressing the suitable key any one of the four indicators may
be actuated without affecting the others. The insertion of a peg in either
the home or multiple switch-springs cuts out the whole of the indicators.
The subscriber who desires a connection rings up on the board to which his
2o4 Divided Board Sysiems.
correspondent is connected, i.e., if the number were 15.361, he would ring up
on the *' B " board. Each subscriber's calls are therefore divided into tour parts,
which, in the aggregate, may be considered equal parts, and therefore each
operator is able to deal with four times as many subscribers' lines upon each
board. So, instead of 200 lines per multiple section, 800 are connected, and
the same result is obtained with the equivalent of four multiple boards of ten
thousand each, plus the extra home switch-springs and indicators as would be
obtained by the construction of a 40,000-line multiple, were such a thing
possible. It will therefore be seen that, by the employment of this method,
every connection is dealt with by one operator only, and thus the waste of
time inseparable from junction-worked systems is avoided.
The Chicago Express System. — This system, which is a non-multiple or
divided board one, has for its object the increase of the number of subscribers
which may be connected to a single exchange, and also the even distribution
of work amongst the various operators. Every call is, however, dealt with by
three operators, and thus a slight increase in the time required to make a con-
nection by this system as against a multiple system is observed.
The first or " Y " operator, as she is termed, does not answer any calls, and
speaking apparatus is not therefore required. Her sole duty consists in
plugging subsciibers who call up through to the operators who answer the
calls, i.e., we may imagine the "Y" operator as evenly distributing the calls
between two A operators. Upon the removal of the subscriber's telephone
from the hook a lamp lights in front of the Y operator, who immediately con-
nects him to a junction to one of the A operators. This A operator answers
his call and extends him, by means of a junction, to the B operator, who has
charge of the circuit to which he desires to be connected. The B operator is
instructed by call wire as to the connection to be made. These junctions are
all worked automatically, the insertion and removal of the pegs giving the
requisite signals. The A operator has complete control of the connections and
has two lamps in her cords, one for the calling and one for the called sub-
scriber, and it is not till both glow that the connection is severed. The
removal of the pegs from the junction to the Y and B operators by actuality
the corresponding signals, advise the conclusion of the conversation.
Sabin's Divided Board System.— The distinctive feature of this system
is that the junctions between the switches are multipled in such a way as to
greatly reduce their number.
The operators are divided into two classes, the "A" and the "B" operators,
and are seated side by side. The A operators answer all calls from the sub-
scriber's lines to which they attend, whilst the B operators only connect
incoming junction calls to the subscribers' lines in iheir charge.
The A operators have junctions to each B section in the room, but the
junctions are multipled to the A operators, i.e.. No. i junction to No. 10 B
operator appears on every A operator's section. Since the B operator selects
General Improvements. 205
the junction it is not possible for an A operator to peg on to an already engaged
junction.
The procedure is as follows : — The subscriber lights his calling lamp by the
removal of his receiver. The A operator extinguishes this by pegging in.
Having ascertained the number of the wanted subscriber, the operator goes on
to the call wire to the B section upon which the subscriber's line is connected,
and asks for the wanted subscriber. The B operator names the junction, upon
which she connects him, and by depressing her ringing key calls him up.
Until he answers, a white lamp glows. The connection is now complete.
Upon conclusion of the conversation the subscribers replace their receivers.
This act lights the A operator's lamp in circuit with the connecting cord and
also a red lamp in circuit frith the B operator's cord. The pegs are all with-
drawn, and all signals and connections return to the normal.
It may perhaps be well to explain that upon the board directly in front of
an A operator appear all the multipled junctions, together with connecting pegs
and cords and call-wires. Directly in front of the B are the subscribers'
switch-springs and indicators, together with the ends of all the multipled
junctions, &c., which terminate in pegs. Thus both the A and B operators
have access to the particular set of subscribers' lines to which they attend — A
to answer the subscribers' calls and pass them forward to the required B
operator, and the B operator to answer and connect all incoming calls over the
office junctions to the subscribers in her charge.
Recent Minor Improvements. — Engaged subscribers. — In order to avoid
the necessity of an operator telling calling subscribers that their correspondent
is " engaged," a switch-spring connected to a telephone and phonograph is
employed. The phonograph incessantly repeats " Line engaged ; call again " :
and instead of the operator telling a subscriber that his corresppndent is engaged,
his line is plugged on to the phonograph for a few seconds. The relief thus
afforded is not inconsiderable.
Automatic ringing. — Instead of calling by a series of rings given by the
repeated depression of a ringing-key, upon depressing the ringing-key the
subscriber's bell rings intermittently for certain definite periods, until the circuit
is broken by the removal of the subscriber's telephone from the hook.
Equal Distribution of Work. — An attempt to secure this object has
been made by providing junctions from each operator to a neighbouring one.
Any operator having more than two calls on hand, pegs the remainder on to
one of these relief junctions. The operator on the section at the end of the
junction answers the call, or pegs it on to another operator until it finds its
level. It is well known that the capacity of an operator's hands largely
exceeds that of her voice, and therefore the passing-on oi calls in excess of her
ability to deal with along automatically-worked junctions presents no difficulty.
By thus distributing the work each operator is able to attend to a large
number of subscribers, and thus the number of multiples necessary is reduced.
2o6 The Telephone System of the British Post Office.
APPENDIX A.
The KR Law.
If a battery be connected to a circuit possessing resistance and capacity,
a certain time elapses before ttie current attains its maximum value. A
telegraphic or telephonic circuit possesses both resistance and capacity,
and it is not possible for a current to attain its maximum value in an
infinitely short time. In telegraphy this means that the speed of signalling
cannot be indefinitely increased. Lord Kelvin (then Sir William Thomson)
investigated the question of the propagation of electric currents along
wires possessing distributed capacity and resistance, and as a result the
speed of signalling upon any cable can be calculated from the dimensions,
or, what is more to the point, the dimensions can be determined which
will permit a given speed of working. This law applies equally to tele-
phonic circuits, but it has to be stated in a far more general way, owing to
the fact that the problem of speech transmission is too complicated, and
requires so many conditions that a complete mathematical statement of
the conditions and a solution has so far been impossible.
Kelvin found that the time taken for a current to arrive at a certain
fraction of its maximum value was inversely proportional to the product
of the total resistance and capacity of the circuit. The limiting values at
which speech was possible were determined experimentally by Preece. The
results obtained, after a large number of experiments, were as follows : —
Conductor.
Speech possible up
to KR (m.f. X
ohms).
Speech commerciaily
practicable up to
KR.
Copper (open wire)
Copper (cables and underground)
Iron (open wire)
15,000
12,000
10,000
10,000
8,000
5,000
This rule is known as the KR law, and is one about which much dis-
parity of opinion exists. These values were determined several years ago,
before the granular transmitters had attained their present perfection, and
The KR Law.
207
there is little doubt that the values given on the previous page will not
hold strictly to-day.
The lower value for iron wire is due to the fact that iron is a magnetic
substance. The currents flowing along it magnetise it and thus increase
its self-inductance, thereby adding a certain amount of retardation (which
is absent in the case of copper). The difference in the KR between open
and covered conductors is largely a matter of insulation resistance. In
the case of open wires the insulation is lower, and the wire has more
points for discharging. It is found in practice that the insulation of a
telephone circuit may fall to one quarter of its conductor resistance
before speech is affected. In this condition the circuit is at its best
working value. The lo\*er the insulation of a circuit the more rapidly
will a current attain its maximum value ; but it has to be recollected that
this will be of little service if the current is thereby rendered so small as
to be unable to produce its effects. The working power upon a telephone
circuit cannot be increased by any known means. The design of a
telephone relay has hitherto proved impracticable.
A single wire circuit with an earth return has theoretically the same
working value as a metallic circuit of the same length and material. This
is due to the fact that the capacity of the metallic circuit is only half
that of the single wire. The capacities of the two wires of the loop are
in cascade and therefore jointly are half either of them. The resistance
of the metallic circuit is, however, double that of the single wire and
therefore the product KR is the same. In practice the capacity is -6
instead of half the single wire value.
The effect of capacity or self-induction and resistance or all three is to
distort speech through a telephone. The distortion is due to the fact that
the length of time that a current takes to arrive at its maximum strength
depends not only upon the values of the resistance, capacity, and self-
inductance, or, as the combination of all three is frequently termed, the
reactance, but also upon the frequency of the currents. A current which
changes in direction 1,000 times per second will be retarded more than
one which reverses 500 times. Now articulate speech gives rise to a series
of most complicated waves of current. If any particular wave be taken
and carefully analysed, it will be found that it is possible to split it up
into a number of waves of the simple sine type. The fundamental heavy
wave is the wave which gives the sound its pitch. Allied with it there
are a large number of smaller waves in octaves and in harmony above
and below the fundamental, and perhaps arranged in a still more com-
plicated fashion. These smaller waves, or overtones as they are termed,
are of much greater frequency, and are therefore retarded more than the
fundamental — i.e., the overtones are displaced with respect to the funda-
mentals, — and when this displacement exceeds a certain amount speech
becomes impossible owing to the speech losing its timbre. Thus it will
2o8 Mutual Induction.
be seen that pitch and volume of sound accurately specified includes
timbre. It is to Helmholtz that much of our modern theory of sound owes
its origin and its development.
A controversy has arisen in reference to the possibility of mutual
induction between the wires constituting a metallic circuit. On the one
hand it is stated that this can occur, and that, by bringing the wires closer
together, a point may be reached where the beneficial effects of the currents
flowing in one wire, and thus inducing currents of opposite sense in the other
wire, thereby assisting the rise of the current, may be sufficient to satisfy
the circuit's capacity and thus produce a distortionless circuit. Upon the
other hand this action, if it exist, is termed diminution of the self-induct-
ance, and the principle stated is the one upon which the most inefficient
circuit would be designed. Experiments will probably settle this matter
conclusively ere long.
The Tdej^hone System of the British Post O'flice. 26^
APPENDIX B.
CAPACITY AND RESISTANCE OF LINE WIRES. (A. EDEN.)
Copper Wire.
Gauge.
Resistance
per mile.
Capacity to
earth.
Capacity
wire to wire.
100 lbs.
8a82»>
•0144 m.f.
'O0864 m.f.
150 ..
■0147 ..
•00882 „
200 „
4-391-
•015 ..
•009
300 „
2-9280)
•0153 ..
-00918 „
400 ,,
2-I95-
■0156 „
-00936 „
600 „
1 ■464M
•0158 „
•00948 „
800 „
1-098"
•016 „
•0096
Note. — It will be seen that the capacity to earth is i-6 times the
capacity measured wire to wire, or in other words a metallic circuit has
■6 of the capacity of a single wire of the same loop. Six per cent, should
be taken off these values for the winter. The table assumes average
distribution of trees, and twelve wires upon the poles at an average height
of thirty feet above the ground. Twenty-eight per cent, should be taken
off if only one circuit upon the poles.
Wires placed underground, No. 7 prepared G.P. wire, averages "3 m.f.
per mile, whilst paper cable averages about '08 only.
Hard Drawn Copper Wire.
Weight per mile (lbs.).
Approxi-
mate
Standard
Wire
Gauge.
Diameter (inches).
Resistance in
Standard
Ohms per
mile at 60° F.
of Standard
size.
Minimum
breaking
stress (lbs.).
Stand.
Max.
Min.
Stand.
Max.
Min.
100
150
aoo
300
400
600
800
205
307i
410
615
820
I46J
195
292J
390
780
14
13
1
6
4i
•079
-097
•112
•137
•158
•194
•224
■080
■098
•11325
•13875
■16025
■196
■226
•078
•0955
■II05
■13525
•1555
•191
■2205
8782
5855
4391
2-928
2195
1-464
i^ogS
330
490
650
950
1.250
1,800
2,400
To convert B.A ohms to standard ohms : — Multiply by -9866.
To convert standard ohms to B.A. ohms: — Multiply by i^oi36.
210 The Telephone Sysiem ofihe British PostOffiee.
APPENDIX C.
CROSS-CONNECTION STRIPS.
Signalling Side.
Tag.
A Section.
B Section.
C Section
I
2
3
4
5
6
7
8
9
lO
II
12
13
14
15
I Local P.C. I
2
1
1
1 Main P.C. I
} .. 2 ..
].
I-ocalP.C. I
2
3
4
3
Main P.C. i
2
3
4
5
Local P.C. I
2
3
4
5
Main P.C. i
2
3
i6
17
1
1
i8
19
1
1
4
20
21
1
•> 5
Transfer Circuit 1
22
23
dis.
Transfer Circuit 2
24
25
dis.
Transfer Circuit 3
2b
27
dis.
Transfer Circuit 4
28
29
dis.
Transfer Circuit 5
3°
31
dis.
Transfer Circuit 6
32
33
dis.
Transfer Circuit 7
34
35
dis.
Transfer Circuit 8
36
dis.
Cross-Connection Strips.
211
Signalling Sidb. — continued.
Tag.
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
A Section.
} Service S.R.I. bat'ry
I S.R.Indctor.baPry
L „ local
I P.I. No. 2
} Dis.
P.I.R.B. 1 local+
>• 2 „ „
3 .. ..
4 .. ..
>• 5 ti fi
P.I.R.B. locals -
[ Trunk Ringing
[ Generator
f ■
Transmitter i +
Split
Transmitter 2 -
dis.
B Section.
Junction Clearing
dis.
Service S.R.I, bat'ry
S.R.Indicator bat'ry
local
P.I. No. 2
Transfer battery
P.I.R.B. I local +
II 2 „ ,,
II 3 fi II
4 .. I.
P.I.R.B. locals' -"
Trunk Ringing
Generator
C Section.
Transmitter i +
Split
Transmitter 2 —
dis.
Junction Clearing
dis.
Transfer Circuit 9
dis.
Transfer Circuit 10
dis.
S.R. Indicator battery
local
P.I. No. 2 local
Transfer battery
P.I.R.B. I local +
II 2 „ „
3 .1 II
4 >i >i
5 .. I.
P.I.R.B. locals -
Trunk Ringing
Generator
Transmitter i +
Split
Transmitter 2 —
dis.
Junction Clearing
dis.
In every case where two tags are bracketed the odd number is a positive
?nd the 9ven number a negative lead.
212
Cross-Cottnection Strips,
Spea-king Side.
Tag.
A Section.
B Section.
C Section.
73
74
75
76
I P.O. Subscriber
I
P.O. Subscriber
I
P.O. Subscriber
I
2
>•
2
..
2
77
78
3
■•
3
ft
3
79
80
4
>>
4
1'
4
81
82
5
..
5
tt
5
83
84
85
86
6
,,
6
Transfer Cct. . .
I
7
„
7
2
87
88
8
*f
8
3
89
90
9
»
9
4
91
92
10
"
10
5
93
Transfer Cot.
6
7
94
95
I
2
96
97
,98
99
100
[ Service
I
Service
I
8
} "
2
it
2
9
lOI
102
} "
3
••
3
10
103
N.T. Co. Junction
I
N. T. Co. Junction
I
104
105
106
107
•■
2
ti
2
108
109
)
"
3
»f
3
110
}
It
4
It
4
III
)
i»
5
It
112
}
5
113
f
P.O. Ex. Junction
"4
1
I
"5
1
„
116
r
2
117
!
II
118
j
3
Cross-Connection Strips.
213
Speaking Side. — continued.
119
120
121
122
123
124
125
126
127
128
I2g
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
A Section.
}. Trunk
-Trunk
B Section.
C Section.
P.O. Ex. Junction 4
5
Trunk
, 2
3
4
5
N. T. Co. Call Wire.
P.O. Ex
Transfer „ „
In every case the odd number is the B line and the even number the
A line.
Where 20 transfers have to be provided two extra cross-connection
strips, one for signalling and one for speaking circuits, are added.
Where secondary batteries are employed tags 61 and 63 should be teed
together.
INDEX
PAGE
A and B lines 36
A and B transfer circuits 107
ABC Wheatstone 57
Absolute units S
Accumulators 113, 152
Ader transmitter 25
,, receiver 18
Agglomerate, Leclanch6 40
Ammeter 157
Amplitude of vibration .: 2
A operators 204
Armature, Siemens 45
Arms, 48-inch 13s
Arresters. See lightning protectors.
A switch section 88
Automatic clearing of junctions 94, 188
, , on service circuits 100
, , Stock Ex. circuits 128
cut-out for generator ... 48
signalling 58
switch 157
Auxiliary receiver 194
B
Batteries, bichromate
,, Leclanch^
E.P.S. K7
Battery ringing telephone station
I, speaking
tablet
,, trunk ringing
Bell Blake station
Bell, Prof.
Bell receiver
,, ,, theory of
,, ,, ,, dimensions
Bells, short-circuiting
,, magneto
,, trembler
Berliner transmitter
Bidwell, Shelford
Blake transmitter
B operator
Bourseul
Breastplate transmitter
Bridge, intermediate station
Bridging coils 94
B switch section
159
40
152
36
42
147
90
52
8,9
35
44
3+
25
15
26
204
I
86
56
8, 128
•• 93
PAGE
Call-office circuits 123
Call wire key loi
,, wires on supervisor's switch ... 192
,, wire system 185
,, ,, to local switch 105
Capacity of line wire 209
Carbon, resistance of 15
Cell voltaic ... ' S
Central battery system 200
Charging switches 158
Chemistry of Leclanch^ cell 40
Chicago express system 204
Coils, bridging 94, 118, 128
Collier-Marr receiver 21
Combination key 109
Concentration of trunks at night ... 189
Condensation waves 2
Condenser 95, 106, 190, 200
Connections —
Battery worked station 36
Bell Blake set 49
Call wire system 175
,, ,. ,, night control
switch 177
Central battery system 202
Counter communication switch
for I cabinet 123
Counter communication switch
for s cabinets 126, 127
Generators, hand and power ... 95
Intermediate stations on trunk
lines 190
Junction clearing ... 94
Junctions to N. T. Co 90
Local contacts 91
Local switch (P.O. ) junctions ... 101;
,, ,, standard board ... 161
Local switch standard board
operating 162
Newcastle system 180
P.O. local switch 68
,, ,, ,, operating ... 6g
, , subscribers on trunk switch 89
,, telephone 55
,, ,, for permanent
currents 61
R.C.J, circuit on trunk switch 120
,, switch section 115
Record table switch section ... iig
Index.
215
Connections — co/itd. ^^r,^
Record table tablet 102
Self-restoring board 171
.1 ,, operating ... 172
Series, multiple i68
Service circuit 89
Superimposed circuit 141
Switch telephone connector ... 84
Transfer board, A circuits ... 113
II ,, B ,, Ill
,, circuits, B sections ... 98
ri ,, C , 103
Trunk engaged test 186
,, line permanent current
signalling 72
Trunk line operating"! 82
Universal trunk battery tablet... 158
,, signalling ... 154
Up service circuit multipled ... 99
Cordeaux's insulator 135
Core of induction coil 16
Cord testing 192
Counter communication switch for
r cabinet 124
Counter communication switch for
S cabinets 126
Cross connections 150
Cross connection strips
72, 92, 112, 148, 159
>, ,, ,, appropriation 210
Crossing system for prevention of in
duction
Crossley transmitter
C switch section
Current, effects of
,, methods of producing
133
24
103
4
S
D
Deckert transmitter 30
Detector for Newcastle system 181
Q. and 1 193
Diaphragm 3
silk 87
Direct junction circuits 118
Distribution of work 205
Divided board systems 204
Double pole receiver 20
Du Moncel 9
Dust extraction 174
Dynamic induction 131
Dynamotor 96
Earths, effect of, on short lines ... 135
Eden 209
Energy, chemical and electrical ... 5
,, transformations of 8
Engaged test 168, 172, 176, i8i, i86, 205
Ericsson switch-board telephone ... 56
,, table set 52
,, transmitter 86
Exchange, principle of 57
Faults in switch sections 195
,, key-board and junctions ... 197
,, local circuits 198
,, trunk signalling circuit ... 196
Flat boards 174
Fundamental tone 207
Fuse tablet 153, 156
,, ,, arrangement 157
Galvanometer telephone exchange
75, 128, 185
Generator, magneto 46, 91
,, three-contact 188
Glasgow junction arrangements ... 187
Gouloubilzky receiver 17
Gower-Bell station 53
Gower receiver 20
,, transmitter 25
Granular transmitters ■ 28
H
Headgear receiver 86
Helmholtz 208
Hughes, Prof. 14
,, ,, microphone 14
Hunnings' cone transmitter 30
,, transmitter 28
Indicator, calling drop 161
Polarised No. 2 60
Relay polarised B 64
,, non-polarised B 140
Ring-off tubular 161
Self-restoring 76,99,171
Visual 9^. 104, 109, lis, "9
Induction coil, principle of 13
,, ,, lor P.O. telephone ... ■ 55
,, ,, examples of use
8s, 102, 163, 187, 201
Induction of electric current 12
Inductive disturbances 130
Insulators 13S
Intermediate stations 56
,, ,, on trunk lines 189
2l6
Index.
Jacks. See switch-springs.
Joint microphonic i6
Junctions —
To N. T. Co 70. 93
Automatic clearing of 94
Central battery system 203
Chicago express system 204
Direct circuits 118
Newcastle system 184
Operating of 90
P.O. local switch 105
Sabin's divided board system ... 204
Junction transfer section 122
K
Kellog system ...
Kelvin
Keyboard special
Keys, call-wire ...
Combination
Plug
Ringing
Speaking
KRlaw
203
206
192
... 101
109
... .-.. 68, 116
... 79, 162, 172
... 81, 162, 172
206
Lamps 115, 201, 204
Leclanch^ cell 40
Lightning protectors 144, 159
Lines of force 4
Line wire, resistance and capacity of 209
Listening keys. See speaking keys.
Local circuit 55, 155
,, contacts 92
,, system N. T. Co 160-178
,, ,, Newcastle 184
Local system permanent current ... 57
,, ,, P.O. trunk exchange 67
Lodge O.J 4
M
Magnetic circuit of indi^ction coil ... 16
,, field 4
,, tick X, 9
Magnetism residual iG
Magneto bell 44
,, generator 46, 91
,, worked telephone station 49
Manchester junction arrangements 185
Maxwell 4
Mechanical telephone 3
Mercadier 8
Microphone 14
Microphonic joint 16
Micro-telephone. See switphboafd
telephone.
Mix and Genest transmitter 25
Mixed circuit working ... 163, 170
Mosley transmitter 29
Motor generator 9S
Multiple switchboards 165
Mutual induction 77. 208
N
National Telephone Company's sys-
tems 160-178
Night arrangements for trunk lines 189
,, ,, >, call-wire sys-
tem 177
Non-polarised indicator relay 140
Ohm's law ..,
Overhearing.. 130.
, , between contiguous indi-
cators
Overhearing on operators' telephones 154
;, on record table circuits 117
Overtones 207
P
Packing diffipulty
Pegs, circular
,, ,, 3-point
„ detector
,, double circular
,, flat old form
Permanent current battery-
Counter switch
Trunk local
,, main
System
Universal working
Permanent current working local sys-
tem
Permanent current working, trunk lines 70
Persistent vibration 2
Phelps pony crown receiver 17
Phonograph, use of, in telephony ... 205
Pitch
Plug-key
Plus circuits
Po'arised indicator relay .
P.O. telephone
Primary circuit ... ...
,, coil
Protectors, Lightning..,
143
77
29
67
167
181
85
S9,
124
71
71
61
iSS
57
116, 191
. ... 138
. ... 64
159
36
54.
R
Rarefaction waves
R.C.J, circuits ...
. 144. 159
2
114
Index.
217
R.C.J, circuits, method of working... 120
„ ,, terminated on switch
sections 119
Receivers —
Ader 18
Bell 17
CoUier-Marr 21
Double pole 20
Gouloubitzky 17
Gower 20
Headgear form 86
Phelp's pony crown 17
Swiss administration 17
Watch „ ... 22, 86
Record table^ .' 100
Circuits 113
,, calling at night 117
,, method of working ... 117
Switch section 115
Tablet 102
Reed ringers 95
Reis I
Relay Board System 200
Relays — SSi ^o, 117, 120
Non-polarised indicator B ... 140
N. T. Co.'s on record circuits ... 117
Polarised indicator B 64
Repeater. See transformer.
Residual magnetism 16
Retard coil 117
Ringing, automatic 205
,, circuit, resistance inserted 156
, , , , trunk switch sections 80
,, keys 79, 162, 172
Sabin's divided board system 204
Secondary cells 113, 152
,, coil 12
Secrecy of communication 58
Self-induction 78, 145, 207
Self-restoring indicator ... 76,99,171
,, ,, differential... 188
,, multiple 171,
Series multiple 165
Service circuits 89,99
,, failure of 188
, , , , automatic signalling on 100
Shelford-Bidwell iS
Siemen's armature 45
Silence cabinets 123
Sine waves 207
Single wire circuits, KR of 207
, N.T. Co. 163, 170
,, ,, trunk circuits 191
Sis-block agglomerate cell .„ ... -^2
PACE
Socket contacts 94, 114, 122, 176
Solid back transmitter 32,86
Sound, definition of, &c i
Sounders, vibrating 106
Speaking battery switches 157
Speaking circuit, skeleton of 37
,, keys 81, 162, 172
„ J>lug 68
Speech, principle of transmission of 3
Square, preservation of 133
Static induction 130
Stock Exchange circuits 127
Sub-exchanges 122
Subscribers connected to trunk switches 89
Subscriber's switch, new form 68
Superimposed circuits 137
,, ,, conditions neces-
sary 139
,, ,, testing of 143
Supervisor's switch section 191
Sur excitateur 18
Switch-arms 36, 38, 39
Switch-boards —
Flat form 174
Newcastle system 179
P.O. local new form 64
,, ,, old form 59
Standard for 50 lines 160
Self-restoring multiple 171
Series multiple 165
Switchboard telephone ... 56, 85
tablet ... 63
Switch-springs — 57
2-point 105, 191
3 17s. 200
S .. 73. 83, 89, go, 94, 105, 115,
126, 149, 161, 171
7 IIS
8 ,, 74, 89, 97, 99, 103, III, 119,
126
Old form 59
Switch section —
A 88
^ 93
C 103
Record table 115
R.C.J 121
Switch telephone connector ... 84,102
Symmetrical twist system 133
Table telephones 52, 56
Tags connecting. See cross connec-
tion strips.
Test boards 144, 149
Test cases ... ,., ... 1
+4
2l8
Index.
Test holes
Test-room appliances
Test tablets, 25 circuit (line)
40 circuit (line)
Test tablets, 60 circuit (line)
32 circuit (battery)
Test wire ...168, 172, 176, i
Theory of Bell telephone ...
Theory of microphone
Through wires
Timbre
Transfer board
Transfer circuits, B sections —
C sections
Junction section
Method of working ...
Necessity for
Number required
Principle of working ...
Transformers
, , serious effects
Transformer circuits
Translator keys
Transmitters —
Ader
Berliner
Blake
Breastplate
Crossley
Deckert
Edison
Ericsson
Gower
Granular
Hunnings
Hunnings' cone
Mi.\ and Genest
Mosley
81.
PAGE
... 144
... 144
... 146
... I^
... 149
... 147
185, 205
6
14
144
... 3. 207
107
96
103
122'
Ill
96
104
97
... 137. 163
of ... 199
142
164
33
Transmitters— «aA/. ''*°'^
Solid back 32
Worked by universal battery ... 153
TremblerbeU 34
Triple switching 182
Trunk engaged tests 186, 187
Trunk line permanent current system
principle 70
Tuning fork 2
Twisting of circuits 133
u
U links 141
U links wire 193
Umbrella spring ... . 173
IJp-call wires 191
Universal battery system 152
,, ,, ,, primary bat-
teries ... 158
Universal permanent current working 155
V
Vibrating sounders 106
Visual indicators 96, 104, 109, 115, 119
Volume of sound 2
w
Watch receiver
Wave, definition of
Wires, capacity and resistance of
Y operators ...
Zincs shallow circular
22, 86
2
... 2og
... 204
... 43
MANUAL OF TELEPHONY. By Sir W. M. Preece,
F.R.S., Past President of the Inst, of Electrical Engineers, late Engineer-
in-Chief and Electrician at the General Post Office ; and A. J. Stubbs,
A.I.E.E., Technical Officer, General Post Office. With Illustrations,
Appendix, Tables, and full Index. 15^.
Contents.
I. Transmitters and Receivers.
II. Apparatus and Circuits.
III. Simple Telephone Exchange
Systems,
IV. Multiple Switches.
V. Miscellaneous Switching and
other Systems.
VI. Construction, Wires and Cables.
' The most complete epitome of present-day telephonic practice."
, Electrical Engineer.
THE PRACTICAL TELEPHONE HANDBOOK,
By J. PboLE ( Wh. Sc. ), A. I. E. E. With 288 Illustrations. 5^.
" This essentially practical book is published at an opportune moment."
Electrician.
ELECTRICITY IN ITS APPLICATION TO
TELEGRAPHY. A Practical Handbook covering the Syllabus of the
new Technical Examinations. Adopted by the Post Office Telegraph
Department. By T. E. Herbert, Engineer of Postal Telegraphs.
With 50 Illustrations. Crown 8vo. Cloth. 3^. 6d.
WIRELESS TELEGRAPHY AND HERTZIAN
WAVES. By S. R. Bottone. With 35 Illustrations. Crown 8vo.
Cloth, y.
ELECTRIC INSTRUMENT MAKING FOR AMA-
TEURS. By S. R. Bottonb. With 96 Illustrations. 3^. 6d.
ELECTRIC BELLS. A Practical Book for Practical Men.
By S. R. Bottonb. With 99 Illustrations. 3^.
Catalogue oj Works on Electricity, &'c., post free.
WHITTAKER & CO.,
"White Hart Street, Paternoster Square, London, E.C.
THE ELECTRIC TELEPHOlsrE. By Edwin J.
Houston, Ph.D., and A. E. Kennelly, D.Sc. With 143 Illustrations.
Crown 8vo. Cloth. S^-
ELECTRIC TELEGRAPHY. By Edwin J. Houston,
Ph.D., and A. E. Kennelly, D.Sc. With 163 Illustrations. Crown
8vo. Cloth. 5s.
MAGNETISM AND ELECTRICITY, AN ELE-
MENTARY COURSE OF PRACTICAL EXERCISES IN.
By J. R. AsHWORTH, M.Sc, F.Phys.Soc., Lecturer in Physics, Roch-
dale Municipal Technical School. Crown 8vo. 2s. 6d.
ELECTRICITY & MAGNETISM, FIRST BOOK
OF. By W. Perren Maycock, M.I.E.E. With 107 Illustrations.
is. 6cl.
' ' Students who purchase a copy and carefully study it will obtain an excellent
groundwork of the science." — Electrical Review.
PRACTICAL ELECTRICAL MEASUREMENTS.
An Introductory Course for Students and Engineers. By E. H.^|
Crapper, Lecturer in Physics and Electrical Engineering, Shefl5el#*f
Technical School. With 56 Illustrations. 2s. 6d.
INDUCTION COILS. A Practical Manual for Amateur
Coil-Makers. By G. E. Bonney. With loi Illustrations. 3^.
THE MODERN SAFETY BICYCLE. By H. A.
Garratt, A.M.I.C.E., Head of the Engineering Department of the
Northern Polytechnic Institute, HoUoway. With J04 Illustrations and
five Working Drawings. Crown Svo. Cloth. 3^.
Catalogue of Works on Electricity, Engineering, Sr'c., post free.
WHITTAKER & CO.,
White Hart Street, Paternoster Square, Londoni E.G.
( 4)
Iron and Steel Structures, Twelve-
trees, 7j. 6d.
Ironfounding, Homer, 4^.
Jack's Cooking, 2s.
Laundry Work, 2s.
Soups, Fish, and Vegetables, is.
each.
Jacobi's Printer's Handbook, y.
Vocabulary, y. 6d.
Jukes-Browne's Geology, 2J. 6d.
Kapp's Electric Transmission of
Energy, los. 6d.
Transformers, jf.
Kemp's Screw Propeller, 2J. 6d.
Kennedy's Electric Lamps, 2s. 6d.
Kingdon's Applied Magnetism, Js. 6d.
Lamps, Electric, Maycock, 6s.
Arc, Smithson, is.
Land Measuring Tables, 2s. 6d.
■ Surveying, Middleton, 4s. 6d.
■ Surveynig and Levelling,
Walmisley, 7j. 6d.
Landolt's Optical Activity and Chemi-
cal Composition, 4^. 6d
Lathes, Horner, 2ij. net.
Laundry Work, Jack, 2s.
Leather Work, Leiand's, 5j.
Leiand's Wood-carving, 5j.
Metal Work, JJ.
Leather Work, Jj.
Drawing and Designing, 2s.
Practical Education, 6s,
Lens Work for Amateurs, Orford, 3^.
Lenses, Photographic, Taylor, 3J. 6d.
Light, Sir H. T. Wood, 2s. 6d.
Lightning Conductors, Lodge, 15^.
Lloyd'."! Mine Manager, is. 6d.
Locomotives, Cooke, Js. 6d.
Locomotives Engineer, McDonnell, is.
Lodge's Lightning Conductors, i^s.
Lopp^ and Bouquet's Alternate
Currents, ICM. 6d.
Lukin's Turning Lathes, y.
Screws, y.
Lunge and Hurter's Alkali Makers'
Handbook, los. 6d.
McDonnell, How to become a Loco.
Engineer, l.r.
Maginnis' Atlantic Ferry, fs. 6d.
Magnetism, Kingdon, 7^. 6d. \6d.
and Electricity, Ashwo'rth, Zs.
Bottone, 2s. 6d.
Magnetism and Electricity, Houston,
y. 6d.
Maycock, 2s. 6d.
Manchester Ship Canal, y. 6d.
Manual Instruction and Training,
Manures, Griffiths, "js. 6d.
Marine Engineers, advice to, Roberts,
2s. 6d.
Drawing and Designing, 6s.
Electric Lighting, Walker, 5/.
Verbal Notes, Sothern, 3^. 6d.
net.
Marshall's Cakes is.
Cookery, is.
Martin's Structures, +?.
Massee's The Plant World, 2s. 6d.
Mathematical Tables, 12s.
Mathematics, Elementary, Halton,
2s. 6d.
May's Ballooning, 2s. 6d.
Maycock's Electricity and Magnetism,
2s. 6d.
Electric Lighting, 6j.
Alternating Current Circuit,
2s. 6d.
Electric Wiring, Fittings,
Switches, and Lamps, 6s.
■ Details Forms, 2s. 6d. net.
Electric Wiring Tables, 3^. 6d.
Mechanical Tables, Foden, is. 6d.
Mechanical Engineer's Pocket Book,
Medical Terms, Iloblyn, loj. 6d.
Merrill's Electric Lighting Speci-
fications, 6s. 6d.
Tables for Street Railway
Engineers, 5*.
Metal Turning, 4?.
Work, Leland, y.
Metric System, Wagstaff, is. 6d.
Measures, Born, y.
Middleton's Surveying, 4J. 6d.
Mill Work, Sutcliffe, 21s.
Millinery, Ortner, 2s. 6d.
Mine, Coal, Boyd, 7j. 6d.
Mineralogy, Hatch, 2J. 6d.
Model Engine Making, los. 6d
Modern Safety Bicycle, Garratt, y.
Moffett's Alternate Currents, 10s. 6d.
Motor Cars, Farman, 'y.
Motors, Electro, Bottone, 3^.
Houston, 5J-.
Hydraulic, Bodmer.
Municipal Electricitj' Supply, Gib-
bings, ly.
Nadifine, Drainage, is.
( 5 )
Naval Tactics, Fitzgerald, is.
Neumann's Electro Analysis, 6s.
NichoU's Agricultural Engineering in
India, y. 6d.
Noll's Wiring, 6s.
Optical Activity and Chemical Com-
position, Landolt, 4$. 6d.
Optical Instruments, Orford, 2s. 6d.
Optics of Photography, Taylor, y. 6d.
Orford's Lens Work, y.
Modern Optical Instruments,
2s. 6d.
Organic Chemical Manipulation,
Hewitt, 4J. 6d.
Ortner's Millinery, 2s. 6d.
Ozone, Andreoli, zs. 6d.
Parham and Shedd's Dynamo Test-
ing, Ss. 6d.
Patenting Inventions, 2j. 6d. net.
Pattern Making, Horner, y. 6d.
Periodic Classification and Chemical
Evolution, Rudorf, 4?. 6d.
Perry's Electric Railway Motors, Jj.
Petroleum, Boyd, zs.
Photographic Lenses, Taylor, y. 6d.
Physical Chemistry, Reychler, 4^. 6d.
Pickworth's Indicator, y. net.
Slide Rule, 2s.
Plant World, Massee, 2s. 6d.
Plante's Electric Storage, I2s.
Ponce de Leon's Spanish Techno-
logical Dictionary. Vol. I., 36J. net.
Vol. II., 32f. net.
Poole's Telephone Handbook, y.
Practical Education, Leiand, 6s.
Preece's Telephony, \y.
Price's Hoblyn's Dictionary of Medical
Terms, \os. 6d. \6d.
Primers of Electricity, Houston, ly.
Printer's Handbook, Jacobi, y.
Pruning, Des Cars, 2J. 6d.
Quadruplex, Mavor, 6s. 6d.
Quantities, Davis,' 3^^. 6d.
Questions in Typography, 6d.
Radiography, Bottone, y.
Railway Management, Findlay, Is. 6d.
■ Material Inspection, Bodmer.
Vocab., French-English, 7J-. 6d.
Electric, Hering, y. [net.
Repousse, Leiand, 5-^-
Reychler's Physical Chemistry, 4?. 6i.
Reynold's Locomotive, 2j, 6d.
Roberts' Drawing for Marine Engin-
eers, 6s.
Advice to Marine Eng., 2S. 6d.
Rontgen's X Rays, Bottone, 3^.
Rudorf's Periodic Classification and
Chemical Evolution, 4r. 6d.
Russell's Electric Cables, loj. 6d.
Safety Bicycle, Garratt, 3^.
Salomons' Electric Light Instal-
lations.
Vol. I., Accumulators, S^.
Vol. II., Apparatus, "Js. 6d.
Vol. III.. Applications, 5^.
Sanitary Drainage, is.
Savage's Sewage Disposal, y.
Screw Propeller, 2S. 6d.
Propulsion, Taylor, ly.
Screws and Screw Making, y.
Segundo's Dom. Elec. Lighting, is. ,
Serraillier's Railway Vocab., French
and English, "Js. 6d. net.
Sewage Treatment, Slater, 6s.
Shafting and Gearing, Bale, 2J. 6d.
Ships' Resistance, Taylor, 15^.
Slater's Sewage Treatment, 6s.
Slide Rule, Pickworth, is.
Sloyd, English, Barter, 7j. 6d.
Smith's Cable "Traction, S'f.
Smith's How to Patent, zs. 6d. net.
Wire, Is. 6d. \ls.
Smithson and Sharpe's Arc Lamps,
Sothern's Verbal Notes and Sketches,
y. 6d. net.
Southam's Elect. Engineering, y. 6d.
Spanish Technological Dictionary,
2 vols., 6%s. net.
Specialists' Series.
Steam Jacket, Fletcher, fs. 6d.
Locomotives, Cooke, 7^. 6d.
Model, Alexander, los. 6d.
Power and Mill Work, Sut-
cliffe, 2IJ.
Steamships, Atlantic, "js. 6d.
Steel Works Analysis, Arnold,
loj. 6d.
and Iron Structures, Twelve-
trees', 7j.
Steinmetz's Alternate Current Phe-
nomena, i%s,
Stevenson's Trees of Commerce,
y. 6d.
Still's Alternating Currents, S^.
Storage Battery, Tread well, 7j. 6d. net.
Street Railways, 5^.
Stresses and Strains, Farrow, y.
(6)
Structural Iron and Steel, Twelve-
trees', 7^.
Structures, Martin, 4s.
Stubbs' Telephony, 15^.
Stuttering, Klindworth, 61. net.
Sugar Machinery, 5j.
Surveying, Middleton, 4s. 6d.
Walmisley, "js. 6d.
Sutcliffe's Steam Power and Mill
Work, 2 1 J.
Switches, Electric, Maycock, 6s.
Tables, Electric Wiring, p. 6d.
Mechanical, Foden, is. 6d.
Taylor's Optic% of Photography,
3^. 6d.
Resistance of Ships, l$s.
Technological Dictionaries.
Telegraphist's Guide, Bell, is. 6d.
Telegraphy, Bell, is. 6d.
— — Herbert, 3^. 6d.
— ■ — ■ Houston, $s.
Wireless, Bottone, y.
Telephone System of the'British Post
Office, Herbert, 3^. 6d.
Handbook, Poole,